diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,45138 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 6444,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004655493482309125,
+      "grad_norm": 4.143864162727672,
+      "learning_rate": 1.5479876160990715e-08,
+      "loss": 0.6798,
+      "step": 1
+    },
+    {
+      "epoch": 0.000931098696461825,
+      "grad_norm": 4.260876342398468,
+      "learning_rate": 3.095975232198143e-08,
+      "loss": 0.6907,
+      "step": 2
+    },
+    {
+      "epoch": 0.0013966480446927375,
+      "grad_norm": 4.2281173067638145,
+      "learning_rate": 4.6439628482972136e-08,
+      "loss": 0.6878,
+      "step": 3
+    },
+    {
+      "epoch": 0.00186219739292365,
+      "grad_norm": 4.432080471415043,
+      "learning_rate": 6.191950464396286e-08,
+      "loss": 0.7012,
+      "step": 4
+    },
+    {
+      "epoch": 0.0023277467411545625,
+      "grad_norm": 4.26280360052213,
+      "learning_rate": 7.739938080495356e-08,
+      "loss": 0.6936,
+      "step": 5
+    },
+    {
+      "epoch": 0.002793296089385475,
+      "grad_norm": 4.210247227564665,
+      "learning_rate": 9.287925696594427e-08,
+      "loss": 0.6937,
+      "step": 6
+    },
+    {
+      "epoch": 0.0032588454376163874,
+      "grad_norm": 4.336581057898776,
+      "learning_rate": 1.08359133126935e-07,
+      "loss": 0.6929,
+      "step": 7
+    },
+    {
+      "epoch": 0.0037243947858473,
+      "grad_norm": 4.311882385298882,
+      "learning_rate": 1.2383900928792572e-07,
+      "loss": 0.7156,
+      "step": 8
+    },
+    {
+      "epoch": 0.004189944134078212,
+      "grad_norm": 4.264004473079579,
+      "learning_rate": 1.393188854489164e-07,
+      "loss": 0.6792,
+      "step": 9
+    },
+    {
+      "epoch": 0.004655493482309125,
+      "grad_norm": 4.1384486803089615,
+      "learning_rate": 1.5479876160990712e-07,
+      "loss": 0.6771,
+      "step": 10
+    },
+    {
+      "epoch": 0.005121042830540037,
+      "grad_norm": 4.177371129234454,
+      "learning_rate": 1.7027863777089783e-07,
+      "loss": 0.6825,
+      "step": 11
+    },
+    {
+      "epoch": 0.00558659217877095,
+      "grad_norm": 3.949145392814891,
+      "learning_rate": 1.8575851393188855e-07,
+      "loss": 0.6755,
+      "step": 12
+    },
+    {
+      "epoch": 0.006052141527001862,
+      "grad_norm": 4.247599686356128,
+      "learning_rate": 2.0123839009287928e-07,
+      "loss": 0.666,
+      "step": 13
+    },
+    {
+      "epoch": 0.006517690875232775,
+      "grad_norm": 4.243636333621182,
+      "learning_rate": 2.1671826625387e-07,
+      "loss": 0.685,
+      "step": 14
+    },
+    {
+      "epoch": 0.006983240223463687,
+      "grad_norm": 4.203951762080091,
+      "learning_rate": 2.321981424148607e-07,
+      "loss": 0.7026,
+      "step": 15
+    },
+    {
+      "epoch": 0.0074487895716946,
+      "grad_norm": 4.1791311562903815,
+      "learning_rate": 2.4767801857585145e-07,
+      "loss": 0.6814,
+      "step": 16
+    },
+    {
+      "epoch": 0.007914338919925512,
+      "grad_norm": 3.952532049148496,
+      "learning_rate": 2.6315789473684213e-07,
+      "loss": 0.676,
+      "step": 17
+    },
+    {
+      "epoch": 0.008379888268156424,
+      "grad_norm": 3.9976880842149773,
+      "learning_rate": 2.786377708978328e-07,
+      "loss": 0.6888,
+      "step": 18
+    },
+    {
+      "epoch": 0.008845437616387336,
+      "grad_norm": 4.029366252118066,
+      "learning_rate": 2.9411764705882356e-07,
+      "loss": 0.7091,
+      "step": 19
+    },
+    {
+      "epoch": 0.00931098696461825,
+      "grad_norm": 4.002934293984722,
+      "learning_rate": 3.0959752321981424e-07,
+      "loss": 0.691,
+      "step": 20
+    },
+    {
+      "epoch": 0.009776536312849162,
+      "grad_norm": 3.8302020485361132,
+      "learning_rate": 3.25077399380805e-07,
+      "loss": 0.6898,
+      "step": 21
+    },
+    {
+      "epoch": 0.010242085661080074,
+      "grad_norm": 3.97278302440018,
+      "learning_rate": 3.4055727554179567e-07,
+      "loss": 0.6748,
+      "step": 22
+    },
+    {
+      "epoch": 0.010707635009310988,
+      "grad_norm": 3.732627729909466,
+      "learning_rate": 3.560371517027864e-07,
+      "loss": 0.6532,
+      "step": 23
+    },
+    {
+      "epoch": 0.0111731843575419,
+      "grad_norm": 3.5770655073661093,
+      "learning_rate": 3.715170278637771e-07,
+      "loss": 0.6778,
+      "step": 24
+    },
+    {
+      "epoch": 0.011638733705772812,
+      "grad_norm": 3.3717040056594865,
+      "learning_rate": 3.869969040247679e-07,
+      "loss": 0.6944,
+      "step": 25
+    },
+    {
+      "epoch": 0.012104283054003724,
+      "grad_norm": 3.1821393756768384,
+      "learning_rate": 4.0247678018575857e-07,
+      "loss": 0.6653,
+      "step": 26
+    },
+    {
+      "epoch": 0.012569832402234637,
+      "grad_norm": 3.123584210799692,
+      "learning_rate": 4.1795665634674925e-07,
+      "loss": 0.6486,
+      "step": 27
+    },
+    {
+      "epoch": 0.01303538175046555,
+      "grad_norm": 3.099810703113951,
+      "learning_rate": 4.3343653250774e-07,
+      "loss": 0.6422,
+      "step": 28
+    },
+    {
+      "epoch": 0.013500931098696461,
+      "grad_norm": 2.994501525927118,
+      "learning_rate": 4.489164086687307e-07,
+      "loss": 0.6544,
+      "step": 29
+    },
+    {
+      "epoch": 0.013966480446927373,
+      "grad_norm": 2.8726706269638607,
+      "learning_rate": 4.643962848297214e-07,
+      "loss": 0.6612,
+      "step": 30
+    },
+    {
+      "epoch": 0.014432029795158287,
+      "grad_norm": 2.894044319251768,
+      "learning_rate": 4.798761609907122e-07,
+      "loss": 0.6428,
+      "step": 31
+    },
+    {
+      "epoch": 0.0148975791433892,
+      "grad_norm": 2.839801376552692,
+      "learning_rate": 4.953560371517029e-07,
+      "loss": 0.6605,
+      "step": 32
+    },
+    {
+      "epoch": 0.015363128491620111,
+      "grad_norm": 2.6345596170152263,
+      "learning_rate": 5.108359133126935e-07,
+      "loss": 0.6612,
+      "step": 33
+    },
+    {
+      "epoch": 0.015828677839851025,
+      "grad_norm": 2.2492857756140707,
+      "learning_rate": 5.263157894736843e-07,
+      "loss": 0.6376,
+      "step": 34
+    },
+    {
+      "epoch": 0.016294227188081937,
+      "grad_norm": 1.9616851257511285,
+      "learning_rate": 5.41795665634675e-07,
+      "loss": 0.6452,
+      "step": 35
+    },
+    {
+      "epoch": 0.01675977653631285,
+      "grad_norm": 1.8608987770524916,
+      "learning_rate": 5.572755417956656e-07,
+      "loss": 0.6116,
+      "step": 36
+    },
+    {
+      "epoch": 0.01722532588454376,
+      "grad_norm": 1.8601672431470884,
+      "learning_rate": 5.727554179566564e-07,
+      "loss": 0.6282,
+      "step": 37
+    },
+    {
+      "epoch": 0.017690875232774673,
+      "grad_norm": 1.7475478992264075,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 0.6418,
+      "step": 38
+    },
+    {
+      "epoch": 0.018156424581005588,
+      "grad_norm": 1.6605997435167843,
+      "learning_rate": 6.037151702786379e-07,
+      "loss": 0.6293,
+      "step": 39
+    },
+    {
+      "epoch": 0.0186219739292365,
+      "grad_norm": 1.654466932159507,
+      "learning_rate": 6.191950464396285e-07,
+      "loss": 0.6409,
+      "step": 40
+    },
+    {
+      "epoch": 0.019087523277467412,
+      "grad_norm": 1.6511624932968363,
+      "learning_rate": 6.346749226006192e-07,
+      "loss": 0.6281,
+      "step": 41
+    },
+    {
+      "epoch": 0.019553072625698324,
+      "grad_norm": 1.5905127643631167,
+      "learning_rate": 6.5015479876161e-07,
+      "loss": 0.6191,
+      "step": 42
+    },
+    {
+      "epoch": 0.020018621973929236,
+      "grad_norm": 1.4686469780923481,
+      "learning_rate": 6.656346749226007e-07,
+      "loss": 0.6206,
+      "step": 43
+    },
+    {
+      "epoch": 0.020484171322160148,
+      "grad_norm": 1.4486201792524727,
+      "learning_rate": 6.811145510835913e-07,
+      "loss": 0.6093,
+      "step": 44
+    },
+    {
+      "epoch": 0.02094972067039106,
+      "grad_norm": 1.3161801403793787,
+      "learning_rate": 6.965944272445821e-07,
+      "loss": 0.607,
+      "step": 45
+    },
+    {
+      "epoch": 0.021415270018621976,
+      "grad_norm": 1.2096109319958612,
+      "learning_rate": 7.120743034055728e-07,
+      "loss": 0.6039,
+      "step": 46
+    },
+    {
+      "epoch": 0.021880819366852888,
+      "grad_norm": 1.1961736862001944,
+      "learning_rate": 7.275541795665634e-07,
+      "loss": 0.5999,
+      "step": 47
+    },
+    {
+      "epoch": 0.0223463687150838,
+      "grad_norm": 1.1924285887633252,
+      "learning_rate": 7.430340557275542e-07,
+      "loss": 0.5959,
+      "step": 48
+    },
+    {
+      "epoch": 0.02281191806331471,
+      "grad_norm": 1.1226628258631968,
+      "learning_rate": 7.585139318885449e-07,
+      "loss": 0.5835,
+      "step": 49
+    },
+    {
+      "epoch": 0.023277467411545624,
+      "grad_norm": 1.1156484547534924,
+      "learning_rate": 7.739938080495358e-07,
+      "loss": 0.5984,
+      "step": 50
+    },
+    {
+      "epoch": 0.023743016759776536,
+      "grad_norm": 1.0741520468626569,
+      "learning_rate": 7.894736842105263e-07,
+      "loss": 0.5944,
+      "step": 51
+    },
+    {
+      "epoch": 0.024208566108007448,
+      "grad_norm": 1.0518814589152363,
+      "learning_rate": 8.049535603715171e-07,
+      "loss": 0.5737,
+      "step": 52
+    },
+    {
+      "epoch": 0.024674115456238363,
+      "grad_norm": 1.0693228936357408,
+      "learning_rate": 8.204334365325079e-07,
+      "loss": 0.5963,
+      "step": 53
+    },
+    {
+      "epoch": 0.025139664804469275,
+      "grad_norm": 1.0101967177725584,
+      "learning_rate": 8.359133126934985e-07,
+      "loss": 0.5882,
+      "step": 54
+    },
+    {
+      "epoch": 0.025605214152700187,
+      "grad_norm": 1.0457722894291805,
+      "learning_rate": 8.513931888544892e-07,
+      "loss": 0.5889,
+      "step": 55
+    },
+    {
+      "epoch": 0.0260707635009311,
+      "grad_norm": 0.9393864932236966,
+      "learning_rate": 8.6687306501548e-07,
+      "loss": 0.5563,
+      "step": 56
+    },
+    {
+      "epoch": 0.02653631284916201,
+      "grad_norm": 0.9066728696474886,
+      "learning_rate": 8.823529411764707e-07,
+      "loss": 0.5771,
+      "step": 57
+    },
+    {
+      "epoch": 0.027001862197392923,
+      "grad_norm": 0.9163958196414529,
+      "learning_rate": 8.978328173374614e-07,
+      "loss": 0.5713,
+      "step": 58
+    },
+    {
+      "epoch": 0.027467411545623835,
+      "grad_norm": 0.8881444942504693,
+      "learning_rate": 9.133126934984521e-07,
+      "loss": 0.5737,
+      "step": 59
+    },
+    {
+      "epoch": 0.027932960893854747,
+      "grad_norm": 0.8124818009559723,
+      "learning_rate": 9.287925696594428e-07,
+      "loss": 0.5903,
+      "step": 60
+    },
+    {
+      "epoch": 0.028398510242085662,
+      "grad_norm": 0.8883942495153572,
+      "learning_rate": 9.442724458204335e-07,
+      "loss": 0.5692,
+      "step": 61
+    },
+    {
+      "epoch": 0.028864059590316574,
+      "grad_norm": 0.8642824372816141,
+      "learning_rate": 9.597523219814243e-07,
+      "loss": 0.594,
+      "step": 62
+    },
+    {
+      "epoch": 0.029329608938547486,
+      "grad_norm": 0.7815875141331973,
+      "learning_rate": 9.75232198142415e-07,
+      "loss": 0.5755,
+      "step": 63
+    },
+    {
+      "epoch": 0.0297951582867784,
+      "grad_norm": 0.8148447993375885,
+      "learning_rate": 9.907120743034058e-07,
+      "loss": 0.5689,
+      "step": 64
+    },
+    {
+      "epoch": 0.03026070763500931,
+      "grad_norm": 0.8349827260798924,
+      "learning_rate": 1.0061919504643964e-06,
+      "loss": 0.556,
+      "step": 65
+    },
+    {
+      "epoch": 0.030726256983240222,
+      "grad_norm": 0.7834269132985276,
+      "learning_rate": 1.021671826625387e-06,
+      "loss": 0.5838,
+      "step": 66
+    },
+    {
+      "epoch": 0.031191806331471134,
+      "grad_norm": 0.8581075641611279,
+      "learning_rate": 1.037151702786378e-06,
+      "loss": 0.5792,
+      "step": 67
+    },
+    {
+      "epoch": 0.03165735567970205,
+      "grad_norm": 0.8534943446132287,
+      "learning_rate": 1.0526315789473685e-06,
+      "loss": 0.5789,
+      "step": 68
+    },
+    {
+      "epoch": 0.03212290502793296,
+      "grad_norm": 0.869067548569647,
+      "learning_rate": 1.0681114551083592e-06,
+      "loss": 0.5564,
+      "step": 69
+    },
+    {
+      "epoch": 0.032588454376163874,
+      "grad_norm": 0.8081733571051437,
+      "learning_rate": 1.08359133126935e-06,
+      "loss": 0.5624,
+      "step": 70
+    },
+    {
+      "epoch": 0.033054003724394786,
+      "grad_norm": 0.7986672538823885,
+      "learning_rate": 1.0990712074303406e-06,
+      "loss": 0.5776,
+      "step": 71
+    },
+    {
+      "epoch": 0.0335195530726257,
+      "grad_norm": 0.8027656558757922,
+      "learning_rate": 1.1145510835913313e-06,
+      "loss": 0.568,
+      "step": 72
+    },
+    {
+      "epoch": 0.03398510242085661,
+      "grad_norm": 0.8293654524550554,
+      "learning_rate": 1.1300309597523221e-06,
+      "loss": 0.5502,
+      "step": 73
+    },
+    {
+      "epoch": 0.03445065176908752,
+      "grad_norm": 0.7485544057330278,
+      "learning_rate": 1.1455108359133127e-06,
+      "loss": 0.5589,
+      "step": 74
+    },
+    {
+      "epoch": 0.034916201117318434,
+      "grad_norm": 0.8561192212306219,
+      "learning_rate": 1.1609907120743034e-06,
+      "loss": 0.5341,
+      "step": 75
+    },
+    {
+      "epoch": 0.035381750465549346,
+      "grad_norm": 0.8012653334970884,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 0.5542,
+      "step": 76
+    },
+    {
+      "epoch": 0.03584729981378026,
+      "grad_norm": 0.7378686176876464,
+      "learning_rate": 1.1919504643962849e-06,
+      "loss": 0.563,
+      "step": 77
+    },
+    {
+      "epoch": 0.036312849162011177,
+      "grad_norm": 0.7232077152309456,
+      "learning_rate": 1.2074303405572757e-06,
+      "loss": 0.549,
+      "step": 78
+    },
+    {
+      "epoch": 0.03677839851024209,
+      "grad_norm": 0.7139157721881457,
+      "learning_rate": 1.2229102167182663e-06,
+      "loss": 0.5577,
+      "step": 79
+    },
+    {
+      "epoch": 0.037243947858473,
+      "grad_norm": 0.6846483931749778,
+      "learning_rate": 1.238390092879257e-06,
+      "loss": 0.5304,
+      "step": 80
+    },
+    {
+      "epoch": 0.03770949720670391,
+      "grad_norm": 0.6929290496675755,
+      "learning_rate": 1.2538699690402478e-06,
+      "loss": 0.5844,
+      "step": 81
+    },
+    {
+      "epoch": 0.038175046554934824,
+      "grad_norm": 0.7158950918585701,
+      "learning_rate": 1.2693498452012384e-06,
+      "loss": 0.548,
+      "step": 82
+    },
+    {
+      "epoch": 0.038640595903165736,
+      "grad_norm": 0.6744895219791348,
+      "learning_rate": 1.2848297213622293e-06,
+      "loss": 0.5442,
+      "step": 83
+    },
+    {
+      "epoch": 0.03910614525139665,
+      "grad_norm": 0.7120557108432657,
+      "learning_rate": 1.30030959752322e-06,
+      "loss": 0.5522,
+      "step": 84
+    },
+    {
+      "epoch": 0.03957169459962756,
+      "grad_norm": 0.7389205049999904,
+      "learning_rate": 1.3157894736842106e-06,
+      "loss": 0.5407,
+      "step": 85
+    },
+    {
+      "epoch": 0.04003724394785847,
+      "grad_norm": 0.729159392936322,
+      "learning_rate": 1.3312693498452014e-06,
+      "loss": 0.5575,
+      "step": 86
+    },
+    {
+      "epoch": 0.040502793296089384,
+      "grad_norm": 0.6637691214344167,
+      "learning_rate": 1.346749226006192e-06,
+      "loss": 0.5554,
+      "step": 87
+    },
+    {
+      "epoch": 0.040968342644320296,
+      "grad_norm": 0.7019745994325189,
+      "learning_rate": 1.3622291021671827e-06,
+      "loss": 0.5435,
+      "step": 88
+    },
+    {
+      "epoch": 0.04143389199255121,
+      "grad_norm": 0.6745468058820847,
+      "learning_rate": 1.3777089783281735e-06,
+      "loss": 0.5498,
+      "step": 89
+    },
+    {
+      "epoch": 0.04189944134078212,
+      "grad_norm": 0.6518959217306766,
+      "learning_rate": 1.3931888544891641e-06,
+      "loss": 0.556,
+      "step": 90
+    },
+    {
+      "epoch": 0.04236499068901303,
+      "grad_norm": 0.6549501893776432,
+      "learning_rate": 1.4086687306501548e-06,
+      "loss": 0.5594,
+      "step": 91
+    },
+    {
+      "epoch": 0.04283054003724395,
+      "grad_norm": 0.6444908146316602,
+      "learning_rate": 1.4241486068111456e-06,
+      "loss": 0.573,
+      "step": 92
+    },
+    {
+      "epoch": 0.04329608938547486,
+      "grad_norm": 0.6461328650378827,
+      "learning_rate": 1.4396284829721363e-06,
+      "loss": 0.5603,
+      "step": 93
+    },
+    {
+      "epoch": 0.043761638733705775,
+      "grad_norm": 0.6907198996102626,
+      "learning_rate": 1.4551083591331269e-06,
+      "loss": 0.5379,
+      "step": 94
+    },
+    {
+      "epoch": 0.04422718808193669,
+      "grad_norm": 0.6476541200623472,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 0.5434,
+      "step": 95
+    },
+    {
+      "epoch": 0.0446927374301676,
+      "grad_norm": 0.6700980024487265,
+      "learning_rate": 1.4860681114551084e-06,
+      "loss": 0.5431,
+      "step": 96
+    },
+    {
+      "epoch": 0.04515828677839851,
+      "grad_norm": 0.6523383648451098,
+      "learning_rate": 1.5015479876160994e-06,
+      "loss": 0.5196,
+      "step": 97
+    },
+    {
+      "epoch": 0.04562383612662942,
+      "grad_norm": 0.6724917680884513,
+      "learning_rate": 1.5170278637770898e-06,
+      "loss": 0.574,
+      "step": 98
+    },
+    {
+      "epoch": 0.046089385474860335,
+      "grad_norm": 0.6971936417526337,
+      "learning_rate": 1.5325077399380805e-06,
+      "loss": 0.5512,
+      "step": 99
+    },
+    {
+      "epoch": 0.04655493482309125,
+      "grad_norm": 0.7382973686012949,
+      "learning_rate": 1.5479876160990715e-06,
+      "loss": 0.5513,
+      "step": 100
+    },
+    {
+      "epoch": 0.04702048417132216,
+      "grad_norm": 0.6602986160252511,
+      "learning_rate": 1.5634674922600622e-06,
+      "loss": 0.5365,
+      "step": 101
+    },
+    {
+      "epoch": 0.04748603351955307,
+      "grad_norm": 0.6481821524615549,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 0.5321,
+      "step": 102
+    },
+    {
+      "epoch": 0.04795158286778398,
+      "grad_norm": 0.6459250768429449,
+      "learning_rate": 1.5944272445820436e-06,
+      "loss": 0.5459,
+      "step": 103
+    },
+    {
+      "epoch": 0.048417132216014895,
+      "grad_norm": 0.6378033055662111,
+      "learning_rate": 1.6099071207430343e-06,
+      "loss": 0.5395,
+      "step": 104
+    },
+    {
+      "epoch": 0.04888268156424581,
+      "grad_norm": 0.6615548261658989,
+      "learning_rate": 1.6253869969040247e-06,
+      "loss": 0.5428,
+      "step": 105
+    },
+    {
+      "epoch": 0.049348230912476726,
+      "grad_norm": 0.6609589516632917,
+      "learning_rate": 1.6408668730650157e-06,
+      "loss": 0.5596,
+      "step": 106
+    },
+    {
+      "epoch": 0.04981378026070764,
+      "grad_norm": 0.6317002822800324,
+      "learning_rate": 1.6563467492260064e-06,
+      "loss": 0.5402,
+      "step": 107
+    },
+    {
+      "epoch": 0.05027932960893855,
+      "grad_norm": 0.6482285938222528,
+      "learning_rate": 1.671826625386997e-06,
+      "loss": 0.5473,
+      "step": 108
+    },
+    {
+      "epoch": 0.05074487895716946,
+      "grad_norm": 0.6250840390044115,
+      "learning_rate": 1.6873065015479879e-06,
+      "loss": 0.5191,
+      "step": 109
+    },
+    {
+      "epoch": 0.051210428305400374,
+      "grad_norm": 0.6340226673717015,
+      "learning_rate": 1.7027863777089785e-06,
+      "loss": 0.5175,
+      "step": 110
+    },
+    {
+      "epoch": 0.051675977653631286,
+      "grad_norm": 0.6536773739435273,
+      "learning_rate": 1.7182662538699693e-06,
+      "loss": 0.5445,
+      "step": 111
+    },
+    {
+      "epoch": 0.0521415270018622,
+      "grad_norm": 0.6423659058057712,
+      "learning_rate": 1.73374613003096e-06,
+      "loss": 0.5437,
+      "step": 112
+    },
+    {
+      "epoch": 0.05260707635009311,
+      "grad_norm": 0.6252643088263633,
+      "learning_rate": 1.7492260061919506e-06,
+      "loss": 0.5234,
+      "step": 113
+    },
+    {
+      "epoch": 0.05307262569832402,
+      "grad_norm": 0.6515986913608977,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 0.5335,
+      "step": 114
+    },
+    {
+      "epoch": 0.053538175046554934,
+      "grad_norm": 0.6273665526512381,
+      "learning_rate": 1.780185758513932e-06,
+      "loss": 0.551,
+      "step": 115
+    },
+    {
+      "epoch": 0.054003724394785846,
+      "grad_norm": 0.6061199108474716,
+      "learning_rate": 1.7956656346749227e-06,
+      "loss": 0.5253,
+      "step": 116
+    },
+    {
+      "epoch": 0.05446927374301676,
+      "grad_norm": 0.6041114244929616,
+      "learning_rate": 1.8111455108359136e-06,
+      "loss": 0.5491,
+      "step": 117
+    },
+    {
+      "epoch": 0.05493482309124767,
+      "grad_norm": 0.6203101491746932,
+      "learning_rate": 1.8266253869969042e-06,
+      "loss": 0.5273,
+      "step": 118
+    },
+    {
+      "epoch": 0.05540037243947858,
+      "grad_norm": 0.6614582574586371,
+      "learning_rate": 1.8421052631578948e-06,
+      "loss": 0.5388,
+      "step": 119
+    },
+    {
+      "epoch": 0.055865921787709494,
+      "grad_norm": 0.6132509149876961,
+      "learning_rate": 1.8575851393188857e-06,
+      "loss": 0.5233,
+      "step": 120
+    },
+    {
+      "epoch": 0.05633147113594041,
+      "grad_norm": 0.6228086642025954,
+      "learning_rate": 1.8730650154798763e-06,
+      "loss": 0.5073,
+      "step": 121
+    },
+    {
+      "epoch": 0.056797020484171325,
+      "grad_norm": 0.6391184445175822,
+      "learning_rate": 1.888544891640867e-06,
+      "loss": 0.5471,
+      "step": 122
+    },
+    {
+      "epoch": 0.05726256983240224,
+      "grad_norm": 0.6597330018091992,
+      "learning_rate": 1.9040247678018578e-06,
+      "loss": 0.5317,
+      "step": 123
+    },
+    {
+      "epoch": 0.05772811918063315,
+      "grad_norm": 0.6472785905751132,
+      "learning_rate": 1.9195046439628486e-06,
+      "loss": 0.5342,
+      "step": 124
+    },
+    {
+      "epoch": 0.05819366852886406,
+      "grad_norm": 0.6411599103959432,
+      "learning_rate": 1.9349845201238395e-06,
+      "loss": 0.5283,
+      "step": 125
+    },
+    {
+      "epoch": 0.05865921787709497,
+      "grad_norm": 0.6911222053032953,
+      "learning_rate": 1.95046439628483e-06,
+      "loss": 0.5215,
+      "step": 126
+    },
+    {
+      "epoch": 0.059124767225325885,
+      "grad_norm": 0.6632556065417751,
+      "learning_rate": 1.9659442724458207e-06,
+      "loss": 0.518,
+      "step": 127
+    },
+    {
+      "epoch": 0.0595903165735568,
+      "grad_norm": 0.6985940643828608,
+      "learning_rate": 1.9814241486068116e-06,
+      "loss": 0.5347,
+      "step": 128
+    },
+    {
+      "epoch": 0.06005586592178771,
+      "grad_norm": 0.6389492743580009,
+      "learning_rate": 1.996904024767802e-06,
+      "loss": 0.5408,
+      "step": 129
+    },
+    {
+      "epoch": 0.06052141527001862,
+      "grad_norm": 0.662049275428993,
+      "learning_rate": 2.012383900928793e-06,
+      "loss": 0.5113,
+      "step": 130
+    },
+    {
+      "epoch": 0.06098696461824953,
+      "grad_norm": 0.6233493112047304,
+      "learning_rate": 2.0278637770897837e-06,
+      "loss": 0.5163,
+      "step": 131
+    },
+    {
+      "epoch": 0.061452513966480445,
+      "grad_norm": 0.6691313944763179,
+      "learning_rate": 2.043343653250774e-06,
+      "loss": 0.5518,
+      "step": 132
+    },
+    {
+      "epoch": 0.06191806331471136,
+      "grad_norm": 0.7149450740838151,
+      "learning_rate": 2.058823529411765e-06,
+      "loss": 0.5289,
+      "step": 133
+    },
+    {
+      "epoch": 0.06238361266294227,
+      "grad_norm": 0.7249416220212493,
+      "learning_rate": 2.074303405572756e-06,
+      "loss": 0.5367,
+      "step": 134
+    },
+    {
+      "epoch": 0.06284916201117319,
+      "grad_norm": 0.6405999066005678,
+      "learning_rate": 2.0897832817337462e-06,
+      "loss": 0.5217,
+      "step": 135
+    },
+    {
+      "epoch": 0.0633147113594041,
+      "grad_norm": 0.6200498171246361,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 0.5255,
+      "step": 136
+    },
+    {
+      "epoch": 0.06378026070763501,
+      "grad_norm": 0.6896144482728684,
+      "learning_rate": 2.120743034055728e-06,
+      "loss": 0.5434,
+      "step": 137
+    },
+    {
+      "epoch": 0.06424581005586592,
+      "grad_norm": 0.6780928522001057,
+      "learning_rate": 2.1362229102167183e-06,
+      "loss": 0.5249,
+      "step": 138
+    },
+    {
+      "epoch": 0.06471135940409684,
+      "grad_norm": 0.703141896091384,
+      "learning_rate": 2.151702786377709e-06,
+      "loss": 0.508,
+      "step": 139
+    },
+    {
+      "epoch": 0.06517690875232775,
+      "grad_norm": 0.6384608914359078,
+      "learning_rate": 2.1671826625387e-06,
+      "loss": 0.4985,
+      "step": 140
+    },
+    {
+      "epoch": 0.06564245810055866,
+      "grad_norm": 0.666489363078141,
+      "learning_rate": 2.1826625386996904e-06,
+      "loss": 0.5432,
+      "step": 141
+    },
+    {
+      "epoch": 0.06610800744878957,
+      "grad_norm": 0.6072351472047566,
+      "learning_rate": 2.1981424148606813e-06,
+      "loss": 0.5429,
+      "step": 142
+    },
+    {
+      "epoch": 0.06657355679702048,
+      "grad_norm": 0.6289577186251796,
+      "learning_rate": 2.213622291021672e-06,
+      "loss": 0.5188,
+      "step": 143
+    },
+    {
+      "epoch": 0.0670391061452514,
+      "grad_norm": 0.6933724474215712,
+      "learning_rate": 2.2291021671826625e-06,
+      "loss": 0.5311,
+      "step": 144
+    },
+    {
+      "epoch": 0.06750465549348231,
+      "grad_norm": 0.6121205885059605,
+      "learning_rate": 2.2445820433436534e-06,
+      "loss": 0.5275,
+      "step": 145
+    },
+    {
+      "epoch": 0.06797020484171322,
+      "grad_norm": 0.7063634578184468,
+      "learning_rate": 2.2600619195046442e-06,
+      "loss": 0.5409,
+      "step": 146
+    },
+    {
+      "epoch": 0.06843575418994413,
+      "grad_norm": 0.739985919570153,
+      "learning_rate": 2.2755417956656347e-06,
+      "loss": 0.5088,
+      "step": 147
+    },
+    {
+      "epoch": 0.06890130353817504,
+      "grad_norm": 0.6442281539712866,
+      "learning_rate": 2.2910216718266255e-06,
+      "loss": 0.5147,
+      "step": 148
+    },
+    {
+      "epoch": 0.06936685288640596,
+      "grad_norm": 0.6480534488589174,
+      "learning_rate": 2.3065015479876163e-06,
+      "loss": 0.529,
+      "step": 149
+    },
+    {
+      "epoch": 0.06983240223463687,
+      "grad_norm": 0.7124060234322813,
+      "learning_rate": 2.3219814241486068e-06,
+      "loss": 0.536,
+      "step": 150
+    },
+    {
+      "epoch": 0.07029795158286778,
+      "grad_norm": 0.6561706094766784,
+      "learning_rate": 2.3374613003095976e-06,
+      "loss": 0.5398,
+      "step": 151
+    },
+    {
+      "epoch": 0.07076350093109869,
+      "grad_norm": 0.6327947831783897,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.5267,
+      "step": 152
+    },
+    {
+      "epoch": 0.0712290502793296,
+      "grad_norm": 0.6772073388107503,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 0.5206,
+      "step": 153
+    },
+    {
+      "epoch": 0.07169459962756052,
+      "grad_norm": 0.6514017462499977,
+      "learning_rate": 2.3839009287925697e-06,
+      "loss": 0.5143,
+      "step": 154
+    },
+    {
+      "epoch": 0.07216014897579143,
+      "grad_norm": 0.6607262248249812,
+      "learning_rate": 2.3993808049535606e-06,
+      "loss": 0.5261,
+      "step": 155
+    },
+    {
+      "epoch": 0.07262569832402235,
+      "grad_norm": 0.7393472839113571,
+      "learning_rate": 2.4148606811145514e-06,
+      "loss": 0.519,
+      "step": 156
+    },
+    {
+      "epoch": 0.07309124767225327,
+      "grad_norm": 0.6276201096593895,
+      "learning_rate": 2.430340557275542e-06,
+      "loss": 0.5454,
+      "step": 157
+    },
+    {
+      "epoch": 0.07355679702048418,
+      "grad_norm": 0.6601284950696961,
+      "learning_rate": 2.4458204334365327e-06,
+      "loss": 0.5103,
+      "step": 158
+    },
+    {
+      "epoch": 0.07402234636871509,
+      "grad_norm": 0.6439383869328791,
+      "learning_rate": 2.4613003095975235e-06,
+      "loss": 0.4969,
+      "step": 159
+    },
+    {
+      "epoch": 0.074487895716946,
+      "grad_norm": 0.6084236642629911,
+      "learning_rate": 2.476780185758514e-06,
+      "loss": 0.5265,
+      "step": 160
+    },
+    {
+      "epoch": 0.07495344506517691,
+      "grad_norm": 0.7086114964933651,
+      "learning_rate": 2.4922600619195048e-06,
+      "loss": 0.5309,
+      "step": 161
+    },
+    {
+      "epoch": 0.07541899441340782,
+      "grad_norm": 0.6433273694541162,
+      "learning_rate": 2.5077399380804956e-06,
+      "loss": 0.5201,
+      "step": 162
+    },
+    {
+      "epoch": 0.07588454376163874,
+      "grad_norm": 0.6357818090987353,
+      "learning_rate": 2.5232198142414865e-06,
+      "loss": 0.5267,
+      "step": 163
+    },
+    {
+      "epoch": 0.07635009310986965,
+      "grad_norm": 0.6931627220347318,
+      "learning_rate": 2.538699690402477e-06,
+      "loss": 0.5006,
+      "step": 164
+    },
+    {
+      "epoch": 0.07681564245810056,
+      "grad_norm": 0.6392003808969955,
+      "learning_rate": 2.5541795665634677e-06,
+      "loss": 0.5151,
+      "step": 165
+    },
+    {
+      "epoch": 0.07728119180633147,
+      "grad_norm": 0.668216129981097,
+      "learning_rate": 2.5696594427244586e-06,
+      "loss": 0.5166,
+      "step": 166
+    },
+    {
+      "epoch": 0.07774674115456238,
+      "grad_norm": 0.6322906137784438,
+      "learning_rate": 2.585139318885449e-06,
+      "loss": 0.523,
+      "step": 167
+    },
+    {
+      "epoch": 0.0782122905027933,
+      "grad_norm": 0.6407821243031374,
+      "learning_rate": 2.60061919504644e-06,
+      "loss": 0.5193,
+      "step": 168
+    },
+    {
+      "epoch": 0.07867783985102421,
+      "grad_norm": 0.6890043933304311,
+      "learning_rate": 2.6160990712074307e-06,
+      "loss": 0.5291,
+      "step": 169
+    },
+    {
+      "epoch": 0.07914338919925512,
+      "grad_norm": 0.6353660283409324,
+      "learning_rate": 2.631578947368421e-06,
+      "loss": 0.5016,
+      "step": 170
+    },
+    {
+      "epoch": 0.07960893854748603,
+      "grad_norm": 0.6529531231880904,
+      "learning_rate": 2.647058823529412e-06,
+      "loss": 0.484,
+      "step": 171
+    },
+    {
+      "epoch": 0.08007448789571694,
+      "grad_norm": 0.709553183402242,
+      "learning_rate": 2.662538699690403e-06,
+      "loss": 0.5179,
+      "step": 172
+    },
+    {
+      "epoch": 0.08054003724394786,
+      "grad_norm": 0.6929636405871185,
+      "learning_rate": 2.6780185758513932e-06,
+      "loss": 0.5386,
+      "step": 173
+    },
+    {
+      "epoch": 0.08100558659217877,
+      "grad_norm": 0.6622376836978119,
+      "learning_rate": 2.693498452012384e-06,
+      "loss": 0.5166,
+      "step": 174
+    },
+    {
+      "epoch": 0.08147113594040968,
+      "grad_norm": 0.6542782078856054,
+      "learning_rate": 2.708978328173375e-06,
+      "loss": 0.5221,
+      "step": 175
+    },
+    {
+      "epoch": 0.08193668528864059,
+      "grad_norm": 0.6391762697733179,
+      "learning_rate": 2.7244582043343653e-06,
+      "loss": 0.5152,
+      "step": 176
+    },
+    {
+      "epoch": 0.0824022346368715,
+      "grad_norm": 0.6629125484497218,
+      "learning_rate": 2.739938080495356e-06,
+      "loss": 0.539,
+      "step": 177
+    },
+    {
+      "epoch": 0.08286778398510242,
+      "grad_norm": 0.6599356049160929,
+      "learning_rate": 2.755417956656347e-06,
+      "loss": 0.5076,
+      "step": 178
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 0.6397452530547824,
+      "learning_rate": 2.7708978328173374e-06,
+      "loss": 0.5415,
+      "step": 179
+    },
+    {
+      "epoch": 0.08379888268156424,
+      "grad_norm": 0.6508893284318623,
+      "learning_rate": 2.7863777089783283e-06,
+      "loss": 0.5044,
+      "step": 180
+    },
+    {
+      "epoch": 0.08426443202979515,
+      "grad_norm": 0.659151246105914,
+      "learning_rate": 2.801857585139319e-06,
+      "loss": 0.5067,
+      "step": 181
+    },
+    {
+      "epoch": 0.08472998137802606,
+      "grad_norm": 0.6803816636055867,
+      "learning_rate": 2.8173374613003096e-06,
+      "loss": 0.5222,
+      "step": 182
+    },
+    {
+      "epoch": 0.08519553072625698,
+      "grad_norm": 0.6592560735388285,
+      "learning_rate": 2.8328173374613004e-06,
+      "loss": 0.5147,
+      "step": 183
+    },
+    {
+      "epoch": 0.0856610800744879,
+      "grad_norm": 0.6749495867515741,
+      "learning_rate": 2.8482972136222912e-06,
+      "loss": 0.5074,
+      "step": 184
+    },
+    {
+      "epoch": 0.08612662942271881,
+      "grad_norm": 0.6433604414273647,
+      "learning_rate": 2.8637770897832817e-06,
+      "loss": 0.5055,
+      "step": 185
+    },
+    {
+      "epoch": 0.08659217877094973,
+      "grad_norm": 0.6502849194337784,
+      "learning_rate": 2.8792569659442725e-06,
+      "loss": 0.5253,
+      "step": 186
+    },
+    {
+      "epoch": 0.08705772811918064,
+      "grad_norm": 0.6533600590485262,
+      "learning_rate": 2.8947368421052634e-06,
+      "loss": 0.5142,
+      "step": 187
+    },
+    {
+      "epoch": 0.08752327746741155,
+      "grad_norm": 0.6950324031389645,
+      "learning_rate": 2.9102167182662538e-06,
+      "loss": 0.5303,
+      "step": 188
+    },
+    {
+      "epoch": 0.08798882681564246,
+      "grad_norm": 0.6404695102000653,
+      "learning_rate": 2.9256965944272446e-06,
+      "loss": 0.5254,
+      "step": 189
+    },
+    {
+      "epoch": 0.08845437616387337,
+      "grad_norm": 0.6303325618100977,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.5171,
+      "step": 190
+    },
+    {
+      "epoch": 0.08891992551210429,
+      "grad_norm": 0.6478460935798706,
+      "learning_rate": 2.9566563467492267e-06,
+      "loss": 0.5043,
+      "step": 191
+    },
+    {
+      "epoch": 0.0893854748603352,
+      "grad_norm": 0.6584382455144241,
+      "learning_rate": 2.9721362229102167e-06,
+      "loss": 0.515,
+      "step": 192
+    },
+    {
+      "epoch": 0.08985102420856611,
+      "grad_norm": 0.687609276618317,
+      "learning_rate": 2.9876160990712076e-06,
+      "loss": 0.5289,
+      "step": 193
+    },
+    {
+      "epoch": 0.09031657355679702,
+      "grad_norm": 0.6649398074628041,
+      "learning_rate": 3.003095975232199e-06,
+      "loss": 0.511,
+      "step": 194
+    },
+    {
+      "epoch": 0.09078212290502793,
+      "grad_norm": 0.6780933277709346,
+      "learning_rate": 3.018575851393189e-06,
+      "loss": 0.5369,
+      "step": 195
+    },
+    {
+      "epoch": 0.09124767225325885,
+      "grad_norm": 0.6943521546172964,
+      "learning_rate": 3.0340557275541797e-06,
+      "loss": 0.5242,
+      "step": 196
+    },
+    {
+      "epoch": 0.09171322160148976,
+      "grad_norm": 0.6499301184934378,
+      "learning_rate": 3.049535603715171e-06,
+      "loss": 0.5077,
+      "step": 197
+    },
+    {
+      "epoch": 0.09217877094972067,
+      "grad_norm": 0.7124156955138994,
+      "learning_rate": 3.065015479876161e-06,
+      "loss": 0.5188,
+      "step": 198
+    },
+    {
+      "epoch": 0.09264432029795158,
+      "grad_norm": 0.6851850619146651,
+      "learning_rate": 3.080495356037152e-06,
+      "loss": 0.5255,
+      "step": 199
+    },
+    {
+      "epoch": 0.0931098696461825,
+      "grad_norm": 0.6998886274436751,
+      "learning_rate": 3.095975232198143e-06,
+      "loss": 0.4991,
+      "step": 200
+    },
+    {
+      "epoch": 0.0935754189944134,
+      "grad_norm": 0.673044814790283,
+      "learning_rate": 3.111455108359133e-06,
+      "loss": 0.524,
+      "step": 201
+    },
+    {
+      "epoch": 0.09404096834264432,
+      "grad_norm": 0.6174749684813496,
+      "learning_rate": 3.1269349845201243e-06,
+      "loss": 0.513,
+      "step": 202
+    },
+    {
+      "epoch": 0.09450651769087523,
+      "grad_norm": 0.6293665102961191,
+      "learning_rate": 3.142414860681115e-06,
+      "loss": 0.5162,
+      "step": 203
+    },
+    {
+      "epoch": 0.09497206703910614,
+      "grad_norm": 0.6658006718629849,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 0.5046,
+      "step": 204
+    },
+    {
+      "epoch": 0.09543761638733705,
+      "grad_norm": 0.6475422123633591,
+      "learning_rate": 3.1733746130030964e-06,
+      "loss": 0.5326,
+      "step": 205
+    },
+    {
+      "epoch": 0.09590316573556797,
+      "grad_norm": 0.6636012423061434,
+      "learning_rate": 3.1888544891640873e-06,
+      "loss": 0.5318,
+      "step": 206
+    },
+    {
+      "epoch": 0.09636871508379888,
+      "grad_norm": 0.6060300066627223,
+      "learning_rate": 3.2043343653250773e-06,
+      "loss": 0.4974,
+      "step": 207
+    },
+    {
+      "epoch": 0.09683426443202979,
+      "grad_norm": 0.7219617195345138,
+      "learning_rate": 3.2198142414860685e-06,
+      "loss": 0.5249,
+      "step": 208
+    },
+    {
+      "epoch": 0.0972998137802607,
+      "grad_norm": 0.6805752871929227,
+      "learning_rate": 3.2352941176470594e-06,
+      "loss": 0.5226,
+      "step": 209
+    },
+    {
+      "epoch": 0.09776536312849161,
+      "grad_norm": 0.667318165248406,
+      "learning_rate": 3.2507739938080494e-06,
+      "loss": 0.5226,
+      "step": 210
+    },
+    {
+      "epoch": 0.09823091247672253,
+      "grad_norm": 0.6092347534993429,
+      "learning_rate": 3.2662538699690407e-06,
+      "loss": 0.4996,
+      "step": 211
+    },
+    {
+      "epoch": 0.09869646182495345,
+      "grad_norm": 0.5979333824039105,
+      "learning_rate": 3.2817337461300315e-06,
+      "loss": 0.5073,
+      "step": 212
+    },
+    {
+      "epoch": 0.09916201117318436,
+      "grad_norm": 0.6283695557057687,
+      "learning_rate": 3.297213622291022e-06,
+      "loss": 0.5268,
+      "step": 213
+    },
+    {
+      "epoch": 0.09962756052141528,
+      "grad_norm": 0.6636638998267982,
+      "learning_rate": 3.3126934984520128e-06,
+      "loss": 0.5103,
+      "step": 214
+    },
+    {
+      "epoch": 0.10009310986964619,
+      "grad_norm": 0.6505174406136378,
+      "learning_rate": 3.3281733746130036e-06,
+      "loss": 0.5166,
+      "step": 215
+    },
+    {
+      "epoch": 0.1005586592178771,
+      "grad_norm": 0.6189439849489848,
+      "learning_rate": 3.343653250773994e-06,
+      "loss": 0.5262,
+      "step": 216
+    },
+    {
+      "epoch": 0.10102420856610801,
+      "grad_norm": 0.6534934194368058,
+      "learning_rate": 3.359133126934985e-06,
+      "loss": 0.5092,
+      "step": 217
+    },
+    {
+      "epoch": 0.10148975791433892,
+      "grad_norm": 0.6480284081656406,
+      "learning_rate": 3.3746130030959757e-06,
+      "loss": 0.5041,
+      "step": 218
+    },
+    {
+      "epoch": 0.10195530726256984,
+      "grad_norm": 0.6771925949186104,
+      "learning_rate": 3.390092879256966e-06,
+      "loss": 0.5034,
+      "step": 219
+    },
+    {
+      "epoch": 0.10242085661080075,
+      "grad_norm": 0.6254280856101838,
+      "learning_rate": 3.405572755417957e-06,
+      "loss": 0.4951,
+      "step": 220
+    },
+    {
+      "epoch": 0.10288640595903166,
+      "grad_norm": 0.6520965098004818,
+      "learning_rate": 3.421052631578948e-06,
+      "loss": 0.5051,
+      "step": 221
+    },
+    {
+      "epoch": 0.10335195530726257,
+      "grad_norm": 0.6254609523284526,
+      "learning_rate": 3.4365325077399387e-06,
+      "loss": 0.5282,
+      "step": 222
+    },
+    {
+      "epoch": 0.10381750465549348,
+      "grad_norm": 0.654370654633827,
+      "learning_rate": 3.452012383900929e-06,
+      "loss": 0.4729,
+      "step": 223
+    },
+    {
+      "epoch": 0.1042830540037244,
+      "grad_norm": 0.6131882901432193,
+      "learning_rate": 3.46749226006192e-06,
+      "loss": 0.5099,
+      "step": 224
+    },
+    {
+      "epoch": 0.10474860335195531,
+      "grad_norm": 0.6726858414132164,
+      "learning_rate": 3.4829721362229108e-06,
+      "loss": 0.5184,
+      "step": 225
+    },
+    {
+      "epoch": 0.10521415270018622,
+      "grad_norm": 0.6028253993089268,
+      "learning_rate": 3.498452012383901e-06,
+      "loss": 0.485,
+      "step": 226
+    },
+    {
+      "epoch": 0.10567970204841713,
+      "grad_norm": 0.7595198192709417,
+      "learning_rate": 3.513931888544892e-06,
+      "loss": 0.5141,
+      "step": 227
+    },
+    {
+      "epoch": 0.10614525139664804,
+      "grad_norm": 0.6347533737700222,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 0.5245,
+      "step": 228
+    },
+    {
+      "epoch": 0.10661080074487896,
+      "grad_norm": 0.6686985540971694,
+      "learning_rate": 3.5448916408668733e-06,
+      "loss": 0.5212,
+      "step": 229
+    },
+    {
+      "epoch": 0.10707635009310987,
+      "grad_norm": 0.6584892320855432,
+      "learning_rate": 3.560371517027864e-06,
+      "loss": 0.5095,
+      "step": 230
+    },
+    {
+      "epoch": 0.10754189944134078,
+      "grad_norm": 0.6351776573657688,
+      "learning_rate": 3.575851393188855e-06,
+      "loss": 0.5132,
+      "step": 231
+    },
+    {
+      "epoch": 0.10800744878957169,
+      "grad_norm": 0.6401789821740734,
+      "learning_rate": 3.5913312693498454e-06,
+      "loss": 0.5168,
+      "step": 232
+    },
+    {
+      "epoch": 0.1084729981378026,
+      "grad_norm": 0.6919531137894394,
+      "learning_rate": 3.6068111455108363e-06,
+      "loss": 0.5359,
+      "step": 233
+    },
+    {
+      "epoch": 0.10893854748603352,
+      "grad_norm": 0.6559190186160263,
+      "learning_rate": 3.622291021671827e-06,
+      "loss": 0.5015,
+      "step": 234
+    },
+    {
+      "epoch": 0.10940409683426443,
+      "grad_norm": 0.6710196697166376,
+      "learning_rate": 3.6377708978328175e-06,
+      "loss": 0.501,
+      "step": 235
+    },
+    {
+      "epoch": 0.10986964618249534,
+      "grad_norm": 0.6569784680366509,
+      "learning_rate": 3.6532507739938084e-06,
+      "loss": 0.52,
+      "step": 236
+    },
+    {
+      "epoch": 0.11033519553072625,
+      "grad_norm": 0.7305168183177061,
+      "learning_rate": 3.6687306501547992e-06,
+      "loss": 0.5082,
+      "step": 237
+    },
+    {
+      "epoch": 0.11080074487895716,
+      "grad_norm": 0.6107498574283965,
+      "learning_rate": 3.6842105263157896e-06,
+      "loss": 0.4921,
+      "step": 238
+    },
+    {
+      "epoch": 0.11126629422718808,
+      "grad_norm": 0.6270160329399098,
+      "learning_rate": 3.6996904024767805e-06,
+      "loss": 0.4844,
+      "step": 239
+    },
+    {
+      "epoch": 0.11173184357541899,
+      "grad_norm": 0.7076279215434673,
+      "learning_rate": 3.7151702786377713e-06,
+      "loss": 0.5198,
+      "step": 240
+    },
+    {
+      "epoch": 0.11219739292364991,
+      "grad_norm": 0.6343523612061386,
+      "learning_rate": 3.7306501547987618e-06,
+      "loss": 0.4948,
+      "step": 241
+    },
+    {
+      "epoch": 0.11266294227188083,
+      "grad_norm": 0.6506913157541835,
+      "learning_rate": 3.7461300309597526e-06,
+      "loss": 0.4981,
+      "step": 242
+    },
+    {
+      "epoch": 0.11312849162011174,
+      "grad_norm": 0.652743479086599,
+      "learning_rate": 3.7616099071207434e-06,
+      "loss": 0.5046,
+      "step": 243
+    },
+    {
+      "epoch": 0.11359404096834265,
+      "grad_norm": 0.7583734022423179,
+      "learning_rate": 3.777089783281734e-06,
+      "loss": 0.5141,
+      "step": 244
+    },
+    {
+      "epoch": 0.11405959031657356,
+      "grad_norm": 0.7046092401744204,
+      "learning_rate": 3.7925696594427247e-06,
+      "loss": 0.5135,
+      "step": 245
+    },
+    {
+      "epoch": 0.11452513966480447,
+      "grad_norm": 0.6583140185160512,
+      "learning_rate": 3.8080495356037155e-06,
+      "loss": 0.5142,
+      "step": 246
+    },
+    {
+      "epoch": 0.11499068901303539,
+      "grad_norm": 0.671365414631487,
+      "learning_rate": 3.8235294117647055e-06,
+      "loss": 0.512,
+      "step": 247
+    },
+    {
+      "epoch": 0.1154562383612663,
+      "grad_norm": 0.6147103980616719,
+      "learning_rate": 3.839009287925697e-06,
+      "loss": 0.4886,
+      "step": 248
+    },
+    {
+      "epoch": 0.11592178770949721,
+      "grad_norm": 0.7117294870346428,
+      "learning_rate": 3.854489164086688e-06,
+      "loss": 0.4936,
+      "step": 249
+    },
+    {
+      "epoch": 0.11638733705772812,
+      "grad_norm": 0.6471003204983427,
+      "learning_rate": 3.869969040247679e-06,
+      "loss": 0.4836,
+      "step": 250
+    },
+    {
+      "epoch": 0.11685288640595903,
+      "grad_norm": 0.6607391810469527,
+      "learning_rate": 3.885448916408669e-06,
+      "loss": 0.504,
+      "step": 251
+    },
+    {
+      "epoch": 0.11731843575418995,
+      "grad_norm": 0.7239843719104224,
+      "learning_rate": 3.90092879256966e-06,
+      "loss": 0.4932,
+      "step": 252
+    },
+    {
+      "epoch": 0.11778398510242086,
+      "grad_norm": 0.6819369960692606,
+      "learning_rate": 3.916408668730651e-06,
+      "loss": 0.5088,
+      "step": 253
+    },
+    {
+      "epoch": 0.11824953445065177,
+      "grad_norm": 0.6688917913895226,
+      "learning_rate": 3.9318885448916415e-06,
+      "loss": 0.522,
+      "step": 254
+    },
+    {
+      "epoch": 0.11871508379888268,
+      "grad_norm": 0.6288219878761805,
+      "learning_rate": 3.947368421052632e-06,
+      "loss": 0.4886,
+      "step": 255
+    },
+    {
+      "epoch": 0.1191806331471136,
+      "grad_norm": 0.6832119126933617,
+      "learning_rate": 3.962848297213623e-06,
+      "loss": 0.4915,
+      "step": 256
+    },
+    {
+      "epoch": 0.1196461824953445,
+      "grad_norm": 0.6621458500784531,
+      "learning_rate": 3.978328173374613e-06,
+      "loss": 0.5067,
+      "step": 257
+    },
+    {
+      "epoch": 0.12011173184357542,
+      "grad_norm": 0.6873499970564896,
+      "learning_rate": 3.993808049535604e-06,
+      "loss": 0.5245,
+      "step": 258
+    },
+    {
+      "epoch": 0.12057728119180633,
+      "grad_norm": 0.6457783547714737,
+      "learning_rate": 4.009287925696595e-06,
+      "loss": 0.5165,
+      "step": 259
+    },
+    {
+      "epoch": 0.12104283054003724,
+      "grad_norm": 0.7261427577233803,
+      "learning_rate": 4.024767801857586e-06,
+      "loss": 0.4907,
+      "step": 260
+    },
+    {
+      "epoch": 0.12150837988826815,
+      "grad_norm": 0.6889294767152678,
+      "learning_rate": 4.0402476780185765e-06,
+      "loss": 0.4919,
+      "step": 261
+    },
+    {
+      "epoch": 0.12197392923649907,
+      "grad_norm": 0.651103152132562,
+      "learning_rate": 4.055727554179567e-06,
+      "loss": 0.4943,
+      "step": 262
+    },
+    {
+      "epoch": 0.12243947858472998,
+      "grad_norm": 0.6327991390902421,
+      "learning_rate": 4.071207430340557e-06,
+      "loss": 0.5098,
+      "step": 263
+    },
+    {
+      "epoch": 0.12290502793296089,
+      "grad_norm": 0.6418085630041567,
+      "learning_rate": 4.086687306501548e-06,
+      "loss": 0.5065,
+      "step": 264
+    },
+    {
+      "epoch": 0.1233705772811918,
+      "grad_norm": 0.7147632545178161,
+      "learning_rate": 4.102167182662539e-06,
+      "loss": 0.505,
+      "step": 265
+    },
+    {
+      "epoch": 0.12383612662942271,
+      "grad_norm": 0.6451576491670913,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.5321,
+      "step": 266
+    },
+    {
+      "epoch": 0.12430167597765363,
+      "grad_norm": 0.727622408346516,
+      "learning_rate": 4.133126934984521e-06,
+      "loss": 0.5211,
+      "step": 267
+    },
+    {
+      "epoch": 0.12476722532588454,
+      "grad_norm": 0.6562824491012875,
+      "learning_rate": 4.148606811145512e-06,
+      "loss": 0.4928,
+      "step": 268
+    },
+    {
+      "epoch": 0.12523277467411545,
+      "grad_norm": 0.6706442530290604,
+      "learning_rate": 4.164086687306502e-06,
+      "loss": 0.5204,
+      "step": 269
+    },
+    {
+      "epoch": 0.12569832402234637,
+      "grad_norm": 0.67778467214186,
+      "learning_rate": 4.1795665634674924e-06,
+      "loss": 0.5049,
+      "step": 270
+    },
+    {
+      "epoch": 0.12616387337057727,
+      "grad_norm": 0.7085865696606708,
+      "learning_rate": 4.195046439628483e-06,
+      "loss": 0.5014,
+      "step": 271
+    },
+    {
+      "epoch": 0.1266294227188082,
+      "grad_norm": 0.6585038582064933,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 0.4858,
+      "step": 272
+    },
+    {
+      "epoch": 0.1270949720670391,
+      "grad_norm": 0.6474206162289781,
+      "learning_rate": 4.226006191950465e-06,
+      "loss": 0.4994,
+      "step": 273
+    },
+    {
+      "epoch": 0.12756052141527002,
+      "grad_norm": 0.656029972746577,
+      "learning_rate": 4.241486068111456e-06,
+      "loss": 0.4921,
+      "step": 274
+    },
+    {
+      "epoch": 0.12802607076350092,
+      "grad_norm": 0.7106191818493574,
+      "learning_rate": 4.256965944272446e-06,
+      "loss": 0.5088,
+      "step": 275
+    },
+    {
+      "epoch": 0.12849162011173185,
+      "grad_norm": 0.6880458850644549,
+      "learning_rate": 4.272445820433437e-06,
+      "loss": 0.498,
+      "step": 276
+    },
+    {
+      "epoch": 0.12895716945996274,
+      "grad_norm": 0.6963145499811602,
+      "learning_rate": 4.2879256965944275e-06,
+      "loss": 0.5299,
+      "step": 277
+    },
+    {
+      "epoch": 0.12942271880819367,
+      "grad_norm": 0.6515509885218885,
+      "learning_rate": 4.303405572755418e-06,
+      "loss": 0.4824,
+      "step": 278
+    },
+    {
+      "epoch": 0.12988826815642457,
+      "grad_norm": 0.681522748828664,
+      "learning_rate": 4.318885448916409e-06,
+      "loss": 0.4864,
+      "step": 279
+    },
+    {
+      "epoch": 0.1303538175046555,
+      "grad_norm": 0.6883827576718019,
+      "learning_rate": 4.3343653250774e-06,
+      "loss": 0.5138,
+      "step": 280
+    },
+    {
+      "epoch": 0.1308193668528864,
+      "grad_norm": 0.7104810035153587,
+      "learning_rate": 4.349845201238391e-06,
+      "loss": 0.4998,
+      "step": 281
+    },
+    {
+      "epoch": 0.13128491620111732,
+      "grad_norm": 0.6801816721757588,
+      "learning_rate": 4.365325077399381e-06,
+      "loss": 0.4844,
+      "step": 282
+    },
+    {
+      "epoch": 0.13175046554934824,
+      "grad_norm": 0.6716431200595535,
+      "learning_rate": 4.380804953560372e-06,
+      "loss": 0.4812,
+      "step": 283
+    },
+    {
+      "epoch": 0.13221601489757914,
+      "grad_norm": 0.6564536718587851,
+      "learning_rate": 4.3962848297213626e-06,
+      "loss": 0.4778,
+      "step": 284
+    },
+    {
+      "epoch": 0.13268156424581007,
+      "grad_norm": 0.6577415181950953,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 0.5219,
+      "step": 285
+    },
+    {
+      "epoch": 0.13314711359404097,
+      "grad_norm": 0.6668219655637413,
+      "learning_rate": 4.427244582043344e-06,
+      "loss": 0.4955,
+      "step": 286
+    },
+    {
+      "epoch": 0.1336126629422719,
+      "grad_norm": 0.678354592318351,
+      "learning_rate": 4.442724458204335e-06,
+      "loss": 0.4967,
+      "step": 287
+    },
+    {
+      "epoch": 0.1340782122905028,
+      "grad_norm": 0.6722643275139306,
+      "learning_rate": 4.458204334365325e-06,
+      "loss": 0.5199,
+      "step": 288
+    },
+    {
+      "epoch": 0.13454376163873372,
+      "grad_norm": 0.7119662244692079,
+      "learning_rate": 4.473684210526316e-06,
+      "loss": 0.4821,
+      "step": 289
+    },
+    {
+      "epoch": 0.13500931098696461,
+      "grad_norm": 0.6844190799590407,
+      "learning_rate": 4.489164086687307e-06,
+      "loss": 0.5126,
+      "step": 290
+    },
+    {
+      "epoch": 0.13547486033519554,
+      "grad_norm": 0.6091939915920533,
+      "learning_rate": 4.504643962848298e-06,
+      "loss": 0.4775,
+      "step": 291
+    },
+    {
+      "epoch": 0.13594040968342644,
+      "grad_norm": 0.6663936681113477,
+      "learning_rate": 4.5201238390092885e-06,
+      "loss": 0.5052,
+      "step": 292
+    },
+    {
+      "epoch": 0.13640595903165736,
+      "grad_norm": 0.6527056354608161,
+      "learning_rate": 4.535603715170279e-06,
+      "loss": 0.5123,
+      "step": 293
+    },
+    {
+      "epoch": 0.13687150837988826,
+      "grad_norm": 0.6471291581021584,
+      "learning_rate": 4.551083591331269e-06,
+      "loss": 0.4875,
+      "step": 294
+    },
+    {
+      "epoch": 0.1373370577281192,
+      "grad_norm": 0.6656967145976497,
+      "learning_rate": 4.56656346749226e-06,
+      "loss": 0.4997,
+      "step": 295
+    },
+    {
+      "epoch": 0.1378026070763501,
+      "grad_norm": 0.6715230345847547,
+      "learning_rate": 4.582043343653251e-06,
+      "loss": 0.52,
+      "step": 296
+    },
+    {
+      "epoch": 0.138268156424581,
+      "grad_norm": 0.6979383060725824,
+      "learning_rate": 4.597523219814242e-06,
+      "loss": 0.5161,
+      "step": 297
+    },
+    {
+      "epoch": 0.1387337057728119,
+      "grad_norm": 0.6441345604379588,
+      "learning_rate": 4.613003095975233e-06,
+      "loss": 0.4972,
+      "step": 298
+    },
+    {
+      "epoch": 0.13919925512104284,
+      "grad_norm": 0.6566783378949751,
+      "learning_rate": 4.6284829721362235e-06,
+      "loss": 0.5017,
+      "step": 299
+    },
+    {
+      "epoch": 0.13966480446927373,
+      "grad_norm": 0.6810237716925174,
+      "learning_rate": 4.6439628482972135e-06,
+      "loss": 0.5141,
+      "step": 300
+    },
+    {
+      "epoch": 0.14013035381750466,
+      "grad_norm": 0.6412338562241248,
+      "learning_rate": 4.659442724458204e-06,
+      "loss": 0.5099,
+      "step": 301
+    },
+    {
+      "epoch": 0.14059590316573556,
+      "grad_norm": 0.686975775182641,
+      "learning_rate": 4.674922600619195e-06,
+      "loss": 0.5181,
+      "step": 302
+    },
+    {
+      "epoch": 0.14106145251396648,
+      "grad_norm": 0.6838838638538765,
+      "learning_rate": 4.690402476780186e-06,
+      "loss": 0.4818,
+      "step": 303
+    },
+    {
+      "epoch": 0.14152700186219738,
+      "grad_norm": 0.6783801141369618,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.5151,
+      "step": 304
+    },
+    {
+      "epoch": 0.1419925512104283,
+      "grad_norm": 0.6660318549296181,
+      "learning_rate": 4.721362229102168e-06,
+      "loss": 0.4814,
+      "step": 305
+    },
+    {
+      "epoch": 0.1424581005586592,
+      "grad_norm": 0.6731708505726399,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 0.483,
+      "step": 306
+    },
+    {
+      "epoch": 0.14292364990689013,
+      "grad_norm": 0.6714165940609655,
+      "learning_rate": 4.752321981424149e-06,
+      "loss": 0.4762,
+      "step": 307
+    },
+    {
+      "epoch": 0.14338919925512103,
+      "grad_norm": 0.6800104751183922,
+      "learning_rate": 4.7678018575851394e-06,
+      "loss": 0.4978,
+      "step": 308
+    },
+    {
+      "epoch": 0.14385474860335196,
+      "grad_norm": 0.6429815051731851,
+      "learning_rate": 4.78328173374613e-06,
+      "loss": 0.4898,
+      "step": 309
+    },
+    {
+      "epoch": 0.14432029795158285,
+      "grad_norm": 0.6912971701261036,
+      "learning_rate": 4.798761609907121e-06,
+      "loss": 0.4908,
+      "step": 310
+    },
+    {
+      "epoch": 0.14478584729981378,
+      "grad_norm": 0.6503474948384449,
+      "learning_rate": 4.814241486068112e-06,
+      "loss": 0.4827,
+      "step": 311
+    },
+    {
+      "epoch": 0.1452513966480447,
+      "grad_norm": 0.6912661589714176,
+      "learning_rate": 4.829721362229103e-06,
+      "loss": 0.5059,
+      "step": 312
+    },
+    {
+      "epoch": 0.1457169459962756,
+      "grad_norm": 0.6931173654076968,
+      "learning_rate": 4.845201238390093e-06,
+      "loss": 0.5035,
+      "step": 313
+    },
+    {
+      "epoch": 0.14618249534450653,
+      "grad_norm": 0.6712536416906297,
+      "learning_rate": 4.860681114551084e-06,
+      "loss": 0.4948,
+      "step": 314
+    },
+    {
+      "epoch": 0.14664804469273743,
+      "grad_norm": 0.6536825539489614,
+      "learning_rate": 4.876160990712075e-06,
+      "loss": 0.4957,
+      "step": 315
+    },
+    {
+      "epoch": 0.14711359404096835,
+      "grad_norm": 0.6622210482413142,
+      "learning_rate": 4.891640866873065e-06,
+      "loss": 0.5114,
+      "step": 316
+    },
+    {
+      "epoch": 0.14757914338919925,
+      "grad_norm": 0.6824219243775093,
+      "learning_rate": 4.907120743034056e-06,
+      "loss": 0.4936,
+      "step": 317
+    },
+    {
+      "epoch": 0.14804469273743018,
+      "grad_norm": 0.6394807786994813,
+      "learning_rate": 4.922600619195047e-06,
+      "loss": 0.4835,
+      "step": 318
+    },
+    {
+      "epoch": 0.14851024208566108,
+      "grad_norm": 0.6769583633580291,
+      "learning_rate": 4.938080495356037e-06,
+      "loss": 0.4832,
+      "step": 319
+    },
+    {
+      "epoch": 0.148975791433892,
+      "grad_norm": 0.659710608385043,
+      "learning_rate": 4.953560371517028e-06,
+      "loss": 0.4789,
+      "step": 320
+    },
+    {
+      "epoch": 0.1494413407821229,
+      "grad_norm": 0.6727337681139334,
+      "learning_rate": 4.9690402476780196e-06,
+      "loss": 0.5041,
+      "step": 321
+    },
+    {
+      "epoch": 0.14990689013035383,
+      "grad_norm": 0.6876864906895455,
+      "learning_rate": 4.9845201238390096e-06,
+      "loss": 0.4954,
+      "step": 322
+    },
+    {
+      "epoch": 0.15037243947858472,
+      "grad_norm": 0.7733655723185404,
+      "learning_rate": 5e-06,
+      "loss": 0.4854,
+      "step": 323
+    },
+    {
+      "epoch": 0.15083798882681565,
+      "grad_norm": 0.7054463481219058,
+      "learning_rate": 4.999999670720276e-06,
+      "loss": 0.5085,
+      "step": 324
+    },
+    {
+      "epoch": 0.15130353817504655,
+      "grad_norm": 0.6833159646275306,
+      "learning_rate": 4.999998682881188e-06,
+      "loss": 0.4731,
+      "step": 325
+    },
+    {
+      "epoch": 0.15176908752327747,
+      "grad_norm": 0.6120810444080651,
+      "learning_rate": 4.999997036483e-06,
