{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 14105, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 7.089684509039348e-05, "grad_norm": 62.79866149959377, "learning_rate": 2.358490566037736e-08, "loss": 2.5128, "step": 1 }, { "epoch": 0.00014179369018078695, "grad_norm": 58.46138963459668, "learning_rate": 4.716981132075472e-08, "loss": 2.4331, "step": 2 }, { "epoch": 0.00021269053527118043, "grad_norm": 65.67932032081204, "learning_rate": 7.075471698113208e-08, "loss": 2.401, "step": 3 }, { "epoch": 0.0002835873803615739, "grad_norm": 69.80669022852773, "learning_rate": 9.433962264150944e-08, "loss": 2.3053, "step": 4 }, { "epoch": 0.0003544842254519674, "grad_norm": 43.58515907966457, "learning_rate": 1.179245283018868e-07, "loss": 2.2627, "step": 5 }, { "epoch": 0.00042538107054236086, "grad_norm": 59.90080904156415, "learning_rate": 1.4150943396226417e-07, "loss": 2.3406, "step": 6 }, { "epoch": 0.0004962779156327543, "grad_norm": 53.16485425526669, "learning_rate": 1.6509433962264153e-07, "loss": 2.2701, "step": 7 }, { "epoch": 0.0005671747607231478, "grad_norm": 60.20825361406952, "learning_rate": 1.886792452830189e-07, "loss": 2.3269, "step": 8 }, { "epoch": 0.0006380716058135413, "grad_norm": 46.47172551621936, "learning_rate": 2.1226415094339622e-07, "loss": 2.3108, "step": 9 }, { "epoch": 0.0007089684509039348, "grad_norm": 49.267087513007326, "learning_rate": 2.358490566037736e-07, "loss": 2.3374, "step": 10 }, { "epoch": 0.0007798652959943283, "grad_norm": 40.3698989073045, "learning_rate": 2.59433962264151e-07, "loss": 2.3069, "step": 11 }, { "epoch": 0.0008507621410847217, "grad_norm": 48.02213643831459, "learning_rate": 2.8301886792452833e-07, "loss": 2.3454, "step": 12 }, { "epoch": 0.0009216589861751152, "grad_norm": 49.29134348826726, "learning_rate": 3.0660377358490567e-07, "loss": 2.2639, "step": 13 }, { "epoch": 0.0009925558312655087, "grad_norm": 50.07808672404782, "learning_rate": 3.3018867924528305e-07, "loss": 2.2252, "step": 14 }, { "epoch": 0.0010634526763559022, "grad_norm": 60.59238920281996, "learning_rate": 3.537735849056604e-07, "loss": 2.4635, "step": 15 }, { "epoch": 0.0011343495214462956, "grad_norm": 48.17131625230766, "learning_rate": 3.773584905660378e-07, "loss": 2.1261, "step": 16 }, { "epoch": 0.0012052463665366891, "grad_norm": 48.08472822986266, "learning_rate": 4.009433962264151e-07, "loss": 2.0969, "step": 17 }, { "epoch": 0.0012761432116270826, "grad_norm": 59.284618914014416, "learning_rate": 4.2452830188679244e-07, "loss": 2.3915, "step": 18 }, { "epoch": 0.001347040056717476, "grad_norm": 52.14857954243755, "learning_rate": 4.4811320754716983e-07, "loss": 2.2715, "step": 19 }, { "epoch": 0.0014179369018078695, "grad_norm": 49.95025563791469, "learning_rate": 4.716981132075472e-07, "loss": 2.1164, "step": 20 }, { "epoch": 0.001488833746898263, "grad_norm": 44.16737542068862, "learning_rate": 4.952830188679246e-07, "loss": 1.9546, "step": 21 }, { "epoch": 0.0015597305919886565, "grad_norm": 40.45043947856083, "learning_rate": 5.18867924528302e-07, "loss": 1.8822, "step": 22 }, { "epoch": 0.00163062743707905, "grad_norm": 43.05759971145575, "learning_rate": 5.424528301886793e-07, "loss": 1.9618, "step": 23 }, { "epoch": 0.0017015242821694435, "grad_norm": 57.66956745773418, "learning_rate": 5.660377358490567e-07, "loss": 2.2903, "step": 24 }, { "epoch": 0.001772421127259837, "grad_norm": 52.59040323804807, "learning_rate": 5.89622641509434e-07, "loss": 1.9838, "step": 25 }, { "epoch": 0.0018433179723502304, "grad_norm": 38.711170342323236, "learning_rate": 6.132075471698113e-07, "loss": 2.0053, "step": 26 }, { "epoch": 0.0019142148174406239, "grad_norm": 48.77000998092091, "learning_rate": 6.367924528301888e-07, "loss": 2.1672, "step": 27 }, { "epoch": 0.0019851116625310174, "grad_norm": 27.05361466488668, "learning_rate": 6.603773584905661e-07, "loss": 1.8275, "step": 28 }, { "epoch": 0.002056008507621411, "grad_norm": 25.389787381234985, "learning_rate": 6.839622641509434e-07, "loss": 1.8396, "step": 29 }, { "epoch": 0.0021269053527118043, "grad_norm": 18.502148725239877, "learning_rate": 7.075471698113208e-07, "loss": 1.7924, "step": 30 }, { "epoch": 0.002197802197802198, "grad_norm": 16.337707419861268, "learning_rate": 7.311320754716981e-07, "loss": 1.7674, "step": 31 }, { "epoch": 0.0022686990428925913, "grad_norm": 20.744666507011257, "learning_rate": 7.547169811320755e-07, "loss": 1.7675, "step": 32 }, { "epoch": 0.0023395958879829848, "grad_norm": 14.19804553407355, "learning_rate": 7.783018867924529e-07, "loss": 1.6546, "step": 33 }, { "epoch": 0.0024104927330733782, "grad_norm": 15.851883863725487, "learning_rate": 8.018867924528302e-07, "loss": 1.7501, "step": 34 }, { "epoch": 0.0024813895781637717, "grad_norm": 18.560751417573933, "learning_rate": 8.254716981132076e-07, "loss": 1.7541, "step": 35 }, { "epoch": 0.002552286423254165, "grad_norm": 20.05122313414433, "learning_rate": 8.490566037735849e-07, "loss": 1.6237, "step": 36 }, { "epoch": 0.0026231832683445587, "grad_norm": 28.56929746008384, "learning_rate": 8.726415094339623e-07, "loss": 1.7239, "step": 37 }, { "epoch": 0.002694080113434952, "grad_norm": 14.359850890658675, "learning_rate": 8.962264150943397e-07, "loss": 1.6827, "step": 38 }, { "epoch": 0.0027649769585253456, "grad_norm": 44.49432151362063, "learning_rate": 9.19811320754717e-07, "loss": 1.6523, "step": 39 }, { "epoch": 0.002835873803615739, "grad_norm": 20.920977848134836, "learning_rate": 9.433962264150944e-07, "loss": 1.4358, "step": 40 }, { "epoch": 0.0029067706487061326, "grad_norm": 29.044587436011057, "learning_rate": 9.669811320754719e-07, "loss": 1.6374, "step": 41 }, { "epoch": 0.002977667493796526, "grad_norm": 25.421473631943297, "learning_rate": 9.90566037735849e-07, "loss": 1.4121, "step": 42 }, { "epoch": 0.0030485643388869195, "grad_norm": 8.89982061599873, "learning_rate": 1.0141509433962265e-06, "loss": 1.4321, "step": 43 }, { "epoch": 0.003119461183977313, "grad_norm": 24.32030288606618, "learning_rate": 1.037735849056604e-06, "loss": 1.3747, "step": 44 }, { "epoch": 0.0031903580290677065, "grad_norm": 8.89569290325764, "learning_rate": 1.0613207547169812e-06, "loss": 1.5516, "step": 45 }, { "epoch": 0.0032612548741581, "grad_norm": 16.673118051763282, "learning_rate": 1.0849056603773587e-06, "loss": 1.5126, "step": 46 }, { "epoch": 0.0033321517192484934, "grad_norm": 8.927409453492615, "learning_rate": 1.1084905660377359e-06, "loss": 1.4704, "step": 47 }, { "epoch": 0.003403048564338887, "grad_norm": 24.303617120742498, "learning_rate": 1.1320754716981133e-06, "loss": 1.272, "step": 48 }, { "epoch": 0.0034739454094292804, "grad_norm": 7.3033016600367375, "learning_rate": 1.1556603773584908e-06, "loss": 1.2006, "step": 49 }, { "epoch": 0.003544842254519674, "grad_norm": 20.032616433563877, "learning_rate": 1.179245283018868e-06, "loss": 1.3505, "step": 50 }, { "epoch": 0.0036157390996100673, "grad_norm": 11.697592398210146, "learning_rate": 1.2028301886792454e-06, "loss": 1.3179, "step": 51 }, { "epoch": 0.003686635944700461, "grad_norm": 10.333000595380067, "learning_rate": 1.2264150943396227e-06, "loss": 1.2341, "step": 52 }, { "epoch": 0.0037575327897908543, "grad_norm": 17.51071218100834, "learning_rate": 1.25e-06, "loss": 1.3152, "step": 53 }, { "epoch": 0.0038284296348812478, "grad_norm": 10.560648884117816, "learning_rate": 1.2735849056603775e-06, "loss": 1.2827, "step": 54 }, { "epoch": 0.0038993264799716413, "grad_norm": 28.051922820832807, "learning_rate": 1.2971698113207548e-06, "loss": 1.4016, "step": 55 }, { "epoch": 0.003970223325062035, "grad_norm": 12.57210962862163, "learning_rate": 1.3207547169811322e-06, "loss": 1.223, "step": 56 }, { "epoch": 0.004041120170152428, "grad_norm": 24.858856962759575, "learning_rate": 1.3443396226415094e-06, "loss": 1.2785, "step": 57 }, { "epoch": 0.004112017015242822, "grad_norm": 13.704282836046433, "learning_rate": 1.3679245283018869e-06, "loss": 1.1879, "step": 58 }, { "epoch": 0.004182913860333215, "grad_norm": 20.687235811745175, "learning_rate": 1.3915094339622643e-06, "loss": 1.2157, "step": 59 }, { "epoch": 0.004253810705423609, "grad_norm": 27.697578235499673, "learning_rate": 1.4150943396226415e-06, "loss": 1.1213, "step": 60 }, { "epoch": 0.004324707550514002, "grad_norm": 8.475529291368074, "learning_rate": 1.438679245283019e-06, "loss": 1.1514, "step": 61 }, { "epoch": 0.004395604395604396, "grad_norm": 35.09840990193332, "learning_rate": 1.4622641509433962e-06, "loss": 1.2533, "step": 62 }, { "epoch": 0.004466501240694789, "grad_norm": 12.604632725989822, "learning_rate": 1.4858490566037737e-06, "loss": 1.3003, "step": 63 }, { "epoch": 0.0045373980857851826, "grad_norm": 12.119145758975556, "learning_rate": 1.509433962264151e-06, "loss": 1.1512, "step": 64 }, { "epoch": 0.004608294930875576, "grad_norm": 21.737884914947646, "learning_rate": 1.5330188679245283e-06, "loss": 1.1825, "step": 65 }, { "epoch": 0.0046791917759659695, "grad_norm": 28.534525783847, "learning_rate": 1.5566037735849058e-06, "loss": 1.2164, "step": 66 }, { "epoch": 0.004750088621056363, "grad_norm": 31.911339240827136, "learning_rate": 1.580188679245283e-06, "loss": 1.1891, "step": 67 }, { "epoch": 0.0048209854661467565, "grad_norm": 34.315853757716404, "learning_rate": 1.6037735849056604e-06, "loss": 1.2654, "step": 68 }, { "epoch": 0.00489188231123715, "grad_norm": 22.086344980591406, "learning_rate": 1.6273584905660379e-06, "loss": 1.2422, "step": 69 }, { "epoch": 0.004962779156327543, "grad_norm": 16.367184423793873, "learning_rate": 1.650943396226415e-06, "loss": 1.0625, "step": 70 }, { "epoch": 0.005033676001417937, "grad_norm": 21.3137651508565, "learning_rate": 1.6745283018867925e-06, "loss": 1.1933, "step": 71 }, { "epoch": 0.00510457284650833, "grad_norm": 13.381216540005344, "learning_rate": 1.6981132075471698e-06, "loss": 1.1367, "step": 72 }, { "epoch": 0.005175469691598724, "grad_norm": 19.115519053929876, "learning_rate": 1.7216981132075472e-06, "loss": 1.1523, "step": 73 }, { "epoch": 0.005246366536689117, "grad_norm": 20.856020823992466, "learning_rate": 1.7452830188679247e-06, "loss": 1.1005, "step": 74 }, { "epoch": 0.005317263381779511, "grad_norm": 17.732799383967993, "learning_rate": 1.7688679245283019e-06, "loss": 1.0958, "step": 75 }, { "epoch": 0.005388160226869904, "grad_norm": 12.411076694458, "learning_rate": 1.7924528301886793e-06, "loss": 1.0331, "step": 76 }, { "epoch": 0.005459057071960298, "grad_norm": 14.667540247880801, "learning_rate": 1.8160377358490566e-06, "loss": 1.101, "step": 77 }, { "epoch": 0.005529953917050691, "grad_norm": 11.722247599664582, "learning_rate": 1.839622641509434e-06, "loss": 1.1878, "step": 78 }, { "epoch": 0.005600850762141085, "grad_norm": 34.23157582409135, "learning_rate": 1.8632075471698114e-06, "loss": 1.1157, "step": 79 }, { "epoch": 0.005671747607231478, "grad_norm": 11.50514430244904, "learning_rate": 1.8867924528301889e-06, "loss": 1.1474, "step": 80 }, { "epoch": 0.005742644452321872, "grad_norm": 24.263372698724414, "learning_rate": 1.9103773584905665e-06, "loss": 1.1698, "step": 81 }, { "epoch": 0.005813541297412265, "grad_norm": 12.618846238702037, "learning_rate": 1.9339622641509438e-06, "loss": 1.0773, "step": 82 }, { "epoch": 0.005884438142502659, "grad_norm": 18.359394455844306, "learning_rate": 1.957547169811321e-06, "loss": 1.0655, "step": 83 }, { "epoch": 0.005955334987593052, "grad_norm": 17.61870856270326, "learning_rate": 1.981132075471698e-06, "loss": 1.0309, "step": 84 }, { "epoch": 0.006026231832683446, "grad_norm": 12.41821409979351, "learning_rate": 2.004716981132076e-06, "loss": 1.0788, "step": 85 }, { "epoch": 0.006097128677773839, "grad_norm": 19.169558381649278, "learning_rate": 2.028301886792453e-06, "loss": 1.0751, "step": 86 }, { "epoch": 0.0061680255228642325, "grad_norm": 2.382052955434158, "learning_rate": 2.0518867924528303e-06, "loss": 0.86, "step": 87 }, { "epoch": 0.006238922367954626, "grad_norm": 12.516435232622063, "learning_rate": 2.075471698113208e-06, "loss": 1.0971, "step": 88 }, { "epoch": 0.0063098192130450195, "grad_norm": 30.491406347963384, "learning_rate": 2.099056603773585e-06, "loss": 0.9636, "step": 89 }, { "epoch": 0.006380716058135413, "grad_norm": 12.401326889705404, "learning_rate": 2.1226415094339624e-06, "loss": 1.1357, "step": 90 }, { "epoch": 0.0064516129032258064, "grad_norm": 19.130494849242638, "learning_rate": 2.14622641509434e-06, "loss": 0.963, "step": 91 }, { "epoch": 0.0065225097483162, "grad_norm": 13.660158390077306, "learning_rate": 2.1698113207547173e-06, "loss": 0.9698, "step": 92 }, { "epoch": 0.006593406593406593, "grad_norm": 13.438776554478084, "learning_rate": 2.1933962264150945e-06, "loss": 0.9903, "step": 93 }, { "epoch": 0.006664303438496987, "grad_norm": 27.775934246264583, "learning_rate": 2.2169811320754718e-06, "loss": 1.0634, "step": 94 }, { "epoch": 0.00673520028358738, "grad_norm": 10.927308145218516, "learning_rate": 2.2405660377358494e-06, "loss": 1.0205, "step": 95 }, { "epoch": 0.006806097128677774, "grad_norm": 13.91790042056528, "learning_rate": 2.2641509433962266e-06, "loss": 1.0211, "step": 96 }, { "epoch": 0.006876993973768167, "grad_norm": 23.169899890917208, "learning_rate": 2.287735849056604e-06, "loss": 1.1108, "step": 97 }, { "epoch": 0.006947890818858561, "grad_norm": 10.820852248578747, "learning_rate": 2.3113207547169815e-06, "loss": 1.1429, "step": 98 }, { "epoch": 0.007018787663948954, "grad_norm": 27.03348699768736, "learning_rate": 2.3349056603773588e-06, "loss": 0.9924, "step": 99 }, { "epoch": 0.007089684509039348, "grad_norm": 19.0035254721748, "learning_rate": 2.358490566037736e-06, "loss": 1.048, "step": 100 }, { "epoch": 0.007160581354129741, "grad_norm": 20.52541655594829, "learning_rate": 2.3820754716981136e-06, "loss": 1.1378, "step": 101 }, { "epoch": 0.007231478199220135, "grad_norm": 15.944564427425536, "learning_rate": 2.405660377358491e-06, "loss": 0.9855, "step": 102 }, { "epoch": 0.007302375044310528, "grad_norm": 9.586982661943322, "learning_rate": 2.429245283018868e-06, "loss": 0.9413, "step": 103 }, { "epoch": 0.007373271889400922, "grad_norm": 10.468648191663519, "learning_rate": 2.4528301886792453e-06, "loss": 0.8885, "step": 104 }, { "epoch": 0.007444168734491315, "grad_norm": 8.895020812186798, "learning_rate": 2.476415094339623e-06, "loss": 0.9454, "step": 105 }, { "epoch": 0.007515065579581709, "grad_norm": 9.06924143653337, "learning_rate": 2.5e-06, "loss": 0.9785, "step": 106 }, { "epoch": 0.007585962424672102, "grad_norm": 9.70522411962086, "learning_rate": 2.523584905660378e-06, "loss": 0.9965, "step": 107 }, { "epoch": 0.0076568592697624956, "grad_norm": 9.360056847367867, "learning_rate": 2.547169811320755e-06, "loss": 0.918, "step": 108 }, { "epoch": 0.007727756114852889, "grad_norm": 11.176837515251304, "learning_rate": 2.5707547169811327e-06, "loss": 1.0032, "step": 109 }, { "epoch": 0.0077986529599432825, "grad_norm": 15.374441313798798, "learning_rate": 2.5943396226415095e-06, "loss": 0.9775, "step": 110 }, { "epoch": 0.007869549805033677, "grad_norm": 10.62724284834002, "learning_rate": 2.617924528301887e-06, "loss": 1.0307, "step": 111 }, { "epoch": 0.00794044665012407, "grad_norm": 16.871741991935526, "learning_rate": 2.6415094339622644e-06, "loss": 0.989, "step": 112 }, { "epoch": 0.008011343495214464, "grad_norm": 17.056364464845966, "learning_rate": 2.665094339622642e-06, "loss": 1.0878, "step": 113 }, { "epoch": 0.008082240340304856, "grad_norm": 10.523978104213334, "learning_rate": 2.688679245283019e-06, "loss": 1.0396, "step": 114 }, { "epoch": 0.00815313718539525, "grad_norm": 14.349237135967359, "learning_rate": 2.7122641509433965e-06, "loss": 0.9614, "step": 115 }, { "epoch": 0.008224034030485643, "grad_norm": 14.48191109238339, "learning_rate": 2.7358490566037738e-06, "loss": 0.9567, "step": 116 }, { "epoch": 0.008294930875576038, "grad_norm": 10.479991456592613, "learning_rate": 2.7594339622641514e-06, "loss": 1.002, "step": 117 }, { "epoch": 0.00836582772066643, "grad_norm": 18.07516329513671, "learning_rate": 2.7830188679245286e-06, "loss": 0.8989, "step": 118 }, { "epoch": 0.008436724565756825, "grad_norm": 10.444927730616367, "learning_rate": 2.8066037735849063e-06, "loss": 0.8998, "step": 119 }, { "epoch": 0.008507621410847217, "grad_norm": 12.777624855372126, "learning_rate": 2.830188679245283e-06, "loss": 0.8698, "step": 120 }, { "epoch": 0.008578518255937612, "grad_norm": 12.697688593951932, "learning_rate": 2.8537735849056608e-06, "loss": 0.9519, "step": 121 }, { "epoch": 0.008649415101028004, "grad_norm": 11.824499695080656, "learning_rate": 2.877358490566038e-06, "loss": 0.8992, "step": 122 }, { "epoch": 0.008720311946118399, "grad_norm": 1.9776425354808076, "learning_rate": 2.9009433962264156e-06, "loss": 0.8117, "step": 123 }, { "epoch": 0.008791208791208791, "grad_norm": 22.52391212593123, "learning_rate": 2.9245283018867924e-06, "loss": 0.9096, "step": 124 }, { "epoch": 0.008862105636299186, "grad_norm": 10.589012313882543, "learning_rate": 2.94811320754717e-06, "loss": 0.9213, "step": 125 }, { "epoch": 0.008933002481389578, "grad_norm": 17.05642997954554, "learning_rate": 2.9716981132075473e-06, "loss": 0.9243, "step": 126 }, { "epoch": 0.009003899326479973, "grad_norm": 11.338659963331889, "learning_rate": 2.995283018867925e-06, "loss": 0.8937, "step": 127 }, { "epoch": 0.009074796171570365, "grad_norm": 9.508349642080637, "learning_rate": 3.018867924528302e-06, "loss": 0.9597, "step": 128 }, { "epoch": 0.00914569301666076, "grad_norm": 10.476059074204278, "learning_rate": 3.04245283018868e-06, "loss": 0.8771, "step": 129 }, { "epoch": 0.009216589861751152, "grad_norm": 17.606396773428465, "learning_rate": 3.0660377358490567e-06, "loss": 0.8245, "step": 130 }, { "epoch": 0.009287486706841546, "grad_norm": 9.193570876320983, "learning_rate": 3.0896226415094343e-06, "loss": 0.838, "step": 131 }, { "epoch": 0.009358383551931939, "grad_norm": 11.66286683994402, "learning_rate": 3.1132075471698115e-06, "loss": 0.9285, "step": 132 }, { "epoch": 0.009429280397022333, "grad_norm": 11.288160526427609, "learning_rate": 3.136792452830189e-06, "loss": 0.7918, "step": 133 }, { "epoch": 0.009500177242112726, "grad_norm": 15.957722012404227, "learning_rate": 3.160377358490566e-06, "loss": 0.8829, "step": 134 }, { "epoch": 0.00957107408720312, "grad_norm": 8.574418984726286, "learning_rate": 3.1839622641509436e-06, "loss": 0.7982, "step": 135 }, { "epoch": 0.009641970932293513, "grad_norm": 11.96855308614916, "learning_rate": 3.207547169811321e-06, "loss": 0.8493, "step": 136 }, { "epoch": 0.009712867777383907, "grad_norm": 8.898283999423308, "learning_rate": 3.2311320754716985e-06, "loss": 0.7366, "step": 137 }, { "epoch": 0.0097837646224743, "grad_norm": 10.446711822352848, "learning_rate": 3.2547169811320758e-06, "loss": 0.7427, "step": 138 }, { "epoch": 0.009854661467564694, "grad_norm": 12.35814515328214, "learning_rate": 3.2783018867924534e-06, "loss": 0.8145, "step": 139 }, { "epoch": 0.009925558312655087, "grad_norm": 13.134880083390238, "learning_rate": 3.30188679245283e-06, "loss": 0.7674, "step": 140 }, { "epoch": 0.009996455157745481, "grad_norm": 8.350764001235486, "learning_rate": 3.325471698113208e-06, "loss": 0.8476, "step": 141 }, { "epoch": 0.010067352002835874, "grad_norm": 8.802547543430865, "learning_rate": 3.349056603773585e-06, "loss": 0.7758, "step": 142 }, { "epoch": 0.010138248847926268, "grad_norm": 9.096273325612495, "learning_rate": 3.3726415094339627e-06, "loss": 0.7485, "step": 143 }, { "epoch": 0.01020914569301666, "grad_norm": 8.135370149297598, "learning_rate": 3.3962264150943395e-06, "loss": 0.8576, "step": 144 }, { "epoch": 0.010280042538107055, "grad_norm": 7.532942705444922, "learning_rate": 3.419811320754717e-06, "loss": 0.7588, "step": 145 }, { "epoch": 0.010350939383197448, "grad_norm": 11.859787490386541, "learning_rate": 3.4433962264150944e-06, "loss": 0.7791, "step": 146 }, { "epoch": 0.010421836228287842, "grad_norm": 6.632383990929329, "learning_rate": 3.466981132075472e-06, "loss": 0.7399, "step": 147 }, { "epoch": 0.010492733073378235, "grad_norm": 9.862786838277295, "learning_rate": 3.4905660377358493e-06, "loss": 0.7595, "step": 148 }, { "epoch": 0.010563629918468629, "grad_norm": 9.089919666598725, "learning_rate": 3.514150943396227e-06, "loss": 0.7672, "step": 149 }, { "epoch": 0.010634526763559022, "grad_norm": 6.460190361158846, "learning_rate": 3.5377358490566038e-06, "loss": 0.7473, "step": 150 }, { "epoch": 0.010705423608649416, "grad_norm": 5.78972219392316, "learning_rate": 3.5613207547169814e-06, "loss": 0.7791, "step": 151 }, { "epoch": 0.010776320453739809, "grad_norm": 7.810378601536955, "learning_rate": 3.5849056603773586e-06, "loss": 0.7039, "step": 152 }, { "epoch": 0.010847217298830203, "grad_norm": 5.91827592366381, "learning_rate": 3.6084905660377363e-06, "loss": 0.6614, "step": 153 }, { "epoch": 0.010918114143920596, "grad_norm": 6.15005508880087, "learning_rate": 3.632075471698113e-06, "loss": 0.7308, "step": 154 }, { "epoch": 0.01098901098901099, "grad_norm": 8.901510147292857, "learning_rate": 3.6556603773584908e-06, "loss": 0.7582, "step": 155 }, { "epoch": 0.011059907834101382, "grad_norm": 4.876877445970411, "learning_rate": 3.679245283018868e-06, "loss": 0.6399, "step": 156 }, { "epoch": 0.011130804679191777, "grad_norm": 6.421263053009188, "learning_rate": 3.7028301886792456e-06, "loss": 0.7528, "step": 157 }, { "epoch": 0.01120170152428217, "grad_norm": 1.8153741603210045, "learning_rate": 3.726415094339623e-06, "loss": 0.8543, "step": 158 }, { "epoch": 0.011272598369372564, "grad_norm": 6.381507673009308, "learning_rate": 3.7500000000000005e-06, "loss": 0.7287, "step": 159 }, { "epoch": 0.011343495214462956, "grad_norm": 4.203507499642998, "learning_rate": 3.7735849056603777e-06, "loss": 0.738, "step": 160 }, { "epoch": 0.01141439205955335, "grad_norm": 6.259571430318609, "learning_rate": 3.797169811320755e-06, "loss": 0.681, "step": 161 }, { "epoch": 0.011485288904643743, "grad_norm": 6.114555521229215, "learning_rate": 3.820754716981133e-06, "loss": 0.7071, "step": 162 }, { "epoch": 0.011556185749734138, "grad_norm": 3.874307365616256, "learning_rate": 3.8443396226415094e-06, "loss": 0.7004, "step": 163 }, { "epoch": 0.01162708259482453, "grad_norm": 7.688161517053674, "learning_rate": 3.8679245283018875e-06, "loss": 0.727, "step": 164 }, { "epoch": 0.011697979439914925, "grad_norm": 4.05944220745164, "learning_rate": 3.891509433962265e-06, "loss": 0.7635, "step": 165 }, { "epoch": 0.011768876285005317, "grad_norm": 4.695782594913331, "learning_rate": 3.915094339622642e-06, "loss": 0.7175, "step": 166 }, { "epoch": 0.011839773130095712, "grad_norm": 7.356338458868027, "learning_rate": 3.938679245283019e-06, "loss": 0.7218, "step": 167 }, { "epoch": 0.011910669975186104, "grad_norm": 6.0890948991002904, "learning_rate": 3.962264150943396e-06, "loss": 0.6463, "step": 168 }, { "epoch": 0.011981566820276499, "grad_norm": 5.704071381245182, "learning_rate": 3.985849056603774e-06, "loss": 0.6864, "step": 169 }, { "epoch": 0.012052463665366891, "grad_norm": 4.775155767030014, "learning_rate": 4.009433962264152e-06, "loss": 0.5972, "step": 170 }, { "epoch": 0.012123360510457286, "grad_norm": 7.518476763691964, "learning_rate": 4.033018867924529e-06, "loss": 0.7012, "step": 171 }, { "epoch": 0.012194257355547678, "grad_norm": 4.594560807282772, "learning_rate": 4.056603773584906e-06, "loss": 0.7164, "step": 172 }, { "epoch": 0.012265154200638072, "grad_norm": 5.9369788061539355, "learning_rate": 4.080188679245283e-06, "loss": 0.6591, "step": 173 }, { "epoch": 0.012336051045728465, "grad_norm": 5.410873110782695, "learning_rate": 4.103773584905661e-06, "loss": 0.6222, "step": 174 }, { "epoch": 0.01240694789081886, "grad_norm": 4.359391110890997, "learning_rate": 4.127358490566038e-06, "loss": 0.6406, "step": 175 }, { "epoch": 0.012477844735909252, "grad_norm": 7.948406580118142, "learning_rate": 4.150943396226416e-06, "loss": 0.6595, "step": 176 }, { "epoch": 0.012548741580999646, "grad_norm": 4.276409391449875, "learning_rate": 4.174528301886792e-06, "loss": 0.5842, "step": 177 }, { "epoch": 0.012619638426090039, "grad_norm": 4.264187638814518, "learning_rate": 4.19811320754717e-06, "loss": 0.6241, "step": 178 }, { "epoch": 0.012690535271180433, "grad_norm": 6.262493034371604, "learning_rate": 4.221698113207548e-06, "loss": 0.756, "step": 179 }, { "epoch": 0.012761432116270826, "grad_norm": 4.814167594563414, "learning_rate": 4.245283018867925e-06, "loss": 0.7269, "step": 180 }, { "epoch": 0.01283232896136122, "grad_norm": 6.727230218718842, "learning_rate": 4.268867924528302e-06, "loss": 0.6345, "step": 181 }, { "epoch": 0.012903225806451613, "grad_norm": 5.414568820101855, "learning_rate": 4.29245283018868e-06, "loss": 0.6112, "step": 182 }, { "epoch": 0.012974122651542007, "grad_norm": 6.302033609194174, "learning_rate": 4.3160377358490565e-06, "loss": 0.6578, "step": 183 }, { "epoch": 0.0130450194966324, "grad_norm": 4.030118768165745, "learning_rate": 4.339622641509435e-06, "loss": 0.6343, "step": 184 }, { "epoch": 0.013115916341722794, "grad_norm": 5.221588355214441, "learning_rate": 4.363207547169812e-06, "loss": 0.5909, "step": 185 }, { "epoch": 0.013186813186813187, "grad_norm": 4.780463478667065, "learning_rate": 4.386792452830189e-06, "loss": 0.6362, "step": 186 }, { "epoch": 0.013257710031903581, "grad_norm": 6.771442122697124, "learning_rate": 4.410377358490566e-06, "loss": 0.641, "step": 187 }, { "epoch": 0.013328606876993974, "grad_norm": 4.593314351815171, "learning_rate": 4.4339622641509435e-06, "loss": 0.638, "step": 188 }, { "epoch": 0.013399503722084368, "grad_norm": 7.23462593819815, "learning_rate": 4.457547169811321e-06, "loss": 0.672, "step": 189 }, { "epoch": 0.01347040056717476, "grad_norm": 11.069764852340805, "learning_rate": 4.481132075471699e-06, "loss": 0.6022, "step": 190 }, { "epoch": 0.013541297412265155, "grad_norm": 7.748840678447438, "learning_rate": 4.504716981132076e-06, "loss": 0.6878, "step": 191 }, { "epoch": 0.013612194257355548, "grad_norm": 5.178649093698924, "learning_rate": 4.528301886792453e-06, "loss": 0.5622, "step": 192 }, { "epoch": 0.013683091102445942, "grad_norm": 7.9902909361557874, "learning_rate": 4.5518867924528305e-06, "loss": 0.5494, "step": 193 }, { "epoch": 0.013753987947536335, "grad_norm": 12.603669731939856, "learning_rate": 4.575471698113208e-06, "loss": 0.5707, "step": 194 }, { "epoch": 0.013824884792626729, "grad_norm": 12.60281756508717, "learning_rate": 4.599056603773585e-06, "loss": 0.6229, "step": 195 }, { "epoch": 0.013895781637717122, "grad_norm": 4.086883687037787, "learning_rate": 4.622641509433963e-06, "loss": 0.5654, "step": 196 }, { "epoch": 0.013966678482807516, "grad_norm": 5.382973479348249, "learning_rate": 4.6462264150943394e-06, "loss": 0.5747, "step": 197 }, { "epoch": 0.014037575327897909, "grad_norm": 5.8451886655777, "learning_rate": 4.6698113207547175e-06, "loss": 0.6386, "step": 198 }, { "epoch": 0.014108472172988303, "grad_norm": 6.4984796747224625, "learning_rate": 4.693396226415095e-06, "loss": 0.5451, "step": 199 }, { "epoch": 0.014179369018078695, "grad_norm": 5.906980062257258, "learning_rate": 4.716981132075472e-06, "loss": 0.6055, "step": 200 }, { "epoch": 0.01425026586316909, "grad_norm": 5.327067730461185, "learning_rate": 4.740566037735849e-06, "loss": 0.572, "step": 201 }, { "epoch": 0.014321162708259482, "grad_norm": 6.851276913312794, "learning_rate": 4.764150943396227e-06, "loss": 0.5726, "step": 202 }, { "epoch": 0.014392059553349877, "grad_norm": 6.052996083132558, "learning_rate": 4.787735849056604e-06, "loss": 0.5386, "step": 203 }, { "epoch": 0.01446295639844027, "grad_norm": 5.7462807050036595, "learning_rate": 4.811320754716982e-06, "loss": 0.5739, "step": 204 }, { "epoch": 0.014533853243530664, "grad_norm": 6.307644339396345, "learning_rate": 4.834905660377359e-06, "loss": 0.6408, "step": 205 }, { "epoch": 0.014604750088621056, "grad_norm": 5.848513638268455, "learning_rate": 4.858490566037736e-06, "loss": 0.5204, "step": 206 }, { "epoch": 0.01467564693371145, "grad_norm": 7.005319711784124, "learning_rate": 4.882075471698113e-06, "loss": 0.5548, "step": 207 }, { "epoch": 0.014746543778801843, "grad_norm": 3.5567530783104613, "learning_rate": 4.905660377358491e-06, "loss": 0.5622, "step": 208 }, { "epoch": 0.014817440623892238, "grad_norm": 4.566123321823704, "learning_rate": 4.929245283018868e-06, "loss": 0.5709, "step": 209 }, { "epoch": 0.01488833746898263, "grad_norm": 7.555419260021845, "learning_rate": 4.952830188679246e-06, "loss": 0.5848, "step": 210 }, { "epoch": 0.014959234314073025, "grad_norm": 5.677719138042076, "learning_rate": 4.976415094339623e-06, "loss": 0.5774, "step": 211 }, { "epoch": 0.015030131159163417, "grad_norm": 5.2544499447746595, "learning_rate": 5e-06, "loss": 0.5807, "step": 212 }, { "epoch": 0.015101028004253812, "grad_norm": 5.300796240282239, "learning_rate": 5.023584905660378e-06, "loss": 0.5487, "step": 213 }, { "epoch": 0.015171924849344204, "grad_norm": 4.086371395446162, "learning_rate": 5.047169811320756e-06, "loss": 0.5837, "step": 214 }, { "epoch": 0.015242821694434599, "grad_norm": 5.60150890663973, "learning_rate": 5.070754716981132e-06, "loss": 0.5692, "step": 215 }, { "epoch": 0.015313718539524991, "grad_norm": 1.547612396211903, "learning_rate": 5.09433962264151e-06, "loss": 0.8184, "step": 216 }, { "epoch": 0.015384615384615385, "grad_norm": 4.376873156951854, "learning_rate": 5.117924528301887e-06, "loss": 0.5451, "step": 217 }, { "epoch": 0.015455512229705778, "grad_norm": 4.242091765895576, "learning_rate": 5.1415094339622655e-06, "loss": 0.5511, "step": 218 }, { "epoch": 0.015526409074796172, "grad_norm": 7.304623267018853, "learning_rate": 5.165094339622642e-06, "loss": 0.5652, "step": 219 }, { "epoch": 0.015597305919886565, "grad_norm": 4.898872364846602, "learning_rate": 5.188679245283019e-06, "loss": 0.5072, "step": 220 }, { "epoch": 0.01566820276497696, "grad_norm": 7.458638100366504, "learning_rate": 5.212264150943397e-06, "loss": 0.5697, "step": 221 }, { "epoch": 0.015739099610067354, "grad_norm": 5.183378477846666, "learning_rate": 5.235849056603774e-06, "loss": 0.5813, "step": 222 }, { "epoch": 0.015809996455157745, "grad_norm": 6.7197914370969265, "learning_rate": 5.259433962264151e-06, "loss": 0.5905, "step": 223 }, { "epoch": 0.01588089330024814, "grad_norm": 6.405184652883935, "learning_rate": 5.283018867924529e-06, "loss": 0.5018, "step": 224 }, { "epoch": 0.015951790145338533, "grad_norm": 4.398213434055678, "learning_rate": 5.306603773584906e-06, "loss": 0.5341, "step": 225 }, { "epoch": 0.016022686990428928, "grad_norm": 17.69292392142621, "learning_rate": 5.330188679245284e-06, "loss": 0.5709, "step": 226 }, { "epoch": 0.01609358383551932, "grad_norm": 6.403155060851229, "learning_rate": 5.3537735849056605e-06, "loss": 0.5478, "step": 227 }, { "epoch": 0.016164480680609713, "grad_norm": 12.507676115252599, "learning_rate": 5.377358490566038e-06, "loss": 0.5079, "step": 228 }, { "epoch": 0.016235377525700107, "grad_norm": 4.748437877766233, "learning_rate": 5.400943396226416e-06, "loss": 0.5867, "step": 229 }, { "epoch": 0.0163062743707905, "grad_norm": 3.571632414442551, "learning_rate": 5.424528301886793e-06, "loss": 0.5304, "step": 230 }, { "epoch": 0.016377171215880892, "grad_norm": 1.3605983943826498, "learning_rate": 5.44811320754717e-06, "loss": 0.6886, "step": 231 }, { "epoch": 0.016448068060971287, "grad_norm": 8.831552751005612, "learning_rate": 5.4716981132075475e-06, "loss": 0.6171, "step": 232 }, { "epoch": 0.01651896490606168, "grad_norm": 4.732842213497534, "learning_rate": 5.495283018867925e-06, "loss": 0.5326, "step": 233 }, { "epoch": 0.016589861751152075, "grad_norm": 5.3952283514884884, "learning_rate": 5.518867924528303e-06, "loss": 0.5461, "step": 234 }, { "epoch": 0.016660758596242466, "grad_norm": 7.277771596774963, "learning_rate": 5.542452830188679e-06, "loss": 0.6182, "step": 235 }, { "epoch": 0.01673165544133286, "grad_norm": 5.9701451374295775, "learning_rate": 5.566037735849057e-06, "loss": 0.6127, "step": 236 }, { "epoch": 0.016802552286423255, "grad_norm": 7.362913131388615, "learning_rate": 5.5896226415094345e-06, "loss": 0.481, "step": 237 }, { "epoch": 0.01687344913151365, "grad_norm": 4.735116392503911, "learning_rate": 5.613207547169813e-06, "loss": 0.4844, "step": 238 }, { "epoch": 0.01694434597660404, "grad_norm": 9.401124448482948, "learning_rate": 5.636792452830189e-06, "loss": 0.5296, "step": 239 }, { "epoch": 0.017015242821694435, "grad_norm": 6.987520761006301, "learning_rate": 5.660377358490566e-06, "loss": 0.5831, "step": 240 }, { "epoch": 0.01708613966678483, "grad_norm": 3.826179603196633, "learning_rate": 5.683962264150944e-06, "loss": 0.5918, "step": 241 }, { "epoch": 0.017157036511875223, "grad_norm": 7.6510693773934095, "learning_rate": 5.7075471698113215e-06, "loss": 0.5456, "step": 242 }, { "epoch": 0.017227933356965614, "grad_norm": 26.810898141954706, "learning_rate": 5.731132075471698e-06, "loss": 0.5603, "step": 243 }, { "epoch": 0.01729883020205601, "grad_norm": 5.007268321556926, "learning_rate": 5.754716981132076e-06, "loss": 0.5659, "step": 244 }, { "epoch": 0.017369727047146403, "grad_norm": 3.541672061710706, "learning_rate": 5.778301886792453e-06, "loss": 0.4792, "step": 245 }, { "epoch": 0.017440623892236797, "grad_norm": 5.021517658200181, "learning_rate": 5.801886792452831e-06, "loss": 0.5648, "step": 246 }, { "epoch": 0.017511520737327188, "grad_norm": 5.875400797159038, "learning_rate": 5.825471698113208e-06, "loss": 0.4785, "step": 247 }, { "epoch": 0.017582417582417582, "grad_norm": 4.130446875254679, "learning_rate": 5.849056603773585e-06, "loss": 0.6105, "step": 248 }, { "epoch": 0.017653314427507977, "grad_norm": 5.607107074696043, "learning_rate": 5.872641509433963e-06, "loss": 0.5205, "step": 249 }, { "epoch": 0.01772421127259837, "grad_norm": 4.459050131707145, "learning_rate": 5.89622641509434e-06, "loss": 0.527, "step": 250 }, { "epoch": 0.017795108117688762, "grad_norm": 5.2284404243970695, "learning_rate": 5.919811320754717e-06, "loss": 0.5611, "step": 251 }, { "epoch": 0.017866004962779156, "grad_norm": 8.644484193186283, "learning_rate": 5.943396226415095e-06, "loss": 0.4904, "step": 252 }, { "epoch": 0.01793690180786955, "grad_norm": 5.815161904827929, "learning_rate": 5.966981132075472e-06, "loss": 0.5548, "step": 253 }, { "epoch": 0.018007798652959945, "grad_norm": 5.015737832257517, "learning_rate": 5.99056603773585e-06, "loss": 0.5112, "step": 254 }, { "epoch": 0.018078695498050336, "grad_norm": 6.21388745979328, "learning_rate": 6.014150943396226e-06, "loss": 0.5281, "step": 255 }, { "epoch": 0.01814959234314073, "grad_norm": 4.106113449672835, "learning_rate": 6.037735849056604e-06, "loss": 0.4922, "step": 256 }, { "epoch": 0.018220489188231125, "grad_norm": 4.15661316818096, "learning_rate": 6.061320754716982e-06, "loss": 0.5276, "step": 257 }, { "epoch": 0.01829138603332152, "grad_norm": 4.492405546283805, "learning_rate": 6.08490566037736e-06, "loss": 0.4871, "step": 258 }, { "epoch": 0.01836228287841191, "grad_norm": 7.469183240517566, "learning_rate": 6.108490566037736e-06, "loss": 0.4995, "step": 259 }, { "epoch": 0.018433179723502304, "grad_norm": 3.8330562033187014, "learning_rate": 6.132075471698113e-06, "loss": 0.4998, "step": 260 }, { "epoch": 0.0185040765685927, "grad_norm": 5.936524697691081, "learning_rate": 6.155660377358491e-06, "loss": 0.4876, "step": 261 }, { "epoch": 0.018574973413683093, "grad_norm": 4.396621265689584, "learning_rate": 6.179245283018869e-06, "loss": 0.5441, "step": 262 }, { "epoch": 0.018645870258773484, "grad_norm": 4.497405426215639, "learning_rate": 6.202830188679245e-06, "loss": 0.5816, "step": 263 }, { "epoch": 0.018716767103863878, "grad_norm": 4.5927464439642245, "learning_rate": 6.226415094339623e-06, "loss": 0.5351, "step": 264 }, { "epoch": 0.018787663948954272, "grad_norm": 5.087522982060394, "learning_rate": 6.25e-06, "loss": 0.5371, "step": 265 }, { "epoch": 0.018858560794044667, "grad_norm": 4.702967354717237, "learning_rate": 6.273584905660378e-06, "loss": 0.4507, "step": 266 }, { "epoch": 0.018929457639135058, "grad_norm": 4.943646165085859, "learning_rate": 6.297169811320756e-06, "loss": 0.5618, "step": 267 }, { "epoch": 0.019000354484225452, "grad_norm": 3.718738371232409, "learning_rate": 6.320754716981132e-06, "loss": 0.4889, "step": 268 }, { "epoch": 0.019071251329315846, "grad_norm": 13.952193853381406, "learning_rate": 6.34433962264151e-06, "loss": 0.5164, "step": 269 }, { "epoch": 0.01914214817440624, "grad_norm": 4.066833777104133, "learning_rate": 6.367924528301887e-06, "loss": 0.5751, "step": 270 }, { "epoch": 0.01921304501949663, "grad_norm": 3.429703549787762, "learning_rate": 6.391509433962265e-06, "loss": 0.5192, "step": 271 }, { "epoch": 0.019283941864587026, "grad_norm": 3.6774495862474716, "learning_rate": 6.415094339622642e-06, "loss": 0.5636, "step": 272 }, { "epoch": 0.01935483870967742, "grad_norm": 2.9517221268582006, "learning_rate": 6.438679245283019e-06, "loss": 0.5276, "step": 273 }, { "epoch": 0.019425735554767815, "grad_norm": 18.069868194123444, "learning_rate": 6.462264150943397e-06, "loss": 0.5297, "step": 274 }, { "epoch": 0.019496632399858205, "grad_norm": 3.2404588509854615, "learning_rate": 6.485849056603774e-06, "loss": 0.5281, "step": 275 }, { "epoch": 0.0195675292449486, "grad_norm": 5.3896981478791455, "learning_rate": 6.5094339622641515e-06, "loss": 0.5171, "step": 276 }, { "epoch": 0.019638426090038994, "grad_norm": 3.808873792656264, "learning_rate": 6.533018867924529e-06, "loss": 0.5099, "step": 277 }, { "epoch": 0.01970932293512939, "grad_norm": 4.448159427563967, "learning_rate": 6.556603773584907e-06, "loss": 0.5404, "step": 278 }, { "epoch": 0.01978021978021978, "grad_norm": 5.407890155687382, "learning_rate": 6.580188679245284e-06, "loss": 0.5474, "step": 279 }, { "epoch": 0.019851116625310174, "grad_norm": 4.5727278777712606, "learning_rate": 6.60377358490566e-06, "loss": 0.5422, "step": 280 }, { "epoch": 0.019922013470400568, "grad_norm": 1.4799286138517767, "learning_rate": 6.6273584905660385e-06, "loss": 0.6767, "step": 281 }, { "epoch": 0.019992910315490962, "grad_norm": 6.300257099487581, "learning_rate": 6.650943396226416e-06, "loss": 0.6049, "step": 282 }, { "epoch": 0.020063807160581353, "grad_norm": 3.46711392275951, "learning_rate": 6.674528301886794e-06, "loss": 0.5182, "step": 283 }, { "epoch": 0.020134704005671748, "grad_norm": 9.205585315225076, "learning_rate": 6.69811320754717e-06, "loss": 0.4737, "step": 284 }, { "epoch": 0.020205600850762142, "grad_norm": 5.9453437373275495, "learning_rate": 6.721698113207547e-06, "loss": 0.5752, "step": 285 }, { "epoch": 0.020276497695852536, "grad_norm": 4.889224166372858, "learning_rate": 6.7452830188679255e-06, "loss": 0.5456, "step": 286 }, { "epoch": 0.020347394540942927, "grad_norm": 6.972625617077045, "learning_rate": 6.768867924528303e-06, "loss": 0.5207, "step": 287 }, { "epoch": 0.02041829138603332, "grad_norm": 3.4908109664036444, "learning_rate": 6.792452830188679e-06, "loss": 0.5146, "step": 288 }, { "epoch": 0.020489188231123716, "grad_norm": 1.3293154432649275, "learning_rate": 6.816037735849057e-06, "loss": 0.7953, "step": 289 }, { "epoch": 0.02056008507621411, "grad_norm": 6.094796056797746, "learning_rate": 6.839622641509434e-06, "loss": 0.5479, "step": 290 }, { "epoch": 0.0206309819213045, "grad_norm": 6.426521407541103, "learning_rate": 6.8632075471698125e-06, "loss": 0.5384, "step": 291 }, { "epoch": 0.020701878766394895, "grad_norm": 4.318970760492421, "learning_rate": 6.886792452830189e-06, "loss": 0.5365, "step": 292 }, { "epoch": 0.02077277561148529, "grad_norm": 6.916444543441868, "learning_rate": 6.910377358490566e-06, "loss": 0.5504, "step": 293 }, { "epoch": 0.020843672456575684, "grad_norm": 5.294494004042308, "learning_rate": 6.933962264150944e-06, "loss": 0.526, "step": 294 }, { "epoch": 0.020914569301666075, "grad_norm": 8.562114837085748, "learning_rate": 6.957547169811321e-06, "loss": 0.4904, "step": 295 }, { "epoch": 0.02098546614675647, "grad_norm": 5.340035026334745, "learning_rate": 6.981132075471699e-06, "loss": 0.5519, "step": 296 }, { "epoch": 0.021056362991846864, "grad_norm": 4.955664370533099, "learning_rate": 7.004716981132076e-06, "loss": 0.4994, "step": 297 }, { "epoch": 0.021127259836937258, "grad_norm": 4.344167880308369, "learning_rate": 7.028301886792454e-06, "loss": 0.5182, "step": 298 }, { "epoch": 0.02119815668202765, "grad_norm": 1.3257885546853445, "learning_rate": 7.051886792452831e-06, "loss": 0.6689, "step": 299 }, { "epoch": 0.021269053527118043, "grad_norm": 28.48273750377888, "learning_rate": 7.0754716981132075e-06, "loss": 0.4936, "step": 300 }, { "epoch": 0.021339950372208438, "grad_norm": 5.475347308003009, "learning_rate": 7.099056603773586e-06, "loss": 0.5337, "step": 301 }, { "epoch": 0.021410847217298832, "grad_norm": 3.8558806957124085, "learning_rate": 7.122641509433963e-06, "loss": 0.5373, "step": 302 }, { "epoch": 0.021481744062389223, "grad_norm": 5.007024209133945, "learning_rate": 7.146226415094341e-06, "loss": 0.5359, "step": 303 }, { "epoch": 0.021552640907479617, "grad_norm": 3.82455824895914, "learning_rate": 7.169811320754717e-06, "loss": 0.5771, "step": 304 }, { "epoch": 0.02162353775257001, "grad_norm": 5.1197305784992935, "learning_rate": 7.1933962264150945e-06, "loss": 0.5738, "step": 305 }, { "epoch": 0.021694434597660406, "grad_norm": 5.0328179834436675, "learning_rate": 7.216981132075473e-06, "loss": 0.5266, "step": 306 }, { "epoch": 0.021765331442750797, "grad_norm": 2.5906430666008124, "learning_rate": 7.24056603773585e-06, "loss": 0.53, "step": 307 }, { "epoch": 0.02183622828784119, "grad_norm": 3.797481525383175, "learning_rate": 7.264150943396226e-06, "loss": 0.5029, "step": 308 }, { "epoch": 0.021907125132931585, "grad_norm": 2.9651381911281507, "learning_rate": 7.287735849056604e-06, "loss": 0.4761, "step": 309 }, { "epoch": 0.02197802197802198, "grad_norm": 2.698225033213811, "learning_rate": 7.3113207547169815e-06, "loss": 0.4885, "step": 310 }, { "epoch": 0.02204891882311237, "grad_norm": 6.662265331335316, "learning_rate": 7.33490566037736e-06, "loss": 0.5247, "step": 311 }, { "epoch": 0.022119815668202765, "grad_norm": 15.152888445673495, "learning_rate": 7.358490566037736e-06, "loss": 0.5153, "step": 312 }, { "epoch": 0.02219071251329316, "grad_norm": 3.4088421554162167, "learning_rate": 7.382075471698113e-06, "loss": 0.4599, "step": 313 }, { "epoch": 0.022261609358383554, "grad_norm": 6.260572407104118, "learning_rate": 7.405660377358491e-06, "loss": 0.5931, "step": 314 }, { "epoch": 0.022332506203473945, "grad_norm": 4.317116095065182, "learning_rate": 7.4292452830188685e-06, "loss": 0.4749, "step": 315 }, { "epoch": 0.02240340304856434, "grad_norm": 12.08632659050467, "learning_rate": 7.452830188679246e-06, "loss": 0.5197, "step": 316 }, { "epoch": 0.022474299893654733, "grad_norm": 5.052135290956749, "learning_rate": 7.476415094339623e-06, "loss": 0.5142, "step": 317 }, { "epoch": 0.022545196738745128, "grad_norm": 4.910144460189018, "learning_rate": 7.500000000000001e-06, "loss": 0.5186, "step": 318 }, { "epoch": 0.02261609358383552, "grad_norm": 5.307629881541831, "learning_rate": 7.523584905660378e-06, "loss": 0.5219, "step": 319 }, { "epoch": 0.022686990428925913, "grad_norm": 4.892031030730444, "learning_rate": 7.5471698113207555e-06, "loss": 0.5259, "step": 320 }, { "epoch": 0.022757887274016307, "grad_norm": 5.908440728705891, "learning_rate": 7.570754716981133e-06, "loss": 0.5895, "step": 321 }, { "epoch": 0.0228287841191067, "grad_norm": 3.390308914725182, "learning_rate": 7.59433962264151e-06, "loss": 0.497, "step": 322 }, { "epoch": 0.022899680964197092, "grad_norm": 5.581991531913085, "learning_rate": 7.617924528301888e-06, "loss": 0.5778, "step": 323 }, { "epoch": 0.022970577809287487, "grad_norm": 4.713498722221476, "learning_rate": 7.641509433962266e-06, "loss": 0.5355, "step": 324 }, { "epoch": 0.02304147465437788, "grad_norm": 8.900035614257007, "learning_rate": 7.665094339622642e-06, "loss": 0.5265, "step": 325 }, { "epoch": 0.023112371499468275, "grad_norm": 7.265379204883509, "learning_rate": 7.688679245283019e-06, "loss": 0.5376, "step": 326 }, { "epoch": 0.023183268344558666, "grad_norm": 7.872498523432269, "learning_rate": 7.712264150943398e-06, "loss": 0.5651, "step": 327 }, { "epoch": 0.02325416518964906, "grad_norm": 5.499282524390253, "learning_rate": 7.735849056603775e-06, "loss": 0.5521, "step": 328 }, { "epoch": 0.023325062034739455, "grad_norm": 4.793924816715041, "learning_rate": 7.75943396226415e-06, "loss": 0.5944, "step": 329 }, { "epoch": 0.02339595887982985, "grad_norm": 4.076215910404997, "learning_rate": 7.78301886792453e-06, "loss": 0.5206, "step": 330 }, { "epoch": 0.02346685572492024, "grad_norm": 4.6434269859595725, "learning_rate": 7.806603773584907e-06, "loss": 0.5169, "step": 331 }, { "epoch": 0.023537752570010635, "grad_norm": 5.013054276522261, "learning_rate": 7.830188679245284e-06, "loss": 0.4996, "step": 332 }, { "epoch": 0.02360864941510103, "grad_norm": 7.835904793493307, "learning_rate": 7.853773584905661e-06, "loss": 0.4848, "step": 333 }, { "epoch": 0.023679546260191423, "grad_norm": 15.940078535686993, "learning_rate": 7.877358490566038e-06, "loss": 0.456, "step": 334 }, { "epoch": 0.023750443105281814, "grad_norm": 3.9442755273777688, "learning_rate": 7.900943396226416e-06, "loss": 0.5056, "step": 335 }, { "epoch": 0.02382133995037221, "grad_norm": 6.225584392504023, "learning_rate": 7.924528301886793e-06, "loss": 0.5465, "step": 336 }, { "epoch": 0.023892236795462603, "grad_norm": 3.2730395706054325, "learning_rate": 7.94811320754717e-06, "loss": 0.5105, "step": 337 }, { "epoch": 0.023963133640552997, "grad_norm": 4.283593530213031, "learning_rate": 7.971698113207547e-06, "loss": 0.5666, "step": 338 }, { "epoch": 0.024034030485643388, "grad_norm": 14.788343545058652, "learning_rate": 7.995283018867925e-06, "loss": 0.4898, "step": 339 }, { "epoch": 0.024104927330733782, "grad_norm": 3.6141072003939456, "learning_rate": 8.018867924528303e-06, "loss": 0.5014, "step": 340 }, { "epoch": 0.024175824175824177, "grad_norm": 4.009152002533376, "learning_rate": 8.042452830188679e-06, "loss": 0.5008, "step": 341 }, { "epoch": 0.02424672102091457, "grad_norm": 3.6667412861131963, "learning_rate": 8.066037735849058e-06, "loss": 0.4636, "step": 342 }, { "epoch": 0.024317617866004962, "grad_norm": 4.509666040305149, "learning_rate": 8.089622641509435e-06, "loss": 0.511, "step": 343 }, { "epoch": 0.024388514711095356, "grad_norm": 8.422824864329481, "learning_rate": 8.113207547169812e-06, "loss": 0.4997, "step": 344 }, { "epoch": 0.02445941155618575, "grad_norm": 5.1307196773240475, "learning_rate": 8.13679245283019e-06, "loss": 0.5316, "step": 345 }, { "epoch": 0.024530308401276145, "grad_norm": 11.814989265045913, "learning_rate": 8.160377358490567e-06, "loss": 0.5204, "step": 346 }, { "epoch": 0.024601205246366536, "grad_norm": 9.340311866179121, "learning_rate": 8.183962264150944e-06, "loss": 0.5135, "step": 347 }, { "epoch": 0.02467210209145693, "grad_norm": 4.270933281810506, "learning_rate": 8.207547169811321e-06, "loss": 0.4991, "step": 348 }, { "epoch": 0.024742998936547324, "grad_norm": 4.494450791020251, "learning_rate": 8.231132075471699e-06, "loss": 0.532, "step": 349 }, { "epoch": 0.02481389578163772, "grad_norm": 3.6533666274028285, "learning_rate": 8.254716981132076e-06, "loss": 0.4907, "step": 350 }, { "epoch": 0.02488479262672811, "grad_norm": 1.088247678050943, "learning_rate": 8.278301886792453e-06, "loss": 0.7474, "step": 351 }, { "epoch": 0.024955689471818504, "grad_norm": 4.433402460860673, "learning_rate": 8.301886792452832e-06, "loss": 0.5698, "step": 352 }, { "epoch": 0.0250265863169089, "grad_norm": 3.0184780786888226, "learning_rate": 8.325471698113207e-06, "loss": 0.537, "step": 353 }, { "epoch": 0.025097483161999293, "grad_norm": 5.852948866730734, "learning_rate": 8.349056603773585e-06, "loss": 0.5274, "step": 354 }, { "epoch": 0.025168380007089684, "grad_norm": 3.3100337686847072, "learning_rate": 8.372641509433964e-06, "loss": 0.4761, "step": 355 }, { "epoch": 0.025239276852180078, "grad_norm": 6.349880254007272, "learning_rate": 8.39622641509434e-06, "loss": 0.5224, "step": 356 }, { "epoch": 0.025310173697270472, "grad_norm": 6.527952167064973, "learning_rate": 8.419811320754718e-06, "loss": 0.531, "step": 357 }, { "epoch": 0.025381070542360867, "grad_norm": 3.5950495656766317, "learning_rate": 8.443396226415095e-06, "loss": 0.5128, "step": 358 }, { "epoch": 0.025451967387451258, "grad_norm": 3.2040798921056877, "learning_rate": 8.466981132075472e-06, "loss": 0.4996, "step": 359 }, { "epoch": 0.025522864232541652, "grad_norm": 8.482661888402008, "learning_rate": 8.49056603773585e-06, "loss": 0.4697, "step": 360 }, { "epoch": 0.025593761077632046, "grad_norm": 5.11041175918556, "learning_rate": 8.514150943396227e-06, "loss": 0.5535, "step": 361 }, { "epoch": 0.02566465792272244, "grad_norm": 4.478311709326187, "learning_rate": 8.537735849056604e-06, "loss": 0.5661, "step": 362 }, { "epoch": 0.02573555476781283, "grad_norm": 2.9188483953697717, "learning_rate": 8.561320754716981e-06, "loss": 0.4904, "step": 363 }, { "epoch": 0.025806451612903226, "grad_norm": 1.4087031126608058, "learning_rate": 8.58490566037736e-06, "loss": 0.7237, "step": 364 }, { "epoch": 0.02587734845799362, "grad_norm": 7.659159879475842, "learning_rate": 8.608490566037736e-06, "loss": 0.5164, "step": 365 }, { "epoch": 0.025948245303084014, "grad_norm": 5.042593475951203, "learning_rate": 8.632075471698113e-06, "loss": 0.6077, "step": 366 }, { "epoch": 0.026019142148174405, "grad_norm": 5.409121840640401, "learning_rate": 8.655660377358492e-06, "loss": 0.5742, "step": 367 }, { "epoch": 0.0260900389932648, "grad_norm": 4.770185350965779, "learning_rate": 8.67924528301887e-06, "loss": 0.476, "step": 368 }, { "epoch": 0.026160935838355194, "grad_norm": 4.593998095099258, "learning_rate": 8.702830188679245e-06, "loss": 0.5316, "step": 369 }, { "epoch": 0.02623183268344559, "grad_norm": 6.007831165304678, "learning_rate": 8.726415094339624e-06, "loss": 0.5564, "step": 370 }, { "epoch": 0.02630272952853598, "grad_norm": 4.085021299732255, "learning_rate": 8.750000000000001e-06, "loss": 0.4599, "step": 371 }, { "epoch": 0.026373626373626374, "grad_norm": 3.7556681898272806, "learning_rate": 8.773584905660378e-06, "loss": 0.5081, "step": 372 }, { "epoch": 0.026444523218716768, "grad_norm": 3.97237969879046, "learning_rate": 8.797169811320755e-06, "loss": 0.5056, "step": 373 }, { "epoch": 0.026515420063807162, "grad_norm": 4.149621727601999, "learning_rate": 8.820754716981133e-06, "loss": 0.5312, "step": 374 }, { "epoch": 0.026586316908897553, "grad_norm": 5.285217804967652, "learning_rate": 8.84433962264151e-06, "loss": 0.4811, "step": 375 }, { "epoch": 0.026657213753987948, "grad_norm": 4.740446741002062, "learning_rate": 8.867924528301887e-06, "loss": 0.5751, "step": 376 }, { "epoch": 0.026728110599078342, "grad_norm": 5.76915663342388, "learning_rate": 8.891509433962266e-06, "loss": 0.5611, "step": 377 }, { "epoch": 0.026799007444168736, "grad_norm": 4.010729099476451, "learning_rate": 8.915094339622642e-06, "loss": 0.5115, "step": 378 }, { "epoch": 0.026869904289259127, "grad_norm": 3.400607309267024, "learning_rate": 8.938679245283019e-06, "loss": 0.5382, "step": 379 }, { "epoch": 0.02694080113434952, "grad_norm": 4.707224587208088, "learning_rate": 8.962264150943398e-06, "loss": 0.5148, "step": 380 }, { "epoch": 0.027011697979439916, "grad_norm": 6.574869626352459, "learning_rate": 8.985849056603775e-06, "loss": 0.5466, "step": 381 }, { "epoch": 0.02708259482453031, "grad_norm": 4.965363831647764, "learning_rate": 9.009433962264152e-06, "loss": 0.5303, "step": 382 }, { "epoch": 0.0271534916696207, "grad_norm": 3.338040074913745, "learning_rate": 9.03301886792453e-06, "loss": 0.4652, "step": 383 }, { "epoch": 0.027224388514711095, "grad_norm": 3.0798613356914775, "learning_rate": 9.056603773584907e-06, "loss": 0.4781, "step": 384 }, { "epoch": 0.02729528535980149, "grad_norm": 3.2361890288244832, "learning_rate": 9.080188679245284e-06, "loss": 0.4991, "step": 385 }, { "epoch": 0.027366182204891884, "grad_norm": 7.143498864064263, "learning_rate": 9.103773584905661e-06, "loss": 0.5122, "step": 386 }, { "epoch": 0.027437079049982275, "grad_norm": 3.8777066010339105, "learning_rate": 9.127358490566038e-06, "loss": 0.4981, "step": 387 }, { "epoch": 0.02750797589507267, "grad_norm": 3.3079583371688477, "learning_rate": 9.150943396226416e-06, "loss": 0.5246, "step": 388 }, { "epoch": 0.027578872740163064, "grad_norm": 3.5270627726933563, "learning_rate": 9.174528301886794e-06, "loss": 0.5063, "step": 389 }, { "epoch": 0.027649769585253458, "grad_norm": 3.563999356176284, "learning_rate": 9.19811320754717e-06, "loss": 0.6013, "step": 390 }, { "epoch": 0.02772066643034385, "grad_norm": 5.736701265246778, "learning_rate": 9.221698113207547e-06, "loss": 0.5351, "step": 391 }, { "epoch": 0.027791563275434243, "grad_norm": 3.2772850193875853, "learning_rate": 9.245283018867926e-06, "loss": 0.5001, "step": 392 }, { "epoch": 0.027862460120524637, "grad_norm": 4.814363203009247, "learning_rate": 9.268867924528303e-06, "loss": 0.5473, "step": 393 }, { "epoch": 0.027933356965615032, "grad_norm": 8.741355594969878, "learning_rate": 9.292452830188679e-06, "loss": 0.4745, "step": 394 }, { "epoch": 0.028004253810705423, "grad_norm": 3.385314346429951, "learning_rate": 9.316037735849058e-06, "loss": 0.4713, "step": 395 }, { "epoch": 0.028075150655795817, "grad_norm": 5.5010348122086805, "learning_rate": 9.339622641509435e-06, "loss": 0.4883, "step": 396 }, { "epoch": 0.02814604750088621, "grad_norm": 3.030180139615956, "learning_rate": 9.363207547169812e-06, "loss": 0.4835, "step": 397 }, { "epoch": 0.028216944345976606, "grad_norm": 7.7097329172224605, "learning_rate": 9.38679245283019e-06, "loss": 0.5567, "step": 398 }, { "epoch": 0.028287841191066997, "grad_norm": 12.786945640802681, "learning_rate": 9.410377358490567e-06, "loss": 0.5138, "step": 399 }, { "epoch": 0.02835873803615739, "grad_norm": 6.543070436638372, "learning_rate": 9.433962264150944e-06, "loss": 0.4948, "step": 400 }, { "epoch": 0.028429634881247785, "grad_norm": 3.826075075459927, "learning_rate": 9.457547169811321e-06, "loss": 0.4725, "step": 401 }, { "epoch": 0.02850053172633818, "grad_norm": 4.42208488072467, "learning_rate": 9.481132075471698e-06, "loss": 0.4291, "step": 402 }, { "epoch": 0.02857142857142857, "grad_norm": 2.206968681234039, "learning_rate": 9.504716981132076e-06, "loss": 0.5259, "step": 403 }, { "epoch": 0.028642325416518965, "grad_norm": 3.813975301237046, "learning_rate": 9.528301886792455e-06, "loss": 0.5094, "step": 404 }, { "epoch": 0.02871322226160936, "grad_norm": 3.906947789935457, "learning_rate": 9.551886792452832e-06, "loss": 0.5046, "step": 405 }, { "epoch": 0.028784119106699754, "grad_norm": 3.545470580960482, "learning_rate": 9.575471698113207e-06, "loss": 0.495, "step": 406 }, { "epoch": 0.028855015951790144, "grad_norm": 4.74109051901093, "learning_rate": 9.599056603773586e-06, "loss": 0.5895, "step": 407 }, { "epoch": 0.02892591279688054, "grad_norm": 3.827185380134332, "learning_rate": 9.622641509433963e-06, "loss": 0.4433, "step": 408 }, { "epoch": 0.028996809641970933, "grad_norm": 5.365282180429837, "learning_rate": 9.64622641509434e-06, "loss": 0.466, "step": 409 }, { "epoch": 0.029067706487061327, "grad_norm": 5.688228253187277, "learning_rate": 9.669811320754718e-06, "loss": 0.5095, "step": 410 }, { "epoch": 0.02913860333215172, "grad_norm": 3.2964892036446742, "learning_rate": 9.693396226415095e-06, "loss": 0.4847, "step": 411 }, { "epoch": 0.029209500177242113, "grad_norm": 4.735925068255346, "learning_rate": 9.716981132075472e-06, "loss": 0.5132, "step": 412 }, { "epoch": 0.029280397022332507, "grad_norm": 4.128734039163789, "learning_rate": 9.74056603773585e-06, "loss": 0.4849, "step": 413 }, { "epoch": 0.0293512938674229, "grad_norm": 5.929681616613668, "learning_rate": 9.764150943396227e-06, "loss": 0.5281, "step": 414 }, { "epoch": 0.029422190712513292, "grad_norm": 4.6243166762957175, "learning_rate": 9.787735849056604e-06, "loss": 0.4883, "step": 415 }, { "epoch": 0.029493087557603687, "grad_norm": 19.91795900062745, "learning_rate": 9.811320754716981e-06, "loss": 0.5017, "step": 416 }, { "epoch": 0.02956398440269408, "grad_norm": 7.515747176159766, "learning_rate": 9.83490566037736e-06, "loss": 0.5229, "step": 417 }, { "epoch": 0.029634881247784475, "grad_norm": 3.7555746098694804, "learning_rate": 9.858490566037736e-06, "loss": 0.507, "step": 418 }, { "epoch": 0.029705778092874866, "grad_norm": 7.72193772956651, "learning_rate": 9.882075471698113e-06, "loss": 0.5292, "step": 419 }, { "epoch": 0.02977667493796526, "grad_norm": 4.446325628342342, "learning_rate": 9.905660377358492e-06, "loss": 0.4847, "step": 420 }, { "epoch": 0.029847571783055655, "grad_norm": 4.638025330503627, "learning_rate": 9.929245283018869e-06, "loss": 0.5147, "step": 421 }, { "epoch": 0.02991846862814605, "grad_norm": 3.8154639559115995, "learning_rate": 9.952830188679246e-06, "loss": 0.4718, "step": 422 }, { "epoch": 0.02998936547323644, "grad_norm": 10.368026887668881, "learning_rate": 9.976415094339624e-06, "loss": 0.4519, "step": 423 }, { "epoch": 0.030060262318326834, "grad_norm": 3.64740120324922, "learning_rate": 1e-05, "loss": 0.4855, "step": 424 }, { "epoch": 0.03013115916341723, "grad_norm": 4.833076164980656, "learning_rate": 9.999999868173092e-06, "loss": 0.4957, "step": 425 }, { "epoch": 0.030202056008507623, "grad_norm": 3.753752605372728, "learning_rate": 9.999999472692374e-06, "loss": 0.414, "step": 426 }, { "epoch": 0.030272952853598014, "grad_norm": 7.087444703247379, "learning_rate": 9.99999881355787e-06, "loss": 0.5083, "step": 427 }, { "epoch": 0.03034384969868841, "grad_norm": 5.35774957181629, "learning_rate": 9.99999789076961e-06, "loss": 0.5035, "step": 428 }, { "epoch": 0.030414746543778803, "grad_norm": 5.616134443885102, "learning_rate": 9.999996704327643e-06, "loss": 0.5905, "step": 429 }, { "epoch": 0.030485643388869197, "grad_norm": 4.28581396755678, "learning_rate": 9.999995254232037e-06, "loss": 0.4706, "step": 430 }, { "epoch": 0.030556540233959588, "grad_norm": 2.8277929468150336, "learning_rate": 9.999993540482864e-06, "loss": 0.5422, "step": 431 }, { "epoch": 0.030627437079049982, "grad_norm": 3.606076546855887, "learning_rate": 9.999991563080213e-06, "loss": 0.4912, "step": 432 }, { "epoch": 0.030698333924140377, "grad_norm": 1.2613203491877625, "learning_rate": 9.999989322024192e-06, "loss": 0.7038, "step": 433 }, { "epoch": 0.03076923076923077, "grad_norm": 2.539203015020222, "learning_rate": 9.999986817314919e-06, "loss": 0.4786, "step": 434 }, { "epoch": 0.030840127614321162, "grad_norm": 3.314193016467025, "learning_rate": 9.999984048952525e-06, "loss": 0.4743, "step": 435 }, { "epoch": 0.030911024459411556, "grad_norm": 2.6188491779484044, "learning_rate": 9.999981016937154e-06, "loss": 0.4493, "step": 436 }, { "epoch": 0.03098192130450195, "grad_norm": 6.244476260445374, "learning_rate": 9.999977721268968e-06, "loss": 0.5393, "step": 437 }, { "epoch": 0.031052818149592345, "grad_norm": 3.288500137377778, "learning_rate": 9.99997416194814e-06, "loss": 0.4704, "step": 438 }, { "epoch": 0.031123714994682736, "grad_norm": 5.8366007010245315, "learning_rate": 9.999970338974861e-06, "loss": 0.4929, "step": 439 }, { "epoch": 0.03119461183977313, "grad_norm": 3.86021725729832, "learning_rate": 9.999966252349327e-06, "loss": 0.5451, "step": 440 }, { "epoch": 0.031265508684863524, "grad_norm": 7.837439914504211, "learning_rate": 9.999961902071755e-06, "loss": 0.5454, "step": 441 }, { "epoch": 0.03133640552995392, "grad_norm": 3.7204164081776354, "learning_rate": 9.999957288142379e-06, "loss": 0.5093, "step": 442 }, { "epoch": 0.03140730237504431, "grad_norm": 2.6860046096479455, "learning_rate": 9.999952410561436e-06, "loss": 0.4641, "step": 443 }, { "epoch": 0.03147819922013471, "grad_norm": 3.6164333528714274, "learning_rate": 9.999947269329188e-06, "loss": 0.5001, "step": 444 }, { "epoch": 0.031549096065225095, "grad_norm": 4.206244501286116, "learning_rate": 9.999941864445903e-06, "loss": 0.5457, "step": 445 }, { "epoch": 0.03161999291031549, "grad_norm": 4.023029885192812, "learning_rate": 9.999936195911867e-06, "loss": 0.5006, "step": 446 }, { "epoch": 0.031690889755405884, "grad_norm": 4.652810315525827, "learning_rate": 9.999930263727382e-06, "loss": 0.5198, "step": 447 }, { "epoch": 0.03176178660049628, "grad_norm": 10.40891343058792, "learning_rate": 9.999924067892755e-06, "loss": 0.5101, "step": 448 }, { "epoch": 0.03183268344558667, "grad_norm": 3.443290131923013, "learning_rate": 9.999917608408317e-06, "loss": 0.5453, "step": 449 }, { "epoch": 0.03190358029067707, "grad_norm": 4.433574612133371, "learning_rate": 9.999910885274406e-06, "loss": 0.4903, "step": 450 }, { "epoch": 0.03197447713576746, "grad_norm": 8.110446115011506, "learning_rate": 9.999903898491378e-06, "loss": 0.4402, "step": 451 }, { "epoch": 0.032045373980857855, "grad_norm": 6.368345271165315, "learning_rate": 9.999896648059603e-06, "loss": 0.4735, "step": 452 }, { "epoch": 0.03211627082594824, "grad_norm": 3.9698442802213707, "learning_rate": 9.999889133979459e-06, "loss": 0.5034, "step": 453 }, { "epoch": 0.03218716767103864, "grad_norm": 4.218662830046309, "learning_rate": 9.999881356251346e-06, "loss": 0.4482, "step": 454 }, { "epoch": 0.03225806451612903, "grad_norm": 2.8980995437355244, "learning_rate": 9.999873314875674e-06, "loss": 0.4734, "step": 455 }, { "epoch": 0.032328961361219426, "grad_norm": 4.440619160384881, "learning_rate": 9.999865009852865e-06, "loss": 0.5335, "step": 456 }, { "epoch": 0.03239985820630982, "grad_norm": 5.950792692739966, "learning_rate": 9.999856441183357e-06, "loss": 0.5747, "step": 457 }, { "epoch": 0.032470755051400214, "grad_norm": 6.058959203833448, "learning_rate": 9.999847608867603e-06, "loss": 0.5049, "step": 458 }, { "epoch": 0.03254165189649061, "grad_norm": 6.362361557816691, "learning_rate": 9.999838512906069e-06, "loss": 0.4998, "step": 459 }, { "epoch": 0.032612548741581, "grad_norm": 4.1660222172367325, "learning_rate": 9.999829153299233e-06, "loss": 0.5273, "step": 460 }, { "epoch": 0.03268344558667139, "grad_norm": 5.427371497816613, "learning_rate": 9.99981953004759e-06, "loss": 0.4787, "step": 461 }, { "epoch": 0.032754342431761785, "grad_norm": 2.8280965696709055, "learning_rate": 9.999809643151646e-06, "loss": 0.5137, "step": 462 }, { "epoch": 0.03282523927685218, "grad_norm": 10.689705007213057, "learning_rate": 9.999799492611925e-06, "loss": 0.4536, "step": 463 }, { "epoch": 0.032896136121942574, "grad_norm": 5.1788525202696905, "learning_rate": 9.99978907842896e-06, "loss": 0.549, "step": 464 }, { "epoch": 0.03296703296703297, "grad_norm": 3.6850876564636033, "learning_rate": 9.999778400603301e-06, "loss": 0.4811, "step": 465 }, { "epoch": 0.03303792981212336, "grad_norm": 6.617178927564454, "learning_rate": 9.99976745913551e-06, "loss": 0.5287, "step": 466 }, { "epoch": 0.03310882665721376, "grad_norm": 8.373552387604521, "learning_rate": 9.999756254026165e-06, "loss": 0.5397, "step": 467 }, { "epoch": 0.03317972350230415, "grad_norm": 5.380304590131256, "learning_rate": 9.999744785275857e-06, "loss": 0.4997, "step": 468 }, { "epoch": 0.03325062034739454, "grad_norm": 3.6175411363906296, "learning_rate": 9.99973305288519e-06, "loss": 0.4666, "step": 469 }, { "epoch": 0.03332151719248493, "grad_norm": 12.359904430055257, "learning_rate": 9.999721056854784e-06, "loss": 0.494, "step": 470 }, { "epoch": 0.03339241403757533, "grad_norm": 3.9186696416818747, "learning_rate": 9.999708797185269e-06, "loss": 0.4942, "step": 471 }, { "epoch": 0.03346331088266572, "grad_norm": 3.784534088079515, "learning_rate": 9.999696273877293e-06, "loss": 0.504, "step": 472 }, { "epoch": 0.033534207727756116, "grad_norm": 4.711536586469255, "learning_rate": 9.999683486931517e-06, "loss": 0.508, "step": 473 }, { "epoch": 0.03360510457284651, "grad_norm": 3.675535224528616, "learning_rate": 9.999670436348614e-06, "loss": 0.5107, "step": 474 }, { "epoch": 0.033676001417936904, "grad_norm": 3.650726048180102, "learning_rate": 9.999657122129274e-06, "loss": 0.495, "step": 475 }, { "epoch": 0.0337468982630273, "grad_norm": 4.438167570206442, "learning_rate": 9.999643544274198e-06, "loss": 0.5104, "step": 476 }, { "epoch": 0.033817795108117686, "grad_norm": 2.7020847127769074, "learning_rate": 9.999629702784101e-06, "loss": 0.5159, "step": 477 }, { "epoch": 0.03388869195320808, "grad_norm": 2.7885269531369667, "learning_rate": 9.999615597659713e-06, "loss": 0.4627, "step": 478 }, { "epoch": 0.033959588798298475, "grad_norm": 18.962018234781375, "learning_rate": 9.999601228901782e-06, "loss": 0.5133, "step": 479 }, { "epoch": 0.03403048564338887, "grad_norm": 3.5661462354299918, "learning_rate": 9.999586596511057e-06, "loss": 0.4959, "step": 480 }, { "epoch": 0.034101382488479264, "grad_norm": 3.4512553301719406, "learning_rate": 9.999571700488318e-06, "loss": 0.4773, "step": 481 }, { "epoch": 0.03417227933356966, "grad_norm": 4.955746979981906, "learning_rate": 9.999556540834348e-06, "loss": 0.5258, "step": 482 }, { "epoch": 0.03424317617866005, "grad_norm": 4.039341675537325, "learning_rate": 9.999541117549942e-06, "loss": 0.5115, "step": 483 }, { "epoch": 0.03431407302375045, "grad_norm": 4.6730618626471205, "learning_rate": 9.999525430635921e-06, "loss": 0.5428, "step": 484 }, { "epoch": 0.034384969868840834, "grad_norm": 4.860039801709536, "learning_rate": 9.999509480093105e-06, "loss": 0.5198, "step": 485 }, { "epoch": 0.03445586671393123, "grad_norm": 3.175762678083821, "learning_rate": 9.99949326592234e-06, "loss": 0.4792, "step": 486 }, { "epoch": 0.03452676355902162, "grad_norm": 5.721243253959769, "learning_rate": 9.999476788124478e-06, "loss": 0.5049, "step": 487 }, { "epoch": 0.03459766040411202, "grad_norm": 5.32791415421032, "learning_rate": 9.99946004670039e-06, "loss": 0.4741, "step": 488 }, { "epoch": 0.03466855724920241, "grad_norm": 2.7548460808072366, "learning_rate": 9.999443041650958e-06, "loss": 0.4901, "step": 489 }, { "epoch": 0.034739454094292806, "grad_norm": 6.003499355888121, "learning_rate": 9.999425772977076e-06, "loss": 0.5125, "step": 490 }, { "epoch": 0.0348103509393832, "grad_norm": 9.87114966300572, "learning_rate": 9.99940824067966e-06, "loss": 0.5206, "step": 491 }, { "epoch": 0.034881247784473594, "grad_norm": 3.910258438074311, "learning_rate": 9.999390444759632e-06, "loss": 0.508, "step": 492 }, { "epoch": 0.03495214462956398, "grad_norm": 4.093673340109355, "learning_rate": 9.999372385217928e-06, "loss": 0.5035, "step": 493 }, { "epoch": 0.035023041474654376, "grad_norm": 3.319302384910709, "learning_rate": 9.999354062055503e-06, "loss": 0.5564, "step": 494 }, { "epoch": 0.03509393831974477, "grad_norm": 3.8674680266482784, "learning_rate": 9.999335475273322e-06, "loss": 0.5355, "step": 495 }, { "epoch": 0.035164835164835165, "grad_norm": 17.836955700282612, "learning_rate": 9.999316624872366e-06, "loss": 0.5425, "step": 496 }, { "epoch": 0.03523573200992556, "grad_norm": 4.457051033674754, "learning_rate": 9.999297510853627e-06, "loss": 0.4742, "step": 497 }, { "epoch": 0.035306628855015953, "grad_norm": 3.443979506292567, "learning_rate": 9.999278133218117e-06, "loss": 0.467, "step": 498 }, { "epoch": 0.03537752570010635, "grad_norm": 6.254522318977552, "learning_rate": 9.999258491966854e-06, "loss": 0.5493, "step": 499 }, { "epoch": 0.03544842254519674, "grad_norm": 3.4883278133409794, "learning_rate": 9.999238587100875e-06, "loss": 0.4878, "step": 500 }, { "epoch": 0.03551931939028713, "grad_norm": 6.95507691080322, "learning_rate": 9.99921841862123e-06, "loss": 0.5217, "step": 501 }, { "epoch": 0.035590216235377524, "grad_norm": 4.4462703329718245, "learning_rate": 9.999197986528981e-06, "loss": 0.5095, "step": 502 }, { "epoch": 0.03566111308046792, "grad_norm": 5.179319142090798, "learning_rate": 9.999177290825208e-06, "loss": 0.5056, "step": 503 }, { "epoch": 0.03573200992555831, "grad_norm": 5.42647907646635, "learning_rate": 9.999156331511e-06, "loss": 0.4946, "step": 504 }, { "epoch": 0.03580290677064871, "grad_norm": 2.1985621420135644, "learning_rate": 9.999135108587464e-06, "loss": 0.4967, "step": 505 }, { "epoch": 0.0358738036157391, "grad_norm": 3.6308707122097914, "learning_rate": 9.999113622055715e-06, "loss": 0.5382, "step": 506 }, { "epoch": 0.035944700460829496, "grad_norm": 4.334092436815336, "learning_rate": 9.999091871916892e-06, "loss": 0.5306, "step": 507 }, { "epoch": 0.03601559730591989, "grad_norm": 3.5391973174038003, "learning_rate": 9.999069858172138e-06, "loss": 0.4988, "step": 508 }, { "epoch": 0.03608649415101028, "grad_norm": 7.152263177866166, "learning_rate": 9.999047580822614e-06, "loss": 0.4742, "step": 509 }, { "epoch": 0.03615739099610067, "grad_norm": 4.9585775268198, "learning_rate": 9.999025039869497e-06, "loss": 0.5198, "step": 510 }, { "epoch": 0.036228287841191066, "grad_norm": 3.9845040648782235, "learning_rate": 9.999002235313973e-06, "loss": 0.4718, "step": 511 }, { "epoch": 0.03629918468628146, "grad_norm": 5.838345557806041, "learning_rate": 9.998979167157246e-06, "loss": 0.5632, "step": 512 }, { "epoch": 0.036370081531371855, "grad_norm": 6.061592441579902, "learning_rate": 9.998955835400533e-06, "loss": 0.4597, "step": 513 }, { "epoch": 0.03644097837646225, "grad_norm": 3.296813937469406, "learning_rate": 9.99893224004506e-06, "loss": 0.4949, "step": 514 }, { "epoch": 0.036511875221552643, "grad_norm": 8.166186165044241, "learning_rate": 9.998908381092077e-06, "loss": 0.4919, "step": 515 }, { "epoch": 0.03658277206664304, "grad_norm": 4.654478696040088, "learning_rate": 9.99888425854284e-06, "loss": 0.503, "step": 516 }, { "epoch": 0.036653668911733425, "grad_norm": 3.5489455111989443, "learning_rate": 9.998859872398619e-06, "loss": 0.4794, "step": 517 }, { "epoch": 0.03672456575682382, "grad_norm": 3.2736384183396208, "learning_rate": 9.998835222660702e-06, "loss": 0.4493, "step": 518 }, { "epoch": 0.036795462601914214, "grad_norm": 4.839872529169615, "learning_rate": 9.998810309330388e-06, "loss": 0.4644, "step": 519 }, { "epoch": 0.03686635944700461, "grad_norm": 3.2241863319739315, "learning_rate": 9.998785132408992e-06, "loss": 0.4642, "step": 520 }, { "epoch": 0.036937256292095, "grad_norm": 3.592908388805739, "learning_rate": 9.99875969189784e-06, "loss": 0.5018, "step": 521 }, { "epoch": 0.0370081531371854, "grad_norm": 5.563549649440812, "learning_rate": 9.998733987798275e-06, "loss": 0.4601, "step": 522 }, { "epoch": 0.03707904998227579, "grad_norm": 3.65305617439303, "learning_rate": 9.998708020111649e-06, "loss": 0.4613, "step": 523 }, { "epoch": 0.037149946827366186, "grad_norm": 3.182299969760936, "learning_rate": 9.998681788839337e-06, "loss": 0.4813, "step": 524 }, { "epoch": 0.03722084367245657, "grad_norm": 3.1498944120259407, "learning_rate": 9.998655293982717e-06, "loss": 0.5149, "step": 525 }, { "epoch": 0.03729174051754697, "grad_norm": 2.9279003529066183, "learning_rate": 9.998628535543187e-06, "loss": 0.5181, "step": 526 }, { "epoch": 0.03736263736263736, "grad_norm": 2.3947849430593244, "learning_rate": 9.998601513522161e-06, "loss": 0.5224, "step": 527 }, { "epoch": 0.037433534207727756, "grad_norm": 3.339387801200832, "learning_rate": 9.99857422792106e-06, "loss": 0.4408, "step": 528 }, { "epoch": 0.03750443105281815, "grad_norm": 1.372455150273396, "learning_rate": 9.998546678741328e-06, "loss": 0.7283, "step": 529 }, { "epoch": 0.037575327897908545, "grad_norm": 5.764501452281906, "learning_rate": 9.998518865984412e-06, "loss": 0.482, "step": 530 }, { "epoch": 0.03764622474299894, "grad_norm": 4.26619470199033, "learning_rate": 9.99849078965178e-06, "loss": 0.5157, "step": 531 }, { "epoch": 0.03771712158808933, "grad_norm": 4.295093067553618, "learning_rate": 9.998462449744915e-06, "loss": 0.5646, "step": 532 }, { "epoch": 0.03778801843317972, "grad_norm": 3.10762106976791, "learning_rate": 9.998433846265311e-06, "loss": 0.5024, "step": 533 }, { "epoch": 0.037858915278270115, "grad_norm": 2.8157872325367035, "learning_rate": 9.998404979214473e-06, "loss": 0.4746, "step": 534 }, { "epoch": 0.03792981212336051, "grad_norm": 2.643195595732194, "learning_rate": 9.998375848593926e-06, "loss": 0.4734, "step": 535 }, { "epoch": 0.038000708968450904, "grad_norm": 2.822331904293088, "learning_rate": 9.998346454405207e-06, "loss": 0.4649, "step": 536 }, { "epoch": 0.0380716058135413, "grad_norm": 4.981861399181221, "learning_rate": 9.998316796649863e-06, "loss": 0.5167, "step": 537 }, { "epoch": 0.03814250265863169, "grad_norm": 5.9867175541694015, "learning_rate": 9.998286875329459e-06, "loss": 0.4873, "step": 538 }, { "epoch": 0.03821339950372209, "grad_norm": 4.516786670266049, "learning_rate": 9.998256690445573e-06, "loss": 0.5076, "step": 539 }, { "epoch": 0.03828429634881248, "grad_norm": 6.449218490031399, "learning_rate": 9.998226241999798e-06, "loss": 0.4777, "step": 540 }, { "epoch": 0.03835519319390287, "grad_norm": 3.026236856128972, "learning_rate": 9.998195529993738e-06, "loss": 0.4579, "step": 541 }, { "epoch": 0.03842609003899326, "grad_norm": 11.984257344507133, "learning_rate": 9.99816455442901e-06, "loss": 0.4719, "step": 542 }, { "epoch": 0.03849698688408366, "grad_norm": 7.021280238583561, "learning_rate": 9.998133315307253e-06, "loss": 0.5032, "step": 543 }, { "epoch": 0.03856788372917405, "grad_norm": 5.371521117042946, "learning_rate": 9.99810181263011e-06, "loss": 0.5187, "step": 544 }, { "epoch": 0.038638780574264446, "grad_norm": 4.750021875202984, "learning_rate": 9.998070046399245e-06, "loss": 0.4454, "step": 545 }, { "epoch": 0.03870967741935484, "grad_norm": 3.8401009516336786, "learning_rate": 9.99803801661633e-06, "loss": 0.4968, "step": 546 }, { "epoch": 0.038780574264445235, "grad_norm": 11.14046651564773, "learning_rate": 9.998005723283057e-06, "loss": 0.4766, "step": 547 }, { "epoch": 0.03885147110953563, "grad_norm": 3.4686892037039847, "learning_rate": 9.997973166401128e-06, "loss": 0.4457, "step": 548 }, { "epoch": 0.038922367954626017, "grad_norm": 4.6354281856970925, "learning_rate": 9.997940345972256e-06, "loss": 0.4781, "step": 549 }, { "epoch": 0.03899326479971641, "grad_norm": 3.3567571586388643, "learning_rate": 9.997907261998177e-06, "loss": 0.5032, "step": 550 }, { "epoch": 0.039064161644806805, "grad_norm": 4.334081450126684, "learning_rate": 9.997873914480633e-06, "loss": 0.5101, "step": 551 }, { "epoch": 0.0391350584898972, "grad_norm": 5.621052961129649, "learning_rate": 9.997840303421383e-06, "loss": 0.5212, "step": 552 }, { "epoch": 0.039205955334987594, "grad_norm": 3.9139504648056804, "learning_rate": 9.997806428822199e-06, "loss": 0.4972, "step": 553 }, { "epoch": 0.03927685218007799, "grad_norm": 3.2896240343041696, "learning_rate": 9.997772290684868e-06, "loss": 0.4717, "step": 554 }, { "epoch": 0.03934774902516838, "grad_norm": 13.67766990256712, "learning_rate": 9.997737889011187e-06, "loss": 0.5358, "step": 555 }, { "epoch": 0.03941864587025878, "grad_norm": 3.125518775886191, "learning_rate": 9.997703223802975e-06, "loss": 0.507, "step": 556 }, { "epoch": 0.039489542715349164, "grad_norm": 3.369166271196928, "learning_rate": 9.997668295062057e-06, "loss": 0.4826, "step": 557 }, { "epoch": 0.03956043956043956, "grad_norm": 2.7614818300269883, "learning_rate": 9.997633102790275e-06, "loss": 0.4669, "step": 558 }, { "epoch": 0.03963133640552995, "grad_norm": 3.806647339462525, "learning_rate": 9.997597646989485e-06, "loss": 0.4584, "step": 559 }, { "epoch": 0.03970223325062035, "grad_norm": 1.2844462305367756, "learning_rate": 9.997561927661557e-06, "loss": 0.7443, "step": 560 }, { "epoch": 0.03977313009571074, "grad_norm": 5.534309904378585, "learning_rate": 9.997525944808373e-06, "loss": 0.5129, "step": 561 }, { "epoch": 0.039844026940801136, "grad_norm": 7.598962892561292, "learning_rate": 9.997489698431832e-06, "loss": 0.4736, "step": 562 }, { "epoch": 0.03991492378589153, "grad_norm": 1.8202505105536053, "learning_rate": 9.997453188533845e-06, "loss": 0.6968, "step": 563 }, { "epoch": 0.039985820630981925, "grad_norm": 6.268173256555975, "learning_rate": 9.997416415116337e-06, "loss": 0.4772, "step": 564 }, { "epoch": 0.04005671747607231, "grad_norm": 4.980844161149229, "learning_rate": 9.997379378181247e-06, "loss": 0.4612, "step": 565 }, { "epoch": 0.040127614321162706, "grad_norm": 4.43214800355919, "learning_rate": 9.997342077730528e-06, "loss": 0.4958, "step": 566 }, { "epoch": 0.0401985111662531, "grad_norm": 2.32750577460637, "learning_rate": 9.997304513766148e-06, "loss": 0.5021, "step": 567 }, { "epoch": 0.040269408011343495, "grad_norm": 8.267969619229751, "learning_rate": 9.997266686290084e-06, "loss": 0.5283, "step": 568 }, { "epoch": 0.04034030485643389, "grad_norm": 1.3218742442390998, "learning_rate": 9.997228595304335e-06, "loss": 0.6774, "step": 569 }, { "epoch": 0.040411201701524284, "grad_norm": 3.8983028804827704, "learning_rate": 9.997190240810909e-06, "loss": 0.5151, "step": 570 }, { "epoch": 0.04048209854661468, "grad_norm": 3.5065844660377077, "learning_rate": 9.997151622811828e-06, "loss": 0.456, "step": 571 }, { "epoch": 0.04055299539170507, "grad_norm": 4.850322822284982, "learning_rate": 9.997112741309125e-06, "loss": 0.5383, "step": 572 }, { "epoch": 0.04062389223679546, "grad_norm": 5.058793037447801, "learning_rate": 9.997073596304853e-06, "loss": 0.4425, "step": 573 }, { "epoch": 0.040694789081885854, "grad_norm": 4.6756894021622, "learning_rate": 9.997034187801078e-06, "loss": 0.5178, "step": 574 }, { "epoch": 0.04076568592697625, "grad_norm": 4.076728374601589, "learning_rate": 9.996994515799876e-06, "loss": 0.4196, "step": 575 }, { "epoch": 0.04083658277206664, "grad_norm": 4.39145810476105, "learning_rate": 9.99695458030334e-06, "loss": 0.5227, "step": 576 }, { "epoch": 0.04090747961715704, "grad_norm": 2.849037136371157, "learning_rate": 9.996914381313573e-06, "loss": 0.4538, "step": 577 }, { "epoch": 0.04097837646224743, "grad_norm": 3.097463426226102, "learning_rate": 9.996873918832696e-06, "loss": 0.3827, "step": 578 }, { "epoch": 0.041049273307337826, "grad_norm": 4.009206748677541, "learning_rate": 9.996833192862845e-06, "loss": 0.4473, "step": 579 }, { "epoch": 0.04112017015242822, "grad_norm": 3.0340467571743894, "learning_rate": 9.996792203406167e-06, "loss": 0.4475, "step": 580 }, { "epoch": 0.04119106699751861, "grad_norm": 3.344082417601275, "learning_rate": 9.99675095046482e-06, "loss": 0.4976, "step": 581 }, { "epoch": 0.041261963842609, "grad_norm": 5.897386881017899, "learning_rate": 9.996709434040983e-06, "loss": 0.4933, "step": 582 }, { "epoch": 0.041332860687699396, "grad_norm": 4.347114044535193, "learning_rate": 9.996667654136842e-06, "loss": 0.5116, "step": 583 }, { "epoch": 0.04140375753278979, "grad_norm": 2.8728923075732307, "learning_rate": 9.996625610754604e-06, "loss": 0.4623, "step": 584 }, { "epoch": 0.041474654377880185, "grad_norm": 4.73147173283241, "learning_rate": 9.996583303896484e-06, "loss": 0.4953, "step": 585 }, { "epoch": 0.04154555122297058, "grad_norm": 4.17388322592462, "learning_rate": 9.99654073356471e-06, "loss": 0.4771, "step": 586 }, { "epoch": 0.041616448068060974, "grad_norm": 5.642592345064658, "learning_rate": 9.996497899761531e-06, "loss": 0.4586, "step": 587 }, { "epoch": 0.04168734491315137, "grad_norm": 3.8987507414129006, "learning_rate": 9.996454802489203e-06, "loss": 0.4759, "step": 588 }, { "epoch": 0.041758241758241756, "grad_norm": 3.6356491382504648, "learning_rate": 9.996411441750003e-06, "loss": 0.4794, "step": 589 }, { "epoch": 0.04182913860333215, "grad_norm": 4.4735458877030165, "learning_rate": 9.99636781754621e-06, "loss": 0.4746, "step": 590 }, { "epoch": 0.041900035448422544, "grad_norm": 3.71382959068629, "learning_rate": 9.996323929880131e-06, "loss": 0.4578, "step": 591 }, { "epoch": 0.04197093229351294, "grad_norm": 3.6571449832604688, "learning_rate": 9.996279778754077e-06, "loss": 0.4854, "step": 592 }, { "epoch": 0.04204182913860333, "grad_norm": 2.9444246258856417, "learning_rate": 9.996235364170376e-06, "loss": 0.4237, "step": 593 }, { "epoch": 0.04211272598369373, "grad_norm": 10.744680404773371, "learning_rate": 9.99619068613137e-06, "loss": 0.4744, "step": 594 }, { "epoch": 0.04218362282878412, "grad_norm": 2.7097508236893364, "learning_rate": 9.996145744639419e-06, "loss": 0.5108, "step": 595 }, { "epoch": 0.042254519673874516, "grad_norm": 6.4487043348311035, "learning_rate": 9.996100539696887e-06, "loss": 0.4935, "step": 596 }, { "epoch": 0.0423254165189649, "grad_norm": 5.024236318674306, "learning_rate": 9.996055071306162e-06, "loss": 0.4958, "step": 597 }, { "epoch": 0.0423963133640553, "grad_norm": 14.260503903485535, "learning_rate": 9.996009339469636e-06, "loss": 0.522, "step": 598 }, { "epoch": 0.04246721020914569, "grad_norm": 4.009283815683314, "learning_rate": 9.995963344189728e-06, "loss": 0.5024, "step": 599 }, { "epoch": 0.042538107054236086, "grad_norm": 4.047062875138479, "learning_rate": 9.995917085468858e-06, "loss": 0.5047, "step": 600 }, { "epoch": 0.04260900389932648, "grad_norm": 4.295713405966012, "learning_rate": 9.995870563309466e-06, "loss": 0.4845, "step": 601 }, { "epoch": 0.042679900744416875, "grad_norm": 4.103096676187277, "learning_rate": 9.995823777714008e-06, "loss": 0.4383, "step": 602 }, { "epoch": 0.04275079758950727, "grad_norm": 3.6317534464991637, "learning_rate": 9.99577672868495e-06, "loss": 0.481, "step": 603 }, { "epoch": 0.042821694434597664, "grad_norm": 3.182925829785973, "learning_rate": 9.99572941622477e-06, "loss": 0.5033, "step": 604 }, { "epoch": 0.04289259127968805, "grad_norm": 3.0092051258363868, "learning_rate": 9.995681840335964e-06, "loss": 0.4898, "step": 605 }, { "epoch": 0.042963488124778446, "grad_norm": 5.119830015718799, "learning_rate": 9.995634001021043e-06, "loss": 0.4241, "step": 606 }, { "epoch": 0.04303438496986884, "grad_norm": 3.986292593927326, "learning_rate": 9.995585898282527e-06, "loss": 0.4772, "step": 607 }, { "epoch": 0.043105281814959234, "grad_norm": 4.786831798747608, "learning_rate": 9.995537532122954e-06, "loss": 0.5148, "step": 608 }, { "epoch": 0.04317617866004963, "grad_norm": 4.485187444494352, "learning_rate": 9.995488902544876e-06, "loss": 0.5041, "step": 609 }, { "epoch": 0.04324707550514002, "grad_norm": 3.7182265250675095, "learning_rate": 9.995440009550852e-06, "loss": 0.4528, "step": 610 }, { "epoch": 0.04331797235023042, "grad_norm": 6.251442070697995, "learning_rate": 9.995390853143467e-06, "loss": 0.516, "step": 611 }, { "epoch": 0.04338886919532081, "grad_norm": 2.6558926047607807, "learning_rate": 9.995341433325308e-06, "loss": 0.4993, "step": 612 }, { "epoch": 0.0434597660404112, "grad_norm": 3.285512785101704, "learning_rate": 9.99529175009898e-06, "loss": 0.515, "step": 613 }, { "epoch": 0.04353066288550159, "grad_norm": 2.170945208215833, "learning_rate": 9.995241803467108e-06, "loss": 0.4706, "step": 614 }, { "epoch": 0.04360155973059199, "grad_norm": 4.47430022460066, "learning_rate": 9.995191593432322e-06, "loss": 0.4732, "step": 615 }, { "epoch": 0.04367245657568238, "grad_norm": 3.8859999390248308, "learning_rate": 9.995141119997271e-06, "loss": 0.4455, "step": 616 }, { "epoch": 0.043743353420772776, "grad_norm": 33.122366065058834, "learning_rate": 9.995090383164616e-06, "loss": 0.5642, "step": 617 }, { "epoch": 0.04381425026586317, "grad_norm": 5.445395154626312, "learning_rate": 9.995039382937033e-06, "loss": 0.4854, "step": 618 }, { "epoch": 0.043885147110953565, "grad_norm": 4.1343360867561, "learning_rate": 9.99498811931721e-06, "loss": 0.5216, "step": 619 }, { "epoch": 0.04395604395604396, "grad_norm": 1.1045590947359476, "learning_rate": 9.994936592307853e-06, "loss": 0.7164, "step": 620 }, { "epoch": 0.04402694080113435, "grad_norm": 4.922966386713366, "learning_rate": 9.994884801911674e-06, "loss": 0.4923, "step": 621 }, { "epoch": 0.04409783764622474, "grad_norm": 1.2667896003202932, "learning_rate": 9.994832748131409e-06, "loss": 0.7635, "step": 622 }, { "epoch": 0.044168734491315136, "grad_norm": 3.97557649542504, "learning_rate": 9.994780430969801e-06, "loss": 0.483, "step": 623 }, { "epoch": 0.04423963133640553, "grad_norm": 3.120918332991783, "learning_rate": 9.994727850429607e-06, "loss": 0.4624, "step": 624 }, { "epoch": 0.044310528181495924, "grad_norm": 3.8148563126849964, "learning_rate": 9.994675006513603e-06, "loss": 0.4639, "step": 625 }, { "epoch": 0.04438142502658632, "grad_norm": 4.303464106611614, "learning_rate": 9.994621899224573e-06, "loss": 0.4991, "step": 626 }, { "epoch": 0.04445232187167671, "grad_norm": 3.8971943836437917, "learning_rate": 9.994568528565317e-06, "loss": 0.5063, "step": 627 }, { "epoch": 0.04452321871676711, "grad_norm": 2.650418462837771, "learning_rate": 9.99451489453865e-06, "loss": 0.4313, "step": 628 }, { "epoch": 0.044594115561857495, "grad_norm": 4.38588289490937, "learning_rate": 9.994460997147402e-06, "loss": 0.4841, "step": 629 }, { "epoch": 0.04466501240694789, "grad_norm": 3.5529955030013016, "learning_rate": 9.994406836394412e-06, "loss": 0.5029, "step": 630 }, { "epoch": 0.04473590925203828, "grad_norm": 1.3140794111397127, "learning_rate": 9.994352412282539e-06, "loss": 0.7042, "step": 631 }, { "epoch": 0.04480680609712868, "grad_norm": 6.425519055681589, "learning_rate": 9.99429772481465e-06, "loss": 0.4822, "step": 632 }, { "epoch": 0.04487770294221907, "grad_norm": 3.723196750989054, "learning_rate": 9.994242773993631e-06, "loss": 0.4918, "step": 633 }, { "epoch": 0.044948599787309466, "grad_norm": 3.9176739859226726, "learning_rate": 9.994187559822379e-06, "loss": 0.5232, "step": 634 }, { "epoch": 0.04501949663239986, "grad_norm": 6.680550723248518, "learning_rate": 9.994132082303805e-06, "loss": 0.4918, "step": 635 }, { "epoch": 0.045090393477490255, "grad_norm": 6.640294594440781, "learning_rate": 9.994076341440833e-06, "loss": 0.4657, "step": 636 }, { "epoch": 0.04516129032258064, "grad_norm": 6.5884926611519035, "learning_rate": 9.994020337236404e-06, "loss": 0.4794, "step": 637 }, { "epoch": 0.04523218716767104, "grad_norm": 3.956768896304974, "learning_rate": 9.993964069693472e-06, "loss": 0.4847, "step": 638 }, { "epoch": 0.04530308401276143, "grad_norm": 4.774364534741239, "learning_rate": 9.993907538815002e-06, "loss": 0.4748, "step": 639 }, { "epoch": 0.045373980857851826, "grad_norm": 3.210324076949567, "learning_rate": 9.993850744603976e-06, "loss": 0.4892, "step": 640 }, { "epoch": 0.04544487770294222, "grad_norm": 6.782934105551227, "learning_rate": 9.993793687063389e-06, "loss": 0.4769, "step": 641 }, { "epoch": 0.045515774548032614, "grad_norm": 2.3388613279772024, "learning_rate": 9.993736366196249e-06, "loss": 0.5189, "step": 642 }, { "epoch": 0.04558667139312301, "grad_norm": 4.623193973345266, "learning_rate": 9.993678782005578e-06, "loss": 0.4524, "step": 643 }, { "epoch": 0.0456575682382134, "grad_norm": 3.881698033108591, "learning_rate": 9.993620934494415e-06, "loss": 0.5044, "step": 644 }, { "epoch": 0.04572846508330379, "grad_norm": 4.290852046316961, "learning_rate": 9.993562823665807e-06, "loss": 0.5203, "step": 645 }, { "epoch": 0.045799361928394185, "grad_norm": 3.8021148192782386, "learning_rate": 9.99350444952282e-06, "loss": 0.5078, "step": 646 }, { "epoch": 0.04587025877348458, "grad_norm": 3.055254695619011, "learning_rate": 9.993445812068533e-06, "loss": 0.4701, "step": 647 }, { "epoch": 0.04594115561857497, "grad_norm": 3.717870384494413, "learning_rate": 9.993386911306036e-06, "loss": 0.5105, "step": 648 }, { "epoch": 0.04601205246366537, "grad_norm": 2.9892387085838847, "learning_rate": 9.993327747238438e-06, "loss": 0.4612, "step": 649 }, { "epoch": 0.04608294930875576, "grad_norm": 3.3362258964875333, "learning_rate": 9.993268319868854e-06, "loss": 0.5096, "step": 650 }, { "epoch": 0.046153846153846156, "grad_norm": 3.8857437896992226, "learning_rate": 9.993208629200422e-06, "loss": 0.4753, "step": 651 }, { "epoch": 0.04622474299893655, "grad_norm": 3.79520157959834, "learning_rate": 9.993148675236289e-06, "loss": 0.4588, "step": 652 }, { "epoch": 0.04629563984402694, "grad_norm": 9.396687947833101, "learning_rate": 9.993088457979611e-06, "loss": 0.4479, "step": 653 }, { "epoch": 0.04636653668911733, "grad_norm": 3.0225757207804276, "learning_rate": 9.99302797743357e-06, "loss": 0.453, "step": 654 }, { "epoch": 0.04643743353420773, "grad_norm": 2.6381430173425944, "learning_rate": 9.992967233601353e-06, "loss": 0.5131, "step": 655 }, { "epoch": 0.04650833037929812, "grad_norm": 3.4011361992206237, "learning_rate": 9.992906226486163e-06, "loss": 0.4633, "step": 656 }, { "epoch": 0.046579227224388516, "grad_norm": 6.817647868185, "learning_rate": 9.992844956091216e-06, "loss": 0.4475, "step": 657 }, { "epoch": 0.04665012406947891, "grad_norm": 5.339723659047055, "learning_rate": 9.992783422419744e-06, "loss": 0.4012, "step": 658 }, { "epoch": 0.046721020914569304, "grad_norm": 4.440388927492446, "learning_rate": 9.992721625474992e-06, "loss": 0.455, "step": 659 }, { "epoch": 0.0467919177596597, "grad_norm": 3.663726010220307, "learning_rate": 9.992659565260218e-06, "loss": 0.4786, "step": 660 }, { "epoch": 0.046862814604750086, "grad_norm": 3.8660996238993586, "learning_rate": 9.992597241778694e-06, "loss": 0.4813, "step": 661 }, { "epoch": 0.04693371144984048, "grad_norm": 3.2091064596104673, "learning_rate": 9.992534655033706e-06, "loss": 0.5394, "step": 662 }, { "epoch": 0.047004608294930875, "grad_norm": 4.5711300169453715, "learning_rate": 9.992471805028557e-06, "loss": 0.4625, "step": 663 }, { "epoch": 0.04707550514002127, "grad_norm": 5.251747048913059, "learning_rate": 9.992408691766556e-06, "loss": 0.5267, "step": 664 }, { "epoch": 0.04714640198511166, "grad_norm": 2.3492542346122063, "learning_rate": 9.992345315251037e-06, "loss": 0.4372, "step": 665 }, { "epoch": 0.04721729883020206, "grad_norm": 5.452419746417745, "learning_rate": 9.992281675485339e-06, "loss": 0.5195, "step": 666 }, { "epoch": 0.04728819567529245, "grad_norm": 3.0753694048778724, "learning_rate": 9.992217772472815e-06, "loss": 0.4709, "step": 667 }, { "epoch": 0.047359092520382846, "grad_norm": 3.937764866221986, "learning_rate": 9.992153606216839e-06, "loss": 0.4675, "step": 668 }, { "epoch": 0.047429989365473234, "grad_norm": 5.851642204789372, "learning_rate": 9.992089176720794e-06, "loss": 0.5103, "step": 669 }, { "epoch": 0.04750088621056363, "grad_norm": 1.3377813899321953, "learning_rate": 9.992024483988074e-06, "loss": 0.7126, "step": 670 }, { "epoch": 0.04757178305565402, "grad_norm": 3.166948575054805, "learning_rate": 9.991959528022096e-06, "loss": 0.4589, "step": 671 }, { "epoch": 0.04764267990074442, "grad_norm": 3.2793661204805233, "learning_rate": 9.99189430882628e-06, "loss": 0.4524, "step": 672 }, { "epoch": 0.04771357674583481, "grad_norm": 6.319530971001784, "learning_rate": 9.991828826404066e-06, "loss": 0.5002, "step": 673 }, { "epoch": 0.047784473590925206, "grad_norm": 1.3451973761320177, "learning_rate": 9.99176308075891e-06, "loss": 0.7408, "step": 674 }, { "epoch": 0.0478553704360156, "grad_norm": 3.1109075432091298, "learning_rate": 9.991697071894276e-06, "loss": 0.4858, "step": 675 }, { "epoch": 0.047926267281105994, "grad_norm": 1.2846348574468769, "learning_rate": 9.991630799813644e-06, "loss": 0.6911, "step": 676 }, { "epoch": 0.04799716412619638, "grad_norm": 4.185108211501638, "learning_rate": 9.991564264520511e-06, "loss": 0.4719, "step": 677 }, { "epoch": 0.048068060971286776, "grad_norm": 3.59460370152704, "learning_rate": 9.991497466018386e-06, "loss": 0.4541, "step": 678 }, { "epoch": 0.04813895781637717, "grad_norm": 3.3283226620317037, "learning_rate": 9.991430404310787e-06, "loss": 0.4448, "step": 679 }, { "epoch": 0.048209854661467565, "grad_norm": 4.212600497125198, "learning_rate": 9.991363079401256e-06, "loss": 0.4537, "step": 680 }, { "epoch": 0.04828075150655796, "grad_norm": 3.863757680047542, "learning_rate": 9.991295491293336e-06, "loss": 0.4671, "step": 681 }, { "epoch": 0.04835164835164835, "grad_norm": 3.4565562376513843, "learning_rate": 9.991227639990598e-06, "loss": 0.4993, "step": 682 }, { "epoch": 0.04842254519673875, "grad_norm": 3.669962344034137, "learning_rate": 9.991159525496616e-06, "loss": 0.5007, "step": 683 }, { "epoch": 0.04849344204182914, "grad_norm": 2.5749502244174276, "learning_rate": 9.991091147814985e-06, "loss": 0.4614, "step": 684 }, { "epoch": 0.04856433888691953, "grad_norm": 4.24514482858627, "learning_rate": 9.991022506949307e-06, "loss": 0.4615, "step": 685 }, { "epoch": 0.048635235732009924, "grad_norm": 3.6297689519576317, "learning_rate": 9.990953602903201e-06, "loss": 0.4939, "step": 686 }, { "epoch": 0.04870613257710032, "grad_norm": 3.102120976768821, "learning_rate": 9.990884435680303e-06, "loss": 0.439, "step": 687 }, { "epoch": 0.04877702942219071, "grad_norm": 11.071210194047175, "learning_rate": 9.99081500528426e-06, "loss": 0.4804, "step": 688 }, { "epoch": 0.04884792626728111, "grad_norm": 3.5128540547367204, "learning_rate": 9.99074531171873e-06, "loss": 0.5188, "step": 689 }, { "epoch": 0.0489188231123715, "grad_norm": 6.397977553137573, "learning_rate": 9.990675354987394e-06, "loss": 0.4844, "step": 690 }, { "epoch": 0.048989719957461896, "grad_norm": 4.186389146724633, "learning_rate": 9.990605135093934e-06, "loss": 0.4578, "step": 691 }, { "epoch": 0.04906061680255229, "grad_norm": 2.7184347967187494, "learning_rate": 9.990534652042058e-06, "loss": 0.4754, "step": 692 }, { "epoch": 0.04913151364764268, "grad_norm": 4.091543839638447, "learning_rate": 9.990463905835481e-06, "loss": 0.4522, "step": 693 }, { "epoch": 0.04920241049273307, "grad_norm": 4.64835678001726, "learning_rate": 9.990392896477933e-06, "loss": 0.4481, "step": 694 }, { "epoch": 0.049273307337823466, "grad_norm": 3.326686537053917, "learning_rate": 9.990321623973159e-06, "loss": 0.4693, "step": 695 }, { "epoch": 0.04934420418291386, "grad_norm": 3.6396099450144446, "learning_rate": 9.990250088324916e-06, "loss": 0.4682, "step": 696 }, { "epoch": 0.049415101028004255, "grad_norm": 3.044925188474691, "learning_rate": 9.990178289536976e-06, "loss": 0.4657, "step": 697 }, { "epoch": 0.04948599787309465, "grad_norm": 5.431191062511538, "learning_rate": 9.990106227613127e-06, "loss": 0.4595, "step": 698 }, { "epoch": 0.04955689471818504, "grad_norm": 6.53749483029299, "learning_rate": 9.990033902557168e-06, "loss": 0.4884, "step": 699 }, { "epoch": 0.04962779156327544, "grad_norm": 3.598852315648863, "learning_rate": 9.989961314372913e-06, "loss": 0.3999, "step": 700 }, { "epoch": 0.049698688408365825, "grad_norm": 3.183448890007699, "learning_rate": 9.989888463064187e-06, "loss": 0.5293, "step": 701 }, { "epoch": 0.04976958525345622, "grad_norm": 2.3705047994259107, "learning_rate": 9.989815348634837e-06, "loss": 0.4489, "step": 702 }, { "epoch": 0.049840482098546614, "grad_norm": 4.4471929354388315, "learning_rate": 9.989741971088714e-06, "loss": 0.4719, "step": 703 }, { "epoch": 0.04991137894363701, "grad_norm": 4.816851857492328, "learning_rate": 9.989668330429687e-06, "loss": 0.4317, "step": 704 }, { "epoch": 0.0499822757887274, "grad_norm": 6.6421079910949725, "learning_rate": 9.989594426661641e-06, "loss": 0.4799, "step": 705 }, { "epoch": 0.0500531726338178, "grad_norm": 35.44452908186764, "learning_rate": 9.989520259788474e-06, "loss": 0.5382, "step": 706 }, { "epoch": 0.05012406947890819, "grad_norm": 2.9791030535929885, "learning_rate": 9.989445829814093e-06, "loss": 0.4765, "step": 707 }, { "epoch": 0.050194966323998585, "grad_norm": 4.382773065986338, "learning_rate": 9.989371136742428e-06, "loss": 0.4834, "step": 708 }, { "epoch": 0.05026586316908897, "grad_norm": 3.857369360419858, "learning_rate": 9.989296180577413e-06, "loss": 0.4536, "step": 709 }, { "epoch": 0.05033676001417937, "grad_norm": 4.43961710465899, "learning_rate": 9.989220961323e-06, "loss": 0.455, "step": 710 }, { "epoch": 0.05040765685926976, "grad_norm": 5.343345131744657, "learning_rate": 9.989145478983161e-06, "loss": 0.5013, "step": 711 }, { "epoch": 0.050478553704360156, "grad_norm": 8.518040101152936, "learning_rate": 9.989069733561872e-06, "loss": 0.471, "step": 712 }, { "epoch": 0.05054945054945055, "grad_norm": 6.5283416042529705, "learning_rate": 9.988993725063128e-06, "loss": 0.5015, "step": 713 }, { "epoch": 0.050620347394540945, "grad_norm": 4.134429886560168, "learning_rate": 9.988917453490937e-06, "loss": 0.5055, "step": 714 }, { "epoch": 0.05069124423963134, "grad_norm": 4.782863881390929, "learning_rate": 9.988840918849321e-06, "loss": 0.4987, "step": 715 }, { "epoch": 0.05076214108472173, "grad_norm": 4.271907259918832, "learning_rate": 9.988764121142313e-06, "loss": 0.4829, "step": 716 }, { "epoch": 0.05083303792981212, "grad_norm": 5.026325690499605, "learning_rate": 9.988687060373968e-06, "loss": 0.4657, "step": 717 }, { "epoch": 0.050903934774902515, "grad_norm": 4.068163528750401, "learning_rate": 9.988609736548346e-06, "loss": 0.5299, "step": 718 }, { "epoch": 0.05097483161999291, "grad_norm": 8.945297572573873, "learning_rate": 9.988532149669526e-06, "loss": 0.4959, "step": 719 }, { "epoch": 0.051045728465083304, "grad_norm": 3.0573167605358798, "learning_rate": 9.988454299741596e-06, "loss": 0.4153, "step": 720 }, { "epoch": 0.0511166253101737, "grad_norm": 2.964942812065989, "learning_rate": 9.988376186768664e-06, "loss": 0.466, "step": 721 }, { "epoch": 0.05118752215526409, "grad_norm": 3.1620755438145838, "learning_rate": 9.988297810754847e-06, "loss": 0.491, "step": 722 }, { "epoch": 0.05125841900035449, "grad_norm": 2.9295376166876976, "learning_rate": 9.988219171704282e-06, "loss": 0.4497, "step": 723 }, { "epoch": 0.05132931584544488, "grad_norm": 3.881449141435777, "learning_rate": 9.988140269621112e-06, "loss": 0.5034, "step": 724 }, { "epoch": 0.05140021269053527, "grad_norm": 14.318792892408991, "learning_rate": 9.988061104509497e-06, "loss": 0.4557, "step": 725 }, { "epoch": 0.05147110953562566, "grad_norm": 8.06460915187992, "learning_rate": 9.987981676373613e-06, "loss": 0.49, "step": 726 }, { "epoch": 0.05154200638071606, "grad_norm": 4.921775008211979, "learning_rate": 9.987901985217649e-06, "loss": 0.518, "step": 727 }, { "epoch": 0.05161290322580645, "grad_norm": 5.239870431755818, "learning_rate": 9.987822031045805e-06, "loss": 0.4408, "step": 728 }, { "epoch": 0.051683800070896846, "grad_norm": 4.833587843231638, "learning_rate": 9.9877418138623e-06, "loss": 0.5142, "step": 729 }, { "epoch": 0.05175469691598724, "grad_norm": 2.5191984327435235, "learning_rate": 9.98766133367136e-06, "loss": 0.478, "step": 730 }, { "epoch": 0.051825593761077635, "grad_norm": 6.860472192210849, "learning_rate": 9.987580590477231e-06, "loss": 0.4602, "step": 731 }, { "epoch": 0.05189649060616803, "grad_norm": 4.435523142054012, "learning_rate": 9.987499584284173e-06, "loss": 0.4767, "step": 732 }, { "epoch": 0.051967387451258416, "grad_norm": 3.9619699484626385, "learning_rate": 9.987418315096452e-06, "loss": 0.4527, "step": 733 }, { "epoch": 0.05203828429634881, "grad_norm": 4.637547585659928, "learning_rate": 9.98733678291836e-06, "loss": 0.4815, "step": 734 }, { "epoch": 0.052109181141439205, "grad_norm": 10.694283987526289, "learning_rate": 9.987254987754191e-06, "loss": 0.4795, "step": 735 }, { "epoch": 0.0521800779865296, "grad_norm": 5.686817498967355, "learning_rate": 9.98717292960826e-06, "loss": 0.4893, "step": 736 }, { "epoch": 0.052250974831619994, "grad_norm": 9.696805161618292, "learning_rate": 9.987090608484894e-06, "loss": 0.4873, "step": 737 }, { "epoch": 0.05232187167671039, "grad_norm": 3.4698051858229033, "learning_rate": 9.987008024388433e-06, "loss": 0.5118, "step": 738 }, { "epoch": 0.05239276852180078, "grad_norm": 3.5673668507291594, "learning_rate": 9.986925177323233e-06, "loss": 0.5336, "step": 739 }, { "epoch": 0.05246366536689118, "grad_norm": 6.811313524028677, "learning_rate": 9.986842067293661e-06, "loss": 0.4838, "step": 740 }, { "epoch": 0.052534562211981564, "grad_norm": 4.848897837367656, "learning_rate": 9.986758694304102e-06, "loss": 0.4468, "step": 741 }, { "epoch": 0.05260545905707196, "grad_norm": 1.2867184494609354, "learning_rate": 9.98667505835895e-06, "loss": 0.7311, "step": 742 }, { "epoch": 0.05267635590216235, "grad_norm": 3.18117790145881, "learning_rate": 9.986591159462618e-06, "loss": 0.499, "step": 743 }, { "epoch": 0.05274725274725275, "grad_norm": 4.282735581905572, "learning_rate": 9.986506997619525e-06, "loss": 0.5215, "step": 744 }, { "epoch": 0.05281814959234314, "grad_norm": 4.760882601285083, "learning_rate": 9.986422572834113e-06, "loss": 0.5005, "step": 745 }, { "epoch": 0.052889046437433536, "grad_norm": 23.296586547372662, "learning_rate": 9.986337885110835e-06, "loss": 0.5102, "step": 746 }, { "epoch": 0.05295994328252393, "grad_norm": 5.234568501371865, "learning_rate": 9.986252934454152e-06, "loss": 0.5074, "step": 747 }, { "epoch": 0.053030840127614325, "grad_norm": 5.017793801582525, "learning_rate": 9.986167720868548e-06, "loss": 0.4326, "step": 748 }, { "epoch": 0.05310173697270471, "grad_norm": 5.098198164566221, "learning_rate": 9.986082244358512e-06, "loss": 0.4935, "step": 749 }, { "epoch": 0.053172633817795106, "grad_norm": 4.690936280867312, "learning_rate": 9.985996504928554e-06, "loss": 0.4334, "step": 750 }, { "epoch": 0.0532435306628855, "grad_norm": 3.7418985911607496, "learning_rate": 9.985910502583195e-06, "loss": 0.5181, "step": 751 }, { "epoch": 0.053314427507975895, "grad_norm": 4.425713369028912, "learning_rate": 9.985824237326969e-06, "loss": 0.4708, "step": 752 }, { "epoch": 0.05338532435306629, "grad_norm": 5.883226141013849, "learning_rate": 9.985737709164426e-06, "loss": 0.4253, "step": 753 }, { "epoch": 0.053456221198156684, "grad_norm": 2.4139421640655745, "learning_rate": 9.98565091810013e-06, "loss": 0.4551, "step": 754 }, { "epoch": 0.05352711804324708, "grad_norm": 5.727644503279649, "learning_rate": 9.985563864138654e-06, "loss": 0.4281, "step": 755 }, { "epoch": 0.05359801488833747, "grad_norm": 3.7449548380864184, "learning_rate": 9.98547654728459e-06, "loss": 0.5062, "step": 756 }, { "epoch": 0.05366891173342786, "grad_norm": 2.3761413568226137, "learning_rate": 9.985388967542542e-06, "loss": 0.4599, "step": 757 }, { "epoch": 0.053739808578518254, "grad_norm": 1.1896162064293672, "learning_rate": 9.98530112491713e-06, "loss": 0.7013, "step": 758 }, { "epoch": 0.05381070542360865, "grad_norm": 15.910436280943847, "learning_rate": 9.985213019412983e-06, "loss": 0.4676, "step": 759 }, { "epoch": 0.05388160226869904, "grad_norm": 3.549909156322418, "learning_rate": 9.98512465103475e-06, "loss": 0.4801, "step": 760 }, { "epoch": 0.05395249911378944, "grad_norm": 4.7858706502514785, "learning_rate": 9.985036019787088e-06, "loss": 0.4938, "step": 761 }, { "epoch": 0.05402339595887983, "grad_norm": 4.156401599346069, "learning_rate": 9.984947125674672e-06, "loss": 0.4173, "step": 762 }, { "epoch": 0.054094292803970226, "grad_norm": 4.019447004795387, "learning_rate": 9.98485796870219e-06, "loss": 0.4907, "step": 763 }, { "epoch": 0.05416518964906062, "grad_norm": 4.8950032287560035, "learning_rate": 9.984768548874342e-06, "loss": 0.4733, "step": 764 }, { "epoch": 0.05423608649415101, "grad_norm": 4.019357121381108, "learning_rate": 9.984678866195842e-06, "loss": 0.4118, "step": 765 }, { "epoch": 0.0543069833392414, "grad_norm": 3.806944202921592, "learning_rate": 9.984588920671423e-06, "loss": 0.5027, "step": 766 }, { "epoch": 0.054377880184331796, "grad_norm": 2.8126120385545796, "learning_rate": 9.984498712305825e-06, "loss": 0.4782, "step": 767 }, { "epoch": 0.05444877702942219, "grad_norm": 1.1407823931531547, "learning_rate": 9.984408241103807e-06, "loss": 0.6751, "step": 768 }, { "epoch": 0.054519673874512585, "grad_norm": 4.741130919180903, "learning_rate": 9.984317507070136e-06, "loss": 0.468, "step": 769 }, { "epoch": 0.05459057071960298, "grad_norm": 4.708061771905874, "learning_rate": 9.984226510209597e-06, "loss": 0.4948, "step": 770 }, { "epoch": 0.054661467564693374, "grad_norm": 2.7864045062767366, "learning_rate": 9.984135250526995e-06, "loss": 0.4827, "step": 771 }, { "epoch": 0.05473236440978377, "grad_norm": 5.692955061133997, "learning_rate": 9.984043728027133e-06, "loss": 0.4491, "step": 772 }, { "epoch": 0.054803261254874155, "grad_norm": 1.3746237003771682, "learning_rate": 9.983951942714842e-06, "loss": 0.7369, "step": 773 }, { "epoch": 0.05487415809996455, "grad_norm": 9.336679887760095, "learning_rate": 9.98385989459496e-06, "loss": 0.5159, "step": 774 }, { "epoch": 0.054945054945054944, "grad_norm": 18.219159211591727, "learning_rate": 9.983767583672342e-06, "loss": 0.4888, "step": 775 }, { "epoch": 0.05501595179014534, "grad_norm": 3.5297148297125176, "learning_rate": 9.983675009951857e-06, "loss": 0.4832, "step": 776 }, { "epoch": 0.05508684863523573, "grad_norm": 2.7175346564265217, "learning_rate": 9.983582173438384e-06, "loss": 0.4985, "step": 777 }, { "epoch": 0.05515774548032613, "grad_norm": 4.625489446600758, "learning_rate": 9.983489074136818e-06, "loss": 0.4762, "step": 778 }, { "epoch": 0.05522864232541652, "grad_norm": 4.52259669862275, "learning_rate": 9.98339571205207e-06, "loss": 0.4511, "step": 779 }, { "epoch": 0.055299539170506916, "grad_norm": 4.038516203249569, "learning_rate": 9.983302087189064e-06, "loss": 0.5183, "step": 780 }, { "epoch": 0.0553704360155973, "grad_norm": 8.184394542782968, "learning_rate": 9.983208199552734e-06, "loss": 0.4413, "step": 781 }, { "epoch": 0.0554413328606877, "grad_norm": 6.681710235479214, "learning_rate": 9.983114049148032e-06, "loss": 0.4455, "step": 782 }, { "epoch": 0.05551222970577809, "grad_norm": 3.1721270204055982, "learning_rate": 9.983019635979924e-06, "loss": 0.453, "step": 783 }, { "epoch": 0.055583126550868486, "grad_norm": 2.7766723294245477, "learning_rate": 9.982924960053386e-06, "loss": 0.474, "step": 784 }, { "epoch": 0.05565402339595888, "grad_norm": 6.065080383449522, "learning_rate": 9.982830021373412e-06, "loss": 0.4805, "step": 785 }, { "epoch": 0.055724920241049275, "grad_norm": 5.822095732173486, "learning_rate": 9.982734819945008e-06, "loss": 0.4214, "step": 786 }, { "epoch": 0.05579581708613967, "grad_norm": 2.893932695872874, "learning_rate": 9.982639355773194e-06, "loss": 0.4178, "step": 787 }, { "epoch": 0.055866713931230064, "grad_norm": 6.824292589212943, "learning_rate": 9.982543628863002e-06, "loss": 0.5028, "step": 788 }, { "epoch": 0.05593761077632045, "grad_norm": 3.256412157313564, "learning_rate": 9.982447639219483e-06, "loss": 0.4355, "step": 789 }, { "epoch": 0.056008507621410845, "grad_norm": 3.0542675184306995, "learning_rate": 9.982351386847698e-06, "loss": 0.5116, "step": 790 }, { "epoch": 0.05607940446650124, "grad_norm": 3.425979017555618, "learning_rate": 9.98225487175272e-06, "loss": 0.4281, "step": 791 }, { "epoch": 0.056150301311591634, "grad_norm": 4.359474296057366, "learning_rate": 9.98215809393964e-06, "loss": 0.4164, "step": 792 }, { "epoch": 0.05622119815668203, "grad_norm": 5.5772204001183106, "learning_rate": 9.98206105341356e-06, "loss": 0.5929, "step": 793 }, { "epoch": 0.05629209500177242, "grad_norm": 3.517416290118714, "learning_rate": 9.9819637501796e-06, "loss": 0.5324, "step": 794 }, { "epoch": 0.05636299184686282, "grad_norm": 6.599053534242153, "learning_rate": 9.981866184242887e-06, "loss": 0.4615, "step": 795 }, { "epoch": 0.05643388869195321, "grad_norm": 2.797830111081043, "learning_rate": 9.981768355608569e-06, "loss": 0.5036, "step": 796 }, { "epoch": 0.0565047855370436, "grad_norm": 2.693104273423434, "learning_rate": 9.981670264281802e-06, "loss": 0.4829, "step": 797 }, { "epoch": 0.05657568238213399, "grad_norm": 4.135769807204175, "learning_rate": 9.98157191026776e-06, "loss": 0.4633, "step": 798 }, { "epoch": 0.05664657922722439, "grad_norm": 3.4756732436342674, "learning_rate": 9.98147329357163e-06, "loss": 0.458, "step": 799 }, { "epoch": 0.05671747607231478, "grad_norm": 4.393701644494579, "learning_rate": 9.981374414198611e-06, "loss": 0.4904, "step": 800 }, { "epoch": 0.056788372917405176, "grad_norm": 2.844898324027347, "learning_rate": 9.981275272153916e-06, "loss": 0.4734, "step": 801 }, { "epoch": 0.05685926976249557, "grad_norm": 4.704027488199398, "learning_rate": 9.981175867442774e-06, "loss": 0.44, "step": 802 }, { "epoch": 0.056930166607585965, "grad_norm": 4.272724577018039, "learning_rate": 9.981076200070428e-06, "loss": 0.4601, "step": 803 }, { "epoch": 0.05700106345267636, "grad_norm": 3.3915646479994987, "learning_rate": 9.98097627004213e-06, "loss": 0.479, "step": 804 }, { "epoch": 0.05707196029776675, "grad_norm": 3.6582503388759466, "learning_rate": 9.980876077363152e-06, "loss": 0.5112, "step": 805 }, { "epoch": 0.05714285714285714, "grad_norm": 4.442610173229398, "learning_rate": 9.980775622038776e-06, "loss": 0.4823, "step": 806 }, { "epoch": 0.057213753987947535, "grad_norm": 3.609476765115333, "learning_rate": 9.9806749040743e-06, "loss": 0.4645, "step": 807 }, { "epoch": 0.05728465083303793, "grad_norm": 2.877770878228515, "learning_rate": 9.980573923475038e-06, "loss": 0.438, "step": 808 }, { "epoch": 0.057355547678128324, "grad_norm": 4.254218736205217, "learning_rate": 9.980472680246309e-06, "loss": 0.4787, "step": 809 }, { "epoch": 0.05742644452321872, "grad_norm": 4.01624989501347, "learning_rate": 9.980371174393454e-06, "loss": 0.5311, "step": 810 }, { "epoch": 0.05749734136830911, "grad_norm": 3.6004989290169185, "learning_rate": 9.980269405921826e-06, "loss": 0.434, "step": 811 }, { "epoch": 0.05756823821339951, "grad_norm": 11.573631461063806, "learning_rate": 9.980167374836792e-06, "loss": 0.4331, "step": 812 }, { "epoch": 0.057639135058489895, "grad_norm": 3.5066402965343046, "learning_rate": 9.980065081143729e-06, "loss": 0.4372, "step": 813 }, { "epoch": 0.05771003190358029, "grad_norm": 2.8540044673803697, "learning_rate": 9.979962524848033e-06, "loss": 0.498, "step": 814 }, { "epoch": 0.05778092874867068, "grad_norm": 5.948838861854879, "learning_rate": 9.979859705955114e-06, "loss": 0.4878, "step": 815 }, { "epoch": 0.05785182559376108, "grad_norm": 6.474995313658146, "learning_rate": 9.979756624470392e-06, "loss": 0.4556, "step": 816 }, { "epoch": 0.05792272243885147, "grad_norm": 4.55636812461726, "learning_rate": 9.9796532803993e-06, "loss": 0.445, "step": 817 }, { "epoch": 0.057993619283941866, "grad_norm": 16.950121994884118, "learning_rate": 9.979549673747291e-06, "loss": 0.4672, "step": 818 }, { "epoch": 0.05806451612903226, "grad_norm": 2.688056596433423, "learning_rate": 9.979445804519828e-06, "loss": 0.4664, "step": 819 }, { "epoch": 0.058135412974122655, "grad_norm": 2.871521987840477, "learning_rate": 9.979341672722384e-06, "loss": 0.4478, "step": 820 }, { "epoch": 0.05820630981921304, "grad_norm": 3.2288179217589774, "learning_rate": 9.979237278360457e-06, "loss": 0.4678, "step": 821 }, { "epoch": 0.05827720666430344, "grad_norm": 2.788512425900849, "learning_rate": 9.979132621439545e-06, "loss": 0.4481, "step": 822 }, { "epoch": 0.05834810350939383, "grad_norm": 3.5901254931455195, "learning_rate": 9.979027701965171e-06, "loss": 0.5351, "step": 823 }, { "epoch": 0.058419000354484225, "grad_norm": 2.536650175088936, "learning_rate": 9.978922519942864e-06, "loss": 0.4752, "step": 824 }, { "epoch": 0.05848989719957462, "grad_norm": 7.628293039186481, "learning_rate": 9.978817075378175e-06, "loss": 0.4251, "step": 825 }, { "epoch": 0.058560794044665014, "grad_norm": 4.798725355550995, "learning_rate": 9.97871136827666e-06, "loss": 0.4779, "step": 826 }, { "epoch": 0.05863169088975541, "grad_norm": 3.0689241864257104, "learning_rate": 9.978605398643895e-06, "loss": 0.4678, "step": 827 }, { "epoch": 0.0587025877348458, "grad_norm": 3.057833423625542, "learning_rate": 9.978499166485466e-06, "loss": 0.4687, "step": 828 }, { "epoch": 0.05877348457993619, "grad_norm": 3.2428765244383477, "learning_rate": 9.978392671806976e-06, "loss": 0.4631, "step": 829 }, { "epoch": 0.058844381425026585, "grad_norm": 4.3524290722568635, "learning_rate": 9.978285914614041e-06, "loss": 0.4423, "step": 830 }, { "epoch": 0.05891527827011698, "grad_norm": 4.799110453489222, "learning_rate": 9.978178894912291e-06, "loss": 0.4701, "step": 831 }, { "epoch": 0.05898617511520737, "grad_norm": 2.5455996894456336, "learning_rate": 9.978071612707368e-06, "loss": 0.4396, "step": 832 }, { "epoch": 0.05905707196029777, "grad_norm": 3.2080238795969724, "learning_rate": 9.97796406800493e-06, "loss": 0.4646, "step": 833 }, { "epoch": 0.05912796880538816, "grad_norm": 10.28406436706162, "learning_rate": 9.977856260810646e-06, "loss": 0.4243, "step": 834 }, { "epoch": 0.059198865650478556, "grad_norm": 4.829328995722104, "learning_rate": 9.9777481911302e-06, "loss": 0.4729, "step": 835 }, { "epoch": 0.05926976249556895, "grad_norm": 7.549943297919198, "learning_rate": 9.977639858969297e-06, "loss": 0.5001, "step": 836 }, { "epoch": 0.05934065934065934, "grad_norm": 1.2419953212761161, "learning_rate": 9.977531264333642e-06, "loss": 0.7878, "step": 837 }, { "epoch": 0.05941155618574973, "grad_norm": 5.193883924127648, "learning_rate": 9.977422407228966e-06, "loss": 0.5537, "step": 838 }, { "epoch": 0.05948245303084013, "grad_norm": 3.30116987982149, "learning_rate": 9.977313287661004e-06, "loss": 0.5356, "step": 839 }, { "epoch": 0.05955334987593052, "grad_norm": 5.566004255166636, "learning_rate": 9.977203905635516e-06, "loss": 0.4851, "step": 840 }, { "epoch": 0.059624246721020915, "grad_norm": 3.616292683387844, "learning_rate": 9.977094261158267e-06, "loss": 0.4569, "step": 841 }, { "epoch": 0.05969514356611131, "grad_norm": 5.028254892969749, "learning_rate": 9.976984354235039e-06, "loss": 0.4973, "step": 842 }, { "epoch": 0.059766040411201704, "grad_norm": 5.686443852323213, "learning_rate": 9.976874184871625e-06, "loss": 0.4892, "step": 843 }, { "epoch": 0.0598369372562921, "grad_norm": 1.8990361670972409, "learning_rate": 9.976763753073838e-06, "loss": 0.4563, "step": 844 }, { "epoch": 0.059907834101382486, "grad_norm": 2.8456626669190994, "learning_rate": 9.9766530588475e-06, "loss": 0.5132, "step": 845 }, { "epoch": 0.05997873094647288, "grad_norm": 3.8748056035340226, "learning_rate": 9.976542102198449e-06, "loss": 0.4818, "step": 846 }, { "epoch": 0.060049627791563275, "grad_norm": 8.560045331235974, "learning_rate": 9.976430883132533e-06, "loss": 0.4799, "step": 847 }, { "epoch": 0.06012052463665367, "grad_norm": 3.110378224868225, "learning_rate": 9.976319401655619e-06, "loss": 0.4458, "step": 848 }, { "epoch": 0.06019142148174406, "grad_norm": 4.556861600524946, "learning_rate": 9.976207657773583e-06, "loss": 0.493, "step": 849 }, { "epoch": 0.06026231832683446, "grad_norm": 3.1813978270289036, "learning_rate": 9.97609565149232e-06, "loss": 0.4415, "step": 850 }, { "epoch": 0.06033321517192485, "grad_norm": 8.40041813000998, "learning_rate": 9.975983382817736e-06, "loss": 0.4975, "step": 851 }, { "epoch": 0.060404112017015246, "grad_norm": 3.4937663513449553, "learning_rate": 9.975870851755749e-06, "loss": 0.4628, "step": 852 }, { "epoch": 0.060475008862105634, "grad_norm": 8.136928168107492, "learning_rate": 9.975758058312295e-06, "loss": 0.4764, "step": 853 }, { "epoch": 0.06054590570719603, "grad_norm": 9.255047339089135, "learning_rate": 9.97564500249332e-06, "loss": 0.5065, "step": 854 }, { "epoch": 0.06061680255228642, "grad_norm": 4.661671388027247, "learning_rate": 9.975531684304787e-06, "loss": 0.4799, "step": 855 }, { "epoch": 0.06068769939737682, "grad_norm": 2.9711632209687466, "learning_rate": 9.97541810375267e-06, "loss": 0.438, "step": 856 }, { "epoch": 0.06075859624246721, "grad_norm": 2.9771266135817194, "learning_rate": 9.97530426084296e-06, "loss": 0.4445, "step": 857 }, { "epoch": 0.060829493087557605, "grad_norm": 2.943179695432369, "learning_rate": 9.975190155581658e-06, "loss": 0.4879, "step": 858 }, { "epoch": 0.060900389932648, "grad_norm": 2.0872994038889714, "learning_rate": 9.975075787974782e-06, "loss": 0.4282, "step": 859 }, { "epoch": 0.060971286777738394, "grad_norm": 3.7004614110988423, "learning_rate": 9.97496115802836e-06, "loss": 0.4579, "step": 860 }, { "epoch": 0.06104218362282878, "grad_norm": 4.377609051431559, "learning_rate": 9.974846265748443e-06, "loss": 0.4918, "step": 861 }, { "epoch": 0.061113080467919176, "grad_norm": 3.369663821293333, "learning_rate": 9.974731111141083e-06, "loss": 0.4744, "step": 862 }, { "epoch": 0.06118397731300957, "grad_norm": 6.161304218331212, "learning_rate": 9.974615694212354e-06, "loss": 0.4557, "step": 863 }, { "epoch": 0.061254874158099964, "grad_norm": 2.438037918728881, "learning_rate": 9.974500014968344e-06, "loss": 0.4359, "step": 864 }, { "epoch": 0.06132577100319036, "grad_norm": 4.341611641216351, "learning_rate": 9.974384073415151e-06, "loss": 0.4352, "step": 865 }, { "epoch": 0.06139666784828075, "grad_norm": 5.686781891184208, "learning_rate": 9.974267869558887e-06, "loss": 0.4103, "step": 866 }, { "epoch": 0.06146756469337115, "grad_norm": 3.8574484752141185, "learning_rate": 9.974151403405685e-06, "loss": 0.4678, "step": 867 }, { "epoch": 0.06153846153846154, "grad_norm": 7.922573418335058, "learning_rate": 9.97403467496168e-06, "loss": 0.4563, "step": 868 }, { "epoch": 0.06160935838355193, "grad_norm": 5.31307733554175, "learning_rate": 9.973917684233033e-06, "loss": 0.4739, "step": 869 }, { "epoch": 0.061680255228642324, "grad_norm": 4.774781876877581, "learning_rate": 9.973800431225906e-06, "loss": 0.4509, "step": 870 }, { "epoch": 0.06175115207373272, "grad_norm": 3.6070232778873836, "learning_rate": 9.973682915946489e-06, "loss": 0.529, "step": 871 }, { "epoch": 0.06182204891882311, "grad_norm": 8.274867181107005, "learning_rate": 9.973565138400975e-06, "loss": 0.4854, "step": 872 }, { "epoch": 0.06189294576391351, "grad_norm": 1.1818146924033242, "learning_rate": 9.973447098595574e-06, "loss": 0.6881, "step": 873 }, { "epoch": 0.0619638426090039, "grad_norm": 4.719539476784723, "learning_rate": 9.973328796536512e-06, "loss": 0.4962, "step": 874 }, { "epoch": 0.062034739454094295, "grad_norm": 3.407597692926925, "learning_rate": 9.973210232230029e-06, "loss": 0.5116, "step": 875 }, { "epoch": 0.06210563629918469, "grad_norm": 3.4935348848695273, "learning_rate": 9.973091405682371e-06, "loss": 0.4337, "step": 876 }, { "epoch": 0.06217653314427508, "grad_norm": 2.645391831869339, "learning_rate": 9.972972316899809e-06, "loss": 0.432, "step": 877 }, { "epoch": 0.06224742998936547, "grad_norm": 3.691861623667809, "learning_rate": 9.972852965888622e-06, "loss": 0.4632, "step": 878 }, { "epoch": 0.062318326834455866, "grad_norm": 10.208387572546203, "learning_rate": 9.9727333526551e-06, "loss": 0.5269, "step": 879 }, { "epoch": 0.06238922367954626, "grad_norm": 4.113386180500243, "learning_rate": 9.972613477205554e-06, "loss": 0.4516, "step": 880 }, { "epoch": 0.062460120524636654, "grad_norm": 9.646916084893347, "learning_rate": 9.972493339546305e-06, "loss": 0.4925, "step": 881 }, { "epoch": 0.06253101736972705, "grad_norm": 5.602870401715015, "learning_rate": 9.972372939683686e-06, "loss": 0.4605, "step": 882 }, { "epoch": 0.06260191421481744, "grad_norm": 3.010703979315049, "learning_rate": 9.972252277624045e-06, "loss": 0.4577, "step": 883 }, { "epoch": 0.06267281105990784, "grad_norm": 5.488571983617461, "learning_rate": 9.972131353373749e-06, "loss": 0.4919, "step": 884 }, { "epoch": 0.06274370790499822, "grad_norm": 3.625039195426439, "learning_rate": 9.972010166939172e-06, "loss": 0.4821, "step": 885 }, { "epoch": 0.06281460475008863, "grad_norm": 4.727362896626685, "learning_rate": 9.971888718326702e-06, "loss": 0.5317, "step": 886 }, { "epoch": 0.06288550159517901, "grad_norm": 4.241725115960502, "learning_rate": 9.971767007542745e-06, "loss": 0.4656, "step": 887 }, { "epoch": 0.06295639844026941, "grad_norm": 4.176603355921239, "learning_rate": 9.97164503459372e-06, "loss": 0.5377, "step": 888 }, { "epoch": 0.0630272952853598, "grad_norm": 3.4554090957022816, "learning_rate": 9.971522799486057e-06, "loss": 0.4923, "step": 889 }, { "epoch": 0.06309819213045019, "grad_norm": 5.533144209903275, "learning_rate": 9.971400302226201e-06, "loss": 0.5203, "step": 890 }, { "epoch": 0.06316908897554059, "grad_norm": 5.4050218694522, "learning_rate": 9.971277542820615e-06, "loss": 0.5478, "step": 891 }, { "epoch": 0.06323998582063098, "grad_norm": 3.1174332380509675, "learning_rate": 9.971154521275768e-06, "loss": 0.4644, "step": 892 }, { "epoch": 0.06331088266572138, "grad_norm": 3.2768937638399067, "learning_rate": 9.97103123759815e-06, "loss": 0.5051, "step": 893 }, { "epoch": 0.06338177951081177, "grad_norm": 1.9480805586465895, "learning_rate": 9.97090769179426e-06, "loss": 0.4756, "step": 894 }, { "epoch": 0.06345267635590217, "grad_norm": 3.218097220300877, "learning_rate": 9.970783883870613e-06, "loss": 0.4487, "step": 895 }, { "epoch": 0.06352357320099256, "grad_norm": 4.002555143490571, "learning_rate": 9.970659813833739e-06, "loss": 0.4882, "step": 896 }, { "epoch": 0.06359447004608294, "grad_norm": 4.646847331299745, "learning_rate": 9.970535481690178e-06, "loss": 0.4743, "step": 897 }, { "epoch": 0.06366536689117334, "grad_norm": 3.107015450193722, "learning_rate": 9.970410887446486e-06, "loss": 0.5132, "step": 898 }, { "epoch": 0.06373626373626373, "grad_norm": 8.132803404078983, "learning_rate": 9.970286031109237e-06, "loss": 0.4372, "step": 899 }, { "epoch": 0.06380716058135413, "grad_norm": 3.092408782829151, "learning_rate": 9.97016091268501e-06, "loss": 0.4838, "step": 900 }, { "epoch": 0.06387805742644452, "grad_norm": 79.5599537410857, "learning_rate": 9.970035532180404e-06, "loss": 0.5038, "step": 901 }, { "epoch": 0.06394895427153492, "grad_norm": 2.506697232270154, "learning_rate": 9.969909889602032e-06, "loss": 0.5413, "step": 902 }, { "epoch": 0.06401985111662531, "grad_norm": 3.6842384903003254, "learning_rate": 9.969783984956519e-06, "loss": 0.4607, "step": 903 }, { "epoch": 0.06409074796171571, "grad_norm": 3.326816573659, "learning_rate": 9.969657818250503e-06, "loss": 0.4525, "step": 904 }, { "epoch": 0.0641616448068061, "grad_norm": 3.0419918373375725, "learning_rate": 9.969531389490635e-06, "loss": 0.5169, "step": 905 }, { "epoch": 0.06423254165189649, "grad_norm": 4.43158013591064, "learning_rate": 9.969404698683586e-06, "loss": 0.5308, "step": 906 }, { "epoch": 0.06430343849698689, "grad_norm": 4.309552728841112, "learning_rate": 9.969277745836032e-06, "loss": 0.496, "step": 907 }, { "epoch": 0.06437433534207727, "grad_norm": 3.327125538654735, "learning_rate": 9.969150530954672e-06, "loss": 0.4908, "step": 908 }, { "epoch": 0.06444523218716768, "grad_norm": 5.724329992621171, "learning_rate": 9.969023054046211e-06, "loss": 0.4597, "step": 909 }, { "epoch": 0.06451612903225806, "grad_norm": 2.3270039853628632, "learning_rate": 9.968895315117371e-06, "loss": 0.4324, "step": 910 }, { "epoch": 0.06458702587734846, "grad_norm": 6.469764323912249, "learning_rate": 9.96876731417489e-06, "loss": 0.4514, "step": 911 }, { "epoch": 0.06465792272243885, "grad_norm": 2.4382710282678572, "learning_rate": 9.968639051225513e-06, "loss": 0.4964, "step": 912 }, { "epoch": 0.06472881956752924, "grad_norm": 4.055918734877608, "learning_rate": 9.968510526276009e-06, "loss": 0.4378, "step": 913 }, { "epoch": 0.06479971641261964, "grad_norm": 2.5781993726620684, "learning_rate": 9.968381739333153e-06, "loss": 0.5035, "step": 914 }, { "epoch": 0.06487061325771003, "grad_norm": 3.872230566335627, "learning_rate": 9.968252690403734e-06, "loss": 0.4815, "step": 915 }, { "epoch": 0.06494151010280043, "grad_norm": 2.612946676968358, "learning_rate": 9.968123379494559e-06, "loss": 0.4291, "step": 916 }, { "epoch": 0.06501240694789082, "grad_norm": 3.3793891917593513, "learning_rate": 9.967993806612446e-06, "loss": 0.4864, "step": 917 }, { "epoch": 0.06508330379298122, "grad_norm": 10.726520763326372, "learning_rate": 9.96786397176423e-06, "loss": 0.4457, "step": 918 }, { "epoch": 0.0651542006380716, "grad_norm": 2.4563718296472996, "learning_rate": 9.96773387495675e-06, "loss": 0.4094, "step": 919 }, { "epoch": 0.065225097483162, "grad_norm": 5.3430877679754945, "learning_rate": 9.967603516196875e-06, "loss": 0.5047, "step": 920 }, { "epoch": 0.0652959943282524, "grad_norm": 3.3829583224203397, "learning_rate": 9.967472895491474e-06, "loss": 0.4937, "step": 921 }, { "epoch": 0.06536689117334278, "grad_norm": 4.083642358361018, "learning_rate": 9.967342012847437e-06, "loss": 0.4744, "step": 922 }, { "epoch": 0.06543778801843318, "grad_norm": 2.4590081342082892, "learning_rate": 9.967210868271662e-06, "loss": 0.4364, "step": 923 }, { "epoch": 0.06550868486352357, "grad_norm": 3.006072331545936, "learning_rate": 9.967079461771069e-06, "loss": 0.5211, "step": 924 }, { "epoch": 0.06557958170861397, "grad_norm": 3.431630334822058, "learning_rate": 9.966947793352585e-06, "loss": 0.4226, "step": 925 }, { "epoch": 0.06565047855370436, "grad_norm": 4.360258016531942, "learning_rate": 9.966815863023149e-06, "loss": 0.4803, "step": 926 }, { "epoch": 0.06572137539879476, "grad_norm": 3.804002626254679, "learning_rate": 9.966683670789724e-06, "loss": 0.4693, "step": 927 }, { "epoch": 0.06579227224388515, "grad_norm": 4.145795364784604, "learning_rate": 9.966551216659278e-06, "loss": 0.4552, "step": 928 }, { "epoch": 0.06586316908897553, "grad_norm": 5.661663889047222, "learning_rate": 9.966418500638796e-06, "loss": 0.4961, "step": 929 }, { "epoch": 0.06593406593406594, "grad_norm": 4.138936874331337, "learning_rate": 9.966285522735276e-06, "loss": 0.395, "step": 930 }, { "epoch": 0.06600496277915632, "grad_norm": 2.271381284841836, "learning_rate": 9.966152282955728e-06, "loss": 0.4996, "step": 931 }, { "epoch": 0.06607585962424672, "grad_norm": 4.28886702113293, "learning_rate": 9.96601878130718e-06, "loss": 0.5015, "step": 932 }, { "epoch": 0.06614675646933711, "grad_norm": 5.052872868068134, "learning_rate": 9.965885017796673e-06, "loss": 0.4034, "step": 933 }, { "epoch": 0.06621765331442751, "grad_norm": 3.0796962209774956, "learning_rate": 9.965750992431257e-06, "loss": 0.4312, "step": 934 }, { "epoch": 0.0662885501595179, "grad_norm": 4.354426671532057, "learning_rate": 9.965616705218002e-06, "loss": 0.5072, "step": 935 }, { "epoch": 0.0663594470046083, "grad_norm": 4.16525901608509, "learning_rate": 9.965482156163989e-06, "loss": 0.4477, "step": 936 }, { "epoch": 0.06643034384969869, "grad_norm": 3.0562260751583374, "learning_rate": 9.96534734527631e-06, "loss": 0.4525, "step": 937 }, { "epoch": 0.06650124069478908, "grad_norm": 3.9560334453176544, "learning_rate": 9.965212272562075e-06, "loss": 0.4671, "step": 938 }, { "epoch": 0.06657213753987948, "grad_norm": 3.8685028142552285, "learning_rate": 9.965076938028409e-06, "loss": 0.4648, "step": 939 }, { "epoch": 0.06664303438496987, "grad_norm": 2.223447623010919, "learning_rate": 9.964941341682446e-06, "loss": 0.4441, "step": 940 }, { "epoch": 0.06671393123006027, "grad_norm": 4.032502333110286, "learning_rate": 9.964805483531337e-06, "loss": 0.4422, "step": 941 }, { "epoch": 0.06678482807515065, "grad_norm": 3.985818862128329, "learning_rate": 9.964669363582243e-06, "loss": 0.5524, "step": 942 }, { "epoch": 0.06685572492024106, "grad_norm": 2.479519464323413, "learning_rate": 9.964532981842346e-06, "loss": 0.4577, "step": 943 }, { "epoch": 0.06692662176533144, "grad_norm": 2.669379234270663, "learning_rate": 9.964396338318835e-06, "loss": 0.4047, "step": 944 }, { "epoch": 0.06699751861042183, "grad_norm": 2.2527888006753183, "learning_rate": 9.964259433018916e-06, "loss": 0.4894, "step": 945 }, { "epoch": 0.06706841545551223, "grad_norm": 3.6979945361640056, "learning_rate": 9.96412226594981e-06, "loss": 0.4476, "step": 946 }, { "epoch": 0.06713931230060262, "grad_norm": 4.139895681346387, "learning_rate": 9.963984837118746e-06, "loss": 0.4783, "step": 947 }, { "epoch": 0.06721020914569302, "grad_norm": 4.934568537966256, "learning_rate": 9.963847146532973e-06, "loss": 0.4609, "step": 948 }, { "epoch": 0.06728110599078341, "grad_norm": 3.849353316003477, "learning_rate": 9.96370919419975e-06, "loss": 0.4904, "step": 949 }, { "epoch": 0.06735200283587381, "grad_norm": 5.829485947751723, "learning_rate": 9.963570980126354e-06, "loss": 0.5104, "step": 950 }, { "epoch": 0.0674228996809642, "grad_norm": 2.912301820904526, "learning_rate": 9.963432504320072e-06, "loss": 0.5203, "step": 951 }, { "epoch": 0.0674937965260546, "grad_norm": 3.2921138471443716, "learning_rate": 9.963293766788204e-06, "loss": 0.4826, "step": 952 }, { "epoch": 0.06756469337114498, "grad_norm": 3.0154614137592852, "learning_rate": 9.963154767538069e-06, "loss": 0.4405, "step": 953 }, { "epoch": 0.06763559021623537, "grad_norm": 2.2888062221906464, "learning_rate": 9.963015506576994e-06, "loss": 0.4684, "step": 954 }, { "epoch": 0.06770648706132577, "grad_norm": 4.065732784076834, "learning_rate": 9.962875983912324e-06, "loss": 0.5337, "step": 955 }, { "epoch": 0.06777738390641616, "grad_norm": 4.931436793077363, "learning_rate": 9.962736199551415e-06, "loss": 0.3931, "step": 956 }, { "epoch": 0.06784828075150656, "grad_norm": 4.0247697844556125, "learning_rate": 9.962596153501638e-06, "loss": 0.4898, "step": 957 }, { "epoch": 0.06791917759659695, "grad_norm": 5.858015079058961, "learning_rate": 9.962455845770379e-06, "loss": 0.5259, "step": 958 }, { "epoch": 0.06799007444168735, "grad_norm": 4.829257457024723, "learning_rate": 9.962315276365035e-06, "loss": 0.4602, "step": 959 }, { "epoch": 0.06806097128677774, "grad_norm": 2.103012397578334, "learning_rate": 9.962174445293018e-06, "loss": 0.4363, "step": 960 }, { "epoch": 0.06813186813186813, "grad_norm": 3.6953914975997164, "learning_rate": 9.962033352561756e-06, "loss": 0.5136, "step": 961 }, { "epoch": 0.06820276497695853, "grad_norm": 4.813085431791682, "learning_rate": 9.961891998178688e-06, "loss": 0.4841, "step": 962 }, { "epoch": 0.06827366182204891, "grad_norm": 2.155980036682444, "learning_rate": 9.961750382151267e-06, "loss": 0.4628, "step": 963 }, { "epoch": 0.06834455866713932, "grad_norm": 2.831653877948743, "learning_rate": 9.961608504486962e-06, "loss": 0.5068, "step": 964 }, { "epoch": 0.0684154555122297, "grad_norm": 0.9649627332531963, "learning_rate": 9.961466365193254e-06, "loss": 0.7259, "step": 965 }, { "epoch": 0.0684863523573201, "grad_norm": 5.990250270197903, "learning_rate": 9.961323964277637e-06, "loss": 0.4952, "step": 966 }, { "epoch": 0.06855724920241049, "grad_norm": 4.266873144026386, "learning_rate": 9.96118130174762e-06, "loss": 0.4184, "step": 967 }, { "epoch": 0.0686281460475009, "grad_norm": 5.957995931398697, "learning_rate": 9.961038377610727e-06, "loss": 0.4728, "step": 968 }, { "epoch": 0.06869904289259128, "grad_norm": 2.332685814873957, "learning_rate": 9.960895191874494e-06, "loss": 0.4729, "step": 969 }, { "epoch": 0.06876993973768167, "grad_norm": 3.230589508786408, "learning_rate": 9.96075174454647e-06, "loss": 0.4138, "step": 970 }, { "epoch": 0.06884083658277207, "grad_norm": 2.6317130949994407, "learning_rate": 9.96060803563422e-06, "loss": 0.488, "step": 971 }, { "epoch": 0.06891173342786246, "grad_norm": 3.501291338819332, "learning_rate": 9.960464065145323e-06, "loss": 0.5278, "step": 972 }, { "epoch": 0.06898263027295286, "grad_norm": 12.423529950941592, "learning_rate": 9.96031983308737e-06, "loss": 0.4523, "step": 973 }, { "epoch": 0.06905352711804325, "grad_norm": 5.2537813434380345, "learning_rate": 9.960175339467966e-06, "loss": 0.4604, "step": 974 }, { "epoch": 0.06912442396313365, "grad_norm": 3.2485413613469523, "learning_rate": 9.96003058429473e-06, "loss": 0.4434, "step": 975 }, { "epoch": 0.06919532080822403, "grad_norm": 2.5966725054289386, "learning_rate": 9.959885567575295e-06, "loss": 0.4829, "step": 976 }, { "epoch": 0.06926621765331442, "grad_norm": 4.6100249284464425, "learning_rate": 9.959740289317308e-06, "loss": 0.4652, "step": 977 }, { "epoch": 0.06933711449840482, "grad_norm": 3.954399278072551, "learning_rate": 9.959594749528432e-06, "loss": 0.5215, "step": 978 }, { "epoch": 0.06940801134349521, "grad_norm": 1.7937216061237358, "learning_rate": 9.959448948216338e-06, "loss": 0.4438, "step": 979 }, { "epoch": 0.06947890818858561, "grad_norm": 2.8671241074820193, "learning_rate": 9.959302885388715e-06, "loss": 0.4051, "step": 980 }, { "epoch": 0.069549805033676, "grad_norm": 2.2085650684783635, "learning_rate": 9.959156561053266e-06, "loss": 0.4394, "step": 981 }, { "epoch": 0.0696207018787664, "grad_norm": 3.992461390414882, "learning_rate": 9.959009975217708e-06, "loss": 0.5153, "step": 982 }, { "epoch": 0.06969159872385679, "grad_norm": 2.321420102069754, "learning_rate": 9.958863127889767e-06, "loss": 0.4704, "step": 983 }, { "epoch": 0.06976249556894719, "grad_norm": 2.75437292686624, "learning_rate": 9.95871601907719e-06, "loss": 0.4664, "step": 984 }, { "epoch": 0.06983339241403758, "grad_norm": 1.7485746575723478, "learning_rate": 9.958568648787731e-06, "loss": 0.4773, "step": 985 }, { "epoch": 0.06990428925912796, "grad_norm": 5.789667399870282, "learning_rate": 9.958421017029162e-06, "loss": 0.4423, "step": 986 }, { "epoch": 0.06997518610421836, "grad_norm": 4.704657097966053, "learning_rate": 9.95827312380927e-06, "loss": 0.5324, "step": 987 }, { "epoch": 0.07004608294930875, "grad_norm": 2.5395974161183346, "learning_rate": 9.958124969135851e-06, "loss": 0.478, "step": 988 }, { "epoch": 0.07011697979439915, "grad_norm": 2.667383432204658, "learning_rate": 9.957976553016716e-06, "loss": 0.4973, "step": 989 }, { "epoch": 0.07018787663948954, "grad_norm": 2.108013007153526, "learning_rate": 9.957827875459695e-06, "loss": 0.4662, "step": 990 }, { "epoch": 0.07025877348457994, "grad_norm": 7.375428831986992, "learning_rate": 9.957678936472628e-06, "loss": 0.4614, "step": 991 }, { "epoch": 0.07032967032967033, "grad_norm": 4.36907715986678, "learning_rate": 9.957529736063364e-06, "loss": 0.4541, "step": 992 }, { "epoch": 0.07040056717476072, "grad_norm": 2.699757861649825, "learning_rate": 9.957380274239774e-06, "loss": 0.4937, "step": 993 }, { "epoch": 0.07047146401985112, "grad_norm": 3.493273540606088, "learning_rate": 9.957230551009737e-06, "loss": 0.4378, "step": 994 }, { "epoch": 0.0705423608649415, "grad_norm": 2.9722435383505283, "learning_rate": 9.95708056638115e-06, "loss": 0.5154, "step": 995 }, { "epoch": 0.07061325771003191, "grad_norm": 3.3943853185063677, "learning_rate": 9.956930320361921e-06, "loss": 0.4364, "step": 996 }, { "epoch": 0.0706841545551223, "grad_norm": 2.0315343182377386, "learning_rate": 9.956779812959975e-06, "loss": 0.4123, "step": 997 }, { "epoch": 0.0707550514002127, "grad_norm": 8.355207797765035, "learning_rate": 9.956629044183244e-06, "loss": 0.4632, "step": 998 }, { "epoch": 0.07082594824530308, "grad_norm": 4.426560065043996, "learning_rate": 9.956478014039681e-06, "loss": 0.4349, "step": 999 }, { "epoch": 0.07089684509039348, "grad_norm": 2.3714351746260145, "learning_rate": 9.956326722537248e-06, "loss": 0.4514, "step": 1000 }, { "epoch": 0.07096774193548387, "grad_norm": 3.214325492309238, "learning_rate": 9.956175169683926e-06, "loss": 0.4165, "step": 1001 }, { "epoch": 0.07103863878057426, "grad_norm": 5.376555696027417, "learning_rate": 9.956023355487702e-06, "loss": 0.4208, "step": 1002 }, { "epoch": 0.07110953562566466, "grad_norm": 2.922365922391431, "learning_rate": 9.955871279956586e-06, "loss": 0.488, "step": 1003 }, { "epoch": 0.07118043247075505, "grad_norm": 2.5458683945816625, "learning_rate": 9.955718943098594e-06, "loss": 0.457, "step": 1004 }, { "epoch": 0.07125132931584545, "grad_norm": 8.787585884849461, "learning_rate": 9.955566344921759e-06, "loss": 0.4612, "step": 1005 }, { "epoch": 0.07132222616093584, "grad_norm": 2.195432585023841, "learning_rate": 9.955413485434128e-06, "loss": 0.4698, "step": 1006 }, { "epoch": 0.07139312300602624, "grad_norm": 3.3709698025178776, "learning_rate": 9.955260364643764e-06, "loss": 0.4937, "step": 1007 }, { "epoch": 0.07146401985111663, "grad_norm": 1.9331838683377047, "learning_rate": 9.955106982558736e-06, "loss": 0.4648, "step": 1008 }, { "epoch": 0.07153491669620701, "grad_norm": 5.014020757629102, "learning_rate": 9.954953339187136e-06, "loss": 0.5509, "step": 1009 }, { "epoch": 0.07160581354129741, "grad_norm": 2.0059973784332263, "learning_rate": 9.954799434537064e-06, "loss": 0.4435, "step": 1010 }, { "epoch": 0.0716767103863878, "grad_norm": 2.1962986176518706, "learning_rate": 9.954645268616636e-06, "loss": 0.4831, "step": 1011 }, { "epoch": 0.0717476072314782, "grad_norm": 3.6896625173562048, "learning_rate": 9.954490841433982e-06, "loss": 0.4684, "step": 1012 }, { "epoch": 0.07181850407656859, "grad_norm": 2.608373601034628, "learning_rate": 9.954336152997244e-06, "loss": 0.4733, "step": 1013 }, { "epoch": 0.07188940092165899, "grad_norm": 3.081048482005368, "learning_rate": 9.954181203314579e-06, "loss": 0.4901, "step": 1014 }, { "epoch": 0.07196029776674938, "grad_norm": 3.3121285285190556, "learning_rate": 9.954025992394157e-06, "loss": 0.4233, "step": 1015 }, { "epoch": 0.07203119461183978, "grad_norm": 1.902916372632216, "learning_rate": 9.953870520244164e-06, "loss": 0.4436, "step": 1016 }, { "epoch": 0.07210209145693017, "grad_norm": 2.7997104856955084, "learning_rate": 9.953714786872797e-06, "loss": 0.431, "step": 1017 }, { "epoch": 0.07217298830202055, "grad_norm": 3.388470299238671, "learning_rate": 9.953558792288268e-06, "loss": 0.4451, "step": 1018 }, { "epoch": 0.07224388514711096, "grad_norm": 2.1655068850391133, "learning_rate": 9.953402536498804e-06, "loss": 0.434, "step": 1019 }, { "epoch": 0.07231478199220134, "grad_norm": 5.916657110201965, "learning_rate": 9.953246019512643e-06, "loss": 0.49, "step": 1020 }, { "epoch": 0.07238567883729174, "grad_norm": 2.6313018385661704, "learning_rate": 9.953089241338038e-06, "loss": 0.4654, "step": 1021 }, { "epoch": 0.07245657568238213, "grad_norm": 3.121743109126661, "learning_rate": 9.952932201983258e-06, "loss": 0.4591, "step": 1022 }, { "epoch": 0.07252747252747253, "grad_norm": 2.3421567742116673, "learning_rate": 9.952774901456584e-06, "loss": 0.4611, "step": 1023 }, { "epoch": 0.07259836937256292, "grad_norm": 3.481808347341085, "learning_rate": 9.952617339766306e-06, "loss": 0.5058, "step": 1024 }, { "epoch": 0.07266926621765331, "grad_norm": 2.2964545529979614, "learning_rate": 9.952459516920738e-06, "loss": 0.4703, "step": 1025 }, { "epoch": 0.07274016306274371, "grad_norm": 2.6495778139366943, "learning_rate": 9.952301432928198e-06, "loss": 0.4515, "step": 1026 }, { "epoch": 0.0728110599078341, "grad_norm": 3.072879632843977, "learning_rate": 9.952143087797023e-06, "loss": 0.5573, "step": 1027 }, { "epoch": 0.0728819567529245, "grad_norm": 2.213401723609293, "learning_rate": 9.951984481535566e-06, "loss": 0.4581, "step": 1028 }, { "epoch": 0.07295285359801489, "grad_norm": 5.197607678148979, "learning_rate": 9.951825614152186e-06, "loss": 0.4923, "step": 1029 }, { "epoch": 0.07302375044310529, "grad_norm": 18.35939482323764, "learning_rate": 9.95166648565526e-06, "loss": 0.4804, "step": 1030 }, { "epoch": 0.07309464728819567, "grad_norm": 4.8557979271579255, "learning_rate": 9.951507096053184e-06, "loss": 0.4105, "step": 1031 }, { "epoch": 0.07316554413328608, "grad_norm": 3.534136904271138, "learning_rate": 9.951347445354356e-06, "loss": 0.4815, "step": 1032 }, { "epoch": 0.07323644097837646, "grad_norm": 3.148102820199614, "learning_rate": 9.951187533567202e-06, "loss": 0.4807, "step": 1033 }, { "epoch": 0.07330733782346685, "grad_norm": 2.98116059388516, "learning_rate": 9.951027360700147e-06, "loss": 0.4637, "step": 1034 }, { "epoch": 0.07337823466855725, "grad_norm": 3.9040404292518613, "learning_rate": 9.950866926761641e-06, "loss": 0.4497, "step": 1035 }, { "epoch": 0.07344913151364764, "grad_norm": 8.413123340617153, "learning_rate": 9.950706231760145e-06, "loss": 0.4591, "step": 1036 }, { "epoch": 0.07352002835873804, "grad_norm": 12.426705646540112, "learning_rate": 9.950545275704129e-06, "loss": 0.4622, "step": 1037 }, { "epoch": 0.07359092520382843, "grad_norm": 3.7757092706215634, "learning_rate": 9.950384058602083e-06, "loss": 0.4485, "step": 1038 }, { "epoch": 0.07366182204891883, "grad_norm": 2.4054863324347235, "learning_rate": 9.950222580462506e-06, "loss": 0.4942, "step": 1039 }, { "epoch": 0.07373271889400922, "grad_norm": 2.921854376231596, "learning_rate": 9.950060841293913e-06, "loss": 0.4634, "step": 1040 }, { "epoch": 0.0738036157390996, "grad_norm": 3.538739058461975, "learning_rate": 9.949898841104837e-06, "loss": 0.4952, "step": 1041 }, { "epoch": 0.07387451258419, "grad_norm": 2.345555337209188, "learning_rate": 9.949736579903815e-06, "loss": 0.4042, "step": 1042 }, { "epoch": 0.07394540942928039, "grad_norm": 2.862361383428773, "learning_rate": 9.949574057699404e-06, "loss": 0.4933, "step": 1043 }, { "epoch": 0.0740163062743708, "grad_norm": 2.91273605282155, "learning_rate": 9.949411274500178e-06, "loss": 0.4775, "step": 1044 }, { "epoch": 0.07408720311946118, "grad_norm": 2.6888773547497786, "learning_rate": 9.949248230314717e-06, "loss": 0.4252, "step": 1045 }, { "epoch": 0.07415809996455158, "grad_norm": 3.3739345297999264, "learning_rate": 9.949084925151617e-06, "loss": 0.4782, "step": 1046 }, { "epoch": 0.07422899680964197, "grad_norm": 2.9241173268992764, "learning_rate": 9.948921359019493e-06, "loss": 0.5147, "step": 1047 }, { "epoch": 0.07429989365473237, "grad_norm": 2.793671892545451, "learning_rate": 9.94875753192697e-06, "loss": 0.4235, "step": 1048 }, { "epoch": 0.07437079049982276, "grad_norm": 5.093765582891687, "learning_rate": 9.948593443882682e-06, "loss": 0.4461, "step": 1049 }, { "epoch": 0.07444168734491315, "grad_norm": 3.170077427876683, "learning_rate": 9.948429094895287e-06, "loss": 0.4513, "step": 1050 }, { "epoch": 0.07451258419000355, "grad_norm": 3.8345606454289665, "learning_rate": 9.948264484973446e-06, "loss": 0.4819, "step": 1051 }, { "epoch": 0.07458348103509393, "grad_norm": 4.366999438031701, "learning_rate": 9.948099614125846e-06, "loss": 0.4476, "step": 1052 }, { "epoch": 0.07465437788018434, "grad_norm": 3.039564986059401, "learning_rate": 9.947934482361174e-06, "loss": 0.5257, "step": 1053 }, { "epoch": 0.07472527472527472, "grad_norm": 3.4466900998812178, "learning_rate": 9.94776908968814e-06, "loss": 0.4817, "step": 1054 }, { "epoch": 0.07479617157036512, "grad_norm": 1.0791035818761, "learning_rate": 9.947603436115466e-06, "loss": 0.6318, "step": 1055 }, { "epoch": 0.07486706841545551, "grad_norm": 4.832475265232096, "learning_rate": 9.947437521651887e-06, "loss": 0.4776, "step": 1056 }, { "epoch": 0.0749379652605459, "grad_norm": 4.654964412682383, "learning_rate": 9.94727134630615e-06, "loss": 0.5, "step": 1057 }, { "epoch": 0.0750088621056363, "grad_norm": 2.696099232898556, "learning_rate": 9.94710491008702e-06, "loss": 0.4722, "step": 1058 }, { "epoch": 0.07507975895072669, "grad_norm": 2.3728924274501284, "learning_rate": 9.946938213003272e-06, "loss": 0.4062, "step": 1059 }, { "epoch": 0.07515065579581709, "grad_norm": 2.95253656237558, "learning_rate": 9.946771255063696e-06, "loss": 0.4581, "step": 1060 }, { "epoch": 0.07522155264090748, "grad_norm": 6.2644757618534, "learning_rate": 9.946604036277096e-06, "loss": 0.4301, "step": 1061 }, { "epoch": 0.07529244948599788, "grad_norm": 5.14064038879472, "learning_rate": 9.94643655665229e-06, "loss": 0.4747, "step": 1062 }, { "epoch": 0.07536334633108827, "grad_norm": 2.8672989888352354, "learning_rate": 9.946268816198108e-06, "loss": 0.4973, "step": 1063 }, { "epoch": 0.07543424317617867, "grad_norm": 5.030367806114156, "learning_rate": 9.946100814923397e-06, "loss": 0.4974, "step": 1064 }, { "epoch": 0.07550514002126905, "grad_norm": 2.440337807901486, "learning_rate": 9.945932552837014e-06, "loss": 0.3997, "step": 1065 }, { "epoch": 0.07557603686635944, "grad_norm": 2.757010347826085, "learning_rate": 9.945764029947831e-06, "loss": 0.4832, "step": 1066 }, { "epoch": 0.07564693371144984, "grad_norm": 2.958994798807274, "learning_rate": 9.94559524626474e-06, "loss": 0.4171, "step": 1067 }, { "epoch": 0.07571783055654023, "grad_norm": 3.0993320843265892, "learning_rate": 9.945426201796634e-06, "loss": 0.439, "step": 1068 }, { "epoch": 0.07578872740163063, "grad_norm": 1.9955888072118106, "learning_rate": 9.945256896552429e-06, "loss": 0.4315, "step": 1069 }, { "epoch": 0.07585962424672102, "grad_norm": 2.9599862677001214, "learning_rate": 9.945087330541055e-06, "loss": 0.4688, "step": 1070 }, { "epoch": 0.07593052109181142, "grad_norm": 2.171977078902972, "learning_rate": 9.944917503771451e-06, "loss": 0.4497, "step": 1071 }, { "epoch": 0.07600141793690181, "grad_norm": 3.4364785916926324, "learning_rate": 9.944747416252573e-06, "loss": 0.5105, "step": 1072 }, { "epoch": 0.0760723147819922, "grad_norm": 2.6426616238013025, "learning_rate": 9.944577067993388e-06, "loss": 0.4555, "step": 1073 }, { "epoch": 0.0761432116270826, "grad_norm": 3.4191329081316892, "learning_rate": 9.944406459002882e-06, "loss": 0.4454, "step": 1074 }, { "epoch": 0.07621410847217298, "grad_norm": 2.5663617286483427, "learning_rate": 9.94423558929005e-06, "loss": 0.4639, "step": 1075 }, { "epoch": 0.07628500531726339, "grad_norm": 2.874586657987844, "learning_rate": 9.9440644588639e-06, "loss": 0.4443, "step": 1076 }, { "epoch": 0.07635590216235377, "grad_norm": 5.292602949041532, "learning_rate": 9.94389306773346e-06, "loss": 0.4531, "step": 1077 }, { "epoch": 0.07642679900744417, "grad_norm": 3.2842097448264687, "learning_rate": 9.943721415907761e-06, "loss": 0.5257, "step": 1078 }, { "epoch": 0.07649769585253456, "grad_norm": 4.376405815019089, "learning_rate": 9.943549503395861e-06, "loss": 0.4623, "step": 1079 }, { "epoch": 0.07656859269762496, "grad_norm": 3.383132380929524, "learning_rate": 9.943377330206822e-06, "loss": 0.4997, "step": 1080 }, { "epoch": 0.07663948954271535, "grad_norm": 3.327373443644337, "learning_rate": 9.943204896349724e-06, "loss": 0.4525, "step": 1081 }, { "epoch": 0.07671038638780574, "grad_norm": 3.9633559958727305, "learning_rate": 9.943032201833657e-06, "loss": 0.462, "step": 1082 }, { "epoch": 0.07678128323289614, "grad_norm": 2.8056030745004814, "learning_rate": 9.942859246667731e-06, "loss": 0.4386, "step": 1083 }, { "epoch": 0.07685218007798653, "grad_norm": 3.2721869016471263, "learning_rate": 9.942686030861063e-06, "loss": 0.4593, "step": 1084 }, { "epoch": 0.07692307692307693, "grad_norm": 7.861399931307453, "learning_rate": 9.942512554422788e-06, "loss": 0.4912, "step": 1085 }, { "epoch": 0.07699397376816731, "grad_norm": 3.8972157493545265, "learning_rate": 9.942338817362054e-06, "loss": 0.4228, "step": 1086 }, { "epoch": 0.07706487061325772, "grad_norm": 2.601763222099773, "learning_rate": 9.942164819688023e-06, "loss": 0.4517, "step": 1087 }, { "epoch": 0.0771357674583481, "grad_norm": 2.825511152849376, "learning_rate": 9.941990561409866e-06, "loss": 0.4133, "step": 1088 }, { "epoch": 0.07720666430343849, "grad_norm": 2.926753526656504, "learning_rate": 9.941816042536779e-06, "loss": 0.4572, "step": 1089 }, { "epoch": 0.07727756114852889, "grad_norm": 3.014925266470665, "learning_rate": 9.941641263077958e-06, "loss": 0.4637, "step": 1090 }, { "epoch": 0.07734845799361928, "grad_norm": 6.379502417785546, "learning_rate": 9.94146622304262e-06, "loss": 0.4964, "step": 1091 }, { "epoch": 0.07741935483870968, "grad_norm": 3.2431192291488666, "learning_rate": 9.941290922439996e-06, "loss": 0.484, "step": 1092 }, { "epoch": 0.07749025168380007, "grad_norm": 5.919328479373313, "learning_rate": 9.941115361279332e-06, "loss": 0.4753, "step": 1093 }, { "epoch": 0.07756114852889047, "grad_norm": 2.398706596647491, "learning_rate": 9.940939539569882e-06, "loss": 0.5394, "step": 1094 }, { "epoch": 0.07763204537398086, "grad_norm": 1.933198856052526, "learning_rate": 9.940763457320922e-06, "loss": 0.4785, "step": 1095 }, { "epoch": 0.07770294221907126, "grad_norm": 2.0225354251229013, "learning_rate": 9.940587114541732e-06, "loss": 0.4484, "step": 1096 }, { "epoch": 0.07777383906416165, "grad_norm": 3.059336412687386, "learning_rate": 9.940410511241612e-06, "loss": 0.4933, "step": 1097 }, { "epoch": 0.07784473590925203, "grad_norm": 4.473389434094355, "learning_rate": 9.940233647429874e-06, "loss": 0.4385, "step": 1098 }, { "epoch": 0.07791563275434243, "grad_norm": 2.59635782146704, "learning_rate": 9.940056523115847e-06, "loss": 0.4429, "step": 1099 }, { "epoch": 0.07798652959943282, "grad_norm": 5.280533652905831, "learning_rate": 9.939879138308868e-06, "loss": 0.4644, "step": 1100 }, { "epoch": 0.07805742644452322, "grad_norm": 2.9101574711951614, "learning_rate": 9.939701493018292e-06, "loss": 0.4427, "step": 1101 }, { "epoch": 0.07812832328961361, "grad_norm": 3.2554076405981767, "learning_rate": 9.939523587253485e-06, "loss": 0.5638, "step": 1102 }, { "epoch": 0.07819922013470401, "grad_norm": 3.9195926422646115, "learning_rate": 9.93934542102383e-06, "loss": 0.439, "step": 1103 }, { "epoch": 0.0782701169797944, "grad_norm": 5.46632765121804, "learning_rate": 9.93916699433872e-06, "loss": 0.4031, "step": 1104 }, { "epoch": 0.07834101382488479, "grad_norm": 2.441506930891397, "learning_rate": 9.938988307207567e-06, "loss": 0.4875, "step": 1105 }, { "epoch": 0.07841191066997519, "grad_norm": 5.87085921315682, "learning_rate": 9.938809359639788e-06, "loss": 0.4425, "step": 1106 }, { "epoch": 0.07848280751506558, "grad_norm": 4.390716414593827, "learning_rate": 9.938630151644824e-06, "loss": 0.5084, "step": 1107 }, { "epoch": 0.07855370436015598, "grad_norm": 4.548980378229315, "learning_rate": 9.93845068323212e-06, "loss": 0.4509, "step": 1108 }, { "epoch": 0.07862460120524636, "grad_norm": 3.826267789124482, "learning_rate": 9.938270954411145e-06, "loss": 0.4355, "step": 1109 }, { "epoch": 0.07869549805033677, "grad_norm": 2.961628993786082, "learning_rate": 9.938090965191372e-06, "loss": 0.4672, "step": 1110 }, { "epoch": 0.07876639489542715, "grad_norm": 2.926918563748918, "learning_rate": 9.937910715582293e-06, "loss": 0.4885, "step": 1111 }, { "epoch": 0.07883729174051755, "grad_norm": 2.48324693703784, "learning_rate": 9.937730205593413e-06, "loss": 0.4076, "step": 1112 }, { "epoch": 0.07890818858560794, "grad_norm": 2.3981543953871127, "learning_rate": 9.937549435234251e-06, "loss": 0.4281, "step": 1113 }, { "epoch": 0.07897908543069833, "grad_norm": 2.6878781873334265, "learning_rate": 9.937368404514338e-06, "loss": 0.4481, "step": 1114 }, { "epoch": 0.07904998227578873, "grad_norm": 2.825909948447235, "learning_rate": 9.93718711344322e-06, "loss": 0.4584, "step": 1115 }, { "epoch": 0.07912087912087912, "grad_norm": 2.7820543264838284, "learning_rate": 9.937005562030459e-06, "loss": 0.4735, "step": 1116 }, { "epoch": 0.07919177596596952, "grad_norm": 12.639678938464728, "learning_rate": 9.936823750285624e-06, "loss": 0.4987, "step": 1117 }, { "epoch": 0.0792626728110599, "grad_norm": 4.065762806332103, "learning_rate": 9.936641678218308e-06, "loss": 0.5348, "step": 1118 }, { "epoch": 0.07933356965615031, "grad_norm": 2.1759522248400422, "learning_rate": 9.936459345838106e-06, "loss": 0.4377, "step": 1119 }, { "epoch": 0.0794044665012407, "grad_norm": 3.3484676303611005, "learning_rate": 9.936276753154635e-06, "loss": 0.447, "step": 1120 }, { "epoch": 0.07947536334633108, "grad_norm": 2.161458987745513, "learning_rate": 9.936093900177523e-06, "loss": 0.479, "step": 1121 }, { "epoch": 0.07954626019142148, "grad_norm": 5.061153593622392, "learning_rate": 9.935910786916413e-06, "loss": 0.4512, "step": 1122 }, { "epoch": 0.07961715703651187, "grad_norm": 3.805912825211518, "learning_rate": 9.93572741338096e-06, "loss": 0.4379, "step": 1123 }, { "epoch": 0.07968805388160227, "grad_norm": 6.76038819160023, "learning_rate": 9.935543779580831e-06, "loss": 0.4759, "step": 1124 }, { "epoch": 0.07975895072669266, "grad_norm": 3.754056788713772, "learning_rate": 9.935359885525713e-06, "loss": 0.4793, "step": 1125 }, { "epoch": 0.07982984757178306, "grad_norm": 3.7986107813075107, "learning_rate": 9.935175731225301e-06, "loss": 0.466, "step": 1126 }, { "epoch": 0.07990074441687345, "grad_norm": 2.5947818431673193, "learning_rate": 9.934991316689306e-06, "loss": 0.4004, "step": 1127 }, { "epoch": 0.07997164126196385, "grad_norm": 3.178894976727988, "learning_rate": 9.934806641927453e-06, "loss": 0.4391, "step": 1128 }, { "epoch": 0.08004253810705424, "grad_norm": 4.01557502251565, "learning_rate": 9.934621706949479e-06, "loss": 0.4881, "step": 1129 }, { "epoch": 0.08011343495214462, "grad_norm": 2.8710461742215703, "learning_rate": 9.934436511765134e-06, "loss": 0.4816, "step": 1130 }, { "epoch": 0.08018433179723503, "grad_norm": 4.4328723173214115, "learning_rate": 9.934251056384188e-06, "loss": 0.4784, "step": 1131 }, { "epoch": 0.08025522864232541, "grad_norm": 1.4158868082059524, "learning_rate": 9.934065340816415e-06, "loss": 0.7341, "step": 1132 }, { "epoch": 0.08032612548741581, "grad_norm": 2.7315770201038783, "learning_rate": 9.933879365071613e-06, "loss": 0.4717, "step": 1133 }, { "epoch": 0.0803970223325062, "grad_norm": 3.4163973641063423, "learning_rate": 9.933693129159585e-06, "loss": 0.5061, "step": 1134 }, { "epoch": 0.0804679191775966, "grad_norm": 2.3176474076358122, "learning_rate": 9.933506633090154e-06, "loss": 0.4481, "step": 1135 }, { "epoch": 0.08053881602268699, "grad_norm": 4.780549961059213, "learning_rate": 9.93331987687315e-06, "loss": 0.4664, "step": 1136 }, { "epoch": 0.08060971286777738, "grad_norm": 2.256000438038709, "learning_rate": 9.933132860518424e-06, "loss": 0.5002, "step": 1137 }, { "epoch": 0.08068060971286778, "grad_norm": 2.8344208450737423, "learning_rate": 9.932945584035838e-06, "loss": 0.4219, "step": 1138 }, { "epoch": 0.08075150655795817, "grad_norm": 2.9159425077750574, "learning_rate": 9.932758047435267e-06, "loss": 0.4698, "step": 1139 }, { "epoch": 0.08082240340304857, "grad_norm": 3.218597556029383, "learning_rate": 9.932570250726599e-06, "loss": 0.415, "step": 1140 }, { "epoch": 0.08089330024813896, "grad_norm": 3.367594802246883, "learning_rate": 9.932382193919736e-06, "loss": 0.4749, "step": 1141 }, { "epoch": 0.08096419709322936, "grad_norm": 3.246829955294623, "learning_rate": 9.932193877024594e-06, "loss": 0.4956, "step": 1142 }, { "epoch": 0.08103509393831974, "grad_norm": 4.35109918034061, "learning_rate": 9.932005300051106e-06, "loss": 0.3979, "step": 1143 }, { "epoch": 0.08110599078341015, "grad_norm": 6.058593522463753, "learning_rate": 9.931816463009214e-06, "loss": 0.4615, "step": 1144 }, { "epoch": 0.08117688762850053, "grad_norm": 3.7845436732072457, "learning_rate": 9.931627365908874e-06, "loss": 0.4639, "step": 1145 }, { "epoch": 0.08124778447359092, "grad_norm": 2.671581666245115, "learning_rate": 9.931438008760062e-06, "loss": 0.4794, "step": 1146 }, { "epoch": 0.08131868131868132, "grad_norm": 3.0265735789430703, "learning_rate": 9.931248391572756e-06, "loss": 0.4616, "step": 1147 }, { "epoch": 0.08138957816377171, "grad_norm": 3.8857559678411744, "learning_rate": 9.931058514356961e-06, "loss": 0.4693, "step": 1148 }, { "epoch": 0.08146047500886211, "grad_norm": 16.67906373035783, "learning_rate": 9.930868377122686e-06, "loss": 0.5004, "step": 1149 }, { "epoch": 0.0815313718539525, "grad_norm": 2.49135712617147, "learning_rate": 9.930677979879958e-06, "loss": 0.4426, "step": 1150 }, { "epoch": 0.0816022686990429, "grad_norm": 2.9886864974139082, "learning_rate": 9.930487322638817e-06, "loss": 0.5107, "step": 1151 }, { "epoch": 0.08167316554413329, "grad_norm": 4.65226686966156, "learning_rate": 9.930296405409316e-06, "loss": 0.5148, "step": 1152 }, { "epoch": 0.08174406238922367, "grad_norm": 4.750558413689454, "learning_rate": 9.930105228201523e-06, "loss": 0.4325, "step": 1153 }, { "epoch": 0.08181495923431407, "grad_norm": 5.8803340091370435, "learning_rate": 9.929913791025518e-06, "loss": 0.397, "step": 1154 }, { "epoch": 0.08188585607940446, "grad_norm": 3.90591255307194, "learning_rate": 9.929722093891396e-06, "loss": 0.5003, "step": 1155 }, { "epoch": 0.08195675292449486, "grad_norm": 4.9502490528931755, "learning_rate": 9.929530136809264e-06, "loss": 0.4607, "step": 1156 }, { "epoch": 0.08202764976958525, "grad_norm": 3.4502701753806932, "learning_rate": 9.929337919789247e-06, "loss": 0.4662, "step": 1157 }, { "epoch": 0.08209854661467565, "grad_norm": 3.177534588657412, "learning_rate": 9.929145442841479e-06, "loss": 0.4753, "step": 1158 }, { "epoch": 0.08216944345976604, "grad_norm": 3.2931712720917328, "learning_rate": 9.92895270597611e-06, "loss": 0.4201, "step": 1159 }, { "epoch": 0.08224034030485644, "grad_norm": 4.727995625065464, "learning_rate": 9.9287597092033e-06, "loss": 0.5011, "step": 1160 }, { "epoch": 0.08231123714994683, "grad_norm": 4.528397891833552, "learning_rate": 9.928566452533231e-06, "loss": 0.4281, "step": 1161 }, { "epoch": 0.08238213399503722, "grad_norm": 3.6142487182203094, "learning_rate": 9.928372935976089e-06, "loss": 0.5034, "step": 1162 }, { "epoch": 0.08245303084012762, "grad_norm": 4.2157218511045755, "learning_rate": 9.928179159542083e-06, "loss": 0.5653, "step": 1163 }, { "epoch": 0.082523927685218, "grad_norm": 3.36691750530843, "learning_rate": 9.927985123241428e-06, "loss": 0.5015, "step": 1164 }, { "epoch": 0.0825948245303084, "grad_norm": 3.055964535724667, "learning_rate": 9.927790827084355e-06, "loss": 0.4327, "step": 1165 }, { "epoch": 0.08266572137539879, "grad_norm": 3.446122729462899, "learning_rate": 9.92759627108111e-06, "loss": 0.4905, "step": 1166 }, { "epoch": 0.0827366182204892, "grad_norm": 4.5431986752636355, "learning_rate": 9.927401455241954e-06, "loss": 0.4784, "step": 1167 }, { "epoch": 0.08280751506557958, "grad_norm": 9.432283142656257, "learning_rate": 9.92720637957716e-06, "loss": 0.5379, "step": 1168 }, { "epoch": 0.08287841191066997, "grad_norm": 2.617294605075661, "learning_rate": 9.92701104409701e-06, "loss": 0.4121, "step": 1169 }, { "epoch": 0.08294930875576037, "grad_norm": 4.609672442722168, "learning_rate": 9.926815448811807e-06, "loss": 0.4381, "step": 1170 }, { "epoch": 0.08302020560085076, "grad_norm": 2.955129842951614, "learning_rate": 9.926619593731865e-06, "loss": 0.4738, "step": 1171 }, { "epoch": 0.08309110244594116, "grad_norm": 2.686799213833263, "learning_rate": 9.926423478867514e-06, "loss": 0.456, "step": 1172 }, { "epoch": 0.08316199929103155, "grad_norm": 2.6406075218223206, "learning_rate": 9.926227104229089e-06, "loss": 0.4231, "step": 1173 }, { "epoch": 0.08323289613612195, "grad_norm": 2.8892061650816907, "learning_rate": 9.926030469826951e-06, "loss": 0.4713, "step": 1174 }, { "epoch": 0.08330379298121234, "grad_norm": 4.418781910461604, "learning_rate": 9.925833575671466e-06, "loss": 0.4836, "step": 1175 }, { "epoch": 0.08337468982630274, "grad_norm": 2.3038004498052955, "learning_rate": 9.925636421773018e-06, "loss": 0.4052, "step": 1176 }, { "epoch": 0.08344558667139312, "grad_norm": 3.9611900705980445, "learning_rate": 9.925439008141999e-06, "loss": 0.4619, "step": 1177 }, { "epoch": 0.08351648351648351, "grad_norm": 3.3191485836300307, "learning_rate": 9.925241334788824e-06, "loss": 0.4741, "step": 1178 }, { "epoch": 0.08358738036157391, "grad_norm": 4.25563270869145, "learning_rate": 9.925043401723915e-06, "loss": 0.5045, "step": 1179 }, { "epoch": 0.0836582772066643, "grad_norm": 1.989712771463771, "learning_rate": 9.924845208957707e-06, "loss": 0.47, "step": 1180 }, { "epoch": 0.0837291740517547, "grad_norm": 3.117441055209465, "learning_rate": 9.92464675650065e-06, "loss": 0.5138, "step": 1181 }, { "epoch": 0.08380007089684509, "grad_norm": 4.377195010443442, "learning_rate": 9.924448044363216e-06, "loss": 0.4609, "step": 1182 }, { "epoch": 0.08387096774193549, "grad_norm": 2.6078262325313877, "learning_rate": 9.924249072555875e-06, "loss": 0.476, "step": 1183 }, { "epoch": 0.08394186458702588, "grad_norm": 2.0861736682174774, "learning_rate": 9.924049841089121e-06, "loss": 0.4107, "step": 1184 }, { "epoch": 0.08401276143211626, "grad_norm": 2.2586794617055586, "learning_rate": 9.923850349973461e-06, "loss": 0.4327, "step": 1185 }, { "epoch": 0.08408365827720667, "grad_norm": 3.28396280466244, "learning_rate": 9.923650599219414e-06, "loss": 0.4378, "step": 1186 }, { "epoch": 0.08415455512229705, "grad_norm": 2.6190311117436544, "learning_rate": 9.923450588837514e-06, "loss": 0.4958, "step": 1187 }, { "epoch": 0.08422545196738745, "grad_norm": 9.234530884732113, "learning_rate": 9.923250318838304e-06, "loss": 0.4067, "step": 1188 }, { "epoch": 0.08429634881247784, "grad_norm": 2.5396713866061615, "learning_rate": 9.923049789232348e-06, "loss": 0.4528, "step": 1189 }, { "epoch": 0.08436724565756824, "grad_norm": 3.296376277901984, "learning_rate": 9.92284900003022e-06, "loss": 0.4569, "step": 1190 }, { "epoch": 0.08443814250265863, "grad_norm": 3.818491107231182, "learning_rate": 9.922647951242507e-06, "loss": 0.4952, "step": 1191 }, { "epoch": 0.08450903934774903, "grad_norm": 2.634248280150172, "learning_rate": 9.922446642879809e-06, "loss": 0.4997, "step": 1192 }, { "epoch": 0.08457993619283942, "grad_norm": 2.9741481188405463, "learning_rate": 9.922245074952744e-06, "loss": 0.4628, "step": 1193 }, { "epoch": 0.0846508330379298, "grad_norm": 2.8851906440585906, "learning_rate": 9.922043247471938e-06, "loss": 0.4114, "step": 1194 }, { "epoch": 0.08472172988302021, "grad_norm": 3.0075023868041186, "learning_rate": 9.921841160448033e-06, "loss": 0.4823, "step": 1195 }, { "epoch": 0.0847926267281106, "grad_norm": 3.0150383343904865, "learning_rate": 9.92163881389169e-06, "loss": 0.4041, "step": 1196 }, { "epoch": 0.084863523573201, "grad_norm": 2.284174199969728, "learning_rate": 9.921436207813575e-06, "loss": 0.4639, "step": 1197 }, { "epoch": 0.08493442041829138, "grad_norm": 2.838901291290282, "learning_rate": 9.921233342224373e-06, "loss": 0.4651, "step": 1198 }, { "epoch": 0.08500531726338179, "grad_norm": 4.33421476657656, "learning_rate": 9.921030217134778e-06, "loss": 0.4362, "step": 1199 }, { "epoch": 0.08507621410847217, "grad_norm": 3.7084486882757943, "learning_rate": 9.920826832555505e-06, "loss": 0.4318, "step": 1200 }, { "epoch": 0.08514711095356256, "grad_norm": 2.7326680059326627, "learning_rate": 9.920623188497278e-06, "loss": 0.4709, "step": 1201 }, { "epoch": 0.08521800779865296, "grad_norm": 3.1608271608326364, "learning_rate": 9.920419284970834e-06, "loss": 0.4771, "step": 1202 }, { "epoch": 0.08528890464374335, "grad_norm": 4.14155130052577, "learning_rate": 9.920215121986925e-06, "loss": 0.4734, "step": 1203 }, { "epoch": 0.08535980148883375, "grad_norm": 2.369978692301208, "learning_rate": 9.920010699556318e-06, "loss": 0.4301, "step": 1204 }, { "epoch": 0.08543069833392414, "grad_norm": 2.3535207538965692, "learning_rate": 9.919806017689793e-06, "loss": 0.4946, "step": 1205 }, { "epoch": 0.08550159517901454, "grad_norm": 3.4675053343514164, "learning_rate": 9.919601076398138e-06, "loss": 0.4591, "step": 1206 }, { "epoch": 0.08557249202410493, "grad_norm": 7.9688040295218725, "learning_rate": 9.919395875692166e-06, "loss": 0.4597, "step": 1207 }, { "epoch": 0.08564338886919533, "grad_norm": 4.287752107683109, "learning_rate": 9.919190415582693e-06, "loss": 0.4486, "step": 1208 }, { "epoch": 0.08571428571428572, "grad_norm": 3.677693504681326, "learning_rate": 9.918984696080555e-06, "loss": 0.4848, "step": 1209 }, { "epoch": 0.0857851825593761, "grad_norm": 2.9264050188632256, "learning_rate": 9.918778717196602e-06, "loss": 0.4787, "step": 1210 }, { "epoch": 0.0858560794044665, "grad_norm": 2.694893332290938, "learning_rate": 9.91857247894169e-06, "loss": 0.4403, "step": 1211 }, { "epoch": 0.08592697624955689, "grad_norm": 4.395541469418124, "learning_rate": 9.918365981326698e-06, "loss": 0.5076, "step": 1212 }, { "epoch": 0.08599787309464729, "grad_norm": 3.9760395855365234, "learning_rate": 9.918159224362513e-06, "loss": 0.4504, "step": 1213 }, { "epoch": 0.08606876993973768, "grad_norm": 4.242909540949698, "learning_rate": 9.91795220806004e-06, "loss": 0.4982, "step": 1214 }, { "epoch": 0.08613966678482808, "grad_norm": 3.068769735615108, "learning_rate": 9.917744932430193e-06, "loss": 0.4823, "step": 1215 }, { "epoch": 0.08621056362991847, "grad_norm": 3.8765692655200428, "learning_rate": 9.917537397483901e-06, "loss": 0.4773, "step": 1216 }, { "epoch": 0.08628146047500886, "grad_norm": 2.3561998525933276, "learning_rate": 9.917329603232108e-06, "loss": 0.3989, "step": 1217 }, { "epoch": 0.08635235732009926, "grad_norm": 4.164110300961899, "learning_rate": 9.917121549685774e-06, "loss": 0.4649, "step": 1218 }, { "epoch": 0.08642325416518964, "grad_norm": 3.6701723742216403, "learning_rate": 9.916913236855868e-06, "loss": 0.4718, "step": 1219 }, { "epoch": 0.08649415101028005, "grad_norm": 7.989407836828944, "learning_rate": 9.916704664753372e-06, "loss": 0.4837, "step": 1220 }, { "epoch": 0.08656504785537043, "grad_norm": 2.1105031995062693, "learning_rate": 9.916495833389286e-06, "loss": 0.4747, "step": 1221 }, { "epoch": 0.08663594470046083, "grad_norm": 3.6305481557487016, "learning_rate": 9.916286742774623e-06, "loss": 0.4689, "step": 1222 }, { "epoch": 0.08670684154555122, "grad_norm": 3.753582069196251, "learning_rate": 9.916077392920409e-06, "loss": 0.4351, "step": 1223 }, { "epoch": 0.08677773839064162, "grad_norm": 2.517015581658319, "learning_rate": 9.91586778383768e-06, "loss": 0.471, "step": 1224 }, { "epoch": 0.08684863523573201, "grad_norm": 3.714486346434294, "learning_rate": 9.91565791553749e-06, "loss": 0.51, "step": 1225 }, { "epoch": 0.0869195320808224, "grad_norm": 1.872167093639576, "learning_rate": 9.915447788030907e-06, "loss": 0.4085, "step": 1226 }, { "epoch": 0.0869904289259128, "grad_norm": 2.553604897109171, "learning_rate": 9.915237401329012e-06, "loss": 0.4872, "step": 1227 }, { "epoch": 0.08706132577100319, "grad_norm": 2.945395564713251, "learning_rate": 9.915026755442895e-06, "loss": 0.4585, "step": 1228 }, { "epoch": 0.08713222261609359, "grad_norm": 3.8798185187956475, "learning_rate": 9.914815850383667e-06, "loss": 0.5215, "step": 1229 }, { "epoch": 0.08720311946118398, "grad_norm": 2.1568888582943857, "learning_rate": 9.914604686162448e-06, "loss": 0.4427, "step": 1230 }, { "epoch": 0.08727401630627438, "grad_norm": 6.22148200929537, "learning_rate": 9.914393262790371e-06, "loss": 0.5291, "step": 1231 }, { "epoch": 0.08734491315136476, "grad_norm": 2.8352376110758133, "learning_rate": 9.914181580278588e-06, "loss": 0.4463, "step": 1232 }, { "epoch": 0.08741580999645515, "grad_norm": 2.644260685904243, "learning_rate": 9.913969638638258e-06, "loss": 0.4915, "step": 1233 }, { "epoch": 0.08748670684154555, "grad_norm": 1.8615419304542464, "learning_rate": 9.91375743788056e-06, "loss": 0.4392, "step": 1234 }, { "epoch": 0.08755760368663594, "grad_norm": 3.6185704064326076, "learning_rate": 9.913544978016682e-06, "loss": 0.4165, "step": 1235 }, { "epoch": 0.08762850053172634, "grad_norm": 2.637542303956858, "learning_rate": 9.913332259057827e-06, "loss": 0.4713, "step": 1236 }, { "epoch": 0.08769939737681673, "grad_norm": 1.9112661639008792, "learning_rate": 9.91311928101521e-06, "loss": 0.4442, "step": 1237 }, { "epoch": 0.08777029422190713, "grad_norm": 2.9981421285531065, "learning_rate": 9.912906043900064e-06, "loss": 0.4336, "step": 1238 }, { "epoch": 0.08784119106699752, "grad_norm": 5.405045880016273, "learning_rate": 9.912692547723632e-06, "loss": 0.444, "step": 1239 }, { "epoch": 0.08791208791208792, "grad_norm": 3.969957406042707, "learning_rate": 9.912478792497173e-06, "loss": 0.4942, "step": 1240 }, { "epoch": 0.0879829847571783, "grad_norm": 0.9997496052593535, "learning_rate": 9.912264778231957e-06, "loss": 0.6767, "step": 1241 }, { "epoch": 0.0880538816022687, "grad_norm": 2.702439353256491, "learning_rate": 9.912050504939272e-06, "loss": 0.5015, "step": 1242 }, { "epoch": 0.0881247784473591, "grad_norm": 3.5404270190337193, "learning_rate": 9.911835972630413e-06, "loss": 0.4226, "step": 1243 }, { "epoch": 0.08819567529244948, "grad_norm": 4.255936614965883, "learning_rate": 9.911621181316694e-06, "loss": 0.4776, "step": 1244 }, { "epoch": 0.08826657213753988, "grad_norm": 2.6721482303559294, "learning_rate": 9.911406131009442e-06, "loss": 0.4455, "step": 1245 }, { "epoch": 0.08833746898263027, "grad_norm": 3.2221863933633075, "learning_rate": 9.911190821719995e-06, "loss": 0.5259, "step": 1246 }, { "epoch": 0.08840836582772067, "grad_norm": 3.4550231211945723, "learning_rate": 9.910975253459707e-06, "loss": 0.4831, "step": 1247 }, { "epoch": 0.08847926267281106, "grad_norm": 3.4772496934443544, "learning_rate": 9.910759426239947e-06, "loss": 0.4679, "step": 1248 }, { "epoch": 0.08855015951790145, "grad_norm": 1.1475043380045569, "learning_rate": 9.910543340072094e-06, "loss": 0.6769, "step": 1249 }, { "epoch": 0.08862105636299185, "grad_norm": 3.9830907699741425, "learning_rate": 9.910326994967543e-06, "loss": 0.4745, "step": 1250 }, { "epoch": 0.08869195320808224, "grad_norm": 3.7845687861607584, "learning_rate": 9.9101103909377e-06, "loss": 0.4806, "step": 1251 }, { "epoch": 0.08876285005317264, "grad_norm": 3.02155734014065, "learning_rate": 9.90989352799399e-06, "loss": 0.4895, "step": 1252 }, { "epoch": 0.08883374689826302, "grad_norm": 5.696656803845103, "learning_rate": 9.909676406147845e-06, "loss": 0.3837, "step": 1253 }, { "epoch": 0.08890464374335343, "grad_norm": 6.384819907916961, "learning_rate": 9.909459025410717e-06, "loss": 0.4466, "step": 1254 }, { "epoch": 0.08897554058844381, "grad_norm": 3.386804321241593, "learning_rate": 9.909241385794068e-06, "loss": 0.4949, "step": 1255 }, { "epoch": 0.08904643743353421, "grad_norm": 3.739771349943868, "learning_rate": 9.909023487309373e-06, "loss": 0.4555, "step": 1256 }, { "epoch": 0.0891173342786246, "grad_norm": 4.768402404548384, "learning_rate": 9.908805329968121e-06, "loss": 0.4731, "step": 1257 }, { "epoch": 0.08918823112371499, "grad_norm": 4.272298663737852, "learning_rate": 9.908586913781819e-06, "loss": 0.4089, "step": 1258 }, { "epoch": 0.08925912796880539, "grad_norm": 4.8837478902079985, "learning_rate": 9.908368238761983e-06, "loss": 0.4545, "step": 1259 }, { "epoch": 0.08933002481389578, "grad_norm": 5.941332590443018, "learning_rate": 9.908149304920143e-06, "loss": 0.4899, "step": 1260 }, { "epoch": 0.08940092165898618, "grad_norm": 6.023649999660912, "learning_rate": 9.907930112267843e-06, "loss": 0.4436, "step": 1261 }, { "epoch": 0.08947181850407657, "grad_norm": 5.537926638730042, "learning_rate": 9.907710660816642e-06, "loss": 0.4914, "step": 1262 }, { "epoch": 0.08954271534916697, "grad_norm": 3.762256638220506, "learning_rate": 9.907490950578114e-06, "loss": 0.4867, "step": 1263 }, { "epoch": 0.08961361219425736, "grad_norm": 4.128508644932729, "learning_rate": 9.90727098156384e-06, "loss": 0.4689, "step": 1264 }, { "epoch": 0.08968450903934774, "grad_norm": 4.372098413906174, "learning_rate": 9.907050753785423e-06, "loss": 0.4644, "step": 1265 }, { "epoch": 0.08975540588443814, "grad_norm": 2.1419451740553237, "learning_rate": 9.906830267254475e-06, "loss": 0.4831, "step": 1266 }, { "epoch": 0.08982630272952853, "grad_norm": 26.36297290785987, "learning_rate": 9.90660952198262e-06, "loss": 0.4907, "step": 1267 }, { "epoch": 0.08989719957461893, "grad_norm": 3.870709909519836, "learning_rate": 9.9063885179815e-06, "loss": 0.5043, "step": 1268 }, { "epoch": 0.08996809641970932, "grad_norm": 3.0209809892905835, "learning_rate": 9.90616725526277e-06, "loss": 0.451, "step": 1269 }, { "epoch": 0.09003899326479972, "grad_norm": 3.4707748919940613, "learning_rate": 9.905945733838094e-06, "loss": 0.4603, "step": 1270 }, { "epoch": 0.09010989010989011, "grad_norm": 3.0405705135310295, "learning_rate": 9.905723953719158e-06, "loss": 0.421, "step": 1271 }, { "epoch": 0.09018078695498051, "grad_norm": 3.2861446668150975, "learning_rate": 9.90550191491765e-06, "loss": 0.4678, "step": 1272 }, { "epoch": 0.0902516838000709, "grad_norm": 5.049099047799353, "learning_rate": 9.905279617445284e-06, "loss": 0.4336, "step": 1273 }, { "epoch": 0.09032258064516129, "grad_norm": 4.6019782541997065, "learning_rate": 9.90505706131378e-06, "loss": 0.5126, "step": 1274 }, { "epoch": 0.09039347749025169, "grad_norm": 2.9806907838301218, "learning_rate": 9.904834246534872e-06, "loss": 0.4927, "step": 1275 }, { "epoch": 0.09046437433534207, "grad_norm": 3.529766084891218, "learning_rate": 9.90461117312031e-06, "loss": 0.4857, "step": 1276 }, { "epoch": 0.09053527118043248, "grad_norm": 3.7317396372509113, "learning_rate": 9.904387841081859e-06, "loss": 0.4469, "step": 1277 }, { "epoch": 0.09060616802552286, "grad_norm": 5.320078338271754, "learning_rate": 9.904164250431294e-06, "loss": 0.4754, "step": 1278 }, { "epoch": 0.09067706487061326, "grad_norm": 2.0959201615766694, "learning_rate": 9.903940401180402e-06, "loss": 0.4611, "step": 1279 }, { "epoch": 0.09074796171570365, "grad_norm": 2.7156675783587976, "learning_rate": 9.903716293340993e-06, "loss": 0.4523, "step": 1280 }, { "epoch": 0.09081885856079404, "grad_norm": 6.4577556313194355, "learning_rate": 9.90349192692488e-06, "loss": 0.4581, "step": 1281 }, { "epoch": 0.09088975540588444, "grad_norm": 1.6353257484133583, "learning_rate": 9.903267301943893e-06, "loss": 0.4652, "step": 1282 }, { "epoch": 0.09096065225097483, "grad_norm": 2.112294588927681, "learning_rate": 9.903042418409881e-06, "loss": 0.4549, "step": 1283 }, { "epoch": 0.09103154909606523, "grad_norm": 7.342798707806429, "learning_rate": 9.9028172763347e-06, "loss": 0.4591, "step": 1284 }, { "epoch": 0.09110244594115562, "grad_norm": 2.9355768724751545, "learning_rate": 9.902591875730221e-06, "loss": 0.4341, "step": 1285 }, { "epoch": 0.09117334278624602, "grad_norm": 3.764286676429223, "learning_rate": 9.90236621660833e-06, "loss": 0.5006, "step": 1286 }, { "epoch": 0.0912442396313364, "grad_norm": 6.630952799338855, "learning_rate": 9.902140298980928e-06, "loss": 0.408, "step": 1287 }, { "epoch": 0.0913151364764268, "grad_norm": 3.3903544126019343, "learning_rate": 9.901914122859926e-06, "loss": 0.5139, "step": 1288 }, { "epoch": 0.0913860333215172, "grad_norm": 4.288161545742007, "learning_rate": 9.90168768825725e-06, "loss": 0.4736, "step": 1289 }, { "epoch": 0.09145693016660758, "grad_norm": 2.5998388385178397, "learning_rate": 9.901460995184841e-06, "loss": 0.5058, "step": 1290 }, { "epoch": 0.09152782701169798, "grad_norm": 3.828433370448413, "learning_rate": 9.901234043654652e-06, "loss": 0.4901, "step": 1291 }, { "epoch": 0.09159872385678837, "grad_norm": 2.571827146653696, "learning_rate": 9.901006833678653e-06, "loss": 0.4017, "step": 1292 }, { "epoch": 0.09166962070187877, "grad_norm": 4.259801368657085, "learning_rate": 9.900779365268823e-06, "loss": 0.5084, "step": 1293 }, { "epoch": 0.09174051754696916, "grad_norm": 2.0285459692257053, "learning_rate": 9.900551638437154e-06, "loss": 0.4107, "step": 1294 }, { "epoch": 0.09181141439205956, "grad_norm": 2.5557418927654996, "learning_rate": 9.900323653195658e-06, "loss": 0.4235, "step": 1295 }, { "epoch": 0.09188231123714995, "grad_norm": 2.921640747406084, "learning_rate": 9.900095409556356e-06, "loss": 0.4482, "step": 1296 }, { "epoch": 0.09195320808224033, "grad_norm": 4.877545575346193, "learning_rate": 9.899866907531281e-06, "loss": 0.4527, "step": 1297 }, { "epoch": 0.09202410492733074, "grad_norm": 5.460752521856847, "learning_rate": 9.899638147132486e-06, "loss": 0.4111, "step": 1298 }, { "epoch": 0.09209500177242112, "grad_norm": 2.930102913348851, "learning_rate": 9.899409128372031e-06, "loss": 0.478, "step": 1299 }, { "epoch": 0.09216589861751152, "grad_norm": 4.01974645893253, "learning_rate": 9.899179851261994e-06, "loss": 0.4942, "step": 1300 }, { "epoch": 0.09223679546260191, "grad_norm": 3.155434665896623, "learning_rate": 9.898950315814464e-06, "loss": 0.4773, "step": 1301 }, { "epoch": 0.09230769230769231, "grad_norm": 3.457078410529994, "learning_rate": 9.898720522041544e-06, "loss": 0.4996, "step": 1302 }, { "epoch": 0.0923785891527827, "grad_norm": 1.867616467786662, "learning_rate": 9.898490469955351e-06, "loss": 0.483, "step": 1303 }, { "epoch": 0.0924494859978731, "grad_norm": 2.7365630372988368, "learning_rate": 9.898260159568017e-06, "loss": 0.4389, "step": 1304 }, { "epoch": 0.09252038284296349, "grad_norm": 2.0786294761026296, "learning_rate": 9.898029590891688e-06, "loss": 0.4209, "step": 1305 }, { "epoch": 0.09259127968805388, "grad_norm": 2.4561155426608168, "learning_rate": 9.897798763938519e-06, "loss": 0.4851, "step": 1306 }, { "epoch": 0.09266217653314428, "grad_norm": 2.22928427794935, "learning_rate": 9.897567678720683e-06, "loss": 0.4901, "step": 1307 }, { "epoch": 0.09273307337823467, "grad_norm": 1.2695881225667158, "learning_rate": 9.897336335250365e-06, "loss": 0.7775, "step": 1308 }, { "epoch": 0.09280397022332507, "grad_norm": 3.8761719423782623, "learning_rate": 9.897104733539765e-06, "loss": 0.4824, "step": 1309 }, { "epoch": 0.09287486706841545, "grad_norm": 2.917728085502531, "learning_rate": 9.896872873601093e-06, "loss": 0.4452, "step": 1310 }, { "epoch": 0.09294576391350585, "grad_norm": 2.942054443187069, "learning_rate": 9.896640755446575e-06, "loss": 0.4677, "step": 1311 }, { "epoch": 0.09301666075859624, "grad_norm": 6.333278675206965, "learning_rate": 9.896408379088456e-06, "loss": 0.4517, "step": 1312 }, { "epoch": 0.09308755760368663, "grad_norm": 2.6330625261652223, "learning_rate": 9.896175744538984e-06, "loss": 0.4325, "step": 1313 }, { "epoch": 0.09315845444877703, "grad_norm": 3.770637550174313, "learning_rate": 9.89594285181043e-06, "loss": 0.4387, "step": 1314 }, { "epoch": 0.09322935129386742, "grad_norm": 4.008719534779182, "learning_rate": 9.895709700915072e-06, "loss": 0.5492, "step": 1315 }, { "epoch": 0.09330024813895782, "grad_norm": 2.578608906984693, "learning_rate": 9.895476291865204e-06, "loss": 0.4592, "step": 1316 }, { "epoch": 0.09337114498404821, "grad_norm": 3.062601359747691, "learning_rate": 9.895242624673134e-06, "loss": 0.4335, "step": 1317 }, { "epoch": 0.09344204182913861, "grad_norm": 4.133736071170074, "learning_rate": 9.895008699351185e-06, "loss": 0.4293, "step": 1318 }, { "epoch": 0.093512938674229, "grad_norm": 2.6092880272318975, "learning_rate": 9.894774515911691e-06, "loss": 0.4214, "step": 1319 }, { "epoch": 0.0935838355193194, "grad_norm": 2.4859467462537213, "learning_rate": 9.894540074367e-06, "loss": 0.44, "step": 1320 }, { "epoch": 0.09365473236440978, "grad_norm": 4.013797373988659, "learning_rate": 9.894305374729476e-06, "loss": 0.4574, "step": 1321 }, { "epoch": 0.09372562920950017, "grad_norm": 3.377461294071985, "learning_rate": 9.894070417011491e-06, "loss": 0.4489, "step": 1322 }, { "epoch": 0.09379652605459057, "grad_norm": 4.600311322424704, "learning_rate": 9.89383520122544e-06, "loss": 0.4852, "step": 1323 }, { "epoch": 0.09386742289968096, "grad_norm": 17.977844167364022, "learning_rate": 9.893599727383725e-06, "loss": 0.4719, "step": 1324 }, { "epoch": 0.09393831974477136, "grad_norm": 2.7340526202122244, "learning_rate": 9.89336399549876e-06, "loss": 0.4392, "step": 1325 }, { "epoch": 0.09400921658986175, "grad_norm": 3.1016034024043835, "learning_rate": 9.893128005582976e-06, "loss": 0.4499, "step": 1326 }, { "epoch": 0.09408011343495215, "grad_norm": 1.973323148661261, "learning_rate": 9.892891757648816e-06, "loss": 0.4788, "step": 1327 }, { "epoch": 0.09415101028004254, "grad_norm": 3.1606482128041335, "learning_rate": 9.892655251708742e-06, "loss": 0.4773, "step": 1328 }, { "epoch": 0.09422190712513293, "grad_norm": 3.584008706957916, "learning_rate": 9.892418487775222e-06, "loss": 0.4398, "step": 1329 }, { "epoch": 0.09429280397022333, "grad_norm": 3.4138522135969525, "learning_rate": 9.89218146586074e-06, "loss": 0.4525, "step": 1330 }, { "epoch": 0.09436370081531371, "grad_norm": 4.98889038988043, "learning_rate": 9.891944185977796e-06, "loss": 0.4632, "step": 1331 }, { "epoch": 0.09443459766040412, "grad_norm": 2.9512914244166195, "learning_rate": 9.891706648138898e-06, "loss": 0.4547, "step": 1332 }, { "epoch": 0.0945054945054945, "grad_norm": 15.993927287352772, "learning_rate": 9.891468852356579e-06, "loss": 0.5104, "step": 1333 }, { "epoch": 0.0945763913505849, "grad_norm": 2.8498962395398206, "learning_rate": 9.891230798643373e-06, "loss": 0.4881, "step": 1334 }, { "epoch": 0.09464728819567529, "grad_norm": 2.990195438569614, "learning_rate": 9.89099248701183e-06, "loss": 0.4392, "step": 1335 }, { "epoch": 0.09471818504076569, "grad_norm": 1.7931308576570726, "learning_rate": 9.890753917474525e-06, "loss": 0.4919, "step": 1336 }, { "epoch": 0.09478908188585608, "grad_norm": 2.3535693409729466, "learning_rate": 9.890515090044031e-06, "loss": 0.4411, "step": 1337 }, { "epoch": 0.09485997873094647, "grad_norm": 2.9281262196126083, "learning_rate": 9.890276004732942e-06, "loss": 0.4683, "step": 1338 }, { "epoch": 0.09493087557603687, "grad_norm": 2.1086085033194086, "learning_rate": 9.890036661553867e-06, "loss": 0.4459, "step": 1339 }, { "epoch": 0.09500177242112726, "grad_norm": 1.1360616488549538, "learning_rate": 9.889797060519428e-06, "loss": 0.7821, "step": 1340 }, { "epoch": 0.09507266926621766, "grad_norm": 2.44753380463033, "learning_rate": 9.88955720164226e-06, "loss": 0.4537, "step": 1341 }, { "epoch": 0.09514356611130804, "grad_norm": 3.376883801880814, "learning_rate": 9.889317084935004e-06, "loss": 0.419, "step": 1342 }, { "epoch": 0.09521446295639845, "grad_norm": 3.9080301131867947, "learning_rate": 9.889076710410328e-06, "loss": 0.5165, "step": 1343 }, { "epoch": 0.09528535980148883, "grad_norm": 5.504885319263615, "learning_rate": 9.888836078080906e-06, "loss": 0.4329, "step": 1344 }, { "epoch": 0.09535625664657922, "grad_norm": 3.3528955494332835, "learning_rate": 9.888595187959427e-06, "loss": 0.5142, "step": 1345 }, { "epoch": 0.09542715349166962, "grad_norm": 4.662152534732398, "learning_rate": 9.88835404005859e-06, "loss": 0.4692, "step": 1346 }, { "epoch": 0.09549805033676001, "grad_norm": 3.128652381578117, "learning_rate": 9.888112634391115e-06, "loss": 0.4655, "step": 1347 }, { "epoch": 0.09556894718185041, "grad_norm": 2.381454546375555, "learning_rate": 9.88787097096973e-06, "loss": 0.4381, "step": 1348 }, { "epoch": 0.0956398440269408, "grad_norm": 5.236686134895902, "learning_rate": 9.887629049807176e-06, "loss": 0.4692, "step": 1349 }, { "epoch": 0.0957107408720312, "grad_norm": 5.299853532354861, "learning_rate": 9.887386870916213e-06, "loss": 0.4946, "step": 1350 }, { "epoch": 0.09578163771712159, "grad_norm": 3.877811484857888, "learning_rate": 9.88714443430961e-06, "loss": 0.4146, "step": 1351 }, { "epoch": 0.09585253456221199, "grad_norm": 14.338251380238567, "learning_rate": 9.886901740000153e-06, "loss": 0.4563, "step": 1352 }, { "epoch": 0.09592343140730238, "grad_norm": 3.1191795869455667, "learning_rate": 9.886658788000634e-06, "loss": 0.4998, "step": 1353 }, { "epoch": 0.09599432825239276, "grad_norm": 4.200331058783322, "learning_rate": 9.886415578323868e-06, "loss": 0.4726, "step": 1354 }, { "epoch": 0.09606522509748316, "grad_norm": 2.3517038523309735, "learning_rate": 9.886172110982679e-06, "loss": 0.4986, "step": 1355 }, { "epoch": 0.09613612194257355, "grad_norm": 2.8461955090489575, "learning_rate": 9.885928385989905e-06, "loss": 0.4745, "step": 1356 }, { "epoch": 0.09620701878766395, "grad_norm": 5.81830885755907, "learning_rate": 9.8856844033584e-06, "loss": 0.5083, "step": 1357 }, { "epoch": 0.09627791563275434, "grad_norm": 3.0294886032044492, "learning_rate": 9.885440163101025e-06, "loss": 0.4333, "step": 1358 }, { "epoch": 0.09634881247784474, "grad_norm": 3.6009706709705953, "learning_rate": 9.885195665230662e-06, "loss": 0.4775, "step": 1359 }, { "epoch": 0.09641970932293513, "grad_norm": 2.527244652381206, "learning_rate": 9.884950909760202e-06, "loss": 0.4075, "step": 1360 }, { "epoch": 0.09649060616802552, "grad_norm": 3.4415507421570384, "learning_rate": 9.884705896702553e-06, "loss": 0.4697, "step": 1361 }, { "epoch": 0.09656150301311592, "grad_norm": 3.3401327341976668, "learning_rate": 9.884460626070634e-06, "loss": 0.4802, "step": 1362 }, { "epoch": 0.0966323998582063, "grad_norm": 5.733632307334381, "learning_rate": 9.884215097877378e-06, "loss": 0.4676, "step": 1363 }, { "epoch": 0.0967032967032967, "grad_norm": 4.178090747169507, "learning_rate": 9.883969312135731e-06, "loss": 0.4782, "step": 1364 }, { "epoch": 0.0967741935483871, "grad_norm": 1.0766861397987282, "learning_rate": 9.883723268858655e-06, "loss": 0.688, "step": 1365 }, { "epoch": 0.0968450903934775, "grad_norm": 3.7552157367988945, "learning_rate": 9.883476968059124e-06, "loss": 0.4382, "step": 1366 }, { "epoch": 0.09691598723856788, "grad_norm": 2.161762388446235, "learning_rate": 9.883230409750124e-06, "loss": 0.4298, "step": 1367 }, { "epoch": 0.09698688408365828, "grad_norm": 3.042901315721617, "learning_rate": 9.882983593944658e-06, "loss": 0.4261, "step": 1368 }, { "epoch": 0.09705778092874867, "grad_norm": 2.279886047645026, "learning_rate": 9.88273652065574e-06, "loss": 0.483, "step": 1369 }, { "epoch": 0.09712867777383906, "grad_norm": 2.266435188765416, "learning_rate": 9.882489189896398e-06, "loss": 0.4368, "step": 1370 }, { "epoch": 0.09719957461892946, "grad_norm": 4.931233725782129, "learning_rate": 9.882241601679677e-06, "loss": 0.44, "step": 1371 }, { "epoch": 0.09727047146401985, "grad_norm": 2.162981749489282, "learning_rate": 9.881993756018626e-06, "loss": 0.4526, "step": 1372 }, { "epoch": 0.09734136830911025, "grad_norm": 3.6895152420105517, "learning_rate": 9.88174565292632e-06, "loss": 0.4683, "step": 1373 }, { "epoch": 0.09741226515420064, "grad_norm": 3.3030859096074656, "learning_rate": 9.88149729241584e-06, "loss": 0.4723, "step": 1374 }, { "epoch": 0.09748316199929104, "grad_norm": 2.589761392892016, "learning_rate": 9.881248674500282e-06, "loss": 0.4708, "step": 1375 }, { "epoch": 0.09755405884438142, "grad_norm": 1.8881175001104435, "learning_rate": 9.880999799192753e-06, "loss": 0.4241, "step": 1376 }, { "epoch": 0.09762495568947181, "grad_norm": 4.052663435482994, "learning_rate": 9.88075066650638e-06, "loss": 0.4939, "step": 1377 }, { "epoch": 0.09769585253456221, "grad_norm": 2.675937115323414, "learning_rate": 9.880501276454302e-06, "loss": 0.4747, "step": 1378 }, { "epoch": 0.0977667493796526, "grad_norm": 3.0234164768547207, "learning_rate": 9.880251629049664e-06, "loss": 0.4389, "step": 1379 }, { "epoch": 0.097837646224743, "grad_norm": 4.150102117217636, "learning_rate": 9.880001724305632e-06, "loss": 0.474, "step": 1380 }, { "epoch": 0.09790854306983339, "grad_norm": 3.3322491329989945, "learning_rate": 9.879751562235387e-06, "loss": 0.4646, "step": 1381 }, { "epoch": 0.09797943991492379, "grad_norm": 2.7654429866137034, "learning_rate": 9.879501142852115e-06, "loss": 0.4589, "step": 1382 }, { "epoch": 0.09805033676001418, "grad_norm": 1.8674686890757397, "learning_rate": 9.879250466169026e-06, "loss": 0.4129, "step": 1383 }, { "epoch": 0.09812123360510458, "grad_norm": 4.4928135657379835, "learning_rate": 9.878999532199334e-06, "loss": 0.4905, "step": 1384 }, { "epoch": 0.09819213045019497, "grad_norm": 2.7706531825574805, "learning_rate": 9.878748340956273e-06, "loss": 0.4538, "step": 1385 }, { "epoch": 0.09826302729528535, "grad_norm": 3.3230479878179238, "learning_rate": 9.87849689245309e-06, "loss": 0.4864, "step": 1386 }, { "epoch": 0.09833392414037576, "grad_norm": 2.7188176071935115, "learning_rate": 9.878245186703039e-06, "loss": 0.494, "step": 1387 }, { "epoch": 0.09840482098546614, "grad_norm": 7.699939273086203, "learning_rate": 9.877993223719398e-06, "loss": 0.4693, "step": 1388 }, { "epoch": 0.09847571783055654, "grad_norm": 4.542419285208761, "learning_rate": 9.87774100351545e-06, "loss": 0.4516, "step": 1389 }, { "epoch": 0.09854661467564693, "grad_norm": 2.6973002898014204, "learning_rate": 9.877488526104497e-06, "loss": 0.4436, "step": 1390 }, { "epoch": 0.09861751152073733, "grad_norm": 3.963208364247192, "learning_rate": 9.87723579149985e-06, "loss": 0.502, "step": 1391 }, { "epoch": 0.09868840836582772, "grad_norm": 9.98825534891498, "learning_rate": 9.876982799714837e-06, "loss": 0.4091, "step": 1392 }, { "epoch": 0.09875930521091811, "grad_norm": 2.825692695024001, "learning_rate": 9.8767295507628e-06, "loss": 0.4322, "step": 1393 }, { "epoch": 0.09883020205600851, "grad_norm": 2.9328447444407955, "learning_rate": 9.87647604465709e-06, "loss": 0.4417, "step": 1394 }, { "epoch": 0.0989010989010989, "grad_norm": 3.0951662414384833, "learning_rate": 9.876222281411076e-06, "loss": 0.4462, "step": 1395 }, { "epoch": 0.0989719957461893, "grad_norm": 3.849097479737599, "learning_rate": 9.875968261038142e-06, "loss": 0.4028, "step": 1396 }, { "epoch": 0.09904289259127969, "grad_norm": 9.326297291490368, "learning_rate": 9.875713983551678e-06, "loss": 0.4108, "step": 1397 }, { "epoch": 0.09911378943637009, "grad_norm": 2.982815148044112, "learning_rate": 9.875459448965094e-06, "loss": 0.4506, "step": 1398 }, { "epoch": 0.09918468628146047, "grad_norm": 2.6225846152472125, "learning_rate": 9.875204657291813e-06, "loss": 0.4485, "step": 1399 }, { "epoch": 0.09925558312655088, "grad_norm": 2.3662180426553787, "learning_rate": 9.874949608545268e-06, "loss": 0.4332, "step": 1400 }, { "epoch": 0.09932647997164126, "grad_norm": 2.8071826797228256, "learning_rate": 9.87469430273891e-06, "loss": 0.4847, "step": 1401 }, { "epoch": 0.09939737681673165, "grad_norm": 2.7268829538147332, "learning_rate": 9.8744387398862e-06, "loss": 0.4431, "step": 1402 }, { "epoch": 0.09946827366182205, "grad_norm": 2.7604236745366997, "learning_rate": 9.874182920000616e-06, "loss": 0.4423, "step": 1403 }, { "epoch": 0.09953917050691244, "grad_norm": 2.071871594094094, "learning_rate": 9.873926843095645e-06, "loss": 0.4887, "step": 1404 }, { "epoch": 0.09961006735200284, "grad_norm": 3.1447257431535345, "learning_rate": 9.873670509184793e-06, "loss": 0.5328, "step": 1405 }, { "epoch": 0.09968096419709323, "grad_norm": 3.9874961595021423, "learning_rate": 9.873413918281574e-06, "loss": 0.4916, "step": 1406 }, { "epoch": 0.09975186104218363, "grad_norm": 5.424403911591371, "learning_rate": 9.87315707039952e-06, "loss": 0.4644, "step": 1407 }, { "epoch": 0.09982275788727402, "grad_norm": 24.791863131836784, "learning_rate": 9.872899965552173e-06, "loss": 0.4171, "step": 1408 }, { "epoch": 0.0998936547323644, "grad_norm": 2.6067241359953415, "learning_rate": 9.872642603753093e-06, "loss": 0.5274, "step": 1409 }, { "epoch": 0.0999645515774548, "grad_norm": 3.2129724708929532, "learning_rate": 9.872384985015848e-06, "loss": 0.4884, "step": 1410 }, { "epoch": 0.10003544842254519, "grad_norm": 3.5113144233133076, "learning_rate": 9.872127109354026e-06, "loss": 0.469, "step": 1411 }, { "epoch": 0.1001063452676356, "grad_norm": 2.27603366291603, "learning_rate": 9.87186897678122e-06, "loss": 0.4256, "step": 1412 }, { "epoch": 0.10017724211272598, "grad_norm": 1.913121535242964, "learning_rate": 9.871610587311046e-06, "loss": 0.4451, "step": 1413 }, { "epoch": 0.10024813895781638, "grad_norm": 3.222574850978485, "learning_rate": 9.871351940957127e-06, "loss": 0.4628, "step": 1414 }, { "epoch": 0.10031903580290677, "grad_norm": 3.6869876543646063, "learning_rate": 9.871093037733102e-06, "loss": 0.4876, "step": 1415 }, { "epoch": 0.10038993264799717, "grad_norm": 3.8145669039533696, "learning_rate": 9.870833877652623e-06, "loss": 0.4855, "step": 1416 }, { "epoch": 0.10046082949308756, "grad_norm": 4.756864957144634, "learning_rate": 9.870574460729356e-06, "loss": 0.4492, "step": 1417 }, { "epoch": 0.10053172633817795, "grad_norm": 3.397793564513188, "learning_rate": 9.87031478697698e-06, "loss": 0.4779, "step": 1418 }, { "epoch": 0.10060262318326835, "grad_norm": 5.5617274202099685, "learning_rate": 9.870054856409188e-06, "loss": 0.4155, "step": 1419 }, { "epoch": 0.10067352002835873, "grad_norm": 5.885479766368831, "learning_rate": 9.869794669039687e-06, "loss": 0.4704, "step": 1420 }, { "epoch": 0.10074441687344914, "grad_norm": 4.011825005648419, "learning_rate": 9.869534224882196e-06, "loss": 0.477, "step": 1421 }, { "epoch": 0.10081531371853952, "grad_norm": 3.2851198223982747, "learning_rate": 9.86927352395045e-06, "loss": 0.4154, "step": 1422 }, { "epoch": 0.10088621056362992, "grad_norm": 2.33187844721757, "learning_rate": 9.86901256625819e-06, "loss": 0.4401, "step": 1423 }, { "epoch": 0.10095710740872031, "grad_norm": 5.199346912770292, "learning_rate": 9.868751351819186e-06, "loss": 0.4733, "step": 1424 }, { "epoch": 0.1010280042538107, "grad_norm": 1.9835064102202076, "learning_rate": 9.868489880647205e-06, "loss": 0.4239, "step": 1425 }, { "epoch": 0.1010989010989011, "grad_norm": 2.5665838724705807, "learning_rate": 9.868228152756036e-06, "loss": 0.4368, "step": 1426 }, { "epoch": 0.10116979794399149, "grad_norm": 2.878001384473534, "learning_rate": 9.86796616815948e-06, "loss": 0.4301, "step": 1427 }, { "epoch": 0.10124069478908189, "grad_norm": 3.1435419938597704, "learning_rate": 9.867703926871354e-06, "loss": 0.446, "step": 1428 }, { "epoch": 0.10131159163417228, "grad_norm": 4.027670237784497, "learning_rate": 9.867441428905484e-06, "loss": 0.4831, "step": 1429 }, { "epoch": 0.10138248847926268, "grad_norm": 3.587029351648749, "learning_rate": 9.867178674275713e-06, "loss": 0.4884, "step": 1430 }, { "epoch": 0.10145338532435307, "grad_norm": 10.079853650946815, "learning_rate": 9.866915662995895e-06, "loss": 0.4648, "step": 1431 }, { "epoch": 0.10152428216944347, "grad_norm": 2.197220563917957, "learning_rate": 9.866652395079899e-06, "loss": 0.4551, "step": 1432 }, { "epoch": 0.10159517901453385, "grad_norm": 2.4343488695630837, "learning_rate": 9.86638887054161e-06, "loss": 0.4425, "step": 1433 }, { "epoch": 0.10166607585962424, "grad_norm": 2.760873309291183, "learning_rate": 9.866125089394919e-06, "loss": 0.4554, "step": 1434 }, { "epoch": 0.10173697270471464, "grad_norm": 6.581710712811045, "learning_rate": 9.865861051653738e-06, "loss": 0.4811, "step": 1435 }, { "epoch": 0.10180786954980503, "grad_norm": 2.1841597711171468, "learning_rate": 9.86559675733199e-06, "loss": 0.4697, "step": 1436 }, { "epoch": 0.10187876639489543, "grad_norm": 2.706388436242509, "learning_rate": 9.865332206443613e-06, "loss": 0.3955, "step": 1437 }, { "epoch": 0.10194966323998582, "grad_norm": 2.750177550867352, "learning_rate": 9.865067399002552e-06, "loss": 0.4625, "step": 1438 }, { "epoch": 0.10202056008507622, "grad_norm": 3.4754663311395224, "learning_rate": 9.864802335022776e-06, "loss": 0.4183, "step": 1439 }, { "epoch": 0.10209145693016661, "grad_norm": 3.791782001460879, "learning_rate": 9.864537014518261e-06, "loss": 0.4593, "step": 1440 }, { "epoch": 0.102162353775257, "grad_norm": 5.204694731461543, "learning_rate": 9.864271437502994e-06, "loss": 0.4897, "step": 1441 }, { "epoch": 0.1022332506203474, "grad_norm": 7.6026373690145705, "learning_rate": 9.864005603990981e-06, "loss": 0.4312, "step": 1442 }, { "epoch": 0.10230414746543778, "grad_norm": 2.503229849164424, "learning_rate": 9.863739513996242e-06, "loss": 0.4394, "step": 1443 }, { "epoch": 0.10237504431052818, "grad_norm": 1.1521428216532328, "learning_rate": 9.863473167532804e-06, "loss": 0.6731, "step": 1444 }, { "epoch": 0.10244594115561857, "grad_norm": 2.857728358781883, "learning_rate": 9.863206564614716e-06, "loss": 0.4755, "step": 1445 }, { "epoch": 0.10251683800070897, "grad_norm": 44.72553736992587, "learning_rate": 9.862939705256032e-06, "loss": 0.4865, "step": 1446 }, { "epoch": 0.10258773484579936, "grad_norm": 2.966290052971148, "learning_rate": 9.862672589470826e-06, "loss": 0.5017, "step": 1447 }, { "epoch": 0.10265863169088976, "grad_norm": 3.1852571545342787, "learning_rate": 9.862405217273183e-06, "loss": 0.5152, "step": 1448 }, { "epoch": 0.10272952853598015, "grad_norm": 2.4503845685105934, "learning_rate": 9.862137588677202e-06, "loss": 0.5109, "step": 1449 }, { "epoch": 0.10280042538107054, "grad_norm": 2.186602678677607, "learning_rate": 9.861869703696995e-06, "loss": 0.467, "step": 1450 }, { "epoch": 0.10287132222616094, "grad_norm": 1.8904889929925577, "learning_rate": 9.861601562346688e-06, "loss": 0.4372, "step": 1451 }, { "epoch": 0.10294221907125133, "grad_norm": 2.904869085811955, "learning_rate": 9.861333164640418e-06, "loss": 0.4099, "step": 1452 }, { "epoch": 0.10301311591634173, "grad_norm": 2.5689576181805314, "learning_rate": 9.861064510592341e-06, "loss": 0.4679, "step": 1453 }, { "epoch": 0.10308401276143211, "grad_norm": 4.441811485655783, "learning_rate": 9.860795600216622e-06, "loss": 0.4715, "step": 1454 }, { "epoch": 0.10315490960652252, "grad_norm": 2.7137064472037844, "learning_rate": 9.86052643352744e-06, "loss": 0.4425, "step": 1455 }, { "epoch": 0.1032258064516129, "grad_norm": 2.8882956685940786, "learning_rate": 9.860257010538992e-06, "loss": 0.4752, "step": 1456 }, { "epoch": 0.10329670329670329, "grad_norm": 3.2066824877551943, "learning_rate": 9.85998733126548e-06, "loss": 0.4733, "step": 1457 }, { "epoch": 0.10336760014179369, "grad_norm": 3.0859469992370228, "learning_rate": 9.859717395721126e-06, "loss": 0.5008, "step": 1458 }, { "epoch": 0.10343849698688408, "grad_norm": 2.135355801786162, "learning_rate": 9.859447203920164e-06, "loss": 0.452, "step": 1459 }, { "epoch": 0.10350939383197448, "grad_norm": 7.747993777325768, "learning_rate": 9.859176755876845e-06, "loss": 0.4408, "step": 1460 }, { "epoch": 0.10358029067706487, "grad_norm": 2.309140941155754, "learning_rate": 9.858906051605426e-06, "loss": 0.4726, "step": 1461 }, { "epoch": 0.10365118752215527, "grad_norm": 2.9393602148544384, "learning_rate": 9.85863509112018e-06, "loss": 0.4229, "step": 1462 }, { "epoch": 0.10372208436724566, "grad_norm": 1.5503670706541492, "learning_rate": 9.858363874435399e-06, "loss": 0.4748, "step": 1463 }, { "epoch": 0.10379298121233606, "grad_norm": 2.2134604933680664, "learning_rate": 9.858092401565382e-06, "loss": 0.4482, "step": 1464 }, { "epoch": 0.10386387805742645, "grad_norm": 4.280831658162354, "learning_rate": 9.857820672524445e-06, "loss": 0.4069, "step": 1465 }, { "epoch": 0.10393477490251683, "grad_norm": 13.27608750271994, "learning_rate": 9.857548687326916e-06, "loss": 0.4565, "step": 1466 }, { "epoch": 0.10400567174760723, "grad_norm": 3.828936470996531, "learning_rate": 9.857276445987137e-06, "loss": 0.4546, "step": 1467 }, { "epoch": 0.10407656859269762, "grad_norm": 3.626289172508771, "learning_rate": 9.857003948519463e-06, "loss": 0.4164, "step": 1468 }, { "epoch": 0.10414746543778802, "grad_norm": 6.695119895740298, "learning_rate": 9.856731194938265e-06, "loss": 0.4075, "step": 1469 }, { "epoch": 0.10421836228287841, "grad_norm": 3.1151360847380234, "learning_rate": 9.856458185257925e-06, "loss": 0.4474, "step": 1470 }, { "epoch": 0.10428925912796881, "grad_norm": 3.318103421714098, "learning_rate": 9.856184919492837e-06, "loss": 0.4425, "step": 1471 }, { "epoch": 0.1043601559730592, "grad_norm": 4.414522500754744, "learning_rate": 9.85591139765741e-06, "loss": 0.4281, "step": 1472 }, { "epoch": 0.10443105281814959, "grad_norm": 2.936435647991984, "learning_rate": 9.85563761976607e-06, "loss": 0.4682, "step": 1473 }, { "epoch": 0.10450194966323999, "grad_norm": 3.769625683027973, "learning_rate": 9.855363585833253e-06, "loss": 0.4802, "step": 1474 }, { "epoch": 0.10457284650833037, "grad_norm": 3.170131071010728, "learning_rate": 9.855089295873406e-06, "loss": 0.4385, "step": 1475 }, { "epoch": 0.10464374335342078, "grad_norm": 2.9494775137661255, "learning_rate": 9.854814749900997e-06, "loss": 0.4282, "step": 1476 }, { "epoch": 0.10471464019851116, "grad_norm": 6.545696999408565, "learning_rate": 9.8545399479305e-06, "loss": 0.4436, "step": 1477 }, { "epoch": 0.10478553704360156, "grad_norm": 3.848985917198309, "learning_rate": 9.854264889976404e-06, "loss": 0.4981, "step": 1478 }, { "epoch": 0.10485643388869195, "grad_norm": 8.398604289393173, "learning_rate": 9.853989576053219e-06, "loss": 0.4242, "step": 1479 }, { "epoch": 0.10492733073378235, "grad_norm": 2.2893831110496783, "learning_rate": 9.853714006175456e-06, "loss": 0.4363, "step": 1480 }, { "epoch": 0.10499822757887274, "grad_norm": 2.5105595847174262, "learning_rate": 9.85343818035765e-06, "loss": 0.4627, "step": 1481 }, { "epoch": 0.10506912442396313, "grad_norm": 5.114542769365885, "learning_rate": 9.853162098614344e-06, "loss": 0.421, "step": 1482 }, { "epoch": 0.10514002126905353, "grad_norm": 2.142133521000662, "learning_rate": 9.852885760960096e-06, "loss": 0.4275, "step": 1483 }, { "epoch": 0.10521091811414392, "grad_norm": 2.7836479973127326, "learning_rate": 9.85260916740948e-06, "loss": 0.4695, "step": 1484 }, { "epoch": 0.10528181495923432, "grad_norm": 3.4137892346766345, "learning_rate": 9.852332317977076e-06, "loss": 0.4718, "step": 1485 }, { "epoch": 0.1053527118043247, "grad_norm": 2.3949317178870038, "learning_rate": 9.852055212677485e-06, "loss": 0.4759, "step": 1486 }, { "epoch": 0.10542360864941511, "grad_norm": 4.505659304082982, "learning_rate": 9.851777851525322e-06, "loss": 0.4725, "step": 1487 }, { "epoch": 0.1054945054945055, "grad_norm": 2.7019984677982385, "learning_rate": 9.851500234535207e-06, "loss": 0.4293, "step": 1488 }, { "epoch": 0.10556540233959588, "grad_norm": 1.7756398803024893, "learning_rate": 9.851222361721783e-06, "loss": 0.402, "step": 1489 }, { "epoch": 0.10563629918468628, "grad_norm": 6.88284204196329, "learning_rate": 9.8509442330997e-06, "loss": 0.4961, "step": 1490 }, { "epoch": 0.10570719602977667, "grad_norm": 1.2779997076400382, "learning_rate": 9.850665848683628e-06, "loss": 0.6527, "step": 1491 }, { "epoch": 0.10577809287486707, "grad_norm": 2.1108683899842613, "learning_rate": 9.85038720848824e-06, "loss": 0.4136, "step": 1492 }, { "epoch": 0.10584898971995746, "grad_norm": 4.146803000776314, "learning_rate": 9.850108312528233e-06, "loss": 0.4676, "step": 1493 }, { "epoch": 0.10591988656504786, "grad_norm": 2.950558428098633, "learning_rate": 9.849829160818313e-06, "loss": 0.4617, "step": 1494 }, { "epoch": 0.10599078341013825, "grad_norm": 2.099816006852762, "learning_rate": 9.849549753373201e-06, "loss": 0.4207, "step": 1495 }, { "epoch": 0.10606168025522865, "grad_norm": 3.3316546883366995, "learning_rate": 9.849270090207629e-06, "loss": 0.4563, "step": 1496 }, { "epoch": 0.10613257710031904, "grad_norm": 8.03874562038121, "learning_rate": 9.848990171336342e-06, "loss": 0.4951, "step": 1497 }, { "epoch": 0.10620347394540942, "grad_norm": 3.1269991655373044, "learning_rate": 9.848709996774104e-06, "loss": 0.4779, "step": 1498 }, { "epoch": 0.10627437079049983, "grad_norm": 2.914707319272395, "learning_rate": 9.848429566535686e-06, "loss": 0.4801, "step": 1499 }, { "epoch": 0.10634526763559021, "grad_norm": 3.740831380594115, "learning_rate": 9.848148880635875e-06, "loss": 0.4496, "step": 1500 }, { "epoch": 0.10641616448068061, "grad_norm": 2.051300634788304, "learning_rate": 9.847867939089475e-06, "loss": 0.4532, "step": 1501 }, { "epoch": 0.106487061325771, "grad_norm": 3.94342971187442, "learning_rate": 9.847586741911299e-06, "loss": 0.4553, "step": 1502 }, { "epoch": 0.1065579581708614, "grad_norm": 4.826871187805627, "learning_rate": 9.847305289116174e-06, "loss": 0.4492, "step": 1503 }, { "epoch": 0.10662885501595179, "grad_norm": 2.84778884730146, "learning_rate": 9.84702358071894e-06, "loss": 0.4715, "step": 1504 }, { "epoch": 0.10669975186104218, "grad_norm": 4.309879340495927, "learning_rate": 9.846741616734452e-06, "loss": 0.4614, "step": 1505 }, { "epoch": 0.10677064870613258, "grad_norm": 5.733491690646196, "learning_rate": 9.846459397177582e-06, "loss": 0.4897, "step": 1506 }, { "epoch": 0.10684154555122297, "grad_norm": 2.5364489355673965, "learning_rate": 9.846176922063207e-06, "loss": 0.4518, "step": 1507 }, { "epoch": 0.10691244239631337, "grad_norm": 3.066598960463324, "learning_rate": 9.845894191406224e-06, "loss": 0.4506, "step": 1508 }, { "epoch": 0.10698333924140375, "grad_norm": 2.4600542337219258, "learning_rate": 9.84561120522154e-06, "loss": 0.4731, "step": 1509 }, { "epoch": 0.10705423608649416, "grad_norm": 4.363048245351467, "learning_rate": 9.845327963524083e-06, "loss": 0.4857, "step": 1510 }, { "epoch": 0.10712513293158454, "grad_norm": 2.8532564269786587, "learning_rate": 9.845044466328782e-06, "loss": 0.4916, "step": 1511 }, { "epoch": 0.10719602977667494, "grad_norm": 2.2898969209706546, "learning_rate": 9.844760713650588e-06, "loss": 0.478, "step": 1512 }, { "epoch": 0.10726692662176533, "grad_norm": 2.606837397153784, "learning_rate": 9.844476705504465e-06, "loss": 0.4741, "step": 1513 }, { "epoch": 0.10733782346685572, "grad_norm": 3.111665286221622, "learning_rate": 9.844192441905384e-06, "loss": 0.4537, "step": 1514 }, { "epoch": 0.10740872031194612, "grad_norm": 5.488985259121631, "learning_rate": 9.843907922868342e-06, "loss": 0.5209, "step": 1515 }, { "epoch": 0.10747961715703651, "grad_norm": 4.383567333366404, "learning_rate": 9.843623148408337e-06, "loss": 0.5032, "step": 1516 }, { "epoch": 0.10755051400212691, "grad_norm": 5.416580980124831, "learning_rate": 9.843338118540386e-06, "loss": 0.5063, "step": 1517 }, { "epoch": 0.1076214108472173, "grad_norm": 2.8091116869456365, "learning_rate": 9.84305283327952e-06, "loss": 0.4518, "step": 1518 }, { "epoch": 0.1076923076923077, "grad_norm": 3.19674379638348, "learning_rate": 9.84276729264078e-06, "loss": 0.524, "step": 1519 }, { "epoch": 0.10776320453739809, "grad_norm": 3.570460984437687, "learning_rate": 9.842481496639226e-06, "loss": 0.3873, "step": 1520 }, { "epoch": 0.10783410138248847, "grad_norm": 2.554462243322921, "learning_rate": 9.842195445289927e-06, "loss": 0.4525, "step": 1521 }, { "epoch": 0.10790499822757887, "grad_norm": 2.6996421659061856, "learning_rate": 9.841909138607964e-06, "loss": 0.4964, "step": 1522 }, { "epoch": 0.10797589507266926, "grad_norm": 1.193621538900233, "learning_rate": 9.841622576608437e-06, "loss": 0.6857, "step": 1523 }, { "epoch": 0.10804679191775966, "grad_norm": 3.403275811302415, "learning_rate": 9.841335759306457e-06, "loss": 0.509, "step": 1524 }, { "epoch": 0.10811768876285005, "grad_norm": 2.357540203813119, "learning_rate": 9.841048686717146e-06, "loss": 0.434, "step": 1525 }, { "epoch": 0.10818858560794045, "grad_norm": 4.1954584173506415, "learning_rate": 9.840761358855644e-06, "loss": 0.4475, "step": 1526 }, { "epoch": 0.10825948245303084, "grad_norm": 2.390321868046923, "learning_rate": 9.840473775737101e-06, "loss": 0.4471, "step": 1527 }, { "epoch": 0.10833037929812124, "grad_norm": 2.8934940269547718, "learning_rate": 9.84018593737668e-06, "loss": 0.5096, "step": 1528 }, { "epoch": 0.10840127614321163, "grad_norm": 3.390033998597313, "learning_rate": 9.839897843789562e-06, "loss": 0.4535, "step": 1529 }, { "epoch": 0.10847217298830202, "grad_norm": 3.011165846031431, "learning_rate": 9.839609494990936e-06, "loss": 0.4323, "step": 1530 }, { "epoch": 0.10854306983339242, "grad_norm": 5.858626151262423, "learning_rate": 9.839320890996008e-06, "loss": 0.4929, "step": 1531 }, { "epoch": 0.1086139666784828, "grad_norm": 2.708523428698348, "learning_rate": 9.839032031819993e-06, "loss": 0.5185, "step": 1532 }, { "epoch": 0.1086848635235732, "grad_norm": 2.9603274835862115, "learning_rate": 9.838742917478127e-06, "loss": 0.4476, "step": 1533 }, { "epoch": 0.10875576036866359, "grad_norm": 2.2093820360253114, "learning_rate": 9.838453547985653e-06, "loss": 0.3858, "step": 1534 }, { "epoch": 0.108826657213754, "grad_norm": 3.226737575703645, "learning_rate": 9.838163923357831e-06, "loss": 0.443, "step": 1535 }, { "epoch": 0.10889755405884438, "grad_norm": 1.775106678414576, "learning_rate": 9.837874043609933e-06, "loss": 0.4889, "step": 1536 }, { "epoch": 0.10896845090393477, "grad_norm": 4.046526799944191, "learning_rate": 9.837583908757245e-06, "loss": 0.465, "step": 1537 }, { "epoch": 0.10903934774902517, "grad_norm": 2.7889876444457857, "learning_rate": 9.837293518815064e-06, "loss": 0.4519, "step": 1538 }, { "epoch": 0.10911024459411556, "grad_norm": 5.478351416577967, "learning_rate": 9.837002873798702e-06, "loss": 0.5072, "step": 1539 }, { "epoch": 0.10918114143920596, "grad_norm": 3.925209233441258, "learning_rate": 9.83671197372349e-06, "loss": 0.4968, "step": 1540 }, { "epoch": 0.10925203828429635, "grad_norm": 3.1610454552501337, "learning_rate": 9.836420818604762e-06, "loss": 0.4943, "step": 1541 }, { "epoch": 0.10932293512938675, "grad_norm": 3.1836664578418814, "learning_rate": 9.836129408457871e-06, "loss": 0.4547, "step": 1542 }, { "epoch": 0.10939383197447713, "grad_norm": 2.5291123986419297, "learning_rate": 9.835837743298187e-06, "loss": 0.4679, "step": 1543 }, { "epoch": 0.10946472881956754, "grad_norm": 2.3676452595059003, "learning_rate": 9.835545823141088e-06, "loss": 0.471, "step": 1544 }, { "epoch": 0.10953562566465792, "grad_norm": 2.2904099115676435, "learning_rate": 9.835253648001967e-06, "loss": 0.4463, "step": 1545 }, { "epoch": 0.10960652250974831, "grad_norm": 2.2062865451366607, "learning_rate": 9.83496121789623e-06, "loss": 0.488, "step": 1546 }, { "epoch": 0.10967741935483871, "grad_norm": 2.968740546233121, "learning_rate": 9.834668532839298e-06, "loss": 0.4583, "step": 1547 }, { "epoch": 0.1097483161999291, "grad_norm": 6.363381229943704, "learning_rate": 9.834375592846603e-06, "loss": 0.5566, "step": 1548 }, { "epoch": 0.1098192130450195, "grad_norm": 4.349444691951875, "learning_rate": 9.834082397933595e-06, "loss": 0.4071, "step": 1549 }, { "epoch": 0.10989010989010989, "grad_norm": 3.8072580081487213, "learning_rate": 9.833788948115731e-06, "loss": 0.4358, "step": 1550 }, { "epoch": 0.10996100673520029, "grad_norm": 4.095298562341299, "learning_rate": 9.833495243408486e-06, "loss": 0.5045, "step": 1551 }, { "epoch": 0.11003190358029068, "grad_norm": 4.5766254289686845, "learning_rate": 9.83320128382735e-06, "loss": 0.465, "step": 1552 }, { "epoch": 0.11010280042538106, "grad_norm": 2.9936974083611534, "learning_rate": 9.832907069387819e-06, "loss": 0.4336, "step": 1553 }, { "epoch": 0.11017369727047147, "grad_norm": 3.8572249269761465, "learning_rate": 9.832612600105409e-06, "loss": 0.4445, "step": 1554 }, { "epoch": 0.11024459411556185, "grad_norm": 2.983804534274985, "learning_rate": 9.832317875995649e-06, "loss": 0.4938, "step": 1555 }, { "epoch": 0.11031549096065225, "grad_norm": 6.039626920079816, "learning_rate": 9.83202289707408e-06, "loss": 0.4594, "step": 1556 }, { "epoch": 0.11038638780574264, "grad_norm": 3.201537558863724, "learning_rate": 9.831727663356255e-06, "loss": 0.435, "step": 1557 }, { "epoch": 0.11045728465083304, "grad_norm": 3.2181629614141927, "learning_rate": 9.83143217485774e-06, "loss": 0.4931, "step": 1558 }, { "epoch": 0.11052818149592343, "grad_norm": 3.3262069240609278, "learning_rate": 9.831136431594122e-06, "loss": 0.4582, "step": 1559 }, { "epoch": 0.11059907834101383, "grad_norm": 2.5716885682084323, "learning_rate": 9.830840433580992e-06, "loss": 0.4154, "step": 1560 }, { "epoch": 0.11066997518610422, "grad_norm": 2.404654591710592, "learning_rate": 9.830544180833957e-06, "loss": 0.4272, "step": 1561 }, { "epoch": 0.1107408720311946, "grad_norm": 3.8652478647349535, "learning_rate": 9.830247673368643e-06, "loss": 0.4132, "step": 1562 }, { "epoch": 0.11081176887628501, "grad_norm": 2.44797537799076, "learning_rate": 9.82995091120068e-06, "loss": 0.4313, "step": 1563 }, { "epoch": 0.1108826657213754, "grad_norm": 3.2852190937606154, "learning_rate": 9.82965389434572e-06, "loss": 0.5327, "step": 1564 }, { "epoch": 0.1109535625664658, "grad_norm": 2.70703857726617, "learning_rate": 9.829356622819424e-06, "loss": 0.5178, "step": 1565 }, { "epoch": 0.11102445941155618, "grad_norm": 3.1739046834099733, "learning_rate": 9.829059096637465e-06, "loss": 0.4285, "step": 1566 }, { "epoch": 0.11109535625664659, "grad_norm": 1.9291378890651087, "learning_rate": 9.828761315815536e-06, "loss": 0.3826, "step": 1567 }, { "epoch": 0.11116625310173697, "grad_norm": 3.0986544457435605, "learning_rate": 9.828463280369338e-06, "loss": 0.4571, "step": 1568 }, { "epoch": 0.11123714994682736, "grad_norm": 7.761958637650186, "learning_rate": 9.828164990314584e-06, "loss": 0.4654, "step": 1569 }, { "epoch": 0.11130804679191776, "grad_norm": 5.11547051360499, "learning_rate": 9.827866445667005e-06, "loss": 0.4774, "step": 1570 }, { "epoch": 0.11137894363700815, "grad_norm": 6.414573933391783, "learning_rate": 9.827567646442345e-06, "loss": 0.4307, "step": 1571 }, { "epoch": 0.11144984048209855, "grad_norm": 4.409179648198286, "learning_rate": 9.827268592656359e-06, "loss": 0.4664, "step": 1572 }, { "epoch": 0.11152073732718894, "grad_norm": 3.234356184058177, "learning_rate": 9.826969284324813e-06, "loss": 0.4522, "step": 1573 }, { "epoch": 0.11159163417227934, "grad_norm": 5.192753314142492, "learning_rate": 9.826669721463493e-06, "loss": 0.4145, "step": 1574 }, { "epoch": 0.11166253101736973, "grad_norm": 3.769935929947971, "learning_rate": 9.826369904088195e-06, "loss": 0.4658, "step": 1575 }, { "epoch": 0.11173342786246013, "grad_norm": 3.165298174014974, "learning_rate": 9.826069832214728e-06, "loss": 0.4823, "step": 1576 }, { "epoch": 0.11180432470755051, "grad_norm": 3.8484016757414303, "learning_rate": 9.825769505858915e-06, "loss": 0.4928, "step": 1577 }, { "epoch": 0.1118752215526409, "grad_norm": 2.884810024117639, "learning_rate": 9.825468925036593e-06, "loss": 0.4168, "step": 1578 }, { "epoch": 0.1119461183977313, "grad_norm": 2.0710311104043875, "learning_rate": 9.825168089763611e-06, "loss": 0.4595, "step": 1579 }, { "epoch": 0.11201701524282169, "grad_norm": 2.5434661386561412, "learning_rate": 9.824867000055832e-06, "loss": 0.4874, "step": 1580 }, { "epoch": 0.11208791208791209, "grad_norm": 5.089591800903534, "learning_rate": 9.824565655929136e-06, "loss": 0.458, "step": 1581 }, { "epoch": 0.11215880893300248, "grad_norm": 3.029018112186499, "learning_rate": 9.824264057399408e-06, "loss": 0.401, "step": 1582 }, { "epoch": 0.11222970577809288, "grad_norm": 2.4690027963415044, "learning_rate": 9.823962204482556e-06, "loss": 0.4646, "step": 1583 }, { "epoch": 0.11230060262318327, "grad_norm": 2.936694595847078, "learning_rate": 9.823660097194494e-06, "loss": 0.4571, "step": 1584 }, { "epoch": 0.11237149946827366, "grad_norm": 2.5337852833330006, "learning_rate": 9.823357735551154e-06, "loss": 0.4389, "step": 1585 }, { "epoch": 0.11244239631336406, "grad_norm": 5.105857029153958, "learning_rate": 9.823055119568478e-06, "loss": 0.4321, "step": 1586 }, { "epoch": 0.11251329315845444, "grad_norm": 27.958473442571844, "learning_rate": 9.822752249262425e-06, "loss": 0.4177, "step": 1587 }, { "epoch": 0.11258419000354485, "grad_norm": 4.595557076511662, "learning_rate": 9.822449124648965e-06, "loss": 0.4653, "step": 1588 }, { "epoch": 0.11265508684863523, "grad_norm": 35.5654096710017, "learning_rate": 9.82214574574408e-06, "loss": 0.4843, "step": 1589 }, { "epoch": 0.11272598369372563, "grad_norm": 4.779803327328721, "learning_rate": 9.821842112563772e-06, "loss": 0.4785, "step": 1590 }, { "epoch": 0.11279688053881602, "grad_norm": 3.2207631588816032, "learning_rate": 9.821538225124047e-06, "loss": 0.4338, "step": 1591 }, { "epoch": 0.11286777738390642, "grad_norm": 3.0656517714253897, "learning_rate": 9.821234083440932e-06, "loss": 0.5095, "step": 1592 }, { "epoch": 0.11293867422899681, "grad_norm": 3.1385347470049636, "learning_rate": 9.820929687530465e-06, "loss": 0.4603, "step": 1593 }, { "epoch": 0.1130095710740872, "grad_norm": 4.4522231257784535, "learning_rate": 9.820625037408696e-06, "loss": 0.4744, "step": 1594 }, { "epoch": 0.1130804679191776, "grad_norm": 6.308856223209965, "learning_rate": 9.820320133091688e-06, "loss": 0.4776, "step": 1595 }, { "epoch": 0.11315136476426799, "grad_norm": 2.7876572616562516, "learning_rate": 9.82001497459552e-06, "loss": 0.4354, "step": 1596 }, { "epoch": 0.11322226160935839, "grad_norm": 4.382066241759971, "learning_rate": 9.819709561936284e-06, "loss": 0.499, "step": 1597 }, { "epoch": 0.11329315845444878, "grad_norm": 4.784817912632191, "learning_rate": 9.819403895130085e-06, "loss": 0.4346, "step": 1598 }, { "epoch": 0.11336405529953918, "grad_norm": 6.631780753028119, "learning_rate": 9.819097974193037e-06, "loss": 0.546, "step": 1599 }, { "epoch": 0.11343495214462956, "grad_norm": 4.155489586030063, "learning_rate": 9.818791799141278e-06, "loss": 0.4106, "step": 1600 }, { "epoch": 0.11350584898971995, "grad_norm": 2.326065942671352, "learning_rate": 9.818485369990947e-06, "loss": 0.4244, "step": 1601 }, { "epoch": 0.11357674583481035, "grad_norm": 2.942573716705073, "learning_rate": 9.818178686758207e-06, "loss": 0.4651, "step": 1602 }, { "epoch": 0.11364764267990074, "grad_norm": 2.791842192665445, "learning_rate": 9.817871749459224e-06, "loss": 0.4653, "step": 1603 }, { "epoch": 0.11371853952499114, "grad_norm": 2.4690616696330228, "learning_rate": 9.817564558110189e-06, "loss": 0.4734, "step": 1604 }, { "epoch": 0.11378943637008153, "grad_norm": 2.180166264244019, "learning_rate": 9.817257112727298e-06, "loss": 0.4035, "step": 1605 }, { "epoch": 0.11386033321517193, "grad_norm": 2.360462067338051, "learning_rate": 9.816949413326762e-06, "loss": 0.4294, "step": 1606 }, { "epoch": 0.11393123006026232, "grad_norm": 2.8176322757642467, "learning_rate": 9.816641459924807e-06, "loss": 0.4589, "step": 1607 }, { "epoch": 0.11400212690535272, "grad_norm": 1.8570751720612264, "learning_rate": 9.81633325253767e-06, "loss": 0.4782, "step": 1608 }, { "epoch": 0.1140730237504431, "grad_norm": 5.283477378915315, "learning_rate": 9.816024791181607e-06, "loss": 0.4137, "step": 1609 }, { "epoch": 0.1141439205955335, "grad_norm": 3.429796958294709, "learning_rate": 9.815716075872878e-06, "loss": 0.49, "step": 1610 }, { "epoch": 0.1142148174406239, "grad_norm": 2.8234109494757096, "learning_rate": 9.815407106627765e-06, "loss": 0.4637, "step": 1611 }, { "epoch": 0.11428571428571428, "grad_norm": 2.9278609820716035, "learning_rate": 9.815097883462562e-06, "loss": 0.4727, "step": 1612 }, { "epoch": 0.11435661113080468, "grad_norm": 2.562739913897779, "learning_rate": 9.814788406393572e-06, "loss": 0.5116, "step": 1613 }, { "epoch": 0.11442750797589507, "grad_norm": 6.890299668880736, "learning_rate": 9.814478675437114e-06, "loss": 0.4253, "step": 1614 }, { "epoch": 0.11449840482098547, "grad_norm": 2.913323308234268, "learning_rate": 9.814168690609519e-06, "loss": 0.4417, "step": 1615 }, { "epoch": 0.11456930166607586, "grad_norm": 1.881242959962745, "learning_rate": 9.813858451927135e-06, "loss": 0.5421, "step": 1616 }, { "epoch": 0.11464019851116625, "grad_norm": 2.378058829488207, "learning_rate": 9.813547959406322e-06, "loss": 0.4836, "step": 1617 }, { "epoch": 0.11471109535625665, "grad_norm": 3.6132734377271403, "learning_rate": 9.813237213063451e-06, "loss": 0.5062, "step": 1618 }, { "epoch": 0.11478199220134704, "grad_norm": 2.4374245214371446, "learning_rate": 9.812926212914908e-06, "loss": 0.4271, "step": 1619 }, { "epoch": 0.11485288904643744, "grad_norm": 4.0364071126202425, "learning_rate": 9.81261495897709e-06, "loss": 0.3716, "step": 1620 }, { "epoch": 0.11492378589152782, "grad_norm": 3.9173424446888925, "learning_rate": 9.812303451266414e-06, "loss": 0.444, "step": 1621 }, { "epoch": 0.11499468273661823, "grad_norm": 3.4535173028643236, "learning_rate": 9.811991689799304e-06, "loss": 0.5019, "step": 1622 }, { "epoch": 0.11506557958170861, "grad_norm": 5.044783486359595, "learning_rate": 9.811679674592198e-06, "loss": 0.4686, "step": 1623 }, { "epoch": 0.11513647642679901, "grad_norm": 2.695183182408353, "learning_rate": 9.811367405661551e-06, "loss": 0.4325, "step": 1624 }, { "epoch": 0.1152073732718894, "grad_norm": 2.8232926612338782, "learning_rate": 9.81105488302383e-06, "loss": 0.5, "step": 1625 }, { "epoch": 0.11527827011697979, "grad_norm": 1.7501396092896782, "learning_rate": 9.810742106695511e-06, "loss": 0.4688, "step": 1626 }, { "epoch": 0.11534916696207019, "grad_norm": 2.926391090176236, "learning_rate": 9.810429076693088e-06, "loss": 0.4616, "step": 1627 }, { "epoch": 0.11542006380716058, "grad_norm": 6.4442349239694945, "learning_rate": 9.810115793033071e-06, "loss": 0.4541, "step": 1628 }, { "epoch": 0.11549096065225098, "grad_norm": 2.943995128229437, "learning_rate": 9.809802255731975e-06, "loss": 0.4708, "step": 1629 }, { "epoch": 0.11556185749734137, "grad_norm": 7.219315317720864, "learning_rate": 9.809488464806335e-06, "loss": 0.4108, "step": 1630 }, { "epoch": 0.11563275434243177, "grad_norm": 2.5790444810594932, "learning_rate": 9.809174420272698e-06, "loss": 0.4721, "step": 1631 }, { "epoch": 0.11570365118752216, "grad_norm": 2.3741161636424226, "learning_rate": 9.80886012214762e-06, "loss": 0.419, "step": 1632 }, { "epoch": 0.11577454803261254, "grad_norm": 2.2260792085455052, "learning_rate": 9.808545570447682e-06, "loss": 0.4322, "step": 1633 }, { "epoch": 0.11584544487770294, "grad_norm": 3.894446393455003, "learning_rate": 9.808230765189464e-06, "loss": 0.4537, "step": 1634 }, { "epoch": 0.11591634172279333, "grad_norm": 3.041243877555194, "learning_rate": 9.807915706389566e-06, "loss": 0.4284, "step": 1635 }, { "epoch": 0.11598723856788373, "grad_norm": 10.41361165029608, "learning_rate": 9.807600394064605e-06, "loss": 0.4756, "step": 1636 }, { "epoch": 0.11605813541297412, "grad_norm": 3.7093741388055728, "learning_rate": 9.807284828231206e-06, "loss": 0.4676, "step": 1637 }, { "epoch": 0.11612903225806452, "grad_norm": 1.1609657708568233, "learning_rate": 9.806969008906009e-06, "loss": 0.6856, "step": 1638 }, { "epoch": 0.11619992910315491, "grad_norm": 4.454744736821673, "learning_rate": 9.806652936105666e-06, "loss": 0.4492, "step": 1639 }, { "epoch": 0.11627082594824531, "grad_norm": 2.882819907935762, "learning_rate": 9.806336609846845e-06, "loss": 0.4827, "step": 1640 }, { "epoch": 0.1163417227933357, "grad_norm": 2.298044631298054, "learning_rate": 9.806020030146227e-06, "loss": 0.3993, "step": 1641 }, { "epoch": 0.11641261963842608, "grad_norm": 2.8241802897373582, "learning_rate": 9.805703197020504e-06, "loss": 0.3927, "step": 1642 }, { "epoch": 0.11648351648351649, "grad_norm": 3.5459810857378953, "learning_rate": 9.805386110486382e-06, "loss": 0.4531, "step": 1643 }, { "epoch": 0.11655441332860687, "grad_norm": 3.8616406452506458, "learning_rate": 9.805068770560585e-06, "loss": 0.5065, "step": 1644 }, { "epoch": 0.11662531017369727, "grad_norm": 1.1002200360082495, "learning_rate": 9.804751177259841e-06, "loss": 0.681, "step": 1645 }, { "epoch": 0.11669620701878766, "grad_norm": 5.010980271639471, "learning_rate": 9.804433330600904e-06, "loss": 0.4892, "step": 1646 }, { "epoch": 0.11676710386387806, "grad_norm": 5.9590224041934015, "learning_rate": 9.804115230600528e-06, "loss": 0.4391, "step": 1647 }, { "epoch": 0.11683800070896845, "grad_norm": 2.1229941147165685, "learning_rate": 9.80379687727549e-06, "loss": 0.4367, "step": 1648 }, { "epoch": 0.11690889755405884, "grad_norm": 3.056862271647135, "learning_rate": 9.803478270642575e-06, "loss": 0.4757, "step": 1649 }, { "epoch": 0.11697979439914924, "grad_norm": 1.1186142093052158, "learning_rate": 9.803159410718586e-06, "loss": 0.6583, "step": 1650 }, { "epoch": 0.11705069124423963, "grad_norm": 3.4234776717426394, "learning_rate": 9.802840297520334e-06, "loss": 0.4437, "step": 1651 }, { "epoch": 0.11712158808933003, "grad_norm": 8.160452700834876, "learning_rate": 9.802520931064648e-06, "loss": 0.4657, "step": 1652 }, { "epoch": 0.11719248493442042, "grad_norm": 6.906163155532012, "learning_rate": 9.802201311368366e-06, "loss": 0.4475, "step": 1653 }, { "epoch": 0.11726338177951082, "grad_norm": 3.7811059302880192, "learning_rate": 9.801881438448346e-06, "loss": 0.4684, "step": 1654 }, { "epoch": 0.1173342786246012, "grad_norm": 3.1858665800343235, "learning_rate": 9.801561312321452e-06, "loss": 0.4285, "step": 1655 }, { "epoch": 0.1174051754696916, "grad_norm": 2.3658489377223106, "learning_rate": 9.801240933004563e-06, "loss": 0.4429, "step": 1656 }, { "epoch": 0.11747607231478199, "grad_norm": 3.1971400762835356, "learning_rate": 9.800920300514578e-06, "loss": 0.5191, "step": 1657 }, { "epoch": 0.11754696915987238, "grad_norm": 2.9194024290201894, "learning_rate": 9.800599414868399e-06, "loss": 0.402, "step": 1658 }, { "epoch": 0.11761786600496278, "grad_norm": 2.5210944892312543, "learning_rate": 9.80027827608295e-06, "loss": 0.4649, "step": 1659 }, { "epoch": 0.11768876285005317, "grad_norm": 2.752717062858412, "learning_rate": 9.799956884175163e-06, "loss": 0.4082, "step": 1660 }, { "epoch": 0.11775965969514357, "grad_norm": 2.3789314099992827, "learning_rate": 9.799635239161986e-06, "loss": 0.4245, "step": 1661 }, { "epoch": 0.11783055654023396, "grad_norm": 5.71653181152187, "learning_rate": 9.79931334106038e-06, "loss": 0.4583, "step": 1662 }, { "epoch": 0.11790145338532436, "grad_norm": 2.4745302221937893, "learning_rate": 9.798991189887317e-06, "loss": 0.4779, "step": 1663 }, { "epoch": 0.11797235023041475, "grad_norm": 1.2444509024064512, "learning_rate": 9.798668785659787e-06, "loss": 0.7389, "step": 1664 }, { "epoch": 0.11804324707550513, "grad_norm": 4.361312048905991, "learning_rate": 9.798346128394789e-06, "loss": 0.4249, "step": 1665 }, { "epoch": 0.11811414392059554, "grad_norm": 2.7670437092301654, "learning_rate": 9.798023218109335e-06, "loss": 0.5025, "step": 1666 }, { "epoch": 0.11818504076568592, "grad_norm": 1.1208690965059362, "learning_rate": 9.797700054820458e-06, "loss": 0.6919, "step": 1667 }, { "epoch": 0.11825593761077632, "grad_norm": 2.5003270175738055, "learning_rate": 9.797376638545194e-06, "loss": 0.4664, "step": 1668 }, { "epoch": 0.11832683445586671, "grad_norm": 2.7620558670378625, "learning_rate": 9.797052969300598e-06, "loss": 0.4701, "step": 1669 }, { "epoch": 0.11839773130095711, "grad_norm": 2.9368466982714105, "learning_rate": 9.796729047103738e-06, "loss": 0.4443, "step": 1670 }, { "epoch": 0.1184686281460475, "grad_norm": 2.5814148237197077, "learning_rate": 9.796404871971692e-06, "loss": 0.461, "step": 1671 }, { "epoch": 0.1185395249911379, "grad_norm": 5.261511246988483, "learning_rate": 9.796080443921558e-06, "loss": 0.437, "step": 1672 }, { "epoch": 0.11861042183622829, "grad_norm": 2.7863917009374606, "learning_rate": 9.795755762970442e-06, "loss": 0.4568, "step": 1673 }, { "epoch": 0.11868131868131868, "grad_norm": 4.606710361956344, "learning_rate": 9.795430829135464e-06, "loss": 0.4157, "step": 1674 }, { "epoch": 0.11875221552640908, "grad_norm": 3.755051495901355, "learning_rate": 9.795105642433757e-06, "loss": 0.4499, "step": 1675 }, { "epoch": 0.11882311237149946, "grad_norm": 3.7867968958990867, "learning_rate": 9.79478020288247e-06, "loss": 0.5044, "step": 1676 }, { "epoch": 0.11889400921658987, "grad_norm": 3.1037037499478815, "learning_rate": 9.794454510498763e-06, "loss": 0.4675, "step": 1677 }, { "epoch": 0.11896490606168025, "grad_norm": 2.8513008208383943, "learning_rate": 9.79412856529981e-06, "loss": 0.4252, "step": 1678 }, { "epoch": 0.11903580290677065, "grad_norm": 2.898277384053136, "learning_rate": 9.7938023673028e-06, "loss": 0.4808, "step": 1679 }, { "epoch": 0.11910669975186104, "grad_norm": 5.246988568726137, "learning_rate": 9.793475916524932e-06, "loss": 0.4877, "step": 1680 }, { "epoch": 0.11917759659695143, "grad_norm": 1.966215262646127, "learning_rate": 9.79314921298342e-06, "loss": 0.4458, "step": 1681 }, { "epoch": 0.11924849344204183, "grad_norm": 9.366610075064177, "learning_rate": 9.792822256695492e-06, "loss": 0.4081, "step": 1682 }, { "epoch": 0.11931939028713222, "grad_norm": 6.462482370360749, "learning_rate": 9.792495047678385e-06, "loss": 0.4857, "step": 1683 }, { "epoch": 0.11939028713222262, "grad_norm": 3.2522941482091223, "learning_rate": 9.792167585949358e-06, "loss": 0.5065, "step": 1684 }, { "epoch": 0.119461183977313, "grad_norm": 4.556845053159463, "learning_rate": 9.791839871525679e-06, "loss": 0.4025, "step": 1685 }, { "epoch": 0.11953208082240341, "grad_norm": 3.641910388132623, "learning_rate": 9.791511904424625e-06, "loss": 0.3825, "step": 1686 }, { "epoch": 0.1196029776674938, "grad_norm": 7.460717881862673, "learning_rate": 9.79118368466349e-06, "loss": 0.4612, "step": 1687 }, { "epoch": 0.1196738745125842, "grad_norm": 5.840149269321743, "learning_rate": 9.790855212259582e-06, "loss": 0.4305, "step": 1688 }, { "epoch": 0.11974477135767458, "grad_norm": 4.138988125449894, "learning_rate": 9.790526487230222e-06, "loss": 0.4439, "step": 1689 }, { "epoch": 0.11981566820276497, "grad_norm": 2.922408206078902, "learning_rate": 9.790197509592744e-06, "loss": 0.4696, "step": 1690 }, { "epoch": 0.11988656504785537, "grad_norm": 3.077580024016014, "learning_rate": 9.789868279364497e-06, "loss": 0.433, "step": 1691 }, { "epoch": 0.11995746189294576, "grad_norm": 3.798902586787128, "learning_rate": 9.789538796562839e-06, "loss": 0.4371, "step": 1692 }, { "epoch": 0.12002835873803616, "grad_norm": 2.8014694476149593, "learning_rate": 9.789209061205143e-06, "loss": 0.4602, "step": 1693 }, { "epoch": 0.12009925558312655, "grad_norm": 2.910692412585922, "learning_rate": 9.788879073308799e-06, "loss": 0.3856, "step": 1694 }, { "epoch": 0.12017015242821695, "grad_norm": 7.879569973126844, "learning_rate": 9.788548832891205e-06, "loss": 0.478, "step": 1695 }, { "epoch": 0.12024104927330734, "grad_norm": 2.0092070279773995, "learning_rate": 9.788218339969777e-06, "loss": 0.4616, "step": 1696 }, { "epoch": 0.12031194611839773, "grad_norm": 4.875274117494004, "learning_rate": 9.787887594561942e-06, "loss": 0.4433, "step": 1697 }, { "epoch": 0.12038284296348813, "grad_norm": 5.204675995746763, "learning_rate": 9.78755659668514e-06, "loss": 0.4203, "step": 1698 }, { "epoch": 0.12045373980857851, "grad_norm": 13.54070066077068, "learning_rate": 9.787225346356822e-06, "loss": 0.5016, "step": 1699 }, { "epoch": 0.12052463665366892, "grad_norm": 3.1201829847929585, "learning_rate": 9.78689384359446e-06, "loss": 0.4337, "step": 1700 }, { "epoch": 0.1205955334987593, "grad_norm": 5.301520292576671, "learning_rate": 9.78656208841553e-06, "loss": 0.4248, "step": 1701 }, { "epoch": 0.1206664303438497, "grad_norm": 2.5547179541022107, "learning_rate": 9.78623008083753e-06, "loss": 0.4487, "step": 1702 }, { "epoch": 0.12073732718894009, "grad_norm": 5.74966978338307, "learning_rate": 9.785897820877963e-06, "loss": 0.4552, "step": 1703 }, { "epoch": 0.12080822403403049, "grad_norm": 6.0838854480793305, "learning_rate": 9.78556530855435e-06, "loss": 0.4149, "step": 1704 }, { "epoch": 0.12087912087912088, "grad_norm": 3.8305509213839426, "learning_rate": 9.785232543884228e-06, "loss": 0.4457, "step": 1705 }, { "epoch": 0.12095001772421127, "grad_norm": 3.392705720828755, "learning_rate": 9.78489952688514e-06, "loss": 0.4343, "step": 1706 }, { "epoch": 0.12102091456930167, "grad_norm": 7.270775138424205, "learning_rate": 9.784566257574649e-06, "loss": 0.5144, "step": 1707 }, { "epoch": 0.12109181141439206, "grad_norm": 3.7295339565133174, "learning_rate": 9.784232735970325e-06, "loss": 0.4451, "step": 1708 }, { "epoch": 0.12116270825948246, "grad_norm": 4.106766765886688, "learning_rate": 9.78389896208976e-06, "loss": 0.4659, "step": 1709 }, { "epoch": 0.12123360510457284, "grad_norm": 2.835490560738352, "learning_rate": 9.78356493595055e-06, "loss": 0.4827, "step": 1710 }, { "epoch": 0.12130450194966325, "grad_norm": 3.2385649762633544, "learning_rate": 9.78323065757031e-06, "loss": 0.466, "step": 1711 }, { "epoch": 0.12137539879475363, "grad_norm": 3.2891237701268405, "learning_rate": 9.782896126966666e-06, "loss": 0.4727, "step": 1712 }, { "epoch": 0.12144629563984402, "grad_norm": 3.8164613103839957, "learning_rate": 9.782561344157259e-06, "loss": 0.5005, "step": 1713 }, { "epoch": 0.12151719248493442, "grad_norm": 5.1072186817206795, "learning_rate": 9.782226309159743e-06, "loss": 0.4202, "step": 1714 }, { "epoch": 0.12158808933002481, "grad_norm": 2.893303996144878, "learning_rate": 9.781891021991783e-06, "loss": 0.4678, "step": 1715 }, { "epoch": 0.12165898617511521, "grad_norm": 4.618799243577655, "learning_rate": 9.781555482671059e-06, "loss": 0.416, "step": 1716 }, { "epoch": 0.1217298830202056, "grad_norm": 3.771827757284567, "learning_rate": 9.781219691215264e-06, "loss": 0.5029, "step": 1717 }, { "epoch": 0.121800779865296, "grad_norm": 5.170268272309375, "learning_rate": 9.780883647642108e-06, "loss": 0.4994, "step": 1718 }, { "epoch": 0.12187167671038639, "grad_norm": 2.8638464784168534, "learning_rate": 9.780547351969306e-06, "loss": 0.3808, "step": 1719 }, { "epoch": 0.12194257355547679, "grad_norm": 2.6222099626652, "learning_rate": 9.780210804214594e-06, "loss": 0.4346, "step": 1720 }, { "epoch": 0.12201347040056718, "grad_norm": 3.612704384387709, "learning_rate": 9.779874004395718e-06, "loss": 0.4632, "step": 1721 }, { "epoch": 0.12208436724565756, "grad_norm": 3.093813697929966, "learning_rate": 9.779536952530437e-06, "loss": 0.4141, "step": 1722 }, { "epoch": 0.12215526409074796, "grad_norm": 4.352976910154667, "learning_rate": 9.779199648636523e-06, "loss": 0.4758, "step": 1723 }, { "epoch": 0.12222616093583835, "grad_norm": 7.530415953513782, "learning_rate": 9.778862092731766e-06, "loss": 0.4491, "step": 1724 }, { "epoch": 0.12229705778092875, "grad_norm": 3.5872519897000172, "learning_rate": 9.778524284833963e-06, "loss": 0.4279, "step": 1725 }, { "epoch": 0.12236795462601914, "grad_norm": 3.3742381387627183, "learning_rate": 9.778186224960927e-06, "loss": 0.4277, "step": 1726 }, { "epoch": 0.12243885147110954, "grad_norm": 2.638590467218637, "learning_rate": 9.777847913130484e-06, "loss": 0.3751, "step": 1727 }, { "epoch": 0.12250974831619993, "grad_norm": 4.396575029610281, "learning_rate": 9.777509349360474e-06, "loss": 0.4629, "step": 1728 }, { "epoch": 0.12258064516129032, "grad_norm": 5.104570317129462, "learning_rate": 9.77717053366875e-06, "loss": 0.4404, "step": 1729 }, { "epoch": 0.12265154200638072, "grad_norm": 4.334076716747197, "learning_rate": 9.776831466073175e-06, "loss": 0.4504, "step": 1730 }, { "epoch": 0.1227224388514711, "grad_norm": 5.299928107106015, "learning_rate": 9.776492146591633e-06, "loss": 0.477, "step": 1731 }, { "epoch": 0.1227933356965615, "grad_norm": 5.9880383784634565, "learning_rate": 9.776152575242014e-06, "loss": 0.4339, "step": 1732 }, { "epoch": 0.1228642325416519, "grad_norm": 4.995495763363755, "learning_rate": 9.775812752042224e-06, "loss": 0.4986, "step": 1733 }, { "epoch": 0.1229351293867423, "grad_norm": 3.6360201136461563, "learning_rate": 9.775472677010184e-06, "loss": 0.5029, "step": 1734 }, { "epoch": 0.12300602623183268, "grad_norm": 3.2927338626601648, "learning_rate": 9.775132350163823e-06, "loss": 0.4692, "step": 1735 }, { "epoch": 0.12307692307692308, "grad_norm": 3.772866074130064, "learning_rate": 9.774791771521089e-06, "loss": 0.4789, "step": 1736 }, { "epoch": 0.12314781992201347, "grad_norm": 7.411752659092565, "learning_rate": 9.77445094109994e-06, "loss": 0.4789, "step": 1737 }, { "epoch": 0.12321871676710386, "grad_norm": 3.87350012493972, "learning_rate": 9.774109858918349e-06, "loss": 0.5017, "step": 1738 }, { "epoch": 0.12328961361219426, "grad_norm": 3.0783092204515636, "learning_rate": 9.7737685249943e-06, "loss": 0.4428, "step": 1739 }, { "epoch": 0.12336051045728465, "grad_norm": 4.1296479334669955, "learning_rate": 9.773426939345796e-06, "loss": 0.4663, "step": 1740 }, { "epoch": 0.12343140730237505, "grad_norm": 2.85433558798632, "learning_rate": 9.773085101990844e-06, "loss": 0.4233, "step": 1741 }, { "epoch": 0.12350230414746544, "grad_norm": 7.352720631438108, "learning_rate": 9.772743012947473e-06, "loss": 0.4658, "step": 1742 }, { "epoch": 0.12357320099255584, "grad_norm": 6.223668535781227, "learning_rate": 9.772400672233719e-06, "loss": 0.4797, "step": 1743 }, { "epoch": 0.12364409783764622, "grad_norm": 6.920409468500795, "learning_rate": 9.772058079867634e-06, "loss": 0.4666, "step": 1744 }, { "epoch": 0.12371499468273661, "grad_norm": 3.4404149810775353, "learning_rate": 9.771715235867286e-06, "loss": 0.4576, "step": 1745 }, { "epoch": 0.12378589152782701, "grad_norm": 4.637604304661856, "learning_rate": 9.77137214025075e-06, "loss": 0.5058, "step": 1746 }, { "epoch": 0.1238567883729174, "grad_norm": 2.5556404790548064, "learning_rate": 9.77102879303612e-06, "loss": 0.3937, "step": 1747 }, { "epoch": 0.1239276852180078, "grad_norm": 5.850113551803512, "learning_rate": 9.770685194241499e-06, "loss": 0.4236, "step": 1748 }, { "epoch": 0.12399858206309819, "grad_norm": 5.314878395485551, "learning_rate": 9.770341343885008e-06, "loss": 0.4374, "step": 1749 }, { "epoch": 0.12406947890818859, "grad_norm": 4.524328493555082, "learning_rate": 9.769997241984775e-06, "loss": 0.4537, "step": 1750 }, { "epoch": 0.12414037575327898, "grad_norm": 3.5376097655099357, "learning_rate": 9.769652888558949e-06, "loss": 0.4664, "step": 1751 }, { "epoch": 0.12421127259836938, "grad_norm": 2.4247212377555445, "learning_rate": 9.769308283625684e-06, "loss": 0.4778, "step": 1752 }, { "epoch": 0.12428216944345977, "grad_norm": 2.2575824988873427, "learning_rate": 9.768963427203154e-06, "loss": 0.473, "step": 1753 }, { "epoch": 0.12435306628855015, "grad_norm": 2.6318866838979416, "learning_rate": 9.768618319309541e-06, "loss": 0.4683, "step": 1754 }, { "epoch": 0.12442396313364056, "grad_norm": 3.2895238221536642, "learning_rate": 9.768272959963045e-06, "loss": 0.4845, "step": 1755 }, { "epoch": 0.12449485997873094, "grad_norm": 4.437218899508378, "learning_rate": 9.767927349181878e-06, "loss": 0.4205, "step": 1756 }, { "epoch": 0.12456575682382134, "grad_norm": 9.191823036977253, "learning_rate": 9.767581486984262e-06, "loss": 0.4412, "step": 1757 }, { "epoch": 0.12463665366891173, "grad_norm": 3.814337685275861, "learning_rate": 9.767235373388433e-06, "loss": 0.519, "step": 1758 }, { "epoch": 0.12470755051400213, "grad_norm": 4.489503176067312, "learning_rate": 9.766889008412646e-06, "loss": 0.4541, "step": 1759 }, { "epoch": 0.12477844735909252, "grad_norm": 2.7405332893060637, "learning_rate": 9.766542392075162e-06, "loss": 0.4429, "step": 1760 }, { "epoch": 0.12484934420418291, "grad_norm": 1.2125582265570374, "learning_rate": 9.766195524394262e-06, "loss": 0.731, "step": 1761 }, { "epoch": 0.12492024104927331, "grad_norm": 2.763665820802212, "learning_rate": 9.765848405388232e-06, "loss": 0.4429, "step": 1762 }, { "epoch": 0.1249911378943637, "grad_norm": 3.161749429527872, "learning_rate": 9.76550103507538e-06, "loss": 0.4267, "step": 1763 }, { "epoch": 0.1250620347394541, "grad_norm": 2.6989023633605065, "learning_rate": 9.765153413474018e-06, "loss": 0.4128, "step": 1764 }, { "epoch": 0.1251329315845445, "grad_norm": 5.8156288918246855, "learning_rate": 9.76480554060248e-06, "loss": 0.4743, "step": 1765 }, { "epoch": 0.12520382842963487, "grad_norm": 5.327484323352017, "learning_rate": 9.764457416479112e-06, "loss": 0.4822, "step": 1766 }, { "epoch": 0.12527472527472527, "grad_norm": 2.8222236120896507, "learning_rate": 9.764109041122265e-06, "loss": 0.4466, "step": 1767 }, { "epoch": 0.12534562211981568, "grad_norm": 4.851053840949719, "learning_rate": 9.763760414550313e-06, "loss": 0.4887, "step": 1768 }, { "epoch": 0.12541651896490605, "grad_norm": 8.268359377747304, "learning_rate": 9.763411536781639e-06, "loss": 0.452, "step": 1769 }, { "epoch": 0.12548741580999645, "grad_norm": 2.8239364743107385, "learning_rate": 9.763062407834639e-06, "loss": 0.4108, "step": 1770 }, { "epoch": 0.12555831265508685, "grad_norm": 4.499947499195205, "learning_rate": 9.762713027727721e-06, "loss": 0.4903, "step": 1771 }, { "epoch": 0.12562920950017725, "grad_norm": 4.244154836968584, "learning_rate": 9.762363396479311e-06, "loss": 0.4402, "step": 1772 }, { "epoch": 0.12570010634526763, "grad_norm": 2.5117048264001682, "learning_rate": 9.762013514107843e-06, "loss": 0.4338, "step": 1773 }, { "epoch": 0.12577100319035803, "grad_norm": 3.5236676836761673, "learning_rate": 9.76166338063177e-06, "loss": 0.4757, "step": 1774 }, { "epoch": 0.12584190003544843, "grad_norm": 3.1331375964265136, "learning_rate": 9.76131299606955e-06, "loss": 0.4196, "step": 1775 }, { "epoch": 0.12591279688053883, "grad_norm": 6.454847744975876, "learning_rate": 9.760962360439663e-06, "loss": 0.4633, "step": 1776 }, { "epoch": 0.1259836937256292, "grad_norm": 3.2397818724826393, "learning_rate": 9.760611473760596e-06, "loss": 0.4457, "step": 1777 }, { "epoch": 0.1260545905707196, "grad_norm": 3.892172645638498, "learning_rate": 9.760260336050853e-06, "loss": 0.4699, "step": 1778 }, { "epoch": 0.12612548741581, "grad_norm": 5.649671264412773, "learning_rate": 9.75990894732895e-06, "loss": 0.4663, "step": 1779 }, { "epoch": 0.12619638426090038, "grad_norm": 4.515306610253973, "learning_rate": 9.759557307613414e-06, "loss": 0.4381, "step": 1780 }, { "epoch": 0.12626728110599078, "grad_norm": 4.567914282618379, "learning_rate": 9.759205416922788e-06, "loss": 0.4561, "step": 1781 }, { "epoch": 0.12633817795108118, "grad_norm": 4.2812402368466245, "learning_rate": 9.758853275275629e-06, "loss": 0.435, "step": 1782 }, { "epoch": 0.12640907479617158, "grad_norm": 1.1792655263995728, "learning_rate": 9.758500882690503e-06, "loss": 0.6682, "step": 1783 }, { "epoch": 0.12647997164126196, "grad_norm": 4.337220331379262, "learning_rate": 9.758148239185996e-06, "loss": 0.4417, "step": 1784 }, { "epoch": 0.12655086848635236, "grad_norm": 3.900355300723764, "learning_rate": 9.757795344780698e-06, "loss": 0.4575, "step": 1785 }, { "epoch": 0.12662176533144276, "grad_norm": 4.032507728798496, "learning_rate": 9.75744219949322e-06, "loss": 0.4705, "step": 1786 }, { "epoch": 0.12669266217653313, "grad_norm": 3.1327513644429694, "learning_rate": 9.757088803342186e-06, "loss": 0.4552, "step": 1787 }, { "epoch": 0.12676355902162353, "grad_norm": 4.1140921972463005, "learning_rate": 9.756735156346227e-06, "loss": 0.4714, "step": 1788 }, { "epoch": 0.12683445586671394, "grad_norm": 3.750631134631233, "learning_rate": 9.756381258523993e-06, "loss": 0.4349, "step": 1789 }, { "epoch": 0.12690535271180434, "grad_norm": 5.519643077178806, "learning_rate": 9.756027109894144e-06, "loss": 0.484, "step": 1790 }, { "epoch": 0.1269762495568947, "grad_norm": 3.4263725557491123, "learning_rate": 9.755672710475357e-06, "loss": 0.3751, "step": 1791 }, { "epoch": 0.1270471464019851, "grad_norm": 13.626880417316306, "learning_rate": 9.755318060286317e-06, "loss": 0.4137, "step": 1792 }, { "epoch": 0.1271180432470755, "grad_norm": 2.6870622445319072, "learning_rate": 9.754963159345725e-06, "loss": 0.4652, "step": 1793 }, { "epoch": 0.1271889400921659, "grad_norm": 5.462760470443516, "learning_rate": 9.754608007672299e-06, "loss": 0.4198, "step": 1794 }, { "epoch": 0.1272598369372563, "grad_norm": 6.062781432806735, "learning_rate": 9.754252605284761e-06, "loss": 0.4129, "step": 1795 }, { "epoch": 0.1273307337823467, "grad_norm": 2.913715279438595, "learning_rate": 9.753896952201856e-06, "loss": 0.3936, "step": 1796 }, { "epoch": 0.1274016306274371, "grad_norm": 4.919771294651539, "learning_rate": 9.753541048442337e-06, "loss": 0.4635, "step": 1797 }, { "epoch": 0.12747252747252746, "grad_norm": 3.5434069710532565, "learning_rate": 9.753184894024968e-06, "loss": 0.4362, "step": 1798 }, { "epoch": 0.12754342431761787, "grad_norm": 5.8786512176221475, "learning_rate": 9.752828488968533e-06, "loss": 0.4837, "step": 1799 }, { "epoch": 0.12761432116270827, "grad_norm": 2.079515553595753, "learning_rate": 9.752471833291824e-06, "loss": 0.4179, "step": 1800 }, { "epoch": 0.12768521800779864, "grad_norm": 4.954470250848991, "learning_rate": 9.752114927013648e-06, "loss": 0.4987, "step": 1801 }, { "epoch": 0.12775611485288904, "grad_norm": 5.105103448271213, "learning_rate": 9.751757770152823e-06, "loss": 0.4637, "step": 1802 }, { "epoch": 0.12782701169797944, "grad_norm": 5.917691868822002, "learning_rate": 9.751400362728186e-06, "loss": 0.5081, "step": 1803 }, { "epoch": 0.12789790854306984, "grad_norm": 3.6956537635832394, "learning_rate": 9.751042704758582e-06, "loss": 0.5002, "step": 1804 }, { "epoch": 0.12796880538816022, "grad_norm": 4.79337810759765, "learning_rate": 9.750684796262867e-06, "loss": 0.4632, "step": 1805 }, { "epoch": 0.12803970223325062, "grad_norm": 5.098147625718408, "learning_rate": 9.750326637259918e-06, "loss": 0.4182, "step": 1806 }, { "epoch": 0.12811059907834102, "grad_norm": 3.097729993177108, "learning_rate": 9.74996822776862e-06, "loss": 0.4827, "step": 1807 }, { "epoch": 0.12818149592343142, "grad_norm": 3.1985504514906697, "learning_rate": 9.749609567807872e-06, "loss": 0.4108, "step": 1808 }, { "epoch": 0.1282523927685218, "grad_norm": 1.3067248280289125, "learning_rate": 9.749250657396587e-06, "loss": 0.663, "step": 1809 }, { "epoch": 0.1283232896136122, "grad_norm": 3.2103957889997634, "learning_rate": 9.748891496553688e-06, "loss": 0.4736, "step": 1810 }, { "epoch": 0.1283941864587026, "grad_norm": 2.7563285398134765, "learning_rate": 9.748532085298118e-06, "loss": 0.438, "step": 1811 }, { "epoch": 0.12846508330379297, "grad_norm": 1.100281375689407, "learning_rate": 9.748172423648825e-06, "loss": 0.6718, "step": 1812 }, { "epoch": 0.12853598014888337, "grad_norm": 3.2105519890452543, "learning_rate": 9.747812511624776e-06, "loss": 0.4648, "step": 1813 }, { "epoch": 0.12860687699397377, "grad_norm": 3.8599680183335843, "learning_rate": 9.74745234924495e-06, "loss": 0.4672, "step": 1814 }, { "epoch": 0.12867777383906417, "grad_norm": 3.125739521426236, "learning_rate": 9.747091936528337e-06, "loss": 0.5024, "step": 1815 }, { "epoch": 0.12874867068415455, "grad_norm": 2.56761395722709, "learning_rate": 9.746731273493943e-06, "loss": 0.4535, "step": 1816 }, { "epoch": 0.12881956752924495, "grad_norm": 2.7252540391784206, "learning_rate": 9.746370360160788e-06, "loss": 0.388, "step": 1817 }, { "epoch": 0.12889046437433535, "grad_norm": 4.536597777765356, "learning_rate": 9.7460091965479e-06, "loss": 0.4839, "step": 1818 }, { "epoch": 0.12896136121942572, "grad_norm": 4.078332035869746, "learning_rate": 9.745647782674324e-06, "loss": 0.446, "step": 1819 }, { "epoch": 0.12903225806451613, "grad_norm": 3.784486494173941, "learning_rate": 9.745286118559118e-06, "loss": 0.4587, "step": 1820 }, { "epoch": 0.12910315490960653, "grad_norm": 3.234699154054033, "learning_rate": 9.744924204221354e-06, "loss": 0.4525, "step": 1821 }, { "epoch": 0.12917405175469693, "grad_norm": 4.060557954460595, "learning_rate": 9.744562039680115e-06, "loss": 0.5238, "step": 1822 }, { "epoch": 0.1292449485997873, "grad_norm": 5.808710434057479, "learning_rate": 9.744199624954498e-06, "loss": 0.4342, "step": 1823 }, { "epoch": 0.1293158454448777, "grad_norm": 4.414894265724918, "learning_rate": 9.743836960063614e-06, "loss": 0.4353, "step": 1824 }, { "epoch": 0.1293867422899681, "grad_norm": 15.424290582422419, "learning_rate": 9.743474045026586e-06, "loss": 0.4165, "step": 1825 }, { "epoch": 0.12945763913505848, "grad_norm": 4.210908617894102, "learning_rate": 9.743110879862552e-06, "loss": 0.4374, "step": 1826 }, { "epoch": 0.12952853598014888, "grad_norm": 4.342932805628042, "learning_rate": 9.742747464590659e-06, "loss": 0.4944, "step": 1827 }, { "epoch": 0.12959943282523928, "grad_norm": 4.515472798645543, "learning_rate": 9.742383799230074e-06, "loss": 0.4609, "step": 1828 }, { "epoch": 0.12967032967032968, "grad_norm": 2.7172561137031455, "learning_rate": 9.742019883799972e-06, "loss": 0.4696, "step": 1829 }, { "epoch": 0.12974122651542006, "grad_norm": 3.786956016744019, "learning_rate": 9.74165571831954e-06, "loss": 0.4657, "step": 1830 }, { "epoch": 0.12981212336051046, "grad_norm": 4.740930807576371, "learning_rate": 9.741291302807985e-06, "loss": 0.4564, "step": 1831 }, { "epoch": 0.12988302020560086, "grad_norm": 4.796524393038825, "learning_rate": 9.740926637284518e-06, "loss": 0.4806, "step": 1832 }, { "epoch": 0.12995391705069123, "grad_norm": 5.559683435686319, "learning_rate": 9.740561721768373e-06, "loss": 0.4865, "step": 1833 }, { "epoch": 0.13002481389578163, "grad_norm": 2.0605819477006317, "learning_rate": 9.74019655627879e-06, "loss": 0.3721, "step": 1834 }, { "epoch": 0.13009571074087203, "grad_norm": 5.574833772960424, "learning_rate": 9.739831140835024e-06, "loss": 0.4757, "step": 1835 }, { "epoch": 0.13016660758596243, "grad_norm": 3.713885782819392, "learning_rate": 9.739465475456344e-06, "loss": 0.4077, "step": 1836 }, { "epoch": 0.1302375044310528, "grad_norm": 2.791750717313671, "learning_rate": 9.739099560162033e-06, "loss": 0.4032, "step": 1837 }, { "epoch": 0.1303084012761432, "grad_norm": 2.860138889434735, "learning_rate": 9.738733394971384e-06, "loss": 0.4483, "step": 1838 }, { "epoch": 0.1303792981212336, "grad_norm": 4.150366650739205, "learning_rate": 9.738366979903706e-06, "loss": 0.525, "step": 1839 }, { "epoch": 0.130450194966324, "grad_norm": 2.103706549778316, "learning_rate": 9.738000314978321e-06, "loss": 0.4381, "step": 1840 }, { "epoch": 0.13052109181141439, "grad_norm": 2.1715663757716603, "learning_rate": 9.737633400214563e-06, "loss": 0.4584, "step": 1841 }, { "epoch": 0.1305919886565048, "grad_norm": 5.341508456524376, "learning_rate": 9.737266235631782e-06, "loss": 0.468, "step": 1842 }, { "epoch": 0.1306628855015952, "grad_norm": 4.552127383353271, "learning_rate": 9.736898821249333e-06, "loss": 0.4586, "step": 1843 }, { "epoch": 0.13073378234668556, "grad_norm": 3.9246565155700734, "learning_rate": 9.736531157086596e-06, "loss": 0.4204, "step": 1844 }, { "epoch": 0.13080467919177596, "grad_norm": 8.466818029750172, "learning_rate": 9.736163243162955e-06, "loss": 0.4238, "step": 1845 }, { "epoch": 0.13087557603686636, "grad_norm": 2.296615675271641, "learning_rate": 9.73579507949781e-06, "loss": 0.4082, "step": 1846 }, { "epoch": 0.13094647288195677, "grad_norm": 3.9969817965701377, "learning_rate": 9.735426666110577e-06, "loss": 0.516, "step": 1847 }, { "epoch": 0.13101736972704714, "grad_norm": 1.098245511750092, "learning_rate": 9.735058003020682e-06, "loss": 0.6958, "step": 1848 }, { "epoch": 0.13108826657213754, "grad_norm": 5.076447658036387, "learning_rate": 9.734689090247564e-06, "loss": 0.4604, "step": 1849 }, { "epoch": 0.13115916341722794, "grad_norm": 3.22332143992883, "learning_rate": 9.734319927810676e-06, "loss": 0.474, "step": 1850 }, { "epoch": 0.13123006026231832, "grad_norm": 2.658862943782776, "learning_rate": 9.733950515729485e-06, "loss": 0.4266, "step": 1851 }, { "epoch": 0.13130095710740872, "grad_norm": 3.056719811713673, "learning_rate": 9.73358085402347e-06, "loss": 0.4378, "step": 1852 }, { "epoch": 0.13137185395249912, "grad_norm": 2.945647207261667, "learning_rate": 9.733210942712123e-06, "loss": 0.4347, "step": 1853 }, { "epoch": 0.13144275079758952, "grad_norm": 2.8053176797395025, "learning_rate": 9.732840781814951e-06, "loss": 0.4065, "step": 1854 }, { "epoch": 0.1315136476426799, "grad_norm": 10.095157270902341, "learning_rate": 9.732470371351472e-06, "loss": 0.4276, "step": 1855 }, { "epoch": 0.1315845444877703, "grad_norm": 2.1364109047786877, "learning_rate": 9.732099711341219e-06, "loss": 0.4812, "step": 1856 }, { "epoch": 0.1316554413328607, "grad_norm": 4.1436178942756925, "learning_rate": 9.731728801803736e-06, "loss": 0.443, "step": 1857 }, { "epoch": 0.13172633817795107, "grad_norm": 4.295616455191043, "learning_rate": 9.73135764275858e-06, "loss": 0.4477, "step": 1858 }, { "epoch": 0.13179723502304147, "grad_norm": 3.0323400246118237, "learning_rate": 9.730986234225326e-06, "loss": 0.455, "step": 1859 }, { "epoch": 0.13186813186813187, "grad_norm": 5.680449922676568, "learning_rate": 9.730614576223557e-06, "loss": 0.4459, "step": 1860 }, { "epoch": 0.13193902871322227, "grad_norm": 5.150881224609417, "learning_rate": 9.73024266877287e-06, "loss": 0.5015, "step": 1861 }, { "epoch": 0.13200992555831265, "grad_norm": 5.349616154450249, "learning_rate": 9.729870511892877e-06, "loss": 0.4912, "step": 1862 }, { "epoch": 0.13208082240340305, "grad_norm": 3.8277483696873076, "learning_rate": 9.729498105603201e-06, "loss": 0.4554, "step": 1863 }, { "epoch": 0.13215171924849345, "grad_norm": 2.7935458458504905, "learning_rate": 9.729125449923481e-06, "loss": 0.4254, "step": 1864 }, { "epoch": 0.13222261609358382, "grad_norm": 2.3017170038110235, "learning_rate": 9.728752544873365e-06, "loss": 0.3954, "step": 1865 }, { "epoch": 0.13229351293867422, "grad_norm": 6.250638144437334, "learning_rate": 9.72837939047252e-06, "loss": 0.4614, "step": 1866 }, { "epoch": 0.13236440978376462, "grad_norm": 3.4022791246951067, "learning_rate": 9.728005986740621e-06, "loss": 0.3998, "step": 1867 }, { "epoch": 0.13243530662885503, "grad_norm": 4.184726808438611, "learning_rate": 9.727632333697355e-06, "loss": 0.4267, "step": 1868 }, { "epoch": 0.1325062034739454, "grad_norm": 9.236623885851877, "learning_rate": 9.727258431362429e-06, "loss": 0.4384, "step": 1869 }, { "epoch": 0.1325771003190358, "grad_norm": 4.438900106568357, "learning_rate": 9.726884279755558e-06, "loss": 0.4094, "step": 1870 }, { "epoch": 0.1326479971641262, "grad_norm": 7.481625339687779, "learning_rate": 9.72650987889647e-06, "loss": 0.4579, "step": 1871 }, { "epoch": 0.1327188940092166, "grad_norm": 4.604635728869877, "learning_rate": 9.72613522880491e-06, "loss": 0.4459, "step": 1872 }, { "epoch": 0.13278979085430698, "grad_norm": 3.1744037421910907, "learning_rate": 9.725760329500631e-06, "loss": 0.4435, "step": 1873 }, { "epoch": 0.13286068769939738, "grad_norm": 5.13077397550718, "learning_rate": 9.725385181003405e-06, "loss": 0.4733, "step": 1874 }, { "epoch": 0.13293158454448778, "grad_norm": 3.0599201872310764, "learning_rate": 9.725009783333009e-06, "loss": 0.4433, "step": 1875 }, { "epoch": 0.13300248138957815, "grad_norm": 1.1114608570467528, "learning_rate": 9.72463413650924e-06, "loss": 0.6946, "step": 1876 }, { "epoch": 0.13307337823466855, "grad_norm": 2.2264472844924557, "learning_rate": 9.72425824055191e-06, "loss": 0.4755, "step": 1877 }, { "epoch": 0.13314427507975896, "grad_norm": 3.983120774648954, "learning_rate": 9.723882095480834e-06, "loss": 0.4495, "step": 1878 }, { "epoch": 0.13321517192484936, "grad_norm": 2.7075426595841403, "learning_rate": 9.723505701315851e-06, "loss": 0.4868, "step": 1879 }, { "epoch": 0.13328606876993973, "grad_norm": 3.811933183219834, "learning_rate": 9.723129058076808e-06, "loss": 0.4531, "step": 1880 }, { "epoch": 0.13335696561503013, "grad_norm": 3.39540697782624, "learning_rate": 9.722752165783565e-06, "loss": 0.4504, "step": 1881 }, { "epoch": 0.13342786246012053, "grad_norm": 3.5094015345172056, "learning_rate": 9.722375024455993e-06, "loss": 0.4453, "step": 1882 }, { "epoch": 0.1334987593052109, "grad_norm": 1.0898387506046097, "learning_rate": 9.721997634113985e-06, "loss": 0.6519, "step": 1883 }, { "epoch": 0.1335696561503013, "grad_norm": 7.816526077370045, "learning_rate": 9.721619994777434e-06, "loss": 0.4473, "step": 1884 }, { "epoch": 0.1336405529953917, "grad_norm": 3.070559824611749, "learning_rate": 9.721242106466259e-06, "loss": 0.4295, "step": 1885 }, { "epoch": 0.1337114498404821, "grad_norm": 1.2459273808167264, "learning_rate": 9.720863969200383e-06, "loss": 0.7343, "step": 1886 }, { "epoch": 0.13378234668557248, "grad_norm": 4.91557537793074, "learning_rate": 9.720485582999749e-06, "loss": 0.4399, "step": 1887 }, { "epoch": 0.13385324353066289, "grad_norm": 6.7510023454070645, "learning_rate": 9.720106947884305e-06, "loss": 0.4675, "step": 1888 }, { "epoch": 0.1339241403757533, "grad_norm": 3.9987845440467304, "learning_rate": 9.719728063874019e-06, "loss": 0.4342, "step": 1889 }, { "epoch": 0.13399503722084366, "grad_norm": 2.6579220259392877, "learning_rate": 9.71934893098887e-06, "loss": 0.5276, "step": 1890 }, { "epoch": 0.13406593406593406, "grad_norm": 2.3394948507734634, "learning_rate": 9.718969549248848e-06, "loss": 0.462, "step": 1891 }, { "epoch": 0.13413683091102446, "grad_norm": 4.295468490629555, "learning_rate": 9.718589918673964e-06, "loss": 0.3939, "step": 1892 }, { "epoch": 0.13420772775611486, "grad_norm": 7.92345986692499, "learning_rate": 9.71821003928423e-06, "loss": 0.4186, "step": 1893 }, { "epoch": 0.13427862460120524, "grad_norm": 2.914053099778914, "learning_rate": 9.717829911099678e-06, "loss": 0.4739, "step": 1894 }, { "epoch": 0.13434952144629564, "grad_norm": 6.5654383928672715, "learning_rate": 9.717449534140356e-06, "loss": 0.4658, "step": 1895 }, { "epoch": 0.13442041829138604, "grad_norm": 2.3879813510708066, "learning_rate": 9.717068908426318e-06, "loss": 0.4278, "step": 1896 }, { "epoch": 0.1344913151364764, "grad_norm": 3.2994223369267037, "learning_rate": 9.716688033977638e-06, "loss": 0.4292, "step": 1897 }, { "epoch": 0.13456221198156681, "grad_norm": 2.542054814410013, "learning_rate": 9.716306910814395e-06, "loss": 0.4344, "step": 1898 }, { "epoch": 0.13463310882665722, "grad_norm": 3.370819187242318, "learning_rate": 9.71592553895669e-06, "loss": 0.4217, "step": 1899 }, { "epoch": 0.13470400567174762, "grad_norm": 3.5968319825164192, "learning_rate": 9.715543918424633e-06, "loss": 0.4696, "step": 1900 }, { "epoch": 0.134774902516838, "grad_norm": 4.325476976761579, "learning_rate": 9.715162049238345e-06, "loss": 0.438, "step": 1901 }, { "epoch": 0.1348457993619284, "grad_norm": 3.5045479783797218, "learning_rate": 9.714779931417962e-06, "loss": 0.5079, "step": 1902 }, { "epoch": 0.1349166962070188, "grad_norm": 6.810104597371626, "learning_rate": 9.714397564983637e-06, "loss": 0.4333, "step": 1903 }, { "epoch": 0.1349875930521092, "grad_norm": 5.461090576662943, "learning_rate": 9.714014949955528e-06, "loss": 0.4497, "step": 1904 }, { "epoch": 0.13505848989719957, "grad_norm": 3.7994988715711755, "learning_rate": 9.713632086353815e-06, "loss": 0.379, "step": 1905 }, { "epoch": 0.13512938674228997, "grad_norm": 2.2363798045313734, "learning_rate": 9.713248974198681e-06, "loss": 0.385, "step": 1906 }, { "epoch": 0.13520028358738037, "grad_norm": 5.571288822868222, "learning_rate": 9.712865613510333e-06, "loss": 0.3985, "step": 1907 }, { "epoch": 0.13527118043247074, "grad_norm": 2.2573194943312034, "learning_rate": 9.712482004308985e-06, "loss": 0.4584, "step": 1908 }, { "epoch": 0.13534207727756115, "grad_norm": 2.7227638224113724, "learning_rate": 9.712098146614863e-06, "loss": 0.389, "step": 1909 }, { "epoch": 0.13541297412265155, "grad_norm": 3.527245801327087, "learning_rate": 9.71171404044821e-06, "loss": 0.4506, "step": 1910 }, { "epoch": 0.13548387096774195, "grad_norm": 3.063653340998221, "learning_rate": 9.711329685829278e-06, "loss": 0.4106, "step": 1911 }, { "epoch": 0.13555476781283232, "grad_norm": 2.466379680225676, "learning_rate": 9.710945082778336e-06, "loss": 0.4225, "step": 1912 }, { "epoch": 0.13562566465792272, "grad_norm": 3.3684831423231745, "learning_rate": 9.710560231315664e-06, "loss": 0.4372, "step": 1913 }, { "epoch": 0.13569656150301312, "grad_norm": 2.2893481715538573, "learning_rate": 9.710175131461557e-06, "loss": 0.4926, "step": 1914 }, { "epoch": 0.1357674583481035, "grad_norm": 3.1850281460887286, "learning_rate": 9.70978978323632e-06, "loss": 0.476, "step": 1915 }, { "epoch": 0.1358383551931939, "grad_norm": 3.5530793010312514, "learning_rate": 9.709404186660271e-06, "loss": 0.4299, "step": 1916 }, { "epoch": 0.1359092520382843, "grad_norm": 5.122965755112162, "learning_rate": 9.709018341753746e-06, "loss": 0.439, "step": 1917 }, { "epoch": 0.1359801488833747, "grad_norm": 2.4528103261809573, "learning_rate": 9.708632248537088e-06, "loss": 0.4569, "step": 1918 }, { "epoch": 0.13605104572846508, "grad_norm": 5.850563450396137, "learning_rate": 9.70824590703066e-06, "loss": 0.4438, "step": 1919 }, { "epoch": 0.13612194257355548, "grad_norm": 3.359235036127092, "learning_rate": 9.70785931725483e-06, "loss": 0.4533, "step": 1920 }, { "epoch": 0.13619283941864588, "grad_norm": 2.862781526959091, "learning_rate": 9.707472479229986e-06, "loss": 0.4295, "step": 1921 }, { "epoch": 0.13626373626373625, "grad_norm": 1.9573887078352916, "learning_rate": 9.707085392976526e-06, "loss": 0.4219, "step": 1922 }, { "epoch": 0.13633463310882665, "grad_norm": 3.147171324380509, "learning_rate": 9.706698058514858e-06, "loss": 0.4651, "step": 1923 }, { "epoch": 0.13640552995391705, "grad_norm": 4.003603748518548, "learning_rate": 9.70631047586541e-06, "loss": 0.4675, "step": 1924 }, { "epoch": 0.13647642679900746, "grad_norm": 2.374135637605359, "learning_rate": 9.705922645048618e-06, "loss": 0.4301, "step": 1925 }, { "epoch": 0.13654732364409783, "grad_norm": 2.832634439476655, "learning_rate": 9.705534566084934e-06, "loss": 0.483, "step": 1926 }, { "epoch": 0.13661822048918823, "grad_norm": 2.6174414913957644, "learning_rate": 9.70514623899482e-06, "loss": 0.468, "step": 1927 }, { "epoch": 0.13668911733427863, "grad_norm": 12.5664625447355, "learning_rate": 9.704757663798755e-06, "loss": 0.4821, "step": 1928 }, { "epoch": 0.136760014179369, "grad_norm": 3.1931593643216125, "learning_rate": 9.704368840517225e-06, "loss": 0.4846, "step": 1929 }, { "epoch": 0.1368309110244594, "grad_norm": 6.6975090361278955, "learning_rate": 9.703979769170738e-06, "loss": 0.4551, "step": 1930 }, { "epoch": 0.1369018078695498, "grad_norm": 3.8512384430231363, "learning_rate": 9.703590449779805e-06, "loss": 0.4308, "step": 1931 }, { "epoch": 0.1369727047146402, "grad_norm": 3.287687153023622, "learning_rate": 9.703200882364959e-06, "loss": 0.3918, "step": 1932 }, { "epoch": 0.13704360155973058, "grad_norm": 2.5135564241533865, "learning_rate": 9.70281106694674e-06, "loss": 0.438, "step": 1933 }, { "epoch": 0.13711449840482098, "grad_norm": 3.369295009604545, "learning_rate": 9.702421003545703e-06, "loss": 0.4468, "step": 1934 }, { "epoch": 0.13718539524991138, "grad_norm": 14.29390769672712, "learning_rate": 9.702030692182419e-06, "loss": 0.4886, "step": 1935 }, { "epoch": 0.1372562920950018, "grad_norm": 3.942367764894751, "learning_rate": 9.701640132877468e-06, "loss": 0.4131, "step": 1936 }, { "epoch": 0.13732718894009216, "grad_norm": 5.336392063887494, "learning_rate": 9.701249325651443e-06, "loss": 0.4593, "step": 1937 }, { "epoch": 0.13739808578518256, "grad_norm": 6.324314259309427, "learning_rate": 9.700858270524953e-06, "loss": 0.4604, "step": 1938 }, { "epoch": 0.13746898263027296, "grad_norm": 1.2027256548502259, "learning_rate": 9.700466967518619e-06, "loss": 0.6004, "step": 1939 }, { "epoch": 0.13753987947536334, "grad_norm": 2.6004790217411053, "learning_rate": 9.700075416653071e-06, "loss": 0.49, "step": 1940 }, { "epoch": 0.13761077632045374, "grad_norm": 2.8410165544283403, "learning_rate": 9.699683617948963e-06, "loss": 0.464, "step": 1941 }, { "epoch": 0.13768167316554414, "grad_norm": 2.092292660226237, "learning_rate": 9.69929157142695e-06, "loss": 0.3926, "step": 1942 }, { "epoch": 0.13775257001063454, "grad_norm": 1.1935831369658885, "learning_rate": 9.698899277107705e-06, "loss": 0.6369, "step": 1943 }, { "epoch": 0.1378234668557249, "grad_norm": 3.8881671177393193, "learning_rate": 9.698506735011913e-06, "loss": 0.4656, "step": 1944 }, { "epoch": 0.13789436370081531, "grad_norm": 2.631620297620248, "learning_rate": 9.698113945160276e-06, "loss": 0.4466, "step": 1945 }, { "epoch": 0.13796526054590572, "grad_norm": 3.892281000374878, "learning_rate": 9.697720907573504e-06, "loss": 0.4715, "step": 1946 }, { "epoch": 0.1380361573909961, "grad_norm": 3.0945114376363585, "learning_rate": 9.697327622272326e-06, "loss": 0.4577, "step": 1947 }, { "epoch": 0.1381070542360865, "grad_norm": 2.8866163947626307, "learning_rate": 9.696934089277474e-06, "loss": 0.4393, "step": 1948 }, { "epoch": 0.1381779510811769, "grad_norm": 2.5526755032951476, "learning_rate": 9.696540308609704e-06, "loss": 0.452, "step": 1949 }, { "epoch": 0.1382488479262673, "grad_norm": 3.7316899021701473, "learning_rate": 9.696146280289778e-06, "loss": 0.4261, "step": 1950 }, { "epoch": 0.13831974477135767, "grad_norm": 3.325256968090211, "learning_rate": 9.695752004338474e-06, "loss": 0.4534, "step": 1951 }, { "epoch": 0.13839064161644807, "grad_norm": 5.520098072816761, "learning_rate": 9.695357480776583e-06, "loss": 0.4759, "step": 1952 }, { "epoch": 0.13846153846153847, "grad_norm": 2.833565971437745, "learning_rate": 9.694962709624908e-06, "loss": 0.4455, "step": 1953 }, { "epoch": 0.13853243530662884, "grad_norm": 3.7130567587677046, "learning_rate": 9.694567690904267e-06, "loss": 0.4392, "step": 1954 }, { "epoch": 0.13860333215171924, "grad_norm": 2.916159040815516, "learning_rate": 9.694172424635487e-06, "loss": 0.4959, "step": 1955 }, { "epoch": 0.13867422899680965, "grad_norm": 2.889762113361714, "learning_rate": 9.693776910839414e-06, "loss": 0.5003, "step": 1956 }, { "epoch": 0.13874512584190005, "grad_norm": 2.397978292271258, "learning_rate": 9.6933811495369e-06, "loss": 0.4331, "step": 1957 }, { "epoch": 0.13881602268699042, "grad_norm": 2.2092241983156113, "learning_rate": 9.692985140748816e-06, "loss": 0.4217, "step": 1958 }, { "epoch": 0.13888691953208082, "grad_norm": 4.859291950050823, "learning_rate": 9.692588884496044e-06, "loss": 0.4405, "step": 1959 }, { "epoch": 0.13895781637717122, "grad_norm": 3.2582262396631556, "learning_rate": 9.692192380799479e-06, "loss": 0.3939, "step": 1960 }, { "epoch": 0.1390287132222616, "grad_norm": 2.981504828375045, "learning_rate": 9.691795629680028e-06, "loss": 0.5093, "step": 1961 }, { "epoch": 0.139099610067352, "grad_norm": 3.112341208915163, "learning_rate": 9.691398631158611e-06, "loss": 0.4626, "step": 1962 }, { "epoch": 0.1391705069124424, "grad_norm": 3.712827959392285, "learning_rate": 9.691001385256165e-06, "loss": 0.4305, "step": 1963 }, { "epoch": 0.1392414037575328, "grad_norm": 3.399675917879026, "learning_rate": 9.690603891993636e-06, "loss": 0.4971, "step": 1964 }, { "epoch": 0.13931230060262317, "grad_norm": 2.98740278841692, "learning_rate": 9.690206151391981e-06, "loss": 0.4108, "step": 1965 }, { "epoch": 0.13938319744771357, "grad_norm": 5.121485019394617, "learning_rate": 9.689808163472177e-06, "loss": 0.4465, "step": 1966 }, { "epoch": 0.13945409429280398, "grad_norm": 3.4390935501829722, "learning_rate": 9.68940992825521e-06, "loss": 0.4101, "step": 1967 }, { "epoch": 0.13952499113789438, "grad_norm": 2.3872639707147525, "learning_rate": 9.689011445762078e-06, "loss": 0.4453, "step": 1968 }, { "epoch": 0.13959588798298475, "grad_norm": 2.7125224529210508, "learning_rate": 9.688612716013793e-06, "loss": 0.3981, "step": 1969 }, { "epoch": 0.13966678482807515, "grad_norm": 5.305216585498361, "learning_rate": 9.68821373903138e-06, "loss": 0.4964, "step": 1970 }, { "epoch": 0.13973768167316555, "grad_norm": 5.192582591505528, "learning_rate": 9.687814514835879e-06, "loss": 0.4591, "step": 1971 }, { "epoch": 0.13980857851825593, "grad_norm": 2.911695488928302, "learning_rate": 9.68741504344834e-06, "loss": 0.4034, "step": 1972 }, { "epoch": 0.13987947536334633, "grad_norm": 4.949892150730437, "learning_rate": 9.68701532488983e-06, "loss": 0.4365, "step": 1973 }, { "epoch": 0.13995037220843673, "grad_norm": 4.504860060482682, "learning_rate": 9.686615359181424e-06, "loss": 0.425, "step": 1974 }, { "epoch": 0.14002126905352713, "grad_norm": 4.1402111281566425, "learning_rate": 9.686215146344212e-06, "loss": 0.4323, "step": 1975 }, { "epoch": 0.1400921658986175, "grad_norm": 2.2371873967540954, "learning_rate": 9.685814686399298e-06, "loss": 0.3943, "step": 1976 }, { "epoch": 0.1401630627437079, "grad_norm": 3.2484705235818354, "learning_rate": 9.685413979367799e-06, "loss": 0.4606, "step": 1977 }, { "epoch": 0.1402339595887983, "grad_norm": 3.6330164646408, "learning_rate": 9.685013025270847e-06, "loss": 0.4107, "step": 1978 }, { "epoch": 0.14030485643388868, "grad_norm": 3.695900164458554, "learning_rate": 9.68461182412958e-06, "loss": 0.438, "step": 1979 }, { "epoch": 0.14037575327897908, "grad_norm": 1.097579123126947, "learning_rate": 9.684210375965156e-06, "loss": 0.6811, "step": 1980 }, { "epoch": 0.14044665012406948, "grad_norm": 4.896770400244484, "learning_rate": 9.683808680798744e-06, "loss": 0.4141, "step": 1981 }, { "epoch": 0.14051754696915988, "grad_norm": 2.7118734551476047, "learning_rate": 9.683406738651525e-06, "loss": 0.375, "step": 1982 }, { "epoch": 0.14058844381425026, "grad_norm": 5.956986946473448, "learning_rate": 9.683004549544694e-06, "loss": 0.4559, "step": 1983 }, { "epoch": 0.14065934065934066, "grad_norm": 4.757242149988023, "learning_rate": 9.68260211349946e-06, "loss": 0.4785, "step": 1984 }, { "epoch": 0.14073023750443106, "grad_norm": 2.5400490319364133, "learning_rate": 9.682199430537041e-06, "loss": 0.4745, "step": 1985 }, { "epoch": 0.14080113434952143, "grad_norm": 3.1796049587296054, "learning_rate": 9.681796500678673e-06, "loss": 0.4616, "step": 1986 }, { "epoch": 0.14087203119461184, "grad_norm": 2.7835354141398208, "learning_rate": 9.681393323945602e-06, "loss": 0.5124, "step": 1987 }, { "epoch": 0.14094292803970224, "grad_norm": 3.2623221564674396, "learning_rate": 9.680989900359088e-06, "loss": 0.4667, "step": 1988 }, { "epoch": 0.14101382488479264, "grad_norm": 2.547001541053247, "learning_rate": 9.680586229940402e-06, "loss": 0.4207, "step": 1989 }, { "epoch": 0.141084721729883, "grad_norm": 3.6217369399232253, "learning_rate": 9.680182312710835e-06, "loss": 0.4735, "step": 1990 }, { "epoch": 0.1411556185749734, "grad_norm": 3.4428719667189047, "learning_rate": 9.67977814869168e-06, "loss": 0.4572, "step": 1991 }, { "epoch": 0.14122651542006381, "grad_norm": 3.720173703926067, "learning_rate": 9.679373737904252e-06, "loss": 0.447, "step": 1992 }, { "epoch": 0.1412974122651542, "grad_norm": 3.371542445226375, "learning_rate": 9.678969080369876e-06, "loss": 0.4553, "step": 1993 }, { "epoch": 0.1413683091102446, "grad_norm": 3.4732992333584045, "learning_rate": 9.678564176109887e-06, "loss": 0.4081, "step": 1994 }, { "epoch": 0.141439205955335, "grad_norm": 3.9112438625268138, "learning_rate": 9.67815902514564e-06, "loss": 0.4466, "step": 1995 }, { "epoch": 0.1415101028004254, "grad_norm": 4.377783463289379, "learning_rate": 9.677753627498494e-06, "loss": 0.4685, "step": 1996 }, { "epoch": 0.14158099964551576, "grad_norm": 3.652930625792541, "learning_rate": 9.677347983189833e-06, "loss": 0.4496, "step": 1997 }, { "epoch": 0.14165189649060617, "grad_norm": 4.9604641417827935, "learning_rate": 9.67694209224104e-06, "loss": 0.4977, "step": 1998 }, { "epoch": 0.14172279333569657, "grad_norm": 4.4431185495455265, "learning_rate": 9.67653595467352e-06, "loss": 0.452, "step": 1999 }, { "epoch": 0.14179369018078697, "grad_norm": 5.29494428354799, "learning_rate": 9.676129570508692e-06, "loss": 0.4409, "step": 2000 }, { "epoch": 0.14186458702587734, "grad_norm": 4.049114719054791, "learning_rate": 9.675722939767983e-06, "loss": 0.4062, "step": 2001 }, { "epoch": 0.14193548387096774, "grad_norm": 6.22109904254236, "learning_rate": 9.675316062472833e-06, "loss": 0.4769, "step": 2002 }, { "epoch": 0.14200638071605814, "grad_norm": 2.779733163358284, "learning_rate": 9.674908938644698e-06, "loss": 0.4255, "step": 2003 }, { "epoch": 0.14207727756114852, "grad_norm": 3.3037889477752, "learning_rate": 9.674501568305048e-06, "loss": 0.4648, "step": 2004 }, { "epoch": 0.14214817440623892, "grad_norm": 4.741343871786943, "learning_rate": 9.67409395147536e-06, "loss": 0.4341, "step": 2005 }, { "epoch": 0.14221907125132932, "grad_norm": 3.7370416174926495, "learning_rate": 9.673686088177133e-06, "loss": 0.4679, "step": 2006 }, { "epoch": 0.14228996809641972, "grad_norm": 2.9774436794914054, "learning_rate": 9.67327797843187e-06, "loss": 0.4529, "step": 2007 }, { "epoch": 0.1423608649415101, "grad_norm": 3.5775769532585318, "learning_rate": 9.672869622261093e-06, "loss": 0.4697, "step": 2008 }, { "epoch": 0.1424317617866005, "grad_norm": 4.9727463735102955, "learning_rate": 9.672461019686334e-06, "loss": 0.4598, "step": 2009 }, { "epoch": 0.1425026586316909, "grad_norm": 3.682847356693891, "learning_rate": 9.672052170729139e-06, "loss": 0.3975, "step": 2010 }, { "epoch": 0.14257355547678127, "grad_norm": 2.4251156907504243, "learning_rate": 9.671643075411065e-06, "loss": 0.4423, "step": 2011 }, { "epoch": 0.14264445232187167, "grad_norm": 2.82655487685971, "learning_rate": 9.671233733753688e-06, "loss": 0.43, "step": 2012 }, { "epoch": 0.14271534916696207, "grad_norm": 3.9802876547443766, "learning_rate": 9.670824145778591e-06, "loss": 0.4282, "step": 2013 }, { "epoch": 0.14278624601205248, "grad_norm": 3.744620603852119, "learning_rate": 9.670414311507372e-06, "loss": 0.4013, "step": 2014 }, { "epoch": 0.14285714285714285, "grad_norm": 2.508358740037614, "learning_rate": 9.67000423096164e-06, "loss": 0.3948, "step": 2015 }, { "epoch": 0.14292803970223325, "grad_norm": 7.8741990124790044, "learning_rate": 9.669593904163023e-06, "loss": 0.4708, "step": 2016 }, { "epoch": 0.14299893654732365, "grad_norm": 3.6594972405334856, "learning_rate": 9.669183331133154e-06, "loss": 0.4358, "step": 2017 }, { "epoch": 0.14306983339241403, "grad_norm": 3.085917596028178, "learning_rate": 9.668772511893682e-06, "loss": 0.4693, "step": 2018 }, { "epoch": 0.14314073023750443, "grad_norm": 5.606716808928554, "learning_rate": 9.668361446466274e-06, "loss": 0.4873, "step": 2019 }, { "epoch": 0.14321162708259483, "grad_norm": 3.077667235062897, "learning_rate": 9.667950134872605e-06, "loss": 0.4039, "step": 2020 }, { "epoch": 0.14328252392768523, "grad_norm": 3.466148890671022, "learning_rate": 9.667538577134361e-06, "loss": 0.3774, "step": 2021 }, { "epoch": 0.1433534207727756, "grad_norm": 2.384677922294861, "learning_rate": 9.667126773273246e-06, "loss": 0.4111, "step": 2022 }, { "epoch": 0.143424317617866, "grad_norm": 4.887888253662324, "learning_rate": 9.666714723310973e-06, "loss": 0.4254, "step": 2023 }, { "epoch": 0.1434952144629564, "grad_norm": 8.671471428469262, "learning_rate": 9.66630242726927e-06, "loss": 0.446, "step": 2024 }, { "epoch": 0.14356611130804678, "grad_norm": 3.828610155171584, "learning_rate": 9.665889885169879e-06, "loss": 0.4194, "step": 2025 }, { "epoch": 0.14363700815313718, "grad_norm": 3.1856561601570856, "learning_rate": 9.665477097034554e-06, "loss": 0.44, "step": 2026 }, { "epoch": 0.14370790499822758, "grad_norm": 2.62289232167241, "learning_rate": 9.66506406288506e-06, "loss": 0.4578, "step": 2027 }, { "epoch": 0.14377880184331798, "grad_norm": 3.361888406204598, "learning_rate": 9.664650782743176e-06, "loss": 0.4416, "step": 2028 }, { "epoch": 0.14384969868840836, "grad_norm": 3.6276133791432277, "learning_rate": 9.6642372566307e-06, "loss": 0.4227, "step": 2029 }, { "epoch": 0.14392059553349876, "grad_norm": 3.1788888008525173, "learning_rate": 9.66382348456943e-06, "loss": 0.4532, "step": 2030 }, { "epoch": 0.14399149237858916, "grad_norm": 3.117543020734669, "learning_rate": 9.663409466581186e-06, "loss": 0.432, "step": 2031 }, { "epoch": 0.14406238922367956, "grad_norm": 3.608598522085734, "learning_rate": 9.662995202687807e-06, "loss": 0.4638, "step": 2032 }, { "epoch": 0.14413328606876993, "grad_norm": 2.6780106794741876, "learning_rate": 9.662580692911128e-06, "loss": 0.4652, "step": 2033 }, { "epoch": 0.14420418291386033, "grad_norm": 2.8502341212919764, "learning_rate": 9.66216593727301e-06, "loss": 0.4485, "step": 2034 }, { "epoch": 0.14427507975895074, "grad_norm": 3.5296041194591057, "learning_rate": 9.661750935795325e-06, "loss": 0.4915, "step": 2035 }, { "epoch": 0.1443459766040411, "grad_norm": 30.900847666820482, "learning_rate": 9.661335688499956e-06, "loss": 0.4914, "step": 2036 }, { "epoch": 0.1444168734491315, "grad_norm": 4.188668080806773, "learning_rate": 9.660920195408796e-06, "loss": 0.474, "step": 2037 }, { "epoch": 0.1444877702942219, "grad_norm": 3.7687358069276775, "learning_rate": 9.660504456543758e-06, "loss": 0.413, "step": 2038 }, { "epoch": 0.1445586671393123, "grad_norm": 2.695522958928373, "learning_rate": 9.660088471926761e-06, "loss": 0.3725, "step": 2039 }, { "epoch": 0.1446295639844027, "grad_norm": 3.4771441019705818, "learning_rate": 9.659672241579745e-06, "loss": 0.386, "step": 2040 }, { "epoch": 0.1447004608294931, "grad_norm": 2.730203896140307, "learning_rate": 9.659255765524653e-06, "loss": 0.3932, "step": 2041 }, { "epoch": 0.1447713576745835, "grad_norm": 3.3920932987340766, "learning_rate": 9.658839043783448e-06, "loss": 0.4311, "step": 2042 }, { "epoch": 0.14484225451967386, "grad_norm": 2.934540892785038, "learning_rate": 9.658422076378105e-06, "loss": 0.445, "step": 2043 }, { "epoch": 0.14491315136476426, "grad_norm": 1.8422806195800088, "learning_rate": 9.65800486333061e-06, "loss": 0.4313, "step": 2044 }, { "epoch": 0.14498404820985467, "grad_norm": 3.249560099019359, "learning_rate": 9.657587404662965e-06, "loss": 0.4792, "step": 2045 }, { "epoch": 0.14505494505494507, "grad_norm": 14.298201039122004, "learning_rate": 9.65716970039718e-06, "loss": 0.4103, "step": 2046 }, { "epoch": 0.14512584190003544, "grad_norm": 3.455446637907929, "learning_rate": 9.656751750555283e-06, "loss": 0.4341, "step": 2047 }, { "epoch": 0.14519673874512584, "grad_norm": 6.624007628075833, "learning_rate": 9.65633355515931e-06, "loss": 0.4694, "step": 2048 }, { "epoch": 0.14526763559021624, "grad_norm": 3.1859873120549556, "learning_rate": 9.655915114231315e-06, "loss": 0.459, "step": 2049 }, { "epoch": 0.14533853243530662, "grad_norm": 2.3702010193641967, "learning_rate": 9.655496427793363e-06, "loss": 0.466, "step": 2050 }, { "epoch": 0.14540942928039702, "grad_norm": 3.8701212889105214, "learning_rate": 9.65507749586753e-06, "loss": 0.4455, "step": 2051 }, { "epoch": 0.14548032612548742, "grad_norm": 3.1125433854043534, "learning_rate": 9.654658318475909e-06, "loss": 0.3952, "step": 2052 }, { "epoch": 0.14555122297057782, "grad_norm": 4.783616609712575, "learning_rate": 9.654238895640599e-06, "loss": 0.4894, "step": 2053 }, { "epoch": 0.1456221198156682, "grad_norm": 3.484607406492356, "learning_rate": 9.653819227383722e-06, "loss": 0.4497, "step": 2054 }, { "epoch": 0.1456930166607586, "grad_norm": 7.41220332195646, "learning_rate": 9.653399313727406e-06, "loss": 0.4537, "step": 2055 }, { "epoch": 0.145763913505849, "grad_norm": 5.343875071254517, "learning_rate": 9.652979154693791e-06, "loss": 0.4499, "step": 2056 }, { "epoch": 0.14583481035093937, "grad_norm": 3.194890562757104, "learning_rate": 9.652558750305034e-06, "loss": 0.4867, "step": 2057 }, { "epoch": 0.14590570719602977, "grad_norm": 2.5100984205381334, "learning_rate": 9.652138100583303e-06, "loss": 0.4053, "step": 2058 }, { "epoch": 0.14597660404112017, "grad_norm": 4.978031130071487, "learning_rate": 9.65171720555078e-06, "loss": 0.4372, "step": 2059 }, { "epoch": 0.14604750088621057, "grad_norm": 3.6810217403194914, "learning_rate": 9.651296065229658e-06, "loss": 0.4635, "step": 2060 }, { "epoch": 0.14611839773130095, "grad_norm": 4.574375176494401, "learning_rate": 9.650874679642144e-06, "loss": 0.4415, "step": 2061 }, { "epoch": 0.14618929457639135, "grad_norm": 2.528287610616369, "learning_rate": 9.650453048810458e-06, "loss": 0.4313, "step": 2062 }, { "epoch": 0.14626019142148175, "grad_norm": 3.3087306799447993, "learning_rate": 9.650031172756832e-06, "loss": 0.4466, "step": 2063 }, { "epoch": 0.14633108826657215, "grad_norm": 3.4997465989621777, "learning_rate": 9.649609051503514e-06, "loss": 0.3985, "step": 2064 }, { "epoch": 0.14640198511166252, "grad_norm": 21.846313291432065, "learning_rate": 9.649186685072764e-06, "loss": 0.4575, "step": 2065 }, { "epoch": 0.14647288195675293, "grad_norm": 3.668320285090597, "learning_rate": 9.648764073486849e-06, "loss": 0.4714, "step": 2066 }, { "epoch": 0.14654377880184333, "grad_norm": 2.6647063695356734, "learning_rate": 9.648341216768056e-06, "loss": 0.4323, "step": 2067 }, { "epoch": 0.1466146756469337, "grad_norm": 3.538126279469846, "learning_rate": 9.647918114938685e-06, "loss": 0.4621, "step": 2068 }, { "epoch": 0.1466855724920241, "grad_norm": 2.12015884355427, "learning_rate": 9.647494768021042e-06, "loss": 0.4512, "step": 2069 }, { "epoch": 0.1467564693371145, "grad_norm": 2.8611754562284535, "learning_rate": 9.647071176037453e-06, "loss": 0.3943, "step": 2070 }, { "epoch": 0.1468273661822049, "grad_norm": 5.369554242392568, "learning_rate": 9.646647339010255e-06, "loss": 0.4797, "step": 2071 }, { "epoch": 0.14689826302729528, "grad_norm": 2.5429466395160802, "learning_rate": 9.646223256961795e-06, "loss": 0.4332, "step": 2072 }, { "epoch": 0.14696915987238568, "grad_norm": 2.9415262164856952, "learning_rate": 9.645798929914438e-06, "loss": 0.3882, "step": 2073 }, { "epoch": 0.14704005671747608, "grad_norm": 8.65387504062547, "learning_rate": 9.645374357890556e-06, "loss": 0.483, "step": 2074 }, { "epoch": 0.14711095356256645, "grad_norm": 3.9867639645053985, "learning_rate": 9.644949540912539e-06, "loss": 0.4283, "step": 2075 }, { "epoch": 0.14718185040765686, "grad_norm": 3.19814341762265, "learning_rate": 9.644524479002788e-06, "loss": 0.4589, "step": 2076 }, { "epoch": 0.14725274725274726, "grad_norm": 2.823415073576907, "learning_rate": 9.644099172183714e-06, "loss": 0.4856, "step": 2077 }, { "epoch": 0.14732364409783766, "grad_norm": 5.780402811374145, "learning_rate": 9.643673620477746e-06, "loss": 0.4279, "step": 2078 }, { "epoch": 0.14739454094292803, "grad_norm": 3.2082278297514253, "learning_rate": 9.643247823907326e-06, "loss": 0.5535, "step": 2079 }, { "epoch": 0.14746543778801843, "grad_norm": 2.189645917920863, "learning_rate": 9.642821782494904e-06, "loss": 0.4237, "step": 2080 }, { "epoch": 0.14753633463310883, "grad_norm": 3.3166300274913585, "learning_rate": 9.642395496262945e-06, "loss": 0.4408, "step": 2081 }, { "epoch": 0.1476072314781992, "grad_norm": 11.919848407970225, "learning_rate": 9.641968965233928e-06, "loss": 0.4367, "step": 2082 }, { "epoch": 0.1476781283232896, "grad_norm": 2.2466304570613436, "learning_rate": 9.641542189430344e-06, "loss": 0.4029, "step": 2083 }, { "epoch": 0.14774902516838, "grad_norm": 3.54990049103763, "learning_rate": 9.641115168874698e-06, "loss": 0.4134, "step": 2084 }, { "epoch": 0.1478199220134704, "grad_norm": 3.674786435030841, "learning_rate": 9.640687903589507e-06, "loss": 0.4443, "step": 2085 }, { "epoch": 0.14789081885856079, "grad_norm": 5.089596545778253, "learning_rate": 9.6402603935973e-06, "loss": 0.3942, "step": 2086 }, { "epoch": 0.1479617157036512, "grad_norm": 3.9062298370121753, "learning_rate": 9.639832638920623e-06, "loss": 0.506, "step": 2087 }, { "epoch": 0.1480326125487416, "grad_norm": 3.242629176512814, "learning_rate": 9.639404639582026e-06, "loss": 0.4609, "step": 2088 }, { "epoch": 0.14810350939383196, "grad_norm": 3.948995826062066, "learning_rate": 9.638976395604083e-06, "loss": 0.4967, "step": 2089 }, { "epoch": 0.14817440623892236, "grad_norm": 2.525012270777662, "learning_rate": 9.638547907009374e-06, "loss": 0.4194, "step": 2090 }, { "epoch": 0.14824530308401276, "grad_norm": 3.0506198648703444, "learning_rate": 9.638119173820495e-06, "loss": 0.4663, "step": 2091 }, { "epoch": 0.14831619992910317, "grad_norm": 3.176012878231604, "learning_rate": 9.63769019606005e-06, "loss": 0.4524, "step": 2092 }, { "epoch": 0.14838709677419354, "grad_norm": 40.26557843795642, "learning_rate": 9.63726097375066e-06, "loss": 0.5063, "step": 2093 }, { "epoch": 0.14845799361928394, "grad_norm": 2.6976945223679007, "learning_rate": 9.63683150691496e-06, "loss": 0.4476, "step": 2094 }, { "epoch": 0.14852889046437434, "grad_norm": 2.8380677367710407, "learning_rate": 9.636401795575598e-06, "loss": 0.4726, "step": 2095 }, { "epoch": 0.14859978730946474, "grad_norm": 3.1225978658183413, "learning_rate": 9.63597183975523e-06, "loss": 0.4905, "step": 2096 }, { "epoch": 0.14867068415455512, "grad_norm": 2.48301471893311, "learning_rate": 9.635541639476527e-06, "loss": 0.4332, "step": 2097 }, { "epoch": 0.14874158099964552, "grad_norm": 3.872703700996478, "learning_rate": 9.635111194762174e-06, "loss": 0.406, "step": 2098 }, { "epoch": 0.14881247784473592, "grad_norm": 3.7828774463870314, "learning_rate": 9.634680505634873e-06, "loss": 0.5076, "step": 2099 }, { "epoch": 0.1488833746898263, "grad_norm": 4.8646343243683265, "learning_rate": 9.63424957211733e-06, "loss": 0.4745, "step": 2100 }, { "epoch": 0.1489542715349167, "grad_norm": 6.447990045861704, "learning_rate": 9.633818394232268e-06, "loss": 0.4703, "step": 2101 }, { "epoch": 0.1490251683800071, "grad_norm": 3.515503929971585, "learning_rate": 9.633386972002428e-06, "loss": 0.4702, "step": 2102 }, { "epoch": 0.1490960652250975, "grad_norm": 2.879955228845537, "learning_rate": 9.632955305450555e-06, "loss": 0.4826, "step": 2103 }, { "epoch": 0.14916696207018787, "grad_norm": 6.388301560494688, "learning_rate": 9.632523394599412e-06, "loss": 0.446, "step": 2104 }, { "epoch": 0.14923785891527827, "grad_norm": 2.77800228536965, "learning_rate": 9.632091239471775e-06, "loss": 0.4609, "step": 2105 }, { "epoch": 0.14930875576036867, "grad_norm": 2.855935976392228, "learning_rate": 9.631658840090433e-06, "loss": 0.3963, "step": 2106 }, { "epoch": 0.14937965260545905, "grad_norm": 3.7768929440324235, "learning_rate": 9.631226196478182e-06, "loss": 0.4923, "step": 2107 }, { "epoch": 0.14945054945054945, "grad_norm": 3.991078058197377, "learning_rate": 9.630793308657842e-06, "loss": 0.4432, "step": 2108 }, { "epoch": 0.14952144629563985, "grad_norm": 4.009054396302052, "learning_rate": 9.630360176652235e-06, "loss": 0.378, "step": 2109 }, { "epoch": 0.14959234314073025, "grad_norm": 5.039705486450688, "learning_rate": 9.629926800484203e-06, "loss": 0.4504, "step": 2110 }, { "epoch": 0.14966323998582062, "grad_norm": 2.759767038545442, "learning_rate": 9.629493180176595e-06, "loss": 0.4387, "step": 2111 }, { "epoch": 0.14973413683091102, "grad_norm": 5.258748162978227, "learning_rate": 9.62905931575228e-06, "loss": 0.4653, "step": 2112 }, { "epoch": 0.14980503367600143, "grad_norm": 3.1608133304327746, "learning_rate": 9.628625207234132e-06, "loss": 0.4616, "step": 2113 }, { "epoch": 0.1498759305210918, "grad_norm": 3.697287037569946, "learning_rate": 9.628190854645046e-06, "loss": 0.4873, "step": 2114 }, { "epoch": 0.1499468273661822, "grad_norm": 2.4056218714773454, "learning_rate": 9.627756258007922e-06, "loss": 0.4484, "step": 2115 }, { "epoch": 0.1500177242112726, "grad_norm": 2.4714870330931964, "learning_rate": 9.627321417345679e-06, "loss": 0.4214, "step": 2116 }, { "epoch": 0.150088621056363, "grad_norm": 3.590316129460957, "learning_rate": 9.626886332681246e-06, "loss": 0.448, "step": 2117 }, { "epoch": 0.15015951790145338, "grad_norm": 2.6329934969699744, "learning_rate": 9.626451004037565e-06, "loss": 0.4643, "step": 2118 }, { "epoch": 0.15023041474654378, "grad_norm": 3.7809983771699174, "learning_rate": 9.62601543143759e-06, "loss": 0.4991, "step": 2119 }, { "epoch": 0.15030131159163418, "grad_norm": 2.5204530630271242, "learning_rate": 9.625579614904292e-06, "loss": 0.3851, "step": 2120 }, { "epoch": 0.15037220843672455, "grad_norm": 2.2989815664870203, "learning_rate": 9.62514355446065e-06, "loss": 0.4552, "step": 2121 }, { "epoch": 0.15044310528181495, "grad_norm": 4.289523727828256, "learning_rate": 9.62470725012966e-06, "loss": 0.4687, "step": 2122 }, { "epoch": 0.15051400212690536, "grad_norm": 3.0353608798441694, "learning_rate": 9.624270701934325e-06, "loss": 0.4604, "step": 2123 }, { "epoch": 0.15058489897199576, "grad_norm": 1.370949396085352, "learning_rate": 9.623833909897664e-06, "loss": 0.7448, "step": 2124 }, { "epoch": 0.15065579581708613, "grad_norm": 3.0993498936172927, "learning_rate": 9.623396874042715e-06, "loss": 0.4858, "step": 2125 }, { "epoch": 0.15072669266217653, "grad_norm": 5.521738585532693, "learning_rate": 9.62295959439252e-06, "loss": 0.4798, "step": 2126 }, { "epoch": 0.15079758950726693, "grad_norm": 2.3383470772503756, "learning_rate": 9.622522070970134e-06, "loss": 0.4584, "step": 2127 }, { "epoch": 0.15086848635235733, "grad_norm": 2.8993077810202705, "learning_rate": 9.622084303798633e-06, "loss": 0.4622, "step": 2128 }, { "epoch": 0.1509393831974477, "grad_norm": 3.778128300474704, "learning_rate": 9.621646292901098e-06, "loss": 0.4727, "step": 2129 }, { "epoch": 0.1510102800425381, "grad_norm": 3.9631130542295634, "learning_rate": 9.62120803830063e-06, "loss": 0.4943, "step": 2130 }, { "epoch": 0.1510811768876285, "grad_norm": 4.370646034794442, "learning_rate": 9.62076954002033e-06, "loss": 0.4541, "step": 2131 }, { "epoch": 0.15115207373271888, "grad_norm": 2.633766101688182, "learning_rate": 9.620330798083329e-06, "loss": 0.4473, "step": 2132 }, { "epoch": 0.15122297057780928, "grad_norm": 2.795241984920613, "learning_rate": 9.619891812512756e-06, "loss": 0.4486, "step": 2133 }, { "epoch": 0.15129386742289969, "grad_norm": 3.571795369260934, "learning_rate": 9.619452583331764e-06, "loss": 0.4674, "step": 2134 }, { "epoch": 0.1513647642679901, "grad_norm": 2.312691377694613, "learning_rate": 9.619013110563508e-06, "loss": 0.4571, "step": 2135 }, { "epoch": 0.15143566111308046, "grad_norm": 2.6069804021036216, "learning_rate": 9.618573394231168e-06, "loss": 0.4086, "step": 2136 }, { "epoch": 0.15150655795817086, "grad_norm": 2.7146786295682253, "learning_rate": 9.618133434357927e-06, "loss": 0.4696, "step": 2137 }, { "epoch": 0.15157745480326126, "grad_norm": 3.9875973237640254, "learning_rate": 9.617693230966984e-06, "loss": 0.4337, "step": 2138 }, { "epoch": 0.15164835164835164, "grad_norm": 2.941253318014692, "learning_rate": 9.617252784081554e-06, "loss": 0.4472, "step": 2139 }, { "epoch": 0.15171924849344204, "grad_norm": 3.824586186044479, "learning_rate": 9.61681209372486e-06, "loss": 0.4611, "step": 2140 }, { "epoch": 0.15179014533853244, "grad_norm": 4.662063104399959, "learning_rate": 9.61637115992014e-06, "loss": 0.4608, "step": 2141 }, { "epoch": 0.15186104218362284, "grad_norm": 2.168296947022114, "learning_rate": 9.615929982690646e-06, "loss": 0.467, "step": 2142 }, { "epoch": 0.15193193902871321, "grad_norm": 2.4506932734390703, "learning_rate": 9.61548856205964e-06, "loss": 0.3807, "step": 2143 }, { "epoch": 0.15200283587380362, "grad_norm": 2.3152983868924584, "learning_rate": 9.6150468980504e-06, "loss": 0.4787, "step": 2144 }, { "epoch": 0.15207373271889402, "grad_norm": 3.1658105175678384, "learning_rate": 9.614604990686213e-06, "loss": 0.3768, "step": 2145 }, { "epoch": 0.1521446295639844, "grad_norm": 2.797552170799452, "learning_rate": 9.614162839990385e-06, "loss": 0.4685, "step": 2146 }, { "epoch": 0.1522155264090748, "grad_norm": 2.9963471695736015, "learning_rate": 9.613720445986227e-06, "loss": 0.4991, "step": 2147 }, { "epoch": 0.1522864232541652, "grad_norm": 2.7706117072898624, "learning_rate": 9.613277808697068e-06, "loss": 0.4942, "step": 2148 }, { "epoch": 0.1523573200992556, "grad_norm": 5.941794552644157, "learning_rate": 9.61283492814625e-06, "loss": 0.3904, "step": 2149 }, { "epoch": 0.15242821694434597, "grad_norm": 1.1998592269166262, "learning_rate": 9.612391804357127e-06, "loss": 0.732, "step": 2150 }, { "epoch": 0.15249911378943637, "grad_norm": 3.1045409830846897, "learning_rate": 9.61194843735306e-06, "loss": 0.4532, "step": 2151 }, { "epoch": 0.15257001063452677, "grad_norm": 4.109184075900486, "learning_rate": 9.611504827157435e-06, "loss": 0.421, "step": 2152 }, { "epoch": 0.15264090747961714, "grad_norm": 3.4300478057081016, "learning_rate": 9.61106097379364e-06, "loss": 0.4414, "step": 2153 }, { "epoch": 0.15271180432470755, "grad_norm": 2.3430047353247616, "learning_rate": 9.610616877285077e-06, "loss": 0.4261, "step": 2154 }, { "epoch": 0.15278270116979795, "grad_norm": 2.2839125877318445, "learning_rate": 9.610172537655171e-06, "loss": 0.4265, "step": 2155 }, { "epoch": 0.15285359801488835, "grad_norm": 1.939855081804976, "learning_rate": 9.609727954927348e-06, "loss": 0.4548, "step": 2156 }, { "epoch": 0.15292449485997872, "grad_norm": 4.2268696676807105, "learning_rate": 9.609283129125048e-06, "loss": 0.4426, "step": 2157 }, { "epoch": 0.15299539170506912, "grad_norm": 3.218650315592088, "learning_rate": 9.608838060271735e-06, "loss": 0.4545, "step": 2158 }, { "epoch": 0.15306628855015952, "grad_norm": 7.421434348977305, "learning_rate": 9.60839274839087e-06, "loss": 0.4781, "step": 2159 }, { "epoch": 0.15313718539524993, "grad_norm": 4.012638063254305, "learning_rate": 9.607947193505939e-06, "loss": 0.4334, "step": 2160 }, { "epoch": 0.1532080822403403, "grad_norm": 3.6283031453031613, "learning_rate": 9.607501395640437e-06, "loss": 0.4591, "step": 2161 }, { "epoch": 0.1532789790854307, "grad_norm": 3.4535194121168042, "learning_rate": 9.607055354817868e-06, "loss": 0.4601, "step": 2162 }, { "epoch": 0.1533498759305211, "grad_norm": 3.4289561571583205, "learning_rate": 9.606609071061755e-06, "loss": 0.5437, "step": 2163 }, { "epoch": 0.15342077277561147, "grad_norm": 4.964128816419777, "learning_rate": 9.606162544395628e-06, "loss": 0.4577, "step": 2164 }, { "epoch": 0.15349166962070188, "grad_norm": 2.7742763063489857, "learning_rate": 9.605715774843034e-06, "loss": 0.4157, "step": 2165 }, { "epoch": 0.15356256646579228, "grad_norm": 2.5332233533242183, "learning_rate": 9.605268762427532e-06, "loss": 0.4322, "step": 2166 }, { "epoch": 0.15363346331088268, "grad_norm": 2.5660532334471275, "learning_rate": 9.604821507172694e-06, "loss": 0.5112, "step": 2167 }, { "epoch": 0.15370436015597305, "grad_norm": 3.5808033626465376, "learning_rate": 9.604374009102104e-06, "loss": 0.4558, "step": 2168 }, { "epoch": 0.15377525700106345, "grad_norm": 3.7784466795144414, "learning_rate": 9.603926268239357e-06, "loss": 0.4336, "step": 2169 }, { "epoch": 0.15384615384615385, "grad_norm": 2.6843898017450756, "learning_rate": 9.603478284608065e-06, "loss": 0.4662, "step": 2170 }, { "epoch": 0.15391705069124423, "grad_norm": 3.526159625308136, "learning_rate": 9.603030058231848e-06, "loss": 0.4248, "step": 2171 }, { "epoch": 0.15398794753633463, "grad_norm": 3.474567050890958, "learning_rate": 9.602581589134344e-06, "loss": 0.4958, "step": 2172 }, { "epoch": 0.15405884438142503, "grad_norm": 2.0388686570306733, "learning_rate": 9.602132877339198e-06, "loss": 0.4315, "step": 2173 }, { "epoch": 0.15412974122651543, "grad_norm": 1.1902393553424995, "learning_rate": 9.601683922870074e-06, "loss": 0.73, "step": 2174 }, { "epoch": 0.1542006380716058, "grad_norm": 3.121099670899724, "learning_rate": 9.601234725750645e-06, "loss": 0.4162, "step": 2175 }, { "epoch": 0.1542715349166962, "grad_norm": 3.115308949989836, "learning_rate": 9.600785286004595e-06, "loss": 0.398, "step": 2176 }, { "epoch": 0.1543424317617866, "grad_norm": 2.7317943785604677, "learning_rate": 9.600335603655625e-06, "loss": 0.4774, "step": 2177 }, { "epoch": 0.15441332860687698, "grad_norm": 2.356133624063734, "learning_rate": 9.59988567872745e-06, "loss": 0.4683, "step": 2178 }, { "epoch": 0.15448422545196738, "grad_norm": 4.153240062737246, "learning_rate": 9.59943551124379e-06, "loss": 0.4464, "step": 2179 }, { "epoch": 0.15455512229705778, "grad_norm": 4.484689430421896, "learning_rate": 9.598985101228387e-06, "loss": 0.42, "step": 2180 }, { "epoch": 0.15462601914214819, "grad_norm": 5.567685625050115, "learning_rate": 9.598534448704988e-06, "loss": 0.4738, "step": 2181 }, { "epoch": 0.15469691598723856, "grad_norm": 2.204853112386386, "learning_rate": 9.598083553697358e-06, "loss": 0.4454, "step": 2182 }, { "epoch": 0.15476781283232896, "grad_norm": 3.8203087965992055, "learning_rate": 9.597632416229272e-06, "loss": 0.4318, "step": 2183 }, { "epoch": 0.15483870967741936, "grad_norm": 2.8696743140699135, "learning_rate": 9.597181036324522e-06, "loss": 0.3904, "step": 2184 }, { "epoch": 0.15490960652250974, "grad_norm": 2.256729787657368, "learning_rate": 9.596729414006903e-06, "loss": 0.3965, "step": 2185 }, { "epoch": 0.15498050336760014, "grad_norm": 2.2002674217893743, "learning_rate": 9.596277549300236e-06, "loss": 0.4365, "step": 2186 }, { "epoch": 0.15505140021269054, "grad_norm": 2.213823835621817, "learning_rate": 9.595825442228345e-06, "loss": 0.4489, "step": 2187 }, { "epoch": 0.15512229705778094, "grad_norm": 2.679835215574931, "learning_rate": 9.595373092815071e-06, "loss": 0.4397, "step": 2188 }, { "epoch": 0.1551931939028713, "grad_norm": 4.372377724152731, "learning_rate": 9.594920501084266e-06, "loss": 0.4438, "step": 2189 }, { "epoch": 0.1552640907479617, "grad_norm": 3.2918707784416505, "learning_rate": 9.594467667059795e-06, "loss": 0.4697, "step": 2190 }, { "epoch": 0.15533498759305212, "grad_norm": 7.206074593579444, "learning_rate": 9.59401459076554e-06, "loss": 0.4553, "step": 2191 }, { "epoch": 0.15540588443814252, "grad_norm": 4.226046673494589, "learning_rate": 9.593561272225387e-06, "loss": 0.4369, "step": 2192 }, { "epoch": 0.1554767812832329, "grad_norm": 3.1175864723134485, "learning_rate": 9.593107711463241e-06, "loss": 0.4529, "step": 2193 }, { "epoch": 0.1555476781283233, "grad_norm": 2.4909230536113602, "learning_rate": 9.592653908503021e-06, "loss": 0.4736, "step": 2194 }, { "epoch": 0.1556185749734137, "grad_norm": 2.3785216623041423, "learning_rate": 9.592199863368655e-06, "loss": 0.4652, "step": 2195 }, { "epoch": 0.15568947181850407, "grad_norm": 5.84310976299953, "learning_rate": 9.591745576084085e-06, "loss": 0.4453, "step": 2196 }, { "epoch": 0.15576036866359447, "grad_norm": 3.1135522427130757, "learning_rate": 9.591291046673265e-06, "loss": 0.4162, "step": 2197 }, { "epoch": 0.15583126550868487, "grad_norm": 2.0821737259078845, "learning_rate": 9.590836275160166e-06, "loss": 0.4393, "step": 2198 }, { "epoch": 0.15590216235377527, "grad_norm": 5.956799266973211, "learning_rate": 9.590381261568763e-06, "loss": 0.4605, "step": 2199 }, { "epoch": 0.15597305919886564, "grad_norm": 8.412840311477785, "learning_rate": 9.589926005923056e-06, "loss": 0.488, "step": 2200 }, { "epoch": 0.15604395604395604, "grad_norm": 2.512299646079177, "learning_rate": 9.589470508247044e-06, "loss": 0.4277, "step": 2201 }, { "epoch": 0.15611485288904645, "grad_norm": 1.1991179813430388, "learning_rate": 9.589014768564753e-06, "loss": 0.707, "step": 2202 }, { "epoch": 0.15618574973413682, "grad_norm": 1.994280033912172, "learning_rate": 9.588558786900208e-06, "loss": 0.442, "step": 2203 }, { "epoch": 0.15625664657922722, "grad_norm": 5.077590160739669, "learning_rate": 9.588102563277456e-06, "loss": 0.4701, "step": 2204 }, { "epoch": 0.15632754342431762, "grad_norm": 2.106172992183369, "learning_rate": 9.587646097720555e-06, "loss": 0.4629, "step": 2205 }, { "epoch": 0.15639844026940802, "grad_norm": 2.9923165516979364, "learning_rate": 9.587189390253573e-06, "loss": 0.4105, "step": 2206 }, { "epoch": 0.1564693371144984, "grad_norm": 4.780802653551714, "learning_rate": 9.586732440900592e-06, "loss": 0.4537, "step": 2207 }, { "epoch": 0.1565402339595888, "grad_norm": 2.241722788389127, "learning_rate": 9.58627524968571e-06, "loss": 0.4176, "step": 2208 }, { "epoch": 0.1566111308046792, "grad_norm": 2.3956732572050474, "learning_rate": 9.585817816633033e-06, "loss": 0.4296, "step": 2209 }, { "epoch": 0.15668202764976957, "grad_norm": 3.538720181180387, "learning_rate": 9.585360141766683e-06, "loss": 0.431, "step": 2210 }, { "epoch": 0.15675292449485997, "grad_norm": 3.9669527903579413, "learning_rate": 9.584902225110792e-06, "loss": 0.4699, "step": 2211 }, { "epoch": 0.15682382133995038, "grad_norm": 2.876745769206115, "learning_rate": 9.584444066689507e-06, "loss": 0.4584, "step": 2212 }, { "epoch": 0.15689471818504078, "grad_norm": 2.456155465390959, "learning_rate": 9.58398566652699e-06, "loss": 0.4927, "step": 2213 }, { "epoch": 0.15696561503013115, "grad_norm": 4.241146032118097, "learning_rate": 9.583527024647405e-06, "loss": 0.4607, "step": 2214 }, { "epoch": 0.15703651187522155, "grad_norm": 2.1495572174626267, "learning_rate": 9.583068141074946e-06, "loss": 0.4551, "step": 2215 }, { "epoch": 0.15710740872031195, "grad_norm": 2.7356404332580215, "learning_rate": 9.582609015833804e-06, "loss": 0.4408, "step": 2216 }, { "epoch": 0.15717830556540233, "grad_norm": 2.918632234828579, "learning_rate": 9.58214964894819e-06, "loss": 0.4316, "step": 2217 }, { "epoch": 0.15724920241049273, "grad_norm": 2.435081105285365, "learning_rate": 9.581690040442328e-06, "loss": 0.45, "step": 2218 }, { "epoch": 0.15732009925558313, "grad_norm": 3.021445472458223, "learning_rate": 9.581230190340454e-06, "loss": 0.4031, "step": 2219 }, { "epoch": 0.15739099610067353, "grad_norm": 3.3416797732657293, "learning_rate": 9.580770098666817e-06, "loss": 0.4387, "step": 2220 }, { "epoch": 0.1574618929457639, "grad_norm": 4.062573604238772, "learning_rate": 9.580309765445673e-06, "loss": 0.4319, "step": 2221 }, { "epoch": 0.1575327897908543, "grad_norm": 4.005037432436096, "learning_rate": 9.579849190701301e-06, "loss": 0.4018, "step": 2222 }, { "epoch": 0.1576036866359447, "grad_norm": 3.8163618665116115, "learning_rate": 9.579388374457984e-06, "loss": 0.4845, "step": 2223 }, { "epoch": 0.1576745834810351, "grad_norm": 4.367469075656581, "learning_rate": 9.578927316740025e-06, "loss": 0.4528, "step": 2224 }, { "epoch": 0.15774548032612548, "grad_norm": 3.3161540309300253, "learning_rate": 9.578466017571733e-06, "loss": 0.4577, "step": 2225 }, { "epoch": 0.15781637717121588, "grad_norm": 5.454204459211929, "learning_rate": 9.578004476977432e-06, "loss": 0.4517, "step": 2226 }, { "epoch": 0.15788727401630628, "grad_norm": 3.166861471406081, "learning_rate": 9.57754269498146e-06, "loss": 0.4874, "step": 2227 }, { "epoch": 0.15795817086139666, "grad_norm": 2.601889138373645, "learning_rate": 9.57708067160817e-06, "loss": 0.4072, "step": 2228 }, { "epoch": 0.15802906770648706, "grad_norm": 2.8049034661400443, "learning_rate": 9.576618406881922e-06, "loss": 0.4032, "step": 2229 }, { "epoch": 0.15809996455157746, "grad_norm": 2.1733637725075763, "learning_rate": 9.576155900827093e-06, "loss": 0.4108, "step": 2230 }, { "epoch": 0.15817086139666786, "grad_norm": 2.1563239329679726, "learning_rate": 9.575693153468068e-06, "loss": 0.4889, "step": 2231 }, { "epoch": 0.15824175824175823, "grad_norm": 2.806021183543985, "learning_rate": 9.575230164829254e-06, "loss": 0.4551, "step": 2232 }, { "epoch": 0.15831265508684864, "grad_norm": 2.7541268905948497, "learning_rate": 9.574766934935059e-06, "loss": 0.3955, "step": 2233 }, { "epoch": 0.15838355193193904, "grad_norm": 5.281795594366705, "learning_rate": 9.574303463809913e-06, "loss": 0.4924, "step": 2234 }, { "epoch": 0.1584544487770294, "grad_norm": 3.663746045464175, "learning_rate": 9.573839751478252e-06, "loss": 0.4712, "step": 2235 }, { "epoch": 0.1585253456221198, "grad_norm": 5.806999207971458, "learning_rate": 9.57337579796453e-06, "loss": 0.4341, "step": 2236 }, { "epoch": 0.1585962424672102, "grad_norm": 2.4466808439735943, "learning_rate": 9.572911603293214e-06, "loss": 0.4671, "step": 2237 }, { "epoch": 0.15866713931230061, "grad_norm": 5.446793641314267, "learning_rate": 9.572447167488778e-06, "loss": 0.4464, "step": 2238 }, { "epoch": 0.158738036157391, "grad_norm": 3.138876882764112, "learning_rate": 9.571982490575711e-06, "loss": 0.4907, "step": 2239 }, { "epoch": 0.1588089330024814, "grad_norm": 1.1935822194544514, "learning_rate": 9.571517572578519e-06, "loss": 0.665, "step": 2240 }, { "epoch": 0.1588798298475718, "grad_norm": 2.7648125035435833, "learning_rate": 9.571052413521716e-06, "loss": 0.4284, "step": 2241 }, { "epoch": 0.15895072669266216, "grad_norm": 1.185418270356841, "learning_rate": 9.57058701342983e-06, "loss": 0.7351, "step": 2242 }, { "epoch": 0.15902162353775257, "grad_norm": 5.717474127973453, "learning_rate": 9.570121372327404e-06, "loss": 0.4182, "step": 2243 }, { "epoch": 0.15909252038284297, "grad_norm": 2.3303354915203482, "learning_rate": 9.569655490238986e-06, "loss": 0.4249, "step": 2244 }, { "epoch": 0.15916341722793337, "grad_norm": 6.992557373981384, "learning_rate": 9.56918936718915e-06, "loss": 0.4959, "step": 2245 }, { "epoch": 0.15923431407302374, "grad_norm": 2.694881754634685, "learning_rate": 9.568723003202469e-06, "loss": 0.3984, "step": 2246 }, { "epoch": 0.15930521091811414, "grad_norm": 1.161291462797548, "learning_rate": 9.568256398303538e-06, "loss": 0.677, "step": 2247 }, { "epoch": 0.15937610776320454, "grad_norm": 2.1306169451496704, "learning_rate": 9.567789552516959e-06, "loss": 0.4303, "step": 2248 }, { "epoch": 0.15944700460829492, "grad_norm": 3.0001741684866987, "learning_rate": 9.567322465867353e-06, "loss": 0.4165, "step": 2249 }, { "epoch": 0.15951790145338532, "grad_norm": 3.2527968061461094, "learning_rate": 9.566855138379346e-06, "loss": 0.4438, "step": 2250 }, { "epoch": 0.15958879829847572, "grad_norm": 3.1143233740636846, "learning_rate": 9.56638757007758e-06, "loss": 0.4526, "step": 2251 }, { "epoch": 0.15965969514356612, "grad_norm": 2.56101493929828, "learning_rate": 9.565919760986712e-06, "loss": 0.4422, "step": 2252 }, { "epoch": 0.1597305919886565, "grad_norm": 2.8058849278814013, "learning_rate": 9.565451711131413e-06, "loss": 0.4087, "step": 2253 }, { "epoch": 0.1598014888337469, "grad_norm": 3.340237747124801, "learning_rate": 9.564983420536359e-06, "loss": 0.4494, "step": 2254 }, { "epoch": 0.1598723856788373, "grad_norm": 3.546251211735958, "learning_rate": 9.564514889226245e-06, "loss": 0.4507, "step": 2255 }, { "epoch": 0.1599432825239277, "grad_norm": 4.423258286536172, "learning_rate": 9.564046117225776e-06, "loss": 0.4522, "step": 2256 }, { "epoch": 0.16001417936901807, "grad_norm": 4.191872044295764, "learning_rate": 9.563577104559673e-06, "loss": 0.4268, "step": 2257 }, { "epoch": 0.16008507621410847, "grad_norm": 3.499309829203131, "learning_rate": 9.563107851252666e-06, "loss": 0.4532, "step": 2258 }, { "epoch": 0.16015597305919888, "grad_norm": 9.964077230624907, "learning_rate": 9.562638357329498e-06, "loss": 0.4095, "step": 2259 }, { "epoch": 0.16022686990428925, "grad_norm": 3.6453633140309707, "learning_rate": 9.562168622814929e-06, "loss": 0.4747, "step": 2260 }, { "epoch": 0.16029776674937965, "grad_norm": 9.474506329441146, "learning_rate": 9.561698647733726e-06, "loss": 0.438, "step": 2261 }, { "epoch": 0.16036866359447005, "grad_norm": 4.497671141197873, "learning_rate": 9.561228432110669e-06, "loss": 0.4826, "step": 2262 }, { "epoch": 0.16043956043956045, "grad_norm": 3.4687139682988066, "learning_rate": 9.560757975970558e-06, "loss": 0.4956, "step": 2263 }, { "epoch": 0.16051045728465083, "grad_norm": 3.205108540448499, "learning_rate": 9.560287279338196e-06, "loss": 0.4487, "step": 2264 }, { "epoch": 0.16058135412974123, "grad_norm": 4.2347782391702955, "learning_rate": 9.559816342238407e-06, "loss": 0.4185, "step": 2265 }, { "epoch": 0.16065225097483163, "grad_norm": 2.3150580122647932, "learning_rate": 9.55934516469602e-06, "loss": 0.4692, "step": 2266 }, { "epoch": 0.160723147819922, "grad_norm": 1.9823514598299194, "learning_rate": 9.558873746735883e-06, "loss": 0.4368, "step": 2267 }, { "epoch": 0.1607940446650124, "grad_norm": 2.33464483002381, "learning_rate": 9.558402088382854e-06, "loss": 0.409, "step": 2268 }, { "epoch": 0.1608649415101028, "grad_norm": 3.0123569992807173, "learning_rate": 9.557930189661804e-06, "loss": 0.4366, "step": 2269 }, { "epoch": 0.1609358383551932, "grad_norm": 2.247866209299271, "learning_rate": 9.557458050597616e-06, "loss": 0.3981, "step": 2270 }, { "epoch": 0.16100673520028358, "grad_norm": 3.7180020800402516, "learning_rate": 9.556985671215187e-06, "loss": 0.438, "step": 2271 }, { "epoch": 0.16107763204537398, "grad_norm": 5.200399192586739, "learning_rate": 9.556513051539425e-06, "loss": 0.4757, "step": 2272 }, { "epoch": 0.16114852889046438, "grad_norm": 3.6222752176051376, "learning_rate": 9.55604019159525e-06, "loss": 0.4081, "step": 2273 }, { "epoch": 0.16121942573555476, "grad_norm": 2.5047038602075316, "learning_rate": 9.5555670914076e-06, "loss": 0.4743, "step": 2274 }, { "epoch": 0.16129032258064516, "grad_norm": 2.275151477916641, "learning_rate": 9.55509375100142e-06, "loss": 0.4773, "step": 2275 }, { "epoch": 0.16136121942573556, "grad_norm": 2.147181406442963, "learning_rate": 9.55462017040167e-06, "loss": 0.4153, "step": 2276 }, { "epoch": 0.16143211627082596, "grad_norm": 3.200298800758293, "learning_rate": 9.554146349633321e-06, "loss": 0.4211, "step": 2277 }, { "epoch": 0.16150301311591633, "grad_norm": 2.8869207165111206, "learning_rate": 9.553672288721361e-06, "loss": 0.4211, "step": 2278 }, { "epoch": 0.16157390996100673, "grad_norm": 4.438888908512783, "learning_rate": 9.553197987690785e-06, "loss": 0.5032, "step": 2279 }, { "epoch": 0.16164480680609714, "grad_norm": 2.805673154655641, "learning_rate": 9.552723446566604e-06, "loss": 0.4817, "step": 2280 }, { "epoch": 0.1617157036511875, "grad_norm": 5.169657886162878, "learning_rate": 9.552248665373838e-06, "loss": 0.4781, "step": 2281 }, { "epoch": 0.1617866004962779, "grad_norm": 4.727477449982133, "learning_rate": 9.551773644137528e-06, "loss": 0.3996, "step": 2282 }, { "epoch": 0.1618574973413683, "grad_norm": 2.744136330279479, "learning_rate": 9.551298382882718e-06, "loss": 0.4836, "step": 2283 }, { "epoch": 0.1619283941864587, "grad_norm": 2.6555934762286717, "learning_rate": 9.550822881634472e-06, "loss": 0.4457, "step": 2284 }, { "epoch": 0.1619992910315491, "grad_norm": 2.532878299949124, "learning_rate": 9.55034714041786e-06, "loss": 0.4519, "step": 2285 }, { "epoch": 0.1620701878766395, "grad_norm": 3.921108816279632, "learning_rate": 9.549871159257971e-06, "loss": 0.5332, "step": 2286 }, { "epoch": 0.1621410847217299, "grad_norm": 3.9166356552004093, "learning_rate": 9.549394938179904e-06, "loss": 0.5086, "step": 2287 }, { "epoch": 0.1622119815668203, "grad_norm": 2.783867283800149, "learning_rate": 9.54891847720877e-06, "loss": 0.4407, "step": 2288 }, { "epoch": 0.16228287841191066, "grad_norm": 4.6124353900992965, "learning_rate": 9.54844177636969e-06, "loss": 0.5106, "step": 2289 }, { "epoch": 0.16235377525700107, "grad_norm": 5.158105938630095, "learning_rate": 9.547964835687804e-06, "loss": 0.4584, "step": 2290 }, { "epoch": 0.16242467210209147, "grad_norm": 6.325846673709071, "learning_rate": 9.547487655188263e-06, "loss": 0.4318, "step": 2291 }, { "epoch": 0.16249556894718184, "grad_norm": 2.3164553366262988, "learning_rate": 9.547010234896225e-06, "loss": 0.4281, "step": 2292 }, { "epoch": 0.16256646579227224, "grad_norm": 2.863640763014485, "learning_rate": 9.546532574836868e-06, "loss": 0.457, "step": 2293 }, { "epoch": 0.16263736263736264, "grad_norm": 1.9434151105492696, "learning_rate": 9.546054675035378e-06, "loss": 0.4917, "step": 2294 }, { "epoch": 0.16270825948245304, "grad_norm": 3.1303160288578527, "learning_rate": 9.545576535516953e-06, "loss": 0.4202, "step": 2295 }, { "epoch": 0.16277915632754342, "grad_norm": 5.024246549710822, "learning_rate": 9.54509815630681e-06, "loss": 0.4425, "step": 2296 }, { "epoch": 0.16285005317263382, "grad_norm": 5.037316594319691, "learning_rate": 9.544619537430173e-06, "loss": 0.4419, "step": 2297 }, { "epoch": 0.16292095001772422, "grad_norm": 12.231249642779616, "learning_rate": 9.544140678912278e-06, "loss": 0.4267, "step": 2298 }, { "epoch": 0.1629918468628146, "grad_norm": 3.1614184950185127, "learning_rate": 9.543661580778376e-06, "loss": 0.4135, "step": 2299 }, { "epoch": 0.163062743707905, "grad_norm": 2.1907141211154317, "learning_rate": 9.54318224305373e-06, "loss": 0.3976, "step": 2300 }, { "epoch": 0.1631336405529954, "grad_norm": 3.635609131415213, "learning_rate": 9.542702665763619e-06, "loss": 0.513, "step": 2301 }, { "epoch": 0.1632045373980858, "grad_norm": 2.6417666388169705, "learning_rate": 9.542222848933327e-06, "loss": 0.4369, "step": 2302 }, { "epoch": 0.16327543424317617, "grad_norm": 2.9441234029769805, "learning_rate": 9.541742792588158e-06, "loss": 0.44, "step": 2303 }, { "epoch": 0.16334633108826657, "grad_norm": 2.3093638967073034, "learning_rate": 9.541262496753426e-06, "loss": 0.4442, "step": 2304 }, { "epoch": 0.16341722793335697, "grad_norm": 2.7828583035767287, "learning_rate": 9.540781961454456e-06, "loss": 0.4789, "step": 2305 }, { "epoch": 0.16348812477844735, "grad_norm": 3.6727123400876844, "learning_rate": 9.540301186716587e-06, "loss": 0.4607, "step": 2306 }, { "epoch": 0.16355902162353775, "grad_norm": 1.9162886034035673, "learning_rate": 9.539820172565171e-06, "loss": 0.4543, "step": 2307 }, { "epoch": 0.16362991846862815, "grad_norm": 4.155501102786716, "learning_rate": 9.539338919025572e-06, "loss": 0.4586, "step": 2308 }, { "epoch": 0.16370081531371855, "grad_norm": 2.949141502008368, "learning_rate": 9.538857426123167e-06, "loss": 0.4691, "step": 2309 }, { "epoch": 0.16377171215880892, "grad_norm": 2.695848649479946, "learning_rate": 9.538375693883346e-06, "loss": 0.4329, "step": 2310 }, { "epoch": 0.16384260900389933, "grad_norm": 13.015062194061871, "learning_rate": 9.53789372233151e-06, "loss": 0.4878, "step": 2311 }, { "epoch": 0.16391350584898973, "grad_norm": 2.5524938945173417, "learning_rate": 9.537411511493076e-06, "loss": 0.5168, "step": 2312 }, { "epoch": 0.1639844026940801, "grad_norm": 2.315612911777675, "learning_rate": 9.536929061393468e-06, "loss": 0.4358, "step": 2313 }, { "epoch": 0.1640552995391705, "grad_norm": 2.9288921705314657, "learning_rate": 9.536446372058128e-06, "loss": 0.3982, "step": 2314 }, { "epoch": 0.1641261963842609, "grad_norm": 4.917812077189256, "learning_rate": 9.535963443512508e-06, "loss": 0.4786, "step": 2315 }, { "epoch": 0.1641970932293513, "grad_norm": 1.9336293463486955, "learning_rate": 9.535480275782075e-06, "loss": 0.4599, "step": 2316 }, { "epoch": 0.16426799007444168, "grad_norm": 3.552176713685478, "learning_rate": 9.534996868892303e-06, "loss": 0.3988, "step": 2317 }, { "epoch": 0.16433888691953208, "grad_norm": 10.342813067465546, "learning_rate": 9.534513222868687e-06, "loss": 0.3853, "step": 2318 }, { "epoch": 0.16440978376462248, "grad_norm": 3.6351681764872628, "learning_rate": 9.534029337736727e-06, "loss": 0.4548, "step": 2319 }, { "epoch": 0.16448068060971288, "grad_norm": 1.7695482486877088, "learning_rate": 9.53354521352194e-06, "loss": 0.4383, "step": 2320 }, { "epoch": 0.16455157745480326, "grad_norm": 3.0550374678511845, "learning_rate": 9.53306085024985e-06, "loss": 0.4902, "step": 2321 }, { "epoch": 0.16462247429989366, "grad_norm": 2.7006024610567545, "learning_rate": 9.532576247946005e-06, "loss": 0.4371, "step": 2322 }, { "epoch": 0.16469337114498406, "grad_norm": 4.332079269504323, "learning_rate": 9.532091406635954e-06, "loss": 0.4729, "step": 2323 }, { "epoch": 0.16476426799007443, "grad_norm": 3.739367402946699, "learning_rate": 9.531606326345266e-06, "loss": 0.4582, "step": 2324 }, { "epoch": 0.16483516483516483, "grad_norm": 2.8701599543055334, "learning_rate": 9.531121007099515e-06, "loss": 0.4316, "step": 2325 }, { "epoch": 0.16490606168025523, "grad_norm": 2.276189324636286, "learning_rate": 9.530635448924295e-06, "loss": 0.4961, "step": 2326 }, { "epoch": 0.16497695852534563, "grad_norm": 3.109648973505426, "learning_rate": 9.53014965184521e-06, "loss": 0.3992, "step": 2327 }, { "epoch": 0.165047855370436, "grad_norm": 3.0721862166973617, "learning_rate": 9.529663615887876e-06, "loss": 0.4091, "step": 2328 }, { "epoch": 0.1651187522155264, "grad_norm": 3.3513390710167608, "learning_rate": 9.529177341077924e-06, "loss": 0.4808, "step": 2329 }, { "epoch": 0.1651896490606168, "grad_norm": 2.7704245843210975, "learning_rate": 9.528690827440992e-06, "loss": 0.4328, "step": 2330 }, { "epoch": 0.16526054590570718, "grad_norm": 2.469720400119436, "learning_rate": 9.528204075002735e-06, "loss": 0.4263, "step": 2331 }, { "epoch": 0.16533144275079759, "grad_norm": 1.9824489425661378, "learning_rate": 9.527717083788823e-06, "loss": 0.425, "step": 2332 }, { "epoch": 0.165402339595888, "grad_norm": 2.7538233416049, "learning_rate": 9.527229853824934e-06, "loss": 0.4204, "step": 2333 }, { "epoch": 0.1654732364409784, "grad_norm": 5.866011176143765, "learning_rate": 9.526742385136757e-06, "loss": 0.4682, "step": 2334 }, { "epoch": 0.16554413328606876, "grad_norm": 2.2096053064577807, "learning_rate": 9.526254677749999e-06, "loss": 0.3942, "step": 2335 }, { "epoch": 0.16561503013115916, "grad_norm": 4.449365303033194, "learning_rate": 9.525766731690378e-06, "loss": 0.4421, "step": 2336 }, { "epoch": 0.16568592697624956, "grad_norm": 5.24310437528426, "learning_rate": 9.525278546983623e-06, "loss": 0.4534, "step": 2337 }, { "epoch": 0.16575682382133994, "grad_norm": 2.556446236901184, "learning_rate": 9.524790123655477e-06, "loss": 0.4339, "step": 2338 }, { "epoch": 0.16582772066643034, "grad_norm": 1.4382598854612674, "learning_rate": 9.524301461731694e-06, "loss": 0.7053, "step": 2339 }, { "epoch": 0.16589861751152074, "grad_norm": 2.759465456422821, "learning_rate": 9.523812561238039e-06, "loss": 0.443, "step": 2340 }, { "epoch": 0.16596951435661114, "grad_norm": 3.3465258901152515, "learning_rate": 9.523323422200297e-06, "loss": 0.4865, "step": 2341 }, { "epoch": 0.16604041120170152, "grad_norm": 2.4401949017529523, "learning_rate": 9.522834044644258e-06, "loss": 0.4462, "step": 2342 }, { "epoch": 0.16611130804679192, "grad_norm": 2.319606173851121, "learning_rate": 9.52234442859573e-06, "loss": 0.431, "step": 2343 }, { "epoch": 0.16618220489188232, "grad_norm": 10.232920659562282, "learning_rate": 9.521854574080524e-06, "loss": 0.4409, "step": 2344 }, { "epoch": 0.1662531017369727, "grad_norm": 3.894638614338247, "learning_rate": 9.521364481124477e-06, "loss": 0.4459, "step": 2345 }, { "epoch": 0.1663239985820631, "grad_norm": 4.170740404040535, "learning_rate": 9.520874149753432e-06, "loss": 0.4785, "step": 2346 }, { "epoch": 0.1663948954271535, "grad_norm": 2.053146489933428, "learning_rate": 9.52038357999324e-06, "loss": 0.4491, "step": 2347 }, { "epoch": 0.1664657922722439, "grad_norm": 2.94302753661241, "learning_rate": 9.519892771869773e-06, "loss": 0.4739, "step": 2348 }, { "epoch": 0.16653668911733427, "grad_norm": 2.2246704522241947, "learning_rate": 9.51940172540891e-06, "loss": 0.3855, "step": 2349 }, { "epoch": 0.16660758596242467, "grad_norm": 2.053288347633817, "learning_rate": 9.518910440636545e-06, "loss": 0.4361, "step": 2350 }, { "epoch": 0.16667848280751507, "grad_norm": 1.2488086416989075, "learning_rate": 9.51841891757858e-06, "loss": 0.6824, "step": 2351 }, { "epoch": 0.16674937965260547, "grad_norm": 3.416206999993811, "learning_rate": 9.51792715626094e-06, "loss": 0.4807, "step": 2352 }, { "epoch": 0.16682027649769585, "grad_norm": 5.479679621697342, "learning_rate": 9.51743515670955e-06, "loss": 0.4702, "step": 2353 }, { "epoch": 0.16689117334278625, "grad_norm": 2.2762352447527943, "learning_rate": 9.516942918950361e-06, "loss": 0.4026, "step": 2354 }, { "epoch": 0.16696207018787665, "grad_norm": 2.839067898595002, "learning_rate": 9.51645044300932e-06, "loss": 0.4283, "step": 2355 }, { "epoch": 0.16703296703296702, "grad_norm": 4.619047954178747, "learning_rate": 9.515957728912401e-06, "loss": 0.4269, "step": 2356 }, { "epoch": 0.16710386387805742, "grad_norm": 2.6547701909425783, "learning_rate": 9.515464776685584e-06, "loss": 0.525, "step": 2357 }, { "epoch": 0.16717476072314782, "grad_norm": 7.543881937935813, "learning_rate": 9.514971586354863e-06, "loss": 0.4093, "step": 2358 }, { "epoch": 0.16724565756823823, "grad_norm": 1.0853280418202842, "learning_rate": 9.514478157946244e-06, "loss": 0.6541, "step": 2359 }, { "epoch": 0.1673165544133286, "grad_norm": 3.200780262321249, "learning_rate": 9.513984491485747e-06, "loss": 0.432, "step": 2360 }, { "epoch": 0.167387451258419, "grad_norm": 1.9330385896773548, "learning_rate": 9.513490586999401e-06, "loss": 0.4337, "step": 2361 }, { "epoch": 0.1674583481035094, "grad_norm": 3.8210389523637893, "learning_rate": 9.512996444513252e-06, "loss": 0.4663, "step": 2362 }, { "epoch": 0.16752924494859978, "grad_norm": 7.80291707477214, "learning_rate": 9.512502064053356e-06, "loss": 0.457, "step": 2363 }, { "epoch": 0.16760014179369018, "grad_norm": 2.389618192277616, "learning_rate": 9.51200744564578e-06, "loss": 0.386, "step": 2364 }, { "epoch": 0.16767103863878058, "grad_norm": 2.151860169246065, "learning_rate": 9.511512589316612e-06, "loss": 0.4735, "step": 2365 }, { "epoch": 0.16774193548387098, "grad_norm": 1.8221405611503012, "learning_rate": 9.511017495091937e-06, "loss": 0.4353, "step": 2366 }, { "epoch": 0.16781283232896135, "grad_norm": 3.4259932237002904, "learning_rate": 9.510522162997869e-06, "loss": 0.4684, "step": 2367 }, { "epoch": 0.16788372917405175, "grad_norm": 2.9227956979791183, "learning_rate": 9.510026593060525e-06, "loss": 0.4591, "step": 2368 }, { "epoch": 0.16795462601914216, "grad_norm": 2.001260521990557, "learning_rate": 9.509530785306034e-06, "loss": 0.3965, "step": 2369 }, { "epoch": 0.16802552286423253, "grad_norm": 2.3969871061967605, "learning_rate": 9.509034739760546e-06, "loss": 0.4296, "step": 2370 }, { "epoch": 0.16809641970932293, "grad_norm": 3.0356273176885127, "learning_rate": 9.508538456450212e-06, "loss": 0.4269, "step": 2371 }, { "epoch": 0.16816731655441333, "grad_norm": 6.3440198703907065, "learning_rate": 9.508041935401204e-06, "loss": 0.4035, "step": 2372 }, { "epoch": 0.16823821339950373, "grad_norm": 8.302265732072348, "learning_rate": 9.507545176639703e-06, "loss": 0.4837, "step": 2373 }, { "epoch": 0.1683091102445941, "grad_norm": 4.517968883572933, "learning_rate": 9.507048180191906e-06, "loss": 0.4807, "step": 2374 }, { "epoch": 0.1683800070896845, "grad_norm": 2.055982356967345, "learning_rate": 9.50655094608402e-06, "loss": 0.4446, "step": 2375 }, { "epoch": 0.1684509039347749, "grad_norm": 2.276533817558042, "learning_rate": 9.50605347434226e-06, "loss": 0.4356, "step": 2376 }, { "epoch": 0.16852180077986528, "grad_norm": 3.0064464551747623, "learning_rate": 9.505555764992863e-06, "loss": 0.415, "step": 2377 }, { "epoch": 0.16859269762495568, "grad_norm": 1.8178608679052648, "learning_rate": 9.505057818062071e-06, "loss": 0.4209, "step": 2378 }, { "epoch": 0.16866359447004609, "grad_norm": 3.1300669798074794, "learning_rate": 9.504559633576142e-06, "loss": 0.4964, "step": 2379 }, { "epoch": 0.1687344913151365, "grad_norm": 3.4616657451520063, "learning_rate": 9.504061211561343e-06, "loss": 0.4245, "step": 2380 }, { "epoch": 0.16880538816022686, "grad_norm": 9.03258254374909, "learning_rate": 9.503562552043961e-06, "loss": 0.5093, "step": 2381 }, { "epoch": 0.16887628500531726, "grad_norm": 3.440651520419813, "learning_rate": 9.503063655050288e-06, "loss": 0.4764, "step": 2382 }, { "epoch": 0.16894718185040766, "grad_norm": 2.7494818545791695, "learning_rate": 9.50256452060663e-06, "loss": 0.4535, "step": 2383 }, { "epoch": 0.16901807869549806, "grad_norm": 3.1221774369365267, "learning_rate": 9.50206514873931e-06, "loss": 0.4759, "step": 2384 }, { "epoch": 0.16908897554058844, "grad_norm": 2.6546669490546564, "learning_rate": 9.501565539474657e-06, "loss": 0.4667, "step": 2385 }, { "epoch": 0.16915987238567884, "grad_norm": 2.182789499945156, "learning_rate": 9.50106569283902e-06, "loss": 0.439, "step": 2386 }, { "epoch": 0.16923076923076924, "grad_norm": 2.0729429168780698, "learning_rate": 9.500565608858749e-06, "loss": 0.4702, "step": 2387 }, { "epoch": 0.1693016660758596, "grad_norm": 2.221506759708443, "learning_rate": 9.50006528756022e-06, "loss": 0.3917, "step": 2388 }, { "epoch": 0.16937256292095001, "grad_norm": 3.479471051853101, "learning_rate": 9.499564728969815e-06, "loss": 0.442, "step": 2389 }, { "epoch": 0.16944345976604042, "grad_norm": 22.801445996565267, "learning_rate": 9.499063933113928e-06, "loss": 0.4763, "step": 2390 }, { "epoch": 0.16951435661113082, "grad_norm": 2.6429737611349644, "learning_rate": 9.498562900018963e-06, "loss": 0.4117, "step": 2391 }, { "epoch": 0.1695852534562212, "grad_norm": 4.874083204772344, "learning_rate": 9.498061629711345e-06, "loss": 0.4322, "step": 2392 }, { "epoch": 0.1696561503013116, "grad_norm": 3.2618592918008487, "learning_rate": 9.497560122217504e-06, "loss": 0.4241, "step": 2393 }, { "epoch": 0.169727047146402, "grad_norm": 3.885344535562466, "learning_rate": 9.497058377563884e-06, "loss": 0.3926, "step": 2394 }, { "epoch": 0.16979794399149237, "grad_norm": 2.3035913638094834, "learning_rate": 9.496556395776944e-06, "loss": 0.4119, "step": 2395 }, { "epoch": 0.16986884083658277, "grad_norm": 3.958598309799832, "learning_rate": 9.496054176883152e-06, "loss": 0.4236, "step": 2396 }, { "epoch": 0.16993973768167317, "grad_norm": 6.954620179216175, "learning_rate": 9.495551720908993e-06, "loss": 0.4568, "step": 2397 }, { "epoch": 0.17001063452676357, "grad_norm": 4.437348553539924, "learning_rate": 9.495049027880959e-06, "loss": 0.4159, "step": 2398 }, { "epoch": 0.17008153137185394, "grad_norm": 3.097212178634219, "learning_rate": 9.494546097825559e-06, "loss": 0.4386, "step": 2399 }, { "epoch": 0.17015242821694435, "grad_norm": 2.9259534093466844, "learning_rate": 9.494042930769314e-06, "loss": 0.4594, "step": 2400 }, { "epoch": 0.17022332506203475, "grad_norm": 2.0739141090959965, "learning_rate": 9.493539526738756e-06, "loss": 0.4471, "step": 2401 }, { "epoch": 0.17029422190712512, "grad_norm": 5.755977175732491, "learning_rate": 9.493035885760426e-06, "loss": 0.4526, "step": 2402 }, { "epoch": 0.17036511875221552, "grad_norm": 5.43635024079191, "learning_rate": 9.492532007860886e-06, "loss": 0.4058, "step": 2403 }, { "epoch": 0.17043601559730592, "grad_norm": 5.40260943999807, "learning_rate": 9.492027893066705e-06, "loss": 0.4324, "step": 2404 }, { "epoch": 0.17050691244239632, "grad_norm": 5.311875339066739, "learning_rate": 9.491523541404464e-06, "loss": 0.4453, "step": 2405 }, { "epoch": 0.1705778092874867, "grad_norm": 4.120605058411537, "learning_rate": 9.491018952900756e-06, "loss": 0.4825, "step": 2406 }, { "epoch": 0.1706487061325771, "grad_norm": 1.9788058986347579, "learning_rate": 9.490514127582193e-06, "loss": 0.4744, "step": 2407 }, { "epoch": 0.1707196029776675, "grad_norm": 6.964466981901991, "learning_rate": 9.490009065475395e-06, "loss": 0.4432, "step": 2408 }, { "epoch": 0.17079049982275787, "grad_norm": 3.6816051567697934, "learning_rate": 9.489503766606987e-06, "loss": 0.4922, "step": 2409 }, { "epoch": 0.17086139666784828, "grad_norm": 2.1437605622259572, "learning_rate": 9.488998231003623e-06, "loss": 0.4514, "step": 2410 }, { "epoch": 0.17093229351293868, "grad_norm": 3.02318315083626, "learning_rate": 9.488492458691953e-06, "loss": 0.4767, "step": 2411 }, { "epoch": 0.17100319035802908, "grad_norm": 2.9844049254308382, "learning_rate": 9.487986449698651e-06, "loss": 0.4439, "step": 2412 }, { "epoch": 0.17107408720311945, "grad_norm": 3.1215119494500243, "learning_rate": 9.487480204050398e-06, "loss": 0.4668, "step": 2413 }, { "epoch": 0.17114498404820985, "grad_norm": 2.027119524468664, "learning_rate": 9.486973721773888e-06, "loss": 0.3918, "step": 2414 }, { "epoch": 0.17121588089330025, "grad_norm": 6.5818617959257, "learning_rate": 9.486467002895829e-06, "loss": 0.4245, "step": 2415 }, { "epoch": 0.17128677773839066, "grad_norm": 5.7157039066333954, "learning_rate": 9.485960047442941e-06, "loss": 0.4377, "step": 2416 }, { "epoch": 0.17135767458348103, "grad_norm": 3.9621381392425703, "learning_rate": 9.485452855441956e-06, "loss": 0.4319, "step": 2417 }, { "epoch": 0.17142857142857143, "grad_norm": 4.810245134531917, "learning_rate": 9.484945426919618e-06, "loss": 0.4674, "step": 2418 }, { "epoch": 0.17149946827366183, "grad_norm": 5.368487733039106, "learning_rate": 9.484437761902685e-06, "loss": 0.4354, "step": 2419 }, { "epoch": 0.1715703651187522, "grad_norm": 10.401549646280595, "learning_rate": 9.483929860417925e-06, "loss": 0.4406, "step": 2420 }, { "epoch": 0.1716412619638426, "grad_norm": 3.6220938692796314, "learning_rate": 9.483421722492121e-06, "loss": 0.4823, "step": 2421 }, { "epoch": 0.171712158808933, "grad_norm": 4.697647880605123, "learning_rate": 9.482913348152069e-06, "loss": 0.4384, "step": 2422 }, { "epoch": 0.1717830556540234, "grad_norm": 3.125719581219329, "learning_rate": 9.482404737424573e-06, "loss": 0.5193, "step": 2423 }, { "epoch": 0.17185395249911378, "grad_norm": 2.9369008541404655, "learning_rate": 9.481895890336454e-06, "loss": 0.4339, "step": 2424 }, { "epoch": 0.17192484934420418, "grad_norm": 3.8915006433482815, "learning_rate": 9.481386806914545e-06, "loss": 0.4875, "step": 2425 }, { "epoch": 0.17199574618929458, "grad_norm": 2.2258475784131013, "learning_rate": 9.480877487185689e-06, "loss": 0.4382, "step": 2426 }, { "epoch": 0.17206664303438496, "grad_norm": 2.5006226635959727, "learning_rate": 9.48036793117674e-06, "loss": 0.4264, "step": 2427 }, { "epoch": 0.17213753987947536, "grad_norm": 3.833292341521467, "learning_rate": 9.479858138914575e-06, "loss": 0.4694, "step": 2428 }, { "epoch": 0.17220843672456576, "grad_norm": 3.0461562687599755, "learning_rate": 9.479348110426068e-06, "loss": 0.4322, "step": 2429 }, { "epoch": 0.17227933356965616, "grad_norm": 4.60281758812033, "learning_rate": 9.478837845738116e-06, "loss": 0.4819, "step": 2430 }, { "epoch": 0.17235023041474654, "grad_norm": 3.9453789301983657, "learning_rate": 9.478327344877627e-06, "loss": 0.4267, "step": 2431 }, { "epoch": 0.17242112725983694, "grad_norm": 7.682976663163015, "learning_rate": 9.477816607871519e-06, "loss": 0.4661, "step": 2432 }, { "epoch": 0.17249202410492734, "grad_norm": 6.797316620761945, "learning_rate": 9.477305634746723e-06, "loss": 0.467, "step": 2433 }, { "epoch": 0.1725629209500177, "grad_norm": 4.723181937028499, "learning_rate": 9.476794425530185e-06, "loss": 0.4483, "step": 2434 }, { "epoch": 0.1726338177951081, "grad_norm": 4.243114754994931, "learning_rate": 9.476282980248858e-06, "loss": 0.4241, "step": 2435 }, { "epoch": 0.17270471464019851, "grad_norm": 3.4531591848358127, "learning_rate": 9.475771298929715e-06, "loss": 0.4276, "step": 2436 }, { "epoch": 0.17277561148528892, "grad_norm": 3.47307565577118, "learning_rate": 9.475259381599734e-06, "loss": 0.4818, "step": 2437 }, { "epoch": 0.1728465083303793, "grad_norm": 2.9408608049064657, "learning_rate": 9.474747228285909e-06, "loss": 0.429, "step": 2438 }, { "epoch": 0.1729174051754697, "grad_norm": 3.4903392127062984, "learning_rate": 9.474234839015248e-06, "loss": 0.4747, "step": 2439 }, { "epoch": 0.1729883020205601, "grad_norm": 1.195022257126424, "learning_rate": 9.473722213814769e-06, "loss": 0.6883, "step": 2440 }, { "epoch": 0.17305919886565047, "grad_norm": 3.0281950895239893, "learning_rate": 9.473209352711503e-06, "loss": 0.4776, "step": 2441 }, { "epoch": 0.17313009571074087, "grad_norm": 2.9489278615122037, "learning_rate": 9.472696255732493e-06, "loss": 0.4069, "step": 2442 }, { "epoch": 0.17320099255583127, "grad_norm": 5.557915439822615, "learning_rate": 9.472182922904795e-06, "loss": 0.4807, "step": 2443 }, { "epoch": 0.17327188940092167, "grad_norm": 3.5774909565691067, "learning_rate": 9.47166935425548e-06, "loss": 0.369, "step": 2444 }, { "epoch": 0.17334278624601204, "grad_norm": 3.6155118496094825, "learning_rate": 9.471155549811626e-06, "loss": 0.3871, "step": 2445 }, { "epoch": 0.17341368309110244, "grad_norm": 2.5395887912408934, "learning_rate": 9.470641509600328e-06, "loss": 0.426, "step": 2446 }, { "epoch": 0.17348457993619285, "grad_norm": 4.819018055731312, "learning_rate": 9.470127233648689e-06, "loss": 0.4622, "step": 2447 }, { "epoch": 0.17355547678128325, "grad_norm": 2.9668915982467063, "learning_rate": 9.469612721983831e-06, "loss": 0.4228, "step": 2448 }, { "epoch": 0.17362637362637362, "grad_norm": 2.993011373839314, "learning_rate": 9.469097974632881e-06, "loss": 0.4045, "step": 2449 }, { "epoch": 0.17369727047146402, "grad_norm": 3.5094241957621164, "learning_rate": 9.468582991622986e-06, "loss": 0.4387, "step": 2450 }, { "epoch": 0.17376816731655442, "grad_norm": 2.545111795731737, "learning_rate": 9.468067772981298e-06, "loss": 0.4369, "step": 2451 }, { "epoch": 0.1738390641616448, "grad_norm": 2.111726910386781, "learning_rate": 9.467552318734986e-06, "loss": 0.4559, "step": 2452 }, { "epoch": 0.1739099610067352, "grad_norm": 2.759278976704051, "learning_rate": 9.467036628911228e-06, "loss": 0.4453, "step": 2453 }, { "epoch": 0.1739808578518256, "grad_norm": 12.927998505719112, "learning_rate": 9.466520703537222e-06, "loss": 0.4506, "step": 2454 }, { "epoch": 0.174051754696916, "grad_norm": 3.0717635987840874, "learning_rate": 9.46600454264017e-06, "loss": 0.4048, "step": 2455 }, { "epoch": 0.17412265154200637, "grad_norm": 2.358583643612874, "learning_rate": 9.465488146247287e-06, "loss": 0.4649, "step": 2456 }, { "epoch": 0.17419354838709677, "grad_norm": 5.115306833995957, "learning_rate": 9.464971514385808e-06, "loss": 0.4554, "step": 2457 }, { "epoch": 0.17426444523218718, "grad_norm": 2.2311200334972754, "learning_rate": 9.464454647082973e-06, "loss": 0.3898, "step": 2458 }, { "epoch": 0.17433534207727755, "grad_norm": 4.58788132355535, "learning_rate": 9.463937544366038e-06, "loss": 0.4714, "step": 2459 }, { "epoch": 0.17440623892236795, "grad_norm": 4.339373878374237, "learning_rate": 9.463420206262267e-06, "loss": 0.4258, "step": 2460 }, { "epoch": 0.17447713576745835, "grad_norm": 4.7500115788671975, "learning_rate": 9.462902632798943e-06, "loss": 0.463, "step": 2461 }, { "epoch": 0.17454803261254875, "grad_norm": 3.103843561863248, "learning_rate": 9.462384824003355e-06, "loss": 0.4517, "step": 2462 }, { "epoch": 0.17461892945763913, "grad_norm": 3.3997163799681673, "learning_rate": 9.46186677990281e-06, "loss": 0.4296, "step": 2463 }, { "epoch": 0.17468982630272953, "grad_norm": 3.4759041816835845, "learning_rate": 9.461348500524625e-06, "loss": 0.4815, "step": 2464 }, { "epoch": 0.17476072314781993, "grad_norm": 2.825121829760864, "learning_rate": 9.460829985896129e-06, "loss": 0.447, "step": 2465 }, { "epoch": 0.1748316199929103, "grad_norm": 5.874104735032461, "learning_rate": 9.460311236044662e-06, "loss": 0.4281, "step": 2466 }, { "epoch": 0.1749025168380007, "grad_norm": 3.4524601883732697, "learning_rate": 9.459792250997579e-06, "loss": 0.4617, "step": 2467 }, { "epoch": 0.1749734136830911, "grad_norm": 2.420799673525759, "learning_rate": 9.459273030782245e-06, "loss": 0.4058, "step": 2468 }, { "epoch": 0.1750443105281815, "grad_norm": 3.526887087996418, "learning_rate": 9.458753575426042e-06, "loss": 0.5198, "step": 2469 }, { "epoch": 0.17511520737327188, "grad_norm": 3.2331859946547667, "learning_rate": 9.458233884956361e-06, "loss": 0.4559, "step": 2470 }, { "epoch": 0.17518610421836228, "grad_norm": 4.522816321723111, "learning_rate": 9.457713959400604e-06, "loss": 0.4347, "step": 2471 }, { "epoch": 0.17525700106345268, "grad_norm": 4.567826538245225, "learning_rate": 9.457193798786187e-06, "loss": 0.4204, "step": 2472 }, { "epoch": 0.17532789790854306, "grad_norm": 7.976832523456668, "learning_rate": 9.456673403140539e-06, "loss": 0.4688, "step": 2473 }, { "epoch": 0.17539879475363346, "grad_norm": 3.078222782806556, "learning_rate": 9.4561527724911e-06, "loss": 0.4157, "step": 2474 }, { "epoch": 0.17546969159872386, "grad_norm": 5.728968206747602, "learning_rate": 9.455631906865326e-06, "loss": 0.4442, "step": 2475 }, { "epoch": 0.17554058844381426, "grad_norm": 3.2397359507223644, "learning_rate": 9.455110806290679e-06, "loss": 0.4327, "step": 2476 }, { "epoch": 0.17561148528890463, "grad_norm": 3.0118831255597454, "learning_rate": 9.45458947079464e-06, "loss": 0.4651, "step": 2477 }, { "epoch": 0.17568238213399504, "grad_norm": 3.530213729769548, "learning_rate": 9.454067900404696e-06, "loss": 0.4173, "step": 2478 }, { "epoch": 0.17575327897908544, "grad_norm": 4.712930879074872, "learning_rate": 9.453546095148354e-06, "loss": 0.4336, "step": 2479 }, { "epoch": 0.17582417582417584, "grad_norm": 3.3172599558285554, "learning_rate": 9.453024055053126e-06, "loss": 0.4535, "step": 2480 }, { "epoch": 0.1758950726692662, "grad_norm": 3.6874813050071187, "learning_rate": 9.45250178014654e-06, "loss": 0.4479, "step": 2481 }, { "epoch": 0.1759659695143566, "grad_norm": 2.779253737849476, "learning_rate": 9.451979270456139e-06, "loss": 0.4585, "step": 2482 }, { "epoch": 0.17603686635944701, "grad_norm": 2.6517855338918572, "learning_rate": 9.451456526009473e-06, "loss": 0.4298, "step": 2483 }, { "epoch": 0.1761077632045374, "grad_norm": 2.7896015128916303, "learning_rate": 9.450933546834106e-06, "loss": 0.4813, "step": 2484 }, { "epoch": 0.1761786600496278, "grad_norm": 3.176249655409078, "learning_rate": 9.450410332957616e-06, "loss": 0.3994, "step": 2485 }, { "epoch": 0.1762495568947182, "grad_norm": 3.677367466399795, "learning_rate": 9.44988688440759e-06, "loss": 0.4677, "step": 2486 }, { "epoch": 0.1763204537398086, "grad_norm": 2.774309538525833, "learning_rate": 9.449363201211635e-06, "loss": 0.433, "step": 2487 }, { "epoch": 0.17639135058489896, "grad_norm": 1.1068396263882399, "learning_rate": 9.448839283397361e-06, "loss": 0.6727, "step": 2488 }, { "epoch": 0.17646224742998937, "grad_norm": 2.692164777718927, "learning_rate": 9.448315130992397e-06, "loss": 0.4469, "step": 2489 }, { "epoch": 0.17653314427507977, "grad_norm": 4.6166032316047, "learning_rate": 9.447790744024382e-06, "loss": 0.4299, "step": 2490 }, { "epoch": 0.17660404112017014, "grad_norm": 4.017170217460148, "learning_rate": 9.447266122520963e-06, "loss": 0.4761, "step": 2491 }, { "epoch": 0.17667493796526054, "grad_norm": 3.312847062203368, "learning_rate": 9.446741266509809e-06, "loss": 0.4275, "step": 2492 }, { "epoch": 0.17674583481035094, "grad_norm": 0.9976501652342846, "learning_rate": 9.446216176018593e-06, "loss": 0.6655, "step": 2493 }, { "epoch": 0.17681673165544134, "grad_norm": 4.1414144741546055, "learning_rate": 9.445690851075005e-06, "loss": 0.4091, "step": 2494 }, { "epoch": 0.17688762850053172, "grad_norm": 3.3816531999671593, "learning_rate": 9.445165291706744e-06, "loss": 0.4459, "step": 2495 }, { "epoch": 0.17695852534562212, "grad_norm": 4.333636755710868, "learning_rate": 9.444639497941526e-06, "loss": 0.4099, "step": 2496 }, { "epoch": 0.17702942219071252, "grad_norm": 4.491471488958757, "learning_rate": 9.444113469807075e-06, "loss": 0.4515, "step": 2497 }, { "epoch": 0.1771003190358029, "grad_norm": 5.929824365061193, "learning_rate": 9.443587207331126e-06, "loss": 0.439, "step": 2498 }, { "epoch": 0.1771712158808933, "grad_norm": 6.72269332638971, "learning_rate": 9.443060710541435e-06, "loss": 0.4661, "step": 2499 }, { "epoch": 0.1772421127259837, "grad_norm": 2.5868040393608376, "learning_rate": 9.44253397946576e-06, "loss": 0.4049, "step": 2500 }, { "epoch": 0.1773130095710741, "grad_norm": 3.947121708514348, "learning_rate": 9.442007014131877e-06, "loss": 0.4068, "step": 2501 }, { "epoch": 0.17738390641616447, "grad_norm": 3.2378406055548914, "learning_rate": 9.441479814567574e-06, "loss": 0.4843, "step": 2502 }, { "epoch": 0.17745480326125487, "grad_norm": 1.6750349240419036, "learning_rate": 9.440952380800651e-06, "loss": 0.4652, "step": 2503 }, { "epoch": 0.17752570010634527, "grad_norm": 10.7057845632491, "learning_rate": 9.44042471285892e-06, "loss": 0.4146, "step": 2504 }, { "epoch": 0.17759659695143565, "grad_norm": 3.645011899455414, "learning_rate": 9.439896810770205e-06, "loss": 0.4088, "step": 2505 }, { "epoch": 0.17766749379652605, "grad_norm": 4.277511716089291, "learning_rate": 9.439368674562339e-06, "loss": 0.3805, "step": 2506 }, { "epoch": 0.17773839064161645, "grad_norm": 3.2405948566558225, "learning_rate": 9.438840304263177e-06, "loss": 0.4368, "step": 2507 }, { "epoch": 0.17780928748670685, "grad_norm": 5.828790316980707, "learning_rate": 9.438311699900577e-06, "loss": 0.4519, "step": 2508 }, { "epoch": 0.17788018433179723, "grad_norm": 4.055320057587629, "learning_rate": 9.437782861502413e-06, "loss": 0.4349, "step": 2509 }, { "epoch": 0.17795108117688763, "grad_norm": 7.411159186175219, "learning_rate": 9.437253789096571e-06, "loss": 0.478, "step": 2510 }, { "epoch": 0.17802197802197803, "grad_norm": 6.105541994085798, "learning_rate": 9.436724482710952e-06, "loss": 0.4649, "step": 2511 }, { "epoch": 0.17809287486706843, "grad_norm": 2.383993665470991, "learning_rate": 9.436194942373464e-06, "loss": 0.4553, "step": 2512 }, { "epoch": 0.1781637717121588, "grad_norm": 3.5511660978305954, "learning_rate": 9.435665168112031e-06, "loss": 0.4258, "step": 2513 }, { "epoch": 0.1782346685572492, "grad_norm": 4.619775516076541, "learning_rate": 9.435135159954587e-06, "loss": 0.4283, "step": 2514 }, { "epoch": 0.1783055654023396, "grad_norm": 4.887925786050473, "learning_rate": 9.434604917929082e-06, "loss": 0.4457, "step": 2515 }, { "epoch": 0.17837646224742998, "grad_norm": 12.19773227484192, "learning_rate": 9.434074442063475e-06, "loss": 0.4283, "step": 2516 }, { "epoch": 0.17844735909252038, "grad_norm": 3.610135296165568, "learning_rate": 9.433543732385736e-06, "loss": 0.4522, "step": 2517 }, { "epoch": 0.17851825593761078, "grad_norm": 3.1073762289471554, "learning_rate": 9.433012788923855e-06, "loss": 0.413, "step": 2518 }, { "epoch": 0.17858915278270118, "grad_norm": 2.8115739100227164, "learning_rate": 9.432481611705823e-06, "loss": 0.4543, "step": 2519 }, { "epoch": 0.17866004962779156, "grad_norm": 3.5857202336479945, "learning_rate": 9.431950200759655e-06, "loss": 0.414, "step": 2520 }, { "epoch": 0.17873094647288196, "grad_norm": 4.240099522991545, "learning_rate": 9.43141855611337e-06, "loss": 0.4608, "step": 2521 }, { "epoch": 0.17880184331797236, "grad_norm": 1.9440533212116933, "learning_rate": 9.430886677794998e-06, "loss": 0.49, "step": 2522 }, { "epoch": 0.17887274016306273, "grad_norm": 4.186107020962193, "learning_rate": 9.430354565832594e-06, "loss": 0.4576, "step": 2523 }, { "epoch": 0.17894363700815313, "grad_norm": 3.3498874454163605, "learning_rate": 9.429822220254212e-06, "loss": 0.4545, "step": 2524 }, { "epoch": 0.17901453385324353, "grad_norm": 5.1806605111827295, "learning_rate": 9.429289641087923e-06, "loss": 0.4705, "step": 2525 }, { "epoch": 0.17908543069833394, "grad_norm": 3.4635542620312316, "learning_rate": 9.428756828361808e-06, "loss": 0.45, "step": 2526 }, { "epoch": 0.1791563275434243, "grad_norm": 3.977759466032798, "learning_rate": 9.428223782103967e-06, "loss": 0.4506, "step": 2527 }, { "epoch": 0.1792272243885147, "grad_norm": 4.687373774171302, "learning_rate": 9.427690502342507e-06, "loss": 0.4639, "step": 2528 }, { "epoch": 0.1792981212336051, "grad_norm": 3.5117642009294268, "learning_rate": 9.427156989105546e-06, "loss": 0.3981, "step": 2529 }, { "epoch": 0.17936901807869549, "grad_norm": 4.3343051524781675, "learning_rate": 9.42662324242122e-06, "loss": 0.3894, "step": 2530 }, { "epoch": 0.1794399149237859, "grad_norm": 2.595252650673224, "learning_rate": 9.426089262317668e-06, "loss": 0.4221, "step": 2531 }, { "epoch": 0.1795108117688763, "grad_norm": 2.6901346824522574, "learning_rate": 9.425555048823054e-06, "loss": 0.4173, "step": 2532 }, { "epoch": 0.1795817086139667, "grad_norm": 2.09580512082657, "learning_rate": 9.425020601965541e-06, "loss": 0.4214, "step": 2533 }, { "epoch": 0.17965260545905706, "grad_norm": 3.4430055667351973, "learning_rate": 9.424485921773318e-06, "loss": 0.4903, "step": 2534 }, { "epoch": 0.17972350230414746, "grad_norm": 7.308206388849076, "learning_rate": 9.423951008274574e-06, "loss": 0.4337, "step": 2535 }, { "epoch": 0.17979439914923787, "grad_norm": 2.340952297156953, "learning_rate": 9.423415861497515e-06, "loss": 0.5676, "step": 2536 }, { "epoch": 0.17986529599432824, "grad_norm": 2.5325564155406552, "learning_rate": 9.422880481470364e-06, "loss": 0.3805, "step": 2537 }, { "epoch": 0.17993619283941864, "grad_norm": 6.055631668446473, "learning_rate": 9.42234486822135e-06, "loss": 0.5065, "step": 2538 }, { "epoch": 0.18000708968450904, "grad_norm": 5.473566693806737, "learning_rate": 9.421809021778711e-06, "loss": 0.429, "step": 2539 }, { "epoch": 0.18007798652959944, "grad_norm": 2.96461087759663, "learning_rate": 9.421272942170711e-06, "loss": 0.422, "step": 2540 }, { "epoch": 0.18014888337468982, "grad_norm": 1.9873157021479193, "learning_rate": 9.420736629425614e-06, "loss": 0.4336, "step": 2541 }, { "epoch": 0.18021978021978022, "grad_norm": 4.1715975651694475, "learning_rate": 9.4202000835717e-06, "loss": 0.4315, "step": 2542 }, { "epoch": 0.18029067706487062, "grad_norm": 4.477151550278575, "learning_rate": 9.41966330463726e-06, "loss": 0.4328, "step": 2543 }, { "epoch": 0.18036157390996102, "grad_norm": 4.535861458492337, "learning_rate": 9.419126292650602e-06, "loss": 0.382, "step": 2544 }, { "epoch": 0.1804324707550514, "grad_norm": 3.2943091277136776, "learning_rate": 9.418589047640043e-06, "loss": 0.4492, "step": 2545 }, { "epoch": 0.1805033676001418, "grad_norm": 8.548733400937751, "learning_rate": 9.418051569633909e-06, "loss": 0.4415, "step": 2546 }, { "epoch": 0.1805742644452322, "grad_norm": 1.2441572056263932, "learning_rate": 9.417513858660544e-06, "loss": 0.7414, "step": 2547 }, { "epoch": 0.18064516129032257, "grad_norm": 3.094025608877498, "learning_rate": 9.416975914748301e-06, "loss": 0.4526, "step": 2548 }, { "epoch": 0.18071605813541297, "grad_norm": 11.731209707708809, "learning_rate": 9.416437737925547e-06, "loss": 0.4315, "step": 2549 }, { "epoch": 0.18078695498050337, "grad_norm": 3.530484436534711, "learning_rate": 9.41589932822066e-06, "loss": 0.4383, "step": 2550 }, { "epoch": 0.18085785182559377, "grad_norm": 3.5068663323040337, "learning_rate": 9.415360685662034e-06, "loss": 0.4602, "step": 2551 }, { "epoch": 0.18092874867068415, "grad_norm": 5.664132149341884, "learning_rate": 9.414821810278066e-06, "loss": 0.3996, "step": 2552 }, { "epoch": 0.18099964551577455, "grad_norm": 3.9174939605735744, "learning_rate": 9.414282702097175e-06, "loss": 0.4433, "step": 2553 }, { "epoch": 0.18107054236086495, "grad_norm": 4.26775421767574, "learning_rate": 9.413743361147786e-06, "loss": 0.4541, "step": 2554 }, { "epoch": 0.18114143920595532, "grad_norm": 3.463367503585087, "learning_rate": 9.413203787458343e-06, "loss": 0.4667, "step": 2555 }, { "epoch": 0.18121233605104572, "grad_norm": 6.999265013990399, "learning_rate": 9.412663981057294e-06, "loss": 0.5002, "step": 2556 }, { "epoch": 0.18128323289613613, "grad_norm": 4.371520615181363, "learning_rate": 9.412123941973108e-06, "loss": 0.4961, "step": 2557 }, { "epoch": 0.18135412974122653, "grad_norm": 7.296389472940014, "learning_rate": 9.411583670234257e-06, "loss": 0.4045, "step": 2558 }, { "epoch": 0.1814250265863169, "grad_norm": 3.734691486292126, "learning_rate": 9.411043165869232e-06, "loss": 0.4692, "step": 2559 }, { "epoch": 0.1814959234314073, "grad_norm": 3.728755711874114, "learning_rate": 9.410502428906533e-06, "loss": 0.4448, "step": 2560 }, { "epoch": 0.1815668202764977, "grad_norm": 4.804567814152279, "learning_rate": 9.409961459374676e-06, "loss": 0.4274, "step": 2561 }, { "epoch": 0.18163771712158808, "grad_norm": 5.14991712956847, "learning_rate": 9.409420257302185e-06, "loss": 0.4874, "step": 2562 }, { "epoch": 0.18170861396667848, "grad_norm": 2.4613717556734653, "learning_rate": 9.408878822717598e-06, "loss": 0.4278, "step": 2563 }, { "epoch": 0.18177951081176888, "grad_norm": 23.17946014301219, "learning_rate": 9.408337155649467e-06, "loss": 0.3998, "step": 2564 }, { "epoch": 0.18185040765685928, "grad_norm": 3.60433954934998, "learning_rate": 9.40779525612635e-06, "loss": 0.4106, "step": 2565 }, { "epoch": 0.18192130450194965, "grad_norm": 2.2532592633030832, "learning_rate": 9.407253124176828e-06, "loss": 0.4475, "step": 2566 }, { "epoch": 0.18199220134704006, "grad_norm": 4.873024980721455, "learning_rate": 9.406710759829484e-06, "loss": 0.4391, "step": 2567 }, { "epoch": 0.18206309819213046, "grad_norm": 6.783681900387299, "learning_rate": 9.406168163112917e-06, "loss": 0.4793, "step": 2568 }, { "epoch": 0.18213399503722083, "grad_norm": 2.611480918935124, "learning_rate": 9.40562533405574e-06, "loss": 0.4197, "step": 2569 }, { "epoch": 0.18220489188231123, "grad_norm": 4.140949050871621, "learning_rate": 9.405082272686577e-06, "loss": 0.4589, "step": 2570 }, { "epoch": 0.18227578872740163, "grad_norm": 2.3628167368087434, "learning_rate": 9.404538979034064e-06, "loss": 0.4518, "step": 2571 }, { "epoch": 0.18234668557249203, "grad_norm": 3.013169581982576, "learning_rate": 9.403995453126848e-06, "loss": 0.4329, "step": 2572 }, { "epoch": 0.1824175824175824, "grad_norm": 2.684715480858234, "learning_rate": 9.403451694993591e-06, "loss": 0.4633, "step": 2573 }, { "epoch": 0.1824884792626728, "grad_norm": 4.176760459587765, "learning_rate": 9.402907704662965e-06, "loss": 0.4252, "step": 2574 }, { "epoch": 0.1825593761077632, "grad_norm": 2.4711459138938863, "learning_rate": 9.402363482163655e-06, "loss": 0.4821, "step": 2575 }, { "epoch": 0.1826302729528536, "grad_norm": 6.458428062365907, "learning_rate": 9.401819027524357e-06, "loss": 0.4604, "step": 2576 }, { "epoch": 0.18270116979794399, "grad_norm": 5.541182710597423, "learning_rate": 9.401274340773782e-06, "loss": 0.4718, "step": 2577 }, { "epoch": 0.1827720666430344, "grad_norm": 7.522279097202247, "learning_rate": 9.400729421940652e-06, "loss": 0.4541, "step": 2578 }, { "epoch": 0.1828429634881248, "grad_norm": 2.5591114958415595, "learning_rate": 9.400184271053702e-06, "loss": 0.4388, "step": 2579 }, { "epoch": 0.18291386033321516, "grad_norm": 4.87315948413085, "learning_rate": 9.399638888141678e-06, "loss": 0.423, "step": 2580 }, { "epoch": 0.18298475717830556, "grad_norm": 4.973951664778142, "learning_rate": 9.399093273233336e-06, "loss": 0.5086, "step": 2581 }, { "epoch": 0.18305565402339596, "grad_norm": 2.3014617007804685, "learning_rate": 9.398547426357448e-06, "loss": 0.4436, "step": 2582 }, { "epoch": 0.18312655086848637, "grad_norm": 2.6537093068642257, "learning_rate": 9.398001347542795e-06, "loss": 0.3996, "step": 2583 }, { "epoch": 0.18319744771357674, "grad_norm": 2.511148259144445, "learning_rate": 9.397455036818176e-06, "loss": 0.4028, "step": 2584 }, { "epoch": 0.18326834455866714, "grad_norm": 2.99467109023316, "learning_rate": 9.396908494212396e-06, "loss": 0.4193, "step": 2585 }, { "epoch": 0.18333924140375754, "grad_norm": 2.80063765304485, "learning_rate": 9.396361719754275e-06, "loss": 0.4123, "step": 2586 }, { "epoch": 0.18341013824884791, "grad_norm": 2.868910837827465, "learning_rate": 9.395814713472645e-06, "loss": 0.4325, "step": 2587 }, { "epoch": 0.18348103509393832, "grad_norm": 3.2039011626339478, "learning_rate": 9.39526747539635e-06, "loss": 0.4255, "step": 2588 }, { "epoch": 0.18355193193902872, "grad_norm": 5.9286639203937375, "learning_rate": 9.394720005554247e-06, "loss": 0.5222, "step": 2589 }, { "epoch": 0.18362282878411912, "grad_norm": 2.3131453060497784, "learning_rate": 9.394172303975204e-06, "loss": 0.4429, "step": 2590 }, { "epoch": 0.1836937256292095, "grad_norm": 3.795065703639395, "learning_rate": 9.393624370688101e-06, "loss": 0.4633, "step": 2591 }, { "epoch": 0.1837646224742999, "grad_norm": 2.6233055892093495, "learning_rate": 9.39307620572183e-06, "loss": 0.4098, "step": 2592 }, { "epoch": 0.1838355193193903, "grad_norm": 2.085476255398813, "learning_rate": 9.3925278091053e-06, "loss": 0.4384, "step": 2593 }, { "epoch": 0.18390641616448067, "grad_norm": 1.8916053719006092, "learning_rate": 9.391979180867423e-06, "loss": 0.4044, "step": 2594 }, { "epoch": 0.18397731300957107, "grad_norm": 3.128805445453407, "learning_rate": 9.391430321037134e-06, "loss": 0.4164, "step": 2595 }, { "epoch": 0.18404820985466147, "grad_norm": 3.0799220103035685, "learning_rate": 9.390881229643371e-06, "loss": 0.4232, "step": 2596 }, { "epoch": 0.18411910669975187, "grad_norm": 8.829101158731346, "learning_rate": 9.39033190671509e-06, "loss": 0.441, "step": 2597 }, { "epoch": 0.18419000354484225, "grad_norm": 15.864543316634371, "learning_rate": 9.389782352281257e-06, "loss": 0.4651, "step": 2598 }, { "epoch": 0.18426090038993265, "grad_norm": 3.2598842223500477, "learning_rate": 9.38923256637085e-06, "loss": 0.4193, "step": 2599 }, { "epoch": 0.18433179723502305, "grad_norm": 4.914955862870672, "learning_rate": 9.38868254901286e-06, "loss": 0.4664, "step": 2600 }, { "epoch": 0.18440269408011342, "grad_norm": 2.604271383353569, "learning_rate": 9.388132300236288e-06, "loss": 0.4778, "step": 2601 }, { "epoch": 0.18447359092520382, "grad_norm": 4.485389575057888, "learning_rate": 9.387581820070152e-06, "loss": 0.47, "step": 2602 }, { "epoch": 0.18454448777029422, "grad_norm": 1.914010699675432, "learning_rate": 9.387031108543476e-06, "loss": 0.4362, "step": 2603 }, { "epoch": 0.18461538461538463, "grad_norm": 3.901558317194455, "learning_rate": 9.3864801656853e-06, "loss": 0.4657, "step": 2604 }, { "epoch": 0.184686281460475, "grad_norm": 2.8156454606123646, "learning_rate": 9.38592899152468e-06, "loss": 0.4324, "step": 2605 }, { "epoch": 0.1847571783055654, "grad_norm": 3.115242342732142, "learning_rate": 9.385377586090675e-06, "loss": 0.47, "step": 2606 }, { "epoch": 0.1848280751506558, "grad_norm": 1.085955445852248, "learning_rate": 9.384825949412363e-06, "loss": 0.7633, "step": 2607 }, { "epoch": 0.1848989719957462, "grad_norm": 6.217860385468807, "learning_rate": 9.384274081518833e-06, "loss": 0.454, "step": 2608 }, { "epoch": 0.18496986884083658, "grad_norm": 4.540050121626392, "learning_rate": 9.383721982439184e-06, "loss": 0.4206, "step": 2609 }, { "epoch": 0.18504076568592698, "grad_norm": 2.9140418563137356, "learning_rate": 9.383169652202529e-06, "loss": 0.4552, "step": 2610 }, { "epoch": 0.18511166253101738, "grad_norm": 1.9620944236020943, "learning_rate": 9.38261709083799e-06, "loss": 0.4407, "step": 2611 }, { "epoch": 0.18518255937610775, "grad_norm": 3.7472866619040732, "learning_rate": 9.382064298374709e-06, "loss": 0.3952, "step": 2612 }, { "epoch": 0.18525345622119815, "grad_norm": 3.330461533121021, "learning_rate": 9.381511274841834e-06, "loss": 0.42, "step": 2613 }, { "epoch": 0.18532435306628856, "grad_norm": 6.508994849912691, "learning_rate": 9.380958020268523e-06, "loss": 0.5211, "step": 2614 }, { "epoch": 0.18539524991137896, "grad_norm": 31.998764220789628, "learning_rate": 9.380404534683953e-06, "loss": 0.4546, "step": 2615 }, { "epoch": 0.18546614675646933, "grad_norm": 3.105313504510621, "learning_rate": 9.379850818117307e-06, "loss": 0.4178, "step": 2616 }, { "epoch": 0.18553704360155973, "grad_norm": 2.8567237095427593, "learning_rate": 9.379296870597786e-06, "loss": 0.3849, "step": 2617 }, { "epoch": 0.18560794044665013, "grad_norm": 2.7716823005897098, "learning_rate": 9.3787426921546e-06, "loss": 0.4367, "step": 2618 }, { "epoch": 0.1856788372917405, "grad_norm": 4.4339225263591535, "learning_rate": 9.378188282816966e-06, "loss": 0.4725, "step": 2619 }, { "epoch": 0.1857497341368309, "grad_norm": 4.970789500596229, "learning_rate": 9.377633642614124e-06, "loss": 0.4401, "step": 2620 }, { "epoch": 0.1858206309819213, "grad_norm": 4.1953951832213505, "learning_rate": 9.377078771575317e-06, "loss": 0.4885, "step": 2621 }, { "epoch": 0.1858915278270117, "grad_norm": 2.248946863730056, "learning_rate": 9.376523669729807e-06, "loss": 0.4777, "step": 2622 }, { "epoch": 0.18596242467210208, "grad_norm": 3.7775333574718455, "learning_rate": 9.375968337106864e-06, "loss": 0.4894, "step": 2623 }, { "epoch": 0.18603332151719248, "grad_norm": 2.781229160723132, "learning_rate": 9.37541277373577e-06, "loss": 0.4431, "step": 2624 }, { "epoch": 0.18610421836228289, "grad_norm": 2.7670061806127677, "learning_rate": 9.374856979645822e-06, "loss": 0.4943, "step": 2625 }, { "epoch": 0.18617511520737326, "grad_norm": 3.758818092705497, "learning_rate": 9.374300954866324e-06, "loss": 0.367, "step": 2626 }, { "epoch": 0.18624601205246366, "grad_norm": 7.0301608724937, "learning_rate": 9.373744699426599e-06, "loss": 0.4112, "step": 2627 }, { "epoch": 0.18631690889755406, "grad_norm": 4.225108754998925, "learning_rate": 9.373188213355977e-06, "loss": 0.4067, "step": 2628 }, { "epoch": 0.18638780574264446, "grad_norm": 3.9134911151200176, "learning_rate": 9.372631496683804e-06, "loss": 0.4268, "step": 2629 }, { "epoch": 0.18645870258773484, "grad_norm": 6.119864938753787, "learning_rate": 9.372074549439434e-06, "loss": 0.473, "step": 2630 }, { "epoch": 0.18652959943282524, "grad_norm": 6.7820191740966616, "learning_rate": 9.371517371652236e-06, "loss": 0.4733, "step": 2631 }, { "epoch": 0.18660049627791564, "grad_norm": 8.554961949635533, "learning_rate": 9.37095996335159e-06, "loss": 0.4291, "step": 2632 }, { "epoch": 0.186671393123006, "grad_norm": 2.9967264589029328, "learning_rate": 9.370402324566889e-06, "loss": 0.4662, "step": 2633 }, { "epoch": 0.18674228996809641, "grad_norm": 3.2953109838742747, "learning_rate": 9.369844455327536e-06, "loss": 0.4961, "step": 2634 }, { "epoch": 0.18681318681318682, "grad_norm": 5.386107054248095, "learning_rate": 9.369286355662953e-06, "loss": 0.4843, "step": 2635 }, { "epoch": 0.18688408365827722, "grad_norm": 7.641747184310772, "learning_rate": 9.368728025602564e-06, "loss": 0.4967, "step": 2636 }, { "epoch": 0.1869549805033676, "grad_norm": 2.982234544291244, "learning_rate": 9.368169465175812e-06, "loss": 0.4611, "step": 2637 }, { "epoch": 0.187025877348458, "grad_norm": 3.135804775689875, "learning_rate": 9.367610674412148e-06, "loss": 0.4274, "step": 2638 }, { "epoch": 0.1870967741935484, "grad_norm": 3.6710059978332006, "learning_rate": 9.367051653341042e-06, "loss": 0.4334, "step": 2639 }, { "epoch": 0.1871676710386388, "grad_norm": 2.8118123795585723, "learning_rate": 9.36649240199197e-06, "loss": 0.4837, "step": 2640 }, { "epoch": 0.18723856788372917, "grad_norm": 6.356046668460806, "learning_rate": 9.365932920394419e-06, "loss": 0.4713, "step": 2641 }, { "epoch": 0.18730946472881957, "grad_norm": 2.970580549602475, "learning_rate": 9.365373208577892e-06, "loss": 0.4562, "step": 2642 }, { "epoch": 0.18738036157390997, "grad_norm": 2.891332381715655, "learning_rate": 9.364813266571903e-06, "loss": 0.4828, "step": 2643 }, { "epoch": 0.18745125841900034, "grad_norm": 6.86000466210727, "learning_rate": 9.364253094405982e-06, "loss": 0.5059, "step": 2644 }, { "epoch": 0.18752215526409075, "grad_norm": 3.457627737670115, "learning_rate": 9.363692692109662e-06, "loss": 0.4521, "step": 2645 }, { "epoch": 0.18759305210918115, "grad_norm": 3.935674968301835, "learning_rate": 9.363132059712497e-06, "loss": 0.4204, "step": 2646 }, { "epoch": 0.18766394895427155, "grad_norm": 3.9707469416509404, "learning_rate": 9.362571197244048e-06, "loss": 0.4298, "step": 2647 }, { "epoch": 0.18773484579936192, "grad_norm": 0.9388478310009095, "learning_rate": 9.36201010473389e-06, "loss": 0.6797, "step": 2648 }, { "epoch": 0.18780574264445232, "grad_norm": 2.5276968640220927, "learning_rate": 9.361448782211609e-06, "loss": 0.4636, "step": 2649 }, { "epoch": 0.18787663948954272, "grad_norm": 1.724331124553739, "learning_rate": 9.360887229706806e-06, "loss": 0.4571, "step": 2650 }, { "epoch": 0.1879475363346331, "grad_norm": 3.5625422145426024, "learning_rate": 9.360325447249091e-06, "loss": 0.4518, "step": 2651 }, { "epoch": 0.1880184331797235, "grad_norm": 2.8595507234985047, "learning_rate": 9.359763434868086e-06, "loss": 0.458, "step": 2652 }, { "epoch": 0.1880893300248139, "grad_norm": 2.5789942346666463, "learning_rate": 9.359201192593426e-06, "loss": 0.4439, "step": 2653 }, { "epoch": 0.1881602268699043, "grad_norm": 3.4192591573978803, "learning_rate": 9.358638720454762e-06, "loss": 0.4544, "step": 2654 }, { "epoch": 0.18823112371499467, "grad_norm": 2.1815748107931703, "learning_rate": 9.358076018481751e-06, "loss": 0.4598, "step": 2655 }, { "epoch": 0.18830202056008508, "grad_norm": 3.3397915643806586, "learning_rate": 9.357513086704065e-06, "loss": 0.4692, "step": 2656 }, { "epoch": 0.18837291740517548, "grad_norm": 3.0153362870449474, "learning_rate": 9.356949925151388e-06, "loss": 0.4359, "step": 2657 }, { "epoch": 0.18844381425026585, "grad_norm": 3.8193938967570493, "learning_rate": 9.356386533853417e-06, "loss": 0.4526, "step": 2658 }, { "epoch": 0.18851471109535625, "grad_norm": 2.203670518913618, "learning_rate": 9.355822912839858e-06, "loss": 0.4396, "step": 2659 }, { "epoch": 0.18858560794044665, "grad_norm": 5.476395399308384, "learning_rate": 9.355259062140432e-06, "loss": 0.4785, "step": 2660 }, { "epoch": 0.18865650478553705, "grad_norm": 2.4765445261444046, "learning_rate": 9.354694981784872e-06, "loss": 0.4555, "step": 2661 }, { "epoch": 0.18872740163062743, "grad_norm": 2.419666467523469, "learning_rate": 9.354130671802921e-06, "loss": 0.4206, "step": 2662 }, { "epoch": 0.18879829847571783, "grad_norm": 1.2092270726761465, "learning_rate": 9.353566132224337e-06, "loss": 0.6447, "step": 2663 }, { "epoch": 0.18886919532080823, "grad_norm": 4.167206109881285, "learning_rate": 9.353001363078887e-06, "loss": 0.5122, "step": 2664 }, { "epoch": 0.1889400921658986, "grad_norm": 7.13174995419706, "learning_rate": 9.352436364396354e-06, "loss": 0.4632, "step": 2665 }, { "epoch": 0.189010989010989, "grad_norm": 5.852678055890818, "learning_rate": 9.35187113620653e-06, "loss": 0.4246, "step": 2666 }, { "epoch": 0.1890818858560794, "grad_norm": 3.1019649761123977, "learning_rate": 9.351305678539217e-06, "loss": 0.4518, "step": 2667 }, { "epoch": 0.1891527827011698, "grad_norm": 3.823646924351898, "learning_rate": 9.350739991424236e-06, "loss": 0.3992, "step": 2668 }, { "epoch": 0.18922367954626018, "grad_norm": 3.0058153445216904, "learning_rate": 9.350174074891414e-06, "loss": 0.4198, "step": 2669 }, { "epoch": 0.18929457639135058, "grad_norm": 3.5082395525112595, "learning_rate": 9.349607928970592e-06, "loss": 0.4859, "step": 2670 }, { "epoch": 0.18936547323644098, "grad_norm": 4.284799684880484, "learning_rate": 9.349041553691624e-06, "loss": 0.4437, "step": 2671 }, { "epoch": 0.18943637008153139, "grad_norm": 4.022549084516074, "learning_rate": 9.348474949084377e-06, "loss": 0.4199, "step": 2672 }, { "epoch": 0.18950726692662176, "grad_norm": 4.820956803477306, "learning_rate": 9.347908115178726e-06, "loss": 0.4648, "step": 2673 }, { "epoch": 0.18957816377171216, "grad_norm": 1.295681587159546, "learning_rate": 9.347341052004562e-06, "loss": 0.7257, "step": 2674 }, { "epoch": 0.18964906061680256, "grad_norm": 6.2714032954071754, "learning_rate": 9.346773759591786e-06, "loss": 0.4196, "step": 2675 }, { "epoch": 0.18971995746189294, "grad_norm": 2.7388240734087486, "learning_rate": 9.346206237970312e-06, "loss": 0.4746, "step": 2676 }, { "epoch": 0.18979085430698334, "grad_norm": 1.9313612900000319, "learning_rate": 9.345638487170066e-06, "loss": 0.3973, "step": 2677 }, { "epoch": 0.18986175115207374, "grad_norm": 3.0506636453640392, "learning_rate": 9.345070507220984e-06, "loss": 0.4267, "step": 2678 }, { "epoch": 0.18993264799716414, "grad_norm": 4.0891671053196905, "learning_rate": 9.344502298153018e-06, "loss": 0.4143, "step": 2679 }, { "epoch": 0.1900035448422545, "grad_norm": 3.0133914231588816, "learning_rate": 9.34393385999613e-06, "loss": 0.4529, "step": 2680 }, { "epoch": 0.1900744416873449, "grad_norm": 3.3401231973783427, "learning_rate": 9.343365192780296e-06, "loss": 0.4224, "step": 2681 }, { "epoch": 0.19014533853243532, "grad_norm": 5.558511091572148, "learning_rate": 9.342796296535498e-06, "loss": 0.3914, "step": 2682 }, { "epoch": 0.1902162353775257, "grad_norm": 5.4695182620474085, "learning_rate": 9.342227171291738e-06, "loss": 0.4705, "step": 2683 }, { "epoch": 0.1902871322226161, "grad_norm": 3.672769696074095, "learning_rate": 9.341657817079024e-06, "loss": 0.4254, "step": 2684 }, { "epoch": 0.1903580290677065, "grad_norm": 2.2795138481308235, "learning_rate": 9.34108823392738e-06, "loss": 0.4207, "step": 2685 }, { "epoch": 0.1904289259127969, "grad_norm": 3.88355425742157, "learning_rate": 9.34051842186684e-06, "loss": 0.4154, "step": 2686 }, { "epoch": 0.19049982275788727, "grad_norm": 1.1908238156266817, "learning_rate": 9.33994838092745e-06, "loss": 0.7366, "step": 2687 }, { "epoch": 0.19057071960297767, "grad_norm": 3.13145018359935, "learning_rate": 9.33937811113927e-06, "loss": 0.4539, "step": 2688 }, { "epoch": 0.19064161644806807, "grad_norm": 3.2785882600582714, "learning_rate": 9.33880761253237e-06, "loss": 0.4211, "step": 2689 }, { "epoch": 0.19071251329315844, "grad_norm": 3.374846242590474, "learning_rate": 9.338236885136833e-06, "loss": 0.4384, "step": 2690 }, { "epoch": 0.19078341013824884, "grad_norm": 4.900997319063155, "learning_rate": 9.337665928982754e-06, "loss": 0.471, "step": 2691 }, { "epoch": 0.19085430698333924, "grad_norm": 7.505362619843957, "learning_rate": 9.337094744100239e-06, "loss": 0.4655, "step": 2692 }, { "epoch": 0.19092520382842965, "grad_norm": 2.970333230886754, "learning_rate": 9.336523330519409e-06, "loss": 0.4074, "step": 2693 }, { "epoch": 0.19099610067352002, "grad_norm": 3.9077478048181327, "learning_rate": 9.335951688270393e-06, "loss": 0.4836, "step": 2694 }, { "epoch": 0.19106699751861042, "grad_norm": 2.4558040853901937, "learning_rate": 9.335379817383334e-06, "loss": 0.4425, "step": 2695 }, { "epoch": 0.19113789436370082, "grad_norm": 2.4626997845751006, "learning_rate": 9.33480771788839e-06, "loss": 0.4413, "step": 2696 }, { "epoch": 0.1912087912087912, "grad_norm": 2.0689902071795525, "learning_rate": 9.334235389815726e-06, "loss": 0.4406, "step": 2697 }, { "epoch": 0.1912796880538816, "grad_norm": 3.645014804192671, "learning_rate": 9.333662833195522e-06, "loss": 0.4639, "step": 2698 }, { "epoch": 0.191350584898972, "grad_norm": 4.998986345086934, "learning_rate": 9.33309004805797e-06, "loss": 0.4131, "step": 2699 }, { "epoch": 0.1914214817440624, "grad_norm": 8.331198487025825, "learning_rate": 9.332517034433269e-06, "loss": 0.406, "step": 2700 }, { "epoch": 0.19149237858915277, "grad_norm": 3.0848973022376116, "learning_rate": 9.33194379235164e-06, "loss": 0.4578, "step": 2701 }, { "epoch": 0.19156327543424317, "grad_norm": 2.886411126001788, "learning_rate": 9.331370321843309e-06, "loss": 0.4744, "step": 2702 }, { "epoch": 0.19163417227933358, "grad_norm": 3.0313568403354987, "learning_rate": 9.330796622938514e-06, "loss": 0.4746, "step": 2703 }, { "epoch": 0.19170506912442398, "grad_norm": 2.537198001809069, "learning_rate": 9.330222695667508e-06, "loss": 0.4232, "step": 2704 }, { "epoch": 0.19177596596951435, "grad_norm": 5.277052430676453, "learning_rate": 9.329648540060554e-06, "loss": 0.4279, "step": 2705 }, { "epoch": 0.19184686281460475, "grad_norm": 2.5203925778700405, "learning_rate": 9.329074156147928e-06, "loss": 0.4161, "step": 2706 }, { "epoch": 0.19191775965969515, "grad_norm": 3.9673476350140575, "learning_rate": 9.328499543959916e-06, "loss": 0.5269, "step": 2707 }, { "epoch": 0.19198865650478553, "grad_norm": 2.90194209603732, "learning_rate": 9.327924703526821e-06, "loss": 0.4211, "step": 2708 }, { "epoch": 0.19205955334987593, "grad_norm": 3.0288098530844, "learning_rate": 9.327349634878953e-06, "loss": 0.4259, "step": 2709 }, { "epoch": 0.19213045019496633, "grad_norm": 3.1244788181599987, "learning_rate": 9.326774338046635e-06, "loss": 0.4624, "step": 2710 }, { "epoch": 0.19220134704005673, "grad_norm": 2.10567268616392, "learning_rate": 9.326198813060204e-06, "loss": 0.3972, "step": 2711 }, { "epoch": 0.1922722438851471, "grad_norm": 10.414837600319332, "learning_rate": 9.325623059950008e-06, "loss": 0.429, "step": 2712 }, { "epoch": 0.1923431407302375, "grad_norm": 3.8806006585831656, "learning_rate": 9.325047078746405e-06, "loss": 0.4412, "step": 2713 }, { "epoch": 0.1924140375753279, "grad_norm": 4.3044250055568, "learning_rate": 9.32447086947977e-06, "loss": 0.4623, "step": 2714 }, { "epoch": 0.19248493442041828, "grad_norm": 3.5970644993048992, "learning_rate": 9.323894432180485e-06, "loss": 0.462, "step": 2715 }, { "epoch": 0.19255583126550868, "grad_norm": 1.0976320232397776, "learning_rate": 9.323317766878945e-06, "loss": 0.6897, "step": 2716 }, { "epoch": 0.19262672811059908, "grad_norm": 4.0956516191921075, "learning_rate": 9.32274087360556e-06, "loss": 0.4419, "step": 2717 }, { "epoch": 0.19269762495568948, "grad_norm": 2.5243054492291686, "learning_rate": 9.322163752390752e-06, "loss": 0.4361, "step": 2718 }, { "epoch": 0.19276852180077986, "grad_norm": 2.3316125397664917, "learning_rate": 9.321586403264947e-06, "loss": 0.4152, "step": 2719 }, { "epoch": 0.19283941864587026, "grad_norm": 4.091355515159923, "learning_rate": 9.321008826258595e-06, "loss": 0.4556, "step": 2720 }, { "epoch": 0.19291031549096066, "grad_norm": 2.1109469825180183, "learning_rate": 9.320431021402147e-06, "loss": 0.4672, "step": 2721 }, { "epoch": 0.19298121233605103, "grad_norm": 3.8804325079161606, "learning_rate": 9.319852988726075e-06, "loss": 0.4544, "step": 2722 }, { "epoch": 0.19305210918114143, "grad_norm": 2.17777144880349, "learning_rate": 9.319274728260857e-06, "loss": 0.4269, "step": 2723 }, { "epoch": 0.19312300602623184, "grad_norm": 2.7850401659034736, "learning_rate": 9.318696240036986e-06, "loss": 0.4049, "step": 2724 }, { "epoch": 0.19319390287132224, "grad_norm": 2.047059845307717, "learning_rate": 9.318117524084966e-06, "loss": 0.4291, "step": 2725 }, { "epoch": 0.1932647997164126, "grad_norm": 2.289238556622891, "learning_rate": 9.317538580435313e-06, "loss": 0.4509, "step": 2726 }, { "epoch": 0.193335696561503, "grad_norm": 5.2429927082497665, "learning_rate": 9.316959409118555e-06, "loss": 0.4249, "step": 2727 }, { "epoch": 0.1934065934065934, "grad_norm": 2.741330605152623, "learning_rate": 9.316380010165234e-06, "loss": 0.4371, "step": 2728 }, { "epoch": 0.1934774902516838, "grad_norm": 2.543239877143066, "learning_rate": 9.3158003836059e-06, "loss": 0.3641, "step": 2729 }, { "epoch": 0.1935483870967742, "grad_norm": 7.3598132869934725, "learning_rate": 9.315220529471117e-06, "loss": 0.4016, "step": 2730 }, { "epoch": 0.1936192839418646, "grad_norm": 3.972425242832599, "learning_rate": 9.31464044779146e-06, "loss": 0.5175, "step": 2731 }, { "epoch": 0.193690180786955, "grad_norm": 2.602396290465624, "learning_rate": 9.314060138597523e-06, "loss": 0.392, "step": 2732 }, { "epoch": 0.19376107763204536, "grad_norm": 4.496833181087726, "learning_rate": 9.313479601919899e-06, "loss": 0.5055, "step": 2733 }, { "epoch": 0.19383197447713577, "grad_norm": 4.06886474652572, "learning_rate": 9.312898837789204e-06, "loss": 0.4193, "step": 2734 }, { "epoch": 0.19390287132222617, "grad_norm": 3.0006118094619687, "learning_rate": 9.312317846236062e-06, "loss": 0.5096, "step": 2735 }, { "epoch": 0.19397376816731657, "grad_norm": 3.456884005246183, "learning_rate": 9.311736627291108e-06, "loss": 0.4475, "step": 2736 }, { "epoch": 0.19404466501240694, "grad_norm": 12.54600637698691, "learning_rate": 9.311155180984991e-06, "loss": 0.4015, "step": 2737 }, { "epoch": 0.19411556185749734, "grad_norm": 3.21525865050327, "learning_rate": 9.310573507348371e-06, "loss": 0.4906, "step": 2738 }, { "epoch": 0.19418645870258774, "grad_norm": 4.190231695041982, "learning_rate": 9.309991606411918e-06, "loss": 0.4306, "step": 2739 }, { "epoch": 0.19425735554767812, "grad_norm": 5.462059031705493, "learning_rate": 9.30940947820632e-06, "loss": 0.4088, "step": 2740 }, { "epoch": 0.19432825239276852, "grad_norm": 4.068373825338094, "learning_rate": 9.30882712276227e-06, "loss": 0.4307, "step": 2741 }, { "epoch": 0.19439914923785892, "grad_norm": 6.155844203243154, "learning_rate": 9.308244540110478e-06, "loss": 0.3848, "step": 2742 }, { "epoch": 0.19447004608294932, "grad_norm": 4.067150023623875, "learning_rate": 9.307661730281664e-06, "loss": 0.451, "step": 2743 }, { "epoch": 0.1945409429280397, "grad_norm": 4.029273327756357, "learning_rate": 9.307078693306556e-06, "loss": 0.4435, "step": 2744 }, { "epoch": 0.1946118397731301, "grad_norm": 5.552374783788194, "learning_rate": 9.306495429215902e-06, "loss": 0.543, "step": 2745 }, { "epoch": 0.1946827366182205, "grad_norm": 6.1769917942423875, "learning_rate": 9.305911938040458e-06, "loss": 0.431, "step": 2746 }, { "epoch": 0.19475363346331087, "grad_norm": 2.7593807670890707, "learning_rate": 9.305328219810991e-06, "loss": 0.4453, "step": 2747 }, { "epoch": 0.19482453030840127, "grad_norm": 1.7617837010473343, "learning_rate": 9.304744274558279e-06, "loss": 0.3784, "step": 2748 }, { "epoch": 0.19489542715349167, "grad_norm": 2.6810553371341723, "learning_rate": 9.304160102313119e-06, "loss": 0.428, "step": 2749 }, { "epoch": 0.19496632399858208, "grad_norm": 3.821647620893192, "learning_rate": 9.303575703106308e-06, "loss": 0.4612, "step": 2750 }, { "epoch": 0.19503722084367245, "grad_norm": 3.8488436405492736, "learning_rate": 9.302991076968669e-06, "loss": 0.4541, "step": 2751 }, { "epoch": 0.19510811768876285, "grad_norm": 2.9457143690213754, "learning_rate": 9.302406223931023e-06, "loss": 0.467, "step": 2752 }, { "epoch": 0.19517901453385325, "grad_norm": 4.189514344062737, "learning_rate": 9.301821144024214e-06, "loss": 0.4072, "step": 2753 }, { "epoch": 0.19524991137894362, "grad_norm": 2.5365989121807297, "learning_rate": 9.301235837279094e-06, "loss": 0.4305, "step": 2754 }, { "epoch": 0.19532080822403403, "grad_norm": 4.594345706948385, "learning_rate": 9.300650303726523e-06, "loss": 0.4598, "step": 2755 }, { "epoch": 0.19539170506912443, "grad_norm": 2.7026093360903847, "learning_rate": 9.30006454339738e-06, "loss": 0.4458, "step": 2756 }, { "epoch": 0.19546260191421483, "grad_norm": 2.3335270993609614, "learning_rate": 9.299478556322552e-06, "loss": 0.4271, "step": 2757 }, { "epoch": 0.1955334987593052, "grad_norm": 2.7883327963352182, "learning_rate": 9.298892342532937e-06, "loss": 0.4908, "step": 2758 }, { "epoch": 0.1956043956043956, "grad_norm": 6.27099126965471, "learning_rate": 9.298305902059449e-06, "loss": 0.4573, "step": 2759 }, { "epoch": 0.195675292449486, "grad_norm": 4.331767145135529, "learning_rate": 9.297719234933009e-06, "loss": 0.446, "step": 2760 }, { "epoch": 0.19574618929457638, "grad_norm": 3.2962463315477364, "learning_rate": 9.297132341184553e-06, "loss": 0.4381, "step": 2761 }, { "epoch": 0.19581708613966678, "grad_norm": 3.130897154067974, "learning_rate": 9.296545220845031e-06, "loss": 0.5081, "step": 2762 }, { "epoch": 0.19588798298475718, "grad_norm": 3.385604785320791, "learning_rate": 9.295957873945398e-06, "loss": 0.457, "step": 2763 }, { "epoch": 0.19595887982984758, "grad_norm": 2.355539139031749, "learning_rate": 9.295370300516628e-06, "loss": 0.3999, "step": 2764 }, { "epoch": 0.19602977667493796, "grad_norm": 2.331470969944342, "learning_rate": 9.294782500589706e-06, "loss": 0.4168, "step": 2765 }, { "epoch": 0.19610067352002836, "grad_norm": 4.364579299604977, "learning_rate": 9.294194474195622e-06, "loss": 0.5161, "step": 2766 }, { "epoch": 0.19617157036511876, "grad_norm": 3.355276507883999, "learning_rate": 9.293606221365386e-06, "loss": 0.3676, "step": 2767 }, { "epoch": 0.19624246721020916, "grad_norm": 2.5185409526410547, "learning_rate": 9.293017742130016e-06, "loss": 0.3957, "step": 2768 }, { "epoch": 0.19631336405529953, "grad_norm": 2.385793429665304, "learning_rate": 9.292429036520544e-06, "loss": 0.4155, "step": 2769 }, { "epoch": 0.19638426090038993, "grad_norm": 3.246117353974633, "learning_rate": 9.291840104568014e-06, "loss": 0.4459, "step": 2770 }, { "epoch": 0.19645515774548034, "grad_norm": 2.4202122153380077, "learning_rate": 9.291250946303478e-06, "loss": 0.4839, "step": 2771 }, { "epoch": 0.1965260545905707, "grad_norm": 3.5326040569557215, "learning_rate": 9.290661561758006e-06, "loss": 0.4085, "step": 2772 }, { "epoch": 0.1965969514356611, "grad_norm": 4.27046966017711, "learning_rate": 9.290071950962674e-06, "loss": 0.4871, "step": 2773 }, { "epoch": 0.1966678482807515, "grad_norm": 2.979255585891229, "learning_rate": 9.289482113948573e-06, "loss": 0.4769, "step": 2774 }, { "epoch": 0.1967387451258419, "grad_norm": 3.9996573924315624, "learning_rate": 9.288892050746806e-06, "loss": 0.4637, "step": 2775 }, { "epoch": 0.1968096419709323, "grad_norm": 3.2973435368035986, "learning_rate": 9.288301761388489e-06, "loss": 0.3839, "step": 2776 }, { "epoch": 0.1968805388160227, "grad_norm": 4.691006055601112, "learning_rate": 9.287711245904746e-06, "loss": 0.4764, "step": 2777 }, { "epoch": 0.1969514356611131, "grad_norm": 2.0696837892255933, "learning_rate": 9.287120504326716e-06, "loss": 0.4226, "step": 2778 }, { "epoch": 0.19702233250620346, "grad_norm": 2.4048024899085845, "learning_rate": 9.28652953668555e-06, "loss": 0.4386, "step": 2779 }, { "epoch": 0.19709322935129386, "grad_norm": 4.332979699345569, "learning_rate": 9.28593834301241e-06, "loss": 0.4927, "step": 2780 }, { "epoch": 0.19716412619638427, "grad_norm": 3.3547790055913054, "learning_rate": 9.28534692333847e-06, "loss": 0.4774, "step": 2781 }, { "epoch": 0.19723502304147467, "grad_norm": 7.731477050324674, "learning_rate": 9.284755277694915e-06, "loss": 0.5168, "step": 2782 }, { "epoch": 0.19730591988656504, "grad_norm": 2.317512817007992, "learning_rate": 9.284163406112945e-06, "loss": 0.4585, "step": 2783 }, { "epoch": 0.19737681673165544, "grad_norm": 3.44173490818737, "learning_rate": 9.283571308623769e-06, "loss": 0.4521, "step": 2784 }, { "epoch": 0.19744771357674584, "grad_norm": 6.292974032049769, "learning_rate": 9.282978985258606e-06, "loss": 0.4464, "step": 2785 }, { "epoch": 0.19751861042183622, "grad_norm": 1.1273365324344076, "learning_rate": 9.282386436048693e-06, "loss": 0.7155, "step": 2786 }, { "epoch": 0.19758950726692662, "grad_norm": 3.0587018634973533, "learning_rate": 9.281793661025275e-06, "loss": 0.4507, "step": 2787 }, { "epoch": 0.19766040411201702, "grad_norm": 3.2818406056523606, "learning_rate": 9.28120066021961e-06, "loss": 0.4557, "step": 2788 }, { "epoch": 0.19773130095710742, "grad_norm": 5.001067248380327, "learning_rate": 9.280607433662967e-06, "loss": 0.435, "step": 2789 }, { "epoch": 0.1978021978021978, "grad_norm": 2.7631945471180197, "learning_rate": 9.280013981386625e-06, "loss": 0.4391, "step": 2790 }, { "epoch": 0.1978730946472882, "grad_norm": 2.3160541369761614, "learning_rate": 9.27942030342188e-06, "loss": 0.4608, "step": 2791 }, { "epoch": 0.1979439914923786, "grad_norm": 1.1905130602240392, "learning_rate": 9.278826399800035e-06, "loss": 0.7431, "step": 2792 }, { "epoch": 0.19801488833746897, "grad_norm": 4.003516873987212, "learning_rate": 9.27823227055241e-06, "loss": 0.4771, "step": 2793 }, { "epoch": 0.19808578518255937, "grad_norm": 3.8044059746797347, "learning_rate": 9.27763791571033e-06, "loss": 0.4564, "step": 2794 }, { "epoch": 0.19815668202764977, "grad_norm": 2.887817286760997, "learning_rate": 9.277043335305138e-06, "loss": 0.3929, "step": 2795 }, { "epoch": 0.19822757887274017, "grad_norm": 4.844350120135338, "learning_rate": 9.276448529368187e-06, "loss": 0.3765, "step": 2796 }, { "epoch": 0.19829847571783055, "grad_norm": 3.0412205641040297, "learning_rate": 9.27585349793084e-06, "loss": 0.4047, "step": 2797 }, { "epoch": 0.19836937256292095, "grad_norm": 15.176406574058113, "learning_rate": 9.275258241024476e-06, "loss": 0.4114, "step": 2798 }, { "epoch": 0.19844026940801135, "grad_norm": 2.9155150244734616, "learning_rate": 9.274662758680481e-06, "loss": 0.4242, "step": 2799 }, { "epoch": 0.19851116625310175, "grad_norm": 2.5339903499410554, "learning_rate": 9.274067050930256e-06, "loss": 0.3936, "step": 2800 }, { "epoch": 0.19858206309819212, "grad_norm": 2.3708147660147203, "learning_rate": 9.273471117805213e-06, "loss": 0.4499, "step": 2801 }, { "epoch": 0.19865295994328253, "grad_norm": 3.6567791243593954, "learning_rate": 9.272874959336774e-06, "loss": 0.4275, "step": 2802 }, { "epoch": 0.19872385678837293, "grad_norm": 3.6673527166558673, "learning_rate": 9.272278575556381e-06, "loss": 0.4161, "step": 2803 }, { "epoch": 0.1987947536334633, "grad_norm": 3.142891510037181, "learning_rate": 9.271681966495477e-06, "loss": 0.4061, "step": 2804 }, { "epoch": 0.1988656504785537, "grad_norm": 4.459776682433028, "learning_rate": 9.271085132185521e-06, "loss": 0.4429, "step": 2805 }, { "epoch": 0.1989365473236441, "grad_norm": 3.9700544552325105, "learning_rate": 9.270488072657986e-06, "loss": 0.4359, "step": 2806 }, { "epoch": 0.1990074441687345, "grad_norm": 2.9062764349052053, "learning_rate": 9.269890787944356e-06, "loss": 0.4723, "step": 2807 }, { "epoch": 0.19907834101382488, "grad_norm": 2.1405668361691013, "learning_rate": 9.269293278076126e-06, "loss": 0.43, "step": 2808 }, { "epoch": 0.19914923785891528, "grad_norm": 2.7163394807954417, "learning_rate": 9.268695543084803e-06, "loss": 0.4319, "step": 2809 }, { "epoch": 0.19922013470400568, "grad_norm": 3.633290079855162, "learning_rate": 9.268097583001905e-06, "loss": 0.4509, "step": 2810 }, { "epoch": 0.19929103154909605, "grad_norm": 2.14019325570438, "learning_rate": 9.267499397858966e-06, "loss": 0.4158, "step": 2811 }, { "epoch": 0.19936192839418646, "grad_norm": 3.3196813666922558, "learning_rate": 9.266900987687524e-06, "loss": 0.4448, "step": 2812 }, { "epoch": 0.19943282523927686, "grad_norm": 3.0266395817080958, "learning_rate": 9.266302352519136e-06, "loss": 0.4652, "step": 2813 }, { "epoch": 0.19950372208436726, "grad_norm": 2.554748270918365, "learning_rate": 9.26570349238537e-06, "loss": 0.4482, "step": 2814 }, { "epoch": 0.19957461892945763, "grad_norm": 1.2620565569829405, "learning_rate": 9.265104407317802e-06, "loss": 0.7144, "step": 2815 }, { "epoch": 0.19964551577454803, "grad_norm": 6.229051613544868, "learning_rate": 9.264505097348024e-06, "loss": 0.4781, "step": 2816 }, { "epoch": 0.19971641261963843, "grad_norm": 2.342989372370436, "learning_rate": 9.263905562507638e-06, "loss": 0.4223, "step": 2817 }, { "epoch": 0.1997873094647288, "grad_norm": 1.1538996367487233, "learning_rate": 9.263305802828255e-06, "loss": 0.7427, "step": 2818 }, { "epoch": 0.1998582063098192, "grad_norm": 2.7749086804252476, "learning_rate": 9.262705818341504e-06, "loss": 0.481, "step": 2819 }, { "epoch": 0.1999291031549096, "grad_norm": 2.088903703666001, "learning_rate": 9.262105609079022e-06, "loss": 0.4653, "step": 2820 }, { "epoch": 0.2, "grad_norm": 3.535337179633026, "learning_rate": 9.261505175072459e-06, "loss": 0.4875, "step": 2821 }, { "epoch": 0.20007089684509038, "grad_norm": 4.588830722427434, "learning_rate": 9.260904516353474e-06, "loss": 0.4235, "step": 2822 }, { "epoch": 0.20014179369018079, "grad_norm": 4.012597915295204, "learning_rate": 9.260303632953742e-06, "loss": 0.4266, "step": 2823 }, { "epoch": 0.2002126905352712, "grad_norm": 3.411572645343914, "learning_rate": 9.259702524904948e-06, "loss": 0.4629, "step": 2824 }, { "epoch": 0.20028358738036156, "grad_norm": 4.09074876856639, "learning_rate": 9.259101192238788e-06, "loss": 0.4439, "step": 2825 }, { "epoch": 0.20035448422545196, "grad_norm": 5.913500986205853, "learning_rate": 9.258499634986972e-06, "loss": 0.4853, "step": 2826 }, { "epoch": 0.20042538107054236, "grad_norm": 3.2044486836482884, "learning_rate": 9.257897853181219e-06, "loss": 0.3989, "step": 2827 }, { "epoch": 0.20049627791563276, "grad_norm": 3.7066037153463336, "learning_rate": 9.257295846853263e-06, "loss": 0.4164, "step": 2828 }, { "epoch": 0.20056717476072314, "grad_norm": 3.05503280109159, "learning_rate": 9.256693616034848e-06, "loss": 0.4411, "step": 2829 }, { "epoch": 0.20063807160581354, "grad_norm": 3.235568571430898, "learning_rate": 9.256091160757729e-06, "loss": 0.4127, "step": 2830 }, { "epoch": 0.20070896845090394, "grad_norm": 4.207983919581094, "learning_rate": 9.255488481053674e-06, "loss": 0.4741, "step": 2831 }, { "epoch": 0.20077986529599434, "grad_norm": 12.435435276541877, "learning_rate": 9.254885576954465e-06, "loss": 0.4584, "step": 2832 }, { "epoch": 0.20085076214108472, "grad_norm": 4.938910153674258, "learning_rate": 9.254282448491888e-06, "loss": 0.4715, "step": 2833 }, { "epoch": 0.20092165898617512, "grad_norm": 3.9710311527923947, "learning_rate": 9.253679095697754e-06, "loss": 0.3981, "step": 2834 }, { "epoch": 0.20099255583126552, "grad_norm": 4.192761710771748, "learning_rate": 9.253075518603875e-06, "loss": 0.4243, "step": 2835 }, { "epoch": 0.2010634526763559, "grad_norm": 4.884496277286719, "learning_rate": 9.252471717242076e-06, "loss": 0.4608, "step": 2836 }, { "epoch": 0.2011343495214463, "grad_norm": 5.0917353385890305, "learning_rate": 9.251867691644197e-06, "loss": 0.4537, "step": 2837 }, { "epoch": 0.2012052463665367, "grad_norm": 2.7785550610761747, "learning_rate": 9.25126344184209e-06, "loss": 0.5069, "step": 2838 }, { "epoch": 0.2012761432116271, "grad_norm": 2.226778762688033, "learning_rate": 9.250658967867618e-06, "loss": 0.3956, "step": 2839 }, { "epoch": 0.20134704005671747, "grad_norm": 3.4804831406930528, "learning_rate": 9.250054269752654e-06, "loss": 0.4526, "step": 2840 }, { "epoch": 0.20141793690180787, "grad_norm": 3.215385050172596, "learning_rate": 9.249449347529082e-06, "loss": 0.4278, "step": 2841 }, { "epoch": 0.20148883374689827, "grad_norm": 1.0900938067537465, "learning_rate": 9.248844201228806e-06, "loss": 0.6376, "step": 2842 }, { "epoch": 0.20155973059198865, "grad_norm": 2.4402448518929636, "learning_rate": 9.248238830883732e-06, "loss": 0.4344, "step": 2843 }, { "epoch": 0.20163062743707905, "grad_norm": 3.6302479258150506, "learning_rate": 9.24763323652578e-06, "loss": 0.4385, "step": 2844 }, { "epoch": 0.20170152428216945, "grad_norm": 4.414151892696169, "learning_rate": 9.247027418186885e-06, "loss": 0.4442, "step": 2845 }, { "epoch": 0.20177242112725985, "grad_norm": 2.1131223654534907, "learning_rate": 9.246421375898993e-06, "loss": 0.4347, "step": 2846 }, { "epoch": 0.20184331797235022, "grad_norm": 6.087371272015467, "learning_rate": 9.245815109694061e-06, "loss": 0.4455, "step": 2847 }, { "epoch": 0.20191421481744062, "grad_norm": 3.3575366209106146, "learning_rate": 9.245208619604058e-06, "loss": 0.3988, "step": 2848 }, { "epoch": 0.20198511166253103, "grad_norm": 4.793451386957781, "learning_rate": 9.244601905660964e-06, "loss": 0.4834, "step": 2849 }, { "epoch": 0.2020560085076214, "grad_norm": 2.7674995653936136, "learning_rate": 9.243994967896772e-06, "loss": 0.4972, "step": 2850 }, { "epoch": 0.2021269053527118, "grad_norm": 2.8623718459335605, "learning_rate": 9.243387806343487e-06, "loss": 0.4424, "step": 2851 }, { "epoch": 0.2021978021978022, "grad_norm": 3.5069136024066494, "learning_rate": 9.242780421033121e-06, "loss": 0.4678, "step": 2852 }, { "epoch": 0.2022686990428926, "grad_norm": 3.3845560100073504, "learning_rate": 9.242172811997707e-06, "loss": 0.4396, "step": 2853 }, { "epoch": 0.20233959588798298, "grad_norm": 5.191983440245069, "learning_rate": 9.241564979269284e-06, "loss": 0.4732, "step": 2854 }, { "epoch": 0.20241049273307338, "grad_norm": 2.130252815103952, "learning_rate": 9.2409569228799e-06, "loss": 0.4358, "step": 2855 }, { "epoch": 0.20248138957816378, "grad_norm": 2.2972876171621595, "learning_rate": 9.240348642861622e-06, "loss": 0.4294, "step": 2856 }, { "epoch": 0.20255228642325415, "grad_norm": 3.826141852384723, "learning_rate": 9.239740139246523e-06, "loss": 0.4817, "step": 2857 }, { "epoch": 0.20262318326834455, "grad_norm": 2.2637038378781966, "learning_rate": 9.239131412066691e-06, "loss": 0.4128, "step": 2858 }, { "epoch": 0.20269408011343495, "grad_norm": 4.936829095412528, "learning_rate": 9.238522461354223e-06, "loss": 0.4555, "step": 2859 }, { "epoch": 0.20276497695852536, "grad_norm": 4.260527865715992, "learning_rate": 9.23791328714123e-06, "loss": 0.4733, "step": 2860 }, { "epoch": 0.20283587380361573, "grad_norm": 4.564776910131845, "learning_rate": 9.237303889459837e-06, "loss": 0.451, "step": 2861 }, { "epoch": 0.20290677064870613, "grad_norm": 9.242700663270277, "learning_rate": 9.236694268342175e-06, "loss": 0.4508, "step": 2862 }, { "epoch": 0.20297766749379653, "grad_norm": 2.1742221720544177, "learning_rate": 9.236084423820388e-06, "loss": 0.4771, "step": 2863 }, { "epoch": 0.20304856433888693, "grad_norm": 3.674026937236388, "learning_rate": 9.23547435592664e-06, "loss": 0.4517, "step": 2864 }, { "epoch": 0.2031194611839773, "grad_norm": 27.079312045455644, "learning_rate": 9.234864064693094e-06, "loss": 0.3928, "step": 2865 }, { "epoch": 0.2031903580290677, "grad_norm": 2.9372496358655575, "learning_rate": 9.234253550151933e-06, "loss": 0.4287, "step": 2866 }, { "epoch": 0.2032612548741581, "grad_norm": 5.318382763284036, "learning_rate": 9.233642812335351e-06, "loss": 0.5062, "step": 2867 }, { "epoch": 0.20333215171924848, "grad_norm": 3.3003113920713356, "learning_rate": 9.233031851275553e-06, "loss": 0.4449, "step": 2868 }, { "epoch": 0.20340304856433888, "grad_norm": 13.683939062831275, "learning_rate": 9.232420667004754e-06, "loss": 0.5102, "step": 2869 }, { "epoch": 0.20347394540942929, "grad_norm": 3.4713535191147935, "learning_rate": 9.231809259555183e-06, "loss": 0.4572, "step": 2870 }, { "epoch": 0.2035448422545197, "grad_norm": 3.1635773612060327, "learning_rate": 9.23119762895908e-06, "loss": 0.4358, "step": 2871 }, { "epoch": 0.20361573909961006, "grad_norm": 1.2988174654990972, "learning_rate": 9.230585775248696e-06, "loss": 0.6974, "step": 2872 }, { "epoch": 0.20368663594470046, "grad_norm": 2.448297878262785, "learning_rate": 9.229973698456295e-06, "loss": 0.4694, "step": 2873 }, { "epoch": 0.20375753278979086, "grad_norm": 1.1101673364035773, "learning_rate": 9.229361398614152e-06, "loss": 0.6438, "step": 2874 }, { "epoch": 0.20382842963488124, "grad_norm": 3.242841964223992, "learning_rate": 9.228748875754555e-06, "loss": 0.4538, "step": 2875 }, { "epoch": 0.20389932647997164, "grad_norm": 5.451609622096032, "learning_rate": 9.228136129909801e-06, "loss": 0.3791, "step": 2876 }, { "epoch": 0.20397022332506204, "grad_norm": 3.2261988366479204, "learning_rate": 9.227523161112203e-06, "loss": 0.4545, "step": 2877 }, { "epoch": 0.20404112017015244, "grad_norm": 3.944262901634496, "learning_rate": 9.226909969394083e-06, "loss": 0.4321, "step": 2878 }, { "epoch": 0.2041120170152428, "grad_norm": 2.0218255695493568, "learning_rate": 9.226296554787772e-06, "loss": 0.4249, "step": 2879 }, { "epoch": 0.20418291386033321, "grad_norm": 4.131863506175647, "learning_rate": 9.225682917325617e-06, "loss": 0.438, "step": 2880 }, { "epoch": 0.20425381070542362, "grad_norm": 3.3928118977226203, "learning_rate": 9.225069057039977e-06, "loss": 0.4338, "step": 2881 }, { "epoch": 0.204324707550514, "grad_norm": 5.124391856721034, "learning_rate": 9.224454973963223e-06, "loss": 0.42, "step": 2882 }, { "epoch": 0.2043956043956044, "grad_norm": 3.151230127775448, "learning_rate": 9.223840668127733e-06, "loss": 0.4316, "step": 2883 }, { "epoch": 0.2044665012406948, "grad_norm": 2.914110900511173, "learning_rate": 9.2232261395659e-06, "loss": 0.4517, "step": 2884 }, { "epoch": 0.2045373980857852, "grad_norm": 2.924345384091771, "learning_rate": 9.222611388310129e-06, "loss": 0.4507, "step": 2885 }, { "epoch": 0.20460829493087557, "grad_norm": 2.1337062251189742, "learning_rate": 9.221996414392838e-06, "loss": 0.4514, "step": 2886 }, { "epoch": 0.20467919177596597, "grad_norm": 2.3403416730210784, "learning_rate": 9.221381217846451e-06, "loss": 0.4709, "step": 2887 }, { "epoch": 0.20475008862105637, "grad_norm": 8.307288162827556, "learning_rate": 9.220765798703413e-06, "loss": 0.4316, "step": 2888 }, { "epoch": 0.20482098546614674, "grad_norm": 3.145841367668703, "learning_rate": 9.220150156996171e-06, "loss": 0.4494, "step": 2889 }, { "epoch": 0.20489188231123714, "grad_norm": 4.3747978957918745, "learning_rate": 9.21953429275719e-06, "loss": 0.4322, "step": 2890 }, { "epoch": 0.20496277915632755, "grad_norm": 21.508446404228266, "learning_rate": 9.218918206018947e-06, "loss": 0.3994, "step": 2891 }, { "epoch": 0.20503367600141795, "grad_norm": 3.8255910285079415, "learning_rate": 9.218301896813926e-06, "loss": 0.4562, "step": 2892 }, { "epoch": 0.20510457284650832, "grad_norm": 4.175300815418422, "learning_rate": 9.217685365174625e-06, "loss": 0.383, "step": 2893 }, { "epoch": 0.20517546969159872, "grad_norm": 2.9011029676511693, "learning_rate": 9.217068611133558e-06, "loss": 0.471, "step": 2894 }, { "epoch": 0.20524636653668912, "grad_norm": 3.646375429851814, "learning_rate": 9.216451634723243e-06, "loss": 0.4342, "step": 2895 }, { "epoch": 0.20531726338177952, "grad_norm": 3.3654549742355417, "learning_rate": 9.215834435976214e-06, "loss": 0.4455, "step": 2896 }, { "epoch": 0.2053881602268699, "grad_norm": 2.758390589648515, "learning_rate": 9.215217014925018e-06, "loss": 0.4473, "step": 2897 }, { "epoch": 0.2054590570719603, "grad_norm": 3.300920884053894, "learning_rate": 9.214599371602213e-06, "loss": 0.4586, "step": 2898 }, { "epoch": 0.2055299539170507, "grad_norm": 5.275884131214821, "learning_rate": 9.213981506040364e-06, "loss": 0.4075, "step": 2899 }, { "epoch": 0.20560085076214107, "grad_norm": 2.8306763354907347, "learning_rate": 9.213363418272055e-06, "loss": 0.4004, "step": 2900 }, { "epoch": 0.20567174760723148, "grad_norm": 2.5560219936445057, "learning_rate": 9.212745108329877e-06, "loss": 0.4489, "step": 2901 }, { "epoch": 0.20574264445232188, "grad_norm": 2.663635506164763, "learning_rate": 9.212126576246433e-06, "loss": 0.4116, "step": 2902 }, { "epoch": 0.20581354129741228, "grad_norm": 2.5384944563801155, "learning_rate": 9.211507822054343e-06, "loss": 0.4572, "step": 2903 }, { "epoch": 0.20588443814250265, "grad_norm": 5.476813911369501, "learning_rate": 9.210888845786227e-06, "loss": 0.4073, "step": 2904 }, { "epoch": 0.20595533498759305, "grad_norm": 3.504147425276632, "learning_rate": 9.21026964747473e-06, "loss": 0.5496, "step": 2905 }, { "epoch": 0.20602623183268345, "grad_norm": 2.986637904632361, "learning_rate": 9.2096502271525e-06, "loss": 0.5011, "step": 2906 }, { "epoch": 0.20609712867777383, "grad_norm": 2.9731926769467023, "learning_rate": 9.209030584852203e-06, "loss": 0.4418, "step": 2907 }, { "epoch": 0.20616802552286423, "grad_norm": 2.6895777150707385, "learning_rate": 9.20841072060651e-06, "loss": 0.4617, "step": 2908 }, { "epoch": 0.20623892236795463, "grad_norm": 3.5955518879472077, "learning_rate": 9.207790634448104e-06, "loss": 0.4326, "step": 2909 }, { "epoch": 0.20630981921304503, "grad_norm": 2.103567507847261, "learning_rate": 9.20717032640969e-06, "loss": 0.4494, "step": 2910 }, { "epoch": 0.2063807160581354, "grad_norm": 2.049299764274864, "learning_rate": 9.20654979652397e-06, "loss": 0.4093, "step": 2911 }, { "epoch": 0.2064516129032258, "grad_norm": 2.457939775496065, "learning_rate": 9.205929044823671e-06, "loss": 0.4395, "step": 2912 }, { "epoch": 0.2065225097483162, "grad_norm": 2.672315694996468, "learning_rate": 9.205308071341523e-06, "loss": 0.4665, "step": 2913 }, { "epoch": 0.20659340659340658, "grad_norm": 2.265243600075359, "learning_rate": 9.204686876110272e-06, "loss": 0.4595, "step": 2914 }, { "epoch": 0.20666430343849698, "grad_norm": 2.178252674403452, "learning_rate": 9.204065459162669e-06, "loss": 0.4437, "step": 2915 }, { "epoch": 0.20673520028358738, "grad_norm": 9.09428472957741, "learning_rate": 9.203443820531489e-06, "loss": 0.4192, "step": 2916 }, { "epoch": 0.20680609712867778, "grad_norm": 7.056779982816904, "learning_rate": 9.202821960249506e-06, "loss": 0.4661, "step": 2917 }, { "epoch": 0.20687699397376816, "grad_norm": 7.884401938260659, "learning_rate": 9.202199878349514e-06, "loss": 0.3876, "step": 2918 }, { "epoch": 0.20694789081885856, "grad_norm": 2.409350780446531, "learning_rate": 9.201577574864316e-06, "loss": 0.4629, "step": 2919 }, { "epoch": 0.20701878766394896, "grad_norm": 2.651611581615528, "learning_rate": 9.200955049826722e-06, "loss": 0.4858, "step": 2920 }, { "epoch": 0.20708968450903933, "grad_norm": 2.801644021119771, "learning_rate": 9.200332303269565e-06, "loss": 0.4244, "step": 2921 }, { "epoch": 0.20716058135412974, "grad_norm": 2.5508136599718907, "learning_rate": 9.19970933522568e-06, "loss": 0.4559, "step": 2922 }, { "epoch": 0.20723147819922014, "grad_norm": 3.2355535686641916, "learning_rate": 9.199086145727915e-06, "loss": 0.4394, "step": 2923 }, { "epoch": 0.20730237504431054, "grad_norm": 4.435399906856687, "learning_rate": 9.198462734809133e-06, "loss": 0.4661, "step": 2924 }, { "epoch": 0.2073732718894009, "grad_norm": 4.945622081415492, "learning_rate": 9.197839102502207e-06, "loss": 0.4531, "step": 2925 }, { "epoch": 0.2074441687344913, "grad_norm": 4.041776717147399, "learning_rate": 9.19721524884002e-06, "loss": 0.5, "step": 2926 }, { "epoch": 0.20751506557958171, "grad_norm": 1.9767590979785339, "learning_rate": 9.196591173855469e-06, "loss": 0.442, "step": 2927 }, { "epoch": 0.20758596242467212, "grad_norm": 5.290658626988147, "learning_rate": 9.195966877581462e-06, "loss": 0.3971, "step": 2928 }, { "epoch": 0.2076568592697625, "grad_norm": 18.577915083244044, "learning_rate": 9.19534236005092e-06, "loss": 0.4807, "step": 2929 }, { "epoch": 0.2077277561148529, "grad_norm": 5.687818034838154, "learning_rate": 9.194717621296773e-06, "loss": 0.4899, "step": 2930 }, { "epoch": 0.2077986529599433, "grad_norm": 2.7458543370286033, "learning_rate": 9.194092661351964e-06, "loss": 0.4632, "step": 2931 }, { "epoch": 0.20786954980503367, "grad_norm": 6.947368707500006, "learning_rate": 9.193467480249447e-06, "loss": 0.4754, "step": 2932 }, { "epoch": 0.20794044665012407, "grad_norm": 8.880815454986493, "learning_rate": 9.192842078022191e-06, "loss": 0.4469, "step": 2933 }, { "epoch": 0.20801134349521447, "grad_norm": 1.0071620889980932, "learning_rate": 9.192216454703172e-06, "loss": 0.6549, "step": 2934 }, { "epoch": 0.20808224034030487, "grad_norm": 2.2861496749551318, "learning_rate": 9.191590610325377e-06, "loss": 0.4614, "step": 2935 }, { "epoch": 0.20815313718539524, "grad_norm": 3.0315380101182323, "learning_rate": 9.190964544921811e-06, "loss": 0.4567, "step": 2936 }, { "epoch": 0.20822403403048564, "grad_norm": 2.293015867662813, "learning_rate": 9.190338258525487e-06, "loss": 0.4431, "step": 2937 }, { "epoch": 0.20829493087557605, "grad_norm": 4.061278165397014, "learning_rate": 9.189711751169427e-06, "loss": 0.4689, "step": 2938 }, { "epoch": 0.20836582772066642, "grad_norm": 2.091527114216211, "learning_rate": 9.189085022886668e-06, "loss": 0.4445, "step": 2939 }, { "epoch": 0.20843672456575682, "grad_norm": 12.99047501257922, "learning_rate": 9.188458073710261e-06, "loss": 0.4068, "step": 2940 }, { "epoch": 0.20850762141084722, "grad_norm": 5.176194661251275, "learning_rate": 9.18783090367326e-06, "loss": 0.4437, "step": 2941 }, { "epoch": 0.20857851825593762, "grad_norm": 1.3659424409017067, "learning_rate": 9.18720351280874e-06, "loss": 0.7303, "step": 2942 }, { "epoch": 0.208649415101028, "grad_norm": 5.6268959045069735, "learning_rate": 9.186575901149782e-06, "loss": 0.4243, "step": 2943 }, { "epoch": 0.2087203119461184, "grad_norm": 7.3851554415123015, "learning_rate": 9.185948068729482e-06, "loss": 0.4622, "step": 2944 }, { "epoch": 0.2087912087912088, "grad_norm": 5.897641854551717, "learning_rate": 9.185320015580945e-06, "loss": 0.4429, "step": 2945 }, { "epoch": 0.20886210563629917, "grad_norm": 2.446947055813246, "learning_rate": 9.184691741737291e-06, "loss": 0.4429, "step": 2946 }, { "epoch": 0.20893300248138957, "grad_norm": 2.6100775092638715, "learning_rate": 9.184063247231644e-06, "loss": 0.491, "step": 2947 }, { "epoch": 0.20900389932647997, "grad_norm": 5.725812095792712, "learning_rate": 9.18343453209715e-06, "loss": 0.4336, "step": 2948 }, { "epoch": 0.20907479617157038, "grad_norm": 2.6636394734782827, "learning_rate": 9.182805596366962e-06, "loss": 0.4771, "step": 2949 }, { "epoch": 0.20914569301666075, "grad_norm": 5.297472220368466, "learning_rate": 9.18217644007424e-06, "loss": 0.4649, "step": 2950 }, { "epoch": 0.20921658986175115, "grad_norm": 2.001634156420959, "learning_rate": 9.181547063252164e-06, "loss": 0.4332, "step": 2951 }, { "epoch": 0.20928748670684155, "grad_norm": 2.531890357061793, "learning_rate": 9.180917465933919e-06, "loss": 0.4972, "step": 2952 }, { "epoch": 0.20935838355193193, "grad_norm": 2.433521276428944, "learning_rate": 9.180287648152703e-06, "loss": 0.4488, "step": 2953 }, { "epoch": 0.20942928039702233, "grad_norm": 3.7511653268301304, "learning_rate": 9.17965760994173e-06, "loss": 0.4204, "step": 2954 }, { "epoch": 0.20950017724211273, "grad_norm": 7.0364165034278034, "learning_rate": 9.179027351334222e-06, "loss": 0.4858, "step": 2955 }, { "epoch": 0.20957107408720313, "grad_norm": 6.3059414988945806, "learning_rate": 9.17839687236341e-06, "loss": 0.4669, "step": 2956 }, { "epoch": 0.2096419709322935, "grad_norm": 3.064352028176949, "learning_rate": 9.177766173062545e-06, "loss": 0.4579, "step": 2957 }, { "epoch": 0.2097128677773839, "grad_norm": 4.1122287049577615, "learning_rate": 9.17713525346488e-06, "loss": 0.4572, "step": 2958 }, { "epoch": 0.2097837646224743, "grad_norm": 0.9583274999976205, "learning_rate": 9.176504113603683e-06, "loss": 0.6533, "step": 2959 }, { "epoch": 0.2098546614675647, "grad_norm": 2.3672903379636008, "learning_rate": 9.175872753512236e-06, "loss": 0.3978, "step": 2960 }, { "epoch": 0.20992555831265508, "grad_norm": 2.3984819248764713, "learning_rate": 9.175241173223832e-06, "loss": 0.4396, "step": 2961 }, { "epoch": 0.20999645515774548, "grad_norm": 2.709111029701677, "learning_rate": 9.174609372771775e-06, "loss": 0.4371, "step": 2962 }, { "epoch": 0.21006735200283588, "grad_norm": 5.312282920348376, "learning_rate": 9.173977352189379e-06, "loss": 0.4922, "step": 2963 }, { "epoch": 0.21013824884792626, "grad_norm": 2.7167715223714586, "learning_rate": 9.173345111509972e-06, "loss": 0.4136, "step": 2964 }, { "epoch": 0.21020914569301666, "grad_norm": 3.1913800441852977, "learning_rate": 9.172712650766891e-06, "loss": 0.4665, "step": 2965 }, { "epoch": 0.21028004253810706, "grad_norm": 1.1322683525905057, "learning_rate": 9.172079969993489e-06, "loss": 0.7257, "step": 2966 }, { "epoch": 0.21035093938319746, "grad_norm": 27.468927816145946, "learning_rate": 9.171447069223122e-06, "loss": 0.3901, "step": 2967 }, { "epoch": 0.21042183622828783, "grad_norm": 2.911539641623348, "learning_rate": 9.17081394848917e-06, "loss": 0.4156, "step": 2968 }, { "epoch": 0.21049273307337824, "grad_norm": 4.8978207957431605, "learning_rate": 9.170180607825015e-06, "loss": 0.4134, "step": 2969 }, { "epoch": 0.21056362991846864, "grad_norm": 5.17415730301948, "learning_rate": 9.169547047264052e-06, "loss": 0.4118, "step": 2970 }, { "epoch": 0.210634526763559, "grad_norm": 2.376624688477135, "learning_rate": 9.168913266839692e-06, "loss": 0.448, "step": 2971 }, { "epoch": 0.2107054236086494, "grad_norm": 1.925213731796088, "learning_rate": 9.168279266585352e-06, "loss": 0.455, "step": 2972 }, { "epoch": 0.2107763204537398, "grad_norm": 2.1688348325553126, "learning_rate": 9.167645046534465e-06, "loss": 0.4441, "step": 2973 }, { "epoch": 0.21084721729883021, "grad_norm": 2.3527076754081833, "learning_rate": 9.167010606720477e-06, "loss": 0.3839, "step": 2974 }, { "epoch": 0.2109181141439206, "grad_norm": 3.5301692707179217, "learning_rate": 9.166375947176835e-06, "loss": 0.4145, "step": 2975 }, { "epoch": 0.210989010989011, "grad_norm": 11.293235983061908, "learning_rate": 9.165741067937013e-06, "loss": 0.4335, "step": 2976 }, { "epoch": 0.2110599078341014, "grad_norm": 3.195886800131238, "learning_rate": 9.16510596903448e-06, "loss": 0.3953, "step": 2977 }, { "epoch": 0.21113080467919176, "grad_norm": 2.1140608483232817, "learning_rate": 9.164470650502735e-06, "loss": 0.4617, "step": 2978 }, { "epoch": 0.21120170152428216, "grad_norm": 4.151691193964577, "learning_rate": 9.163835112375273e-06, "loss": 0.4687, "step": 2979 }, { "epoch": 0.21127259836937257, "grad_norm": 2.135686774049568, "learning_rate": 9.163199354685605e-06, "loss": 0.4248, "step": 2980 }, { "epoch": 0.21134349521446297, "grad_norm": 2.0731269176079135, "learning_rate": 9.16256337746726e-06, "loss": 0.4618, "step": 2981 }, { "epoch": 0.21141439205955334, "grad_norm": 2.642810966161312, "learning_rate": 9.16192718075377e-06, "loss": 0.4113, "step": 2982 }, { "epoch": 0.21148528890464374, "grad_norm": 3.1162774814363337, "learning_rate": 9.161290764578682e-06, "loss": 0.4344, "step": 2983 }, { "epoch": 0.21155618574973414, "grad_norm": 4.442000397412193, "learning_rate": 9.160654128975556e-06, "loss": 0.4624, "step": 2984 }, { "epoch": 0.21162708259482452, "grad_norm": 2.4837043572624213, "learning_rate": 9.160017273977964e-06, "loss": 0.4365, "step": 2985 }, { "epoch": 0.21169797943991492, "grad_norm": 2.7609332831976925, "learning_rate": 9.159380199619485e-06, "loss": 0.3985, "step": 2986 }, { "epoch": 0.21176887628500532, "grad_norm": 2.48517383521568, "learning_rate": 9.158742905933713e-06, "loss": 0.4924, "step": 2987 }, { "epoch": 0.21183977313009572, "grad_norm": 3.504226302904738, "learning_rate": 9.158105392954253e-06, "loss": 0.4285, "step": 2988 }, { "epoch": 0.2119106699751861, "grad_norm": 2.8369064359821987, "learning_rate": 9.157467660714721e-06, "loss": 0.4088, "step": 2989 }, { "epoch": 0.2119815668202765, "grad_norm": 4.686450502842966, "learning_rate": 9.156829709248748e-06, "loss": 0.4906, "step": 2990 }, { "epoch": 0.2120524636653669, "grad_norm": 3.3460918660158905, "learning_rate": 9.156191538589971e-06, "loss": 0.4504, "step": 2991 }, { "epoch": 0.2121233605104573, "grad_norm": 3.8401832019141087, "learning_rate": 9.155553148772043e-06, "loss": 0.3843, "step": 2992 }, { "epoch": 0.21219425735554767, "grad_norm": 2.5417510370329395, "learning_rate": 9.154914539828623e-06, "loss": 0.4331, "step": 2993 }, { "epoch": 0.21226515420063807, "grad_norm": 2.636722594667918, "learning_rate": 9.15427571179339e-06, "loss": 0.5089, "step": 2994 }, { "epoch": 0.21233605104572847, "grad_norm": 2.5942960431917297, "learning_rate": 9.153636664700027e-06, "loss": 0.4879, "step": 2995 }, { "epoch": 0.21240694789081885, "grad_norm": 5.2204326397086405, "learning_rate": 9.152997398582235e-06, "loss": 0.4656, "step": 2996 }, { "epoch": 0.21247784473590925, "grad_norm": 4.302600027186246, "learning_rate": 9.152357913473717e-06, "loss": 0.4658, "step": 2997 }, { "epoch": 0.21254874158099965, "grad_norm": 9.10204021998151, "learning_rate": 9.151718209408198e-06, "loss": 0.418, "step": 2998 }, { "epoch": 0.21261963842609005, "grad_norm": 3.474414169981547, "learning_rate": 9.15107828641941e-06, "loss": 0.4354, "step": 2999 }, { "epoch": 0.21269053527118043, "grad_norm": 2.928873988097836, "learning_rate": 9.150438144541094e-06, "loss": 0.4969, "step": 3000 }, { "epoch": 0.21276143211627083, "grad_norm": 4.29723704198127, "learning_rate": 9.149797783807007e-06, "loss": 0.4269, "step": 3001 }, { "epoch": 0.21283232896136123, "grad_norm": 6.683240474192202, "learning_rate": 9.149157204250915e-06, "loss": 0.4498, "step": 3002 }, { "epoch": 0.2129032258064516, "grad_norm": 6.31978797296038, "learning_rate": 9.148516405906599e-06, "loss": 0.4394, "step": 3003 }, { "epoch": 0.212974122651542, "grad_norm": 3.6800967962837845, "learning_rate": 9.147875388807846e-06, "loss": 0.4099, "step": 3004 }, { "epoch": 0.2130450194966324, "grad_norm": 3.4978949869276237, "learning_rate": 9.147234152988456e-06, "loss": 0.3879, "step": 3005 }, { "epoch": 0.2131159163417228, "grad_norm": 3.1377757929393635, "learning_rate": 9.146592698482243e-06, "loss": 0.4618, "step": 3006 }, { "epoch": 0.21318681318681318, "grad_norm": 3.860066042696554, "learning_rate": 9.145951025323034e-06, "loss": 0.4451, "step": 3007 }, { "epoch": 0.21325771003190358, "grad_norm": 5.28614041847842, "learning_rate": 9.145309133544663e-06, "loss": 0.4022, "step": 3008 }, { "epoch": 0.21332860687699398, "grad_norm": 6.313475726749499, "learning_rate": 9.144667023180978e-06, "loss": 0.441, "step": 3009 }, { "epoch": 0.21339950372208435, "grad_norm": 7.439264008953669, "learning_rate": 9.144024694265835e-06, "loss": 0.4462, "step": 3010 }, { "epoch": 0.21347040056717476, "grad_norm": 1.1275227564577033, "learning_rate": 9.14338214683311e-06, "loss": 0.7145, "step": 3011 }, { "epoch": 0.21354129741226516, "grad_norm": 2.7298105522197025, "learning_rate": 9.14273938091668e-06, "loss": 0.4091, "step": 3012 }, { "epoch": 0.21361219425735556, "grad_norm": 3.70402766463013, "learning_rate": 9.14209639655044e-06, "loss": 0.4439, "step": 3013 }, { "epoch": 0.21368309110244593, "grad_norm": 2.2685551296104145, "learning_rate": 9.141453193768296e-06, "loss": 0.3929, "step": 3014 }, { "epoch": 0.21375398794753633, "grad_norm": 5.179716610127118, "learning_rate": 9.140809772604164e-06, "loss": 0.4321, "step": 3015 }, { "epoch": 0.21382488479262673, "grad_norm": 5.1826719437442925, "learning_rate": 9.140166133091973e-06, "loss": 0.4233, "step": 3016 }, { "epoch": 0.2138957816377171, "grad_norm": 3.7830001763056265, "learning_rate": 9.13952227526566e-06, "loss": 0.4609, "step": 3017 }, { "epoch": 0.2139666784828075, "grad_norm": 4.964410877227326, "learning_rate": 9.138878199159179e-06, "loss": 0.3937, "step": 3018 }, { "epoch": 0.2140375753278979, "grad_norm": 2.8961433451240293, "learning_rate": 9.138233904806492e-06, "loss": 0.4133, "step": 3019 }, { "epoch": 0.2141084721729883, "grad_norm": 4.6235120012471755, "learning_rate": 9.137589392241571e-06, "loss": 0.4353, "step": 3020 }, { "epoch": 0.21417936901807869, "grad_norm": 5.9291517069100985, "learning_rate": 9.136944661498404e-06, "loss": 0.5048, "step": 3021 }, { "epoch": 0.2142502658631691, "grad_norm": 3.024947162806642, "learning_rate": 9.136299712610987e-06, "loss": 0.3894, "step": 3022 }, { "epoch": 0.2143211627082595, "grad_norm": 3.3596838391781843, "learning_rate": 9.135654545613331e-06, "loss": 0.4309, "step": 3023 }, { "epoch": 0.2143920595533499, "grad_norm": 2.9350438735779685, "learning_rate": 9.135009160539453e-06, "loss": 0.4417, "step": 3024 }, { "epoch": 0.21446295639844026, "grad_norm": 13.17739557429971, "learning_rate": 9.134363557423386e-06, "loss": 0.4432, "step": 3025 }, { "epoch": 0.21453385324353066, "grad_norm": 8.857338168756517, "learning_rate": 9.133717736299173e-06, "loss": 0.4079, "step": 3026 }, { "epoch": 0.21460475008862107, "grad_norm": 2.14247077145818, "learning_rate": 9.133071697200869e-06, "loss": 0.4699, "step": 3027 }, { "epoch": 0.21467564693371144, "grad_norm": 5.863544803921452, "learning_rate": 9.132425440162539e-06, "loss": 0.4362, "step": 3028 }, { "epoch": 0.21474654377880184, "grad_norm": 2.696749342008429, "learning_rate": 9.131778965218261e-06, "loss": 0.475, "step": 3029 }, { "epoch": 0.21481744062389224, "grad_norm": 21.0935165519619, "learning_rate": 9.131132272402125e-06, "loss": 0.5224, "step": 3030 }, { "epoch": 0.21488833746898264, "grad_norm": 4.8324838767394525, "learning_rate": 9.130485361748231e-06, "loss": 0.4675, "step": 3031 }, { "epoch": 0.21495923431407302, "grad_norm": 3.6410663407455366, "learning_rate": 9.129838233290693e-06, "loss": 0.4881, "step": 3032 }, { "epoch": 0.21503013115916342, "grad_norm": 2.8434700277845817, "learning_rate": 9.129190887063631e-06, "loss": 0.4157, "step": 3033 }, { "epoch": 0.21510102800425382, "grad_norm": 5.431475813329878, "learning_rate": 9.128543323101182e-06, "loss": 0.4443, "step": 3034 }, { "epoch": 0.2151719248493442, "grad_norm": 2.4506511786869463, "learning_rate": 9.127895541437493e-06, "loss": 0.4199, "step": 3035 }, { "epoch": 0.2152428216944346, "grad_norm": 5.357418082974827, "learning_rate": 9.12724754210672e-06, "loss": 0.4381, "step": 3036 }, { "epoch": 0.215313718539525, "grad_norm": 2.6639990980420936, "learning_rate": 9.126599325143035e-06, "loss": 0.4602, "step": 3037 }, { "epoch": 0.2153846153846154, "grad_norm": 2.616800223773587, "learning_rate": 9.125950890580619e-06, "loss": 0.4757, "step": 3038 }, { "epoch": 0.21545551222970577, "grad_norm": 2.937766991449456, "learning_rate": 9.125302238453662e-06, "loss": 0.4175, "step": 3039 }, { "epoch": 0.21552640907479617, "grad_norm": 28.135321570792673, "learning_rate": 9.124653368796369e-06, "loss": 0.4259, "step": 3040 }, { "epoch": 0.21559730591988657, "grad_norm": 2.3225486504598836, "learning_rate": 9.124004281642955e-06, "loss": 0.4096, "step": 3041 }, { "epoch": 0.21566820276497695, "grad_norm": 3.5269472365549666, "learning_rate": 9.12335497702765e-06, "loss": 0.4591, "step": 3042 }, { "epoch": 0.21573909961006735, "grad_norm": 3.5327044563675023, "learning_rate": 9.122705454984688e-06, "loss": 0.4259, "step": 3043 }, { "epoch": 0.21580999645515775, "grad_norm": 3.1939302539080683, "learning_rate": 9.122055715548322e-06, "loss": 0.4559, "step": 3044 }, { "epoch": 0.21588089330024815, "grad_norm": 2.5852266266629123, "learning_rate": 9.121405758752811e-06, "loss": 0.4452, "step": 3045 }, { "epoch": 0.21595179014533852, "grad_norm": 2.1371861997558543, "learning_rate": 9.120755584632428e-06, "loss": 0.4109, "step": 3046 }, { "epoch": 0.21602268699042892, "grad_norm": 2.3022113446378922, "learning_rate": 9.120105193221457e-06, "loss": 0.4261, "step": 3047 }, { "epoch": 0.21609358383551933, "grad_norm": 3.1005881431479203, "learning_rate": 9.119454584554197e-06, "loss": 0.4648, "step": 3048 }, { "epoch": 0.2161644806806097, "grad_norm": 2.3364624158694456, "learning_rate": 9.11880375866495e-06, "loss": 0.432, "step": 3049 }, { "epoch": 0.2162353775257001, "grad_norm": 2.9901861512002483, "learning_rate": 9.11815271558804e-06, "loss": 0.4632, "step": 3050 }, { "epoch": 0.2163062743707905, "grad_norm": 2.710644934444615, "learning_rate": 9.117501455357792e-06, "loss": 0.4495, "step": 3051 }, { "epoch": 0.2163771712158809, "grad_norm": 3.947443949892295, "learning_rate": 9.11684997800855e-06, "loss": 0.3987, "step": 3052 }, { "epoch": 0.21644806806097128, "grad_norm": 3.228265343950032, "learning_rate": 9.116198283574666e-06, "loss": 0.4532, "step": 3053 }, { "epoch": 0.21651896490606168, "grad_norm": 15.104343797719494, "learning_rate": 9.115546372090505e-06, "loss": 0.4249, "step": 3054 }, { "epoch": 0.21658986175115208, "grad_norm": 3.1902540223730447, "learning_rate": 9.114894243590443e-06, "loss": 0.4688, "step": 3055 }, { "epoch": 0.21666075859624248, "grad_norm": 3.5807121380077334, "learning_rate": 9.114241898108867e-06, "loss": 0.4908, "step": 3056 }, { "epoch": 0.21673165544133285, "grad_norm": 6.476672263079115, "learning_rate": 9.113589335680175e-06, "loss": 0.4223, "step": 3057 }, { "epoch": 0.21680255228642326, "grad_norm": 4.6598058202343475, "learning_rate": 9.112936556338779e-06, "loss": 0.4273, "step": 3058 }, { "epoch": 0.21687344913151366, "grad_norm": 2.8730817801525603, "learning_rate": 9.112283560119096e-06, "loss": 0.4806, "step": 3059 }, { "epoch": 0.21694434597660403, "grad_norm": 4.255012907295759, "learning_rate": 9.111630347055565e-06, "loss": 0.4581, "step": 3060 }, { "epoch": 0.21701524282169443, "grad_norm": 4.590695896481188, "learning_rate": 9.110976917182626e-06, "loss": 0.456, "step": 3061 }, { "epoch": 0.21708613966678483, "grad_norm": 2.700699317597091, "learning_rate": 9.110323270534735e-06, "loss": 0.5109, "step": 3062 }, { "epoch": 0.21715703651187523, "grad_norm": 4.775646618027997, "learning_rate": 9.109669407146363e-06, "loss": 0.4421, "step": 3063 }, { "epoch": 0.2172279333569656, "grad_norm": 3.208278280162353, "learning_rate": 9.109015327051986e-06, "loss": 0.4544, "step": 3064 }, { "epoch": 0.217298830202056, "grad_norm": 3.0437639876570004, "learning_rate": 9.108361030286094e-06, "loss": 0.3937, "step": 3065 }, { "epoch": 0.2173697270471464, "grad_norm": 11.686577458048134, "learning_rate": 9.10770651688319e-06, "loss": 0.4155, "step": 3066 }, { "epoch": 0.21744062389223678, "grad_norm": 1.6745976628371286, "learning_rate": 9.107051786877786e-06, "loss": 0.461, "step": 3067 }, { "epoch": 0.21751152073732719, "grad_norm": 10.757637040176677, "learning_rate": 9.106396840304404e-06, "loss": 0.4124, "step": 3068 }, { "epoch": 0.2175824175824176, "grad_norm": 3.7592896163553045, "learning_rate": 9.105741677197587e-06, "loss": 0.4122, "step": 3069 }, { "epoch": 0.217653314427508, "grad_norm": 7.50041453158234, "learning_rate": 9.105086297591874e-06, "loss": 0.4329, "step": 3070 }, { "epoch": 0.21772421127259836, "grad_norm": 4.214374361315134, "learning_rate": 9.104430701521827e-06, "loss": 0.4757, "step": 3071 }, { "epoch": 0.21779510811768876, "grad_norm": 3.111015510304076, "learning_rate": 9.103774889022018e-06, "loss": 0.4578, "step": 3072 }, { "epoch": 0.21786600496277916, "grad_norm": 2.730073614812885, "learning_rate": 9.103118860127027e-06, "loss": 0.4044, "step": 3073 }, { "epoch": 0.21793690180786954, "grad_norm": 2.626232712996172, "learning_rate": 9.102462614871446e-06, "loss": 0.3915, "step": 3074 }, { "epoch": 0.21800779865295994, "grad_norm": 7.66415182093164, "learning_rate": 9.10180615328988e-06, "loss": 0.3937, "step": 3075 }, { "epoch": 0.21807869549805034, "grad_norm": 9.982406876307472, "learning_rate": 9.101149475416944e-06, "loss": 0.4136, "step": 3076 }, { "epoch": 0.21814959234314074, "grad_norm": 5.117568076326015, "learning_rate": 9.100492581287268e-06, "loss": 0.441, "step": 3077 }, { "epoch": 0.21822048918823111, "grad_norm": 3.8928791593836025, "learning_rate": 9.099835470935487e-06, "loss": 0.4347, "step": 3078 }, { "epoch": 0.21829138603332152, "grad_norm": 3.9376187358164123, "learning_rate": 9.099178144396252e-06, "loss": 0.5096, "step": 3079 }, { "epoch": 0.21836228287841192, "grad_norm": 4.422681082815805, "learning_rate": 9.098520601704226e-06, "loss": 0.4256, "step": 3080 }, { "epoch": 0.2184331797235023, "grad_norm": 14.903531162630689, "learning_rate": 9.097862842894079e-06, "loss": 0.437, "step": 3081 }, { "epoch": 0.2185040765685927, "grad_norm": 2.4366442630873766, "learning_rate": 9.097204868000498e-06, "loss": 0.451, "step": 3082 }, { "epoch": 0.2185749734136831, "grad_norm": 4.896805161212825, "learning_rate": 9.096546677058176e-06, "loss": 0.4613, "step": 3083 }, { "epoch": 0.2186458702587735, "grad_norm": 6.581007125270683, "learning_rate": 9.09588827010182e-06, "loss": 0.4366, "step": 3084 }, { "epoch": 0.21871676710386387, "grad_norm": 2.6912081612021934, "learning_rate": 9.09522964716615e-06, "loss": 0.4997, "step": 3085 }, { "epoch": 0.21878766394895427, "grad_norm": 2.9628203373592608, "learning_rate": 9.094570808285897e-06, "loss": 0.4521, "step": 3086 }, { "epoch": 0.21885856079404467, "grad_norm": 4.039803277881627, "learning_rate": 9.093911753495798e-06, "loss": 0.4867, "step": 3087 }, { "epoch": 0.21892945763913507, "grad_norm": 3.7855911931595907, "learning_rate": 9.093252482830608e-06, "loss": 0.4037, "step": 3088 }, { "epoch": 0.21900035448422545, "grad_norm": 2.8714200945217967, "learning_rate": 9.09259299632509e-06, "loss": 0.4732, "step": 3089 }, { "epoch": 0.21907125132931585, "grad_norm": 2.6732743129168353, "learning_rate": 9.09193329401402e-06, "loss": 0.4449, "step": 3090 }, { "epoch": 0.21914214817440625, "grad_norm": 1.6649759481256459, "learning_rate": 9.091273375932184e-06, "loss": 0.3859, "step": 3091 }, { "epoch": 0.21921304501949662, "grad_norm": 1.059038425256094, "learning_rate": 9.090613242114381e-06, "loss": 0.665, "step": 3092 }, { "epoch": 0.21928394186458702, "grad_norm": 3.1404301828786454, "learning_rate": 9.089952892595418e-06, "loss": 0.4462, "step": 3093 }, { "epoch": 0.21935483870967742, "grad_norm": 3.857578282162236, "learning_rate": 9.089292327410119e-06, "loss": 0.5208, "step": 3094 }, { "epoch": 0.21942573555476783, "grad_norm": 3.1894529333239734, "learning_rate": 9.088631546593313e-06, "loss": 0.4765, "step": 3095 }, { "epoch": 0.2194966323998582, "grad_norm": 10.053447656212331, "learning_rate": 9.087970550179846e-06, "loss": 0.4182, "step": 3096 }, { "epoch": 0.2195675292449486, "grad_norm": 2.623740500761365, "learning_rate": 9.08730933820457e-06, "loss": 0.4768, "step": 3097 }, { "epoch": 0.219638426090039, "grad_norm": 5.796518979856382, "learning_rate": 9.086647910702353e-06, "loss": 0.4412, "step": 3098 }, { "epoch": 0.21970932293512938, "grad_norm": 1.6214521934922033, "learning_rate": 9.085986267708074e-06, "loss": 0.4007, "step": 3099 }, { "epoch": 0.21978021978021978, "grad_norm": 2.196630134763447, "learning_rate": 9.08532440925662e-06, "loss": 0.4181, "step": 3100 }, { "epoch": 0.21985111662531018, "grad_norm": 2.9193578012838293, "learning_rate": 9.08466233538289e-06, "loss": 0.4266, "step": 3101 }, { "epoch": 0.21992201347040058, "grad_norm": 3.229559410106871, "learning_rate": 9.0840000461218e-06, "loss": 0.4532, "step": 3102 }, { "epoch": 0.21999291031549095, "grad_norm": 3.7519786134489967, "learning_rate": 9.083337541508268e-06, "loss": 0.4269, "step": 3103 }, { "epoch": 0.22006380716058135, "grad_norm": 4.054174168464488, "learning_rate": 9.082674821577233e-06, "loss": 0.4179, "step": 3104 }, { "epoch": 0.22013470400567176, "grad_norm": 3.647554214886108, "learning_rate": 9.082011886363637e-06, "loss": 0.4024, "step": 3105 }, { "epoch": 0.22020560085076213, "grad_norm": 1.8462034529042863, "learning_rate": 9.08134873590244e-06, "loss": 0.4719, "step": 3106 }, { "epoch": 0.22027649769585253, "grad_norm": 2.995659048835128, "learning_rate": 9.080685370228608e-06, "loss": 0.5387, "step": 3107 }, { "epoch": 0.22034739454094293, "grad_norm": 3.6430854073993033, "learning_rate": 9.08002178937712e-06, "loss": 0.4455, "step": 3108 }, { "epoch": 0.22041829138603333, "grad_norm": 2.4693710873039696, "learning_rate": 9.079357993382972e-06, "loss": 0.45, "step": 3109 }, { "epoch": 0.2204891882311237, "grad_norm": 3.6956276088745854, "learning_rate": 9.078693982281162e-06, "loss": 0.4628, "step": 3110 }, { "epoch": 0.2205600850762141, "grad_norm": 2.1341461991655146, "learning_rate": 9.078029756106706e-06, "loss": 0.4613, "step": 3111 }, { "epoch": 0.2206309819213045, "grad_norm": 6.334670795166778, "learning_rate": 9.077365314894627e-06, "loss": 0.4184, "step": 3112 }, { "epoch": 0.22070187876639488, "grad_norm": 2.205072525292545, "learning_rate": 9.076700658679963e-06, "loss": 0.4829, "step": 3113 }, { "epoch": 0.22077277561148528, "grad_norm": 4.178679829287648, "learning_rate": 9.076035787497763e-06, "loss": 0.3925, "step": 3114 }, { "epoch": 0.22084367245657568, "grad_norm": 2.418511284813459, "learning_rate": 9.075370701383083e-06, "loss": 0.4009, "step": 3115 }, { "epoch": 0.22091456930166609, "grad_norm": 4.22424649417998, "learning_rate": 9.074705400370995e-06, "loss": 0.4488, "step": 3116 }, { "epoch": 0.22098546614675646, "grad_norm": 3.297870852739308, "learning_rate": 9.074039884496584e-06, "loss": 0.4323, "step": 3117 }, { "epoch": 0.22105636299184686, "grad_norm": 1.157144173281214, "learning_rate": 9.073374153794938e-06, "loss": 0.7135, "step": 3118 }, { "epoch": 0.22112725983693726, "grad_norm": 3.1819784558389563, "learning_rate": 9.072708208301164e-06, "loss": 0.4072, "step": 3119 }, { "epoch": 0.22119815668202766, "grad_norm": 5.1763722877193254, "learning_rate": 9.072042048050379e-06, "loss": 0.4557, "step": 3120 }, { "epoch": 0.22126905352711804, "grad_norm": 2.2966480947522303, "learning_rate": 9.071375673077709e-06, "loss": 0.4715, "step": 3121 }, { "epoch": 0.22133995037220844, "grad_norm": 3.0139980630332834, "learning_rate": 9.07070908341829e-06, "loss": 0.4885, "step": 3122 }, { "epoch": 0.22141084721729884, "grad_norm": 10.905009383527219, "learning_rate": 9.070042279107274e-06, "loss": 0.4952, "step": 3123 }, { "epoch": 0.2214817440623892, "grad_norm": 3.175742261761657, "learning_rate": 9.069375260179824e-06, "loss": 0.4239, "step": 3124 }, { "epoch": 0.22155264090747961, "grad_norm": 8.106052307703616, "learning_rate": 9.06870802667111e-06, "loss": 0.4678, "step": 3125 }, { "epoch": 0.22162353775257002, "grad_norm": 2.7573408800418586, "learning_rate": 9.068040578616315e-06, "loss": 0.4333, "step": 3126 }, { "epoch": 0.22169443459766042, "grad_norm": 9.714821007226634, "learning_rate": 9.067372916050638e-06, "loss": 0.4989, "step": 3127 }, { "epoch": 0.2217653314427508, "grad_norm": 3.3762196031873555, "learning_rate": 9.06670503900928e-06, "loss": 0.4514, "step": 3128 }, { "epoch": 0.2218362282878412, "grad_norm": 2.2804412203229285, "learning_rate": 9.066036947527463e-06, "loss": 0.4173, "step": 3129 }, { "epoch": 0.2219071251329316, "grad_norm": 3.3805691711384553, "learning_rate": 9.065368641640413e-06, "loss": 0.4051, "step": 3130 }, { "epoch": 0.22197802197802197, "grad_norm": 2.6682238015137423, "learning_rate": 9.064700121383372e-06, "loss": 0.4705, "step": 3131 }, { "epoch": 0.22204891882311237, "grad_norm": 2.656381204328177, "learning_rate": 9.064031386791591e-06, "loss": 0.4718, "step": 3132 }, { "epoch": 0.22211981566820277, "grad_norm": 6.318051220433762, "learning_rate": 9.063362437900335e-06, "loss": 0.4819, "step": 3133 }, { "epoch": 0.22219071251329317, "grad_norm": 2.81829354560376, "learning_rate": 9.062693274744874e-06, "loss": 0.4519, "step": 3134 }, { "epoch": 0.22226160935838354, "grad_norm": 9.717122230120053, "learning_rate": 9.062023897360494e-06, "loss": 0.4168, "step": 3135 }, { "epoch": 0.22233250620347395, "grad_norm": 2.343649552385114, "learning_rate": 9.061354305782497e-06, "loss": 0.3857, "step": 3136 }, { "epoch": 0.22240340304856435, "grad_norm": 1.0088983918929584, "learning_rate": 9.060684500046186e-06, "loss": 0.669, "step": 3137 }, { "epoch": 0.22247429989365472, "grad_norm": 3.2697205962506413, "learning_rate": 9.060014480186882e-06, "loss": 0.4284, "step": 3138 }, { "epoch": 0.22254519673874512, "grad_norm": 3.6467132809523903, "learning_rate": 9.059344246239917e-06, "loss": 0.4544, "step": 3139 }, { "epoch": 0.22261609358383552, "grad_norm": 2.5561977191655396, "learning_rate": 9.058673798240629e-06, "loss": 0.4178, "step": 3140 }, { "epoch": 0.22268699042892592, "grad_norm": 2.2848753609458177, "learning_rate": 9.058003136224376e-06, "loss": 0.3948, "step": 3141 }, { "epoch": 0.2227578872740163, "grad_norm": 3.1511902460624857, "learning_rate": 9.057332260226519e-06, "loss": 0.4478, "step": 3142 }, { "epoch": 0.2228287841191067, "grad_norm": 3.9981117779138935, "learning_rate": 9.056661170282436e-06, "loss": 0.4614, "step": 3143 }, { "epoch": 0.2228996809641971, "grad_norm": 26.585888779222508, "learning_rate": 9.055989866427512e-06, "loss": 0.507, "step": 3144 }, { "epoch": 0.22297057780928747, "grad_norm": 2.650123322401632, "learning_rate": 9.05531834869715e-06, "loss": 0.4512, "step": 3145 }, { "epoch": 0.22304147465437787, "grad_norm": 3.2916038075999294, "learning_rate": 9.05464661712675e-06, "loss": 0.4354, "step": 3146 }, { "epoch": 0.22311237149946828, "grad_norm": 2.3206649269297372, "learning_rate": 9.053974671751744e-06, "loss": 0.4407, "step": 3147 }, { "epoch": 0.22318326834455868, "grad_norm": 8.943113651837344, "learning_rate": 9.053302512607557e-06, "loss": 0.5091, "step": 3148 }, { "epoch": 0.22325416518964905, "grad_norm": 2.411738256509649, "learning_rate": 9.052630139729637e-06, "loss": 0.4612, "step": 3149 }, { "epoch": 0.22332506203473945, "grad_norm": 2.1247560175900375, "learning_rate": 9.051957553153434e-06, "loss": 0.4505, "step": 3150 }, { "epoch": 0.22339595887982985, "grad_norm": 1.5728592449833612, "learning_rate": 9.051284752914418e-06, "loss": 0.4364, "step": 3151 }, { "epoch": 0.22346685572492025, "grad_norm": 2.5129996734326276, "learning_rate": 9.050611739048065e-06, "loss": 0.4026, "step": 3152 }, { "epoch": 0.22353775257001063, "grad_norm": 2.578398886861346, "learning_rate": 9.049938511589862e-06, "loss": 0.4109, "step": 3153 }, { "epoch": 0.22360864941510103, "grad_norm": 2.7092446682294713, "learning_rate": 9.049265070575311e-06, "loss": 0.4628, "step": 3154 }, { "epoch": 0.22367954626019143, "grad_norm": 2.3851115529578197, "learning_rate": 9.048591416039923e-06, "loss": 0.4177, "step": 3155 }, { "epoch": 0.2237504431052818, "grad_norm": 3.0956966785280065, "learning_rate": 9.047917548019219e-06, "loss": 0.4357, "step": 3156 }, { "epoch": 0.2238213399503722, "grad_norm": 2.5111045741573412, "learning_rate": 9.047243466548732e-06, "loss": 0.3978, "step": 3157 }, { "epoch": 0.2238922367954626, "grad_norm": 3.307760833997562, "learning_rate": 9.046569171664008e-06, "loss": 0.4243, "step": 3158 }, { "epoch": 0.223963133640553, "grad_norm": 2.920377544786675, "learning_rate": 9.045894663400603e-06, "loss": 0.4797, "step": 3159 }, { "epoch": 0.22403403048564338, "grad_norm": 2.3862809489722343, "learning_rate": 9.045219941794084e-06, "loss": 0.4545, "step": 3160 }, { "epoch": 0.22410492733073378, "grad_norm": 2.037645493991718, "learning_rate": 9.044545006880031e-06, "loss": 0.4056, "step": 3161 }, { "epoch": 0.22417582417582418, "grad_norm": 6.879403784179113, "learning_rate": 9.043869858694031e-06, "loss": 0.4269, "step": 3162 }, { "epoch": 0.22424672102091456, "grad_norm": 2.3965073541146977, "learning_rate": 9.043194497271688e-06, "loss": 0.3908, "step": 3163 }, { "epoch": 0.22431761786600496, "grad_norm": 2.6544327309638556, "learning_rate": 9.042518922648615e-06, "loss": 0.4405, "step": 3164 }, { "epoch": 0.22438851471109536, "grad_norm": 2.508849339559146, "learning_rate": 9.041843134860432e-06, "loss": 0.454, "step": 3165 }, { "epoch": 0.22445941155618576, "grad_norm": 2.215544927728453, "learning_rate": 9.041167133942776e-06, "loss": 0.4447, "step": 3166 }, { "epoch": 0.22453030840127614, "grad_norm": 11.391958168766523, "learning_rate": 9.040490919931291e-06, "loss": 0.4844, "step": 3167 }, { "epoch": 0.22460120524636654, "grad_norm": 5.305959884783501, "learning_rate": 9.039814492861637e-06, "loss": 0.4416, "step": 3168 }, { "epoch": 0.22467210209145694, "grad_norm": 1.8674229731390561, "learning_rate": 9.039137852769482e-06, "loss": 0.4199, "step": 3169 }, { "epoch": 0.2247429989365473, "grad_norm": 1.8870604008313325, "learning_rate": 9.038460999690504e-06, "loss": 0.453, "step": 3170 }, { "epoch": 0.2248138957816377, "grad_norm": 2.2422943068867003, "learning_rate": 9.037783933660394e-06, "loss": 0.4129, "step": 3171 }, { "epoch": 0.2248847926267281, "grad_norm": 3.244116916688168, "learning_rate": 9.037106654714859e-06, "loss": 0.4544, "step": 3172 }, { "epoch": 0.22495568947181852, "grad_norm": 6.862704950940976, "learning_rate": 9.036429162889604e-06, "loss": 0.4538, "step": 3173 }, { "epoch": 0.2250265863169089, "grad_norm": 3.9120088286492, "learning_rate": 9.03575145822036e-06, "loss": 0.4144, "step": 3174 }, { "epoch": 0.2250974831619993, "grad_norm": 2.589288205944333, "learning_rate": 9.035073540742863e-06, "loss": 0.4437, "step": 3175 }, { "epoch": 0.2251683800070897, "grad_norm": 7.930194543989803, "learning_rate": 9.034395410492858e-06, "loss": 0.4743, "step": 3176 }, { "epoch": 0.22523927685218006, "grad_norm": 5.6458631671637445, "learning_rate": 9.033717067506103e-06, "loss": 0.4366, "step": 3177 }, { "epoch": 0.22531017369727047, "grad_norm": 6.557015751181291, "learning_rate": 9.033038511818368e-06, "loss": 0.4826, "step": 3178 }, { "epoch": 0.22538107054236087, "grad_norm": 3.820373725119855, "learning_rate": 9.032359743465434e-06, "loss": 0.478, "step": 3179 }, { "epoch": 0.22545196738745127, "grad_norm": 2.5824614985812033, "learning_rate": 9.031680762483093e-06, "loss": 0.4433, "step": 3180 }, { "epoch": 0.22552286423254164, "grad_norm": 1.8996662080181894, "learning_rate": 9.031001568907148e-06, "loss": 0.4194, "step": 3181 }, { "epoch": 0.22559376107763204, "grad_norm": 10.266975125354946, "learning_rate": 9.030322162773414e-06, "loss": 0.4231, "step": 3182 }, { "epoch": 0.22566465792272244, "grad_norm": 2.8987072896505253, "learning_rate": 9.029642544117716e-06, "loss": 0.3984, "step": 3183 }, { "epoch": 0.22573555476781285, "grad_norm": 5.889276325271632, "learning_rate": 9.028962712975891e-06, "loss": 0.4398, "step": 3184 }, { "epoch": 0.22580645161290322, "grad_norm": 2.3590876734574837, "learning_rate": 9.028282669383787e-06, "loss": 0.4661, "step": 3185 }, { "epoch": 0.22587734845799362, "grad_norm": 3.8368258560039683, "learning_rate": 9.027602413377262e-06, "loss": 0.3939, "step": 3186 }, { "epoch": 0.22594824530308402, "grad_norm": 1.9499604947096831, "learning_rate": 9.02692194499219e-06, "loss": 0.4744, "step": 3187 }, { "epoch": 0.2260191421481744, "grad_norm": 2.2597660027548496, "learning_rate": 9.02624126426445e-06, "loss": 0.4823, "step": 3188 }, { "epoch": 0.2260900389932648, "grad_norm": 12.640357410307217, "learning_rate": 9.025560371229934e-06, "loss": 0.4094, "step": 3189 }, { "epoch": 0.2261609358383552, "grad_norm": 1.9296439438189528, "learning_rate": 9.024879265924547e-06, "loss": 0.4021, "step": 3190 }, { "epoch": 0.2262318326834456, "grad_norm": 2.2263247330410425, "learning_rate": 9.024197948384203e-06, "loss": 0.4337, "step": 3191 }, { "epoch": 0.22630272952853597, "grad_norm": 2.2737468220970105, "learning_rate": 9.023516418644831e-06, "loss": 0.4446, "step": 3192 }, { "epoch": 0.22637362637362637, "grad_norm": 3.7665115304435903, "learning_rate": 9.022834676742368e-06, "loss": 0.4538, "step": 3193 }, { "epoch": 0.22644452321871678, "grad_norm": 3.8030347983862196, "learning_rate": 9.02215272271276e-06, "loss": 0.4259, "step": 3194 }, { "epoch": 0.22651542006380715, "grad_norm": 2.4257845338402624, "learning_rate": 9.021470556591971e-06, "loss": 0.4478, "step": 3195 }, { "epoch": 0.22658631690889755, "grad_norm": 1.8425821574939922, "learning_rate": 9.02078817841597e-06, "loss": 0.4736, "step": 3196 }, { "epoch": 0.22665721375398795, "grad_norm": 2.0727365299127807, "learning_rate": 9.020105588220738e-06, "loss": 0.473, "step": 3197 }, { "epoch": 0.22672811059907835, "grad_norm": 3.8368215744236513, "learning_rate": 9.019422786042271e-06, "loss": 0.4256, "step": 3198 }, { "epoch": 0.22679900744416873, "grad_norm": 2.2748643358878207, "learning_rate": 9.018739771916572e-06, "loss": 0.4115, "step": 3199 }, { "epoch": 0.22686990428925913, "grad_norm": 3.6373349067524754, "learning_rate": 9.018056545879658e-06, "loss": 0.424, "step": 3200 }, { "epoch": 0.22694080113434953, "grad_norm": 9.281277830930126, "learning_rate": 9.017373107967555e-06, "loss": 0.4876, "step": 3201 }, { "epoch": 0.2270116979794399, "grad_norm": 6.796504065543167, "learning_rate": 9.016689458216303e-06, "loss": 0.4282, "step": 3202 }, { "epoch": 0.2270825948245303, "grad_norm": 3.1247709820695513, "learning_rate": 9.016005596661948e-06, "loss": 0.4603, "step": 3203 }, { "epoch": 0.2271534916696207, "grad_norm": 3.913955043456296, "learning_rate": 9.015321523340554e-06, "loss": 0.4438, "step": 3204 }, { "epoch": 0.2272243885147111, "grad_norm": 3.864345422500888, "learning_rate": 9.014637238288191e-06, "loss": 0.4425, "step": 3205 }, { "epoch": 0.22729528535980148, "grad_norm": 2.0870029431645065, "learning_rate": 9.01395274154094e-06, "loss": 0.4462, "step": 3206 }, { "epoch": 0.22736618220489188, "grad_norm": 4.500765708814183, "learning_rate": 9.0132680331349e-06, "loss": 0.4481, "step": 3207 }, { "epoch": 0.22743707904998228, "grad_norm": 2.0199768608968722, "learning_rate": 9.012583113106174e-06, "loss": 0.4093, "step": 3208 }, { "epoch": 0.22750797589507266, "grad_norm": 3.7537592305394583, "learning_rate": 9.011897981490875e-06, "loss": 0.4425, "step": 3209 }, { "epoch": 0.22757887274016306, "grad_norm": 2.600650288006387, "learning_rate": 9.011212638325136e-06, "loss": 0.4868, "step": 3210 }, { "epoch": 0.22764976958525346, "grad_norm": 2.954502915385285, "learning_rate": 9.010527083645091e-06, "loss": 0.5104, "step": 3211 }, { "epoch": 0.22772066643034386, "grad_norm": 2.307640795099099, "learning_rate": 9.009841317486894e-06, "loss": 0.4307, "step": 3212 }, { "epoch": 0.22779156327543423, "grad_norm": 3.6209966580252213, "learning_rate": 9.009155339886702e-06, "loss": 0.3844, "step": 3213 }, { "epoch": 0.22786246012052463, "grad_norm": 2.5498530846170513, "learning_rate": 9.00846915088069e-06, "loss": 0.4844, "step": 3214 }, { "epoch": 0.22793335696561504, "grad_norm": 5.553847965113211, "learning_rate": 9.007782750505038e-06, "loss": 0.433, "step": 3215 }, { "epoch": 0.22800425381070544, "grad_norm": 2.56461744888354, "learning_rate": 9.007096138795945e-06, "loss": 0.386, "step": 3216 }, { "epoch": 0.2280751506557958, "grad_norm": 2.747146055162643, "learning_rate": 9.006409315789612e-06, "loss": 0.4519, "step": 3217 }, { "epoch": 0.2281460475008862, "grad_norm": 2.393883033642813, "learning_rate": 9.00572228152226e-06, "loss": 0.5183, "step": 3218 }, { "epoch": 0.2282169443459766, "grad_norm": 3.57212867907339, "learning_rate": 9.005035036030114e-06, "loss": 0.3889, "step": 3219 }, { "epoch": 0.228287841191067, "grad_norm": 3.5803270288318476, "learning_rate": 9.004347579349412e-06, "loss": 0.4198, "step": 3220 }, { "epoch": 0.2283587380361574, "grad_norm": 3.210843202972818, "learning_rate": 9.003659911516406e-06, "loss": 0.4542, "step": 3221 }, { "epoch": 0.2284296348812478, "grad_norm": 4.310718709206904, "learning_rate": 9.002972032567358e-06, "loss": 0.4554, "step": 3222 }, { "epoch": 0.2285005317263382, "grad_norm": 3.081483353419941, "learning_rate": 9.00228394253854e-06, "loss": 0.4268, "step": 3223 }, { "epoch": 0.22857142857142856, "grad_norm": 2.709595083966468, "learning_rate": 9.001595641466235e-06, "loss": 0.4266, "step": 3224 }, { "epoch": 0.22864232541651897, "grad_norm": 2.044766916748749, "learning_rate": 9.000907129386735e-06, "loss": 0.4299, "step": 3225 }, { "epoch": 0.22871322226160937, "grad_norm": 4.038502289957011, "learning_rate": 9.000218406336352e-06, "loss": 0.4242, "step": 3226 }, { "epoch": 0.22878411910669974, "grad_norm": 2.777251831138083, "learning_rate": 8.999529472351398e-06, "loss": 0.4366, "step": 3227 }, { "epoch": 0.22885501595179014, "grad_norm": 2.852101045982972, "learning_rate": 8.9988403274682e-06, "loss": 0.5011, "step": 3228 }, { "epoch": 0.22892591279688054, "grad_norm": 1.5075756236319393, "learning_rate": 8.9981509717231e-06, "loss": 0.3776, "step": 3229 }, { "epoch": 0.22899680964197094, "grad_norm": 3.1016090056037737, "learning_rate": 8.99746140515245e-06, "loss": 0.4477, "step": 3230 }, { "epoch": 0.22906770648706132, "grad_norm": 2.9510032690942642, "learning_rate": 8.996771627792607e-06, "loss": 0.4562, "step": 3231 }, { "epoch": 0.22913860333215172, "grad_norm": 3.6729533627666378, "learning_rate": 8.996081639679946e-06, "loss": 0.4773, "step": 3232 }, { "epoch": 0.22920950017724212, "grad_norm": 7.754592722089802, "learning_rate": 8.99539144085085e-06, "loss": 0.3523, "step": 3233 }, { "epoch": 0.2292803970223325, "grad_norm": 3.018735744499898, "learning_rate": 8.994701031341715e-06, "loss": 0.4117, "step": 3234 }, { "epoch": 0.2293512938674229, "grad_norm": 3.055228736558539, "learning_rate": 8.994010411188944e-06, "loss": 0.4158, "step": 3235 }, { "epoch": 0.2294221907125133, "grad_norm": 4.7763967247378245, "learning_rate": 8.993319580428956e-06, "loss": 0.4461, "step": 3236 }, { "epoch": 0.2294930875576037, "grad_norm": 2.2007786979443296, "learning_rate": 8.992628539098177e-06, "loss": 0.3751, "step": 3237 }, { "epoch": 0.22956398440269407, "grad_norm": 2.7068382333828094, "learning_rate": 8.99193728723305e-06, "loss": 0.4111, "step": 3238 }, { "epoch": 0.22963488124778447, "grad_norm": 3.007498112768853, "learning_rate": 8.991245824870023e-06, "loss": 0.416, "step": 3239 }, { "epoch": 0.22970577809287487, "grad_norm": 2.611684328782293, "learning_rate": 8.990554152045556e-06, "loss": 0.4493, "step": 3240 }, { "epoch": 0.22977667493796525, "grad_norm": 3.8183922117373745, "learning_rate": 8.989862268796124e-06, "loss": 0.4053, "step": 3241 }, { "epoch": 0.22984757178305565, "grad_norm": 5.135612835090105, "learning_rate": 8.989170175158207e-06, "loss": 0.4483, "step": 3242 }, { "epoch": 0.22991846862814605, "grad_norm": 7.198771365415722, "learning_rate": 8.988477871168305e-06, "loss": 0.4877, "step": 3243 }, { "epoch": 0.22998936547323645, "grad_norm": 3.0647989193699314, "learning_rate": 8.987785356862917e-06, "loss": 0.5323, "step": 3244 }, { "epoch": 0.23006026231832682, "grad_norm": 2.039769876357869, "learning_rate": 8.987092632278566e-06, "loss": 0.4727, "step": 3245 }, { "epoch": 0.23013115916341723, "grad_norm": 5.106916416439665, "learning_rate": 8.986399697451775e-06, "loss": 0.4401, "step": 3246 }, { "epoch": 0.23020205600850763, "grad_norm": 4.125751484143472, "learning_rate": 8.985706552419088e-06, "loss": 0.5109, "step": 3247 }, { "epoch": 0.23027295285359803, "grad_norm": 2.309389470923157, "learning_rate": 8.985013197217051e-06, "loss": 0.4563, "step": 3248 }, { "epoch": 0.2303438496986884, "grad_norm": 1.9461348895144452, "learning_rate": 8.984319631882225e-06, "loss": 0.4533, "step": 3249 }, { "epoch": 0.2304147465437788, "grad_norm": 2.64090085058567, "learning_rate": 8.983625856451186e-06, "loss": 0.4833, "step": 3250 }, { "epoch": 0.2304856433888692, "grad_norm": 5.520314656239803, "learning_rate": 8.982931870960515e-06, "loss": 0.4299, "step": 3251 }, { "epoch": 0.23055654023395958, "grad_norm": 1.1649313228407496, "learning_rate": 8.982237675446806e-06, "loss": 0.6385, "step": 3252 }, { "epoch": 0.23062743707904998, "grad_norm": 2.0362393569147357, "learning_rate": 8.981543269946664e-06, "loss": 0.4561, "step": 3253 }, { "epoch": 0.23069833392414038, "grad_norm": 1.6543213458895498, "learning_rate": 8.980848654496707e-06, "loss": 0.4317, "step": 3254 }, { "epoch": 0.23076923076923078, "grad_norm": 2.095943068475245, "learning_rate": 8.980153829133563e-06, "loss": 0.4061, "step": 3255 }, { "epoch": 0.23084012761432116, "grad_norm": 2.859947345242407, "learning_rate": 8.979458793893869e-06, "loss": 0.4813, "step": 3256 }, { "epoch": 0.23091102445941156, "grad_norm": 2.8538504161005966, "learning_rate": 8.978763548814275e-06, "loss": 0.418, "step": 3257 }, { "epoch": 0.23098192130450196, "grad_norm": 1.5255251860723613, "learning_rate": 8.978068093931443e-06, "loss": 0.4587, "step": 3258 }, { "epoch": 0.23105281814959233, "grad_norm": 1.5380689322022978, "learning_rate": 8.977372429282044e-06, "loss": 0.447, "step": 3259 }, { "epoch": 0.23112371499468273, "grad_norm": 3.287711161876007, "learning_rate": 8.976676554902759e-06, "loss": 0.4547, "step": 3260 }, { "epoch": 0.23119461183977313, "grad_norm": 1.9116922785286898, "learning_rate": 8.975980470830286e-06, "loss": 0.4322, "step": 3261 }, { "epoch": 0.23126550868486354, "grad_norm": 3.739875474462698, "learning_rate": 8.975284177101326e-06, "loss": 0.4769, "step": 3262 }, { "epoch": 0.2313364055299539, "grad_norm": 2.1234308372776005, "learning_rate": 8.9745876737526e-06, "loss": 0.4532, "step": 3263 }, { "epoch": 0.2314073023750443, "grad_norm": 3.0535289737635916, "learning_rate": 8.97389096082083e-06, "loss": 0.3724, "step": 3264 }, { "epoch": 0.2314781992201347, "grad_norm": 2.2261098823371137, "learning_rate": 8.973194038342757e-06, "loss": 0.4574, "step": 3265 }, { "epoch": 0.23154909606522509, "grad_norm": 1.8533437969542952, "learning_rate": 8.97249690635513e-06, "loss": 0.3979, "step": 3266 }, { "epoch": 0.2316199929103155, "grad_norm": 2.2140494353321016, "learning_rate": 8.97179956489471e-06, "loss": 0.4024, "step": 3267 }, { "epoch": 0.2316908897554059, "grad_norm": 3.2926356460717123, "learning_rate": 8.971102013998268e-06, "loss": 0.4502, "step": 3268 }, { "epoch": 0.2317617866004963, "grad_norm": 1.094728821011881, "learning_rate": 8.970404253702582e-06, "loss": 0.6342, "step": 3269 }, { "epoch": 0.23183268344558666, "grad_norm": 3.704690568061812, "learning_rate": 8.969706284044453e-06, "loss": 0.4449, "step": 3270 }, { "epoch": 0.23190358029067706, "grad_norm": 3.1015769210432054, "learning_rate": 8.96900810506068e-06, "loss": 0.4303, "step": 3271 }, { "epoch": 0.23197447713576747, "grad_norm": 4.493730611291958, "learning_rate": 8.96830971678808e-06, "loss": 0.486, "step": 3272 }, { "epoch": 0.23204537398085784, "grad_norm": 3.425706642369872, "learning_rate": 8.967611119263481e-06, "loss": 0.4007, "step": 3273 }, { "epoch": 0.23211627082594824, "grad_norm": 2.517916490559249, "learning_rate": 8.966912312523719e-06, "loss": 0.3985, "step": 3274 }, { "epoch": 0.23218716767103864, "grad_norm": 2.7506918319182696, "learning_rate": 8.966213296605642e-06, "loss": 0.4369, "step": 3275 }, { "epoch": 0.23225806451612904, "grad_norm": 2.0382374291026433, "learning_rate": 8.965514071546113e-06, "loss": 0.4662, "step": 3276 }, { "epoch": 0.23232896136121942, "grad_norm": 2.4496659521979693, "learning_rate": 8.964814637381997e-06, "loss": 0.4921, "step": 3277 }, { "epoch": 0.23239985820630982, "grad_norm": 2.454854968234827, "learning_rate": 8.964114994150181e-06, "loss": 0.4043, "step": 3278 }, { "epoch": 0.23247075505140022, "grad_norm": 2.416911656434514, "learning_rate": 8.963415141887555e-06, "loss": 0.4512, "step": 3279 }, { "epoch": 0.23254165189649062, "grad_norm": 2.190508363083441, "learning_rate": 8.962715080631024e-06, "loss": 0.4477, "step": 3280 }, { "epoch": 0.232612548741581, "grad_norm": 2.4443214703492604, "learning_rate": 8.9620148104175e-06, "loss": 0.4311, "step": 3281 }, { "epoch": 0.2326834455866714, "grad_norm": 3.9295108585678804, "learning_rate": 8.961314331283915e-06, "loss": 0.4562, "step": 3282 }, { "epoch": 0.2327543424317618, "grad_norm": 2.7757199389702194, "learning_rate": 8.960613643267199e-06, "loss": 0.4589, "step": 3283 }, { "epoch": 0.23282523927685217, "grad_norm": 1.914616450594456, "learning_rate": 8.959912746404303e-06, "loss": 0.4158, "step": 3284 }, { "epoch": 0.23289613612194257, "grad_norm": 3.8370243813481055, "learning_rate": 8.959211640732187e-06, "loss": 0.4793, "step": 3285 }, { "epoch": 0.23296703296703297, "grad_norm": 1.147183366830313, "learning_rate": 8.958510326287816e-06, "loss": 0.6814, "step": 3286 }, { "epoch": 0.23303792981212337, "grad_norm": 2.8225781244816814, "learning_rate": 8.957808803108176e-06, "loss": 0.4294, "step": 3287 }, { "epoch": 0.23310882665721375, "grad_norm": 2.969673547069707, "learning_rate": 8.957107071230258e-06, "loss": 0.4351, "step": 3288 }, { "epoch": 0.23317972350230415, "grad_norm": 1.999678830313094, "learning_rate": 8.956405130691061e-06, "loss": 0.4037, "step": 3289 }, { "epoch": 0.23325062034739455, "grad_norm": 3.0792988187361674, "learning_rate": 8.955702981527603e-06, "loss": 0.4065, "step": 3290 }, { "epoch": 0.23332151719248492, "grad_norm": 2.737315539130356, "learning_rate": 8.955000623776909e-06, "loss": 0.4392, "step": 3291 }, { "epoch": 0.23339241403757532, "grad_norm": 3.415533933885833, "learning_rate": 8.954298057476011e-06, "loss": 0.4206, "step": 3292 }, { "epoch": 0.23346331088266573, "grad_norm": 2.9018697878627804, "learning_rate": 8.95359528266196e-06, "loss": 0.3957, "step": 3293 }, { "epoch": 0.23353420772775613, "grad_norm": 1.8409014948409839, "learning_rate": 8.952892299371813e-06, "loss": 0.4539, "step": 3294 }, { "epoch": 0.2336051045728465, "grad_norm": 1.7138531588654726, "learning_rate": 8.952189107642636e-06, "loss": 0.4255, "step": 3295 }, { "epoch": 0.2336760014179369, "grad_norm": 3.509893491083699, "learning_rate": 8.951485707511511e-06, "loss": 0.3858, "step": 3296 }, { "epoch": 0.2337468982630273, "grad_norm": 3.0635781190650273, "learning_rate": 8.950782099015529e-06, "loss": 0.4371, "step": 3297 }, { "epoch": 0.23381779510811768, "grad_norm": 2.0172088740177925, "learning_rate": 8.950078282191792e-06, "loss": 0.4402, "step": 3298 }, { "epoch": 0.23388869195320808, "grad_norm": 2.39217287968822, "learning_rate": 8.949374257077412e-06, "loss": 0.481, "step": 3299 }, { "epoch": 0.23395958879829848, "grad_norm": 2.9999862681205958, "learning_rate": 8.948670023709512e-06, "loss": 0.4214, "step": 3300 }, { "epoch": 0.23403048564338888, "grad_norm": 1.7887464772861381, "learning_rate": 8.947965582125228e-06, "loss": 0.4612, "step": 3301 }, { "epoch": 0.23410138248847925, "grad_norm": 5.724067700262907, "learning_rate": 8.947260932361705e-06, "loss": 0.5064, "step": 3302 }, { "epoch": 0.23417227933356966, "grad_norm": 2.4007812361749825, "learning_rate": 8.946556074456103e-06, "loss": 0.4497, "step": 3303 }, { "epoch": 0.23424317617866006, "grad_norm": 2.654198266866823, "learning_rate": 8.945851008445585e-06, "loss": 0.4414, "step": 3304 }, { "epoch": 0.23431407302375043, "grad_norm": 2.9181617162969795, "learning_rate": 8.945145734367331e-06, "loss": 0.4663, "step": 3305 }, { "epoch": 0.23438496986884083, "grad_norm": 3.07076453227066, "learning_rate": 8.944440252258532e-06, "loss": 0.4506, "step": 3306 }, { "epoch": 0.23445586671393123, "grad_norm": 2.5574010204282795, "learning_rate": 8.943734562156387e-06, "loss": 0.4623, "step": 3307 }, { "epoch": 0.23452676355902163, "grad_norm": 2.4908548189379953, "learning_rate": 8.94302866409811e-06, "loss": 0.4344, "step": 3308 }, { "epoch": 0.234597660404112, "grad_norm": 2.3580855740185, "learning_rate": 8.942322558120921e-06, "loss": 0.5091, "step": 3309 }, { "epoch": 0.2346685572492024, "grad_norm": 2.6166013991699733, "learning_rate": 8.941616244262055e-06, "loss": 0.3891, "step": 3310 }, { "epoch": 0.2347394540942928, "grad_norm": 1.5197247788207688, "learning_rate": 8.940909722558754e-06, "loss": 0.441, "step": 3311 }, { "epoch": 0.2348103509393832, "grad_norm": 3.316020141325918, "learning_rate": 8.940202993048278e-06, "loss": 0.4293, "step": 3312 }, { "epoch": 0.23488124778447358, "grad_norm": 4.097213994614725, "learning_rate": 8.93949605576789e-06, "loss": 0.456, "step": 3313 }, { "epoch": 0.23495214462956399, "grad_norm": 1.806548018156555, "learning_rate": 8.938788910754867e-06, "loss": 0.4165, "step": 3314 }, { "epoch": 0.2350230414746544, "grad_norm": 2.425605864974392, "learning_rate": 8.9380815580465e-06, "loss": 0.4605, "step": 3315 }, { "epoch": 0.23509393831974476, "grad_norm": 3.1828439124066947, "learning_rate": 8.937373997680086e-06, "loss": 0.4486, "step": 3316 }, { "epoch": 0.23516483516483516, "grad_norm": 1.150351539798644, "learning_rate": 8.936666229692935e-06, "loss": 0.7457, "step": 3317 }, { "epoch": 0.23523573200992556, "grad_norm": 3.3892205295202515, "learning_rate": 8.93595825412237e-06, "loss": 0.4456, "step": 3318 }, { "epoch": 0.23530662885501596, "grad_norm": 3.0274588226323824, "learning_rate": 8.93525007100572e-06, "loss": 0.4261, "step": 3319 }, { "epoch": 0.23537752570010634, "grad_norm": 2.9119328793410513, "learning_rate": 8.934541680380332e-06, "loss": 0.4102, "step": 3320 }, { "epoch": 0.23544842254519674, "grad_norm": 4.7799427130843455, "learning_rate": 8.93383308228356e-06, "loss": 0.4694, "step": 3321 }, { "epoch": 0.23551931939028714, "grad_norm": 1.6720682845538934, "learning_rate": 8.933124276752765e-06, "loss": 0.3826, "step": 3322 }, { "epoch": 0.23559021623537751, "grad_norm": 2.4381403175939695, "learning_rate": 8.932415263825325e-06, "loss": 0.471, "step": 3323 }, { "epoch": 0.23566111308046792, "grad_norm": 2.9482486689445175, "learning_rate": 8.931706043538626e-06, "loss": 0.4507, "step": 3324 }, { "epoch": 0.23573200992555832, "grad_norm": 2.299587142140729, "learning_rate": 8.930996615930066e-06, "loss": 0.4055, "step": 3325 }, { "epoch": 0.23580290677064872, "grad_norm": 12.125718556273778, "learning_rate": 8.930286981037056e-06, "loss": 0.446, "step": 3326 }, { "epoch": 0.2358738036157391, "grad_norm": 4.479625718319469, "learning_rate": 8.929577138897012e-06, "loss": 0.4588, "step": 3327 }, { "epoch": 0.2359447004608295, "grad_norm": 1.7506560186258822, "learning_rate": 8.928867089547366e-06, "loss": 0.4457, "step": 3328 }, { "epoch": 0.2360155973059199, "grad_norm": 3.238518905574864, "learning_rate": 8.92815683302556e-06, "loss": 0.4878, "step": 3329 }, { "epoch": 0.23608649415101027, "grad_norm": 2.7383646993991366, "learning_rate": 8.927446369369047e-06, "loss": 0.3833, "step": 3330 }, { "epoch": 0.23615739099610067, "grad_norm": 2.2902247587815427, "learning_rate": 8.926735698615288e-06, "loss": 0.45, "step": 3331 }, { "epoch": 0.23622828784119107, "grad_norm": 2.272495523479487, "learning_rate": 8.92602482080176e-06, "loss": 0.4525, "step": 3332 }, { "epoch": 0.23629918468628147, "grad_norm": 2.2204453006984664, "learning_rate": 8.925313735965946e-06, "loss": 0.4034, "step": 3333 }, { "epoch": 0.23637008153137185, "grad_norm": 2.057602552745968, "learning_rate": 8.924602444145342e-06, "loss": 0.4118, "step": 3334 }, { "epoch": 0.23644097837646225, "grad_norm": 2.210124345706488, "learning_rate": 8.923890945377456e-06, "loss": 0.3939, "step": 3335 }, { "epoch": 0.23651187522155265, "grad_norm": 1.8502959874509874, "learning_rate": 8.923179239699804e-06, "loss": 0.4064, "step": 3336 }, { "epoch": 0.23658277206664302, "grad_norm": 2.6824568152760055, "learning_rate": 8.92246732714992e-06, "loss": 0.463, "step": 3337 }, { "epoch": 0.23665366891173342, "grad_norm": 2.2376842329046864, "learning_rate": 8.921755207765335e-06, "loss": 0.4653, "step": 3338 }, { "epoch": 0.23672456575682382, "grad_norm": 3.5540633245391726, "learning_rate": 8.921042881583609e-06, "loss": 0.4712, "step": 3339 }, { "epoch": 0.23679546260191423, "grad_norm": 2.288947577402525, "learning_rate": 8.920330348642297e-06, "loss": 0.4846, "step": 3340 }, { "epoch": 0.2368663594470046, "grad_norm": 3.173075638531163, "learning_rate": 8.919617608978972e-06, "loss": 0.4784, "step": 3341 }, { "epoch": 0.236937256292095, "grad_norm": 2.8273253985499895, "learning_rate": 8.91890466263122e-06, "loss": 0.3903, "step": 3342 }, { "epoch": 0.2370081531371854, "grad_norm": 2.8028186448442347, "learning_rate": 8.918191509636635e-06, "loss": 0.4668, "step": 3343 }, { "epoch": 0.2370790499822758, "grad_norm": 7.432943536392129, "learning_rate": 8.91747815003282e-06, "loss": 0.4511, "step": 3344 }, { "epoch": 0.23714994682736618, "grad_norm": 2.402516685512204, "learning_rate": 8.916764583857394e-06, "loss": 0.4249, "step": 3345 }, { "epoch": 0.23722084367245658, "grad_norm": 4.6217803533300605, "learning_rate": 8.91605081114798e-06, "loss": 0.4288, "step": 3346 }, { "epoch": 0.23729174051754698, "grad_norm": 6.063110575124011, "learning_rate": 8.915336831942218e-06, "loss": 0.4019, "step": 3347 }, { "epoch": 0.23736263736263735, "grad_norm": 2.950395682245944, "learning_rate": 8.914622646277758e-06, "loss": 0.3905, "step": 3348 }, { "epoch": 0.23743353420772775, "grad_norm": 2.733211545744963, "learning_rate": 8.913908254192254e-06, "loss": 0.4273, "step": 3349 }, { "epoch": 0.23750443105281815, "grad_norm": 2.773705675876374, "learning_rate": 8.913193655723384e-06, "loss": 0.4087, "step": 3350 }, { "epoch": 0.23757532789790856, "grad_norm": 4.429586444016617, "learning_rate": 8.912478850908826e-06, "loss": 0.4359, "step": 3351 }, { "epoch": 0.23764622474299893, "grad_norm": 3.6572651648891994, "learning_rate": 8.911763839786269e-06, "loss": 0.5131, "step": 3352 }, { "epoch": 0.23771712158808933, "grad_norm": 2.812580214365729, "learning_rate": 8.91104862239342e-06, "loss": 0.4388, "step": 3353 }, { "epoch": 0.23778801843317973, "grad_norm": 3.7863768174431383, "learning_rate": 8.910333198767994e-06, "loss": 0.4637, "step": 3354 }, { "epoch": 0.2378589152782701, "grad_norm": 4.685665615533194, "learning_rate": 8.909617568947713e-06, "loss": 0.4536, "step": 3355 }, { "epoch": 0.2379298121233605, "grad_norm": 2.23108560474174, "learning_rate": 8.908901732970311e-06, "loss": 0.4713, "step": 3356 }, { "epoch": 0.2380007089684509, "grad_norm": 1.9326625474904826, "learning_rate": 8.908185690873541e-06, "loss": 0.4056, "step": 3357 }, { "epoch": 0.2380716058135413, "grad_norm": 1.956488462164062, "learning_rate": 8.907469442695154e-06, "loss": 0.48, "step": 3358 }, { "epoch": 0.23814250265863168, "grad_norm": 5.913952452169974, "learning_rate": 8.906752988472922e-06, "loss": 0.4301, "step": 3359 }, { "epoch": 0.23821339950372208, "grad_norm": 2.8594897540286333, "learning_rate": 8.906036328244623e-06, "loss": 0.4905, "step": 3360 }, { "epoch": 0.23828429634881249, "grad_norm": 3.6150899269376553, "learning_rate": 8.905319462048046e-06, "loss": 0.409, "step": 3361 }, { "epoch": 0.23835519319390286, "grad_norm": 2.7015075823015655, "learning_rate": 8.904602389920994e-06, "loss": 0.4283, "step": 3362 }, { "epoch": 0.23842609003899326, "grad_norm": 5.651535861789076, "learning_rate": 8.903885111901276e-06, "loss": 0.4087, "step": 3363 }, { "epoch": 0.23849698688408366, "grad_norm": 7.494369747295909, "learning_rate": 8.903167628026718e-06, "loss": 0.4307, "step": 3364 }, { "epoch": 0.23856788372917406, "grad_norm": 2.2843245855128327, "learning_rate": 8.90244993833515e-06, "loss": 0.4553, "step": 3365 }, { "epoch": 0.23863878057426444, "grad_norm": 2.420975636401514, "learning_rate": 8.901732042864421e-06, "loss": 0.4344, "step": 3366 }, { "epoch": 0.23870967741935484, "grad_norm": 9.490241861909492, "learning_rate": 8.901013941652383e-06, "loss": 0.421, "step": 3367 }, { "epoch": 0.23878057426444524, "grad_norm": 2.6188463383090057, "learning_rate": 8.9002956347369e-06, "loss": 0.424, "step": 3368 }, { "epoch": 0.2388514711095356, "grad_norm": 4.076606162523838, "learning_rate": 8.899577122155852e-06, "loss": 0.4069, "step": 3369 }, { "epoch": 0.238922367954626, "grad_norm": 4.876295569582291, "learning_rate": 8.898858403947126e-06, "loss": 0.4442, "step": 3370 }, { "epoch": 0.23899326479971642, "grad_norm": 2.586469706151139, "learning_rate": 8.898139480148623e-06, "loss": 0.4608, "step": 3371 }, { "epoch": 0.23906416164480682, "grad_norm": 3.452601604355121, "learning_rate": 8.897420350798246e-06, "loss": 0.4681, "step": 3372 }, { "epoch": 0.2391350584898972, "grad_norm": 3.103535945769528, "learning_rate": 8.896701015933921e-06, "loss": 0.4364, "step": 3373 }, { "epoch": 0.2392059553349876, "grad_norm": 1.853377845728422, "learning_rate": 8.895981475593578e-06, "loss": 0.3893, "step": 3374 }, { "epoch": 0.239276852180078, "grad_norm": 3.6027655784020025, "learning_rate": 8.895261729815157e-06, "loss": 0.4793, "step": 3375 }, { "epoch": 0.2393477490251684, "grad_norm": 2.947551692554008, "learning_rate": 8.894541778636612e-06, "loss": 0.4497, "step": 3376 }, { "epoch": 0.23941864587025877, "grad_norm": 2.4443647653515685, "learning_rate": 8.893821622095906e-06, "loss": 0.4173, "step": 3377 }, { "epoch": 0.23948954271534917, "grad_norm": 2.320017358437325, "learning_rate": 8.893101260231016e-06, "loss": 0.4148, "step": 3378 }, { "epoch": 0.23956043956043957, "grad_norm": 2.618007606448261, "learning_rate": 8.892380693079923e-06, "loss": 0.4422, "step": 3379 }, { "epoch": 0.23963133640552994, "grad_norm": 6.223462286920893, "learning_rate": 8.891659920680625e-06, "loss": 0.4128, "step": 3380 }, { "epoch": 0.23970223325062034, "grad_norm": 2.866150673950492, "learning_rate": 8.89093894307113e-06, "loss": 0.4491, "step": 3381 }, { "epoch": 0.23977313009571075, "grad_norm": 1.7461417800749577, "learning_rate": 8.890217760289455e-06, "loss": 0.4575, "step": 3382 }, { "epoch": 0.23984402694080115, "grad_norm": 2.6979782121721065, "learning_rate": 8.889496372373626e-06, "loss": 0.4623, "step": 3383 }, { "epoch": 0.23991492378589152, "grad_norm": 4.3413850977701, "learning_rate": 8.888774779361687e-06, "loss": 0.4301, "step": 3384 }, { "epoch": 0.23998582063098192, "grad_norm": 2.006368971963676, "learning_rate": 8.888052981291685e-06, "loss": 0.3966, "step": 3385 }, { "epoch": 0.24005671747607232, "grad_norm": 1.719728144022738, "learning_rate": 8.887330978201682e-06, "loss": 0.3865, "step": 3386 }, { "epoch": 0.2401276143211627, "grad_norm": 4.606839126238176, "learning_rate": 8.88660877012975e-06, "loss": 0.4087, "step": 3387 }, { "epoch": 0.2401985111662531, "grad_norm": 2.448047058318288, "learning_rate": 8.885886357113969e-06, "loss": 0.4127, "step": 3388 }, { "epoch": 0.2402694080113435, "grad_norm": 3.825059911481931, "learning_rate": 8.885163739192436e-06, "loss": 0.4505, "step": 3389 }, { "epoch": 0.2403403048564339, "grad_norm": 3.027506747598425, "learning_rate": 8.884440916403252e-06, "loss": 0.431, "step": 3390 }, { "epoch": 0.24041120170152427, "grad_norm": 2.5724458874226226, "learning_rate": 8.883717888784535e-06, "loss": 0.4196, "step": 3391 }, { "epoch": 0.24048209854661468, "grad_norm": 2.8441696931670912, "learning_rate": 8.88299465637441e-06, "loss": 0.4352, "step": 3392 }, { "epoch": 0.24055299539170508, "grad_norm": 1.5605817262552744, "learning_rate": 8.882271219211012e-06, "loss": 0.418, "step": 3393 }, { "epoch": 0.24062389223679545, "grad_norm": 1.6850425391836976, "learning_rate": 8.881547577332488e-06, "loss": 0.4206, "step": 3394 }, { "epoch": 0.24069478908188585, "grad_norm": 2.389804025605416, "learning_rate": 8.880823730777e-06, "loss": 0.4283, "step": 3395 }, { "epoch": 0.24076568592697625, "grad_norm": 2.370829218356665, "learning_rate": 8.880099679582715e-06, "loss": 0.4253, "step": 3396 }, { "epoch": 0.24083658277206665, "grad_norm": 2.656690851428999, "learning_rate": 8.87937542378781e-06, "loss": 0.439, "step": 3397 }, { "epoch": 0.24090747961715703, "grad_norm": 2.0735440797847287, "learning_rate": 8.87865096343048e-06, "loss": 0.4032, "step": 3398 }, { "epoch": 0.24097837646224743, "grad_norm": 2.2873008864090605, "learning_rate": 8.877926298548922e-06, "loss": 0.4104, "step": 3399 }, { "epoch": 0.24104927330733783, "grad_norm": 3.4421363032302565, "learning_rate": 8.877201429181352e-06, "loss": 0.467, "step": 3400 }, { "epoch": 0.2411201701524282, "grad_norm": 3.725302309711422, "learning_rate": 8.87647635536599e-06, "loss": 0.4613, "step": 3401 }, { "epoch": 0.2411910669975186, "grad_norm": 3.1364751508788196, "learning_rate": 8.875751077141073e-06, "loss": 0.4453, "step": 3402 }, { "epoch": 0.241261963842609, "grad_norm": 2.3722908036359325, "learning_rate": 8.875025594544842e-06, "loss": 0.4134, "step": 3403 }, { "epoch": 0.2413328606876994, "grad_norm": 3.004490053993718, "learning_rate": 8.874299907615554e-06, "loss": 0.4783, "step": 3404 }, { "epoch": 0.24140375753278978, "grad_norm": 1.9778602777712104, "learning_rate": 8.873574016391475e-06, "loss": 0.3848, "step": 3405 }, { "epoch": 0.24147465437788018, "grad_norm": 7.41487460483142, "learning_rate": 8.87284792091088e-06, "loss": 0.4097, "step": 3406 }, { "epoch": 0.24154555122297058, "grad_norm": 5.1859834370930695, "learning_rate": 8.872121621212061e-06, "loss": 0.4161, "step": 3407 }, { "epoch": 0.24161644806806098, "grad_norm": 2.1353771003340354, "learning_rate": 8.871395117333312e-06, "loss": 0.4129, "step": 3408 }, { "epoch": 0.24168734491315136, "grad_norm": 2.638990634293985, "learning_rate": 8.870668409312943e-06, "loss": 0.4076, "step": 3409 }, { "epoch": 0.24175824175824176, "grad_norm": 3.7418399634333053, "learning_rate": 8.869941497189277e-06, "loss": 0.4358, "step": 3410 }, { "epoch": 0.24182913860333216, "grad_norm": 2.3432713469220454, "learning_rate": 8.869214381000639e-06, "loss": 0.3534, "step": 3411 }, { "epoch": 0.24190003544842253, "grad_norm": 2.286046266202036, "learning_rate": 8.868487060785374e-06, "loss": 0.4631, "step": 3412 }, { "epoch": 0.24197093229351294, "grad_norm": 1.7258528833108275, "learning_rate": 8.867759536581834e-06, "loss": 0.4239, "step": 3413 }, { "epoch": 0.24204182913860334, "grad_norm": 2.852851594997272, "learning_rate": 8.867031808428382e-06, "loss": 0.389, "step": 3414 }, { "epoch": 0.24211272598369374, "grad_norm": 5.520276409388905, "learning_rate": 8.866303876363391e-06, "loss": 0.4673, "step": 3415 }, { "epoch": 0.2421836228287841, "grad_norm": 5.591710191409889, "learning_rate": 8.865575740425245e-06, "loss": 0.4676, "step": 3416 }, { "epoch": 0.2422545196738745, "grad_norm": 5.560369622158736, "learning_rate": 8.864847400652341e-06, "loss": 0.4284, "step": 3417 }, { "epoch": 0.24232541651896491, "grad_norm": 2.70300687624693, "learning_rate": 8.864118857083083e-06, "loss": 0.4496, "step": 3418 }, { "epoch": 0.2423963133640553, "grad_norm": 1.749424777132604, "learning_rate": 8.863390109755886e-06, "loss": 0.3992, "step": 3419 }, { "epoch": 0.2424672102091457, "grad_norm": 4.696403925548912, "learning_rate": 8.862661158709184e-06, "loss": 0.4134, "step": 3420 }, { "epoch": 0.2425381070542361, "grad_norm": 2.9184689592422957, "learning_rate": 8.861932003981409e-06, "loss": 0.4687, "step": 3421 }, { "epoch": 0.2426090038993265, "grad_norm": 2.0143846534960237, "learning_rate": 8.86120264561101e-06, "loss": 0.4286, "step": 3422 }, { "epoch": 0.24267990074441687, "grad_norm": 5.020995489609952, "learning_rate": 8.86047308363645e-06, "loss": 0.4099, "step": 3423 }, { "epoch": 0.24275079758950727, "grad_norm": 2.3012931240405865, "learning_rate": 8.859743318096198e-06, "loss": 0.417, "step": 3424 }, { "epoch": 0.24282169443459767, "grad_norm": 5.441882044811627, "learning_rate": 8.859013349028735e-06, "loss": 0.4454, "step": 3425 }, { "epoch": 0.24289259127968804, "grad_norm": 2.3599471049128202, "learning_rate": 8.858283176472552e-06, "loss": 0.4552, "step": 3426 }, { "epoch": 0.24296348812477844, "grad_norm": 1.680076028438383, "learning_rate": 8.857552800466152e-06, "loss": 0.4438, "step": 3427 }, { "epoch": 0.24303438496986884, "grad_norm": 2.0444936120220145, "learning_rate": 8.856822221048048e-06, "loss": 0.4789, "step": 3428 }, { "epoch": 0.24310528181495925, "grad_norm": 2.2542953415162583, "learning_rate": 8.856091438256766e-06, "loss": 0.3851, "step": 3429 }, { "epoch": 0.24317617866004962, "grad_norm": 2.8824609902302325, "learning_rate": 8.85536045213084e-06, "loss": 0.4803, "step": 3430 }, { "epoch": 0.24324707550514002, "grad_norm": 3.5869017578165145, "learning_rate": 8.854629262708811e-06, "loss": 0.4304, "step": 3431 }, { "epoch": 0.24331797235023042, "grad_norm": 2.147528400194839, "learning_rate": 8.853897870029241e-06, "loss": 0.4292, "step": 3432 }, { "epoch": 0.2433888691953208, "grad_norm": 6.258542741131767, "learning_rate": 8.853166274130696e-06, "loss": 0.4232, "step": 3433 }, { "epoch": 0.2434597660404112, "grad_norm": 5.744322664923311, "learning_rate": 8.85243447505175e-06, "loss": 0.4616, "step": 3434 }, { "epoch": 0.2435306628855016, "grad_norm": 3.017566182557771, "learning_rate": 8.851702472830997e-06, "loss": 0.4224, "step": 3435 }, { "epoch": 0.243601559730592, "grad_norm": 3.0891975044829723, "learning_rate": 8.85097026750703e-06, "loss": 0.4747, "step": 3436 }, { "epoch": 0.24367245657568237, "grad_norm": 2.713511681740667, "learning_rate": 8.850237859118462e-06, "loss": 0.4441, "step": 3437 }, { "epoch": 0.24374335342077277, "grad_norm": 2.077016115608556, "learning_rate": 8.849505247703914e-06, "loss": 0.4311, "step": 3438 }, { "epoch": 0.24381425026586317, "grad_norm": 2.4170117227275765, "learning_rate": 8.848772433302012e-06, "loss": 0.4039, "step": 3439 }, { "epoch": 0.24388514711095358, "grad_norm": 2.590856317253881, "learning_rate": 8.848039415951408e-06, "loss": 0.43, "step": 3440 }, { "epoch": 0.24395604395604395, "grad_norm": 3.053880713930957, "learning_rate": 8.847306195690744e-06, "loss": 0.4237, "step": 3441 }, { "epoch": 0.24402694080113435, "grad_norm": 2.593158979710797, "learning_rate": 8.846572772558691e-06, "loss": 0.3898, "step": 3442 }, { "epoch": 0.24409783764622475, "grad_norm": 2.482425185417696, "learning_rate": 8.845839146593918e-06, "loss": 0.4896, "step": 3443 }, { "epoch": 0.24416873449131513, "grad_norm": 3.561721164853681, "learning_rate": 8.845105317835112e-06, "loss": 0.4329, "step": 3444 }, { "epoch": 0.24423963133640553, "grad_norm": 5.179615060107677, "learning_rate": 8.844371286320967e-06, "loss": 0.4554, "step": 3445 }, { "epoch": 0.24431052818149593, "grad_norm": 2.5178169614424126, "learning_rate": 8.843637052090191e-06, "loss": 0.4729, "step": 3446 }, { "epoch": 0.24438142502658633, "grad_norm": 3.952431985561848, "learning_rate": 8.842902615181497e-06, "loss": 0.4604, "step": 3447 }, { "epoch": 0.2444523218716767, "grad_norm": 7.391881470848358, "learning_rate": 8.842167975633617e-06, "loss": 0.4234, "step": 3448 }, { "epoch": 0.2445232187167671, "grad_norm": 1.1950508978301424, "learning_rate": 8.841433133485286e-06, "loss": 0.7069, "step": 3449 }, { "epoch": 0.2445941155618575, "grad_norm": 6.4215544838642975, "learning_rate": 8.840698088775254e-06, "loss": 0.4477, "step": 3450 }, { "epoch": 0.24466501240694788, "grad_norm": 3.9342272974336057, "learning_rate": 8.839962841542283e-06, "loss": 0.4159, "step": 3451 }, { "epoch": 0.24473590925203828, "grad_norm": 2.108380408968342, "learning_rate": 8.839227391825136e-06, "loss": 0.4021, "step": 3452 }, { "epoch": 0.24480680609712868, "grad_norm": 6.5701981447356514, "learning_rate": 8.838491739662602e-06, "loss": 0.4452, "step": 3453 }, { "epoch": 0.24487770294221908, "grad_norm": 5.823504717693977, "learning_rate": 8.837755885093467e-06, "loss": 0.4222, "step": 3454 }, { "epoch": 0.24494859978730946, "grad_norm": 3.4764522499889225, "learning_rate": 8.837019828156534e-06, "loss": 0.438, "step": 3455 }, { "epoch": 0.24501949663239986, "grad_norm": 3.3305503606680626, "learning_rate": 8.83628356889062e-06, "loss": 0.4872, "step": 3456 }, { "epoch": 0.24509039347749026, "grad_norm": 2.5800047012208007, "learning_rate": 8.835547107334544e-06, "loss": 0.3762, "step": 3457 }, { "epoch": 0.24516129032258063, "grad_norm": 2.17736394485937, "learning_rate": 8.83481044352714e-06, "loss": 0.4322, "step": 3458 }, { "epoch": 0.24523218716767103, "grad_norm": 2.0237379774846223, "learning_rate": 8.834073577507255e-06, "loss": 0.4132, "step": 3459 }, { "epoch": 0.24530308401276144, "grad_norm": 4.509271618393153, "learning_rate": 8.833336509313744e-06, "loss": 0.4811, "step": 3460 }, { "epoch": 0.24537398085785184, "grad_norm": 1.8491015125867785, "learning_rate": 8.832599238985474e-06, "loss": 0.4401, "step": 3461 }, { "epoch": 0.2454448777029422, "grad_norm": 4.004786156702921, "learning_rate": 8.831861766561318e-06, "loss": 0.4066, "step": 3462 }, { "epoch": 0.2455157745480326, "grad_norm": 3.029425132701892, "learning_rate": 8.83112409208017e-06, "loss": 0.4794, "step": 3463 }, { "epoch": 0.245586671393123, "grad_norm": 1.0893236341085524, "learning_rate": 8.830386215580921e-06, "loss": 0.661, "step": 3464 }, { "epoch": 0.2456575682382134, "grad_norm": 1.5033909818957014, "learning_rate": 8.829648137102486e-06, "loss": 0.3945, "step": 3465 }, { "epoch": 0.2457284650833038, "grad_norm": 3.4255108820237075, "learning_rate": 8.82890985668378e-06, "loss": 0.4476, "step": 3466 }, { "epoch": 0.2457993619283942, "grad_norm": 5.552759488062566, "learning_rate": 8.828171374363736e-06, "loss": 0.4539, "step": 3467 }, { "epoch": 0.2458702587734846, "grad_norm": 3.282514831205256, "learning_rate": 8.827432690181292e-06, "loss": 0.3777, "step": 3468 }, { "epoch": 0.24594115561857496, "grad_norm": 3.778035583862983, "learning_rate": 8.826693804175403e-06, "loss": 0.4432, "step": 3469 }, { "epoch": 0.24601205246366536, "grad_norm": 3.4410434824971583, "learning_rate": 8.825954716385028e-06, "loss": 0.4597, "step": 3470 }, { "epoch": 0.24608294930875577, "grad_norm": 2.481930546637839, "learning_rate": 8.82521542684914e-06, "loss": 0.4413, "step": 3471 }, { "epoch": 0.24615384615384617, "grad_norm": 2.486865210254036, "learning_rate": 8.824475935606723e-06, "loss": 0.4447, "step": 3472 }, { "epoch": 0.24622474299893654, "grad_norm": 2.994199665577017, "learning_rate": 8.823736242696773e-06, "loss": 0.4052, "step": 3473 }, { "epoch": 0.24629563984402694, "grad_norm": 2.758157192327815, "learning_rate": 8.822996348158288e-06, "loss": 0.5268, "step": 3474 }, { "epoch": 0.24636653668911734, "grad_norm": 2.4298692471974874, "learning_rate": 8.82225625203029e-06, "loss": 0.4534, "step": 3475 }, { "epoch": 0.24643743353420772, "grad_norm": 2.385621087921703, "learning_rate": 8.821515954351803e-06, "loss": 0.4276, "step": 3476 }, { "epoch": 0.24650833037929812, "grad_norm": 2.271391462473539, "learning_rate": 8.820775455161864e-06, "loss": 0.4081, "step": 3477 }, { "epoch": 0.24657922722438852, "grad_norm": 2.4284758183094475, "learning_rate": 8.820034754499516e-06, "loss": 0.4934, "step": 3478 }, { "epoch": 0.24665012406947892, "grad_norm": 2.625063695109953, "learning_rate": 8.81929385240382e-06, "loss": 0.4299, "step": 3479 }, { "epoch": 0.2467210209145693, "grad_norm": 2.3204405582638943, "learning_rate": 8.818552748913846e-06, "loss": 0.4157, "step": 3480 }, { "epoch": 0.2467919177596597, "grad_norm": 2.657344593417817, "learning_rate": 8.81781144406867e-06, "loss": 0.4162, "step": 3481 }, { "epoch": 0.2468628146047501, "grad_norm": 5.562247729957768, "learning_rate": 8.817069937907383e-06, "loss": 0.443, "step": 3482 }, { "epoch": 0.24693371144984047, "grad_norm": 2.170759699335824, "learning_rate": 8.816328230469085e-06, "loss": 0.4404, "step": 3483 }, { "epoch": 0.24700460829493087, "grad_norm": 3.008298592875387, "learning_rate": 8.815586321792884e-06, "loss": 0.3879, "step": 3484 }, { "epoch": 0.24707550514002127, "grad_norm": 3.0632385944582503, "learning_rate": 8.814844211917907e-06, "loss": 0.4529, "step": 3485 }, { "epoch": 0.24714640198511167, "grad_norm": 2.8063854765705774, "learning_rate": 8.814101900883282e-06, "loss": 0.4503, "step": 3486 }, { "epoch": 0.24721729883020205, "grad_norm": 4.991278386507632, "learning_rate": 8.813359388728152e-06, "loss": 0.4576, "step": 3487 }, { "epoch": 0.24728819567529245, "grad_norm": 2.253034893473186, "learning_rate": 8.812616675491672e-06, "loss": 0.4733, "step": 3488 }, { "epoch": 0.24735909252038285, "grad_norm": 1.8208340763849682, "learning_rate": 8.811873761213003e-06, "loss": 0.4388, "step": 3489 }, { "epoch": 0.24742998936547322, "grad_norm": 2.8117756642813214, "learning_rate": 8.811130645931323e-06, "loss": 0.4758, "step": 3490 }, { "epoch": 0.24750088621056363, "grad_norm": 2.362755426296798, "learning_rate": 8.810387329685813e-06, "loss": 0.4618, "step": 3491 }, { "epoch": 0.24757178305565403, "grad_norm": 2.2365064587063754, "learning_rate": 8.809643812515672e-06, "loss": 0.4569, "step": 3492 }, { "epoch": 0.24764267990074443, "grad_norm": 2.0307390194531854, "learning_rate": 8.808900094460106e-06, "loss": 0.4442, "step": 3493 }, { "epoch": 0.2477135767458348, "grad_norm": 2.838576078451795, "learning_rate": 8.808156175558328e-06, "loss": 0.4084, "step": 3494 }, { "epoch": 0.2477844735909252, "grad_norm": 3.9128084457580363, "learning_rate": 8.80741205584957e-06, "loss": 0.4149, "step": 3495 }, { "epoch": 0.2478553704360156, "grad_norm": 4.6472588528097205, "learning_rate": 8.80666773537307e-06, "loss": 0.4439, "step": 3496 }, { "epoch": 0.24792626728110598, "grad_norm": 3.862027873987669, "learning_rate": 8.805923214168072e-06, "loss": 0.4494, "step": 3497 }, { "epoch": 0.24799716412619638, "grad_norm": 2.1296277364276888, "learning_rate": 8.80517849227384e-06, "loss": 0.4138, "step": 3498 }, { "epoch": 0.24806806097128678, "grad_norm": 5.579732801348267, "learning_rate": 8.80443356972964e-06, "loss": 0.4527, "step": 3499 }, { "epoch": 0.24813895781637718, "grad_norm": 2.302918174937007, "learning_rate": 8.803688446574754e-06, "loss": 0.3999, "step": 3500 }, { "epoch": 0.24820985466146755, "grad_norm": 2.883110666209527, "learning_rate": 8.802943122848474e-06, "loss": 0.4035, "step": 3501 }, { "epoch": 0.24828075150655796, "grad_norm": 2.568738921373575, "learning_rate": 8.802197598590102e-06, "loss": 0.4788, "step": 3502 }, { "epoch": 0.24835164835164836, "grad_norm": 1.2067543192333636, "learning_rate": 8.801451873838946e-06, "loss": 0.686, "step": 3503 }, { "epoch": 0.24842254519673876, "grad_norm": 2.0503242008629363, "learning_rate": 8.800705948634331e-06, "loss": 0.3676, "step": 3504 }, { "epoch": 0.24849344204182913, "grad_norm": 2.688118613842857, "learning_rate": 8.79995982301559e-06, "loss": 0.4374, "step": 3505 }, { "epoch": 0.24856433888691953, "grad_norm": 2.032152602197654, "learning_rate": 8.79921349702207e-06, "loss": 0.398, "step": 3506 }, { "epoch": 0.24863523573200993, "grad_norm": 2.3166729733419205, "learning_rate": 8.798466970693121e-06, "loss": 0.4343, "step": 3507 }, { "epoch": 0.2487061325771003, "grad_norm": 1.9707960728309084, "learning_rate": 8.79772024406811e-06, "loss": 0.4329, "step": 3508 }, { "epoch": 0.2487770294221907, "grad_norm": 5.165661126108426, "learning_rate": 8.796973317186412e-06, "loss": 0.4517, "step": 3509 }, { "epoch": 0.2488479262672811, "grad_norm": 2.188869948737533, "learning_rate": 8.79622619008741e-06, "loss": 0.4295, "step": 3510 }, { "epoch": 0.2489188231123715, "grad_norm": 2.235759533092055, "learning_rate": 8.795478862810507e-06, "loss": 0.4593, "step": 3511 }, { "epoch": 0.24898971995746189, "grad_norm": 3.1573677550030212, "learning_rate": 8.794731335395105e-06, "loss": 0.422, "step": 3512 }, { "epoch": 0.2490606168025523, "grad_norm": 3.1839170255735847, "learning_rate": 8.793983607880622e-06, "loss": 0.4342, "step": 3513 }, { "epoch": 0.2491315136476427, "grad_norm": 2.3191271860545095, "learning_rate": 8.793235680306492e-06, "loss": 0.4585, "step": 3514 }, { "epoch": 0.24920241049273306, "grad_norm": 2.061685422318976, "learning_rate": 8.792487552712146e-06, "loss": 0.4138, "step": 3515 }, { "epoch": 0.24927330733782346, "grad_norm": 6.359532426568577, "learning_rate": 8.791739225137037e-06, "loss": 0.4011, "step": 3516 }, { "epoch": 0.24934420418291386, "grad_norm": 1.8883347749954043, "learning_rate": 8.790990697620623e-06, "loss": 0.435, "step": 3517 }, { "epoch": 0.24941510102800427, "grad_norm": 8.442729472773337, "learning_rate": 8.790241970202378e-06, "loss": 0.4682, "step": 3518 }, { "epoch": 0.24948599787309464, "grad_norm": 6.936986417857813, "learning_rate": 8.789493042921782e-06, "loss": 0.4302, "step": 3519 }, { "epoch": 0.24955689471818504, "grad_norm": 2.5155282845806455, "learning_rate": 8.788743915818323e-06, "loss": 0.4451, "step": 3520 }, { "epoch": 0.24962779156327544, "grad_norm": 4.552859736522893, "learning_rate": 8.787994588931507e-06, "loss": 0.3664, "step": 3521 }, { "epoch": 0.24969868840836582, "grad_norm": 2.7333866136656315, "learning_rate": 8.787245062300845e-06, "loss": 0.4207, "step": 3522 }, { "epoch": 0.24976958525345622, "grad_norm": 4.69002755558559, "learning_rate": 8.786495335965858e-06, "loss": 0.4673, "step": 3523 }, { "epoch": 0.24984048209854662, "grad_norm": 4.128272932083557, "learning_rate": 8.785745409966084e-06, "loss": 0.4259, "step": 3524 }, { "epoch": 0.24991137894363702, "grad_norm": 2.920456759792705, "learning_rate": 8.784995284341065e-06, "loss": 0.4375, "step": 3525 }, { "epoch": 0.2499822757887274, "grad_norm": 2.1184405830047215, "learning_rate": 8.784244959130357e-06, "loss": 0.3837, "step": 3526 }, { "epoch": 0.2500531726338178, "grad_norm": 1.918999745268729, "learning_rate": 8.78349443437352e-06, "loss": 0.3738, "step": 3527 }, { "epoch": 0.2501240694789082, "grad_norm": 1.6967154417356913, "learning_rate": 8.782743710110138e-06, "loss": 0.3862, "step": 3528 }, { "epoch": 0.25019496632399857, "grad_norm": 1.7561595684041706, "learning_rate": 8.781992786379789e-06, "loss": 0.4044, "step": 3529 }, { "epoch": 0.250265863169089, "grad_norm": 3.439906876136479, "learning_rate": 8.781241663222077e-06, "loss": 0.4119, "step": 3530 }, { "epoch": 0.25033676001417937, "grad_norm": 4.7444197830528125, "learning_rate": 8.780490340676603e-06, "loss": 