+      "loss": 0.4842,
+      "step": 326
+    },
+    {
+      "epoch": 0.15223463687150837,
+      "grad_norm": 0.7014925350910873,
+      "learning_rate": 4.999994731526141e-06,
+      "loss": 0.4963,
+      "step": 327
+    },
+    {
+      "epoch": 0.1527001862197393,
+      "grad_norm": 0.6392372164610517,
+      "learning_rate": 4.999991768011222e-06,
+      "loss": 0.5055,
+      "step": 328
+    },
+    {
+      "epoch": 0.1531657355679702,
+      "grad_norm": 0.6438100687632863,
+      "learning_rate": 4.999988145939022e-06,
+      "loss": 0.5079,
+      "step": 329
+    },
+    {
+      "epoch": 0.15363128491620112,
+      "grad_norm": 0.7498908088615899,
+      "learning_rate": 4.999983865310496e-06,
+      "loss": 0.4996,
+      "step": 330
+    },
+    {
+      "epoch": 0.15409683426443202,
+      "grad_norm": 0.7195324117074359,
+      "learning_rate": 4.99997892612677e-06,
+      "loss": 0.5054,
+      "step": 331
+    },
+    {
+      "epoch": 0.15456238361266295,
+      "grad_norm": 0.6923353580221735,
+      "learning_rate": 4.999973328389147e-06,
+      "loss": 0.4903,
+      "step": 332
+    },
+    {
+      "epoch": 0.15502793296089384,
+      "grad_norm": 0.7088302805312129,
+      "learning_rate": 4.9999670720991e-06,
+      "loss": 0.515,
+      "step": 333
+    },
+    {
+      "epoch": 0.15549348230912477,
+      "grad_norm": 0.6724608869200307,
+      "learning_rate": 4.999960157258278e-06,
+      "loss": 0.4809,
+      "step": 334
+    },
+    {
+      "epoch": 0.15595903165735567,
+      "grad_norm": 0.6751595838266834,
+      "learning_rate": 4.999952583868503e-06,
+      "loss": 0.5169,
+      "step": 335
+    },
+    {
+      "epoch": 0.1564245810055866,
+      "grad_norm": 0.6403231325519236,
+      "learning_rate": 4.999944351931768e-06,
+      "loss": 0.4957,
+      "step": 336
+    },
+    {
+      "epoch": 0.1568901303538175,
+      "grad_norm": 0.6894255839639198,
+      "learning_rate": 4.999935461450244e-06,
+      "loss": 0.5005,
+      "step": 337
+    },
+    {
+      "epoch": 0.15735567970204842,
+      "grad_norm": 0.6558215241310459,
+      "learning_rate": 4.999925912426271e-06,
+      "loss": 0.5047,
+      "step": 338
+    },
+    {
+      "epoch": 0.15782122905027932,
+      "grad_norm": 0.6817162015929153,
+      "learning_rate": 4.9999157048623655e-06,
+      "loss": 0.5107,
+      "step": 339
+    },
+    {
+      "epoch": 0.15828677839851024,
+      "grad_norm": 0.6426216392912043,
+      "learning_rate": 4.999904838761216e-06,
+      "loss": 0.5182,
+      "step": 340
+    },
+    {
+      "epoch": 0.15875232774674117,
+      "grad_norm": 0.6688424548411167,
+      "learning_rate": 4.999893314125686e-06,
+      "loss": 0.4929,
+      "step": 341
+    },
+    {
+      "epoch": 0.15921787709497207,
+      "grad_norm": 0.6487311603112154,
+      "learning_rate": 4.999881130958808e-06,
+      "loss": 0.5098,
+      "step": 342
+    },
+    {
+      "epoch": 0.159683426443203,
+      "grad_norm": 0.7058126612897556,
+      "learning_rate": 4.999868289263796e-06,
+      "loss": 0.5118,
+      "step": 343
+    },
+    {
+      "epoch": 0.1601489757914339,
+      "grad_norm": 0.6994037652904751,
+      "learning_rate": 4.99985478904403e-06,
+      "loss": 0.5021,
+      "step": 344
+    },
+    {
+      "epoch": 0.16061452513966482,
+      "grad_norm": 0.7387477415665471,
+      "learning_rate": 4.999840630303066e-06,
+      "loss": 0.4868,
+      "step": 345
+    },
+    {
+      "epoch": 0.1610800744878957,
+      "grad_norm": 0.6763057273539088,
+      "learning_rate": 4.9998258130446356e-06,
+      "loss": 0.5224,
+      "step": 346
+    },
+    {
+      "epoch": 0.16154562383612664,
+      "grad_norm": 0.6703028186376043,
+      "learning_rate": 4.99981033727264e-06,
+      "loss": 0.4937,
+      "step": 347
+    },
+    {
+      "epoch": 0.16201117318435754,
+      "grad_norm": 0.7131330889276251,
+      "learning_rate": 4.999794202991157e-06,
+      "loss": 0.4917,
+      "step": 348
+    },
+    {
+      "epoch": 0.16247672253258846,
+      "grad_norm": 0.6846540176481088,
+      "learning_rate": 4.999777410204437e-06,
+      "loss": 0.5127,
+      "step": 349
+    },
+    {
+      "epoch": 0.16294227188081936,
+      "grad_norm": 0.6951639070171981,
+      "learning_rate": 4.999759958916904e-06,
+      "loss": 0.4999,
+      "step": 350
+    },
+    {
+      "epoch": 0.1634078212290503,
+      "grad_norm": 0.6888163329311681,
+      "learning_rate": 4.999741849133153e-06,
+      "loss": 0.5025,
+      "step": 351
+    },
+    {
+      "epoch": 0.16387337057728119,
+      "grad_norm": 0.675768213550412,
+      "learning_rate": 4.999723080857956e-06,
+      "loss": 0.5066,
+      "step": 352
+    },
+    {
+      "epoch": 0.1643389199255121,
+      "grad_norm": 0.6449509976119283,
+      "learning_rate": 4.999703654096257e-06,
+      "loss": 0.4959,
+      "step": 353
+    },
+    {
+      "epoch": 0.164804469273743,
+      "grad_norm": 0.6847489660575293,
+      "learning_rate": 4.999683568853174e-06,
+      "loss": 0.5013,
+      "step": 354
+    },
+    {
+      "epoch": 0.16527001862197394,
+      "grad_norm": 0.635254271862752,
+      "learning_rate": 4.999662825133996e-06,
+      "loss": 0.4791,
+      "step": 355
+    },
+    {
+      "epoch": 0.16573556797020483,
+      "grad_norm": 0.7261586765022625,
+      "learning_rate": 4.99964142294419e-06,
+      "loss": 0.5055,
+      "step": 356
+    },
+    {
+      "epoch": 0.16620111731843576,
+      "grad_norm": 0.6449298112045118,
+      "learning_rate": 4.9996193622893914e-06,
+      "loss": 0.4887,
+      "step": 357
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.6504320148514781,
+      "learning_rate": 4.999596643175414e-06,
+      "loss": 0.5077,
+      "step": 358
+    },
+    {
+      "epoch": 0.16713221601489758,
+      "grad_norm": 0.6870747384179126,
+      "learning_rate": 4.999573265608239e-06,
+      "loss": 0.4749,
+      "step": 359
+    },
+    {
+      "epoch": 0.16759776536312848,
+      "grad_norm": 0.6290730990044666,
+      "learning_rate": 4.999549229594029e-06,
+      "loss": 0.4924,
+      "step": 360
+    },
+    {
+      "epoch": 0.1680633147113594,
+      "grad_norm": 0.6490831449756712,
+      "learning_rate": 4.999524535139112e-06,
+      "loss": 0.4933,
+      "step": 361
+    },
+    {
+      "epoch": 0.1685288640595903,
+      "grad_norm": 0.61268263086085,
+      "learning_rate": 4.999499182249996e-06,
+      "loss": 0.4824,
+      "step": 362
+    },
+    {
+      "epoch": 0.16899441340782123,
+      "grad_norm": 0.6197948707048391,
+      "learning_rate": 4.999473170933357e-06,
+      "loss": 0.4827,
+      "step": 363
+    },
+    {
+      "epoch": 0.16945996275605213,
+      "grad_norm": 0.6738944292031096,
+      "learning_rate": 4.999446501196048e-06,
+      "loss": 0.4679,
+      "step": 364
+    },
+    {
+      "epoch": 0.16992551210428306,
+      "grad_norm": 0.7515436412991718,
+      "learning_rate": 4.999419173045095e-06,
+      "loss": 0.4927,
+      "step": 365
+    },
+    {
+      "epoch": 0.17039106145251395,
+      "grad_norm": 0.6306099073198599,
+      "learning_rate": 4.999391186487695e-06,
+      "loss": 0.4849,
+      "step": 366
+    },
+    {
+      "epoch": 0.17085661080074488,
+      "grad_norm": 0.6465596087101093,
+      "learning_rate": 4.999362541531224e-06,
+      "loss": 0.4839,
+      "step": 367
+    },
+    {
+      "epoch": 0.1713221601489758,
+      "grad_norm": 0.655520105851241,
+      "learning_rate": 4.999333238183224e-06,
+      "loss": 0.4988,
+      "step": 368
+    },
+    {
+      "epoch": 0.1717877094972067,
+      "grad_norm": 0.6588921059043852,
+      "learning_rate": 4.999303276451416e-06,
+      "loss": 0.5007,
+      "step": 369
+    },
+    {
+      "epoch": 0.17225325884543763,
+      "grad_norm": 0.6340996939724594,
+      "learning_rate": 4.9992726563436926e-06,
+      "loss": 0.4945,
+      "step": 370
+    },
+    {
+      "epoch": 0.17271880819366853,
+      "grad_norm": 0.6833307114320069,
+      "learning_rate": 4.999241377868119e-06,
+      "loss": 0.523,
+      "step": 371
+    },
+    {
+      "epoch": 0.17318435754189945,
+      "grad_norm": 0.6749689815942786,
+      "learning_rate": 4.999209441032936e-06,
+      "loss": 0.4912,
+      "step": 372
+    },
+    {
+      "epoch": 0.17364990689013035,
+      "grad_norm": 0.6951711720217982,
+      "learning_rate": 4.999176845846555e-06,
+      "loss": 0.494,
+      "step": 373
+    },
+    {
+      "epoch": 0.17411545623836128,
+      "grad_norm": 0.6619038154204162,
+      "learning_rate": 4.999143592317564e-06,
+      "loss": 0.4781,
+      "step": 374
+    },
+    {
+      "epoch": 0.17458100558659218,
+      "grad_norm": 0.6732735714684929,
+      "learning_rate": 4.999109680454721e-06,
+      "loss": 0.4907,
+      "step": 375
+    },
+    {
+      "epoch": 0.1750465549348231,
+      "grad_norm": 0.6831830901130143,
+      "learning_rate": 4.999075110266961e-06,
+      "loss": 0.5031,
+      "step": 376
+    },
+    {
+      "epoch": 0.175512104283054,
+      "grad_norm": 0.6304792790353789,
+      "learning_rate": 4.99903988176339e-06,
+      "loss": 0.4935,
+      "step": 377
+    },
+    {
+      "epoch": 0.17597765363128492,
+      "grad_norm": 0.7449223269280735,
+      "learning_rate": 4.999003994953287e-06,
+      "loss": 0.4941,
+      "step": 378
+    },
+    {
+      "epoch": 0.17644320297951582,
+      "grad_norm": 0.667115267021414,
+      "learning_rate": 4.998967449846106e-06,
+      "loss": 0.4727,
+      "step": 379
+    },
+    {
+      "epoch": 0.17690875232774675,
+      "grad_norm": 0.6318085527954493,
+      "learning_rate": 4.998930246451474e-06,
+      "loss": 0.5071,
+      "step": 380
+    },
+    {
+      "epoch": 0.17737430167597765,
+      "grad_norm": 0.6937558592654792,
+      "learning_rate": 4.998892384779192e-06,
+      "loss": 0.5029,
+      "step": 381
+    },
+    {
+      "epoch": 0.17783985102420857,
+      "grad_norm": 0.6942891419410026,
+      "learning_rate": 4.9988538648392335e-06,
+      "loss": 0.4796,
+      "step": 382
+    },
+    {
+      "epoch": 0.17830540037243947,
+      "grad_norm": 0.6484866937943841,
+      "learning_rate": 4.998814686641744e-06,
+      "loss": 0.4841,
+      "step": 383
+    },
+    {
+      "epoch": 0.1787709497206704,
+      "grad_norm": 0.7017284502701508,
+      "learning_rate": 4.998774850197045e-06,
+      "loss": 0.4969,
+      "step": 384
+    },
+    {
+      "epoch": 0.1792364990689013,
+      "grad_norm": 0.7045096673993267,
+      "learning_rate": 4.998734355515631e-06,
+      "loss": 0.4752,
+      "step": 385
+    },
+    {
+      "epoch": 0.17970204841713222,
+      "grad_norm": 0.6689137153968578,
+      "learning_rate": 4.998693202608168e-06,
+      "loss": 0.4958,
+      "step": 386
+    },
+    {
+      "epoch": 0.18016759776536312,
+      "grad_norm": 0.6857754884701179,
+      "learning_rate": 4.998651391485497e-06,
+      "loss": 0.4942,
+      "step": 387
+    },
+    {
+      "epoch": 0.18063314711359404,
+      "grad_norm": 0.6819049210632732,
+      "learning_rate": 4.998608922158633e-06,
+      "loss": 0.4779,
+      "step": 388
+    },
+    {
+      "epoch": 0.18109869646182494,
+      "grad_norm": 0.659705015951754,
+      "learning_rate": 4.998565794638763e-06,
+      "loss": 0.5036,
+      "step": 389
+    },
+    {
+      "epoch": 0.18156424581005587,
+      "grad_norm": 0.6201528634837709,
+      "learning_rate": 4.998522008937247e-06,
+      "loss": 0.4952,
+      "step": 390
+    },
+    {
+      "epoch": 0.18202979515828677,
+      "grad_norm": 0.6832495037697356,
+      "learning_rate": 4.99847756506562e-06,
+      "loss": 0.5164,
+      "step": 391
+    },
+    {
+      "epoch": 0.1824953445065177,
+      "grad_norm": 0.6609310580321621,
+      "learning_rate": 4.99843246303559e-06,
+      "loss": 0.4896,
+      "step": 392
+    },
+    {
+      "epoch": 0.1829608938547486,
+      "grad_norm": 0.6209165607141458,
+      "learning_rate": 4.998386702859036e-06,
+      "loss": 0.4953,
+      "step": 393
+    },
+    {
+      "epoch": 0.18342644320297952,
+      "grad_norm": 0.7214250945591565,
+      "learning_rate": 4.998340284548014e-06,
+      "loss": 0.5081,
+      "step": 394
+    },
+    {
+      "epoch": 0.18389199255121041,
+      "grad_norm": 0.6839061659745944,
+      "learning_rate": 4.998293208114752e-06,
+      "loss": 0.4915,
+      "step": 395
+    },
+    {
+      "epoch": 0.18435754189944134,
+      "grad_norm": 0.6737940403394574,
+      "learning_rate": 4.99824547357165e-06,
+      "loss": 0.4914,
+      "step": 396
+    },
+    {
+      "epoch": 0.18482309124767227,
+      "grad_norm": 0.6467505522679918,
+      "learning_rate": 4.998197080931282e-06,
+      "loss": 0.5016,
+      "step": 397
+    },
+    {
+      "epoch": 0.18528864059590316,
+      "grad_norm": 0.7130574959793906,
+      "learning_rate": 4.998148030206397e-06,
+      "loss": 0.4945,
+      "step": 398
+    },
+    {
+      "epoch": 0.1857541899441341,
+      "grad_norm": 0.7084613884637921,
+      "learning_rate": 4.998098321409916e-06,
+      "loss": 0.5052,
+      "step": 399
+    },
+    {
+      "epoch": 0.186219739292365,
+      "grad_norm": 0.659999578590095,
+      "learning_rate": 4.998047954554933e-06,
+      "loss": 0.488,
+      "step": 400
+    },
+    {
+      "epoch": 0.18668528864059591,
+      "grad_norm": 0.7191894751523055,
+      "learning_rate": 4.997996929654716e-06,
+      "loss": 0.4848,
+      "step": 401
+    },
+    {
+      "epoch": 0.1871508379888268,
+      "grad_norm": 0.6865959693166541,
+      "learning_rate": 4.997945246722706e-06,
+      "loss": 0.5057,
+      "step": 402
+    },
+    {
+      "epoch": 0.18761638733705774,
+      "grad_norm": 0.7156556594238739,
+      "learning_rate": 4.997892905772518e-06,
+      "loss": 0.4785,
+      "step": 403
+    },
+    {
+      "epoch": 0.18808193668528864,
+      "grad_norm": 0.6499452018277444,
+      "learning_rate": 4.9978399068179394e-06,
+      "loss": 0.4817,
+      "step": 404
+    },
+    {
+      "epoch": 0.18854748603351956,
+      "grad_norm": 0.6664182203985279,
+      "learning_rate": 4.997786249872931e-06,
+      "loss": 0.4739,
+      "step": 405
+    },
+    {
+      "epoch": 0.18901303538175046,
+      "grad_norm": 0.6603174793886444,
+      "learning_rate": 4.997731934951628e-06,
+      "loss": 0.4625,
+      "step": 406
+    },
+    {
+      "epoch": 0.1894785847299814,
+      "grad_norm": 0.6671052839404236,
+      "learning_rate": 4.997676962068338e-06,
+      "loss": 0.4871,
+      "step": 407
+    },
+    {
+      "epoch": 0.18994413407821228,
+      "grad_norm": 0.7317189789441665,
+      "learning_rate": 4.997621331237543e-06,
+      "loss": 0.466,
+      "step": 408
+    },
+    {
+      "epoch": 0.1904096834264432,
+      "grad_norm": 0.6753202014890578,
+      "learning_rate": 4.997565042473896e-06,
+      "loss": 0.5063,
+      "step": 409
+    },
+    {
+      "epoch": 0.1908752327746741,
+      "grad_norm": 0.7090463463390854,
+      "learning_rate": 4.997508095792226e-06,
+      "loss": 0.5007,
+      "step": 410
+    },
+    {
+      "epoch": 0.19134078212290503,
+      "grad_norm": 0.6893555776942853,
+      "learning_rate": 4.9974504912075325e-06,
+      "loss": 0.4837,
+      "step": 411
+    },
+    {
+      "epoch": 0.19180633147113593,
+      "grad_norm": 0.6374064924537906,
+      "learning_rate": 4.997392228734992e-06,
+      "loss": 0.4914,
+      "step": 412
+    },
+    {
+      "epoch": 0.19227188081936686,
+      "grad_norm": 0.6648025296895367,
+      "learning_rate": 4.997333308389951e-06,
+      "loss": 0.4953,
+      "step": 413
+    },
+    {
+      "epoch": 0.19273743016759776,
+      "grad_norm": 0.6985436621939908,
+      "learning_rate": 4.99727373018793e-06,
+      "loss": 0.4985,
+      "step": 414
+    },
+    {
+      "epoch": 0.19320297951582868,
+      "grad_norm": 0.6886379718794371,
+      "learning_rate": 4.997213494144625e-06,
+      "loss": 0.5202,
+      "step": 415
+    },
+    {
+      "epoch": 0.19366852886405958,
+      "grad_norm": 0.6172911984024713,
+      "learning_rate": 4.997152600275903e-06,
+      "loss": 0.4905,
+      "step": 416
+    },
+    {
+      "epoch": 0.1941340782122905,
+      "grad_norm": 0.6122242527848365,
+      "learning_rate": 4.997091048597804e-06,
+      "loss": 0.5134,
+      "step": 417
+    },
+    {
+      "epoch": 0.1945996275605214,
+      "grad_norm": 0.6810274411949823,
+      "learning_rate": 4.997028839126543e-06,
+      "loss": 0.5044,
+      "step": 418
+    },
+    {
+      "epoch": 0.19506517690875233,
+      "grad_norm": 0.6259240415588101,
+      "learning_rate": 4.996965971878507e-06,
+      "loss": 0.4796,
+      "step": 419
+    },
+    {
+      "epoch": 0.19553072625698323,
+      "grad_norm": 0.6348191328243723,
+      "learning_rate": 4.996902446870256e-06,
+      "loss": 0.4985,
+      "step": 420
+    },
+    {
+      "epoch": 0.19599627560521415,
+      "grad_norm": 0.6050367868790869,
+      "learning_rate": 4.996838264118526e-06,
+      "loss": 0.4955,
+      "step": 421
+    },
+    {
+      "epoch": 0.19646182495344505,
+      "grad_norm": 0.6541561685139006,
+      "learning_rate": 4.996773423640224e-06,
+      "loss": 0.492,
+      "step": 422
+    },
+    {
+      "epoch": 0.19692737430167598,
+      "grad_norm": 0.6515627327435464,
+      "learning_rate": 4.996707925452429e-06,
+      "loss": 0.5103,
+      "step": 423
+    },
+    {
+      "epoch": 0.1973929236499069,
+      "grad_norm": 0.679533968446411,
+      "learning_rate": 4.996641769572394e-06,
+      "loss": 0.4975,
+      "step": 424
+    },
+    {
+      "epoch": 0.1978584729981378,
+      "grad_norm": 0.6692905152624291,
+      "learning_rate": 4.99657495601755e-06,
+      "loss": 0.4854,
+      "step": 425
+    },
+    {
+      "epoch": 0.19832402234636873,
+      "grad_norm": 0.6732491814981746,
+      "learning_rate": 4.996507484805494e-06,
+      "loss": 0.4942,
+      "step": 426
+    },
+    {
+      "epoch": 0.19878957169459963,
+      "grad_norm": 0.7062025442566869,
+      "learning_rate": 4.9964393559539995e-06,
+      "loss": 0.5166,
+      "step": 427
+    },
+    {
+      "epoch": 0.19925512104283055,
+      "grad_norm": 0.6240669377316027,
+      "learning_rate": 4.996370569481014e-06,
+      "loss": 0.4669,
+      "step": 428
+    },
+    {
+      "epoch": 0.19972067039106145,
+      "grad_norm": 0.6658998893732727,
+      "learning_rate": 4.996301125404659e-06,
+      "loss": 0.4831,
+      "step": 429
+    },
+    {
+      "epoch": 0.20018621973929238,
+      "grad_norm": 0.7168653964220361,
+      "learning_rate": 4.996231023743226e-06,
+      "loss": 0.5067,
+      "step": 430
+    },
+    {
+      "epoch": 0.20065176908752327,
+      "grad_norm": 0.7185681322404474,
+      "learning_rate": 4.99616026451518e-06,
+      "loss": 0.4919,
+      "step": 431
+    },
+    {
+      "epoch": 0.2011173184357542,
+      "grad_norm": 0.6653392532306012,
+      "learning_rate": 4.9960888477391635e-06,
+      "loss": 0.4732,
+      "step": 432
+    },
+    {
+      "epoch": 0.2015828677839851,
+      "grad_norm": 0.7325743293042596,
+      "learning_rate": 4.996016773433989e-06,
+      "loss": 0.4807,
+      "step": 433
+    },
+    {
+      "epoch": 0.20204841713221602,
+      "grad_norm": 0.7524259886340442,
+      "learning_rate": 4.995944041618641e-06,
+      "loss": 0.4934,
+      "step": 434
+    },
+    {
+      "epoch": 0.20251396648044692,
+      "grad_norm": 0.6860661393180427,
+      "learning_rate": 4.9958706523122805e-06,
+      "loss": 0.4965,
+      "step": 435
+    },
+    {
+      "epoch": 0.20297951582867785,
+      "grad_norm": 0.6875792725289399,
+      "learning_rate": 4.995796605534238e-06,
+      "loss": 0.4842,
+      "step": 436
+    },
+    {
+      "epoch": 0.20344506517690875,
+      "grad_norm": 0.6426304344190528,
+      "learning_rate": 4.995721901304021e-06,
+      "loss": 0.4872,
+      "step": 437
+    },
+    {
+      "epoch": 0.20391061452513967,
+      "grad_norm": 0.6482117932614719,
+      "learning_rate": 4.995646539641308e-06,
+      "loss": 0.4928,
+      "step": 438
+    },
+    {
+      "epoch": 0.20437616387337057,
+      "grad_norm": 0.667394575626564,
+      "learning_rate": 4.99557052056595e-06,
+      "loss": 0.4789,
+      "step": 439
+    },
+    {
+      "epoch": 0.2048417132216015,
+      "grad_norm": 0.6794615191076304,
+      "learning_rate": 4.995493844097974e-06,
+      "loss": 0.5016,
+      "step": 440
+    },
+    {
+      "epoch": 0.2053072625698324,
+      "grad_norm": 0.6753680331828084,
+      "learning_rate": 4.995416510257577e-06,
+      "loss": 0.4926,
+      "step": 441
+    },
+    {
+      "epoch": 0.20577281191806332,
+      "grad_norm": 0.6436924138374562,
+      "learning_rate": 4.995338519065131e-06,
+      "loss": 0.5012,
+      "step": 442
+    },
+    {
+      "epoch": 0.20623836126629422,
+      "grad_norm": 0.6522443921629036,
+      "learning_rate": 4.99525987054118e-06,
+      "loss": 0.4904,
+      "step": 443
+    },
+    {
+      "epoch": 0.20670391061452514,
+      "grad_norm": 0.652564359464448,
+      "learning_rate": 4.9951805647064435e-06,
+      "loss": 0.486,
+      "step": 444
+    },
+    {
+      "epoch": 0.20716945996275604,
+      "grad_norm": 0.6784283661754015,
+      "learning_rate": 4.995100601581811e-06,
+      "loss": 0.4918,
+      "step": 445
+    },
+    {
+      "epoch": 0.20763500931098697,
+      "grad_norm": 0.6383250899241992,
+      "learning_rate": 4.995019981188347e-06,
+      "loss": 0.4796,
+      "step": 446
+    },
+    {
+      "epoch": 0.20810055865921787,
+      "grad_norm": 0.6583277571414673,
+      "learning_rate": 4.9949387035472906e-06,
+      "loss": 0.4636,
+      "step": 447
+    },
+    {
+      "epoch": 0.2085661080074488,
+      "grad_norm": 0.7327305977163997,
+      "learning_rate": 4.99485676868005e-06,
+      "loss": 0.5006,
+      "step": 448
+    },
+    {
+      "epoch": 0.2090316573556797,
+      "grad_norm": 0.6304610794178843,
+      "learning_rate": 4.994774176608209e-06,
+      "loss": 0.472,
+      "step": 449
+    },
+    {
+      "epoch": 0.20949720670391062,
+      "grad_norm": 0.653431385965446,
+      "learning_rate": 4.994690927353526e-06,
+      "loss": 0.5038,
+      "step": 450
+    },
+    {
+      "epoch": 0.2099627560521415,
+      "grad_norm": 0.6897855711118853,
+      "learning_rate": 4.994607020937929e-06,
+      "loss": 0.4939,
+      "step": 451
+    },
+    {
+      "epoch": 0.21042830540037244,
+      "grad_norm": 0.6166162707022773,
+      "learning_rate": 4.994522457383522e-06,
+      "loss": 0.5038,
+      "step": 452
+    },
+    {
+      "epoch": 0.21089385474860337,
+      "grad_norm": 0.6315427234102929,
+      "learning_rate": 4.994437236712581e-06,
+      "loss": 0.5036,
+      "step": 453
+    },
+    {
+      "epoch": 0.21135940409683426,
+      "grad_norm": 0.6347797068447566,
+      "learning_rate": 4.9943513589475545e-06,
+      "loss": 0.4857,
+      "step": 454
+    },
+    {
+      "epoch": 0.2118249534450652,
+      "grad_norm": 0.6778531751315957,
+      "learning_rate": 4.994264824111065e-06,
+      "loss": 0.4635,
+      "step": 455
+    },
+    {
+      "epoch": 0.2122905027932961,
+      "grad_norm": 0.6692283764991555,
+      "learning_rate": 4.9941776322259085e-06,
+      "loss": 0.4685,
+      "step": 456
+    },
+    {
+      "epoch": 0.212756052141527,
+      "grad_norm": 0.6415553314154355,
+      "learning_rate": 4.994089783315052e-06,
+      "loss": 0.5004,
+      "step": 457
+    },
+    {
+      "epoch": 0.2132216014897579,
+      "grad_norm": 0.6594948451358219,
+      "learning_rate": 4.994001277401639e-06,
+      "loss": 0.477,
+      "step": 458
+    },
+    {
+      "epoch": 0.21368715083798884,
+      "grad_norm": 0.6795163188764092,
+      "learning_rate": 4.993912114508982e-06,
+      "loss": 0.5169,
+      "step": 459
+    },
+    {
+      "epoch": 0.21415270018621974,
+      "grad_norm": 0.6736210843021158,
+      "learning_rate": 4.993822294660571e-06,
+      "loss": 0.5062,
+      "step": 460
+    },
+    {
+      "epoch": 0.21461824953445066,
+      "grad_norm": 0.6615279691648641,
+      "learning_rate": 4.9937318178800645e-06,
+      "loss": 0.4999,
+      "step": 461
+    },
+    {
+      "epoch": 0.21508379888268156,
+      "grad_norm": 0.6152779749627438,
+      "learning_rate": 4.993640684191296e-06,
+      "loss": 0.4969,
+      "step": 462
+    },
+    {
+      "epoch": 0.21554934823091249,
+      "grad_norm": 0.6539578787279585,
+      "learning_rate": 4.993548893618275e-06,
+      "loss": 0.4998,
+      "step": 463
+    },
+    {
+      "epoch": 0.21601489757914338,
+      "grad_norm": 0.6699286795450498,
+      "learning_rate": 4.993456446185179e-06,
+      "loss": 0.4793,
+      "step": 464
+    },
+    {
+      "epoch": 0.2164804469273743,
+      "grad_norm": 0.6275688865189731,
+      "learning_rate": 4.993363341916361e-06,
+      "loss": 0.4971,
+      "step": 465
+    },
+    {
+      "epoch": 0.2169459962756052,
+      "grad_norm": 0.6715806260412706,
+      "learning_rate": 4.993269580836347e-06,
+      "loss": 0.4794,
+      "step": 466
+    },
+    {
+      "epoch": 0.21741154562383613,
+      "grad_norm": 0.6606501409784871,
+      "learning_rate": 4.993175162969837e-06,
+      "loss": 0.4949,
+      "step": 467
+    },
+    {
+      "epoch": 0.21787709497206703,
+      "grad_norm": 0.6451469192225622,
+      "learning_rate": 4.993080088341702e-06,
+      "loss": 0.4926,
+      "step": 468
+    },
+    {
+      "epoch": 0.21834264432029796,
+      "grad_norm": 0.6159885378404277,
+      "learning_rate": 4.992984356976987e-06,
+      "loss": 0.4864,
+      "step": 469
+    },
+    {
+      "epoch": 0.21880819366852886,
+      "grad_norm": 0.6418577732370576,
+      "learning_rate": 4.99288796890091e-06,
+      "loss": 0.4716,
+      "step": 470
+    },
+    {
+      "epoch": 0.21927374301675978,
+      "grad_norm": 0.637703342293294,
+      "learning_rate": 4.992790924138862e-06,
+      "loss": 0.4739,
+      "step": 471
+    },
+    {
+      "epoch": 0.21973929236499068,
+      "grad_norm": 0.7791746332164188,
+      "learning_rate": 4.992693222716407e-06,
+      "loss": 0.4971,
+      "step": 472
+    },
+    {
+      "epoch": 0.2202048417132216,
+      "grad_norm": 0.6047876723673562,
+      "learning_rate": 4.992594864659282e-06,
+      "loss": 0.4758,
+      "step": 473
+    },
+    {
+      "epoch": 0.2206703910614525,
+      "grad_norm": 0.6387426662181985,
+      "learning_rate": 4.992495849993395e-06,
+      "loss": 0.5052,
+      "step": 474
+    },
+    {
+      "epoch": 0.22113594040968343,
+      "grad_norm": 0.6551420721576494,
+      "learning_rate": 4.992396178744832e-06,
+      "loss": 0.4972,
+      "step": 475
+    },
+    {
+      "epoch": 0.22160148975791433,
+      "grad_norm": 0.6354400963796127,
+      "learning_rate": 4.992295850939845e-06,
+      "loss": 0.4777,
+      "step": 476
+    },
+    {
+      "epoch": 0.22206703910614525,
+      "grad_norm": 0.6505352085757535,
+      "learning_rate": 4.992194866604867e-06,
+      "loss": 0.4626,
+      "step": 477
+    },
+    {
+      "epoch": 0.22253258845437615,
+      "grad_norm": 0.6215891863930396,
+      "learning_rate": 4.9920932257664965e-06,
+      "loss": 0.4739,
+      "step": 478
+    },
+    {
+      "epoch": 0.22299813780260708,
+      "grad_norm": 0.637431603289288,
+      "learning_rate": 4.99199092845151e-06,
+      "loss": 0.4953,
+      "step": 479
+    },
+    {
+      "epoch": 0.22346368715083798,
+      "grad_norm": 0.6676939594429065,
+      "learning_rate": 4.991887974686853e-06,
+      "loss": 0.511,
+      "step": 480
+    },
+    {
+      "epoch": 0.2239292364990689,
+      "grad_norm": 0.6521730199345667,
+      "learning_rate": 4.991784364499648e-06,
+      "loss": 0.49,
+      "step": 481
+    },
+    {
+      "epoch": 0.22439478584729983,
+      "grad_norm": 0.6302531478489631,
+      "learning_rate": 4.991680097917188e-06,
+      "loss": 0.4766,
+      "step": 482
+    },
+    {
+      "epoch": 0.22486033519553073,
+      "grad_norm": 0.6367027591042698,
+      "learning_rate": 4.991575174966937e-06,
+      "loss": 0.4984,
+      "step": 483
+    },
+    {
+      "epoch": 0.22532588454376165,
+      "grad_norm": 0.6276569868409256,
+      "learning_rate": 4.991469595676537e-06,
+      "loss": 0.4782,
+      "step": 484
+    },
+    {
+      "epoch": 0.22579143389199255,
+      "grad_norm": 0.6509407526165827,
+      "learning_rate": 4.9913633600738e-06,
+      "loss": 0.471,
+      "step": 485
+    },
+    {
+      "epoch": 0.22625698324022347,
+      "grad_norm": 0.6555654047138948,
+      "learning_rate": 4.991256468186709e-06,
+      "loss": 0.4823,
+      "step": 486
+    },
+    {
+      "epoch": 0.22672253258845437,
+      "grad_norm": 0.6440061168377553,
+      "learning_rate": 4.991148920043423e-06,
+      "loss": 0.4695,
+      "step": 487
+    },
+    {
+      "epoch": 0.2271880819366853,
+      "grad_norm": 0.6323558817884143,
+      "learning_rate": 4.991040715672273e-06,
+      "loss": 0.4703,
+      "step": 488
+    },
+    {
+      "epoch": 0.2276536312849162,
+      "grad_norm": 0.6536912045812083,
+      "learning_rate": 4.9909318551017625e-06,
+      "loss": 0.456,
+      "step": 489
+    },
+    {
+      "epoch": 0.22811918063314712,
+      "grad_norm": 0.681254024988898,
+      "learning_rate": 4.9908223383605684e-06,
+      "loss": 0.4708,
+      "step": 490
+    },
+    {
+      "epoch": 0.22858472998137802,
+      "grad_norm": 0.6359100793630226,
+      "learning_rate": 4.990712165477538e-06,
+      "loss": 0.4904,
+      "step": 491
+    },
+    {
+      "epoch": 0.22905027932960895,
+      "grad_norm": 0.6361206725377583,
+      "learning_rate": 4.990601336481695e-06,
+      "loss": 0.5157,
+      "step": 492
+    },
+    {
+      "epoch": 0.22951582867783984,
+      "grad_norm": 0.7095087343996155,
+      "learning_rate": 4.990489851402235e-06,
+      "loss": 0.4751,
+      "step": 493
+    },
+    {
+      "epoch": 0.22998137802607077,
+      "grad_norm": 0.6166836653640029,
+      "learning_rate": 4.990377710268525e-06,
+      "loss": 0.4619,
+      "step": 494
+    },
+    {
+      "epoch": 0.23044692737430167,
+      "grad_norm": 0.658070535150144,
+      "learning_rate": 4.990264913110105e-06,
+      "loss": 0.481,
+      "step": 495
+    },
+    {
+      "epoch": 0.2309124767225326,
+      "grad_norm": 0.6631235323945147,
+      "learning_rate": 4.99015145995669e-06,
+      "loss": 0.5048,
+      "step": 496
+    },
+    {
+      "epoch": 0.2313780260707635,
+      "grad_norm": 0.6709499359990856,
+      "learning_rate": 4.990037350838165e-06,
+      "loss": 0.4936,
+      "step": 497
+    },
+    {
+      "epoch": 0.23184357541899442,
+      "grad_norm": 0.6375115747023777,
+      "learning_rate": 4.98992258578459e-06,
+      "loss": 0.4753,
+      "step": 498
+    },
+    {
+      "epoch": 0.23230912476722532,
+      "grad_norm": 0.6785801503131028,
+      "learning_rate": 4.989807164826196e-06,
+      "loss": 0.4709,
+      "step": 499
+    },
+    {
+      "epoch": 0.23277467411545624,
+      "grad_norm": 0.7176623685104356,
+      "learning_rate": 4.989691087993388e-06,
+      "loss": 0.4954,
+      "step": 500
+    },
+    {
+      "epoch": 0.23324022346368714,
+      "grad_norm": 0.6813347425299445,
+      "learning_rate": 4.989574355316743e-06,
+      "loss": 0.4818,
+      "step": 501
+    },
+    {
+      "epoch": 0.23370577281191807,
+      "grad_norm": 0.6766094035644477,
+      "learning_rate": 4.989456966827012e-06,
+      "loss": 0.5094,
+      "step": 502
+    },
+    {
+      "epoch": 0.23417132216014896,
+      "grad_norm": 0.669049920443907,
+      "learning_rate": 4.989338922555117e-06,
+      "loss": 0.4839,
+      "step": 503
+    },
+    {
+      "epoch": 0.2346368715083799,
+      "grad_norm": 0.665206528445216,
+      "learning_rate": 4.989220222532154e-06,
+      "loss": 0.4743,
+      "step": 504
+    },
+    {
+      "epoch": 0.2351024208566108,
+      "grad_norm": 0.6615828363603158,
+      "learning_rate": 4.989100866789391e-06,
+      "loss": 0.4641,
+      "step": 505
+    },
+    {
+      "epoch": 0.23556797020484171,
+      "grad_norm": 0.6663630775551603,
+      "learning_rate": 4.988980855358271e-06,
+      "loss": 0.4863,
+      "step": 506
+    },
+    {
+      "epoch": 0.2360335195530726,
+      "grad_norm": 0.69650119310072,
+      "learning_rate": 4.988860188270406e-06,
+      "loss": 0.4924,
+      "step": 507
+    },
+    {
+      "epoch": 0.23649906890130354,
+      "grad_norm": 0.6353741420704654,
+      "learning_rate": 4.988738865557583e-06,
+      "loss": 0.4911,
+      "step": 508
+    },
+    {
+      "epoch": 0.23696461824953446,
+      "grad_norm": 0.6523302108329064,
+      "learning_rate": 4.988616887251762e-06,
+      "loss": 0.4552,
+      "step": 509
+    },
+    {
+      "epoch": 0.23743016759776536,
+      "grad_norm": 0.7645894490571482,
+      "learning_rate": 4.988494253385073e-06,
+      "loss": 0.4889,
+      "step": 510
+    },
+    {
+      "epoch": 0.2378957169459963,
+      "grad_norm": 0.6297523524848463,
+      "learning_rate": 4.9883709639898226e-06,
+      "loss": 0.4928,
+      "step": 511
+    },
+    {
+      "epoch": 0.2383612662942272,
+      "grad_norm": 0.791137602865106,
+      "learning_rate": 4.988247019098487e-06,
+      "loss": 0.4669,
+      "step": 512
+    },
+    {
+      "epoch": 0.2388268156424581,
+      "grad_norm": 0.5981568003039297,
+      "learning_rate": 4.988122418743718e-06,
+      "loss": 0.4793,
+      "step": 513
+    },
+    {
+      "epoch": 0.239292364990689,
+      "grad_norm": 0.661918623699415,
+      "learning_rate": 4.987997162958337e-06,
+      "loss": 0.5045,
+      "step": 514
+    },
+    {
+      "epoch": 0.23975791433891994,
+      "grad_norm": 0.6789403973722047,
+      "learning_rate": 4.987871251775339e-06,
+      "loss": 0.4741,
+      "step": 515
+    },
+    {
+      "epoch": 0.24022346368715083,
+      "grad_norm": 0.6530708813606093,
+      "learning_rate": 4.987744685227893e-06,
+      "loss": 0.4825,
+      "step": 516
+    },
+    {
+      "epoch": 0.24068901303538176,
+      "grad_norm": 0.666927549654768,
+      "learning_rate": 4.987617463349339e-06,
+      "loss": 0.4865,
+      "step": 517
+    },
+    {
+      "epoch": 0.24115456238361266,
+      "grad_norm": 0.6591159664268077,
+      "learning_rate": 4.98748958617319e-06,
+      "loss": 0.4998,
+      "step": 518
+    },
+    {
+      "epoch": 0.24162011173184358,
+      "grad_norm": 0.6450689726108075,
+      "learning_rate": 4.987361053733133e-06,
+      "loss": 0.4946,
+      "step": 519
+    },
+    {
+      "epoch": 0.24208566108007448,
+      "grad_norm": 0.6352737300147594,
+      "learning_rate": 4.987231866063025e-06,
+      "loss": 0.4872,
+      "step": 520
+    },
+    {
+      "epoch": 0.2425512104283054,
+      "grad_norm": 0.6243147296091692,
+      "learning_rate": 4.987102023196898e-06,
+      "loss": 0.4643,
+      "step": 521
+    },
+    {
+      "epoch": 0.2430167597765363,
+      "grad_norm": 0.6878451059093437,
+      "learning_rate": 4.986971525168956e-06,
+      "loss": 0.4837,
+      "step": 522
+    },
+    {
+      "epoch": 0.24348230912476723,
+      "grad_norm": 0.6561267363912592,
+      "learning_rate": 4.986840372013574e-06,
+      "loss": 0.4781,
+      "step": 523
+    },
+    {
+      "epoch": 0.24394785847299813,
+      "grad_norm": 0.7108321527885036,
+      "learning_rate": 4.986708563765303e-06,
+      "loss": 0.4902,
+      "step": 524
+    },
+    {
+      "epoch": 0.24441340782122906,
+      "grad_norm": 0.6889210869235715,
+      "learning_rate": 4.986576100458862e-06,
+      "loss": 0.4886,
+      "step": 525
+    },
+    {
+      "epoch": 0.24487895716945995,
+      "grad_norm": 0.6257291149874021,
+      "learning_rate": 4.986442982129147e-06,
+      "loss": 0.4915,
+      "step": 526
+    },
+    {
+      "epoch": 0.24534450651769088,
+      "grad_norm": 0.6711549833676407,
+      "learning_rate": 4.986309208811223e-06,
+      "loss": 0.4981,
+      "step": 527
+    },
+    {
+      "epoch": 0.24581005586592178,
+      "grad_norm": 0.6600164225540271,
+      "learning_rate": 4.98617478054033e-06,
+      "loss": 0.4784,
+      "step": 528
+    },
+    {
+      "epoch": 0.2462756052141527,
+      "grad_norm": 0.6477459232519147,
+      "learning_rate": 4.986039697351878e-06,
+      "loss": 0.4802,
+      "step": 529
+    },
+    {
+      "epoch": 0.2467411545623836,
+      "grad_norm": 0.6820295489884041,
+      "learning_rate": 4.985903959281455e-06,
+      "loss": 0.5147,
+      "step": 530
+    },
+    {
+      "epoch": 0.24720670391061453,
+      "grad_norm": 0.6362703668945694,
+      "learning_rate": 4.985767566364814e-06,
+      "loss": 0.4656,
+      "step": 531
+    },
+    {
+      "epoch": 0.24767225325884543,
+      "grad_norm": 0.6345796536972468,
+      "learning_rate": 4.985630518637885e-06,
+      "loss": 0.4825,
+      "step": 532
+    },
+    {
+      "epoch": 0.24813780260707635,
+      "grad_norm": 0.67575870040398,
+      "learning_rate": 4.98549281613677e-06,
+      "loss": 0.5017,
+      "step": 533
+    },
+    {
+      "epoch": 0.24860335195530725,
+      "grad_norm": 0.6252249407194558,
+      "learning_rate": 4.9853544588977435e-06,
+      "loss": 0.4817,
+      "step": 534
+    },
+    {
+      "epoch": 0.24906890130353818,
+      "grad_norm": 0.622924366505422,
+      "learning_rate": 4.98521544695725e-06,
+      "loss": 0.4794,
+      "step": 535
+    },
+    {
+      "epoch": 0.24953445065176907,
+      "grad_norm": 0.6927500422702694,
+      "learning_rate": 4.9850757803519114e-06,
+      "loss": 0.4873,
+      "step": 536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6617434696140359,
+      "learning_rate": 4.984935459118518e-06,
+      "loss": 0.4859,
+      "step": 537
+    },
+    {
+      "epoch": 0.2504655493482309,
+      "grad_norm": 0.6248074911898552,
+      "learning_rate": 4.984794483294034e-06,
+      "loss": 0.4897,
+      "step": 538
+    },
+    {
+      "epoch": 0.25093109869646185,
+      "grad_norm": 0.6634656187277846,
+      "learning_rate": 4.984652852915595e-06,
+      "loss": 0.4816,
+      "step": 539
+    },
+    {
+      "epoch": 0.25139664804469275,
+      "grad_norm": 0.6447967488370139,
+      "learning_rate": 4.98451056802051e-06,
+      "loss": 0.4823,
+      "step": 540
+    },
+    {
+      "epoch": 0.25186219739292365,
+      "grad_norm": 0.6745772871346939,
+      "learning_rate": 4.984367628646261e-06,
+      "loss": 0.4748,
+      "step": 541
+    },
+    {
+      "epoch": 0.25232774674115455,
+      "grad_norm": 0.6708454823025314,
+      "learning_rate": 4.984224034830501e-06,
+      "loss": 0.4918,
+      "step": 542
+    },
+    {
+      "epoch": 0.2527932960893855,
+      "grad_norm": 0.8064207425324753,
+      "learning_rate": 4.984079786611054e-06,
+      "loss": 0.5194,
+      "step": 543
+    },
+    {
+      "epoch": 0.2532588454376164,
+      "grad_norm": 0.6661568404681422,
+      "learning_rate": 4.983934884025924e-06,
+      "loss": 0.4743,
+      "step": 544
+    },
+    {
+      "epoch": 0.2537243947858473,
+      "grad_norm": 0.6151847539209632,
+      "learning_rate": 4.983789327113277e-06,
+      "loss": 0.4752,
+      "step": 545
+    },
+    {
+      "epoch": 0.2541899441340782,
+      "grad_norm": 0.6124010556084097,
+      "learning_rate": 4.983643115911456e-06,
+      "loss": 0.4689,
+      "step": 546
+    },
+    {
+      "epoch": 0.25465549348230915,
+      "grad_norm": 0.6871886931628992,
+      "learning_rate": 4.9834962504589804e-06,
+      "loss": 0.4821,
+      "step": 547
+    },
+    {
+      "epoch": 0.25512104283054005,
+      "grad_norm": 0.6539867966013401,
+      "learning_rate": 4.983348730794534e-06,
+      "loss": 0.4955,
+      "step": 548
+    },
+    {
+      "epoch": 0.25558659217877094,
+      "grad_norm": 0.6775295863297914,
+      "learning_rate": 4.983200556956979e-06,
+      "loss": 0.4705,
+      "step": 549
+    },
+    {
+      "epoch": 0.25605214152700184,
+      "grad_norm": 0.6041066905821963,
+      "learning_rate": 4.983051728985348e-06,
+      "loss": 0.4739,
+      "step": 550
+    },
+    {
+      "epoch": 0.2565176908752328,
+      "grad_norm": 0.6248931365138353,
+      "learning_rate": 4.982902246918846e-06,
+      "loss": 0.4863,
+      "step": 551
+    },
+    {
+      "epoch": 0.2569832402234637,
+      "grad_norm": 0.6524292269320814,
+      "learning_rate": 4.982752110796848e-06,
+      "loss": 0.4881,
+      "step": 552
+    },
+    {
+      "epoch": 0.2574487895716946,
+      "grad_norm": 0.6161600956732515,
+      "learning_rate": 4.982601320658906e-06,
+      "loss": 0.4632,
+      "step": 553
+    },
+    {
+      "epoch": 0.2579143389199255,
+      "grad_norm": 0.6312963748965409,
+      "learning_rate": 4.9824498765447405e-06,
+      "loss": 0.4644,
+      "step": 554
+    },
+    {
+      "epoch": 0.25837988826815644,
+      "grad_norm": 0.7381674787640551,
+      "learning_rate": 4.982297778494245e-06,
+      "loss": 0.4853,
+      "step": 555
+    },
+    {
+      "epoch": 0.25884543761638734,
+      "grad_norm": 0.6304032789916956,
+      "learning_rate": 4.982145026547487e-06,
+      "loss": 0.4759,
+      "step": 556
+    },
+    {
+      "epoch": 0.25931098696461824,
+      "grad_norm": 0.6326575520725387,
+      "learning_rate": 4.981991620744704e-06,
+      "loss": 0.4801,
+      "step": 557
+    },
+    {
+      "epoch": 0.25977653631284914,
+      "grad_norm": 0.6531995529616806,
+      "learning_rate": 4.981837561126308e-06,
+      "loss": 0.4804,
+      "step": 558
+    },
+    {
+      "epoch": 0.2602420856610801,
+      "grad_norm": 0.6715261128500106,
+      "learning_rate": 4.98168284773288e-06,
+      "loss": 0.4738,
+      "step": 559
+    },
+    {
+      "epoch": 0.260707635009311,
+      "grad_norm": 0.6098855594756762,
+      "learning_rate": 4.981527480605177e-06,
+      "loss": 0.4694,
+      "step": 560
+    },
+    {
+      "epoch": 0.2611731843575419,
+      "grad_norm": 0.7770343352684186,
+      "learning_rate": 4.981371459784125e-06,
+      "loss": 0.4867,
+      "step": 561
+    },
+    {
+      "epoch": 0.2616387337057728,
+      "grad_norm": 0.6712811918989601,
+      "learning_rate": 4.981214785310825e-06,
+      "loss": 0.4942,
+      "step": 562
+    },
+    {
+      "epoch": 0.26210428305400374,
+      "grad_norm": 0.693179004938577,
+      "learning_rate": 4.981057457226547e-06,
+      "loss": 0.4962,
+      "step": 563
+    },
+    {
+      "epoch": 0.26256983240223464,
+      "grad_norm": 0.6775156315936247,
+      "learning_rate": 4.9808994755727365e-06,
+      "loss": 0.4856,
+      "step": 564
+    },
+    {
+      "epoch": 0.26303538175046554,
+      "grad_norm": 0.641443627449384,
+      "learning_rate": 4.980740840391009e-06,
+      "loss": 0.4779,
+      "step": 565
+    },
+    {
+      "epoch": 0.2635009310986965,
+      "grad_norm": 0.5903116669668282,
+      "learning_rate": 4.9805815517231534e-06,
+      "loss": 0.485,
+      "step": 566
+    },
+    {
+      "epoch": 0.2639664804469274,
+      "grad_norm": 0.6550720016849867,
+      "learning_rate": 4.980421609611131e-06,
+      "loss": 0.4758,
+      "step": 567
+    },
+    {
+      "epoch": 0.2644320297951583,
+      "grad_norm": 0.6625796092992569,
+      "learning_rate": 4.980261014097071e-06,
+      "loss": 0.462,
+      "step": 568
+    },
+    {
+      "epoch": 0.2648975791433892,
+      "grad_norm": 0.7027010739406956,
+      "learning_rate": 4.9800997652232794e-06,
+      "loss": 0.4921,
+      "step": 569
+    },
+    {
+      "epoch": 0.26536312849162014,
+      "grad_norm": 0.630543819176205,
+      "learning_rate": 4.979937863032235e-06,
+      "loss": 0.5159,
+      "step": 570
+    },
+    {
+      "epoch": 0.26582867783985104,
+      "grad_norm": 0.6479667078311206,
+      "learning_rate": 4.979775307566584e-06,
+      "loss": 0.4825,
+      "step": 571
+    },
+    {
+      "epoch": 0.26629422718808193,
+      "grad_norm": 0.6486539278614667,
+      "learning_rate": 4.97961209886915e-06,
+      "loss": 0.4612,
+      "step": 572
+    },
+    {
+      "epoch": 0.26675977653631283,
+      "grad_norm": 0.6599894885684692,
+      "learning_rate": 4.979448236982925e-06,
+      "loss": 0.4709,
+      "step": 573
+    },
+    {
+      "epoch": 0.2672253258845438,
+      "grad_norm": 0.6084028385241658,
+      "learning_rate": 4.979283721951073e-06,
+      "loss": 0.4726,
+      "step": 574
+    },
+    {
+      "epoch": 0.2676908752327747,
+      "grad_norm": 0.6594637359511528,
+      "learning_rate": 4.979118553816933e-06,
+      "loss": 0.4751,
+      "step": 575
+    },
+    {
+      "epoch": 0.2681564245810056,
+      "grad_norm": 0.6580972799083414,
+      "learning_rate": 4.978952732624012e-06,
+      "loss": 0.4838,
+      "step": 576
+    },
+    {
+      "epoch": 0.2686219739292365,
+      "grad_norm": 0.6297063563474037,
+      "learning_rate": 4.978786258415993e-06,
+      "loss": 0.4603,
+      "step": 577
+    },
+    {
+      "epoch": 0.26908752327746743,
+      "grad_norm": 0.6573586360175507,
+      "learning_rate": 4.97861913123673e-06,
+      "loss": 0.4775,
+      "step": 578
+    },
+    {
+      "epoch": 0.26955307262569833,
+      "grad_norm": 0.6748412394381442,
+      "learning_rate": 4.978451351130245e-06,
+      "loss": 0.4957,
+      "step": 579
+    },
+    {
+      "epoch": 0.27001862197392923,
+      "grad_norm": 0.6580108562858471,
+      "learning_rate": 4.978282918140738e-06,
+      "loss": 0.4801,
+      "step": 580
+    },
+    {
+      "epoch": 0.2704841713221601,
+      "grad_norm": 0.6117247103086192,
+      "learning_rate": 4.978113832312578e-06,
+      "loss": 0.4733,
+      "step": 581
+    },
+    {
+      "epoch": 0.2709497206703911,
+      "grad_norm": 0.6698962337472555,
+      "learning_rate": 4.977944093690306e-06,
+      "loss": 0.474,
+      "step": 582
+    },
+    {
+      "epoch": 0.271415270018622,
+      "grad_norm": 0.660055030962005,
+      "learning_rate": 4.977773702318635e-06,
+      "loss": 0.4594,
+      "step": 583
+    },
+    {
+      "epoch": 0.2718808193668529,
+      "grad_norm": 0.6738824553276879,
+      "learning_rate": 4.9776026582424485e-06,
+      "loss": 0.479,
+      "step": 584
+    },
+    {
+      "epoch": 0.2723463687150838,
+      "grad_norm": 0.6429100798886866,
+      "learning_rate": 4.977430961506807e-06,
+      "loss": 0.4773,
+      "step": 585
+    },
+    {
+      "epoch": 0.27281191806331473,
+      "grad_norm": 0.6849203226758223,
+      "learning_rate": 4.9772586121569365e-06,
+      "loss": 0.49,
+      "step": 586
+    },
+    {
+      "epoch": 0.2732774674115456,
+      "grad_norm": 0.6481713569191861,
+      "learning_rate": 4.97708561023824e-06,
+      "loss": 0.5025,
+      "step": 587
+    },
+    {
+      "epoch": 0.2737430167597765,
+      "grad_norm": 0.6860496413511674,
+      "learning_rate": 4.976911955796289e-06,
+      "loss": 0.46,
+      "step": 588
+    },
+    {
+      "epoch": 0.2742085661080074,
+      "grad_norm": 0.6384986310127626,
+      "learning_rate": 4.976737648876829e-06,
+      "loss": 0.4632,
+      "step": 589
+    },
+    {
+      "epoch": 0.2746741154562384,
+      "grad_norm": 0.699218886430182,
+      "learning_rate": 4.976562689525777e-06,
+      "loss": 0.4984,
+      "step": 590
+    },
+    {
+      "epoch": 0.2751396648044693,
+      "grad_norm": 0.6536269654380361,
+      "learning_rate": 4.976387077789219e-06,
+      "loss": 0.4623,
+      "step": 591
+    },
+    {
+      "epoch": 0.2756052141527002,
+      "grad_norm": 0.6422736499760907,
+      "learning_rate": 4.976210813713418e-06,
+      "loss": 0.4835,
+      "step": 592
+    },
+    {
+      "epoch": 0.27607076350093107,
+      "grad_norm": 0.6712896612596988,
+      "learning_rate": 4.976033897344804e-06,
+      "loss": 0.4629,
+      "step": 593
+    },
+    {
+      "epoch": 0.276536312849162,
+      "grad_norm": 0.6556567484301513,
+      "learning_rate": 4.975856328729983e-06,
+      "loss": 0.4902,
+      "step": 594
+    },
+    {
+      "epoch": 0.2770018621973929,
+      "grad_norm": 0.6319327178772516,
+      "learning_rate": 4.97567810791573e-06,
+      "loss": 0.4847,
+      "step": 595
+    },
+    {
+      "epoch": 0.2774674115456238,
+      "grad_norm": 0.6365348242576997,
+      "learning_rate": 4.9754992349489924e-06,
+      "loss": 0.475,
+      "step": 596
+    },
+    {
+      "epoch": 0.2779329608938548,
+      "grad_norm": 0.6752902343235576,
+      "learning_rate": 4.97531970987689e-06,
+      "loss": 0.4604,
+      "step": 597
+    },
+    {
+      "epoch": 0.2783985102420857,
+      "grad_norm": 0.6555270472810107,
+      "learning_rate": 4.975139532746714e-06,
+      "loss": 0.4664,
+      "step": 598
+    },
+    {
+      "epoch": 0.27886405959031657,
+      "grad_norm": 0.614456233494169,
+      "learning_rate": 4.974958703605926e-06,
+      "loss": 0.471,
+      "step": 599
+    },
+    {
+      "epoch": 0.27932960893854747,
+      "grad_norm": 0.663237449748165,
+      "learning_rate": 4.974777222502163e-06,
+      "loss": 0.4628,
+      "step": 600
+    },
+    {
+      "epoch": 0.2797951582867784,
+      "grad_norm": 0.6471454592300786,
+      "learning_rate": 4.974595089483228e-06,
+      "loss": 0.4742,
+      "step": 601
+    },
+    {
+      "epoch": 0.2802607076350093,
+      "grad_norm": 0.6876395682174765,
+      "learning_rate": 4.974412304597104e-06,
+      "loss": 0.5116,
+      "step": 602
+    },
+    {
+      "epoch": 0.2807262569832402,
+      "grad_norm": 0.6633046793992097,
+      "learning_rate": 4.974228867891936e-06,
+      "loss": 0.4977,
+      "step": 603
+    },
+    {
+      "epoch": 0.2811918063314711,
+      "grad_norm": 0.6356378174533838,
+      "learning_rate": 4.97404477941605e-06,
+      "loss": 0.4821,
+      "step": 604
+    },
+    {
+      "epoch": 0.28165735567970207,
+      "grad_norm": 0.6498808159784766,
+      "learning_rate": 4.973860039217935e-06,
+      "loss": 0.4643,
+      "step": 605
+    },
+    {
+      "epoch": 0.28212290502793297,
+      "grad_norm": 0.6254821857544629,
+      "learning_rate": 4.973674647346259e-06,
+      "loss": 0.4641,
+      "step": 606
+    },
+    {
+      "epoch": 0.28258845437616387,
+      "grad_norm": 0.5950436326383921,
+      "learning_rate": 4.973488603849857e-06,
+      "loss": 0.4612,
+      "step": 607
+    },
+    {
+      "epoch": 0.28305400372439476,
+      "grad_norm": 0.6293655186723873,
+      "learning_rate": 4.973301908777739e-06,
+      "loss": 0.4522,
+      "step": 608
+    },
+    {
+      "epoch": 0.2835195530726257,
+      "grad_norm": 0.6827391440686623,
+      "learning_rate": 4.973114562179083e-06,
+      "loss": 0.4909,
+      "step": 609
+    },
+    {
+      "epoch": 0.2839851024208566,
+      "grad_norm": 0.6510156670246277,
+      "learning_rate": 4.972926564103242e-06,
+      "loss": 0.483,
+      "step": 610
+    },
+    {
+      "epoch": 0.2844506517690875,
+      "grad_norm": 0.6144715841092306,
+      "learning_rate": 4.972737914599739e-06,
+      "loss": 0.475,
+      "step": 611
+    },
+    {
+      "epoch": 0.2849162011173184,
+      "grad_norm": 0.6440471730209673,
+      "learning_rate": 4.972548613718267e-06,
+      "loss": 0.4808,
+      "step": 612
+    },
+    {
+      "epoch": 0.28538175046554937,
+      "grad_norm": 0.678769380375081,
+      "learning_rate": 4.972358661508695e-06,
+      "loss": 0.4615,
+      "step": 613
+    },
+    {
+      "epoch": 0.28584729981378026,
+      "grad_norm": 0.6302304095829889,
+      "learning_rate": 4.972168058021059e-06,
+      "loss": 0.4532,
+      "step": 614
+    },
+    {
+      "epoch": 0.28631284916201116,
+      "grad_norm": 0.6428247972746336,
+      "learning_rate": 4.9719768033055695e-06,
+      "loss": 0.4879,
+      "step": 615
+    },
+    {
+      "epoch": 0.28677839851024206,
+      "grad_norm": 0.6843346418534972,
+      "learning_rate": 4.971784897412607e-06,
+      "loss": 0.4815,
+      "step": 616
+    },
+    {
+      "epoch": 0.287243947858473,
+      "grad_norm": 0.7076639518324503,
+      "learning_rate": 4.971592340392723e-06,
+      "loss": 0.4861,
+      "step": 617
+    },
+    {
+      "epoch": 0.2877094972067039,
+      "grad_norm": 0.6210754973872052,
+      "learning_rate": 4.971399132296645e-06,
+      "loss": 0.4584,
+      "step": 618
+    },
+    {
+      "epoch": 0.2881750465549348,
+      "grad_norm": 0.6273882289087482,
+      "learning_rate": 4.971205273175265e-06,
+      "loss": 0.4849,
+      "step": 619
+    },
+    {
+      "epoch": 0.2886405959031657,
+      "grad_norm": 0.6574455982172519,
+      "learning_rate": 4.971010763079652e-06,
+      "loss": 0.4967,
+      "step": 620
+    },
+    {
+      "epoch": 0.28910614525139666,
+      "grad_norm": 0.6219891923611582,
+      "learning_rate": 4.970815602061044e-06,
+      "loss": 0.4811,
+      "step": 621
+    },
+    {
+      "epoch": 0.28957169459962756,
+      "grad_norm": 0.6225391454851527,
+      "learning_rate": 4.970619790170852e-06,
+      "loss": 0.4748,
+      "step": 622
+    },
+    {
+      "epoch": 0.29003724394785846,
+      "grad_norm": 0.6678756351380153,
+      "learning_rate": 4.970423327460656e-06,
+      "loss": 0.4544,
+      "step": 623
+    },
+    {
+      "epoch": 0.2905027932960894,
+      "grad_norm": 0.727703025156093,
+      "learning_rate": 4.97022621398221e-06,
+      "loss": 0.5151,
+      "step": 624
+    },
+    {
+      "epoch": 0.2909683426443203,
+      "grad_norm": 0.5875471193237926,
+      "learning_rate": 4.9700284497874374e-06,
+      "loss": 0.4628,
+      "step": 625
+    },
+    {
+      "epoch": 0.2914338919925512,
+      "grad_norm": 0.6207287017779713,
+      "learning_rate": 4.969830034928435e-06,
+      "loss": 0.4749,
+      "step": 626
+    },
+    {
+      "epoch": 0.2918994413407821,
+      "grad_norm": 0.6122223565367133,
+      "learning_rate": 4.96963096945747e-06,
+      "loss": 0.4749,
+      "step": 627
+    },
+    {
+      "epoch": 0.29236499068901306,
+      "grad_norm": 0.5888367576874958,
+      "learning_rate": 4.969431253426979e-06,
+      "loss": 0.4642,
+      "step": 628
+    },
+    {
+      "epoch": 0.29283054003724396,
+      "grad_norm": 0.633230676457288,
+      "learning_rate": 4.969230886889575e-06,
+      "loss": 0.4611,
+      "step": 629
+    },
+    {
+      "epoch": 0.29329608938547486,
+      "grad_norm": 0.6386451383634858,
+      "learning_rate": 4.969029869898038e-06,
+      "loss": 0.4945,
+      "step": 630
+    },
+    {
+      "epoch": 0.29376163873370575,
+      "grad_norm": 0.63092744102819,
+      "learning_rate": 4.968828202505321e-06,
+      "loss": 0.4899,
+      "step": 631
+    },
+    {
+      "epoch": 0.2942271880819367,
+      "grad_norm": 0.6425525692510823,
+      "learning_rate": 4.968625884764545e-06,
+      "loss": 0.465,
+      "step": 632
+    },
+    {
+      "epoch": 0.2946927374301676,
+      "grad_norm": 0.6032843937466621,
+      "learning_rate": 4.96842291672901e-06,
+      "loss": 0.4931,
+      "step": 633
+    },
+    {
+      "epoch": 0.2951582867783985,
+      "grad_norm": 0.6661519573265277,
+      "learning_rate": 4.96821929845218e-06,
+      "loss": 0.4834,
+      "step": 634
+    },
+    {
+      "epoch": 0.2956238361266294,
+      "grad_norm": 0.657128637988324,
+      "learning_rate": 4.968015029987693e-06,
+      "loss": 0.47,
+      "step": 635
+    },
+    {
+      "epoch": 0.29608938547486036,
+      "grad_norm": 0.6390004735905854,
+      "learning_rate": 4.967810111389358e-06,
+      "loss": 0.4737,
+      "step": 636
+    },
+    {
+      "epoch": 0.29655493482309125,
+      "grad_norm": 0.704586690206763,
+      "learning_rate": 4.967604542711157e-06,
+      "loss": 0.4794,
+      "step": 637
+    },
+    {
+      "epoch": 0.29702048417132215,
+      "grad_norm": 0.610120482778731,
+      "learning_rate": 4.96739832400724e-06,
+      "loss": 0.4688,
+      "step": 638
+    },
+    {
+      "epoch": 0.29748603351955305,
+      "grad_norm": 0.6279081339886109,
+      "learning_rate": 4.96719145533193e-06,
+      "loss": 0.4762,
+      "step": 639
+    },
+    {
+      "epoch": 0.297951582867784,
+      "grad_norm": 0.6036410725452863,
+      "learning_rate": 4.9669839367397226e-06,
+      "loss": 0.4742,
+      "step": 640
+    },
+    {
+      "epoch": 0.2984171322160149,
+      "grad_norm": 0.6578092694575863,
+      "learning_rate": 4.966775768285281e-06,
+      "loss": 0.4767,
+      "step": 641
+    },
+    {
+      "epoch": 0.2988826815642458,
+      "grad_norm": 0.7334142682160777,
+      "learning_rate": 4.966566950023444e-06,
+      "loss": 0.4858,
+      "step": 642
+    },
+    {
+      "epoch": 0.2993482309124767,
+      "grad_norm": 0.6849406987359092,
+      "learning_rate": 4.966357482009218e-06,
+      "loss": 0.4912,
+      "step": 643
+    },
+    {
+      "epoch": 0.29981378026070765,
+      "grad_norm": 0.6306698971005936,
+      "learning_rate": 4.966147364297782e-06,
+      "loss": 0.4688,
+      "step": 644
+    },
+    {
+      "epoch": 0.30027932960893855,
+      "grad_norm": 0.6444742898307945,
+      "learning_rate": 4.965936596944485e-06,
+      "loss": 0.4555,
+      "step": 645
+    },
+    {
+      "epoch": 0.30074487895716945,
+      "grad_norm": 0.6411179493681214,
+      "learning_rate": 4.965725180004851e-06,
+      "loss": 0.4711,
+      "step": 646
+    },
+    {
+      "epoch": 0.30121042830540035,
+      "grad_norm": 0.63326161019324,
+      "learning_rate": 4.965513113534569e-06,
+      "loss": 0.5166,
+      "step": 647
+    },
+    {
+      "epoch": 0.3016759776536313,
+      "grad_norm": 0.6423995515142575,
+      "learning_rate": 4.965300397589505e-06,
+      "loss": 0.4596,
+      "step": 648
+    },
+    {
+      "epoch": 0.3021415270018622,
+      "grad_norm": 0.6228325633227516,
+      "learning_rate": 4.965087032225691e-06,
+      "loss": 0.4535,
+      "step": 649
+    },
+    {
+      "epoch": 0.3026070763500931,
+      "grad_norm": 0.5961431906625267,
+      "learning_rate": 4.964873017499333e-06,
+      "loss": 0.4576,
+      "step": 650
+    },
+    {
+      "epoch": 0.30307262569832405,
+      "grad_norm": 0.6769735876941761,
+      "learning_rate": 4.96465835346681e-06,
+      "loss": 0.4859,
+      "step": 651
+    },
+    {
+      "epoch": 0.30353817504655495,
+      "grad_norm": 0.6573101746060285,
+      "learning_rate": 4.964443040184667e-06,
+      "loss": 0.4631,
+      "step": 652
+    },
+    {
+      "epoch": 0.30400372439478585,
+      "grad_norm": 0.6618302172773146,