0.494, "step": 3531 }, { "epoch": 0.25040765685926974, "grad_norm": 4.273675638329505, "learning_rate": 8.77973881878299e-06, "loss": 0.4269, "step": 3532 }, { "epoch": 0.2504785537043602, "grad_norm": 2.598782620871862, "learning_rate": 8.778987097580865e-06, "loss": 0.4311, "step": 3533 }, { "epoch": 0.25054945054945055, "grad_norm": 2.606147092558575, "learning_rate": 8.778235177109864e-06, "loss": 0.483, "step": 3534 }, { "epoch": 0.2506203473945409, "grad_norm": 7.903621815632829, "learning_rate": 8.777483057409639e-06, "loss": 0.4325, "step": 3535 }, { "epoch": 0.25069124423963135, "grad_norm": 1.912961178332686, "learning_rate": 8.776730738519848e-06, "loss": 0.4151, "step": 3536 }, { "epoch": 0.2507621410847217, "grad_norm": 1.5850503496617527, "learning_rate": 8.775978220480164e-06, "loss": 0.4815, "step": 3537 }, { "epoch": 0.2508330379298121, "grad_norm": 3.5940725016679242, "learning_rate": 8.775225503330266e-06, "loss": 0.4507, "step": 3538 }, { "epoch": 0.2509039347749025, "grad_norm": 6.666065225127014, "learning_rate": 8.774472587109846e-06, "loss": 0.4076, "step": 3539 }, { "epoch": 0.2509748316199929, "grad_norm": 2.940530770696674, "learning_rate": 8.773719471858605e-06, "loss": 0.4266, "step": 3540 }, { "epoch": 0.25104572846508333, "grad_norm": 6.644497572729141, "learning_rate": 8.772966157616256e-06, "loss": 0.4482, "step": 3541 }, { "epoch": 0.2511166253101737, "grad_norm": 2.8850407878633106, "learning_rate": 8.77221264442252e-06, "loss": 0.4714, "step": 3542 }, { "epoch": 0.2511875221552641, "grad_norm": 2.951201209470134, "learning_rate": 8.771458932317135e-06, "loss": 0.5226, "step": 3543 }, { "epoch": 0.2512584190003545, "grad_norm": 2.064618276197783, "learning_rate": 8.77070502133984e-06, "loss": 0.4529, "step": 3544 }, { "epoch": 0.2513293158454449, "grad_norm": 4.973455094961624, "learning_rate": 8.769950911530392e-06, "loss": 0.4326, "step": 3545 }, { "epoch": 0.25140021269053525, "grad_norm": 9.023526201223117, "learning_rate": 8.769196602928556e-06, "loss": 0.4189, "step": 3546 }, { "epoch": 0.2514711095356257, "grad_norm": 2.4153009614259524, "learning_rate": 8.768442095574104e-06, "loss": 0.4767, "step": 3547 }, { "epoch": 0.25154200638071605, "grad_norm": 2.703746429751278, "learning_rate": 8.767687389506825e-06, "loss": 0.4218, "step": 3548 }, { "epoch": 0.25161290322580643, "grad_norm": 2.1500489927314534, "learning_rate": 8.766932484766515e-06, "loss": 0.4261, "step": 3549 }, { "epoch": 0.25168380007089686, "grad_norm": 3.536208467698112, "learning_rate": 8.76617738139298e-06, "loss": 0.4509, "step": 3550 }, { "epoch": 0.25175469691598723, "grad_norm": 1.9355631555246482, "learning_rate": 8.765422079426037e-06, "loss": 0.4322, "step": 3551 }, { "epoch": 0.25182559376107766, "grad_norm": 2.2551590980061325, "learning_rate": 8.764666578905511e-06, "loss": 0.4186, "step": 3552 }, { "epoch": 0.25189649060616803, "grad_norm": 2.508999836816528, "learning_rate": 8.763910879871244e-06, "loss": 0.4189, "step": 3553 }, { "epoch": 0.2519673874512584, "grad_norm": 2.3341854977808927, "learning_rate": 8.763154982363085e-06, "loss": 0.4935, "step": 3554 }, { "epoch": 0.25203828429634884, "grad_norm": 4.837237948545586, "learning_rate": 8.762398886420891e-06, "loss": 0.4538, "step": 3555 }, { "epoch": 0.2521091811414392, "grad_norm": 1.9222466485559055, "learning_rate": 8.761642592084532e-06, "loss": 0.3996, "step": 3556 }, { "epoch": 0.2521800779865296, "grad_norm": 11.721448510541535, "learning_rate": 8.760886099393888e-06, "loss": 0.4538, "step": 3557 }, { "epoch": 0.25225097483162, "grad_norm": 6.261657305674597, "learning_rate": 8.760129408388848e-06, "loss": 0.4217, "step": 3558 }, { "epoch": 0.2523218716767104, "grad_norm": 2.8771422610260395, "learning_rate": 8.759372519109317e-06, "loss": 0.4226, "step": 3559 }, { "epoch": 0.25239276852180076, "grad_norm": 2.4887115453572126, "learning_rate": 8.758615431595202e-06, "loss": 0.4404, "step": 3560 }, { "epoch": 0.2524636653668912, "grad_norm": 2.476167868746466, "learning_rate": 8.757858145886426e-06, "loss": 0.3658, "step": 3561 }, { "epoch": 0.25253456221198156, "grad_norm": 3.048357757320896, "learning_rate": 8.757100662022921e-06, "loss": 0.4772, "step": 3562 }, { "epoch": 0.25260545905707193, "grad_norm": 2.554170136337277, "learning_rate": 8.756342980044633e-06, "loss": 0.41, "step": 3563 }, { "epoch": 0.25267635590216236, "grad_norm": 2.4308356733626946, "learning_rate": 8.755585099991511e-06, "loss": 0.4858, "step": 3564 }, { "epoch": 0.25274725274725274, "grad_norm": 5.155197607105434, "learning_rate": 8.754827021903522e-06, "loss": 0.4882, "step": 3565 }, { "epoch": 0.25281814959234317, "grad_norm": 3.0761222043060696, "learning_rate": 8.754068745820637e-06, "loss": 0.424, "step": 3566 }, { "epoch": 0.25288904643743354, "grad_norm": 4.1586754033778295, "learning_rate": 8.75331027178284e-06, "loss": 0.4268, "step": 3567 }, { "epoch": 0.2529599432825239, "grad_norm": 3.2686816715582627, "learning_rate": 8.75255159983013e-06, "loss": 0.4447, "step": 3568 }, { "epoch": 0.25303084012761434, "grad_norm": 2.2510389492588034, "learning_rate": 8.75179273000251e-06, "loss": 0.4144, "step": 3569 }, { "epoch": 0.2531017369727047, "grad_norm": 2.4999570362744503, "learning_rate": 8.751033662339993e-06, "loss": 0.4504, "step": 3570 }, { "epoch": 0.2531726338177951, "grad_norm": 4.264055598054252, "learning_rate": 8.75027439688261e-06, "loss": 0.4464, "step": 3571 }, { "epoch": 0.2532435306628855, "grad_norm": 2.8591751353970376, "learning_rate": 8.749514933670396e-06, "loss": 0.4624, "step": 3572 }, { "epoch": 0.2533144275079759, "grad_norm": 2.0315017827279584, "learning_rate": 8.748755272743396e-06, "loss": 0.4072, "step": 3573 }, { "epoch": 0.25338532435306627, "grad_norm": 4.104104777299614, "learning_rate": 8.74799541414167e-06, "loss": 0.4516, "step": 3574 }, { "epoch": 0.2534562211981567, "grad_norm": 4.977101380190131, "learning_rate": 8.747235357905285e-06, "loss": 0.4921, "step": 3575 }, { "epoch": 0.25352711804324707, "grad_norm": 2.736398435874533, "learning_rate": 8.74647510407432e-06, "loss": 0.4367, "step": 3576 }, { "epoch": 0.25359801488833744, "grad_norm": 2.477514997383167, "learning_rate": 8.745714652688864e-06, "loss": 0.4399, "step": 3577 }, { "epoch": 0.25366891173342787, "grad_norm": 9.4338660695207, "learning_rate": 8.744954003789013e-06, "loss": 0.4687, "step": 3578 }, { "epoch": 0.25373980857851824, "grad_norm": 5.548463534294726, "learning_rate": 8.744193157414879e-06, "loss": 0.4182, "step": 3579 }, { "epoch": 0.2538107054236087, "grad_norm": 6.738768444219186, "learning_rate": 8.743432113606584e-06, "loss": 0.4261, "step": 3580 }, { "epoch": 0.25388160226869905, "grad_norm": 3.0259371926143275, "learning_rate": 8.742670872404257e-06, "loss": 0.4356, "step": 3581 }, { "epoch": 0.2539524991137894, "grad_norm": 3.1912443710468725, "learning_rate": 8.741909433848037e-06, "loss": 0.4867, "step": 3582 }, { "epoch": 0.25402339595887985, "grad_norm": 1.8768255778025544, "learning_rate": 8.741147797978076e-06, "loss": 0.4412, "step": 3583 }, { "epoch": 0.2540942928039702, "grad_norm": 4.056185232953842, "learning_rate": 8.740385964834536e-06, "loss": 0.429, "step": 3584 }, { "epoch": 0.2541651896490606, "grad_norm": 2.3228304195375316, "learning_rate": 8.739623934457591e-06, "loss": 0.3837, "step": 3585 }, { "epoch": 0.254236086494151, "grad_norm": 2.2046745391071405, "learning_rate": 8.73886170688742e-06, "loss": 0.3837, "step": 3586 }, { "epoch": 0.2543069833392414, "grad_norm": 4.910005087996593, "learning_rate": 8.738099282164216e-06, "loss": 0.4762, "step": 3587 }, { "epoch": 0.2543778801843318, "grad_norm": 3.6533169525336717, "learning_rate": 8.737336660328187e-06, "loss": 0.4269, "step": 3588 }, { "epoch": 0.2544487770294222, "grad_norm": 2.3595766371987197, "learning_rate": 8.736573841419541e-06, "loss": 0.453, "step": 3589 }, { "epoch": 0.2545196738745126, "grad_norm": 7.470652974231138, "learning_rate": 8.735810825478506e-06, "loss": 0.4259, "step": 3590 }, { "epoch": 0.254590570719603, "grad_norm": 5.682776173102075, "learning_rate": 8.735047612545314e-06, "loss": 0.3781, "step": 3591 }, { "epoch": 0.2546614675646934, "grad_norm": 3.610860095290555, "learning_rate": 8.734284202660208e-06, "loss": 0.4455, "step": 3592 }, { "epoch": 0.25473236440978375, "grad_norm": 6.8347250988424895, "learning_rate": 8.73352059586345e-06, "loss": 0.4172, "step": 3593 }, { "epoch": 0.2548032612548742, "grad_norm": 3.411963562295503, "learning_rate": 8.732756792195297e-06, "loss": 0.4132, "step": 3594 }, { "epoch": 0.25487415809996455, "grad_norm": 2.1034212852062164, "learning_rate": 8.731992791696033e-06, "loss": 0.4188, "step": 3595 }, { "epoch": 0.2549450549450549, "grad_norm": 3.9537054996492733, "learning_rate": 8.73122859440594e-06, "loss": 0.3844, "step": 3596 }, { "epoch": 0.25501595179014536, "grad_norm": 3.0495034003676764, "learning_rate": 8.730464200365315e-06, "loss": 0.4206, "step": 3597 }, { "epoch": 0.25508684863523573, "grad_norm": 2.3045170663268526, "learning_rate": 8.729699609614465e-06, "loss": 0.4002, "step": 3598 }, { "epoch": 0.2551577454803261, "grad_norm": 2.915220972399393, "learning_rate": 8.728934822193708e-06, "loss": 0.4149, "step": 3599 }, { "epoch": 0.25522864232541653, "grad_norm": 2.9156372734839784, "learning_rate": 8.728169838143371e-06, "loss": 0.4549, "step": 3600 }, { "epoch": 0.2552995391705069, "grad_norm": 4.63430866695299, "learning_rate": 8.727404657503792e-06, "loss": 0.3657, "step": 3601 }, { "epoch": 0.2553704360155973, "grad_norm": 3.9754795264259006, "learning_rate": 8.726639280315324e-06, "loss": 0.4203, "step": 3602 }, { "epoch": 0.2554413328606877, "grad_norm": 3.248296534658843, "learning_rate": 8.72587370661832e-06, "loss": 0.4313, "step": 3603 }, { "epoch": 0.2555122297057781, "grad_norm": 3.426333731236058, "learning_rate": 8.725107936453151e-06, "loss": 0.5168, "step": 3604 }, { "epoch": 0.2555831265508685, "grad_norm": 3.1058367855589526, "learning_rate": 8.724341969860196e-06, "loss": 0.4766, "step": 3605 }, { "epoch": 0.2556540233959589, "grad_norm": 2.496539074277095, "learning_rate": 8.72357580687985e-06, "loss": 0.4631, "step": 3606 }, { "epoch": 0.25572492024104926, "grad_norm": 2.9723775090664493, "learning_rate": 8.722809447552508e-06, "loss": 0.4611, "step": 3607 }, { "epoch": 0.2557958170861397, "grad_norm": 2.3179206018783214, "learning_rate": 8.722042891918582e-06, "loss": 0.4244, "step": 3608 }, { "epoch": 0.25586671393123006, "grad_norm": 2.4258329266035785, "learning_rate": 8.721276140018493e-06, "loss": 0.4048, "step": 3609 }, { "epoch": 0.25593761077632043, "grad_norm": 2.4739963438782717, "learning_rate": 8.720509191892675e-06, "loss": 0.4509, "step": 3610 }, { "epoch": 0.25600850762141086, "grad_norm": 3.7460447728697437, "learning_rate": 8.719742047581565e-06, "loss": 0.431, "step": 3611 }, { "epoch": 0.25607940446650124, "grad_norm": 2.0409670472876233, "learning_rate": 8.71897470712562e-06, "loss": 0.3477, "step": 3612 }, { "epoch": 0.2561503013115916, "grad_norm": 2.6649047085063327, "learning_rate": 8.718207170565298e-06, "loss": 0.413, "step": 3613 }, { "epoch": 0.25622119815668204, "grad_norm": 2.8333935486299966, "learning_rate": 8.717439437941076e-06, "loss": 0.4273, "step": 3614 }, { "epoch": 0.2562920950017724, "grad_norm": 2.3775004018048564, "learning_rate": 8.716671509293433e-06, "loss": 0.493, "step": 3615 }, { "epoch": 0.25636299184686284, "grad_norm": 2.308233963788947, "learning_rate": 8.715903384662865e-06, "loss": 0.4298, "step": 3616 }, { "epoch": 0.2564338886919532, "grad_norm": 40.939808752286986, "learning_rate": 8.715135064089875e-06, "loss": 0.43, "step": 3617 }, { "epoch": 0.2565047855370436, "grad_norm": 4.130651414728113, "learning_rate": 8.714366547614979e-06, "loss": 0.4182, "step": 3618 }, { "epoch": 0.256575682382134, "grad_norm": 2.2069665330474177, "learning_rate": 8.713597835278697e-06, "loss": 0.4193, "step": 3619 }, { "epoch": 0.2566465792272244, "grad_norm": 4.419014211482669, "learning_rate": 8.71282892712157e-06, "loss": 0.4609, "step": 3620 }, { "epoch": 0.25671747607231477, "grad_norm": 1.1383537678265165, "learning_rate": 8.712059823184139e-06, "loss": 0.6698, "step": 3621 }, { "epoch": 0.2567883729174052, "grad_norm": 1.8122786748333326, "learning_rate": 8.711290523506958e-06, "loss": 0.4149, "step": 3622 }, { "epoch": 0.25685926976249557, "grad_norm": 4.0502291912294295, "learning_rate": 8.710521028130598e-06, "loss": 0.4412, "step": 3623 }, { "epoch": 0.25693016660758594, "grad_norm": 2.945548506816414, "learning_rate": 8.70975133709563e-06, "loss": 0.482, "step": 3624 }, { "epoch": 0.25700106345267637, "grad_norm": 4.472783333111875, "learning_rate": 8.708981450442644e-06, "loss": 0.4664, "step": 3625 }, { "epoch": 0.25707196029776674, "grad_norm": 1.9852846806369595, "learning_rate": 8.708211368212233e-06, "loss": 0.4429, "step": 3626 }, { "epoch": 0.2571428571428571, "grad_norm": 4.834177616596805, "learning_rate": 8.707441090445009e-06, "loss": 0.4705, "step": 3627 }, { "epoch": 0.25721375398794755, "grad_norm": 1.8142892474213834, "learning_rate": 8.706670617181586e-06, "loss": 0.4608, "step": 3628 }, { "epoch": 0.2572846508330379, "grad_norm": 2.8515016869898537, "learning_rate": 8.705899948462591e-06, "loss": 0.4099, "step": 3629 }, { "epoch": 0.25735554767812835, "grad_norm": 2.9635573710822882, "learning_rate": 8.705129084328662e-06, "loss": 0.5132, "step": 3630 }, { "epoch": 0.2574264445232187, "grad_norm": 3.0383644961020404, "learning_rate": 8.70435802482045e-06, "loss": 0.4491, "step": 3631 }, { "epoch": 0.2574973413683091, "grad_norm": 2.6679602282868755, "learning_rate": 8.703586769978612e-06, "loss": 0.4462, "step": 3632 }, { "epoch": 0.2575682382133995, "grad_norm": 3.0862493604007253, "learning_rate": 8.702815319843815e-06, "loss": 0.4837, "step": 3633 }, { "epoch": 0.2576391350584899, "grad_norm": 3.6382441167534285, "learning_rate": 8.702043674456744e-06, "loss": 0.4145, "step": 3634 }, { "epoch": 0.25771003190358027, "grad_norm": 2.6062406443471686, "learning_rate": 8.70127183385808e-06, "loss": 0.4059, "step": 3635 }, { "epoch": 0.2577809287486707, "grad_norm": 2.9134140931292767, "learning_rate": 8.70049979808853e-06, "loss": 0.4034, "step": 3636 }, { "epoch": 0.2578518255937611, "grad_norm": 1.1536017458929033, "learning_rate": 8.6997275671888e-06, "loss": 0.6961, "step": 3637 }, { "epoch": 0.25792272243885145, "grad_norm": 8.129566532617583, "learning_rate": 8.698955141199612e-06, "loss": 0.4946, "step": 3638 }, { "epoch": 0.2579936192839419, "grad_norm": 2.5971353386747, "learning_rate": 8.698182520161696e-06, "loss": 0.4308, "step": 3639 }, { "epoch": 0.25806451612903225, "grad_norm": 3.8953260938339787, "learning_rate": 8.697409704115792e-06, "loss": 0.4654, "step": 3640 }, { "epoch": 0.2581354129741226, "grad_norm": 3.4194399789342356, "learning_rate": 8.696636693102654e-06, "loss": 0.4509, "step": 3641 }, { "epoch": 0.25820630981921305, "grad_norm": 2.8864113929758104, "learning_rate": 8.69586348716304e-06, "loss": 0.4351, "step": 3642 }, { "epoch": 0.2582772066643034, "grad_norm": 2.95423676244574, "learning_rate": 8.695090086337725e-06, "loss": 0.4568, "step": 3643 }, { "epoch": 0.25834810350939386, "grad_norm": 2.8006793281510873, "learning_rate": 8.694316490667488e-06, "loss": 0.4519, "step": 3644 }, { "epoch": 0.25841900035448423, "grad_norm": 2.7226730856222567, "learning_rate": 8.693542700193124e-06, "loss": 0.4429, "step": 3645 }, { "epoch": 0.2584898971995746, "grad_norm": 11.570439207671628, "learning_rate": 8.692768714955434e-06, "loss": 0.4329, "step": 3646 }, { "epoch": 0.25856079404466503, "grad_norm": 9.3808955176479, "learning_rate": 8.69199453499523e-06, "loss": 0.4244, "step": 3647 }, { "epoch": 0.2586316908897554, "grad_norm": 4.724084068931552, "learning_rate": 8.691220160353336e-06, "loss": 0.4225, "step": 3648 }, { "epoch": 0.2587025877348458, "grad_norm": 4.475815645386767, "learning_rate": 8.690445591070587e-06, "loss": 0.4182, "step": 3649 }, { "epoch": 0.2587734845799362, "grad_norm": 2.9590558825864055, "learning_rate": 8.689670827187825e-06, "loss": 0.4552, "step": 3650 }, { "epoch": 0.2588443814250266, "grad_norm": 3.2401270675717764, "learning_rate": 8.688895868745904e-06, "loss": 0.4295, "step": 3651 }, { "epoch": 0.25891527827011696, "grad_norm": 2.760405761259781, "learning_rate": 8.688120715785687e-06, "loss": 0.4286, "step": 3652 }, { "epoch": 0.2589861751152074, "grad_norm": 2.171953133607803, "learning_rate": 8.687345368348051e-06, "loss": 0.4029, "step": 3653 }, { "epoch": 0.25905707196029776, "grad_norm": 2.157226606786255, "learning_rate": 8.686569826473878e-06, "loss": 0.4223, "step": 3654 }, { "epoch": 0.2591279688053882, "grad_norm": 4.453432145660338, "learning_rate": 8.685794090204065e-06, "loss": 0.5068, "step": 3655 }, { "epoch": 0.25919886565047856, "grad_norm": 2.1870540953835826, "learning_rate": 8.685018159579516e-06, "loss": 0.4836, "step": 3656 }, { "epoch": 0.25926976249556893, "grad_norm": 3.5091561795227024, "learning_rate": 8.684242034641148e-06, "loss": 0.4789, "step": 3657 }, { "epoch": 0.25934065934065936, "grad_norm": 4.502164483722967, "learning_rate": 8.683465715429886e-06, "loss": 0.497, "step": 3658 }, { "epoch": 0.25941155618574974, "grad_norm": 4.487915585422095, "learning_rate": 8.682689201986662e-06, "loss": 0.4181, "step": 3659 }, { "epoch": 0.2594824530308401, "grad_norm": 2.8742432023849376, "learning_rate": 8.68191249435243e-06, "loss": 0.4233, "step": 3660 }, { "epoch": 0.25955334987593054, "grad_norm": 3.3894841463676624, "learning_rate": 8.681135592568139e-06, "loss": 0.4765, "step": 3661 }, { "epoch": 0.2596242467210209, "grad_norm": 2.7510475239506906, "learning_rate": 8.68035849667476e-06, "loss": 0.4959, "step": 3662 }, { "epoch": 0.2596951435661113, "grad_norm": 2.642047689210476, "learning_rate": 8.679581206713265e-06, "loss": 0.3962, "step": 3663 }, { "epoch": 0.2597660404112017, "grad_norm": 3.272219155328443, "learning_rate": 8.678803722724647e-06, "loss": 0.3695, "step": 3664 }, { "epoch": 0.2598369372562921, "grad_norm": 4.754564247773268, "learning_rate": 8.678026044749901e-06, "loss": 0.4804, "step": 3665 }, { "epoch": 0.25990783410138246, "grad_norm": 3.7121620306821876, "learning_rate": 8.677248172830036e-06, "loss": 0.3961, "step": 3666 }, { "epoch": 0.2599787309464729, "grad_norm": 2.2909729309633664, "learning_rate": 8.676470107006064e-06, "loss": 0.4218, "step": 3667 }, { "epoch": 0.26004962779156326, "grad_norm": 2.655433437388311, "learning_rate": 8.675691847319018e-06, "loss": 0.4675, "step": 3668 }, { "epoch": 0.2601205246366537, "grad_norm": 3.1283978048282592, "learning_rate": 8.674913393809938e-06, "loss": 0.4194, "step": 3669 }, { "epoch": 0.26019142148174407, "grad_norm": 2.8075898595829054, "learning_rate": 8.674134746519869e-06, "loss": 0.4447, "step": 3670 }, { "epoch": 0.26026231832683444, "grad_norm": 2.534865428003742, "learning_rate": 8.67335590548987e-06, "loss": 0.453, "step": 3671 }, { "epoch": 0.26033321517192487, "grad_norm": 4.129017302283529, "learning_rate": 8.67257687076101e-06, "loss": 0.4933, "step": 3672 }, { "epoch": 0.26040411201701524, "grad_norm": 4.0887579425960485, "learning_rate": 8.671797642374368e-06, "loss": 0.4769, "step": 3673 }, { "epoch": 0.2604750088621056, "grad_norm": 1.129884655171148, "learning_rate": 8.671018220371035e-06, "loss": 0.7012, "step": 3674 }, { "epoch": 0.26054590570719605, "grad_norm": 1.7552452082332932, "learning_rate": 8.67023860479211e-06, "loss": 0.4426, "step": 3675 }, { "epoch": 0.2606168025522864, "grad_norm": 4.754729933110968, "learning_rate": 8.6694587956787e-06, "loss": 0.4583, "step": 3676 }, { "epoch": 0.2606876993973768, "grad_norm": 2.4354752521192573, "learning_rate": 8.668678793071928e-06, "loss": 0.4137, "step": 3677 }, { "epoch": 0.2607585962424672, "grad_norm": 1.9357892565660888, "learning_rate": 8.667898597012923e-06, "loss": 0.4063, "step": 3678 }, { "epoch": 0.2608294930875576, "grad_norm": 1.167004240921997, "learning_rate": 8.667118207542825e-06, "loss": 0.6727, "step": 3679 }, { "epoch": 0.260900389932648, "grad_norm": 2.3922556654724865, "learning_rate": 8.666337624702785e-06, "loss": 0.3975, "step": 3680 }, { "epoch": 0.2609712867777384, "grad_norm": 1.7761881139899045, "learning_rate": 8.665556848533963e-06, "loss": 0.4335, "step": 3681 }, { "epoch": 0.26104218362282877, "grad_norm": 2.520695433288951, "learning_rate": 8.664775879077532e-06, "loss": 0.4259, "step": 3682 }, { "epoch": 0.2611130804679192, "grad_norm": 3.281305458779872, "learning_rate": 8.66399471637467e-06, "loss": 0.4518, "step": 3683 }, { "epoch": 0.2611839773130096, "grad_norm": 3.3939595200311516, "learning_rate": 8.66321336046657e-06, "loss": 0.5016, "step": 3684 }, { "epoch": 0.26125487415809995, "grad_norm": 2.2119478880314403, "learning_rate": 8.662431811394436e-06, "loss": 0.4364, "step": 3685 }, { "epoch": 0.2613257710031904, "grad_norm": 4.833809508739736, "learning_rate": 8.661650069199477e-06, "loss": 0.3965, "step": 3686 }, { "epoch": 0.26139666784828075, "grad_norm": 2.5053828974172463, "learning_rate": 8.660868133922913e-06, "loss": 0.4175, "step": 3687 }, { "epoch": 0.2614675646933711, "grad_norm": 3.859696119314345, "learning_rate": 8.66008600560598e-06, "loss": 0.4657, "step": 3688 }, { "epoch": 0.26153846153846155, "grad_norm": 1.1924269477219702, "learning_rate": 8.659303684289915e-06, "loss": 0.7359, "step": 3689 }, { "epoch": 0.2616093583835519, "grad_norm": 3.264072496058802, "learning_rate": 8.658521170015976e-06, "loss": 0.42, "step": 3690 }, { "epoch": 0.2616802552286423, "grad_norm": 3.694564223284315, "learning_rate": 8.657738462825425e-06, "loss": 0.4389, "step": 3691 }, { "epoch": 0.26175115207373273, "grad_norm": 3.0802198056408465, "learning_rate": 8.656955562759531e-06, "loss": 0.4185, "step": 3692 }, { "epoch": 0.2618220489188231, "grad_norm": 6.0168593694554255, "learning_rate": 8.656172469859578e-06, "loss": 0.4372, "step": 3693 }, { "epoch": 0.26189294576391353, "grad_norm": 1.6607797177259809, "learning_rate": 8.655389184166862e-06, "loss": 0.4254, "step": 3694 }, { "epoch": 0.2619638426090039, "grad_norm": 3.922057440498304, "learning_rate": 8.654605705722683e-06, "loss": 0.4444, "step": 3695 }, { "epoch": 0.2620347394540943, "grad_norm": 1.0921047956398025, "learning_rate": 8.653822034568355e-06, "loss": 0.6396, "step": 3696 }, { "epoch": 0.2621056362991847, "grad_norm": 4.049377853890335, "learning_rate": 8.653038170745205e-06, "loss": 0.4157, "step": 3697 }, { "epoch": 0.2621765331442751, "grad_norm": 4.884696689254304, "learning_rate": 8.652254114294563e-06, "loss": 0.416, "step": 3698 }, { "epoch": 0.26224742998936545, "grad_norm": 4.055681438290379, "learning_rate": 8.651469865257773e-06, "loss": 0.4516, "step": 3699 }, { "epoch": 0.2623183268344559, "grad_norm": 2.45594771475814, "learning_rate": 8.650685423676189e-06, "loss": 0.4304, "step": 3700 }, { "epoch": 0.26238922367954626, "grad_norm": 2.377888782333126, "learning_rate": 8.649900789591178e-06, "loss": 0.4382, "step": 3701 }, { "epoch": 0.26246012052463663, "grad_norm": 3.046351751438687, "learning_rate": 8.649115963044112e-06, "loss": 0.4345, "step": 3702 }, { "epoch": 0.26253101736972706, "grad_norm": 2.5657049204965903, "learning_rate": 8.648330944076376e-06, "loss": 0.4445, "step": 3703 }, { "epoch": 0.26260191421481743, "grad_norm": 2.081133454973232, "learning_rate": 8.647545732729366e-06, "loss": 0.4013, "step": 3704 }, { "epoch": 0.2626728110599078, "grad_norm": 3.477035057362018, "learning_rate": 8.646760329044484e-06, "loss": 0.3967, "step": 3705 }, { "epoch": 0.26274370790499824, "grad_norm": 1.0492044535982048, "learning_rate": 8.645974733063147e-06, "loss": 0.6853, "step": 3706 }, { "epoch": 0.2628146047500886, "grad_norm": 2.9462536011033964, "learning_rate": 8.645188944826777e-06, "loss": 0.4517, "step": 3707 }, { "epoch": 0.26288550159517904, "grad_norm": 2.377065429270448, "learning_rate": 8.644402964376814e-06, "loss": 0.4761, "step": 3708 }, { "epoch": 0.2629563984402694, "grad_norm": 2.6576051312948605, "learning_rate": 8.643616791754701e-06, "loss": 0.377, "step": 3709 }, { "epoch": 0.2630272952853598, "grad_norm": 1.9201917423116779, "learning_rate": 8.642830427001894e-06, "loss": 0.413, "step": 3710 }, { "epoch": 0.2630981921304502, "grad_norm": 1.668931078296906, "learning_rate": 8.642043870159858e-06, "loss": 0.3782, "step": 3711 }, { "epoch": 0.2631690889755406, "grad_norm": 2.181341402042442, "learning_rate": 8.641257121270068e-06, "loss": 0.4145, "step": 3712 }, { "epoch": 0.26323998582063096, "grad_norm": 4.984982981797072, "learning_rate": 8.64047018037401e-06, "loss": 0.4222, "step": 3713 }, { "epoch": 0.2633108826657214, "grad_norm": 2.986039358681062, "learning_rate": 8.639683047513183e-06, "loss": 0.4576, "step": 3714 }, { "epoch": 0.26338177951081176, "grad_norm": 1.9573749130554141, "learning_rate": 8.63889572272909e-06, "loss": 0.4075, "step": 3715 }, { "epoch": 0.26345267635590214, "grad_norm": 1.9901635875135926, "learning_rate": 8.638108206063248e-06, "loss": 0.4326, "step": 3716 }, { "epoch": 0.26352357320099257, "grad_norm": 4.600099902112117, "learning_rate": 8.637320497557183e-06, "loss": 0.4154, "step": 3717 }, { "epoch": 0.26359447004608294, "grad_norm": 4.1212617427832825, "learning_rate": 8.636532597252432e-06, "loss": 0.3926, "step": 3718 }, { "epoch": 0.26366536689117337, "grad_norm": 4.210305864202268, "learning_rate": 8.635744505190543e-06, "loss": 0.4701, "step": 3719 }, { "epoch": 0.26373626373626374, "grad_norm": 11.498646340504171, "learning_rate": 8.634956221413068e-06, "loss": 0.4171, "step": 3720 }, { "epoch": 0.2638071605813541, "grad_norm": 2.3183810192444936, "learning_rate": 8.634167745961579e-06, "loss": 0.3904, "step": 3721 }, { "epoch": 0.26387805742644455, "grad_norm": 2.1344535519738765, "learning_rate": 8.633379078877651e-06, "loss": 0.4528, "step": 3722 }, { "epoch": 0.2639489542715349, "grad_norm": 2.7881610448156677, "learning_rate": 8.632590220202871e-06, "loss": 0.4311, "step": 3723 }, { "epoch": 0.2640198511166253, "grad_norm": 3.457285283442153, "learning_rate": 8.631801169978835e-06, "loss": 0.4194, "step": 3724 }, { "epoch": 0.2640907479617157, "grad_norm": 2.1400202034382168, "learning_rate": 8.631011928247153e-06, "loss": 0.4227, "step": 3725 }, { "epoch": 0.2641616448068061, "grad_norm": 3.378956830421491, "learning_rate": 8.630222495049439e-06, "loss": 0.474, "step": 3726 }, { "epoch": 0.26423254165189647, "grad_norm": 2.583670274289406, "learning_rate": 8.629432870427323e-06, "loss": 0.3754, "step": 3727 }, { "epoch": 0.2643034384969869, "grad_norm": 4.188787692129581, "learning_rate": 8.628643054422441e-06, "loss": 0.4145, "step": 3728 }, { "epoch": 0.26437433534207727, "grad_norm": 2.9702155097386647, "learning_rate": 8.627853047076442e-06, "loss": 0.4572, "step": 3729 }, { "epoch": 0.26444523218716764, "grad_norm": 4.056896653679439, "learning_rate": 8.627062848430982e-06, "loss": 0.45, "step": 3730 }, { "epoch": 0.2645161290322581, "grad_norm": 3.2955058634119343, "learning_rate": 8.62627245852773e-06, "loss": 0.4721, "step": 3731 }, { "epoch": 0.26458702587734845, "grad_norm": 4.080762096859471, "learning_rate": 8.625481877408364e-06, "loss": 0.446, "step": 3732 }, { "epoch": 0.2646579227224389, "grad_norm": 2.1530832713716612, "learning_rate": 8.624691105114569e-06, "loss": 0.4613, "step": 3733 }, { "epoch": 0.26472881956752925, "grad_norm": 23.468728859755334, "learning_rate": 8.623900141688048e-06, "loss": 0.4345, "step": 3734 }, { "epoch": 0.2647997164126196, "grad_norm": 2.600443786737737, "learning_rate": 8.623108987170505e-06, "loss": 0.4786, "step": 3735 }, { "epoch": 0.26487061325771005, "grad_norm": 1.9795921267253636, "learning_rate": 8.62231764160366e-06, "loss": 0.4234, "step": 3736 }, { "epoch": 0.2649415101028004, "grad_norm": 2.813616206487581, "learning_rate": 8.621526105029243e-06, "loss": 0.4238, "step": 3737 }, { "epoch": 0.2650124069478908, "grad_norm": 2.0043249494199147, "learning_rate": 8.620734377488988e-06, "loss": 0.4117, "step": 3738 }, { "epoch": 0.26508330379298123, "grad_norm": 8.334202660708678, "learning_rate": 8.619942459024645e-06, "loss": 0.388, "step": 3739 }, { "epoch": 0.2651542006380716, "grad_norm": 2.4190404476834853, "learning_rate": 8.619150349677974e-06, "loss": 0.3761, "step": 3740 }, { "epoch": 0.265225097483162, "grad_norm": 4.343132124806151, "learning_rate": 8.618358049490745e-06, "loss": 0.4514, "step": 3741 }, { "epoch": 0.2652959943282524, "grad_norm": 2.561663558255462, "learning_rate": 8.617565558504731e-06, "loss": 0.4091, "step": 3742 }, { "epoch": 0.2653668911733428, "grad_norm": 3.1319403001834454, "learning_rate": 8.616772876761725e-06, "loss": 0.405, "step": 3743 }, { "epoch": 0.2654377880184332, "grad_norm": 2.4091728998597306, "learning_rate": 8.615980004303526e-06, "loss": 0.5155, "step": 3744 }, { "epoch": 0.2655086848635236, "grad_norm": 2.8136829770088467, "learning_rate": 8.61518694117194e-06, "loss": 0.3975, "step": 3745 }, { "epoch": 0.26557958170861395, "grad_norm": 2.460285415134826, "learning_rate": 8.614393687408788e-06, "loss": 0.4267, "step": 3746 }, { "epoch": 0.2656504785537044, "grad_norm": 11.21265324417665, "learning_rate": 8.613600243055898e-06, "loss": 0.4283, "step": 3747 }, { "epoch": 0.26572137539879476, "grad_norm": 1.7690645595794525, "learning_rate": 8.612806608155109e-06, "loss": 0.3972, "step": 3748 }, { "epoch": 0.26579227224388513, "grad_norm": 2.64858608013787, "learning_rate": 8.612012782748269e-06, "loss": 0.4641, "step": 3749 }, { "epoch": 0.26586316908897556, "grad_norm": 1.9927703600351905, "learning_rate": 8.61121876687724e-06, "loss": 0.4478, "step": 3750 }, { "epoch": 0.26593406593406593, "grad_norm": 4.376020030568845, "learning_rate": 8.610424560583888e-06, "loss": 0.4365, "step": 3751 }, { "epoch": 0.2660049627791563, "grad_norm": 2.6789003070524777, "learning_rate": 8.609630163910093e-06, "loss": 0.4889, "step": 3752 }, { "epoch": 0.26607585962424674, "grad_norm": 2.4445538377715255, "learning_rate": 8.608835576897744e-06, "loss": 0.4414, "step": 3753 }, { "epoch": 0.2661467564693371, "grad_norm": 2.4816655153367297, "learning_rate": 8.608040799588742e-06, "loss": 0.4791, "step": 3754 }, { "epoch": 0.2662176533144275, "grad_norm": 2.425190047366002, "learning_rate": 8.607245832024993e-06, "loss": 0.4434, "step": 3755 }, { "epoch": 0.2662885501595179, "grad_norm": 2.744858482050589, "learning_rate": 8.606450674248419e-06, "loss": 0.524, "step": 3756 }, { "epoch": 0.2663594470046083, "grad_norm": 4.6426755189737765, "learning_rate": 8.605655326300949e-06, "loss": 0.3667, "step": 3757 }, { "epoch": 0.2664303438496987, "grad_norm": 2.675163286503852, "learning_rate": 8.60485978822452e-06, "loss": 0.4046, "step": 3758 }, { "epoch": 0.2665012406947891, "grad_norm": 2.0045745273036792, "learning_rate": 8.604064060061083e-06, "loss": 0.4338, "step": 3759 }, { "epoch": 0.26657213753987946, "grad_norm": 6.045245799917603, "learning_rate": 8.603268141852598e-06, "loss": 0.4578, "step": 3760 }, { "epoch": 0.2666430343849699, "grad_norm": 16.547725214650065, "learning_rate": 8.602472033641034e-06, "loss": 0.4324, "step": 3761 }, { "epoch": 0.26671393123006026, "grad_norm": 3.512746897402119, "learning_rate": 8.601675735468368e-06, "loss": 0.4709, "step": 3762 }, { "epoch": 0.26678482807515064, "grad_norm": 3.9546712130984587, "learning_rate": 8.600879247376594e-06, "loss": 0.4355, "step": 3763 }, { "epoch": 0.26685572492024107, "grad_norm": 2.929563077185251, "learning_rate": 8.600082569407707e-06, "loss": 0.3807, "step": 3764 }, { "epoch": 0.26692662176533144, "grad_norm": 3.1742700060940434, "learning_rate": 8.599285701603719e-06, "loss": 0.416, "step": 3765 }, { "epoch": 0.2669975186104218, "grad_norm": 2.90196594830788, "learning_rate": 8.598488644006649e-06, "loss": 0.4246, "step": 3766 }, { "epoch": 0.26706841545551224, "grad_norm": 2.288193576858396, "learning_rate": 8.597691396658527e-06, "loss": 0.4386, "step": 3767 }, { "epoch": 0.2671393123006026, "grad_norm": 4.233527464294461, "learning_rate": 8.59689395960139e-06, "loss": 0.4352, "step": 3768 }, { "epoch": 0.267210209145693, "grad_norm": 3.1150584794519958, "learning_rate": 8.59609633287729e-06, "loss": 0.4205, "step": 3769 }, { "epoch": 0.2672811059907834, "grad_norm": 2.4444672216707373, "learning_rate": 8.595298516528286e-06, "loss": 0.4613, "step": 3770 }, { "epoch": 0.2673520028358738, "grad_norm": 4.284834999721448, "learning_rate": 8.594500510596445e-06, "loss": 0.4082, "step": 3771 }, { "epoch": 0.2674228996809642, "grad_norm": 6.3850141014635895, "learning_rate": 8.59370231512385e-06, "loss": 0.45, "step": 3772 }, { "epoch": 0.2674937965260546, "grad_norm": 7.151741242403757, "learning_rate": 8.59290393015259e-06, "loss": 0.4306, "step": 3773 }, { "epoch": 0.26756469337114497, "grad_norm": 2.678621662041746, "learning_rate": 8.592105355724762e-06, "loss": 0.4681, "step": 3774 }, { "epoch": 0.2676355902162354, "grad_norm": 1.6937845436067718, "learning_rate": 8.591306591882476e-06, "loss": 0.4247, "step": 3775 }, { "epoch": 0.26770648706132577, "grad_norm": 3.6497287641557286, "learning_rate": 8.590507638667853e-06, "loss": 0.4427, "step": 3776 }, { "epoch": 0.26777738390641614, "grad_norm": 3.0757416510745106, "learning_rate": 8.589708496123023e-06, "loss": 0.4396, "step": 3777 }, { "epoch": 0.2678482807515066, "grad_norm": 3.7709467148631526, "learning_rate": 8.588909164290124e-06, "loss": 0.4491, "step": 3778 }, { "epoch": 0.26791917759659695, "grad_norm": 2.7406492402986564, "learning_rate": 8.588109643211304e-06, "loss": 0.4569, "step": 3779 }, { "epoch": 0.2679900744416873, "grad_norm": 3.9733605540201937, "learning_rate": 8.587309932928725e-06, "loss": 0.4441, "step": 3780 }, { "epoch": 0.26806097128677775, "grad_norm": 2.0363823505720475, "learning_rate": 8.586510033484555e-06, "loss": 0.4252, "step": 3781 }, { "epoch": 0.2681318681318681, "grad_norm": 2.7780849683521884, "learning_rate": 8.585709944920973e-06, "loss": 0.4878, "step": 3782 }, { "epoch": 0.26820276497695855, "grad_norm": 2.5456065820618448, "learning_rate": 8.58490966728017e-06, "loss": 0.455, "step": 3783 }, { "epoch": 0.2682736618220489, "grad_norm": 2.917472599524592, "learning_rate": 8.584109200604342e-06, "loss": 0.4543, "step": 3784 }, { "epoch": 0.2683445586671393, "grad_norm": 4.808490932632921, "learning_rate": 8.5833085449357e-06, "loss": 0.4373, "step": 3785 }, { "epoch": 0.26841545551222973, "grad_norm": 11.497252462733627, "learning_rate": 8.582507700316467e-06, "loss": 0.4405, "step": 3786 }, { "epoch": 0.2684863523573201, "grad_norm": 7.002956367274451, "learning_rate": 8.581706666788865e-06, "loss": 0.4507, "step": 3787 }, { "epoch": 0.2685572492024105, "grad_norm": 141.21822299567268, "learning_rate": 8.580905444395138e-06, "loss": 0.3866, "step": 3788 }, { "epoch": 0.2686281460475009, "grad_norm": 1.9910456884646626, "learning_rate": 8.580104033177534e-06, "loss": 0.4158, "step": 3789 }, { "epoch": 0.2686990428925913, "grad_norm": 3.9418494610663415, "learning_rate": 8.579302433178312e-06, "loss": 0.4651, "step": 3790 }, { "epoch": 0.26876993973768165, "grad_norm": 4.960054600138471, "learning_rate": 8.578500644439738e-06, "loss": 0.4371, "step": 3791 }, { "epoch": 0.2688408365827721, "grad_norm": 5.28961114149398, "learning_rate": 8.577698667004098e-06, "loss": 0.4324, "step": 3792 }, { "epoch": 0.26891173342786245, "grad_norm": 4.030108705935263, "learning_rate": 8.576896500913674e-06, "loss": 0.3975, "step": 3793 }, { "epoch": 0.2689826302729528, "grad_norm": 2.264577145114201, "learning_rate": 8.576094146210768e-06, "loss": 0.4034, "step": 3794 }, { "epoch": 0.26905352711804326, "grad_norm": 2.0096426589216483, "learning_rate": 8.575291602937687e-06, "loss": 0.4762, "step": 3795 }, { "epoch": 0.26912442396313363, "grad_norm": 2.337014821330302, "learning_rate": 8.574488871136753e-06, "loss": 0.4804, "step": 3796 }, { "epoch": 0.26919532080822406, "grad_norm": 3.528964898672145, "learning_rate": 8.57368595085029e-06, "loss": 0.4282, "step": 3797 }, { "epoch": 0.26926621765331443, "grad_norm": 3.6778127109260774, "learning_rate": 8.57288284212064e-06, "loss": 0.4224, "step": 3798 }, { "epoch": 0.2693371144984048, "grad_norm": 17.39469933743398, "learning_rate": 8.572079544990151e-06, "loss": 0.4399, "step": 3799 }, { "epoch": 0.26940801134349524, "grad_norm": 2.904357369824627, "learning_rate": 8.571276059501181e-06, "loss": 0.4736, "step": 3800 }, { "epoch": 0.2694789081885856, "grad_norm": 4.395965158824309, "learning_rate": 8.5704723856961e-06, "loss": 0.4339, "step": 3801 }, { "epoch": 0.269549805033676, "grad_norm": 3.028702823680094, "learning_rate": 8.569668523617283e-06, "loss": 0.483, "step": 3802 }, { "epoch": 0.2696207018787664, "grad_norm": 4.733308722308152, "learning_rate": 8.568864473307122e-06, "loss": 0.4197, "step": 3803 }, { "epoch": 0.2696915987238568, "grad_norm": 4.9521989596454254, "learning_rate": 8.568060234808012e-06, "loss": 0.4403, "step": 3804 }, { "epoch": 0.26976249556894716, "grad_norm": 3.7697126419085905, "learning_rate": 8.567255808162362e-06, "loss": 0.4386, "step": 3805 }, { "epoch": 0.2698333924140376, "grad_norm": 2.9664133407067426, "learning_rate": 8.56645119341259e-06, "loss": 0.4469, "step": 3806 }, { "epoch": 0.26990428925912796, "grad_norm": 2.240469651004504, "learning_rate": 8.565646390601126e-06, "loss": 0.4229, "step": 3807 }, { "epoch": 0.2699751861042184, "grad_norm": 3.4085385064545504, "learning_rate": 8.564841399770407e-06, "loss": 0.4837, "step": 3808 }, { "epoch": 0.27004608294930876, "grad_norm": 7.642906646718317, "learning_rate": 8.56403622096288e-06, "loss": 0.4319, "step": 3809 }, { "epoch": 0.27011697979439914, "grad_norm": 2.5658490655696666, "learning_rate": 8.563230854221001e-06, "loss": 0.462, "step": 3810 }, { "epoch": 0.27018787663948957, "grad_norm": 7.8763331273809225, "learning_rate": 8.56242529958724e-06, "loss": 0.4214, "step": 3811 }, { "epoch": 0.27025877348457994, "grad_norm": 3.549998220501411, "learning_rate": 8.561619557104072e-06, "loss": 0.4597, "step": 3812 }, { "epoch": 0.2703296703296703, "grad_norm": 6.521418289612187, "learning_rate": 8.56081362681399e-06, "loss": 0.3646, "step": 3813 }, { "epoch": 0.27040056717476074, "grad_norm": 5.0885545613346155, "learning_rate": 8.560007508759484e-06, "loss": 0.4317, "step": 3814 }, { "epoch": 0.2704714640198511, "grad_norm": 7.882821529533713, "learning_rate": 8.559201202983067e-06, "loss": 0.4718, "step": 3815 }, { "epoch": 0.2705423608649415, "grad_norm": 3.372252630248183, "learning_rate": 8.558394709527255e-06, "loss": 0.4021, "step": 3816 }, { "epoch": 0.2706132577100319, "grad_norm": 4.978331413401471, "learning_rate": 8.557588028434572e-06, "loss": 0.4476, "step": 3817 }, { "epoch": 0.2706841545551223, "grad_norm": 3.1648215565388544, "learning_rate": 8.556781159747558e-06, "loss": 0.3888, "step": 3818 }, { "epoch": 0.27075505140021267, "grad_norm": 3.921865836056083, "learning_rate": 8.555974103508757e-06, "loss": 0.4525, "step": 3819 }, { "epoch": 0.2708259482453031, "grad_norm": 2.7407350418696006, "learning_rate": 8.555166859760728e-06, "loss": 0.3854, "step": 3820 }, { "epoch": 0.27089684509039347, "grad_norm": 2.085011905424081, "learning_rate": 8.554359428546039e-06, "loss": 0.417, "step": 3821 }, { "epoch": 0.2709677419354839, "grad_norm": 4.426582796051045, "learning_rate": 8.553551809907263e-06, "loss": 0.4206, "step": 3822 }, { "epoch": 0.27103863878057427, "grad_norm": 23.742911905411614, "learning_rate": 8.552744003886987e-06, "loss": 0.5209, "step": 3823 }, { "epoch": 0.27110953562566464, "grad_norm": 2.6204651817206814, "learning_rate": 8.551936010527808e-06, "loss": 0.421, "step": 3824 }, { "epoch": 0.2711804324707551, "grad_norm": 5.334202205434184, "learning_rate": 8.551127829872333e-06, "loss": 0.4904, "step": 3825 }, { "epoch": 0.27125132931584545, "grad_norm": 3.334921383362823, "learning_rate": 8.550319461963178e-06, "loss": 0.4226, "step": 3826 }, { "epoch": 0.2713222261609358, "grad_norm": 3.8854394758347723, "learning_rate": 8.549510906842966e-06, "loss": 0.4724, "step": 3827 }, { "epoch": 0.27139312300602625, "grad_norm": 2.9428788128758705, "learning_rate": 8.548702164554335e-06, "loss": 0.4085, "step": 3828 }, { "epoch": 0.2714640198511166, "grad_norm": 4.143645044855115, "learning_rate": 8.54789323513993e-06, "loss": 0.4028, "step": 3829 }, { "epoch": 0.271534916696207, "grad_norm": 2.585834226920265, "learning_rate": 8.547084118642408e-06, "loss": 0.4222, "step": 3830 }, { "epoch": 0.2716058135412974, "grad_norm": 4.780987149503425, "learning_rate": 8.546274815104432e-06, "loss": 0.4776, "step": 3831 }, { "epoch": 0.2716767103863878, "grad_norm": 2.533621261511198, "learning_rate": 8.545465324568679e-06, "loss": 0.4243, "step": 3832 }, { "epoch": 0.27174760723147817, "grad_norm": 2.772030916902765, "learning_rate": 8.544655647077833e-06, "loss": 0.4243, "step": 3833 }, { "epoch": 0.2718185040765686, "grad_norm": 1.1388888745525125, "learning_rate": 8.543845782674589e-06, "loss": 0.6943, "step": 3834 }, { "epoch": 0.271889400921659, "grad_norm": 3.586497335824366, "learning_rate": 8.543035731401652e-06, "loss": 0.4011, "step": 3835 }, { "epoch": 0.2719602977667494, "grad_norm": 4.696918657089864, "learning_rate": 8.542225493301737e-06, "loss": 0.4584, "step": 3836 }, { "epoch": 0.2720311946118398, "grad_norm": 3.9757906006226844, "learning_rate": 8.541415068417567e-06, "loss": 0.4621, "step": 3837 }, { "epoch": 0.27210209145693015, "grad_norm": 3.164544137096463, "learning_rate": 8.540604456791878e-06, "loss": 0.3614, "step": 3838 }, { "epoch": 0.2721729883020206, "grad_norm": 3.441741298511541, "learning_rate": 8.539793658467414e-06, "loss": 0.4428, "step": 3839 }, { "epoch": 0.27224388514711095, "grad_norm": 2.2595133640544476, "learning_rate": 8.53898267348693e-06, "loss": 0.4011, "step": 3840 }, { "epoch": 0.2723147819922013, "grad_norm": 2.292057958250701, "learning_rate": 8.538171501893186e-06, "loss": 0.4688, "step": 3841 }, { "epoch": 0.27238567883729176, "grad_norm": 1.1546999592023288, "learning_rate": 8.53736014372896e-06, "loss": 0.6167, "step": 3842 }, { "epoch": 0.27245657568238213, "grad_norm": 3.074658814383589, "learning_rate": 8.536548599037032e-06, "loss": 0.4499, "step": 3843 }, { "epoch": 0.2725274725274725, "grad_norm": 2.016415363143357, "learning_rate": 8.535736867860199e-06, "loss": 0.413, "step": 3844 }, { "epoch": 0.27259836937256293, "grad_norm": 2.9408372984214677, "learning_rate": 8.534924950241258e-06, "loss": 0.4616, "step": 3845 }, { "epoch": 0.2726692662176533, "grad_norm": 4.4439965772834205, "learning_rate": 8.534112846223031e-06, "loss": 0.4087, "step": 3846 }, { "epoch": 0.27274016306274373, "grad_norm": 2.802247794053311, "learning_rate": 8.533300555848335e-06, "loss": 0.4411, "step": 3847 }, { "epoch": 0.2728110599078341, "grad_norm": 3.7124450268636595, "learning_rate": 8.532488079160003e-06, "loss": 0.3992, "step": 3848 }, { "epoch": 0.2728819567529245, "grad_norm": 3.5418208903001727, "learning_rate": 8.531675416200879e-06, "loss": 0.4245, "step": 3849 }, { "epoch": 0.2729528535980149, "grad_norm": 4.103344930181323, "learning_rate": 8.530862567013814e-06, "loss": 0.4974, "step": 3850 }, { "epoch": 0.2730237504431053, "grad_norm": 2.0018140462091947, "learning_rate": 8.530049531641671e-06, "loss": 0.4945, "step": 3851 }, { "epoch": 0.27309464728819566, "grad_norm": 2.2667329513488115, "learning_rate": 8.529236310127322e-06, "loss": 0.4659, "step": 3852 }, { "epoch": 0.2731655441332861, "grad_norm": 3.182285825940515, "learning_rate": 8.528422902513649e-06, "loss": 0.5245, "step": 3853 }, { "epoch": 0.27323644097837646, "grad_norm": 2.649419957894227, "learning_rate": 8.527609308843543e-06, "loss": 0.432, "step": 3854 }, { "epoch": 0.27330733782346683, "grad_norm": 4.646178869117364, "learning_rate": 8.526795529159906e-06, "loss": 0.4015, "step": 3855 }, { "epoch": 0.27337823466855726, "grad_norm": 2.0982966936956853, "learning_rate": 8.525981563505647e-06, "loss": 0.435, "step": 3856 }, { "epoch": 0.27344913151364764, "grad_norm": 4.603198511316982, "learning_rate": 8.525167411923692e-06, "loss": 0.4861, "step": 3857 }, { "epoch": 0.273520028358738, "grad_norm": 1.502398302934728, "learning_rate": 8.524353074456967e-06, "loss": 0.4351, "step": 3858 }, { "epoch": 0.27359092520382844, "grad_norm": 11.678293100439575, "learning_rate": 8.523538551148414e-06, "loss": 0.4723, "step": 3859 }, { "epoch": 0.2736618220489188, "grad_norm": 3.813185560948735, "learning_rate": 8.522723842040982e-06, "loss": 0.4244, "step": 3860 }, { "epoch": 0.27373271889400924, "grad_norm": 4.935741465450288, "learning_rate": 8.521908947177636e-06, "loss": 0.4838, "step": 3861 }, { "epoch": 0.2738036157390996, "grad_norm": 3.9001533219947753, "learning_rate": 8.521093866601343e-06, "loss": 0.4204, "step": 3862 }, { "epoch": 0.27387451258419, "grad_norm": 3.1182423875646506, "learning_rate": 8.520278600355083e-06, "loss": 0.4656, "step": 3863 }, { "epoch": 0.2739454094292804, "grad_norm": 2.871656794397924, "learning_rate": 8.519463148481844e-06, "loss": 0.4278, "step": 3864 }, { "epoch": 0.2740163062743708, "grad_norm": 2.0350632855597675, "learning_rate": 8.518647511024627e-06, "loss": 0.4334, "step": 3865 }, { "epoch": 0.27408720311946116, "grad_norm": 2.9291076044129585, "learning_rate": 8.517831688026442e-06, "loss": 0.3858, "step": 3866 }, { "epoch": 0.2741580999645516, "grad_norm": 4.515458435106886, "learning_rate": 8.517015679530306e-06, "loss": 0.4648, "step": 3867 }, { "epoch": 0.27422899680964197, "grad_norm": 5.2046531509272045, "learning_rate": 8.51619948557925e-06, "loss": 0.5052, "step": 3868 }, { "epoch": 0.27429989365473234, "grad_norm": 4.315449612199264, "learning_rate": 8.51538310621631e-06, "loss": 0.4014, "step": 3869 }, { "epoch": 0.27437079049982277, "grad_norm": 2.247054721094288, "learning_rate": 8.514566541484536e-06, "loss": 0.4209, "step": 3870 }, { "epoch": 0.27444168734491314, "grad_norm": 1.1002227880412412, "learning_rate": 8.513749791426988e-06, "loss": 0.7229, "step": 3871 }, { "epoch": 0.2745125841900036, "grad_norm": 2.583195899211733, "learning_rate": 8.512932856086729e-06, "loss": 0.4154, "step": 3872 }, { "epoch": 0.27458348103509395, "grad_norm": 2.329857317754452, "learning_rate": 8.51211573550684e-06, "loss": 0.3782, "step": 3873 }, { "epoch": 0.2746543778801843, "grad_norm": 5.2693778917404845, "learning_rate": 8.511298429730408e-06, "loss": 0.4547, "step": 3874 }, { "epoch": 0.27472527472527475, "grad_norm": 2.550096159593497, "learning_rate": 8.51048093880053e-06, "loss": 0.4294, "step": 3875 }, { "epoch": 0.2747961715703651, "grad_norm": 6.118971965717874, "learning_rate": 8.509663262760311e-06, "loss": 0.4561, "step": 3876 }, { "epoch": 0.2748670684154555, "grad_norm": 2.3386256899248834, "learning_rate": 8.50884540165287e-06, "loss": 0.3801, "step": 3877 }, { "epoch": 0.2749379652605459, "grad_norm": 2.323275821919223, "learning_rate": 8.508027355521333e-06, "loss": 0.4469, "step": 3878 }, { "epoch": 0.2750088621056363, "grad_norm": 2.604827472158324, "learning_rate": 8.507209124408837e-06, "loss": 0.449, "step": 3879 }, { "epoch": 0.27507975895072667, "grad_norm": 3.1558436592754844, "learning_rate": 8.506390708358525e-06, "loss": 0.4596, "step": 3880 }, { "epoch": 0.2751506557958171, "grad_norm": 2.2183646275799656, "learning_rate": 8.505572107413556e-06, "loss": 0.4528, "step": 3881 }, { "epoch": 0.2752215526409075, "grad_norm": 3.3456258064711597, "learning_rate": 8.504753321617094e-06, "loss": 0.4167, "step": 3882 }, { "epoch": 0.27529244948599785, "grad_norm": 4.465866427022707, "learning_rate": 8.503934351012315e-06, "loss": 0.4736, "step": 3883 }, { "epoch": 0.2753633463310883, "grad_norm": 3.0730865625229535, "learning_rate": 8.503115195642403e-06, "loss": 0.4365, "step": 3884 }, { "epoch": 0.27543424317617865, "grad_norm": 3.3756999688915044, "learning_rate": 8.502295855550551e-06, "loss": 0.424, "step": 3885 }, { "epoch": 0.2755051400212691, "grad_norm": 3.4301100172348264, "learning_rate": 8.501476330779965e-06, "loss": 0.4348, "step": 3886 }, { "epoch": 0.27557603686635945, "grad_norm": 3.710036798919165, "learning_rate": 8.500656621373862e-06, "loss": 0.4754, "step": 3887 }, { "epoch": 0.2756469337114498, "grad_norm": 2.9959619183263007, "learning_rate": 8.499836727375464e-06, "loss": 0.391, "step": 3888 }, { "epoch": 0.27571783055654026, "grad_norm": 2.466331420101046, "learning_rate": 8.499016648828003e-06, "loss": 0.4455, "step": 3889 }, { "epoch": 0.27578872740163063, "grad_norm": 3.7098448368436734, "learning_rate": 8.498196385774722e-06, "loss": 0.4102, "step": 3890 }, { "epoch": 0.275859624246721, "grad_norm": 3.0107949519461905, "learning_rate": 8.497375938258876e-06, "loss": 0.4111, "step": 3891 }, { "epoch": 0.27593052109181143, "grad_norm": 2.6504258184021983, "learning_rate": 8.496555306323729e-06, "loss": 0.421, "step": 3892 }, { "epoch": 0.2760014179369018, "grad_norm": 4.102295477788429, "learning_rate": 8.49573449001255e-06, "loss": 0.4308, "step": 3893 }, { "epoch": 0.2760723147819922, "grad_norm": 2.3997608911587345, "learning_rate": 8.494913489368624e-06, "loss": 0.4182, "step": 3894 }, { "epoch": 0.2761432116270826, "grad_norm": 3.4581438031209104, "learning_rate": 8.494092304435243e-06, "loss": 0.4983, "step": 3895 }, { "epoch": 0.276214108472173, "grad_norm": 2.9077210212163895, "learning_rate": 8.493270935255706e-06, "loss": 0.4273, "step": 3896 }, { "epoch": 0.27628500531726335, "grad_norm": 4.0458363507877255, "learning_rate": 8.492449381873326e-06, "loss": 0.453, "step": 3897 }, { "epoch": 0.2763559021623538, "grad_norm": 2.935195280076421, "learning_rate": 8.491627644331426e-06, "loss": 0.4295, "step": 3898 }, { "epoch": 0.27642679900744416, "grad_norm": 7.898313133071877, "learning_rate": 8.490805722673336e-06, "loss": 0.3796, "step": 3899 }, { "epoch": 0.2764976958525346, "grad_norm": 2.263991680623728, "learning_rate": 8.489983616942395e-06, "loss": 0.433, "step": 3900 }, { "epoch": 0.27656859269762496, "grad_norm": 4.3315432754004055, "learning_rate": 8.489161327181955e-06, "loss": 0.4437, "step": 3901 }, { "epoch": 0.27663948954271533, "grad_norm": 4.002567589445557, "learning_rate": 8.488338853435371e-06, "loss": 0.465, "step": 3902 }, { "epoch": 0.27671038638780576, "grad_norm": 3.6067799862716576, "learning_rate": 8.487516195746021e-06, "loss": 0.4486, "step": 3903 }, { "epoch": 0.27678128323289614, "grad_norm": 2.9328424482588504, "learning_rate": 8.486693354157278e-06, "loss": 0.4759, "step": 3904 }, { "epoch": 0.2768521800779865, "grad_norm": 4.191057290542392, "learning_rate": 8.485870328712535e-06, "loss": 0.4492, "step": 3905 }, { "epoch": 0.27692307692307694, "grad_norm": 3.1649023734285806, "learning_rate": 8.485047119455187e-06, "loss": 0.4166, "step": 3906 }, { "epoch": 0.2769939737681673, "grad_norm": 3.2661569527621674, "learning_rate": 8.484223726428646e-06, "loss": 0.413, "step": 3907 }, { "epoch": 0.2770648706132577, "grad_norm": 3.6337800389825783, "learning_rate": 8.483400149676328e-06, "loss": 0.4236, "step": 3908 }, { "epoch": 0.2771357674583481, "grad_norm": 3.0945062444838345, "learning_rate": 8.482576389241662e-06, "loss": 0.4054, "step": 3909 }, { "epoch": 0.2772066643034385, "grad_norm": 2.4362340644174623, "learning_rate": 8.481752445168084e-06, "loss": 0.3963, "step": 3910 }, { "epoch": 0.2772775611485289, "grad_norm": 2.401779320147754, "learning_rate": 8.48092831749904e-06, "loss": 0.3887, "step": 3911 }, { "epoch": 0.2773484579936193, "grad_norm": 4.356334810475466, "learning_rate": 8.480104006277991e-06, "loss": 0.3842, "step": 3912 }, { "epoch": 0.27741935483870966, "grad_norm": 3.342064228516198, "learning_rate": 8.479279511548402e-06, "loss": 0.4909, "step": 3913 }, { "epoch": 0.2774902516838001, "grad_norm": 4.188158338660009, "learning_rate": 8.478454833353749e-06, "loss": 0.4181, "step": 3914 }, { "epoch": 0.27756114852889047, "grad_norm": 5.615831254875536, "learning_rate": 8.477629971737516e-06, "loss": 0.4167, "step": 3915 }, { "epoch": 0.27763204537398084, "grad_norm": 5.290953509664018, "learning_rate": 8.476804926743202e-06, "loss": 0.4645, "step": 3916 }, { "epoch": 0.27770294221907127, "grad_norm": 3.8064538871499956, "learning_rate": 8.475979698414309e-06, "loss": 0.4809, "step": 3917 }, { "epoch": 0.27777383906416164, "grad_norm": 2.7287593396530814, "learning_rate": 8.475154286794354e-06, "loss": 0.429, "step": 3918 }, { "epoch": 0.277844735909252, "grad_norm": 1.9562741245453772, "learning_rate": 8.47432869192686e-06, "loss": 0.4415, "step": 3919 }, { "epoch": 0.27791563275434245, "grad_norm": 3.4511437880001914, "learning_rate": 8.473502913855365e-06, "loss": 0.4047, "step": 3920 }, { "epoch": 0.2779865295994328, "grad_norm": 1.1496130820350263, "learning_rate": 8.472676952623409e-06, "loss": 0.7595, "step": 3921 }, { "epoch": 0.2780574264445232, "grad_norm": 4.316055097520317, "learning_rate": 8.471850808274545e-06, "loss": 0.4581, "step": 3922 }, { "epoch": 0.2781283232896136, "grad_norm": 1.0967822686109978, "learning_rate": 8.471024480852338e-06, "loss": 0.6533, "step": 3923 }, { "epoch": 0.278199220134704, "grad_norm": 6.164638690543583, "learning_rate": 8.470197970400364e-06, "loss": 0.5133, "step": 3924 }, { "epoch": 0.2782701169797944, "grad_norm": 4.407056588562634, "learning_rate": 8.469371276962199e-06, "loss": 0.4458, "step": 3925 }, { "epoch": 0.2783410138248848, "grad_norm": 4.449260251865846, "learning_rate": 8.46854440058144e-06, "loss": 0.4435, "step": 3926 }, { "epoch": 0.27841191066997517, "grad_norm": 4.915006275959993, "learning_rate": 8.467717341301687e-06, "loss": 0.4682, "step": 3927 }, { "epoch": 0.2784828075150656, "grad_norm": 4.692343790571802, "learning_rate": 8.466890099166551e-06, "loss": 0.4531, "step": 3928 }, { "epoch": 0.278553704360156, "grad_norm": 3.1475889112271704, "learning_rate": 8.466062674219657e-06, "loss": 0.5425, "step": 3929 }, { "epoch": 0.27862460120524635, "grad_norm": 3.1740649194056276, "learning_rate": 8.46523506650463e-06, "loss": 0.4511, "step": 3930 }, { "epoch": 0.2786954980503368, "grad_norm": 6.625569389479653, "learning_rate": 8.464407276065113e-06, "loss": 0.4696, "step": 3931 }, { "epoch": 0.27876639489542715, "grad_norm": 2.7069715072103215, "learning_rate": 8.463579302944758e-06, "loss": 0.4481, "step": 3932 }, { "epoch": 0.2788372917405175, "grad_norm": 2.6752077639282357, "learning_rate": 8.46275114718722e-06, "loss": 0.3953, "step": 3933 }, { "epoch": 0.27890818858560795, "grad_norm": 4.663283542378205, "learning_rate": 8.461922808836172e-06, "loss": 0.469, "step": 3934 }, { "epoch": 0.2789790854306983, "grad_norm": 3.535469653618836, "learning_rate": 8.461094287935292e-06, "loss": 0.4312, "step": 3935 }, { "epoch": 0.27904998227578875, "grad_norm": 6.583337446195491, "learning_rate": 8.460265584528269e-06, "loss": 0.4547, "step": 3936 }, { "epoch": 0.27912087912087913, "grad_norm": 30.60307428723508, "learning_rate": 8.4594366986588e-06, "loss": 0.4552, "step": 3937 }, { "epoch": 0.2791917759659695, "grad_norm": 2.608134936721008, "learning_rate": 8.458607630370594e-06, "loss": 0.4165, "step": 3938 }, { "epoch": 0.27926267281105993, "grad_norm": 7.808247957603168, "learning_rate": 8.45777837970737e-06, "loss": 0.4759, "step": 3939 }, { "epoch": 0.2793335696561503, "grad_norm": 3.0353294658078456, "learning_rate": 8.456948946712848e-06, "loss": 0.4503, "step": 3940 }, { "epoch": 0.2794044665012407, "grad_norm": 3.30720380284568, "learning_rate": 8.456119331430772e-06, "loss": 0.4053, "step": 3941 }, { "epoch": 0.2794753633463311, "grad_norm": 3.343219321050844, "learning_rate": 8.455289533904886e-06, "loss": 0.4061, "step": 3942 }, { "epoch": 0.2795462601914215, "grad_norm": 4.019862511742824, "learning_rate": 8.454459554178944e-06, "loss": 0.4659, "step": 3943 }, { "epoch": 0.27961715703651185, "grad_norm": 4.122219216359014, "learning_rate": 8.453629392296715e-06, "loss": 0.4549, "step": 3944 }, { "epoch": 0.2796880538816023, "grad_norm": 2.8764639928221256, "learning_rate": 8.452799048301971e-06, "loss": 0.4146, "step": 3945 }, { "epoch": 0.27975895072669266, "grad_norm": 3.6674618846105758, "learning_rate": 8.451968522238498e-06, "loss": 0.4025, "step": 3946 }, { "epoch": 0.27982984757178303, "grad_norm": 2.460524404393393, "learning_rate": 8.45113781415009e-06, "loss": 0.3984, "step": 3947 }, { "epoch": 0.27990074441687346, "grad_norm": 3.3899503708247685, "learning_rate": 8.45030692408055e-06, "loss": 0.4222, "step": 3948 }, { "epoch": 0.27997164126196383, "grad_norm": 10.095620021865196, "learning_rate": 8.449475852073695e-06, "loss": 0.4116, "step": 3949 }, { "epoch": 0.28004253810705426, "grad_norm": 6.408037997897546, "learning_rate": 8.448644598173344e-06, "loss": 0.4666, "step": 3950 }, { "epoch": 0.28011343495214464, "grad_norm": 4.274276333542408, "learning_rate": 8.447813162423331e-06, "loss": 0.4581, "step": 3951 }, { "epoch": 0.280184331797235, "grad_norm": 3.35177434000022, "learning_rate": 8.446981544867498e-06, "loss": 0.3913, "step": 3952 }, { "epoch": 0.28025522864232544, "grad_norm": 8.400164624522157, "learning_rate": 8.446149745549699e-06, "loss": 0.496, "step": 3953 }, { "epoch": 0.2803261254874158, "grad_norm": 2.4450134714168286, "learning_rate": 8.445317764513791e-06, "loss": 0.3816, "step": 3954 }, { "epoch": 0.2803970223325062, "grad_norm": 2.9215705369781424, "learning_rate": 8.444485601803649e-06, "loss": 0.419, "step": 3955 }, { "epoch": 0.2804679191775966, "grad_norm": 2.321231591464734, "learning_rate": 8.443653257463152e-06, "loss": 0.4191, "step": 3956 }, { "epoch": 0.280538816022687, "grad_norm": 1.6930276641348934, "learning_rate": 8.44282073153619e-06, "loss": 0.5071, "step": 3957 }, { "epoch": 0.28060971286777736, "grad_norm": 1.942470880786037, "learning_rate": 8.441988024066663e-06, "loss": 0.411, "step": 3958 }, { "epoch": 0.2806806097128678, "grad_norm": 2.270232568880339, "learning_rate": 8.44115513509848e-06, "loss": 0.4353, "step": 3959 }, { "epoch": 0.28075150655795816, "grad_norm": 1.9176485175004556, "learning_rate": 8.440322064675561e-06, "loss": 0.446, "step": 3960 }, { "epoch": 0.28082240340304854, "grad_norm": 1.5129980613919478, "learning_rate": 8.439488812841834e-06, "loss": 0.7695, "step": 3961 }, { "epoch": 0.28089330024813897, "grad_norm": 3.7502077041558324, "learning_rate": 8.438655379641236e-06, "loss": 0.4725, "step": 3962 }, { "epoch": 0.28096419709322934, "grad_norm": 5.2510598730013305, "learning_rate": 8.437821765117716e-06, "loss": 0.4079, "step": 3963 }, { "epoch": 0.28103509393831977, "grad_norm": 2.0644733764230434, "learning_rate": 8.43698796931523e-06, "loss": 0.3943, "step": 3964 }, { "epoch": 0.28110599078341014, "grad_norm": 2.431021417274822, "learning_rate": 8.436153992277744e-06, "loss": 0.4379, "step": 3965 }, { "epoch": 0.2811768876285005, "grad_norm": 4.402042172354927, "learning_rate": 8.435319834049236e-06, "loss": 0.4389, "step": 3966 }, { "epoch": 0.28124778447359094, "grad_norm": 2.2054584595544346, "learning_rate": 8.434485494673691e-06, "loss": 0.4335, "step": 3967 }, { "epoch": 0.2813186813186813, "grad_norm": 2.945459605739018, "learning_rate": 8.433650974195105e-06, "loss": 0.4631, "step": 3968 }, { "epoch": 0.2813895781637717, "grad_norm": 2.024999363615156, "learning_rate": 8.432816272657482e-06, "loss": 0.4511, "step": 3969 }, { "epoch": 0.2814604750088621, "grad_norm": 19.987504174152416, "learning_rate": 8.431981390104837e-06, "loss": 0.3904, "step": 3970 }, { "epoch": 0.2815313718539525, "grad_norm": 2.0411169768593895, "learning_rate": 8.431146326581192e-06, "loss": 0.4506, "step": 3971 }, { "epoch": 0.28160226869904287, "grad_norm": 2.6410894839449393, "learning_rate": 8.430311082130584e-06, "loss": 0.4772, "step": 3972 }, { "epoch": 0.2816731655441333, "grad_norm": 2.982866657239814, "learning_rate": 8.429475656797055e-06, "loss": 0.4849, "step": 3973 }, { "epoch": 0.28174406238922367, "grad_norm": 4.530267584754893, "learning_rate": 8.428640050624657e-06, "loss": 0.412, "step": 3974 }, { "epoch": 0.2818149592343141, "grad_norm": 5.767208078082445, "learning_rate": 8.42780426365745e-06, "loss": 0.4451, "step": 3975 }, { "epoch": 0.2818858560794045, "grad_norm": 4.077763321222555, "learning_rate": 8.42696829593951e-06, "loss": 0.4903, "step": 3976 }, { "epoch": 0.28195675292449485, "grad_norm": 2.661337710850634, "learning_rate": 8.426132147514914e-06, "loss": 0.4153, "step": 3977 }, { "epoch": 0.2820276497695853, "grad_norm": 2.5383667978850553, "learning_rate": 8.425295818427757e-06, "loss": 0.4597, "step": 3978 }, { "epoch": 0.28209854661467565, "grad_norm": 7.058366132699663, "learning_rate": 8.424459308722137e-06, "loss": 0.4502, "step": 3979 }, { "epoch": 0.282169443459766, "grad_norm": 4.693815653243089, "learning_rate": 8.423622618442162e-06, "loss": 0.4702, "step": 3980 }, { "epoch": 0.28224034030485645, "grad_norm": 2.17356098756551, "learning_rate": 8.422785747631954e-06, "loss": 0.4376, "step": 3981 }, { "epoch": 0.2823112371499468, "grad_norm": 2.7602301653994705, "learning_rate": 8.421948696335643e-06, "loss": 0.4232, "step": 3982 }, { "epoch": 0.2823821339950372, "grad_norm": 3.034380488663498, "learning_rate": 8.421111464597363e-06, "loss": 0.4343, "step": 3983 }, { "epoch": 0.28245303084012763, "grad_norm": 3.4887855747415166, "learning_rate": 8.420274052461265e-06, "loss": 0.4391, "step": 3984 }, { "epoch": 0.282523927685218, "grad_norm": 3.6427283038242395, "learning_rate": 8.419436459971504e-06, "loss": 0.4199, "step": 3985 }, { "epoch": 0.2825948245303084, "grad_norm": 3.7235986687719693, "learning_rate": 8.418598687172252e-06, "loss": 0.4408, "step": 3986 }, { "epoch": 0.2826657213753988, "grad_norm": 2.742356047516637, "learning_rate": 8.41776073410768e-06, "loss": 0.3746, "step": 3987 }, { "epoch": 0.2827366182204892, "grad_norm": 4.070320533350951, "learning_rate": 8.416922600821975e-06, "loss": 0.4439, "step": 3988 }, { "epoch": 0.2828075150655796, "grad_norm": 2.2909508250099364, "learning_rate": 8.416084287359334e-06, "loss": 0.3815, "step": 3989 }, { "epoch": 0.28287841191067, "grad_norm": 3.4026080749679113, "learning_rate": 8.415245793763962e-06, "loss": 0.4219, "step": 3990 }, { "epoch": 0.28294930875576035, "grad_norm": 4.159178384416251, "learning_rate": 8.41440712008007e-06, "loss": 0.3945, "step": 3991 }, { "epoch": 0.2830202056008508, "grad_norm": 2.4960078855061454, "learning_rate": 8.413568266351888e-06, "loss": 0.4156, "step": 3992 }, { "epoch": 0.28309110244594116, "grad_norm": 2.397357539772105, "learning_rate": 8.412729232623644e-06, "loss": 0.4266, "step": 3993 }, { "epoch": 0.28316199929103153, "grad_norm": 3.6150493584236325, "learning_rate": 8.411890018939583e-06, "loss": 0.4007, "step": 3994 }, { "epoch": 0.28323289613612196, "grad_norm": 2.8825956726956456, "learning_rate": 8.411050625343957e-06, "loss": 0.4305, "step": 3995 }, { "epoch": 0.28330379298121233, "grad_norm": 2.4522420147949546, "learning_rate": 8.410211051881027e-06, "loss": 0.3673, "step": 3996 }, { "epoch": 0.2833746898263027, "grad_norm": 4.658225896591436, "learning_rate": 8.409371298595067e-06, "loss": 0.3897, "step": 3997 }, { "epoch": 0.28344558667139313, "grad_norm": 3.292899987530094, "learning_rate": 8.408531365530357e-06, "loss": 0.4306, "step": 3998 }, { "epoch": 0.2835164835164835, "grad_norm": 3.1049638734161884, "learning_rate": 8.407691252731183e-06, "loss": 0.4392, "step": 3999 }, { "epoch": 0.28358738036157394, "grad_norm": 3.206466448527993, "learning_rate": 8.40685096024185e-06, "loss": 0.418, "step": 4000 }, { "epoch": 0.2836582772066643, "grad_norm": 3.3290619829614303, "learning_rate": 8.406010488106666e-06, "loss": 0.4411, "step": 4001 }, { "epoch": 0.2837291740517547, "grad_norm": 3.6444085798675045, "learning_rate": 8.40516983636995e-06, "loss": 0.4095, "step": 4002 }, { "epoch": 0.2838000708968451, "grad_norm": 6.1578113744505965, "learning_rate": 8.404329005076029e-06, "loss": 0.3785, "step": 4003 }, { "epoch": 0.2838709677419355, "grad_norm": 4.5146620305862335, "learning_rate": 8.40348799426924e-06, "loss": 0.4463, "step": 4004 }, { "epoch": 0.28394186458702586, "grad_norm": 3.584262585900306, "learning_rate": 8.402646803993932e-06, "loss": 0.4512, "step": 4005 }, { "epoch": 0.2840127614321163, "grad_norm": 3.3662287344288844, "learning_rate": 8.401805434294461e-06, "loss": 0.485, "step": 4006 }, { "epoch": 0.28408365827720666, "grad_norm": 3.020759401831044, "learning_rate": 8.400963885215192e-06, "loss": 0.4113, "step": 4007 }, { "epoch": 0.28415455512229704, "grad_norm": 14.512622059994339, "learning_rate": 8.400122156800503e-06, "loss": 0.5073, "step": 4008 }, { "epoch": 0.28422545196738747, "grad_norm": 3.0214040804708766, "learning_rate": 8.399280249094776e-06, "loss": 0.4001, "step": 4009 }, { "epoch": 0.28429634881247784, "grad_norm": 3.040088326005994, "learning_rate": 8.398438162142407e-06, "loss": 0.4428, "step": 4010 }, { "epoch": 0.2843672456575682, "grad_norm": 1.8622130400425836, "learning_rate": 8.397595895987803e-06, "loss": 0.4439, "step": 4011 }, { "epoch": 0.28443814250265864, "grad_norm": 6.837282789233108, "learning_rate": 8.396753450675371e-06, "loss": 0.4486, "step": 4012 }, { "epoch": 0.284509039347749, "grad_norm": 2.927901124217217, "learning_rate": 8.395910826249538e-06, "loss": 0.4514, "step": 4013 }, { "epoch": 0.28457993619283944, "grad_norm": 4.065169623242783, "learning_rate": 8.395068022754735e-06, "loss": 0.4647, "step": 4014 }, { "epoch": 0.2846508330379298, "grad_norm": 6.646001112536041, "learning_rate": 8.394225040235403e-06, "loss": 0.4425, "step": 4015 }, { "epoch": 0.2847217298830202, "grad_norm": 3.9172942581367347, "learning_rate": 8.393381878735996e-06, "loss": 0.4343, "step": 4016 }, { "epoch": 0.2847926267281106, "grad_norm": 4.283974085539611, "learning_rate": 8.39253853830097e-06, "loss": 0.4254, "step": 4017 }, { "epoch": 0.284863523573201, "grad_norm": 3.5395110039862674, "learning_rate": 8.391695018974798e-06, "loss": 0.461, "step": 4018 }, { "epoch": 0.28493442041829137, "grad_norm": 9.958131571622115, "learning_rate": 8.39085132080196e-06, "loss": 0.4123, "step": 4019 }, { "epoch": 0.2850053172633818, "grad_norm": 4.229149269633057, "learning_rate": 8.390007443826941e-06, "loss": 0.5155, "step": 4020 }, { "epoch": 0.28507621410847217, "grad_norm": 1.7625124812977064, "learning_rate": 8.389163388094245e-06, "loss": 0.4239, "step": 4021 }, { "epoch": 0.28514711095356254, "grad_norm": 2.956845065800002, "learning_rate": 8.388319153648376e-06, "loss": 0.4099, "step": 4022 }, { "epoch": 0.285218007798653, "grad_norm": 3.0193079957465305, "learning_rate": 8.387474740533853e-06, "loss": 0.4608, "step": 4023 }, { "epoch": 0.28528890464374335, "grad_norm": 3.6013857459545853, "learning_rate": 8.386630148795199e-06, "loss": 0.4588, "step": 4024 }, { "epoch": 0.2853598014888337, "grad_norm": 2.9297159470523484, "learning_rate": 8.385785378476954e-06, "loss": 0.3987, "step": 4025 }, { "epoch": 0.28543069833392415, "grad_norm": 2.9015282347256606, "learning_rate": 8.38494042962366e-06, "loss": 0.4425, "step": 4026 }, { "epoch": 0.2855015951790145, "grad_norm": 3.246118731861232, "learning_rate": 8.384095302279875e-06, "loss": 0.4366, "step": 4027 }, { "epoch": 0.28557249202410495, "grad_norm": 9.681420530501898, "learning_rate": 8.383249996490161e-06, "loss": 0.4498, "step": 4028 }, { "epoch": 0.2856433888691953, "grad_norm": 4.320086397444856, "learning_rate": 8.382404512299093e-06, "loss": 0.3818, "step": 4029 }, { "epoch": 0.2857142857142857, "grad_norm": 3.4186786285737707, "learning_rate": 8.381558849751253e-06, "loss": 0.3951, "step": 4030 }, { "epoch": 0.2857851825593761, "grad_norm": 3.687511936488752, "learning_rate": 8.380713008891235e-06, "loss": 0.4728, "step": 4031 }, { "epoch": 0.2858560794044665, "grad_norm": 2.281760211595193, "learning_rate": 8.379866989763638e-06, "loss": 0.4148, "step": 4032 }, { "epoch": 0.2859269762495569, "grad_norm": 3.0959056356464556, "learning_rate": 8.379020792413077e-06, "loss": 0.4304, "step": 4033 }, { "epoch": 0.2859978730946473, "grad_norm": 3.5131709327764677, "learning_rate": 8.378174416884168e-06, "loss": 0.4251, "step": 4034 }, { "epoch": 0.2860687699397377, "grad_norm": 2.3339415265819445, "learning_rate": 8.377327863221545e-06, "loss": 0.4289, "step": 4035 }, { "epoch": 0.28613966678482805, "grad_norm": 3.138895496203069, "learning_rate": 8.376481131469845e-06, "loss": 0.4126, "step": 4036 }, { "epoch": 0.2862105636299185, "grad_norm": 3.3013917381770153, "learning_rate": 8.37563422167372e-06, "loss": 0.4168, "step": 4037 }, { "epoch": 0.28628146047500885, "grad_norm": 3.130079477453383, "learning_rate": 8.374787133877824e-06, "loss": 0.4271, "step": 4038 }, { "epoch": 0.2863523573200993, "grad_norm": 5.174715566966407, "learning_rate": 8.373939868126825e-06, "loss": 0.4601, "step": 4039 }, { "epoch": 0.28642325416518966, "grad_norm": 3.3043869100098644, "learning_rate": 8.373092424465406e-06, "loss": 0.4268, "step": 4040 }, { "epoch": 0.28649415101028003, "grad_norm": 3.8571303203117706, "learning_rate": 8.372244802938245e-06, "loss": 0.451, "step": 4041 }, { "epoch": 0.28656504785537046, "grad_norm": 4.496001154471186, "learning_rate": 8.371397003590043e-06, "loss": 0.4594, "step": 4042 }, { "epoch": 0.28663594470046083, "grad_norm": 5.456481055904411, "learning_rate": 8.370549026465505e-06, "loss": 0.4572, "step": 4043 }, { "epoch": 0.2867068415455512, "grad_norm": 2.609387069186749, "learning_rate": 8.369700871609343e-06, "loss": 0.4659, "step": 4044 }, { "epoch": 0.28677773839064163, "grad_norm": 7.350052055821908, "learning_rate": 8.368852539066282e-06, "loss": 0.458, "step": 4045 }, { "epoch": 0.286848635235732, "grad_norm": 2.80448383525971, "learning_rate": 8.368004028881053e-06, "loss": 0.466, "step": 4046 }, { "epoch": 0.2869195320808224, "grad_norm": 2.9948031442879683, "learning_rate": 8.367155341098405e-06, "loss": 0.3631, "step": 4047 }, { "epoch": 0.2869904289259128, "grad_norm": 5.755142670367085, "learning_rate": 8.366306475763085e-06, "loss": 0.4471, "step": 4048 }, { "epoch": 0.2870613257710032, "grad_norm": 3.036192341768454, "learning_rate": 8.365457432919853e-06, "loss": 0.4559, "step": 4049 }, { "epoch": 0.28713222261609356, "grad_norm": 2.226512745781533, "learning_rate": 8.364608212613482e-06, "loss": 0.4059, "step": 4050 }, { "epoch": 0.287203119461184, "grad_norm": 5.079592547243241, "learning_rate": 8.363758814888755e-06, "loss": 0.4098, "step": 4051 }, { "epoch": 0.28727401630627436, "grad_norm": 2.109541817962501, "learning_rate": 8.362909239790456e-06, "loss": 0.4129, "step": 4052 }, { "epoch": 0.2873449131513648, "grad_norm": 1.8656926684031838, "learning_rate": 8.362059487363385e-06, "loss": 0.3875, "step": 4053 }, { "epoch": 0.28741580999645516, "grad_norm": 3.758829286045512, "learning_rate": 8.361209557652352e-06, "loss": 0.4704, "step": 4054 }, { "epoch": 0.28748670684154554, "grad_norm": 3.967403877988882, "learning_rate": 8.360359450702174e-06, "loss": 0.4554, "step": 4055 }, { "epoch": 0.28755760368663597, "grad_norm": 3.13142680685087, "learning_rate": 8.359509166557677e-06, "loss": 0.4474, "step": 4056 }, { "epoch": 0.28762850053172634, "grad_norm": 3.1255105739437017, "learning_rate": 8.358658705263698e-06, "loss": 0.4132, "step": 4057 }, { "epoch": 0.2876993973768167, "grad_norm": 3.073654238011467, "learning_rate": 8.357808066865082e-06, "loss": 0.4381, "step": 4058 }, { "epoch": 0.28777029422190714, "grad_norm": 2.9213376249358087, "learning_rate": 8.356957251406682e-06, "loss": 0.3778, "step": 4059 }, { "epoch": 0.2878411910669975, "grad_norm": 2.8188043690179945, "learning_rate": 8.356106258933365e-06, "loss": 0.4337, "step": 4060 }, { "epoch": 0.2879120879120879, "grad_norm": 2.948687310492613, "learning_rate": 8.355255089490005e-06, "loss": 0.4784, "step": 4061 }, { "epoch": 0.2879829847571783, "grad_norm": 5.204301771501341, "learning_rate": 8.35440374312148e-06, "loss": 0.46, "step": 4062 }, { "epoch": 0.2880538816022687, "grad_norm": 3.1177448991729997, "learning_rate": 8.353552219872687e-06, "loss": 0.4053, "step": 4063 }, { "epoch": 0.2881247784473591, "grad_norm": 3.465007315064814, "learning_rate": 8.352700519788526e-06, "loss": 0.4343, "step": 4064 }, { "epoch": 0.2881956752924495, "grad_norm": 3.6455735265863556, "learning_rate": 8.351848642913907e-06, "loss": 0.4346, "step": 4065 }, { "epoch": 0.28826657213753987, "grad_norm": 2.974798415848635, "learning_rate": 8.350996589293751e-06, "loss": 0.4385, "step": 4066 }, { "epoch": 0.2883374689826303, "grad_norm": 1.9908772475694807, "learning_rate": 8.350144358972987e-06, "loss": 0.4274, "step": 4067 }, { "epoch": 0.28840836582772067, "grad_norm": 3.896538797300659, "learning_rate": 8.349291951996554e-06, "loss": 0.3858, "step": 4068 }, { "epoch": 0.28847926267281104, "grad_norm": 4.8507838742312535, "learning_rate": 8.3484393684094e-06, "loss": 0.4137, "step": 4069 }, { "epoch": 0.28855015951790147, "grad_norm": 2.126975478917356, "learning_rate": 8.347586608256483e-06, "loss": 0.4329, "step": 4070 }, { "epoch": 0.28862105636299185, "grad_norm": 1.9567859754449963, "learning_rate": 8.346733671582769e-06, "loss": 0.3904, "step": 4071 }, { "epoch": 0.2886919532080822, "grad_norm": 6.479619326275139, "learning_rate": 8.345880558433234e-06, "loss": 0.4679, "step": 4072 }, { "epoch": 0.28876285005317265, "grad_norm": 3.9525757323349304, "learning_rate": 8.345027268852861e-06, "loss": 0.4053, "step": 4073 }, { "epoch": 0.288833746898263, "grad_norm": 6.7303800900304775, "learning_rate": 8.34417380288665e-06, "loss": 0.4547, "step": 4074 }, { "epoch": 0.2889046437433534, "grad_norm": 2.517985219147756, "learning_rate": 8.3433201605796e-06, "loss": 0.4403, "step": 4075 }, { "epoch": 0.2889755405884438, "grad_norm": 4.569545772959858, "learning_rate": 8.342466341976727e-06, "loss": 0.3974, "step": 4076 }, { "epoch": 0.2890464374335342, "grad_norm": 3.2393667551278273, "learning_rate": 8.341612347123054e-06, "loss": 0.4377, "step": 4077 }, { "epoch": 0.2891173342786246, "grad_norm": 6.913881479960455, "learning_rate": 8.340758176063608e-06, "loss": 0.4124, "step": 4078 }, { "epoch": 0.289188231123715, "grad_norm": 2.1343646271036385, "learning_rate": 8.339903828843436e-06, "loss": 0.4038, "step": 4079 }, { "epoch": 0.2892591279688054, "grad_norm": 2.3063045017709043, "learning_rate": 8.339049305507584e-06, "loss": 0.4405, "step": 4080 }, { "epoch": 0.2893300248138958, "grad_norm": 2.385134803225663, "learning_rate": 8.338194606101115e-06, "loss": 0.4729, "step": 4081 }, { "epoch": 0.2894009216589862, "grad_norm": 2.4423470579398434, "learning_rate": 8.337339730669098e-06, "loss": 0.4789, "step": 4082 }, { "epoch": 0.28947181850407655, "grad_norm": 3.1284787419449067, "learning_rate": 8.336484679256607e-06, "loss": 0.4047, "step": 4083 }, { "epoch": 0.289542715349167, "grad_norm": 4.340791508269043, "learning_rate": 8.335629451908733e-06, "loss": 0.4041, "step": 4084 }, { "epoch": 0.28961361219425735, "grad_norm": 2.2687141440007377, "learning_rate": 8.334774048670573e-06, "loss": 0.415, "step": 4085 }, { "epoch": 0.2896845090393477, "grad_norm": 4.570130854729522, "learning_rate": 8.33391846958723e-06, "loss": 0.3913, "step": 4086 }, { "epoch": 0.28975540588443816, "grad_norm": 2.970667792759305, "learning_rate": 8.333062714703823e-06, "loss": 0.4374, "step": 4087 }, { "epoch": 0.28982630272952853, "grad_norm": 1.8030826922968908, "learning_rate": 8.332206784065474e-06, "loss": 0.4334, "step": 4088 }, { "epoch": 0.2898971995746189, "grad_norm": 12.027824800396571, "learning_rate": 8.33135067771732e-06, "loss": 0.4168, "step": 4089 }, { "epoch": 0.28996809641970933, "grad_norm": 2.362217373450418, "learning_rate": 8.3304943957045e-06, "loss": 0.4082, "step": 4090 }, { "epoch": 0.2900389932647997, "grad_norm": 1.771359529810565, "learning_rate": 8.32963793807217e-06, "loss": 0.3881, "step": 4091 }, { "epoch": 0.29010989010989013, "grad_norm": 3.3859431817571473, "learning_rate": 8.32878130486549e-06, "loss": 0.4617, "step": 4092 }, { "epoch": 0.2901807869549805, "grad_norm": 2.99486075057964, "learning_rate": 8.32792449612963e-06, "loss": 0.4368, "step": 4093 }, { "epoch": 0.2902516838000709, "grad_norm": 3.3683418401606033, "learning_rate": 8.32706751190977e-06, "loss": 0.4493, "step": 4094 }, { "epoch": 0.2903225806451613, "grad_norm": 3.6186950296614513, "learning_rate": 8.326210352251103e-06, "loss": 0.4555, "step": 4095 }, { "epoch": 0.2903934774902517, "grad_norm": 3.094218583922072, "learning_rate": 8.325353017198826e-06, "loss": 0.4974, "step": 4096 }, { "epoch": 0.29046437433534206, "grad_norm": 2.73753247087576, "learning_rate": 8.324495506798143e-06, "loss": 0.4829, "step": 4097 }, { "epoch": 0.2905352711804325, "grad_norm": 10.098531098716677, "learning_rate": 8.323637821094277e-06, "loss": 0.4101, "step": 4098 }, { "epoch": 0.29060616802552286, "grad_norm": 2.888239279485028, "learning_rate": 8.322779960132451e-06, "loss": 0.4639, "step": 4099 }, { "epoch": 0.29067706487061323, "grad_norm": 2.028257790187587, "learning_rate": 8.321921923957903e-06, "loss": 0.4418, "step": 4100 }, { "epoch": 0.29074796171570366, "grad_norm": 4.071352047806492, "learning_rate": 8.321063712615875e-06, "loss": 0.4564, "step": 4101 }, { "epoch": 0.29081885856079404, "grad_norm": 4.96141259603085, "learning_rate": 8.320205326151623e-06, "loss": 0.4926, "step": 4102 }, { "epoch": 0.29088975540588446, "grad_norm": 3.3324356314060077, "learning_rate": 8.31934676461041e-06, "loss": 0.4139, "step": 4103 }, { "epoch": 0.29096065225097484, "grad_norm": 2.9357354920451324, "learning_rate": 8.318488028037508e-06, "loss": 0.4014, "step": 4104 }, { "epoch": 0.2910315490960652, "grad_norm": 2.365301246948629, "learning_rate": 8.317629116478198e-06, "loss": 0.4696, "step": 4105 }, { "epoch": 0.29110244594115564, "grad_norm": 2.8963273351820136, "learning_rate": 8.316770029977776e-06, "loss": 0.4542, "step": 4106 }, { "epoch": 0.291173342786246, "grad_norm": 2.904928247932392, "learning_rate": 8.315910768581537e-06, "loss": 0.4573, "step": 4107 }, { "epoch": 0.2912442396313364, "grad_norm": 3.1236284327148733, "learning_rate": 8.315051332334792e-06, "loss": 0.455, "step": 4108 }, { "epoch": 0.2913151364764268, "grad_norm": 2.9855106454701863, "learning_rate": 8.314191721282859e-06, "loss": 0.4538, "step": 4109 }, { "epoch": 0.2913860333215172, "grad_norm": 2.9272538558981167, "learning_rate": 8.313331935471068e-06, "loss": 0.4012, "step": 4110 }, { "epoch": 0.29145693016660756, "grad_norm": 3.0205258274666216, "learning_rate": 8.312471974944756e-06, "loss": 0.3946, "step": 4111 }, { "epoch": 0.291527827011698, "grad_norm": 3.0476600752149046, "learning_rate": 8.311611839749267e-06, "loss": 0.4586, "step": 4112 }, { "epoch": 0.29159872385678837, "grad_norm": 6.855752902039328, "learning_rate": 8.310751529929959e-06, "loss": 0.451, "step": 4113 }, { "epoch": 0.29166962070187874, "grad_norm": 2.5977152766454856, "learning_rate": 8.309891045532195e-06, "loss": 0.4704, "step": 4114 }, { "epoch": 0.29174051754696917, "grad_norm": 2.7091286781056287, "learning_rate": 8.309030386601351e-06, "loss": 0.4023, "step": 4115 }, { "epoch": 0.29181141439205954, "grad_norm": 2.7569359725860223, "learning_rate": 8.308169553182808e-06, "loss": 0.4234, "step": 4116 }, { "epoch": 0.29188231123714997, "grad_norm": 2.362296117560633, "learning_rate": 8.307308545321959e-06, "loss": 0.3986, "step": 4117 }, { "epoch": 0.29195320808224035, "grad_norm": 2.110139027878358, "learning_rate": 8.306447363064209e-06, "loss": 0.4557, "step": 4118 }, { "epoch": 0.2920241049273307, "grad_norm": 7.179230560523705, "learning_rate": 8.305586006454963e-06, "loss": 0.4612, "step": 4119 }, { "epoch": 0.29209500177242115, "grad_norm": 2.4190237034671886, "learning_rate": 8.304724475539646e-06, "loss": 0.4406, "step": 4120 }, { "epoch": 0.2921658986175115, "grad_norm": 2.565346889773989, "learning_rate": 8.303862770363682e-06, "loss": 0.3781, "step": 4121 }, { "epoch": 0.2922367954626019, "grad_norm": 3.6921399626288465, "learning_rate": 8.303000890972515e-06, "loss": 0.4356, "step": 4122 }, { "epoch": 0.2923076923076923, "grad_norm": 4.374659488692205, "learning_rate": 8.302138837411588e-06, "loss": 0.414, "step": 4123 }, { "epoch": 0.2923785891527827, "grad_norm": 2.0287481549520234, "learning_rate": 8.301276609726361e-06, "loss": 0.404, "step": 4124 }, { "epoch": 0.29244948599787307, "grad_norm": 2.235584981084608, "learning_rate": 8.300414207962297e-06, "loss": 0.4678, "step": 4125 }, { "epoch": 0.2925203828429635, "grad_norm": 5.027528402613566, "learning_rate": 8.299551632164875e-06, "loss": 0.3767, "step": 4126 }, { "epoch": 0.2925912796880539, "grad_norm": 3.111546959784037, "learning_rate": 8.298688882379576e-06, "loss": 0.4012, "step": 4127 }, { "epoch": 0.2926621765331443, "grad_norm": 3.4576064859690554, "learning_rate": 8.297825958651895e-06, "loss": 0.3882, "step": 4128 }, { "epoch": 0.2927330733782347, "grad_norm": 2.5721765011471107, "learning_rate": 8.296962861027334e-06, "loss": 0.4013, "step": 4129 }, { "epoch": 0.29280397022332505, "grad_norm": 4.08075526463606, "learning_rate": 8.296099589551403e-06, "loss": 0.4577, "step": 4130 }, { "epoch": 0.2928748670684155, "grad_norm": 4.292624531660992, "learning_rate": 8.295236144269628e-06, "loss": 0.4352, "step": 4131 }, { "epoch": 0.29294576391350585, "grad_norm": 2.703480959405316, "learning_rate": 8.294372525227532e-06, "loss": 0.4525, "step": 4132 }, { "epoch": 0.2930166607585962, "grad_norm": 7.211977526203766, "learning_rate": 8.293508732470662e-06, "loss": 0.3867, "step": 4133 }, { "epoch": 0.29308755760368665, "grad_norm": 3.3221397650071864, "learning_rate": 8.29264476604456e-06, "loss": 0.4181, "step": 4134 }, { "epoch": 0.29315845444877703, "grad_norm": 2.7063155950935713, "learning_rate": 8.291780625994788e-06, "loss": 0.4314, "step": 4135 }, { "epoch": 0.2932293512938674, "grad_norm": 9.653443542971218, "learning_rate": 8.29091631236691e-06, "loss": 0.447, "step": 4136 }, { "epoch": 0.29330024813895783, "grad_norm": 2.603631157544556, "learning_rate": 8.290051825206504e-06, "loss": 0.4567, "step": 4137 }, { "epoch": 0.2933711449840482, "grad_norm": 4.512963977829055, "learning_rate": 8.289187164559155e-06, "loss": 0.4682, "step": 4138 }, { "epoch": 0.2934420418291386, "grad_norm": 2.042363686406774, "learning_rate": 8.288322330470455e-06, "loss": 0.4526, "step": 4139 }, { "epoch": 0.293512938674229, "grad_norm": 2.82978106255257, "learning_rate": 8.287457322986008e-06, "loss": 0.4236, "step": 4140 }, { "epoch": 0.2935838355193194, "grad_norm": 3.1971469089532536, "learning_rate": 8.286592142151428e-06, "loss": 0.4164, "step": 4141 }, { "epoch": 0.2936547323644098, "grad_norm": 8.186964404980941, "learning_rate": 8.285726788012336e-06, "loss": 0.4382, "step": 4142 }, { "epoch": 0.2937256292095002, "grad_norm": 2.8056478480590075, "learning_rate": 8.284861260614361e-06, "loss": 0.4164, "step": 4143 }, { "epoch": 0.29379652605459056, "grad_norm": 3.1968569923798404, "learning_rate": 8.283995560003147e-06, "loss": 0.4294, "step": 4144 }, { "epoch": 0.293867422899681, "grad_norm": 1.1787475952945972, "learning_rate": 8.283129686224338e-06, "loss": 0.7525, "step": 4145 }, { "epoch": 0.29393831974477136, "grad_norm": 2.4312747582903405, "learning_rate": 8.282263639323596e-06, "loss": 0.4507, "step": 4146 }, { "epoch": 0.29400921658986173, "grad_norm": 3.4444314494190826, "learning_rate": 8.281397419346588e-06, "loss": 0.4248, "step": 4147 }, { "epoch": 0.29408011343495216, "grad_norm": 3.9548269475941913, "learning_rate": 8.280531026338989e-06, "loss": 0.4172, "step": 4148 }, { "epoch": 0.29415101028004254, "grad_norm": 4.829179715551938, "learning_rate": 8.279664460346484e-06, "loss": 0.4226, "step": 4149 }, { "epoch": 0.2942219071251329, "grad_norm": 1.6144162930637218, "learning_rate": 8.278797721414769e-06, "loss": 0.3938, "step": 4150 }, { "epoch": 0.29429280397022334, "grad_norm": 2.3438853486447906, "learning_rate": 8.277930809589549e-06, "loss": 0.4363, "step": 4151 }, { "epoch": 0.2943637008153137, "grad_norm": 1.9237093411428854, "learning_rate": 8.277063724916534e-06, "loss": 0.3972, "step": 4152 }, { "epoch": 0.2944345976604041, "grad_norm": 2.249275563000083, "learning_rate": 8.276196467441449e-06, "loss": 0.4228, "step": 4153 }, { "epoch": 0.2945054945054945, "grad_norm": 2.6864451448222573, "learning_rate": 8.275329037210021e-06, "loss": 0.4568, "step": 4154 }, { "epoch": 0.2945763913505849, "grad_norm": 4.1239635328169, "learning_rate": 8.274461434267995e-06, "loss": 0.4709, "step": 4155 }, { "epoch": 0.2946472881956753, "grad_norm": 2.4018412198267822, "learning_rate": 8.273593658661118e-06, "loss": 0.3677, "step": 4156 }, { "epoch": 0.2947181850407657, "grad_norm": 6.807779939945064, "learning_rate": 8.272725710435147e-06, "loss": 0.3988, "step": 4157 }, { "epoch": 0.29478908188585606, "grad_norm": 1.8609180130268566, "learning_rate": 8.271857589635854e-06, "loss": 0.4477, "step": 4158 }, { "epoch": 0.2948599787309465, "grad_norm": 2.099295582007752, "learning_rate": 8.270989296309014e-06, "loss": 0.4002, "step": 4159 }, { "epoch": 0.29493087557603687, "grad_norm": 2.5320439444567917, "learning_rate": 8.270120830500408e-06, "loss": 0.4034, "step": 4160 }, { "epoch": 0.29500177242112724, "grad_norm": 2.77759576470571, "learning_rate": 8.269252192255836e-06, "loss": 0.4187, "step": 4161 }, { "epoch": 0.29507266926621767, "grad_norm": 1.0138170943497578, "learning_rate": 8.268383381621101e-06, "loss": 0.6749, "step": 4162 }, { "epoch": 0.29514356611130804, "grad_norm": 2.109664525839495, "learning_rate": 8.267514398642016e-06, "loss": 0.4203, "step": 4163 }, { "epoch": 0.2952144629563984, "grad_norm": 5.620090587120745, "learning_rate": 8.266645243364402e-06, "loss": 0.4191, "step": 4164 }, { "epoch": 0.29528535980148884, "grad_norm": 4.655365731474357, "learning_rate": 8.26577591583409e-06, "loss": 0.4065, "step": 4165 }, { "epoch": 0.2953562566465792, "grad_norm": 4.443165257130664, "learning_rate": 8.264906416096925e-06, "loss": 0.4582, "step": 4166 }, { "epoch": 0.29542715349166965, "grad_norm": 3.893976637759605, "learning_rate": 8.26403674419875e-06, "loss": 0.4457, "step": 4167 }, { "epoch": 0.29549805033676, "grad_norm": 6.9943878143241625, "learning_rate": 8.263166900185426e-06, "loss": 0.4605, "step": 4168 }, { "epoch": 0.2955689471818504, "grad_norm": 3.424038428260617, "learning_rate": 8.262296884102822e-06, "loss": 0.4407, "step": 4169 }, { "epoch": 0.2956398440269408, "grad_norm": 1.818562905507247, "learning_rate": 8.261426695996811e-06, "loss": 0.412, "step": 4170 }, { "epoch": 0.2957107408720312, "grad_norm": 2.8339081763381397, "learning_rate": 8.260556335913283e-06, "loss": 0.4247, "step": 4171 }, { "epoch": 0.29578163771712157, "grad_norm": 4.67370907452418, "learning_rate": 8.25968580389813e-06, "loss": 0.4738, "step": 4172 }, { "epoch": 0.295852534562212, "grad_norm": 2.9275531131770234, "learning_rate": 8.258815099997258e-06, "loss": 0.3926, "step": 4173 }, { "epoch": 0.2959234314073024, "grad_norm": 3.88180540886569, "learning_rate": 8.257944224256577e-06, "loss": 0.4475, "step": 4174 }, { "epoch": 0.29599432825239275, "grad_norm": 2.8350792498907236, "learning_rate": 8.25707317672201e-06, "loss": 0.4416, "step": 4175 }, { "epoch": 0.2960652250974832, "grad_norm": 3.56041018362667, "learning_rate": 8.256201957439489e-06, "loss": 0.4632, "step": 4176 }, { "epoch": 0.29613612194257355, "grad_norm": 3.633262961124076, "learning_rate": 8.255330566454953e-06, "loss": 0.4409, "step": 4177 }, { "epoch": 0.2962070187876639, "grad_norm": 3.448396515530614, "learning_rate": 8.254459003814352e-06, "loss": 0.4651, "step": 4178 }, { "epoch": 0.29627791563275435, "grad_norm": 4.212688454444231, "learning_rate": 8.253587269563646e-06, "loss": 0.4409, "step": 4179 }, { "epoch": 0.2963488124778447, "grad_norm": 2.159684889337408, "learning_rate": 8.252715363748796e-06, "loss": 0.399, "step": 4180 }, { "epoch": 0.29641970932293515, "grad_norm": 2.579298695329257, "learning_rate": 8.251843286415784e-06, "loss": 0.4673, "step": 4181 }, { "epoch": 0.29649060616802553, "grad_norm": 3.3571649616416894, "learning_rate": 8.250971037610594e-06, "loss": 0.4607, "step": 4182 }, { "epoch": 0.2965615030131159, "grad_norm": 4.237577470980649, "learning_rate": 8.250098617379219e-06, "loss": 0.4439, "step": 4183 }, { "epoch": 0.29663239985820633, "grad_norm": 1.856017368190985, "learning_rate": 8.249226025767662e-06, "loss": 0.3699, "step": 4184 }, { "epoch": 0.2967032967032967, "grad_norm": 2.3276898408761566, "learning_rate": 8.248353262821938e-06, "loss": 0.4003, "step": 4185 }, { "epoch": 0.2967741935483871, "grad_norm": 2.766600906533573, "learning_rate": 8.247480328588067e-06, "loss": 0.4669, "step": 4186 }, { "epoch": 0.2968450903934775, "grad_norm": 29.71971984872655, "learning_rate": 8.246607223112079e-06, "loss": 0.3887, "step": 4187 }, { "epoch": 0.2969159872385679, "grad_norm": 2.467262885267654, "learning_rate": 8.245733946440016e-06, "loss": 0.443, "step": 4188 }, { "epoch": 0.29698688408365825, "grad_norm": 7.790488227946199, "learning_rate": 8.244860498617923e-06, "loss": 0.4446, "step": 4189 }, { "epoch": 0.2970577809287487, "grad_norm": 2.2870791582806826, "learning_rate": 8.24398687969186e-06, "loss": 0.4631, "step": 4190 }, { "epoch": 0.29712867777383906, "grad_norm": 2.5915770114760965, "learning_rate": 8.243113089707892e-06, "loss": 0.3775, "step": 4191 }, { "epoch": 0.2971995746189295, "grad_norm": 2.107145765346304, "learning_rate": 8.242239128712096e-06, "loss": 0.4206, "step": 4192 }, { "epoch": 0.29727047146401986, "grad_norm": 4.35049760884588, "learning_rate": 8.241364996750556e-06, "loss": 0.464, "step": 4193 }, { "epoch": 0.29734136830911023, "grad_norm": 2.171507935205832, "learning_rate": 8.240490693869366e-06, "loss": 0.4146, "step": 4194 }, { "epoch": 0.29741226515420066, "grad_norm": 2.7242997748885927, "learning_rate": 8.239616220114627e-06, "loss": 0.4563, "step": 4195 }, { "epoch": 0.29748316199929103, "grad_norm": 2.3531286936035913, "learning_rate": 8.238741575532454e-06, "loss": 0.4268, "step": 4196 }, { "epoch": 0.2975540588443814, "grad_norm": 3.4115478415190954, "learning_rate": 8.237866760168964e-06, "loss": 0.4468, "step": 4197 }, { "epoch": 0.29762495568947184, "grad_norm": 2.8142543862506724, "learning_rate": 8.23699177407029e-06, "loss": 0.4412, "step": 4198 }, { "epoch": 0.2976958525345622, "grad_norm": 2.748726746103931, "learning_rate": 8.236116617282568e-06, "loss": 0.4681, "step": 4199 }, { "epoch": 0.2977667493796526, "grad_norm": 1.690788115790645, "learning_rate": 8.235241289851946e-06, "loss": 0.4254, "step": 4200 }, { "epoch": 0.297837646224743, "grad_norm": 3.348941336547608, "learning_rate": 8.234365791824583e-06, "loss": 0.4079, "step": 4201 }, { "epoch": 0.2979085430698334, "grad_norm": 2.0193906905728567, "learning_rate": 8.23349012324664e-06, "loss": 0.4313, "step": 4202 }, { "epoch": 0.29797943991492376, "grad_norm": 5.234512395148607, "learning_rate": 8.2326142841643e-06, "loss": 0.4132, "step": 4203 }, { "epoch": 0.2980503367600142, "grad_norm": 2.9424785020695925, "learning_rate": 8.231738274623738e-06, "loss": 0.4506, "step": 4204 }, { "epoch": 0.29812123360510456, "grad_norm": 3.534611352328448, "learning_rate": 8.23086209467115e-06, "loss": 0.4294, "step": 4205 }, { "epoch": 0.298192130450195, "grad_norm": 4.842309330794614, "learning_rate": 8.22998574435274e-06, "loss": 0.4, "step": 4206 }, { "epoch": 0.29826302729528537, "grad_norm": 3.144278617269253, "learning_rate": 8.229109223714715e-06, "loss": 0.4233, "step": 4207 }, { "epoch": 0.29833392414037574, "grad_norm": 2.583354584886363, "learning_rate": 8.228232532803295e-06, "loss": 0.4313, "step": 4208 }, { "epoch": 0.29840482098546617, "grad_norm": 4.400346055810531, "learning_rate": 8.22735567166471e-06, "loss": 0.3905, "step": 4209 }, { "epoch": 0.29847571783055654, "grad_norm": 3.2649707525526894, "learning_rate": 8.226478640345198e-06, "loss": 0.5091, "step": 4210 }, { "epoch": 0.2985466146756469, "grad_norm": 2.2423451189026258, "learning_rate": 8.225601438891005e-06, "loss": 0.4071, "step": 4211 }, { "epoch": 0.29861751152073734, "grad_norm": 2.1201469271665214, "learning_rate": 8.224724067348384e-06, "loss": 0.4044, "step": 4212 }, { "epoch": 0.2986884083658277, "grad_norm": 3.7291450110396913, "learning_rate": 8.223846525763603e-06, "loss": 0.4406, "step": 4213 }, { "epoch": 0.2987593052109181, "grad_norm": 2.723019738513933, "learning_rate": 8.222968814182934e-06, "loss": 0.4113, "step": 4214 }, { "epoch": 0.2988302020560085, "grad_norm": 2.4362860065215455, "learning_rate": 8.22209093265266e-06, "loss": 0.4248, "step": 4215 }, { "epoch": 0.2989010989010989, "grad_norm": 4.008589274855329, "learning_rate": 8.221212881219071e-06, "loss": 0.457, "step": 4216 }, { "epoch": 0.29897199574618927, "grad_norm": 6.48719905498558, "learning_rate": 8.220334659928468e-06, "loss": 0.4723, "step": 4217 }, { "epoch": 0.2990428925912797, "grad_norm": 7.640560339596575, "learning_rate": 8.219456268827162e-06, "loss": 0.4142, "step": 4218 }, { "epoch": 0.29911378943637007, "grad_norm": 3.7371509678628927, "learning_rate": 8.218577707961469e-06, "loss": 0.4474, "step": 4219 }, { "epoch": 0.2991846862814605, "grad_norm": 3.9001880174434382, "learning_rate": 8.217698977377715e-06, "loss": 0.4411, "step": 4220 }, { "epoch": 0.2992555831265509, "grad_norm": 6.890967346634894, "learning_rate": 8.216820077122238e-06, "loss": 0.4704, "step": 4221 }, { "epoch": 0.29932647997164125, "grad_norm": 4.3442420626315, "learning_rate": 8.215941007241386e-06, "loss": 0.3975, "step": 4222 }, { "epoch": 0.2993973768167317, "grad_norm": 4.162007964454888, "learning_rate": 8.215061767781508e-06, "loss": 0.409, "step": 4223 }, { "epoch": 0.29946827366182205, "grad_norm": 5.433518700634399, "learning_rate": 8.21418235878897e-06, "loss": 0.3916, "step": 4224 }, { "epoch": 0.2995391705069124, "grad_norm": 4.134593698843794, "learning_rate": 8.213302780310141e-06, "loss": 0.3968, "step": 4225 }, { "epoch": 0.29961006735200285, "grad_norm": 3.815951781342616, "learning_rate": 8.212423032391406e-06, "loss": 0.4723, "step": 4226 }, { "epoch": 0.2996809641970932, "grad_norm": 5.276272643877615, "learning_rate": 8.211543115079152e-06, "loss": 0.3962, "step": 4227 }, { "epoch": 0.2997518610421836, "grad_norm": 2.437406983879806, "learning_rate": 8.210663028419776e-06, "loss": 0.4604, "step": 4228 }, { "epoch": 0.299822757887274, "grad_norm": 2.144128836578964, "learning_rate": 8.209782772459692e-06, "loss": 0.4201, "step": 4229 }, { "epoch": 0.2998936547323644, "grad_norm": 3.2856384921645416, "learning_rate": 8.20890234724531e-06, "loss": 0.3925, "step": 4230 }, { "epoch": 0.29996455157745483, "grad_norm": 6.3052626364033, "learning_rate": 8.208021752823057e-06, "loss": 0.467, "step": 4231 }, { "epoch": 0.3000354484225452, "grad_norm": 4.947292972033771, "learning_rate": 8.20714098923937e-06, "loss": 0.4102, "step": 4232 }, { "epoch": 0.3001063452676356, "grad_norm": 2.9252242686071286, "learning_rate": 8.206260056540691e-06, "loss": 0.393, "step": 4233 }, { "epoch": 0.300177242112726, "grad_norm": 3.933021664702246, "learning_rate": 8.20537895477347e-06, "loss": 0.4899, "step": 4234 }, { "epoch": 0.3002481389578164, "grad_norm": 4.5600451068686025, "learning_rate": 8.204497683984174e-06, "loss": 0.4492, "step": 4235 }, { "epoch": 0.30031903580290675, "grad_norm": 2.1723756418702918, "learning_rate": 8.203616244219266e-06, "loss": 0.4444, "step": 4236 }, { "epoch": 0.3003899326479972, "grad_norm": 2.5923304621715744, "learning_rate": 8.202734635525229e-06, "loss": 0.3834, "step": 4237 }, { "epoch": 0.30046082949308756, "grad_norm": 3.6364463409596763, "learning_rate": 8.201852857948551e-06, "loss": 0.4133, "step": 4238 }, { "epoch": 0.30053172633817793, "grad_norm": 5.304030470594581, "learning_rate": 8.200970911535725e-06, "loss": 0.4651, "step": 4239 }, { "epoch": 0.30060262318326836, "grad_norm": 4.061819368267481, "learning_rate": 8.200088796333263e-06, "loss": 0.4702, "step": 4240 }, { "epoch": 0.30067352002835873, "grad_norm": 2.224417719792942, "learning_rate": 8.199206512387676e-06, "loss": 0.4205, "step": 4241 }, { "epoch": 0.3007444168734491, "grad_norm": 4.57843029132075, "learning_rate": 8.198324059745486e-06, "loss": 0.4777, "step": 4242 }, { "epoch": 0.30081531371853953, "grad_norm": 5.373981335457964, "learning_rate": 8.197441438453228e-06, "loss": 0.4047, "step": 4243 }, { "epoch": 0.3008862105636299, "grad_norm": 2.275225811696343, "learning_rate": 8.196558648557443e-06, "loss": 0.4506, "step": 4244 }, { "epoch": 0.30095710740872034, "grad_norm": 2.316937559831753, "learning_rate": 8.195675690104681e-06, "loss": 0.44, "step": 4245 }, { "epoch": 0.3010280042538107, "grad_norm": 2.2685804102772464, "learning_rate": 8.1947925631415e-06, "loss": 0.412, "step": 4246 }, { "epoch": 0.3010989010989011, "grad_norm": 2.6156241467091648, "learning_rate": 8.193909267714467e-06, "loss": 0.4079, "step": 4247 }, { "epoch": 0.3011697979439915, "grad_norm": 4.0104680313082985, "learning_rate": 8.193025803870162e-06, "loss": 0.4539, "step": 4248 }, { "epoch": 0.3012406947890819, "grad_norm": 2.013373326492561, "learning_rate": 8.192142171655168e-06, "loss": 0.4067, "step": 4249 }, { "epoch": 0.30131159163417226, "grad_norm": 6.92337332153215, "learning_rate": 8.191258371116084e-06, "loss": 0.4479, "step": 4250 }, { "epoch": 0.3013824884792627, "grad_norm": 5.737910425953589, "learning_rate": 8.190374402299509e-06, "loss": 0.4223, "step": 4251 }, { "epoch": 0.30145338532435306, "grad_norm": 7.304741024570455, "learning_rate": 8.189490265252055e-06, "loss": 0.4226, "step": 4252 }, { "epoch": 0.30152428216944344, "grad_norm": 5.53395367322932, "learning_rate": 8.188605960020346e-06, "loss": 0.3879, "step": 4253 }, { "epoch": 0.30159517901453387, "grad_norm": 5.633826238228409, "learning_rate": 8.187721486651012e-06, "loss": 0.4528, "step": 4254 }, { "epoch": 0.30166607585962424, "grad_norm": 4.8149631454902675, "learning_rate": 8.186836845190689e-06, "loss": 0.384, "step": 4255 }, { "epoch": 0.30173697270471467, "grad_norm": 1.8770561660441947, "learning_rate": 8.185952035686026e-06, "loss": 0.4035, "step": 4256 }, { "epoch": 0.30180786954980504, "grad_norm": 1.0164157151054236, "learning_rate": 8.185067058183683e-06, "loss": 0.6645, "step": 4257 }, { "epoch": 0.3018787663948954, "grad_norm": 1.9152139586763877, "learning_rate": 8.184181912730322e-06, "loss": 0.4643, "step": 4258 }, { "epoch": 0.30194966323998584, "grad_norm": 14.97874247088619, "learning_rate": 8.183296599372618e-06, "loss": 0.4711, "step": 4259 }, { "epoch": 0.3020205600850762, "grad_norm": 2.6614000796829447, "learning_rate": 8.182411118157256e-06, "loss": 0.4436, "step": 4260 }, { "epoch": 0.3020914569301666, "grad_norm": 2.5637607558898554, "learning_rate": 8.181525469130925e-06, "loss": 0.441, "step": 4261 }, { "epoch": 0.302162353775257, "grad_norm": 2.344255629481317, "learning_rate": 8.18063965234033e-06, "loss": 0.406, "step": 4262 }, { "epoch": 0.3022332506203474, "grad_norm": 2.494561678582594, "learning_rate": 8.179753667832176e-06, "loss": 0.4306, "step": 4263 }, { "epoch": 0.30230414746543777, "grad_norm": 4.3749860393999125, "learning_rate": 8.178867515653184e-06, "loss": 0.4606, "step": 4264 }, { "epoch": 0.3023750443105282, "grad_norm": 6.8859124830700935, "learning_rate": 8.177981195850082e-06, "loss": 0.4828, "step": 4265 }, { "epoch": 0.30244594115561857, "grad_norm": 2.1307647328828168, "learning_rate": 8.177094708469608e-06, "loss": 0.3959, "step": 4266 }, { "epoch": 0.30251683800070894, "grad_norm": 12.848381191934653, "learning_rate": 8.176208053558503e-06, "loss": 0.4669, "step": 4267 }, { "epoch": 0.30258773484579937, "grad_norm": 5.5055745710651065, "learning_rate": 8.175321231163523e-06, "loss": 0.4116, "step": 4268 }, { "epoch": 0.30265863169088975, "grad_norm": 2.4282172794710655, "learning_rate": 8.174434241331432e-06, "loss": 0.3966, "step": 4269 }, { "epoch": 0.3027295285359802, "grad_norm": 2.592477696286291, "learning_rate": 8.173547084108999e-06, "loss": 0.4221, "step": 4270 }, { "epoch": 0.30280042538107055, "grad_norm": 2.4709530775517097, "learning_rate": 8.172659759543007e-06, "loss": 0.4306, "step": 4271 }, { "epoch": 0.3028713222261609, "grad_norm": 10.218649875294709, "learning_rate": 8.171772267680245e-06, "loss": 0.4551, "step": 4272 }, { "epoch": 0.30294221907125135, "grad_norm": 2.89576204537876, "learning_rate": 8.170884608567509e-06, "loss": 0.4371, "step": 4273 }, { "epoch": 0.3030131159163417, "grad_norm": 3.4613746615296934, "learning_rate": 8.169996782251609e-06, "loss": 0.4861, "step": 4274 }, { "epoch": 0.3030840127614321, "grad_norm": 2.311878110336269, "learning_rate": 8.169108788779359e-06, "loss": 0.4139, "step": 4275 }, { "epoch": 0.3031549096065225, "grad_norm": 2.2065818938395987, "learning_rate": 8.168220628197581e-06, "loss": 0.3677, "step": 4276 }, { "epoch": 0.3032258064516129, "grad_norm": 5.093347339330436, "learning_rate": 8.167332300553112e-06, "loss": 0.4124, "step": 4277 }, { "epoch": 0.3032967032967033, "grad_norm": 16.09097662066432, "learning_rate": 8.166443805892794e-06, "loss": 0.4125, "step": 4278 }, { "epoch": 0.3033676001417937, "grad_norm": 2.184361466374769, "learning_rate": 8.165555144263479e-06, "loss": 0.4205, "step": 4279 }, { "epoch": 0.3034384969868841, "grad_norm": 2.3483914279933975, "learning_rate": 8.164666315712022e-06, "loss": 0.3975, "step": 4280 }, { "epoch": 0.30350939383197445, "grad_norm": 2.987887687887539, "learning_rate": 8.163777320285297e-06, "loss": 0.4311, "step": 4281 }, { "epoch": 0.3035802906770649, "grad_norm": 5.5415914738687375, "learning_rate": 8.162888158030178e-06, "loss": 0.4189, "step": 4282 }, { "epoch": 0.30365118752215525, "grad_norm": 2.635703705706551, "learning_rate": 8.161998828993553e-06, "loss": 0.4228, "step": 4283 }, { "epoch": 0.3037220843672457, "grad_norm": 2.600527925636766, "learning_rate": 8.161109333222316e-06, "loss": 0.4135, "step": 4284 }, { "epoch": 0.30379298121233606, "grad_norm": 3.3719104068753607, "learning_rate": 8.160219670763372e-06, "loss": 0.405, "step": 4285 }, { "epoch": 0.30386387805742643, "grad_norm": 2.1273626356101825, "learning_rate": 8.15932984166363e-06, "loss": 0.4501, "step": 4286 }, { "epoch": 0.30393477490251686, "grad_norm": 3.2679329915306075, "learning_rate": 8.158439845970017e-06, "loss": 0.4213, "step": 4287 }, { "epoch": 0.30400567174760723, "grad_norm": 10.63112108222499, "learning_rate": 8.157549683729459e-06, "loss": 0.4649, "step": 4288 }, { "epoch": 0.3040765685926976, "grad_norm": 3.491058431023091, "learning_rate": 8.156659354988896e-06, "loss": 0.4312, "step": 4289 }, { "epoch": 0.30414746543778803, "grad_norm": 5.249228125018128, "learning_rate": 8.155768859795276e-06, "loss": 0.4392, "step": 4290 }, { "epoch": 0.3042183622828784, "grad_norm": 2.71918112839692, "learning_rate": 8.154878198195556e-06, "loss": 0.4181, "step": 4291 }, { "epoch": 0.3042892591279688, "grad_norm": 5.86392570835715, "learning_rate": 8.153987370236698e-06, "loss": 0.4528, "step": 4292 }, { "epoch": 0.3043601559730592, "grad_norm": 5.266362175627095, "learning_rate": 8.153096375965681e-06, "loss": 0.4642, "step": 4293 }, { "epoch": 0.3044310528181496, "grad_norm": 3.687496105468022, "learning_rate": 8.152205215429484e-06, "loss": 0.4324, "step": 4294 }, { "epoch": 0.30450194966324, "grad_norm": 2.648285575015972, "learning_rate": 8.151313888675101e-06, "loss": 0.4194, "step": 4295 }, { "epoch": 0.3045728465083304, "grad_norm": 2.924506036524129, "learning_rate": 8.150422395749532e-06, "loss": 0.3918, "step": 4296 }, { "epoch": 0.30464374335342076, "grad_norm": 3.950468838874824, "learning_rate": 8.149530736699785e-06, "loss": 0.4389, "step": 4297 }, { "epoch": 0.3047146401985112, "grad_norm": 2.5636822985846544, "learning_rate": 8.148638911572878e-06, "loss": 0.4455, "step": 4298 }, { "epoch": 0.30478553704360156, "grad_norm": 2.462985686512328, "learning_rate": 8.147746920415837e-06, "loss": 0.3657, "step": 4299 }, { "epoch": 0.30485643388869194, "grad_norm": 5.987595933838253, "learning_rate": 8.1468547632757e-06, "loss": 0.4044, "step": 4300 }, { "epoch": 0.30492733073378236, "grad_norm": 4.145217884875537, "learning_rate": 8.145962440199508e-06, "loss": 0.4751, "step": 4301 }, { "epoch": 0.30499822757887274, "grad_norm": 4.451187660038548, "learning_rate": 8.145069951234317e-06, "loss": 0.4626, "step": 4302 }, { "epoch": 0.3050691244239631, "grad_norm": 2.4909942317967957, "learning_rate": 8.144177296427187e-06, "loss": 0.4476, "step": 4303 }, { "epoch": 0.30514002126905354, "grad_norm": 2.8792800430765597, "learning_rate": 8.143284475825186e-06, "loss": 0.4537, "step": 4304 }, { "epoch": 0.3052109181141439, "grad_norm": 2.9438467685076954, "learning_rate": 8.142391489475397e-06, "loss": 0.412, "step": 4305 }, { "epoch": 0.3052818149592343, "grad_norm": 3.7323222106929532, "learning_rate": 8.141498337424907e-06, "loss": 0.438, "step": 4306 }, { "epoch": 0.3053527118043247, "grad_norm": 4.339725051356815, "learning_rate": 8.14060501972081e-06, "loss": 0.4393, "step": 4307 }, { "epoch": 0.3054236086494151, "grad_norm": 4.2150406967540475, "learning_rate": 8.139711536410215e-06, "loss": 0.4356, "step": 4308 }, { "epoch": 0.3054945054945055, "grad_norm": 1.5131072162569565, "learning_rate": 8.138817887540233e-06, "loss": 0.4511, "step": 4309 }, { "epoch": 0.3055654023395959, "grad_norm": 1.9323892754743746, "learning_rate": 8.137924073157988e-06, "loss": 0.4503, "step": 4310 }, { "epoch": 0.30563629918468627, "grad_norm": 13.13540487807439, "learning_rate": 8.137030093310612e-06, "loss": 0.5035, "step": 4311 }, { "epoch": 0.3057071960297767, "grad_norm": 3.7958774239727875, "learning_rate": 8.136135948045245e-06, "loss": 0.3973, "step": 4312 }, { "epoch": 0.30577809287486707, "grad_norm": 3.7297384669077855, "learning_rate": 8.135241637409035e-06, "loss": 0.4468, "step": 4313 }, { "epoch": 0.30584898971995744, "grad_norm": 4.419864253528918, "learning_rate": 8.134347161449141e-06, "loss": 0.4509, "step": 4314 }, { "epoch": 0.30591988656504787, "grad_norm": 2.9417736083947883, "learning_rate": 8.13345252021273e-06, "loss": 0.4117, "step": 4315 }, { "epoch": 0.30599078341013825, "grad_norm": 3.615565008530074, "learning_rate": 8.132557713746975e-06, "loss": 0.4575, "step": 4316 }, { "epoch": 0.3060616802552286, "grad_norm": 2.541144591147789, "learning_rate": 8.13166274209906e-06, "loss": 0.3795, "step": 4317 }, { "epoch": 0.30613257710031905, "grad_norm": 2.658681444518653, "learning_rate": 8.130767605316178e-06, "loss": 0.4577, "step": 4318 }, { "epoch": 0.3062034739454094, "grad_norm": 5.1069431495733415, "learning_rate": 8.129872303445532e-06, "loss": 0.4423, "step": 4319 }, { "epoch": 0.30627437079049985, "grad_norm": 2.4207240745298244, "learning_rate": 8.12897683653433e-06, "loss": 0.4598, "step": 4320 }, { "epoch": 0.3063452676355902, "grad_norm": 2.707101009876607, "learning_rate": 8.128081204629791e-06, "loss": 0.4748, "step": 4321 }, { "epoch": 0.3064161644806806, "grad_norm": 3.8324058910472596, "learning_rate": 8.127185407779143e-06, "loss": 0.4293, "step": 4322 }, { "epoch": 0.306487061325771, "grad_norm": 3.6683492984620054, "learning_rate": 8.12628944602962e-06, "loss": 0.4144, "step": 4323 }, { "epoch": 0.3065579581708614, "grad_norm": 3.4740423685977144, "learning_rate": 8.12539331942847e-06, "loss": 0.3906, "step": 4324 }, { "epoch": 0.3066288550159518, "grad_norm": 3.6927575406680533, "learning_rate": 8.124497028022945e-06, "loss": 0.4206, "step": 4325 }, { "epoch": 0.3066997518610422, "grad_norm": 5.078612008717184, "learning_rate": 8.123600571860306e-06, "loss": 0.4248, "step": 4326 }, { "epoch": 0.3067706487061326, "grad_norm": 1.1904652900336736, "learning_rate": 8.122703950987825e-06, "loss": 0.6973, "step": 4327 }, { "epoch": 0.30684154555122295, "grad_norm": 2.7855145466404387, "learning_rate": 8.121807165452782e-06, "loss": 0.4313, "step": 4328 }, { "epoch": 0.3069124423963134, "grad_norm": 2.1772742174817017, "learning_rate": 8.120910215302463e-06, "loss": 0.4445, "step": 4329 }, { "epoch": 0.30698333924140375, "grad_norm": 3.793520704328535, "learning_rate": 8.120013100584167e-06, "loss": 0.4734, "step": 4330 }, { "epoch": 0.3070542360864941, "grad_norm": 3.304669585122268, "learning_rate": 8.119115821345199e-06, "loss": 0.4366, "step": 4331 }, { "epoch": 0.30712513293158455, "grad_norm": 2.2125038974662266, "learning_rate": 8.118218377632872e-06, "loss": 0.413, "step": 4332 }, { "epoch": 0.30719602977667493, "grad_norm": 4.417502429178966, "learning_rate": 8.11732076949451e-06, "loss": 0.5108, "step": 4333 }, { "epoch": 0.30726692662176536, "grad_norm": 2.6731556850848626, "learning_rate": 8.116422996977446e-06, "loss": 0.431, "step": 4334 }, { "epoch": 0.30733782346685573, "grad_norm": 2.9175781920823685, "learning_rate": 8.115525060129016e-06, "loss": 0.4513, "step": 4335 }, { "epoch": 0.3074087203119461, "grad_norm": 3.890045112169412, "learning_rate": 8.114626958996576e-06, "loss": 0.3662, "step": 4336 }, { "epoch": 0.30747961715703653, "grad_norm": 4.344400705981229, "learning_rate": 8.113728693627476e-06, "loss": 0.4274, "step": 4337 }, { "epoch": 0.3075505140021269, "grad_norm": 5.16366494271386, "learning_rate": 8.112830264069084e-06, "loss": 0.448, "step": 4338 }, { "epoch": 0.3076214108472173, "grad_norm": 3.1925802599776945, "learning_rate": 8.111931670368779e-06, "loss": 0.4676, "step": 4339 }, { "epoch": 0.3076923076923077, "grad_norm": 2.689195103420484, "learning_rate": 8.111032912573941e-06, "loss": 0.455, "step": 4340 }, { "epoch": 0.3077632045373981, "grad_norm": 1.0673557830768772, "learning_rate": 8.110133990731964e-06, "loss": 0.6994, "step": 4341 }, { "epoch": 0.30783410138248846, "grad_norm": 2.5468957924805413, "learning_rate": 8.109234904890246e-06, "loss": 0.4249, "step": 4342 }, { "epoch": 0.3079049982275789, "grad_norm": 2.6587459189258262, "learning_rate": 8.108335655096199e-06, "loss": 0.4193, "step": 4343 }, { "epoch": 0.30797589507266926, "grad_norm": 2.0729263150626114, "learning_rate": 8.107436241397239e-06, "loss": 0.4142, "step": 4344 }, { "epoch": 0.30804679191775963, "grad_norm": 3.051746728986865, "learning_rate": 8.106536663840796e-06, "loss": 0.4219, "step": 4345 }, { "epoch": 0.30811768876285006, "grad_norm": 4.048488879781016, "learning_rate": 8.105636922474305e-06, "loss": 0.4093, "step": 4346 }, { "epoch": 0.30818858560794044, "grad_norm": 6.803902636946537, "learning_rate": 8.104737017345205e-06, "loss": 0.4306, "step": 4347 }, { "epoch": 0.30825948245303086, "grad_norm": 1.0829993190205678, "learning_rate": 8.103836948500956e-06, "loss": 0.6688, "step": 4348 }, { "epoch": 0.30833037929812124, "grad_norm": 2.6088009481404786, "learning_rate": 8.102936715989014e-06, "loss": 0.464, "step": 4349 }, { "epoch": 0.3084012761432116, "grad_norm": 3.945959017502613, "learning_rate": 8.102036319856849e-06, "loss": 0.4078, "step": 4350 }, { "epoch": 0.30847217298830204, "grad_norm": 2.7292961926315567, "learning_rate": 8.101135760151943e-06, "loss": 0.4316, "step": 4351 }, { "epoch": 0.3085430698333924, "grad_norm": 3.0305703643704622, "learning_rate": 8.100235036921781e-06, "loss": 0.386, "step": 4352 }, { "epoch": 0.3086139666784828, "grad_norm": 2.3633905307293306, "learning_rate": 8.09933415021386e-06, "loss": 0.3809, "step": 4353 }, { "epoch": 0.3086848635235732, "grad_norm": 2.1708164772321457, "learning_rate": 8.098433100075683e-06, "loss": 0.4654, "step": 4354 }, { "epoch": 0.3087557603686636, "grad_norm": 4.788490832348547, "learning_rate": 8.097531886554764e-06, "loss": 0.512, "step": 4355 }, { "epoch": 0.30882665721375396, "grad_norm": 26.630491839159536, "learning_rate": 8.096630509698624e-06, "loss": 0.4403, "step": 4356 }, { "epoch": 0.3088975540588444, "grad_norm": 2.578227086303425, "learning_rate": 8.095728969554792e-06, "loss": 0.4203, "step": 4357 }, { "epoch": 0.30896845090393477, "grad_norm": 4.94394576464999, "learning_rate": 8.09482726617081e-06, "loss": 0.4305, "step": 4358 }, { "epoch": 0.3090393477490252, "grad_norm": 4.3950907114307585, "learning_rate": 8.093925399594226e-06, "loss": 0.4685, "step": 4359 }, { "epoch": 0.30911024459411557, "grad_norm": 6.144450170319319, "learning_rate": 8.093023369872593e-06, "loss": 0.4396, "step": 4360 }, { "epoch": 0.30918114143920594, "grad_norm": 2.638856170003793, "learning_rate": 8.092121177053475e-06, "loss": 0.4121, "step": 4361 }, { "epoch": 0.30925203828429637, "grad_norm": 2.842904614120381, "learning_rate": 8.091218821184448e-06, "loss": 0.4735, "step": 4362 }, { "epoch": 0.30932293512938674, "grad_norm": 5.982341722913011, "learning_rate": 8.090316302313094e-06, "loss": 0.4514, "step": 4363 }, { "epoch": 0.3093938319744771, "grad_norm": 4.202702710607053, "learning_rate": 8.089413620487004e-06, "loss": 0.4408, "step": 4364 }, { "epoch": 0.30946472881956755, "grad_norm": 4.6141492483375535, "learning_rate": 8.088510775753772e-06, "loss": 0.4973, "step": 4365 }, { "epoch": 0.3095356256646579, "grad_norm": 17.7547690901753, "learning_rate": 8.087607768161013e-06, "loss": 0.4664, "step": 4366 }, { "epoch": 0.3096065225097483, "grad_norm": 3.273137446596358, "learning_rate": 8.086704597756339e-06, "loss": 0.4407, "step": 4367 }, { "epoch": 0.3096774193548387, "grad_norm": 2.7541963961826395, "learning_rate": 8.085801264587374e-06, "loss": 0.4413, "step": 4368 }, { "epoch": 0.3097483161999291, "grad_norm": 2.407097114489497, "learning_rate": 8.084897768701754e-06, "loss": 0.3931, "step": 4369 }, { "epoch": 0.30981921304501947, "grad_norm": 3.165203083899408, "learning_rate": 8.08399411014712e-06, "loss": 0.4244, "step": 4370 }, { "epoch": 0.3098901098901099, "grad_norm": 2.047918242802955, "learning_rate": 8.083090288971123e-06, "loss": 0.4158, "step": 4371 }, { "epoch": 0.3099610067352003, "grad_norm": 2.598612302743412, "learning_rate": 8.082186305221423e-06, "loss": 0.398, "step": 4372 }, { "epoch": 0.3100319035802907, "grad_norm": 2.1822369909197623, "learning_rate": 8.081282158945684e-06, "loss": 0.4163, "step": 4373 }, { "epoch": 0.3101028004253811, "grad_norm": 3.4537642840268545, "learning_rate": 8.080377850191586e-06, "loss": 0.4825, "step": 4374 }, { "epoch": 0.31017369727047145, "grad_norm": 3.5406890629967314, "learning_rate": 8.079473379006812e-06, "loss": 0.459, "step": 4375 }, { "epoch": 0.3102445941155619, "grad_norm": 2.3303182351248957, "learning_rate": 8.078568745439057e-06, "loss": 0.3958, "step": 4376 }, { "epoch": 0.31031549096065225, "grad_norm": 3.382057677966748, "learning_rate": 8.077663949536024e-06, "loss": 0.4277, "step": 4377 }, { "epoch": 0.3103863878057426, "grad_norm": 3.2118900443832183, "learning_rate": 8.076758991345419e-06, "loss": 0.4248, "step": 4378 }, { "epoch": 0.31045728465083305, "grad_norm": 5.133455603830768, "learning_rate": 8.075853870914966e-06, "loss": 0.4825, "step": 4379 }, { "epoch": 0.3105281814959234, "grad_norm": 4.354108168196335, "learning_rate": 8.07494858829239e-06, "loss": 0.426, "step": 4380 }, { "epoch": 0.3105990783410138, "grad_norm": 3.9813945323448334, "learning_rate": 8.074043143525427e-06, "loss": 0.4101, "step": 4381 }, { "epoch": 0.31066997518610423, "grad_norm": 1.7730377912799389, "learning_rate": 8.073137536661824e-06, "loss": 0.4041, "step": 4382 }, { "epoch": 0.3107408720311946, "grad_norm": 3.4227216290572593, "learning_rate": 8.072231767749334e-06, "loss": 0.4004, "step": 4383 }, { "epoch": 0.31081176887628503, "grad_norm": 2.7495255883846683, "learning_rate": 8.071325836835714e-06, "loss": 0.4375, "step": 4384 }, { "epoch": 0.3108826657213754, "grad_norm": 5.300192506662993, "learning_rate": 8.070419743968743e-06, "loss": 0.4528, "step": 4385 }, { "epoch": 0.3109535625664658, "grad_norm": 2.6424629794334944, "learning_rate": 8.069513489196193e-06, "loss": 0.4674, "step": 4386 }, { "epoch": 0.3110244594115562, "grad_norm": 3.1722128708562094, "learning_rate": 8.068607072565855e-06, "loss": 0.467, "step": 4387 }, { "epoch": 0.3110953562566466, "grad_norm": 3.248699026544778, "learning_rate": 8.067700494125524e-06, "loss": 0.4028, "step": 4388 }, { "epoch": 0.31116625310173696, "grad_norm": 4.8496698997234295, "learning_rate": 8.066793753923001e-06, "loss": 0.4699, "step": 4389 }, { "epoch": 0.3112371499468274, "grad_norm": 2.618286330459951, "learning_rate": 8.065886852006107e-06, "loss": 0.4182, "step": 4390 }, { "epoch": 0.31130804679191776, "grad_norm": 2.1515613007233525, "learning_rate": 8.064979788422657e-06, "loss": 0.4311, "step": 4391 }, { "epoch": 0.31137894363700813, "grad_norm": 7.427173282957745, "learning_rate": 8.064072563220484e-06, "loss": 0.4243, "step": 4392 }, { "epoch": 0.31144984048209856, "grad_norm": 3.1024561465737217, "learning_rate": 8.063165176447426e-06, "loss": 0.4473, "step": 4393 }, { "epoch": 0.31152073732718893, "grad_norm": 3.4771475671557788, "learning_rate": 8.06225762815133e-06, "loss": 0.4303, "step": 4394 }, { "epoch": 0.3115916341722793, "grad_norm": 1.916989491780574, "learning_rate": 8.061349918380053e-06, "loss": 0.4262, "step": 4395 }, { "epoch": 0.31166253101736974, "grad_norm": 3.162183531686685, "learning_rate": 8.060442047181456e-06, "loss": 0.4445, "step": 4396 }, { "epoch": 0.3117334278624601, "grad_norm": 4.095316495489047, "learning_rate": 8.059534014603415e-06, "loss": 0.4973, "step": 4397 }, { "epoch": 0.31180432470755054, "grad_norm": 5.118946503431806, "learning_rate": 8.05862582069381e-06, "loss": 0.4576, "step": 4398 }, { "epoch": 0.3118752215526409, "grad_norm": 9.040740931859597, "learning_rate": 8.057717465500531e-06, "loss": 0.4297, "step": 4399 }, { "epoch": 0.3119461183977313, "grad_norm": 4.245945056714226, "learning_rate": 8.056808949071477e-06, "loss": 0.4261, "step": 4400 }, { "epoch": 0.3120170152428217, "grad_norm": 1.0627855682885181, "learning_rate": 8.055900271454554e-06, "loss": 0.609, "step": 4401 }, { "epoch": 0.3120879120879121, "grad_norm": 4.155406932530074, "learning_rate": 8.054991432697676e-06, "loss": 0.4749, "step": 4402 }, { "epoch": 0.31215880893300246, "grad_norm": 6.191198820091904, "learning_rate": 8.054082432848768e-06, "loss": 0.4551, "step": 4403 }, { "epoch": 0.3122297057780929, "grad_norm": 2.6116252272626594, "learning_rate": 8.053173271955764e-06, "loss": 0.4428, "step": 4404 }, { "epoch": 0.31230060262318327, "grad_norm": 2.072371482165456, "learning_rate": 8.0522639500666e-06, "loss": 0.4553, "step": 4405 }, { "epoch": 0.31237149946827364, "grad_norm": 6.823601723788955, "learning_rate": 8.051354467229231e-06, "loss": 0.4621, "step": 4406 }, { "epoch": 0.31244239631336407, "grad_norm": 4.961530353470188, "learning_rate": 8.050444823491612e-06, "loss": 0.4448, "step": 4407 }, { "epoch": 0.31251329315845444, "grad_norm": 3.214018157492258, "learning_rate": 8.049535018901708e-06, "loss": 0.4245, "step": 4408 }, { "epoch": 0.3125841900035448, "grad_norm": 2.798589758254771, "learning_rate": 8.048625053507494e-06, "loss": 0.4361, "step": 4409 }, { "epoch": 0.31265508684863524, "grad_norm": 2.239098014130397, "learning_rate": 8.047714927356957e-06, "loss": 0.4438, "step": 4410 }, { "epoch": 0.3127259836937256, "grad_norm": 1.9770285281784417, "learning_rate": 8.046804640498083e-06, "loss": 0.4303, "step": 4411 }, { "epoch": 0.31279688053881605, "grad_norm": 6.883321135766541, "learning_rate": 8.045894192978877e-06, "loss": 0.4668, "step": 4412 }, { "epoch": 0.3128677773839064, "grad_norm": 2.8003773537815517, "learning_rate": 8.044983584847343e-06, "loss": 0.425, "step": 4413 }, { "epoch": 0.3129386742289968, "grad_norm": 5.417802027128339, "learning_rate": 8.0440728161515e-06, "loss": 0.5016, "step": 4414 }, { "epoch": 0.3130095710740872, "grad_norm": 5.212491117730924, "learning_rate": 8.043161886939376e-06, "loss": 0.3965, "step": 4415 }, { "epoch": 0.3130804679191776, "grad_norm": 10.797130227378238, "learning_rate": 8.042250797259002e-06, "loss": 0.4173, "step": 4416 }, { "epoch": 0.31315136476426797, "grad_norm": 1.049848361080311, "learning_rate": 8.041339547158422e-06, "loss": 0.6848, "step": 4417 }, { "epoch": 0.3132222616093584, "grad_norm": 3.410201711700601, "learning_rate": 8.040428136685685e-06, "loss": 0.4259, "step": 4418 }, { "epoch": 0.3132931584544488, "grad_norm": 3.085405736410312, "learning_rate": 8.039516565888855e-06, "loss": 0.4577, "step": 4419 }, { "epoch": 0.31336405529953915, "grad_norm": 4.378371766929315, "learning_rate": 8.038604834815993e-06, "loss": 0.4488, "step": 4420 }, { "epoch": 0.3134349521446296, "grad_norm": 4.13562656875136, "learning_rate": 8.03769294351518e-06, "loss": 0.4162, "step": 4421 }, { "epoch": 0.31350584898971995, "grad_norm": 2.8082694191394766, "learning_rate": 8.0367808920345e-06, "loss": 0.4283, "step": 4422 }, { "epoch": 0.3135767458348104, "grad_norm": 2.6101442890952877, "learning_rate": 8.035868680422045e-06, "loss": 0.3816, "step": 4423 }, { "epoch": 0.31364764267990075, "grad_norm": 1.1639001415632269, "learning_rate": 8.034956308725918e-06, "loss": 0.745, "step": 4424 }, { "epoch": 0.3137185395249911, "grad_norm": 3.3134426412997873, "learning_rate": 8.034043776994226e-06, "loss": 0.3747, "step": 4425 }, { "epoch": 0.31378943637008155, "grad_norm": 4.656318910625192, "learning_rate": 8.033131085275092e-06, "loss": 0.406, "step": 4426 }, { "epoch": 0.3138603332151719, "grad_norm": 2.8707523333704934, "learning_rate": 8.032218233616642e-06, "loss": 0.4209, "step": 4427 }, { "epoch": 0.3139312300602623, "grad_norm": 2.1252347735324437, "learning_rate": 8.031305222067008e-06, "loss": 0.4226, "step": 4428 }, { "epoch": 0.31400212690535273, "grad_norm": 3.8695361455866712, "learning_rate": 8.030392050674336e-06, "loss": 0.4599, "step": 4429 }, { "epoch": 0.3140730237504431, "grad_norm": 5.44002909184335, "learning_rate": 8.02947871948678e-06, "loss": 0.4427, "step": 4430 }, { "epoch": 0.3141439205955335, "grad_norm": 4.67417492958606, "learning_rate": 8.028565228552499e-06, "loss": 0.4569, "step": 4431 }, { "epoch": 0.3142148174406239, "grad_norm": 4.380504628052526, "learning_rate": 8.02765157791966e-06, "loss": 0.3888, "step": 4432 }, { "epoch": 0.3142857142857143, "grad_norm": 2.8535353821096168, "learning_rate": 8.026737767636443e-06, "loss": 0.3866, "step": 4433 }, { "epoch": 0.31435661113080465, "grad_norm": 2.883801295202891, "learning_rate": 8.02582379775103e-06, "loss": 0.3834, "step": 4434 }, { "epoch": 0.3144275079758951, "grad_norm": 5.962234961237425, "learning_rate": 8.024909668311624e-06, "loss": 0.4829, "step": 4435 }, { "epoch": 0.31449840482098546, "grad_norm": 3.5364801373374752, "learning_rate": 8.02399537936642e-06, "loss": 0.4143, "step": 4436 }, { "epoch": 0.3145693016660759, "grad_norm": 3.0990087704957956, "learning_rate": 8.02308093096363e-06, "loss": 0.4559, "step": 4437 }, { "epoch": 0.31464019851116626, "grad_norm": 11.422736785856333, "learning_rate": 8.022166323151478e-06, "loss": 0.4035, "step": 4438 }, { "epoch": 0.31471109535625663, "grad_norm": 5.146309245779919, "learning_rate": 8.02125155597819e-06, "loss": 0.4372, "step": 4439 }, { "epoch": 0.31478199220134706, "grad_norm": 2.329216251449128, "learning_rate": 8.020336629491998e-06, "loss": 0.4223, "step": 4440 }, { "epoch": 0.31485288904643743, "grad_norm": 4.404832696998901, "learning_rate": 8.019421543741153e-06, "loss": 0.4236, "step": 4441 }, { "epoch": 0.3149237858915278, "grad_norm": 3.6208068588119984, "learning_rate": 8.018506298773902e-06, "loss": 0.4691, "step": 4442 }, { "epoch": 0.31499468273661824, "grad_norm": 3.796194615929156, "learning_rate": 8.017590894638513e-06, "loss": 0.4547, "step": 4443 }, { "epoch": 0.3150655795817086, "grad_norm": 3.958787442555528, "learning_rate": 8.016675331383252e-06, "loss": 0.3975, "step": 4444 }, { "epoch": 0.315136476426799, "grad_norm": 4.3140696602892445, "learning_rate": 8.015759609056398e-06, "loss": 0.4583, "step": 4445 }, { "epoch": 0.3152073732718894, "grad_norm": 4.9897213836734675, "learning_rate": 8.014843727706238e-06, "loss": 0.4244, "step": 4446 }, { "epoch": 0.3152782701169798, "grad_norm": 2.1852574438728176, "learning_rate": 8.013927687381067e-06, "loss": 0.4412, "step": 4447 }, { "epoch": 0.3153491669620702, "grad_norm": 5.371560624019603, "learning_rate": 8.01301148812919e-06, "loss": 0.4563, "step": 4448 }, { "epoch": 0.3154200638071606, "grad_norm": 3.2229618415452332, "learning_rate": 8.012095129998916e-06, "loss": 0.4091, "step": 4449 }, { "epoch": 0.31549096065225096, "grad_norm": 3.5553023395555052, "learning_rate": 8.011178613038566e-06, "loss": 0.4469, "step": 4450 }, { "epoch": 0.3155618574973414, "grad_norm": 5.802320094248876, "learning_rate": 8.010261937296471e-06, "loss": 0.4177, "step": 4451 }, { "epoch": 0.31563275434243176, "grad_norm": 2.947193565210172, "learning_rate": 8.009345102820966e-06, "loss": 0.4411, "step": 4452 }, { "epoch": 0.31570365118752214, "grad_norm": 3.222803287632686, "learning_rate": 8.008428109660396e-06, "loss": 0.4175, "step": 4453 }, { "epoch": 0.31577454803261257, "grad_norm": 3.8734865044867144, "learning_rate": 8.007510957863117e-06, "loss": 0.4548, "step": 4454 }, { "epoch": 0.31584544487770294, "grad_norm": 3.183609215034004, "learning_rate": 8.00659364747749e-06, "loss": 0.4658, "step": 4455 }, { "epoch": 0.3159163417227933, "grad_norm": 3.4856391601263548, "learning_rate": 8.005676178551884e-06, "loss": 0.3776, "step": 4456 }, { "epoch": 0.31598723856788374, "grad_norm": 2.7597726071693254, "learning_rate": 8.00475855113468e-06, "loss": 0.3911, "step": 4457 }, { "epoch": 0.3160581354129741, "grad_norm": 1.6826154358081302, "learning_rate": 8.003840765274262e-06, "loss": 0.386, "step": 4458 }, { "epoch": 0.3161290322580645, "grad_norm": 3.3607136346409034, "learning_rate": 8.002922821019028e-06, "loss": 0.4317, "step": 4459 }, { "epoch": 0.3161999291031549, "grad_norm": 7.710859771281011, "learning_rate": 8.002004718417382e-06, "loss": 0.3819, "step": 4460 }, { "epoch": 0.3162708259482453, "grad_norm": 1.197651110837546, "learning_rate": 8.001086457517737e-06, "loss": 0.7219, "step": 4461 }, { "epoch": 0.3163417227933357, "grad_norm": 3.056819650535828, "learning_rate": 8.000168038368511e-06, "loss": 0.4412, "step": 4462 }, { "epoch": 0.3164126196384261, "grad_norm": 3.197579476413708, "learning_rate": 7.999249461018138e-06, "loss": 0.4546, "step": 4463 }, { "epoch": 0.31648351648351647, "grad_norm": 44.96995149054972, "learning_rate": 7.998330725515047e-06, "loss": 0.4094, "step": 4464 }, { "epoch": 0.3165544133286069, "grad_norm": 2.931961206784025, "learning_rate": 7.99741183190769e-06, "loss": 0.4195, "step": 4465 }, { "epoch": 0.31662531017369727, "grad_norm": 2.5472733123399807, "learning_rate": 7.99649278024452e-06, "loss": 0.4407, "step": 4466 }, { "epoch": 0.31669620701878765, "grad_norm": 2.6615240729842737, "learning_rate": 7.995573570573998e-06, "loss": 0.4101, "step": 4467 }, { "epoch": 0.3167671038638781, "grad_norm": 3.1934453141465453, "learning_rate": 7.994654202944595e-06, "loss": 0.4025, "step": 4468 }, { "epoch": 0.31683800070896845, "grad_norm": 2.9738578295126605, "learning_rate": 7.993734677404792e-06, "loss": 0.4495, "step": 4469 }, { "epoch": 0.3169088975540588, "grad_norm": 3.9874252045692953, "learning_rate": 7.992814994003073e-06, "loss": 0.4277, "step": 4470 }, { "epoch": 0.31697979439914925, "grad_norm": 1.2083863459899107, "learning_rate": 7.991895152787934e-06, "loss": 0.696, "step": 4471 }, { "epoch": 0.3170506912442396, "grad_norm": 5.433422990700881, "learning_rate": 7.990975153807883e-06, "loss": 0.4706, "step": 4472 }, { "epoch": 0.31712158808933, "grad_norm": 39.08421287557648, "learning_rate": 7.990054997111426e-06, "loss": 0.4315, "step": 4473 }, { "epoch": 0.3171924849344204, "grad_norm": 2.660639863027117, "learning_rate": 7.989134682747088e-06, "loss": 0.3758, "step": 4474 }, { "epoch": 0.3172633817795108, "grad_norm": 2.9564244112267857, "learning_rate": 7.988214210763397e-06, "loss": 0.4579, "step": 4475 }, { "epoch": 0.31733427862460123, "grad_norm": 3.5065239029174315, "learning_rate": 7.98729358120889e-06, "loss": 0.4276, "step": 4476 }, { "epoch": 0.3174051754696916, "grad_norm": 2.915120359832332, "learning_rate": 7.986372794132111e-06, "loss": 0.4215, "step": 4477 }, { "epoch": 0.317476072314782, "grad_norm": 4.152854117936152, "learning_rate": 7.985451849581616e-06, "loss": 0.4435, "step": 4478 }, { "epoch": 0.3175469691598724, "grad_norm": 3.4919977686110912, "learning_rate": 7.984530747605967e-06, "loss": 0.5025, "step": 4479 }, { "epoch": 0.3176178660049628, "grad_norm": 6.625305640045716, "learning_rate": 7.983609488253734e-06, "loss": 0.3745, "step": 4480 }, { "epoch": 0.31768876285005315, "grad_norm": 5.830770777173941, "learning_rate": 7.982688071573494e-06, "loss": 0.414, "step": 4481 }, { "epoch": 0.3177596596951436, "grad_norm": 2.6188863223146504, "learning_rate": 7.981766497613836e-06, "loss": 0.376, "step": 4482 }, { "epoch": 0.31783055654023395, "grad_norm": 3.631332070010235, "learning_rate": 7.980844766423355e-06, "loss": 0.5445, "step": 4483 }, { "epoch": 0.31790145338532433, "grad_norm": 6.539134216390247, "learning_rate": 7.979922878050653e-06, "loss": 0.3952, "step": 4484 }, { "epoch": 0.31797235023041476, "grad_norm": 3.857781052324733, "learning_rate": 7.979000832544345e-06, "loss": 0.3777, "step": 4485 }, { "epoch": 0.31804324707550513, "grad_norm": 3.843900721790667, "learning_rate": 7.978078629953047e-06, "loss": 0.4083, "step": 4486 }, { "epoch": 0.31811414392059556, "grad_norm": 12.13203892593874, "learning_rate": 7.977156270325393e-06, "loss": 0.4022, "step": 4487 }, { "epoch": 0.31818504076568593, "grad_norm": 7.511060284747351, "learning_rate": 7.976233753710015e-06, "loss": 0.4221, "step": 4488 }, { "epoch": 0.3182559376107763, "grad_norm": 2.5730642436230977, "learning_rate": 7.97531108015556e-06, "loss": 0.4415, "step": 4489 }, { "epoch": 0.31832683445586674, "grad_norm": 2.2693905737707327, "learning_rate": 7.974388249710678e-06, "loss": 0.4636, "step": 4490 }, { "epoch": 0.3183977313009571, "grad_norm": 3.611941271082135, "learning_rate": 7.973465262424036e-06, "loss": 0.4821, "step": 4491 }, { "epoch": 0.3184686281460475, "grad_norm": 7.532312135776835, "learning_rate": 7.9725421183443e-06, "loss": 0.4443, "step": 4492 }, { "epoch": 0.3185395249911379, "grad_norm": 3.9646845530692376, "learning_rate": 7.97161881752015e-06, "loss": 0.3942, "step": 4493 }, { "epoch": 0.3186104218362283, "grad_norm": 4.977722038559661, "learning_rate": 7.970695360000271e-06, "loss": 0.5079, "step": 4494 }, { "epoch": 0.31868131868131866, "grad_norm": 2.7555962003813863, "learning_rate": 7.969771745833359e-06, "loss": 0.3759, "step": 4495 }, { "epoch": 0.3187522155264091, "grad_norm": 9.519264149369945, "learning_rate": 7.968847975068116e-06, "loss": 0.4238, "step": 4496 }, { "epoch": 0.31882311237149946, "grad_norm": 2.1739035068391814, "learning_rate": 7.967924047753252e-06, "loss": 0.4436, "step": 4497 }, { "epoch": 0.31889400921658984, "grad_norm": 4.155546654708447, "learning_rate": 7.96699996393749e-06, "loss": 0.4513, "step": 4498 }, { "epoch": 0.31896490606168026, "grad_norm": 5.2650152504924295, "learning_rate": 7.966075723669552e-06, "loss": 0.4435, "step": 4499 }, { "epoch": 0.31903580290677064, "grad_norm": 5.145929905225928, "learning_rate": 7.965151326998181e-06, "loss": 0.5146, "step": 4500 }, { "epoch": 0.31910669975186107, "grad_norm": 3.8034711121482188, "learning_rate": 7.964226773972114e-06, "loss": 0.4265, "step": 4501 }, { "epoch": 0.31917759659695144, "grad_norm": 2.58657840951963, "learning_rate": 7.963302064640109e-06, "loss": 0.4124, "step": 4502 }, { "epoch": 0.3192484934420418, "grad_norm": 4.2056984284864525, "learning_rate": 7.962377199050924e-06, "loss": 0.4545, "step": 4503 }, { "epoch": 0.31931939028713224, "grad_norm": 1.6186015774352323, "learning_rate": 7.961452177253328e-06, "loss": 0.7788, "step": 4504 }, { "epoch": 0.3193902871322226, "grad_norm": 2.4941777510098535, "learning_rate": 7.960526999296099e-06, "loss": 0.4422, "step": 4505 }, { "epoch": 0.319461183977313, "grad_norm": 4.93407577557712, "learning_rate": 7.95960166522802e-06, "loss": 0.4465, "step": 4506 }, { "epoch": 0.3195320808224034, "grad_norm": 1.6949716282726177, "learning_rate": 7.958676175097885e-06, "loss": 0.3932, "step": 4507 }, { "epoch": 0.3196029776674938, "grad_norm": 2.4835224829777824, "learning_rate": 7.9577505289545e-06, "loss": 0.412, "step": 4508 }, { "epoch": 0.31967387451258417, "grad_norm": 4.249633204358734, "learning_rate": 7.956824726846672e-06, "loss": 0.4207, "step": 4509 }, { "epoch": 0.3197447713576746, "grad_norm": 2.462134747849003, "learning_rate": 7.955898768823218e-06, "loss": 0.401, "step": 4510 }, { "epoch": 0.31981566820276497, "grad_norm": 6.408423644755208, "learning_rate": 7.954972654932965e-06, "loss": 0.4785, "step": 4511 }, { "epoch": 0.3198865650478554, "grad_norm": 1.1369234630857774, "learning_rate": 7.954046385224751e-06, "loss": 0.7181, "step": 4512 }, { "epoch": 0.31995746189294577, "grad_norm": 3.371900470866223, "learning_rate": 7.953119959747415e-06, "loss": 0.3897, "step": 4513 }, { "epoch": 0.32002835873803614, "grad_norm": 3.0702080644887966, "learning_rate": 7.95219337854981e-06, "loss": 0.402, "step": 4514 }, { "epoch": 0.3200992555831266, "grad_norm": 3.3938572935517657, "learning_rate": 7.951266641680793e-06, "loss": 0.449, "step": 4515 }, { "epoch": 0.32017015242821695, "grad_norm": 2.4291462389207985, "learning_rate": 7.950339749189235e-06, "loss": 0.376, "step": 4516 }, { "epoch": 0.3202410492733073, "grad_norm": 5.655457570012143, "learning_rate": 7.94941270112401e-06, "loss": 0.3942, "step": 4517 }, { "epoch": 0.32031194611839775, "grad_norm": 3.8567848904160673, "learning_rate": 7.948485497534001e-06, "loss": 0.4361, "step": 4518 }, { "epoch": 0.3203828429634881, "grad_norm": 7.7078700612640185, "learning_rate": 7.947558138468103e-06, "loss": 0.4394, "step": 4519 }, { "epoch": 0.3204537398085785, "grad_norm": 2.9704944001688953, "learning_rate": 7.946630623975215e-06, "loss": 0.4479, "step": 4520 }, { "epoch": 0.3205246366536689, "grad_norm": 3.641552810069247, "learning_rate": 7.945702954104242e-06, "loss": 0.4722, "step": 4521 }, { "epoch": 0.3205955334987593, "grad_norm": 2.8536407459464566, "learning_rate": 7.944775128904107e-06, "loss": 0.3784, "step": 4522 }, { "epoch": 0.3206664303438497, "grad_norm": 4.523119547535437, "learning_rate": 7.94384714842373e-06, "loss": 0.3918, "step": 4523 }, { "epoch": 0.3207373271889401, "grad_norm": 4.710217718468519, "learning_rate": 7.942919012712047e-06, "loss": 0.4205, "step": 4524 }, { "epoch": 0.3208082240340305, "grad_norm": 4.612177309619189, "learning_rate": 7.941990721817998e-06, "loss": 0.4461, "step": 4525 }, { "epoch": 0.3208791208791209, "grad_norm": 4.270660072580472, "learning_rate": 7.941062275790531e-06, "loss": 0.5082, "step": 4526 }, { "epoch": 0.3209500177242113, "grad_norm": 2.855948984062428, "learning_rate": 7.940133674678608e-06, "loss": 0.4841, "step": 4527 }, { "epoch": 0.32102091456930165, "grad_norm": 2.5128406973743447, "learning_rate": 7.939204918531191e-06, "loss": 0.4637, "step": 4528 }, { "epoch": 0.3210918114143921, "grad_norm": 13.047197205151784, "learning_rate": 7.938276007397255e-06, "loss": 0.397, "step": 4529 }, { "epoch": 0.32116270825948245, "grad_norm": 5.899331061987916, "learning_rate": 7.937346941325783e-06, "loss": 0.412, "step": 4530 }, { "epoch": 0.32123360510457283, "grad_norm": 2.5078137473229325, "learning_rate": 7.936417720365766e-06, "loss": 0.4871, "step": 4531 }, { "epoch": 0.32130450194966326, "grad_norm": 2.4596573298150965, "learning_rate": 7.935488344566199e-06, "loss": 0.3733, "step": 4532 }, { "epoch": 0.32137539879475363, "grad_norm": 1.018899933128228, "learning_rate": 7.934558813976093e-06, "loss": 0.6382, "step": 4533 }, { "epoch": 0.321446295639844, "grad_norm": 4.262895296448533, "learning_rate": 7.93362912864446e-06, "loss": 0.4341, "step": 4534 }, { "epoch": 0.32151719248493443, "grad_norm": 3.240310028031968, "learning_rate": 7.932699288620322e-06, "loss": 0.3996, "step": 4535 }, { "epoch": 0.3215880893300248, "grad_norm": 3.920092378118111, "learning_rate": 7.931769293952714e-06, "loss": 0.3885, "step": 4536 }, { "epoch": 0.3216589861751152, "grad_norm": 3.80073380128583, "learning_rate": 7.930839144690673e-06, "loss": 0.4415, "step": 4537 }, { "epoch": 0.3217298830202056, "grad_norm": 4.177216948751558, "learning_rate": 7.929908840883248e-06, "loss": 0.4491, "step": 4538 }, { "epoch": 0.321800779865296, "grad_norm": 2.280694064955698, "learning_rate": 7.928978382579493e-06, "loss": 0.3573, "step": 4539 }, { "epoch": 0.3218716767103864, "grad_norm": 4.6956859037828345, "learning_rate": 7.92804776982847e-06, "loss": 0.4388, "step": 4540 }, { "epoch": 0.3219425735554768, "grad_norm": 8.589506922103903, "learning_rate": 7.927117002679257e-06, "loss": 0.4431, "step": 4541 }, { "epoch": 0.32201347040056716, "grad_norm": 2.372592199944966, "learning_rate": 7.926186081180929e-06, "loss": 0.4455, "step": 4542 }, { "epoch": 0.3220843672456576, "grad_norm": 3.520041921279615, "learning_rate": 7.925255005382575e-06, "loss": 0.4427, "step": 4543 }, { "epoch": 0.32215526409074796, "grad_norm": 3.336463304131963, "learning_rate": 7.924323775333291e-06, "loss": 0.4193, "step": 4544 }, { "epoch": 0.32222616093583833, "grad_norm": 8.933708494371977, "learning_rate": 7.923392391082184e-06, "loss": 0.4827, "step": 4545 }, { "epoch": 0.32229705778092876, "grad_norm": 8.606136252922047, "learning_rate": 7.922460852678362e-06, "loss": 0.4233, "step": 4546 }, { "epoch": 0.32236795462601914, "grad_norm": 3.4646348317585307, "learning_rate": 7.921529160170952e-06, "loss": 0.3845, "step": 4547 }, { "epoch": 0.3224388514711095, "grad_norm": 3.4825015075507713, "learning_rate": 7.920597313609077e-06, "loss": 0.4115, "step": 4548 }, { "epoch": 0.32250974831619994, "grad_norm": 3.3917440726169485, "learning_rate": 7.919665313041879e-06, "loss": 0.4521, "step": 4549 }, { "epoch": 0.3225806451612903, "grad_norm": 3.983595734790177, "learning_rate": 7.918733158518499e-06, "loss": 0.4253, "step": 4550 }, { "epoch": 0.32265154200638074, "grad_norm": 9.35696039310186, "learning_rate": 7.91780085008809e-06, "loss": 0.3857, "step": 4551 }, { "epoch": 0.3227224388514711, "grad_norm": 4.553914472771268, "learning_rate": 7.91686838779982e-06, "loss": 0.4449, "step": 4552 }, { "epoch": 0.3227933356965615, "grad_norm": 2.4704470174895676, "learning_rate": 7.91593577170285e-06, "loss": 0.3868, "step": 4553 }, { "epoch": 0.3228642325416519, "grad_norm": 5.146788954544983, "learning_rate": 7.915003001846362e-06, "loss": 0.4416, "step": 4554 }, { "epoch": 0.3229351293867423, "grad_norm": 6.389022351021478, "learning_rate": 7.91407007827954e-06, "loss": 0.4316, "step": 4555 }, { "epoch": 0.32300602623183267, "grad_norm": 4.460512285143542, "learning_rate": 7.913137001051578e-06, "loss": 0.4358, "step": 4556 }, { "epoch": 0.3230769230769231, "grad_norm": 2.140868577802574, "learning_rate": 7.91220377021168e-06, "loss": 0.4539, "step": 4557 }, { "epoch": 0.32314781992201347, "grad_norm": 4.121892997433282, "learning_rate": 7.911270385809053e-06, "loss": 0.4444, "step": 4558 }, { "epoch": 0.32321871676710384, "grad_norm": 25.2507099588986, "learning_rate": 7.910336847892919e-06, "loss": 0.4256, "step": 4559 }, { "epoch": 0.32328961361219427, "grad_norm": 3.597505230159245, "learning_rate": 7.909403156512499e-06, "loss": 0.4377, "step": 4560 }, { "epoch": 0.32336051045728464, "grad_norm": 3.6582607728375787, "learning_rate": 7.908469311717031e-06, "loss": 0.4236, "step": 4561 }, { "epoch": 0.323431407302375, "grad_norm": 3.163064349695943, "learning_rate": 7.907535313555757e-06, "loss": 0.4116, "step": 4562 }, { "epoch": 0.32350230414746545, "grad_norm": 1.9556621722336125, "learning_rate": 7.906601162077926e-06, "loss": 0.4156, "step": 4563 }, { "epoch": 0.3235732009925558, "grad_norm": 1.939252920703339, "learning_rate": 7.905666857332797e-06, "loss": 0.4287, "step": 4564 }, { "epoch": 0.32364409783764625, "grad_norm": 3.194628899474156, "learning_rate": 7.904732399369636e-06, "loss": 0.4426, "step": 4565 }, { "epoch": 0.3237149946827366, "grad_norm": 4.755848330187274, "learning_rate": 7.903797788237721e-06, "loss": 0.413, "step": 4566 }, { "epoch": 0.323785891527827, "grad_norm": 4.430293572235195, "learning_rate": 7.90286302398633e-06, "loss": 0.4792, "step": 4567 }, { "epoch": 0.3238567883729174, "grad_norm": 4.417081879830345, "learning_rate": 7.901928106664758e-06, "loss": 0.4049, "step": 4568 }, { "epoch": 0.3239276852180078, "grad_norm": 5.699413538020315, "learning_rate": 7.9009930363223e-06, "loss": 0.4245, "step": 4569 }, { "epoch": 0.3239985820630982, "grad_norm": 4.805344010795319, "learning_rate": 7.900057813008265e-06, "loss": 0.4655, "step": 4570 }, { "epoch": 0.3240694789081886, "grad_norm": 3.5606129129375774, "learning_rate": 7.89912243677197e-06, "loss": 0.4607, "step": 4571 }, { "epoch": 0.324140375753279, "grad_norm": 6.088639965028092, "learning_rate": 7.898186907662736e-06, "loss": 0.4605, "step": 4572 }, { "epoch": 0.32421127259836935, "grad_norm": 2.2180176858302505, "learning_rate": 7.897251225729892e-06, "loss": 0.3763, "step": 4573 }, { "epoch": 0.3242821694434598, "grad_norm": 2.024452447265839, "learning_rate": 7.896315391022779e-06, "loss": 0.4317, "step": 4574 }, { "epoch": 0.32435306628855015, "grad_norm": 2.9463434254725263, "learning_rate": 7.895379403590747e-06, "loss": 0.3941, "step": 4575 }, { "epoch": 0.3244239631336406, "grad_norm": 3.2556737037546055, "learning_rate": 7.894443263483149e-06, "loss": 0.4655, "step": 4576 }, { "epoch": 0.32449485997873095, "grad_norm": 8.463348824254753, "learning_rate": 7.893506970749348e-06, "loss": 0.4839, "step": 4577 }, { "epoch": 0.3245657568238213, "grad_norm": 3.8844704891353268, "learning_rate": 7.892570525438716e-06, "loss": 0.448, "step": 4578 }, { "epoch": 0.32463665366891176, "grad_norm": 3.7256953214096282, "learning_rate": 7.891633927600632e-06, "loss": 0.4444, "step": 4579 }, { "epoch": 0.32470755051400213, "grad_norm": 5.328514740320017, "learning_rate": 7.890697177284486e-06, "loss": 0.4287, "step": 4580 }, { "epoch": 0.3247784473590925, "grad_norm": 2.516676500173288, "learning_rate": 7.88976027453967e-06, "loss": 0.4785, "step": 4581 }, { "epoch": 0.32484934420418293, "grad_norm": 3.764442031230846, "learning_rate": 7.888823219415587e-06, "loss": 0.3949, "step": 4582 }, { "epoch": 0.3249202410492733, "grad_norm": 2.951146148923433, "learning_rate": 7.887886011961653e-06, "loss": 0.4559, "step": 4583 }, { "epoch": 0.3249911378943637, "grad_norm": 2.359797451897681, "learning_rate": 7.886948652227284e-06, "loss": 0.3965, "step": 4584 }, { "epoch": 0.3250620347394541, "grad_norm": 4.138540093140528, "learning_rate": 7.88601114026191e-06, "loss": 0.4114, "step": 4585 }, { "epoch": 0.3251329315845445, "grad_norm": 2.696038344489269, "learning_rate": 7.885073476114966e-06, "loss": 0.3954, "step": 4586 }, { "epoch": 0.32520382842963486, "grad_norm": 4.089568487852079, "learning_rate": 7.884135659835895e-06, "loss": 0.4267, "step": 4587 }, { "epoch": 0.3252747252747253, "grad_norm": 2.7981472466099615, "learning_rate": 7.88319769147415e-06, "loss": 0.3883, "step": 4588 }, { "epoch": 0.32534562211981566, "grad_norm": 3.6711426280442807, "learning_rate": 7.88225957107919e-06, "loss": 0.3868, "step": 4589 }, { "epoch": 0.3254165189649061, "grad_norm": 2.914190768032365, "learning_rate": 7.881321298700481e-06, "loss": 0.4294, "step": 4590 }, { "epoch": 0.32548741580999646, "grad_norm": 3.0694352232777113, "learning_rate": 7.880382874387503e-06, "loss": 0.4011, "step": 4591 }, { "epoch": 0.32555831265508683, "grad_norm": 3.5489461066147445, "learning_rate": 7.879444298189738e-06, "loss": 0.4273, "step": 4592 }, { "epoch": 0.32562920950017726, "grad_norm": 3.228787371719364, "learning_rate": 7.878505570156675e-06, "loss": 0.456, "step": 4593 }, { "epoch": 0.32570010634526764, "grad_norm": 4.054858603375314, "learning_rate": 7.877566690337818e-06, "loss": 0.4435, "step": 4594 }, { "epoch": 0.325771003190358, "grad_norm": 4.334438022149405, "learning_rate": 7.876627658782672e-06, "loss": 0.5037, "step": 4595 }, { "epoch": 0.32584190003544844, "grad_norm": 4.380191955232061, "learning_rate": 7.875688475540755e-06, "loss": 0.4277, "step": 4596 }, { "epoch": 0.3259127968805388, "grad_norm": 5.125930707515628, "learning_rate": 7.874749140661588e-06, "loss": 0.464, "step": 4597 }, { "epoch": 0.3259836937256292, "grad_norm": 2.5252210597186835, "learning_rate": 7.873809654194706e-06, "loss": 0.3932, "step": 4598 }, { "epoch": 0.3260545905707196, "grad_norm": 4.110601472729308, "learning_rate": 7.872870016189647e-06, "loss": 0.5054, "step": 4599 }, { "epoch": 0.32612548741581, "grad_norm": 6.113132345946712, "learning_rate": 7.871930226695961e-06, "loss": 0.3853, "step": 4600 }, { "epoch": 0.32619638426090036, "grad_norm": 3.8270618555257876, "learning_rate": 7.8709902857632e-06, "loss": 0.4591, "step": 4601 }, { "epoch": 0.3262672811059908, "grad_norm": 2.5810926596460857, "learning_rate": 7.87005019344093e-06, "loss": 0.3929, "step": 4602 }, { "epoch": 0.32633817795108117, "grad_norm": 4.2348083994360755, "learning_rate": 7.869109949778723e-06, "loss": 0.433, "step": 4603 }, { "epoch": 0.3264090747961716, "grad_norm": 3.5472473572854955, "learning_rate": 7.86816955482616e-06, "loss": 0.4296, "step": 4604 }, { "epoch": 0.32647997164126197, "grad_norm": 5.5420779151078925, "learning_rate": 7.867229008632823e-06, "loss": 0.4965, "step": 4605 }, { "epoch": 0.32655086848635234, "grad_norm": 1.164202232261254, "learning_rate": 7.866288311248316e-06, "loss": 0.6898, "step": 4606 }, { "epoch": 0.32662176533144277, "grad_norm": 3.794278176048673, "learning_rate": 7.865347462722237e-06, "loss": 0.4064, "step": 4607 }, { "epoch": 0.32669266217653314, "grad_norm": 4.153224079370209, "learning_rate": 7.8644064631042e-06, "loss": 0.4417, "step": 4608 }, { "epoch": 0.3267635590216235, "grad_norm": 3.4419590336016697, "learning_rate": 7.863465312443823e-06, "loss": 0.4132, "step": 4609 }, { "epoch": 0.32683445586671395, "grad_norm": 2.582079375070871, "learning_rate": 7.862524010790736e-06, "loss": 0.3862, "step": 4610 }, { "epoch": 0.3269053527118043, "grad_norm": 2.688604653225798, "learning_rate": 7.861582558194572e-06, "loss": 0.5008, "step": 4611 }, { "epoch": 0.3269762495568947, "grad_norm": 23.664587927568, "learning_rate": 7.860640954704975e-06, "loss": 0.4035, "step": 4612 }, { "epoch": 0.3270471464019851, "grad_norm": 2.205937071673611, "learning_rate": 7.8596992003716e-06, "loss": 0.4142, "step": 4613 }, { "epoch": 0.3271180432470755, "grad_norm": 1.17457843242274, "learning_rate": 7.858757295244099e-06, "loss": 0.7247, "step": 4614 }, { "epoch": 0.3271889400921659, "grad_norm": 1.3703584077794384, "learning_rate": 7.857815239372147e-06, "loss": 0.7288, "step": 4615 }, { "epoch": 0.3272598369372563, "grad_norm": 3.7302059255222537, "learning_rate": 7.856873032805414e-06, "loss": 0.4547, "step": 4616 }, { "epoch": 0.3273307337823467, "grad_norm": 3.175238064012444, "learning_rate": 7.855930675593586e-06, "loss": 0.3906, "step": 4617 }, { "epoch": 0.3274016306274371, "grad_norm": 9.112213563693743, "learning_rate": 7.854988167786355e-06, "loss": 0.4687, "step": 4618 }, { "epoch": 0.3274725274725275, "grad_norm": 3.1111633590952916, "learning_rate": 7.854045509433417e-06, "loss": 0.4514, "step": 4619 }, { "epoch": 0.32754342431761785, "grad_norm": 3.1136934662173656, "learning_rate": 7.853102700584483e-06, "loss": 0.4342, "step": 4620 }, { "epoch": 0.3276143211627083, "grad_norm": 2.8049659990335165, "learning_rate": 7.852159741289265e-06, "loss": 0.4289, "step": 4621 }, { "epoch": 0.32768521800779865, "grad_norm": 3.901435641865376, "learning_rate": 7.851216631597485e-06, "loss": 0.4497, "step": 4622 }, { "epoch": 0.327756114852889, "grad_norm": 2.1601985677384983, "learning_rate": 7.85027337155888e-06, "loss": 0.478, "step": 4623 }, { "epoch": 0.32782701169797945, "grad_norm": 3.2331683303561234, "learning_rate": 7.84932996122318e-06, "loss": 0.4605, "step": 4624 }, { "epoch": 0.3278979085430698, "grad_norm": 19.93064377935946, "learning_rate": 7.84838640064014e-06, "loss": 0.469, "step": 4625 }, { "epoch": 0.3279688053881602, "grad_norm": 2.093718164394236, "learning_rate": 7.84744268985951e-06, "loss": 0.4312, "step": 4626 }, { "epoch": 0.32803970223325063, "grad_norm": 1.6934534850133047, "learning_rate": 7.84649882893105e-06, "loss": 0.4191, "step": 4627 }, { "epoch": 0.328110599078341, "grad_norm": 4.675734215924519, "learning_rate": 7.845554817904539e-06, "loss": 0.4632, "step": 4628 }, { "epoch": 0.32818149592343143, "grad_norm": 3.198464760610885, "learning_rate": 7.84461065682975e-06, "loss": 0.4404, "step": 4629 }, { "epoch": 0.3282523927685218, "grad_norm": 8.966769735478943, "learning_rate": 7.843666345756467e-06, "loss": 0.4227, "step": 4630 }, { "epoch": 0.3283232896136122, "grad_norm": 2.818010969596728, "learning_rate": 7.842721884734488e-06, "loss": 0.4272, "step": 4631 }, { "epoch": 0.3283941864587026, "grad_norm": 5.736996388678994, "learning_rate": 7.841777273813616e-06, "loss": 0.4608, "step": 4632 }, { "epoch": 0.328465083303793, "grad_norm": 2.4430849322607884, "learning_rate": 7.840832513043657e-06, "loss": 0.3959, "step": 4633 }, { "epoch": 0.32853598014888336, "grad_norm": 3.3862481821092656, "learning_rate": 7.839887602474434e-06, "loss": 0.4448, "step": 4634 }, { "epoch": 0.3286068769939738, "grad_norm": 3.896996332660037, "learning_rate": 7.838942542155768e-06, "loss": 0.455, "step": 4635 }, { "epoch": 0.32867777383906416, "grad_norm": 3.6245509603081523, "learning_rate": 7.837997332137494e-06, "loss": 0.4175, "step": 4636 }, { "epoch": 0.32874867068415453, "grad_norm": 2.4185084551742433, "learning_rate": 7.837051972469458e-06, "loss": 0.4395, "step": 4637 }, { "epoch": 0.32881956752924496, "grad_norm": 4.269981310827187, "learning_rate": 7.836106463201505e-06, "loss": 0.3752, "step": 4638 }, { "epoch": 0.32889046437433533, "grad_norm": 4.037901481951159, "learning_rate": 7.835160804383491e-06, "loss": 0.4272, "step": 4639 }, { "epoch": 0.32896136121942576, "grad_norm": 2.515674543721822, "learning_rate": 7.834214996065286e-06, "loss": 0.432, "step": 4640 }, { "epoch": 0.32903225806451614, "grad_norm": 4.449725989771304, "learning_rate": 7.83326903829676e-06, "loss": 0.4402, "step": 4641 }, { "epoch": 0.3291031549096065, "grad_norm": 3.648856758424005, "learning_rate": 7.832322931127796e-06, "loss": 0.4191, "step": 4642 }, { "epoch": 0.32917405175469694, "grad_norm": 3.604083840017503, "learning_rate": 7.831376674608281e-06, "loss": 0.4159, "step": 4643 }, { "epoch": 0.3292449485997873, "grad_norm": 7.047984186154638, "learning_rate": 7.830430268788114e-06, "loss": 0.4655, "step": 4644 }, { "epoch": 0.3293158454448777, "grad_norm": 5.951297107386688, "learning_rate": 7.829483713717199e-06, "loss": 0.4621, "step": 4645 }, { "epoch": 0.3293867422899681, "grad_norm": 5.854176414640361, "learning_rate": 7.828537009445447e-06, "loss": 0.393, "step": 4646 }, { "epoch": 0.3294576391350585, "grad_norm": 2.605049065751092, "learning_rate": 7.82759015602278e-06, "loss": 0.4371, "step": 4647 }, { "epoch": 0.32952853598014886, "grad_norm": 4.18243346804372, "learning_rate": 7.826643153499124e-06, "loss": 0.4651, "step": 4648 }, { "epoch": 0.3295994328252393, "grad_norm": 3.534218683462921, "learning_rate": 7.82569600192442e-06, "loss": 0.3904, "step": 4649 }, { "epoch": 0.32967032967032966, "grad_norm": 7.5921585095734905, "learning_rate": 7.824748701348608e-06, "loss": 0.4277, "step": 4650 }, { "epoch": 0.32974122651542004, "grad_norm": 2.5679174641499554, "learning_rate": 7.823801251821639e-06, "loss": 0.3856, "step": 4651 }, { "epoch": 0.32981212336051047, "grad_norm": 5.625723189022281, "learning_rate": 7.822853653393476e-06, "loss": 0.4792, "step": 4652 }, { "epoch": 0.32988302020560084, "grad_norm": 3.479101734715476, "learning_rate": 7.821905906114086e-06, "loss": 0.5023, "step": 4653 }, { "epoch": 0.32995391705069127, "grad_norm": 3.481015996534996, "learning_rate": 7.820958010033444e-06, "loss": 0.3964, "step": 4654 }, { "epoch": 0.33002481389578164, "grad_norm": 3.7569659720347834, "learning_rate": 7.820009965201532e-06, "loss": 0.46, "step": 4655 }, { "epoch": 0.330095710740872, "grad_norm": 3.0244432304640005, "learning_rate": 7.819061771668341e-06, "loss": 0.4349, "step": 4656 }, { "epoch": 0.33016660758596245, "grad_norm": 3.2492928726594483, "learning_rate": 7.818113429483873e-06, "loss": 0.445, "step": 4657 }, { "epoch": 0.3302375044310528, "grad_norm": 3.5784247870631547, "learning_rate": 7.817164938698133e-06, "loss": 0.4111, "step": 4658 }, { "epoch": 0.3303084012761432, "grad_norm": 3.2117181214486346, "learning_rate": 7.816216299361134e-06, "loss": 0.4135, "step": 4659 }, { "epoch": 0.3303792981212336, "grad_norm": 5.976814531873193, "learning_rate": 7.815267511522901e-06, "loss": 0.3761, "step": 4660 }, { "epoch": 0.330450194966324, "grad_norm": 5.700875283742716, "learning_rate": 7.814318575233463e-06, "loss": 0.4012, "step": 4661 }, { "epoch": 0.33052109181141437, "grad_norm": 6.753027164843201, "learning_rate": 7.813369490542859e-06, "loss": 0.499, "step": 4662 }, { "epoch": 0.3305919886565048, "grad_norm": 3.437860709835309, "learning_rate": 7.812420257501134e-06, "loss": 0.4334, "step": 4663 }, { "epoch": 0.33066288550159517, "grad_norm": 4.712714303989625, "learning_rate": 7.811470876158342e-06, "loss": 0.3921, "step": 4664 }, { "epoch": 0.33073378234668555, "grad_norm": 3.3550344010308453, "learning_rate": 7.810521346564545e-06, "loss": 0.4453, "step": 4665 }, { "epoch": 0.330804679191776, "grad_norm": 3.726601761544386, "learning_rate": 7.809571668769813e-06, "loss": 0.4337, "step": 4666 }, { "epoch": 0.33087557603686635, "grad_norm": 2.660148289539854, "learning_rate": 7.808621842824221e-06, "loss": 0.3901, "step": 4667 }, { "epoch": 0.3309464728819568, "grad_norm": 5.9046916634874, "learning_rate": 7.807671868777857e-06, "loss": 0.392, "step": 4668 }, { "epoch": 0.33101736972704715, "grad_norm": 3.3086280646114257, "learning_rate": 7.806721746680812e-06, "loss": 0.4284, "step": 4669 }, { "epoch": 0.3310882665721375, "grad_norm": 2.8796480623268446, "learning_rate": 7.805771476583186e-06, "loss": 0.452, "step": 4670 }, { "epoch": 0.33115916341722795, "grad_norm": 6.130327513944081, "learning_rate": 7.804821058535089e-06, "loss": 0.3874, "step": 4671 }, { "epoch": 0.3312300602623183, "grad_norm": 2.861252856422674, "learning_rate": 7.803870492586636e-06, "loss": 0.4405, "step": 4672 }, { "epoch": 0.3313009571074087, "grad_norm": 3.100785581696416, "learning_rate": 7.802919778787955e-06, "loss": 0.4351, "step": 4673 }, { "epoch": 0.33137185395249913, "grad_norm": 3.9187953653215573, "learning_rate": 7.801968917189171e-06, "loss": 0.4078, "step": 4674 }, { "epoch": 0.3314427507975895, "grad_norm": 1.991683619259027, "learning_rate": 7.80101790784043e-06, "loss": 0.4293, "step": 4675 }, { "epoch": 0.3315136476426799, "grad_norm": 2.8702192343127217, "learning_rate": 7.800066750791875e-06, "loss": 0.4484, "step": 4676 }, { "epoch": 0.3315845444877703, "grad_norm": 2.381841709888876, "learning_rate": 7.799115446093663e-06, "loss": 0.4706, "step": 4677 }, { "epoch": 0.3316554413328607, "grad_norm": 4.263320072661332, "learning_rate": 7.798163993795957e-06, "loss": 0.4406, "step": 4678 }, { "epoch": 0.3317263381779511, "grad_norm": 2.8783523129039357, "learning_rate": 7.79721239394893e-06, "loss": 0.4315, "step": 4679 }, { "epoch": 0.3317972350230415, "grad_norm": 5.246169359450789, "learning_rate": 7.796260646602755e-06, "loss": 0.4487, "step": 4680 }, { "epoch": 0.33186813186813185, "grad_norm": 2.0684728712986704, "learning_rate": 7.795308751807625e-06, "loss": 0.3845, "step": 4681 }, { "epoch": 0.3319390287132223, "grad_norm": 2.9490180786428257, "learning_rate": 7.794356709613731e-06, "loss": 0.455, "step": 4682 }, { "epoch": 0.33200992555831266, "grad_norm": 2.6321354768846317, "learning_rate": 7.793404520071272e-06, "loss": 0.4291, "step": 4683 }, { "epoch": 0.33208082240340303, "grad_norm": 3.127844107756317, "learning_rate": 7.792452183230463e-06, "loss": 0.3898, "step": 4684 }, { "epoch": 0.33215171924849346, "grad_norm": 3.230661876403859, "learning_rate": 7.791499699141519e-06, "loss": 0.5099, "step": 4685 }, { "epoch": 0.33222261609358383, "grad_norm": 3.6183338249367925, "learning_rate": 7.790547067854664e-06, "loss": 0.425, "step": 4686 }, { "epoch": 0.3322935129386742, "grad_norm": 1.90895159395894, "learning_rate": 7.789594289420132e-06, "loss": 0.4376, "step": 4687 }, { "epoch": 0.33236440978376464, "grad_norm": 3.2284260398206572, "learning_rate": 7.788641363888164e-06, "loss": 0.4326, "step": 4688 }, { "epoch": 0.332435306628855, "grad_norm": 3.7950153143131438, "learning_rate": 7.78768829130901e-06, "loss": 0.4024, "step": 4689 }, { "epoch": 0.3325062034739454, "grad_norm": 2.8299457478003847, "learning_rate": 7.786735071732923e-06, "loss": 0.4154, "step": 4690 }, { "epoch": 0.3325771003190358, "grad_norm": 7.3084841515563275, "learning_rate": 7.78578170521017e-06, "loss": 0.3979, "step": 4691 }, { "epoch": 0.3326479971641262, "grad_norm": 2.7869163896419216, "learning_rate": 7.784828191791021e-06, "loss": 0.3906, "step": 4692 }, { "epoch": 0.3327188940092166, "grad_norm": 4.3479532935490175, "learning_rate": 7.783874531525755e-06, "loss": 0.4223, "step": 4693 }, { "epoch": 0.332789790854307, "grad_norm": 5.367723992738072, "learning_rate": 7.78292072446466e-06, "loss": 0.404, "step": 4694 }, { "epoch": 0.33286068769939736, "grad_norm": 4.7536754933452725, "learning_rate": 7.78196677065803e-06, "loss": 0.4331, "step": 4695 }, { "epoch": 0.3329315845444878, "grad_norm": 4.009974729639407, "learning_rate": 7.781012670156171e-06, "loss": 0.4974, "step": 4696 }, { "epoch": 0.33300248138957816, "grad_norm": 1.9593296526097943, "learning_rate": 7.780058423009391e-06, "loss": 0.4263, "step": 4697 }, { "epoch": 0.33307337823466854, "grad_norm": 2.621307982456548, "learning_rate": 7.779104029268008e-06, "loss": 0.443, "step": 4698 }, { "epoch": 0.33314427507975897, "grad_norm": 2.910972586656098, "learning_rate": 7.778149488982349e-06, "loss": 0.4179, "step": 4699 }, { "epoch": 0.33321517192484934, "grad_norm": 7.432044687411545, "learning_rate": 7.777194802202747e-06, "loss": 0.4669, "step": 4700 }, { "epoch": 0.3332860687699397, "grad_norm": 1.31645656762181, "learning_rate": 7.776239968979542e-06, "loss": 0.6869, "step": 4701 }, { "epoch": 0.33335696561503014, "grad_norm": 2.350792436784367, "learning_rate": 7.775284989363086e-06, "loss": 0.459, "step": 4702 }, { "epoch": 0.3334278624601205, "grad_norm": 1.859536329846472, "learning_rate": 7.774329863403733e-06, "loss": 0.4104, "step": 4703 }, { "epoch": 0.33349875930521095, "grad_norm": 3.998986704123604, "learning_rate": 7.773374591151849e-06, "loss": 0.3684, "step": 4704 }, { "epoch": 0.3335696561503013, "grad_norm": 4.128476808647899, "learning_rate": 7.772419172657807e-06, "loss": 0.409, "step": 4705 }, { "epoch": 0.3336405529953917, "grad_norm": 5.254131529197511, "learning_rate": 7.771463607971985e-06, "loss": 0.491, "step": 4706 }, { "epoch": 0.3337114498404821, "grad_norm": 3.0504708528914564, "learning_rate": 7.770507897144772e-06, "loss": 0.4403, "step": 4707 }, { "epoch": 0.3337823466855725, "grad_norm": 3.149085034774711, "learning_rate": 7.769552040226562e-06, "loss": 0.404, "step": 4708 }, { "epoch": 0.33385324353066287, "grad_norm": 3.35280247278883, "learning_rate": 7.768596037267759e-06, "loss": 0.4151, "step": 4709 }, { "epoch": 0.3339241403757533, "grad_norm": 3.6188173510028734, "learning_rate": 7.767639888318776e-06, "loss": 0.4794, "step": 4710 }, { "epoch": 0.33399503722084367, "grad_norm": 3.0958679063778676, "learning_rate": 7.766683593430029e-06, "loss": 0.4346, "step": 4711 }, { "epoch": 0.33406593406593404, "grad_norm": 2.547479878183925, "learning_rate": 7.765727152651942e-06, "loss": 0.3904, "step": 4712 }, { "epoch": 0.3341368309110245, "grad_norm": 4.006240681162526, "learning_rate": 7.764770566034953e-06, "loss": 0.4174, "step": 4713 }, { "epoch": 0.33420772775611485, "grad_norm": 2.371565573634698, "learning_rate": 7.7638138336295e-06, "loss": 0.427, "step": 4714 }, { "epoch": 0.3342786246012052, "grad_norm": 1.7881765457939474, "learning_rate": 7.762856955486036e-06, "loss": 0.3701, "step": 4715 }, { "epoch": 0.33434952144629565, "grad_norm": 2.4109795116321027, "learning_rate": 7.761899931655015e-06, "loss": 0.4551, "step": 4716 }, { "epoch": 0.334420418291386, "grad_norm": 5.342058855090381, "learning_rate": 7.760942762186902e-06, "loss": 0.4105, "step": 4717 }, { "epoch": 0.33449131513647645, "grad_norm": 2.387268189588131, "learning_rate": 7.759985447132172e-06, "loss": 0.4365, "step": 4718 }, { "epoch": 0.3345622119815668, "grad_norm": 4.082539880175432, "learning_rate": 7.759027986541302e-06, "loss": 0.4338, "step": 4719 }, { "epoch": 0.3346331088266572, "grad_norm": 5.750147507083218, "learning_rate": 7.758070380464778e-06, "loss": 0.4606, "step": 4720 }, { "epoch": 0.33470400567174763, "grad_norm": 3.3443240656438324, "learning_rate": 7.7571126289531e-06, "loss": 0.43, "step": 4721 }, { "epoch": 0.334774902516838, "grad_norm": 2.3312023085691767, "learning_rate": 7.756154732056767e-06, "loss": 0.4268, "step": 4722 }, { "epoch": 0.3348457993619284, "grad_norm": 3.106939868207488, "learning_rate": 7.755196689826293e-06, "loss": 0.4018, "step": 4723 }, { "epoch": 0.3349166962070188, "grad_norm": 1.8994392227513264, "learning_rate": 7.754238502312195e-06, "loss": 0.3719, "step": 4724 }, { "epoch": 0.3349875930521092, "grad_norm": 3.4007493977855328, "learning_rate": 7.753280169564996e-06, "loss": 0.4281, "step": 4725 }, { "epoch": 0.33505848989719955, "grad_norm": 11.268876962617862, "learning_rate": 7.752321691635234e-06, "loss": 0.5283, "step": 4726 }, { "epoch": 0.33512938674229, "grad_norm": 2.796821324867597, "learning_rate": 7.751363068573448e-06, "loss": 0.4068, "step": 4727 }, { "epoch": 0.33520028358738035, "grad_norm": 4.821008924901002, "learning_rate": 7.750404300430188e-06, "loss": 0.4746, "step": 4728 }, { "epoch": 0.33527118043247073, "grad_norm": 4.619881218941458, "learning_rate": 7.74944538725601e-06, "loss": 0.4172, "step": 4729 }, { "epoch": 0.33534207727756116, "grad_norm": 4.909206914730786, "learning_rate": 7.748486329101477e-06, "loss": 0.4536, "step": 4730 }, { "epoch": 0.33541297412265153, "grad_norm": 3.1973623627874974, "learning_rate": 7.747527126017163e-06, "loss": 0.4206, "step": 4731 }, { "epoch": 0.33548387096774196, "grad_norm": 3.1983387036442794, "learning_rate": 7.746567778053646e-06, "loss": 0.39, "step": 4732 }, { "epoch": 0.33555476781283233, "grad_norm": 4.952085817798908, "learning_rate": 7.745608285261514e-06, "loss": 0.4317, "step": 4733 }, { "epoch": 0.3356256646579227, "grad_norm": 2.6216663188564926, "learning_rate": 7.744648647691361e-06, "loss": 0.4498, "step": 4734 }, { "epoch": 0.33569656150301314, "grad_norm": 7.75522293187235, "learning_rate": 7.74368886539379e-06, "loss": 0.4201, "step": 4735 }, { "epoch": 0.3357674583481035, "grad_norm": 2.2236523359119644, "learning_rate": 7.742728938419413e-06, "loss": 0.4031, "step": 4736 }, { "epoch": 0.3358383551931939, "grad_norm": 1.9969130181691463, "learning_rate": 7.741768866818844e-06, "loss": 0.4041, "step": 4737 }, { "epoch": 0.3359092520382843, "grad_norm": 4.2302215834801, "learning_rate": 7.74080865064271e-06, "loss": 0.46, "step": 4738 }, { "epoch": 0.3359801488833747, "grad_norm": 16.751285933768777, "learning_rate": 7.739848289941642e-06, "loss": 0.394, "step": 4739 }, { "epoch": 0.33605104572846506, "grad_norm": 3.317060840510915, "learning_rate": 7.738887784766286e-06, "loss": 0.4161, "step": 4740 }, { "epoch": 0.3361219425735555, "grad_norm": 4.184064120605504, "learning_rate": 7.737927135167285e-06, "loss": 0.4482, "step": 4741 }, { "epoch": 0.33619283941864586, "grad_norm": 11.184638974112792, "learning_rate": 7.736966341195296e-06, "loss": 0.4171, "step": 4742 }, { "epoch": 0.3362637362637363, "grad_norm": 2.853324305627257, "learning_rate": 7.736005402900983e-06, "loss": 0.4452, "step": 4743 }, { "epoch": 0.33633463310882666, "grad_norm": 4.038055191714402, "learning_rate": 7.735044320335016e-06, "loss": 0.4138, "step": 4744 }, { "epoch": 0.33640552995391704, "grad_norm": 4.349186835699139, "learning_rate": 7.734083093548077e-06, "loss": 0.4451, "step": 4745 }, { "epoch": 0.33647642679900747, "grad_norm": 3.1452028461136203, "learning_rate": 7.733121722590846e-06, "loss": 0.3958, "step": 4746 }, { "epoch": 0.33654732364409784, "grad_norm": 6.448420871646004, "learning_rate": 7.732160207514023e-06, "loss": 0.3943, "step": 4747 }, { "epoch": 0.3366182204891882, "grad_norm": 3.8037600213504947, "learning_rate": 7.731198548368306e-06, "loss": 0.4227, "step": 4748 }, { "epoch": 0.33668911733427864, "grad_norm": 11.904118691423715, "learning_rate": 7.730236745204406e-06, "loss": 0.3947, "step": 4749 }, { "epoch": 0.336760014179369, "grad_norm": 3.3294706618336924, "learning_rate": 7.729274798073038e-06, "loss": 0.4259, "step": 4750 }, { "epoch": 0.3368309110244594, "grad_norm": 5.46921846912596, "learning_rate": 7.728312707024925e-06, "loss": 0.4263, "step": 4751 }, { "epoch": 0.3369018078695498, "grad_norm": 4.377878169932648, "learning_rate": 7.727350472110804e-06, "loss": 0.4284, "step": 4752 }, { "epoch": 0.3369727047146402, "grad_norm": 1.1519511458193288, "learning_rate": 7.726388093381408e-06, "loss": 0.718, "step": 4753 }, { "epoch": 0.33704360155973057, "grad_norm": 7.828968435831155, "learning_rate": 7.725425570887488e-06, "loss": 0.4507, "step": 4754 }, { "epoch": 0.337114498404821, "grad_norm": 2.666628964726796, "learning_rate": 7.724462904679797e-06, "loss": 0.4272, "step": 4755 }, { "epoch": 0.33718539524991137, "grad_norm": 3.9781725704121706, "learning_rate": 7.723500094809097e-06, "loss": 0.4374, "step": 4756 }, { "epoch": 0.3372562920950018, "grad_norm": 3.008120001448175, "learning_rate": 7.72253714132616e-06, "loss": 0.3947, "step": 4757 }, { "epoch": 0.33732718894009217, "grad_norm": 4.430969621961241, "learning_rate": 7.72157404428176e-06, "loss": 0.4521, "step": 4758 }, { "epoch": 0.33739808578518254, "grad_norm": 2.99984214680682, "learning_rate": 7.720610803726685e-06, "loss": 0.4278, "step": 4759 }, { "epoch": 0.337468982630273, "grad_norm": 3.3238802432975576, "learning_rate": 7.719647419711724e-06, "loss": 0.4178, "step": 4760 }, { "epoch": 0.33753987947536335, "grad_norm": 3.0001194574130725, "learning_rate": 7.71868389228768e-06, "loss": 0.4134, "step": 4761 }, { "epoch": 0.3376107763204537, "grad_norm": 1.814085044919389, "learning_rate": 7.717720221505359e-06, "loss": 0.3795, "step": 4762 }, { "epoch": 0.33768167316554415, "grad_norm": 2.7733857757067577, "learning_rate": 7.716756407415575e-06, "loss": 0.4072, "step": 4763 }, { "epoch": 0.3377525700106345, "grad_norm": 5.77596048234758, "learning_rate": 7.715792450069152e-06, "loss": 0.4105, "step": 4764 }, { "epoch": 0.3378234668557249, "grad_norm": 3.7516183303854236, "learning_rate": 7.714828349516924e-06, "loss": 0.4061, "step": 4765 }, { "epoch": 0.3378943637008153, "grad_norm": 5.721240470121473, "learning_rate": 7.713864105809722e-06, "loss": 0.4485, "step": 4766 }, { "epoch": 0.3379652605459057, "grad_norm": 1.1310037921547882, "learning_rate": 7.712899718998394e-06, "loss": 0.6474, "step": 4767 }, { "epoch": 0.33803615739099613, "grad_norm": 6.01619247727875, "learning_rate": 7.711935189133794e-06, "loss": 0.4144, "step": 4768 }, { "epoch": 0.3381070542360865, "grad_norm": 3.551034939835278, "learning_rate": 7.710970516266782e-06, "loss": 0.4096, "step": 4769 }, { "epoch": 0.3381779510811769, "grad_norm": 2.28487207717512, "learning_rate": 7.710005700448225e-06, "loss": 0.3699, "step": 4770 }, { "epoch": 0.3382488479262673, "grad_norm": 2.7210673203078333, "learning_rate": 7.709040741728998e-06, "loss": 0.4313, "step": 4771 }, { "epoch": 0.3383197447713577, "grad_norm": 6.50555210898806, "learning_rate": 7.708075640159985e-06, "loss": 0.398, "step": 4772 }, { "epoch": 0.33839064161644805, "grad_norm": 4.822908070996737, "learning_rate": 7.707110395792077e-06, "loss": 0.3851, "step": 4773 }, { "epoch": 0.3384615384615385, "grad_norm": 6.452288106937592, "learning_rate": 7.706145008676174e-06, "loss": 0.4765, "step": 4774 }, { "epoch": 0.33853243530662885, "grad_norm": 4.160208629940918, "learning_rate": 7.705179478863176e-06, "loss": 0.43, "step": 4775 }, { "epoch": 0.3386033321517192, "grad_norm": 5.575665282320722, "learning_rate": 7.704213806404e-06, "loss": 0.4461, "step": 4776 }, { "epoch": 0.33867422899680966, "grad_norm": 3.915205714074848, "learning_rate": 7.703247991349567e-06, "loss": 0.4361, "step": 4777 }, { "epoch": 0.33874512584190003, "grad_norm": 2.4474322651399785, "learning_rate": 7.702282033750804e-06, "loss": 0.453, "step": 4778 }, { "epoch": 0.3388160226869904, "grad_norm": 2.8636914067025105, "learning_rate": 7.701315933658646e-06, "loss": 0.4393, "step": 4779 }, { "epoch": 0.33888691953208083, "grad_norm": 4.908709111624847, "learning_rate": 7.700349691124037e-06, "loss": 0.4671, "step": 4780 }, { "epoch": 0.3389578163771712, "grad_norm": 2.921684993774227, "learning_rate": 7.699383306197928e-06, "loss": 0.4065, "step": 4781 }, { "epoch": 0.33902871322226164, "grad_norm": 3.9102502920684135, "learning_rate": 7.698416778931278e-06, "loss": 0.436, "step": 4782 }, { "epoch": 0.339099610067352, "grad_norm": 2.1961320803562296, "learning_rate": 7.697450109375049e-06, "loss": 0.3911, "step": 4783 }, { "epoch": 0.3391705069124424, "grad_norm": 1.8339291602616596, "learning_rate": 7.696483297580218e-06, "loss": 0.438, "step": 4784 }, { "epoch": 0.3392414037575328, "grad_norm": 1.9826982938442854, "learning_rate": 7.695516343597765e-06, "loss": 0.3816, "step": 4785 }, { "epoch": 0.3393123006026232, "grad_norm": 4.809546782053114, "learning_rate": 7.694549247478678e-06, "loss": 0.3942, "step": 4786 }, { "epoch": 0.33938319744771356, "grad_norm": 2.8597142144631054, "learning_rate": 7.693582009273953e-06, "loss": 0.4009, "step": 4787 }, { "epoch": 0.339454094292804, "grad_norm": 9.427027808344532, "learning_rate": 7.692614629034591e-06, "loss": 0.3922, "step": 4788 }, { "epoch": 0.33952499113789436, "grad_norm": 3.48307526835814, "learning_rate": 7.691647106811606e-06, "loss": 0.3975, "step": 4789 }, { "epoch": 0.33959588798298473, "grad_norm": 3.218830143714081, "learning_rate": 7.690679442656015e-06, "loss": 0.4234, "step": 4790 }, { "epoch": 0.33966678482807516, "grad_norm": 4.468155390639087, "learning_rate": 7.689711636618842e-06, "loss": 0.4175, "step": 4791 }, { "epoch": 0.33973768167316554, "grad_norm": 2.9099797459668717, "learning_rate": 7.688743688751122e-06, "loss": 0.4651, "step": 4792 }, { "epoch": 0.3398085785182559, "grad_norm": 4.980780845056034, "learning_rate": 7.687775599103894e-06, "loss": 0.4416, "step": 4793 }, { "epoch": 0.33987947536334634, "grad_norm": 2.6454167777902367, "learning_rate": 7.686807367728209e-06, "loss": 0.5012, "step": 4794 }, { "epoch": 0.3399503722084367, "grad_norm": 2.5815370241436346, "learning_rate": 7.685838994675119e-06, "loss": 0.3632, "step": 4795 }, { "epoch": 0.34002126905352714, "grad_norm": 6.44887451953276, "learning_rate": 7.68487047999569e-06, "loss": 0.4483, "step": 4796 }, { "epoch": 0.3400921658986175, "grad_norm": 1.1782184012859953, "learning_rate": 7.683901823740992e-06, "loss": 0.6719, "step": 4797 }, { "epoch": 0.3401630627437079, "grad_norm": 4.49403955347649, "learning_rate": 7.682933025962101e-06, "loss": 0.4015, "step": 4798 }, { "epoch": 0.3402339595887983, "grad_norm": 4.278341722511935, "learning_rate": 7.681964086710102e-06, "loss": 0.4054, "step": 4799 }, { "epoch": 0.3403048564338887, "grad_norm": 5.357001069305836, "learning_rate": 7.680995006036092e-06, "loss": 0.4002, "step": 4800 }, { "epoch": 0.34037575327897907, "grad_norm": 2.043417668830008, "learning_rate": 7.680025783991168e-06, "loss": 0.4025, "step": 4801 }, { "epoch": 0.3404466501240695, "grad_norm": 3.336931822015867, "learning_rate": 7.67905642062644e-06, "loss": 0.4146, "step": 4802 }, { "epoch": 0.34051754696915987, "grad_norm": 2.70281663736924, "learning_rate": 7.678086915993022e-06, "loss": 0.3785, "step": 4803 }, { "epoch": 0.34058844381425024, "grad_norm": 9.24252568178107, "learning_rate": 7.677117270142035e-06, "loss": 0.4691, "step": 4804 }, { "epoch": 0.34065934065934067, "grad_norm": 5.517945191073497, "learning_rate": 7.676147483124614e-06, "loss": 0.3846, "step": 4805 }, { "epoch": 0.34073023750443104, "grad_norm": 10.947904343172207, "learning_rate": 7.675177554991892e-06, "loss": 0.4538, "step": 4806 }, { "epoch": 0.3408011343495215, "grad_norm": 4.760297229798904, "learning_rate": 7.674207485795016e-06, "loss": 0.4736, "step": 4807 }, { "epoch": 0.34087203119461185, "grad_norm": 2.622407519315574, "learning_rate": 7.673237275585138e-06, "loss": 0.4613, "step": 4808 }, { "epoch": 0.3409429280397022, "grad_norm": 4.243326147221805, "learning_rate": 7.672266924413417e-06, "loss": 0.4715, "step": 4809 }, { "epoch": 0.34101382488479265, "grad_norm": 2.4218105302040556, "learning_rate": 7.671296432331023e-06, "loss": 0.4518, "step": 4810 }, { "epoch": 0.341084721729883, "grad_norm": 2.956328764966956, "learning_rate": 7.67032579938913e-06, "loss": 0.432, "step": 4811 }, { "epoch": 0.3411556185749734, "grad_norm": 3.3749799989928477, "learning_rate": 7.669355025638917e-06, "loss": 0.3533, "step": 4812 }, { "epoch": 0.3412265154200638, "grad_norm": 3.703953443633458, "learning_rate": 7.668384111131575e-06, "loss": 0.4591, "step": 4813 }, { "epoch": 0.3412974122651542, "grad_norm": 3.5313987001138916, "learning_rate": 7.667413055918307e-06, "loss": 0.4484, "step": 4814 }, { "epoch": 0.34136830911024457, "grad_norm": 5.511478702199945, "learning_rate": 7.66644186005031e-06, "loss": 0.4351, "step": 4815 }, { "epoch": 0.341439205955335, "grad_norm": 1.1445547369817701, "learning_rate": 7.665470523578798e-06, "loss": 0.6612, "step": 4816 }, { "epoch": 0.3415101028004254, "grad_norm": 0.946368395492316, "learning_rate": 7.664499046554989e-06, "loss": 0.7237, "step": 4817 }, { "epoch": 0.34158099964551575, "grad_norm": 3.2138011882840725, "learning_rate": 7.663527429030115e-06, "loss": 0.4624, "step": 4818 }, { "epoch": 0.3416518964906062, "grad_norm": 3.5325050822458643, "learning_rate": 7.662555671055407e-06, "loss": 0.4485, "step": 4819 }, { "epoch": 0.34172279333569655, "grad_norm": 2.435781601745427, "learning_rate": 7.661583772682103e-06, "loss": 0.4694, "step": 4820 }, { "epoch": 0.341793690180787, "grad_norm": 3.936913023191963, "learning_rate": 7.660611733961458e-06, "loss": 0.4501, "step": 4821 }, { "epoch": 0.34186458702587735, "grad_norm": 4.588811689306067, "learning_rate": 7.659639554944722e-06, "loss": 0.4801, "step": 4822 }, { "epoch": 0.3419354838709677, "grad_norm": 7.736275418182444, "learning_rate": 7.658667235683167e-06, "loss": 0.4913, "step": 4823 }, { "epoch": 0.34200638071605816, "grad_norm": 2.7236750234214777, "learning_rate": 7.657694776228055e-06, "loss": 0.5178, "step": 4824 }, { "epoch": 0.34207727756114853, "grad_norm": 2.3171913612022026, "learning_rate": 7.656722176630672e-06, "loss": 0.415, "step": 4825 }, { "epoch": 0.3421481744062389, "grad_norm": 2.1501885729028856, "learning_rate": 7.6557494369423e-06, "loss": 0.4061, "step": 4826 }, { "epoch": 0.34221907125132933, "grad_norm": 2.6443511044347017, "learning_rate": 7.654776557214232e-06, "loss": 0.4577, "step": 4827 }, { "epoch": 0.3422899680964197, "grad_norm": 28.475505718724904, "learning_rate": 7.65380353749777e-06, "loss": 0.4165, "step": 4828 }, { "epoch": 0.3423608649415101, "grad_norm": 11.656775601673301, "learning_rate": 7.652830377844224e-06, "loss": 0.4038, "step": 4829 }, { "epoch": 0.3424317617866005, "grad_norm": 4.415190080128498, "learning_rate": 7.651857078304906e-06, "loss": 0.406, "step": 4830 }, { "epoch": 0.3425026586316909, "grad_norm": 5.365387203135361, "learning_rate": 7.65088363893114e-06, "loss": 0.4096, "step": 4831 }, { "epoch": 0.3425735554767813, "grad_norm": 6.444218284435629, "learning_rate": 7.649910059774256e-06, "loss": 0.475, "step": 4832 }, { "epoch": 0.3426444523218717, "grad_norm": 4.66114619931385, "learning_rate": 7.648936340885591e-06, "loss": 0.4443, "step": 4833 }, { "epoch": 0.34271534916696206, "grad_norm": 4.3935879793860915, "learning_rate": 7.647962482316494e-06, "loss": 0.4552, "step": 4834 }, { "epoch": 0.3427862460120525, "grad_norm": 4.244904136773934, "learning_rate": 7.646988484118314e-06, "loss": 0.5226, "step": 4835 }, { "epoch": 0.34285714285714286, "grad_norm": 4.189361422253154, "learning_rate": 7.646014346342408e-06, "loss": 0.4492, "step": 4836 }, { "epoch": 0.34292803970223323, "grad_norm": 2.915391185947659, "learning_rate": 7.645040069040148e-06, "loss": 0.449, "step": 4837 }, { "epoch": 0.34299893654732366, "grad_norm": 3.1861395844332003, "learning_rate": 7.644065652262906e-06, "loss": 0.4267, "step": 4838 }, { "epoch": 0.34306983339241404, "grad_norm": 3.8192931531724037, "learning_rate": 7.643091096062063e-06, "loss": 0.4428, "step": 4839 }, { "epoch": 0.3431407302375044, "grad_norm": 5.191967689513851, "learning_rate": 7.64211640048901e-06, "loss": 0.449, "step": 4840 }, { "epoch": 0.34321162708259484, "grad_norm": 3.086021765494287, "learning_rate": 7.641141565595143e-06, "loss": 0.4405, "step": 4841 }, { "epoch": 0.3432825239276852, "grad_norm": 2.2721950178560992, "learning_rate": 7.640166591431864e-06, "loss": 0.375, "step": 4842 }, { "epoch": 0.3433534207727756, "grad_norm": 5.141288447435788, "learning_rate": 7.639191478050588e-06, "loss": 0.47, "step": 4843 }, { "epoch": 0.343424317617866, "grad_norm": 2.762009711943025, "learning_rate": 7.638216225502728e-06, "loss": 0.4315, "step": 4844 }, { "epoch": 0.3434952144629564, "grad_norm": 2.7227580653397445, "learning_rate": 7.637240833839714e-06, "loss": 0.4115, "step": 4845 }, { "epoch": 0.3435661113080468, "grad_norm": 2.8137366791950913, "learning_rate": 7.636265303112977e-06, "loss": 0.3811, "step": 4846 }, { "epoch": 0.3436370081531372, "grad_norm": 3.8131899322329637, "learning_rate": 7.635289633373961e-06, "loss": 0.4484, "step": 4847 }, { "epoch": 0.34370790499822756, "grad_norm": 4.007248597213708, "learning_rate": 7.63431382467411e-06, "loss": 0.3894, "step": 4848 }, { "epoch": 0.343778801843318, "grad_norm": 5.733701574674063, "learning_rate": 7.63333787706488e-06, "loss": 0.4694, "step": 4849 }, { "epoch": 0.34384969868840837, "grad_norm": 3.0132017454675517, "learning_rate": 7.632361790597732e-06, "loss": 0.418, "step": 4850 }, { "epoch": 0.34392059553349874, "grad_norm": 9.870112660432094, "learning_rate": 7.631385565324138e-06, "loss": 0.4588, "step": 4851 }, { "epoch": 0.34399149237858917, "grad_norm": 2.7246248254250793, "learning_rate": 7.630409201295576e-06, "loss": 0.425, "step": 4852 }, { "epoch": 0.34406238922367954, "grad_norm": 2.5491632066015244, "learning_rate": 7.629432698563531e-06, "loss": 0.4239, "step": 4853 }, { "epoch": 0.3441332860687699, "grad_norm": 5.07961113910545, "learning_rate": 7.62845605717949e-06, "loss": 0.4248, "step": 4854 }, { "epoch": 0.34420418291386035, "grad_norm": 3.025312740292474, "learning_rate": 7.627479277194957e-06, "loss": 0.4041, "step": 4855 }, { "epoch": 0.3442750797589507, "grad_norm": 4.279050010660419, "learning_rate": 7.626502358661434e-06, "loss": 0.4207, "step": 4856 }, { "epoch": 0.3443459766040411, "grad_norm": 2.4428502259783444, "learning_rate": 7.6255253016304384e-06, "loss": 0.4677, "step": 4857 }, { "epoch": 0.3444168734491315, "grad_norm": 3.531409424700501, "learning_rate": 7.6245481061534885e-06, "loss": 0.427, "step": 4858 }, { "epoch": 0.3444877702942219, "grad_norm": 2.469304784267372, "learning_rate": 7.623570772282115e-06, "loss": 0.4207, "step": 4859 }, { "epoch": 0.3445586671393123, "grad_norm": 1.1634607214874744, "learning_rate": 7.622593300067851e-06, "loss": 0.673, "step": 4860 }, { "epoch": 0.3446295639844027, "grad_norm": 3.9962343732758017, "learning_rate": 7.621615689562243e-06, "loss": 0.4665, "step": 4861 }, { "epoch": 0.34470046082949307, "grad_norm": 2.586478435553788, "learning_rate": 7.620637940816837e-06, "loss": 0.4506, "step": 4862 }, { "epoch": 0.3447713576745835, "grad_norm": 3.5378543495989545, "learning_rate": 7.619660053883191e-06, "loss": 0.3989, "step": 4863 }, { "epoch": 0.3448422545196739, "grad_norm": 2.2894840876126032, "learning_rate": 7.618682028812871e-06, "loss": 0.4455, "step": 4864 }, { "epoch": 0.34491315136476425, "grad_norm": 2.198472475887757, "learning_rate": 7.617703865657451e-06, "loss": 0.4482, "step": 4865 }, { "epoch": 0.3449840482098547, "grad_norm": 2.9527946560618066, "learning_rate": 7.616725564468508e-06, "loss": 0.4457, "step": 4866 }, { "epoch": 0.34505494505494505, "grad_norm": 2.662416031425578, "learning_rate": 7.6157471252976266e-06, "loss": 0.407, "step": 4867 }, { "epoch": 0.3451258419000354, "grad_norm": 5.020889840230455, "learning_rate": 7.614768548196405e-06, "loss": 0.4462, "step": 4868 }, { "epoch": 0.34519673874512585, "grad_norm": 3.893326388889299, "learning_rate": 7.61378983321644e-06, "loss": 0.4266, "step": 4869 }, { "epoch": 0.3452676355902162, "grad_norm": 3.1180203775405255, "learning_rate": 7.612810980409343e-06, "loss": 0.4116, "step": 4870 }, { "epoch": 0.34533853243530666, "grad_norm": 3.3432493206964002, "learning_rate": 7.611831989826729e-06, "loss": 0.4367, "step": 4871 }, { "epoch": 0.34540942928039703, "grad_norm": 3.65012985527572, "learning_rate": 7.610852861520222e-06, "loss": 0.3992, "step": 4872 }, { "epoch": 0.3454803261254874, "grad_norm": 4.209521086945963, "learning_rate": 7.609873595541448e-06, "loss": 0.4182, "step": 4873 }, { "epoch": 0.34555122297057783, "grad_norm": 7.703567727388106, "learning_rate": 7.608894191942049e-06, "loss": 0.4418, "step": 4874 }, { "epoch": 0.3456221198156682, "grad_norm": 2.398755589308795, "learning_rate": 7.607914650773666e-06, "loss": 0.3833, "step": 4875 }, { "epoch": 0.3456930166607586, "grad_norm": 1.9388113700745555, "learning_rate": 7.606934972087956e-06, "loss": 0.4299, "step": 4876 }, { "epoch": 0.345763913505849, "grad_norm": 3.973285537927678, "learning_rate": 7.6059551559365725e-06, "loss": 0.414, "step": 4877 }, { "epoch": 0.3458348103509394, "grad_norm": 2.914834832437174, "learning_rate": 7.6049752023711864e-06, "loss": 0.4543, "step": 4878 }, { "epoch": 0.34590570719602975, "grad_norm": 5.150991584421727, "learning_rate": 7.603995111443469e-06, "loss": 0.4543, "step": 4879 }, { "epoch": 0.3459766040411202, "grad_norm": 7.812466163707538, "learning_rate": 7.603014883205102e-06, "loss": 0.4183, "step": 4880 }, { "epoch": 0.34604750088621056, "grad_norm": 4.862460510438236, "learning_rate": 7.602034517707772e-06, "loss": 0.4607, "step": 4881 }, { "epoch": 0.34611839773130093, "grad_norm": 4.297525421432949, "learning_rate": 7.601054015003178e-06, "loss": 0.4491, "step": 4882 }, { "epoch": 0.34618929457639136, "grad_norm": 2.3705932235645726, "learning_rate": 7.60007337514302e-06, "loss": 0.3924, "step": 4883 }, { "epoch": 0.34626019142148173, "grad_norm": 15.02394739001483, "learning_rate": 7.599092598179008e-06, "loss": 0.4099, "step": 4884 }, { "epoch": 0.34633108826657216, "grad_norm": 15.925019850860899, "learning_rate": 7.598111684162859e-06, "loss": 0.4113, "step": 4885 }, { "epoch": 0.34640198511166254, "grad_norm": 2.0850752530781103, "learning_rate": 7.597130633146297e-06, "loss": 0.3945, "step": 4886 }, { "epoch": 0.3464728819567529, "grad_norm": 4.0351399023124435, "learning_rate": 7.596149445181057e-06, "loss": 0.4598, "step": 4887 }, { "epoch": 0.34654377880184334, "grad_norm": 6.642712469049589, "learning_rate": 7.5951681203188724e-06, "loss": 0.4176, "step": 4888 }, { "epoch": 0.3466146756469337, "grad_norm": 7.120723082053105, "learning_rate": 7.594186658611494e-06, "loss": 0.41, "step": 4889 }, { "epoch": 0.3466855724920241, "grad_norm": 7.600898072959622, "learning_rate": 7.593205060110673e-06, "loss": 0.4061, "step": 4890 }, { "epoch": 0.3467564693371145, "grad_norm": 2.8492076771439248, "learning_rate": 7.592223324868168e-06, "loss": 0.4464, "step": 4891 }, { "epoch": 0.3468273661822049, "grad_norm": 4.31624844970505, "learning_rate": 7.5912414529357515e-06, "loss": 0.3893, "step": 4892 }, { "epoch": 0.34689826302729526, "grad_norm": 4.986466948374489, "learning_rate": 7.590259444365193e-06, "loss": 0.4519, "step": 4893 }, { "epoch": 0.3469691598723857, "grad_norm": 3.4583559936506107, "learning_rate": 7.589277299208279e-06, "loss": 0.4087, "step": 4894 }, { "epoch": 0.34704005671747606, "grad_norm": 2.7706909497925425, "learning_rate": 7.588295017516795e-06, "loss": 0.4278, "step": 4895 }, { "epoch": 0.3471109535625665, "grad_norm": 2.5227421512135764, "learning_rate": 7.58731259934254e-06, "loss": 0.3875, "step": 4896 }, { "epoch": 0.34718185040765687, "grad_norm": 7.271838629240954, "learning_rate": 7.586330044737316e-06, "loss": 0.3549, "step": 4897 }, { "epoch": 0.34725274725274724, "grad_norm": 3.3080066154911334, "learning_rate": 7.585347353752935e-06, "loss": 0.4105, "step": 4898 }, { "epoch": 0.34732364409783767, "grad_norm": 2.409310415260885, "learning_rate": 7.584364526441214e-06, "loss": 0.4074, "step": 4899 }, { "epoch": 0.34739454094292804, "grad_norm": 6.715157355296827, "learning_rate": 7.583381562853981e-06, "loss": 0.4311, "step": 4900 }, { "epoch": 0.3474654377880184, "grad_norm": 4.9491544282976605, "learning_rate": 7.582398463043066e-06, "loss": 0.4205, "step": 4901 }, { "epoch": 0.34753633463310885, "grad_norm": 5.824529564285067, "learning_rate": 7.581415227060307e-06, "loss": 0.4073, "step": 4902 }, { "epoch": 0.3476072314781992, "grad_norm": 3.946038141181459, "learning_rate": 7.580431854957554e-06, "loss": 0.4708, "step": 4903 }, { "epoch": 0.3476781283232896, "grad_norm": 3.3685644177934027, "learning_rate": 7.5794483467866595e-06, "loss": 0.4127, "step": 4904 }, { "epoch": 0.34774902516838, "grad_norm": 6.099448988718645, "learning_rate": 7.5784647025994855e-06, "loss": 0.3876, "step": 4905 }, { "epoch": 0.3478199220134704, "grad_norm": 3.904862334935757, "learning_rate": 7.5774809224479e-06, "loss": 0.435, "step": 4906 }, { "epoch": 0.34789081885856077, "grad_norm": 3.9270651726919206, "learning_rate": 7.576497006383776e-06, "loss": 0.4227, "step": 4907 }, { "epoch": 0.3479617157036512, "grad_norm": 2.6416326710467217, "learning_rate": 7.575512954459e-06, "loss": 0.391, "step": 4908 }, { "epoch": 0.34803261254874157, "grad_norm": 3.237883398659406, "learning_rate": 7.57452876672546e-06, "loss": 0.4543, "step": 4909 }, { "epoch": 0.348103509393832, "grad_norm": 2.6346060149007515, "learning_rate": 7.5735444432350524e-06, "loss": 0.4241, "step": 4910 }, { "epoch": 0.3481744062389224, "grad_norm": 2.06125235016552, "learning_rate": 7.572559984039683e-06, "loss": 0.405, "step": 4911 }, { "epoch": 0.34824530308401275, "grad_norm": 4.880881735847494, "learning_rate": 7.571575389191261e-06, "loss": 0.4165, "step": 4912 }, { "epoch": 0.3483161999291032, "grad_norm": 3.7051542156717034, "learning_rate": 7.5705906587417076e-06, "loss": 0.376, "step": 4913 }, { "epoch": 0.34838709677419355, "grad_norm": 0.9930809817428364, "learning_rate": 7.5696057927429455e-06, "loss": 0.6494, "step": 4914 }, { "epoch": 0.3484579936192839, "grad_norm": 6.419707232669014, "learning_rate": 7.568620791246908e-06, "loss": 0.4086, "step": 4915 }, { "epoch": 0.34852889046437435, "grad_norm": 2.298230400486469, "learning_rate": 7.567635654305536e-06, "loss": 0.4724, "step": 4916 }, { "epoch": 0.3485997873094647, "grad_norm": 2.8815517503950687, "learning_rate": 7.566650381970777e-06, "loss": 0.4216, "step": 4917 }, { "epoch": 0.3486706841545551, "grad_norm": 11.421772538967538, "learning_rate": 7.565664974294584e-06, "loss": 0.4119, "step": 4918 }, { "epoch": 0.34874158099964553, "grad_norm": 2.893799949827121, "learning_rate": 7.564679431328919e-06, "loss": 0.4145, "step": 4919 }, { "epoch": 0.3488124778447359, "grad_norm": 2.5582501731163942, "learning_rate": 7.5636937531257494e-06, "loss": 0.415, "step": 4920 }, { "epoch": 0.3488833746898263, "grad_norm": 3.4217934049465413, "learning_rate": 7.562707939737053e-06, "loss": 0.4502, "step": 4921 }, { "epoch": 0.3489542715349167, "grad_norm": 4.15537501446647, "learning_rate": 7.561721991214809e-06, "loss": 0.4721, "step": 4922 }, { "epoch": 0.3490251683800071, "grad_norm": 1.0852107522042393, "learning_rate": 7.56073590761101e-06, "loss": 0.6873, "step": 4923 }, { "epoch": 0.3490960652250975, "grad_norm": 5.871262503904771, "learning_rate": 7.559749688977651e-06, "loss": 0.4206, "step": 4924 }, { "epoch": 0.3491669620701879, "grad_norm": 2.5918082398755007, "learning_rate": 7.5587633353667365e-06, "loss": 0.4034, "step": 4925 }, { "epoch": 0.34923785891527825, "grad_norm": 3.622417068545556, "learning_rate": 7.557776846830281e-06, "loss": 0.4403, "step": 4926 }, { "epoch": 0.3493087557603687, "grad_norm": 2.470125608378168, "learning_rate": 7.556790223420299e-06, "loss": 0.4062, "step": 4927 }, { "epoch": 0.34937965260545906, "grad_norm": 6.223941491302505, "learning_rate": 7.555803465188817e-06, "loss": 0.4254, "step": 4928 }, { "epoch": 0.34945054945054943, "grad_norm": 2.2457183952807833, "learning_rate": 7.5548165721878664e-06, "loss": 0.3973, "step": 4929 }, { "epoch": 0.34952144629563986, "grad_norm": 2.7364910992189726, "learning_rate": 7.553829544469491e-06, "loss": 0.4218, "step": 4930 }, { "epoch": 0.34959234314073023, "grad_norm": 4.16012031232902, "learning_rate": 7.55284238208573e-06, "loss": 0.4315, "step": 4931 }, { "epoch": 0.3496632399858206, "grad_norm": 41.577426275384845, "learning_rate": 7.5518550850886455e-06, "loss": 0.3947, "step": 4932 }, { "epoch": 0.34973413683091104, "grad_norm": 2.5480922638010184, "learning_rate": 7.550867653530291e-06, "loss": 0.4069, "step": 4933 }, { "epoch": 0.3498050336760014, "grad_norm": 2.9809038980337315, "learning_rate": 7.549880087462742e-06, "loss": 0.3737, "step": 4934 }, { "epoch": 0.34987593052109184, "grad_norm": 1.1286037096424977, "learning_rate": 7.548892386938068e-06, "loss": 0.6544, "step": 4935 }, { "epoch": 0.3499468273661822, "grad_norm": 5.668956845053906, "learning_rate": 7.547904552008352e-06, "loss": 0.4287, "step": 4936 }, { "epoch": 0.3500177242112726, "grad_norm": 3.58314621058381, "learning_rate": 7.546916582725685e-06, "loss": 0.4083, "step": 4937 }, { "epoch": 0.350088621056363, "grad_norm": 5.504098021902171, "learning_rate": 7.545928479142164e-06, "loss": 0.4094, "step": 4938 }, { "epoch": 0.3501595179014534, "grad_norm": 3.501454599473204, "learning_rate": 7.5449402413098905e-06, "loss": 0.4122, "step": 4939 }, { "epoch": 0.35023041474654376, "grad_norm": 4.295236830863111, "learning_rate": 7.543951869280974e-06, "loss": 0.4194, "step": 4940 }, { "epoch": 0.3503013115916342, "grad_norm": 14.996571984541514, "learning_rate": 7.542963363107536e-06, "loss": 0.4025, "step": 4941 }, { "epoch": 0.35037220843672456, "grad_norm": 3.7016649532379864, "learning_rate": 7.541974722841697e-06, "loss": 0.4549, "step": 4942 }, { "epoch": 0.35044310528181494, "grad_norm": 2.8033467987811522, "learning_rate": 7.540985948535593e-06, "loss": 0.4419, "step": 4943 }, { "epoch": 0.35051400212690537, "grad_norm": 2.7478314248155526, "learning_rate": 7.539997040241357e-06, "loss": 0.4438, "step": 4944 }, { "epoch": 0.35058489897199574, "grad_norm": 2.5054199870901113, "learning_rate": 7.539007998011141e-06, "loss": 0.4024, "step": 4945 }, { "epoch": 0.3506557958170861, "grad_norm": 4.315059571650027, "learning_rate": 7.538018821897095e-06, "loss": 0.4405, "step": 4946 }, { "epoch": 0.35072669266217654, "grad_norm": 2.7409238606085315, "learning_rate": 7.537029511951378e-06, "loss": 0.4623, "step": 4947 }, { "epoch": 0.3507975895072669, "grad_norm": 5.028397233229989, "learning_rate": 7.536040068226159e-06, "loss": 0.4498, "step": 4948 }, { "epoch": 0.35086848635235734, "grad_norm": 6.568296758494112, "learning_rate": 7.535050490773612e-06, "loss": 0.4539, "step": 4949 }, { "epoch": 0.3509393831974477, "grad_norm": 2.1441660356028094, "learning_rate": 7.534060779645917e-06, "loss": 0.4218, "step": 4950 }, { "epoch": 0.3510102800425381, "grad_norm": 3.4117204165793966, "learning_rate": 7.5330709348952644e-06, "loss": 0.4031, "step": 4951 }, { "epoch": 0.3510811768876285, "grad_norm": 3.2933870834739594, "learning_rate": 7.532080956573846e-06, "loss": 0.469, "step": 4952 }, { "epoch": 0.3511520737327189, "grad_norm": 2.1649772183409346, "learning_rate": 7.5310908447338685e-06, "loss": 0.4554, "step": 4953 }, { "epoch": 0.35122297057780927, "grad_norm": 4.425893118534474, "learning_rate": 7.530100599427537e-06, "loss": 0.4244, "step": 4954 }, { "epoch": 0.3512938674228997, "grad_norm": 3.282754399220069, "learning_rate": 7.52911022070707e-06, "loss": 0.4538, "step": 4955 }, { "epoch": 0.35136476426799007, "grad_norm": 3.581112667811766, "learning_rate": 7.528119708624693e-06, "loss": 0.402, "step": 4956 }, { "epoch": 0.35143566111308044, "grad_norm": 5.3928904646006615, "learning_rate": 7.5271290632326295e-06, "loss": 0.3891, "step": 4957 }, { "epoch": 0.3515065579581709, "grad_norm": 4.57005976621545, "learning_rate": 7.526138284583126e-06, "loss": 0.4396, "step": 4958 }, { "epoch": 0.35157745480326125, "grad_norm": 5.83196164719632, "learning_rate": 7.525147372728421e-06, "loss": 0.3811, "step": 4959 }, { "epoch": 0.3516483516483517, "grad_norm": 3.2548074557371733, "learning_rate": 7.524156327720766e-06, "loss": 0.4155, "step": 4960 }, { "epoch": 0.35171924849344205, "grad_norm": 3.670559774411452, "learning_rate": 7.523165149612423e-06, "loss": 0.3718, "step": 4961 }, { "epoch": 0.3517901453385324, "grad_norm": 4.507712408179845, "learning_rate": 7.522173838455655e-06, "loss": 0.4044, "step": 4962 }, { "epoch": 0.35186104218362285, "grad_norm": 6.495657882239023, "learning_rate": 7.521182394302738e-06, "loss": 0.4554, "step": 4963 }, { "epoch": 0.3519319390287132, "grad_norm": 4.991862550309622, "learning_rate": 7.520190817205946e-06, "loss": 0.4695, "step": 4964 }, { "epoch": 0.3520028358738036, "grad_norm": 5.629942698839832, "learning_rate": 7.51919910721757e-06, "loss": 0.4006, "step": 4965 }, { "epoch": 0.35207373271889403, "grad_norm": 5.565484850769735, "learning_rate": 7.5182072643899015e-06, "loss": 0.3703, "step": 4966 }, { "epoch": 0.3521446295639844, "grad_norm": 10.288343674814596, "learning_rate": 7.517215288775243e-06, "loss": 0.4982, "step": 4967 }, { "epoch": 0.3522155264090748, "grad_norm": 8.930293756860966, "learning_rate": 7.5162231804259e-06, "loss": 0.3954, "step": 4968 }, { "epoch": 0.3522864232541652, "grad_norm": 2.193799392239354, "learning_rate": 7.515230939394188e-06, "loss": 0.4574, "step": 4969 }, { "epoch": 0.3523573200992556, "grad_norm": 10.301793243891677, "learning_rate": 7.514238565732429e-06, "loss": 0.4807, "step": 4970 }, { "epoch": 0.35242821694434595, "grad_norm": 6.896620728119907, "learning_rate": 7.5132460594929535e-06, "loss": 0.4244, "step": 4971 }, { "epoch": 0.3524991137894364, "grad_norm": 3.5097809358478407, "learning_rate": 7.512253420728093e-06, "loss": 0.3948, "step": 4972 }, { "epoch": 0.35257001063452675, "grad_norm": 4.256307440798068, "learning_rate": 7.511260649490191e-06, "loss": 0.4368, "step": 4973 }, { "epoch": 0.3526409074796172, "grad_norm": 10.934671614710394, "learning_rate": 7.5102677458316e-06, "loss": 0.4091, "step": 4974 }, { "epoch": 0.35271180432470756, "grad_norm": 2.291943001565668, "learning_rate": 7.509274709804674e-06, "loss": 0.3934, "step": 4975 }, { "epoch": 0.35278270116979793, "grad_norm": 3.231503938882334, "learning_rate": 7.508281541461777e-06, "loss": 0.4854, "step": 4976 }, { "epoch": 0.35285359801488836, "grad_norm": 6.398121907919886, "learning_rate": 7.507288240855282e-06, "loss": 0.4255, "step": 4977 }, { "epoch": 0.35292449485997873, "grad_norm": 5.842995144190368, "learning_rate": 7.506294808037561e-06, "loss": 0.4984, "step": 4978 }, { "epoch": 0.3529953917050691, "grad_norm": 1.1630944482391097, "learning_rate": 7.505301243061005e-06, "loss": 0.6962, "step": 4979 }, { "epoch": 0.35306628855015953, "grad_norm": 4.88277535594464, "learning_rate": 7.504307545978e-06, "loss": 0.375, "step": 4980 }, { "epoch": 0.3531371853952499, "grad_norm": 3.885792258542172, "learning_rate": 7.503313716840948e-06, "loss": 0.4371, "step": 4981 }, { "epoch": 0.3532080822403403, "grad_norm": 5.4651867493999955, "learning_rate": 7.502319755702252e-06, "loss": 0.3986, "step": 4982 }, { "epoch": 0.3532789790854307, "grad_norm": 2.7175276626158165, "learning_rate": 7.501325662614326e-06, "loss": 0.3998, "step": 4983 }, { "epoch": 0.3533498759305211, "grad_norm": 4.710759694865944, "learning_rate": 7.500331437629589e-06, "loss": 0.4589, "step": 4984 }, { "epoch": 0.35342077277561146, "grad_norm": 3.3008954879758043, "learning_rate": 7.499337080800466e-06, "loss": 0.3909, "step": 4985 }, { "epoch": 0.3534916696207019, "grad_norm": 3.268681783617637, "learning_rate": 7.498342592179391e-06, "loss": 0.428, "step": 4986 }, { "epoch": 0.35356256646579226, "grad_norm": 2.706749700899116, "learning_rate": 7.4973479718188046e-06, "loss": 0.3783, "step": 4987 }, { "epoch": 0.3536334633108827, "grad_norm": 4.71336108742166, "learning_rate": 7.496353219771154e-06, "loss": 0.4189, "step": 4988 }, { "epoch": 0.35370436015597306, "grad_norm": 2.1573854751197232, "learning_rate": 7.495358336088892e-06, "loss": 0.4033, "step": 4989 }, { "epoch": 0.35377525700106344, "grad_norm": 3.7469994117256773, "learning_rate": 7.49436332082448e-06, "loss": 0.4122, "step": 4990 }, { "epoch": 0.35384615384615387, "grad_norm": 6.4910439592319085, "learning_rate": 7.4933681740303865e-06, "loss": 0.4303, "step": 4991 }, { "epoch": 0.35391705069124424, "grad_norm": 5.372590494104972, "learning_rate": 7.492372895759087e-06, "loss": 0.4192, "step": 4992 }, { "epoch": 0.3539879475363346, "grad_norm": 5.223008134791671, "learning_rate": 7.4913774860630614e-06, "loss": 0.4364, "step": 4993 }, { "epoch": 0.35405884438142504, "grad_norm": 2.323733236038454, "learning_rate": 7.4903819449948004e-06, "loss": 0.4332, "step": 4994 }, { "epoch": 0.3541297412265154, "grad_norm": 5.4561805784612725, "learning_rate": 7.489386272606797e-06, "loss": 0.4096, "step": 4995 }, { "epoch": 0.3542006380716058, "grad_norm": 1.9554479995382954, "learning_rate": 7.488390468951557e-06, "loss": 0.4398, "step": 4996 }, { "epoch": 0.3542715349166962, "grad_norm": 2.9312710095716654, "learning_rate": 7.487394534081587e-06, "loss": 0.3987, "step": 4997 }, { "epoch": 0.3543424317617866, "grad_norm": 2.8717200452949068, "learning_rate": 7.486398468049405e-06, "loss": 0.4621, "step": 4998 }, { "epoch": 0.354413328606877, "grad_norm": 3.104247227097183, "learning_rate": 7.485402270907534e-06, "loss": 0.3989, "step": 4999 }, { "epoch": 0.3544842254519674, "grad_norm": 3.996431193905585, "learning_rate": 7.4844059427085046e-06, "loss": 0.4085, "step": 5000 }, { "epoch": 0.35455512229705777, "grad_norm": 3.1691128308450573, "learning_rate": 7.483409483504853e-06, "loss": 0.3619, "step": 5001 }, { "epoch": 0.3546260191421482, "grad_norm": 57.19466791897932, "learning_rate": 7.482412893349125e-06, "loss": 0.4405, "step": 5002 }, { "epoch": 0.35469691598723857, "grad_norm": 5.718877513018744, "learning_rate": 7.48141617229387e-06, "loss": 0.4579, "step": 5003 }, { "epoch": 0.35476781283232894, "grad_norm": 3.928129120946671, "learning_rate": 7.480419320391646e-06, "loss": 0.4111, "step": 5004 }, { "epoch": 0.3548387096774194, "grad_norm": 2.3602162310416235, "learning_rate": 7.4794223376950184e-06, "loss": 0.4377, "step": 5005 }, { "epoch": 0.35490960652250975, "grad_norm": 6.1247117626981575, "learning_rate": 7.478425224256558e-06, "loss": 0.3787, "step": 5006 }, { "epoch": 0.3549805033676001, "grad_norm": 4.520274884804672, "learning_rate": 7.477427980128843e-06, "loss": 0.4122, "step": 5007 }, { "epoch": 0.35505140021269055, "grad_norm": 6.5704336926097024, "learning_rate": 7.4764306053644606e-06, "loss": 0.44, "step": 5008 }, { "epoch": 0.3551222970577809, "grad_norm": 4.80838760093929, "learning_rate": 7.475433100016003e-06, "loss": 0.4678, "step": 5009 }, { "epoch": 0.3551931939028713, "grad_norm": 4.383699899985414, "learning_rate": 7.474435464136067e-06, "loss": 0.4301, "step": 5010 }, { "epoch": 0.3552640907479617, "grad_norm": 3.1375742918983383, "learning_rate": 7.473437697777261e-06, "loss": 0.4057, "step": 5011 }, { "epoch": 0.3553349875930521, "grad_norm": 3.279835330780214, "learning_rate": 7.472439800992198e-06, "loss": 0.4524, "step": 5012 }, { "epoch": 0.3554058844381425, "grad_norm": 4.298794414008096, "learning_rate": 7.4714417738334965e-06, "loss": 0.3941, "step": 5013 }, { "epoch": 0.3554767812832329, "grad_norm": 2.633678788974699, "learning_rate": 7.470443616353784e-06, "loss": 0.4389, "step": 5014 }, { "epoch": 0.3555476781283233, "grad_norm": 1.0731729614656509, "learning_rate": 7.469445328605693e-06, "loss": 0.7002, "step": 5015 }, { "epoch": 0.3556185749734137, "grad_norm": 5.135783140877987, "learning_rate": 7.468446910641867e-06, "loss": 0.4243, "step": 5016 }, { "epoch": 0.3556894718185041, "grad_norm": 6.824649721808097, "learning_rate": 7.467448362514951e-06, "loss": 0.483, "step": 5017 }, { "epoch": 0.35576036866359445, "grad_norm": 2.644072505250653, "learning_rate": 7.466449684277598e-06, "loss": 0.4465, "step": 5018 }, { "epoch": 0.3558312655086849, "grad_norm": 2.7608140977585047, "learning_rate": 7.4654508759824715e-06, "loss": 0.4867, "step": 5019 }, { "epoch": 0.35590216235377525, "grad_norm": 2.9303616602580482, "learning_rate": 7.4644519376822375e-06, "loss": 0.3905, "step": 5020 }, { "epoch": 0.3559730591988656, "grad_norm": 2.0729102372842867, "learning_rate": 7.4634528694295726e-06, "loss": 0.3754, "step": 5021 }, { "epoch": 0.35604395604395606, "grad_norm": 2.8564504005963007, "learning_rate": 7.462453671277159e-06, "loss": 0.4191, "step": 5022 }, { "epoch": 0.35611485288904643, "grad_norm": 2.4862242918839725, "learning_rate": 7.461454343277681e-06, "loss": 0.3937, "step": 5023 }, { "epoch": 0.35618574973413686, "grad_norm": 6.538958194350777, "learning_rate": 7.46045488548384e-06, "loss": 0.4296, "step": 5024 }, { "epoch": 0.35625664657922723, "grad_norm": 5.340080194096767, "learning_rate": 7.459455297948332e-06, "loss": 0.4291, "step": 5025 }, { "epoch": 0.3563275434243176, "grad_norm": 3.4489898612270653, "learning_rate": 7.45845558072387e-06, "loss": 0.4718, "step": 5026 }, { "epoch": 0.35639844026940803, "grad_norm": 3.6343865802201876, "learning_rate": 7.457455733863169e-06, "loss": 0.4018, "step": 5027 }, { "epoch": 0.3564693371144984, "grad_norm": 3.928980194460475, "learning_rate": 7.456455757418951e-06, "loss": 0.4299, "step": 5028 }, { "epoch": 0.3565402339595888, "grad_norm": 3.143238445917009, "learning_rate": 7.455455651443947e-06, "loss": 0.4122, "step": 5029 }, { "epoch": 0.3566111308046792, "grad_norm": 3.6249463744643102, "learning_rate": 7.454455415990892e-06, "loss": 0.4696, "step": 5030 }, { "epoch": 0.3566820276497696, "grad_norm": 3.900025351398966, "learning_rate": 7.45345505111253e-06, "loss": 0.4318, "step": 5031 }, { "epoch": 0.35675292449485996, "grad_norm": 4.773407689375681, "learning_rate": 7.452454556861608e-06, "loss": 0.4835, "step": 5032 }, { "epoch": 0.3568238213399504, "grad_norm": 4.1491442161314325, "learning_rate": 7.4514539332908874e-06, "loss": 0.4481, "step": 5033 }, { "epoch": 0.35689471818504076, "grad_norm": 2.5928180075068097, "learning_rate": 7.45045318045313e-06, "loss": 0.4481, "step": 5034 }, { "epoch": 0.35696561503013113, "grad_norm": 2.642746304379687, "learning_rate": 7.449452298401106e-06, "loss": 0.4532, "step": 5035 }, { "epoch": 0.35703651187522156, "grad_norm": 5.049050388284994, "learning_rate": 7.448451287187591e-06, "loss": 0.3994, "step": 5036 }, { "epoch": 0.35710740872031194, "grad_norm": 3.0208452591211215, "learning_rate": 7.4474501468653725e-06, "loss": 0.4005, "step": 5037 }, { "epoch": 0.35717830556540237, "grad_norm": 3.9939558788673963, "learning_rate": 7.446448877487239e-06, "loss": 0.4375, "step": 5038 }, { "epoch": 0.35724920241049274, "grad_norm": 4.606974634372394, "learning_rate": 7.445447479105989e-06, "loss": 0.4664, "step": 5039 }, { "epoch": 0.3573200992555831, "grad_norm": 2.961245885361878, "learning_rate": 7.4444459517744276e-06, "loss": 0.3763, "step": 5040 }, { "epoch": 0.35739099610067354, "grad_norm": 3.0425017867333044, "learning_rate": 7.4434442955453656e-06, "loss": 0.414, "step": 5041 }, { "epoch": 0.3574618929457639, "grad_norm": 3.036366657746825, "learning_rate": 7.442442510471619e-06, "loss": 0.4396, "step": 5042 }, { "epoch": 0.3575327897908543, "grad_norm": 4.730797164191311, "learning_rate": 7.4414405966060156e-06, "loss": 0.438, "step": 5043 }, { "epoch": 0.3576036866359447, "grad_norm": 4.0201257327931055, "learning_rate": 7.440438554001386e-06, "loss": 0.4292, "step": 5044 }, { "epoch": 0.3576745834810351, "grad_norm": 4.021284379710365, "learning_rate": 7.4394363827105695e-06, "loss": 0.4246, "step": 5045 }, { "epoch": 0.35774548032612546, "grad_norm": 5.846084450604305, "learning_rate": 7.4384340827864095e-06, "loss": 0.3799, "step": 5046 }, { "epoch": 0.3578163771712159, "grad_norm": 8.418168817207102, "learning_rate": 7.4374316542817596e-06, "loss": 0.4402, "step": 5047 }, { "epoch": 0.35788727401630627, "grad_norm": 3.564043494561858, "learning_rate": 7.436429097249479e-06, "loss": 0.4627, "step": 5048 }, { "epoch": 0.35795817086139664, "grad_norm": 4.479245496687386, "learning_rate": 7.435426411742431e-06, "loss": 0.4053, "step": 5049 }, { "epoch": 0.35802906770648707, "grad_norm": 2.844839240765583, "learning_rate": 7.434423597813491e-06, "loss": 0.4847, "step": 5050 }, { "epoch": 0.35809996455157744, "grad_norm": 3.4715981456295566, "learning_rate": 7.4334206555155355e-06, "loss": 0.4502, "step": 5051 }, { "epoch": 0.3581708613966679, "grad_norm": 1.8374324881402708, "learning_rate": 7.432417584901451e-06, "loss": 0.4719, "step": 5052 }, { "epoch": 0.35824175824175825, "grad_norm": 4.248121587824838, "learning_rate": 7.431414386024132e-06, "loss": 0.4097, "step": 5053 }, { "epoch": 0.3583126550868486, "grad_norm": 4.270787290261621, "learning_rate": 7.430411058936476e-06, "loss": 0.4683, "step": 5054 }, { "epoch": 0.35838355193193905, "grad_norm": 2.1535268369369858, "learning_rate": 7.429407603691389e-06, "loss": 0.4064, "step": 5055 }, { "epoch": 0.3584544487770294, "grad_norm": 3.568151463634671, "learning_rate": 7.428404020341786e-06, "loss": 0.4268, "step": 5056 }, { "epoch": 0.3585253456221198, "grad_norm": 3.1430657190378737, "learning_rate": 7.4274003089405844e-06, "loss": 0.4142, "step": 5057 }, { "epoch": 0.3585962424672102, "grad_norm": 1.7459593162630653, "learning_rate": 7.426396469540713e-06, "loss": 0.3989, "step": 5058 }, { "epoch": 0.3586671393123006, "grad_norm": 2.904178759209505, "learning_rate": 7.425392502195103e-06, "loss": 0.425, "step": 5059 }, { "epoch": 0.35873803615739097, "grad_norm": 5.502469090141166, "learning_rate": 7.424388406956695e-06, "loss": 0.4077, "step": 5060 }, { "epoch": 0.3588089330024814, "grad_norm": 1.776319610725979, "learning_rate": 7.423384183878436e-06, "loss": 0.422, "step": 5061 }, { "epoch": 0.3588798298475718, "grad_norm": 3.5100163920953036, "learning_rate": 7.42237983301328e-06, "loss": 0.4228, "step": 5062 }, { "epoch": 0.3589507266926622, "grad_norm": 1.465608622180679, "learning_rate": 7.421375354414186e-06, "loss": 0.4491, "step": 5063 }, { "epoch": 0.3590216235377526, "grad_norm": 2.4199287695268294, "learning_rate": 7.420370748134121e-06, "loss": 0.4122, "step": 5064 }, { "epoch": 0.35909252038284295, "grad_norm": 0.9933891423333513, "learning_rate": 7.419366014226059e-06, "loss": 0.642, "step": 5065 }, { "epoch": 0.3591634172279334, "grad_norm": 3.3019424163722735, "learning_rate": 7.418361152742981e-06, "loss": 0.414, "step": 5066 }, { "epoch": 0.35923431407302375, "grad_norm": 3.4919402344178785, "learning_rate": 7.417356163737875e-06, "loss": 0.4131, "step": 5067 }, { "epoch": 0.3593052109181141, "grad_norm": 4.589061981816697, "learning_rate": 7.41635104726373e-06, "loss": 0.4275, "step": 5068 }, { "epoch": 0.35937610776320456, "grad_norm": 23.090408284600347, "learning_rate": 7.415345803373554e-06, "loss": 0.4568, "step": 5069 }, { "epoch": 0.35944700460829493, "grad_norm": 3.154858556139782, "learning_rate": 7.414340432120347e-06, "loss": 0.45, "step": 5070 }, { "epoch": 0.3595179014533853, "grad_norm": 34.11785687433073, "learning_rate": 7.413334933557128e-06, "loss": 0.3666, "step": 5071 }, { "epoch": 0.35958879829847573, "grad_norm": 0.9511878087312318, "learning_rate": 7.412329307736914e-06, "loss": 0.6442, "step": 5072 }, { "epoch": 0.3596596951435661, "grad_norm": 5.078230857025755, "learning_rate": 7.411323554712736e-06, "loss": 0.4032, "step": 5073 }, { "epoch": 0.3597305919886565, "grad_norm": 2.7700818463736545, "learning_rate": 7.410317674537626e-06, "loss": 0.4608, "step": 5074 }, { "epoch": 0.3598014888337469, "grad_norm": 5.450863724760973, "learning_rate": 7.409311667264626e-06, "loss": 0.4665, "step": 5075 }, { "epoch": 0.3598723856788373, "grad_norm": 3.5353923927272928, "learning_rate": 7.408305532946782e-06, "loss": 0.4116, "step": 5076 }, { "epoch": 0.3599432825239277, "grad_norm": 6.041032311367413, "learning_rate": 7.40729927163715e-06, "loss": 0.4271, "step": 5077 }, { "epoch": 0.3600141793690181, "grad_norm": 2.4540638465186015, "learning_rate": 7.406292883388788e-06, "loss": 0.3546, "step": 5078 }, { "epoch": 0.36008507621410846, "grad_norm": 7.918316835281466, "learning_rate": 7.405286368254767e-06, "loss": 0.4431, "step": 5079 }, { "epoch": 0.3601559730591989, "grad_norm": 10.234510936113935, "learning_rate": 7.4042797262881605e-06, "loss": 0.4335, "step": 5080 }, { "epoch": 0.36022686990428926, "grad_norm": 2.7740819539274204, "learning_rate": 7.403272957542047e-06, "loss": 0.463, "step": 5081 }, { "epoch": 0.36029776674937963, "grad_norm": 2.44078936859303, "learning_rate": 7.402266062069519e-06, "loss": 0.4043, "step": 5082 }, { "epoch": 0.36036866359447006, "grad_norm": 4.957431085652598, "learning_rate": 7.4012590399236665e-06, "loss": 0.4519, "step": 5083 }, { "epoch": 0.36043956043956044, "grad_norm": 2.479151624772821, "learning_rate": 7.400251891157591e-06, "loss": 0.4035, "step": 5084 }, { "epoch": 0.3605104572846508, "grad_norm": 2.5249847348999417, "learning_rate": 7.399244615824403e-06, "loss": 0.3865, "step": 5085 }, { "epoch": 0.36058135412974124, "grad_norm": 3.7957115426749786, "learning_rate": 7.3982372139772125e-06, "loss": 0.408, "step": 5086 }, { "epoch": 0.3606522509748316, "grad_norm": 3.3301659140095436, "learning_rate": 7.397229685669145e-06, "loss": 0.4346, "step": 5087 }, { "epoch": 0.36072314781992204, "grad_norm": 6.171854956013064, "learning_rate": 7.396222030953328e-06, "loss": 0.4463, "step": 5088 }, { "epoch": 0.3607940446650124, "grad_norm": 1.1199526657858139, "learning_rate": 7.3952142498828915e-06, "loss": 0.6385, "step": 5089 }, { "epoch": 0.3608649415101028, "grad_norm": 3.881642092476874, "learning_rate": 7.394206342510981e-06, "loss": 0.4098, "step": 5090 }, { "epoch": 0.3609358383551932, "grad_norm": 11.135902405182417, "learning_rate": 7.393198308890743e-06, "loss": 0.4583, "step": 5091 }, { "epoch": 0.3610067352002836, "grad_norm": 3.075718211444098, "learning_rate": 7.392190149075331e-06, "loss": 0.4642, "step": 5092 }, { "epoch": 0.36107763204537396, "grad_norm": 3.3402956628827307, "learning_rate": 7.391181863117907e-06, "loss": 0.3884, "step": 5093 }, { "epoch": 0.3611485288904644, "grad_norm": 3.273744266628999, "learning_rate": 7.3901734510716365e-06, "loss": 0.3972, "step": 5094 }, { "epoch": 0.36121942573555477, "grad_norm": 2.7706075345307104, "learning_rate": 7.389164912989697e-06, "loss": 0.4733, "step": 5095 }, { "epoch": 0.36129032258064514, "grad_norm": 7.194419634223055, "learning_rate": 7.388156248925268e-06, "loss": 0.4513, "step": 5096 }, { "epoch": 0.36136121942573557, "grad_norm": 2.200527393966701, "learning_rate": 7.387147458931537e-06, "loss": 0.4183, "step": 5097 }, { "epoch": 0.36143211627082594, "grad_norm": 3.1327030261573228, "learning_rate": 7.386138543061698e-06, "loss": 0.414, "step": 5098 }, { "epoch": 0.3615030131159163, "grad_norm": 2.5041949055649004, "learning_rate": 7.385129501368954e-06, "loss": 0.4531, "step": 5099 }, { "epoch": 0.36157390996100675, "grad_norm": 2.676315446322033, "learning_rate": 7.384120333906509e-06, "loss": 0.4172, "step": 5100 }, { "epoch": 0.3616448068060971, "grad_norm": 2.3811644513178196, "learning_rate": 7.383111040727581e-06, "loss": 0.4495, "step": 5101 }, { "epoch": 0.36171570365118755, "grad_norm": 5.608558087839665, "learning_rate": 7.382101621885387e-06, "loss": 0.4618, "step": 5102 }, { "epoch": 0.3617866004962779, "grad_norm": 3.90394383021018, "learning_rate": 7.381092077433158e-06, "loss": 0.436, "step": 5103 }, { "epoch": 0.3618574973413683, "grad_norm": 2.4774770429046344, "learning_rate": 7.380082407424125e-06, "loss": 0.4516, "step": 5104 }, { "epoch": 0.3619283941864587, "grad_norm": 4.104352564174088, "learning_rate": 7.3790726119115295e-06, "loss": 0.4728, "step": 5105 }, { "epoch": 0.3619992910315491, "grad_norm": 3.5798156153182146, "learning_rate": 7.37806269094862e-06, "loss": 0.3925, "step": 5106 }, { "epoch": 0.36207018787663947, "grad_norm": 2.498847620863414, "learning_rate": 7.377052644588651e-06, "loss": 0.4456, "step": 5107 }, { "epoch": 0.3621410847217299, "grad_norm": 3.781953230412729, "learning_rate": 7.376042472884881e-06, "loss": 0.4494, "step": 5108 }, { "epoch": 0.3622119815668203, "grad_norm": 2.345585079286526, "learning_rate": 7.375032175890577e-06, "loss": 0.4243, "step": 5109 }, { "epoch": 0.36228287841191065, "grad_norm": 4.1308127097703355, "learning_rate": 7.374021753659014e-06, "loss": 0.4777, "step": 5110 }, { "epoch": 0.3623537752570011, "grad_norm": 1.1063155615877895, "learning_rate": 7.373011206243473e-06, "loss": 0.697, "step": 5111 }, { "epoch": 0.36242467210209145, "grad_norm": 3.1267695993838336, "learning_rate": 7.372000533697239e-06, "loss": 0.4231, "step": 5112 }, { "epoch": 0.3624955689471818, "grad_norm": 3.562390716194599, "learning_rate": 7.370989736073605e-06, "loss": 0.4323, "step": 5113 }, { "epoch": 0.36256646579227225, "grad_norm": 2.239817814691622, "learning_rate": 7.3699788134258745e-06, "loss": 0.4213, "step": 5114 }, { "epoch": 0.3626373626373626, "grad_norm": 5.262494745251277, "learning_rate": 7.368967765807351e-06, "loss": 0.423, "step": 5115 }, { "epoch": 0.36270825948245305, "grad_norm": 29.996887172134493, "learning_rate": 7.36795659327135e-06, "loss": 0.4569, "step": 5116 }, { "epoch": 0.36277915632754343, "grad_norm": 3.3571175212288957, "learning_rate": 7.36694529587119e-06, "loss": 0.4363, "step": 5117 }, { "epoch": 0.3628500531726338, "grad_norm": 2.8693992956772814, "learning_rate": 7.365933873660198e-06, "loss": 0.4024, "step": 5118 }, { "epoch": 0.36292095001772423, "grad_norm": 6.121761601817363, "learning_rate": 7.364922326691706e-06, "loss": 0.4453, "step": 5119 }, { "epoch": 0.3629918468628146, "grad_norm": 3.209720476555913, "learning_rate": 7.3639106550190575e-06, "loss": 0.448, "step": 5120 }, { "epoch": 0.363062743707905, "grad_norm": 3.541705986547793, "learning_rate": 7.362898858695593e-06, "loss": 0.439, "step": 5121 }, { "epoch": 0.3631336405529954, "grad_norm": 6.94785274053988, "learning_rate": 7.361886937774669e-06, "loss": 0.456, "step": 5122 }, { "epoch": 0.3632045373980858, "grad_norm": 2.6120712896249816, "learning_rate": 7.360874892309644e-06, "loss": 0.4361, "step": 5123 }, { "epoch": 0.36327543424317615, "grad_norm": 3.423465889412323, "learning_rate": 7.359862722353884e-06, "loss": 0.3907, "step": 5124 }, { "epoch": 0.3633463310882666, "grad_norm": 1.9617382960204754, "learning_rate": 7.358850427960763e-06, "loss": 0.4444, "step": 5125 }, { "epoch": 0.36341722793335696, "grad_norm": 4.869982091588737, "learning_rate": 7.357838009183656e-06, "loss": 0.4786, "step": 5126 }, { "epoch": 0.3634881247784474, "grad_norm": 2.3961355675042593, "learning_rate": 7.356825466075953e-06, "loss": 0.4185, "step": 5127 }, { "epoch": 0.36355902162353776, "grad_norm": 4.628911135755714, "learning_rate": 7.355812798691043e-06, "loss": 0.4478, "step": 5128 }, { "epoch": 0.36362991846862813, "grad_norm": 2.64772139577575, "learning_rate": 7.3548000070823265e-06, "loss": 0.3444, "step": 5129 }, { "epoch": 0.36370081531371856, "grad_norm": 3.351612895790579, "learning_rate": 7.353787091303209e-06, "loss": 0.3827, "step": 5130 }, { "epoch": 0.36377171215880894, "grad_norm": 3.317701741638246, "learning_rate": 7.352774051407101e-06, "loss": 0.4685, "step": 5131 }, { "epoch": 0.3638426090038993, "grad_norm": 2.6200563098658383, "learning_rate": 7.351760887447421e-06, "loss": 0.3518, "step": 5132 }, { "epoch": 0.36391350584898974, "grad_norm": 4.161909576000416, "learning_rate": 7.3507475994775955e-06, "loss": 0.4367, "step": 5133 }, { "epoch": 0.3639844026940801, "grad_norm": 5.693963056190147, "learning_rate": 7.349734187551053e-06, "loss": 0.4454, "step": 5134 }, { "epoch": 0.3640552995391705, "grad_norm": 3.060024888518509, "learning_rate": 7.348720651721235e-06, "loss": 0.4284, "step": 5135 }, { "epoch": 0.3641261963842609, "grad_norm": 2.7999281419221766, "learning_rate": 7.347706992041584e-06, "loss": 0.3996, "step": 5136 }, { "epoch": 0.3641970932293513, "grad_norm": 4.793823720802062, "learning_rate": 7.3466932085655514e-06, "loss": 0.4494, "step": 5137 }, { "epoch": 0.36426799007444166, "grad_norm": 2.732728677647168, "learning_rate": 7.345679301346594e-06, "loss": 0.4732, "step": 5138 }, { "epoch": 0.3643388869195321, "grad_norm": 6.334031340813675, "learning_rate": 7.3446652704381775e-06, "loss": 0.4054, "step": 5139 }, { "epoch": 0.36440978376462246, "grad_norm": 7.009084750261492, "learning_rate": 7.343651115893772e-06, "loss": 0.3896, "step": 5140 }, { "epoch": 0.3644806806097129, "grad_norm": 4.843426570601106, "learning_rate": 7.342636837766853e-06, "loss": 0.4631, "step": 5141 }, { "epoch": 0.36455157745480327, "grad_norm": 1.1838538136459373, "learning_rate": 7.341622436110907e-06, "loss": 0.6647, "step": 5142 }, { "epoch": 0.36462247429989364, "grad_norm": 6.684986483110961, "learning_rate": 7.340607910979421e-06, "loss": 0.425, "step": 5143 }, { "epoch": 0.36469337114498407, "grad_norm": 2.255259146158266, "learning_rate": 7.339593262425896e-06, "loss": 0.4011, "step": 5144 }, { "epoch": 0.36476426799007444, "grad_norm": 4.450501111873637, "learning_rate": 7.33857849050383e-06, "loss": 0.4041, "step": 5145 }, { "epoch": 0.3648351648351648, "grad_norm": 3.2770992239236536, "learning_rate": 7.337563595266737e-06, "loss": 0.4415, "step": 5146 }, { "epoch": 0.36490606168025524, "grad_norm": 3.19508938338094, "learning_rate": 7.3365485767681284e-06, "loss": 0.4355, "step": 5147 }, { "epoch": 0.3649769585253456, "grad_norm": 4.0452654471240495, "learning_rate": 7.3355334350615325e-06, "loss": 0.3743, "step": 5148 }, { "epoch": 0.365047855370436, "grad_norm": 3.2996747800436874, "learning_rate": 7.334518170200475e-06, "loss": 0.4679, "step": 5149 }, { "epoch": 0.3651187522155264, "grad_norm": 10.763033727669798, "learning_rate": 7.333502782238493e-06, "loss": 0.4536, "step": 5150 }, { "epoch": 0.3651896490606168, "grad_norm": 2.5770780035667618, "learning_rate": 7.332487271229128e-06, "loss": 0.4502, "step": 5151 }, { "epoch": 0.3652605459057072, "grad_norm": 2.6212496679062416, "learning_rate": 7.331471637225931e-06, "loss": 0.4028, "step": 5152 }, { "epoch": 0.3653314427507976, "grad_norm": 3.349719597925582, "learning_rate": 7.330455880282452e-06, "loss": 0.4308, "step": 5153 }, { "epoch": 0.36540233959588797, "grad_norm": 4.682392162937273, "learning_rate": 7.329440000452257e-06, "loss": 0.4855, "step": 5154 }, { "epoch": 0.3654732364409784, "grad_norm": 2.6768307693587348, "learning_rate": 7.328423997788913e-06, "loss": 0.4445, "step": 5155 }, { "epoch": 0.3655441332860688, "grad_norm": 2.5629663760767945, "learning_rate": 7.327407872345993e-06, "loss": 0.3962, "step": 5156 }, { "epoch": 0.36561503013115915, "grad_norm": 3.6812773984617264, "learning_rate": 7.3263916241770815e-06, "loss": 0.3926, "step": 5157 }, { "epoch": 0.3656859269762496, "grad_norm": 2.48199450421285, "learning_rate": 7.325375253335762e-06, "loss": 0.3873, "step": 5158 }, { "epoch": 0.36575682382133995, "grad_norm": 3.3920192409961043, "learning_rate": 7.324358759875633e-06, "loss": 0.4182, "step": 5159 }, { "epoch": 0.3658277206664303, "grad_norm": 3.00615114346473, "learning_rate": 7.32334214385029e-06, "loss": 0.4619, "step": 5160 }, { "epoch": 0.36589861751152075, "grad_norm": 3.6773945837015907, "learning_rate": 7.322325405313343e-06, "loss": 0.3865, "step": 5161 }, { "epoch": 0.3659695143566111, "grad_norm": 2.8910865844171, "learning_rate": 7.321308544318405e-06, "loss": 0.4157, "step": 5162 }, { "epoch": 0.3660404112017015, "grad_norm": 1.1678384348473483, "learning_rate": 7.320291560919096e-06, "loss": 0.7043, "step": 5163 }, { "epoch": 0.36611130804679193, "grad_norm": 2.971698896062009, "learning_rate": 7.319274455169042e-06, "loss": 0.4223, "step": 5164 }, { "epoch": 0.3661822048918823, "grad_norm": 2.241962178811814, "learning_rate": 7.318257227121877e-06, "loss": 0.4059, "step": 5165 }, { "epoch": 0.36625310173697273, "grad_norm": 3.05738367201898, "learning_rate": 7.317239876831236e-06, "loss": 0.391, "step": 5166 }, { "epoch": 0.3663239985820631, "grad_norm": 5.0644106617011255, "learning_rate": 7.31622240435077e-06, "loss": 0.3674, "step": 5167 }, { "epoch": 0.3663948954271535, "grad_norm": 8.735428611026489, "learning_rate": 7.3152048097341275e-06, "loss": 0.4405, "step": 5168 }, { "epoch": 0.3664657922722439, "grad_norm": 5.212571755753624, "learning_rate": 7.314187093034969e-06, "loss": 0.471, "step": 5169 }, { "epoch": 0.3665366891173343, "grad_norm": 6.121925599805049, "learning_rate": 7.313169254306957e-06, "loss": 0.4519, "step": 5170 }, { "epoch": 0.36660758596242465, "grad_norm": 5.89294547992929, "learning_rate": 7.312151293603768e-06, "loss": 0.4719, "step": 5171 }, { "epoch": 0.3666784828075151, "grad_norm": 2.8815734437622305, "learning_rate": 7.3111332109790745e-06, "loss": 0.4317, "step": 5172 }, { "epoch": 0.36674937965260546, "grad_norm": 3.2042407458351607, "learning_rate": 7.310115006486563e-06, "loss": 0.4636, "step": 5173 }, { "epoch": 0.36682027649769583, "grad_norm": 2.4070978384832107, "learning_rate": 7.309096680179923e-06, "loss": 0.4322, "step": 5174 }, { "epoch": 0.36689117334278626, "grad_norm": 3.8461154920845546, "learning_rate": 7.308078232112854e-06, "loss": 0.4237, "step": 5175 }, { "epoch": 0.36696207018787663, "grad_norm": 1.8682454179670696, "learning_rate": 7.3070596623390574e-06, "loss": 0.4328, "step": 5176 }, { "epoch": 0.367032967032967, "grad_norm": 3.0065537968328746, "learning_rate": 7.306040970912244e-06, "loss": 0.4325, "step": 5177 }, { "epoch": 0.36710386387805743, "grad_norm": 4.462880383467964, "learning_rate": 7.305022157886131e-06, "loss": 0.4606, "step": 5178 }, { "epoch": 0.3671747607231478, "grad_norm": 2.2608162817721706, "learning_rate": 7.304003223314438e-06, "loss": 0.4258, "step": 5179 }, { "epoch": 0.36724565756823824, "grad_norm": 2.616502238036251, "learning_rate": 7.3029841672509e-06, "loss": 0.4372, "step": 5180 }, { "epoch": 0.3673165544133286, "grad_norm": 2.1866669221993758, "learning_rate": 7.301964989749246e-06, "loss": 0.3482, "step": 5181 }, { "epoch": 0.367387451258419, "grad_norm": 5.294748162205524, "learning_rate": 7.3009456908632225e-06, "loss": 0.4516, "step": 5182 }, { "epoch": 0.3674583481035094, "grad_norm": 3.137142065834388, "learning_rate": 7.299926270646576e-06, "loss": 0.3796, "step": 5183 }, { "epoch": 0.3675292449485998, "grad_norm": 5.2249204215580125, "learning_rate": 7.298906729153061e-06, "loss": 0.454, "step": 5184 }, { "epoch": 0.36760014179369016, "grad_norm": 3.3019388595657535, "learning_rate": 7.297887066436442e-06, "loss": 0.4455, "step": 5185 }, { "epoch": 0.3676710386387806, "grad_norm": 11.203642221557642, "learning_rate": 7.296867282550482e-06, "loss": 0.5013, "step": 5186 }, { "epoch": 0.36774193548387096, "grad_norm": 1.9483188328645178, "learning_rate": 7.295847377548957e-06, "loss": 0.3831, "step": 5187 }, { "epoch": 0.36781283232896134, "grad_norm": 2.3112237961617303, "learning_rate": 7.294827351485648e-06, "loss": 0.4364, "step": 5188 }, { "epoch": 0.36788372917405177, "grad_norm": 2.344940631761939, "learning_rate": 7.293807204414341e-06, "loss": 0.4448, "step": 5189 }, { "epoch": 0.36795462601914214, "grad_norm": 3.089698300273454, "learning_rate": 7.2927869363888295e-06, "loss": 0.4463, "step": 5190 }, { "epoch": 0.36802552286423257, "grad_norm": 2.755514440009989, "learning_rate": 7.291766547462914e-06, "loss": 0.4018, "step": 5191 }, { "epoch": 0.36809641970932294, "grad_norm": 1.9477601797580175, "learning_rate": 7.2907460376903974e-06, "loss": 0.4435, "step": 5192 }, { "epoch": 0.3681673165544133, "grad_norm": 3.815537704120675, "learning_rate": 7.289725407125094e-06, "loss": 0.4218, "step": 5193 }, { "epoch": 0.36823821339950374, "grad_norm": 6.066985395799159, "learning_rate": 7.288704655820823e-06, "loss": 0.457, "step": 5194 }, { "epoch": 0.3683091102445941, "grad_norm": 2.203596576920176, "learning_rate": 7.2876837838314084e-06, "loss": 0.4392, "step": 5195 }, { "epoch": 0.3683800070896845, "grad_norm": 2.6177994299600056, "learning_rate": 7.286662791210681e-06, "loss": 0.4174, "step": 5196 }, { "epoch": 0.3684509039347749, "grad_norm": 2.432957443018611, "learning_rate": 7.285641678012481e-06, "loss": 0.4263, "step": 5197 }, { "epoch": 0.3685218007798653, "grad_norm": 7.209774025510694, "learning_rate": 7.284620444290649e-06, "loss": 0.4013, "step": 5198 }, { "epoch": 0.36859269762495567, "grad_norm": 1.9742577991228984, "learning_rate": 7.283599090099038e-06, "loss": 0.4134, "step": 5199 }, { "epoch": 0.3686635944700461, "grad_norm": 3.288290970884958, "learning_rate": 7.282577615491504e-06, "loss": 0.3648, "step": 5200 }, { "epoch": 0.36873449131513647, "grad_norm": 4.71670126737167, "learning_rate": 7.28155602052191e-06, "loss": 0.4445, "step": 5201 }, { "epoch": 0.36880538816022684, "grad_norm": 1.169650135145528, "learning_rate": 7.280534305244126e-06, "loss": 0.6212, "step": 5202 }, { "epoch": 0.3688762850053173, "grad_norm": 1.7468345020484963, "learning_rate": 7.279512469712027e-06, "loss": 0.3777, "step": 5203 }, { "epoch": 0.36894718185040765, "grad_norm": 3.5692272240056306, "learning_rate": 7.278490513979497e-06, "loss": 0.4717, "step": 5204 }, { "epoch": 0.3690180786954981, "grad_norm": 2.217354397069472, "learning_rate": 7.27746843810042e-06, "loss": 0.417, "step": 5205 }, { "epoch": 0.36908897554058845, "grad_norm": 2.2363082668511107, "learning_rate": 7.276446242128697e-06, "loss": 0.4058, "step": 5206 }, { "epoch": 0.3691598723856788, "grad_norm": 6.490711059768602, "learning_rate": 7.275423926118223e-06, "loss": 0.4148, "step": 5207 }, { "epoch": 0.36923076923076925, "grad_norm": 2.253648861475427, "learning_rate": 7.274401490122911e-06, "loss": 0.4056, "step": 5208 }, { "epoch": 0.3693016660758596, "grad_norm": 2.768839112394564, "learning_rate": 7.273378934196672e-06, "loss": 0.39, "step": 5209 }, { "epoch": 0.36937256292095, "grad_norm": 3.055014168263267, "learning_rate": 7.272356258393427e-06, "loss": 0.4444, "step": 5210 }, { "epoch": 0.3694434597660404, "grad_norm": 2.7930605726287103, "learning_rate": 7.271333462767101e-06, "loss": 0.3929, "step": 5211 }, { "epoch": 0.3695143566111308, "grad_norm": 5.906051104984663, "learning_rate": 7.27031054737163e-06, "loss": 0.4252, "step": 5212 }, { "epoch": 0.3695852534562212, "grad_norm": 2.612398898238841, "learning_rate": 7.2692875122609484e-06, "loss": 0.4171, "step": 5213 }, { "epoch": 0.3696561503013116, "grad_norm": 4.766913094016657, "learning_rate": 7.268264357489005e-06, "loss": 0.402, "step": 5214 }, { "epoch": 0.369727047146402, "grad_norm": 2.130082122824771, "learning_rate": 7.2672410831097515e-06, "loss": 0.3693, "step": 5215 }, { "epoch": 0.3697979439914924, "grad_norm": 2.285478255925834, "learning_rate": 7.266217689177144e-06, "loss": 0.4063, "step": 5216 }, { "epoch": 0.3698688408365828, "grad_norm": 5.4294945626625095, "learning_rate": 7.26519417574515e-06, "loss": 0.4347, "step": 5217 }, { "epoch": 0.36993973768167315, "grad_norm": 2.2468467887237837, "learning_rate": 7.264170542867737e-06, "loss": 0.4028, "step": 5218 }, { "epoch": 0.3700106345267636, "grad_norm": 3.289824170505583, "learning_rate": 7.2631467905988824e-06, "loss": 0.436, "step": 5219 }, { "epoch": 0.37008153137185396, "grad_norm": 1.7589598650845368, "learning_rate": 7.262122918992572e-06, "loss": 0.3854, "step": 5220 }, { "epoch": 0.37015242821694433, "grad_norm": 3.8308591245505403, "learning_rate": 7.261098928102791e-06, "loss": 0.4747, "step": 5221 }, { "epoch": 0.37022332506203476, "grad_norm": 3.0347300336611345, "learning_rate": 7.260074817983539e-06, "loss": 0.3656, "step": 5222 }, { "epoch": 0.37029422190712513, "grad_norm": 2.283036402967709, "learning_rate": 7.259050588688818e-06, "loss": 0.382, "step": 5223 }, { "epoch": 0.3703651187522155, "grad_norm": 2.613867589332025, "learning_rate": 7.258026240272632e-06, "loss": 0.3984, "step": 5224 }, { "epoch": 0.37043601559730593, "grad_norm": 10.406920578858065, "learning_rate": 7.257001772789002e-06, "loss": 0.4339, "step": 5225 }, { "epoch": 0.3705069124423963, "grad_norm": 2.3768525338585227, "learning_rate": 7.255977186291944e-06, "loss": 0.4076, "step": 5226 }, { "epoch": 0.3705778092874867, "grad_norm": 2.310897974243063, "learning_rate": 7.254952480835489e-06, "loss": 0.3801, "step": 5227 }, { "epoch": 0.3706487061325771, "grad_norm": 5.799956133158973, "learning_rate": 7.253927656473666e-06, "loss": 0.4706, "step": 5228 }, { "epoch": 0.3707196029776675, "grad_norm": 2.0272534994316724, "learning_rate": 7.252902713260519e-06, "loss": 0.3955, "step": 5229 }, { "epoch": 0.3707904998227579, "grad_norm": 4.749599502876075, "learning_rate": 7.251877651250092e-06, "loss": 0.3631, "step": 5230 }, { "epoch": 0.3708613966678483, "grad_norm": 2.4550972034706238, "learning_rate": 7.250852470496438e-06, "loss": 0.4018, "step": 5231 }, { "epoch": 0.37093229351293866, "grad_norm": 2.4451012911649137, "learning_rate": 7.249827171053615e-06, "loss": 0.444, "step": 5232 }, { "epoch": 0.3710031903580291, "grad_norm": 3.1081882015875473, "learning_rate": 7.2488017529756875e-06, "loss": 0.4179, "step": 5233 }, { "epoch": 0.37107408720311946, "grad_norm": 1.948833848937106, "learning_rate": 7.247776216316727e-06, "loss": 0.462, "step": 5234 }, { "epoch": 0.37114498404820984, "grad_norm": 5.370220722815198, "learning_rate": 7.246750561130812e-06, "loss": 0.3955, "step": 5235 }, { "epoch": 0.37121588089330027, "grad_norm": 3.0950599744688922, "learning_rate": 7.2457247874720245e-06, "loss": 0.3955, "step": 5236 }, { "epoch": 0.37128677773839064, "grad_norm": 2.529110895483362, "learning_rate": 7.244698895394455e-06, "loss": 0.4171, "step": 5237 }, { "epoch": 0.371357674583481, "grad_norm": 2.566754086320382, "learning_rate": 7.2436728849522e-06, "loss": 0.4316, "step": 5238 }, { "epoch": 0.37142857142857144, "grad_norm": 2.857605403849145, "learning_rate": 7.242646756199361e-06, "loss": 0.3999, "step": 5239 }, { "epoch": 0.3714994682736618, "grad_norm": 2.412521993076319, "learning_rate": 7.241620509190046e-06, "loss": 0.4164, "step": 5240 }, { "epoch": 0.3715703651187522, "grad_norm": 19.07995756979956, "learning_rate": 7.24059414397837e-06, "loss": 0.4201, "step": 5241 }, { "epoch": 0.3716412619638426, "grad_norm": 1.2400166510946609, "learning_rate": 7.239567660618456e-06, "loss": 0.6636, "step": 5242 }, { "epoch": 0.371712158808933, "grad_norm": 4.451334257207289, "learning_rate": 7.2385410591644305e-06, "loss": 0.3756, "step": 5243 }, { "epoch": 0.3717830556540234, "grad_norm": 2.125608852982195, "learning_rate": 7.237514339670425e-06, "loss": 0.4108, "step": 5244 }, { "epoch": 0.3718539524991138, "grad_norm": 2.702978864794342, "learning_rate": 7.236487502190581e-06, "loss": 0.42, "step": 5245 }, { "epoch": 0.37192484934420417, "grad_norm": 3.8757116913211043, "learning_rate": 7.2354605467790436e-06, "loss": 0.4109, "step": 5246 }, { "epoch": 0.3719957461892946, "grad_norm": 3.1829538439495586, "learning_rate": 7.234433473489966e-06, "loss": 0.498, "step": 5247 }, { "epoch": 0.37206664303438497, "grad_norm": 4.366904195950588, "learning_rate": 7.233406282377505e-06, "loss": 0.4375, "step": 5248 }, { "epoch": 0.37213753987947534, "grad_norm": 2.975483692224294, "learning_rate": 7.232378973495828e-06, "loss": 0.4833, "step": 5249 }, { "epoch": 0.37220843672456577, "grad_norm": 4.49896274716361, "learning_rate": 7.231351546899101e-06, "loss": 0.4039, "step": 5250 }, { "epoch": 0.37227933356965615, "grad_norm": 2.8108966264677964, "learning_rate": 7.230324002641507e-06, "loss": 0.4191, "step": 5251 }, { "epoch": 0.3723502304147465, "grad_norm": 1.1608502447059221, "learning_rate": 7.229296340777223e-06, "loss": 0.7268, "step": 5252 }, { "epoch": 0.37242112725983695, "grad_norm": 1.7413426515480914, "learning_rate": 7.228268561360443e-06, "loss": 0.4631, "step": 5253 }, { "epoch": 0.3724920241049273, "grad_norm": 4.70774122997891, "learning_rate": 7.227240664445361e-06, "loss": 0.4615, "step": 5254 }, { "epoch": 0.37256292095001775, "grad_norm": 3.007122160146786, "learning_rate": 7.2262126500861774e-06, "loss": 0.4407, "step": 5255 }, { "epoch": 0.3726338177951081, "grad_norm": 2.6177063461726897, "learning_rate": 7.225184518337103e-06, "loss": 0.491, "step": 5256 }, { "epoch": 0.3727047146401985, "grad_norm": 3.564311540597476, "learning_rate": 7.2241562692523505e-06, "loss": 0.4477, "step": 5257 }, { "epoch": 0.3727756114852889, "grad_norm": 6.203216149515918, "learning_rate": 7.223127902886139e-06, "loss": 0.3856, "step": 5258 }, { "epoch": 0.3728465083303793, "grad_norm": 3.3943421387484563, "learning_rate": 7.2220994192926986e-06, "loss": 0.447, "step": 5259 }, { "epoch": 0.3729174051754697, "grad_norm": 2.632132474769688, "learning_rate": 7.221070818526258e-06, "loss": 0.3886, "step": 5260 }, { "epoch": 0.3729883020205601, "grad_norm": 2.902130141277176, "learning_rate": 7.220042100641058e-06, "loss": 0.4638, "step": 5261 }, { "epoch": 0.3730591988656505, "grad_norm": 3.472384313916136, "learning_rate": 7.219013265691345e-06, "loss": 0.4115, "step": 5262 }, { "epoch": 0.37313009571074085, "grad_norm": 8.433639095463118, "learning_rate": 7.217984313731365e-06, "loss": 0.39, "step": 5263 }, { "epoch": 0.3732009925558313, "grad_norm": 2.816301995918583, "learning_rate": 7.216955244815385e-06, "loss": 0.4617, "step": 5264 }, { "epoch": 0.37327188940092165, "grad_norm": 6.077907564894095, "learning_rate": 7.215926058997658e-06, "loss": 0.4659, "step": 5265 }, { "epoch": 0.373342786246012, "grad_norm": 2.621616865282536, "learning_rate": 7.214896756332461e-06, "loss": 0.4214, "step": 5266 }, { "epoch": 0.37341368309110246, "grad_norm": 9.222478528635577, "learning_rate": 7.213867336874067e-06, "loss": 0.4294, "step": 5267 }, { "epoch": 0.37348457993619283, "grad_norm": 5.317035685969939, "learning_rate": 7.212837800676759e-06, "loss": 0.4284, "step": 5268 }, { "epoch": 0.37355547678128326, "grad_norm": 8.689857270558726, "learning_rate": 7.211808147794822e-06, "loss": 0.3645, "step": 5269 }, { "epoch": 0.37362637362637363, "grad_norm": 1.9739965338711576, "learning_rate": 7.210778378282557e-06, "loss": 0.4186, "step": 5270 }, { "epoch": 0.373697270471464, "grad_norm": 1.8863645422063842, "learning_rate": 7.209748492194258e-06, "loss": 0.4085, "step": 5271 }, { "epoch": 0.37376816731655443, "grad_norm": 8.923546184585874, "learning_rate": 7.208718489584236e-06, "loss": 0.4528, "step": 5272 }, { "epoch": 0.3738390641616448, "grad_norm": 4.569783173836002, "learning_rate": 7.207688370506801e-06, "loss": 0.4366, "step": 5273 }, { "epoch": 0.3739099610067352, "grad_norm": 5.364567105497519, "learning_rate": 7.2066581350162734e-06, "loss": 0.4258, "step": 5274 }, { "epoch": 0.3739808578518256, "grad_norm": 2.2462717448475638, "learning_rate": 7.205627783166979e-06, "loss": 0.372, "step": 5275 }, { "epoch": 0.374051754696916, "grad_norm": 4.035021787671319, "learning_rate": 7.2045973150132486e-06, "loss": 0.4316, "step": 5276 }, { "epoch": 0.37412265154200636, "grad_norm": 4.063027343436262, "learning_rate": 7.203566730609417e-06, "loss": 0.4134, "step": 5277 }, { "epoch": 0.3741935483870968, "grad_norm": 8.149613112966449, "learning_rate": 7.202536030009831e-06, "loss": 0.4251, "step": 5278 }, { "epoch": 0.37426444523218716, "grad_norm": 3.873781637524249, "learning_rate": 7.201505213268839e-06, "loss": 0.4159, "step": 5279 }, { "epoch": 0.3743353420772776, "grad_norm": 4.546410097743101, "learning_rate": 7.200474280440797e-06, "loss": 0.4538, "step": 5280 }, { "epoch": 0.37440623892236796, "grad_norm": 3.1797242397785253, "learning_rate": 7.199443231580067e-06, "loss": 0.4247, "step": 5281 }, { "epoch": 0.37447713576745834, "grad_norm": 2.4055185915356265, "learning_rate": 7.198412066741014e-06, "loss": 0.4336, "step": 5282 }, { "epoch": 0.37454803261254876, "grad_norm": 1.9024307275637737, "learning_rate": 7.197380785978017e-06, "loss": 0.4463, "step": 5283 }, { "epoch": 0.37461892945763914, "grad_norm": 3.1931412115195013, "learning_rate": 7.196349389345454e-06, "loss": 0.4632, "step": 5284 }, { "epoch": 0.3746898263027295, "grad_norm": 1.8125332889021335, "learning_rate": 7.1953178768977105e-06, "loss": 0.4052, "step": 5285 }, { "epoch": 0.37476072314781994, "grad_norm": 3.234301051826098, "learning_rate": 7.194286248689181e-06, "loss": 0.3911, "step": 5286 }, { "epoch": 0.3748316199929103, "grad_norm": 2.6382122030747324, "learning_rate": 7.1932545047742605e-06, "loss": 0.415, "step": 5287 }, { "epoch": 0.3749025168380007, "grad_norm": 3.8644331964992995, "learning_rate": 7.192222645207358e-06, "loss": 0.3923, "step": 5288 }, { "epoch": 0.3749734136830911, "grad_norm": 3.3817272305624355, "learning_rate": 7.191190670042882e-06, "loss": 0.4306, "step": 5289 }, { "epoch": 0.3750443105281815, "grad_norm": 3.299301340530032, "learning_rate": 7.190158579335249e-06, "loss": 0.4503, "step": 5290 }, { "epoch": 0.37511520737327186, "grad_norm": 4.033291916419501, "learning_rate": 7.189126373138883e-06, "loss": 0.4233, "step": 5291 }, { "epoch": 0.3751861042183623, "grad_norm": 2.869995673485527, "learning_rate": 7.188094051508212e-06, "loss": 0.4367, "step": 5292 }, { "epoch": 0.37525700106345267, "grad_norm": 3.10266920058993, "learning_rate": 7.187061614497671e-06, "loss": 0.3679, "step": 5293 }, { "epoch": 0.3753278979085431, "grad_norm": 2.1669910680893163, "learning_rate": 7.186029062161704e-06, "loss": 0.4608, "step": 5294 }, { "epoch": 0.37539879475363347, "grad_norm": 2.5573638129422887, "learning_rate": 7.184996394554752e-06, "loss": 0.4542, "step": 5295 }, { "epoch": 0.37546969159872384, "grad_norm": 4.034485835761873, "learning_rate": 7.183963611731276e-06, "loss": 0.474, "step": 5296 }, { "epoch": 0.37554058844381427, "grad_norm": 2.8799206847286687, "learning_rate": 7.1829307137457306e-06, "loss": 0.396, "step": 5297 }, { "epoch": 0.37561148528890465, "grad_norm": 2.7874192913811933, "learning_rate": 7.181897700652582e-06, "loss": 0.3849, "step": 5298 }, { "epoch": 0.375682382133995, "grad_norm": 2.1420910842420855, "learning_rate": 7.180864572506302e-06, "loss": 0.3982, "step": 5299 }, { "epoch": 0.37575327897908545, "grad_norm": 6.170546574186303, "learning_rate": 7.1798313293613696e-06, "loss": 0.4089, "step": 5300 }, { "epoch": 0.3758241758241758, "grad_norm": 2.137497351506363, "learning_rate": 7.178797971272267e-06, "loss": 0.4229, "step": 5301 }, { "epoch": 0.3758950726692662, "grad_norm": 2.266975558842344, "learning_rate": 7.177764498293485e-06, "loss": 0.4077, "step": 5302 }, { "epoch": 0.3759659695143566, "grad_norm": 1.879718412250209, "learning_rate": 7.176730910479517e-06, "loss": 0.43, "step": 5303 }, { "epoch": 0.376036866359447, "grad_norm": 21.55527885755962, "learning_rate": 7.1756972078848685e-06, "loss": 0.3703, "step": 5304 }, { "epoch": 0.37610776320453737, "grad_norm": 2.179521353716291, "learning_rate": 7.174663390564045e-06, "loss": 0.4317, "step": 5305 }, { "epoch": 0.3761786600496278, "grad_norm": 6.497287523246236, "learning_rate": 7.17362945857156e-06, "loss": 0.4516, "step": 5306 }, { "epoch": 0.3762495568947182, "grad_norm": 2.695869958746582, "learning_rate": 7.1725954119619365e-06, "loss": 0.4435, "step": 5307 }, { "epoch": 0.3763204537398086, "grad_norm": 2.5519918597524387, "learning_rate": 7.171561250789698e-06, "loss": 0.4025, "step": 5308 }, { "epoch": 0.376391350584899, "grad_norm": 2.2676929970062307, "learning_rate": 7.170526975109378e-06, "loss": 0.4542, "step": 5309 }, { "epoch": 0.37646224742998935, "grad_norm": 2.553668962006785, "learning_rate": 7.1694925849755125e-06, "loss": 0.4287, "step": 5310 }, { "epoch": 0.3765331442750798, "grad_norm": 2.2080424130769383, "learning_rate": 7.168458080442647e-06, "loss": 0.4515, "step": 5311 }, { "epoch": 0.37660404112017015, "grad_norm": 2.3216156698697707, "learning_rate": 7.167423461565332e-06, "loss": 0.497, "step": 5312 }, { "epoch": 0.3766749379652605, "grad_norm": 6.511656864978721, "learning_rate": 7.1663887283981244e-06, "loss": 0.4256, "step": 5313 }, { "epoch": 0.37674583481035095, "grad_norm": 2.7827491369478023, "learning_rate": 7.165353880995584e-06, "loss": 0.3816, "step": 5314 }, { "epoch": 0.37681673165544133, "grad_norm": 3.9140424765909025, "learning_rate": 7.164318919412283e-06, "loss": 0.3951, "step": 5315 }, { "epoch": 0.3768876285005317, "grad_norm": 3.035664911642187, "learning_rate": 7.163283843702792e-06, "loss": 0.4387, "step": 5316 }, { "epoch": 0.37695852534562213, "grad_norm": 3.677314894351405, "learning_rate": 7.162248653921693e-06, "loss": 0.4326, "step": 5317 }, { "epoch": 0.3770294221907125, "grad_norm": 6.175785634564935, "learning_rate": 7.1612133501235724e-06, "loss": 0.4054, "step": 5318 }, { "epoch": 0.37710031903580293, "grad_norm": 2.9115449872398056, "learning_rate": 7.160177932363022e-06, "loss": 0.3763, "step": 5319 }, { "epoch": 0.3771712158808933, "grad_norm": 3.461237106473155, "learning_rate": 7.159142400694641e-06, "loss": 0.3492, "step": 5320 }, { "epoch": 0.3772421127259837, "grad_norm": 1.8016208013334423, "learning_rate": 7.158106755173033e-06, "loss": 0.46, "step": 5321 }, { "epoch": 0.3773130095710741, "grad_norm": 4.519550571790003, "learning_rate": 7.157070995852808e-06, "loss": 0.4241, "step": 5322 }, { "epoch": 0.3773839064161645, "grad_norm": 2.4997966476925138, "learning_rate": 7.1560351227885825e-06, "loss": 0.3727, "step": 5323 }, { "epoch": 0.37745480326125486, "grad_norm": 2.0366132868533935, "learning_rate": 7.15499913603498e-06, "loss": 0.4424, "step": 5324 }, { "epoch": 0.3775257001063453, "grad_norm": 2.987902199504847, "learning_rate": 7.153963035646629e-06, "loss": 0.4456, "step": 5325 }, { "epoch": 0.37759659695143566, "grad_norm": 3.6138319856815966, "learning_rate": 7.152926821678164e-06, "loss": 0.4216, "step": 5326 }, { "epoch": 0.37766749379652603, "grad_norm": 1.8746759832871231, "learning_rate": 7.15189049418422e-06, "loss": 0.4427, "step": 5327 }, { "epoch": 0.37773839064161646, "grad_norm": 6.910690978950205, "learning_rate": 7.1508540532194515e-06, "loss": 0.4297, "step": 5328 }, { "epoch": 0.37780928748670684, "grad_norm": 4.599948909903436, "learning_rate": 7.149817498838505e-06, "loss": 0.4474, "step": 5329 }, { "epoch": 0.3778801843317972, "grad_norm": 3.724354349966599, "learning_rate": 7.148780831096044e-06, "loss": 0.4344, "step": 5330 }, { "epoch": 0.37795108117688764, "grad_norm": 3.3227542817990856, "learning_rate": 7.147744050046728e-06, "loss": 0.4, "step": 5331 }, { "epoch": 0.378021978021978, "grad_norm": 2.436224673085045, "learning_rate": 7.1467071557452276e-06, "loss": 0.4239, "step": 5332 }, { "epoch": 0.37809287486706844, "grad_norm": 3.1202505135014102, "learning_rate": 7.145670148246222e-06, "loss": 0.414, "step": 5333 }, { "epoch": 0.3781637717121588, "grad_norm": 4.9971492298430435, "learning_rate": 7.144633027604392e-06, "loss": 0.4181, "step": 5334 }, { "epoch": 0.3782346685572492, "grad_norm": 2.6564265155311526, "learning_rate": 7.1435957938744234e-06, "loss": 0.4486, "step": 5335 }, { "epoch": 0.3783055654023396, "grad_norm": 3.295158304245221, "learning_rate": 7.142558447111014e-06, "loss": 0.4135, "step": 5336 }, { "epoch": 0.37837646224743, "grad_norm": 3.1237786211148215, "learning_rate": 7.141520987368863e-06, "loss": 0.4303, "step": 5337 }, { "epoch": 0.37844735909252036, "grad_norm": 1.1790328335011258, "learning_rate": 7.140483414702675e-06, "loss": 0.6274, "step": 5338 }, { "epoch": 0.3785182559376108, "grad_norm": 2.21364440415905, "learning_rate": 7.139445729167163e-06, "loss": 0.4284, "step": 5339 }, { "epoch": 0.37858915278270117, "grad_norm": 2.991927844363288, "learning_rate": 7.138407930817044e-06, "loss": 0.4214, "step": 5340 }, { "epoch": 0.37866004962779154, "grad_norm": 2.885557139037406, "learning_rate": 7.137370019707044e-06, "loss": 0.3725, "step": 5341 }, { "epoch": 0.37873094647288197, "grad_norm": 2.4495206469979847, "learning_rate": 7.13633199589189e-06, "loss": 0.411, "step": 5342 }, { "epoch": 0.37880184331797234, "grad_norm": 4.4723414496260885, "learning_rate": 7.13529385942632e-06, "loss": 0.4469, "step": 5343 }, { "epoch": 0.37887274016306277, "grad_norm": 5.243913600665227, "learning_rate": 7.134255610365075e-06, "loss": 0.4181, "step": 5344 }, { "epoch": 0.37894363700815314, "grad_norm": 1.4999695619352866, "learning_rate": 7.133217248762903e-06, "loss": 0.3683, "step": 5345 }, { "epoch": 0.3790145338532435, "grad_norm": 2.8642788371910557, "learning_rate": 7.132178774674556e-06, "loss": 0.4286, "step": 5346 }, { "epoch": 0.37908543069833395, "grad_norm": 3.439805547587592, "learning_rate": 7.131140188154796e-06, "loss": 0.3964, "step": 5347 }, { "epoch": 0.3791563275434243, "grad_norm": 2.756198525374127, "learning_rate": 7.130101489258386e-06, "loss": 0.4453, "step": 5348 }, { "epoch": 0.3792272243885147, "grad_norm": 2.761616784330104, "learning_rate": 7.129062678040101e-06, "loss": 0.3957, "step": 5349 }, { "epoch": 0.3792981212336051, "grad_norm": 2.79185361358991, "learning_rate": 7.128023754554714e-06, "loss": 0.4173, "step": 5350 }, { "epoch": 0.3793690180786955, "grad_norm": 3.748235061993849, "learning_rate": 7.1269847188570105e-06, "loss": 0.3952, "step": 5351 }, { "epoch": 0.37943991492378587, "grad_norm": 1.8559037129578397, "learning_rate": 7.125945571001779e-06, "loss": 0.445, "step": 5352 }, { "epoch": 0.3795108117688763, "grad_norm": 2.654380187197914, "learning_rate": 7.124906311043815e-06, "loss": 0.4303, "step": 5353 }, { "epoch": 0.3795817086139667, "grad_norm": 3.3092399402007504, "learning_rate": 7.123866939037921e-06, "loss": 0.4039, "step": 5354 }, { "epoch": 0.37965260545905705, "grad_norm": 2.1144283679725944, "learning_rate": 7.122827455038901e-06, "loss": 0.3964, "step": 5355 }, { "epoch": 0.3797235023041475, "grad_norm": 3.659871057258316, "learning_rate": 7.121787859101568e-06, "loss": 0.3993, "step": 5356 }, { "epoch": 0.37979439914923785, "grad_norm": 2.247750027132024, "learning_rate": 7.120748151280743e-06, "loss": 0.3744, "step": 5357 }, { "epoch": 0.3798652959943283, "grad_norm": 2.424068619487423, "learning_rate": 7.119708331631248e-06, "loss": 0.4461, "step": 5358 }, { "epoch": 0.37993619283941865, "grad_norm": 1.931143062779508, "learning_rate": 7.118668400207917e-06, "loss": 0.3842, "step": 5359 }, { "epoch": 0.380007089684509, "grad_norm": 2.513852941769287, "learning_rate": 7.117628357065582e-06, "loss": 0.4286, "step": 5360 }, { "epoch": 0.38007798652959945, "grad_norm": 2.2036884520117934, "learning_rate": 7.116588202259088e-06, "loss": 0.3638, "step": 5361 }, { "epoch": 0.3801488833746898, "grad_norm": 2.647469296620241, "learning_rate": 7.115547935843283e-06, "loss": 0.4537, "step": 5362 }, { "epoch": 0.3802197802197802, "grad_norm": 2.9560985231592567, "learning_rate": 7.11450755787302e-06, "loss": 0.4038, "step": 5363 }, { "epoch": 0.38029067706487063, "grad_norm": 4.542980972841368, "learning_rate": 7.1134670684031616e-06, "loss": 0.425, "step": 5364 }, { "epoch": 0.380361573909961, "grad_norm": 2.8174429150757607, "learning_rate": 7.112426467488569e-06, "loss": 0.4402, "step": 5365 }, { "epoch": 0.3804324707550514, "grad_norm": 1.1559195559015008, "learning_rate": 7.111385755184117e-06, "loss": 0.7109, "step": 5366 }, { "epoch": 0.3805033676001418, "grad_norm": 1.9497114879466158, "learning_rate": 7.110344931544684e-06, "loss": 0.3929, "step": 5367 }, { "epoch": 0.3805742644452322, "grad_norm": 1.980975220674774, "learning_rate": 7.1093039966251505e-06, "loss": 0.3974, "step": 5368 }, { "epoch": 0.38064516129032255, "grad_norm": 2.431251622976346, "learning_rate": 7.108262950480408e-06, "loss": 0.4463, "step": 5369 }, { "epoch": 0.380716058135413, "grad_norm": 5.812004267324304, "learning_rate": 7.107221793165351e-06, "loss": 0.4332, "step": 5370 }, { "epoch": 0.38078695498050336, "grad_norm": 2.7329276774179587, "learning_rate": 7.106180524734881e-06, "loss": 0.44, "step": 5371 }, { "epoch": 0.3808578518255938, "grad_norm": 2.7643295078037133, "learning_rate": 7.105139145243902e-06, "loss": 0.3808, "step": 5372 }, { "epoch": 0.38092874867068416, "grad_norm": 1.861016574646936, "learning_rate": 7.104097654747332e-06, "loss": 0.3992, "step": 5373 }, { "epoch": 0.38099964551577453, "grad_norm": 2.1457311689402596, "learning_rate": 7.103056053300084e-06, "loss": 0.3926, "step": 5374 }, { "epoch": 0.38107054236086496, "grad_norm": 2.061125857348114, "learning_rate": 7.102014340957088e-06, "loss": 0.4576, "step": 5375 }, { "epoch": 0.38114143920595533, "grad_norm": 4.004680898964017, "learning_rate": 7.1009725177732695e-06, "loss": 0.3843, "step": 5376 }, { "epoch": 0.3812123360510457, "grad_norm": 6.873208627861662, "learning_rate": 7.099930583803566e-06, "loss": 0.4083, "step": 5377 }, { "epoch": 0.38128323289613614, "grad_norm": 3.4908818592406807, "learning_rate": 7.09888853910292e-06, "loss": 0.4547, "step": 5378 }, { "epoch": 0.3813541297412265, "grad_norm": 2.9944429946379953, "learning_rate": 7.097846383726283e-06, "loss": 0.4416, "step": 5379 }, { "epoch": 0.3814250265863169, "grad_norm": 2.5894576714288515, "learning_rate": 7.0968041177286e-06, "loss": 0.3969, "step": 5380 }, { "epoch": 0.3814959234314073, "grad_norm": 4.045092332774967, "learning_rate": 7.0957617411648385e-06, "loss": 0.4385, "step": 5381 }, { "epoch": 0.3815668202764977, "grad_norm": 3.268368858158725, "learning_rate": 7.09471925408996e-06, "loss": 0.4245, "step": 5382 }, { "epoch": 0.3816377171215881, "grad_norm": 3.860844245504445, "learning_rate": 7.093676656558936e-06, "loss": 0.4472, "step": 5383 }, { "epoch": 0.3817086139666785, "grad_norm": 2.717101866958117, "learning_rate": 7.092633948626744e-06, "loss": 0.4526, "step": 5384 }, { "epoch": 0.38177951081176886, "grad_norm": 3.5925311321686615, "learning_rate": 7.091591130348367e-06, "loss": 0.3771, "step": 5385 }, { "epoch": 0.3818504076568593, "grad_norm": 3.5590275472870463, "learning_rate": 7.090548201778794e-06, "loss": 0.4389, "step": 5386 }, { "epoch": 0.38192130450194967, "grad_norm": 1.8966234831552686, "learning_rate": 7.089505162973017e-06, "loss": 0.407, "step": 5387 }, { "epoch": 0.38199220134704004, "grad_norm": 2.1997735513571284, "learning_rate": 7.08846201398604e-06, "loss": 0.409, "step": 5388 }, { "epoch": 0.38206309819213047, "grad_norm": 3.919844305732151, "learning_rate": 7.087418754872864e-06, "loss": 0.4524, "step": 5389 }, { "epoch": 0.38213399503722084, "grad_norm": 4.25801881783859, "learning_rate": 7.086375385688505e-06, "loss": 0.3957, "step": 5390 }, { "epoch": 0.3822048918823112, "grad_norm": 2.579083905951233, "learning_rate": 7.08533190648798e-06, "loss": 0.4024, "step": 5391 }, { "epoch": 0.38227578872740164, "grad_norm": 2.269022109079093, "learning_rate": 7.0842883173263135e-06, "loss": 0.4332, "step": 5392 }, { "epoch": 0.382346685572492, "grad_norm": 3.622624856593421, "learning_rate": 7.08324461825853e-06, "loss": 0.4025, "step": 5393 }, { "epoch": 0.3824175824175824, "grad_norm": 5.854060839198888, "learning_rate": 7.08220080933967e-06, "loss": 0.4874, "step": 5394 }, { "epoch": 0.3824884792626728, "grad_norm": 3.6807587514912603, "learning_rate": 7.08115689062477e-06, "loss": 0.3598, "step": 5395 }, { "epoch": 0.3825593761077632, "grad_norm": 3.0115680962609996, "learning_rate": 7.080112862168881e-06, "loss": 0.422, "step": 5396 }, { "epoch": 0.3826302729528536, "grad_norm": 2.115571461121791, "learning_rate": 7.079068724027052e-06, "loss": 0.3704, "step": 5397 }, { "epoch": 0.382701169797944, "grad_norm": 3.679884209569866, "learning_rate": 7.078024476254341e-06, "loss": 0.386, "step": 5398 }, { "epoch": 0.38277206664303437, "grad_norm": 3.0354617248660407, "learning_rate": 7.076980118905815e-06, "loss": 0.4313, "step": 5399 }, { "epoch": 0.3828429634881248, "grad_norm": 3.55926643561753, "learning_rate": 7.075935652036543e-06, "loss": 0.3939, "step": 5400 }, { "epoch": 0.3829138603332152, "grad_norm": 2.807705530325691, "learning_rate": 7.074891075701598e-06, "loss": 0.421, "step": 5401 }, { "epoch": 0.38298475717830555, "grad_norm": 13.670374804158461, "learning_rate": 7.0738463899560626e-06, "loss": 0.4061, "step": 5402 }, { "epoch": 0.383055654023396, "grad_norm": 2.4666168393399346, "learning_rate": 7.0728015948550255e-06, "loss": 0.4412, "step": 5403 }, { "epoch": 0.38312655086848635, "grad_norm": 3.488196207720511, "learning_rate": 7.071756690453578e-06, "loss": 0.4424, "step": 5404 }, { "epoch": 0.3831974477135767, "grad_norm": 2.8609086549344807, "learning_rate": 7.070711676806819e-06, "loss": 0.408, "step": 5405 }, { "epoch": 0.38326834455866715, "grad_norm": 1.910296100665788, "learning_rate": 7.06966655396985e-06, "loss": 0.4412, "step": 5406 }, { "epoch": 0.3833392414037575, "grad_norm": 2.712537789819786, "learning_rate": 7.068621321997786e-06, "loss": 0.4336, "step": 5407 }, { "epoch": 0.38341013824884795, "grad_norm": 2.7401111636310413, "learning_rate": 7.0675759809457415e-06, "loss": 0.4069, "step": 5408 }, { "epoch": 0.3834810350939383, "grad_norm": 2.787206125701918, "learning_rate": 7.066530530868836e-06, "loss": 0.4146, "step": 5409 }, { "epoch": 0.3835519319390287, "grad_norm": 6.054791014923113, "learning_rate": 7.065484971822198e-06, "loss": 0.4151, "step": 5410 }, { "epoch": 0.38362282878411913, "grad_norm": 3.2921104710120206, "learning_rate": 7.064439303860963e-06, "loss": 0.4723, "step": 5411 }, { "epoch": 0.3836937256292095, "grad_norm": 2.5212626492211254, "learning_rate": 7.063393527040266e-06, "loss": 0.4336, "step": 5412 }, { "epoch": 0.3837646224742999, "grad_norm": 4.546158061940378, "learning_rate": 7.062347641415255e-06, "loss": 0.4153, "step": 5413 }, { "epoch": 0.3838355193193903, "grad_norm": 4.042029386938385, "learning_rate": 7.061301647041078e-06, "loss": 0.463, "step": 5414 }, { "epoch": 0.3839064161644807, "grad_norm": 2.4930327773828718, "learning_rate": 7.060255543972891e-06, "loss": 0.4027, "step": 5415 }, { "epoch": 0.38397731300957105, "grad_norm": 2.33475564375694, "learning_rate": 7.059209332265857e-06, "loss": 0.456, "step": 5416 }, { "epoch": 0.3840482098546615, "grad_norm": 3.5101841986128077, "learning_rate": 7.058163011975144e-06, "loss": 0.4144, "step": 5417 }, { "epoch": 0.38411910669975186, "grad_norm": 3.192460960275983, "learning_rate": 7.0571165831559255e-06, "loss": 0.3778, "step": 5418 }, { "epoch": 0.38419000354484223, "grad_norm": 4.582821763922537, "learning_rate": 7.056070045863375e-06, "loss": 0.4206, "step": 5419 }, { "epoch": 0.38426090038993266, "grad_norm": 3.4996998673603916, "learning_rate": 7.0550234001526866e-06, "loss": 0.4137, "step": 5420 }, { "epoch": 0.38433179723502303, "grad_norm": 3.6409710510595272, "learning_rate": 7.053976646079043e-06, "loss": 0.387, "step": 5421 }, { "epoch": 0.38440269408011346, "grad_norm": 2.4352302033868214, "learning_rate": 7.052929783697644e-06, "loss": 0.3856, "step": 5422 }, { "epoch": 0.38447359092520383, "grad_norm": 3.5763063063696694, "learning_rate": 7.0518828130636905e-06, "loss": 0.4478, "step": 5423 }, { "epoch": 0.3845444877702942, "grad_norm": 3.3210352465156734, "learning_rate": 7.05083573423239e-06, "loss": 0.4308, "step": 5424 }, { "epoch": 0.38461538461538464, "grad_norm": 3.4333509718581476, "learning_rate": 7.049788547258956e-06, "loss": 0.4495, "step": 5425 }, { "epoch": 0.384686281460475, "grad_norm": 3.7574113539437897, "learning_rate": 7.048741252198607e-06, "loss": 0.4619, "step": 5426 }, { "epoch": 0.3847571783055654, "grad_norm": 3.2713516034693626, "learning_rate": 7.047693849106568e-06, "loss": 0.4272, "step": 5427 }, { "epoch": 0.3848280751506558, "grad_norm": 3.713396826623444, "learning_rate": 7.046646338038071e-06, "loss": 0.4585, "step": 5428 }, { "epoch": 0.3848989719957462, "grad_norm": 6.634398078730633, "learning_rate": 7.045598719048349e-06, "loss": 0.4282, "step": 5429 }, { "epoch": 0.38496986884083656, "grad_norm": 4.556418408529556, "learning_rate": 7.0445509921926445e-06, "loss": 0.416, "step": 5430 }, { "epoch": 0.385040765685927, "grad_norm": 4.728705573372178, "learning_rate": 7.043503157526208e-06, "loss": 0.448, "step": 5431 }, { "epoch": 0.38511166253101736, "grad_norm": 2.045384437355744, "learning_rate": 7.042455215104288e-06, "loss": 0.3868, "step": 5432 }, { "epoch": 0.38518255937610774, "grad_norm": 4.7661307409448765, "learning_rate": 7.041407164982147e-06, "loss": 0.4037, "step": 5433 }, { "epoch": 0.38525345622119817, "grad_norm": 1.973676768556265, "learning_rate": 7.040359007215048e-06, "loss": 0.4128, "step": 5434 }, { "epoch": 0.38532435306628854, "grad_norm": 4.066626299766035, "learning_rate": 7.039310741858261e-06, "loss": 0.4129, "step": 5435 }, { "epoch": 0.38539524991137897, "grad_norm": 2.660505443389274, "learning_rate": 7.038262368967062e-06, "loss": 0.4056, "step": 5436 }, { "epoch": 0.38546614675646934, "grad_norm": 4.450129488264331, "learning_rate": 7.037213888596732e-06, "loss": 0.4104, "step": 5437 }, { "epoch": 0.3855370436015597, "grad_norm": 2.519193045022775, "learning_rate": 7.036165300802557e-06, "loss": 0.4418, "step": 5438 }, { "epoch": 0.38560794044665014, "grad_norm": 2.3049390716856397, "learning_rate": 7.035116605639836e-06, "loss": 0.4111, "step": 5439 }, { "epoch": 0.3856788372917405, "grad_norm": 2.3928456332544843, "learning_rate": 7.034067803163859e-06, "loss": 0.3964, "step": 5440 }, { "epoch": 0.3857497341368309, "grad_norm": 2.4544575602830325, "learning_rate": 7.033018893429937e-06, "loss": 0.3994, "step": 5441 }, { "epoch": 0.3858206309819213, "grad_norm": 3.419481119236445, "learning_rate": 7.031969876493375e-06, "loss": 0.3832, "step": 5442 }, { "epoch": 0.3858915278270117, "grad_norm": 4.008106266932976, "learning_rate": 7.030920752409491e-06, "loss": 0.4528, "step": 5443 }, { "epoch": 0.38596242467210207, "grad_norm": 2.790947925990463, "learning_rate": 7.029871521233606e-06, "loss": 0.4323, "step": 5444 }, { "epoch": 0.3860333215171925, "grad_norm": 2.0566037338091374, "learning_rate": 7.028822183021047e-06, "loss": 0.41, "step": 5445 }, { "epoch": 0.38610421836228287, "grad_norm": 2.1801848020147148, "learning_rate": 7.027772737827145e-06, "loss": 0.406, "step": 5446 }, { "epoch": 0.3861751152073733, "grad_norm": 3.062583740986227, "learning_rate": 7.026723185707239e-06, "loss": 0.4534, "step": 5447 }, { "epoch": 0.38624601205246367, "grad_norm": 9.141460608150442, "learning_rate": 7.025673526716672e-06, "loss": 0.3875, "step": 5448 }, { "epoch": 0.38631690889755405, "grad_norm": 2.9177981189247, "learning_rate": 7.024623760910794e-06, "loss": 0.3619, "step": 5449 }, { "epoch": 0.3863878057426445, "grad_norm": 3.5462927602689756, "learning_rate": 7.023573888344961e-06, "loss": 0.5315, "step": 5450 }, { "epoch": 0.38645870258773485, "grad_norm": 4.0322967468787345, "learning_rate": 7.02252390907453e-06, "loss": 0.4471, "step": 5451 }, { "epoch": 0.3865295994328252, "grad_norm": 6.73678610452412, "learning_rate": 7.0214738231548715e-06, "loss": 0.4461, "step": 5452 }, { "epoch": 0.38660049627791565, "grad_norm": 3.179901961455468, "learning_rate": 7.020423630641356e-06, "loss": 0.445, "step": 5453 }, { "epoch": 0.386671393123006, "grad_norm": 2.3872600492669735, "learning_rate": 7.019373331589358e-06, "loss": 0.405, "step": 5454 }, { "epoch": 0.3867422899680964, "grad_norm": 3.061131087932302, "learning_rate": 7.018322926054264e-06, "loss": 0.3967, "step": 5455 }, { "epoch": 0.3868131868131868, "grad_norm": 3.7751802263376923, "learning_rate": 7.017272414091461e-06, "loss": 0.4325, "step": 5456 }, { "epoch": 0.3868840836582772, "grad_norm": 3.059732886158707, "learning_rate": 7.016221795756344e-06, "loss": 0.4262, "step": 5457 }, { "epoch": 0.3869549805033676, "grad_norm": 7.10401620634118, "learning_rate": 7.015171071104314e-06, "loss": 0.3839, "step": 5458 }, { "epoch": 0.387025877348458, "grad_norm": 2.308014012361717, "learning_rate": 7.014120240190775e-06, "loss": 0.4183, "step": 5459 }, { "epoch": 0.3870967741935484, "grad_norm": 1.3784976807605107, "learning_rate": 7.013069303071137e-06, "loss": 0.684, "step": 5460 }, { "epoch": 0.3871676710386388, "grad_norm": 4.245177979683411, "learning_rate": 7.012018259800819e-06, "loss": 0.3852, "step": 5461 }, { "epoch": 0.3872385678837292, "grad_norm": 3.1834072728491747, "learning_rate": 7.010967110435243e-06, "loss": 0.4123, "step": 5462 }, { "epoch": 0.38730946472881955, "grad_norm": 4.049561866701871, "learning_rate": 7.009915855029836e-06, "loss": 0.4592, "step": 5463 }, { "epoch": 0.38738036157391, "grad_norm": 6.490169852778818, "learning_rate": 7.0088644936400306e-06, "loss": 0.3954, "step": 5464 }, { "epoch": 0.38745125841900036, "grad_norm": 4.774078352366998, "learning_rate": 7.00781302632127e-06, "loss": 0.4244, "step": 5465 }, { "epoch": 0.38752215526409073, "grad_norm": 4.66485765734932, "learning_rate": 7.006761453128994e-06, "loss": 0.4995, "step": 5466 }, { "epoch": 0.38759305210918116, "grad_norm": 5.723450276740201, "learning_rate": 7.0057097741186555e-06, "loss": 0.4309, "step": 5467 }, { "epoch": 0.38766394895427153, "grad_norm": 5.05741167537727, "learning_rate": 7.004657989345709e-06, "loss": 0.4234, "step": 5468 }, { "epoch": 0.3877348457993619, "grad_norm": 3.2389556131527653, "learning_rate": 7.003606098865618e-06, "loss": 0.4213, "step": 5469 }, { "epoch": 0.38780574264445233, "grad_norm": 2.7322498359976652, "learning_rate": 7.002554102733846e-06, "loss": 0.4183, "step": 5470 }, { "epoch": 0.3878766394895427, "grad_norm": 2.9572312249161232, "learning_rate": 7.00150200100587e-06, "loss": 0.3987, "step": 5471 }, { "epoch": 0.38794753633463314, "grad_norm": 2.039531951824729, "learning_rate": 7.000449793737163e-06, "loss": 0.3997, "step": 5472 }, { "epoch": 0.3880184331797235, "grad_norm": 7.207651685547178, "learning_rate": 6.999397480983214e-06, "loss": 0.4448, "step": 5473 }, { "epoch": 0.3880893300248139, "grad_norm": 2.811110988461488, "learning_rate": 6.99834506279951e-06, "loss": 0.4357, "step": 5474 }, { "epoch": 0.3881602268699043, "grad_norm": 5.19511779431636, "learning_rate": 6.997292539241543e-06, "loss": 0.3984, "step": 5475 }, { "epoch": 0.3882311237149947, "grad_norm": 3.985717336972039, "learning_rate": 6.996239910364818e-06, "loss": 0.427, "step": 5476 }, { "epoch": 0.38830202056008506, "grad_norm": 4.184857176503865, "learning_rate": 6.995187176224838e-06, "loss": 0.4097, "step": 5477 }, { "epoch": 0.3883729174051755, "grad_norm": 3.8456046265414585, "learning_rate": 6.994134336877117e-06, "loss": 0.3972, "step": 5478 }, { "epoch": 0.38844381425026586, "grad_norm": 2.390371733260541, "learning_rate": 6.993081392377169e-06, "loss": 0.4542, "step": 5479 }, { "epoch": 0.38851471109535624, "grad_norm": 3.1890647694364698, "learning_rate": 6.992028342780518e-06, "loss": 0.421, "step": 5480 }, { "epoch": 0.38858560794044666, "grad_norm": 3.120862064532439, "learning_rate": 6.990975188142691e-06, "loss": 0.4401, "step": 5481 }, { "epoch": 0.38865650478553704, "grad_norm": 5.166896546201213, "learning_rate": 6.989921928519226e-06, "loss": 0.4535, "step": 5482 }, { "epoch": 0.3887274016306274, "grad_norm": 2.4691753234028107, "learning_rate": 6.988868563965655e-06, "loss": 0.4485, "step": 5483 }, { "epoch": 0.38879829847571784, "grad_norm": 2.616875699136418, "learning_rate": 6.987815094537529e-06, "loss": 0.4222, "step": 5484 }, { "epoch": 0.3888691953208082, "grad_norm": 6.852259762581904, "learning_rate": 6.986761520290394e-06, "loss": 0.4391, "step": 5485 }, { "epoch": 0.38894009216589864, "grad_norm": 2.1563089966797846, "learning_rate": 6.9857078412798096e-06, "loss": 0.4777, "step": 5486 }, { "epoch": 0.389010989010989, "grad_norm": 2.653959713471806, "learning_rate": 6.984654057561333e-06, "loss": 0.4009, "step": 5487 }, { "epoch": 0.3890818858560794, "grad_norm": 6.643094005791345, "learning_rate": 6.9836001691905346e-06, "loss": 0.406, "step": 5488 }, { "epoch": 0.3891527827011698, "grad_norm": 3.2047213823210328, "learning_rate": 6.982546176222985e-06, "loss": 0.3962, "step": 5489 }, { "epoch": 0.3892236795462602, "grad_norm": 2.880859410215992, "learning_rate": 6.9814920787142625e-06, "loss": 0.3632, "step": 5490 }, { "epoch": 0.38929457639135057, "grad_norm": 11.786425952912023, "learning_rate": 6.98043787671995e-06, "loss": 0.384, "step": 5491 }, { "epoch": 0.389365473236441, "grad_norm": 4.539902250942735, "learning_rate": 6.979383570295637e-06, "loss": 0.4605, "step": 5492 }, { "epoch": 0.38943637008153137, "grad_norm": 3.048418184557269, "learning_rate": 6.978329159496917e-06, "loss": 0.4367, "step": 5493 }, { "epoch": 0.38950726692662174, "grad_norm": 4.4302899346326825, "learning_rate": 6.97727464437939e-06, "loss": 0.4303, "step": 5494 }, { "epoch": 0.38957816377171217, "grad_norm": 5.646057993950811, "learning_rate": 6.976220024998664e-06, "loss": 0.4126, "step": 5495 }, { "epoch": 0.38964906061680255, "grad_norm": 3.707078005468769, "learning_rate": 6.975165301410345e-06, "loss": 0.4305, "step": 5496 }, { "epoch": 0.3897199574618929, "grad_norm": 2.870744249524984, "learning_rate": 6.974110473670054e-06, "loss": 0.4382, "step": 5497 }, { "epoch": 0.38979085430698335, "grad_norm": 3.3565247413379495, "learning_rate": 6.973055541833409e-06, "loss": 0.4016, "step": 5498 }, { "epoch": 0.3898617511520737, "grad_norm": 6.300559602261966, "learning_rate": 6.972000505956042e-06, "loss": 0.412, "step": 5499 }, { "epoch": 0.38993264799716415, "grad_norm": 2.6869070777965156, "learning_rate": 6.970945366093581e-06, "loss": 0.3787, "step": 5500 }, { "epoch": 0.3900035448422545, "grad_norm": 4.0277303482851785, "learning_rate": 6.969890122301667e-06, "loss": 0.4204, "step": 5501 }, { "epoch": 0.3900744416873449, "grad_norm": 3.1300606601132483, "learning_rate": 6.968834774635943e-06, "loss": 0.4029, "step": 5502 }, { "epoch": 0.3901453385324353, "grad_norm": 4.37402293430478, "learning_rate": 6.9677793231520594e-06, "loss": 0.435, "step": 5503 }, { "epoch": 0.3902162353775257, "grad_norm": 5.469747679194209, "learning_rate": 6.96672376790567e-06, "loss": 0.4098, "step": 5504 }, { "epoch": 0.3902871322226161, "grad_norm": 5.784663375178671, "learning_rate": 6.965668108952436e-06, "loss": 0.3926, "step": 5505 }, { "epoch": 0.3903580290677065, "grad_norm": 15.899925180622644, "learning_rate": 6.964612346348021e-06, "loss": 0.4748, "step": 5506 }, { "epoch": 0.3904289259127969, "grad_norm": 2.2901346649593393, "learning_rate": 6.963556480148097e-06, "loss": 0.4268, "step": 5507 }, { "epoch": 0.39049982275788725, "grad_norm": 6.440170315596347, "learning_rate": 6.962500510408342e-06, "loss": 0.356, "step": 5508 }, { "epoch": 0.3905707196029777, "grad_norm": 2.653676318839232, "learning_rate": 6.961444437184437e-06, "loss": 0.3956, "step": 5509 }, { "epoch": 0.39064161644806805, "grad_norm": 15.205569961771321, "learning_rate": 6.96038826053207e-06, "loss": 0.4309, "step": 5510 }, { "epoch": 0.3907125132931585, "grad_norm": 2.0101802486879037, "learning_rate": 6.959331980506934e-06, "loss": 0.3902, "step": 5511 }, { "epoch": 0.39078341013824885, "grad_norm": 6.851231284628757, "learning_rate": 6.958275597164726e-06, "loss": 0.4242, "step": 5512 }, { "epoch": 0.39085430698333923, "grad_norm": 3.0945025164064424, "learning_rate": 6.957219110561152e-06, "loss": 0.424, "step": 5513 }, { "epoch": 0.39092520382842966, "grad_norm": 8.19826839196917, "learning_rate": 6.9561625207519204e-06, "loss": 0.4801, "step": 5514 }, { "epoch": 0.39099610067352003, "grad_norm": 4.199571222435286, "learning_rate": 6.9551058277927455e-06, "loss": 0.3722, "step": 5515 }, { "epoch": 0.3910669975186104, "grad_norm": 2.647952962313968, "learning_rate": 6.954049031739349e-06, "loss": 0.4031, "step": 5516 }, { "epoch": 0.39113789436370083, "grad_norm": 2.3555789869790664, "learning_rate": 6.952992132647453e-06, "loss": 0.3633, "step": 5517 }, { "epoch": 0.3912087912087912, "grad_norm": 1.1000146323079496, "learning_rate": 6.951935130572795e-06, "loss": 0.7284, "step": 5518 }, { "epoch": 0.3912796880538816, "grad_norm": 2.7649546121525552, "learning_rate": 6.950878025571105e-06, "loss": 0.4056, "step": 5519 }, { "epoch": 0.391350584898972, "grad_norm": 3.249096086915111, "learning_rate": 6.949820817698128e-06, "loss": 0.4001, "step": 5520 }, { "epoch": 0.3914214817440624, "grad_norm": 3.767214337327614, "learning_rate": 6.9487635070096114e-06, "loss": 0.4464, "step": 5521 }, { "epoch": 0.39149237858915276, "grad_norm": 2.675796841440316, "learning_rate": 6.947706093561308e-06, "loss": 0.3995, "step": 5522 }, { "epoch": 0.3915632754342432, "grad_norm": 1.0871704735861436, "learning_rate": 6.946648577408976e-06, "loss": 0.6844, "step": 5523 }, { "epoch": 0.39163417227933356, "grad_norm": 3.0437773145822806, "learning_rate": 6.945590958608378e-06, "loss": 0.4635, "step": 5524 }, { "epoch": 0.391705069124424, "grad_norm": 2.8559047613842825, "learning_rate": 6.944533237215284e-06, "loss": 0.435, "step": 5525 }, { "epoch": 0.39177596596951436, "grad_norm": 2.977198795916163, "learning_rate": 6.943475413285467e-06, "loss": 0.4366, "step": 5526 }, { "epoch": 0.39184686281460474, "grad_norm": 4.410703765201396, "learning_rate": 6.9424174868747105e-06, "loss": 0.4214, "step": 5527 }, { "epoch": 0.39191775965969516, "grad_norm": 3.9769419294060895, "learning_rate": 6.941359458038795e-06, "loss": 0.4514, "step": 5528 }, { "epoch": 0.39198865650478554, "grad_norm": 4.073786495018158, "learning_rate": 6.940301326833516e-06, "loss": 0.45, "step": 5529 }, { "epoch": 0.3920595533498759, "grad_norm": 2.8131047299575633, "learning_rate": 6.939243093314665e-06, "loss": 0.4277, "step": 5530 }, { "epoch": 0.39213045019496634, "grad_norm": 2.701085909574361, "learning_rate": 6.938184757538047e-06, "loss": 0.4099, "step": 5531 }, { "epoch": 0.3922013470400567, "grad_norm": 3.578906241763545, "learning_rate": 6.9371263195594664e-06, "loss": 0.4193, "step": 5532 }, { "epoch": 0.3922722438851471, "grad_norm": 1.337423913378543, "learning_rate": 6.936067779434737e-06, "loss": 0.7056, "step": 5533 }, { "epoch": 0.3923431407302375, "grad_norm": 4.647323270389333, "learning_rate": 6.935009137219675e-06, "loss": 0.3992, "step": 5534 }, { "epoch": 0.3924140375753279, "grad_norm": 4.841131120650471, "learning_rate": 6.933950392970104e-06, "loss": 0.3981, "step": 5535 }, { "epoch": 0.3924849344204183, "grad_norm": 4.234680922773508, "learning_rate": 6.932891546741853e-06, "loss": 0.4349, "step": 5536 }, { "epoch": 0.3925558312655087, "grad_norm": 6.963489885899931, "learning_rate": 6.9318325985907565e-06, "loss": 0.4598, "step": 5537 }, { "epoch": 0.39262672811059907, "grad_norm": 3.9986811567300857, "learning_rate": 6.930773548572651e-06, "loss": 0.4386, "step": 5538 }, { "epoch": 0.3926976249556895, "grad_norm": 5.072805126514667, "learning_rate": 6.929714396743384e-06, "loss": 0.4371, "step": 5539 }, { "epoch": 0.39276852180077987, "grad_norm": 4.386241803920276, "learning_rate": 6.928655143158804e-06, "loss": 0.4208, "step": 5540 }, { "epoch": 0.39283941864587024, "grad_norm": 4.563356575357596, "learning_rate": 6.927595787874764e-06, "loss": 0.452, "step": 5541 }, { "epoch": 0.39291031549096067, "grad_norm": 2.7924397531934066, "learning_rate": 6.92653633094713e-06, "loss": 0.3738, "step": 5542 }, { "epoch": 0.39298121233605104, "grad_norm": 5.908734737370971, "learning_rate": 6.925476772431763e-06, "loss": 0.3954, "step": 5543 }, { "epoch": 0.3930521091811414, "grad_norm": 2.260278143637564, "learning_rate": 6.924417112384536e-06, "loss": 0.3927, "step": 5544 }, { "epoch": 0.39312300602623185, "grad_norm": 13.506880430284028, "learning_rate": 6.923357350861326e-06, "loss": 0.4139, "step": 5545 }, { "epoch": 0.3931939028713222, "grad_norm": 2.754726167490118, "learning_rate": 6.922297487918016e-06, "loss": 0.4249, "step": 5546 }, { "epoch": 0.3932647997164126, "grad_norm": 4.065845430122717, "learning_rate": 6.921237523610491e-06, "loss": 0.3991, "step": 5547 }, { "epoch": 0.393335696561503, "grad_norm": 2.060905099250869, "learning_rate": 6.9201774579946455e-06, "loss": 0.4362, "step": 5548 }, { "epoch": 0.3934065934065934, "grad_norm": 5.259496284508251, "learning_rate": 6.919117291126376e-06, "loss": 0.4143, "step": 5549 }, { "epoch": 0.3934774902516838, "grad_norm": 1.8994158454811725, "learning_rate": 6.918057023061588e-06, "loss": 0.4593, "step": 5550 }, { "epoch": 0.3935483870967742, "grad_norm": 4.917629753956283, "learning_rate": 6.91699665385619e-06, "loss": 0.4295, "step": 5551 }, { "epoch": 0.3936192839418646, "grad_norm": 4.223381864338879, "learning_rate": 6.915936183566094e-06, "loss": 0.3582, "step": 5552 }, { "epoch": 0.393690180786955, "grad_norm": 2.7108686633509222, "learning_rate": 6.914875612247221e-06, "loss": 0.3736, "step": 5553 }, { "epoch": 0.3937610776320454, "grad_norm": 2.9924458484187735, "learning_rate": 6.9138149399554945e-06, "loss": 0.4311, "step": 5554 }, { "epoch": 0.39383197447713575, "grad_norm": 3.5113938026800358, "learning_rate": 6.912754166746847e-06, "loss": 0.4245, "step": 5555 }, { "epoch": 0.3939028713222262, "grad_norm": 0.9259365558880532, "learning_rate": 6.911693292677211e-06, "loss": 0.6955, "step": 5556 }, { "epoch": 0.39397376816731655, "grad_norm": 2.872374800902008, "learning_rate": 6.9106323178025304e-06, "loss": 0.3967, "step": 5557 }, { "epoch": 0.3940446650124069, "grad_norm": 3.1196078130311498, "learning_rate": 6.909571242178747e-06, "loss": 0.3924, "step": 5558 }, { "epoch": 0.39411556185749735, "grad_norm": 2.6619210047920547, "learning_rate": 6.908510065861816e-06, "loss": 0.3932, "step": 5559 }, { "epoch": 0.3941864587025877, "grad_norm": 3.120414241239408, "learning_rate": 6.907448788907692e-06, "loss": 0.4069, "step": 5560 }, { "epoch": 0.3942573555476781, "grad_norm": 50.10983942124344, "learning_rate": 6.9063874113723386e-06, "loss": 0.4056, "step": 5561 }, { "epoch": 0.39432825239276853, "grad_norm": 5.001568169784329, "learning_rate": 6.90532593331172e-06, "loss": 0.4042, "step": 5562 }, { "epoch": 0.3943991492378589, "grad_norm": 2.1731135323583457, "learning_rate": 6.9042643547818125e-06, "loss": 0.3921, "step": 5563 }, { "epoch": 0.39447004608294933, "grad_norm": 2.6213169023379383, "learning_rate": 6.90320267583859e-06, "loss": 0.4653, "step": 5564 }, { "epoch": 0.3945409429280397, "grad_norm": 3.7171708472362575, "learning_rate": 6.90214089653804e-06, "loss": 0.413, "step": 5565 }, { "epoch": 0.3946118397731301, "grad_norm": 3.5881894120600757, "learning_rate": 6.901079016936148e-06, "loss": 0.4798, "step": 5566 }, { "epoch": 0.3946827366182205, "grad_norm": 2.151517598292493, "learning_rate": 6.900017037088908e-06, "loss": 0.4296, "step": 5567 }, { "epoch": 0.3947536334633109, "grad_norm": 5.155391738352437, "learning_rate": 6.898954957052318e-06, "loss": 0.442, "step": 5568 }, { "epoch": 0.39482453030840126, "grad_norm": 2.149874879753799, "learning_rate": 6.897892776882387e-06, "loss": 0.4288, "step": 5569 }, { "epoch": 0.3948954271534917, "grad_norm": 2.102229572722177, "learning_rate": 6.896830496635118e-06, "loss": 0.4582, "step": 5570 }, { "epoch": 0.39496632399858206, "grad_norm": 2.405485293383411, "learning_rate": 6.8957681163665316e-06, "loss": 0.3605, "step": 5571 }, { "epoch": 0.39503722084367243, "grad_norm": 2.1874386455856265, "learning_rate": 6.894705636132644e-06, "loss": 0.4515, "step": 5572 }, { "epoch": 0.39510811768876286, "grad_norm": 2.3254336945201506, "learning_rate": 6.893643055989483e-06, "loss": 0.362, "step": 5573 }, { "epoch": 0.39517901453385323, "grad_norm": 3.03124967608746, "learning_rate": 6.892580375993079e-06, "loss": 0.4128, "step": 5574 }, { "epoch": 0.39524991137894366, "grad_norm": 3.020855067048681, "learning_rate": 6.891517596199464e-06, "loss": 0.4122, "step": 5575 }, { "epoch": 0.39532080822403404, "grad_norm": 2.6853488046781653, "learning_rate": 6.890454716664685e-06, "loss": 0.4441, "step": 5576 }, { "epoch": 0.3953917050691244, "grad_norm": 4.034143120533322, "learning_rate": 6.889391737444786e-06, "loss": 0.4184, "step": 5577 }, { "epoch": 0.39546260191421484, "grad_norm": 8.750371355908248, "learning_rate": 6.8883286585958175e-06, "loss": 0.364, "step": 5578 }, { "epoch": 0.3955334987593052, "grad_norm": 5.490454862170599, "learning_rate": 6.8872654801738385e-06, "loss": 0.4521, "step": 5579 }, { "epoch": 0.3956043956043956, "grad_norm": 3.686078786036114, "learning_rate": 6.886202202234908e-06, "loss": 0.434, "step": 5580 }, { "epoch": 0.395675292449486, "grad_norm": 3.805688316502408, "learning_rate": 6.885138824835098e-06, "loss": 0.4407, "step": 5581 }, { "epoch": 0.3957461892945764, "grad_norm": 2.5005262246830395, "learning_rate": 6.884075348030479e-06, "loss": 0.4179, "step": 5582 }, { "epoch": 0.39581708613966676, "grad_norm": 8.777822437652645, "learning_rate": 6.8830117718771285e-06, "loss": 0.4609, "step": 5583 }, { "epoch": 0.3958879829847572, "grad_norm": 3.0932153855897013, "learning_rate": 6.881948096431129e-06, "loss": 0.4447, "step": 5584 }, { "epoch": 0.39595887982984757, "grad_norm": 2.0061490039138463, "learning_rate": 6.880884321748571e-06, "loss": 0.4574, "step": 5585 }, { "epoch": 0.39602977667493794, "grad_norm": 3.5201386136192006, "learning_rate": 6.879820447885547e-06, "loss": 0.4106, "step": 5586 }, { "epoch": 0.39610067352002837, "grad_norm": 18.9622620596516, "learning_rate": 6.878756474898156e-06, "loss": 0.4022, "step": 5587 }, { "epoch": 0.39617157036511874, "grad_norm": 3.001122669891751, "learning_rate": 6.877692402842501e-06, "loss": 0.4109, "step": 5588 }, { "epoch": 0.39624246721020917, "grad_norm": 6.392747583558897, "learning_rate": 6.876628231774694e-06, "loss": 0.4724, "step": 5589 }, { "epoch": 0.39631336405529954, "grad_norm": 1.772742628665159, "learning_rate": 6.875563961750847e-06, "loss": 0.3975, "step": 5590 }, { "epoch": 0.3963842609003899, "grad_norm": 3.0476798295954133, "learning_rate": 6.8744995928270804e-06, "loss": 0.4714, "step": 5591 }, { "epoch": 0.39645515774548035, "grad_norm": 2.242994400226948, "learning_rate": 6.87343512505952e-06, "loss": 0.3887, "step": 5592 }, { "epoch": 0.3965260545905707, "grad_norm": 2.333777727739821, "learning_rate": 6.872370558504295e-06, "loss": 0.399, "step": 5593 }, { "epoch": 0.3965969514356611, "grad_norm": 7.767495909673519, "learning_rate": 6.8713058932175425e-06, "loss": 0.468, "step": 5594 }, { "epoch": 0.3966678482807515, "grad_norm": 3.9153295010296127, "learning_rate": 6.870241129255403e-06, "loss": 0.4252, "step": 5595 }, { "epoch": 0.3967387451258419, "grad_norm": 8.658237035770389, "learning_rate": 6.869176266674018e-06, "loss": 0.4331, "step": 5596 }, { "epoch": 0.39680964197093227, "grad_norm": 5.240747362666961, "learning_rate": 6.868111305529545e-06, "loss": 0.4472, "step": 5597 }, { "epoch": 0.3968805388160227, "grad_norm": 4.84141548480063, "learning_rate": 6.867046245878135e-06, "loss": 0.4921, "step": 5598 }, { "epoch": 0.3969514356611131, "grad_norm": 2.2862274979786976, "learning_rate": 6.865981087775951e-06, "loss": 0.3863, "step": 5599 }, { "epoch": 0.3970223325062035, "grad_norm": 3.204274671873896, "learning_rate": 6.864915831279162e-06, "loss": 0.4077, "step": 5600 }, { "epoch": 0.3970932293512939, "grad_norm": 3.181940806008864, "learning_rate": 6.863850476443935e-06, "loss": 0.4255, "step": 5601 }, { "epoch": 0.39716412619638425, "grad_norm": 5.754592107635311, "learning_rate": 6.862785023326453e-06, "loss": 0.4384, "step": 5602 }, { "epoch": 0.3972350230414747, "grad_norm": 1.0564202733426378, "learning_rate": 6.861719471982893e-06, "loss": 0.6616, "step": 5603 }, { "epoch": 0.39730591988656505, "grad_norm": 2.7782974995974272, "learning_rate": 6.860653822469444e-06, "loss": 0.379, "step": 5604 }, { "epoch": 0.3973768167316554, "grad_norm": 5.691223679943281, "learning_rate": 6.859588074842299e-06, "loss": 0.3963, "step": 5605 }, { "epoch": 0.39744771357674585, "grad_norm": 3.9660931713865746, "learning_rate": 6.858522229157657e-06, "loss": 0.4044, "step": 5606 }, { "epoch": 0.3975186104218362, "grad_norm": 6.269382023745322, "learning_rate": 6.8574562854717174e-06, "loss": 0.4055, "step": 5607 }, { "epoch": 0.3975895072669266, "grad_norm": 4.0935268817233155, "learning_rate": 6.856390243840692e-06, "loss": 0.4491, "step": 5608 }, { "epoch": 0.39766040411201703, "grad_norm": 3.2790724700825837, "learning_rate": 6.855324104320789e-06, "loss": 0.4075, "step": 5609 }, { "epoch": 0.3977313009571074, "grad_norm": 4.07891316612055, "learning_rate": 6.854257866968232e-06, "loss": 0.4447, "step": 5610 }, { "epoch": 0.3978021978021978, "grad_norm": 5.026003434615423, "learning_rate": 6.8531915318392425e-06, "loss": 0.4069, "step": 5611 }, { "epoch": 0.3978730946472882, "grad_norm": 6.756015663361198, "learning_rate": 6.852125098990049e-06, "loss": 0.45, "step": 5612 }, { "epoch": 0.3979439914923786, "grad_norm": 2.7340202097104997, "learning_rate": 6.851058568476885e-06, "loss": 0.3808, "step": 5613 }, { "epoch": 0.398014888337469, "grad_norm": 4.225794417274699, "learning_rate": 6.8499919403559915e-06, "loss": 0.4277, "step": 5614 }, { "epoch": 0.3980857851825594, "grad_norm": 4.770129183927175, "learning_rate": 6.848925214683608e-06, "loss": 0.4072, "step": 5615 }, { "epoch": 0.39815668202764976, "grad_norm": 6.258680661031884, "learning_rate": 6.847858391515989e-06, "loss": 0.4472, "step": 5616 }, { "epoch": 0.3982275788727402, "grad_norm": 4.4187687967357085, "learning_rate": 6.846791470909386e-06, "loss": 0.4253, "step": 5617 }, { "epoch": 0.39829847571783056, "grad_norm": 3.218798402620337, "learning_rate": 6.845724452920059e-06, "loss": 0.4681, "step": 5618 }, { "epoch": 0.39836937256292093, "grad_norm": 5.009564644159081, "learning_rate": 6.844657337604272e-06, "loss": 0.4869, "step": 5619 }, { "epoch": 0.39844026940801136, "grad_norm": 2.800241971848261, "learning_rate": 6.8435901250182956e-06, "loss": 0.4611, "step": 5620 }, { "epoch": 0.39851116625310173, "grad_norm": 4.47680233040745, "learning_rate": 6.842522815218406e-06, "loss": 0.3682, "step": 5621 }, { "epoch": 0.3985820630981921, "grad_norm": 2.8153116342373585, "learning_rate": 6.8414554082608816e-06, "loss": 0.3787, "step": 5622 }, { "epoch": 0.39865295994328254, "grad_norm": 2.090759067641805, "learning_rate": 6.840387904202008e-06, "loss": 0.3762, "step": 5623 }, { "epoch": 0.3987238567883729, "grad_norm": 2.0792333006171955, "learning_rate": 6.839320303098076e-06, "loss": 0.3909, "step": 5624 }, { "epoch": 0.3987947536334633, "grad_norm": 2.2530768904562377, "learning_rate": 6.83825260500538e-06, "loss": 0.4075, "step": 5625 }, { "epoch": 0.3988656504785537, "grad_norm": 2.6834821099304262, "learning_rate": 6.837184809980219e-06, "loss": 0.3989, "step": 5626 }, { "epoch": 0.3989365473236441, "grad_norm": 4.339788242685937, "learning_rate": 6.836116918078905e-06, "loss": 0.4324, "step": 5627 }, { "epoch": 0.3990074441687345, "grad_norm": 4.975550225061824, "learning_rate": 6.8350489293577415e-06, "loss": 0.3911, "step": 5628 }, { "epoch": 0.3990783410138249, "grad_norm": 3.4771585200442137, "learning_rate": 6.833980843873048e-06, "loss": 0.404, "step": 5629 }, { "epoch": 0.39914923785891526, "grad_norm": 6.451694699319634, "learning_rate": 6.8329126616811444e-06, "loss": 0.3946, "step": 5630 }, { "epoch": 0.3992201347040057, "grad_norm": 3.358722824891993, "learning_rate": 6.8318443828383574e-06, "loss": 0.3883, "step": 5631 }, { "epoch": 0.39929103154909606, "grad_norm": 4.423374235727431, "learning_rate": 6.83077600740102e-06, "loss": 0.4303, "step": 5632 }, { "epoch": 0.39936192839418644, "grad_norm": 4.781384451804024, "learning_rate": 6.829707535425463e-06, "loss": 0.4257, "step": 5633 }, { "epoch": 0.39943282523927687, "grad_norm": 4.485564011434575, "learning_rate": 6.828638966968033e-06, "loss": 0.4441, "step": 5634 }, { "epoch": 0.39950372208436724, "grad_norm": 2.431078581450714, "learning_rate": 6.827570302085074e-06, "loss": 0.4467, "step": 5635 }, { "epoch": 0.3995746189294576, "grad_norm": 37.35027221919818, "learning_rate": 6.826501540832938e-06, "loss": 0.3577, "step": 5636 }, { "epoch": 0.39964551577454804, "grad_norm": 1.8169904796350027, "learning_rate": 6.825432683267982e-06, "loss": 0.4018, "step": 5637 }, { "epoch": 0.3997164126196384, "grad_norm": 4.08744844602923, "learning_rate": 6.824363729446567e-06, "loss": 0.4613, "step": 5638 }, { "epoch": 0.39978730946472885, "grad_norm": 2.188658619620782, "learning_rate": 6.82329467942506e-06, "loss": 0.345, "step": 5639 }, { "epoch": 0.3998582063098192, "grad_norm": 7.761960123255524, "learning_rate": 6.822225533259834e-06, "loss": 0.4282, "step": 5640 }, { "epoch": 0.3999291031549096, "grad_norm": 1.073166301589039, "learning_rate": 6.821156291007262e-06, "loss": 0.6924, "step": 5641 }, { "epoch": 0.4, "grad_norm": 5.420746053549114, "learning_rate": 6.820086952723732e-06, "loss": 0.4361, "step": 5642 }, { "epoch": 0.4000708968450904, "grad_norm": 3.016419554887126, "learning_rate": 6.819017518465626e-06, "loss": 0.4587, "step": 5643 }, { "epoch": 0.40014179369018077, "grad_norm": 4.040366158399714, "learning_rate": 6.8179479882893376e-06, "loss": 0.4567, "step": 5644 }, { "epoch": 0.4002126905352712, "grad_norm": 2.639818889558232, "learning_rate": 6.816878362251265e-06, "loss": 0.4619, "step": 5645 }, { "epoch": 0.40028358738036157, "grad_norm": 2.715457660543708, "learning_rate": 6.815808640407809e-06, "loss": 0.4595, "step": 5646 }, { "epoch": 0.40035448422545195, "grad_norm": 2.1410658627109895, "learning_rate": 6.814738822815379e-06, "loss": 0.4551, "step": 5647 }, { "epoch": 0.4004253810705424, "grad_norm": 5.004845366491596, "learning_rate": 6.813668909530384e-06, "loss": 0.4499, "step": 5648 }, { "epoch": 0.40049627791563275, "grad_norm": 2.4280038882913924, "learning_rate": 6.812598900609244e-06, "loss": 0.3973, "step": 5649 }, { "epoch": 0.4005671747607231, "grad_norm": 7.3946046687169416, "learning_rate": 6.81152879610838e-06, "loss": 0.4642, "step": 5650 }, { "epoch": 0.40063807160581355, "grad_norm": 3.0678596187980456, "learning_rate": 6.810458596084222e-06, "loss": 0.4101, "step": 5651 }, { "epoch": 0.4007089684509039, "grad_norm": 1.9671235575435595, "learning_rate": 6.809388300593196e-06, "loss": 0.4309, "step": 5652 }, { "epoch": 0.40077986529599435, "grad_norm": 4.325128793631886, "learning_rate": 6.808317909691747e-06, "loss": 0.412, "step": 5653 }, { "epoch": 0.4008507621410847, "grad_norm": 3.047849493853129, "learning_rate": 6.807247423436314e-06, "loss": 0.4016, "step": 5654 }, { "epoch": 0.4009216589861751, "grad_norm": 3.1569436600776815, "learning_rate": 6.8061768418833456e-06, "loss": 0.4243, "step": 5655 }, { "epoch": 0.40099255583126553, "grad_norm": 7.132404242492918, "learning_rate": 6.805106165089292e-06, "loss": 0.4243, "step": 5656 }, { "epoch": 0.4010634526763559, "grad_norm": 2.5834927618048376, "learning_rate": 6.804035393110613e-06, "loss": 0.4513, "step": 5657 }, { "epoch": 0.4011343495214463, "grad_norm": 4.392800911540687, "learning_rate": 6.802964526003771e-06, "loss": 0.3847, "step": 5658 }, { "epoch": 0.4012052463665367, "grad_norm": 5.072837659095561, "learning_rate": 6.801893563825235e-06, "loss": 0.4641, "step": 5659 }, { "epoch": 0.4012761432116271, "grad_norm": 3.1534486761624554, "learning_rate": 6.8008225066314735e-06, "loss": 0.3727, "step": 5660 }, { "epoch": 0.40134704005671745, "grad_norm": 3.9444831752733336, "learning_rate": 6.799751354478968e-06, "loss": 0.4043, "step": 5661 }, { "epoch": 0.4014179369018079, "grad_norm": 2.0116257173323233, "learning_rate": 6.798680107424199e-06, "loss": 0.436, "step": 5662 }, { "epoch": 0.40148883374689825, "grad_norm": 3.479660357691865, "learning_rate": 6.7976087655236555e-06, "loss": 0.3968, "step": 5663 }, { "epoch": 0.4015597305919887, "grad_norm": 3.8961802669536807, "learning_rate": 6.796537328833832e-06, "loss": 0.402, "step": 5664 }, { "epoch": 0.40163062743707906, "grad_norm": 1.5548703248730873, "learning_rate": 6.79546579741122e-06, "loss": 0.4415, "step": 5665 }, { "epoch": 0.40170152428216943, "grad_norm": 4.200097789234065, "learning_rate": 6.79439417131233e-06, "loss": 0.4323, "step": 5666 }, { "epoch": 0.40177242112725986, "grad_norm": 2.8143917817974784, "learning_rate": 6.793322450593662e-06, "loss": 0.3927, "step": 5667 }, { "epoch": 0.40184331797235023, "grad_norm": 2.9410633199042215, "learning_rate": 6.792250635311734e-06, "loss": 0.4155, "step": 5668 }, { "epoch": 0.4019142148174406, "grad_norm": 2.764520605749477, "learning_rate": 6.791178725523063e-06, "loss": 0.4523, "step": 5669 }, { "epoch": 0.40198511166253104, "grad_norm": 58.468239470780304, "learning_rate": 6.79010672128417e-06, "loss": 0.4499, "step": 5670 }, { "epoch": 0.4020560085076214, "grad_norm": 3.0160915675318836, "learning_rate": 6.789034622651583e-06, "loss": 0.4143, "step": 5671 }, { "epoch": 0.4021269053527118, "grad_norm": 2.584574604777591, "learning_rate": 6.787962429681836e-06, "loss": 0.4531, "step": 5672 }, { "epoch": 0.4021978021978022, "grad_norm": 2.1797237893369195, "learning_rate": 6.786890142431463e-06, "loss": 0.4498, "step": 5673 }, { "epoch": 0.4022686990428926, "grad_norm": 5.981881631175128, "learning_rate": 6.785817760957011e-06, "loss": 0.3935, "step": 5674 }, { "epoch": 0.40233959588798296, "grad_norm": 9.125895653657684, "learning_rate": 6.7847452853150245e-06, "loss": 0.4084, "step": 5675 }, { "epoch": 0.4024104927330734, "grad_norm": 3.7711575176594585, "learning_rate": 6.7836727155620565e-06, "loss": 0.4121, "step": 5676 }, { "epoch": 0.40248138957816376, "grad_norm": 5.142920675999007, "learning_rate": 6.782600051754665e-06, "loss": 0.4885, "step": 5677 }, { "epoch": 0.4025522864232542, "grad_norm": 2.3889656772048258, "learning_rate": 6.781527293949413e-06, "loss": 0.4632, "step": 5678 }, { "epoch": 0.40262318326834456, "grad_norm": 3.301247276703917, "learning_rate": 6.780454442202867e-06, "loss": 0.3803, "step": 5679 }, { "epoch": 0.40269408011343494, "grad_norm": 2.4079329783061794, "learning_rate": 6.779381496571599e-06, "loss": 0.4221, "step": 5680 }, { "epoch": 0.40276497695852537, "grad_norm": 2.57390838660571, "learning_rate": 6.778308457112186e-06, "loss": 0.3937, "step": 5681 }, { "epoch": 0.40283587380361574, "grad_norm": 4.636748925132673, "learning_rate": 6.77723532388121e-06, "loss": 0.4219, "step": 5682 }, { "epoch": 0.4029067706487061, "grad_norm": 2.7401944379905467, "learning_rate": 6.77616209693526e-06, "loss": 0.3822, "step": 5683 }, { "epoch": 0.40297766749379654, "grad_norm": 2.517832376488459, "learning_rate": 6.775088776330927e-06, "loss": 0.4124, "step": 5684 }, { "epoch": 0.4030485643388869, "grad_norm": 3.3887162100941834, "learning_rate": 6.77401536212481e-06, "loss": 0.4136, "step": 5685 }, { "epoch": 0.4031194611839773, "grad_norm": 3.8328580493077076, "learning_rate": 6.772941854373505e-06, "loss": 0.4226, "step": 5686 }, { "epoch": 0.4031903580290677, "grad_norm": 3.6370156032836682, "learning_rate": 6.771868253133626e-06, "loss": 0.4122, "step": 5687 }, { "epoch": 0.4032612548741581, "grad_norm": 4.3091308937093125, "learning_rate": 6.77079455846178e-06, "loss": 0.4644, "step": 5688 }, { "epoch": 0.40333215171924847, "grad_norm": 2.6214074160306042, "learning_rate": 6.769720770414585e-06, "loss": 0.4517, "step": 5689 }, { "epoch": 0.4034030485643389, "grad_norm": 2.836239481777675, "learning_rate": 6.768646889048662e-06, "loss": 0.4078, "step": 5690 }, { "epoch": 0.40347394540942927, "grad_norm": 3.2108940378893487, "learning_rate": 6.767572914420642e-06, "loss": 0.4686, "step": 5691 }, { "epoch": 0.4035448422545197, "grad_norm": 6.143577220796909, "learning_rate": 6.76649884658715e-06, "loss": 0.4049, "step": 5692 }, { "epoch": 0.40361573909961007, "grad_norm": 4.0047866066643225, "learning_rate": 6.765424685604827e-06, "loss": 0.4426, "step": 5693 }, { "epoch": 0.40368663594470044, "grad_norm": 5.652574940441517, "learning_rate": 6.764350431530313e-06, "loss": 0.425, "step": 5694 }, { "epoch": 0.4037575327897909, "grad_norm": 5.128894358873319, "learning_rate": 6.763276084420254e-06, "loss": 0.3774, "step": 5695 }, { "epoch": 0.40382842963488125, "grad_norm": 6.413232792308942, "learning_rate": 6.762201644331301e-06, "loss": 0.4425, "step": 5696 }, { "epoch": 0.4038993264799716, "grad_norm": 3.3905312241416263, "learning_rate": 6.761127111320109e-06, "loss": 0.4237, "step": 5697 }, { "epoch": 0.40397022332506205, "grad_norm": 2.983211886256957, "learning_rate": 6.760052485443342e-06, "loss": 0.4121, "step": 5698 }, { "epoch": 0.4040411201701524, "grad_norm": 4.127570617434794, "learning_rate": 6.7589777667576625e-06, "loss": 0.3918, "step": 5699 }, { "epoch": 0.4041120170152428, "grad_norm": 2.5716194962901895, "learning_rate": 6.757902955319745e-06, "loss": 0.371, "step": 5700 }, { "epoch": 0.4041829138603332, "grad_norm": 2.0635510941355877, "learning_rate": 6.756828051186261e-06, "loss": 0.436, "step": 5701 }, { "epoch": 0.4042538107054236, "grad_norm": 3.0846190296430938, "learning_rate": 6.755753054413894e-06, "loss": 0.4626, "step": 5702 }, { "epoch": 0.40432470755051403, "grad_norm": 2.8719536898955482, "learning_rate": 6.754677965059327e-06, "loss": 0.4791, "step": 5703 }, { "epoch": 0.4043956043956044, "grad_norm": 4.40398138783608, "learning_rate": 6.753602783179253e-06, "loss": 0.384, "step": 5704 }, { "epoch": 0.4044665012406948, "grad_norm": 3.759174973451293, "learning_rate": 6.7525275088303654e-06, "loss": 0.4272, "step": 5705 }, { "epoch": 0.4045373980857852, "grad_norm": 6.508491778048779, "learning_rate": 6.751452142069366e-06, "loss": 0.4019, "step": 5706 }, { "epoch": 0.4046082949308756, "grad_norm": 3.294875555034441, "learning_rate": 6.750376682952956e-06, "loss": 0.4496, "step": 5707 }, { "epoch": 0.40467919177596595, "grad_norm": 2.891318172594753, "learning_rate": 6.749301131537848e-06, "loss": 0.4405, "step": 5708 }, { "epoch": 0.4047500886210564, "grad_norm": 2.9748086671529013, "learning_rate": 6.748225487880757e-06, "loss": 0.4118, "step": 5709 }, { "epoch": 0.40482098546614675, "grad_norm": 5.129739099487142, "learning_rate": 6.7471497520384e-06, "loss": 0.4473, "step": 5710 }, { "epoch": 0.40489188231123713, "grad_norm": 3.072799471712157, "learning_rate": 6.7460739240675055e-06, "loss": 0.4047, "step": 5711 }, { "epoch": 0.40496277915632756, "grad_norm": 6.741621863485896, "learning_rate": 6.7449980040247975e-06, "loss": 0.4317, "step": 5712 }, { "epoch": 0.40503367600141793, "grad_norm": 1.363462016776575, "learning_rate": 6.743921991967015e-06, "loss": 0.7013, "step": 5713 }, { "epoch": 0.4051045728465083, "grad_norm": 2.319621830602907, "learning_rate": 6.742845887950894e-06, "loss": 0.386, "step": 5714 }, { "epoch": 0.40517546969159873, "grad_norm": 5.716544587334313, "learning_rate": 6.74176969203318e-06, "loss": 0.3859, "step": 5715 }, { "epoch": 0.4052463665366891, "grad_norm": 2.782528793979443, "learning_rate": 6.740693404270619e-06, "loss": 0.3842, "step": 5716 }, { "epoch": 0.40531726338177954, "grad_norm": 2.7340519374794003, "learning_rate": 6.7396170247199676e-06, "loss": 0.4297, "step": 5717 }, { "epoch": 0.4053881602268699, "grad_norm": 7.571745804013218, "learning_rate": 6.738540553437981e-06, "loss": 0.4088, "step": 5718 }, { "epoch": 0.4054590570719603, "grad_norm": 3.4878745995104588, "learning_rate": 6.737463990481427e-06, "loss": 0.4692, "step": 5719 }, { "epoch": 0.4055299539170507, "grad_norm": 2.17990596684057, "learning_rate": 6.736387335907069e-06, "loss": 0.3834, "step": 5720 }, { "epoch": 0.4056008507621411, "grad_norm": 15.555371119963317, "learning_rate": 6.73531058977168e-06, "loss": 0.4273, "step": 5721 }, { "epoch": 0.40567174760723146, "grad_norm": 2.6806272420759343, "learning_rate": 6.73423375213204e-06, "loss": 0.4007, "step": 5722 }, { "epoch": 0.4057426444523219, "grad_norm": 3.8113420388123584, "learning_rate": 6.733156823044932e-06, "loss": 0.4237, "step": 5723 }, { "epoch": 0.40581354129741226, "grad_norm": 5.132084802911918, "learning_rate": 6.732079802567141e-06, "loss": 0.37, "step": 5724 }, { "epoch": 0.40588443814250263, "grad_norm": 13.071400148095451, "learning_rate": 6.731002690755459e-06, "loss": 0.3938, "step": 5725 }, { "epoch": 0.40595533498759306, "grad_norm": 4.266856059659535, "learning_rate": 6.729925487666686e-06, "loss": 0.3914, "step": 5726 }, { "epoch": 0.40602623183268344, "grad_norm": 2.8090533206162194, "learning_rate": 6.72884819335762e-06, "loss": 0.425, "step": 5727 }, { "epoch": 0.40609712867777387, "grad_norm": 2.3703301812347455, "learning_rate": 6.727770807885069e-06, "loss": 0.375, "step": 5728 }, { "epoch": 0.40616802552286424, "grad_norm": 2.6366113297560823, "learning_rate": 6.726693331305845e-06, "loss": 0.4419, "step": 5729 }, { "epoch": 0.4062389223679546, "grad_norm": 1.8086079865186429, "learning_rate": 6.725615763676765e-06, "loss": 0.3747, "step": 5730 }, { "epoch": 0.40630981921304504, "grad_norm": 2.2201627185169834, "learning_rate": 6.724538105054646e-06, "loss": 0.4254, "step": 5731 }, { "epoch": 0.4063807160581354, "grad_norm": 4.755251822876024, "learning_rate": 6.7234603554963185e-06, "loss": 0.4133, "step": 5732 }, { "epoch": 0.4064516129032258, "grad_norm": 1.8214671038954615, "learning_rate": 6.72238251505861e-06, "loss": 0.4112, "step": 5733 }, { "epoch": 0.4065225097483162, "grad_norm": 3.337015159912737, "learning_rate": 6.721304583798358e-06, "loss": 0.4057, "step": 5734 }, { "epoch": 0.4065934065934066, "grad_norm": 3.4496920149448944, "learning_rate": 6.720226561772401e-06, "loss": 0.4166, "step": 5735 }, { "epoch": 0.40666430343849697, "grad_norm": 5.7227767050863845, "learning_rate": 6.719148449037583e-06, "loss": 0.4173, "step": 5736 }, { "epoch": 0.4067352002835874, "grad_norm": 3.149472328451312, "learning_rate": 6.718070245650757e-06, "loss": 0.4277, "step": 5737 }, { "epoch": 0.40680609712867777, "grad_norm": 5.490036551523552, "learning_rate": 6.716991951668775e-06, "loss": 0.4692, "step": 5738 }, { "epoch": 0.40687699397376814, "grad_norm": 2.1614378050810887, "learning_rate": 6.715913567148497e-06, "loss": 0.4082, "step": 5739 }, { "epoch": 0.40694789081885857, "grad_norm": 3.691374195824579, "learning_rate": 6.714835092146786e-06, "loss": 0.4534, "step": 5740 }, { "epoch": 0.40701878766394894, "grad_norm": 1.13306244898437, "learning_rate": 6.7137565267205125e-06, "loss": 0.7179, "step": 5741 }, { "epoch": 0.4070896845090394, "grad_norm": 1.8253807207025856, "learning_rate": 6.7126778709265495e-06, "loss": 0.4305, "step": 5742 }, { "epoch": 0.40716058135412975, "grad_norm": 3.389263883843445, "learning_rate": 6.711599124821775e-06, "loss": 0.4179, "step": 5743 }, { "epoch": 0.4072314781992201, "grad_norm": 1.9833709833125222, "learning_rate": 6.7105202884630715e-06, "loss": 0.4059, "step": 5744 }, { "epoch": 0.40730237504431055, "grad_norm": 2.4871984368009694, "learning_rate": 6.709441361907328e-06, "loss": 0.4295, "step": 5745 }, { "epoch": 0.4073732718894009, "grad_norm": 3.125768006905109, "learning_rate": 6.708362345211437e-06, "loss": 0.3947, "step": 5746 }, { "epoch": 0.4074441687344913, "grad_norm": 5.2552515905003085, "learning_rate": 6.7072832384322965e-06, "loss": 0.4824, "step": 5747 }, { "epoch": 0.4075150655795817, "grad_norm": 3.9884530147700965, "learning_rate": 6.706204041626806e-06, "loss": 0.375, "step": 5748 }, { "epoch": 0.4075859624246721, "grad_norm": 3.564628357382889, "learning_rate": 6.705124754851875e-06, "loss": 0.3985, "step": 5749 }, { "epoch": 0.4076568592697625, "grad_norm": 2.7572389630360266, "learning_rate": 6.704045378164413e-06, "loss": 0.4004, "step": 5750 }, { "epoch": 0.4077277561148529, "grad_norm": 3.2550908427047016, "learning_rate": 6.70296591162134e-06, "loss": 0.4377, "step": 5751 }, { "epoch": 0.4077986529599433, "grad_norm": 2.2575008851947858, "learning_rate": 6.701886355279573e-06, "loss": 0.464, "step": 5752 }, { "epoch": 0.40786954980503365, "grad_norm": 2.2558502187711142, "learning_rate": 6.700806709196039e-06, "loss": 0.4948, "step": 5753 }, { "epoch": 0.4079404466501241, "grad_norm": 2.8183681076891247, "learning_rate": 6.69972697342767e-06, "loss": 0.375, "step": 5754 }, { "epoch": 0.40801134349521445, "grad_norm": 4.184740337131072, "learning_rate": 6.6986471480314e-06, "loss": 0.4653, "step": 5755 }, { "epoch": 0.4080822403403049, "grad_norm": 3.5107079142632336, "learning_rate": 6.6975672330641705e-06, "loss": 0.4015, "step": 5756 }, { "epoch": 0.40815313718539525, "grad_norm": 3.3532153192534473, "learning_rate": 6.696487228582923e-06, "loss": 0.3977, "step": 5757 }, { "epoch": 0.4082240340304856, "grad_norm": 1.1524275870253844, "learning_rate": 6.695407134644612e-06, "loss": 0.668, "step": 5758 }, { "epoch": 0.40829493087557606, "grad_norm": 3.182282507508217, "learning_rate": 6.694326951306186e-06, "loss": 0.357, "step": 5759 }, { "epoch": 0.40836582772066643, "grad_norm": 2.584579827048674, "learning_rate": 6.693246678624608e-06, "loss": 0.4373, "step": 5760 }, { "epoch": 0.4084367245657568, "grad_norm": 3.911131450187629, "learning_rate": 6.69216631665684e-06, "loss": 0.513, "step": 5761 }, { "epoch": 0.40850762141084723, "grad_norm": 4.449109469240552, "learning_rate": 6.69108586545985e-06, "loss": 0.4343, "step": 5762 }, { "epoch": 0.4085785182559376, "grad_norm": 2.9051318475445402, "learning_rate": 6.690005325090611e-06, "loss": 0.397, "step": 5763 }, { "epoch": 0.408649415101028, "grad_norm": 2.4604167466808056, "learning_rate": 6.688924695606104e-06, "loss": 0.4231, "step": 5764 }, { "epoch": 0.4087203119461184, "grad_norm": 2.0311890330926876, "learning_rate": 6.687843977063305e-06, "loss": 0.3954, "step": 5765 }, { "epoch": 0.4087912087912088, "grad_norm": 5.723801823292545, "learning_rate": 6.686763169519208e-06, "loss": 0.4435, "step": 5766 }, { "epoch": 0.4088621056362992, "grad_norm": 4.872279578094734, "learning_rate": 6.6856822730308e-06, "loss": 0.4048, "step": 5767 }, { "epoch": 0.4089330024813896, "grad_norm": 3.3207304865181286, "learning_rate": 6.684601287655079e-06, "loss": 0.4253, "step": 5768 }, { "epoch": 0.40900389932647996, "grad_norm": 2.427173211833054, "learning_rate": 6.683520213449049e-06, "loss": 0.4181, "step": 5769 }, { "epoch": 0.4090747961715704, "grad_norm": 3.6982667219682823, "learning_rate": 6.682439050469709e-06, "loss": 0.4225, "step": 5770 }, { "epoch": 0.40914569301666076, "grad_norm": 3.7435737148886408, "learning_rate": 6.681357798774076e-06, "loss": 0.424, "step": 5771 }, { "epoch": 0.40921658986175113, "grad_norm": 2.8848319579741957, "learning_rate": 6.680276458419163e-06, "loss": 0.4443, "step": 5772 }, { "epoch": 0.40928748670684156, "grad_norm": 1.1367976275621392, "learning_rate": 6.679195029461989e-06, "loss": 0.673, "step": 5773 }, { "epoch": 0.40935838355193194, "grad_norm": 4.074716242422497, "learning_rate": 6.678113511959581e-06, "loss": 0.4193, "step": 5774 }, { "epoch": 0.4094292803970223, "grad_norm": 2.905470032666464, "learning_rate": 6.677031905968966e-06, "loss": 0.4062, "step": 5775 }, { "epoch": 0.40950017724211274, "grad_norm": 1.3691159041772072, "learning_rate": 6.6759502115471765e-06, "loss": 0.7238, "step": 5776 }, { "epoch": 0.4095710740872031, "grad_norm": 3.482998263722296, "learning_rate": 6.674868428751255e-06, "loss": 0.4291, "step": 5777 }, { "epoch": 0.4096419709322935, "grad_norm": 4.264344854717868, "learning_rate": 6.6737865576382425e-06, "loss": 0.423, "step": 5778 }, { "epoch": 0.4097128677773839, "grad_norm": 4.708680527965177, "learning_rate": 6.672704598265189e-06, "loss": 0.4086, "step": 5779 }, { "epoch": 0.4097837646224743, "grad_norm": 3.1315428436509807, "learning_rate": 6.671622550689143e-06, "loss": 0.4195, "step": 5780 }, { "epoch": 0.4098546614675647, "grad_norm": 3.918803562561408, "learning_rate": 6.670540414967165e-06, "loss": 0.4023, "step": 5781 }, { "epoch": 0.4099255583126551, "grad_norm": 2.513797686919288, "learning_rate": 6.669458191156315e-06, "loss": 0.4173, "step": 5782 }, { "epoch": 0.40999645515774547, "grad_norm": 2.4883397031772625, "learning_rate": 6.668375879313662e-06, "loss": 0.4217, "step": 5783 }, { "epoch": 0.4100673520028359, "grad_norm": 2.395509143481896, "learning_rate": 6.667293479496274e-06, "loss": 0.3903, "step": 5784 }, { "epoch": 0.41013824884792627, "grad_norm": 8.258842642027687, "learning_rate": 6.666210991761228e-06, "loss": 0.4574, "step": 5785 }, { "epoch": 0.41020914569301664, "grad_norm": 8.136886639275565, "learning_rate": 6.665128416165606e-06, "loss": 0.451, "step": 5786 }, { "epoch": 0.41028004253810707, "grad_norm": 3.9294168766625446, "learning_rate": 6.66404575276649e-06, "loss": 0.4453, "step": 5787 }, { "epoch": 0.41035093938319744, "grad_norm": 2.6302058222405105, "learning_rate": 6.662963001620975e-06, "loss": 0.4246, "step": 5788 }, { "epoch": 0.4104218362282878, "grad_norm": 4.128268070701896, "learning_rate": 6.661880162786148e-06, "loss": 0.3893, "step": 5789 }, { "epoch": 0.41049273307337825, "grad_norm": 2.5972574193478115, "learning_rate": 6.660797236319112e-06, "loss": 0.4329, "step": 5790 }, { "epoch": 0.4105636299184686, "grad_norm": 2.3983899069879464, "learning_rate": 6.6597142222769705e-06, "loss": 0.4222, "step": 5791 }, { "epoch": 0.41063452676355905, "grad_norm": 2.812263678074997, "learning_rate": 6.658631120716832e-06, "loss": 0.4366, "step": 5792 }, { "epoch": 0.4107054236086494, "grad_norm": 2.2588897428282704, "learning_rate": 6.657547931695807e-06, "loss": 0.3953, "step": 5793 }, { "epoch": 0.4107763204537398, "grad_norm": 3.1260698080228253, "learning_rate": 6.656464655271016e-06, "loss": 0.3952, "step": 5794 }, { "epoch": 0.4108472172988302, "grad_norm": 2.6913960748916406, "learning_rate": 6.655381291499578e-06, "loss": 0.385, "step": 5795 }, { "epoch": 0.4109181141439206, "grad_norm": 3.8683521863357884, "learning_rate": 6.654297840438623e-06, "loss": 0.485, "step": 5796 }, { "epoch": 0.41098901098901097, "grad_norm": 2.4864133382325715, "learning_rate": 6.6532143021452785e-06, "loss": 0.4285, "step": 5797 }, { "epoch": 0.4110599078341014, "grad_norm": 2.1163580180439023, "learning_rate": 6.652130676676682e-06, "loss": 0.4204, "step": 5798 }, { "epoch": 0.4111308046791918, "grad_norm": 2.589757756665158, "learning_rate": 6.651046964089975e-06, "loss": 0.369, "step": 5799 }, { "epoch": 0.41120170152428215, "grad_norm": 2.604685454824944, "learning_rate": 6.649963164442302e-06, "loss": 0.3879, "step": 5800 }, { "epoch": 0.4112725983693726, "grad_norm": 7.569365357188654, "learning_rate": 6.64887927779081e-06, "loss": 0.4211, "step": 5801 }, { "epoch": 0.41134349521446295, "grad_norm": 5.074866699129863, "learning_rate": 6.647795304192654e-06, "loss": 0.3697, "step": 5802 }, { "epoch": 0.4114143920595533, "grad_norm": 2.2350364202401165, "learning_rate": 6.646711243704997e-06, "loss": 0.4332, "step": 5803 }, { "epoch": 0.41148528890464375, "grad_norm": 3.032183424227302, "learning_rate": 6.645627096384999e-06, "loss": 0.4479, "step": 5804 }, { "epoch": 0.4115561857497341, "grad_norm": 2.0696903964834443, "learning_rate": 6.644542862289827e-06, "loss": 0.4261, "step": 5805 }, { "epoch": 0.41162708259482456, "grad_norm": 2.7384175555351122, "learning_rate": 6.643458541476655e-06, "loss": 0.3935, "step": 5806 }, { "epoch": 0.41169797943991493, "grad_norm": 3.288468177001182, "learning_rate": 6.642374134002659e-06, "loss": 0.41, "step": 5807 }, { "epoch": 0.4117688762850053, "grad_norm": 2.4214862873943526, "learning_rate": 6.6412896399250214e-06, "loss": 0.4074, "step": 5808 }, { "epoch": 0.41183977313009573, "grad_norm": 3.460871856015357, "learning_rate": 6.640205059300931e-06, "loss": 0.3862, "step": 5809 }, { "epoch": 0.4119106699751861, "grad_norm": 5.73555760959815, "learning_rate": 6.639120392187572e-06, "loss": 0.4056, "step": 5810 }, { "epoch": 0.4119815668202765, "grad_norm": 2.3632932910965403, "learning_rate": 6.638035638642146e-06, "loss": 0.387, "step": 5811 }, { "epoch": 0.4120524636653669, "grad_norm": 2.3225540229780326, "learning_rate": 6.63695079872185e-06, "loss": 0.3763, "step": 5812 }, { "epoch": 0.4121233605104573, "grad_norm": 2.718175705318534, "learning_rate": 6.635865872483889e-06, "loss": 0.4698, "step": 5813 }, { "epoch": 0.41219425735554766, "grad_norm": 2.5003098696029915, "learning_rate": 6.634780859985471e-06, "loss": 0.4121, "step": 5814 }, { "epoch": 0.4122651542006381, "grad_norm": 5.284087822218048, "learning_rate": 6.633695761283812e-06, "loss": 0.4634, "step": 5815 }, { "epoch": 0.41233605104572846, "grad_norm": 1.7262952351328293, "learning_rate": 6.632610576436128e-06, "loss": 0.3785, "step": 5816 }, { "epoch": 0.41240694789081883, "grad_norm": 5.242529086616619, "learning_rate": 6.631525305499642e-06, "loss": 0.4359, "step": 5817 }, { "epoch": 0.41247784473590926, "grad_norm": 2.88893022975295, "learning_rate": 6.6304399485315815e-06, "loss": 0.3791, "step": 5818 }, { "epoch": 0.41254874158099963, "grad_norm": 2.5506955593838874, "learning_rate": 6.629354505589178e-06, "loss": 0.3924, "step": 5819 }, { "epoch": 0.41261963842609006, "grad_norm": 2.6852876128372567, "learning_rate": 6.628268976729668e-06, "loss": 0.3978, "step": 5820 }, { "epoch": 0.41269053527118044, "grad_norm": 2.7352495453893275, "learning_rate": 6.627183362010291e-06, "loss": 0.4484, "step": 5821 }, { "epoch": 0.4127614321162708, "grad_norm": 4.564914769417034, "learning_rate": 6.626097661488296e-06, "loss": 0.4244, "step": 5822 }, { "epoch": 0.41283232896136124, "grad_norm": 3.5912890643263538, "learning_rate": 6.625011875220927e-06, "loss": 0.4288, "step": 5823 }, { "epoch": 0.4129032258064516, "grad_norm": 2.3617214387539547, "learning_rate": 6.623926003265444e-06, "loss": 0.4123, "step": 5824 }, { "epoch": 0.412974122651542, "grad_norm": 3.3831081004364867, "learning_rate": 6.622840045679102e-06, "loss": 0.411, "step": 5825 }, { "epoch": 0.4130450194966324, "grad_norm": 2.7169823557249533, "learning_rate": 6.621754002519165e-06, "loss": 0.4308, "step": 5826 }, { "epoch": 0.4131159163417228, "grad_norm": 2.389668330626286, "learning_rate": 6.6206678738429035e-06, "loss": 0.3882, "step": 5827 }, { "epoch": 0.41318681318681316, "grad_norm": 2.9947382455092675, "learning_rate": 6.6195816597075865e-06, "loss": 0.4623, "step": 5828 }, { "epoch": 0.4132577100319036, "grad_norm": 2.7996423248486515, "learning_rate": 6.618495360170494e-06, "loss": 0.4598, "step": 5829 }, { "epoch": 0.41332860687699396, "grad_norm": 8.952453022791513, "learning_rate": 6.617408975288904e-06, "loss": 0.3924, "step": 5830 }, { "epoch": 0.4133995037220844, "grad_norm": 2.5300717284102174, "learning_rate": 6.616322505120105e-06, "loss": 0.4009, "step": 5831 }, { "epoch": 0.41347040056717477, "grad_norm": 2.800389015873425, "learning_rate": 6.615235949721387e-06, "loss": 0.3642, "step": 5832 }, { "epoch": 0.41354129741226514, "grad_norm": 3.309821693323526, "learning_rate": 6.614149309150044e-06, "loss": 0.3862, "step": 5833 }, { "epoch": 0.41361219425735557, "grad_norm": 1.7985933664011016, "learning_rate": 6.613062583463376e-06, "loss": 0.4119, "step": 5834 }, { "epoch": 0.41368309110244594, "grad_norm": 2.3864704235314282, "learning_rate": 6.611975772718687e-06, "loss": 0.3873, "step": 5835 }, { "epoch": 0.4137539879475363, "grad_norm": 1.9108496763934155, "learning_rate": 6.610888876973284e-06, "loss": 0.4221, "step": 5836 }, { "epoch": 0.41382488479262675, "grad_norm": 2.694523544650799, "learning_rate": 6.6098018962844835e-06, "loss": 0.4136, "step": 5837 }, { "epoch": 0.4138957816377171, "grad_norm": 5.50486382715327, "learning_rate": 6.608714830709598e-06, "loss": 0.3645, "step": 5838 }, { "epoch": 0.4139666784828075, "grad_norm": 2.7326092124876777, "learning_rate": 6.607627680305952e-06, "loss": 0.439, "step": 5839 }, { "epoch": 0.4140375753278979, "grad_norm": 2.895401578367038, "learning_rate": 6.606540445130872e-06, "loss": 0.4021, "step": 5840 }, { "epoch": 0.4141084721729883, "grad_norm": 6.842587418035003, "learning_rate": 6.605453125241689e-06, "loss": 0.4025, "step": 5841 }, { "epoch": 0.41417936901807867, "grad_norm": 6.182336253283411, "learning_rate": 6.604365720695734e-06, "loss": 0.4274, "step": 5842 }, { "epoch": 0.4142502658631691, "grad_norm": 3.598203464199621, "learning_rate": 6.603278231550354e-06, "loss": 0.4747, "step": 5843 }, { "epoch": 0.41432116270825947, "grad_norm": 3.181302173826043, "learning_rate": 6.602190657862887e-06, "loss": 0.4945, "step": 5844 }, { "epoch": 0.4143920595533499, "grad_norm": 3.6165801496077097, "learning_rate": 6.601102999690684e-06, "loss": 0.4091, "step": 5845 }, { "epoch": 0.4144629563984403, "grad_norm": 2.7328507151669603, "learning_rate": 6.600015257091097e-06, "loss": 0.428, "step": 5846 }, { "epoch": 0.41453385324353065, "grad_norm": 2.3599038278634663, "learning_rate": 6.598927430121486e-06, "loss": 0.4397, "step": 5847 }, { "epoch": 0.4146047500886211, "grad_norm": 3.4156357600018246, "learning_rate": 6.597839518839211e-06, "loss": 0.4115, "step": 5848 }, { "epoch": 0.41467564693371145, "grad_norm": 2.639828975693704, "learning_rate": 6.596751523301637e-06, "loss": 0.428, "step": 5849 }, { "epoch": 0.4147465437788018, "grad_norm": 2.9675051268021937, "learning_rate": 6.595663443566137e-06, "loss": 0.43, "step": 5850 }, { "epoch": 0.41481744062389225, "grad_norm": 2.006990472993884, "learning_rate": 6.594575279690086e-06, "loss": 0.4114, "step": 5851 }, { "epoch": 0.4148883374689826, "grad_norm": 3.293505607291973, "learning_rate": 6.593487031730863e-06, "loss": 0.4283, "step": 5852 }, { "epoch": 0.414959234314073, "grad_norm": 3.402860855718075, "learning_rate": 6.592398699745853e-06, "loss": 0.3458, "step": 5853 }, { "epoch": 0.41503013115916343, "grad_norm": 3.7599068293496254, "learning_rate": 6.591310283792445e-06, "loss": 0.4468, "step": 5854 }, { "epoch": 0.4151010280042538, "grad_norm": 7.651012469304478, "learning_rate": 6.590221783928029e-06, "loss": 0.4449, "step": 5855 }, { "epoch": 0.41517192484934423, "grad_norm": 3.119295542026712, "learning_rate": 6.589133200210007e-06, "loss": 0.4054, "step": 5856 }, { "epoch": 0.4152428216944346, "grad_norm": 5.0535572141638925, "learning_rate": 6.588044532695776e-06, "loss": 0.4438, "step": 5857 }, { "epoch": 0.415313718539525, "grad_norm": 2.669215595207985, "learning_rate": 6.586955781442747e-06, "loss": 0.4532, "step": 5858 }, { "epoch": 0.4153846153846154, "grad_norm": 3.840915178192567, "learning_rate": 6.5858669465083266e-06, "loss": 0.4421, "step": 5859 }, { "epoch": 0.4154555122297058, "grad_norm": 5.5352727435935565, "learning_rate": 6.584778027949933e-06, "loss": 0.4148, "step": 5860 }, { "epoch": 0.41552640907479615, "grad_norm": 2.8326557563643706, "learning_rate": 6.583689025824984e-06, "loss": 0.4175, "step": 5861 }, { "epoch": 0.4155973059198866, "grad_norm": 2.388492189324524, "learning_rate": 6.582599940190904e-06, "loss": 0.3845, "step": 5862 }, { "epoch": 0.41566820276497696, "grad_norm": 4.443158328592987, "learning_rate": 6.581510771105121e-06, "loss": 0.4071, "step": 5863 }, { "epoch": 0.41573909961006733, "grad_norm": 1.9786691622558723, "learning_rate": 6.580421518625067e-06, "loss": 0.4272, "step": 5864 }, { "epoch": 0.41580999645515776, "grad_norm": 8.261761933112851, "learning_rate": 6.57933218280818e-06, "loss": 0.421, "step": 5865 }, { "epoch": 0.41588089330024813, "grad_norm": 1.7275947323600407, "learning_rate": 6.578242763711903e-06, "loss": 0.4169, "step": 5866 }, { "epoch": 0.4159517901453385, "grad_norm": 2.352328304957264, "learning_rate": 6.577153261393681e-06, "loss": 0.4316, "step": 5867 }, { "epoch": 0.41602268699042894, "grad_norm": 2.4068055556958514, "learning_rate": 6.576063675910961e-06, "loss": 0.3988, "step": 5868 }, { "epoch": 0.4160935838355193, "grad_norm": 3.4088714911611735, "learning_rate": 6.574974007321203e-06, "loss": 0.4287, "step": 5869 }, { "epoch": 0.41616448068060974, "grad_norm": 4.647500951715144, "learning_rate": 6.573884255681862e-06, "loss": 0.4396, "step": 5870 }, { "epoch": 0.4162353775257001, "grad_norm": 3.076265081155189, "learning_rate": 6.5727944210504035e-06, "loss": 0.4057, "step": 5871 }, { "epoch": 0.4163062743707905, "grad_norm": 4.59033032736685, "learning_rate": 6.5717045034842945e-06, "loss": 0.4249, "step": 5872 }, { "epoch": 0.4163771712158809, "grad_norm": 4.908332896560718, "learning_rate": 6.5706145030410074e-06, "loss": 0.4307, "step": 5873 }, { "epoch": 0.4164480680609713, "grad_norm": 2.2314692396705484, "learning_rate": 6.5695244197780185e-06, "loss": 0.4128, "step": 5874 }, { "epoch": 0.41651896490606166, "grad_norm": 17.856633962327834, "learning_rate": 6.56843425375281e-06, "loss": 0.3877, "step": 5875 }, { "epoch": 0.4165898617511521, "grad_norm": 3.4101031534104553, "learning_rate": 6.567344005022866e-06, "loss": 0.4618, "step": 5876 }, { "epoch": 0.41666075859624246, "grad_norm": 2.7964931791977965, "learning_rate": 6.566253673645675e-06, "loss": 0.451, "step": 5877 }, { "epoch": 0.41673165544133284, "grad_norm": 4.0056291758370195, "learning_rate": 6.565163259678734e-06, "loss": 0.4425, "step": 5878 }, { "epoch": 0.41680255228642327, "grad_norm": 3.525758122177924, "learning_rate": 6.564072763179539e-06, "loss": 0.4213, "step": 5879 }, { "epoch": 0.41687344913151364, "grad_norm": 4.218301672290926, "learning_rate": 6.562982184205593e-06, "loss": 0.4257, "step": 5880 }, { "epoch": 0.416944345976604, "grad_norm": 2.7257880455450607, "learning_rate": 6.561891522814403e-06, "loss": 0.4642, "step": 5881 }, { "epoch": 0.41701524282169444, "grad_norm": 3.8555848046539185, "learning_rate": 6.560800779063482e-06, "loss": 0.4127, "step": 5882 }, { "epoch": 0.4170861396667848, "grad_norm": 1.9144915802124367, "learning_rate": 6.559709953010344e-06, "loss": 0.419, "step": 5883 }, { "epoch": 0.41715703651187525, "grad_norm": 3.756354792023302, "learning_rate": 6.55861904471251e-06, "loss": 0.4592, "step": 5884 }, { "epoch": 0.4172279333569656, "grad_norm": 3.685326220586681, "learning_rate": 6.5575280542275035e-06, "loss": 0.4055, "step": 5885 }, { "epoch": 0.417298830202056, "grad_norm": 2.895837458725322, "learning_rate": 6.556436981612854e-06, "loss": 0.4063, "step": 5886 }, { "epoch": 0.4173697270471464, "grad_norm": 2.117541948630994, "learning_rate": 6.555345826926095e-06, "loss": 0.4326, "step": 5887 }, { "epoch": 0.4174406238922368, "grad_norm": 5.455921931125559, "learning_rate": 6.554254590224764e-06, "loss": 0.4608, "step": 5888 }, { "epoch": 0.41751152073732717, "grad_norm": 1.969277972814474, "learning_rate": 6.553163271566402e-06, "loss": 0.4151, "step": 5889 }, { "epoch": 0.4175824175824176, "grad_norm": 6.412068965192165, "learning_rate": 6.552071871008554e-06, "loss": 0.4618, "step": 5890 }, { "epoch": 0.41765331442750797, "grad_norm": 5.688178131472987, "learning_rate": 6.550980388608772e-06, "loss": 0.4262, "step": 5891 }, { "epoch": 0.41772421127259834, "grad_norm": 4.347674751804074, "learning_rate": 6.549888824424611e-06, "loss": 0.4094, "step": 5892 }, { "epoch": 0.4177951081176888, "grad_norm": 3.7265517657308154, "learning_rate": 6.548797178513629e-06, "loss": 0.4441, "step": 5893 }, { "epoch": 0.41786600496277915, "grad_norm": 6.597397756450971, "learning_rate": 6.547705450933389e-06, "loss": 0.4717, "step": 5894 }, { "epoch": 0.4179369018078696, "grad_norm": 2.829110064985926, "learning_rate": 6.5466136417414596e-06, "loss": 0.4329, "step": 5895 }, { "epoch": 0.41800779865295995, "grad_norm": 2.767148533804691, "learning_rate": 6.5455217509954125e-06, "loss": 0.4507, "step": 5896 }, { "epoch": 0.4180786954980503, "grad_norm": 3.4370857091715816, "learning_rate": 6.544429778752824e-06, "loss": 0.4073, "step": 5897 }, { "epoch": 0.41814959234314075, "grad_norm": 5.827341605887368, "learning_rate": 6.5433377250712745e-06, "loss": 0.3941, "step": 5898 }, { "epoch": 0.4182204891882311, "grad_norm": 7.632522960219075, "learning_rate": 6.5422455900083494e-06, "loss": 0.4137, "step": 5899 }, { "epoch": 0.4182913860333215, "grad_norm": 3.0455683556953557, "learning_rate": 6.541153373621634e-06, "loss": 0.4229, "step": 5900 }, { "epoch": 0.41836228287841193, "grad_norm": 3.0826020762031217, "learning_rate": 6.5400610759687285e-06, "loss": 0.3901, "step": 5901 }, { "epoch": 0.4184331797235023, "grad_norm": 9.189450604106977, "learning_rate": 6.538968697107225e-06, "loss": 0.4311, "step": 5902 }, { "epoch": 0.4185040765685927, "grad_norm": 1.0835264762442747, "learning_rate": 6.537876237094729e-06, "loss": 0.6814, "step": 5903 }, { "epoch": 0.4185749734136831, "grad_norm": 2.9971316363651312, "learning_rate": 6.5367836959888445e-06, "loss": 0.3784, "step": 5904 }, { "epoch": 0.4186458702587735, "grad_norm": 2.107919180061643, "learning_rate": 6.535691073847182e-06, "loss": 0.3736, "step": 5905 }, { "epoch": 0.41871676710386385, "grad_norm": 2.550395499998072, "learning_rate": 6.53459837072736e-06, "loss": 0.3643, "step": 5906 }, { "epoch": 0.4187876639489543, "grad_norm": 3.871765972752724, "learning_rate": 6.533505586686991e-06, "loss": 0.4221, "step": 5907 }, { "epoch": 0.41885856079404465, "grad_norm": 2.435785749325456, "learning_rate": 6.532412721783703e-06, "loss": 0.4515, "step": 5908 }, { "epoch": 0.4189294576391351, "grad_norm": 2.4760721609961434, "learning_rate": 6.5313197760751236e-06, "loss": 0.4631, "step": 5909 }, { "epoch": 0.41900035448422546, "grad_norm": 5.194868885208579, "learning_rate": 6.530226749618882e-06, "loss": 0.4829, "step": 5910 }, { "epoch": 0.41907125132931583, "grad_norm": 3.002498806087402, "learning_rate": 6.529133642472617e-06, "loss": 0.4828, "step": 5911 }, { "epoch": 0.41914214817440626, "grad_norm": 3.0915201545518802, "learning_rate": 6.5280404546939694e-06, "loss": 0.4286, "step": 5912 }, { "epoch": 0.41921304501949663, "grad_norm": 3.058647505322334, "learning_rate": 6.526947186340578e-06, "loss": 0.4032, "step": 5913 }, { "epoch": 0.419283941864587, "grad_norm": 7.881914237659394, "learning_rate": 6.5258538374701e-06, "loss": 0.3964, "step": 5914 }, { "epoch": 0.41935483870967744, "grad_norm": 1.4202987028932408, "learning_rate": 6.524760408140182e-06, "loss": 0.3816, "step": 5915 }, { "epoch": 0.4194257355547678, "grad_norm": 1.8228671023983885, "learning_rate": 6.523666898408486e-06, "loss": 0.4229, "step": 5916 }, { "epoch": 0.4194966323998582, "grad_norm": 2.370460146938976, "learning_rate": 6.522573308332671e-06, "loss": 0.4324, "step": 5917 }, { "epoch": 0.4195675292449486, "grad_norm": 3.3398979678612366, "learning_rate": 6.5214796379704015e-06, "loss": 0.4275, "step": 5918 }, { "epoch": 0.419638426090039, "grad_norm": 6.561716406501354, "learning_rate": 6.52038588737935e-06, "loss": 0.4475, "step": 5919 }, { "epoch": 0.4197093229351294, "grad_norm": 5.123270032835533, "learning_rate": 6.519292056617193e-06, "loss": 0.427, "step": 5920 }, { "epoch": 0.4197802197802198, "grad_norm": 2.41153335642399, "learning_rate": 6.518198145741604e-06, "loss": 0.409, "step": 5921 }, { "epoch": 0.41985111662531016, "grad_norm": 1.0632856452804726, "learning_rate": 6.517104154810268e-06, "loss": 0.6798, "step": 5922 }, { "epoch": 0.4199220134704006, "grad_norm": 2.4636283740820963, "learning_rate": 6.516010083880871e-06, "loss": 0.4051, "step": 5923 }, { "epoch": 0.41999291031549096, "grad_norm": 3.147558552676651, "learning_rate": 6.514915933011106e-06, "loss": 0.3946, "step": 5924 }, { "epoch": 0.42006380716058134, "grad_norm": 2.345019366245377, "learning_rate": 6.513821702258669e-06, "loss": 0.3803, "step": 5925 }, { "epoch": 0.42013470400567177, "grad_norm": 1.8976382582961402, "learning_rate": 6.512727391681256e-06, "loss": 0.6544, "step": 5926 }, { "epoch": 0.42020560085076214, "grad_norm": 6.493792141513512, "learning_rate": 6.511633001336575e-06, "loss": 0.4137, "step": 5927 }, { "epoch": 0.4202764976958525, "grad_norm": 3.912031500321209, "learning_rate": 6.510538531282331e-06, "loss": 0.3673, "step": 5928 }, { "epoch": 0.42034739454094294, "grad_norm": 3.709213386151362, "learning_rate": 6.509443981576237e-06, "loss": 0.4532, "step": 5929 }, { "epoch": 0.4204182913860333, "grad_norm": 2.7865376480198516, "learning_rate": 6.50834935227601e-06, "loss": 0.3901, "step": 5930 }, { "epoch": 0.4204891882311237, "grad_norm": 2.3040725932293427, "learning_rate": 6.50725464343937e-06, "loss": 0.4171, "step": 5931 }, { "epoch": 0.4205600850762141, "grad_norm": 3.4334428336312506, "learning_rate": 6.506159855124042e-06, "loss": 0.4361, "step": 5932 }, { "epoch": 0.4206309819213045, "grad_norm": 3.2468781148120653, "learning_rate": 6.505064987387757e-06, "loss": 0.4178, "step": 5933 }, { "epoch": 0.4207018787663949, "grad_norm": 5.765598263252258, "learning_rate": 6.503970040288246e-06, "loss": 0.3948, "step": 5934 }, { "epoch": 0.4207727756114853, "grad_norm": 1.6378240161574162, "learning_rate": 6.502875013883245e-06, "loss": 0.3961, "step": 5935 }, { "epoch": 0.42084367245657567, "grad_norm": 2.6174110676010023, "learning_rate": 6.5017799082304985e-06, "loss": 0.4443, "step": 5936 }, { "epoch": 0.4209145693016661, "grad_norm": 3.5242901705552696, "learning_rate": 6.500684723387752e-06, "loss": 0.3717, "step": 5937 }, { "epoch": 0.42098546614675647, "grad_norm": 2.8959359379474776, "learning_rate": 6.499589459412755e-06, "loss": 0.427, "step": 5938 }, { "epoch": 0.42105636299184684, "grad_norm": 2.292381596486996, "learning_rate": 6.4984941163632585e-06, "loss": 0.3777, "step": 5939 }, { "epoch": 0.4211272598369373, "grad_norm": 2.6668330939083957, "learning_rate": 6.497398694297027e-06, "loss": 0.4084, "step": 5940 }, { "epoch": 0.42119815668202765, "grad_norm": 5.422451058372606, "learning_rate": 6.496303193271817e-06, "loss": 0.4175, "step": 5941 }, { "epoch": 0.421269053527118, "grad_norm": 2.7931140766986013, "learning_rate": 6.495207613345398e-06, "loss": 0.4475, "step": 5942 }, { "epoch": 0.42133995037220845, "grad_norm": 2.2079940316143505, "learning_rate": 6.49411195457554e-06, "loss": 0.3659, "step": 5943 }, { "epoch": 0.4214108472172988, "grad_norm": 2.305680333280948, "learning_rate": 6.493016217020021e-06, "loss": 0.4071, "step": 5944 }, { "epoch": 0.4214817440623892, "grad_norm": 7.699137313869878, "learning_rate": 6.491920400736613e-06, "loss": 0.4388, "step": 5945 }, { "epoch": 0.4215526409074796, "grad_norm": 2.8840111517402662, "learning_rate": 6.4908245057831066e-06, "loss": 0.3812, "step": 5946 }, { "epoch": 0.42162353775257, "grad_norm": 2.6842625778436213, "learning_rate": 6.489728532217283e-06, "loss": 0.4238, "step": 5947 }, { "epoch": 0.42169443459766043, "grad_norm": 3.3134894648643054, "learning_rate": 6.4886324800969405e-06, "loss": 0.3764, "step": 5948 }, { "epoch": 0.4217653314427508, "grad_norm": 2.0660754948740525, "learning_rate": 6.487536349479869e-06, "loss": 0.4449, "step": 5949 }, { "epoch": 0.4218362282878412, "grad_norm": 3.1123299844732175, "learning_rate": 6.486440140423871e-06, "loss": 0.36, "step": 5950 }, { "epoch": 0.4219071251329316, "grad_norm": 2.7229142934049144, "learning_rate": 6.48534385298675e-06, "loss": 0.4221, "step": 5951 }, { "epoch": 0.421978021978022, "grad_norm": 1.976138015771006, "learning_rate": 6.484247487226314e-06, "loss": 0.4392, "step": 5952 }, { "epoch": 0.42204891882311235, "grad_norm": 3.6265360561901088, "learning_rate": 6.483151043200375e-06, "loss": 0.4405, "step": 5953 }, { "epoch": 0.4221198156682028, "grad_norm": 2.488818324246205, "learning_rate": 6.482054520966749e-06, "loss": 0.3885, "step": 5954 }, { "epoch": 0.42219071251329315, "grad_norm": 4.047729020549644, "learning_rate": 6.480957920583258e-06, "loss": 0.4376, "step": 5955 }, { "epoch": 0.4222616093583835, "grad_norm": 4.337955538166807, "learning_rate": 6.479861242107723e-06, "loss": 0.402, "step": 5956 }, { "epoch": 0.42233250620347396, "grad_norm": 12.309418687903907, "learning_rate": 6.4787644855979785e-06, "loss": 0.4348, "step": 5957 }, { "epoch": 0.42240340304856433, "grad_norm": 3.603060969643288, "learning_rate": 6.477667651111851e-06, "loss": 0.3961, "step": 5958 }, { "epoch": 0.42247429989365476, "grad_norm": 3.579632422742303, "learning_rate": 6.476570738707182e-06, "loss": 0.4696, "step": 5959 }, { "epoch": 0.42254519673874513, "grad_norm": 3.39397291750655, "learning_rate": 6.475473748441811e-06, "loss": 0.4085, "step": 5960 }, { "epoch": 0.4226160935838355, "grad_norm": 5.279887342517349, "learning_rate": 6.474376680373582e-06, "loss": 0.4086, "step": 5961 }, { "epoch": 0.42268699042892593, "grad_norm": 3.652296099207983, "learning_rate": 6.473279534560346e-06, "loss": 0.4048, "step": 5962 }, { "epoch": 0.4227578872740163, "grad_norm": 2.2160088068071158, "learning_rate": 6.472182311059956e-06, "loss": 0.4393, "step": 5963 }, { "epoch": 0.4228287841191067, "grad_norm": 2.408913335480993, "learning_rate": 6.471085009930268e-06, "loss": 0.3764, "step": 5964 }, { "epoch": 0.4228996809641971, "grad_norm": 2.1592193575451315, "learning_rate": 6.4699876312291466e-06, "loss": 0.4565, "step": 5965 }, { "epoch": 0.4229705778092875, "grad_norm": 11.735038850623544, "learning_rate": 6.468890175014454e-06, "loss": 0.4462, "step": 5966 }, { "epoch": 0.42304147465437786, "grad_norm": 2.770463811080144, "learning_rate": 6.467792641344061e-06, "loss": 0.4383, "step": 5967 }, { "epoch": 0.4231123714994683, "grad_norm": 3.142444281782224, "learning_rate": 6.4666950302758425e-06, "loss": 0.3408, "step": 5968 }, { "epoch": 0.42318326834455866, "grad_norm": 4.785102315853053, "learning_rate": 6.465597341867675e-06, "loss": 0.3799, "step": 5969 }, { "epoch": 0.42325416518964903, "grad_norm": 5.262163144313121, "learning_rate": 6.464499576177442e-06, "loss": 0.4399, "step": 5970 }, { "epoch": 0.42332506203473946, "grad_norm": 2.530330847424012, "learning_rate": 6.463401733263026e-06, "loss": 0.386, "step": 5971 }, { "epoch": 0.42339595887982984, "grad_norm": 2.0519595910984023, "learning_rate": 6.462303813182323e-06, "loss": 0.3585, "step": 5972 }, { "epoch": 0.42346685572492027, "grad_norm": 2.4750914606614614, "learning_rate": 6.461205815993222e-06, "loss": 0.4732, "step": 5973 }, { "epoch": 0.42353775257001064, "grad_norm": 2.37883877996801, "learning_rate": 6.460107741753623e-06, "loss": 0.4759, "step": 5974 }, { "epoch": 0.423608649415101, "grad_norm": 3.0640325732483933, "learning_rate": 6.459009590521429e-06, "loss": 0.4665, "step": 5975 }, { "epoch": 0.42367954626019144, "grad_norm": 2.3936564145182193, "learning_rate": 6.457911362354545e-06, "loss": 0.4521, "step": 5976 }, { "epoch": 0.4237504431052818, "grad_norm": 2.8461103099238083, "learning_rate": 6.456813057310882e-06, "loss": 0.4258, "step": 5977 }, { "epoch": 0.4238213399503722, "grad_norm": 2.080559618035561, "learning_rate": 6.4557146754483556e-06, "loss": 0.3547, "step": 5978 }, { "epoch": 0.4238922367954626, "grad_norm": 2.44440904630067, "learning_rate": 6.454616216824882e-06, "loss": 0.4456, "step": 5979 }, { "epoch": 0.423963133640553, "grad_norm": 4.86818037673053, "learning_rate": 6.453517681498387e-06, "loss": 0.4718, "step": 5980 }, { "epoch": 0.42403403048564337, "grad_norm": 2.4561952695085147, "learning_rate": 6.452419069526793e-06, "loss": 0.438, "step": 5981 }, { "epoch": 0.4241049273307338, "grad_norm": 3.55321925211313, "learning_rate": 6.4513203809680335e-06, "loss": 0.3996, "step": 5982 }, { "epoch": 0.42417582417582417, "grad_norm": 3.7179976631709883, "learning_rate": 6.450221615880043e-06, "loss": 0.418, "step": 5983 }, { "epoch": 0.4242467210209146, "grad_norm": 2.2827534904278983, "learning_rate": 6.44912277432076e-06, "loss": 0.4241, "step": 5984 }, { "epoch": 0.42431761786600497, "grad_norm": 2.7817469101088865, "learning_rate": 6.448023856348126e-06, "loss": 0.4391, "step": 5985 }, { "epoch": 0.42438851471109534, "grad_norm": 2.208928195463799, "learning_rate": 6.44692486202009e-06, "loss": 0.3984, "step": 5986 }, { "epoch": 0.4244594115561858, "grad_norm": 1.88138574990725, "learning_rate": 6.445825791394602e-06, "loss": 0.3952, "step": 5987 }, { "epoch": 0.42453030840127615, "grad_norm": 2.7166878645570285, "learning_rate": 6.444726644529615e-06, "loss": 0.4159, "step": 5988 }, { "epoch": 0.4246012052463665, "grad_norm": 3.7219729747121497, "learning_rate": 6.443627421483091e-06, "loss": 0.5031, "step": 5989 }, { "epoch": 0.42467210209145695, "grad_norm": 2.389838093954661, "learning_rate": 6.442528122312989e-06, "loss": 0.4109, "step": 5990 }, { "epoch": 0.4247429989365473, "grad_norm": 2.7377167653751036, "learning_rate": 6.44142874707728e-06, "loss": 0.3942, "step": 5991 }, { "epoch": 0.4248138957816377, "grad_norm": 3.200466298220261, "learning_rate": 6.44032929583393e-06, "loss": 0.4233, "step": 5992 }, { "epoch": 0.4248847926267281, "grad_norm": 2.640174764748553, "learning_rate": 6.43922976864092e-06, "loss": 0.4526, "step": 5993 }, { "epoch": 0.4249556894718185, "grad_norm": 2.8923196191788088, "learning_rate": 6.438130165556226e-06, "loss": 0.4418, "step": 5994 }, { "epoch": 0.42502658631690887, "grad_norm": 2.4710582081383743, "learning_rate": 6.437030486637829e-06, "loss": 0.4489, "step": 5995 }, { "epoch": 0.4250974831619993, "grad_norm": 1.482179240516291, "learning_rate": 6.435930731943718e-06, "loss": 0.4384, "step": 5996 }, { "epoch": 0.4251683800070897, "grad_norm": 2.381037826147955, "learning_rate": 6.434830901531884e-06, "loss": 0.427, "step": 5997 }, { "epoch": 0.4252392768521801, "grad_norm": 5.652997411114793, "learning_rate": 6.433730995460321e-06, "loss": 0.416, "step": 5998 }, { "epoch": 0.4253101736972705, "grad_norm": 2.6401062354543505, "learning_rate": 6.432631013787028e-06, "loss": 0.4892, "step": 5999 }, { "epoch": 0.42538107054236085, "grad_norm": 2.086515297321212, "learning_rate": 6.431530956570009e-06, "loss": 0.3788, "step": 6000 }, { "epoch": 0.4254519673874513, "grad_norm": 5.0512674937217135, "learning_rate": 6.4304308238672685e-06, "loss": 0.3889, "step": 6001 }, { "epoch": 0.42552286423254165, "grad_norm": 6.057882292181343, "learning_rate": 6.4293306157368215e-06, "loss": 0.4866, "step": 6002 }, { "epoch": 0.425593761077632, "grad_norm": 2.2086732895564065, "learning_rate": 6.428230332236678e-06, "loss": 0.4332, "step": 6003 }, { "epoch": 0.42566465792272246, "grad_norm": 1.8652705612355662, "learning_rate": 6.427129973424859e-06, "loss": 0.4166, "step": 6004 }, { "epoch": 0.42573555476781283, "grad_norm": 2.6205348959387345, "learning_rate": 6.426029539359387e-06, "loss": 0.4002, "step": 6005 }, { "epoch": 0.4258064516129032, "grad_norm": 1.1885134046842554, "learning_rate": 6.42492903009829e-06, "loss": 0.6767, "step": 6006 }, { "epoch": 0.42587734845799363, "grad_norm": 1.7591814198748237, "learning_rate": 6.423828445699598e-06, "loss": 0.389, "step": 6007 }, { "epoch": 0.425948245303084, "grad_norm": 5.566354907717736, "learning_rate": 6.422727786221344e-06, "loss": 0.4638, "step": 6008 }, { "epoch": 0.4260191421481744, "grad_norm": 2.914652368857765, "learning_rate": 6.421627051721568e-06, "loss": 0.395, "step": 6009 }, { "epoch": 0.4260900389932648, "grad_norm": 2.910698364675771, "learning_rate": 6.4205262422583144e-06, "loss": 0.3786, "step": 6010 }, { "epoch": 0.4261609358383552, "grad_norm": 2.435583067032225, "learning_rate": 6.419425357889626e-06, "loss": 0.4081, "step": 6011 }, { "epoch": 0.4262318326834456, "grad_norm": 13.713794109321645, "learning_rate": 6.418324398673557e-06, "loss": 0.4402, "step": 6012 }, { "epoch": 0.426302729528536, "grad_norm": 4.225039746636047, "learning_rate": 6.417223364668157e-06, "loss": 0.3753, "step": 6013 }, { "epoch": 0.42637362637362636, "grad_norm": 3.062508505772062, "learning_rate": 6.416122255931489e-06, "loss": 0.4348, "step": 6014 }, { "epoch": 0.4264445232187168, "grad_norm": 1.5975970237793709, "learning_rate": 6.415021072521614e-06, "loss": 0.3899, "step": 6015 }, { "epoch": 0.42651542006380716, "grad_norm": 2.575475226373744, "learning_rate": 6.4139198144965966e-06, "loss": 0.4503, "step": 6016 }, { "epoch": 0.42658631690889753, "grad_norm": 4.0987294404147105, "learning_rate": 6.412818481914509e-06, "loss": 0.3923, "step": 6017 }, { "epoch": 0.42665721375398796, "grad_norm": 1.2445376549242526, "learning_rate": 6.4117170748334235e-06, "loss": 0.6925, "step": 6018 }, { "epoch": 0.42672811059907834, "grad_norm": 2.196136143680894, "learning_rate": 6.41061559331142e-06, "loss": 0.3782, "step": 6019 }, { "epoch": 0.4267990074441687, "grad_norm": 4.539335705027374, "learning_rate": 6.4095140374065786e-06, "loss": 0.5384, "step": 6020 }, { "epoch": 0.42686990428925914, "grad_norm": 2.6924091239233423, "learning_rate": 6.408412407176987e-06, "loss": 0.459, "step": 6021 }, { "epoch": 0.4269408011343495, "grad_norm": 1.950856350982198, "learning_rate": 6.407310702680735e-06, "loss": 0.4096, "step": 6022 }, { "epoch": 0.42701169797943994, "grad_norm": 2.359914493864203, "learning_rate": 6.406208923975915e-06, "loss": 0.4068, "step": 6023 }, { "epoch": 0.4270825948245303, "grad_norm": 4.413438877163208, "learning_rate": 6.405107071120624e-06, "loss": 0.3884, "step": 6024 }, { "epoch": 0.4271534916696207, "grad_norm": 2.3605488579794636, "learning_rate": 6.404005144172966e-06, "loss": 0.4622, "step": 6025 }, { "epoch": 0.4272243885147111, "grad_norm": 4.051738831837637, "learning_rate": 6.402903143191043e-06, "loss": 0.3973, "step": 6026 }, { "epoch": 0.4272952853598015, "grad_norm": 2.411881269405768, "learning_rate": 6.401801068232969e-06, "loss": 0.4264, "step": 6027 }, { "epoch": 0.42736618220489186, "grad_norm": 2.5360467774422193, "learning_rate": 6.400698919356854e-06, "loss": 0.4568, "step": 6028 }, { "epoch": 0.4274370790499823, "grad_norm": 5.57963686314793, "learning_rate": 6.399596696620815e-06, "loss": 0.4873, "step": 6029 }, { "epoch": 0.42750797589507267, "grad_norm": 2.4626607498543915, "learning_rate": 6.398494400082975e-06, "loss": 0.4148, "step": 6030 }, { "epoch": 0.42757887274016304, "grad_norm": 0.9912313583589716, "learning_rate": 6.397392029801458e-06, "loss": 0.6409, "step": 6031 }, { "epoch": 0.42764976958525347, "grad_norm": 3.0650650348210977, "learning_rate": 6.396289585834392e-06, "loss": 0.3947, "step": 6032 }, { "epoch": 0.42772066643034384, "grad_norm": 5.005622364969182, "learning_rate": 6.39518706823991e-06, "loss": 0.3673, "step": 6033 }, { "epoch": 0.4277915632754342, "grad_norm": 2.1762978862470845, "learning_rate": 6.39408447707615e-06, "loss": 0.371, "step": 6034 }, { "epoch": 0.42786246012052465, "grad_norm": 2.3897908845803117, "learning_rate": 6.39298181240125e-06, "loss": 0.4035, "step": 6035 }, { "epoch": 0.427933356965615, "grad_norm": 4.1562378545156875, "learning_rate": 6.3918790742733586e-06, "loss": 0.4117, "step": 6036 }, { "epoch": 0.42800425381070545, "grad_norm": 3.2262273408711306, "learning_rate": 6.390776262750617e-06, "loss": 0.4468, "step": 6037 }, { "epoch": 0.4280751506557958, "grad_norm": 5.7685481839367485, "learning_rate": 6.389673377891186e-06, "loss": 0.4157, "step": 6038 }, { "epoch": 0.4281460475008862, "grad_norm": 2.447486487911199, "learning_rate": 6.3885704197532135e-06, "loss": 0.402, "step": 6039 }, { "epoch": 0.4282169443459766, "grad_norm": 5.692836137684119, "learning_rate": 6.387467388394866e-06, "loss": 0.441, "step": 6040 }, { "epoch": 0.428287841191067, "grad_norm": 4.498610752647964, "learning_rate": 6.386364283874301e-06, "loss": 0.4312, "step": 6041 }, { "epoch": 0.42835873803615737, "grad_norm": 5.983579875137271, "learning_rate": 6.385261106249692e-06, "loss": 0.3902, "step": 6042 }, { "epoch": 0.4284296348812478, "grad_norm": 2.833429124334824, "learning_rate": 6.384157855579205e-06, "loss": 0.4021, "step": 6043 }, { "epoch": 0.4285005317263382, "grad_norm": 3.1576337788420936, "learning_rate": 6.38305453192102e-06, "loss": 0.3778, "step": 6044 }, { "epoch": 0.42857142857142855, "grad_norm": 3.4306740412543935, "learning_rate": 6.381951135333313e-06, "loss": 0.4371, "step": 6045 }, { "epoch": 0.428642325416519, "grad_norm": 11.60467592378525, "learning_rate": 6.380847665874268e-06, "loss": 0.4383, "step": 6046 }, { "epoch": 0.42871322226160935, "grad_norm": 2.4928340847753945, "learning_rate": 6.379744123602072e-06, "loss": 0.4184, "step": 6047 }, { "epoch": 0.4287841191066998, "grad_norm": 2.9639545257733784, "learning_rate": 6.378640508574914e-06, "loss": 0.4097, "step": 6048 }, { "epoch": 0.42885501595179015, "grad_norm": 1.108482004924472, "learning_rate": 6.37753682085099e-06, "loss": 0.6399, "step": 6049 }, { "epoch": 0.4289259127968805, "grad_norm": 3.59892954372332, "learning_rate": 6.376433060488498e-06, "loss": 0.3899, "step": 6050 }, { "epoch": 0.42899680964197096, "grad_norm": 4.098102675482753, "learning_rate": 6.375329227545642e-06, "loss": 0.4265, "step": 6051 }, { "epoch": 0.42906770648706133, "grad_norm": 2.1484841395658747, "learning_rate": 6.374225322080624e-06, "loss": 0.4355, "step": 6052 }, { "epoch": 0.4291386033321517, "grad_norm": 4.1113141264074695, "learning_rate": 6.373121344151656e-06, "loss": 0.4262, "step": 6053 }, { "epoch": 0.42920950017724213, "grad_norm": 1.9656244930702487, "learning_rate": 6.372017293816952e-06, "loss": 0.4492, "step": 6054 }, { "epoch": 0.4292803970223325, "grad_norm": 4.024280247335564, "learning_rate": 6.37091317113473e-06, "loss": 0.407, "step": 6055 }, { "epoch": 0.4293512938674229, "grad_norm": 2.963072265052312, "learning_rate": 6.369808976163208e-06, "loss": 0.4203, "step": 6056 }, { "epoch": 0.4294221907125133, "grad_norm": 1.886598726320976, "learning_rate": 6.3687047089606135e-06, "loss": 0.4109, "step": 6057 }, { "epoch": 0.4294930875576037, "grad_norm": 3.070384880119506, "learning_rate": 6.367600369585176e-06, "loss": 0.4158, "step": 6058 }, { "epoch": 0.42956398440269405, "grad_norm": 2.368473661045708, "learning_rate": 6.366495958095127e-06, "loss": 0.4503, "step": 6059 }, { "epoch": 0.4296348812477845, "grad_norm": 2.8469208205261314, "learning_rate": 6.365391474548702e-06, "loss": 0.3921, "step": 6060 }, { "epoch": 0.42970577809287486, "grad_norm": 3.478871916170839, "learning_rate": 6.3642869190041436e-06, "loss": 0.4048, "step": 6061 }, { "epoch": 0.4297766749379653, "grad_norm": 5.925116627891919, "learning_rate": 6.363182291519695e-06, "loss": 0.4355, "step": 6062 }, { "epoch": 0.42984757178305566, "grad_norm": 1.7373398544887924, "learning_rate": 6.362077592153604e-06, "loss": 0.4077, "step": 6063 }, { "epoch": 0.42991846862814603, "grad_norm": 3.4867808404357237, "learning_rate": 6.36097282096412e-06, "loss": 0.4287, "step": 6064 }, { "epoch": 0.42998936547323646, "grad_norm": 2.562038912661104, "learning_rate": 6.3598679780095016e-06, "loss": 0.6859, "step": 6065 }, { "epoch": 0.43006026231832684, "grad_norm": 3.6270814857733535, "learning_rate": 6.3587630633480056e-06, "loss": 0.4118, "step": 6066 }, { "epoch": 0.4301311591634172, "grad_norm": 3.688564824547691, "learning_rate": 6.357658077037897e-06, "loss": 0.4242, "step": 6067 }, { "epoch": 0.43020205600850764, "grad_norm": 2.615942873818789, "learning_rate": 6.356553019137444e-06, "loss": 0.3852, "step": 6068 }, { "epoch": 0.430272952853598, "grad_norm": 2.5446736473139873, "learning_rate": 6.355447889704911e-06, "loss": 0.4005, "step": 6069 }, { "epoch": 0.4303438496986884, "grad_norm": 3.5076034051581244, "learning_rate": 6.3543426887985786e-06, "loss": 0.3849, "step": 6070 }, { "epoch": 0.4304147465437788, "grad_norm": 2.519342503218612, "learning_rate": 6.353237416476721e-06, "loss": 0.4642, "step": 6071 }, { "epoch": 0.4304856433888692, "grad_norm": 3.457705944025552, "learning_rate": 6.352132072797623e-06, "loss": 0.4396, "step": 6072 }, { "epoch": 0.43055654023395956, "grad_norm": 2.748558136567107, "learning_rate": 6.351026657819568e-06, "loss": 0.4391, "step": 6073 }, { "epoch": 0.43062743707905, "grad_norm": 3.597650247424484, "learning_rate": 6.349921171600846e-06, "loss": 0.4057, "step": 6074 }, { "epoch": 0.43069833392414036, "grad_norm": 2.936591003805226, "learning_rate": 6.348815614199752e-06, "loss": 0.4175, "step": 6075 }, { "epoch": 0.4307692307692308, "grad_norm": 2.2116838183515557, "learning_rate": 6.347709985674579e-06, "loss": 0.431, "step": 6076 }, { "epoch": 0.43084012761432117, "grad_norm": 3.160601389159736, "learning_rate": 6.346604286083631e-06, "loss": 0.355, "step": 6077 }, { "epoch": 0.43091102445941154, "grad_norm": 3.461596890937984, "learning_rate": 6.34549851548521e-06, "loss": 0.4621, "step": 6078 }, { "epoch": 0.43098192130450197, "grad_norm": 2.761824956447069, "learning_rate": 6.344392673937626e-06, "loss": 0.3243, "step": 6079 }, { "epoch": 0.43105281814959234, "grad_norm": 2.437594724511213, "learning_rate": 6.34328676149919e-06, "loss": 0.3969, "step": 6080 }, { "epoch": 0.4311237149946827, "grad_norm": 2.104716105287619, "learning_rate": 6.342180778228218e-06, "loss": 0.3928, "step": 6081 }, { "epoch": 0.43119461183977315, "grad_norm": 2.7824921209538434, "learning_rate": 6.341074724183028e-06, "loss": 0.4228, "step": 6082 }, { "epoch": 0.4312655086848635, "grad_norm": 3.433158939641639, "learning_rate": 6.3399685994219465e-06, "loss": 0.3972, "step": 6083 }, { "epoch": 0.4313364055299539, "grad_norm": 2.501309360359537, "learning_rate": 6.338862404003297e-06, "loss": 0.4374, "step": 6084 }, { "epoch": 0.4314073023750443, "grad_norm": 5.75311505973521, "learning_rate": 6.337756137985411e-06, "loss": 0.4196, "step": 6085 }, { "epoch": 0.4314781992201347, "grad_norm": 2.0236553852243566, "learning_rate": 6.336649801426622e-06, "loss": 0.3528, "step": 6086 }, { "epoch": 0.4315490960652251, "grad_norm": 2.320267305890605, "learning_rate": 6.33554339438527e-06, "loss": 0.4216, "step": 6087 }, { "epoch": 0.4316199929103155, "grad_norm": 2.926333979513692, "learning_rate": 6.334436916919694e-06, "loss": 0.4145, "step": 6088 }, { "epoch": 0.43169088975540587, "grad_norm": 1.6508137335220454, "learning_rate": 6.333330369088242e-06, "loss": 0.383, "step": 6089 }, { "epoch": 0.4317617866004963, "grad_norm": 3.041614506450271, "learning_rate": 6.332223750949262e-06, "loss": 0.3952, "step": 6090 }, { "epoch": 0.4318326834455867, "grad_norm": 2.386479057184535, "learning_rate": 6.3311170625611065e-06, "loss": 0.3607, "step": 6091 }, { "epoch": 0.43190358029067705, "grad_norm": 5.395590490697909, "learning_rate": 6.3300103039821315e-06, "loss": 0.4098, "step": 6092 }, { "epoch": 0.4319744771357675, "grad_norm": 4.9287861877688925, "learning_rate": 6.328903475270699e-06, "loss": 0.3978, "step": 6093 }, { "epoch": 0.43204537398085785, "grad_norm": 2.556337093874214, "learning_rate": 6.327796576485172e-06, "loss": 0.3894, "step": 6094 }, { "epoch": 0.4321162708259482, "grad_norm": 3.2651505262424543, "learning_rate": 6.3266896076839164e-06, "loss": 0.4699, "step": 6095 }, { "epoch": 0.43218716767103865, "grad_norm": 2.5311114132271637, "learning_rate": 6.3255825689253085e-06, "loss": 0.4094, "step": 6096 }, { "epoch": 0.432258064516129, "grad_norm": 3.022345717236632, "learning_rate": 6.324475460267717e-06, "loss": 0.4375, "step": 6097 }, { "epoch": 0.4323289613612194, "grad_norm": 2.000086414369012, "learning_rate": 6.323368281769524e-06, "loss": 0.4046, "step": 6098 }, { "epoch": 0.43239985820630983, "grad_norm": 2.837979766989498, "learning_rate": 6.322261033489112e-06, "loss": 0.4482, "step": 6099 }, { "epoch": 0.4324707550514002, "grad_norm": 2.612032970672807, "learning_rate": 6.321153715484867e-06, "loss": 0.4069, "step": 6100 }, { "epoch": 0.43254165189649063, "grad_norm": 2.445636307434933, "learning_rate": 6.3200463278151765e-06, "loss": 0.3875, "step": 6101 }, { "epoch": 0.432612548741581, "grad_norm": 4.066413295702399, "learning_rate": 6.318938870538438e-06, "loss": 0.4417, "step": 6102 }, { "epoch": 0.4326834455866714, "grad_norm": 3.1776033361699834, "learning_rate": 6.317831343713045e-06, "loss": 0.4336, "step": 6103 }, { "epoch": 0.4327543424317618, "grad_norm": 4.619742899907135, "learning_rate": 6.316723747397399e-06, "loss": 0.4596, "step": 6104 }, { "epoch": 0.4328252392768522, "grad_norm": 2.1512617137611163, "learning_rate": 6.315616081649904e-06, "loss": 0.3705, "step": 6105 }, { "epoch": 0.43289613612194255, "grad_norm": 3.2285987177281976, "learning_rate": 6.31450834652897e-06, "loss": 0.4154, "step": 6106 }, { "epoch": 0.432967032967033, "grad_norm": 1.9770682734138347, "learning_rate": 6.3134005420930085e-06, "loss": 0.4344, "step": 6107 }, { "epoch": 0.43303792981212336, "grad_norm": 3.8399770065302343, "learning_rate": 6.312292668400431e-06, "loss": 0.4009, "step": 6108 }, { "epoch": 0.43310882665721373, "grad_norm": 3.2096973368217503, "learning_rate": 6.3111847255096625e-06, "loss": 0.4183, "step": 6109 }, { "epoch": 0.43317972350230416, "grad_norm": 1.0677509389718067, "learning_rate": 6.310076713479122e-06, "loss": 0.7381, "step": 6110 }, { "epoch": 0.43325062034739453, "grad_norm": 3.070857167602233, "learning_rate": 6.308968632367235e-06, "loss": 0.4173, "step": 6111 }, { "epoch": 0.43332151719248496, "grad_norm": 2.957389822282657, "learning_rate": 6.307860482232434e-06, "loss": 0.3625, "step": 6112 }, { "epoch": 0.43339241403757534, "grad_norm": 2.536172377011186, "learning_rate": 6.306752263133152e-06, "loss": 0.4155, "step": 6113 }, { "epoch": 0.4334633108826657, "grad_norm": 2.268587510977493, "learning_rate": 6.305643975127824e-06, "loss": 0.4117, "step": 6114 }, { "epoch": 0.43353420772775614, "grad_norm": 2.915622931951984, "learning_rate": 6.304535618274895e-06, "loss": 0.4317, "step": 6115 }, { "epoch": 0.4336051045728465, "grad_norm": 2.927569830120911, "learning_rate": 6.303427192632805e-06, "loss": 0.4454, "step": 6116 }, { "epoch": 0.4336760014179369, "grad_norm": 2.6016339526681223, "learning_rate": 6.302318698260006e-06, "loss": 0.4591, "step": 6117 }, { "epoch": 0.4337468982630273, "grad_norm": 3.4794303835022347, "learning_rate": 6.301210135214947e-06, "loss": 0.4472, "step": 6118 }, { "epoch": 0.4338177951081177, "grad_norm": 3.055333606773503, "learning_rate": 6.300101503556084e-06, "loss": 0.3926, "step": 6119 }, { "epoch": 0.43388869195320806, "grad_norm": 3.1325480250650264, "learning_rate": 6.2989928033418765e-06, "loss": 0.4741, "step": 6120 }, { "epoch": 0.4339595887982985, "grad_norm": 1.9016046147669219, "learning_rate": 6.297884034630789e-06, "loss": 0.4115, "step": 6121 }, { "epoch": 0.43403048564338886, "grad_norm": 3.126905366594088, "learning_rate": 6.296775197481284e-06, "loss": 0.4177, "step": 6122 }, { "epoch": 0.43410138248847924, "grad_norm": 5.28689914703813, "learning_rate": 6.295666291951833e-06, "loss": 0.3478, "step": 6123 }, { "epoch": 0.43417227933356967, "grad_norm": 3.068840033123465, "learning_rate": 6.29455731810091e-06, "loss": 0.4151, "step": 6124 }, { "epoch": 0.43424317617866004, "grad_norm": 3.1923305495292387, "learning_rate": 6.293448275986991e-06, "loss": 0.3667, "step": 6125 }, { "epoch": 0.43431407302375047, "grad_norm": 7.899073437605946, "learning_rate": 6.292339165668558e-06, "loss": 0.4454, "step": 6126 }, { "epoch": 0.43438496986884084, "grad_norm": 3.249240601588513, "learning_rate": 6.291229987204093e-06, "loss": 0.4682, "step": 6127 }, { "epoch": 0.4344558667139312, "grad_norm": 3.018657177705846, "learning_rate": 6.290120740652087e-06, "loss": 0.395, "step": 6128 }, { "epoch": 0.43452676355902164, "grad_norm": 3.1725454161754287, "learning_rate": 6.289011426071029e-06, "loss": 0.4434, "step": 6129 }, { "epoch": 0.434597660404112, "grad_norm": 3.8361885486015628, "learning_rate": 6.287902043519414e-06, "loss": 0.4279, "step": 6130 }, { "epoch": 0.4346685572492024, "grad_norm": 2.164147760315507, "learning_rate": 6.286792593055742e-06, "loss": 0.4248, "step": 6131 }, { "epoch": 0.4347394540942928, "grad_norm": 4.027771825244459, "learning_rate": 6.285683074738514e-06, "loss": 0.4399, "step": 6132 }, { "epoch": 0.4348103509393832, "grad_norm": 2.54569417655965, "learning_rate": 6.2845734886262355e-06, "loss": 0.4008, "step": 6133 }, { "epoch": 0.43488124778447357, "grad_norm": 6.573568986526615, "learning_rate": 6.283463834777418e-06, "loss": 0.4272, "step": 6134 }, { "epoch": 0.434952144629564, "grad_norm": 2.370044056367089, "learning_rate": 6.282354113250572e-06, "loss": 0.4157, "step": 6135 }, { "epoch": 0.43502304147465437, "grad_norm": 1.54014446971334, "learning_rate": 6.281244324104214e-06, "loss": 0.4112, "step": 6136 }, { "epoch": 0.43509393831974474, "grad_norm": 3.6420824375827774, "learning_rate": 6.2801344673968654e-06, "loss": 0.4112, "step": 6137 }, { "epoch": 0.4351648351648352, "grad_norm": 11.98134637229678, "learning_rate": 6.27902454318705e-06, "loss": 0.4151, "step": 6138 }, { "epoch": 0.43523573200992555, "grad_norm": 2.106198519574816, "learning_rate": 6.277914551533295e-06, "loss": 0.4114, "step": 6139 }, { "epoch": 0.435306628855016, "grad_norm": 2.244211256467716, "learning_rate": 6.276804492494127e-06, "loss": 0.4284, "step": 6140 }, { "epoch": 0.43537752570010635, "grad_norm": 3.291512624455305, "learning_rate": 6.275694366128087e-06, "loss": 0.3652, "step": 6141 }, { "epoch": 0.4354484225451967, "grad_norm": 2.1893030014343506, "learning_rate": 6.274584172493707e-06, "loss": 0.4779, "step": 6142 }, { "epoch": 0.43551931939028715, "grad_norm": 2.8577070728679064, "learning_rate": 6.273473911649532e-06, "loss": 0.3774, "step": 6143 }, { "epoch": 0.4355902162353775, "grad_norm": 4.908070983951429, "learning_rate": 6.272363583654106e-06, "loss": 0.4105, "step": 6144 }, { "epoch": 0.4356611130804679, "grad_norm": 3.1426588284058012, "learning_rate": 6.271253188565975e-06, "loss": 0.4424, "step": 6145 }, { "epoch": 0.43573200992555833, "grad_norm": 8.091531900212956, "learning_rate": 6.270142726443695e-06, "loss": 0.4162, "step": 6146 }, { "epoch": 0.4358029067706487, "grad_norm": 4.357700677818638, "learning_rate": 6.26903219734582e-06, "loss": 0.3693, "step": 6147 }, { "epoch": 0.4358738036157391, "grad_norm": 2.132230735114735, "learning_rate": 6.267921601330907e-06, "loss": 0.3948, "step": 6148 }, { "epoch": 0.4359447004608295, "grad_norm": 2.668623590130363, "learning_rate": 6.266810938457522e-06, "loss": 0.4282, "step": 6149 }, { "epoch": 0.4360155973059199, "grad_norm": 3.2864135625476316, "learning_rate": 6.265700208784228e-06, "loss": 0.4224, "step": 6150 }, { "epoch": 0.4360864941510103, "grad_norm": 7.800280513357666, "learning_rate": 6.264589412369596e-06, "loss": 0.4544, "step": 6151 }, { "epoch": 0.4361573909961007, "grad_norm": 3.313120079420529, "learning_rate": 6.263478549272199e-06, "loss": 0.4001, "step": 6152 }, { "epoch": 0.43622828784119105, "grad_norm": 4.452551067444001, "learning_rate": 6.262367619550614e-06, "loss": 0.4068, "step": 6153 }, { "epoch": 0.4362991846862815, "grad_norm": 2.5408068390973915, "learning_rate": 6.261256623263422e-06, "loss": 0.4447, "step": 6154 }, { "epoch": 0.43637008153137186, "grad_norm": 4.850975797066562, "learning_rate": 6.260145560469204e-06, "loss": 0.4272, "step": 6155 }, { "epoch": 0.43644097837646223, "grad_norm": 3.2216137730550285, "learning_rate": 6.25903443122655e-06, "loss": 0.4081, "step": 6156 }, { "epoch": 0.43651187522155266, "grad_norm": 2.0143448236331154, "learning_rate": 6.2579232355940474e-06, "loss": 0.4077, "step": 6157 }, { "epoch": 0.43658277206664303, "grad_norm": 2.8415989973189077, "learning_rate": 6.256811973630294e-06, "loss": 0.4743, "step": 6158 }, { "epoch": 0.4366536689117334, "grad_norm": 4.310029932047435, "learning_rate": 6.255700645393885e-06, "loss": 0.4022, "step": 6159 }, { "epoch": 0.43672456575682383, "grad_norm": 5.76725801614875, "learning_rate": 6.2545892509434244e-06, "loss": 0.5131, "step": 6160 }, { "epoch": 0.4367954626019142, "grad_norm": 1.8714726904178813, "learning_rate": 6.253477790337512e-06, "loss": 0.4601, "step": 6161 }, { "epoch": 0.4368663594470046, "grad_norm": 2.6577637777871983, "learning_rate": 6.252366263634763e-06, "loss": 0.4235, "step": 6162 }, { "epoch": 0.436937256292095, "grad_norm": 3.338812881169432, "learning_rate": 6.251254670893783e-06, "loss": 0.3766, "step": 6163 }, { "epoch": 0.4370081531371854, "grad_norm": 2.5129312967810127, "learning_rate": 6.250143012173188e-06, "loss": 0.3889, "step": 6164 }, { "epoch": 0.4370790499822758, "grad_norm": 3.1166596108600655, "learning_rate": 6.249031287531598e-06, "loss": 0.4263, "step": 6165 }, { "epoch": 0.4371499468273662, "grad_norm": 1.8612200553330542, "learning_rate": 6.247919497027637e-06, "loss": 0.4633, "step": 6166 }, { "epoch": 0.43722084367245656, "grad_norm": 2.434808363283465, "learning_rate": 6.246807640719927e-06, "loss": 0.457, "step": 6167 }, { "epoch": 0.437291740517547, "grad_norm": 2.1892749020874565, "learning_rate": 6.245695718667099e-06, "loss": 0.3849, "step": 6168 }, { "epoch": 0.43736263736263736, "grad_norm": 0.9730869579319249, "learning_rate": 6.2445837309277835e-06, "loss": 0.7197, "step": 6169 }, { "epoch": 0.43743353420772774, "grad_norm": 3.4672638304374392, "learning_rate": 6.24347167756062e-06, "loss": 0.4331, "step": 6170 }, { "epoch": 0.43750443105281817, "grad_norm": 2.55472180973366, "learning_rate": 6.242359558624244e-06, "loss": 0.4322, "step": 6171 }, { "epoch": 0.43757532789790854, "grad_norm": 5.585645742931132, "learning_rate": 6.241247374177301e-06, "loss": 0.4139, "step": 6172 }, { "epoch": 0.4376462247429989, "grad_norm": 7.893030416541228, "learning_rate": 6.240135124278437e-06, "loss": 0.3989, "step": 6173 }, { "epoch": 0.43771712158808934, "grad_norm": 2.3399159968198306, "learning_rate": 6.2390228089863e-06, "loss": 0.4251, "step": 6174 }, { "epoch": 0.4377880184331797, "grad_norm": 3.3663356842638823, "learning_rate": 6.237910428359546e-06, "loss": 0.3955, "step": 6175 }, { "epoch": 0.43785891527827014, "grad_norm": 2.795951429541169, "learning_rate": 6.23679798245683e-06, "loss": 0.3946, "step": 6176 }, { "epoch": 0.4379298121233605, "grad_norm": 1.9959097182978276, "learning_rate": 6.235685471336812e-06, "loss": 0.3615, "step": 6177 }, { "epoch": 0.4380007089684509, "grad_norm": 2.554289678083592, "learning_rate": 6.234572895058156e-06, "loss": 0.4036, "step": 6178 }, { "epoch": 0.4380716058135413, "grad_norm": 2.261181267806157, "learning_rate": 6.23346025367953e-06, "loss": 0.4033, "step": 6179 }, { "epoch": 0.4381425026586317, "grad_norm": 3.5651240828753794, "learning_rate": 6.232347547259601e-06, "loss": 0.4515, "step": 6180 }, { "epoch": 0.43821339950372207, "grad_norm": 4.535991291578405, "learning_rate": 6.231234775857047e-06, "loss": 0.3953, "step": 6181 }, { "epoch": 0.4382842963488125, "grad_norm": 4.988137647937287, "learning_rate": 6.230121939530543e-06, "loss": 0.3602, "step": 6182 }, { "epoch": 0.43835519319390287, "grad_norm": 2.6835305664210725, "learning_rate": 6.22900903833877e-06, "loss": 0.4147, "step": 6183 }, { "epoch": 0.43842609003899324, "grad_norm": 4.392164562990537, "learning_rate": 6.227896072340413e-06, "loss": 0.4496, "step": 6184 }, { "epoch": 0.4384969868840837, "grad_norm": 2.4937183737931226, "learning_rate": 6.226783041594158e-06, "loss": 0.4006, "step": 6185 }, { "epoch": 0.43856788372917405, "grad_norm": 9.327170308036461, "learning_rate": 6.225669946158699e-06, "loss": 0.4758, "step": 6186 }, { "epoch": 0.4386387805742644, "grad_norm": 2.3699161258295716, "learning_rate": 6.224556786092726e-06, "loss": 0.4129, "step": 6187 }, { "epoch": 0.43870967741935485, "grad_norm": 2.9240688582679852, "learning_rate": 6.223443561454939e-06, "loss": 0.4659, "step": 6188 }, { "epoch": 0.4387805742644452, "grad_norm": 2.2211618836534837, "learning_rate": 6.22233027230404e-06, "loss": 0.4093, "step": 6189 }, { "epoch": 0.43885147110953565, "grad_norm": 2.0482748158394535, "learning_rate": 6.221216918698732e-06, "loss": 0.443, "step": 6190 }, { "epoch": 0.438922367954626, "grad_norm": 1.45288990324726, "learning_rate": 6.220103500697725e-06, "loss": 0.3749, "step": 6191 }, { "epoch": 0.4389932647997164, "grad_norm": 2.2747946252999105, "learning_rate": 6.21899001835973e-06, "loss": 0.3906, "step": 6192 }, { "epoch": 0.4390641616448068, "grad_norm": 34.654342285592634, "learning_rate": 6.217876471743458e-06, "loss": 0.4151, "step": 6193 }, { "epoch": 0.4391350584898972, "grad_norm": 3.142871126043376, "learning_rate": 6.216762860907632e-06, "loss": 0.4057, "step": 6194 }, { "epoch": 0.4392059553349876, "grad_norm": 2.383121916125203, "learning_rate": 6.2156491859109715e-06, "loss": 0.4474, "step": 6195 }, { "epoch": 0.439276852180078, "grad_norm": 1.9910993936810524, "learning_rate": 6.214535446812201e-06, "loss": 0.4496, "step": 6196 }, { "epoch": 0.4393477490251684, "grad_norm": 1.899558649205468, "learning_rate": 6.213421643670049e-06, "loss": 0.3866, "step": 6197 }, { "epoch": 0.43941864587025875, "grad_norm": 4.707901110168214, "learning_rate": 6.212307776543248e-06, "loss": 0.4555, "step": 6198 }, { "epoch": 0.4394895427153492, "grad_norm": 3.795007677407544, "learning_rate": 6.211193845490534e-06, "loss": 0.3887, "step": 6199 }, { "epoch": 0.43956043956043955, "grad_norm": 1.0395268399059532, "learning_rate": 6.210079850570643e-06, "loss": 0.645, "step": 6200 }, { "epoch": 0.4396313364055299, "grad_norm": 2.4390689360797433, "learning_rate": 6.208965791842318e-06, "loss": 0.4181, "step": 6201 }, { "epoch": 0.43970223325062036, "grad_norm": 3.023889935655745, "learning_rate": 6.207851669364303e-06, "loss": 0.4619, "step": 6202 }, { "epoch": 0.43977313009571073, "grad_norm": 4.2586971827390885, "learning_rate": 6.206737483195349e-06, "loss": 0.3972, "step": 6203 }, { "epoch": 0.43984402694080116, "grad_norm": 3.24569870230072, "learning_rate": 6.205623233394204e-06, "loss": 0.3915, "step": 6204 }, { "epoch": 0.43991492378589153, "grad_norm": 3.635261890107923, "learning_rate": 6.20450892001963e-06, "loss": 0.4038, "step": 6205 }, { "epoch": 0.4399858206309819, "grad_norm": 2.2170401111550775, "learning_rate": 6.203394543130377e-06, "loss": 0.4493, "step": 6206 }, { "epoch": 0.44005671747607233, "grad_norm": 2.9577335690693825, "learning_rate": 6.202280102785213e-06, "loss": 0.3735, "step": 6207 }, { "epoch": 0.4401276143211627, "grad_norm": 6.90251330051408, "learning_rate": 6.201165599042902e-06, "loss": 0.4456, "step": 6208 }, { "epoch": 0.4401985111662531, "grad_norm": 3.334999665048235, "learning_rate": 6.200051031962211e-06, "loss": 0.3871, "step": 6209 }, { "epoch": 0.4402694080113435, "grad_norm": 3.024315540465282, "learning_rate": 6.198936401601914e-06, "loss": 0.4307, "step": 6210 }, { "epoch": 0.4403403048564339, "grad_norm": 2.950826713209929, "learning_rate": 6.197821708020785e-06, "loss": 0.3681, "step": 6211 }, { "epoch": 0.44041120170152426, "grad_norm": 2.037629527537283, "learning_rate": 6.196706951277604e-06, "loss": 0.4575, "step": 6212 }, { "epoch": 0.4404820985466147, "grad_norm": 6.778203913480892, "learning_rate": 6.195592131431152e-06, "loss": 0.3918, "step": 6213 }, { "epoch": 0.44055299539170506, "grad_norm": 4.137630761555231, "learning_rate": 6.194477248540212e-06, "loss": 0.4197, "step": 6214 }, { "epoch": 0.4406238922367955, "grad_norm": 2.5156105496816044, "learning_rate": 6.193362302663578e-06, "loss": 0.4905, "step": 6215 }, { "epoch": 0.44069478908188586, "grad_norm": 3.247466138631752, "learning_rate": 6.192247293860038e-06, "loss": 0.4148, "step": 6216 }, { "epoch": 0.44076568592697624, "grad_norm": 2.6924450507142765, "learning_rate": 6.191132222188387e-06, "loss": 0.4844, "step": 6217 }, { "epoch": 0.44083658277206667, "grad_norm": 1.9759830880779363, "learning_rate": 6.1900170877074265e-06, "loss": 0.3837, "step": 6218 }, { "epoch": 0.44090747961715704, "grad_norm": 7.21569960648171, "learning_rate": 6.188901890475955e-06, "loss": 0.4246, "step": 6219 }, { "epoch": 0.4409783764622474, "grad_norm": 2.836319212032771, "learning_rate": 6.18778663055278e-06, "loss": 0.4165, "step": 6220 }, { "epoch": 0.44104927330733784, "grad_norm": 4.293708263356249, "learning_rate": 6.18667130799671e-06, "loss": 0.4481, "step": 6221 }, { "epoch": 0.4411201701524282, "grad_norm": 2.2102891458575575, "learning_rate": 6.185555922866555e-06, "loss": 0.3932, "step": 6222 }, { "epoch": 0.4411910669975186, "grad_norm": 2.003507055620526, "learning_rate": 6.1844404752211315e-06, "loss": 0.4078, "step": 6223 }, { "epoch": 0.441261963842609, "grad_norm": 3.299890254475551, "learning_rate": 6.183324965119259e-06, "loss": 0.4091, "step": 6224 }, { "epoch": 0.4413328606876994, "grad_norm": 4.0425638879909584, "learning_rate": 6.182209392619757e-06, "loss": 0.3424, "step": 6225 }, { "epoch": 0.44140375753278976, "grad_norm": 2.1448698737704315, "learning_rate": 6.181093757781452e-06, "loss": 0.4117, "step": 6226 }, { "epoch": 0.4414746543778802, "grad_norm": 8.031404515114168, "learning_rate": 6.179978060663171e-06, "loss": 0.4283, "step": 6227 }, { "epoch": 0.44154555122297057, "grad_norm": 3.925766440924221, "learning_rate": 6.178862301323746e-06, "loss": 0.4466, "step": 6228 }, { "epoch": 0.441616448068061, "grad_norm": 2.679103499102769, "learning_rate": 6.177746479822012e-06, "loss": 0.4079, "step": 6229 }, { "epoch": 0.44168734491315137, "grad_norm": 4.292721544460541, "learning_rate": 6.176630596216808e-06, "loss": 0.3876, "step": 6230 }, { "epoch": 0.44175824175824174, "grad_norm": 2.2026951258864083, "learning_rate": 6.175514650566975e-06, "loss": 0.3953, "step": 6231 }, { "epoch": 0.44182913860333217, "grad_norm": 3.329697052385944, "learning_rate": 6.174398642931356e-06, "loss": 0.3827, "step": 6232 }, { "epoch": 0.44190003544842255, "grad_norm": 1.9313246226656733, "learning_rate": 6.173282573368801e-06, "loss": 0.3637, "step": 6233 }, { "epoch": 0.4419709322935129, "grad_norm": 3.123698660178552, "learning_rate": 6.17216644193816e-06, "loss": 0.457, "step": 6234 }, { "epoch": 0.44204182913860335, "grad_norm": 1.1683952780052747, "learning_rate": 6.171050248698287e-06, "loss": 0.6879, "step": 6235 }, { "epoch": 0.4421127259836937, "grad_norm": 3.728299254489906, "learning_rate": 6.1699339937080416e-06, "loss": 0.4259, "step": 6236 }, { "epoch": 0.4421836228287841, "grad_norm": 2.7675366977895686, "learning_rate": 6.168817677026284e-06, "loss": 0.4451, "step": 6237 }, { "epoch": 0.4422545196738745, "grad_norm": 4.271353123667232, "learning_rate": 6.167701298711877e-06, "loss": 0.4585, "step": 6238 }, { "epoch": 0.4423254165189649, "grad_norm": 2.9707829938029793, "learning_rate": 6.166584858823691e-06, "loss": 0.4348, "step": 6239 }, { "epoch": 0.4423963133640553, "grad_norm": 2.4833656952824463, "learning_rate": 6.165468357420593e-06, "loss": 0.4329, "step": 6240 }, { "epoch": 0.4424672102091457, "grad_norm": 2.612828858111381, "learning_rate": 6.164351794561461e-06, "loss": 0.3799, "step": 6241 }, { "epoch": 0.4425381070542361, "grad_norm": 3.7964612916487432, "learning_rate": 6.163235170305168e-06, "loss": 0.4791, "step": 6242 }, { "epoch": 0.4426090038993265, "grad_norm": 4.98122689988152, "learning_rate": 6.1621184847105975e-06, "loss": 0.4096, "step": 6243 }, { "epoch": 0.4426799007444169, "grad_norm": 3.270093750937916, "learning_rate": 6.161001737836634e-06, "loss": 0.408, "step": 6244 }, { "epoch": 0.44275079758950725, "grad_norm": 4.732187824400281, "learning_rate": 6.159884929742162e-06, "loss": 0.3739, "step": 6245 }, { "epoch": 0.4428216944345977, "grad_norm": 2.6940018248277764, "learning_rate": 6.15876806048607e-06, "loss": 0.407, "step": 6246 }, { "epoch": 0.44289259127968805, "grad_norm": 2.582393740147656, "learning_rate": 6.157651130127256e-06, "loss": 0.4296, "step": 6247 }, { "epoch": 0.4429634881247784, "grad_norm": 4.617308309695455, "learning_rate": 6.156534138724612e-06, "loss": 0.433, "step": 6248 }, { "epoch": 0.44303438496986886, "grad_norm": 3.162962830849623, "learning_rate": 6.155417086337042e-06, "loss": 0.4375, "step": 6249 }, { "epoch": 0.44310528181495923, "grad_norm": 2.0647766338050357, "learning_rate": 6.154299973023448e-06, "loss": 0.4056, "step": 6250 }, { "epoch": 0.4431761786600496, "grad_norm": 2.0507617885625313, "learning_rate": 6.153182798842733e-06, "loss": 0.4223, "step": 6251 }, { "epoch": 0.44324707550514003, "grad_norm": 1.845294402968902, "learning_rate": 6.1520655638538105e-06, "loss": 0.4221, "step": 6252 }, { "epoch": 0.4433179723502304, "grad_norm": 1.0253663983486148, "learning_rate": 6.150948268115592e-06, "loss": 0.6913, "step": 6253 }, { "epoch": 0.44338886919532083, "grad_norm": 3.104761148581146, "learning_rate": 6.149830911686991e-06, "loss": 0.379, "step": 6254 }, { "epoch": 0.4434597660404112, "grad_norm": 2.575341159762955, "learning_rate": 6.14871349462693e-06, "loss": 0.3473, "step": 6255 }, { "epoch": 0.4435306628855016, "grad_norm": 4.445097880232542, "learning_rate": 6.147596016994328e-06, "loss": 0.3917, "step": 6256 }, { "epoch": 0.443601559730592, "grad_norm": 15.45312086798662, "learning_rate": 6.146478478848113e-06, "loss": 0.3968, "step": 6257 }, { "epoch": 0.4436724565756824, "grad_norm": 3.1268600393898045, "learning_rate": 6.145360880247214e-06, "loss": 0.4077, "step": 6258 }, { "epoch": 0.44374335342077276, "grad_norm": 2.795612513768045, "learning_rate": 6.1442432212505595e-06, "loss": 0.4033, "step": 6259 }, { "epoch": 0.4438142502658632, "grad_norm": 2.3380337976008123, "learning_rate": 6.143125501917087e-06, "loss": 0.453, "step": 6260 }, { "epoch": 0.44388514711095356, "grad_norm": 2.7045488194909937, "learning_rate": 6.142007722305736e-06, "loss": 0.4526, "step": 6261 }, { "epoch": 0.44395604395604393, "grad_norm": 2.233910744524494, "learning_rate": 6.140889882475445e-06, "loss": 0.3993, "step": 6262 }, { "epoch": 0.44402694080113436, "grad_norm": 1.3085283516273782, "learning_rate": 6.13977198248516e-06, "loss": 0.7417, "step": 6263 }, { "epoch": 0.44409783764622474, "grad_norm": 2.0101436499811767, "learning_rate": 6.138654022393826e-06, "loss": 0.4029, "step": 6264 }, { "epoch": 0.4441687344913151, "grad_norm": 2.0962921465197444, "learning_rate": 6.1375360022604e-06, "loss": 0.3882, "step": 6265 }, { "epoch": 0.44423963133640554, "grad_norm": 2.3516183085160196, "learning_rate": 6.13641792214383e-06, "loss": 0.3629, "step": 6266 }, { "epoch": 0.4443105281814959, "grad_norm": 2.4313868897728725, "learning_rate": 6.135299782103077e-06, "loss": 0.3601, "step": 6267 }, { "epoch": 0.44438142502658634, "grad_norm": 3.034113859167718, "learning_rate": 6.134181582197099e-06, "loss": 0.4026, "step": 6268 }, { "epoch": 0.4444523218716767, "grad_norm": 2.8489249171739925, "learning_rate": 6.13306332248486e-06, "loss": 0.4165, "step": 6269 }, { "epoch": 0.4445232187167671, "grad_norm": 1.6426232846424076, "learning_rate": 6.131945003025328e-06, "loss": 0.4194, "step": 6270 }, { "epoch": 0.4445941155618575, "grad_norm": 2.420224995161739, "learning_rate": 6.130826623877473e-06, "loss": 0.3658, "step": 6271 }, { "epoch": 0.4446650124069479, "grad_norm": 9.247492765368433, "learning_rate": 6.129708185100265e-06, "loss": 0.3978, "step": 6272 }, { "epoch": 0.44473590925203826, "grad_norm": 3.452708691151653, "learning_rate": 6.128589686752683e-06, "loss": 0.4603, "step": 6273 }, { "epoch": 0.4448068060971287, "grad_norm": 3.3277864654276135, "learning_rate": 6.127471128893705e-06, "loss": 0.4055, "step": 6274 }, { "epoch": 0.44487770294221907, "grad_norm": 3.8646191748259793, "learning_rate": 6.1263525115823145e-06, "loss": 0.4786, "step": 6275 }, { "epoch": 0.44494859978730944, "grad_norm": 2.0341055092313303, "learning_rate": 6.1252338348774964e-06, "loss": 0.3686, "step": 6276 }, { "epoch": 0.44501949663239987, "grad_norm": 2.9993977137408945, "learning_rate": 6.124115098838238e-06, "loss": 0.4467, "step": 6277 }, { "epoch": 0.44509039347749024, "grad_norm": 6.935974640847557, "learning_rate": 6.1229963035235344e-06, "loss": 0.4724, "step": 6278 }, { "epoch": 0.44516129032258067, "grad_norm": 1.6979391020661279, "learning_rate": 6.121877448992377e-06, "loss": 0.4156, "step": 6279 }, { "epoch": 0.44523218716767105, "grad_norm": 3.129063680326572, "learning_rate": 6.120758535303767e-06, "loss": 0.455, "step": 6280 }, { "epoch": 0.4453030840127614, "grad_norm": 4.968983689309812, "learning_rate": 6.119639562516703e-06, "loss": 0.4422, "step": 6281 }, { "epoch": 0.44537398085785185, "grad_norm": 2.543914523953785, "learning_rate": 6.118520530690191e-06, "loss": 0.3675, "step": 6282 }, { "epoch": 0.4454448777029422, "grad_norm": 2.3449917261347926, "learning_rate": 6.117401439883235e-06, "loss": 0.4239, "step": 6283 }, { "epoch": 0.4455157745480326, "grad_norm": 8.360801004645104, "learning_rate": 6.116282290154851e-06, "loss": 0.4305, "step": 6284 }, { "epoch": 0.445586671393123, "grad_norm": 3.6058463382646018, "learning_rate": 6.115163081564048e-06, "loss": 0.4403, "step": 6285 }, { "epoch": 0.4456575682382134, "grad_norm": 2.876176237642009, "learning_rate": 6.114043814169846e-06, "loss": 0.4012, "step": 6286 }, { "epoch": 0.44572846508330377, "grad_norm": 4.6822646568524995, "learning_rate": 6.112924488031264e-06, "loss": 0.449, "step": 6287 }, { "epoch": 0.4457993619283942, "grad_norm": 3.09588438515164, "learning_rate": 6.111805103207322e-06, "loss": 0.4067, "step": 6288 }, { "epoch": 0.4458702587734846, "grad_norm": 4.19764794723872, "learning_rate": 6.11068565975705e-06, "loss": 0.3952, "step": 6289 }, { "epoch": 0.44594115561857495, "grad_norm": 11.458222962410513, "learning_rate": 6.109566157739476e-06, "loss": 0.4611, "step": 6290 }, { "epoch": 0.4460120524636654, "grad_norm": 6.720955273724729, "learning_rate": 6.108446597213631e-06, "loss": 0.481, "step": 6291 }, { "epoch": 0.44608294930875575, "grad_norm": 2.7159665728736178, "learning_rate": 6.10732697823855e-06, "loss": 0.3911, "step": 6292 }, { "epoch": 0.4461538461538462, "grad_norm": 2.883921121229784, "learning_rate": 6.106207300873272e-06, "loss": 0.4594, "step": 6293 }, { "epoch": 0.44622474299893655, "grad_norm": 3.782764783340411, "learning_rate": 6.105087565176839e-06, "loss": 0.4351, "step": 6294 }, { "epoch": 0.4462956398440269, "grad_norm": 2.811100647754134, "learning_rate": 6.103967771208297e-06, "loss": 0.4739, "step": 6295 }, { "epoch": 0.44636653668911735, "grad_norm": 1.9911986939852675, "learning_rate": 6.10284791902669e-06, "loss": 0.4114, "step": 6296 }, { "epoch": 0.44643743353420773, "grad_norm": 4.669528712230306, "learning_rate": 6.101728008691072e-06, "loss": 0.3859, "step": 6297 }, { "epoch": 0.4465083303792981, "grad_norm": 3.135507869259564, "learning_rate": 6.100608040260493e-06, "loss": 0.3956, "step": 6298 }, { "epoch": 0.44657922722438853, "grad_norm": 2.275796500252263, "learning_rate": 6.099488013794012e-06, "loss": 0.3778, "step": 6299 }, { "epoch": 0.4466501240694789, "grad_norm": 1.8521137526410063, "learning_rate": 6.09836792935069e-06, "loss": 0.416, "step": 6300 }, { "epoch": 0.4467210209145693, "grad_norm": 3.47865002729167, "learning_rate": 6.09724778698959e-06, "loss": 0.4533, "step": 6301 }, { "epoch": 0.4467919177596597, "grad_norm": 2.039828106117823, "learning_rate": 6.096127586769774e-06, "loss": 0.4296, "step": 6302 }, { "epoch": 0.4468628146047501, "grad_norm": 2.671523654100131, "learning_rate": 6.095007328750317e-06, "loss": 0.4118, "step": 6303 }, { "epoch": 0.4469337114498405, "grad_norm": 6.689716544269652, "learning_rate": 6.093887012990285e-06, "loss": 0.4444, "step": 6304 }, { "epoch": 0.4470046082949309, "grad_norm": 3.0023156481602635, "learning_rate": 6.092766639548757e-06, "loss": 0.4421, "step": 6305 }, { "epoch": 0.44707550514002126, "grad_norm": 2.357105366889589, "learning_rate": 6.09164620848481e-06, "loss": 0.4697, "step": 6306 }, { "epoch": 0.4471464019851117, "grad_norm": 4.134489393625459, "learning_rate": 6.090525719857524e-06, "loss": 0.4198, "step": 6307 }, { "epoch": 0.44721729883020206, "grad_norm": 2.5135938058859, "learning_rate": 6.0894051737259866e-06, "loss": 0.4472, "step": 6308 }, { "epoch": 0.44728819567529243, "grad_norm": 3.853375808112122, "learning_rate": 6.08828457014928e-06, "loss": 0.4213, "step": 6309 }, { "epoch": 0.44735909252038286, "grad_norm": 5.620255932411714, "learning_rate": 6.0871639091865e-06, "loss": 0.3628, "step": 6310 }, { "epoch": 0.44742998936547324, "grad_norm": 2.6192385941920904, "learning_rate": 6.086043190896737e-06, "loss": 0.4345, "step": 6311 }, { "epoch": 0.4475008862105636, "grad_norm": 1.7951485474403504, "learning_rate": 6.084922415339086e-06, "loss": 0.3759, "step": 6312 }, { "epoch": 0.44757178305565404, "grad_norm": 2.6797697560529827, "learning_rate": 6.083801582572649e-06, "loss": 0.3995, "step": 6313 }, { "epoch": 0.4476426799007444, "grad_norm": 4.706058541112822, "learning_rate": 6.082680692656526e-06, "loss": 0.4618, "step": 6314 }, { "epoch": 0.4477135767458348, "grad_norm": 1.8773952781740761, "learning_rate": 6.081559745649825e-06, "loss": 0.4055, "step": 6315 }, { "epoch": 0.4477844735909252, "grad_norm": 3.203867190836182, "learning_rate": 6.080438741611654e-06, "loss": 0.4058, "step": 6316 }, { "epoch": 0.4478553704360156, "grad_norm": 4.097032470308999, "learning_rate": 6.079317680601121e-06, "loss": 0.4191, "step": 6317 }, { "epoch": 0.447926267281106, "grad_norm": 3.9519304979002654, "learning_rate": 6.078196562677346e-06, "loss": 0.4441, "step": 6318 }, { "epoch": 0.4479971641261964, "grad_norm": 3.774776649589039, "learning_rate": 6.077075387899442e-06, "loss": 0.4474, "step": 6319 }, { "epoch": 0.44806806097128676, "grad_norm": 2.867728964858271, "learning_rate": 6.0759541563265304e-06, "loss": 0.4085, "step": 6320 }, { "epoch": 0.4481389578163772, "grad_norm": 3.005445147222388, "learning_rate": 6.0748328680177335e-06, "loss": 0.4174, "step": 6321 }, { "epoch": 0.44820985466146757, "grad_norm": 6.075412393110051, "learning_rate": 6.073711523032181e-06, "loss": 0.45, "step": 6322 }, { "epoch": 0.44828075150655794, "grad_norm": 2.3261055661000194, "learning_rate": 6.072590121429e-06, "loss": 0.3755, "step": 6323 }, { "epoch": 0.44835164835164837, "grad_norm": 3.3758486577248696, "learning_rate": 6.071468663267323e-06, "loss": 0.4342, "step": 6324 }, { "epoch": 0.44842254519673874, "grad_norm": 2.6499427733605474, "learning_rate": 6.070347148606286e-06, "loss": 0.4403, "step": 6325 }, { "epoch": 0.4484934420418291, "grad_norm": 2.10561388775904, "learning_rate": 6.069225577505026e-06, "loss": 0.4533, "step": 6326 }, { "epoch": 0.44856433888691954, "grad_norm": 2.1511437989027202, "learning_rate": 6.068103950022687e-06, "loss": 0.4562, "step": 6327 }, { "epoch": 0.4486352357320099, "grad_norm": 2.1266185196144343, "learning_rate": 6.066982266218409e-06, "loss": 0.4197, "step": 6328 }, { "epoch": 0.4487061325771003, "grad_norm": 3.513297908110209, "learning_rate": 6.065860526151344e-06, "loss": 0.4253, "step": 6329 }, { "epoch": 0.4487770294221907, "grad_norm": 2.48207847087344, "learning_rate": 6.064738729880637e-06, "loss": 0.4348, "step": 6330 }, { "epoch": 0.4488479262672811, "grad_norm": 3.020188137369177, "learning_rate": 6.063616877465447e-06, "loss": 0.4553, "step": 6331 }, { "epoch": 0.4489188231123715, "grad_norm": 6.319006386703077, "learning_rate": 6.062494968964926e-06, "loss": 0.432, "step": 6332 }, { "epoch": 0.4489897199574619, "grad_norm": 2.1949706589568163, "learning_rate": 6.0613730044382345e-06, "loss": 0.3958, "step": 6333 }, { "epoch": 0.44906061680255227, "grad_norm": 2.8572314854773135, "learning_rate": 6.060250983944533e-06, "loss": 0.4302, "step": 6334 }, { "epoch": 0.4491315136476427, "grad_norm": 1.217889559536943, "learning_rate": 6.0591289075429905e-06, "loss": 0.6478, "step": 6335 }, { "epoch": 0.4492024104927331, "grad_norm": 2.8120423462017365, "learning_rate": 6.058006775292772e-06, "loss": 0.4741, "step": 6336 }, { "epoch": 0.44927330733782345, "grad_norm": 2.665097191582749, "learning_rate": 6.0568845872530465e-06, "loss": 0.3989, "step": 6337 }, { "epoch": 0.4493442041829139, "grad_norm": 4.843907792328806, "learning_rate": 6.055762343482992e-06, "loss": 0.4333, "step": 6338 }, { "epoch": 0.44941510102800425, "grad_norm": 2.8323967772081824, "learning_rate": 6.054640044041782e-06, "loss": 0.4198, "step": 6339 }, { "epoch": 0.4494859978730946, "grad_norm": 3.2187285024441654, "learning_rate": 6.0535176889886e-06, "loss": 0.4725, "step": 6340 }, { "epoch": 0.44955689471818505, "grad_norm": 3.7290384002857366, "learning_rate": 6.052395278382622e-06, "loss": 0.3729, "step": 6341 }, { "epoch": 0.4496277915632754, "grad_norm": 3.588006240281654, "learning_rate": 6.05127281228304e-06, "loss": 0.3634, "step": 6342 }, { "epoch": 0.44969868840836585, "grad_norm": 1.01634387910409, "learning_rate": 6.05015029074904e-06, "loss": 0.6618, "step": 6343 }, { "epoch": 0.4497695852534562, "grad_norm": 2.948109744863814, "learning_rate": 6.049027713839815e-06, "loss": 0.4017, "step": 6344 }, { "epoch": 0.4498404820985466, "grad_norm": 4.892706731444971, "learning_rate": 6.047905081614556e-06, "loss": 0.4225, "step": 6345 }, { "epoch": 0.44991137894363703, "grad_norm": 2.4370844150636946, "learning_rate": 6.046782394132462e-06, "loss": 0.372, "step": 6346 }, { "epoch": 0.4499822757887274, "grad_norm": 2.2382268304187334, "learning_rate": 6.045659651452735e-06, "loss": 0.417, "step": 6347 }, { "epoch": 0.4500531726338178, "grad_norm": 3.4814885901890036, "learning_rate": 6.044536853634576e-06, "loss": 0.4173, "step": 6348 }, { "epoch": 0.4501240694789082, "grad_norm": 3.1463805062197427, "learning_rate": 6.04341400073719e-06, "loss": 0.3856, "step": 6349 }, { "epoch": 0.4501949663239986, "grad_norm": 1.077905438745743, "learning_rate": 6.04229109281979e-06, "loss": 0.6445, "step": 6350 }, { "epoch": 0.45026586316908895, "grad_norm": 3.2008392265950687, "learning_rate": 6.041168129941584e-06, "loss": 0.4189, "step": 6351 }, { "epoch": 0.4503367600141794, "grad_norm": 3.6934426619793146, "learning_rate": 6.040045112161787e-06, "loss": 0.4155, "step": 6352 }, { "epoch": 0.45040765685926976, "grad_norm": 2.9545171780500414, "learning_rate": 6.038922039539618e-06, "loss": 0.4081, "step": 6353 }, { "epoch": 0.45047855370436013, "grad_norm": 3.0565144477329396, "learning_rate": 6.0377989121342965e-06, "loss": 0.4441, "step": 6354 }, { "epoch": 0.45054945054945056, "grad_norm": 2.7932184619416947, "learning_rate": 6.0366757300050475e-06, "loss": 0.3867, "step": 6355 }, { "epoch": 0.45062034739454093, "grad_norm": 2.637220540842559, "learning_rate": 6.035552493211094e-06, "loss": 0.4615, "step": 6356 }, { "epoch": 0.45069124423963136, "grad_norm": 1.576220564791717, "learning_rate": 6.034429201811668e-06, "loss": 0.4348, "step": 6357 }, { "epoch": 0.45076214108472173, "grad_norm": 2.8180016819685108, "learning_rate": 6.033305855866001e-06, "loss": 0.4205, "step": 6358 }, { "epoch": 0.4508330379298121, "grad_norm": 2.851974504899385, "learning_rate": 6.032182455433326e-06, "loss": 0.3591, "step": 6359 }, { "epoch": 0.45090393477490254, "grad_norm": 3.54378952630179, "learning_rate": 6.031059000572884e-06, "loss": 0.4263, "step": 6360 }, { "epoch": 0.4509748316199929, "grad_norm": 2.133495269898456, "learning_rate": 6.029935491343915e-06, "loss": 0.4353, "step": 6361 }, { "epoch": 0.4510457284650833, "grad_norm": 2.2807793023404432, "learning_rate": 6.028811927805659e-06, "loss": 0.4241, "step": 6362 }, { "epoch": 0.4511166253101737, "grad_norm": 3.572119403838476, "learning_rate": 6.027688310017367e-06, "loss": 0.4186, "step": 6363 }, { "epoch": 0.4511875221552641, "grad_norm": 5.01733327296076, "learning_rate": 6.026564638038283e-06, "loss": 0.4044, "step": 6364 }, { "epoch": 0.45125841900035446, "grad_norm": 3.994294648762542, "learning_rate": 6.025440911927664e-06, "loss": 0.3688, "step": 6365 }, { "epoch": 0.4513293158454449, "grad_norm": 1.907763513127076, "learning_rate": 6.0243171317447635e-06, "loss": 0.3963, "step": 6366 }, { "epoch": 0.45140021269053526, "grad_norm": 7.445037461361229, "learning_rate": 6.023193297548838e-06, "loss": 0.4371, "step": 6367 }, { "epoch": 0.4514711095356257, "grad_norm": 3.2876806714849742, "learning_rate": 6.0220694093991505e-06, "loss": 0.4467, "step": 6368 }, { "epoch": 0.45154200638071607, "grad_norm": 3.123077733117571, "learning_rate": 6.020945467354962e-06, "loss": 0.4449, "step": 6369 }, { "epoch": 0.45161290322580644, "grad_norm": 3.330793288304786, "learning_rate": 6.019821471475539e-06, "loss": 0.4062, "step": 6370 }, { "epoch": 0.45168380007089687, "grad_norm": 12.898940596519283, "learning_rate": 6.018697421820151e-06, "loss": 0.4253, "step": 6371 }, { "epoch": 0.45175469691598724, "grad_norm": 5.583185076810456, "learning_rate": 6.017573318448071e-06, "loss": 0.4276, "step": 6372 }, { "epoch": 0.4518255937610776, "grad_norm": 2.0895659631240036, "learning_rate": 6.016449161418574e-06, "loss": 0.3451, "step": 6373 }, { "epoch": 0.45189649060616804, "grad_norm": 1.7451899988940178, "learning_rate": 6.015324950790937e-06, "loss": 0.3806, "step": 6374 }, { "epoch": 0.4519673874512584, "grad_norm": 4.126599143274692, "learning_rate": 6.014200686624439e-06, "loss": 0.4071, "step": 6375 }, { "epoch": 0.4520382842963488, "grad_norm": 3.2158821220095386, "learning_rate": 6.013076368978365e-06, "loss": 0.3826, "step": 6376 }, { "epoch": 0.4521091811414392, "grad_norm": 3.484270606286142, "learning_rate": 6.011951997912001e-06, "loss": 0.3831, "step": 6377 }, { "epoch": 0.4521800779865296, "grad_norm": 2.8343193357960494, "learning_rate": 6.010827573484637e-06, "loss": 0.3994, "step": 6378 }, { "epoch": 0.45225097483161997, "grad_norm": 4.202583666155559, "learning_rate": 6.009703095755561e-06, "loss": 0.4399, "step": 6379 }, { "epoch": 0.4523218716767104, "grad_norm": 2.0573858277454558, "learning_rate": 6.008578564784072e-06, "loss": 0.4202, "step": 6380 }, { "epoch": 0.45239276852180077, "grad_norm": 4.458787955295566, "learning_rate": 6.007453980629465e-06, "loss": 0.4524, "step": 6381 }, { "epoch": 0.4524636653668912, "grad_norm": 2.311042877657065, "learning_rate": 6.006329343351042e-06, "loss": 0.4276, "step": 6382 }, { "epoch": 0.4525345622119816, "grad_norm": 1.5197208404390736, "learning_rate": 6.005204653008103e-06, "loss": 0.6815, "step": 6383 }, { "epoch": 0.45260545905707195, "grad_norm": 7.859455540194065, "learning_rate": 6.004079909659955e-06, "loss": 0.3884, "step": 6384 }, { "epoch": 0.4526763559021624, "grad_norm": 2.0571699145275444, "learning_rate": 6.002955113365909e-06, "loss": 0.4026, "step": 6385 }, { "epoch": 0.45274725274725275, "grad_norm": 2.179150083529179, "learning_rate": 6.001830264185272e-06, "loss": 0.4101, "step": 6386 }, { "epoch": 0.4528181495923431, "grad_norm": 2.9192823165188035, "learning_rate": 6.0007053621773625e-06, "loss": 0.3894, "step": 6387 }, { "epoch": 0.45288904643743355, "grad_norm": 2.8514164606756505, "learning_rate": 5.999580407401493e-06, "loss": 0.3859, "step": 6388 }, { "epoch": 0.4529599432825239, "grad_norm": 2.6424381411738747, "learning_rate": 5.9984553999169885e-06, "loss": 0.4364, "step": 6389 }, { "epoch": 0.4530308401276143, "grad_norm": 5.290862515779205, "learning_rate": 5.997330339783167e-06, "loss": 0.4413, "step": 6390 }, { "epoch": 0.4531017369727047, "grad_norm": 3.6105236043986113, "learning_rate": 5.9962052270593566e-06, "loss": 0.3742, "step": 6391 }, { "epoch": 0.4531726338177951, "grad_norm": 2.8530730940989177, "learning_rate": 5.995080061804882e-06, "loss": 0.3842, "step": 6392 }, { "epoch": 0.4532435306628855, "grad_norm": 3.175075351912588, "learning_rate": 5.99395484407908e-06, "loss": 0.434, "step": 6393 }, { "epoch": 0.4533144275079759, "grad_norm": 2.6599251627959886, "learning_rate": 5.992829573941276e-06, "loss": 0.4566, "step": 6394 }, { "epoch": 0.4533853243530663, "grad_norm": 2.948385062857341, "learning_rate": 5.991704251450814e-06, "loss": 0.4139, "step": 6395 }, { "epoch": 0.4534562211981567, "grad_norm": 1.8202409323200002, "learning_rate": 5.990578876667028e-06, "loss": 0.3695, "step": 6396 }, { "epoch": 0.4535271180432471, "grad_norm": 2.140944628724265, "learning_rate": 5.989453449649263e-06, "loss": 0.4142, "step": 6397 }, { "epoch": 0.45359801488833745, "grad_norm": 4.852808234248758, "learning_rate": 5.988327970456862e-06, "loss": 0.4264, "step": 6398 }, { "epoch": 0.4536689117334279, "grad_norm": 2.745564102665982, "learning_rate": 5.987202439149172e-06, "loss": 0.4157, "step": 6399 }, { "epoch": 0.45373980857851826, "grad_norm": 3.755101815449731, "learning_rate": 5.986076855785546e-06, "loss": 0.4128, "step": 6400 }, { "epoch": 0.45381070542360863, "grad_norm": 6.547548191577122, "learning_rate": 5.9849512204253325e-06, "loss": 0.4627, "step": 6401 }, { "epoch": 0.45388160226869906, "grad_norm": 6.824274499386209, "learning_rate": 5.9838255331278905e-06, "loss": 0.4312, "step": 6402 }, { "epoch": 0.45395249911378943, "grad_norm": 2.839039133070023, "learning_rate": 5.9826997939525765e-06, "loss": 0.3933, "step": 6403 }, { "epoch": 0.4540233959588798, "grad_norm": 3.6377252428189832, "learning_rate": 5.981574002958752e-06, "loss": 0.4548, "step": 6404 }, { "epoch": 0.45409429280397023, "grad_norm": 2.6776620006441125, "learning_rate": 5.980448160205782e-06, "loss": 0.3685, "step": 6405 }, { "epoch": 0.4541651896490606, "grad_norm": 6.441260817344004, "learning_rate": 5.979322265753034e-06, "loss": 0.4156, "step": 6406 }, { "epoch": 0.45423608649415104, "grad_norm": 2.414952198700234, "learning_rate": 5.978196319659871e-06, "loss": 0.4577, "step": 6407 }, { "epoch": 0.4543069833392414, "grad_norm": 2.77579733481464, "learning_rate": 5.977070321985673e-06, "loss": 0.3908, "step": 6408 }, { "epoch": 0.4543778801843318, "grad_norm": 4.202523194574116, "learning_rate": 5.975944272789809e-06, "loss": 0.4277, "step": 6409 }, { "epoch": 0.4544487770294222, "grad_norm": 2.390716266363835, "learning_rate": 5.97481817213166e-06, "loss": 0.4476, "step": 6410 }, { "epoch": 0.4545196738745126, "grad_norm": 3.242120710220599, "learning_rate": 5.973692020070605e-06, "loss": 0.4132, "step": 6411 }, { "epoch": 0.45459057071960296, "grad_norm": 2.817623838989906, "learning_rate": 5.972565816666025e-06, "loss": 0.4242, "step": 6412 }, { "epoch": 0.4546614675646934, "grad_norm": 2.13560174240804, "learning_rate": 5.971439561977309e-06, "loss": 0.4172, "step": 6413 }, { "epoch": 0.45473236440978376, "grad_norm": 3.739254817413229, "learning_rate": 5.9703132560638435e-06, "loss": 0.407, "step": 6414 }, { "epoch": 0.45480326125487414, "grad_norm": 1.908259046333264, "learning_rate": 5.969186898985018e-06, "loss": 0.3866, "step": 6415 }, { "epoch": 0.45487415809996457, "grad_norm": 3.4846249793491277, "learning_rate": 5.968060490800227e-06, "loss": 0.3582, "step": 6416 }, { "epoch": 0.45494505494505494, "grad_norm": 4.435617260836646, "learning_rate": 5.96693403156887e-06, "loss": 0.417, "step": 6417 }, { "epoch": 0.4550159517901453, "grad_norm": 7.368607369834597, "learning_rate": 5.965807521350341e-06, "loss": 0.4428, "step": 6418 }, { "epoch": 0.45508684863523574, "grad_norm": 2.832495837951794, "learning_rate": 5.964680960204045e-06, "loss": 0.3615, "step": 6419 }, { "epoch": 0.4551577454803261, "grad_norm": 5.0528855332537566, "learning_rate": 5.9635543481893855e-06, "loss": 0.3794, "step": 6420 }, { "epoch": 0.45522864232541654, "grad_norm": 2.9778827162623682, "learning_rate": 5.962427685365771e-06, "loss": 0.4063, "step": 6421 }, { "epoch": 0.4552995391705069, "grad_norm": 3.102049907032933, "learning_rate": 5.96130097179261e-06, "loss": 0.418, "step": 6422 }, { "epoch": 0.4553704360155973, "grad_norm": 2.956865102104056, "learning_rate": 5.960174207529314e-06, "loss": 0.4192, "step": 6423 }, { "epoch": 0.4554413328606877, "grad_norm": 1.8769287455196255, "learning_rate": 5.9590473926352984e-06, "loss": 0.4157, "step": 6424 }, { "epoch": 0.4555122297057781, "grad_norm": 4.024546454488852, "learning_rate": 5.957920527169982e-06, "loss": 0.4258, "step": 6425 }, { "epoch": 0.45558312655086847, "grad_norm": 1.822648621234324, "learning_rate": 5.956793611192786e-06, "loss": 0.4008, "step": 6426 }, { "epoch": 0.4556540233959589, "grad_norm": 16.96316802897396, "learning_rate": 5.955666644763134e-06, "loss": 0.4359, "step": 6427 }, { "epoch": 0.45572492024104927, "grad_norm": 3.590410135365956, "learning_rate": 5.954539627940447e-06, "loss": 0.4694, "step": 6428 }, { "epoch": 0.45579581708613964, "grad_norm": 3.9149500920539855, "learning_rate": 5.95341256078416e-06, "loss": 0.3976, "step": 6429 }, { "epoch": 0.45586671393123007, "grad_norm": 4.447681656311218, "learning_rate": 5.952285443353699e-06, "loss": 0.3665, "step": 6430 }, { "epoch": 0.45593761077632045, "grad_norm": 3.7474780665766323, "learning_rate": 5.951158275708501e-06, "loss": 0.409, "step": 6431 }, { "epoch": 0.4560085076214109, "grad_norm": 4.831703327857799, "learning_rate": 5.950031057908001e-06, "loss": 0.4137, "step": 6432 }, { "epoch": 0.45607940446650125, "grad_norm": 3.221851684616268, "learning_rate": 5.948903790011636e-06, "loss": 0.4143, "step": 6433 }, { "epoch": 0.4561503013115916, "grad_norm": 4.683876140436455, "learning_rate": 5.947776472078852e-06, "loss": 0.4388, "step": 6434 }, { "epoch": 0.45622119815668205, "grad_norm": 29.0026713657131, "learning_rate": 5.946649104169091e-06, "loss": 0.4254, "step": 6435 }, { "epoch": 0.4562920950017724, "grad_norm": 3.6550479221629697, "learning_rate": 5.9455216863418e-06, "loss": 0.3926, "step": 6436 }, { "epoch": 0.4563629918468628, "grad_norm": 8.830976820341103, "learning_rate": 5.944394218656429e-06, "loss": 0.3691, "step": 6437 }, { "epoch": 0.4564338886919532, "grad_norm": 3.4077549053544667, "learning_rate": 5.94326670117243e-06, "loss": 0.3737, "step": 6438 }, { "epoch": 0.4565047855370436, "grad_norm": 5.9308305879156045, "learning_rate": 5.942139133949256e-06, "loss": 0.4065, "step": 6439 }, { "epoch": 0.456575682382134, "grad_norm": 2.0899383667129174, "learning_rate": 5.94101151704637e-06, "loss": 0.4028, "step": 6440 }, { "epoch": 0.4566465792272244, "grad_norm": 5.029282721458251, "learning_rate": 5.939883850523226e-06, "loss": 0.4558, "step": 6441 }, { "epoch": 0.4567174760723148, "grad_norm": 5.108407279236821, "learning_rate": 5.93875613443929e-06, "loss": 0.4437, "step": 6442 }, { "epoch": 0.45678837291740515, "grad_norm": 2.742535414158367, "learning_rate": 5.937628368854026e-06, "loss": 0.4197, "step": 6443 }, { "epoch": 0.4568592697624956, "grad_norm": 7.685134370148257, "learning_rate": 5.936500553826903e-06, "loss": 0.4097, "step": 6444 }, { "epoch": 0.45693016660758595, "grad_norm": 2.582689842437986, "learning_rate": 5.935372689417391e-06, "loss": 0.3914, "step": 6445 }, { "epoch": 0.4570010634526764, "grad_norm": 2.3501964594492497, "learning_rate": 5.934244775684961e-06, "loss": 0.4254, "step": 6446 }, { "epoch": 0.45707196029776676, "grad_norm": 2.288418391803619, "learning_rate": 5.933116812689094e-06, "loss": 0.4513, "step": 6447 }, { "epoch": 0.45714285714285713, "grad_norm": 2.11468826749536, "learning_rate": 5.931988800489263e-06, "loss": 0.3904, "step": 6448 }, { "epoch": 0.45721375398794756, "grad_norm": 11.934394010554039, "learning_rate": 5.930860739144952e-06, "loss": 0.43, "step": 6449 }, { "epoch": 0.45728465083303793, "grad_norm": 2.511216135496121, "learning_rate": 5.929732628715644e-06, "loss": 0.3578, "step": 6450 }, { "epoch": 0.4573555476781283, "grad_norm": 2.5512687431349055, "learning_rate": 5.9286044692608255e-06, "loss": 0.4083, "step": 6451 }, { "epoch": 0.45742644452321873, "grad_norm": 14.527089523718084, "learning_rate": 5.9274762608399825e-06, "loss": 0.4276, "step": 6452 }, { "epoch": 0.4574973413683091, "grad_norm": 3.0673070971452123, "learning_rate": 5.926348003512611e-06, "loss": 0.4381, "step": 6453 }, { "epoch": 0.4575682382133995, "grad_norm": 2.6572140977527092, "learning_rate": 5.9252196973382e-06, "loss": 0.4058, "step": 6454 }, { "epoch": 0.4576391350584899, "grad_norm": 1.9498768883834685, "learning_rate": 5.92409134237625e-06, "loss": 0.3719, "step": 6455 }, { "epoch": 0.4577100319035803, "grad_norm": 5.4951144327108565, "learning_rate": 5.922962938686257e-06, "loss": 0.4106, "step": 6456 }, { "epoch": 0.45778092874867066, "grad_norm": 3.4808709001404607, "learning_rate": 5.9218344863277236e-06, "loss": 0.3889, "step": 6457 }, { "epoch": 0.4578518255937611, "grad_norm": 7.377131756551472, "learning_rate": 5.920705985360154e-06, "loss": 0.4279, "step": 6458 }, { "epoch": 0.45792272243885146, "grad_norm": 4.093705761253048, "learning_rate": 5.919577435843057e-06, "loss": 0.4198, "step": 6459 }, { "epoch": 0.4579936192839419, "grad_norm": 4.8935078896897055, "learning_rate": 5.9184488378359385e-06, "loss": 0.4199, "step": 6460 }, { "epoch": 0.45806451612903226, "grad_norm": 2.5883095995901586, "learning_rate": 5.9173201913983115e-06, "loss": 0.4086, "step": 6461 }, { "epoch": 0.45813541297412264, "grad_norm": 2.337493447523663, "learning_rate": 5.916191496589691e-06, "loss": 0.4504, "step": 6462 }, { "epoch": 0.45820630981921306, "grad_norm": 69.23304740730102, "learning_rate": 5.915062753469594e-06, "loss": 0.4418, "step": 6463 }, { "epoch": 0.45827720666430344, "grad_norm": 4.286759688929057, "learning_rate": 5.913933962097539e-06, "loss": 0.424, "step": 6464 }, { "epoch": 0.4583481035093938, "grad_norm": 2.208710635704432, "learning_rate": 5.912805122533047e-06, "loss": 0.4246, "step": 6465 }, { "epoch": 0.45841900035448424, "grad_norm": 2.905525910087631, "learning_rate": 5.911676234835647e-06, "loss": 0.4344, "step": 6466 }, { "epoch": 0.4584898971995746, "grad_norm": 9.672090217042411, "learning_rate": 5.9105472990648625e-06, "loss": 0.3902, "step": 6467 }, { "epoch": 0.458560794044665, "grad_norm": 3.414389859038868, "learning_rate": 5.909418315280223e-06, "loss": 0.4599, "step": 6468 }, { "epoch": 0.4586316908897554, "grad_norm": 3.654768311130717, "learning_rate": 5.908289283541262e-06, "loss": 0.4608, "step": 6469 }, { "epoch": 0.4587025877348458, "grad_norm": 1.5757232871828146, "learning_rate": 5.907160203907514e-06, "loss": 0.3769, "step": 6470 }, { "epoch": 0.4587734845799362, "grad_norm": 3.7336668890916154, "learning_rate": 5.906031076438516e-06, "loss": 0.4274, "step": 6471 }, { "epoch": 0.4588443814250266, "grad_norm": 5.484154694135516, "learning_rate": 5.904901901193809e-06, "loss": 0.3719, "step": 6472 }, { "epoch": 0.45891527827011697, "grad_norm": 3.1871554107288813, "learning_rate": 5.903772678232933e-06, "loss": 0.3991, "step": 6473 }, { "epoch": 0.4589861751152074, "grad_norm": 2.643901261659843, "learning_rate": 5.9026434076154334e-06, "loss": 0.4309, "step": 6474 }, { "epoch": 0.45905707196029777, "grad_norm": 2.345473963272652, "learning_rate": 5.901514089400858e-06, "loss": 0.4275, "step": 6475 }, { "epoch": 0.45912796880538814, "grad_norm": 2.367104219060825, "learning_rate": 5.900384723648757e-06, "loss": 0.4053, "step": 6476 }, { "epoch": 0.45919886565047857, "grad_norm": 3.259793592875148, "learning_rate": 5.899255310418684e-06, "loss": 0.4278, "step": 6477 }, { "epoch": 0.45926976249556895, "grad_norm": 2.627060320893947, "learning_rate": 5.898125849770188e-06, "loss": 0.3821, "step": 6478 }, { "epoch": 0.4593406593406593, "grad_norm": 1.9958730159783244, "learning_rate": 5.896996341762834e-06, "loss": 0.3894, "step": 6479 }, { "epoch": 0.45941155618574975, "grad_norm": 2.3181543302103904, "learning_rate": 5.8958667864561765e-06, "loss": 0.3881, "step": 6480 }, { "epoch": 0.4594824530308401, "grad_norm": 2.1233643281903576, "learning_rate": 5.89473718390978e-06, "loss": 0.4132, "step": 6481 }, { "epoch": 0.4595533498759305, "grad_norm": 3.3583587408294218, "learning_rate": 5.893607534183209e-06, "loss": 0.4777, "step": 6482 }, { "epoch": 0.4596242467210209, "grad_norm": 9.146661259784494, "learning_rate": 5.8924778373360306e-06, "loss": 0.4161, "step": 6483 }, { "epoch": 0.4596951435661113, "grad_norm": 4.241270273102516, "learning_rate": 5.891348093427815e-06, "loss": 0.3933, "step": 6484 }, { "epoch": 0.4597660404112017, "grad_norm": 2.9749462471156285, "learning_rate": 5.890218302518134e-06, "loss": 0.3745, "step": 6485 }, { "epoch": 0.4598369372562921, "grad_norm": 3.502871236111512, "learning_rate": 5.8890884646665605e-06, "loss": 0.3874, "step": 6486 }, { "epoch": 0.4599078341013825, "grad_norm": 2.1937310292699723, "learning_rate": 5.8879585799326764e-06, "loss": 0.4109, "step": 6487 }, { "epoch": 0.4599787309464729, "grad_norm": 4.76149929771379, "learning_rate": 5.886828648376057e-06, "loss": 0.4036, "step": 6488 }, { "epoch": 0.4600496277915633, "grad_norm": 2.20186693449444, "learning_rate": 5.885698670056287e-06, "loss": 0.4474, "step": 6489 }, { "epoch": 0.46012052463665365, "grad_norm": 2.752407341686958, "learning_rate": 5.88456864503295e-06, "loss": 0.3956, "step": 6490 }, { "epoch": 0.4601914214817441, "grad_norm": 2.9452986553313845, "learning_rate": 5.8834385733656314e-06, "loss": 0.3941, "step": 6491 }, { "epoch": 0.46026231832683445, "grad_norm": 3.9950185760803962, "learning_rate": 5.882308455113925e-06, "loss": 0.3919, "step": 6492 }, { "epoch": 0.4603332151719248, "grad_norm": 3.0513902698643385, "learning_rate": 5.88117829033742e-06, "loss": 0.447, "step": 6493 }, { "epoch": 0.46040411201701525, "grad_norm": 3.6492705509780503, "learning_rate": 5.8800480790957105e-06, "loss": 0.4147, "step": 6494 }, { "epoch": 0.46047500886210563, "grad_norm": 2.6903490719073595, "learning_rate": 5.878917821448395e-06, "loss": 0.4104, "step": 6495 }, { "epoch": 0.46054590570719606, "grad_norm": 4.449794739932028, "learning_rate": 5.877787517455071e-06, "loss": 0.4084, "step": 6496 }, { "epoch": 0.46061680255228643, "grad_norm": 3.6052493482828707, "learning_rate": 5.87665716717534e-06, "loss": 0.4128, "step": 6497 }, { "epoch": 0.4606876993973768, "grad_norm": 2.974809409303497, "learning_rate": 5.87552677066881e-06, "loss": 0.4361, "step": 6498 }, { "epoch": 0.46075859624246723, "grad_norm": 3.3230423885865323, "learning_rate": 5.874396327995082e-06, "loss": 0.3961, "step": 6499 }, { "epoch": 0.4608294930875576, "grad_norm": 4.940784557098265, "learning_rate": 5.873265839213771e-06, "loss": 0.4134, "step": 6500 }, { "epoch": 0.460900389932648, "grad_norm": 3.1184086063511516, "learning_rate": 5.8721353043844854e-06, "loss": 0.4606, "step": 6501 }, { "epoch": 0.4609712867777384, "grad_norm": 2.122172927234987, "learning_rate": 5.871004723566839e-06, "loss": 0.4281, "step": 6502 }, { "epoch": 0.4610421836228288, "grad_norm": 5.139931201013526, "learning_rate": 5.869874096820448e-06, "loss": 0.3944, "step": 6503 }, { "epoch": 0.46111308046791916, "grad_norm": 2.529720916795999, "learning_rate": 5.868743424204934e-06, "loss": 0.4353, "step": 6504 }, { "epoch": 0.4611839773130096, "grad_norm": 5.581937223011334, "learning_rate": 5.867612705779915e-06, "loss": 0.35, "step": 6505 }, { "epoch": 0.46125487415809996, "grad_norm": 2.527632684462726, "learning_rate": 5.8664819416050146e-06, "loss": 0.4296, "step": 6506 }, { "epoch": 0.46132577100319033, "grad_norm": 3.3987547375300857, "learning_rate": 5.86535113173986e-06, "loss": 0.43, "step": 6507 }, { "epoch": 0.46139666784828076, "grad_norm": 3.8479640364107017, "learning_rate": 5.864220276244082e-06, "loss": 0.4154, "step": 6508 }, { "epoch": 0.46146756469337114, "grad_norm": 1.7532867161864587, "learning_rate": 5.8630893751773085e-06, "loss": 0.3813, "step": 6509 }, { "epoch": 0.46153846153846156, "grad_norm": 3.1926111580289827, "learning_rate": 5.861958428599171e-06, "loss": 0.3984, "step": 6510 }, { "epoch": 0.46160935838355194, "grad_norm": 1.8579600918033314, "learning_rate": 5.860827436569312e-06, "loss": 0.3659, "step": 6511 }, { "epoch": 0.4616802552286423, "grad_norm": 1.9946607336352968, "learning_rate": 5.859696399147361e-06, "loss": 0.3962, "step": 6512 }, { "epoch": 0.46175115207373274, "grad_norm": 2.4823423735345203, "learning_rate": 5.858565316392966e-06, "loss": 0.426, "step": 6513 }, { "epoch": 0.4618220489188231, "grad_norm": 2.668475997099804, "learning_rate": 5.857434188365765e-06, "loss": 0.4239, "step": 6514 }, { "epoch": 0.4618929457639135, "grad_norm": 2.6852277480531566, "learning_rate": 5.856303015125404e-06, "loss": 0.4503, "step": 6515 }, { "epoch": 0.4619638426090039, "grad_norm": 4.350369963169682, "learning_rate": 5.8551717967315324e-06, "loss": 0.4029, "step": 6516 }, { "epoch": 0.4620347394540943, "grad_norm": 6.009671957502745, "learning_rate": 5.8540405332437995e-06, "loss": 0.3668, "step": 6517 }, { "epoch": 0.46210563629918466, "grad_norm": 4.639479233493881, "learning_rate": 5.852909224721857e-06, "loss": 0.397, "step": 6518 }, { "epoch": 0.4621765331442751, "grad_norm": 2.761303229757619, "learning_rate": 5.85177787122536e-06, "loss": 0.4331, "step": 6519 }, { "epoch": 0.46224742998936547, "grad_norm": 3.4502821841087736, "learning_rate": 5.850646472813967e-06, "loss": 0.4235, "step": 6520 }, { "epoch": 0.46231832683445584, "grad_norm": 3.960716308016247, "learning_rate": 5.849515029547335e-06, "loss": 0.4415, "step": 6521 }, { "epoch": 0.46238922367954627, "grad_norm": 2.0368023525006937, "learning_rate": 5.848383541485127e-06, "loss": 0.4293, "step": 6522 }, { "epoch": 0.46246012052463664, "grad_norm": 6.0193980508558145, "learning_rate": 5.847252008687007e-06, "loss": 0.3572, "step": 6523 }, { "epoch": 0.46253101736972707, "grad_norm": 0.927439234573415, "learning_rate": 5.846120431212643e-06, "loss": 0.6719, "step": 6524 }, { "epoch": 0.46260191421481744, "grad_norm": 3.1536503050404527, "learning_rate": 5.844988809121702e-06, "loss": 0.4006, "step": 6525 }, { "epoch": 0.4626728110599078, "grad_norm": 2.6794986479565277, "learning_rate": 5.843857142473856e-06, "loss": 0.4292, "step": 6526 }, { "epoch": 0.46274370790499825, "grad_norm": 3.3071066369237143, "learning_rate": 5.84272543132878e-06, "loss": 0.3931, "step": 6527 }, { "epoch": 0.4628146047500886, "grad_norm": 3.0760777541649427, "learning_rate": 5.841593675746148e-06, "loss": 0.4325, "step": 6528 }, { "epoch": 0.462885501595179, "grad_norm": 3.7616369972992856, "learning_rate": 5.8404618757856376e-06, "loss": 0.417, "step": 6529 }, { "epoch": 0.4629563984402694, "grad_norm": 1.991568046622942, "learning_rate": 5.839330031506934e-06, "loss": 0.4491, "step": 6530 }, { "epoch": 0.4630272952853598, "grad_norm": 4.632777432760801, "learning_rate": 5.838198142969714e-06, "loss": 0.4324, "step": 6531 }, { "epoch": 0.46309819213045017, "grad_norm": 2.006783370563514, "learning_rate": 5.83706621023367e-06, "loss": 0.4075, "step": 6532 }, { "epoch": 0.4631690889755406, "grad_norm": 6.580176198287074, "learning_rate": 5.8359342333584824e-06, "loss": 0.478, "step": 6533 }, { "epoch": 0.463239985820631, "grad_norm": 2.508430472598018, "learning_rate": 5.834802212403846e-06, "loss": 0.4082, "step": 6534 }, { "epoch": 0.4633108826657214, "grad_norm": 2.5945513928712467, "learning_rate": 5.8336701474294524e-06, "loss": 0.4037, "step": 6535 }, { "epoch": 0.4633817795108118, "grad_norm": 4.671733621870014, "learning_rate": 5.832538038494994e-06, "loss": 0.4209, "step": 6536 }, { "epoch": 0.46345267635590215, "grad_norm": 2.4273503371542646, "learning_rate": 5.831405885660172e-06, "loss": 0.4213, "step": 6537 }, { "epoch": 0.4635235732009926, "grad_norm": 1.072060792177539, "learning_rate": 5.83027368898468e-06, "loss": 0.7162, "step": 6538 }, { "epoch": 0.46359447004608295, "grad_norm": 4.5660638563452896, "learning_rate": 5.8291414485282235e-06, "loss": 0.4036, "step": 6539 }, { "epoch": 0.4636653668911733, "grad_norm": 3.8312616029985915, "learning_rate": 5.828009164350506e-06, "loss": 0.4116, "step": 6540 }, { "epoch": 0.46373626373626375, "grad_norm": 2.507052991669049, "learning_rate": 5.826876836511234e-06, "loss": 0.4496, "step": 6541 }, { "epoch": 0.4638071605813541, "grad_norm": 5.168510409856455, "learning_rate": 5.825744465070115e-06, "loss": 0.3703, "step": 6542 }, { "epoch": 0.4638780574264445, "grad_norm": 4.540789947598253, "learning_rate": 5.824612050086859e-06, "loss": 0.3978, "step": 6543 }, { "epoch": 0.46394895427153493, "grad_norm": 4.446782908658184, "learning_rate": 5.8234795916211805e-06, "loss": 0.4111, "step": 6544 }, { "epoch": 0.4640198511166253, "grad_norm": 2.0404719115561756, "learning_rate": 5.822347089732796e-06, "loss": 0.3977, "step": 6545 }, { "epoch": 0.4640907479617157, "grad_norm": 3.3087403794502626, "learning_rate": 5.821214544481422e-06, "loss": 0.4328, "step": 6546 }, { "epoch": 0.4641616448068061, "grad_norm": 12.45340341399952, "learning_rate": 5.820081955926776e-06, "loss": 0.4464, "step": 6547 }, { "epoch": 0.4642325416518965, "grad_norm": 3.4908687379904304, "learning_rate": 5.818949324128583e-06, "loss": 0.3822, "step": 6548 }, { "epoch": 0.4643034384969869, "grad_norm": 2.685383678145482, "learning_rate": 5.817816649146568e-06, "loss": 0.3997, "step": 6549 }, { "epoch": 0.4643743353420773, "grad_norm": 2.871866605500912, "learning_rate": 5.816683931040458e-06, "loss": 0.4119, "step": 6550 }, { "epoch": 0.46444523218716766, "grad_norm": 2.6865733259169353, "learning_rate": 5.815551169869978e-06, "loss": 0.5047, "step": 6551 }, { "epoch": 0.4645161290322581, "grad_norm": 2.193622723144028, "learning_rate": 5.8144183656948645e-06, "loss": 0.4156, "step": 6552 }, { "epoch": 0.46458702587734846, "grad_norm": 1.5562131213042492, "learning_rate": 5.813285518574848e-06, "loss": 0.4104, "step": 6553 }, { "epoch": 0.46465792272243883, "grad_norm": 4.125660205039695, "learning_rate": 5.8121526285696675e-06, "loss": 0.4139, "step": 6554 }, { "epoch": 0.46472881956752926, "grad_norm": 2.450285866181186, "learning_rate": 5.811019695739056e-06, "loss": 0.4071, "step": 6555 }, { "epoch": 0.46479971641261963, "grad_norm": 3.2335871447224247, "learning_rate": 5.809886720142757e-06, "loss": 0.4235, "step": 6556 }, { "epoch": 0.46487061325771, "grad_norm": 5.830884291026778, "learning_rate": 5.808753701840514e-06, "loss": 0.4484, "step": 6557 }, { "epoch": 0.46494151010280044, "grad_norm": 2.2016594597806014, "learning_rate": 5.807620640892072e-06, "loss": 0.3824, "step": 6558 }, { "epoch": 0.4650124069478908, "grad_norm": 3.25328831089141, "learning_rate": 5.806487537357176e-06, "loss": 0.4204, "step": 6559 }, { "epoch": 0.46508330379298124, "grad_norm": 2.508789326484156, "learning_rate": 5.805354391295577e-06, "loss": 0.3797, "step": 6560 }, { "epoch": 0.4651542006380716, "grad_norm": 1.9001161646723723, "learning_rate": 5.804221202767026e-06, "loss": 0.3571, "step": 6561 }, { "epoch": 0.465225097483162, "grad_norm": 4.032296935138821, "learning_rate": 5.803087971831278e-06, "loss": 0.4238, "step": 6562 }, { "epoch": 0.4652959943282524, "grad_norm": 3.1764648356335807, "learning_rate": 5.8019546985480865e-06, "loss": 0.3718, "step": 6563 }, { "epoch": 0.4653668911733428, "grad_norm": 1.8932791701415577, "learning_rate": 5.8008213829772146e-06, "loss": 0.3953, "step": 6564 }, { "epoch": 0.46543778801843316, "grad_norm": 2.0999994572746945, "learning_rate": 5.799688025178418e-06, "loss": 0.4144, "step": 6565 }, { "epoch": 0.4655086848635236, "grad_norm": 2.2628313778212537, "learning_rate": 5.798554625211461e-06, "loss": 0.4169, "step": 6566 }, { "epoch": 0.46557958170861397, "grad_norm": 2.915830939472831, "learning_rate": 5.79742118313611e-06, "loss": 0.3913, "step": 6567 }, { "epoch": 0.46565047855370434, "grad_norm": 2.4200690353693712, "learning_rate": 5.796287699012131e-06, "loss": 0.4122, "step": 6568 }, { "epoch": 0.46572137539879477, "grad_norm": 3.9123775366531666, "learning_rate": 5.795154172899294e-06, "loss": 0.4356, "step": 6569 }, { "epoch": 0.46579227224388514, "grad_norm": 2.3277693743683985, "learning_rate": 5.794020604857371e-06, "loss": 0.36, "step": 6570 }, { "epoch": 0.4658631690889755, "grad_norm": 2.8602409362612002, "learning_rate": 5.792886994946134e-06, "loss": 0.4462, "step": 6571 }, { "epoch": 0.46593406593406594, "grad_norm": 2.0948134775166096, "learning_rate": 5.791753343225362e-06, "loss": 0.3888, "step": 6572 }, { "epoch": 0.4660049627791563, "grad_norm": 3.4203401004660887, "learning_rate": 5.7906196497548306e-06, "loss": 0.4435, "step": 6573 }, { "epoch": 0.46607585962424675, "grad_norm": 1.9148097214717559, "learning_rate": 5.789485914594323e-06, "loss": 0.3479, "step": 6574 }, { "epoch": 0.4661467564693371, "grad_norm": 4.357690915121403, "learning_rate": 5.78835213780362e-06, "loss": 0.4317, "step": 6575 }, { "epoch": 0.4662176533144275, "grad_norm": 2.5038803402945287, "learning_rate": 5.787218319442506e-06, "loss": 0.4233, "step": 6576 }, { "epoch": 0.4662885501595179, "grad_norm": 4.838224502688288, "learning_rate": 5.786084459570771e-06, "loss": 0.4067, "step": 6577 }, { "epoch": 0.4663594470046083, "grad_norm": 3.613476294430824, "learning_rate": 5.7849505582482e-06, "loss": 0.3921, "step": 6578 }, { "epoch": 0.46643034384969867, "grad_norm": 5.295781892170858, "learning_rate": 5.78381661553459e-06, "loss": 0.4183, "step": 6579 }, { "epoch": 0.4665012406947891, "grad_norm": 5.902985585081151, "learning_rate": 5.78268263148973e-06, "loss": 0.3825, "step": 6580 }, { "epoch": 0.4665721375398795, "grad_norm": 2.6716256314820375, "learning_rate": 5.781548606173417e-06, "loss": 0.3913, "step": 6581 }, { "epoch": 0.46664303438496985, "grad_norm": 2.2207661909926797, "learning_rate": 5.780414539645451e-06, "loss": 0.4188, "step": 6582 }, { "epoch": 0.4667139312300603, "grad_norm": 2.2648627778402526, "learning_rate": 5.77928043196563e-06, "loss": 0.3886, "step": 6583 }, { "epoch": 0.46678482807515065, "grad_norm": 4.088269765317342, "learning_rate": 5.778146283193756e-06, "loss": 0.4925, "step": 6584 }, { "epoch": 0.466855724920241, "grad_norm": 1.1303896852714335, "learning_rate": 5.777012093389636e-06, "loss": 0.6862, "step": 6585 }, { "epoch": 0.46692662176533145, "grad_norm": 2.578665280269674, "learning_rate": 5.775877862613074e-06, "loss": 0.354, "step": 6586 }, { "epoch": 0.4669975186104218, "grad_norm": 3.0500405180778154, "learning_rate": 5.77474359092388e-06, "loss": 0.3954, "step": 6587 }, { "epoch": 0.46706841545551225, "grad_norm": 2.810350612677862, "learning_rate": 5.773609278381866e-06, "loss": 0.4196, "step": 6588 }, { "epoch": 0.4671393123006026, "grad_norm": 4.24171302222257, "learning_rate": 5.772474925046844e-06, "loss": 0.3614, "step": 6589 }, { "epoch": 0.467210209145693, "grad_norm": 2.782392375734976, "learning_rate": 5.771340530978631e-06, "loss": 0.4354, "step": 6590 }, { "epoch": 0.46728110599078343, "grad_norm": 4.195252084133469, "learning_rate": 5.770206096237041e-06, "loss": 0.4389, "step": 6591 }, { "epoch": 0.4673520028358738, "grad_norm": 4.300863305981838, "learning_rate": 5.769071620881896e-06, "loss": 0.4484, "step": 6592 }, { "epoch": 0.4674228996809642, "grad_norm": 7.785318405742378, "learning_rate": 5.767937104973018e-06, "loss": 0.4265, "step": 6593 }, { "epoch": 0.4674937965260546, "grad_norm": 4.197722145239323, "learning_rate": 5.766802548570229e-06, "loss": 0.4124, "step": 6594 }, { "epoch": 0.467564693371145, "grad_norm": 2.076310558537186, "learning_rate": 5.765667951733358e-06, "loss": 0.4161, "step": 6595 }, { "epoch": 0.46763559021623535, "grad_norm": 3.063847102624944, "learning_rate": 5.764533314522231e-06, "loss": 0.425, "step": 6596 }, { "epoch": 0.4677064870613258, "grad_norm": 3.702746050483697, "learning_rate": 5.76339863699668e-06, "loss": 0.4147, "step": 6597 }, { "epoch": 0.46777738390641616, "grad_norm": 3.5122341504067265, "learning_rate": 5.762263919216534e-06, "loss": 0.4431, "step": 6598 }, { "epoch": 0.4678482807515066, "grad_norm": 3.106544585985087, "learning_rate": 5.7611291612416305e-06, "loss": 0.4466, "step": 6599 }, { "epoch": 0.46791917759659696, "grad_norm": 1.168730388156946, "learning_rate": 5.759994363131806e-06, "loss": 0.6985, "step": 6600 }, { "epoch": 0.46799007444168733, "grad_norm": 3.888865859135301, "learning_rate": 5.758859524946898e-06, "loss": 0.398, "step": 6601 }, { "epoch": 0.46806097128677776, "grad_norm": 2.309358486980621, "learning_rate": 5.757724646746747e-06, "loss": 0.387, "step": 6602 }, { "epoch": 0.46813186813186813, "grad_norm": 7.176090978868378, "learning_rate": 5.7565897285911994e-06, "loss": 0.4332, "step": 6603 }, { "epoch": 0.4682027649769585, "grad_norm": 3.5550462649455117, "learning_rate": 5.755454770540097e-06, "loss": 0.394, "step": 6604 }, { "epoch": 0.46827366182204894, "grad_norm": 3.088035444120877, "learning_rate": 5.754319772653287e-06, "loss": 0.4271, "step": 6605 }, { "epoch": 0.4683445586671393, "grad_norm": 2.3270168731366843, "learning_rate": 5.753184734990621e-06, "loss": 0.4654, "step": 6606 }, { "epoch": 0.4684154555122297, "grad_norm": 3.4773753522801587, "learning_rate": 5.752049657611948e-06, "loss": 0.4267, "step": 6607 }, { "epoch": 0.4684863523573201, "grad_norm": 4.3071296401877515, "learning_rate": 5.750914540577123e-06, "loss": 0.482, "step": 6608 }, { "epoch": 0.4685572492024105, "grad_norm": 3.5363981013108874, "learning_rate": 5.749779383946002e-06, "loss": 0.4612, "step": 6609 }, { "epoch": 0.46862814604750086, "grad_norm": 4.656974310209833, "learning_rate": 5.748644187778441e-06, "loss": 0.4562, "step": 6610 }, { "epoch": 0.4686990428925913, "grad_norm": 1.7242610021914264, "learning_rate": 5.7475089521342995e-06, "loss": 0.3955, "step": 6611 }, { "epoch": 0.46876993973768166, "grad_norm": 18.20583652760366, "learning_rate": 5.746373677073443e-06, "loss": 0.4136, "step": 6612 }, { "epoch": 0.4688408365827721, "grad_norm": 3.0630357889321647, "learning_rate": 5.745238362655729e-06, "loss": 0.389, "step": 6613 }, { "epoch": 0.46891173342786246, "grad_norm": 3.1293285335997436, "learning_rate": 5.744103008941031e-06, "loss": 0.4192, "step": 6614 }, { "epoch": 0.46898263027295284, "grad_norm": 2.4698786747808716, "learning_rate": 5.7429676159892115e-06, "loss": 0.3638, "step": 6615 }, { "epoch": 0.46905352711804327, "grad_norm": 4.918574314994574, "learning_rate": 5.7418321838601435e-06, "loss": 0.4423, "step": 6616 }, { "epoch": 0.46912442396313364, "grad_norm": 3.321516366750308, "learning_rate": 5.740696712613697e-06, "loss": 0.4465, "step": 6617 }, { "epoch": 0.469195320808224, "grad_norm": 5.972010433028452, "learning_rate": 5.739561202309748e-06, "loss": 0.3937, "step": 6618 }, { "epoch": 0.46926621765331444, "grad_norm": 3.1603408175638132, "learning_rate": 5.738425653008173e-06, "loss": 0.4334, "step": 6619 }, { "epoch": 0.4693371144984048, "grad_norm": 5.66936683466154, "learning_rate": 5.73729006476885e-06, "loss": 0.456, "step": 6620 }, { "epoch": 0.4694080113434952, "grad_norm": 1.5536610035420608, "learning_rate": 5.736154437651656e-06, "loss": 0.4207, "step": 6621 }, { "epoch": 0.4694789081885856, "grad_norm": 3.4965063851603997, "learning_rate": 5.73501877171648e-06, "loss": 0.4385, "step": 6622 }, { "epoch": 0.469549805033676, "grad_norm": 2.3946433383089483, "learning_rate": 5.733883067023202e-06, "loss": 0.3964, "step": 6623 }, { "epoch": 0.4696207018787664, "grad_norm": 2.283548297157251, "learning_rate": 5.7327473236317096e-06, "loss": 0.4195, "step": 6624 }, { "epoch": 0.4696915987238568, "grad_norm": 3.044348825029181, "learning_rate": 5.731611541601893e-06, "loss": 0.3999, "step": 6625 }, { "epoch": 0.46976249556894717, "grad_norm": 3.324006196493643, "learning_rate": 5.730475720993639e-06, "loss": 0.4281, "step": 6626 }, { "epoch": 0.4698333924140376, "grad_norm": 1.663421401589331, "learning_rate": 5.729339861866844e-06, "loss": 0.3961, "step": 6627 }, { "epoch": 0.46990428925912797, "grad_norm": 2.901746984927909, "learning_rate": 5.728203964281402e-06, "loss": 0.383, "step": 6628 }, { "epoch": 0.46997518610421835, "grad_norm": 1.9264214432875362, "learning_rate": 5.727068028297208e-06, "loss": 0.3788, "step": 6629 }, { "epoch": 0.4700460829493088, "grad_norm": 2.19600839492433, "learning_rate": 5.725932053974162e-06, "loss": 0.4048, "step": 6630 }, { "epoch": 0.47011697979439915, "grad_norm": 2.5074201328996764, "learning_rate": 5.724796041372164e-06, "loss": 0.431, "step": 6631 }, { "epoch": 0.4701878766394895, "grad_norm": 5.56305885330821, "learning_rate": 5.72365999055112e-06, "loss": 0.4146, "step": 6632 }, { "epoch": 0.47025877348457995, "grad_norm": 5.267984469150673, "learning_rate": 5.7225239015709324e-06, "loss": 0.395, "step": 6633 }, { "epoch": 0.4703296703296703, "grad_norm": 7.354930854213737, "learning_rate": 5.721387774491505e-06, "loss": 0.3979, "step": 6634 }, { "epoch": 0.4704005671747607, "grad_norm": 3.6602114177103724, "learning_rate": 5.7202516093727515e-06, "loss": 0.4087, "step": 6635 }, { "epoch": 0.4704714640198511, "grad_norm": 3.461226158659251, "learning_rate": 5.71911540627458e-06, "loss": 0.3613, "step": 6636 }, { "epoch": 0.4705423608649415, "grad_norm": 2.6029380727721905, "learning_rate": 5.717979165256905e-06, "loss": 0.4398, "step": 6637 }, { "epoch": 0.47061325771003193, "grad_norm": 3.1929039276758693, "learning_rate": 5.716842886379641e-06, "loss": 0.4176, "step": 6638 }, { "epoch": 0.4706841545551223, "grad_norm": 5.745526484359771, "learning_rate": 5.715706569702703e-06, "loss": 0.4578, "step": 6639 }, { "epoch": 0.4707550514002127, "grad_norm": 13.193599483977945, "learning_rate": 5.714570215286013e-06, "loss": 0.4169, "step": 6640 }, { "epoch": 0.4708259482453031, "grad_norm": 1.7805813543500364, "learning_rate": 5.7134338231894895e-06, "loss": 0.3555, "step": 6641 }, { "epoch": 0.4708968450903935, "grad_norm": 5.4705750913286915, "learning_rate": 5.712297393473055e-06, "loss": 0.4223, "step": 6642 }, { "epoch": 0.47096774193548385, "grad_norm": 4.185714057698237, "learning_rate": 5.711160926196636e-06, "loss": 0.3669, "step": 6643 }, { "epoch": 0.4710386387805743, "grad_norm": 2.083671833999676, "learning_rate": 5.71002442142016e-06, "loss": 0.3764, "step": 6644 }, { "epoch": 0.47110953562566465, "grad_norm": 3.3426755351709434, "learning_rate": 5.708887879203552e-06, "loss": 0.4781, "step": 6645 }, { "epoch": 0.47118043247075503, "grad_norm": 2.380546780146447, "learning_rate": 5.707751299606746e-06, "loss": 0.4305, "step": 6646 }, { "epoch": 0.47125132931584546, "grad_norm": 6.035140587020505, "learning_rate": 5.706614682689673e-06, "loss": 0.4123, "step": 6647 }, { "epoch": 0.47132222616093583, "grad_norm": 3.405179468154623, "learning_rate": 5.7054780285122705e-06, "loss": 0.4519, "step": 6648 }, { "epoch": 0.4713931230060262, "grad_norm": 1.0152837827461927, "learning_rate": 5.704341337134472e-06, "loss": 0.6844, "step": 6649 }, { "epoch": 0.47146401985111663, "grad_norm": 5.493662205854508, "learning_rate": 5.703204608616216e-06, "loss": 0.4221, "step": 6650 }, { "epoch": 0.471534916696207, "grad_norm": 1.1403698986871642, "learning_rate": 5.702067843017445e-06, "loss": 0.7233, "step": 6651 }, { "epoch": 0.47160581354129744, "grad_norm": 1.4976245764483498, "learning_rate": 5.700931040398101e-06, "loss": 0.3847, "step": 6652 }, { "epoch": 0.4716767103863878, "grad_norm": 2.031803097656337, "learning_rate": 5.699794200818128e-06, "loss": 0.437, "step": 6653 }, { "epoch": 0.4717476072314782, "grad_norm": 3.0436415442898856, "learning_rate": 5.698657324337474e-06, "loss": 0.4084, "step": 6654 }, { "epoch": 0.4718185040765686, "grad_norm": 2.2630278768143945, "learning_rate": 5.697520411016084e-06, "loss": 0.4116, "step": 6655 }, { "epoch": 0.471889400921659, "grad_norm": 4.0091582109501624, "learning_rate": 5.696383460913912e-06, "loss": 0.3937, "step": 6656 }, { "epoch": 0.47196029776674936, "grad_norm": 5.998506438546807, "learning_rate": 5.695246474090908e-06, "loss": 0.3607, "step": 6657 }, { "epoch": 0.4720311946118398, "grad_norm": 2.735940986216912, "learning_rate": 5.694109450607026e-06, "loss": 0.3905, "step": 6658 }, { "epoch": 0.47210209145693016, "grad_norm": 2.0979839665550464, "learning_rate": 5.692972390522223e-06, "loss": 0.3897, "step": 6659 }, { "epoch": 0.47217298830202054, "grad_norm": 4.290611827171453, "learning_rate": 5.691835293896457e-06, "loss": 0.4305, "step": 6660 }, { "epoch": 0.47224388514711096, "grad_norm": 2.804311798318843, "learning_rate": 5.69069816078969e-06, "loss": 0.4353, "step": 6661 }, { "epoch": 0.47231478199220134, "grad_norm": 14.892844679696497, "learning_rate": 5.68956099126188e-06, "loss": 0.4182, "step": 6662 }, { "epoch": 0.47238567883729177, "grad_norm": 3.697384400471206, "learning_rate": 5.688423785372993e-06, "loss": 0.4121, "step": 6663 }, { "epoch": 0.47245657568238214, "grad_norm": 3.0173014400139886, "learning_rate": 5.687286543182994e-06, "loss": 0.4547, "step": 6664 }, { "epoch": 0.4725274725274725, "grad_norm": 4.005368758851699, "learning_rate": 5.6861492647518515e-06, "loss": 0.3914, "step": 6665 }, { "epoch": 0.47259836937256294, "grad_norm": 2.452602391962201, "learning_rate": 5.685011950139533e-06, "loss": 0.3567, "step": 6666 }, { "epoch": 0.4726692662176533, "grad_norm": 6.3299656889452205, "learning_rate": 5.683874599406014e-06, "loss": 0.4244, "step": 6667 }, { "epoch": 0.4727401630627437, "grad_norm": 2.183624533110317, "learning_rate": 5.682737212611263e-06, "loss": 0.4789, "step": 6668 }, { "epoch": 0.4728110599078341, "grad_norm": 2.429783339987987, "learning_rate": 5.681599789815259e-06, "loss": 0.3958, "step": 6669 }, { "epoch": 0.4728819567529245, "grad_norm": 2.1378173810528813, "learning_rate": 5.680462331077977e-06, "loss": 0.3959, "step": 6670 }, { "epoch": 0.47295285359801487, "grad_norm": 3.4323917823546064, "learning_rate": 5.6793248364593975e-06, "loss": 0.4174, "step": 6671 }, { "epoch": 0.4730237504431053, "grad_norm": 2.5440388457981724, "learning_rate": 5.6781873060195e-06, "loss": 0.418, "step": 6672 }, { "epoch": 0.47309464728819567, "grad_norm": 3.256779872469603, "learning_rate": 5.677049739818271e-06, "loss": 0.4039, "step": 6673 }, { "epoch": 0.47316554413328604, "grad_norm": 2.0745518476275215, "learning_rate": 5.67591213791569e-06, "loss": 0.4009, "step": 6674 }, { "epoch": 0.47323644097837647, "grad_norm": 4.62677716169007, "learning_rate": 5.674774500371747e-06, "loss": 0.4538, "step": 6675 }, { "epoch": 0.47330733782346684, "grad_norm": 2.2945742398138314, "learning_rate": 5.673636827246429e-06, "loss": 0.3996, "step": 6676 }, { "epoch": 0.4733782346685573, "grad_norm": 2.395464796146269, "learning_rate": 5.672499118599727e-06, "loss": 0.468, "step": 6677 }, { "epoch": 0.47344913151364765, "grad_norm": 8.443961692704919, "learning_rate": 5.671361374491634e-06, "loss": 0.4206, "step": 6678 }, { "epoch": 0.473520028358738, "grad_norm": 3.300502684689342, "learning_rate": 5.670223594982142e-06, "loss": 0.4075, "step": 6679 }, { "epoch": 0.47359092520382845, "grad_norm": 2.5192729984104867, "learning_rate": 5.66908578013125e-06, "loss": 0.4064, "step": 6680 }, { "epoch": 0.4736618220489188, "grad_norm": 2.651696562962927, "learning_rate": 5.66794792999895e-06, "loss": 0.4322, "step": 6681 }, { "epoch": 0.4737327188940092, "grad_norm": 3.932354065609856, "learning_rate": 5.66681004464525e-06, "loss": 0.4197, "step": 6682 }, { "epoch": 0.4738036157390996, "grad_norm": 3.0535126367087213, "learning_rate": 5.665672124130146e-06, "loss": 0.4128, "step": 6683 }, { "epoch": 0.47387451258419, "grad_norm": 2.0998637599546295, "learning_rate": 5.664534168513643e-06, "loss": 0.3951, "step": 6684 }, { "epoch": 0.4739454094292804, "grad_norm": 3.085153308577265, "learning_rate": 5.663396177855747e-06, "loss": 0.3981, "step": 6685 }, { "epoch": 0.4740163062743708, "grad_norm": 3.8319080612462444, "learning_rate": 5.662258152216463e-06, "loss": 0.3794, "step": 6686 }, { "epoch": 0.4740872031194612, "grad_norm": 3.4826276921657207, "learning_rate": 5.661120091655801e-06, "loss": 0.4493, "step": 6687 }, { "epoch": 0.4741580999645516, "grad_norm": 2.757574200599288, "learning_rate": 5.659981996233773e-06, "loss": 0.42, "step": 6688 }, { "epoch": 0.474228996809642, "grad_norm": 2.978427930138782, "learning_rate": 5.658843866010389e-06, "loss": 0.4286, "step": 6689 }, { "epoch": 0.47429989365473235, "grad_norm": 2.5035854704090905, "learning_rate": 5.6577057010456655e-06, "loss": 0.391, "step": 6690 }, { "epoch": 0.4743707904998228, "grad_norm": 1.823974550653835, "learning_rate": 5.656567501399622e-06, "loss": 0.4296, "step": 6691 }, { "epoch": 0.47444168734491315, "grad_norm": 2.940286143547894, "learning_rate": 5.655429267132267e-06, "loss": 0.4164, "step": 6692 }, { "epoch": 0.47451258419000353, "grad_norm": 3.5432657619714765, "learning_rate": 5.654290998303631e-06, "loss": 0.4183, "step": 6693 }, { "epoch": 0.47458348103509396, "grad_norm": 3.649626495557927, "learning_rate": 5.653152694973729e-06, "loss": 0.3987, "step": 6694 }, { "epoch": 0.47465437788018433, "grad_norm": 2.339169946753468, "learning_rate": 5.652014357202588e-06, "loss": 0.429, "step": 6695 }, { "epoch": 0.4747252747252747, "grad_norm": 3.1266664807608073, "learning_rate": 5.650875985050232e-06, "loss": 0.4108, "step": 6696 }, { "epoch": 0.47479617157036513, "grad_norm": 2.019474404658189, "learning_rate": 5.649737578576687e-06, "loss": 0.3947, "step": 6697 }, { "epoch": 0.4748670684154555, "grad_norm": 2.857820893079517, "learning_rate": 5.648599137841986e-06, "loss": 0.409, "step": 6698 }, { "epoch": 0.4749379652605459, "grad_norm": 7.208223177296553, "learning_rate": 5.647460662906157e-06, "loss": 0.4321, "step": 6699 }, { "epoch": 0.4750088621056363, "grad_norm": 1.784114597163373, "learning_rate": 5.6463221538292315e-06, "loss": 0.389, "step": 6700 }, { "epoch": 0.4750797589507267, "grad_norm": 1.7368074864228, "learning_rate": 5.645183610671247e-06, "loss": 0.4484, "step": 6701 }, { "epoch": 0.4751506557958171, "grad_norm": 3.1382737662687634, "learning_rate": 5.644045033492237e-06, "loss": 0.3938, "step": 6702 }, { "epoch": 0.4752215526409075, "grad_norm": 3.33329227817666, "learning_rate": 5.642906422352241e-06, "loss": 0.4015, "step": 6703 }, { "epoch": 0.47529244948599786, "grad_norm": 1.9833106196459445, "learning_rate": 5.641767777311299e-06, "loss": 0.385, "step": 6704 }, { "epoch": 0.4753633463310883, "grad_norm": 2.648630734594857, "learning_rate": 5.6406290984294535e-06, "loss": 0.4194, "step": 6705 }, { "epoch": 0.47543424317617866, "grad_norm": 2.2097637275405777, "learning_rate": 5.639490385766746e-06, "loss": 0.3913, "step": 6706 }, { "epoch": 0.47550514002126903, "grad_norm": 3.724804288379878, "learning_rate": 5.638351639383222e-06, "loss": 0.3704, "step": 6707 }, { "epoch": 0.47557603686635946, "grad_norm": 2.8552238793824487, "learning_rate": 5.637212859338928e-06, "loss": 0.4127, "step": 6708 }, { "epoch": 0.47564693371144984, "grad_norm": 10.751731780593675, "learning_rate": 5.636074045693913e-06, "loss": 0.4669, "step": 6709 }, { "epoch": 0.4757178305565402, "grad_norm": 3.304161010216155, "learning_rate": 5.63493519850823e-06, "loss": 0.3874, "step": 6710 }, { "epoch": 0.47578872740163064, "grad_norm": 2.601695493444215, "learning_rate": 5.633796317841929e-06, "loss": 0.4316, "step": 6711 }, { "epoch": 0.475859624246721, "grad_norm": 4.268152972670025, "learning_rate": 5.6326574037550655e-06, "loss": 0.4632, "step": 6712 }, { "epoch": 0.4759305210918114, "grad_norm": 4.958440859095518, "learning_rate": 5.631518456307691e-06, "loss": 0.4512, "step": 6713 }, { "epoch": 0.4760014179369018, "grad_norm": 1.258407772341926, "learning_rate": 5.630379475559868e-06, "loss": 0.738, "step": 6714 }, { "epoch": 0.4760723147819922, "grad_norm": 9.072975087003233, "learning_rate": 5.6292404615716544e-06, "loss": 0.392, "step": 6715 }, { "epoch": 0.4761432116270826, "grad_norm": 1.9219547070359444, "learning_rate": 5.628101414403111e-06, "loss": 0.3692, "step": 6716 }, { "epoch": 0.476214108472173, "grad_norm": 4.089398964887026, "learning_rate": 5.626962334114301e-06, "loss": 0.419, "step": 6717 }, { "epoch": 0.47628500531726337, "grad_norm": 4.201295993517288, "learning_rate": 5.625823220765287e-06, "loss": 0.3893, "step": 6718 }, { "epoch": 0.4763559021623538, "grad_norm": 2.433633750513231, "learning_rate": 5.624684074416139e-06, "loss": 0.4303, "step": 6719 }, { "epoch": 0.47642679900744417, "grad_norm": 4.572980856176571, "learning_rate": 5.623544895126922e-06, "loss": 0.3814, "step": 6720 }, { "epoch": 0.47649769585253454, "grad_norm": 5.255191999923158, "learning_rate": 5.622405682957707e-06, "loss": 0.4515, "step": 6721 }, { "epoch": 0.47656859269762497, "grad_norm": 4.12373372751152, "learning_rate": 5.621266437968564e-06, "loss": 0.3881, "step": 6722 }, { "epoch": 0.47663948954271534, "grad_norm": 6.32092875840614, "learning_rate": 5.62012716021957e-06, "loss": 0.4204, "step": 6723 }, { "epoch": 0.4767103863878057, "grad_norm": 11.407962203664546, "learning_rate": 5.618987849770794e-06, "loss": 0.4469, "step": 6724 }, { "epoch": 0.47678128323289615, "grad_norm": 3.1724485992109557, "learning_rate": 5.61784850668232e-06, "loss": 0.466, "step": 6725 }, { "epoch": 0.4768521800779865, "grad_norm": 2.2748393335590977, "learning_rate": 5.6167091310142184e-06, "loss": 0.404, "step": 6726 }, { "epoch": 0.47692307692307695, "grad_norm": 10.05757059778203, "learning_rate": 5.615569722826577e-06, "loss": 0.4293, "step": 6727 }, { "epoch": 0.4769939737681673, "grad_norm": 4.301300246247498, "learning_rate": 5.614430282179474e-06, "loss": 0.424, "step": 6728 }, { "epoch": 0.4770648706132577, "grad_norm": 2.658453309058149, "learning_rate": 5.613290809132992e-06, "loss": 0.4456, "step": 6729 }, { "epoch": 0.4771357674583481, "grad_norm": 3.5419121305293, "learning_rate": 5.612151303747217e-06, "loss": 0.4582, "step": 6730 }, { "epoch": 0.4772066643034385, "grad_norm": 3.9939094204978973, "learning_rate": 5.611011766082239e-06, "loss": 0.4031, "step": 6731 }, { "epoch": 0.4772775611485289, "grad_norm": 2.343882133609441, "learning_rate": 5.609872196198142e-06, "loss": 0.4394, "step": 6732 }, { "epoch": 0.4773484579936193, "grad_norm": 3.8306547152812085, "learning_rate": 5.608732594155021e-06, "loss": 0.4669, "step": 6733 }, { "epoch": 0.4774193548387097, "grad_norm": 2.4849847862492567, "learning_rate": 5.607592960012963e-06, "loss": 0.4579, "step": 6734 }, { "epoch": 0.47749025168380005, "grad_norm": 2.7890446204364747, "learning_rate": 5.606453293832066e-06, "loss": 0.4006, "step": 6735 }, { "epoch": 0.4775611485288905, "grad_norm": 2.338515997427861, "learning_rate": 5.605313595672423e-06, "loss": 0.3993, "step": 6736 }, { "epoch": 0.47763204537398085, "grad_norm": 2.8895403424185138, "learning_rate": 5.604173865594133e-06, "loss": 0.377, "step": 6737 }, { "epoch": 0.4777029422190712, "grad_norm": 1.9378474564818347, "learning_rate": 5.603034103657294e-06, "loss": 0.4182, "step": 6738 }, { "epoch": 0.47777383906416165, "grad_norm": 6.2705646869252245, "learning_rate": 5.601894309922006e-06, "loss": 0.442, "step": 6739 }, { "epoch": 0.477844735909252, "grad_norm": 2.8647956348310997, "learning_rate": 5.60075448444837e-06, "loss": 0.4132, "step": 6740 }, { "epoch": 0.47791563275434246, "grad_norm": 2.2994158692417175, "learning_rate": 5.599614627296493e-06, "loss": 0.4156, "step": 6741 }, { "epoch": 0.47798652959943283, "grad_norm": 6.519252125412405, "learning_rate": 5.5984747385264804e-06, "loss": 0.3645, "step": 6742 }, { "epoch": 0.4780574264445232, "grad_norm": 1.970650638720974, "learning_rate": 5.597334818198437e-06, "loss": 0.449, "step": 6743 }, { "epoch": 0.47812832328961363, "grad_norm": 2.0763983164054762, "learning_rate": 5.596194866372473e-06, "loss": 0.4105, "step": 6744 }, { "epoch": 0.478199220134704, "grad_norm": 11.190788011621486, "learning_rate": 5.595054883108697e-06, "loss": 0.4453, "step": 6745 }, { "epoch": 0.4782701169797944, "grad_norm": 3.1165935048774993, "learning_rate": 5.593914868467224e-06, "loss": 0.42, "step": 6746 }, { "epoch": 0.4783410138248848, "grad_norm": 5.362984674789947, "learning_rate": 5.592774822508167e-06, "loss": 0.3994, "step": 6747 }, { "epoch": 0.4784119106699752, "grad_norm": 2.4732707142813424, "learning_rate": 5.591634745291641e-06, "loss": 0.4106, "step": 6748 }, { "epoch": 0.47848280751506556, "grad_norm": 3.0797981131115373, "learning_rate": 5.590494636877762e-06, "loss": 0.4314, "step": 6749 }, { "epoch": 0.478553704360156, "grad_norm": 2.7283619444003424, "learning_rate": 5.589354497326651e-06, "loss": 0.4924, "step": 6750 }, { "epoch": 0.47862460120524636, "grad_norm": 6.791347149556834, "learning_rate": 5.588214326698429e-06, "loss": 0.4296, "step": 6751 }, { "epoch": 0.4786954980503368, "grad_norm": 2.4947484817519663, "learning_rate": 5.587074125053216e-06, "loss": 0.4088, "step": 6752 }, { "epoch": 0.47876639489542716, "grad_norm": 1.7840942693657642, "learning_rate": 5.585933892451134e-06, "loss": 0.3863, "step": 6753 }, { "epoch": 0.47883729174051753, "grad_norm": 1.8634512976674058, "learning_rate": 5.584793628952313e-06, "loss": 0.347, "step": 6754 }, { "epoch": 0.47890818858560796, "grad_norm": 3.8753596731091475, "learning_rate": 5.583653334616876e-06, "loss": 0.4438, "step": 6755 }, { "epoch": 0.47897908543069834, "grad_norm": 1.8613460053515805, "learning_rate": 5.582513009504954e-06, "loss": 0.3512, "step": 6756 }, { "epoch": 0.4790499822757887, "grad_norm": 2.417722667136902, "learning_rate": 5.581372653676678e-06, "loss": 0.3938, "step": 6757 }, { "epoch": 0.47912087912087914, "grad_norm": 13.700623467586876, "learning_rate": 5.580232267192175e-06, "loss": 0.4318, "step": 6758 }, { "epoch": 0.4791917759659695, "grad_norm": 7.681927812033932, "learning_rate": 5.579091850111584e-06, "loss": 0.4311, "step": 6759 }, { "epoch": 0.4792626728110599, "grad_norm": 2.9710933556030095, "learning_rate": 5.577951402495037e-06, "loss": 0.4971, "step": 6760 }, { "epoch": 0.4793335696561503, "grad_norm": 3.065790495115028, "learning_rate": 5.576810924402672e-06, "loss": 0.4507, "step": 6761 }, { "epoch": 0.4794044665012407, "grad_norm": 2.9026032103246218, "learning_rate": 5.575670415894625e-06, "loss": 0.4085, "step": 6762 }, { "epoch": 0.47947536334633106, "grad_norm": 4.372072915068248, "learning_rate": 5.574529877031038e-06, "loss": 0.4094, "step": 6763 }, { "epoch": 0.4795462601914215, "grad_norm": 5.711287881730295, "learning_rate": 5.573389307872054e-06, "loss": 0.417, "step": 6764 }, { "epoch": 0.47961715703651187, "grad_norm": 2.655521079179631, "learning_rate": 5.572248708477813e-06, "loss": 0.3977, "step": 6765 }, { "epoch": 0.4796880538816023, "grad_norm": 1.4216553529843716, "learning_rate": 5.57110807890846e-06, "loss": 0.3434, "step": 6766 }, { "epoch": 0.47975895072669267, "grad_norm": 2.55841528173058, "learning_rate": 5.569967419224144e-06, "loss": 0.4329, "step": 6767 }, { "epoch": 0.47982984757178304, "grad_norm": 5.968082330502619, "learning_rate": 5.568826729485009e-06, "loss": 0.4225, "step": 6768 }, { "epoch": 0.47990074441687347, "grad_norm": 7.226284182721181, "learning_rate": 5.567686009751208e-06, "loss": 0.4714, "step": 6769 }, { "epoch": 0.47997164126196384, "grad_norm": 4.686227422843736, "learning_rate": 5.566545260082889e-06, "loss": 0.4272, "step": 6770 }, { "epoch": 0.4800425381070542, "grad_norm": 2.360426952263747, "learning_rate": 5.5654044805402065e-06, "loss": 0.4243, "step": 6771 }, { "epoch": 0.48011343495214465, "grad_norm": 2.329415796391348, "learning_rate": 5.564263671183315e-06, "loss": 0.4466, "step": 6772 }, { "epoch": 0.480184331797235, "grad_norm": 2.1366061701095407, "learning_rate": 5.563122832072369e-06, "loss": 0.443, "step": 6773 }, { "epoch": 0.4802552286423254, "grad_norm": 5.301702828482194, "learning_rate": 5.561981963267525e-06, "loss": 0.4077, "step": 6774 }, { "epoch": 0.4803261254874158, "grad_norm": 3.98942287258083, "learning_rate": 5.560841064828945e-06, "loss": 0.4246, "step": 6775 }, { "epoch": 0.4803970223325062, "grad_norm": 3.9728783813472788, "learning_rate": 5.559700136816785e-06, "loss": 0.3992, "step": 6776 }, { "epoch": 0.48046791917759657, "grad_norm": 2.422771871244631, "learning_rate": 5.558559179291212e-06, "loss": 0.3869, "step": 6777 }, { "epoch": 0.480538816022687, "grad_norm": 8.529181419861215, "learning_rate": 5.557418192312387e-06, "loss": 0.3984, "step": 6778 }, { "epoch": 0.48060971286777737, "grad_norm": 3.1398867646992388, "learning_rate": 5.5562771759404755e-06, "loss": 0.4224, "step": 6779 }, { "epoch": 0.4806806097128678, "grad_norm": 6.1273738956900194, "learning_rate": 5.555136130235642e-06, "loss": 0.4024, "step": 6780 }, { "epoch": 0.4807515065579582, "grad_norm": 5.09584549960333, "learning_rate": 5.553995055258058e-06, "loss": 0.4192, "step": 6781 }, { "epoch": 0.48082240340304855, "grad_norm": 2.2357943950904624, "learning_rate": 5.5528539510678915e-06, "loss": 0.3969, "step": 6782 }, { "epoch": 0.480893300248139, "grad_norm": 2.945197757281154, "learning_rate": 5.551712817725315e-06, "loss": 0.4185, "step": 6783 }, { "epoch": 0.48096419709322935, "grad_norm": 3.349029483313423, "learning_rate": 5.5505716552904985e-06, "loss": 0.4457, "step": 6784 }, { "epoch": 0.4810350939383197, "grad_norm": 4.804293075029286, "learning_rate": 5.549430463823621e-06, "loss": 0.5052, "step": 6785 }, { "epoch": 0.48110599078341015, "grad_norm": 2.8223979902736374, "learning_rate": 5.5482892433848544e-06, "loss": 0.435, "step": 6786 }, { "epoch": 0.4811768876285005, "grad_norm": 1.9983662112602596, "learning_rate": 5.547147994034378e-06, "loss": 0.3841, "step": 6787 }, { "epoch": 0.4812477844735909, "grad_norm": 6.010007056858514, "learning_rate": 5.54600671583237e-06, "loss": 0.4525, "step": 6788 }, { "epoch": 0.48131868131868133, "grad_norm": 5.368063797022613, "learning_rate": 5.544865408839011e-06, "loss": 0.3646, "step": 6789 }, { "epoch": 0.4813895781637717, "grad_norm": 2.773395179758406, "learning_rate": 5.543724073114483e-06, "loss": 0.4313, "step": 6790 }, { "epoch": 0.48146047500886213, "grad_norm": 19.387377788953707, "learning_rate": 5.54258270871897e-06, "loss": 0.4087, "step": 6791 }, { "epoch": 0.4815313718539525, "grad_norm": 5.4498729363131275, "learning_rate": 5.541441315712655e-06, "loss": 0.3697, "step": 6792 }, { "epoch": 0.4816022686990429, "grad_norm": 2.9341017990020206, "learning_rate": 5.540299894155729e-06, "loss": 0.451, "step": 6793 }, { "epoch": 0.4816731655441333, "grad_norm": 2.213235468642137, "learning_rate": 5.539158444108374e-06, "loss": 0.3969, "step": 6794 }, { "epoch": 0.4817440623892237, "grad_norm": 6.069104027702328, "learning_rate": 5.538016965630786e-06, "loss": 0.3578, "step": 6795 }, { "epoch": 0.48181495923431406, "grad_norm": 3.498773644163328, "learning_rate": 5.5368754587831506e-06, "loss": 0.3991, "step": 6796 }, { "epoch": 0.4818858560794045, "grad_norm": 2.6627162228932884, "learning_rate": 5.535733923625663e-06, "loss": 0.3628, "step": 6797 }, { "epoch": 0.48195675292449486, "grad_norm": 2.426068163550886, "learning_rate": 5.534592360218517e-06, "loss": 0.3974, "step": 6798 }, { "epoch": 0.48202764976958523, "grad_norm": 2.743409386257284, "learning_rate": 5.5334507686219074e-06, "loss": 0.4075, "step": 6799 }, { "epoch": 0.48209854661467566, "grad_norm": 4.4185402206916535, "learning_rate": 5.532309148896032e-06, "loss": 0.4215, "step": 6800 }, { "epoch": 0.48216944345976603, "grad_norm": 2.4601877564627093, "learning_rate": 5.531167501101089e-06, "loss": 0.4118, "step": 6801 }, { "epoch": 0.4822403403048564, "grad_norm": 3.120836223522387, "learning_rate": 5.5300258252972785e-06, "loss": 0.43, "step": 6802 }, { "epoch": 0.48231123714994684, "grad_norm": 2.5961221582542673, "learning_rate": 5.5288841215448e-06, "loss": 0.4549, "step": 6803 }, { "epoch": 0.4823821339950372, "grad_norm": 2.0948169741322538, "learning_rate": 5.5277423899038595e-06, "loss": 0.4101, "step": 6804 }, { "epoch": 0.48245303084012764, "grad_norm": 1.8581574404711296, "learning_rate": 5.52660063043466e-06, "loss": 0.4124, "step": 6805 }, { "epoch": 0.482523927685218, "grad_norm": 2.8135928831354478, "learning_rate": 5.525458843197407e-06, "loss": 0.4473, "step": 6806 }, { "epoch": 0.4825948245303084, "grad_norm": 4.734502193181236, "learning_rate": 5.5243170282523075e-06, "loss": 0.3734, "step": 6807 }, { "epoch": 0.4826657213753988, "grad_norm": 3.1611546804040396, "learning_rate": 5.523175185659572e-06, "loss": 0.4416, "step": 6808 }, { "epoch": 0.4827366182204892, "grad_norm": 2.867778303506973, "learning_rate": 5.522033315479409e-06, "loss": 0.3811, "step": 6809 }, { "epoch": 0.48280751506557956, "grad_norm": 2.891888560190596, "learning_rate": 5.520891417772032e-06, "loss": 0.3966, "step": 6810 }, { "epoch": 0.48287841191067, "grad_norm": 2.446061583527292, "learning_rate": 5.5197494925976514e-06, "loss": 0.4257, "step": 6811 }, { "epoch": 0.48294930875576036, "grad_norm": 2.291058097283361, "learning_rate": 5.518607540016484e-06, "loss": 0.3937, "step": 6812 }, { "epoch": 0.48302020560085074, "grad_norm": 5.262457571374499, "learning_rate": 5.517465560088745e-06, "loss": 0.4059, "step": 6813 }, { "epoch": 0.48309110244594117, "grad_norm": 3.1384415325733603, "learning_rate": 5.516323552874652e-06, "loss": 0.4131, "step": 6814 }, { "epoch": 0.48316199929103154, "grad_norm": 3.66417454381241, "learning_rate": 5.515181518434425e-06, "loss": 0.4167, "step": 6815 }, { "epoch": 0.48323289613612197, "grad_norm": 2.3156735817329945, "learning_rate": 5.5140394568282805e-06, "loss": 0.4057, "step": 6816 }, { "epoch": 0.48330379298121234, "grad_norm": 4.895869733481389, "learning_rate": 5.512897368116447e-06, "loss": 0.4091, "step": 6817 }, { "epoch": 0.4833746898263027, "grad_norm": 3.3198141586099577, "learning_rate": 5.511755252359141e-06, "loss": 0.4435, "step": 6818 }, { "epoch": 0.48344558667139315, "grad_norm": 3.4154695431939786, "learning_rate": 5.51061310961659e-06, "loss": 0.4151, "step": 6819 }, { "epoch": 0.4835164835164835, "grad_norm": 2.472138499978638, "learning_rate": 5.509470939949021e-06, "loss": 0.4269, "step": 6820 }, { "epoch": 0.4835873803615739, "grad_norm": 5.728157418509995, "learning_rate": 5.50832874341666e-06, "loss": 0.4132, "step": 6821 }, { "epoch": 0.4836582772066643, "grad_norm": 2.6493626113205897, "learning_rate": 5.507186520079737e-06, "loss": 0.4312, "step": 6822 }, { "epoch": 0.4837291740517547, "grad_norm": 2.253790037055245, "learning_rate": 5.5060442699984814e-06, "loss": 0.3903, "step": 6823 }, { "epoch": 0.48380007089684507, "grad_norm": 2.1402294763565224, "learning_rate": 5.504901993233125e-06, "loss": 0.4071, "step": 6824 }, { "epoch": 0.4838709677419355, "grad_norm": 2.2489924757246413, "learning_rate": 5.503759689843902e-06, "loss": 0.4418, "step": 6825 }, { "epoch": 0.48394186458702587, "grad_norm": 2.4355071717508476, "learning_rate": 5.502617359891045e-06, "loss": 0.3493, "step": 6826 }, { "epoch": 0.48401276143211625, "grad_norm": 2.325407844875994, "learning_rate": 5.501475003434791e-06, "loss": 0.4314, "step": 6827 }, { "epoch": 0.4840836582772067, "grad_norm": 4.852027742383635, "learning_rate": 5.500332620535377e-06, "loss": 0.3658, "step": 6828 }, { "epoch": 0.48415455512229705, "grad_norm": 3.120554968470892, "learning_rate": 5.499190211253043e-06, "loss": 0.4081, "step": 6829 }, { "epoch": 0.4842254519673875, "grad_norm": 2.7620471629684027, "learning_rate": 5.4980477756480275e-06, "loss": 0.3831, "step": 6830 }, { "epoch": 0.48429634881247785, "grad_norm": 2.402427761102229, "learning_rate": 5.496905313780574e-06, "loss": 0.4107, "step": 6831 }, { "epoch": 0.4843672456575682, "grad_norm": 2.0865399238521825, "learning_rate": 5.495762825710922e-06, "loss": 0.4303, "step": 6832 }, { "epoch": 0.48443814250265865, "grad_norm": 2.866561998456905, "learning_rate": 5.494620311499319e-06, "loss": 0.402, "step": 6833 }, { "epoch": 0.484509039347749, "grad_norm": 5.121186004351229, "learning_rate": 5.493477771206009e-06, "loss": 0.415, "step": 6834 }, { "epoch": 0.4845799361928394, "grad_norm": 7.890223501724758, "learning_rate": 5.492335204891239e-06, "loss": 0.3771, "step": 6835 }, { "epoch": 0.48465083303792983, "grad_norm": 5.219545015856912, "learning_rate": 5.491192612615259e-06, "loss": 0.4145, "step": 6836 }, { "epoch": 0.4847217298830202, "grad_norm": 2.925604013937996, "learning_rate": 5.4900499944383165e-06, "loss": 0.3655, "step": 6837 }, { "epoch": 0.4847926267281106, "grad_norm": 82.22832295434712, "learning_rate": 5.488907350420664e-06, "loss": 0.3762, "step": 6838 }, { "epoch": 0.484863523573201, "grad_norm": 4.089874901457668, "learning_rate": 5.487764680622555e-06, "loss": 0.3941, "step": 6839 }, { "epoch": 0.4849344204182914, "grad_norm": 8.015647943184176, "learning_rate": 5.4866219851042415e-06, "loss": 0.3967, "step": 6840 }, { "epoch": 0.48500531726338175, "grad_norm": 4.076204519140611, "learning_rate": 5.485479263925979e-06, "loss": 0.3731, "step": 6841 }, { "epoch": 0.4850762141084722, "grad_norm": 1.1158050251539169, "learning_rate": 5.4843365171480245e-06, "loss": 0.6534, "step": 6842 }, { "epoch": 0.48514711095356255, "grad_norm": 1.1478991324745738, "learning_rate": 5.483193744830636e-06, "loss": 0.763, "step": 6843 }, { "epoch": 0.485218007798653, "grad_norm": 2.4963720997227745, "learning_rate": 5.482050947034072e-06, "loss": 0.4215, "step": 6844 }, { "epoch": 0.48528890464374336, "grad_norm": 3.0902180636545733, "learning_rate": 5.480908123818593e-06, "loss": 0.4341, "step": 6845 }, { "epoch": 0.48535980148883373, "grad_norm": 3.8359458465198553, "learning_rate": 5.479765275244463e-06, "loss": 0.3901, "step": 6846 }, { "epoch": 0.48543069833392416, "grad_norm": 4.476337077777161, "learning_rate": 5.478622401371944e-06, "loss": 0.4306, "step": 6847 }, { "epoch": 0.48550159517901453, "grad_norm": 5.775402879323432, "learning_rate": 5.477479502261298e-06, "loss": 0.3986, "step": 6848 }, { "epoch": 0.4855724920241049, "grad_norm": 2.8996170276396787, "learning_rate": 5.476336577972797e-06, "loss": 0.3848, "step": 6849 }, { "epoch": 0.48564338886919534, "grad_norm": 2.6553847763578395, "learning_rate": 5.4751936285667005e-06, "loss": 0.4255, "step": 6850 }, { "epoch": 0.4857142857142857, "grad_norm": 6.649053379210927, "learning_rate": 5.474050654103284e-06, "loss": 0.3962, "step": 6851 }, { "epoch": 0.4857851825593761, "grad_norm": 7.106609635035912, "learning_rate": 5.472907654642815e-06, "loss": 0.3936, "step": 6852 }, { "epoch": 0.4858560794044665, "grad_norm": 1.9576801885228494, "learning_rate": 5.471764630245564e-06, "loss": 0.3931, "step": 6853 }, { "epoch": 0.4859269762495569, "grad_norm": 1.8624075724874463, "learning_rate": 5.470621580971803e-06, "loss": 0.3946, "step": 6854 }, { "epoch": 0.4859978730946473, "grad_norm": 2.5348431249053065, "learning_rate": 5.4694785068818075e-06, "loss": 0.4485, "step": 6855 }, { "epoch": 0.4860687699397377, "grad_norm": 3.548772575556365, "learning_rate": 5.468335408035852e-06, "loss": 0.3707, "step": 6856 }, { "epoch": 0.48613966678482806, "grad_norm": 1.9871916656266462, "learning_rate": 5.4671922844942136e-06, "loss": 0.4412, "step": 6857 }, { "epoch": 0.4862105636299185, "grad_norm": 4.185380229435831, "learning_rate": 5.466049136317168e-06, "loss": 0.375, "step": 6858 }, { "epoch": 0.48628146047500886, "grad_norm": 2.891026829149933, "learning_rate": 5.464905963564997e-06, "loss": 0.4022, "step": 6859 }, { "epoch": 0.48635235732009924, "grad_norm": 2.0680610398729224, "learning_rate": 5.463762766297979e-06, "loss": 0.381, "step": 6860 }, { "epoch": 0.48642325416518967, "grad_norm": 2.932968964323134, "learning_rate": 5.462619544576396e-06, "loss": 0.4078, "step": 6861 }, { "epoch": 0.48649415101028004, "grad_norm": 3.103667208043571, "learning_rate": 5.461476298460533e-06, "loss": 0.3766, "step": 6862 }, { "epoch": 0.4865650478553704, "grad_norm": 4.613445241472457, "learning_rate": 5.460333028010672e-06, "loss": 0.4477, "step": 6863 }, { "epoch": 0.48663594470046084, "grad_norm": 2.479942514295342, "learning_rate": 5.459189733287097e-06, "loss": 0.4403, "step": 6864 }, { "epoch": 0.4867068415455512, "grad_norm": 3.0672977849523817, "learning_rate": 5.458046414350099e-06, "loss": 0.4559, "step": 6865 }, { "epoch": 0.4867777383906416, "grad_norm": 5.3694066390791715, "learning_rate": 5.456903071259962e-06, "loss": 0.4076, "step": 6866 }, { "epoch": 0.486848635235732, "grad_norm": 4.861055183524354, "learning_rate": 5.455759704076979e-06, "loss": 0.4601, "step": 6867 }, { "epoch": 0.4869195320808224, "grad_norm": 1.2161802371113013, "learning_rate": 5.454616312861439e-06, "loss": 0.713, "step": 6868 }, { "epoch": 0.4869904289259128, "grad_norm": 3.397577214334364, "learning_rate": 5.453472897673633e-06, "loss": 0.4086, "step": 6869 }, { "epoch": 0.4870613257710032, "grad_norm": 2.5494118963721553, "learning_rate": 5.4523294585738566e-06, "loss": 0.3942, "step": 6870 }, { "epoch": 0.48713222261609357, "grad_norm": 3.200689922961028, "learning_rate": 5.4511859956224e-06, "loss": 0.4185, "step": 6871 }, { "epoch": 0.487203119461184, "grad_norm": 2.2681371369142598, "learning_rate": 5.450042508879562e-06, "loss": 0.3943, "step": 6872 }, { "epoch": 0.48727401630627437, "grad_norm": 3.2292164373851215, "learning_rate": 5.44889899840564e-06, "loss": 0.4338, "step": 6873 }, { "epoch": 0.48734491315136474, "grad_norm": 3.572216530052383, "learning_rate": 5.447755464260932e-06, "loss": 0.4418, "step": 6874 }, { "epoch": 0.4874158099964552, "grad_norm": 2.632335892805738, "learning_rate": 5.4466119065057366e-06, "loss": 0.373, "step": 6875 }, { "epoch": 0.48748670684154555, "grad_norm": 2.59849064262224, "learning_rate": 5.445468325200353e-06, "loss": 0.3927, "step": 6876 }, { "epoch": 0.4875576036866359, "grad_norm": 3.3635796663506934, "learning_rate": 5.4443247204050855e-06, "loss": 0.4845, "step": 6877 }, { "epoch": 0.48762850053172635, "grad_norm": 7.979101499736228, "learning_rate": 5.443181092180235e-06, "loss": 0.4109, "step": 6878 }, { "epoch": 0.4876993973768167, "grad_norm": 2.1933091906761994, "learning_rate": 5.442037440586109e-06, "loss": 0.3807, "step": 6879 }, { "epoch": 0.48777029422190715, "grad_norm": 3.132512844090813, "learning_rate": 5.440893765683009e-06, "loss": 0.4092, "step": 6880 }, { "epoch": 0.4878411910669975, "grad_norm": 2.348818464251379, "learning_rate": 5.4397500675312474e-06, "loss": 0.421, "step": 6881 }, { "epoch": 0.4879120879120879, "grad_norm": 2.8660204496003123, "learning_rate": 5.438606346191127e-06, "loss": 0.4454, "step": 6882 }, { "epoch": 0.48798298475717833, "grad_norm": 1.9447816265797644, "learning_rate": 5.437462601722961e-06, "loss": 0.4256, "step": 6883 }, { "epoch": 0.4880538816022687, "grad_norm": 2.670865323167659, "learning_rate": 5.436318834187056e-06, "loss": 0.4127, "step": 6884 }, { "epoch": 0.4881247784473591, "grad_norm": 2.517489370214937, "learning_rate": 5.435175043643727e-06, "loss": 0.4135, "step": 6885 }, { "epoch": 0.4881956752924495, "grad_norm": 2.293561520756111, "learning_rate": 5.434031230153285e-06, "loss": 0.3948, "step": 6886 }, { "epoch": 0.4882665721375399, "grad_norm": 2.7565231919351785, "learning_rate": 5.432887393776045e-06, "loss": 0.4272, "step": 6887 }, { "epoch": 0.48833746898263025, "grad_norm": 2.8339717665298245, "learning_rate": 5.431743534572324e-06, "loss": 0.411, "step": 6888 }, { "epoch": 0.4884083658277207, "grad_norm": 5.153518223507155, "learning_rate": 5.430599652602435e-06, "loss": 0.4603, "step": 6889 }, { "epoch": 0.48847926267281105, "grad_norm": 1.7162434349765874, "learning_rate": 5.429455747926699e-06, "loss": 0.465, "step": 6890 }, { "epoch": 0.48855015951790143, "grad_norm": 2.2604327576635894, "learning_rate": 5.428311820605432e-06, "loss": 0.4246, "step": 6891 }, { "epoch": 0.48862105636299186, "grad_norm": 3.5593103582099994, "learning_rate": 5.427167870698957e-06, "loss": 0.4393, "step": 6892 }, { "epoch": 0.48869195320808223, "grad_norm": 2.195968186535572, "learning_rate": 5.426023898267592e-06, "loss": 0.4239, "step": 6893 }, { "epoch": 0.48876285005317266, "grad_norm": 2.7496102301820984, "learning_rate": 5.424879903371666e-06, "loss": 0.3678, "step": 6894 }, { "epoch": 0.48883374689826303, "grad_norm": 1.251237993883999, "learning_rate": 5.423735886071494e-06, "loss": 0.691, "step": 6895 }, { "epoch": 0.4889046437433534, "grad_norm": 5.275367511023467, "learning_rate": 5.422591846427408e-06, "loss": 0.3748, "step": 6896 }, { "epoch": 0.48897554058844384, "grad_norm": 3.0933220118569666, "learning_rate": 5.421447784499731e-06, "loss": 0.4259, "step": 6897 }, { "epoch": 0.4890464374335342, "grad_norm": 3.6076399556299865, "learning_rate": 5.42030370034879e-06, "loss": 0.3938, "step": 6898 }, { "epoch": 0.4891173342786246, "grad_norm": 4.6998809236203725, "learning_rate": 5.419159594034914e-06, "loss": 0.423, "step": 6899 }, { "epoch": 0.489188231123715, "grad_norm": 5.737682039142857, "learning_rate": 5.418015465618433e-06, "loss": 0.4258, "step": 6900 }, { "epoch": 0.4892591279688054, "grad_norm": 5.195464865569685, "learning_rate": 5.4168713151596776e-06, "loss": 0.4069, "step": 6901 }, { "epoch": 0.48933002481389576, "grad_norm": 4.260071596361965, "learning_rate": 5.4157271427189794e-06, "loss": 0.4469, "step": 6902 }, { "epoch": 0.4894009216589862, "grad_norm": 3.098211588476596, "learning_rate": 5.414582948356671e-06, "loss": 0.4254, "step": 6903 }, { "epoch": 0.48947181850407656, "grad_norm": 3.216079266894384, "learning_rate": 5.413438732133088e-06, "loss": 0.4549, "step": 6904 }, { "epoch": 0.48954271534916693, "grad_norm": 12.22560722212904, "learning_rate": 5.4122944941085655e-06, "loss": 0.3949, "step": 6905 }, { "epoch": 0.48961361219425736, "grad_norm": 6.490672577698815, "learning_rate": 5.4111502343434385e-06, "loss": 0.429, "step": 6906 }, { "epoch": 0.48968450903934774, "grad_norm": 2.8766128648681466, "learning_rate": 5.4100059528980475e-06, "loss": 0.3866, "step": 6907 }, { "epoch": 0.48975540588443817, "grad_norm": 4.814593558518622, "learning_rate": 5.408861649832728e-06, "loss": 0.4498, "step": 6908 }, { "epoch": 0.48982630272952854, "grad_norm": 2.5871072989315973, "learning_rate": 5.407717325207823e-06, "loss": 0.4218, "step": 6909 }, { "epoch": 0.4898971995746189, "grad_norm": 3.3056237780785382, "learning_rate": 5.406572979083672e-06, "loss": 0.3689, "step": 6910 }, { "epoch": 0.48996809641970934, "grad_norm": 4.3709677315181565, "learning_rate": 5.405428611520617e-06, "loss": 0.3752, "step": 6911 }, { "epoch": 0.4900389932647997, "grad_norm": 2.2703151611738854, "learning_rate": 5.404284222579002e-06, "loss": 0.4073, "step": 6912 }, { "epoch": 0.4901098901098901, "grad_norm": 2.720298047741237, "learning_rate": 5.403139812319171e-06, "loss": 0.4125, "step": 6913 }, { "epoch": 0.4901807869549805, "grad_norm": 3.6691481866644398, "learning_rate": 5.401995380801469e-06, "loss": 0.3713, "step": 6914 }, { "epoch": 0.4902516838000709, "grad_norm": 3.3428511976002526, "learning_rate": 5.400850928086247e-06, "loss": 0.3877, "step": 6915 }, { "epoch": 0.49032258064516127, "grad_norm": 5.455302114528444, "learning_rate": 5.399706454233846e-06, "loss": 0.4358, "step": 6916 }, { "epoch": 0.4903934774902517, "grad_norm": 1.8379976994634382, "learning_rate": 5.39856195930462e-06, "loss": 0.4133, "step": 6917 }, { "epoch": 0.49046437433534207, "grad_norm": 2.695378921003641, "learning_rate": 5.397417443358918e-06, "loss": 0.399, "step": 6918 }, { "epoch": 0.4905352711804325, "grad_norm": 2.2535506013693127, "learning_rate": 5.39627290645709e-06, "loss": 0.3942, "step": 6919 }, { "epoch": 0.49060616802552287, "grad_norm": 4.1171702603290035, "learning_rate": 5.3951283486594915e-06, "loss": 0.4281, "step": 6920 }, { "epoch": 0.49067706487061324, "grad_norm": 2.22188688985835, "learning_rate": 5.393983770026472e-06, "loss": 0.4427, "step": 6921 }, { "epoch": 0.4907479617157037, "grad_norm": 4.5524052580210155, "learning_rate": 5.392839170618387e-06, "loss": 0.4478, "step": 6922 }, { "epoch": 0.49081885856079405, "grad_norm": 4.4364722067841305, "learning_rate": 5.391694550495594e-06, "loss": 0.3657, "step": 6923 }, { "epoch": 0.4908897554058844, "grad_norm": 3.7058443356392643, "learning_rate": 5.390549909718447e-06, "loss": 0.4205, "step": 6924 }, { "epoch": 0.49096065225097485, "grad_norm": 2.9061263726147852, "learning_rate": 5.3894052483473055e-06, "loss": 0.4502, "step": 6925 }, { "epoch": 0.4910315490960652, "grad_norm": 1.1630681035226451, "learning_rate": 5.388260566442529e-06, "loss": 0.6507, "step": 6926 }, { "epoch": 0.4911024459411556, "grad_norm": 2.2359617712086437, "learning_rate": 5.387115864064475e-06, "loss": 0.4116, "step": 6927 }, { "epoch": 0.491173342786246, "grad_norm": 4.157857046871674, "learning_rate": 5.385971141273509e-06, "loss": 0.4449, "step": 6928 }, { "epoch": 0.4912442396313364, "grad_norm": 2.2881249226233624, "learning_rate": 5.3848263981299865e-06, "loss": 0.4289, "step": 6929 }, { "epoch": 0.4913151364764268, "grad_norm": 3.986184715804066, "learning_rate": 5.3836816346942755e-06, "loss": 0.3818, "step": 6930 }, { "epoch": 0.4913860333215172, "grad_norm": 3.3515804929695094, "learning_rate": 5.38253685102674e-06, "loss": 0.4216, "step": 6931 }, { "epoch": 0.4914569301666076, "grad_norm": 2.573686927832087, "learning_rate": 5.3813920471877435e-06, "loss": 0.4313, "step": 6932 }, { "epoch": 0.491527827011698, "grad_norm": 2.716756688547788, "learning_rate": 5.380247223237653e-06, "loss": 0.4137, "step": 6933 }, { "epoch": 0.4915987238567884, "grad_norm": 5.8134755295866345, "learning_rate": 5.379102379236838e-06, "loss": 0.4278, "step": 6934 }, { "epoch": 0.49166962070187875, "grad_norm": 3.4965398114076316, "learning_rate": 5.3779575152456634e-06, "loss": 0.3798, "step": 6935 }, { "epoch": 0.4917405175469692, "grad_norm": 3.98454591198917, "learning_rate": 5.3768126313245005e-06, "loss": 0.46, "step": 6936 }, { "epoch": 0.49181141439205955, "grad_norm": 2.1949545462666666, "learning_rate": 5.375667727533721e-06, "loss": 0.4089, "step": 6937 }, { "epoch": 0.4918823112371499, "grad_norm": 0.9620871463473301, "learning_rate": 5.374522803933695e-06, "loss": 0.6879, "step": 6938 }, { "epoch": 0.49195320808224036, "grad_norm": 4.001447455757116, "learning_rate": 5.3733778605847964e-06, "loss": 0.3984, "step": 6939 }, { "epoch": 0.49202410492733073, "grad_norm": 1.0124934586444918, "learning_rate": 5.372232897547396e-06, "loss": 0.6562, "step": 6940 }, { "epoch": 0.4920950017724211, "grad_norm": 2.0807069096909108, "learning_rate": 5.371087914881873e-06, "loss": 0.4135, "step": 6941 }, { "epoch": 0.49216589861751153, "grad_norm": 3.956067778276048, "learning_rate": 5.3699429126486e-06, "loss": 0.4082, "step": 6942 }, { "epoch": 0.4922367954626019, "grad_norm": 2.2382156848602293, "learning_rate": 5.368797890907955e-06, "loss": 0.3771, "step": 6943 }, { "epoch": 0.49230769230769234, "grad_norm": 3.2024643722522232, "learning_rate": 5.367652849720315e-06, "loss": 0.4008, "step": 6944 }, { "epoch": 0.4923785891527827, "grad_norm": 16.782116566266254, "learning_rate": 5.36650778914606e-06, "loss": 0.4426, "step": 6945 }, { "epoch": 0.4924494859978731, "grad_norm": 2.011528952398564, "learning_rate": 5.365362709245571e-06, "loss": 0.435, "step": 6946 }, { "epoch": 0.4925203828429635, "grad_norm": 4.127977743341431, "learning_rate": 5.364217610079228e-06, "loss": 0.4358, "step": 6947 }, { "epoch": 0.4925912796880539, "grad_norm": 7.504313872383651, "learning_rate": 5.3630724917074095e-06, "loss": 0.4275, "step": 6948 }, { "epoch": 0.49266217653314426, "grad_norm": 3.7517137452250893, "learning_rate": 5.361927354190503e-06, "loss": 0.4228, "step": 6949 }, { "epoch": 0.4927330733782347, "grad_norm": 3.7300494730208515, "learning_rate": 5.3607821975888904e-06, "loss": 0.4416, "step": 6950 }, { "epoch": 0.49280397022332506, "grad_norm": 2.224082369502897, "learning_rate": 5.359637021962957e-06, "loss": 0.4006, "step": 6951 }, { "epoch": 0.49287486706841543, "grad_norm": 3.8797417347366294, "learning_rate": 5.35849182737309e-06, "loss": 0.393, "step": 6952 }, { "epoch": 0.49294576391350586, "grad_norm": 5.8432345031565225, "learning_rate": 5.357346613879673e-06, "loss": 0.407, "step": 6953 }, { "epoch": 0.49301666075859624, "grad_norm": 3.2974152677194497, "learning_rate": 5.3562013815431e-06, "loss": 0.3665, "step": 6954 }, { "epoch": 0.4930875576036866, "grad_norm": 1.7467904868753694, "learning_rate": 5.355056130423753e-06, "loss": 0.4367, "step": 6955 }, { "epoch": 0.49315845444877704, "grad_norm": 2.67372463991532, "learning_rate": 5.3539108605820265e-06, "loss": 0.4498, "step": 6956 }, { "epoch": 0.4932293512938674, "grad_norm": 6.228764913133697, "learning_rate": 5.35276557207831e-06, "loss": 0.3787, "step": 6957 }, { "epoch": 0.49330024813895784, "grad_norm": 4.078807787463632, "learning_rate": 5.351620264972997e-06, "loss": 0.4128, "step": 6958 }, { "epoch": 0.4933711449840482, "grad_norm": 2.1787493971267256, "learning_rate": 5.350474939326477e-06, "loss": 0.4021, "step": 6959 }, { "epoch": 0.4934420418291386, "grad_norm": 1.783279036655512, "learning_rate": 5.349329595199147e-06, "loss": 0.4565, "step": 6960 }, { "epoch": 0.493512938674229, "grad_norm": 4.250568969503124, "learning_rate": 5.3481842326514e-06, "loss": 0.4033, "step": 6961 }, { "epoch": 0.4935838355193194, "grad_norm": 2.305391053215918, "learning_rate": 5.347038851743634e-06, "loss": 0.374, "step": 6962 }, { "epoch": 0.49365473236440977, "grad_norm": 3.051615047984937, "learning_rate": 5.345893452536244e-06, "loss": 0.4276, "step": 6963 }, { "epoch": 0.4937256292095002, "grad_norm": 22.63727639848638, "learning_rate": 5.344748035089628e-06, "loss": 0.4153, "step": 6964 }, { "epoch": 0.49379652605459057, "grad_norm": 2.648123937613984, "learning_rate": 5.343602599464184e-06, "loss": 0.3587, "step": 6965 }, { "epoch": 0.49386742289968094, "grad_norm": 2.765967787932687, "learning_rate": 5.342457145720314e-06, "loss": 0.3596, "step": 6966 }, { "epoch": 0.49393831974477137, "grad_norm": 1.6661344438337173, "learning_rate": 5.341311673918417e-06, "loss": 0.4512, "step": 6967 }, { "epoch": 0.49400921658986174, "grad_norm": 1.3212095017823584, "learning_rate": 5.340166184118895e-06, "loss": 0.4091, "step": 6968 }, { "epoch": 0.4940801134349521, "grad_norm": 2.4474737169648244, "learning_rate": 5.339020676382151e-06, "loss": 0.4372, "step": 6969 }, { "epoch": 0.49415101028004255, "grad_norm": 5.575160177982982, "learning_rate": 5.337875150768587e-06, "loss": 0.3584, "step": 6970 }, { "epoch": 0.4942219071251329, "grad_norm": 4.575320474017995, "learning_rate": 5.336729607338609e-06, "loss": 0.3772, "step": 6971 }, { "epoch": 0.49429280397022335, "grad_norm": 2.608861346497674, "learning_rate": 5.33558404615262e-06, "loss": 0.4332, "step": 6972 }, { "epoch": 0.4943637008153137, "grad_norm": 3.1577826234078437, "learning_rate": 5.33443846727103e-06, "loss": 0.4372, "step": 6973 }, { "epoch": 0.4944345976604041, "grad_norm": 2.383974972771783, "learning_rate": 5.333292870754244e-06, "loss": 0.4178, "step": 6974 }, { "epoch": 0.4945054945054945, "grad_norm": 2.140511706691109, "learning_rate": 5.332147256662671e-06, "loss": 0.3587, "step": 6975 }, { "epoch": 0.4945763913505849, "grad_norm": 2.801384442776326, "learning_rate": 5.331001625056719e-06, "loss": 0.3959, "step": 6976 }, { "epoch": 0.49464728819567527, "grad_norm": 1.7559419860791228, "learning_rate": 5.329855975996798e-06, "loss": 0.3979, "step": 6977 }, { "epoch": 0.4947181850407657, "grad_norm": 3.5346955680706604, "learning_rate": 5.32871030954332e-06, "loss": 0.422, "step": 6978 }, { "epoch": 0.4947890818858561, "grad_norm": 14.916628105235912, "learning_rate": 5.327564625756698e-06, "loss": 0.433, "step": 6979 }, { "epoch": 0.49485997873094645, "grad_norm": 4.069427196804566, "learning_rate": 5.326418924697342e-06, "loss": 0.38, "step": 6980 }, { "epoch": 0.4949308755760369, "grad_norm": 2.741320303022988, "learning_rate": 5.325273206425667e-06, "loss": 0.402, "step": 6981 }, { "epoch": 0.49500177242112725, "grad_norm": 2.6970126655563735, "learning_rate": 5.324127471002088e-06, "loss": 0.357, "step": 6982 }, { "epoch": 0.4950726692662177, "grad_norm": 1.6835047240462242, "learning_rate": 5.322981718487019e-06, "loss": 0.3431, "step": 6983 }, { "epoch": 0.49514356611130805, "grad_norm": 2.36647736037379, "learning_rate": 5.321835948940879e-06, "loss": 0.3999, "step": 6984 }, { "epoch": 0.4952144629563984, "grad_norm": 2.7187165610503494, "learning_rate": 5.3206901624240825e-06, "loss": 0.4055, "step": 6985 }, { "epoch": 0.49528535980148886, "grad_norm": 4.779696830412368, "learning_rate": 5.31954435899705e-06, "loss": 0.4204, "step": 6986 }, { "epoch": 0.49535625664657923, "grad_norm": 2.8568735945832002, "learning_rate": 5.318398538720198e-06, "loss": 0.43, "step": 6987 }, { "epoch": 0.4954271534916696, "grad_norm": 2.2373304372260674, "learning_rate": 5.31725270165395e-06, "loss": 0.3914, "step": 6988 }, { "epoch": 0.49549805033676003, "grad_norm": 3.385176821292801, "learning_rate": 5.316106847858723e-06, "loss": 0.4064, "step": 6989 }, { "epoch": 0.4955689471818504, "grad_norm": 1.67088424071396, "learning_rate": 5.314960977394942e-06, "loss": 0.4296, "step": 6990 }, { "epoch": 0.4956398440269408, "grad_norm": 8.353616703768363, "learning_rate": 5.313815090323027e-06, "loss": 0.418, "step": 6991 }, { "epoch": 0.4957107408720312, "grad_norm": 2.4719452722141937, "learning_rate": 5.312669186703406e-06, "loss": 0.4001, "step": 6992 }, { "epoch": 0.4957816377171216, "grad_norm": 3.962734038642295, "learning_rate": 5.311523266596495e-06, "loss": 0.4162, "step": 6993 }, { "epoch": 0.49585253456221196, "grad_norm": 2.736645152843618, "learning_rate": 5.3103773300627284e-06, "loss": 0.3947, "step": 6994 }, { "epoch": 0.4959234314073024, "grad_norm": 2.265508829669254, "learning_rate": 5.309231377162527e-06, "loss": 0.4132, "step": 6995 }, { "epoch": 0.49599432825239276, "grad_norm": 2.5851991106869248, "learning_rate": 5.308085407956318e-06, "loss": 0.3869, "step": 6996 }, { "epoch": 0.4960652250974832, "grad_norm": 2.799024321301535, "learning_rate": 5.306939422504531e-06, "loss": 0.4071, "step": 6997 }, { "epoch": 0.49613612194257356, "grad_norm": 2.370800109469119, "learning_rate": 5.305793420867594e-06, "loss": 0.3788, "step": 6998 }, { "epoch": 0.49620701878766393, "grad_norm": 3.0509598200147083, "learning_rate": 5.304647403105938e-06, "loss": 0.3844, "step": 6999 }, { "epoch": 0.49627791563275436, "grad_norm": 2.9388835263886337, "learning_rate": 5.3035013692799905e-06, "loss": 0.3849, "step": 7000 }, { "epoch": 0.49634881247784474, "grad_norm": 2.675986476492587, "learning_rate": 5.302355319450184e-06, "loss": 0.3901, "step": 7001 }, { "epoch": 0.4964197093229351, "grad_norm": 4.502549429038118, "learning_rate": 5.301209253676951e-06, "loss": 0.4561, "step": 7002 }, { "epoch": 0.49649060616802554, "grad_norm": 1.73461448012478, "learning_rate": 5.300063172020726e-06, "loss": 0.4161, "step": 7003 }, { "epoch": 0.4965615030131159, "grad_norm": 2.9100944606705617, "learning_rate": 5.298917074541937e-06, "loss": 0.3668, "step": 7004 }, { "epoch": 0.4966323998582063, "grad_norm": 3.1595214160473484, "learning_rate": 5.2977709613010265e-06, "loss": 0.426, "step": 7005 }, { "epoch": 0.4967032967032967, "grad_norm": 6.854816957670791, "learning_rate": 5.2966248323584225e-06, "loss": 0.4007, "step": 7006 }, { "epoch": 0.4967741935483871, "grad_norm": 2.4874522722930217, "learning_rate": 5.2954786877745666e-06, "loss": 0.4075, "step": 7007 }, { "epoch": 0.4968450903934775, "grad_norm": 21.3936768518738, "learning_rate": 5.2943325276098944e-06, "loss": 0.4273, "step": 7008 }, { "epoch": 0.4969159872385679, "grad_norm": 2.578106140696819, "learning_rate": 5.293186351924842e-06, "loss": 0.3962, "step": 7009 }, { "epoch": 0.49698688408365826, "grad_norm": 2.4255027160502745, "learning_rate": 5.29204016077985e-06, "loss": 0.3664, "step": 7010 }, { "epoch": 0.4970577809287487, "grad_norm": 1.0512733318445604, "learning_rate": 5.290893954235359e-06, "loss": 0.6634, "step": 7011 }, { "epoch": 0.49712867777383907, "grad_norm": 6.360103003564674, "learning_rate": 5.289747732351805e-06, "loss": 0.4128, "step": 7012 }, { "epoch": 0.49719957461892944, "grad_norm": 0.9873522201209997, "learning_rate": 5.288601495189633e-06, "loss": 0.6722, "step": 7013 }, { "epoch": 0.49727047146401987, "grad_norm": 2.4748314332440073, "learning_rate": 5.287455242809284e-06, "loss": 0.3861, "step": 7014 }, { "epoch": 0.49734136830911024, "grad_norm": 2.5992205741444736, "learning_rate": 5.2863089752712e-06, "loss": 0.3764, "step": 7015 }, { "epoch": 0.4974122651542006, "grad_norm": 2.8979765894899625, "learning_rate": 5.285162692635828e-06, "loss": 0.382, "step": 7016 }, { "epoch": 0.49748316199929105, "grad_norm": 2.4574867767104536, "learning_rate": 5.2840163949636055e-06, "loss": 0.4136, "step": 7017 }, { "epoch": 0.4975540588443814, "grad_norm": 6.554000152387931, "learning_rate": 5.282870082314985e-06, "loss": 0.4635, "step": 7018 }, { "epoch": 0.4976249556894718, "grad_norm": 3.8732592279777385, "learning_rate": 5.281723754750406e-06, "loss": 0.3868, "step": 7019 }, { "epoch": 0.4976958525345622, "grad_norm": 2.680454579464664, "learning_rate": 5.280577412330322e-06, "loss": 0.4158, "step": 7020 }, { "epoch": 0.4977667493796526, "grad_norm": 3.9051472887172434, "learning_rate": 5.279431055115175e-06, "loss": 0.4083, "step": 7021 }, { "epoch": 0.497837646224743, "grad_norm": 3.3525990124795304, "learning_rate": 5.278284683165414e-06, "loss": 0.4199, "step": 7022 }, { "epoch": 0.4979085430698334, "grad_norm": 1.9190205340185664, "learning_rate": 5.277138296541491e-06, "loss": 0.3906, "step": 7023 }, { "epoch": 0.49797943991492377, "grad_norm": 2.926047548077274, "learning_rate": 5.275991895303854e-06, "loss": 0.3783, "step": 7024 }, { "epoch": 0.4980503367600142, "grad_norm": 4.041993489616516, "learning_rate": 5.274845479512953e-06, "loss": 0.4503, "step": 7025 }, { "epoch": 0.4981212336051046, "grad_norm": 1.9692309384906928, "learning_rate": 5.273699049229242e-06, "loss": 0.3794, "step": 7026 }, { "epoch": 0.49819213045019495, "grad_norm": 2.978845222537558, "learning_rate": 5.2725526045131696e-06, "loss": 0.4972, "step": 7027 }, { "epoch": 0.4982630272952854, "grad_norm": 2.3025725971791076, "learning_rate": 5.27140614542519e-06, "loss": 0.4056, "step": 7028 }, { "epoch": 0.49833392414037575, "grad_norm": 2.6893149659311852, "learning_rate": 5.270259672025758e-06, "loss": 0.3467, "step": 7029 }, { "epoch": 0.4984048209854661, "grad_norm": 14.039165236347227, "learning_rate": 5.269113184375325e-06, "loss": 0.4809, "step": 7030 }, { "epoch": 0.49847571783055655, "grad_norm": 8.463997353070528, "learning_rate": 5.267966682534351e-06, "loss": 0.4063, "step": 7031 }, { "epoch": 0.4985466146756469, "grad_norm": 2.123674983487077, "learning_rate": 5.2668201665632886e-06, "loss": 0.3599, "step": 7032 }, { "epoch": 0.4986175115207373, "grad_norm": 2.0401653148931813, "learning_rate": 5.265673636522595e-06, "loss": 0.4295, "step": 7033 }, { "epoch": 0.49868840836582773, "grad_norm": 5.210912621546304, "learning_rate": 5.264527092472728e-06, "loss": 0.3733, "step": 7034 }, { "epoch": 0.4987593052109181, "grad_norm": 2.6852128208229953, "learning_rate": 5.263380534474145e-06, "loss": 0.3983, "step": 7035 }, { "epoch": 0.49883020205600853, "grad_norm": 3.575840350827192, "learning_rate": 5.262233962587305e-06, "loss": 0.3949, "step": 7036 }, { "epoch": 0.4989010989010989, "grad_norm": 2.514741813646079, "learning_rate": 5.26108737687267e-06, "loss": 0.4003, "step": 7037 }, { "epoch": 0.4989719957461893, "grad_norm": 3.462854234559453, "learning_rate": 5.259940777390696e-06, "loss": 0.4384, "step": 7038 }, { "epoch": 0.4990428925912797, "grad_norm": 2.925429179841174, "learning_rate": 5.258794164201848e-06, "loss": 0.4095, "step": 7039 }, { "epoch": 0.4991137894363701, "grad_norm": 3.265053007239825, "learning_rate": 5.257647537366586e-06, "loss": 0.4219, "step": 7040 }, { "epoch": 0.49918468628146045, "grad_norm": 3.82893615785814, "learning_rate": 5.256500896945372e-06, "loss": 0.4545, "step": 7041 }, { "epoch": 0.4992555831265509, "grad_norm": 3.8695484203616695, "learning_rate": 5.25535424299867e-06, "loss": 0.3849, "step": 7042 }, { "epoch": 0.49932647997164126, "grad_norm": 2.97104350276389, "learning_rate": 5.2542075755869435e-06, "loss": 0.4023, "step": 7043 }, { "epoch": 0.49939737681673163, "grad_norm": 4.037846671603316, "learning_rate": 5.25306089477066e-06, "loss": 0.4353, "step": 7044 }, { "epoch": 0.49946827366182206, "grad_norm": 5.824831613272179, "learning_rate": 5.25191420061028e-06, "loss": 0.3689, "step": 7045 }, { "epoch": 0.49953917050691243, "grad_norm": 1.1481228305540758, "learning_rate": 5.250767493166272e-06, "loss": 0.6417, "step": 7046 }, { "epoch": 0.49961006735200286, "grad_norm": 1.073732718452789, "learning_rate": 5.249620772499103e-06, "loss": 0.6232, "step": 7047 }, { "epoch": 0.49968096419709324, "grad_norm": 7.905064617397779, "learning_rate": 5.24847403866924e-06, "loss": 0.4698, "step": 7048 }, { "epoch": 0.4997518610421836, "grad_norm": 6.588677835805558, "learning_rate": 5.247327291737152e-06, "loss": 0.4038, "step": 7049 }, { "epoch": 0.49982275788727404, "grad_norm": 3.186887033639524, "learning_rate": 5.246180531763308e-06, "loss": 0.4195, "step": 7050 }, { "epoch": 0.4998936547323644, "grad_norm": 3.9847156987484484, "learning_rate": 5.245033758808173e-06, "loss": 0.3777, "step": 7051 }, { "epoch": 0.4999645515774548, "grad_norm": 5.313735985526722, "learning_rate": 5.243886972932225e-06, "loss": 0.4304, "step": 7052 }, { "epoch": 0.5000354484225452, "grad_norm": 3.7750254007933064, "learning_rate": 5.242740174195928e-06, "loss": 0.4471, "step": 7053 }, { "epoch": 0.5001063452676356, "grad_norm": 2.950648546497201, "learning_rate": 5.241593362659758e-06, "loss": 0.4503, "step": 7054 }, { "epoch": 0.500177242112726, "grad_norm": 4.260379675436961, "learning_rate": 5.240446538384184e-06, "loss": 0.4051, "step": 7055 }, { "epoch": 0.5002481389578164, "grad_norm": 2.899189331951099, "learning_rate": 5.239299701429682e-06, "loss": 0.4841, "step": 7056 }, { "epoch": 0.5003190358029068, "grad_norm": 3.4445502034154662, "learning_rate": 5.238152851856724e-06, "loss": 0.3509, "step": 7057 }, { "epoch": 0.5003899326479971, "grad_norm": 4.477607320167519, "learning_rate": 5.237005989725784e-06, "loss": 0.4152, "step": 7058 }, { "epoch": 0.5004608294930876, "grad_norm": 1.9813255857179755, "learning_rate": 5.235859115097337e-06, "loss": 0.3889, "step": 7059 }, { "epoch": 0.500531726338178, "grad_norm": 1.7818875125676297, "learning_rate": 5.2347122280318585e-06, "loss": 0.4247, "step": 7060 }, { "epoch": 0.5006026231832683, "grad_norm": 2.916567116299638, "learning_rate": 5.233565328589827e-06, "loss": 0.4761, "step": 7061 }, { "epoch": 0.5006735200283587, "grad_norm": 2.3845863000508483, "learning_rate": 5.2324184168317135e-06, "loss": 0.4106, "step": 7062 }, { "epoch": 0.5007444168734492, "grad_norm": 2.3255397986099005, "learning_rate": 5.231271492818003e-06, "loss": 0.4754, "step": 7063 }, { "epoch": 0.5008153137185395, "grad_norm": 4.367708627016974, "learning_rate": 5.230124556609168e-06, "loss": 0.3549, "step": 7064 }, { "epoch": 0.5008862105636299, "grad_norm": 3.0362593555947344, "learning_rate": 5.228977608265692e-06, "loss": 0.4403, "step": 7065 }, { "epoch": 0.5009571074087203, "grad_norm": 3.499978681988792, "learning_rate": 