+      "learning_rate": 4.9642270777096236e-06,
+      "loss": 0.468,
+      "step": 653
+    },
+    {
+      "epoch": 0.30446927374301674,
+      "grad_norm": 0.6247044284420775,
+      "learning_rate": 4.96401046609857e-06,
+      "loss": 0.4701,
+      "step": 654
+    },
+    {
+      "epoch": 0.3049348230912477,
+      "grad_norm": 0.6042399958140557,
+      "learning_rate": 4.963793205408566e-06,
+      "loss": 0.4749,
+      "step": 655
+    },
+    {
+      "epoch": 0.3054003724394786,
+      "grad_norm": 0.6312241010945346,
+      "learning_rate": 4.963575295696843e-06,
+      "loss": 0.4668,
+      "step": 656
+    },
+    {
+      "epoch": 0.3058659217877095,
+      "grad_norm": 0.6401576027489495,
+      "learning_rate": 4.963356737020804e-06,
+      "loss": 0.4678,
+      "step": 657
+    },
+    {
+      "epoch": 0.3063314711359404,
+      "grad_norm": 0.6762519825693567,
+      "learning_rate": 4.963137529438023e-06,
+      "loss": 0.4879,
+      "step": 658
+    },
+    {
+      "epoch": 0.30679702048417135,
+      "grad_norm": 0.6463039052905999,
+      "learning_rate": 4.962917673006244e-06,
+      "loss": 0.46,
+      "step": 659
+    },
+    {
+      "epoch": 0.30726256983240224,
+      "grad_norm": 0.6766936316867374,
+      "learning_rate": 4.962697167783382e-06,
+      "loss": 0.475,
+      "step": 660
+    },
+    {
+      "epoch": 0.30772811918063314,
+      "grad_norm": 0.631717107843128,
+      "learning_rate": 4.962476013827525e-06,
+      "loss": 0.4694,
+      "step": 661
+    },
+    {
+      "epoch": 0.30819366852886404,
+      "grad_norm": 0.6311740717005658,
+      "learning_rate": 4.962254211196928e-06,
+      "loss": 0.4775,
+      "step": 662
+    },
+    {
+      "epoch": 0.308659217877095,
+      "grad_norm": 0.6024420932649136,
+      "learning_rate": 4.962031759950021e-06,
+      "loss": 0.4676,
+      "step": 663
+    },
+    {
+      "epoch": 0.3091247672253259,
+      "grad_norm": 0.7217831156738872,
+      "learning_rate": 4.961808660145402e-06,
+      "loss": 0.4772,
+      "step": 664
+    },
+    {
+      "epoch": 0.3095903165735568,
+      "grad_norm": 0.6650601132587651,
+      "learning_rate": 4.96158491184184e-06,
+      "loss": 0.4671,
+      "step": 665
+    },
+    {
+      "epoch": 0.3100558659217877,
+      "grad_norm": 0.6346662523289833,
+      "learning_rate": 4.961360515098277e-06,
+      "loss": 0.4699,
+      "step": 666
+    },
+    {
+      "epoch": 0.31052141527001864,
+      "grad_norm": 0.694664436359774,
+      "learning_rate": 4.961135469973824e-06,
+      "loss": 0.4932,
+      "step": 667
+    },
+    {
+      "epoch": 0.31098696461824954,
+      "grad_norm": 0.6284190364086008,
+      "learning_rate": 4.9609097765277635e-06,
+      "loss": 0.4754,
+      "step": 668
+    },
+    {
+      "epoch": 0.31145251396648044,
+      "grad_norm": 0.6779212871923593,
+      "learning_rate": 4.960683434819547e-06,
+      "loss": 0.4683,
+      "step": 669
+    },
+    {
+      "epoch": 0.31191806331471134,
+      "grad_norm": 0.6760045342173218,
+      "learning_rate": 4.960456444908801e-06,
+      "loss": 0.4957,
+      "step": 670
+    },
+    {
+      "epoch": 0.3123836126629423,
+      "grad_norm": 0.628580075974176,
+      "learning_rate": 4.960228806855317e-06,
+      "loss": 0.4762,
+      "step": 671
+    },
+    {
+      "epoch": 0.3128491620111732,
+      "grad_norm": 0.6517745008606406,
+      "learning_rate": 4.960000520719063e-06,
+      "loss": 0.4763,
+      "step": 672
+    },
+    {
+      "epoch": 0.3133147113594041,
+      "grad_norm": 0.6534370918898651,
+      "learning_rate": 4.959771586560174e-06,
+      "loss": 0.476,
+      "step": 673
+    },
+    {
+      "epoch": 0.313780260707635,
+      "grad_norm": 0.6581330987146,
+      "learning_rate": 4.959542004438955e-06,
+      "loss": 0.4665,
+      "step": 674
+    },
+    {
+      "epoch": 0.31424581005586594,
+      "grad_norm": 0.6772929811030347,
+      "learning_rate": 4.959311774415886e-06,
+      "loss": 0.4775,
+      "step": 675
+    },
+    {
+      "epoch": 0.31471135940409684,
+      "grad_norm": 0.6149844188528272,
+      "learning_rate": 4.959080896551614e-06,
+      "loss": 0.4494,
+      "step": 676
+    },
+    {
+      "epoch": 0.31517690875232773,
+      "grad_norm": 0.6760026469691609,
+      "learning_rate": 4.958849370906957e-06,
+      "loss": 0.4736,
+      "step": 677
+    },
+    {
+      "epoch": 0.31564245810055863,
+      "grad_norm": 0.6311607329895665,
+      "learning_rate": 4.958617197542906e-06,
+      "loss": 0.4891,
+      "step": 678
+    },
+    {
+      "epoch": 0.3161080074487896,
+      "grad_norm": 0.598685628155277,
+      "learning_rate": 4.958384376520619e-06,
+      "loss": 0.4712,
+      "step": 679
+    },
+    {
+      "epoch": 0.3165735567970205,
+      "grad_norm": 0.698847786256506,
+      "learning_rate": 4.958150907901429e-06,
+      "loss": 0.4832,
+      "step": 680
+    },
+    {
+      "epoch": 0.3170391061452514,
+      "grad_norm": 0.6937627800610745,
+      "learning_rate": 4.957916791746834e-06,
+      "loss": 0.4781,
+      "step": 681
+    },
+    {
+      "epoch": 0.31750465549348234,
+      "grad_norm": 0.6628695606827939,
+      "learning_rate": 4.957682028118509e-06,
+      "loss": 0.4597,
+      "step": 682
+    },
+    {
+      "epoch": 0.31797020484171323,
+      "grad_norm": 0.6692053162275723,
+      "learning_rate": 4.957446617078294e-06,
+      "loss": 0.4782,
+      "step": 683
+    },
+    {
+      "epoch": 0.31843575418994413,
+      "grad_norm": 0.6393777733436068,
+      "learning_rate": 4.957210558688203e-06,
+      "loss": 0.4755,
+      "step": 684
+    },
+    {
+      "epoch": 0.31890130353817503,
+      "grad_norm": 0.6609496359016922,
+      "learning_rate": 4.956973853010419e-06,
+      "loss": 0.4695,
+      "step": 685
+    },
+    {
+      "epoch": 0.319366852886406,
+      "grad_norm": 0.6159968482127901,
+      "learning_rate": 4.956736500107296e-06,
+      "loss": 0.4634,
+      "step": 686
+    },
+    {
+      "epoch": 0.3198324022346369,
+      "grad_norm": 0.5990356467963264,
+      "learning_rate": 4.95649850004136e-06,
+      "loss": 0.4708,
+      "step": 687
+    },
+    {
+      "epoch": 0.3202979515828678,
+      "grad_norm": 0.6880588625027413,
+      "learning_rate": 4.956259852875303e-06,
+      "loss": 0.465,
+      "step": 688
+    },
+    {
+      "epoch": 0.3207635009310987,
+      "grad_norm": 0.6489086996172685,
+      "learning_rate": 4.956020558671991e-06,
+      "loss": 0.4733,
+      "step": 689
+    },
+    {
+      "epoch": 0.32122905027932963,
+      "grad_norm": 0.6400860706614486,
+      "learning_rate": 4.955780617494461e-06,
+      "loss": 0.4748,
+      "step": 690
+    },
+    {
+      "epoch": 0.32169459962756053,
+      "grad_norm": 0.6752921160951707,
+      "learning_rate": 4.955540029405919e-06,
+      "loss": 0.4633,
+      "step": 691
+    },
+    {
+      "epoch": 0.3221601489757914,
+      "grad_norm": 0.6403002116850084,
+      "learning_rate": 4.955298794469741e-06,
+      "loss": 0.4846,
+      "step": 692
+    },
+    {
+      "epoch": 0.3226256983240223,
+      "grad_norm": 0.6146440292482699,
+      "learning_rate": 4.955056912749475e-06,
+      "loss": 0.4792,
+      "step": 693
+    },
+    {
+      "epoch": 0.3230912476722533,
+      "grad_norm": 0.7219674906871857,
+      "learning_rate": 4.954814384308837e-06,
+      "loss": 0.4694,
+      "step": 694
+    },
+    {
+      "epoch": 0.3235567970204842,
+      "grad_norm": 0.7656022274893002,
+      "learning_rate": 4.954571209211715e-06,
+      "loss": 0.5137,
+      "step": 695
+    },
+    {
+      "epoch": 0.3240223463687151,
+      "grad_norm": 0.6135809161611907,
+      "learning_rate": 4.954327387522168e-06,
+      "loss": 0.4753,
+      "step": 696
+    },
+    {
+      "epoch": 0.324487895716946,
+      "grad_norm": 0.6344658067097666,
+      "learning_rate": 4.954082919304425e-06,
+      "loss": 0.4827,
+      "step": 697
+    },
+    {
+      "epoch": 0.3249534450651769,
+      "grad_norm": 0.5986892313614238,
+      "learning_rate": 4.953837804622882e-06,
+      "loss": 0.4579,
+      "step": 698
+    },
+    {
+      "epoch": 0.3254189944134078,
+      "grad_norm": 0.6893354134986406,
+      "learning_rate": 4.95359204354211e-06,
+      "loss": 0.4516,
+      "step": 699
+    },
+    {
+      "epoch": 0.3258845437616387,
+      "grad_norm": 0.6379193422985298,
+      "learning_rate": 4.9533456361268496e-06,
+      "loss": 0.4647,
+      "step": 700
+    },
+    {
+      "epoch": 0.3263500931098696,
+      "grad_norm": 0.6660698313962877,
+      "learning_rate": 4.953098582442007e-06,
+      "loss": 0.4738,
+      "step": 701
+    },
+    {
+      "epoch": 0.3268156424581006,
+      "grad_norm": 0.6216656794326816,
+      "learning_rate": 4.952850882552665e-06,
+      "loss": 0.4761,
+      "step": 702
+    },
+    {
+      "epoch": 0.3272811918063315,
+      "grad_norm": 0.6577258660237737,
+      "learning_rate": 4.9526025365240725e-06,
+      "loss": 0.4893,
+      "step": 703
+    },
+    {
+      "epoch": 0.32774674115456237,
+      "grad_norm": 0.6716369250161917,
+      "learning_rate": 4.95235354442165e-06,
+      "loss": 0.463,
+      "step": 704
+    },
+    {
+      "epoch": 0.32821229050279327,
+      "grad_norm": 0.6538711131573695,
+      "learning_rate": 4.952103906310986e-06,
+      "loss": 0.4849,
+      "step": 705
+    },
+    {
+      "epoch": 0.3286778398510242,
+      "grad_norm": 0.7053243585399566,
+      "learning_rate": 4.951853622257845e-06,
+      "loss": 0.4703,
+      "step": 706
+    },
+    {
+      "epoch": 0.3291433891992551,
+      "grad_norm": 0.6552583097322,
+      "learning_rate": 4.951602692328154e-06,
+      "loss": 0.4647,
+      "step": 707
+    },
+    {
+      "epoch": 0.329608938547486,
+      "grad_norm": 0.6284469542603011,
+      "learning_rate": 4.951351116588017e-06,
+      "loss": 0.4693,
+      "step": 708
+    },
+    {
+      "epoch": 0.330074487895717,
+      "grad_norm": 0.715567700635531,
+      "learning_rate": 4.951098895103702e-06,
+      "loss": 0.4686,
+      "step": 709
+    },
+    {
+      "epoch": 0.33054003724394787,
+      "grad_norm": 0.6540098129868619,
+      "learning_rate": 4.950846027941653e-06,
+      "loss": 0.4854,
+      "step": 710
+    },
+    {
+      "epoch": 0.33100558659217877,
+      "grad_norm": 0.6471406522608392,
+      "learning_rate": 4.950592515168479e-06,
+      "loss": 0.4741,
+      "step": 711
+    },
+    {
+      "epoch": 0.33147113594040967,
+      "grad_norm": 0.6671897124997558,
+      "learning_rate": 4.950338356850962e-06,
+      "loss": 0.4805,
+      "step": 712
+    },
+    {
+      "epoch": 0.3319366852886406,
+      "grad_norm": 0.6391431111232587,
+      "learning_rate": 4.9500835530560535e-06,
+      "loss": 0.4933,
+      "step": 713
+    },
+    {
+      "epoch": 0.3324022346368715,
+      "grad_norm": 0.6420709725985829,
+      "learning_rate": 4.9498281038508755e-06,
+      "loss": 0.4787,
+      "step": 714
+    },
+    {
+      "epoch": 0.3328677839851024,
+      "grad_norm": 0.641114378528786,
+      "learning_rate": 4.949572009302719e-06,
+      "loss": 0.4927,
+      "step": 715
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.6514024737874371,
+      "learning_rate": 4.949315269479045e-06,
+      "loss": 0.4843,
+      "step": 716
+    },
+    {
+      "epoch": 0.33379888268156427,
+      "grad_norm": 0.6412126329526349,
+      "learning_rate": 4.949057884447485e-06,
+      "loss": 0.4641,
+      "step": 717
+    },
+    {
+      "epoch": 0.33426443202979517,
+      "grad_norm": 0.641408568986961,
+      "learning_rate": 4.94879985427584e-06,
+      "loss": 0.4832,
+      "step": 718
+    },
+    {
+      "epoch": 0.33472998137802606,
+      "grad_norm": 0.6049862514481887,
+      "learning_rate": 4.948541179032081e-06,
+      "loss": 0.4871,
+      "step": 719
+    },
+    {
+      "epoch": 0.33519553072625696,
+      "grad_norm": 0.6245779838421307,
+      "learning_rate": 4.948281858784351e-06,
+      "loss": 0.4862,
+      "step": 720
+    },
+    {
+      "epoch": 0.3356610800744879,
+      "grad_norm": 0.6097940353727755,
+      "learning_rate": 4.94802189360096e-06,
+      "loss": 0.4677,
+      "step": 721
+    },
+    {
+      "epoch": 0.3361266294227188,
+      "grad_norm": 0.5964302300176993,
+      "learning_rate": 4.947761283550389e-06,
+      "loss": 0.4586,
+      "step": 722
+    },
+    {
+      "epoch": 0.3365921787709497,
+      "grad_norm": 0.6808780197596234,
+      "learning_rate": 4.947500028701288e-06,
+      "loss": 0.4916,
+      "step": 723
+    },
+    {
+      "epoch": 0.3370577281191806,
+      "grad_norm": 0.6478084542614099,
+      "learning_rate": 4.9472381291224805e-06,
+      "loss": 0.482,
+      "step": 724
+    },
+    {
+      "epoch": 0.33752327746741156,
+      "grad_norm": 0.620217429644143,
+      "learning_rate": 4.946975584882954e-06,
+      "loss": 0.4753,
+      "step": 725
+    },
+    {
+      "epoch": 0.33798882681564246,
+      "grad_norm": 0.665445023826721,
+      "learning_rate": 4.946712396051871e-06,
+      "loss": 0.4574,
+      "step": 726
+    },
+    {
+      "epoch": 0.33845437616387336,
+      "grad_norm": 0.6337269383708976,
+      "learning_rate": 4.9464485626985595e-06,
+      "loss": 0.4954,
+      "step": 727
+    },
+    {
+      "epoch": 0.33891992551210426,
+      "grad_norm": 0.6639194865926016,
+      "learning_rate": 4.946184084892522e-06,
+      "loss": 0.4861,
+      "step": 728
+    },
+    {
+      "epoch": 0.3393854748603352,
+      "grad_norm": 0.702603850058697,
+      "learning_rate": 4.945918962703426e-06,
+      "loss": 0.4788,
+      "step": 729
+    },
+    {
+      "epoch": 0.3398510242085661,
+      "grad_norm": 0.6380145120337463,
+      "learning_rate": 4.945653196201113e-06,
+      "loss": 0.4604,
+      "step": 730
+    },
+    {
+      "epoch": 0.340316573556797,
+      "grad_norm": 0.7004992418791834,
+      "learning_rate": 4.94538678545559e-06,
+      "loss": 0.4786,
+      "step": 731
+    },
+    {
+      "epoch": 0.3407821229050279,
+      "grad_norm": 0.6247540945973492,
+      "learning_rate": 4.945119730537038e-06,
+      "loss": 0.4897,
+      "step": 732
+    },
+    {
+      "epoch": 0.34124767225325886,
+      "grad_norm": 0.6763335148501232,
+      "learning_rate": 4.944852031515804e-06,
+      "loss": 0.4569,
+      "step": 733
+    },
+    {
+      "epoch": 0.34171322160148976,
+      "grad_norm": 0.6599412465588316,
+      "learning_rate": 4.944583688462408e-06,
+      "loss": 0.4905,
+      "step": 734
+    },
+    {
+      "epoch": 0.34217877094972066,
+      "grad_norm": 0.6873364688177377,
+      "learning_rate": 4.944314701447537e-06,
+      "loss": 0.4934,
+      "step": 735
+    },
+    {
+      "epoch": 0.3426443202979516,
+      "grad_norm": 0.6803271985860735,
+      "learning_rate": 4.944045070542048e-06,
+      "loss": 0.4559,
+      "step": 736
+    },
+    {
+      "epoch": 0.3431098696461825,
+      "grad_norm": 0.6213121687451557,
+      "learning_rate": 4.943774795816969e-06,
+      "loss": 0.4491,
+      "step": 737
+    },
+    {
+      "epoch": 0.3435754189944134,
+      "grad_norm": 0.6488943505507518,
+      "learning_rate": 4.943503877343497e-06,
+      "loss": 0.4742,
+      "step": 738
+    },
+    {
+      "epoch": 0.3440409683426443,
+      "grad_norm": 0.7004187518604765,
+      "learning_rate": 4.943232315192998e-06,
+      "loss": 0.4661,
+      "step": 739
+    },
+    {
+      "epoch": 0.34450651769087526,
+      "grad_norm": 0.6141455365115432,
+      "learning_rate": 4.942960109437008e-06,
+      "loss": 0.4655,
+      "step": 740
+    },
+    {
+      "epoch": 0.34497206703910616,
+      "grad_norm": 0.6260328511000656,
+      "learning_rate": 4.942687260147232e-06,
+      "loss": 0.4897,
+      "step": 741
+    },
+    {
+      "epoch": 0.34543761638733705,
+      "grad_norm": 0.6185197569899994,
+      "learning_rate": 4.942413767395545e-06,
+      "loss": 0.4745,
+      "step": 742
+    },
+    {
+      "epoch": 0.34590316573556795,
+      "grad_norm": 0.6228988625250969,
+      "learning_rate": 4.942139631253993e-06,
+      "loss": 0.4594,
+      "step": 743
+    },
+    {
+      "epoch": 0.3463687150837989,
+      "grad_norm": 0.6922795300575871,
+      "learning_rate": 4.941864851794787e-06,
+      "loss": 0.4575,
+      "step": 744
+    },
+    {
+      "epoch": 0.3468342644320298,
+      "grad_norm": 0.6871842216786143,
+      "learning_rate": 4.941589429090313e-06,
+      "loss": 0.4795,
+      "step": 745
+    },
+    {
+      "epoch": 0.3472998137802607,
+      "grad_norm": 0.6836216052372917,
+      "learning_rate": 4.9413133632131236e-06,
+      "loss": 0.4725,
+      "step": 746
+    },
+    {
+      "epoch": 0.3477653631284916,
+      "grad_norm": 0.6256032152995039,
+      "learning_rate": 4.941036654235941e-06,
+      "loss": 0.4713,
+      "step": 747
+    },
+    {
+      "epoch": 0.34823091247672255,
+      "grad_norm": 0.6325486018398588,
+      "learning_rate": 4.940759302231656e-06,
+      "loss": 0.4741,
+      "step": 748
+    },
+    {
+      "epoch": 0.34869646182495345,
+      "grad_norm": 0.6843542353365593,
+      "learning_rate": 4.9404813072733295e-06,
+      "loss": 0.4768,
+      "step": 749
+    },
+    {
+      "epoch": 0.34916201117318435,
+      "grad_norm": 0.6731714968055886,
+      "learning_rate": 4.940202669434193e-06,
+      "loss": 0.4839,
+      "step": 750
+    },
+    {
+      "epoch": 0.34962756052141525,
+      "grad_norm": 0.6752220853024181,
+      "learning_rate": 4.939923388787647e-06,
+      "loss": 0.4713,
+      "step": 751
+    },
+    {
+      "epoch": 0.3500931098696462,
+      "grad_norm": 0.6198323772173456,
+      "learning_rate": 4.93964346540726e-06,
+      "loss": 0.4688,
+      "step": 752
+    },
+    {
+      "epoch": 0.3505586592178771,
+      "grad_norm": 0.6193831771381659,
+      "learning_rate": 4.939362899366769e-06,
+      "loss": 0.4656,
+      "step": 753
+    },
+    {
+      "epoch": 0.351024208566108,
+      "grad_norm": 0.6752488180070185,
+      "learning_rate": 4.939081690740084e-06,
+      "loss": 0.4581,
+      "step": 754
+    },
+    {
+      "epoch": 0.3514897579143389,
+      "grad_norm": 0.6356504250906698,
+      "learning_rate": 4.938799839601281e-06,
+      "loss": 0.4621,
+      "step": 755
+    },
+    {
+      "epoch": 0.35195530726256985,
+      "grad_norm": 0.6720696691922603,
+      "learning_rate": 4.938517346024606e-06,
+      "loss": 0.4678,
+      "step": 756
+    },
+    {
+      "epoch": 0.35242085661080075,
+      "grad_norm": 0.6283989578899462,
+      "learning_rate": 4.938234210084475e-06,
+      "loss": 0.4606,
+      "step": 757
+    },
+    {
+      "epoch": 0.35288640595903165,
+      "grad_norm": 0.6457686975489778,
+      "learning_rate": 4.937950431855473e-06,
+      "loss": 0.4677,
+      "step": 758
+    },
+    {
+      "epoch": 0.35335195530726254,
+      "grad_norm": 0.6198459343491703,
+      "learning_rate": 4.937666011412353e-06,
+      "loss": 0.4619,
+      "step": 759
+    },
+    {
+      "epoch": 0.3538175046554935,
+      "grad_norm": 0.6023724585012881,
+      "learning_rate": 4.937380948830038e-06,
+      "loss": 0.4748,
+      "step": 760
+    },
+    {
+      "epoch": 0.3542830540037244,
+      "grad_norm": 0.6359267790475659,
+      "learning_rate": 4.937095244183622e-06,
+      "loss": 0.468,
+      "step": 761
+    },
+    {
+      "epoch": 0.3547486033519553,
+      "grad_norm": 0.6425819826639435,
+      "learning_rate": 4.936808897548365e-06,
+      "loss": 0.4752,
+      "step": 762
+    },
+    {
+      "epoch": 0.3552141527001862,
+      "grad_norm": 0.682873988398687,
+      "learning_rate": 4.936521908999698e-06,
+      "loss": 0.4763,
+      "step": 763
+    },
+    {
+      "epoch": 0.35567970204841715,
+      "grad_norm": 0.6756082839940482,
+      "learning_rate": 4.936234278613221e-06,
+      "loss": 0.4755,
+      "step": 764
+    },
+    {
+      "epoch": 0.35614525139664804,
+      "grad_norm": 0.5946847549477619,
+      "learning_rate": 4.935946006464702e-06,
+      "loss": 0.4578,
+      "step": 765
+    },
+    {
+      "epoch": 0.35661080074487894,
+      "grad_norm": 0.6257258245479131,
+      "learning_rate": 4.935657092630077e-06,
+      "loss": 0.4932,
+      "step": 766
+    },
+    {
+      "epoch": 0.3570763500931099,
+      "grad_norm": 0.6402958700217235,
+      "learning_rate": 4.935367537185456e-06,
+      "loss": 0.4902,
+      "step": 767
+    },
+    {
+      "epoch": 0.3575418994413408,
+      "grad_norm": 0.6508099202487825,
+      "learning_rate": 4.935077340207114e-06,
+      "loss": 0.4688,
+      "step": 768
+    },
+    {
+      "epoch": 0.3580074487895717,
+      "grad_norm": 0.6117521758930059,
+      "learning_rate": 4.934786501771494e-06,
+      "loss": 0.4592,
+      "step": 769
+    },
+    {
+      "epoch": 0.3584729981378026,
+      "grad_norm": 0.6683504869144088,
+      "learning_rate": 4.934495021955212e-06,
+      "loss": 0.465,
+      "step": 770
+    },
+    {
+      "epoch": 0.35893854748603354,
+      "grad_norm": 0.6243968025705827,
+      "learning_rate": 4.934202900835049e-06,
+      "loss": 0.4801,
+      "step": 771
+    },
+    {
+      "epoch": 0.35940409683426444,
+      "grad_norm": 0.6510232891935009,
+      "learning_rate": 4.9339101384879575e-06,
+      "loss": 0.4893,
+      "step": 772
+    },
+    {
+      "epoch": 0.35986964618249534,
+      "grad_norm": 0.6336895528155869,
+      "learning_rate": 4.933616734991058e-06,
+      "loss": 0.459,
+      "step": 773
+    },
+    {
+      "epoch": 0.36033519553072624,
+      "grad_norm": 0.6142248830867765,
+      "learning_rate": 4.93332269042164e-06,
+      "loss": 0.4791,
+      "step": 774
+    },
+    {
+      "epoch": 0.3608007448789572,
+      "grad_norm": 0.6215367993931932,
+      "learning_rate": 4.933028004857161e-06,
+      "loss": 0.4566,
+      "step": 775
+    },
+    {
+      "epoch": 0.3612662942271881,
+      "grad_norm": 0.5970046903120558,
+      "learning_rate": 4.9327326783752495e-06,
+      "loss": 0.4631,
+      "step": 776
+    },
+    {
+      "epoch": 0.361731843575419,
+      "grad_norm": 0.6854213101092897,
+      "learning_rate": 4.9324367110537005e-06,
+      "loss": 0.4814,
+      "step": 777
+    },
+    {
+      "epoch": 0.3621973929236499,
+      "grad_norm": 0.636608115502853,
+      "learning_rate": 4.932140102970479e-06,
+      "loss": 0.473,
+      "step": 778
+    },
+    {
+      "epoch": 0.36266294227188084,
+      "grad_norm": 0.6182774685030656,
+      "learning_rate": 4.93184285420372e-06,
+      "loss": 0.4497,
+      "step": 779
+    },
+    {
+      "epoch": 0.36312849162011174,
+      "grad_norm": 0.6050249689758551,
+      "learning_rate": 4.931544964831725e-06,
+      "loss": 0.4802,
+      "step": 780
+    },
+    {
+      "epoch": 0.36359404096834264,
+      "grad_norm": 0.6202075693518087,
+      "learning_rate": 4.931246434932964e-06,
+      "loss": 0.4762,
+      "step": 781
+    },
+    {
+      "epoch": 0.36405959031657353,
+      "grad_norm": 0.6601711018247509,
+      "learning_rate": 4.9309472645860765e-06,
+      "loss": 0.4811,
+      "step": 782
+    },
+    {
+      "epoch": 0.3645251396648045,
+      "grad_norm": 0.7011441478591383,
+      "learning_rate": 4.930647453869875e-06,
+      "loss": 0.4948,
+      "step": 783
+    },
+    {
+      "epoch": 0.3649906890130354,
+      "grad_norm": 0.6338582535779955,
+      "learning_rate": 4.930347002863332e-06,
+      "loss": 0.4675,
+      "step": 784
+    },
+    {
+      "epoch": 0.3654562383612663,
+      "grad_norm": 0.6013331988622372,
+      "learning_rate": 4.930045911645596e-06,
+      "loss": 0.4855,
+      "step": 785
+    },
+    {
+      "epoch": 0.3659217877094972,
+      "grad_norm": 0.6295836635361738,
+      "learning_rate": 4.9297441802959814e-06,
+      "loss": 0.4747,
+      "step": 786
+    },
+    {
+      "epoch": 0.36638733705772814,
+      "grad_norm": 0.7488483085895042,
+      "learning_rate": 4.929441808893971e-06,
+      "loss": 0.4672,
+      "step": 787
+    },
+    {
+      "epoch": 0.36685288640595903,
+      "grad_norm": 0.7053123652741422,
+      "learning_rate": 4.929138797519218e-06,
+      "loss": 0.479,
+      "step": 788
+    },
+    {
+      "epoch": 0.36731843575418993,
+      "grad_norm": 0.6336131140568002,
+      "learning_rate": 4.9288351462515395e-06,
+      "loss": 0.441,
+      "step": 789
+    },
+    {
+      "epoch": 0.36778398510242083,
+      "grad_norm": 0.6028238030619381,
+      "learning_rate": 4.928530855170928e-06,
+      "loss": 0.4581,
+      "step": 790
+    },
+    {
+      "epoch": 0.3682495344506518,
+      "grad_norm": 0.6706998296988638,
+      "learning_rate": 4.928225924357538e-06,
+      "loss": 0.4682,
+      "step": 791
+    },
+    {
+      "epoch": 0.3687150837988827,
+      "grad_norm": 0.6453500078345706,
+      "learning_rate": 4.927920353891699e-06,
+      "loss": 0.4452,
+      "step": 792
+    },
+    {
+      "epoch": 0.3691806331471136,
+      "grad_norm": 0.6875657057824942,
+      "learning_rate": 4.927614143853903e-06,
+      "loss": 0.4579,
+      "step": 793
+    },
+    {
+      "epoch": 0.36964618249534453,
+      "grad_norm": 0.6964829278628596,
+      "learning_rate": 4.927307294324812e-06,
+      "loss": 0.4837,
+      "step": 794
+    },
+    {
+      "epoch": 0.37011173184357543,
+      "grad_norm": 0.6303388253006122,
+      "learning_rate": 4.926999805385261e-06,
+      "loss": 0.4733,
+      "step": 795
+    },
+    {
+      "epoch": 0.37057728119180633,
+      "grad_norm": 0.687021800760328,
+      "learning_rate": 4.926691677116248e-06,
+      "loss": 0.4823,
+      "step": 796
+    },
+    {
+      "epoch": 0.3710428305400372,
+      "grad_norm": 0.624130557461175,
+      "learning_rate": 4.92638290959894e-06,
+      "loss": 0.4872,
+      "step": 797
+    },
+    {
+      "epoch": 0.3715083798882682,
+      "grad_norm": 0.6699404787548625,
+      "learning_rate": 4.926073502914675e-06,
+      "loss": 0.4777,
+      "step": 798
+    },
+    {
+      "epoch": 0.3719739292364991,
+      "grad_norm": 0.623712846961961,
+      "learning_rate": 4.925763457144959e-06,
+      "loss": 0.4501,
+      "step": 799
+    },
+    {
+      "epoch": 0.37243947858473,
+      "grad_norm": 0.6441637153921923,
+      "learning_rate": 4.925452772371464e-06,
+      "loss": 0.4808,
+      "step": 800
+    },
+    {
+      "epoch": 0.3729050279329609,
+      "grad_norm": 0.6394514644620384,
+      "learning_rate": 4.925141448676033e-06,
+      "loss": 0.48,
+      "step": 801
+    },
+    {
+      "epoch": 0.37337057728119183,
+      "grad_norm": 0.6176604317284,
+      "learning_rate": 4.924829486140674e-06,
+      "loss": 0.4782,
+      "step": 802
+    },
+    {
+      "epoch": 0.3738361266294227,
+      "grad_norm": 0.6533605365085523,
+      "learning_rate": 4.924516884847568e-06,
+      "loss": 0.4662,
+      "step": 803
+    },
+    {
+      "epoch": 0.3743016759776536,
+      "grad_norm": 0.6579876461604806,
+      "learning_rate": 4.9242036448790596e-06,
+      "loss": 0.4691,
+      "step": 804
+    },
+    {
+      "epoch": 0.3747672253258845,
+      "grad_norm": 0.6105474712247841,
+      "learning_rate": 4.923889766317664e-06,
+      "loss": 0.4722,
+      "step": 805
+    },
+    {
+      "epoch": 0.3752327746741155,
+      "grad_norm": 0.6531607502318059,
+      "learning_rate": 4.923575249246066e-06,
+      "loss": 0.4622,
+      "step": 806
+    },
+    {
+      "epoch": 0.3756983240223464,
+      "grad_norm": 0.6307103865885226,
+      "learning_rate": 4.9232600937471135e-06,
+      "loss": 0.4646,
+      "step": 807
+    },
+    {
+      "epoch": 0.3761638733705773,
+      "grad_norm": 0.640719236597497,
+      "learning_rate": 4.922944299903829e-06,
+      "loss": 0.4847,
+      "step": 808
+    },
+    {
+      "epoch": 0.37662942271880817,
+      "grad_norm": 0.6563038488523362,
+      "learning_rate": 4.922627867799399e-06,
+      "loss": 0.4911,
+      "step": 809
+    },
+    {
+      "epoch": 0.3770949720670391,
+      "grad_norm": 0.5999719902971525,
+      "learning_rate": 4.922310797517179e-06,
+      "loss": 0.477,
+      "step": 810
+    },
+    {
+      "epoch": 0.37756052141527,
+      "grad_norm": 0.6021424164537907,
+      "learning_rate": 4.921993089140693e-06,
+      "loss": 0.448,
+      "step": 811
+    },
+    {
+      "epoch": 0.3780260707635009,
+      "grad_norm": 0.6831159365894415,
+      "learning_rate": 4.921674742753634e-06,
+      "loss": 0.4566,
+      "step": 812
+    },
+    {
+      "epoch": 0.3784916201117318,
+      "grad_norm": 0.613722508199464,
+      "learning_rate": 4.921355758439861e-06,
+      "loss": 0.4802,
+      "step": 813
+    },
+    {
+      "epoch": 0.3789571694599628,
+      "grad_norm": 0.63509099057191,
+      "learning_rate": 4.921036136283402e-06,
+      "loss": 0.4726,
+      "step": 814
+    },
+    {
+      "epoch": 0.37942271880819367,
+      "grad_norm": 0.6260138389863249,
+      "learning_rate": 4.920715876368452e-06,
+      "loss": 0.4667,
+      "step": 815
+    },
+    {
+      "epoch": 0.37988826815642457,
+      "grad_norm": 0.5929916742753283,
+      "learning_rate": 4.920394978779377e-06,
+      "loss": 0.4613,
+      "step": 816
+    },
+    {
+      "epoch": 0.38035381750465547,
+      "grad_norm": 0.6455034431680491,
+      "learning_rate": 4.920073443600709e-06,
+      "loss": 0.4712,
+      "step": 817
+    },
+    {
+      "epoch": 0.3808193668528864,
+      "grad_norm": 0.6211553937997829,
+      "learning_rate": 4.9197512709171465e-06,
+      "loss": 0.4656,
+      "step": 818
+    },
+    {
+      "epoch": 0.3812849162011173,
+      "grad_norm": 0.7222523397111195,
+      "learning_rate": 4.919428460813559e-06,
+      "loss": 0.4915,
+      "step": 819
+    },
+    {
+      "epoch": 0.3817504655493482,
+      "grad_norm": 0.6565840326972144,
+      "learning_rate": 4.919105013374981e-06,
+      "loss": 0.4893,
+      "step": 820
+    },
+    {
+      "epoch": 0.38221601489757917,
+      "grad_norm": 0.6260345648734299,
+      "learning_rate": 4.918780928686617e-06,
+      "loss": 0.4574,
+      "step": 821
+    },
+    {
+      "epoch": 0.38268156424581007,
+      "grad_norm": 0.6433673963181513,
+      "learning_rate": 4.918456206833839e-06,
+      "loss": 0.4904,
+      "step": 822
+    },
+    {
+      "epoch": 0.38314711359404097,
+      "grad_norm": 0.6611382319134068,
+      "learning_rate": 4.918130847902186e-06,
+      "loss": 0.4584,
+      "step": 823
+    },
+    {
+      "epoch": 0.38361266294227186,
+      "grad_norm": 0.6389219473761231,
+      "learning_rate": 4.9178048519773656e-06,
+      "loss": 0.4822,
+      "step": 824
+    },
+    {
+      "epoch": 0.3840782122905028,
+      "grad_norm": 0.6093325406022321,
+      "learning_rate": 4.9174782191452515e-06,
+      "loss": 0.4577,
+      "step": 825
+    },
+    {
+      "epoch": 0.3845437616387337,
+      "grad_norm": 0.5961096948114254,
+      "learning_rate": 4.917150949491888e-06,
+      "loss": 0.4739,
+      "step": 826
+    },
+    {
+      "epoch": 0.3850093109869646,
+      "grad_norm": 0.606488348761175,
+      "learning_rate": 4.916823043103486e-06,
+      "loss": 0.4714,
+      "step": 827
+    },
+    {
+      "epoch": 0.3854748603351955,
+      "grad_norm": 0.587655258816685,
+      "learning_rate": 4.916494500066422e-06,
+      "loss": 0.4531,
+      "step": 828
+    },
+    {
+      "epoch": 0.38594040968342647,
+      "grad_norm": 0.6215997035557835,
+      "learning_rate": 4.9161653204672446e-06,
+      "loss": 0.4706,
+      "step": 829
+    },
+    {
+      "epoch": 0.38640595903165736,
+      "grad_norm": 0.6328979872702324,
+      "learning_rate": 4.9158355043926655e-06,
+      "loss": 0.471,
+      "step": 830
+    },
+    {
+      "epoch": 0.38687150837988826,
+      "grad_norm": 0.6455050218000562,
+      "learning_rate": 4.915505051929566e-06,
+      "loss": 0.4771,
+      "step": 831
+    },
+    {
+      "epoch": 0.38733705772811916,
+      "grad_norm": 0.622870845176216,
+      "learning_rate": 4.915173963164997e-06,
+      "loss": 0.4698,
+      "step": 832
+    },
+    {
+      "epoch": 0.3878026070763501,
+      "grad_norm": 0.6241258460848491,
+      "learning_rate": 4.914842238186173e-06,
+      "loss": 0.473,
+      "step": 833
+    },
+    {
+      "epoch": 0.388268156424581,
+      "grad_norm": 0.6302120194715802,
+      "learning_rate": 4.91450987708048e-06,
+      "loss": 0.4809,
+      "step": 834
+    },
+    {
+      "epoch": 0.3887337057728119,
+      "grad_norm": 0.6657727994045914,
+      "learning_rate": 4.914176879935468e-06,
+      "loss": 0.4705,
+      "step": 835
+    },
+    {
+      "epoch": 0.3891992551210428,
+      "grad_norm": 0.6632602815857767,
+      "learning_rate": 4.913843246838859e-06,
+      "loss": 0.5011,
+      "step": 836
+    },
+    {
+      "epoch": 0.38966480446927376,
+      "grad_norm": 0.6185437343701294,
+      "learning_rate": 4.913508977878537e-06,
+      "loss": 0.4674,
+      "step": 837
+    },
+    {
+      "epoch": 0.39013035381750466,
+      "grad_norm": 0.6367920582295674,
+      "learning_rate": 4.913174073142558e-06,
+      "loss": 0.4612,
+      "step": 838
+    },
+    {
+      "epoch": 0.39059590316573556,
+      "grad_norm": 0.66341297216136,
+      "learning_rate": 4.912838532719144e-06,
+      "loss": 0.451,
+      "step": 839
+    },
+    {
+      "epoch": 0.39106145251396646,
+      "grad_norm": 0.638815227198975,
+      "learning_rate": 4.912502356696684e-06,
+      "loss": 0.4628,
+      "step": 840
+    },
+    {
+      "epoch": 0.3915270018621974,
+      "grad_norm": 0.623634754010245,
+      "learning_rate": 4.912165545163735e-06,
+      "loss": 0.4781,
+      "step": 841
+    },
+    {
+      "epoch": 0.3919925512104283,
+      "grad_norm": 0.6042303677839573,
+      "learning_rate": 4.911828098209021e-06,
+      "loss": 0.4521,
+      "step": 842
+    },
+    {
+      "epoch": 0.3924581005586592,
+      "grad_norm": 0.6074560531500909,
+      "learning_rate": 4.9114900159214314e-06,
+      "loss": 0.4753,
+      "step": 843
+    },
+    {
+      "epoch": 0.3929236499068901,
+      "grad_norm": 0.6857332671495816,
+      "learning_rate": 4.911151298390029e-06,
+      "loss": 0.4775,
+      "step": 844
+    },
+    {
+      "epoch": 0.39338919925512106,
+      "grad_norm": 0.6539579769998742,
+      "learning_rate": 4.910811945704038e-06,
+      "loss": 0.4475,
+      "step": 845
+    },
+    {
+      "epoch": 0.39385474860335196,
+      "grad_norm": 0.6857315863781245,
+      "learning_rate": 4.910471957952852e-06,
+      "loss": 0.4741,
+      "step": 846
+    },
+    {
+      "epoch": 0.39432029795158285,
+      "grad_norm": 0.6630708218348811,
+      "learning_rate": 4.910131335226031e-06,
+      "loss": 0.4755,
+      "step": 847
+    },
+    {
+      "epoch": 0.3947858472998138,
+      "grad_norm": 0.621853743858161,
+      "learning_rate": 4.9097900776133055e-06,
+      "loss": 0.4584,
+      "step": 848
+    },
+    {
+      "epoch": 0.3952513966480447,
+      "grad_norm": 0.6399763025824953,
+      "learning_rate": 4.909448185204569e-06,
+      "loss": 0.4609,
+      "step": 849
+    },
+    {
+      "epoch": 0.3957169459962756,
+      "grad_norm": 0.6732785183938688,
+      "learning_rate": 4.909105658089885e-06,
+      "loss": 0.4755,
+      "step": 850
+    },
+    {
+      "epoch": 0.3961824953445065,
+      "grad_norm": 0.6704497149706169,
+      "learning_rate": 4.908762496359482e-06,
+      "loss": 0.4515,
+      "step": 851
+    },
+    {
+      "epoch": 0.39664804469273746,
+      "grad_norm": 0.5882343141858953,
+      "learning_rate": 4.908418700103759e-06,
+      "loss": 0.4589,
+      "step": 852
+    },
+    {
+      "epoch": 0.39711359404096835,
+      "grad_norm": 0.6762046886295933,
+      "learning_rate": 4.908074269413278e-06,
+      "loss": 0.4788,
+      "step": 853
+    },
+    {
+      "epoch": 0.39757914338919925,
+      "grad_norm": 0.6366475726264113,
+      "learning_rate": 4.907729204378772e-06,
+      "loss": 0.4628,
+      "step": 854
+    },
+    {
+      "epoch": 0.39804469273743015,
+      "grad_norm": 0.6601570677256395,
+      "learning_rate": 4.907383505091137e-06,
+      "loss": 0.4692,
+      "step": 855
+    },
+    {
+      "epoch": 0.3985102420856611,
+      "grad_norm": 0.6491321905307778,
+      "learning_rate": 4.907037171641441e-06,
+      "loss": 0.4662,
+      "step": 856
+    },
+    {
+      "epoch": 0.398975791433892,
+      "grad_norm": 0.67863105954598,
+      "learning_rate": 4.906690204120915e-06,
+      "loss": 0.4861,
+      "step": 857
+    },
+    {
+      "epoch": 0.3994413407821229,
+      "grad_norm": 0.6147308899870986,
+      "learning_rate": 4.90634260262096e-06,
+      "loss": 0.4666,
+      "step": 858
+    },
+    {
+      "epoch": 0.3999068901303538,
+      "grad_norm": 0.6273296401373223,
+      "learning_rate": 4.9059943672331404e-06,
+      "loss": 0.4593,
+      "step": 859
+    },
+    {
+      "epoch": 0.40037243947858475,
+      "grad_norm": 0.6469429233663354,
+      "learning_rate": 4.905645498049191e-06,
+      "loss": 0.4534,
+      "step": 860
+    },
+    {
+      "epoch": 0.40083798882681565,
+      "grad_norm": 0.6479578316328013,
+      "learning_rate": 4.905295995161011e-06,
+      "loss": 0.4608,
+      "step": 861
+    },
+    {
+      "epoch": 0.40130353817504655,
+      "grad_norm": 0.7029117757994059,
+      "learning_rate": 4.90494585866067e-06,
+      "loss": 0.4824,
+      "step": 862
+    },
+    {
+      "epoch": 0.40176908752327745,
+      "grad_norm": 0.6294910687391004,
+      "learning_rate": 4.9045950886404005e-06,
+      "loss": 0.4879,
+      "step": 863
+    },
+    {
+      "epoch": 0.4022346368715084,
+      "grad_norm": 0.6939372259253429,
+      "learning_rate": 4.904243685192604e-06,
+      "loss": 0.4535,
+      "step": 864
+    },
+    {
+      "epoch": 0.4027001862197393,
+      "grad_norm": 0.6565101457694527,
+      "learning_rate": 4.903891648409849e-06,
+      "loss": 0.4697,
+      "step": 865
+    },
+    {
+      "epoch": 0.4031657355679702,
+      "grad_norm": 0.6775202537280691,
+      "learning_rate": 4.90353897838487e-06,
+      "loss": 0.4643,
+      "step": 866
+    },
+    {
+      "epoch": 0.4036312849162011,
+      "grad_norm": 0.6570190880915651,
+      "learning_rate": 4.903185675210568e-06,
+      "loss": 0.4547,
+      "step": 867
+    },
+    {
+      "epoch": 0.40409683426443205,
+      "grad_norm": 0.6714681324184795,
+      "learning_rate": 4.9028317389800125e-06,
+      "loss": 0.4753,
+      "step": 868
+    },
+    {
+      "epoch": 0.40456238361266295,
+      "grad_norm": 0.704031990507195,
+      "learning_rate": 4.9024771697864385e-06,
+      "loss": 0.4996,
+      "step": 869
+    },
+    {
+      "epoch": 0.40502793296089384,
+      "grad_norm": 0.6571866708571896,
+      "learning_rate": 4.902121967723248e-06,
+      "loss": 0.4685,
+      "step": 870
+    },
+    {
+      "epoch": 0.40549348230912474,
+      "grad_norm": 0.6618671080238566,
+      "learning_rate": 4.901766132884009e-06,
+      "loss": 0.4753,
+      "step": 871
+    },
+    {
+      "epoch": 0.4059590316573557,
+      "grad_norm": 0.6993405942298184,
+      "learning_rate": 4.901409665362457e-06,
+      "loss": 0.4682,
+      "step": 872
+    },
+    {
+      "epoch": 0.4064245810055866,
+      "grad_norm": 0.6335788946603675,
+      "learning_rate": 4.901052565252496e-06,
+      "loss": 0.4835,
+      "step": 873
+    },
+    {
+      "epoch": 0.4068901303538175,
+      "grad_norm": 0.5994395685067679,
+      "learning_rate": 4.900694832648191e-06,
+      "loss": 0.4827,
+      "step": 874
+    },
+    {
+      "epoch": 0.4073556797020484,
+      "grad_norm": 0.6556302256320401,
+      "learning_rate": 4.90033646764378e-06,
+      "loss": 0.4655,
+      "step": 875
+    },
+    {
+      "epoch": 0.40782122905027934,
+      "grad_norm": 0.6323069611134717,
+      "learning_rate": 4.899977470333665e-06,
+      "loss": 0.4789,
+      "step": 876
+    },
+    {
+      "epoch": 0.40828677839851024,
+      "grad_norm": 0.6308988427345963,
+      "learning_rate": 4.899617840812413e-06,
+      "loss": 0.4665,
+      "step": 877
+    },
+    {
+      "epoch": 0.40875232774674114,
+      "grad_norm": 0.6177318779993888,
+      "learning_rate": 4.89925757917476e-06,
+      "loss": 0.4633,
+      "step": 878
+    },
+    {
+      "epoch": 0.4092178770949721,
+      "grad_norm": 0.6286224298080071,
+      "learning_rate": 4.898896685515606e-06,
+      "loss": 0.4722,
+      "step": 879
+    },
+    {
+      "epoch": 0.409683426443203,
+      "grad_norm": 0.607869906647734,
+      "learning_rate": 4.898535159930021e-06,
+      "loss": 0.4517,
+      "step": 880
+    },
+    {
+      "epoch": 0.4101489757914339,
+      "grad_norm": 0.6435283809424427,
+      "learning_rate": 4.898173002513238e-06,
+      "loss": 0.4551,
+      "step": 881
+    },
+    {
+      "epoch": 0.4106145251396648,
+      "grad_norm": 0.5918363455500879,
+      "learning_rate": 4.897810213360659e-06,
+      "loss": 0.4361,
+      "step": 882
+    },
+    {
+      "epoch": 0.41108007448789574,
+      "grad_norm": 0.5802650345128577,
+      "learning_rate": 4.8974467925678494e-06,
+      "loss": 0.4573,
+      "step": 883
+    },
+    {
+      "epoch": 0.41154562383612664,
+      "grad_norm": 0.6269729979693888,
+      "learning_rate": 4.897082740230545e-06,
+      "loss": 0.4936,
+      "step": 884
+    },
+    {
+      "epoch": 0.41201117318435754,
+      "grad_norm": 0.5926155922243164,
+      "learning_rate": 4.896718056444645e-06,
+      "loss": 0.4663,
+      "step": 885
+    },
+    {
+      "epoch": 0.41247672253258844,
+      "grad_norm": 0.6409816604957148,
+      "learning_rate": 4.896352741306215e-06,
+      "loss": 0.4665,
+      "step": 886
+    },
+    {
+      "epoch": 0.4129422718808194,
+      "grad_norm": 0.6128658919782248,
+      "learning_rate": 4.895986794911489e-06,
+      "loss": 0.4656,
+      "step": 887
+    },
+    {
+      "epoch": 0.4134078212290503,
+      "grad_norm": 0.617797953791939,
+      "learning_rate": 4.895620217356865e-06,
+      "loss": 0.4563,
+      "step": 888
+    },
+    {
+      "epoch": 0.4138733705772812,
+      "grad_norm": 0.6121806380520051,
+      "learning_rate": 4.895253008738909e-06,
+      "loss": 0.4794,
+      "step": 889
+    },
+    {
+      "epoch": 0.4143389199255121,
+      "grad_norm": 0.6619383646662328,
+      "learning_rate": 4.8948851691543525e-06,
+      "loss": 0.4564,
+      "step": 890
+    },
+    {
+      "epoch": 0.41480446927374304,
+      "grad_norm": 0.590063127796581,
+      "learning_rate": 4.894516698700093e-06,
+      "loss": 0.4793,
+      "step": 891
+    },
+    {
+      "epoch": 0.41527001862197394,
+      "grad_norm": 0.6553795415221206,
+      "learning_rate": 4.894147597473193e-06,
+      "loss": 0.4717,
+      "step": 892
+    },
+    {
+      "epoch": 0.41573556797020483,
+      "grad_norm": 0.6080761268948223,
+      "learning_rate": 4.893777865570884e-06,
+      "loss": 0.4593,
+      "step": 893
+    },
+    {
+      "epoch": 0.41620111731843573,
+      "grad_norm": 0.6368887371104103,
+      "learning_rate": 4.8934075030905625e-06,
+      "loss": 0.4575,
+      "step": 894
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 0.6178683964518161,
+      "learning_rate": 4.893036510129791e-06,
+      "loss": 0.4456,
+      "step": 895
+    },
+    {
+      "epoch": 0.4171322160148976,
+      "grad_norm": 0.6358788556303248,
+      "learning_rate": 4.892664886786295e-06,
+      "loss": 0.4589,
+      "step": 896
+    },
+    {
+      "epoch": 0.4175977653631285,
+      "grad_norm": 0.6220119748005581,
+      "learning_rate": 4.892292633157972e-06,
+      "loss": 0.4517,
+      "step": 897
+    },
+    {
+      "epoch": 0.4180633147113594,
+      "grad_norm": 0.6409082584082681,
+      "learning_rate": 4.8919197493428816e-06,
+      "loss": 0.4928,
+      "step": 898
+    },
+    {
+      "epoch": 0.41852886405959033,
+      "grad_norm": 0.647607663595771,
+      "learning_rate": 4.89154623543925e-06,
+      "loss": 0.4542,
+      "step": 899
+    },
+    {
+      "epoch": 0.41899441340782123,
+      "grad_norm": 0.6755132551975458,
+      "learning_rate": 4.891172091545469e-06,
+      "loss": 0.4656,
+      "step": 900
+    },
+    {
+      "epoch": 0.41945996275605213,
+      "grad_norm": 0.6447392643737889,
+      "learning_rate": 4.8907973177600996e-06,
+      "loss": 0.4563,
+      "step": 901
+    },
+    {
+      "epoch": 0.419925512104283,
+      "grad_norm": 0.6465046840631848,
+      "learning_rate": 4.890421914181863e-06,
+      "loss": 0.4458,
+      "step": 902
+    },
+    {
+      "epoch": 0.420391061452514,
+      "grad_norm": 0.5700815426526978,
+      "learning_rate": 4.890045880909651e-06,
+      "loss": 0.4686,
+      "step": 903
+    },
+    {
+      "epoch": 0.4208566108007449,
+      "grad_norm": 0.6149587473299961,
+      "learning_rate": 4.889669218042519e-06,
+      "loss": 0.4601,
+      "step": 904
+    },
+    {
+      "epoch": 0.4213221601489758,
+      "grad_norm": 0.6265990150567091,
+      "learning_rate": 4.88929192567969e-06,
+      "loss": 0.4732,
+      "step": 905
+    },
+    {
+      "epoch": 0.42178770949720673,
+      "grad_norm": 0.6316732760978735,
+      "learning_rate": 4.88891400392055e-06,
+      "loss": 0.4641,
+      "step": 906
+    },
+    {
+      "epoch": 0.42225325884543763,
+      "grad_norm": 0.6316104874922882,
+      "learning_rate": 4.888535452864656e-06,
+      "loss": 0.4497,
+      "step": 907
+    },
+    {
+      "epoch": 0.4227188081936685,
+      "grad_norm": 0.6309068529313834,
+      "learning_rate": 4.888156272611723e-06,
+      "loss": 0.466,
+      "step": 908
+    },
+    {
+      "epoch": 0.4231843575418994,
+      "grad_norm": 0.6080201398535497,
+      "learning_rate": 4.887776463261639e-06,
+      "loss": 0.4798,
+      "step": 909
+    },
+    {
+      "epoch": 0.4236499068901304,
+      "grad_norm": 0.6289175990012926,
+      "learning_rate": 4.887396024914453e-06,
+      "loss": 0.4623,
+      "step": 910
+    },
+    {
+      "epoch": 0.4241154562383613,
+      "grad_norm": 0.618617734152923,
+      "learning_rate": 4.887014957670384e-06,
+      "loss": 0.4757,
+      "step": 911
+    },
+    {
+      "epoch": 0.4245810055865922,
+      "grad_norm": 0.592765099679565,
+      "learning_rate": 4.886633261629812e-06,
+      "loss": 0.4705,
+      "step": 912
+    },
+    {
+      "epoch": 0.4250465549348231,
+      "grad_norm": 0.6265603026088671,
+      "learning_rate": 4.886250936893286e-06,
+      "loss": 0.4636,
+      "step": 913
+    },
+    {
+      "epoch": 0.425512104283054,
+      "grad_norm": 0.6564729256584824,
+      "learning_rate": 4.885867983561519e-06,
+      "loss": 0.471,
+      "step": 914
+    },
+    {
+      "epoch": 0.4259776536312849,
+      "grad_norm": 0.6174575483910285,
+      "learning_rate": 4.885484401735389e-06,
+      "loss": 0.4719,
+      "step": 915
+    },
+    {
+      "epoch": 0.4264432029795158,
+      "grad_norm": 0.5840560710607847,
+      "learning_rate": 4.885100191515942e-06,
+      "loss": 0.4558,
+      "step": 916
+    },
+    {
+      "epoch": 0.4269087523277467,
+      "grad_norm": 0.6159138702448153,
+      "learning_rate": 4.884715353004387e-06,
+      "loss": 0.4652,
+      "step": 917
+    },
+    {
+      "epoch": 0.4273743016759777,
+      "grad_norm": 0.6385761426849687,
+      "learning_rate": 4.884329886302101e-06,
+      "loss": 0.4811,
+      "step": 918
+    },
+    {
+      "epoch": 0.4278398510242086,
+      "grad_norm": 0.6285406086324946,
+      "learning_rate": 4.883943791510625e-06,
+      "loss": 0.4598,
+      "step": 919
+    },
+    {
+      "epoch": 0.42830540037243947,
+      "grad_norm": 0.6145526288963175,
+      "learning_rate": 4.8835570687316635e-06,
+      "loss": 0.4491,
+      "step": 920
+    },
+    {
+      "epoch": 0.42877094972067037,
+      "grad_norm": 0.629894588319512,
+      "learning_rate": 4.88316971806709e-06,
+      "loss": 0.4806,
+      "step": 921
+    },
+    {
+      "epoch": 0.4292364990689013,
+      "grad_norm": 0.6331970878978456,
+      "learning_rate": 4.882781739618942e-06,
+      "loss": 0.4549,
+      "step": 922
+    },
+    {
+      "epoch": 0.4297020484171322,
+      "grad_norm": 0.6878080727251829,
+      "learning_rate": 4.882393133489423e-06,
+      "loss": 0.474,
+      "step": 923
+    },
+    {
+      "epoch": 0.4301675977653631,
+      "grad_norm": 0.6247418883339388,
+      "learning_rate": 4.882003899780898e-06,
+      "loss": 0.4826,
+      "step": 924
+    },
+    {
+      "epoch": 0.430633147113594,
+      "grad_norm": 0.6204393338574391,
+      "learning_rate": 4.881614038595904e-06,
+      "loss": 0.4579,
+      "step": 925
+    },
+    {
+      "epoch": 0.43109869646182497,
+      "grad_norm": 0.6055730380996902,
+      "learning_rate": 4.881223550037137e-06,
+      "loss": 0.4479,
+      "step": 926
+    },
+    {
+      "epoch": 0.43156424581005587,
+      "grad_norm": 0.5981630965151316,
+      "learning_rate": 4.880832434207463e-06,
+      "loss": 0.4455,
+      "step": 927
+    },
+    {
+      "epoch": 0.43202979515828677,
+      "grad_norm": 0.6643021917449464,
+      "learning_rate": 4.880440691209911e-06,
+      "loss": 0.4832,
+      "step": 928
+    },
+    {
+      "epoch": 0.43249534450651766,
+      "grad_norm": 0.6607592307047513,
+      "learning_rate": 4.880048321147674e-06,
+      "loss": 0.4666,
+      "step": 929
+    },
+    {
+      "epoch": 0.4329608938547486,
+      "grad_norm": 0.6859365218937872,
+      "learning_rate": 4.8796553241241125e-06,
+      "loss": 0.4558,
+      "step": 930
+    },
+    {
+      "epoch": 0.4334264432029795,
+      "grad_norm": 0.6106939367865185,
+      "learning_rate": 4.879261700242751e-06,
+      "loss": 0.479,
+      "step": 931
+    },
+    {
+      "epoch": 0.4338919925512104,
+      "grad_norm": 0.6467244461565984,
+      "learning_rate": 4.87886744960728e-06,
+      "loss": 0.4737,
+      "step": 932
+    },
+    {
+      "epoch": 0.43435754189944137,
+      "grad_norm": 0.6462758959321117,
+      "learning_rate": 4.878472572321554e-06,
+      "loss": 0.4985,
+      "step": 933
+    },
+    {
+      "epoch": 0.43482309124767227,
+      "grad_norm": 0.6428829599903008,
+      "learning_rate": 4.878077068489593e-06,
+      "loss": 0.4715,
+      "step": 934
+    },
+    {
+      "epoch": 0.43528864059590316,
+      "grad_norm": 0.7017918123636155,
+      "learning_rate": 4.877680938215582e-06,
+      "loss": 0.4775,
+      "step": 935
+    },
+    {
+      "epoch": 0.43575418994413406,
+      "grad_norm": 0.6595246758208536,
+      "learning_rate": 4.877284181603872e-06,
+      "loss": 0.4655,
+      "step": 936
+    },
+    {
+      "epoch": 0.436219739292365,
+      "grad_norm": 0.7094107452136772,
+      "learning_rate": 4.876886798758977e-06,
+      "loss": 0.481,
+      "step": 937
+    },
+    {
+      "epoch": 0.4366852886405959,
+      "grad_norm": 0.6475295392093389,
+      "learning_rate": 4.876488789785577e-06,
+      "loss": 0.4649,
+      "step": 938
+    },
+    {
+      "epoch": 0.4371508379888268,
+      "grad_norm": 0.6385349206364859,
+      "learning_rate": 4.8760901547885185e-06,
+      "loss": 0.4861,
+      "step": 939
+    },
+    {
+      "epoch": 0.4376163873370577,
+      "grad_norm": 0.6823365739495671,
+      "learning_rate": 4.8756908938728095e-06,
+      "loss": 0.4761,
+      "step": 940
+    },
+    {
+      "epoch": 0.43808193668528866,
+      "grad_norm": 0.6526004601417323,
+      "learning_rate": 4.875291007143628e-06,
+      "loss": 0.4877,
+      "step": 941
+    },
+    {
+      "epoch": 0.43854748603351956,
+      "grad_norm": 0.6228142658847875,
+      "learning_rate": 4.874890494706309e-06,
+      "loss": 0.4637,
+      "step": 942
+    },
+    {
+      "epoch": 0.43901303538175046,
+      "grad_norm": 0.6638612734868641,
+      "learning_rate": 4.874489356666361e-06,
+      "loss": 0.4687,
+      "step": 943
+    },
+    {
+      "epoch": 0.43947858472998136,
+      "grad_norm": 0.6666769067227107,
+      "learning_rate": 4.874087593129452e-06,
+      "loss": 0.4568,
+      "step": 944
+    },
+    {
+      "epoch": 0.4399441340782123,
+      "grad_norm": 0.647207177584227,
+      "learning_rate": 4.873685204201415e-06,
+      "loss": 0.46,
+      "step": 945
+    },
+    {
+      "epoch": 0.4404096834264432,
+      "grad_norm": 0.6038950777434363,
+      "learning_rate": 4.8732821899882495e-06,
+      "loss": 0.4518,
+      "step": 946
+    },
+    {
+      "epoch": 0.4408752327746741,
+      "grad_norm": 0.6338889126090352,
+      "learning_rate": 4.8728785505961205e-06,
+      "loss": 0.4664,
+      "step": 947
+    },
+    {
+      "epoch": 0.441340782122905,
+      "grad_norm": 0.6526424968862293,
+      "learning_rate": 4.872474286131355e-06,
+      "loss": 0.4613,
+      "step": 948
+    },
+    {
+      "epoch": 0.44180633147113596,
+      "grad_norm": 0.6437636690745716,
+      "learning_rate": 4.872069396700445e-06,
+      "loss": 0.4762,
+      "step": 949
+    },
+    {
+      "epoch": 0.44227188081936686,
+      "grad_norm": 0.5960113602842982,
+      "learning_rate": 4.871663882410049e-06,
+      "loss": 0.4495,
+      "step": 950
+    },
+    {
+      "epoch": 0.44273743016759776,
+      "grad_norm": 0.6294924685582955,
+      "learning_rate": 4.871257743366988e-06,
+      "loss": 0.4605,
+      "step": 951
+    },
+    {
+      "epoch": 0.44320297951582865,
+      "grad_norm": 0.6274421448065481,
+      "learning_rate": 4.870850979678251e-06,
+      "loss": 0.4665,
+      "step": 952
+    },
+    {
+      "epoch": 0.4436685288640596,
+      "grad_norm": 0.6606383655522561,
+      "learning_rate": 4.870443591450988e-06,
+      "loss": 0.4469,
+      "step": 953
+    },
+    {
+      "epoch": 0.4441340782122905,
+      "grad_norm": 0.6129960933420363,
+      "learning_rate": 4.870035578792514e-06,
+      "loss": 0.4755,
+      "step": 954
+    },
+    {
+      "epoch": 0.4445996275605214,
+      "grad_norm": 0.6369724460083941,
+      "learning_rate": 4.86962694181031e-06,
+      "loss": 0.4486,
+      "step": 955
+    },
+    {
+      "epoch": 0.4450651769087523,
+      "grad_norm": 0.6250413258772751,
+      "learning_rate": 4.869217680612021e-06,
+      "loss": 0.473,
+      "step": 956
+    },
+    {
+      "epoch": 0.44553072625698326,
+      "grad_norm": 0.591782125259715,
+      "learning_rate": 4.868807795305455e-06,
+      "loss": 0.4616,
+      "step": 957
+    },
+    {
+      "epoch": 0.44599627560521415,
+      "grad_norm": 0.6678636022426143,
+      "learning_rate": 4.868397285998586e-06,
+      "loss": 0.4609,
+      "step": 958
+    },
+    {
+      "epoch": 0.44646182495344505,
+      "grad_norm": 0.6155899866237423,
+      "learning_rate": 4.867986152799552e-06,
+      "loss": 0.456,
+      "step": 959
+    },
+    {
+      "epoch": 0.44692737430167595,
+      "grad_norm": 0.63690917928669,
+      "learning_rate": 4.867574395816656e-06,
+      "loss": 0.4609,
+      "step": 960
+    },
+    {
+      "epoch": 0.4473929236499069,
+      "grad_norm": 0.6407777062475025,
+      "learning_rate": 4.8671620151583635e-06,
+      "loss": 0.4822,
+      "step": 961
+    },
+    {
+      "epoch": 0.4478584729981378,
+      "grad_norm": 0.7012642709110991,
+      "learning_rate": 4.866749010933306e-06,
+      "loss": 0.4457,
+      "step": 962
+    },
+    {
+      "epoch": 0.4483240223463687,
+      "grad_norm": 0.6081723882436926,
+      "learning_rate": 4.866335383250279e-06,
+      "loss": 0.4696,
+      "step": 963
+    },
+    {
+      "epoch": 0.44878957169459965,
+      "grad_norm": 0.6706177130531238,
+      "learning_rate": 4.865921132218241e-06,
+      "loss": 0.4631,
+      "step": 964
+    },
+    {
+      "epoch": 0.44925512104283055,
+      "grad_norm": 0.6288030846361611,
+      "learning_rate": 4.865506257946316e-06,
+      "loss": 0.477,
+      "step": 965
+    },
+    {
+      "epoch": 0.44972067039106145,
+      "grad_norm": 0.65743201374991,
+      "learning_rate": 4.8650907605437915e-06,
+      "loss": 0.4575,
+      "step": 966
+    },
+    {
+      "epoch": 0.45018621973929235,
+      "grad_norm": 0.6112264025753583,
+      "learning_rate": 4.864674640120119e-06,
+      "loss": 0.4721,
+      "step": 967
+    },
+    {
+      "epoch": 0.4506517690875233,
+      "grad_norm": 0.6271638685553242,
+      "learning_rate": 4.864257896784916e-06,
+      "loss": 0.4704,
+      "step": 968
+    },
+    {
+      "epoch": 0.4511173184357542,
+      "grad_norm": 0.6108988783880878,
+      "learning_rate": 4.863840530647961e-06,
+      "loss": 0.4489,
+      "step": 969
+    },
+    {
+      "epoch": 0.4515828677839851,
+      "grad_norm": 0.6236208145969833,
+      "learning_rate": 4.8634225418192e-06,
+      "loss": 0.4953,
+      "step": 970
+    },
+    {
+      "epoch": 0.452048417132216,
+      "grad_norm": 0.6218031523219798,
+      "learning_rate": 4.863003930408739e-06,
+      "loss": 0.4696,
+      "step": 971
+    },
+    {
+      "epoch": 0.45251396648044695,
+      "grad_norm": 0.6472521020475792,
+      "learning_rate": 4.8625846965268505e-06,
+      "loss": 0.4753,
+      "step": 972
+    },
+    {
+      "epoch": 0.45297951582867785,
+      "grad_norm": 0.6371397178081865,
+      "learning_rate": 4.862164840283973e-06,
+      "loss": 0.4681,
+      "step": 973
+    },
+    {
+      "epoch": 0.45344506517690875,
+      "grad_norm": 0.6900324026263875,
+      "learning_rate": 4.861744361790704e-06,
+      "loss": 0.4738,
+      "step": 974
+    },
+    {
+      "epoch": 0.45391061452513964,
+      "grad_norm": 0.6233439105851637,
+      "learning_rate": 4.861323261157809e-06,
+      "loss": 0.4741,
+      "step": 975
+    },
+    {
+      "epoch": 0.4543761638733706,
+      "grad_norm": 0.6098810467134975,
+      "learning_rate": 4.860901538496215e-06,
+      "loss": 0.4615,
+      "step": 976
+    },
+    {
+      "epoch": 0.4548417132216015,
+      "grad_norm": 0.6440815817021965,
+      "learning_rate": 4.8604791939170145e-06,
+      "loss": 0.4569,
+      "step": 977
+    },
+    {
+      "epoch": 0.4553072625698324,
+      "grad_norm": 0.6195155757154539,
+      "learning_rate": 4.860056227531463e-06,
+      "loss": 0.4669,
+      "step": 978
+    },
+    {
+      "epoch": 0.4557728119180633,
+      "grad_norm": 0.6337129386407777,
+      "learning_rate": 4.859632639450979e-06,
+      "loss": 0.4646,
+      "step": 979
+    },
+    {
+      "epoch": 0.45623836126629425,
+      "grad_norm": 0.6310116338854407,
+      "learning_rate": 4.859208429787148e-06,
+      "loss": 0.4582,
+      "step": 980
+    },
+    {
+      "epoch": 0.45670391061452514,
+      "grad_norm": 0.6958230539868767,
+      "learning_rate": 4.858783598651715e-06,
+      "loss": 0.4525,
+      "step": 981
+    },
+    {
+      "epoch": 0.45716945996275604,
+      "grad_norm": 0.6489165872540826,
+      "learning_rate": 4.85835814615659e-06,
+      "loss": 0.467,
+      "step": 982
+    },
+    {
+      "epoch": 0.45763500931098694,
+      "grad_norm": 0.6378545552012003,
+      "learning_rate": 4.857932072413849e-06,
+      "loss": 0.4504,
+      "step": 983
+    },
+    {
+      "epoch": 0.4581005586592179,
+      "grad_norm": 0.6240125261851802,
+      "learning_rate": 4.857505377535729e-06,
+      "loss": 0.4829,
+      "step": 984
+    },
+    {
+      "epoch": 0.4585661080074488,
+      "grad_norm": 0.684144951037981,
+      "learning_rate": 4.857078061634632e-06,
+      "loss": 0.4887,
+      "step": 985
+    },
+    {
+      "epoch": 0.4590316573556797,
+      "grad_norm": 0.6178412383240028,
+      "learning_rate": 4.856650124823122e-06,
+      "loss": 0.4582,
+      "step": 986
+    },
+    {
+      "epoch": 0.4594972067039106,
+      "grad_norm": 0.6662532254742324,
+      "learning_rate": 4.85622156721393e-06,
+      "loss": 0.4717,
+      "step": 987
+    },
+    {
+      "epoch": 0.45996275605214154,
+      "grad_norm": 0.6713954425058134,
+      "learning_rate": 4.8557923889199465e-06,
+      "loss": 0.4744,
+      "step": 988
+    },
+    {
+      "epoch": 0.46042830540037244,
+      "grad_norm": 0.6354599719702383,
+      "learning_rate": 4.855362590054228e-06,
+      "loss": 0.4758,
+      "step": 989
+    },
+    {
+      "epoch": 0.46089385474860334,
+      "grad_norm": 0.6761550272536654,
+      "learning_rate": 4.854932170729994e-06,
+      "loss": 0.4698,
+      "step": 990
+    },
+    {
+      "epoch": 0.4613594040968343,
+      "grad_norm": 0.6375630191128813,
+      "learning_rate": 4.854501131060625e-06,
+      "loss": 0.4654,
+      "step": 991
+    },
+    {
+      "epoch": 0.4618249534450652,
+      "grad_norm": 0.6411253860995384,
+      "learning_rate": 4.8540694711596705e-06,
+      "loss": 0.4631,
+      "step": 992
+    },
+    {
+      "epoch": 0.4622905027932961,
+      "grad_norm": 0.6286122279882985,
+      "learning_rate": 4.853637191140837e-06,
+      "loss": 0.4611,
+      "step": 993
+    },
+    {
+      "epoch": 0.462756052141527,
+      "grad_norm": 0.6043872602336748,
+      "learning_rate": 4.8532042911179984e-06,
+      "loss": 0.4556,
+      "step": 994
+    },
+    {
+      "epoch": 0.46322160148975794,
+      "grad_norm": 0.6635622721884372,
+      "learning_rate": 4.852770771205192e-06,
+      "loss": 0.4821,
+      "step": 995
+    },
+    {
+      "epoch": 0.46368715083798884,
+      "grad_norm": 0.6601542385867899,
+      "learning_rate": 4.852336631516616e-06,
+      "loss": 0.4663,
+      "step": 996
+    },
+    {
+      "epoch": 0.46415270018621974,
+      "grad_norm": 0.6396928760167109,
+      "learning_rate": 4.851901872166634e-06,
+      "loss": 0.4676,
+      "step": 997
+    },
+    {
+      "epoch": 0.46461824953445063,
+      "grad_norm": 0.6325407742854606,
+      "learning_rate": 4.851466493269771e-06,
+      "loss": 0.473,
+      "step": 998
+    },
+    {
+      "epoch": 0.4650837988826816,
+      "grad_norm": 0.6729596014877671,
+      "learning_rate": 4.851030494940716e-06,
+      "loss": 0.4697,
+      "step": 999
+    },
+    {
+      "epoch": 0.4655493482309125,
+      "grad_norm": 0.6254203451360453,
+      "learning_rate": 4.850593877294323e-06,
+      "loss": 0.455,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4660148975791434,
+      "grad_norm": 0.6023825673803682,
+      "learning_rate": 4.850156640445605e-06,
+      "loss": 0.4459,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4664804469273743,
+      "grad_norm": 0.631699134087582,
+      "learning_rate": 4.849718784509743e-06,
+      "loss": 0.4638,
+      "step": 1002
+    },
+    {
+      "epoch": 0.46694599627560524,
+      "grad_norm": 0.9296322256720971,
+      "learning_rate": 4.849280309602077e-06,
+      "loss": 0.4608,
+      "step": 1003
+    },
+    {
+      "epoch": 0.46741154562383613,
+      "grad_norm": 0.6385132646289368,
+      "learning_rate": 4.848841215838112e-06,
+      "loss": 0.4614,
+      "step": 1004
+    },
+    {
+      "epoch": 0.46787709497206703,
+      "grad_norm": 0.587137805200072,
+      "learning_rate": 4.848401503333516e-06,
+      "loss": 0.4613,
+      "step": 1005
+    },
+    {
+      "epoch": 0.46834264432029793,
+      "grad_norm": 0.6554289367932853,
+      "learning_rate": 4.847961172204121e-06,
+      "loss": 0.4686,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4688081936685289,
+      "grad_norm": 0.6478241632621988,
+      "learning_rate": 4.8475202225659184e-06,
+      "loss": 0.4583,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4692737430167598,
+      "grad_norm": 0.6350581574268478,
+      "learning_rate": 4.847078654535066e-06,
+      "loss": 0.4462,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4697392923649907,
+      "grad_norm": 0.6657395478268403,
+      "learning_rate": 4.846636468227883e-06,
+      "loss": 0.4654,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4702048417132216,
+      "grad_norm": 0.6516058315638855,
+      "learning_rate": 4.846193663760852e-06,
+      "loss": 0.4563,
+      "step": 1010
+    },
+    {
+      "epoch": 0.47067039106145253,
+      "grad_norm": 0.7131778412227324,
+      "learning_rate": 4.845750241250619e-06,
+      "loss": 0.4888,
+      "step": 1011
+    },
+    {
+      "epoch": 0.47113594040968343,
+      "grad_norm": 0.6428812161611224,
+      "learning_rate": 4.845306200813991e-06,
+      "loss": 0.4487,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4716014897579143,
+      "grad_norm": 0.6347077067081834,
+      "learning_rate": 4.8448615425679394e-06,
+      "loss": 0.4745,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4720670391061452,
+      "grad_norm": 0.591944851660107,
+      "learning_rate": 4.844416266629596e-06,
+      "loss": 0.4642,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4725325884543762,
+      "grad_norm": 0.7067353406235319,
+      "learning_rate": 4.843970373116259e-06,
+      "loss": 0.47,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4729981378026071,
+      "grad_norm": 0.7012382214647699,
+      "learning_rate": 4.843523862145387e-06,
+      "loss": 0.4701,
+      "step": 1016
+    },
+    {
+      "epoch": 0.473463687150838,
+      "grad_norm": 0.6359667958782377,
+      "learning_rate": 4.8430767338346015e-06,
+      "loss": 0.4498,
+      "step": 1017
+    },
+    {
+      "epoch": 0.47392923649906893,
+      "grad_norm": 0.6176525535656018,
+      "learning_rate": 4.8426289883016874e-06,
+      "loss": 0.4499,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4743947858472998,
+      "grad_norm": 0.630435571931979,
+      "learning_rate": 4.84218062566459e-06,
+      "loss": 0.4562,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4748603351955307,
+      "grad_norm": 0.6300344228713561,
+      "learning_rate": 4.8417316460414195e-06,
+      "loss": 0.4751,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4753258845437616,
+      "grad_norm": 0.6118041088632397,
+      "learning_rate": 4.841282049550449e-06,
+      "loss": 0.4717,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4757914338919926,
+      "grad_norm": 0.6036750722706604,
+      "learning_rate": 4.84083183631011e-06,
+      "loss": 0.4767,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4762569832402235,
+      "grad_norm": 0.6331915670787607,
+      "learning_rate": 4.840381006439001e-06,
+      "loss": 0.4594,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4767225325884544,
+      "grad_norm": 0.5900479995540634,
+      "learning_rate": 4.839929560055883e-06,
+      "loss": 0.455,
+      "step": 1024
+    },
+    {
+      "epoch": 0.47718808193668527,
+      "grad_norm": 0.6689097556220146,
+      "learning_rate": 4.839477497279675e-06,
+      "loss": 0.4442,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4776536312849162,
+      "grad_norm": 0.6328005759955856,
+      "learning_rate": 4.839024818229462e-06,
+      "loss": 0.4629,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4781191806331471,
+      "grad_norm": 0.5955973066941195,
+      "learning_rate": 4.838571523024491e-06,
+      "loss": 0.4514,
+      "step": 1027
+    },
+    {
+      "epoch": 0.478584729981378,
+      "grad_norm": 0.6357885313669589,
+      "learning_rate": 4.83811761178417e-06,
+      "loss": 0.4814,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4790502793296089,
+      "grad_norm": 0.6155979174799342,
+      "learning_rate": 4.8376630846280705e-06,
+      "loss": 0.4668,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4795158286778399,
+      "grad_norm": 0.6427731688673556,
+      "learning_rate": 4.837207941675926e-06,
+      "loss": 0.4749,
+      "step": 1030
+    },
+    {
+      "epoch": 0.47998137802607077,
+      "grad_norm": 0.6345791279551041,
+      "learning_rate": 4.836752183047632e-06,
+      "loss": 0.4516,
+      "step": 1031
+    },
+    {
+      "epoch": 0.48044692737430167,
+      "grad_norm": 0.6712083330290028,
+      "learning_rate": 4.836295808863244e-06,
+      "loss": 0.4802,
+      "step": 1032
+    },
+    {
+      "epoch": 0.48091247672253257,
+      "grad_norm": 0.6505454334807397,
+      "learning_rate": 4.835838819242985e-06,
+      "loss": 0.4559,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4813780260707635,
+      "grad_norm": 0.6370674858313976,
+      "learning_rate": 4.835381214307235e-06,
+      "loss": 0.4586,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4818435754189944,
+      "grad_norm": 0.6379537241370451,
+      "learning_rate": 4.834922994176539e-06,
+      "loss": 0.4794,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4823091247672253,
+      "grad_norm": 0.6370175931303382,
+      "learning_rate": 4.834464158971602e-06,
+      "loss": 0.4583,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4827746741154562,
+      "grad_norm": 0.6798975856648598,
+      "learning_rate": 4.834004708813293e-06,
+      "loss": 0.4844,
+      "step": 1037
+    },
+    {
+      "epoch": 0.48324022346368717,
+      "grad_norm": 0.6476182000849392,
+      "learning_rate": 4.833544643822642e-06,
+      "loss": 0.4813,
+      "step": 1038
+    },
+    {
+      "epoch": 0.48370577281191807,
+      "grad_norm": 0.6338445283142492,
+      "learning_rate": 4.83308396412084e-06,
+      "loss": 0.4542,
+      "step": 1039
+    },
+    {
+      "epoch": 0.48417132216014896,
+      "grad_norm": 0.6286214432140195,
+      "learning_rate": 4.832622669829242e-06,
+      "loss": 0.4756,
+      "step": 1040
+    },
+    {
+      "epoch": 0.48463687150837986,
+      "grad_norm": 0.6685571716215326,
+      "learning_rate": 4.832160761069365e-06,
+      "loss": 0.4708,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4851024208566108,
+      "grad_norm": 0.6681162758627683,
+      "learning_rate": 4.831698237962884e-06,
+      "loss": 0.4728,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4855679702048417,
+      "grad_norm": 0.600001686109816,
+      "learning_rate": 4.83123510063164e-06,
+      "loss": 0.4676,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4860335195530726,
+      "grad_norm": 0.6266138527342523,
+      "learning_rate": 4.830771349197636e-06,
+      "loss": 0.4712,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4864990689013035,
+      "grad_norm": 0.6330878221524688,
+      "learning_rate": 4.830306983783033e-06,
+      "loss": 0.4698,
+      "step": 1045
+    },
+    {
+      "epoch": 0.48696461824953446,
+      "grad_norm": 0.670838148027699,
+      "learning_rate": 4.8298420045101555e-06,
+      "loss": 0.4729,
+      "step": 1046
+    },
+    {
+      "epoch": 0.48743016759776536,
+      "grad_norm": 0.6127963028459029,
+      "learning_rate": 4.829376411501492e-06,
+      "loss": 0.4426,
+      "step": 1047
+    },
+    {
+      "epoch": 0.48789571694599626,
+      "grad_norm": 0.6479565675839415,
+      "learning_rate": 4.82891020487969e-06,
+      "loss": 0.4672,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4883612662942272,
+      "grad_norm": 0.6052027358207316,
+      "learning_rate": 4.8284433847675584e-06,
+      "loss": 0.4731,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4888268156424581,
+      "grad_norm": 0.6187259978024328,
+      "learning_rate": 4.827975951288071e-06,
+      "loss": 0.4652,
+      "step": 1050
+    },
+    {
+      "epoch": 0.489292364990689,
+      "grad_norm": 0.6172987368097329,
+      "learning_rate": 4.827507904564358e-06,
+      "loss": 0.4662,
+      "step": 1051
+    },
+    {
+      "epoch": 0.4897579143389199,
+      "grad_norm": 0.6162502024079254,
+      "learning_rate": 4.8270392447197164e-06,
+      "loss": 0.4645,
+      "step": 1052
+    },
+    {
+      "epoch": 0.49022346368715086,
+      "grad_norm": 0.6146328193036777,
+      "learning_rate": 4.826569971877601e-06,
+      "loss": 0.4473,
+      "step": 1053
+    },
+    {
+      "epoch": 0.49068901303538176,
+      "grad_norm": 0.605177279123611,
+      "learning_rate": 4.82610008616163e-06,
+      "loss": 0.4466,
+      "step": 1054
+    },
+    {
+      "epoch": 0.49115456238361266,
+      "grad_norm": 0.6943367323820276,
+      "learning_rate": 4.8256295876955825e-06,
+      "loss": 0.4638,
+      "step": 1055
+    },
+    {
+      "epoch": 0.49162011173184356,
+      "grad_norm": 0.6573674459780595,
+      "learning_rate": 4.825158476603399e-06,
+      "loss": 0.4657,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4920856610800745,
+      "grad_norm": 0.6411373374354522,
+      "learning_rate": 4.824686753009181e-06,
+      "loss": 0.4682,
+      "step": 1057
+    },
+    {
+      "epoch": 0.4925512104283054,
+      "grad_norm": 0.5844450972577233,
+      "learning_rate": 4.824214417037192e-06,
+      "loss": 0.4496,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4930167597765363,
+      "grad_norm": 0.6289707217802873,
+      "learning_rate": 4.823741468811857e-06,
+      "loss": 0.4768,
+      "step": 1059
+    },
+    {
+      "epoch": 0.4934823091247672,
+      "grad_norm": 0.7100796118706829,
+      "learning_rate": 4.8232679084577604e-06,
+      "loss": 0.4634,
+      "step": 1060
+    },
+    {
+      "epoch": 0.49394785847299816,
+      "grad_norm": 0.6157654064617827,
+      "learning_rate": 4.82279373609965e-06,
+      "loss": 0.4587,
+      "step": 1061
+    },
+    {
+      "epoch": 0.49441340782122906,
+      "grad_norm": 0.5875501191603181,
+      "learning_rate": 4.822318951862435e-06,
+      "loss": 0.4316,
+      "step": 1062
+    },
+    {
+      "epoch": 0.49487895716945995,
+      "grad_norm": 0.622427778455474,
+      "learning_rate": 4.821843555871184e-06,
+      "loss": 0.4462,
+      "step": 1063
+    },
+    {
+      "epoch": 0.49534450651769085,
+      "grad_norm": 0.6429549688573667,
+      "learning_rate": 4.821367548251127e-06,
+      "loss": 0.4733,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4958100558659218,
+      "grad_norm": 0.6760741721482619,
+      "learning_rate": 4.8208909291276575e-06,
+      "loss": 0.4496,
+      "step": 1065
+    },
+    {
+      "epoch": 0.4962756052141527,
+      "grad_norm": 0.6660305997744997,
+      "learning_rate": 4.820413698626326e-06,
+      "loss": 0.4542,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4967411545623836,
+      "grad_norm": 0.608153930720617,
+      "learning_rate": 4.819935856872849e-06,
+      "loss": 0.4637,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4972067039106145,
+      "grad_norm": 0.6046127238861214,
+      "learning_rate": 4.819457403993098e-06,
+      "loss": 0.4713,
+      "step": 1068
+    },
+    {
+      "epoch": 0.49767225325884545,
+      "grad_norm": 0.60454700528616,
+      "learning_rate": 4.818978340113112e-06,
+      "loss": 0.4601,
+      "step": 1069
+    },
+    {
+      "epoch": 0.49813780260707635,
+      "grad_norm": 0.6313782873894814,
+      "learning_rate": 4.818498665359086e-06,
+      "loss": 0.4421,
+      "step": 1070
+    },
+    {
+      "epoch": 0.49860335195530725,
+      "grad_norm": 0.7039492009670417,
+      "learning_rate": 4.81801837985738e-06,
+      "loss": 0.4687,
+      "step": 1071
+    },
+    {
+      "epoch": 0.49906890130353815,
+      "grad_norm": 0.6460011028118299,
+      "learning_rate": 4.8175374837345094e-06,
+      "loss": 0.4514,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4995344506517691,
+      "grad_norm": 0.648442847846222,
+      "learning_rate": 4.8170559771171555e-06,
+      "loss": 0.4572,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6915468802066607,
+      "learning_rate": 4.8165738601321585e-06,
+      "loss": 0.4606,
+      "step": 1074
+    },
+    {
+      "epoch": 0.500465549348231,
+      "grad_norm": 0.6187603449564465,
+      "learning_rate": 4.81609113290652e-06,
+      "loss": 0.4801,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5009310986964618,
+      "grad_norm": 0.6319283415803244,
+      "learning_rate": 4.8156077955674e-06,
+      "loss": 0.4582,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5013966480446927,
+      "grad_norm": 0.6352328997809649,
+      "learning_rate": 4.815123848242123e-06,
+      "loss": 0.4611,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5018621973929237,
+      "grad_norm": 0.6206515917644025,
+      "learning_rate": 4.814639291058171e-06,
+      "loss": 0.4518,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5023277467411545,
+      "grad_norm": 0.6279898849818845,
+      "learning_rate": 4.814154124143188e-06,
+      "loss": 0.4401,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5027932960893855,
+      "grad_norm": 0.6149343890205446,
+      "learning_rate": 4.8136683476249805e-06,
+      "loss": 0.4815,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5032588454376163,
+      "grad_norm": 0.6249795446935299,
+      "learning_rate": 4.81318196163151e-06,
+      "loss": 0.4774,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5037243947858473,
+      "grad_norm": 0.6444020606719267,
+      "learning_rate": 4.812694966290905e-06,
+      "loss": 0.4538,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5041899441340782,
+      "grad_norm": 0.6279622554682457,
+      "learning_rate": 4.81220736173145e-06,
+      "loss": 0.4484,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5046554934823091,
+      "grad_norm": 0.6569319514548273,
+      "learning_rate": 4.811719148081594e-06,
+      "loss": 0.4654,
+      "step": 1084
+    },
+    {
+      "epoch": 0.50512104283054,
+      "grad_norm": 0.6470558764514603,
+      "learning_rate": 4.811230325469941e-06,
+      "loss": 0.465,
+      "step": 1085
+    },
+    {
+      "epoch": 0.505586592178771,
+      "grad_norm": 0.6450288353345118,
+      "learning_rate": 4.81074089402526e-06,
+      "loss": 0.4527,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5060521415270018,
+      "grad_norm": 0.6258311455281796,
+      "learning_rate": 4.810250853876479e-06,
+      "loss": 0.4631,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5065176908752328,
+      "grad_norm": 0.6530437093286651,
+      "learning_rate": 4.809760205152687e-06,
+      "loss": 0.4431,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5069832402234636,
+      "grad_norm": 0.7205773715551725,
+      "learning_rate": 4.80926894798313e-06,
+      "loss": 0.475,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5074487895716946,
+      "grad_norm": 0.6636230519077175,
+      "learning_rate": 4.8087770824972195e-06,
+      "loss": 0.4801,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5079143389199255,
+      "grad_norm": 0.6143632027362663,
+      "learning_rate": 4.8082846088245225e-06,
+      "loss": 0.4598,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5083798882681564,
+      "grad_norm": 0.6338501529755833,
+      "learning_rate": 4.8077915270947705e-06,
+      "loss": 0.4451,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5088454376163873,
+      "grad_norm": 0.667013498757164,
+      "learning_rate": 4.807297837437851e-06,
+      "loss": 0.4439,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5093109869646183,
+      "grad_norm": 0.6231787304564216,
+      "learning_rate": 4.806803539983814e-06,
+      "loss": 0.4493,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5097765363128491,
+      "grad_norm": 0.6273342398772456,
+      "learning_rate": 4.80630863486287e-06,
+      "loss": 0.4521,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5102420856610801,
+      "grad_norm": 0.6369459200134047,
+      "learning_rate": 4.805813122205388e-06,
+      "loss": 0.4608,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5107076350093109,
+      "grad_norm": 0.6558159254711259,
+      "learning_rate": 4.805317002141898e-06,
+      "loss": 0.4589,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5111731843575419,
+      "grad_norm": 0.6271087194786253,
+      "learning_rate": 4.804820274803091e-06,
+      "loss": 0.4601,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5116387337057728,
+      "grad_norm": 0.6415882571358437,
+      "learning_rate": 4.804322940319815e-06,
+      "loss": 0.46,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5121042830540037,
+      "grad_norm": 0.6880185027064674,
+      "learning_rate": 4.8038249988230805e-06,
+      "loss": 0.4589,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5125698324022346,
+      "grad_norm": 0.6720399207070452,
+      "learning_rate": 4.8033264504440564e-06,
+      "loss": 0.4862,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5130353817504656,
+      "grad_norm": 0.6423259198585777,
+      "learning_rate": 4.8028272953140735e-06,
+      "loss": 0.4594,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5135009310986964,
+      "grad_norm": 0.6960417193664723,
+      "learning_rate": 4.80232753356462e-06,
+      "loss": 0.4556,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5139664804469274,
+      "grad_norm": 0.6311406000487755,
+      "learning_rate": 4.8018271653273466e-06,
+      "loss": 0.4568,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5144320297951583,
+      "grad_norm": 0.6435130313758896,
+      "learning_rate": 4.80132619073406e-06,
+      "loss": 0.4594,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5148975791433892,
+      "grad_norm": 0.6533552620243155,
+      "learning_rate": 4.80082460991673e-06,
+      "loss": 0.473,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5153631284916201,
+      "grad_norm": 0.6519706933364636,
+      "learning_rate": 4.800322423007485e-06,
+      "loss": 0.4567,
+      "step": 1107
+    },
+    {
+      "epoch": 0.515828677839851,
+      "grad_norm": 0.6371704862167704,
+      "learning_rate": 4.799819630138613e-06,
+      "loss": 0.4615,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5162942271880819,
+      "grad_norm": 0.6173985438054354,
+      "learning_rate": 4.799316231442562e-06,
+      "loss": 0.4634,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5167597765363129,
+      "grad_norm": 0.6451980573951619,
+      "learning_rate": 4.798812227051938e-06,
+      "loss": 0.4662,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5172253258845437,
+      "grad_norm": 0.6468883880901448,
+      "learning_rate": 4.798307617099509e-06,
+      "loss": 0.4594,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5176908752327747,
+      "grad_norm": 0.6601977852751778,
+      "learning_rate": 4.7978024017182e-06,
+      "loss": 0.4616,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5181564245810056,
+      "grad_norm": 0.6698778065271487,
+      "learning_rate": 4.797296581041097e-06,
+      "loss": 0.4535,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5186219739292365,
+      "grad_norm": 0.6452982680372866,
+      "learning_rate": 4.796790155201446e-06,
+      "loss": 0.4496,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5190875232774674,
+      "grad_norm": 0.6137075973059919,
+      "learning_rate": 4.796283124332651e-06,
+      "loss": 0.4746,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5195530726256983,
+      "grad_norm": 0.6115685142940461,
+      "learning_rate": 4.795775488568276e-06,
+      "loss": 0.4593,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5200186219739292,
+      "grad_norm": 0.626172055236667,
+      "learning_rate": 4.795267248042045e-06,
+      "loss": 0.4701,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5204841713221602,
+      "grad_norm": 0.6359672750438967,
+      "learning_rate": 4.79475840288784e-06,
+      "loss": 0.4606,
+      "step": 1118
+    },
+    {
+      "epoch": 0.520949720670391,
+      "grad_norm": 0.578846542541125,
+      "learning_rate": 4.794248953239703e-06,
+      "loss": 0.4714,
+      "step": 1119
+    },
+    {
+      "epoch": 0.521415270018622,
+      "grad_norm": 0.6303357028164904,
+      "learning_rate": 4.793738899231836e-06,
+      "loss": 0.4648,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5218808193668529,
+      "grad_norm": 0.6230369103149863,
+      "learning_rate": 4.793228240998597e-06,
+      "loss": 0.4677,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5223463687150838,
+      "grad_norm": 0.6716575788497521,
+      "learning_rate": 4.792716978674508e-06,
+      "loss": 0.4547,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5228119180633147,
+      "grad_norm": 0.6506438654483642,
+      "learning_rate": 4.792205112394246e-06,
+      "loss": 0.4649,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5232774674115456,
+      "grad_norm": 0.638462575920576,
+      "learning_rate": 4.79169264229265e-06,
+      "loss": 0.4497,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5237430167597765,
+      "grad_norm": 0.6550293857787693,
+      "learning_rate": 4.791179568504716e-06,
+      "loss": 0.4766,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5242085661080075,
+      "grad_norm": 0.6448593422517506,
+      "learning_rate": 4.790665891165601e-06,
+      "loss": 0.4579,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5246741154562383,
+      "grad_norm": 0.627378331500031,
+      "learning_rate": 4.790151610410619e-06,
+      "loss": 0.4891,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5251396648044693,
+      "grad_norm": 0.5957098729753422,
+      "learning_rate": 4.789636726375242e-06,
+      "loss": 0.459,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5256052141527002,
+      "grad_norm": 0.695888039903327,
+      "learning_rate": 4.789121239195106e-06,
+      "loss": 0.4533,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5260707635009311,
+      "grad_norm": 0.6733509277015868,
+      "learning_rate": 4.7886051490060005e-06,
+      "loss": 0.4858,
+      "step": 1130
+    },
+    {
+      "epoch": 0.526536312849162,
+      "grad_norm": 0.6224196359677772,
+      "learning_rate": 4.788088455943876e-06,
+      "loss": 0.4585,
+      "step": 1131
+    },
+    {
+      "epoch": 0.527001862197393,
+      "grad_norm": 0.6363555149811734,
+      "learning_rate": 4.787571160144843e-06,
+      "loss": 0.4703,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5274674115456238,
+      "grad_norm": 0.632133671006589,
+      "learning_rate": 4.787053261745168e-06,
+      "loss": 0.4536,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5279329608938548,
+      "grad_norm": 0.6367184585125033,
+      "learning_rate": 4.786534760881278e-06,
+      "loss": 0.4629,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5283985102420856,
+      "grad_norm": 0.6500797399973317,
+      "learning_rate": 4.78601565768976e-06,
+      "loss": 0.4746,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5288640595903166,
+      "grad_norm": 0.6192547076593807,
+      "learning_rate": 4.785495952307357e-06,
+      "loss": 0.4711,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5293296089385475,
+      "grad_norm": 0.6604615070910439,
+      "learning_rate": 4.784975644870971e-06,
+      "loss": 0.4669,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5297951582867784,
+      "grad_norm": 0.6318049599588975,
+      "learning_rate": 4.7844547355176645e-06,
+      "loss": 0.4639,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5302607076350093,
+      "grad_norm": 0.6341989642934386,
+      "learning_rate": 4.783933224384657e-06,
+      "loss": 0.4604,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5307262569832403,
+      "grad_norm": 0.6001955950516837,
+      "learning_rate": 4.7834111116093275e-06,
+      "loss": 0.4547,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5311918063314711,
+      "grad_norm": 0.65469664707481,
+      "learning_rate": 4.782888397329212e-06,
+      "loss": 0.4712,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5316573556797021,
+      "grad_norm": 0.6635251010253044,
+      "learning_rate": 4.782365081682006e-06,
+      "loss": 0.4571,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5321229050279329,
+      "grad_norm": 0.6433905803984069,
+      "learning_rate": 4.781841164805564e-06,
+      "loss": 0.4638,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5325884543761639,
+      "grad_norm": 0.6194018057267306,
+      "learning_rate": 4.781316646837898e-06,
+      "loss": 0.461,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5330540037243948,
+      "grad_norm": 0.6152733910535632,
+      "learning_rate": 4.780791527917177e-06,
+      "loss": 0.4633,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5335195530726257,
+      "grad_norm": 0.6124171502784678,
+      "learning_rate": 4.780265808181732e-06,
+      "loss": 0.463,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5339851024208566,
+      "grad_norm": 0.6454834053768772,
+      "learning_rate": 4.779739487770049e-06,
+      "loss": 0.4814,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5344506517690876,
+      "grad_norm": 0.6652528738975773,
+      "learning_rate": 4.779212566820773e-06,
+      "loss": 0.4504,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5349162011173184,
+      "grad_norm": 0.6286332266936884,
+      "learning_rate": 4.778685045472708e-06,
+      "loss": 0.4668,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5353817504655494,
+      "grad_norm": 0.6496359064752516,
+      "learning_rate": 4.778156923864816e-06,
+      "loss": 0.4718,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5358472998137802,
+      "grad_norm": 0.6339066719452602,
+      "learning_rate": 4.777628202136217e-06,
+      "loss": 0.4514,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5363128491620112,
+      "grad_norm": 0.5948315904146868,
+      "learning_rate": 4.777098880426187e-06,
+      "loss": 0.4601,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5367783985102421,
+      "grad_norm": 0.653027213413644,
+      "learning_rate": 4.776568958874164e-06,
+      "loss": 0.4668,
+      "step": 1153
+    },
+    {
+      "epoch": 0.537243947858473,
+      "grad_norm": 0.6434767471255578,
+      "learning_rate": 4.77603843761974e-06,
+      "loss": 0.4767,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5377094972067039,
+      "grad_norm": 0.670995119519536,
+      "learning_rate": 4.77550731680267e-06,
+      "loss": 0.4929,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5381750465549349,
+      "grad_norm": 0.6620447767530563,
+      "learning_rate": 4.77497559656286e-06,
+      "loss": 0.4506,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5386405959031657,
+      "grad_norm": 0.6115039510055779,
+      "learning_rate": 4.77444327704038e-06,
+      "loss": 0.4618,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5391061452513967,
+      "grad_norm": 0.6395903737911849,
+      "learning_rate": 4.7739103583754565e-06,
+      "loss": 0.4673,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5395716945996276,
+      "grad_norm": 0.6172634354861881,
+      "learning_rate": 4.773376840708472e-06,
+      "loss": 0.46,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5400372439478585,
+      "grad_norm": 0.6599911490705223,
+      "learning_rate": 4.7728427241799655e-06,
+      "loss": 0.4853,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5405027932960894,
+      "grad_norm": 0.6280722761360461,
+      "learning_rate": 4.772308008930639e-06,
+      "loss": 0.4664,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5409683426443203,
+      "grad_norm": 0.5968189893261003,
+      "learning_rate": 4.771772695101348e-06,
+      "loss": 0.4558,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5414338919925512,
+      "grad_norm": 0.6216382842673579,
+      "learning_rate": 4.771236782833107e-06,
+      "loss": 0.4495,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5418994413407822,
+      "grad_norm": 0.6308727367554738,
+      "learning_rate": 4.770700272267088e-06,
+      "loss": 0.4584,
+      "step": 1164
+    },
+    {
+      "epoch": 0.542364990689013,
+      "grad_norm": 0.6139329408838153,
+      "learning_rate": 4.770163163544621e-06,
+      "loss": 0.4715,
+      "step": 1165
+    },
+    {
+      "epoch": 0.542830540037244,
+      "grad_norm": 0.6394878176507541,
+      "learning_rate": 4.769625456807193e-06,
+      "loss": 0.4403,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5432960893854749,
+      "grad_norm": 0.6341634494098936,
+      "learning_rate": 4.769087152196449e-06,
+      "loss": 0.4759,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5437616387337058,
+      "grad_norm": 0.6278319162057319,
+      "learning_rate": 4.7685482498541905e-06,
+      "loss": 0.4696,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5442271880819367,
+      "grad_norm": 0.6432922087386866,
+      "learning_rate": 4.768008749922377e-06,
+      "loss": 0.4599,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5446927374301676,
+      "grad_norm": 0.634911131460627,
+      "learning_rate": 4.767468652543127e-06,
+      "loss": 0.4519,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5451582867783985,
+      "grad_norm": 0.6382868795784808,
+      "learning_rate": 4.766927957858715e-06,
+      "loss": 0.4625,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5456238361266295,
+      "grad_norm": 0.6418453850234203,
+      "learning_rate": 4.766386666011571e-06,
+      "loss": 0.4425,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5460893854748603,
+      "grad_norm": 0.6110969440342874,
+      "learning_rate": 4.765844777144285e-06,
+      "loss": 0.4421,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5465549348230913,
+      "grad_norm": 0.646946468086923,
+      "learning_rate": 4.765302291399604e-06,
+      "loss": 0.4462,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5470204841713222,
+      "grad_norm": 0.6529338091441416,
+      "learning_rate": 4.764759208920431e-06,
+      "loss": 0.4396,
+      "step": 1175
+    },
+    {
+      "epoch": 0.547486033519553,
+      "grad_norm": 0.6262577934345215,
+      "learning_rate": 4.764215529849827e-06,
+      "loss": 0.4425,
+      "step": 1176
+    },
+    {
+      "epoch": 0.547951582867784,
+      "grad_norm": 0.6360616634119529,
+      "learning_rate": 4.763671254331011e-06,
+      "loss": 0.4463,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5484171322160148,
+      "grad_norm": 0.6232371023302613,
+      "learning_rate": 4.763126382507356e-06,
+      "loss": 0.4728,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5488826815642458,
+      "grad_norm": 0.6417254130525217,
+      "learning_rate": 4.7625809145223965e-06,
+      "loss": 0.4508,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5493482309124768,
+      "grad_norm": 0.6649777735146172,
+      "learning_rate": 4.76203485051982e-06,
+      "loss": 0.4592,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5498137802607076,
+      "grad_norm": 0.5838674205202774,
+      "learning_rate": 4.761488190643474e-06,
+      "loss": 0.4601,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5502793296089385,
+      "grad_norm": 0.6368695774877664,
+      "learning_rate": 4.7609409350373605e-06,
+      "loss": 0.4673,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5507448789571695,
+      "grad_norm": 0.6154106666980449,
+      "learning_rate": 4.76039308384564e-06,
+      "loss": 0.4563,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5512104283054003,
+      "grad_norm": 0.5963965868541832,
+      "learning_rate": 4.7598446372126304e-06,
+      "loss": 0.4521,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5516759776536313,
+      "grad_norm": 0.6015505525657838,
+      "learning_rate": 4.759295595282805e-06,
+      "loss": 0.4523,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5521415270018621,
+      "grad_norm": 0.643251603813869,
+      "learning_rate": 4.758745958200794e-06,
+      "loss": 0.4694,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5526070763500931,
+      "grad_norm": 0.6284437039185229,
+      "learning_rate": 4.758195726111386e-06,
+      "loss": 0.4612,
+      "step": 1187
+    },
+    {
+      "epoch": 0.553072625698324,
+      "grad_norm": 0.607566594794309,
+      "learning_rate": 4.757644899159524e-06,
+      "loss": 0.4608,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5535381750465549,
+      "grad_norm": 0.6733866014963974,
+      "learning_rate": 4.75709347749031e-06,
+      "loss": 0.4513,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5540037243947858,
+      "grad_norm": 0.6185082827888421,
+      "learning_rate": 4.756541461249e-06,
+      "loss": 0.4545,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5544692737430168,
+      "grad_norm": 0.6771316580612468,
+      "learning_rate": 4.75598885058101e-06,
+      "loss": 0.4361,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5549348230912476,
+      "grad_norm": 0.5977953602562018,
+      "learning_rate": 4.75543564563191e-06,
+      "loss": 0.4435,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5554003724394786,
+      "grad_norm": 0.6122271990059766,
+      "learning_rate": 4.754881846547428e-06,
+      "loss": 0.4505,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5558659217877095,
+      "grad_norm": 0.5907866856053179,
+      "learning_rate": 4.7543274534734464e-06,
+      "loss": 0.4764,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5563314711359404,
+      "grad_norm": 0.7048075086826702,
+      "learning_rate": 4.753772466556006e-06,
+      "loss": 0.4693,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5567970204841713,
+      "grad_norm": 0.5770699529707782,
+      "learning_rate": 4.753216885941305e-06,
+      "loss": 0.4445,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5572625698324022,
+      "grad_norm": 0.635949159809215,
+      "learning_rate": 4.752660711775695e-06,
+      "loss": 0.4593,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5577281191806331,
+      "grad_norm": 0.6439981332661128,
+      "learning_rate": 4.7521039442056855e-06,
+      "loss": 0.4473,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5581936685288641,
+      "grad_norm": 0.6376119504407562,
+      "learning_rate": 4.751546583377943e-06,
+      "loss": 0.4328,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5586592178770949,
+      "grad_norm": 0.649067616908171,
+      "learning_rate": 4.750988629439289e-06,
+      "loss": 0.4446,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5591247672253259,
+      "grad_norm": 0.6744347995587144,
+      "learning_rate": 4.750430082536704e-06,
+      "loss": 0.4714,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5595903165735568,
+      "grad_norm": 0.6244997337527148,
+      "learning_rate": 4.749870942817318e-06,
+      "loss": 0.4678,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5600558659217877,
+      "grad_norm": 0.6428077304710175,
+      "learning_rate": 4.749311210428426e-06,
+      "loss": 0.4358,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5605214152700186,
+      "grad_norm": 0.6322163386582066,
+      "learning_rate": 4.748750885517473e-06,
+      "loss": 0.4722,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5609869646182495,
+      "grad_norm": 0.5880583333197749,
+      "learning_rate": 4.748189968232062e-06,
+      "loss": 0.4661,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5614525139664804,
+      "grad_norm": 0.654603569108245,
+      "learning_rate": 4.747628458719953e-06,
+      "loss": 0.4458,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5619180633147114,
+      "grad_norm": 0.6394340836488789,
+      "learning_rate": 4.7470663571290595e-06,
+      "loss": 0.4709,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5623836126629422,
+      "grad_norm": 0.6090058190471943,
+      "learning_rate": 4.746503663607453e-06,
+      "loss": 0.4772,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5628491620111732,
+      "grad_norm": 0.6402397940044264,
+      "learning_rate": 4.74594037830336e-06,
+      "loss": 0.4562,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5633147113594041,
+      "grad_norm": 0.6057608769664429,
+      "learning_rate": 4.745376501365164e-06,
+      "loss": 0.4657,
+      "step": 1210
+    },
+    {
+      "epoch": 0.563780260707635,
+      "grad_norm": 0.6258102060749121,
+      "learning_rate": 4.744812032941403e-06,
+      "loss": 0.4533,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5642458100558659,
+      "grad_norm": 0.5723636335058497,
+      "learning_rate": 4.7442469731807715e-06,
+      "loss": 0.4367,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5647113594040968,
+      "grad_norm": 0.6532254727862885,
+      "learning_rate": 4.74368132223212e-06,
+      "loss": 0.4649,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5651769087523277,
+      "grad_norm": 0.6377193368702198,
+      "learning_rate": 4.743115080244454e-06,
+      "loss": 0.4667,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5656424581005587,
+      "grad_norm": 0.6685276902938069,
+      "learning_rate": 4.742548247366936e-06,
+      "loss": 0.4477,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5661080074487895,
+      "grad_norm": 0.6093200747781764,
+      "learning_rate": 4.741980823748881e-06,
+      "loss": 0.4454,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5665735567970205,
+      "grad_norm": 0.646065074220935,
+      "learning_rate": 4.741412809539765e-06,
+      "loss": 0.4467,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5670391061452514,
+      "grad_norm": 0.6496740938137532,
+      "learning_rate": 4.740844204889214e-06,
+      "loss": 0.4597,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5675046554934823,
+      "grad_norm": 0.7162422852026953,
+      "learning_rate": 4.740275009947013e-06,
+      "loss": 0.4552,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5679702048417132,
+      "grad_norm": 0.7457935002822215,
+      "learning_rate": 4.739705224863101e-06,
+      "loss": 0.474,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5684357541899442,
+      "grad_norm": 1.1624087049113492,
+      "learning_rate": 4.7391348497875735e-06,
+      "loss": 0.4683,
+      "step": 1221
+    },
+    {
+      "epoch": 0.568901303538175,
+      "grad_norm": 0.6457635889850405,
+      "learning_rate": 4.738563884870682e-06,
+      "loss": 0.469,
+      "step": 1222
+    },
+    {
+      "epoch": 0.569366852886406,
+      "grad_norm": 0.6317093815282149,
+      "learning_rate": 4.7379923302628285e-06,
+      "loss": 0.4578,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5698324022346368,
+      "grad_norm": 0.6211765943244856,
+      "learning_rate": 4.737420186114579e-06,
+      "loss": 0.4404,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5702979515828678,
+      "grad_norm": 0.6775802182318915,
+      "learning_rate": 4.736847452576645e-06,
+      "loss": 0.4596,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5707635009310987,
+      "grad_norm": 0.6771639547038202,
+      "learning_rate": 4.736274129799902e-06,
+      "loss": 0.4587,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5712290502793296,
+      "grad_norm": 0.7284814029805603,
+      "learning_rate": 4.735700217935374e-06,
+      "loss": 0.4878,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5716945996275605,
+      "grad_norm": 0.6001418697819199,
+      "learning_rate": 4.735125717134245e-06,
+      "loss": 0.4459,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5721601489757915,
+      "grad_norm": 0.6465052145400169,
+      "learning_rate": 4.734550627547852e-06,
+      "loss": 0.4653,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5726256983240223,
+      "grad_norm": 0.6614383481032058,
+      "learning_rate": 4.733974949327686e-06,
+      "loss": 0.4882,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5730912476722533,
+      "grad_norm": 0.6253454036652443,
+      "learning_rate": 4.733398682625396e-06,
+      "loss": 0.4604,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5735567970204841,
+      "grad_norm": 0.6030834380453912,
+      "learning_rate": 4.732821827592783e-06,
+      "loss": 0.4377,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5740223463687151,
+      "grad_norm": 0.6736220282665687,
+      "learning_rate": 4.732244384381805e-06,
+      "loss": 0.4694,
+      "step": 1233
+    },
+    {
+      "epoch": 0.574487895716946,
+      "grad_norm": 0.6606893990211111,
+      "learning_rate": 4.7316663531445735e-06,
+      "loss": 0.4653,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5749534450651769,
+      "grad_norm": 0.6390483417466403,
+      "learning_rate": 4.731087734033356e-06,
+      "loss": 0.4499,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5754189944134078,
+      "grad_norm": 0.6430040887922893,
+      "learning_rate": 4.730508527200576e-06,
+      "loss": 0.4821,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5758845437616388,
+      "grad_norm": 0.7205011003722405,
+      "learning_rate": 4.729928732798809e-06,
+      "loss": 0.4894,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5763500931098696,
+      "grad_norm": 0.6554106714608859,
+      "learning_rate": 4.729348350980785e-06,
+      "loss": 0.4813,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5768156424581006,
+      "grad_norm": 0.6392150007392716,
+      "learning_rate": 4.728767381899394e-06,
+      "loss": 0.4572,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5772811918063314,
+      "grad_norm": 0.6308524579057007,
+      "learning_rate": 4.728185825707675e-06,
+      "loss": 0.4659,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5777467411545624,
+      "grad_norm": 0.6078489371026907,
+      "learning_rate": 4.727603682558823e-06,
+      "loss": 0.4608,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5782122905027933,
+      "grad_norm": 0.6297708501918517,
+      "learning_rate": 4.7270209526061906e-06,
+      "loss": 0.4376,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5786778398510242,
+      "grad_norm": 0.6073914983840603,
+      "learning_rate": 4.7264376360032805e-06,
+      "loss": 0.446,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5791433891992551,
+      "grad_norm": 0.6046736913920674,
+      "learning_rate": 4.725853732903754e-06,
+      "loss": 0.4459,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5796089385474861,
+      "grad_norm": 0.666475948933972,
+      "learning_rate": 4.725269243461423e-06,
+      "loss": 0.4584,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5800744878957169,
+      "grad_norm": 0.5844823568537969,
+      "learning_rate": 4.724684167830257e-06,
+      "loss": 0.4349,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5805400372439479,
+      "grad_norm": 0.6202871999725382,
+      "learning_rate": 4.72409850616438e-06,
+      "loss": 0.4553,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5810055865921788,
+      "grad_norm": 0.6286800642196431,
+      "learning_rate": 4.723512258618067e-06,
+      "loss": 0.4424,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5814711359404097,
+      "grad_norm": 0.7011815958735079,
+      "learning_rate": 4.722925425345752e-06,
+      "loss": 0.4507,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5819366852886406,
+      "grad_norm": 0.637755074669698,
+      "learning_rate": 4.722338006502018e-06,
+      "loss": 0.4539,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5824022346368715,
+      "grad_norm": 0.6538212916860098,
+      "learning_rate": 4.721750002241608e-06,
+      "loss": 0.4851,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5828677839851024,
+      "grad_norm": 0.5809472342630466,
+      "learning_rate": 4.721161412719414e-06,
+      "loss": 0.4178,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 0.6646773768099377,
+      "learning_rate": 4.720572238090485e-06,
+      "loss": 0.4599,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5837988826815642,
+      "grad_norm": 0.7254812738581334,
+      "learning_rate": 4.719982478510025e-06,
+      "loss": 0.4758,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5842644320297952,
+      "grad_norm": 0.5904463811139601,
+      "learning_rate": 4.719392134133389e-06,
+      "loss": 0.4485,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5847299813780261,
+      "grad_norm": 0.6252312550861696,
+      "learning_rate": 4.718801205116088e-06,
+      "loss": 0.4502,
+      "step": 1256
+    },
+    {
+      "epoch": 0.585195530726257,
+      "grad_norm": 0.6155100501969357,
+      "learning_rate": 4.718209691613788e-06,
+      "loss": 0.463,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5856610800744879,
+      "grad_norm": 0.6079423302957528,
+      "learning_rate": 4.717617593782306e-06,
+      "loss": 0.4475,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5861266294227188,
+      "grad_norm": 0.6381198499587649,
+      "learning_rate": 4.717024911777617e-06,
+      "loss": 0.4644,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5865921787709497,
+      "grad_norm": 0.65606321553362,
+      "learning_rate": 4.7164316457558445e-06,
+      "loss": 0.491,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5870577281191807,
+      "grad_norm": 0.6058836158254673,
+      "learning_rate": 4.715837795873271e-06,
+      "loss": 0.469,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5875232774674115,
+      "grad_norm": 0.6286557963074868,
+      "learning_rate": 4.715243362286331e-06,
+      "loss": 0.4648,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5879888268156425,
+      "grad_norm": 0.6608843335480807,
+      "learning_rate": 4.714648345151611e-06,
+      "loss": 0.4528,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5884543761638734,
+      "grad_norm": 0.6296239388868986,
+      "learning_rate": 4.714052744625853e-06,
+      "loss": 0.4615,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5889199255121043,
+      "grad_norm": 0.6623823266794809,
+      "learning_rate": 4.713456560865953e-06,
+      "loss": 0.4673,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5893854748603352,
+      "grad_norm": 0.6507645600043505,
+      "learning_rate": 4.712859794028959e-06,
+      "loss": 0.4608,
+      "step": 1266
+    },
+    {
+      "epoch": 0.589851024208566,
+      "grad_norm": 0.6194691533420319,
+      "learning_rate": 4.712262444272074e-06,
+      "loss": 0.4503,
+      "step": 1267
+    },
+    {
+      "epoch": 0.590316573556797,
+      "grad_norm": 0.5857038902927265,
+      "learning_rate": 4.711664511752655e-06,
+      "loss": 0.4447,
+      "step": 1268
+    },
+    {
+      "epoch": 0.590782122905028,
+      "grad_norm": 0.6446664619336238,
+      "learning_rate": 4.711065996628212e-06,
+      "loss": 0.453,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5912476722532588,
+      "grad_norm": 0.6408043560344777,
+      "learning_rate": 4.710466899056404e-06,
+      "loss": 0.4527,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5917132216014898,
+      "grad_norm": 0.6132635735958419,
+      "learning_rate": 4.709867219195052e-06,
+      "loss": 0.47,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5921787709497207,
+      "grad_norm": 0.5791024872799583,
+      "learning_rate": 4.709266957202125e-06,
+      "loss": 0.4407,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5926443202979516,
+      "grad_norm": 0.6547732099231949,
+      "learning_rate": 4.708666113235746e-06,
+      "loss": 0.4502,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5931098696461825,
+      "grad_norm": 0.6330305204460372,
+      "learning_rate": 4.708064687454191e-06,
+      "loss": 0.4474,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5935754189944135,
+      "grad_norm": 0.6129246228915791,
+      "learning_rate": 4.70746268001589e-06,
+      "loss": 0.4394,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5940409683426443,
+      "grad_norm": 0.6174383086142775,
+      "learning_rate": 4.706860091079426e-06,
+      "loss": 0.4583,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5945065176908753,
+      "grad_norm": 0.61987130951028,
+      "learning_rate": 4.706256920803537e-06,
+      "loss": 0.4468,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5949720670391061,
+      "grad_norm": 0.6559870193169687,
+      "learning_rate": 4.705653169347109e-06,
+      "loss": 0.4776,
+      "step": 1278
+    },
+    {
+      "epoch": 0.595437616387337,
+      "grad_norm": 0.6257105190859283,
+      "learning_rate": 4.705048836869188e-06,
+      "loss": 0.4479,
+      "step": 1279
+    },
+    {
+      "epoch": 0.595903165735568,
+      "grad_norm": 0.6152069266754636,
+      "learning_rate": 4.704443923528967e-06,
+      "loss": 0.4598,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5963687150837989,
+      "grad_norm": 0.6139372138999329,
+      "learning_rate": 4.703838429485796e-06,
+      "loss": 0.4584,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5968342644320298,
+      "grad_norm": 0.6623035386598503,
+      "learning_rate": 4.703232354899177e-06,
+      "loss": 0.4624,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5972998137802608,
+      "grad_norm": 0.6309417457661927,
+      "learning_rate": 4.702625699928762e-06,
+      "loss": 0.4418,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5977653631284916,
+      "grad_norm": 0.6305254232277467,
+      "learning_rate": 4.70201846473436e-06,
+      "loss": 0.4468,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5982309124767226,
+      "grad_norm": 0.642316698011433,
+      "learning_rate": 4.701410649475932e-06,
+      "loss": 0.463,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5986964618249534,
+      "grad_norm": 0.6061957211424545,
+      "learning_rate": 4.70080225431359e-06,
+      "loss": 0.4568,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5991620111731844,
+      "grad_norm": 0.6508741648016327,
+      "learning_rate": 4.700193279407599e-06,
+      "loss": 0.4435,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5996275605214153,
+      "grad_norm": 0.6143353994466992,
+      "learning_rate": 4.699583724918378e-06,
+      "loss": 0.4558,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6000931098696461,
+      "grad_norm": 0.6071515443983306,
+      "learning_rate": 4.698973591006501e-06,
+      "loss": 0.457,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6005586592178771,
+      "grad_norm": 0.640371321224654,
+      "learning_rate": 4.6983628778326864e-06,
+      "loss": 0.4691,
+      "step": 1290
+    },
+    {
+      "epoch": 0.601024208566108,
+      "grad_norm": 0.6369779849965985,
+      "learning_rate": 4.6977515855578146e-06,
+      "loss": 0.4619,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6014897579143389,
+      "grad_norm": 0.6188310902243358,
+      "learning_rate": 4.697139714342913e-06,
+      "loss": 0.4442,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6019553072625698,
+      "grad_norm": 0.6054881063266444,
+      "learning_rate": 4.6965272643491636e-06,
+      "loss": 0.4579,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6024208566108007,
+      "grad_norm": 0.6527580215140913,
+      "learning_rate": 4.6959142357379e-06,
+      "loss": 0.4627,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6028864059590316,
+      "grad_norm": 0.6167640846678931,
+      "learning_rate": 4.695300628670609e-06,
+      "loss": 0.4608,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6033519553072626,
+      "grad_norm": 0.639195694848929,
+      "learning_rate": 4.6946864433089275e-06,
+      "loss": 0.4675,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6038175046554934,
+      "grad_norm": 0.617608170811796,
+      "learning_rate": 4.694071679814649e-06,
+      "loss": 0.4302,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6042830540037244,
+      "grad_norm": 0.7022130249999666,
+      "learning_rate": 4.693456338349715e-06,
+      "loss": 0.4687,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6047486033519553,
+      "grad_norm": 0.6529914045319422,
+      "learning_rate": 4.692840419076222e-06,
+      "loss": 0.4721,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6052141527001862,
+      "grad_norm": 0.6893415523366596,
+      "learning_rate": 4.692223922156418e-06,
+      "loss": 0.488,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6056797020484171,
+      "grad_norm": 0.6242211233755117,
+      "learning_rate": 4.691606847752702e-06,
+      "loss": 0.4582,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6061452513966481,
+      "grad_norm": 0.5783904534131188,
+      "learning_rate": 4.6909891960276275e-06,
+      "loss": 0.4538,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6066108007448789,
+      "grad_norm": 0.6939526254232516,
+      "learning_rate": 4.690370967143897e-06,
+      "loss": 0.4563,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6070763500931099,
+      "grad_norm": 0.715485594189787,
+      "learning_rate": 4.6897521612643674e-06,
+      "loss": 0.4629,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6075418994413407,
+      "grad_norm": 0.63958997326545,
+      "learning_rate": 4.689132778552047e-06,
+      "loss": 0.4507,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6080074487895717,
+      "grad_norm": 0.6225026974527192,
+      "learning_rate": 4.688512819170095e-06,
+      "loss": 0.4727,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6084729981378026,
+      "grad_norm": 0.6208207196854802,
+      "learning_rate": 4.687892283281825e-06,