5.22783064784805e-06, "loss": 0.4219, "step": 7066 }, { "epoch": 0.5010280042538107, "grad_norm": 3.9588947177196188, "learning_rate": 5.226683675416725e-06, "loss": 0.4057, "step": 7067 }, { "epoch": 0.5010989010989011, "grad_norm": 8.71416021961352, "learning_rate": 5.225536691032196e-06, "loss": 0.4315, "step": 7068 }, { "epoch": 0.5011697979439915, "grad_norm": 3.2455031975103443, "learning_rate": 5.2243896947549465e-06, "loss": 0.4423, "step": 7069 }, { "epoch": 0.5012406947890818, "grad_norm": 5.975414082353539, "learning_rate": 5.223242686645455e-06, "loss": 0.3918, "step": 7070 }, { "epoch": 0.5013115916341723, "grad_norm": 4.552566412666664, "learning_rate": 5.222095666764207e-06, "loss": 0.4045, "step": 7071 }, { "epoch": 0.5013824884792627, "grad_norm": 4.7573816904177315, "learning_rate": 5.220948635171687e-06, "loss": 0.4202, "step": 7072 }, { "epoch": 0.501453385324353, "grad_norm": 5.825862171830802, "learning_rate": 5.219801591928374e-06, "loss": 0.4159, "step": 7073 }, { "epoch": 0.5015242821694434, "grad_norm": 2.349975690451426, "learning_rate": 5.218654537094756e-06, "loss": 0.4378, "step": 7074 }, { "epoch": 0.5015951790145339, "grad_norm": 2.331734267188722, "learning_rate": 5.217507470731318e-06, "loss": 0.4481, "step": 7075 }, { "epoch": 0.5016660758596242, "grad_norm": 2.983973166229428, "learning_rate": 5.216360392898546e-06, "loss": 0.4052, "step": 7076 }, { "epoch": 0.5017369727047146, "grad_norm": 5.5178029943319435, "learning_rate": 5.215213303656924e-06, "loss": 0.3903, "step": 7077 }, { "epoch": 0.501807869549805, "grad_norm": 2.7216474782724025, "learning_rate": 5.214066203066941e-06, "loss": 0.4285, "step": 7078 }, { "epoch": 0.5018787663948955, "grad_norm": 2.14978259437032, "learning_rate": 5.212919091189083e-06, "loss": 0.3993, "step": 7079 }, { "epoch": 0.5019496632399858, "grad_norm": 4.27597742943235, "learning_rate": 5.211771968083838e-06, "loss": 0.4106, "step": 7080 }, { "epoch": 0.5020205600850762, "grad_norm": 4.308192654837373, "learning_rate": 5.210624833811696e-06, "loss": 0.4314, "step": 7081 }, { "epoch": 0.5020914569301667, "grad_norm": 2.730540913963135, "learning_rate": 5.2094776884331475e-06, "loss": 0.3744, "step": 7082 }, { "epoch": 0.502162353775257, "grad_norm": 2.5061624337661685, "learning_rate": 5.208330532008679e-06, "loss": 0.4212, "step": 7083 }, { "epoch": 0.5022332506203474, "grad_norm": 2.6167818308566515, "learning_rate": 5.207183364598784e-06, "loss": 0.4301, "step": 7084 }, { "epoch": 0.5023041474654378, "grad_norm": 4.114026941988515, "learning_rate": 5.206036186263949e-06, "loss": 0.4199, "step": 7085 }, { "epoch": 0.5023750443105282, "grad_norm": 2.4242529562150885, "learning_rate": 5.204888997064671e-06, "loss": 0.3835, "step": 7086 }, { "epoch": 0.5024459411556186, "grad_norm": 1.2794648045656642, "learning_rate": 5.203741797061439e-06, "loss": 0.649, "step": 7087 }, { "epoch": 0.502516838000709, "grad_norm": 3.6359267156489885, "learning_rate": 5.202594586314748e-06, "loss": 0.3715, "step": 7088 }, { "epoch": 0.5025877348457993, "grad_norm": 4.882532127409277, "learning_rate": 5.201447364885089e-06, "loss": 0.4415, "step": 7089 }, { "epoch": 0.5026586316908898, "grad_norm": 2.7157702641959336, "learning_rate": 5.200300132832956e-06, "loss": 0.3934, "step": 7090 }, { "epoch": 0.5027295285359802, "grad_norm": 2.781684733663149, "learning_rate": 5.199152890218845e-06, "loss": 0.4084, "step": 7091 }, { "epoch": 0.5028004253810705, "grad_norm": 3.3619535515387495, "learning_rate": 5.198005637103248e-06, "loss": 0.4483, "step": 7092 }, { "epoch": 0.5028713222261609, "grad_norm": 3.301560128892921, "learning_rate": 5.196858373546665e-06, "loss": 0.4413, "step": 7093 }, { "epoch": 0.5029422190712514, "grad_norm": 2.552139691749558, "learning_rate": 5.195711099609588e-06, "loss": 0.3903, "step": 7094 }, { "epoch": 0.5030131159163417, "grad_norm": 2.2346742889022133, "learning_rate": 5.194563815352518e-06, "loss": 0.4358, "step": 7095 }, { "epoch": 0.5030840127614321, "grad_norm": 3.08601066966567, "learning_rate": 5.193416520835947e-06, "loss": 0.4489, "step": 7096 }, { "epoch": 0.5031549096065225, "grad_norm": 4.350046143387101, "learning_rate": 5.192269216120376e-06, "loss": 0.3912, "step": 7097 }, { "epoch": 0.5032258064516129, "grad_norm": 5.0615942451219125, "learning_rate": 5.191121901266303e-06, "loss": 0.4003, "step": 7098 }, { "epoch": 0.5032967032967033, "grad_norm": 2.5597376707165287, "learning_rate": 5.189974576334226e-06, "loss": 0.3624, "step": 7099 }, { "epoch": 0.5033676001417937, "grad_norm": 3.495327634393523, "learning_rate": 5.188827241384643e-06, "loss": 0.4263, "step": 7100 }, { "epoch": 0.503438496986884, "grad_norm": 1.85075872201907, "learning_rate": 5.187679896478057e-06, "loss": 0.3374, "step": 7101 }, { "epoch": 0.5035093938319745, "grad_norm": 2.375319536963551, "learning_rate": 5.186532541674967e-06, "loss": 0.4084, "step": 7102 }, { "epoch": 0.5035802906770649, "grad_norm": 2.4515369574835195, "learning_rate": 5.185385177035875e-06, "loss": 0.3709, "step": 7103 }, { "epoch": 0.5036511875221553, "grad_norm": 5.562233681619045, "learning_rate": 5.184237802621278e-06, "loss": 0.4359, "step": 7104 }, { "epoch": 0.5037220843672456, "grad_norm": 2.7057704521916586, "learning_rate": 5.183090418491682e-06, "loss": 0.4003, "step": 7105 }, { "epoch": 0.5037929812123361, "grad_norm": 10.882637218214139, "learning_rate": 5.181943024707589e-06, "loss": 0.4314, "step": 7106 }, { "epoch": 0.5038638780574265, "grad_norm": 2.53678636168163, "learning_rate": 5.180795621329502e-06, "loss": 0.4068, "step": 7107 }, { "epoch": 0.5039347749025168, "grad_norm": 3.7987772727000912, "learning_rate": 5.179648208417923e-06, "loss": 0.3996, "step": 7108 }, { "epoch": 0.5040056717476072, "grad_norm": 2.222628629226957, "learning_rate": 5.178500786033357e-06, "loss": 0.4387, "step": 7109 }, { "epoch": 0.5040765685926977, "grad_norm": 3.158873064932399, "learning_rate": 5.177353354236308e-06, "loss": 0.4442, "step": 7110 }, { "epoch": 0.504147465437788, "grad_norm": 4.2904526981749695, "learning_rate": 5.176205913087282e-06, "loss": 0.3578, "step": 7111 }, { "epoch": 0.5042183622828784, "grad_norm": 5.328492181069249, "learning_rate": 5.1750584626467835e-06, "loss": 0.4207, "step": 7112 }, { "epoch": 0.5042892591279688, "grad_norm": 2.977187779004918, "learning_rate": 5.173911002975318e-06, "loss": 0.4167, "step": 7113 }, { "epoch": 0.5043601559730592, "grad_norm": 1.7562016104811844, "learning_rate": 5.172763534133394e-06, "loss": 0.3649, "step": 7114 }, { "epoch": 0.5044310528181496, "grad_norm": 2.1474022717113894, "learning_rate": 5.171616056181515e-06, "loss": 0.4046, "step": 7115 }, { "epoch": 0.50450194966324, "grad_norm": 2.916527941178054, "learning_rate": 5.170468569180192e-06, "loss": 0.4064, "step": 7116 }, { "epoch": 0.5045728465083303, "grad_norm": 2.6955238598067246, "learning_rate": 5.169321073189931e-06, "loss": 0.4328, "step": 7117 }, { "epoch": 0.5046437433534208, "grad_norm": 3.6999080757658773, "learning_rate": 5.168173568271241e-06, "loss": 0.42, "step": 7118 }, { "epoch": 0.5047146401985112, "grad_norm": 4.528256766795519, "learning_rate": 5.16702605448463e-06, "loss": 0.3514, "step": 7119 }, { "epoch": 0.5047855370436015, "grad_norm": 9.28562105265379, "learning_rate": 5.165878531890607e-06, "loss": 0.3875, "step": 7120 }, { "epoch": 0.504856433888692, "grad_norm": 1.0888249219348103, "learning_rate": 5.164731000549683e-06, "loss": 0.714, "step": 7121 }, { "epoch": 0.5049273307337824, "grad_norm": 4.1994679149011525, "learning_rate": 5.163583460522366e-06, "loss": 0.4535, "step": 7122 }, { "epoch": 0.5049982275788727, "grad_norm": 3.870986735273533, "learning_rate": 5.16243591186917e-06, "loss": 0.4294, "step": 7123 }, { "epoch": 0.5050691244239631, "grad_norm": 3.6703948460434472, "learning_rate": 5.161288354650604e-06, "loss": 0.4248, "step": 7124 }, { "epoch": 0.5051400212690536, "grad_norm": 3.8845867557007865, "learning_rate": 5.16014078892718e-06, "loss": 0.3678, "step": 7125 }, { "epoch": 0.5052109181141439, "grad_norm": 2.5095781732668714, "learning_rate": 5.15899321475941e-06, "loss": 0.4216, "step": 7126 }, { "epoch": 0.5052818149592343, "grad_norm": 5.554812547733243, "learning_rate": 5.157845632207807e-06, "loss": 0.4436, "step": 7127 }, { "epoch": 0.5053527118043247, "grad_norm": 3.358128675387739, "learning_rate": 5.1566980413328815e-06, "loss": 0.3997, "step": 7128 }, { "epoch": 0.505423608649415, "grad_norm": 2.0002601945206075, "learning_rate": 5.155550442195151e-06, "loss": 0.3818, "step": 7129 }, { "epoch": 0.5054945054945055, "grad_norm": 2.4138306362419515, "learning_rate": 5.154402834855125e-06, "loss": 0.4327, "step": 7130 }, { "epoch": 0.5055654023395959, "grad_norm": 5.530363350373073, "learning_rate": 5.153255219373321e-06, "loss": 0.3888, "step": 7131 }, { "epoch": 0.5056362991846863, "grad_norm": 2.3097548829764016, "learning_rate": 5.152107595810252e-06, "loss": 0.3763, "step": 7132 }, { "epoch": 0.5057071960297767, "grad_norm": 5.12720236077063, "learning_rate": 5.150959964226433e-06, "loss": 0.4036, "step": 7133 }, { "epoch": 0.5057780928748671, "grad_norm": 6.518515154730229, "learning_rate": 5.1498123246823795e-06, "loss": 0.4118, "step": 7134 }, { "epoch": 0.5058489897199575, "grad_norm": 2.088128907774272, "learning_rate": 5.148664677238609e-06, "loss": 0.4033, "step": 7135 }, { "epoch": 0.5059198865650478, "grad_norm": 3.8477259619755393, "learning_rate": 5.147517021955636e-06, "loss": 0.4246, "step": 7136 }, { "epoch": 0.5059907834101383, "grad_norm": 2.2853702196486436, "learning_rate": 5.146369358893978e-06, "loss": 0.4166, "step": 7137 }, { "epoch": 0.5060616802552287, "grad_norm": 3.0825947137103804, "learning_rate": 5.1452216881141514e-06, "loss": 0.4173, "step": 7138 }, { "epoch": 0.506132577100319, "grad_norm": 2.2198081189439036, "learning_rate": 5.144074009676675e-06, "loss": 0.4197, "step": 7139 }, { "epoch": 0.5062034739454094, "grad_norm": 3.4150435233170553, "learning_rate": 5.1429263236420664e-06, "loss": 0.3959, "step": 7140 }, { "epoch": 0.5062743707904999, "grad_norm": 4.631914276362772, "learning_rate": 5.141778630070842e-06, "loss": 0.4176, "step": 7141 }, { "epoch": 0.5063452676355902, "grad_norm": 2.22545118033296, "learning_rate": 5.140630929023523e-06, "loss": 0.431, "step": 7142 }, { "epoch": 0.5064161644806806, "grad_norm": 5.564931305403875, "learning_rate": 5.1394832205606284e-06, "loss": 0.3866, "step": 7143 }, { "epoch": 0.506487061325771, "grad_norm": 2.956227804856209, "learning_rate": 5.138335504742677e-06, "loss": 0.3896, "step": 7144 }, { "epoch": 0.5065579581708614, "grad_norm": 4.0613198374288455, "learning_rate": 5.1371877816301884e-06, "loss": 0.3979, "step": 7145 }, { "epoch": 0.5066288550159518, "grad_norm": 3.106095876781769, "learning_rate": 5.136040051283684e-06, "loss": 0.3569, "step": 7146 }, { "epoch": 0.5066997518610422, "grad_norm": 3.091228579474655, "learning_rate": 5.134892313763683e-06, "loss": 0.3742, "step": 7147 }, { "epoch": 0.5067706487061325, "grad_norm": 2.3199568690027603, "learning_rate": 5.133744569130709e-06, "loss": 0.369, "step": 7148 }, { "epoch": 0.506841545551223, "grad_norm": 2.6843042709502463, "learning_rate": 5.13259681744528e-06, "loss": 0.4147, "step": 7149 }, { "epoch": 0.5069124423963134, "grad_norm": 3.6776849126308395, "learning_rate": 5.13144905876792e-06, "loss": 0.4625, "step": 7150 }, { "epoch": 0.5069833392414037, "grad_norm": 2.0104341588068224, "learning_rate": 5.130301293159151e-06, "loss": 0.425, "step": 7151 }, { "epoch": 0.5070542360864941, "grad_norm": 2.8003154265374706, "learning_rate": 5.129153520679495e-06, "loss": 0.436, "step": 7152 }, { "epoch": 0.5071251329315846, "grad_norm": 2.8205971730622768, "learning_rate": 5.128005741389477e-06, "loss": 0.4063, "step": 7153 }, { "epoch": 0.5071960297766749, "grad_norm": 2.584158474478066, "learning_rate": 5.1268579553496154e-06, "loss": 0.4049, "step": 7154 }, { "epoch": 0.5072669266217653, "grad_norm": 6.532018288651301, "learning_rate": 5.12571016262044e-06, "loss": 0.3848, "step": 7155 }, { "epoch": 0.5073378234668557, "grad_norm": 7.863327097658782, "learning_rate": 5.12456236326247e-06, "loss": 0.4379, "step": 7156 }, { "epoch": 0.5074087203119462, "grad_norm": 4.509483387333518, "learning_rate": 5.1234145573362315e-06, "loss": 0.3752, "step": 7157 }, { "epoch": 0.5074796171570365, "grad_norm": 2.2175640913285406, "learning_rate": 5.12226674490225e-06, "loss": 0.4089, "step": 7158 }, { "epoch": 0.5075505140021269, "grad_norm": 2.5889867964389137, "learning_rate": 5.121118926021049e-06, "loss": 0.4506, "step": 7159 }, { "epoch": 0.5076214108472173, "grad_norm": 1.9981922469847464, "learning_rate": 5.119971100753155e-06, "loss": 0.4025, "step": 7160 }, { "epoch": 0.5076923076923077, "grad_norm": 3.427412740735332, "learning_rate": 5.1188232691590944e-06, "loss": 0.3807, "step": 7161 }, { "epoch": 0.5077632045373981, "grad_norm": 2.924662504502846, "learning_rate": 5.117675431299389e-06, "loss": 0.3433, "step": 7162 }, { "epoch": 0.5078341013824885, "grad_norm": 2.9532739168589703, "learning_rate": 5.1165275872345725e-06, "loss": 0.4257, "step": 7163 }, { "epoch": 0.5079049982275788, "grad_norm": 2.9465565990441025, "learning_rate": 5.115379737025164e-06, "loss": 0.4091, "step": 7164 }, { "epoch": 0.5079758950726693, "grad_norm": 2.9937906139933275, "learning_rate": 5.1142318807316956e-06, "loss": 0.4346, "step": 7165 }, { "epoch": 0.5080467919177597, "grad_norm": 3.280327212597752, "learning_rate": 5.113084018414693e-06, "loss": 0.4103, "step": 7166 }, { "epoch": 0.50811768876285, "grad_norm": 4.117348015957243, "learning_rate": 5.111936150134682e-06, "loss": 0.4483, "step": 7167 }, { "epoch": 0.5081885856079404, "grad_norm": 7.323052918483795, "learning_rate": 5.110788275952194e-06, "loss": 0.3858, "step": 7168 }, { "epoch": 0.5082594824530309, "grad_norm": 8.87278008914507, "learning_rate": 5.109640395927756e-06, "loss": 0.4091, "step": 7169 }, { "epoch": 0.5083303792981212, "grad_norm": 3.9419406012915275, "learning_rate": 5.108492510121895e-06, "loss": 0.4176, "step": 7170 }, { "epoch": 0.5084012761432116, "grad_norm": 3.2205429071149188, "learning_rate": 5.1073446185951415e-06, "loss": 0.3864, "step": 7171 }, { "epoch": 0.508472172988302, "grad_norm": 5.66366099817177, "learning_rate": 5.106196721408026e-06, "loss": 0.3918, "step": 7172 }, { "epoch": 0.5085430698333924, "grad_norm": 7.792488918944731, "learning_rate": 5.1050488186210735e-06, "loss": 0.4368, "step": 7173 }, { "epoch": 0.5086139666784828, "grad_norm": 2.2397855825605344, "learning_rate": 5.103900910294819e-06, "loss": 0.3922, "step": 7174 }, { "epoch": 0.5086848635235732, "grad_norm": 4.103351308559366, "learning_rate": 5.102752996489787e-06, "loss": 0.3618, "step": 7175 }, { "epoch": 0.5087557603686635, "grad_norm": 4.34916428005877, "learning_rate": 5.101605077266515e-06, "loss": 0.3894, "step": 7176 }, { "epoch": 0.508826657213754, "grad_norm": 12.712635400164178, "learning_rate": 5.100457152685528e-06, "loss": 0.3956, "step": 7177 }, { "epoch": 0.5088975540588444, "grad_norm": 3.4581698244838113, "learning_rate": 5.099309222807357e-06, "loss": 0.3949, "step": 7178 }, { "epoch": 0.5089684509039347, "grad_norm": 53.00523321085515, "learning_rate": 5.098161287692537e-06, "loss": 0.4551, "step": 7179 }, { "epoch": 0.5090393477490252, "grad_norm": 1.0144517883572115, "learning_rate": 5.0970133474015985e-06, "loss": 0.6835, "step": 7180 }, { "epoch": 0.5091102445941156, "grad_norm": 2.397400809719563, "learning_rate": 5.09586540199507e-06, "loss": 0.4365, "step": 7181 }, { "epoch": 0.509181141439206, "grad_norm": 1.1407719077173806, "learning_rate": 5.0947174515334855e-06, "loss": 0.6783, "step": 7182 }, { "epoch": 0.5092520382842963, "grad_norm": 2.7310241779793616, "learning_rate": 5.0935694960773774e-06, "loss": 0.4667, "step": 7183 }, { "epoch": 0.5093229351293868, "grad_norm": 3.8031450721039257, "learning_rate": 5.09242153568728e-06, "loss": 0.4375, "step": 7184 }, { "epoch": 0.5093938319744772, "grad_norm": 2.643837180469065, "learning_rate": 5.0912735704237235e-06, "loss": 0.4581, "step": 7185 }, { "epoch": 0.5094647288195675, "grad_norm": 2.0242135191271937, "learning_rate": 5.090125600347241e-06, "loss": 0.395, "step": 7186 }, { "epoch": 0.5095356256646579, "grad_norm": 3.6563699170882025, "learning_rate": 5.0889776255183684e-06, "loss": 0.4358, "step": 7187 }, { "epoch": 0.5096065225097484, "grad_norm": 5.0050334789338775, "learning_rate": 5.087829645997636e-06, "loss": 0.413, "step": 7188 }, { "epoch": 0.5096774193548387, "grad_norm": 16.554696811142204, "learning_rate": 5.086681661845581e-06, "loss": 0.392, "step": 7189 }, { "epoch": 0.5097483161999291, "grad_norm": 2.4503927650957618, "learning_rate": 5.085533673122735e-06, "loss": 0.4191, "step": 7190 }, { "epoch": 0.5098192130450195, "grad_norm": 5.252853210358294, "learning_rate": 5.084385679889634e-06, "loss": 0.4134, "step": 7191 }, { "epoch": 0.5098901098901099, "grad_norm": 4.811440503788418, "learning_rate": 5.083237682206811e-06, "loss": 0.4194, "step": 7192 }, { "epoch": 0.5099610067352003, "grad_norm": 3.766076475500288, "learning_rate": 5.082089680134803e-06, "loss": 0.3963, "step": 7193 }, { "epoch": 0.5100319035802907, "grad_norm": 4.529564183822903, "learning_rate": 5.08094167373414e-06, "loss": 0.4959, "step": 7194 }, { "epoch": 0.510102800425381, "grad_norm": 3.0019693598765755, "learning_rate": 5.079793663065364e-06, "loss": 0.3951, "step": 7195 }, { "epoch": 0.5101736972704715, "grad_norm": 7.6461115138989, "learning_rate": 5.078645648189006e-06, "loss": 0.3978, "step": 7196 }, { "epoch": 0.5102445941155619, "grad_norm": 13.161192493641735, "learning_rate": 5.077497629165603e-06, "loss": 0.3809, "step": 7197 }, { "epoch": 0.5103154909606522, "grad_norm": 3.2998080481463368, "learning_rate": 5.076349606055691e-06, "loss": 0.4243, "step": 7198 }, { "epoch": 0.5103863878057426, "grad_norm": 2.0149242990070935, "learning_rate": 5.075201578919805e-06, "loss": 0.4405, "step": 7199 }, { "epoch": 0.5104572846508331, "grad_norm": 4.138889233835835, "learning_rate": 5.074053547818485e-06, "loss": 0.4058, "step": 7200 }, { "epoch": 0.5105281814959234, "grad_norm": 7.42355696350636, "learning_rate": 5.072905512812263e-06, "loss": 0.4669, "step": 7201 }, { "epoch": 0.5105990783410138, "grad_norm": 3.6448537612149323, "learning_rate": 5.071757473961677e-06, "loss": 0.3909, "step": 7202 }, { "epoch": 0.5106699751861042, "grad_norm": 5.611689478440256, "learning_rate": 5.070609431327266e-06, "loss": 0.3901, "step": 7203 }, { "epoch": 0.5107408720311946, "grad_norm": 3.8642493756459726, "learning_rate": 5.069461384969565e-06, "loss": 0.3826, "step": 7204 }, { "epoch": 0.510811768876285, "grad_norm": 2.5744395897856744, "learning_rate": 5.068313334949112e-06, "loss": 0.4617, "step": 7205 }, { "epoch": 0.5108826657213754, "grad_norm": 3.44182343076547, "learning_rate": 5.067165281326446e-06, "loss": 0.3843, "step": 7206 }, { "epoch": 0.5109535625664658, "grad_norm": 2.4407568275476508, "learning_rate": 5.066017224162101e-06, "loss": 0.4469, "step": 7207 }, { "epoch": 0.5110244594115562, "grad_norm": 10.631748835161732, "learning_rate": 5.06486916351662e-06, "loss": 0.442, "step": 7208 }, { "epoch": 0.5110953562566466, "grad_norm": 3.0863707677539396, "learning_rate": 5.063721099450537e-06, "loss": 0.3692, "step": 7209 }, { "epoch": 0.511166253101737, "grad_norm": 5.2885500716496, "learning_rate": 5.062573032024393e-06, "loss": 0.4117, "step": 7210 }, { "epoch": 0.5112371499468273, "grad_norm": 3.13203501850263, "learning_rate": 5.061424961298723e-06, "loss": 0.3926, "step": 7211 }, { "epoch": 0.5113080467919178, "grad_norm": 4.777532744589399, "learning_rate": 5.06027688733407e-06, "loss": 0.4858, "step": 7212 }, { "epoch": 0.5113789436370082, "grad_norm": 3.202835545213897, "learning_rate": 5.059128810190971e-06, "loss": 0.4178, "step": 7213 }, { "epoch": 0.5114498404820985, "grad_norm": 3.3026157691114375, "learning_rate": 5.057980729929963e-06, "loss": 0.4097, "step": 7214 }, { "epoch": 0.511520737327189, "grad_norm": 4.653142048177227, "learning_rate": 5.056832646611588e-06, "loss": 0.3661, "step": 7215 }, { "epoch": 0.5115916341722794, "grad_norm": 3.3099274444873137, "learning_rate": 5.055684560296383e-06, "loss": 0.4107, "step": 7216 }, { "epoch": 0.5116625310173697, "grad_norm": 3.1200169131459816, "learning_rate": 5.05453647104489e-06, "loss": 0.4508, "step": 7217 }, { "epoch": 0.5117334278624601, "grad_norm": 6.410457920887237, "learning_rate": 5.0533883789176464e-06, "loss": 0.392, "step": 7218 }, { "epoch": 0.5118043247075506, "grad_norm": 2.8182007021015423, "learning_rate": 5.0522402839751935e-06, "loss": 0.4336, "step": 7219 }, { "epoch": 0.5118752215526409, "grad_norm": 2.2422906068358754, "learning_rate": 5.051092186278069e-06, "loss": 0.467, "step": 7220 }, { "epoch": 0.5119461183977313, "grad_norm": 2.52694505691983, "learning_rate": 5.0499440858868175e-06, "loss": 0.3846, "step": 7221 }, { "epoch": 0.5120170152428217, "grad_norm": 4.81461016438799, "learning_rate": 5.048795982861974e-06, "loss": 0.3971, "step": 7222 }, { "epoch": 0.512087912087912, "grad_norm": 7.902069270127829, "learning_rate": 5.047647877264082e-06, "loss": 0.4063, "step": 7223 }, { "epoch": 0.5121588089330025, "grad_norm": 4.378526364269726, "learning_rate": 5.04649976915368e-06, "loss": 0.4738, "step": 7224 }, { "epoch": 0.5122297057780929, "grad_norm": 11.896885994520673, "learning_rate": 5.04535165859131e-06, "loss": 0.3532, "step": 7225 }, { "epoch": 0.5123006026231832, "grad_norm": 2.220824557644997, "learning_rate": 5.044203545637514e-06, "loss": 0.3289, "step": 7226 }, { "epoch": 0.5123714994682737, "grad_norm": 8.511739293110953, "learning_rate": 5.0430554303528286e-06, "loss": 0.4156, "step": 7227 }, { "epoch": 0.5124423963133641, "grad_norm": 3.0679166498173283, "learning_rate": 5.041907312797799e-06, "loss": 0.4078, "step": 7228 }, { "epoch": 0.5125132931584544, "grad_norm": 2.210322611855946, "learning_rate": 5.040759193032965e-06, "loss": 0.4271, "step": 7229 }, { "epoch": 0.5125841900035448, "grad_norm": 4.737332545509095, "learning_rate": 5.039611071118868e-06, "loss": 0.4266, "step": 7230 }, { "epoch": 0.5126550868486353, "grad_norm": 3.337449222126819, "learning_rate": 5.038462947116047e-06, "loss": 0.4181, "step": 7231 }, { "epoch": 0.5127259836937257, "grad_norm": 3.3381250573513586, "learning_rate": 5.037314821085048e-06, "loss": 0.4124, "step": 7232 }, { "epoch": 0.512796880538816, "grad_norm": 3.2800313613706322, "learning_rate": 5.036166693086407e-06, "loss": 0.4316, "step": 7233 }, { "epoch": 0.5128677773839064, "grad_norm": 3.4919330756076525, "learning_rate": 5.035018563180673e-06, "loss": 0.4112, "step": 7234 }, { "epoch": 0.5129386742289969, "grad_norm": 5.009161206653452, "learning_rate": 5.0338704314283795e-06, "loss": 0.3765, "step": 7235 }, { "epoch": 0.5130095710740872, "grad_norm": 1.0313284029557874, "learning_rate": 5.032722297890073e-06, "loss": 0.6653, "step": 7236 }, { "epoch": 0.5130804679191776, "grad_norm": 2.455693925886979, "learning_rate": 5.031574162626296e-06, "loss": 0.3799, "step": 7237 }, { "epoch": 0.513151364764268, "grad_norm": 4.009249958918336, "learning_rate": 5.03042602569759e-06, "loss": 0.3951, "step": 7238 }, { "epoch": 0.5132222616093584, "grad_norm": 4.7756049830342535, "learning_rate": 5.029277887164493e-06, "loss": 0.4299, "step": 7239 }, { "epoch": 0.5132931584544488, "grad_norm": 4.978430218881542, "learning_rate": 5.028129747087554e-06, "loss": 0.4276, "step": 7240 }, { "epoch": 0.5133640552995392, "grad_norm": 4.162199257919421, "learning_rate": 5.0269816055273114e-06, "loss": 0.3953, "step": 7241 }, { "epoch": 0.5134349521446295, "grad_norm": 8.196756501629011, "learning_rate": 5.0258334625443075e-06, "loss": 0.4496, "step": 7242 }, { "epoch": 0.51350584898972, "grad_norm": 3.554328906602454, "learning_rate": 5.024685318199085e-06, "loss": 0.3728, "step": 7243 }, { "epoch": 0.5135767458348104, "grad_norm": 3.8169535460403083, "learning_rate": 5.023537172552187e-06, "loss": 0.3696, "step": 7244 }, { "epoch": 0.5136476426799007, "grad_norm": 8.369275891576047, "learning_rate": 5.0223890256641576e-06, "loss": 0.3898, "step": 7245 }, { "epoch": 0.5137185395249911, "grad_norm": 3.700005742171208, "learning_rate": 5.0212408775955356e-06, "loss": 0.3971, "step": 7246 }, { "epoch": 0.5137894363700816, "grad_norm": 2.844341421891407, "learning_rate": 5.020092728406867e-06, "loss": 0.4371, "step": 7247 }, { "epoch": 0.5138603332151719, "grad_norm": 4.347093475290958, "learning_rate": 5.018944578158693e-06, "loss": 0.3891, "step": 7248 }, { "epoch": 0.5139312300602623, "grad_norm": 2.547661928337361, "learning_rate": 5.0177964269115565e-06, "loss": 0.3832, "step": 7249 }, { "epoch": 0.5140021269053527, "grad_norm": 2.1149430345245124, "learning_rate": 5.016648274726001e-06, "loss": 0.335, "step": 7250 }, { "epoch": 0.5140730237504431, "grad_norm": 6.214479401332781, "learning_rate": 5.015500121662571e-06, "loss": 0.4517, "step": 7251 }, { "epoch": 0.5141439205955335, "grad_norm": 9.581739666306897, "learning_rate": 5.014351967781805e-06, "loss": 0.4261, "step": 7252 }, { "epoch": 0.5142148174406239, "grad_norm": 4.732815718331975, "learning_rate": 5.013203813144252e-06, "loss": 0.387, "step": 7253 }, { "epoch": 0.5142857142857142, "grad_norm": 4.268680188059669, "learning_rate": 5.01205565781045e-06, "loss": 0.4117, "step": 7254 }, { "epoch": 0.5143566111308047, "grad_norm": 6.5831569160889964, "learning_rate": 5.010907501840943e-06, "loss": 0.4257, "step": 7255 }, { "epoch": 0.5144275079758951, "grad_norm": 4.2576685662282925, "learning_rate": 5.009759345296277e-06, "loss": 0.4035, "step": 7256 }, { "epoch": 0.5144984048209854, "grad_norm": 2.9949246174945428, "learning_rate": 5.0086111882369915e-06, "loss": 0.446, "step": 7257 }, { "epoch": 0.5145693016660758, "grad_norm": 3.4211241982486436, "learning_rate": 5.007463030723634e-06, "loss": 0.4549, "step": 7258 }, { "epoch": 0.5146401985111663, "grad_norm": 6.747102175626513, "learning_rate": 5.0063148728167435e-06, "loss": 0.4259, "step": 7259 }, { "epoch": 0.5147110953562567, "grad_norm": 4.240854950285769, "learning_rate": 5.005166714576864e-06, "loss": 0.4504, "step": 7260 }, { "epoch": 0.514781992201347, "grad_norm": 3.331782590299743, "learning_rate": 5.004018556064541e-06, "loss": 0.4109, "step": 7261 }, { "epoch": 0.5148528890464374, "grad_norm": 5.42815231011848, "learning_rate": 5.002870397340316e-06, "loss": 0.4357, "step": 7262 }, { "epoch": 0.5149237858915279, "grad_norm": 4.262615473120086, "learning_rate": 5.001722238464733e-06, "loss": 0.3899, "step": 7263 }, { "epoch": 0.5149946827366182, "grad_norm": 2.5746147065267913, "learning_rate": 5.000574079498336e-06, "loss": 0.37, "step": 7264 }, { "epoch": 0.5150655795817086, "grad_norm": 3.1047562417825394, "learning_rate": 4.999425920501665e-06, "loss": 0.4152, "step": 7265 }, { "epoch": 0.515136476426799, "grad_norm": 6.1419583538592075, "learning_rate": 4.998277761535267e-06, "loss": 0.4067, "step": 7266 }, { "epoch": 0.5152073732718894, "grad_norm": 2.1134270529867876, "learning_rate": 4.997129602659685e-06, "loss": 0.4085, "step": 7267 }, { "epoch": 0.5152782701169798, "grad_norm": 2.6915760592186895, "learning_rate": 4.9959814439354605e-06, "loss": 0.451, "step": 7268 }, { "epoch": 0.5153491669620702, "grad_norm": 2.4077401064019064, "learning_rate": 4.9948332854231365e-06, "loss": 0.4072, "step": 7269 }, { "epoch": 0.5154200638071605, "grad_norm": 3.0435630415181447, "learning_rate": 4.993685127183259e-06, "loss": 0.4312, "step": 7270 }, { "epoch": 0.515490960652251, "grad_norm": 3.6025591477295285, "learning_rate": 4.992536969276367e-06, "loss": 0.382, "step": 7271 }, { "epoch": 0.5155618574973414, "grad_norm": 6.045416849858858, "learning_rate": 4.9913888117630085e-06, "loss": 0.4609, "step": 7272 }, { "epoch": 0.5156327543424317, "grad_norm": 2.006490173526585, "learning_rate": 4.990240654703725e-06, "loss": 0.4231, "step": 7273 }, { "epoch": 0.5157036511875221, "grad_norm": 5.4342854900862765, "learning_rate": 4.9890924981590574e-06, "loss": 0.3928, "step": 7274 }, { "epoch": 0.5157745480326126, "grad_norm": 9.020877205242911, "learning_rate": 4.987944342189552e-06, "loss": 0.4613, "step": 7275 }, { "epoch": 0.5158454448777029, "grad_norm": 2.200817737312844, "learning_rate": 4.98679618685575e-06, "loss": 0.4203, "step": 7276 }, { "epoch": 0.5159163417227933, "grad_norm": 2.8394642625061, "learning_rate": 4.9856480322181965e-06, "loss": 0.4274, "step": 7277 }, { "epoch": 0.5159872385678838, "grad_norm": 3.096464162473246, "learning_rate": 4.98449987833743e-06, "loss": 0.368, "step": 7278 }, { "epoch": 0.5160581354129741, "grad_norm": 5.486192971271958, "learning_rate": 4.983351725274001e-06, "loss": 0.3784, "step": 7279 }, { "epoch": 0.5161290322580645, "grad_norm": 1.7696536078377645, "learning_rate": 4.982203573088445e-06, "loss": 0.3972, "step": 7280 }, { "epoch": 0.5161999291031549, "grad_norm": 2.56794471728527, "learning_rate": 4.981055421841309e-06, "loss": 0.4235, "step": 7281 }, { "epoch": 0.5162708259482452, "grad_norm": 3.418727594314245, "learning_rate": 4.979907271593135e-06, "loss": 0.3635, "step": 7282 }, { "epoch": 0.5163417227933357, "grad_norm": 1.215300732183215, "learning_rate": 4.978759122404467e-06, "loss": 0.6943, "step": 7283 }, { "epoch": 0.5164126196384261, "grad_norm": 2.458962223634352, "learning_rate": 4.977610974335844e-06, "loss": 0.3447, "step": 7284 }, { "epoch": 0.5164835164835165, "grad_norm": 4.9933021190629265, "learning_rate": 4.976462827447813e-06, "loss": 0.4373, "step": 7285 }, { "epoch": 0.5165544133286069, "grad_norm": 2.742842759015117, "learning_rate": 4.975314681800917e-06, "loss": 0.4579, "step": 7286 }, { "epoch": 0.5166253101736973, "grad_norm": 2.5861576843144483, "learning_rate": 4.974166537455694e-06, "loss": 0.4021, "step": 7287 }, { "epoch": 0.5166962070187877, "grad_norm": 4.7155105336817424, "learning_rate": 4.97301839447269e-06, "loss": 0.4236, "step": 7288 }, { "epoch": 0.516767103863878, "grad_norm": 5.702231684045815, "learning_rate": 4.971870252912448e-06, "loss": 0.4334, "step": 7289 }, { "epoch": 0.5168380007089685, "grad_norm": 0.8880259422450781, "learning_rate": 4.970722112835508e-06, "loss": 0.5913, "step": 7290 }, { "epoch": 0.5169088975540589, "grad_norm": 3.830705360422026, "learning_rate": 4.969573974302412e-06, "loss": 0.3886, "step": 7291 }, { "epoch": 0.5169797943991492, "grad_norm": 3.8129259000915385, "learning_rate": 4.968425837373705e-06, "loss": 0.4176, "step": 7292 }, { "epoch": 0.5170506912442396, "grad_norm": 1.119987627001459, "learning_rate": 4.967277702109928e-06, "loss": 0.735, "step": 7293 }, { "epoch": 0.5171215880893301, "grad_norm": 3.572244416602984, "learning_rate": 4.966129568571621e-06, "loss": 0.4138, "step": 7294 }, { "epoch": 0.5171924849344204, "grad_norm": 2.8156474764182047, "learning_rate": 4.964981436819328e-06, "loss": 0.3965, "step": 7295 }, { "epoch": 0.5172633817795108, "grad_norm": 6.819393362221504, "learning_rate": 4.963833306913594e-06, "loss": 0.4123, "step": 7296 }, { "epoch": 0.5173342786246012, "grad_norm": 3.2545490765453398, "learning_rate": 4.962685178914953e-06, "loss": 0.4466, "step": 7297 }, { "epoch": 0.5174051754696916, "grad_norm": 4.338097171984916, "learning_rate": 4.9615370528839526e-06, "loss": 0.3887, "step": 7298 }, { "epoch": 0.517476072314782, "grad_norm": 4.492598431688351, "learning_rate": 4.960388928881134e-06, "loss": 0.4115, "step": 7299 }, { "epoch": 0.5175469691598724, "grad_norm": 2.4405590329349773, "learning_rate": 4.959240806967036e-06, "loss": 0.3675, "step": 7300 }, { "epoch": 0.5176178660049627, "grad_norm": 2.56936134402807, "learning_rate": 4.9580926872022014e-06, "loss": 0.4372, "step": 7301 }, { "epoch": 0.5176887628500532, "grad_norm": 2.6141819246463887, "learning_rate": 4.956944569647173e-06, "loss": 0.4169, "step": 7302 }, { "epoch": 0.5177596596951436, "grad_norm": 6.841606492275031, "learning_rate": 4.955796454362488e-06, "loss": 0.4027, "step": 7303 }, { "epoch": 0.5178305565402339, "grad_norm": 10.332191004963377, "learning_rate": 4.954648341408691e-06, "loss": 0.406, "step": 7304 }, { "epoch": 0.5179014533853243, "grad_norm": 7.378694048698476, "learning_rate": 4.953500230846322e-06, "loss": 0.3629, "step": 7305 }, { "epoch": 0.5179723502304148, "grad_norm": 2.9021922026987443, "learning_rate": 4.95235212273592e-06, "loss": 0.4576, "step": 7306 }, { "epoch": 0.5180432470755051, "grad_norm": 5.7301469105976635, "learning_rate": 4.951204017138027e-06, "loss": 0.4697, "step": 7307 }, { "epoch": 0.5181141439205955, "grad_norm": 2.7745209221378078, "learning_rate": 4.950055914113186e-06, "loss": 0.3474, "step": 7308 }, { "epoch": 0.518185040765686, "grad_norm": 1.6486793731264178, "learning_rate": 4.948907813721932e-06, "loss": 0.4316, "step": 7309 }, { "epoch": 0.5182559376107764, "grad_norm": 2.3616698040197917, "learning_rate": 4.947759716024807e-06, "loss": 0.3933, "step": 7310 }, { "epoch": 0.5183268344558667, "grad_norm": 5.067324346569982, "learning_rate": 4.9466116210823535e-06, "loss": 0.3935, "step": 7311 }, { "epoch": 0.5183977313009571, "grad_norm": 3.8632890822884454, "learning_rate": 4.945463528955112e-06, "loss": 0.3593, "step": 7312 }, { "epoch": 0.5184686281460475, "grad_norm": 3.174393177084013, "learning_rate": 4.9443154397036184e-06, "loss": 0.4699, "step": 7313 }, { "epoch": 0.5185395249911379, "grad_norm": 9.980415542820698, "learning_rate": 4.943167353388414e-06, "loss": 0.3631, "step": 7314 }, { "epoch": 0.5186104218362283, "grad_norm": 2.182061664629003, "learning_rate": 4.9420192700700396e-06, "loss": 0.4311, "step": 7315 }, { "epoch": 0.5186813186813187, "grad_norm": 8.216871473929455, "learning_rate": 4.940871189809031e-06, "loss": 0.4348, "step": 7316 }, { "epoch": 0.518752215526409, "grad_norm": 3.733775548959504, "learning_rate": 4.939723112665931e-06, "loss": 0.3948, "step": 7317 }, { "epoch": 0.5188231123714995, "grad_norm": 2.4993112776357775, "learning_rate": 4.938575038701278e-06, "loss": 0.4158, "step": 7318 }, { "epoch": 0.5188940092165899, "grad_norm": 4.532085659620926, "learning_rate": 4.93742696797561e-06, "loss": 0.4307, "step": 7319 }, { "epoch": 0.5189649060616802, "grad_norm": 3.3165636273006363, "learning_rate": 4.936278900549464e-06, "loss": 0.3771, "step": 7320 }, { "epoch": 0.5190358029067706, "grad_norm": 7.102625996540156, "learning_rate": 4.935130836483383e-06, "loss": 0.3965, "step": 7321 }, { "epoch": 0.5191066997518611, "grad_norm": 5.581783149863423, "learning_rate": 4.933982775837901e-06, "loss": 0.4787, "step": 7322 }, { "epoch": 0.5191775965969514, "grad_norm": 0.912867007874324, "learning_rate": 4.932834718673555e-06, "loss": 0.6695, "step": 7323 }, { "epoch": 0.5192484934420418, "grad_norm": 2.7113569001126843, "learning_rate": 4.931686665050888e-06, "loss": 0.4412, "step": 7324 }, { "epoch": 0.5193193902871323, "grad_norm": 5.465874374993374, "learning_rate": 4.930538615030437e-06, "loss": 0.4206, "step": 7325 }, { "epoch": 0.5193902871322226, "grad_norm": 2.376055662645629, "learning_rate": 4.929390568672736e-06, "loss": 0.4358, "step": 7326 }, { "epoch": 0.519461183977313, "grad_norm": 3.174441943422549, "learning_rate": 4.928242526038323e-06, "loss": 0.3899, "step": 7327 }, { "epoch": 0.5195320808224034, "grad_norm": 3.572990212517394, "learning_rate": 4.92709448718774e-06, "loss": 0.3975, "step": 7328 }, { "epoch": 0.5196029776674937, "grad_norm": 2.931590007024244, "learning_rate": 4.925946452181517e-06, "loss": 0.4302, "step": 7329 }, { "epoch": 0.5196738745125842, "grad_norm": 4.4669670127277055, "learning_rate": 4.9247984210801945e-06, "loss": 0.4333, "step": 7330 }, { "epoch": 0.5197447713576746, "grad_norm": 2.196485568507288, "learning_rate": 4.923650393944311e-06, "loss": 0.3863, "step": 7331 }, { "epoch": 0.5198156682027649, "grad_norm": 3.79590171216918, "learning_rate": 4.922502370834398e-06, "loss": 0.4066, "step": 7332 }, { "epoch": 0.5198865650478554, "grad_norm": 1.6119945467094965, "learning_rate": 4.921354351810995e-06, "loss": 0.3337, "step": 7333 }, { "epoch": 0.5199574618929458, "grad_norm": 6.09650536076815, "learning_rate": 4.920206336934638e-06, "loss": 0.3981, "step": 7334 }, { "epoch": 0.5200283587380362, "grad_norm": 3.0396323906792233, "learning_rate": 4.919058326265861e-06, "loss": 0.4418, "step": 7335 }, { "epoch": 0.5200992555831265, "grad_norm": 2.5347098219447566, "learning_rate": 4.917910319865199e-06, "loss": 0.405, "step": 7336 }, { "epoch": 0.520170152428217, "grad_norm": 36.311930369760034, "learning_rate": 4.916762317793191e-06, "loss": 0.439, "step": 7337 }, { "epoch": 0.5202410492733074, "grad_norm": 3.032458224194749, "learning_rate": 4.915614320110367e-06, "loss": 0.3967, "step": 7338 }, { "epoch": 0.5203119461183977, "grad_norm": 1.8675261691212917, "learning_rate": 4.914466326877265e-06, "loss": 0.3233, "step": 7339 }, { "epoch": 0.5203828429634881, "grad_norm": 2.2454365690304545, "learning_rate": 4.913318338154419e-06, "loss": 0.3555, "step": 7340 }, { "epoch": 0.5204537398085786, "grad_norm": 3.473048603486785, "learning_rate": 4.912170354002365e-06, "loss": 0.3271, "step": 7341 }, { "epoch": 0.5205246366536689, "grad_norm": 2.9864377781706977, "learning_rate": 4.911022374481632e-06, "loss": 0.387, "step": 7342 }, { "epoch": 0.5205955334987593, "grad_norm": 2.992491276429168, "learning_rate": 4.909874399652759e-06, "loss": 0.3811, "step": 7343 }, { "epoch": 0.5206664303438497, "grad_norm": 2.758824278489231, "learning_rate": 4.908726429576279e-06, "loss": 0.445, "step": 7344 }, { "epoch": 0.5207373271889401, "grad_norm": 2.6175368534746255, "learning_rate": 4.907578464312722e-06, "loss": 0.4343, "step": 7345 }, { "epoch": 0.5208082240340305, "grad_norm": 3.3408471950396654, "learning_rate": 4.906430503922623e-06, "loss": 0.4675, "step": 7346 }, { "epoch": 0.5208791208791209, "grad_norm": 3.8605275686322202, "learning_rate": 4.905282548466517e-06, "loss": 0.4108, "step": 7347 }, { "epoch": 0.5209500177242112, "grad_norm": 1.9950516606846946, "learning_rate": 4.904134598004932e-06, "loss": 0.4026, "step": 7348 }, { "epoch": 0.5210209145693017, "grad_norm": 1.960682345012993, "learning_rate": 4.902986652598404e-06, "loss": 0.383, "step": 7349 }, { "epoch": 0.5210918114143921, "grad_norm": 13.27162101050347, "learning_rate": 4.9018387123074645e-06, "loss": 0.4194, "step": 7350 }, { "epoch": 0.5211627082594824, "grad_norm": 4.053023666136212, "learning_rate": 4.9006907771926435e-06, "loss": 0.4734, "step": 7351 }, { "epoch": 0.5212336051045728, "grad_norm": 2.2805728807603893, "learning_rate": 4.899542847314474e-06, "loss": 0.3695, "step": 7352 }, { "epoch": 0.5213045019496633, "grad_norm": 2.4808980452971463, "learning_rate": 4.898394922733486e-06, "loss": 0.3548, "step": 7353 }, { "epoch": 0.5213753987947536, "grad_norm": 2.5272734683592653, "learning_rate": 4.897247003510214e-06, "loss": 0.4195, "step": 7354 }, { "epoch": 0.521446295639844, "grad_norm": 2.3208195275451495, "learning_rate": 4.896099089705183e-06, "loss": 0.4168, "step": 7355 }, { "epoch": 0.5215171924849344, "grad_norm": 4.000833138000572, "learning_rate": 4.8949511813789265e-06, "loss": 0.3872, "step": 7356 }, { "epoch": 0.5215880893300248, "grad_norm": 4.08049940128835, "learning_rate": 4.893803278591977e-06, "loss": 0.4116, "step": 7357 }, { "epoch": 0.5216589861751152, "grad_norm": 1.1694307443214866, "learning_rate": 4.892655381404859e-06, "loss": 0.6547, "step": 7358 }, { "epoch": 0.5217298830202056, "grad_norm": 3.171314984952052, "learning_rate": 4.891507489878105e-06, "loss": 0.4517, "step": 7359 }, { "epoch": 0.521800779865296, "grad_norm": 2.349423961935264, "learning_rate": 4.890359604072247e-06, "loss": 0.3767, "step": 7360 }, { "epoch": 0.5218716767103864, "grad_norm": 6.080709994153754, "learning_rate": 4.889211724047807e-06, "loss": 0.3681, "step": 7361 }, { "epoch": 0.5219425735554768, "grad_norm": 2.6597887111109793, "learning_rate": 4.888063849865318e-06, "loss": 0.3806, "step": 7362 }, { "epoch": 0.5220134704005672, "grad_norm": 3.346755715007875, "learning_rate": 4.88691598158531e-06, "loss": 0.3548, "step": 7363 }, { "epoch": 0.5220843672456575, "grad_norm": 3.6329876996595756, "learning_rate": 4.885768119268306e-06, "loss": 0.4048, "step": 7364 }, { "epoch": 0.522155264090748, "grad_norm": 4.1735903913904755, "learning_rate": 4.8846202629748365e-06, "loss": 0.4144, "step": 7365 }, { "epoch": 0.5222261609358384, "grad_norm": 2.3551732159558103, "learning_rate": 4.88347241276543e-06, "loss": 0.3945, "step": 7366 }, { "epoch": 0.5222970577809287, "grad_norm": 3.2262011860059348, "learning_rate": 4.8823245687006115e-06, "loss": 0.458, "step": 7367 }, { "epoch": 0.5223679546260191, "grad_norm": 2.1147285230516006, "learning_rate": 4.881176730840907e-06, "loss": 0.4512, "step": 7368 }, { "epoch": 0.5224388514711096, "grad_norm": 4.1529246351699, "learning_rate": 4.8800288992468455e-06, "loss": 0.4145, "step": 7369 }, { "epoch": 0.5225097483161999, "grad_norm": 4.022682207643078, "learning_rate": 4.878881073978952e-06, "loss": 0.3545, "step": 7370 }, { "epoch": 0.5225806451612903, "grad_norm": 2.8086515918264285, "learning_rate": 4.877733255097751e-06, "loss": 0.4299, "step": 7371 }, { "epoch": 0.5226515420063808, "grad_norm": 2.3869801829365023, "learning_rate": 4.8765854426637684e-06, "loss": 0.4029, "step": 7372 }, { "epoch": 0.5227224388514711, "grad_norm": 2.988274949049086, "learning_rate": 4.8754376367375324e-06, "loss": 0.4191, "step": 7373 }, { "epoch": 0.5227933356965615, "grad_norm": 2.9484702646137526, "learning_rate": 4.874289837379562e-06, "loss": 0.4123, "step": 7374 }, { "epoch": 0.5228642325416519, "grad_norm": 2.1584656446395285, "learning_rate": 4.873142044650385e-06, "loss": 0.3464, "step": 7375 }, { "epoch": 0.5229351293867422, "grad_norm": 2.9418151468338682, "learning_rate": 4.871994258610526e-06, "loss": 0.3744, "step": 7376 }, { "epoch": 0.5230060262318327, "grad_norm": 3.6181216351900587, "learning_rate": 4.870846479320507e-06, "loss": 0.4097, "step": 7377 }, { "epoch": 0.5230769230769231, "grad_norm": 3.148550739900927, "learning_rate": 4.869698706840851e-06, "loss": 0.4178, "step": 7378 }, { "epoch": 0.5231478199220134, "grad_norm": 3.1998727665503157, "learning_rate": 4.868550941232083e-06, "loss": 0.4139, "step": 7379 }, { "epoch": 0.5232187167671039, "grad_norm": 5.173146730991606, "learning_rate": 4.867403182554722e-06, "loss": 0.4491, "step": 7380 }, { "epoch": 0.5232896136121943, "grad_norm": 4.32285355699558, "learning_rate": 4.866255430869293e-06, "loss": 0.4385, "step": 7381 }, { "epoch": 0.5233605104572846, "grad_norm": 7.38270222540682, "learning_rate": 4.865107686236319e-06, "loss": 0.3792, "step": 7382 }, { "epoch": 0.523431407302375, "grad_norm": 3.7159868991636813, "learning_rate": 4.863959948716318e-06, "loss": 0.3943, "step": 7383 }, { "epoch": 0.5235023041474655, "grad_norm": 3.426106078308635, "learning_rate": 4.862812218369813e-06, "loss": 0.3608, "step": 7384 }, { "epoch": 0.5235732009925558, "grad_norm": 2.8975644734300166, "learning_rate": 4.8616644952573236e-06, "loss": 0.4065, "step": 7385 }, { "epoch": 0.5236440978376462, "grad_norm": 3.444314462171744, "learning_rate": 4.860516779439374e-06, "loss": 0.4222, "step": 7386 }, { "epoch": 0.5237149946827366, "grad_norm": 3.0578417457787554, "learning_rate": 4.859369070976478e-06, "loss": 0.4107, "step": 7387 }, { "epoch": 0.5237858915278271, "grad_norm": 2.479630973195126, "learning_rate": 4.858221369929159e-06, "loss": 0.413, "step": 7388 }, { "epoch": 0.5238567883729174, "grad_norm": 2.09713719668816, "learning_rate": 4.857073676357936e-06, "loss": 0.3744, "step": 7389 }, { "epoch": 0.5239276852180078, "grad_norm": 4.932791178572307, "learning_rate": 4.855925990323327e-06, "loss": 0.4728, "step": 7390 }, { "epoch": 0.5239985820630982, "grad_norm": 2.9760087972994884, "learning_rate": 4.8547783118858485e-06, "loss": 0.3601, "step": 7391 }, { "epoch": 0.5240694789081886, "grad_norm": 3.136730482712141, "learning_rate": 4.853630641106025e-06, "loss": 0.4175, "step": 7392 }, { "epoch": 0.524140375753279, "grad_norm": 7.260452479011679, "learning_rate": 4.852482978044365e-06, "loss": 0.4008, "step": 7393 }, { "epoch": 0.5242112725983694, "grad_norm": 3.3643941305421787, "learning_rate": 4.851335322761392e-06, "loss": 0.4144, "step": 7394 }, { "epoch": 0.5242821694434597, "grad_norm": 5.356372853958541, "learning_rate": 4.850187675317622e-06, "loss": 0.4155, "step": 7395 }, { "epoch": 0.5243530662885502, "grad_norm": 3.0715230894075423, "learning_rate": 4.849040035773568e-06, "loss": 0.4788, "step": 7396 }, { "epoch": 0.5244239631336406, "grad_norm": 4.730709078718969, "learning_rate": 4.8478924041897495e-06, "loss": 0.4239, "step": 7397 }, { "epoch": 0.5244948599787309, "grad_norm": 2.7650311165233186, "learning_rate": 4.846744780626679e-06, "loss": 0.4022, "step": 7398 }, { "epoch": 0.5245657568238213, "grad_norm": 2.0035806369057347, "learning_rate": 4.845597165144877e-06, "loss": 0.4428, "step": 7399 }, { "epoch": 0.5246366536689118, "grad_norm": 3.608494610297226, "learning_rate": 4.84444955780485e-06, "loss": 0.3513, "step": 7400 }, { "epoch": 0.5247075505140021, "grad_norm": 2.481392265947678, "learning_rate": 4.8433019586671185e-06, "loss": 0.3807, "step": 7401 }, { "epoch": 0.5247784473590925, "grad_norm": 4.2731016652683795, "learning_rate": 4.842154367792195e-06, "loss": 0.3657, "step": 7402 }, { "epoch": 0.5248493442041829, "grad_norm": 3.685631006657145, "learning_rate": 4.8410067852405915e-06, "loss": 0.4127, "step": 7403 }, { "epoch": 0.5249202410492733, "grad_norm": 7.591364599369113, "learning_rate": 4.83985921107282e-06, "loss": 0.3802, "step": 7404 }, { "epoch": 0.5249911378943637, "grad_norm": 3.059272005532194, "learning_rate": 4.838711645349398e-06, "loss": 0.4588, "step": 7405 }, { "epoch": 0.5250620347394541, "grad_norm": 2.028793726306411, "learning_rate": 4.83756408813083e-06, "loss": 0.4114, "step": 7406 }, { "epoch": 0.5251329315845444, "grad_norm": 2.1613656838379076, "learning_rate": 4.836416539477634e-06, "loss": 0.3373, "step": 7407 }, { "epoch": 0.5252038284296349, "grad_norm": 3.0231304709133138, "learning_rate": 4.835268999450319e-06, "loss": 0.4317, "step": 7408 }, { "epoch": 0.5252747252747253, "grad_norm": 3.478636085491393, "learning_rate": 4.834121468109395e-06, "loss": 0.4216, "step": 7409 }, { "epoch": 0.5253456221198156, "grad_norm": 3.7896904515347383, "learning_rate": 4.832973945515372e-06, "loss": 0.4207, "step": 7410 }, { "epoch": 0.525416518964906, "grad_norm": 4.712040753915515, "learning_rate": 4.8318264317287625e-06, "loss": 0.4307, "step": 7411 }, { "epoch": 0.5254874158099965, "grad_norm": 5.526525243021539, "learning_rate": 4.83067892681007e-06, "loss": 0.4055, "step": 7412 }, { "epoch": 0.5255583126550869, "grad_norm": 6.264447870615668, "learning_rate": 4.8295314308198084e-06, "loss": 0.4243, "step": 7413 }, { "epoch": 0.5256292095001772, "grad_norm": 4.124752967614414, "learning_rate": 4.828383943818485e-06, "loss": 0.403, "step": 7414 }, { "epoch": 0.5257001063452676, "grad_norm": 2.2346401611876336, "learning_rate": 4.8272364658666085e-06, "loss": 0.412, "step": 7415 }, { "epoch": 0.5257710031903581, "grad_norm": 3.167634726951928, "learning_rate": 4.826088997024683e-06, "loss": 0.327, "step": 7416 }, { "epoch": 0.5258419000354484, "grad_norm": 2.4362637044350173, "learning_rate": 4.8249415373532165e-06, "loss": 0.3836, "step": 7417 }, { "epoch": 0.5259127968805388, "grad_norm": 3.8870387309149, "learning_rate": 4.823794086912721e-06, "loss": 0.4306, "step": 7418 }, { "epoch": 0.5259836937256293, "grad_norm": 3.683522630984317, "learning_rate": 4.8226466457636924e-06, "loss": 0.4123, "step": 7419 }, { "epoch": 0.5260545905707196, "grad_norm": 1.829737476711933, "learning_rate": 4.821499213966644e-06, "loss": 0.394, "step": 7420 }, { "epoch": 0.52612548741581, "grad_norm": 1.1427924523657293, "learning_rate": 4.820351791582079e-06, "loss": 0.718, "step": 7421 }, { "epoch": 0.5261963842609004, "grad_norm": 2.781003681179096, "learning_rate": 4.8192043786705e-06, "loss": 0.4434, "step": 7422 }, { "epoch": 0.5262672811059907, "grad_norm": 3.6538845253661023, "learning_rate": 4.818056975292412e-06, "loss": 0.445, "step": 7423 }, { "epoch": 0.5263381779510812, "grad_norm": 2.7135266688020288, "learning_rate": 4.81690958150832e-06, "loss": 0.4398, "step": 7424 }, { "epoch": 0.5264090747961716, "grad_norm": 5.390554924455453, "learning_rate": 4.815762197378723e-06, "loss": 0.4054, "step": 7425 }, { "epoch": 0.5264799716412619, "grad_norm": 2.756508338552331, "learning_rate": 4.814614822964128e-06, "loss": 0.4123, "step": 7426 }, { "epoch": 0.5265508684863524, "grad_norm": 2.4751051876702994, "learning_rate": 4.813467458325034e-06, "loss": 0.3663, "step": 7427 }, { "epoch": 0.5266217653314428, "grad_norm": 2.218194958615236, "learning_rate": 4.8123201035219445e-06, "loss": 0.3922, "step": 7428 }, { "epoch": 0.5266926621765331, "grad_norm": 2.7378932672022525, "learning_rate": 4.811172758615358e-06, "loss": 0.4194, "step": 7429 }, { "epoch": 0.5267635590216235, "grad_norm": 2.620896386481113, "learning_rate": 4.810025423665775e-06, "loss": 0.3781, "step": 7430 }, { "epoch": 0.526834455866714, "grad_norm": 2.3914708707492522, "learning_rate": 4.8088780987337e-06, "loss": 0.4076, "step": 7431 }, { "epoch": 0.5269053527118043, "grad_norm": 3.4363005434604603, "learning_rate": 4.807730783879625e-06, "loss": 0.3665, "step": 7432 }, { "epoch": 0.5269762495568947, "grad_norm": 3.5036206411516746, "learning_rate": 4.806583479164054e-06, "loss": 0.3988, "step": 7433 }, { "epoch": 0.5270471464019851, "grad_norm": 3.8877036045703566, "learning_rate": 4.805436184647484e-06, "loss": 0.4115, "step": 7434 }, { "epoch": 0.5271180432470755, "grad_norm": 1.182183470934946, "learning_rate": 4.8042889003904134e-06, "loss": 0.6487, "step": 7435 }, { "epoch": 0.5271889400921659, "grad_norm": 6.65956689362575, "learning_rate": 4.803141626453335e-06, "loss": 0.4443, "step": 7436 }, { "epoch": 0.5272598369372563, "grad_norm": 2.51651422143522, "learning_rate": 4.801994362896753e-06, "loss": 0.4325, "step": 7437 }, { "epoch": 0.5273307337823467, "grad_norm": 3.256573748249805, "learning_rate": 4.800847109781157e-06, "loss": 0.3922, "step": 7438 }, { "epoch": 0.5274016306274371, "grad_norm": 7.4011599361575735, "learning_rate": 4.7996998671670455e-06, "loss": 0.3621, "step": 7439 }, { "epoch": 0.5274725274725275, "grad_norm": 2.8744550027665907, "learning_rate": 4.798552635114914e-06, "loss": 0.3746, "step": 7440 }, { "epoch": 0.5275434243176179, "grad_norm": 2.6753950283011787, "learning_rate": 4.797405413685254e-06, "loss": 0.3575, "step": 7441 }, { "epoch": 0.5276143211627082, "grad_norm": 4.595318322170815, "learning_rate": 4.7962582029385616e-06, "loss": 0.504, "step": 7442 }, { "epoch": 0.5276852180077987, "grad_norm": 4.367293424535773, "learning_rate": 4.795111002935329e-06, "loss": 0.4332, "step": 7443 }, { "epoch": 0.5277561148528891, "grad_norm": 2.1945894712438214, "learning_rate": 4.793963813736051e-06, "loss": 0.4303, "step": 7444 }, { "epoch": 0.5278270116979794, "grad_norm": 4.442532523753377, "learning_rate": 4.792816635401219e-06, "loss": 0.456, "step": 7445 }, { "epoch": 0.5278979085430698, "grad_norm": 3.582855529702034, "learning_rate": 4.791669467991322e-06, "loss": 0.39, "step": 7446 }, { "epoch": 0.5279688053881603, "grad_norm": 4.141382947755364, "learning_rate": 4.790522311566855e-06, "loss": 0.4033, "step": 7447 }, { "epoch": 0.5280397022332506, "grad_norm": 1.6368118238650675, "learning_rate": 4.789375166188305e-06, "loss": 0.4179, "step": 7448 }, { "epoch": 0.528110599078341, "grad_norm": 3.18100278809447, "learning_rate": 4.788228031916163e-06, "loss": 0.3812, "step": 7449 }, { "epoch": 0.5281814959234314, "grad_norm": 2.359704646809443, "learning_rate": 4.78708090881092e-06, "loss": 0.3494, "step": 7450 }, { "epoch": 0.5282523927685218, "grad_norm": 4.332548869377891, "learning_rate": 4.785933796933061e-06, "loss": 0.3599, "step": 7451 }, { "epoch": 0.5283232896136122, "grad_norm": 11.221874739444537, "learning_rate": 4.7847866963430775e-06, "loss": 0.4174, "step": 7452 }, { "epoch": 0.5283941864587026, "grad_norm": 3.318998443675841, "learning_rate": 4.783639607101456e-06, "loss": 0.3492, "step": 7453 }, { "epoch": 0.5284650833037929, "grad_norm": 3.5728678393949074, "learning_rate": 4.782492529268683e-06, "loss": 0.3573, "step": 7454 }, { "epoch": 0.5285359801488834, "grad_norm": 4.751736819896907, "learning_rate": 4.781345462905245e-06, "loss": 0.3961, "step": 7455 }, { "epoch": 0.5286068769939738, "grad_norm": 2.445330567335358, "learning_rate": 4.7801984080716265e-06, "loss": 0.4094, "step": 7456 }, { "epoch": 0.5286777738390641, "grad_norm": 1.9645892198209036, "learning_rate": 4.779051364828315e-06, "loss": 0.4557, "step": 7457 }, { "epoch": 0.5287486706841545, "grad_norm": 4.464294725614864, "learning_rate": 4.777904333235794e-06, "loss": 0.4664, "step": 7458 }, { "epoch": 0.528819567529245, "grad_norm": 7.289814135175239, "learning_rate": 4.776757313354545e-06, "loss": 0.3597, "step": 7459 }, { "epoch": 0.5288904643743353, "grad_norm": 5.260710724291462, "learning_rate": 4.775610305245056e-06, "loss": 0.4518, "step": 7460 }, { "epoch": 0.5289613612194257, "grad_norm": 3.1105992866545633, "learning_rate": 4.7744633089678055e-06, "loss": 0.3921, "step": 7461 }, { "epoch": 0.5290322580645161, "grad_norm": 11.777852820127185, "learning_rate": 4.773316324583276e-06, "loss": 0.3614, "step": 7462 }, { "epoch": 0.5291031549096066, "grad_norm": 1.6376073302089384, "learning_rate": 4.772169352151952e-06, "loss": 0.39, "step": 7463 }, { "epoch": 0.5291740517546969, "grad_norm": 3.6906404718334165, "learning_rate": 4.77102239173431e-06, "loss": 0.3847, "step": 7464 }, { "epoch": 0.5292449485997873, "grad_norm": 2.614751132494193, "learning_rate": 4.769875443390832e-06, "loss": 0.4322, "step": 7465 }, { "epoch": 0.5293158454448778, "grad_norm": 3.0280077307270954, "learning_rate": 4.7687285071819985e-06, "loss": 0.4275, "step": 7466 }, { "epoch": 0.5293867422899681, "grad_norm": 3.659268138231666, "learning_rate": 4.767581583168287e-06, "loss": 0.4073, "step": 7467 }, { "epoch": 0.5294576391350585, "grad_norm": 3.700798104372713, "learning_rate": 4.766434671410175e-06, "loss": 0.4227, "step": 7468 }, { "epoch": 0.5295285359801489, "grad_norm": 3.4365910605160166, "learning_rate": 4.765287771968144e-06, "loss": 0.456, "step": 7469 }, { "epoch": 0.5295994328252392, "grad_norm": 3.2992069019967634, "learning_rate": 4.764140884902665e-06, "loss": 0.3814, "step": 7470 }, { "epoch": 0.5296703296703297, "grad_norm": 2.9297175345983333, "learning_rate": 4.762994010274218e-06, "loss": 0.3929, "step": 7471 }, { "epoch": 0.5297412265154201, "grad_norm": 3.015476330636843, "learning_rate": 4.7618471481432775e-06, "loss": 0.4169, "step": 7472 }, { "epoch": 0.5298121233605104, "grad_norm": 2.7558081079960797, "learning_rate": 4.76070029857032e-06, "loss": 0.4026, "step": 7473 }, { "epoch": 0.5298830202056009, "grad_norm": 3.552600996583656, "learning_rate": 4.759553461615817e-06, "loss": 0.4081, "step": 7474 }, { "epoch": 0.5299539170506913, "grad_norm": 2.348732594540904, "learning_rate": 4.758406637340242e-06, "loss": 0.447, "step": 7475 }, { "epoch": 0.5300248138957816, "grad_norm": 3.6960101164634405, "learning_rate": 4.7572598258040744e-06, "loss": 0.4201, "step": 7476 }, { "epoch": 0.530095710740872, "grad_norm": 3.3283944123219125, "learning_rate": 4.756113027067777e-06, "loss": 0.3784, "step": 7477 }, { "epoch": 0.5301666075859625, "grad_norm": 2.781700994756198, "learning_rate": 4.754966241191827e-06, "loss": 0.4325, "step": 7478 }, { "epoch": 0.5302375044310528, "grad_norm": 4.326784867620635, "learning_rate": 4.753819468236695e-06, "loss": 0.4647, "step": 7479 }, { "epoch": 0.5303084012761432, "grad_norm": 4.611854529444895, "learning_rate": 4.7526727082628495e-06, "loss": 0.4236, "step": 7480 }, { "epoch": 0.5303792981212336, "grad_norm": 3.92395304899727, "learning_rate": 4.75152596133076e-06, "loss": 0.4049, "step": 7481 }, { "epoch": 0.530450194966324, "grad_norm": 3.6296591863386736, "learning_rate": 4.750379227500899e-06, "loss": 0.3843, "step": 7482 }, { "epoch": 0.5305210918114144, "grad_norm": 2.9668153336855876, "learning_rate": 4.749232506833729e-06, "loss": 0.4203, "step": 7483 }, { "epoch": 0.5305919886565048, "grad_norm": 3.651562467605739, "learning_rate": 4.748085799389721e-06, "loss": 0.4025, "step": 7484 }, { "epoch": 0.5306628855015951, "grad_norm": 3.5582179735181394, "learning_rate": 4.746939105229342e-06, "loss": 0.4306, "step": 7485 }, { "epoch": 0.5307337823466856, "grad_norm": 2.6315452510176716, "learning_rate": 4.745792424413057e-06, "loss": 0.3889, "step": 7486 }, { "epoch": 0.530804679191776, "grad_norm": 3.2210211361627294, "learning_rate": 4.744645757001332e-06, "loss": 0.3995, "step": 7487 }, { "epoch": 0.5308755760368664, "grad_norm": 2.330781422881005, "learning_rate": 4.743499103054629e-06, "loss": 0.4113, "step": 7488 }, { "epoch": 0.5309464728819567, "grad_norm": 2.6962336401447597, "learning_rate": 4.742352462633416e-06, "loss": 0.3976, "step": 7489 }, { "epoch": 0.5310173697270472, "grad_norm": 2.128470200420766, "learning_rate": 4.741205835798153e-06, "loss": 0.4371, "step": 7490 }, { "epoch": 0.5310882665721376, "grad_norm": 3.6279287878644015, "learning_rate": 4.740059222609305e-06, "loss": 0.4765, "step": 7491 }, { "epoch": 0.5311591634172279, "grad_norm": 3.5764608959836672, "learning_rate": 4.738912623127332e-06, "loss": 0.4203, "step": 7492 }, { "epoch": 0.5312300602623183, "grad_norm": 4.274116608152871, "learning_rate": 4.737766037412696e-06, "loss": 0.431, "step": 7493 }, { "epoch": 0.5313009571074088, "grad_norm": 4.626941635671307, "learning_rate": 4.736619465525855e-06, "loss": 0.3776, "step": 7494 }, { "epoch": 0.5313718539524991, "grad_norm": 2.4932190358462942, "learning_rate": 4.7354729075272745e-06, "loss": 0.3677, "step": 7495 }, { "epoch": 0.5314427507975895, "grad_norm": 3.3177346293716803, "learning_rate": 4.734326363477406e-06, "loss": 0.4473, "step": 7496 }, { "epoch": 0.5315136476426799, "grad_norm": 3.3366970532024163, "learning_rate": 4.733179833436712e-06, "loss": 0.3995, "step": 7497 }, { "epoch": 0.5315845444877703, "grad_norm": 2.705019593379156, "learning_rate": 4.732033317465651e-06, "loss": 0.4441, "step": 7498 }, { "epoch": 0.5316554413328607, "grad_norm": 5.614446609868234, "learning_rate": 4.730886815624676e-06, "loss": 0.4166, "step": 7499 }, { "epoch": 0.5317263381779511, "grad_norm": 2.7718915465157328, "learning_rate": 4.729740327974244e-06, "loss": 0.4025, "step": 7500 }, { "epoch": 0.5317972350230414, "grad_norm": 2.189285853433323, "learning_rate": 4.728593854574811e-06, "loss": 0.3881, "step": 7501 }, { "epoch": 0.5318681318681319, "grad_norm": 3.243312191553793, "learning_rate": 4.727447395486833e-06, "loss": 0.3934, "step": 7502 }, { "epoch": 0.5319390287132223, "grad_norm": 2.336644224886772, "learning_rate": 4.72630095077076e-06, "loss": 0.3885, "step": 7503 }, { "epoch": 0.5320099255583126, "grad_norm": 9.414713348290936, "learning_rate": 4.725154520487048e-06, "loss": 0.3811, "step": 7504 }, { "epoch": 0.532080822403403, "grad_norm": 8.830084442355444, "learning_rate": 4.724008104696148e-06, "loss": 0.3883, "step": 7505 }, { "epoch": 0.5321517192484935, "grad_norm": 8.26684150752061, "learning_rate": 4.72286170345851e-06, "loss": 0.4459, "step": 7506 }, { "epoch": 0.5322226160935838, "grad_norm": 3.2708634207629395, "learning_rate": 4.721715316834585e-06, "loss": 0.3626, "step": 7507 }, { "epoch": 0.5322935129386742, "grad_norm": 3.0884558089434173, "learning_rate": 4.720568944884828e-06, "loss": 0.3844, "step": 7508 }, { "epoch": 0.5323644097837646, "grad_norm": 3.122154157660464, "learning_rate": 4.7194225876696795e-06, "loss": 0.3625, "step": 7509 }, { "epoch": 0.532435306628855, "grad_norm": 2.8268023421227975, "learning_rate": 4.718276245249593e-06, "loss": 0.4459, "step": 7510 }, { "epoch": 0.5325062034739454, "grad_norm": 2.0835018097763944, "learning_rate": 4.717129917685017e-06, "loss": 0.3631, "step": 7511 }, { "epoch": 0.5325771003190358, "grad_norm": 3.8388942429915818, "learning_rate": 4.715983605036395e-06, "loss": 0.3918, "step": 7512 }, { "epoch": 0.5326479971641261, "grad_norm": 2.596515251830574, "learning_rate": 4.714837307364173e-06, "loss": 0.4023, "step": 7513 }, { "epoch": 0.5327188940092166, "grad_norm": 2.195759268969757, "learning_rate": 4.7136910247288015e-06, "loss": 0.4129, "step": 7514 }, { "epoch": 0.532789790854307, "grad_norm": 2.409529945574899, "learning_rate": 4.712544757190717e-06, "loss": 0.3433, "step": 7515 }, { "epoch": 0.5328606876993974, "grad_norm": 2.4941178947532037, "learning_rate": 4.711398504810367e-06, "loss": 0.3861, "step": 7516 }, { "epoch": 0.5329315845444877, "grad_norm": 2.6838503569697414, "learning_rate": 4.710252267648196e-06, "loss": 0.3797, "step": 7517 }, { "epoch": 0.5330024813895782, "grad_norm": 4.9520821527800845, "learning_rate": 4.709106045764644e-06, "loss": 0.4592, "step": 7518 }, { "epoch": 0.5330733782346686, "grad_norm": 4.93807276312928, "learning_rate": 4.707959839220152e-06, "loss": 0.4123, "step": 7519 }, { "epoch": 0.5331442750797589, "grad_norm": 1.8439012094307965, "learning_rate": 4.706813648075158e-06, "loss": 0.3904, "step": 7520 }, { "epoch": 0.5332151719248494, "grad_norm": 1.0963574964620133, "learning_rate": 4.705667472390107e-06, "loss": 0.73, "step": 7521 }, { "epoch": 0.5332860687699398, "grad_norm": 1.8924020002287778, "learning_rate": 4.704521312225434e-06, "loss": 0.3998, "step": 7522 }, { "epoch": 0.5333569656150301, "grad_norm": 2.916696183238012, "learning_rate": 4.703375167641578e-06, "loss": 0.4055, "step": 7523 }, { "epoch": 0.5334278624601205, "grad_norm": 3.0707309962852896, "learning_rate": 4.702229038698976e-06, "loss": 0.3999, "step": 7524 }, { "epoch": 0.533498759305211, "grad_norm": 2.7833681501347978, "learning_rate": 4.701082925458064e-06, "loss": 0.3763, "step": 7525 }, { "epoch": 0.5335696561503013, "grad_norm": 3.4164044193927143, "learning_rate": 4.6999368279792755e-06, "loss": 0.4437, "step": 7526 }, { "epoch": 0.5336405529953917, "grad_norm": 7.241618197517226, "learning_rate": 4.698790746323051e-06, "loss": 0.4281, "step": 7527 }, { "epoch": 0.5337114498404821, "grad_norm": 1.8856197275525153, "learning_rate": 4.697644680549817e-06, "loss": 0.3683, "step": 7528 }, { "epoch": 0.5337823466855725, "grad_norm": 7.82459511594823, "learning_rate": 4.69649863072001e-06, "loss": 0.4443, "step": 7529 }, { "epoch": 0.5338532435306629, "grad_norm": 3.4482270529290595, "learning_rate": 4.695352596894063e-06, "loss": 0.4222, "step": 7530 }, { "epoch": 0.5339241403757533, "grad_norm": 0.9985045999887074, "learning_rate": 4.694206579132407e-06, "loss": 0.7321, "step": 7531 }, { "epoch": 0.5339950372208436, "grad_norm": 2.6784484285926946, "learning_rate": 4.69306057749547e-06, "loss": 0.3818, "step": 7532 }, { "epoch": 0.5340659340659341, "grad_norm": 2.614935404849433, "learning_rate": 4.691914592043682e-06, "loss": 0.3774, "step": 7533 }, { "epoch": 0.5341368309110245, "grad_norm": 28.29326248893312, "learning_rate": 4.690768622837475e-06, "loss": 0.4625, "step": 7534 }, { "epoch": 0.5342077277561148, "grad_norm": 1.883371091755868, "learning_rate": 4.689622669937273e-06, "loss": 0.3475, "step": 7535 }, { "epoch": 0.5342786246012052, "grad_norm": 2.9603352729542904, "learning_rate": 4.688476733403506e-06, "loss": 0.4311, "step": 7536 }, { "epoch": 0.5343495214462957, "grad_norm": 2.068942831446888, "learning_rate": 4.687330813296598e-06, "loss": 0.4426, "step": 7537 }, { "epoch": 0.534420418291386, "grad_norm": 4.330102047550978, "learning_rate": 4.686184909676974e-06, "loss": 0.398, "step": 7538 }, { "epoch": 0.5344913151364764, "grad_norm": 8.37251026772884, "learning_rate": 4.685039022605059e-06, "loss": 0.3781, "step": 7539 }, { "epoch": 0.5345622119815668, "grad_norm": 2.8298326169808843, "learning_rate": 4.683893152141279e-06, "loss": 0.438, "step": 7540 }, { "epoch": 0.5346331088266573, "grad_norm": 2.4930953664294826, "learning_rate": 4.682747298346052e-06, "loss": 0.4391, "step": 7541 }, { "epoch": 0.5347040056717476, "grad_norm": 3.2913323271881203, "learning_rate": 4.681601461279802e-06, "loss": 0.3933, "step": 7542 }, { "epoch": 0.534774902516838, "grad_norm": 5.151377116912571, "learning_rate": 4.680455641002953e-06, "loss": 0.4311, "step": 7543 }, { "epoch": 0.5348457993619284, "grad_norm": 8.066534581085186, "learning_rate": 4.679309837575919e-06, "loss": 0.4206, "step": 7544 }, { "epoch": 0.5349166962070188, "grad_norm": 3.3564083808225664, "learning_rate": 4.678164051059122e-06, "loss": 0.3935, "step": 7545 }, { "epoch": 0.5349875930521092, "grad_norm": 4.8941463817547595, "learning_rate": 4.677018281512981e-06, "loss": 0.3974, "step": 7546 }, { "epoch": 0.5350584898971996, "grad_norm": 3.219770376949551, "learning_rate": 4.675872528997913e-06, "loss": 0.4217, "step": 7547 }, { "epoch": 0.5351293867422899, "grad_norm": 3.2710264960424733, "learning_rate": 4.674726793574334e-06, "loss": 0.4144, "step": 7548 }, { "epoch": 0.5352002835873804, "grad_norm": 3.0981914007649354, "learning_rate": 4.673581075302659e-06, "loss": 0.384, "step": 7549 }, { "epoch": 0.5352711804324708, "grad_norm": 3.6781316331026206, "learning_rate": 4.6724353742433044e-06, "loss": 0.4554, "step": 7550 }, { "epoch": 0.5353420772775611, "grad_norm": 2.8673227450667413, "learning_rate": 4.6712896904566805e-06, "loss": 0.4031, "step": 7551 }, { "epoch": 0.5354129741226515, "grad_norm": 5.372445390592694, "learning_rate": 4.670144024003202e-06, "loss": 0.3952, "step": 7552 }, { "epoch": 0.535483870967742, "grad_norm": 5.480830113738312, "learning_rate": 4.668998374943283e-06, "loss": 0.389, "step": 7553 }, { "epoch": 0.5355547678128323, "grad_norm": 6.0881399876288995, "learning_rate": 4.667852743337331e-06, "loss": 0.4154, "step": 7554 }, { "epoch": 0.5356256646579227, "grad_norm": 5.055058721191126, "learning_rate": 4.666707129245757e-06, "loss": 0.4283, "step": 7555 }, { "epoch": 0.5356965615030131, "grad_norm": 7.327874993972264, "learning_rate": 4.665561532728972e-06, "loss": 0.4139, "step": 7556 }, { "epoch": 0.5357674583481035, "grad_norm": 2.34213848438779, "learning_rate": 4.6644159538473814e-06, "loss": 0.4477, "step": 7557 }, { "epoch": 0.5358383551931939, "grad_norm": 3.1935356467690714, "learning_rate": 4.663270392661392e-06, "loss": 0.4318, "step": 7558 }, { "epoch": 0.5359092520382843, "grad_norm": 4.1654769990007345, "learning_rate": 4.662124849231414e-06, "loss": 0.423, "step": 7559 }, { "epoch": 0.5359801488833746, "grad_norm": 10.001275900206926, "learning_rate": 4.660979323617852e-06, "loss": 0.4253, "step": 7560 }, { "epoch": 0.5360510457284651, "grad_norm": 2.5667046878621016, "learning_rate": 4.659833815881106e-06, "loss": 0.3736, "step": 7561 }, { "epoch": 0.5361219425735555, "grad_norm": 3.2687167192072657, "learning_rate": 4.658688326081584e-06, "loss": 0.432, "step": 7562 }, { "epoch": 0.5361928394186458, "grad_norm": 2.3541783082385606, "learning_rate": 4.657542854279688e-06, "loss": 0.3957, "step": 7563 }, { "epoch": 0.5362637362637362, "grad_norm": 5.633220270157474, "learning_rate": 4.656397400535818e-06, "loss": 0.45, "step": 7564 }, { "epoch": 0.5363346331088267, "grad_norm": 2.9304910142037985, "learning_rate": 4.655251964910373e-06, "loss": 0.3969, "step": 7565 }, { "epoch": 0.5364055299539171, "grad_norm": 2.973033876399142, "learning_rate": 4.654106547463758e-06, "loss": 0.3927, "step": 7566 }, { "epoch": 0.5364764267990074, "grad_norm": 2.399729889474943, "learning_rate": 4.6529611482563675e-06, "loss": 0.4077, "step": 7567 }, { "epoch": 0.5365473236440979, "grad_norm": 4.9335402315488075, "learning_rate": 4.6518157673486e-06, "loss": 0.4594, "step": 7568 }, { "epoch": 0.5366182204891883, "grad_norm": 4.517755574253455, "learning_rate": 4.650670404800855e-06, "loss": 0.4241, "step": 7569 }, { "epoch": 0.5366891173342786, "grad_norm": 4.89546348095558, "learning_rate": 4.649525060673525e-06, "loss": 0.4288, "step": 7570 }, { "epoch": 0.536760014179369, "grad_norm": 2.5465079683751175, "learning_rate": 4.648379735027004e-06, "loss": 0.4182, "step": 7571 }, { "epoch": 0.5368309110244595, "grad_norm": 1.0663176107093548, "learning_rate": 4.647234427921692e-06, "loss": 0.6705, "step": 7572 }, { "epoch": 0.5369018078695498, "grad_norm": 1.0679816511636262, "learning_rate": 4.646089139417974e-06, "loss": 0.6286, "step": 7573 }, { "epoch": 0.5369727047146402, "grad_norm": 3.5499837489598587, "learning_rate": 4.6449438695762475e-06, "loss": 0.4534, "step": 7574 }, { "epoch": 0.5370436015597306, "grad_norm": 3.2132198906270397, "learning_rate": 4.643798618456902e-06, "loss": 0.4281, "step": 7575 }, { "epoch": 0.537114498404821, "grad_norm": 3.036370282430828, "learning_rate": 4.6426533861203275e-06, "loss": 0.407, "step": 7576 }, { "epoch": 0.5371853952499114, "grad_norm": 2.841478056355765, "learning_rate": 4.641508172626911e-06, "loss": 0.4282, "step": 7577 }, { "epoch": 0.5372562920950018, "grad_norm": 4.734337300493095, "learning_rate": 4.640362978037043e-06, "loss": 0.3882, "step": 7578 }, { "epoch": 0.5373271889400921, "grad_norm": 2.741904095509444, "learning_rate": 4.639217802411111e-06, "loss": 0.4256, "step": 7579 }, { "epoch": 0.5373980857851826, "grad_norm": 6.174987956754735, "learning_rate": 4.638072645809498e-06, "loss": 0.386, "step": 7580 }, { "epoch": 0.537468982630273, "grad_norm": 4.215864322963932, "learning_rate": 4.636927508292592e-06, "loss": 0.3772, "step": 7581 }, { "epoch": 0.5375398794753633, "grad_norm": 2.0850247834139695, "learning_rate": 4.635782389920776e-06, "loss": 0.4421, "step": 7582 }, { "epoch": 0.5376107763204537, "grad_norm": 4.915849172557878, "learning_rate": 4.634637290754431e-06, "loss": 0.4723, "step": 7583 }, { "epoch": 0.5376816731655442, "grad_norm": 8.93704007398213, "learning_rate": 4.633492210853939e-06, "loss": 0.3989, "step": 7584 }, { "epoch": 0.5377525700106345, "grad_norm": 12.557909361931282, "learning_rate": 4.632347150279686e-06, "loss": 0.3971, "step": 7585 }, { "epoch": 0.5378234668557249, "grad_norm": 2.925437278031993, "learning_rate": 4.631202109092047e-06, "loss": 0.3637, "step": 7586 }, { "epoch": 0.5378943637008153, "grad_norm": 3.40031481278666, "learning_rate": 4.630057087351402e-06, "loss": 0.4263, "step": 7587 }, { "epoch": 0.5379652605459057, "grad_norm": 8.023321302223007, "learning_rate": 4.628912085118128e-06, "loss": 0.3731, "step": 7588 }, { "epoch": 0.5380361573909961, "grad_norm": 3.7030141383066515, "learning_rate": 4.6277671024526056e-06, "loss": 0.395, "step": 7589 }, { "epoch": 0.5381070542360865, "grad_norm": 3.2183992455578063, "learning_rate": 4.626622139415205e-06, "loss": 0.3682, "step": 7590 }, { "epoch": 0.5381779510811769, "grad_norm": 7.961021829739338, "learning_rate": 4.625477196066306e-06, "loss": 0.4844, "step": 7591 }, { "epoch": 0.5382488479262673, "grad_norm": 4.763756902066807, "learning_rate": 4.624332272466281e-06, "loss": 0.368, "step": 7592 }, { "epoch": 0.5383197447713577, "grad_norm": 3.13025967972331, "learning_rate": 4.623187368675501e-06, "loss": 0.3821, "step": 7593 }, { "epoch": 0.5383906416164481, "grad_norm": 5.614730058953161, "learning_rate": 4.622042484754337e-06, "loss": 0.4205, "step": 7594 }, { "epoch": 0.5384615384615384, "grad_norm": 2.885278042363604, "learning_rate": 4.620897620763165e-06, "loss": 0.3991, "step": 7595 }, { "epoch": 0.5385324353066289, "grad_norm": 2.5836736403990592, "learning_rate": 4.619752776762348e-06, "loss": 0.3804, "step": 7596 }, { "epoch": 0.5386033321517193, "grad_norm": 7.28428429171648, "learning_rate": 4.6186079528122565e-06, "loss": 0.4211, "step": 7597 }, { "epoch": 0.5386742289968096, "grad_norm": 5.279745829364038, "learning_rate": 4.617463148973262e-06, "loss": 0.4114, "step": 7598 }, { "epoch": 0.5387451258419, "grad_norm": 4.191265008069887, "learning_rate": 4.616318365305725e-06, "loss": 0.3876, "step": 7599 }, { "epoch": 0.5388160226869905, "grad_norm": 0.9893599194844662, "learning_rate": 4.615173601870014e-06, "loss": 0.6352, "step": 7600 }, { "epoch": 0.5388869195320808, "grad_norm": 2.39345621093259, "learning_rate": 4.614028858726495e-06, "loss": 0.3602, "step": 7601 }, { "epoch": 0.5389578163771712, "grad_norm": 2.6748933927696683, "learning_rate": 4.612884135935527e-06, "loss": 0.3495, "step": 7602 }, { "epoch": 0.5390287132222616, "grad_norm": 2.219985982691205, "learning_rate": 4.611739433557472e-06, "loss": 0.4231, "step": 7603 }, { "epoch": 0.539099610067352, "grad_norm": 2.827586549419929, "learning_rate": 4.6105947516526944e-06, "loss": 0.3879, "step": 7604 }, { "epoch": 0.5391705069124424, "grad_norm": 1.6144007112063086, "learning_rate": 4.609450090281554e-06, "loss": 0.3472, "step": 7605 }, { "epoch": 0.5392414037575328, "grad_norm": 1.7066066576803542, "learning_rate": 4.6083054495044074e-06, "loss": 0.4105, "step": 7606 }, { "epoch": 0.5393123006026231, "grad_norm": 5.482572411433925, "learning_rate": 4.6071608293816136e-06, "loss": 0.4034, "step": 7607 }, { "epoch": 0.5393831974477136, "grad_norm": 3.167268887181656, "learning_rate": 4.6060162299735305e-06, "loss": 0.4232, "step": 7608 }, { "epoch": 0.539454094292804, "grad_norm": 3.60972101378658, "learning_rate": 4.604871651340509e-06, "loss": 0.4433, "step": 7609 }, { "epoch": 0.5395249911378943, "grad_norm": 3.434220260380578, "learning_rate": 4.6037270935429096e-06, "loss": 0.4226, "step": 7610 }, { "epoch": 0.5395958879829847, "grad_norm": 2.6726088003123767, "learning_rate": 4.602582556641083e-06, "loss": 0.3957, "step": 7611 }, { "epoch": 0.5396667848280752, "grad_norm": 3.42304127414541, "learning_rate": 4.601438040695381e-06, "loss": 0.3938, "step": 7612 }, { "epoch": 0.5397376816731655, "grad_norm": 3.6983869463295345, "learning_rate": 4.6002935457661544e-06, "loss": 0.4264, "step": 7613 }, { "epoch": 0.5398085785182559, "grad_norm": 4.118290493009285, "learning_rate": 4.599149071913757e-06, "loss": 0.3985, "step": 7614 }, { "epoch": 0.5398794753633464, "grad_norm": 2.9926997333908107, "learning_rate": 4.5980046191985324e-06, "loss": 0.3776, "step": 7615 }, { "epoch": 0.5399503722084368, "grad_norm": 3.1262788706637976, "learning_rate": 4.59686018768083e-06, "loss": 0.4442, "step": 7616 }, { "epoch": 0.5400212690535271, "grad_norm": 4.131453220026484, "learning_rate": 4.5957157774210015e-06, "loss": 0.4789, "step": 7617 }, { "epoch": 0.5400921658986175, "grad_norm": 2.232025228386134, "learning_rate": 4.594571388479385e-06, "loss": 0.4124, "step": 7618 }, { "epoch": 0.540163062743708, "grad_norm": 2.557037249481928, "learning_rate": 4.59342702091633e-06, "loss": 0.3917, "step": 7619 }, { "epoch": 0.5402339595887983, "grad_norm": 4.487805677640229, "learning_rate": 4.592282674792179e-06, "loss": 0.4501, "step": 7620 }, { "epoch": 0.5403048564338887, "grad_norm": 5.868282129102756, "learning_rate": 4.591138350167273e-06, "loss": 0.4048, "step": 7621 }, { "epoch": 0.5403757532789791, "grad_norm": 3.335240163881629, "learning_rate": 4.589994047101953e-06, "loss": 0.4354, "step": 7622 }, { "epoch": 0.5404466501240694, "grad_norm": 2.080061554210406, "learning_rate": 4.5888497656565615e-06, "loss": 0.4055, "step": 7623 }, { "epoch": 0.5405175469691599, "grad_norm": 2.2452079754674683, "learning_rate": 4.587705505891436e-06, "loss": 0.4373, "step": 7624 }, { "epoch": 0.5405884438142503, "grad_norm": 5.01175627910322, "learning_rate": 4.586561267866913e-06, "loss": 0.4462, "step": 7625 }, { "epoch": 0.5406593406593406, "grad_norm": 3.7851957816872694, "learning_rate": 4.5854170516433294e-06, "loss": 0.3734, "step": 7626 }, { "epoch": 0.540730237504431, "grad_norm": 13.281595408877035, "learning_rate": 4.584272857281023e-06, "loss": 0.3901, "step": 7627 }, { "epoch": 0.5408011343495215, "grad_norm": 2.597788837213602, "learning_rate": 4.583128684840324e-06, "loss": 0.4289, "step": 7628 }, { "epoch": 0.5408720311946118, "grad_norm": 2.432354561700674, "learning_rate": 4.5819845343815675e-06, "loss": 0.4104, "step": 7629 }, { "epoch": 0.5409429280397022, "grad_norm": 5.463504562180108, "learning_rate": 4.5808404059650875e-06, "loss": 0.3936, "step": 7630 }, { "epoch": 0.5410138248847927, "grad_norm": 3.8241413618756352, "learning_rate": 4.579696299651211e-06, "loss": 0.4639, "step": 7631 }, { "epoch": 0.541084721729883, "grad_norm": 2.8801335110470396, "learning_rate": 4.578552215500271e-06, "loss": 0.4036, "step": 7632 }, { "epoch": 0.5411556185749734, "grad_norm": 3.0922770373497, "learning_rate": 4.577408153572592e-06, "loss": 0.4393, "step": 7633 }, { "epoch": 0.5412265154200638, "grad_norm": 2.7117118359555903, "learning_rate": 4.576264113928508e-06, "loss": 0.394, "step": 7634 }, { "epoch": 0.5412974122651542, "grad_norm": 4.13832533929574, "learning_rate": 4.575120096628336e-06, "loss": 0.3931, "step": 7635 }, { "epoch": 0.5413683091102446, "grad_norm": 3.1661756770036806, "learning_rate": 4.5739761017324076e-06, "loss": 0.3956, "step": 7636 }, { "epoch": 0.541439205955335, "grad_norm": 4.410522817151446, "learning_rate": 4.572832129301045e-06, "loss": 0.4143, "step": 7637 }, { "epoch": 0.5415101028004253, "grad_norm": 4.143828219911563, "learning_rate": 4.571688179394569e-06, "loss": 0.4468, "step": 7638 }, { "epoch": 0.5415809996455158, "grad_norm": 3.1135113727355095, "learning_rate": 4.570544252073304e-06, "loss": 0.3706, "step": 7639 }, { "epoch": 0.5416518964906062, "grad_norm": 2.534466256027491, "learning_rate": 4.569400347397567e-06, "loss": 0.466, "step": 7640 }, { "epoch": 0.5417227933356965, "grad_norm": 3.975896023185135, "learning_rate": 4.568256465427678e-06, "loss": 0.4945, "step": 7641 }, { "epoch": 0.5417936901807869, "grad_norm": 3.8034542508785303, "learning_rate": 4.5671126062239556e-06, "loss": 0.3887, "step": 7642 }, { "epoch": 0.5418645870258774, "grad_norm": 2.6635693890047705, "learning_rate": 4.565968769846717e-06, "loss": 0.3917, "step": 7643 }, { "epoch": 0.5419354838709678, "grad_norm": 3.112576138092579, "learning_rate": 4.564824956356275e-06, "loss": 0.4369, "step": 7644 }, { "epoch": 0.5420063807160581, "grad_norm": 1.8255406445285645, "learning_rate": 4.563681165812945e-06, "loss": 0.4058, "step": 7645 }, { "epoch": 0.5420772775611485, "grad_norm": 6.537817464720588, "learning_rate": 4.562537398277042e-06, "loss": 0.389, "step": 7646 }, { "epoch": 0.542148174406239, "grad_norm": 2.967665450405963, "learning_rate": 4.561393653808875e-06, "loss": 0.4159, "step": 7647 }, { "epoch": 0.5422190712513293, "grad_norm": 3.742809739790105, "learning_rate": 4.560249932468754e-06, "loss": 0.3809, "step": 7648 }, { "epoch": 0.5422899680964197, "grad_norm": 3.3939292736816373, "learning_rate": 4.55910623431699e-06, "loss": 0.3616, "step": 7649 }, { "epoch": 0.5423608649415101, "grad_norm": 2.9519263998329492, "learning_rate": 4.557962559413893e-06, "loss": 0.3817, "step": 7650 }, { "epoch": 0.5424317617866005, "grad_norm": 3.5844049335876242, "learning_rate": 4.556818907819766e-06, "loss": 0.4105, "step": 7651 }, { "epoch": 0.5425026586316909, "grad_norm": 1.9401003989331431, "learning_rate": 4.555675279594916e-06, "loss": 0.4076, "step": 7652 }, { "epoch": 0.5425735554767813, "grad_norm": 18.119955495062946, "learning_rate": 4.55453167479965e-06, "loss": 0.4335, "step": 7653 }, { "epoch": 0.5426444523218716, "grad_norm": 1.0556209148442552, "learning_rate": 4.553388093494265e-06, "loss": 0.6883, "step": 7654 }, { "epoch": 0.5427153491669621, "grad_norm": 3.804591419070862, "learning_rate": 4.552244535739069e-06, "loss": 0.4282, "step": 7655 }, { "epoch": 0.5427862460120525, "grad_norm": 1.8733159878850696, "learning_rate": 4.5511010015943605e-06, "loss": 0.4131, "step": 7656 }, { "epoch": 0.5428571428571428, "grad_norm": 1.1023755152440429, "learning_rate": 4.5499574911204385e-06, "loss": 0.6701, "step": 7657 }, { "epoch": 0.5429280397022332, "grad_norm": 3.052601658708382, "learning_rate": 4.5488140043776005e-06, "loss": 0.3515, "step": 7658 }, { "epoch": 0.5429989365473237, "grad_norm": 5.44396734799697, "learning_rate": 4.547670541426147e-06, "loss": 0.4055, "step": 7659 }, { "epoch": 0.543069833392414, "grad_norm": 4.077711109281783, "learning_rate": 4.546527102326369e-06, "loss": 0.422, "step": 7660 }, { "epoch": 0.5431407302375044, "grad_norm": 2.3399341139477707, "learning_rate": 4.545383687138562e-06, "loss": 0.3602, "step": 7661 }, { "epoch": 0.5432116270825949, "grad_norm": 4.230870715926978, "learning_rate": 4.544240295923022e-06, "loss": 0.3919, "step": 7662 }, { "epoch": 0.5432825239276852, "grad_norm": 1.158909118747762, "learning_rate": 4.543096928740039e-06, "loss": 0.694, "step": 7663 }, { "epoch": 0.5433534207727756, "grad_norm": 8.42293287837455, "learning_rate": 4.541953585649903e-06, "loss": 0.4392, "step": 7664 }, { "epoch": 0.543424317617866, "grad_norm": 2.516068281714834, "learning_rate": 4.540810266712903e-06, "loss": 0.3882, "step": 7665 }, { "epoch": 0.5434952144629563, "grad_norm": 4.4672722607521935, "learning_rate": 4.539666971989332e-06, "loss": 0.3999, "step": 7666 }, { "epoch": 0.5435661113080468, "grad_norm": 4.323302830487481, "learning_rate": 4.538523701539469e-06, "loss": 0.3916, "step": 7667 }, { "epoch": 0.5436370081531372, "grad_norm": 1.5667463744685695, "learning_rate": 4.5373804554236045e-06, "loss": 0.3624, "step": 7668 }, { "epoch": 0.5437079049982276, "grad_norm": 2.448972700786554, "learning_rate": 4.536237233702022e-06, "loss": 0.4461, "step": 7669 }, { "epoch": 0.543778801843318, "grad_norm": 5.041909409253747, "learning_rate": 4.535094036435005e-06, "loss": 0.4334, "step": 7670 }, { "epoch": 0.5438496986884084, "grad_norm": 2.3639005429530124, "learning_rate": 4.533950863682833e-06, "loss": 0.414, "step": 7671 }, { "epoch": 0.5439205955334988, "grad_norm": 2.918206822926192, "learning_rate": 4.532807715505789e-06, "loss": 0.3963, "step": 7672 }, { "epoch": 0.5439914923785891, "grad_norm": 6.616829809984152, "learning_rate": 4.53166459196415e-06, "loss": 0.4221, "step": 7673 }, { "epoch": 0.5440623892236796, "grad_norm": 2.717166372279145, "learning_rate": 4.530521493118193e-06, "loss": 0.3814, "step": 7674 }, { "epoch": 0.54413328606877, "grad_norm": 6.416386209771965, "learning_rate": 4.5293784190281985e-06, "loss": 0.4285, "step": 7675 }, { "epoch": 0.5442041829138603, "grad_norm": 2.9787183391052428, "learning_rate": 4.528235369754438e-06, "loss": 0.3923, "step": 7676 }, { "epoch": 0.5442750797589507, "grad_norm": 5.774099314730563, "learning_rate": 4.527092345357186e-06, "loss": 0.4211, "step": 7677 }, { "epoch": 0.5443459766040412, "grad_norm": 0.9965027721608063, "learning_rate": 4.5259493458967156e-06, "loss": 0.7063, "step": 7678 }, { "epoch": 0.5444168734491315, "grad_norm": 4.269040758109572, "learning_rate": 4.5248063714333e-06, "loss": 0.4707, "step": 7679 }, { "epoch": 0.5444877702942219, "grad_norm": 4.484182018516646, "learning_rate": 4.523663422027204e-06, "loss": 0.4044, "step": 7680 }, { "epoch": 0.5445586671393123, "grad_norm": 4.290414028379997, "learning_rate": 4.522520497738701e-06, "loss": 0.3996, "step": 7681 }, { "epoch": 0.5446295639844027, "grad_norm": 3.1479851271402297, "learning_rate": 4.521377598628059e-06, "loss": 0.3671, "step": 7682 }, { "epoch": 0.5447004608294931, "grad_norm": 3.976335185026888, "learning_rate": 4.520234724755539e-06, "loss": 0.4999, "step": 7683 }, { "epoch": 0.5447713576745835, "grad_norm": 2.4938501934528134, "learning_rate": 4.519091876181408e-06, "loss": 0.4508, "step": 7684 }, { "epoch": 0.5448422545196738, "grad_norm": 1.087138658130267, "learning_rate": 4.51794905296593e-06, "loss": 0.6407, "step": 7685 }, { "epoch": 0.5449131513647643, "grad_norm": 3.3161642944784853, "learning_rate": 4.516806255169365e-06, "loss": 0.3851, "step": 7686 }, { "epoch": 0.5449840482098547, "grad_norm": 0.9399323872458187, "learning_rate": 4.515663482851976e-06, "loss": 0.6172, "step": 7687 }, { "epoch": 0.545054945054945, "grad_norm": 3.1774684501891923, "learning_rate": 4.514520736074023e-06, "loss": 0.4658, "step": 7688 }, { "epoch": 0.5451258419000354, "grad_norm": 3.7849400876075263, "learning_rate": 4.51337801489576e-06, "loss": 0.3817, "step": 7689 }, { "epoch": 0.5451967387451259, "grad_norm": 5.497989716696259, "learning_rate": 4.512235319377446e-06, "loss": 0.4852, "step": 7690 }, { "epoch": 0.5452676355902162, "grad_norm": 3.14723559242485, "learning_rate": 4.511092649579335e-06, "loss": 0.3885, "step": 7691 }, { "epoch": 0.5453385324353066, "grad_norm": 2.8989493302287497, "learning_rate": 4.509950005561685e-06, "loss": 0.4365, "step": 7692 }, { "epoch": 0.545409429280397, "grad_norm": 2.2456901857155738, "learning_rate": 4.5088073873847416e-06, "loss": 0.4036, "step": 7693 }, { "epoch": 0.5454803261254875, "grad_norm": 5.565664766550308, "learning_rate": 4.507664795108761e-06, "loss": 0.4277, "step": 7694 }, { "epoch": 0.5455512229705778, "grad_norm": 3.4202945163202187, "learning_rate": 4.506522228793992e-06, "loss": 0.4327, "step": 7695 }, { "epoch": 0.5456221198156682, "grad_norm": 4.0434372032735455, "learning_rate": 4.505379688500683e-06, "loss": 0.3742, "step": 7696 }, { "epoch": 0.5456930166607586, "grad_norm": 2.2705813980588645, "learning_rate": 4.504237174289078e-06, "loss": 0.4141, "step": 7697 }, { "epoch": 0.545763913505849, "grad_norm": 3.0018664768699908, "learning_rate": 4.5030946862194295e-06, "loss": 0.4514, "step": 7698 }, { "epoch": 0.5458348103509394, "grad_norm": 5.182869176764151, "learning_rate": 4.501952224351973e-06, "loss": 0.4365, "step": 7699 }, { "epoch": 0.5459057071960298, "grad_norm": 2.9162370370996786, "learning_rate": 4.500809788746957e-06, "loss": 0.3672, "step": 7700 }, { "epoch": 0.5459766040411201, "grad_norm": 3.4897264698574264, "learning_rate": 4.4996673794646245e-06, "loss": 0.3807, "step": 7701 }, { "epoch": 0.5460475008862106, "grad_norm": 3.4312403635338398, "learning_rate": 4.4985249965652105e-06, "loss": 0.3827, "step": 7702 }, { "epoch": 0.546118397731301, "grad_norm": 1.0565548185027076, "learning_rate": 4.497382640108957e-06, "loss": 0.6845, "step": 7703 }, { "epoch": 0.5461892945763913, "grad_norm": 3.3858453089392815, "learning_rate": 4.496240310156101e-06, "loss": 0.4092, "step": 7704 }, { "epoch": 0.5462601914214817, "grad_norm": 4.955397727029775, "learning_rate": 4.495098006766877e-06, "loss": 0.4188, "step": 7705 }, { "epoch": 0.5463310882665722, "grad_norm": 4.8056762893622915, "learning_rate": 4.493955730001519e-06, "loss": 0.4188, "step": 7706 }, { "epoch": 0.5464019851116625, "grad_norm": 4.478639923324033, "learning_rate": 4.492813479920264e-06, "loss": 0.446, "step": 7707 }, { "epoch": 0.5464728819567529, "grad_norm": 6.56686437720445, "learning_rate": 4.491671256583341e-06, "loss": 0.3573, "step": 7708 }, { "epoch": 0.5465437788018433, "grad_norm": 166.6832361443238, "learning_rate": 4.49052906005098e-06, "loss": 0.4075, "step": 7709 }, { "epoch": 0.5466146756469337, "grad_norm": 3.6582715023930725, "learning_rate": 4.48938689038341e-06, "loss": 0.4096, "step": 7710 }, { "epoch": 0.5466855724920241, "grad_norm": 3.418597402248156, "learning_rate": 4.488244747640861e-06, "loss": 0.407, "step": 7711 }, { "epoch": 0.5467564693371145, "grad_norm": 1.8542968255606331, "learning_rate": 4.487102631883555e-06, "loss": 0.4153, "step": 7712 }, { "epoch": 0.5468273661822048, "grad_norm": 3.030562403488188, "learning_rate": 4.4859605431717194e-06, "loss": 0.4037, "step": 7713 }, { "epoch": 0.5468982630272953, "grad_norm": 2.948971419661273, "learning_rate": 4.484818481565578e-06, "loss": 0.3726, "step": 7714 }, { "epoch": 0.5469691598723857, "grad_norm": 2.6259299424730793, "learning_rate": 4.48367644712535e-06, "loss": 0.4253, "step": 7715 }, { "epoch": 0.547040056717476, "grad_norm": 19.21054817206627, "learning_rate": 4.482534439911257e-06, "loss": 0.4468, "step": 7716 }, { "epoch": 0.5471109535625664, "grad_norm": 4.17880702351715, "learning_rate": 4.481392459983519e-06, "loss": 0.4286, "step": 7717 }, { "epoch": 0.5471818504076569, "grad_norm": 3.356310812403281, "learning_rate": 4.48025050740235e-06, "loss": 0.4498, "step": 7718 }, { "epoch": 0.5472527472527473, "grad_norm": 6.456707692439357, "learning_rate": 4.4791085822279705e-06, "loss": 0.404, "step": 7719 }, { "epoch": 0.5473236440978376, "grad_norm": 5.306059669610961, "learning_rate": 4.477966684520592e-06, "loss": 0.3775, "step": 7720 }, { "epoch": 0.547394540942928, "grad_norm": 1.929361346937581, "learning_rate": 4.47682481434043e-06, "loss": 0.4161, "step": 7721 }, { "epoch": 0.5474654377880185, "grad_norm": 4.040778072414759, "learning_rate": 4.475682971747693e-06, "loss": 0.3866, "step": 7722 }, { "epoch": 0.5475363346331088, "grad_norm": 3.2540771723925914, "learning_rate": 4.474541156802594e-06, "loss": 0.3673, "step": 7723 }, { "epoch": 0.5476072314781992, "grad_norm": 3.227442701309202, "learning_rate": 4.473399369565343e-06, "loss": 0.3488, "step": 7724 }, { "epoch": 0.5476781283232897, "grad_norm": 29.470748772442832, "learning_rate": 4.472257610096141e-06, "loss": 0.3866, "step": 7725 }, { "epoch": 0.54774902516838, "grad_norm": 3.5028408043803108, "learning_rate": 4.4711158784552e-06, "loss": 0.4438, "step": 7726 }, { "epoch": 0.5478199220134704, "grad_norm": 2.4581214658636283, "learning_rate": 4.469974174702724e-06, "loss": 0.371, "step": 7727 }, { "epoch": 0.5478908188585608, "grad_norm": 2.820671619035366, "learning_rate": 4.468832498898913e-06, "loss": 0.4206, "step": 7728 }, { "epoch": 0.5479617157036512, "grad_norm": 4.915798082763673, "learning_rate": 4.4676908511039695e-06, "loss": 0.46, "step": 7729 }, { "epoch": 0.5480326125487416, "grad_norm": 2.57437984964446, "learning_rate": 4.466549231378094e-06, "loss": 0.3852, "step": 7730 }, { "epoch": 0.548103509393832, "grad_norm": 1.7880956131286851, "learning_rate": 4.4654076397814835e-06, "loss": 0.3622, "step": 7731 }, { "epoch": 0.5481744062389223, "grad_norm": 1.9528521198851299, "learning_rate": 4.464266076374338e-06, "loss": 0.3781, "step": 7732 }, { "epoch": 0.5482453030840128, "grad_norm": 1.8513311928237766, "learning_rate": 4.463124541216851e-06, "loss": 0.426, "step": 7733 }, { "epoch": 0.5483161999291032, "grad_norm": 5.25193409646767, "learning_rate": 4.4619830343692165e-06, "loss": 0.4102, "step": 7734 }, { "epoch": 0.5483870967741935, "grad_norm": 3.13385469133039, "learning_rate": 4.4608415558916265e-06, "loss": 0.4347, "step": 7735 }, { "epoch": 0.5484579936192839, "grad_norm": 2.302152830157069, "learning_rate": 4.459700105844272e-06, "loss": 0.4471, "step": 7736 }, { "epoch": 0.5485288904643744, "grad_norm": 3.3978667196078223, "learning_rate": 4.458558684287346e-06, "loss": 0.4028, "step": 7737 }, { "epoch": 0.5485997873094647, "grad_norm": 1.7683151356532922, "learning_rate": 4.457417291281031e-06, "loss": 0.3843, "step": 7738 }, { "epoch": 0.5486706841545551, "grad_norm": 3.3928545038503164, "learning_rate": 4.456275926885518e-06, "loss": 0.4532, "step": 7739 }, { "epoch": 0.5487415809996455, "grad_norm": 2.492427475657866, "learning_rate": 4.455134591160991e-06, "loss": 0.4069, "step": 7740 }, { "epoch": 0.5488124778447359, "grad_norm": 3.2726805592393555, "learning_rate": 4.453993284167632e-06, "loss": 0.3921, "step": 7741 }, { "epoch": 0.5488833746898263, "grad_norm": 2.2199659083426564, "learning_rate": 4.452852005965623e-06, "loss": 0.3866, "step": 7742 }, { "epoch": 0.5489542715349167, "grad_norm": 2.8407846556601024, "learning_rate": 4.451710756615148e-06, "loss": 0.3685, "step": 7743 }, { "epoch": 0.5490251683800071, "grad_norm": 4.597076271595619, "learning_rate": 4.45056953617638e-06, "loss": 0.4344, "step": 7744 }, { "epoch": 0.5490960652250975, "grad_norm": 2.6049945140977715, "learning_rate": 4.4494283447095015e-06, "loss": 0.3662, "step": 7745 }, { "epoch": 0.5491669620701879, "grad_norm": 4.675212899632339, "learning_rate": 4.448287182274687e-06, "loss": 0.3759, "step": 7746 }, { "epoch": 0.5492378589152783, "grad_norm": 2.361578101655057, "learning_rate": 4.44714604893211e-06, "loss": 0.393, "step": 7747 }, { "epoch": 0.5493087557603686, "grad_norm": 2.5110046852102497, "learning_rate": 4.446004944741944e-06, "loss": 0.441, "step": 7748 }, { "epoch": 0.5493796526054591, "grad_norm": 30.020437085466277, "learning_rate": 4.4448638697643605e-06, "loss": 0.4172, "step": 7749 }, { "epoch": 0.5494505494505495, "grad_norm": 2.8139159047103477, "learning_rate": 4.443722824059526e-06, "loss": 0.3934, "step": 7750 }, { "epoch": 0.5495214462956398, "grad_norm": 2.1670220018275623, "learning_rate": 4.442581807687614e-06, "loss": 0.3696, "step": 7751 }, { "epoch": 0.5495923431407302, "grad_norm": 2.859810459013069, "learning_rate": 4.4414408207087885e-06, "loss": 0.44, "step": 7752 }, { "epoch": 0.5496632399858207, "grad_norm": 1.1627443424516466, "learning_rate": 4.4402998631832155e-06, "loss": 0.6909, "step": 7753 }, { "epoch": 0.549734136830911, "grad_norm": 3.6595371209312146, "learning_rate": 4.439158935171057e-06, "loss": 0.4268, "step": 7754 }, { "epoch": 0.5498050336760014, "grad_norm": 2.293288306896857, "learning_rate": 4.438018036732475e-06, "loss": 0.4008, "step": 7755 }, { "epoch": 0.5498759305210918, "grad_norm": 1.8874217664776984, "learning_rate": 4.436877167927634e-06, "loss": 0.3749, "step": 7756 }, { "epoch": 0.5499468273661822, "grad_norm": 16.464364321566386, "learning_rate": 4.4357363288166865e-06, "loss": 0.4461, "step": 7757 }, { "epoch": 0.5500177242112726, "grad_norm": 4.01788166492554, "learning_rate": 4.434595519459794e-06, "loss": 0.3907, "step": 7758 }, { "epoch": 0.550088621056363, "grad_norm": 3.8686909513999757, "learning_rate": 4.433454739917113e-06, "loss": 0.4626, "step": 7759 }, { "epoch": 0.5501595179014533, "grad_norm": 2.2070342493853654, "learning_rate": 4.432313990248795e-06, "loss": 0.3871, "step": 7760 }, { "epoch": 0.5502304147465438, "grad_norm": 2.8882219038014147, "learning_rate": 4.4311732705149925e-06, "loss": 0.4335, "step": 7761 }, { "epoch": 0.5503013115916342, "grad_norm": 3.6317536701469386, "learning_rate": 4.430032580775858e-06, "loss": 0.4393, "step": 7762 }, { "epoch": 0.5503722084367245, "grad_norm": 6.538718953414375, "learning_rate": 4.42889192109154e-06, "loss": 0.4048, "step": 7763 }, { "epoch": 0.550443105281815, "grad_norm": 2.7321522490541845, "learning_rate": 4.427751291522188e-06, "loss": 0.4012, "step": 7764 }, { "epoch": 0.5505140021269054, "grad_norm": 6.737380693209335, "learning_rate": 4.426610692127947e-06, "loss": 0.4778, "step": 7765 }, { "epoch": 0.5505848989719957, "grad_norm": 2.7232581443753783, "learning_rate": 4.4254701229689626e-06, "loss": 0.4229, "step": 7766 }, { "epoch": 0.5506557958170861, "grad_norm": 2.3465130511627366, "learning_rate": 4.424329584105376e-06, "loss": 0.4042, "step": 7767 }, { "epoch": 0.5507266926621766, "grad_norm": 3.2659429381897938, "learning_rate": 4.423189075597329e-06, "loss": 0.379, "step": 7768 }, { "epoch": 0.5507975895072669, "grad_norm": 8.825467491578193, "learning_rate": 4.422048597504966e-06, "loss": 0.4438, "step": 7769 }, { "epoch": 0.5508684863523573, "grad_norm": 3.6503666646282285, "learning_rate": 4.420908149888417e-06, "loss": 0.4464, "step": 7770 }, { "epoch": 0.5509393831974477, "grad_norm": 2.659812735552267, "learning_rate": 4.419767732807825e-06, "loss": 0.3511, "step": 7771 }, { "epoch": 0.5510102800425382, "grad_norm": 2.7004789392037427, "learning_rate": 4.418627346323325e-06, "loss": 0.3664, "step": 7772 }, { "epoch": 0.5510811768876285, "grad_norm": 6.638969782422434, "learning_rate": 4.417486990495048e-06, "loss": 0.3823, "step": 7773 }, { "epoch": 0.5511520737327189, "grad_norm": 6.482055176926585, "learning_rate": 4.416346665383124e-06, "loss": 0.3628, "step": 7774 }, { "epoch": 0.5512229705778093, "grad_norm": 1.6582379506530294, "learning_rate": 4.41520637104769e-06, "loss": 0.3544, "step": 7775 }, { "epoch": 0.5512938674228997, "grad_norm": 4.130131189724383, "learning_rate": 4.414066107548867e-06, "loss": 0.3746, "step": 7776 }, { "epoch": 0.5513647642679901, "grad_norm": 1.91367834583521, "learning_rate": 4.4129258749467865e-06, "loss": 0.3986, "step": 7777 }, { "epoch": 0.5514356611130805, "grad_norm": 3.0456201995763763, "learning_rate": 4.411785673301573e-06, "loss": 0.3641, "step": 7778 }, { "epoch": 0.5515065579581708, "grad_norm": 2.323144177997009, "learning_rate": 4.41064550267335e-06, "loss": 0.3857, "step": 7779 }, { "epoch": 0.5515774548032613, "grad_norm": 3.195081316844613, "learning_rate": 4.409505363122239e-06, "loss": 0.3915, "step": 7780 }, { "epoch": 0.5516483516483517, "grad_norm": 3.140772685325762, "learning_rate": 4.40836525470836e-06, "loss": 0.4111, "step": 7781 }, { "epoch": 0.551719248493442, "grad_norm": 4.997127211681789, "learning_rate": 4.407225177491835e-06, "loss": 0.4385, "step": 7782 }, { "epoch": 0.5517901453385324, "grad_norm": 4.5156393118549545, "learning_rate": 4.4060851315327765e-06, "loss": 0.4313, "step": 7783 }, { "epoch": 0.5518610421836229, "grad_norm": 2.2240031832166833, "learning_rate": 4.404945116891304e-06, "loss": 0.3638, "step": 7784 }, { "epoch": 0.5519319390287132, "grad_norm": 4.9783681934625115, "learning_rate": 4.40380513362753e-06, "loss": 0.4429, "step": 7785 }, { "epoch": 0.5520028358738036, "grad_norm": 2.291978874166371, "learning_rate": 4.402665181801565e-06, "loss": 0.3914, "step": 7786 }, { "epoch": 0.552073732718894, "grad_norm": 2.4222335867623284, "learning_rate": 4.40152526147352e-06, "loss": 0.3917, "step": 7787 }, { "epoch": 0.5521446295639844, "grad_norm": 3.4800050528165976, "learning_rate": 4.4003853727035084e-06, "loss": 0.4515, "step": 7788 }, { "epoch": 0.5522155264090748, "grad_norm": 2.198701199246029, "learning_rate": 4.39924551555163e-06, "loss": 0.4176, "step": 7789 }, { "epoch": 0.5522864232541652, "grad_norm": 4.350618344558, "learning_rate": 4.398105690077995e-06, "loss": 0.4017, "step": 7790 }, { "epoch": 0.5523573200992555, "grad_norm": 3.408845018633914, "learning_rate": 4.396965896342709e-06, "loss": 0.4473, "step": 7791 }, { "epoch": 0.552428216944346, "grad_norm": 2.6406878176486854, "learning_rate": 4.395826134405869e-06, "loss": 0.4097, "step": 7792 }, { "epoch": 0.5524991137894364, "grad_norm": 7.148718321671551, "learning_rate": 4.394686404327578e-06, "loss": 0.4402, "step": 7793 }, { "epoch": 0.5525700106345267, "grad_norm": 2.8502035178940255, "learning_rate": 4.393546706167935e-06, "loss": 0.4146, "step": 7794 }, { "epoch": 0.5526409074796171, "grad_norm": 25.20489457147978, "learning_rate": 4.3924070399870385e-06, "loss": 0.4206, "step": 7795 }, { "epoch": 0.5527118043247076, "grad_norm": 1.7759116181224488, "learning_rate": 4.3912674058449815e-06, "loss": 0.3832, "step": 7796 }, { "epoch": 0.552782701169798, "grad_norm": 3.847400245538117, "learning_rate": 4.390127803801859e-06, "loss": 0.3872, "step": 7797 }, { "epoch": 0.5528535980148883, "grad_norm": 2.4616510931846185, "learning_rate": 4.388988233917763e-06, "loss": 0.4096, "step": 7798 }, { "epoch": 0.5529244948599787, "grad_norm": 3.2616886060118073, "learning_rate": 4.387848696252784e-06, "loss": 0.4351, "step": 7799 }, { "epoch": 0.5529953917050692, "grad_norm": 3.722897240226096, "learning_rate": 4.386709190867008e-06, "loss": 0.4007, "step": 7800 }, { "epoch": 0.5530662885501595, "grad_norm": 1.9664033719006713, "learning_rate": 4.3855697178205294e-06, "loss": 0.3673, "step": 7801 }, { "epoch": 0.5531371853952499, "grad_norm": 2.6173817385775155, "learning_rate": 4.384430277173424e-06, "loss": 0.3608, "step": 7802 }, { "epoch": 0.5532080822403403, "grad_norm": 2.1325258266078886, "learning_rate": 4.3832908689857815e-06, "loss": 0.3861, "step": 7803 }, { "epoch": 0.5532789790854307, "grad_norm": 3.658561992678761, "learning_rate": 4.382151493317683e-06, "loss": 0.4541, "step": 7804 }, { "epoch": 0.5533498759305211, "grad_norm": 3.096421635937217, "learning_rate": 4.381012150229207e-06, "loss": 0.3792, "step": 7805 }, { "epoch": 0.5534207727756115, "grad_norm": 2.9114517510882387, "learning_rate": 4.379872839780432e-06, "loss": 0.3595, "step": 7806 }, { "epoch": 0.5534916696207018, "grad_norm": 3.1114536682889, "learning_rate": 4.378733562031439e-06, "loss": 0.448, "step": 7807 }, { "epoch": 0.5535625664657923, "grad_norm": 4.624594255445958, "learning_rate": 4.377594317042296e-06, "loss": 0.4725, "step": 7808 }, { "epoch": 0.5536334633108827, "grad_norm": 2.6341295737296715, "learning_rate": 4.376455104873079e-06, "loss": 0.4063, "step": 7809 }, { "epoch": 0.553704360155973, "grad_norm": 2.1222752711852766, "learning_rate": 4.375315925583863e-06, "loss": 0.4059, "step": 7810 }, { "epoch": 0.5537752570010634, "grad_norm": 2.630547490073835, "learning_rate": 4.374176779234714e-06, "loss": 0.4145, "step": 7811 }, { "epoch": 0.5538461538461539, "grad_norm": 3.4215130157499325, "learning_rate": 4.373037665885701e-06, "loss": 0.4194, "step": 7812 }, { "epoch": 0.5539170506912442, "grad_norm": 3.2164661264425685, "learning_rate": 4.37189858559689e-06, "loss": 0.3807, "step": 7813 }, { "epoch": 0.5539879475363346, "grad_norm": 7.664975001103411, "learning_rate": 4.370759538428347e-06, "loss": 0.3909, "step": 7814 }, { "epoch": 0.554058844381425, "grad_norm": 4.136047990857018, "learning_rate": 4.369620524440133e-06, "loss": 0.3887, "step": 7815 }, { "epoch": 0.5541297412265154, "grad_norm": 3.5210735578180077, "learning_rate": 4.368481543692309e-06, "loss": 0.4307, "step": 7816 }, { "epoch": 0.5542006380716058, "grad_norm": 4.3270084711985906, "learning_rate": 4.367342596244938e-06, "loss": 0.3844, "step": 7817 }, { "epoch": 0.5542715349166962, "grad_norm": 3.627784269893231, "learning_rate": 4.366203682158073e-06, "loss": 0.4251, "step": 7818 }, { "epoch": 0.5543424317617865, "grad_norm": 2.9418270874108376, "learning_rate": 4.36506480149177e-06, "loss": 0.3469, "step": 7819 }, { "epoch": 0.554413328606877, "grad_norm": 6.461412782255725, "learning_rate": 4.363925954306089e-06, "loss": 0.3871, "step": 7820 }, { "epoch": 0.5544842254519674, "grad_norm": 4.684839396377548, "learning_rate": 4.362787140661073e-06, "loss": 0.3993, "step": 7821 }, { "epoch": 0.5545551222970578, "grad_norm": 7.478222719995005, "learning_rate": 4.3616483606167806e-06, "loss": 0.4647, "step": 7822 }, { "epoch": 0.5546260191421482, "grad_norm": 2.0339094977979637, "learning_rate": 4.360509614233256e-06, "loss": 0.3566, "step": 7823 }, { "epoch": 0.5546969159872386, "grad_norm": 2.875449853051551, "learning_rate": 4.359370901570549e-06, "loss": 0.3825, "step": 7824 }, { "epoch": 0.554767812832329, "grad_norm": 3.8293283441833808, "learning_rate": 4.3582322226887016e-06, "loss": 0.3827, "step": 7825 }, { "epoch": 0.5548387096774193, "grad_norm": 2.6779474634910114, "learning_rate": 4.357093577647759e-06, "loss": 0.3932, "step": 7826 }, { "epoch": 0.5549096065225098, "grad_norm": 2.1800976593441175, "learning_rate": 4.355954966507765e-06, "loss": 0.4337, "step": 7827 }, { "epoch": 0.5549805033676002, "grad_norm": 2.565269338532889, "learning_rate": 4.354816389328755e-06, "loss": 0.358, "step": 7828 }, { "epoch": 0.5550514002126905, "grad_norm": 2.782861663720808, "learning_rate": 4.353677846170769e-06, "loss": 0.3581, "step": 7829 }, { "epoch": 0.5551222970577809, "grad_norm": 2.8840980594352428, "learning_rate": 4.3525393370938454e-06, "loss": 0.4277, "step": 7830 }, { "epoch": 0.5551931939028714, "grad_norm": 6.0667506637885165, "learning_rate": 4.351400862158016e-06, "loss": 0.4304, "step": 7831 }, { "epoch": 0.5552640907479617, "grad_norm": 3.0133274165648167, "learning_rate": 4.350262421423313e-06, "loss": 0.4063, "step": 7832 }, { "epoch": 0.5553349875930521, "grad_norm": 3.765286168298084, "learning_rate": 4.349124014949771e-06, "loss": 0.4012, "step": 7833 }, { "epoch": 0.5554058844381425, "grad_norm": 2.556662061943192, "learning_rate": 4.347985642797414e-06, "loss": 0.3813, "step": 7834 }, { "epoch": 0.5554767812832329, "grad_norm": 4.1960158220280945, "learning_rate": 4.346847305026272e-06, "loss": 0.4035, "step": 7835 }, { "epoch": 0.5555476781283233, "grad_norm": 3.8652043848424094, "learning_rate": 4.345709001696372e-06, "loss": 0.4031, "step": 7836 }, { "epoch": 0.5556185749734137, "grad_norm": 2.9529112231757653, "learning_rate": 4.344570732867734e-06, "loss": 0.4424, "step": 7837 }, { "epoch": 0.555689471818504, "grad_norm": 4.355497509678632, "learning_rate": 4.343432498600381e-06, "loss": 0.3892, "step": 7838 }, { "epoch": 0.5557603686635945, "grad_norm": 2.0985671016786056, "learning_rate": 4.342294298954334e-06, "loss": 0.3628, "step": 7839 }, { "epoch": 0.5558312655086849, "grad_norm": 21.526010455169256, "learning_rate": 4.3411561339896115e-06, "loss": 0.4163, "step": 7840 }, { "epoch": 0.5559021623537752, "grad_norm": 8.902607728071878, "learning_rate": 4.340018003766229e-06, "loss": 0.4063, "step": 7841 }, { "epoch": 0.5559730591988656, "grad_norm": 4.278634180735949, "learning_rate": 4.3388799083442e-06, "loss": 0.4571, "step": 7842 }, { "epoch": 0.5560439560439561, "grad_norm": 2.29599860478324, "learning_rate": 4.337741847783539e-06, "loss": 0.3687, "step": 7843 }, { "epoch": 0.5561148528890464, "grad_norm": 10.14888367316105, "learning_rate": 4.3366038221442554e-06, "loss": 0.3753, "step": 7844 }, { "epoch": 0.5561857497341368, "grad_norm": 3.760081048625243, "learning_rate": 4.3354658314863575e-06, "loss": 0.4277, "step": 7845 }, { "epoch": 0.5562566465792272, "grad_norm": 2.9520398129835623, "learning_rate": 4.334327875869856e-06, "loss": 0.4306, "step": 7846 }, { "epoch": 0.5563275434243177, "grad_norm": 10.441799580607253, "learning_rate": 4.333189955354751e-06, "loss": 0.4401, "step": 7847 }, { "epoch": 0.556398440269408, "grad_norm": 3.481031389767924, "learning_rate": 4.33205207000105e-06, "loss": 0.4165, "step": 7848 }, { "epoch": 0.5564693371144984, "grad_norm": 2.895350751545099, "learning_rate": 4.330914219868754e-06, "loss": 0.4355, "step": 7849 }, { "epoch": 0.5565402339595888, "grad_norm": 16.08039529509877, "learning_rate": 4.32977640501786e-06, "loss": 0.4635, "step": 7850 }, { "epoch": 0.5566111308046792, "grad_norm": 2.6164874550914914, "learning_rate": 4.328638625508368e-06, "loss": 0.3562, "step": 7851 }, { "epoch": 0.5566820276497696, "grad_norm": 3.069345652324453, "learning_rate": 4.327500881400276e-06, "loss": 0.397, "step": 7852 }, { "epoch": 0.55675292449486, "grad_norm": 2.895245442700753, "learning_rate": 4.326363172753573e-06, "loss": 0.4071, "step": 7853 }, { "epoch": 0.5568238213399503, "grad_norm": 2.6816125014578955, "learning_rate": 4.325225499628255e-06, "loss": 0.4321, "step": 7854 }, { "epoch": 0.5568947181850408, "grad_norm": 1.9364978690726404, "learning_rate": 4.324087862084311e-06, "loss": 0.4068, "step": 7855 }, { "epoch": 0.5569656150301312, "grad_norm": 2.329319716339936, "learning_rate": 4.322950260181732e-06, "loss": 0.4012, "step": 7856 }, { "epoch": 0.5570365118752215, "grad_norm": 2.0938366857557833, "learning_rate": 4.321812693980501e-06, "loss": 0.3822, "step": 7857 }, { "epoch": 0.557107408720312, "grad_norm": 4.056274910986311, "learning_rate": 4.3206751635406025e-06, "loss": 0.3646, "step": 7858 }, { "epoch": 0.5571783055654024, "grad_norm": 5.052531411030714, "learning_rate": 4.319537668922024e-06, "loss": 0.4367, "step": 7859 }, { "epoch": 0.5572492024104927, "grad_norm": 3.8823065404765, "learning_rate": 4.318400210184742e-06, "loss": 0.4287, "step": 7860 }, { "epoch": 0.5573200992555831, "grad_norm": 3.079933216069578, "learning_rate": 4.317262787388738e-06, "loss": 0.3709, "step": 7861 }, { "epoch": 0.5573909961006736, "grad_norm": 2.2483768693593067, "learning_rate": 4.316125400593989e-06, "loss": 0.3476, "step": 7862 }, { "epoch": 0.5574618929457639, "grad_norm": 4.715161240473044, "learning_rate": 4.314988049860468e-06, "loss": 0.3789, "step": 7863 }, { "epoch": 0.5575327897908543, "grad_norm": 3.5316297492302504, "learning_rate": 4.313850735248149e-06, "loss": 0.4116, "step": 7864 }, { "epoch": 0.5576036866359447, "grad_norm": 2.966380911825684, "learning_rate": 4.312713456817009e-06, "loss": 0.4534, "step": 7865 }, { "epoch": 0.557674583481035, "grad_norm": 3.9149083040532937, "learning_rate": 4.311576214627009e-06, "loss": 0.3639, "step": 7866 }, { "epoch": 0.5577454803261255, "grad_norm": 3.88367219275354, "learning_rate": 4.310439008738122e-06, "loss": 0.4074, "step": 7867 }, { "epoch": 0.5578163771712159, "grad_norm": 10.058576714795192, "learning_rate": 4.3093018392103115e-06, "loss": 0.4007, "step": 7868 }, { "epoch": 0.5578872740163062, "grad_norm": 1.8163879149513134, "learning_rate": 4.308164706103544e-06, "loss": 0.3982, "step": 7869 }, { "epoch": 0.5579581708613967, "grad_norm": 3.5806014631898204, "learning_rate": 4.307027609477778e-06, "loss": 0.4684, "step": 7870 }, { "epoch": 0.5580290677064871, "grad_norm": 2.7015512138763285, "learning_rate": 4.305890549392974e-06, "loss": 0.4282, "step": 7871 }, { "epoch": 0.5580999645515775, "grad_norm": 10.941106453717284, "learning_rate": 4.304753525909094e-06, "loss": 0.4305, "step": 7872 }, { "epoch": 0.5581708613966678, "grad_norm": 3.258219533797991, "learning_rate": 4.30361653908609e-06, "loss": 0.4153, "step": 7873 }, { "epoch": 0.5582417582417583, "grad_norm": 1.8618496705788898, "learning_rate": 4.302479588983917e-06, "loss": 0.4149, "step": 7874 }, { "epoch": 0.5583126550868487, "grad_norm": 3.704306392342101, "learning_rate": 4.3013426756625285e-06, "loss": 0.3752, "step": 7875 }, { "epoch": 0.558383551931939, "grad_norm": 1.8780890348437125, "learning_rate": 4.3002057991818735e-06, "loss": 0.4525, "step": 7876 }, { "epoch": 0.5584544487770294, "grad_norm": 3.1200527083828375, "learning_rate": 4.299068959601899e-06, "loss": 0.383, "step": 7877 }, { "epoch": 0.5585253456221199, "grad_norm": 2.7309394103315783, "learning_rate": 4.2979321569825575e-06, "loss": 0.4009, "step": 7878 }, { "epoch": 0.5585962424672102, "grad_norm": 1.9392324595876818, "learning_rate": 4.296795391383785e-06, "loss": 0.3755, "step": 7879 }, { "epoch": 0.5586671393123006, "grad_norm": 2.7491178581010014, "learning_rate": 4.29565866286553e-06, "loss": 0.4004, "step": 7880 }, { "epoch": 0.558738036157391, "grad_norm": 3.5157319751688054, "learning_rate": 4.294521971487732e-06, "loss": 0.4631, "step": 7881 }, { "epoch": 0.5588089330024814, "grad_norm": 4.357482858413467, "learning_rate": 4.2933853173103286e-06, "loss": 0.4174, "step": 7882 }, { "epoch": 0.5588798298475718, "grad_norm": 2.336829676690532, "learning_rate": 4.292248700393255e-06, "loss": 0.3582, "step": 7883 }, { "epoch": 0.5589507266926622, "grad_norm": 6.579642239144495, "learning_rate": 4.291112120796449e-06, "loss": 0.4019, "step": 7884 }, { "epoch": 0.5590216235377525, "grad_norm": 3.992630443687115, "learning_rate": 4.289975578579843e-06, "loss": 0.4189, "step": 7885 }, { "epoch": 0.559092520382843, "grad_norm": 2.438203533250925, "learning_rate": 4.288839073803365e-06, "loss": 0.4368, "step": 7886 }, { "epoch": 0.5591634172279334, "grad_norm": 2.823789743726917, "learning_rate": 4.287702606526946e-06, "loss": 0.3795, "step": 7887 }, { "epoch": 0.5592343140730237, "grad_norm": 3.7373317033570443, "learning_rate": 4.286566176810513e-06, "loss": 0.4052, "step": 7888 }, { "epoch": 0.5593052109181141, "grad_norm": 3.367199053374572, "learning_rate": 4.285429784713989e-06, "loss": 0.4138, "step": 7889 }, { "epoch": 0.5593761077632046, "grad_norm": 6.690607816718074, "learning_rate": 4.2842934302972974e-06, "loss": 0.3839, "step": 7890 }, { "epoch": 0.5594470046082949, "grad_norm": 3.5272012035586693, "learning_rate": 4.283157113620361e-06, "loss": 0.3773, "step": 7891 }, { "epoch": 0.5595179014533853, "grad_norm": 3.5308481372687033, "learning_rate": 4.282020834743097e-06, "loss": 0.401, "step": 7892 }, { "epoch": 0.5595887982984757, "grad_norm": 2.1230657581131385, "learning_rate": 4.2808845937254205e-06, "loss": 0.3756, "step": 7893 }, { "epoch": 0.5596596951435661, "grad_norm": 2.8662810536235654, "learning_rate": 4.279748390627251e-06, "loss": 0.4019, "step": 7894 }, { "epoch": 0.5597305919886565, "grad_norm": 3.35024333401475, "learning_rate": 4.278612225508497e-06, "loss": 0.3035, "step": 7895 }, { "epoch": 0.5598014888337469, "grad_norm": 9.695589234707576, "learning_rate": 4.27747609842907e-06, "loss": 0.3779, "step": 7896 }, { "epoch": 0.5598723856788372, "grad_norm": 2.7765496345189167, "learning_rate": 4.276340009448881e-06, "loss": 0.4198, "step": 7897 }, { "epoch": 0.5599432825239277, "grad_norm": 2.4094976399169736, "learning_rate": 4.275203958627836e-06, "loss": 0.3535, "step": 7898 }, { "epoch": 0.5600141793690181, "grad_norm": 2.0574249135695446, "learning_rate": 4.274067946025839e-06, "loss": 0.3722, "step": 7899 }, { "epoch": 0.5600850762141085, "grad_norm": 2.4088704189916137, "learning_rate": 4.272931971702793e-06, "loss": 0.3859, "step": 7900 }, { "epoch": 0.5601559730591988, "grad_norm": 2.612383220983172, "learning_rate": 4.2717960357186e-06, "loss": 0.4428, "step": 7901 }, { "epoch": 0.5602268699042893, "grad_norm": 2.173013530977005, "learning_rate": 4.2706601381331575e-06, "loss": 0.3836, "step": 7902 }, { "epoch": 0.5602977667493797, "grad_norm": 2.666052304909778, "learning_rate": 4.269524279006361e-06, "loss": 0.4591, "step": 7903 }, { "epoch": 0.56036866359447, "grad_norm": 22.98615105706165, "learning_rate": 4.26838845839811e-06, "loss": 0.3834, "step": 7904 }, { "epoch": 0.5604395604395604, "grad_norm": 1.9497665241426256, "learning_rate": 4.267252676368291e-06, "loss": 0.3532, "step": 7905 }, { "epoch": 0.5605104572846509, "grad_norm": 2.304456118193217, "learning_rate": 4.266116932976799e-06, "loss": 0.3679, "step": 7906 }, { "epoch": 0.5605813541297412, "grad_norm": 5.500407552704017, "learning_rate": 4.264981228283522e-06, "loss": 0.3762, "step": 7907 }, { "epoch": 0.5606522509748316, "grad_norm": 2.9022306114341165, "learning_rate": 4.2638455623483445e-06, "loss": 0.3611, "step": 7908 }, { "epoch": 0.560723147819922, "grad_norm": 2.4986780663416304, "learning_rate": 4.262709935231152e-06, "loss": 0.474, "step": 7909 }, { "epoch": 0.5607940446650124, "grad_norm": 2.408157074899837, "learning_rate": 4.26157434699183e-06, "loss": 0.3504, "step": 7910 }, { "epoch": 0.5608649415101028, "grad_norm": 2.5748388672028977, "learning_rate": 4.260438797690253e-06, "loss": 0.4069, "step": 7911 }, { "epoch": 0.5609358383551932, "grad_norm": 6.3240828023456634, "learning_rate": 4.2593032873863036e-06, "loss": 0.3762, "step": 7912 }, { "epoch": 0.5610067352002835, "grad_norm": 3.959500297909677, "learning_rate": 4.258167816139858e-06, "loss": 0.4192, "step": 7913 }, { "epoch": 0.561077632045374, "grad_norm": 3.056316059158024, "learning_rate": 4.257032384010791e-06, "loss": 0.4542, "step": 7914 }, { "epoch": 0.5611485288904644, "grad_norm": 3.292605161695474, "learning_rate": 4.2558969910589696e-06, "loss": 0.3833, "step": 7915 }, { "epoch": 0.5612194257355547, "grad_norm": 3.4462257532884926, "learning_rate": 4.254761637344271e-06, "loss": 0.3644, "step": 7916 }, { "epoch": 0.5612903225806452, "grad_norm": 2.9256184482638403, "learning_rate": 4.25362632292656e-06, "loss": 0.3709, "step": 7917 }, { "epoch": 0.5613612194257356, "grad_norm": 5.382888221533896, "learning_rate": 4.252491047865701e-06, "loss": 0.3677, "step": 7918 }, { "epoch": 0.5614321162708259, "grad_norm": 6.521675560935786, "learning_rate": 4.25135581222156e-06, "loss": 0.4277, "step": 7919 }, { "epoch": 0.5615030131159163, "grad_norm": 2.6921600766889515, "learning_rate": 4.250220616054001e-06, "loss": 0.4527, "step": 7920 }, { "epoch": 0.5615739099610068, "grad_norm": 2.9426981440232916, "learning_rate": 4.249085459422878e-06, "loss": 0.4248, "step": 7921 }, { "epoch": 0.5616448068060971, "grad_norm": 7.489918680293593, "learning_rate": 4.2479503423880515e-06, "loss": 0.4199, "step": 7922 }, { "epoch": 0.5617157036511875, "grad_norm": 2.7094662996191405, "learning_rate": 4.246815265009382e-06, "loss": 0.3022, "step": 7923 }, { "epoch": 0.5617866004962779, "grad_norm": 4.702845548418686, "learning_rate": 4.245680227346714e-06, "loss": 0.3492, "step": 7924 }, { "epoch": 0.5618574973413684, "grad_norm": 2.694713407733457, "learning_rate": 4.244545229459904e-06, "loss": 0.4217, "step": 7925 }, { "epoch": 0.5619283941864587, "grad_norm": 2.8005045097279506, "learning_rate": 4.243410271408801e-06, "loss": 0.4554, "step": 7926 }, { "epoch": 0.5619992910315491, "grad_norm": 4.310277122740637, "learning_rate": 4.242275353253253e-06, "loss": 0.3552, "step": 7927 }, { "epoch": 0.5620701878766395, "grad_norm": 4.612561442168191, "learning_rate": 4.2411404750531025e-06, "loss": 0.3726, "step": 7928 }, { "epoch": 0.5621410847217299, "grad_norm": 5.674405163196491, "learning_rate": 4.240005636868195e-06, "loss": 0.345, "step": 7929 }, { "epoch": 0.5622119815668203, "grad_norm": 3.035237886988049, "learning_rate": 4.238870838758371e-06, "loss": 0.3993, "step": 7930 }, { "epoch": 0.5622828784119107, "grad_norm": 3.334511806006606, "learning_rate": 4.237736080783467e-06, "loss": 0.3995, "step": 7931 }, { "epoch": 0.562353775257001, "grad_norm": 2.786153055501271, "learning_rate": 4.236601363003322e-06, "loss": 0.4056, "step": 7932 }, { "epoch": 0.5624246721020915, "grad_norm": 5.433195038189313, "learning_rate": 4.23546668547777e-06, "loss": 0.3965, "step": 7933 }, { "epoch": 0.5624955689471819, "grad_norm": 6.454410573284594, "learning_rate": 4.234332048266643e-06, "loss": 0.4022, "step": 7934 }, { "epoch": 0.5625664657922722, "grad_norm": 2.3074941003567995, "learning_rate": 4.233197451429771e-06, "loss": 0.3932, "step": 7935 }, { "epoch": 0.5626373626373626, "grad_norm": 8.891634186139843, "learning_rate": 4.232062895026983e-06, "loss": 0.4454, "step": 7936 }, { "epoch": 0.5627082594824531, "grad_norm": 3.0748367037060955, "learning_rate": 4.230928379118105e-06, "loss": 0.3888, "step": 7937 }, { "epoch": 0.5627791563275434, "grad_norm": 3.5119527836183453, "learning_rate": 4.22979390376296e-06, "loss": 0.4096, "step": 7938 }, { "epoch": 0.5628500531726338, "grad_norm": 3.579747885932074, "learning_rate": 4.2286594690213724e-06, "loss": 0.3815, "step": 7939 }, { "epoch": 0.5629209500177242, "grad_norm": 2.525517957542741, "learning_rate": 4.2275250749531576e-06, "loss": 0.4296, "step": 7940 }, { "epoch": 0.5629918468628146, "grad_norm": 3.263164141446425, "learning_rate": 4.226390721618135e-06, "loss": 0.4699, "step": 7941 }, { "epoch": 0.563062743707905, "grad_norm": 3.1530008800862266, "learning_rate": 4.22525640907612e-06, "loss": 0.3886, "step": 7942 }, { "epoch": 0.5631336405529954, "grad_norm": 3.533548820546675, "learning_rate": 4.224122137386928e-06, "loss": 0.4172, "step": 7943 }, { "epoch": 0.5632045373980857, "grad_norm": 2.386116290803424, "learning_rate": 4.222987906610366e-06, "loss": 0.4581, "step": 7944 }, { "epoch": 0.5632754342431762, "grad_norm": 4.365099702293361, "learning_rate": 4.221853716806245e-06, "loss": 0.4112, "step": 7945 }, { "epoch": 0.5633463310882666, "grad_norm": 3.9351778025701107, "learning_rate": 4.2207195680343725e-06, "loss": 0.4629, "step": 7946 }, { "epoch": 0.5634172279333569, "grad_norm": 2.814763265389179, "learning_rate": 4.219585460354551e-06, "loss": 0.5077, "step": 7947 }, { "epoch": 0.5634881247784473, "grad_norm": 2.202129167594295, "learning_rate": 4.218451393826583e-06, "loss": 0.428, "step": 7948 }, { "epoch": 0.5635590216235378, "grad_norm": 1.7495937427136723, "learning_rate": 4.217317368510271e-06, "loss": 0.4042, "step": 7949 }, { "epoch": 0.5636299184686282, "grad_norm": 2.075852701721945, "learning_rate": 4.216183384465412e-06, "loss": 0.3963, "step": 7950 }, { "epoch": 0.5637008153137185, "grad_norm": 3.5578639934714213, "learning_rate": 4.2150494417518e-06, "loss": 0.3625, "step": 7951 }, { "epoch": 0.563771712158809, "grad_norm": 6.192737649534769, "learning_rate": 4.213915540429231e-06, "loss": 0.4392, "step": 7952 }, { "epoch": 0.5638426090038994, "grad_norm": 4.827678112946491, "learning_rate": 4.212781680557495e-06, "loss": 0.4028, "step": 7953 }, { "epoch": 0.5639135058489897, "grad_norm": 2.0876233838055946, "learning_rate": 4.211647862196381e-06, "loss": 0.3724, "step": 7954 }, { "epoch": 0.5639844026940801, "grad_norm": 3.291537913360389, "learning_rate": 4.21051408540568e-06, "loss": 0.3617, "step": 7955 }, { "epoch": 0.5640552995391706, "grad_norm": 2.4035302843221267, "learning_rate": 4.209380350245171e-06, "loss": 0.4059, "step": 7956 }, { "epoch": 0.5641261963842609, "grad_norm": 5.2216623001606814, "learning_rate": 4.20824665677464e-06, "loss": 0.43, "step": 7957 }, { "epoch": 0.5641970932293513, "grad_norm": 2.2758359175369525, "learning_rate": 4.2071130050538675e-06, "loss": 0.3732, "step": 7958 }, { "epoch": 0.5642679900744417, "grad_norm": 10.0250400297067, "learning_rate": 4.205979395142632e-06, "loss": 0.3813, "step": 7959 }, { "epoch": 0.564338886919532, "grad_norm": 2.9085502773466305, "learning_rate": 4.2048458271007066e-06, "loss": 0.427, "step": 7960 }, { "epoch": 0.5644097837646225, "grad_norm": 2.1142016873832326, "learning_rate": 4.20371230098787e-06, "loss": 0.3794, "step": 7961 }, { "epoch": 0.5644806806097129, "grad_norm": 2.4538409454886194, "learning_rate": 4.202578816863892e-06, "loss": 0.415, "step": 7962 }, { "epoch": 0.5645515774548032, "grad_norm": 3.6524724471544925, "learning_rate": 4.20144537478854e-06, "loss": 0.3877, "step": 7963 }, { "epoch": 0.5646224742998937, "grad_norm": 3.2672358900491942, "learning_rate": 4.200311974821584e-06, "loss": 0.3903, "step": 7964 }, { "epoch": 0.5646933711449841, "grad_norm": 3.0162344881362464, "learning_rate": 4.199178617022788e-06, "loss": 0.413, "step": 7965 }, { "epoch": 0.5647642679900744, "grad_norm": 9.137829503065808, "learning_rate": 4.198045301451914e-06, "loss": 0.3721, "step": 7966 }, { "epoch": 0.5648351648351648, "grad_norm": 5.37967447806997, "learning_rate": 4.196912028168723e-06, "loss": 0.3195, "step": 7967 }, { "epoch": 0.5649060616802553, "grad_norm": 4.57175707145111, "learning_rate": 4.195778797232975e-06, "loss": 0.4248, "step": 7968 }, { "epoch": 0.5649769585253456, "grad_norm": 3.172522900107053, "learning_rate": 4.194645608704424e-06, "loss": 0.4224, "step": 7969 }, { "epoch": 0.565047855370436, "grad_norm": 2.862215865575891, "learning_rate": 4.193512462642825e-06, "loss": 0.3967, "step": 7970 }, { "epoch": 0.5651187522155264, "grad_norm": 4.365639593307073, "learning_rate": 4.192379359107928e-06, "loss": 0.4448, "step": 7971 }, { "epoch": 0.5651896490606168, "grad_norm": 2.0599978195386224, "learning_rate": 4.191246298159487e-06, "loss": 0.3619, "step": 7972 }, { "epoch": 0.5652605459057072, "grad_norm": 3.3894954998891946, "learning_rate": 4.190113279857243e-06, "loss": 0.3951, "step": 7973 }, { "epoch": 0.5653314427507976, "grad_norm": 1.9915512071993298, "learning_rate": 4.188980304260945e-06, "loss": 0.4305, "step": 7974 }, { "epoch": 0.565402339595888, "grad_norm": 4.060017258787415, "learning_rate": 4.187847371430336e-06, "loss": 0.4203, "step": 7975 }, { "epoch": 0.5654732364409784, "grad_norm": 1.8019850682112346, "learning_rate": 4.186714481425153e-06, "loss": 0.3988, "step": 7976 }, { "epoch": 0.5655441332860688, "grad_norm": 4.0302934568181925, "learning_rate": 4.185581634305137e-06, "loss": 0.4652, "step": 7977 }, { "epoch": 0.5656150301311592, "grad_norm": 4.285707309946061, "learning_rate": 4.184448830130023e-06, "loss": 0.3629, "step": 7978 }, { "epoch": 0.5656859269762495, "grad_norm": 5.335081788717629, "learning_rate": 4.183316068959544e-06, "loss": 0.4135, "step": 7979 }, { "epoch": 0.56575682382134, "grad_norm": 1.125396887037215, "learning_rate": 4.182183350853433e-06, "loss": 0.7004, "step": 7980 }, { "epoch": 0.5658277206664304, "grad_norm": 2.5180446475187543, "learning_rate": 4.181050675871418e-06, "loss": 0.3947, "step": 7981 }, { "epoch": 0.5658986175115207, "grad_norm": 1.032809471304899, "learning_rate": 4.179918044073225e-06, "loss": 0.5878, "step": 7982 }, { "epoch": 0.5659695143566111, "grad_norm": 6.009512456225618, "learning_rate": 4.178785455518581e-06, "loss": 0.3921, "step": 7983 }, { "epoch": 0.5660404112017016, "grad_norm": 3.54622604032574, "learning_rate": 4.177652910267206e-06, "loss": 0.3651, "step": 7984 }, { "epoch": 0.5661113080467919, "grad_norm": 4.296577212743226, "learning_rate": 4.17652040837882e-06, "loss": 0.4639, "step": 7985 }, { "epoch": 0.5661822048918823, "grad_norm": 2.2372698310435135, "learning_rate": 4.175387949913141e-06, "loss": 0.3857, "step": 7986 }, { "epoch": 0.5662531017369727, "grad_norm": 2.2327578024530177, "learning_rate": 4.174255534929886e-06, "loss": 0.4124, "step": 7987 }, { "epoch": 0.5663239985820631, "grad_norm": 2.3087326181108754, "learning_rate": 4.173123163488768e-06, "loss": 0.44, "step": 7988 }, { "epoch": 0.5663948954271535, "grad_norm": 1.5875653416042126, "learning_rate": 4.171990835649495e-06, "loss": 0.4523, "step": 7989 }, { "epoch": 0.5664657922722439, "grad_norm": 4.363833874305166, "learning_rate": 4.170858551471777e-06, "loss": 0.4099, "step": 7990 }, { "epoch": 0.5665366891173342, "grad_norm": 5.767787200328498, "learning_rate": 4.1697263110153215e-06, "loss": 0.3959, "step": 7991 }, { "epoch": 0.5666075859624247, "grad_norm": 2.2469820343503435, "learning_rate": 4.16859411433983e-06, "loss": 0.3973, "step": 7992 }, { "epoch": 0.5666784828075151, "grad_norm": 13.624363102005903, "learning_rate": 4.167461961505006e-06, "loss": 0.4201, "step": 7993 }, { "epoch": 0.5667493796526054, "grad_norm": 3.7692978629748533, "learning_rate": 4.16632985257055e-06, "loss": 0.4259, "step": 7994 }, { "epoch": 0.5668202764976958, "grad_norm": 4.149759198873255, "learning_rate": 4.165197787596155e-06, "loss": 0.3917, "step": 7995 }, { "epoch": 0.5668911733427863, "grad_norm": 4.354842549434478, "learning_rate": 4.164065766641518e-06, "loss": 0.4504, "step": 7996 }, { "epoch": 0.5669620701878766, "grad_norm": 3.0368406999088897, "learning_rate": 4.162933789766333e-06, "loss": 0.4023, "step": 7997 }, { "epoch": 0.567032967032967, "grad_norm": 2.1264871046562126, "learning_rate": 4.161801857030287e-06, "loss": 0.3868, "step": 7998 }, { "epoch": 0.5671038638780574, "grad_norm": 2.2392058173869107, "learning_rate": 4.160669968493067e-06, "loss": 0.3946, "step": 7999 }, { "epoch": 0.5671747607231479, "grad_norm": 3.327074352450163, "learning_rate": 4.1595381242143616e-06, "loss": 0.4122, "step": 8000 }, { "epoch": 0.5672456575682382, "grad_norm": 2.7110766089822613, "learning_rate": 4.1584063242538544e-06, "loss": 0.391, "step": 8001 }, { "epoch": 0.5673165544133286, "grad_norm": 2.1851499941802506, "learning_rate": 4.157274568671222e-06, "loss": 0.3927, "step": 8002 }, { "epoch": 0.567387451258419, "grad_norm": 3.0829198528032435, "learning_rate": 4.1561428575261445e-06, "loss": 0.4359, "step": 8003 }, { "epoch": 0.5674583481035094, "grad_norm": 4.2662120833128645, "learning_rate": 4.1550111908783005e-06, "loss": 0.34, "step": 8004 }, { "epoch": 0.5675292449485998, "grad_norm": 5.902258103338128, "learning_rate": 4.153879568787358e-06, "loss": 0.3857, "step": 8005 }, { "epoch": 0.5676001417936902, "grad_norm": 3.296920548235937, "learning_rate": 4.152747991312993e-06, "loss": 0.4211, "step": 8006 }, { "epoch": 0.5676710386387805, "grad_norm": 3.3630225416640944, "learning_rate": 4.151616458514875e-06, "loss": 0.4111, "step": 8007 }, { "epoch": 0.567741935483871, "grad_norm": 3.2790200866709664, "learning_rate": 4.150484970452667e-06, "loss": 0.4049, "step": 8008 }, { "epoch": 0.5678128323289614, "grad_norm": 6.25123939097628, "learning_rate": 4.149353527186036e-06, "loss": 0.4146, "step": 8009 }, { "epoch": 0.5678837291740517, "grad_norm": 3.9727024862493576, "learning_rate": 4.148222128774642e-06, "loss": 0.3816, "step": 8010 }, { "epoch": 0.5679546260191422, "grad_norm": 3.8906167537533722, "learning_rate": 4.147090775278145e-06, "loss": 0.4072, "step": 8011 }, { "epoch": 0.5680255228642326, "grad_norm": 3.0208817726972024, "learning_rate": 4.145959466756201e-06, "loss": 0.4191, "step": 8012 }, { "epoch": 0.5680964197093229, "grad_norm": 2.653306018530406, "learning_rate": 4.144828203268469e-06, "loss": 0.3998, "step": 8013 }, { "epoch": 0.5681673165544133, "grad_norm": 6.752039527759998, "learning_rate": 4.143696984874597e-06, "loss": 0.3863, "step": 8014 }, { "epoch": 0.5682382133995038, "grad_norm": 2.866126209545088, "learning_rate": 4.142565811634237e-06, "loss": 0.4127, "step": 8015 }, { "epoch": 0.5683091102445941, "grad_norm": 9.049033928012364, "learning_rate": 4.141434683607035e-06, "loss": 0.4188, "step": 8016 }, { "epoch": 0.5683800070896845, "grad_norm": 2.438374395754783, "learning_rate": 4.140303600852641e-06, "loss": 0.4072, "step": 8017 }, { "epoch": 0.5684509039347749, "grad_norm": 4.302520671849468, "learning_rate": 4.13917256343069e-06, "loss": 0.3942, "step": 8018 }, { "epoch": 0.5685218007798652, "grad_norm": 3.796134440385531, "learning_rate": 4.138041571400829e-06, "loss": 0.3879, "step": 8019 }, { "epoch": 0.5685926976249557, "grad_norm": 2.173333639501705, "learning_rate": 4.136910624822694e-06, "loss": 0.4795, "step": 8020 }, { "epoch": 0.5686635944700461, "grad_norm": 2.95673821606098, "learning_rate": 4.13577972375592e-06, "loss": 0.4208, "step": 8021 }, { "epoch": 0.5687344913151364, "grad_norm": 5.29573127346897, "learning_rate": 4.134648868260141e-06, "loss": 0.3861, "step": 8022 }, { "epoch": 0.5688053881602269, "grad_norm": 3.0954025415425703, "learning_rate": 4.133518058394988e-06, "loss": 0.4326, "step": 8023 }, { "epoch": 0.5688762850053173, "grad_norm": 3.7541157502400755, "learning_rate": 4.132387294220087e-06, "loss": 0.4629, "step": 8024 }, { "epoch": 0.5689471818504076, "grad_norm": 2.794919528732113, "learning_rate": 4.131256575795068e-06, "loss": 0.3767, "step": 8025 }, { "epoch": 0.569018078695498, "grad_norm": 4.89243896684742, "learning_rate": 4.130125903179553e-06, "loss": 0.4514, "step": 8026 }, { "epoch": 0.5690889755405885, "grad_norm": 5.282939321708824, "learning_rate": 4.128995276433163e-06, "loss": 0.4198, "step": 8027 }, { "epoch": 0.5691598723856789, "grad_norm": 2.2375990160198134, "learning_rate": 4.127864695615516e-06, "loss": 0.4209, "step": 8028 }, { "epoch": 0.5692307692307692, "grad_norm": 2.1399493860214, "learning_rate": 4.126734160786229e-06, "loss": 0.3932, "step": 8029 }, { "epoch": 0.5693016660758596, "grad_norm": 2.6927303594454792, "learning_rate": 4.125603672004919e-06, "loss": 0.3696, "step": 8030 }, { "epoch": 0.5693725629209501, "grad_norm": 12.710000481672234, "learning_rate": 4.124473229331192e-06, "loss": 0.3965, "step": 8031 }, { "epoch": 0.5694434597660404, "grad_norm": 2.913376172893076, "learning_rate": 4.123342832824661e-06, "loss": 0.411, "step": 8032 }, { "epoch": 0.5695143566111308, "grad_norm": 4.043109592624872, "learning_rate": 4.122212482544931e-06, "loss": 0.4553, "step": 8033 }, { "epoch": 0.5695852534562212, "grad_norm": 2.1411438185514813, "learning_rate": 4.121082178551608e-06, "loss": 0.4178, "step": 8034 }, { "epoch": 0.5696561503013116, "grad_norm": 2.856674981210477, "learning_rate": 4.11995192090429e-06, "loss": 0.3987, "step": 8035 }, { "epoch": 0.569727047146402, "grad_norm": 0.9602623785511659, "learning_rate": 4.118821709662582e-06, "loss": 0.6623, "step": 8036 }, { "epoch": 0.5697979439914924, "grad_norm": 3.0729241471555064, "learning_rate": 4.1176915448860755e-06, "loss": 0.4058, "step": 8037 }, { "epoch": 0.5698688408365827, "grad_norm": 3.541715684160882, "learning_rate": 4.116561426634368e-06, "loss": 0.4007, "step": 8038 }, { "epoch": 0.5699397376816732, "grad_norm": 2.7167427054512436, "learning_rate": 4.115431354967052e-06, "loss": 0.4028, "step": 8039 }, { "epoch": 0.5700106345267636, "grad_norm": 5.879016299160996, "learning_rate": 4.114301329943715e-06, "loss": 0.3986, "step": 8040 }, { "epoch": 0.5700815313718539, "grad_norm": 3.5820631647816423, "learning_rate": 4.113171351623944e-06, "loss": 0.4505, "step": 8041 }, { "epoch": 0.5701524282169443, "grad_norm": 7.504415495000673, "learning_rate": 4.112041420067326e-06, "loss": 0.4098, "step": 8042 }, { "epoch": 0.5702233250620348, "grad_norm": 6.808998731488472, "learning_rate": 4.11091153533344e-06, "loss": 0.4618, "step": 8043 }, { "epoch": 0.5702942219071251, "grad_norm": 2.581023661838305, "learning_rate": 4.109781697481868e-06, "loss": 0.4041, "step": 8044 }, { "epoch": 0.5703651187522155, "grad_norm": 3.5329956937478992, "learning_rate": 4.108651906572186e-06, "loss": 0.4025, "step": 8045 }, { "epoch": 0.570436015597306, "grad_norm": 3.000504826050475, "learning_rate": 4.107522162663971e-06, "loss": 0.4176, "step": 8046 }, { "epoch": 0.5705069124423963, "grad_norm": 4.117326046532102, "learning_rate": 4.1063924658167925e-06, "loss": 0.419, "step": 8047 }, { "epoch": 0.5705778092874867, "grad_norm": 2.267347645422314, "learning_rate": 4.10526281609022e-06, "loss": 0.3686, "step": 8048 }, { "epoch": 0.5706487061325771, "grad_norm": 2.2013477836098723, "learning_rate": 4.104133213543826e-06, "loss": 0.4264, "step": 8049 }, { "epoch": 0.5707196029776674, "grad_norm": 2.186262461154148, "learning_rate": 4.103003658237168e-06, "loss": 0.3582, "step": 8050 }, { "epoch": 0.5707904998227579, "grad_norm": 2.5389183145221668, "learning_rate": 4.101874150229812e-06, "loss": 0.395, "step": 8051 }, { "epoch": 0.5708613966678483, "grad_norm": 3.3700700127170506, "learning_rate": 4.10074468958132e-06, "loss": 0.4342, "step": 8052 }, { "epoch": 0.5709322935129387, "grad_norm": 3.09810206197035, "learning_rate": 4.099615276351244e-06, "loss": 0.4337, "step": 8053 }, { "epoch": 0.571003190358029, "grad_norm": 2.3271013439721955, "learning_rate": 4.098485910599143e-06, "loss": 0.4137, "step": 8054 }, { "epoch": 0.5710740872031195, "grad_norm": 2.6907979043198127, "learning_rate": 4.097356592384569e-06, "loss": 0.4172, "step": 8055 }, { "epoch": 0.5711449840482099, "grad_norm": 3.1699113873104348, "learning_rate": 4.096227321767069e-06, "loss": 0.4122, "step": 8056 }, { "epoch": 0.5712158808933002, "grad_norm": 4.095120845191336, "learning_rate": 4.095098098806192e-06, "loss": 0.3973, "step": 8057 }, { "epoch": 0.5712867777383906, "grad_norm": 3.6282773986494106, "learning_rate": 4.0939689235614856e-06, "loss": 0.4482, "step": 8058 }, { "epoch": 0.5713576745834811, "grad_norm": 3.3034130879889587, "learning_rate": 4.092839796092487e-06, "loss": 0.4457, "step": 8059 }, { "epoch": 0.5714285714285714, "grad_norm": 4.776386398479351, "learning_rate": 4.091710716458739e-06, "loss": 0.4053, "step": 8060 }, { "epoch": 0.5714994682736618, "grad_norm": 3.3756400145440257, "learning_rate": 4.090581684719777e-06, "loss": 0.3931, "step": 8061 }, { "epoch": 0.5715703651187523, "grad_norm": 2.3260345865744125, "learning_rate": 4.08945270093514e-06, "loss": 0.421, "step": 8062 }, { "epoch": 0.5716412619638426, "grad_norm": 2.4432921273805683, "learning_rate": 4.088323765164354e-06, "loss": 0.4036, "step": 8063 }, { "epoch": 0.571712158808933, "grad_norm": 3.859986209312135, "learning_rate": 4.087194877466953e-06, "loss": 0.4535, "step": 8064 }, { "epoch": 0.5717830556540234, "grad_norm": 2.664048666750494, "learning_rate": 4.086066037902463e-06, "loss": 0.4086, "step": 8065 }, { "epoch": 0.5718539524991137, "grad_norm": 2.7131419336594105, "learning_rate": 4.084937246530408e-06, "loss": 0.383, "step": 8066 }, { "epoch": 0.5719248493442042, "grad_norm": 4.283209702868319, "learning_rate": 4.08380850341031e-06, "loss": 0.3926, "step": 8067 }, { "epoch": 0.5719957461892946, "grad_norm": 3.5905104250478317, "learning_rate": 4.082679808601691e-06, "loss": 0.3741, "step": 8068 }, { "epoch": 0.5720666430343849, "grad_norm": 2.10503150388175, "learning_rate": 4.081551162164063e-06, "loss": 0.4457, "step": 8069 }, { "epoch": 0.5721375398794754, "grad_norm": 6.643611334008778, "learning_rate": 4.080422564156944e-06, "loss": 0.4239, "step": 8070 }, { "epoch": 0.5722084367245658, "grad_norm": 2.6202175336503704, "learning_rate": 4.079294014639847e-06, "loss": 0.4499, "step": 8071 }, { "epoch": 0.5722793335696561, "grad_norm": 2.941602725878959, "learning_rate": 4.078165513672277e-06, "loss": 0.4087, "step": 8072 }, { "epoch": 0.5723502304147465, "grad_norm": 2.4030678021399043, "learning_rate": 4.077037061313744e-06, "loss": 0.3585, "step": 8073 }, { "epoch": 0.572421127259837, "grad_norm": 2.8129883230824797, "learning_rate": 4.075908657623751e-06, "loss": 0.4165, "step": 8074 }, { "epoch": 0.5724920241049273, "grad_norm": 2.862399390635587, "learning_rate": 4.0747803026618015e-06, "loss": 0.3981, "step": 8075 }, { "epoch": 0.5725629209500177, "grad_norm": 10.893012439513756, "learning_rate": 4.07365199648739e-06, "loss": 0.4202, "step": 8076 }, { "epoch": 0.5726338177951081, "grad_norm": 10.355775947280781, "learning_rate": 4.072523739160017e-06, "loss": 0.3918, "step": 8077 }, { "epoch": 0.5727047146401986, "grad_norm": 3.3472775457862904, "learning_rate": 4.071395530739176e-06, "loss": 0.3517, "step": 8078 }, { "epoch": 0.5727756114852889, "grad_norm": 2.0177318427820015, "learning_rate": 4.070267371284357e-06, "loss": 0.3921, "step": 8079 }, { "epoch": 0.5728465083303793, "grad_norm": 3.4133720955738127, "learning_rate": 4.069139260855047e-06, "loss": 0.38, "step": 8080 }, { "epoch": 0.5729174051754697, "grad_norm": 7.525854935990821, "learning_rate": 4.068011199510738e-06, "loss": 0.4156, "step": 8081 }, { "epoch": 0.5729883020205601, "grad_norm": 4.15180130148602, "learning_rate": 4.066883187310908e-06, "loss": 0.4399, "step": 8082 }, { "epoch": 0.5730591988656505, "grad_norm": 4.68005903053647, "learning_rate": 4.065755224315039e-06, "loss": 0.4034, "step": 8083 }, { "epoch": 0.5731300957107409, "grad_norm": 3.7599656418091034, "learning_rate": 4.0646273105826115e-06, "loss": 0.437, "step": 8084 }, { "epoch": 0.5732009925558312, "grad_norm": 4.312638434994536, "learning_rate": 4.063499446173099e-06, "loss": 0.4324, "step": 8085 }, { "epoch": 0.5732718894009217, "grad_norm": 2.9560307991957586, "learning_rate": 4.062371631145976e-06, "loss": 0.3969, "step": 8086 }, { "epoch": 0.5733427862460121, "grad_norm": 1.8903351786559766, "learning_rate": 4.0612438655607125e-06, "loss": 0.4386, "step": 8087 }, { "epoch": 0.5734136830911024, "grad_norm": 4.2361622227252145, "learning_rate": 4.060116149476776e-06, "loss": 0.3999, "step": 8088 }, { "epoch": 0.5734845799361928, "grad_norm": 9.6958246232787, "learning_rate": 4.058988482953632e-06, "loss": 0.3729, "step": 8089 }, { "epoch": 0.5735554767812833, "grad_norm": 2.854335323166369, "learning_rate": 4.057860866050743e-06, "loss": 0.4055, "step": 8090 }, { "epoch": 0.5736263736263736, "grad_norm": 4.75379982381243, "learning_rate": 4.056733298827572e-06, "loss": 0.4232, "step": 8091 }, { "epoch": 0.573697270471464, "grad_norm": 8.762821146025452, "learning_rate": 4.055605781343573e-06, "loss": 0.4228, "step": 8092 }, { "epoch": 0.5737681673165544, "grad_norm": 2.7410556175406016, "learning_rate": 4.0544783136582e-06, "loss": 0.3481, "step": 8093 }, { "epoch": 0.5738390641616448, "grad_norm": 3.4037382059089154, "learning_rate": 4.053350895830912e-06, "loss": 0.3873, "step": 8094 }, { "epoch": 0.5739099610067352, "grad_norm": 3.8408813541821907, "learning_rate": 4.052223527921149e-06, "loss": 0.3833, "step": 8095 }, { "epoch": 0.5739808578518256, "grad_norm": 3.2263612208688928, "learning_rate": 4.051096209988365e-06, "loss": 0.3652, "step": 8096 }, { "epoch": 0.5740517546969159, "grad_norm": 4.12694576059227, "learning_rate": 4.049968942092002e-06, "loss": 0.4003, "step": 8097 }, { "epoch": 0.5741226515420064, "grad_norm": 4.072058862765123, "learning_rate": 4.0488417242915016e-06, "loss": 0.4481, "step": 8098 }, { "epoch": 0.5741935483870968, "grad_norm": 7.401433954322066, "learning_rate": 4.047714556646302e-06, "loss": 0.493, "step": 8099 }, { "epoch": 0.5742644452321871, "grad_norm": 3.0437509843518407, "learning_rate": 4.046587439215843e-06, "loss": 0.3816, "step": 8100 }, { "epoch": 0.5743353420772775, "grad_norm": 4.748242175877497, "learning_rate": 4.0454603720595535e-06, "loss": 0.3743, "step": 8101 }, { "epoch": 0.574406238922368, "grad_norm": 2.2155200788563225, "learning_rate": 4.044333355236868e-06, "loss": 0.4005, "step": 8102 }, { "epoch": 0.5744771357674584, "grad_norm": 2.7375795244027437, "learning_rate": 4.043206388807214e-06, "loss": 0.41, "step": 8103 }, { "epoch": 0.5745480326125487, "grad_norm": 7.097928826023674, "learning_rate": 4.0420794728300185e-06, "loss": 0.3926, "step": 8104 }, { "epoch": 0.5746189294576391, "grad_norm": 2.4074591364621534, "learning_rate": 4.040952607364703e-06, "loss": 0.431, "step": 8105 }, { "epoch": 0.5746898263027296, "grad_norm": 3.9524906993790245, "learning_rate": 4.039825792470687e-06, "loss": 0.4008, "step": 8106 }, { "epoch": 0.5747607231478199, "grad_norm": 2.4812179360895947, "learning_rate": 4.038699028207394e-06, "loss": 0.3743, "step": 8107 }, { "epoch": 0.5748316199929103, "grad_norm": 6.740377079327435, "learning_rate": 4.0375723146342306e-06, "loss": 0.4059, "step": 8108 }, { "epoch": 0.5749025168380008, "grad_norm": 2.7092322563806395, "learning_rate": 4.0364456518106145e-06, "loss": 0.3768, "step": 8109 }, { "epoch": 0.5749734136830911, "grad_norm": 3.7108945713420307, "learning_rate": 4.035319039795956e-06, "loss": 0.3565, "step": 8110 }, { "epoch": 0.5750443105281815, "grad_norm": 2.9300490073548637, "learning_rate": 4.03419247864966e-06, "loss": 0.3714, "step": 8111 }, { "epoch": 0.5751152073732719, "grad_norm": 4.587483879970613, "learning_rate": 4.033065968431131e-06, "loss": 0.3811, "step": 8112 }, { "epoch": 0.5751861042183622, "grad_norm": 2.4240030416568112, "learning_rate": 4.031939509199775e-06, "loss": 0.3599, "step": 8113 }, { "epoch": 0.5752570010634527, "grad_norm": 6.49113756014435, "learning_rate": 4.030813101014984e-06, "loss": 0.4004, "step": 8114 }, { "epoch": 0.5753278979085431, "grad_norm": 4.2237790240829804, "learning_rate": 4.029686743936159e-06, "loss": 0.4331, "step": 8115 }, { "epoch": 0.5753987947536334, "grad_norm": 3.3074499736700735, "learning_rate": 4.0285604380226926e-06, "loss": 0.3644, "step": 8116 }, { "epoch": 0.5754696915987239, "grad_norm": 7.95652757446387, "learning_rate": 4.0274341833339755e-06, "loss": 0.3578, "step": 8117 }, { "epoch": 0.5755405884438143, "grad_norm": 2.026964948084903, "learning_rate": 4.026307979929397e-06, "loss": 0.385, "step": 8118 }, { "epoch": 0.5756114852889046, "grad_norm": 2.6650853989060024, "learning_rate": 4.0251818278683395e-06, "loss": 0.4303, "step": 8119 }, { "epoch": 0.575682382133995, "grad_norm": 3.7916675473465085, "learning_rate": 4.024055727210193e-06, "loss": 0.4018, "step": 8120 }, { "epoch": 0.5757532789790855, "grad_norm": 6.581802102695931, "learning_rate": 4.022929678014329e-06, "loss": 0.3592, "step": 8121 }, { "epoch": 0.5758241758241758, "grad_norm": 3.5275146735300806, "learning_rate": 4.021803680340129e-06, "loss": 0.4102, "step": 8122 }, { "epoch": 0.5758950726692662, "grad_norm": 3.3134656292762017, "learning_rate": 4.02067773424697e-06, "loss": 0.4228, "step": 8123 }, { "epoch": 0.5759659695143566, "grad_norm": 8.810779956278619, "learning_rate": 4.019551839794219e-06, "loss": 0.4433, "step": 8124 }, { "epoch": 0.576036866359447, "grad_norm": 4.585731424758444, "learning_rate": 4.018425997041248e-06, "loss": 0.4036, "step": 8125 }, { "epoch": 0.5761077632045374, "grad_norm": 4.134753789154607, "learning_rate": 4.017300206047426e-06, "loss": 0.4233, "step": 8126 }, { "epoch": 0.5761786600496278, "grad_norm": 2.4467138984717023, "learning_rate": 4.016174466872111e-06, "loss": 0.3453, "step": 8127 }, { "epoch": 0.5762495568947182, "grad_norm": 1.90332811485734, "learning_rate": 4.015048779574668e-06, "loss": 0.3862, "step": 8128 }, { "epoch": 0.5763204537398086, "grad_norm": 3.448209955161496, "learning_rate": 4.013923144214457e-06, "loss": 0.4105, "step": 8129 }, { "epoch": 0.576391350584899, "grad_norm": 2.108065942815853, "learning_rate": 4.01279756085083e-06, "loss": 0.3913, "step": 8130 }, { "epoch": 0.5764622474299894, "grad_norm": 3.3945916554401894, "learning_rate": 4.01167202954314e-06, "loss": 0.4045, "step": 8131 }, { "epoch": 0.5765331442750797, "grad_norm": 3.4037897253038305, "learning_rate": 4.010546550350737e-06, "loss": 0.3916, "step": 8132 }, { "epoch": 0.5766040411201702, "grad_norm": 2.956628837751207, "learning_rate": 4.009421123332973e-06, "loss": 0.3952, "step": 8133 }, { "epoch": 0.5766749379652606, "grad_norm": 6.67671172960082, "learning_rate": 4.008295748549188e-06, "loss": 0.4096, "step": 8134 }, { "epoch": 0.5767458348103509, "grad_norm": 2.720371227341403, "learning_rate": 4.007170426058724e-06, "loss": 0.4236, "step": 8135 }, { "epoch": 0.5768167316554413, "grad_norm": 2.660572174136159, "learning_rate": 4.006045155920924e-06, "loss": 0.3728, "step": 8136 }, { "epoch": 0.5768876285005318, "grad_norm": 4.125384650709846, "learning_rate": 4.004919938195119e-06, "loss": 0.3873, "step": 8137 }, { "epoch": 0.5769585253456221, "grad_norm": 6.559205813201291, "learning_rate": 4.003794772940644e-06, "loss": 0.4145, "step": 8138 }, { "epoch": 0.5770294221907125, "grad_norm": 2.1620725844343895, "learning_rate": 4.002669660216836e-06, "loss": 0.3915, "step": 8139 }, { "epoch": 0.5771003190358029, "grad_norm": 4.3960158292853295, "learning_rate": 4.001544600083012e-06, "loss": 0.5007, "step": 8140 }, { "epoch": 0.5771712158808933, "grad_norm": 5.18944524085628, "learning_rate": 4.0004195925985065e-06, "loss": 0.4416, "step": 8141 }, { "epoch": 0.5772421127259837, "grad_norm": 3.6622480804247357, "learning_rate": 3.99929463782264e-06, "loss": 0.3678, "step": 8142 }, { "epoch": 0.5773130095710741, "grad_norm": 5.983464486370944, "learning_rate": 3.99816973581473e-06, "loss": 0.3752, "step": 8143 }, { "epoch": 0.5773839064161644, "grad_norm": 2.0413605362242313, "learning_rate": 3.997044886634092e-06, "loss": 0.412, "step": 8144 }, { "epoch": 0.5774548032612549, "grad_norm": 2.817734029820766, "learning_rate": 3.995920090340047e-06, "loss": 0.4202, "step": 8145 }, { "epoch": 0.5775257001063453, "grad_norm": 7.184121417643234, "learning_rate": 3.9947953469918986e-06, "loss": 0.4287, "step": 8146 }, { "epoch": 0.5775965969514356, "grad_norm": 2.2369275846548042, "learning_rate": 3.99367065664896e-06, "loss": 0.4435, "step": 8147 }, { "epoch": 0.577667493796526, "grad_norm": 3.781291038726825, "learning_rate": 3.992546019370535e-06, "loss": 0.4426, "step": 8148 }, { "epoch": 0.5777383906416165, "grad_norm": 2.922770914638083, "learning_rate": 3.991421435215929e-06, "loss": 0.4005, "step": 8149 }, { "epoch": 0.5778092874867068, "grad_norm": 3.323441744180459, "learning_rate": 3.99029690424444e-06, "loss": 0.4154, "step": 8150 }, { "epoch": 0.5778801843317972, "grad_norm": 3.435282112161302, "learning_rate": 3.989172426515364e-06, "loss": 0.4002, "step": 8151 }, { "epoch": 0.5779510811768876, "grad_norm": 2.5236674441206453, "learning_rate": 3.988048002088001e-06, "loss": 0.4175, "step": 8152 }, { "epoch": 0.578021978021978, "grad_norm": 88.98703454819734, "learning_rate": 3.9869236310216355e-06, "loss": 0.4045, "step": 8153 }, { "epoch": 0.5780928748670684, "grad_norm": 5.335994744980931, "learning_rate": 3.985799313375562e-06, "loss": 0.4377, "step": 8154 }, { "epoch": 0.5781637717121588, "grad_norm": 10.115958598204486, "learning_rate": 3.984675049209066e-06, "loss": 0.3724, "step": 8155 }, { "epoch": 0.5782346685572493, "grad_norm": 2.8448100670375394, "learning_rate": 3.983550838581428e-06, "loss": 0.4552, "step": 8156 }, { "epoch": 0.5783055654023396, "grad_norm": 2.4834506256646254, "learning_rate": 3.982426681551929e-06, "loss": 0.3397, "step": 8157 }, { "epoch": 0.57837646224743, "grad_norm": 5.217175108998783, "learning_rate": 3.9813025781798515e-06, "loss": 0.4492, "step": 8158 }, { "epoch": 0.5784473590925204, "grad_norm": 2.726748276718323, "learning_rate": 3.980178528524463e-06, "loss": 0.3966, "step": 8159 }, { "epoch": 0.5785182559376107, "grad_norm": 3.6980809756060564, "learning_rate": 3.9790545326450405e-06, "loss": 0.4072, "step": 8160 }, { "epoch": 0.5785891527827012, "grad_norm": 6.611721870825477, "learning_rate": 3.977930590600851e-06, "loss": 0.3722, "step": 8161 }, { "epoch": 0.5786600496277916, "grad_norm": 2.6859005800637665, "learning_rate": 3.976806702451163e-06, "loss": 0.3662, "step": 8162 }, { "epoch": 0.5787309464728819, "grad_norm": 3.5896798660226805, "learning_rate": 3.975682868255238e-06, "loss": 0.4548, "step": 8163 }, { "epoch": 0.5788018433179724, "grad_norm": 3.0639118982196987, "learning_rate": 3.974559088072336e-06, "loss": 0.3595, "step": 8164 }, { "epoch": 0.5788727401630628, "grad_norm": 4.384590119615808, "learning_rate": 3.973435361961718e-06, "loss": 0.489, "step": 8165 }, { "epoch": 0.5789436370081531, "grad_norm": 8.124112652910226, "learning_rate": 3.972311689982635e-06, "loss": 0.4372, "step": 8166 }, { "epoch": 0.5790145338532435, "grad_norm": 4.2078119640941445, "learning_rate": 3.971188072194342e-06, "loss": 0.4036, "step": 8167 }, { "epoch": 0.579085430698334, "grad_norm": 4.889160720165039, "learning_rate": 3.970064508656087e-06, "loss": 0.421, "step": 8168 }, { "epoch": 0.5791563275434243, "grad_norm": 3.697291721206034, "learning_rate": 3.968940999427117e-06, "loss": 0.3448, "step": 8169 }, { "epoch": 0.5792272243885147, "grad_norm": 2.5199614116935387, "learning_rate": 3.967817544566673e-06, "loss": 0.3677, "step": 8170 }, { "epoch": 0.5792981212336051, "grad_norm": 3.30993769728929, "learning_rate": 3.966694144134002e-06, "loss": 0.3702, "step": 8171 }, { "epoch": 0.5793690180786955, "grad_norm": 3.8339222896538994, "learning_rate": 3.965570798188333e-06, "loss": 0.4333, "step": 8172 }, { "epoch": 0.5794399149237859, "grad_norm": 5.3482361656810085, "learning_rate": 3.964447506788907e-06, "loss": 0.3891, "step": 8173 }, { "epoch": 0.5795108117688763, "grad_norm": 4.212563367962336, "learning_rate": 3.963324269994956e-06, "loss": 0.4434, "step": 8174 }, { "epoch": 0.5795817086139666, "grad_norm": 2.059195768501828, "learning_rate": 3.962201087865706e-06, "loss": 0.3947, "step": 8175 }, { "epoch": 0.5796526054590571, "grad_norm": 2.392988466527437, "learning_rate": 3.961077960460383e-06, "loss": 0.4565, "step": 8176 }, { "epoch": 0.5797235023041475, "grad_norm": 4.385673213456838, "learning_rate": 3.959954887838214e-06, "loss": 0.372, "step": 8177 }, { "epoch": 0.5797943991492378, "grad_norm": 4.069339919407216, "learning_rate": 3.958831870058418e-06, "loss": 0.424, "step": 8178 }, { "epoch": 0.5798652959943282, "grad_norm": 8.005279731787075, "learning_rate": 3.957708907180212e-06, "loss": 0.382, "step": 8179 }, { "epoch": 0.5799361928394187, "grad_norm": 2.3966038610103277, "learning_rate": 3.956585999262811e-06, "loss": 0.4391, "step": 8180 }, { "epoch": 0.5800070896845091, "grad_norm": 2.294527243151129, "learning_rate": 3.9554631463654265e-06, "loss": 0.3713, "step": 8181 }, { "epoch": 0.5800779865295994, "grad_norm": 2.309475481026726, "learning_rate": 3.954340348547267e-06, "loss": 0.3548, "step": 8182 }, { "epoch": 0.5801488833746898, "grad_norm": 8.644102049849657, "learning_rate": 3.953217605867538e-06, "loss": 0.3964, "step": 8183 }, { "epoch": 0.5802197802197803, "grad_norm": 2.7429672190715024, "learning_rate": 3.952094918385447e-06, "loss": 0.4087, "step": 8184 }, { "epoch": 0.5802906770648706, "grad_norm": 3.4012145134687897, "learning_rate": 3.950972286160187e-06, "loss": 0.3432, "step": 8185 }, { "epoch": 0.580361573909961, "grad_norm": 10.040079830550932, "learning_rate": 3.949849709250961e-06, "loss": 0.422, "step": 8186 }, { "epoch": 0.5804324707550514, "grad_norm": 2.526487825081847, "learning_rate": 3.948727187716962e-06, "loss": 0.3959, "step": 8187 }, { "epoch": 0.5805033676001418, "grad_norm": 2.633147629922923, "learning_rate": 3.947604721617379e-06, "loss": 0.4165, "step": 8188 }, { "epoch": 0.5805742644452322, "grad_norm": 2.748939932544293, "learning_rate": 3.946482311011403e-06, "loss": 0.3954, "step": 8189 }, { "epoch": 0.5806451612903226, "grad_norm": 1.807584689957945, "learning_rate": 3.9453599559582205e-06, "loss": 0.4267, "step": 8190 }, { "epoch": 0.5807160581354129, "grad_norm": 3.853665477436209, "learning_rate": 3.94423765651701e-06, "loss": 0.3868, "step": 8191 }, { "epoch": 0.5807869549805034, "grad_norm": 2.924331932131493, "learning_rate": 3.943115412746954e-06, "loss": 0.4273, "step": 8192 }, { "epoch": 0.5808578518255938, "grad_norm": 4.230983555362442, "learning_rate": 3.94199322470723e-06, "loss": 0.3607, "step": 8193 }, { "epoch": 0.5809287486706841, "grad_norm": 5.139535860016358, "learning_rate": 3.940871092457011e-06, "loss": 0.431, "step": 8194 }, { "epoch": 0.5809996455157745, "grad_norm": 2.419764477726964, "learning_rate": 3.939749016055468e-06, "loss": 0.419, "step": 8195 }, { "epoch": 0.581070542360865, "grad_norm": 5.647549351307392, "learning_rate": 3.938626995561766e-06, "loss": 0.3963, "step": 8196 }, { "epoch": 0.5811414392059553, "grad_norm": 3.9293182998387155, "learning_rate": 3.937505031035076e-06, "loss": 0.4093, "step": 8197 }, { "epoch": 0.5812123360510457, "grad_norm": 5.6262781186460264, "learning_rate": 3.9363831225345545e-06, "loss": 0.399, "step": 8198 }, { "epoch": 0.5812832328961361, "grad_norm": 5.457915384377983, "learning_rate": 3.935261270119364e-06, "loss": 0.4219, "step": 8199 }, { "epoch": 0.5813541297412265, "grad_norm": 3.1750844592779393, "learning_rate": 3.934139473848659e-06, "loss": 0.4043, "step": 8200 }, { "epoch": 0.5814250265863169, "grad_norm": 3.6831229502327973, "learning_rate": 3.933017733781593e-06, "loss": 0.4015, "step": 8201 }, { "epoch": 0.5814959234314073, "grad_norm": 3.982153430319152, "learning_rate": 3.931896049977314e-06, "loss": 0.4025, "step": 8202 }, { "epoch": 0.5815668202764976, "grad_norm": 9.655636444094517, "learning_rate": 3.930774422494976e-06, "loss": 0.4047, "step": 8203 }, { "epoch": 0.5816377171215881, "grad_norm": 4.173305981925287, "learning_rate": 3.929652851393715e-06, "loss": 0.3521, "step": 8204 }, { "epoch": 0.5817086139666785, "grad_norm": 8.30576819441716, "learning_rate": 3.928531336732678e-06, "loss": 0.4665, "step": 8205 }, { "epoch": 0.5817795108117689, "grad_norm": 5.803132951202807, "learning_rate": 3.9274098785710005e-06, "loss": 0.4025, "step": 8206 }, { "epoch": 0.5818504076568592, "grad_norm": 3.2665927120140683, "learning_rate": 3.926288476967821e-06, "loss": 0.3826, "step": 8207 }, { "epoch": 0.5819213045019497, "grad_norm": 2.9587592762799724, "learning_rate": 3.925167131982267e-06, "loss": 0.3807, "step": 8208 }, { "epoch": 0.5819922013470401, "grad_norm": 3.9564237746032065, "learning_rate": 3.92404584367347e-06, "loss": 0.3858, "step": 8209 }, { "epoch": 0.5820630981921304, "grad_norm": 4.491617440639917, "learning_rate": 3.92292461210056e-06, "loss": 0.4413, "step": 8210 }, { "epoch": 0.5821339950372209, "grad_norm": 1.1281161280865668, "learning_rate": 3.921803437322655e-06, "loss": 0.6562, "step": 8211 }, { "epoch": 0.5822048918823113, "grad_norm": 2.8527233541312786, "learning_rate": 3.920682319398878e-06, "loss": 0.4333, "step": 8212 }, { "epoch": 0.5822757887274016, "grad_norm": 4.183739635177192, "learning_rate": 3.919561258388348e-06, "loss": 0.3929, "step": 8213 }, { "epoch": 0.582346685572492, "grad_norm": 6.335021711411532, "learning_rate": 3.918440254350176e-06, "loss": 0.4325, "step": 8214 }, { "epoch": 0.5824175824175825, "grad_norm": 7.265623202044664, "learning_rate": 3.917319307343473e-06, "loss": 0.3894, "step": 8215 }, { "epoch": 0.5824884792626728, "grad_norm": 4.285470647751079, "learning_rate": 3.916198417427354e-06, "loss": 0.3787, "step": 8216 }, { "epoch": 0.5825593761077632, "grad_norm": 3.9467126212516894, "learning_rate": 3.915077584660916e-06, "loss": 0.4277, "step": 8217 }, { "epoch": 0.5826302729528536, "grad_norm": 5.780971771198572, "learning_rate": 3.913956809103266e-06, "loss": 0.4548, "step": 8218 }, { "epoch": 0.582701169797944, "grad_norm": 13.220677707025219, "learning_rate": 3.912836090813502e-06, "loss": 0.409, "step": 8219 }, { "epoch": 0.5827720666430344, "grad_norm": 4.376259258900217, "learning_rate": 3.911715429850721e-06, "loss": 0.3865, "step": 8220 }, { "epoch": 0.5828429634881248, "grad_norm": 4.333585650495044, "learning_rate": 3.910594826274015e-06, "loss": 0.4339, "step": 8221 }, { "epoch": 0.5829138603332151, "grad_norm": 2.148061024594918, "learning_rate": 3.909474280142476e-06, "loss": 0.4472, "step": 8222 }, { "epoch": 0.5829847571783056, "grad_norm": 4.061372065598716, "learning_rate": 3.908353791515192e-06, "loss": 0.4213, "step": 8223 }, { "epoch": 0.583055654023396, "grad_norm": 5.432773519858466, "learning_rate": 3.9072333604512445e-06, "loss": 0.3656, "step": 8224 }, { "epoch": 0.5831265508684863, "grad_norm": 4.226074631913365, "learning_rate": 3.906112987009717e-06, "loss": 0.4231, "step": 8225 }, { "epoch": 0.5831974477135767, "grad_norm": 2.1609725131262367, "learning_rate": 3.904992671249686e-06, "loss": 0.3997, "step": 8226 }, { "epoch": 0.5832683445586672, "grad_norm": 3.03556732286595, "learning_rate": 3.903872413230227e-06, "loss": 0.4181, "step": 8227 }, { "epoch": 0.5833392414037575, "grad_norm": 5.165494924394864, "learning_rate": 3.902752213010411e-06, "loss": 0.4325, "step": 8228 }, { "epoch": 0.5834101382488479, "grad_norm": 3.4366245401200173, "learning_rate": 3.901632070649311e-06, "loss": 0.3933, "step": 8229 }, { "epoch": 0.5834810350939383, "grad_norm": 4.298952084426781, "learning_rate": 3.9005119862059885e-06, "loss": 0.4581, "step": 8230 }, { "epoch": 0.5835519319390288, "grad_norm": 2.178703458375035, "learning_rate": 3.899391959739508e-06, "loss": 0.4196, "step": 8231 }, { "epoch": 0.5836228287841191, "grad_norm": 1.9424612842358966, "learning_rate": 3.898271991308931e-06, "loss": 0.4457, "step": 8232 }, { "epoch": 0.5836937256292095, "grad_norm": 3.810908369923133, "learning_rate": 3.897152080973312e-06, "loss": 0.4053, "step": 8233 }, { "epoch": 0.5837646224742999, "grad_norm": 2.181776262221453, "learning_rate": 3.896032228791704e-06, "loss": 0.4052, "step": 8234 }, { "epoch": 0.5838355193193903, "grad_norm": 2.8099369797614653, "learning_rate": 3.894912434823161e-06, "loss": 0.3833, "step": 8235 }, { "epoch": 0.5839064161644807, "grad_norm": 9.326210714063418, "learning_rate": 3.893792699126729e-06, "loss": 0.3956, "step": 8236 }, { "epoch": 0.5839773130095711, "grad_norm": 3.468083143128059, "learning_rate": 3.892673021761452e-06, "loss": 0.3621, "step": 8237 }, { "epoch": 0.5840482098546614, "grad_norm": 22.998030641808104, "learning_rate": 3.891553402786371e-06, "loss": 0.3781, "step": 8238 }, { "epoch": 0.5841191066997519, "grad_norm": 2.543860829430976, "learning_rate": 3.890433842260527e-06, "loss": 0.4054, "step": 8239 }, { "epoch": 0.5841900035448423, "grad_norm": 3.365690005126428, "learning_rate": 3.889314340242951e-06, "loss": 0.4197, "step": 8240 }, { "epoch": 0.5842609003899326, "grad_norm": 2.9150107018840923, "learning_rate": 3.888194896792678e-06, "loss": 0.4405, "step": 8241 }, { "epoch": 0.584331797235023, "grad_norm": 3.108315855222717, "learning_rate": 3.887075511968737e-06, "loss": 0.392, "step": 8242 }, { "epoch": 0.5844026940801135, "grad_norm": 6.520234962428372, "learning_rate": 3.8859561858301545e-06, "loss": 0.4155, "step": 8243 }, { "epoch": 0.5844735909252038, "grad_norm": 4.144028072255337, "learning_rate": 3.884836918435952e-06, "loss": 0.4294, "step": 8244 }, { "epoch": 0.5845444877702942, "grad_norm": 11.398140633204283, "learning_rate": 3.88371770984515e-06, "loss": 0.4393, "step": 8245 }, { "epoch": 0.5846153846153846, "grad_norm": 5.736699075517698, "learning_rate": 3.882598560116766e-06, "loss": 0.3574, "step": 8246 }, { "epoch": 0.584686281460475, "grad_norm": 2.195797732120259, "learning_rate": 3.881479469309811e-06, "loss": 0.3972, "step": 8247 }, { "epoch": 0.5847571783055654, "grad_norm": 3.9082272251581576, "learning_rate": 3.880360437483301e-06, "loss": 0.3892, "step": 8248 }, { "epoch": 0.5848280751506558, "grad_norm": 4.731983330175642, "learning_rate": 3.879241464696235e-06, "loss": 0.4231, "step": 8249 }, { "epoch": 0.5848989719957461, "grad_norm": 3.3139795928217235, "learning_rate": 3.878122551007624e-06, "loss": 0.4124, "step": 8250 }, { "epoch": 0.5849698688408366, "grad_norm": 2.820124552927357, "learning_rate": 3.877003696476466e-06, "loss": 0.3953, "step": 8251 }, { "epoch": 0.585040765685927, "grad_norm": 3.3945989262274723, "learning_rate": 3.875884901161764e-06, "loss": 0.4357, "step": 8252 }, { "epoch": 0.5851116625310173, "grad_norm": 2.2440952690981386, "learning_rate": 3.874766165122504e-06, "loss": 0.4305, "step": 8253 }, { "epoch": 0.5851825593761077, "grad_norm": 3.93380891672032, "learning_rate": 3.873647488417686e-06, "loss": 0.4107, "step": 8254 }, { "epoch": 0.5852534562211982, "grad_norm": 4.355453081503113, "learning_rate": 3.872528871106297e-06, "loss": 0.3947, "step": 8255 }, { "epoch": 0.5853243530662886, "grad_norm": 4.7387750509497035, "learning_rate": 3.871410313247319e-06, "loss": 0.3977, "step": 8256 }, { "epoch": 0.5853952499113789, "grad_norm": 2.7922018918324434, "learning_rate": 3.870291814899736e-06, "loss": 0.3895, "step": 8257 }, { "epoch": 0.5854661467564694, "grad_norm": 2.9799555980824173, "learning_rate": 3.86917337612253e-06, "loss": 0.4282, "step": 8258 }, { "epoch": 0.5855370436015598, "grad_norm": 2.9191459899811787, "learning_rate": 3.868054996974674e-06, "loss": 0.3821, "step": 8259 }, { "epoch": 0.5856079404466501, "grad_norm": 3.878879846462103, "learning_rate": 3.866936677515141e-06, "loss": 0.3359, "step": 8260 }, { "epoch": 0.5856788372917405, "grad_norm": 2.414714186558498, "learning_rate": 3.865818417802904e-06, "loss": 0.4618, "step": 8261 }, { "epoch": 0.585749734136831, "grad_norm": 3.5493839509896064, "learning_rate": 3.864700217896925e-06, "loss": 0.3706, "step": 8262 }, { "epoch": 0.5858206309819213, "grad_norm": 2.6978109332073377, "learning_rate": 3.863582077856171e-06, "loss": 0.393, "step": 8263 }, { "epoch": 0.5858915278270117, "grad_norm": 1.9836542371695156, "learning_rate": 3.862463997739602e-06, "loss": 0.3805, "step": 8264 }, { "epoch": 0.5859624246721021, "grad_norm": 5.818099848755827, "learning_rate": 3.861345977606175e-06, "loss": 0.3507, "step": 8265 }, { "epoch": 0.5860333215171925, "grad_norm": 3.429980140338987, "learning_rate": 3.860228017514841e-06, "loss": 0.4151, "step": 8266 }, { "epoch": 0.5861042183622829, "grad_norm": 6.799689724919513, "learning_rate": 3.859110117524556e-06, "loss": 0.4414, "step": 8267 }, { "epoch": 0.5861751152073733, "grad_norm": 4.015006709277881, "learning_rate": 3.8579922776942655e-06, "loss": 0.4408, "step": 8268 }, { "epoch": 0.5862460120524636, "grad_norm": 3.7936072326159405, "learning_rate": 3.856874498082913e-06, "loss": 0.3429, "step": 8269 }, { "epoch": 0.5863169088975541, "grad_norm": 5.030303278401094, "learning_rate": 3.855756778749441e-06, "loss": 0.3886, "step": 8270 }, { "epoch": 0.5863878057426445, "grad_norm": 2.504372009401073, "learning_rate": 3.854639119752789e-06, "loss": 0.4244, "step": 8271 }, { "epoch": 0.5864587025877348, "grad_norm": 4.176050233568907, "learning_rate": 3.853521521151888e-06, "loss": 0.3898, "step": 8272 }, { "epoch": 0.5865295994328252, "grad_norm": 1.8895769307621293, "learning_rate": 3.852403983005672e-06, "loss": 0.3791, "step": 8273 }, { "epoch": 0.5866004962779157, "grad_norm": 3.1033283804291303, "learning_rate": 3.851286505373072e-06, "loss": 0.358, "step": 8274 }, { "epoch": 0.586671393123006, "grad_norm": 3.107137063381799, "learning_rate": 3.85016908831301e-06, "loss": 0.3869, "step": 8275 }, { "epoch": 0.5867422899680964, "grad_norm": 3.849730120761904, "learning_rate": 3.8490517318844104e-06, "loss": 0.3885, "step": 8276 }, { "epoch": 0.5868131868131868, "grad_norm": 2.2949703970509487, "learning_rate": 3.847934436146191e-06, "loss": 0.4436, "step": 8277 }, { "epoch": 0.5868840836582772, "grad_norm": 3.8275159891854185, "learning_rate": 3.846817201157268e-06, "loss": 0.3495, "step": 8278 }, { "epoch": 0.5869549805033676, "grad_norm": 4.160356219790741, "learning_rate": 3.845700026976553e-06, "loss": 0.355, "step": 8279 }, { "epoch": 0.587025877348458, "grad_norm": 3.290557161794031, "learning_rate": 3.844582913662957e-06, "loss": 0.387, "step": 8280 }, { "epoch": 0.5870967741935483, "grad_norm": 2.6490973017993986, "learning_rate": 3.8434658612753885e-06, "loss": 0.4387, "step": 8281 }, { "epoch": 0.5871676710386388, "grad_norm": 4.404356011498707, "learning_rate": 3.842348869872746e-06, "loss": 0.4293, "step": 8282 }, { "epoch": 0.5872385678837292, "grad_norm": 2.507007146949349, "learning_rate": 3.841231939513931e-06, "loss": 0.4065, "step": 8283 }, { "epoch": 0.5873094647288196, "grad_norm": 0.9703757525938674, "learning_rate": 3.840115070257842e-06, "loss": 0.6561, "step": 8284 }, { "epoch": 0.5873803615739099, "grad_norm": 3.4268231775977718, "learning_rate": 3.838998262163368e-06, "loss": 0.4943, "step": 8285 }, { "epoch": 0.5874512584190004, "grad_norm": 2.768978681986434, "learning_rate": 3.8378815152894025e-06, "loss": 0.4702, "step": 8286 }, { "epoch": 0.5875221552640908, "grad_norm": 14.05573447290265, "learning_rate": 3.836764829694833e-06, "loss": 0.4139, "step": 8287 }, { "epoch": 0.5875930521091811, "grad_norm": 4.698255678242245, "learning_rate": 3.835648205438541e-06, "loss": 0.3697, "step": 8288 }, { "epoch": 0.5876639489542715, "grad_norm": 4.991761533041898, "learning_rate": 3.8345316425794075e-06, "loss": 0.407, "step": 8289 }, { "epoch": 0.587734845799362, "grad_norm": 10.147761898701203, "learning_rate": 3.833415141176312e-06, "loss": 0.3792, "step": 8290 }, { "epoch": 0.5878057426444523, "grad_norm": 3.6057631773589525, "learning_rate": 3.832298701288124e-06, "loss": 0.3588, "step": 8291 }, { "epoch": 0.5878766394895427, "grad_norm": 6.519843523674866, "learning_rate": 3.831182322973717e-06, "loss": 0.3637, "step": 8292 }, { "epoch": 0.5879475363346331, "grad_norm": 4.304774632828232, "learning_rate": 3.830066006291961e-06, "loss": 0.4078, "step": 8293 }, { "epoch": 0.5880184331797235, "grad_norm": 2.9165643775799603, "learning_rate": 3.828949751301714e-06, "loss": 0.4394, "step": 8294 }, { "epoch": 0.5880893300248139, "grad_norm": 4.1203948933395695, "learning_rate": 3.827833558061842e-06, "loss": 0.3972, "step": 8295 }, { "epoch": 0.5881602268699043, "grad_norm": 3.517119854623109, "learning_rate": 3.8267174266312e-06, "loss": 0.4113, "step": 8296 }, { "epoch": 0.5882311237149946, "grad_norm": 4.8055959890559965, "learning_rate": 3.825601357068646e-06, "loss": 0.4267, "step": 8297 }, { "epoch": 0.5883020205600851, "grad_norm": 2.357243682285161, "learning_rate": 3.824485349433027e-06, "loss": 0.4058, "step": 8298 }, { "epoch": 0.5883729174051755, "grad_norm": 2.103616691270321, "learning_rate": 3.823369403783193e-06, "loss": 0.3679, "step": 8299 }, { "epoch": 0.5884438142502658, "grad_norm": 4.7742388299072385, "learning_rate": 3.8222535201779885e-06, "loss": 0.3757, "step": 8300 }, { "epoch": 0.5885147110953562, "grad_norm": 2.2653564505949566, "learning_rate": 3.821137698676255e-06, "loss": 0.4264, "step": 8301 }, { "epoch": 0.5885856079404467, "grad_norm": 5.044983080702189, "learning_rate": 3.82002193933683e-06, "loss": 0.437, "step": 8302 }, { "epoch": 0.588656504785537, "grad_norm": 2.57413518139029, "learning_rate": 3.81890624221855e-06, "loss": 0.3719, "step": 8303 }, { "epoch": 0.5887274016306274, "grad_norm": 3.8887069415833686, "learning_rate": 3.8177906073802445e-06, "loss": 0.4334, "step": 8304 }, { "epoch": 0.5887982984757179, "grad_norm": 4.038104316193361, "learning_rate": 3.8166750348807416e-06, "loss": 0.4501, "step": 8305 }, { "epoch": 0.5888691953208082, "grad_norm": 3.4874637039394125, "learning_rate": 3.815559524778869e-06, "loss": 0.3844, "step": 8306 }, { "epoch": 0.5889400921658986, "grad_norm": 2.3839406079703593, "learning_rate": 3.8144440771334458e-06, "loss": 0.3845, "step": 8307 }, { "epoch": 0.589010989010989, "grad_norm": 1.7994008229634535, "learning_rate": 3.8133286920032914e-06, "loss": 0.3915, "step": 8308 }, { "epoch": 0.5890818858560795, "grad_norm": 3.001504356293782, "learning_rate": 3.8122133694472196e-06, "loss": 0.4202, "step": 8309 }, { "epoch": 0.5891527827011698, "grad_norm": 2.8054958195894004, "learning_rate": 3.8110981095240466e-06, "loss": 0.456, "step": 8310 }, { "epoch": 0.5892236795462602, "grad_norm": 2.001686466420049, "learning_rate": 3.8099829122925747e-06, "loss": 0.3914, "step": 8311 }, { "epoch": 0.5892945763913506, "grad_norm": 2.847227261661988, "learning_rate": 3.8088677778116134e-06, "loss": 0.386, "step": 8312 }, { "epoch": 0.589365473236441, "grad_norm": 2.359104260563819, "learning_rate": 3.8077527061399645e-06, "loss": 0.4222, "step": 8313 }, { "epoch": 0.5894363700815314, "grad_norm": 4.568337045756176, "learning_rate": 3.8066376973364237e-06, "loss": 0.3962, "step": 8314 }, { "epoch": 0.5895072669266218, "grad_norm": 2.0755851346062264, "learning_rate": 3.805522751459788e-06, "loss": 0.4217, "step": 8315 }, { "epoch": 0.5895781637717121, "grad_norm": 2.8855894160001014, "learning_rate": 3.8044078685688505e-06, "loss": 0.3844, "step": 8316 }, { "epoch": 0.5896490606168026, "grad_norm": 6.851139251009476, "learning_rate": 3.8032930487223987e-06, "loss": 0.4063, "step": 8317 }, { "epoch": 0.589719957461893, "grad_norm": 2.7419949646099555, "learning_rate": 3.802178291979215e-06, "loss": 0.431, "step": 8318 }, { "epoch": 0.5897908543069833, "grad_norm": 3.1940522027379505, "learning_rate": 3.801063598398087e-06, "loss": 0.4415, "step": 8319 }, { "epoch": 0.5898617511520737, "grad_norm": 3.6854425474583623, "learning_rate": 3.79994896803779e-06, "loss": 0.3977, "step": 8320 }, { "epoch": 0.5899326479971642, "grad_norm": 1.6991916426948313, "learning_rate": 3.798834400957099e-06, "loss": 0.401, "step": 8321 }, { "epoch": 0.5900035448422545, "grad_norm": 35.24383792098601, "learning_rate": 3.7977198972147886e-06, "loss": 0.402, "step": 8322 }, { "epoch": 0.5900744416873449, "grad_norm": 6.396496730567279, "learning_rate": 3.7966054568696247e-06, "loss": 0.4417, "step": 8323 }, { "epoch": 0.5901453385324353, "grad_norm": 5.257811531161307, "learning_rate": 3.7954910799803723e-06, "loss": 0.3779, "step": 8324 }, { "epoch": 0.5902162353775257, "grad_norm": 6.905766213316744, "learning_rate": 3.7943767666057957e-06, "loss": 0.4196, "step": 8325 }, { "epoch": 0.5902871322226161, "grad_norm": 6.237899897960535, "learning_rate": 3.7932625168046535e-06, "loss": 0.4253, "step": 8326 }, { "epoch": 0.5903580290677065, "grad_norm": 1.9240428419297009, "learning_rate": 3.792148330635699e-06, "loss": 0.3873, "step": 8327 }, { "epoch": 0.5904289259127968, "grad_norm": 3.107536873329639, "learning_rate": 3.791034208157684e-06, "loss": 0.3991, "step": 8328 }, { "epoch": 0.5904998227578873, "grad_norm": 2.5782999114606353, "learning_rate": 3.7899201494293598e-06, "loss": 0.3974, "step": 8329 }, { "epoch": 0.5905707196029777, "grad_norm": 2.9268883092033926, "learning_rate": 3.788806154509467e-06, "loss": 0.4067, "step": 8330 }, { "epoch": 0.590641616448068, "grad_norm": 4.018102871619392, "learning_rate": 3.787692223456752e-06, "loss": 0.3827, "step": 8331 }, { "epoch": 0.5907125132931584, "grad_norm": 2.2314937382546196, "learning_rate": 3.7865783563299525e-06, "loss": 0.3818, "step": 8332 }, { "epoch": 0.5907834101382489, "grad_norm": 1.959665190867283, "learning_rate": 3.785464553187801e-06, "loss": 0.3501, "step": 8333 }, { "epoch": 0.5908543069833393, "grad_norm": 1.7191678606166578, "learning_rate": 3.78435081408903e-06, "loss": 0.4056, "step": 8334 }, { "epoch": 0.5909252038284296, "grad_norm": 1.4808959359424583, "learning_rate": 3.7832371390923704e-06, "loss": 0.4443, "step": 8335 }, { "epoch": 0.59099610067352, "grad_norm": 20.716985329809408, "learning_rate": 3.782123528256544e-06, "loss": 0.3805, "step": 8336 }, { "epoch": 0.5910669975186105, "grad_norm": 3.3136375688570987, "learning_rate": 3.7810099816402718e-06, "loss": 0.3606, "step": 8337 }, { "epoch": 0.5911378943637008, "grad_norm": 2.5322805975548115, "learning_rate": 3.7798964993022753e-06, "loss": 0.4507, "step": 8338 }, { "epoch": 0.5912087912087912, "grad_norm": 2.4180409296692895, "learning_rate": 3.778783081301269e-06, "loss": 0.3753, "step": 8339 }, { "epoch": 0.5912796880538816, "grad_norm": 3.8104977531659845, "learning_rate": 3.777669727695961e-06, "loss": 0.4262, "step": 8340 }, { "epoch": 0.591350584898972, "grad_norm": 3.733786179579143, "learning_rate": 3.7765564385450603e-06, "loss": 0.4274, "step": 8341 }, { "epoch": 0.5914214817440624, "grad_norm": 5.295117224338661, "learning_rate": 3.775443213907276e-06, "loss": 0.4225, "step": 8342 }, { "epoch": 0.5914923785891528, "grad_norm": 2.831171539086841, "learning_rate": 3.774330053841303e-06, "loss": 0.4017, "step": 8343 }, { "epoch": 0.5915632754342431, "grad_norm": 3.5858084747557926, "learning_rate": 3.773216958405842e-06, "loss": 0.3758, "step": 8344 }, { "epoch": 0.5916341722793336, "grad_norm": 2.9354793860517394, "learning_rate": 3.772103927659588e-06, "loss": 0.5196, "step": 8345 }, { "epoch": 0.591705069124424, "grad_norm": 2.290823198075332, "learning_rate": 3.7709909616612313e-06, "loss": 0.4308, "step": 8346 }, { "epoch": 0.5917759659695143, "grad_norm": 2.7253201642384193, "learning_rate": 3.7698780604694584e-06, "loss": 0.4129, "step": 8347 }, { "epoch": 0.5918468628146047, "grad_norm": 3.5610266186932025, "learning_rate": 3.768765224142955e-06, "loss": 0.3803, "step": 8348 }, { "epoch": 0.5919177596596952, "grad_norm": 8.404085192179963, "learning_rate": 3.7676524527404007e-06, "loss": 0.3935, "step": 8349 }, { "epoch": 0.5919886565047855, "grad_norm": 6.030403616323722, "learning_rate": 3.7665397463204718e-06, "loss": 0.4092, "step": 8350 }, { "epoch": 0.5920595533498759, "grad_norm": 2.4890311121769932, "learning_rate": 3.7654271049418456e-06, "loss": 0.3849, "step": 8351 }, { "epoch": 0.5921304501949664, "grad_norm": 1.634191694616664, "learning_rate": 3.76431452866319e-06, "loss": 0.3902, "step": 8352 }, { "epoch": 0.5922013470400567, "grad_norm": 3.437102557219554, "learning_rate": 3.7632020175431716e-06, "loss": 0.3805, "step": 8353 }, { "epoch": 0.5922722438851471, "grad_norm": 0.9987745876882558, "learning_rate": 3.762089571640454e-06, "loss": 0.7015, "step": 8354 }, { "epoch": 0.5923431407302375, "grad_norm": 2.374604485188252, "learning_rate": 3.7609771910137016e-06, "loss": 0.4082, "step": 8355 }, { "epoch": 0.5924140375753278, "grad_norm": 3.215851022540079, "learning_rate": 3.759864875721564e-06, "loss": 0.41, "step": 8356 }, { "epoch": 0.5924849344204183, "grad_norm": 2.2399853987639355, "learning_rate": 3.7587526258226997e-06, "loss": 0.4146, "step": 8357 }, { "epoch": 0.5925558312655087, "grad_norm": 2.1127366453790293, "learning_rate": 3.7576404413757572e-06, "loss": 0.3895, "step": 8358 }, { "epoch": 0.5926267281105991, "grad_norm": 2.733511858733474, "learning_rate": 3.7565283224393822e-06, "loss": 0.4324, "step": 8359 }, { "epoch": 0.5926976249556895, "grad_norm": 4.259352771147079, "learning_rate": 3.7554162690722173e-06, "loss": 0.3809, "step": 8360 }, { "epoch": 0.5927685218007799, "grad_norm": 2.791899306703329, "learning_rate": 3.754304281332904e-06, "loss": 0.3796, "step": 8361 }, { "epoch": 0.5928394186458703, "grad_norm": 2.4365890498731395, "learning_rate": 3.7531923592800747e-06, "loss": 0.3748, "step": 8362 }, { "epoch": 0.5929103154909606, "grad_norm": 2.0598845822720993, "learning_rate": 3.752080502972364e-06, "loss": 0.4223, "step": 8363 }, { "epoch": 0.5929812123360511, "grad_norm": 2.791631752769499, "learning_rate": 3.7509687124684026e-06, "loss": 0.3554, "step": 8364 }, { "epoch": 0.5930521091811415, "grad_norm": 3.709909503580613, "learning_rate": 3.7498569878268132e-06, "loss": 0.4134, "step": 8365 }, { "epoch": 0.5931230060262318, "grad_norm": 6.11502076415367, "learning_rate": 3.7487453291062194e-06, "loss": 0.3961, "step": 8366 }, { "epoch": 0.5931939028713222, "grad_norm": 4.915984763906213, "learning_rate": 3.7476337363652384e-06, "loss": 0.4243, "step": 8367 }, { "epoch": 0.5932647997164127, "grad_norm": 2.9794760287068383, "learning_rate": 3.7465222096624886e-06, "loss": 0.4308, "step": 8368 }, { "epoch": 0.593335696561503, "grad_norm": 2.338383212228572, "learning_rate": 3.7454107490565772e-06, "loss": 0.3768, "step": 8369 }, { "epoch": 0.5934065934065934, "grad_norm": 3.1085431923514713, "learning_rate": 3.7442993546061146e-06, "loss": 0.4173, "step": 8370 }, { "epoch": 0.5934774902516838, "grad_norm": 3.0367467749120904, "learning_rate": 3.7431880263697074e-06, "loss": 0.3712, "step": 8371 }, { "epoch": 0.5935483870967742, "grad_norm": 1.8225390640401684, "learning_rate": 3.7420767644059542e-06, "loss": 0.3845, "step": 8372 }, { "epoch": 0.5936192839418646, "grad_norm": 1.907321602444361, "learning_rate": 3.7409655687734515e-06, "loss": 0.4324, "step": 8373 }, { "epoch": 0.593690180786955, "grad_norm": 2.236542101310148, "learning_rate": 3.7398544395307983e-06, "loss": 0.4008, "step": 8374 }, { "epoch": 0.5937610776320453, "grad_norm": 2.756859697701813, "learning_rate": 3.73874337673658e-06, "loss": 0.4203, "step": 8375 }, { "epoch": 0.5938319744771358, "grad_norm": 3.0464030709966625, "learning_rate": 3.7376323804493863e-06, "loss": 0.3876, "step": 8376 }, { "epoch": 0.5939028713222262, "grad_norm": 2.0279609922471904, "learning_rate": 3.7365214507278023e-06, "loss": 0.4205, "step": 8377 }, { "epoch": 0.5939737681673165, "grad_norm": 2.9187170676678247, "learning_rate": 3.7354105876304054e-06, "loss": 0.3842, "step": 8378 }, { "epoch": 0.5940446650124069, "grad_norm": 1.7690700692305499, "learning_rate": 3.7342997912157738e-06, "loss": 0.3593, "step": 8379 }, { "epoch": 0.5941155618574974, "grad_norm": 2.6775405557248724, "learning_rate": 3.7331890615424803e-06, "loss": 0.3756, "step": 8380 }, { "epoch": 0.5941864587025877, "grad_norm": 3.2094466001161384, "learning_rate": 3.7320783986690946e-06, "loss": 0.3576, "step": 8381 }, { "epoch": 0.5942573555476781, "grad_norm": 3.0051217791792397, "learning_rate": 3.7309678026541807e-06, "loss": 0.4279, "step": 8382 }, { "epoch": 0.5943282523927685, "grad_norm": 2.9701315351772983, "learning_rate": 3.729857273556305e-06, "loss": 0.3785, "step": 8383 }, { "epoch": 0.594399149237859, "grad_norm": 9.494114794650573, "learning_rate": 3.7287468114340252e-06, "loss": 0.4351, "step": 8384 }, { "epoch": 0.5944700460829493, "grad_norm": 2.290979360833984, "learning_rate": 3.7276364163458957e-06, "loss": 0.4714, "step": 8385 }, { "epoch": 0.5945409429280397, "grad_norm": 6.506430999477394, "learning_rate": 3.726526088350468e-06, "loss": 0.4141, "step": 8386 }, { "epoch": 0.5946118397731301, "grad_norm": 2.0254528444519244, "learning_rate": 3.7254158275062947e-06, "loss": 0.3734, "step": 8387 }, { "epoch": 0.5946827366182205, "grad_norm": 3.4041279850966664, "learning_rate": 3.724305633871914e-06, "loss": 0.3697, "step": 8388 }, { "epoch": 0.5947536334633109, "grad_norm": 3.7200194589024806, "learning_rate": 3.7231955075058727e-06, "loss": 0.4043, "step": 8389 }, { "epoch": 0.5948245303084013, "grad_norm": 1.7702616297455107, "learning_rate": 3.722085448466708e-06, "loss": 0.3641, "step": 8390 }, { "epoch": 0.5948954271534916, "grad_norm": 2.522479217176453, "learning_rate": 3.720975456812952e-06, "loss": 0.4523, "step": 8391 }, { "epoch": 0.5949663239985821, "grad_norm": 2.90459852973915, "learning_rate": 3.7198655326031354e-06, "loss": 0.4201, "step": 8392 }, { "epoch": 0.5950372208436725, "grad_norm": 5.446949991163924, "learning_rate": 3.7187556758957883e-06, "loss": 0.3956, "step": 8393 }, { "epoch": 0.5951081176887628, "grad_norm": 3.679003361672454, "learning_rate": 3.7176458867494298e-06, "loss": 0.3736, "step": 8394 }, { "epoch": 0.5951790145338532, "grad_norm": 3.3738244363602283, "learning_rate": 3.716536165222584e-06, "loss": 0.4687, "step": 8395 }, { "epoch": 0.5952499113789437, "grad_norm": 2.4095499365684874, "learning_rate": 3.7154265113737666e-06, "loss": 0.3898, "step": 8396 }, { "epoch": 0.595320808224034, "grad_norm": 2.122362909486192, "learning_rate": 3.7143169252614884e-06, "loss": 0.3986, "step": 8397 }, { "epoch": 0.5953917050691244, "grad_norm": 2.500226808963436, "learning_rate": 3.7132074069442603e-06, "loss": 0.3769, "step": 8398 }, { "epoch": 0.5954626019142149, "grad_norm": 2.6365378658511687, "learning_rate": 3.7120979564805864e-06, "loss": 0.4125, "step": 8399 }, { "epoch": 0.5955334987593052, "grad_norm": 2.751213911350303, "learning_rate": 3.710988573928974e-06, "loss": 0.3632, "step": 8400 }, { "epoch": 0.5956043956043956, "grad_norm": 14.867124815460372, "learning_rate": 3.7098792593479142e-06, "loss": 0.4267, "step": 8401 }, { "epoch": 0.595675292449486, "grad_norm": 2.1877940292763576, "learning_rate": 3.708770012795907e-06, "loss": 0.4025, "step": 8402 }, { "epoch": 0.5957461892945763, "grad_norm": 2.4195717901959064, "learning_rate": 3.707660834331444e-06, "loss": 0.4202, "step": 8403 }, { "epoch": 0.5958170861396668, "grad_norm": 5.012909595721611, "learning_rate": 3.70655172401301e-06, "loss": 0.3574, "step": 8404 }, { "epoch": 0.5958879829847572, "grad_norm": 1.124350682165721, "learning_rate": 3.705442681899092e-06, "loss": 0.6447, "step": 8405 }, { "epoch": 0.5959588798298475, "grad_norm": 0.9770609205326336, "learning_rate": 3.7043337080481688e-06, "loss": 0.7252, "step": 8406 }, { "epoch": 0.596029776674938, "grad_norm": 2.6170541947220203, "learning_rate": 3.7032248025187174e-06, "loss": 0.4348, "step": 8407 }, { "epoch": 0.5961006735200284, "grad_norm": 3.1674886102509276, "learning_rate": 3.7021159653692122e-06, "loss": 0.435, "step": 8408 }, { "epoch": 0.5961715703651187, "grad_norm": 1.9988749606363552, "learning_rate": 3.701007196658124e-06, "loss": 0.3512, "step": 8409 }, { "epoch": 0.5962424672102091, "grad_norm": 9.879335955016073, "learning_rate": 3.6998984964439172e-06, "loss": 0.4067, "step": 8410 }, { "epoch": 0.5963133640552996, "grad_norm": 53.948351914316966, "learning_rate": 3.6987898647850544e-06, "loss": 0.4175, "step": 8411 }, { "epoch": 0.59638426090039, "grad_norm": 6.577817171226669, "learning_rate": 3.6976813017399947e-06, "loss": 0.4266, "step": 8412 }, { "epoch": 0.5964551577454803, "grad_norm": 4.591205087078315, "learning_rate": 3.696572807367197e-06, "loss": 0.3953, "step": 8413 }, { "epoch": 0.5965260545905707, "grad_norm": 2.984094595704296, "learning_rate": 3.6954643817251067e-06, "loss": 0.4403, "step": 8414 }, { "epoch": 0.5965969514356612, "grad_norm": 3.4749933608317742, "learning_rate": 3.694356024872176e-06, "loss": 0.3912, "step": 8415 }, { "epoch": 0.5966678482807515, "grad_norm": 2.830807887087703, "learning_rate": 3.69324773686685e-06, "loss": 0.4307, "step": 8416 }, { "epoch": 0.5967387451258419, "grad_norm": 5.468004675976067, "learning_rate": 3.692139517767568e-06, "loss": 0.3838, "step": 8417 }, { "epoch": 0.5968096419709323, "grad_norm": 2.200624736537567, "learning_rate": 3.691031367632765e-06, "loss": 0.4166, "step": 8418 }, { "epoch": 0.5968805388160227, "grad_norm": 4.249951840245618, "learning_rate": 3.6899232865208812e-06, "loss": 0.4, "step": 8419 }, { "epoch": 0.5969514356611131, "grad_norm": 5.846037711810964, "learning_rate": 3.6888152744903388e-06, "loss": 0.3784, "step": 8420 }, { "epoch": 0.5970223325062035, "grad_norm": 11.619924030227843, "learning_rate": 3.687707331599569e-06, "loss": 0.3449, "step": 8421 }, { "epoch": 0.5970932293512938, "grad_norm": 2.9519754202483313, "learning_rate": 3.686599457906994e-06, "loss": 0.4297, "step": 8422 }, { "epoch": 0.5971641261963843, "grad_norm": 3.125848180800389, "learning_rate": 3.685491653471032e-06, "loss": 0.4546, "step": 8423 }, { "epoch": 0.5972350230414747, "grad_norm": 6.761443319752294, "learning_rate": 3.6843839183500975e-06, "loss": 0.4042, "step": 8424 }, { "epoch": 0.597305919886565, "grad_norm": 2.614945334338458, "learning_rate": 3.683276252602604e-06, "loss": 0.3843, "step": 8425 }, { "epoch": 0.5973768167316554, "grad_norm": 2.9115887203030844, "learning_rate": 3.682168656286957e-06, "loss": 0.4017, "step": 8426 }, { "epoch": 0.5974477135767459, "grad_norm": 1.0186395482089357, "learning_rate": 3.681061129461564e-06, "loss": 0.6145, "step": 8427 }, { "epoch": 0.5975186104218362, "grad_norm": 3.6714732945819315, "learning_rate": 3.679953672184824e-06, "loss": 0.3514, "step": 8428 }, { "epoch": 0.5975895072669266, "grad_norm": 5.373443093561021, "learning_rate": 3.678846284515135e-06, "loss": 0.4146, "step": 8429 }, { "epoch": 0.597660404112017, "grad_norm": 7.6154285523536585, "learning_rate": 3.6777389665108893e-06, "loss": 0.4223, "step": 8430 }, { "epoch": 0.5977313009571074, "grad_norm": 2.893082465549817, "learning_rate": 3.6766317182304757e-06, "loss": 0.4658, "step": 8431 }, { "epoch": 0.5978021978021978, "grad_norm": 3.5233632535714783, "learning_rate": 3.675524539732286e-06, "loss": 0.4111, "step": 8432 }, { "epoch": 0.5978730946472882, "grad_norm": 3.3579091208790004, "learning_rate": 3.6744174310746936e-06, "loss": 0.4171, "step": 8433 }, { "epoch": 0.5979439914923785, "grad_norm": 3.3324358006211283, "learning_rate": 3.6733103923160835e-06, "loss": 0.3777, "step": 8434 }, { "epoch": 0.598014888337469, "grad_norm": 2.1905730523056137, "learning_rate": 3.67220342351483e-06, "loss": 0.3992, "step": 8435 }, { "epoch": 0.5980857851825594, "grad_norm": 2.654904740489847, "learning_rate": 3.671096524729303e-06, "loss": 0.4395, "step": 8436 }, { "epoch": 0.5981566820276498, "grad_norm": 3.1893827906151953, "learning_rate": 3.6699896960178693e-06, "loss": 0.431, "step": 8437 }, { "epoch": 0.5982275788727401, "grad_norm": 2.562835303217383, "learning_rate": 3.668882937438896e-06, "loss": 0.4012, "step": 8438 }, { "epoch": 0.5982984757178306, "grad_norm": 2.7798865271955386, "learning_rate": 3.66777624905074e-06, "loss": 0.4188, "step": 8439 }, { "epoch": 0.598369372562921, "grad_norm": 4.430705789559662, "learning_rate": 3.6666696309117587e-06, "loss": 0.426, "step": 8440 }, { "epoch": 0.5984402694080113, "grad_norm": 3.9900191847508286, "learning_rate": 3.6655630830803065e-06, "loss": 0.3615, "step": 8441 }, { "epoch": 0.5985111662531017, "grad_norm": 3.521793737175042, "learning_rate": 3.6644566056147323e-06, "loss": 0.3907, "step": 8442 }, { "epoch": 0.5985820630981922, "grad_norm": 3.682812102685932, "learning_rate": 3.6633501985733794e-06, "loss": 0.4409, "step": 8443 }, { "epoch": 0.5986529599432825, "grad_norm": 2.3853554454206227, "learning_rate": 3.6622438620145896e-06, "loss": 0.4042, "step": 8444 }, { "epoch": 0.5987238567883729, "grad_norm": 2.189081180916301, "learning_rate": 3.6611375959967055e-06, "loss": 0.3971, "step": 8445 }, { "epoch": 0.5987947536334634, "grad_norm": 3.1750778711620993, "learning_rate": 3.6600314005780548e-06, "loss": 0.4737, "step": 8446 }, { "epoch": 0.5988656504785537, "grad_norm": 2.2904349377309794, "learning_rate": 3.6589252758169715e-06, "loss": 0.3801, "step": 8447 }, { "epoch": 0.5989365473236441, "grad_norm": 4.813711525916893, "learning_rate": 3.6578192217717835e-06, "loss": 0.3764, "step": 8448 }, { "epoch": 0.5990074441687345, "grad_norm": 2.266069191396138, "learning_rate": 3.6567132385008112e-06, "loss": 0.3609, "step": 8449 }, { "epoch": 0.5990783410138248, "grad_norm": 2.72968453913867, "learning_rate": 3.655607326062375e-06, "loss": 0.4116, "step": 8450 }, { "epoch": 0.5991492378589153, "grad_norm": 1.801887798177187, "learning_rate": 3.6545014845147925e-06, "loss": 0.3798, "step": 8451 }, { "epoch": 0.5992201347040057, "grad_norm": 6.124292940930106, "learning_rate": 3.653395713916371e-06, "loss": 0.4343, "step": 8452 }, { "epoch": 0.599291031549096, "grad_norm": 3.043462155224594, "learning_rate": 3.6522900143254226e-06, "loss": 0.389, "step": 8453 }, { "epoch": 0.5993619283941864, "grad_norm": 3.6459347503656203, "learning_rate": 3.651184385800251e-06, "loss": 0.4017, "step": 8454 }, { "epoch": 0.5994328252392769, "grad_norm": 6.0562594383384, "learning_rate": 3.6500788283991552e-06, "loss": 0.4173, "step": 8455 }, { "epoch": 0.5995037220843672, "grad_norm": 3.1258014670581304, "learning_rate": 3.6489733421804335e-06, "loss": 0.3728, "step": 8456 }, { "epoch": 0.5995746189294576, "grad_norm": 2.133346152122346, "learning_rate": 3.6478679272023774e-06, "loss": 0.4221, "step": 8457 }, { "epoch": 0.599645515774548, "grad_norm": 1.8396963295414304, "learning_rate": 3.646762583523281e-06, "loss": 0.3736, "step": 8458 }, { "epoch": 0.5997164126196384, "grad_norm": 1.042497902313278, "learning_rate": 3.6456573112014227e-06, "loss": 0.7133, "step": 8459 }, { "epoch": 0.5997873094647288, "grad_norm": 2.0876322689868267, "learning_rate": 3.6445521102950897e-06, "loss": 0.3651, "step": 8460 }, { "epoch": 0.5998582063098192, "grad_norm": 2.3421521510911947, "learning_rate": 3.6434469808625594e-06, "loss": 0.3997, "step": 8461 }, { "epoch": 0.5999291031549097, "grad_norm": 2.2132344960724932, "learning_rate": 3.6423419229621037e-06, "loss": 0.4249, "step": 8462 }, { "epoch": 0.6, "grad_norm": 3.879689187813273, "learning_rate": 3.641236936651994e-06, "loss": 0.4485, "step": 8463 }, { "epoch": 0.6000708968450904, "grad_norm": 3.9420527707413124, "learning_rate": 3.6401320219905014e-06, "loss": 0.3534, "step": 8464 }, { "epoch": 0.6001417936901808, "grad_norm": 2.7653504545617684, "learning_rate": 3.639027179035881e-06, "loss": 0.3817, "step": 8465 }, { "epoch": 0.6002126905352712, "grad_norm": 3.515906860350891, "learning_rate": 3.637922407846398e-06, "loss": 0.4248, "step": 8466 }, { "epoch": 0.6002835873803616, "grad_norm": 5.603113006698517, "learning_rate": 3.636817708480307e-06, "loss": 0.4378, "step": 8467 }, { "epoch": 0.600354484225452, "grad_norm": 1.7748356562517635, "learning_rate": 3.6357130809958573e-06, "loss": 0.4181, "step": 8468 }, { "epoch": 0.6004253810705423, "grad_norm": 1.6111182887846565, "learning_rate": 3.6346085254512985e-06, "loss": 0.4125, "step": 8469 }, { "epoch": 0.6004962779156328, "grad_norm": 2.023715417839834, "learning_rate": 3.633504041904874e-06, "loss": 0.41, "step": 8470 }, { "epoch": 0.6005671747607232, "grad_norm": 2.550649336669861, "learning_rate": 3.6323996304148258e-06, "loss": 0.3782, "step": 8471 }, { "epoch": 0.6006380716058135, "grad_norm": 3.4635368372592987, "learning_rate": 3.6312952910393873e-06, "loss": 0.4004, "step": 8472 }, { "epoch": 0.6007089684509039, "grad_norm": 2.1973656860739648, "learning_rate": 3.6301910238367933e-06, "loss": 0.361, "step": 8473 }, { "epoch": 0.6007798652959944, "grad_norm": 2.6577106545355837, "learning_rate": 3.629086828865273e-06, "loss": 0.3876, "step": 8474 }, { "epoch": 0.6008507621410847, "grad_norm": 4.4670567340331, "learning_rate": 3.6279827061830497e-06, "loss": 0.4171, "step": 8475 }, { "epoch": 0.6009216589861751, "grad_norm": 3.3326161036330215, "learning_rate": 3.6268786558483447e-06, "loss": 0.4048, "step": 8476 }, { "epoch": 0.6009925558312655, "grad_norm": 3.4344184410790164, "learning_rate": 3.6257746779193793e-06, "loss": 0.4564, "step": 8477 }, { "epoch": 0.6010634526763559, "grad_norm": 3.7795686137951097, "learning_rate": 3.624670772454361e-06, "loss": 0.5001, "step": 8478 }, { "epoch": 0.6011343495214463, "grad_norm": 11.164802671251522, "learning_rate": 3.6235669395115024e-06, "loss": 0.442, "step": 8479 }, { "epoch": 0.6012052463665367, "grad_norm": 1.5451859751413317, "learning_rate": 3.6224631791490117e-06, "loss": 0.3912, "step": 8480 }, { "epoch": 0.601276143211627, "grad_norm": 4.761675478354997, "learning_rate": 3.6213594914250884e-06, "loss": 0.3701, "step": 8481 }, { "epoch": 0.6013470400567175, "grad_norm": 1.7508632957550634, "learning_rate": 3.6202558763979296e-06, "loss": 0.4325, "step": 8482 }, { "epoch": 0.6014179369018079, "grad_norm": 14.871647983450085, "learning_rate": 3.619152334125735e-06, "loss": 0.3633, "step": 8483 }, { "epoch": 0.6014888337468982, "grad_norm": 2.0630888218410437, "learning_rate": 3.6180488646666888e-06, "loss": 0.4149, "step": 8484 }, { "epoch": 0.6015597305919886, "grad_norm": 2.4493751919861197, "learning_rate": 3.616945468078982e-06, "loss": 0.3673, "step": 8485 }, { "epoch": 0.6016306274370791, "grad_norm": 2.856279462877311, "learning_rate": 3.615842144420795e-06, "loss": 0.4161, "step": 8486 }, { "epoch": 0.6017015242821695, "grad_norm": 4.180843375426767, "learning_rate": 3.61473889375031e-06, "loss": 0.3786, "step": 8487 }, { "epoch": 0.6017724211272598, "grad_norm": 2.6951613369046177, "learning_rate": 3.6136357161256997e-06, "loss": 0.3848, "step": 8488 }, { "epoch": 0.6018433179723502, "grad_norm": 2.2630940836545252, "learning_rate": 3.612532611605135e-06, "loss": 0.3892, "step": 8489 }, { "epoch": 0.6019142148174407, "grad_norm": 2.025586009267927, "learning_rate": 3.611429580246788e-06, "loss": 0.446, "step": 8490 }, { "epoch": 0.601985111662531, "grad_norm": 4.188796154043448, "learning_rate": 3.6103266221088164e-06, "loss": 0.4032, "step": 8491 }, { "epoch": 0.6020560085076214, "grad_norm": 2.6704496342621304, "learning_rate": 3.6092237372493827e-06, "loss": 0.3683, "step": 8492 }, { "epoch": 0.6021269053527118, "grad_norm": 3.1920801232964866, "learning_rate": 3.608120925726645e-06, "loss": 0.4072, "step": 8493 }, { "epoch": 0.6021978021978022, "grad_norm": 5.190493482431072, "learning_rate": 3.607018187598751e-06, "loss": 0.3756, "step": 8494 }, { "epoch": 0.6022686990428926, "grad_norm": 2.6959199560247584, "learning_rate": 3.6059155229238506e-06, "loss": 0.4048, "step": 8495 }, { "epoch": 0.602339595887983, "grad_norm": 2.7985663868406, "learning_rate": 3.604812931760092e-06, "loss": 0.4456, "step": 8496 }, { "epoch": 0.6024104927330733, "grad_norm": 2.1162220715970137, "learning_rate": 3.60371041416561e-06, "loss": 0.3759, "step": 8497 }, { "epoch": 0.6024813895781638, "grad_norm": 4.017145423893351, "learning_rate": 3.602607970198544e-06, "loss": 0.4405, "step": 8498 }, { "epoch": 0.6025522864232542, "grad_norm": 2.5257800587671646, "learning_rate": 3.6015055999170256e-06, "loss": 0.3996, "step": 8499 }, { "epoch": 0.6026231832683445, "grad_norm": 2.762751418854339, "learning_rate": 3.6004033033791863e-06, "loss": 0.3536, "step": 8500 }, { "epoch": 0.602694080113435, "grad_norm": 2.291917714726916, "learning_rate": 3.5993010806431476e-06, "loss": 0.4085, "step": 8501 }, { "epoch": 0.6027649769585254, "grad_norm": 3.6035557624933467, "learning_rate": 3.5981989317670313e-06, "loss": 0.4195, "step": 8502 }, { "epoch": 0.6028358738036157, "grad_norm": 2.2803796069349485, "learning_rate": 3.597096856808957e-06, "loss": 0.462, "step": 8503 }, { "epoch": 0.6029067706487061, "grad_norm": 1.779634127176986, "learning_rate": 3.5959948558270354e-06, "loss": 0.3869, "step": 8504 }, { "epoch": 0.6029776674937966, "grad_norm": 1.9757159822905415, "learning_rate": 3.5948929288793767e-06, "loss": 0.3983, "step": 8505 }, { "epoch": 0.6030485643388869, "grad_norm": 2.7080340787915937, "learning_rate": 3.5937910760240875e-06, "loss": 0.4249, "step": 8506 }, { "epoch": 0.6031194611839773, "grad_norm": 3.6695202245701135, "learning_rate": 3.5926892973192674e-06, "loss": 0.3809, "step": 8507 }, { "epoch": 0.6031903580290677, "grad_norm": 2.0995277476103213, "learning_rate": 3.5915875928230127e-06, "loss": 0.3569, "step": 8508 }, { "epoch": 0.603261254874158, "grad_norm": 4.778562096004595, "learning_rate": 3.590485962593423e-06, "loss": 0.4363, "step": 8509 }, { "epoch": 0.6033321517192485, "grad_norm": 2.0908547308254524, "learning_rate": 3.5893844066885813e-06, "loss": 0.4436, "step": 8510 }, { "epoch": 0.6034030485643389, "grad_norm": 1.922112080581458, "learning_rate": 3.5882829251665773e-06, "loss": 0.3893, "step": 8511 }, { "epoch": 0.6034739454094293, "grad_norm": 3.6194711127046997, "learning_rate": 3.5871815180854934e-06, "loss": 0.4173, "step": 8512 }, { "epoch": 0.6035448422545197, "grad_norm": 2.1481564265942974, "learning_rate": 3.5860801855034055e-06, "loss": 0.4376, "step": 8513 }, { "epoch": 0.6036157390996101, "grad_norm": 6.702550940551526, "learning_rate": 3.5849789274783884e-06, "loss": 0.3427, "step": 8514 }, { "epoch": 0.6036866359447005, "grad_norm": 2.0434118068221583, "learning_rate": 3.5838777440685113e-06, "loss": 0.3751, "step": 8515 }, { "epoch": 0.6037575327897908, "grad_norm": 3.9656432229177403, "learning_rate": 3.5827766353318437e-06, "loss": 0.4077, "step": 8516 }, { "epoch": 0.6038284296348813, "grad_norm": 2.372421165209539, "learning_rate": 3.5816756013264455e-06, "loss": 0.4147, "step": 8517 }, { "epoch": 0.6038993264799717, "grad_norm": 3.8425623025652107, "learning_rate": 3.580574642110375e-06, "loss": 0.4281, "step": 8518 }, { "epoch": 0.603970223325062, "grad_norm": 1.500391871367408, "learning_rate": 3.5794737577416877e-06, "loss": 0.6523, "step": 8519 }, { "epoch": 0.6040411201701524, "grad_norm": 5.311322012999223, "learning_rate": 3.5783729482784325e-06, "loss": 0.4018, "step": 8520 }, { "epoch": 0.6041120170152429, "grad_norm": 2.555012780114217, "learning_rate": 3.5772722137786557e-06, "loss": 0.3668, "step": 8521 }, { "epoch": 0.6041829138603332, "grad_norm": 2.561925760709283, "learning_rate": 3.5761715543004046e-06, "loss": 0.3902, "step": 8522 }, { "epoch": 0.6042538107054236, "grad_norm": 1.0669889956068743, "learning_rate": 3.5750709699017105e-06, "loss": 0.6493, "step": 8523 }, { "epoch": 0.604324707550514, "grad_norm": 2.6785959811985816, "learning_rate": 3.5739704606406136e-06, "loss": 0.4417, "step": 8524 }, { "epoch": 0.6043956043956044, "grad_norm": 2.7855216203494506, "learning_rate": 3.5728700265751426e-06, "loss": 0.39, "step": 8525 }, { "epoch": 0.6044665012406948, "grad_norm": 16.41759440001616, "learning_rate": 3.571769667763324e-06, "loss": 0.4238, "step": 8526 }, { "epoch": 0.6045373980857852, "grad_norm": 19.37470356071557, "learning_rate": 3.57066938426318e-06, "loss": 0.4516, "step": 8527 }, { "epoch": 0.6046082949308755, "grad_norm": 3.631835970581996, "learning_rate": 3.5695691761327335e-06, "loss": 0.3744, "step": 8528 }, { "epoch": 0.604679191775966, "grad_norm": 1.8313550585954885, "learning_rate": 3.5684690434299927e-06, "loss": 0.4133, "step": 8529 }, { "epoch": 0.6047500886210564, "grad_norm": 2.9504869973932775, "learning_rate": 3.5673689862129736e-06, "loss": 0.3689, "step": 8530 }, { "epoch": 0.6048209854661467, "grad_norm": 2.5215980546751533, "learning_rate": 3.56626900453968e-06, "loss": 0.4287, "step": 8531 }, { "epoch": 0.6048918823112371, "grad_norm": 2.855876975139866, "learning_rate": 3.565169098468118e-06, "loss": 0.4155, "step": 8532 }, { "epoch": 0.6049627791563276, "grad_norm": 2.837308265289163, "learning_rate": 3.564069268056284e-06, "loss": 0.4416, "step": 8533 }, { "epoch": 0.6050336760014179, "grad_norm": 2.8693919987580756, "learning_rate": 3.5629695133621717e-06, "loss": 0.337, "step": 8534 }, { "epoch": 0.6051045728465083, "grad_norm": 2.392391060652426, "learning_rate": 3.5618698344437764e-06, "loss": 0.3519, "step": 8535 }, { "epoch": 0.6051754696915987, "grad_norm": 4.995703472495934, "learning_rate": 3.560770231359081e-06, "loss": 0.4267, "step": 8536 }, { "epoch": 0.6052463665366891, "grad_norm": 3.3807370521199163, "learning_rate": 3.5596707041660693e-06, "loss": 0.379, "step": 8537 }, { "epoch": 0.6053172633817795, "grad_norm": 2.665313504793707, "learning_rate": 3.558571252922723e-06, "loss": 0.4059, "step": 8538 }, { "epoch": 0.6053881602268699, "grad_norm": 4.831509512217473, "learning_rate": 3.5574718776870134e-06, "loss": 0.4348, "step": 8539 }, { "epoch": 0.6054590570719603, "grad_norm": 1.980688479334254, "learning_rate": 3.5563725785169105e-06, "loss": 0.3913, "step": 8540 }, { "epoch": 0.6055299539170507, "grad_norm": 2.8414757505810373, "learning_rate": 3.5552733554703876e-06, "loss": 0.4247, "step": 8541 }, { "epoch": 0.6056008507621411, "grad_norm": 5.708418974593406, "learning_rate": 3.5541742086053997e-06, "loss": 0.3538, "step": 8542 }, { "epoch": 0.6056717476072315, "grad_norm": 4.521254944598117, "learning_rate": 3.553075137979911e-06, "loss": 0.319, "step": 8543 }, { "epoch": 0.6057426444523218, "grad_norm": 2.201941674327824, "learning_rate": 3.551976143651874e-06, "loss": 0.4519, "step": 8544 }, { "epoch": 0.6058135412974123, "grad_norm": 1.1492855571665095, "learning_rate": 3.550877225679242e-06, "loss": 0.6661, "step": 8545 }, { "epoch": 0.6058844381425027, "grad_norm": 5.282536852963733, "learning_rate": 3.5497783841199583e-06, "loss": 0.4024, "step": 8546 }, { "epoch": 0.605955334987593, "grad_norm": 3.4532106552887467, "learning_rate": 3.548679619031966e-06, "loss": 0.404, "step": 8547 }, { "epoch": 0.6060262318326834, "grad_norm": 3.2551860792040683, "learning_rate": 3.547580930473208e-06, "loss": 0.415, "step": 8548 }, { "epoch": 0.6060971286777739, "grad_norm": 2.344265605462187, "learning_rate": 3.546482318501615e-06, "loss": 0.371, "step": 8549 }, { "epoch": 0.6061680255228642, "grad_norm": 3.7767869490707455, "learning_rate": 3.5453837831751183e-06, "loss": 0.4265, "step": 8550 }, { "epoch": 0.6062389223679546, "grad_norm": 3.3718603062093404, "learning_rate": 3.544285324551646e-06, "loss": 0.3963, "step": 8551 }, { "epoch": 0.606309819213045, "grad_norm": 6.228682261041553, "learning_rate": 3.543186942689119e-06, "loss": 0.378, "step": 8552 }, { "epoch": 0.6063807160581354, "grad_norm": 3.1262776926191163, "learning_rate": 3.5420886376454553e-06, "loss": 0.4292, "step": 8553 }, { "epoch": 0.6064516129032258, "grad_norm": 2.398566516639634, "learning_rate": 3.5409904094785735e-06, "loss": 0.3743, "step": 8554 }, { "epoch": 0.6065225097483162, "grad_norm": 3.0620447553496053, "learning_rate": 3.539892258246378e-06, "loss": 0.4647, "step": 8555 }, { "epoch": 0.6065934065934065, "grad_norm": 1.919798706820701, "learning_rate": 3.538794184006779e-06, "loss": 0.4165, "step": 8556 }, { "epoch": 0.606664303438497, "grad_norm": 2.206929791859223, "learning_rate": 3.5376961868176794e-06, "loss": 0.3606, "step": 8557 }, { "epoch": 0.6067352002835874, "grad_norm": 2.8123260407744426, "learning_rate": 3.5365982667369746e-06, "loss": 0.4531, "step": 8558 }, { "epoch": 0.6068060971286777, "grad_norm": 2.297339602590643, "learning_rate": 3.5355004238225594e-06, "loss": 0.4163, "step": 8559 }, { "epoch": 0.6068769939737682, "grad_norm": 3.2591750673304873, "learning_rate": 3.5344026581323253e-06, "loss": 0.4039, "step": 8560 }, { "epoch": 0.6069478908188586, "grad_norm": 3.303752743457488, "learning_rate": 3.5333049697241588e-06, "loss": 0.3655, "step": 8561 }, { "epoch": 0.6070187876639489, "grad_norm": 4.497572215735525, "learning_rate": 3.5322073586559402e-06, "loss": 0.3918, "step": 8562 }, { "epoch": 0.6070896845090393, "grad_norm": 3.795209292611652, "learning_rate": 3.5311098249855474e-06, "loss": 0.3362, "step": 8563 }, { "epoch": 0.6071605813541298, "grad_norm": 3.9977466027968864, "learning_rate": 3.530012368770856e-06, "loss": 0.3644, "step": 8564 }, { "epoch": 0.6072314781992202, "grad_norm": 5.109200690988357, "learning_rate": 3.528914990069733e-06, "loss": 0.3466, "step": 8565 }, { "epoch": 0.6073023750443105, "grad_norm": 1.0216750630174845, "learning_rate": 3.5278176889400445e-06, "loss": 0.6433, "step": 8566 }, { "epoch": 0.6073732718894009, "grad_norm": 3.2822241652404758, "learning_rate": 3.526720465439656e-06, "loss": 0.387, "step": 8567 }, { "epoch": 0.6074441687344914, "grad_norm": 3.475399866676772, "learning_rate": 3.5256233196264187e-06, "loss": 0.4398, "step": 8568 }, { "epoch": 0.6075150655795817, "grad_norm": 2.625728040295798, "learning_rate": 3.52452625155819e-06, "loss": 0.3756, "step": 8569 }, { "epoch": 0.6075859624246721, "grad_norm": 3.2177670066936748, "learning_rate": 3.52342926129282e-06, "loss": 0.3941, "step": 8570 }, { "epoch": 0.6076568592697625, "grad_norm": 3.826366314696578, "learning_rate": 3.5223323488881506e-06, "loss": 0.3794, "step": 8571 }, { "epoch": 0.6077277561148529, "grad_norm": 2.1487884544510782, "learning_rate": 3.521235514402023e-06, "loss": 0.3823, "step": 8572 }, { "epoch": 0.6077986529599433, "grad_norm": 1.9078107558201978, "learning_rate": 3.520138757892277e-06, "loss": 0.386, "step": 8573 }, { "epoch": 0.6078695498050337, "grad_norm": 3.758953901783147, "learning_rate": 3.5190420794167444e-06, "loss": 0.4188, "step": 8574 }, { "epoch": 0.607940446650124, "grad_norm": 2.332898323528347, "learning_rate": 3.5179454790332525e-06, "loss": 0.3682, "step": 8575 }, { "epoch": 0.6080113434952145, "grad_norm": 8.60401227522604, "learning_rate": 3.5168489567996267e-06, "loss": 0.4128, "step": 8576 }, { "epoch": 0.6080822403403049, "grad_norm": 2.0808231919040603, "learning_rate": 3.515752512773688e-06, "loss": 0.406, "step": 8577 }, { "epoch": 0.6081531371853952, "grad_norm": 5.01221845074275, "learning_rate": 3.514656147013252e-06, "loss": 0.4043, "step": 8578 }, { "epoch": 0.6082240340304856, "grad_norm": 5.3733859211384996, "learning_rate": 3.5135598595761292e-06, "loss": 0.4109, "step": 8579 }, { "epoch": 0.6082949308755761, "grad_norm": 3.389591190897529, "learning_rate": 3.5124636505201328e-06, "loss": 0.4102, "step": 8580 }, { "epoch": 0.6083658277206664, "grad_norm": 2.6299166849264286, "learning_rate": 3.511367519903061e-06, "loss": 0.3837, "step": 8581 }, { "epoch": 0.6084367245657568, "grad_norm": 2.905214729325753, "learning_rate": 3.5102714677827165e-06, "loss": 0.394, "step": 8582 }, { "epoch": 0.6085076214108472, "grad_norm": 2.86578053031272, "learning_rate": 3.5091754942168964e-06, "loss": 0.3493, "step": 8583 }, { "epoch": 0.6085785182559376, "grad_norm": 1.2310742958855492, "learning_rate": 3.508079599263389e-06, "loss": 0.6723, "step": 8584 }, { "epoch": 0.608649415101028, "grad_norm": 4.677078503000642, "learning_rate": 3.5069837829799814e-06, "loss": 0.3978, "step": 8585 }, { "epoch": 0.6087203119461184, "grad_norm": 3.3262312430997274, "learning_rate": 3.505888045424462e-06, "loss": 0.3746, "step": 8586 }, { "epoch": 0.6087912087912087, "grad_norm": 5.584589593939243, "learning_rate": 3.5047923866546036e-06, "loss": 0.3447, "step": 8587 }, { "epoch": 0.6088621056362992, "grad_norm": 3.1109190708282, "learning_rate": 3.5036968067281846e-06, "loss": 0.4063, "step": 8588 }, { "epoch": 0.6089330024813896, "grad_norm": 9.128924926047526, "learning_rate": 3.502601305702975e-06, "loss": 0.4341, "step": 8589 }, { "epoch": 0.60900389932648, "grad_norm": 1.8143000757455834, "learning_rate": 3.5015058836367423e-06, "loss": 0.4087, "step": 8590 }, { "epoch": 0.6090747961715703, "grad_norm": 14.377721570209903, "learning_rate": 3.5004105405872468e-06, "loss": 0.4119, "step": 8591 }, { "epoch": 0.6091456930166608, "grad_norm": 4.938305971697354, "learning_rate": 3.4993152766122486e-06, "loss": 0.4523, "step": 8592 }, { "epoch": 0.6092165898617512, "grad_norm": 2.362888566971844, "learning_rate": 3.498220091769502e-06, "loss": 0.3709, "step": 8593 }, { "epoch": 0.6092874867068415, "grad_norm": 3.2069157094078156, "learning_rate": 3.4971249861167556e-06, "loss": 0.3541, "step": 8594 }, { "epoch": 0.609358383551932, "grad_norm": 8.657784323502986, "learning_rate": 3.4960299597117554e-06, "loss": 0.4243, "step": 8595 }, { "epoch": 0.6094292803970224, "grad_norm": 2.9326003212815577, "learning_rate": 3.494935012612245e-06, "loss": 0.4086, "step": 8596 }, { "epoch": 0.6095001772421127, "grad_norm": 1.9028277147171162, "learning_rate": 3.4938401448759585e-06, "loss": 0.4087, "step": 8597 }, { "epoch": 0.6095710740872031, "grad_norm": 3.613677251850271, "learning_rate": 3.4927453565606306e-06, "loss": 0.3975, "step": 8598 }, { "epoch": 0.6096419709322936, "grad_norm": 8.779883681736184, "learning_rate": 3.491650647723993e-06, "loss": 0.3538, "step": 8599 }, { "epoch": 0.6097128677773839, "grad_norm": 3.5905090484385993, "learning_rate": 3.4905560184237647e-06, "loss": 0.3771, "step": 8600 }, { "epoch": 0.6097837646224743, "grad_norm": 3.078464628437057, "learning_rate": 3.489461468717671e-06, "loss": 0.4078, "step": 8601 }, { "epoch": 0.6098546614675647, "grad_norm": 5.005151478021415, "learning_rate": 3.4883669986634265e-06, "loss": 0.4114, "step": 8602 }, { "epoch": 0.609925558312655, "grad_norm": 1.9198927185258028, "learning_rate": 3.487272608318746e-06, "loss": 0.3958, "step": 8603 }, { "epoch": 0.6099964551577455, "grad_norm": 2.170766213709893, "learning_rate": 3.4861782977413327e-06, "loss": 0.3989, "step": 8604 }, { "epoch": 0.6100673520028359, "grad_norm": 3.903068921078165, "learning_rate": 3.485084066988894e-06, "loss": 0.4566, "step": 8605 }, { "epoch": 0.6101382488479262, "grad_norm": 3.6373385570227734, "learning_rate": 3.4839899161191304e-06, "loss": 0.3936, "step": 8606 }, { "epoch": 0.6102091456930167, "grad_norm": 2.976733270774797, "learning_rate": 3.4828958451897343e-06, "loss": 0.4142, "step": 8607 }, { "epoch": 0.6102800425381071, "grad_norm": 2.4374566864065437, "learning_rate": 3.481801854258398e-06, "loss": 0.4021, "step": 8608 }, { "epoch": 0.6103509393831974, "grad_norm": 7.705966939136202, "learning_rate": 3.48070794338281e-06, "loss": 0.4138, "step": 8609 }, { "epoch": 0.6104218362282878, "grad_norm": 2.8114882116123607, "learning_rate": 3.4796141126206507e-06, "loss": 0.413, "step": 8610 }, { "epoch": 0.6104927330733783, "grad_norm": 5.2100457326080365, "learning_rate": 3.478520362029599e-06, "loss": 0.4772, "step": 8611 }, { "epoch": 0.6105636299184686, "grad_norm": 2.7124263464753757, "learning_rate": 3.4774266916673314e-06, "loss": 0.3796, "step": 8612 }, { "epoch": 0.610634526763559, "grad_norm": 2.1687290491358793, "learning_rate": 3.476333101591516e-06, "loss": 0.4728, "step": 8613 }, { "epoch": 0.6107054236086494, "grad_norm": 3.6707322610292703, "learning_rate": 3.475239591859819e-06, "loss": 0.3823, "step": 8614 }, { "epoch": 0.6107763204537399, "grad_norm": 2.4660581257557492, "learning_rate": 3.4741461625299024e-06, "loss": 0.3669, "step": 8615 }, { "epoch": 0.6108472172988302, "grad_norm": 3.733010803554711, "learning_rate": 3.4730528136594233e-06, "loss": 0.3987, "step": 8616 }, { "epoch": 0.6109181141439206, "grad_norm": 2.5816668826008375, "learning_rate": 3.4719595453060326e-06, "loss": 0.4313, "step": 8617 }, { "epoch": 0.610989010989011, "grad_norm": 3.8385061865786416, "learning_rate": 3.4708663575273834e-06, "loss": 0.4148, "step": 8618 }, { "epoch": 0.6110599078341014, "grad_norm": 2.644861430297398, "learning_rate": 3.4697732503811186e-06, "loss": 0.4359, "step": 8619 }, { "epoch": 0.6111308046791918, "grad_norm": 4.906658647277082, "learning_rate": 3.4686802239248785e-06, "loss": 0.3758, "step": 8620 }, { "epoch": 0.6112017015242822, "grad_norm": 2.0329197826229572, "learning_rate": 3.4675872782162978e-06, "loss": 0.3978, "step": 8621 }, { "epoch": 0.6112725983693725, "grad_norm": 2.8842420838533274, "learning_rate": 3.4664944133130106e-06, "loss": 0.3401, "step": 8622 }, { "epoch": 0.611343495214463, "grad_norm": 3.493364104516652, "learning_rate": 3.4654016292726423e-06, "loss": 0.3767, "step": 8623 }, { "epoch": 0.6114143920595534, "grad_norm": 2.13924521985368, "learning_rate": 3.4643089261528174e-06, "loss": 0.4008, "step": 8624 }, { "epoch": 0.6114852889046437, "grad_norm": 3.356508868156579, "learning_rate": 3.4632163040111567e-06, "loss": 0.3829, "step": 8625 }, { "epoch": 0.6115561857497341, "grad_norm": 4.934528304676936, "learning_rate": 3.462123762905272e-06, "loss": 0.4524, "step": 8626 }, { "epoch": 0.6116270825948246, "grad_norm": 2.1833873119487532, "learning_rate": 3.461031302892775e-06, "loss": 0.4462, "step": 8627 }, { "epoch": 0.6116979794399149, "grad_norm": 16.72509886615462, "learning_rate": 3.459938924031273e-06, "loss": 0.4153, "step": 8628 }, { "epoch": 0.6117688762850053, "grad_norm": 1.0578561451564106, "learning_rate": 3.4588466263783664e-06, "loss": 0.6298, "step": 8629 }, { "epoch": 0.6118397731300957, "grad_norm": 6.504198754244295, "learning_rate": 3.4577544099916522e-06, "loss": 0.4226, "step": 8630 }, { "epoch": 0.6119106699751861, "grad_norm": 4.029792969880598, "learning_rate": 3.456662274928728e-06, "loss": 0.4068, "step": 8631 }, { "epoch": 0.6119815668202765, "grad_norm": 2.5803901392736166, "learning_rate": 3.4555702212471776e-06, "loss": 0.3762, "step": 8632 }, { "epoch": 0.6120524636653669, "grad_norm": 4.0663505720497835, "learning_rate": 3.4544782490045887e-06, "loss": 0.3876, "step": 8633 }, { "epoch": 0.6121233605104572, "grad_norm": 0.9470636184643785, "learning_rate": 3.4533863582585412e-06, "loss": 0.6415, "step": 8634 }, { "epoch": 0.6121942573555477, "grad_norm": 2.102120943583878, "learning_rate": 3.4522945490666128e-06, "loss": 0.3741, "step": 8635 }, { "epoch": 0.6122651542006381, "grad_norm": 4.471860884909385, "learning_rate": 3.4512028214863723e-06, "loss": 0.4396, "step": 8636 }, { "epoch": 0.6123360510457284, "grad_norm": 4.172757510708574, "learning_rate": 3.45011117557539e-06, "loss": 0.4027, "step": 8637 }, { "epoch": 0.6124069478908188, "grad_norm": 3.0727889605267644, "learning_rate": 3.4490196113912293e-06, "loss": 0.3885, "step": 8638 }, { "epoch": 0.6124778447359093, "grad_norm": 5.402829114019177, "learning_rate": 3.4479281289914474e-06, "loss": 0.3995, "step": 8639 }, { "epoch": 0.6125487415809997, "grad_norm": 6.336694815132926, "learning_rate": 3.4468367284336e-06, "loss": 0.4039, "step": 8640 }, { "epoch": 0.61261963842609, "grad_norm": 2.3293771997280226, "learning_rate": 3.4457454097752386e-06, "loss": 0.4001, "step": 8641 }, { "epoch": 0.6126905352711804, "grad_norm": 2.8517878550140088, "learning_rate": 3.4446541730739065e-06, "loss": 0.4218, "step": 8642 }, { "epoch": 0.6127614321162709, "grad_norm": 8.76496551210963, "learning_rate": 3.443563018387146e-06, "loss": 0.3867, "step": 8643 }, { "epoch": 0.6128323289613612, "grad_norm": 4.418957126447753, "learning_rate": 3.4424719457724977e-06, "loss": 0.3994, "step": 8644 }, { "epoch": 0.6129032258064516, "grad_norm": 3.0949366018991973, "learning_rate": 3.4413809552874917e-06, "loss": 0.3632, "step": 8645 }, { "epoch": 0.612974122651542, "grad_norm": 6.018594656318208, "learning_rate": 3.4402900469896572e-06, "loss": 0.3979, "step": 8646 }, { "epoch": 0.6130450194966324, "grad_norm": 2.542317124278798, "learning_rate": 3.4391992209365183e-06, "loss": 0.3934, "step": 8647 }, { "epoch": 0.6131159163417228, "grad_norm": 3.0071873755113008, "learning_rate": 3.438108477185599e-06, "loss": 0.363, "step": 8648 }, { "epoch": 0.6131868131868132, "grad_norm": 2.693247653901383, "learning_rate": 3.437017815794408e-06, "loss": 0.3971, "step": 8649 }, { "epoch": 0.6132577100319035, "grad_norm": 6.65566953083326, "learning_rate": 3.435927236820462e-06, "loss": 0.4361, "step": 8650 }, { "epoch": 0.613328606876994, "grad_norm": 3.349517415178398, "learning_rate": 3.4348367403212677e-06, "loss": 0.399, "step": 8651 }, { "epoch": 0.6133995037220844, "grad_norm": 3.980864935826332, "learning_rate": 3.4337463263543257e-06, "loss": 0.4484, "step": 8652 }, { "epoch": 0.6134704005671747, "grad_norm": 10.662523917170635, "learning_rate": 3.4326559949771355e-06, "loss": 0.3872, "step": 8653 }, { "epoch": 0.6135412974122652, "grad_norm": 2.309194200477795, "learning_rate": 3.4315657462471917e-06, "loss": 0.4124, "step": 8654 }, { "epoch": 0.6136121942573556, "grad_norm": 3.789369855176038, "learning_rate": 3.4304755802219823e-06, "loss": 0.3769, "step": 8655 }, { "epoch": 0.6136830911024459, "grad_norm": 2.9119553814187173, "learning_rate": 3.429385496958993e-06, "loss": 0.4063, "step": 8656 }, { "epoch": 0.6137539879475363, "grad_norm": 5.127593728247667, "learning_rate": 3.4282954965157067e-06, "loss": 0.4077, "step": 8657 }, { "epoch": 0.6138248847926268, "grad_norm": 2.9096251076188446, "learning_rate": 3.4272055789495977e-06, "loss": 0.4376, "step": 8658 }, { "epoch": 0.6138957816377171, "grad_norm": 2.349744163439165, "learning_rate": 3.4261157443181393e-06, "loss": 0.4129, "step": 8659 }, { "epoch": 0.6139666784828075, "grad_norm": 3.4177299618988375, "learning_rate": 3.425025992678799e-06, "loss": 0.4287, "step": 8660 }, { "epoch": 0.6140375753278979, "grad_norm": 4.278780383270141, "learning_rate": 3.42393632408904e-06, "loss": 0.4044, "step": 8661 }, { "epoch": 0.6141084721729883, "grad_norm": 4.021617567134226, "learning_rate": 3.42284673860632e-06, "loss": 0.4384, "step": 8662 }, { "epoch": 0.6141793690180787, "grad_norm": 2.7162729978063944, "learning_rate": 3.421757236288097e-06, "loss": 0.355, "step": 8663 }, { "epoch": 0.6142502658631691, "grad_norm": 7.538998664029722, "learning_rate": 3.4206678171918205e-06, "loss": 0.4009, "step": 8664 }, { "epoch": 0.6143211627082594, "grad_norm": 2.472629434392548, "learning_rate": 3.4195784813749345e-06, "loss": 0.4125, "step": 8665 }, { "epoch": 0.6143920595533499, "grad_norm": 3.529600484440477, "learning_rate": 3.4184892288948813e-06, "loss": 0.3757, "step": 8666 }, { "epoch": 0.6144629563984403, "grad_norm": 2.951286201158492, "learning_rate": 3.4174000598090986e-06, "loss": 0.3862, "step": 8667 }, { "epoch": 0.6145338532435307, "grad_norm": 3.7434794023317877, "learning_rate": 3.4163109741750165e-06, "loss": 0.3885, "step": 8668 }, { "epoch": 0.614604750088621, "grad_norm": 3.3241749294662126, "learning_rate": 3.4152219720500677e-06, "loss": 0.3706, "step": 8669 }, { "epoch": 0.6146756469337115, "grad_norm": 3.105410737820696, "learning_rate": 3.4141330534916742e-06, "loss": 0.3886, "step": 8670 }, { "epoch": 0.6147465437788019, "grad_norm": 2.498796845517573, "learning_rate": 3.4130442185572544e-06, "loss": 0.3975, "step": 8671 }, { "epoch": 0.6148174406238922, "grad_norm": 3.2075187104044924, "learning_rate": 3.411955467304224e-06, "loss": 0.4099, "step": 8672 }, { "epoch": 0.6148883374689826, "grad_norm": 2.6873790947604594, "learning_rate": 3.4108667997899956e-06, "loss": 0.3979, "step": 8673 }, { "epoch": 0.6149592343140731, "grad_norm": 3.1326421505454585, "learning_rate": 3.409778216071972e-06, "loss": 0.3981, "step": 8674 }, { "epoch": 0.6150301311591634, "grad_norm": 5.708634661109359, "learning_rate": 3.408689716207556e-06, "loss": 0.3657, "step": 8675 }, { "epoch": 0.6151010280042538, "grad_norm": 2.2684427290511726, "learning_rate": 3.4076013002541474e-06, "loss": 0.4038, "step": 8676 }, { "epoch": 0.6151719248493442, "grad_norm": 3.042182358878035, "learning_rate": 3.406512968269138e-06, "loss": 0.3896, "step": 8677 }, { "epoch": 0.6152428216944346, "grad_norm": 2.170628081688066, "learning_rate": 3.4054247203099157e-06, "loss": 0.4069, "step": 8678 }, { "epoch": 0.615313718539525, "grad_norm": 8.91888497135757, "learning_rate": 3.4043365564338628e-06, "loss": 0.3719, "step": 8679 }, { "epoch": 0.6153846153846154, "grad_norm": 2.3241261280314003, "learning_rate": 3.4032484766983655e-06, "loss": 0.3631, "step": 8680 }, { "epoch": 0.6154555122297057, "grad_norm": 65.21603351840552, "learning_rate": 3.4021604811607913e-06, "loss": 0.4123, "step": 8681 }, { "epoch": 0.6155264090747962, "grad_norm": 4.13297722251145, "learning_rate": 3.4010725698785153e-06, "loss": 0.4745, "step": 8682 }, { "epoch": 0.6155973059198866, "grad_norm": 3.2451811980081633, "learning_rate": 3.399984742908904e-06, "loss": 0.4132, "step": 8683 }, { "epoch": 0.6156682027649769, "grad_norm": 2.0132999892376464, "learning_rate": 3.3988970003093173e-06, "loss": 0.4016, "step": 8684 }, { "epoch": 0.6157390996100673, "grad_norm": 4.991684976356959, "learning_rate": 3.3978093421371144e-06, "loss": 0.377, "step": 8685 }, { "epoch": 0.6158099964551578, "grad_norm": 3.6882255824769064, "learning_rate": 3.396721768449649e-06, "loss": 0.4162, "step": 8686 }, { "epoch": 0.6158808933002481, "grad_norm": 3.2355299574709226, "learning_rate": 3.395634279304266e-06, "loss": 0.386, "step": 8687 }, { "epoch": 0.6159517901453385, "grad_norm": 2.6286448762664953, "learning_rate": 3.394546874758313e-06, "loss": 0.3936, "step": 8688 }, { "epoch": 0.616022686990429, "grad_norm": 2.64155992623511, "learning_rate": 3.3934595548691297e-06, "loss": 0.4071, "step": 8689 }, { "epoch": 0.6160935838355193, "grad_norm": 3.5436256236126846, "learning_rate": 3.392372319694049e-06, "loss": 0.3728, "step": 8690 }, { "epoch": 0.6161644806806097, "grad_norm": 3.0441973067403354, "learning_rate": 3.3912851692904035e-06, "loss": 0.4018, "step": 8691 }, { "epoch": 0.6162353775257001, "grad_norm": 3.52536614862416, "learning_rate": 3.3901981037155173e-06, "loss": 0.4387, "step": 8692 }, { "epoch": 0.6163062743707906, "grad_norm": 2.70192686971709, "learning_rate": 3.389111123026717e-06, "loss": 0.4412, "step": 8693 }, { "epoch": 0.6163771712158809, "grad_norm": 3.265078947575407, "learning_rate": 3.388024227281314e-06, "loss": 0.4459, "step": 8694 }, { "epoch": 0.6164480680609713, "grad_norm": 9.034251750647156, "learning_rate": 3.386937416536625e-06, "loss": 0.4065, "step": 8695 }, { "epoch": 0.6165189649060617, "grad_norm": 3.735826562656741, "learning_rate": 3.385850690849958e-06, "loss": 0.3293, "step": 8696 }, { "epoch": 0.616589861751152, "grad_norm": 3.376705213432192, "learning_rate": 3.384764050278615e-06, "loss": 0.3538, "step": 8697 }, { "epoch": 0.6166607585962425, "grad_norm": 3.265602471020776, "learning_rate": 3.3836774948798968e-06, "loss": 0.3943, "step": 8698 }, { "epoch": 0.6167316554413329, "grad_norm": 1.0132015817460567, "learning_rate": 3.382591024711098e-06, "loss": 0.6442, "step": 8699 }, { "epoch": 0.6168025522864232, "grad_norm": 3.509203880420946, "learning_rate": 3.381504639829508e-06, "loss": 0.3987, "step": 8700 }, { "epoch": 0.6168734491315137, "grad_norm": 2.349180745760729, "learning_rate": 3.380418340292414e-06, "loss": 0.4109, "step": 8701 }, { "epoch": 0.6169443459766041, "grad_norm": 4.05509392557093, "learning_rate": 3.3793321261570986e-06, "loss": 0.3766, "step": 8702 }, { "epoch": 0.6170152428216944, "grad_norm": 3.8986479608388978, "learning_rate": 3.3782459974808365e-06, "loss": 0.4019, "step": 8703 }, { "epoch": 0.6170861396667848, "grad_norm": 3.9065116788248204, "learning_rate": 3.3771599543209e-06, "loss": 0.408, "step": 8704 }, { "epoch": 0.6171570365118753, "grad_norm": 4.934251092486923, "learning_rate": 3.376073996734557e-06, "loss": 0.3844, "step": 8705 }, { "epoch": 0.6172279333569656, "grad_norm": 3.3353190146780167, "learning_rate": 3.374988124779075e-06, "loss": 0.4213, "step": 8706 }, { "epoch": 0.617298830202056, "grad_norm": 5.442194578406769, "learning_rate": 3.373902338511706e-06, "loss": 0.4479, "step": 8707 }, { "epoch": 0.6173697270471464, "grad_norm": 3.7446605889501705, "learning_rate": 3.3728166379897087e-06, "loss": 0.3821, "step": 8708 }, { "epoch": 0.6174406238922368, "grad_norm": 6.958195527316637, "learning_rate": 3.3717310232703338e-06, "loss": 0.4121, "step": 8709 }, { "epoch": 0.6175115207373272, "grad_norm": 3.6864832589825776, "learning_rate": 3.3706454944108237e-06, "loss": 0.3746, "step": 8710 }, { "epoch": 0.6175824175824176, "grad_norm": 3.0376832502664826, "learning_rate": 3.36956005146842e-06, "loss": 0.3594, "step": 8711 }, { "epoch": 0.6176533144275079, "grad_norm": 4.11398180688828, "learning_rate": 3.36847469450036e-06, "loss": 0.4137, "step": 8712 }, { "epoch": 0.6177242112725984, "grad_norm": 3.9066793340280532, "learning_rate": 3.367389423563873e-06, "loss": 0.3913, "step": 8713 }, { "epoch": 0.6177951081176888, "grad_norm": 2.8717103610252797, "learning_rate": 3.3663042387161886e-06, "loss": 0.4136, "step": 8714 }, { "epoch": 0.6178660049627791, "grad_norm": 2.659639319159052, "learning_rate": 3.36521914001453e-06, "loss": 0.3345, "step": 8715 }, { "epoch": 0.6179369018078695, "grad_norm": 2.8118638801531515, "learning_rate": 3.3641341275161128e-06, "loss": 0.4201, "step": 8716 }, { "epoch": 0.61800779865296, "grad_norm": 2.5657057491659514, "learning_rate": 3.363049201278151e-06, "loss": 0.4079, "step": 8717 }, { "epoch": 0.6180786954980504, "grad_norm": 9.321243171337347, "learning_rate": 3.361964361357856e-06, "loss": 0.3679, "step": 8718 }, { "epoch": 0.6181495923431407, "grad_norm": 6.6146222942688695, "learning_rate": 3.3608796078124297e-06, "loss": 0.4356, "step": 8719 }, { "epoch": 0.6182204891882311, "grad_norm": 3.206947610972011, "learning_rate": 3.3597949406990714e-06, "loss": 0.4034, "step": 8720 }, { "epoch": 0.6182913860333216, "grad_norm": 3.3785568899723195, "learning_rate": 3.358710360074978e-06, "loss": 0.376, "step": 8721 }, { "epoch": 0.6183622828784119, "grad_norm": 3.6430682656132296, "learning_rate": 3.3576258659973427e-06, "loss": 0.4261, "step": 8722 }, { "epoch": 0.6184331797235023, "grad_norm": 3.96586356653281, "learning_rate": 3.3565414585233467e-06, "loss": 0.4107, "step": 8723 }, { "epoch": 0.6185040765685927, "grad_norm": 2.5943750329832556, "learning_rate": 3.3554571377101738e-06, "loss": 0.3863, "step": 8724 }, { "epoch": 0.6185749734136831, "grad_norm": 4.964630745832801, "learning_rate": 3.354372903615004e-06, "loss": 0.4017, "step": 8725 }, { "epoch": 0.6186458702587735, "grad_norm": 3.993267211947137, "learning_rate": 3.3532887562950034e-06, "loss": 0.3497, "step": 8726 }, { "epoch": 0.6187167671038639, "grad_norm": 3.0378282066855147, "learning_rate": 3.3522046958073453e-06, "loss": 0.4038, "step": 8727 }, { "epoch": 0.6187876639489542, "grad_norm": 2.861498566434304, "learning_rate": 3.3511207222091926e-06, "loss": 0.3785, "step": 8728 }, { "epoch": 0.6188585607940447, "grad_norm": 0.9931312663044394, "learning_rate": 3.3500368355577005e-06, "loss": 0.6442, "step": 8729 }, { "epoch": 0.6189294576391351, "grad_norm": 2.621313188149418, "learning_rate": 3.3489530359100264e-06, "loss": 0.3998, "step": 8730 }, { "epoch": 0.6190003544842254, "grad_norm": 6.04977152005698, "learning_rate": 3.3478693233233196e-06, "loss": 0.3913, "step": 8731 }, { "epoch": 0.6190712513293158, "grad_norm": 2.516298521610143, "learning_rate": 3.3467856978547227e-06, "loss": 0.4127, "step": 8732 }, { "epoch": 0.6191421481744063, "grad_norm": 5.307187980416792, "learning_rate": 3.345702159561378e-06, "loss": 0.3461, "step": 8733 }, { "epoch": 0.6192130450194966, "grad_norm": 3.5757276455391835, "learning_rate": 3.3446187085004222e-06, "loss": 0.4147, "step": 8734 }, { "epoch": 0.619283941864587, "grad_norm": 2.7071511417490304, "learning_rate": 3.3435353447289855e-06, "loss": 0.4074, "step": 8735 }, { "epoch": 0.6193548387096774, "grad_norm": 3.3814796393889672, "learning_rate": 3.342452068304194e-06, "loss": 0.3956, "step": 8736 }, { "epoch": 0.6194257355547678, "grad_norm": 3.959061237519692, "learning_rate": 3.341368879283169e-06, "loss": 0.367, "step": 8737 }, { "epoch": 0.6194966323998582, "grad_norm": 7.279627655592422, "learning_rate": 3.340285777723031e-06, "loss": 0.4118, "step": 8738 }, { "epoch": 0.6195675292449486, "grad_norm": 7.597942897270638, "learning_rate": 3.339202763680889e-06, "loss": 0.3111, "step": 8739 }, { "epoch": 0.6196384260900389, "grad_norm": 4.393530937598777, "learning_rate": 3.3381198372138535e-06, "loss": 0.4216, "step": 8740 }, { "epoch": 0.6197093229351294, "grad_norm": 6.734843773294506, "learning_rate": 3.3370369983790283e-06, "loss": 0.3947, "step": 8741 }, { "epoch": 0.6197802197802198, "grad_norm": 4.480646805076901, "learning_rate": 3.3359542472335105e-06, "loss": 0.4204, "step": 8742 }, { "epoch": 0.6198511166253102, "grad_norm": 2.37430984757821, "learning_rate": 3.334871583834395e-06, "loss": 0.3625, "step": 8743 }, { "epoch": 0.6199220134704005, "grad_norm": 3.717838051925801, "learning_rate": 3.3337890082387736e-06, "loss": 0.4203, "step": 8744 }, { "epoch": 0.619992910315491, "grad_norm": 13.583811137560245, "learning_rate": 3.3327065205037273e-06, "loss": 0.351, "step": 8745 }, { "epoch": 0.6200638071605814, "grad_norm": 2.5869270524007173, "learning_rate": 3.3316241206863396e-06, "loss": 0.3317, "step": 8746 }, { "epoch": 0.6201347040056717, "grad_norm": 3.247512338185431, "learning_rate": 3.3305418088436865e-06, "loss": 0.4138, "step": 8747 }, { "epoch": 0.6202056008507622, "grad_norm": 2.7225822115692853, "learning_rate": 3.3294595850328372e-06, "loss": 0.3778, "step": 8748 }, { "epoch": 0.6202764976958526, "grad_norm": 3.016958454519698, "learning_rate": 3.328377449310859e-06, "loss": 0.4298, "step": 8749 }, { "epoch": 0.6203473945409429, "grad_norm": 3.900026032734709, "learning_rate": 3.3272954017348126e-06, "loss": 0.4517, "step": 8750 }, { "epoch": 0.6204182913860333, "grad_norm": 4.220579270623329, "learning_rate": 3.3262134423617587e-06, "loss": 0.4349, "step": 8751 }, { "epoch": 0.6204891882311238, "grad_norm": 9.9633642406399, "learning_rate": 3.325131571248745e-06, "loss": 0.3968, "step": 8752 }, { "epoch": 0.6205600850762141, "grad_norm": 2.5233147538385117, "learning_rate": 3.324049788452824e-06, "loss": 0.3943, "step": 8753 }, { "epoch": 0.6206309819213045, "grad_norm": 2.393710800846362, "learning_rate": 3.3229680940310373e-06, "loss": 0.3639, "step": 8754 }, { "epoch": 0.6207018787663949, "grad_norm": 2.677089984743061, "learning_rate": 3.3218864880404213e-06, "loss": 0.4456, "step": 8755 }, { "epoch": 0.6207727756114853, "grad_norm": 3.867147441892003, "learning_rate": 3.320804970538011e-06, "loss": 0.345, "step": 8756 }, { "epoch": 0.6208436724565757, "grad_norm": 3.645183256478628, "learning_rate": 3.319723541580839e-06, "loss": 0.3795, "step": 8757 }, { "epoch": 0.6209145693016661, "grad_norm": 1.1647718716081297, "learning_rate": 3.3186422012259244e-06, "loss": 0.7345, "step": 8758 }, { "epoch": 0.6209854661467564, "grad_norm": 5.219849316326054, "learning_rate": 3.3175609495302917e-06, "loss": 0.3942, "step": 8759 }, { "epoch": 0.6210563629918469, "grad_norm": 2.6353620183067723, "learning_rate": 3.3164797865509548e-06, "loss": 0.3641, "step": 8760 }, { "epoch": 0.6211272598369373, "grad_norm": 3.1588112790961453, "learning_rate": 3.3153987123449216e-06, "loss": 0.3678, "step": 8761 }, { "epoch": 0.6211981566820276, "grad_norm": 5.178813279602526, "learning_rate": 3.3143177269692017e-06, "loss": 0.4076, "step": 8762 }, { "epoch": 0.621269053527118, "grad_norm": 2.861471402460168, "learning_rate": 3.313236830480795e-06, "loss": 0.3743, "step": 8763 }, { "epoch": 0.6213399503722085, "grad_norm": 5.59269754690614, "learning_rate": 3.312156022936696e-06, "loss": 0.3571, "step": 8764 }, { "epoch": 0.6214108472172988, "grad_norm": 7.860166903049076, "learning_rate": 3.311075304393898e-06, "loss": 0.4442, "step": 8765 }, { "epoch": 0.6214817440623892, "grad_norm": 21.291832618619225, "learning_rate": 3.309994674909389e-06, "loss": 0.3495, "step": 8766 }, { "epoch": 0.6215526409074796, "grad_norm": 4.746637652845608, "learning_rate": 3.308914134540152e-06, "loss": 0.4158, "step": 8767 }, { "epoch": 0.6216235377525701, "grad_norm": 2.676347097223002, "learning_rate": 3.3078336833431625e-06, "loss": 0.462, "step": 8768 }, { "epoch": 0.6216944345976604, "grad_norm": 5.392794655319586, "learning_rate": 3.3067533213753925e-06, "loss": 0.4436, "step": 8769 }, { "epoch": 0.6217653314427508, "grad_norm": 5.754883753565773, "learning_rate": 3.305673048693816e-06, "loss": 0.3868, "step": 8770 }, { "epoch": 0.6218362282878412, "grad_norm": 5.641678338287083, "learning_rate": 3.30459286535539e-06, "loss": 0.3819, "step": 8771 }, { "epoch": 0.6219071251329316, "grad_norm": 5.9542659477792315, "learning_rate": 3.3035127714170765e-06, "loss": 0.4172, "step": 8772 }, { "epoch": 0.621978021978022, "grad_norm": 2.8943955542874225, "learning_rate": 3.302432766935831e-06, "loss": 0.4244, "step": 8773 }, { "epoch": 0.6220489188231124, "grad_norm": 3.30193720265099, "learning_rate": 3.3013528519686013e-06, "loss": 0.3445, "step": 8774 }, { "epoch": 0.6221198156682027, "grad_norm": 2.202042860293348, "learning_rate": 3.300273026572331e-06, "loss": 0.4193, "step": 8775 }, { "epoch": 0.6221907125132932, "grad_norm": 4.012600974362955, "learning_rate": 3.2991932908039627e-06, "loss": 0.4049, "step": 8776 }, { "epoch": 0.6222616093583836, "grad_norm": 2.427395187994888, "learning_rate": 3.298113644720429e-06, "loss": 0.423, "step": 8777 }, { "epoch": 0.6223325062034739, "grad_norm": 1.742851396986875, "learning_rate": 3.297034088378661e-06, "loss": 0.361, "step": 8778 }, { "epoch": 0.6224034030485643, "grad_norm": 2.606952078089131, "learning_rate": 3.295954621835587e-06, "loss": 0.3754, "step": 8779 }, { "epoch": 0.6224742998936548, "grad_norm": 1.023489773283869, "learning_rate": 3.294875245148127e-06, "loss": 0.6464, "step": 8780 }, { "epoch": 0.6225451967387451, "grad_norm": 3.7245742852717223, "learning_rate": 3.2937959583731956e-06, "loss": 0.4294, "step": 8781 }, { "epoch": 0.6226160935838355, "grad_norm": 4.215724510597015, "learning_rate": 3.2927167615677047e-06, "loss": 0.4228, "step": 8782 }, { "epoch": 0.622686990428926, "grad_norm": 3.6771117238795266, "learning_rate": 3.291637654788565e-06, "loss": 0.3903, "step": 8783 }, { "epoch": 0.6227578872740163, "grad_norm": 10.413845146865418, "learning_rate": 3.2905586380926726e-06, "loss": 0.3842, "step": 8784 }, { "epoch": 0.6228287841191067, "grad_norm": 3.47240463466783, "learning_rate": 3.2894797115369293e-06, "loss": 0.3952, "step": 8785 }, { "epoch": 0.6228996809641971, "grad_norm": 3.2384621196964676, "learning_rate": 3.288400875178227e-06, "loss": 0.3912, "step": 8786 }, { "epoch": 0.6229705778092874, "grad_norm": 6.2136113740466925, "learning_rate": 3.2873221290734526e-06, "loss": 0.4066, "step": 8787 }, { "epoch": 0.6230414746543779, "grad_norm": 2.3856330188142008, "learning_rate": 3.286243473279488e-06, "loss": 0.4147, "step": 8788 }, { "epoch": 0.6231123714994683, "grad_norm": 2.791275440513632, "learning_rate": 3.285164907853216e-06, "loss": 0.4147, "step": 8789 }, { "epoch": 0.6231832683445586, "grad_norm": 1.9342699713567217, "learning_rate": 3.2840864328515044e-06, "loss": 0.3864, "step": 8790 }, { "epoch": 0.623254165189649, "grad_norm": 2.715964395842066, "learning_rate": 3.2830080483312264e-06, "loss": 0.408, "step": 8791 }, { "epoch": 0.6233250620347395, "grad_norm": 5.532238209728573, "learning_rate": 3.2819297543492453e-06, "loss": 0.3682, "step": 8792 }, { "epoch": 0.6233959588798298, "grad_norm": 2.7019105615272303, "learning_rate": 3.2808515509624183e-06, "loss": 0.3888, "step": 8793 }, { "epoch": 0.6234668557249202, "grad_norm": 5.743629611117142, "learning_rate": 3.2797734382276015e-06, "loss": 0.3931, "step": 8794 }, { "epoch": 0.6235377525700107, "grad_norm": 4.026609882666755, "learning_rate": 3.2786954162016425e-06, "loss": 0.3728, "step": 8795 }, { "epoch": 0.6236086494151011, "grad_norm": 2.0368937954810504, "learning_rate": 3.277617484941392e-06, "loss": 0.3483, "step": 8796 }, { "epoch": 0.6236795462601914, "grad_norm": 4.186011100899595, "learning_rate": 3.2765396445036823e-06, "loss": 0.361, "step": 8797 }, { "epoch": 0.6237504431052818, "grad_norm": 2.16013723610519, "learning_rate": 3.2754618949453543e-06, "loss": 0.4015, "step": 8798 }, { "epoch": 0.6238213399503723, "grad_norm": 5.556129679572505, "learning_rate": 3.2743842363232383e-06, "loss": 0.4552, "step": 8799 }, { "epoch": 0.6238922367954626, "grad_norm": 4.325302052446093, "learning_rate": 3.2733066686941563e-06, "loss": 0.4087, "step": 8800 }, { "epoch": 0.623963133640553, "grad_norm": 1.0261061609604496, "learning_rate": 3.272229192114931e-06, "loss": 0.618, "step": 8801 }, { "epoch": 0.6240340304856434, "grad_norm": 2.7288923091954196, "learning_rate": 3.271151806642383e-06, "loss": 0.3571, "step": 8802 }, { "epoch": 0.6241049273307337, "grad_norm": 3.5411222076172417, "learning_rate": 3.270074512333316e-06, "loss": 0.4233, "step": 8803 }, { "epoch": 0.6241758241758242, "grad_norm": 3.1976009609482086, "learning_rate": 3.2689973092445416e-06, "loss": 0.4128, "step": 8804 }, { "epoch": 0.6242467210209146, "grad_norm": 5.368789049397965, "learning_rate": 3.2679201974328613e-06, "loss": 0.4026, "step": 8805 }, { "epoch": 0.6243176178660049, "grad_norm": 2.9392769999312836, "learning_rate": 3.26684317695507e-06, "loss": 0.3804, "step": 8806 }, { "epoch": 0.6243885147110954, "grad_norm": 2.3708407630197503, "learning_rate": 3.265766247867961e-06, "loss": 0.3926, "step": 8807 }, { "epoch": 0.6244594115561858, "grad_norm": 1.7354170679077785, "learning_rate": 3.26468941022832e-06, "loss": 0.3652, "step": 8808 }, { "epoch": 0.6245303084012761, "grad_norm": 4.505801735651291, "learning_rate": 3.2636126640929332e-06, "loss": 0.4592, "step": 8809 }, { "epoch": 0.6246012052463665, "grad_norm": 2.649066772988138, "learning_rate": 3.262536009518575e-06, "loss": 0.4656, "step": 8810 }, { "epoch": 0.624672102091457, "grad_norm": 2.1328924883971756, "learning_rate": 3.2614594465620187e-06, "loss": 0.417, "step": 8811 }, { "epoch": 0.6247429989365473, "grad_norm": 2.7022932826756256, "learning_rate": 3.260382975280034e-06, "loss": 0.4317, "step": 8812 }, { "epoch": 0.6248138957816377, "grad_norm": 2.579482836957691, "learning_rate": 3.2593065957293824e-06, "loss": 0.4, "step": 8813 }, { "epoch": 0.6248847926267281, "grad_norm": 6.922751148217892, "learning_rate": 3.258230307966821e-06, "loss": 0.406, "step": 8814 }, { "epoch": 0.6249556894718185, "grad_norm": 5.35108551816197, "learning_rate": 3.257154112049108e-06, "loss": 0.3585, "step": 8815 }, { "epoch": 0.6250265863169089, "grad_norm": 3.5211233597768214, "learning_rate": 3.2560780080329855e-06, "loss": 0.3498, "step": 8816 }, { "epoch": 0.6250974831619993, "grad_norm": 3.5221672533852373, "learning_rate": 3.255001995975203e-06, "loss": 0.3515, "step": 8817 }, { "epoch": 0.6251683800070896, "grad_norm": 7.33933065736067, "learning_rate": 3.253926075932497e-06, "loss": 0.3974, "step": 8818 }, { "epoch": 0.6252392768521801, "grad_norm": 6.096508161853162, "learning_rate": 3.2528502479616008e-06, "loss": 0.375, "step": 8819 }, { "epoch": 0.6253101736972705, "grad_norm": 4.3347576896405045, "learning_rate": 3.251774512119244e-06, "loss": 0.4235, "step": 8820 }, { "epoch": 0.6253810705423609, "grad_norm": 2.5237305097426157, "learning_rate": 3.2506988684621545e-06, "loss": 0.4341, "step": 8821 }, { "epoch": 0.6254519673874512, "grad_norm": 4.453296913034211, "learning_rate": 3.249623317047046e-06, "loss": 0.4446, "step": 8822 }, { "epoch": 0.6255228642325417, "grad_norm": 8.720754635995858, "learning_rate": 3.248547857930636e-06, "loss": 0.451, "step": 8823 }, { "epoch": 0.6255937610776321, "grad_norm": 26.147512410484364, "learning_rate": 3.2474724911696354e-06, "loss": 0.4429, "step": 8824 }, { "epoch": 0.6256646579227224, "grad_norm": 17.602572836547786, "learning_rate": 3.2463972168207487e-06, "loss": 0.3509, "step": 8825 }, { "epoch": 0.6257355547678128, "grad_norm": 4.438894849283746, "learning_rate": 3.245322034940674e-06, "loss": 0.3838, "step": 8826 }, { "epoch": 0.6258064516129033, "grad_norm": 2.846670774062782, "learning_rate": 3.2442469455861063e-06, "loss": 0.3556, "step": 8827 }, { "epoch": 0.6258773484579936, "grad_norm": 4.5671842657441655, "learning_rate": 3.2431719488137416e-06, "loss": 0.3322, "step": 8828 }, { "epoch": 0.625948245303084, "grad_norm": 2.369747102221818, "learning_rate": 3.242097044680257e-06, "loss": 0.3844, "step": 8829 }, { "epoch": 0.6260191421481744, "grad_norm": 6.418120217013829, "learning_rate": 3.2410222332423374e-06, "loss": 0.4008, "step": 8830 }, { "epoch": 0.6260900389932648, "grad_norm": 6.657508582800971, "learning_rate": 3.2399475145566593e-06, "loss": 0.3947, "step": 8831 }, { "epoch": 0.6261609358383552, "grad_norm": 1.938140326771451, "learning_rate": 3.238872888679892e-06, "loss": 0.426, "step": 8832 }, { "epoch": 0.6262318326834456, "grad_norm": 2.085932256931798, "learning_rate": 3.2377983556686997e-06, "loss": 0.3711, "step": 8833 }, { "epoch": 0.6263027295285359, "grad_norm": 6.25851812097181, "learning_rate": 3.236723915579749e-06, "loss": 0.4424, "step": 8834 }, { "epoch": 0.6263736263736264, "grad_norm": 5.184453512720953, "learning_rate": 3.235649568469689e-06, "loss": 0.4178, "step": 8835 }, { "epoch": 0.6264445232187168, "grad_norm": 3.6465179772668455, "learning_rate": 3.2345753143951743e-06, "loss": 0.4518, "step": 8836 }, { "epoch": 0.6265154200638071, "grad_norm": 