+      "loss": 0.4446,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6089385474860335,
+      "grad_norm": 0.6634339340901023,
+      "learning_rate": 4.6872711710507006e-06,
+      "loss": 0.461,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6094040968342644,
+      "grad_norm": 0.6143852738816725,
+      "learning_rate": 4.6866494826403364e-06,
+      "loss": 0.4405,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6098696461824954,
+      "grad_norm": 0.6955251911906477,
+      "learning_rate": 4.686027218214502e-06,
+      "loss": 0.4634,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6103351955307262,
+      "grad_norm": 0.6609488499883357,
+      "learning_rate": 4.685404377937114e-06,
+      "loss": 0.4704,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6108007448789572,
+      "grad_norm": 0.5791667814086793,
+      "learning_rate": 4.684780961972244e-06,
+      "loss": 0.4639,
+      "step": 1312
+    },
+    {
+      "epoch": 0.611266294227188,
+      "grad_norm": 0.6633283356755534,
+      "learning_rate": 4.684156970484116e-06,
+      "loss": 0.4428,
+      "step": 1313
+    },
+    {
+      "epoch": 0.611731843575419,
+      "grad_norm": 0.6420739510492356,
+      "learning_rate": 4.6835324036371035e-06,
+      "loss": 0.4716,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6121973929236499,
+      "grad_norm": 0.6019006937660961,
+      "learning_rate": 4.682907261595732e-06,
+      "loss": 0.4616,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6126629422718808,
+      "grad_norm": 0.6797656410833135,
+      "learning_rate": 4.682281544524679e-06,
+      "loss": 0.4954,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6131284916201117,
+      "grad_norm": 0.702603103460103,
+      "learning_rate": 4.681655252588774e-06,
+      "loss": 0.47,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6135940409683427,
+      "grad_norm": 0.6429405716063072,
+      "learning_rate": 4.681028385952996e-06,
+      "loss": 0.451,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6140595903165735,
+      "grad_norm": 0.6380166846466787,
+      "learning_rate": 4.6804009447824765e-06,
+      "loss": 0.465,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6145251396648045,
+      "grad_norm": 0.6209303134496597,
+      "learning_rate": 4.679772929242499e-06,
+      "loss": 0.447,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6149906890130353,
+      "grad_norm": 0.6204343483981415,
+      "learning_rate": 4.679144339498497e-06,
+      "loss": 0.4498,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6154562383612663,
+      "grad_norm": 0.6843558816277137,
+      "learning_rate": 4.678515175716058e-06,
+      "loss": 0.4735,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6159217877094972,
+      "grad_norm": 0.6342169171427405,
+      "learning_rate": 4.6778854380609165e-06,
+      "loss": 0.4397,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6163873370577281,
+      "grad_norm": 0.6141184494411762,
+      "learning_rate": 4.677255126698961e-06,
+      "loss": 0.4391,
+      "step": 1324
+    },
+    {
+      "epoch": 0.616852886405959,
+      "grad_norm": 0.7254540880492282,
+      "learning_rate": 4.676624241796231e-06,
+      "loss": 0.4733,
+      "step": 1325
+    },
+    {
+      "epoch": 0.61731843575419,
+      "grad_norm": 0.6202173645652218,
+      "learning_rate": 4.675992783518916e-06,
+      "loss": 0.4536,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6177839851024208,
+      "grad_norm": 0.5955694052129314,
+      "learning_rate": 4.675360752033358e-06,
+      "loss": 0.4488,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6182495344506518,
+      "grad_norm": 0.6046358166985646,
+      "learning_rate": 4.674728147506048e-06,
+      "loss": 0.4607,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6187150837988827,
+      "grad_norm": 0.624735558015132,
+      "learning_rate": 4.6740949701036294e-06,
+      "loss": 0.4681,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6191806331471136,
+      "grad_norm": 0.6342878218546344,
+      "learning_rate": 4.6734612199928965e-06,
+      "loss": 0.4728,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6196461824953445,
+      "grad_norm": 0.6296839615377358,
+      "learning_rate": 4.672826897340794e-06,
+      "loss": 0.4507,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6201117318435754,
+      "grad_norm": 0.6395577470060281,
+      "learning_rate": 4.672192002314418e-06,
+      "loss": 0.4651,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6205772811918063,
+      "grad_norm": 0.6950328309438776,
+      "learning_rate": 4.671556535081014e-06,
+      "loss": 0.4575,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6210428305400373,
+      "grad_norm": 0.6314894445844184,
+      "learning_rate": 4.6709204958079804e-06,
+      "loss": 0.4537,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6215083798882681,
+      "grad_norm": 0.6574821154491728,
+      "learning_rate": 4.670283884662864e-06,
+      "loss": 0.4602,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6219739292364991,
+      "grad_norm": 0.6507394729263953,
+      "learning_rate": 4.669646701813364e-06,
+      "loss": 0.4745,
+      "step": 1336
+    },
+    {
+      "epoch": 0.62243947858473,
+      "grad_norm": 0.6264580137146262,
+      "learning_rate": 4.669008947427329e-06,
+      "loss": 0.4657,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6229050279329609,
+      "grad_norm": 0.6085447437697249,
+      "learning_rate": 4.6683706216727586e-06,
+      "loss": 0.4522,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6233705772811918,
+      "grad_norm": 0.6908429748619673,
+      "learning_rate": 4.667731724717803e-06,
+      "loss": 0.4547,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6238361266294227,
+      "grad_norm": 0.6687105784135421,
+      "learning_rate": 4.667092256730765e-06,
+      "loss": 0.4507,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6243016759776536,
+      "grad_norm": 0.6395142835285098,
+      "learning_rate": 4.666452217880092e-06,
+      "loss": 0.4669,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6247672253258846,
+      "grad_norm": 0.6212638008753567,
+      "learning_rate": 4.6658116083343875e-06,
+      "loss": 0.4629,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6252327746741154,
+      "grad_norm": 0.5697451835572693,
+      "learning_rate": 4.665170428262403e-06,
+      "loss": 0.4435,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6256983240223464,
+      "grad_norm": 0.6292231211567642,
+      "learning_rate": 4.664528677833041e-06,
+      "loss": 0.4623,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6261638733705773,
+      "grad_norm": 0.6286000600674378,
+      "learning_rate": 4.663886357215353e-06,
+      "loss": 0.475,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6266294227188082,
+      "grad_norm": 0.6771323493258429,
+      "learning_rate": 4.6632434665785426e-06,
+      "loss": 0.442,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6270949720670391,
+      "grad_norm": 0.660286635467813,
+      "learning_rate": 4.662600006091961e-06,
+      "loss": 0.4678,
+      "step": 1347
+    },
+    {
+      "epoch": 0.62756052141527,
+      "grad_norm": 0.6036886177730679,
+      "learning_rate": 4.661955975925113e-06,
+      "loss": 0.4613,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6280260707635009,
+      "grad_norm": 0.6574326357244963,
+      "learning_rate": 4.661311376247649e-06,
+      "loss": 0.4093,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6284916201117319,
+      "grad_norm": 0.6390495678840302,
+      "learning_rate": 4.660666207229374e-06,
+      "loss": 0.4475,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6289571694599627,
+      "grad_norm": 0.6166461739986022,
+      "learning_rate": 4.660020469040241e-06,
+      "loss": 0.477,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6294227188081937,
+      "grad_norm": 0.6413842760658303,
+      "learning_rate": 4.6593741618503495e-06,
+      "loss": 0.4695,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6298882681564246,
+      "grad_norm": 0.6496971902447728,
+      "learning_rate": 4.658727285829956e-06,
+      "loss": 0.4616,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6303538175046555,
+      "grad_norm": 0.5927085995377865,
+      "learning_rate": 4.658079841149462e-06,
+      "loss": 0.4364,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6308193668528864,
+      "grad_norm": 0.6337256770060141,
+      "learning_rate": 4.657431827979418e-06,
+      "loss": 0.4697,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6312849162011173,
+      "grad_norm": 0.6547606819235585,
+      "learning_rate": 4.656783246490528e-06,
+      "loss": 0.4573,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6317504655493482,
+      "grad_norm": 0.6727889589845278,
+      "learning_rate": 4.656134096853644e-06,
+      "loss": 0.4594,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6322160148975792,
+      "grad_norm": 0.6147030506787781,
+      "learning_rate": 4.655484379239766e-06,
+      "loss": 0.4522,
+      "step": 1358
+    },
+    {
+      "epoch": 0.63268156424581,
+      "grad_norm": 0.6433921526840369,
+      "learning_rate": 4.654834093820045e-06,
+      "loss": 0.4421,
+      "step": 1359
+    },
+    {
+      "epoch": 0.633147113594041,
+      "grad_norm": 0.6333373734526426,
+      "learning_rate": 4.6541832407657835e-06,
+      "loss": 0.4589,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6336126629422719,
+      "grad_norm": 0.6060866339276634,
+      "learning_rate": 4.65353182024843e-06,
+      "loss": 0.4495,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6340782122905028,
+      "grad_norm": 0.6181268781457576,
+      "learning_rate": 4.652879832439586e-06,
+      "loss": 0.4555,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6345437616387337,
+      "grad_norm": 0.6475032553914516,
+      "learning_rate": 4.652227277510998e-06,
+      "loss": 0.4622,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6350093109869647,
+      "grad_norm": 0.6554169785180821,
+      "learning_rate": 4.651574155634566e-06,
+      "loss": 0.4459,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6354748603351955,
+      "grad_norm": 0.6679722332141554,
+      "learning_rate": 4.650920466982338e-06,
+      "loss": 0.4668,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6359404096834265,
+      "grad_norm": 0.6737047989189343,
+      "learning_rate": 4.650266211726511e-06,
+      "loss": 0.4603,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6364059590316573,
+      "grad_norm": 0.6001983228380583,
+      "learning_rate": 4.649611390039431e-06,
+      "loss": 0.489,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6368715083798883,
+      "grad_norm": 0.6735600311364232,
+      "learning_rate": 4.648956002093593e-06,
+      "loss": 0.4686,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6373370577281192,
+      "grad_norm": 0.734491973947864,
+      "learning_rate": 4.648300048061643e-06,
+      "loss": 0.4493,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6378026070763501,
+      "grad_norm": 0.6268556814832238,
+      "learning_rate": 4.647643528116374e-06,
+      "loss": 0.4559,
+      "step": 1370
+    },
+    {
+      "epoch": 0.638268156424581,
+      "grad_norm": 0.623236316240112,
+      "learning_rate": 4.646986442430731e-06,
+      "loss": 0.4631,
+      "step": 1371
+    },
+    {
+      "epoch": 0.638733705772812,
+      "grad_norm": 0.741613778213252,
+      "learning_rate": 4.646328791177803e-06,
+      "loss": 0.458,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6391992551210428,
+      "grad_norm": 0.636650252119821,
+      "learning_rate": 4.645670574530833e-06,
+      "loss": 0.447,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6396648044692738,
+      "grad_norm": 0.613705561803619,
+      "learning_rate": 4.64501179266321e-06,
+      "loss": 0.4746,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6401303538175046,
+      "grad_norm": 0.6509821979045428,
+      "learning_rate": 4.644352445748473e-06,
+      "loss": 0.4382,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6405959031657356,
+      "grad_norm": 0.6791850491177673,
+      "learning_rate": 4.64369253396031e-06,
+      "loss": 0.4471,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6410614525139665,
+      "grad_norm": 0.7003862057938174,
+      "learning_rate": 4.643032057472557e-06,
+      "loss": 0.4503,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6415270018621974,
+      "grad_norm": 0.679557444592612,
+      "learning_rate": 4.642371016459199e-06,
+      "loss": 0.4454,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6419925512104283,
+      "grad_norm": 0.6169021217008905,
+      "learning_rate": 4.641709411094371e-06,
+      "loss": 0.4639,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6424581005586593,
+      "grad_norm": 0.6601092573631575,
+      "learning_rate": 4.641047241552355e-06,
+      "loss": 0.4469,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6429236499068901,
+      "grad_norm": 0.6796880542994572,
+      "learning_rate": 4.6403845080075816e-06,
+      "loss": 0.4726,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6433891992551211,
+      "grad_norm": 0.6816698182002072,
+      "learning_rate": 4.639721210634632e-06,
+      "loss": 0.4891,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6438547486033519,
+      "grad_norm": 0.7050837928684983,
+      "learning_rate": 4.639057349608233e-06,
+      "loss": 0.467,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6443202979515829,
+      "grad_norm": 0.6246565783903887,
+      "learning_rate": 4.638392925103263e-06,
+      "loss": 0.4388,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6447858472998138,
+      "grad_norm": 0.6254951186717878,
+      "learning_rate": 4.637727937294744e-06,
+      "loss": 0.4694,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6452513966480447,
+      "grad_norm": 0.6680790280770922,
+      "learning_rate": 4.637062386357853e-06,
+      "loss": 0.4561,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6457169459962756,
+      "grad_norm": 0.601337335377994,
+      "learning_rate": 4.636396272467911e-06,
+      "loss": 0.4499,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6461824953445066,
+      "grad_norm": 0.6386407085668507,
+      "learning_rate": 4.635729595800388e-06,
+      "loss": 0.4451,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6466480446927374,
+      "grad_norm": 0.635942757286654,
+      "learning_rate": 4.635062356530903e-06,
+      "loss": 0.4522,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6471135940409684,
+      "grad_norm": 0.6135882192863278,
+      "learning_rate": 4.634394554835221e-06,
+      "loss": 0.4399,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6475791433891993,
+      "grad_norm": 0.5697188111655511,
+      "learning_rate": 4.633726190889258e-06,
+      "loss": 0.4729,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6480446927374302,
+      "grad_norm": 0.592272043007675,
+      "learning_rate": 4.6330572648690775e-06,
+      "loss": 0.4568,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6485102420856611,
+      "grad_norm": 0.6209672677018321,
+      "learning_rate": 4.63238777695089e-06,
+      "loss": 0.4771,
+      "step": 1393
+    },
+    {
+      "epoch": 0.648975791433892,
+      "grad_norm": 0.6795098597339252,
+      "learning_rate": 4.631717727311055e-06,
+      "loss": 0.4739,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6494413407821229,
+      "grad_norm": 0.6158440724117644,
+      "learning_rate": 4.631047116126079e-06,
+      "loss": 0.4701,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6499068901303539,
+      "grad_norm": 0.6219494280317508,
+      "learning_rate": 4.630375943572616e-06,
+      "loss": 0.4725,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6503724394785847,
+      "grad_norm": 0.6376130175127319,
+      "learning_rate": 4.62970420982747e-06,
+      "loss": 0.4729,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6508379888268156,
+      "grad_norm": 0.5575124334048182,
+      "learning_rate": 4.629031915067592e-06,
+      "loss": 0.4399,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6513035381750466,
+      "grad_norm": 0.6389713923817938,
+      "learning_rate": 4.62835905947008e-06,
+      "loss": 0.4484,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6517690875232774,
+      "grad_norm": 0.6289218612663521,
+      "learning_rate": 4.6276856432121795e-06,
+      "loss": 0.4788,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6522346368715084,
+      "grad_norm": 0.6562908907622022,
+      "learning_rate": 4.6270116664712846e-06,
+      "loss": 0.4576,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6527001862197392,
+      "grad_norm": 0.6226479824576934,
+      "learning_rate": 4.626337129424937e-06,
+      "loss": 0.4529,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6531657355679702,
+      "grad_norm": 0.6629961549633652,
+      "learning_rate": 4.6256620322508276e-06,
+      "loss": 0.4641,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6536312849162011,
+      "grad_norm": 0.6352109185810781,
+      "learning_rate": 4.624986375126789e-06,
+      "loss": 0.4686,
+      "step": 1404
+    },
+    {
+      "epoch": 0.654096834264432,
+      "grad_norm": 0.6227881272019218,
+      "learning_rate": 4.624310158230809e-06,
+      "loss": 0.4439,
+      "step": 1405
+    },
+    {
+      "epoch": 0.654562383612663,
+      "grad_norm": 0.6294182319866858,
+      "learning_rate": 4.623633381741018e-06,
+      "loss": 0.4553,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6550279329608939,
+      "grad_norm": 0.6275866726623328,
+      "learning_rate": 4.622956045835694e-06,
+      "loss": 0.4646,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6554934823091247,
+      "grad_norm": 0.6673422358323398,
+      "learning_rate": 4.6222781506932655e-06,
+      "loss": 0.4421,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6559590316573557,
+      "grad_norm": 0.6202243341489292,
+      "learning_rate": 4.6215996964923045e-06,
+      "loss": 0.4604,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6564245810055865,
+      "grad_norm": 0.6170156636987876,
+      "learning_rate": 4.620920683411532e-06,
+      "loss": 0.4553,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6568901303538175,
+      "grad_norm": 0.6326155658704142,
+      "learning_rate": 4.620241111629817e-06,
+      "loss": 0.445,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6573556797020484,
+      "grad_norm": 0.6001747552939201,
+      "learning_rate": 4.619560981326174e-06,
+      "loss": 0.4588,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6578212290502793,
+      "grad_norm": 0.6243679177345168,
+      "learning_rate": 4.618880292679767e-06,
+      "loss": 0.4414,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6582867783985102,
+      "grad_norm": 0.6407925080276731,
+      "learning_rate": 4.6181990458699044e-06,
+      "loss": 0.4453,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6587523277467412,
+      "grad_norm": 0.6428887650862052,
+      "learning_rate": 4.617517241076043e-06,
+      "loss": 0.4573,
+      "step": 1415
+    },
+    {
+      "epoch": 0.659217877094972,
+      "grad_norm": 0.5731359498183629,
+      "learning_rate": 4.616834878477786e-06,
+      "loss": 0.4642,
+      "step": 1416
+    },
+    {
+      "epoch": 0.659683426443203,
+      "grad_norm": 0.5938043791968518,
+      "learning_rate": 4.6161519582548845e-06,
+      "loss": 0.4621,
+      "step": 1417
+    },
+    {
+      "epoch": 0.660148975791434,
+      "grad_norm": 0.6038538793223667,
+      "learning_rate": 4.615468480587235e-06,
+      "loss": 0.462,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6606145251396648,
+      "grad_norm": 0.661206596433322,
+      "learning_rate": 4.614784445654884e-06,
+      "loss": 0.4445,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6610800744878957,
+      "grad_norm": 0.6072623352799922,
+      "learning_rate": 4.61409985363802e-06,
+      "loss": 0.4668,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6615456238361266,
+      "grad_norm": 0.6254993761412769,
+      "learning_rate": 4.613414704716982e-06,
+      "loss": 0.4382,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6620111731843575,
+      "grad_norm": 0.6881218347163,
+      "learning_rate": 4.612728999072254e-06,
+      "loss": 0.4536,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6624767225325885,
+      "grad_norm": 0.7236767047804453,
+      "learning_rate": 4.6120427368844685e-06,
+      "loss": 0.455,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6629422718808193,
+      "grad_norm": 0.633875876173579,
+      "learning_rate": 4.6113559183344035e-06,
+      "loss": 0.4613,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6634078212290503,
+      "grad_norm": 0.6053576537367039,
+      "learning_rate": 4.61066854360298e-06,
+      "loss": 0.4741,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6638733705772812,
+      "grad_norm": 0.632033459314569,
+      "learning_rate": 4.609980612871272e-06,
+      "loss": 0.464,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6643389199255121,
+      "grad_norm": 0.6212273243425901,
+      "learning_rate": 4.609292126320497e-06,
+      "loss": 0.4725,
+      "step": 1427
+    },
+    {
+      "epoch": 0.664804469273743,
+      "grad_norm": 0.6271861141877657,
+      "learning_rate": 4.608603084132016e-06,
+      "loss": 0.4697,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6652700186219739,
+      "grad_norm": 0.5843520857633653,
+      "learning_rate": 4.607913486487341e-06,
+      "loss": 0.4531,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6657355679702048,
+      "grad_norm": 0.6713203812751943,
+      "learning_rate": 4.607223333568128e-06,
+      "loss": 0.4485,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6662011173184358,
+      "grad_norm": 0.5967927020475347,
+      "learning_rate": 4.60653262555618e-06,
+      "loss": 0.4435,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.6470418472805233,
+      "learning_rate": 4.605841362633447e-06,
+      "loss": 0.4495,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6671322160148976,
+      "grad_norm": 0.597639826196871,
+      "learning_rate": 4.6051495449820215e-06,
+      "loss": 0.4603,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6675977653631285,
+      "grad_norm": 0.6212729783147667,
+      "learning_rate": 4.6044571727841465e-06,
+      "loss": 0.4435,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6680633147113594,
+      "grad_norm": 0.5915917236008019,
+      "learning_rate": 4.603764246222209e-06,
+      "loss": 0.4414,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6685288640595903,
+      "grad_norm": 0.6093978686433713,
+      "learning_rate": 4.603070765478741e-06,
+      "loss": 0.4681,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6689944134078212,
+      "grad_norm": 0.6129102019444536,
+      "learning_rate": 4.602376730736424e-06,
+      "loss": 0.4493,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6694599627560521,
+      "grad_norm": 0.6114040167889849,
+      "learning_rate": 4.601682142178082e-06,
+      "loss": 0.4761,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6699255121042831,
+      "grad_norm": 0.6310473787243082,
+      "learning_rate": 4.600986999986687e-06,
+      "loss": 0.4463,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6703910614525139,
+      "grad_norm": 0.5907047377038447,
+      "learning_rate": 4.6002913043453544e-06,
+      "loss": 0.4513,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6708566108007449,
+      "grad_norm": 0.657184260063503,
+      "learning_rate": 4.599595055437349e-06,
+      "loss": 0.4441,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6713221601489758,
+      "grad_norm": 0.6330169129471134,
+      "learning_rate": 4.598898253446078e-06,
+      "loss": 0.466,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6717877094972067,
+      "grad_norm": 0.5950502498421253,
+      "learning_rate": 4.598200898555095e-06,
+      "loss": 0.4196,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6722532588454376,
+      "grad_norm": 0.6140696163462285,
+      "learning_rate": 4.597502990948101e-06,
+      "loss": 0.4514,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6727188081936686,
+      "grad_norm": 0.6333178856778301,
+      "learning_rate": 4.596804530808943e-06,
+      "loss": 0.4546,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6731843575418994,
+      "grad_norm": 0.6125196309334658,
+      "learning_rate": 4.596105518321608e-06,
+      "loss": 0.455,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6736499068901304,
+      "grad_norm": 0.6187777788352221,
+      "learning_rate": 4.595405953670236e-06,
+      "loss": 0.4497,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6741154562383612,
+      "grad_norm": 0.6448198688394354,
+      "learning_rate": 4.594705837039108e-06,
+      "loss": 0.4617,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6745810055865922,
+      "grad_norm": 0.6350074015044287,
+      "learning_rate": 4.594005168612649e-06,
+      "loss": 0.4532,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6750465549348231,
+      "grad_norm": 0.6186033499913858,
+      "learning_rate": 4.593303948575436e-06,
+      "loss": 0.4655,
+      "step": 1450
+    },
+    {
+      "epoch": 0.675512104283054,
+      "grad_norm": 0.6223301687945506,
+      "learning_rate": 4.592602177112184e-06,
+      "loss": 0.4643,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6759776536312849,
+      "grad_norm": 0.646283838325953,
+      "learning_rate": 4.591899854407756e-06,
+      "loss": 0.4565,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6764432029795159,
+      "grad_norm": 0.6281536589210466,
+      "learning_rate": 4.5911969806471635e-06,
+      "loss": 0.4507,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6769087523277467,
+      "grad_norm": 0.6068748660052273,
+      "learning_rate": 4.590493556015557e-06,
+      "loss": 0.4632,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6773743016759777,
+      "grad_norm": 0.6622847430010408,
+      "learning_rate": 4.589789580698238e-06,
+      "loss": 0.4645,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6778398510242085,
+      "grad_norm": 0.6288728918828311,
+      "learning_rate": 4.5890850548806475e-06,
+      "loss": 0.4725,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6783054003724395,
+      "grad_norm": 0.6465117217546712,
+      "learning_rate": 4.5883799787483765e-06,
+      "loss": 0.4551,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6787709497206704,
+      "grad_norm": 0.6335749366197827,
+      "learning_rate": 4.587674352487158e-06,
+      "loss": 0.4564,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6792364990689013,
+      "grad_norm": 0.6935216887876474,
+      "learning_rate": 4.58696817628287e-06,
+      "loss": 0.4603,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6797020484171322,
+      "grad_norm": 0.612632021854194,
+      "learning_rate": 4.586261450321538e-06,
+      "loss": 0.4264,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6801675977653632,
+      "grad_norm": 0.6036044105947811,
+      "learning_rate": 4.585554174789329e-06,
+      "loss": 0.459,
+      "step": 1461
+    },
+    {
+      "epoch": 0.680633147113594,
+      "grad_norm": 0.6180356922237538,
+      "learning_rate": 4.584846349872556e-06,
+      "loss": 0.4544,
+      "step": 1462
+    },
+    {
+      "epoch": 0.681098696461825,
+      "grad_norm": 0.6157232489789518,
+      "learning_rate": 4.584137975757678e-06,
+      "loss": 0.459,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6815642458100558,
+      "grad_norm": 0.6425707220975451,
+      "learning_rate": 4.583429052631298e-06,
+      "loss": 0.4316,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6820297951582868,
+      "grad_norm": 0.6926505147524119,
+      "learning_rate": 4.58271958068016e-06,
+      "loss": 0.4634,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6824953445065177,
+      "grad_norm": 0.6207574911300954,
+      "learning_rate": 4.58200956009116e-06,
+      "loss": 0.4558,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6829608938547486,
+      "grad_norm": 0.6084928598484677,
+      "learning_rate": 4.581298991051331e-06,
+      "loss": 0.452,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6834264432029795,
+      "grad_norm": 0.6044050957184218,
+      "learning_rate": 4.580587873747856e-06,
+      "loss": 0.4703,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6838919925512105,
+      "grad_norm": 0.6417059213058935,
+      "learning_rate": 4.579876208368059e-06,
+      "loss": 0.4393,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6843575418994413,
+      "grad_norm": 0.5995758042731036,
+      "learning_rate": 4.5791639950994096e-06,
+      "loss": 0.4311,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6848230912476723,
+      "grad_norm": 0.6700601498841262,
+      "learning_rate": 4.5784512341295226e-06,
+      "loss": 0.4435,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6852886405959032,
+      "grad_norm": 0.6155203136219803,
+      "learning_rate": 4.577737925646155e-06,
+      "loss": 0.4538,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6857541899441341,
+      "grad_norm": 0.6889470033546843,
+      "learning_rate": 4.57702406983721e-06,
+      "loss": 0.4664,
+      "step": 1473
+    },
+    {
+      "epoch": 0.686219739292365,
+      "grad_norm": 0.6498144685115433,
+      "learning_rate": 4.576309666890735e-06,
+      "loss": 0.4512,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6866852886405959,
+      "grad_norm": 0.6692026281633254,
+      "learning_rate": 4.575594716994918e-06,
+      "loss": 0.4651,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6871508379888268,
+      "grad_norm": 0.6575220500616599,
+      "learning_rate": 4.574879220338096e-06,
+      "loss": 0.4473,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6876163873370578,
+      "grad_norm": 0.6036988798812007,
+      "learning_rate": 4.574163177108748e-06,
+      "loss": 0.4659,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6880819366852886,
+      "grad_norm": 0.60434961045253,
+      "learning_rate": 4.573446587495496e-06,
+      "loss": 0.4704,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6885474860335196,
+      "grad_norm": 0.6533801956471739,
+      "learning_rate": 4.572729451687107e-06,
+      "loss": 0.4583,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6890130353817505,
+      "grad_norm": 0.6711702327792058,
+      "learning_rate": 4.572011769872492e-06,
+      "loss": 0.4631,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6894785847299814,
+      "grad_norm": 0.6062126670396872,
+      "learning_rate": 4.571293542240704e-06,
+      "loss": 0.4576,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6899441340782123,
+      "grad_norm": 0.6821174705904024,
+      "learning_rate": 4.570574768980943e-06,
+      "loss": 0.4481,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6904096834264432,
+      "grad_norm": 0.6570983144933045,
+      "learning_rate": 4.56985545028255e-06,
+      "loss": 0.4597,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6908752327746741,
+      "grad_norm": 0.6486995878866532,
+      "learning_rate": 4.569135586335011e-06,
+      "loss": 0.4672,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6913407821229051,
+      "grad_norm": 0.6402150530351998,
+      "learning_rate": 4.568415177327955e-06,
+      "loss": 0.4524,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6918063314711359,
+      "grad_norm": 0.6493272715461044,
+      "learning_rate": 4.567694223451155e-06,
+      "loss": 0.4756,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6922718808193669,
+      "grad_norm": 0.6459658698287775,
+      "learning_rate": 4.566972724894528e-06,
+      "loss": 0.4694,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6927374301675978,
+      "grad_norm": 0.6421661986439344,
+      "learning_rate": 4.566250681848132e-06,
+      "loss": 0.4433,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6932029795158287,
+      "grad_norm": 0.6532143357448965,
+      "learning_rate": 4.565528094502173e-06,
+      "loss": 0.4627,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6936685288640596,
+      "grad_norm": 0.6094885900656264,
+      "learning_rate": 4.564804963046996e-06,
+      "loss": 0.4324,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6941340782122905,
+      "grad_norm": 0.6306111123933191,
+      "learning_rate": 4.564081287673091e-06,
+      "loss": 0.4581,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6945996275605214,
+      "grad_norm": 0.5972561267050885,
+      "learning_rate": 4.5633570685710915e-06,
+      "loss": 0.463,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6950651769087524,
+      "grad_norm": 0.6086062196321059,
+      "learning_rate": 4.562632305931774e-06,
+      "loss": 0.4416,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6955307262569832,
+      "grad_norm": 0.5944675408631049,
+      "learning_rate": 4.561906999946059e-06,
+      "loss": 0.4647,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6959962756052142,
+      "grad_norm": 0.6493493713971286,
+      "learning_rate": 4.561181150805008e-06,
+      "loss": 0.4548,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6964618249534451,
+      "grad_norm": 0.6308205895794278,
+      "learning_rate": 4.560454758699828e-06,
+      "loss": 0.4504,
+      "step": 1496
+    },
+    {
+      "epoch": 0.696927374301676,
+      "grad_norm": 0.5976274736457617,
+      "learning_rate": 4.559727823821868e-06,
+      "loss": 0.4469,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6973929236499069,
+      "grad_norm": 0.6459524123525178,
+      "learning_rate": 4.559000346362618e-06,
+      "loss": 0.4459,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6978584729981379,
+      "grad_norm": 0.6400094987312115,
+      "learning_rate": 4.558272326513716e-06,
+      "loss": 0.4589,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6983240223463687,
+      "grad_norm": 0.6182236287151086,
+      "learning_rate": 4.557543764466938e-06,
+      "loss": 0.46,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6987895716945997,
+      "grad_norm": 0.5911218884177076,
+      "learning_rate": 4.556814660414204e-06,
+      "loss": 0.436,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6992551210428305,
+      "grad_norm": 0.6213286972075303,
+      "learning_rate": 4.556085014547578e-06,
+      "loss": 0.4536,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6997206703910615,
+      "grad_norm": 0.6213985503169623,
+      "learning_rate": 4.555354827059267e-06,
+      "loss": 0.4483,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7001862197392924,
+      "grad_norm": 0.5691716364860002,
+      "learning_rate": 4.554624098141618e-06,
+      "loss": 0.4515,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7006517690875232,
+      "grad_norm": 0.6289137725953262,
+      "learning_rate": 4.553892827987124e-06,
+      "loss": 0.4285,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7011173184357542,
+      "grad_norm": 0.6983242743230622,
+      "learning_rate": 4.553161016788417e-06,
+      "loss": 0.4369,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7015828677839852,
+      "grad_norm": 0.6678092394795032,
+      "learning_rate": 4.552428664738276e-06,
+      "loss": 0.4642,
+      "step": 1507
+    },
+    {
+      "epoch": 0.702048417132216,
+      "grad_norm": 0.6141049688841526,
+      "learning_rate": 4.551695772029618e-06,
+      "loss": 0.4507,
+      "step": 1508
+    },
+    {
+      "epoch": 0.702513966480447,
+      "grad_norm": 0.578865346551071,
+      "learning_rate": 4.550962338855505e-06,
+      "loss": 0.4645,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7029795158286778,
+      "grad_norm": 0.5809031901710795,
+      "learning_rate": 4.550228365409141e-06,
+      "loss": 0.4166,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7034450651769087,
+      "grad_norm": 0.6516489631514671,
+      "learning_rate": 4.5494938518838705e-06,
+      "loss": 0.4576,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7039106145251397,
+      "grad_norm": 0.6363250461253412,
+      "learning_rate": 4.548758798473184e-06,
+      "loss": 0.4642,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7043761638733705,
+      "grad_norm": 0.6089696896476706,
+      "learning_rate": 4.548023205370711e-06,
+      "loss": 0.4464,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7048417132216015,
+      "grad_norm": 0.6006689820724969,
+      "learning_rate": 4.547287072770224e-06,
+      "loss": 0.4574,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7053072625698324,
+      "grad_norm": 0.6044752391346596,
+      "learning_rate": 4.54655040086564e-06,
+      "loss": 0.4481,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7057728119180633,
+      "grad_norm": 0.6225554617230171,
+      "learning_rate": 4.545813189851012e-06,
+      "loss": 0.4481,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7062383612662942,
+      "grad_norm": 0.6561691612605753,
+      "learning_rate": 4.545075439920542e-06,
+      "loss": 0.4542,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7067039106145251,
+      "grad_norm": 0.5876331757576423,
+      "learning_rate": 4.54433715126857e-06,
+      "loss": 0.4506,
+      "step": 1518
+    },
+    {
+      "epoch": 0.707169459962756,
+      "grad_norm": 0.58379574542077,
+      "learning_rate": 4.543598324089577e-06,
+      "loss": 0.4495,
+      "step": 1519
+    },
+    {
+      "epoch": 0.707635009310987,
+      "grad_norm": 0.6212299803759317,
+      "learning_rate": 4.542858958578191e-06,
+      "loss": 0.4535,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7081005586592178,
+      "grad_norm": 0.5874483088478114,
+      "learning_rate": 4.5421190549291765e-06,
+      "loss": 0.4591,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7085661080074488,
+      "grad_norm": 0.63183225478042,
+      "learning_rate": 4.541378613337442e-06,
+      "loss": 0.4576,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7090316573556797,
+      "grad_norm": 0.6259975101559755,
+      "learning_rate": 4.5406376339980376e-06,
+      "loss": 0.4778,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7094972067039106,
+      "grad_norm": 0.5732050010400295,
+      "learning_rate": 4.539896117106154e-06,
+      "loss": 0.4499,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7099627560521415,
+      "grad_norm": 0.5963015318089787,
+      "learning_rate": 4.5391540628571255e-06,
+      "loss": 0.4555,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7104283054003724,
+      "grad_norm": 0.6141680519351512,
+      "learning_rate": 4.5384114714464256e-06,
+      "loss": 0.4405,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7108938547486033,
+      "grad_norm": 0.5865515324551687,
+      "learning_rate": 4.5376683430696725e-06,
+      "loss": 0.4479,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7113594040968343,
+      "grad_norm": 0.5926353197387066,
+      "learning_rate": 4.5369246779226215e-06,
+      "loss": 0.4766,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7118249534450651,
+      "grad_norm": 0.6239788768992308,
+      "learning_rate": 4.536180476201174e-06,
+      "loss": 0.4368,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7122905027932961,
+      "grad_norm": 0.6139439507495891,
+      "learning_rate": 4.535435738101369e-06,
+      "loss": 0.4369,
+      "step": 1530
+    },
+    {
+      "epoch": 0.712756052141527,
+      "grad_norm": 0.7285130615901765,
+      "learning_rate": 4.534690463819388e-06,
+      "loss": 0.446,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7132216014897579,
+      "grad_norm": 0.6508107924488056,
+      "learning_rate": 4.533944653551555e-06,
+      "loss": 0.4605,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7136871508379888,
+      "grad_norm": 0.598929716701949,
+      "learning_rate": 4.533198307494334e-06,
+      "loss": 0.4487,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7141527001862198,
+      "grad_norm": 0.6190216822283837,
+      "learning_rate": 4.53245142584433e-06,
+      "loss": 0.4561,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7146182495344506,
+      "grad_norm": 0.6046860620242067,
+      "learning_rate": 4.531704008798289e-06,
+      "loss": 0.4407,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7150837988826816,
+      "grad_norm": 0.6540633782208415,
+      "learning_rate": 4.530956056553098e-06,
+      "loss": 0.4359,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7155493482309124,
+      "grad_norm": 0.6706613627325443,
+      "learning_rate": 4.5302075693057875e-06,
+      "loss": 0.4548,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7160148975791434,
+      "grad_norm": 0.6372252317331645,
+      "learning_rate": 4.5294585472535255e-06,
+      "loss": 0.4547,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7164804469273743,
+      "grad_norm": 0.6076783714229193,
+      "learning_rate": 4.528708990593621e-06,
+      "loss": 0.4342,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7169459962756052,
+      "grad_norm": 0.6261782303641648,
+      "learning_rate": 4.527958899523528e-06,
+      "loss": 0.4594,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7174115456238361,
+      "grad_norm": 0.5987684528578318,
+      "learning_rate": 4.527208274240835e-06,
+      "loss": 0.4358,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7178770949720671,
+      "grad_norm": 0.6635918418059569,
+      "learning_rate": 4.526457114943277e-06,
+      "loss": 0.4595,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7183426443202979,
+      "grad_norm": 0.6133550915030701,
+      "learning_rate": 4.525705421828727e-06,
+      "loss": 0.4512,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7188081936685289,
+      "grad_norm": 0.6312967255998547,
+      "learning_rate": 4.524953195095198e-06,
+      "loss": 0.4806,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7192737430167597,
+      "grad_norm": 0.6083300941175687,
+      "learning_rate": 4.524200434940844e-06,
+      "loss": 0.4417,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7197392923649907,
+      "grad_norm": 0.5866720984139175,
+      "learning_rate": 4.52344714156396e-06,
+      "loss": 0.4557,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7202048417132216,
+      "grad_norm": 0.6389953706525204,
+      "learning_rate": 4.522693315162984e-06,
+      "loss": 0.4815,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7206703910614525,
+      "grad_norm": 0.5834898118090258,
+      "learning_rate": 4.521938955936488e-06,
+      "loss": 0.4734,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7211359404096834,
+      "grad_norm": 0.6016150428025172,
+      "learning_rate": 4.52118406408319e-06,
+      "loss": 0.4449,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7216014897579144,
+      "grad_norm": 0.6370404901153996,
+      "learning_rate": 4.520428639801947e-06,
+      "loss": 0.4486,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7220670391061452,
+      "grad_norm": 0.6127809734290122,
+      "learning_rate": 4.519672683291756e-06,
+      "loss": 0.4579,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7225325884543762,
+      "grad_norm": 0.6442427141826185,
+      "learning_rate": 4.518916194751751e-06,
+      "loss": 0.4738,
+      "step": 1552
+    },
+    {
+      "epoch": 0.722998137802607,
+      "grad_norm": 0.5800279187628676,
+      "learning_rate": 4.518159174381213e-06,
+      "loss": 0.454,
+      "step": 1553
+    },
+    {
+      "epoch": 0.723463687150838,
+      "grad_norm": 0.6073838717287422,
+      "learning_rate": 4.517401622379557e-06,
+      "loss": 0.4384,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7239292364990689,
+      "grad_norm": 0.6452369151361437,
+      "learning_rate": 4.516643538946339e-06,
+      "loss": 0.4335,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7243947858472998,
+      "grad_norm": 0.6565295458533179,
+      "learning_rate": 4.515884924281258e-06,
+      "loss": 0.4471,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7248603351955307,
+      "grad_norm": 0.6383540551694179,
+      "learning_rate": 4.515125778584152e-06,
+      "loss": 0.4521,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7253258845437617,
+      "grad_norm": 0.6176634137500076,
+      "learning_rate": 4.5143661020549966e-06,
+      "loss": 0.4632,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7257914338919925,
+      "grad_norm": 0.6182649649154249,
+      "learning_rate": 4.5136058948939085e-06,
+      "loss": 0.4493,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7262569832402235,
+      "grad_norm": 0.728546941087777,
+      "learning_rate": 4.512845157301145e-06,
+      "loss": 0.4503,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7267225325884544,
+      "grad_norm": 0.6362024806124175,
+      "learning_rate": 4.512083889477102e-06,
+      "loss": 0.4503,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7271880819366853,
+      "grad_norm": 0.6274764706847753,
+      "learning_rate": 4.511322091622316e-06,
+      "loss": 0.4583,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7276536312849162,
+      "grad_norm": 0.5927636988175695,
+      "learning_rate": 4.5105597639374625e-06,
+      "loss": 0.4518,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7281191806331471,
+      "grad_norm": 0.6357650250511669,
+      "learning_rate": 4.509796906623357e-06,
+      "loss": 0.4624,
+      "step": 1564
+    },
+    {
+      "epoch": 0.728584729981378,
+      "grad_norm": 0.627057779465511,
+      "learning_rate": 4.509033519880953e-06,
+      "loss": 0.4639,
+      "step": 1565
+    },
+    {
+      "epoch": 0.729050279329609,
+      "grad_norm": 0.6885126316165354,
+      "learning_rate": 4.508269603911346e-06,
+      "loss": 0.4566,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7295158286778398,
+      "grad_norm": 0.6658586503048138,
+      "learning_rate": 4.50750515891577e-06,
+      "loss": 0.4587,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7299813780260708,
+      "grad_norm": 0.6773269144925671,
+      "learning_rate": 4.506740185095597e-06,
+      "loss": 0.4502,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7304469273743017,
+      "grad_norm": 0.6540111182103675,
+      "learning_rate": 4.505974682652339e-06,
+      "loss": 0.446,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7309124767225326,
+      "grad_norm": 0.5973875048091001,
+      "learning_rate": 4.505208651787648e-06,
+      "loss": 0.4513,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7313780260707635,
+      "grad_norm": 0.6402234493894431,
+      "learning_rate": 4.504442092703316e-06,
+      "loss": 0.4635,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7318435754189944,
+      "grad_norm": 0.6280056189637372,
+      "learning_rate": 4.503675005601271e-06,
+      "loss": 0.4623,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7323091247672253,
+      "grad_norm": 0.6434603807076872,
+      "learning_rate": 4.502907390683583e-06,
+      "loss": 0.4639,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7327746741154563,
+      "grad_norm": 0.6078616344414993,
+      "learning_rate": 4.50213924815246e-06,
+      "loss": 0.4387,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7332402234636871,
+      "grad_norm": 0.5927776455254404,
+      "learning_rate": 4.5013705782102476e-06,
+      "loss": 0.4643,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7337057728119181,
+      "grad_norm": 0.5928006454626441,
+      "learning_rate": 4.500601381059434e-06,
+      "loss": 0.4387,
+      "step": 1576
+    },
+    {
+      "epoch": 0.734171322160149,
+      "grad_norm": 0.6182493669177985,
+      "learning_rate": 4.499831656902643e-06,
+      "loss": 0.4624,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7346368715083799,
+      "grad_norm": 0.6086887442382125,
+      "learning_rate": 4.499061405942639e-06,
+      "loss": 0.4523,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7351024208566108,
+      "grad_norm": 0.5974781998121903,
+      "learning_rate": 4.498290628382322e-06,
+      "loss": 0.4574,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7355679702048417,
+      "grad_norm": 0.6051615241369099,
+      "learning_rate": 4.497519324424735e-06,
+      "loss": 0.4554,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7360335195530726,
+      "grad_norm": 0.6510920986010081,
+      "learning_rate": 4.496747494273058e-06,
+      "loss": 0.4531,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7364990689013036,
+      "grad_norm": 0.6074713726636868,
+      "learning_rate": 4.495975138130609e-06,
+      "loss": 0.4689,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7369646182495344,
+      "grad_norm": 0.6231262600487805,
+      "learning_rate": 4.4952022562008445e-06,
+      "loss": 0.4374,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7374301675977654,
+      "grad_norm": 0.5722599714188367,
+      "learning_rate": 4.4944288486873615e-06,
+      "loss": 0.4652,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7378957169459963,
+      "grad_norm": 0.6015440160525088,
+      "learning_rate": 4.493654915793892e-06,
+      "loss": 0.456,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7383612662942272,
+      "grad_norm": 0.5942521661263738,
+      "learning_rate": 4.492880457724309e-06,
+      "loss": 0.4432,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7388268156424581,
+      "grad_norm": 0.6543669320971646,
+      "learning_rate": 4.492105474682625e-06,
+      "loss": 0.4484,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7392923649906891,
+      "grad_norm": 0.6522376903441123,
+      "learning_rate": 4.4913299668729845e-06,
+      "loss": 0.4613,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7397579143389199,
+      "grad_norm": 0.6030869166188872,
+      "learning_rate": 4.49055393449968e-06,
+      "loss": 0.4558,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7402234636871509,
+      "grad_norm": 0.6925091658201907,
+      "learning_rate": 4.4897773777671325e-06,
+      "loss": 0.4693,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7406890130353817,
+      "grad_norm": 0.6933145706722915,
+      "learning_rate": 4.489000296879909e-06,
+      "loss": 0.4579,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7411545623836127,
+      "grad_norm": 0.6016567256237618,
+      "learning_rate": 4.4882226920427085e-06,
+      "loss": 0.4731,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7416201117318436,
+      "grad_norm": 0.6745445864604722,
+      "learning_rate": 4.487444563460372e-06,
+      "loss": 0.4485,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7420856610800745,
+      "grad_norm": 0.6848261267048017,
+      "learning_rate": 4.486665911337877e-06,
+      "loss": 0.4571,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7425512104283054,
+      "grad_norm": 0.6247412716331107,
+      "learning_rate": 4.485886735880338e-06,
+      "loss": 0.4426,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7430167597765364,
+      "grad_norm": 0.6187575446003535,
+      "learning_rate": 4.4851070372930095e-06,
+      "loss": 0.4481,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7434823091247672,
+      "grad_norm": 0.5912784470117102,
+      "learning_rate": 4.484326815781281e-06,
+      "loss": 0.4394,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7439478584729982,
+      "grad_norm": 0.6200363668746407,
+      "learning_rate": 4.483546071550684e-06,
+      "loss": 0.4257,
+      "step": 1598
+    },
+    {
+      "epoch": 0.744413407821229,
+      "grad_norm": 0.601234674192418,
+      "learning_rate": 4.482764804806883e-06,
+      "loss": 0.458,
+      "step": 1599
+    },
+    {
+      "epoch": 0.74487895716946,
+      "grad_norm": 0.6205980665162832,
+      "learning_rate": 4.481983015755683e-06,
+      "loss": 0.4333,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7453445065176909,
+      "grad_norm": 0.6628114777850789,
+      "learning_rate": 4.481200704603026e-06,
+      "loss": 0.4636,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7458100558659218,
+      "grad_norm": 0.6412900665270993,
+      "learning_rate": 4.480417871554991e-06,
+      "loss": 0.4713,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7462756052141527,
+      "grad_norm": 0.5921579866872017,
+      "learning_rate": 4.479634516817795e-06,
+      "loss": 0.455,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7467411545623837,
+      "grad_norm": 0.619072196434505,
+      "learning_rate": 4.478850640597792e-06,
+      "loss": 0.4713,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7472067039106145,
+      "grad_norm": 0.6404717581603446,
+      "learning_rate": 4.478066243101473e-06,
+      "loss": 0.449,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7476722532588455,
+      "grad_norm": 0.6532809972955499,
+      "learning_rate": 4.477281324535469e-06,
+      "loss": 0.4692,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7481378026070763,
+      "grad_norm": 0.6225254712167666,
+      "learning_rate": 4.4764958851065445e-06,
+      "loss": 0.4472,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7486033519553073,
+      "grad_norm": 0.5800683696673175,
+      "learning_rate": 4.475709925021603e-06,
+      "loss": 0.4355,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7490689013035382,
+      "grad_norm": 0.6024499135009621,
+      "learning_rate": 4.474923444487685e-06,
+      "loss": 0.4464,
+      "step": 1609
+    },
+    {
+      "epoch": 0.749534450651769,
+      "grad_norm": 0.6036586074086892,
+      "learning_rate": 4.47413644371197e-06,
+      "loss": 0.4546,
+      "step": 1610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.642397793965458,
+      "learning_rate": 4.473348922901771e-06,
+      "loss": 0.441,
+      "step": 1611
+    },
+    {
+      "epoch": 0.750465549348231,
+      "grad_norm": 0.6806536063360468,
+      "learning_rate": 4.472560882264541e-06,
+      "loss": 0.4452,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7509310986964618,
+      "grad_norm": 0.5900390227449652,
+      "learning_rate": 4.471772322007866e-06,
+      "loss": 0.4427,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7513966480446927,
+      "grad_norm": 0.6145909735665207,
+      "learning_rate": 4.470983242339474e-06,
+      "loss": 0.4689,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7518621973929237,
+      "grad_norm": 0.6386486094022614,
+      "learning_rate": 4.470193643467227e-06,
+      "loss": 0.4596,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7523277467411545,
+      "grad_norm": 0.6235877911866973,
+      "learning_rate": 4.469403525599123e-06,
+      "loss": 0.4567,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7527932960893855,
+      "grad_norm": 0.6196019599991636,
+      "learning_rate": 4.468612888943298e-06,
+      "loss": 0.4907,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7532588454376163,
+      "grad_norm": 0.5897836985605567,
+      "learning_rate": 4.467821733708026e-06,
+      "loss": 0.4635,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7537243947858473,
+      "grad_norm": 0.6070994204892002,
+      "learning_rate": 4.467030060101715e-06,
+      "loss": 0.4561,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7541899441340782,
+      "grad_norm": 0.6325029756062818,
+      "learning_rate": 4.466237868332911e-06,
+      "loss": 0.4438,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7546554934823091,
+      "grad_norm": 0.626176897288853,
+      "learning_rate": 4.465445158610297e-06,
+      "loss": 0.4453,
+      "step": 1621
+    },
+    {
+      "epoch": 0.75512104283054,
+      "grad_norm": 0.6213400587951801,
+      "learning_rate": 4.464651931142689e-06,
+      "loss": 0.4786,
+      "step": 1622
+    },
+    {
+      "epoch": 0.755586592178771,
+      "grad_norm": 0.5983355054967282,
+      "learning_rate": 4.4638581861390425e-06,
+      "loss": 0.4246,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7560521415270018,
+      "grad_norm": 0.5880799404065172,
+      "learning_rate": 4.463063923808452e-06,
+      "loss": 0.4347,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7565176908752328,
+      "grad_norm": 0.6474856800546467,
+      "learning_rate": 4.462269144360142e-06,
+      "loss": 0.4541,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7569832402234636,
+      "grad_norm": 0.6314012781075757,
+      "learning_rate": 4.4614738480034775e-06,
+      "loss": 0.458,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7574487895716946,
+      "grad_norm": 0.5968582905055606,
+      "learning_rate": 4.460678034947959e-06,
+      "loss": 0.4534,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7579143389199255,
+      "grad_norm": 0.6470473013017238,
+      "learning_rate": 4.459881705403219e-06,
+      "loss": 0.4352,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7583798882681564,
+      "grad_norm": 0.6468217758960803,
+      "learning_rate": 4.459084859579035e-06,
+      "loss": 0.4651,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7588454376163873,
+      "grad_norm": 0.6109856847867328,
+      "learning_rate": 4.4582874976853115e-06,
+      "loss": 0.4617,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7593109869646183,
+      "grad_norm": 0.5952099919224507,
+      "learning_rate": 4.457489619932094e-06,
+      "loss": 0.4419,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7597765363128491,
+      "grad_norm": 0.6253063508946968,
+      "learning_rate": 4.456691226529563e-06,
+      "loss": 0.4632,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7602420856610801,
+      "grad_norm": 0.6530720118082691,
+      "learning_rate": 4.455892317688032e-06,
+      "loss": 0.4552,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7607076350093109,
+      "grad_norm": 0.6119194073309658,
+      "learning_rate": 4.455092893617955e-06,
+      "loss": 0.4694,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7611731843575419,