2.7253015724924903, "learning_rate": 3.23350115341285e-06, "loss": 0.3884, "step": 8837 }, { "epoch": 0.6265863169088975, "grad_norm": 2.783088064532297, "learning_rate": 3.2324270855793604e-06, "loss": 0.3474, "step": 8838 }, { "epoch": 0.626657213753988, "grad_norm": 5.207064262637931, "learning_rate": 3.2313531109513383e-06, "loss": 0.345, "step": 8839 }, { "epoch": 0.6267281105990783, "grad_norm": 2.769210362747082, "learning_rate": 3.2302792295854155e-06, "loss": 0.4337, "step": 8840 }, { "epoch": 0.6267990074441687, "grad_norm": 2.461691624690482, "learning_rate": 3.229205441538222e-06, "loss": 0.3467, "step": 8841 }, { "epoch": 0.6268699042892591, "grad_norm": 3.8648327090087284, "learning_rate": 3.2281317468663764e-06, "loss": 0.4239, "step": 8842 }, { "epoch": 0.6269408011343495, "grad_norm": 3.0958143818106327, "learning_rate": 3.227058145626496e-06, "loss": 0.3909, "step": 8843 }, { "epoch": 0.6270116979794399, "grad_norm": 6.322105934250334, "learning_rate": 3.2259846378751936e-06, "loss": 0.3768, "step": 8844 }, { "epoch": 0.6270825948245303, "grad_norm": 4.416049771245759, "learning_rate": 3.224911223669074e-06, "loss": 0.3545, "step": 8845 }, { "epoch": 0.6271534916696208, "grad_norm": 3.2049079492147716, "learning_rate": 3.22383790306474e-06, "loss": 0.4148, "step": 8846 }, { "epoch": 0.6272243885147111, "grad_norm": 1.788208218449262, "learning_rate": 3.2227646761187914e-06, "loss": 0.3547, "step": 8847 }, { "epoch": 0.6272952853598015, "grad_norm": 1.8917568631094643, "learning_rate": 3.221691542887816e-06, "loss": 0.3831, "step": 8848 }, { "epoch": 0.6273661822048919, "grad_norm": 6.607902682220276, "learning_rate": 3.220618503428403e-06, "loss": 0.4051, "step": 8849 }, { "epoch": 0.6274370790499822, "grad_norm": 4.768538618817456, "learning_rate": 3.2195455577971357e-06, "loss": 0.4442, "step": 8850 }, { "epoch": 0.6275079758950727, "grad_norm": 2.4761533010899686, "learning_rate": 3.2184727060505887e-06, "loss": 0.3704, "step": 8851 }, { "epoch": 0.6275788727401631, "grad_norm": 2.555997910245496, "learning_rate": 3.2173999482453364e-06, "loss": 0.3972, "step": 8852 }, { "epoch": 0.6276497695852534, "grad_norm": 4.085084482815014, "learning_rate": 3.216327284437944e-06, "loss": 0.4091, "step": 8853 }, { "epoch": 0.6277206664303439, "grad_norm": 3.399511310441433, "learning_rate": 3.215254714684977e-06, "loss": 0.4949, "step": 8854 }, { "epoch": 0.6277915632754343, "grad_norm": 6.991906068840576, "learning_rate": 3.2141822390429913e-06, "loss": 0.4151, "step": 8855 }, { "epoch": 0.6278624601205246, "grad_norm": 9.267883537439328, "learning_rate": 3.213109857568538e-06, "loss": 0.3978, "step": 8856 }, { "epoch": 0.627933356965615, "grad_norm": 3.4127443752076876, "learning_rate": 3.212037570318167e-06, "loss": 0.428, "step": 8857 }, { "epoch": 0.6280042538107055, "grad_norm": 6.13170889639586, "learning_rate": 3.210965377348419e-06, "loss": 0.4459, "step": 8858 }, { "epoch": 0.6280751506557958, "grad_norm": 2.6168075063929823, "learning_rate": 3.209893278715831e-06, "loss": 0.4093, "step": 8859 }, { "epoch": 0.6281460475008862, "grad_norm": 6.46039244635598, "learning_rate": 3.2088212744769397e-06, "loss": 0.406, "step": 8860 }, { "epoch": 0.6282169443459766, "grad_norm": 2.1176041256886404, "learning_rate": 3.207749364688266e-06, "loss": 0.4019, "step": 8861 }, { "epoch": 0.628287841191067, "grad_norm": 3.211576982787828, "learning_rate": 3.2066775494063386e-06, "loss": 0.3884, "step": 8862 }, { "epoch": 0.6283587380361574, "grad_norm": 2.368393308319115, "learning_rate": 3.2056058286876733e-06, "loss": 0.3884, "step": 8863 }, { "epoch": 0.6284296348812478, "grad_norm": 3.887040994979243, "learning_rate": 3.204534202588781e-06, "loss": 0.3663, "step": 8864 }, { "epoch": 0.6285005317263381, "grad_norm": 3.049519247629884, "learning_rate": 3.2034626711661698e-06, "loss": 0.4155, "step": 8865 }, { "epoch": 0.6285714285714286, "grad_norm": 2.5440277525473176, "learning_rate": 3.2023912344763466e-06, "loss": 0.4046, "step": 8866 }, { "epoch": 0.628642325416519, "grad_norm": 5.158069689714607, "learning_rate": 3.2013198925758015e-06, "loss": 0.4013, "step": 8867 }, { "epoch": 0.6287132222616093, "grad_norm": 0.983232696269567, "learning_rate": 3.2002486455210337e-06, "loss": 0.6744, "step": 8868 }, { "epoch": 0.6287841191066997, "grad_norm": 3.7869815862357323, "learning_rate": 3.1991774933685273e-06, "loss": 0.3947, "step": 8869 }, { "epoch": 0.6288550159517902, "grad_norm": 2.604884756368431, "learning_rate": 3.1981064361747683e-06, "loss": 0.3795, "step": 8870 }, { "epoch": 0.6289259127968806, "grad_norm": 2.65468893321254, "learning_rate": 3.1970354739962305e-06, "loss": 0.3804, "step": 8871 }, { "epoch": 0.6289968096419709, "grad_norm": 4.062143359112605, "learning_rate": 3.1959646068893873e-06, "loss": 0.4204, "step": 8872 }, { "epoch": 0.6290677064870613, "grad_norm": 1.0661188615992998, "learning_rate": 3.1948938349107095e-06, "loss": 0.6912, "step": 8873 }, { "epoch": 0.6291386033321518, "grad_norm": 3.391340560169466, "learning_rate": 3.193823158116657e-06, "loss": 0.4152, "step": 8874 }, { "epoch": 0.6292095001772421, "grad_norm": 4.537795095862733, "learning_rate": 3.192752576563687e-06, "loss": 0.415, "step": 8875 }, { "epoch": 0.6292803970223325, "grad_norm": 2.7691685726154125, "learning_rate": 3.1916820903082547e-06, "loss": 0.3656, "step": 8876 }, { "epoch": 0.629351293867423, "grad_norm": 4.94936650608029, "learning_rate": 3.1906116994068047e-06, "loss": 0.4181, "step": 8877 }, { "epoch": 0.6294221907125133, "grad_norm": 4.671257360967721, "learning_rate": 3.1895414039157803e-06, "loss": 0.3981, "step": 8878 }, { "epoch": 0.6294930875576037, "grad_norm": 2.1675617312925524, "learning_rate": 3.1884712038916223e-06, "loss": 0.393, "step": 8879 }, { "epoch": 0.6295639844026941, "grad_norm": 2.4683503388039774, "learning_rate": 3.187401099390757e-06, "loss": 0.4033, "step": 8880 }, { "epoch": 0.6296348812477844, "grad_norm": 4.989425576354702, "learning_rate": 3.1863310904696165e-06, "loss": 0.435, "step": 8881 }, { "epoch": 0.6297057780928749, "grad_norm": 6.001910938118061, "learning_rate": 3.185261177184622e-06, "loss": 0.4406, "step": 8882 }, { "epoch": 0.6297766749379653, "grad_norm": 2.0649745418083048, "learning_rate": 3.184191359592192e-06, "loss": 0.409, "step": 8883 }, { "epoch": 0.6298475717830556, "grad_norm": 2.504407765905998, "learning_rate": 3.183121637748736e-06, "loss": 0.3798, "step": 8884 }, { "epoch": 0.629918468628146, "grad_norm": 3.8384971342672447, "learning_rate": 3.1820520117106624e-06, "loss": 0.4587, "step": 8885 }, { "epoch": 0.6299893654732365, "grad_norm": 4.3001001659812506, "learning_rate": 3.1809824815343753e-06, "loss": 0.3703, "step": 8886 }, { "epoch": 0.6300602623183268, "grad_norm": 2.1136771842009883, "learning_rate": 3.1799130472762698e-06, "loss": 0.3757, "step": 8887 }, { "epoch": 0.6301311591634172, "grad_norm": 1.7666174127349548, "learning_rate": 3.1788437089927384e-06, "loss": 0.4248, "step": 8888 }, { "epoch": 0.6302020560085076, "grad_norm": 3.0635725306921944, "learning_rate": 3.177774466740169e-06, "loss": 0.4064, "step": 8889 }, { "epoch": 0.630272952853598, "grad_norm": 2.777372849352696, "learning_rate": 3.176705320574942e-06, "loss": 0.3669, "step": 8890 }, { "epoch": 0.6303438496986884, "grad_norm": 3.645634893473727, "learning_rate": 3.1756362705534337e-06, "loss": 0.4753, "step": 8891 }, { "epoch": 0.6304147465437788, "grad_norm": 4.306721976488494, "learning_rate": 3.1745673167320208e-06, "loss": 0.378, "step": 8892 }, { "epoch": 0.6304856433888691, "grad_norm": 12.197670497720834, "learning_rate": 3.1734984591670635e-06, "loss": 0.3786, "step": 8893 }, { "epoch": 0.6305565402339596, "grad_norm": 6.940934892783418, "learning_rate": 3.1724296979149272e-06, "loss": 0.4355, "step": 8894 }, { "epoch": 0.63062743707905, "grad_norm": 2.1877622386625144, "learning_rate": 3.1713610330319694e-06, "loss": 0.4132, "step": 8895 }, { "epoch": 0.6306983339241404, "grad_norm": 3.659916566192379, "learning_rate": 3.1702924645745388e-06, "loss": 0.3921, "step": 8896 }, { "epoch": 0.6307692307692307, "grad_norm": 4.944757202029228, "learning_rate": 3.169223992598982e-06, "loss": 0.4036, "step": 8897 }, { "epoch": 0.6308401276143212, "grad_norm": 2.547914800890468, "learning_rate": 3.1681556171616425e-06, "loss": 0.3743, "step": 8898 }, { "epoch": 0.6309110244594116, "grad_norm": 4.666556740119561, "learning_rate": 3.167087338318857e-06, "loss": 0.4068, "step": 8899 }, { "epoch": 0.6309819213045019, "grad_norm": 4.438255061205562, "learning_rate": 3.166019156126953e-06, "loss": 0.4076, "step": 8900 }, { "epoch": 0.6310528181495924, "grad_norm": 3.4291247356160866, "learning_rate": 3.1649510706422593e-06, "loss": 0.4047, "step": 8901 }, { "epoch": 0.6311237149946828, "grad_norm": 2.8013537471994363, "learning_rate": 3.1638830819210973e-06, "loss": 0.4093, "step": 8902 }, { "epoch": 0.6311946118397731, "grad_norm": 3.56612950876919, "learning_rate": 3.162815190019781e-06, "loss": 0.3718, "step": 8903 }, { "epoch": 0.6312655086848635, "grad_norm": 6.213049010744029, "learning_rate": 3.1617473949946207e-06, "loss": 0.4061, "step": 8904 }, { "epoch": 0.631336405529954, "grad_norm": 3.870303275766597, "learning_rate": 3.160679696901927e-06, "loss": 0.3564, "step": 8905 }, { "epoch": 0.6314073023750443, "grad_norm": 3.353637323792199, "learning_rate": 3.1596120957979933e-06, "loss": 0.421, "step": 8906 }, { "epoch": 0.6314781992201347, "grad_norm": 4.988545082197253, "learning_rate": 3.1585445917391192e-06, "loss": 0.4311, "step": 8907 }, { "epoch": 0.6315490960652251, "grad_norm": 2.298706801845157, "learning_rate": 3.1574771847815955e-06, "loss": 0.463, "step": 8908 }, { "epoch": 0.6316199929103155, "grad_norm": 3.3653511103340947, "learning_rate": 3.1564098749817053e-06, "loss": 0.4416, "step": 8909 }, { "epoch": 0.6316908897554059, "grad_norm": 4.642525892366475, "learning_rate": 3.1553426623957283e-06, "loss": 0.4024, "step": 8910 }, { "epoch": 0.6317617866004963, "grad_norm": 3.9978169362262217, "learning_rate": 3.1542755470799425e-06, "loss": 0.3775, "step": 8911 }, { "epoch": 0.6318326834455866, "grad_norm": 3.2714147349491354, "learning_rate": 3.153208529090616e-06, "loss": 0.4055, "step": 8912 }, { "epoch": 0.6319035802906771, "grad_norm": 3.5377188443777694, "learning_rate": 3.152141608484013e-06, "loss": 0.3804, "step": 8913 }, { "epoch": 0.6319744771357675, "grad_norm": 2.566870944116224, "learning_rate": 3.1510747853163928e-06, "loss": 0.3632, "step": 8914 }, { "epoch": 0.6320453739808578, "grad_norm": 2.6023559822105358, "learning_rate": 3.1500080596440115e-06, "loss": 0.3895, "step": 8915 }, { "epoch": 0.6321162708259482, "grad_norm": 3.406417414817228, "learning_rate": 3.1489414315231158e-06, "loss": 0.4091, "step": 8916 }, { "epoch": 0.6321871676710387, "grad_norm": 2.925729888460617, "learning_rate": 3.1478749010099512e-06, "loss": 0.3833, "step": 8917 }, { "epoch": 0.632258064516129, "grad_norm": 2.8316595051307556, "learning_rate": 3.146808468160758e-06, "loss": 0.3879, "step": 8918 }, { "epoch": 0.6323289613612194, "grad_norm": 2.92945749824791, "learning_rate": 3.1457421330317685e-06, "loss": 0.3928, "step": 8919 }, { "epoch": 0.6323998582063098, "grad_norm": 3.9292742128755487, "learning_rate": 3.1446758956792112e-06, "loss": 0.4154, "step": 8920 }, { "epoch": 0.6324707550514002, "grad_norm": 10.719266439996755, "learning_rate": 3.143609756159311e-06, "loss": 0.3977, "step": 8921 }, { "epoch": 0.6325416518964906, "grad_norm": 1.7252710655907222, "learning_rate": 3.1425437145282842e-06, "loss": 0.3557, "step": 8922 }, { "epoch": 0.632612548741581, "grad_norm": 4.123940205101547, "learning_rate": 3.1414777708423437e-06, "loss": 0.3669, "step": 8923 }, { "epoch": 0.6326834455866714, "grad_norm": 3.5379553155209003, "learning_rate": 3.1404119251577027e-06, "loss": 0.3239, "step": 8924 }, { "epoch": 0.6327543424317618, "grad_norm": 5.943900186715446, "learning_rate": 3.1393461775305567e-06, "loss": 0.3909, "step": 8925 }, { "epoch": 0.6328252392768522, "grad_norm": 3.839587577266639, "learning_rate": 3.1382805280171086e-06, "loss": 0.338, "step": 8926 }, { "epoch": 0.6328961361219426, "grad_norm": 2.988184592484265, "learning_rate": 3.1372149766735487e-06, "loss": 0.4043, "step": 8927 }, { "epoch": 0.6329670329670329, "grad_norm": 20.81315611617188, "learning_rate": 3.136149523556066e-06, "loss": 0.3969, "step": 8928 }, { "epoch": 0.6330379298121234, "grad_norm": 3.7904409058762245, "learning_rate": 3.1350841687208395e-06, "loss": 0.3805, "step": 8929 }, { "epoch": 0.6331088266572138, "grad_norm": 7.5739909653226025, "learning_rate": 3.1340189122240495e-06, "loss": 0.3922, "step": 8930 }, { "epoch": 0.6331797235023041, "grad_norm": 2.6749928425946186, "learning_rate": 3.132953754121867e-06, "loss": 0.3849, "step": 8931 }, { "epoch": 0.6332506203473945, "grad_norm": 18.50191901244568, "learning_rate": 3.1318886944704572e-06, "loss": 0.3771, "step": 8932 }, { "epoch": 0.633321517192485, "grad_norm": 6.349087498502989, "learning_rate": 3.130823733325983e-06, "loss": 0.4521, "step": 8933 }, { "epoch": 0.6333924140375753, "grad_norm": 4.945804235710883, "learning_rate": 3.1297588707446e-06, "loss": 0.4217, "step": 8934 }, { "epoch": 0.6334633108826657, "grad_norm": 2.986878782103402, "learning_rate": 3.1286941067824583e-06, "loss": 0.3542, "step": 8935 }, { "epoch": 0.6335342077277561, "grad_norm": 2.496857045679403, "learning_rate": 3.1276294414957043e-06, "loss": 0.3589, "step": 8936 }, { "epoch": 0.6336051045728465, "grad_norm": 10.895148478041788, "learning_rate": 3.126564874940482e-06, "loss": 0.3988, "step": 8937 }, { "epoch": 0.6336760014179369, "grad_norm": 10.398081349389814, "learning_rate": 3.1255004071729204e-06, "loss": 0.4322, "step": 8938 }, { "epoch": 0.6337468982630273, "grad_norm": 7.020656301796696, "learning_rate": 3.1244360382491545e-06, "loss": 0.454, "step": 8939 }, { "epoch": 0.6338177951081176, "grad_norm": 0.8880894309464464, "learning_rate": 3.1233717682253075e-06, "loss": 0.6231, "step": 8940 }, { "epoch": 0.6338886919532081, "grad_norm": 4.561584366915096, "learning_rate": 3.1223075971575008e-06, "loss": 0.415, "step": 8941 }, { "epoch": 0.6339595887982985, "grad_norm": 4.13817147284011, "learning_rate": 3.1212435251018452e-06, "loss": 0.3729, "step": 8942 }, { "epoch": 0.6340304856433888, "grad_norm": 4.584550574181111, "learning_rate": 3.120179552114454e-06, "loss": 0.3623, "step": 8943 }, { "epoch": 0.6341013824884792, "grad_norm": 4.410593031281424, "learning_rate": 3.1191156782514305e-06, "loss": 0.4199, "step": 8944 }, { "epoch": 0.6341722793335697, "grad_norm": 6.890605252499741, "learning_rate": 3.1180519035688715e-06, "loss": 0.4001, "step": 8945 }, { "epoch": 0.63424317617866, "grad_norm": 7.480736185544558, "learning_rate": 3.116988228122873e-06, "loss": 0.3989, "step": 8946 }, { "epoch": 0.6343140730237504, "grad_norm": 4.839161924706743, "learning_rate": 3.115924651969523e-06, "loss": 0.3621, "step": 8947 }, { "epoch": 0.6343849698688409, "grad_norm": 5.340100797081629, "learning_rate": 3.1148611751649032e-06, "loss": 0.388, "step": 8948 }, { "epoch": 0.6344558667139313, "grad_norm": 2.3412721882236367, "learning_rate": 3.113797797765091e-06, "loss": 0.3697, "step": 8949 }, { "epoch": 0.6345267635590216, "grad_norm": 6.567082059434663, "learning_rate": 3.112734519826164e-06, "loss": 0.4034, "step": 8950 }, { "epoch": 0.634597660404112, "grad_norm": 4.104536252892669, "learning_rate": 3.1116713414041837e-06, "loss": 0.4182, "step": 8951 }, { "epoch": 0.6346685572492025, "grad_norm": 2.9693145437415978, "learning_rate": 3.1106082625552157e-06, "loss": 0.4399, "step": 8952 }, { "epoch": 0.6347394540942928, "grad_norm": 3.2636821758429537, "learning_rate": 3.1095452833353167e-06, "loss": 0.3976, "step": 8953 }, { "epoch": 0.6348103509393832, "grad_norm": 3.092292719768067, "learning_rate": 3.108482403800537e-06, "loss": 0.3995, "step": 8954 }, { "epoch": 0.6348812477844736, "grad_norm": 2.4925194760785576, "learning_rate": 3.1074196240069234e-06, "loss": 0.4898, "step": 8955 }, { "epoch": 0.634952144629564, "grad_norm": 1.7671107802332093, "learning_rate": 3.1063569440105177e-06, "loss": 0.3933, "step": 8956 }, { "epoch": 0.6350230414746544, "grad_norm": 4.207311796916053, "learning_rate": 3.1052943638673572e-06, "loss": 0.4541, "step": 8957 }, { "epoch": 0.6350939383197448, "grad_norm": 2.2844251839888985, "learning_rate": 3.1042318836334697e-06, "loss": 0.3727, "step": 8958 }, { "epoch": 0.6351648351648351, "grad_norm": 45.0531904028355, "learning_rate": 3.103169503364882e-06, "loss": 0.4022, "step": 8959 }, { "epoch": 0.6352357320099256, "grad_norm": 11.204805594362789, "learning_rate": 3.102107223117616e-06, "loss": 0.3301, "step": 8960 }, { "epoch": 0.635306628855016, "grad_norm": 2.623811754971691, "learning_rate": 3.1010450429476822e-06, "loss": 0.4193, "step": 8961 }, { "epoch": 0.6353775257001063, "grad_norm": 2.0969051917494905, "learning_rate": 3.0999829629110934e-06, "loss": 0.389, "step": 8962 }, { "epoch": 0.6354484225451967, "grad_norm": 5.95320833997059, "learning_rate": 3.0989209830638545e-06, "loss": 0.3705, "step": 8963 }, { "epoch": 0.6355193193902872, "grad_norm": 2.834576642733472, "learning_rate": 3.0978591034619616e-06, "loss": 0.4313, "step": 8964 }, { "epoch": 0.6355902162353775, "grad_norm": 1.9215724356643815, "learning_rate": 3.0967973241614104e-06, "loss": 0.3601, "step": 8965 }, { "epoch": 0.6356611130804679, "grad_norm": 4.466471743476474, "learning_rate": 3.0957356452181896e-06, "loss": 0.3935, "step": 8966 }, { "epoch": 0.6357320099255583, "grad_norm": 4.734633596641022, "learning_rate": 3.094674066688281e-06, "loss": 0.4382, "step": 8967 }, { "epoch": 0.6358029067706487, "grad_norm": 3.15906869193565, "learning_rate": 3.0936125886276623e-06, "loss": 0.3411, "step": 8968 }, { "epoch": 0.6358738036157391, "grad_norm": 2.668677799986045, "learning_rate": 3.0925512110923105e-06, "loss": 0.4081, "step": 8969 }, { "epoch": 0.6359447004608295, "grad_norm": 2.55105612522745, "learning_rate": 3.0914899341381855e-06, "loss": 0.3823, "step": 8970 }, { "epoch": 0.6360155973059198, "grad_norm": 2.6056954358177227, "learning_rate": 3.090428757821254e-06, "loss": 0.4232, "step": 8971 }, { "epoch": 0.6360864941510103, "grad_norm": 2.523449348768366, "learning_rate": 3.0893676821974716e-06, "loss": 0.3567, "step": 8972 }, { "epoch": 0.6361573909961007, "grad_norm": 5.219982224980806, "learning_rate": 3.0883067073227903e-06, "loss": 0.4333, "step": 8973 }, { "epoch": 0.6362282878411911, "grad_norm": 3.869592011339222, "learning_rate": 3.087245833253154e-06, "loss": 0.4139, "step": 8974 }, { "epoch": 0.6362991846862814, "grad_norm": 3.9715882543055123, "learning_rate": 3.0861850600445054e-06, "loss": 0.4087, "step": 8975 }, { "epoch": 0.6363700815313719, "grad_norm": 3.461187974749585, "learning_rate": 3.085124387752781e-06, "loss": 0.3653, "step": 8976 }, { "epoch": 0.6364409783764623, "grad_norm": 3.2914825770791842, "learning_rate": 3.0840638164339077e-06, "loss": 0.41, "step": 8977 }, { "epoch": 0.6365118752215526, "grad_norm": 3.743410675907087, "learning_rate": 3.0830033461438114e-06, "loss": 0.4344, "step": 8978 }, { "epoch": 0.636582772066643, "grad_norm": 3.5193261180909663, "learning_rate": 3.0819429769384135e-06, "loss": 0.3862, "step": 8979 }, { "epoch": 0.6366536689117335, "grad_norm": 4.302007127680192, "learning_rate": 3.0808827088736247e-06, "loss": 0.3741, "step": 8980 }, { "epoch": 0.6367245657568238, "grad_norm": 3.0872171455390656, "learning_rate": 3.0798225420053553e-06, "loss": 0.3831, "step": 8981 }, { "epoch": 0.6367954626019142, "grad_norm": 1.6579030487779083, "learning_rate": 3.078762476389511e-06, "loss": 0.4251, "step": 8982 }, { "epoch": 0.6368663594470046, "grad_norm": 4.320296548487447, "learning_rate": 3.077702512081986e-06, "loss": 0.4057, "step": 8983 }, { "epoch": 0.636937256292095, "grad_norm": 14.506006699506292, "learning_rate": 3.076642649138675e-06, "loss": 0.4038, "step": 8984 }, { "epoch": 0.6370081531371854, "grad_norm": 1.9993610902259378, "learning_rate": 3.0755828876154636e-06, "loss": 0.3673, "step": 8985 }, { "epoch": 0.6370790499822758, "grad_norm": 3.8523438607686327, "learning_rate": 3.074523227568239e-06, "loss": 0.3763, "step": 8986 }, { "epoch": 0.6371499468273661, "grad_norm": 2.47759455458469, "learning_rate": 3.0734636690528715e-06, "loss": 0.3552, "step": 8987 }, { "epoch": 0.6372208436724566, "grad_norm": 3.0840677664422254, "learning_rate": 3.0724042121252356e-06, "loss": 0.3707, "step": 8988 }, { "epoch": 0.637291740517547, "grad_norm": 2.9669151743733413, "learning_rate": 3.071344856841198e-06, "loss": 0.4042, "step": 8989 }, { "epoch": 0.6373626373626373, "grad_norm": 3.5752429437548647, "learning_rate": 3.0702856032566176e-06, "loss": 0.4275, "step": 8990 }, { "epoch": 0.6374335342077277, "grad_norm": 5.847766224074296, "learning_rate": 3.0692264514273494e-06, "loss": 0.4395, "step": 8991 }, { "epoch": 0.6375044310528182, "grad_norm": 2.991334702560793, "learning_rate": 3.0681674014092456e-06, "loss": 0.3742, "step": 8992 }, { "epoch": 0.6375753278979085, "grad_norm": 3.168760708246972, "learning_rate": 3.0671084532581484e-06, "loss": 0.3848, "step": 8993 }, { "epoch": 0.6376462247429989, "grad_norm": 3.7345034972416906, "learning_rate": 3.066049607029897e-06, "loss": 0.3449, "step": 8994 }, { "epoch": 0.6377171215880894, "grad_norm": 4.655341981262078, "learning_rate": 3.064990862780327e-06, "loss": 0.3655, "step": 8995 }, { "epoch": 0.6377880184331797, "grad_norm": 3.423936014759023, "learning_rate": 3.0639322205652653e-06, "loss": 0.4316, "step": 8996 }, { "epoch": 0.6378589152782701, "grad_norm": 2.8046610318086604, "learning_rate": 3.0628736804405357e-06, "loss": 0.3851, "step": 8997 }, { "epoch": 0.6379298121233605, "grad_norm": 2.2221812011172446, "learning_rate": 3.061815242461956e-06, "loss": 0.3748, "step": 8998 }, { "epoch": 0.638000708968451, "grad_norm": 10.425154689280149, "learning_rate": 3.060756906685337e-06, "loss": 0.3881, "step": 8999 }, { "epoch": 0.6380716058135413, "grad_norm": 7.558393273254552, "learning_rate": 3.059698673166486e-06, "loss": 0.4, "step": 9000 }, { "epoch": 0.6381425026586317, "grad_norm": 3.444643857433175, "learning_rate": 3.0586405419612048e-06, "loss": 0.3552, "step": 9001 }, { "epoch": 0.6382133995037221, "grad_norm": 9.217870771364828, "learning_rate": 3.057582513125291e-06, "loss": 0.4635, "step": 9002 }, { "epoch": 0.6382842963488125, "grad_norm": 6.261326171050179, "learning_rate": 3.056524586714533e-06, "loss": 0.4276, "step": 9003 }, { "epoch": 0.6383551931939029, "grad_norm": 4.880379902399091, "learning_rate": 3.055466762784717e-06, "loss": 0.4233, "step": 9004 }, { "epoch": 0.6384260900389933, "grad_norm": 4.887390589930814, "learning_rate": 3.054409041391624e-06, "loss": 0.3803, "step": 9005 }, { "epoch": 0.6384969868840836, "grad_norm": 3.6941062154960482, "learning_rate": 3.053351422591025e-06, "loss": 0.3562, "step": 9006 }, { "epoch": 0.6385678837291741, "grad_norm": 2.4522576691312037, "learning_rate": 3.0522939064386924e-06, "loss": 0.3851, "step": 9007 }, { "epoch": 0.6386387805742645, "grad_norm": 2.5553809699370924, "learning_rate": 3.0512364929903894e-06, "loss": 0.3759, "step": 9008 }, { "epoch": 0.6387096774193548, "grad_norm": 3.173395770026555, "learning_rate": 3.050179182301873e-06, "loss": 0.3561, "step": 9009 }, { "epoch": 0.6387805742644452, "grad_norm": 2.494597369869727, "learning_rate": 3.0491219744288964e-06, "loss": 0.4071, "step": 9010 }, { "epoch": 0.6388514711095357, "grad_norm": 2.9231742809433365, "learning_rate": 3.048064869427208e-06, "loss": 0.4105, "step": 9011 }, { "epoch": 0.638922367954626, "grad_norm": 16.390120872164015, "learning_rate": 3.0470078673525483e-06, "loss": 0.4563, "step": 9012 }, { "epoch": 0.6389932647997164, "grad_norm": 2.238519878699258, "learning_rate": 3.0459509682606525e-06, "loss": 0.435, "step": 9013 }, { "epoch": 0.6390641616448068, "grad_norm": 2.909879024983952, "learning_rate": 3.0448941722072557e-06, "loss": 0.3743, "step": 9014 }, { "epoch": 0.6391350584898972, "grad_norm": 3.946373194410407, "learning_rate": 3.0438374792480816e-06, "loss": 0.445, "step": 9015 }, { "epoch": 0.6392059553349876, "grad_norm": 3.9464846598357055, "learning_rate": 3.0427808894388494e-06, "loss": 0.4108, "step": 9016 }, { "epoch": 0.639276852180078, "grad_norm": 3.0823649732782545, "learning_rate": 3.0417244028352743e-06, "loss": 0.391, "step": 9017 }, { "epoch": 0.6393477490251683, "grad_norm": 3.5378580697588813, "learning_rate": 3.040668019493069e-06, "loss": 0.4218, "step": 9018 }, { "epoch": 0.6394186458702588, "grad_norm": 3.140545654207945, "learning_rate": 3.039611739467931e-06, "loss": 0.3996, "step": 9019 }, { "epoch": 0.6394895427153492, "grad_norm": 4.23783962575225, "learning_rate": 3.038555562815564e-06, "loss": 0.4451, "step": 9020 }, { "epoch": 0.6395604395604395, "grad_norm": 7.620676951186637, "learning_rate": 3.03749948959166e-06, "loss": 0.4062, "step": 9021 }, { "epoch": 0.6396313364055299, "grad_norm": 2.237232924684321, "learning_rate": 3.0364435198519037e-06, "loss": 0.3578, "step": 9022 }, { "epoch": 0.6397022332506204, "grad_norm": 11.40800172628229, "learning_rate": 3.035387653651981e-06, "loss": 0.4345, "step": 9023 }, { "epoch": 0.6397731300957108, "grad_norm": 5.607671465433153, "learning_rate": 3.0343318910475673e-06, "loss": 0.4237, "step": 9024 }, { "epoch": 0.6398440269408011, "grad_norm": 3.032847172386618, "learning_rate": 3.033276232094332e-06, "loss": 0.3986, "step": 9025 }, { "epoch": 0.6399149237858915, "grad_norm": 4.496537076816901, "learning_rate": 3.0322206768479414e-06, "loss": 0.4404, "step": 9026 }, { "epoch": 0.639985820630982, "grad_norm": 3.511801348175991, "learning_rate": 3.031165225364058e-06, "loss": 0.405, "step": 9027 }, { "epoch": 0.6400567174760723, "grad_norm": 3.0263073772710762, "learning_rate": 3.0301098776983344e-06, "loss": 0.4297, "step": 9028 }, { "epoch": 0.6401276143211627, "grad_norm": 1.7620121682999301, "learning_rate": 3.029054633906421e-06, "loss": 0.3951, "step": 9029 }, { "epoch": 0.6401985111662531, "grad_norm": 2.1577992326534914, "learning_rate": 3.027999494043959e-06, "loss": 0.3927, "step": 9030 }, { "epoch": 0.6402694080113435, "grad_norm": 2.69141827824211, "learning_rate": 3.0269444581665926e-06, "loss": 0.4136, "step": 9031 }, { "epoch": 0.6403403048564339, "grad_norm": 3.074264991255804, "learning_rate": 3.0258895263299477e-06, "loss": 0.4282, "step": 9032 }, { "epoch": 0.6404112017015243, "grad_norm": 5.3691005535476215, "learning_rate": 3.0248346985896557e-06, "loss": 0.4132, "step": 9033 }, { "epoch": 0.6404820985466146, "grad_norm": 10.701053450784427, "learning_rate": 3.0237799750013384e-06, "loss": 0.4018, "step": 9034 }, { "epoch": 0.6405529953917051, "grad_norm": 7.969606978949925, "learning_rate": 3.022725355620611e-06, "loss": 0.4082, "step": 9035 }, { "epoch": 0.6406238922367955, "grad_norm": 3.8662376379725054, "learning_rate": 3.021670840503085e-06, "loss": 0.4109, "step": 9036 }, { "epoch": 0.6406947890818858, "grad_norm": 4.130978198472263, "learning_rate": 3.0206164297043656e-06, "loss": 0.3869, "step": 9037 }, { "epoch": 0.6407656859269762, "grad_norm": 2.4298793753911982, "learning_rate": 3.019562123280051e-06, "loss": 0.3961, "step": 9038 }, { "epoch": 0.6408365827720667, "grad_norm": 3.4648928121540012, "learning_rate": 3.0185079212857383e-06, "loss": 0.3911, "step": 9039 }, { "epoch": 0.640907479617157, "grad_norm": 2.805794400313046, "learning_rate": 3.0174538237770166e-06, "loss": 0.433, "step": 9040 }, { "epoch": 0.6409783764622474, "grad_norm": 3.5349196731682357, "learning_rate": 3.0163998308094667e-06, "loss": 0.4575, "step": 9041 }, { "epoch": 0.6410492733073379, "grad_norm": 1.7011243867228243, "learning_rate": 3.015345942438668e-06, "loss": 0.4156, "step": 9042 }, { "epoch": 0.6411201701524282, "grad_norm": 2.794014493596408, "learning_rate": 3.0142921587201913e-06, "loss": 0.3984, "step": 9043 }, { "epoch": 0.6411910669975186, "grad_norm": 3.102068029041847, "learning_rate": 3.0132384797096084e-06, "loss": 0.4026, "step": 9044 }, { "epoch": 0.641261963842609, "grad_norm": 13.573730776484352, "learning_rate": 3.012184905462473e-06, "loss": 0.494, "step": 9045 }, { "epoch": 0.6413328606876993, "grad_norm": 2.0107175534400175, "learning_rate": 3.011131436034346e-06, "loss": 0.4143, "step": 9046 }, { "epoch": 0.6414037575327898, "grad_norm": 3.121474021091634, "learning_rate": 3.0100780714807776e-06, "loss": 0.4058, "step": 9047 }, { "epoch": 0.6414746543778802, "grad_norm": 9.72326556086572, "learning_rate": 3.0090248118573094e-06, "loss": 0.4017, "step": 9048 }, { "epoch": 0.6415455512229705, "grad_norm": 4.195597963643243, "learning_rate": 3.007971657219484e-06, "loss": 0.4293, "step": 9049 }, { "epoch": 0.641616448068061, "grad_norm": 2.852485559336189, "learning_rate": 3.0069186076228333e-06, "loss": 0.4054, "step": 9050 }, { "epoch": 0.6416873449131514, "grad_norm": 2.7717099249547443, "learning_rate": 3.0058656631228845e-06, "loss": 0.3992, "step": 9051 }, { "epoch": 0.6417582417582418, "grad_norm": 2.961027398720073, "learning_rate": 3.004812823775162e-06, "loss": 0.4348, "step": 9052 }, { "epoch": 0.6418291386033321, "grad_norm": 3.19004181257303, "learning_rate": 3.003760089635183e-06, "loss": 0.3885, "step": 9053 }, { "epoch": 0.6419000354484226, "grad_norm": 6.790628671004153, "learning_rate": 3.002707460758457e-06, "loss": 0.3811, "step": 9054 }, { "epoch": 0.641970932293513, "grad_norm": 2.5014091528297007, "learning_rate": 3.001654937200492e-06, "loss": 0.4437, "step": 9055 }, { "epoch": 0.6420418291386033, "grad_norm": 9.954545264656131, "learning_rate": 3.0006025190167874e-06, "loss": 0.3959, "step": 9056 }, { "epoch": 0.6421127259836937, "grad_norm": 3.4609642486751175, "learning_rate": 2.9995502062628378e-06, "loss": 0.4058, "step": 9057 }, { "epoch": 0.6421836228287842, "grad_norm": 2.750940528892236, "learning_rate": 2.9984979989941316e-06, "loss": 0.4445, "step": 9058 }, { "epoch": 0.6422545196738745, "grad_norm": 3.75088015448877, "learning_rate": 2.9974458972661545e-06, "loss": 0.3709, "step": 9059 }, { "epoch": 0.6423254165189649, "grad_norm": 41.7655413437757, "learning_rate": 2.9963939011343848e-06, "loss": 0.4028, "step": 9060 }, { "epoch": 0.6423963133640553, "grad_norm": 2.3982538924579977, "learning_rate": 2.9953420106542924e-06, "loss": 0.4306, "step": 9061 }, { "epoch": 0.6424672102091457, "grad_norm": 3.8537342489397215, "learning_rate": 2.9942902258813457e-06, "loss": 0.401, "step": 9062 }, { "epoch": 0.6425381070542361, "grad_norm": 3.2929472670862623, "learning_rate": 2.993238546871008e-06, "loss": 0.4235, "step": 9063 }, { "epoch": 0.6426090038993265, "grad_norm": 3.912260498334874, "learning_rate": 2.9921869736787313e-06, "loss": 0.508, "step": 9064 }, { "epoch": 0.6426799007444168, "grad_norm": 9.559640953070902, "learning_rate": 2.991135506359969e-06, "loss": 0.3697, "step": 9065 }, { "epoch": 0.6427507975895073, "grad_norm": 1.8878309687458181, "learning_rate": 2.9900841449701658e-06, "loss": 0.4281, "step": 9066 }, { "epoch": 0.6428216944345977, "grad_norm": 4.5751876144519334, "learning_rate": 2.9890328895647587e-06, "loss": 0.3826, "step": 9067 }, { "epoch": 0.642892591279688, "grad_norm": 3.038547963331694, "learning_rate": 2.9879817401991818e-06, "loss": 0.4047, "step": 9068 }, { "epoch": 0.6429634881247784, "grad_norm": 3.6391473985867777, "learning_rate": 2.986930696928865e-06, "loss": 0.4105, "step": 9069 }, { "epoch": 0.6430343849698689, "grad_norm": 3.0643046892322148, "learning_rate": 2.985879759809227e-06, "loss": 0.4284, "step": 9070 }, { "epoch": 0.6431052818149592, "grad_norm": 3.086700385939654, "learning_rate": 2.984828928895687e-06, "loss": 0.4395, "step": 9071 }, { "epoch": 0.6431761786600496, "grad_norm": 3.056462892722413, "learning_rate": 2.9837782042436556e-06, "loss": 0.364, "step": 9072 }, { "epoch": 0.64324707550514, "grad_norm": 3.9613905468254234, "learning_rate": 2.98272758590854e-06, "loss": 0.3682, "step": 9073 }, { "epoch": 0.6433179723502304, "grad_norm": 3.0800222761732603, "learning_rate": 2.981677073945738e-06, "loss": 0.3935, "step": 9074 }, { "epoch": 0.6433888691953208, "grad_norm": 4.332051185153716, "learning_rate": 2.980626668410643e-06, "loss": 0.409, "step": 9075 }, { "epoch": 0.6434597660404112, "grad_norm": 1.0578252920133082, "learning_rate": 2.9795763693586477e-06, "loss": 0.6897, "step": 9076 }, { "epoch": 0.6435306628855016, "grad_norm": 6.081791883703699, "learning_rate": 2.9785261768451297e-06, "loss": 0.4252, "step": 9077 }, { "epoch": 0.643601559730592, "grad_norm": 6.233840667947866, "learning_rate": 2.97747609092547e-06, "loss": 0.4491, "step": 9078 }, { "epoch": 0.6436724565756824, "grad_norm": 2.0599624020488974, "learning_rate": 2.9764261116550414e-06, "loss": 0.3727, "step": 9079 }, { "epoch": 0.6437433534207728, "grad_norm": 2.8979548208205923, "learning_rate": 2.975376239089207e-06, "loss": 0.4287, "step": 9080 }, { "epoch": 0.6438142502658631, "grad_norm": 2.852408978123809, "learning_rate": 2.974326473283329e-06, "loss": 0.4184, "step": 9081 }, { "epoch": 0.6438851471109536, "grad_norm": 1.0920774990238775, "learning_rate": 2.9732768142927637e-06, "loss": 0.7057, "step": 9082 }, { "epoch": 0.643956043956044, "grad_norm": 2.6438127935627262, "learning_rate": 2.972227262172856e-06, "loss": 0.3621, "step": 9083 }, { "epoch": 0.6440269408011343, "grad_norm": 1.1067345831993731, "learning_rate": 2.971177816978954e-06, "loss": 0.6573, "step": 9084 }, { "epoch": 0.6440978376462247, "grad_norm": 3.620283223068823, "learning_rate": 2.9701284787663953e-06, "loss": 0.4034, "step": 9085 }, { "epoch": 0.6441687344913152, "grad_norm": 1.990570732032317, "learning_rate": 2.96907924759051e-06, "loss": 0.4151, "step": 9086 }, { "epoch": 0.6442396313364055, "grad_norm": 7.164255448406742, "learning_rate": 2.968030123506626e-06, "loss": 0.4796, "step": 9087 }, { "epoch": 0.6443105281814959, "grad_norm": 2.8293411323575515, "learning_rate": 2.9669811065700638e-06, "loss": 0.4084, "step": 9088 }, { "epoch": 0.6443814250265864, "grad_norm": 3.4710076290554017, "learning_rate": 2.965932196836143e-06, "loss": 0.4299, "step": 9089 }, { "epoch": 0.6444523218716767, "grad_norm": 4.123571974979717, "learning_rate": 2.9648833943601663e-06, "loss": 0.421, "step": 9090 }, { "epoch": 0.6445232187167671, "grad_norm": 2.6608953021100157, "learning_rate": 2.963834699197442e-06, "loss": 0.424, "step": 9091 }, { "epoch": 0.6445941155618575, "grad_norm": 2.617830634489653, "learning_rate": 2.9627861114032702e-06, "loss": 0.4645, "step": 9092 }, { "epoch": 0.6446650124069478, "grad_norm": 8.210977438323427, "learning_rate": 2.9617376310329405e-06, "loss": 0.3903, "step": 9093 }, { "epoch": 0.6447359092520383, "grad_norm": 4.3672189085593685, "learning_rate": 2.96068925814174e-06, "loss": 0.3401, "step": 9094 }, { "epoch": 0.6448068060971287, "grad_norm": 4.199491046890729, "learning_rate": 2.9596409927849544e-06, "loss": 0.3703, "step": 9095 }, { "epoch": 0.644877702942219, "grad_norm": 4.182430017625867, "learning_rate": 2.958592835017854e-06, "loss": 0.38, "step": 9096 }, { "epoch": 0.6449485997873095, "grad_norm": 0.9744448280667826, "learning_rate": 2.9575447848957122e-06, "loss": 0.654, "step": 9097 }, { "epoch": 0.6450194966323999, "grad_norm": 3.7874230732073255, "learning_rate": 2.956496842473794e-06, "loss": 0.3732, "step": 9098 }, { "epoch": 0.6450903934774902, "grad_norm": 3.042493287018864, "learning_rate": 2.955449007807356e-06, "loss": 0.3936, "step": 9099 }, { "epoch": 0.6451612903225806, "grad_norm": 3.993106326814443, "learning_rate": 2.954401280951652e-06, "loss": 0.4198, "step": 9100 }, { "epoch": 0.6452321871676711, "grad_norm": 2.3683125113541164, "learning_rate": 2.953353661961932e-06, "loss": 0.3612, "step": 9101 }, { "epoch": 0.6453030840127615, "grad_norm": 2.2712235441091306, "learning_rate": 2.952306150893433e-06, "loss": 0.3931, "step": 9102 }, { "epoch": 0.6453739808578518, "grad_norm": 3.7177325404893096, "learning_rate": 2.951258747801393e-06, "loss": 0.4095, "step": 9103 }, { "epoch": 0.6454448777029422, "grad_norm": 36.68402200179178, "learning_rate": 2.9502114527410443e-06, "loss": 0.3948, "step": 9104 }, { "epoch": 0.6455157745480327, "grad_norm": 2.3856402932303937, "learning_rate": 2.9491642657676113e-06, "loss": 0.3645, "step": 9105 }, { "epoch": 0.645586671393123, "grad_norm": 3.109176112962547, "learning_rate": 2.948117186936311e-06, "loss": 0.3961, "step": 9106 }, { "epoch": 0.6456575682382134, "grad_norm": 3.270511082267422, "learning_rate": 2.9470702163023557e-06, "loss": 0.3661, "step": 9107 }, { "epoch": 0.6457284650833038, "grad_norm": 3.190640067693148, "learning_rate": 2.9460233539209588e-06, "loss": 0.4748, "step": 9108 }, { "epoch": 0.6457993619283942, "grad_norm": 2.2897426285104934, "learning_rate": 2.9449765998473147e-06, "loss": 0.383, "step": 9109 }, { "epoch": 0.6458702587734846, "grad_norm": 7.901061592920957, "learning_rate": 2.9439299541366238e-06, "loss": 0.3929, "step": 9110 }, { "epoch": 0.645941155618575, "grad_norm": 13.827118125328843, "learning_rate": 2.942883416844078e-06, "loss": 0.3473, "step": 9111 }, { "epoch": 0.6460120524636653, "grad_norm": 2.6437170699445707, "learning_rate": 2.9418369880248577e-06, "loss": 0.3838, "step": 9112 }, { "epoch": 0.6460829493087558, "grad_norm": 3.730044712850997, "learning_rate": 2.9407906677341434e-06, "loss": 0.3763, "step": 9113 }, { "epoch": 0.6461538461538462, "grad_norm": 3.502416830547675, "learning_rate": 2.9397444560271105e-06, "loss": 0.4115, "step": 9114 }, { "epoch": 0.6462247429989365, "grad_norm": 4.972375279428144, "learning_rate": 2.938698352958923e-06, "loss": 0.4017, "step": 9115 }, { "epoch": 0.6462956398440269, "grad_norm": 4.013129286107946, "learning_rate": 2.9376523585847463e-06, "loss": 0.3964, "step": 9116 }, { "epoch": 0.6463665366891174, "grad_norm": 4.942408040658475, "learning_rate": 2.9366064729597342e-06, "loss": 0.4338, "step": 9117 }, { "epoch": 0.6464374335342077, "grad_norm": 2.77563921816087, "learning_rate": 2.935560696139038e-06, "loss": 0.3563, "step": 9118 }, { "epoch": 0.6465083303792981, "grad_norm": 3.133814751822861, "learning_rate": 2.9345150281778023e-06, "loss": 0.4343, "step": 9119 }, { "epoch": 0.6465792272243885, "grad_norm": 2.662957890162864, "learning_rate": 2.933469469131164e-06, "loss": 0.4101, "step": 9120 }, { "epoch": 0.6466501240694789, "grad_norm": 3.937914975953112, "learning_rate": 2.9324240190542606e-06, "loss": 0.4133, "step": 9121 }, { "epoch": 0.6467210209145693, "grad_norm": 3.588187185765087, "learning_rate": 2.9313786780022145e-06, "loss": 0.3891, "step": 9122 }, { "epoch": 0.6467919177596597, "grad_norm": 3.2290515149189165, "learning_rate": 2.9303334460301503e-06, "loss": 0.3993, "step": 9123 }, { "epoch": 0.64686281460475, "grad_norm": 1.167195611933436, "learning_rate": 2.9292883231931847e-06, "loss": 0.7064, "step": 9124 }, { "epoch": 0.6469337114498405, "grad_norm": 1.9934628794551121, "learning_rate": 2.928243309546425e-06, "loss": 0.3796, "step": 9125 }, { "epoch": 0.6470046082949309, "grad_norm": 22.062070776553913, "learning_rate": 2.9271984051449753e-06, "loss": 0.3896, "step": 9126 }, { "epoch": 0.6470755051400213, "grad_norm": 3.7069690189376656, "learning_rate": 2.926153610043939e-06, "loss": 0.3228, "step": 9127 }, { "epoch": 0.6471464019851116, "grad_norm": 2.436408012496581, "learning_rate": 2.9251089242984033e-06, "loss": 0.4467, "step": 9128 }, { "epoch": 0.6472172988302021, "grad_norm": 2.6829046383627024, "learning_rate": 2.9240643479634585e-06, "loss": 0.3615, "step": 9129 }, { "epoch": 0.6472881956752925, "grad_norm": 3.425787436036893, "learning_rate": 2.9230198810941856e-06, "loss": 0.4124, "step": 9130 }, { "epoch": 0.6473590925203828, "grad_norm": 3.492489112984726, "learning_rate": 2.921975523745659e-06, "loss": 0.3513, "step": 9131 }, { "epoch": 0.6474299893654732, "grad_norm": 4.643178421188256, "learning_rate": 2.92093127597295e-06, "loss": 0.3992, "step": 9132 }, { "epoch": 0.6475008862105637, "grad_norm": 8.885008166451993, "learning_rate": 2.9198871378311193e-06, "loss": 0.3941, "step": 9133 }, { "epoch": 0.647571783055654, "grad_norm": 3.1180570107734074, "learning_rate": 2.91884310937523e-06, "loss": 0.3834, "step": 9134 }, { "epoch": 0.6476426799007444, "grad_norm": 3.180939985495392, "learning_rate": 2.917799190660332e-06, "loss": 0.4208, "step": 9135 }, { "epoch": 0.6477135767458349, "grad_norm": 2.7999470951066217, "learning_rate": 2.9167553817414706e-06, "loss": 0.4047, "step": 9136 }, { "epoch": 0.6477844735909252, "grad_norm": 0.9315792071319541, "learning_rate": 2.915711682673689e-06, "loss": 0.6474, "step": 9137 }, { "epoch": 0.6478553704360156, "grad_norm": 4.426222397932618, "learning_rate": 2.9146680935120196e-06, "loss": 0.428, "step": 9138 }, { "epoch": 0.647926267281106, "grad_norm": 4.209731782997777, "learning_rate": 2.9136246143114956e-06, "loss": 0.3779, "step": 9139 }, { "epoch": 0.6479971641261963, "grad_norm": 2.1215043252041883, "learning_rate": 2.912581245127138e-06, "loss": 0.3802, "step": 9140 }, { "epoch": 0.6480680609712868, "grad_norm": 3.340413966392649, "learning_rate": 2.9115379860139624e-06, "loss": 0.356, "step": 9141 }, { "epoch": 0.6481389578163772, "grad_norm": 11.922643542609919, "learning_rate": 2.9104948370269832e-06, "loss": 0.4453, "step": 9142 }, { "epoch": 0.6482098546614675, "grad_norm": 1.8559739112715326, "learning_rate": 2.90945179822121e-06, "loss": 0.4331, "step": 9143 }, { "epoch": 0.648280751506558, "grad_norm": 4.985088594188475, "learning_rate": 2.9084088696516346e-06, "loss": 0.4278, "step": 9144 }, { "epoch": 0.6483516483516484, "grad_norm": 2.6725175045688663, "learning_rate": 2.907366051373256e-06, "loss": 0.466, "step": 9145 }, { "epoch": 0.6484225451967387, "grad_norm": 4.936744436447393, "learning_rate": 2.9063233434410654e-06, "loss": 0.3899, "step": 9146 }, { "epoch": 0.6484934420418291, "grad_norm": 2.350223104089575, "learning_rate": 2.9052807459100428e-06, "loss": 0.3983, "step": 9147 }, { "epoch": 0.6485643388869196, "grad_norm": 2.1254381631433397, "learning_rate": 2.9042382588351624e-06, "loss": 0.3897, "step": 9148 }, { "epoch": 0.6486352357320099, "grad_norm": 2.968560030836939, "learning_rate": 2.903195882271399e-06, "loss": 0.3302, "step": 9149 }, { "epoch": 0.6487061325771003, "grad_norm": 2.5365684470888623, "learning_rate": 2.902153616273722e-06, "loss": 0.3718, "step": 9150 }, { "epoch": 0.6487770294221907, "grad_norm": 3.189284071701203, "learning_rate": 2.9011114608970804e-06, "loss": 0.4008, "step": 9151 }, { "epoch": 0.6488479262672812, "grad_norm": 2.8128109227166074, "learning_rate": 2.900069416196434e-06, "loss": 0.3556, "step": 9152 }, { "epoch": 0.6489188231123715, "grad_norm": 8.386448802525097, "learning_rate": 2.899027482226733e-06, "loss": 0.4154, "step": 9153 }, { "epoch": 0.6489897199574619, "grad_norm": 4.056861588880337, "learning_rate": 2.8979856590429156e-06, "loss": 0.3558, "step": 9154 }, { "epoch": 0.6490606168025523, "grad_norm": 3.299233753882249, "learning_rate": 2.8969439466999166e-06, "loss": 0.3956, "step": 9155 }, { "epoch": 0.6491315136476427, "grad_norm": 6.098383889169015, "learning_rate": 2.895902345252671e-06, "loss": 0.3843, "step": 9156 }, { "epoch": 0.6492024104927331, "grad_norm": 2.447779519437115, "learning_rate": 2.8948608547560986e-06, "loss": 0.3604, "step": 9157 }, { "epoch": 0.6492733073378235, "grad_norm": 2.265242228472219, "learning_rate": 2.8938194752651215e-06, "loss": 0.3975, "step": 9158 }, { "epoch": 0.6493442041829138, "grad_norm": 3.8405960919288624, "learning_rate": 2.8927782068346514e-06, "loss": 0.3525, "step": 9159 }, { "epoch": 0.6494151010280043, "grad_norm": 4.265464567954636, "learning_rate": 2.8917370495195927e-06, "loss": 0.3798, "step": 9160 }, { "epoch": 0.6494859978730947, "grad_norm": 0.9061923819032801, "learning_rate": 2.8906960033748516e-06, "loss": 0.6485, "step": 9161 }, { "epoch": 0.649556894718185, "grad_norm": 4.954006681200267, "learning_rate": 2.889655068455317e-06, "loss": 0.3914, "step": 9162 }, { "epoch": 0.6496277915632754, "grad_norm": 1.8339804799474824, "learning_rate": 2.888614244815884e-06, "loss": 0.364, "step": 9163 }, { "epoch": 0.6496986884083659, "grad_norm": 4.388753978369117, "learning_rate": 2.887573532511431e-06, "loss": 0.4109, "step": 9164 }, { "epoch": 0.6497695852534562, "grad_norm": 5.60533121556441, "learning_rate": 2.8865329315968405e-06, "loss": 0.4035, "step": 9165 }, { "epoch": 0.6498404820985466, "grad_norm": 2.7149357546437076, "learning_rate": 2.8854924421269813e-06, "loss": 0.3718, "step": 9166 }, { "epoch": 0.649911378943637, "grad_norm": 28.419354678076818, "learning_rate": 2.8844520641567174e-06, "loss": 0.3817, "step": 9167 }, { "epoch": 0.6499822757887274, "grad_norm": 2.6160551789708166, "learning_rate": 2.8834117977409117e-06, "loss": 0.3836, "step": 9168 }, { "epoch": 0.6500531726338178, "grad_norm": 3.111471375336957, "learning_rate": 2.8823716429344205e-06, "loss": 0.3755, "step": 9169 }, { "epoch": 0.6501240694789082, "grad_norm": 2.2697817496498987, "learning_rate": 2.8813315997920853e-06, "loss": 0.4128, "step": 9170 }, { "epoch": 0.6501949663239985, "grad_norm": 1.4800916380081528, "learning_rate": 2.880291668368752e-06, "loss": 0.4642, "step": 9171 }, { "epoch": 0.650265863169089, "grad_norm": 2.7979466949308653, "learning_rate": 2.879251848719259e-06, "loss": 0.4042, "step": 9172 }, { "epoch": 0.6503367600141794, "grad_norm": 1.7118984150892333, "learning_rate": 2.878212140898434e-06, "loss": 0.3299, "step": 9173 }, { "epoch": 0.6504076568592697, "grad_norm": 2.1572450289925773, "learning_rate": 2.877172544961101e-06, "loss": 0.392, "step": 9174 }, { "epoch": 0.6504785537043601, "grad_norm": 2.093279906346017, "learning_rate": 2.87613306096208e-06, "loss": 0.4223, "step": 9175 }, { "epoch": 0.6505494505494506, "grad_norm": 5.0346634220464574, "learning_rate": 2.875093688956185e-06, "loss": 0.3785, "step": 9176 }, { "epoch": 0.6506203473945409, "grad_norm": 3.407540704059891, "learning_rate": 2.874054428998222e-06, "loss": 0.3802, "step": 9177 }, { "epoch": 0.6506912442396313, "grad_norm": 7.6487646489029935, "learning_rate": 2.8730152811429903e-06, "loss": 0.3518, "step": 9178 }, { "epoch": 0.6507621410847217, "grad_norm": 2.7408974782399733, "learning_rate": 2.8719762454452877e-06, "loss": 0.4015, "step": 9179 }, { "epoch": 0.6508330379298122, "grad_norm": 2.7009094305737524, "learning_rate": 2.8709373219599023e-06, "loss": 0.4211, "step": 9180 }, { "epoch": 0.6509039347749025, "grad_norm": 3.030065273983999, "learning_rate": 2.8698985107416143e-06, "loss": 0.4624, "step": 9181 }, { "epoch": 0.6509748316199929, "grad_norm": 3.2457132823078876, "learning_rate": 2.868859811845206e-06, "loss": 0.374, "step": 9182 }, { "epoch": 0.6510457284650834, "grad_norm": 2.7646551986982013, "learning_rate": 2.8678212253254446e-06, "loss": 0.4191, "step": 9183 }, { "epoch": 0.6511166253101737, "grad_norm": 2.7656465897257596, "learning_rate": 2.8667827512370994e-06, "loss": 0.4048, "step": 9184 }, { "epoch": 0.6511875221552641, "grad_norm": 2.0127310973417862, "learning_rate": 2.8657443896349278e-06, "loss": 0.4115, "step": 9185 }, { "epoch": 0.6512584190003545, "grad_norm": 2.9722255117046035, "learning_rate": 2.8647061405736816e-06, "loss": 0.4155, "step": 9186 }, { "epoch": 0.6513293158454448, "grad_norm": 2.4142602620758393, "learning_rate": 2.86366800410811e-06, "loss": 0.3826, "step": 9187 }, { "epoch": 0.6514002126905353, "grad_norm": 7.203719347366485, "learning_rate": 2.8626299802929597e-06, "loss": 0.3959, "step": 9188 }, { "epoch": 0.6514711095356257, "grad_norm": 2.3786250191603173, "learning_rate": 2.861592069182958e-06, "loss": 0.4142, "step": 9189 }, { "epoch": 0.651542006380716, "grad_norm": 5.776102407566139, "learning_rate": 2.860554270832838e-06, "loss": 0.3371, "step": 9190 }, { "epoch": 0.6516129032258065, "grad_norm": 3.4364359325996494, "learning_rate": 2.8595165852973266e-06, "loss": 0.4022, "step": 9191 }, { "epoch": 0.6516838000708969, "grad_norm": 2.3938169656678605, "learning_rate": 2.8584790126311394e-06, "loss": 0.411, "step": 9192 }, { "epoch": 0.6517546969159872, "grad_norm": 3.117571278694873, "learning_rate": 2.8574415528889864e-06, "loss": 0.4396, "step": 9193 }, { "epoch": 0.6518255937610776, "grad_norm": 3.143498332508449, "learning_rate": 2.8564042061255757e-06, "loss": 0.3931, "step": 9194 }, { "epoch": 0.651896490606168, "grad_norm": 3.625453112296623, "learning_rate": 2.8553669723956095e-06, "loss": 0.3924, "step": 9195 }, { "epoch": 0.6519673874512584, "grad_norm": 2.801043471421096, "learning_rate": 2.85432985175378e-06, "loss": 0.3262, "step": 9196 }, { "epoch": 0.6520382842963488, "grad_norm": 2.056409267689862, "learning_rate": 2.8532928442547724e-06, "loss": 0.3993, "step": 9197 }, { "epoch": 0.6521091811414392, "grad_norm": 4.074717928077819, "learning_rate": 2.852255949953274e-06, "loss": 0.4201, "step": 9198 }, { "epoch": 0.6521800779865295, "grad_norm": 3.7090311214552054, "learning_rate": 2.8512191689039586e-06, "loss": 0.4057, "step": 9199 }, { "epoch": 0.65225097483162, "grad_norm": 2.9071688801450586, "learning_rate": 2.8501825011614947e-06, "loss": 0.397, "step": 9200 }, { "epoch": 0.6523218716767104, "grad_norm": 2.4797640845404847, "learning_rate": 2.8491459467805506e-06, "loss": 0.3859, "step": 9201 }, { "epoch": 0.6523927685218007, "grad_norm": 4.814420909376624, "learning_rate": 2.84810950581578e-06, "loss": 0.4498, "step": 9202 }, { "epoch": 0.6524636653668912, "grad_norm": 3.6207446349655616, "learning_rate": 2.8470731783218396e-06, "loss": 0.3514, "step": 9203 }, { "epoch": 0.6525345622119816, "grad_norm": 5.909649052969266, "learning_rate": 2.8460369643533734e-06, "loss": 0.3784, "step": 9204 }, { "epoch": 0.652605459057072, "grad_norm": 6.244414840826891, "learning_rate": 2.8450008639650205e-06, "loss": 0.3816, "step": 9205 }, { "epoch": 0.6526763559021623, "grad_norm": 14.581211804843601, "learning_rate": 2.843964877211419e-06, "loss": 0.4045, "step": 9206 }, { "epoch": 0.6527472527472528, "grad_norm": 2.736383413789721, "learning_rate": 2.8429290041471937e-06, "loss": 0.3975, "step": 9207 }, { "epoch": 0.6528181495923432, "grad_norm": 6.543134429955491, "learning_rate": 2.84189324482697e-06, "loss": 0.4023, "step": 9208 }, { "epoch": 0.6528890464374335, "grad_norm": 3.2306611908501295, "learning_rate": 2.84085759930536e-06, "loss": 0.4467, "step": 9209 }, { "epoch": 0.6529599432825239, "grad_norm": 2.0012836279594115, "learning_rate": 2.8398220676369794e-06, "loss": 0.4065, "step": 9210 }, { "epoch": 0.6530308401276144, "grad_norm": 2.8707308791016106, "learning_rate": 2.83878664987643e-06, "loss": 0.3994, "step": 9211 }, { "epoch": 0.6531017369727047, "grad_norm": 4.124823251137532, "learning_rate": 2.837751346078308e-06, "loss": 0.3667, "step": 9212 }, { "epoch": 0.6531726338177951, "grad_norm": 3.028954959199864, "learning_rate": 2.836716156297208e-06, "loss": 0.3999, "step": 9213 }, { "epoch": 0.6532435306628855, "grad_norm": 3.122883275544985, "learning_rate": 2.83568108058772e-06, "loss": 0.3756, "step": 9214 }, { "epoch": 0.6533144275079759, "grad_norm": 3.4066592518095717, "learning_rate": 2.8346461190044165e-06, "loss": 0.4405, "step": 9215 }, { "epoch": 0.6533853243530663, "grad_norm": 4.94906829422244, "learning_rate": 2.833611271601876e-06, "loss": 0.3983, "step": 9216 }, { "epoch": 0.6534562211981567, "grad_norm": 17.794436599701925, "learning_rate": 2.8325765384346694e-06, "loss": 0.3672, "step": 9217 }, { "epoch": 0.653527118043247, "grad_norm": 8.175841354606021, "learning_rate": 2.8315419195573553e-06, "loss": 0.3415, "step": 9218 }, { "epoch": 0.6535980148883375, "grad_norm": 3.0183878048771957, "learning_rate": 2.830507415024489e-06, "loss": 0.4177, "step": 9219 }, { "epoch": 0.6536689117334279, "grad_norm": 6.080442682198868, "learning_rate": 2.829473024890623e-06, "loss": 0.3476, "step": 9220 }, { "epoch": 0.6537398085785182, "grad_norm": 3.9607871982327665, "learning_rate": 2.8284387492103028e-06, "loss": 0.375, "step": 9221 }, { "epoch": 0.6538107054236086, "grad_norm": 2.2537474587571196, "learning_rate": 2.827404588038065e-06, "loss": 0.4346, "step": 9222 }, { "epoch": 0.6538816022686991, "grad_norm": 5.075964174966043, "learning_rate": 2.8263705414284393e-06, "loss": 0.4232, "step": 9223 }, { "epoch": 0.6539524991137894, "grad_norm": 10.229887273524701, "learning_rate": 2.8253366094359564e-06, "loss": 0.3679, "step": 9224 }, { "epoch": 0.6540233959588798, "grad_norm": 26.54759826999521, "learning_rate": 2.8243027921151336e-06, "loss": 0.4034, "step": 9225 }, { "epoch": 0.6540942928039702, "grad_norm": 2.8701486352976646, "learning_rate": 2.8232690895204832e-06, "loss": 0.4154, "step": 9226 }, { "epoch": 0.6541651896490606, "grad_norm": 2.9189685840480077, "learning_rate": 2.822235501706517e-06, "loss": 0.3915, "step": 9227 }, { "epoch": 0.654236086494151, "grad_norm": 4.66812747032081, "learning_rate": 2.821202028727734e-06, "loss": 0.4254, "step": 9228 }, { "epoch": 0.6543069833392414, "grad_norm": 2.116245657609992, "learning_rate": 2.820168670638632e-06, "loss": 0.371, "step": 9229 }, { "epoch": 0.6543778801843319, "grad_norm": 2.840860221617529, "learning_rate": 2.8191354274937e-06, "loss": 0.4723, "step": 9230 }, { "epoch": 0.6544487770294222, "grad_norm": 3.0755805733064316, "learning_rate": 2.8181022993474196e-06, "loss": 0.3887, "step": 9231 }, { "epoch": 0.6545196738745126, "grad_norm": 4.149335418628888, "learning_rate": 2.8170692862542703e-06, "loss": 0.4057, "step": 9232 }, { "epoch": 0.654590570719603, "grad_norm": 3.855970826502491, "learning_rate": 2.8160363882687276e-06, "loss": 0.4375, "step": 9233 }, { "epoch": 0.6546614675646933, "grad_norm": 2.099215967048427, "learning_rate": 2.815003605445249e-06, "loss": 0.3677, "step": 9234 }, { "epoch": 0.6547323644097838, "grad_norm": 4.158338908562258, "learning_rate": 2.8139709378382983e-06, "loss": 0.4124, "step": 9235 }, { "epoch": 0.6548032612548742, "grad_norm": 6.104196631872696, "learning_rate": 2.8129383855023308e-06, "loss": 0.4217, "step": 9236 }, { "epoch": 0.6548741580999645, "grad_norm": 1.8457832290956846, "learning_rate": 2.811905948491791e-06, "loss": 0.4171, "step": 9237 }, { "epoch": 0.654945054945055, "grad_norm": 7.466326563685787, "learning_rate": 2.8108736268611186e-06, "loss": 0.3881, "step": 9238 }, { "epoch": 0.6550159517901454, "grad_norm": 8.56099462044793, "learning_rate": 2.809841420664751e-06, "loss": 0.4714, "step": 9239 }, { "epoch": 0.6550868486352357, "grad_norm": 2.170040603646778, "learning_rate": 2.808809329957119e-06, "loss": 0.3683, "step": 9240 }, { "epoch": 0.6551577454803261, "grad_norm": 2.7571011633022855, "learning_rate": 2.8077773547926435e-06, "loss": 0.4144, "step": 9241 }, { "epoch": 0.6552286423254166, "grad_norm": 2.0719045641575544, "learning_rate": 2.806745495225739e-06, "loss": 0.394, "step": 9242 }, { "epoch": 0.6552995391705069, "grad_norm": 2.433033194013758, "learning_rate": 2.8057137513108214e-06, "loss": 0.39, "step": 9243 }, { "epoch": 0.6553704360155973, "grad_norm": 2.499222342332951, "learning_rate": 2.8046821231022916e-06, "loss": 0.4279, "step": 9244 }, { "epoch": 0.6554413328606877, "grad_norm": 3.386561661721885, "learning_rate": 2.803650610654547e-06, "loss": 0.4412, "step": 9245 }, { "epoch": 0.655512229705778, "grad_norm": 6.130899147989895, "learning_rate": 2.8026192140219844e-06, "loss": 0.3691, "step": 9246 }, { "epoch": 0.6555831265508685, "grad_norm": 3.773791327334006, "learning_rate": 2.8015879332589864e-06, "loss": 0.3889, "step": 9247 }, { "epoch": 0.6556540233959589, "grad_norm": 15.157049924890746, "learning_rate": 2.8005567684199363e-06, "loss": 0.3827, "step": 9248 }, { "epoch": 0.6557249202410492, "grad_norm": 3.4587528794999822, "learning_rate": 2.7995257195592042e-06, "loss": 0.3816, "step": 9249 }, { "epoch": 0.6557958170861397, "grad_norm": 1.8121197235598852, "learning_rate": 2.798494786731163e-06, "loss": 0.4172, "step": 9250 }, { "epoch": 0.6558667139312301, "grad_norm": 2.8828930641792576, "learning_rate": 2.7974639699901696e-06, "loss": 0.4401, "step": 9251 }, { "epoch": 0.6559376107763204, "grad_norm": 5.531283332922435, "learning_rate": 2.7964332693905837e-06, "loss": 0.3676, "step": 9252 }, { "epoch": 0.6560085076214108, "grad_norm": 2.357130798263041, "learning_rate": 2.7954026849867544e-06, "loss": 0.4562, "step": 9253 }, { "epoch": 0.6560794044665013, "grad_norm": 2.872852260349122, "learning_rate": 2.7943722168330216e-06, "loss": 0.3807, "step": 9254 }, { "epoch": 0.6561503013115917, "grad_norm": 2.2759102234518145, "learning_rate": 2.7933418649837274e-06, "loss": 0.3448, "step": 9255 }, { "epoch": 0.656221198156682, "grad_norm": 2.933649816799828, "learning_rate": 2.792311629493201e-06, "loss": 0.3663, "step": 9256 }, { "epoch": 0.6562920950017724, "grad_norm": 4.462095464040134, "learning_rate": 2.791281510415765e-06, "loss": 0.3596, "step": 9257 }, { "epoch": 0.6563629918468629, "grad_norm": 3.9209411556805356, "learning_rate": 2.790251507805742e-06, "loss": 0.4483, "step": 9258 }, { "epoch": 0.6564338886919532, "grad_norm": 3.552896535222556, "learning_rate": 2.7892216217174465e-06, "loss": 0.3769, "step": 9259 }, { "epoch": 0.6565047855370436, "grad_norm": 3.2248901809313546, "learning_rate": 2.788191852205179e-06, "loss": 0.3637, "step": 9260 }, { "epoch": 0.656575682382134, "grad_norm": 2.1725802695494454, "learning_rate": 2.787162199323243e-06, "loss": 0.348, "step": 9261 }, { "epoch": 0.6566465792272244, "grad_norm": 2.256787097821445, "learning_rate": 2.786132663125935e-06, "loss": 0.42, "step": 9262 }, { "epoch": 0.6567174760723148, "grad_norm": 3.3748913852250793, "learning_rate": 2.785103243667542e-06, "loss": 0.4488, "step": 9263 }, { "epoch": 0.6567883729174052, "grad_norm": 10.73450042382524, "learning_rate": 2.7840739410023433e-06, "loss": 0.4112, "step": 9264 }, { "epoch": 0.6568592697624955, "grad_norm": 1.13989298932772, "learning_rate": 2.7830447551846165e-06, "loss": 0.6737, "step": 9265 }, { "epoch": 0.656930166607586, "grad_norm": 2.009196931359012, "learning_rate": 2.782015686268634e-06, "loss": 0.4018, "step": 9266 }, { "epoch": 0.6570010634526764, "grad_norm": 8.824828649003305, "learning_rate": 2.7809867343086576e-06, "loss": 0.3911, "step": 9267 }, { "epoch": 0.6570719602977667, "grad_norm": 13.34295379635369, "learning_rate": 2.779957899358943e-06, "loss": 0.3649, "step": 9268 }, { "epoch": 0.6571428571428571, "grad_norm": 2.478782124798409, "learning_rate": 2.778929181473744e-06, "loss": 0.374, "step": 9269 }, { "epoch": 0.6572137539879476, "grad_norm": 10.026767695411285, "learning_rate": 2.7779005807073044e-06, "loss": 0.3769, "step": 9270 }, { "epoch": 0.6572846508330379, "grad_norm": 2.4055781286251268, "learning_rate": 2.776872097113862e-06, "loss": 0.4151, "step": 9271 }, { "epoch": 0.6573555476781283, "grad_norm": 3.779348036510491, "learning_rate": 2.7758437307476516e-06, "loss": 0.3389, "step": 9272 }, { "epoch": 0.6574264445232187, "grad_norm": 2.735563242460252, "learning_rate": 2.774815481662898e-06, "loss": 0.3505, "step": 9273 }, { "epoch": 0.6574973413683091, "grad_norm": 2.4357640208098075, "learning_rate": 2.7737873499138242e-06, "loss": 0.387, "step": 9274 }, { "epoch": 0.6575682382133995, "grad_norm": 2.6270295689813805, "learning_rate": 2.772759335554642e-06, "loss": 0.3786, "step": 9275 }, { "epoch": 0.6576391350584899, "grad_norm": 2.3106857943900256, "learning_rate": 2.771731438639559e-06, "loss": 0.4106, "step": 9276 }, { "epoch": 0.6577100319035802, "grad_norm": 21.098071706479754, "learning_rate": 2.7707036592227775e-06, "loss": 0.3741, "step": 9277 }, { "epoch": 0.6577809287486707, "grad_norm": 9.547676058851701, "learning_rate": 2.7696759973584953e-06, "loss": 0.4635, "step": 9278 }, { "epoch": 0.6578518255937611, "grad_norm": 7.142802188363802, "learning_rate": 2.7686484531009006e-06, "loss": 0.3763, "step": 9279 }, { "epoch": 0.6579227224388515, "grad_norm": 2.6824084022813177, "learning_rate": 2.7676210265041737e-06, "loss": 0.4089, "step": 9280 }, { "epoch": 0.6579936192839418, "grad_norm": 5.8105613024347464, "learning_rate": 2.7665937176224962e-06, "loss": 0.3784, "step": 9281 }, { "epoch": 0.6580645161290323, "grad_norm": 13.707006601145377, "learning_rate": 2.7655665265100364e-06, "loss": 0.3521, "step": 9282 }, { "epoch": 0.6581354129741227, "grad_norm": 5.159521045770542, "learning_rate": 2.7645394532209573e-06, "loss": 0.3315, "step": 9283 }, { "epoch": 0.658206309819213, "grad_norm": 3.4822966975828513, "learning_rate": 2.763512497809419e-06, "loss": 0.3577, "step": 9284 }, { "epoch": 0.6582772066643034, "grad_norm": 11.922807185760597, "learning_rate": 2.7624856603295757e-06, "loss": 0.3818, "step": 9285 }, { "epoch": 0.6583481035093939, "grad_norm": 1.9578645655983151, "learning_rate": 2.761458940835572e-06, "loss": 0.41, "step": 9286 }, { "epoch": 0.6584190003544842, "grad_norm": 3.7566046289844772, "learning_rate": 2.7604323393815443e-06, "loss": 0.3961, "step": 9287 }, { "epoch": 0.6584898971995746, "grad_norm": 1.7302741169661056, "learning_rate": 2.759405856021631e-06, "loss": 0.4091, "step": 9288 }, { "epoch": 0.658560794044665, "grad_norm": 14.627938546319742, "learning_rate": 2.758379490809957e-06, "loss": 0.3834, "step": 9289 }, { "epoch": 0.6586316908897554, "grad_norm": 2.9778912780452518, "learning_rate": 2.7573532438006413e-06, "loss": 0.4061, "step": 9290 }, { "epoch": 0.6587025877348458, "grad_norm": 2.5349987670928487, "learning_rate": 2.7563271150478023e-06, "loss": 0.4118, "step": 9291 }, { "epoch": 0.6587734845799362, "grad_norm": 2.5011045925564863, "learning_rate": 2.7553011046055454e-06, "loss": 0.3976, "step": 9292 }, { "epoch": 0.6588443814250265, "grad_norm": 4.471261599878, "learning_rate": 2.754275212527977e-06, "loss": 0.4381, "step": 9293 }, { "epoch": 0.658915278270117, "grad_norm": 2.9137738275850222, "learning_rate": 2.753249438869189e-06, "loss": 0.3925, "step": 9294 }, { "epoch": 0.6589861751152074, "grad_norm": 13.852694338625515, "learning_rate": 2.7522237836832736e-06, "loss": 0.4173, "step": 9295 }, { "epoch": 0.6590570719602977, "grad_norm": 3.0850278679478826, "learning_rate": 2.7511982470243125e-06, "loss": 0.3895, "step": 9296 }, { "epoch": 0.6591279688053882, "grad_norm": 3.9476719704084946, "learning_rate": 2.750172828946387e-06, "loss": 0.4437, "step": 9297 }, { "epoch": 0.6591988656504786, "grad_norm": 6.1969698165540095, "learning_rate": 2.749147529503564e-06, "loss": 0.3589, "step": 9298 }, { "epoch": 0.6592697624955689, "grad_norm": 2.6774576118536286, "learning_rate": 2.7481223487499085e-06, "loss": 0.392, "step": 9299 }, { "epoch": 0.6593406593406593, "grad_norm": 4.23315355240092, "learning_rate": 2.747097286739483e-06, "loss": 0.4102, "step": 9300 }, { "epoch": 0.6594115561857498, "grad_norm": 2.916064185641667, "learning_rate": 2.7460723435263357e-06, "loss": 0.3562, "step": 9301 }, { "epoch": 0.6594824530308401, "grad_norm": 2.4970663468978618, "learning_rate": 2.7450475191645134e-06, "loss": 0.3879, "step": 9302 }, { "epoch": 0.6595533498759305, "grad_norm": 4.4140553725177645, "learning_rate": 2.744022813708056e-06, "loss": 0.3714, "step": 9303 }, { "epoch": 0.6596242467210209, "grad_norm": 2.1295959885857325, "learning_rate": 2.7429982272109996e-06, "loss": 0.35, "step": 9304 }, { "epoch": 0.6596951435661113, "grad_norm": 2.3217445089027247, "learning_rate": 2.7419737597273688e-06, "loss": 0.3741, "step": 9305 }, { "epoch": 0.6597660404112017, "grad_norm": 2.7474885571721797, "learning_rate": 2.740949411311184e-06, "loss": 0.4412, "step": 9306 }, { "epoch": 0.6598369372562921, "grad_norm": 2.546294986113062, "learning_rate": 2.7399251820164624e-06, "loss": 0.4075, "step": 9307 }, { "epoch": 0.6599078341013825, "grad_norm": 6.69632401274794, "learning_rate": 2.7389010718972115e-06, "loss": 0.4009, "step": 9308 }, { "epoch": 0.6599787309464729, "grad_norm": 3.1584592498666857, "learning_rate": 2.737877081007431e-06, "loss": 0.4131, "step": 9309 }, { "epoch": 0.6600496277915633, "grad_norm": 2.43160046941571, "learning_rate": 2.736853209401118e-06, "loss": 0.389, "step": 9310 }, { "epoch": 0.6601205246366537, "grad_norm": 2.525656671310385, "learning_rate": 2.735829457132264e-06, "loss": 0.4524, "step": 9311 }, { "epoch": 0.660191421481744, "grad_norm": 2.5101462236680243, "learning_rate": 2.7348058242548523e-06, "loss": 0.3976, "step": 9312 }, { "epoch": 0.6602623183268345, "grad_norm": 7.755570458027349, "learning_rate": 2.7337823108228556e-06, "loss": 0.3681, "step": 9313 }, { "epoch": 0.6603332151719249, "grad_norm": 7.694571879648364, "learning_rate": 2.7327589168902502e-06, "loss": 0.3817, "step": 9314 }, { "epoch": 0.6604041120170152, "grad_norm": 2.4088880381839304, "learning_rate": 2.7317356425109963e-06, "loss": 0.3402, "step": 9315 }, { "epoch": 0.6604750088621056, "grad_norm": 3.4994960874510332, "learning_rate": 2.7307124877390524e-06, "loss": 0.4076, "step": 9316 }, { "epoch": 0.6605459057071961, "grad_norm": 4.05610696124965, "learning_rate": 2.729689452628373e-06, "loss": 0.4094, "step": 9317 }, { "epoch": 0.6606168025522864, "grad_norm": 3.1321044767307056, "learning_rate": 2.728666537232899e-06, "loss": 0.3899, "step": 9318 }, { "epoch": 0.6606876993973768, "grad_norm": 2.3492646342294035, "learning_rate": 2.727643741606575e-06, "loss": 0.4347, "step": 9319 }, { "epoch": 0.6607585962424672, "grad_norm": 6.948807521996115, "learning_rate": 2.72662106580333e-06, "loss": 0.4047, "step": 9320 }, { "epoch": 0.6608294930875576, "grad_norm": 5.2710249961072195, "learning_rate": 2.7255985098770897e-06, "loss": 0.3512, "step": 9321 }, { "epoch": 0.660900389932648, "grad_norm": 2.966228986562263, "learning_rate": 2.724576073881776e-06, "loss": 0.3893, "step": 9322 }, { "epoch": 0.6609712867777384, "grad_norm": 7.8563437911888325, "learning_rate": 2.7235537578713046e-06, "loss": 0.4153, "step": 9323 }, { "epoch": 0.6610421836228287, "grad_norm": 1.9089571455779228, "learning_rate": 2.722531561899582e-06, "loss": 0.4004, "step": 9324 }, { "epoch": 0.6611130804679192, "grad_norm": 2.3962080015667944, "learning_rate": 2.721509486020505e-06, "loss": 0.4089, "step": 9325 }, { "epoch": 0.6611839773130096, "grad_norm": 3.7775790950568373, "learning_rate": 2.7204875302879746e-06, "loss": 0.4394, "step": 9326 }, { "epoch": 0.6612548741580999, "grad_norm": 3.5320914407417288, "learning_rate": 2.7194656947558764e-06, "loss": 0.3932, "step": 9327 }, { "epoch": 0.6613257710031903, "grad_norm": 3.3217487208638237, "learning_rate": 2.718443979478091e-06, "loss": 0.3785, "step": 9328 }, { "epoch": 0.6613966678482808, "grad_norm": 2.531287184256055, "learning_rate": 2.717422384508496e-06, "loss": 0.4492, "step": 9329 }, { "epoch": 0.6614675646933711, "grad_norm": 4.020679703831741, "learning_rate": 2.7164009099009624e-06, "loss": 0.3959, "step": 9330 }, { "epoch": 0.6615384615384615, "grad_norm": 3.4196475571965452, "learning_rate": 2.715379555709352e-06, "loss": 0.4899, "step": 9331 }, { "epoch": 0.661609358383552, "grad_norm": 3.481356277845962, "learning_rate": 2.7143583219875203e-06, "loss": 0.3622, "step": 9332 }, { "epoch": 0.6616802552286424, "grad_norm": 5.062517876116679, "learning_rate": 2.7133372087893193e-06, "loss": 0.391, "step": 9333 }, { "epoch": 0.6617511520737327, "grad_norm": 9.376666390769994, "learning_rate": 2.712316216168594e-06, "loss": 0.4114, "step": 9334 }, { "epoch": 0.6618220489188231, "grad_norm": 2.2093583930184435, "learning_rate": 2.7112953441791778e-06, "loss": 0.3727, "step": 9335 }, { "epoch": 0.6618929457639136, "grad_norm": 10.214261541938265, "learning_rate": 2.7102745928749074e-06, "loss": 0.4016, "step": 9336 }, { "epoch": 0.6619638426090039, "grad_norm": 2.488190723397766, "learning_rate": 2.709253962309604e-06, "loss": 0.4076, "step": 9337 }, { "epoch": 0.6620347394540943, "grad_norm": 3.2080956727641556, "learning_rate": 2.7082334525370883e-06, "loss": 0.3491, "step": 9338 }, { "epoch": 0.6621056362991847, "grad_norm": 3.019870622903945, "learning_rate": 2.7072130636111704e-06, "loss": 0.4173, "step": 9339 }, { "epoch": 0.662176533144275, "grad_norm": 2.9568593646737895, "learning_rate": 2.7061927955856604e-06, "loss": 0.4061, "step": 9340 }, { "epoch": 0.6622474299893655, "grad_norm": 2.409094137929332, "learning_rate": 2.705172648514352e-06, "loss": 0.3733, "step": 9341 }, { "epoch": 0.6623183268344559, "grad_norm": 6.886737951842518, "learning_rate": 2.704152622451044e-06, "loss": 0.4049, "step": 9342 }, { "epoch": 0.6623892236795462, "grad_norm": 4.129934511559491, "learning_rate": 2.7031327174495204e-06, "loss": 0.3908, "step": 9343 }, { "epoch": 0.6624601205246367, "grad_norm": 5.242102179003458, "learning_rate": 2.70211293356356e-06, "loss": 0.3987, "step": 9344 }, { "epoch": 0.6625310173697271, "grad_norm": 3.7108139079416467, "learning_rate": 2.7010932708469396e-06, "loss": 0.3359, "step": 9345 }, { "epoch": 0.6626019142148174, "grad_norm": 3.0260557588371153, "learning_rate": 2.7000737293534263e-06, "loss": 0.4037, "step": 9346 }, { "epoch": 0.6626728110599078, "grad_norm": 4.44714874102705, "learning_rate": 2.699054309136779e-06, "loss": 0.4145, "step": 9347 }, { "epoch": 0.6627437079049983, "grad_norm": 5.855265603254551, "learning_rate": 2.698035010250754e-06, "loss": 0.4754, "step": 9348 }, { "epoch": 0.6628146047500886, "grad_norm": 6.079192992014887, "learning_rate": 2.697015832749102e-06, "loss": 0.3818, "step": 9349 }, { "epoch": 0.662885501595179, "grad_norm": 2.4267328038041023, "learning_rate": 2.695996776685563e-06, "loss": 0.3717, "step": 9350 }, { "epoch": 0.6629563984402694, "grad_norm": 3.7583429261830466, "learning_rate": 2.69497784211387e-06, "loss": 0.4227, "step": 9351 }, { "epoch": 0.6630272952853598, "grad_norm": 5.035956037352134, "learning_rate": 2.693959029087755e-06, "loss": 0.3968, "step": 9352 }, { "epoch": 0.6630981921304502, "grad_norm": 2.622465805284335, "learning_rate": 2.6929403376609447e-06, "loss": 0.3865, "step": 9353 }, { "epoch": 0.6631690889755406, "grad_norm": 3.1779041194828728, "learning_rate": 2.6919217678871475e-06, "loss": 0.4162, "step": 9354 }, { "epoch": 0.6632399858206309, "grad_norm": 2.724115630167031, "learning_rate": 2.6909033198200768e-06, "loss": 0.3782, "step": 9355 }, { "epoch": 0.6633108826657214, "grad_norm": 1.1728792961433077, "learning_rate": 2.6898849935134387e-06, "loss": 0.7659, "step": 9356 }, { "epoch": 0.6633817795108118, "grad_norm": 4.171073954125746, "learning_rate": 2.6888667890209276e-06, "loss": 0.395, "step": 9357 }, { "epoch": 0.6634526763559022, "grad_norm": 2.2713543620221603, "learning_rate": 2.6878487063962333e-06, "loss": 0.3836, "step": 9358 }, { "epoch": 0.6635235732009925, "grad_norm": 2.6780622583603555, "learning_rate": 2.686830745693043e-06, "loss": 0.4086, "step": 9359 }, { "epoch": 0.663594470046083, "grad_norm": 2.5124530532229, "learning_rate": 2.6858129069650316e-06, "loss": 0.3573, "step": 9360 }, { "epoch": 0.6636653668911734, "grad_norm": 2.6518522006475256, "learning_rate": 2.6847951902658737e-06, "loss": 0.399, "step": 9361 }, { "epoch": 0.6637362637362637, "grad_norm": 3.183176986953042, "learning_rate": 2.6837775956492323e-06, "loss": 0.3643, "step": 9362 }, { "epoch": 0.6638071605813541, "grad_norm": 2.206945952634911, "learning_rate": 2.6827601231687644e-06, "loss": 0.3588, "step": 9363 }, { "epoch": 0.6638780574264446, "grad_norm": 3.7427216602759596, "learning_rate": 2.681742772878126e-06, "loss": 0.4604, "step": 9364 }, { "epoch": 0.6639489542715349, "grad_norm": 4.797407194220383, "learning_rate": 2.6807255448309604e-06, "loss": 0.423, "step": 9365 }, { "epoch": 0.6640198511166253, "grad_norm": 2.3998329376876533, "learning_rate": 2.6797084390809045e-06, "loss": 0.389, "step": 9366 }, { "epoch": 0.6640907479617157, "grad_norm": 7.663644687299985, "learning_rate": 2.6786914556815946e-06, "loss": 0.4031, "step": 9367 }, { "epoch": 0.6641616448068061, "grad_norm": 4.364178362856189, "learning_rate": 2.6776745946866576e-06, "loss": 0.3925, "step": 9368 }, { "epoch": 0.6642325416518965, "grad_norm": 5.037361028262885, "learning_rate": 2.6766578561497124e-06, "loss": 0.397, "step": 9369 }, { "epoch": 0.6643034384969869, "grad_norm": 3.608199439365831, "learning_rate": 2.6756412401243694e-06, "loss": 0.4143, "step": 9370 }, { "epoch": 0.6643743353420772, "grad_norm": 4.149027717923313, "learning_rate": 2.6746247466642394e-06, "loss": 0.4502, "step": 9371 }, { "epoch": 0.6644452321871677, "grad_norm": 2.109097059317909, "learning_rate": 2.6736083758229214e-06, "loss": 0.3925, "step": 9372 }, { "epoch": 0.6645161290322581, "grad_norm": 5.695411220680348, "learning_rate": 2.672592127654008e-06, "loss": 0.3368, "step": 9373 }, { "epoch": 0.6645870258773484, "grad_norm": 1.0142302158434349, "learning_rate": 2.6715760022110878e-06, "loss": 0.6542, "step": 9374 }, { "epoch": 0.6646579227224388, "grad_norm": 2.368265249690558, "learning_rate": 2.6705599995477438e-06, "loss": 0.4017, "step": 9375 }, { "epoch": 0.6647288195675293, "grad_norm": 2.5443441521834544, "learning_rate": 2.6695441197175496e-06, "loss": 0.3889, "step": 9376 }, { "epoch": 0.6647997164126196, "grad_norm": 2.340975385116243, "learning_rate": 2.668528362774071e-06, "loss": 0.4235, "step": 9377 }, { "epoch": 0.66487061325771, "grad_norm": 3.289542146315753, "learning_rate": 2.6675127287708723e-06, "loss": 0.4065, "step": 9378 }, { "epoch": 0.6649415101028004, "grad_norm": 2.740277564594231, "learning_rate": 2.666497217761509e-06, "loss": 0.3604, "step": 9379 }, { "epoch": 0.6650124069478908, "grad_norm": 5.211336683608997, "learning_rate": 2.6654818297995256e-06, "loss": 0.3823, "step": 9380 }, { "epoch": 0.6650833037929812, "grad_norm": 14.358378125413443, "learning_rate": 2.664466564938467e-06, "loss": 0.4028, "step": 9381 }, { "epoch": 0.6651542006380716, "grad_norm": 2.6692438638046845, "learning_rate": 2.663451423231872e-06, "loss": 0.4006, "step": 9382 }, { "epoch": 0.665225097483162, "grad_norm": 2.618655435985214, "learning_rate": 2.6624364047332663e-06, "loss": 0.419, "step": 9383 }, { "epoch": 0.6652959943282524, "grad_norm": 3.4962355197716417, "learning_rate": 2.6614215094961716e-06, "loss": 0.4299, "step": 9384 }, { "epoch": 0.6653668911733428, "grad_norm": 2.33868913612184, "learning_rate": 2.6604067375741072e-06, "loss": 0.3651, "step": 9385 }, { "epoch": 0.6654377880184332, "grad_norm": 3.9937673623217735, "learning_rate": 2.6593920890205786e-06, "loss": 0.3398, "step": 9386 }, { "epoch": 0.6655086848635235, "grad_norm": 6.056182900934351, "learning_rate": 2.6583775638890948e-06, "loss": 0.3527, "step": 9387 }, { "epoch": 0.665579581708614, "grad_norm": 2.703262045873138, "learning_rate": 2.6573631622331485e-06, "loss": 0.4131, "step": 9388 }, { "epoch": 0.6656504785537044, "grad_norm": 2.359421070204861, "learning_rate": 2.656348884106229e-06, "loss": 0.4093, "step": 9389 }, { "epoch": 0.6657213753987947, "grad_norm": 3.5076718754360328, "learning_rate": 2.655334729561824e-06, "loss": 0.4099, "step": 9390 }, { "epoch": 0.6657922722438852, "grad_norm": 3.391097218197575, "learning_rate": 2.654320698653408e-06, "loss": 0.3935, "step": 9391 }, { "epoch": 0.6658631690889756, "grad_norm": 4.480658005281796, "learning_rate": 2.6533067914344502e-06, "loss": 0.4009, "step": 9392 }, { "epoch": 0.6659340659340659, "grad_norm": 2.362634217769024, "learning_rate": 2.6522930079584162e-06, "loss": 0.3872, "step": 9393 }, { "epoch": 0.6660049627791563, "grad_norm": 4.997185618846738, "learning_rate": 2.6512793482787657e-06, "loss": 0.4041, "step": 9394 }, { "epoch": 0.6660758596242468, "grad_norm": 1.8004568629060698, "learning_rate": 2.650265812448948e-06, "loss": 0.3577, "step": 9395 }, { "epoch": 0.6661467564693371, "grad_norm": 4.0498936407236386, "learning_rate": 2.6492524005224058e-06, "loss": 0.3386, "step": 9396 }, { "epoch": 0.6662176533144275, "grad_norm": 3.2317970444825455, "learning_rate": 2.6482391125525788e-06, "loss": 0.3827, "step": 9397 }, { "epoch": 0.6662885501595179, "grad_norm": 2.7775623702002528, "learning_rate": 2.647225948592902e-06, "loss": 0.3885, "step": 9398 }, { "epoch": 0.6663594470046083, "grad_norm": 10.13558020898901, "learning_rate": 2.6462129086967926e-06, "loss": 0.3434, "step": 9399 }, { "epoch": 0.6664303438496987, "grad_norm": 4.789018897082056, "learning_rate": 2.645199992917673e-06, "loss": 0.397, "step": 9400 }, { "epoch": 0.6665012406947891, "grad_norm": 2.6563086056331633, "learning_rate": 2.6441872013089575e-06, "loss": 0.3569, "step": 9401 }, { "epoch": 0.6665721375398794, "grad_norm": 2.175981766217507, "learning_rate": 2.6431745339240488e-06, "loss": 0.3903, "step": 9402 }, { "epoch": 0.6666430343849699, "grad_norm": 23.533284197547072, "learning_rate": 2.6421619908163444e-06, "loss": 0.4279, "step": 9403 }, { "epoch": 0.6667139312300603, "grad_norm": 4.9514029034595275, "learning_rate": 2.6411495720392395e-06, "loss": 0.4104, "step": 9404 }, { "epoch": 0.6667848280751506, "grad_norm": 3.0467094916430195, "learning_rate": 2.6401372776461154e-06, "loss": 0.4325, "step": 9405 }, { "epoch": 0.666855724920241, "grad_norm": 2.297765534848772, "learning_rate": 2.639125107690357e-06, "loss": 0.3958, "step": 9406 }, { "epoch": 0.6669266217653315, "grad_norm": 3.8141451602936374, "learning_rate": 2.6381130622253327e-06, "loss": 0.4202, "step": 9407 }, { "epoch": 0.6669975186104219, "grad_norm": 4.648950935664325, "learning_rate": 2.637101141304408e-06, "loss": 0.443, "step": 9408 }, { "epoch": 0.6670684154555122, "grad_norm": 4.1057323156525936, "learning_rate": 2.6360893449809454e-06, "loss": 0.377, "step": 9409 }, { "epoch": 0.6671393123006026, "grad_norm": 4.333554146445143, "learning_rate": 2.635077673308294e-06, "loss": 0.3827, "step": 9410 }, { "epoch": 0.6672102091456931, "grad_norm": 2.9382460857156927, "learning_rate": 2.634066126339804e-06, "loss": 0.3765, "step": 9411 }, { "epoch": 0.6672811059907834, "grad_norm": 2.6227329624785725, "learning_rate": 2.6330547041288107e-06, "loss": 0.4045, "step": 9412 }, { "epoch": 0.6673520028358738, "grad_norm": 3.1570863998006247, "learning_rate": 2.6320434067286517e-06, "loss": 0.4197, "step": 9413 }, { "epoch": 0.6674228996809642, "grad_norm": 2.246266345387002, "learning_rate": 2.631032234192651e-06, "loss": 0.3389, "step": 9414 }, { "epoch": 0.6674937965260546, "grad_norm": 2.2505525102803285, "learning_rate": 2.6300211865741263e-06, "loss": 0.3971, "step": 9415 }, { "epoch": 0.667564693371145, "grad_norm": 4.819973904257208, "learning_rate": 2.6290102639263946e-06, "loss": 0.3832, "step": 9416 }, { "epoch": 0.6676355902162354, "grad_norm": 4.8993420747435605, "learning_rate": 2.6279994663027648e-06, "loss": 0.3629, "step": 9417 }, { "epoch": 0.6677064870613257, "grad_norm": 6.786955260522161, "learning_rate": 2.6269887937565295e-06, "loss": 0.4105, "step": 9418 }, { "epoch": 0.6677773839064162, "grad_norm": 3.584440116457379, "learning_rate": 2.6259782463409867e-06, "loss": 0.4095, "step": 9419 }, { "epoch": 0.6678482807515066, "grad_norm": 2.9213238651642603, "learning_rate": 2.6249678241094244e-06, "loss": 0.4331, "step": 9420 }, { "epoch": 0.6679191775965969, "grad_norm": 3.486942359158623, "learning_rate": 2.623957527115122e-06, "loss": 0.4121, "step": 9421 }, { "epoch": 0.6679900744416873, "grad_norm": 2.4288789092409866, "learning_rate": 2.6229473554113504e-06, "loss": 0.4101, "step": 9422 }, { "epoch": 0.6680609712867778, "grad_norm": 2.7599284206001746, "learning_rate": 2.621937309051381e-06, "loss": 0.4032, "step": 9423 }, { "epoch": 0.6681318681318681, "grad_norm": 2.203919453090799, "learning_rate": 2.620927388088472e-06, "loss": 0.44, "step": 9424 }, { "epoch": 0.6682027649769585, "grad_norm": 3.189986201665339, "learning_rate": 2.619917592575877e-06, "loss": 0.3822, "step": 9425 }, { "epoch": 0.668273661822049, "grad_norm": 2.2737099956134097, "learning_rate": 2.6189079225668435e-06, "loss": 0.3178, "step": 9426 }, { "epoch": 0.6683445586671393, "grad_norm": 0.9401929634867303, "learning_rate": 2.617898378114614e-06, "loss": 0.6563, "step": 9427 }, { "epoch": 0.6684154555122297, "grad_norm": 24.35429733350833, "learning_rate": 2.6168889592724216e-06, "loss": 0.3848, "step": 9428 }, { "epoch": 0.6684863523573201, "grad_norm": 3.0804626948855454, "learning_rate": 2.6158796660934916e-06, "loss": 0.4182, "step": 9429 }, { "epoch": 0.6685572492024104, "grad_norm": 3.5642552544381676, "learning_rate": 2.614870498631048e-06, "loss": 0.3856, "step": 9430 }, { "epoch": 0.6686281460475009, "grad_norm": 2.927245818095535, "learning_rate": 2.613861456938302e-06, "loss": 0.3766, "step": 9431 }, { "epoch": 0.6686990428925913, "grad_norm": 2.4321302913462555, "learning_rate": 2.612852541068464e-06, "loss": 0.4177, "step": 9432 }, { "epoch": 0.6687699397376816, "grad_norm": 3.2913218679561527, "learning_rate": 2.6118437510747343e-06, "loss": 0.3971, "step": 9433 }, { "epoch": 0.668840836582772, "grad_norm": 2.7222411655043515, "learning_rate": 2.6108350870103037e-06, "loss": 0.4382, "step": 9434 }, { "epoch": 0.6689117334278625, "grad_norm": 3.165859899289396, "learning_rate": 2.6098265489283648e-06, "loss": 0.3761, "step": 9435 }, { "epoch": 0.6689826302729529, "grad_norm": 4.2119222354143755, "learning_rate": 2.6088181368820964e-06, "loss": 0.3937, "step": 9436 }, { "epoch": 0.6690535271180432, "grad_norm": 5.561894891122236, "learning_rate": 2.607809850924671e-06, "loss": 0.4016, "step": 9437 }, { "epoch": 0.6691244239631337, "grad_norm": 1.135991519995978, "learning_rate": 2.6068016911092576e-06, "loss": 0.6837, "step": 9438 }, { "epoch": 0.6691953208082241, "grad_norm": 4.811721656451192, "learning_rate": 2.6057936574890195e-06, "loss": 0.4468, "step": 9439 }, { "epoch": 0.6692662176533144, "grad_norm": 5.921112936970381, "learning_rate": 2.6047857501171097e-06, "loss": 0.4031, "step": 9440 }, { "epoch": 0.6693371144984048, "grad_norm": 3.6153768849188954, "learning_rate": 2.6037779690466736e-06, "loss": 0.3771, "step": 9441 }, { "epoch": 0.6694080113434953, "grad_norm": 2.386120446861068, "learning_rate": 2.6027703143308536e-06, "loss": 0.3669, "step": 9442 }, { "epoch": 0.6694789081885856, "grad_norm": 3.5002793751500327, "learning_rate": 2.601762786022789e-06, "loss": 0.3785, "step": 9443 }, { "epoch": 0.669549805033676, "grad_norm": 4.005766496694907, "learning_rate": 2.6007553841755993e-06, "loss": 0.4156, "step": 9444 }, { "epoch": 0.6696207018787664, "grad_norm": 4.236440229255541, "learning_rate": 2.5997481088424093e-06, "loss": 0.3799, "step": 9445 }, { "epoch": 0.6696915987238568, "grad_norm": 2.3662918389541643, "learning_rate": 2.598740960076336e-06, "loss": 0.4118, "step": 9446 }, { "epoch": 0.6697624955689472, "grad_norm": 3.5631508521194486, "learning_rate": 2.5977339379304834e-06, "loss": 0.3544, "step": 9447 }, { "epoch": 0.6698333924140376, "grad_norm": 3.698438764026046, "learning_rate": 2.5967270424579527e-06, "loss": 0.3466, "step": 9448 }, { "epoch": 0.6699042892591279, "grad_norm": 2.1422183725184354, "learning_rate": 2.5957202737118416e-06, "loss": 0.3801, "step": 9449 }, { "epoch": 0.6699751861042184, "grad_norm": 2.429568549060402, "learning_rate": 2.594713631745233e-06, "loss": 0.473, "step": 9450 }, { "epoch": 0.6700460829493088, "grad_norm": 2.891980684421691, "learning_rate": 2.593707116611213e-06, "loss": 0.428, "step": 9451 }, { "epoch": 0.6701169797943991, "grad_norm": 3.8094365061591766, "learning_rate": 2.592700728362853e-06, "loss": 0.4393, "step": 9452 }, { "epoch": 0.6701878766394895, "grad_norm": 3.608937125228932, "learning_rate": 2.5916944670532195e-06, "loss": 0.3709, "step": 9453 }, { "epoch": 0.67025877348458, "grad_norm": 1.9924343411343297, "learning_rate": 2.590688332735376e-06, "loss": 0.4052, "step": 9454 }, { "epoch": 0.6703296703296703, "grad_norm": 4.154333264922482, "learning_rate": 2.589682325462375e-06, "loss": 0.3977, "step": 9455 }, { "epoch": 0.6704005671747607, "grad_norm": 10.107937019673361, "learning_rate": 2.588676445287265e-06, "loss": 0.4157, "step": 9456 }, { "epoch": 0.6704714640198511, "grad_norm": 3.8374812381010472, "learning_rate": 2.5876706922630857e-06, "loss": 0.4083, "step": 9457 }, { "epoch": 0.6705423608649415, "grad_norm": 2.9914619663576176, "learning_rate": 2.586665066442874e-06, "loss": 0.3904, "step": 9458 }, { "epoch": 0.6706132577100319, "grad_norm": 15.701128290537586, "learning_rate": 2.5856595678796548e-06, "loss": 0.3889, "step": 9459 }, { "epoch": 0.6706841545551223, "grad_norm": 5.071090913636016, "learning_rate": 2.584654196626448e-06, "loss": 0.3761, "step": 9460 }, { "epoch": 0.6707550514002127, "grad_norm": 2.3130036929377864, "learning_rate": 2.5836489527362684e-06, "loss": 0.4418, "step": 9461 }, { "epoch": 0.6708259482453031, "grad_norm": 6.945147415213716, "learning_rate": 2.5826438362621286e-06, "loss": 0.3859, "step": 9462 }, { "epoch": 0.6708968450903935, "grad_norm": 2.300307265655125, "learning_rate": 2.58163884725702e-06, "loss": 0.3788, "step": 9463 }, { "epoch": 0.6709677419354839, "grad_norm": 5.977620493151962, "learning_rate": 2.5806339857739406e-06, "loss": 0.3573, "step": 9464 }, { "epoch": 0.6710386387805742, "grad_norm": 3.302175102322108, "learning_rate": 2.5796292518658796e-06, "loss": 0.4175, "step": 9465 }, { "epoch": 0.6711095356256647, "grad_norm": 3.78952751933142, "learning_rate": 2.5786246455858153e-06, "loss": 0.367, "step": 9466 }, { "epoch": 0.6711804324707551, "grad_norm": 4.632329939034832, "learning_rate": 2.5776201669867207e-06, "loss": 0.4352, "step": 9467 }, { "epoch": 0.6712513293158454, "grad_norm": 3.307050067885282, "learning_rate": 2.5766158161215647e-06, "loss": 0.3876, "step": 9468 }, { "epoch": 0.6713222261609358, "grad_norm": 7.86627644562822, "learning_rate": 2.575611593043305e-06, "loss": 0.3876, "step": 9469 }, { "epoch": 0.6713931230060263, "grad_norm": 2.236242643909801, "learning_rate": 2.5746074978048984e-06, "loss": 0.3643, "step": 9470 }, { "epoch": 0.6714640198511166, "grad_norm": 3.921133359132214, "learning_rate": 2.5736035304592876e-06, "loss": 0.4095, "step": 9471 }, { "epoch": 0.671534916696207, "grad_norm": 5.918279469979719, "learning_rate": 2.5725996910594164e-06, "loss": 0.4112, "step": 9472 }, { "epoch": 0.6716058135412974, "grad_norm": 2.9814615297998617, "learning_rate": 2.5715959796582157e-06, "loss": 0.4249, "step": 9473 }, { "epoch": 0.6716767103863878, "grad_norm": 2.2804627413389644, "learning_rate": 2.5705923963086117e-06, "loss": 0.3947, "step": 9474 }, { "epoch": 0.6717476072314782, "grad_norm": 7.861727692480676, "learning_rate": 2.569588941063526e-06, "loss": 0.3854, "step": 9475 }, { "epoch": 0.6718185040765686, "grad_norm": 4.2025411300211255, "learning_rate": 2.5685856139758693e-06, "loss": 0.4445, "step": 9476 }, { "epoch": 0.6718894009216589, "grad_norm": 3.6305009285535994, "learning_rate": 2.56758241509855e-06, "loss": 0.3715, "step": 9477 }, { "epoch": 0.6719602977667494, "grad_norm": 4.925447055583414, "learning_rate": 2.566579344484467e-06, "loss": 0.3778, "step": 9478 }, { "epoch": 0.6720311946118398, "grad_norm": 2.0354593929018674, "learning_rate": 2.5655764021865105e-06, "loss": 0.3772, "step": 9479 }, { "epoch": 0.6721020914569301, "grad_norm": 2.6601603800387625, "learning_rate": 2.5645735882575686e-06, "loss": 0.3774, "step": 9480 }, { "epoch": 0.6721729883020205, "grad_norm": 9.34849532134273, "learning_rate": 2.563570902750524e-06, "loss": 0.3901, "step": 9481 }, { "epoch": 0.672243885147111, "grad_norm": 3.5177810929266093, "learning_rate": 2.5625683457182417e-06, "loss": 0.4142, "step": 9482 }, { "epoch": 0.6723147819922013, "grad_norm": 1.8968296767881412, "learning_rate": 2.561565917213591e-06, "loss": 0.4271, "step": 9483 }, { "epoch": 0.6723856788372917, "grad_norm": 2.8674687229184355, "learning_rate": 2.560563617289432e-06, "loss": 0.3699, "step": 9484 }, { "epoch": 0.6724565756823822, "grad_norm": 4.9094850869733735, "learning_rate": 2.559561445998616e-06, "loss": 0.4033, "step": 9485 }, { "epoch": 0.6725274725274726, "grad_norm": 2.411100676795487, "learning_rate": 2.5585594033939853e-06, "loss": 0.3448, "step": 9486 }, { "epoch": 0.6725983693725629, "grad_norm": 3.1032473794004174, "learning_rate": 2.5575574895283806e-06, "loss": 0.4017, "step": 9487 }, { "epoch": 0.6726692662176533, "grad_norm": 7.200212259904704, "learning_rate": 2.5565557044546386e-06, "loss": 0.3632, "step": 9488 }, { "epoch": 0.6727401630627438, "grad_norm": 1.0874938037384894, "learning_rate": 2.555554048225574e-06, "loss": 0.6711, "step": 9489 }, { "epoch": 0.6728110599078341, "grad_norm": 6.4491568010937455, "learning_rate": 2.5545525208940113e-06, "loss": 0.3825, "step": 9490 }, { "epoch": 0.6728819567529245, "grad_norm": 2.6263914998776743, "learning_rate": 2.553551122512762e-06, "loss": 0.3684, "step": 9491 }, { "epoch": 0.6729528535980149, "grad_norm": 3.8625693416618145, "learning_rate": 2.552549853134629e-06, "loss": 0.4186, "step": 9492 }, { "epoch": 0.6730237504431053, "grad_norm": 3.5664799895532218, "learning_rate": 2.5515487128124094e-06, "loss": 0.3955, "step": 9493 }, { "epoch": 0.6730946472881957, "grad_norm": 2.786112464977455, "learning_rate": 2.5505477015988964e-06, "loss": 0.4169, "step": 9494 }, { "epoch": 0.6731655441332861, "grad_norm": 18.707422267722276, "learning_rate": 2.5495468195468707e-06, "loss": 0.3914, "step": 9495 }, { "epoch": 0.6732364409783764, "grad_norm": 4.126412469180832, "learning_rate": 2.548546066709114e-06, "loss": 0.4295, "step": 9496 }, { "epoch": 0.6733073378234669, "grad_norm": 3.1088629338681493, "learning_rate": 2.547545443138394e-06, "loss": 0.446, "step": 9497 }, { "epoch": 0.6733782346685573, "grad_norm": 3.3805582793663653, "learning_rate": 2.546544948887472e-06, "loss": 0.4218, "step": 9498 }, { "epoch": 0.6734491315136476, "grad_norm": 11.335938243032842, "learning_rate": 2.5455445840091102e-06, "loss": 0.383, "step": 9499 }, { "epoch": 0.673520028358738, "grad_norm": 2.8126516310265566, "learning_rate": 2.5445443485560537e-06, "loss": 0.3989, "step": 9500 }, { "epoch": 0.6735909252038285, "grad_norm": 2.4384581806452976, "learning_rate": 2.54354424258105e-06, "loss": 0.3973, "step": 9501 }, { "epoch": 0.6736618220489188, "grad_norm": 2.475426318037749, "learning_rate": 2.542544266136831e-06, "loss": 0.4049, "step": 9502 }, { "epoch": 0.6737327188940092, "grad_norm": 4.947519811287735, "learning_rate": 2.5415444192761306e-06, "loss": 0.4601, "step": 9503 }, { "epoch": 0.6738036157390996, "grad_norm": 2.360030486595569, "learning_rate": 2.54054470205167e-06, "loss": 0.4076, "step": 9504 }, { "epoch": 0.67387451258419, "grad_norm": 2.5696663364078414, "learning_rate": 2.539545114516162e-06, "loss": 0.3972, "step": 9505 }, { "epoch": 0.6739454094292804, "grad_norm": 3.938550607456351, "learning_rate": 2.538545656722318e-06, "loss": 0.3807, "step": 9506 }, { "epoch": 0.6740163062743708, "grad_norm": 5.894712188443413, "learning_rate": 2.537546328722844e-06, "loss": 0.3637, "step": 9507 }, { "epoch": 0.6740872031194611, "grad_norm": 2.5729755716561606, "learning_rate": 2.5365471305704283e-06, "loss": 0.3907, "step": 9508 }, { "epoch": 0.6741580999645516, "grad_norm": 4.935024500858645, "learning_rate": 2.5355480623177625e-06, "loss": 0.4059, "step": 9509 }, { "epoch": 0.674228996809642, "grad_norm": 3.3417739905675425, "learning_rate": 2.53454912401753e-06, "loss": 0.4809, "step": 9510 }, { "epoch": 0.6742998936547324, "grad_norm": 2.4822266653614156, "learning_rate": 2.5335503157224044e-06, "loss": 0.4099, "step": 9511 }, { "epoch": 0.6743707904998227, "grad_norm": 2.5294610490838907, "learning_rate": 2.532551637485051e-06, "loss": 0.4035, "step": 9512 }, { "epoch": 0.6744416873449132, "grad_norm": 1.8907897060677208, "learning_rate": 2.531553089358133e-06, "loss": 0.4091, "step": 9513 }, { "epoch": 0.6745125841900036, "grad_norm": 7.270273421389624, "learning_rate": 2.5305546713943074e-06, "loss": 0.3784, "step": 9514 }, { "epoch": 0.6745834810350939, "grad_norm": 2.8636657234354956, "learning_rate": 2.5295563836462177e-06, "loss": 0.4248, "step": 9515 }, { "epoch": 0.6746543778801843, "grad_norm": 2.6767027624543482, "learning_rate": 2.5285582261665043e-06, "loss": 0.3935, "step": 9516 }, { "epoch": 0.6747252747252748, "grad_norm": 2.3206087483539357, "learning_rate": 2.5275601990078036e-06, "loss": 0.3773, "step": 9517 }, { "epoch": 0.6747961715703651, "grad_norm": 2.5107934108946286, "learning_rate": 2.5265623022227408e-06, "loss": 0.4017, "step": 9518 }, { "epoch": 0.6748670684154555, "grad_norm": 2.8531499616899456, "learning_rate": 2.5255645358639336e-06, "loss": 0.3902, "step": 9519 }, { "epoch": 0.674937965260546, "grad_norm": 2.225402061652643, "learning_rate": 2.5245668999839997e-06, "loss": 0.3421, "step": 9520 }, { "epoch": 0.6750088621056363, "grad_norm": 5.2530382074619455, "learning_rate": 2.52356939463554e-06, "loss": 0.3845, "step": 9521 }, { "epoch": 0.6750797589507267, "grad_norm": 2.8850264580501355, "learning_rate": 2.5225720198711577e-06, "loss": 0.3846, "step": 9522 }, { "epoch": 0.6751506557958171, "grad_norm": 2.3321999645554903, "learning_rate": 2.521574775743445e-06, "loss": 0.4265, "step": 9523 }, { "epoch": 0.6752215526409074, "grad_norm": 6.447048189276249, "learning_rate": 2.5205776623049837e-06, "loss": 0.3866, "step": 9524 }, { "epoch": 0.6752924494859979, "grad_norm": 16.055355533256584, "learning_rate": 2.519580679608354e-06, "loss": 0.4283, "step": 9525 }, { "epoch": 0.6753633463310883, "grad_norm": 10.94263891213372, "learning_rate": 2.5185838277061326e-06, "loss": 0.455, "step": 9526 }, { "epoch": 0.6754342431761786, "grad_norm": 5.146780613068981, "learning_rate": 2.517587106650876e-06, "loss": 0.4114, "step": 9527 }, { "epoch": 0.675505140021269, "grad_norm": 3.2575032708304894, "learning_rate": 2.5165905164951463e-06, "loss": 0.3887, "step": 9528 }, { "epoch": 0.6755760368663595, "grad_norm": 35.172062519716306, "learning_rate": 2.5155940572914963e-06, "loss": 0.3871, "step": 9529 }, { "epoch": 0.6756469337114498, "grad_norm": 2.3789419952578617, "learning_rate": 2.5145977290924678e-06, "loss": 0.3736, "step": 9530 }, { "epoch": 0.6757178305565402, "grad_norm": 1.922215912647628, "learning_rate": 2.513601531950596e-06, "loss": 0.3171, "step": 9531 }, { "epoch": 0.6757887274016307, "grad_norm": 2.9549054510582504, "learning_rate": 2.512605465918413e-06, "loss": 0.393, "step": 9532 }, { "epoch": 0.675859624246721, "grad_norm": 7.2836208178259145, "learning_rate": 2.5116095310484445e-06, "loss": 0.4213, "step": 9533 }, { "epoch": 0.6759305210918114, "grad_norm": 3.304672479272064, "learning_rate": 2.510613727393204e-06, "loss": 0.37, "step": 9534 }, { "epoch": 0.6760014179369018, "grad_norm": 4.878782807546249, "learning_rate": 2.509618055005201e-06, "loss": 0.3967, "step": 9535 }, { "epoch": 0.6760723147819923, "grad_norm": 3.4582981451602994, "learning_rate": 2.5086225139369402e-06, "loss": 0.4358, "step": 9536 }, { "epoch": 0.6761432116270826, "grad_norm": 2.7459856218650573, "learning_rate": 2.5076271042409156e-06, "loss": 0.3851, "step": 9537 }, { "epoch": 0.676214108472173, "grad_norm": 3.9116120798648866, "learning_rate": 2.5066318259696143e-06, "loss": 0.3408, "step": 9538 }, { "epoch": 0.6762850053172634, "grad_norm": 3.2702429272974882, "learning_rate": 2.5056366791755217e-06, "loss": 0.4703, "step": 9539 }, { "epoch": 0.6763559021623538, "grad_norm": 2.909632962917071, "learning_rate": 2.5046416639111095e-06, "loss": 0.3706, "step": 9540 }, { "epoch": 0.6764267990074442, "grad_norm": 3.7662865394680236, "learning_rate": 2.5036467802288483e-06, "loss": 0.3662, "step": 9541 }, { "epoch": 0.6764976958525346, "grad_norm": 4.247005754649202, "learning_rate": 2.5026520281811984e-06, "loss": 0.4013, "step": 9542 }, { "epoch": 0.6765685926976249, "grad_norm": 2.2456009695685566, "learning_rate": 2.50165740782061e-06, "loss": 0.3805, "step": 9543 }, { "epoch": 0.6766394895427154, "grad_norm": 3.6143523130100963, "learning_rate": 2.5006629191995367e-06, "loss": 0.3989, "step": 9544 }, { "epoch": 0.6767103863878058, "grad_norm": 2.286515904789272, "learning_rate": 2.499668562370413e-06, "loss": 0.3417, "step": 9545 }, { "epoch": 0.6767812832328961, "grad_norm": 3.4935778569232965, "learning_rate": 2.498674337385676e-06, "loss": 0.412, "step": 9546 }, { "epoch": 0.6768521800779865, "grad_norm": 2.659477258646566, "learning_rate": 2.4976802442977488e-06, "loss": 0.3907, "step": 9547 }, { "epoch": 0.676923076923077, "grad_norm": 7.463463082098815, "learning_rate": 2.496686283159054e-06, "loss": 0.381, "step": 9548 }, { "epoch": 0.6769939737681673, "grad_norm": 3.0291957405210184, "learning_rate": 2.495692454022002e-06, "loss": 0.3613, "step": 9549 }, { "epoch": 0.6770648706132577, "grad_norm": 5.744044223504445, "learning_rate": 2.4946987569389967e-06, "loss": 0.4155, "step": 9550 }, { "epoch": 0.6771357674583481, "grad_norm": 2.9073745366056216, "learning_rate": 2.4937051919624385e-06, "loss": 0.4458, "step": 9551 }, { "epoch": 0.6772066643034385, "grad_norm": 2.737420717286633, "learning_rate": 2.4927117591447218e-06, "loss": 0.3635, "step": 9552 }, { "epoch": 0.6772775611485289, "grad_norm": 3.5934859997308206, "learning_rate": 2.4917184585382243e-06, "loss": 0.3883, "step": 9553 }, { "epoch": 0.6773484579936193, "grad_norm": 7.673417647158929, "learning_rate": 2.4907252901953264e-06, "loss": 0.3556, "step": 9554 }, { "epoch": 0.6774193548387096, "grad_norm": 3.029631618079353, "learning_rate": 2.4897322541684015e-06, "loss": 0.3928, "step": 9555 }, { "epoch": 0.6774902516838001, "grad_norm": 4.419339764118007, "learning_rate": 2.4887393505098107e-06, "loss": 0.3997, "step": 9556 }, { "epoch": 0.6775611485288905, "grad_norm": 2.775768000648895, "learning_rate": 2.487746579271909e-06, "loss": 0.3498, "step": 9557 }, { "epoch": 0.6776320453739808, "grad_norm": 4.23095546057433, "learning_rate": 2.4867539405070477e-06, "loss": 0.4097, "step": 9558 }, { "epoch": 0.6777029422190712, "grad_norm": 1.920800362693557, "learning_rate": 2.4857614342675714e-06, "loss": 0.413, "step": 9559 }, { "epoch": 0.6777738390641617, "grad_norm": 4.056317619498693, "learning_rate": 2.4847690606058135e-06, "loss": 0.3882, "step": 9560 }, { "epoch": 0.677844735909252, "grad_norm": 4.352877771531661, "learning_rate": 2.483776819574101e-06, "loss": 0.4362, "step": 9561 }, { "epoch": 0.6779156327543424, "grad_norm": 4.567095559869581, "learning_rate": 2.482784711224759e-06, "loss": 0.4163, "step": 9562 }, { "epoch": 0.6779865295994328, "grad_norm": 2.237405181067523, "learning_rate": 2.4817927356101e-06, "loss": 0.4013, "step": 9563 }, { "epoch": 0.6780574264445233, "grad_norm": 3.438465455501367, "learning_rate": 2.4808008927824313e-06, "loss": 0.3807, "step": 9564 }, { "epoch": 0.6781283232896136, "grad_norm": 5.156375036940389, "learning_rate": 2.4798091827940557e-06, "loss": 0.4277, "step": 9565 }, { "epoch": 0.678199220134704, "grad_norm": 8.629811429737194, "learning_rate": 2.4788176056972637e-06, "loss": 0.4921, "step": 9566 }, { "epoch": 0.6782701169797944, "grad_norm": 3.5127956034377164, "learning_rate": 2.4778261615443455e-06, "loss": 0.3655, "step": 9567 }, { "epoch": 0.6783410138248848, "grad_norm": 4.96224937767504, "learning_rate": 2.4768348503875787e-06, "loss": 0.391, "step": 9568 }, { "epoch": 0.6784119106699752, "grad_norm": 2.4639598405178686, "learning_rate": 2.4758436722792347e-06, "loss": 0.3773, "step": 9569 }, { "epoch": 0.6784828075150656, "grad_norm": 3.3739118053080945, "learning_rate": 2.47485262727158e-06, "loss": 0.4112, "step": 9570 }, { "epoch": 0.6785537043601559, "grad_norm": 2.9158025005319117, "learning_rate": 2.4738617154168777e-06, "loss": 0.4449, "step": 9571 }, { "epoch": 0.6786246012052464, "grad_norm": 1.531616022013363, "learning_rate": 2.4728709367673713e-06, "loss": 0.3881, "step": 9572 }, { "epoch": 0.6786954980503368, "grad_norm": 5.9160240526328485, "learning_rate": 2.4718802913753095e-06, "loss": 0.4234, "step": 9573 }, { "epoch": 0.6787663948954271, "grad_norm": 0.9378293491534304, "learning_rate": 2.470889779292931e-06, "loss": 0.6138, "step": 9574 }, { "epoch": 0.6788372917405175, "grad_norm": 2.694812768226434, "learning_rate": 2.4698994005724653e-06, "loss": 0.399, "step": 9575 }, { "epoch": 0.678908188585608, "grad_norm": 1.342672615367106, "learning_rate": 2.468909155266133e-06, "loss": 0.3863, "step": 9576 }, { "epoch": 0.6789790854306983, "grad_norm": 3.0253122289047014, "learning_rate": 2.467919043426153e-06, "loss": 0.3396, "step": 9577 }, { "epoch": 0.6790499822757887, "grad_norm": 2.6622729470968642, "learning_rate": 2.466929065104737e-06, "loss": 0.4581, "step": 9578 }, { "epoch": 0.6791208791208792, "grad_norm": 3.9182526395494635, "learning_rate": 2.465939220354084e-06, "loss": 0.3839, "step": 9579 }, { "epoch": 0.6791917759659695, "grad_norm": 2.542860227733223, "learning_rate": 2.464949509226388e-06, "loss": 0.3708, "step": 9580 }, { "epoch": 0.6792626728110599, "grad_norm": 5.202340911941215, "learning_rate": 2.463959931773842e-06, "loss": 0.3561, "step": 9581 }, { "epoch": 0.6793335696561503, "grad_norm": 1.0231834981692098, "learning_rate": 2.4629704880486244e-06, "loss": 0.6642, "step": 9582 }, { "epoch": 0.6794044665012406, "grad_norm": 2.8828749931077895, "learning_rate": 2.461981178102907e-06, "loss": 0.3496, "step": 9583 }, { "epoch": 0.6794753633463311, "grad_norm": 3.479794663343542, "learning_rate": 2.4609920019888613e-06, "loss": 0.4092, "step": 9584 }, { "epoch": 0.6795462601914215, "grad_norm": 3.951063289520757, "learning_rate": 2.4600029597586434e-06, "loss": 0.3828, "step": 9585 }, { "epoch": 0.6796171570365118, "grad_norm": 5.781551863034399, "learning_rate": 2.45901405146441e-06, "loss": 0.3923, "step": 9586 }, { "epoch": 0.6796880538816022, "grad_norm": 4.344038724974512, "learning_rate": 2.4580252771583034e-06, "loss": 0.3655, "step": 9587 }, { "epoch": 0.6797589507266927, "grad_norm": 2.791729524817896, "learning_rate": 2.457036636892466e-06, "loss": 0.4251, "step": 9588 }, { "epoch": 0.6798298475717831, "grad_norm": 2.134614346157033, "learning_rate": 2.4560481307190253e-06, "loss": 0.4138, "step": 9589 }, { "epoch": 0.6799007444168734, "grad_norm": 5.602504443526653, "learning_rate": 2.455059758690111e-06, "loss": 0.4042, "step": 9590 }, { "epoch": 0.6799716412619639, "grad_norm": 4.113675455857003, "learning_rate": 2.4540715208578377e-06, "loss": 0.3971, "step": 9591 }, { "epoch": 0.6800425381070543, "grad_norm": 2.2912681119940066, "learning_rate": 2.4530834172743147e-06, "loss": 0.4157, "step": 9592 }, { "epoch": 0.6801134349521446, "grad_norm": 4.061415319980971, "learning_rate": 2.4520954479916485e-06, "loss": 0.3955, "step": 9593 }, { "epoch": 0.680184331797235, "grad_norm": 6.889879516749443, "learning_rate": 2.4511076130619343e-06, "loss": 0.3671, "step": 9594 }, { "epoch": 0.6802552286423255, "grad_norm": 3.034619966274231, "learning_rate": 2.4501199125372593e-06, "loss": 0.3818, "step": 9595 }, { "epoch": 0.6803261254874158, "grad_norm": 3.4658341008873155, "learning_rate": 2.4491323464697075e-06, "loss": 0.3834, "step": 9596 }, { "epoch": 0.6803970223325062, "grad_norm": 3.413810657927633, "learning_rate": 2.448144914911358e-06, "loss": 0.3844, "step": 9597 }, { "epoch": 0.6804679191775966, "grad_norm": 3.0911813115411357, "learning_rate": 2.4471576179142708e-06, "loss": 0.4007, "step": 9598 }, { "epoch": 0.680538816022687, "grad_norm": 3.8337671017316763, "learning_rate": 2.446170455530511e-06, "loss": 0.427, "step": 9599 }, { "epoch": 0.6806097128677774, "grad_norm": 2.3314686386237504, "learning_rate": 2.4451834278121344e-06, "loss": 0.3557, "step": 9600 }, { "epoch": 0.6806806097128678, "grad_norm": 1.7842276003751913, "learning_rate": 2.444196534811186e-06, "loss": 0.3903, "step": 9601 }, { "epoch": 0.6807515065579581, "grad_norm": 4.13026008733355, "learning_rate": 2.4432097765797026e-06, "loss": 0.3824, "step": 9602 }, { "epoch": 0.6808224034030486, "grad_norm": 2.9617326455579627, "learning_rate": 2.442223153169719e-06, "loss": 0.4059, "step": 9603 }, { "epoch": 0.680893300248139, "grad_norm": 2.0850152679682274, "learning_rate": 2.441236664633263e-06, "loss": 0.393, "step": 9604 }, { "epoch": 0.6809641970932293, "grad_norm": 2.7825740557414447, "learning_rate": 2.4402503110223507e-06, "loss": 0.4602, "step": 9605 }, { "epoch": 0.6810350939383197, "grad_norm": 3.2560850906452465, "learning_rate": 2.4392640923889917e-06, "loss": 0.3936, "step": 9606 }, { "epoch": 0.6811059907834102, "grad_norm": 3.462318381468347, "learning_rate": 2.4382780087851926e-06, "loss": 0.4127, "step": 9607 }, { "epoch": 0.6811768876285005, "grad_norm": 4.771333119463003, "learning_rate": 2.4372920602629503e-06, "loss": 0.4244, "step": 9608 }, { "epoch": 0.6812477844735909, "grad_norm": 2.5516307638010134, "learning_rate": 2.4363062468742514e-06, "loss": 0.4079, "step": 9609 }, { "epoch": 0.6813186813186813, "grad_norm": 2.6582753096335807, "learning_rate": 2.435320568671083e-06, "loss": 0.3754, "step": 9610 }, { "epoch": 0.6813895781637717, "grad_norm": 3.1873766549987614, "learning_rate": 2.434335025705416e-06, "loss": 0.3853, "step": 9611 }, { "epoch": 0.6814604750088621, "grad_norm": 6.868356632032778, "learning_rate": 2.4333496180292238e-06, "loss": 0.4732, "step": 9612 }, { "epoch": 0.6815313718539525, "grad_norm": 2.6786463920001635, "learning_rate": 2.4323643456944657e-06, "loss": 0.337, "step": 9613 }, { "epoch": 0.681602268699043, "grad_norm": 14.02044176820418, "learning_rate": 2.431379208753093e-06, "loss": 0.3911, "step": 9614 }, { "epoch": 0.6816731655441333, "grad_norm": 2.84378765194414, "learning_rate": 2.4303942072570553e-06, "loss": 0.4453, "step": 9615 }, { "epoch": 0.6817440623892237, "grad_norm": 2.1850262335830877, "learning_rate": 2.4294093412582937e-06, "loss": 0.3865, "step": 9616 }, { "epoch": 0.6818149592343141, "grad_norm": 2.319816388706004, "learning_rate": 2.42842461080874e-06, "loss": 0.3759, "step": 9617 }, { "epoch": 0.6818858560794044, "grad_norm": 2.8146735405188457, "learning_rate": 2.427440015960318e-06, "loss": 0.3955, "step": 9618 }, { "epoch": 0.6819567529244949, "grad_norm": 3.7771720309454686, "learning_rate": 2.426455556764949e-06, "loss": 0.4091, "step": 9619 }, { "epoch": 0.6820276497695853, "grad_norm": 0.9879251160221333, "learning_rate": 2.425471233274542e-06, "loss": 0.5973, "step": 9620 }, { "epoch": 0.6820985466146756, "grad_norm": 2.9266891061127858, "learning_rate": 2.4244870455410014e-06, "loss": 0.4594, "step": 9621 }, { "epoch": 0.682169443459766, "grad_norm": 3.9470470725060443, "learning_rate": 2.4235029936162236e-06, "loss": 0.3898, "step": 9622 }, { "epoch": 0.6822403403048565, "grad_norm": 2.317348403369152, "learning_rate": 2.4225190775521017e-06, "loss": 0.4281, "step": 9623 }, { "epoch": 0.6823112371499468, "grad_norm": 5.621661689453937, "learning_rate": 2.421535297400516e-06, "loss": 0.3951, "step": 9624 }, { "epoch": 0.6823821339950372, "grad_norm": 5.369820928957669, "learning_rate": 2.420551653213341e-06, "loss": 0.414, "step": 9625 }, { "epoch": 0.6824530308401277, "grad_norm": 2.3905970210509726, "learning_rate": 2.4195681450424474e-06, "loss": 0.3986, "step": 9626 }, { "epoch": 0.682523927685218, "grad_norm": 5.731325370043687, "learning_rate": 2.418584772939695e-06, "loss": 0.3671, "step": 9627 }, { "epoch": 0.6825948245303084, "grad_norm": 6.741288910601119, "learning_rate": 2.417601536956936e-06, "loss": 0.3422, "step": 9628 }, { "epoch": 0.6826657213753988, "grad_norm": 1.8926444813743253, "learning_rate": 2.416618437146021e-06, "loss": 0.4106, "step": 9629 }, { "epoch": 0.6827366182204891, "grad_norm": 4.836819031561167, "learning_rate": 2.415635473558786e-06, "loss": 0.4112, "step": 9630 }, { "epoch": 0.6828075150655796, "grad_norm": 6.214757900859927, "learning_rate": 2.4146526462470667e-06, "loss": 0.3609, "step": 9631 }, { "epoch": 0.68287841191067, "grad_norm": 2.220459915889747, "learning_rate": 2.4136699552626846e-06, "loss": 0.3327, "step": 9632 }, { "epoch": 0.6829493087557603, "grad_norm": 2.2689955409288913, "learning_rate": 2.4126874006574617e-06, "loss": 0.4146, "step": 9633 }, { "epoch": 0.6830202056008507, "grad_norm": 3.197297534631032, "learning_rate": 2.411704982483206e-06, "loss": 0.4032, "step": 9634 }, { "epoch": 0.6830911024459412, "grad_norm": 6.42717413767818, "learning_rate": 2.4107227007917228e-06, "loss": 0.3772, "step": 9635 }, { "epoch": 0.6831619992910315, "grad_norm": 2.7405749602010676, "learning_rate": 2.4097405556348087e-06, "loss": 0.4358, "step": 9636 }, { "epoch": 0.6832328961361219, "grad_norm": 1.7462656915344432, "learning_rate": 2.4087585470642494e-06, "loss": 0.3836, "step": 9637 }, { "epoch": 0.6833037929812124, "grad_norm": 0.9080212831960239, "learning_rate": 2.407776675131832e-06, "loss": 0.6498, "step": 9638 }, { "epoch": 0.6833746898263028, "grad_norm": 11.506933043292136, "learning_rate": 2.406794939889329e-06, "loss": 0.4497, "step": 9639 }, { "epoch": 0.6834455866713931, "grad_norm": 2.0439624715868403, "learning_rate": 2.4058133413885064e-06, "loss": 0.3692, "step": 9640 }, { "epoch": 0.6835164835164835, "grad_norm": 3.5743675294650656, "learning_rate": 2.4048318796811267e-06, "loss": 0.4085, "step": 9641 }, { "epoch": 0.683587380361574, "grad_norm": 2.1493256738820437, "learning_rate": 2.4038505548189445e-06, "loss": 0.3948, "step": 9642 }, { "epoch": 0.6836582772066643, "grad_norm": 3.621587726055771, "learning_rate": 2.402869366853704e-06, "loss": 0.4084, "step": 9643 }, { "epoch": 0.6837291740517547, "grad_norm": 3.174217212806004, "learning_rate": 2.401888315837142e-06, "loss": 0.4325, "step": 9644 }, { "epoch": 0.6838000708968451, "grad_norm": 3.341409016330197, "learning_rate": 2.4009074018209945e-06, "loss": 0.3729, "step": 9645 }, { "epoch": 0.6838709677419355, "grad_norm": 2.440075199537822, "learning_rate": 2.399926624856983e-06, "loss": 0.3464, "step": 9646 }, { "epoch": 0.6839418645870259, "grad_norm": 3.8777476891370606, "learning_rate": 2.3989459849968235e-06, "loss": 0.3873, "step": 9647 }, { "epoch": 0.6840127614321163, "grad_norm": 6.848724973781554, "learning_rate": 2.3979654822922273e-06, "loss": 0.3851, "step": 9648 }, { "epoch": 0.6840836582772066, "grad_norm": 4.009915399625133, "learning_rate": 2.3969851167948992e-06, "loss": 0.3869, "step": 9649 }, { "epoch": 0.6841545551222971, "grad_norm": 2.178350918325336, "learning_rate": 2.396004888556533e-06, "loss": 0.355, "step": 9650 }, { "epoch": 0.6842254519673875, "grad_norm": 11.56266642409139, "learning_rate": 2.3950247976288144e-06, "loss": 0.4263, "step": 9651 }, { "epoch": 0.6842963488124778, "grad_norm": 5.413118861993935, "learning_rate": 2.3940448440634283e-06, "loss": 0.3956, "step": 9652 }, { "epoch": 0.6843672456575682, "grad_norm": 3.053040548152535, "learning_rate": 2.3930650279120464e-06, "loss": 0.3934, "step": 9653 }, { "epoch": 0.6844381425026587, "grad_norm": 3.3966891026786747, "learning_rate": 2.392085349226334e-06, "loss": 0.3881, "step": 9654 }, { "epoch": 0.684509039347749, "grad_norm": 3.152069512756881, "learning_rate": 2.3911058080579537e-06, "loss": 0.4113, "step": 9655 }, { "epoch": 0.6845799361928394, "grad_norm": 10.306741457432981, "learning_rate": 2.390126404458553e-06, "loss": 0.4304, "step": 9656 }, { "epoch": 0.6846508330379298, "grad_norm": 6.485919159666973, "learning_rate": 2.389147138479781e-06, "loss": 0.5227, "step": 9657 }, { "epoch": 0.6847217298830202, "grad_norm": 3.7792153408396256, "learning_rate": 2.3881680101732734e-06, "loss": 0.4025, "step": 9658 }, { "epoch": 0.6847926267281106, "grad_norm": 2.2534154524982046, "learning_rate": 2.3871890195906575e-06, "loss": 0.3991, "step": 9659 }, { "epoch": 0.684863523573201, "grad_norm": 4.900078298050718, "learning_rate": 2.38621016678356e-06, "loss": 0.408, "step": 9660 }, { "epoch": 0.6849344204182913, "grad_norm": 2.6830836664654285, "learning_rate": 2.385231451803597e-06, "loss": 0.402, "step": 9661 }, { "epoch": 0.6850053172633818, "grad_norm": 10.00267292723257, "learning_rate": 2.3842528747023747e-06, "loss": 0.4506, "step": 9662 }, { "epoch": 0.6850762141084722, "grad_norm": 4.375130723339771, "learning_rate": 2.3832744355314935e-06, "loss": 0.4082, "step": 9663 }, { "epoch": 0.6851471109535626, "grad_norm": 3.552503007968666, "learning_rate": 2.3822961343425505e-06, "loss": 0.393, "step": 9664 }, { "epoch": 0.6852180077986529, "grad_norm": 2.2078694334981024, "learning_rate": 2.38131797118713e-06, "loss": 0.4074, "step": 9665 }, { "epoch": 0.6852889046437434, "grad_norm": 6.871780387263573, "learning_rate": 2.38033994611681e-06, "loss": 0.3675, "step": 9666 }, { "epoch": 0.6853598014888338, "grad_norm": 5.7779974392416795, "learning_rate": 2.379362059183164e-06, "loss": 0.4482, "step": 9667 }, { "epoch": 0.6854306983339241, "grad_norm": 3.1547314049358137, "learning_rate": 2.3783843104377585e-06, "loss": 0.4126, "step": 9668 }, { "epoch": 0.6855015951790145, "grad_norm": 3.174880693058294, "learning_rate": 2.3774066999321497e-06, "loss": 0.4484, "step": 9669 }, { "epoch": 0.685572492024105, "grad_norm": 3.3871092840385466, "learning_rate": 2.376429227717885e-06, "loss": 0.3888, "step": 9670 }, { "epoch": 0.6856433888691953, "grad_norm": 4.657973703973481, "learning_rate": 2.3754518938465123e-06, "loss": 0.3867, "step": 9671 }, { "epoch": 0.6857142857142857, "grad_norm": 2.931962299799606, "learning_rate": 2.3744746983695637e-06, "loss": 0.3906, "step": 9672 }, { "epoch": 0.6857851825593761, "grad_norm": 3.17521842508621, "learning_rate": 2.3734976413385667e-06, "loss": 0.4006, "step": 9673 }, { "epoch": 0.6858560794044665, "grad_norm": 1.9336864178312356, "learning_rate": 2.3725207228050457e-06, "loss": 0.3315, "step": 9674 }, { "epoch": 0.6859269762495569, "grad_norm": 3.9711315454118674, "learning_rate": 2.3715439428205107e-06, "loss": 0.4328, "step": 9675 }, { "epoch": 0.6859978730946473, "grad_norm": 3.4300356397084664, "learning_rate": 2.370567301436471e-06, "loss": 0.4175, "step": 9676 }, { "epoch": 0.6860687699397376, "grad_norm": 1.8129130480074482, "learning_rate": 2.369590798704423e-06, "loss": 0.3477, "step": 9677 }, { "epoch": 0.6861396667848281, "grad_norm": 3.5015201400686533, "learning_rate": 2.3686144346758617e-06, "loss": 0.3809, "step": 9678 }, { "epoch": 0.6862105636299185, "grad_norm": 2.2285398043956253, "learning_rate": 2.3676382094022682e-06, "loss": 0.3673, "step": 9679 }, { "epoch": 0.6862814604750088, "grad_norm": 11.204719616259322, "learning_rate": 2.366662122935123e-06, "loss": 0.3724, "step": 9680 }, { "epoch": 0.6863523573200992, "grad_norm": 2.836458933098625, "learning_rate": 2.365686175325893e-06, "loss": 0.3637, "step": 9681 }, { "epoch": 0.6864232541651897, "grad_norm": 2.556582048190005, "learning_rate": 2.364710366626041e-06, "loss": 0.4152, "step": 9682 }, { "epoch": 0.68649415101028, "grad_norm": 4.698539649408908, "learning_rate": 2.363734696887024e-06, "loss": 0.4062, "step": 9683 }, { "epoch": 0.6865650478553704, "grad_norm": 7.664986105513075, "learning_rate": 2.3627591661602882e-06, "loss": 0.3619, "step": 9684 }, { "epoch": 0.6866359447004609, "grad_norm": 5.545841472848135, "learning_rate": 2.3617837744972733e-06, "loss": 0.4043, "step": 9685 }, { "epoch": 0.6867068415455512, "grad_norm": 7.888736106190476, "learning_rate": 2.3608085219494136e-06, "loss": 0.4157, "step": 9686 }, { "epoch": 0.6867777383906416, "grad_norm": 5.190462505380251, "learning_rate": 2.3598334085681367e-06, "loss": 0.3974, "step": 9687 }, { "epoch": 0.686848635235732, "grad_norm": 2.1269520208035804, "learning_rate": 2.3588584344048595e-06, "loss": 0.4175, "step": 9688 }, { "epoch": 0.6869195320808223, "grad_norm": 1.8706099825915137, "learning_rate": 2.3578835995109906e-06, "loss": 0.361, "step": 9689 }, { "epoch": 0.6869904289259128, "grad_norm": 4.548302037578225, "learning_rate": 2.3569089039379366e-06, "loss": 0.4348, "step": 9690 }, { "epoch": 0.6870613257710032, "grad_norm": 3.0165302744770477, "learning_rate": 2.3559343477370973e-06, "loss": 0.4444, "step": 9691 }, { "epoch": 0.6871322226160936, "grad_norm": 4.290601878167679, "learning_rate": 2.3549599309598535e-06, "loss": 0.3936, "step": 9692 }, { "epoch": 0.687203119461184, "grad_norm": 3.683444612425175, "learning_rate": 2.353985653657592e-06, "loss": 0.3692, "step": 9693 }, { "epoch": 0.6872740163062744, "grad_norm": 2.9932657810722425, "learning_rate": 2.3530115158816886e-06, "loss": 0.3473, "step": 9694 }, { "epoch": 0.6873449131513648, "grad_norm": 2.609118036628548, "learning_rate": 2.3520375176835074e-06, "loss": 0.3836, "step": 9695 }, { "epoch": 0.6874158099964551, "grad_norm": 2.9198726915591724, "learning_rate": 2.351063659114408e-06, "loss": 0.3244, "step": 9696 }, { "epoch": 0.6874867068415456, "grad_norm": 2.9695000324619727, "learning_rate": 2.3500899402257454e-06, "loss": 0.4026, "step": 9697 }, { "epoch": 0.687557603686636, "grad_norm": 2.5800802986083573, "learning_rate": 2.3491163610688613e-06, "loss": 0.3973, "step": 9698 }, { "epoch": 0.6876285005317263, "grad_norm": 3.052933130174304, "learning_rate": 2.3481429216950957e-06, "loss": 0.4117, "step": 9699 }, { "epoch": 0.6876993973768167, "grad_norm": 5.2652057259072444, "learning_rate": 2.347169622155779e-06, "loss": 0.4538, "step": 9700 }, { "epoch": 0.6877702942219072, "grad_norm": 2.906765303818849, "learning_rate": 2.3461964625022306e-06, "loss": 0.4117, "step": 9701 }, { "epoch": 0.6878411910669975, "grad_norm": 3.4668441705002317, "learning_rate": 2.34522344278577e-06, "loss": 0.4158, "step": 9702 }, { "epoch": 0.6879120879120879, "grad_norm": 5.1284029663219615, "learning_rate": 2.3442505630577033e-06, "loss": 0.3926, "step": 9703 }, { "epoch": 0.6879829847571783, "grad_norm": 10.088040953106054, "learning_rate": 2.3432778233693303e-06, "loss": 0.4009, "step": 9704 }, { "epoch": 0.6880538816022687, "grad_norm": 2.34328881881985, "learning_rate": 2.342305223771945e-06, "loss": 0.3321, "step": 9705 }, { "epoch": 0.6881247784473591, "grad_norm": 4.4144695494988015, "learning_rate": 2.341332764316836e-06, "loss": 0.4006, "step": 9706 }, { "epoch": 0.6881956752924495, "grad_norm": 1.8537094912348824, "learning_rate": 2.340360445055279e-06, "loss": 0.3683, "step": 9707 }, { "epoch": 0.6882665721375398, "grad_norm": 2.814521306423334, "learning_rate": 2.3393882660385442e-06, "loss": 0.3488, "step": 9708 }, { "epoch": 0.6883374689826303, "grad_norm": 2.8149666632137773, "learning_rate": 2.338416227317899e-06, "loss": 0.413, "step": 9709 }, { "epoch": 0.6884083658277207, "grad_norm": 3.382439678285888, "learning_rate": 2.3374443289445973e-06, "loss": 0.376, "step": 9710 }, { "epoch": 0.688479262672811, "grad_norm": 3.5984384256443462, "learning_rate": 2.3364725709698864e-06, "loss": 0.374, "step": 9711 }, { "epoch": 0.6885501595179014, "grad_norm": 2.5704733060936475, "learning_rate": 2.3355009534450102e-06, "loss": 0.3479, "step": 9712 }, { "epoch": 0.6886210563629919, "grad_norm": 4.526861889161354, "learning_rate": 2.334529476421204e-06, "loss": 0.403, "step": 9713 }, { "epoch": 0.6886919532080822, "grad_norm": 2.062235151464842, "learning_rate": 2.333558139949693e-06, "loss": 0.3974, "step": 9714 }, { "epoch": 0.6887628500531726, "grad_norm": 2.9968245835208265, "learning_rate": 2.332586944081695e-06, "loss": 0.4484, "step": 9715 }, { "epoch": 0.688833746898263, "grad_norm": 2.39685325725264, "learning_rate": 2.3316158888684246e-06, "loss": 0.3481, "step": 9716 }, { "epoch": 0.6889046437433535, "grad_norm": 2.5109806004969966, "learning_rate": 2.330644974361086e-06, "loss": 0.3926, "step": 9717 }, { "epoch": 0.6889755405884438, "grad_norm": 2.6737429235551256, "learning_rate": 2.329674200610873e-06, "loss": 0.4279, "step": 9718 }, { "epoch": 0.6890464374335342, "grad_norm": 4.5725506560881595, "learning_rate": 2.3287035676689775e-06, "loss": 0.3506, "step": 9719 }, { "epoch": 0.6891173342786246, "grad_norm": 3.0679507924373475, "learning_rate": 2.3277330755865834e-06, "loss": 0.318, "step": 9720 }, { "epoch": 0.689188231123715, "grad_norm": 2.7185006147976916, "learning_rate": 2.3267627244148642e-06, "loss": 0.3876, "step": 9721 }, { "epoch": 0.6892591279688054, "grad_norm": 3.8147871635939947, "learning_rate": 2.325792514204985e-06, "loss": 0.3801, "step": 9722 }, { "epoch": 0.6893300248138958, "grad_norm": 1.0216119414218507, "learning_rate": 2.3248224450081092e-06, "loss": 0.6245, "step": 9723 }, { "epoch": 0.6894009216589861, "grad_norm": 4.701697764166465, "learning_rate": 2.323852516875386e-06, "loss": 0.4429, "step": 9724 }, { "epoch": 0.6894718185040766, "grad_norm": 2.6333327062531153, "learning_rate": 2.322882729857965e-06, "loss": 0.4569, "step": 9725 }, { "epoch": 0.689542715349167, "grad_norm": 2.4094077947411514, "learning_rate": 2.3219130840069797e-06, "loss": 0.3462, "step": 9726 }, { "epoch": 0.6896136121942573, "grad_norm": 4.102121597309321, "learning_rate": 2.32094357937356e-06, "loss": 0.404, "step": 9727 }, { "epoch": 0.6896845090393477, "grad_norm": 3.0464727713250355, "learning_rate": 2.3199742160088327e-06, "loss": 0.4093, "step": 9728 }, { "epoch": 0.6897554058844382, "grad_norm": 4.491265842913566, "learning_rate": 2.3190049939639103e-06, "loss": 0.3737, "step": 9729 }, { "epoch": 0.6898263027295285, "grad_norm": 3.646513321275043, "learning_rate": 2.318035913289899e-06, "loss": 0.4068, "step": 9730 }, { "epoch": 0.6898971995746189, "grad_norm": 5.001876878539244, "learning_rate": 2.3170669740379006e-06, "loss": 0.4236, "step": 9731 }, { "epoch": 0.6899680964197094, "grad_norm": 2.64084882956935, "learning_rate": 2.3160981762590107e-06, "loss": 0.3677, "step": 9732 }, { "epoch": 0.6900389932647997, "grad_norm": 5.0942881012411885, "learning_rate": 2.3151295200043123e-06, "loss": 0.4356, "step": 9733 }, { "epoch": 0.6901098901098901, "grad_norm": 4.160068214054469, "learning_rate": 2.314161005324882e-06, "loss": 0.417, "step": 9734 }, { "epoch": 0.6901807869549805, "grad_norm": 3.1744322777435947, "learning_rate": 2.313192632271792e-06, "loss": 0.365, "step": 9735 }, { "epoch": 0.6902516838000708, "grad_norm": 2.962172121154923, "learning_rate": 2.312224400896108e-06, "loss": 0.4443, "step": 9736 }, { "epoch": 0.6903225806451613, "grad_norm": 0.9501782116895362, "learning_rate": 2.31125631124888e-06, "loss": 0.6583, "step": 9737 }, { "epoch": 0.6903934774902517, "grad_norm": 3.308504909952215, "learning_rate": 2.310288363381159e-06, "loss": 0.3564, "step": 9738 }, { "epoch": 0.690464374335342, "grad_norm": 4.546936783269387, "learning_rate": 2.309320557343987e-06, "loss": 0.4361, "step": 9739 }, { "epoch": 0.6905352711804325, "grad_norm": 67.44035748408076, "learning_rate": 2.3083528931883957e-06, "loss": 0.4313, "step": 9740 }, { "epoch": 0.6906061680255229, "grad_norm": 2.305306898053508, "learning_rate": 2.3073853709654094e-06, "loss": 0.3565, "step": 9741 }, { "epoch": 0.6906770648706133, "grad_norm": 3.3655125474173575, "learning_rate": 2.3064179907260494e-06, "loss": 0.4215, "step": 9742 }, { "epoch": 0.6907479617157036, "grad_norm": 3.9357030596199434, "learning_rate": 2.305450752521323e-06, "loss": 0.3833, "step": 9743 }, { "epoch": 0.6908188585607941, "grad_norm": 5.2786217086779175, "learning_rate": 2.304483656402236e-06, "loss": 0.397, "step": 9744 }, { "epoch": 0.6908897554058845, "grad_norm": 2.2987135941399495, "learning_rate": 2.3035167024197835e-06, "loss": 0.3593, "step": 9745 }, { "epoch": 0.6909606522509748, "grad_norm": 3.425413799502903, "learning_rate": 2.3025498906249522e-06, "loss": 0.4453, "step": 9746 }, { "epoch": 0.6910315490960652, "grad_norm": 3.3803382829997974, "learning_rate": 2.3015832210687254e-06, "loss": 0.3656, "step": 9747 }, { "epoch": 0.6911024459411557, "grad_norm": 3.0619410401565497, "learning_rate": 2.300616693802073e-06, "loss": 0.444, "step": 9748 }, { "epoch": 0.691173342786246, "grad_norm": 2.1531821782289344, "learning_rate": 2.2996503088759646e-06, "loss": 0.4007, "step": 9749 }, { "epoch": 0.6912442396313364, "grad_norm": 2.8537901907633243, "learning_rate": 2.2986840663413553e-06, "loss": 0.4345, "step": 9750 }, { "epoch": 0.6913151364764268, "grad_norm": 16.865776501005406, "learning_rate": 2.297717966249198e-06, "loss": 0.3964, "step": 9751 }, { "epoch": 0.6913860333215172, "grad_norm": 4.058629812341572, "learning_rate": 2.296752008650435e-06, "loss": 0.4216, "step": 9752 }, { "epoch": 0.6914569301666076, "grad_norm": 6.850393072336841, "learning_rate": 2.2957861935960007e-06, "loss": 0.4002, "step": 9753 }, { "epoch": 0.691527827011698, "grad_norm": 2.1751200503373957, "learning_rate": 2.294820521136824e-06, "loss": 0.3881, "step": 9754 }, { "epoch": 0.6915987238567883, "grad_norm": 2.1428354431677175, "learning_rate": 2.29385499132383e-06, "loss": 0.3773, "step": 9755 }, { "epoch": 0.6916696207018788, "grad_norm": 2.784607398814449, "learning_rate": 2.2928896042079234e-06, "loss": 0.3591, "step": 9756 }, { "epoch": 0.6917405175469692, "grad_norm": 3.6528400964329655, "learning_rate": 2.291924359840015e-06, "loss": 0.3905, "step": 9757 }, { "epoch": 0.6918114143920595, "grad_norm": 2.1010961521959812, "learning_rate": 2.290959258271003e-06, "loss": 0.3469, "step": 9758 }, { "epoch": 0.6918823112371499, "grad_norm": 2.6793814268759792, "learning_rate": 2.2899942995517783e-06, "loss": 0.3883, "step": 9759 }, { "epoch": 0.6919532080822404, "grad_norm": 7.15966847476014, "learning_rate": 2.28902948373322e-06, "loss": 0.4146, "step": 9760 }, { "epoch": 0.6920241049273307, "grad_norm": 4.315021879168341, "learning_rate": 2.288064810866208e-06, "loss": 0.3496, "step": 9761 }, { "epoch": 0.6920950017724211, "grad_norm": 2.462409997144724, "learning_rate": 2.2871002810016087e-06, "loss": 0.3289, "step": 9762 }, { "epoch": 0.6921658986175115, "grad_norm": 4.315711682092965, "learning_rate": 2.28613589419028e-06, "loss": 0.366, "step": 9763 }, { "epoch": 0.6922367954626019, "grad_norm": 4.388479417694525, "learning_rate": 2.2851716504830773e-06, "loss": 0.3907, "step": 9764 }, { "epoch": 0.6923076923076923, "grad_norm": 3.064926209336052, "learning_rate": 2.2842075499308475e-06, "loss": 0.4021, "step": 9765 }, { "epoch": 0.6923785891527827, "grad_norm": 6.95671741617765, "learning_rate": 2.2832435925844266e-06, "loss": 0.4171, "step": 9766 }, { "epoch": 0.6924494859978731, "grad_norm": 3.6479134815065253, "learning_rate": 2.2822797784946425e-06, "loss": 0.3774, "step": 9767 }, { "epoch": 0.6925203828429635, "grad_norm": 2.561952759408402, "learning_rate": 2.281316107712322e-06, "loss": 0.3362, "step": 9768 }, { "epoch": 0.6925912796880539, "grad_norm": 3.8926934686567467, "learning_rate": 2.2803525802882764e-06, "loss": 0.3991, "step": 9769 }, { "epoch": 0.6926621765331443, "grad_norm": 3.7421865895757054, "learning_rate": 2.2793891962733167e-06, "loss": 0.3992, "step": 9770 }, { "epoch": 0.6927330733782346, "grad_norm": 1.8714678642106228, "learning_rate": 2.2784259557182415e-06, "loss": 0.3845, "step": 9771 }, { "epoch": 0.6928039702233251, "grad_norm": 6.958875200405998, "learning_rate": 2.2774628586738413e-06, "loss": 0.3442, "step": 9772 }, { "epoch": 0.6928748670684155, "grad_norm": 3.87996725189249, "learning_rate": 2.2764999051909036e-06, "loss": 0.3632, "step": 9773 }, { "epoch": 0.6929457639135058, "grad_norm": 1.058942901385652, "learning_rate": 2.2755370953202056e-06, "loss": 0.6626, "step": 9774 }, { "epoch": 0.6930166607585962, "grad_norm": 4.429797214096931, "learning_rate": 2.2745744291125136e-06, "loss": 0.3947, "step": 9775 }, { "epoch": 0.6930875576036867, "grad_norm": 7.082124237720656, "learning_rate": 2.2736119066185923e-06, "loss": 0.4271, "step": 9776 }, { "epoch": 0.693158454448777, "grad_norm": 6.519134244703244, "learning_rate": 2.272649527889198e-06, "loss": 0.358, "step": 9777 }, { "epoch": 0.6932293512938674, "grad_norm": 2.3506820057353193, "learning_rate": 2.2716872929750762e-06, "loss": 0.4167, "step": 9778 }, { "epoch": 0.6933002481389579, "grad_norm": 38.593130977993695, "learning_rate": 2.270725201926963e-06, "loss": 0.4299, "step": 9779 }, { "epoch": 0.6933711449840482, "grad_norm": 2.7225286625359613, "learning_rate": 2.269763254795594e-06, "loss": 0.4124, "step": 9780 }, { "epoch": 0.6934420418291386, "grad_norm": 3.2064454074611963, "learning_rate": 2.268801451631696e-06, "loss": 0.3651, "step": 9781 }, { "epoch": 0.693512938674229, "grad_norm": 2.2802141517682415, "learning_rate": 2.2678397924859784e-06, "loss": 0.4393, "step": 9782 }, { "epoch": 0.6935838355193193, "grad_norm": 5.059307032077088, "learning_rate": 2.2668782774091535e-06, "loss": 0.387, "step": 9783 }, { "epoch": 0.6936547323644098, "grad_norm": 5.616021652177827, "learning_rate": 2.2659169064519254e-06, "loss": 0.3553, "step": 9784 }, { "epoch": 0.6937256292095002, "grad_norm": 3.7939480945054216, "learning_rate": 2.264955679664985e-06, "loss": 0.3939, "step": 9785 }, { "epoch": 0.6937965260545905, "grad_norm": 2.323892078692441, "learning_rate": 2.2639945970990186e-06, "loss": 0.3645, "step": 9786 }, { "epoch": 0.693867422899681, "grad_norm": 2.498805247400102, "learning_rate": 2.263033658804706e-06, "loss": 0.3313, "step": 9787 }, { "epoch": 0.6939383197447714, "grad_norm": 5.379955677003238, "learning_rate": 2.262072864832716e-06, "loss": 0.3955, "step": 9788 }, { "epoch": 0.6940092165898617, "grad_norm": 17.1239237718211, "learning_rate": 2.261112215233716e-06, "loss": 0.4033, "step": 9789 }, { "epoch": 0.6940801134349521, "grad_norm": 4.413553327958214, "learning_rate": 2.2601517100583592e-06, "loss": 0.4174, "step": 9790 }, { "epoch": 0.6941510102800426, "grad_norm": 1.9588554853613824, "learning_rate": 2.259191349357292e-06, "loss": 0.3855, "step": 9791 }, { "epoch": 0.694221907125133, "grad_norm": 2.3181726168267813, "learning_rate": 2.2582311331811583e-06, "loss": 0.4075, "step": 9792 }, { "epoch": 0.6942928039702233, "grad_norm": 4.98733665735425, "learning_rate": 2.257271061580588e-06, "loss": 0.381, "step": 9793 }, { "epoch": 0.6943637008153137, "grad_norm": 3.693080805791954, "learning_rate": 2.2563111346062107e-06, "loss": 0.3877, "step": 9794 }, { "epoch": 0.6944345976604042, "grad_norm": 5.893819989327174, "learning_rate": 2.255351352308639e-06, "loss": 0.3867, "step": 9795 }, { "epoch": 0.6945054945054945, "grad_norm": 5.572143207364209, "learning_rate": 2.254391714738487e-06, "loss": 0.413, "step": 9796 }, { "epoch": 0.6945763913505849, "grad_norm": 21.012302520052792, "learning_rate": 2.2534322219463554e-06, "loss": 0.4088, "step": 9797 }, { "epoch": 0.6946472881956753, "grad_norm": 5.527488157795009, "learning_rate": 2.252472873982838e-06, "loss": 0.3738, "step": 9798 }, { "epoch": 0.6947181850407657, "grad_norm": 3.141158556049301, "learning_rate": 2.2515136708985226e-06, "loss": 0.3395, "step": 9799 }, { "epoch": 0.6947890818858561, "grad_norm": 2.421113100061405, "learning_rate": 2.2505546127439933e-06, "loss": 0.3742, "step": 9800 }, { "epoch": 0.6948599787309465, "grad_norm": 3.4412125013376795, "learning_rate": 2.2495956995698132e-06, "loss": 0.4089, "step": 9801 }, { "epoch": 0.6949308755760368, "grad_norm": 2.675001939401511, "learning_rate": 2.2486369314265516e-06, "loss": 0.3727, "step": 9802 }, { "epoch": 0.6950017724211273, "grad_norm": 2.6947832771577556, "learning_rate": 2.247678308364767e-06, "loss": 0.3819, "step": 9803 }, { "epoch": 0.6950726692662177, "grad_norm": 3.9141209492056226, "learning_rate": 2.2467198304350054e-06, "loss": 0.3808, "step": 9804 }, { "epoch": 0.695143566111308, "grad_norm": 4.246173325682569, "learning_rate": 2.2457614976878066e-06, "loss": 0.4076, "step": 9805 }, { "epoch": 0.6952144629563984, "grad_norm": 2.3153639899530987, "learning_rate": 2.2448033101737087e-06, "loss": 0.4183, "step": 9806 }, { "epoch": 0.6952853598014889, "grad_norm": 3.889272287109167, "learning_rate": 2.2438452679432326e-06, "loss": 0.4173, "step": 9807 }, { "epoch": 0.6953562566465792, "grad_norm": 2.8957006147193303, "learning_rate": 2.2428873710469017e-06, "loss": 0.4259, "step": 9808 }, { "epoch": 0.6954271534916696, "grad_norm": 5.338717259640101, "learning_rate": 2.241929619535222e-06, "loss": 0.4102, "step": 9809 }, { "epoch": 0.69549805033676, "grad_norm": 3.412123611524942, "learning_rate": 2.2409720134587003e-06, "loss": 0.3442, "step": 9810 }, { "epoch": 0.6955689471818504, "grad_norm": 2.4131908670848743, "learning_rate": 2.2400145528678306e-06, "loss": 0.3786, "step": 9811 }, { "epoch": 0.6956398440269408, "grad_norm": 2.444359462516729, "learning_rate": 2.2390572378130983e-06, "loss": 0.4281, "step": 9812 }, { "epoch": 0.6957107408720312, "grad_norm": 2.7163932094848215, "learning_rate": 2.2381000683449864e-06, "loss": 0.3966, "step": 9813 }, { "epoch": 0.6957816377171215, "grad_norm": 4.675831861096543, "learning_rate": 2.2371430445139648e-06, "loss": 0.3874, "step": 9814 }, { "epoch": 0.695852534562212, "grad_norm": 3.6054376436791102, "learning_rate": 2.236186166370501e-06, "loss": 0.4127, "step": 9815 }, { "epoch": 0.6959234314073024, "grad_norm": 4.210330900604331, "learning_rate": 2.2352294339650498e-06, "loss": 0.3672, "step": 9816 }, { "epoch": 0.6959943282523927, "grad_norm": 3.4849393941616715, "learning_rate": 2.234272847348059e-06, "loss": 0.3481, "step": 9817 }, { "epoch": 0.6960652250974831, "grad_norm": 8.758168519361233, "learning_rate": 2.2333164065699743e-06, "loss": 0.3995, "step": 9818 }, { "epoch": 0.6961361219425736, "grad_norm": 2.5997207755113187, "learning_rate": 2.2323601116812265e-06, "loss": 0.3758, "step": 9819 }, { "epoch": 0.696207018787664, "grad_norm": 3.2887625836143406, "learning_rate": 2.2314039627322415e-06, "loss": 0.4432, "step": 9820 }, { "epoch": 0.6962779156327543, "grad_norm": 3.6860251212318165, "learning_rate": 2.230447959773438e-06, "loss": 0.3941, "step": 9821 }, { "epoch": 0.6963488124778447, "grad_norm": 3.4915870695445683, "learning_rate": 2.229492102855229e-06, "loss": 0.3799, "step": 9822 }, { "epoch": 0.6964197093229352, "grad_norm": 8.073467753844485, "learning_rate": 2.2285363920280167e-06, "loss": 0.4054, "step": 9823 }, { "epoch": 0.6964906061680255, "grad_norm": 2.2093266638801357, "learning_rate": 2.2275808273421938e-06, "loss": 0.3855, "step": 9824 }, { "epoch": 0.6965615030131159, "grad_norm": 2.0115517280682953, "learning_rate": 2.2266254088481504e-06, "loss": 0.3534, "step": 9825 }, { "epoch": 0.6966323998582064, "grad_norm": 2.4837722934950857, "learning_rate": 2.225670136596269e-06, "loss": 0.371, "step": 9826 }, { "epoch": 0.6967032967032967, "grad_norm": 2.7884487139893985, "learning_rate": 2.2247150106369155e-06, "loss": 0.3334, "step": 9827 }, { "epoch": 0.6967741935483871, "grad_norm": 8.6315582239361, "learning_rate": 2.2237600310204577e-06, "loss": 0.3754, "step": 9828 }, { "epoch": 0.6968450903934775, "grad_norm": 2.207472142334234, "learning_rate": 2.2228051977972543e-06, "loss": 0.4062, "step": 9829 }, { "epoch": 0.6969159872385678, "grad_norm": 2.541933183808696, "learning_rate": 2.2218505110176526e-06, "loss": 0.368, "step": 9830 }, { "epoch": 0.6969868840836583, "grad_norm": 3.336992188206696, "learning_rate": 2.220895970731992e-06, "loss": 0.3687, "step": 9831 }, { "epoch": 0.6970577809287487, "grad_norm": 2.8466565791006824, "learning_rate": 2.2199415769906103e-06, "loss": 0.3557, "step": 9832 }, { "epoch": 0.697128677773839, "grad_norm": 4.546734079287675, "learning_rate": 2.2189873298438286e-06, "loss": 0.394, "step": 9833 }, { "epoch": 0.6971995746189295, "grad_norm": 2.6365253938643143, "learning_rate": 2.21803322934197e-06, "loss": 0.4374, "step": 9834 }, { "epoch": 0.6972704714640199, "grad_norm": 3.4331987153221952, "learning_rate": 2.217079275535343e-06, "loss": 0.3813, "step": 9835 }, { "epoch": 0.6973413683091102, "grad_norm": 3.344073725683698, "learning_rate": 2.216125468474247e-06, "loss": 0.3539, "step": 9836 }, { "epoch": 0.6974122651542006, "grad_norm": 1.8487832625886997, "learning_rate": 2.2151718082089823e-06, "loss": 0.3926, "step": 9837 }, { "epoch": 0.6974831619992911, "grad_norm": 4.499785036831906, "learning_rate": 2.2142182947898315e-06, "loss": 0.4537, "step": 9838 }, { "epoch": 0.6975540588443814, "grad_norm": 4.403430875411328, "learning_rate": 2.2132649282670787e-06, "loss": 0.388, "step": 9839 }, { "epoch": 0.6976249556894718, "grad_norm": 2.1988755934169952, "learning_rate": 2.2123117086909914e-06, "loss": 0.3803, "step": 9840 }, { "epoch": 0.6976958525345622, "grad_norm": 23.855801592485818, "learning_rate": 2.2113586361118367e-06, "loss": 0.4195, "step": 9841 }, { "epoch": 0.6977667493796526, "grad_norm": 3.3162072296446032, "learning_rate": 2.2104057105798705e-06, "loss": 0.3459, "step": 9842 }, { "epoch": 0.697837646224743, "grad_norm": 4.054292285946148, "learning_rate": 2.209452932145338e-06, "loss": 0.3921, "step": 9843 }, { "epoch": 0.6979085430698334, "grad_norm": 2.956786634722759, "learning_rate": 2.2085003008584822e-06, "loss": 0.4017, "step": 9844 }, { "epoch": 0.6979794399149238, "grad_norm": 3.1837400568369087, "learning_rate": 2.20754781676954e-06, "loss": 0.3667, "step": 9845 }, { "epoch": 0.6980503367600142, "grad_norm": 4.619775621753586, "learning_rate": 2.206595479928729e-06, "loss": 0.443, "step": 9846 }, { "epoch": 0.6981212336051046, "grad_norm": 3.091945915307521, "learning_rate": 2.2056432903862705e-06, "loss": 0.4225, "step": 9847 }, { "epoch": 0.698192130450195, "grad_norm": 4.038595227635156, "learning_rate": 2.204691248192376e-06, "loss": 0.3857, "step": 9848 }, { "epoch": 0.6982630272952853, "grad_norm": 5.663864067796684, "learning_rate": 2.203739353397246e-06, "loss": 0.3401, "step": 9849 }, { "epoch": 0.6983339241403758, "grad_norm": 2.791054681114976, "learning_rate": 2.2027876060510715e-06, "loss": 0.364, "step": 9850 }, { "epoch": 0.6984048209854662, "grad_norm": 4.0320141502986155, "learning_rate": 2.2018360062040424e-06, "loss": 0.4013, "step": 9851 }, { "epoch": 0.6984757178305565, "grad_norm": 2.347924731917598, "learning_rate": 2.2008845539063373e-06, "loss": 0.4262, "step": 9852 }, { "epoch": 0.6985466146756469, "grad_norm": 4.8700498603291855, "learning_rate": 2.199933249208127e-06, "loss": 0.4031, "step": 9853 }, { "epoch": 0.6986175115207374, "grad_norm": 28.07771071011709, "learning_rate": 2.198982092159571e-06, "loss": 0.3993, "step": 9854 }, { "epoch": 0.6986884083658277, "grad_norm": 3.65103172346538, "learning_rate": 2.1980310828108298e-06, "loss": 0.3982, "step": 9855 }, { "epoch": 0.6987593052109181, "grad_norm": 4.732475487992845, "learning_rate": 2.1970802212120477e-06, "loss": 0.414, "step": 9856 }, { "epoch": 0.6988302020560085, "grad_norm": 2.574075769590678, "learning_rate": 2.1961295074133635e-06, "loss": 0.4129, "step": 9857 }, { "epoch": 0.6989010989010989, "grad_norm": 4.32430462465226, "learning_rate": 2.1951789414649126e-06, "loss": 0.3676, "step": 9858 }, { "epoch": 0.6989719957461893, "grad_norm": 4.359985161525858, "learning_rate": 2.1942285234168147e-06, "loss": 0.47, "step": 9859 }, { "epoch": 0.6990428925912797, "grad_norm": 2.3112830433977236, "learning_rate": 2.19327825331919e-06, "loss": 0.3992, "step": 9860 }, { "epoch": 0.69911378943637, "grad_norm": 2.2804952836905823, "learning_rate": 2.1923281312221457e-06, "loss": 0.3955, "step": 9861 }, { "epoch": 0.6991846862814605, "grad_norm": 4.203756910562116, "learning_rate": 2.19137815717578e-06, "loss": 0.3994, "step": 9862 }, { "epoch": 0.6992555831265509, "grad_norm": 2.6659991785246913, "learning_rate": 2.1904283312301873e-06, "loss": 0.3864, "step": 9863 }, { "epoch": 0.6993264799716412, "grad_norm": 3.0442365784965237, "learning_rate": 2.1894786534354574e-06, "loss": 0.4251, "step": 9864 }, { "epoch": 0.6993973768167316, "grad_norm": 2.3256537704243105, "learning_rate": 2.1885291238416595e-06, "loss": 0.4222, "step": 9865 }, { "epoch": 0.6994682736618221, "grad_norm": 3.106290677986557, "learning_rate": 2.1875797424988664e-06, "loss": 0.3734, "step": 9866 }, { "epoch": 0.6995391705069124, "grad_norm": 2.7975260253938665, "learning_rate": 2.186630509457142e-06, "loss": 0.4939, "step": 9867 }, { "epoch": 0.6996100673520028, "grad_norm": 5.622705229651305, "learning_rate": 2.1856814247665386e-06, "loss": 0.4155, "step": 9868 }, { "epoch": 0.6996809641970932, "grad_norm": 5.420208378600189, "learning_rate": 2.1847324884771e-06, "loss": 0.402, "step": 9869 }, { "epoch": 0.6997518610421837, "grad_norm": 3.383811228626702, "learning_rate": 2.1837837006388656e-06, "loss": 0.3829, "step": 9870 }, { "epoch": 0.699822757887274, "grad_norm": 6.47653501003328, "learning_rate": 2.18283506130187e-06, "loss": 0.3829, "step": 9871 }, { "epoch": 0.6998936547323644, "grad_norm": 6.341388443019632, "learning_rate": 2.1818865705161285e-06, "loss": 0.4166, "step": 9872 }, { "epoch": 0.6999645515774549, "grad_norm": 5.077723537868396, "learning_rate": 2.1809382283316586e-06, "loss": 0.3565, "step": 9873 }, { "epoch": 0.7000354484225452, "grad_norm": 5.408535912790977, "learning_rate": 2.17999003479847e-06, "loss": 0.3856, "step": 9874 }, { "epoch": 0.7001063452676356, "grad_norm": 5.849732318952005, "learning_rate": 2.179041989966559e-06, "loss": 0.3696, "step": 9875 }, { "epoch": 0.700177242112726, "grad_norm": 4.128896693740259, "learning_rate": 2.178094093885915e-06, "loss": 0.4064, "step": 9876 }, { "epoch": 0.7002481389578163, "grad_norm": 2.096152908488047, "learning_rate": 2.1771463466065256e-06, "loss": 0.3664, "step": 9877 }, { "epoch": 0.7003190358029068, "grad_norm": 3.0568066072101754, "learning_rate": 2.176198748178362e-06, "loss": 0.349, "step": 9878 }, { "epoch": 0.7003899326479972, "grad_norm": 6.215120270011294, "learning_rate": 2.1752512986513953e-06, "loss": 0.3744, "step": 9879 }, { "epoch": 0.7004608294930875, "grad_norm": 6.320030814795496, "learning_rate": 2.1743039980755837e-06, "loss": 0.3881, "step": 9880 }, { "epoch": 0.700531726338178, "grad_norm": 3.7931161782244933, "learning_rate": 2.173356846500877e-06, "loss": 0.3805, "step": 9881 }, { "epoch": 0.7006026231832684, "grad_norm": 0.9627414990081395, "learning_rate": 2.1724098439772236e-06, "loss": 0.606, "step": 9882 }, { "epoch": 0.7006735200283587, "grad_norm": 1.698521451095992, "learning_rate": 2.1714629905545547e-06, "loss": 0.3847, "step": 9883 }, { "epoch": 0.7007444168734491, "grad_norm": 3.7947856033278335, "learning_rate": 2.1705162862828038e-06, "loss": 0.4757, "step": 9884 }, { "epoch": 0.7008153137185396, "grad_norm": 2.085820438144672, "learning_rate": 2.169569731211887e-06, "loss": 0.323, "step": 9885 }, { "epoch": 0.7008862105636299, "grad_norm": 2.7397502482791287, "learning_rate": 2.16862332539172e-06, "loss": 0.3903, "step": 9886 }, { "epoch": 0.7009571074087203, "grad_norm": 4.6433158595846065, "learning_rate": 2.1676770688722066e-06, "loss": 0.4065, "step": 9887 }, { "epoch": 0.7010280042538107, "grad_norm": 3.8757525372346278, "learning_rate": 2.1667309617032405e-06, "loss": 0.4575, "step": 9888 }, { "epoch": 0.701098901098901, "grad_norm": 2.708692266102888, "learning_rate": 2.1657850039347144e-06, "loss": 0.3942, "step": 9889 }, { "epoch": 0.7011697979439915, "grad_norm": 3.438724161374003, "learning_rate": 2.164839195616511e-06, "loss": 0.4244, "step": 9890 }, { "epoch": 0.7012406947890819, "grad_norm": 2.697179225036324, "learning_rate": 2.163893536798498e-06, "loss": 0.3952, "step": 9891 }, { "epoch": 0.7013115916341722, "grad_norm": 2.9959500813694784, "learning_rate": 2.1629480275305432e-06, "loss": 0.3648, "step": 9892 }, { "epoch": 0.7013824884792627, "grad_norm": 5.142511063540977, "learning_rate": 2.1620026678625055e-06, "loss": 0.3994, "step": 9893 }, { "epoch": 0.7014533853243531, "grad_norm": 3.4791302970715883, "learning_rate": 2.161057457844234e-06, "loss": 0.4332, "step": 9894 }, { "epoch": 0.7015242821694435, "grad_norm": 2.591104822020831, "learning_rate": 2.1601123975255677e-06, "loss": 0.3978, "step": 9895 }, { "epoch": 0.7015951790145338, "grad_norm": 4.034528209480697, "learning_rate": 2.159167486956342e-06, "loss": 0.403, "step": 9896 }, { "epoch": 0.7016660758596243, "grad_norm": 2.596313503602076, "learning_rate": 2.1582227261863854e-06, "loss": 0.3278, "step": 9897 }, { "epoch": 0.7017369727047147, "grad_norm": 2.603730480811454, "learning_rate": 2.157278115265513e-06, "loss": 0.3115, "step": 9898 }, { "epoch": 0.701807869549805, "grad_norm": 1.9163089359317318, "learning_rate": 2.1563336542435337e-06, "loss": 0.362, "step": 9899 }, { "epoch": 0.7018787663948954, "grad_norm": 3.1410400015103983, "learning_rate": 2.155389343170253e-06, "loss": 0.3937, "step": 9900 }, { "epoch": 0.7019496632399859, "grad_norm": 3.4881162861468558, "learning_rate": 2.154445182095463e-06, "loss": 0.431, "step": 9901 }, { "epoch": 0.7020205600850762, "grad_norm": 4.90449261505384, "learning_rate": 2.1535011710689497e-06, "loss": 0.3833, "step": 9902 }, { "epoch": 0.7020914569301666, "grad_norm": 5.613374163853578, "learning_rate": 2.1525573101404934e-06, "loss": 0.357, "step": 9903 }, { "epoch": 0.702162353775257, "grad_norm": 2.744345159167678, "learning_rate": 2.151613599359862e-06, "loss": 0.3764, "step": 9904 }, { "epoch": 0.7022332506203474, "grad_norm": 3.052064033343795, "learning_rate": 2.150670038776821e-06, "loss": 0.4013, "step": 9905 }, { "epoch": 0.7023041474654378, "grad_norm": 4.9675213803049045, "learning_rate": 2.1497266284411237e-06, "loss": 0.4018, "step": 9906 }, { "epoch": 0.7023750443105282, "grad_norm": 2.5342305405297103, "learning_rate": 2.148783368402515e-06, "loss": 0.3856, "step": 9907 }, { "epoch": 0.7024459411556185, "grad_norm": 3.337045951961332, "learning_rate": 2.1478402587107357e-06, "loss": 0.415, "step": 9908 }, { "epoch": 0.702516838000709, "grad_norm": 3.6666707243934638, "learning_rate": 2.1468972994155203e-06, "loss": 0.4082, "step": 9909 }, { "epoch": 0.7025877348457994, "grad_norm": 6.500395035568646, "learning_rate": 2.1459544905665835e-06, "loss": 0.3754, "step": 9910 }, { "epoch": 0.7026586316908897, "grad_norm": 3.8493275532715217, "learning_rate": 2.1450118322136458e-06, "loss": 0.4089, "step": 9911 }, { "epoch": 0.7027295285359801, "grad_norm": 3.2005848350950625, "learning_rate": 2.144069324406415e-06, "loss": 0.403, "step": 9912 }, { "epoch": 0.7028004253810706, "grad_norm": 5.948838534572933, "learning_rate": 2.143126967194588e-06, "loss": 0.354, "step": 9913 }, { "epoch": 0.7028713222261609, "grad_norm": 4.66988150139051, "learning_rate": 2.142184760627855e-06, "loss": 0.4054, "step": 9914 }, { "epoch": 0.7029422190712513, "grad_norm": 3.013313827657884, "learning_rate": 2.141242704755901e-06, "loss": 0.4278, "step": 9915 }, { "epoch": 0.7030131159163417, "grad_norm": 3.953113470933818, "learning_rate": 2.1403007996284024e-06, "loss": 0.4048, "step": 9916 }, { "epoch": 0.7030840127614321, "grad_norm": 2.168018350657282, "learning_rate": 2.1393590452950264e-06, "loss": 0.3658, "step": 9917 }, { "epoch": 0.7031549096065225, "grad_norm": 3.3173088548401233, "learning_rate": 2.1384174418054286e-06, "loss": 0.3347, "step": 9918 }, { "epoch": 0.7032258064516129, "grad_norm": 3.537651814529127, "learning_rate": 2.1374759892092657e-06, "loss": 0.4078, "step": 9919 }, { "epoch": 0.7032967032967034, "grad_norm": 4.878667363276165, "learning_rate": 2.1365346875561783e-06, "loss": 0.3936, "step": 9920 }, { "epoch": 0.7033676001417937, "grad_norm": 5.512347071134759, "learning_rate": 2.1355935368958006e-06, "loss": 0.386, "step": 9921 }, { "epoch": 0.7034384969868841, "grad_norm": 2.6002555578757334, "learning_rate": 2.1346525372777643e-06, "loss": 0.3622, "step": 9922 }, { "epoch": 0.7035093938319745, "grad_norm": 4.521768308830408, "learning_rate": 2.1337116887516846e-06, "loss": 0.3936, "step": 9923 }, { "epoch": 0.7035802906770648, "grad_norm": 3.129213998541183, "learning_rate": 2.132770991367177e-06, "loss": 0.379, "step": 9924 }, { "epoch": 0.7036511875221553, "grad_norm": 7.1070596575603115, "learning_rate": 2.131830445173842e-06, "loss": 0.405, "step": 9925 }, { "epoch": 0.7037220843672457, "grad_norm": 3.531201063666007, "learning_rate": 2.130890050221278e-06, "loss": 0.3944, "step": 9926 }, { "epoch": 0.703792981212336, "grad_norm": 2.6912684325063467, "learning_rate": 2.1299498065590703e-06, "loss": 0.3966, "step": 9927 }, { "epoch": 0.7038638780574265, "grad_norm": 2.6452614303306676, "learning_rate": 2.1290097142368018e-06, "loss": 0.3586, "step": 9928 }, { "epoch": 0.7039347749025169, "grad_norm": 3.2168663702447122, "learning_rate": 2.1280697733040416e-06, "loss": 0.4555, "step": 9929 }, { "epoch": 0.7040056717476072, "grad_norm": 3.637514719951444, "learning_rate": 2.127129983810353e-06, "loss": 0.3747, "step": 9930 }, { "epoch": 0.7040765685926976, "grad_norm": 7.8776294867184475, "learning_rate": 2.126190345805295e-06, "loss": 0.4326, "step": 9931 }, { "epoch": 0.7041474654377881, "grad_norm": 2.558782538787852, "learning_rate": 2.1252508593384137e-06, "loss": 0.4089, "step": 9932 }, { "epoch": 0.7042183622828784, "grad_norm": 3.590908103117127, "learning_rate": 2.1243115244592465e-06, "loss": 0.3853, "step": 9933 }, { "epoch": 0.7042892591279688, "grad_norm": 2.9091163661094837, "learning_rate": 2.123372341217328e-06, "loss": 0.4197, "step": 9934 }, { "epoch": 0.7043601559730592, "grad_norm": 4.303552820873344, "learning_rate": 2.122433309662185e-06, "loss": 0.3481, "step": 9935 }, { "epoch": 0.7044310528181496, "grad_norm": 3.4576887244737433, "learning_rate": 2.1214944298433267e-06, "loss": 0.3611, "step": 9936 }, { "epoch": 0.70450194966324, "grad_norm": 2.196120843193509, "learning_rate": 2.1205557018102635e-06, "loss": 0.3845, "step": 9937 }, { "epoch": 0.7045728465083304, "grad_norm": 2.526118738639766, "learning_rate": 2.119617125612498e-06, "loss": 0.3909, "step": 9938 }, { "epoch": 0.7046437433534207, "grad_norm": 2.913267146508503, "learning_rate": 2.1186787012995204e-06, "loss": 0.4281, "step": 9939 }, { "epoch": 0.7047146401985112, "grad_norm": 6.042816794692635, "learning_rate": 2.1177404289208114e-06, "loss": 0.4115, "step": 9940 }, { "epoch": 0.7047855370436016, "grad_norm": 4.516579615031672, "learning_rate": 2.11680230852585e-06, "loss": 0.4067, "step": 9941 }, { "epoch": 0.7048564338886919, "grad_norm": 3.2129208545739663, "learning_rate": 2.1158643401641056e-06, "loss": 0.4118, "step": 9942 }, { "epoch": 0.7049273307337823, "grad_norm": 2.4566920554669833, "learning_rate": 2.114926523885035e-06, "loss": 0.3732, "step": 9943 }, { "epoch": 0.7049982275788728, "grad_norm": 5.562439512960948, "learning_rate": 2.1139888597380905e-06, "loss": 0.3747, "step": 9944 }, { "epoch": 0.7050691244239631, "grad_norm": 16.609247895283538, "learning_rate": 2.1130513477727168e-06, "loss": 0.4025, "step": 9945 }, { "epoch": 0.7051400212690535, "grad_norm": 4.284955684814792, "learning_rate": 2.1121139880383494e-06, "loss": 0.3927, "step": 9946 }, { "epoch": 0.7052109181141439, "grad_norm": 1.7390766747812418, "learning_rate": 2.111176780584414e-06, "loss": 0.3762, "step": 9947 }, { "epoch": 0.7052818149592344, "grad_norm": 3.244039103743792, "learning_rate": 2.1102397254603333e-06, "loss": 0.4561, "step": 9948 }, { "epoch": 0.7053527118043247, "grad_norm": 2.53985551470046, "learning_rate": 2.109302822715516e-06, "loss": 0.3792, "step": 9949 }, { "epoch": 0.7054236086494151, "grad_norm": 9.822977245541251, "learning_rate": 2.1083660723993694e-06, "loss": 0.434, "step": 9950 }, { "epoch": 0.7054945054945055, "grad_norm": 3.686882648387879, "learning_rate": 2.1074294745612863e-06, "loss": 0.4308, "step": 9951 }, { "epoch": 0.7055654023395959, "grad_norm": 2.5293852729190176, "learning_rate": 2.106493029250653e-06, "loss": 0.4004, "step": 9952 }, { "epoch": 0.7056362991846863, "grad_norm": 6.593468569944344, "learning_rate": 2.105556736516851e-06, "loss": 0.3783, "step": 9953 }, { "epoch": 0.7057071960297767, "grad_norm": 3.5583975558578635, "learning_rate": 2.1046205964092536e-06, "loss": 0.3537, "step": 9954 }, { "epoch": 0.705778092874867, "grad_norm": 2.5782997157653607, "learning_rate": 2.103684608977222e-06, "loss": 0.3857, "step": 9955 }, { "epoch": 0.7058489897199575, "grad_norm": 2.9799552431740084, "learning_rate": 2.102748774270109e-06, "loss": 0.4206, "step": 9956 }, { "epoch": 0.7059198865650479, "grad_norm": 2.25543530979004, "learning_rate": 2.101813092337267e-06, "loss": 0.3979, "step": 9957 }, { "epoch": 0.7059907834101382, "grad_norm": 2.268658702670884, "learning_rate": 2.1008775632280332e-06, "loss": 0.4189, "step": 9958 }, { "epoch": 0.7060616802552286, "grad_norm": 3.3489470770225545, "learning_rate": 2.0999421869917358e-06, "loss": 0.3784, "step": 9959 }, { "epoch": 0.7061325771003191, "grad_norm": 4.207193447193068, "learning_rate": 2.0990069636777006e-06, "loss": 0.3617, "step": 9960 }, { "epoch": 0.7062034739454094, "grad_norm": 3.239823119485556, "learning_rate": 2.098071893335244e-06, "loss": 0.4243, "step": 9961 }, { "epoch": 0.7062743707904998, "grad_norm": 2.508303022859958, "learning_rate": 2.0971369760136716e-06, "loss": 0.396, "step": 9962 }, { "epoch": 0.7063452676355902, "grad_norm": 4.230941001976125, "learning_rate": 2.0962022117622805e-06, "loss": 0.4031, "step": 9963 }, { "epoch": 0.7064161644806806, "grad_norm": 6.950235786524293, "learning_rate": 2.0952676006303647e-06, "loss": 0.3958, "step": 9964 }, { "epoch": 0.706487061325771, "grad_norm": 16.923830044655393, "learning_rate": 2.0943331426672055e-06, "loss": 0.4419, "step": 9965 }, { "epoch": 0.7065579581708614, "grad_norm": 1.792277062783636, "learning_rate": 2.0933988379220755e-06, "loss": 0.3845, "step": 9966 }, { "epoch": 0.7066288550159517, "grad_norm": 1.473760847693492, "learning_rate": 2.092464686444245e-06, "loss": 0.367, "step": 9967 }, { "epoch": 0.7066997518610422, "grad_norm": 2.9686020844275416, "learning_rate": 2.0915306882829693e-06, "loss": 0.3862, "step": 9968 }, { "epoch": 0.7067706487061326, "grad_norm": 10.349146392832893, "learning_rate": 2.090596843487502e-06, "loss": 0.4011, "step": 9969 }, { "epoch": 0.7068415455512229, "grad_norm": 4.246439998432063, "learning_rate": 2.089663152107082e-06, "loss": 0.3729, "step": 9970 }, { "epoch": 0.7069124423963133, "grad_norm": 2.9101386674951546, "learning_rate": 2.0887296141909476e-06, "loss": 0.3857, "step": 9971 }, { "epoch": 0.7069833392414038, "grad_norm": 3.294187709032082, "learning_rate": 2.0877962297883204e-06, "loss": 0.3808, "step": 9972 }, { "epoch": 0.7070542360864942, "grad_norm": 4.98066860859318, "learning_rate": 2.0868629989484225e-06, "loss": 0.4378, "step": 9973 }, { "epoch": 0.7071251329315845, "grad_norm": 9.670063958881418, "learning_rate": 2.085929921720462e-06, "loss": 0.4844, "step": 9974 }, { "epoch": 0.707196029776675, "grad_norm": 20.464881437568483, "learning_rate": 2.0849969981536393e-06, "loss": 0.417, "step": 9975 }, { "epoch": 0.7072669266217654, "grad_norm": 3.902571983800735, "learning_rate": 2.084064228297152e-06, "loss": 0.3812, "step": 9976 }, { "epoch": 0.7073378234668557, "grad_norm": 4.378313159075003, "learning_rate": 2.083131612200184e-06, "loss": 0.3821, "step": 9977 }, { "epoch": 0.7074087203119461, "grad_norm": 2.3554655523234787, "learning_rate": 2.0821991499119092e-06, "loss": 0.3684, "step": 9978 }, { "epoch": 0.7074796171570366, "grad_norm": 3.9484753012430986, "learning_rate": 2.081266841481501e-06, "loss": 0.3827, "step": 9979 }, { "epoch": 0.7075505140021269, "grad_norm": 46.2420329931822, "learning_rate": 2.0803346869581226e-06, "loss": 0.3968, "step": 9980 }, { "epoch": 0.7076214108472173, "grad_norm": 2.7879907516849216, "learning_rate": 2.079402686390924e-06, "loss": 0.4245, "step": 9981 }, { "epoch": 0.7076923076923077, "grad_norm": 2.4295048703889166, "learning_rate": 2.0784708398290486e-06, "loss": 0.3491, "step": 9982 }, { "epoch": 0.707763204537398, "grad_norm": 2.470900829608823, "learning_rate": 2.077539147321638e-06, "loss": 0.4151, "step": 9983 }, { "epoch": 0.7078341013824885, "grad_norm": 2.0735236201367284, "learning_rate": 2.076607608917819e-06, "loss": 0.4181, "step": 9984 }, { "epoch": 0.7079049982275789, "grad_norm": 2.584339961726559, "learning_rate": 2.07567622466671e-06, "loss": 0.3813, "step": 9985 }, { "epoch": 0.7079758950726692, "grad_norm": 2.2557927666847846, "learning_rate": 2.0747449946174258e-06, "loss": 0.3909, "step": 9986 }, { "epoch": 0.7080467919177597, "grad_norm": 2.058089504674244, "learning_rate": 2.0738139188190726e-06, "loss": 0.3855, "step": 9987 }, { "epoch": 0.7081176887628501, "grad_norm": 2.8484067685039602, "learning_rate": 2.072882997320745e-06, "loss": 0.4017, "step": 9988 }, { "epoch": 0.7081885856079404, "grad_norm": 2.2321419161853937, "learning_rate": 2.0719522301715288e-06, "loss": 0.3348, "step": 9989 }, { "epoch": 0.7082594824530308, "grad_norm": 2.7570219803534335, "learning_rate": 2.0710216174205088e-06, "loss": 0.3783, "step": 9990 }, { "epoch": 0.7083303792981213, "grad_norm": 2.9920298931400398, "learning_rate": 2.0700911591167543e-06, "loss": 0.3793, "step": 9991 }, { "epoch": 0.7084012761432116, "grad_norm": 4.945441837393666, "learning_rate": 2.069160855309327e-06, "loss": 0.3853, "step": 9992 }, { "epoch": 0.708472172988302, "grad_norm": 4.366317258627529, "learning_rate": 2.0682307060472875e-06, "loss": 0.4216, "step": 9993 }, { "epoch": 0.7085430698333924, "grad_norm": 3.3347875286369577, "learning_rate": 2.0673007113796785e-06, "loss": 0.4166, "step": 9994 }, { "epoch": 0.7086139666784828, "grad_norm": 2.6560181177662288, "learning_rate": 2.066370871355543e-06, "loss": 0.3907, "step": 9995 }, { "epoch": 0.7086848635235732, "grad_norm": 3.8236702807761205, "learning_rate": 2.0654411860239103e-06, "loss": 0.4011, "step": 9996 }, { "epoch": 0.7087557603686636, "grad_norm": 2.2225330536970147, "learning_rate": 2.0645116554338023e-06, "loss": 0.3697, "step": 9997 }, { "epoch": 0.708826657213754, "grad_norm": 2.73173413181438, "learning_rate": 2.063582279634235e-06, "loss": 0.3895, "step": 9998 }, { "epoch": 0.7088975540588444, "grad_norm": 1.9551921858103813, "learning_rate": 2.062653058674218e-06, "loss": 0.3746, "step": 9999 }, { "epoch": 0.7089684509039348, "grad_norm": 2.666792985513505, "learning_rate": 2.061723992602746e-06, "loss": 0.3784, "step": 10000 }, { "epoch": 0.7090393477490252, "grad_norm": 2.913887845747468, "learning_rate": 2.0607950814688095e-06, "loss": 0.3826, "step": 10001 }, { "epoch": 0.7091102445941155, "grad_norm": 1.9537069689417996, "learning_rate": 2.0598663253213936e-06, "loss": 0.3466, "step": 10002 }, { "epoch": 0.709181141439206, "grad_norm": 5.194344666806766, "learning_rate": 2.05893772420947e-06, "loss": 0.3779, "step": 10003 }, { "epoch": 0.7092520382842964, "grad_norm": 2.9009840293186793, "learning_rate": 2.0580092781820037e-06, "loss": 0.3416, "step": 10004 }, { "epoch": 0.7093229351293867, "grad_norm": 3.5349042777351856, "learning_rate": 2.0570809872879534e-06, "loss": 0.4052, "step": 10005 }, { "epoch": 0.7093938319744771, "grad_norm": 3.016437381106504, "learning_rate": 2.056152851576271e-06, "loss": 0.4299, "step": 10006 }, { "epoch": 0.7094647288195676, "grad_norm": 3.121108322196524, "learning_rate": 2.055224871095895e-06, "loss": 0.4018, "step": 10007 }, { "epoch": 0.7095356256646579, "grad_norm": 6.296566971054356, "learning_rate": 2.054297045895758e-06, "loss": 0.3834, "step": 10008 }, { "epoch": 0.7096065225097483, "grad_norm": 8.834450455779182, "learning_rate": 2.0533693760247876e-06, "loss": 0.3733, "step": 10009 }, { "epoch": 0.7096774193548387, "grad_norm": 3.462424763585618, "learning_rate": 2.052441861531899e-06, "loss": 0.3815, "step": 10010 }, { "epoch": 0.7097483161999291, "grad_norm": 2.1418797059885235, "learning_rate": 2.0515145024659995e-06, "loss": 0.3836, "step": 10011 }, { "epoch": 0.7098192130450195, "grad_norm": 5.226245074734119, "learning_rate": 2.050587298875992e-06, "loss": 0.361, "step": 10012 }, { "epoch": 0.7098901098901099, "grad_norm": 11.063866269933726, "learning_rate": 2.049660250810766e-06, "loss": 0.3989, "step": 10013 }, { "epoch": 0.7099610067352002, "grad_norm": 2.7472521144896485, "learning_rate": 2.0487333583192082e-06, "loss": 0.4047, "step": 10014 }, { "epoch": 0.7100319035802907, "grad_norm": 5.699299129706608, "learning_rate": 2.047806621450192e-06, "loss": 0.3527, "step": 10015 }, { "epoch": 0.7101028004253811, "grad_norm": 3.4003647301828663, "learning_rate": 2.0468800402525868e-06, "loss": 0.3864, "step": 10016 }, { "epoch": 0.7101736972704714, "grad_norm": 2.6836428042164773, "learning_rate": 2.04595361477525e-06, "loss": 0.3309, "step": 10017 }, { "epoch": 0.7102445941155618, "grad_norm": 3.4681648862944483, "learning_rate": 2.0450273450670353e-06, "loss": 0.4076, "step": 10018 }, { "epoch": 0.7103154909606523, "grad_norm": 4.97595222938758, "learning_rate": 2.044101231176784e-06, "loss": 0.41, "step": 10019 }, { "epoch": 0.7103863878057426, "grad_norm": 2.472368573339734, "learning_rate": 2.0431752731533292e-06, "loss": 0.4049, "step": 10020 }, { "epoch": 0.710457284650833, "grad_norm": 2.342422889178057, "learning_rate": 2.0422494710455014e-06, "loss": 0.3577, "step": 10021 }, { "epoch": 0.7105281814959234, "grad_norm": 3.729356872352532, "learning_rate": 2.041323824902116e-06, "loss": 0.4065, "step": 10022 }, { "epoch": 0.7105990783410139, "grad_norm": 5.225820891825142, "learning_rate": 2.040398334771982e-06, "loss": 0.3847, "step": 10023 }, { "epoch": 0.7106699751861042, "grad_norm": 2.2728178637754946, "learning_rate": 2.0394730007039026e-06, "loss": 0.381, "step": 10024 }, { "epoch": 0.7107408720311946, "grad_norm": 10.71246080266473, "learning_rate": 2.0385478227466736e-06, "loss": 0.3968, "step": 10025 }, { "epoch": 0.710811768876285, "grad_norm": 3.7228525672685873, "learning_rate": 2.0376228009490783e-06, "loss": 0.4096, "step": 10026 }, { "epoch": 0.7108826657213754, "grad_norm": 2.5421871313544644, "learning_rate": 2.0366979353598913e-06, "loss": 0.3947, "step": 10027 }, { "epoch": 0.7109535625664658, "grad_norm": 3.047919549781417, "learning_rate": 2.035773226027885e-06, "loss": 0.3741, "step": 10028 }, { "epoch": 0.7110244594115562, "grad_norm": 2.7460968709025817, "learning_rate": 2.034848673001822e-06, "loss": 0.4365, "step": 10029 }, { "epoch": 0.7110953562566465, "grad_norm": 3.3286332624381507, "learning_rate": 2.033924276330448e-06, "loss": 0.3911, "step": 10030 }, { "epoch": 0.711166253101737, "grad_norm": 2.3185628663505806, "learning_rate": 2.0330000360625106e-06, "loss": 0.3807, "step": 10031 }, { "epoch": 0.7112371499468274, "grad_norm": 9.774420418793675, "learning_rate": 2.0320759522467483e-06, "loss": 0.4008, "step": 10032 }, { "epoch": 0.7113080467919177, "grad_norm": 2.6287357606800485, "learning_rate": 2.0311520249318863e-06, "loss": 0.4009, "step": 10033 }, { "epoch": 0.7113789436370082, "grad_norm": 6.476030048892034, "learning_rate": 2.0302282541666417e-06, "loss": 0.3889, "step": 10034 }, { "epoch": 0.7114498404820986, "grad_norm": 3.818457815634464, "learning_rate": 2.0293046399997303e-06, "loss": 0.4175, "step": 10035 }, { "epoch": 0.7115207373271889, "grad_norm": 2.4428346588486307, "learning_rate": 2.0283811824798506e-06, "loss": 0.3835, "step": 10036 }, { "epoch": 0.7115916341722793, "grad_norm": 3.353955439890619, "learning_rate": 2.027457881655701e-06, "loss": 0.3459, "step": 10037 }, { "epoch": 0.7116625310173698, "grad_norm": 1.1735853899184745, "learning_rate": 2.0265347375759664e-06, "loss": 0.6169, "step": 10038 }, { "epoch": 0.7117334278624601, "grad_norm": 16.96239709119752, "learning_rate": 2.025611750289323e-06, "loss": 0.3922, "step": 10039 }, { "epoch": 0.7118043247075505, "grad_norm": 3.113555768997182, "learning_rate": 2.0246889198444436e-06, "loss": 0.3818, "step": 10040 }, { "epoch": 0.7118752215526409, "grad_norm": 6.51615423165202, "learning_rate": 2.0237662462899887e-06, "loss": 0.4001, "step": 10041 }, { "epoch": 0.7119461183977313, "grad_norm": 2.3071209203543876, "learning_rate": 2.022843729674609e-06, "loss": 0.3992, "step": 10042 }, { "epoch": 0.7120170152428217, "grad_norm": 1.9363732362187165, "learning_rate": 2.0219213700469526e-06, "loss": 0.3405, "step": 10043 }, { "epoch": 0.7120879120879121, "grad_norm": 2.6316027739776136, "learning_rate": 2.0209991674556566e-06, "loss": 0.4078, "step": 10044 }, { "epoch": 0.7121588089330024, "grad_norm": 2.4165344774613797, "learning_rate": 2.020077121949348e-06, "loss": 0.4474, "step": 10045 }, { "epoch": 0.7122297057780929, "grad_norm": 2.6412154332713116, "learning_rate": 2.0191552335766463e-06, "loss": 0.3502, "step": 10046 }, { "epoch": 0.7123006026231833, "grad_norm": 12.385776115117531, "learning_rate": 2.0182335023861655e-06, "loss": 0.4361, "step": 10047 }, { "epoch": 0.7123714994682737, "grad_norm": 2.4822017875486293, "learning_rate": 2.0173119284265084e-06, "loss": 0.386, "step": 10048 }, { "epoch": 0.712442396313364, "grad_norm": 4.105267752383594, "learning_rate": 2.016390511746268e-06, "loss": 0.3918, "step": 10049 }, { "epoch": 0.7125132931584545, "grad_norm": 5.0082317831790055, "learning_rate": 2.015469252394033e-06, "loss": 0.41, "step": 10050 }, { "epoch": 0.7125841900035449, "grad_norm": 1.4867920463747217, "learning_rate": 2.0145481504183838e-06, "loss": 0.6463, "step": 10051 }, { "epoch": 0.7126550868486352, "grad_norm": 2.46612539447885, "learning_rate": 2.0136272058678903e-06, "loss": 0.4001, "step": 10052 }, { "epoch": 0.7127259836937256, "grad_norm": 11.115726486513227, "learning_rate": 2.012706418791111e-06, "loss": 0.4065, "step": 10053 }, { "epoch": 0.7127968805388161, "grad_norm": 3.4254752756035196, "learning_rate": 2.0117857892366042e-06, "loss": 0.4001, "step": 10054 }, { "epoch": 0.7128677773839064, "grad_norm": 2.515923746860648, "learning_rate": 2.0108653172529143e-06, "loss": 0.4162, "step": 10055 }, { "epoch": 0.7129386742289968, "grad_norm": 3.1321465247952993, "learning_rate": 2.009945002888575e-06, "loss": 0.4357, "step": 10056 }, { "epoch": 0.7130095710740872, "grad_norm": 2.2616994469975054, "learning_rate": 2.009024846192118e-06, "loss": 0.3746, "step": 10057 }, { "epoch": 0.7130804679191776, "grad_norm": 1.9124059422557738, "learning_rate": 2.0081048472120663e-06, "loss": 0.4291, "step": 10058 }, { "epoch": 0.713151364764268, "grad_norm": 3.410466283244755, "learning_rate": 2.0071850059969287e-06, "loss": 0.3745, "step": 10059 }, { "epoch": 0.7132222616093584, "grad_norm": 4.074909718024443, "learning_rate": 2.0062653225952087e-06, "loss": 0.4162, "step": 10060 }, { "epoch": 0.7132931584544487, "grad_norm": 3.3139109105002413, "learning_rate": 2.0053457970554054e-06, "loss": 0.4217, "step": 10061 }, { "epoch": 0.7133640552995392, "grad_norm": 1.0499166867322238, "learning_rate": 2.004426429426002e-06, "loss": 0.6753, "step": 10062 }, { "epoch": 0.7134349521446296, "grad_norm": 3.4906335069578107, "learning_rate": 2.0035072197554805e-06, "loss": 0.3602, "step": 10063 }, { "epoch": 0.7135058489897199, "grad_norm": 4.4272172087373445, "learning_rate": 2.002588168092311e-06, "loss": 0.337, "step": 10064 }, { "epoch": 0.7135767458348103, "grad_norm": 5.217956635731819, "learning_rate": 2.001669274484953e-06, "loss": 0.4017, "step": 10065 }, { "epoch": 0.7136476426799008, "grad_norm": 2.068690894127035, "learning_rate": 2.000750538981865e-06, "loss": 0.3382, "step": 10066 }, { "epoch": 0.7137185395249911, "grad_norm": 2.8865856184764365, "learning_rate": 1.99983196163149e-06, "loss": 0.3765, "step": 10067 }, { "epoch": 0.7137894363700815, "grad_norm": 4.144598641020667, "learning_rate": 1.9989135424822644e-06, "loss": 0.3985, "step": 10068 }, { "epoch": 0.713860333215172, "grad_norm": 4.368787321939017, "learning_rate": 1.9979952815826175e-06, "loss": 0.4, "step": 10069 }, { "epoch": 0.7139312300602623, "grad_norm": 4.5937935914808055, "learning_rate": 1.997077178980973e-06, "loss": 0.3902, "step": 10070 }, { "epoch": 0.7140021269053527, "grad_norm": 2.310342771935261, "learning_rate": 1.99615923472574e-06, "loss": 0.4283, "step": 10071 }, { "epoch": 0.7140730237504431, "grad_norm": 2.660558876455043, "learning_rate": 1.9952414488653225e-06, "loss": 0.3577, "step": 10072 }, { "epoch": 0.7141439205955334, "grad_norm": 3.455515965713673, "learning_rate": 1.9943238214481165e-06, "loss": 0.4316, "step": 10073 }, { "epoch": 0.7142148174406239, "grad_norm": 7.807955208949989, "learning_rate": 1.9934063525225135e-06, "loss": 0.3875, "step": 10074 }, { "epoch": 0.7142857142857143, "grad_norm": 2.6141423910414336, "learning_rate": 1.992489042136884e-06, "loss": 0.4014, "step": 10075 }, { "epoch": 0.7143566111308047, "grad_norm": 2.565192329856273, "learning_rate": 1.9915718903396035e-06, "loss": 0.3993, "step": 10076 }, { "epoch": 0.714427507975895, "grad_norm": 1.0017400525680646, "learning_rate": 1.9906548971790344e-06, "loss": 0.6175, "step": 10077 }, { "epoch": 0.7144984048209855, "grad_norm": 2.9345885362514115, "learning_rate": 1.9897380627035307e-06, "loss": 0.3886, "step": 10078 }, { "epoch": 0.7145693016660759, "grad_norm": 3.3469818604852097, "learning_rate": 1.9888213869614343e-06, "loss": 0.3969, "step": 10079 }, { "epoch": 0.7146401985111662, "grad_norm": 2.644920238635374, "learning_rate": 1.987904870001086e-06, "loss": 0.3935, "step": 10080 }, { "epoch": 0.7147110953562567, "grad_norm": 5.161100405785014, "learning_rate": 1.986988511870811e-06, "loss": 0.3935, "step": 10081 }, { "epoch": 0.7147819922013471, "grad_norm": 2.267229979790739, "learning_rate": 1.9860723126189336e-06, "loss": 0.4432, "step": 10082 }, { "epoch": 0.7148528890464374, "grad_norm": 2.730338578530427, "learning_rate": 1.985156272293764e-06, "loss": 0.4019, "step": 10083 }, { "epoch": 0.7149237858915278, "grad_norm": 2.8156471046256644, "learning_rate": 1.9842403909436025e-06, "loss": 0.4434, "step": 10084 }, { "epoch": 0.7149946827366183, "grad_norm": 4.807056497311379, "learning_rate": 1.9833246686167495e-06, "loss": 0.3403, "step": 10085 }, { "epoch": 0.7150655795817086, "grad_norm": 6.1820979774883815, "learning_rate": 1.9824091053614876e-06, "loss": 0.4002, "step": 10086 }, { "epoch": 0.715136476426799, "grad_norm": 3.2516362237927026, "learning_rate": 1.981493701226098e-06, "loss": 0.3697, "step": 10087 }, { "epoch": 0.7152073732718894, "grad_norm": 3.170407092879066, "learning_rate": 1.980578456258848e-06, "loss": 0.3858, "step": 10088 }, { "epoch": 0.7152782701169798, "grad_norm": 3.499713852286878, "learning_rate": 1.9796633705080027e-06, "loss": 0.4293, "step": 10089 }, { "epoch": 0.7153491669620702, "grad_norm": 5.016274144100739, "learning_rate": 1.978748444021813e-06, "loss": 0.4006, "step": 10090 }, { "epoch": 0.7154200638071606, "grad_norm": 3.016107866483143, "learning_rate": 1.9778336768485217e-06, "loss": 0.3763, "step": 10091 }, { "epoch": 0.7154909606522509, "grad_norm": 2.647960852487525, "learning_rate": 1.9769190690363675e-06, "loss": 0.3384, "step": 10092 }, { "epoch": 0.7155618574973414, "grad_norm": 4.893387906832971, "learning_rate": 1.976004620633582e-06, "loss": 0.4105, "step": 10093 }, { "epoch": 0.7156327543424318, "grad_norm": 2.227286509843461, "learning_rate": 1.9750903316883775e-06, "loss": 0.3843, "step": 10094 }, { "epoch": 0.7157036511875221, "grad_norm": 7.781577324522566, "learning_rate": 1.974176202248968e-06, "loss": 0.4045, "step": 10095 }, { "epoch": 0.7157745480326125, "grad_norm": 1.627793604672912, "learning_rate": 1.9732622323635593e-06, "loss": 0.3682, "step": 10096 }, { "epoch": 0.715845444877703, "grad_norm": 1.9589956970971867, "learning_rate": 1.972348422080343e-06, "loss": 0.3634, "step": 10097 }, { "epoch": 0.7159163417227933, "grad_norm": 2.578961725889032, "learning_rate": 1.9714347714475035e-06, "loss": 0.3269, "step": 10098 }, { "epoch": 0.7159872385678837, "grad_norm": 4.774029734711214, "learning_rate": 1.9705212805132216e-06, "loss": 0.3833, "step": 10099 }, { "epoch": 0.7160581354129741, "grad_norm": 2.8936511779084335, "learning_rate": 1.969607949325665e-06, "loss": 0.3943, "step": 10100 }, { "epoch": 0.7161290322580646, "grad_norm": 2.9351701329563906, "learning_rate": 1.9686947779329927e-06, "loss": 0.3849, "step": 10101 }, { "epoch": 0.7161999291031549, "grad_norm": 4.754539983604805, "learning_rate": 1.967781766383359e-06, "loss": 0.452, "step": 10102 }, { "epoch": 0.7162708259482453, "grad_norm": 3.311889769827193, "learning_rate": 1.966868914724908e-06, "loss": 0.3966, "step": 10103 }, { "epoch": 0.7163417227933357, "grad_norm": 2.7291794734200567, "learning_rate": 1.9659562230057742e-06, "loss": 0.3982, "step": 10104 }, { "epoch": 0.7164126196384261, "grad_norm": 1.816779303118432, "learning_rate": 1.9650436912740834e-06, "loss": 0.3774, "step": 10105 }, { "epoch": 0.7164835164835165, "grad_norm": 4.387705771934962, "learning_rate": 1.9641313195779566e-06, "loss": 0.3864, "step": 10106 }, { "epoch": 0.7165544133286069, "grad_norm": 3.385513896783345, "learning_rate": 1.963219107965501e-06, "loss": 0.3633, "step": 10107 }, { "epoch": 0.7166253101736972, "grad_norm": 2.023913366722203, "learning_rate": 1.9623070564848213e-06, "loss": 0.4077, "step": 10108 }, { "epoch": 0.7166962070187877, "grad_norm": 4.252616215130473, "learning_rate": 1.961395165184009e-06, "loss": 0.3425, "step": 10109 }, { "epoch": 0.7167671038638781, "grad_norm": 2.6610013543933584, "learning_rate": 1.960483434111147e-06, "loss": 0.3881, "step": 10110 }, { "epoch": 0.7168380007089684, "grad_norm": 2.899402266108133, "learning_rate": 1.9595718633143157e-06, "loss": 0.4049, "step": 10111 }, { "epoch": 0.7169088975540588, "grad_norm": 2.762443245646647, "learning_rate": 1.9586604528415803e-06, "loss": 0.3543, "step": 10112 }, { "epoch": 0.7169797943991493, "grad_norm": 5.637877114672377, "learning_rate": 1.9577492027409987e-06, "loss": 0.4103, "step": 10113 }, { "epoch": 0.7170506912442396, "grad_norm": 5.068241579008313, "learning_rate": 1.9568381130606244e-06, "loss": 0.3822, "step": 10114 }, { "epoch": 0.71712158808933, "grad_norm": 3.7882440559424677, "learning_rate": 1.9559271838485007e-06, "loss": 0.3948, "step": 10115 }, { "epoch": 0.7171924849344204, "grad_norm": 1.9667742141808995, "learning_rate": 1.9550164151526598e-06, "loss": 0.4036, "step": 10116 }, { "epoch": 0.7172633817795108, "grad_norm": 2.4738629519634636, "learning_rate": 1.9541058070211255e-06, "loss": 0.4027, "step": 10117 }, { "epoch": 0.7173342786246012, "grad_norm": 4.668667505477127, "learning_rate": 1.953195359501917e-06, "loss": 0.3707, "step": 10118 }, { "epoch": 0.7174051754696916, "grad_norm": 4.507829217413173, "learning_rate": 1.952285072643046e-06, "loss": 0.4151, "step": 10119 }, { "epoch": 0.7174760723147819, "grad_norm": 4.471637228798666, "learning_rate": 1.9513749464925065e-06, "loss": 0.4245, "step": 10120 }, { "epoch": 0.7175469691598724, "grad_norm": 2.137871488418786, "learning_rate": 1.9504649810982924e-06, "loss": 0.3677, "step": 10121 }, { "epoch": 0.7176178660049628, "grad_norm": 3.028977883954806, "learning_rate": 1.949555176508389e-06, "loss": 0.412, "step": 10122 }, { "epoch": 0.7176887628500531, "grad_norm": 3.521476719102583, "learning_rate": 1.9486455327707698e-06, "loss": 0.3838, "step": 10123 }, { "epoch": 0.7177596596951435, "grad_norm": 1.9531900371791413, "learning_rate": 1.947736049933399e-06, "loss": 0.3681, "step": 10124 }, { "epoch": 0.717830556540234, "grad_norm": 4.89909927724089, "learning_rate": 1.9468267280442377e-06, "loss": 0.3973, "step": 10125 }, { "epoch": 0.7179014533853244, "grad_norm": 10.52614008430377, "learning_rate": 1.9459175671512316e-06, "loss": 0.4241, "step": 10126 }, { "epoch": 0.7179723502304147, "grad_norm": 2.9930546488320204, "learning_rate": 1.945008567302325e-06, "loss": 0.3484, "step": 10127 }, { "epoch": 0.7180432470755052, "grad_norm": 7.968231112814937, "learning_rate": 1.944099728545448e-06, "loss": 0.3342, "step": 10128 }, { "epoch": 0.7181141439205956, "grad_norm": 8.649685546460255, "learning_rate": 1.9431910509285233e-06, "loss": 0.374, "step": 10129 }, { "epoch": 0.7181850407656859, "grad_norm": 5.125433581547316, "learning_rate": 1.94228253449947e-06, "loss": 0.4612, "step": 10130 }, { "epoch": 0.7182559376107763, "grad_norm": 0.9830940711846066, "learning_rate": 1.9413741793061903e-06, "loss": 0.6721, "step": 10131 }, { "epoch": 0.7183268344558668, "grad_norm": 3.121779338113928, "learning_rate": 1.9404659853965864e-06, "loss": 0.3808, "step": 10132 }, { "epoch": 0.7183977313009571, "grad_norm": 2.0224780659676362, "learning_rate": 1.9395579528185442e-06, "loss": 0.3411, "step": 10133 }, { "epoch": 0.7184686281460475, "grad_norm": 3.2481270955474906, "learning_rate": 1.9386500816199496e-06, "loss": 0.3862, "step": 10134 }, { "epoch": 0.7185395249911379, "grad_norm": 3.958531654203962, "learning_rate": 1.937742371848672e-06, "loss": 0.4206, "step": 10135 }, { "epoch": 0.7186104218362283, "grad_norm": 2.3019159032942405, "learning_rate": 1.936834823552575e-06, "loss": 0.401, "step": 10136 }, { "epoch": 0.7186813186813187, "grad_norm": 2.2571826170192395, "learning_rate": 1.935927436779516e-06, "loss": 0.4188, "step": 10137 }, { "epoch": 0.7187522155264091, "grad_norm": 2.8343995497157333, "learning_rate": 1.935020211577345e-06, "loss": 0.4395, "step": 10138 }, { "epoch": 0.7188231123714994, "grad_norm": 2.9213745040140395, "learning_rate": 1.9341131479938947e-06, "loss": 0.3723, "step": 10139 }, { "epoch": 0.7188940092165899, "grad_norm": 4.184888538466917, "learning_rate": 1.933206246076998e-06, "loss": 0.4426, "step": 10140 }, { "epoch": 0.7189649060616803, "grad_norm": 13.51967340613285, "learning_rate": 1.932299505874479e-06, "loss": 0.3659, "step": 10141 }, { "epoch": 0.7190358029067706, "grad_norm": 2.9451218136146227, "learning_rate": 1.9313929274341477e-06, "loss": 0.4068, "step": 10142 }, { "epoch": 0.719106699751861, "grad_norm": 41.59124914884022, "learning_rate": 1.930486510803808e-06, "loss": 0.4404, "step": 10143 }, { "epoch": 0.7191775965969515, "grad_norm": 9.511997421878311, "learning_rate": 1.9295802560312586e-06, "loss": 0.3942, "step": 10144 }, { "epoch": 0.7192484934420418, "grad_norm": 2.7800744277884393, "learning_rate": 1.928674163164285e-06, "loss": 0.3816, "step": 10145 }, { "epoch": 0.7193193902871322, "grad_norm": 5.201235013632607, "learning_rate": 1.9277682322506688e-06, "loss": 0.4425, "step": 10146 }, { "epoch": 0.7193902871322226, "grad_norm": 4.383236972871427, "learning_rate": 1.9268624633381762e-06, "loss": 0.3794, "step": 10147 }, { "epoch": 0.719461183977313, "grad_norm": 1.8921273540815666, "learning_rate": 1.9259568564745736e-06, "loss": 0.3524, "step": 10148 }, { "epoch": 0.7195320808224034, "grad_norm": 1.8093950494660027, "learning_rate": 1.9250514117076118e-06, "loss": 0.3092, "step": 10149 }, { "epoch": 0.7196029776674938, "grad_norm": 4.062323345928806, "learning_rate": 1.9241461290850348e-06, "loss": 0.4028, "step": 10150 }, { "epoch": 0.7196738745125842, "grad_norm": 3.9671711524958115, "learning_rate": 1.923241008654582e-06, "loss": 0.3733, "step": 10151 }, { "epoch": 0.7197447713576746, "grad_norm": 4.9287460479588105, "learning_rate": 1.9223360504639774e-06, "loss": 0.3685, "step": 10152 }, { "epoch": 0.719815668202765, "grad_norm": 24.421644259944394, "learning_rate": 1.9214312545609438e-06, "loss": 0.3736, "step": 10153 }, { "epoch": 0.7198865650478554, "grad_norm": 3.5493088450015766, "learning_rate": 1.9205266209931894e-06, "loss": 0.3905, "step": 10154 }, { "epoch": 0.7199574618929457, "grad_norm": 4.119974397036144, "learning_rate": 1.9196221498084154e-06, "loss": 0.4508, "step": 10155 }, { "epoch": 0.7200283587380362, "grad_norm": 3.1955592341033556, "learning_rate": 1.9187178410543183e-06, "loss": 0.3896, "step": 10156 }, { "epoch": 0.7200992555831266, "grad_norm": 2.364515913467834, "learning_rate": 1.917813694778581e-06, "loss": 0.3377, "step": 10157 }, { "epoch": 0.7201701524282169, "grad_norm": 22.487586116407684, "learning_rate": 1.9169097110288787e-06, "loss": 0.4155, "step": 10158 }, { "epoch": 0.7202410492733073, "grad_norm": 2.8554151732905426, "learning_rate": 1.91600588985288e-06, "loss": 0.4091, "step": 10159 }, { "epoch": 0.7203119461183978, "grad_norm": 1.7200785680200565, "learning_rate": 1.915102231298247e-06, "loss": 0.3642, "step": 10160 }, { "epoch": 0.7203828429634881, "grad_norm": 3.077096678897873, "learning_rate": 1.914198735412628e-06, "loss": 0.3895, "step": 10161 }, { "epoch": 0.7204537398085785, "grad_norm": 3.018591993013431, "learning_rate": 1.913295402243663e-06, "loss": 0.4144, "step": 10162 }, { "epoch": 0.720524636653669, "grad_norm": 3.423220806721023, "learning_rate": 1.912392231838987e-06, "loss": 0.4445, "step": 10163 }, { "epoch": 0.7205955334987593, "grad_norm": 35.41920260544328, "learning_rate": 1.9114892242462292e-06, "loss": 0.3681, "step": 10164 }, { "epoch": 0.7206664303438497, "grad_norm": 2.375102540100727, "learning_rate": 1.910586379512998e-06, "loss": 0.3964, "step": 10165 }, { "epoch": 0.7207373271889401, "grad_norm": 5.094978278762005, "learning_rate": 1.9096836976869053e-06, "loss": 0.4442, "step": 10166 }, { "epoch": 0.7208082240340304, "grad_norm": 7.238433816413216, "learning_rate": 1.908781178815552e-06, "loss": 0.3708, "step": 10167 }, { "epoch": 0.7208791208791209, "grad_norm": 4.891163387658773, "learning_rate": 1.9078788229465267e-06, "loss": 0.3663, "step": 10168 }, { "epoch": 0.7209500177242113, "grad_norm": 2.1914932147776156, "learning_rate": 1.906976630127409e-06, "loss": 0.3607, "step": 10169 }, { "epoch": 0.7210209145693016, "grad_norm": 3.44844766564664, "learning_rate": 1.9060746004057762e-06, "loss": 0.4002, "step": 10170 }, { "epoch": 0.721091811414392, "grad_norm": 3.512583137898187, "learning_rate": 1.9051727338291898e-06, "loss": 0.3493, "step": 10171 }, { "epoch": 0.7211627082594825, "grad_norm": 2.1299255118597853, "learning_rate": 1.9042710304452084e-06, "loss": 0.4686, "step": 10172 }, { "epoch": 0.7212336051045728, "grad_norm": 2.8054972558756957, "learning_rate": 1.903369490301379e-06, "loss": 0.3459, "step": 10173 }, { "epoch": 0.7213045019496632, "grad_norm": 8.621951529088992, "learning_rate": 1.9024681134452378e-06, "loss": 0.4, "step": 10174 }, { "epoch": 0.7213753987947537, "grad_norm": 3.696026580284071, "learning_rate": 1.9015668999243192e-06, "loss": 0.4149, "step": 10175 }, { "epoch": 0.7214462956398441, "grad_norm": 3.3853430305062093, "learning_rate": 1.9006658497861414e-06, "loss": 0.3769, "step": 10176 }, { "epoch": 0.7215171924849344, "grad_norm": 2.0211879008900713, "learning_rate": 1.89976496307822e-06, "loss": 0.4068, "step": 10177 }, { "epoch": 0.7215880893300248, "grad_norm": 2.667173674596708, "learning_rate": 1.8988642398480571e-06, "loss": 0.3463, "step": 10178 }, { "epoch": 0.7216589861751153, "grad_norm": 4.235594281543383, "learning_rate": 1.8979636801431518e-06, "loss": 0.3918, "step": 10179 }, { "epoch": 0.7217298830202056, "grad_norm": 2.8504677549754978, "learning_rate": 1.8970632840109888e-06, "loss": 0.4571, "step": 10180 }, { "epoch": 0.721800779865296, "grad_norm": 1.1576871157265791, "learning_rate": 1.8961630514990458e-06, "loss": 0.626, "step": 10181 }, { "epoch": 0.7218716767103864, "grad_norm": 3.936644350116971, "learning_rate": 1.8952629826547942e-06, "loss": 0.4112, "step": 10182 }, { "epoch": 0.7219425735554768, "grad_norm": 4.17149898906343, "learning_rate": 1.8943630775256983e-06, "loss": 0.3993, "step": 10183 }, { "epoch": 0.7220134704005672, "grad_norm": 2.4833350484533114, "learning_rate": 1.8934633361592047e-06, "loss": 0.3845, "step": 10184 }, { "epoch": 0.7220843672456576, "grad_norm": 1.7422510047634843, "learning_rate": 1.89256375860276e-06, "loss": 0.3682, "step": 10185 }, { "epoch": 0.7221552640907479, "grad_norm": 5.124313625348452, "learning_rate": 1.8916643449038024e-06, "loss": 0.3926, "step": 10186 }, { "epoch": 0.7222261609358384, "grad_norm": 2.4915811187736554, "learning_rate": 1.890765095109756e-06, "loss": 0.3902, "step": 10187 }, { "epoch": 0.7222970577809288, "grad_norm": 4.359088956224518, "learning_rate": 1.8898660092680376e-06, "loss": 0.4294, "step": 10188 }, { "epoch": 0.7223679546260191, "grad_norm": 8.577932568129896, "learning_rate": 1.888967087426059e-06, "loss": 0.3847, "step": 10189 }, { "epoch": 0.7224388514711095, "grad_norm": 9.24959154561745, "learning_rate": 1.8880683296312214e-06, "loss": 0.427, "step": 10190 }, { "epoch": 0.7225097483162, "grad_norm": 2.649129395210073, "learning_rate": 1.8871697359309167e-06, "loss": 0.3572, "step": 10191 }, { "epoch": 0.7225806451612903, "grad_norm": 1.7619663664314065, "learning_rate": 1.8862713063725252e-06, "loss": 0.3975, "step": 10192 }, { "epoch": 0.7226515420063807, "grad_norm": 6.05426715929262, "learning_rate": 1.8853730410034266e-06, "loss": 0.4288, "step": 10193 }, { "epoch": 0.7227224388514711, "grad_norm": 2.9801365059341167, "learning_rate": 1.8844749398709845e-06, "loss": 0.3884, "step": 10194 }, { "epoch": 0.7227933356965615, "grad_norm": 8.138723462215808, "learning_rate": 1.8835770030225554e-06, "loss": 0.4056, "step": 10195 }, { "epoch": 0.7228642325416519, "grad_norm": 2.287646901399709, "learning_rate": 1.8826792305054913e-06, "loss": 0.3407, "step": 10196 }, { "epoch": 0.7229351293867423, "grad_norm": 3.6486899406197586, "learning_rate": 1.8817816223671286e-06, "loss": 0.4059, "step": 10197 }, { "epoch": 0.7230060262318326, "grad_norm": 3.21460988720629, "learning_rate": 1.8808841786548031e-06, "loss": 0.4044, "step": 10198 }, { "epoch": 0.7230769230769231, "grad_norm": 3.1840062940053917, "learning_rate": 1.8799868994158354e-06, "loss": 0.4336, "step": 10199 }, { "epoch": 0.7231478199220135, "grad_norm": 2.843102281460356, "learning_rate": 1.8790897846975381e-06, "loss": 0.4017, "step": 10200 }, { "epoch": 0.7232187167671038, "grad_norm": 2.9387168745304626, "learning_rate": 1.878192834547219e-06, "loss": 0.3973, "step": 10201 }, { "epoch": 0.7232896136121942, "grad_norm": 4.3778708769419845, "learning_rate": 1.8772960490121771e-06, "loss": 0.3875, "step": 10202 }, { "epoch": 0.7233605104572847, "grad_norm": 2.1551291841314244, "learning_rate": 1.8763994281396952e-06, "loss": 0.4009, "step": 10203 }, { "epoch": 0.7234314073023751, "grad_norm": 6.140459592406519, "learning_rate": 1.8755029719770556e-06, "loss": 0.4143, "step": 10204 }, { "epoch": 0.7235023041474654, "grad_norm": 3.7294495511262706, "learning_rate": 1.8746066805715307e-06, "loss": 0.401, "step": 10205 }, { "epoch": 0.7235732009925558, "grad_norm": 2.161030678046518, "learning_rate": 1.8737105539703809e-06, "loss": 0.4138, "step": 10206 }, { "epoch": 0.7236440978376463, "grad_norm": 64.93763619058048, "learning_rate": 1.872814592220858e-06, "loss": 0.3575, "step": 10207 }, { "epoch": 0.7237149946827366, "grad_norm": 3.668087749303973, "learning_rate": 1.871918795370209e-06, "loss": 0.3962, "step": 10208 }, { "epoch": 0.723785891527827, "grad_norm": 4.083702557511285, "learning_rate": 1.8710231634656723e-06, "loss": 0.4042, "step": 10209 }, { "epoch": 0.7238567883729174, "grad_norm": 2.9187616501423297, "learning_rate": 1.870127696554469e-06, "loss": 0.3974, "step": 10210 }, { "epoch": 0.7239276852180078, "grad_norm": 4.360889870954963, "learning_rate": 1.8692323946838214e-06, "loss": 0.3713, "step": 10211 }, { "epoch": 0.7239985820630982, "grad_norm": 4.022680868894691, "learning_rate": 1.8683372579009412e-06, "loss": 0.3833, "step": 10212 }, { "epoch": 0.7240694789081886, "grad_norm": 2.3900017489487824, "learning_rate": 1.8674422862530278e-06, "loss": 0.4175, "step": 10213 }, { "epoch": 0.7241403757532789, "grad_norm": 5.575838786426267, "learning_rate": 1.8665474797872712e-06, "loss": 0.3775, "step": 10214 }, { "epoch": 0.7242112725983694, "grad_norm": 4.752847113012837, "learning_rate": 1.8656528385508598e-06, "loss": 0.4076, "step": 10215 }, { "epoch": 0.7242821694434598, "grad_norm": 3.88666031971844, "learning_rate": 1.8647583625909648e-06, "loss": 0.394, "step": 10216 }, { "epoch": 0.7243530662885501, "grad_norm": 2.8402582314046088, "learning_rate": 1.8638640519547563e-06, "loss": 0.348, "step": 10217 }, { "epoch": 0.7244239631336405, "grad_norm": 5.391138517392902, "learning_rate": 1.8629699066893896e-06, "loss": 0.3927, "step": 10218 }, { "epoch": 0.724494859978731, "grad_norm": 3.1486780983336913, "learning_rate": 1.862075926842013e-06, "loss": 0.432, "step": 10219 }, { "epoch": 0.7245657568238213, "grad_norm": 2.2333449614386667, "learning_rate": 1.861182112459769e-06, "loss": 0.379, "step": 10220 }, { "epoch": 0.7246366536689117, "grad_norm": 2.852874386962634, "learning_rate": 1.860288463589786e-06, "loss": 0.3613, "step": 10221 }, { "epoch": 0.7247075505140022, "grad_norm": 2.176844362229613, "learning_rate": 1.859394980279191e-06, "loss": 0.3426, "step": 10222 }, { "epoch": 0.7247784473590925, "grad_norm": 2.317201405227442, "learning_rate": 1.858501662575094e-06, "loss": 0.4228, "step": 10223 }, { "epoch": 0.7248493442041829, "grad_norm": 5.394557483360892, "learning_rate": 1.8576085105246039e-06, "loss": 0.3497, "step": 10224 }, { "epoch": 0.7249202410492733, "grad_norm": 5.061522710717225, "learning_rate": 1.8567155241748159e-06, "loss": 0.38, "step": 10225 }, { "epoch": 0.7249911378943636, "grad_norm": 2.308866176295044, "learning_rate": 1.8558227035728149e-06, "loss": 0.4034, "step": 10226 }, { "epoch": 0.7250620347394541, "grad_norm": 2.59824182223808, "learning_rate": 1.854930048765683e-06, "loss": 0.448, "step": 10227 }, { "epoch": 0.7251329315845445, "grad_norm": 4.2777740368334065, "learning_rate": 1.8540375598004933e-06, "loss": 0.4615, "step": 10228 }, { "epoch": 0.7252038284296349, "grad_norm": 20.76651565609787, "learning_rate": 1.8531452367243013e-06, "loss": 0.4244, "step": 10229 }, { "epoch": 0.7252747252747253, "grad_norm": 2.106913330878564, "learning_rate": 1.8522530795841626e-06, "loss": 0.4125, "step": 10230 }, { "epoch": 0.7253456221198157, "grad_norm": 2.6376886663576964, "learning_rate": 1.8513610884271232e-06, "loss": 0.4181, "step": 10231 }, { "epoch": 0.7254165189649061, "grad_norm": 3.6628735020651333, "learning_rate": 1.8504692633002174e-06, "loss": 0.421, "step": 10232 }, { "epoch": 0.7254874158099964, "grad_norm": 5.162273943363029, "learning_rate": 1.8495776042504692e-06, "loss": 0.4304, "step": 10233 }, { "epoch": 0.7255583126550869, "grad_norm": 2.194548860347661, "learning_rate": 1.8486861113248982e-06, "loss": 0.3804, "step": 10234 }, { "epoch": 0.7256292095001773, "grad_norm": 6.058233603257383, "learning_rate": 1.8477947845705157e-06, "loss": 0.4, "step": 10235 }, { "epoch": 0.7257001063452676, "grad_norm": 4.688613268378458, "learning_rate": 1.8469036240343208e-06, "loss": 0.4172, "step": 10236 }, { "epoch": 0.725771003190358, "grad_norm": 3.6517384532353043, "learning_rate": 1.8460126297633018e-06, "loss": 0.4243, "step": 10237 }, { "epoch": 0.7258419000354485, "grad_norm": 15.276618076519174, "learning_rate": 1.8451218018044464e-06, "loss": 0.378, "step": 10238 }, { "epoch": 0.7259127968805388, "grad_norm": 3.8512439767023565, "learning_rate": 1.8442311402047263e-06, "loss": 0.4301, "step": 10239 }, { "epoch": 0.7259836937256292, "grad_norm": 0.9706564845095521, "learning_rate": 1.8433406450111058e-06, "loss": 0.6644, "step": 10240 }, { "epoch": 0.7260545905707196, "grad_norm": 2.59686023391419, "learning_rate": 1.842450316270543e-06, "loss": 0.3895, "step": 10241 }, { "epoch": 0.72612548741581, "grad_norm": 3.4024513909995484, "learning_rate": 1.8415601540299843e-06, "loss": 0.3632, "step": 10242 }, { "epoch": 0.7261963842609004, "grad_norm": 6.129411166951057, "learning_rate": 1.8406701583363707e-06, "loss": 0.4382, "step": 10243 }, { "epoch": 0.7262672811059908, "grad_norm": 2.960014761104922, "learning_rate": 1.8397803292366312e-06, "loss": 0.4412, "step": 10244 }, { "epoch": 0.7263381779510811, "grad_norm": 2.3603275941828454, "learning_rate": 1.8388906667776852e-06, "loss": 0.3415, "step": 10245 }, { "epoch": 0.7264090747961716, "grad_norm": 2.9568002755626885, "learning_rate": 1.8380011710064471e-06, "loss": 0.412, "step": 10246 }, { "epoch": 0.726479971641262, "grad_norm": 21.804931909903736, "learning_rate": 1.8371118419698242e-06, "loss": 0.3626, "step": 10247 }, { "epoch": 0.7265508684863523, "grad_norm": 2.17959941355361, "learning_rate": 1.8362226797147043e-06, "loss": 0.3578, "step": 10248 }, { "epoch": 0.7266217653314427, "grad_norm": 3.0041281422441126, "learning_rate": 1.8353336842879777e-06, "loss": 0.4437, "step": 10249 }, { "epoch": 0.7266926621765332, "grad_norm": 3.371452123606664, "learning_rate": 1.834444855736523e-06, "loss": 0.3675, "step": 10250 }, { "epoch": 0.7267635590216235, "grad_norm": 4.572259167933255, "learning_rate": 1.8335561941072067e-06, "loss": 0.3848, "step": 10251 }, { "epoch": 0.7268344558667139, "grad_norm": 2.695106633813575, "learning_rate": 1.8326676994468879e-06, "loss": 0.3908, "step": 10252 }, { "epoch": 0.7269053527118043, "grad_norm": 5.681010262153256, "learning_rate": 1.831779371802419e-06, "loss": 0.44, "step": 10253 }, { "epoch": 0.7269762495568948, "grad_norm": 2.6652529951448307, "learning_rate": 1.8308912112206433e-06, "loss": 0.3879, "step": 10254 }, { "epoch": 0.7270471464019851, "grad_norm": 2.774541554060757, "learning_rate": 1.8300032177483934e-06, "loss": 0.4032, "step": 10255 }, { "epoch": 0.7271180432470755, "grad_norm": 2.0117325490764117, "learning_rate": 1.829115391432491e-06, "loss": 0.388, "step": 10256 }, { "epoch": 0.727188940092166, "grad_norm": 6.2683794093983884, "learning_rate": 1.8282277323197568e-06, "loss": 0.3807, "step": 10257 }, { "epoch": 0.7272598369372563, "grad_norm": 4.549854365274063, "learning_rate": 1.8273402404569945e-06, "loss": 0.3986, "step": 10258 }, { "epoch": 0.7273307337823467, "grad_norm": 5.906824778706833, "learning_rate": 1.8264529158910016e-06, "loss": 0.3944, "step": 10259 }, { "epoch": 0.7274016306274371, "grad_norm": 2.544181545023727, "learning_rate": 1.82556575866857e-06, "loss": 0.4095, "step": 10260 }, { "epoch": 0.7274725274725274, "grad_norm": 2.7667450412674186, "learning_rate": 1.8246787688364776e-06, "loss": 0.3753, "step": 10261 }, { "epoch": 0.7275434243176179, "grad_norm": 7.751754119638844, "learning_rate": 1.8237919464414988e-06, "loss": 0.3668, "step": 10262 }, { "epoch": 0.7276143211627083, "grad_norm": 3.8087885356651383, "learning_rate": 1.822905291530393e-06, "loss": 0.343, "step": 10263 }, { "epoch": 0.7276852180077986, "grad_norm": 5.264029872378583, "learning_rate": 1.8220188041499182e-06, "loss": 0.3533, "step": 10264 }, { "epoch": 0.727756114852889, "grad_norm": 7.3731877835640605, "learning_rate": 1.8211324843468175e-06, "loss": 0.3634, "step": 10265 }, { "epoch": 0.7278270116979795, "grad_norm": 5.2179988022018495, "learning_rate": 1.8202463321678255e-06, "loss": 0.3586, "step": 10266 }, { "epoch": 0.7278979085430698, "grad_norm": 3.5666006882673083, "learning_rate": 1.8193603476596728e-06, "loss": 0.4088, "step": 10267 }, { "epoch": 0.7279688053881602, "grad_norm": 2.1034171068661083, "learning_rate": 1.8184745308690755e-06, "loss": 0.4319, "step": 10268 }, { "epoch": 0.7280397022332507, "grad_norm": 2.437231156491873, "learning_rate": 1.8175888818427457e-06, "loss": 0.4309, "step": 10269 }, { "epoch": 0.728110599078341, "grad_norm": 1.8377070103172892, "learning_rate": 1.816703400627383e-06, "loss": 0.3771, "step": 10270 }, { "epoch": 0.7281814959234314, "grad_norm": 2.4372252851162837, "learning_rate": 1.815818087269679e-06, "loss": 0.3713, "step": 10271 }, { "epoch": 0.7282523927685218, "grad_norm": 5.648671022120072, "learning_rate": 1.814932941816317e-06, "loss": 0.4004, "step": 10272 }, { "epoch": 0.7283232896136121, "grad_norm": 2.5594801491095898, "learning_rate": 1.8140479643139753e-06, "loss": 0.3793, "step": 10273 }, { "epoch": 0.7283941864587026, "grad_norm": 1.859639000732912, "learning_rate": 1.8131631548093127e-06, "loss": 0.3463, "step": 10274 }, { "epoch": 0.728465083303793, "grad_norm": 4.096960758438121, "learning_rate": 1.8122785133489895e-06, "loss": 0.4031, "step": 10275 }, { "epoch": 0.7285359801488833, "grad_norm": 2.9938593932857027, "learning_rate": 1.8113940399796554e-06, "loss": 0.3602, "step": 10276 }, { "epoch": 0.7286068769939738, "grad_norm": 3.308016225665887, "learning_rate": 1.8105097347479473e-06, "loss": 0.3728, "step": 10277 }, { "epoch": 0.7286777738390642, "grad_norm": 3.720510025413351, "learning_rate": 1.8096255977004929e-06, "loss": 0.4127, "step": 10278 }, { "epoch": 0.7287486706841546, "grad_norm": 2.1690515013842417, "learning_rate": 1.8087416288839165e-06, "loss": 0.3877, "step": 10279 }, { "epoch": 0.7288195675292449, "grad_norm": 3.4103491206663215, "learning_rate": 1.8078578283448311e-06, "loss": 0.3879, "step": 10280 }, { "epoch": 0.7288904643743354, "grad_norm": 3.8227973640359476, "learning_rate": 1.8069741961298394e-06, "loss": 0.3951, "step": 10281 }, { "epoch": 0.7289613612194258, "grad_norm": 3.407616724456566, "learning_rate": 1.8060907322855331e-06, "loss": 0.4142, "step": 10282 }, { "epoch": 0.7290322580645161, "grad_norm": 4.048756875497789, "learning_rate": 1.8052074368585026e-06, "loss": 0.4048, "step": 10283 }, { "epoch": 0.7291031549096065, "grad_norm": 4.444929614578321, "learning_rate": 1.8043243098953217e-06, "loss": 0.3858, "step": 10284 }, { "epoch": 0.729174051754697, "grad_norm": 1.0764980860139592, "learning_rate": 1.8034413514425581e-06, "loss": 0.6626, "step": 10285 }, { "epoch": 0.7292449485997873, "grad_norm": 0.9142095523319193, "learning_rate": 1.8025585615467734e-06, "loss": 0.6536, "step": 10286 }, { "epoch": 0.7293158454448777, "grad_norm": 3.75833789891779, "learning_rate": 1.8016759402545141e-06, "loss": 0.3865, "step": 10287 }, { "epoch": 0.7293867422899681, "grad_norm": 3.693162501693541, "learning_rate": 1.8007934876123256e-06, "loss": 0.3978, "step": 10288 }, { "epoch": 0.7294576391350585, "grad_norm": 3.250922293739258, "learning_rate": 1.7999112036667388e-06, "loss": 0.4279, "step": 10289 }, { "epoch": 0.7295285359801489, "grad_norm": 2.4242033113480246, "learning_rate": 1.7990290884642748e-06, "loss": 0.3958, "step": 10290 }, { "epoch": 0.7295994328252393, "grad_norm": 2.798990775491441, "learning_rate": 1.7981471420514495e-06, "loss": 0.4063, "step": 10291 }, { "epoch": 0.7296703296703296, "grad_norm": 2.7841159618394644, "learning_rate": 1.7972653644747718e-06, "loss": 0.4039, "step": 10292 }, { "epoch": 0.7297412265154201, "grad_norm": 4.742099896960437, "learning_rate": 1.7963837557807356e-06, "loss": 0.4076, "step": 10293 }, { "epoch": 0.7298121233605105, "grad_norm": 3.202753306086872, "learning_rate": 1.7955023160158274e-06, "loss": 0.3766, "step": 10294 }, { "epoch": 0.7298830202056008, "grad_norm": 3.2771692566906743, "learning_rate": 1.79462104522653e-06, "loss": 0.3421, "step": 10295 }, { "epoch": 0.7299539170506912, "grad_norm": 3.729776241622137, "learning_rate": 1.793739943459311e-06, "loss": 0.4306, "step": 10296 }, { "epoch": 0.7300248138957817, "grad_norm": 2.055847346463706, "learning_rate": 1.7928590107606303e-06, "loss": 0.3658, "step": 10297 }, { "epoch": 0.730095710740872, "grad_norm": 8.874807724748143, "learning_rate": 1.7919782471769426e-06, "loss": 0.3634, "step": 10298 }, { "epoch": 0.7301666075859624, "grad_norm": 3.8730539232109553, "learning_rate": 1.7910976527546914e-06, "loss": 0.3843, "step": 10299 }, { "epoch": 0.7302375044310528, "grad_norm": 3.8199314322633966, "learning_rate": 1.7902172275403106e-06, "loss": 0.3928, "step": 10300 }, { "epoch": 0.7303084012761432, "grad_norm": 3.252394491780848, "learning_rate": 1.7893369715802234e-06, "loss": 0.4158, "step": 10301 }, { "epoch": 0.7303792981212336, "grad_norm": 2.3588022792966066, "learning_rate": 1.78845688492085e-06, "loss": 0.3845, "step": 10302 }, { "epoch": 0.730450194966324, "grad_norm": 4.304635286731874, "learning_rate": 1.7875769676085964e-06, "loss": 0.4014, "step": 10303 }, { "epoch": 0.7305210918114144, "grad_norm": 2.971233166390049, "learning_rate": 1.7866972196898596e-06, "loss": 0.4112, "step": 10304 }, { "epoch": 0.7305919886565048, "grad_norm": 2.4349001102941457, "learning_rate": 1.7858176412110328e-06, "loss": 0.4176, "step": 10305 }, { "epoch": 0.7306628855015952, "grad_norm": 0.9385793302287503, "learning_rate": 1.7849382322184926e-06, "loss": 0.6695, "step": 10306 }, { "epoch": 0.7307337823466856, "grad_norm": 2.3491000896006025, "learning_rate": 1.7840589927586162e-06, "loss": 0.374, "step": 10307 }, { "epoch": 0.7308046791917759, "grad_norm": 2.33149314351406, "learning_rate": 1.7831799228777613e-06, "loss": 0.4567, "step": 10308 }, { "epoch": 0.7308755760368664, "grad_norm": 7.630167903256192, "learning_rate": 1.7823010226222865e-06, "loss": 0.439, "step": 10309 }, { "epoch": 0.7309464728819568, "grad_norm": 2.9299819880161904, "learning_rate": 1.7814222920385328e-06, "loss": 0.3705, "step": 10310 }, { "epoch": 0.7310173697270471, "grad_norm": 2.0888302056938732, "learning_rate": 1.7805437311728397e-06, "loss": 0.4188, "step": 10311 }, { "epoch": 0.7310882665721375, "grad_norm": 3.594266050424145, "learning_rate": 1.779665340071533e-06, "loss": 0.375, "step": 10312 }, { "epoch": 0.731159163417228, "grad_norm": 5.359942180309438, "learning_rate": 1.7787871187809297e-06, "loss": 0.4095, "step": 10313 }, { "epoch": 0.7312300602623183, "grad_norm": 4.462864872403553, "learning_rate": 1.7779090673473414e-06, "loss": 0.4283, "step": 10314 }, { "epoch": 0.7313009571074087, "grad_norm": 8.49611732871907, "learning_rate": 1.7770311858170675e-06, "loss": 0.3628, "step": 10315 }, { "epoch": 0.7313718539524992, "grad_norm": 2.255194537672861, "learning_rate": 1.7761534742363972e-06, "loss": 0.3725, "step": 10316 }, { "epoch": 0.7314427507975895, "grad_norm": 1.166866022659205, "learning_rate": 1.7752759326516156e-06, "loss": 0.6722, "step": 10317 }, { "epoch": 0.7315136476426799, "grad_norm": 2.9689286016420873, "learning_rate": 1.7743985611089965e-06, "loss": 0.4031, "step": 10318 }, { "epoch": 0.7315845444877703, "grad_norm": 2.445295703953615, "learning_rate": 1.7735213596548034e-06, "loss": 0.4025, "step": 10319 }, { "epoch": 0.7316554413328606, "grad_norm": 9.564533150329735, "learning_rate": 1.77264432833529e-06, "loss": 0.3909, "step": 10320 }, { "epoch": 0.7317263381779511, "grad_norm": 5.4452814031937065, "learning_rate": 1.7717674671967062e-06, "loss": 0.4293, "step": 10321 }, { "epoch": 0.7317972350230415, "grad_norm": 6.38252702862695, "learning_rate": 1.7708907762852877e-06, "loss": 0.4312, "step": 10322 }, { "epoch": 0.7318681318681318, "grad_norm": 3.146494991668284, "learning_rate": 1.7700142556472616e-06, "loss": 0.4144, "step": 10323 }, { "epoch": 0.7319390287132223, "grad_norm": 1.853417803275931, "learning_rate": 1.7691379053288494e-06, "loss": 0.4032, "step": 10324 }, { "epoch": 0.7320099255583127, "grad_norm": 2.4161817402202397, "learning_rate": 1.7682617253762629e-06, "loss": 0.4605, "step": 10325 }, { "epoch": 0.732080822403403, "grad_norm": 5.7713408721869435, "learning_rate": 1.7673857158357022e-06, "loss": 0.4086, "step": 10326 }, { "epoch": 0.7321517192484934, "grad_norm": 2.415163981007564, "learning_rate": 1.7665098767533583e-06, "loss": 0.3357, "step": 10327 }, { "epoch": 0.7322226160935839, "grad_norm": 1.4293881182126673, "learning_rate": 1.7656342081754185e-06, "loss": 0.3703, "step": 10328 }, { "epoch": 0.7322935129386742, "grad_norm": 3.197790568428672, "learning_rate": 1.7647587101480556e-06, "loss": 0.3862, "step": 10329 }, { "epoch": 0.7323644097837646, "grad_norm": 3.108751749596631, "learning_rate": 1.7638833827174333e-06, "loss": 0.3753, "step": 10330 }, { "epoch": 0.732435306628855, "grad_norm": 3.454744385621726, "learning_rate": 1.7630082259297127e-06, "loss": 0.4111, "step": 10331 }, { "epoch": 0.7325062034739455, "grad_norm": 2.708920552178013, "learning_rate": 1.7621332398310365e-06, "loss": 0.4202, "step": 10332 }, { "epoch": 0.7325771003190358, "grad_norm": 3.1143122454328775, "learning_rate": 1.761258424467548e-06, "loss": 0.4362, "step": 10333 }, { "epoch": 0.7326479971641262, "grad_norm": 1.0760648187727213, "learning_rate": 1.7603837798853746e-06, "loss": 0.6321, "step": 10334 }, { "epoch": 0.7327188940092166, "grad_norm": 2.0223042111584046, "learning_rate": 1.7595093061306357e-06, "loss": 0.357, "step": 10335 }, { "epoch": 0.732789790854307, "grad_norm": 2.642377810719935, "learning_rate": 1.7586350032494442e-06, "loss": 0.3396, "step": 10336 }, { "epoch": 0.7328606876993974, "grad_norm": 2.9345229864583953, "learning_rate": 1.7577608712879046e-06, "loss": 0.3951, "step": 10337 }, { "epoch": 0.7329315845444878, "grad_norm": 2.84742247375894, "learning_rate": 1.7568869102921093e-06, "loss": 0.4242, "step": 10338 }, { "epoch": 0.7330024813895781, "grad_norm": 2.0472370506990147, "learning_rate": 1.756013120308141e-06, "loss": 0.4253, "step": 10339 }, { "epoch": 0.7330733782346686, "grad_norm": 2.3474365791364677, "learning_rate": 1.7551395013820783e-06, "loss": 0.3767, "step": 10340 }, { "epoch": 0.733144275079759, "grad_norm": 2.857451601729729, "learning_rate": 1.7542660535599865e-06, "loss": 0.3657, "step": 10341 }, { "epoch": 0.7332151719248493, "grad_norm": 2.4631710051358224, "learning_rate": 1.753392776887921e-06, "loss": 0.423, "step": 10342 }, { "epoch": 0.7332860687699397, "grad_norm": 2.7298096306944717, "learning_rate": 1.752519671411933e-06, "loss": 0.3872, "step": 10343 }, { "epoch": 0.7333569656150302, "grad_norm": 2.6711416382684043, "learning_rate": 1.7516467371780628e-06, "loss": 0.3517, "step": 10344 }, { "epoch": 0.7334278624601205, "grad_norm": 4.6241906813012035, "learning_rate": 1.7507739742323393e-06, "loss": 0.3762, "step": 10345 }, { "epoch": 0.7334987593052109, "grad_norm": 5.738908315022081, "learning_rate": 1.7499013826207827e-06, "loss": 0.3819, "step": 10346 }, { "epoch": 0.7335696561503013, "grad_norm": 2.850610807246807, "learning_rate": 1.7490289623894085e-06, "loss": 0.3819, "step": 10347 }, { "epoch": 0.7336405529953917, "grad_norm": 2.8548692260098414, "learning_rate": 1.7481567135842182e-06, "loss": 0.474, "step": 10348 }, { "epoch": 0.7337114498404821, "grad_norm": 2.8153597522753273, "learning_rate": 1.7472846362512054e-06, "loss": 0.3339, "step": 10349 }, { "epoch": 0.7337823466855725, "grad_norm": 4.171111556605831, "learning_rate": 1.7464127304363576e-06, "loss": 0.4201, "step": 10350 }, { "epoch": 0.7338532435306628, "grad_norm": 1.086838816868837, "learning_rate": 1.7455409961856484e-06, "loss": 0.694, "step": 10351 }, { "epoch": 0.7339241403757533, "grad_norm": 4.734414510391371, "learning_rate": 1.744669433545048e-06, "loss": 0.3801, "step": 10352 }, { "epoch": 0.7339950372208437, "grad_norm": 3.5412718122694833, "learning_rate": 1.7437980425605112e-06, "loss": 0.4529, "step": 10353 }, { "epoch": 0.734065934065934, "grad_norm": 2.7427838006413605, "learning_rate": 1.7429268232779912e-06, "loss": 0.3897, "step": 10354 }, { "epoch": 0.7341368309110244, "grad_norm": 2.8803094110083993, "learning_rate": 1.7420557757434237e-06, "loss": 0.3938, "step": 10355 }, { "epoch": 0.7342077277561149, "grad_norm": 3.6141341816567847, "learning_rate": 1.7411849000027436e-06, "loss": 0.4064, "step": 10356 }, { "epoch": 0.7342786246012053, "grad_norm": 3.6417048315263996, "learning_rate": 1.740314196101871e-06, "loss": 0.376, "step": 10357 }, { "epoch": 0.7343495214462956, "grad_norm": 7.906730467666766, "learning_rate": 1.7394436640867173e-06, "loss": 0.3642, "step": 10358 }, { "epoch": 0.734420418291386, "grad_norm": 3.188744351152074, "learning_rate": 1.7385733040031894e-06, "loss": 0.3843, "step": 10359 }, { "epoch": 0.7344913151364765, "grad_norm": 1.7425110467599683, "learning_rate": 1.7377031158971808e-06, "loss": 0.4069, "step": 10360 }, { "epoch": 0.7345622119815668, "grad_norm": 4.214749283741804, "learning_rate": 1.736833099814575e-06, "loss": 0.3594, "step": 10361 }, { "epoch": 0.7346331088266572, "grad_norm": 4.005062391683634, "learning_rate": 1.7359632558012506e-06, "loss": 0.3886, "step": 10362 }, { "epoch": 0.7347040056717477, "grad_norm": 10.694598104486097, "learning_rate": 1.7350935839030763e-06, "loss": 0.4228, "step": 10363 }, { "epoch": 0.734774902516838, "grad_norm": 4.398677505686363, "learning_rate": 1.7342240841659098e-06, "loss": 0.4282, "step": 10364 }, { "epoch": 0.7348457993619284, "grad_norm": 2.67260782188661, "learning_rate": 1.7333547566355985e-06, "loss": 0.3899, "step": 10365 }, { "epoch": 0.7349166962070188, "grad_norm": 1.111995350582759, "learning_rate": 1.7324856013579838e-06, "loss": 0.6497, "step": 10366 }, { "epoch": 0.7349875930521091, "grad_norm": 2.7301103867583394, "learning_rate": 1.7316166183789007e-06, "loss": 0.3785, "step": 10367 }, { "epoch": 0.7350584898971996, "grad_norm": 2.4500816757381285, "learning_rate": 1.7307478077441647e-06, "loss": 0.3543, "step": 10368 }, { "epoch": 0.73512938674229, "grad_norm": 7.470360579357245, "learning_rate": 1.729879169499592e-06, "loss": 0.4207, "step": 10369 }, { "epoch": 0.7352002835873803, "grad_norm": 2.414166646599455, "learning_rate": 1.7290107036909886e-06, "loss": 0.4271, "step": 10370 }, { "epoch": 0.7352711804324708, "grad_norm": 3.603009411663456, "learning_rate": 1.728142410364147e-06, "loss": 0.3638, "step": 10371 }, { "epoch": 0.7353420772775612, "grad_norm": 2.170102136991277, "learning_rate": 1.727274289564852e-06, "loss": 0.3719, "step": 10372 }, { "epoch": 0.7354129741226515, "grad_norm": 4.633478097650681, "learning_rate": 1.7264063413388832e-06, "loss": 0.4306, "step": 10373 }, { "epoch": 0.7354838709677419, "grad_norm": 3.482325204447108, "learning_rate": 1.725538565732005e-06, "loss": 0.3701, "step": 10374 }, { "epoch": 0.7355547678128324, "grad_norm": 3.2322123544420607, "learning_rate": 1.7246709627899794e-06, "loss": 0.4217, "step": 10375 }, { "epoch": 0.7356256646579227, "grad_norm": 2.6795064192141442, "learning_rate": 1.7238035325585539e-06, "loss": 0.3825, "step": 10376 }, { "epoch": 0.7356965615030131, "grad_norm": 4.932290402799753, "learning_rate": 1.7229362750834667e-06, "loss": 0.3933, "step": 10377 }, { "epoch": 0.7357674583481035, "grad_norm": 6.423473271222328, "learning_rate": 1.7220691904104531e-06, "loss": 0.393, "step": 10378 }, { "epoch": 0.7358383551931938, "grad_norm": 3.0333484851685113, "learning_rate": 1.7212022785852328e-06, "loss": 0.3424, "step": 10379 }, { "epoch": 0.7359092520382843, "grad_norm": 3.6890475589464606, "learning_rate": 1.720335539653517e-06, "loss": 0.3867, "step": 10380 }, { "epoch": 0.7359801488833747, "grad_norm": 13.679916083626875, "learning_rate": 1.719468973661012e-06, "loss": 0.3947, "step": 10381 }, { "epoch": 0.7360510457284651, "grad_norm": 3.8768741941833307, "learning_rate": 1.7186025806534128e-06, "loss": 0.4018, "step": 10382 }, { "epoch": 0.7361219425735555, "grad_norm": 2.7767043801423363, "learning_rate": 1.7177363606764048e-06, "loss": 0.4042, "step": 10383 }, { "epoch": 0.7361928394186459, "grad_norm": 3.7879251149665096, "learning_rate": 1.716870313775662e-06, "loss": 0.3843, "step": 10384 }, { "epoch": 0.7362637362637363, "grad_norm": 3.868156412709206, "learning_rate": 1.7160044399968552e-06, "loss": 0.3974, "step": 10385 }, { "epoch": 0.7363346331088266, "grad_norm": 0.9034677606217509, "learning_rate": 1.7151387393856407e-06, "loss": 0.6464, "step": 10386 }, { "epoch": 0.7364055299539171, "grad_norm": 3.4096781024282405, "learning_rate": 1.714273211987666e-06, "loss": 0.4393, "step": 10387 }, { "epoch": 0.7364764267990075, "grad_norm": 2.7029243982597997, "learning_rate": 1.7134078578485724e-06, "loss": 0.4612, "step": 10388 }, { "epoch": 0.7365473236440978, "grad_norm": 2.1769518795703338, "learning_rate": 1.712542677013993e-06, "loss": 0.3737, "step": 10389 }, { "epoch": 0.7366182204891882, "grad_norm": 5.42733146228149, "learning_rate": 1.7116776695295473e-06, "loss": 0.4175, "step": 10390 }, { "epoch": 0.7366891173342787, "grad_norm": 4.013876964081573, "learning_rate": 1.7108128354408466e-06, "loss": 0.3606, "step": 10391 }, { "epoch": 0.736760014179369, "grad_norm": 2.372556517084739, "learning_rate": 1.7099481747934964e-06, "loss": 0.3608, "step": 10392 }, { "epoch": 0.7368309110244594, "grad_norm": 6.0876515726389355, "learning_rate": 1.7090836876330914e-06, "loss": 0.4155, "step": 10393 }, { "epoch": 0.7369018078695498, "grad_norm": 4.9186674190379085, "learning_rate": 1.7082193740052128e-06, "loss": 0.3977, "step": 10394 }, { "epoch": 0.7369727047146402, "grad_norm": 2.7000133053068898, "learning_rate": 1.7073552339554395e-06, "loss": 0.3728, "step": 10395 }, { "epoch": 0.7370436015597306, "grad_norm": 1.7850982675583171, "learning_rate": 1.7064912675293393e-06, "loss": 0.3744, "step": 10396 }, { "epoch": 0.737114498404821, "grad_norm": 4.44547819930173, "learning_rate": 1.7056274747724688e-06, "loss": 0.4495, "step": 10397 }, { "epoch": 0.7371853952499113, "grad_norm": 2.896068186344916, "learning_rate": 1.7047638557303737e-06, "loss": 0.3811, "step": 10398 }, { "epoch": 0.7372562920950018, "grad_norm": 3.0365787392714916, "learning_rate": 1.7039004104485979e-06, "loss": 0.3973, "step": 10399 }, { "epoch": 0.7373271889400922, "grad_norm": 2.066947249255698, "learning_rate": 1.7030371389726675e-06, "loss": 0.3491, "step": 10400 }, { "epoch": 0.7373980857851825, "grad_norm": 2.517129249201248, "learning_rate": 1.7021740413481063e-06, "loss": 0.4141, "step": 10401 }, { "epoch": 0.7374689826302729, "grad_norm": 2.3657913704688847, "learning_rate": 1.7013111176204255e-06, "loss": 0.3718, "step": 10402 }, { "epoch": 0.7375398794753634, "grad_norm": 2.8895132112565913, "learning_rate": 1.7004483678351258e-06, "loss": 0.335, "step": 10403 }, { "epoch": 0.7376107763204537, "grad_norm": 2.0889426373897315, "learning_rate": 1.6995857920377035e-06, "loss": 0.4071, "step": 10404 }, { "epoch": 0.7376816731655441, "grad_norm": 2.801854177530377, "learning_rate": 1.6987233902736417e-06, "loss": 0.4037, "step": 10405 }, { "epoch": 0.7377525700106345, "grad_norm": 2.4880040175841844, "learning_rate": 1.6978611625884135e-06, "loss": 0.3616, "step": 10406 }, { "epoch": 0.737823466855725, "grad_norm": 3.131106318520753, "learning_rate": 1.6969991090274863e-06, "loss": 0.3728, "step": 10407 }, { "epoch": 0.7378943637008153, "grad_norm": 3.7581981799461306, "learning_rate": 1.6961372296363187e-06, "loss": 0.3784, "step": 10408 }, { "epoch": 0.7379652605459057, "grad_norm": 2.3878040452664147, "learning_rate": 1.6952755244603576e-06, "loss": 0.4167, "step": 10409 }, { "epoch": 0.7380361573909962, "grad_norm": 3.132184855860041, "learning_rate": 1.6944139935450382e-06, "loss": 0.4743, "step": 10410 }, { "epoch": 0.7381070542360865, "grad_norm": 2.261236483984638, "learning_rate": 1.693552636935792e-06, "loss": 0.3895, "step": 10411 }, { "epoch": 0.7381779510811769, "grad_norm": 2.5427033503983982, "learning_rate": 1.692691454678042e-06, "loss": 0.3598, "step": 10412 }, { "epoch": 0.7382488479262673, "grad_norm": 3.067503642863374, "learning_rate": 1.6918304468171932e-06, "loss": 0.3882, "step": 10413 }, { "epoch": 0.7383197447713576, "grad_norm": 2.279678154556136, "learning_rate": 1.6909696133986493e-06, "loss": 0.3509, "step": 10414 }, { "epoch": 0.7383906416164481, "grad_norm": 6.693610193754575, "learning_rate": 1.6901089544678057e-06, "loss": 0.3775, "step": 10415 }, { "epoch": 0.7384615384615385, "grad_norm": 2.5112214733543965, "learning_rate": 1.689248470070043e-06, "loss": 0.3806, "step": 10416 }, { "epoch": 0.7385324353066288, "grad_norm": 3.3249887662077042, "learning_rate": 1.6883881602507334e-06, "loss": 0.3776, "step": 10417 }, { "epoch": 0.7386033321517192, "grad_norm": 2.4681189097897476, "learning_rate": 1.6875280250552457e-06, "loss": 0.3589, "step": 10418 }, { "epoch": 0.7386742289968097, "grad_norm": 4.882849708888928, "learning_rate": 1.686668064528932e-06, "loss": 0.3626, "step": 10419 }, { "epoch": 0.7387451258419, "grad_norm": 2.9237407375049287, "learning_rate": 1.6858082787171421e-06, "loss": 0.356, "step": 10420 }, { "epoch": 0.7388160226869904, "grad_norm": 2.7479121782841447, "learning_rate": 1.6849486676652105e-06, "loss": 0.3722, "step": 10421 }, { "epoch": 0.7388869195320809, "grad_norm": 3.6425927623166348, "learning_rate": 1.6840892314184649e-06, "loss": 0.3553, "step": 10422 }, { "epoch": 0.7389578163771712, "grad_norm": 2.5581939401531244, "learning_rate": 1.6832299700222265e-06, "loss": 0.3771, "step": 10423 }, { "epoch": 0.7390287132222616, "grad_norm": 3.436497514434243, "learning_rate": 1.6823708835218016e-06, "loss": 0.3608, "step": 10424 }, { "epoch": 0.739099610067352, "grad_norm": 3.082365346450737, "learning_rate": 1.681511971962494e-06, "loss": 0.4235, "step": 10425 }, { "epoch": 0.7391705069124423, "grad_norm": 5.533404605514323, "learning_rate": 1.6806532353895917e-06, "loss": 0.3695, "step": 10426 }, { "epoch": 0.7392414037575328, "grad_norm": 3.0838043016107326, "learning_rate": 1.679794673848379e-06, "loss": 0.4036, "step": 10427 }, { "epoch": 0.7393123006026232, "grad_norm": 5.6789603764833965, "learning_rate": 1.6789362873841275e-06, "loss": 0.3942, "step": 10428 }, { "epoch": 0.7393831974477135, "grad_norm": 2.0191068799610705, "learning_rate": 1.6780780760420983e-06, "loss": 0.3984, "step": 10429 }, { "epoch": 0.739454094292804, "grad_norm": 6.844203505145276, "learning_rate": 1.6772200398675487e-06, "loss": 0.3861, "step": 10430 }, { "epoch": 0.7395249911378944, "grad_norm": 2.2437000687185367, "learning_rate": 1.6763621789057249e-06, "loss": 0.386, "step": 10431 }, { "epoch": 0.7395958879829848, "grad_norm": 4.124032641976636, "learning_rate": 1.6755044932018577e-06, "loss": 0.3925, "step": 10432 }, { "epoch": 0.7396667848280751, "grad_norm": 2.674067970448956, "learning_rate": 1.6746469828011752e-06, "loss": 0.3926, "step": 10433 }, { "epoch": 0.7397376816731656, "grad_norm": 4.462564130878272, "learning_rate": 1.673789647748898e-06, "loss": 0.3551, "step": 10434 }, { "epoch": 0.739808578518256, "grad_norm": 12.765470337562993, "learning_rate": 1.6729324880902308e-06, "loss": 0.3723, "step": 10435 }, { "epoch": 0.7398794753633463, "grad_norm": 2.590272324863309, "learning_rate": 1.6720755038703713e-06, "loss": 0.3944, "step": 10436 }, { "epoch": 0.7399503722084367, "grad_norm": 2.4650456674943526, "learning_rate": 1.6712186951345128e-06, "loss": 0.3591, "step": 10437 }, { "epoch": 0.7400212690535272, "grad_norm": 1.9371615642651634, "learning_rate": 1.6703620619278327e-06, "loss": 0.4117, "step": 10438 }, { "epoch": 0.7400921658986175, "grad_norm": 2.9096807058182628, "learning_rate": 1.669505604295501e-06, "loss": 0.4248, "step": 10439 }, { "epoch": 0.7401630627437079, "grad_norm": 8.52103988395587, "learning_rate": 1.6686493222826805e-06, "loss": 0.365, "step": 10440 }, { "epoch": 0.7402339595887983, "grad_norm": 6.936186074918194, "learning_rate": 1.6677932159345262e-06, "loss": 0.3851, "step": 10441 }, { "epoch": 0.7403048564338887, "grad_norm": 6.121295789199004, "learning_rate": 1.6669372852961785e-06, "loss": 0.4089, "step": 10442 }, { "epoch": 0.7403757532789791, "grad_norm": 4.268237211194446, "learning_rate": 1.6660815304127703e-06, "loss": 0.4223, "step": 10443 }, { "epoch": 0.7404466501240695, "grad_norm": 2.367049188397197, "learning_rate": 1.6652259513294295e-06, "loss": 0.3834, "step": 10444 }, { "epoch": 0.7405175469691598, "grad_norm": 1.599849854183569, "learning_rate": 1.6643705480912676e-06, "loss": 0.3759, "step": 10445 }, { "epoch": 0.7405884438142503, "grad_norm": 2.779841174469385, "learning_rate": 1.6635153207433942e-06, "loss": 0.3799, "step": 10446 }, { "epoch": 0.7406593406593407, "grad_norm": 5.038979314941201, "learning_rate": 1.6626602693309052e-06, "loss": 0.3828, "step": 10447 }, { "epoch": 0.740730237504431, "grad_norm": 1.1935218560488436, "learning_rate": 1.6618053938988854e-06, "loss": 0.6825, "step": 10448 }, { "epoch": 0.7408011343495214, "grad_norm": 4.120644894472635, "learning_rate": 1.6609506944924164e-06, "loss": 0.4036, "step": 10449 }, { "epoch": 0.7408720311946119, "grad_norm": 1.9598115544007966, "learning_rate": 1.6600961711565666e-06, "loss": 0.3656, "step": 10450 }, { "epoch": 0.7409429280397022, "grad_norm": 3.0285573298157353, "learning_rate": 1.6592418239363927e-06, "loss": 0.3899, "step": 10451 }, { "epoch": 0.7410138248847926, "grad_norm": 4.471492269262128, "learning_rate": 1.6583876528769478e-06, "loss": 0.3625, "step": 10452 }, { "epoch": 0.741084721729883, "grad_norm": 2.1331611613578585, "learning_rate": 1.6575336580232736e-06, "loss": 0.3628, "step": 10453 }, { "epoch": 0.7411556185749734, "grad_norm": 2.6535134971313097, "learning_rate": 1.6566798394204015e-06, "loss": 0.39, "step": 10454 }, { "epoch": 0.7412265154200638, "grad_norm": 3.1035647196196248, "learning_rate": 1.655826197113351e-06, "loss": 0.4171, "step": 10455 }, { "epoch": 0.7412974122651542, "grad_norm": 3.854971337116575, "learning_rate": 1.654972731147138e-06, "loss": 0.4323, "step": 10456 }, { "epoch": 0.7413683091102445, "grad_norm": 17.252428828600976, "learning_rate": 1.6541194415667693e-06, "loss": 0.3888, "step": 10457 }, { "epoch": 0.741439205955335, "grad_norm": 5.820831661705933, "learning_rate": 1.653266328417233e-06, "loss": 0.4138, "step": 10458 }, { "epoch": 0.7415101028004254, "grad_norm": 1.2540383329256608, "learning_rate": 1.652413391743517e-06, "loss": 0.6691, "step": 10459 }, { "epoch": 0.7415809996455158, "grad_norm": 3.145549031769501, "learning_rate": 1.6515606315906003e-06, "loss": 0.4019, "step": 10460 }, { "epoch": 0.7416518964906061, "grad_norm": 3.1872256383935498, "learning_rate": 1.6507080480034471e-06, "loss": 0.4106, "step": 10461 }, { "epoch": 0.7417227933356966, "grad_norm": 6.529768425187225, "learning_rate": 1.6498556410270134e-06, "loss": 0.3621, "step": 10462 }, { "epoch": 0.741793690180787, "grad_norm": 2.1279348369687114, "learning_rate": 1.6490034107062502e-06, "loss": 0.4046, "step": 10463 }, { "epoch": 0.7418645870258773, "grad_norm": 3.85118629165019, "learning_rate": 1.648151357086093e-06, "loss": 0.3813, "step": 10464 }, { "epoch": 0.7419354838709677, "grad_norm": 3.2267954759443156, "learning_rate": 1.647299480211475e-06, "loss": 0.4067, "step": 10465 }, { "epoch": 0.7420063807160582, "grad_norm": 4.224296750839083, "learning_rate": 1.6464477801273143e-06, "loss": 0.4144, "step": 10466 }, { "epoch": 0.7420772775611485, "grad_norm": 2.7143788638795896, "learning_rate": 1.6455962568785205e-06, "loss": 0.3866, "step": 10467 }, { "epoch": 0.7421481744062389, "grad_norm": 4.195098270838241, "learning_rate": 1.6447449105099977e-06, "loss": 0.3958, "step": 10468 }, { "epoch": 0.7422190712513294, "grad_norm": 7.633056870327458, "learning_rate": 1.6438937410666356e-06, "loss": 0.3931, "step": 10469 }, { "epoch": 0.7422899680964197, "grad_norm": 5.617405429128436, "learning_rate": 1.6430427485933193e-06, "loss": 0.3355, "step": 10470 }, { "epoch": 0.7423608649415101, "grad_norm": 1.1183141652055237, "learning_rate": 1.6421919331349196e-06, "loss": 0.7165, "step": 10471 }, { "epoch": 0.7424317617866005, "grad_norm": 5.197558505619296, "learning_rate": 1.6413412947363032e-06, "loss": 0.4093, "step": 10472 }, { "epoch": 0.7425026586316908, "grad_norm": 2.3025653845910607, "learning_rate": 1.6404908334423247e-06, "loss": 0.4195, "step": 10473 }, { "epoch": 0.7425735554767813, "grad_norm": 2.385325503392096, "learning_rate": 1.6396405492978267e-06, "loss": 0.3929, "step": 10474 }, { "epoch": 0.7426444523218717, "grad_norm": 2.4819272187440964, "learning_rate": 1.6387904423476475e-06, "loss": 0.4031, "step": 10475 }, { "epoch": 0.742715349166962, "grad_norm": 3.4222723521793195, "learning_rate": 1.637940512636617e-06, "loss": 0.3916, "step": 10476 }, { "epoch": 0.7427862460120525, "grad_norm": 2.6527916247996504, "learning_rate": 1.6370907602095465e-06, "loss": 0.379, "step": 10477 }, { "epoch": 0.7428571428571429, "grad_norm": 1.069307445580149, "learning_rate": 1.6362411851112465e-06, "loss": 0.715, "step": 10478 }, { "epoch": 0.7429280397022332, "grad_norm": 5.743595849510704, "learning_rate": 1.6353917873865183e-06, "loss": 0.3926, "step": 10479 }, { "epoch": 0.7429989365473236, "grad_norm": 3.2341523006264055, "learning_rate": 1.6345425670801485e-06, "loss": 0.3711, "step": 10480 }, { "epoch": 0.7430698333924141, "grad_norm": 2.7801054196100075, "learning_rate": 1.6336935242369167e-06, "loss": 0.3967, "step": 10481 }, { "epoch": 0.7431407302375044, "grad_norm": 3.671910226290651, "learning_rate": 1.6328446589015968e-06, "loss": 0.3417, "step": 10482 }, { "epoch": 0.7432116270825948, "grad_norm": 6.11387712386513, "learning_rate": 1.631995971118946e-06, "loss": 0.4575, "step": 10483 }, { "epoch": 0.7432825239276852, "grad_norm": 3.7604010721353323, "learning_rate": 1.6311474609337201e-06, "loss": 0.4248, "step": 10484 }, { "epoch": 0.7433534207727757, "grad_norm": 2.0678132809247867, "learning_rate": 1.6302991283906583e-06, "loss": 0.3551, "step": 10485 }, { "epoch": 0.743424317617866, "grad_norm": 5.917713962572539, "learning_rate": 1.6294509735344966e-06, "loss": 0.3441, "step": 10486 }, { "epoch": 0.7434952144629564, "grad_norm": 1.949519413740679, "learning_rate": 1.628602996409958e-06, "loss": 0.4035, "step": 10487 }, { "epoch": 0.7435661113080468, "grad_norm": 4.060871249554586, "learning_rate": 1.6277551970617556e-06, "loss": 0.362, "step": 10488 }, { "epoch": 0.7436370081531372, "grad_norm": 6.016956293559916, "learning_rate": 1.6269075755345965e-06, "loss": 0.3803, "step": 10489 }, { "epoch": 0.7437079049982276, "grad_norm": 3.5244735858461853, "learning_rate": 1.626060131873174e-06, "loss": 0.3601, "step": 10490 }, { "epoch": 0.743778801843318, "grad_norm": 2.96516372044201, "learning_rate": 1.6252128661221777e-06, "loss": 0.3745, "step": 10491 }, { "epoch": 0.7438496986884083, "grad_norm": 2.2238841133925873, "learning_rate": 1.6243657783262828e-06, "loss": 0.3622, "step": 10492 }, { "epoch": 0.7439205955334988, "grad_norm": 5.362312133418126, "learning_rate": 1.6235188685301556e-06, "loss": 0.3989, "step": 10493 }, { "epoch": 0.7439914923785892, "grad_norm": 3.428419802707832, "learning_rate": 1.6226721367784564e-06, "loss": 0.3323, "step": 10494 }, { "epoch": 0.7440623892236795, "grad_norm": 6.0794529993618225, "learning_rate": 1.6218255831158341e-06, "loss": 0.3295, "step": 10495 }, { "epoch": 0.7441332860687699, "grad_norm": 2.772687359099948, "learning_rate": 1.620979207586925e-06, "loss": 0.3848, "step": 10496 }, { "epoch": 0.7442041829138604, "grad_norm": 2.352561087798288, "learning_rate": 1.620133010236362e-06, "loss": 0.3944, "step": 10497 }, { "epoch": 0.7442750797589507, "grad_norm": 2.4671514076445, "learning_rate": 1.619286991108766e-06, "loss": 0.4478, "step": 10498 }, { "epoch": 0.7443459766040411, "grad_norm": 2.48407400729136, "learning_rate": 1.618441150248748e-06, "loss": 0.3614, "step": 10499 }, { "epoch": 0.7444168734491315, "grad_norm": 2.1307878068670076, "learning_rate": 1.617595487700907e-06, "loss": 0.4001, "step": 10500 }, { "epoch": 0.7444877702942219, "grad_norm": 3.6217563438633062, "learning_rate": 1.6167500035098382e-06, "loss": 0.382, "step": 10501 }, { "epoch": 0.7445586671393123, "grad_norm": 3.01344860248831, "learning_rate": 1.6159046977201271e-06, "loss": 0.3954, "step": 10502 }, { "epoch": 0.7446295639844027, "grad_norm": 2.0226574123437886, "learning_rate": 1.6150595703763404e-06, "loss": 0.3807, "step": 10503 }, { "epoch": 0.744700460829493, "grad_norm": 16.322184504546676, "learning_rate": 1.614214621523047e-06, "loss": 0.4707, "step": 10504 }, { "epoch": 0.7447713576745835, "grad_norm": 2.8580326200742396, "learning_rate": 1.6133698512048019e-06, "loss": 0.439, "step": 10505 }, { "epoch": 0.7448422545196739, "grad_norm": 3.2275884356422395, "learning_rate": 1.6125252594661494e-06, "loss": 0.431, "step": 10506 }, { "epoch": 0.7449131513647642, "grad_norm": 9.424054185350213, "learning_rate": 1.6116808463516243e-06, "loss": 0.3858, "step": 10507 }, { "epoch": 0.7449840482098546, "grad_norm": 2.5916181449440914, "learning_rate": 1.610836611905756e-06, "loss": 0.3785, "step": 10508 }, { "epoch": 0.7450549450549451, "grad_norm": 4.445221738952391, "learning_rate": 1.6099925561730583e-06, "loss": 0.3992, "step": 10509 }, { "epoch": 0.7451258419000355, "grad_norm": 1.1230886941212281, "learning_rate": 1.609148679198042e-06, "loss": 0.6305, "step": 10510 }, { "epoch": 0.7451967387451258, "grad_norm": 3.500357065943704, "learning_rate": 1.6083049810252038e-06, "loss": 0.3939, "step": 10511 }, { "epoch": 0.7452676355902162, "grad_norm": 3.262609895130327, "learning_rate": 1.6074614616990314e-06, "loss": 0.3848, "step": 10512 }, { "epoch": 0.7453385324353067, "grad_norm": 2.32357500952149, "learning_rate": 1.6066181212640074e-06, "loss": 0.3983, "step": 10513 }, { "epoch": 0.745409429280397, "grad_norm": 4.097275870773392, "learning_rate": 1.6057749597645976e-06, "loss": 0.3115, "step": 10514 }, { "epoch": 0.7454803261254874, "grad_norm": 8.174854205603138, "learning_rate": 1.604931977245267e-06, "loss": 0.4453, "step": 10515 }, { "epoch": 0.7455512229705779, "grad_norm": 12.828129418605487, "learning_rate": 1.6040891737504628e-06, "loss": 0.394, "step": 10516 }, { "epoch": 0.7456221198156682, "grad_norm": 5.018282472948102, "learning_rate": 1.6032465493246302e-06, "loss": 0.3582, "step": 10517 }, { "epoch": 0.7456930166607586, "grad_norm": 7.7822234909160795, "learning_rate": 1.6024041040121996e-06, "loss": 0.3719, "step": 10518 }, { "epoch": 0.745763913505849, "grad_norm": 2.993617660520728, "learning_rate": 1.6015618378575925e-06, "loss": 0.3605, "step": 10519 }, { "epoch": 0.7458348103509393, "grad_norm": 6.997705641193549, "learning_rate": 1.600719750905223e-06, "loss": 0.4225, "step": 10520 }, { "epoch": 0.7459057071960298, "grad_norm": 5.54213528919501, "learning_rate": 1.5998778431994992e-06, "loss": 0.444, "step": 10521 }, { "epoch": 0.7459766040411202, "grad_norm": 2.959470643225685, "learning_rate": 1.5990361147848088e-06, "loss": 0.3548, "step": 10522 }, { "epoch": 0.7460475008862105, "grad_norm": 2.6062043955932155, "learning_rate": 1.5981945657055398e-06, "loss": 0.4032, "step": 10523 }, { "epoch": 0.746118397731301, "grad_norm": 3.641084029151408, "learning_rate": 1.5973531960060696e-06, "loss": 0.3902, "step": 10524 }, { "epoch": 0.7461892945763914, "grad_norm": 2.6019521102548353, "learning_rate": 1.5965120057307621e-06, "loss": 0.408, "step": 10525 }, { "epoch": 0.7462601914214817, "grad_norm": 2.4876948290760215, "learning_rate": 1.595670994923973e-06, "loss": 0.361, "step": 10526 }, { "epoch": 0.7463310882665721, "grad_norm": 2.7062351228545194, "learning_rate": 1.5948301636300506e-06, "loss": 0.4032, "step": 10527 }, { "epoch": 0.7464019851116626, "grad_norm": 2.6396563112619753, "learning_rate": 1.593989511893334e-06, "loss": 0.3962, "step": 10528 }, { "epoch": 0.7464728819567529, "grad_norm": 2.121282727005269, "learning_rate": 1.5931490397581506e-06, "loss": 0.3466, "step": 10529 }, { "epoch": 0.7465437788018433, "grad_norm": 1.841146694899664, "learning_rate": 1.592308747268817e-06, "loss": 0.4166, "step": 10530 }, { "epoch": 0.7466146756469337, "grad_norm": 3.739780309413025, "learning_rate": 1.5914686344696455e-06, "loss": 0.4322, "step": 10531 }, { "epoch": 0.746685572492024, "grad_norm": 3.647195077567298, "learning_rate": 1.5906287014049348e-06, "loss": 0.3625, "step": 10532 }, { "epoch": 0.7467564693371145, "grad_norm": 3.1794982978720943, "learning_rate": 1.5897889481189731e-06, "loss": 0.3967, "step": 10533 }, { "epoch": 0.7468273661822049, "grad_norm": 3.131650237657732, "learning_rate": 1.5889493746560447e-06, "loss": 0.4215, "step": 10534 }, { "epoch": 0.7468982630272953, "grad_norm": 0.8987443711055783, "learning_rate": 1.5881099810604178e-06, "loss": 0.6313, "step": 10535 }, { "epoch": 0.7469691598723857, "grad_norm": 4.891182772294153, "learning_rate": 1.5872707673763577e-06, "loss": 0.3352, "step": 10536 }, { "epoch": 0.7470400567174761, "grad_norm": 3.9529712854716235, "learning_rate": 1.5864317336481144e-06, "loss": 0.4622, "step": 10537 }, { "epoch": 0.7471109535625665, "grad_norm": 3.1025695202318078, "learning_rate": 1.5855928799199299e-06, "loss": 0.3999, "step": 10538 }, { "epoch": 0.7471818504076568, "grad_norm": 3.635460322528092, "learning_rate": 1.5847542062360388e-06, "loss": 0.4177, "step": 10539 }, { "epoch": 0.7472527472527473, "grad_norm": 2.884789173360861, "learning_rate": 1.5839157126406684e-06, "loss": 0.4108, "step": 10540 }, { "epoch": 0.7473236440978377, "grad_norm": 2.4399104585540425, "learning_rate": 1.5830773991780262e-06, "loss": 0.3661, "step": 10541 }, { "epoch": 0.747394540942928, "grad_norm": 3.2996759844353885, "learning_rate": 1.5822392658923214e-06, "loss": 0.4282, "step": 10542 }, { "epoch": 0.7474654377880184, "grad_norm": 3.4180842340798665, "learning_rate": 1.58140131282775e-06, "loss": 0.3758, "step": 10543 }, { "epoch": 0.7475363346331089, "grad_norm": 2.6377557969828613, "learning_rate": 1.580563540028497e-06, "loss": 0.4189, "step": 10544 }, { "epoch": 0.7476072314781992, "grad_norm": 2.9478124574982214, "learning_rate": 1.5797259475387366e-06, "loss": 0.3946, "step": 10545 }, { "epoch": 0.7476781283232896, "grad_norm": 3.164098368923599, "learning_rate": 1.5788885354026374e-06, "loss": 0.4148, "step": 10546 }, { "epoch": 0.74774902516838, "grad_norm": 2.8069888936040286, "learning_rate": 1.5780513036643608e-06, "loss": 0.4139, "step": 10547 }, { "epoch": 0.7478199220134704, "grad_norm": 2.486726131355444, "learning_rate": 1.5772142523680468e-06, "loss": 0.3536, "step": 10548 }, { "epoch": 0.7478908188585608, "grad_norm": 3.5400597419758775, "learning_rate": 1.5763773815578382e-06, "loss": 0.3852, "step": 10549 }, { "epoch": 0.7479617157036512, "grad_norm": 3.3696753149250798, "learning_rate": 1.5755406912778655e-06, "loss": 0.342, "step": 10550 }, { "epoch": 0.7480326125487415, "grad_norm": 3.1762359976271224, "learning_rate": 1.5747041815722453e-06, "loss": 0.3989, "step": 10551 }, { "epoch": 0.748103509393832, "grad_norm": 2.8635372418210925, "learning_rate": 1.5738678524850864e-06, "loss": 0.4418, "step": 10552 }, { "epoch": 0.7481744062389224, "grad_norm": 2.095429566269673, "learning_rate": 1.573031704060492e-06, "loss": 0.4017, "step": 10553 }, { "epoch": 0.7482453030840127, "grad_norm": 2.669325741039801, "learning_rate": 1.5721957363425504e-06, "loss": 0.4006, "step": 10554 }, { "epoch": 0.7483161999291031, "grad_norm": 3.8749532884306683, "learning_rate": 1.5713599493753457e-06, "loss": 0.3989, "step": 10555 }, { "epoch": 0.7483870967741936, "grad_norm": 2.1508077169473463, "learning_rate": 1.5705243432029478e-06, "loss": 0.372, "step": 10556 }, { "epoch": 0.7484579936192839, "grad_norm": 2.534897880739306, "learning_rate": 1.569688917869417e-06, "loss": 0.4176, "step": 10557 }, { "epoch": 0.7485288904643743, "grad_norm": 3.1588138664833227, "learning_rate": 1.5688536734188092e-06, "loss": 0.3811, "step": 10558 }, { "epoch": 0.7485997873094647, "grad_norm": 3.2354145051013212, "learning_rate": 1.5680186098951649e-06, "loss": 0.3956, "step": 10559 }, { "epoch": 0.7486706841545552, "grad_norm": 2.644422339923547, "learning_rate": 1.56718372734252e-06, "loss": 0.362, "step": 10560 }, { "epoch": 0.7487415809996455, "grad_norm": 3.187953469904196, "learning_rate": 1.5663490258048958e-06, "loss": 0.3883, "step": 10561 }, { "epoch": 0.7488124778447359, "grad_norm": 2.24459887941657, "learning_rate": 1.5655145053263104e-06, "loss": 0.4375, "step": 10562 }, { "epoch": 0.7488833746898264, "grad_norm": 3.729256929237694, "learning_rate": 1.564680165950766e-06, "loss": 0.4401, "step": 10563 }, { "epoch": 0.7489542715349167, "grad_norm": 2.1796551449489385, "learning_rate": 1.5638460077222573e-06, "loss": 0.3832, "step": 10564 }, { "epoch": 0.7490251683800071, "grad_norm": 1.9931196179677066, "learning_rate": 1.5630120306847708e-06, "loss": 0.3411, "step": 10565 }, { "epoch": 0.7490960652250975, "grad_norm": 4.338191060107026, "learning_rate": 1.5621782348822868e-06, "loss": 0.4181, "step": 10566 }, { "epoch": 0.7491669620701878, "grad_norm": 4.467360194136639, "learning_rate": 1.561344620358765e-06, "loss": 0.3791, "step": 10567 }, { "epoch": 0.7492378589152783, "grad_norm": 2.85698453737401, "learning_rate": 1.5605111871581662e-06, "loss": 0.4198, "step": 10568 }, { "epoch": 0.7493087557603687, "grad_norm": 3.006539968118666, "learning_rate": 1.5596779353244396e-06, "loss": 0.382, "step": 10569 }, { "epoch": 0.749379652605459, "grad_norm": 1.577623272694338, "learning_rate": 1.5588448649015208e-06, "loss": 0.3709, "step": 10570 }, { "epoch": 0.7494505494505495, "grad_norm": 2.8996722378649973, "learning_rate": 1.5580119759333379e-06, "loss": 0.3534, "step": 10571 }, { "epoch": 0.7495214462956399, "grad_norm": 2.5860078880103132, "learning_rate": 1.55717926846381e-06, "loss": 0.4202, "step": 10572 }, { "epoch": 0.7495923431407302, "grad_norm": 3.613963146100154, "learning_rate": 1.5563467425368488e-06, "loss": 0.3908, "step": 10573 }, { "epoch": 0.7496632399858206, "grad_norm": 2.583243384090236, "learning_rate": 1.5555143981963527e-06, "loss": 0.4337, "step": 10574 }, { "epoch": 0.7497341368309111, "grad_norm": 14.329111665256313, "learning_rate": 1.5546822354862096e-06, "loss": 0.3714, "step": 10575 }, { "epoch": 0.7498050336760014, "grad_norm": 23.443417084567443, "learning_rate": 1.553850254450303e-06, "loss": 0.422, "step": 10576 }, { "epoch": 0.7498759305210918, "grad_norm": 4.815660300229752, "learning_rate": 1.5530184551325034e-06, "loss": 0.4165, "step": 10577 }, { "epoch": 0.7499468273661822, "grad_norm": 2.1068525344664724, "learning_rate": 1.5521868375766697e-06, "loss": 0.3611, "step": 10578 }, { "epoch": 0.7500177242112726, "grad_norm": 4.784775405497615, "learning_rate": 1.5513554018266576e-06, "loss": 0.3672, "step": 10579 }, { "epoch": 0.750088621056363, "grad_norm": 1.9549881240531168, "learning_rate": 1.5505241479263055e-06, "loss": 0.39, "step": 10580 }, { "epoch": 0.7501595179014534, "grad_norm": 2.7079215851350584, "learning_rate": 1.5496930759194495e-06, "loss": 0.4383, "step": 10581 }, { "epoch": 0.7502304147465437, "grad_norm": 10.172549799412279, "learning_rate": 1.5488621858499114e-06, "loss": 0.4325, "step": 10582 }, { "epoch": 0.7503013115916342, "grad_norm": 3.0941272951302357, "learning_rate": 1.5480314777615025e-06, "loss": 0.3567, "step": 10583 }, { "epoch": 0.7503722084367246, "grad_norm": 0.9922065304166248, "learning_rate": 1.5472009516980286e-06, "loss": 0.6347, "step": 10584 }, { "epoch": 0.7504431052818149, "grad_norm": 21.944419976015368, "learning_rate": 1.5463706077032876e-06, "loss": 0.4006, "step": 10585 }, { "epoch": 0.7505140021269053, "grad_norm": 2.4112078150257483, "learning_rate": 1.5455404458210565e-06, "loss": 0.3639, "step": 10586 }, { "epoch": 0.7505848989719958, "grad_norm": 2.768664956642565, "learning_rate": 1.544710466095115e-06, "loss": 0.3677, "step": 10587 }, { "epoch": 0.7506557958170862, "grad_norm": 5.036681365752753, "learning_rate": 1.5438806685692293e-06, "loss": 0.4337, "step": 10588 }, { "epoch": 0.7507266926621765, "grad_norm": 4.420391598427122, "learning_rate": 1.5430510532871534e-06, "loss": 0.4378, "step": 10589 }, { "epoch": 0.7507975895072669, "grad_norm": 2.8572708046784547, "learning_rate": 1.5422216202926333e-06, "loss": 0.3442, "step": 10590 }, { "epoch": 0.7508684863523574, "grad_norm": 2.510263620960419, "learning_rate": 1.5413923696294058e-06, "loss": 0.3831, "step": 10591 }, { "epoch": 0.7509393831974477, "grad_norm": 2.036959948110912, "learning_rate": 1.5405633013412002e-06, "loss": 0.4108, "step": 10592 }, { "epoch": 0.7510102800425381, "grad_norm": 4.429081668882502, "learning_rate": 1.539734415471732e-06, "loss": 0.3768, "step": 10593 }, { "epoch": 0.7510811768876285, "grad_norm": 21.27469715696904, "learning_rate": 1.5389057120647082e-06, "loss": 0.3592, "step": 10594 }, { "epoch": 0.7511520737327189, "grad_norm": 1.7808968945851007, "learning_rate": 1.5380771911638288e-06, "loss": 0.3622, "step": 10595 }, { "epoch": 0.7512229705778093, "grad_norm": 3.141487195596528, "learning_rate": 1.5372488528127822e-06, "loss": 0.4207, "step": 10596 }, { "epoch": 0.7512938674228997, "grad_norm": 3.050414011658805, "learning_rate": 1.5364206970552442e-06, "loss": 0.3768, "step": 10597 }, { "epoch": 