+      "grad_norm": 0.6307278091548539,
+      "learning_rate": 4.454292954529918e-06,
+      "loss": 0.4253,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7616387337057728,
+      "grad_norm": 0.6082601549678214,
+      "learning_rate": 4.453492500634644e-06,
+      "loss": 0.4365,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7621042830540037,
+      "grad_norm": 0.6653149937522248,
+      "learning_rate": 4.452691532142992e-06,
+      "loss": 0.4301,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7625698324022346,
+      "grad_norm": 0.6021963321592186,
+      "learning_rate": 4.451890049265956e-06,
+      "loss": 0.453,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7630353817504656,
+      "grad_norm": 0.6352767869187491,
+      "learning_rate": 4.451088052214665e-06,
+      "loss": 0.4864,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7635009310986964,
+      "grad_norm": 0.6776130476886011,
+      "learning_rate": 4.450285541200386e-06,
+      "loss": 0.4397,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7639664804469274,
+      "grad_norm": 0.6023283005867275,
+      "learning_rate": 4.449482516434518e-06,
+      "loss": 0.4605,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7644320297951583,
+      "grad_norm": 0.646194466381472,
+      "learning_rate": 4.448678978128597e-06,
+      "loss": 0.4667,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7648975791433892,
+      "grad_norm": 0.6471097578775568,
+      "learning_rate": 4.447874926494294e-06,
+      "loss": 0.4595,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7653631284916201,
+      "grad_norm": 0.6185558276706056,
+      "learning_rate": 4.447070361743415e-06,
+      "loss": 0.4219,
+      "step": 1644
+    },
+    {
+      "epoch": 0.765828677839851,
+      "grad_norm": 0.6551327483223297,
+      "learning_rate": 4.4462652840879025e-06,
+      "loss": 0.4624,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7662942271880819,
+      "grad_norm": 0.5972384631693491,
+      "learning_rate": 4.445459693739833e-06,
+      "loss": 0.4403,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7667597765363129,
+      "grad_norm": 0.6075503899304543,
+      "learning_rate": 4.444653590911417e-06,
+      "loss": 0.4483,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7672253258845437,
+      "grad_norm": 0.657689014329737,
+      "learning_rate": 4.443846975815003e-06,
+      "loss": 0.4571,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7676908752327747,
+      "grad_norm": 0.6051066449590474,
+      "learning_rate": 4.443039848663071e-06,
+      "loss": 0.4567,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7681564245810056,
+      "grad_norm": 0.6270857516421009,
+      "learning_rate": 4.442232209668238e-06,
+      "loss": 0.4609,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7686219739292365,
+      "grad_norm": 0.5812148701802032,
+      "learning_rate": 4.441424059043254e-06,
+      "loss": 0.4336,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7690875232774674,
+      "grad_norm": 0.6258186671772897,
+      "learning_rate": 4.440615397001008e-06,
+      "loss": 0.4384,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7695530726256983,
+      "grad_norm": 0.6099242617887485,
+      "learning_rate": 4.439806223754519e-06,
+      "loss": 0.4536,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7700186219739292,
+      "grad_norm": 0.6362345400222942,
+      "learning_rate": 4.438996539516942e-06,
+      "loss": 0.4619,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7704841713221602,
+      "grad_norm": 0.6246648908929086,
+      "learning_rate": 4.438186344501569e-06,
+      "loss": 0.4604,
+      "step": 1655
+    },
+    {
+      "epoch": 0.770949720670391,
+      "grad_norm": 0.6156190467878425,
+      "learning_rate": 4.437375638921822e-06,
+      "loss": 0.4565,
+      "step": 1656
+    },
+    {
+      "epoch": 0.771415270018622,
+      "grad_norm": 0.5861448841165354,
+      "learning_rate": 4.436564422991262e-06,
+      "loss": 0.4373,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7718808193668529,
+      "grad_norm": 0.6089504610594776,
+      "learning_rate": 4.435752696923582e-06,
+      "loss": 0.4602,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7723463687150838,
+      "grad_norm": 0.6167733635248662,
+      "learning_rate": 4.43494046093261e-06,
+      "loss": 0.4458,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7728119180633147,
+      "grad_norm": 0.6655249473021535,
+      "learning_rate": 4.434127715232308e-06,
+      "loss": 0.449,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7732774674115456,
+      "grad_norm": 0.6169945259897918,
+      "learning_rate": 4.433314460036774e-06,
+      "loss": 0.4545,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7737430167597765,
+      "grad_norm": 0.6188505079245957,
+      "learning_rate": 4.432500695560237e-06,
+      "loss": 0.4652,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7742085661080075,
+      "grad_norm": 0.5932176533473167,
+      "learning_rate": 4.431686422017062e-06,
+      "loss": 0.449,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7746741154562383,
+      "grad_norm": 0.6149629669301985,
+      "learning_rate": 4.43087163962175e-06,
+      "loss": 0.4528,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7751396648044693,
+      "grad_norm": 0.6385984303513973,
+      "learning_rate": 4.430056348588931e-06,
+      "loss": 0.4536,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7756052141527002,
+      "grad_norm": 0.600851698882398,
+      "learning_rate": 4.429240549133374e-06,
+      "loss": 0.4402,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7760707635009311,
+      "grad_norm": 0.6061385368418547,
+      "learning_rate": 4.42842424146998e-06,
+      "loss": 0.4668,
+      "step": 1667
+    },
+    {
+      "epoch": 0.776536312849162,
+      "grad_norm": 0.6139753668627268,
+      "learning_rate": 4.427607425813783e-06,
+      "loss": 0.4399,
+      "step": 1668
+    },
+    {
+      "epoch": 0.777001862197393,
+      "grad_norm": 0.6157504560449415,
+      "learning_rate": 4.426790102379952e-06,
+      "loss": 0.467,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7774674115456238,
+      "grad_norm": 0.6347909438366657,
+      "learning_rate": 4.42597227138379e-06,
+      "loss": 0.4435,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7779329608938548,
+      "grad_norm": 0.6522822807552915,
+      "learning_rate": 4.425153933040733e-06,
+      "loss": 0.4666,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7783985102420856,
+      "grad_norm": 0.6108770309343851,
+      "learning_rate": 4.424335087566351e-06,
+      "loss": 0.445,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7788640595903166,
+      "grad_norm": 0.6122590762650924,
+      "learning_rate": 4.423515735176346e-06,
+      "loss": 0.4394,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7793296089385475,
+      "grad_norm": 0.6022176164264043,
+      "learning_rate": 4.422695876086556e-06,
+      "loss": 0.4311,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7797951582867784,
+      "grad_norm": 0.632865236890516,
+      "learning_rate": 4.421875510512951e-06,
+      "loss": 0.4816,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7802607076350093,
+      "grad_norm": 0.6264258635910939,
+      "learning_rate": 4.421054638671634e-06,
+      "loss": 0.4557,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7807262569832403,
+      "grad_norm": 0.6470017124800459,
+      "learning_rate": 4.420233260778845e-06,
+      "loss": 0.4579,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7811918063314711,
+      "grad_norm": 0.584109450285105,
+      "learning_rate": 4.41941137705095e-06,
+      "loss": 0.4448,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7816573556797021,
+      "grad_norm": 0.6128006568873308,
+      "learning_rate": 4.418588987704457e-06,
+      "loss": 0.4638,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7821229050279329,
+      "grad_norm": 0.6403089077802973,
+      "learning_rate": 4.417766092956002e-06,
+      "loss": 0.4485,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7825884543761639,
+      "grad_norm": 0.6481500945251241,
+      "learning_rate": 4.416942693022352e-06,
+      "loss": 0.4516,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7830540037243948,
+      "grad_norm": 0.6208307061683566,
+      "learning_rate": 4.416118788120414e-06,
+      "loss": 0.448,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7835195530726257,
+      "grad_norm": 0.5842362319187616,
+      "learning_rate": 4.415294378467222e-06,
+      "loss": 0.4556,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7839851024208566,
+      "grad_norm": 0.5855179818714212,
+      "learning_rate": 4.414469464279946e-06,
+      "loss": 0.4566,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7844506517690876,
+      "grad_norm": 0.6161372917592242,
+      "learning_rate": 4.413644045775887e-06,
+      "loss": 0.4447,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7849162011173184,
+      "grad_norm": 0.6439268788223765,
+      "learning_rate": 4.412818123172481e-06,
+      "loss": 0.459,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7853817504655494,
+      "grad_norm": 0.5817458386689888,
+      "learning_rate": 4.411991696687295e-06,
+      "loss": 0.4573,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7858472998137802,
+      "grad_norm": 0.6154307405231313,
+      "learning_rate": 4.41116476653803e-06,
+      "loss": 0.4529,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7863128491620112,
+      "grad_norm": 0.6012006925269711,
+      "learning_rate": 4.410337332942518e-06,
+      "loss": 0.432,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7867783985102421,
+      "grad_norm": 0.6356400330885408,
+      "learning_rate": 4.409509396118725e-06,
+      "loss": 0.4588,
+      "step": 1690
+    },
+    {
+      "epoch": 0.787243947858473,
+      "grad_norm": 0.6130925703969043,
+      "learning_rate": 4.40868095628475e-06,
+      "loss": 0.4597,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7877094972067039,
+      "grad_norm": 0.6472348192753804,
+      "learning_rate": 4.407852013658823e-06,
+      "loss": 0.4631,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7881750465549349,
+      "grad_norm": 0.713563243556077,
+      "learning_rate": 4.407022568459308e-06,
+      "loss": 0.4573,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7886405959031657,
+      "grad_norm": 0.6126970956889409,
+      "learning_rate": 4.406192620904699e-06,
+      "loss": 0.4598,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7891061452513967,
+      "grad_norm": 0.6377459125586648,
+      "learning_rate": 4.405362171213626e-06,
+      "loss": 0.4377,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7895716945996276,
+      "grad_norm": 0.6227903455847454,
+      "learning_rate": 4.404531219604848e-06,
+      "loss": 0.4327,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7900372439478585,
+      "grad_norm": 0.6404436895398397,
+      "learning_rate": 4.403699766297257e-06,
+      "loss": 0.4217,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7905027932960894,
+      "grad_norm": 0.647405096232577,
+      "learning_rate": 4.402867811509879e-06,
+      "loss": 0.4367,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7909683426443203,
+      "grad_norm": 0.5991065947042978,
+      "learning_rate": 4.402035355461869e-06,
+      "loss": 0.4335,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7914338919925512,
+      "grad_norm": 0.6230038544710319,
+      "learning_rate": 4.401202398372517e-06,
+      "loss": 0.448,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7918994413407822,
+      "grad_norm": 0.6303072825926265,
+      "learning_rate": 4.400368940461244e-06,
+      "loss": 0.444,
+      "step": 1701
+    },
+    {
+      "epoch": 0.792364990689013,
+      "grad_norm": 0.6398898820638341,
+      "learning_rate": 4.399534981947601e-06,
+      "loss": 0.4532,
+      "step": 1702
+    },
+    {
+      "epoch": 0.792830540037244,
+      "grad_norm": 0.6421871444517112,
+      "learning_rate": 4.398700523051274e-06,
+      "loss": 0.4527,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7932960893854749,
+      "grad_norm": 0.623820317835018,
+      "learning_rate": 4.397865563992078e-06,
+      "loss": 0.4502,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7937616387337058,
+      "grad_norm": 0.7189900029820113,
+      "learning_rate": 4.3970301049899625e-06,
+      "loss": 0.4559,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7942271880819367,
+      "grad_norm": 0.5780027079085902,
+      "learning_rate": 4.396194146265006e-06,
+      "loss": 0.4343,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7946927374301676,
+      "grad_norm": 0.5838304864728459,
+      "learning_rate": 4.39535768803742e-06,
+      "loss": 0.4632,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7951582867783985,
+      "grad_norm": 0.5904232974988907,
+      "learning_rate": 4.394520730527549e-06,
+      "loss": 0.4378,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7956238361266295,
+      "grad_norm": 0.6242892850395138,
+      "learning_rate": 4.393683273955866e-06,
+      "loss": 0.4559,
+      "step": 1709
+    },
+    {
+      "epoch": 0.7960893854748603,
+      "grad_norm": 0.67931932973126,
+      "learning_rate": 4.392845318542978e-06,
+      "loss": 0.479,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7965549348230913,
+      "grad_norm": 0.6333883047396841,
+      "learning_rate": 4.392006864509621e-06,
+      "loss": 0.4322,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7970204841713222,
+      "grad_norm": 0.6021433085611275,
+      "learning_rate": 4.391167912076665e-06,
+      "loss": 0.4434,
+      "step": 1712
+    },
+    {
+      "epoch": 0.797486033519553,
+      "grad_norm": 0.6359384616651578,
+      "learning_rate": 4.390328461465109e-06,
+      "loss": 0.4443,
+      "step": 1713
+    },
+    {
+      "epoch": 0.797951582867784,
+      "grad_norm": 0.569977753507062,
+      "learning_rate": 4.389488512896085e-06,
+      "loss": 0.4343,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7984171322160148,
+      "grad_norm": 0.6093523508266672,
+      "learning_rate": 4.388648066590856e-06,
+      "loss": 0.4382,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7988826815642458,
+      "grad_norm": 0.6253992698411142,
+      "learning_rate": 4.387807122770814e-06,
+      "loss": 0.4423,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7993482309124768,
+      "grad_norm": 0.6238175478901843,
+      "learning_rate": 4.386965681657485e-06,
+      "loss": 0.4545,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7998137802607076,
+      "grad_norm": 0.6263642030828224,
+      "learning_rate": 4.386123743472524e-06,
+      "loss": 0.4331,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8002793296089385,
+      "grad_norm": 0.6751993700989362,
+      "learning_rate": 4.3852813084377175e-06,
+      "loss": 0.449,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8007448789571695,
+      "grad_norm": 0.58859555789158,
+      "learning_rate": 4.384438376774983e-06,
+      "loss": 0.4601,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8012104283054003,
+      "grad_norm": 0.6289135143084267,
+      "learning_rate": 4.383594948706368e-06,
+      "loss": 0.4421,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8016759776536313,
+      "grad_norm": 0.6202479583080812,
+      "learning_rate": 4.382751024454053e-06,
+      "loss": 0.4462,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8021415270018621,
+      "grad_norm": 0.6244258975888667,
+      "learning_rate": 4.381906604240346e-06,
+      "loss": 0.4445,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8026070763500931,
+      "grad_norm": 0.5975710444640409,
+      "learning_rate": 4.381061688287689e-06,
+      "loss": 0.4683,
+      "step": 1724
+    },
+    {
+      "epoch": 0.803072625698324,
+      "grad_norm": 0.6309055005752486,
+      "learning_rate": 4.380216276818652e-06,
+      "loss": 0.4535,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8035381750465549,
+      "grad_norm": 0.6508171958195068,
+      "learning_rate": 4.379370370055937e-06,
+      "loss": 0.4564,
+      "step": 1726
+    },
+    {
+      "epoch": 0.8040037243947858,
+      "grad_norm": 0.5835794537042276,
+      "learning_rate": 4.378523968222376e-06,
+      "loss": 0.4413,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8044692737430168,
+      "grad_norm": 0.5851877237611682,
+      "learning_rate": 4.37767707154093e-06,
+      "loss": 0.4412,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8049348230912476,
+      "grad_norm": 0.5935572726225521,
+      "learning_rate": 4.376829680234694e-06,
+      "loss": 0.4554,
+      "step": 1729
+    },
+    {
+      "epoch": 0.8054003724394786,
+      "grad_norm": 0.6072606066657357,
+      "learning_rate": 4.375981794526889e-06,
+      "loss": 0.4333,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8058659217877095,
+      "grad_norm": 0.5813360918143734,
+      "learning_rate": 4.375133414640869e-06,
+      "loss": 0.4396,
+      "step": 1731
+    },
+    {
+      "epoch": 0.8063314711359404,
+      "grad_norm": 0.591396485091668,
+      "learning_rate": 4.3742845408001175e-06,
+      "loss": 0.4528,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8067970204841713,
+      "grad_norm": 0.6382500778955114,
+      "learning_rate": 4.373435173228249e-06,
+      "loss": 0.4489,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8072625698324022,
+      "grad_norm": 0.5944047384755891,
+      "learning_rate": 4.3725853121490045e-06,
+      "loss": 0.4404,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8077281191806331,
+      "grad_norm": 0.5918885390110288,
+      "learning_rate": 4.3717349577862605e-06,
+      "loss": 0.4467,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8081936685288641,
+      "grad_norm": 0.6244155805395547,
+      "learning_rate": 4.370884110364019e-06,
+      "loss": 0.4449,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8086592178770949,
+      "grad_norm": 0.587350395986124,
+      "learning_rate": 4.370032770106414e-06,
+      "loss": 0.4546,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8091247672253259,
+      "grad_norm": 0.6472262795553687,
+      "learning_rate": 4.369180937237709e-06,
+      "loss": 0.4619,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8095903165735568,
+      "grad_norm": 0.610187907479253,
+      "learning_rate": 4.3683286119822955e-06,
+      "loss": 0.4399,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8100558659217877,
+      "grad_norm": 0.611269926214896,
+      "learning_rate": 4.3674757945646976e-06,
+      "loss": 0.4652,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8105214152700186,
+      "grad_norm": 0.5896563856134243,
+      "learning_rate": 4.366622485209568e-06,
+      "loss": 0.4449,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8109869646182495,
+      "grad_norm": 0.6218039008286244,
+      "learning_rate": 4.365768684141688e-06,
+      "loss": 0.453,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8114525139664804,
+      "grad_norm": 0.6601978989834907,
+      "learning_rate": 4.364914391585969e-06,
+      "loss": 0.4699,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8119180633147114,
+      "grad_norm": 0.6689840271446245,
+      "learning_rate": 4.364059607767452e-06,
+      "loss": 0.4629,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8123836126629422,
+      "grad_norm": 0.603084450427529,
+      "learning_rate": 4.3632043329113075e-06,
+      "loss": 0.446,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8128491620111732,
+      "grad_norm": 0.6230623112070284,
+      "learning_rate": 4.362348567242835e-06,
+      "loss": 0.4468,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8133147113594041,
+      "grad_norm": 0.6357545584843707,
+      "learning_rate": 4.361492310987465e-06,
+      "loss": 0.4404,
+      "step": 1747
+    },
+    {
+      "epoch": 0.813780260707635,
+      "grad_norm": 0.5713553680205491,
+      "learning_rate": 4.360635564370754e-06,
+      "loss": 0.4601,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8142458100558659,
+      "grad_norm": 0.5919454395359237,
+      "learning_rate": 4.35977832761839e-06,
+      "loss": 0.4511,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8147113594040968,
+      "grad_norm": 0.6601120204233901,
+      "learning_rate": 4.358920600956189e-06,
+      "loss": 0.458,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8151769087523277,
+      "grad_norm": 0.5976642915942892,
+      "learning_rate": 4.358062384610098e-06,
+      "loss": 0.4434,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8156424581005587,
+      "grad_norm": 0.6977255247543199,
+      "learning_rate": 4.35720367880619e-06,
+      "loss": 0.482,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8161080074487895,
+      "grad_norm": 0.6244674184240739,
+      "learning_rate": 4.356344483770669e-06,
+      "loss": 0.4576,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8165735567970205,
+      "grad_norm": 0.6622321580434667,
+      "learning_rate": 4.3554847997298685e-06,
+      "loss": 0.4642,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8170391061452514,
+      "grad_norm": 0.6244923697383659,
+      "learning_rate": 4.3546246269102495e-06,
+      "loss": 0.4416,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8175046554934823,
+      "grad_norm": 0.586201287648749,
+      "learning_rate": 4.3537639655384e-06,
+      "loss": 0.4536,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8179702048417132,
+      "grad_norm": 0.5860481248339019,
+      "learning_rate": 4.352902815841041e-06,
+      "loss": 0.4635,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8184357541899442,
+      "grad_norm": 0.5914360661546066,
+      "learning_rate": 4.352041178045019e-06,
+      "loss": 0.466,
+      "step": 1758
+    },
+    {
+      "epoch": 0.818901303538175,
+      "grad_norm": 0.6479530271802175,
+      "learning_rate": 4.351179052377309e-06,
+      "loss": 0.4445,
+      "step": 1759
+    },
+    {
+      "epoch": 0.819366852886406,
+      "grad_norm": 0.6235921928413678,
+      "learning_rate": 4.350316439065016e-06,
+      "loss": 0.4762,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8198324022346368,
+      "grad_norm": 0.6561961365396338,
+      "learning_rate": 4.349453338335374e-06,
+      "loss": 0.4465,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8202979515828678,
+      "grad_norm": 0.6050601707812917,
+      "learning_rate": 4.348589750415743e-06,
+      "loss": 0.4339,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8207635009310987,
+      "grad_norm": 0.6227527284693957,
+      "learning_rate": 4.347725675533612e-06,
+      "loss": 0.4606,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8212290502793296,
+      "grad_norm": 0.6268555528211757,
+      "learning_rate": 4.3468611139166e-06,
+      "loss": 0.4505,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8216945996275605,
+      "grad_norm": 0.5910972531978342,
+      "learning_rate": 4.345996065792453e-06,
+      "loss": 0.474,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8221601489757915,
+      "grad_norm": 0.6847059443922806,
+      "learning_rate": 4.345130531389045e-06,
+      "loss": 0.4635,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8226256983240223,
+      "grad_norm": 0.6337513693766312,
+      "learning_rate": 4.344264510934379e-06,
+      "loss": 0.4414,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8230912476722533,
+      "grad_norm": 0.6466448881310796,
+      "learning_rate": 4.343398004656584e-06,
+      "loss": 0.4432,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8235567970204841,
+      "grad_norm": 0.7530949149456304,
+      "learning_rate": 4.3425310127839194e-06,
+      "loss": 0.4578,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8240223463687151,
+      "grad_norm": 0.6353149763803244,
+      "learning_rate": 4.341663535544771e-06,
+      "loss": 0.4469,
+      "step": 1770
+    },
+    {
+      "epoch": 0.824487895716946,
+      "grad_norm": 0.6599334923049609,
+      "learning_rate": 4.340795573167654e-06,
+      "loss": 0.4443,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8249534450651769,
+      "grad_norm": 0.6676597546913451,
+      "learning_rate": 4.339927125881209e-06,
+      "loss": 0.4501,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8254189944134078,
+      "grad_norm": 0.6116040077743841,
+      "learning_rate": 4.339058193914206e-06,
+      "loss": 0.4253,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8258845437616388,
+      "grad_norm": 0.703459747719486,
+      "learning_rate": 4.338188777495542e-06,
+      "loss": 0.4528,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8263500931098696,
+      "grad_norm": 0.6459496628116007,
+      "learning_rate": 4.337318876854243e-06,
+      "loss": 0.4349,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8268156424581006,
+      "grad_norm": 0.6378400636549293,
+      "learning_rate": 4.336448492219462e-06,
+      "loss": 0.4652,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8272811918063314,
+      "grad_norm": 0.5656071078990436,
+      "learning_rate": 4.335577623820476e-06,
+      "loss": 0.4276,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8277467411545624,
+      "grad_norm": 0.614921063230537,
+      "learning_rate": 4.334706271886696e-06,
+      "loss": 0.4375,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8282122905027933,
+      "grad_norm": 0.6363266705627771,
+      "learning_rate": 4.333834436647654e-06,
+      "loss": 0.4482,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8286778398510242,
+      "grad_norm": 0.5853066165775855,
+      "learning_rate": 4.332962118333014e-06,
+      "loss": 0.4513,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8291433891992551,
+      "grad_norm": 0.6366786536685473,
+      "learning_rate": 4.332089317172566e-06,
+      "loss": 0.4469,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8296089385474861,
+      "grad_norm": 0.6062786230632675,
+      "learning_rate": 4.331216033396223e-06,
+      "loss": 0.4451,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8300744878957169,
+      "grad_norm": 0.6848382242889092,
+      "learning_rate": 4.330342267234033e-06,
+      "loss": 0.4555,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8305400372439479,
+      "grad_norm": 0.6150633767149212,
+      "learning_rate": 4.329468018916163e-06,
+      "loss": 0.4425,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8310055865921788,
+      "grad_norm": 0.6338325317733304,
+      "learning_rate": 4.328593288672914e-06,
+      "loss": 0.4551,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8314711359404097,
+      "grad_norm": 0.604720874736086,
+      "learning_rate": 4.327718076734709e-06,
+      "loss": 0.4624,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8319366852886406,
+      "grad_norm": 0.6166235053926764,
+      "learning_rate": 4.3268423833320995e-06,
+      "loss": 0.4393,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8324022346368715,
+      "grad_norm": 0.6178517868721008,
+      "learning_rate": 4.3259662086957656e-06,
+      "loss": 0.4758,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8328677839851024,
+      "grad_norm": 0.6167945992511485,
+      "learning_rate": 4.325089553056511e-06,
+      "loss": 0.4495,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.6285460175614471,
+      "learning_rate": 4.324212416645267e-06,
+      "loss": 0.4573,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8337988826815642,
+      "grad_norm": 0.632608878813843,
+      "learning_rate": 4.323334799693094e-06,
+      "loss": 0.463,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8342644320297952,
+      "grad_norm": 0.6518971137463367,
+      "learning_rate": 4.3224567024311774e-06,
+      "loss": 0.4553,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8347299813780261,
+      "grad_norm": 0.5948557158868556,
+      "learning_rate": 4.321578125090827e-06,
+      "loss": 0.4502,
+      "step": 1793
+    },
+    {
+      "epoch": 0.835195530726257,
+      "grad_norm": 0.6299616875215642,
+      "learning_rate": 4.320699067903482e-06,
+      "loss": 0.4288,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8356610800744879,
+      "grad_norm": 0.5872607881176437,
+      "learning_rate": 4.3198195311007075e-06,
+      "loss": 0.4504,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8361266294227188,
+      "grad_norm": 0.5867993703696585,
+      "learning_rate": 4.318939514914193e-06,
+      "loss": 0.4418,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8365921787709497,
+      "grad_norm": 0.5949348588235428,
+      "learning_rate": 4.318059019575757e-06,
+      "loss": 0.4674,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8370577281191807,
+      "grad_norm": 0.6240609737504457,
+      "learning_rate": 4.3171780453173415e-06,
+      "loss": 0.4267,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8375232774674115,
+      "grad_norm": 0.6036546699874551,
+      "learning_rate": 4.316296592371017e-06,
+      "loss": 0.4522,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8379888268156425,
+      "grad_norm": 0.5971311098446692,
+      "learning_rate": 4.31541466096898e-06,
+      "loss": 0.462,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8384543761638734,
+      "grad_norm": 0.5745531831490462,
+      "learning_rate": 4.314532251343551e-06,
+      "loss": 0.4535,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8389199255121043,
+      "grad_norm": 0.6157156174441237,
+      "learning_rate": 4.313649363727177e-06,
+      "loss": 0.4299,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8393854748603352,
+      "grad_norm": 0.5861227087869527,
+      "learning_rate": 4.312765998352434e-06,
+      "loss": 0.4559,
+      "step": 1803
+    },
+    {
+      "epoch": 0.839851024208566,
+      "grad_norm": 0.6190081862146654,
+      "learning_rate": 4.311882155452018e-06,
+      "loss": 0.4534,
+      "step": 1804
+    },
+    {
+      "epoch": 0.840316573556797,
+      "grad_norm": 0.5892819174446567,
+      "learning_rate": 4.310997835258759e-06,
+      "loss": 0.4533,
+      "step": 1805
+    },
+    {
+      "epoch": 0.840782122905028,
+      "grad_norm": 0.6206886139071368,
+      "learning_rate": 4.310113038005604e-06,
+      "loss": 0.4598,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8412476722532588,
+      "grad_norm": 0.6418209216787121,
+      "learning_rate": 4.30922776392563e-06,
+      "loss": 0.4611,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8417132216014898,
+      "grad_norm": 0.5955395004728082,
+      "learning_rate": 4.308342013252041e-06,
+      "loss": 0.4422,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8421787709497207,
+      "grad_norm": 0.5912266986719438,
+      "learning_rate": 4.307455786218164e-06,
+      "loss": 0.418,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8426443202979516,
+      "grad_norm": 0.6690501933091115,
+      "learning_rate": 4.306569083057452e-06,
+      "loss": 0.4513,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8431098696461825,
+      "grad_norm": 0.6077183288076048,
+      "learning_rate": 4.3056819040034836e-06,
+      "loss": 0.4507,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8435754189944135,
+      "grad_norm": 0.6339816843207547,
+      "learning_rate": 4.304794249289963e-06,
+      "loss": 0.4561,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8440409683426443,
+      "grad_norm": 0.5787165386541564,
+      "learning_rate": 4.3039061191507205e-06,
+      "loss": 0.4439,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8445065176908753,
+      "grad_norm": 0.6079428753825755,
+      "learning_rate": 4.30301751381971e-06,
+      "loss": 0.4518,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8449720670391061,
+      "grad_norm": 0.6332087287704601,
+      "learning_rate": 4.30212843353101e-06,
+      "loss": 0.458,
+      "step": 1815
+    },
+    {
+      "epoch": 0.845437616387337,
+      "grad_norm": 0.5877622291873877,
+      "learning_rate": 4.301238878518827e-06,
+      "loss": 0.4357,
+      "step": 1816
+    },
+    {
+      "epoch": 0.845903165735568,
+      "grad_norm": 0.650793008490777,
+      "learning_rate": 4.30034884901749e-06,
+      "loss": 0.4306,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8463687150837989,
+      "grad_norm": 0.7064797331449562,
+      "learning_rate": 4.2994583452614554e-06,
+      "loss": 0.4532,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8468342644320298,
+      "grad_norm": 0.5847508245122853,
+      "learning_rate": 4.298567367485302e-06,
+      "loss": 0.4416,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8472998137802608,
+      "grad_norm": 0.5654947657235613,
+      "learning_rate": 4.297675915923735e-06,
+      "loss": 0.4534,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8477653631284916,
+      "grad_norm": 0.6046970777638769,
+      "learning_rate": 4.296783990811583e-06,
+      "loss": 0.4526,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8482309124767226,
+      "grad_norm": 0.6302765548944284,
+      "learning_rate": 4.295891592383801e-06,
+      "loss": 0.453,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8486964618249534,
+      "grad_norm": 0.6379848285936752,
+      "learning_rate": 4.294998720875467e-06,
+      "loss": 0.4401,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8491620111731844,
+      "grad_norm": 0.588922519352572,
+      "learning_rate": 4.294105376521787e-06,
+      "loss": 0.4406,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8496275605214153,
+      "grad_norm": 0.5942433713920421,
+      "learning_rate": 4.293211559558086e-06,
+      "loss": 0.4711,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8500931098696461,
+      "grad_norm": 0.9765503982571141,
+      "learning_rate": 4.292317270219819e-06,
+      "loss": 0.4397,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8505586592178771,
+      "grad_norm": 0.6404712986953999,
+      "learning_rate": 4.2914225087425625e-06,
+      "loss": 0.4351,
+      "step": 1827
+    },
+    {
+      "epoch": 0.851024208566108,
+      "grad_norm": 0.5826317953950368,
+      "learning_rate": 4.2905272753620166e-06,
+      "loss": 0.4493,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8514897579143389,
+      "grad_norm": 0.7925227454629159,
+      "learning_rate": 4.289631570314009e-06,
+      "loss": 0.4514,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8519553072625698,
+      "grad_norm": 0.5897572053315293,
+      "learning_rate": 4.288735393834489e-06,
+      "loss": 0.4411,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8524208566108007,
+      "grad_norm": 0.5934825099154837,
+      "learning_rate": 4.287838746159531e-06,
+      "loss": 0.4548,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8528864059590316,
+      "grad_norm": 0.5996861638935727,
+      "learning_rate": 4.286941627525332e-06,
+      "loss": 0.445,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8533519553072626,
+      "grad_norm": 0.5986390689878738,
+      "learning_rate": 4.2860440381682165e-06,
+      "loss": 0.4562,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8538175046554934,
+      "grad_norm": 0.6350877384295208,
+      "learning_rate": 4.285145978324629e-06,
+      "loss": 0.4755,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8542830540037244,
+      "grad_norm": 0.5471851653366328,
+      "learning_rate": 4.284247448231141e-06,
+      "loss": 0.4354,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8547486033519553,
+      "grad_norm": 0.6312431710619636,
+      "learning_rate": 4.283348448124446e-06,
+      "loss": 0.4291,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8552141527001862,
+      "grad_norm": 0.613244452239879,
+      "learning_rate": 4.282448978241362e-06,
+      "loss": 0.4436,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8556797020484171,
+      "grad_norm": 0.6274188284857346,
+      "learning_rate": 4.281549038818832e-06,
+      "loss": 0.4395,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8561452513966481,
+      "grad_norm": 0.6140904644063618,
+      "learning_rate": 4.28064863009392e-06,
+      "loss": 0.4675,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8566108007448789,
+      "grad_norm": 0.5764906871818898,
+      "learning_rate": 4.279747752303816e-06,
+      "loss": 0.4217,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8570763500931099,
+      "grad_norm": 0.57744269842736,
+      "learning_rate": 4.278846405685832e-06,
+      "loss": 0.4781,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8575418994413407,
+      "grad_norm": 0.6256358444681728,
+      "learning_rate": 4.277944590477403e-06,
+      "loss": 0.4586,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8580074487895717,
+      "grad_norm": 0.5957074064347515,
+      "learning_rate": 4.2770423069160905e-06,
+      "loss": 0.4593,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8584729981378026,
+      "grad_norm": 0.6066114106603888,
+      "learning_rate": 4.276139555239578e-06,
+      "loss": 0.4606,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8589385474860335,
+      "grad_norm": 0.6147198463639563,
+      "learning_rate": 4.275236335685669e-06,
+      "loss": 0.4586,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8594040968342644,
+      "grad_norm": 0.5912113861825036,
+      "learning_rate": 4.274332648492296e-06,
+      "loss": 0.4441,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8598696461824954,
+      "grad_norm": 0.5988161135223754,
+      "learning_rate": 4.2734284938975105e-06,
+      "loss": 0.46,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8603351955307262,
+      "grad_norm": 0.6277463253945906,
+      "learning_rate": 4.272523872139487e-06,
+      "loss": 0.4382,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8608007448789572,
+      "grad_norm": 0.6289721506964272,
+      "learning_rate": 4.271618783456526e-06,
+      "loss": 0.4397,
+      "step": 1849
+    },
+    {
+      "epoch": 0.861266294227188,
+      "grad_norm": 0.6041479767885163,
+      "learning_rate": 4.270713228087049e-06,
+      "loss": 0.4426,
+      "step": 1850
+    },
+    {
+      "epoch": 0.861731843575419,
+      "grad_norm": 0.6284884920425592,
+      "learning_rate": 4.269807206269601e-06,
+      "loss": 0.4638,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8621973929236499,
+      "grad_norm": 0.6108115239127135,
+      "learning_rate": 4.268900718242848e-06,
+      "loss": 0.4382,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8626629422718808,
+      "grad_norm": 0.6188899280601757,
+      "learning_rate": 4.267993764245583e-06,
+      "loss": 0.4628,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8631284916201117,
+      "grad_norm": 0.6017327735109989,
+      "learning_rate": 4.267086344516718e-06,
+      "loss": 0.4422,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8635940409683427,
+      "grad_norm": 0.6073594551995746,
+      "learning_rate": 4.266178459295289e-06,
+      "loss": 0.4388,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8640595903165735,
+      "grad_norm": 0.583234857995289,
+      "learning_rate": 4.265270108820455e-06,
+      "loss": 0.4358,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8645251396648045,
+      "grad_norm": 0.6437709428477506,
+      "learning_rate": 4.264361293331497e-06,
+      "loss": 0.4669,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8649906890130353,
+      "grad_norm": 0.6029349511430228,
+      "learning_rate": 4.263452013067818e-06,
+      "loss": 0.4492,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8654562383612663,
+      "grad_norm": 0.5792770436756591,
+      "learning_rate": 4.262542268268945e-06,
+      "loss": 0.4459,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8659217877094972,
+      "grad_norm": 0.619199383276732,
+      "learning_rate": 4.261632059174524e-06,
+      "loss": 0.4536,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8663873370577281,
+      "grad_norm": 0.6547307514737046,
+      "learning_rate": 4.260721386024329e-06,
+      "loss": 0.4599,
+      "step": 1861
+    },
+    {
+      "epoch": 0.866852886405959,
+      "grad_norm": 0.5917993859814602,
+      "learning_rate": 4.259810249058252e-06,
+      "loss": 0.4397,
+      "step": 1862
+    },
+    {
+      "epoch": 0.86731843575419,
+      "grad_norm": 0.6595187244761728,
+      "learning_rate": 4.2588986485163066e-06,
+      "loss": 0.4459,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8677839851024208,
+      "grad_norm": 0.5935758993659553,
+      "learning_rate": 4.257986584638631e-06,
+      "loss": 0.4624,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8682495344506518,
+      "grad_norm": 0.6302946089746101,
+      "learning_rate": 4.257074057665485e-06,
+      "loss": 0.4394,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8687150837988827,
+      "grad_norm": 0.6270381497483177,
+      "learning_rate": 4.256161067837249e-06,
+      "loss": 0.4358,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8691806331471136,
+      "grad_norm": 0.6061148008771401,
+      "learning_rate": 4.255247615394428e-06,
+      "loss": 0.4425,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8696461824953445,
+      "grad_norm": 0.6473964343401685,
+      "learning_rate": 4.254333700577645e-06,
+      "loss": 0.4466,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8701117318435754,
+      "grad_norm": 0.6295004249975784,
+      "learning_rate": 4.253419323627648e-06,
+      "loss": 0.4567,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8705772811918063,
+      "grad_norm": 0.6239265889564688,
+      "learning_rate": 4.2525044847853055e-06,
+      "loss": 0.4559,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8710428305400373,
+      "grad_norm": 0.6493730869285529,
+      "learning_rate": 4.251589184291608e-06,
+      "loss": 0.446,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8715083798882681,
+      "grad_norm": 0.6282528515444047,
+      "learning_rate": 4.250673422387667e-06,
+      "loss": 0.4369,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8719739292364991,
+      "grad_norm": 0.596048476435912,
+      "learning_rate": 4.249757199314715e-06,
+      "loss": 0.4477,
+      "step": 1873
+    },
+    {
+      "epoch": 0.87243947858473,
+      "grad_norm": 0.5947854057999189,
+      "learning_rate": 4.248840515314109e-06,
+      "loss": 0.4307,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8729050279329609,
+      "grad_norm": 0.6187364556480418,
+      "learning_rate": 4.247923370627324e-06,
+      "loss": 0.4508,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8733705772811918,
+      "grad_norm": 0.6098069459566919,
+      "learning_rate": 4.247005765495959e-06,
+      "loss": 0.4466,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8738361266294227,
+      "grad_norm": 0.6114276434813217,
+      "learning_rate": 4.2460877001617305e-06,
+      "loss": 0.4521,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8743016759776536,
+      "grad_norm": 0.5843815890969277,
+      "learning_rate": 4.2451691748664815e-06,
+      "loss": 0.4489,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8747672253258846,
+      "grad_norm": 0.6293871781583027,
+      "learning_rate": 4.244250189852172e-06,
+      "loss": 0.4396,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8752327746741154,
+      "grad_norm": 0.5927128277042962,
+      "learning_rate": 4.243330745360885e-06,
+      "loss": 0.4602,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8756983240223464,
+      "grad_norm": 0.6118307878089636,
+      "learning_rate": 4.242410841634823e-06,
+      "loss": 0.4363,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8761638733705773,
+      "grad_norm": 0.5903430356583703,
+      "learning_rate": 4.241490478916312e-06,
+      "loss": 0.4427,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8766294227188082,
+      "grad_norm": 0.5894800048985359,
+      "learning_rate": 4.2405696574477976e-06,
+      "loss": 0.4511,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8770949720670391,
+      "grad_norm": 0.6502709064427371,
+      "learning_rate": 4.239648377471844e-06,
+      "loss": 0.4426,
+      "step": 1884
+    },
+    {
+      "epoch": 0.87756052141527,
+      "grad_norm": 0.6297975074328278,
+      "learning_rate": 4.2387266392311396e-06,
+      "loss": 0.4525,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8780260707635009,
+      "grad_norm": 0.6359409044068398,
+      "learning_rate": 4.2378044429684925e-06,
+      "loss": 0.4591,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8784916201117319,
+      "grad_norm": 0.5844248998375748,
+      "learning_rate": 4.23688178892683e-06,
+      "loss": 0.4402,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8789571694599627,
+      "grad_norm": 0.5936185840161105,
+      "learning_rate": 4.2359586773492026e-06,
+      "loss": 0.4331,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8794227188081937,
+      "grad_norm": 0.6286077114735295,
+      "learning_rate": 4.235035108478779e-06,
+      "loss": 0.4445,
+      "step": 1889
+    },
+    {
+      "epoch": 0.8798882681564246,
+      "grad_norm": 0.6223668019828312,
+      "learning_rate": 4.234111082558849e-06,
+      "loss": 0.4699,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8803538175046555,
+      "grad_norm": 0.7036864448765909,
+      "learning_rate": 4.233186599832823e-06,
+      "loss": 0.4786,
+      "step": 1891
+    },
+    {
+      "epoch": 0.8808193668528864,
+      "grad_norm": 0.6221179756479911,
+      "learning_rate": 4.232261660544231e-06,
+      "loss": 0.4651,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8812849162011173,
+      "grad_norm": 0.6205819271757068,
+      "learning_rate": 4.231336264936726e-06,
+      "loss": 0.449,
+      "step": 1893
+    },
+    {
+      "epoch": 0.8817504655493482,
+      "grad_norm": 0.5958639711443685,
+      "learning_rate": 4.230410413254078e-06,
+      "loss": 0.4391,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8822160148975792,
+      "grad_norm": 0.664839393886841,
+      "learning_rate": 4.229484105740179e-06,
+      "loss": 0.4579,
+      "step": 1895
+    },
+    {
+      "epoch": 0.88268156424581,
+      "grad_norm": 0.6061549134725484,
+      "learning_rate": 4.22855734263904e-06,
+      "loss": 0.4224,
+      "step": 1896
+    },
+    {
+      "epoch": 0.883147113594041,
+      "grad_norm": 0.6127064424279364,
+      "learning_rate": 4.227630124194792e-06,
+      "loss": 0.4448,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8836126629422719,
+      "grad_norm": 0.5914538503030132,
+      "learning_rate": 4.2267024506516876e-06,
+      "loss": 0.4373,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8840782122905028,
+      "grad_norm": 0.611368304283315,
+      "learning_rate": 4.2257743222540956e-06,
+      "loss": 0.465,
+      "step": 1899
+    },
+    {
+      "epoch": 0.8845437616387337,
+      "grad_norm": 0.663144832469451,
+      "learning_rate": 4.22484573924651e-06,
+      "loss": 0.4657,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8850093109869647,
+      "grad_norm": 0.6363540993761571,
+      "learning_rate": 4.22391670187354e-06,
+      "loss": 0.4459,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8854748603351955,
+      "grad_norm": 0.6079920421489645,
+      "learning_rate": 4.222987210379916e-06,
+      "loss": 0.4558,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8859404096834265,
+      "grad_norm": 0.6005777473596958,
+      "learning_rate": 4.22205726501049e-06,
+      "loss": 0.4745,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8864059590316573,
+      "grad_norm": 0.6322067310642275,
+      "learning_rate": 4.22112686601023e-06,
+      "loss": 0.4398,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8868715083798883,
+      "grad_norm": 0.637155881513649,
+      "learning_rate": 4.220196013624225e-06,
+      "loss": 0.454,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8873370577281192,
+      "grad_norm": 0.6271573998367475,
+      "learning_rate": 4.219264708097685e-06,
+      "loss": 0.4306,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8878026070763501,
+      "grad_norm": 0.5716364294580859,
+      "learning_rate": 4.2183329496759364e-06,
+      "loss": 0.4427,
+      "step": 1907
+    },
+    {
+      "epoch": 0.888268156424581,
+      "grad_norm": 0.6188132589937421,
+      "learning_rate": 4.217400738604428e-06,
+      "loss": 0.447,
+      "step": 1908
+    },
+    {
+      "epoch": 0.888733705772812,
+      "grad_norm": 0.6339959259266003,
+      "learning_rate": 4.216468075128726e-06,
+      "loss": 0.4484,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8891992551210428,
+      "grad_norm": 0.6443624823548685,
+      "learning_rate": 4.215534959494515e-06,
+      "loss": 0.4228,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8896648044692738,
+      "grad_norm": 0.6373467991553858,
+      "learning_rate": 4.214601391947601e-06,
+      "loss": 0.4786,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8901303538175046,
+      "grad_norm": 0.6131404060505202,
+      "learning_rate": 4.213667372733909e-06,
+      "loss": 0.4623,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8905959031657356,
+      "grad_norm": 0.5659485691150472,
+      "learning_rate": 4.212732902099479e-06,
+      "loss": 0.4494,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8910614525139665,
+      "grad_norm": 0.5991689418048453,
+      "learning_rate": 4.2117979802904755e-06,
+      "loss": 0.4377,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8915270018621974,
+      "grad_norm": 0.5893317943871255,
+      "learning_rate": 4.210862607553178e-06,
+      "loss": 0.4449,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8919925512104283,
+      "grad_norm": 0.5984553173926012,
+      "learning_rate": 4.209926784133985e-06,
+      "loss": 0.4395,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8924581005586593,
+      "grad_norm": 0.6022561727905493,
+      "learning_rate": 4.208990510279416e-06,
+      "loss": 0.4374,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8929236499068901,
+      "grad_norm": 0.6197700551553612,
+      "learning_rate": 4.208053786236107e-06,
+      "loss": 0.4671,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8933891992551211,
+      "grad_norm": 0.6793952299558116,
+      "learning_rate": 4.2071166122508145e-06,
+      "loss": 0.4622,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8938547486033519,
+      "grad_norm": 0.63501431240461,
+      "learning_rate": 4.206178988570411e-06,
+      "loss": 0.4437,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8943202979515829,
+      "grad_norm": 0.6037664931850022,
+      "learning_rate": 4.20524091544189e-06,
+      "loss": 0.4493,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8947858472998138,
+      "grad_norm": 0.6208787025030736,
+      "learning_rate": 4.204302393112361e-06,
+      "loss": 0.4495,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8952513966480447,
+      "grad_norm": 0.6493773958594822,
+      "learning_rate": 4.203363421829054e-06,
+      "loss": 0.4655,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8957169459962756,
+      "grad_norm": 0.6165884604706069,
+      "learning_rate": 4.202424001839316e-06,
+      "loss": 0.4473,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8961824953445066,
+      "grad_norm": 0.6226155741502372,
+      "learning_rate": 4.201484133390615e-06,
+      "loss": 0.4473,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8966480446927374,
+      "grad_norm": 0.5997701407992959,
+      "learning_rate": 4.200543816730529e-06,
+      "loss": 0.4417,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8971135940409684,
+      "grad_norm": 0.6125131084324605,
+      "learning_rate": 4.199603052106765e-06,
+      "loss": 0.4445,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8975791433891993,
+      "grad_norm": 0.6286940855211806,
+      "learning_rate": 4.198661839767142e-06,
+      "loss": 0.458,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8980446927374302,
+      "grad_norm": 0.6375910365798136,
+      "learning_rate": 4.197720179959596e-06,
+      "loss": 0.4499,
+      "step": 1929
+    },
+    {
+      "epoch": 0.8985102420856611,
+      "grad_norm": 0.6104975204976724,
+      "learning_rate": 4.196778072932183e-06,
+      "loss": 0.4499,
+      "step": 1930
+    },
+    {
+      "epoch": 0.898975791433892,
+      "grad_norm": 0.6125434181966316,
+      "learning_rate": 4.195835518933077e-06,
+      "loss": 0.4302,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8994413407821229,
+      "grad_norm": 0.5617533286358537,
+      "learning_rate": 4.194892518210569e-06,
+      "loss": 0.4483,
+      "step": 1932
+    },
+    {
+      "epoch": 0.8999068901303539,
+      "grad_norm": 0.6355143650331884,
+      "learning_rate": 4.193949071013067e-06,
+      "loss": 0.444,
+      "step": 1933
+    },
+    {
+      "epoch": 0.9003724394785847,
+      "grad_norm": 0.5868153141431426,
+      "learning_rate": 4.193005177589099e-06,
+      "loss": 0.4393,
+      "step": 1934
+    },
+    {
+      "epoch": 0.9008379888268156,
+      "grad_norm": 0.6465390053961134,
+      "learning_rate": 4.1920608381873075e-06,
+      "loss": 0.4479,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9013035381750466,
+      "grad_norm": 0.6577469100419505,
+      "learning_rate": 4.191116053056454e-06,
+      "loss": 0.442,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9017690875232774,
+      "grad_norm": 0.6000218778923752,
+      "learning_rate": 4.1901708224454194e-06,
+      "loss": 0.4575,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9022346368715084,
+      "grad_norm": 0.58817463578685,
+      "learning_rate": 4.189225146603198e-06,
+      "loss": 0.4464,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9027001862197392,
+      "grad_norm": 0.6367050555297056,
+      "learning_rate": 4.188279025778903e-06,
+      "loss": 0.4328,
+      "step": 1939
+    },
+    {
+      "epoch": 0.9031657355679702,
+      "grad_norm": 0.636866354434664,
+      "learning_rate": 4.187332460221766e-06,
+      "loss": 0.4697,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9036312849162011,
+      "grad_norm": 0.6373917095720932,
+      "learning_rate": 4.186385450181135e-06,
+      "loss": 0.4441,
+      "step": 1941
+    },
+    {
+      "epoch": 0.904096834264432,
+      "grad_norm": 0.6112088962320124,
+      "learning_rate": 4.185437995906475e-06,
+      "loss": 0.4518,
+      "step": 1942
+    },
+    {
+      "epoch": 0.904562383612663,
+      "grad_norm": 0.6172233650718835,
+      "learning_rate": 4.184490097647368e-06,
+      "loss": 0.4572,
+      "step": 1943
+    },
+    {
+      "epoch": 0.9050279329608939,
+      "grad_norm": 0.5911524396289121,
+      "learning_rate": 4.183541755653512e-06,
+      "loss": 0.4776,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9054934823091247,
+      "grad_norm": 0.6184470274267406,
+      "learning_rate": 4.1825929701747235e-06,
+      "loss": 0.4264,
+      "step": 1945
+    },
+    {
+      "epoch": 0.9059590316573557,
+      "grad_norm": 0.6099800342951365,
+      "learning_rate": 4.181643741460935e-06,
+      "loss": 0.4547,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9064245810055865,
+      "grad_norm": 0.6919920270697215,
+      "learning_rate": 4.180694069762196e-06,
+      "loss": 0.4381,
+      "step": 1947
+    },
+    {
+      "epoch": 0.9068901303538175,
+      "grad_norm": 0.689719524592407,
+      "learning_rate": 4.179743955328673e-06,
+      "loss": 0.4375,
+      "step": 1948
+    },
+    {
+      "epoch": 0.9073556797020484,
+      "grad_norm": 0.6491404270647646,
+      "learning_rate": 4.178793398410648e-06,
+      "loss": 0.457,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9078212290502793,
+      "grad_norm": 0.6208206029588188,
+      "learning_rate": 4.177842399258521e-06,
+      "loss": 0.4451,
+      "step": 1950
+    },
+    {
+      "epoch": 0.9082867783985102,
+      "grad_norm": 0.5785600723489407,
+      "learning_rate": 4.176890958122807e-06,
+      "loss": 0.449,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9087523277467412,
+      "grad_norm": 0.6036664435923181,
+      "learning_rate": 4.175939075254139e-06,
+      "loss": 0.4403,
+      "step": 1952
+    },
+    {
+      "epoch": 0.909217877094972,
+      "grad_norm": 0.6727359618655931,
+      "learning_rate": 4.174986750903265e-06,
+      "loss": 0.4331,
+      "step": 1953
+    },
+    {
+      "epoch": 0.909683426443203,
+      "grad_norm": 0.6392465358358111,
+      "learning_rate": 4.174033985321051e-06,
+      "loss": 0.4416,
+      "step": 1954
+    },
+    {
+      "epoch": 0.910148975791434,
+      "grad_norm": 0.5892262564282884,
+      "learning_rate": 4.173080778758477e-06,
+      "loss": 0.436,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9106145251396648,
+      "grad_norm": 0.6113678851794414,
+      "learning_rate": 4.17212713146664e-06,
+      "loss": 0.432,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9110800744878957,
+      "grad_norm": 0.6129984711384145,
+      "learning_rate": 4.171173043696754e-06,
+      "loss": 0.4361,
+      "step": 1957
+    },
+    {
+      "epoch": 0.9115456238361266,
+      "grad_norm": 0.6715591248926025,
+      "learning_rate": 4.170218515700149e-06,
+      "loss": 0.4491,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9120111731843575,
+      "grad_norm": 0.635286041936838,
+      "learning_rate": 4.169263547728269e-06,
+      "loss": 0.4446,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9124767225325885,
+      "grad_norm": 0.6386302574693,
+      "learning_rate": 4.168308140032677e-06,
+      "loss": 0.4625,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9129422718808193,
+      "grad_norm": 0.5848727149865285,
+      "learning_rate": 4.167352292865047e-06,
+      "loss": 0.4374,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9134078212290503,
+      "grad_norm": 0.6273543051686267,
+      "learning_rate": 4.166396006477175e-06,