0.75136476426799, "grad_norm": 2.7258787527174375, "learning_rate": 1.5355927239348884e-06, "loss": 0.3093, "step": 10598 }, { "epoch": 0.7514356611130805, "grad_norm": 24.886518337406724, "learning_rate": 1.5347649334953712e-06, "loss": 0.3553, "step": 10599 }, { "epoch": 0.7515065579581709, "grad_norm": 2.1554581481374178, "learning_rate": 1.5339373257803452e-06, "loss": 0.41, "step": 10600 }, { "epoch": 0.7515774548032612, "grad_norm": 2.042471451873637, "learning_rate": 1.5331099008334483e-06, "loss": 0.3863, "step": 10601 }, { "epoch": 0.7516483516483516, "grad_norm": 4.250861747185829, "learning_rate": 1.5322826586983136e-06, "loss": 0.4413, "step": 10602 }, { "epoch": 0.7517192484934421, "grad_norm": 3.6652886618433795, "learning_rate": 1.5314555994185614e-06, "loss": 0.3674, "step": 10603 }, { "epoch": 0.7517901453385324, "grad_norm": 2.4032543145191614, "learning_rate": 1.5306287230378015e-06, "loss": 0.3498, "step": 10604 }, { "epoch": 0.7518610421836228, "grad_norm": 3.7585563062333605, "learning_rate": 1.529802029599638e-06, "loss": 0.4319, "step": 10605 }, { "epoch": 0.7519319390287132, "grad_norm": 2.2657450519548417, "learning_rate": 1.5289755191476613e-06, "loss": 0.4112, "step": 10606 }, { "epoch": 0.7520028358738036, "grad_norm": 2.176331565765014, "learning_rate": 1.528149191725456e-06, "loss": 0.3801, "step": 10607 }, { "epoch": 0.752073732718894, "grad_norm": 3.305726120355579, "learning_rate": 1.5273230473765938e-06, "loss": 0.4102, "step": 10608 }, { "epoch": 0.7521446295639844, "grad_norm": 9.758397481523637, "learning_rate": 1.5264970861446365e-06, "loss": 0.3739, "step": 10609 }, { "epoch": 0.7522155264090747, "grad_norm": 1.929491471853171, "learning_rate": 1.525671308073139e-06, "loss": 0.3432, "step": 10610 }, { "epoch": 0.7522864232541652, "grad_norm": 2.2856907961854422, "learning_rate": 1.5248457132056476e-06, "loss": 0.3465, "step": 10611 }, { "epoch": 0.7523573200992556, "grad_norm": 2.0917417291085734, "learning_rate": 1.524020301585692e-06, "loss": 0.3841, "step": 10612 }, { "epoch": 0.752428216944346, "grad_norm": 2.0338585161245732, "learning_rate": 1.5231950732567991e-06, "loss": 0.3492, "step": 10613 }, { "epoch": 0.7524991137894363, "grad_norm": 9.80822181117419, "learning_rate": 1.5223700282624849e-06, "loss": 0.3615, "step": 10614 }, { "epoch": 0.7525700106345268, "grad_norm": 3.7280972452180596, "learning_rate": 1.5215451666462532e-06, "loss": 0.3375, "step": 10615 }, { "epoch": 0.7526409074796172, "grad_norm": 2.1272337628070206, "learning_rate": 1.5207204884515987e-06, "loss": 0.3748, "step": 10616 }, { "epoch": 0.7527118043247075, "grad_norm": 2.112340138563736, "learning_rate": 1.5198959937220082e-06, "loss": 0.3346, "step": 10617 }, { "epoch": 0.752782701169798, "grad_norm": 11.662098465649663, "learning_rate": 1.51907168250096e-06, "loss": 0.3559, "step": 10618 }, { "epoch": 0.7528535980148884, "grad_norm": 2.2672054572415252, "learning_rate": 1.5182475548319188e-06, "loss": 0.3999, "step": 10619 }, { "epoch": 0.7529244948599787, "grad_norm": 2.6250152850472963, "learning_rate": 1.5174236107583394e-06, "loss": 0.3851, "step": 10620 }, { "epoch": 0.7529953917050691, "grad_norm": 4.098127281491715, "learning_rate": 1.5165998503236734e-06, "loss": 0.4161, "step": 10621 }, { "epoch": 0.7530662885501596, "grad_norm": 4.6268165080005526, "learning_rate": 1.515776273571356e-06, "loss": 0.4713, "step": 10622 }, { "epoch": 0.7531371853952499, "grad_norm": 3.420884685566047, "learning_rate": 1.5149528805448134e-06, "loss": 0.3896, "step": 10623 }, { "epoch": 0.7532080822403403, "grad_norm": 1.8474962069923344, "learning_rate": 1.514129671287467e-06, "loss": 0.4115, "step": 10624 }, { "epoch": 0.7532789790854307, "grad_norm": 3.2866222152923252, "learning_rate": 1.513306645842722e-06, "loss": 0.4236, "step": 10625 }, { "epoch": 0.753349875930521, "grad_norm": 2.7333208386475505, "learning_rate": 1.5124838042539803e-06, "loss": 0.3653, "step": 10626 }, { "epoch": 0.7534207727756115, "grad_norm": 2.725150464592404, "learning_rate": 1.5116611465646298e-06, "loss": 0.4032, "step": 10627 }, { "epoch": 0.7534916696207019, "grad_norm": 3.2920885802043944, "learning_rate": 1.510838672818048e-06, "loss": 0.4312, "step": 10628 }, { "epoch": 0.7535625664657922, "grad_norm": 4.275297397757005, "learning_rate": 1.5100163830576058e-06, "loss": 0.3945, "step": 10629 }, { "epoch": 0.7536334633108827, "grad_norm": 3.3420431831708974, "learning_rate": 1.5091942773266654e-06, "loss": 0.4298, "step": 10630 }, { "epoch": 0.7537043601559731, "grad_norm": 3.516653517937034, "learning_rate": 1.5083723556685748e-06, "loss": 0.3624, "step": 10631 }, { "epoch": 0.7537752570010634, "grad_norm": 3.6809248893447055, "learning_rate": 1.5075506181266736e-06, "loss": 0.4336, "step": 10632 }, { "epoch": 0.7538461538461538, "grad_norm": 2.7087690330671386, "learning_rate": 1.5067290647442956e-06, "loss": 0.3696, "step": 10633 }, { "epoch": 0.7539170506912443, "grad_norm": 4.1741948342566095, "learning_rate": 1.5059076955647606e-06, "loss": 0.3833, "step": 10634 }, { "epoch": 0.7539879475363346, "grad_norm": 1.9957746851051554, "learning_rate": 1.5050865106313777e-06, "loss": 0.3705, "step": 10635 }, { "epoch": 0.754058844381425, "grad_norm": 2.294437243939212, "learning_rate": 1.5042655099874503e-06, "loss": 0.3895, "step": 10636 }, { "epoch": 0.7541297412265154, "grad_norm": 3.6794209450845567, "learning_rate": 1.5034446936762731e-06, "loss": 0.4607, "step": 10637 }, { "epoch": 0.7542006380716059, "grad_norm": 3.4092920553210484, "learning_rate": 1.5026240617411258e-06, "loss": 0.3898, "step": 10638 }, { "epoch": 0.7542715349166962, "grad_norm": 10.160766065061084, "learning_rate": 1.5018036142252796e-06, "loss": 0.3575, "step": 10639 }, { "epoch": 0.7543424317617866, "grad_norm": 3.2325591829757503, "learning_rate": 1.5009833511720002e-06, "loss": 0.3434, "step": 10640 }, { "epoch": 0.754413328606877, "grad_norm": 2.6759686988706037, "learning_rate": 1.500163272624539e-06, "loss": 0.3534, "step": 10641 }, { "epoch": 0.7544842254519674, "grad_norm": 2.6346692206275413, "learning_rate": 1.499343378626139e-06, "loss": 0.3722, "step": 10642 }, { "epoch": 0.7545551222970578, "grad_norm": 3.2319565359316393, "learning_rate": 1.4985236692200356e-06, "loss": 0.382, "step": 10643 }, { "epoch": 0.7546260191421482, "grad_norm": 3.2438989509648457, "learning_rate": 1.4977041444494506e-06, "loss": 0.3923, "step": 10644 }, { "epoch": 0.7546969159872385, "grad_norm": 3.2559265586270976, "learning_rate": 1.4968848043576001e-06, "loss": 0.3746, "step": 10645 }, { "epoch": 0.754767812832329, "grad_norm": 2.741593973057339, "learning_rate": 1.4960656489876869e-06, "loss": 0.3911, "step": 10646 }, { "epoch": 0.7548387096774194, "grad_norm": 6.784611794329825, "learning_rate": 1.4952466783829072e-06, "loss": 0.3482, "step": 10647 }, { "epoch": 0.7549096065225097, "grad_norm": 2.186820991702669, "learning_rate": 1.4944278925864441e-06, "loss": 0.3776, "step": 10648 }, { "epoch": 0.7549805033676001, "grad_norm": 2.2598695038690844, "learning_rate": 1.4936092916414757e-06, "loss": 0.378, "step": 10649 }, { "epoch": 0.7550514002126906, "grad_norm": 5.545126927228335, "learning_rate": 1.4927908755911652e-06, "loss": 0.3704, "step": 10650 }, { "epoch": 0.7551222970577809, "grad_norm": 6.979046091043564, "learning_rate": 1.4919726444786676e-06, "loss": 0.4043, "step": 10651 }, { "epoch": 0.7551931939028713, "grad_norm": 4.464625706362203, "learning_rate": 1.491154598347131e-06, "loss": 0.4605, "step": 10652 }, { "epoch": 0.7552640907479617, "grad_norm": 2.623815082034839, "learning_rate": 1.490336737239691e-06, "loss": 0.343, "step": 10653 }, { "epoch": 0.7553349875930521, "grad_norm": 3.113007485445459, "learning_rate": 1.489519061199472e-06, "loss": 0.4461, "step": 10654 }, { "epoch": 0.7554058844381425, "grad_norm": 2.5900366377923696, "learning_rate": 1.4887015702695928e-06, "loss": 0.3574, "step": 10655 }, { "epoch": 0.7554767812832329, "grad_norm": 3.238287552630849, "learning_rate": 1.4878842644931623e-06, "loss": 0.4145, "step": 10656 }, { "epoch": 0.7555476781283232, "grad_norm": 2.541698963073196, "learning_rate": 1.4870671439132722e-06, "loss": 0.3596, "step": 10657 }, { "epoch": 0.7556185749734137, "grad_norm": 2.6984238006484147, "learning_rate": 1.4862502085730134e-06, "loss": 0.3836, "step": 10658 }, { "epoch": 0.7556894718185041, "grad_norm": 2.328241013973121, "learning_rate": 1.485433458515464e-06, "loss": 0.3432, "step": 10659 }, { "epoch": 0.7557603686635944, "grad_norm": 2.82413323403893, "learning_rate": 1.4846168937836913e-06, "loss": 0.4075, "step": 10660 }, { "epoch": 0.7558312655086848, "grad_norm": 3.517667442074206, "learning_rate": 1.4838005144207512e-06, "loss": 0.4335, "step": 10661 }, { "epoch": 0.7559021623537753, "grad_norm": 2.033180335101841, "learning_rate": 1.4829843204696936e-06, "loss": 0.4052, "step": 10662 }, { "epoch": 0.7559730591988657, "grad_norm": 2.2728532006398505, "learning_rate": 1.4821683119735592e-06, "loss": 0.3708, "step": 10663 }, { "epoch": 0.756043956043956, "grad_norm": 1.8682224512301768, "learning_rate": 1.4813524889753745e-06, "loss": 0.3854, "step": 10664 }, { "epoch": 0.7561148528890465, "grad_norm": 3.4211555198410446, "learning_rate": 1.4805368515181574e-06, "loss": 0.3552, "step": 10665 }, { "epoch": 0.7561857497341369, "grad_norm": 3.540152792929924, "learning_rate": 1.4797213996449195e-06, "loss": 0.38, "step": 10666 }, { "epoch": 0.7562566465792272, "grad_norm": 2.667383543032251, "learning_rate": 1.4789061333986593e-06, "loss": 0.3883, "step": 10667 }, { "epoch": 0.7563275434243176, "grad_norm": 3.731319222397985, "learning_rate": 1.4780910528223646e-06, "loss": 0.3775, "step": 10668 }, { "epoch": 0.7563984402694081, "grad_norm": 1.9208352508622777, "learning_rate": 1.4772761579590184e-06, "loss": 0.3366, "step": 10669 }, { "epoch": 0.7564693371144984, "grad_norm": 2.7141120994750385, "learning_rate": 1.4764614488515877e-06, "loss": 0.365, "step": 10670 }, { "epoch": 0.7565402339595888, "grad_norm": 16.27172444987192, "learning_rate": 1.4756469255430355e-06, "loss": 0.3525, "step": 10671 }, { "epoch": 0.7566111308046792, "grad_norm": 2.9688883097764474, "learning_rate": 1.4748325880763114e-06, "loss": 0.3937, "step": 10672 }, { "epoch": 0.7566820276497696, "grad_norm": 2.9006254826866154, "learning_rate": 1.4740184364943538e-06, "loss": 0.4302, "step": 10673 }, { "epoch": 0.75675292449486, "grad_norm": 2.3785937514706577, "learning_rate": 1.4732044708400956e-06, "loss": 0.389, "step": 10674 }, { "epoch": 0.7568238213399504, "grad_norm": 5.44133388715653, "learning_rate": 1.4723906911564583e-06, "loss": 0.4424, "step": 10675 }, { "epoch": 0.7568947181850407, "grad_norm": 2.6183756386158383, "learning_rate": 1.471577097486353e-06, "loss": 0.368, "step": 10676 }, { "epoch": 0.7569656150301312, "grad_norm": 3.1621398612109886, "learning_rate": 1.470763689872679e-06, "loss": 0.3619, "step": 10677 }, { "epoch": 0.7570365118752216, "grad_norm": 2.381999572045265, "learning_rate": 1.4699504683583305e-06, "loss": 0.3825, "step": 10678 }, { "epoch": 0.7571074087203119, "grad_norm": 3.3207888851046463, "learning_rate": 1.4691374329861885e-06, "loss": 0.3974, "step": 10679 }, { "epoch": 0.7571783055654023, "grad_norm": 12.823195560965246, "learning_rate": 1.4683245837991228e-06, "loss": 0.4228, "step": 10680 }, { "epoch": 0.7572492024104928, "grad_norm": 2.6433824674404014, "learning_rate": 1.4675119208399975e-06, "loss": 0.3695, "step": 10681 }, { "epoch": 0.7573200992555831, "grad_norm": 4.459860933089323, "learning_rate": 1.4666994441516663e-06, "loss": 0.3983, "step": 10682 }, { "epoch": 0.7573909961006735, "grad_norm": 3.831623951555148, "learning_rate": 1.4658871537769703e-06, "loss": 0.4086, "step": 10683 }, { "epoch": 0.7574618929457639, "grad_norm": 1.9937011388151178, "learning_rate": 1.4650750497587407e-06, "loss": 0.4013, "step": 10684 }, { "epoch": 0.7575327897908543, "grad_norm": 4.509979791835468, "learning_rate": 1.4642631321398038e-06, "loss": 0.4177, "step": 10685 }, { "epoch": 0.7576036866359447, "grad_norm": 10.65301950126395, "learning_rate": 1.4634514009629697e-06, "loss": 0.4228, "step": 10686 }, { "epoch": 0.7576745834810351, "grad_norm": 5.988732960855769, "learning_rate": 1.4626398562710414e-06, "loss": 0.4351, "step": 10687 }, { "epoch": 0.7577454803261255, "grad_norm": 2.578586835268186, "learning_rate": 1.4618284981068155e-06, "loss": 0.3799, "step": 10688 }, { "epoch": 0.7578163771712159, "grad_norm": 1.740159631741305, "learning_rate": 1.4610173265130712e-06, "loss": 0.4005, "step": 10689 }, { "epoch": 0.7578872740163063, "grad_norm": 1.9445439565349703, "learning_rate": 1.4602063415325862e-06, "loss": 0.3522, "step": 10690 }, { "epoch": 0.7579581708613967, "grad_norm": 2.9063868616197692, "learning_rate": 1.4593955432081215e-06, "loss": 0.3989, "step": 10691 }, { "epoch": 0.758029067706487, "grad_norm": 2.404038543441837, "learning_rate": 1.4585849315824336e-06, "loss": 0.3712, "step": 10692 }, { "epoch": 0.7580999645515775, "grad_norm": 2.8266541046093274, "learning_rate": 1.4577745066982636e-06, "loss": 0.374, "step": 10693 }, { "epoch": 0.7581708613966679, "grad_norm": 3.182377880115738, "learning_rate": 1.4569642685983487e-06, "loss": 0.3841, "step": 10694 }, { "epoch": 0.7582417582417582, "grad_norm": 4.2667401896269475, "learning_rate": 1.4561542173254122e-06, "loss": 0.4141, "step": 10695 }, { "epoch": 0.7583126550868486, "grad_norm": 5.0485728445032425, "learning_rate": 1.4553443529221678e-06, "loss": 0.3812, "step": 10696 }, { "epoch": 0.7583835519319391, "grad_norm": 7.2115187043425015, "learning_rate": 1.4545346754313222e-06, "loss": 0.3985, "step": 10697 }, { "epoch": 0.7584544487770294, "grad_norm": 4.141979396417947, "learning_rate": 1.4537251848955697e-06, "loss": 0.3789, "step": 10698 }, { "epoch": 0.7585253456221198, "grad_norm": 3.2445270262776256, "learning_rate": 1.4529158813575934e-06, "loss": 0.3787, "step": 10699 }, { "epoch": 0.7585962424672102, "grad_norm": 2.2387423103503266, "learning_rate": 1.4521067648600694e-06, "loss": 0.3701, "step": 10700 }, { "epoch": 0.7586671393123006, "grad_norm": 3.019639075446063, "learning_rate": 1.4512978354456658e-06, "loss": 0.4372, "step": 10701 }, { "epoch": 0.758738036157391, "grad_norm": 2.9205080464140085, "learning_rate": 1.450489093157036e-06, "loss": 0.3572, "step": 10702 }, { "epoch": 0.7588089330024814, "grad_norm": 1.6047229574245372, "learning_rate": 1.449680538036824e-06, "loss": 0.3767, "step": 10703 }, { "epoch": 0.7588798298475717, "grad_norm": 2.690180303257766, "learning_rate": 1.4488721701276664e-06, "loss": 0.3921, "step": 10704 }, { "epoch": 0.7589507266926622, "grad_norm": 4.065002461050032, "learning_rate": 1.4480639894721936e-06, "loss": 0.4379, "step": 10705 }, { "epoch": 0.7590216235377526, "grad_norm": 3.361411507020213, "learning_rate": 1.4472559961130145e-06, "loss": 0.4085, "step": 10706 }, { "epoch": 0.7590925203828429, "grad_norm": 2.223395220251396, "learning_rate": 1.446448190092738e-06, "loss": 0.4304, "step": 10707 }, { "epoch": 0.7591634172279333, "grad_norm": 3.5032734267777035, "learning_rate": 1.4456405714539623e-06, "loss": 0.4271, "step": 10708 }, { "epoch": 0.7592343140730238, "grad_norm": 3.8445446426996077, "learning_rate": 1.4448331402392729e-06, "loss": 0.4214, "step": 10709 }, { "epoch": 0.7593052109181141, "grad_norm": 2.091634217524893, "learning_rate": 1.4440258964912435e-06, "loss": 0.3267, "step": 10710 }, { "epoch": 0.7593761077632045, "grad_norm": 3.5623150329218296, "learning_rate": 1.443218840252444e-06, "loss": 0.4377, "step": 10711 }, { "epoch": 0.759447004608295, "grad_norm": 1.7335987966161353, "learning_rate": 1.4424119715654284e-06, "loss": 0.3974, "step": 10712 }, { "epoch": 0.7595179014533853, "grad_norm": 6.118680707668369, "learning_rate": 1.4416052904727467e-06, "loss": 0.3805, "step": 10713 }, { "epoch": 0.7595887982984757, "grad_norm": 2.4936702448370367, "learning_rate": 1.4407987970169341e-06, "loss": 0.3383, "step": 10714 }, { "epoch": 0.7596596951435661, "grad_norm": 2.362128543160505, "learning_rate": 1.4399924912405156e-06, "loss": 0.3805, "step": 10715 }, { "epoch": 0.7597305919886566, "grad_norm": 0.9556523705396829, "learning_rate": 1.4391863731860124e-06, "loss": 0.6175, "step": 10716 }, { "epoch": 0.7598014888337469, "grad_norm": 2.2973620044818928, "learning_rate": 1.4383804428959293e-06, "loss": 0.3478, "step": 10717 }, { "epoch": 0.7598723856788373, "grad_norm": 9.760583502350892, "learning_rate": 1.437574700412762e-06, "loss": 0.3882, "step": 10718 }, { "epoch": 0.7599432825239277, "grad_norm": 2.0138415883618577, "learning_rate": 1.436769145779e-06, "loss": 0.39, "step": 10719 }, { "epoch": 0.760014179369018, "grad_norm": 6.6475895522778075, "learning_rate": 1.4359637790371222e-06, "loss": 0.3949, "step": 10720 }, { "epoch": 0.7600850762141085, "grad_norm": 2.597511533239825, "learning_rate": 1.4351586002295953e-06, "loss": 0.3863, "step": 10721 }, { "epoch": 0.7601559730591989, "grad_norm": 4.293347879920355, "learning_rate": 1.4343536093988741e-06, "loss": 0.3792, "step": 10722 }, { "epoch": 0.7602268699042892, "grad_norm": 3.6747592775622646, "learning_rate": 1.4335488065874103e-06, "loss": 0.4383, "step": 10723 }, { "epoch": 0.7602977667493797, "grad_norm": 2.0886401327195707, "learning_rate": 1.43274419183764e-06, "loss": 0.4, "step": 10724 }, { "epoch": 0.7603686635944701, "grad_norm": 5.807157848331524, "learning_rate": 1.4319397651919897e-06, "loss": 0.4504, "step": 10725 }, { "epoch": 0.7604395604395604, "grad_norm": 2.4407825710969644, "learning_rate": 1.431135526692879e-06, "loss": 0.4223, "step": 10726 }, { "epoch": 0.7605104572846508, "grad_norm": 4.776340455414632, "learning_rate": 1.4303314763827176e-06, "loss": 0.4694, "step": 10727 }, { "epoch": 0.7605813541297413, "grad_norm": 9.76729624351117, "learning_rate": 1.4295276143039017e-06, "loss": 0.4355, "step": 10728 }, { "epoch": 0.7606522509748316, "grad_norm": 2.392564394899517, "learning_rate": 1.428723940498819e-06, "loss": 0.3501, "step": 10729 }, { "epoch": 0.760723147819922, "grad_norm": 2.2759635426482037, "learning_rate": 1.4279204550098496e-06, "loss": 0.3821, "step": 10730 }, { "epoch": 0.7607940446650124, "grad_norm": 3.164523726137801, "learning_rate": 1.4271171578793614e-06, "loss": 0.4123, "step": 10731 }, { "epoch": 0.7608649415101028, "grad_norm": 2.6872194208997877, "learning_rate": 1.4263140491497112e-06, "loss": 0.3391, "step": 10732 }, { "epoch": 0.7609358383551932, "grad_norm": 5.585976374509451, "learning_rate": 1.4255111288632484e-06, "loss": 0.3282, "step": 10733 }, { "epoch": 0.7610067352002836, "grad_norm": 2.842385436432799, "learning_rate": 1.4247083970623138e-06, "loss": 0.3469, "step": 10734 }, { "epoch": 0.7610776320453739, "grad_norm": 2.5163343501775195, "learning_rate": 1.4239058537892348e-06, "loss": 0.3882, "step": 10735 }, { "epoch": 0.7611485288904644, "grad_norm": 2.340984575144142, "learning_rate": 1.4231034990863273e-06, "loss": 0.3892, "step": 10736 }, { "epoch": 0.7612194257355548, "grad_norm": 2.0870902717987336, "learning_rate": 1.4223013329959046e-06, "loss": 0.3859, "step": 10737 }, { "epoch": 0.7612903225806451, "grad_norm": 6.917880518742827, "learning_rate": 1.4214993555602612e-06, "loss": 0.3711, "step": 10738 }, { "epoch": 0.7613612194257355, "grad_norm": 2.26181881198224, "learning_rate": 1.42069756682169e-06, "loss": 0.4217, "step": 10739 }, { "epoch": 0.761432116270826, "grad_norm": 1.9114539117717209, "learning_rate": 1.4198959668224683e-06, "loss": 0.3473, "step": 10740 }, { "epoch": 0.7615030131159164, "grad_norm": 4.881949454313286, "learning_rate": 1.419094555604863e-06, "loss": 0.3869, "step": 10741 }, { "epoch": 0.7615739099610067, "grad_norm": 1.2531258311023143, "learning_rate": 1.4182933332111364e-06, "loss": 0.7255, "step": 10742 }, { "epoch": 0.7616448068060971, "grad_norm": 2.9705706336211044, "learning_rate": 1.417492299683536e-06, "loss": 0.3548, "step": 10743 }, { "epoch": 0.7617157036511876, "grad_norm": 4.912500931635019, "learning_rate": 1.4166914550643001e-06, "loss": 0.3828, "step": 10744 }, { "epoch": 0.7617866004962779, "grad_norm": 3.4309495035540043, "learning_rate": 1.4158907993956582e-06, "loss": 0.3933, "step": 10745 }, { "epoch": 0.7618574973413683, "grad_norm": 2.1755086899475162, "learning_rate": 1.4150903327198318e-06, "loss": 0.402, "step": 10746 }, { "epoch": 0.7619283941864587, "grad_norm": 2.373345457185438, "learning_rate": 1.414290055079029e-06, "loss": 0.3617, "step": 10747 }, { "epoch": 0.7619992910315491, "grad_norm": 2.691329668079918, "learning_rate": 1.4134899665154461e-06, "loss": 0.3829, "step": 10748 }, { "epoch": 0.7620701878766395, "grad_norm": 11.394924740544333, "learning_rate": 1.4126900670712752e-06, "loss": 0.3696, "step": 10749 }, { "epoch": 0.7621410847217299, "grad_norm": 3.0668625829854133, "learning_rate": 1.4118903567886977e-06, "loss": 0.3621, "step": 10750 }, { "epoch": 0.7622119815668202, "grad_norm": 4.341128417769025, "learning_rate": 1.4110908357098774e-06, "loss": 0.3561, "step": 10751 }, { "epoch": 0.7622828784119107, "grad_norm": 5.540318361674195, "learning_rate": 1.4102915038769771e-06, "loss": 0.4102, "step": 10752 }, { "epoch": 0.7623537752570011, "grad_norm": 3.814197461787872, "learning_rate": 1.409492361332147e-06, "loss": 0.3784, "step": 10753 }, { "epoch": 0.7624246721020914, "grad_norm": 1.8031626417740145, "learning_rate": 1.4086934081175252e-06, "loss": 0.4089, "step": 10754 }, { "epoch": 0.7624955689471818, "grad_norm": 3.4668336883411137, "learning_rate": 1.4078946442752395e-06, "loss": 0.3836, "step": 10755 }, { "epoch": 0.7625664657922723, "grad_norm": 2.9256920056585685, "learning_rate": 1.4070960698474123e-06, "loss": 0.3753, "step": 10756 }, { "epoch": 0.7626373626373626, "grad_norm": 3.7826506953990138, "learning_rate": 1.4062976848761506e-06, "loss": 0.4503, "step": 10757 }, { "epoch": 0.762708259482453, "grad_norm": 2.084101446862321, "learning_rate": 1.4054994894035562e-06, "loss": 0.41, "step": 10758 }, { "epoch": 0.7627791563275435, "grad_norm": 3.6124914849027348, "learning_rate": 1.4047014834717172e-06, "loss": 0.3914, "step": 10759 }, { "epoch": 0.7628500531726338, "grad_norm": 6.656344521333597, "learning_rate": 1.4039036671227112e-06, "loss": 0.4061, "step": 10760 }, { "epoch": 0.7629209500177242, "grad_norm": 2.7987887756326577, "learning_rate": 1.4031060403986119e-06, "loss": 0.384, "step": 10761 }, { "epoch": 0.7629918468628146, "grad_norm": 2.056283888258137, "learning_rate": 1.4023086033414746e-06, "loss": 0.3762, "step": 10762 }, { "epoch": 0.763062743707905, "grad_norm": 2.426860040983483, "learning_rate": 1.4015113559933524e-06, "loss": 0.3739, "step": 10763 }, { "epoch": 0.7631336405529954, "grad_norm": 1.7992490268863255, "learning_rate": 1.4007142983962812e-06, "loss": 0.3826, "step": 10764 }, { "epoch": 0.7632045373980858, "grad_norm": 2.460048884509459, "learning_rate": 1.3999174305922942e-06, "loss": 0.4142, "step": 10765 }, { "epoch": 0.7632754342431762, "grad_norm": 7.528814265760977, "learning_rate": 1.3991207526234086e-06, "loss": 0.38, "step": 10766 }, { "epoch": 0.7633463310882665, "grad_norm": 0.937883883880503, "learning_rate": 1.3983242645316326e-06, "loss": 0.619, "step": 10767 }, { "epoch": 0.763417227933357, "grad_norm": 2.971173045630271, "learning_rate": 1.3975279663589668e-06, "loss": 0.3916, "step": 10768 }, { "epoch": 0.7634881247784474, "grad_norm": 1.6908121522165158, "learning_rate": 1.396731858147405e-06, "loss": 0.3616, "step": 10769 }, { "epoch": 0.7635590216235377, "grad_norm": 2.30966658682689, "learning_rate": 1.3959359399389183e-06, "loss": 0.3599, "step": 10770 }, { "epoch": 0.7636299184686282, "grad_norm": 2.208134702654018, "learning_rate": 1.3951402117754814e-06, "loss": 0.4022, "step": 10771 }, { "epoch": 0.7637008153137186, "grad_norm": 4.07425906266448, "learning_rate": 1.3943446736990534e-06, "loss": 0.3814, "step": 10772 }, { "epoch": 0.7637717121588089, "grad_norm": 7.395266647209873, "learning_rate": 1.393549325751583e-06, "loss": 0.3468, "step": 10773 }, { "epoch": 0.7638426090038993, "grad_norm": 1.9165631534155463, "learning_rate": 1.3927541679750083e-06, "loss": 0.3629, "step": 10774 }, { "epoch": 0.7639135058489898, "grad_norm": 6.0207374341845865, "learning_rate": 1.391959200411261e-06, "loss": 0.3739, "step": 10775 }, { "epoch": 0.7639844026940801, "grad_norm": 3.325369350379283, "learning_rate": 1.391164423102258e-06, "loss": 0.3756, "step": 10776 }, { "epoch": 0.7640552995391705, "grad_norm": 2.0980193593478655, "learning_rate": 1.3903698360899086e-06, "loss": 0.4056, "step": 10777 }, { "epoch": 0.7641261963842609, "grad_norm": 2.7235281045498594, "learning_rate": 1.3895754394161131e-06, "loss": 0.4031, "step": 10778 }, { "epoch": 0.7641970932293513, "grad_norm": 2.8628858899384997, "learning_rate": 1.388781233122762e-06, "loss": 0.4021, "step": 10779 }, { "epoch": 0.7642679900744417, "grad_norm": 3.4260592108267724, "learning_rate": 1.3879872172517323e-06, "loss": 0.3947, "step": 10780 }, { "epoch": 0.7643388869195321, "grad_norm": 3.687573158154552, "learning_rate": 1.3871933918448922e-06, "loss": 0.4086, "step": 10781 }, { "epoch": 0.7644097837646224, "grad_norm": 3.0304961162755926, "learning_rate": 1.386399756944104e-06, "loss": 0.4051, "step": 10782 }, { "epoch": 0.7644806806097129, "grad_norm": 2.6423250775113174, "learning_rate": 1.3856063125912128e-06, "loss": 0.3446, "step": 10783 }, { "epoch": 0.7645515774548033, "grad_norm": 2.0252341335100743, "learning_rate": 1.384813058828061e-06, "loss": 0.386, "step": 10784 }, { "epoch": 0.7646224742998936, "grad_norm": 2.3708118547391264, "learning_rate": 1.3840199956964762e-06, "loss": 0.3657, "step": 10785 }, { "epoch": 0.764693371144984, "grad_norm": 2.6207756229848695, "learning_rate": 1.383227123238275e-06, "loss": 0.4176, "step": 10786 }, { "epoch": 0.7647642679900745, "grad_norm": 3.757283033922569, "learning_rate": 1.38243444149527e-06, "loss": 0.3777, "step": 10787 }, { "epoch": 0.7648351648351648, "grad_norm": 2.1838077356758867, "learning_rate": 1.3816419505092582e-06, "loss": 0.366, "step": 10788 }, { "epoch": 0.7649060616802552, "grad_norm": 4.340960577846673, "learning_rate": 1.3808496503220263e-06, "loss": 0.3871, "step": 10789 }, { "epoch": 0.7649769585253456, "grad_norm": 2.5314001044942267, "learning_rate": 1.3800575409753547e-06, "loss": 0.4484, "step": 10790 }, { "epoch": 0.7650478553704361, "grad_norm": 2.9257854067677327, "learning_rate": 1.379265622511014e-06, "loss": 0.4144, "step": 10791 }, { "epoch": 0.7651187522155264, "grad_norm": 2.649314716128769, "learning_rate": 1.3784738949707599e-06, "loss": 0.3766, "step": 10792 }, { "epoch": 0.7651896490606168, "grad_norm": 2.8866670610239207, "learning_rate": 1.3776823583963406e-06, "loss": 0.3917, "step": 10793 }, { "epoch": 0.7652605459057072, "grad_norm": 2.5102872042275255, "learning_rate": 1.3768910128294948e-06, "loss": 0.3938, "step": 10794 }, { "epoch": 0.7653314427507976, "grad_norm": 3.3850009665134726, "learning_rate": 1.3760998583119546e-06, "loss": 0.4, "step": 10795 }, { "epoch": 0.765402339595888, "grad_norm": 2.489413101334232, "learning_rate": 1.3753088948854316e-06, "loss": 0.4154, "step": 10796 }, { "epoch": 0.7654732364409784, "grad_norm": 3.556404365029406, "learning_rate": 1.3745181225916377e-06, "loss": 0.3554, "step": 10797 }, { "epoch": 0.7655441332860687, "grad_norm": 12.512825609341661, "learning_rate": 1.3737275414722712e-06, "loss": 0.3815, "step": 10798 }, { "epoch": 0.7656150301311592, "grad_norm": 3.618949621172501, "learning_rate": 1.3729371515690198e-06, "loss": 0.3873, "step": 10799 }, { "epoch": 0.7656859269762496, "grad_norm": 1.8158508941236176, "learning_rate": 1.3721469529235592e-06, "loss": 0.3875, "step": 10800 }, { "epoch": 0.7657568238213399, "grad_norm": 2.7543030892037517, "learning_rate": 1.37135694557756e-06, "loss": 0.3801, "step": 10801 }, { "epoch": 0.7658277206664303, "grad_norm": 2.5080262376794646, "learning_rate": 1.3705671295726774e-06, "loss": 0.4271, "step": 10802 }, { "epoch": 0.7658986175115208, "grad_norm": 2.875163713322062, "learning_rate": 1.3697775049505619e-06, "loss": 0.4015, "step": 10803 }, { "epoch": 0.7659695143566111, "grad_norm": 4.049489022248205, "learning_rate": 1.3689880717528492e-06, "loss": 0.399, "step": 10804 }, { "epoch": 0.7660404112017015, "grad_norm": 2.4601623744099546, "learning_rate": 1.368198830021165e-06, "loss": 0.3983, "step": 10805 }, { "epoch": 0.766111308046792, "grad_norm": 1.331793578007696, "learning_rate": 1.3674097797971309e-06, "loss": 0.3635, "step": 10806 }, { "epoch": 0.7661822048918823, "grad_norm": 2.977812149747354, "learning_rate": 1.3666209211223496e-06, "loss": 0.3651, "step": 10807 }, { "epoch": 0.7662531017369727, "grad_norm": 3.458312010240945, "learning_rate": 1.3658322540384223e-06, "loss": 0.3679, "step": 10808 }, { "epoch": 0.7663239985820631, "grad_norm": 3.925034144815688, "learning_rate": 1.3650437785869325e-06, "loss": 0.3997, "step": 10809 }, { "epoch": 0.7663948954271534, "grad_norm": 2.7969818610122217, "learning_rate": 1.36425549480946e-06, "loss": 0.3761, "step": 10810 }, { "epoch": 0.7664657922722439, "grad_norm": 2.4395329892093605, "learning_rate": 1.3634674027475702e-06, "loss": 0.3595, "step": 10811 }, { "epoch": 0.7665366891173343, "grad_norm": 3.6188142145217164, "learning_rate": 1.3626795024428185e-06, "loss": 0.3832, "step": 10812 }, { "epoch": 0.7666075859624246, "grad_norm": 3.7642235986783064, "learning_rate": 1.3618917939367526e-06, "loss": 0.3976, "step": 10813 }, { "epoch": 0.766678482807515, "grad_norm": 3.5434731863878013, "learning_rate": 1.3611042772709126e-06, "loss": 0.3928, "step": 10814 }, { "epoch": 0.7667493796526055, "grad_norm": 257.55023283033205, "learning_rate": 1.3603169524868186e-06, "loss": 0.4017, "step": 10815 }, { "epoch": 0.7668202764976959, "grad_norm": 3.576159880631145, "learning_rate": 1.3595298196259899e-06, "loss": 0.3545, "step": 10816 }, { "epoch": 0.7668911733427862, "grad_norm": 4.19241706082119, "learning_rate": 1.3587428787299333e-06, "loss": 0.4119, "step": 10817 }, { "epoch": 0.7669620701878767, "grad_norm": 4.730154417867445, "learning_rate": 1.3579561298401444e-06, "loss": 0.3999, "step": 10818 }, { "epoch": 0.7670329670329671, "grad_norm": 2.7847558865058093, "learning_rate": 1.357169572998107e-06, "loss": 0.4132, "step": 10819 }, { "epoch": 0.7671038638780574, "grad_norm": 2.1790214475568757, "learning_rate": 1.3563832082452998e-06, "loss": 0.3711, "step": 10820 }, { "epoch": 0.7671747607231478, "grad_norm": 3.022909277385963, "learning_rate": 1.3555970356231856e-06, "loss": 0.3664, "step": 10821 }, { "epoch": 0.7672456575682383, "grad_norm": 3.905155201745027, "learning_rate": 1.3548110551732234e-06, "loss": 0.3787, "step": 10822 }, { "epoch": 0.7673165544133286, "grad_norm": 1.9863357096724752, "learning_rate": 1.3540252669368547e-06, "loss": 0.387, "step": 10823 }, { "epoch": 0.767387451258419, "grad_norm": 3.9200212564575585, "learning_rate": 1.3532396709555179e-06, "loss": 0.4072, "step": 10824 }, { "epoch": 0.7674583481035094, "grad_norm": 2.2459633503676324, "learning_rate": 1.3524542672706365e-06, "loss": 0.4279, "step": 10825 }, { "epoch": 0.7675292449485998, "grad_norm": 2.051876455180929, "learning_rate": 1.3516690559236246e-06, "loss": 0.3844, "step": 10826 }, { "epoch": 0.7676001417936902, "grad_norm": 3.506611587962008, "learning_rate": 1.350884036955889e-06, "loss": 0.3999, "step": 10827 }, { "epoch": 0.7676710386387806, "grad_norm": 4.502610029661848, "learning_rate": 1.3500992104088222e-06, "loss": 0.4335, "step": 10828 }, { "epoch": 0.7677419354838709, "grad_norm": 2.8175519259902084, "learning_rate": 1.3493145763238113e-06, "loss": 0.3588, "step": 10829 }, { "epoch": 0.7678128323289614, "grad_norm": 3.7394409788354697, "learning_rate": 1.3485301347422297e-06, "loss": 0.4105, "step": 10830 }, { "epoch": 0.7678837291740518, "grad_norm": 1.7809596999524346, "learning_rate": 1.3477458857054393e-06, "loss": 0.3506, "step": 10831 }, { "epoch": 0.7679546260191421, "grad_norm": 8.892050116279307, "learning_rate": 1.346961829254797e-06, "loss": 0.4249, "step": 10832 }, { "epoch": 0.7680255228642325, "grad_norm": 9.926823768527703, "learning_rate": 1.3461779654316464e-06, "loss": 0.3822, "step": 10833 }, { "epoch": 0.768096419709323, "grad_norm": 1.2232710521923524, "learning_rate": 1.3453942942773185e-06, "loss": 0.6484, "step": 10834 }, { "epoch": 0.7681673165544133, "grad_norm": 1.5206903871896826, "learning_rate": 1.3446108158331388e-06, "loss": 0.3863, "step": 10835 }, { "epoch": 0.7682382133995037, "grad_norm": 15.345498972549748, "learning_rate": 1.3438275301404229e-06, "loss": 0.4139, "step": 10836 }, { "epoch": 0.7683091102445941, "grad_norm": 2.121556685467305, "learning_rate": 1.3430444372404716e-06, "loss": 0.3698, "step": 10837 }, { "epoch": 0.7683800070896845, "grad_norm": 84.88581683111784, "learning_rate": 1.3422615371745773e-06, "loss": 0.361, "step": 10838 }, { "epoch": 0.7684509039347749, "grad_norm": 2.6452577202590217, "learning_rate": 1.3414788299840231e-06, "loss": 0.3907, "step": 10839 }, { "epoch": 0.7685218007798653, "grad_norm": 5.192372780600639, "learning_rate": 1.3406963157100866e-06, "loss": 0.4098, "step": 10840 }, { "epoch": 0.7685926976249556, "grad_norm": 1.8997884871209598, "learning_rate": 1.3399139943940226e-06, "loss": 0.3969, "step": 10841 }, { "epoch": 0.7686635944700461, "grad_norm": 5.220902487668434, "learning_rate": 1.3391318660770874e-06, "loss": 0.4617, "step": 10842 }, { "epoch": 0.7687344913151365, "grad_norm": 2.168082077548947, "learning_rate": 1.338349930800525e-06, "loss": 0.3844, "step": 10843 }, { "epoch": 0.7688053881602269, "grad_norm": 1.7858105897248582, "learning_rate": 1.3375681886055652e-06, "loss": 0.3736, "step": 10844 }, { "epoch": 0.7688762850053172, "grad_norm": 2.4000252562345086, "learning_rate": 1.336786639533429e-06, "loss": 0.3911, "step": 10845 }, { "epoch": 0.7689471818504077, "grad_norm": 1.0378481088137996, "learning_rate": 1.3360052836253308e-06, "loss": 0.6107, "step": 10846 }, { "epoch": 0.7690180786954981, "grad_norm": 2.2751653371926763, "learning_rate": 1.3352241209224693e-06, "loss": 0.3534, "step": 10847 }, { "epoch": 0.7690889755405884, "grad_norm": 3.3290408647830194, "learning_rate": 1.3344431514660378e-06, "loss": 0.3704, "step": 10848 }, { "epoch": 0.7691598723856788, "grad_norm": 2.6532432167667586, "learning_rate": 1.3336623752972171e-06, "loss": 0.3701, "step": 10849 }, { "epoch": 0.7692307692307693, "grad_norm": 5.152030187672682, "learning_rate": 1.332881792457177e-06, "loss": 0.4091, "step": 10850 }, { "epoch": 0.7693016660758596, "grad_norm": 3.5530214947756855, "learning_rate": 1.3321014029870794e-06, "loss": 0.3917, "step": 10851 }, { "epoch": 0.76937256292095, "grad_norm": 2.001508832871924, "learning_rate": 1.3313212069280734e-06, "loss": 0.3639, "step": 10852 }, { "epoch": 0.7694434597660404, "grad_norm": 10.962524488110343, "learning_rate": 1.3305412043213018e-06, "loss": 0.3918, "step": 10853 }, { "epoch": 0.7695143566111308, "grad_norm": 2.4801774402545913, "learning_rate": 1.329761395207892e-06, "loss": 0.3842, "step": 10854 }, { "epoch": 0.7695852534562212, "grad_norm": 2.204973039861028, "learning_rate": 1.3289817796289662e-06, "loss": 0.3659, "step": 10855 }, { "epoch": 0.7696561503013116, "grad_norm": 4.052116694597465, "learning_rate": 1.3282023576256332e-06, "loss": 0.3409, "step": 10856 }, { "epoch": 0.7697270471464019, "grad_norm": 3.3147249483322825, "learning_rate": 1.3274231292389911e-06, "loss": 0.4178, "step": 10857 }, { "epoch": 0.7697979439914924, "grad_norm": 3.0189428201475934, "learning_rate": 1.3266440945101305e-06, "loss": 0.3987, "step": 10858 }, { "epoch": 0.7698688408365828, "grad_norm": 11.912586603279902, "learning_rate": 1.3258652534801336e-06, "loss": 0.3941, "step": 10859 }, { "epoch": 0.7699397376816731, "grad_norm": 2.1782607364497157, "learning_rate": 1.3250866061900631e-06, "loss": 0.4137, "step": 10860 }, { "epoch": 0.7700106345267635, "grad_norm": 3.277380042104304, "learning_rate": 1.324308152680981e-06, "loss": 0.3916, "step": 10861 }, { "epoch": 0.770081531371854, "grad_norm": 2.138026225266625, "learning_rate": 1.3235298929939365e-06, "loss": 0.3806, "step": 10862 }, { "epoch": 0.7701524282169443, "grad_norm": 2.5063906780075653, "learning_rate": 1.3227518271699675e-06, "loss": 0.4149, "step": 10863 }, { "epoch": 0.7702233250620347, "grad_norm": 1.6157173886544156, "learning_rate": 1.3219739552501e-06, "loss": 0.3788, "step": 10864 }, { "epoch": 0.7702942219071252, "grad_norm": 2.147029355073953, "learning_rate": 1.3211962772753527e-06, "loss": 0.3769, "step": 10865 }, { "epoch": 0.7703651187522155, "grad_norm": 3.6772289983652247, "learning_rate": 1.3204187932867352e-06, "loss": 0.4563, "step": 10866 }, { "epoch": 0.7704360155973059, "grad_norm": 4.470636674126657, "learning_rate": 1.319641503325243e-06, "loss": 0.3785, "step": 10867 }, { "epoch": 0.7705069124423963, "grad_norm": 4.29540599211152, "learning_rate": 1.3188644074318624e-06, "loss": 0.4414, "step": 10868 }, { "epoch": 0.7705778092874868, "grad_norm": 2.591075886307014, "learning_rate": 1.3180875056475722e-06, "loss": 0.3695, "step": 10869 }, { "epoch": 0.7706487061325771, "grad_norm": 3.102946073723184, "learning_rate": 1.3173107980133387e-06, "loss": 0.4312, "step": 10870 }, { "epoch": 0.7707196029776675, "grad_norm": 2.589572589142018, "learning_rate": 1.3165342845701157e-06, "loss": 0.3727, "step": 10871 }, { "epoch": 0.7707904998227579, "grad_norm": 2.094991552300572, "learning_rate": 1.3157579653588532e-06, "loss": 0.4029, "step": 10872 }, { "epoch": 0.7708613966678483, "grad_norm": 1.0640038503770253, "learning_rate": 1.3149818404204833e-06, "loss": 0.7301, "step": 10873 }, { "epoch": 0.7709322935129387, "grad_norm": 3.6171707115938396, "learning_rate": 1.3142059097959359e-06, "loss": 0.4094, "step": 10874 }, { "epoch": 0.7710031903580291, "grad_norm": 3.945533486557765, "learning_rate": 1.3134301735261234e-06, "loss": 0.3953, "step": 10875 }, { "epoch": 0.7710740872031194, "grad_norm": 2.4362763122072195, "learning_rate": 1.31265463165195e-06, "loss": 0.4026, "step": 10876 }, { "epoch": 0.7711449840482099, "grad_norm": 7.851507882286301, "learning_rate": 1.311879284214313e-06, "loss": 0.4049, "step": 10877 }, { "epoch": 0.7712158808933003, "grad_norm": 4.658753794952379, "learning_rate": 1.311104131254099e-06, "loss": 0.413, "step": 10878 }, { "epoch": 0.7712867777383906, "grad_norm": 1.8833404683870214, "learning_rate": 1.310329172812177e-06, "loss": 0.3817, "step": 10879 }, { "epoch": 0.771357674583481, "grad_norm": 2.923215260863174, "learning_rate": 1.3095544089294138e-06, "loss": 0.3964, "step": 10880 }, { "epoch": 0.7714285714285715, "grad_norm": 11.37484864045233, "learning_rate": 1.3087798396466644e-06, "loss": 0.4289, "step": 10881 }, { "epoch": 0.7714994682736618, "grad_norm": 2.2607221962480866, "learning_rate": 1.3080054650047718e-06, "loss": 0.3509, "step": 10882 }, { "epoch": 0.7715703651187522, "grad_norm": 2.9109467806111233, "learning_rate": 1.307231285044568e-06, "loss": 0.4491, "step": 10883 }, { "epoch": 0.7716412619638426, "grad_norm": 2.993524390282383, "learning_rate": 1.3064572998068765e-06, "loss": 0.399, "step": 10884 }, { "epoch": 0.771712158808933, "grad_norm": 3.4751700865963344, "learning_rate": 1.3056835093325138e-06, "loss": 0.3387, "step": 10885 }, { "epoch": 0.7717830556540234, "grad_norm": 1.7196564370984189, "learning_rate": 1.3049099136622767e-06, "loss": 0.4124, "step": 10886 }, { "epoch": 0.7718539524991138, "grad_norm": 2.271898068210717, "learning_rate": 1.3041365128369598e-06, "loss": 0.3961, "step": 10887 }, { "epoch": 0.7719248493442041, "grad_norm": 2.867996921513719, "learning_rate": 1.3033633068973473e-06, "loss": 0.4332, "step": 10888 }, { "epoch": 0.7719957461892946, "grad_norm": 3.290303164679166, "learning_rate": 1.302590295884209e-06, "loss": 0.3585, "step": 10889 }, { "epoch": 0.772066643034385, "grad_norm": 5.161842075795049, "learning_rate": 1.3018174798383049e-06, "loss": 0.4095, "step": 10890 }, { "epoch": 0.7721375398794753, "grad_norm": 1.6962266758870135, "learning_rate": 1.3010448588003894e-06, "loss": 0.4079, "step": 10891 }, { "epoch": 0.7722084367245657, "grad_norm": 4.368396429690976, "learning_rate": 1.3002724328112005e-06, "loss": 0.4586, "step": 10892 }, { "epoch": 0.7722793335696562, "grad_norm": 2.4984316942839127, "learning_rate": 1.2995002019114716e-06, "loss": 0.3503, "step": 10893 }, { "epoch": 0.7723502304147466, "grad_norm": 5.359091976780791, "learning_rate": 1.298728166141921e-06, "loss": 0.4217, "step": 10894 }, { "epoch": 0.7724211272598369, "grad_norm": 3.665245324391374, "learning_rate": 1.297956325543258e-06, "loss": 0.4199, "step": 10895 }, { "epoch": 0.7724920241049273, "grad_norm": 3.935138390676282, "learning_rate": 1.2971846801561854e-06, "loss": 0.4148, "step": 10896 }, { "epoch": 0.7725629209500178, "grad_norm": 3.44845428751658, "learning_rate": 1.2964132300213888e-06, "loss": 0.4102, "step": 10897 }, { "epoch": 0.7726338177951081, "grad_norm": 3.0257790918480585, "learning_rate": 1.2956419751795513e-06, "loss": 0.4058, "step": 10898 }, { "epoch": 0.7727047146401985, "grad_norm": 2.2967660269927213, "learning_rate": 1.2948709156713386e-06, "loss": 0.3655, "step": 10899 }, { "epoch": 0.772775611485289, "grad_norm": 3.05688617118667, "learning_rate": 1.2941000515374114e-06, "loss": 0.4128, "step": 10900 }, { "epoch": 0.7728465083303793, "grad_norm": 3.4015889643605783, "learning_rate": 1.2933293828184175e-06, "loss": 0.3178, "step": 10901 }, { "epoch": 0.7729174051754697, "grad_norm": 3.139790768512296, "learning_rate": 1.2925589095549928e-06, "loss": 0.3913, "step": 10902 }, { "epoch": 0.7729883020205601, "grad_norm": 6.19888068336876, "learning_rate": 1.2917886317877665e-06, "loss": 0.3567, "step": 10903 }, { "epoch": 0.7730591988656504, "grad_norm": 3.1412002588289343, "learning_rate": 1.291018549557359e-06, "loss": 0.4185, "step": 10904 }, { "epoch": 0.7731300957107409, "grad_norm": 2.782205037328913, "learning_rate": 1.2902486629043708e-06, "loss": 0.3748, "step": 10905 }, { "epoch": 0.7732009925558313, "grad_norm": 4.635634293148375, "learning_rate": 1.2894789718694028e-06, "loss": 0.3762, "step": 10906 }, { "epoch": 0.7732718894009216, "grad_norm": 4.211041646421752, "learning_rate": 1.2887094764930425e-06, "loss": 0.3477, "step": 10907 }, { "epoch": 0.773342786246012, "grad_norm": 2.922410144379966, "learning_rate": 1.2879401768158634e-06, "loss": 0.386, "step": 10908 }, { "epoch": 0.7734136830911025, "grad_norm": 2.502998925074107, "learning_rate": 1.287171072878431e-06, "loss": 0.4417, "step": 10909 }, { "epoch": 0.7734845799361928, "grad_norm": 1.8429516786914886, "learning_rate": 1.286402164721302e-06, "loss": 0.4062, "step": 10910 }, { "epoch": 0.7735554767812832, "grad_norm": 2.3185967148339204, "learning_rate": 1.2856334523850222e-06, "loss": 0.3728, "step": 10911 }, { "epoch": 0.7736263736263737, "grad_norm": 3.3799136087304316, "learning_rate": 1.2848649359101262e-06, "loss": 0.3816, "step": 10912 }, { "epoch": 0.773697270471464, "grad_norm": 2.299990444243096, "learning_rate": 1.2840966153371359e-06, "loss": 0.3857, "step": 10913 }, { "epoch": 0.7737681673165544, "grad_norm": 5.589687092918818, "learning_rate": 1.2833284907065685e-06, "loss": 0.399, "step": 10914 }, { "epoch": 0.7738390641616448, "grad_norm": 3.7737955317068166, "learning_rate": 1.2825605620589271e-06, "loss": 0.3953, "step": 10915 }, { "epoch": 0.7739099610067351, "grad_norm": 2.270762940139143, "learning_rate": 1.2817928294347031e-06, "loss": 0.3864, "step": 10916 }, { "epoch": 0.7739808578518256, "grad_norm": 4.734898624294436, "learning_rate": 1.2810252928743827e-06, "loss": 0.3762, "step": 10917 }, { "epoch": 0.774051754696916, "grad_norm": 2.7866640879461837, "learning_rate": 1.2802579524184355e-06, "loss": 0.3933, "step": 10918 }, { "epoch": 0.7741226515420064, "grad_norm": 2.3202981997682848, "learning_rate": 1.279490808107327e-06, "loss": 0.3343, "step": 10919 }, { "epoch": 0.7741935483870968, "grad_norm": 22.810335495355428, "learning_rate": 1.2787238599815083e-06, "loss": 0.4455, "step": 10920 }, { "epoch": 0.7742644452321872, "grad_norm": 2.61750913059403, "learning_rate": 1.277957108081419e-06, "loss": 0.3761, "step": 10921 }, { "epoch": 0.7743353420772776, "grad_norm": 3.4319285147168084, "learning_rate": 1.277190552447492e-06, "loss": 0.4139, "step": 10922 }, { "epoch": 0.7744062389223679, "grad_norm": 3.2768935625563804, "learning_rate": 1.276424193120152e-06, "loss": 0.3998, "step": 10923 }, { "epoch": 0.7744771357674584, "grad_norm": 3.712871711764263, "learning_rate": 1.2756580301398036e-06, "loss": 0.418, "step": 10924 }, { "epoch": 0.7745480326125488, "grad_norm": 3.149520940935242, "learning_rate": 1.2748920635468497e-06, "loss": 0.3793, "step": 10925 }, { "epoch": 0.7746189294576391, "grad_norm": 4.74697489048457, "learning_rate": 1.2741262933816824e-06, "loss": 0.3726, "step": 10926 }, { "epoch": 0.7746898263027295, "grad_norm": 2.293948109528895, "learning_rate": 1.2733607196846786e-06, "loss": 0.4492, "step": 10927 }, { "epoch": 0.77476072314782, "grad_norm": 2.4698203868483883, "learning_rate": 1.272595342496208e-06, "loss": 0.4325, "step": 10928 }, { "epoch": 0.7748316199929103, "grad_norm": 1.7369257970873235, "learning_rate": 1.2718301618566293e-06, "loss": 0.3665, "step": 10929 }, { "epoch": 0.7749025168380007, "grad_norm": 2.60236402386418, "learning_rate": 1.2710651778062932e-06, "loss": 0.3924, "step": 10930 }, { "epoch": 0.7749734136830911, "grad_norm": 2.443146916499714, "learning_rate": 1.2703003903855364e-06, "loss": 0.4056, "step": 10931 }, { "epoch": 0.7750443105281815, "grad_norm": 1.1886824677680838, "learning_rate": 1.2695357996346858e-06, "loss": 0.6747, "step": 10932 }, { "epoch": 0.7751152073732719, "grad_norm": 3.004138826228966, "learning_rate": 1.2687714055940614e-06, "loss": 0.4079, "step": 10933 }, { "epoch": 0.7751861042183623, "grad_norm": 3.378946157913116, "learning_rate": 1.268007208303968e-06, "loss": 0.3693, "step": 10934 }, { "epoch": 0.7752570010634526, "grad_norm": 3.634551913397478, "learning_rate": 1.267243207804702e-06, "loss": 0.378, "step": 10935 }, { "epoch": 0.7753278979085431, "grad_norm": 3.674790416193692, "learning_rate": 1.266479404136552e-06, "loss": 0.3601, "step": 10936 }, { "epoch": 0.7753987947536335, "grad_norm": 3.6449336592864903, "learning_rate": 1.2657157973397915e-06, "loss": 0.3744, "step": 10937 }, { "epoch": 0.7754696915987238, "grad_norm": 2.0939645116759076, "learning_rate": 1.2649523874546883e-06, "loss": 0.3732, "step": 10938 }, { "epoch": 0.7755405884438142, "grad_norm": 3.444119126115472, "learning_rate": 1.2641891745214951e-06, "loss": 0.3568, "step": 10939 }, { "epoch": 0.7756114852889047, "grad_norm": 1.9150203421975796, "learning_rate": 1.2634261585804603e-06, "loss": 0.358, "step": 10940 }, { "epoch": 0.775682382133995, "grad_norm": 2.6739299100516134, "learning_rate": 1.2626633396718158e-06, "loss": 0.366, "step": 10941 }, { "epoch": 0.7757532789790854, "grad_norm": 2.211317458240796, "learning_rate": 1.261900717835784e-06, "loss": 0.4273, "step": 10942 }, { "epoch": 0.7758241758241758, "grad_norm": 2.5936007930744696, "learning_rate": 1.2611382931125827e-06, "loss": 0.4325, "step": 10943 }, { "epoch": 0.7758950726692663, "grad_norm": 3.3324847033468146, "learning_rate": 1.2603760655424108e-06, "loss": 0.3607, "step": 10944 }, { "epoch": 0.7759659695143566, "grad_norm": 2.4956605737444835, "learning_rate": 1.259614035165465e-06, "loss": 0.3609, "step": 10945 }, { "epoch": 0.776036866359447, "grad_norm": 3.621876108797226, "learning_rate": 1.2588522020219263e-06, "loss": 0.3619, "step": 10946 }, { "epoch": 0.7761077632045374, "grad_norm": 2.2553787108637895, "learning_rate": 1.2580905661519649e-06, "loss": 0.3676, "step": 10947 }, { "epoch": 0.7761786600496278, "grad_norm": 3.7359657009216614, "learning_rate": 1.2573291275957438e-06, "loss": 0.4086, "step": 10948 }, { "epoch": 0.7762495568947182, "grad_norm": 3.069901381884863, "learning_rate": 1.2565678863934177e-06, "loss": 0.3555, "step": 10949 }, { "epoch": 0.7763204537398086, "grad_norm": 4.512355048855873, "learning_rate": 1.255806842585121e-06, "loss": 0.4031, "step": 10950 }, { "epoch": 0.7763913505848989, "grad_norm": 3.287701365250557, "learning_rate": 1.2550459962109873e-06, "loss": 0.3846, "step": 10951 }, { "epoch": 0.7764622474299894, "grad_norm": 2.037132124698426, "learning_rate": 1.2542853473111382e-06, "loss": 0.4101, "step": 10952 }, { "epoch": 0.7765331442750798, "grad_norm": 4.249340260998421, "learning_rate": 1.2535248959256818e-06, "loss": 0.4261, "step": 10953 }, { "epoch": 0.7766040411201701, "grad_norm": 3.6821840636295646, "learning_rate": 1.2527646420947154e-06, "loss": 0.399, "step": 10954 }, { "epoch": 0.7766749379652605, "grad_norm": 4.645175388569147, "learning_rate": 1.25200458585833e-06, "loss": 0.3936, "step": 10955 }, { "epoch": 0.776745834810351, "grad_norm": 3.726044423385327, "learning_rate": 1.2512447272566047e-06, "loss": 0.4058, "step": 10956 }, { "epoch": 0.7768167316554413, "grad_norm": 2.5834841800571304, "learning_rate": 1.2504850663296063e-06, "loss": 0.412, "step": 10957 }, { "epoch": 0.7768876285005317, "grad_norm": 5.337954754107657, "learning_rate": 1.2497256031173904e-06, "loss": 0.3578, "step": 10958 }, { "epoch": 0.7769585253456222, "grad_norm": 2.342151188207441, "learning_rate": 1.248966337660008e-06, "loss": 0.3893, "step": 10959 }, { "epoch": 0.7770294221907125, "grad_norm": 2.320184814482524, "learning_rate": 1.2482072699974935e-06, "loss": 0.3759, "step": 10960 }, { "epoch": 0.7771003190358029, "grad_norm": 2.5467571587454483, "learning_rate": 1.2474484001698718e-06, "loss": 0.3983, "step": 10961 }, { "epoch": 0.7771712158808933, "grad_norm": 3.0891582049494763, "learning_rate": 1.2466897282171613e-06, "loss": 0.3576, "step": 10962 }, { "epoch": 0.7772421127259836, "grad_norm": 2.9876959086535333, "learning_rate": 1.245931254179365e-06, "loss": 0.37, "step": 10963 }, { "epoch": 0.7773130095710741, "grad_norm": 5.316377145574539, "learning_rate": 1.24517297809648e-06, "loss": 0.3656, "step": 10964 }, { "epoch": 0.7773839064161645, "grad_norm": 3.657060281394585, "learning_rate": 1.2444149000084904e-06, "loss": 0.3681, "step": 10965 }, { "epoch": 0.7774548032612548, "grad_norm": 4.407614945598461, "learning_rate": 1.2436570199553677e-06, "loss": 0.3766, "step": 10966 }, { "epoch": 0.7775257001063453, "grad_norm": 3.827694700311364, "learning_rate": 1.2428993379770777e-06, "loss": 0.3758, "step": 10967 }, { "epoch": 0.7775965969514357, "grad_norm": 2.3510619874619354, "learning_rate": 1.242141854113575e-06, "loss": 0.3921, "step": 10968 }, { "epoch": 0.777667493796526, "grad_norm": 4.7912521615992585, "learning_rate": 1.2413845684047998e-06, "loss": 0.4027, "step": 10969 }, { "epoch": 0.7777383906416164, "grad_norm": 3.8384240969543777, "learning_rate": 1.2406274808906843e-06, "loss": 0.4222, "step": 10970 }, { "epoch": 0.7778092874867069, "grad_norm": 2.4276131359813986, "learning_rate": 1.2398705916111526e-06, "loss": 0.3342, "step": 10971 }, { "epoch": 0.7778801843317973, "grad_norm": 3.0717532914020196, "learning_rate": 1.2391139006061143e-06, "loss": 0.4254, "step": 10972 }, { "epoch": 0.7779510811768876, "grad_norm": 3.1494798328507696, "learning_rate": 1.2383574079154692e-06, "loss": 0.3702, "step": 10973 }, { "epoch": 0.778021978021978, "grad_norm": 2.8249894067916244, "learning_rate": 1.2376011135791094e-06, "loss": 0.402, "step": 10974 }, { "epoch": 0.7780928748670685, "grad_norm": 2.4362237755464604, "learning_rate": 1.2368450176369157e-06, "loss": 0.3772, "step": 10975 }, { "epoch": 0.7781637717121588, "grad_norm": 3.1645731447315066, "learning_rate": 1.2360891201287567e-06, "loss": 0.3894, "step": 10976 }, { "epoch": 0.7782346685572492, "grad_norm": 2.06049280796509, "learning_rate": 1.2353334210944895e-06, "loss": 0.3529, "step": 10977 }, { "epoch": 0.7783055654023396, "grad_norm": 2.7261222741278424, "learning_rate": 1.234577920573966e-06, "loss": 0.3621, "step": 10978 }, { "epoch": 0.77837646224743, "grad_norm": 3.898736502815473, "learning_rate": 1.2338226186070228e-06, "loss": 0.3899, "step": 10979 }, { "epoch": 0.7784473590925204, "grad_norm": 2.7944843492065883, "learning_rate": 1.2330675152334864e-06, "loss": 0.432, "step": 10980 }, { "epoch": 0.7785182559376108, "grad_norm": 2.8273826264973785, "learning_rate": 1.2323126104931759e-06, "loss": 0.3872, "step": 10981 }, { "epoch": 0.7785891527827011, "grad_norm": 2.6098437775836962, "learning_rate": 1.2315579044258964e-06, "loss": 0.3941, "step": 10982 }, { "epoch": 0.7786600496277916, "grad_norm": 3.785953168278964, "learning_rate": 1.230803397071446e-06, "loss": 0.3591, "step": 10983 }, { "epoch": 0.778730946472882, "grad_norm": 1.7934777922084417, "learning_rate": 1.230049088469608e-06, "loss": 0.367, "step": 10984 }, { "epoch": 0.7788018433179723, "grad_norm": 5.852208251733818, "learning_rate": 1.2292949786601604e-06, "loss": 0.3902, "step": 10985 }, { "epoch": 0.7788727401630627, "grad_norm": 3.1284668649254406, "learning_rate": 1.2285410676828652e-06, "loss": 0.4116, "step": 10986 }, { "epoch": 0.7789436370081532, "grad_norm": 7.357732074617166, "learning_rate": 1.22778735557748e-06, "loss": 0.3867, "step": 10987 }, { "epoch": 0.7790145338532435, "grad_norm": 2.021226146857321, "learning_rate": 1.2270338423837464e-06, "loss": 0.366, "step": 10988 }, { "epoch": 0.7790854306983339, "grad_norm": 2.7662032754808306, "learning_rate": 1.2262805281413964e-06, "loss": 0.4334, "step": 10989 }, { "epoch": 0.7791563275434243, "grad_norm": 1.9918775223207956, "learning_rate": 1.2255274128901561e-06, "loss": 0.383, "step": 10990 }, { "epoch": 0.7792272243885147, "grad_norm": 2.3011964574769133, "learning_rate": 1.2247744966697362e-06, "loss": 0.3541, "step": 10991 }, { "epoch": 0.7792981212336051, "grad_norm": 2.0585624120722774, "learning_rate": 1.2240217795198372e-06, "loss": 0.3536, "step": 10992 }, { "epoch": 0.7793690180786955, "grad_norm": 2.54770988172358, "learning_rate": 1.2232692614801523e-06, "loss": 0.4177, "step": 10993 }, { "epoch": 0.7794399149237858, "grad_norm": 1.8122430578930506, "learning_rate": 1.2225169425903644e-06, "loss": 0.3719, "step": 10994 }, { "epoch": 0.7795108117688763, "grad_norm": 2.409252266808779, "learning_rate": 1.221764822890138e-06, "loss": 0.3871, "step": 10995 }, { "epoch": 0.7795817086139667, "grad_norm": 5.430844039596564, "learning_rate": 1.221012902419137e-06, "loss": 0.3929, "step": 10996 }, { "epoch": 0.7796526054590571, "grad_norm": 1.046223469339351, "learning_rate": 1.2202611812170112e-06, "loss": 0.6847, "step": 10997 }, { "epoch": 0.7797235023041474, "grad_norm": 5.448333468429737, "learning_rate": 1.219509659323398e-06, "loss": 0.3829, "step": 10998 }, { "epoch": 0.7797943991492379, "grad_norm": 2.7079676562924484, "learning_rate": 1.218758336777925e-06, "loss": 0.3359, "step": 10999 }, { "epoch": 0.7798652959943283, "grad_norm": 3.325415344331361, "learning_rate": 1.2180072136202103e-06, "loss": 0.3805, "step": 11000 }, { "epoch": 0.7799361928394186, "grad_norm": 3.3483094811117624, "learning_rate": 1.2172562898898637e-06, "loss": 0.4011, "step": 11001 }, { "epoch": 0.780007089684509, "grad_norm": 2.442008961610186, "learning_rate": 1.2165055656264807e-06, "loss": 0.3817, "step": 11002 }, { "epoch": 0.7800779865295995, "grad_norm": 2.1285467572147887, "learning_rate": 1.215755040869645e-06, "loss": 0.3649, "step": 11003 }, { "epoch": 0.7801488833746898, "grad_norm": 1.8877648032931593, "learning_rate": 1.2150047156589356e-06, "loss": 0.4103, "step": 11004 }, { "epoch": 0.7802197802197802, "grad_norm": 3.2246779784222115, "learning_rate": 1.214254590033917e-06, "loss": 0.4055, "step": 11005 }, { "epoch": 0.7802906770648707, "grad_norm": 2.953259781752509, "learning_rate": 1.2135046640341419e-06, "loss": 0.3892, "step": 11006 }, { "epoch": 0.780361573909961, "grad_norm": 3.0902717824423442, "learning_rate": 1.212754937699157e-06, "loss": 0.4125, "step": 11007 }, { "epoch": 0.7804324707550514, "grad_norm": 8.815782188686503, "learning_rate": 1.2120054110684943e-06, "loss": 0.3722, "step": 11008 }, { "epoch": 0.7805033676001418, "grad_norm": 16.977881141973302, "learning_rate": 1.2112560841816783e-06, "loss": 0.4029, "step": 11009 }, { "epoch": 0.7805742644452321, "grad_norm": 2.4187544435444517, "learning_rate": 1.2105069570782208e-06, "loss": 0.3214, "step": 11010 }, { "epoch": 0.7806451612903226, "grad_norm": 2.192632531876802, "learning_rate": 1.2097580297976225e-06, "loss": 0.4139, "step": 11011 }, { "epoch": 0.780716058135413, "grad_norm": 3.4003906082178124, "learning_rate": 1.2090093023793764e-06, "loss": 0.3844, "step": 11012 }, { "epoch": 0.7807869549805033, "grad_norm": 2.2320225883354374, "learning_rate": 1.208260774862965e-06, "loss": 0.3517, "step": 11013 }, { "epoch": 0.7808578518255938, "grad_norm": 2.3475428941192313, "learning_rate": 1.2075124472878563e-06, "loss": 0.3926, "step": 11014 }, { "epoch": 0.7809287486706842, "grad_norm": 2.6248000937367277, "learning_rate": 1.2067643196935098e-06, "loss": 0.3692, "step": 11015 }, { "epoch": 0.7809996455157745, "grad_norm": 5.137234652611149, "learning_rate": 1.2060163921193774e-06, "loss": 0.4378, "step": 11016 }, { "epoch": 0.7810705423608649, "grad_norm": 1.8771978212153688, "learning_rate": 1.2052686646048967e-06, "loss": 0.4053, "step": 11017 }, { "epoch": 0.7811414392059554, "grad_norm": 5.333480212661932, "learning_rate": 1.2045211371894944e-06, "loss": 0.4467, "step": 11018 }, { "epoch": 0.7812123360510457, "grad_norm": 2.7246424429704073, "learning_rate": 1.203773809912589e-06, "loss": 0.3859, "step": 11019 }, { "epoch": 0.7812832328961361, "grad_norm": 3.4791586241239987, "learning_rate": 1.2030266828135895e-06, "loss": 0.3876, "step": 11020 }, { "epoch": 0.7813541297412265, "grad_norm": 3.7887132124769147, "learning_rate": 1.2022797559318915e-06, "loss": 0.3727, "step": 11021 }, { "epoch": 0.781425026586317, "grad_norm": 3.0109493971755454, "learning_rate": 1.2015330293068795e-06, "loss": 0.3967, "step": 11022 }, { "epoch": 0.7814959234314073, "grad_norm": 3.5785110702630587, "learning_rate": 1.2007865029779308e-06, "loss": 0.4036, "step": 11023 }, { "epoch": 0.7815668202764977, "grad_norm": 7.019602869143007, "learning_rate": 1.20004017698441e-06, "loss": 0.4145, "step": 11024 }, { "epoch": 0.7816377171215881, "grad_norm": 7.374720308337481, "learning_rate": 1.1992940513656699e-06, "loss": 0.3626, "step": 11025 }, { "epoch": 0.7817086139666785, "grad_norm": 4.966463421450765, "learning_rate": 1.198548126161056e-06, "loss": 0.4007, "step": 11026 }, { "epoch": 0.7817795108117689, "grad_norm": 3.567911483369872, "learning_rate": 1.1978024014099e-06, "loss": 0.4238, "step": 11027 }, { "epoch": 0.7818504076568593, "grad_norm": 2.276246153399918, "learning_rate": 1.1970568771515267e-06, "loss": 0.4111, "step": 11028 }, { "epoch": 0.7819213045019496, "grad_norm": 3.26708984566627, "learning_rate": 1.196311553425246e-06, "loss": 0.3926, "step": 11029 }, { "epoch": 0.7819922013470401, "grad_norm": 6.15086807995158, "learning_rate": 1.1955664302703613e-06, "loss": 0.4165, "step": 11030 }, { "epoch": 0.7820630981921305, "grad_norm": 3.8350474119208178, "learning_rate": 1.194821507726161e-06, "loss": 0.3671, "step": 11031 }, { "epoch": 0.7821339950372208, "grad_norm": 2.8736669322964628, "learning_rate": 1.1940767858319292e-06, "loss": 0.3499, "step": 11032 }, { "epoch": 0.7822048918823112, "grad_norm": 6.612799972325301, "learning_rate": 1.1933322646269325e-06, "loss": 0.3835, "step": 11033 }, { "epoch": 0.7822757887274017, "grad_norm": 4.60329560058887, "learning_rate": 1.1925879441504302e-06, "loss": 0.4116, "step": 11034 }, { "epoch": 0.782346685572492, "grad_norm": 6.120223276091013, "learning_rate": 1.1918438244416725e-06, "loss": 0.3572, "step": 11035 }, { "epoch": 0.7824175824175824, "grad_norm": 3.3208781393576494, "learning_rate": 1.191099905539897e-06, "loss": 0.3714, "step": 11036 }, { "epoch": 0.7824884792626728, "grad_norm": 3.489438434654944, "learning_rate": 1.1903561874843288e-06, "loss": 0.403, "step": 11037 }, { "epoch": 0.7825593761077632, "grad_norm": 3.8245184067567246, "learning_rate": 1.189612670314187e-06, "loss": 0.4205, "step": 11038 }, { "epoch": 0.7826302729528536, "grad_norm": 2.835867897281602, "learning_rate": 1.1888693540686785e-06, "loss": 0.4074, "step": 11039 }, { "epoch": 0.782701169797944, "grad_norm": 3.8705717302295755, "learning_rate": 1.1881262387869984e-06, "loss": 0.379, "step": 11040 }, { "epoch": 0.7827720666430343, "grad_norm": 7.414377490216134, "learning_rate": 1.1873833245083294e-06, "loss": 0.4051, "step": 11041 }, { "epoch": 0.7828429634881248, "grad_norm": 2.1249010126304473, "learning_rate": 1.1866406112718476e-06, "loss": 0.4417, "step": 11042 }, { "epoch": 0.7829138603332152, "grad_norm": 6.898157643867465, "learning_rate": 1.1858980991167202e-06, "loss": 0.3552, "step": 11043 }, { "epoch": 0.7829847571783055, "grad_norm": 2.667713655054967, "learning_rate": 1.185155788082094e-06, "loss": 0.3695, "step": 11044 }, { "epoch": 0.7830556540233959, "grad_norm": 2.6272324236786506, "learning_rate": 1.1844136782071153e-06, "loss": 0.4229, "step": 11045 }, { "epoch": 0.7831265508684864, "grad_norm": 2.4943382959424243, "learning_rate": 1.1836717695309168e-06, "loss": 0.4251, "step": 11046 }, { "epoch": 0.7831974477135768, "grad_norm": 24.707783777134107, "learning_rate": 1.1829300620926187e-06, "loss": 0.3833, "step": 11047 }, { "epoch": 0.7832683445586671, "grad_norm": 2.9869775588453575, "learning_rate": 1.1821885559313307e-06, "loss": 0.4069, "step": 11048 }, { "epoch": 0.7833392414037575, "grad_norm": 21.413066771212943, "learning_rate": 1.1814472510861551e-06, "loss": 0.3301, "step": 11049 }, { "epoch": 0.783410138248848, "grad_norm": 2.488151638398353, "learning_rate": 1.1807061475961811e-06, "loss": 0.394, "step": 11050 }, { "epoch": 0.7834810350939383, "grad_norm": 3.00126224106497, "learning_rate": 1.179965245500485e-06, "loss": 0.3991, "step": 11051 }, { "epoch": 0.7835519319390287, "grad_norm": 1.9434035833702747, "learning_rate": 1.179224544838139e-06, "loss": 0.394, "step": 11052 }, { "epoch": 0.7836228287841192, "grad_norm": 5.873213421445933, "learning_rate": 1.1784840456481977e-06, "loss": 0.3607, "step": 11053 }, { "epoch": 0.7836937256292095, "grad_norm": 4.217062849183596, "learning_rate": 1.1777437479697108e-06, "loss": 0.4637, "step": 11054 }, { "epoch": 0.7837646224742999, "grad_norm": 4.607263042468937, "learning_rate": 1.1770036518417133e-06, "loss": 0.3972, "step": 11055 }, { "epoch": 0.7838355193193903, "grad_norm": 3.663317094409722, "learning_rate": 1.1762637573032304e-06, "loss": 0.4357, "step": 11056 }, { "epoch": 0.7839064161644806, "grad_norm": 2.8947619476136803, "learning_rate": 1.1755240643932775e-06, "loss": 0.3511, "step": 11057 }, { "epoch": 0.7839773130095711, "grad_norm": 2.2325768525045966, "learning_rate": 1.1747845731508612e-06, "loss": 0.3697, "step": 11058 }, { "epoch": 0.7840482098546615, "grad_norm": 2.2823692173227026, "learning_rate": 1.174045283614974e-06, "loss": 0.3108, "step": 11059 }, { "epoch": 0.7841191066997518, "grad_norm": 1.9572852441267812, "learning_rate": 1.1733061958245984e-06, "loss": 0.4163, "step": 11060 }, { "epoch": 0.7841900035448423, "grad_norm": 2.8638082302807635, "learning_rate": 1.1725673098187085e-06, "loss": 0.4213, "step": 11061 }, { "epoch": 0.7842609003899327, "grad_norm": 5.234724629333447, "learning_rate": 1.171828625636266e-06, "loss": 0.4135, "step": 11062 }, { "epoch": 0.784331797235023, "grad_norm": 2.491383159763119, "learning_rate": 1.171090143316221e-06, "loss": 0.4001, "step": 11063 }, { "epoch": 0.7844026940801134, "grad_norm": 9.291813087014976, "learning_rate": 1.1703518628975148e-06, "loss": 0.3692, "step": 11064 }, { "epoch": 0.7844735909252039, "grad_norm": 2.447928344599975, "learning_rate": 1.169613784419079e-06, "loss": 0.3626, "step": 11065 }, { "epoch": 0.7845444877702942, "grad_norm": 2.533305425098833, "learning_rate": 1.1688759079198326e-06, "loss": 0.3448, "step": 11066 }, { "epoch": 0.7846153846153846, "grad_norm": 2.16717161390687, "learning_rate": 1.1681382334386821e-06, "loss": 0.3967, "step": 11067 }, { "epoch": 0.784686281460475, "grad_norm": 2.553099193006476, "learning_rate": 1.1674007610145282e-06, "loss": 0.4549, "step": 11068 }, { "epoch": 0.7847571783055654, "grad_norm": 3.3739817194711885, "learning_rate": 1.166663490686258e-06, "loss": 0.3811, "step": 11069 }, { "epoch": 0.7848280751506558, "grad_norm": 2.8779934828321307, "learning_rate": 1.1659264224927464e-06, "loss": 0.3707, "step": 11070 }, { "epoch": 0.7848989719957462, "grad_norm": 4.955755601712267, "learning_rate": 1.1651895564728605e-06, "loss": 0.3584, "step": 11071 }, { "epoch": 0.7849698688408366, "grad_norm": 3.6231611105072385, "learning_rate": 1.164452892665458e-06, "loss": 0.4093, "step": 11072 }, { "epoch": 0.785040765685927, "grad_norm": 1.910360669581855, "learning_rate": 1.1637164311093824e-06, "loss": 0.3711, "step": 11073 }, { "epoch": 0.7851116625310174, "grad_norm": 5.993890556807313, "learning_rate": 1.1629801718434653e-06, "loss": 0.3833, "step": 11074 }, { "epoch": 0.7851825593761078, "grad_norm": 5.132895225384651, "learning_rate": 1.1622441149065345e-06, "loss": 0.4207, "step": 11075 }, { "epoch": 0.7852534562211981, "grad_norm": 9.534016084432226, "learning_rate": 1.161508260337399e-06, "loss": 0.3695, "step": 11076 }, { "epoch": 0.7853243530662886, "grad_norm": 2.862593588247629, "learning_rate": 1.1607726081748637e-06, "loss": 0.3788, "step": 11077 }, { "epoch": 0.785395249911379, "grad_norm": 2.2157305716883196, "learning_rate": 1.1600371584577197e-06, "loss": 0.408, "step": 11078 }, { "epoch": 0.7854661467564693, "grad_norm": 6.5695417537966865, "learning_rate": 1.1593019112247461e-06, "loss": 0.3321, "step": 11079 }, { "epoch": 0.7855370436015597, "grad_norm": 5.027486484895746, "learning_rate": 1.158566866514715e-06, "loss": 0.3799, "step": 11080 }, { "epoch": 0.7856079404466502, "grad_norm": 2.1854161598552952, "learning_rate": 1.1578320243663854e-06, "loss": 0.3801, "step": 11081 }, { "epoch": 0.7856788372917405, "grad_norm": 4.518759281425999, "learning_rate": 1.1570973848185042e-06, "loss": 0.3658, "step": 11082 }, { "epoch": 0.7857497341368309, "grad_norm": 3.499581481624851, "learning_rate": 1.156362947909811e-06, "loss": 0.3979, "step": 11083 }, { "epoch": 0.7858206309819213, "grad_norm": 3.1893877250129377, "learning_rate": 1.1556287136790344e-06, "loss": 0.4038, "step": 11084 }, { "epoch": 0.7858915278270117, "grad_norm": 25.98516768783144, "learning_rate": 1.1548946821648903e-06, "loss": 0.3998, "step": 11085 }, { "epoch": 0.7859624246721021, "grad_norm": 4.380010288477939, "learning_rate": 1.1541608534060833e-06, "loss": 0.361, "step": 11086 }, { "epoch": 0.7860333215171925, "grad_norm": 3.4128250491547547, "learning_rate": 1.15342722744131e-06, "loss": 0.3766, "step": 11087 }, { "epoch": 0.7861042183622828, "grad_norm": 10.577895791117657, "learning_rate": 1.1526938043092572e-06, "loss": 0.4128, "step": 11088 }, { "epoch": 0.7861751152073733, "grad_norm": 10.939055789075, "learning_rate": 1.1519605840485942e-06, "loss": 0.4235, "step": 11089 }, { "epoch": 0.7862460120524637, "grad_norm": 8.988318177665809, "learning_rate": 1.1512275666979866e-06, "loss": 0.3893, "step": 11090 }, { "epoch": 0.786316908897554, "grad_norm": 4.11066505301219, "learning_rate": 1.150494752296088e-06, "loss": 0.3554, "step": 11091 }, { "epoch": 0.7863878057426444, "grad_norm": 0.9240816603693497, "learning_rate": 1.1497621408815397e-06, "loss": 0.6737, "step": 11092 }, { "epoch": 0.7864587025877349, "grad_norm": 1.1913999325844868, "learning_rate": 1.149029732492971e-06, "loss": 0.6462, "step": 11093 }, { "epoch": 0.7865295994328252, "grad_norm": 5.248555393611991, "learning_rate": 1.1482975271690056e-06, "loss": 0.4442, "step": 11094 }, { "epoch": 0.7866004962779156, "grad_norm": 1.133320396154212, "learning_rate": 1.1475655249482497e-06, "loss": 0.6252, "step": 11095 }, { "epoch": 0.786671393123006, "grad_norm": 1.8329264198714954, "learning_rate": 1.1468337258693053e-06, "loss": 0.4043, "step": 11096 }, { "epoch": 0.7867422899680964, "grad_norm": 3.9314617281605084, "learning_rate": 1.1461021299707598e-06, "loss": 0.3456, "step": 11097 }, { "epoch": 0.7868131868131868, "grad_norm": 2.509905091944428, "learning_rate": 1.1453707372911893e-06, "loss": 0.3733, "step": 11098 }, { "epoch": 0.7868840836582772, "grad_norm": 8.569278036626692, "learning_rate": 1.144639547869163e-06, "loss": 0.4004, "step": 11099 }, { "epoch": 0.7869549805033677, "grad_norm": 3.3272447067704465, "learning_rate": 1.1439085617432344e-06, "loss": 0.3676, "step": 11100 }, { "epoch": 0.787025877348458, "grad_norm": 2.610605460778413, "learning_rate": 1.1431777789519528e-06, "loss": 0.3368, "step": 11101 }, { "epoch": 0.7870967741935484, "grad_norm": 3.074672710915694, "learning_rate": 1.1424471995338487e-06, "loss": 0.4135, "step": 11102 }, { "epoch": 0.7871676710386388, "grad_norm": 3.0574408612598165, "learning_rate": 1.1417168235274495e-06, "loss": 0.3684, "step": 11103 }, { "epoch": 0.7872385678837291, "grad_norm": 2.3993614554614333, "learning_rate": 1.1409866509712669e-06, "loss": 0.3768, "step": 11104 }, { "epoch": 0.7873094647288196, "grad_norm": 2.934490724925577, "learning_rate": 1.1402566819038024e-06, "loss": 0.4215, "step": 11105 }, { "epoch": 0.78738036157391, "grad_norm": 6.554640659251424, "learning_rate": 1.1395269163635492e-06, "loss": 0.4081, "step": 11106 }, { "epoch": 0.7874512584190003, "grad_norm": 2.7831600347054772, "learning_rate": 1.1387973543889913e-06, "loss": 0.3476, "step": 11107 }, { "epoch": 0.7875221552640908, "grad_norm": 3.50805276598759, "learning_rate": 1.1380679960185931e-06, "loss": 0.3827, "step": 11108 }, { "epoch": 0.7875930521091812, "grad_norm": 9.743301476570629, "learning_rate": 1.137338841290817e-06, "loss": 0.4244, "step": 11109 }, { "epoch": 0.7876639489542715, "grad_norm": 2.0177340480554777, "learning_rate": 1.1366098902441131e-06, "loss": 0.3372, "step": 11110 }, { "epoch": 0.7877348457993619, "grad_norm": 1.9264790801486136, "learning_rate": 1.1358811429169193e-06, "loss": 0.3574, "step": 11111 }, { "epoch": 0.7878057426444524, "grad_norm": 2.2162106065615217, "learning_rate": 1.1351525993476599e-06, "loss": 0.4011, "step": 11112 }, { "epoch": 0.7878766394895427, "grad_norm": 4.554649913852681, "learning_rate": 1.1344242595747556e-06, "loss": 0.3705, "step": 11113 }, { "epoch": 0.7879475363346331, "grad_norm": 22.058572281141004, "learning_rate": 1.1336961236366105e-06, "loss": 0.3726, "step": 11114 }, { "epoch": 0.7880184331797235, "grad_norm": 4.06150882163595, "learning_rate": 1.132968191571619e-06, "loss": 0.4074, "step": 11115 }, { "epoch": 0.7880893300248138, "grad_norm": 2.9644114539661484, "learning_rate": 1.1322404634181654e-06, "loss": 0.3782, "step": 11116 }, { "epoch": 0.7881602268699043, "grad_norm": 2.2461976789329072, "learning_rate": 1.1315129392146262e-06, "loss": 0.3951, "step": 11117 }, { "epoch": 0.7882311237149947, "grad_norm": 2.4157445871119574, "learning_rate": 1.1307856189993627e-06, "loss": 0.4279, "step": 11118 }, { "epoch": 0.788302020560085, "grad_norm": 5.918508260982466, "learning_rate": 1.1300585028107247e-06, "loss": 0.4125, "step": 11119 }, { "epoch": 0.7883729174051755, "grad_norm": 2.0829862801765318, "learning_rate": 1.1293315906870572e-06, "loss": 0.3703, "step": 11120 }, { "epoch": 0.7884438142502659, "grad_norm": 7.6446328319536265, "learning_rate": 1.1286048826666884e-06, "loss": 0.4007, "step": 11121 }, { "epoch": 0.7885147110953562, "grad_norm": 2.1762898897248633, "learning_rate": 1.1278783787879399e-06, "loss": 0.3463, "step": 11122 }, { "epoch": 0.7885856079404466, "grad_norm": 2.458743243861403, "learning_rate": 1.12715207908912e-06, "loss": 0.4085, "step": 11123 }, { "epoch": 0.7886565047855371, "grad_norm": 3.1403169874825116, "learning_rate": 1.1264259836085257e-06, "loss": 0.3435, "step": 11124 }, { "epoch": 0.7887274016306275, "grad_norm": 1.9434744921577485, "learning_rate": 1.1257000923844475e-06, "loss": 0.3711, "step": 11125 }, { "epoch": 0.7887982984757178, "grad_norm": 1.716170167195514, "learning_rate": 1.1249744054551598e-06, "loss": 0.3565, "step": 11126 }, { "epoch": 0.7888691953208082, "grad_norm": 3.783394424150523, "learning_rate": 1.1242489228589287e-06, "loss": 0.3804, "step": 11127 }, { "epoch": 0.7889400921658987, "grad_norm": 2.495765229452712, "learning_rate": 1.1235236446340098e-06, "loss": 0.3976, "step": 11128 }, { "epoch": 0.789010989010989, "grad_norm": 1.9146156107733454, "learning_rate": 1.1227985708186495e-06, "loss": 0.3587, "step": 11129 }, { "epoch": 0.7890818858560794, "grad_norm": 2.5632997205661696, "learning_rate": 1.1220737014510796e-06, "loss": 0.383, "step": 11130 }, { "epoch": 0.7891527827011698, "grad_norm": 2.764589242993444, "learning_rate": 1.1213490365695218e-06, "loss": 0.3775, "step": 11131 }, { "epoch": 0.7892236795462602, "grad_norm": 2.564604350545542, "learning_rate": 1.1206245762121903e-06, "loss": 0.404, "step": 11132 }, { "epoch": 0.7892945763913506, "grad_norm": 6.5813511206450475, "learning_rate": 1.1199003204172882e-06, "loss": 0.4309, "step": 11133 }, { "epoch": 0.789365473236441, "grad_norm": 2.698939236460957, "learning_rate": 1.1191762692230013e-06, "loss": 0.3515, "step": 11134 }, { "epoch": 0.7894363700815313, "grad_norm": 4.0449523407984795, "learning_rate": 1.1184524226675115e-06, "loss": 0.3023, "step": 11135 }, { "epoch": 0.7895072669266218, "grad_norm": 8.83043487329725, "learning_rate": 1.1177287807889897e-06, "loss": 0.399, "step": 11136 }, { "epoch": 0.7895781637717122, "grad_norm": 5.697680991994047, "learning_rate": 1.1170053436255923e-06, "loss": 0.3778, "step": 11137 }, { "epoch": 0.7896490606168025, "grad_norm": 3.9032354857598133, "learning_rate": 1.116282111215466e-06, "loss": 0.3728, "step": 11138 }, { "epoch": 0.7897199574618929, "grad_norm": 1.9615105962455741, "learning_rate": 1.115559083596749e-06, "loss": 0.3699, "step": 11139 }, { "epoch": 0.7897908543069834, "grad_norm": 3.2200528438042917, "learning_rate": 1.1148362608075652e-06, "loss": 0.4024, "step": 11140 }, { "epoch": 0.7898617511520737, "grad_norm": 3.6957438063902983, "learning_rate": 1.1141136428860322e-06, "loss": 0.4008, "step": 11141 }, { "epoch": 0.7899326479971641, "grad_norm": 2.46400514044617, "learning_rate": 1.1133912298702531e-06, "loss": 0.3937, "step": 11142 }, { "epoch": 0.7900035448422545, "grad_norm": 3.979654231149216, "learning_rate": 1.1126690217983194e-06, "loss": 0.4001, "step": 11143 }, { "epoch": 0.7900744416873449, "grad_norm": 3.382229049997501, "learning_rate": 1.1119470187083165e-06, "loss": 0.4083, "step": 11144 }, { "epoch": 0.7901453385324353, "grad_norm": 2.010075291048208, "learning_rate": 1.1112252206383135e-06, "loss": 0.3479, "step": 11145 }, { "epoch": 0.7902162353775257, "grad_norm": 2.3870593013887387, "learning_rate": 1.1105036276263742e-06, "loss": 0.3708, "step": 11146 }, { "epoch": 0.790287132222616, "grad_norm": 2.5141532276984577, "learning_rate": 1.1097822397105463e-06, "loss": 0.3534, "step": 11147 }, { "epoch": 0.7903580290677065, "grad_norm": 5.07055866886595, "learning_rate": 1.1090610569288712e-06, "loss": 0.3928, "step": 11148 }, { "epoch": 0.7904289259127969, "grad_norm": 6.8045372688053565, "learning_rate": 1.1083400793193766e-06, "loss": 0.4158, "step": 11149 }, { "epoch": 0.7904998227578873, "grad_norm": 3.371798115594589, "learning_rate": 1.107619306920078e-06, "loss": 0.3727, "step": 11150 }, { "epoch": 0.7905707196029776, "grad_norm": 2.753679706742842, "learning_rate": 1.1068987397689851e-06, "loss": 0.3556, "step": 11151 }, { "epoch": 0.7906416164480681, "grad_norm": 2.6347420001297603, "learning_rate": 1.1061783779040951e-06, "loss": 0.4017, "step": 11152 }, { "epoch": 0.7907125132931585, "grad_norm": 3.9634046299739567, "learning_rate": 1.105458221363389e-06, "loss": 0.3544, "step": 11153 }, { "epoch": 0.7907834101382488, "grad_norm": 5.663845782031348, "learning_rate": 1.1047382701848437e-06, "loss": 0.3967, "step": 11154 }, { "epoch": 0.7908543069833393, "grad_norm": 2.0511908080987586, "learning_rate": 1.1040185244064234e-06, "loss": 0.3496, "step": 11155 }, { "epoch": 0.7909252038284297, "grad_norm": 5.024279472191045, "learning_rate": 1.1032989840660802e-06, "loss": 0.3977, "step": 11156 }, { "epoch": 0.79099610067352, "grad_norm": 9.581292989249796, "learning_rate": 1.1025796492017544e-06, "loss": 0.4115, "step": 11157 }, { "epoch": 0.7910669975186104, "grad_norm": 1.891651808449297, "learning_rate": 1.1018605198513793e-06, "loss": 0.3751, "step": 11158 }, { "epoch": 0.7911378943637009, "grad_norm": 4.407495836791141, "learning_rate": 1.1011415960528737e-06, "loss": 0.4268, "step": 11159 }, { "epoch": 0.7912087912087912, "grad_norm": 2.4032724419034914, "learning_rate": 1.1004228778441483e-06, "loss": 0.4186, "step": 11160 }, { "epoch": 0.7912796880538816, "grad_norm": 3.1857942838938103, "learning_rate": 1.0997043652631002e-06, "loss": 0.3948, "step": 11161 }, { "epoch": 0.791350584898972, "grad_norm": 2.15905697079989, "learning_rate": 1.0989860583476192e-06, "loss": 0.3806, "step": 11162 }, { "epoch": 0.7914214817440623, "grad_norm": 6.085173592469963, "learning_rate": 1.0982679571355804e-06, "loss": 0.374, "step": 11163 }, { "epoch": 0.7914923785891528, "grad_norm": 2.2869946921225464, "learning_rate": 1.0975500616648494e-06, "loss": 0.375, "step": 11164 }, { "epoch": 0.7915632754342432, "grad_norm": 5.979656238321163, "learning_rate": 1.0968323719732832e-06, "loss": 0.4052, "step": 11165 }, { "epoch": 0.7916341722793335, "grad_norm": 2.5066065906544077, "learning_rate": 1.0961148880987244e-06, "loss": 0.3806, "step": 11166 }, { "epoch": 0.791705069124424, "grad_norm": 4.527636824046247, "learning_rate": 1.0953976100790082e-06, "loss": 0.3755, "step": 11167 }, { "epoch": 0.7917759659695144, "grad_norm": 3.080446212729223, "learning_rate": 1.0946805379519559e-06, "loss": 0.3931, "step": 11168 }, { "epoch": 0.7918468628146047, "grad_norm": 2.4888276866350063, "learning_rate": 1.0939636717553788e-06, "loss": 0.3983, "step": 11169 }, { "epoch": 0.7919177596596951, "grad_norm": 2.703738235753405, "learning_rate": 1.09324701152708e-06, "loss": 0.3896, "step": 11170 }, { "epoch": 0.7919886565047856, "grad_norm": 10.135431377087066, "learning_rate": 1.092530557304848e-06, "loss": 0.3542, "step": 11171 }, { "epoch": 0.7920595533498759, "grad_norm": 3.0356291979343815, "learning_rate": 1.091814309126461e-06, "loss": 0.4414, "step": 11172 }, { "epoch": 0.7921304501949663, "grad_norm": 4.471164422918901, "learning_rate": 1.0910982670296883e-06, "loss": 0.3762, "step": 11173 }, { "epoch": 0.7922013470400567, "grad_norm": 3.635407877157, "learning_rate": 1.0903824310522887e-06, "loss": 0.3762, "step": 11174 }, { "epoch": 0.7922722438851472, "grad_norm": 4.3502893502081905, "learning_rate": 1.0896668012320077e-06, "loss": 0.4304, "step": 11175 }, { "epoch": 0.7923431407302375, "grad_norm": 3.1358652414698964, "learning_rate": 1.0889513776065795e-06, "loss": 0.3755, "step": 11176 }, { "epoch": 0.7924140375753279, "grad_norm": 3.249128388674701, "learning_rate": 1.0882361602137304e-06, "loss": 0.3559, "step": 11177 }, { "epoch": 0.7924849344204183, "grad_norm": 4.558476932276919, "learning_rate": 1.0875211490911769e-06, "loss": 0.4319, "step": 11178 }, { "epoch": 0.7925558312655087, "grad_norm": 3.001528160785795, "learning_rate": 1.0868063442766169e-06, "loss": 0.3791, "step": 11179 }, { "epoch": 0.7926267281105991, "grad_norm": 4.211156848825953, "learning_rate": 1.0860917458077453e-06, "loss": 0.3696, "step": 11180 }, { "epoch": 0.7926976249556895, "grad_norm": 2.6354432450424725, "learning_rate": 1.0853773537222446e-06, "loss": 0.3538, "step": 11181 }, { "epoch": 0.7927685218007798, "grad_norm": 4.045185957430991, "learning_rate": 1.0846631680577835e-06, "loss": 0.3825, "step": 11182 }, { "epoch": 0.7928394186458703, "grad_norm": 1.931158995039543, "learning_rate": 1.083949188852021e-06, "loss": 0.4088, "step": 11183 }, { "epoch": 0.7929103154909607, "grad_norm": 2.147412818856256, "learning_rate": 1.0832354161426084e-06, "loss": 0.3613, "step": 11184 }, { "epoch": 0.792981212336051, "grad_norm": 49.6095542838911, "learning_rate": 1.0825218499671797e-06, "loss": 0.4263, "step": 11185 }, { "epoch": 0.7930521091811414, "grad_norm": 4.4168763238805075, "learning_rate": 1.0818084903633658e-06, "loss": 0.4086, "step": 11186 }, { "epoch": 0.7931230060262319, "grad_norm": 2.9390727163402413, "learning_rate": 1.081095337368781e-06, "loss": 0.353, "step": 11187 }, { "epoch": 0.7931939028713222, "grad_norm": 2.1366694970058, "learning_rate": 1.0803823910210287e-06, "loss": 0.3976, "step": 11188 }, { "epoch": 0.7932647997164126, "grad_norm": 9.42901137138157, "learning_rate": 1.0796696513577053e-06, "loss": 0.3865, "step": 11189 }, { "epoch": 0.793335696561503, "grad_norm": 2.026562816393432, "learning_rate": 1.0789571184163932e-06, "loss": 0.4274, "step": 11190 }, { "epoch": 0.7934065934065934, "grad_norm": 3.5361552132972673, "learning_rate": 1.0782447922346656e-06, "loss": 0.3489, "step": 11191 }, { "epoch": 0.7934774902516838, "grad_norm": 2.5392088660931575, "learning_rate": 1.0775326728500823e-06, "loss": 0.4041, "step": 11192 }, { "epoch": 0.7935483870967742, "grad_norm": 3.4324083939462913, "learning_rate": 1.0768207603001969e-06, "loss": 0.3687, "step": 11193 }, { "epoch": 0.7936192839418645, "grad_norm": 3.500158654814327, "learning_rate": 1.0761090546225467e-06, "loss": 0.3902, "step": 11194 }, { "epoch": 0.793690180786955, "grad_norm": 4.670010984933197, "learning_rate": 1.0753975558546597e-06, "loss": 0.392, "step": 11195 }, { "epoch": 0.7937610776320454, "grad_norm": 3.5880095932519884, "learning_rate": 1.074686264034055e-06, "loss": 0.4181, "step": 11196 }, { "epoch": 0.7938319744771357, "grad_norm": 7.518350001746747, "learning_rate": 1.0739751791982428e-06, "loss": 0.3622, "step": 11197 }, { "epoch": 0.7939028713222261, "grad_norm": 2.364175345536718, "learning_rate": 1.0732643013847128e-06, "loss": 0.3879, "step": 11198 }, { "epoch": 0.7939737681673166, "grad_norm": 4.645718980254178, "learning_rate": 1.0725536306309536e-06, "loss": 0.3552, "step": 11199 }, { "epoch": 0.794044665012407, "grad_norm": 2.6921062832993012, "learning_rate": 1.0718431669744406e-06, "loss": 0.3443, "step": 11200 }, { "epoch": 0.7941155618574973, "grad_norm": 3.409138919292584, "learning_rate": 1.0711329104526353e-06, "loss": 0.4842, "step": 11201 }, { "epoch": 0.7941864587025877, "grad_norm": 3.5426630696561183, "learning_rate": 1.0704228611029892e-06, "loss": 0.4024, "step": 11202 }, { "epoch": 0.7942573555476782, "grad_norm": 1.9893676341810123, "learning_rate": 1.069713018962945e-06, "loss": 0.3337, "step": 11203 }, { "epoch": 0.7943282523927685, "grad_norm": 12.24370496720531, "learning_rate": 1.0690033840699343e-06, "loss": 0.3979, "step": 11204 }, { "epoch": 0.7943991492378589, "grad_norm": 5.056182533043505, "learning_rate": 1.068293956461376e-06, "loss": 0.4504, "step": 11205 }, { "epoch": 0.7944700460829494, "grad_norm": 2.3026300298372915, "learning_rate": 1.0675847361746767e-06, "loss": 0.3908, "step": 11206 }, { "epoch": 0.7945409429280397, "grad_norm": 2.136249782411877, "learning_rate": 1.066875723247237e-06, "loss": 0.4078, "step": 11207 }, { "epoch": 0.7946118397731301, "grad_norm": 3.071668683260851, "learning_rate": 1.0661669177164425e-06, "loss": 0.3568, "step": 11208 }, { "epoch": 0.7946827366182205, "grad_norm": 2.8882260837487066, "learning_rate": 1.0654583196196678e-06, "loss": 0.4037, "step": 11209 }, { "epoch": 0.7947536334633108, "grad_norm": 2.545512536975335, "learning_rate": 1.0647499289942797e-06, "loss": 0.4021, "step": 11210 }, { "epoch": 0.7948245303084013, "grad_norm": 1.7465725533709462, "learning_rate": 1.064041745877631e-06, "loss": 0.3756, "step": 11211 }, { "epoch": 0.7948954271534917, "grad_norm": 3.695199234026198, "learning_rate": 1.0633337703070661e-06, "loss": 0.3973, "step": 11212 }, { "epoch": 0.794966323998582, "grad_norm": 5.632008305950739, "learning_rate": 1.0626260023199164e-06, "loss": 0.3573, "step": 11213 }, { "epoch": 0.7950372208436725, "grad_norm": 2.642223841723704, "learning_rate": 1.0619184419535012e-06, "loss": 0.3703, "step": 11214 }, { "epoch": 0.7951081176887629, "grad_norm": 2.4440984773308636, "learning_rate": 1.0612110892451328e-06, "loss": 0.3727, "step": 11215 }, { "epoch": 0.7951790145338532, "grad_norm": 8.260647935954204, "learning_rate": 1.0605039442321124e-06, "loss": 0.3987, "step": 11216 }, { "epoch": 0.7952499113789436, "grad_norm": 1.9713291797050583, "learning_rate": 1.0597970069517232e-06, "loss": 0.3643, "step": 11217 }, { "epoch": 0.7953208082240341, "grad_norm": 4.155367326083204, "learning_rate": 1.0590902774412454e-06, "loss": 0.3813, "step": 11218 }, { "epoch": 0.7953917050691244, "grad_norm": 2.1411644639599485, "learning_rate": 1.058383755737946e-06, "loss": 0.4314, "step": 11219 }, { "epoch": 0.7954626019142148, "grad_norm": 2.154268290575518, "learning_rate": 1.057677441879081e-06, "loss": 0.3889, "step": 11220 }, { "epoch": 0.7955334987593052, "grad_norm": 3.3063162969723088, "learning_rate": 1.0569713359018907e-06, "loss": 0.3701, "step": 11221 }, { "epoch": 0.7956043956043956, "grad_norm": 9.704907812191212, "learning_rate": 1.056265437843612e-06, "loss": 0.3535, "step": 11222 }, { "epoch": 0.795675292449486, "grad_norm": 2.9371369096113598, "learning_rate": 1.0555597477414697e-06, "loss": 0.3833, "step": 11223 }, { "epoch": 0.7957461892945764, "grad_norm": 4.502209341160603, "learning_rate": 1.05485426563267e-06, "loss": 0.3776, "step": 11224 }, { "epoch": 0.7958170861396667, "grad_norm": 3.1823986366673203, "learning_rate": 1.0541489915544162e-06, "loss": 0.3961, "step": 11225 }, { "epoch": 0.7958879829847572, "grad_norm": 1.8988755472411956, "learning_rate": 1.0534439255438988e-06, "loss": 0.4092, "step": 11226 }, { "epoch": 0.7959588798298476, "grad_norm": 1.0420554525801877, "learning_rate": 1.0527390676382953e-06, "loss": 0.6486, "step": 11227 }, { "epoch": 0.796029776674938, "grad_norm": 2.1778899737767947, "learning_rate": 1.0520344178747727e-06, "loss": 0.3574, "step": 11228 }, { "epoch": 0.7961006735200283, "grad_norm": 3.0537022534822795, "learning_rate": 1.0513299762904894e-06, "loss": 0.405, "step": 11229 }, { "epoch": 0.7961715703651188, "grad_norm": 6.141510700209365, "learning_rate": 1.050625742922589e-06, "loss": 0.4041, "step": 11230 }, { "epoch": 0.7962424672102092, "grad_norm": 1.9909761620780726, "learning_rate": 1.0499217178082094e-06, "loss": 0.3508, "step": 11231 }, { "epoch": 0.7963133640552995, "grad_norm": 3.7669384917901643, "learning_rate": 1.0492179009844723e-06, "loss": 0.3225, "step": 11232 }, { "epoch": 0.7963842609003899, "grad_norm": 1.9728531847717672, "learning_rate": 1.0485142924884894e-06, "loss": 0.3484, "step": 11233 }, { "epoch": 0.7964551577454804, "grad_norm": 3.797169768310203, "learning_rate": 1.0478108923573654e-06, "loss": 0.4136, "step": 11234 }, { "epoch": 0.7965260545905707, "grad_norm": 2.121460949533898, "learning_rate": 1.0471077006281883e-06, "loss": 0.3633, "step": 11235 }, { "epoch": 0.7965969514356611, "grad_norm": 3.7950911431390506, "learning_rate": 1.0464047173380404e-06, "loss": 0.4478, "step": 11236 }, { "epoch": 0.7966678482807515, "grad_norm": 4.799701949458896, "learning_rate": 1.0457019425239884e-06, "loss": 0.4449, "step": 11237 }, { "epoch": 0.7967387451258419, "grad_norm": 3.0847993353051804, "learning_rate": 1.044999376223092e-06, "loss": 0.43, "step": 11238 }, { "epoch": 0.7968096419709323, "grad_norm": 4.126384817767868, "learning_rate": 1.0442970184723978e-06, "loss": 0.4295, "step": 11239 }, { "epoch": 0.7968805388160227, "grad_norm": 3.7773830062442135, "learning_rate": 1.0435948693089394e-06, "loss": 0.4273, "step": 11240 }, { "epoch": 0.796951435661113, "grad_norm": 2.584334770698197, "learning_rate": 1.042892928769743e-06, "loss": 0.3708, "step": 11241 }, { "epoch": 0.7970223325062035, "grad_norm": 3.6870829369476223, "learning_rate": 1.0421911968918257e-06, "loss": 0.3961, "step": 11242 }, { "epoch": 0.7970932293512939, "grad_norm": 3.3279046561799785, "learning_rate": 1.041489673712185e-06, "loss": 0.4182, "step": 11243 }, { "epoch": 0.7971641261963842, "grad_norm": 2.7695842856580852, "learning_rate": 1.0407883592678154e-06, "loss": 0.3564, "step": 11244 }, { "epoch": 0.7972350230414746, "grad_norm": 8.841467766280942, "learning_rate": 1.0400872535956986e-06, "loss": 0.3726, "step": 11245 }, { "epoch": 0.7973059198865651, "grad_norm": 1.9440100705552092, "learning_rate": 1.0393863567328034e-06, "loss": 0.3876, "step": 11246 }, { "epoch": 0.7973768167316554, "grad_norm": 5.102598434032156, "learning_rate": 1.0386856687160867e-06, "loss": 0.4011, "step": 11247 }, { "epoch": 0.7974477135767458, "grad_norm": 3.6246977256989936, "learning_rate": 1.0379851895824989e-06, "loss": 0.3698, "step": 11248 }, { "epoch": 0.7975186104218362, "grad_norm": 2.7959797044978876, "learning_rate": 1.0372849193689771e-06, "loss": 0.396, "step": 11249 }, { "epoch": 0.7975895072669266, "grad_norm": 3.731323809834239, "learning_rate": 1.0365848581124467e-06, "loss": 0.3691, "step": 11250 }, { "epoch": 0.797660404112017, "grad_norm": 4.774076006847411, "learning_rate": 1.03588500584982e-06, "loss": 0.4379, "step": 11251 }, { "epoch": 0.7977313009571074, "grad_norm": 2.7758627695094344, "learning_rate": 1.0351853626180036e-06, "loss": 0.4177, "step": 11252 }, { "epoch": 0.7978021978021979, "grad_norm": 2.53876016067792, "learning_rate": 1.03448592845389e-06, "loss": 0.4084, "step": 11253 }, { "epoch": 0.7978730946472882, "grad_norm": 6.7038035536061615, "learning_rate": 1.0337867033943583e-06, "loss": 0.3884, "step": 11254 }, { "epoch": 0.7979439914923786, "grad_norm": 7.176243573359601, "learning_rate": 1.0330876874762824e-06, "loss": 0.4192, "step": 11255 }, { "epoch": 0.798014888337469, "grad_norm": 2.1565788224753186, "learning_rate": 1.0323888807365195e-06, "loss": 0.3942, "step": 11256 }, { "epoch": 0.7980857851825593, "grad_norm": 2.6544972865454297, "learning_rate": 1.0316902832119202e-06, "loss": 0.4027, "step": 11257 }, { "epoch": 0.7981566820276498, "grad_norm": 36.31182808482535, "learning_rate": 1.0309918949393216e-06, "loss": 0.4149, "step": 11258 }, { "epoch": 0.7982275788727402, "grad_norm": 4.57640661930184, "learning_rate": 1.0302937159555481e-06, "loss": 0.4134, "step": 11259 }, { "epoch": 0.7982984757178305, "grad_norm": 3.6564864079028054, "learning_rate": 1.029595746297417e-06, "loss": 0.3635, "step": 11260 }, { "epoch": 0.798369372562921, "grad_norm": 3.1423112296005167, "learning_rate": 1.0288979860017357e-06, "loss": 0.4085, "step": 11261 }, { "epoch": 0.7984402694080114, "grad_norm": 2.0726769155661207, "learning_rate": 1.0282004351052916e-06, "loss": 0.3666, "step": 11262 }, { "epoch": 0.7985111662531017, "grad_norm": 5.453058909899725, "learning_rate": 1.02750309364487e-06, "loss": 0.3787, "step": 11263 }, { "epoch": 0.7985820630981921, "grad_norm": 13.929284659418416, "learning_rate": 1.0268059616572435e-06, "loss": 0.4192, "step": 11264 }, { "epoch": 0.7986529599432826, "grad_norm": 2.9560306914336314, "learning_rate": 1.0261090391791717e-06, "loss": 0.4038, "step": 11265 }, { "epoch": 0.7987238567883729, "grad_norm": 2.217068731425085, "learning_rate": 1.025412326247402e-06, "loss": 0.3606, "step": 11266 }, { "epoch": 0.7987947536334633, "grad_norm": 2.6837932434045513, "learning_rate": 1.0247158228986737e-06, "loss": 0.4129, "step": 11267 }, { "epoch": 0.7988656504785537, "grad_norm": 2.3015734634129164, "learning_rate": 1.0240195291697158e-06, "loss": 0.3514, "step": 11268 }, { "epoch": 0.798936547323644, "grad_norm": 6.401970576108142, "learning_rate": 1.0233234450972428e-06, "loss": 0.403, "step": 11269 }, { "epoch": 0.7990074441687345, "grad_norm": 2.4140676368255973, "learning_rate": 1.0226275707179583e-06, "loss": 0.405, "step": 11270 }, { "epoch": 0.7990783410138249, "grad_norm": 3.4410879718608425, "learning_rate": 1.0219319060685595e-06, "loss": 0.3677, "step": 11271 }, { "epoch": 0.7991492378589152, "grad_norm": 2.712618511159078, "learning_rate": 1.0212364511857275e-06, "loss": 0.4013, "step": 11272 }, { "epoch": 0.7992201347040057, "grad_norm": 2.8723012956661487, "learning_rate": 1.0205412061061326e-06, "loss": 0.4026, "step": 11273 }, { "epoch": 0.7992910315490961, "grad_norm": 5.222760881090595, "learning_rate": 1.0198461708664393e-06, "loss": 0.415, "step": 11274 }, { "epoch": 0.7993619283941864, "grad_norm": 2.68391289218665, "learning_rate": 1.0191513455032931e-06, "loss": 0.4183, "step": 11275 }, { "epoch": 0.7994328252392768, "grad_norm": 1.9917850408667293, "learning_rate": 1.018456730053337e-06, "loss": 0.3663, "step": 11276 }, { "epoch": 0.7995037220843673, "grad_norm": 3.4729636995996294, "learning_rate": 1.0177623245531948e-06, "loss": 0.4003, "step": 11277 }, { "epoch": 0.7995746189294577, "grad_norm": 2.334545618351202, "learning_rate": 1.0170681290394862e-06, "loss": 0.3441, "step": 11278 }, { "epoch": 0.799645515774548, "grad_norm": 2.6078465888975813, "learning_rate": 1.0163741435488151e-06, "loss": 0.3787, "step": 11279 }, { "epoch": 0.7997164126196384, "grad_norm": 2.7662954397862483, "learning_rate": 1.0156803681177746e-06, "loss": 0.4046, "step": 11280 }, { "epoch": 0.7997873094647289, "grad_norm": 1.8799670149756746, "learning_rate": 1.0149868027829507e-06, "loss": 0.3342, "step": 11281 }, { "epoch": 0.7998582063098192, "grad_norm": 3.028592120013366, "learning_rate": 1.014293447580913e-06, "loss": 0.3717, "step": 11282 }, { "epoch": 0.7999291031549096, "grad_norm": 3.2852569469355846, "learning_rate": 1.0136003025482255e-06, "loss": 0.3852, "step": 11283 }, { "epoch": 0.8, "grad_norm": 0.9711518793216811, "learning_rate": 1.0129073677214362e-06, "loss": 0.6583, "step": 11284 }, { "epoch": 0.8000708968450904, "grad_norm": 2.56661482673391, "learning_rate": 1.012214643137084e-06, "loss": 0.3735, "step": 11285 }, { "epoch": 0.8001417936901808, "grad_norm": 3.789976996763963, "learning_rate": 1.0115221288316968e-06, "loss": 0.3839, "step": 11286 }, { "epoch": 0.8002126905352712, "grad_norm": 4.804610931829048, "learning_rate": 1.0108298248417952e-06, "loss": 0.3897, "step": 11287 }, { "epoch": 0.8002835873803615, "grad_norm": 3.7933504838626133, "learning_rate": 1.0101377312038785e-06, "loss": 0.3855, "step": 11288 }, { "epoch": 0.800354484225452, "grad_norm": 2.5461363887890722, "learning_rate": 1.009445847954445e-06, "loss": 0.4129, "step": 11289 }, { "epoch": 0.8004253810705424, "grad_norm": 8.391457291134628, "learning_rate": 1.008754175129979e-06, "loss": 0.4121, "step": 11290 }, { "epoch": 0.8004962779156327, "grad_norm": 3.3847613142462003, "learning_rate": 1.0080627127669517e-06, "loss": 0.3499, "step": 11291 }, { "epoch": 0.8005671747607231, "grad_norm": 3.171879486869523, "learning_rate": 1.0073714609018232e-06, "loss": 0.3926, "step": 11292 }, { "epoch": 0.8006380716058136, "grad_norm": 3.3346996242433207, "learning_rate": 1.006680419571045e-06, "loss": 0.3659, "step": 11293 }, { "epoch": 0.8007089684509039, "grad_norm": 6.330051063184622, "learning_rate": 1.005989588811057e-06, "loss": 0.3932, "step": 11294 }, { "epoch": 0.8007798652959943, "grad_norm": 3.889268535244967, "learning_rate": 1.0052989686582875e-06, "loss": 0.3969, "step": 11295 }, { "epoch": 0.8008507621410847, "grad_norm": 0.9579539796361533, "learning_rate": 1.0046085591491505e-06, "loss": 0.6029, "step": 11296 }, { "epoch": 0.8009216589861751, "grad_norm": 5.531232950412959, "learning_rate": 1.0039183603200547e-06, "loss": 0.3604, "step": 11297 }, { "epoch": 0.8009925558312655, "grad_norm": 2.7220493873258578, "learning_rate": 1.0032283722073943e-06, "loss": 0.3917, "step": 11298 }, { "epoch": 0.8010634526763559, "grad_norm": 2.0717085777337756, "learning_rate": 1.0025385948475513e-06, "loss": 0.3282, "step": 11299 }, { "epoch": 0.8011343495214462, "grad_norm": 6.499001470129486, "learning_rate": 1.0018490282769006e-06, "loss": 0.3803, "step": 11300 }, { "epoch": 0.8012052463665367, "grad_norm": 1.7494517211739744, "learning_rate": 1.0011596725318006e-06, "loss": 0.3953, "step": 11301 }, { "epoch": 0.8012761432116271, "grad_norm": 2.5027023372725092, "learning_rate": 1.0004705276486044e-06, "loss": 0.3872, "step": 11302 }, { "epoch": 0.8013470400567175, "grad_norm": 2.9763080884205997, "learning_rate": 9.997815936636507e-07, "loss": 0.364, "step": 11303 }, { "epoch": 0.8014179369018078, "grad_norm": 3.8172178363057543, "learning_rate": 9.990928706132646e-07, "loss": 0.3855, "step": 11304 }, { "epoch": 0.8014888337468983, "grad_norm": 1.7298540890937584, "learning_rate": 9.98404358533766e-07, "loss": 0.3868, "step": 11305 }, { "epoch": 0.8015597305919887, "grad_norm": 3.2907424092341864, "learning_rate": 9.97716057461461e-07, "loss": 0.3494, "step": 11306 }, { "epoch": 0.801630627437079, "grad_norm": 2.034224543472106, "learning_rate": 9.970279674326427e-07, "loss": 0.4, "step": 11307 }, { "epoch": 0.8017015242821695, "grad_norm": 4.239300895847797, "learning_rate": 9.963400884835938e-07, "loss": 0.3734, "step": 11308 }, { "epoch": 0.8017724211272599, "grad_norm": 3.006823704809022, "learning_rate": 9.956524206505891e-07, "loss": 0.3634, "step": 11309 }, { "epoch": 0.8018433179723502, "grad_norm": 2.8424187407211687, "learning_rate": 9.949649639698888e-07, "loss": 0.4253, "step": 11310 }, { "epoch": 0.8019142148174406, "grad_norm": 1.7778641802075967, "learning_rate": 9.942777184777418e-07, "loss": 0.3503, "step": 11311 }, { "epoch": 0.8019851116625311, "grad_norm": 4.998766355576271, "learning_rate": 9.935906842103881e-07, "loss": 0.3744, "step": 11312 }, { "epoch": 0.8020560085076214, "grad_norm": 4.166795373907942, "learning_rate": 9.929038612040564e-07, "loss": 0.4023, "step": 11313 }, { "epoch": 0.8021269053527118, "grad_norm": 2.663472990000012, "learning_rate": 9.922172494949629e-07, "loss": 0.3948, "step": 11314 }, { "epoch": 0.8021978021978022, "grad_norm": 1.9652998686586574, "learning_rate": 9.915308491193115e-07, "loss": 0.3963, "step": 11315 }, { "epoch": 0.8022686990428926, "grad_norm": 3.227475892251327, "learning_rate": 9.908446601132992e-07, "loss": 0.355, "step": 11316 }, { "epoch": 0.802339595887983, "grad_norm": 1.9573703871835404, "learning_rate": 9.90158682513108e-07, "loss": 0.3412, "step": 11317 }, { "epoch": 0.8024104927330734, "grad_norm": 1.8486090745448778, "learning_rate": 9.894729163549087e-07, "loss": 0.387, "step": 11318 }, { "epoch": 0.8024813895781637, "grad_norm": 3.3037546431129305, "learning_rate": 9.887873616748656e-07, "loss": 0.3708, "step": 11319 }, { "epoch": 0.8025522864232542, "grad_norm": 0.9834122901086663, "learning_rate": 9.881020185091244e-07, "loss": 0.682, "step": 11320 }, { "epoch": 0.8026231832683446, "grad_norm": 5.6563085170060114, "learning_rate": 9.874168868938278e-07, "loss": 0.452, "step": 11321 }, { "epoch": 0.8026940801134349, "grad_norm": 2.9322703693349337, "learning_rate": 9.867319668650999e-07, "loss": 0.4045, "step": 11322 }, { "epoch": 0.8027649769585253, "grad_norm": 3.414722869723994, "learning_rate": 9.860472584590597e-07, "loss": 0.368, "step": 11323 }, { "epoch": 0.8028358738036158, "grad_norm": 4.281102959922009, "learning_rate": 9.853627617118101e-07, "loss": 0.3992, "step": 11324 }, { "epoch": 0.8029067706487061, "grad_norm": 3.0882321225797265, "learning_rate": 9.846784766594475e-07, "loss": 0.3925, "step": 11325 }, { "epoch": 0.8029776674937965, "grad_norm": 1.8389289466305683, "learning_rate": 9.839944033380533e-07, "loss": 0.4027, "step": 11326 }, { "epoch": 0.8030485643388869, "grad_norm": 2.2480538428757586, "learning_rate": 9.833105417836986e-07, "loss": 0.4225, "step": 11327 }, { "epoch": 0.8031194611839774, "grad_norm": 30.85140058233882, "learning_rate": 9.82626892032446e-07, "loss": 0.4484, "step": 11328 }, { "epoch": 0.8031903580290677, "grad_norm": 0.9602182878460167, "learning_rate": 9.819434541203438e-07, "loss": 0.6995, "step": 11329 }, { "epoch": 0.8032612548741581, "grad_norm": 4.442361560239552, "learning_rate": 9.81260228083429e-07, "loss": 0.3511, "step": 11330 }, { "epoch": 0.8033321517192485, "grad_norm": 2.0639044321866726, "learning_rate": 9.805772139577296e-07, "loss": 0.4032, "step": 11331 }, { "epoch": 0.8034030485643389, "grad_norm": 3.0882421277961205, "learning_rate": 9.79894411779264e-07, "loss": 0.3912, "step": 11332 }, { "epoch": 0.8034739454094293, "grad_norm": 2.435816793328423, "learning_rate": 9.792118215840318e-07, "loss": 0.3646, "step": 11333 }, { "epoch": 0.8035448422545197, "grad_norm": 2.6785730970973223, "learning_rate": 9.785294434080295e-07, "loss": 0.4204, "step": 11334 }, { "epoch": 0.80361573909961, "grad_norm": 2.9051421228603393, "learning_rate": 9.778472772872406e-07, "loss": 0.3899, "step": 11335 }, { "epoch": 0.8036866359447005, "grad_norm": 2.819241766126015, "learning_rate": 9.77165323257634e-07, "loss": 0.3867, "step": 11336 }, { "epoch": 0.8037575327897909, "grad_norm": 2.603207062675757, "learning_rate": 9.7648358135517e-07, "loss": 0.3365, "step": 11337 }, { "epoch": 0.8038284296348812, "grad_norm": 4.913967397334171, "learning_rate": 9.758020516157967e-07, "loss": 0.4104, "step": 11338 }, { "epoch": 0.8038993264799716, "grad_norm": 4.221762365616162, "learning_rate": 9.751207340754548e-07, "loss": 0.3704, "step": 11339 }, { "epoch": 0.8039702233250621, "grad_norm": 3.2678626140127154, "learning_rate": 9.744396287700687e-07, "loss": 0.3328, "step": 11340 }, { "epoch": 0.8040411201701524, "grad_norm": 1.999449481587312, "learning_rate": 9.737587357355516e-07, "loss": 0.3673, "step": 11341 }, { "epoch": 0.8041120170152428, "grad_norm": 1.3387803163757477, "learning_rate": 9.73078055007811e-07, "loss": 0.3404, "step": 11342 }, { "epoch": 0.8041829138603332, "grad_norm": 5.9175557460489765, "learning_rate": 9.723975866227385e-07, "loss": 0.3632, "step": 11343 }, { "epoch": 0.8042538107054236, "grad_norm": 4.156587733534527, "learning_rate": 9.717173306162143e-07, "loss": 0.3801, "step": 11344 }, { "epoch": 0.804324707550514, "grad_norm": 1.8527015632506798, "learning_rate": 9.710372870241103e-07, "loss": 0.4068, "step": 11345 }, { "epoch": 0.8043956043956044, "grad_norm": 5.53915982306498, "learning_rate": 9.703574558822848e-07, "loss": 0.4057, "step": 11346 }, { "epoch": 0.8044665012406947, "grad_norm": 4.045028025733685, "learning_rate": 9.696778372265874e-07, "loss": 0.3595, "step": 11347 }, { "epoch": 0.8045373980857852, "grad_norm": 4.354126894395889, "learning_rate": 9.689984310928535e-07, "loss": 0.4135, "step": 11348 }, { "epoch": 0.8046082949308756, "grad_norm": 2.510325266735776, "learning_rate": 9.683192375169081e-07, "loss": 0.3494, "step": 11349 }, { "epoch": 0.8046791917759659, "grad_norm": 5.290641589262908, "learning_rate": 9.676402565345667e-07, "loss": 0.3661, "step": 11350 }, { "epoch": 0.8047500886210563, "grad_norm": 2.846923939474411, "learning_rate": 9.669614881816336e-07, "loss": 0.3775, "step": 11351 }, { "epoch": 0.8048209854661468, "grad_norm": 5.670753389890236, "learning_rate": 9.662829324938993e-07, "loss": 0.3769, "step": 11352 }, { "epoch": 0.8048918823112371, "grad_norm": 4.442698462502354, "learning_rate": 9.656045895071435e-07, "loss": 0.3933, "step": 11353 }, { "epoch": 0.8049627791563275, "grad_norm": 2.899060200423377, "learning_rate": 9.649264592571383e-07, "loss": 0.4297, "step": 11354 }, { "epoch": 0.805033676001418, "grad_norm": 2.2347532849200227, "learning_rate": 9.64248541779641e-07, "loss": 0.4077, "step": 11355 }, { "epoch": 0.8051045728465084, "grad_norm": 2.6802237448475044, "learning_rate": 9.63570837110397e-07, "loss": 0.4141, "step": 11356 }, { "epoch": 0.8051754696915987, "grad_norm": 2.2483135455243466, "learning_rate": 9.62893345285143e-07, "loss": 0.3744, "step": 11357 }, { "epoch": 0.8052463665366891, "grad_norm": 2.322695474974867, "learning_rate": 9.622160663396063e-07, "loss": 0.3556, "step": 11358 }, { "epoch": 0.8053172633817796, "grad_norm": 3.0292326321496796, "learning_rate": 9.615390003094983e-07, "loss": 0.426, "step": 11359 }, { "epoch": 0.8053881602268699, "grad_norm": 2.7517214970328654, "learning_rate": 9.608621472305196e-07, "loss": 0.3645, "step": 11360 }, { "epoch": 0.8054590570719603, "grad_norm": 3.280896013183468, "learning_rate": 9.601855071383643e-07, "loss": 0.406, "step": 11361 }, { "epoch": 0.8055299539170507, "grad_norm": 3.9399534728225007, "learning_rate": 9.595090800687107e-07, "loss": 0.4121, "step": 11362 }, { "epoch": 0.805600850762141, "grad_norm": 4.5297349235303, "learning_rate": 9.58832866057226e-07, "loss": 0.3684, "step": 11363 }, { "epoch": 0.8056717476072315, "grad_norm": 4.422848323620975, "learning_rate": 9.5815686513957e-07, "loss": 0.391, "step": 11364 }, { "epoch": 0.8057426444523219, "grad_norm": 1.7771135744223134, "learning_rate": 9.57481077351386e-07, "loss": 0.3565, "step": 11365 }, { "epoch": 0.8058135412974122, "grad_norm": 1.0475241165600255, "learning_rate": 9.56805502728312e-07, "loss": 0.666, "step": 11366 }, { "epoch": 0.8058844381425027, "grad_norm": 1.561609449811065, "learning_rate": 9.561301413059682e-07, "loss": 0.3724, "step": 11367 }, { "epoch": 0.8059553349875931, "grad_norm": 6.651625407127297, "learning_rate": 9.554549931199703e-07, "loss": 0.3983, "step": 11368 }, { "epoch": 0.8060262318326834, "grad_norm": 3.693351702726471, "learning_rate": 9.547800582059159e-07, "loss": 0.3484, "step": 11369 }, { "epoch": 0.8060971286777738, "grad_norm": 2.380939805264567, "learning_rate": 9.54105336599398e-07, "loss": 0.3637, "step": 11370 }, { "epoch": 0.8061680255228643, "grad_norm": 5.543938538761397, "learning_rate": 9.534308283359939e-07, "loss": 0.4216, "step": 11371 }, { "epoch": 0.8062389223679546, "grad_norm": 5.121817133195775, "learning_rate": 9.5275653345127e-07, "loss": 0.3935, "step": 11372 }, { "epoch": 0.806309819213045, "grad_norm": 3.5478308886084005, "learning_rate": 9.520824519807836e-07, "loss": 0.4403, "step": 11373 }, { "epoch": 0.8063807160581354, "grad_norm": 4.983387226246107, "learning_rate": 9.514085839600801e-07, "loss": 0.4041, "step": 11374 }, { "epoch": 0.8064516129032258, "grad_norm": 4.068414328901669, "learning_rate": 9.507349294246904e-07, "loss": 0.4184, "step": 11375 }, { "epoch": 0.8065225097483162, "grad_norm": 3.6281454644549034, "learning_rate": 9.500614884101383e-07, "loss": 0.3736, "step": 11376 }, { "epoch": 0.8065934065934066, "grad_norm": 3.353737421856581, "learning_rate": 9.493882609519361e-07, "loss": 0.3784, "step": 11377 }, { "epoch": 0.8066643034384969, "grad_norm": 2.748134269358954, "learning_rate": 9.487152470855832e-07, "loss": 0.3915, "step": 11378 }, { "epoch": 0.8067352002835874, "grad_norm": 2.2446042898183833, "learning_rate": 9.48042446846566e-07, "loss": 0.3345, "step": 11379 }, { "epoch": 0.8068060971286778, "grad_norm": 2.16609260739476, "learning_rate": 9.473698602703635e-07, "loss": 0.3556, "step": 11380 }, { "epoch": 0.8068769939737682, "grad_norm": 3.1546543869313646, "learning_rate": 9.46697487392444e-07, "loss": 0.3953, "step": 11381 }, { "epoch": 0.8069478908188585, "grad_norm": 1.7592295142221495, "learning_rate": 9.460253282482573e-07, "loss": 0.3849, "step": 11382 }, { "epoch": 0.807018787663949, "grad_norm": 1.0404236053185836, "learning_rate": 9.453533828732492e-07, "loss": 0.6144, "step": 11383 }, { "epoch": 0.8070896845090394, "grad_norm": 2.4150029493451224, "learning_rate": 9.446816513028528e-07, "loss": 0.4004, "step": 11384 }, { "epoch": 0.8071605813541297, "grad_norm": 2.7428936735882266, "learning_rate": 9.440101335724888e-07, "loss": 0.5086, "step": 11385 }, { "epoch": 0.8072314781992201, "grad_norm": 2.401814821416194, "learning_rate": 9.43338829717565e-07, "loss": 0.3782, "step": 11386 }, { "epoch": 0.8073023750443106, "grad_norm": 2.9800481948148825, "learning_rate": 9.426677397734818e-07, "loss": 0.4063, "step": 11387 }, { "epoch": 0.8073732718894009, "grad_norm": 2.4340640124150745, "learning_rate": 9.41996863775626e-07, "loss": 0.3661, "step": 11388 }, { "epoch": 0.8074441687344913, "grad_norm": 2.0585376152257657, "learning_rate": 9.41326201759371e-07, "loss": 0.3437, "step": 11389 }, { "epoch": 0.8075150655795817, "grad_norm": 2.4256309728162897, "learning_rate": 9.406557537600852e-07, "loss": 0.4075, "step": 11390 }, { "epoch": 0.8075859624246721, "grad_norm": 2.8733639007800127, "learning_rate": 9.39985519813118e-07, "loss": 0.3337, "step": 11391 }, { "epoch": 0.8076568592697625, "grad_norm": 2.6977691450993673, "learning_rate": 9.393154999538151e-07, "loss": 0.3363, "step": 11392 }, { "epoch": 0.8077277561148529, "grad_norm": 5.067322428180986, "learning_rate": 9.386456942175049e-07, "loss": 0.3943, "step": 11393 }, { "epoch": 0.8077986529599432, "grad_norm": 2.04867534747846, "learning_rate": 9.379761026395056e-07, "loss": 0.4255, "step": 11394 }, { "epoch": 0.8078695498050337, "grad_norm": 1.6538757903691745, "learning_rate": 9.373067252551271e-07, "loss": 0.3947, "step": 11395 }, { "epoch": 0.8079404466501241, "grad_norm": 3.5186665053090667, "learning_rate": 9.366375620996676e-07, "loss": 0.351, "step": 11396 }, { "epoch": 0.8080113434952144, "grad_norm": 3.3635899970872645, "learning_rate": 9.3596861320841e-07, "loss": 0.4046, "step": 11397 }, { "epoch": 0.8080822403403048, "grad_norm": 2.733089069869359, "learning_rate": 9.352998786166289e-07, "loss": 0.401, "step": 11398 }, { "epoch": 0.8081531371853953, "grad_norm": 2.4321967476326996, "learning_rate": 9.346313583595884e-07, "loss": 0.4119, "step": 11399 }, { "epoch": 0.8082240340304856, "grad_norm": 3.3981679098740347, "learning_rate": 9.339630524725396e-07, "loss": 0.4165, "step": 11400 }, { "epoch": 0.808294930875576, "grad_norm": 3.1786648488823217, "learning_rate": 9.33294960990721e-07, "loss": 0.3776, "step": 11401 }, { "epoch": 0.8083658277206665, "grad_norm": 3.5347555934688346, "learning_rate": 9.326270839493634e-07, "loss": 0.3852, "step": 11402 }, { "epoch": 0.8084367245657568, "grad_norm": 4.082013306601372, "learning_rate": 9.319594213836852e-07, "loss": 0.3963, "step": 11403 }, { "epoch": 0.8085076214108472, "grad_norm": 2.935517064274075, "learning_rate": 9.312919733288917e-07, "loss": 0.4057, "step": 11404 }, { "epoch": 0.8085785182559376, "grad_norm": 3.562707797300734, "learning_rate": 9.306247398201768e-07, "loss": 0.3863, "step": 11405 }, { "epoch": 0.8086494151010281, "grad_norm": 2.3124793995706985, "learning_rate": 9.299577208927263e-07, "loss": 0.4047, "step": 11406 }, { "epoch": 0.8087203119461184, "grad_norm": 3.0712203497250803, "learning_rate": 9.292909165817121e-07, "loss": 0.3988, "step": 11407 }, { "epoch": 0.8087912087912088, "grad_norm": 3.0229740725087253, "learning_rate": 9.286243269222933e-07, "loss": 0.3983, "step": 11408 }, { "epoch": 0.8088621056362992, "grad_norm": 4.338960116393547, "learning_rate": 9.279579519496212e-07, "loss": 0.3513, "step": 11409 }, { "epoch": 0.8089330024813896, "grad_norm": 2.1159561479755586, "learning_rate": 9.272917916988361e-07, "loss": 0.3486, "step": 11410 }, { "epoch": 0.80900389932648, "grad_norm": 2.9628550681899424, "learning_rate": 9.266258462050631e-07, "loss": 0.3373, "step": 11411 }, { "epoch": 0.8090747961715704, "grad_norm": 2.737756662996263, "learning_rate": 9.25960115503417e-07, "loss": 0.3741, "step": 11412 }, { "epoch": 0.8091456930166607, "grad_norm": 3.955479377733785, "learning_rate": 9.252945996290052e-07, "loss": 0.395, "step": 11413 }, { "epoch": 0.8092165898617512, "grad_norm": 9.484499736168273, "learning_rate": 9.246292986169175e-07, "loss": 0.4064, "step": 11414 }, { "epoch": 0.8092874867068416, "grad_norm": 3.5465465606065254, "learning_rate": 9.239642125022391e-07, "loss": 0.4611, "step": 11415 }, { "epoch": 0.8093583835519319, "grad_norm": 6.0340816328774105, "learning_rate": 9.232993413200386e-07, "loss": 0.4076, "step": 11416 }, { "epoch": 0.8094292803970223, "grad_norm": 4.742083048889673, "learning_rate": 9.226346851053742e-07, "loss": 0.378, "step": 11417 }, { "epoch": 0.8095001772421128, "grad_norm": 2.1292160409590597, "learning_rate": 9.219702438932965e-07, "loss": 0.412, "step": 11418 }, { "epoch": 0.8095710740872031, "grad_norm": 3.8894988804513204, "learning_rate": 9.213060177188404e-07, "loss": 0.4021, "step": 11419 }, { "epoch": 0.8096419709322935, "grad_norm": 2.087564806052274, "learning_rate": 9.206420066170297e-07, "loss": 0.3561, "step": 11420 }, { "epoch": 0.8097128677773839, "grad_norm": 4.149622430644911, "learning_rate": 9.199782106228794e-07, "loss": 0.2959, "step": 11421 }, { "epoch": 0.8097837646224743, "grad_norm": 1.8586004659387485, "learning_rate": 9.193146297713939e-07, "loss": 0.3829, "step": 11422 }, { "epoch": 0.8098546614675647, "grad_norm": 4.884190774045079, "learning_rate": 9.186512640975625e-07, "loss": 0.4008, "step": 11423 }, { "epoch": 0.8099255583126551, "grad_norm": 2.25493654028715, "learning_rate": 9.179881136363639e-07, "loss": 0.3854, "step": 11424 }, { "epoch": 0.8099964551577454, "grad_norm": 3.398452953562656, "learning_rate": 9.173251784227671e-07, "loss": 0.4138, "step": 11425 }, { "epoch": 0.8100673520028359, "grad_norm": 4.309626320551884, "learning_rate": 9.166624584917333e-07, "loss": 0.4072, "step": 11426 }, { "epoch": 0.8101382488479263, "grad_norm": 4.312858155665843, "learning_rate": 9.159999538782016e-07, "loss": 0.4249, "step": 11427 }, { "epoch": 0.8102091456930166, "grad_norm": 3.3146311123781436, "learning_rate": 9.153376646171098e-07, "loss": 0.3966, "step": 11428 }, { "epoch": 0.810280042538107, "grad_norm": 11.186807668502041, "learning_rate": 9.146755907433813e-07, "loss": 0.4108, "step": 11429 }, { "epoch": 0.8103509393831975, "grad_norm": 4.059557413852386, "learning_rate": 9.140137322919279e-07, "loss": 0.3648, "step": 11430 }, { "epoch": 0.8104218362282879, "grad_norm": 5.56807029540422, "learning_rate": 9.133520892976471e-07, "loss": 0.3704, "step": 11431 }, { "epoch": 0.8104927330733782, "grad_norm": 3.36598476770747, "learning_rate": 9.126906617954318e-07, "loss": 0.4305, "step": 11432 }, { "epoch": 0.8105636299184686, "grad_norm": 2.602469582159809, "learning_rate": 9.12029449820156e-07, "loss": 0.3717, "step": 11433 }, { "epoch": 0.8106345267635591, "grad_norm": 2.6014810899712506, "learning_rate": 9.113684534066886e-07, "loss": 0.3504, "step": 11434 }, { "epoch": 0.8107054236086494, "grad_norm": 1.7325243464378008, "learning_rate": 9.107076725898834e-07, "loss": 0.3617, "step": 11435 }, { "epoch": 0.8107763204537398, "grad_norm": 1.9294563077113105, "learning_rate": 9.100471074045825e-07, "loss": 0.3679, "step": 11436 }, { "epoch": 0.8108472172988302, "grad_norm": 2.613121526141402, "learning_rate": 9.093867578856208e-07, "loss": 0.3816, "step": 11437 }, { "epoch": 0.8109181141439206, "grad_norm": 2.417653808768323, "learning_rate": 9.087266240678166e-07, "loss": 0.368, "step": 11438 }, { "epoch": 0.810989010989011, "grad_norm": 3.707937043943761, "learning_rate": 9.08066705985981e-07, "loss": 0.4039, "step": 11439 }, { "epoch": 0.8110599078341014, "grad_norm": 2.720849652341017, "learning_rate": 9.0740700367491e-07, "loss": 0.3556, "step": 11440 }, { "epoch": 0.8111308046791917, "grad_norm": 7.593246248971131, "learning_rate": 9.067475171693929e-07, "loss": 0.3657, "step": 11441 }, { "epoch": 0.8112017015242822, "grad_norm": 11.612435646657552, "learning_rate": 9.060882465042037e-07, "loss": 0.4455, "step": 11442 }, { "epoch": 0.8112725983693726, "grad_norm": 2.5879681444524234, "learning_rate": 9.054291917141045e-07, "loss": 0.4276, "step": 11443 }, { "epoch": 0.8113434952144629, "grad_norm": 2.47016640338455, "learning_rate": 9.047703528338503e-07, "loss": 0.3352, "step": 11444 }, { "epoch": 0.8114143920595533, "grad_norm": 2.019535952986153, "learning_rate": 9.041117298981816e-07, "loss": 0.3436, "step": 11445 }, { "epoch": 0.8114852889046438, "grad_norm": 8.22329893536287, "learning_rate": 9.034533229418258e-07, "loss": 0.3537, "step": 11446 }, { "epoch": 0.8115561857497341, "grad_norm": 2.592067648444064, "learning_rate": 9.027951319995037e-07, "loss": 0.3839, "step": 11447 }, { "epoch": 0.8116270825948245, "grad_norm": 3.1155670343224546, "learning_rate": 9.021371571059223e-07, "loss": 0.3968, "step": 11448 }, { "epoch": 0.811697979439915, "grad_norm": 6.696234433050048, "learning_rate": 9.014793982957765e-07, "loss": 0.4137, "step": 11449 }, { "epoch": 0.8117688762850053, "grad_norm": 2.307069762070259, "learning_rate": 9.00821855603749e-07, "loss": 0.364, "step": 11450 }, { "epoch": 0.8118397731300957, "grad_norm": 7.116496796745324, "learning_rate": 9.001645290645151e-07, "loss": 0.4151, "step": 11451 }, { "epoch": 0.8119106699751861, "grad_norm": 2.0088415706227054, "learning_rate": 8.995074187127351e-07, "loss": 0.3363, "step": 11452 }, { "epoch": 0.8119815668202764, "grad_norm": 2.946911505673864, "learning_rate": 8.98850524583057e-07, "loss": 0.3625, "step": 11453 }, { "epoch": 0.8120524636653669, "grad_norm": 3.0468321593019483, "learning_rate": 8.981938467101209e-07, "loss": 0.377, "step": 11454 }, { "epoch": 0.8121233605104573, "grad_norm": 4.284221065498558, "learning_rate": 8.975373851285557e-07, "loss": 0.4036, "step": 11455 }, { "epoch": 0.8121942573555477, "grad_norm": 2.9752963896852793, "learning_rate": 8.968811398729755e-07, "loss": 0.4101, "step": 11456 }, { "epoch": 0.812265154200638, "grad_norm": 2.0511091119849914, "learning_rate": 8.962251109779829e-07, "loss": 0.4023, "step": 11457 }, { "epoch": 0.8123360510457285, "grad_norm": 3.3956970291241015, "learning_rate": 8.955692984781739e-07, "loss": 0.4214, "step": 11458 }, { "epoch": 0.8124069478908189, "grad_norm": 2.6998872727347316, "learning_rate": 8.949137024081272e-07, "loss": 0.3959, "step": 11459 }, { "epoch": 0.8124778447359092, "grad_norm": 1.9345338409465505, "learning_rate": 8.94258322802416e-07, "loss": 0.3816, "step": 11460 }, { "epoch": 0.8125487415809997, "grad_norm": 2.7653175818049403, "learning_rate": 8.936031596955968e-07, "loss": 0.3668, "step": 11461 }, { "epoch": 0.8126196384260901, "grad_norm": 3.791820210061319, "learning_rate": 8.929482131222161e-07, "loss": 0.461, "step": 11462 }, { "epoch": 0.8126905352711804, "grad_norm": 1.8144498625846333, "learning_rate": 8.922934831168118e-07, "loss": 0.4027, "step": 11463 }, { "epoch": 0.8127614321162708, "grad_norm": 2.2091509309288813, "learning_rate": 8.916389697139077e-07, "loss": 0.3354, "step": 11464 }, { "epoch": 0.8128323289613613, "grad_norm": 2.593022123329501, "learning_rate": 8.909846729480154e-07, "loss": 0.3576, "step": 11465 }, { "epoch": 0.8129032258064516, "grad_norm": 4.325129178050547, "learning_rate": 8.903305928536377e-07, "loss": 0.3561, "step": 11466 }, { "epoch": 0.812974122651542, "grad_norm": 3.3987934696825395, "learning_rate": 8.896767294652652e-07, "loss": 0.3733, "step": 11467 }, { "epoch": 0.8130450194966324, "grad_norm": 4.761449498706776, "learning_rate": 8.890230828173768e-07, "loss": 0.4121, "step": 11468 }, { "epoch": 0.8131159163417228, "grad_norm": 1.1005450236255818, "learning_rate": 8.883696529444369e-07, "loss": 0.7227, "step": 11469 }, { "epoch": 0.8131868131868132, "grad_norm": 2.011611790351758, "learning_rate": 8.87716439880904e-07, "loss": 0.4371, "step": 11470 }, { "epoch": 0.8132577100319036, "grad_norm": 2.636287190778027, "learning_rate": 8.870634436612246e-07, "loss": 0.3432, "step": 11471 }, { "epoch": 0.8133286068769939, "grad_norm": 9.267303330401427, "learning_rate": 8.864106643198261e-07, "loss": 0.373, "step": 11472 }, { "epoch": 0.8133995037220844, "grad_norm": 1.879517131424285, "learning_rate": 8.857581018911337e-07, "loss": 0.371, "step": 11473 }, { "epoch": 0.8134704005671748, "grad_norm": 2.513552473512139, "learning_rate": 8.851057564095577e-07, "loss": 0.4119, "step": 11474 }, { "epoch": 0.8135412974122651, "grad_norm": 2.446086281610813, "learning_rate": 8.844536279094956e-07, "loss": 0.3912, "step": 11475 }, { "epoch": 0.8136121942573555, "grad_norm": 3.3087912375602606, "learning_rate": 8.838017164253343e-07, "loss": 0.363, "step": 11476 }, { "epoch": 0.813683091102446, "grad_norm": 13.995250855426402, "learning_rate": 8.831500219914513e-07, "loss": 0.42, "step": 11477 }, { "epoch": 0.8137539879475363, "grad_norm": 2.11358258223208, "learning_rate": 8.824985446422086e-07, "loss": 0.4306, "step": 11478 }, { "epoch": 0.8138248847926267, "grad_norm": 3.1742291645747858, "learning_rate": 8.818472844119613e-07, "loss": 0.3798, "step": 11479 }, { "epoch": 0.8138957816377171, "grad_norm": 4.351370454851838, "learning_rate": 8.811962413350505e-07, "loss": 0.4186, "step": 11480 }, { "epoch": 0.8139666784828075, "grad_norm": 3.492784212117891, "learning_rate": 8.805454154458038e-07, "loss": 0.4009, "step": 11481 }, { "epoch": 0.8140375753278979, "grad_norm": 2.2985070112024286, "learning_rate": 8.798948067785435e-07, "loss": 0.3622, "step": 11482 }, { "epoch": 0.8141084721729883, "grad_norm": 3.3114618384505006, "learning_rate": 8.792444153675733e-07, "loss": 0.3725, "step": 11483 }, { "epoch": 0.8141793690180787, "grad_norm": 2.103234149073842, "learning_rate": 8.785942412471909e-07, "loss": 0.3677, "step": 11484 }, { "epoch": 0.8142502658631691, "grad_norm": 5.704849498633646, "learning_rate": 8.779442844516795e-07, "loss": 0.3383, "step": 11485 }, { "epoch": 0.8143211627082595, "grad_norm": 2.0304035832878866, "learning_rate": 8.77294545015313e-07, "loss": 0.4209, "step": 11486 }, { "epoch": 0.8143920595533499, "grad_norm": 2.342143402399532, "learning_rate": 8.766450229723522e-07, "loss": 0.3716, "step": 11487 }, { "epoch": 0.8144629563984402, "grad_norm": 1.6652314762633027, "learning_rate": 8.75995718357045e-07, "loss": 0.3769, "step": 11488 }, { "epoch": 0.8145338532435307, "grad_norm": 2.1613984180673835, "learning_rate": 8.753466312036312e-07, "loss": 0.3638, "step": 11489 }, { "epoch": 0.8146047500886211, "grad_norm": 13.260719084029898, "learning_rate": 8.746977615463404e-07, "loss": 0.4357, "step": 11490 }, { "epoch": 0.8146756469337114, "grad_norm": 1.652199740965039, "learning_rate": 8.740491094193831e-07, "loss": 0.3745, "step": 11491 }, { "epoch": 0.8147465437788018, "grad_norm": 1.7546131126653208, "learning_rate": 8.734006748569651e-07, "loss": 0.3386, "step": 11492 }, { "epoch": 0.8148174406238923, "grad_norm": 7.802872133803628, "learning_rate": 8.727524578932806e-07, "loss": 0.3727, "step": 11493 }, { "epoch": 0.8148883374689826, "grad_norm": 9.4515642195047, "learning_rate": 8.721044585625093e-07, "loss": 0.3892, "step": 11494 }, { "epoch": 0.814959234314073, "grad_norm": 2.9486487991951065, "learning_rate": 8.714566768988192e-07, "loss": 0.3803, "step": 11495 }, { "epoch": 0.8150301311591635, "grad_norm": 2.519764473557668, "learning_rate": 8.708091129363705e-07, "loss": 0.358, "step": 11496 }, { "epoch": 0.8151010280042538, "grad_norm": 6.021852811210047, "learning_rate": 8.701617667093082e-07, "loss": 0.4328, "step": 11497 }, { "epoch": 0.8151719248493442, "grad_norm": 3.428010569970171, "learning_rate": 8.695146382517694e-07, "loss": 0.4398, "step": 11498 }, { "epoch": 0.8152428216944346, "grad_norm": 2.263351120481213, "learning_rate": 8.688677275978752e-07, "loss": 0.4211, "step": 11499 }, { "epoch": 0.815313718539525, "grad_norm": 3.1496785648471666, "learning_rate": 8.682210347817399e-07, "loss": 0.3947, "step": 11500 }, { "epoch": 0.8153846153846154, "grad_norm": 2.6681678139520573, "learning_rate": 8.67574559837463e-07, "loss": 0.3774, "step": 11501 }, { "epoch": 0.8154555122297058, "grad_norm": 1.889497783484148, "learning_rate": 8.669283027991327e-07, "loss": 0.3945, "step": 11502 }, { "epoch": 0.8155264090747961, "grad_norm": 1.9926546534087763, "learning_rate": 8.662822637008284e-07, "loss": 0.3743, "step": 11503 }, { "epoch": 0.8155973059198866, "grad_norm": 1.9739139984271523, "learning_rate": 8.656364425766145e-07, "loss": 0.381, "step": 11504 }, { "epoch": 0.815668202764977, "grad_norm": 3.016339316193515, "learning_rate": 8.649908394605477e-07, "loss": 0.3575, "step": 11505 }, { "epoch": 0.8157390996100673, "grad_norm": 3.0343356554345635, "learning_rate": 8.643454543866703e-07, "loss": 0.3694, "step": 11506 }, { "epoch": 0.8158099964551577, "grad_norm": 17.29462035362298, "learning_rate": 8.637002873890127e-07, "loss": 0.4063, "step": 11507 }, { "epoch": 0.8158808933002482, "grad_norm": 2.9295549517844446, "learning_rate": 8.630553385015966e-07, "loss": 0.4251, "step": 11508 }, { "epoch": 0.8159517901453386, "grad_norm": 7.047086844402952, "learning_rate": 8.624106077584305e-07, "loss": 0.3727, "step": 11509 }, { "epoch": 0.8160226869904289, "grad_norm": 3.980156086443459, "learning_rate": 8.617660951935097e-07, "loss": 0.4054, "step": 11510 }, { "epoch": 0.8160935838355193, "grad_norm": 4.30653161672188, "learning_rate": 8.611218008408217e-07, "loss": 0.4323, "step": 11511 }, { "epoch": 0.8161644806806098, "grad_norm": 8.873661644809378, "learning_rate": 8.60477724734341e-07, "loss": 0.3951, "step": 11512 }, { "epoch": 0.8162353775257001, "grad_norm": 2.4937220452739206, "learning_rate": 8.598338669080297e-07, "loss": 0.3675, "step": 11513 }, { "epoch": 0.8163062743707905, "grad_norm": 4.324974901924789, "learning_rate": 8.591902273958374e-07, "loss": 0.318, "step": 11514 }, { "epoch": 0.8163771712158809, "grad_norm": 3.0314045193343926, "learning_rate": 8.585468062317048e-07, "loss": 0.39, "step": 11515 }, { "epoch": 0.8164480680609713, "grad_norm": 3.073915205651239, "learning_rate": 8.579036034495624e-07, "loss": 0.3369, "step": 11516 }, { "epoch": 0.8165189649060617, "grad_norm": 7.390775725700312, "learning_rate": 8.572606190833221e-07, "loss": 0.4055, "step": 11517 }, { "epoch": 0.8165898617511521, "grad_norm": 1.4361200530155274, "learning_rate": 8.566178531668917e-07, "loss": 0.3641, "step": 11518 }, { "epoch": 0.8166607585962424, "grad_norm": 1.9422942197443989, "learning_rate": 8.559753057341657e-07, "loss": 0.3758, "step": 11519 }, { "epoch": 0.8167316554413329, "grad_norm": 6.568641936196822, "learning_rate": 8.553329768190244e-07, "loss": 0.4004, "step": 11520 }, { "epoch": 0.8168025522864233, "grad_norm": 5.337937945528435, "learning_rate": 8.546908664553378e-07, "loss": 0.3566, "step": 11521 }, { "epoch": 0.8168734491315136, "grad_norm": 2.2432977769890905, "learning_rate": 8.54048974676967e-07, "loss": 0.4068, "step": 11522 }, { "epoch": 0.816944345976604, "grad_norm": 6.686366854383373, "learning_rate": 8.534073015177569e-07, "loss": 0.3767, "step": 11523 }, { "epoch": 0.8170152428216945, "grad_norm": 8.494869304956401, "learning_rate": 8.527658470115457e-07, "loss": 0.3658, "step": 11524 }, { "epoch": 0.8170861396667848, "grad_norm": 3.8245164585371403, "learning_rate": 8.521246111921571e-07, "loss": 0.4131, "step": 11525 }, { "epoch": 0.8171570365118752, "grad_norm": 2.9705166930822156, "learning_rate": 8.514835940934024e-07, "loss": 0.3737, "step": 11526 }, { "epoch": 0.8172279333569656, "grad_norm": 2.4645394199629314, "learning_rate": 8.508427957490855e-07, "loss": 0.4055, "step": 11527 }, { "epoch": 0.817298830202056, "grad_norm": 4.991556289932192, "learning_rate": 8.502022161929935e-07, "loss": 0.3448, "step": 11528 }, { "epoch": 0.8173697270471464, "grad_norm": 2.2606522878752537, "learning_rate": 8.495618554589074e-07, "loss": 0.3937, "step": 11529 }, { "epoch": 0.8174406238922368, "grad_norm": 2.0102568325114603, "learning_rate": 8.489217135805916e-07, "loss": 0.4473, "step": 11530 }, { "epoch": 0.8175115207373271, "grad_norm": 5.407249080484359, "learning_rate": 8.482817905918029e-07, "loss": 0.3478, "step": 11531 }, { "epoch": 0.8175824175824176, "grad_norm": 3.4119145523749896, "learning_rate": 8.476420865262847e-07, "loss": 0.4288, "step": 11532 }, { "epoch": 0.817653314427508, "grad_norm": 3.134251172336933, "learning_rate": 8.470026014177668e-07, "loss": 0.3421, "step": 11533 }, { "epoch": 0.8177242112725984, "grad_norm": 3.41498470993352, "learning_rate": 8.463633352999723e-07, "loss": 0.4154, "step": 11534 }, { "epoch": 0.8177951081176887, "grad_norm": 3.9616018539752065, "learning_rate": 8.457242882066114e-07, "loss": 0.402, "step": 11535 }, { "epoch": 0.8178660049627792, "grad_norm": 2.8087693394096975, "learning_rate": 8.450854601713776e-07, "loss": 0.3841, "step": 11536 }, { "epoch": 0.8179369018078696, "grad_norm": 2.7432714339325512, "learning_rate": 8.444468512279585e-07, "loss": 0.3948, "step": 11537 }, { "epoch": 0.8180077986529599, "grad_norm": 3.4236912382362474, "learning_rate": 8.438084614100301e-07, "loss": 0.4086, "step": 11538 }, { "epoch": 0.8180786954980503, "grad_norm": 4.723233896978941, "learning_rate": 8.431702907512535e-07, "loss": 0.3816, "step": 11539 }, { "epoch": 0.8181495923431408, "grad_norm": 2.5426431302679227, "learning_rate": 8.425323392852791e-07, "loss": 0.4066, "step": 11540 }, { "epoch": 0.8182204891882311, "grad_norm": 2.95462652500773, "learning_rate": 8.418946070457479e-07, "loss": 0.3818, "step": 11541 }, { "epoch": 0.8182913860333215, "grad_norm": 2.3355354476624823, "learning_rate": 8.412570940662884e-07, "loss": 0.3291, "step": 11542 }, { "epoch": 0.818362282878412, "grad_norm": 2.6955549372386494, "learning_rate": 8.406198003805171e-07, "loss": 0.387, "step": 11543 }, { "epoch": 0.8184331797235023, "grad_norm": 2.4346178930599454, "learning_rate": 8.399827260220373e-07, "loss": 0.3591, "step": 11544 }, { "epoch": 0.8185040765685927, "grad_norm": 3.4664650190230875, "learning_rate": 8.393458710244445e-07, "loss": 0.4035, "step": 11545 }, { "epoch": 0.8185749734136831, "grad_norm": 2.893426441364348, "learning_rate": 8.3870923542132e-07, "loss": 0.3767, "step": 11546 }, { "epoch": 0.8186458702587734, "grad_norm": 2.368906266323311, "learning_rate": 8.380728192462323e-07, "loss": 0.3351, "step": 11547 }, { "epoch": 0.8187167671038639, "grad_norm": 4.24028495174384, "learning_rate": 8.374366225327424e-07, "loss": 0.3585, "step": 11548 }, { "epoch": 0.8187876639489543, "grad_norm": 5.418471623090761, "learning_rate": 8.368006453143956e-07, "loss": 0.3343, "step": 11549 }, { "epoch": 0.8188585607940446, "grad_norm": 1.025354432813058, "learning_rate": 8.361648876247297e-07, "loss": 0.6764, "step": 11550 }, { "epoch": 0.818929457639135, "grad_norm": 2.3908924485991263, "learning_rate": 8.355293494972671e-07, "loss": 0.4133, "step": 11551 }, { "epoch": 0.8190003544842255, "grad_norm": 3.540167111636743, "learning_rate": 8.348940309655196e-07, "loss": 0.3555, "step": 11552 }, { "epoch": 0.8190712513293158, "grad_norm": 5.774999563606069, "learning_rate": 8.342589320629884e-07, "loss": 0.3473, "step": 11553 }, { "epoch": 0.8191421481744062, "grad_norm": 2.606183705951661, "learning_rate": 8.336240528231665e-07, "loss": 0.3623, "step": 11554 }, { "epoch": 0.8192130450194967, "grad_norm": 2.140182426982303, "learning_rate": 8.329893932795252e-07, "loss": 0.3756, "step": 11555 }, { "epoch": 0.819283941864587, "grad_norm": 3.327791798619425, "learning_rate": 8.32354953465534e-07, "loss": 0.432, "step": 11556 }, { "epoch": 0.8193548387096774, "grad_norm": 13.252290083389992, "learning_rate": 8.317207334146482e-07, "loss": 0.3647, "step": 11557 }, { "epoch": 0.8194257355547678, "grad_norm": 2.619677666039613, "learning_rate": 8.310867331603101e-07, "loss": 0.3869, "step": 11558 }, { "epoch": 0.8194966323998583, "grad_norm": 1.7960524673110332, "learning_rate": 8.30452952735949e-07, "loss": 0.3894, "step": 11559 }, { "epoch": 0.8195675292449486, "grad_norm": 2.6928499269159656, "learning_rate": 8.298193921749859e-07, "loss": 0.3636, "step": 11560 }, { "epoch": 0.819638426090039, "grad_norm": 3.4731503041090743, "learning_rate": 8.291860515108319e-07, "loss": 0.3456, "step": 11561 }, { "epoch": 0.8197093229351294, "grad_norm": 4.4779865571715565, "learning_rate": 8.285529307768786e-07, "loss": 0.3992, "step": 11562 }, { "epoch": 0.8197802197802198, "grad_norm": 3.9564258027554615, "learning_rate": 8.279200300065132e-07, "loss": 0.3589, "step": 11563 }, { "epoch": 0.8198511166253102, "grad_norm": 0.9472149427011496, "learning_rate": 8.272873492331102e-07, "loss": 0.643, "step": 11564 }, { "epoch": 0.8199220134704006, "grad_norm": 2.305493718831155, "learning_rate": 8.2665488849003e-07, "loss": 0.3513, "step": 11565 }, { "epoch": 0.8199929103154909, "grad_norm": 3.665153049544116, "learning_rate": 8.260226478106214e-07, "loss": 0.3768, "step": 11566 }, { "epoch": 0.8200638071605814, "grad_norm": 2.6615246263547974, "learning_rate": 8.253906272282258e-07, "loss": 0.3713, "step": 11567 }, { "epoch": 0.8201347040056718, "grad_norm": 2.17231267671711, "learning_rate": 8.24758826776168e-07, "loss": 0.3513, "step": 11568 }, { "epoch": 0.8202056008507621, "grad_norm": 2.354978750804693, "learning_rate": 8.241272464877648e-07, "loss": 0.383, "step": 11569 }, { "epoch": 0.8202764976958525, "grad_norm": 3.522459045902424, "learning_rate": 8.234958863963193e-07, "loss": 0.4162, "step": 11570 }, { "epoch": 0.820347394540943, "grad_norm": 4.125954288046893, "learning_rate": 8.228647465351225e-07, "loss": 0.37, "step": 11571 }, { "epoch": 0.8204182913860333, "grad_norm": 2.6715311167494193, "learning_rate": 8.222338269374569e-07, "loss": 0.3603, "step": 11572 }, { "epoch": 0.8204891882311237, "grad_norm": 2.220562462284317, "learning_rate": 8.216031276365893e-07, "loss": 0.4236, "step": 11573 }, { "epoch": 0.8205600850762141, "grad_norm": 3.316427900272732, "learning_rate": 8.209726486657787e-07, "loss": 0.3803, "step": 11574 }, { "epoch": 0.8206309819213045, "grad_norm": 9.906883546773244, "learning_rate": 8.203423900582697e-07, "loss": 0.3832, "step": 11575 }, { "epoch": 0.8207018787663949, "grad_norm": 2.179192622373021, "learning_rate": 8.197123518472971e-07, "loss": 0.4078, "step": 11576 }, { "epoch": 0.8207727756114853, "grad_norm": 2.4030954536311486, "learning_rate": 8.19082534066083e-07, "loss": 0.3861, "step": 11577 }, { "epoch": 0.8208436724565756, "grad_norm": 7.100288811396575, "learning_rate": 8.184529367478372e-07, "loss": 0.4062, "step": 11578 }, { "epoch": 0.8209145693016661, "grad_norm": 2.8545493099123984, "learning_rate": 8.178235599257594e-07, "loss": 0.3532, "step": 11579 }, { "epoch": 0.8209854661467565, "grad_norm": 2.1093045832760673, "learning_rate": 8.171944036330398e-07, "loss": 0.3786, "step": 11580 }, { "epoch": 0.8210563629918468, "grad_norm": 2.3394019280712555, "learning_rate": 8.165654679028495e-07, "loss": 0.3244, "step": 11581 }, { "epoch": 0.8211272598369372, "grad_norm": 3.7317795173234742, "learning_rate": 8.159367527683553e-07, "loss": 0.3916, "step": 11582 }, { "epoch": 0.8211981566820277, "grad_norm": 3.5263005671124286, "learning_rate": 8.153082582627109e-07, "loss": 0.4043, "step": 11583 }, { "epoch": 0.8212690535271181, "grad_norm": 3.269724526304821, "learning_rate": 8.146799844190556e-07, "loss": 0.3971, "step": 11584 }, { "epoch": 0.8213399503722084, "grad_norm": 5.3838479020983785, "learning_rate": 8.140519312705186e-07, "loss": 0.3443, "step": 11585 }, { "epoch": 0.8214108472172988, "grad_norm": 2.5682031352927543, "learning_rate": 8.134240988502179e-07, "loss": 0.4044, "step": 11586 }, { "epoch": 0.8214817440623893, "grad_norm": 6.379835502271424, "learning_rate": 8.127964871912614e-07, "loss": 0.3877, "step": 11587 }, { "epoch": 0.8215526409074796, "grad_norm": 5.8194887352500375, "learning_rate": 8.121690963267421e-07, "loss": 0.3962, "step": 11588 }, { "epoch": 0.82162353775257, "grad_norm": 2.4974163370927975, "learning_rate": 8.115419262897412e-07, "loss": 0.3677, "step": 11589 }, { "epoch": 0.8216944345976605, "grad_norm": 4.502180284040434, "learning_rate": 8.109149771133329e-07, "loss": 0.4016, "step": 11590 }, { "epoch": 0.8217653314427508, "grad_norm": 2.1575051883510628, "learning_rate": 8.102882488305752e-07, "loss": 0.3702, "step": 11591 }, { "epoch": 0.8218362282878412, "grad_norm": 3.1004904633136947, "learning_rate": 8.096617414745145e-07, "loss": 0.3337, "step": 11592 }, { "epoch": 0.8219071251329316, "grad_norm": 2.354239151771453, "learning_rate": 8.090354550781903e-07, "loss": 0.3746, "step": 11593 }, { "epoch": 0.8219780219780219, "grad_norm": 2.972757819807618, "learning_rate": 8.084093896746237e-07, "loss": 0.3636, "step": 11594 }, { "epoch": 0.8220489188231124, "grad_norm": 4.6103390712651615, "learning_rate": 8.077835452968308e-07, "loss": 0.3419, "step": 11595 }, { "epoch": 0.8221198156682028, "grad_norm": 3.3600409536926934, "learning_rate": 8.071579219778108e-07, "loss": 0.3916, "step": 11596 }, { "epoch": 0.8221907125132931, "grad_norm": 2.4168759440494236, "learning_rate": 8.06532519750553e-07, "loss": 0.3702, "step": 11597 }, { "epoch": 0.8222616093583835, "grad_norm": 4.349991461433409, "learning_rate": 8.059073386480365e-07, "loss": 0.4139, "step": 11598 }, { "epoch": 0.822332506203474, "grad_norm": 2.2800812140124926, "learning_rate": 8.052823787032294e-07, "loss": 0.3609, "step": 11599 }, { "epoch": 0.8224034030485643, "grad_norm": 2.2997882050139378, "learning_rate": 8.046576399490813e-07, "loss": 0.4122, "step": 11600 }, { "epoch": 0.8224742998936547, "grad_norm": 1.9568148650301105, "learning_rate": 8.040331224185383e-07, "loss": 0.3559, "step": 11601 }, { "epoch": 0.8225451967387452, "grad_norm": 2.3293521639630046, "learning_rate": 8.034088261445328e-07, "loss": 0.3715, "step": 11602 }, { "epoch": 0.8226160935838355, "grad_norm": 3.1956039387995654, "learning_rate": 8.027847511599828e-07, "loss": 0.3885, "step": 11603 }, { "epoch": 0.8226869904289259, "grad_norm": 3.0143806248197422, "learning_rate": 8.021608974977952e-07, "loss": 0.3878, "step": 11604 }, { "epoch": 0.8227578872740163, "grad_norm": 4.981322957802372, "learning_rate": 8.015372651908676e-07, "loss": 0.4104, "step": 11605 }, { "epoch": 0.8228287841191066, "grad_norm": 6.593186755025935, "learning_rate": 8.009138542720862e-07, "loss": 0.4092, "step": 11606 }, { "epoch": 0.8228996809641971, "grad_norm": 1.954210388724742, "learning_rate": 8.002906647743219e-07, "loss": 0.4042, "step": 11607 }, { "epoch": 0.8229705778092875, "grad_norm": 5.697058500765892, "learning_rate": 7.996676967304351e-07, "loss": 0.4547, "step": 11608 }, { "epoch": 0.8230414746543778, "grad_norm": 2.901674016638141, "learning_rate": 7.990449501732777e-07, "loss": 0.3532, "step": 11609 }, { "epoch": 0.8231123714994683, "grad_norm": 2.70984665740842, "learning_rate": 7.984224251356865e-07, "loss": 0.3717, "step": 11610 }, { "epoch": 0.8231832683445587, "grad_norm": 4.446956801996336, "learning_rate": 7.978001216504866e-07, "loss": 0.4013, "step": 11611 }, { "epoch": 0.8232541651896491, "grad_norm": 3.6081032824925696, "learning_rate": 7.971780397504953e-07, "loss": 0.3769, "step": 11612 }, { "epoch": 0.8233250620347394, "grad_norm": 2.257678973049913, "learning_rate": 7.965561794685122e-07, "loss": 0.3679, "step": 11613 }, { "epoch": 0.8233959588798299, "grad_norm": 3.7669248427128426, "learning_rate": 7.959345408373315e-07, "loss": 0.3514, "step": 11614 }, { "epoch": 0.8234668557249203, "grad_norm": 3.8263629498618066, "learning_rate": 7.9531312388973e-07, "loss": 0.3965, "step": 11615 }, { "epoch": 0.8235377525700106, "grad_norm": 2.893342849122809, "learning_rate": 7.946919286584782e-07, "loss": 0.3659, "step": 11616 }, { "epoch": 0.823608649415101, "grad_norm": 2.4344730174474707, "learning_rate": 7.940709551763304e-07, "loss": 0.3369, "step": 11617 }, { "epoch": 0.8236795462601915, "grad_norm": 2.8428934647530846, "learning_rate": 7.9345020347603e-07, "loss": 0.3595, "step": 11618 }, { "epoch": 0.8237504431052818, "grad_norm": 2.58424057853878, "learning_rate": 7.92829673590313e-07, "loss": 0.3865, "step": 11619 }, { "epoch": 0.8238213399503722, "grad_norm": 1.9683465712378405, "learning_rate": 7.922093655518964e-07, "loss": 0.3836, "step": 11620 }, { "epoch": 0.8238922367954626, "grad_norm": 2.7323297646952787, "learning_rate": 7.915892793934937e-07, "loss": 0.3542, "step": 11621 }, { "epoch": 0.823963133640553, "grad_norm": 5.243648728335735, "learning_rate": 7.909694151477998e-07, "loss": 0.3566, "step": 11622 }, { "epoch": 0.8240340304856434, "grad_norm": 3.4511585427383684, "learning_rate": 7.903497728475002e-07, "loss": 0.4045, "step": 11623 }, { "epoch": 0.8241049273307338, "grad_norm": 3.073751120994107, "learning_rate": 7.897303525252704e-07, "loss": 0.3656, "step": 11624 }, { "epoch": 0.8241758241758241, "grad_norm": 3.0977786991605685, "learning_rate": 7.891111542137747e-07, "loss": 0.4465, "step": 11625 }, { "epoch": 0.8242467210209146, "grad_norm": 14.684094523736583, "learning_rate": 7.8849217794566e-07, "loss": 0.4231, "step": 11626 }, { "epoch": 0.824317617866005, "grad_norm": 3.3450389466084234, "learning_rate": 7.878734237535668e-07, "loss": 0.3391, "step": 11627 }, { "epoch": 0.8243885147110953, "grad_norm": 4.423019621308905, "learning_rate": 7.872548916701239e-07, "loss": 0.3828, "step": 11628 }, { "epoch": 0.8244594115561857, "grad_norm": 2.7098640650044445, "learning_rate": 7.866365817279464e-07, "loss": 0.3818, "step": 11629 }, { "epoch": 0.8245303084012762, "grad_norm": 5.4874925187950145, "learning_rate": 7.860184939596366e-07, "loss": 0.346, "step": 11630 }, { "epoch": 0.8246012052463665, "grad_norm": 2.1955608664415425, "learning_rate": 7.854006283977883e-07, "loss": 0.4369, "step": 11631 }, { "epoch": 0.8246721020914569, "grad_norm": 3.623681492414072, "learning_rate": 7.847829850749827e-07, "loss": 0.3471, "step": 11632 }, { "epoch": 0.8247429989365473, "grad_norm": 3.3411883768214308, "learning_rate": 7.841655640237877e-07, "loss": 0.3998, "step": 11633 }, { "epoch": 0.8248138957816377, "grad_norm": 3.8055143102325215, "learning_rate": 7.835483652767589e-07, "loss": 0.343, "step": 11634 }, { "epoch": 0.8248847926267281, "grad_norm": 3.157011403090001, "learning_rate": 7.829313888664442e-07, "loss": 0.3564, "step": 11635 }, { "epoch": 0.8249556894718185, "grad_norm": 2.0588770132666183, "learning_rate": 7.823146348253763e-07, "loss": 0.345, "step": 11636 }, { "epoch": 0.825026586316909, "grad_norm": 4.253695705621032, "learning_rate": 7.816981031860754e-07, "loss": 0.3936, "step": 11637 }, { "epoch": 0.8250974831619993, "grad_norm": 3.0984957067884586, "learning_rate": 7.81081793981055e-07, "loss": 0.4144, "step": 11638 }, { "epoch": 0.8251683800070897, "grad_norm": 3.08214501048754, "learning_rate": 7.804657072428101e-07, "loss": 0.3665, "step": 11639 }, { "epoch": 0.8252392768521801, "grad_norm": 2.5054631746856164, "learning_rate": 7.798498430038304e-07, "loss": 0.3612, "step": 11640 }, { "epoch": 0.8253101736972704, "grad_norm": 3.625905615601636, "learning_rate": 7.792342012965898e-07, "loss": 0.3765, "step": 11641 }, { "epoch": 0.8253810705423609, "grad_norm": 3.6574474357068603, "learning_rate": 7.786187821535496e-07, "loss": 0.3778, "step": 11642 }, { "epoch": 0.8254519673874513, "grad_norm": 4.934302501958551, "learning_rate": 7.780035856071633e-07, "loss": 0.3924, "step": 11643 }, { "epoch": 0.8255228642325416, "grad_norm": 3.6657505842073106, "learning_rate": 7.773886116898722e-07, "loss": 0.3845, "step": 11644 }, { "epoch": 0.825593761077632, "grad_norm": 2.5397920096332305, "learning_rate": 7.767738604341018e-07, "loss": 0.3991, "step": 11645 }, { "epoch": 0.8256646579227225, "grad_norm": 1.0931148471692438, "learning_rate": 7.761593318722682e-07, "loss": 0.6605, "step": 11646 }, { "epoch": 0.8257355547678128, "grad_norm": 2.7614680160687355, "learning_rate": 7.755450260367781e-07, "loss": 0.4004, "step": 11647 }, { "epoch": 0.8258064516129032, "grad_norm": 3.0564824701054096, "learning_rate": 7.749309429600232e-07, "loss": 0.3923, "step": 11648 }, { "epoch": 0.8258773484579937, "grad_norm": 2.133219678136151, "learning_rate": 7.743170826743835e-07, "loss": 0.4091, "step": 11649 }, { "epoch": 0.825948245303084, "grad_norm": 8.219445215456314, "learning_rate": 7.737034452122294e-07, "loss": 0.3561, "step": 11650 }, { "epoch": 0.8260191421481744, "grad_norm": 3.386763721761116, "learning_rate": 7.730900306059191e-07, "loss": 0.3751, "step": 11651 }, { "epoch": 0.8260900389932648, "grad_norm": 2.23265000854745, "learning_rate": 7.724768388877985e-07, "loss": 0.3705, "step": 11652 }, { "epoch": 0.8261609358383551, "grad_norm": 2.0109056402442698, "learning_rate": 7.718638700901993e-07, "loss": 0.3184, "step": 11653 }, { "epoch": 0.8262318326834456, "grad_norm": 2.1788463919465606, "learning_rate": 7.712511242454468e-07, "loss": 0.3798, "step": 11654 }, { "epoch": 0.826302729528536, "grad_norm": 3.172081057036096, "learning_rate": 7.706386013858497e-07, "loss": 0.3803, "step": 11655 }, { "epoch": 0.8263736263736263, "grad_norm": 3.180345274670837, "learning_rate": 7.700263015437065e-07, "loss": 0.3824, "step": 11656 }, { "epoch": 0.8264445232187168, "grad_norm": 2.9072548110913083, "learning_rate": 7.694142247513059e-07, "loss": 0.3832, "step": 11657 }, { "epoch": 0.8265154200638072, "grad_norm": 2.5347230796204583, "learning_rate": 7.688023710409215e-07, "loss": 0.4374, "step": 11658 }, { "epoch": 0.8265863169088975, "grad_norm": 2.307530767140682, "learning_rate": 7.681907404448186e-07, "loss": 0.4275, "step": 11659 }, { "epoch": 0.8266572137539879, "grad_norm": 3.271215359060331, "learning_rate": 7.675793329952463e-07, "loss": 0.4011, "step": 11660 }, { "epoch": 0.8267281105990784, "grad_norm": 3.737579309657153, "learning_rate": 7.669681487244479e-07, "loss": 0.4716, "step": 11661 }, { "epoch": 0.8267990074441688, "grad_norm": 2.8108998185939003, "learning_rate": 7.663571876646492e-07, "loss": 0.3563, "step": 11662 }, { "epoch": 0.8268699042892591, "grad_norm": 3.7764828721907957, "learning_rate": 7.657464498480677e-07, "loss": 0.4179, "step": 11663 }, { "epoch": 0.8269408011343495, "grad_norm": 1.0952202663222532, "learning_rate": 7.651359353069082e-07, "loss": 0.6616, "step": 11664 }, { "epoch": 0.82701169797944, "grad_norm": 3.0936215857702325, "learning_rate": 7.645256440733617e-07, "loss": 0.3939, "step": 11665 }, { "epoch": 0.8270825948245303, "grad_norm": 2.5328479707750855, "learning_rate": 7.639155761796125e-07, "loss": 0.4294, "step": 11666 }, { "epoch": 0.8271534916696207, "grad_norm": 2.986781714193911, "learning_rate": 7.633057316578279e-07, "loss": 0.3888, "step": 11667 }, { "epoch": 0.8272243885147111, "grad_norm": 3.2882430508868463, "learning_rate": 7.626961105401648e-07, "loss": 0.4565, "step": 11668 }, { "epoch": 0.8272952853598015, "grad_norm": 2.054808593679912, "learning_rate": 7.620867128587695e-07, "loss": 0.3338, "step": 11669 }, { "epoch": 0.8273661822048919, "grad_norm": 9.0697172050362, "learning_rate": 7.61477538645779e-07, "loss": 0.3877, "step": 11670 }, { "epoch": 0.8274370790499823, "grad_norm": 2.8056705421753367, "learning_rate": 7.608685879333111e-07, "loss": 0.3773, "step": 11671 }, { "epoch": 0.8275079758950726, "grad_norm": 2.581891956807049, "learning_rate": 7.602598607534778e-07, "loss": 0.3969, "step": 11672 }, { "epoch": 0.8275788727401631, "grad_norm": 5.176390243579225, "learning_rate": 7.59651357138379e-07, "loss": 0.4116, "step": 11673 }, { "epoch": 0.8276497695852535, "grad_norm": 2.525135892139217, "learning_rate": 7.590430771201013e-07, "loss": 0.3769, "step": 11674 }, { "epoch": 0.8277206664303438, "grad_norm": 2.5863379771618704, "learning_rate": 7.584350207307173e-07, "loss": 0.4501, "step": 11675 }, { "epoch": 0.8277915632754342, "grad_norm": 3.866223509948356, "learning_rate": 7.578271880022925e-07, "loss": 0.3735, "step": 11676 }, { "epoch": 0.8278624601205247, "grad_norm": 3.06181117463025, "learning_rate": 7.57219578966879e-07, "loss": 0.3831, "step": 11677 }, { "epoch": 0.827933356965615, "grad_norm": 5.229845701225887, "learning_rate": 7.566121936565152e-07, "loss": 0.4081, "step": 11678 }, { "epoch": 0.8280042538107054, "grad_norm": 4.1627494838580335, "learning_rate": 7.560050321032286e-07, "loss": 0.3874, "step": 11679 }, { "epoch": 0.8280751506557958, "grad_norm": 5.654014460969767, "learning_rate": 7.553980943390371e-07, "loss": 0.3686, "step": 11680 }, { "epoch": 0.8281460475008862, "grad_norm": 2.287886845204201, "learning_rate": 7.547913803959434e-07, "loss": 0.3496, "step": 11681 }, { "epoch": 0.8282169443459766, "grad_norm": 2.5994986808892464, "learning_rate": 7.541848903059396e-07, "loss": 0.4131, "step": 11682 }, { "epoch": 0.828287841191067, "grad_norm": 4.961405599494663, "learning_rate": 7.535786241010085e-07, "loss": 0.4052, "step": 11683 }, { "epoch": 0.8283587380361573, "grad_norm": 2.745585418690606, "learning_rate": 7.529725818131162e-07, "loss": 0.4354, "step": 11684 }, { "epoch": 0.8284296348812478, "grad_norm": 3.120907473080804, "learning_rate": 7.523667634742226e-07, "loss": 0.3822, "step": 11685 }, { "epoch": 0.8285005317263382, "grad_norm": 2.861967285795772, "learning_rate": 7.517611691162713e-07, "loss": 0.4187, "step": 11686 }, { "epoch": 0.8285714285714286, "grad_norm": 5.456463124458905, "learning_rate": 7.511557987711954e-07, "loss": 0.3388, "step": 11687 }, { "epoch": 0.8286423254165189, "grad_norm": 2.8789748656198517, "learning_rate": 7.505506524709172e-07, "loss": 0.399, "step": 11688 }, { "epoch": 0.8287132222616094, "grad_norm": 5.082808967913016, "learning_rate": 7.499457302473473e-07, "loss": 0.3799, "step": 11689 }, { "epoch": 0.8287841191066998, "grad_norm": 6.927292986619105, "learning_rate": 7.493410321323829e-07, "loss": 0.4208, "step": 11690 }, { "epoch": 0.8288550159517901, "grad_norm": 1.898063514862738, "learning_rate": 7.487365581579098e-07, "loss": 0.4396, "step": 11691 }, { "epoch": 0.8289259127968805, "grad_norm": 2.7232334881786997, "learning_rate": 7.481323083558035e-07, "loss": 0.3684, "step": 11692 }, { "epoch": 0.828996809641971, "grad_norm": 3.6014233946846677, "learning_rate": 7.475282827579261e-07, "loss": 0.3835, "step": 11693 }, { "epoch": 0.8290677064870613, "grad_norm": 4.631588173200103, "learning_rate": 7.469244813961268e-07, "loss": 0.4195, "step": 11694 }, { "epoch": 0.8291386033321517, "grad_norm": 2.7941145934370457, "learning_rate": 7.463209043022457e-07, "loss": 0.4344, "step": 11695 }, { "epoch": 0.8292095001772422, "grad_norm": 2.770540184876653, "learning_rate": 7.457175515081116e-07, "loss": 0.4555, "step": 11696 }, { "epoch": 0.8292803970223325, "grad_norm": 2.1599833557153305, "learning_rate": 7.451144230455376e-07, "loss": 0.3844, "step": 11697 }, { "epoch": 0.8293512938674229, "grad_norm": 3.24035698807784, "learning_rate": 7.44511518946327e-07, "loss": 0.4143, "step": 11698 }, { "epoch": 0.8294221907125133, "grad_norm": 2.679972406493072, "learning_rate": 7.439088392422728e-07, "loss": 0.3541, "step": 11699 }, { "epoch": 0.8294930875576036, "grad_norm": 4.2313120103304795, "learning_rate": 7.433063839651545e-07, "loss": 0.4189, "step": 11700 }, { "epoch": 0.8295639844026941, "grad_norm": 2.0223683832924455, "learning_rate": 7.427041531467383e-07, "loss": 0.4087, "step": 11701 }, { "epoch": 0.8296348812477845, "grad_norm": 2.880558687765957, "learning_rate": 7.421021468187822e-07, "loss": 0.3607, "step": 11702 }, { "epoch": 0.8297057780928748, "grad_norm": 4.121901296329311, "learning_rate": 7.415003650130292e-07, "loss": 0.4158, "step": 11703 }, { "epoch": 0.8297766749379653, "grad_norm": 4.476318215232743, "learning_rate": 7.408988077612134e-07, "loss": 0.3553, "step": 11704 }, { "epoch": 0.8298475717830557, "grad_norm": 4.130811442148341, "learning_rate": 7.402974750950532e-07, "loss": 0.4089, "step": 11705 }, { "epoch": 0.829918468628146, "grad_norm": 2.9092586171836334, "learning_rate": 7.396963670462593e-07, "loss": 0.3409, "step": 11706 }, { "epoch": 0.8299893654732364, "grad_norm": 4.035815239256117, "learning_rate": 7.390954836465263e-07, "loss": 0.3926, "step": 11707 }, { "epoch": 0.8300602623183269, "grad_norm": 3.1510431908956003, "learning_rate": 7.384948249275425e-07, "loss": 0.3673, "step": 11708 }, { "epoch": 0.8301311591634172, "grad_norm": 4.0319582449199824, "learning_rate": 7.378943909209791e-07, "loss": 0.3634, "step": 11709 }, { "epoch": 0.8302020560085076, "grad_norm": 2.402943798150778, "learning_rate": 7.372941816584961e-07, "loss": 0.3633, "step": 11710 }, { "epoch": 0.830272952853598, "grad_norm": 2.9622936700631977, "learning_rate": 7.366941971717457e-07, "loss": 0.4105, "step": 11711 }, { "epoch": 0.8303438496986885, "grad_norm": 2.2630043972989737, "learning_rate": 7.360944374923646e-07, "loss": 0.3822, "step": 11712 }, { "epoch": 0.8304147465437788, "grad_norm": 5.963292674148335, "learning_rate": 7.354949026519764e-07, "loss": 0.3556, "step": 11713 }, { "epoch": 0.8304856433888692, "grad_norm": 2.4972577944773082, "learning_rate": 7.348955926821976e-07, "loss": 0.3898, "step": 11714 }, { "epoch": 0.8305565402339596, "grad_norm": 3.574603745276263, "learning_rate": 7.342965076146308e-07, "loss": 0.341, "step": 11715 }, { "epoch": 0.83062743707905, "grad_norm": 2.8415375270670284, "learning_rate": 7.336976474808649e-07, "loss": 0.407, "step": 11716 }, { "epoch": 0.8306983339241404, "grad_norm": 3.2338805255309127, "learning_rate": 7.330990123124771e-07, "loss": 0.3968, "step": 11717 }, { "epoch": 0.8307692307692308, "grad_norm": 5.2270660572999805, "learning_rate": 7.325006021410353e-07, "loss": 0.3775, "step": 11718 }, { "epoch": 0.8308401276143211, "grad_norm": 4.088177489470425, "learning_rate": 7.31902416998096e-07, "loss": 0.3504, "step": 11719 }, { "epoch": 0.8309110244594116, "grad_norm": 3.1769397848623924, "learning_rate": 7.313044569151978e-07, "loss": 0.429, "step": 11720 }, { "epoch": 0.830981921304502, "grad_norm": 1.0443039990205287, "learning_rate": 7.30706721923874e-07, "loss": 0.7312, "step": 11721 }, { "epoch": 0.8310528181495923, "grad_norm": 2.785540868163895, "learning_rate": 7.301092120556441e-07, "loss": 0.3589, "step": 11722 }, { "epoch": 0.8311237149946827, "grad_norm": 3.1410521107036087, "learning_rate": 7.295119273420148e-07, "loss": 0.3395, "step": 11723 }, { "epoch": 0.8311946118397732, "grad_norm": 2.881376086243491, "learning_rate": 7.289148678144797e-07, "loss": 0.408, "step": 11724 }, { "epoch": 0.8312655086848635, "grad_norm": 2.9574448930086974, "learning_rate": 7.283180335045248e-07, "loss": 0.3572, "step": 11725 }, { "epoch": 0.8313364055299539, "grad_norm": 2.296825334329151, "learning_rate": 7.277214244436204e-07, "loss": 0.386, "step": 11726 }, { "epoch": 0.8314073023750443, "grad_norm": 4.1926551066492825, "learning_rate": 7.271250406632252e-07, "loss": 0.3674, "step": 11727 }, { "epoch": 0.8314781992201347, "grad_norm": 2.161866604927363, "learning_rate": 7.265288821947891e-07, "loss": 0.3914, "step": 11728 }, { "epoch": 0.8315490960652251, "grad_norm": 2.25069824272051, "learning_rate": 7.259329490697459e-07, "loss": 0.4171, "step": 11729 }, { "epoch": 0.8316199929103155, "grad_norm": 2.2253036975359897, "learning_rate": 7.253372413195209e-07, "loss": 0.3748, "step": 11730 }, { "epoch": 0.8316908897554058, "grad_norm": 2.987506808249861, "learning_rate": 7.247417589755268e-07, "loss": 0.4004, "step": 11731 }, { "epoch": 0.8317617866004963, "grad_norm": 3.3371741074041106, "learning_rate": 7.241465020691613e-07, "loss": 0.4193, "step": 11732 }, { "epoch": 0.8318326834455867, "grad_norm": 1.6071161559398144, "learning_rate": 7.235514706318142e-07, "loss": 0.375, "step": 11733 }, { "epoch": 0.831903580290677, "grad_norm": 5.1439861596725365, "learning_rate": 7.229566646948633e-07, "loss": 0.3888, "step": 11734 }, { "epoch": 0.8319744771357674, "grad_norm": 2.203654213431938, "learning_rate": 7.223620842896722e-07, "loss": 0.3676, "step": 11735 }, { "epoch": 0.8320453739808579, "grad_norm": 2.2792131103128455, "learning_rate": 7.217677294475922e-07, "loss": 0.4199, "step": 11736 }, { "epoch": 0.8321162708259482, "grad_norm": 2.2995284012726325, "learning_rate": 7.211736001999664e-07, "loss": 0.4087, "step": 11737 }, { "epoch": 0.8321871676710386, "grad_norm": 3.896044393252404, "learning_rate": 7.205796965781225e-07, "loss": 0.381, "step": 11738 }, { "epoch": 0.832258064516129, "grad_norm": 3.1123925284255196, "learning_rate": 7.199860186133761e-07, "loss": 0.3434, "step": 11739 }, { "epoch": 0.8323289613612195, "grad_norm": 2.5862497509917763, "learning_rate": 7.193925663370339e-07, "loss": 0.3476, "step": 11740 }, { "epoch": 0.8323998582063098, "grad_norm": 3.27006868391829, "learning_rate": 7.187993397803906e-07, "loss": 0.3351, "step": 11741 }, { "epoch": 0.8324707550514002, "grad_norm": 2.120965914350263, "learning_rate": 7.182063389747252e-07, "loss": 0.3612, "step": 11742 }, { "epoch": 0.8325416518964907, "grad_norm": 10.765165194667471, "learning_rate": 7.176135639513065e-07, "loss": 0.3676, "step": 11743 }, { "epoch": 0.832612548741581, "grad_norm": 2.0694998104808846, "learning_rate": 7.170210147413948e-07, "loss": 0.3953, "step": 11744 }, { "epoch": 0.8326834455866714, "grad_norm": 3.1794875911879403, "learning_rate": 7.164286913762336e-07, "loss": 0.3779, "step": 11745 }, { "epoch": 0.8327543424317618, "grad_norm": 3.5582982905608582, "learning_rate": 7.15836593887056e-07, "loss": 0.3868, "step": 11746 }, { "epoch": 0.8328252392768521, "grad_norm": 2.875368106798703, "learning_rate": 7.152447223050846e-07, "loss": 0.3903, "step": 11747 }, { "epoch": 0.8328961361219426, "grad_norm": 5.124370744650477, "learning_rate": 7.146530766615306e-07, "loss": 0.4301, "step": 11748 }, { "epoch": 0.832967032967033, "grad_norm": 6.694112286679663, "learning_rate": 7.140616569875913e-07, "loss": 0.3849, "step": 11749 }, { "epoch": 0.8330379298121233, "grad_norm": 1.0245550669682242, "learning_rate": 7.134704633144501e-07, "loss": 0.6566, "step": 11750 }, { "epoch": 0.8331088266572138, "grad_norm": 2.7527186768772456, "learning_rate": 7.128794956732848e-07, "loss": 0.4097, "step": 11751 }, { "epoch": 0.8331797235023042, "grad_norm": 2.52490332765136, "learning_rate": 7.12288754095255e-07, "loss": 0.4208, "step": 11752 }, { "epoch": 0.8332506203473945, "grad_norm": 1.5949796462648613, "learning_rate": 7.116982386115123e-07, "loss": 0.3825, "step": 11753 }, { "epoch": 0.8333215171924849, "grad_norm": 2.8157890465573643, "learning_rate": 7.111079492531953e-07, "loss": 0.399, "step": 11754 }, { "epoch": 0.8333924140375754, "grad_norm": 2.7120899279369413, "learning_rate": 7.105178860514283e-07, "loss": 0.3796, "step": 11755 }, { "epoch": 0.8334633108826657, "grad_norm": 6.2311013928099594, "learning_rate": 7.099280490373279e-07, "loss": 0.34, "step": 11756 }, { "epoch": 0.8335342077277561, "grad_norm": 4.859458421970635, "learning_rate": 7.093384382419966e-07, "loss": 0.3928, "step": 11757 }, { "epoch": 0.8336051045728465, "grad_norm": 2.5877465032276548, "learning_rate": 7.087490536965224e-07, "loss": 0.3656, "step": 11758 }, { "epoch": 0.8336760014179369, "grad_norm": 1.99509247050934, "learning_rate": 7.081598954319869e-07, "loss": 0.3827, "step": 11759 }, { "epoch": 0.8337468982630273, "grad_norm": 3.253786690151941, "learning_rate": 7.075709634794565e-07, "loss": 0.3804, "step": 11760 }, { "epoch": 0.8338177951081177, "grad_norm": 3.711416208730111, "learning_rate": 7.069822578699853e-07, "loss": 0.3768, "step": 11761 }, { "epoch": 0.833888691953208, "grad_norm": 4.7304228224322955, "learning_rate": 7.063937786346159e-07, "loss": 0.393, "step": 11762 }, { "epoch": 0.8339595887982985, "grad_norm": 2.99602090314548, "learning_rate": 7.058055258043794e-07, "loss": 0.3924, "step": 11763 }, { "epoch": 0.8340304856433889, "grad_norm": 6.994493334631682, "learning_rate": 7.052174994102972e-07, "loss": 0.3743, "step": 11764 }, { "epoch": 0.8341013824884793, "grad_norm": 2.136395431404228, "learning_rate": 7.046296994833723e-07, "loss": 0.3608, "step": 11765 }, { "epoch": 0.8341722793335696, "grad_norm": 3.011435794660662, "learning_rate": 7.040421260546015e-07, "loss": 0.3816, "step": 11766 }, { "epoch": 0.8342431761786601, "grad_norm": 10.350460371707422, "learning_rate": 7.034547791549695e-07, "loss": 0.404, "step": 11767 }, { "epoch": 0.8343140730237505, "grad_norm": 2.227134055868359, "learning_rate": 7.028676588154471e-07, "loss": 0.3837, "step": 11768 }, { "epoch": 0.8343849698688408, "grad_norm": 1.0197957253449879, "learning_rate": 7.02280765066991e-07, "loss": 0.6609, "step": 11769 }, { "epoch": 0.8344558667139312, "grad_norm": 4.022457342016143, "learning_rate": 7.016940979405518e-07, "loss": 0.3742, "step": 11770 }, { "epoch": 0.8345267635590217, "grad_norm": 3.1148213258309316, "learning_rate": 7.011076574670627e-07, "loss": 0.4124, "step": 11771 }, { "epoch": 0.834597660404112, "grad_norm": 2.276852481362487, "learning_rate": 7.005214436774488e-07, "loss": 0.3703, "step": 11772 }, { "epoch": 0.8346685572492024, "grad_norm": 2.214658560700111, "learning_rate": 6.999354566026212e-07, "loss": 0.4018, "step": 11773 }, { "epoch": 0.8347394540942928, "grad_norm": 1.8253113722905117, "learning_rate": 6.993496962734775e-07, "loss": 0.3925, "step": 11774 }, { "epoch": 0.8348103509393832, "grad_norm": 12.210228974727725, "learning_rate": 6.987641627209079e-07, "loss": 0.3924, "step": 11775 }, { "epoch": 0.8348812477844736, "grad_norm": 3.4149274139352563, "learning_rate": 6.981788559757863e-07, "loss": 0.3855, "step": 11776 }, { "epoch": 0.834952144629564, "grad_norm": 2.3787446762574005, "learning_rate": 6.975937760689783e-07, "loss": 0.4027, "step": 11777 }, { "epoch": 0.8350230414746543, "grad_norm": 3.6010420992572874, "learning_rate": 6.970089230313326e-07, "loss": 0.3902, "step": 11778 }, { "epoch": 0.8350939383197448, "grad_norm": 2.4635078218977844, "learning_rate": 6.964242968936925e-07, "loss": 0.355, "step": 11779 }, { "epoch": 0.8351648351648352, "grad_norm": 3.954082773558522, "learning_rate": 6.958398976868835e-07, "loss": 0.3598, "step": 11780 }, { "epoch": 0.8352357320099255, "grad_norm": 9.72338425856971, "learning_rate": 6.952557254417208e-07, "loss": 0.449, "step": 11781 }, { "epoch": 0.8353066288550159, "grad_norm": 1.0039023501838467, "learning_rate": 6.946717801890108e-07, "loss": 0.6549, "step": 11782 }, { "epoch": 0.8353775257001064, "grad_norm": 2.359901666291681, "learning_rate": 6.940880619595436e-07, "loss": 0.3609, "step": 11783 }, { "epoch": 0.8354484225451967, "grad_norm": 4.5602584259839, "learning_rate": 6.935045707840987e-07, "loss": 0.3618, "step": 11784 }, { "epoch": 0.8355193193902871, "grad_norm": 3.5636412733404303, "learning_rate": 6.929213066934449e-07, "loss": 0.3692, "step": 11785 }, { "epoch": 0.8355902162353775, "grad_norm": 2.5726927256223386, "learning_rate": 6.923382697183384e-07, "loss": 0.3955, "step": 11786 }, { "epoch": 0.8356611130804679, "grad_norm": 7.0661259929441265, "learning_rate": 6.917554598895237e-07, "loss": 0.3808, "step": 11787 }, { "epoch": 0.8357320099255583, "grad_norm": 3.184146052939567, "learning_rate": 6.911728772377302e-07, "loss": 0.3951, "step": 11788 }, { "epoch": 0.8358029067706487, "grad_norm": 2.3917801306836495, "learning_rate": 6.905905217936809e-07, "loss": 0.39, "step": 11789 }, { "epoch": 0.8358738036157392, "grad_norm": 2.7352436417034642, "learning_rate": 6.90008393588083e-07, "loss": 0.4005, "step": 11790 }, { "epoch": 0.8359447004608295, "grad_norm": 1.6911981229926472, "learning_rate": 6.894264926516303e-07, "loss": 0.3678, "step": 11791 }, { "epoch": 0.8360155973059199, "grad_norm": 4.112160653865726, "learning_rate": 6.888448190150093e-07, "loss": 0.387, "step": 11792 }, { "epoch": 0.8360864941510103, "grad_norm": 6.258698319514365, "learning_rate": 6.882633727088927e-07, "loss": 0.4099, "step": 11793 }, { "epoch": 0.8361573909961006, "grad_norm": 2.6233306445222486, "learning_rate": 6.876821537639394e-07, "loss": 0.4028, "step": 11794 }, { "epoch": 0.8362282878411911, "grad_norm": 2.3448147458061275, "learning_rate": 6.871011622107965e-07, "loss": 0.3787, "step": 11795 }, { "epoch": 0.8362991846862815, "grad_norm": 2.044972092991318, "learning_rate": 6.865203980801022e-07, "loss": 0.3409, "step": 11796 }, { "epoch": 0.8363700815313718, "grad_norm": 2.6973990486463575, "learning_rate": 6.859398614024781e-07, "loss": 0.4315, "step": 11797 }, { "epoch": 0.8364409783764623, "grad_norm": 3.4776786993387736, "learning_rate": 6.853595522085399e-07, "loss": 0.3872, "step": 11798 }, { "epoch": 0.8365118752215527, "grad_norm": 2.6256222686646513, "learning_rate": 6.847794705288852e-07, "loss": 0.3581, "step": 11799 }, { "epoch": 0.836582772066643, "grad_norm": 3.866992820403618, "learning_rate": 6.841996163941012e-07, "loss": 0.3427, "step": 11800 }, { "epoch": 0.8366536689117334, "grad_norm": 4.41306534933814, "learning_rate": 6.836199898347673e-07, "loss": 0.3855, "step": 11801 }, { "epoch": 0.8367245657568239, "grad_norm": 4.942055857111354, "learning_rate": 6.830405908814458e-07, "loss": 0.3655, "step": 11802 }, { "epoch": 0.8367954626019142, "grad_norm": 2.2994261899032287, "learning_rate": 6.824614195646873e-07, "loss": 0.383, "step": 11803 }, { "epoch": 0.8368663594470046, "grad_norm": 2.42610816453095, "learning_rate": 6.818824759150339e-07, "loss": 0.3484, "step": 11804 }, { "epoch": 0.836937256292095, "grad_norm": 3.6757648922964896, "learning_rate": 6.813037599630145e-07, "loss": 0.3925, "step": 11805 }, { "epoch": 0.8370081531371854, "grad_norm": 3.2899633337428633, "learning_rate": 6.807252717391443e-07, "loss": 0.4136, "step": 11806 }, { "epoch": 0.8370790499822758, "grad_norm": 3.6142958526355424, "learning_rate": 6.801470112739261e-07, "loss": 0.4302, "step": 11807 }, { "epoch": 0.8371499468273662, "grad_norm": 1.739223814656366, "learning_rate": 6.795689785978532e-07, "loss": 0.4164, "step": 11808 }, { "epoch": 0.8372208436724565, "grad_norm": 2.2012299447473356, "learning_rate": 6.789911737414079e-07, "loss": 0.3843, "step": 11809 }, { "epoch": 0.837291740517547, "grad_norm": 2.672292921123381, "learning_rate": 6.784135967350536e-07, "loss": 0.3788, "step": 11810 }, { "epoch": 0.8373626373626374, "grad_norm": 2.3362363203039678, "learning_rate": 6.778362476092493e-07, "loss": 0.3814, "step": 11811 }, { "epoch": 0.8374335342077277, "grad_norm": 4.62744925401355, "learning_rate": 6.772591263944394e-07, "loss": 0.3461, "step": 11812 }, { "epoch": 0.8375044310528181, "grad_norm": 5.815499028594005, "learning_rate": 6.766822331210555e-07, "loss": 0.3495, "step": 11813 }, { "epoch": 0.8375753278979086, "grad_norm": 3.6656070659397244, "learning_rate": 6.761055678195161e-07, "loss": 0.3351, "step": 11814 }, { "epoch": 0.837646224742999, "grad_norm": 1.9636025773141443, "learning_rate": 6.75529130520231e-07, "loss": 0.3359, "step": 11815 }, { "epoch": 0.8377171215880893, "grad_norm": 4.328510117261925, "learning_rate": 6.749529212535949e-07, "loss": 0.3638, "step": 11816 }, { "epoch": 0.8377880184331797, "grad_norm": 2.873320613749568, "learning_rate": 6.743769400499933e-07, "loss": 0.3787, "step": 11817 }, { "epoch": 0.8378589152782702, "grad_norm": 3.8002472555295377, "learning_rate": 6.738011869397975e-07, "loss": 0.3652, "step": 11818 }, { "epoch": 0.8379298121233605, "grad_norm": 10.57395383624398, "learning_rate": 6.732256619533662e-07, "loss": 0.4299, "step": 11819 }, { "epoch": 0.8380007089684509, "grad_norm": 3.094956736616375, "learning_rate": 6.726503651210486e-07, "loss": 0.3752, "step": 11820 }, { "epoch": 0.8380716058135413, "grad_norm": 3.0696218393370556, "learning_rate": 6.720752964731797e-07, "loss": 0.4232, "step": 11821 }, { "epoch": 0.8381425026586317, "grad_norm": 2.9284702841674455, "learning_rate": 6.715004560400845e-07, "loss": 0.4073, "step": 11822 }, { "epoch": 0.8382133995037221, "grad_norm": 19.157419836246397, "learning_rate": 6.709258438520733e-07, "loss": 0.3674, "step": 11823 }, { "epoch": 0.8382842963488125, "grad_norm": 1.9883486665660794, "learning_rate": 6.703514599394478e-07, "loss": 0.321, "step": 11824 }, { "epoch": 0.8383551931939028, "grad_norm": 2.9574475115884273, "learning_rate": 6.69777304332494e-07, "loss": 0.4032, "step": 11825 }, { "epoch": 0.8384260900389933, "grad_norm": 5.299075610517412, "learning_rate": 6.692033770614869e-07, "loss": 0.4316, "step": 11826 }, { "epoch": 0.8384969868840837, "grad_norm": 3.725041644899337, "learning_rate": 6.68629678156692e-07, "loss": 0.4404, "step": 11827 }, { "epoch": 0.838567883729174, "grad_norm": 2.599281633571061, "learning_rate": 6.680562076483615e-07, "loss": 0.3786, "step": 11828 }, { "epoch": 0.8386387805742644, "grad_norm": 1.936342307839413, "learning_rate": 6.67482965566732e-07, "loss": 0.4081, "step": 11829 }, { "epoch": 0.8387096774193549, "grad_norm": 2.353281896104394, "learning_rate": 6.669099519420325e-07, "loss": 0.3815, "step": 11830 }, { "epoch": 0.8387805742644452, "grad_norm": 0.9726687796538791, "learning_rate": 6.663371668044794e-07, "loss": 0.6562, "step": 11831 }, { "epoch": 0.8388514711095356, "grad_norm": 3.841399822440774, "learning_rate": 6.657646101842752e-07, "loss": 0.3583, "step": 11832 }, { "epoch": 0.838922367954626, "grad_norm": 5.4540976505218275, "learning_rate": 6.651922821116102e-07, "loss": 0.3352, "step": 11833 }, { "epoch": 0.8389932647997164, "grad_norm": 3.3595031042393466, "learning_rate": 6.646201826166665e-07, "loss": 0.3735, "step": 11834 }, { "epoch": 0.8390641616448068, "grad_norm": 2.1600714509217696, "learning_rate": 6.640483117296087e-07, "loss": 0.3798, "step": 11835 }, { "epoch": 0.8391350584898972, "grad_norm": 2.440594012725561, "learning_rate": 6.634766694805921e-07, "loss": 0.4204, "step": 11836 }, { "epoch": 0.8392059553349875, "grad_norm": 2.4974432495839793, "learning_rate": 6.629052558997612e-07, "loss": 0.369, "step": 11837 }, { "epoch": 0.839276852180078, "grad_norm": 5.748775365155301, "learning_rate": 6.623340710172471e-07, "loss": 0.4414, "step": 11838 }, { "epoch": 0.8393477490251684, "grad_norm": 4.133283956523825, "learning_rate": 6.617631148631682e-07, "loss": 0.3585, "step": 11839 }, { "epoch": 0.8394186458702588, "grad_norm": 5.77762829140209, "learning_rate": 6.611923874676307e-07, "loss": 0.3779, "step": 11840 }, { "epoch": 0.8394895427153491, "grad_norm": 2.2586611419928535, "learning_rate": 6.606218888607307e-07, "loss": 0.3984, "step": 11841 }, { "epoch": 0.8395604395604396, "grad_norm": 2.4899834896305633, "learning_rate": 6.600516190725504e-07, "loss": 0.393, "step": 11842 }, { "epoch": 0.83963133640553, "grad_norm": 2.3522935791967328, "learning_rate": 6.59481578133161e-07, "loss": 0.4105, "step": 11843 }, { "epoch": 0.8397022332506203, "grad_norm": 6.441465757780926, "learning_rate": 6.589117660726213e-07, "loss": 0.3419, "step": 11844 }, { "epoch": 0.8397731300957108, "grad_norm": 1.6440295703576133, "learning_rate": 6.583421829209768e-07, "loss": 0.3803, "step": 11845 }, { "epoch": 0.8398440269408012, "grad_norm": 5.596855730067703, "learning_rate": 6.577728287082635e-07, "loss": 0.4069, "step": 11846 }, { "epoch": 0.8399149237858915, "grad_norm": 6.634004752689254, "learning_rate": 6.572037034645035e-07, "loss": 0.3805, "step": 11847 }, { "epoch": 0.8399858206309819, "grad_norm": 3.0312011551592035, "learning_rate": 6.566348072197054e-07, "loss": 0.349, "step": 11848 }, { "epoch": 0.8400567174760724, "grad_norm": 3.386011856673343, "learning_rate": 6.560661400038693e-07, "loss": 0.3667, "step": 11849 }, { "epoch": 0.8401276143211627, "grad_norm": 2.192672266863389, "learning_rate": 6.554977018469827e-07, "loss": 0.3847, "step": 11850 }, { "epoch": 0.8401985111662531, "grad_norm": 2.8178865322050135, "learning_rate": 6.549294927790178e-07, "loss": 0.4029, "step": 11851 }, { "epoch": 0.8402694080113435, "grad_norm": 5.564992592036709, "learning_rate": 6.543615128299363e-07, "loss": 0.3196, "step": 11852 }, { "epoch": 0.8403403048564339, "grad_norm": 3.0628500837605275, "learning_rate": 6.537937620296892e-07, "loss": 0.3724, "step": 11853 }, { "epoch": 0.8404112017015243, "grad_norm": 5.270175289676022, "learning_rate": 6.532262404082163e-07, "loss": 0.3951, "step": 11854 }, { "epoch": 0.8404820985466147, "grad_norm": 5.067192297631987, "learning_rate": 6.526589479954392e-07, "loss": 0.3834, "step": 11855 }, { "epoch": 0.840552995391705, "grad_norm": 3.292478319665185, "learning_rate": 6.520918848212743e-07, "loss": 0.4078, "step": 11856 }, { "epoch": 0.8406238922367955, "grad_norm": 2.8486351601131883, "learning_rate": 6.515250509156235e-07, "loss": 0.3771, "step": 11857 }, { "epoch": 0.8406947890818859, "grad_norm": 2.51522656254088, "learning_rate": 6.509584463083762e-07, "loss": 0.4132, "step": 11858 }, { "epoch": 0.8407656859269762, "grad_norm": 2.7542286826193845, "learning_rate": 6.503920710294081e-07, "loss": 0.4041, "step": 11859 }, { "epoch": 0.8408365827720666, "grad_norm": 3.0162444829729993, "learning_rate": 6.498259251085875e-07, "loss": 0.3714, "step": 11860 }, { "epoch": 0.8409074796171571, "grad_norm": 2.50571133305944, "learning_rate": 6.49260008575765e-07, "loss": 0.4048, "step": 11861 }, { "epoch": 0.8409783764622474, "grad_norm": 3.3897627288120575, "learning_rate": 6.486943214607838e-07, "loss": 0.4286, "step": 11862 }, { "epoch": 0.8410492733073378, "grad_norm": 3.4978007287912845, "learning_rate": 6.481288637934724e-07, "loss": 0.3977, "step": 11863 }, { "epoch": 0.8411201701524282, "grad_norm": 2.041948823784223, "learning_rate": 6.475636356036469e-07, "loss": 0.3547, "step": 11864 }, { "epoch": 0.8411910669975186, "grad_norm": 3.176338399634628, "learning_rate": 6.46998636921114e-07, "loss": 0.3606, "step": 11865 }, { "epoch": 0.841261963842609, "grad_norm": 4.753108147878254, "learning_rate": 6.464338677756637e-07, "loss": 0.3769, "step": 11866 }, { "epoch": 0.8413328606876994, "grad_norm": 3.583801470054296, "learning_rate": 6.458693281970802e-07, "loss": 0.39, "step": 11867 }, { "epoch": 0.8414037575327898, "grad_norm": 4.205733991422201, "learning_rate": 6.453050182151294e-07, "loss": 0.4119, "step": 11868 }, { "epoch": 0.8414746543778802, "grad_norm": 6.700254415080033, "learning_rate": 6.447409378595692e-07, "loss": 0.4333, "step": 11869 }, { "epoch": 0.8415455512229706, "grad_norm": 1.9775153261614122, "learning_rate": 6.441770871601439e-07, "loss": 0.449, "step": 11870 }, { "epoch": 0.841616448068061, "grad_norm": 2.254908673987422, "learning_rate": 6.436134661465848e-07, "loss": 0.4056, "step": 11871 }, { "epoch": 0.8416873449131513, "grad_norm": 3.7328904962095812, "learning_rate": 6.430500748486118e-07, "loss": 0.3754, "step": 11872 }, { "epoch": 0.8417582417582418, "grad_norm": 18.469193965260267, "learning_rate": 6.424869132959366e-07, "loss": 0.3718, "step": 11873 }, { "epoch": 0.8418291386033322, "grad_norm": 3.078219611869783, "learning_rate": 6.419239815182504e-07, "loss": 0.4025, "step": 11874 }, { "epoch": 0.8419000354484225, "grad_norm": 3.143158065071129, "learning_rate": 6.41361279545239e-07, "loss": 0.3903, "step": 11875 }, { "epoch": 0.8419709322935129, "grad_norm": 3.0178116140280773, "learning_rate": 6.407988074065746e-07, "loss": 0.4035, "step": 11876 }, { "epoch": 0.8420418291386034, "grad_norm": 3.1151712373429388, "learning_rate": 6.40236565131917e-07, "loss": 0.4576, "step": 11877 }, { "epoch": 0.8421127259836937, "grad_norm": 2.7254567459119374, "learning_rate": 6.396745527509112e-07, "loss": 0.3943, "step": 11878 }, { "epoch": 0.8421836228287841, "grad_norm": 2.4634781393755145, "learning_rate": 6.391127702931948e-07, "loss": 0.3753, "step": 11879 }, { "epoch": 0.8422545196738745, "grad_norm": 4.210956611506751, "learning_rate": 6.385512177883918e-07, "loss": 0.3823, "step": 11880 }, { "epoch": 0.8423254165189649, "grad_norm": 2.358313202518975, "learning_rate": 6.379898952661118e-07, "loss": 0.3654, "step": 11881 }, { "epoch": 0.8423963133640553, "grad_norm": 2.8231562775094816, "learning_rate": 6.374288027559528e-07, "loss": 0.3906, "step": 11882 }, { "epoch": 0.8424672102091457, "grad_norm": 2.7644144225670506, "learning_rate": 6.368679402875039e-07, "loss": 0.4284, "step": 11883 }, { "epoch": 0.842538107054236, "grad_norm": 4.028897161925092, "learning_rate": 6.363073078903392e-07, "loss": 0.4036, "step": 11884 }, { "epoch": 0.8426090038993265, "grad_norm": 2.0970602117634893, "learning_rate": 6.357469055940191e-07, "loss": 0.3313, "step": 11885 }, { "epoch": 0.8426799007444169, "grad_norm": 4.426415459043875, "learning_rate": 6.35186733428097e-07, "loss": 0.4289, "step": 11886 }, { "epoch": 0.8427507975895072, "grad_norm": 4.477995425310677, "learning_rate": 6.346267914221094e-07, "loss": 0.3735, "step": 11887 }, { "epoch": 0.8428216944345976, "grad_norm": 2.9838181066804053, "learning_rate": 6.340670796055837e-07, "loss": 0.3621, "step": 11888 }, { "epoch": 0.8428925912796881, "grad_norm": 2.5680502597010553, "learning_rate": 6.335075980080329e-07, "loss": 0.3332, "step": 11889 }, { "epoch": 0.8429634881247784, "grad_norm": 3.57760604020476, "learning_rate": 6.329483466589586e-07, "loss": 0.3693, "step": 11890 }, { "epoch": 0.8430343849698688, "grad_norm": 3.139865919854295, "learning_rate": 6.32389325587851e-07, "loss": 0.3993, "step": 11891 }, { "epoch": 0.8431052818149593, "grad_norm": 2.866932406949758, "learning_rate": 6.318305348241904e-07, "loss": 0.3965, "step": 11892 }, { "epoch": 0.8431761786600497, "grad_norm": 2.7830530281714396, "learning_rate": 6.312719743974372e-07, "loss": 0.3628, "step": 11893 }, { "epoch": 0.84324707550514, "grad_norm": 2.5387090439617475, "learning_rate": 6.307136443370476e-07, "loss": 0.4254, "step": 11894 }, { "epoch": 0.8433179723502304, "grad_norm": 7.4144559412565725, "learning_rate": 6.301555446724633e-07, "loss": 0.3435, "step": 11895 }, { "epoch": 0.8433888691953209, "grad_norm": 3.8191143715181726, "learning_rate": 6.295976754331129e-07, "loss": 0.4024, "step": 11896 }, { "epoch": 0.8434597660404112, "grad_norm": 2.6584308807874444, "learning_rate": 6.290400366484111e-07, "loss": 0.39, "step": 11897 }, { "epoch": 0.8435306628855016, "grad_norm": 3.5679138424572665, "learning_rate": 6.284826283477652e-07, "loss": 0.3401, "step": 11898 }, { "epoch": 0.843601559730592, "grad_norm": 2.994001651671421, "learning_rate": 6.27925450560568e-07, "loss": 0.3426, "step": 11899 }, { "epoch": 0.8436724565756824, "grad_norm": 7.976765848106166, "learning_rate": 6.273685033161975e-07, "loss": 0.4052, "step": 11900 }, { "epoch": 0.8437433534207728, "grad_norm": 6.01176303049626, "learning_rate": 6.268117866440232e-07, "loss": 0.3481, "step": 11901 }, { "epoch": 0.8438142502658632, "grad_norm": 2.9934939707851003, "learning_rate": 6.262553005734023e-07, "loss": 0.3646, "step": 11902 }, { "epoch": 0.8438851471109535, "grad_norm": 20.3035711000554, "learning_rate": 6.256990451336775e-07, "loss": 0.3698, "step": 11903 }, { "epoch": 0.843956043956044, "grad_norm": 2.892639662968782, "learning_rate": 6.251430203541803e-07, "loss": 0.4411, "step": 11904 }, { "epoch": 0.8440269408011344, "grad_norm": 2.2983266638134223, "learning_rate": 6.245872262642317e-07, "loss": 0.4001, "step": 11905 }, { "epoch": 0.8440978376462247, "grad_norm": 2.7860265105130053, "learning_rate": 6.24031662893137e-07, "loss": 0.3902, "step": 11906 }, { "epoch": 0.8441687344913151, "grad_norm": 4.153855503434697, "learning_rate": 6.234763302701935e-07, "loss": 0.4106, "step": 11907 }, { "epoch": 0.8442396313364056, "grad_norm": 0.9978130268416876, "learning_rate": 6.229212284246838e-07, "loss": 0.6763, "step": 11908 }, { "epoch": 0.8443105281814959, "grad_norm": 4.162041278378059, "learning_rate": 6.223663573858779e-07, "loss": 0.3853, "step": 11909 }, { "epoch": 0.8443814250265863, "grad_norm": 3.320499224870894, "learning_rate": 6.218117171830363e-07, "loss": 0.3502, "step": 11910 }, { "epoch": 0.8444523218716767, "grad_norm": 2.891585277028761, "learning_rate": 6.212573078454032e-07, "loss": 0.4113, "step": 11911 }, { "epoch": 0.844523218716767, "grad_norm": 2.5118940756516293, "learning_rate": 6.207031294022154e-07, "loss": 0.3953, "step": 11912 }, { "epoch": 0.8445941155618575, "grad_norm": 2.484155792273769, "learning_rate": 6.201491818826932e-07, "loss": 0.384, "step": 11913 }, { "epoch": 0.8446650124069479, "grad_norm": 2.144996902490435, "learning_rate": 6.195954653160486e-07, "loss": 0.3713, "step": 11914 }, { "epoch": 0.8447359092520382, "grad_norm": 6.918504853572868, "learning_rate": 6.190419797314784e-07, "loss": 0.3894, "step": 11915 }, { "epoch": 0.8448068060971287, "grad_norm": 3.3320091729178998, "learning_rate": 6.184887251581678e-07, "loss": 0.4029, "step": 11916 }, { "epoch": 0.8448777029422191, "grad_norm": 2.4220910073304993, "learning_rate": 6.179357016252907e-07, "loss": 0.3941, "step": 11917 }, { "epoch": 0.8449485997873095, "grad_norm": 2.931177541128547, "learning_rate": 6.173829091620109e-07, "loss": 0.3696, "step": 11918 }, { "epoch": 0.8450194966323998, "grad_norm": 2.1598157883885443, "learning_rate": 6.168303477974735e-07, "loss": 0.378, "step": 11919 }, { "epoch": 0.8450903934774903, "grad_norm": 2.6510092260309475, "learning_rate": 6.162780175608168e-07, "loss": 0.3841, "step": 11920 }, { "epoch": 0.8451612903225807, "grad_norm": 5.644468477288241, "learning_rate": 6.157259184811682e-07, "loss": 0.3319, "step": 11921 }, { "epoch": 0.845232187167671, "grad_norm": 4.213758015241807, "learning_rate": 6.151740505876374e-07, "loss": 0.3721, "step": 11922 }, { "epoch": 0.8453030840127614, "grad_norm": 2.371581666600866, "learning_rate": 6.146224139093249e-07, "loss": 0.3638, "step": 11923 }, { "epoch": 0.8453739808578519, "grad_norm": 3.2485839526841875, "learning_rate": 6.140710084753199e-07, "loss": 0.3688, "step": 11924 }, { "epoch": 0.8454448777029422, "grad_norm": 3.3757318638171525, "learning_rate": 6.135198343146992e-07, "loss": 0.4074, "step": 11925 }, { "epoch": 0.8455157745480326, "grad_norm": 2.8347975889426755, "learning_rate": 6.129688914565257e-07, "loss": 0.3716, "step": 11926 }, { "epoch": 0.845586671393123, "grad_norm": 2.090921820952921, "learning_rate": 6.1241817992985e-07, "loss": 0.4067, "step": 11927 }, { "epoch": 0.8456575682382134, "grad_norm": 3.345304584743749, "learning_rate": 6.118676997637135e-07, "loss": 0.4036, "step": 11928 }, { "epoch": 0.8457284650833038, "grad_norm": 0.9795927092735935, "learning_rate": 6.113174509871422e-07, "loss": 0.7135, "step": 11929 }, { "epoch": 0.8457993619283942, "grad_norm": 2.9713446288103578, "learning_rate": 6.107674336291508e-07, "loss": 0.3734, "step": 11930 }, { "epoch": 0.8458702587734845, "grad_norm": 3.7880423973172026, "learning_rate": 6.102176477187438e-07, "loss": 0.436, "step": 11931 }, { "epoch": 0.845941155618575, "grad_norm": 2.0514284386304498, "learning_rate": 6.096680932849097e-07, "loss": 0.3995, "step": 11932 }, { "epoch": 0.8460120524636654, "grad_norm": 2.8086930517468023, "learning_rate": 6.091187703566293e-07, "loss": 0.3353, "step": 11933 }, { "epoch": 0.8460829493087557, "grad_norm": 2.5285880037321844, "learning_rate": 6.085696789628676e-07, "loss": 0.4278, "step": 11934 }, { "epoch": 0.8461538461538461, "grad_norm": 1.9818373236004787, "learning_rate": 6.08020819132577e-07, "loss": 0.3895, "step": 11935 }, { "epoch": 0.8462247429989366, "grad_norm": 8.525953322381278, "learning_rate": 6.074721908947012e-07, "loss": 0.3758, "step": 11936 }, { "epoch": 0.8462956398440269, "grad_norm": 2.5812999306207876, "learning_rate": 6.069237942781714e-07, "loss": 0.377, "step": 11937 }, { "epoch": 0.8463665366891173, "grad_norm": 11.076417160479643, "learning_rate": 6.063756293119011e-07, "loss": 0.3686, "step": 11938 }, { "epoch": 0.8464374335342078, "grad_norm": 5.4638972817075055, "learning_rate": 6.058276960247972e-07, "loss": 0.4477, "step": 11939 }, { "epoch": 0.8465083303792981, "grad_norm": 2.5594794848192053, "learning_rate": 6.052799944457539e-07, "loss": 0.3674, "step": 11940 }, { "epoch": 0.8465792272243885, "grad_norm": 6.732984084448268, "learning_rate": 6.047325246036507e-07, "loss": 0.4078, "step": 11941 }, { "epoch": 0.8466501240694789, "grad_norm": 4.58731332089795, "learning_rate": 6.04185286527355e-07, "loss": 0.4591, "step": 11942 }, { "epoch": 0.8467210209145694, "grad_norm": 3.215091406838719, "learning_rate": 6.036382802457247e-07, "loss": 0.3534, "step": 11943 }, { "epoch": 0.8467919177596597, "grad_norm": 6.786412082535489, "learning_rate": 6.030915057876047e-07, "loss": 0.3986, "step": 11944 }, { "epoch": 0.8468628146047501, "grad_norm": 2.364886568787498, "learning_rate": 6.025449631818254e-07, "loss": 0.37, "step": 11945 }, { "epoch": 0.8469337114498405, "grad_norm": 1.9634987830389383, "learning_rate": 6.019986524572052e-07, "loss": 0.3862, "step": 11946 }, { "epoch": 0.8470046082949308, "grad_norm": 2.6180566597837602, "learning_rate": 6.014525736425542e-07, "loss": 0.4087, "step": 11947 }, { "epoch": 0.8470755051400213, "grad_norm": 3.2814199327303744, "learning_rate": 6.009067267666662e-07, "loss": 0.4009, "step": 11948 }, { "epoch": 0.8471464019851117, "grad_norm": 3.132283608692511, "learning_rate": 6.003611118583235e-07, "loss": 0.3694, "step": 11949 }, { "epoch": 0.847217298830202, "grad_norm": 3.157363783135426, "learning_rate": 5.998157289462981e-07, "loss": 0.3264, "step": 11950 }, { "epoch": 0.8472881956752925, "grad_norm": 2.1415096320277818, "learning_rate": 5.992705780593472e-07, "loss": 0.3936, "step": 11951 }, { "epoch": 0.8473590925203829, "grad_norm": 2.7214530024562698, "learning_rate": 5.98725659226218e-07, "loss": 0.388, "step": 11952 }, { "epoch": 0.8474299893654732, "grad_norm": 2.3542832803599216, "learning_rate": 5.981809724756438e-07, "loss": 0.4269, "step": 11953 }, { "epoch": 0.8475008862105636, "grad_norm": 2.7858668374743605, "learning_rate": 5.976365178363474e-07, "loss": 0.3582, "step": 11954 }, { "epoch": 0.8475717830556541, "grad_norm": 2.5701199313729504, "learning_rate": 5.970922953370372e-07, "loss": 0.3816, "step": 11955 }, { "epoch": 0.8476426799007444, "grad_norm": 5.127960294912728, "learning_rate": 5.965483050064103e-07, "loss": 0.4076, "step": 11956 }, { "epoch": 0.8477135767458348, "grad_norm": 3.281234431899607, "learning_rate": 5.960045468731535e-07, "loss": 0.4013, "step": 11957 }, { "epoch": 0.8477844735909252, "grad_norm": 2.5832106337637586, "learning_rate": 5.954610209659367e-07, "loss": 0.3499, "step": 11958 }, { "epoch": 0.8478553704360156, "grad_norm": 2.8016575116748323, "learning_rate": 5.949177273134238e-07, "loss": 0.4057, "step": 11959 }, { "epoch": 0.847926267281106, "grad_norm": 8.765262749345748, "learning_rate": 5.943746659442612e-07, "loss": 0.3943, "step": 11960 }, { "epoch": 0.8479971641261964, "grad_norm": 2.7406023963737134, "learning_rate": 5.938318368870837e-07, "loss": 0.3548, "step": 11961 }, { "epoch": 0.8480680609712867, "grad_norm": 4.620267880650766, "learning_rate": 5.932892401705171e-07, "loss": 0.4089, "step": 11962 }, { "epoch": 0.8481389578163772, "grad_norm": 3.042860700404717, "learning_rate": 5.927468758231736e-07, "loss": 0.4229, "step": 11963 }, { "epoch": 0.8482098546614676, "grad_norm": 3.462756268249043, "learning_rate": 5.9220474387365e-07, "loss": 0.4208, "step": 11964 }, { "epoch": 0.8482807515065579, "grad_norm": 7.164693379794074, "learning_rate": 5.91662844350534e-07, "loss": 0.3163, "step": 11965 }, { "epoch": 0.8483516483516483, "grad_norm": 2.2256693162061487, "learning_rate": 5.911211772824022e-07, "loss": 0.3407, "step": 11966 }, { "epoch": 0.8484225451967388, "grad_norm": 3.118756759636775, "learning_rate": 5.90579742697816e-07, "loss": 0.3635, "step": 11967 }, { "epoch": 0.8484934420418292, "grad_norm": 1.1864114871207354, "learning_rate": 5.900385406253245e-07, "loss": 0.6891, "step": 11968 }, { "epoch": 0.8485643388869195, "grad_norm": 5.047413512725389, "learning_rate": 5.894975710934664e-07, "loss": 0.3871, "step": 11969 }, { "epoch": 0.8486352357320099, "grad_norm": 2.7839209174365696, "learning_rate": 5.889568341307689e-07, "loss": 0.4574, "step": 11970 }, { "epoch": 0.8487061325771004, "grad_norm": 3.0393348146822907, "learning_rate": 5.884163297657447e-07, "loss": 0.3802, "step": 11971 }, { "epoch": 0.8487770294221907, "grad_norm": 3.0316699641042932, "learning_rate": 5.878760580268933e-07, "loss": 0.387, "step": 11972 }, { "epoch": 0.8488479262672811, "grad_norm": 3.1077640253751877, "learning_rate": 5.873360189427057e-07, "loss": 0.4059, "step": 11973 }, { "epoch": 0.8489188231123715, "grad_norm": 1.0203441855773205, "learning_rate": 5.867962125416588e-07, "loss": 0.6473, "step": 11974 }, { "epoch": 0.8489897199574619, "grad_norm": 2.614185722547053, "learning_rate": 5.86256638852214e-07, "loss": 0.3535, "step": 11975 }, { "epoch": 0.8490606168025523, "grad_norm": 3.1499381608572925, "learning_rate": 5.857172979028269e-07, "loss": 0.421, "step": 11976 }, { "epoch": 0.8491315136476427, "grad_norm": 2.0736590533385826, "learning_rate": 5.85178189721935e-07, "loss": 0.3767, "step": 11977 }, { "epoch": 0.849202410492733, "grad_norm": 3.3188312019239667, "learning_rate": 5.84639314337968e-07, "loss": 0.4049, "step": 11978 }, { "epoch": 0.8492733073378235, "grad_norm": 3.711097204242251, "learning_rate": 5.841006717793401e-07, "loss": 0.4184, "step": 11979 }, { "epoch": 0.8493442041829139, "grad_norm": 1.8654614671090268, "learning_rate": 5.835622620744529e-07, "loss": 0.3508, "step": 11980 }, { "epoch": 0.8494151010280042, "grad_norm": 3.58154976189598, "learning_rate": 5.830240852516989e-07, "loss": 0.4021, "step": 11981 }, { "epoch": 0.8494859978730946, "grad_norm": 4.3764009186229424, "learning_rate": 5.824861413394567e-07, "loss": 0.3595, "step": 11982 }, { "epoch": 0.8495568947181851, "grad_norm": 3.268637755568239, "learning_rate": 5.819484303660922e-07, "loss": 0.3565, "step": 11983 }, { "epoch": 0.8496277915632754, "grad_norm": 3.2193902293608043, "learning_rate": 5.814109523599582e-07, "loss": 0.4125, "step": 11984 }, { "epoch": 0.8496986884083658, "grad_norm": 3.0823412874664347, "learning_rate": 5.808737073493986e-07, "loss": 0.4332, "step": 11985 }, { "epoch": 0.8497695852534562, "grad_norm": 2.978198029059727, "learning_rate": 5.803366953627409e-07, "loss": 0.3756, "step": 11986 }, { "epoch": 0.8498404820985466, "grad_norm": 5.033195450841446, "learning_rate": 5.79799916428302e-07, "loss": 0.3814, "step": 11987 }, { "epoch": 0.849911378943637, "grad_norm": 3.1282946417136834, "learning_rate": 5.792633705743866e-07, "loss": 0.3684, "step": 11988 }, { "epoch": 0.8499822757887274, "grad_norm": 3.904070551468405, "learning_rate": 5.787270578292892e-07, "loss": 0.371, "step": 11989 }, { "epoch": 0.8500531726338177, "grad_norm": 3.68294422964288, "learning_rate": 5.781909782212891e-07, "loss": 0.3914, "step": 11990 }, { "epoch": 0.8501240694789082, "grad_norm": 3.4984077020810704, "learning_rate": 5.776551317786522e-07, "loss": 0.4022, "step": 11991 }, { "epoch": 0.8501949663239986, "grad_norm": 2.554780298139119, "learning_rate": 5.771195185296369e-07, "loss": 0.3987, "step": 11992 }, { "epoch": 0.8502658631690889, "grad_norm": 2.103081910234636, "learning_rate": 5.765841385024851e-07, "loss": 0.4063, "step": 11993 }, { "epoch": 0.8503367600141793, "grad_norm": 7.163813687075956, "learning_rate": 5.76048991725427e-07, "loss": 0.4036, "step": 11994 }, { "epoch": 0.8504076568592698, "grad_norm": 3.318111053521649, "learning_rate": 5.755140782266832e-07, "loss": 0.4209, "step": 11995 }, { "epoch": 0.8504785537043602, "grad_norm": 3.651150524634007, "learning_rate": 5.749793980344582e-07, "loss": 0.3812, "step": 11996 }, { "epoch": 0.8505494505494505, "grad_norm": 2.318351215979387, "learning_rate": 5.744449511769479e-07, "loss": 0.3699, "step": 11997 }, { "epoch": 0.850620347394541, "grad_norm": 1.7740211878611227, "learning_rate": 5.739107376823322e-07, "loss": 0.3617, "step": 11998 }, { "epoch": 0.8506912442396314, "grad_norm": 2.417793971368908, "learning_rate": 5.733767575787824e-07, "loss": 0.3982, "step": 11999 }, { "epoch": 0.8507621410847217, "grad_norm": 1.9302889049630731, "learning_rate": 5.728430108944538e-07, "loss": 0.3969, "step": 12000 }, { "epoch": 0.8508330379298121, "grad_norm": 2.0025706066266356, "learning_rate": 5.723094976574939e-07, "loss": 0.4178, "step": 12001 }, { "epoch": 0.8509039347749026, "grad_norm": 5.565004289204387, "learning_rate": 5.717762178960335e-07, "loss": 0.3625, "step": 12002 }, { "epoch": 0.8509748316199929, "grad_norm": 2.146797004201279, "learning_rate": 5.712431716381916e-07, "loss": 0.3222, "step": 12003 }, { "epoch": 0.8510457284650833, "grad_norm": 2.42983124456852, "learning_rate": 5.707103589120788e-07, "loss": 0.35, "step": 12004 }, { "epoch": 0.8511166253101737, "grad_norm": 3.076169952122293, "learning_rate": 5.701777797457897e-07, "loss": 0.3982, "step": 12005 }, { "epoch": 0.851187522155264, "grad_norm": 6.023342741186371, "learning_rate": 5.696454341674063e-07, "loss": 0.3799, "step": 12006 }, { "epoch": 0.8512584190003545, "grad_norm": 2.9674813038209726, "learning_rate": 5.691133222050005e-07, "loss": 0.3985, "step": 12007 }, { "epoch": 0.8513293158454449, "grad_norm": 21.46022289876217, "learning_rate": 5.685814438866333e-07, "loss": 0.3924, "step": 12008 }, { "epoch": 0.8514002126905352, "grad_norm": 1.911239965421612, "learning_rate": 5.680497992403472e-07, "loss": 0.3507, "step": 12009 }, { "epoch": 0.8514711095356257, "grad_norm": 5.207395313549287, "learning_rate": 5.675183882941776e-07, "loss": 0.4008, "step": 12010 }, { "epoch": 0.8515420063807161, "grad_norm": 2.827442850774147, "learning_rate": 5.669872110761476e-07, "loss": 0.414, "step": 12011 }, { "epoch": 0.8516129032258064, "grad_norm": 1.9026306912247535, "learning_rate": 5.664562676142654e-07, "loss": 0.4131, "step": 12012 }, { "epoch": 0.8516838000708968, "grad_norm": 1.0602342810089564, "learning_rate": 5.659255579365275e-07, "loss": 0.6394, "step": 12013 }, { "epoch": 0.8517546969159873, "grad_norm": 2.0414918812128984, "learning_rate": 5.653950820709186e-07, "loss": 0.346, "step": 12014 }, { "epoch": 0.8518255937610776, "grad_norm": 14.308365564141868, "learning_rate": 5.648648400454138e-07, "loss": 0.405, "step": 12015 }, { "epoch": 0.851896490606168, "grad_norm": 2.0479498950360244, "learning_rate": 5.643348318879704e-07, "loss": 0.3654, "step": 12016 }, { "epoch": 0.8519673874512584, "grad_norm": 2.266847602478353, "learning_rate": 5.638050576265363e-07, "loss": 0.3808, "step": 12017 }, { "epoch": 0.8520382842963488, "grad_norm": 1.7957141882541066, "learning_rate": 5.632755172890486e-07, "loss": 0.3282, "step": 12018 }, { "epoch": 0.8521091811414392, "grad_norm": 3.2969004071628167, "learning_rate": 5.627462109034288e-07, "loss": 0.3595, "step": 12019 }, { "epoch": 0.8521800779865296, "grad_norm": 3.066122368980971, "learning_rate": 5.622171384975877e-07, "loss": 0.4431, "step": 12020 }, { "epoch": 0.85225097483162, "grad_norm": 2.9735286294737073, "learning_rate": 5.616883000994245e-07, "loss": 0.3609, "step": 12021 }, { "epoch": 0.8523218716767104, "grad_norm": 2.0900521709125823, "learning_rate": 5.611596957368243e-07, "loss": 0.3781, "step": 12022 }, { "epoch": 0.8523927685218008, "grad_norm": 2.28834329631564, "learning_rate": 5.606313254376621e-07, "loss": 0.3726, "step": 12023 }, { "epoch": 0.8524636653668912, "grad_norm": 4.293355600418315, "learning_rate": 5.601031892297986e-07, "loss": 0.4089, "step": 12024 }, { "epoch": 0.8525345622119815, "grad_norm": 2.396888323607007, "learning_rate": 5.595752871410814e-07, "loss": 0.4291, "step": 12025 }, { "epoch": 0.852605459057072, "grad_norm": 2.020503918890044, "learning_rate": 5.590476191993493e-07, "loss": 0.3531, "step": 12026 }, { "epoch": 0.8526763559021624, "grad_norm": 15.784300749409816, "learning_rate": 5.585201854324263e-07, "loss": 0.4176, "step": 12027 }, { "epoch": 0.8527472527472527, "grad_norm": 3.120603266556674, "learning_rate": 5.579929858681243e-07, "loss": 0.354, "step": 12028 }, { "epoch": 0.8528181495923431, "grad_norm": 2.783700277452206, "learning_rate": 5.574660205342414e-07, "loss": 0.3998, "step": 12029 }, { "epoch": 0.8528890464374336, "grad_norm": 2.9307350503868332, "learning_rate": 5.569392894585668e-07, "loss": 0.3737, "step": 12030 }, { "epoch": 0.8529599432825239, "grad_norm": 1.0545584561432217, "learning_rate": 5.56412792668875e-07, "loss": 0.6595, "step": 12031 }, { "epoch": 0.8530308401276143, "grad_norm": 3.2421754398824527, "learning_rate": 5.558865301929273e-07, "loss": 0.3329, "step": 12032 }, { "epoch": 0.8531017369727047, "grad_norm": 2.919922512889765, "learning_rate": 5.553605020584746e-07, "loss": 0.3929, "step": 12033 }, { "epoch": 0.8531726338177951, "grad_norm": 3.40257843541502, "learning_rate": 5.548347082932564e-07, "loss": 0.3927, "step": 12034 }, { "epoch": 0.8532435306628855, "grad_norm": 3.919945855622228, "learning_rate": 5.543091489249969e-07, "loss": 0.4145, "step": 12035 }, { "epoch": 0.8533144275079759, "grad_norm": 1.9281025932738525, "learning_rate": 5.53783823981408e-07, "loss": 0.368, "step": 12036 }, { "epoch": 0.8533853243530662, "grad_norm": 3.0612966496841865, "learning_rate": 5.532587334901929e-07, "loss": 0.4326, "step": 12037 }, { "epoch": 0.8534562211981567, "grad_norm": 2.012465392348268, "learning_rate": 5.527338774790386e-07, "loss": 0.3927, "step": 12038 }, { "epoch": 0.8535271180432471, "grad_norm": 3.1860706240601067, "learning_rate": 5.522092559756204e-07, "loss": 0.4204, "step": 12039 }, { "epoch": 0.8535980148883374, "grad_norm": 5.51595766214358, "learning_rate": 5.516848690076044e-07, "loss": 0.407, "step": 12040 }, { "epoch": 0.8536689117334278, "grad_norm": 2.5610219974645445, "learning_rate": 5.511607166026389e-07, "loss": 0.4117, "step": 12041 }, { "epoch": 0.8537398085785183, "grad_norm": 2.1009012317692424, "learning_rate": 5.50636798788366e-07, "loss": 0.3499, "step": 12042 }, { "epoch": 0.8538107054236086, "grad_norm": 3.1822455804309993, "learning_rate": 5.501131155924094e-07, "loss": 0.4302, "step": 12043 }, { "epoch": 0.853881602268699, "grad_norm": 2.4643385057127447, "learning_rate": 5.495896670423856e-07, "loss": 0.3965, "step": 12044 }, { "epoch": 0.8539524991137895, "grad_norm": 2.4851578223715367, "learning_rate": 5.490664531658946e-07, "loss": 0.4205, "step": 12045 }, { "epoch": 0.8540233959588799, "grad_norm": 2.5394559204882916, "learning_rate": 5.485434739905282e-07, "loss": 0.3895, "step": 12046 }, { "epoch": 0.8540942928039702, "grad_norm": 2.807652393562855, "learning_rate": 5.480207295438616e-07, "loss": 0.4466, "step": 12047 }, { "epoch": 0.8541651896490606, "grad_norm": 2.9186628688543323, "learning_rate": 5.47498219853459e-07, "loss": 0.3959, "step": 12048 }, { "epoch": 0.8542360864941511, "grad_norm": 3.653473667238764, "learning_rate": 5.469759449468748e-07, "loss": 0.4052, "step": 12049 }, { "epoch": 0.8543069833392414, "grad_norm": 3.785977099640968, "learning_rate": 5.464539048516476e-07, "loss": 0.4413, "step": 12050 }, { "epoch": 0.8543778801843318, "grad_norm": 2.3917721428273695, "learning_rate": 5.459320995953044e-07, "loss": 0.3998, "step": 12051 }, { "epoch": 0.8544487770294222, "grad_norm": 3.2425229771417365, "learning_rate": 5.454105292053613e-07, "loss": 0.4436, "step": 12052 }, { "epoch": 0.8545196738745126, "grad_norm": 8.897234479454118, "learning_rate": 5.448891937093221e-07, "loss": 0.3675, "step": 12053 }, { "epoch": 0.854590570719603, "grad_norm": 2.724744053976437, "learning_rate": 5.443680931346762e-07, "loss": 0.3693, "step": 12054 }, { "epoch": 0.8546614675646934, "grad_norm": 3.174936236457067, "learning_rate": 5.438472275089001e-07, "loss": 0.4549, "step": 12055 }, { "epoch": 0.8547323644097837, "grad_norm": 2.9983608723037984, "learning_rate": 5.433265968594614e-07, "loss": 0.3713, "step": 12056 }, { "epoch": 0.8548032612548742, "grad_norm": 3.804760250717113, "learning_rate": 5.428062012138147e-07, "loss": 0.3585, "step": 12057 }, { "epoch": 0.8548741580999646, "grad_norm": 5.632035183546246, "learning_rate": 5.422860405993968e-07, "loss": 0.3637, "step": 12058 }, { "epoch": 0.8549450549450549, "grad_norm": 4.053807877156784, "learning_rate": 5.417661150436388e-07, "loss": 0.3674, "step": 12059 }, { "epoch": 0.8550159517901453, "grad_norm": 2.232569571638673, "learning_rate": 5.412464245739574e-07, "loss": 0.3458, "step": 12060 }, { "epoch": 0.8550868486352358, "grad_norm": 2.8930888321932837, "learning_rate": 5.407269692177547e-07, "loss": 0.3616, "step": 12061 }, { "epoch": 0.8551577454803261, "grad_norm": 2.612593875871934, "learning_rate": 5.402077490024221e-07, "loss": 0.4111, "step": 12062 }, { "epoch": 0.8552286423254165, "grad_norm": 2.386579868489094, "learning_rate": 5.3968876395534e-07, "loss": 0.3644, "step": 12063 }, { "epoch": 0.8552995391705069, "grad_norm": 3.1202388635840266, "learning_rate": 5.391700141038731e-07, "loss": 0.352, "step": 12064 }, { "epoch": 0.8553704360155973, "grad_norm": 3.3557775309093243, "learning_rate": 5.386514994753755e-07, "loss": 0.4401, "step": 12065 }, { "epoch": 0.8554413328606877, "grad_norm": 3.543643465138023, "learning_rate": 5.381332200971906e-07, "loss": 0.4709, "step": 12066 }, { "epoch": 0.8555122297057781, "grad_norm": 2.4537881249235354, "learning_rate": 5.376151759966458e-07, "loss": 0.3897, "step": 12067 }, { "epoch": 0.8555831265508684, "grad_norm": 4.114700411263086, "learning_rate": 5.370973672010594e-07, "loss": 0.3697, "step": 12068 }, { "epoch": 0.8556540233959589, "grad_norm": 1.990746025550626, "learning_rate": 5.365797937377354e-07, "loss": 0.358, "step": 12069 }, { "epoch": 0.8557249202410493, "grad_norm": 3.9842027092133003, "learning_rate": 5.360624556339645e-07, "loss": 0.3646, "step": 12070 }, { "epoch": 0.8557958170861397, "grad_norm": 3.196874212682496, "learning_rate": 5.355453529170274e-07, "loss": 0.4338, "step": 12071 }, { "epoch": 0.85586671393123, "grad_norm": 2.743727579208113, "learning_rate": 5.350284856141924e-07, "loss": 0.3284, "step": 12072 }, { "epoch": 0.8559376107763205, "grad_norm": 3.030928976227068, "learning_rate": 5.345118537527133e-07, "loss": 0.396, "step": 12073 }, { "epoch": 0.8560085076214109, "grad_norm": 3.867089616862768, "learning_rate": 5.339954573598316e-07, "loss": 0.382, "step": 12074 }, { "epoch": 0.8560794044665012, "grad_norm": 4.09906759420234, "learning_rate": 5.334792964627794e-07, "loss": 0.3704, "step": 12075 }, { "epoch": 0.8561503013115916, "grad_norm": 2.3554044595987427, "learning_rate": 5.329633710887727e-07, "loss": 0.4269, "step": 12076 }, { "epoch": 0.8562211981566821, "grad_norm": 2.204812553378105, "learning_rate": 5.324476812650164e-07, "loss": 0.3726, "step": 12077 }, { "epoch": 0.8562920950017724, "grad_norm": 2.6409992833307427, "learning_rate": 5.319322270187033e-07, "loss": 0.3763, "step": 12078 }, { "epoch": 0.8563629918468628, "grad_norm": 9.21658738668758, "learning_rate": 5.314170083770154e-07, "loss": 0.3473, "step": 12079 }, { "epoch": 0.8564338886919532, "grad_norm": 2.373969616493854, "learning_rate": 5.309020253671193e-07, "loss": 0.3782, "step": 12080 }, { "epoch": 0.8565047855370436, "grad_norm": 3.720752181510557, "learning_rate": 5.303872780161701e-07, "loss": 0.3506, "step": 12081 }, { "epoch": 0.856575682382134, "grad_norm": 2.356690840154622, "learning_rate": 5.298727663513115e-07, "loss": 0.3926, "step": 12082 }, { "epoch": 0.8566465792272244, "grad_norm": 2.280140089695998, "learning_rate": 5.29358490399674e-07, "loss": 0.3798, "step": 12083 }, { "epoch": 0.8567174760723147, "grad_norm": 3.1731423972610355, "learning_rate": 5.28844450188375e-07, "loss": 0.363, "step": 12084 }, { "epoch": 0.8567883729174052, "grad_norm": 2.0640038899938706, "learning_rate": 5.283306457445203e-07, "loss": 0.385, "step": 12085 }, { "epoch": 0.8568592697624956, "grad_norm": 2.9093024845937636, "learning_rate": 5.278170770952046e-07, "loss": 0.4015, "step": 12086 }, { "epoch": 0.8569301666075859, "grad_norm": 2.0065590655233145, "learning_rate": 5.273037442675083e-07, "loss": 0.3698, "step": 12087 }, { "epoch": 0.8570010634526763, "grad_norm": 9.609649546843585, "learning_rate": 5.267906472884982e-07, "loss": 0.3856, "step": 12088 }, { "epoch": 0.8570719602977668, "grad_norm": 4.097026729231317, "learning_rate": 5.262777861852325e-07, "loss": 0.3629, "step": 12089 }, { "epoch": 0.8571428571428571, "grad_norm": 2.453397371164957, "learning_rate": 5.25765160984753e-07, "loss": 0.3904, "step": 12090 }, { "epoch": 0.8572137539879475, "grad_norm": 3.2772305345128845, "learning_rate": 5.252527717140926e-07, "loss": 0.3559, "step": 12091 }, { "epoch": 0.857284650833038, "grad_norm": 2.523212798435147, "learning_rate": 5.247406184002685e-07, "loss": 0.3881, "step": 12092 }, { "epoch": 0.8573555476781283, "grad_norm": 3.151601951110506, "learning_rate": 5.242287010702868e-07, "loss": 0.4023, "step": 12093 }, { "epoch": 0.8574264445232187, "grad_norm": 1.902908448452722, "learning_rate": 5.237170197511432e-07, "loss": 0.3739, "step": 12094 }, { "epoch": 0.8574973413683091, "grad_norm": 4.16460503554556, "learning_rate": 5.232055744698172e-07, "loss": 0.4047, "step": 12095 }, { "epoch": 0.8575682382133996, "grad_norm": 3.0712681045359833, "learning_rate": 5.226943652532773e-07, "loss": 0.3556, "step": 12096 }, { "epoch": 0.8576391350584899, "grad_norm": 3.9951432564291043, "learning_rate": 5.221833921284808e-07, "loss": 0.4058, "step": 12097 }, { "epoch": 0.8577100319035803, "grad_norm": 9.33675483353625, "learning_rate": 5.216726551223733e-07, "loss": 0.3993, "step": 12098 }, { "epoch": 0.8577809287486707, "grad_norm": 2.654570818912397, "learning_rate": 5.211621542618844e-07, "loss": 0.344, "step": 12099 }, { "epoch": 0.857851825593761, "grad_norm": 3.317136372310722, "learning_rate": 5.206518895739326e-07, "loss": 0.3744, "step": 12100 }, { "epoch": 0.8579227224388515, "grad_norm": 4.945790583684705, "learning_rate": 5.201418610854258e-07, "loss": 0.326, "step": 12101 }, { "epoch": 0.8579936192839419, "grad_norm": 2.7954717915396072, "learning_rate": 5.196320688232604e-07, "loss": 0.3567, "step": 12102 }, { "epoch": 0.8580645161290322, "grad_norm": 3.3515786806253756, "learning_rate": 5.191225128143129e-07, "loss": 0.3949, "step": 12103 }, { "epoch": 0.8581354129741227, "grad_norm": 4.154403153390791, "learning_rate": 5.18613193085456e-07, "loss": 0.3922, "step": 12104 }, { "epoch": 0.8582063098192131, "grad_norm": 2.108652344276536, "learning_rate": 5.181041096635464e-07, "loss": 0.3595, "step": 12105 }, { "epoch": 0.8582772066643034, "grad_norm": 2.8744821548943817, "learning_rate": 5.175952625754282e-07, "loss": 0.402, "step": 12106 }, { "epoch": 0.8583481035093938, "grad_norm": 2.484471554572816, "learning_rate": 5.170866518479322e-07, "loss": 0.4224, "step": 12107 }, { "epoch": 0.8584190003544843, "grad_norm": 3.5479086581358255, "learning_rate": 5.165782775078798e-07, "loss": 0.4036, "step": 12108 }, { "epoch": 0.8584898971995746, "grad_norm": 4.354688877087414, "learning_rate": 5.160701395820756e-07, "loss": 0.3814, "step": 12109 }, { "epoch": 0.858560794044665, "grad_norm": 1.9485689847144467, "learning_rate": 5.155622380973169e-07, "loss": 0.421, "step": 12110 }, { "epoch": 0.8586316908897554, "grad_norm": 1.5733035155016932, "learning_rate": 5.150545730803835e-07, "loss": 0.4129, "step": 12111 }, { "epoch": 0.8587025877348458, "grad_norm": 28.650832056326617, "learning_rate": 5.145471445580452e-07, "loss": 0.3855, "step": 12112 }, { "epoch": 0.8587734845799362, "grad_norm": 7.270418118278558, "learning_rate": 5.1403995255706e-07, "loss": 0.3661, "step": 12113 }, { "epoch": 0.8588443814250266, "grad_norm": 5.451381945227618, "learning_rate": 5.135329971041714e-07, "loss": 0.3653, "step": 12114 }, { "epoch": 0.8589152782701169, "grad_norm": 2.212408311116208, "learning_rate": 5.13026278226113e-07, "loss": 0.3864, "step": 12115 }, { "epoch": 0.8589861751152074, "grad_norm": 5.907612728045674, "learning_rate": 5.125197959496031e-07, "loss": 0.3974, "step": 12116 }, { "epoch": 0.8590570719602978, "grad_norm": 1.8851166601676357, "learning_rate": 5.120135503013501e-07, "loss": 0.4138, "step": 12117 }, { "epoch": 0.8591279688053881, "grad_norm": 2.246052998117173, "learning_rate": 5.115075413080483e-07, "loss": 0.3896, "step": 12118 }, { "epoch": 0.8591988656504785, "grad_norm": 2.195435943730013, "learning_rate": 5.110017689963792e-07, "loss": 0.3465, "step": 12119 }, { "epoch": 0.859269762495569, "grad_norm": 2.766946298384836, "learning_rate": 5.104962333930136e-07, "loss": 0.4142, "step": 12120 }, { "epoch": 0.8593406593406593, "grad_norm": 3.219037195857546, "learning_rate": 5.099909345246085e-07, "loss": 0.4447, "step": 12121 }, { "epoch": 0.8594115561857497, "grad_norm": 2.9830174945026333, "learning_rate": 5.09485872417807e-07, "loss": 0.3836, "step": 12122 }, { "epoch": 0.8594824530308401, "grad_norm": 2.38775796548222, "learning_rate": 5.089810470992434e-07, "loss": 0.3792, "step": 12123 }, { "epoch": 0.8595533498759306, "grad_norm": 3.310465309288287, "learning_rate": 5.084764585955382e-07, "loss": 0.3815, "step": 12124 }, { "epoch": 0.8596242467210209, "grad_norm": 1.7062570351946258, "learning_rate": 5.079721069332971e-07, "loss": 0.3495, "step": 12125 }, { "epoch": 0.8596951435661113, "grad_norm": 3.0518018641600335, "learning_rate": 5.074679921391146e-07, "loss": 0.4064, "step": 12126 }, { "epoch": 0.8597660404112017, "grad_norm": 1.5742388428485876, "learning_rate": 5.069641142395748e-07, "loss": 0.4104, "step": 12127 }, { "epoch": 0.8598369372562921, "grad_norm": 2.5898235758725234, "learning_rate": 5.064604732612466e-07, "loss": 0.3666, "step": 12128 }, { "epoch": 0.8599078341013825, "grad_norm": 1.9353991536445032, "learning_rate": 5.059570692306865e-07, "loss": 0.3727, "step": 12129 }, { "epoch": 0.8599787309464729, "grad_norm": 2.945393007387022, "learning_rate": 5.054539021744403e-07, "loss": 0.3712, "step": 12130 }, { "epoch": 0.8600496277915632, "grad_norm": 5.918214728404182, "learning_rate": 5.049509721190415e-07, "loss": 0.3835, "step": 12131 }, { "epoch": 0.8601205246366537, "grad_norm": 2.5893587001232055, "learning_rate": 5.044482790910088e-07, "loss": 0.401, "step": 12132 }, { "epoch": 0.8601914214817441, "grad_norm": 2.198368108771131, "learning_rate": 5.039458231168481e-07, "loss": 0.3774, "step": 12133 }, { "epoch": 0.8602623183268344, "grad_norm": 2.995872673381463, "learning_rate": 5.034436042230578e-07, "loss": 0.4001, "step": 12134 }, { "epoch": 0.8603332151719248, "grad_norm": 2.4327930750959563, "learning_rate": 5.029416224361167e-07, "loss": 0.3628, "step": 12135 }, { "epoch": 0.8604041120170153, "grad_norm": 2.030639898570073, "learning_rate": 5.024398777824974e-07, "loss": 0.4307, "step": 12136 }, { "epoch": 0.8604750088621056, "grad_norm": 1.0014239676812988, "learning_rate": 5.01938370288656e-07, "loss": 0.598, "step": 12137 }, { "epoch": 0.860545905707196, "grad_norm": 2.0555760657923745, "learning_rate": 5.014370999810369e-07, "loss": 0.3399, "step": 12138 }, { "epoch": 0.8606168025522865, "grad_norm": 3.958533947517364, "learning_rate": 5.009360668860736e-07, "loss": 0.3453, "step": 12139 }, { "epoch": 0.8606876993973768, "grad_norm": 3.4829369301882056, "learning_rate": 5.004352710301858e-07, "loss": 0.406, "step": 12140 }, { "epoch": 0.8607585962424672, "grad_norm": 3.2135070595334425, "learning_rate": 4.999347124397796e-07, "loss": 0.3943, "step": 12141 }, { "epoch": 0.8608294930875576, "grad_norm": 4.297451588708681, "learning_rate": 4.994343911412509e-07, "loss": 0.4016, "step": 12142 }, { "epoch": 0.860900389932648, "grad_norm": 3.5888675871514004, "learning_rate": 4.989343071609826e-07, "loss": 0.3679, "step": 12143 }, { "epoch": 0.8609712867777384, "grad_norm": 2.244919881967979, "learning_rate": 4.984344605253438e-07, "loss": 0.3815, "step": 12144 }, { "epoch": 0.8610421836228288, "grad_norm": 2.8396086524957305, "learning_rate": 4.979348512606907e-07, "loss": 0.383, "step": 12145 }, { "epoch": 0.8611130804679191, "grad_norm": 1.8282311983551114, "learning_rate": 4.974354793933695e-07, "loss": 0.3962, "step": 12146 }, { "epoch": 0.8611839773130096, "grad_norm": 1.9654268157146462, "learning_rate": 4.969363449497139e-07, "loss": 0.3574, "step": 12147 }, { "epoch": 0.8612548741581, "grad_norm": 2.5107883561681232, "learning_rate": 4.964374479560396e-07, "loss": 0.3676, "step": 12148 }, { "epoch": 0.8613257710031904, "grad_norm": 1.621712290520417, "learning_rate": 4.959387884386568e-07, "loss": 0.3955, "step": 12149 }, { "epoch": 0.8613966678482807, "grad_norm": 2.5201799378271734, "learning_rate": 4.9544036642386e-07, "loss": 0.4109, "step": 12150 }, { "epoch": 0.8614675646933712, "grad_norm": 1.0847537144351433, "learning_rate": 4.949421819379307e-07, "loss": 0.5995, "step": 12151 }, { "epoch": 0.8615384615384616, "grad_norm": 3.138643741730857, "learning_rate": 4.944442350071377e-07, "loss": 0.4534, "step": 12152 }, { "epoch": 0.8616093583835519, "grad_norm": 3.506941140271849, "learning_rate": 4.939465256577403e-07, "loss": 0.381, "step": 12153 }, { "epoch": 0.8616802552286423, "grad_norm": 2.1230473292217895, "learning_rate": 4.934490539159814e-07, "loss": 0.3993, "step": 12154 }, { "epoch": 0.8617511520737328, "grad_norm": 2.797849011280974, "learning_rate": 4.92951819808094e-07, "loss": 0.3951, "step": 12155 }, { "epoch": 0.8618220489188231, "grad_norm": 2.11740218006182, "learning_rate": 4.924548233602972e-07, "loss": 0.4416, "step": 12156 }, { "epoch": 0.8618929457639135, "grad_norm": 5.4841481449070795, "learning_rate": 4.919580645987971e-07, "loss": 0.4089, "step": 12157 }, { "epoch": 0.8619638426090039, "grad_norm": 7.279498537092295, "learning_rate": 4.914615435497905e-07, "loss": 0.4012, "step": 12158 }, { "epoch": 0.8620347394540943, "grad_norm": 3.4056839753891137, "learning_rate": 4.909652602394566e-07, "loss": 0.3029, "step": 12159 }, { "epoch": 0.8621056362991847, "grad_norm": 2.397853852098814, "learning_rate": 4.904692146939666e-07, "loss": 0.401, "step": 12160 }, { "epoch": 0.8621765331442751, "grad_norm": 1.7555569489318366, "learning_rate": 4.899734069394763e-07, "loss": 0.4092, "step": 12161 }, { "epoch": 0.8622474299893654, "grad_norm": 2.545512317518731, "learning_rate": 4.894778370021319e-07, "loss": 0.3695, "step": 12162 }, { "epoch": 0.8623183268344559, "grad_norm": 3.5284489999711184, "learning_rate": 4.889825049080637e-07, "loss": 0.3671, "step": 12163 }, { "epoch": 0.8623892236795463, "grad_norm": 2.963922596365535, "learning_rate": 4.884874106833898e-07, "loss": 0.3994, "step": 12164 }, { "epoch": 0.8624601205246366, "grad_norm": 4.113426598466947, "learning_rate": 4.879925543542185e-07, "loss": 0.3853, "step": 12165 }, { "epoch": 0.862531017369727, "grad_norm": 2.019768155560093, "learning_rate": 4.874979359466458e-07, "loss": 0.3884, "step": 12166 }, { "epoch": 0.8626019142148175, "grad_norm": 1.9187171105281113, "learning_rate": 4.870035554867491e-07, "loss": 0.3577, "step": 12167 }, { "epoch": 0.8626728110599078, "grad_norm": 3.72993978696458, "learning_rate": 4.865094130005992e-07, "loss": 0.4014, "step": 12168 }, { "epoch": 0.8627437079049982, "grad_norm": 2.051355237583712, "learning_rate": 4.860155085142543e-07, "loss": 0.361, "step": 12169 }, { "epoch": 0.8628146047500886, "grad_norm": 3.122047532821343, "learning_rate": 4.855218420537566e-07, "loss": 0.3922, "step": 12170 }, { "epoch": 0.862885501595179, "grad_norm": 3.7333419901780633, "learning_rate": 4.850284136451377e-07, "loss": 0.3951, "step": 12171 }, { "epoch": 0.8629563984402694, "grad_norm": 1.7356365931700617, "learning_rate": 4.845352233144168e-07, "loss": 0.3446, "step": 12172 }, { "epoch": 0.8630272952853598, "grad_norm": 2.7954529794531924, "learning_rate": 4.840422710876009e-07, "loss": 0.4076, "step": 12173 }, { "epoch": 0.8630981921304502, "grad_norm": 5.99455981681808, "learning_rate": 4.835495569906811e-07, "loss": 0.3446, "step": 12174 }, { "epoch": 0.8631690889755406, "grad_norm": 2.4768768794289953, "learning_rate": 4.83057081049641e-07, "loss": 0.3922, "step": 12175 }, { "epoch": 0.863239985820631, "grad_norm": 2.3700239024009666, "learning_rate": 4.825648432904495e-07, "loss": 0.3542, "step": 12176 }, { "epoch": 0.8633108826657214, "grad_norm": 2.6625593956304248, "learning_rate": 4.820728437390615e-07, "loss": 0.3574, "step": 12177 }, { "epoch": 0.8633817795108117, "grad_norm": 28.81164012131464, "learning_rate": 4.815810824214201e-07, "loss": 0.3228, "step": 12178 }, { "epoch": 0.8634526763559022, "grad_norm": 1.0658335240334593, "learning_rate": 4.810895593634579e-07, "loss": 0.6383, "step": 12179 }, { "epoch": 0.8635235732009926, "grad_norm": 2.892462839758424, "learning_rate": 4.805982745910914e-07, "loss": 0.4171, "step": 12180 }, { "epoch": 0.8635944700460829, "grad_norm": 2.828086441019917, "learning_rate": 4.801072281302283e-07, "loss": 0.3885, "step": 12181 }, { "epoch": 0.8636653668911733, "grad_norm": 2.779283937348885, "learning_rate": 4.79616420006761e-07, "loss": 0.3735, "step": 12182 }, { "epoch": 0.8637362637362638, "grad_norm": 2.3959125079952557, "learning_rate": 4.79125850246569e-07, "loss": 0.4131, "step": 12183 }, { "epoch": 0.8638071605813541, "grad_norm": 2.8690387338703145, "learning_rate": 4.786355188755232e-07, "loss": 0.3767, "step": 12184 }, { "epoch": 0.8638780574264445, "grad_norm": 2.247155255670823, "learning_rate": 4.781454259194768e-07, "loss": 0.3938, "step": 12185 }, { "epoch": 0.863948954271535, "grad_norm": 8.023048588437787, "learning_rate": 4.776555714042724e-07, "loss": 0.3453, "step": 12186 }, { "epoch": 0.8640198511166253, "grad_norm": 3.0418922866152482, "learning_rate": 4.771659553557417e-07, "loss": 0.3957, "step": 12187 }, { "epoch": 0.8640907479617157, "grad_norm": 4.36044727064035, "learning_rate": 4.766765777997034e-07, "loss": 0.3529, "step": 12188 }, { "epoch": 0.8641616448068061, "grad_norm": 2.064354894051516, "learning_rate": 4.761874387619614e-07, "loss": 0.368, "step": 12189 }, { "epoch": 0.8642325416518964, "grad_norm": 2.7878919339835098, "learning_rate": 4.7569853826830827e-07, "loss": 0.4315, "step": 12190 }, { "epoch": 0.8643034384969869, "grad_norm": 11.011733736394943, "learning_rate": 4.75209876344524e-07, "loss": 0.3738, "step": 12191 }, { "epoch": 0.8643743353420773, "grad_norm": 14.047200032060267, "learning_rate": 4.7472145301637795e-07, "loss": 0.3842, "step": 12192 }, { "epoch": 0.8644452321871676, "grad_norm": 3.273705645410609, "learning_rate": 4.7423326830962233e-07, "loss": 0.3367, "step": 12193 }, { "epoch": 0.864516129032258, "grad_norm": 2.5933282861757716, "learning_rate": 4.73745322250001e-07, "loss": 0.3777, "step": 12194 }, { "epoch": 0.8645870258773485, "grad_norm": 3.3403697378919674, "learning_rate": 4.732576148632445e-07, "loss": 0.4234, "step": 12195 }, { "epoch": 0.8646579227224388, "grad_norm": 2.675213615772737, "learning_rate": 4.7277014617506886e-07, "loss": 0.3374, "step": 12196 }, { "epoch": 0.8647288195675292, "grad_norm": 1.956819846568877, "learning_rate": 4.7228291621117805e-07, "loss": 0.3535, "step": 12197 }, { "epoch": 0.8647997164126197, "grad_norm": 2.384754259045957, "learning_rate": 4.717959249972653e-07, "loss": 0.4128, "step": 12198 }, { "epoch": 0.8648706132577101, "grad_norm": 3.2675268948782916, "learning_rate": 4.7130917255900963e-07, "loss": 0.3974, "step": 12199 }, { "epoch": 0.8649415101028004, "grad_norm": 2.290694070635714, "learning_rate": 4.7082265892207825e-07, "loss": 0.37, "step": 12200 }, { "epoch": 0.8650124069478908, "grad_norm": 3.3586367798629206, "learning_rate": 4.70336384112125e-07, "loss": 0.3728, "step": 12201 }, { "epoch": 0.8650833037929813, "grad_norm": 1.9938499072892268, "learning_rate": 4.698503481547906e-07, "loss": 0.3675, "step": 12202 }, { "epoch": 0.8651542006380716, "grad_norm": 3.3965683834776375, "learning_rate": 4.693645510757061e-07, "loss": 0.3586, "step": 12203 }, { "epoch": 0.865225097483162, "grad_norm": 3.3120368184224582, "learning_rate": 4.688789929004861e-07, "loss": 0.3718, "step": 12204 }, { "epoch": 0.8652959943282524, "grad_norm": 2.127852198846689, "learning_rate": 4.683936736547362e-07, "loss": 0.3775, "step": 12205 }, { "epoch": 0.8653668911733428, "grad_norm": 3.1851524121094883, "learning_rate": 4.679085933640459e-07, "loss": 0.3866, "step": 12206 }, { "epoch": 0.8654377880184332, "grad_norm": 2.1900743231013875, "learning_rate": 4.674237520539954e-07, "loss": 0.4107, "step": 12207 }, { "epoch": 0.8655086848635236, "grad_norm": 2.6293237435513688, "learning_rate": 4.6693914975014964e-07, "loss": 0.3777, "step": 12208 }, { "epoch": 0.8655795817086139, "grad_norm": 5.126045975169509, "learning_rate": 4.6645478647806163e-07, "loss": 0.3753, "step": 12209 }, { "epoch": 0.8656504785537044, "grad_norm": 3.410151600760528, "learning_rate": 4.659706622632737e-07, "loss": 0.4165, "step": 12210 }, { "epoch": 0.8657213753987948, "grad_norm": 2.6015925950601355, "learning_rate": 4.6548677713131486e-07, "loss": 0.353, "step": 12211 }, { "epoch": 0.8657922722438851, "grad_norm": 3.0847954350930777, "learning_rate": 4.6500313110769693e-07, "loss": 0.3873, "step": 12212 }, { "epoch": 0.8658631690889755, "grad_norm": 5.2005695081544205, "learning_rate": 4.645197242179261e-07, "loss": 0.3911, "step": 12213 }, { "epoch": 0.865934065934066, "grad_norm": 3.5133214407779096, "learning_rate": 4.640365564874927e-07, "loss": 0.3397, "step": 12214 }, { "epoch": 0.8660049627791563, "grad_norm": 4.110197986895499, "learning_rate": 4.635536279418734e-07, "loss": 0.3663, "step": 12215 }, { "epoch": 0.8660758596242467, "grad_norm": 2.3908621405050976, "learning_rate": 4.630709386065335e-07, "loss": 0.3906, "step": 12216 }, { "epoch": 0.8661467564693371, "grad_norm": 2.1666717022068056, "learning_rate": 4.625884885069254e-07, "loss": 0.3789, "step": 12217 }, { "epoch": 0.8662176533144275, "grad_norm": 1.74362929350856, "learning_rate": 4.6210627766849036e-07, "loss": 0.3907, "step": 12218 }, { "epoch": 0.8662885501595179, "grad_norm": 2.355153642070289, "learning_rate": 4.616243061166553e-07, "loss": 0.3861, "step": 12219 }, { "epoch": 0.8663594470046083, "grad_norm": 5.187311246874218, "learning_rate": 4.611425738768338e-07, "loss": 0.3738, "step": 12220 }, { "epoch": 0.8664303438496986, "grad_norm": 2.3836525716870054, "learning_rate": 4.606610809744294e-07, "loss": 0.3586, "step": 12221 }, { "epoch": 0.8665012406947891, "grad_norm": 2.3283217682969815, "learning_rate": 4.601798274348307e-07, "loss": 0.3637, "step": 12222 }, { "epoch": 0.8665721375398795, "grad_norm": 1.7513957476286546, "learning_rate": 4.5969881328341405e-07, "loss": 0.3792, "step": 12223 }, { "epoch": 0.8666430343849699, "grad_norm": 1.9161376339217038, "learning_rate": 4.592180385455458e-07, "loss": 0.3628, "step": 12224 }, { "epoch": 0.8667139312300602, "grad_norm": 7.833870201558196, "learning_rate": 4.587375032465746e-07, "loss": 0.4015, "step": 12225 }, { "epoch": 0.8667848280751507, "grad_norm": 2.774419465834336, "learning_rate": 4.582572074118424e-07, "loss": 0.3945, "step": 12226 }, { "epoch": 0.8668557249202411, "grad_norm": 2.1333437068384784, "learning_rate": 4.57777151066674e-07, "loss": 0.3781, "step": 12227 }, { "epoch": 0.8669266217653314, "grad_norm": 0.9517903873243522, "learning_rate": 4.5729733423638235e-07, "loss": 0.6079, "step": 12228 }, { "epoch": 0.8669975186104218, "grad_norm": 3.898954284504874, "learning_rate": 4.5681775694627064e-07, "loss": 0.3651, "step": 12229 }, { "epoch": 0.8670684154555123, "grad_norm": 2.2081056429186936, "learning_rate": 4.563384192216264e-07, "loss": 0.3994, "step": 12230 }, { "epoch": 0.8671393123006026, "grad_norm": 2.8370600819131617, "learning_rate": 4.558593210877238e-07, "loss": 0.3938, "step": 12231 }, { "epoch": 0.867210209145693, "grad_norm": 3.867319354305392, "learning_rate": 4.5538046256982826e-07, "loss": 0.4719, "step": 12232 }, { "epoch": 0.8672811059907835, "grad_norm": 2.430873390414021, "learning_rate": 4.549018436931901e-07, "loss": 0.3712, "step": 12233 }, { "epoch": 0.8673520028358738, "grad_norm": 3.727889873866343, "learning_rate": 4.54423464483047e-07, "loss": 0.3866, "step": 12234 }, { "epoch": 0.8674228996809642, "grad_norm": 3.534804200228291, "learning_rate": 4.539453249646236e-07, "loss": 0.411, "step": 12235 }, { "epoch": 0.8674937965260546, "grad_norm": 3.465680113574492, "learning_rate": 4.534674251631327e-07, "loss": 0.3619, "step": 12236 }, { "epoch": 0.867564693371145, "grad_norm": 2.904021646280859, "learning_rate": 4.5298976510377623e-07, "loss": 0.3702, "step": 12237 }, { "epoch": 0.8676355902162354, "grad_norm": 2.1185086091029723, "learning_rate": 4.5251234481173857e-07, "loss": 0.4049, "step": 12238 }, { "epoch": 0.8677064870613258, "grad_norm": 4.8542108039447855, "learning_rate": 4.5203516431219617e-07, "loss": 0.3916, "step": 12239 }, { "epoch": 0.8677773839064161, "grad_norm": 5.734336572966718, "learning_rate": 4.515582236303112e-07, "loss": 0.38, "step": 12240 }, { "epoch": 0.8678482807515066, "grad_norm": 3.7538206260830207, "learning_rate": 4.5108152279123296e-07, "loss": 0.3621, "step": 12241 }, { "epoch": 0.867919177596597, "grad_norm": 5.624734077564679, "learning_rate": 4.506050618200974e-07, "loss": 0.3441, "step": 12242 }, { "epoch": 0.8679900744416873, "grad_norm": 8.602191009784805, "learning_rate": 4.5012884074203013e-07, "loss": 0.3711, "step": 12243 }, { "epoch": 0.8680609712867777, "grad_norm": 16.663305099984576, "learning_rate": 4.4965285958214033e-07, "loss": 0.4184, "step": 12244 }, { "epoch": 0.8681318681318682, "grad_norm": 2.663129370552459, "learning_rate": 4.491771183655302e-07, "loss": 0.372, "step": 12245 }, { "epoch": 0.8682027649769585, "grad_norm": 2.2565634266049224, "learning_rate": 4.4870161711728365e-07, "loss": 0.3542, "step": 12246 }, { "epoch": 0.8682736618220489, "grad_norm": 5.59509993634053, "learning_rate": 4.482263558624739e-07, "loss": 0.4136, "step": 12247 }, { "epoch": 0.8683445586671393, "grad_norm": 2.0533403492014823, "learning_rate": 4.4775133462616306e-07, "loss": 0.3694, "step": 12248 }, { "epoch": 0.8684154555122297, "grad_norm": 3.2426165251726755, "learning_rate": 4.472765534333984e-07, "loss": 0.4397, "step": 12249 }, { "epoch": 0.8684863523573201, "grad_norm": 5.407367079021894, "learning_rate": 4.4680201230921705e-07, "loss": 0.3819, "step": 12250 }, { "epoch": 0.8685572492024105, "grad_norm": 2.4758162502285193, "learning_rate": 4.463277112786396e-07, "loss": 0.3841, "step": 12251 }, { "epoch": 0.8686281460475009, "grad_norm": 3.0533482845138287, "learning_rate": 4.458536503666788e-07, "loss": 0.3467, "step": 12252 }, { "epoch": 0.8686990428925913, "grad_norm": 2.667601746212301, "learning_rate": 4.4537982959833084e-07, "loss": 0.3804, "step": 12253 }, { "epoch": 0.8687699397376817, "grad_norm": 3.98069842668347, "learning_rate": 4.4490624899858003e-07, "loss": 0.3953, "step": 12254 }, { "epoch": 0.8688408365827721, "grad_norm": 2.401868288108814, "learning_rate": 4.4443290859239984e-07, "loss": 0.3554, "step": 12255 }, { "epoch": 0.8689117334278624, "grad_norm": 2.622769105962337, "learning_rate": 4.4395980840475083e-07, "loss": 0.3644, "step": 12256 }, { "epoch": 0.8689826302729529, "grad_norm": 4.266101956155307, "learning_rate": 4.434869484605769e-07, "loss": 0.4455, "step": 12257 }, { "epoch": 0.8690535271180433, "grad_norm": 3.756960934968378, "learning_rate": 4.4301432878481374e-07, "loss": 0.4119, "step": 12258 }, { "epoch": 0.8691244239631336, "grad_norm": 11.094714538932665, "learning_rate": 4.425419494023847e-07, "loss": 0.4108, "step": 12259 }, { "epoch": 0.869195320808224, "grad_norm": 2.5168366159123186, "learning_rate": 4.4206981033819705e-07, "loss": 0.3628, "step": 12260 }, { "epoch": 0.8692662176533145, "grad_norm": 1.908963908220197, "learning_rate": 4.4159791161714596e-07, "loss": 0.3865, "step": 12261 }, { "epoch": 0.8693371144984048, "grad_norm": 2.3981494751152166, "learning_rate": 4.411262532641164e-07, "loss": 0.3913, "step": 12262 }, { "epoch": 0.8694080113434952, "grad_norm": 2.511191443790944, "learning_rate": 4.4065483530398025e-07, "loss": 0.3594, "step": 12263 }, { "epoch": 0.8694789081885856, "grad_norm": 6.438376017724412, "learning_rate": 4.401836577615942e-07, "loss": 0.4144, "step": 12264 }, { "epoch": 0.869549805033676, "grad_norm": 3.1887665193028396, "learning_rate": 4.397127206618035e-07, "loss": 0.3816, "step": 12265 }, { "epoch": 0.8696207018787664, "grad_norm": 5.546396952145764, "learning_rate": 4.392420240294432e-07, "loss": 0.3702, "step": 12266 }, { "epoch": 0.8696915987238568, "grad_norm": 2.580228430167997, "learning_rate": 4.3877156788933115e-07, "loss": 0.3629, "step": 12267 }, { "epoch": 0.8697624955689471, "grad_norm": 3.9389249934098203, "learning_rate": 4.383013522662755e-07, "loss": 0.4333, "step": 12268 }, { "epoch": 0.8698333924140376, "grad_norm": 2.2275728588210355, "learning_rate": 4.3783137718507227e-07, "loss": 0.4076, "step": 12269 }, { "epoch": 0.869904289259128, "grad_norm": 1.9961975180946043, "learning_rate": 4.373616426705013e-07, "loss": 0.3343, "step": 12270 }, { "epoch": 0.8699751861042183, "grad_norm": 2.410283374585408, "learning_rate": 4.368921487473343e-07, "loss": 0.3936, "step": 12271 }, { "epoch": 0.8700460829493087, "grad_norm": 4.948937459534262, "learning_rate": 4.364228954403277e-07, "loss": 0.3728, "step": 12272 }, { "epoch": 0.8701169797943992, "grad_norm": 2.8905693214092687, "learning_rate": 4.359538827742238e-07, "loss": 0.3954, "step": 12273 }, { "epoch": 0.8701878766394895, "grad_norm": 2.5874940217262625, "learning_rate": 4.354851107737551e-07, "loss": 0.368, "step": 12274 }, { "epoch": 0.8702587734845799, "grad_norm": 2.6920838745300277, "learning_rate": 4.350165794636424e-07, "loss": 0.3132, "step": 12275 }, { "epoch": 0.8703296703296703, "grad_norm": 3.930530875069629, "learning_rate": 4.3454828886858816e-07, "loss": 0.3766, "step": 12276 }, { "epoch": 0.8704005671747608, "grad_norm": 2.1763516897597173, "learning_rate": 4.34080239013287e-07, "loss": 0.3932, "step": 12277 }, { "epoch": 0.8704714640198511, "grad_norm": 3.476780361720285, "learning_rate": 4.3361242992242094e-07, "loss": 0.3907, "step": 12278 }, { "epoch": 0.8705423608649415, "grad_norm": 2.5071268752650826, "learning_rate": 4.331448616206568e-07, "loss": 0.3589, "step": 12279 }, { "epoch": 0.870613257710032, "grad_norm": 5.9412894675080485, "learning_rate": 4.326775341326489e-07, "loss": 0.4686, "step": 12280 }, { "epoch": 0.8706841545551223, "grad_norm": 2.476034238869649, "learning_rate": 4.3221044748304077e-07, "loss": 0.4087, "step": 12281 }, { "epoch": 0.8707550514002127, "grad_norm": 2.5321692614082174, "learning_rate": 4.317436016964638e-07, "loss": 0.3854, "step": 12282 }, { "epoch": 0.8708259482453031, "grad_norm": 2.822255856344168, "learning_rate": 4.3127699679753164e-07, "loss": 0.3684, "step": 12283 }, { "epoch": 0.8708968450903934, "grad_norm": 2.9977192646319626, "learning_rate": 4.3081063281085076e-07, "loss": 0.4285, "step": 12284 }, { "epoch": 0.8709677419354839, "grad_norm": 3.6651684780402003, "learning_rate": 4.303445097610137e-07, "loss": 0.3679, "step": 12285 }, { "epoch": 0.8710386387805743, "grad_norm": 1.6505689977174778, "learning_rate": 4.2987862767259803e-07, "loss": 0.3895, "step": 12286 }, { "epoch": 0.8711095356256646, "grad_norm": 15.712455768033475, "learning_rate": 4.294129865701702e-07, "loss": 0.3703, "step": 12287 }, { "epoch": 0.871180432470755, "grad_norm": 2.4069413167418117, "learning_rate": 4.2894758647828495e-07, "loss": 0.3824, "step": 12288 }, { "epoch": 0.8712513293158455, "grad_norm": 2.2771854508497618, "learning_rate": 4.284824274214816e-07, "loss": 0.4382, "step": 12289 }, { "epoch": 0.8713222261609358, "grad_norm": 1.9606250823298605, "learning_rate": 4.2801750942428945e-07, "loss": 0.3828, "step": 12290 }, { "epoch": 0.8713931230060262, "grad_norm": 2.8427997294535987, "learning_rate": 4.275528325112233e-07, "loss": 0.4135, "step": 12291 }, { "epoch": 0.8714640198511167, "grad_norm": 7.4075489884267185, "learning_rate": 4.270883967067868e-07, "loss": 0.4352, "step": 12292 }, { "epoch": 0.871534916696207, "grad_norm": 2.0422569264498134, "learning_rate": 4.2662420203546997e-07, "loss": 0.328, "step": 12293 }, { "epoch": 0.8716058135412974, "grad_norm": 3.9911201896537225, "learning_rate": 4.2616024852174863e-07, "loss": 0.3472, "step": 12294 }, { "epoch": 0.8716767103863878, "grad_norm": 2.286951803837129, "learning_rate": 4.256965361900889e-07, "loss": 0.416, "step": 12295 }, { "epoch": 0.8717476072314781, "grad_norm": 2.347066733870281, "learning_rate": 4.2523306506494165e-07, "loss": 0.3722, "step": 12296 }, { "epoch": 0.8718185040765686, "grad_norm": 2.6141481101538195, "learning_rate": 4.2476983517074744e-07, "loss": 0.4003, "step": 12297 }, { "epoch": 0.871889400921659, "grad_norm": 2.9107315127944697, "learning_rate": 4.243068465319322e-07, "loss": 0.4188, "step": 12298 }, { "epoch": 0.8719602977667493, "grad_norm": 3.551340321251803, "learning_rate": 4.238440991729081e-07, "loss": 0.3644, "step": 12299 }, { "epoch": 0.8720311946118398, "grad_norm": 3.646496205612789, "learning_rate": 4.233815931180779e-07, "loss": 0.4146, "step": 12300 }, { "epoch": 0.8721020914569302, "grad_norm": 3.245311167641278, "learning_rate": 4.229193283918309e-07, "loss": 0.4204, "step": 12301 }, { "epoch": 0.8721729883020206, "grad_norm": 2.220959004857215, "learning_rate": 4.2245730501853986e-07, "loss": 0.3477, "step": 12302 }, { "epoch": 0.8722438851471109, "grad_norm": 1.9764351533519362, "learning_rate": 4.219955230225686e-07, "loss": 0.3917, "step": 12303 }, { "epoch": 0.8723147819922014, "grad_norm": 5.4492006519249925, "learning_rate": 4.2153398242826884e-07, "loss": 0.3473, "step": 12304 }, { "epoch": 0.8723856788372918, "grad_norm": 1.9526736589354479, "learning_rate": 4.2107268325997653e-07, "loss": 0.3572, "step": 12305 }, { "epoch": 0.8724565756823821, "grad_norm": 2.0382657971693803, "learning_rate": 4.206116255420162e-07, "loss": 0.3987, "step": 12306 }, { "epoch": 0.8725274725274725, "grad_norm": 2.093289595050384, "learning_rate": 4.2015080929869943e-07, "loss": 0.3852, "step": 12307 }, { "epoch": 0.872598369372563, "grad_norm": 2.6688400745295393, "learning_rate": 4.1969023455432745e-07, "loss": 0.4028, "step": 12308 }, { "epoch": 0.8726692662176533, "grad_norm": 4.101669520147423, "learning_rate": 4.192299013331852e-07, "loss": 0.3541, "step": 12309 }, { "epoch": 0.8727401630627437, "grad_norm": 1.855216456315884, "learning_rate": 4.187698096595455e-07, "loss": 0.4186, "step": 12310 }, { "epoch": 0.8728110599078341, "grad_norm": 2.7584646354794273, "learning_rate": 4.183099595576717e-07, "loss": 0.4104, "step": 12311 }, { "epoch": 0.8728819567529245, "grad_norm": 2.498582913996158, "learning_rate": 4.1785035105181057e-07, "loss": 0.3625, "step": 12312 }, { "epoch": 0.8729528535980149, "grad_norm": 2.1341296534110494, "learning_rate": 4.1739098416619716e-07, "loss": 0.4115, "step": 12313 }, { "epoch": 0.8730237504431053, "grad_norm": 2.570457681697267, "learning_rate": 4.1693185892505594e-07, "loss": 0.3258, "step": 12314 }, { "epoch": 0.8730946472881956, "grad_norm": 0.9436738308564567, "learning_rate": 4.164729753525942e-07, "loss": 0.6294, "step": 12315 }, { "epoch": 0.8731655441332861, "grad_norm": 3.400314088692568, "learning_rate": 4.1601433347301266e-07, "loss": 0.3832, "step": 12316 }, { "epoch": 0.8732364409783765, "grad_norm": 2.2882125675630074, "learning_rate": 4.155559333104936e-07, "loss": 0.3598, "step": 12317 }, { "epoch": 0.8733073378234668, "grad_norm": 1.0695562746303064, "learning_rate": 4.1509777488920876e-07, "loss": 0.6803, "step": 12318 }, { "epoch": 0.8733782346685572, "grad_norm": 2.8382939689759987, "learning_rate": 4.1463985823331774e-07, "loss": 0.423, "step": 12319 }, { "epoch": 0.8734491315136477, "grad_norm": 3.4153433306065093, "learning_rate": 4.1418218336696733e-07, "loss": 0.4, "step": 12320 }, { "epoch": 0.873520028358738, "grad_norm": 3.3122451757745504, "learning_rate": 4.1372475031429105e-07, "loss": 0.4034, "step": 12321 }, { "epoch": 0.8735909252038284, "grad_norm": 5.344214734651649, "learning_rate": 4.1326755909940787e-07, "loss": 0.4091, "step": 12322 }, { "epoch": 0.8736618220489188, "grad_norm": 1.6073669384337357, "learning_rate": 4.128106097464285e-07, "loss": 0.4263, "step": 12323 }, { "epoch": 0.8737327188940092, "grad_norm": 5.042627958707039, "learning_rate": 4.1235390227944705e-07, "loss": 0.4079, "step": 12324 }, { "epoch": 0.8738036157390996, "grad_norm": 8.941737818751843, "learning_rate": 4.1189743672254424e-07, "loss": 0.3984, "step": 12325 }, { "epoch": 0.87387451258419, "grad_norm": 4.376082232416517, "learning_rate": 4.1144121309979244e-07, "loss": 0.3821, "step": 12326 }, { "epoch": 0.8739454094292805, "grad_norm": 3.669134915224328, "learning_rate": 4.1098523143524805e-07, "loss": 0.3688, "step": 12327 }, { "epoch": 0.8740163062743708, "grad_norm": 5.317669899056497, "learning_rate": 4.105294917529551e-07, "loss": 0.3983, "step": 12328 }, { "epoch": 0.8740872031194612, "grad_norm": 1.0873764408839117, "learning_rate": 4.1007399407694437e-07, "loss": 0.604, "step": 12329 }, { "epoch": 0.8741580999645516, "grad_norm": 3.6619925899498855, "learning_rate": 4.0961873843123613e-07, "loss": 0.3342, "step": 12330 }, { "epoch": 0.874228996809642, "grad_norm": 2.6255625802800138, "learning_rate": 4.0916372483983555e-07, "loss": 0.3468, "step": 12331 }, { "epoch": 0.8742998936547324, "grad_norm": 2.4551217445880313, "learning_rate": 4.0870895332673467e-07, "loss": 0.4079, "step": 12332 }, { "epoch": 0.8743707904998228, "grad_norm": 2.373031572793139, "learning_rate": 4.082544239159164e-07, "loss": 0.4041, "step": 12333 }, { "epoch": 0.8744416873449131, "grad_norm": 2.333532538351931, "learning_rate": 4.0780013663134554e-07, "loss": 0.3873, "step": 12334 }, { "epoch": 0.8745125841900036, "grad_norm": 2.5762273679381376, "learning_rate": 4.0734609149697955e-07, "loss": 0.3335, "step": 12335 }, { "epoch": 0.874583481035094, "grad_norm": 2.06328929472542, "learning_rate": 4.0689228853675933e-07, "loss": 0.39, "step": 12336 }, { "epoch": 0.8746543778801843, "grad_norm": 2.7889778104421503, "learning_rate": 4.0643872777461456e-07, "loss": 0.3681, "step": 12337 }, { "epoch": 0.8747252747252747, "grad_norm": 2.3325852668877416, "learning_rate": 4.059854092344617e-07, "loss": 0.4119, "step": 12338 }, { "epoch": 0.8747961715703652, "grad_norm": 2.1823471382097854, "learning_rate": 4.05532332940205e-07, "loss": 0.4264, "step": 12339 }, { "epoch": 0.8748670684154555, "grad_norm": 2.426633441073714, "learning_rate": 4.0507949891573526e-07, "loss": 0.3729, "step": 12340 }, { "epoch": 0.8749379652605459, "grad_norm": 3.607948801064118, "learning_rate": 4.046269071849296e-07, "loss": 0.3961, "step": 12341 }, { "epoch": 0.8750088621056363, "grad_norm": 4.077522649227048, "learning_rate": 4.0417455777165603e-07, "loss": 0.3588, "step": 12342 }, { "epoch": 0.8750797589507266, "grad_norm": 3.423938955939617, "learning_rate": 4.037224506997656e-07, "loss": 0.4036, "step": 12343 }, { "epoch": 0.8751506557958171, "grad_norm": 2.0287969440855314, "learning_rate": 4.032705859930974e-07, "loss": 0.378, "step": 12344 }, { "epoch": 0.8752215526409075, "grad_norm": 2.8281458336967376, "learning_rate": 4.0281896367547976e-07, "loss": 0.4095, "step": 12345 }, { "epoch": 0.8752924494859978, "grad_norm": 2.1723080313312972, "learning_rate": 4.0236758377072914e-07, "loss": 0.3947, "step": 12346 }, { "epoch": 0.8753633463310883, "grad_norm": 2.2661064517430316, "learning_rate": 4.0191644630264314e-07, "loss": 0.4165, "step": 12347 }, { "epoch": 0.8754342431761787, "grad_norm": 5.875970239673494, "learning_rate": 4.0146555129501217e-07, "loss": 0.364, "step": 12348 }, { "epoch": 0.875505140021269, "grad_norm": 2.1270225480732514, "learning_rate": 4.010148987716139e-07, "loss": 0.3516, "step": 12349 }, { "epoch": 0.8755760368663594, "grad_norm": 5.250203205321188, "learning_rate": 4.005644887562099e-07, "loss": 0.4432, "step": 12350 }, { "epoch": 0.8756469337114499, "grad_norm": 4.46525184139007, "learning_rate": 4.001143212725506e-07, "loss": 0.386, "step": 12351 }, { "epoch": 0.8757178305565403, "grad_norm": 2.3205786996521356, "learning_rate": 3.996643963443736e-07, "loss": 0.327, "step": 12352 }, { "epoch": 0.8757887274016306, "grad_norm": 3.8359158346130746, "learning_rate": 3.992147139954056e-07, "loss": 0.4084, "step": 12353 }, { "epoch": 