+      "loss": 0.4524,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9138733705772812,
+      "grad_norm": 0.6348552492937299,
+      "learning_rate": 4.165439281120969e-06,
+      "loss": 0.4509,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9143389199255121,
+      "grad_norm": 0.6057957398490826,
+      "learning_rate": 4.164482117048453e-06,
+      "loss": 0.4579,
+      "step": 1964
+    },
+    {
+      "epoch": 0.914804469273743,
+      "grad_norm": 0.6296144187961839,
+      "learning_rate": 4.163524514511765e-06,
+      "loss": 0.4489,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9152700186219739,
+      "grad_norm": 0.6082131643638563,
+      "learning_rate": 4.1625664737631636e-06,
+      "loss": 0.4508,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9157355679702048,
+      "grad_norm": 0.6497037083429005,
+      "learning_rate": 4.161607995055016e-06,
+      "loss": 0.4632,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9162011173184358,
+      "grad_norm": 0.5952305162648981,
+      "learning_rate": 4.160649078639811e-06,
+      "loss": 0.4616,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 0.5927266671883491,
+      "learning_rate": 4.159689724770148e-06,
+      "loss": 0.4578,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9171322160148976,
+      "grad_norm": 0.5874575887270123,
+      "learning_rate": 4.158729933698746e-06,
+      "loss": 0.4873,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9175977653631285,
+      "grad_norm": 0.5979652996021081,
+      "learning_rate": 4.157769705678435e-06,
+      "loss": 0.4547,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9180633147113594,
+      "grad_norm": 0.6394439375508122,
+      "learning_rate": 4.156809040962162e-06,
+      "loss": 0.463,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9185288640595903,
+      "grad_norm": 0.5911302460848695,
+      "learning_rate": 4.1558479398029885e-06,
+      "loss": 0.4444,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9189944134078212,
+      "grad_norm": 0.6172829483335525,
+      "learning_rate": 4.154886402454094e-06,
+      "loss": 0.4615,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9194599627560521,
+      "grad_norm": 0.6152875176849535,
+      "learning_rate": 4.153924429168766e-06,
+      "loss": 0.4483,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9199255121042831,
+      "grad_norm": 0.5970976855058866,
+      "learning_rate": 4.152962020200415e-06,
+      "loss": 0.4434,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9203910614525139,
+      "grad_norm": 0.6471926196542052,
+      "learning_rate": 4.151999175802561e-06,
+      "loss": 0.4632,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9208566108007449,
+      "grad_norm": 0.6019404167418193,
+      "learning_rate": 4.15103589622884e-06,
+      "loss": 0.4702,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9213221601489758,
+      "grad_norm": 0.5850242945615971,
+      "learning_rate": 4.150072181733003e-06,
+      "loss": 0.4617,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9217877094972067,
+      "grad_norm": 0.5982523059453437,
+      "learning_rate": 4.149108032568914e-06,
+      "loss": 0.4289,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9222532588454376,
+      "grad_norm": 0.6081974250180797,
+      "learning_rate": 4.148143448990554e-06,
+      "loss": 0.4405,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9227188081936686,
+      "grad_norm": 0.6051922081473894,
+      "learning_rate": 4.147178431252018e-06,
+      "loss": 0.4243,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9231843575418994,
+      "grad_norm": 0.5841216409620398,
+      "learning_rate": 4.146212979607514e-06,
+      "loss": 0.4452,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9236499068901304,
+      "grad_norm": 0.6381855539367836,
+      "learning_rate": 4.145247094311365e-06,
+      "loss": 0.4382,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9241154562383612,
+      "grad_norm": 0.6059154987326982,
+      "learning_rate": 4.144280775618007e-06,
+      "loss": 0.4516,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9245810055865922,
+      "grad_norm": 0.6302165335583384,
+      "learning_rate": 4.143314023781992e-06,
+      "loss": 0.4493,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9250465549348231,
+      "grad_norm": 0.583994187657475,
+      "learning_rate": 4.1423468390579865e-06,
+      "loss": 0.4543,
+      "step": 1987
+    },
+    {
+      "epoch": 0.925512104283054,
+      "grad_norm": 0.6126151331762183,
+      "learning_rate": 4.141379221700769e-06,
+      "loss": 0.4418,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9259776536312849,
+      "grad_norm": 0.598575978664457,
+      "learning_rate": 4.140411171965233e-06,
+      "loss": 0.4175,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9264432029795159,
+      "grad_norm": 0.5886582044450241,
+      "learning_rate": 4.139442690106386e-06,
+      "loss": 0.4591,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9269087523277467,
+      "grad_norm": 0.5869523083876385,
+      "learning_rate": 4.138473776379348e-06,
+      "loss": 0.4292,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9273743016759777,
+      "grad_norm": 0.6111895297994935,
+      "learning_rate": 4.137504431039356e-06,
+      "loss": 0.4476,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9278398510242085,
+      "grad_norm": 0.6016741329438929,
+      "learning_rate": 4.1365346543417565e-06,
+      "loss": 0.455,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9283054003724395,
+      "grad_norm": 0.6821715986733581,
+      "learning_rate": 4.135564446542014e-06,
+      "loss": 0.4361,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9287709497206704,
+      "grad_norm": 0.7837515177534649,
+      "learning_rate": 4.134593807895701e-06,
+      "loss": 0.4425,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9292364990689013,
+      "grad_norm": 0.620571309537726,
+      "learning_rate": 4.133622738658511e-06,
+      "loss": 0.44,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9297020484171322,
+      "grad_norm": 0.5784382790479756,
+      "learning_rate": 4.132651239086244e-06,
+      "loss": 0.4378,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9301675977653632,
+      "grad_norm": 0.6222877760941171,
+      "learning_rate": 4.131679309434816e-06,
+      "loss": 0.4603,
+      "step": 1998
+    },
+    {
+      "epoch": 0.930633147113594,
+      "grad_norm": 0.6220509952761436,
+      "learning_rate": 4.130706949960258e-06,
+      "loss": 0.4466,
+      "step": 1999
+    },
+    {
+      "epoch": 0.931098696461825,
+      "grad_norm": 0.6882605656940682,
+      "learning_rate": 4.129734160918711e-06,
+      "loss": 0.4471,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9315642458100558,
+      "grad_norm": 0.619221843591534,
+      "learning_rate": 4.128760942566432e-06,
+      "loss": 0.4724,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9320297951582868,
+      "grad_norm": 0.6176732962275371,
+      "learning_rate": 4.127787295159788e-06,
+      "loss": 0.4494,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9324953445065177,
+      "grad_norm": 0.6355946733788417,
+      "learning_rate": 4.126813218955265e-06,
+      "loss": 0.4429,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9329608938547486,
+      "grad_norm": 0.5999758590561899,
+      "learning_rate": 4.125838714209453e-06,
+      "loss": 0.4359,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9334264432029795,
+      "grad_norm": 0.6306307650370275,
+      "learning_rate": 4.124863781179062e-06,
+      "loss": 0.4684,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9338919925512105,
+      "grad_norm": 0.5810290414294176,
+      "learning_rate": 4.123888420120912e-06,
+      "loss": 0.4542,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9343575418994413,
+      "grad_norm": 0.5834172799872004,
+      "learning_rate": 4.122912631291936e-06,
+      "loss": 0.4472,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9348230912476723,
+      "grad_norm": 0.6147835047140829,
+      "learning_rate": 4.121936414949182e-06,
+      "loss": 0.4503,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9352886405959032,
+      "grad_norm": 0.6299333545690985,
+      "learning_rate": 4.120959771349807e-06,
+      "loss": 0.4452,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9357541899441341,
+      "grad_norm": 0.593985724807021,
+      "learning_rate": 4.119982700751082e-06,
+      "loss": 0.4506,
+      "step": 2010
+    },
+    {
+      "epoch": 0.936219739292365,
+      "grad_norm": 0.6046189889065604,
+      "learning_rate": 4.11900520341039e-06,
+      "loss": 0.4344,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9366852886405959,
+      "grad_norm": 0.5863640091266009,
+      "learning_rate": 4.118027279585229e-06,
+      "loss": 0.453,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9371508379888268,
+      "grad_norm": 0.6326321228831493,
+      "learning_rate": 4.117048929533206e-06,
+      "loss": 0.4492,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9376163873370578,
+      "grad_norm": 0.5981763739565863,
+      "learning_rate": 4.116070153512042e-06,
+      "loss": 0.4135,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9380819366852886,
+      "grad_norm": 0.6345822909851115,
+      "learning_rate": 4.115090951779571e-06,
+      "loss": 0.4543,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9385474860335196,
+      "grad_norm": 0.6721056939035224,
+      "learning_rate": 4.1141113245937355e-06,
+      "loss": 0.4511,
+      "step": 2016
+    },
+    {
+      "epoch": 0.9390130353817505,
+      "grad_norm": 0.6162219369631704,
+      "learning_rate": 4.113131272212595e-06,
+      "loss": 0.4613,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9394785847299814,
+      "grad_norm": 0.5691241434372432,
+      "learning_rate": 4.112150794894317e-06,
+      "loss": 0.4553,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9399441340782123,
+      "grad_norm": 0.5946357995205136,
+      "learning_rate": 4.111169892897183e-06,
+      "loss": 0.4371,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9404096834264432,
+      "grad_norm": 0.6932293176763914,
+      "learning_rate": 4.110188566479587e-06,
+      "loss": 0.4308,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9408752327746741,
+      "grad_norm": 0.6134752866722138,
+      "learning_rate": 4.109206815900032e-06,
+      "loss": 0.442,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9413407821229051,
+      "grad_norm": 0.5925677972427282,
+      "learning_rate": 4.108224641417136e-06,
+      "loss": 0.4447,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9418063314711359,
+      "grad_norm": 0.6127070675864984,
+      "learning_rate": 4.1072420432896255e-06,
+      "loss": 0.4422,
+      "step": 2023
+    },
+    {
+      "epoch": 0.9422718808193669,
+      "grad_norm": 0.6209900221822975,
+      "learning_rate": 4.106259021776342e-06,
+      "loss": 0.4399,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9427374301675978,
+      "grad_norm": 0.5935366562320886,
+      "learning_rate": 4.105275577136235e-06,
+      "loss": 0.4544,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9432029795158287,
+      "grad_norm": 0.6025344517990489,
+      "learning_rate": 4.104291709628368e-06,
+      "loss": 0.458,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9436685288640596,
+      "grad_norm": 0.600944397088997,
+      "learning_rate": 4.103307419511916e-06,
+      "loss": 0.4561,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9441340782122905,
+      "grad_norm": 0.6570262902524859,
+      "learning_rate": 4.102322707046163e-06,
+      "loss": 0.4495,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9445996275605214,
+      "grad_norm": 0.6337244688375989,
+      "learning_rate": 4.101337572490507e-06,
+      "loss": 0.4316,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9450651769087524,
+      "grad_norm": 0.5759518476370782,
+      "learning_rate": 4.100352016104454e-06,
+      "loss": 0.4614,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9455307262569832,
+      "grad_norm": 0.603534431378814,
+      "learning_rate": 4.099366038147625e-06,
+      "loss": 0.4493,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9459962756052142,
+      "grad_norm": 0.606646548415215,
+      "learning_rate": 4.098379638879748e-06,
+      "loss": 0.4398,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9464618249534451,
+      "grad_norm": 0.663193399837009,
+      "learning_rate": 4.097392818560666e-06,
+      "loss": 0.4344,
+      "step": 2033
+    },
+    {
+      "epoch": 0.946927374301676,
+      "grad_norm": 0.636856904829439,
+      "learning_rate": 4.09640557745033e-06,
+      "loss": 0.4371,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9473929236499069,
+      "grad_norm": 0.6070222141607133,
+      "learning_rate": 4.095417915808803e-06,
+      "loss": 0.4633,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9478584729981379,
+      "grad_norm": 0.5809154398732909,
+      "learning_rate": 4.094429833896258e-06,
+      "loss": 0.462,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9483240223463687,
+      "grad_norm": 0.6488697922627221,
+      "learning_rate": 4.09344133197298e-06,
+      "loss": 0.4446,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9487895716945997,
+      "grad_norm": 0.6667152323358488,
+      "learning_rate": 4.092452410299363e-06,
+      "loss": 0.4346,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9492551210428305,
+      "grad_norm": 0.6135814819229224,
+      "learning_rate": 4.0914630691359135e-06,
+      "loss": 0.4571,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9497206703910615,
+      "grad_norm": 0.6081235390898522,
+      "learning_rate": 4.090473308743247e-06,
+      "loss": 0.4643,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9501862197392924,
+      "grad_norm": 0.5730400292533671,
+      "learning_rate": 4.08948312938209e-06,
+      "loss": 0.4581,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9506517690875232,
+      "grad_norm": 0.5669651968436368,
+      "learning_rate": 4.08849253131328e-06,
+      "loss": 0.449,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9511173184357542,
+      "grad_norm": 0.5894895605565966,
+      "learning_rate": 4.087501514797762e-06,
+      "loss": 0.4346,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9515828677839852,
+      "grad_norm": 0.6112640706318471,
+      "learning_rate": 4.086510080096596e-06,
+      "loss": 0.4568,
+      "step": 2044
+    },
+    {
+      "epoch": 0.952048417132216,
+      "grad_norm": 0.6005645250312277,
+      "learning_rate": 4.085518227470948e-06,
+      "loss": 0.4575,
+      "step": 2045
+    },
+    {
+      "epoch": 0.952513966480447,
+      "grad_norm": 0.6796528648250435,
+      "learning_rate": 4.0845259571820965e-06,
+      "loss": 0.449,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9529795158286778,
+      "grad_norm": 0.6204158205611118,
+      "learning_rate": 4.083533269491427e-06,
+      "loss": 0.4468,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9534450651769087,
+      "grad_norm": 0.6192573104551874,
+      "learning_rate": 4.082540164660439e-06,
+      "loss": 0.4759,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9539106145251397,
+      "grad_norm": 0.5633432357128983,
+      "learning_rate": 4.081546642950739e-06,
+      "loss": 0.4286,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9543761638733705,
+      "grad_norm": 0.5634565538893145,
+      "learning_rate": 4.080552704624045e-06,
+      "loss": 0.4395,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9548417132216015,
+      "grad_norm": 0.6274690441881213,
+      "learning_rate": 4.079558349942183e-06,
+      "loss": 0.4433,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9553072625698324,
+      "grad_norm": 0.5965802628874185,
+      "learning_rate": 4.07856357916709e-06,
+      "loss": 0.4204,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9557728119180633,
+      "grad_norm": 0.5889944677505153,
+      "learning_rate": 4.077568392560813e-06,
+      "loss": 0.4493,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9562383612662942,
+      "grad_norm": 0.6364493851375587,
+      "learning_rate": 4.076572790385508e-06,
+      "loss": 0.4172,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9567039106145251,
+      "grad_norm": 0.593087413985067,
+      "learning_rate": 4.075576772903438e-06,
+      "loss": 0.4516,
+      "step": 2055
+    },
+    {
+      "epoch": 0.957169459962756,
+      "grad_norm": 0.6276781054138691,
+      "learning_rate": 4.07458034037698e-06,
+      "loss": 0.4343,
+      "step": 2056
+    },
+    {
+      "epoch": 0.957635009310987,
+      "grad_norm": 0.631037248937793,
+      "learning_rate": 4.073583493068618e-06,
+      "loss": 0.4477,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9581005586592178,
+      "grad_norm": 0.5860396891409251,
+      "learning_rate": 4.072586231240944e-06,
+      "loss": 0.4507,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9585661080074488,
+      "grad_norm": 0.6658662377656706,
+      "learning_rate": 4.0715885551566615e-06,
+      "loss": 0.4351,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9590316573556797,
+      "grad_norm": 0.6323615662479214,
+      "learning_rate": 4.070590465078581e-06,
+      "loss": 0.4454,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9594972067039106,
+      "grad_norm": 0.6629971346905114,
+      "learning_rate": 4.069591961269624e-06,
+      "loss": 0.4536,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9599627560521415,
+      "grad_norm": 0.6729058463719864,
+      "learning_rate": 4.068593043992821e-06,
+      "loss": 0.4531,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9604283054003724,
+      "grad_norm": 0.6720348764395483,
+      "learning_rate": 4.067593713511308e-06,
+      "loss": 0.4419,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9608938547486033,
+      "grad_norm": 0.6178164481755847,
+      "learning_rate": 4.066593970088335e-06,
+      "loss": 0.4399,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9613594040968343,
+      "grad_norm": 0.5714319443091412,
+      "learning_rate": 4.065593813987257e-06,
+      "loss": 0.4289,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9618249534450651,
+      "grad_norm": 0.7263899923503643,
+      "learning_rate": 4.064593245471539e-06,
+      "loss": 0.4613,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9622905027932961,
+      "grad_norm": 0.6181457861136274,
+      "learning_rate": 4.063592264804754e-06,
+      "loss": 0.4336,
+      "step": 2067
+    },
+    {
+      "epoch": 0.962756052141527,
+      "grad_norm": 0.5866152961722146,
+      "learning_rate": 4.062590872250586e-06,
+      "loss": 0.4597,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9632216014897579,
+      "grad_norm": 0.6176366167707104,
+      "learning_rate": 4.061589068072823e-06,
+      "loss": 0.421,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9636871508379888,
+      "grad_norm": 0.626834834607465,
+      "learning_rate": 4.060586852535366e-06,
+      "loss": 0.4144,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9641527001862198,
+      "grad_norm": 0.622436933932794,
+      "learning_rate": 4.059584225902221e-06,
+      "loss": 0.4406,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9646182495344506,
+      "grad_norm": 0.631322887216571,
+      "learning_rate": 4.058581188437506e-06,
+      "loss": 0.4323,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9650837988826816,
+      "grad_norm": 0.6426582672049326,
+      "learning_rate": 4.0575777404054425e-06,
+      "loss": 0.4451,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9655493482309124,
+      "grad_norm": 0.5958320498555161,
+      "learning_rate": 4.056573882070364e-06,
+      "loss": 0.4545,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9660148975791434,
+      "grad_norm": 0.6396872171770792,
+      "learning_rate": 4.0555696136967096e-06,
+      "loss": 0.4337,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9664804469273743,
+      "grad_norm": 0.6289464052643968,
+      "learning_rate": 4.054564935549028e-06,
+      "loss": 0.4386,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9669459962756052,
+      "grad_norm": 0.6209190136731547,
+      "learning_rate": 4.053559847891975e-06,
+      "loss": 0.4522,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9674115456238361,
+      "grad_norm": 0.6776024007865836,
+      "learning_rate": 4.052554350990315e-06,
+      "loss": 0.4338,
+      "step": 2078
+    },
+    {
+      "epoch": 0.9678770949720671,
+      "grad_norm": 0.6496713636065933,
+      "learning_rate": 4.05154844510892e-06,
+      "loss": 0.4465,
+      "step": 2079
+    },
+    {
+      "epoch": 0.9683426443202979,
+      "grad_norm": 0.6726536749731362,
+      "learning_rate": 4.05054213051277e-06,
+      "loss": 0.452,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9688081936685289,
+      "grad_norm": 0.6211168154706808,
+      "learning_rate": 4.04953540746695e-06,
+      "loss": 0.4253,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9692737430167597,
+      "grad_norm": 0.6135285906716463,
+      "learning_rate": 4.0485282762366576e-06,
+      "loss": 0.4497,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9697392923649907,
+      "grad_norm": 0.6159024795662376,
+      "learning_rate": 4.047520737087193e-06,
+      "loss": 0.4434,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9702048417132216,
+      "grad_norm": 0.6004644321192614,
+      "learning_rate": 4.046512790283966e-06,
+      "loss": 0.431,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9706703910614525,
+      "grad_norm": 0.5910754930226851,
+      "learning_rate": 4.045504436092495e-06,
+      "loss": 0.449,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9711359404096834,
+      "grad_norm": 0.5918525954446335,
+      "learning_rate": 4.044495674778404e-06,
+      "loss": 0.4125,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9716014897579144,
+      "grad_norm": 0.5832121663847467,
+      "learning_rate": 4.0434865066074246e-06,
+      "loss": 0.436,
+      "step": 2087
+    },
+    {
+      "epoch": 0.9720670391061452,
+      "grad_norm": 0.6361853304550873,
+      "learning_rate": 4.042476931845396e-06,
+      "loss": 0.4498,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9725325884543762,
+      "grad_norm": 0.6329191152898948,
+      "learning_rate": 4.041466950758264e-06,
+      "loss": 0.449,
+      "step": 2089
+    },
+    {
+      "epoch": 0.972998137802607,
+      "grad_norm": 0.6312342052974993,
+      "learning_rate": 4.040456563612081e-06,
+      "loss": 0.4219,
+      "step": 2090
+    },
+    {
+      "epoch": 0.973463687150838,
+      "grad_norm": 0.5882175202536143,
+      "learning_rate": 4.039445770673008e-06,
+      "loss": 0.4343,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9739292364990689,
+      "grad_norm": 0.6042853212147463,
+      "learning_rate": 4.03843457220731e-06,
+      "loss": 0.4463,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9743947858472998,
+      "grad_norm": 0.6010646450932734,
+      "learning_rate": 4.0374229684813646e-06,
+      "loss": 0.458,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9748603351955307,
+      "grad_norm": 0.6169434070247283,
+      "learning_rate": 4.036410959761648e-06,
+      "loss": 0.4234,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9753258845437617,
+      "grad_norm": 0.637677694920564,
+      "learning_rate": 4.035398546314751e-06,
+      "loss": 0.4463,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9757914338919925,
+      "grad_norm": 0.6380063857414677,
+      "learning_rate": 4.034385728407364e-06,
+      "loss": 0.4576,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9762569832402235,
+      "grad_norm": 0.5933499856945663,
+      "learning_rate": 4.03337250630629e-06,
+      "loss": 0.4308,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9767225325884544,
+      "grad_norm": 0.5589721216227449,
+      "learning_rate": 4.032358880278434e-06,
+      "loss": 0.4321,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9771880819366853,
+      "grad_norm": 0.6192927246872337,
+      "learning_rate": 4.031344850590809e-06,
+      "loss": 0.4426,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9776536312849162,
+      "grad_norm": 0.5705304841436102,
+      "learning_rate": 4.030330417510537e-06,
+      "loss": 0.4534,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9781191806331471,
+      "grad_norm": 0.5888807991527564,
+      "learning_rate": 4.029315581304842e-06,
+      "loss": 0.4492,
+      "step": 2101
+    },
+    {
+      "epoch": 0.978584729981378,
+      "grad_norm": 0.5849117865122486,
+      "learning_rate": 4.028300342241057e-06,
+      "loss": 0.423,
+      "step": 2102
+    },
+    {
+      "epoch": 0.979050279329609,
+      "grad_norm": 0.6254411401194053,
+      "learning_rate": 4.027284700586618e-06,
+      "loss": 0.4447,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9795158286778398,
+      "grad_norm": 0.6047558541659995,
+      "learning_rate": 4.026268656609071e-06,
+      "loss": 0.4708,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9799813780260708,
+      "grad_norm": 0.594495832681331,
+      "learning_rate": 4.025252210576066e-06,
+      "loss": 0.4423,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9804469273743017,
+      "grad_norm": 0.6205986244006005,
+      "learning_rate": 4.024235362755357e-06,
+      "loss": 0.4395,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9809124767225326,
+      "grad_norm": 0.6420708552957556,
+      "learning_rate": 4.02321811341481e-06,
+      "loss": 0.462,
+      "step": 2107
+    },
+    {
+      "epoch": 0.9813780260707635,
+      "grad_norm": 0.634173586588534,
+      "learning_rate": 4.022200462822389e-06,
+      "loss": 0.4455,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9818435754189944,
+      "grad_norm": 0.5880481736009339,
+      "learning_rate": 4.021182411246169e-06,
+      "loss": 0.4411,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9823091247672253,
+      "grad_norm": 0.645163467480518,
+      "learning_rate": 4.020163958954328e-06,
+      "loss": 0.4371,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9827746741154563,
+      "grad_norm": 0.6167476708153812,
+      "learning_rate": 4.0191451062151524e-06,
+      "loss": 0.4416,
+      "step": 2111
+    },
+    {
+      "epoch": 0.9832402234636871,
+      "grad_norm": 0.6439379344641438,
+      "learning_rate": 4.018125853297029e-06,
+      "loss": 0.4392,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9837057728119181,
+      "grad_norm": 0.6402362838616752,
+      "learning_rate": 4.017106200468457e-06,
+      "loss": 0.4356,
+      "step": 2113
+    },
+    {
+      "epoch": 0.984171322160149,
+      "grad_norm": 0.6620336844387272,
+      "learning_rate": 4.0160861479980354e-06,
+      "loss": 0.4494,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9846368715083799,
+      "grad_norm": 0.6614441775827302,
+      "learning_rate": 4.01506569615447e-06,
+      "loss": 0.4602,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9851024208566108,
+      "grad_norm": 0.5804885007310221,
+      "learning_rate": 4.014044845206572e-06,
+      "loss": 0.4418,
+      "step": 2116
+    },
+    {
+      "epoch": 0.9855679702048417,
+      "grad_norm": 0.5707495071401645,
+      "learning_rate": 4.013023595423259e-06,
+      "loss": 0.4479,
+      "step": 2117
+    },
+    {
+      "epoch": 0.9860335195530726,
+      "grad_norm": 0.6084740478896158,
+      "learning_rate": 4.0120019470735515e-06,
+      "loss": 0.4625,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9864990689013036,
+      "grad_norm": 0.5887017889911642,
+      "learning_rate": 4.010979900426576e-06,
+      "loss": 0.4636,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9869646182495344,
+      "grad_norm": 0.6017720778694695,
+      "learning_rate": 4.009957455751564e-06,
+      "loss": 0.4227,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9874301675977654,
+      "grad_norm": 0.7005399333919606,
+      "learning_rate": 4.0089346133178524e-06,
+      "loss": 0.4355,
+      "step": 2121
+    },
+    {
+      "epoch": 0.9878957169459963,
+      "grad_norm": 0.6078127273028366,
+      "learning_rate": 4.0079113733948815e-06,
+      "loss": 0.4396,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9883612662942272,
+      "grad_norm": 0.6150818227904749,
+      "learning_rate": 4.0068877362521965e-06,
+      "loss": 0.4367,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9888268156424581,
+      "grad_norm": 0.5937642461259379,
+      "learning_rate": 4.00586370215945e-06,
+      "loss": 0.4466,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9892923649906891,
+      "grad_norm": 0.620887801832486,
+      "learning_rate": 4.0048392713863936e-06,
+      "loss": 0.4345,
+      "step": 2125
+    },
+    {
+      "epoch": 0.9897579143389199,
+      "grad_norm": 0.6157104457835274,
+      "learning_rate": 4.003814444202889e-06,
+      "loss": 0.4655,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9902234636871509,
+      "grad_norm": 0.6023908581552272,
+      "learning_rate": 4.0027892208789e-06,
+      "loss": 0.4264,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9906890130353817,
+      "grad_norm": 0.6145382696088436,
+      "learning_rate": 4.001763601684494e-06,
+      "loss": 0.4553,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9911545623836127,
+      "grad_norm": 0.6054370061830587,
+      "learning_rate": 4.000737586889844e-06,
+      "loss": 0.4508,
+      "step": 2129
+    },
+    {
+      "epoch": 0.9916201117318436,
+      "grad_norm": 0.5981059254537149,
+      "learning_rate": 3.999711176765227e-06,
+      "loss": 0.4542,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9920856610800745,
+      "grad_norm": 0.5982076778278181,
+      "learning_rate": 3.998684371581023e-06,
+      "loss": 0.4313,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9925512104283054,
+      "grad_norm": 0.6057809099667069,
+      "learning_rate": 3.997657171607717e-06,
+      "loss": 0.4376,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9930167597765364,
+      "grad_norm": 0.6147494016133986,
+      "learning_rate": 3.996629577115898e-06,
+      "loss": 0.437,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9934823091247672,
+      "grad_norm": 0.6057334850375249,
+      "learning_rate": 3.99560158837626e-06,
+      "loss": 0.4391,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9939478584729982,
+      "grad_norm": 0.607389399873129,
+      "learning_rate": 3.994573205659597e-06,
+      "loss": 0.4476,
+      "step": 2135
+    },
+    {
+      "epoch": 0.994413407821229,
+      "grad_norm": 0.6655750005164427,
+      "learning_rate": 3.993544429236813e-06,
+      "loss": 0.4607,
+      "step": 2136
+    },
+    {
+      "epoch": 0.99487895716946,
+      "grad_norm": 0.6928767404431654,
+      "learning_rate": 3.99251525937891e-06,
+      "loss": 0.4514,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9953445065176909,
+      "grad_norm": 0.6041900134147651,
+      "learning_rate": 3.991485696356995e-06,
+      "loss": 0.438,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9958100558659218,
+      "grad_norm": 0.6082567154669749,
+      "learning_rate": 3.990455740442282e-06,
+      "loss": 0.4435,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9962756052141527,
+      "grad_norm": 0.6181716249178792,
+      "learning_rate": 3.989425391906084e-06,
+      "loss": 0.4367,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9967411545623837,
+      "grad_norm": 0.6321127974966788,
+      "learning_rate": 3.988394651019819e-06,
+      "loss": 0.4354,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9972067039106145,
+      "grad_norm": 0.6795111573897884,
+      "learning_rate": 3.98736351805501e-06,
+      "loss": 0.4446,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9976722532588455,
+      "grad_norm": 0.62297430963798,
+      "learning_rate": 3.98633199328328e-06,
+      "loss": 0.4554,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9981378026070763,
+      "grad_norm": 0.59918887669031,
+      "learning_rate": 3.985300076976359e-06,
+      "loss": 0.4502,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9986033519553073,
+      "grad_norm": 0.6397276544797912,
+      "learning_rate": 3.984267769406079e-06,
+      "loss": 0.4603,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9990689013035382,
+      "grad_norm": 0.6184867861282326,
+      "learning_rate": 3.983235070844372e-06,
+      "loss": 0.4544,
+      "step": 2146
+    },
+    {
+      "epoch": 0.999534450651769,
+      "grad_norm": 0.6161180643238597,
+      "learning_rate": 3.982201981563277e-06,
+      "loss": 0.4584,
+      "step": 2147
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6012326840449546,
+      "learning_rate": 3.981168501834933e-06,
+      "loss": 0.4455,
+      "step": 2148
+    },
+    {
+      "epoch": 1.0004655493482308,
+      "grad_norm": 0.883429673764452,
+      "learning_rate": 3.980134631931585e-06,
+      "loss": 0.3685,
+      "step": 2149
+    },
+    {
+      "epoch": 1.000931098696462,
+      "grad_norm": 0.871976828609972,
+      "learning_rate": 3.9791003721255775e-06,
+      "loss": 0.4155,
+      "step": 2150
+    },
+    {
+      "epoch": 1.0013966480446927,
+      "grad_norm": 0.7190509765617679,
+      "learning_rate": 3.978065722689359e-06,
+      "loss": 0.4028,
+      "step": 2151
+    },
+    {
+      "epoch": 1.0018621973929236,
+      "grad_norm": 0.6742004538585304,
+      "learning_rate": 3.9770306838954805e-06,
+      "loss": 0.3927,
+      "step": 2152
+    },
+    {
+      "epoch": 1.0023277467411547,
+      "grad_norm": 0.7336208349680067,
+      "learning_rate": 3.975995256016598e-06,
+      "loss": 0.3924,
+      "step": 2153
+    },
+    {
+      "epoch": 1.0027932960893855,
+      "grad_norm": 0.7277949148838316,
+      "learning_rate": 3.974959439325466e-06,
+      "loss": 0.3862,
+      "step": 2154
+    },
+    {
+      "epoch": 1.0032588454376163,
+      "grad_norm": 0.7035445620423814,
+      "learning_rate": 3.973923234094943e-06,
+      "loss": 0.3859,
+      "step": 2155
+    },
+    {
+      "epoch": 1.0037243947858474,
+      "grad_norm": 0.6572594505065166,
+      "learning_rate": 3.972886640597991e-06,
+      "loss": 0.4048,
+      "step": 2156
+    },
+    {
+      "epoch": 1.0041899441340782,
+      "grad_norm": 0.6497232675368749,
+      "learning_rate": 3.971849659107673e-06,
+      "loss": 0.397,
+      "step": 2157
+    },
+    {
+      "epoch": 1.004655493482309,
+      "grad_norm": 0.7035604277898471,
+      "learning_rate": 3.9708122898971554e-06,
+      "loss": 0.402,
+      "step": 2158
+    },
+    {
+      "epoch": 1.00512104283054,
+      "grad_norm": 0.6764240996288524,
+      "learning_rate": 3.969774533239704e-06,
+      "loss": 0.3927,
+      "step": 2159
+    },
+    {
+      "epoch": 1.005586592178771,
+      "grad_norm": 0.7345640141233657,
+      "learning_rate": 3.96873638940869e-06,
+      "loss": 0.3875,
+      "step": 2160
+    },
+    {
+      "epoch": 1.0060521415270018,
+      "grad_norm": 0.6560565862618178,
+      "learning_rate": 3.967697858677586e-06,
+      "loss": 0.3962,
+      "step": 2161
+    },
+    {
+      "epoch": 1.0065176908752327,
+      "grad_norm": 0.671407079650755,
+      "learning_rate": 3.966658941319964e-06,
+      "loss": 0.3968,
+      "step": 2162
+    },
+    {
+      "epoch": 1.0069832402234637,
+      "grad_norm": 0.6370765675548594,
+      "learning_rate": 3.9656196376095e-06,
+      "loss": 0.3801,
+      "step": 2163
+    },
+    {
+      "epoch": 1.0074487895716946,
+      "grad_norm": 0.6501638692769072,
+      "learning_rate": 3.9645799478199714e-06,
+      "loss": 0.3906,
+      "step": 2164
+    },
+    {
+      "epoch": 1.0079143389199254,
+      "grad_norm": 0.6565677124428211,
+      "learning_rate": 3.9635398722252574e-06,
+      "loss": 0.3883,
+      "step": 2165
+    },
+    {
+      "epoch": 1.0083798882681565,
+      "grad_norm": 0.6969045275079409,
+      "learning_rate": 3.962499411099339e-06,
+      "loss": 0.3837,
+      "step": 2166
+    },
+    {
+      "epoch": 1.0088454376163873,
+      "grad_norm": 0.627110702936665,
+      "learning_rate": 3.961458564716297e-06,
+      "loss": 0.3881,
+      "step": 2167
+    },
+    {
+      "epoch": 1.0093109869646182,
+      "grad_norm": 0.6083890493541673,
+      "learning_rate": 3.960417333350317e-06,
+      "loss": 0.3798,
+      "step": 2168
+    },
+    {
+      "epoch": 1.0097765363128492,
+      "grad_norm": 0.6399766667139741,
+      "learning_rate": 3.959375717275683e-06,
+      "loss": 0.3813,
+      "step": 2169
+    },
+    {
+      "epoch": 1.01024208566108,
+      "grad_norm": 0.6425283357816867,
+      "learning_rate": 3.958333716766781e-06,
+      "loss": 0.4119,
+      "step": 2170
+    },
+    {
+      "epoch": 1.010707635009311,
+      "grad_norm": 0.6381976894271216,
+      "learning_rate": 3.9572913320981e-06,
+      "loss": 0.3862,
+      "step": 2171
+    },
+    {
+      "epoch": 1.011173184357542,
+      "grad_norm": 0.6838803569314552,
+      "learning_rate": 3.9562485635442276e-06,
+      "loss": 0.4086,
+      "step": 2172
+    },
+    {
+      "epoch": 1.0116387337057728,
+      "grad_norm": 0.6173973272329338,
+      "learning_rate": 3.955205411379854e-06,
+      "loss": 0.4071,
+      "step": 2173
+    },
+    {
+      "epoch": 1.0121042830540037,
+      "grad_norm": 0.6389589741318543,
+      "learning_rate": 3.954161875879771e-06,
+      "loss": 0.3735,
+      "step": 2174
+    },
+    {
+      "epoch": 1.0125698324022347,
+      "grad_norm": 0.641959472880965,
+      "learning_rate": 3.953117957318871e-06,
+      "loss": 0.3823,
+      "step": 2175
+    },
+    {
+      "epoch": 1.0130353817504656,
+      "grad_norm": 0.5935206911457699,
+      "learning_rate": 3.952073655972145e-06,
+      "loss": 0.3829,
+      "step": 2176
+    },
+    {
+      "epoch": 1.0135009310986964,
+      "grad_norm": 0.5623229769949024,
+      "learning_rate": 3.9510289721146895e-06,
+      "loss": 0.3796,
+      "step": 2177
+    },
+    {
+      "epoch": 1.0139664804469273,
+      "grad_norm": 0.6048043288154447,
+      "learning_rate": 3.949983906021697e-06,
+      "loss": 0.3916,
+      "step": 2178
+    },
+    {
+      "epoch": 1.0144320297951583,
+      "grad_norm": 0.6280266421991307,
+      "learning_rate": 3.948938457968465e-06,
+      "loss": 0.3925,
+      "step": 2179
+    },
+    {
+      "epoch": 1.0148975791433892,
+      "grad_norm": 0.6004678613302431,
+      "learning_rate": 3.947892628230386e-06,
+      "loss": 0.3863,
+      "step": 2180
+    },
+    {
+      "epoch": 1.01536312849162,
+      "grad_norm": 0.6493677369649239,
+      "learning_rate": 3.946846417082959e-06,
+      "loss": 0.4079,
+      "step": 2181
+    },
+    {
+      "epoch": 1.015828677839851,
+      "grad_norm": 0.6030007869978194,
+      "learning_rate": 3.94579982480178e-06,
+      "loss": 0.3889,
+      "step": 2182
+    },
+    {
+      "epoch": 1.016294227188082,
+      "grad_norm": 0.5745404924209552,
+      "learning_rate": 3.9447528516625476e-06,
+      "loss": 0.397,
+      "step": 2183
+    },
+    {
+      "epoch": 1.0167597765363128,
+      "grad_norm": 0.6688380188937231,
+      "learning_rate": 3.943705497941057e-06,
+      "loss": 0.3859,
+      "step": 2184
+    },
+    {
+      "epoch": 1.0172253258845438,
+      "grad_norm": 0.6698790611785655,
+      "learning_rate": 3.942657763913208e-06,
+      "loss": 0.3661,
+      "step": 2185
+    },
+    {
+      "epoch": 1.0176908752327747,
+      "grad_norm": 0.599531174824093,
+      "learning_rate": 3.941609649854997e-06,
+      "loss": 0.3858,
+      "step": 2186
+    },
+    {
+      "epoch": 1.0181564245810055,
+      "grad_norm": 0.6024523391114207,
+      "learning_rate": 3.940561156042525e-06,
+      "loss": 0.3948,
+      "step": 2187
+    },
+    {
+      "epoch": 1.0186219739292366,
+      "grad_norm": 0.5933210430074157,
+      "learning_rate": 3.939512282751987e-06,
+      "loss": 0.3987,
+      "step": 2188
+    },
+    {
+      "epoch": 1.0190875232774674,
+      "grad_norm": 0.6252243413364676,
+      "learning_rate": 3.938463030259683e-06,
+      "loss": 0.3777,
+      "step": 2189
+    },
+    {
+      "epoch": 1.0195530726256983,
+      "grad_norm": 0.6057313402593021,
+      "learning_rate": 3.93741339884201e-06,
+      "loss": 0.3886,
+      "step": 2190
+    },
+    {
+      "epoch": 1.0200186219739293,
+      "grad_norm": 1.3528860475748905,
+      "learning_rate": 3.936363388775468e-06,
+      "loss": 0.3958,
+      "step": 2191
+    },
+    {
+      "epoch": 1.0204841713221602,
+      "grad_norm": 0.6257962556988487,
+      "learning_rate": 3.9353130003366515e-06,
+      "loss": 0.3828,
+      "step": 2192
+    },
+    {
+      "epoch": 1.020949720670391,
+      "grad_norm": 0.6168036111836858,
+      "learning_rate": 3.934262233802261e-06,
+      "loss": 0.3692,
+      "step": 2193
+    },
+    {
+      "epoch": 1.0214152700186219,
+      "grad_norm": 0.6887306910820064,
+      "learning_rate": 3.933211089449091e-06,
+      "loss": 0.3966,
+      "step": 2194
+    },
+    {
+      "epoch": 1.021880819366853,
+      "grad_norm": 0.7722218601122998,
+      "learning_rate": 3.932159567554038e-06,
+      "loss": 0.3894,
+      "step": 2195
+    },
+    {
+      "epoch": 1.0223463687150838,
+      "grad_norm": 0.6168538039814763,
+      "learning_rate": 3.931107668394099e-06,
+      "loss": 0.385,
+      "step": 2196
+    },
+    {
+      "epoch": 1.0228119180633146,
+      "grad_norm": 0.6450366180540914,
+      "learning_rate": 3.9300553922463695e-06,
+      "loss": 0.3907,
+      "step": 2197
+    },
+    {
+      "epoch": 1.0232774674115457,
+      "grad_norm": 0.6506574280986851,
+      "learning_rate": 3.929002739388043e-06,
+      "loss": 0.3887,
+      "step": 2198
+    },
+    {
+      "epoch": 1.0237430167597765,
+      "grad_norm": 0.5551263976064659,
+      "learning_rate": 3.927949710096414e-06,
+      "loss": 0.3818,
+      "step": 2199
+    },
+    {
+      "epoch": 1.0242085661080074,
+      "grad_norm": 0.6544566739298661,
+      "learning_rate": 3.9268963046488744e-06,
+      "loss": 0.4024,
+      "step": 2200
+    },
+    {
+      "epoch": 1.0246741154562384,
+      "grad_norm": 0.5830451588876331,
+      "learning_rate": 3.925842523322918e-06,
+      "loss": 0.3669,
+      "step": 2201
+    },
+    {
+      "epoch": 1.0251396648044693,
+      "grad_norm": 0.6271651079907353,
+      "learning_rate": 3.924788366396134e-06,
+      "loss": 0.3809,
+      "step": 2202
+    },
+    {
+      "epoch": 1.0256052141527001,
+      "grad_norm": 0.5746677697498828,
+      "learning_rate": 3.923733834146214e-06,
+      "loss": 0.3785,
+      "step": 2203
+    },
+    {
+      "epoch": 1.0260707635009312,
+      "grad_norm": 0.6043907360815733,
+      "learning_rate": 3.922678926850946e-06,
+      "loss": 0.3748,
+      "step": 2204
+    },
+    {
+      "epoch": 1.026536312849162,
+      "grad_norm": 0.586746342396397,
+      "learning_rate": 3.921623644788217e-06,
+      "loss": 0.3912,
+      "step": 2205
+    },
+    {
+      "epoch": 1.0270018621973929,
+      "grad_norm": 0.6075737433083932,
+      "learning_rate": 3.920567988236014e-06,
+      "loss": 0.3906,
+      "step": 2206
+    },
+    {
+      "epoch": 1.027467411545624,
+      "grad_norm": 0.5837623957356346,
+      "learning_rate": 3.919511957472424e-06,
+      "loss": 0.3899,
+      "step": 2207
+    },
+    {
+      "epoch": 1.0279329608938548,
+      "grad_norm": 0.6107850944928549,
+      "learning_rate": 3.918455552775628e-06,
+      "loss": 0.3784,
+      "step": 2208
+    },
+    {
+      "epoch": 1.0283985102420856,
+      "grad_norm": 0.6190853349782715,
+      "learning_rate": 3.917398774423909e-06,
+      "loss": 0.3939,
+      "step": 2209
+    },
+    {
+      "epoch": 1.0288640595903167,
+      "grad_norm": 0.683407812144473,
+      "learning_rate": 3.916341622695647e-06,
+      "loss": 0.3796,
+      "step": 2210
+    },
+    {
+      "epoch": 1.0293296089385475,
+      "grad_norm": 0.5764906461785154,
+      "learning_rate": 3.915284097869322e-06,
+      "loss": 0.3679,
+      "step": 2211
+    },
+    {
+      "epoch": 1.0297951582867784,
+      "grad_norm": 0.6362049069937029,
+      "learning_rate": 3.91422620022351e-06,
+      "loss": 0.3977,
+      "step": 2212
+    },
+    {
+      "epoch": 1.0302607076350092,
+      "grad_norm": 0.6229506561368741,
+      "learning_rate": 3.913167930036888e-06,
+      "loss": 0.3851,
+      "step": 2213
+    },
+    {
+      "epoch": 1.0307262569832403,
+      "grad_norm": 0.6020780401797396,
+      "learning_rate": 3.912109287588227e-06,
+      "loss": 0.4045,
+      "step": 2214
+    },
+    {
+      "epoch": 1.0311918063314711,
+      "grad_norm": 0.6380131646063837,
+      "learning_rate": 3.911050273156401e-06,
+      "loss": 0.3963,
+      "step": 2215
+    },
+    {
+      "epoch": 1.031657355679702,
+      "grad_norm": 0.57856060620169,
+      "learning_rate": 3.909990887020379e-06,
+      "loss": 0.3933,
+      "step": 2216
+    },
+    {
+      "epoch": 1.032122905027933,
+      "grad_norm": 0.573127490846321,
+      "learning_rate": 3.908931129459228e-06,
+      "loss": 0.3885,
+      "step": 2217
+    },
+    {
+      "epoch": 1.0325884543761639,
+      "grad_norm": 0.605612014053347,
+      "learning_rate": 3.907871000752114e-06,
+      "loss": 0.3902,
+      "step": 2218
+    },
+    {
+      "epoch": 1.0330540037243947,
+      "grad_norm": 0.5872588072369137,
+      "learning_rate": 3.906810501178299e-06,
+      "loss": 0.4024,
+      "step": 2219
+    },
+    {
+      "epoch": 1.0335195530726258,
+      "grad_norm": 0.6311015420802459,
+      "learning_rate": 3.905749631017145e-06,
+      "loss": 0.3864,
+      "step": 2220
+    },
+    {
+      "epoch": 1.0339851024208566,
+      "grad_norm": 0.6613964711896778,
+      "learning_rate": 3.90468839054811e-06,
+      "loss": 0.4001,
+      "step": 2221
+    },
+    {
+      "epoch": 1.0344506517690875,
+      "grad_norm": 0.6419792753622416,
+      "learning_rate": 3.903626780050749e-06,
+      "loss": 0.3723,
+      "step": 2222
+    },
+    {
+      "epoch": 1.0349162011173185,
+      "grad_norm": 0.5850155734971118,
+      "learning_rate": 3.9025647998047165e-06,
+      "loss": 0.3894,
+      "step": 2223
+    },
+    {
+      "epoch": 1.0353817504655494,
+      "grad_norm": 0.5815352086123357,
+      "learning_rate": 3.9015024500897635e-06,
+      "loss": 0.3794,
+      "step": 2224
+    },
+    {
+      "epoch": 1.0358472998137802,
+      "grad_norm": 0.832532516123101,
+      "learning_rate": 3.900439731185738e-06,
+      "loss": 0.3862,
+      "step": 2225
+    },
+    {
+      "epoch": 1.0363128491620113,
+      "grad_norm": 0.5969495760918224,
+      "learning_rate": 3.899376643372585e-06,
+      "loss": 0.3766,
+      "step": 2226
+    },
+    {
+      "epoch": 1.0367783985102421,
+      "grad_norm": 0.5849089865195266,
+      "learning_rate": 3.898313186930347e-06,
+      "loss": 0.3852,
+      "step": 2227
+    },
+    {
+      "epoch": 1.037243947858473,
+      "grad_norm": 0.6054830081693764,
+      "learning_rate": 3.897249362139165e-06,
+      "loss": 0.384,
+      "step": 2228
+    },
+    {
+      "epoch": 1.0377094972067038,
+      "grad_norm": 0.5825885076693488,
+      "learning_rate": 3.896185169279274e-06,
+      "loss": 0.3795,
+      "step": 2229
+    },
+    {
+      "epoch": 1.0381750465549349,
+      "grad_norm": 0.6324841109156564,
+      "learning_rate": 3.8951206086310095e-06,
+      "loss": 0.4026,
+      "step": 2230
+    },
+    {
+      "epoch": 1.0386405959031657,
+      "grad_norm": 0.6384965711182182,
+      "learning_rate": 3.894055680474801e-06,
+      "loss": 0.4059,
+      "step": 2231
+    },
+    {
+      "epoch": 1.0391061452513966,
+      "grad_norm": 0.6784049565112068,
+      "learning_rate": 3.8929903850911755e-06,
+      "loss": 0.3761,
+      "step": 2232
+    },
+    {
+      "epoch": 1.0395716945996276,
+      "grad_norm": 0.653511254684347,
+      "learning_rate": 3.891924722760758e-06,
+      "loss": 0.3708,
+      "step": 2233
+    },
+    {
+      "epoch": 1.0400372439478585,
+      "grad_norm": 0.5857245018382524,
+      "learning_rate": 3.890858693764269e-06,
+      "loss": 0.3876,
+      "step": 2234
+    },
+    {
+      "epoch": 1.0405027932960893,
+      "grad_norm": 0.6272579961295238,
+      "learning_rate": 3.889792298382525e-06,
+      "loss": 0.4058,
+      "step": 2235
+    },
+    {
+      "epoch": 1.0409683426443204,
+      "grad_norm": 0.608397244515973,
+      "learning_rate": 3.888725536896443e-06,
+      "loss": 0.3817,
+      "step": 2236
+    },
+    {
+      "epoch": 1.0414338919925512,
+      "grad_norm": 0.6269090821468665,
+      "learning_rate": 3.887658409587029e-06,
+      "loss": 0.3792,
+      "step": 2237
+    },
+    {
+      "epoch": 1.041899441340782,
+      "grad_norm": 0.6121586264546627,
+      "learning_rate": 3.886590916735393e-06,
+      "loss": 0.3843,
+      "step": 2238
+    },
+    {
+      "epoch": 1.0423649906890131,
+      "grad_norm": 0.5800102978441041,
+      "learning_rate": 3.885523058622735e-06,
+      "loss": 0.3988,
+      "step": 2239
+    },
+    {
+      "epoch": 1.042830540037244,
+      "grad_norm": 0.6204927078992784,
+      "learning_rate": 3.884454835530358e-06,
+      "loss": 0.3896,
+      "step": 2240
+    },
+    {
+      "epoch": 1.0432960893854748,
+      "grad_norm": 0.6160582274890373,
+      "learning_rate": 3.883386247739655e-06,
+      "loss": 0.3787,
+      "step": 2241
+    },
+    {
+      "epoch": 1.0437616387337059,
+      "grad_norm": 0.5854706354066542,
+      "learning_rate": 3.882317295532118e-06,
+      "loss": 0.3829,
+      "step": 2242
+    },
+    {
+      "epoch": 1.0442271880819367,
+      "grad_norm": 0.5724938943572893,
+      "learning_rate": 3.881247979189334e-06,
+      "loss": 0.3961,
+      "step": 2243
+    },
+    {
+      "epoch": 1.0446927374301676,
+      "grad_norm": 0.6048260099006953,
+      "learning_rate": 3.8801782989929864e-06,
+      "loss": 0.3892,
+      "step": 2244
+    },
+    {
+      "epoch": 1.0451582867783986,
+      "grad_norm": 0.5817565747146728,
+      "learning_rate": 3.879108255224855e-06,
+      "loss": 0.3767,
+      "step": 2245
+    },
+    {
+      "epoch": 1.0456238361266295,
+      "grad_norm": 0.5914068039571075,
+      "learning_rate": 3.878037848166815e-06,
+      "loss": 0.3971,
+      "step": 2246
+    },
+    {
+      "epoch": 1.0460893854748603,
+      "grad_norm": 0.6591610574285767,
+      "learning_rate": 3.876967078100835e-06,
+      "loss": 0.3922,
+      "step": 2247
+    },
+    {
+      "epoch": 1.0465549348230911,
+      "grad_norm": 0.5779984485407371,
+      "learning_rate": 3.875895945308984e-06,
+      "loss": 0.3854,
+      "step": 2248
+    },
+    {
+      "epoch": 1.0470204841713222,
+      "grad_norm": 0.5963464751674495,
+      "learning_rate": 3.874824450073421e-06,
+      "loss": 0.3956,
+      "step": 2249
+    },
+    {
+      "epoch": 1.047486033519553,
+      "grad_norm": 0.6548578297761539,
+      "learning_rate": 3.873752592676407e-06,
+      "loss": 0.399,
+      "step": 2250
+    },
+    {
+      "epoch": 1.047951582867784,
+      "grad_norm": 0.6257717752128018,
+      "learning_rate": 3.872680373400292e-06,
+      "loss": 0.3822,
+      "step": 2251
+    },
+    {
+      "epoch": 1.048417132216015,
+      "grad_norm": 0.6093821325628432,
+      "learning_rate": 3.871607792527524e-06,
+      "loss": 0.3768,
+      "step": 2252
+    },
+    {
+      "epoch": 1.0488826815642458,
+      "grad_norm": 0.5941855624912107,
+      "learning_rate": 3.870534850340647e-06,
+      "loss": 0.3868,
+      "step": 2253
+    },
+    {
+      "epoch": 1.0493482309124766,
+      "grad_norm": 0.5841640492447651,
+      "learning_rate": 3.869461547122301e-06,
+      "loss": 0.369,
+      "step": 2254
+    },
+    {
+      "epoch": 1.0498137802607077,
+      "grad_norm": 0.6261329583657304,
+      "learning_rate": 3.8683878831552165e-06,
+      "loss": 0.3874,
+      "step": 2255
+    },
+    {
+      "epoch": 1.0502793296089385,
+      "grad_norm": 0.5706596655079026,
+      "learning_rate": 3.867313858722225e-06,
+      "loss": 0.3811,
+      "step": 2256
+    },
+    {
+      "epoch": 1.0507448789571694,
+      "grad_norm": 0.5797670971310801,
+      "learning_rate": 3.866239474106248e-06,
+      "loss": 0.3933,
+      "step": 2257
+    },
+    {
+      "epoch": 1.0512104283054005,
+      "grad_norm": 0.6251884767962488,
+      "learning_rate": 3.865164729590305e-06,
+      "loss": 0.379,
+      "step": 2258
+    },
+    {
+      "epoch": 1.0516759776536313,
+      "grad_norm": 0.610168439964162,
+      "learning_rate": 3.864089625457509e-06,
+      "loss": 0.392,
+      "step": 2259
+    },
+    {
+      "epoch": 1.0521415270018621,
+      "grad_norm": 0.5907604001104855,
+      "learning_rate": 3.8630141619910685e-06,
+      "loss": 0.3945,
+      "step": 2260
+    },
+    {
+      "epoch": 1.0526070763500932,
+      "grad_norm": 0.5712721268788196,
+      "learning_rate": 3.861938339474285e-06,
+      "loss": 0.3798,
+      "step": 2261
+    },
+    {
+      "epoch": 1.053072625698324,
+      "grad_norm": 0.5860212028183598,
+      "learning_rate": 3.860862158190556e-06,
+      "loss": 0.3801,
+      "step": 2262
+    },
+    {
+      "epoch": 1.053538175046555,
+      "grad_norm": 0.6328884975219718,
+      "learning_rate": 3.859785618423375e-06,
+      "loss": 0.4226,
+      "step": 2263
+    },
+    {
+      "epoch": 1.0540037243947857,
+      "grad_norm": 0.6173651244692683,
+      "learning_rate": 3.858708720456326e-06,
+      "loss": 0.3725,
+      "step": 2264
+    },
+    {
+      "epoch": 1.0544692737430168,
+      "grad_norm": 0.6051360916892318,
+      "learning_rate": 3.8576314645730906e-06,
+      "loss": 0.3987,
+      "step": 2265
+    },
+    {
+      "epoch": 1.0549348230912476,
+      "grad_norm": 0.6168708065115964,
+      "learning_rate": 3.856553851057443e-06,
+      "loss": 0.3831,
+      "step": 2266
+    },
+    {
+      "epoch": 1.0554003724394785,
+      "grad_norm": 0.6371905784411118,
+      "learning_rate": 3.8554758801932535e-06,
+      "loss": 0.3854,
+      "step": 2267
+    },
+    {
+      "epoch": 1.0558659217877095,
+      "grad_norm": 0.6736862592277953,
+      "learning_rate": 3.854397552264484e-06,
+      "loss": 0.3839,
+      "step": 2268
+    },
+    {
+      "epoch": 1.0563314711359404,
+      "grad_norm": 0.6019674633171703,
+      "learning_rate": 3.853318867555191e-06,
+      "loss": 0.4099,
+      "step": 2269
+    },
+    {
+      "epoch": 1.0567970204841712,
+      "grad_norm": 0.6212441229151344,
+      "learning_rate": 3.852239826349528e-06,
+      "loss": 0.4108,
+      "step": 2270
+    },
+    {
+      "epoch": 1.0572625698324023,
+      "grad_norm": 0.605565765494463,
+      "learning_rate": 3.8511604289317385e-06,
+      "loss": 0.3918,
+      "step": 2271
+    },
+    {
+      "epoch": 1.0577281191806331,
+      "grad_norm": 0.6137895249575104,
+      "learning_rate": 3.850080675586162e-06,
+      "loss": 0.3942,
+      "step": 2272
+    },
+    {
+      "epoch": 1.058193668528864,
+      "grad_norm": 0.5840265118950945,
+      "learning_rate": 3.849000566597231e-06,
+      "loss": 0.391,
+      "step": 2273
+    },
+    {
+      "epoch": 1.058659217877095,
+      "grad_norm": 0.651998696853593,
+      "learning_rate": 3.847920102249472e-06,
+      "loss": 0.3811,
+      "step": 2274
+    },
+    {
+      "epoch": 1.059124767225326,
+      "grad_norm": 0.6125996032093363,
+      "learning_rate": 3.846839282827505e-06,
+      "loss": 0.3871,
+      "step": 2275
+    },
+    {
+      "epoch": 1.0595903165735567,
+      "grad_norm": 0.5720176056781047,
+      "learning_rate": 3.8457581086160426e-06,
+      "loss": 0.3815,
+      "step": 2276
+    },
+    {
+      "epoch": 1.0600558659217878,
+      "grad_norm": 0.6346588900149914,
+      "learning_rate": 3.844676579899894e-06,
+      "loss": 0.3877,
+      "step": 2277
+    },
+    {
+      "epoch": 1.0605214152700186,
+      "grad_norm": 0.6535997963943597,
+      "learning_rate": 3.8435946969639575e-06,
+      "loss": 0.394,
+      "step": 2278
+    },
+    {
+      "epoch": 1.0609869646182495,
+      "grad_norm": 0.5806137306877823,
+      "learning_rate": 3.842512460093228e-06,
+      "loss": 0.3778,
+      "step": 2279
+    },
+    {
+      "epoch": 1.0614525139664805,
+      "grad_norm": 0.6412626678674298,
+      "learning_rate": 3.841429869572792e-06,
+      "loss": 0.3895,
+      "step": 2280
+    },
+    {
+      "epoch": 1.0619180633147114,
+      "grad_norm": 0.5997622023312044,
+      "learning_rate": 3.840346925687829e-06,
+      "loss": 0.382,
+      "step": 2281
+    },
+    {
+      "epoch": 1.0623836126629422,
+      "grad_norm": 0.5992239282610446,
+      "learning_rate": 3.839263628723613e-06,
+      "loss": 0.3909,
+      "step": 2282
+    },
+    {
+      "epoch": 1.0628491620111733,
+      "grad_norm": 0.6750518530223801,
+      "learning_rate": 3.838179978965509e-06,
+      "loss": 0.3911,
+      "step": 2283
+    },
+    {
+      "epoch": 1.0633147113594041,
+      "grad_norm": 0.632094302711144,
+      "learning_rate": 3.837095976698979e-06,
+      "loss": 0.3916,
+      "step": 2284
+    },
+    {
+      "epoch": 1.063780260707635,
+      "grad_norm": 0.5990551863770709,
+      "learning_rate": 3.836011622209572e-06,
+      "loss": 0.3981,
+      "step": 2285
+    },
+    {
+      "epoch": 1.0642458100558658,
+      "grad_norm": 0.6144779092472826,
+      "learning_rate": 3.8349269157829335e-06,
+      "loss": 0.3787,
+      "step": 2286
+    },
+    {
+      "epoch": 1.064711359404097,
+      "grad_norm": 0.6267162010134683,
+      "learning_rate": 3.833841857704801e-06,
+      "loss": 0.3956,
+      "step": 2287
+    },
+    {
+      "epoch": 1.0651769087523277,
+      "grad_norm": 0.5756729319990394,
+      "learning_rate": 3.832756448261006e-06,
+      "loss": 0.3948,
+      "step": 2288
+    },
+    {
+      "epoch": 1.0656424581005586,
+      "grad_norm": 0.5987123365262415,
+      "learning_rate": 3.831670687737469e-06,
+      "loss": 0.4035,
+      "step": 2289
+    },
+    {
+      "epoch": 1.0661080074487896,
+      "grad_norm": 0.6735755456260883,
+      "learning_rate": 3.830584576420207e-06,
+      "loss": 0.392,
+      "step": 2290
+    },
+    {
+      "epoch": 1.0665735567970205,
+      "grad_norm": 0.5919778100709616,
+      "learning_rate": 3.829498114595326e-06,
+      "loss": 0.4008,
+      "step": 2291
+    },
+    {
+      "epoch": 1.0670391061452513,
+      "grad_norm": 0.6476949891846233,
+      "learning_rate": 3.828411302549027e-06,
+      "loss": 0.416,
+      "step": 2292
+    },
+    {
+      "epoch": 1.0675046554934824,
+      "grad_norm": 0.6112876904082581,
+      "learning_rate": 3.827324140567601e-06,
+      "loss": 0.3979,
+      "step": 2293
+    },
+    {
+      "epoch": 1.0679702048417132,
+      "grad_norm": 0.5854044774998407,
+      "learning_rate": 3.826236628937434e-06,
+      "loss": 0.3915,
+      "step": 2294
+    },
+    {
+      "epoch": 1.068435754189944,
+      "grad_norm": 0.5990800000866627,
+      "learning_rate": 3.825148767945e-06,
+      "loss": 0.3818,
+      "step": 2295
+    },
+    {
+      "epoch": 1.0689013035381751,
+      "grad_norm": 0.5884383185971739,
+      "learning_rate": 3.82406055787687e-06,
+      "loss": 0.3923,
+      "step": 2296
+    },
+    {
+      "epoch": 1.069366852886406,
+      "grad_norm": 0.5795784814316307,
+      "learning_rate": 3.822971999019704e-06,
+      "loss": 0.3942,
+      "step": 2297
+    },
+    {
+      "epoch": 1.0698324022346368,
+      "grad_norm": 0.6400584667966389,
+      "learning_rate": 3.821883091660252e-06,
+      "loss": 0.376,
+      "step": 2298
+    },
+    {
+      "epoch": 1.0702979515828677,
+      "grad_norm": 0.6664557194337816,
+      "learning_rate": 3.820793836085361e-06,
+      "loss": 0.3792,
+      "step": 2299
+    },
+    {
+      "epoch": 1.0707635009310987,
+      "grad_norm": 0.6111854434204874,
+      "learning_rate": 3.819704232581964e-06,
+      "loss": 0.3583,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0712290502793296,
+      "grad_norm": 0.5969180730182829,
+      "learning_rate": 3.818614281437092e-06,
+      "loss": 0.39,
+      "step": 2301
+    },
+    {
+      "epoch": 1.0716945996275604,
+      "grad_norm": 0.6240385310444203,
+      "learning_rate": 3.8175239829378605e-06,
+      "loss": 0.388,
+      "step": 2302
+    },
+    {
+      "epoch": 1.0721601489757915,
+      "grad_norm": 0.6091223101549338,
+      "learning_rate": 3.816433337371482e-06,
+      "loss": 0.381,
+      "step": 2303
+    },
+    {
+      "epoch": 1.0726256983240223,
+      "grad_norm": 0.6553603942760294,
+      "learning_rate": 3.815342345025258e-06,
+      "loss": 0.3977,
+      "step": 2304
+    },
+    {
+      "epoch": 1.0730912476722532,
+      "grad_norm": 0.6247870461685063,
+      "learning_rate": 3.814251006186581e-06,
+      "loss": 0.4065,
+      "step": 2305
+    },
+    {
+      "epoch": 1.0735567970204842,
+      "grad_norm": 0.6138614569092468,
+      "learning_rate": 3.8131593211429384e-06,
+      "loss": 0.392,
+      "step": 2306
+    },
+    {
+      "epoch": 1.074022346368715,
+      "grad_norm": 0.6355659680214093,
+      "learning_rate": 3.8120672901819033e-06,
+      "loss": 0.3693,
+      "step": 2307
+    },
+    {
+      "epoch": 1.074487895716946,
+      "grad_norm": 0.5875555399061254,
+      "learning_rate": 3.8109749135911434e-06,
+      "loss": 0.4186,
+      "step": 2308
+    },
+    {
+      "epoch": 1.074953445065177,
+      "grad_norm": 0.5997740310573976,
+      "learning_rate": 3.809882191658416e-06,
+      "loss": 0.3705,
+      "step": 2309
+    },
+    {
+      "epoch": 1.0754189944134078,
+      "grad_norm": 0.6301331086628683,
+      "learning_rate": 3.8087891246715716e-06,
+      "loss": 0.3754,
+      "step": 2310
+    },
+    {
+      "epoch": 1.0758845437616387,
+      "grad_norm": 0.6256914602803356,
+      "learning_rate": 3.807695712918549e-06,
+      "loss": 0.3977,
+      "step": 2311
+    },
+    {
+      "epoch": 1.0763500931098697,
+      "grad_norm": 0.6065265853648849,
+      "learning_rate": 3.8066019566873787e-06,
+      "loss": 0.3922,
+      "step": 2312
+    },
+    {
+      "epoch": 1.0768156424581006,
+      "grad_norm": 0.6328470219004756,
+      "learning_rate": 3.805507856266182e-06,
+      "loss": 0.3976,
+      "step": 2313
+    },
+    {
+      "epoch": 1.0772811918063314,
+      "grad_norm": 0.5919892855062752,
+      "learning_rate": 3.8044134119431725e-06,
+      "loss": 0.3882,
+      "step": 2314
+    },
+    {
+      "epoch": 1.0777467411545625,
+      "grad_norm": 0.6154820990421553,
+      "learning_rate": 3.803318624006651e-06,
+      "loss": 0.3876,
+      "step": 2315
+    },
+    {
+      "epoch": 1.0782122905027933,
+      "grad_norm": 0.6097143145292566,
+      "learning_rate": 3.802223492745012e-06,
+      "loss": 0.3835,
+      "step": 2316
+    },
+    {
+      "epoch": 1.0786778398510242,
+      "grad_norm": 0.7061668927987823,
+      "learning_rate": 3.801128018446738e-06,
+      "loss": 0.4076,
+      "step": 2317
+    },
+    {
+      "epoch": 1.0791433891992552,
+      "grad_norm": 0.6256242854012491,
+      "learning_rate": 3.8000322014004038e-06,
+      "loss": 0.3902,
+      "step": 2318
+    },
+    {
+      "epoch": 1.079608938547486,
+      "grad_norm": 0.6526599299418107,
+      "learning_rate": 3.798936041894673e-06,
+      "loss": 0.397,
+      "step": 2319
+    },
+    {
+      "epoch": 1.080074487895717,
+      "grad_norm": 0.5941717167538063,
+      "learning_rate": 3.797839540218301e-06,
+      "loss": 0.3896,
+      "step": 2320
+    },
+    {
+      "epoch": 1.0805400372439478,
+      "grad_norm": 0.6014854542271993,
+      "learning_rate": 3.7967426966601316e-06,
+      "loss": 0.3885,
+      "step": 2321
+    },
+    {
+      "epoch": 1.0810055865921788,
+      "grad_norm": 0.6256489689590992,
+      "learning_rate": 3.7956455115090996e-06,
+      "loss": 0.3759,
+      "step": 2322
+    },
+    {
+      "epoch": 1.0814711359404097,
+      "grad_norm": 0.5950387652854947,
+      "learning_rate": 3.7945479850542297e-06,
+      "loss": 0.3901,
+      "step": 2323
+    },
+    {
+      "epoch": 1.0819366852886405,
+      "grad_norm": 0.5867080492718049,
+      "learning_rate": 3.7934501175846374e-06,
+      "loss": 0.3843,
+      "step": 2324
+    },
+    {
+      "epoch": 1.0824022346368716,
+      "grad_norm": 0.6015322484188937,
+      "learning_rate": 3.7923519093895257e-06,
+      "loss": 0.3987,
+      "step": 2325
+    },
+    {
+      "epoch": 1.0828677839851024,
+      "grad_norm": 0.6053721238793681,
+      "learning_rate": 3.791253360758189e-06,
+      "loss": 0.3866,
+      "step": 2326
+    },
+    {
+      "epoch": 1.0833333333333333,
+      "grad_norm": 0.6453696545669964,
+      "learning_rate": 3.7901544719800127e-06,
+      "loss": 0.3936,
+      "step": 2327
+    },
+    {
+      "epoch": 1.0837988826815643,
+      "grad_norm": 0.6428847955746725,
+      "learning_rate": 3.789055243344468e-06,
+      "loss": 0.3848,
+      "step": 2328
+    },
+    {
+      "epoch": 1.0842644320297952,
+      "grad_norm": 0.5936190355254658,
+      "learning_rate": 3.787955675141119e-06,
+      "loss": 0.3819,
+      "step": 2329
+    },
+    {
+      "epoch": 1.084729981378026,
+      "grad_norm": 0.6658609433580461,
+      "learning_rate": 3.7868557676596183e-06,
+      "loss": 0.386,
+      "step": 2330
+    },
+    {
+      "epoch": 1.085195530726257,
+      "grad_norm": 0.599786695242776,
+      "learning_rate": 3.7857555211897067e-06,
+      "loss": 0.3965,
+      "step": 2331
+    },
+    {
+      "epoch": 1.085661080074488,
+      "grad_norm": 0.6412752887105176,
+      "learning_rate": 3.784654936021217e-06,
+      "loss": 0.3933,
+      "step": 2332
+    },
+    {
+      "epoch": 1.0861266294227188,
+      "grad_norm": 0.6501627770276697,
+      "learning_rate": 3.7835540124440677e-06,
+      "loss": 0.3779,
+      "step": 2333
+    },
+    {
+      "epoch": 1.0865921787709498,
+      "grad_norm": 0.5869580344863473,
+      "learning_rate": 3.7824527507482694e-06,
+      "loss": 0.3972,
+      "step": 2334
+    },
+    {
+      "epoch": 1.0870577281191807,
+      "grad_norm": 0.5908454544915633,
+      "learning_rate": 3.7813511512239198e-06,
+      "loss": 0.3834,
+      "step": 2335
+    },
+    {
+      "epoch": 1.0875232774674115,
+      "grad_norm": 0.6013330848396993,
+      "learning_rate": 3.780249214161208e-06,
+      "loss": 0.3827,
+      "step": 2336
+    },
+    {
+      "epoch": 1.0879888268156424,
+      "grad_norm": 0.609707725791012,
+      "learning_rate": 3.7791469398504076e-06,
+      "loss": 0.407,
+      "step": 2337
+    },
+    {
+      "epoch": 1.0884543761638734,
+      "grad_norm": 0.6789493070006057,
+      "learning_rate": 3.778044328581886e-06,
+      "loss": 0.3837,
+      "step": 2338
+    },
+    {
+      "epoch": 1.0889199255121043,
+      "grad_norm": 0.6534517545710322,
+      "learning_rate": 3.7769413806460976e-06,
+      "loss": 0.3931,
+      "step": 2339
+    },
+    {
+      "epoch": 1.089385474860335,
+      "grad_norm": 0.6062220773835597,
+      "learning_rate": 3.775838096333584e-06,
+      "loss": 0.3756,
+      "step": 2340
+    },
+    {
+      "epoch": 1.0898510242085662,
+      "grad_norm": 0.6144473068840883,
+      "learning_rate": 3.7747344759349756e-06,
+      "loss": 0.3887,
+      "step": 2341
+    },
+    {
+      "epoch": 1.090316573556797,
+      "grad_norm": 0.6332948871226642,
+      "learning_rate": 3.7736305197409946e-06,
+      "loss": 0.371,
+      "step": 2342
+    },
+    {
+      "epoch": 1.0907821229050279,
+      "grad_norm": 0.6136533810789556,
+      "learning_rate": 3.772526228042448e-06,
+      "loss": 0.3763,
+      "step": 2343
+    },
+    {
+      "epoch": 1.091247672253259,
+      "grad_norm": 0.6192122097631174,
+      "learning_rate": 3.771421601130233e-06,
+      "loss": 0.3935,
+      "step": 2344
+    },
+    {
+      "epoch": 1.0917132216014898,
+      "grad_norm": 0.649201625745524,
+      "learning_rate": 3.770316639295333e-06,
+      "loss": 0.3734,
+      "step": 2345
+    },
+    {
+      "epoch": 1.0921787709497206,
+      "grad_norm": 0.638883211063833,
+      "learning_rate": 3.7692113428288236e-06,
+      "loss": 0.3851,
+      "step": 2346
+    },
+    {
+      "epoch": 1.0926443202979517,
+      "grad_norm": 0.6259450139845935,
+      "learning_rate": 3.7681057120218644e-06,
+      "loss": 0.3797,
+      "step": 2347
+    },
+    {
+      "epoch": 1.0931098696461825,
+      "grad_norm": 0.6328836345242561,
+      "learning_rate": 3.766999747165706e-06,
+      "loss": 0.3947,
+      "step": 2348
+    },
+    {
+      "epoch": 1.0935754189944134,
+      "grad_norm": 0.661802160879681,
+      "learning_rate": 3.765893448551685e-06,
+      "loss": 0.3926,
+      "step": 2349
+    },
+    {
+      "epoch": 1.0940409683426444,
+      "grad_norm": 0.6232875552442119,
+      "learning_rate": 3.7647868164712277e-06,
+      "loss": 0.3976,
+      "step": 2350
+    },
+    {
+      "epoch": 1.0945065176908753,
+      "grad_norm": 0.6209330652006276,
+      "learning_rate": 3.7636798512158455e-06,
+      "loss": 0.3986,
+      "step": 2351
+    },
+    {
+      "epoch": 1.094972067039106,
+      "grad_norm": 0.6338077307344969,
+      "learning_rate": 3.762572553077142e-06,
+      "loss": 0.3756,
+      "step": 2352
+    },
+    {
+      "epoch": 1.0954376163873372,
+      "grad_norm": 0.6147348352310245,
+      "learning_rate": 3.7614649223468036e-06,
+      "loss": 0.3929,
+      "step": 2353
+    },
+    {
+      "epoch": 1.095903165735568,
+      "grad_norm": 0.6186795229387809,
+      "learning_rate": 3.7603569593166072e-06,
+      "loss": 0.3966,
+      "step": 2354
+    },
+    {
+      "epoch": 1.0963687150837989,
+      "grad_norm": 0.5829736764014936,
+      "learning_rate": 3.7592486642784175e-06,
+      "loss": 0.3776,
+      "step": 2355
+    },
+    {
+      "epoch": 1.0968342644320297,
+      "grad_norm": 0.5842054186099753,
+      "learning_rate": 3.758140037524185e-06,
+      "loss": 0.3705,
+      "step": 2356
+    },
+    {
+      "epoch": 1.0972998137802608,
+      "grad_norm": 0.6521980174419296,
+      "learning_rate": 3.757031079345948e-06,
+      "loss": 0.3848,
+      "step": 2357
+    },
+    {
+      "epoch": 1.0977653631284916,
+      "grad_norm": 0.594459317362286,
+      "learning_rate": 3.7559217900358324e-06,
+      "loss": 0.3797,
+      "step": 2358
+    },
+    {
+      "epoch": 1.0982309124767224,
+      "grad_norm": 0.5817053661882128,
+      "learning_rate": 3.754812169886054e-06,
+      "loss": 0.397,
+      "step": 2359
+    },
+    {
+      "epoch": 1.0986964618249535,
+      "grad_norm": 0.6127800700471905,
+      "learning_rate": 3.753702219188909e-06,
+      "loss": 0.4113,
+      "step": 2360
+    },
+    {
+      "epoch": 1.0991620111731844,
+      "grad_norm": 0.5851899854635639,
+      "learning_rate": 3.752591938236787e-06,
+      "loss": 0.3905,
+      "step": 2361
+    },
+    {
+      "epoch": 1.0996275605214152,
+      "grad_norm": 0.5719532251674811,
+      "learning_rate": 3.7514813273221628e-06,
+      "loss": 0.3654,
+      "step": 2362
+    },
+    {
+      "epoch": 1.1000931098696463,
+      "grad_norm": 0.5843259187008193,
+      "learning_rate": 3.750370386737597e-06,
+      "loss": 0.3862,
+      "step": 2363
+    },
+    {
+      "epoch": 1.100558659217877,
+      "grad_norm": 0.6540015611214748,
+      "learning_rate": 3.749259116775738e-06,
+      "loss": 0.3949,
+      "step": 2364
+    },
+    {
+      "epoch": 1.101024208566108,
+      "grad_norm": 0.6148220056145619,
+      "learning_rate": 3.748147517729321e-06,
+      "loss": 0.3927,
+      "step": 2365
+    },
+    {
+      "epoch": 1.101489757914339,
+      "grad_norm": 0.6124519310422861,
+      "learning_rate": 3.7470355898911664e-06,
+      "loss": 0.3845,
+      "step": 2366
+    },
+    {
+      "epoch": 1.1019553072625698,
+      "grad_norm": 0.6421715030051303,
+      "learning_rate": 3.745923333554184e-06,
+      "loss": 0.3999,
+      "step": 2367
+    },
+    {
+      "epoch": 1.1024208566108007,
+      "grad_norm": 0.601463482956146,
+      "learning_rate": 3.7448107490113672e-06,
+      "loss": 0.3668,
+      "step": 2368
+    },
+    {
+      "epoch": 1.1028864059590318,
+      "grad_norm": 0.6130035673725254,
+      "learning_rate": 3.743697836555799e-06,
+      "loss": 0.3969,
+      "step": 2369
+    },
+    {
+      "epoch": 1.1033519553072626,
+      "grad_norm": 0.5809462688211146,
+      "learning_rate": 3.742584596480645e-06,
+      "loss": 0.3908,
+      "step": 2370
+    },
+    {
+      "epoch": 1.1038175046554934,
+      "grad_norm": 0.5885255647244889,
+      "learning_rate": 3.7414710290791602e-06,
+      "loss": 0.3899,
+      "step": 2371
+    },
+    {
+      "epoch": 1.1042830540037243,
+      "grad_norm": 0.6057587571097205,
+      "learning_rate": 3.7403571346446843e-06,
+      "loss": 0.3714,
+      "step": 2372
+    },
+    {
+      "epoch": 1.1047486033519553,
+      "grad_norm": 0.6608833959191581,
+      "learning_rate": 3.7392429134706444e-06,
+      "loss": 0.3796,
+      "step": 2373
+    },
+    {
+      "epoch": 1.1052141527001862,
+      "grad_norm": 0.661493246244545,
+      "learning_rate": 3.7381283658505517e-06,
+      "loss": 0.4033,
+      "step": 2374
+    },
+    {
+      "epoch": 1.105679702048417,
+      "grad_norm": 0.6148173289177613,
+      "learning_rate": 3.737013492078006e-06,
+      "loss": 0.3955,
+      "step": 2375
+    },
+    {
+      "epoch": 1.106145251396648,
+      "grad_norm": 0.6277488823935631,
+      "learning_rate": 3.73589829244669e-06,
+      "loss": 0.3745,
+      "step": 2376
+    },
+    {
+      "epoch": 1.106610800744879,
+      "grad_norm": 0.6007861765970887,
+      "learning_rate": 3.7347827672503745e-06,
+      "loss": 0.3945,
+      "step": 2377
+    },
+    {
+      "epoch": 1.1070763500931098,
+      "grad_norm": 0.663664288598696,
+      "learning_rate": 3.7336669167829153e-06,
+      "loss": 0.3831,
+      "step": 2378
+    },
+    {
+      "epoch": 1.1075418994413408,
+      "grad_norm": 0.572526731113263,
+      "learning_rate": 3.7325507413382546e-06,
+      "loss": 0.3821,
+      "step": 2379
+    },
+    {
+      "epoch": 1.1080074487895717,
+      "grad_norm": 0.57723617135398,
+      "learning_rate": 3.731434241210418e-06,
+      "loss": 0.3802,
+      "step": 2380
+    },
+    {
+      "epoch": 1.1084729981378025,
+      "grad_norm": 0.6038562755266872,
+      "learning_rate": 3.73031741669352e-06,
+      "loss": 0.3897,
+      "step": 2381
+    },
+    {
+      "epoch": 1.1089385474860336,
+      "grad_norm": 0.588789185925787,
+      "learning_rate": 3.7292002680817572e-06,
+      "loss": 0.3889,
+      "step": 2382
+    },
+    {
+      "epoch": 1.1094040968342644,
+      "grad_norm": 0.6123296522007534,
+      "learning_rate": 3.7280827956694143e-06,
+      "loss": 0.3785,
+      "step": 2383
+    },
+    {
+      "epoch": 1.1098696461824953,
+      "grad_norm": 0.6286944955673213,
+      "learning_rate": 3.7269649997508594e-06,
+      "loss": 0.3775,
+      "step": 2384
+    },
+    {
+      "epoch": 1.1103351955307263,
+      "grad_norm": 0.6163744238714421,
+      "learning_rate": 3.725846880620547e-06,
+      "loss": 0.3856,
+      "step": 2385
+    },
+    {
+      "epoch": 1.1108007448789572,
+      "grad_norm": 0.6563813528888299,
+      "learning_rate": 3.7247284385730154e-06,
+      "loss": 0.413,
+      "step": 2386
+    },
+    {
+      "epoch": 1.111266294227188,
+      "grad_norm": 0.5987721632000497,
+      "learning_rate": 3.7236096739028906e-06,
+      "loss": 0.3747,
+      "step": 2387
+    },
+    {
+      "epoch": 1.111731843575419,
+      "grad_norm": 0.6174544270877214,
+      "learning_rate": 3.7224905869048792e-06,
+      "loss": 0.3811,
+      "step": 2388
+    },
+    {
+      "epoch": 1.11219739292365,
+      "grad_norm": 0.6125059409799302,
+      "learning_rate": 3.7213711778737776e-06,
+      "loss": 0.3978,
+      "step": 2389
+    },
+    {
+      "epoch": 1.1126629422718808,
+      "grad_norm": 0.6202494389618071,
+      "learning_rate": 3.7202514471044632e-06,
+      "loss": 0.3642,
+      "step": 2390
+    },
+    {
+      "epoch": 1.1131284916201118,
+      "grad_norm": 0.6541228640711623,
+      "learning_rate": 3.719131394891901e-06,
+      "loss": 0.3905,
+      "step": 2391
+    },
+    {
+      "epoch": 1.1135940409683427,
+      "grad_norm": 0.6130340360395656,
+      "learning_rate": 3.7180110215311395e-06,
+      "loss": 0.396,
+      "step": 2392
+    },
+    {
+      "epoch": 1.1140595903165735,
+      "grad_norm": 0.6229502742140445,
+      "learning_rate": 3.7168903273173097e-06,
+      "loss": 0.3964,
+      "step": 2393
+    },
+    {
+      "epoch": 1.1145251396648044,
+      "grad_norm": 0.5842656808494603,
+      "learning_rate": 3.715769312545631e-06,
+      "loss": 0.3941,
+      "step": 2394
+    },
+    {
+      "epoch": 1.1149906890130354,
+      "grad_norm": 0.6194046206697349,
+      "learning_rate": 3.714647977511405e-06,
+      "loss": 0.4064,
+      "step": 2395
+    },
+    {
+      "epoch": 1.1154562383612663,
+      "grad_norm": 0.6312924566467233,
+      "learning_rate": 3.7135263225100173e-06,
+      "loss": 0.3791,
+      "step": 2396
+    },
+    {
+      "epoch": 1.1159217877094971,
+      "grad_norm": 0.5989162158021278,
+      "learning_rate": 3.712404347836939e-06,
+      "loss": 0.3868,
+      "step": 2397
+    },
+    {
+      "epoch": 1.1163873370577282,
+      "grad_norm": 0.6183269222451595,
+      "learning_rate": 3.7112820537877253e-06,
+      "loss": 0.3871,
+      "step": 2398
+    },
+    {
+      "epoch": 1.116852886405959,
+      "grad_norm": 0.5975545092970598,
+      "learning_rate": 3.7101594406580145e-06,
+      "loss": 0.3924,
+      "step": 2399
+    },
+    {
+      "epoch": 1.1173184357541899,
+      "grad_norm": 0.5838720477517484,
+      "learning_rate": 3.70903650874353e-06,
+      "loss": 0.3741,
+      "step": 2400
+    },
+    {
+      "epoch": 1.117783985102421,
+      "grad_norm": 0.639381416795744,
+      "learning_rate": 3.707913258340078e-06,
+      "loss": 0.3704,
+      "step": 2401
+    },
+    {
+      "epoch": 1.1182495344506518,
+      "grad_norm": 0.6275961947068434,
+      "learning_rate": 3.7067896897435505e-06,
+      "loss": 0.3882,
+      "step": 2402
+    },
+    {
+      "epoch": 1.1187150837988826,
+      "grad_norm": 0.6278103931379224,
+      "learning_rate": 3.705665803249921e-06,
+      "loss": 0.3766,
+      "step": 2403
+    },
+    {
+      "epoch": 1.1191806331471137,
+      "grad_norm": 0.5998401481285707,
+      "learning_rate": 3.704541599155249e-06,
+      "loss": 0.3802,
+      "step": 2404
+    },
+    {
+      "epoch": 1.1196461824953445,
+      "grad_norm": 0.5839080730193925,
+      "learning_rate": 3.7034170777556766e-06,
+      "loss": 0.3781,
+      "step": 2405
+    },
+    {
+      "epoch": 1.1201117318435754,
+      "grad_norm": 0.5913981647867366,
+      "learning_rate": 3.702292239347429e-06,
+      "loss": 0.3841,
+      "step": 2406
+    },
+    {
+      "epoch": 1.1205772811918062,
+      "grad_norm": 0.6390870766046306,
+      "learning_rate": 3.7011670842268145e-06,
+      "loss": 0.3769,
+      "step": 2407
+    },
+    {
+      "epoch": 1.1210428305400373,
+      "grad_norm": 0.5932151945666609,
+      "learning_rate": 3.700041612690227e-06,
+      "loss": 0.3778,
+      "step": 2408
+    },
+    {
+      "epoch": 1.1215083798882681,
+      "grad_norm": 0.6289078049249276,
+      "learning_rate": 3.698915825034142e-06,
+      "loss": 0.404,
+      "step": 2409
+    },
+    {
+      "epoch": 1.121973929236499,
+      "grad_norm": 0.6057362251624532,
+      "learning_rate": 3.6977897215551194e-06,
+      "loss": 0.3784,
+      "step": 2410
+    },
+    {
+      "epoch": 1.12243947858473,
+      "grad_norm": 0.5834952965338346,
+      "learning_rate": 3.6966633025498012e-06,
+      "loss": 0.3811,
+      "step": 2411
+    },
+    {
+      "epoch": 1.1229050279329609,
+      "grad_norm": 0.6165606646494006,
+      "learning_rate": 3.6955365683149126e-06,
+      "loss": 0.3901,
+      "step": 2412
+    },
+    {
+      "epoch": 1.1233705772811917,
+      "grad_norm": 0.6943125177404206,
+      "learning_rate": 3.694409519147262e-06,
+      "loss": 0.4006,
+      "step": 2413
+    },
+    {
+      "epoch": 1.1238361266294228,
+      "grad_norm": 0.5998981026855151,
+      "learning_rate": 3.693282155343742e-06,
+      "loss": 0.3922,
+      "step": 2414
+    },
+    {
+      "epoch": 1.1243016759776536,
+      "grad_norm": 0.6081947992688457,
+      "learning_rate": 3.6921544772013257e-06,
+      "loss": 0.3778,
+      "step": 2415
+    },
+    {
+      "epoch": 1.1247672253258845,
+      "grad_norm": 0.5838774473106033,
+      "learning_rate": 3.6910264850170713e-06,
+      "loss": 0.3776,
+      "step": 2416
+    },
+    {
+      "epoch": 1.1252327746741155,
+      "grad_norm": 0.5924286294707599,
+      "learning_rate": 3.689898179088119e-06,
+      "loss": 0.394,
+      "step": 2417
+    },
+    {
+      "epoch": 1.1256983240223464,
+      "grad_norm": 0.5988459517417096,
+      "learning_rate": 3.68876955971169e-06,
+      "loss": 0.3843,
+      "step": 2418
+    },
+    {
+      "epoch": 1.1261638733705772,
+      "grad_norm": 0.5898112071891567,
+      "learning_rate": 3.6876406271850913e-06,
+      "loss": 0.3782,
+      "step": 2419
+    },
+    {
+      "epoch": 1.1266294227188083,
+      "grad_norm": 0.6007290495347126,
+      "learning_rate": 3.686511381805709e-06,
+      "loss": 0.3842,
+      "step": 2420
+    },
+    {
+      "epoch": 1.1270949720670391,
+      "grad_norm": 0.5997740814108702,
+      "learning_rate": 3.6853818238710143e-06,
+      "loss": 0.3861,
+      "step": 2421
+    },
+    {
+      "epoch": 1.12756052141527,
+      "grad_norm": 0.6237815473301633,
+      "learning_rate": 3.684251953678559e-06,
+      "loss": 0.3939,
+      "step": 2422
+    },
+    {
+      "epoch": 1.128026070763501,
+      "grad_norm": 0.6157141660849706,
+      "learning_rate": 3.6831217715259776e-06,
+      "loss": 0.383,
+      "step": 2423
+    },
+    {
+      "epoch": 1.1284916201117319,
+      "grad_norm": 0.6596458392350197,
+      "learning_rate": 3.6819912777109885e-06,
+      "loss": 0.4022,
+      "step": 2424
+    },
+    {
+      "epoch": 1.1289571694599627,
+      "grad_norm": 0.6123069097584597,
+      "learning_rate": 3.6808604725313886e-06,
+      "loss": 0.3867,
+      "step": 2425
+    },
+    {
+      "epoch": 1.1294227188081938,
+      "grad_norm": 0.6383124772909652,
+      "learning_rate": 3.6797293562850595e-06,
+      "loss": 0.4033,
+      "step": 2426
+    },
+    {
+      "epoch": 1.1298882681564246,
+      "grad_norm": 0.6064357407494705,
+      "learning_rate": 3.678597929269965e-06,
+      "loss": 0.3835,
+      "step": 2427
+    },
+    {
+      "epoch": 1.1303538175046555,
+      "grad_norm": 0.5785162134328699,
+      "learning_rate": 3.6774661917841486e-06,
+      "loss": 0.3698,
+      "step": 2428
+    },
+    {
+      "epoch": 1.1308193668528863,
+      "grad_norm": 0.6349637198135673,
+      "learning_rate": 3.676334144125738e-06,
+      "loss": 0.394,
+      "step": 2429
+    },
+    {
+      "epoch": 1.1312849162011174,
+      "grad_norm": 0.6034410558086074,
+      "learning_rate": 3.6752017865929407e-06,
+      "loss": 0.3878,
+      "step": 2430
+    },
+    {
+      "epoch": 1.1317504655493482,
+      "grad_norm": 0.5439434155157072,
+      "learning_rate": 3.674069119484047e-06,
+      "loss": 0.3932,
+      "step": 2431
+    },
+    {
+      "epoch": 1.132216014897579,
+      "grad_norm": 0.6020218609053011,
+      "learning_rate": 3.6729361430974282e-06,
+      "loss": 0.3857,
+      "step": 2432
+    },
+    {
+      "epoch": 1.1326815642458101,
+      "grad_norm": 0.6114365316262818,
+      "learning_rate": 3.671802857731537e-06,
+      "loss": 0.3751,
+      "step": 2433
+    },
+    {
+      "epoch": 1.133147113594041,
+      "grad_norm": 0.6320301598158548,
+      "learning_rate": 3.670669263684909e-06,
+      "loss": 0.38,
+      "step": 2434
+    },
+    {
+      "epoch": 1.1336126629422718,
+      "grad_norm": 0.606112669544099,
+      "learning_rate": 3.669535361256157e-06,
+      "loss": 0.3843,
+      "step": 2435
+    },
+    {
+      "epoch": 1.1340782122905029,
+      "grad_norm": 0.5734373055493598,
+      "learning_rate": 3.668401150743981e-06,
+      "loss": 0.3551,
+      "step": 2436
+    },
+    {
+      "epoch": 1.1345437616387337,
+      "grad_norm": 0.5860001495088283,
+      "learning_rate": 3.6672666324471573e-06,
+      "loss": 0.3925,
+      "step": 2437
+    },
+    {
+      "epoch": 1.1350093109869646,
+      "grad_norm": 0.6054888270526806,
+      "learning_rate": 3.6661318066645455e-06,
+      "loss": 0.384,
+      "step": 2438
+    },
+    {
+      "epoch": 1.1354748603351956,
+      "grad_norm": 0.5894893477470247,
+      "learning_rate": 3.6649966736950856e-06,
+      "loss": 0.3792,
+      "step": 2439
+    },
+    {
+      "epoch": 1.1359404096834265,
+      "grad_norm": 0.6100653855027889,
+      "learning_rate": 3.6638612338377987e-06,
+      "loss": 0.3867,
+      "step": 2440
+    },
+    {
+      "epoch": 1.1364059590316573,
+      "grad_norm": 0.6149461072020268,
+      "learning_rate": 3.662725487391786e-06,
+      "loss": 0.4036,
+      "step": 2441
+    },
+    {
+      "epoch": 1.1368715083798882,
+      "grad_norm": 0.6268705777772924,
+      "learning_rate": 3.661589434656231e-06,
+      "loss": 0.3849,
+      "step": 2442
+    },
+    {
+      "epoch": 1.1373370577281192,
+      "grad_norm": 0.6586493566035647,
+      "learning_rate": 3.6604530759303967e-06,
+      "loss": 0.4028,
+      "step": 2443
+    },
+    {
+      "epoch": 1.13780260707635,
+      "grad_norm": 0.6243703514164627,
+      "learning_rate": 3.659316411513627e-06,
+      "loss": 0.3771,
+      "step": 2444
+    },
+    {
+      "epoch": 1.138268156424581,
+      "grad_norm": 0.5856479894883921,
+      "learning_rate": 3.658179441705346e-06,
+      "loss": 0.3909,
+      "step": 2445
+    },
+    {
+      "epoch": 1.138733705772812,
+      "grad_norm": 0.624409081291879,
+      "learning_rate": 3.6570421668050592e-06,
+      "loss": 0.3888,
+      "step": 2446
+    },
+    {
+      "epoch": 1.1391992551210428,
+      "grad_norm": 0.5750073929440698,
+      "learning_rate": 3.6559045871123524e-06,
+      "loss": 0.3871,
+      "step": 2447
+    },
+    {
+      "epoch": 1.1396648044692737,
+      "grad_norm": 0.6387440030490878,
+      "learning_rate": 3.654766702926889e-06,
+      "loss": 0.3837,
+      "step": 2448
+    },
+    {
+      "epoch": 1.1401303538175047,
+      "grad_norm": 0.6434373776771272,
+      "learning_rate": 3.653628514548417e-06,
+      "loss": 0.4012,
+      "step": 2449
+    },
+    {
+      "epoch": 1.1405959031657356,
+      "grad_norm": 0.6249826953648545,
+      "learning_rate": 3.6524900222767606e-06,
+      "loss": 0.3717,
+      "step": 2450
+    },
+    {
+      "epoch": 1.1410614525139664,
+      "grad_norm": 0.5667241240237776,
+      "learning_rate": 3.6513512264118268e-06,
+      "loss": 0.3785,
+      "step": 2451
+    },
+    {
+      "epoch": 1.1415270018621975,
+      "grad_norm": 0.5779801720173531,
+      "learning_rate": 3.6502121272536007e-06,
+      "loss": 0.3992,
+      "step": 2452
+    },
+    {
+      "epoch": 1.1419925512104283,
+      "grad_norm": 0.6002494649157084,
+      "learning_rate": 3.6490727251021496e-06,
+      "loss": 0.3995,
+      "step": 2453
+    },
+    {
+      "epoch": 1.1424581005586592,
+      "grad_norm": 0.6132352339445165,
+      "learning_rate": 3.647933020257617e-06,
+      "loss": 0.3741,
+      "step": 2454
+    },
+    {
+      "epoch": 1.1429236499068902,
+      "grad_norm": 0.5605117779660849,
+      "learning_rate": 3.6467930130202305e-06,
+      "loss": 0.3901,
+      "step": 2455
+    },
+    {
+      "epoch": 1.143389199255121,
+      "grad_norm": 0.6144519984583274,
+      "learning_rate": 3.6456527036902935e-06,
+      "loss": 0.3742,
+      "step": 2456
+    },
+    {
+      "epoch": 1.143854748603352,
+      "grad_norm": 0.6006201223383391,
+      "learning_rate": 3.644512092568191e-06,
+      "loss": 0.3943,
+      "step": 2457
+    },
+    {
+      "epoch": 1.144320297951583,
+      "grad_norm": 0.619556888590478,
+      "learning_rate": 3.643371179954387e-06,
+      "loss": 0.3955,
+      "step": 2458
+    },
+    {
+      "epoch": 1.1447858472998138,
+      "grad_norm": 0.6214638831595659,
+      "learning_rate": 3.642229966149425e-06,
+      "loss": 0.3852,
+      "step": 2459
+    },
+    {
+      "epoch": 1.1452513966480447,
+      "grad_norm": 0.6730562302380475,
+      "learning_rate": 3.641088451453929e-06,
+      "loss": 0.4022,
+      "step": 2460
+    },
+    {
+      "epoch": 1.1457169459962757,
+      "grad_norm": 0.6059221119653482,
+      "learning_rate": 3.6399466361685996e-06,
+      "loss": 0.377,
+      "step": 2461
+    },
+    {
+      "epoch": 1.1461824953445066,
+      "grad_norm": 0.6126875342986506,
+      "learning_rate": 3.6388045205942187e-06,
+      "loss": 0.4081,
+      "step": 2462
+    },
+    {
+      "epoch": 1.1466480446927374,
+      "grad_norm": 0.6083141659419482,
+      "learning_rate": 3.637662105031647e-06,
+      "loss": 0.3872,
+      "step": 2463
+    },
+    {
+      "epoch": 1.1471135940409685,
+      "grad_norm": 0.608603847936783,
+      "learning_rate": 3.636519389781823e-06,
+      "loss": 0.3946,
+      "step": 2464
+    },
+    {
+      "epoch": 1.1475791433891993,
+      "grad_norm": 0.5883836851916848,
+      "learning_rate": 3.6353763751457654e-06,
+      "loss": 0.3973,
+      "step": 2465
+    },
+    {
+      "epoch": 1.1480446927374302,
+      "grad_norm": 0.5834318433971643,
+      "learning_rate": 3.6342330614245726e-06,
+      "loss": 0.3932,
+      "step": 2466
+    },
+    {
+      "epoch": 1.148510242085661,
+      "grad_norm": 0.6035282993254334,
+      "learning_rate": 3.633089448919419e-06,
+      "loss": 0.3852,
+      "step": 2467
+    },
+    {
+      "epoch": 1.148975791433892,
+      "grad_norm": 0.6039796757551313,
+      "learning_rate": 3.63194553793156e-06,
+      "loss": 0.3785,
+      "step": 2468
+    },
+    {
+      "epoch": 1.149441340782123,
+      "grad_norm": 0.5839974668346636,
+      "learning_rate": 3.6308013287623294e-06,
+      "loss": 0.3831,
+      "step": 2469
+    },
+    {
+      "epoch": 1.1499068901303537,
+      "grad_norm": 0.6059746208076118,
+      "learning_rate": 3.6296568217131388e-06,
+      "loss": 0.3785,
+      "step": 2470
+    },
+    {
+      "epoch": 1.1503724394785848,
+      "grad_norm": 0.5922907041679946,
+      "learning_rate": 3.6285120170854783e-06,
+      "loss": 0.3879,
+      "step": 2471
+    },
+    {
+      "epoch": 1.1508379888268156,
+      "grad_norm": 0.6129948498438077,
+      "learning_rate": 3.6273669151809166e-06,
+      "loss": 0.407,
+      "step": 2472
+    },
+    {
+      "epoch": 1.1513035381750465,
+      "grad_norm": 0.5720903442337816,
+      "learning_rate": 3.626221516301102e-06,
+      "loss": 0.3755,
+      "step": 2473
+    },
+    {
+      "epoch": 1.1517690875232776,
+      "grad_norm": 0.5603663495863912,
+      "learning_rate": 3.625075820747757e-06,
+      "loss": 0.374,
+      "step": 2474
+    },
+    {
+      "epoch": 1.1522346368715084,
+      "grad_norm": 0.6242585765152858,
+      "learning_rate": 3.6239298288226883e-06,
+      "loss": 0.3731,
+      "step": 2475
+    },
+    {
+      "epoch": 1.1527001862197392,
+      "grad_norm": 0.6090403214165724,
+      "learning_rate": 3.6227835408277755e-06,
+      "loss": 0.408,
+      "step": 2476
+    },
+    {
+      "epoch": 1.15316573556797,
+      "grad_norm": 0.6018474565505014,
+      "learning_rate": 3.621636957064979e-06,
+      "loss": 0.3839,
+      "step": 2477
+    },
+    {
+      "epoch": 1.1536312849162011,
+      "grad_norm": 0.598510991711065,
+      "learning_rate": 3.620490077836335e-06,
+      "loss": 0.4038,
+      "step": 2478
+    },
+    {
+      "epoch": 1.154096834264432,
+      "grad_norm": 0.6274704298708073,
+      "learning_rate": 3.61934290344396e-06,
+      "loss": 0.4062,
+      "step": 2479
+    },
+    {
+      "epoch": 1.1545623836126628,
+      "grad_norm": 0.593831749583663,
+      "learning_rate": 3.618195434190047e-06,
+      "loss": 0.3776,
+      "step": 2480
+    },
+    {
+      "epoch": 1.155027932960894,
+      "grad_norm": 0.6133499915286649,
+      "learning_rate": 3.617047670376866e-06,
+      "loss": 0.3953,
+      "step": 2481
+    },
+    {
+      "epoch": 1.1554934823091247,
+      "grad_norm": 0.5979916232904934,
+      "learning_rate": 3.615899612306765e-06,
+      "loss": 0.3856,
+      "step": 2482
+    },
+    {
+      "epoch": 1.1559590316573556,
+      "grad_norm": 0.6409386174443789,
+      "learning_rate": 3.6147512602821716e-06,
+      "loss": 0.4089,
+      "step": 2483
+    },
+    {
+      "epoch": 1.1564245810055866,
+      "grad_norm": 0.6068815412328666,
+      "learning_rate": 3.613602614605587e-06,
+      "loss": 0.3614,
+      "step": 2484
+    },
+    {
+      "epoch": 1.1568901303538175,
+      "grad_norm": 0.6092062765208821,
+      "learning_rate": 3.6124536755795924e-06,
+      "loss": 0.387,
+      "step": 2485
+    },
+    {
+      "epoch": 1.1573556797020483,
+      "grad_norm": 0.7980923231233188,
+      "learning_rate": 3.611304443506847e-06,
+      "loss": 0.3925,
+      "step": 2486
+    },
+    {
+      "epoch": 1.1578212290502794,
+      "grad_norm": 0.6255526771341606,
+      "learning_rate": 3.610154918690083e-06,
+      "loss": 0.3871,
+      "step": 2487
+    },
+    {
+      "epoch": 1.1582867783985102,
+      "grad_norm": 0.6041674129738982,
+      "learning_rate": 3.609005101432116e-06,
+      "loss": 0.3624,
+      "step": 2488
+    },
+    {
+      "epoch": 1.158752327746741,
+      "grad_norm": 0.5619520574960076,
+      "learning_rate": 3.6078549920358325e-06,
+      "loss": 0.376,
+      "step": 2489
+    },
+    {
+      "epoch": 1.1592178770949721,
+      "grad_norm": 0.6221899864209056,
+      "learning_rate": 3.6067045908042e-06,
+      "loss": 0.3903,
+      "step": 2490
+    },
+    {
+      "epoch": 1.159683426443203,
+      "grad_norm": 0.6130860671155997,
+      "learning_rate": 3.6055538980402604e-06,
+      "loss": 0.3705,
+      "step": 2491
+    },
+    {
+      "epoch": 1.1601489757914338,
+      "grad_norm": 0.5904305318275997,
+      "learning_rate": 3.604402914047136e-06,
+      "loss": 0.3866,
+      "step": 2492
+    },
+    {
+      "epoch": 1.160614525139665,
+      "grad_norm": 0.5617471734313527,
+      "learning_rate": 3.6032516391280196e-06,
+      "loss": 0.3839,
+      "step": 2493
+    },
+    {
+      "epoch": 1.1610800744878957,
+      "grad_norm": 0.6044707842361149,
+      "learning_rate": 3.602100073586188e-06,
+      "loss": 0.3902,
+      "step": 2494
+    },
+    {
+      "epoch": 1.1615456238361266,
+      "grad_norm": 0.5997908091197678,
+      "learning_rate": 3.6009482177249886e-06,
+      "loss": 0.3754,
+      "step": 2495
+    },
+    {
+      "epoch": 1.1620111731843576,
+      "grad_norm": 0.6580703135714482,
+      "learning_rate": 3.599796071847849e-06,
+      "loss": 0.394,
+      "step": 2496
+    },
+    {
+      "epoch": 1.1624767225325885,
+      "grad_norm": 0.6714227314651797,
+      "learning_rate": 3.598643636258271e-06,
+      "loss": 0.3788,
+      "step": 2497
+    },
+    {
+      "epoch": 1.1629422718808193,
+      "grad_norm": 0.5930803910500599,
+      "learning_rate": 3.5974909112598337e-06,
+      "loss": 0.3846,
+      "step": 2498
+    },
+    {
+      "epoch": 1.1634078212290504,
+      "grad_norm": 0.6165805744967451,
+      "learning_rate": 3.5963378971561924e-06,
+      "loss": 0.3897,
+      "step": 2499
+    },
+    {
+      "epoch": 1.1638733705772812,
+      "grad_norm": 0.6600361680611864,
+      "learning_rate": 3.595184594251079e-06,
+      "loss": 0.4058,
+      "step": 2500
+    },
+    {
+      "epoch": 1.164338919925512,
+      "grad_norm": 0.6511610709033261,
+      "learning_rate": 3.5940310028482995e-06,
+      "loss": 0.404,
+      "step": 2501
+    },
+    {
+      "epoch": 1.164804469273743,
+      "grad_norm": 0.6004514146187995,
+      "learning_rate": 3.592877123251739e-06,
+      "loss": 0.3736,
+      "step": 2502
+    },
+    {
+      "epoch": 1.165270018621974,
+      "grad_norm": 0.6179644544977088,
+      "learning_rate": 3.5917229557653555e-06,
+      "loss": 0.3849,
+      "step": 2503
+    },
+    {
+      "epoch": 1.1657355679702048,
+      "grad_norm": 0.6778712461232503,
+      "learning_rate": 3.5905685006931847e-06,
+      "loss": 0.4096,
+      "step": 2504
+    },
+    {
+      "epoch": 1.1662011173184357,
+      "grad_norm": 0.6670092275148858,
+      "learning_rate": 3.5894137583393375e-06,
+      "loss": 0.3974,
+      "step": 2505
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 0.6479638325167982,
+      "learning_rate": 3.588258729008e-06,
+      "loss": 0.39,
+      "step": 2506
+    },
+    {
+      "epoch": 1.1671322160148976,
+      "grad_norm": 0.6039757402906047,
+      "learning_rate": 3.5871034130034342e-06,
+      "loss": 0.3856,
+      "step": 2507
+    },
+    {
+      "epoch": 1.1675977653631284,
+      "grad_norm": 0.6283503761426772,
+      "learning_rate": 3.58594781062998e-06,
+      "loss": 0.3836,
+      "step": 2508
+    },
+    {
+      "epoch": 1.1680633147113595,
+      "grad_norm": 0.6122736548248648,
+      "learning_rate": 3.584791922192048e-06,
+      "loss": 0.3951,
+      "step": 2509
+    },
+    {
+      "epoch": 1.1685288640595903,
+      "grad_norm": 0.6094468684053473,
+      "learning_rate": 3.5836357479941287e-06,
+      "loss": 0.3765,
+      "step": 2510
+    },
+    {
+      "epoch": 1.1689944134078212,
+      "grad_norm": 0.5783977370266052,
+      "learning_rate": 3.5824792883407832e-06,
+      "loss": 0.4103,
+      "step": 2511
+    },
+    {
+      "epoch": 1.169459962756052,
+      "grad_norm": 0.614614346309832,
+      "learning_rate": 3.581322543536654e-06,
+      "loss": 0.3892,
+      "step": 2512
+    },
+    {
+      "epoch": 1.169925512104283,
+      "grad_norm": 0.66746365589234,
+      "learning_rate": 3.5801655138864524e-06,
+      "loss": 0.3946,
+      "step": 2513
+    },
+    {
+      "epoch": 1.170391061452514,
+      "grad_norm": 0.636405116371045,
+      "learning_rate": 3.579008199694968e-06,
+      "loss": 0.3927,
+      "step": 2514
+    },
+    {
+      "epoch": 1.1708566108007448,
+      "grad_norm": 0.6306896349292754,
+      "learning_rate": 3.5778506012670656e-06,
+      "loss": 0.3768,
+      "step": 2515
+    },
+    {
+      "epoch": 1.1713221601489758,
+      "grad_norm": 0.6056243505002231,
+      "learning_rate": 3.5766927189076844e-06,
+      "loss": 0.3906,
+      "step": 2516
+    },
+    {
+      "epoch": 1.1717877094972067,
+      "grad_norm": 0.6065697136474241,
+      "learning_rate": 3.5755345529218364e-06,
+      "loss": 0.3716,
+      "step": 2517
+    },
+    {
+      "epoch": 1.1722532588454375,
+      "grad_norm": 0.6043913639458873,
+      "learning_rate": 3.574376103614613e-06,
+      "loss": 0.383,
+      "step": 2518
+    },
+    {
+      "epoch": 1.1727188081936686,
+      "grad_norm": 0.5865091141544224,
+      "learning_rate": 3.5732173712911743e-06,
+      "loss": 0.3817,
+      "step": 2519
+    },
+    {
+      "epoch": 1.1731843575418994,
+      "grad_norm": 0.66993516213978,
+      "learning_rate": 3.57205835625676e-06,
+      "loss": 0.407,
+      "step": 2520
+    },
+    {
+      "epoch": 1.1736499068901303,
+      "grad_norm": 0.616635013167286,
+      "learning_rate": 3.570899058816681e-06,
+      "loss": 0.3973,
+      "step": 2521
+    },
+    {
+      "epoch": 1.1741154562383613,
+      "grad_norm": 0.6069136518924783,
+      "learning_rate": 3.5697394792763247e-06,
+      "loss": 0.3847,
+      "step": 2522
+    },
+    {
+      "epoch": 1.1745810055865922,
+      "grad_norm": 0.6568469079631032,
+      "learning_rate": 3.5685796179411507e-06,
+      "loss": 0.4134,
+      "step": 2523
+    },
+    {
+      "epoch": 1.175046554934823,
+      "grad_norm": 0.6117560652948956,
+      "learning_rate": 3.5674194751166953e-06,
+      "loss": 0.3832,
+      "step": 2524
+    },
+    {
+      "epoch": 1.175512104283054,
+      "grad_norm": 0.6191556489898604,
+      "learning_rate": 3.566259051108567e-06,
+      "loss": 0.4049,
+      "step": 2525
+    },
+    {
+      "epoch": 1.175977653631285,
+      "grad_norm": 0.6338360402707311,
+      "learning_rate": 3.5650983462224496e-06,
+      "loss": 0.3939,
+      "step": 2526
+    },
+    {
+      "epoch": 1.1764432029795158,
+      "grad_norm": 0.7141184140501322,
+      "learning_rate": 3.5639373607641003e-06,
+      "loss": 0.3912,
+      "step": 2527
+    },
+    {
+      "epoch": 1.1769087523277468,
+      "grad_norm": 0.6023110503316683,
+      "learning_rate": 3.5627760950393496e-06,
+      "loss": 0.4141,
+      "step": 2528
+    },
+    {
+      "epoch": 1.1773743016759777,
+      "grad_norm": 0.6414877514087256,
+      "learning_rate": 3.561614549354103e-06,
+      "loss": 0.3509,
+      "step": 2529
+    },
+    {
+      "epoch": 1.1778398510242085,
+      "grad_norm": 0.5990087620151914,
+      "learning_rate": 3.5604527240143395e-06,
+      "loss": 0.3982,
+      "step": 2530
+    },
+    {
+      "epoch": 1.1783054003724396,
+      "grad_norm": 0.597737048515768,
+      "learning_rate": 3.559290619326112e-06,
+      "loss": 0.391,
+      "step": 2531
+    },
+    {
+      "epoch": 1.1787709497206704,
+      "grad_norm": 0.6045697911323872,
+      "learning_rate": 3.558128235595545e-06,
+      "loss": 0.392,
+      "step": 2532
+    },
+    {
+      "epoch": 1.1792364990689013,
+      "grad_norm": 0.5965908169787647,
+      "learning_rate": 3.5569655731288383e-06,
+      "loss": 0.3856,
+      "step": 2533
+    },
+    {
+      "epoch": 1.1797020484171323,
+      "grad_norm": 0.5880135613930128,
+      "learning_rate": 3.555802632232266e-06,
+      "loss": 0.3916,
+      "step": 2534
+    },
+    {
+      "epoch": 1.1801675977653632,
+      "grad_norm": 0.5966113480522977,
+      "learning_rate": 3.5546394132121737e-06,
+      "loss": 0.3862,
+      "step": 2535
+    },
+    {
+      "epoch": 1.180633147113594,
+      "grad_norm": 0.6075693853015715,
+      "learning_rate": 3.5534759163749805e-06,
+      "loss": 0.3816,
+      "step": 2536
+    },
+    {
+      "epoch": 1.1810986964618249,
+      "grad_norm": 0.6047161937563307,
+      "learning_rate": 3.55231214202718e-06,
+      "loss": 0.3794,
+      "step": 2537
+    },
+    {
+      "epoch": 1.181564245810056,
+      "grad_norm": 0.6130475955462242,
+      "learning_rate": 3.5511480904753375e-06,
+      "loss": 0.3982,
+      "step": 2538
+    },
+    {
+      "epoch": 1.1820297951582868,
+      "grad_norm": 0.5931335311462086,
+      "learning_rate": 3.5499837620260925e-06,
+      "loss": 0.39,
+      "step": 2539
+    },
+    {
+      "epoch": 1.1824953445065176,
+      "grad_norm": 0.6809870450163736,
+      "learning_rate": 3.5488191569861554e-06,
+      "loss": 0.3722,
+      "step": 2540
+    },
+    {
+      "epoch": 1.1829608938547487,
+      "grad_norm": 0.6097583425046854,
+      "learning_rate": 3.547654275662312e-06,
+      "loss": 0.3845,
+      "step": 2541
+    },
+    {
+      "epoch": 1.1834264432029795,
+      "grad_norm": 0.5672777574295526,
+      "learning_rate": 3.5464891183614196e-06,
+      "loss": 0.3802,
+      "step": 2542
+    },
+    {
+      "epoch": 1.1838919925512104,
+      "grad_norm": 0.5957664093092113,
+      "learning_rate": 3.545323685390408e-06,
+      "loss": 0.3899,
+      "step": 2543
+    },
+    {
+      "epoch": 1.1843575418994414,
+      "grad_norm": 0.5866742323119019,
+      "learning_rate": 3.5441579770562813e-06,
+      "loss": 0.3765,
+      "step": 2544
+    },
+    {
+      "epoch": 1.1848230912476723,
+      "grad_norm": 0.6147479647744674,
+      "learning_rate": 3.542991993666113e-06,
+      "loss": 0.3919,
+      "step": 2545
+    },
+    {
+      "epoch": 1.185288640595903,
+      "grad_norm": 0.5941195400582184,
+      "learning_rate": 3.541825735527051e-06,
+      "loss": 0.3862,
+      "step": 2546
+    },
+    {
+      "epoch": 1.1857541899441342,
+      "grad_norm": 0.6101830101959842,
+      "learning_rate": 3.540659202946316e-06,
+      "loss": 0.3818,
+      "step": 2547
+    },
+    {
+      "epoch": 1.186219739292365,
+      "grad_norm": 0.6120234652664106,
+      "learning_rate": 3.539492396231201e-06,
+      "loss": 0.3804,
+      "step": 2548
+    },
+    {
+      "epoch": 1.1866852886405959,
+      "grad_norm": 0.6152546236190237,
+      "learning_rate": 3.5383253156890703e-06,
+      "loss": 0.3775,
+      "step": 2549
+    },
+    {
+      "epoch": 1.1871508379888267,
+      "grad_norm": 0.617167529598945,
+      "learning_rate": 3.5371579616273593e-06,
+      "loss": 0.3901,
+      "step": 2550
+    },
+    {
+      "epoch": 1.1876163873370578,
+      "grad_norm": 0.5911722731183269,
+      "learning_rate": 3.5359903343535795e-06,
+      "loss": 0.3972,
+      "step": 2551
+    },
+    {
+      "epoch": 1.1880819366852886,
+      "grad_norm": 0.6346036321106977,
+      "learning_rate": 3.5348224341753086e-06,
+      "loss": 0.3849,
+      "step": 2552
+    },
+    {
+      "epoch": 1.1885474860335195,
+      "grad_norm": 0.5724381210861079,
+      "learning_rate": 3.533654261400201e-06,
+      "loss": 0.3875,
+      "step": 2553
+    },
+    {
+      "epoch": 1.1890130353817505,
+      "grad_norm": 0.6756446514301236,
+      "learning_rate": 3.5324858163359814e-06,
+      "loss": 0.4116,
+      "step": 2554
+    },
+    {
+      "epoch": 1.1894785847299814,
+      "grad_norm": 0.5895936968138906,
+      "learning_rate": 3.531317099290446e-06,
+      "loss": 0.398,
+      "step": 2555
+    },
+    {
+      "epoch": 1.1899441340782122,
+      "grad_norm": 0.6441090752443792,
+      "learning_rate": 3.5301481105714615e-06,
+      "loss": 0.3891,
+      "step": 2556
+    },
+    {
+      "epoch": 1.1904096834264433,
+      "grad_norm": 0.6524707793713223,
+      "learning_rate": 3.5289788504869683e-06,
+      "loss": 0.3774,
+      "step": 2557
+    },
+    {
+      "epoch": 1.190875232774674,
+      "grad_norm": 0.6386368997442124,
+      "learning_rate": 3.527809319344977e-06,
+      "loss": 0.3783,
+      "step": 2558
+    },
+    {
+      "epoch": 1.191340782122905,
+      "grad_norm": 0.6468513054104826,
+      "learning_rate": 3.52663951745357e-06,
+      "loss": 0.3852,
+      "step": 2559
+    },
+    {
+      "epoch": 1.191806331471136,
+      "grad_norm": 0.6598101802642365,
+      "learning_rate": 3.525469445120901e-06,
+      "loss": 0.4062,
+      "step": 2560
+    },
+    {
+      "epoch": 1.1922718808193669,
+      "grad_norm": 0.6380889881090474,
+      "learning_rate": 3.5242991026551954e-06,
+      "loss": 0.3913,
+      "step": 2561
+    },
+    {
+      "epoch": 1.1927374301675977,
+      "grad_norm": 0.6319608670778535,
+      "learning_rate": 3.5231284903647476e-06,
+      "loss": 0.3687,
+      "step": 2562
+    },
+    {
+      "epoch": 1.1932029795158288,
+      "grad_norm": 0.6271912214677057,
+      "learning_rate": 3.5219576085579265e-06,
+      "loss": 0.3911,
+      "step": 2563
+    },
+    {
+      "epoch": 1.1936685288640596,
+      "grad_norm": 0.6784432670646744,
+      "learning_rate": 3.5207864575431693e-06,
+      "loss": 0.4019,
+      "step": 2564
+    },
+    {
+      "epoch": 1.1941340782122905,
+      "grad_norm": 0.6257573665931558,
+      "learning_rate": 3.519615037628985e-06,
+      "loss": 0.3815,
+      "step": 2565
+    },
+    {
+      "epoch": 1.1945996275605215,
+      "grad_norm": 0.615251537428129,
+      "learning_rate": 3.518443349123953e-06,
+      "loss": 0.388,
+      "step": 2566
+    },
+    {
+      "epoch": 1.1950651769087524,
+      "grad_norm": 0.6848097307556417,
+      "learning_rate": 3.517271392336726e-06,
+      "loss": 0.3708,
+      "step": 2567
+    },
+    {
+      "epoch": 1.1955307262569832,
+      "grad_norm": 0.6079259556073903,
+      "learning_rate": 3.516099167576023e-06,
+      "loss": 0.3874,
+      "step": 2568
+    },
+    {
+      "epoch": 1.1959962756052143,
+      "grad_norm": 0.58242202872711,
+      "learning_rate": 3.5149266751506367e-06,
+      "loss": 0.3877,
+      "step": 2569
+    },
+    {
+      "epoch": 1.196461824953445,
+      "grad_norm": 0.600288499679151,
+      "learning_rate": 3.5137539153694304e-06,
+      "loss": 0.3825,
+      "step": 2570
+    },
+    {
+      "epoch": 1.196927374301676,
+      "grad_norm": 0.6328890003585049,
+      "learning_rate": 3.5125808885413355e-06,
+      "loss": 0.3765,
+      "step": 2571
+    },
+    {
+      "epoch": 1.197392923649907,
+      "grad_norm": 0.6542817685575688,
+      "learning_rate": 3.5114075949753557e-06,
+      "loss": 0.3881,
+      "step": 2572
+    },
+    {
+      "epoch": 1.1978584729981379,
+      "grad_norm": 0.5865330584831816,
+      "learning_rate": 3.510234034980564e-06,
+      "loss": 0.376,
+      "step": 2573
+    },
+    {
+      "epoch": 1.1983240223463687,
+      "grad_norm": 0.587805433460954,
+      "learning_rate": 3.5090602088661052e-06,
+      "loss": 0.3861,
+      "step": 2574
+    },
+    {
+      "epoch": 1.1987895716945995,
+      "grad_norm": 0.6080948514835197,
+      "learning_rate": 3.5078861169411916e-06,
+      "loss": 0.3873,
+      "step": 2575
+    },
+    {
+      "epoch": 1.1992551210428306,
+      "grad_norm": 0.6051533709394068,
+      "learning_rate": 3.506711759515108e-06,
+      "loss": 0.3989,
+      "step": 2576
+    },
+    {
+      "epoch": 1.1997206703910615,
+      "grad_norm": 0.5770307609166867,
+      "learning_rate": 3.5055371368972067e-06,
+      "loss": 0.3823,
+      "step": 2577
+    },
+    {
+      "epoch": 1.2001862197392923,
+      "grad_norm": 0.6316287272919617,
+      "learning_rate": 3.5043622493969137e-06,
+      "loss": 0.3724,
+      "step": 2578
+    },
+    {
+      "epoch": 1.2006517690875234,
+      "grad_norm": 0.6361034228338776,
+      "learning_rate": 3.5031870973237193e-06,
+      "loss": 0.3933,
+      "step": 2579
+    },
+    {
+      "epoch": 1.2011173184357542,
+      "grad_norm": 0.6284276735564999,
+      "learning_rate": 3.5020116809871895e-06,
+      "loss": 0.3593,
+      "step": 2580
+    },
+    {
+      "epoch": 1.201582867783985,
+      "grad_norm": 0.5950379785010348,
+      "learning_rate": 3.5008360006969545e-06,
+      "loss": 0.3743,
+      "step": 2581
+    },
+    {
+      "epoch": 1.202048417132216,
+      "grad_norm": 0.6079517512627873,
+      "learning_rate": 3.499660056762717e-06,
+      "loss": 0.3808,
+      "step": 2582
+    },
+    {
+      "epoch": 1.202513966480447,
+      "grad_norm": 0.5933919684876672,
+      "learning_rate": 3.49848384949425e-06,
+      "loss": 0.3925,
+      "step": 2583
+    },
+    {
+      "epoch": 1.2029795158286778,
+      "grad_norm": 0.5923417701930579,
+      "learning_rate": 3.497307379201393e-06,
+      "loss": 0.3926,
+      "step": 2584
+    },
+    {
+      "epoch": 1.2034450651769086,
+      "grad_norm": 0.6473903343633247,
+      "learning_rate": 3.496130646194056e-06,
+      "loss": 0.404,
+      "step": 2585
+    },
+    {
+      "epoch": 1.2039106145251397,
+      "grad_norm": 0.5848926609438242,
+      "learning_rate": 3.4949536507822195e-06,
+      "loss": 0.3929,
+      "step": 2586
+    },
+    {
+      "epoch": 1.2043761638733705,
+      "grad_norm": 0.6274440533964614,
+      "learning_rate": 3.493776393275932e-06,
+      "loss": 0.3936,
+      "step": 2587
+    },
+    {
+      "epoch": 1.2048417132216014,
+      "grad_norm": 0.6263677418334606,
+      "learning_rate": 3.4925988739853105e-06,
+      "loss": 0.3921,
+      "step": 2588
+    },
+    {
+      "epoch": 1.2053072625698324,
+      "grad_norm": 0.6233624114599329,
+      "learning_rate": 3.491421093220542e-06,
+      "loss": 0.3915,
+      "step": 2589
+    },
+    {
+      "epoch": 1.2057728119180633,
+      "grad_norm": 0.6414818906207919,
+      "learning_rate": 3.490243051291882e-06,
+      "loss": 0.3696,
+      "step": 2590
+    },
+    {
+      "epoch": 1.2062383612662941,
+      "grad_norm": 0.5876726079728635,
+      "learning_rate": 3.4890647485096544e-06,
+      "loss": 0.3802,
+      "step": 2591
+    },
+    {
+      "epoch": 1.2067039106145252,
+      "grad_norm": 0.6012170577239156,
+      "learning_rate": 3.487886185184253e-06,
+      "loss": 0.3883,
+      "step": 2592
+    },
+    {
+      "epoch": 1.207169459962756,
+      "grad_norm": 0.6185003649948034,
+      "learning_rate": 3.486707361626138e-06,
+      "loss": 0.3811,
+      "step": 2593
+    },
+    {
+      "epoch": 1.2076350093109869,
+      "grad_norm": 0.6289660278496021,
+      "learning_rate": 3.485528278145841e-06,
+      "loss": 0.3971,
+      "step": 2594
+    },
+    {
+      "epoch": 1.208100558659218,
+      "grad_norm": 0.5537656153161206,
+      "learning_rate": 3.4843489350539594e-06,
+      "loss": 0.3668,
+      "step": 2595
+    },
+    {
+      "epoch": 1.2085661080074488,
+      "grad_norm": 0.6207185711573387,
+      "learning_rate": 3.483169332661161e-06,
+      "loss": 0.3787,
+      "step": 2596
+    },
+    {
+      "epoch": 1.2090316573556796,
+      "grad_norm": 0.602436998525542,
+      "learning_rate": 3.481989471278181e-06,
+      "loss": 0.3784,
+      "step": 2597
+    },
+    {
+      "epoch": 1.2094972067039107,
+      "grad_norm": 0.597759046964699,
+      "learning_rate": 3.4808093512158226e-06,
+      "loss": 0.3924,
+      "step": 2598
+    },
+    {
+      "epoch": 1.2099627560521415,
+      "grad_norm": 0.6048097133007926,
+      "learning_rate": 3.4796289727849575e-06,
+      "loss": 0.4048,
+      "step": 2599
+    },
+    {
+      "epoch": 1.2104283054003724,
+      "grad_norm": 0.6111645983119767,
+      "learning_rate": 3.478448336296527e-06,
+      "loss": 0.3862,
+      "step": 2600
+    },
+    {
+      "epoch": 1.2108938547486034,
+      "grad_norm": 0.6032448692031139,
+      "learning_rate": 3.477267442061536e-06,
+      "loss": 0.3973,
+      "step": 2601
+    },
+    {
+      "epoch": 1.2113594040968343,
+      "grad_norm": 0.6249607131233852,
+      "learning_rate": 3.476086290391063e-06,
+      "loss": 0.3991,
+      "step": 2602
+    },
+    {
+      "epoch": 1.2118249534450651,
+      "grad_norm": 0.5980424157914778,
+      "learning_rate": 3.4749048815962496e-06,
+      "loss": 0.3903,
+      "step": 2603
+    },
+    {
+      "epoch": 1.2122905027932962,
+      "grad_norm": 0.6176332200730631,
+      "learning_rate": 3.4737232159883074e-06,
+      "loss": 0.39,
+      "step": 2604
+    },
+    {
+      "epoch": 1.212756052141527,
+      "grad_norm": 0.6047090887236449,
+      "learning_rate": 3.472541293878514e-06,
+      "loss": 0.3949,
+      "step": 2605
+    },
+    {
+      "epoch": 1.2132216014897579,
+      "grad_norm": 0.5812193332210615,
+      "learning_rate": 3.471359115578219e-06,
+      "loss": 0.3755,
+      "step": 2606
+    },
+    {
+      "epoch": 1.213687150837989,
+      "grad_norm": 0.5763641591994688,
+      "learning_rate": 3.4701766813988338e-06,
+      "loss": 0.3816,
+      "step": 2607
+    },
+    {
+      "epoch": 1.2141527001862198,
+      "grad_norm": 0.6315463915344897,
+      "learning_rate": 3.4689939916518394e-06,
+      "loss": 0.3792,
+      "step": 2608
+    },
+    {
+      "epoch": 1.2146182495344506,
+      "grad_norm": 0.6181793111021477,
+      "learning_rate": 3.467811046648786e-06,
+      "loss": 0.4124,
+      "step": 2609
+    },
+    {
+      "epoch": 1.2150837988826815,
+      "grad_norm": 0.6381106834893612,
+      "learning_rate": 3.4666278467012886e-06,
+      "loss": 0.3755,
+      "step": 2610
+    },
+    {
+      "epoch": 1.2155493482309125,
+      "grad_norm": 0.6176561785560859,
+      "learning_rate": 3.4654443921210296e-06,
+      "loss": 0.3929,
+      "step": 2611
+    },
+    {
+      "epoch": 1.2160148975791434,
+      "grad_norm": 0.6116219419055932,
+      "learning_rate": 3.464260683219761e-06,
+      "loss": 0.3944,
+      "step": 2612
+    },
+    {
+      "epoch": 1.2164804469273742,
+      "grad_norm": 0.6103980379029469,
+      "learning_rate": 3.4630767203092972e-06,
+      "loss": 0.4025,
+      "step": 2613
+    },
+    {
+      "epoch": 1.2169459962756053,
+      "grad_norm": 0.652449750962784,
+      "learning_rate": 3.461892503701524e-06,
+      "loss": 0.3816,
+      "step": 2614
+    },
+    {
+      "epoch": 1.2174115456238361,