0.875859624246721, "grad_norm": 2.69169714004679, "learning_rate": 3.9876527424935694e-07, "loss": 0.4208, "step": 12354 }, { "epoch": 0.8759305210918115, "grad_norm": 3.0002510671159346, "learning_rate": 3.983160771299266e-07, "loss": 0.4206, "step": 12355 }, { "epoch": 0.8760014179369018, "grad_norm": 2.0495163985777416, "learning_rate": 3.9786712266080274e-07, "loss": 0.364, "step": 12356 }, { "epoch": 0.8760723147819922, "grad_norm": 5.031048515632338, "learning_rate": 3.974184108656581e-07, "loss": 0.4183, "step": 12357 }, { "epoch": 0.8761432116270826, "grad_norm": 2.32982681187142, "learning_rate": 3.969699417681527e-07, "loss": 0.3506, "step": 12358 }, { "epoch": 0.876214108472173, "grad_norm": 2.9092306408132806, "learning_rate": 3.965217153919365e-07, "loss": 0.431, "step": 12359 }, { "epoch": 0.8762850053172634, "grad_norm": 2.002366040351009, "learning_rate": 3.960737317606433e-07, "loss": 0.4103, "step": 12360 }, { "epoch": 0.8763559021623538, "grad_norm": 2.911781543844271, "learning_rate": 3.9562599089789655e-07, "loss": 0.3332, "step": 12361 }, { "epoch": 0.8764267990074441, "grad_norm": 2.435556155417158, "learning_rate": 3.9517849282730614e-07, "loss": 0.4411, "step": 12362 }, { "epoch": 0.8764976958525346, "grad_norm": 2.2708330946033706, "learning_rate": 3.9473123757246767e-07, "loss": 0.3788, "step": 12363 }, { "epoch": 0.876568592697625, "grad_norm": 1.9906593164168813, "learning_rate": 3.942842251569662e-07, "loss": 0.4167, "step": 12364 }, { "epoch": 0.8766394895427153, "grad_norm": 12.862028356436172, "learning_rate": 3.938374556043734e-07, "loss": 0.3812, "step": 12365 }, { "epoch": 0.8767103863878057, "grad_norm": 3.4775288845679917, "learning_rate": 3.9339092893824705e-07, "loss": 0.3641, "step": 12366 }, { "epoch": 0.8767812832328962, "grad_norm": 2.6863321420007558, "learning_rate": 3.929446451821328e-07, "loss": 0.3801, "step": 12367 }, { "epoch": 0.8768521800779865, "grad_norm": 2.3806080976098447, "learning_rate": 3.9249860435956466e-07, "loss": 0.351, "step": 12368 }, { "epoch": 0.8769230769230769, "grad_norm": 5.699354851847223, "learning_rate": 3.9205280649406154e-07, "loss": 0.3844, "step": 12369 }, { "epoch": 0.8769939737681673, "grad_norm": 3.3265306351514754, "learning_rate": 3.9160725160913017e-07, "loss": 0.4134, "step": 12370 }, { "epoch": 0.8770648706132577, "grad_norm": 3.9659068412176453, "learning_rate": 3.9116193972826623e-07, "loss": 0.3972, "step": 12371 }, { "epoch": 0.8771357674583481, "grad_norm": 2.6681837296123283, "learning_rate": 3.907168708749515e-07, "loss": 0.4138, "step": 12372 }, { "epoch": 0.8772066643034385, "grad_norm": 1.8822909974269368, "learning_rate": 3.902720450726544e-07, "loss": 0.3457, "step": 12373 }, { "epoch": 0.8772775611485288, "grad_norm": 2.589795481826891, "learning_rate": 3.898274623448295e-07, "loss": 0.3861, "step": 12374 }, { "epoch": 0.8773484579936193, "grad_norm": 3.7526646869354714, "learning_rate": 3.893831227149225e-07, "loss": 0.3803, "step": 12375 }, { "epoch": 0.8774193548387097, "grad_norm": 2.8924627078277543, "learning_rate": 3.889390262063625e-07, "loss": 0.3999, "step": 12376 }, { "epoch": 0.8774902516838, "grad_norm": 2.587948645931187, "learning_rate": 3.884951728425668e-07, "loss": 0.3818, "step": 12377 }, { "epoch": 0.8775611485288904, "grad_norm": 2.259265937961221, "learning_rate": 3.8805156264694056e-07, "loss": 0.3948, "step": 12378 }, { "epoch": 0.8776320453739809, "grad_norm": 4.120871679628182, "learning_rate": 3.876081956428751e-07, "loss": 0.4016, "step": 12379 }, { "epoch": 0.8777029422190713, "grad_norm": 12.05705234368467, "learning_rate": 3.871650718537506e-07, "loss": 0.3306, "step": 12380 }, { "epoch": 0.8777738390641616, "grad_norm": 5.974845957254328, "learning_rate": 3.8672219130293177e-07, "loss": 0.3674, "step": 12381 }, { "epoch": 0.877844735909252, "grad_norm": 2.7051154979764025, "learning_rate": 3.8627955401377425e-07, "loss": 0.372, "step": 12382 }, { "epoch": 0.8779156327543425, "grad_norm": 4.32995587322603, "learning_rate": 3.8583716000961614e-07, "loss": 0.4315, "step": 12383 }, { "epoch": 0.8779865295994328, "grad_norm": 4.016387558910458, "learning_rate": 3.8539500931378706e-07, "loss": 0.4059, "step": 12384 }, { "epoch": 0.8780574264445232, "grad_norm": 2.134181870349658, "learning_rate": 3.849531019496017e-07, "loss": 0.3632, "step": 12385 }, { "epoch": 0.8781283232896137, "grad_norm": 7.209346183850885, "learning_rate": 3.8451143794036086e-07, "loss": 0.4184, "step": 12386 }, { "epoch": 0.878199220134704, "grad_norm": 2.2069616566819024, "learning_rate": 3.840700173093553e-07, "loss": 0.3616, "step": 12387 }, { "epoch": 0.8782701169797944, "grad_norm": 5.311600053701383, "learning_rate": 3.836288400798616e-07, "loss": 0.3898, "step": 12388 }, { "epoch": 0.8783410138248848, "grad_norm": 3.5301940907276594, "learning_rate": 3.8318790627514145e-07, "loss": 0.3892, "step": 12389 }, { "epoch": 0.8784119106699751, "grad_norm": 6.811267266927151, "learning_rate": 3.827472159184464e-07, "loss": 0.3619, "step": 12390 }, { "epoch": 0.8784828075150656, "grad_norm": 2.420354008704248, "learning_rate": 3.823067690330162e-07, "loss": 0.3726, "step": 12391 }, { "epoch": 0.878553704360156, "grad_norm": 1.0314563007614488, "learning_rate": 3.818665656420745e-07, "loss": 0.7011, "step": 12392 }, { "epoch": 0.8786246012052463, "grad_norm": 5.593804798486295, "learning_rate": 3.814266057688326e-07, "loss": 0.3688, "step": 12393 }, { "epoch": 0.8786954980503368, "grad_norm": 3.1016233303565617, "learning_rate": 3.8098688943649155e-07, "loss": 0.4282, "step": 12394 }, { "epoch": 0.8787663948954272, "grad_norm": 3.21269038406795, "learning_rate": 3.8054741666823827e-07, "loss": 0.3932, "step": 12395 }, { "epoch": 0.8788372917405175, "grad_norm": 2.699678111909758, "learning_rate": 3.801081874872442e-07, "loss": 0.4044, "step": 12396 }, { "epoch": 0.8789081885856079, "grad_norm": 5.801498352024839, "learning_rate": 3.79669201916672e-07, "loss": 0.4121, "step": 12397 }, { "epoch": 0.8789790854306984, "grad_norm": 3.0956219341071813, "learning_rate": 3.792304599796698e-07, "loss": 0.453, "step": 12398 }, { "epoch": 0.8790499822757887, "grad_norm": 2.7835333067223984, "learning_rate": 3.787919616993724e-07, "loss": 0.364, "step": 12399 }, { "epoch": 0.8791208791208791, "grad_norm": 5.464297765427272, "learning_rate": 3.7835370709890073e-07, "loss": 0.3359, "step": 12400 }, { "epoch": 0.8791917759659695, "grad_norm": 2.3895293627534024, "learning_rate": 3.7791569620136693e-07, "loss": 0.3711, "step": 12401 }, { "epoch": 0.8792626728110599, "grad_norm": 4.924434934227834, "learning_rate": 3.7747792902986635e-07, "loss": 0.381, "step": 12402 }, { "epoch": 0.8793335696561503, "grad_norm": 3.8034625639798847, "learning_rate": 3.770404056074822e-07, "loss": 0.3884, "step": 12403 }, { "epoch": 0.8794044665012407, "grad_norm": 2.611320538735024, "learning_rate": 3.7660312595728665e-07, "loss": 0.3469, "step": 12404 }, { "epoch": 0.8794753633463311, "grad_norm": 2.8922529485758552, "learning_rate": 3.7616609010233616e-07, "loss": 0.3952, "step": 12405 }, { "epoch": 0.8795462601914215, "grad_norm": 1.7292654765695554, "learning_rate": 3.757292980656779e-07, "loss": 0.3787, "step": 12406 }, { "epoch": 0.8796171570365119, "grad_norm": 1.3211906646340439, "learning_rate": 3.752927498703429e-07, "loss": 0.3436, "step": 12407 }, { "epoch": 0.8796880538816023, "grad_norm": 1.9584782950771167, "learning_rate": 3.7485644553935054e-07, "loss": 0.411, "step": 12408 }, { "epoch": 0.8797589507266926, "grad_norm": 3.5413746153439774, "learning_rate": 3.7442038509570843e-07, "loss": 0.393, "step": 12409 }, { "epoch": 0.8798298475717831, "grad_norm": 1.0895705023517666, "learning_rate": 3.739845685624105e-07, "loss": 0.6838, "step": 12410 }, { "epoch": 0.8799007444168735, "grad_norm": 2.076243477480486, "learning_rate": 3.7354899596243665e-07, "loss": 0.3897, "step": 12411 }, { "epoch": 0.8799716412619638, "grad_norm": 4.122655774928503, "learning_rate": 3.731136673187552e-07, "loss": 0.3694, "step": 12412 }, { "epoch": 0.8800425381070542, "grad_norm": 4.135808374676547, "learning_rate": 3.726785826543222e-07, "loss": 0.3723, "step": 12413 }, { "epoch": 0.8801134349521447, "grad_norm": 3.414022447037839, "learning_rate": 3.722437419920799e-07, "loss": 0.4166, "step": 12414 }, { "epoch": 0.880184331797235, "grad_norm": 6.8166967429238285, "learning_rate": 3.7180914535495605e-07, "loss": 0.3953, "step": 12415 }, { "epoch": 0.8802552286423254, "grad_norm": 5.569044311187444, "learning_rate": 3.713747927658684e-07, "loss": 0.3628, "step": 12416 }, { "epoch": 0.8803261254874158, "grad_norm": 4.251264815772616, "learning_rate": 3.7094068424772146e-07, "loss": 0.3708, "step": 12417 }, { "epoch": 0.8803970223325062, "grad_norm": 3.0888393058353674, "learning_rate": 3.7050681982340576e-07, "loss": 0.381, "step": 12418 }, { "epoch": 0.8804679191775966, "grad_norm": 2.8220805761278633, "learning_rate": 3.70073199515798e-07, "loss": 0.3889, "step": 12419 }, { "epoch": 0.880538816022687, "grad_norm": 2.6962750025606046, "learning_rate": 3.696398233477655e-07, "loss": 0.3759, "step": 12420 }, { "epoch": 0.8806097128677773, "grad_norm": 4.378115695952178, "learning_rate": 3.692066913421588e-07, "loss": 0.3707, "step": 12421 }, { "epoch": 0.8806806097128678, "grad_norm": 1.9801444659673022, "learning_rate": 3.6877380352181747e-07, "loss": 0.365, "step": 12422 }, { "epoch": 0.8807515065579582, "grad_norm": 4.113590402188184, "learning_rate": 3.683411599095682e-07, "loss": 0.3821, "step": 12423 }, { "epoch": 0.8808224034030485, "grad_norm": 2.956034606099359, "learning_rate": 3.6790876052822554e-07, "loss": 0.4384, "step": 12424 }, { "epoch": 0.8808933002481389, "grad_norm": 2.333608606202001, "learning_rate": 3.6747660540058906e-07, "loss": 0.3832, "step": 12425 }, { "epoch": 0.8809641970932294, "grad_norm": 2.247097434616372, "learning_rate": 3.6704469454944657e-07, "loss": 0.3747, "step": 12426 }, { "epoch": 0.8810350939383197, "grad_norm": 3.7433044333566925, "learning_rate": 3.6661302799757436e-07, "loss": 0.4114, "step": 12427 }, { "epoch": 0.8811059907834101, "grad_norm": 2.4772851731713486, "learning_rate": 3.661816057677325e-07, "loss": 0.3684, "step": 12428 }, { "epoch": 0.8811768876285005, "grad_norm": 5.022239822216108, "learning_rate": 3.657504278826729e-07, "loss": 0.3992, "step": 12429 }, { "epoch": 0.881247784473591, "grad_norm": 2.912674981601799, "learning_rate": 3.653194943651295e-07, "loss": 0.4069, "step": 12430 }, { "epoch": 0.8813186813186813, "grad_norm": 2.5055299667926008, "learning_rate": 3.6488880523782635e-07, "loss": 0.3464, "step": 12431 }, { "epoch": 0.8813895781637717, "grad_norm": 0.9557243645090909, "learning_rate": 3.6445836052347526e-07, "loss": 0.6346, "step": 12432 }, { "epoch": 0.8814604750088622, "grad_norm": 1.7980221167466048, "learning_rate": 3.6402816024477315e-07, "loss": 0.3857, "step": 12433 }, { "epoch": 0.8815313718539525, "grad_norm": 1.7331494021991698, "learning_rate": 3.6359820442440294e-07, "loss": 0.3598, "step": 12434 }, { "epoch": 0.8816022686990429, "grad_norm": 3.251471915694995, "learning_rate": 3.6316849308503875e-07, "loss": 0.4055, "step": 12435 }, { "epoch": 0.8816731655441333, "grad_norm": 5.778467419257107, "learning_rate": 3.6273902624934023e-07, "loss": 0.3991, "step": 12436 }, { "epoch": 0.8817440623892236, "grad_norm": 2.38609421609928, "learning_rate": 3.62309803939952e-07, "loss": 0.4045, "step": 12437 }, { "epoch": 0.8818149592343141, "grad_norm": 3.37159381582094, "learning_rate": 3.618808261795065e-07, "loss": 0.3785, "step": 12438 }, { "epoch": 0.8818858560794045, "grad_norm": 2.8975172881523785, "learning_rate": 3.614520929906257e-07, "loss": 0.4092, "step": 12439 }, { "epoch": 0.8819567529244948, "grad_norm": 3.74282219808129, "learning_rate": 3.6102360439591753e-07, "loss": 0.3865, "step": 12440 }, { "epoch": 0.8820276497695853, "grad_norm": 4.511383210896751, "learning_rate": 3.6059536041797406e-07, "loss": 0.2988, "step": 12441 }, { "epoch": 0.8820985466146757, "grad_norm": 22.530089776659462, "learning_rate": 3.601673610793788e-07, "loss": 0.3914, "step": 12442 }, { "epoch": 0.882169443459766, "grad_norm": 3.7119412696968173, "learning_rate": 3.5973960640270034e-07, "loss": 0.4069, "step": 12443 }, { "epoch": 0.8822403403048564, "grad_norm": 4.658634399242142, "learning_rate": 3.59312096410494e-07, "loss": 0.358, "step": 12444 }, { "epoch": 0.8823112371499469, "grad_norm": 2.4229042898741646, "learning_rate": 3.588848311253024e-07, "loss": 0.3765, "step": 12445 }, { "epoch": 0.8823821339950372, "grad_norm": 8.632014990871488, "learning_rate": 3.584578105696568e-07, "loss": 0.405, "step": 12446 }, { "epoch": 0.8824530308401276, "grad_norm": 5.0664110907706466, "learning_rate": 3.580310347660726e-07, "loss": 0.3605, "step": 12447 }, { "epoch": 0.882523927685218, "grad_norm": 4.717330034024938, "learning_rate": 3.5760450373705626e-07, "loss": 0.4059, "step": 12448 }, { "epoch": 0.8825948245303084, "grad_norm": 2.6582868404755016, "learning_rate": 3.571782175050975e-07, "loss": 0.343, "step": 12449 }, { "epoch": 0.8826657213753988, "grad_norm": 2.629104234534223, "learning_rate": 3.56752176092674e-07, "loss": 0.3536, "step": 12450 }, { "epoch": 0.8827366182204892, "grad_norm": 2.3558258885518257, "learning_rate": 3.563263795222538e-07, "loss": 0.4131, "step": 12451 }, { "epoch": 0.8828075150655795, "grad_norm": 3.683456662639358, "learning_rate": 3.559008278162862e-07, "loss": 0.3407, "step": 12452 }, { "epoch": 0.88287841191067, "grad_norm": 3.1534277078298474, "learning_rate": 3.554755209972144e-07, "loss": 0.4059, "step": 12453 }, { "epoch": 0.8829493087557604, "grad_norm": 3.602544401138646, "learning_rate": 3.5505045908746204e-07, "loss": 0.3065, "step": 12454 }, { "epoch": 0.8830202056008508, "grad_norm": 3.843475333574604, "learning_rate": 3.546256421094452e-07, "loss": 0.4034, "step": 12455 }, { "epoch": 0.8830911024459411, "grad_norm": 3.347836787007703, "learning_rate": 3.542010700855636e-07, "loss": 0.4235, "step": 12456 }, { "epoch": 0.8831619992910316, "grad_norm": 3.3786551949101025, "learning_rate": 3.5377674303820497e-07, "loss": 0.3617, "step": 12457 }, { "epoch": 0.883232896136122, "grad_norm": 2.3089451642292027, "learning_rate": 3.5335266098974586e-07, "loss": 0.3733, "step": 12458 }, { "epoch": 0.8833037929812123, "grad_norm": 1.9979373339973185, "learning_rate": 3.529288239625478e-07, "loss": 0.3418, "step": 12459 }, { "epoch": 0.8833746898263027, "grad_norm": 5.587285258969492, "learning_rate": 3.5250523197895847e-07, "loss": 0.3702, "step": 12460 }, { "epoch": 0.8834455866713932, "grad_norm": 2.320034005533903, "learning_rate": 3.5208188506131615e-07, "loss": 0.3596, "step": 12461 }, { "epoch": 0.8835164835164835, "grad_norm": 3.4721689185126716, "learning_rate": 3.5165878323194403e-07, "loss": 0.3747, "step": 12462 }, { "epoch": 0.8835873803615739, "grad_norm": 3.6009764877646706, "learning_rate": 3.512359265131521e-07, "loss": 0.4066, "step": 12463 }, { "epoch": 0.8836582772066643, "grad_norm": 1.7936031051447283, "learning_rate": 3.5081331492723745e-07, "loss": 0.406, "step": 12464 }, { "epoch": 0.8837291740517547, "grad_norm": 4.028538236085072, "learning_rate": 3.5039094849648625e-07, "loss": 0.4089, "step": 12465 }, { "epoch": 0.8838000708968451, "grad_norm": 2.483577901110982, "learning_rate": 3.4996882724316895e-07, "loss": 0.3697, "step": 12466 }, { "epoch": 0.8838709677419355, "grad_norm": 0.9424000371210494, "learning_rate": 3.4954695118954327e-07, "loss": 0.6822, "step": 12467 }, { "epoch": 0.8839418645870258, "grad_norm": 4.718624945019413, "learning_rate": 3.4912532035785704e-07, "loss": 0.4228, "step": 12468 }, { "epoch": 0.8840127614321163, "grad_norm": 3.588520654316209, "learning_rate": 3.4870393477034303e-07, "loss": 0.3996, "step": 12469 }, { "epoch": 0.8840836582772067, "grad_norm": 6.443777492219273, "learning_rate": 3.482827944492212e-07, "loss": 0.371, "step": 12470 }, { "epoch": 0.884154555122297, "grad_norm": 1.7469019724628643, "learning_rate": 3.478618994166971e-07, "loss": 0.3782, "step": 12471 }, { "epoch": 0.8842254519673874, "grad_norm": 10.85578612879679, "learning_rate": 3.474412496949664e-07, "loss": 0.3827, "step": 12472 }, { "epoch": 0.8842963488124779, "grad_norm": 4.887802117949076, "learning_rate": 3.4702084530620907e-07, "loss": 0.3784, "step": 12473 }, { "epoch": 0.8843672456575682, "grad_norm": 1.042354728601916, "learning_rate": 3.4660068627259457e-07, "loss": 0.6772, "step": 12474 }, { "epoch": 0.8844381425026586, "grad_norm": 3.110297496902434, "learning_rate": 3.46180772616278e-07, "loss": 0.3454, "step": 12475 }, { "epoch": 0.884509039347749, "grad_norm": 4.251873491248612, "learning_rate": 3.4576110435940055e-07, "loss": 0.3116, "step": 12476 }, { "epoch": 0.8845799361928394, "grad_norm": 2.915916764314453, "learning_rate": 3.4534168152409343e-07, "loss": 0.3895, "step": 12477 }, { "epoch": 0.8846508330379298, "grad_norm": 4.704460306518821, "learning_rate": 3.4492250413247164e-07, "loss": 0.3863, "step": 12478 }, { "epoch": 0.8847217298830202, "grad_norm": 4.229219795277819, "learning_rate": 3.4450357220663923e-07, "loss": 0.3647, "step": 12479 }, { "epoch": 0.8847926267281107, "grad_norm": 2.8630383849315892, "learning_rate": 3.4408488576868624e-07, "loss": 0.359, "step": 12480 }, { "epoch": 0.884863523573201, "grad_norm": 3.3299333933595388, "learning_rate": 3.4366644484069177e-07, "loss": 0.3971, "step": 12481 }, { "epoch": 0.8849344204182914, "grad_norm": 2.496502782798458, "learning_rate": 3.432482494447198e-07, "loss": 0.3826, "step": 12482 }, { "epoch": 0.8850053172633818, "grad_norm": 2.217807750287631, "learning_rate": 3.428302996028215e-07, "loss": 0.3568, "step": 12483 }, { "epoch": 0.8850762141084721, "grad_norm": 3.2790689221241993, "learning_rate": 3.424125953370355e-07, "loss": 0.3702, "step": 12484 }, { "epoch": 0.8851471109535626, "grad_norm": 2.2729272323543612, "learning_rate": 3.4199513666939077e-07, "loss": 0.3378, "step": 12485 }, { "epoch": 0.885218007798653, "grad_norm": 2.8613001657172643, "learning_rate": 3.4157792362189524e-07, "loss": 0.3892, "step": 12486 }, { "epoch": 0.8852889046437433, "grad_norm": 1.9705361592327595, "learning_rate": 3.4116095621655185e-07, "loss": 0.3762, "step": 12487 }, { "epoch": 0.8853598014888338, "grad_norm": 4.043891982767512, "learning_rate": 3.407442344753481e-07, "loss": 0.3831, "step": 12488 }, { "epoch": 0.8854306983339242, "grad_norm": 5.785125297905116, "learning_rate": 3.403277584202569e-07, "loss": 0.422, "step": 12489 }, { "epoch": 0.8855015951790145, "grad_norm": 2.0026681461099805, "learning_rate": 3.3991152807323904e-07, "loss": 0.3573, "step": 12490 }, { "epoch": 0.8855724920241049, "grad_norm": 2.4424635443937666, "learning_rate": 3.3949554345624357e-07, "loss": 0.3786, "step": 12491 }, { "epoch": 0.8856433888691954, "grad_norm": 4.983578098629294, "learning_rate": 3.390798045912047e-07, "loss": 0.3492, "step": 12492 }, { "epoch": 0.8857142857142857, "grad_norm": 5.445277878833157, "learning_rate": 3.3866431150004655e-07, "loss": 0.3912, "step": 12493 }, { "epoch": 0.8857851825593761, "grad_norm": 2.2591044425328266, "learning_rate": 3.382490642046765e-07, "loss": 0.3838, "step": 12494 }, { "epoch": 0.8858560794044665, "grad_norm": 3.2369566984846103, "learning_rate": 3.378340627269905e-07, "loss": 0.3452, "step": 12495 }, { "epoch": 0.8859269762495569, "grad_norm": 2.3552775600340445, "learning_rate": 3.374193070888743e-07, "loss": 0.3548, "step": 12496 }, { "epoch": 0.8859978730946473, "grad_norm": 2.4015800542698162, "learning_rate": 3.370047973121954e-07, "loss": 0.3584, "step": 12497 }, { "epoch": 0.8860687699397377, "grad_norm": 1.973626030395185, "learning_rate": 3.3659053341881353e-07, "loss": 0.3734, "step": 12498 }, { "epoch": 0.886139666784828, "grad_norm": 2.7362047179179068, "learning_rate": 3.3617651543057183e-07, "loss": 0.3965, "step": 12499 }, { "epoch": 0.8862105636299185, "grad_norm": 2.420334582015741, "learning_rate": 3.357627433693028e-07, "loss": 0.3448, "step": 12500 }, { "epoch": 0.8862814604750089, "grad_norm": 2.6078681001553488, "learning_rate": 3.35349217256824e-07, "loss": 0.4008, "step": 12501 }, { "epoch": 0.8863523573200992, "grad_norm": 6.855850555632697, "learning_rate": 3.349359371149413e-07, "loss": 0.3288, "step": 12502 }, { "epoch": 0.8864232541651896, "grad_norm": 2.7353132218575715, "learning_rate": 3.345229029654468e-07, "loss": 0.3618, "step": 12503 }, { "epoch": 0.8864941510102801, "grad_norm": 2.270429267256203, "learning_rate": 3.3411011483012237e-07, "loss": 0.3394, "step": 12504 }, { "epoch": 0.8865650478553704, "grad_norm": 7.353796782972092, "learning_rate": 3.336975727307312e-07, "loss": 0.3891, "step": 12505 }, { "epoch": 0.8866359447004608, "grad_norm": 2.7567473345566462, "learning_rate": 3.3328527668902875e-07, "loss": 0.4255, "step": 12506 }, { "epoch": 0.8867068415455512, "grad_norm": 2.9314221798790965, "learning_rate": 3.328732267267559e-07, "loss": 0.3906, "step": 12507 }, { "epoch": 0.8867777383906417, "grad_norm": 5.404227764117769, "learning_rate": 3.3246142286564086e-07, "loss": 0.4011, "step": 12508 }, { "epoch": 0.886848635235732, "grad_norm": 2.561224275367087, "learning_rate": 3.3204986512739624e-07, "loss": 0.3619, "step": 12509 }, { "epoch": 0.8869195320808224, "grad_norm": 2.87881641532706, "learning_rate": 3.3163855353372633e-07, "loss": 0.3698, "step": 12510 }, { "epoch": 0.8869904289259128, "grad_norm": 2.1133413322282726, "learning_rate": 3.3122748810631823e-07, "loss": 0.3813, "step": 12511 }, { "epoch": 0.8870613257710032, "grad_norm": 2.5655402996542196, "learning_rate": 3.3081666886684735e-07, "loss": 0.319, "step": 12512 }, { "epoch": 0.8871322226160936, "grad_norm": 5.302658950011422, "learning_rate": 3.304060958369776e-07, "loss": 0.4126, "step": 12513 }, { "epoch": 0.887203119461184, "grad_norm": 3.5385342363749728, "learning_rate": 3.299957690383598e-07, "loss": 0.3568, "step": 12514 }, { "epoch": 0.8872740163062743, "grad_norm": 3.554911242532258, "learning_rate": 3.2958568849262907e-07, "loss": 0.3785, "step": 12515 }, { "epoch": 0.8873449131513648, "grad_norm": 14.349559249586108, "learning_rate": 3.291758542214091e-07, "loss": 0.3769, "step": 12516 }, { "epoch": 0.8874158099964552, "grad_norm": 7.0050559635994984, "learning_rate": 3.2876626624631205e-07, "loss": 0.4362, "step": 12517 }, { "epoch": 0.8874867068415455, "grad_norm": 1.8820945831160372, "learning_rate": 3.28356924588934e-07, "loss": 0.3889, "step": 12518 }, { "epoch": 0.8875576036866359, "grad_norm": 4.037316692826163, "learning_rate": 3.2794782927086264e-07, "loss": 0.3847, "step": 12519 }, { "epoch": 0.8876285005317264, "grad_norm": 0.9655002803665446, "learning_rate": 3.275389803136675e-07, "loss": 0.6681, "step": 12520 }, { "epoch": 0.8876993973768167, "grad_norm": 2.479531411002106, "learning_rate": 3.271303777389079e-07, "loss": 0.3682, "step": 12521 }, { "epoch": 0.8877702942219071, "grad_norm": 2.1192685947855336, "learning_rate": 3.26722021568131e-07, "loss": 0.4034, "step": 12522 }, { "epoch": 0.8878411910669975, "grad_norm": 2.1174539542617525, "learning_rate": 3.2631391182286865e-07, "loss": 0.3783, "step": 12523 }, { "epoch": 0.8879120879120879, "grad_norm": 6.37568957437779, "learning_rate": 3.2590604852464013e-07, "loss": 0.3965, "step": 12524 }, { "epoch": 0.8879829847571783, "grad_norm": 1.7905454990065546, "learning_rate": 3.2549843169495334e-07, "loss": 0.3717, "step": 12525 }, { "epoch": 0.8880538816022687, "grad_norm": 2.337389439260536, "learning_rate": 3.250910613553032e-07, "loss": 0.3889, "step": 12526 }, { "epoch": 0.888124778447359, "grad_norm": 2.2173260196256637, "learning_rate": 3.2468393752716933e-07, "loss": 0.3827, "step": 12527 }, { "epoch": 0.8881956752924495, "grad_norm": 2.6411256419955147, "learning_rate": 3.2427706023201886e-07, "loss": 0.4678, "step": 12528 }, { "epoch": 0.8882665721375399, "grad_norm": 3.1149588078159938, "learning_rate": 3.238704294913081e-07, "loss": 0.409, "step": 12529 }, { "epoch": 0.8883374689826302, "grad_norm": 2.621740634850357, "learning_rate": 3.2346404532648035e-07, "loss": 0.4029, "step": 12530 }, { "epoch": 0.8884083658277206, "grad_norm": 4.297577195575555, "learning_rate": 3.230579077589613e-07, "loss": 0.3812, "step": 12531 }, { "epoch": 0.8884792626728111, "grad_norm": 3.746846782938928, "learning_rate": 3.2265201681016875e-07, "loss": 0.3671, "step": 12532 }, { "epoch": 0.8885501595179015, "grad_norm": 4.988500236445186, "learning_rate": 3.222463725015057e-07, "loss": 0.3899, "step": 12533 }, { "epoch": 0.8886210563629918, "grad_norm": 1.856541972153388, "learning_rate": 3.218409748543622e-07, "loss": 0.3648, "step": 12534 }, { "epoch": 0.8886919532080823, "grad_norm": 2.8038422933859484, "learning_rate": 3.214358238901139e-07, "loss": 0.3711, "step": 12535 }, { "epoch": 0.8887628500531727, "grad_norm": 3.193250122998399, "learning_rate": 3.2103091963012645e-07, "loss": 0.3757, "step": 12536 }, { "epoch": 0.888833746898263, "grad_norm": 1.797014801749069, "learning_rate": 3.20626262095749e-07, "loss": 0.3709, "step": 12537 }, { "epoch": 0.8889046437433534, "grad_norm": 2.149939365682769, "learning_rate": 3.202218513083211e-07, "loss": 0.3878, "step": 12538 }, { "epoch": 0.8889755405884439, "grad_norm": 3.695911025173691, "learning_rate": 3.198176872891667e-07, "loss": 0.3637, "step": 12539 }, { "epoch": 0.8890464374335342, "grad_norm": 1.9859070186131251, "learning_rate": 3.194137700595973e-07, "loss": 0.373, "step": 12540 }, { "epoch": 0.8891173342786246, "grad_norm": 2.440833645153415, "learning_rate": 3.1901009964091345e-07, "loss": 0.4209, "step": 12541 }, { "epoch": 0.889188231123715, "grad_norm": 7.525557895723498, "learning_rate": 3.1860667605439876e-07, "loss": 0.3689, "step": 12542 }, { "epoch": 0.8892591279688054, "grad_norm": 2.9441480197783445, "learning_rate": 3.182034993213279e-07, "loss": 0.3721, "step": 12543 }, { "epoch": 0.8893300248138958, "grad_norm": 1.750710432651098, "learning_rate": 3.1780056946295936e-07, "loss": 0.4689, "step": 12544 }, { "epoch": 0.8894009216589862, "grad_norm": 2.1343084410872204, "learning_rate": 3.1739788650054113e-07, "loss": 0.3986, "step": 12545 }, { "epoch": 0.8894718185040765, "grad_norm": 3.0982641077478044, "learning_rate": 3.169954504553069e-07, "loss": 0.4428, "step": 12546 }, { "epoch": 0.889542715349167, "grad_norm": 6.771247105891925, "learning_rate": 3.165932613484757e-07, "loss": 0.418, "step": 12547 }, { "epoch": 0.8896136121942574, "grad_norm": 6.719665326118863, "learning_rate": 3.1619131920125666e-07, "loss": 0.3971, "step": 12548 }, { "epoch": 0.8896845090393477, "grad_norm": 2.6716815772066402, "learning_rate": 3.157896240348457e-07, "loss": 0.3974, "step": 12549 }, { "epoch": 0.8897554058844381, "grad_norm": 3.6545970990666663, "learning_rate": 3.1538817587042193e-07, "loss": 0.3902, "step": 12550 }, { "epoch": 0.8898263027295286, "grad_norm": 4.748679240174452, "learning_rate": 3.1498697472915505e-07, "loss": 0.434, "step": 12551 }, { "epoch": 0.8898971995746189, "grad_norm": 5.335809874787455, "learning_rate": 3.1458602063220154e-07, "loss": 0.3299, "step": 12552 }, { "epoch": 0.8899680964197093, "grad_norm": 5.331255772115141, "learning_rate": 3.1418531360070394e-07, "loss": 0.4414, "step": 12553 }, { "epoch": 0.8900389932647997, "grad_norm": 2.7580409200472253, "learning_rate": 3.137848536557897e-07, "loss": 0.3656, "step": 12554 }, { "epoch": 0.8901098901098901, "grad_norm": 6.822550339291597, "learning_rate": 3.133846408185776e-07, "loss": 0.3983, "step": 12555 }, { "epoch": 0.8901807869549805, "grad_norm": 13.04426410753084, "learning_rate": 3.1298467511017115e-07, "loss": 0.342, "step": 12556 }, { "epoch": 0.8902516838000709, "grad_norm": 3.823070879262644, "learning_rate": 3.125849565516598e-07, "loss": 0.438, "step": 12557 }, { "epoch": 0.8903225806451613, "grad_norm": 2.906059172047899, "learning_rate": 3.1218548516412097e-07, "loss": 0.3553, "step": 12558 }, { "epoch": 0.8903934774902517, "grad_norm": 3.5242974687989745, "learning_rate": 3.117862609686206e-07, "loss": 0.3898, "step": 12559 }, { "epoch": 0.8904643743353421, "grad_norm": 1.0321139158666512, "learning_rate": 3.1138728398620864e-07, "loss": 0.5921, "step": 12560 }, { "epoch": 0.8905352711804325, "grad_norm": 3.984717819798944, "learning_rate": 3.109885542379232e-07, "loss": 0.372, "step": 12561 }, { "epoch": 0.8906061680255228, "grad_norm": 3.236325856600108, "learning_rate": 3.105900717447913e-07, "loss": 0.3463, "step": 12562 }, { "epoch": 0.8906770648706133, "grad_norm": 1.7308569974031218, "learning_rate": 3.101918365278228e-07, "loss": 0.3582, "step": 12563 }, { "epoch": 0.8907479617157037, "grad_norm": 10.575389832316324, "learning_rate": 3.0979384860801984e-07, "loss": 0.408, "step": 12564 }, { "epoch": 0.890818858560794, "grad_norm": 3.0543188358697546, "learning_rate": 3.0939610800636674e-07, "loss": 0.394, "step": 12565 }, { "epoch": 0.8908897554058844, "grad_norm": 1.144555470359824, "learning_rate": 3.089986147438362e-07, "loss": 0.6534, "step": 12566 }, { "epoch": 0.8909606522509749, "grad_norm": 22.54314443835229, "learning_rate": 3.086013688413897e-07, "loss": 0.3885, "step": 12567 }, { "epoch": 0.8910315490960652, "grad_norm": 2.454847700807308, "learning_rate": 3.082043703199744e-07, "loss": 0.4141, "step": 12568 }, { "epoch": 0.8911024459411556, "grad_norm": 3.5613502830004147, "learning_rate": 3.078076192005225e-07, "loss": 0.404, "step": 12569 }, { "epoch": 0.891173342786246, "grad_norm": 3.5874064126389964, "learning_rate": 3.0741111550395664e-07, "loss": 0.3857, "step": 12570 }, { "epoch": 0.8912442396313364, "grad_norm": 2.2417881968245053, "learning_rate": 3.070148592511846e-07, "loss": 0.3841, "step": 12571 }, { "epoch": 0.8913151364764268, "grad_norm": 5.550171826260561, "learning_rate": 3.066188504631012e-07, "loss": 0.4073, "step": 12572 }, { "epoch": 0.8913860333215172, "grad_norm": 3.0566115923739607, "learning_rate": 3.0622308916058774e-07, "loss": 0.4124, "step": 12573 }, { "epoch": 0.8914569301666075, "grad_norm": 3.7791937083374463, "learning_rate": 3.058275753645129e-07, "loss": 0.3673, "step": 12574 }, { "epoch": 0.891527827011698, "grad_norm": 2.47334241116987, "learning_rate": 3.0543230909573507e-07, "loss": 0.3783, "step": 12575 }, { "epoch": 0.8915987238567884, "grad_norm": 2.284568843107838, "learning_rate": 3.0503729037509255e-07, "loss": 0.3753, "step": 12576 }, { "epoch": 0.8916696207018787, "grad_norm": 2.677080877273175, "learning_rate": 3.046425192234176e-07, "loss": 0.3623, "step": 12577 }, { "epoch": 0.8917405175469691, "grad_norm": 2.2807282243970777, "learning_rate": 3.042479956615269e-07, "loss": 0.3978, "step": 12578 }, { "epoch": 0.8918114143920596, "grad_norm": 14.409057356748805, "learning_rate": 3.038537197102237e-07, "loss": 0.3851, "step": 12579 }, { "epoch": 0.8918823112371499, "grad_norm": 4.074830035427227, "learning_rate": 3.034596913902976e-07, "loss": 0.4039, "step": 12580 }, { "epoch": 0.8919532080822403, "grad_norm": 3.5672625067113333, "learning_rate": 3.0306591072252753e-07, "loss": 0.4154, "step": 12581 }, { "epoch": 0.8920241049273308, "grad_norm": 3.405173480405693, "learning_rate": 3.026723777276758e-07, "loss": 0.3886, "step": 12582 }, { "epoch": 0.8920950017724212, "grad_norm": 1.8997046540330318, "learning_rate": 3.0227909242649567e-07, "loss": 0.4059, "step": 12583 }, { "epoch": 0.8921658986175115, "grad_norm": 1.626946818065104, "learning_rate": 3.018860548397251e-07, "loss": 0.3682, "step": 12584 }, { "epoch": 0.8922367954626019, "grad_norm": 5.679211494079793, "learning_rate": 3.014932649880875e-07, "loss": 0.3729, "step": 12585 }, { "epoch": 0.8923076923076924, "grad_norm": 3.3307703669970103, "learning_rate": 3.011007228922974e-07, "loss": 0.3749, "step": 12586 }, { "epoch": 0.8923785891527827, "grad_norm": 2.0473942979635735, "learning_rate": 3.007084285730516e-07, "loss": 0.3626, "step": 12587 }, { "epoch": 0.8924494859978731, "grad_norm": 3.8780219216082004, "learning_rate": 3.0031638205103795e-07, "loss": 0.3881, "step": 12588 }, { "epoch": 0.8925203828429635, "grad_norm": 2.2980210063545963, "learning_rate": 2.999245833469283e-07, "loss": 0.4162, "step": 12589 }, { "epoch": 0.8925912796880539, "grad_norm": 2.4155873477638257, "learning_rate": 2.9953303248138277e-07, "loss": 0.3792, "step": 12590 }, { "epoch": 0.8926621765331443, "grad_norm": 13.218451324981485, "learning_rate": 2.9914172947504873e-07, "loss": 0.3882, "step": 12591 }, { "epoch": 0.8927330733782347, "grad_norm": 3.5805868175509876, "learning_rate": 2.987506743485585e-07, "loss": 0.4556, "step": 12592 }, { "epoch": 0.892803970223325, "grad_norm": 7.9743738094326115, "learning_rate": 2.983598671225335e-07, "loss": 0.4117, "step": 12593 }, { "epoch": 0.8928748670684155, "grad_norm": 1.765335928627929, "learning_rate": 2.979693078175827e-07, "loss": 0.4064, "step": 12594 }, { "epoch": 0.8929457639135059, "grad_norm": 2.7471650674896715, "learning_rate": 2.9757899645429745e-07, "loss": 0.3796, "step": 12595 }, { "epoch": 0.8930166607585962, "grad_norm": 3.509439894698031, "learning_rate": 2.9718893305326124e-07, "loss": 0.4119, "step": 12596 }, { "epoch": 0.8930875576036866, "grad_norm": 2.4712551697702954, "learning_rate": 2.9679911763504266e-07, "loss": 0.3762, "step": 12597 }, { "epoch": 0.8931584544487771, "grad_norm": 3.1714767417661625, "learning_rate": 2.9640955022019633e-07, "loss": 0.4078, "step": 12598 }, { "epoch": 0.8932293512938674, "grad_norm": 4.026660042758413, "learning_rate": 2.960202308292642e-07, "loss": 0.3426, "step": 12599 }, { "epoch": 0.8933002481389578, "grad_norm": 3.6979555560549193, "learning_rate": 2.956311594827749e-07, "loss": 0.3791, "step": 12600 }, { "epoch": 0.8933711449840482, "grad_norm": 3.05647811156858, "learning_rate": 2.9524233620124687e-07, "loss": 0.4217, "step": 12601 }, { "epoch": 0.8934420418291386, "grad_norm": 2.7558709414583054, "learning_rate": 2.9485376100518104e-07, "loss": 0.3556, "step": 12602 }, { "epoch": 0.893512938674229, "grad_norm": 5.228303778312955, "learning_rate": 2.9446543391506655e-07, "loss": 0.3362, "step": 12603 }, { "epoch": 0.8935838355193194, "grad_norm": 6.113273188762362, "learning_rate": 2.9407735495138257e-07, "loss": 0.4174, "step": 12604 }, { "epoch": 0.8936547323644097, "grad_norm": 2.535760220832407, "learning_rate": 2.9368952413459163e-07, "loss": 0.367, "step": 12605 }, { "epoch": 0.8937256292095002, "grad_norm": 3.882288145889578, "learning_rate": 2.933019414851429e-07, "loss": 0.3549, "step": 12606 }, { "epoch": 0.8937965260545906, "grad_norm": 6.663225061889594, "learning_rate": 2.929146070234762e-07, "loss": 0.3971, "step": 12607 }, { "epoch": 0.893867422899681, "grad_norm": 4.795091492651858, "learning_rate": 2.9252752077001456e-07, "loss": 0.4262, "step": 12608 }, { "epoch": 0.8939383197447713, "grad_norm": 5.556714326853047, "learning_rate": 2.921406827451706e-07, "loss": 0.4144, "step": 12609 }, { "epoch": 0.8940092165898618, "grad_norm": 3.708385827395407, "learning_rate": 2.917540929693413e-07, "loss": 0.4032, "step": 12610 }, { "epoch": 0.8940801134349522, "grad_norm": 3.4783874271722697, "learning_rate": 2.9136775146291207e-07, "loss": 0.3488, "step": 12611 }, { "epoch": 0.8941510102800425, "grad_norm": 2.539228228891901, "learning_rate": 2.909816582462549e-07, "loss": 0.385, "step": 12612 }, { "epoch": 0.8942219071251329, "grad_norm": 1.9952730879283447, "learning_rate": 2.9059581333973075e-07, "loss": 0.3455, "step": 12613 }, { "epoch": 0.8942928039702234, "grad_norm": 7.188765207737226, "learning_rate": 2.902102167636822e-07, "loss": 0.4224, "step": 12614 }, { "epoch": 0.8943637008153137, "grad_norm": 8.033110206999702, "learning_rate": 2.8982486853844414e-07, "loss": 0.4115, "step": 12615 }, { "epoch": 0.8944345976604041, "grad_norm": 2.8571151632051563, "learning_rate": 2.8943976868433643e-07, "loss": 0.3644, "step": 12616 }, { "epoch": 0.8945054945054945, "grad_norm": 2.134731935035808, "learning_rate": 2.8905491722166503e-07, "loss": 0.4051, "step": 12617 }, { "epoch": 0.8945763913505849, "grad_norm": 2.097678122896026, "learning_rate": 2.886703141707231e-07, "loss": 0.4025, "step": 12618 }, { "epoch": 0.8946472881956753, "grad_norm": 4.589555523012023, "learning_rate": 2.882859595517912e-07, "loss": 0.4456, "step": 12619 }, { "epoch": 0.8947181850407657, "grad_norm": 2.6715581107453477, "learning_rate": 2.8790185338513855e-07, "loss": 0.3802, "step": 12620 }, { "epoch": 0.894789081885856, "grad_norm": 2.851515380466037, "learning_rate": 2.8751799569101626e-07, "loss": 0.4035, "step": 12621 }, { "epoch": 0.8948599787309465, "grad_norm": 2.7398411542836425, "learning_rate": 2.8713438648966697e-07, "loss": 0.3671, "step": 12622 }, { "epoch": 0.8949308755760369, "grad_norm": 2.757860500491518, "learning_rate": 2.8675102580131896e-07, "loss": 0.3388, "step": 12623 }, { "epoch": 0.8950017724211272, "grad_norm": 3.1836122195236136, "learning_rate": 2.8636791364618724e-07, "loss": 0.381, "step": 12624 }, { "epoch": 0.8950726692662176, "grad_norm": 2.366586212637946, "learning_rate": 2.8598505004447217e-07, "loss": 0.3639, "step": 12625 }, { "epoch": 0.8951435661113081, "grad_norm": 5.7592955250327575, "learning_rate": 2.856024350163644e-07, "loss": 0.3581, "step": 12626 }, { "epoch": 0.8952144629563984, "grad_norm": 2.2305852587774657, "learning_rate": 2.852200685820383e-07, "loss": 0.3712, "step": 12627 }, { "epoch": 0.8952853598014888, "grad_norm": 3.6222376279260944, "learning_rate": 2.8483795076165666e-07, "loss": 0.3978, "step": 12628 }, { "epoch": 0.8953562566465793, "grad_norm": 4.767696848413595, "learning_rate": 2.844560815753683e-07, "loss": 0.3628, "step": 12629 }, { "epoch": 0.8954271534916696, "grad_norm": 3.0592218985510766, "learning_rate": 2.840744610433105e-07, "loss": 0.3905, "step": 12630 }, { "epoch": 0.89549805033676, "grad_norm": 2.7539104221657253, "learning_rate": 2.83693089185606e-07, "loss": 0.3995, "step": 12631 }, { "epoch": 0.8955689471818504, "grad_norm": 3.746470360748079, "learning_rate": 2.833119660223643e-07, "loss": 0.3505, "step": 12632 }, { "epoch": 0.8956398440269407, "grad_norm": 2.306771640275476, "learning_rate": 2.8293109157368323e-07, "loss": 0.3835, "step": 12633 }, { "epoch": 0.8957107408720312, "grad_norm": 2.531394823052774, "learning_rate": 2.8255046585964505e-07, "loss": 0.4146, "step": 12634 }, { "epoch": 0.8957816377171216, "grad_norm": 3.4918040490465327, "learning_rate": 2.821700889003226e-07, "loss": 0.3237, "step": 12635 }, { "epoch": 0.895852534562212, "grad_norm": 2.2762589063123273, "learning_rate": 2.81789960715772e-07, "loss": 0.3305, "step": 12636 }, { "epoch": 0.8959234314073024, "grad_norm": 2.6959310743164107, "learning_rate": 2.8141008132603733e-07, "loss": 0.4354, "step": 12637 }, { "epoch": 0.8959943282523928, "grad_norm": 4.492143431570406, "learning_rate": 2.8103045075115074e-07, "loss": 0.3807, "step": 12638 }, { "epoch": 0.8960652250974832, "grad_norm": 2.1587060279032397, "learning_rate": 2.8065106901113195e-07, "loss": 0.3692, "step": 12639 }, { "epoch": 0.8961361219425735, "grad_norm": 11.926721497768922, "learning_rate": 2.802719361259826e-07, "loss": 0.3759, "step": 12640 }, { "epoch": 0.896207018787664, "grad_norm": 4.085139475197685, "learning_rate": 2.798930521156967e-07, "loss": 0.3798, "step": 12641 }, { "epoch": 0.8962779156327544, "grad_norm": 5.460873455999376, "learning_rate": 2.795144170002534e-07, "loss": 0.4073, "step": 12642 }, { "epoch": 0.8963488124778447, "grad_norm": 4.359141969637336, "learning_rate": 2.7913603079961826e-07, "loss": 0.3468, "step": 12643 }, { "epoch": 0.8964197093229351, "grad_norm": 2.1177735480005655, "learning_rate": 2.7875789353374206e-07, "loss": 0.3595, "step": 12644 }, { "epoch": 0.8964906061680256, "grad_norm": 5.3538146805394025, "learning_rate": 2.783800052225666e-07, "loss": 0.374, "step": 12645 }, { "epoch": 0.8965615030131159, "grad_norm": 5.016133750651593, "learning_rate": 2.7800236588601703e-07, "loss": 0.3973, "step": 12646 }, { "epoch": 0.8966323998582063, "grad_norm": 1.0551454355725305, "learning_rate": 2.7762497554400745e-07, "loss": 0.6086, "step": 12647 }, { "epoch": 0.8967032967032967, "grad_norm": 3.2793963936619077, "learning_rate": 2.772478342164364e-07, "loss": 0.4365, "step": 12648 }, { "epoch": 0.896774193548387, "grad_norm": 2.2737467030912577, "learning_rate": 2.7687094192319295e-07, "loss": 0.3945, "step": 12649 }, { "epoch": 0.8968450903934775, "grad_norm": 2.452158743416558, "learning_rate": 2.7649429868414955e-07, "loss": 0.3526, "step": 12650 }, { "epoch": 0.8969159872385679, "grad_norm": 5.150659638111997, "learning_rate": 2.7611790451916585e-07, "loss": 0.4345, "step": 12651 }, { "epoch": 0.8969868840836582, "grad_norm": 4.16772106045706, "learning_rate": 2.757417594480916e-07, "loss": 0.3981, "step": 12652 }, { "epoch": 0.8970577809287487, "grad_norm": 5.4312581375920805, "learning_rate": 2.7536586349075965e-07, "loss": 0.3962, "step": 12653 }, { "epoch": 0.8971286777738391, "grad_norm": 6.548731136076083, "learning_rate": 2.7499021666699265e-07, "loss": 0.3723, "step": 12654 }, { "epoch": 0.8971995746189294, "grad_norm": 4.454892442008003, "learning_rate": 2.7461481899659806e-07, "loss": 0.3844, "step": 12655 }, { "epoch": 0.8972704714640198, "grad_norm": 2.5811304872571825, "learning_rate": 2.742396704993694e-07, "loss": 0.3458, "step": 12656 }, { "epoch": 0.8973413683091103, "grad_norm": 5.017218960850631, "learning_rate": 2.7386477119509036e-07, "loss": 0.3827, "step": 12657 }, { "epoch": 0.8974122651542006, "grad_norm": 2.7770916812540842, "learning_rate": 2.7349012110353015e-07, "loss": 0.3309, "step": 12658 }, { "epoch": 0.897483161999291, "grad_norm": 2.4592979544168707, "learning_rate": 2.731157202444434e-07, "loss": 0.3618, "step": 12659 }, { "epoch": 0.8975540588443814, "grad_norm": 4.661553754616766, "learning_rate": 2.727415686375717e-07, "loss": 0.3771, "step": 12660 }, { "epoch": 0.8976249556894719, "grad_norm": 2.902395757166593, "learning_rate": 2.723676663026459e-07, "loss": 0.3972, "step": 12661 }, { "epoch": 0.8976958525345622, "grad_norm": 1.7084619935076706, "learning_rate": 2.719940132593812e-07, "loss": 0.3572, "step": 12662 }, { "epoch": 0.8977667493796526, "grad_norm": 4.404734135712125, "learning_rate": 2.716206095274809e-07, "loss": 0.3889, "step": 12663 }, { "epoch": 0.897837646224743, "grad_norm": 4.46626787506566, "learning_rate": 2.712474551266342e-07, "loss": 0.3676, "step": 12664 }, { "epoch": 0.8979085430698334, "grad_norm": 4.194816217616172, "learning_rate": 2.7087455007651977e-07, "loss": 0.4741, "step": 12665 }, { "epoch": 0.8979794399149238, "grad_norm": 6.035738724169445, "learning_rate": 2.7050189439679966e-07, "loss": 0.4007, "step": 12666 }, { "epoch": 0.8980503367600142, "grad_norm": 3.060668703591331, "learning_rate": 2.7012948810712367e-07, "loss": 0.3466, "step": 12667 }, { "epoch": 0.8981212336051045, "grad_norm": 5.2271337092214205, "learning_rate": 2.697573312271307e-07, "loss": 0.4025, "step": 12668 }, { "epoch": 0.898192130450195, "grad_norm": 3.1168949123273917, "learning_rate": 2.6938542377644426e-07, "loss": 0.4474, "step": 12669 }, { "epoch": 0.8982630272952854, "grad_norm": 4.21492092478491, "learning_rate": 2.690137657746744e-07, "loss": 0.4019, "step": 12670 }, { "epoch": 0.8983339241403757, "grad_norm": 2.490564251575088, "learning_rate": 2.686423572414204e-07, "loss": 0.3916, "step": 12671 }, { "epoch": 0.8984048209854661, "grad_norm": 3.0068097240260805, "learning_rate": 2.682711981962649e-07, "loss": 0.3848, "step": 12672 }, { "epoch": 0.8984757178305566, "grad_norm": 33.139544387908, "learning_rate": 2.6790028865878225e-07, "loss": 0.3953, "step": 12673 }, { "epoch": 0.8985466146756469, "grad_norm": 2.4626257291936495, "learning_rate": 2.6752962864852796e-07, "loss": 0.3504, "step": 12674 }, { "epoch": 0.8986175115207373, "grad_norm": 1.729417799819082, "learning_rate": 2.671592181850491e-07, "loss": 0.3585, "step": 12675 }, { "epoch": 0.8986884083658278, "grad_norm": 4.101305713571953, "learning_rate": 2.6678905728787785e-07, "loss": 0.3564, "step": 12676 }, { "epoch": 0.8987593052109181, "grad_norm": 2.4858139420351626, "learning_rate": 2.6641914597653083e-07, "loss": 0.3594, "step": 12677 }, { "epoch": 0.8988302020560085, "grad_norm": 1.6996458554913496, "learning_rate": 2.6604948427051634e-07, "loss": 0.3423, "step": 12678 }, { "epoch": 0.8989010989010989, "grad_norm": 2.2709404967550277, "learning_rate": 2.656800721893249e-07, "loss": 0.3456, "step": 12679 }, { "epoch": 0.8989719957461892, "grad_norm": 2.2008511608346217, "learning_rate": 2.653109097524376e-07, "loss": 0.4297, "step": 12680 }, { "epoch": 0.8990428925912797, "grad_norm": 4.133248641311236, "learning_rate": 2.6494199697931935e-07, "loss": 0.4259, "step": 12681 }, { "epoch": 0.8991137894363701, "grad_norm": 2.4945539698831705, "learning_rate": 2.645733338894235e-07, "loss": 0.3958, "step": 12682 }, { "epoch": 0.8991846862814604, "grad_norm": 6.153850555051838, "learning_rate": 2.6420492050219017e-07, "loss": 0.401, "step": 12683 }, { "epoch": 0.8992555831265509, "grad_norm": 2.3716126571597336, "learning_rate": 2.638367568370476e-07, "loss": 0.351, "step": 12684 }, { "epoch": 0.8993264799716413, "grad_norm": 2.1909289504650684, "learning_rate": 2.634688429134058e-07, "loss": 0.3831, "step": 12685 }, { "epoch": 0.8993973768167317, "grad_norm": 2.149598456995416, "learning_rate": 2.6310117875066766e-07, "loss": 0.4076, "step": 12686 }, { "epoch": 0.899468273661822, "grad_norm": 1.1062398364028756, "learning_rate": 2.627337643682204e-07, "loss": 0.6756, "step": 12687 }, { "epoch": 0.8995391705069125, "grad_norm": 2.2510821346162526, "learning_rate": 2.6236659978543744e-07, "loss": 0.4135, "step": 12688 }, { "epoch": 0.8996100673520029, "grad_norm": 2.615553967160515, "learning_rate": 2.619996850216794e-07, "loss": 0.3692, "step": 12689 }, { "epoch": 0.8996809641970932, "grad_norm": 1.9839313439881847, "learning_rate": 2.616330200962941e-07, "loss": 0.3855, "step": 12690 }, { "epoch": 0.8997518610421836, "grad_norm": 2.0073033684448895, "learning_rate": 2.6126660502861667e-07, "loss": 0.3497, "step": 12691 }, { "epoch": 0.8998227578872741, "grad_norm": 3.449078630836799, "learning_rate": 2.6090043983796834e-07, "loss": 0.3756, "step": 12692 }, { "epoch": 0.8998936547323644, "grad_norm": 3.258269184189014, "learning_rate": 2.6053452454365636e-07, "loss": 0.4002, "step": 12693 }, { "epoch": 0.8999645515774548, "grad_norm": 2.282413298151459, "learning_rate": 2.6016885916497705e-07, "loss": 0.404, "step": 12694 }, { "epoch": 0.9000354484225452, "grad_norm": 2.220472916971408, "learning_rate": 2.598034437212116e-07, "loss": 0.348, "step": 12695 }, { "epoch": 0.9001063452676356, "grad_norm": 2.4471587382095668, "learning_rate": 2.5943827823162794e-07, "loss": 0.4288, "step": 12696 }, { "epoch": 0.900177242112726, "grad_norm": 3.687088373147748, "learning_rate": 2.5907336271548235e-07, "loss": 0.324, "step": 12697 }, { "epoch": 0.9002481389578164, "grad_norm": 2.219519340757293, "learning_rate": 2.5870869719201666e-07, "loss": 0.415, "step": 12698 }, { "epoch": 0.9003190358029067, "grad_norm": 2.2926117095119376, "learning_rate": 2.58344281680461e-07, "loss": 0.3866, "step": 12699 }, { "epoch": 0.9003899326479972, "grad_norm": 3.8385258280972048, "learning_rate": 2.579801162000306e-07, "loss": 0.3947, "step": 12700 }, { "epoch": 0.9004608294930876, "grad_norm": 2.9753608400066516, "learning_rate": 2.576162007699268e-07, "loss": 0.3822, "step": 12701 }, { "epoch": 0.9005317263381779, "grad_norm": 4.007528962242834, "learning_rate": 2.572525354093408e-07, "loss": 0.3762, "step": 12702 }, { "epoch": 0.9006026231832683, "grad_norm": 85.72614411708635, "learning_rate": 2.568891201374496e-07, "loss": 0.4338, "step": 12703 }, { "epoch": 0.9006735200283588, "grad_norm": 2.1413266439085006, "learning_rate": 2.5652595497341505e-07, "loss": 0.3879, "step": 12704 }, { "epoch": 0.9007444168734491, "grad_norm": 2.7733063380030614, "learning_rate": 2.561630399363868e-07, "loss": 0.3675, "step": 12705 }, { "epoch": 0.9008153137185395, "grad_norm": 2.893855566557803, "learning_rate": 2.558003750455029e-07, "loss": 0.3441, "step": 12706 }, { "epoch": 0.9008862105636299, "grad_norm": 2.623534829910332, "learning_rate": 2.554379603198859e-07, "loss": 0.3742, "step": 12707 }, { "epoch": 0.9009571074087203, "grad_norm": 3.5530272808547716, "learning_rate": 2.550757957786465e-07, "loss": 0.3786, "step": 12708 }, { "epoch": 0.9010280042538107, "grad_norm": 3.0072191902515657, "learning_rate": 2.5471388144088174e-07, "loss": 0.3518, "step": 12709 }, { "epoch": 0.9010989010989011, "grad_norm": 2.958975898710063, "learning_rate": 2.5435221732567686e-07, "loss": 0.381, "step": 12710 }, { "epoch": 0.9011697979439915, "grad_norm": 2.8717636812786322, "learning_rate": 2.539908034521016e-07, "loss": 0.3996, "step": 12711 }, { "epoch": 0.9012406947890819, "grad_norm": 5.792390257083034, "learning_rate": 2.5362963983921297e-07, "loss": 0.3827, "step": 12712 }, { "epoch": 0.9013115916341723, "grad_norm": 1.9732860611249223, "learning_rate": 2.5326872650605684e-07, "loss": 0.374, "step": 12713 }, { "epoch": 0.9013824884792627, "grad_norm": 1.423058484830935, "learning_rate": 2.5290806347166407e-07, "loss": 0.6638, "step": 12714 }, { "epoch": 0.901453385324353, "grad_norm": 4.063115150409562, "learning_rate": 2.525476507550517e-07, "loss": 0.3684, "step": 12715 }, { "epoch": 0.9015242821694435, "grad_norm": 2.2813690305121015, "learning_rate": 2.5218748837522567e-07, "loss": 0.3349, "step": 12716 }, { "epoch": 0.9015951790145339, "grad_norm": 3.3365427194535084, "learning_rate": 2.5182757635117625e-07, "loss": 0.3926, "step": 12717 }, { "epoch": 0.9016660758596242, "grad_norm": 11.69264861524789, "learning_rate": 2.514679147018839e-07, "loss": 0.4254, "step": 12718 }, { "epoch": 0.9017369727047146, "grad_norm": 2.6006713720715844, "learning_rate": 2.511085034463123e-07, "loss": 0.4039, "step": 12719 }, { "epoch": 0.9018078695498051, "grad_norm": 1.9093876570285762, "learning_rate": 2.5074934260341467e-07, "loss": 0.4123, "step": 12720 }, { "epoch": 0.9018787663948954, "grad_norm": 1.8883837154548155, "learning_rate": 2.50390432192128e-07, "loss": 0.3339, "step": 12721 }, { "epoch": 0.9019496632399858, "grad_norm": 2.4567164074042744, "learning_rate": 2.5003177223138054e-07, "loss": 0.3836, "step": 12722 }, { "epoch": 0.9020205600850763, "grad_norm": 3.449040533078064, "learning_rate": 2.496733627400827e-07, "loss": 0.4098, "step": 12723 }, { "epoch": 0.9020914569301666, "grad_norm": 3.9759539204950047, "learning_rate": 2.4931520373713326e-07, "loss": 0.3412, "step": 12724 }, { "epoch": 0.902162353775257, "grad_norm": 2.665294667815948, "learning_rate": 2.4895729524141986e-07, "loss": 0.3784, "step": 12725 }, { "epoch": 0.9022332506203474, "grad_norm": 3.2397740485551574, "learning_rate": 2.4859963727181525e-07, "loss": 0.4057, "step": 12726 }, { "epoch": 0.9023041474654377, "grad_norm": 2.97432829050466, "learning_rate": 2.4824222984717704e-07, "loss": 0.3974, "step": 12727 }, { "epoch": 0.9023750443105282, "grad_norm": 1.931124226146261, "learning_rate": 2.4788507298635303e-07, "loss": 0.3861, "step": 12728 }, { "epoch": 0.9024459411556186, "grad_norm": 2.15880251117073, "learning_rate": 2.4752816670817694e-07, "loss": 0.3542, "step": 12729 }, { "epoch": 0.9025168380007089, "grad_norm": 2.5920289058840686, "learning_rate": 2.471715110314682e-07, "loss": 0.4007, "step": 12730 }, { "epoch": 0.9025877348457993, "grad_norm": 2.844734187653917, "learning_rate": 2.4681510597503234e-07, "loss": 0.3772, "step": 12731 }, { "epoch": 0.9026586316908898, "grad_norm": 2.725517441480941, "learning_rate": 2.464589515576643e-07, "loss": 0.3801, "step": 12732 }, { "epoch": 0.9027295285359801, "grad_norm": 2.156160622713012, "learning_rate": 2.461030477981452e-07, "loss": 0.3594, "step": 12733 }, { "epoch": 0.9028004253810705, "grad_norm": 4.421421951205446, "learning_rate": 2.4574739471523945e-07, "loss": 0.4016, "step": 12734 }, { "epoch": 0.902871322226161, "grad_norm": 2.918533099613172, "learning_rate": 2.4539199232770204e-07, "loss": 0.3964, "step": 12735 }, { "epoch": 0.9029422190712514, "grad_norm": 10.81827082892932, "learning_rate": 2.4503684065427525e-07, "loss": 0.4592, "step": 12736 }, { "epoch": 0.9030131159163417, "grad_norm": 0.9626721131106871, "learning_rate": 2.446819397136846e-07, "loss": 0.6068, "step": 12737 }, { "epoch": 0.9030840127614321, "grad_norm": 3.5137628008140163, "learning_rate": 2.4432728952464403e-07, "loss": 0.3703, "step": 12738 }, { "epoch": 0.9031549096065226, "grad_norm": 3.51126495522381, "learning_rate": 2.4397289010585634e-07, "loss": 0.4088, "step": 12739 }, { "epoch": 0.9032258064516129, "grad_norm": 2.9488933786447373, "learning_rate": 2.4361874147600826e-07, "loss": 0.4142, "step": 12740 }, { "epoch": 0.9032967032967033, "grad_norm": 2.3502969410938954, "learning_rate": 2.4326484365377377e-07, "loss": 0.3221, "step": 12741 }, { "epoch": 0.9033676001417937, "grad_norm": 9.861904714912585, "learning_rate": 2.4291119665781515e-07, "loss": 0.4012, "step": 12742 }, { "epoch": 0.903438496986884, "grad_norm": 3.1876092132035128, "learning_rate": 2.425578005067797e-07, "loss": 0.4149, "step": 12743 }, { "epoch": 0.9035093938319745, "grad_norm": 4.767469107586367, "learning_rate": 2.42204655219303e-07, "loss": 0.3986, "step": 12744 }, { "epoch": 0.9035802906770649, "grad_norm": 5.233958271914836, "learning_rate": 2.4185176081400585e-07, "loss": 0.4614, "step": 12745 }, { "epoch": 0.9036511875221552, "grad_norm": 3.9236362905866264, "learning_rate": 2.4149911730949717e-07, "loss": 0.4231, "step": 12746 }, { "epoch": 0.9037220843672457, "grad_norm": 1.7973295881839373, "learning_rate": 2.411467247243715e-07, "loss": 0.3539, "step": 12747 }, { "epoch": 0.9037929812123361, "grad_norm": 6.937179156898808, "learning_rate": 2.407945830772124e-07, "loss": 0.4392, "step": 12748 }, { "epoch": 0.9038638780574264, "grad_norm": 2.313218710023332, "learning_rate": 2.4044269238658713e-07, "loss": 0.4196, "step": 12749 }, { "epoch": 0.9039347749025168, "grad_norm": 1.7881229031220938, "learning_rate": 2.400910526710515e-07, "loss": 0.3713, "step": 12750 }, { "epoch": 0.9040056717476073, "grad_norm": 6.484825401761797, "learning_rate": 2.3973966394914785e-07, "loss": 0.3789, "step": 12751 }, { "epoch": 0.9040765685926976, "grad_norm": 1.6056213555046124, "learning_rate": 2.393885262394047e-07, "loss": 0.3822, "step": 12752 }, { "epoch": 0.904147465437788, "grad_norm": 3.4509907908570274, "learning_rate": 2.390376395603383e-07, "loss": 0.3712, "step": 12753 }, { "epoch": 0.9042183622828784, "grad_norm": 2.2611456146041324, "learning_rate": 2.386870039304506e-07, "loss": 0.3884, "step": 12754 }, { "epoch": 0.9042892591279688, "grad_norm": 2.4734111976954964, "learning_rate": 2.3833661936823182e-07, "loss": 0.3789, "step": 12755 }, { "epoch": 0.9043601559730592, "grad_norm": 2.6639251167169182, "learning_rate": 2.3798648589215766e-07, "loss": 0.3576, "step": 12756 }, { "epoch": 0.9044310528181496, "grad_norm": 2.3859220098581235, "learning_rate": 2.3763660352069062e-07, "loss": 0.3526, "step": 12757 }, { "epoch": 0.9045019496632399, "grad_norm": 2.3454214815048973, "learning_rate": 2.3728697227228036e-07, "loss": 0.3646, "step": 12758 }, { "epoch": 0.9045728465083304, "grad_norm": 2.218096490266667, "learning_rate": 2.3693759216536384e-07, "loss": 0.3622, "step": 12759 }, { "epoch": 0.9046437433534208, "grad_norm": 1.9969126972435414, "learning_rate": 2.3658846321836238e-07, "loss": 0.379, "step": 12760 }, { "epoch": 0.9047146401985111, "grad_norm": 6.04255703346345, "learning_rate": 2.362395854496874e-07, "loss": 0.4068, "step": 12761 }, { "epoch": 0.9047855370436015, "grad_norm": 3.6859366416214403, "learning_rate": 2.3589095887773584e-07, "loss": 0.382, "step": 12762 }, { "epoch": 0.904856433888692, "grad_norm": 3.3139795340550995, "learning_rate": 2.3554258352088966e-07, "loss": 0.4332, "step": 12763 }, { "epoch": 0.9049273307337824, "grad_norm": 2.4353561798515684, "learning_rate": 2.3519445939751971e-07, "loss": 0.4219, "step": 12764 }, { "epoch": 0.9049982275788727, "grad_norm": 13.821546180194119, "learning_rate": 2.3484658652598302e-07, "loss": 0.3624, "step": 12765 }, { "epoch": 0.9050691244239631, "grad_norm": 3.5169749636693073, "learning_rate": 2.344989649246221e-07, "loss": 0.3562, "step": 12766 }, { "epoch": 0.9051400212690536, "grad_norm": 3.477387751248373, "learning_rate": 2.3415159461176894e-07, "loss": 0.4068, "step": 12767 }, { "epoch": 0.9052109181141439, "grad_norm": 2.393046318533641, "learning_rate": 2.338044756057395e-07, "loss": 0.388, "step": 12768 }, { "epoch": 0.9052818149592343, "grad_norm": 3.410989435035459, "learning_rate": 2.3345760792483795e-07, "loss": 0.3974, "step": 12769 }, { "epoch": 0.9053527118043248, "grad_norm": 3.0670894901930326, "learning_rate": 2.3311099158735472e-07, "loss": 0.3781, "step": 12770 }, { "epoch": 0.9054236086494151, "grad_norm": 1.030806065459666, "learning_rate": 2.3276462661156795e-07, "loss": 0.585, "step": 12771 }, { "epoch": 0.9054945054945055, "grad_norm": 4.031856043694275, "learning_rate": 2.3241851301574025e-07, "loss": 0.3784, "step": 12772 }, { "epoch": 0.9055654023395959, "grad_norm": 2.5996059819483057, "learning_rate": 2.3207265081812315e-07, "loss": 0.3681, "step": 12773 }, { "epoch": 0.9056362991846862, "grad_norm": 4.71273047855319, "learning_rate": 2.3172704003695478e-07, "loss": 0.3341, "step": 12774 }, { "epoch": 0.9057071960297767, "grad_norm": 3.9739472927602923, "learning_rate": 2.3138168069045952e-07, "loss": 0.3817, "step": 12775 }, { "epoch": 0.9057780928748671, "grad_norm": 2.205132224888412, "learning_rate": 2.3103657279684721e-07, "loss": 0.3806, "step": 12776 }, { "epoch": 0.9058489897199574, "grad_norm": 5.897884843796471, "learning_rate": 2.3069171637431608e-07, "loss": 0.3776, "step": 12777 }, { "epoch": 0.9059198865650478, "grad_norm": 3.2101823640019815, "learning_rate": 2.303471114410527e-07, "loss": 0.3925, "step": 12778 }, { "epoch": 0.9059907834101383, "grad_norm": 2.1510631091522074, "learning_rate": 2.300027580152253e-07, "loss": 0.358, "step": 12779 }, { "epoch": 0.9060616802552286, "grad_norm": 7.105383198119988, "learning_rate": 2.2965865611499327e-07, "loss": 0.3748, "step": 12780 }, { "epoch": 0.906132577100319, "grad_norm": 2.279377379733364, "learning_rate": 2.293148057585015e-07, "loss": 0.3025, "step": 12781 }, { "epoch": 0.9062034739454095, "grad_norm": 2.8048315039656053, "learning_rate": 2.289712069638822e-07, "loss": 0.3557, "step": 12782 }, { "epoch": 0.9062743707904998, "grad_norm": 4.717656443693006, "learning_rate": 2.286278597492514e-07, "loss": 0.3862, "step": 12783 }, { "epoch": 0.9063452676355902, "grad_norm": 2.619789864479681, "learning_rate": 2.2828476413271627e-07, "loss": 0.3216, "step": 12784 }, { "epoch": 0.9064161644806806, "grad_norm": 3.0701399159601266, "learning_rate": 2.2794192013236684e-07, "loss": 0.3582, "step": 12785 }, { "epoch": 0.906487061325771, "grad_norm": 6.627898260923181, "learning_rate": 2.2759932776628301e-07, "loss": 0.4689, "step": 12786 }, { "epoch": 0.9065579581708614, "grad_norm": 1.8296148283470857, "learning_rate": 2.2725698705252874e-07, "loss": 0.3756, "step": 12787 }, { "epoch": 0.9066288550159518, "grad_norm": 2.155869207891634, "learning_rate": 2.2691489800915623e-07, "loss": 0.6599, "step": 12788 }, { "epoch": 0.9066997518610422, "grad_norm": 3.979216200609797, "learning_rate": 2.2657306065420492e-07, "loss": 0.3288, "step": 12789 }, { "epoch": 0.9067706487061326, "grad_norm": 2.4944055803116405, "learning_rate": 2.262314750056993e-07, "loss": 0.4025, "step": 12790 }, { "epoch": 0.906841545551223, "grad_norm": 2.1889135307791268, "learning_rate": 2.2589014108165165e-07, "loss": 0.3867, "step": 12791 }, { "epoch": 0.9069124423963134, "grad_norm": 2.9663010738813123, "learning_rate": 2.2554905890006028e-07, "loss": 0.3546, "step": 12792 }, { "epoch": 0.9069833392414037, "grad_norm": 7.326347502707644, "learning_rate": 2.25208228478912e-07, "loss": 0.3219, "step": 12793 }, { "epoch": 0.9070542360864942, "grad_norm": 3.0090062903687005, "learning_rate": 2.2486764983617848e-07, "loss": 0.4012, "step": 12794 }, { "epoch": 0.9071251329315846, "grad_norm": 2.0201746323021634, "learning_rate": 2.2452732298981707e-07, "loss": 0.3936, "step": 12795 }, { "epoch": 0.9071960297766749, "grad_norm": 3.106648709163757, "learning_rate": 2.2418724795777614e-07, "loss": 0.3767, "step": 12796 }, { "epoch": 0.9072669266217653, "grad_norm": 2.3298012343628005, "learning_rate": 2.2384742475798692e-07, "loss": 0.3579, "step": 12797 }, { "epoch": 0.9073378234668558, "grad_norm": 5.091367827942263, "learning_rate": 2.2350785340836733e-07, "loss": 0.4185, "step": 12798 }, { "epoch": 0.9074087203119461, "grad_norm": 2.019337550154336, "learning_rate": 2.231685339268247e-07, "loss": 0.3627, "step": 12799 }, { "epoch": 0.9074796171570365, "grad_norm": 2.925923937620432, "learning_rate": 2.2282946633125191e-07, "loss": 0.3996, "step": 12800 }, { "epoch": 0.9075505140021269, "grad_norm": 1.056294346487268, "learning_rate": 2.2249065063952746e-07, "loss": 0.6318, "step": 12801 }, { "epoch": 0.9076214108472173, "grad_norm": 5.71707726675591, "learning_rate": 2.2215208686951705e-07, "loss": 0.3507, "step": 12802 }, { "epoch": 0.9076923076923077, "grad_norm": 3.4005266570182493, "learning_rate": 2.218137750390742e-07, "loss": 0.3801, "step": 12803 }, { "epoch": 0.9077632045373981, "grad_norm": 2.9297406047379435, "learning_rate": 2.2147571516603848e-07, "loss": 0.4383, "step": 12804 }, { "epoch": 0.9078341013824884, "grad_norm": 2.8032739987144337, "learning_rate": 2.211379072682346e-07, "loss": 0.3629, "step": 12805 }, { "epoch": 0.9079049982275789, "grad_norm": 4.392486902562104, "learning_rate": 2.2080035136347655e-07, "loss": 0.4038, "step": 12806 }, { "epoch": 0.9079758950726693, "grad_norm": 11.253565241300503, "learning_rate": 2.2046304746956404e-07, "loss": 0.4137, "step": 12807 }, { "epoch": 0.9080467919177596, "grad_norm": 2.5492711222396314, "learning_rate": 2.201259956042834e-07, "loss": 0.4175, "step": 12808 }, { "epoch": 0.90811768876285, "grad_norm": 4.216845307808668, "learning_rate": 2.1978919578540702e-07, "loss": 0.3948, "step": 12809 }, { "epoch": 0.9081885856079405, "grad_norm": 2.67004464700993, "learning_rate": 2.1945264803069522e-07, "loss": 0.3785, "step": 12810 }, { "epoch": 0.9082594824530308, "grad_norm": 2.321411985750933, "learning_rate": 2.191163523578932e-07, "loss": 0.3902, "step": 12811 }, { "epoch": 0.9083303792981212, "grad_norm": 2.2030055557275956, "learning_rate": 2.1878030878473567e-07, "loss": 0.412, "step": 12812 }, { "epoch": 0.9084012761432116, "grad_norm": 5.323156436540164, "learning_rate": 2.1844451732894235e-07, "loss": 0.4042, "step": 12813 }, { "epoch": 0.9084721729883021, "grad_norm": 3.7733912690937212, "learning_rate": 2.1810897800821795e-07, "loss": 0.4083, "step": 12814 }, { "epoch": 0.9085430698333924, "grad_norm": 2.792446514728324, "learning_rate": 2.1777369084025834e-07, "loss": 0.4213, "step": 12815 }, { "epoch": 0.9086139666784828, "grad_norm": 1.6147578304797103, "learning_rate": 2.174386558427416e-07, "loss": 0.3508, "step": 12816 }, { "epoch": 0.9086848635235732, "grad_norm": 5.339187336908831, "learning_rate": 2.1710387303333413e-07, "loss": 0.4465, "step": 12817 }, { "epoch": 0.9087557603686636, "grad_norm": 3.3588956871401816, "learning_rate": 2.1676934242969073e-07, "loss": 0.3696, "step": 12818 }, { "epoch": 0.908826657213754, "grad_norm": 2.1109957904026464, "learning_rate": 2.164350640494506e-07, "loss": 0.4222, "step": 12819 }, { "epoch": 0.9088975540588444, "grad_norm": 3.3301469118769935, "learning_rate": 2.1610103791024128e-07, "loss": 0.4254, "step": 12820 }, { "epoch": 0.9089684509039347, "grad_norm": 4.076527780343795, "learning_rate": 2.1576726402967485e-07, "loss": 0.3773, "step": 12821 }, { "epoch": 0.9090393477490252, "grad_norm": 1.7448798793105942, "learning_rate": 2.154337424253522e-07, "loss": 0.4161, "step": 12822 }, { "epoch": 0.9091102445941156, "grad_norm": 7.120917824887397, "learning_rate": 2.151004731148615e-07, "loss": 0.3697, "step": 12823 }, { "epoch": 0.9091811414392059, "grad_norm": 3.181802983090881, "learning_rate": 2.1476745611577365e-07, "loss": 0.3651, "step": 12824 }, { "epoch": 0.9092520382842963, "grad_norm": 3.083597194610436, "learning_rate": 2.144346914456502e-07, "loss": 0.3994, "step": 12825 }, { "epoch": 0.9093229351293868, "grad_norm": 2.497087469495731, "learning_rate": 2.141021791220388e-07, "loss": 0.3915, "step": 12826 }, { "epoch": 0.9093938319744771, "grad_norm": 3.492227365161118, "learning_rate": 2.1376991916247202e-07, "loss": 0.3472, "step": 12827 }, { "epoch": 0.9094647288195675, "grad_norm": 4.349070561244417, "learning_rate": 2.1343791158447037e-07, "loss": 0.3757, "step": 12828 }, { "epoch": 0.909535625664658, "grad_norm": 2.7018794640483823, "learning_rate": 2.13106156405542e-07, "loss": 0.3695, "step": 12829 }, { "epoch": 0.9096065225097483, "grad_norm": 1.0568716083471437, "learning_rate": 2.1277465364317852e-07, "loss": 0.694, "step": 12830 }, { "epoch": 0.9096774193548387, "grad_norm": 2.025042743833343, "learning_rate": 2.1244340331486258e-07, "loss": 0.4112, "step": 12831 }, { "epoch": 0.9097483161999291, "grad_norm": 4.117093942715615, "learning_rate": 2.1211240543805966e-07, "loss": 0.4233, "step": 12832 }, { "epoch": 0.9098192130450194, "grad_norm": 2.4428634328905297, "learning_rate": 2.117816600302236e-07, "loss": 0.4219, "step": 12833 }, { "epoch": 0.9098901098901099, "grad_norm": 2.5274052639318576, "learning_rate": 2.1145116710879598e-07, "loss": 0.3614, "step": 12834 }, { "epoch": 0.9099610067352003, "grad_norm": 2.199612061366446, "learning_rate": 2.111209266912023e-07, "loss": 0.365, "step": 12835 }, { "epoch": 0.9100319035802906, "grad_norm": 6.367376040574198, "learning_rate": 2.107909387948581e-07, "loss": 0.3657, "step": 12836 }, { "epoch": 0.910102800425381, "grad_norm": 0.981461199830449, "learning_rate": 2.104612034371628e-07, "loss": 0.6567, "step": 12837 }, { "epoch": 0.9101736972704715, "grad_norm": 2.842042682457716, "learning_rate": 2.101317206355047e-07, "loss": 0.375, "step": 12838 }, { "epoch": 0.9102445941155619, "grad_norm": 2.386422031564866, "learning_rate": 2.0980249040725653e-07, "loss": 0.3923, "step": 12839 }, { "epoch": 0.9103154909606522, "grad_norm": 4.134435840398535, "learning_rate": 2.0947351276977834e-07, "loss": 0.356, "step": 12840 }, { "epoch": 0.9103863878057427, "grad_norm": 2.9553536359931716, "learning_rate": 2.09144787740419e-07, "loss": 0.4054, "step": 12841 }, { "epoch": 0.9104572846508331, "grad_norm": 2.0542348883719876, "learning_rate": 2.0881631533651247e-07, "loss": 0.3769, "step": 12842 }, { "epoch": 0.9105281814959234, "grad_norm": 1.9011792292022698, "learning_rate": 2.0848809557537707e-07, "loss": 0.3537, "step": 12843 }, { "epoch": 0.9105990783410138, "grad_norm": 2.4404132505943132, "learning_rate": 2.0816012847432232e-07, "loss": 0.3369, "step": 12844 }, { "epoch": 0.9106699751861043, "grad_norm": 3.855064427418571, "learning_rate": 2.0783241405064158e-07, "loss": 0.3903, "step": 12845 }, { "epoch": 0.9107408720311946, "grad_norm": 3.6893598808268346, "learning_rate": 2.075049523216155e-07, "loss": 0.4068, "step": 12846 }, { "epoch": 0.910811768876285, "grad_norm": 6.796153370188289, "learning_rate": 2.0717774330451025e-07, "loss": 0.3682, "step": 12847 }, { "epoch": 0.9108826657213754, "grad_norm": 2.4746811003032727, "learning_rate": 2.0685078701658202e-07, "loss": 0.3408, "step": 12848 }, { "epoch": 0.9109535625664658, "grad_norm": 2.0716896683067363, "learning_rate": 2.065240834750698e-07, "loss": 0.403, "step": 12849 }, { "epoch": 0.9110244594115562, "grad_norm": 3.9281743597404306, "learning_rate": 2.0619763269720094e-07, "loss": 0.417, "step": 12850 }, { "epoch": 0.9110953562566466, "grad_norm": 2.470173054151637, "learning_rate": 2.0587143470018945e-07, "loss": 0.4073, "step": 12851 }, { "epoch": 0.9111662531017369, "grad_norm": 7.0154596804848355, "learning_rate": 2.0554548950123765e-07, "loss": 0.4117, "step": 12852 }, { "epoch": 0.9112371499468274, "grad_norm": 2.124234293420688, "learning_rate": 2.0521979711753126e-07, "loss": 0.366, "step": 12853 }, { "epoch": 0.9113080467919178, "grad_norm": 3.1884969060685187, "learning_rate": 2.0489435756624375e-07, "loss": 0.4114, "step": 12854 }, { "epoch": 0.9113789436370081, "grad_norm": 1.0010436702969172, "learning_rate": 2.0456917086453754e-07, "loss": 0.6196, "step": 12855 }, { "epoch": 0.9114498404820985, "grad_norm": 4.506695618246778, "learning_rate": 2.0424423702955885e-07, "loss": 0.3895, "step": 12856 }, { "epoch": 0.911520737327189, "grad_norm": 4.124156636433509, "learning_rate": 2.039195560784424e-07, "loss": 0.4343, "step": 12857 }, { "epoch": 0.9115916341722793, "grad_norm": 2.3631706585692873, "learning_rate": 2.0359512802830827e-07, "loss": 0.387, "step": 12858 }, { "epoch": 0.9116625310173697, "grad_norm": 3.6024954942439225, "learning_rate": 2.032709528962634e-07, "loss": 0.3874, "step": 12859 }, { "epoch": 0.9117334278624601, "grad_norm": 5.649863436879278, "learning_rate": 2.0294703069940358e-07, "loss": 0.4057, "step": 12860 }, { "epoch": 0.9118043247075505, "grad_norm": 2.18315120063768, "learning_rate": 2.026233614548073e-07, "loss": 0.3601, "step": 12861 }, { "epoch": 0.9118752215526409, "grad_norm": 2.789414181835786, "learning_rate": 2.0229994517954265e-07, "loss": 0.4037, "step": 12862 }, { "epoch": 0.9119461183977313, "grad_norm": 4.663824811861674, "learning_rate": 2.0197678189066429e-07, "loss": 0.3659, "step": 12863 }, { "epoch": 0.9120170152428217, "grad_norm": 2.131955333811728, "learning_rate": 2.0165387160521245e-07, "loss": 0.343, "step": 12864 }, { "epoch": 0.9120879120879121, "grad_norm": 0.9913429587578941, "learning_rate": 2.0133121434021407e-07, "loss": 0.6384, "step": 12865 }, { "epoch": 0.9121588089330025, "grad_norm": 1.9179824509866887, "learning_rate": 2.0100881011268336e-07, "loss": 0.3499, "step": 12866 }, { "epoch": 0.9122297057780929, "grad_norm": 3.3586987708080245, "learning_rate": 2.0068665893962115e-07, "loss": 0.3329, "step": 12867 }, { "epoch": 0.9123006026231832, "grad_norm": 3.571620502979732, "learning_rate": 2.0036476083801547e-07, "loss": 0.3938, "step": 12868 }, { "epoch": 0.9123714994682737, "grad_norm": 4.143919443206629, "learning_rate": 2.0004311582483783e-07, "loss": 0.3551, "step": 12869 }, { "epoch": 0.9124423963133641, "grad_norm": 3.572684060963753, "learning_rate": 1.997217239170507e-07, "loss": 0.3646, "step": 12870 }, { "epoch": 0.9125132931584544, "grad_norm": 6.290055897260897, "learning_rate": 1.9940058513160166e-07, "loss": 0.3733, "step": 12871 }, { "epoch": 0.9125841900035448, "grad_norm": 2.5628562151320717, "learning_rate": 1.9907969948542384e-07, "loss": 0.4059, "step": 12872 }, { "epoch": 0.9126550868486353, "grad_norm": 5.068511870705152, "learning_rate": 1.98759066995437e-07, "loss": 0.3494, "step": 12873 }, { "epoch": 0.9127259836937256, "grad_norm": 2.715448889304648, "learning_rate": 1.984386876785499e-07, "loss": 0.374, "step": 12874 }, { "epoch": 0.912796880538816, "grad_norm": 3.509746634069484, "learning_rate": 1.9811856155165509e-07, "loss": 0.3883, "step": 12875 }, { "epoch": 0.9128677773839065, "grad_norm": 2.5786549021777723, "learning_rate": 1.977986886316341e-07, "loss": 0.4116, "step": 12876 }, { "epoch": 0.9129386742289968, "grad_norm": 2.392214700931099, "learning_rate": 1.9747906893535395e-07, "loss": 0.3761, "step": 12877 }, { "epoch": 0.9130095710740872, "grad_norm": 2.5926446995860744, "learning_rate": 1.9715970247966677e-07, "loss": 0.3578, "step": 12878 }, { "epoch": 0.9130804679191776, "grad_norm": 46.57635528425906, "learning_rate": 1.9684058928141569e-07, "loss": 0.3709, "step": 12879 }, { "epoch": 0.913151364764268, "grad_norm": 1.7773119418616403, "learning_rate": 1.9652172935742563e-07, "loss": 0.3714, "step": 12880 }, { "epoch": 0.9132222616093584, "grad_norm": 4.71160389788485, "learning_rate": 1.9620312272451148e-07, "loss": 0.3874, "step": 12881 }, { "epoch": 0.9132931584544488, "grad_norm": 3.333296986959628, "learning_rate": 1.958847693994731e-07, "loss": 0.4197, "step": 12882 }, { "epoch": 0.9133640552995391, "grad_norm": 4.543352383117036, "learning_rate": 1.9556666939909764e-07, "loss": 0.4091, "step": 12883 }, { "epoch": 0.9134349521446296, "grad_norm": 3.0134965209982187, "learning_rate": 1.9524882274015888e-07, "loss": 0.3711, "step": 12884 }, { "epoch": 0.91350584898972, "grad_norm": 2.5359575108191645, "learning_rate": 1.949312294394168e-07, "loss": 0.3673, "step": 12885 }, { "epoch": 0.9135767458348103, "grad_norm": 6.354333869403629, "learning_rate": 1.9461388951361848e-07, "loss": 0.409, "step": 12886 }, { "epoch": 0.9136476426799007, "grad_norm": 10.048148040003344, "learning_rate": 1.9429680297949837e-07, "loss": 0.3466, "step": 12887 }, { "epoch": 0.9137185395249912, "grad_norm": 6.974936000739779, "learning_rate": 1.9397996985377476e-07, "loss": 0.3683, "step": 12888 }, { "epoch": 0.9137894363700815, "grad_norm": 4.547158709558648, "learning_rate": 1.9366339015315594e-07, "loss": 0.4327, "step": 12889 }, { "epoch": 0.9138603332151719, "grad_norm": 2.270519006726749, "learning_rate": 1.9334706389433522e-07, "loss": 0.3676, "step": 12890 }, { "epoch": 0.9139312300602623, "grad_norm": 2.900143734469424, "learning_rate": 1.9303099109399315e-07, "loss": 0.3431, "step": 12891 }, { "epoch": 0.9140021269053528, "grad_norm": 3.144555405314902, "learning_rate": 1.9271517176879474e-07, "loss": 0.4012, "step": 12892 }, { "epoch": 0.9140730237504431, "grad_norm": 5.4471689898246645, "learning_rate": 1.9239960593539497e-07, "loss": 0.4116, "step": 12893 }, { "epoch": 0.9141439205955335, "grad_norm": 2.0789709276648147, "learning_rate": 1.920842936104339e-07, "loss": 0.3843, "step": 12894 }, { "epoch": 0.9142148174406239, "grad_norm": 3.7099679431356436, "learning_rate": 1.9176923481053766e-07, "loss": 0.4124, "step": 12895 }, { "epoch": 0.9142857142857143, "grad_norm": 8.356739359787987, "learning_rate": 1.9145442955231907e-07, "loss": 0.4165, "step": 12896 }, { "epoch": 0.9143566111308047, "grad_norm": 6.849696092632819, "learning_rate": 1.9113987785237932e-07, "loss": 0.3677, "step": 12897 }, { "epoch": 0.9144275079758951, "grad_norm": 1.9914823203415344, "learning_rate": 1.9082557972730397e-07, "loss": 0.3763, "step": 12898 }, { "epoch": 0.9144984048209854, "grad_norm": 3.718335769782334, "learning_rate": 1.9051153519366594e-07, "loss": 0.384, "step": 12899 }, { "epoch": 0.9145693016660759, "grad_norm": 2.649608749143749, "learning_rate": 1.901977442680264e-07, "loss": 0.3514, "step": 12900 }, { "epoch": 0.9146401985111663, "grad_norm": 3.994844535393166, "learning_rate": 1.8988420696693045e-07, "loss": 0.3875, "step": 12901 }, { "epoch": 0.9147110953562566, "grad_norm": 2.5937518643506947, "learning_rate": 1.8957092330691207e-07, "loss": 0.3437, "step": 12902 }, { "epoch": 0.914781992201347, "grad_norm": 4.3063965695408974, "learning_rate": 1.892578933044903e-07, "loss": 0.4523, "step": 12903 }, { "epoch": 0.9148528890464375, "grad_norm": 4.536477556745734, "learning_rate": 1.8894511697617135e-07, "loss": 0.3615, "step": 12904 }, { "epoch": 0.9149237858915278, "grad_norm": 3.0617657526829403, "learning_rate": 1.8863259433844927e-07, "loss": 0.3429, "step": 12905 }, { "epoch": 0.9149946827366182, "grad_norm": 3.6561244499975736, "learning_rate": 1.8832032540780255e-07, "loss": 0.3726, "step": 12906 }, { "epoch": 0.9150655795817086, "grad_norm": 3.0834793987237155, "learning_rate": 1.8800831020069687e-07, "loss": 0.3521, "step": 12907 }, { "epoch": 0.915136476426799, "grad_norm": 2.8209888537259786, "learning_rate": 1.8769654873358635e-07, "loss": 0.352, "step": 12908 }, { "epoch": 0.9152073732718894, "grad_norm": 5.457292348530053, "learning_rate": 1.8738504102291e-07, "loss": 0.3749, "step": 12909 }, { "epoch": 0.9152782701169798, "grad_norm": 3.0258588650169616, "learning_rate": 1.8707378708509417e-07, "loss": 0.3679, "step": 12910 }, { "epoch": 0.9153491669620701, "grad_norm": 2.102350754646756, "learning_rate": 1.8676278693655015e-07, "loss": 0.4121, "step": 12911 }, { "epoch": 0.9154200638071606, "grad_norm": 2.095865315997929, "learning_rate": 1.8645204059367816e-07, "loss": 0.4043, "step": 12912 }, { "epoch": 0.915490960652251, "grad_norm": 2.096184146596724, "learning_rate": 1.8614154807286567e-07, "loss": 0.3806, "step": 12913 }, { "epoch": 0.9155618574973413, "grad_norm": 3.489125581871821, "learning_rate": 1.8583130939048177e-07, "loss": 0.3981, "step": 12914 }, { "epoch": 0.9156327543424317, "grad_norm": 4.482944988751022, "learning_rate": 1.8552132456288785e-07, "loss": 0.4136, "step": 12915 }, { "epoch": 0.9157036511875222, "grad_norm": 4.461287860590642, "learning_rate": 1.852115936064297e-07, "loss": 0.4096, "step": 12916 }, { "epoch": 0.9157745480326126, "grad_norm": 2.8471880052628245, "learning_rate": 1.8490211653743928e-07, "loss": 0.4282, "step": 12917 }, { "epoch": 0.9158454448777029, "grad_norm": 3.860989192092635, "learning_rate": 1.8459289337223464e-07, "loss": 0.369, "step": 12918 }, { "epoch": 0.9159163417227933, "grad_norm": 2.3350256954445032, "learning_rate": 1.8428392412712326e-07, "loss": 0.3811, "step": 12919 }, { "epoch": 0.9159872385678838, "grad_norm": 2.9361888697579066, "learning_rate": 1.8397520881839491e-07, "loss": 0.3903, "step": 12920 }, { "epoch": 0.9160581354129741, "grad_norm": 9.03540874837153, "learning_rate": 1.8366674746233105e-07, "loss": 0.3845, "step": 12921 }, { "epoch": 0.9161290322580645, "grad_norm": 1.9972718061312, "learning_rate": 1.8335854007519526e-07, "loss": 0.3483, "step": 12922 }, { "epoch": 0.916199929103155, "grad_norm": 5.1838511360447725, "learning_rate": 1.830505866732396e-07, "loss": 0.353, "step": 12923 }, { "epoch": 0.9162708259482453, "grad_norm": 4.119247876823417, "learning_rate": 1.827428872727033e-07, "loss": 0.3679, "step": 12924 }, { "epoch": 0.9163417227933357, "grad_norm": 14.816154652755008, "learning_rate": 1.824354418898111e-07, "loss": 0.4221, "step": 12925 }, { "epoch": 0.9164126196384261, "grad_norm": 8.996543142944128, "learning_rate": 1.8212825054077565e-07, "loss": 0.3552, "step": 12926 }, { "epoch": 0.9164835164835164, "grad_norm": 6.068968887796016, "learning_rate": 1.8182131324179452e-07, "loss": 0.3535, "step": 12927 }, { "epoch": 0.9165544133286069, "grad_norm": 2.67187508362437, "learning_rate": 1.8151463000905367e-07, "loss": 0.384, "step": 12928 }, { "epoch": 0.9166253101736973, "grad_norm": 2.4804992841211946, "learning_rate": 1.8120820085872404e-07, "loss": 0.3373, "step": 12929 }, { "epoch": 0.9166962070187876, "grad_norm": 2.606863368667774, "learning_rate": 1.8090202580696326e-07, "loss": 0.3579, "step": 12930 }, { "epoch": 0.916767103863878, "grad_norm": 2.877458290140064, "learning_rate": 1.805961048699173e-07, "loss": 0.3801, "step": 12931 }, { "epoch": 0.9168380007089685, "grad_norm": 3.965233106033843, "learning_rate": 1.8029043806371827e-07, "loss": 0.3691, "step": 12932 }, { "epoch": 0.9169088975540588, "grad_norm": 3.606940936356862, "learning_rate": 1.799850254044816e-07, "loss": 0.3554, "step": 12933 }, { "epoch": 0.9169797943991492, "grad_norm": 5.369032002354081, "learning_rate": 1.7967986690831384e-07, "loss": 0.4009, "step": 12934 }, { "epoch": 0.9170506912442397, "grad_norm": 3.0362835140414246, "learning_rate": 1.793749625913066e-07, "loss": 0.3608, "step": 12935 }, { "epoch": 0.91712158808933, "grad_norm": 4.267345990478246, "learning_rate": 1.7907031246953643e-07, "loss": 0.3648, "step": 12936 }, { "epoch": 0.9171924849344204, "grad_norm": 1.9756023492482329, "learning_rate": 1.7876591655906828e-07, "loss": 0.3765, "step": 12937 }, { "epoch": 0.9172633817795108, "grad_norm": 3.9515411025798173, "learning_rate": 1.7846177487595317e-07, "loss": 0.3359, "step": 12938 }, { "epoch": 0.9173342786246012, "grad_norm": 3.1330431533043166, "learning_rate": 1.7815788743622886e-07, "loss": 0.3997, "step": 12939 }, { "epoch": 0.9174051754696916, "grad_norm": 3.373787147037398, "learning_rate": 1.7785425425592028e-07, "loss": 0.3755, "step": 12940 }, { "epoch": 0.917476072314782, "grad_norm": 2.837757010813316, "learning_rate": 1.775508753510363e-07, "loss": 0.39, "step": 12941 }, { "epoch": 0.9175469691598724, "grad_norm": 3.3323423565023567, "learning_rate": 1.7724775073757638e-07, "loss": 0.3862, "step": 12942 }, { "epoch": 0.9176178660049628, "grad_norm": 4.354381135241593, "learning_rate": 1.7694488043152325e-07, "loss": 0.3893, "step": 12943 }, { "epoch": 0.9176887628500532, "grad_norm": 2.798174398963302, "learning_rate": 1.7664226444884748e-07, "loss": 0.4187, "step": 12944 }, { "epoch": 0.9177596596951436, "grad_norm": 2.085650498959174, "learning_rate": 1.7633990280550683e-07, "loss": 0.4026, "step": 12945 }, { "epoch": 0.9178305565402339, "grad_norm": 3.5186520144758404, "learning_rate": 1.760377955174447e-07, "loss": 0.3757, "step": 12946 }, { "epoch": 0.9179014533853244, "grad_norm": 2.523770895765658, "learning_rate": 1.7573594260059224e-07, "loss": 0.4176, "step": 12947 }, { "epoch": 0.9179723502304148, "grad_norm": 3.867403112335072, "learning_rate": 1.7543434407086557e-07, "loss": 0.3661, "step": 12948 }, { "epoch": 0.9180432470755051, "grad_norm": 3.4305145160429786, "learning_rate": 1.7513299994416754e-07, "loss": 0.4276, "step": 12949 }, { "epoch": 0.9181141439205955, "grad_norm": 2.4298535957716703, "learning_rate": 1.7483191023638935e-07, "loss": 0.364, "step": 12950 }, { "epoch": 0.918185040765686, "grad_norm": 2.768709778048575, "learning_rate": 1.7453107496340825e-07, "loss": 0.4039, "step": 12951 }, { "epoch": 0.9182559376107763, "grad_norm": 2.7013423579632847, "learning_rate": 1.7423049414108605e-07, "loss": 0.3923, "step": 12952 }, { "epoch": 0.9183268344558667, "grad_norm": 3.3860053000096397, "learning_rate": 1.7393016778527283e-07, "loss": 0.3988, "step": 12953 }, { "epoch": 0.9183977313009571, "grad_norm": 4.9120701239262194, "learning_rate": 1.7363009591180646e-07, "loss": 0.3759, "step": 12954 }, { "epoch": 0.9184686281460475, "grad_norm": 2.0823841344897374, "learning_rate": 1.733302785365082e-07, "loss": 0.3535, "step": 12955 }, { "epoch": 0.9185395249911379, "grad_norm": 18.37156909534258, "learning_rate": 1.7303071567518815e-07, "loss": 0.3904, "step": 12956 }, { "epoch": 0.9186104218362283, "grad_norm": 1.8607952622434216, "learning_rate": 1.7273140734364313e-07, "loss": 0.4174, "step": 12957 }, { "epoch": 0.9186813186813186, "grad_norm": 10.460352432105593, "learning_rate": 1.7243235355765608e-07, "loss": 0.3982, "step": 12958 }, { "epoch": 0.9187522155264091, "grad_norm": 1.9929865691135285, "learning_rate": 1.7213355433299496e-07, "loss": 0.422, "step": 12959 }, { "epoch": 0.9188231123714995, "grad_norm": 2.5721464786312658, "learning_rate": 1.7183500968541656e-07, "loss": 0.3558, "step": 12960 }, { "epoch": 0.9188940092165898, "grad_norm": 2.7537355275501723, "learning_rate": 1.7153671963066333e-07, "loss": 0.3615, "step": 12961 }, { "epoch": 0.9189649060616802, "grad_norm": 2.4594578311685473, "learning_rate": 1.7123868418446433e-07, "loss": 0.3541, "step": 12962 }, { "epoch": 0.9190358029067707, "grad_norm": 2.7840228019177737, "learning_rate": 1.7094090336253477e-07, "loss": 0.4251, "step": 12963 }, { "epoch": 0.919106699751861, "grad_norm": 1.962805092660941, "learning_rate": 1.706433771805782e-07, "loss": 0.3744, "step": 12964 }, { "epoch": 0.9191775965969514, "grad_norm": 10.455939287463915, "learning_rate": 1.70346105654281e-07, "loss": 0.362, "step": 12965 }, { "epoch": 0.9192484934420418, "grad_norm": 4.6741404754224245, "learning_rate": 1.7004908879932113e-07, "loss": 0.462, "step": 12966 }, { "epoch": 0.9193193902871323, "grad_norm": 4.169749667355418, "learning_rate": 1.6975232663135887e-07, "loss": 0.4436, "step": 12967 }, { "epoch": 0.9193902871322226, "grad_norm": 3.9089240710861577, "learning_rate": 1.6945581916604338e-07, "loss": 0.402, "step": 12968 }, { "epoch": 0.919461183977313, "grad_norm": 3.8929474915120066, "learning_rate": 1.691595664190099e-07, "loss": 0.3727, "step": 12969 }, { "epoch": 0.9195320808224035, "grad_norm": 2.9618920415056005, "learning_rate": 1.688635684058787e-07, "loss": 0.425, "step": 12970 }, { "epoch": 0.9196029776674938, "grad_norm": 3.4482729096304263, "learning_rate": 1.6856782514225956e-07, "loss": 0.419, "step": 12971 }, { "epoch": 0.9196738745125842, "grad_norm": 2.9367084988887786, "learning_rate": 1.6827233664374664e-07, "loss": 0.4275, "step": 12972 }, { "epoch": 0.9197447713576746, "grad_norm": 3.5157036033243676, "learning_rate": 1.679771029259214e-07, "loss": 0.3641, "step": 12973 }, { "epoch": 0.919815668202765, "grad_norm": 2.136540857811727, "learning_rate": 1.676821240043519e-07, "loss": 0.3981, "step": 12974 }, { "epoch": 0.9198865650478554, "grad_norm": 3.628566178199402, "learning_rate": 1.6738739989459129e-07, "loss": 0.4048, "step": 12975 }, { "epoch": 0.9199574618929458, "grad_norm": 1.9698191542811896, "learning_rate": 1.6709293061218211e-07, "loss": 0.4024, "step": 12976 }, { "epoch": 0.9200283587380361, "grad_norm": 4.79613842287605, "learning_rate": 1.6679871617265254e-07, "loss": 0.3993, "step": 12977 }, { "epoch": 0.9200992555831266, "grad_norm": 2.3067607150546325, "learning_rate": 1.6650475659151454e-07, "loss": 0.4015, "step": 12978 }, { "epoch": 0.920170152428217, "grad_norm": 5.25082935825151, "learning_rate": 1.6621105188426966e-07, "loss": 0.3737, "step": 12979 }, { "epoch": 0.9202410492733073, "grad_norm": 2.989218354597603, "learning_rate": 1.6591760206640662e-07, "loss": 0.4153, "step": 12980 }, { "epoch": 0.9203119461183977, "grad_norm": 3.8640101074796087, "learning_rate": 1.6562440715339745e-07, "loss": 0.3534, "step": 12981 }, { "epoch": 0.9203828429634882, "grad_norm": 2.014685507257128, "learning_rate": 1.6533146716070315e-07, "loss": 0.3594, "step": 12982 }, { "epoch": 0.9204537398085785, "grad_norm": 4.238538541196625, "learning_rate": 1.6503878210377078e-07, "loss": 0.3714, "step": 12983 }, { "epoch": 0.9205246366536689, "grad_norm": 2.266576657445239, "learning_rate": 1.6474635199803413e-07, "loss": 0.3472, "step": 12984 }, { "epoch": 0.9205955334987593, "grad_norm": 7.770486600107294, "learning_rate": 1.644541768589125e-07, "loss": 0.3383, "step": 12985 }, { "epoch": 0.9206664303438497, "grad_norm": 2.252500636539358, "learning_rate": 1.6416225670181308e-07, "loss": 0.3812, "step": 12986 }, { "epoch": 0.9207373271889401, "grad_norm": 2.7472724398708697, "learning_rate": 1.6387059154212904e-07, "loss": 0.332, "step": 12987 }, { "epoch": 0.9208082240340305, "grad_norm": 6.290884692300387, "learning_rate": 1.6357918139524032e-07, "loss": 0.3585, "step": 12988 }, { "epoch": 0.9208791208791208, "grad_norm": 3.315173014099143, "learning_rate": 1.6328802627651186e-07, "loss": 0.3694, "step": 12989 }, { "epoch": 0.9209500177242113, "grad_norm": 10.582611683192397, "learning_rate": 1.6299712620129804e-07, "loss": 0.375, "step": 12990 }, { "epoch": 0.9210209145693017, "grad_norm": 4.501443336032129, "learning_rate": 1.627064811849377e-07, "loss": 0.4336, "step": 12991 }, { "epoch": 0.9210918114143921, "grad_norm": 7.371771049508445, "learning_rate": 1.6241609124275692e-07, "loss": 0.4171, "step": 12992 }, { "epoch": 0.9211627082594824, "grad_norm": 1.7457536594293177, "learning_rate": 1.621259563900679e-07, "loss": 0.3622, "step": 12993 }, { "epoch": 0.9212336051045729, "grad_norm": 2.5140959263904215, "learning_rate": 1.6183607664216893e-07, "loss": 0.3984, "step": 12994 }, { "epoch": 0.9213045019496633, "grad_norm": 2.6189362762185175, "learning_rate": 1.615464520143467e-07, "loss": 0.3052, "step": 12995 }, { "epoch": 0.9213753987947536, "grad_norm": 4.286334354029299, "learning_rate": 1.61257082521874e-07, "loss": 0.419, "step": 12996 }, { "epoch": 0.921446295639844, "grad_norm": 6.651294568308483, "learning_rate": 1.609679681800086e-07, "loss": 0.3968, "step": 12997 }, { "epoch": 0.9215171924849345, "grad_norm": 3.1879364854406735, "learning_rate": 1.6067910900399443e-07, "loss": 0.3877, "step": 12998 }, { "epoch": 0.9215880893300248, "grad_norm": 3.616755956320459, "learning_rate": 1.6039050500906595e-07, "loss": 0.4024, "step": 12999 }, { "epoch": 0.9216589861751152, "grad_norm": 3.64776261162644, "learning_rate": 1.6010215621043935e-07, "loss": 0.3305, "step": 13000 }, { "epoch": 0.9217298830202056, "grad_norm": 5.811765988499836, "learning_rate": 1.5981406262331968e-07, "loss": 0.377, "step": 13001 }, { "epoch": 0.921800779865296, "grad_norm": 2.572280682366605, "learning_rate": 1.5952622426289921e-07, "loss": 0.3345, "step": 13002 }, { "epoch": 0.9218716767103864, "grad_norm": 6.853982522223061, "learning_rate": 1.5923864114435584e-07, "loss": 0.3818, "step": 13003 }, { "epoch": 0.9219425735554768, "grad_norm": 2.961810905262125, "learning_rate": 1.5895131328285408e-07, "loss": 0.3969, "step": 13004 }, { "epoch": 0.9220134704005671, "grad_norm": 2.464864737416441, "learning_rate": 1.5866424069354348e-07, "loss": 0.3707, "step": 13005 }, { "epoch": 0.9220843672456576, "grad_norm": 1.0917541377314026, "learning_rate": 1.583774233915636e-07, "loss": 0.6736, "step": 13006 }, { "epoch": 0.922155264090748, "grad_norm": 2.7719599986991543, "learning_rate": 1.5809086139203732e-07, "loss": 0.3716, "step": 13007 }, { "epoch": 0.9222261609358383, "grad_norm": 3.2618600339999837, "learning_rate": 1.578045547100754e-07, "loss": 0.3904, "step": 13008 }, { "epoch": 0.9222970577809287, "grad_norm": 5.471065822685341, "learning_rate": 1.5751850336077568e-07, "loss": 0.3882, "step": 13009 }, { "epoch": 0.9223679546260192, "grad_norm": 5.342402896756196, "learning_rate": 1.5723270735922057e-07, "loss": 0.3897, "step": 13010 }, { "epoch": 0.9224388514711095, "grad_norm": 2.8140276495213086, "learning_rate": 1.5694716672048194e-07, "loss": 0.3811, "step": 13011 }, { "epoch": 0.9225097483161999, "grad_norm": 2.962525064065388, "learning_rate": 1.5666188145961492e-07, "loss": 0.4317, "step": 13012 }, { "epoch": 0.9225806451612903, "grad_norm": 3.2883286962110656, "learning_rate": 1.5637685159166415e-07, "loss": 0.3528, "step": 13013 }, { "epoch": 0.9226515420063807, "grad_norm": 2.7600903910660333, "learning_rate": 1.5609207713165931e-07, "loss": 0.3427, "step": 13014 }, { "epoch": 0.9227224388514711, "grad_norm": 2.165338893023302, "learning_rate": 1.5580755809461556e-07, "loss": 0.3918, "step": 13015 }, { "epoch": 0.9227933356965615, "grad_norm": 2.146266602122227, "learning_rate": 1.555232944955376e-07, "loss": 0.3935, "step": 13016 }, { "epoch": 0.9228642325416518, "grad_norm": 5.4875182595299, "learning_rate": 1.5523928634941288e-07, "loss": 0.4179, "step": 13017 }, { "epoch": 0.9229351293867423, "grad_norm": 4.209537133812101, "learning_rate": 1.549555336712194e-07, "loss": 0.396, "step": 13018 }, { "epoch": 0.9230060262318327, "grad_norm": 3.6021864425511914, "learning_rate": 1.5467203647591856e-07, "loss": 0.4022, "step": 13019 }, { "epoch": 0.9230769230769231, "grad_norm": 2.4639860023395674, "learning_rate": 1.5438879477845892e-07, "loss": 0.3242, "step": 13020 }, { "epoch": 0.9231478199220134, "grad_norm": 3.1164515654391214, "learning_rate": 1.5410580859377633e-07, "loss": 0.3551, "step": 13021 }, { "epoch": 0.9232187167671039, "grad_norm": 4.936008430854362, "learning_rate": 1.5382307793679496e-07, "loss": 0.4022, "step": 13022 }, { "epoch": 0.9232896136121943, "grad_norm": 2.2867493747591534, "learning_rate": 1.5354060282242013e-07, "loss": 0.4052, "step": 13023 }, { "epoch": 0.9233605104572846, "grad_norm": 2.3630760822967516, "learning_rate": 1.532583832655482e-07, "loss": 0.3746, "step": 13024 }, { "epoch": 0.923431407302375, "grad_norm": 2.4461189273792057, "learning_rate": 1.5297641928106178e-07, "loss": 0.3674, "step": 13025 }, { "epoch": 0.9235023041474655, "grad_norm": 1.9706140122875688, "learning_rate": 1.5269471088382836e-07, "loss": 0.4043, "step": 13026 }, { "epoch": 0.9235732009925558, "grad_norm": 4.788021071886885, "learning_rate": 1.5241325808870223e-07, "loss": 0.4272, "step": 13027 }, { "epoch": 0.9236440978376462, "grad_norm": 2.8812214103794513, "learning_rate": 1.521320609105248e-07, "loss": 0.3455, "step": 13028 }, { "epoch": 0.9237149946827367, "grad_norm": 2.8327992448920516, "learning_rate": 1.5185111936412478e-07, "loss": 0.3595, "step": 13029 }, { "epoch": 0.923785891527827, "grad_norm": 11.527150839353279, "learning_rate": 1.5157043346431533e-07, "loss": 0.3794, "step": 13030 }, { "epoch": 0.9238567883729174, "grad_norm": 6.54537765698812, "learning_rate": 1.5129000322589737e-07, "loss": 0.3423, "step": 13031 }, { "epoch": 0.9239276852180078, "grad_norm": 7.737650731059887, "learning_rate": 1.5100982866365853e-07, "loss": 0.4205, "step": 13032 }, { "epoch": 0.9239985820630982, "grad_norm": 3.186048795655923, "learning_rate": 1.5072990979237313e-07, "loss": 0.43, "step": 13033 }, { "epoch": 0.9240694789081886, "grad_norm": 5.523021464831251, "learning_rate": 1.5045024662679985e-07, "loss": 0.4337, "step": 13034 }, { "epoch": 0.924140375753279, "grad_norm": 2.5634641452766616, "learning_rate": 1.5017083918168695e-07, "loss": 0.3911, "step": 13035 }, { "epoch": 0.9242112725983693, "grad_norm": 2.1700223483818064, "learning_rate": 1.4989168747176708e-07, "loss": 0.3797, "step": 13036 }, { "epoch": 0.9242821694434598, "grad_norm": 2.41148676200985, "learning_rate": 1.4961279151176123e-07, "loss": 0.3512, "step": 13037 }, { "epoch": 0.9243530662885502, "grad_norm": 2.382800987429851, "learning_rate": 1.4933415131637485e-07, "loss": 0.3802, "step": 13038 }, { "epoch": 0.9244239631336405, "grad_norm": 4.000600170700469, "learning_rate": 1.490557669003001e-07, "loss": 0.4319, "step": 13039 }, { "epoch": 0.9244948599787309, "grad_norm": 5.530020756133766, "learning_rate": 1.48777638278218e-07, "loss": 0.3872, "step": 13040 }, { "epoch": 0.9245657568238214, "grad_norm": 2.6956582913205236, "learning_rate": 1.484997654647935e-07, "loss": 0.3549, "step": 13041 }, { "epoch": 0.9246366536689117, "grad_norm": 2.2081380238794917, "learning_rate": 1.4822214847467985e-07, "loss": 0.3986, "step": 13042 }, { "epoch": 0.9247075505140021, "grad_norm": 3.311616566235567, "learning_rate": 1.4794478732251538e-07, "loss": 0.3505, "step": 13043 }, { "epoch": 0.9247784473590925, "grad_norm": 3.5744187723850276, "learning_rate": 1.476676820229256e-07, "loss": 0.3767, "step": 13044 }, { "epoch": 0.924849344204183, "grad_norm": 2.2290523842862777, "learning_rate": 1.4739083259052268e-07, "loss": 0.3205, "step": 13045 }, { "epoch": 0.9249202410492733, "grad_norm": 3.224893901057584, "learning_rate": 1.471142390399044e-07, "loss": 0.4518, "step": 13046 }, { "epoch": 0.9249911378943637, "grad_norm": 2.9528038136679444, "learning_rate": 1.4683790138565634e-07, "loss": 0.3707, "step": 13047 }, { "epoch": 0.9250620347394541, "grad_norm": 3.236816323650438, "learning_rate": 1.4656181964235074e-07, "loss": 0.394, "step": 13048 }, { "epoch": 0.9251329315845445, "grad_norm": 4.455123766897408, "learning_rate": 1.4628599382454477e-07, "loss": 0.3563, "step": 13049 }, { "epoch": 0.9252038284296349, "grad_norm": 7.5451606153346615, "learning_rate": 1.4601042394678245e-07, "loss": 0.4225, "step": 13050 }, { "epoch": 0.9252747252747253, "grad_norm": 2.629268038923434, "learning_rate": 1.4573511002359596e-07, "loss": 0.3511, "step": 13051 }, { "epoch": 0.9253456221198156, "grad_norm": 3.7961865779956963, "learning_rate": 1.4546005206950208e-07, "loss": 0.3769, "step": 13052 }, { "epoch": 0.9254165189649061, "grad_norm": 5.249716992482161, "learning_rate": 1.4518525009900475e-07, "loss": 0.3161, "step": 13053 }, { "epoch": 0.9254874158099965, "grad_norm": 2.3464981455141567, "learning_rate": 1.4491070412659458e-07, "loss": 0.3443, "step": 13054 }, { "epoch": 0.9255583126550868, "grad_norm": 10.976976594931227, "learning_rate": 1.4463641416674888e-07, "loss": 0.3831, "step": 13055 }, { "epoch": 0.9256292095001772, "grad_norm": 3.368935717084608, "learning_rate": 1.4436238023393113e-07, "loss": 0.393, "step": 13056 }, { "epoch": 0.9257001063452677, "grad_norm": 3.6731796236999807, "learning_rate": 1.440886023425908e-07, "loss": 0.4075, "step": 13057 }, { "epoch": 0.925771003190358, "grad_norm": 3.212936082716253, "learning_rate": 1.438150805071653e-07, "loss": 0.3511, "step": 13058 }, { "epoch": 0.9258419000354484, "grad_norm": 2.33714633514536, "learning_rate": 1.435418147420764e-07, "loss": 0.4068, "step": 13059 }, { "epoch": 0.9259127968805388, "grad_norm": 3.0422078638457593, "learning_rate": 1.4326880506173535e-07, "loss": 0.33, "step": 13060 }, { "epoch": 0.9259836937256292, "grad_norm": 3.727328946317248, "learning_rate": 1.4299605148053674e-07, "loss": 0.3603, "step": 13061 }, { "epoch": 0.9260545905707196, "grad_norm": 2.8970255046392057, "learning_rate": 1.4272355401286352e-07, "loss": 0.3893, "step": 13062 }, { "epoch": 0.92612548741581, "grad_norm": 1.9006426625725252, "learning_rate": 1.4245131267308476e-07, "loss": 0.368, "step": 13063 }, { "epoch": 0.9261963842609003, "grad_norm": 4.223079089404493, "learning_rate": 1.4217932747555618e-07, "loss": 0.3572, "step": 13064 }, { "epoch": 0.9262672811059908, "grad_norm": 4.2584719847074854, "learning_rate": 1.4190759843461853e-07, "loss": 0.4252, "step": 13065 }, { "epoch": 0.9263381779510812, "grad_norm": 26.680372244746383, "learning_rate": 1.416361255646015e-07, "loss": 0.4247, "step": 13066 }, { "epoch": 0.9264090747961715, "grad_norm": 1.3293254107787302, "learning_rate": 1.413649088798208e-07, "loss": 0.572, "step": 13067 }, { "epoch": 0.926479971641262, "grad_norm": 2.256606423939029, "learning_rate": 1.4109394839457614e-07, "loss": 0.3652, "step": 13068 }, { "epoch": 0.9265508684863524, "grad_norm": 3.3971657691303596, "learning_rate": 1.4082324412315552e-07, "loss": 0.3564, "step": 13069 }, { "epoch": 0.9266217653314428, "grad_norm": 2.5670243549561707, "learning_rate": 1.405527960798353e-07, "loss": 0.3704, "step": 13070 }, { "epoch": 0.9266926621765331, "grad_norm": 5.780375721859328, "learning_rate": 1.4028260427887465e-07, "loss": 0.3844, "step": 13071 }, { "epoch": 0.9267635590216236, "grad_norm": 3.069855725521296, "learning_rate": 1.4001266873452157e-07, "loss": 0.3961, "step": 13072 }, { "epoch": 0.926834455866714, "grad_norm": 3.6440826213995448, "learning_rate": 1.3974298946100972e-07, "loss": 0.3702, "step": 13073 }, { "epoch": 0.9269053527118043, "grad_norm": 6.4471361788277255, "learning_rate": 1.3947356647255993e-07, "loss": 0.3799, "step": 13074 }, { "epoch": 0.9269762495568947, "grad_norm": 3.3167182654610636, "learning_rate": 1.392043997833792e-07, "loss": 0.4048, "step": 13075 }, { "epoch": 0.9270471464019852, "grad_norm": 3.4457829182810062, "learning_rate": 1.3893548940765998e-07, "loss": 0.3869, "step": 13076 }, { "epoch": 0.9271180432470755, "grad_norm": 1.0907759730239768, "learning_rate": 1.386668353595827e-07, "loss": 0.6631, "step": 13077 }, { "epoch": 0.9271889400921659, "grad_norm": 2.726241816109428, "learning_rate": 1.3839843765331428e-07, "loss": 0.3449, "step": 13078 }, { "epoch": 0.9272598369372563, "grad_norm": 2.454864652645044, "learning_rate": 1.3813029630300623e-07, "loss": 0.4472, "step": 13079 }, { "epoch": 0.9273307337823467, "grad_norm": 2.408706475011837, "learning_rate": 1.3786241132279889e-07, "loss": 0.3388, "step": 13080 }, { "epoch": 0.9274016306274371, "grad_norm": 3.8839768949713602, "learning_rate": 1.375947827268176e-07, "loss": 0.359, "step": 13081 }, { "epoch": 0.9274725274725275, "grad_norm": 4.4979323738543675, "learning_rate": 1.3732741052917497e-07, "loss": 0.3639, "step": 13082 }, { "epoch": 0.9275434243176178, "grad_norm": 4.245320047344512, "learning_rate": 1.370602947439692e-07, "loss": 0.3901, "step": 13083 }, { "epoch": 0.9276143211627083, "grad_norm": 2.2971029740044098, "learning_rate": 1.3679343538528566e-07, "loss": 0.3934, "step": 13084 }, { "epoch": 0.9276852180077987, "grad_norm": 1.006723412790265, "learning_rate": 1.3652683246719644e-07, "loss": 0.6354, "step": 13085 }, { "epoch": 0.927756114852889, "grad_norm": 2.988205490152237, "learning_rate": 1.3626048600375973e-07, "loss": 0.3887, "step": 13086 }, { "epoch": 0.9278270116979794, "grad_norm": 2.628621671255978, "learning_rate": 1.3599439600901987e-07, "loss": 0.4086, "step": 13087 }, { "epoch": 0.9278979085430699, "grad_norm": 6.605474490644868, "learning_rate": 1.3572856249700729e-07, "loss": 0.3619, "step": 13088 }, { "epoch": 0.9279688053881602, "grad_norm": 4.788232678686464, "learning_rate": 1.354629854817413e-07, "loss": 0.4376, "step": 13089 }, { "epoch": 0.9280397022332506, "grad_norm": 2.9422178500182583, "learning_rate": 1.3519766497722465e-07, "loss": 0.4038, "step": 13090 }, { "epoch": 0.928110599078341, "grad_norm": 6.605217324450161, "learning_rate": 1.3493260099744775e-07, "loss": 0.3739, "step": 13091 }, { "epoch": 0.9281814959234314, "grad_norm": 2.5101268562053587, "learning_rate": 1.3466779355638838e-07, "loss": 0.3797, "step": 13092 }, { "epoch": 0.9282523927685218, "grad_norm": 3.5632532833728563, "learning_rate": 1.344032426680103e-07, "loss": 0.3745, "step": 13093 }, { "epoch": 0.9283232896136122, "grad_norm": 2.197013702679169, "learning_rate": 1.3413894834626294e-07, "loss": 0.4149, "step": 13094 }, { "epoch": 0.9283941864587026, "grad_norm": 3.033267819634251, "learning_rate": 1.3387491060508184e-07, "loss": 0.4207, "step": 13095 }, { "epoch": 0.928465083303793, "grad_norm": 1.8956298496535724, "learning_rate": 1.3361112945839195e-07, "loss": 0.3777, "step": 13096 }, { "epoch": 0.9285359801488834, "grad_norm": 0.9640669150872612, "learning_rate": 1.33347604920101e-07, "loss": 0.6393, "step": 13097 }, { "epoch": 0.9286068769939738, "grad_norm": 2.5594065606006575, "learning_rate": 1.3308433700410518e-07, "loss": 0.4201, "step": 13098 }, { "epoch": 0.9286777738390641, "grad_norm": 3.253615772844122, "learning_rate": 1.328213257242872e-07, "loss": 0.4005, "step": 13099 }, { "epoch": 0.9287486706841546, "grad_norm": 2.592868903935568, "learning_rate": 1.3255857109451598e-07, "loss": 0.355, "step": 13100 }, { "epoch": 0.928819567529245, "grad_norm": 1.8591221405197567, "learning_rate": 1.3229607312864655e-07, "loss": 0.3682, "step": 13101 }, { "epoch": 0.9288904643743353, "grad_norm": 2.9312017756033995, "learning_rate": 1.320338318405201e-07, "loss": 0.4061, "step": 13102 }, { "epoch": 0.9289613612194257, "grad_norm": 2.782120981125245, "learning_rate": 1.3177184724396552e-07, "loss": 0.3641, "step": 13103 }, { "epoch": 0.9290322580645162, "grad_norm": 2.322345338422893, "learning_rate": 1.315101193527968e-07, "loss": 0.3626, "step": 13104 }, { "epoch": 0.9291031549096065, "grad_norm": 1.6356436909277272, "learning_rate": 1.3124864818081563e-07, "loss": 0.3399, "step": 13105 }, { "epoch": 0.9291740517546969, "grad_norm": 2.278348351716593, "learning_rate": 1.309874337418099e-07, "loss": 0.3597, "step": 13106 }, { "epoch": 0.9292449485997873, "grad_norm": 2.4078110029978443, "learning_rate": 1.3072647604955247e-07, "loss": 0.3909, "step": 13107 }, { "epoch": 0.9293158454448777, "grad_norm": 2.1874484300001633, "learning_rate": 1.304657751178051e-07, "loss": 0.352, "step": 13108 }, { "epoch": 0.9293867422899681, "grad_norm": 2.934883439406614, "learning_rate": 1.3020533096031407e-07, "loss": 0.3735, "step": 13109 }, { "epoch": 0.9294576391350585, "grad_norm": 3.6053204276447945, "learning_rate": 1.2994514359081224e-07, "loss": 0.382, "step": 13110 }, { "epoch": 0.9295285359801488, "grad_norm": 3.4225683800139897, "learning_rate": 1.2968521302302028e-07, "loss": 0.3728, "step": 13111 }, { "epoch": 0.9295994328252393, "grad_norm": 2.3943956182526516, "learning_rate": 1.294255392706445e-07, "loss": 0.4113, "step": 13112 }, { "epoch": 0.9296703296703297, "grad_norm": 1.3473958672966764, "learning_rate": 1.291661223473778e-07, "loss": 0.3495, "step": 13113 }, { "epoch": 0.92974122651542, "grad_norm": 4.362885366041083, "learning_rate": 1.289069622668987e-07, "loss": 0.3915, "step": 13114 }, { "epoch": 0.9298121233605104, "grad_norm": 2.7836337252468155, "learning_rate": 1.2864805904287348e-07, "loss": 0.3751, "step": 13115 }, { "epoch": 0.9298830202056009, "grad_norm": 2.5500591022930874, "learning_rate": 1.2838941268895567e-07, "loss": 0.4091, "step": 13116 }, { "epoch": 0.9299539170506912, "grad_norm": 2.4873034561152036, "learning_rate": 1.2813102321878102e-07, "loss": 0.37, "step": 13117 }, { "epoch": 0.9300248138957816, "grad_norm": 2.587035458130597, "learning_rate": 1.2787289064597585e-07, "loss": 0.3638, "step": 13118 }, { "epoch": 0.930095710740872, "grad_norm": 11.90393851895066, "learning_rate": 1.2761501498415262e-07, "loss": 0.4108, "step": 13119 }, { "epoch": 0.9301666075859625, "grad_norm": 4.194451802035667, "learning_rate": 1.2735739624690878e-07, "loss": 0.3526, "step": 13120 }, { "epoch": 0.9302375044310528, "grad_norm": 2.3101585243805136, "learning_rate": 1.2710003444782792e-07, "loss": 0.3665, "step": 13121 }, { "epoch": 0.9303084012761432, "grad_norm": 6.864043299206382, "learning_rate": 1.2684292960048193e-07, "loss": 0.4096, "step": 13122 }, { "epoch": 0.9303792981212337, "grad_norm": 18.04002676917642, "learning_rate": 1.2658608171842724e-07, "loss": 0.3476, "step": 13123 }, { "epoch": 0.930450194966324, "grad_norm": 2.9174101293752552, "learning_rate": 1.263294908152085e-07, "loss": 0.3846, "step": 13124 }, { "epoch": 0.9305210918114144, "grad_norm": 16.366034253109703, "learning_rate": 1.2607315690435606e-07, "loss": 0.3414, "step": 13125 }, { "epoch": 0.9305919886565048, "grad_norm": 6.7449001962865385, "learning_rate": 1.258170799993852e-07, "loss": 0.3688, "step": 13126 }, { "epoch": 0.9306628855015951, "grad_norm": 7.409956367246212, "learning_rate": 1.2556126011380122e-07, "loss": 0.3975, "step": 13127 }, { "epoch": 0.9307337823466856, "grad_norm": 2.663069177449912, "learning_rate": 1.2530569726109109e-07, "loss": 0.4135, "step": 13128 }, { "epoch": 0.930804679191776, "grad_norm": 2.16550295316571, "learning_rate": 1.2505039145473352e-07, "loss": 0.3833, "step": 13129 }, { "epoch": 0.9308755760368663, "grad_norm": 3.2110930147905896, "learning_rate": 1.2479534270818884e-07, "loss": 0.409, "step": 13130 }, { "epoch": 0.9309464728819568, "grad_norm": 1.9539155440686875, "learning_rate": 1.2454055103490736e-07, "loss": 0.372, "step": 13131 }, { "epoch": 0.9310173697270472, "grad_norm": 3.1395172213571376, "learning_rate": 1.2428601644832393e-07, "loss": 0.3988, "step": 13132 }, { "epoch": 0.9310882665721375, "grad_norm": 2.0366531120330964, "learning_rate": 1.240317389618595e-07, "loss": 0.3346, "step": 13133 }, { "epoch": 0.9311591634172279, "grad_norm": 3.080227607344346, "learning_rate": 1.2377771858892384e-07, "loss": 0.3934, "step": 13134 }, { "epoch": 0.9312300602623184, "grad_norm": 6.606243054841625, "learning_rate": 1.2352395534291074e-07, "loss": 0.3888, "step": 13135 }, { "epoch": 0.9313009571074087, "grad_norm": 7.848954438149989, "learning_rate": 1.2327044923720166e-07, "loss": 0.3992, "step": 13136 }, { "epoch": 0.9313718539524991, "grad_norm": 3.285614376408469, "learning_rate": 1.2301720028516318e-07, "loss": 0.4276, "step": 13137 }, { "epoch": 0.9314427507975895, "grad_norm": 3.906642542197146, "learning_rate": 1.2276420850015124e-07, "loss": 0.391, "step": 13138 }, { "epoch": 0.9315136476426799, "grad_norm": 3.0065311586306125, "learning_rate": 1.2251147389550523e-07, "loss": 0.3688, "step": 13139 }, { "epoch": 0.9315845444877703, "grad_norm": 2.0331286952560106, "learning_rate": 1.222589964845511e-07, "loss": 0.3693, "step": 13140 }, { "epoch": 0.9316554413328607, "grad_norm": 7.255638127275695, "learning_rate": 1.2200677628060375e-07, "loss": 0.3221, "step": 13141 }, { "epoch": 0.931726338177951, "grad_norm": 1.88540179737103, "learning_rate": 1.2175481329696258e-07, "loss": 0.3869, "step": 13142 }, { "epoch": 0.9317972350230415, "grad_norm": 1.4699183393526545, "learning_rate": 1.2150310754691307e-07, "loss": 0.3666, "step": 13143 }, { "epoch": 0.9318681318681319, "grad_norm": 0.8769143710696362, "learning_rate": 1.2125165904372793e-07, "loss": 0.6456, "step": 13144 }, { "epoch": 0.9319390287132222, "grad_norm": 2.617212613740076, "learning_rate": 1.210004678006671e-07, "loss": 0.4122, "step": 13145 }, { "epoch": 0.9320099255583126, "grad_norm": 5.414632482713646, "learning_rate": 1.2074953383097555e-07, "loss": 0.3278, "step": 13146 }, { "epoch": 0.9320808224034031, "grad_norm": 2.501256380232558, "learning_rate": 1.204988571478849e-07, "loss": 0.4073, "step": 13147 }, { "epoch": 0.9321517192484935, "grad_norm": 3.549729395939968, "learning_rate": 1.2024843776461402e-07, "loss": 0.3594, "step": 13148 }, { "epoch": 0.9322226160935838, "grad_norm": 3.129869275741744, "learning_rate": 1.199982756943674e-07, "loss": 0.3133, "step": 13149 }, { "epoch": 0.9322935129386742, "grad_norm": 2.0583458119565416, "learning_rate": 1.197483709503372e-07, "loss": 0.3497, "step": 13150 }, { "epoch": 0.9323644097837647, "grad_norm": 9.689547948275004, "learning_rate": 1.1949872354570013e-07, "loss": 0.3704, "step": 13151 }, { "epoch": 0.932435306628855, "grad_norm": 2.7675267038638425, "learning_rate": 1.1924933349361957e-07, "loss": 0.3835, "step": 13152 }, { "epoch": 0.9325062034739454, "grad_norm": 2.14028761611385, "learning_rate": 1.1900020080724772e-07, "loss": 0.3541, "step": 13153 }, { "epoch": 0.9325771003190358, "grad_norm": 3.575349982736204, "learning_rate": 1.1875132549972079e-07, "loss": 0.3846, "step": 13154 }, { "epoch": 0.9326479971641262, "grad_norm": 2.725199536645636, "learning_rate": 1.185027075841616e-07, "loss": 0.4281, "step": 13155 }, { "epoch": 0.9327188940092166, "grad_norm": 3.2387491178887653, "learning_rate": 1.1825434707368077e-07, "loss": 0.3616, "step": 13156 }, { "epoch": 0.932789790854307, "grad_norm": 3.0421173699903568, "learning_rate": 1.180062439813745e-07, "loss": 0.3833, "step": 13157 }, { "epoch": 0.9328606876993973, "grad_norm": 4.822080986342874, "learning_rate": 1.1775839832032566e-07, "loss": 0.3789, "step": 13158 }, { "epoch": 0.9329315845444878, "grad_norm": 1.0605713381947781, "learning_rate": 1.1751081010360211e-07, "loss": 0.6425, "step": 13159 }, { "epoch": 0.9330024813895782, "grad_norm": 3.128344318148671, "learning_rate": 1.1726347934426008e-07, "loss": 0.3989, "step": 13160 }, { "epoch": 0.9330733782346685, "grad_norm": 4.778318030203111, "learning_rate": 1.1701640605534248e-07, "loss": 0.3862, "step": 13161 }, { "epoch": 0.933144275079759, "grad_norm": 5.29174885429334, "learning_rate": 1.1676959024987611e-07, "loss": 0.3485, "step": 13162 }, { "epoch": 0.9332151719248494, "grad_norm": 1.73777306009119, "learning_rate": 1.1652303194087666e-07, "loss": 0.3332, "step": 13163 }, { "epoch": 0.9332860687699397, "grad_norm": 2.347163035295462, "learning_rate": 1.1627673114134541e-07, "loss": 0.3943, "step": 13164 }, { "epoch": 0.9333569656150301, "grad_norm": 2.588511293298896, "learning_rate": 1.1603068786426974e-07, "loss": 0.4014, "step": 13165 }, { "epoch": 0.9334278624601205, "grad_norm": 3.086649219506354, "learning_rate": 1.1578490212262261e-07, "loss": 0.3578, "step": 13166 }, { "epoch": 0.9334987593052109, "grad_norm": 2.1333599575689526, "learning_rate": 1.1553937392936699e-07, "loss": 0.425, "step": 13167 }, { "epoch": 0.9335696561503013, "grad_norm": 1.863561962377123, "learning_rate": 1.1529410329744695e-07, "loss": 0.3909, "step": 13168 }, { "epoch": 0.9336405529953917, "grad_norm": 3.1868707405841197, "learning_rate": 1.1504909023979827e-07, "loss": 0.3906, "step": 13169 }, { "epoch": 0.933711449840482, "grad_norm": 2.2504468560098467, "learning_rate": 1.1480433476933894e-07, "loss": 0.3608, "step": 13170 }, { "epoch": 0.9337823466855725, "grad_norm": 2.121887288465848, "learning_rate": 1.1455983689897587e-07, "loss": 0.343, "step": 13171 }, { "epoch": 0.9338532435306629, "grad_norm": 2.9564805275645507, "learning_rate": 1.1431559664160208e-07, "loss": 0.4317, "step": 13172 }, { "epoch": 0.9339241403757533, "grad_norm": 3.1851182548298067, "learning_rate": 1.1407161401009504e-07, "loss": 0.3636, "step": 13173 }, { "epoch": 0.9339950372208436, "grad_norm": 5.605791352408856, "learning_rate": 1.1382788901732166e-07, "loss": 0.3972, "step": 13174 }, { "epoch": 0.9340659340659341, "grad_norm": 3.3821072014723623, "learning_rate": 1.1358442167613226e-07, "loss": 0.3934, "step": 13175 }, { "epoch": 0.9341368309110245, "grad_norm": 5.8947521410099615, "learning_rate": 1.1334121199936709e-07, "loss": 0.3689, "step": 13176 }, { "epoch": 0.9342077277561148, "grad_norm": 3.9759923677577484, "learning_rate": 1.1309825999984924e-07, "loss": 0.3962, "step": 13177 }, { "epoch": 0.9342786246012053, "grad_norm": 4.766634281022346, "learning_rate": 1.1285556569038958e-07, "loss": 0.3986, "step": 13178 }, { "epoch": 0.9343495214462957, "grad_norm": 4.098794699218059, "learning_rate": 1.1261312908378675e-07, "loss": 0.422, "step": 13179 }, { "epoch": 0.934420418291386, "grad_norm": 1.6542360243890435, "learning_rate": 1.123709501928244e-07, "loss": 0.3198, "step": 13180 }, { "epoch": 0.9344913151364764, "grad_norm": 8.635976376501237, "learning_rate": 1.1212902903027123e-07, "loss": 0.3875, "step": 13181 }, { "epoch": 0.9345622119815669, "grad_norm": 3.23849034874048, "learning_rate": 1.1188736560888592e-07, "loss": 0.449, "step": 13182 }, { "epoch": 0.9346331088266572, "grad_norm": 2.1616672231553, "learning_rate": 1.1164595994141047e-07, "loss": 0.4069, "step": 13183 }, { "epoch": 0.9347040056717476, "grad_norm": 2.1083316092731677, "learning_rate": 1.1140481204057473e-07, "loss": 0.3801, "step": 13184 }, { "epoch": 0.934774902516838, "grad_norm": 2.089778895795927, "learning_rate": 1.1116392191909464e-07, "loss": 0.3582, "step": 13185 }, { "epoch": 0.9348457993619284, "grad_norm": 4.515420181489118, "learning_rate": 1.1092328958967224e-07, "loss": 0.4167, "step": 13186 }, { "epoch": 0.9349166962070188, "grad_norm": 3.4570010832395863, "learning_rate": 1.1068291506499684e-07, "loss": 0.4586, "step": 13187 }, { "epoch": 0.9349875930521092, "grad_norm": 3.3696153571182514, "learning_rate": 1.1044279835774274e-07, "loss": 0.3883, "step": 13188 }, { "epoch": 0.9350584898971995, "grad_norm": 2.689030772741752, "learning_rate": 1.1020293948057148e-07, "loss": 0.4086, "step": 13189 }, { "epoch": 0.93512938674229, "grad_norm": 3.7020644911095437, "learning_rate": 1.0996333844613238e-07, "loss": 0.4159, "step": 13190 }, { "epoch": 0.9352002835873804, "grad_norm": 2.7806406327411106, "learning_rate": 1.0972399526705868e-07, "loss": 0.3958, "step": 13191 }, { "epoch": 0.9352711804324707, "grad_norm": 2.863975232333579, "learning_rate": 1.0948490995597083e-07, "loss": 0.3929, "step": 13192 }, { "epoch": 0.9353420772775611, "grad_norm": 0.9506510172015425, "learning_rate": 1.0924608252547652e-07, "loss": 0.6611, "step": 13193 }, { "epoch": 0.9354129741226516, "grad_norm": 4.25828376759602, "learning_rate": 1.0900751298816958e-07, "loss": 0.3954, "step": 13194 }, { "epoch": 0.9354838709677419, "grad_norm": 4.788024786683888, "learning_rate": 1.0876920135662938e-07, "loss": 0.3415, "step": 13195 }, { "epoch": 0.9355547678128323, "grad_norm": 2.4467344921667133, "learning_rate": 1.0853114764342254e-07, "loss": 0.3401, "step": 13196 }, { "epoch": 0.9356256646579227, "grad_norm": 4.546404470736384, "learning_rate": 1.0829335186110179e-07, "loss": 0.4083, "step": 13197 }, { "epoch": 0.9356965615030132, "grad_norm": 2.3564344815689924, "learning_rate": 1.08055814022206e-07, "loss": 0.4424, "step": 13198 }, { "epoch": 0.9357674583481035, "grad_norm": 3.4919020516320516, "learning_rate": 1.078185341392618e-07, "loss": 0.404, "step": 13199 }, { "epoch": 0.9358383551931939, "grad_norm": 2.6727119698494413, "learning_rate": 1.0758151222477919e-07, "loss": 0.3982, "step": 13200 }, { "epoch": 0.9359092520382843, "grad_norm": 18.485771387892004, "learning_rate": 1.0734474829125818e-07, "loss": 0.4179, "step": 13201 }, { "epoch": 0.9359801488833747, "grad_norm": 22.860411448537874, "learning_rate": 1.0710824235118322e-07, "loss": 0.4235, "step": 13202 }, { "epoch": 0.9360510457284651, "grad_norm": 2.018769465455654, "learning_rate": 1.0687199441702489e-07, "loss": 0.3854, "step": 13203 }, { "epoch": 0.9361219425735555, "grad_norm": 2.2415438140623167, "learning_rate": 1.06636004501241e-07, "loss": 0.3971, "step": 13204 }, { "epoch": 0.9361928394186458, "grad_norm": 3.486055008775941, "learning_rate": 1.064002726162755e-07, "loss": 0.3562, "step": 13205 }, { "epoch": 0.9362637362637363, "grad_norm": 5.241604001792649, "learning_rate": 1.0616479877455954e-07, "loss": 0.3773, "step": 13206 }, { "epoch": 0.9363346331088267, "grad_norm": 3.9693116469557594, "learning_rate": 1.0592958298850819e-07, "loss": 0.4155, "step": 13207 }, { "epoch": 0.936405529953917, "grad_norm": 2.8012544100409618, "learning_rate": 1.0569462527052543e-07, "loss": 0.3498, "step": 13208 }, { "epoch": 0.9364764267990074, "grad_norm": 5.991072734825486, "learning_rate": 1.0545992563300134e-07, "loss": 0.4082, "step": 13209 }, { "epoch": 0.9365473236440979, "grad_norm": 2.823426557831522, "learning_rate": 1.0522548408831102e-07, "loss": 0.4144, "step": 13210 }, { "epoch": 0.9366182204891882, "grad_norm": 2.6113508272178905, "learning_rate": 1.0499130064881625e-07, "loss": 0.3784, "step": 13211 }, { "epoch": 0.9366891173342786, "grad_norm": 2.489238245813274, "learning_rate": 1.0475737532686714e-07, "loss": 0.4128, "step": 13212 }, { "epoch": 0.936760014179369, "grad_norm": 3.699285036820814, "learning_rate": 1.0452370813479717e-07, "loss": 0.3852, "step": 13213 }, { "epoch": 0.9368309110244594, "grad_norm": 2.6289291717389833, "learning_rate": 1.0429029908492983e-07, "loss": 0.346, "step": 13214 }, { "epoch": 0.9369018078695498, "grad_norm": 43.16579916747073, "learning_rate": 1.040571481895708e-07, "loss": 0.3741, "step": 13215 }, { "epoch": 0.9369727047146402, "grad_norm": 2.147444803505253, "learning_rate": 1.0382425546101526e-07, "loss": 0.3482, "step": 13216 }, { "epoch": 0.9370436015597305, "grad_norm": 3.331101359529364, "learning_rate": 1.0359162091154452e-07, "loss": 0.3398, "step": 13217 }, { "epoch": 0.937114498404821, "grad_norm": 2.42763888840312, "learning_rate": 1.0335924455342428e-07, "loss": 0.3655, "step": 13218 }, { "epoch": 0.9371853952499114, "grad_norm": 2.448756161016195, "learning_rate": 1.0312712639890865e-07, "loss": 0.4491, "step": 13219 }, { "epoch": 0.9372562920950017, "grad_norm": 2.699475701030992, "learning_rate": 1.0289526646023729e-07, "loss": 0.379, "step": 13220 }, { "epoch": 0.9373271889400921, "grad_norm": 2.2859464700502388, "learning_rate": 1.0266366474963652e-07, "loss": 0.3571, "step": 13221 }, { "epoch": 0.9373980857851826, "grad_norm": 4.386583996958745, "learning_rate": 1.0243232127931823e-07, "loss": 0.4101, "step": 13222 }, { "epoch": 0.937468982630273, "grad_norm": 1.9399808350487522, "learning_rate": 1.0220123606148214e-07, "loss": 0.3349, "step": 13223 }, { "epoch": 0.9375398794753633, "grad_norm": 2.142574275068638, "learning_rate": 1.019704091083129e-07, "loss": 0.4058, "step": 13224 }, { "epoch": 0.9376107763204538, "grad_norm": 2.1667158275976472, "learning_rate": 1.0173984043198304e-07, "loss": 0.3623, "step": 13225 }, { "epoch": 0.9376816731655442, "grad_norm": 3.378256640460287, "learning_rate": 1.0150953004464948e-07, "loss": 0.3701, "step": 13226 }, { "epoch": 0.9377525700106345, "grad_norm": 1.0990758600019852, "learning_rate": 1.0127947795845694e-07, "loss": 0.6279, "step": 13227 }, { "epoch": 0.9378234668557249, "grad_norm": 4.3239472371653855, "learning_rate": 1.010496841855374e-07, "loss": 0.3527, "step": 13228 }, { "epoch": 0.9378943637008154, "grad_norm": 2.8641478137608196, "learning_rate": 1.0082014873800727e-07, "loss": 0.3838, "step": 13229 }, { "epoch": 0.9379652605459057, "grad_norm": 3.2152201762418264, "learning_rate": 1.005908716279691e-07, "loss": 0.4045, "step": 13230 }, { "epoch": 0.9380361573909961, "grad_norm": 2.915332661510461, "learning_rate": 1.0036185286751432e-07, "loss": 0.3558, "step": 13231 }, { "epoch": 0.9381070542360865, "grad_norm": 8.739199980215924, "learning_rate": 1.0013309246871883e-07, "loss": 0.3959, "step": 13232 }, { "epoch": 0.9381779510811769, "grad_norm": 9.574064859426986, "learning_rate": 9.990459044364575e-08, "loss": 0.3823, "step": 13233 }, { "epoch": 0.9382488479262673, "grad_norm": 2.5715515102178537, "learning_rate": 9.967634680434268e-08, "loss": 0.4155, "step": 13234 }, { "epoch": 0.9383197447713577, "grad_norm": 4.127211284848097, "learning_rate": 9.944836156284665e-08, "loss": 0.3578, "step": 13235 }, { "epoch": 0.938390641616448, "grad_norm": 3.604320826221127, "learning_rate": 9.922063473117915e-08, "loss": 0.3345, "step": 13236 }, { "epoch": 0.9384615384615385, "grad_norm": 1.7687408805614282, "learning_rate": 9.899316632134781e-08, "loss": 0.3709, "step": 13237 }, { "epoch": 0.9385324353066289, "grad_norm": 3.9435075577981937, "learning_rate": 9.876595634534802e-08, "loss": 0.3212, "step": 13238 }, { "epoch": 0.9386033321517192, "grad_norm": 3.078960804300364, "learning_rate": 9.853900481515966e-08, "loss": 0.3766, "step": 13239 }, { "epoch": 0.9386742289968096, "grad_norm": 2.76003859270138, "learning_rate": 9.831231174275091e-08, "loss": 0.3751, "step": 13240 }, { "epoch": 0.9387451258419001, "grad_norm": 3.4951518377532866, "learning_rate": 9.808587714007556e-08, "loss": 0.364, "step": 13241 }, { "epoch": 0.9388160226869904, "grad_norm": 4.308829449114419, "learning_rate": 9.785970101907294e-08, "loss": 0.3655, "step": 13242 }, { "epoch": 0.9388869195320808, "grad_norm": 3.428480323274237, "learning_rate": 9.763378339167017e-08, "loss": 0.3421, "step": 13243 }, { "epoch": 0.9389578163771712, "grad_norm": 3.113288946798594, "learning_rate": 9.740812426977997e-08, "loss": 0.3783, "step": 13244 }, { "epoch": 0.9390287132222616, "grad_norm": 2.357542783345006, "learning_rate": 9.718272366530112e-08, "loss": 0.3772, "step": 13245 }, { "epoch": 0.939099610067352, "grad_norm": 3.057168748081461, "learning_rate": 9.695758159011915e-08, "loss": 0.3376, "step": 13246 }, { "epoch": 0.9391705069124424, "grad_norm": 2.333942927756524, "learning_rate": 9.673269805610675e-08, "loss": 0.3866, "step": 13247 }, { "epoch": 0.9392414037575328, "grad_norm": 0.9819366961002459, "learning_rate": 9.650807307512166e-08, "loss": 0.5704, "step": 13248 }, { "epoch": 0.9393123006026232, "grad_norm": 4.070928270251704, "learning_rate": 9.628370665900778e-08, "loss": 0.4215, "step": 13249 }, { "epoch": 0.9393831974477136, "grad_norm": 3.9449381761597895, "learning_rate": 9.605959881959726e-08, "loss": 0.3548, "step": 13250 }, { "epoch": 0.939454094292804, "grad_norm": 1.6828945845120344, "learning_rate": 9.583574956870789e-08, "loss": 0.3562, "step": 13251 }, { "epoch": 0.9395249911378943, "grad_norm": 2.638763320447133, "learning_rate": 9.56121589181419e-08, "loss": 0.4039, "step": 13252 }, { "epoch": 0.9395958879829848, "grad_norm": 8.097485244201481, "learning_rate": 9.538882687968986e-08, "loss": 0.372, "step": 13253 }, { "epoch": 0.9396667848280752, "grad_norm": 3.3487659984925493, "learning_rate": 9.516575346512902e-08, "loss": 0.4029, "step": 13254 }, { "epoch": 0.9397376816731655, "grad_norm": 12.390823760152442, "learning_rate": 9.494293868622162e-08, "loss": 0.4056, "step": 13255 }, { "epoch": 0.9398085785182559, "grad_norm": 4.205519748404095, "learning_rate": 9.472038255471661e-08, "loss": 0.3738, "step": 13256 }, { "epoch": 0.9398794753633464, "grad_norm": 2.8590890442326784, "learning_rate": 9.44980850823507e-08, "loss": 0.4095, "step": 13257 }, { "epoch": 0.9399503722084367, "grad_norm": 3.3462747222771196, "learning_rate": 9.427604628084396e-08, "loss": 0.3428, "step": 13258 }, { "epoch": 0.9400212690535271, "grad_norm": 0.9780478916212967, "learning_rate": 9.405426616190594e-08, "loss": 0.6502, "step": 13259 }, { "epoch": 0.9400921658986175, "grad_norm": 1.8407190874733637, "learning_rate": 9.38327447372317e-08, "loss": 0.3373, "step": 13260 }, { "epoch": 0.9401630627437079, "grad_norm": 4.739379535425824, "learning_rate": 9.361148201850024e-08, "loss": 0.3696, "step": 13261 }, { "epoch": 0.9402339595887983, "grad_norm": 3.13302845258217, "learning_rate": 9.339047801738111e-08, "loss": 0.3322, "step": 13262 }, { "epoch": 0.9403048564338887, "grad_norm": 2.247765706714054, "learning_rate": 9.316973274552665e-08, "loss": 0.374, "step": 13263 }, { "epoch": 0.940375753278979, "grad_norm": 2.3559305788056393, "learning_rate": 9.294924621457813e-08, "loss": 0.3841, "step": 13264 }, { "epoch": 0.9404466501240695, "grad_norm": 2.116076985993355, "learning_rate": 9.272901843616067e-08, "loss": 0.395, "step": 13265 }, { "epoch": 0.9405175469691599, "grad_norm": 5.178510575041906, "learning_rate": 9.250904942188777e-08, "loss": 0.3851, "step": 13266 }, { "epoch": 0.9405884438142502, "grad_norm": 6.00381968956699, "learning_rate": 9.228933918335848e-08, "loss": 0.4492, "step": 13267 }, { "epoch": 0.9406593406593406, "grad_norm": 3.046717856083282, "learning_rate": 9.206988773215797e-08, "loss": 0.4584, "step": 13268 }, { "epoch": 0.9407302375044311, "grad_norm": 3.846637121115079, "learning_rate": 9.18506950798581e-08, "loss": 0.4089, "step": 13269 }, { "epoch": 0.9408011343495214, "grad_norm": 2.477102290400168, "learning_rate": 9.163176123801854e-08, "loss": 0.379, "step": 13270 }, { "epoch": 0.9408720311946118, "grad_norm": 2.273502947948745, "learning_rate": 9.141308621818112e-08, "loss": 0.3175, "step": 13271 }, { "epoch": 0.9409429280397023, "grad_norm": 2.597753127997177, "learning_rate": 9.119467003187888e-08, "loss": 0.3644, "step": 13272 }, { "epoch": 0.9410138248847926, "grad_norm": 3.1351169576102262, "learning_rate": 9.097651269062868e-08, "loss": 0.3436, "step": 13273 }, { "epoch": 0.941084721729883, "grad_norm": 2.2667697198668657, "learning_rate": 9.075861420593358e-08, "loss": 0.3561, "step": 13274 }, { "epoch": 0.9411556185749734, "grad_norm": 2.4033767693264028, "learning_rate": 9.054097458928323e-08, "loss": 0.375, "step": 13275 }, { "epoch": 0.9412265154200639, "grad_norm": 3.700568879953092, "learning_rate": 9.032359385215517e-08, "loss": 0.3581, "step": 13276 }, { "epoch": 0.9412974122651542, "grad_norm": 1.7972065110885136, "learning_rate": 9.01064720060113e-08, "loss": 0.3527, "step": 13277 }, { "epoch": 0.9413683091102446, "grad_norm": 2.722268027065909, "learning_rate": 8.988960906230083e-08, "loss": 0.3691, "step": 13278 }, { "epoch": 0.941439205955335, "grad_norm": 4.66243840048977, "learning_rate": 8.967300503245846e-08, "loss": 0.3902, "step": 13279 }, { "epoch": 0.9415101028004254, "grad_norm": 3.0172339620903292, "learning_rate": 8.945665992790731e-08, "loss": 0.3867, "step": 13280 }, { "epoch": 0.9415809996455158, "grad_norm": 3.5740287011596155, "learning_rate": 8.924057376005379e-08, "loss": 0.3844, "step": 13281 }, { "epoch": 0.9416518964906062, "grad_norm": 9.125941728960777, "learning_rate": 8.902474654029325e-08, "loss": 0.3987, "step": 13282 }, { "epoch": 0.9417227933356965, "grad_norm": 2.432597455423047, "learning_rate": 8.880917828000602e-08, "loss": 0.3889, "step": 13283 }, { "epoch": 0.941793690180787, "grad_norm": 2.6104618161999733, "learning_rate": 8.859386899055967e-08, "loss": 0.3456, "step": 13284 }, { "epoch": 0.9418645870258774, "grad_norm": 3.051863869393726, "learning_rate": 8.837881868330733e-08, "loss": 0.4407, "step": 13285 }, { "epoch": 0.9419354838709677, "grad_norm": 3.52673456762784, "learning_rate": 8.816402736958885e-08, "loss": 0.4023, "step": 13286 }, { "epoch": 0.9420063807160581, "grad_norm": 2.719981529465757, "learning_rate": 8.794949506072959e-08, "loss": 0.3994, "step": 13287 }, { "epoch": 0.9420772775611486, "grad_norm": 2.396616356558578, "learning_rate": 8.773522176804272e-08, "loss": 0.4102, "step": 13288 }, { "epoch": 0.9421481744062389, "grad_norm": 2.3924504607386305, "learning_rate": 8.752120750282812e-08, "loss": 0.4085, "step": 13289 }, { "epoch": 0.9422190712513293, "grad_norm": 2.9870668447012556, "learning_rate": 8.730745227636839e-08, "loss": 0.3595, "step": 13290 }, { "epoch": 0.9422899680964197, "grad_norm": 3.1611503648555535, "learning_rate": 8.709395609993676e-08, "loss": 0.3651, "step": 13291 }, { "epoch": 0.9423608649415101, "grad_norm": 2.438783321314552, "learning_rate": 8.688071898479144e-08, "loss": 0.4218, "step": 13292 }, { "epoch": 0.9424317617866005, "grad_norm": 2.9652806100657623, "learning_rate": 8.666774094217512e-08, "loss": 0.3919, "step": 13293 }, { "epoch": 0.9425026586316909, "grad_norm": 2.1366451845728167, "learning_rate": 8.645502198331879e-08, "loss": 0.4227, "step": 13294 }, { "epoch": 0.9425735554767812, "grad_norm": 7.926083102302876, "learning_rate": 8.62425621194396e-08, "loss": 0.3778, "step": 13295 }, { "epoch": 0.9426444523218717, "grad_norm": 2.5258389976402307, "learning_rate": 8.603036136174192e-08, "loss": 0.393, "step": 13296 }, { "epoch": 0.9427153491669621, "grad_norm": 2.515529480509988, "learning_rate": 8.581841972141236e-08, "loss": 0.3713, "step": 13297 }, { "epoch": 0.9427862460120524, "grad_norm": 3.8618076778028914, "learning_rate": 8.560673720962864e-08, "loss": 0.4228, "step": 13298 }, { "epoch": 0.9428571428571428, "grad_norm": 3.57885539944195, "learning_rate": 8.539531383755296e-08, "loss": 0.4586, "step": 13299 }, { "epoch": 0.9429280397022333, "grad_norm": 4.698217014732196, "learning_rate": 8.518414961633359e-08, "loss": 0.3461, "step": 13300 }, { "epoch": 0.9429989365473237, "grad_norm": 16.62382730530149, "learning_rate": 8.497324455710498e-08, "loss": 0.3986, "step": 13301 }, { "epoch": 0.943069833392414, "grad_norm": 4.879721395374517, "learning_rate": 8.476259867098935e-08, "loss": 0.371, "step": 13302 }, { "epoch": 0.9431407302375044, "grad_norm": 4.298968925252542, "learning_rate": 8.45522119690928e-08, "loss": 0.3592, "step": 13303 }, { "epoch": 0.9432116270825949, "grad_norm": 4.82660155477177, "learning_rate": 8.43420844625098e-08, "loss": 0.3519, "step": 13304 }, { "epoch": 0.9432825239276852, "grad_norm": 3.9177419046923934, "learning_rate": 8.413221616232093e-08, "loss": 0.3914, "step": 13305 }, { "epoch": 0.9433534207727756, "grad_norm": 1.038732753601263, "learning_rate": 8.392260707959287e-08, "loss": 0.635, "step": 13306 }, { "epoch": 0.943424317617866, "grad_norm": 3.755861898448779, "learning_rate": 8.371325722537737e-08, "loss": 0.3973, "step": 13307 }, { "epoch": 0.9434952144629564, "grad_norm": 5.0891266978642, "learning_rate": 8.350416661071447e-08, "loss": 0.3921, "step": 13308 }, { "epoch": 0.9435661113080468, "grad_norm": 2.417573729781909, "learning_rate": 8.32953352466298e-08, "loss": 0.3826, "step": 13309 }, { "epoch": 0.9436370081531372, "grad_norm": 1.7122351594927272, "learning_rate": 8.3086763144134e-08, "loss": 0.3879, "step": 13310 }, { "epoch": 0.9437079049982275, "grad_norm": 5.148263135563777, "learning_rate": 8.287845031422659e-08, "loss": 0.3842, "step": 13311 }, { "epoch": 0.943778801843318, "grad_norm": 4.366066608597086, "learning_rate": 8.267039676789212e-08, "loss": 0.3359, "step": 13312 }, { "epoch": 0.9438496986884084, "grad_norm": 3.271254639948415, "learning_rate": 8.246260251609961e-08, "loss": 0.4055, "step": 13313 }, { "epoch": 0.9439205955334987, "grad_norm": 2.7409642467401776, "learning_rate": 8.225506756980806e-08, "loss": 0.4016, "step": 13314 }, { "epoch": 0.9439914923785891, "grad_norm": 2.183047835010534, "learning_rate": 8.204779193996094e-08, "loss": 0.3343, "step": 13315 }, { "epoch": 0.9440623892236796, "grad_norm": 2.421133330705327, "learning_rate": 8.184077563748671e-08, "loss": 0.4029, "step": 13316 }, { "epoch": 0.9441332860687699, "grad_norm": 2.127210050958343, "learning_rate": 8.163401867330223e-08, "loss": 0.3574, "step": 13317 }, { "epoch": 0.9442041829138603, "grad_norm": 2.122960112230667, "learning_rate": 8.142752105831043e-08, "loss": 0.3828, "step": 13318 }, { "epoch": 0.9442750797589508, "grad_norm": 2.3267723147135038, "learning_rate": 8.12212828033998e-08, "loss": 0.3712, "step": 13319 }, { "epoch": 0.9443459766040411, "grad_norm": 4.407981868345585, "learning_rate": 8.101530391944446e-08, "loss": 0.3333, "step": 13320 }, { "epoch": 0.9444168734491315, "grad_norm": 2.827221083107052, "learning_rate": 8.08095844173068e-08, "loss": 0.4216, "step": 13321 }, { "epoch": 0.9444877702942219, "grad_norm": 3.150163940574865, "learning_rate": 8.060412430783481e-08, "loss": 0.3888, "step": 13322 }, { "epoch": 0.9445586671393122, "grad_norm": 1.1169477021643637, "learning_rate": 8.039892360186208e-08, "loss": 0.6938, "step": 13323 }, { "epoch": 0.9446295639844027, "grad_norm": 1.0843908846824573, "learning_rate": 8.019398231020881e-08, "loss": 0.6157, "step": 13324 }, { "epoch": 0.9447004608294931, "grad_norm": 4.976036443728694, "learning_rate": 7.998930044368247e-08, "loss": 0.3924, "step": 13325 }, { "epoch": 0.9447713576745835, "grad_norm": 2.571790189958794, "learning_rate": 7.978487801307555e-08, "loss": 0.36, "step": 13326 }, { "epoch": 0.9448422545196739, "grad_norm": 2.2088578186650665, "learning_rate": 7.958071502916664e-08, "loss": 0.3666, "step": 13327 }, { "epoch": 0.9449131513647643, "grad_norm": 2.678618572033771, "learning_rate": 7.93768115027227e-08, "loss": 0.4096, "step": 13328 }, { "epoch": 0.9449840482098547, "grad_norm": 2.511024703922841, "learning_rate": 7.91731674444951e-08, "loss": 0.3532, "step": 13329 }, { "epoch": 0.945054945054945, "grad_norm": 3.478163892841377, "learning_rate": 7.89697828652225e-08, "loss": 0.362, "step": 13330 }, { "epoch": 0.9451258419000355, "grad_norm": 1.9632777636708756, "learning_rate": 7.876665777562964e-08, "loss": 0.3851, "step": 13331 }, { "epoch": 0.9451967387451259, "grad_norm": 3.9950548065821585, "learning_rate": 7.856379218642629e-08, "loss": 0.37, "step": 13332 }, { "epoch": 0.9452676355902162, "grad_norm": 2.1886808987263207, "learning_rate": 7.836118610831056e-08, "loss": 0.4085, "step": 13333 }, { "epoch": 0.9453385324353066, "grad_norm": 2.4491802003115226, "learning_rate": 7.815883955196668e-08, "loss": 0.3813, "step": 13334 }, { "epoch": 0.9454094292803971, "grad_norm": 4.02644219273384, "learning_rate": 7.79567525280639e-08, "loss": 0.42, "step": 13335 }, { "epoch": 0.9454803261254874, "grad_norm": 0.9336533638207639, "learning_rate": 7.775492504725757e-08, "loss": 0.5951, "step": 13336 }, { "epoch": 0.9455512229705778, "grad_norm": 6.557344139810762, "learning_rate": 7.755335712019197e-08, "loss": 0.3626, "step": 13337 }, { "epoch": 0.9456221198156682, "grad_norm": 1.980059943866189, "learning_rate": 7.735204875749469e-08, "loss": 0.3672, "step": 13338 }, { "epoch": 0.9456930166607586, "grad_norm": 2.384142184410747, "learning_rate": 7.715099996978059e-08, "loss": 0.4045, "step": 13339 }, { "epoch": 0.945763913505849, "grad_norm": 3.313235207532158, "learning_rate": 7.695021076765175e-08, "loss": 0.4043, "step": 13340 }, { "epoch": 0.9458348103509394, "grad_norm": 1.9120962162047628, "learning_rate": 7.67496811616969e-08, "loss": 0.415, "step": 13341 }, { "epoch": 0.9459057071960297, "grad_norm": 3.86904159879551, "learning_rate": 7.654941116248816e-08, "loss": 0.3565, "step": 13342 }, { "epoch": 0.9459766040411202, "grad_norm": 2.3861357445441502, "learning_rate": 7.634940078058706e-08, "loss": 0.333, "step": 13343 }, { "epoch": 0.9460475008862106, "grad_norm": 3.21391039721311, "learning_rate": 7.614965002654074e-08, "loss": 0.4068, "step": 13344 }, { "epoch": 0.9461183977313009, "grad_norm": 2.847923597319999, "learning_rate": 7.595015891088076e-08, "loss": 0.3372, "step": 13345 }, { "epoch": 0.9461892945763913, "grad_norm": 4.165172790790379, "learning_rate": 7.575092744412759e-08, "loss": 0.3777, "step": 13346 }, { "epoch": 0.9462601914214818, "grad_norm": 2.621237296323775, "learning_rate": 7.555195563678674e-08, "loss": 0.3439, "step": 13347 }, { "epoch": 0.9463310882665721, "grad_norm": 4.194875049424927, "learning_rate": 7.535324349934925e-08, "loss": 0.4096, "step": 13348 }, { "epoch": 0.9464019851116625, "grad_norm": 2.7080637881190985, "learning_rate": 7.515479104229451e-08, "loss": 0.3679, "step": 13349 }, { "epoch": 0.9464728819567529, "grad_norm": 3.2095220066181978, "learning_rate": 7.495659827608636e-08, "loss": 0.3301, "step": 13350 }, { "epoch": 0.9465437788018434, "grad_norm": 6.17033486114281, "learning_rate": 7.475866521117647e-08, "loss": 0.3917, "step": 13351 }, { "epoch": 0.9466146756469337, "grad_norm": 4.267953807263858, "learning_rate": 7.456099185800092e-08, "loss": 0.4181, "step": 13352 }, { "epoch": 0.9466855724920241, "grad_norm": 2.124552526343456, "learning_rate": 7.43635782269836e-08, "loss": 0.3771, "step": 13353 }, { "epoch": 0.9467564693371145, "grad_norm": 11.689266399314462, "learning_rate": 7.416642432853505e-08, "loss": 0.3922, "step": 13354 }, { "epoch": 0.9468273661822049, "grad_norm": 3.211215985049574, "learning_rate": 7.396953017304975e-08, "loss": 0.3734, "step": 13355 }, { "epoch": 0.9468982630272953, "grad_norm": 2.557980853023893, "learning_rate": 7.377289577091163e-08, "loss": 0.3772, "step": 13356 }, { "epoch": 0.9469691598723857, "grad_norm": 6.896101608601327, "learning_rate": 7.357652113248904e-08, "loss": 0.4201, "step": 13357 }, { "epoch": 0.947040056717476, "grad_norm": 3.537256369336701, "learning_rate": 7.338040626813592e-08, "loss": 0.3984, "step": 13358 }, { "epoch": 0.9471109535625665, "grad_norm": 20.68750408861715, "learning_rate": 7.318455118819401e-08, "loss": 0.358, "step": 13359 }, { "epoch": 0.9471818504076569, "grad_norm": 4.039352657912062, "learning_rate": 7.298895590299226e-08, "loss": 0.3535, "step": 13360 }, { "epoch": 0.9472527472527472, "grad_norm": 4.879159709296776, "learning_rate": 7.2793620422843e-08, "loss": 0.3375, "step": 13361 }, { "epoch": 0.9473236440978376, "grad_norm": 9.558201149989824, "learning_rate": 7.259854475804629e-08, "loss": 0.3205, "step": 13362 }, { "epoch": 0.9473945409429281, "grad_norm": 3.630987374907896, "learning_rate": 7.240372891889003e-08, "loss": 0.3807, "step": 13363 }, { "epoch": 0.9474654377880184, "grad_norm": 7.322544862639144, "learning_rate": 7.2209172915646e-08, "loss": 0.3511, "step": 13364 }, { "epoch": 0.9475363346331088, "grad_norm": 2.8196063234731295, "learning_rate": 7.201487675857322e-08, "loss": 0.3594, "step": 13365 }, { "epoch": 0.9476072314781993, "grad_norm": 2.4578642336114944, "learning_rate": 7.182084045791737e-08, "loss": 0.3808, "step": 13366 }, { "epoch": 0.9476781283232896, "grad_norm": 2.445625875615389, "learning_rate": 7.162706402391084e-08, "loss": 0.3483, "step": 13367 }, { "epoch": 0.94774902516838, "grad_norm": 14.310209249452068, "learning_rate": 7.143354746677045e-08, "loss": 0.3747, "step": 13368 }, { "epoch": 0.9478199220134704, "grad_norm": 2.673419006893418, "learning_rate": 7.124029079670025e-08, "loss": 0.3224, "step": 13369 }, { "epoch": 0.9478908188585607, "grad_norm": 3.98181429079822, "learning_rate": 7.104729402389265e-08, "loss": 0.4145, "step": 13370 }, { "epoch": 0.9479617157036512, "grad_norm": 2.8336094449924945, "learning_rate": 7.085455715852285e-08, "loss": 0.3583, "step": 13371 }, { "epoch": 0.9480326125487416, "grad_norm": 2.179511746598623, "learning_rate": 7.066208021075383e-08, "loss": 0.3616, "step": 13372 }, { "epoch": 0.9481035093938319, "grad_norm": 3.0367401800372504, "learning_rate": 7.046986319073635e-08, "loss": 0.4094, "step": 13373 }, { "epoch": 0.9481744062389224, "grad_norm": 1.8195408879320343, "learning_rate": 7.027790610860563e-08, "loss": 0.3675, "step": 13374 }, { "epoch": 0.9482453030840128, "grad_norm": 2.500333391272794, "learning_rate": 7.008620897448304e-08, "loss": 0.3664, "step": 13375 }, { "epoch": 0.9483161999291032, "grad_norm": 12.08907580128158, "learning_rate": 6.989477179847826e-08, "loss": 0.4444, "step": 13376 }, { "epoch": 0.9483870967741935, "grad_norm": 7.989308228491881, "learning_rate": 6.970359459068432e-08, "loss": 0.4484, "step": 13377 }, { "epoch": 0.948457993619284, "grad_norm": 2.5545935892152776, "learning_rate": 6.951267736118316e-08, "loss": 0.3481, "step": 13378 }, { "epoch": 0.9485288904643744, "grad_norm": 2.401629326560575, "learning_rate": 6.93220201200423e-08, "loss": 0.3625, "step": 13379 }, { "epoch": 0.9485997873094647, "grad_norm": 3.343523342428423, "learning_rate": 6.913162287731479e-08, "loss": 0.3414, "step": 13380 }, { "epoch": 0.9486706841545551, "grad_norm": 2.7683988761509903, "learning_rate": 6.894148564303981e-08, "loss": 0.4537, "step": 13381 }, { "epoch": 0.9487415809996456, "grad_norm": 2.1842511286751063, "learning_rate": 6.875160842724437e-08, "loss": 0.3596, "step": 13382 }, { "epoch": 0.9488124778447359, "grad_norm": 11.237177579658622, "learning_rate": 6.856199123994045e-08, "loss": 0.3944, "step": 13383 }, { "epoch": 0.9488833746898263, "grad_norm": 3.604167161984974, "learning_rate": 6.837263409112616e-08, "loss": 0.3993, "step": 13384 }, { "epoch": 0.9489542715349167, "grad_norm": 3.106332275671749, "learning_rate": 6.818353699078739e-08, "loss": 0.416, "step": 13385 }, { "epoch": 0.9490251683800071, "grad_norm": 4.525702910195066, "learning_rate": 6.799469994889507e-08, "loss": 0.3682, "step": 13386 }, { "epoch": 0.9490960652250975, "grad_norm": 6.394573636259875, "learning_rate": 6.780612297540678e-08, "loss": 0.3495, "step": 13387 }, { "epoch": 0.9491669620701879, "grad_norm": 3.5586101966878556, "learning_rate": 6.761780608026569e-08, "loss": 0.401, "step": 13388 }, { "epoch": 0.9492378589152782, "grad_norm": 5.142731162075058, "learning_rate": 6.742974927340273e-08, "loss": 0.374, "step": 13389 }, { "epoch": 0.9493087557603687, "grad_norm": 2.4878480846493836, "learning_rate": 6.724195256473443e-08, "loss": 0.381, "step": 13390 }, { "epoch": 0.9493796526054591, "grad_norm": 4.0078096214978824, "learning_rate": 6.705441596416229e-08, "loss": 0.3663, "step": 13391 }, { "epoch": 0.9494505494505494, "grad_norm": 4.988492370395618, "learning_rate": 6.686713948157619e-08, "loss": 0.4046, "step": 13392 }, { "epoch": 0.9495214462956398, "grad_norm": 2.230174901343421, "learning_rate": 6.668012312685101e-08, "loss": 0.364, "step": 13393 }, { "epoch": 0.9495923431407303, "grad_norm": 1.7576567016421067, "learning_rate": 6.649336690984832e-08, "loss": 0.3743, "step": 13394 }, { "epoch": 0.9496632399858206, "grad_norm": 7.253899736148565, "learning_rate": 6.630687084041631e-08, "loss": 0.3802, "step": 13395 }, { "epoch": 0.949734136830911, "grad_norm": 4.01660598170745, "learning_rate": 6.612063492838827e-08, "loss": 0.3613, "step": 13396 }, { "epoch": 0.9498050336760014, "grad_norm": 2.6084446123568408, "learning_rate": 6.593465918358521e-08, "loss": 0.4198, "step": 13397 }, { "epoch": 0.9498759305210918, "grad_norm": 1.9776818868318564, "learning_rate": 6.574894361581374e-08, "loss": 0.3735, "step": 13398 }, { "epoch": 0.9499468273661822, "grad_norm": 6.142814038595987, "learning_rate": 6.556348823486658e-08, "loss": 0.4359, "step": 13399 }, { "epoch": 0.9500177242112726, "grad_norm": 2.9378800115679007, "learning_rate": 6.537829305052257e-08, "loss": 0.3952, "step": 13400 }, { "epoch": 0.9500886210563629, "grad_norm": 2.2319073310920508, "learning_rate": 6.519335807254834e-08, "loss": 0.3933, "step": 13401 }, { "epoch": 0.9501595179014534, "grad_norm": 2.607750123619575, "learning_rate": 6.500868331069444e-08, "loss": 0.375, "step": 13402 }, { "epoch": 0.9502304147465438, "grad_norm": 7.509869845729916, "learning_rate": 6.482426877469916e-08, "loss": 0.3607, "step": 13403 }, { "epoch": 0.9503013115916342, "grad_norm": 2.755264776805318, "learning_rate": 6.464011447428697e-08, "loss": 0.3738, "step": 13404 }, { "epoch": 0.9503722084367245, "grad_norm": 3.908615223255401, "learning_rate": 6.445622041916955e-08, "loss": 0.3948, "step": 13405 }, { "epoch": 0.950443105281815, "grad_norm": 3.1758564375457805, "learning_rate": 6.42725866190419e-08, "loss": 0.4489, "step": 13406 }, { "epoch": 0.9505140021269054, "grad_norm": 4.178271980090329, "learning_rate": 6.408921308358795e-08, "loss": 0.4232, "step": 13407 }, { "epoch": 0.9505848989719957, "grad_norm": 4.090667132869782, "learning_rate": 6.390609982247776e-08, "loss": 0.3995, "step": 13408 }, { "epoch": 0.9506557958170861, "grad_norm": 5.451633564311433, "learning_rate": 6.372324684536579e-08, "loss": 0.4108, "step": 13409 }, { "epoch": 0.9507266926621766, "grad_norm": 14.032696052406106, "learning_rate": 6.354065416189492e-08, "loss": 0.4631, "step": 13410 }, { "epoch": 0.9507975895072669, "grad_norm": 6.939389535555888, "learning_rate": 6.335832178169299e-08, "loss": 0.3993, "step": 13411 }, { "epoch": 0.9508684863523573, "grad_norm": 1.973618875194866, "learning_rate": 6.317624971437564e-08, "loss": 0.4098, "step": 13412 }, { "epoch": 0.9509393831974478, "grad_norm": 3.0488642203141327, "learning_rate": 6.299443796954185e-08, "loss": 0.3862, "step": 13413 }, { "epoch": 0.9510102800425381, "grad_norm": 3.2264353851839793, "learning_rate": 6.281288655678008e-08, "loss": 0.4024, "step": 13414 }, { "epoch": 0.9510811768876285, "grad_norm": 2.6377337938596774, "learning_rate": 6.263159548566322e-08, "loss": 0.348, "step": 13415 }, { "epoch": 0.9511520737327189, "grad_norm": 4.12676942857791, "learning_rate": 6.245056476575085e-08, "loss": 0.419, "step": 13416 }, { "epoch": 0.9512229705778092, "grad_norm": 2.918333560625321, "learning_rate": 6.226979440658809e-08, "loss": 0.3785, "step": 13417 }, { "epoch": 0.9512938674228997, "grad_norm": 8.377226658589837, "learning_rate": 6.208928441770845e-08, "loss": 0.3576, "step": 13418 }, { "epoch": 0.9513647642679901, "grad_norm": 3.0240780882071268, "learning_rate": 6.190903480862987e-08, "loss": 0.3852, "step": 13419 }, { "epoch": 0.9514356611130804, "grad_norm": 3.525372793004181, "learning_rate": 6.172904558885695e-08, "loss": 0.3894, "step": 13420 }, { "epoch": 0.9515065579581709, "grad_norm": 2.0045264947225268, "learning_rate": 6.154931676788046e-08, "loss": 0.3671, "step": 13421 }, { "epoch": 0.9515774548032613, "grad_norm": 2.829896192456069, "learning_rate": 6.136984835517778e-08, "loss": 0.4441, "step": 13422 }, { "epoch": 0.9516483516483516, "grad_norm": 2.124124432968594, "learning_rate": 6.11906403602125e-08, "loss": 0.3969, "step": 13423 }, { "epoch": 0.951719248493442, "grad_norm": 4.31248800514738, "learning_rate": 6.101169279243479e-08, "loss": 0.3764, "step": 13424 }, { "epoch": 0.9517901453385325, "grad_norm": 2.189296134213098, "learning_rate": 6.083300566127992e-08, "loss": 0.3896, "step": 13425 }, { "epoch": 0.9518610421836228, "grad_norm": 2.0602766575137803, "learning_rate": 6.065457897617089e-08, "loss": 0.3443, "step": 13426 }, { "epoch": 0.9519319390287132, "grad_norm": 3.13950504569475, "learning_rate": 6.047641274651572e-08, "loss": 0.3738, "step": 13427 }, { "epoch": 0.9520028358738036, "grad_norm": 3.534128575192455, "learning_rate": 6.029850698170969e-08, "loss": 0.3956, "step": 13428 }, { "epoch": 0.9520737327188941, "grad_norm": 3.331543639253344, "learning_rate": 6.012086169113306e-08, "loss": 0.4003, "step": 13429 }, { "epoch": 0.9521446295639844, "grad_norm": 1.0886886775469191, "learning_rate": 5.994347688415392e-08, "loss": 0.6883, "step": 13430 }, { "epoch": 0.9522155264090748, "grad_norm": 2.7582691622976747, "learning_rate": 5.976635257012642e-08, "loss": 0.3394, "step": 13431 }, { "epoch": 0.9522864232541652, "grad_norm": 1.9867138457961737, "learning_rate": 5.9589488758389767e-08, "loss": 0.3852, "step": 13432 }, { "epoch": 0.9523573200992556, "grad_norm": 2.237376878159046, "learning_rate": 5.941288545826984e-08, "loss": 0.3855, "step": 13433 }, { "epoch": 0.952428216944346, "grad_norm": 6.87158927213616, "learning_rate": 5.923654267907974e-08, "loss": 0.353, "step": 13434 }, { "epoch": 0.9524991137894364, "grad_norm": 2.1555732025922345, "learning_rate": 5.9060460430117595e-08, "loss": 0.4312, "step": 13435 }, { "epoch": 0.9525700106345267, "grad_norm": 3.5295488696217485, "learning_rate": 5.888463872066874e-08, "loss": 0.3626, "step": 13436 }, { "epoch": 0.9526409074796172, "grad_norm": 1.7940826203260827, "learning_rate": 5.87090775600041e-08, "loss": 0.3994, "step": 13437 }, { "epoch": 0.9527118043247076, "grad_norm": 5.976037989150239, "learning_rate": 5.853377695738127e-08, "loss": 0.3423, "step": 13438 }, { "epoch": 0.9527827011697979, "grad_norm": 2.7748608774659855, "learning_rate": 5.8358736922044526e-08, "loss": 0.3598, "step": 13439 }, { "epoch": 0.9528535980148883, "grad_norm": 4.144479747699645, "learning_rate": 5.818395746322314e-08, "loss": 0.3351, "step": 13440 }, { "epoch": 0.9529244948599788, "grad_norm": 2.266487073374523, "learning_rate": 5.800943859013364e-08, "loss": 0.3771, "step": 13441 }, { "epoch": 0.9529953917050691, "grad_norm": 2.3933341376387913, "learning_rate": 5.78351803119781e-08, "loss": 0.3987, "step": 13442 }, { "epoch": 0.9530662885501595, "grad_norm": 1.7682175302042185, "learning_rate": 5.76611826379464e-08, "loss": 0.3566, "step": 13443 }, { "epoch": 0.9531371853952499, "grad_norm": 1.8887433178459463, "learning_rate": 5.7487445577212306e-08, "loss": 0.3673, "step": 13444 }, { "epoch": 0.9532080822403403, "grad_norm": 3.731685182446286, "learning_rate": 5.7313969138937385e-08, "loss": 0.3709, "step": 13445 }, { "epoch": 0.9532789790854307, "grad_norm": 2.897536682080664, "learning_rate": 5.7140753332270424e-08, "loss": 0.3951, "step": 13446 }, { "epoch": 0.9533498759305211, "grad_norm": 2.4961374625505464, "learning_rate": 5.6967798166343566e-08, "loss": 0.3655, "step": 13447 }, { "epoch": 0.9534207727756114, "grad_norm": 3.392096872856035, "learning_rate": 5.6795103650277295e-08, "loss": 0.4191, "step": 13448 }, { "epoch": 0.9534916696207019, "grad_norm": 2.1469165717808036, "learning_rate": 5.6622669793178765e-08, "loss": 0.3832, "step": 13449 }, { "epoch": 0.9535625664657923, "grad_norm": 2.4874026519467716, "learning_rate": 5.645049660413959e-08, "loss": 0.3421, "step": 13450 }, { "epoch": 0.9536334633108826, "grad_norm": 2.184675449451661, "learning_rate": 5.627858409223974e-08, "loss": 0.3643, "step": 13451 }, { "epoch": 0.953704360155973, "grad_norm": 3.2883982796053792, "learning_rate": 5.6106932266542514e-08, "loss": 0.4055, "step": 13452 }, { "epoch": 0.9537752570010635, "grad_norm": 2.6292735975559696, "learning_rate": 5.5935541136100666e-08, "loss": 0.3468, "step": 13453 }, { "epoch": 0.9538461538461539, "grad_norm": 1.359626508999938, "learning_rate": 5.5764410709951425e-08, "loss": 0.6849, "step": 13454 }, { "epoch": 0.9539170506912442, "grad_norm": 3.8063047434161086, "learning_rate": 5.559354099711867e-08, "loss": 0.3572, "step": 13455 }, { "epoch": 0.9539879475363346, "grad_norm": 4.018738375467934, "learning_rate": 5.5422932006611864e-08, "loss": 0.3518, "step": 13456 }, { "epoch": 0.9540588443814251, "grad_norm": 4.328201743353288, "learning_rate": 5.525258374742826e-08, "loss": 0.3973, "step": 13457 }, { "epoch": 0.9541297412265154, "grad_norm": 2.70545202078125, "learning_rate": 5.508249622855011e-08, "loss": 0.3714, "step": 13458 }, { "epoch": 0.9542006380716058, "grad_norm": 2.9723458740280977, "learning_rate": 5.4912669458946354e-08, "loss": 0.4152, "step": 13459 }, { "epoch": 0.9542715349166963, "grad_norm": 1.0164235192839508, "learning_rate": 5.47431034475715e-08, "loss": 0.7026, "step": 13460 }, { "epoch": 0.9543424317617866, "grad_norm": 2.5033212969272602, "learning_rate": 5.457379820336783e-08, "loss": 0.3447, "step": 13461 }, { "epoch": 0.954413328606877, "grad_norm": 0.9603150830212224, "learning_rate": 5.4404753735262106e-08, "loss": 0.6649, "step": 13462 }, { "epoch": 0.9544842254519674, "grad_norm": 6.963461710585848, "learning_rate": 5.423597005216885e-08, "loss": 0.3741, "step": 13463 }, { "epoch": 0.9545551222970577, "grad_norm": 6.521325941480643, "learning_rate": 5.406744716298762e-08, "loss": 0.4108, "step": 13464 }, { "epoch": 0.9546260191421482, "grad_norm": 2.7523637328057498, "learning_rate": 5.3899185076605184e-08, "loss": 0.3588, "step": 13465 }, { "epoch": 0.9546969159872386, "grad_norm": 2.531886135503789, "learning_rate": 5.373118380189335e-08, "loss": 0.3964, "step": 13466 }, { "epoch": 0.9547678128323289, "grad_norm": 2.0974453081861704, "learning_rate": 5.3563443347711684e-08, "loss": 0.3644, "step": 13467 }, { "epoch": 0.9548387096774194, "grad_norm": 2.7897135137917553, "learning_rate": 5.339596372290534e-08, "loss": 0.3918, "step": 13468 }, { "epoch": 0.9549096065225098, "grad_norm": 2.0033248402668384, "learning_rate": 5.3228744936305585e-08, "loss": 0.3893, "step": 13469 }, { "epoch": 0.9549805033676001, "grad_norm": 3.799188419599319, "learning_rate": 5.306178699672926e-08, "loss": 0.4384, "step": 13470 }, { "epoch": 0.9550514002126905, "grad_norm": 2.3682334165697645, "learning_rate": 5.289508991298098e-08, "loss": 0.3261, "step": 13471 }, { "epoch": 0.955122297057781, "grad_norm": 2.519957125968388, "learning_rate": 5.272865369385094e-08, "loss": 0.3704, "step": 13472 }, { "epoch": 0.9551931939028713, "grad_norm": 3.320073248174442, "learning_rate": 5.2562478348114896e-08, "loss": 0.41, "step": 13473 }, { "epoch": 0.9552640907479617, "grad_norm": 3.2598118647744183, "learning_rate": 5.2396563884534734e-08, "loss": 0.3697, "step": 13474 }, { "epoch": 0.9553349875930521, "grad_norm": 3.003203529308843, "learning_rate": 5.2230910311860674e-08, "loss": 0.3688, "step": 13475 }, { "epoch": 0.9554058844381424, "grad_norm": 5.909997644578322, "learning_rate": 5.206551763882739e-08, "loss": 0.3559, "step": 13476 }, { "epoch": 0.9554767812832329, "grad_norm": 3.4240577887941397, "learning_rate": 5.190038587415625e-08, "loss": 0.3535, "step": 13477 }, { "epoch": 0.9555476781283233, "grad_norm": 2.407403389012, "learning_rate": 5.1735515026553605e-08, "loss": 0.4034, "step": 13478 }, { "epoch": 0.9556185749734137, "grad_norm": 1.8337790987230178, "learning_rate": 5.1570905104714744e-08, "loss": 0.4164, "step": 13479 }, { "epoch": 0.955689471818504, "grad_norm": 2.102338386036052, "learning_rate": 5.140655611731882e-08, "loss": 0.4044, "step": 13480 }, { "epoch": 0.9557603686635945, "grad_norm": 1.0664999615093496, "learning_rate": 5.124246807303224e-08, "loss": 0.6364, "step": 13481 }, { "epoch": 0.9558312655086849, "grad_norm": 4.107023790409798, "learning_rate": 5.107864098050752e-08, "loss": 0.3318, "step": 13482 }, { "epoch": 0.9559021623537752, "grad_norm": 2.5620904677092464, "learning_rate": 5.0915074848383316e-08, "loss": 0.382, "step": 13483 }, { "epoch": 0.9559730591988657, "grad_norm": 3.6969037623577243, "learning_rate": 5.0751769685284944e-08, "loss": 0.4216, "step": 13484 }, { "epoch": 0.9560439560439561, "grad_norm": 5.143071456791778, "learning_rate": 5.05887254998233e-08, "loss": 0.4596, "step": 13485 }, { "epoch": 0.9561148528890464, "grad_norm": 3.0611239132149586, "learning_rate": 5.0425942300595945e-08, "loss": 0.3853, "step": 13486 }, { "epoch": 0.9561857497341368, "grad_norm": 2.2382456341449655, "learning_rate": 5.026342009618601e-08, "loss": 0.4172, "step": 13487 }, { "epoch": 0.9562566465792273, "grad_norm": 2.1851507326863375, "learning_rate": 5.010115889516443e-08, "loss": 0.3836, "step": 13488 }, { "epoch": 0.9563275434243176, "grad_norm": 3.299138516391258, "learning_rate": 4.9939158706086564e-08, "loss": 0.4407, "step": 13489 }, { "epoch": 0.956398440269408, "grad_norm": 21.912082013072148, "learning_rate": 4.977741953749504e-08, "loss": 0.4007, "step": 13490 }, { "epoch": 0.9564693371144984, "grad_norm": 10.54971502595728, "learning_rate": 4.9615941397918586e-08, "loss": 0.4388, "step": 13491 }, { "epoch": 0.9565402339595888, "grad_norm": 1.92060787326266, "learning_rate": 4.9454724295872056e-08, "loss": 0.3691, "step": 13492 }, { "epoch": 0.9566111308046792, "grad_norm": 3.8001375573326714, "learning_rate": 4.9293768239856434e-08, "loss": 0.4006, "step": 13493 }, { "epoch": 0.9566820276497696, "grad_norm": 2.4628138531526154, "learning_rate": 4.913307323835881e-08, "loss": 0.3445, "step": 13494 }, { "epoch": 0.9567529244948599, "grad_norm": 2.9879558960350807, "learning_rate": 4.897263929985352e-08, "loss": 0.3329, "step": 13495 }, { "epoch": 0.9568238213399504, "grad_norm": 2.189895109585326, "learning_rate": 4.881246643279991e-08, "loss": 0.3983, "step": 13496 }, { "epoch": 0.9568947181850408, "grad_norm": 8.7749662559019, "learning_rate": 4.8652554645644004e-08, "loss": 0.3926, "step": 13497 }, { "epoch": 0.9569656150301311, "grad_norm": 2.430874835838801, "learning_rate": 4.849290394681738e-08, "loss": 0.3521, "step": 13498 }, { "epoch": 0.9570365118752215, "grad_norm": 2.9548251444446914, "learning_rate": 4.833351434474054e-08, "loss": 0.3742, "step": 13499 }, { "epoch": 0.957107408720312, "grad_norm": 2.6191071312639216, "learning_rate": 4.8174385847815644e-08, "loss": 0.3922, "step": 13500 }, { "epoch": 0.9571783055654023, "grad_norm": 2.9578483484638354, "learning_rate": 4.801551846443542e-08, "loss": 0.3982, "step": 13501 }, { "epoch": 0.9572492024104927, "grad_norm": 2.885817838927958, "learning_rate": 4.7856912202977054e-08, "loss": 0.3823, "step": 13502 }, { "epoch": 0.9573200992555831, "grad_norm": 1.916983071102845, "learning_rate": 4.7698567071803314e-08, "loss": 0.3899, "step": 13503 }, { "epoch": 0.9573909961006736, "grad_norm": 1.0496343910563133, "learning_rate": 4.754048307926362e-08, "loss": 0.6668, "step": 13504 }, { "epoch": 0.9574618929457639, "grad_norm": 3.0005409932463167, "learning_rate": 4.738266023369464e-08, "loss": 0.3842, "step": 13505 }, { "epoch": 0.9575327897908543, "grad_norm": 4.594542874203959, "learning_rate": 4.72250985434175e-08, "loss": 0.3802, "step": 13506 }, { "epoch": 0.9576036866359448, "grad_norm": 2.9913921155504606, "learning_rate": 4.706779801674222e-08, "loss": 0.4198, "step": 13507 }, { "epoch": 0.9576745834810351, "grad_norm": 2.079978135247388, "learning_rate": 4.69107586619616e-08, "loss": 0.3361, "step": 13508 }, { "epoch": 0.9577454803261255, "grad_norm": 2.354228372517575, "learning_rate": 4.675398048735735e-08, "loss": 0.3761, "step": 13509 }, { "epoch": 0.9578163771712159, "grad_norm": 4.048966079903029, "learning_rate": 4.659746350119676e-08, "loss": 0.3669, "step": 13510 }, { "epoch": 0.9578872740163062, "grad_norm": 3.981668020042385, "learning_rate": 4.644120771173211e-08, "loss": 0.3595, "step": 13511 }, { "epoch": 0.9579581708613967, "grad_norm": 2.738561014017078, "learning_rate": 4.628521312720402e-08, "loss": 0.3863, "step": 13512 }, { "epoch": 0.9580290677064871, "grad_norm": 2.161636295113279, "learning_rate": 4.612947975583704e-08, "loss": 0.4308, "step": 13513 }, { "epoch": 0.9580999645515774, "grad_norm": 3.030237926366276, "learning_rate": 4.597400760584403e-08, "loss": 0.4014, "step": 13514 }, { "epoch": 0.9581708613966679, "grad_norm": 2.3942039532199275, "learning_rate": 4.581879668542288e-08, "loss": 0.3944, "step": 13515 }, { "epoch": 0.9582417582417583, "grad_norm": 2.9217998423274407, "learning_rate": 4.5663847002757614e-08, "loss": 0.3612, "step": 13516 }, { "epoch": 0.9583126550868486, "grad_norm": 1.1402889467802364, "learning_rate": 4.550915856601945e-08, "loss": 0.6261, "step": 13517 }, { "epoch": 0.958383551931939, "grad_norm": 2.0925522512934385, "learning_rate": 4.5354731383365214e-08, "loss": 0.4471, "step": 13518 }, { "epoch": 0.9584544487770295, "grad_norm": 7.872758316288547, "learning_rate": 4.5200565462937274e-08, "loss": 0.3573, "step": 13519 }, { "epoch": 0.9585253456221198, "grad_norm": 2.9218761230305375, "learning_rate": 4.504666081286524e-08, "loss": 0.3872, "step": 13520 }, { "epoch": 0.9585962424672102, "grad_norm": 3.288170470924299, "learning_rate": 4.489301744126539e-08, "loss": 0.3407, "step": 13521 }, { "epoch": 0.9586671393123006, "grad_norm": 1.730000145275318, "learning_rate": 4.473963535623849e-08, "loss": 0.368, "step": 13522 }, { "epoch": 0.958738036157391, "grad_norm": 3.2124334336133016, "learning_rate": 4.458651456587249e-08, "loss": 0.3932, "step": 13523 }, { "epoch": 0.9588089330024814, "grad_norm": 2.090338954016072, "learning_rate": 4.4433655078242047e-08, "loss": 0.4337, "step": 13524 }, { "epoch": 0.9588798298475718, "grad_norm": 4.077395755070726, "learning_rate": 4.428105690140738e-08, "loss": 0.3754, "step": 13525 }, { "epoch": 0.9589507266926621, "grad_norm": 2.542228523751608, "learning_rate": 4.412872004341484e-08, "loss": 0.3437, "step": 13526 }, { "epoch": 0.9590216235377526, "grad_norm": 1.9720171104055453, "learning_rate": 4.3976644512298e-08, "loss": 0.3837, "step": 13527 }, { "epoch": 0.959092520382843, "grad_norm": 3.225537061501417, "learning_rate": 4.382483031607543e-08, "loss": 0.3504, "step": 13528 }, { "epoch": 0.9591634172279333, "grad_norm": 17.077831010910444, "learning_rate": 4.3673277462752405e-08, "loss": 0.4255, "step": 13529 }, { "epoch": 0.9592343140730237, "grad_norm": 31.742273256378763, "learning_rate": 4.3521985960319755e-08, "loss": 0.4076, "step": 13530 }, { "epoch": 0.9593052109181142, "grad_norm": 2.7126780156138395, "learning_rate": 4.3370955816756655e-08, "loss": 0.3214, "step": 13531 }, { "epoch": 0.9593761077632046, "grad_norm": 2.0262069004448793, "learning_rate": 4.3220187040026174e-08, "loss": 0.4069, "step": 13532 }, { "epoch": 0.9594470046082949, "grad_norm": 2.657341526202647, "learning_rate": 4.306967963807862e-08, "loss": 0.4301, "step": 13533 }, { "epoch": 0.9595179014533853, "grad_norm": 1.9295402900707055, "learning_rate": 4.291943361885043e-08, "loss": 0.3803, "step": 13534 }, { "epoch": 0.9595887982984758, "grad_norm": 3.011518817460933, "learning_rate": 4.2769448990263586e-08, "loss": 0.3748, "step": 13535 }, { "epoch": 0.9596596951435661, "grad_norm": 4.051576865935264, "learning_rate": 4.2619725760227326e-08, "loss": 0.4337, "step": 13536 }, { "epoch": 0.9597305919886565, "grad_norm": 6.198774681583598, "learning_rate": 4.2470263936637555e-08, "loss": 0.4413, "step": 13537 }, { "epoch": 0.9598014888337469, "grad_norm": 4.05447965147101, "learning_rate": 4.232106352737353e-08, "loss": 0.3327, "step": 13538 }, { "epoch": 0.9598723856788373, "grad_norm": 3.4021468048525025, "learning_rate": 4.2172124540304505e-08, "loss": 0.3641, "step": 13539 }, { "epoch": 0.9599432825239277, "grad_norm": 2.4144723748348618, "learning_rate": 4.2023446983283645e-08, "loss": 0.376, "step": 13540 }, { "epoch": 0.9600141793690181, "grad_norm": 7.032396266923617, "learning_rate": 4.187503086415079e-08, "loss": 0.3596, "step": 13541 }, { "epoch": 0.9600850762141084, "grad_norm": 4.610452674701658, "learning_rate": 4.1726876190731346e-08, "loss": 0.3931, "step": 13542 }, { "epoch": 0.9601559730591989, "grad_norm": 2.289258032220423, "learning_rate": 4.1578982970837955e-08, "loss": 0.4129, "step": 13543 }, { "epoch": 0.9602268699042893, "grad_norm": 8.903604999940196, "learning_rate": 4.1431351212270485e-08, "loss": 0.3668, "step": 13544 }, { "epoch": 0.9602977667493796, "grad_norm": 3.0849023510416433, "learning_rate": 4.1283980922811606e-08, "loss": 0.4368, "step": 13545 }, { "epoch": 0.96036866359447, "grad_norm": 2.399980991775055, "learning_rate": 4.113687211023343e-08, "loss": 0.3706, "step": 13546 }, { "epoch": 0.9604395604395605, "grad_norm": 4.756622457801916, "learning_rate": 4.099002478229308e-08, "loss": 0.4172, "step": 13547 }, { "epoch": 0.9605104572846508, "grad_norm": 2.4227471672184957, "learning_rate": 4.084343894673437e-08, "loss": 0.4043, "step": 13548 }, { "epoch": 0.9605813541297412, "grad_norm": 2.507729086431293, "learning_rate": 4.0697114611285004e-08, "loss": 0.3772, "step": 13549 }, { "epoch": 0.9606522509748316, "grad_norm": 3.0271259192745745, "learning_rate": 4.055105178366325e-08, "loss": 0.395, "step": 13550 }, { "epoch": 0.960723147819922, "grad_norm": 3.2475750271361763, "learning_rate": 4.040525047156907e-08, "loss": 0.3596, "step": 13551 }, { "epoch": 0.9607940446650124, "grad_norm": 3.654269043100525, "learning_rate": 4.025971068269186e-08, "loss": 0.3564, "step": 13552 }, { "epoch": 0.9608649415101028, "grad_norm": 3.376342725323071, "learning_rate": 4.011443242470603e-08, "loss": 0.344, "step": 13553 }, { "epoch": 0.9609358383551931, "grad_norm": 2.461701236225907, "learning_rate": 3.996941570527102e-08, "loss": 0.3903, "step": 13554 }, { "epoch": 0.9610067352002836, "grad_norm": 2.946342259708985, "learning_rate": 3.982466053203515e-08, "loss": 0.4415, "step": 13555 }, { "epoch": 0.961077632045374, "grad_norm": 3.46742189366167, "learning_rate": 3.9680166912630655e-08, "loss": 0.3825, "step": 13556 }, { "epoch": 0.9611485288904644, "grad_norm": 1.074063982920619, "learning_rate": 3.953593485467699e-08, "loss": 0.6254, "step": 13557 }, { "epoch": 0.9612194257355547, "grad_norm": 2.1368092588350325, "learning_rate": 3.9391964365779745e-08, "loss": 0.3886, "step": 13558 }, { "epoch": 0.9612903225806452, "grad_norm": 1.8594003607258032, "learning_rate": 3.9248255453530617e-08, "loss": 0.4062, "step": 13559 }, { "epoch": 0.9613612194257356, "grad_norm": 2.267769242187634, "learning_rate": 3.910480812550743e-08, "loss": 0.4042, "step": 13560 }, { "epoch": 0.9614321162708259, "grad_norm": 2.6393064617704654, "learning_rate": 3.896162238927359e-08, "loss": 0.4235, "step": 13561 }, { "epoch": 0.9615030131159163, "grad_norm": 5.884104642643234, "learning_rate": 3.881869825238027e-08, "loss": 0.3416, "step": 13562 }, { "epoch": 0.9615739099610068, "grad_norm": 3.785745621837283, "learning_rate": 3.8676035722364226e-08, "loss": 0.3917, "step": 13563 }, { "epoch": 0.9616448068060971, "grad_norm": 3.0958832423341494, "learning_rate": 3.853363480674721e-08, "loss": 0.3949, "step": 13564 }, { "epoch": 0.9617157036511875, "grad_norm": 3.013500992610383, "learning_rate": 3.8391495513038226e-08, "loss": 0.3633, "step": 13565 }, { "epoch": 0.961786600496278, "grad_norm": 4.8001937214861865, "learning_rate": 3.824961784873293e-08, "loss": 0.3781, "step": 13566 }, { "epoch": 0.9618574973413683, "grad_norm": 2.510210710535243, "learning_rate": 3.8108001821312577e-08, "loss": 0.3616, "step": 13567 }, { "epoch": 0.9619283941864587, "grad_norm": 3.2318952418455575, "learning_rate": 3.796664743824452e-08, "loss": 0.3989, "step": 13568 }, { "epoch": 0.9619992910315491, "grad_norm": 3.5354325402257567, "learning_rate": 3.782555470698224e-08, "loss": 0.3306, "step": 13569 }, { "epoch": 0.9620701878766394, "grad_norm": 7.793956361542555, "learning_rate": 3.768472363496589e-08, "loss": 0.4247, "step": 13570 }, { "epoch": 0.9621410847217299, "grad_norm": 2.8225921307831014, "learning_rate": 3.754415422962232e-08, "loss": 0.3522, "step": 13571 }, { "epoch": 0.9622119815668203, "grad_norm": 3.5788009501685223, "learning_rate": 3.740384649836226e-08, "loss": 0.3855, "step": 13572 }, { "epoch": 0.9622828784119106, "grad_norm": 2.4973821330319788, "learning_rate": 3.726380044858591e-08, "loss": 0.394, "step": 13573 }, { "epoch": 0.962353775257001, "grad_norm": 3.0516779278750352, "learning_rate": 3.71240160876768e-08, "loss": 0.3702, "step": 13574 }, { "epoch": 0.9624246721020915, "grad_norm": 2.0272583579096177, "learning_rate": 3.698449342300625e-08, "loss": 0.3565, "step": 13575 }, { "epoch": 0.9624955689471818, "grad_norm": 8.324395455775537, "learning_rate": 3.6845232461931725e-08, "loss": 0.3729, "step": 13576 }, { "epoch": 0.9625664657922722, "grad_norm": 2.799840730922362, "learning_rate": 3.670623321179623e-08, "loss": 0.3608, "step": 13577 }, { "epoch": 0.9626373626373627, "grad_norm": 1.8366024426031333, "learning_rate": 3.6567495679929455e-08, "loss": 0.3695, "step": 13578 }, { "epoch": 0.962708259482453, "grad_norm": 3.52845483149897, "learning_rate": 3.642901987364667e-08, "loss": 0.3718, "step": 13579 }, { "epoch": 0.9627791563275434, "grad_norm": 2.3087092482746554, "learning_rate": 3.629080580025035e-08, "loss": 0.3926, "step": 13580 }, { "epoch": 0.9628500531726338, "grad_norm": 2.5670784077066617, "learning_rate": 3.615285346702857e-08, "loss": 0.3638, "step": 13581 }, { "epoch": 0.9629209500177243, "grad_norm": 4.9756646520019014, "learning_rate": 3.601516288125551e-08, "loss": 0.3964, "step": 13582 }, { "epoch": 0.9629918468628146, "grad_norm": 3.3175807165964533, "learning_rate": 3.5877734050192016e-08, "loss": 0.3975, "step": 13583 }, { "epoch": 0.963062743707905, "grad_norm": 57.983159145075234, "learning_rate": 3.574056698108397e-08, "loss": 0.3867, "step": 13584 }, { "epoch": 0.9631336405529954, "grad_norm": 2.5446097017970017, "learning_rate": 3.560366168116558e-08, "loss": 0.3735, "step": 13585 }, { "epoch": 0.9632045373980858, "grad_norm": 4.007773350799314, "learning_rate": 3.546701815765496e-08, "loss": 0.4185, "step": 13586 }, { "epoch": 0.9632754342431762, "grad_norm": 2.221804441021839, "learning_rate": 3.533063641775747e-08, "loss": 0.3687, "step": 13587 }, { "epoch": 0.9633463310882666, "grad_norm": 3.751664454076803, "learning_rate": 3.5194516468664564e-08, "loss": 0.3838, "step": 13588 }, { "epoch": 0.9634172279333569, "grad_norm": 2.3851106391798154, "learning_rate": 3.5058658317555504e-08, "loss": 0.337, "step": 13589 }, { "epoch": 0.9634881247784474, "grad_norm": 2.4732276987729938, "learning_rate": 3.4923061971591787e-08, "loss": 0.3486, "step": 13590 }, { "epoch": 0.9635590216235378, "grad_norm": 4.203546698323872, "learning_rate": 3.478772743792491e-08, "loss": 0.3706, "step": 13591 }, { "epoch": 0.9636299184686281, "grad_norm": 2.4454107379420313, "learning_rate": 3.465265472369139e-08, "loss": 0.36, "step": 13592 }, { "epoch": 0.9637008153137185, "grad_norm": 1.73558480730073, "learning_rate": 3.451784383601331e-08, "loss": 0.3643, "step": 13593 }, { "epoch": 0.963771712158809, "grad_norm": 2.9592409376324307, "learning_rate": 3.438329478199887e-08, "loss": 0.4336, "step": 13594 }, { "epoch": 0.9638426090038993, "grad_norm": 2.7000489773438945, "learning_rate": 3.42490075687435e-08, "loss": 0.3505, "step": 13595 }, { "epoch": 0.9639135058489897, "grad_norm": 3.8595296678472066, "learning_rate": 3.4114982203328207e-08, "loss": 0.3752, "step": 13596 }, { "epoch": 0.9639844026940801, "grad_norm": 4.627132089806962, "learning_rate": 3.398121869282012e-08, "loss": 0.3883, "step": 13597 }, { "epoch": 0.9640552995391705, "grad_norm": 2.758835407687275, "learning_rate": 3.3847717044273035e-08, "loss": 0.3847, "step": 13598 }, { "epoch": 0.9641261963842609, "grad_norm": 8.917307095197996, "learning_rate": 3.3714477264725765e-08, "loss": 0.3659, "step": 13599 }, { "epoch": 0.9641970932293513, "grad_norm": 5.659661105644643, "learning_rate": 3.358149936120547e-08, "loss": 0.3991, "step": 13600 }, { "epoch": 0.9642679900744416, "grad_norm": 3.3683959642405843, "learning_rate": 3.344878334072266e-08, "loss": 0.3897, "step": 13601 }, { "epoch": 0.9643388869195321, "grad_norm": 2.2512029214821445, "learning_rate": 3.3316329210276724e-08, "loss": 0.4221, "step": 13602 }, { "epoch": 0.9644097837646225, "grad_norm": 3.3862235226501407, "learning_rate": 3.3184136976851524e-08, "loss": 0.3613, "step": 13603 }, { "epoch": 0.9644806806097128, "grad_norm": 2.192654045980162, "learning_rate": 3.305220664741759e-08, "loss": 0.3727, "step": 13604 }, { "epoch": 0.9645515774548032, "grad_norm": 2.4602154163152092, "learning_rate": 3.292053822893215e-08, "loss": 0.3714, "step": 13605 }, { "epoch": 0.9646224742998937, "grad_norm": 2.27009357545534, "learning_rate": 3.278913172833798e-08, "loss": 0.3405, "step": 13606 }, { "epoch": 0.9646933711449841, "grad_norm": 3.503403725775974, "learning_rate": 3.265798715256396e-08, "loss": 0.4023, "step": 13607 }, { "epoch": 0.9647642679900744, "grad_norm": 1.9495542075132442, "learning_rate": 3.252710450852625e-08, "loss": 0.4039, "step": 13608 }, { "epoch": 0.9648351648351648, "grad_norm": 3.1111750478308036, "learning_rate": 3.2396483803124876e-08, "loss": 0.3821, "step": 13609 }, { "epoch": 0.9649060616802553, "grad_norm": 3.3874174329083755, "learning_rate": 3.226612504324933e-08, "loss": 0.347, "step": 13610 }, { "epoch": 0.9649769585253456, "grad_norm": 3.092511648870534, "learning_rate": 3.213602823577244e-08, "loss": 0.3853, "step": 13611 }, { "epoch": 0.965047855370436, "grad_norm": 2.2312063672381703, "learning_rate": 3.200619338755429e-08, "loss": 0.3301, "step": 13612 }, { "epoch": 0.9651187522155265, "grad_norm": 4.136655419398041, "learning_rate": 3.18766205054416e-08, "loss": 0.4055, "step": 13613 }, { "epoch": 0.9651896490606168, "grad_norm": 2.677647734106075, "learning_rate": 3.174730959626671e-08, "loss": 0.3418, "step": 13614 }, { "epoch": 0.9652605459057072, "grad_norm": 2.3656731948733882, "learning_rate": 3.161826066684859e-08, "loss": 0.3772, "step": 13615 }, { "epoch": 0.9653314427507976, "grad_norm": 2.4182127463704908, "learning_rate": 3.14894737239918e-08, "loss": 0.3808, "step": 13616 }, { "epoch": 0.965402339595888, "grad_norm": 3.2775912023516023, "learning_rate": 3.136094877448703e-08, "loss": 0.3818, "step": 13617 }, { "epoch": 0.9654732364409784, "grad_norm": 7.938859881092702, "learning_rate": 3.123268582511163e-08, "loss": 0.3457, "step": 13618 }, { "epoch": 0.9655441332860688, "grad_norm": 6.180840990993821, "learning_rate": 3.110468488262963e-08, "loss": 0.4055, "step": 13619 }, { "epoch": 0.9656150301311591, "grad_norm": 2.214383665774968, "learning_rate": 3.097694595379008e-08, "loss": 0.3917, "step": 13620 }, { "epoch": 0.9656859269762496, "grad_norm": 2.127973003230454, "learning_rate": 3.0849469045328705e-08, "loss": 0.3995, "step": 13621 }, { "epoch": 0.96575682382134, "grad_norm": 2.9566998244312805, "learning_rate": 3.07222541639679e-08, "loss": 0.3903, "step": 13622 }, { "epoch": 0.9658277206664303, "grad_norm": 5.943522023803472, "learning_rate": 3.0595301316415084e-08, "loss": 0.4121, "step": 13623 }, { "epoch": 0.9658986175115207, "grad_norm": 35.80341851693396, "learning_rate": 3.046861050936545e-08, "loss": 0.3883, "step": 13624 }, { "epoch": 0.9659695143566112, "grad_norm": 2.5787113450299244, "learning_rate": 3.0342181749498655e-08, "loss": 0.369, "step": 13625 }, { "epoch": 0.9660404112017015, "grad_norm": 3.1715311955360654, "learning_rate": 3.021601504348215e-08, "loss": 0.3793, "step": 13626 }, { "epoch": 0.9661113080467919, "grad_norm": 2.3809955107540413, "learning_rate": 3.009011039796839e-08, "loss": 0.3034, "step": 13627 }, { "epoch": 0.9661822048918823, "grad_norm": 2.4207759730882197, "learning_rate": 2.9964467819595945e-08, "loss": 0.3753, "step": 13628 }, { "epoch": 0.9662531017369727, "grad_norm": 7.159084365436711, "learning_rate": 2.983908731499119e-08, "loss": 0.3826, "step": 13629 }, { "epoch": 0.9663239985820631, "grad_norm": 1.8414890266527235, "learning_rate": 2.971396889076439e-08, "loss": 0.3648, "step": 13630 }, { "epoch": 0.9663948954271535, "grad_norm": 5.475370822206053, "learning_rate": 2.9589112553514154e-08, "loss": 0.3934, "step": 13631 }, { "epoch": 0.9664657922722439, "grad_norm": 2.8520009647202538, "learning_rate": 2.9464518309823554e-08, "loss": 0.3607, "step": 13632 }, { "epoch": 0.9665366891173343, "grad_norm": 2.5364961034718836, "learning_rate": 2.934018616626233e-08, "loss": 0.3654, "step": 13633 }, { "epoch": 0.9666075859624247, "grad_norm": 2.530416803106305, "learning_rate": 2.9216116129388017e-08, "loss": 0.3521, "step": 13634 }, { "epoch": 0.9666784828075151, "grad_norm": 3.9567452415592412, "learning_rate": 2.9092308205740937e-08, "loss": 0.4453, "step": 13635 }, { "epoch": 0.9667493796526054, "grad_norm": 2.4943141309435686, "learning_rate": 2.896876240185087e-08, "loss": 0.3868, "step": 13636 }, { "epoch": 0.9668202764976959, "grad_norm": 1.9982637860310293, "learning_rate": 2.88454787242326e-08, "loss": 0.3459, "step": 13637 }, { "epoch": 0.9668911733427863, "grad_norm": 11.568379411045102, "learning_rate": 2.8722457179386488e-08, "loss": 0.3664, "step": 13638 }, { "epoch": 0.9669620701878766, "grad_norm": 3.5681599731036995, "learning_rate": 2.8599697773799005e-08, "loss": 0.3568, "step": 13639 }, { "epoch": 0.967032967032967, "grad_norm": 3.450821038987954, "learning_rate": 2.8477200513944426e-08, "loss": 0.3751, "step": 13640 }, { "epoch": 0.9671038638780575, "grad_norm": 0.9386608025675216, "learning_rate": 2.8354965406281464e-08, "loss": 0.6418, "step": 13641 }, { "epoch": 0.9671747607231478, "grad_norm": 5.798770280014338, "learning_rate": 2.8232992457256082e-08, "loss": 0.3755, "step": 13642 }, { "epoch": 0.9672456575682382, "grad_norm": 11.584376337654586, "learning_rate": 2.811128167329924e-08, "loss": 0.3595, "step": 13643 }, { "epoch": 0.9673165544133286, "grad_norm": 33.080216091263786, "learning_rate": 2.7989833060829695e-08, "loss": 0.4041, "step": 13644 }, { "epoch": 0.967387451258419, "grad_norm": 7.310202082677523, "learning_rate": 2.786864662625177e-08, "loss": 0.3905, "step": 13645 }, { "epoch": 0.9674583481035094, "grad_norm": 1.5878948872834806, "learning_rate": 2.7747722375954246e-08, "loss": 0.3414, "step": 13646 }, { "epoch": 0.9675292449485998, "grad_norm": 2.512214408721387, "learning_rate": 2.7627060316315348e-08, "loss": 0.344, "step": 13647 }, { "epoch": 0.9676001417936901, "grad_norm": 2.231397797570761, "learning_rate": 2.75066604536961e-08, "loss": 0.4479, "step": 13648 }, { "epoch": 0.9676710386387806, "grad_norm": 1.915610141143032, "learning_rate": 2.7386522794446424e-08, "loss": 0.4028, "step": 13649 }, { "epoch": 0.967741935483871, "grad_norm": 2.6796449965553717, "learning_rate": 2.7266647344900697e-08, "loss": 0.3924, "step": 13650 }, { "epoch": 0.9678128323289613, "grad_norm": 3.768380250743667, "learning_rate": 2.7147034111379978e-08, "loss": 0.3658, "step": 13651 }, { "epoch": 0.9678837291740517, "grad_norm": 4.550153903050674, "learning_rate": 2.7027683100191438e-08, "loss": 0.4127, "step": 13652 }, { "epoch": 0.9679546260191422, "grad_norm": 3.572271094185448, "learning_rate": 2.6908594317630042e-08, "loss": 0.3527, "step": 13653 }, { "epoch": 0.9680255228642325, "grad_norm": 2.629417276105126, "learning_rate": 2.6789767769972995e-08, "loss": 0.3485, "step": 13654 }, { "epoch": 0.9680964197093229, "grad_norm": 2.2613680756381553, "learning_rate": 2.667120346348806e-08, "loss": 0.3387, "step": 13655 }, { "epoch": 0.9681673165544133, "grad_norm": 2.63688819956979, "learning_rate": 2.6552901404426344e-08, "loss": 0.3922, "step": 13656 }, { "epoch": 0.9682382133995037, "grad_norm": 1.0932943124302525, "learning_rate": 2.6434861599026197e-08, "loss": 0.6686, "step": 13657 }, { "epoch": 0.9683091102445941, "grad_norm": 2.4685617621668516, "learning_rate": 2.6317084053512078e-08, "loss": 0.3623, "step": 13658 }, { "epoch": 0.9683800070896845, "grad_norm": 2.2818643284975217, "learning_rate": 2.619956877409402e-08, "loss": 0.3985, "step": 13659 }, { "epoch": 0.968450903934775, "grad_norm": 3.5627599577278724, "learning_rate": 2.608231576696929e-08, "loss": 0.4144, "step": 13660 }, { "epoch": 0.9685218007798653, "grad_norm": 2.9282106523448235, "learning_rate": 2.5965325038320722e-08, "loss": 0.3801, "step": 13661 }, { "epoch": 0.9685926976249557, "grad_norm": 2.648557408272835, "learning_rate": 2.584859659431671e-08, "loss": 0.4022, "step": 13662 }, { "epoch": 0.9686635944700461, "grad_norm": 2.6281059232574457, "learning_rate": 2.5732130441112892e-08, "loss": 0.3988, "step": 13663 }, { "epoch": 0.9687344913151364, "grad_norm": 5.641869805981173, "learning_rate": 2.5615926584850457e-08, "loss": 0.4032, "step": 13664 }, { "epoch": 0.9688053881602269, "grad_norm": 5.749590060924393, "learning_rate": 2.5499985031657292e-08, "loss": 0.3862, "step": 13665 }, { "epoch": 0.9688762850053173, "grad_norm": 2.441042014684303, "learning_rate": 2.5384305787646278e-08, "loss": 0.4117, "step": 13666 }, { "epoch": 0.9689471818504076, "grad_norm": 1.0745437730381155, "learning_rate": 2.5268888858918094e-08, "loss": 0.6223, "step": 13667 }, { "epoch": 0.969018078695498, "grad_norm": 2.3569232178817314, "learning_rate": 2.5153734251558425e-08, "loss": 0.4178, "step": 13668 }, { "epoch": 0.9690889755405885, "grad_norm": 3.3897715214511, "learning_rate": 2.503884197163964e-08, "loss": 0.4656, "step": 13669 }, { "epoch": 0.9691598723856788, "grad_norm": 1.0603953772822174, "learning_rate": 2.4924212025219662e-08, "loss": 0.6937, "step": 13670 }, { "epoch": 0.9692307692307692, "grad_norm": 2.079873499010514, "learning_rate": 2.480984441834311e-08, "loss": 0.3619, "step": 13671 }, { "epoch": 0.9693016660758597, "grad_norm": 3.174929093156603, "learning_rate": 2.4695739157041266e-08, "loss": 0.3755, "step": 13672 }, { "epoch": 0.96937256292095, "grad_norm": 5.074510316983064, "learning_rate": 2.458189624733043e-08, "loss": 0.3654, "step": 13673 }, { "epoch": 0.9694434597660404, "grad_norm": 2.992824813066851, "learning_rate": 2.4468315695213573e-08, "loss": 0.3566, "step": 13674 }, { "epoch": 0.9695143566111308, "grad_norm": 10.856993263238769, "learning_rate": 2.4354997506680357e-08, "loss": 0.3891, "step": 13675 }, { "epoch": 0.9695852534562212, "grad_norm": 8.914417871776642, "learning_rate": 2.4241941687705994e-08, "loss": 0.3634, "step": 13676 }, { "epoch": 0.9696561503013116, "grad_norm": 24.528715836956344, "learning_rate": 2.4129148244251276e-08, "loss": 0.3708, "step": 13677 }, { "epoch": 0.969727047146402, "grad_norm": 10.671944514802798, "learning_rate": 2.4016617182264777e-08, "loss": 0.3858, "step": 13678 }, { "epoch": 0.9697979439914923, "grad_norm": 1.967509735747705, "learning_rate": 2.3904348507680088e-08, "loss": 0.378, "step": 13679 }, { "epoch": 0.9698688408365828, "grad_norm": 2.8887528019132818, "learning_rate": 2.3792342226417464e-08, "loss": 0.3347, "step": 13680 }, { "epoch": 0.9699397376816732, "grad_norm": 3.2323690224164072, "learning_rate": 2.368059834438219e-08, "loss": 0.3965, "step": 13681 }, { "epoch": 0.9700106345267635, "grad_norm": 2.748215961896629, "learning_rate": 2.356911686746788e-08, "loss": 0.3414, "step": 13682 }, { "epoch": 0.9700815313718539, "grad_norm": 2.6847596571888124, "learning_rate": 2.3457897801552053e-08, "loss": 0.3287, "step": 13683 }, { "epoch": 0.9701524282169444, "grad_norm": 1.8809010387497513, "learning_rate": 2.334694115249947e-08, "loss": 0.3929, "step": 13684 }, { "epoch": 0.9702233250620348, "grad_norm": 0.9127766100973104, "learning_rate": 2.3236246926161554e-08, "loss": 0.5408, "step": 13685 }, { "epoch": 0.9702942219071251, "grad_norm": 2.571123874687733, "learning_rate": 2.312581512837475e-08, "loss": 0.4185, "step": 13686 }, { "epoch": 0.9703651187522155, "grad_norm": 2.453995351373854, "learning_rate": 2.3015645764962736e-08, "loss": 0.3628, "step": 13687 }, { "epoch": 0.970436015597306, "grad_norm": 3.309853603016762, "learning_rate": 2.2905738841733637e-08, "loss": 0.4069, "step": 13688 }, { "epoch": 0.9705069124423963, "grad_norm": 2.198283205368158, "learning_rate": 2.279609436448449e-08, "loss": 0.3854, "step": 13689 }, { "epoch": 0.9705778092874867, "grad_norm": 3.601521403414337, "learning_rate": 2.2686712338996218e-08, "loss": 0.3596, "step": 13690 }, { "epoch": 0.9706487061325771, "grad_norm": 2.3546578850343183, "learning_rate": 2.2577592771036438e-08, "loss": 0.4117, "step": 13691 }, { "epoch": 0.9707196029776675, "grad_norm": 2.316393528276101, "learning_rate": 2.2468735666359432e-08, "loss": 0.3912, "step": 13692 }, { "epoch": 0.9707904998227579, "grad_norm": 2.210580663575585, "learning_rate": 2.2360141030704496e-08, "loss": 0.3604, "step": 13693 }, { "epoch": 0.9708613966678483, "grad_norm": 1.8024302623381017, "learning_rate": 2.2251808869799275e-08, "loss": 0.3627, "step": 13694 }, { "epoch": 0.9709322935129386, "grad_norm": 1.0481668934452102, "learning_rate": 2.2143739189355306e-08, "loss": 0.6043, "step": 13695 }, { "epoch": 0.9710031903580291, "grad_norm": 1.8558420459994918, "learning_rate": 2.2035931995071368e-08, "loss": 0.3515, "step": 13696 }, { "epoch": 0.9710740872031195, "grad_norm": 5.354336907949236, "learning_rate": 2.1928387292632357e-08, "loss": 0.3695, "step": 13697 }, { "epoch": 0.9711449840482098, "grad_norm": 4.090653151592574, "learning_rate": 2.1821105087709293e-08, "loss": 0.3848, "step": 13698 }, { "epoch": 0.9712158808933002, "grad_norm": 3.5654367785290795, "learning_rate": 2.171408538595876e-08, "loss": 0.3924, "step": 13699 }, { "epoch": 0.9712867777383907, "grad_norm": 2.6674644961162035, "learning_rate": 2.1607328193024024e-08, "loss": 0.3756, "step": 13700 }, { "epoch": 0.971357674583481, "grad_norm": 3.2567615593785932, "learning_rate": 2.1500833514535026e-08, "loss": 0.4242, "step": 13701 }, { "epoch": 0.9714285714285714, "grad_norm": 7.138239847476546, "learning_rate": 2.1394601356107268e-08, "loss": 0.3735, "step": 13702 }, { "epoch": 0.9714994682736618, "grad_norm": 4.022263442567783, "learning_rate": 2.128863172334128e-08, "loss": 0.4077, "step": 13703 }, { "epoch": 0.9715703651187522, "grad_norm": 2.198360736133017, "learning_rate": 2.1182924621826472e-08, "loss": 0.3867, "step": 13704 }, { "epoch": 0.9716412619638426, "grad_norm": 2.530049267987343, "learning_rate": 2.107748005713617e-08, "loss": 0.3376, "step": 13705 }, { "epoch": 0.971712158808933, "grad_norm": 1.6983312674613151, "learning_rate": 2.0972298034830362e-08, "loss": 0.369, "step": 13706 }, { "epoch": 0.9717830556540233, "grad_norm": 5.347446876304135, "learning_rate": 2.086737856045573e-08, "loss": 0.3498, "step": 13707 }, { "epoch": 0.9718539524991138, "grad_norm": 2.5633318709230863, "learning_rate": 2.076272163954507e-08, "loss": 0.4074, "step": 13708 }, { "epoch": 0.9719248493442042, "grad_norm": 5.783026132487666, "learning_rate": 2.0658327277616187e-08, "loss": 0.3267, "step": 13709 }, { "epoch": 0.9719957461892946, "grad_norm": 2.5572555506284806, "learning_rate": 2.0554195480174123e-08, "loss": 0.412, "step": 13710 }, { "epoch": 0.972066643034385, "grad_norm": 2.724705005762428, "learning_rate": 2.045032625271004e-08, "loss": 0.3558, "step": 13711 }, { "epoch": 0.9721375398794754, "grad_norm": 2.5229655429067717, "learning_rate": 2.0346719600700672e-08, "loss": 0.4077, "step": 13712 }, { "epoch": 0.9722084367245658, "grad_norm": 19.36218930432296, "learning_rate": 2.0243375529609976e-08, "loss": 0.3896, "step": 13713 }, { "epoch": 0.9722793335696561, "grad_norm": 2.005989362804283, "learning_rate": 2.014029404488693e-08, "loss": 0.3415, "step": 13714 }, { "epoch": 0.9723502304147466, "grad_norm": 3.0027925992070865, "learning_rate": 2.003747515196719e-08, "loss": 0.333, "step": 13715 }, { "epoch": 0.972421127259837, "grad_norm": 4.588998886017439, "learning_rate": 1.993491885627197e-08, "loss": 0.3748, "step": 13716 }, { "epoch": 0.9724920241049273, "grad_norm": 9.894915069285702, "learning_rate": 1.983262516321027e-08, "loss": 0.3731, "step": 13717 }, { "epoch": 0.9725629209500177, "grad_norm": 5.205361746743145, "learning_rate": 1.9730594078175013e-08, "loss": 0.4271, "step": 13718 }, { "epoch": 0.9726338177951082, "grad_norm": 4.269620792728895, "learning_rate": 1.9628825606546887e-08, "loss": 0.3626, "step": 13719 }, { "epoch": 0.9727047146401985, "grad_norm": 2.0417365875624927, "learning_rate": 1.9527319753692152e-08, "loss": 0.3482, "step": 13720 }, { "epoch": 0.9727756114852889, "grad_norm": 5.47849340546481, "learning_rate": 1.94260765249632e-08, "loss": 0.3519, "step": 13721 }, { "epoch": 0.9728465083303793, "grad_norm": 2.463684962995116, "learning_rate": 1.9325095925699087e-08, "loss": 0.4021, "step": 13722 }, { "epoch": 0.9729174051754697, "grad_norm": 2.5542150920198203, "learning_rate": 1.9224377961223896e-08, "loss": 0.3827, "step": 13723 }, { "epoch": 0.9729883020205601, "grad_norm": 4.28204165495906, "learning_rate": 1.9123922636848922e-08, "loss": 0.3254, "step": 13724 }, { "epoch": 0.9730591988656505, "grad_norm": 3.6618420074012348, "learning_rate": 1.902372995787105e-08, "loss": 0.431, "step": 13725 }, { "epoch": 0.9731300957107408, "grad_norm": 1.022017259587597, "learning_rate": 1.8923799929573826e-08, "loss": 0.6948, "step": 13726 }, { "epoch": 0.9732009925558313, "grad_norm": 2.410022039014832, "learning_rate": 1.8824132557226927e-08, "loss": 0.3819, "step": 13727 }, { "epoch": 0.9732718894009217, "grad_norm": 2.5976082103914964, "learning_rate": 1.8724727846085034e-08, "loss": 0.3988, "step": 13728 }, { "epoch": 0.973342786246012, "grad_norm": 2.5567650266506377, "learning_rate": 1.862558580139062e-08, "loss": 0.3837, "step": 13729 }, { "epoch": 0.9734136830911024, "grad_norm": 2.1854716237527096, "learning_rate": 1.852670642837062e-08, "loss": 0.3701, "step": 13730 }, { "epoch": 0.9734845799361929, "grad_norm": 2.8872680018850416, "learning_rate": 1.8428089732240306e-08, "loss": 0.3809, "step": 13731 }, { "epoch": 0.9735554767812832, "grad_norm": 3.637925725748109, "learning_rate": 1.8329735718198294e-08, "loss": 0.3638, "step": 13732 }, { "epoch": 0.9736263736263736, "grad_norm": 2.0149350806469304, "learning_rate": 1.8231644391432103e-08, "loss": 0.3181, "step": 13733 }, { "epoch": 0.973697270471464, "grad_norm": 4.264304068140363, "learning_rate": 1.8133815757113706e-08, "loss": 0.3786, "step": 13734 }, { "epoch": 0.9737681673165545, "grad_norm": 2.0795781303862233, "learning_rate": 1.8036249820401752e-08, "loss": 0.4095, "step": 13735 }, { "epoch": 0.9738390641616448, "grad_norm": 6.555278338541481, "learning_rate": 1.793894658644102e-08, "loss": 0.402, "step": 13736 }, { "epoch": 0.9739099610067352, "grad_norm": 3.333956047831174, "learning_rate": 1.7841906060361846e-08, "loss": 0.3712, "step": 13737 }, { "epoch": 0.9739808578518256, "grad_norm": 2.7827128424865433, "learning_rate": 1.7745128247281805e-08, "loss": 0.3762, "step": 13738 }, { "epoch": 0.974051754696916, "grad_norm": 3.3754699373204815, "learning_rate": 1.764861315230404e-08, "loss": 0.3641, "step": 13739 }, { "epoch": 0.9741226515420064, "grad_norm": 2.349728029205113, "learning_rate": 1.755236078051781e-08, "loss": 0.3344, "step": 13740 }, { "epoch": 0.9741935483870968, "grad_norm": 4.6972349507718825, "learning_rate": 1.7456371136998497e-08, "loss": 0.435, "step": 13741 }, { "epoch": 0.9742644452321871, "grad_norm": 2.6647745607875217, "learning_rate": 1.736064422680761e-08, "loss": 0.3826, "step": 13742 }, { "epoch": 0.9743353420772776, "grad_norm": 2.5260695206205077, "learning_rate": 1.7265180054993337e-08, "loss": 0.3718, "step": 13743 }, { "epoch": 0.974406238922368, "grad_norm": 2.6931743541579536, "learning_rate": 1.7169978626588868e-08, "loss": 0.349, "step": 13744 }, { "epoch": 0.9744771357674583, "grad_norm": 4.647829621435967, "learning_rate": 1.7075039946614635e-08, "loss": 0.3807, "step": 13745 }, { "epoch": 0.9745480326125487, "grad_norm": 3.549967773615805, "learning_rate": 1.698036402007719e-08, "loss": 0.4046, "step": 13746 }, { "epoch": 0.9746189294576392, "grad_norm": 1.0058916946950638, "learning_rate": 1.6885950851968092e-08, "loss": 0.6198, "step": 13747 }, { "epoch": 0.9746898263027295, "grad_norm": 2.089569473162889, "learning_rate": 1.6791800447266693e-08, "loss": 0.3681, "step": 13748 }, { "epoch": 0.9747607231478199, "grad_norm": 2.088044129404202, "learning_rate": 1.66979128109368e-08, "loss": 0.356, "step": 13749 }, { "epoch": 0.9748316199929103, "grad_norm": 5.669240111827427, "learning_rate": 1.660428794793001e-08, "loss": 0.3903, "step": 13750 }, { "epoch": 0.9749025168380007, "grad_norm": 2.3605055791928247, "learning_rate": 1.651092586318237e-08, "loss": 0.4259, "step": 13751 }, { "epoch": 0.9749734136830911, "grad_norm": 6.475582476830836, "learning_rate": 1.6417826561617168e-08, "loss": 0.3812, "step": 13752 }, { "epoch": 0.9750443105281815, "grad_norm": 2.8177084948964235, "learning_rate": 1.6324990048144362e-08, "loss": 0.4173, "step": 13753 }, { "epoch": 0.9751152073732718, "grad_norm": 3.4538737134804363, "learning_rate": 1.6232416327658373e-08, "loss": 0.3714, "step": 13754 }, { "epoch": 0.9751861042183623, "grad_norm": 2.2110302497683816, "learning_rate": 1.614010540504085e-08, "loss": 0.4023, "step": 13755 }, { "epoch": 0.9752570010634527, "grad_norm": 2.77372400323016, "learning_rate": 1.6048057285160124e-08, "loss": 0.435, "step": 13756 }, { "epoch": 0.975327897908543, "grad_norm": 2.487271889928733, "learning_rate": 1.595627197286842e-08, "loss": 0.3635, "step": 13757 }, { "epoch": 0.9753987947536334, "grad_norm": 3.6268498380129026, "learning_rate": 1.5864749473007424e-08, "loss": 0.4065, "step": 13758 }, { "epoch": 0.9754696915987239, "grad_norm": 2.4536314070948393, "learning_rate": 1.5773489790402163e-08, "loss": 0.4146, "step": 13759 }, { "epoch": 0.9755405884438143, "grad_norm": 3.329908759840934, "learning_rate": 1.5682492929865456e-08, "loss": 0.4309, "step": 13760 }, { "epoch": 0.9756114852889046, "grad_norm": 4.319072875701399, "learning_rate": 1.5591758896194575e-08, "loss": 0.4083, "step": 13761 }, { "epoch": 0.975682382133995, "grad_norm": 2.6819171259249037, "learning_rate": 1.5501287694175138e-08, "loss": 0.3568, "step": 13762 }, { "epoch": 0.9757532789790855, "grad_norm": 4.0922860486864545, "learning_rate": 1.5411079328577216e-08, "loss": 0.3949, "step": 13763 }, { "epoch": 0.9758241758241758, "grad_norm": 3.7972310793021062, "learning_rate": 1.532113380415756e-08, "loss": 0.4235, "step": 13764 }, { "epoch": 0.9758950726692662, "grad_norm": 3.406475866535495, "learning_rate": 1.52314511256596e-08, "loss": 0.4036, "step": 13765 }, { "epoch": 0.9759659695143567, "grad_norm": 9.619330639263302, "learning_rate": 1.5142031297811223e-08, "loss": 0.4404, "step": 13766 }, { "epoch": 0.976036866359447, "grad_norm": 2.839998858699673, "learning_rate": 1.5052874325328648e-08, "loss": 0.4027, "step": 13767 }, { "epoch": 0.9761077632045374, "grad_norm": 4.715996930306832, "learning_rate": 1.4963980212912565e-08, "loss": 0.3825, "step": 13768 }, { "epoch": 0.9761786600496278, "grad_norm": 2.478193103314847, "learning_rate": 1.4875348965251446e-08, "loss": 0.3946, "step": 13769 }, { "epoch": 0.9762495568947182, "grad_norm": 3.4974382186595543, "learning_rate": 1.4786980587017662e-08, "loss": 0.42, "step": 13770 }, { "epoch": 0.9763204537398086, "grad_norm": 4.51783349094667, "learning_rate": 1.4698875082871377e-08, "loss": 0.3929, "step": 13771 }, { "epoch": 0.976391350584899, "grad_norm": 3.861096238648495, "learning_rate": 1.461103245745832e-08, "loss": 0.3967, "step": 13772 }, { "epoch": 0.9764622474299893, "grad_norm": 3.4040694869994637, "learning_rate": 1.4523452715411446e-08, "loss": 0.3735, "step": 13773 }, { "epoch": 0.9765331442750798, "grad_norm": 8.381785620672721, "learning_rate": 1.4436135861347067e-08, "loss": 0.456, "step": 13774 }, { "epoch": 0.9766040411201702, "grad_norm": 4.512861803163515, "learning_rate": 1.4349081899871497e-08, "loss": 0.4583, "step": 13775 }, { "epoch": 0.9766749379652605, "grad_norm": 1.7290109450581963, "learning_rate": 1.4262290835573844e-08, "loss": 0.3659, "step": 13776 }, { "epoch": 0.9767458348103509, "grad_norm": 4.738220232990185, "learning_rate": 1.4175762673031001e-08, "loss": 0.4077, "step": 13777 }, { "epoch": 0.9768167316554414, "grad_norm": 2.757823532257356, "learning_rate": 1.4089497416805431e-08, "loss": 0.3796, "step": 13778 }, { "epoch": 0.9768876285005317, "grad_norm": 3.4074108920668005, "learning_rate": 1.4003495071446826e-08, "loss": 0.3431, "step": 13779 }, { "epoch": 0.9769585253456221, "grad_norm": 1.9228524433368226, "learning_rate": 1.3917755641489339e-08, "loss": 0.3587, "step": 13780 }, { "epoch": 0.9770294221907125, "grad_norm": 4.735059171201493, "learning_rate": 1.383227913145435e-08, "loss": 0.3559, "step": 13781 }, { "epoch": 0.9771003190358029, "grad_norm": 2.435736964072987, "learning_rate": 1.3747065545848815e-08, "loss": 0.4162, "step": 13782 }, { "epoch": 0.9771712158808933, "grad_norm": 3.49895299331733, "learning_rate": 1.3662114889166356e-08, "loss": 0.3568, "step": 13783 }, { "epoch": 0.9772421127259837, "grad_norm": 2.7741425737618295, "learning_rate": 1.3577427165886724e-08, "loss": 0.4019, "step": 13784 }, { "epoch": 0.977313009571074, "grad_norm": 3.8196735823684556, "learning_rate": 1.349300238047524e-08, "loss": 0.4068, "step": 13785 }, { "epoch": 0.9773839064161645, "grad_norm": 3.1660171053978794, "learning_rate": 1.3408840537383893e-08, "loss": 0.39, "step": 13786 }, { "epoch": 0.9774548032612549, "grad_norm": 7.157487397469424, "learning_rate": 1.3324941641050249e-08, "loss": 0.4025, "step": 13787 }, { "epoch": 0.9775257001063453, "grad_norm": 2.811262752339016, "learning_rate": 1.3241305695898543e-08, "loss": 0.3893, "step": 13788 }, { "epoch": 0.9775965969514356, "grad_norm": 4.199767996728984, "learning_rate": 1.3157932706339138e-08, "loss": 0.3832, "step": 13789 }, { "epoch": 0.9776674937965261, "grad_norm": 1.9764415744998776, "learning_rate": 1.3074822676767962e-08, "loss": 0.4085, "step": 13790 }, { "epoch": 0.9777383906416165, "grad_norm": 2.151082055017336, "learning_rate": 1.2991975611568176e-08, "loss": 0.3923, "step": 13791 }, { "epoch": 0.9778092874867068, "grad_norm": 2.375137019186037, "learning_rate": 1.290939151510795e-08, "loss": 0.3751, "step": 13792 }, { "epoch": 0.9778801843317972, "grad_norm": 3.250551442758732, "learning_rate": 1.282707039174158e-08, "loss": 0.4115, "step": 13793 }, { "epoch": 0.9779510811768877, "grad_norm": 1.6498547609262273, "learning_rate": 1.2745012245810596e-08, "loss": 0.3884, "step": 13794 }, { "epoch": 0.978021978021978, "grad_norm": 4.005341872532222, "learning_rate": 1.2663217081641532e-08, "loss": 0.3979, "step": 13795 }, { "epoch": 0.9780928748670684, "grad_norm": 2.492654470711437, "learning_rate": 1.2581684903547609e-08, "loss": 0.4099, "step": 13796 }, { "epoch": 0.9781637717121588, "grad_norm": 1.8132721898913866, "learning_rate": 1.2500415715828718e-08, "loss": 0.38, "step": 13797 }, { "epoch": 0.9782346685572492, "grad_norm": 4.050957432084815, "learning_rate": 1.241940952276921e-08, "loss": 0.3853, "step": 13798 }, { "epoch": 0.9783055654023396, "grad_norm": 2.2980845053784362, "learning_rate": 1.2338666328641225e-08, "loss": 0.4115, "step": 13799 }, { "epoch": 0.97837646224743, "grad_norm": 2.0167519024827247, "learning_rate": 1.225818613770191e-08, "loss": 0.4151, "step": 13800 }, { "epoch": 0.9784473590925203, "grad_norm": 3.408708413656375, "learning_rate": 1.2177968954196207e-08, "loss": 0.4272, "step": 13801 }, { "epoch": 0.9785182559376108, "grad_norm": 5.7226383769217, "learning_rate": 1.2098014782352396e-08, "loss": 0.4026, "step": 13802 }, { "epoch": 0.9785891527827012, "grad_norm": 2.5386460709302767, "learning_rate": 1.2018323626387663e-08, "loss": 0.3611, "step": 13803 }, { "epoch": 0.9786600496277915, "grad_norm": 3.303779350019887, "learning_rate": 1.1938895490504198e-08, "loss": 0.3554, "step": 13804 }, { "epoch": 0.978730946472882, "grad_norm": 1.9939501174308012, "learning_rate": 1.1859730378889767e-08, "loss": 0.3924, "step": 13805 }, { "epoch": 0.9788018433179724, "grad_norm": 6.14805229180065, "learning_rate": 1.1780828295718804e-08, "loss": 0.3674, "step": 13806 }, { "epoch": 0.9788727401630627, "grad_norm": 3.421580038833483, "learning_rate": 1.1702189245152428e-08, "loss": 0.3489, "step": 13807 }, { "epoch": 0.9789436370081531, "grad_norm": 2.0213467523324002, "learning_rate": 1.1623813231336767e-08, "loss": 0.3392, "step": 13808 }, { "epoch": 0.9790145338532436, "grad_norm": 2.3307688712853243, "learning_rate": 1.154570025840518e-08, "loss": 0.3635, "step": 13809 }, { "epoch": 0.9790854306983339, "grad_norm": 2.6273427055123446, "learning_rate": 1.1467850330476038e-08, "loss": 0.3556, "step": 13810 }, { "epoch": 0.9791563275434243, "grad_norm": 4.333738909870112, "learning_rate": 1.1390263451654949e-08, "loss": 0.3702, "step": 13811 }, { "epoch": 0.9792272243885147, "grad_norm": 3.1154241918629952, "learning_rate": 1.1312939626032526e-08, "loss": 0.4157, "step": 13812 }, { "epoch": 0.9792981212336052, "grad_norm": 4.932753855675955, "learning_rate": 1.1235878857686621e-08, "loss": 0.385, "step": 13813 }, { "epoch": 0.9793690180786955, "grad_norm": 3.5599136344746807, "learning_rate": 1.115908115068065e-08, "loss": 0.383, "step": 13814 }, { "epoch": 0.9794399149237859, "grad_norm": 2.7015045530388946, "learning_rate": 1.1082546509064707e-08, "loss": 0.3871, "step": 13815 }, { "epoch": 0.9795108117688763, "grad_norm": 3.679077079390222, "learning_rate": 1.100627493687334e-08, "loss": 0.4587, "step": 13816 }, { "epoch": 0.9795817086139667, "grad_norm": 6.138455404820929, "learning_rate": 1.0930266438128889e-08, "loss": 0.3674, "step": 13817 }, { "epoch": 0.9796526054590571, "grad_norm": 5.975513642012442, "learning_rate": 1.0854521016839814e-08, "loss": 0.3938, "step": 13818 }, { "epoch": 0.9797235023041475, "grad_norm": 4.8849944566440096, "learning_rate": 1.0779038676999587e-08, "loss": 0.3264, "step": 13819 }, { "epoch": 0.9797943991492378, "grad_norm": 3.9030002776769415, "learning_rate": 1.0703819422588913e-08, "loss": 0.4249, "step": 13820 }, { "epoch": 0.9798652959943283, "grad_norm": 2.6971655729717527, "learning_rate": 1.0628863257574062e-08, "loss": 0.3888, "step": 13821 }, { "epoch": 0.9799361928394187, "grad_norm": 1.97846484498325, "learning_rate": 1.0554170185907431e-08, "loss": 0.3912, "step": 13822 }, { "epoch": 0.980007089684509, "grad_norm": 2.034022169977493, "learning_rate": 1.0479740211527535e-08, "loss": 0.3857, "step": 13823 }, { "epoch": 0.9800779865295994, "grad_norm": 2.464884130145501, "learning_rate": 1.0405573338359565e-08, "loss": 0.3591, "step": 13824 }, { "epoch": 0.9801488833746899, "grad_norm": 3.093846024731189, "learning_rate": 1.0331669570313729e-08, "loss": 0.3893, "step": 13825 }, { "epoch": 0.9802197802197802, "grad_norm": 3.1231904302153417, "learning_rate": 1.025802891128802e-08, "loss": 0.386, "step": 13826 }, { "epoch": 0.9802906770648706, "grad_norm": 2.2886326249900395, "learning_rate": 1.0184651365164334e-08, "loss": 0.4046, "step": 13827 }, { "epoch": 0.980361573909961, "grad_norm": 3.9985624053002105, "learning_rate": 1.0111536935812905e-08, "loss": 0.4017, "step": 13828 }, { "epoch": 0.9804324707550514, "grad_norm": 4.216578770811241, "learning_rate": 1.0038685627088428e-08, "loss": 0.3721, "step": 13829 }, { "epoch": 0.9805033676001418, "grad_norm": 4.889169184108709, "learning_rate": 9.966097442832834e-09, "loss": 0.3638, "step": 13830 }, { "epoch": 0.9805742644452322, "grad_norm": 2.4296306813512194, "learning_rate": 9.893772386873613e-09, "loss": 0.382, "step": 13831 }, { "epoch": 0.9806451612903225, "grad_norm": 2.33963194999144, "learning_rate": 9.82171046302438e-09, "loss": 0.3754, "step": 13832 }, { "epoch": 0.980716058135413, "grad_norm": 3.6308398245937363, "learning_rate": 9.749911675085432e-09, "loss": 0.427, "step": 13833 }, { "epoch": 0.9807869549805034, "grad_norm": 5.017177525391985, "learning_rate": 9.678376026842628e-09, "loss": 0.4332, "step": 13834 }, { "epoch": 0.9808578518255937, "grad_norm": 10.078963045267153, "learning_rate": 9.60710352206795e-09, "loss": 0.346, "step": 13835 }, { "epoch": 0.9809287486706841, "grad_norm": 3.7744587562321157, "learning_rate": 9.536094164519504e-09, "loss": 0.3384, "step": 13836 }, { "epoch": 0.9809996455157746, "grad_norm": 3.2398959325314602, "learning_rate": 9.465347957942072e-09, "loss": 0.3614, "step": 13837 }, { "epoch": 0.981070542360865, "grad_norm": 2.740453232638714, "learning_rate": 9.394864906066004e-09, "loss": 0.3617, "step": 13838 }, { "epoch": 0.9811414392059553, "grad_norm": 2.1727545576690135, "learning_rate": 9.324645012607214e-09, "loss": 0.353, "step": 13839 }, { "epoch": 0.9812123360510457, "grad_norm": 2.0614744066467794, "learning_rate": 9.254688281269963e-09, "loss": 0.3528, "step": 13840 }, { "epoch": 0.9812832328961362, "grad_norm": 5.0806638448189, "learning_rate": 9.184994715741857e-09, "loss": 0.4071, "step": 13841 }, { "epoch": 0.9813541297412265, "grad_norm": 2.2083103927662413, "learning_rate": 9.11556431969829e-09, "loss": 0.3488, "step": 13842 }, { "epoch": 0.9814250265863169, "grad_norm": 2.9700473814729804, "learning_rate": 9.046397096800218e-09, "loss": 0.3769, "step": 13843 }, { "epoch": 0.9814959234314073, "grad_norm": 2.912404919452179, "learning_rate": 8.977493050695286e-09, "loss": 0.4024, "step": 13844 }, { "epoch": 0.9815668202764977, "grad_norm": 3.184186129925968, "learning_rate": 8.908852185016692e-09, "loss": 0.4127, "step": 13845 }, { "epoch": 0.9816377171215881, "grad_norm": 3.4157379002367487, "learning_rate": 8.840474503383767e-09, "loss": 0.3934, "step": 13846 }, { "epoch": 0.9817086139666785, "grad_norm": 2.3314804975338936, "learning_rate": 8.772360009402514e-09, "loss": 0.3873, "step": 13847 }, { "epoch": 0.9817795108117688, "grad_norm": 4.143357573401539, "learning_rate": 8.70450870666395e-09, "loss": 0.3514, "step": 13848 }, { "epoch": 0.9818504076568593, "grad_norm": 2.5284939402483877, "learning_rate": 8.636920598746323e-09, "loss": 0.3654, "step": 13849 }, { "epoch": 0.9819213045019497, "grad_norm": 2.1638228134227218, "learning_rate": 8.56959568921345e-09, "loss": 0.3542, "step": 13850 }, { "epoch": 0.98199220134704, "grad_norm": 5.782237796911165, "learning_rate": 8.502533981615824e-09, "loss": 0.4211, "step": 13851 }, { "epoch": 0.9820630981921304, "grad_norm": 3.8889588457712647, "learning_rate": 8.435735479489504e-09, "loss": 0.408, "step": 13852 }, { "epoch": 0.9821339950372209, "grad_norm": 4.696497068339736, "learning_rate": 8.369200186356675e-09, "loss": 0.3539, "step": 13853 }, { "epoch": 0.9822048918823112, "grad_norm": 4.302954310686548, "learning_rate": 8.302928105725639e-09, "loss": 0.3282, "step": 13854 }, { "epoch": 0.9822757887274016, "grad_norm": 6.503965137176424, "learning_rate": 8.236919241091378e-09, "loss": 0.3844, "step": 13855 }, { "epoch": 0.982346685572492, "grad_norm": 2.7439536744307222, "learning_rate": 8.171173595934445e-09, "loss": 0.3921, "step": 13856 }, { "epoch": 0.9824175824175824, "grad_norm": 2.3599940469858516, "learning_rate": 8.105691173721508e-09, "loss": 0.3976, "step": 13857 }, { "epoch": 0.9824884792626728, "grad_norm": 2.225968800025123, "learning_rate": 8.040471977905918e-09, "loss": 0.3723, "step": 13858 }, { "epoch": 0.9825593761077632, "grad_norm": 3.4015628402203455, "learning_rate": 7.975516011926033e-09, "loss": 0.3743, "step": 13859 }, { "epoch": 0.9826302729528535, "grad_norm": 6.830289057803105, "learning_rate": 7.910823279207448e-09, "loss": 0.4159, "step": 13860 }, { "epoch": 0.982701169797944, "grad_norm": 8.227073457949675, "learning_rate": 7.846393783161876e-09, "loss": 0.3502, "step": 13861 }, { "epoch": 0.9827720666430344, "grad_norm": 4.047722855150864, "learning_rate": 7.782227527186048e-09, "loss": 0.3669, "step": 13862 }, { "epoch": 0.9828429634881248, "grad_norm": 2.2398365302438163, "learning_rate": 7.718324514663921e-09, "loss": 0.3485, "step": 13863 }, { "epoch": 0.9829138603332152, "grad_norm": 1.0951191731649002, "learning_rate": 7.65468474896447e-09, "loss": 0.7106, "step": 13864 }, { "epoch": 0.9829847571783056, "grad_norm": 3.5196015552634656, "learning_rate": 7.591308233443895e-09, "loss": 0.3707, "step": 13865 }, { "epoch": 0.983055654023396, "grad_norm": 2.350895211452495, "learning_rate": 7.528194971445079e-09, "loss": 0.3834, "step": 13866 }, { "epoch": 0.9831265508684863, "grad_norm": 3.061110969686258, "learning_rate": 7.465344966294253e-09, "loss": 0.3446, "step": 13867 }, { "epoch": 0.9831974477135768, "grad_norm": 2.755590220435814, "learning_rate": 7.402758221307094e-09, "loss": 0.3856, "step": 13868 }, { "epoch": 0.9832683445586672, "grad_norm": 2.674658173062498, "learning_rate": 7.3404347397826315e-09, "loss": 0.3372, "step": 13869 }, { "epoch": 0.9833392414037575, "grad_norm": 2.6911785897273974, "learning_rate": 7.278374525007681e-09, "loss": 0.424, "step": 13870 }, { "epoch": 0.9834101382488479, "grad_norm": 3.559968488006248, "learning_rate": 7.216577580255179e-09, "loss": 0.3826, "step": 13871 }, { "epoch": 0.9834810350939384, "grad_norm": 3.144719058722874, "learning_rate": 7.155043908783632e-09, "loss": 0.3371, "step": 13872 }, { "epoch": 0.9835519319390287, "grad_norm": 1.8675389882814069, "learning_rate": 7.09377351383711e-09, "loss": 0.4049, "step": 13873 }, { "epoch": 0.9836228287841191, "grad_norm": 2.6803621359073473, "learning_rate": 7.032766398646918e-09, "loss": 0.387, "step": 13874 }, { "epoch": 0.9836937256292095, "grad_norm": 4.57825580325399, "learning_rate": 6.9720225664299254e-09, "loss": 0.3685, "step": 13875 }, { "epoch": 0.9837646224742999, "grad_norm": 5.012478879125402, "learning_rate": 6.911542020388573e-09, "loss": 0.3755, "step": 13876 }, { "epoch": 0.9838355193193903, "grad_norm": 5.995513806390957, "learning_rate": 6.851324763713085e-09, "loss": 0.3506, "step": 13877 }, { "epoch": 0.9839064161644807, "grad_norm": 2.0983313202289633, "learning_rate": 6.7913707995787005e-09, "loss": 0.3819, "step": 13878 }, { "epoch": 0.983977313009571, "grad_norm": 5.079572733992615, "learning_rate": 6.731680131146223e-09, "loss": 0.4142, "step": 13879 }, { "epoch": 0.9840482098546615, "grad_norm": 1.86357278038067, "learning_rate": 6.672252761563691e-09, "loss": 0.4122, "step": 13880 }, { "epoch": 0.9841191066997519, "grad_norm": 2.3905237746539187, "learning_rate": 6.613088693964154e-09, "loss": 0.4085, "step": 13881 }, { "epoch": 0.9841900035448422, "grad_norm": 1.0610189385426325, "learning_rate": 6.5541879314678926e-09, "loss": 0.6906, "step": 13882 }, { "epoch": 0.9842609003899326, "grad_norm": 6.0948074933338905, "learning_rate": 6.495550477180202e-09, "loss": 0.4117, "step": 13883 }, { "epoch": 0.9843317972350231, "grad_norm": 2.019773840049691, "learning_rate": 6.437176334194162e-09, "loss": 0.3456, "step": 13884 }, { "epoch": 0.9844026940801134, "grad_norm": 3.9486270881810626, "learning_rate": 6.379065505586757e-09, "loss": 0.4022, "step": 13885 }, { "epoch": 0.9844735909252038, "grad_norm": 1.0771159766036964, "learning_rate": 6.321217994422757e-09, "loss": 0.6452, "step": 13886 }, { "epoch": 0.9845444877702942, "grad_norm": 4.504130471492585, "learning_rate": 6.2636338037525005e-09, "loss": 0.3899, "step": 13887 }, { "epoch": 0.9846153846153847, "grad_norm": 1.9539269939062938, "learning_rate": 6.2063129366124465e-09, "loss": 0.4017, "step": 13888 }, { "epoch": 0.984686281460475, "grad_norm": 9.456678840575016, "learning_rate": 6.149255396025178e-09, "loss": 0.3923, "step": 13889 }, { "epoch": 0.9847571783055654, "grad_norm": 4.586708396436504, "learning_rate": 6.092461184999399e-09, "loss": 0.3863, "step": 13890 }, { "epoch": 0.9848280751506558, "grad_norm": 1.8201725638986777, "learning_rate": 6.03593030652938e-09, "loss": 0.4079, "step": 13891 }, { "epoch": 0.9848989719957462, "grad_norm": 2.781457053663783, "learning_rate": 5.979662763596628e-09, "loss": 0.4064, "step": 13892 }, { "epoch": 0.9849698688408366, "grad_norm": 3.2234294765404625, "learning_rate": 5.92365855916821e-09, "loss": 0.4355, "step": 13893 }, { "epoch": 0.985040765685927, "grad_norm": 2.195383984194511, "learning_rate": 5.867917696196768e-09, "loss": 0.3881, "step": 13894 }, { "epoch": 0.9851116625310173, "grad_norm": 6.203392313071921, "learning_rate": 5.8124401776221695e-09, "loss": 0.3471, "step": 13895 }, { "epoch": 0.9851825593761078, "grad_norm": 1.9560744892654423, "learning_rate": 5.757226006369299e-09, "loss": 0.4241, "step": 13896 }, { "epoch": 0.9852534562211982, "grad_norm": 4.956692071502931, "learning_rate": 5.70227518535027e-09, "loss": 0.3815, "step": 13897 }, { "epoch": 0.9853243530662885, "grad_norm": 3.364441502778, "learning_rate": 5.647587717461656e-09, "loss": 0.3666, "step": 13898 }, { "epoch": 0.985395249911379, "grad_norm": 3.676754570626519, "learning_rate": 5.593163605587814e-09, "loss": 0.363, "step": 13899 }, { "epoch": 0.9854661467564694, "grad_norm": 2.2498761068267483, "learning_rate": 5.539002852598674e-09, "loss": 0.3795, "step": 13900 }, { "epoch": 0.9855370436015597, "grad_norm": 3.910786391532955, "learning_rate": 5.485105461350282e-09, "loss": 0.3689, "step": 13901 }, { "epoch": 0.9856079404466501, "grad_norm": 3.129481624311721, "learning_rate": 5.431471434684254e-09, "loss": 0.3785, "step": 13902 }, { "epoch": 0.9856788372917406, "grad_norm": 1.607575946839372, "learning_rate": 5.3781007754288846e-09, "loss": 0.3838, "step": 13903 }, { "epoch": 0.9857497341368309, "grad_norm": 2.9211131792071376, "learning_rate": 5.324993486398588e-09, "loss": 0.3669, "step": 13904 }, { "epoch": 0.9858206309819213, "grad_norm": 5.7671812037336725, "learning_rate": 5.272149570393348e-09, "loss": 0.4042, "step": 13905 }, { "epoch": 0.9858915278270117, "grad_norm": 5.411099519867758, "learning_rate": 5.219569030199823e-09, "loss": 0.3984, "step": 13906 }, { "epoch": 0.985962424672102, "grad_norm": 2.0559624583417166, "learning_rate": 5.16725186859135e-09, "loss": 0.3847, "step": 13907 }, { "epoch": 0.9860333215171925, "grad_norm": 4.1244053257756175, "learning_rate": 5.115198088325724e-09, "loss": 0.3755, "step": 13908 }, { "epoch": 0.9861042183622829, "grad_norm": 3.229356405238062, "learning_rate": 5.06340769214797e-09, "loss": 0.3688, "step": 13909 }, { "epoch": 0.9861751152073732, "grad_norm": 4.858019573827676, "learning_rate": 5.011880682789794e-09, "loss": 0.4407, "step": 13910 }, { "epoch": 0.9862460120524636, "grad_norm": 3.6470136967650792, "learning_rate": 4.960617062967355e-09, "loss": 0.3878, "step": 13911 }, { "epoch": 0.9863169088975541, "grad_norm": 1.1347038608501272, "learning_rate": 4.909616835384045e-09, "loss": 0.6821, "step": 13912 }, { "epoch": 0.9863878057426444, "grad_norm": 0.9905209311486511, "learning_rate": 4.858880002728827e-09, "loss": 0.6174, "step": 13913 }, { "epoch": 0.9864587025877348, "grad_norm": 2.7453619197395405, "learning_rate": 4.8084065676778924e-09, "loss": 0.4159, "step": 13914 }, { "epoch": 0.9865295994328253, "grad_norm": 10.201470220414677, "learning_rate": 4.758196532892445e-09, "loss": 0.3478, "step": 13915 }, { "epoch": 0.9866004962779157, "grad_norm": 2.135932993472358, "learning_rate": 4.708249901019812e-09, "loss": 0.3625, "step": 13916 }, { "epoch": 0.986671393123006, "grad_norm": 3.2967204588667953, "learning_rate": 4.658566674693999e-09, "loss": 0.3754, "step": 13917 }, { "epoch": 0.9867422899680964, "grad_norm": 5.782456590535425, "learning_rate": 4.609146856534574e-09, "loss": 0.3588, "step": 13918 }, { "epoch": 0.9868131868131869, "grad_norm": 2.15600234115949, "learning_rate": 4.559990449147788e-09, "loss": 0.3442, "step": 13919 }, { "epoch": 0.9868840836582772, "grad_norm": 2.1090045300385265, "learning_rate": 4.511097455125457e-09, "loss": 0.4007, "step": 13920 }, { "epoch": 0.9869549805033676, "grad_norm": 1.0195713825293937, "learning_rate": 4.462467877046073e-09, "loss": 0.6743, "step": 13921 }, { "epoch": 0.987025877348458, "grad_norm": 1.9009703259269937, "learning_rate": 4.414101717473141e-09, "loss": 0.3904, "step": 13922 }, { "epoch": 0.9870967741935484, "grad_norm": 1.916338982640754, "learning_rate": 4.365998978957953e-09, "loss": 0.3684, "step": 13923 }, { "epoch": 0.9871676710386388, "grad_norm": 2.5816675371720934, "learning_rate": 4.318159664036814e-09, "loss": 0.3855, "step": 13924 }, { "epoch": 0.9872385678837292, "grad_norm": 3.1716415449038418, "learning_rate": 4.270583775231596e-09, "loss": 0.3561, "step": 13925 }, { "epoch": 0.9873094647288195, "grad_norm": 4.496169483899141, "learning_rate": 4.223271315051958e-09, "loss": 0.3514, "step": 13926 }, { "epoch": 0.98738036157391, "grad_norm": 5.839636176859919, "learning_rate": 4.176222285992571e-09, "loss": 0.4085, "step": 13927 }, { "epoch": 0.9874512584190004, "grad_norm": 2.18782918717487, "learning_rate": 4.129436690533673e-09, "loss": 0.3694, "step": 13928 }, { "epoch": 0.9875221552640907, "grad_norm": 6.064199557904675, "learning_rate": 4.082914531142734e-09, "loss": 0.3876, "step": 13929 }, { "epoch": 0.9875930521091811, "grad_norm": 17.102880560917512, "learning_rate": 4.036655810273349e-09, "loss": 0.3952, "step": 13930 }, { "epoch": 0.9876639489542716, "grad_norm": 3.7678155962456255, "learning_rate": 3.990660530364121e-09, "loss": 0.4034, "step": 13931 }, { "epoch": 0.9877348457993619, "grad_norm": 4.1633466267366375, "learning_rate": 3.944928693840333e-09, "loss": 0.3712, "step": 13932 }, { "epoch": 0.9878057426444523, "grad_norm": 2.3253358715930554, "learning_rate": 3.8994603031145e-09, "loss": 0.3833, "step": 13933 }, { "epoch": 0.9878766394895427, "grad_norm": 7.821777353093405, "learning_rate": 3.854255360583037e-09, "loss": 0.3877, "step": 13934 }, { "epoch": 0.9879475363346331, "grad_norm": 4.1207898162150425, "learning_rate": 3.809313868629594e-09, "loss": 0.3626, "step": 13935 }, { "epoch": 0.9880184331797235, "grad_norm": 3.378844207804095, "learning_rate": 3.764635829625052e-09, "loss": 0.3616, "step": 13936 }, { "epoch": 0.9880893300248139, "grad_norm": 3.1545236351250376, "learning_rate": 3.720221245924749e-09, "loss": 0.4089, "step": 13937 }, { "epoch": 0.9881602268699042, "grad_norm": 2.1628102760973076, "learning_rate": 3.6760701198707003e-09, "loss": 0.4349, "step": 13938 }, { "epoch": 0.9882311237149947, "grad_norm": 3.653584593905469, "learning_rate": 3.632182453790489e-09, "loss": 0.3811, "step": 13939 }, { "epoch": 0.9883020205600851, "grad_norm": 0.9619253872181899, "learning_rate": 3.5885582499994852e-09, "loss": 0.6722, "step": 13940 }, { "epoch": 0.9883729174051755, "grad_norm": 2.691986590472399, "learning_rate": 3.5451975107969604e-09, "loss": 0.3408, "step": 13941 }, { "epoch": 0.9884438142502658, "grad_norm": 3.3741002877788353, "learning_rate": 3.5021002384699743e-09, "loss": 0.4377, "step": 13942 }, { "epoch": 0.9885147110953563, "grad_norm": 2.169998716446183, "learning_rate": 3.459266435290598e-09, "loss": 0.3808, "step": 13943 }, { "epoch": 0.9885856079404467, "grad_norm": 8.051490859016365, "learning_rate": 3.4166961035181356e-09, "loss": 0.3868, "step": 13944 }, { "epoch": 0.988656504785537, "grad_norm": 5.58828320498023, "learning_rate": 3.374389245396903e-09, "loss": 0.4352, "step": 13945 }, { "epoch": 0.9887274016306274, "grad_norm": 8.55906101278955, "learning_rate": 3.3323458631578932e-09, "loss": 0.3956, "step": 13946 }, { "epoch": 0.9887982984757179, "grad_norm": 2.8488033357166067, "learning_rate": 3.2905659590176663e-09, "loss": 0.3764, "step": 13947 }, { "epoch": 0.9888691953208082, "grad_norm": 1.90756591295526, "learning_rate": 3.2490495351800153e-09, "loss": 0.3512, "step": 13948 }, { "epoch": 0.9889400921658986, "grad_norm": 3.030372713294231, "learning_rate": 3.2077965938343e-09, "loss": 0.4169, "step": 13949 }, { "epoch": 0.989010989010989, "grad_norm": 2.080715701070826, "learning_rate": 3.1668071371548927e-09, "loss": 0.3583, "step": 13950 }, { "epoch": 0.9890818858560794, "grad_norm": 2.3559718320743652, "learning_rate": 3.1260811673033962e-09, "loss": 0.3903, "step": 13951 }, { "epoch": 0.9891527827011698, "grad_norm": 9.60483108879235, "learning_rate": 3.0856186864280936e-09, "loss": 0.4006, "step": 13952 }, { "epoch": 0.9892236795462602, "grad_norm": 5.717374525254459, "learning_rate": 3.0454196966617223e-09, "loss": 0.4599, "step": 13953 }, { "epoch": 0.9892945763913505, "grad_norm": 3.733232006566429, "learning_rate": 3.0054842001248087e-09, "loss": 0.4267, "step": 13954 }, { "epoch": 0.989365473236441, "grad_norm": 2.645973338057285, "learning_rate": 2.9658121989223354e-09, "loss": 0.3576, "step": 13955 }, { "epoch": 0.9894363700815314, "grad_norm": 5.23001099068078, "learning_rate": 2.926403695146518e-09, "loss": 0.3822, "step": 13956 }, { "epoch": 0.9895072669266217, "grad_norm": 2.4962025862399457, "learning_rate": 2.887258690876249e-09, "loss": 0.3633, "step": 13957 }, { "epoch": 0.9895781637717121, "grad_norm": 9.067465408563415, "learning_rate": 2.8483771881743228e-09, "loss": 0.3398, "step": 13958 }, { "epoch": 0.9896490606168026, "grad_norm": 5.383130980101902, "learning_rate": 2.8097591890918764e-09, "loss": 0.3882, "step": 13959 }, { "epoch": 0.9897199574618929, "grad_norm": 2.5411819367476927, "learning_rate": 2.7714046956650587e-09, "loss": 0.3338, "step": 13960 }, { "epoch": 0.9897908543069833, "grad_norm": 2.545688521156586, "learning_rate": 2.733313709916141e-09, "loss": 0.3695, "step": 13961 }, { "epoch": 0.9898617511520738, "grad_norm": 1.8880320970852638, "learning_rate": 2.695486233853517e-09, "loss": 0.373, "step": 13962 }, { "epoch": 0.9899326479971641, "grad_norm": 2.88724183266279, "learning_rate": 2.6579222694728124e-09, "loss": 0.385, "step": 13963 }, { "epoch": 0.9900035448422545, "grad_norm": 2.6386824052799156, "learning_rate": 2.620621818753555e-09, "loss": 0.3351, "step": 13964 }, { "epoch": 0.9900744416873449, "grad_norm": 2.6898183171625765, "learning_rate": 2.5835848836636145e-09, "loss": 0.3872, "step": 13965 }, { "epoch": 0.9901453385324354, "grad_norm": 2.9219721360072053, "learning_rate": 2.546811466155319e-09, "loss": 0.3947, "step": 13966 }, { "epoch": 0.9902162353775257, "grad_norm": 3.852411557000206, "learning_rate": 2.5103015681682276e-09, "loss": 0.338, "step": 13967 }, { "epoch": 0.9902871322226161, "grad_norm": 3.234923336168675, "learning_rate": 2.474055191627467e-09, "loss": 0.3901, "step": 13968 }, { "epoch": 0.9903580290677065, "grad_norm": 2.898349657476146, "learning_rate": 2.438072338444286e-09, "loss": 0.4336, "step": 13969 }, { "epoch": 0.9904289259127969, "grad_norm": 4.88552396657854, "learning_rate": 2.4023530105155015e-09, "loss": 0.3594, "step": 13970 }, { "epoch": 0.9904998227578873, "grad_norm": 3.7007893478733265, "learning_rate": 2.366897209725716e-09, "loss": 0.3854, "step": 13971 }, { "epoch": 0.9905707196029777, "grad_norm": 2.631398238011966, "learning_rate": 2.3317049379434352e-09, "loss": 0.3865, "step": 13972 }, { "epoch": 0.990641616448068, "grad_norm": 3.8161733257737587, "learning_rate": 2.2967761970255078e-09, "loss": 0.368, "step": 13973 }, { "epoch": 0.9907125132931585, "grad_norm": 3.630703335167116, "learning_rate": 2.2621109888126826e-09, "loss": 0.3888, "step": 13974 }, { "epoch": 0.9907834101382489, "grad_norm": 1.8346928127929685, "learning_rate": 2.2277093151334973e-09, "loss": 0.3514, "step": 13975 }, { "epoch": 0.9908543069833392, "grad_norm": 5.155389440311556, "learning_rate": 2.1935711778020564e-09, "loss": 0.3518, "step": 13976 }, { "epoch": 0.9909252038284296, "grad_norm": 2.933167810778698, "learning_rate": 2.1596965786174763e-09, "loss": 0.4203, "step": 13977 }, { "epoch": 0.9909961006735201, "grad_norm": 2.503083883300588, "learning_rate": 2.126085519367771e-09, "loss": 0.4131, "step": 13978 }, { "epoch": 0.9910669975186104, "grad_norm": 2.2419628082029313, "learning_rate": 2.0927380018237464e-09, "loss": 0.4069, "step": 13979 }, { "epoch": 0.9911378943637008, "grad_norm": 2.530938847861037, "learning_rate": 2.059654027744551e-09, "loss": 0.35, "step": 13980 }, { "epoch": 0.9912087912087912, "grad_norm": 2.0402671458446173, "learning_rate": 2.026833598874345e-09, "loss": 0.3627, "step": 13981 }, { "epoch": 0.9912796880538816, "grad_norm": 3.7268770438758674, "learning_rate": 1.9942767169439656e-09, "loss": 0.4026, "step": 13982 }, { "epoch": 0.991350584898972, "grad_norm": 4.536575432601436, "learning_rate": 1.9619833836703737e-09, "loss": 0.438, "step": 13983 }, { "epoch": 0.9914214817440624, "grad_norm": 2.750653913210833, "learning_rate": 1.929953600756096e-09, "loss": 0.3874, "step": 13984 }, { "epoch": 0.9914923785891527, "grad_norm": 3.733881045675348, "learning_rate": 1.8981873698903362e-09, "loss": 0.341, "step": 13985 }, { "epoch": 0.9915632754342432, "grad_norm": 77.5172030456584, "learning_rate": 1.8666846927478667e-09, "loss": 0.3984, "step": 13986 }, { "epoch": 0.9916341722793336, "grad_norm": 1.6895784717576405, "learning_rate": 1.8354455709901354e-09, "loss": 0.4068, "step": 13987 }, { "epoch": 0.9917050691244239, "grad_norm": 4.38224489434617, "learning_rate": 1.8044700062641584e-09, "loss": 0.3994, "step": 13988 }, { "epoch": 0.9917759659695143, "grad_norm": 2.7876372735536976, "learning_rate": 1.7737580002036292e-09, "loss": 0.3945, "step": 13989 }, { "epoch": 0.9918468628146048, "grad_norm": 3.200157268817077, "learning_rate": 1.7433095544278078e-09, "loss": 0.4008, "step": 13990 }, { "epoch": 0.9919177596596952, "grad_norm": 2.985977056008023, "learning_rate": 1.7131246705420768e-09, "loss": 0.3478, "step": 13991 }, { "epoch": 0.9919886565047855, "grad_norm": 4.088894188729038, "learning_rate": 1.6832033501384959e-09, "loss": 0.3998, "step": 13992 }, { "epoch": 0.9920595533498759, "grad_norm": 3.6855654759536716, "learning_rate": 1.653545594794137e-09, "loss": 0.432, "step": 13993 }, { "epoch": 0.9921304501949664, "grad_norm": 2.1567884411764116, "learning_rate": 1.6241514060738595e-09, "loss": 0.3679, "step": 13994 }, { "epoch": 0.9922013470400567, "grad_norm": 3.249846652222815, "learning_rate": 1.5950207855275347e-09, "loss": 0.4476, "step": 13995 }, { "epoch": 0.9922722438851471, "grad_norm": 3.864766582483322, "learning_rate": 1.5661537346900457e-09, "loss": 0.3509, "step": 13996 }, { "epoch": 0.9923431407302375, "grad_norm": 3.6121814021927685, "learning_rate": 1.5375502550851739e-09, "loss": 0.3872, "step": 13997 }, { "epoch": 0.9924140375753279, "grad_norm": 7.73355630088488, "learning_rate": 1.509210348220047e-09, "loss": 0.4018, "step": 13998 }, { "epoch": 0.9924849344204183, "grad_norm": 1.755260500474949, "learning_rate": 1.4811340155890252e-09, "loss": 0.3346, "step": 13999 }, { "epoch": 0.9925558312655087, "grad_norm": 3.0819058640256602, "learning_rate": 1.453321258673701e-09, "loss": 0.3692, "step": 14000 }, { "epoch": 0.992626728110599, "grad_norm": 2.783718662740812, "learning_rate": 1.425772078939569e-09, "loss": 0.3937, "step": 14001 }, { "epoch": 0.9926976249556895, "grad_norm": 3.4255414076406683, "learning_rate": 1.398486477839911e-09, "loss": 0.3443, "step": 14002 }, { "epoch": 0.9927685218007799, "grad_norm": 2.9996921628314057, "learning_rate": 1.371464456813021e-09, "loss": 0.3653, "step": 14003 }, { "epoch": 0.9928394186458702, "grad_norm": 2.4620572317481613, "learning_rate": 1.344706017284425e-09, "loss": 0.4567, "step": 14004 }, { "epoch": 0.9929103154909606, "grad_norm": 2.6672473740395186, "learning_rate": 1.3182111606646619e-09, "loss": 0.3559, "step": 14005 }, { "epoch": 0.9929812123360511, "grad_norm": 11.368682585674435, "learning_rate": 1.291979888350947e-09, "loss": 0.3474, "step": 14006 }, { "epoch": 0.9930521091811414, "grad_norm": 3.838403023845749, "learning_rate": 1.2660122017266185e-09, "loss": 0.4094, "step": 14007 }, { "epoch": 0.9931230060262318, "grad_norm": 21.220548731760474, "learning_rate": 1.2403081021605811e-09, "loss": 0.2919, "step": 14008 }, { "epoch": 0.9931939028713223, "grad_norm": 2.4617536673125517, "learning_rate": 1.2148675910084173e-09, "loss": 0.3632, "step": 14009 }, { "epoch": 0.9932647997164126, "grad_norm": 4.368131660699398, "learning_rate": 1.1896906696118316e-09, "loss": 0.3949, "step": 14010 }, { "epoch": 0.993335696561503, "grad_norm": 2.1457361730658384, "learning_rate": 1.1647773392986506e-09, "loss": 0.4043, "step": 14011 }, { "epoch": 0.9934065934065934, "grad_norm": 4.773136710263812, "learning_rate": 1.1401276013817131e-09, "loss": 0.3931, "step": 14012 }, { "epoch": 0.9934774902516837, "grad_norm": 3.0174884831502804, "learning_rate": 1.1157414571616453e-09, "loss": 0.4186, "step": 14013 }, { "epoch": 0.9935483870967742, "grad_norm": 3.1793345711183183, "learning_rate": 1.0916189079235307e-09, "loss": 0.4117, "step": 14014 }, { "epoch": 0.9936192839418646, "grad_norm": 4.261111414313225, "learning_rate": 1.0677599549402395e-09, "loss": 0.3468, "step": 14015 }, { "epoch": 0.993690180786955, "grad_norm": 3.3238927754244965, "learning_rate": 1.0441645994690996e-09, "loss": 0.4393, "step": 14016 }, { "epoch": 0.9937610776320454, "grad_norm": 2.764037488684613, "learning_rate": 1.020832842755226e-09, "loss": 0.3763, "step": 14017 }, { "epoch": 0.9938319744771358, "grad_norm": 2.0435972135424563, "learning_rate": 9.977646860276358e-10, "loss": 0.3762, "step": 14018 }, { "epoch": 0.9939028713222262, "grad_norm": 2.5777930717688355, "learning_rate": 9.749601305042434e-10, "loss": 0.3525, "step": 14019 }, { "epoch": 0.9939737681673165, "grad_norm": 18.42625435959263, "learning_rate": 9.5241917738631e-10, "loss": 0.3383, "step": 14020 }, { "epoch": 0.994044665012407, "grad_norm": 6.2589471729344694, "learning_rate": 9.3014182786344e-10, "loss": 0.3894, "step": 14021 }, { "epoch": 0.9941155618574974, "grad_norm": 7.745281140090991, "learning_rate": 9.081280831091388e-10, "loss": 0.3772, "step": 14022 }, { "epoch": 0.9941864587025877, "grad_norm": 10.453563790728401, "learning_rate": 8.863779442852549e-10, "loss": 0.3613, "step": 14023 }, { "epoch": 0.9942573555476781, "grad_norm": 4.293447677964531, "learning_rate": 8.648914125380936e-10, "loss": 0.3391, "step": 14024 }, { "epoch": 0.9943282523927686, "grad_norm": 5.470782217591835, "learning_rate": 8.436684890011926e-10, "loss": 0.4012, "step": 14025 }, { "epoch": 0.9943991492378589, "grad_norm": 6.405186561822989, "learning_rate": 8.227091747931015e-10, "loss": 0.337, "step": 14026 }, { "epoch": 0.9944700460829493, "grad_norm": 2.0176140680249937, "learning_rate": 8.020134710190475e-10, "loss": 0.3591, "step": 14027 }, { "epoch": 0.9945409429280397, "grad_norm": 1.907371798877426, "learning_rate": 7.815813787709348e-10, "loss": 0.4041, "step": 14028 }, { "epoch": 0.9946118397731301, "grad_norm": 3.172719664334865, "learning_rate": 7.614128991256798e-10, "loss": 0.4216, "step": 14029 }, { "epoch": 0.9946827366182205, "grad_norm": 3.1185662696796794, "learning_rate": 7.415080331468761e-10, "loss": 0.3473, "step": 14030 }, { "epoch": 0.9947536334633109, "grad_norm": 4.797788283216489, "learning_rate": 7.218667818842395e-10, "loss": 0.4085, "step": 14031 }, { "epoch": 0.9948245303084012, "grad_norm": 3.7404991102416787, "learning_rate": 7.024891463730532e-10, "loss": 0.4198, "step": 14032 }, { "epoch": 0.9948954271534917, "grad_norm": 2.4922032631228586, "learning_rate": 6.833751276352774e-10, "loss": 0.3722, "step": 14033 }, { "epoch": 0.9949663239985821, "grad_norm": 3.064513195586148, "learning_rate": 6.645247266795496e-10, "loss": 0.3867, "step": 14034 }, { "epoch": 0.9950372208436724, "grad_norm": 4.24841963859999, "learning_rate": 6.459379444989644e-10, "loss": 0.3817, "step": 14035 }, { "epoch": 0.9951081176887628, "grad_norm": 2.3560860184242696, "learning_rate": 6.276147820738488e-10, "loss": 0.3928, "step": 14036 }, { "epoch": 0.9951790145338533, "grad_norm": 4.387600830120685, "learning_rate": 6.095552403700966e-10, "loss": 0.4126, "step": 14037 }, { "epoch": 0.9952499113789436, "grad_norm": 3.3341984781133114, "learning_rate": 5.917593203408345e-10, "loss": 0.3684, "step": 14038 }, { "epoch": 0.995320808224034, "grad_norm": 4.486578533650279, "learning_rate": 5.742270229236457e-10, "loss": 0.3653, "step": 14039 }, { "epoch": 0.9953917050691244, "grad_norm": 8.805737866767753, "learning_rate": 5.569583490439012e-10, "loss": 0.4248, "step": 14040 }, { "epoch": 0.9954626019142148, "grad_norm": 2.4949410791482296, "learning_rate": 5.399532996108737e-10, "loss": 0.3743, "step": 14041 }, { "epoch": 0.9955334987593052, "grad_norm": 3.6865529995988875, "learning_rate": 5.232118755227333e-10, "loss": 0.365, "step": 14042 }, { "epoch": 0.9956043956043956, "grad_norm": 3.1284596451624576, "learning_rate": 5.067340776609975e-10, "loss": 0.444, "step": 14043 }, { "epoch": 0.995675292449486, "grad_norm": 2.332240955747141, "learning_rate": 4.905199068955258e-10, "loss": 0.3585, "step": 14044 }, { "epoch": 0.9957461892945764, "grad_norm": 2.7514213130695198, "learning_rate": 4.745693640806348e-10, "loss": 0.3461, "step": 14045 }, { "epoch": 0.9958170861396668, "grad_norm": 15.259278900514063, "learning_rate": 4.5888245005787367e-10, "loss": 0.4593, "step": 14046 }, { "epoch": 0.9958879829847572, "grad_norm": 2.681952333324452, "learning_rate": 4.4345916565380344e-10, "loss": 0.3378, "step": 14047 }, { "epoch": 0.9959588798298475, "grad_norm": 2.764530245579099, "learning_rate": 4.282995116827726e-10, "loss": 0.4121, "step": 14048 }, { "epoch": 0.996029776674938, "grad_norm": 2.7859014450307886, "learning_rate": 4.1340348894303164e-10, "loss": 0.3881, "step": 14049 }, { "epoch": 0.9961006735200284, "grad_norm": 10.138113583504788, "learning_rate": 3.9877109822061833e-10, "loss": 0.3932, "step": 14050 }, { "epoch": 0.9961715703651187, "grad_norm": 5.618221067224342, "learning_rate": 3.844023402871377e-10, "loss": 0.4068, "step": 14051 }, { "epoch": 0.9962424672102091, "grad_norm": 1.7495529229374942, "learning_rate": 3.702972159003171e-10, "loss": 0.3835, "step": 14052 }, { "epoch": 0.9963133640552996, "grad_norm": 9.05596929963388, "learning_rate": 3.5645572580345067e-10, "loss": 0.3101, "step": 14053 }, { "epoch": 0.9963842609003899, "grad_norm": 3.844363640985124, "learning_rate": 3.4287787072651013e-10, "loss": 0.4062, "step": 14054 }, { "epoch": 0.9964551577454803, "grad_norm": 2.5969545357744606, "learning_rate": 3.295636513861444e-10, "loss": 0.3431, "step": 14055 }, { "epoch": 0.9965260545905708, "grad_norm": 3.0820148863204313, "learning_rate": 3.165130684840145e-10, "loss": 0.367, "step": 14056 }, { "epoch": 0.9965969514356611, "grad_norm": 2.417313334269479, "learning_rate": 3.0372612270790356e-10, "loss": 0.4282, "step": 14057 }, { "epoch": 0.9966678482807515, "grad_norm": 4.515228006143381, "learning_rate": 2.912028147328272e-10, "loss": 0.3286, "step": 14058 }, { "epoch": 0.9967387451258419, "grad_norm": 2.5777027909807537, "learning_rate": 2.7894314521825783e-10, "loss": 0.3901, "step": 14059 }, { "epoch": 0.9968096419709322, "grad_norm": 2.5744559096468342, "learning_rate": 2.669471148114555e-10, "loss": 0.3906, "step": 14060 }, { "epoch": 0.9968805388160227, "grad_norm": 3.2414873767876964, "learning_rate": 2.5521472414413716e-10, "loss": 0.3901, "step": 14061 }, { "epoch": 0.9969514356611131, "grad_norm": 1.9691166003177698, "learning_rate": 2.437459738358072e-10, "loss": 0.3705, "step": 14062 }, { "epoch": 0.9970223325062034, "grad_norm": 3.175817370690114, "learning_rate": 2.3254086449098211e-10, "loss": 0.3852, "step": 14063 }, { "epoch": 0.9970932293512939, "grad_norm": 4.125286479440437, "learning_rate": 2.2159939670030052e-10, "loss": 0.3873, "step": 14064 }, { "epoch": 0.9971641261963843, "grad_norm": 3.973150032586099, "learning_rate": 2.1092157104107835e-10, "loss": 0.3843, "step": 14065 }, { "epoch": 0.9972350230414746, "grad_norm": 3.982744779949936, "learning_rate": 2.0050738807619875e-10, "loss": 0.4211, "step": 14066 }, { "epoch": 0.997305919886565, "grad_norm": 3.367229505804895, "learning_rate": 1.9035684835466694e-10, "loss": 0.4188, "step": 14067 }, { "epoch": 0.9973768167316555, "grad_norm": 2.460752236473942, "learning_rate": 1.8046995241161047e-10, "loss": 0.4054, "step": 14068 }, { "epoch": 0.9974477135767459, "grad_norm": 2.0429472245443683, "learning_rate": 1.7084670076827904e-10, "loss": 0.3893, "step": 14069 }, { "epoch": 0.9975186104218362, "grad_norm": 2.301619182075425, "learning_rate": 1.6148709393259964e-10, "loss": 0.3996, "step": 14070 }, { "epoch": 0.9975895072669266, "grad_norm": 5.15117263616229, "learning_rate": 1.5239113239806646e-10, "loss": 0.4082, "step": 14071 }, { "epoch": 0.9976604041120171, "grad_norm": 2.112907288332812, "learning_rate": 1.4355881664429582e-10, "loss": 0.3596, "step": 14072 }, { "epoch": 0.9977313009571074, "grad_norm": 2.660379867900426, "learning_rate": 1.3499014713647118e-10, "loss": 0.3881, "step": 14073 }, { "epoch": 0.9978021978021978, "grad_norm": 2.504378109051838, "learning_rate": 1.2668512432756353e-10, "loss": 0.3297, "step": 14074 }, { "epoch": 0.9978730946472882, "grad_norm": 3.3987093281584055, "learning_rate": 1.186437486538905e-10, "loss": 0.3934, "step": 14075 }, { "epoch": 0.9979439914923786, "grad_norm": 2.6674858932884997, "learning_rate": 1.1086602054122264e-10, "loss": 0.3857, "step": 14076 }, { "epoch": 0.998014888337469, "grad_norm": 2.5189912056459254, "learning_rate": 1.0335194039812202e-10, "loss": 0.39, "step": 14077 }, { "epoch": 0.9980857851825594, "grad_norm": 2.97528094973253, "learning_rate": 9.610150862204848e-11, "loss": 0.3939, "step": 14078 }, { "epoch": 0.9981566820276497, "grad_norm": 2.042442716881628, "learning_rate": 8.911472559436362e-11, "loss": 0.3654, "step": 14079 }, { "epoch": 0.9982275788727402, "grad_norm": 2.692783516739927, "learning_rate": 8.239159168421662e-11, "loss": 0.4013, "step": 14080 }, { "epoch": 0.9982984757178306, "grad_norm": 5.600316336013042, "learning_rate": 7.593210724521349e-11, "loss": 0.4245, "step": 14081 }, { "epoch": 0.9983693725629209, "grad_norm": 4.604429829160288, "learning_rate": 6.973627261930294e-11, "loss": 0.3919, "step": 14082 }, { "epoch": 0.9984402694080113, "grad_norm": 2.6679287164938654, "learning_rate": 6.38040881323354e-11, "loss": 0.387, "step": 14083 }, { "epoch": 0.9985111662531018, "grad_norm": 3.49433956893338, "learning_rate": 5.8135554096838685e-11, "loss": 0.44, "step": 14084 }, { "epoch": 0.9985820630981921, "grad_norm": 5.851859069418354, "learning_rate": 5.273067081257299e-11, "loss": 0.3911, "step": 14085 }, { "epoch": 0.9986529599432825, "grad_norm": 4.6645564653540985, "learning_rate": 4.758943856431053e-11, "loss": 0.3811, "step": 14086 }, { "epoch": 0.9987238567883729, "grad_norm": 3.1331629407571597, "learning_rate": 4.271185762239061e-11, "loss": 0.3751, "step": 14087 }, { "epoch": 0.9987947536334633, "grad_norm": 2.725925321583524, "learning_rate": 3.809792824494007e-11, "loss": 0.3994, "step": 14088 }, { "epoch": 0.9988656504785537, "grad_norm": 3.526260238181396, "learning_rate": 3.374765067454267e-11, "loss": 0.4134, "step": 14089 }, { "epoch": 0.9989365473236441, "grad_norm": 1.8780254564087442, "learning_rate": 2.966102514101454e-11, "loss": 0.352, "step": 14090 }, { "epoch": 0.9990074441687344, "grad_norm": 7.506072846079449, "learning_rate": 2.5838051860294088e-11, "loss": 0.3683, "step": 14091 }, { "epoch": 0.9990783410138249, "grad_norm": 2.600694722076343, "learning_rate": 2.2278731032776557e-11, "loss": 0.3591, "step": 14092 }, { "epoch": 0.9991492378589153, "grad_norm": 8.558459928134608, "learning_rate": 1.8983062847199862e-11, "loss": 0.4252, "step": 14093 }, { "epoch": 0.9992201347040057, "grad_norm": 2.863423243512842, "learning_rate": 1.5951047476758798e-11, "loss": 0.3371, "step": 14094 }, { "epoch": 0.999291031549096, "grad_norm": 2.2928573640394374, "learning_rate": 1.3182685081325474e-11, "loss": 0.3465, "step": 14095 }, { "epoch": 0.9993619283941865, "grad_norm": 1.7930061209036081, "learning_rate": 1.0677975807449337e-11, "loss": 0.3781, "step": 14096 }, { "epoch": 0.9994328252392769, "grad_norm": 2.623645210194336, "learning_rate": 8.43691978669181e-12, "loss": 0.3415, "step": 14097 }, { "epoch": 0.9995037220843672, "grad_norm": 3.8024433430458986, "learning_rate": 6.459517137291649e-12, "loss": 0.3785, "step": 14098 }, { "epoch": 0.9995746189294576, "grad_norm": 3.9180433502793295, "learning_rate": 4.745767964164927e-12, "loss": 0.3801, "step": 14099 }, { "epoch": 0.9996455157745481, "grad_norm": 2.3341311117771344, "learning_rate": 3.2956723566845984e-12, "loss": 0.3858, "step": 14100 }, { "epoch": 0.9997164126196384, "grad_norm": 3.159687876436403, "learning_rate": 2.109230391456052e-12, "loss": 0.3814, "step": 14101 }, { "epoch": 0.9997873094647288, "grad_norm": 3.409977825797542, "learning_rate": 1.1864421317620001e-12, "loss": 0.3712, "step": 14102 }, { "epoch": 0.9998582063098193, "grad_norm": 2.7722383689587184, "learning_rate": 5.273076258971444e-13, "loss": 0.3694, "step": 14103 }, { "epoch": 0.9999291031549096, "grad_norm": 2.586797895630909, "learning_rate": 1.318269082783985e-13, "loss": 0.3696, "step": 14104 }, { "epoch": 1.0, "grad_norm": 2.054940695008144, "learning_rate": 0.0, "loss": 0.417, "step": 14105 }, { "epoch": 1.0, "step": 14105, "total_flos": 4628114056822784.0, "train_loss": 0.4341994650606482, "train_runtime": 186096.2196, "train_samples_per_second": 9.702, "train_steps_per_second": 0.076 } ], "logging_steps": 1.0, "max_steps": 14105, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 4628114056822784.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }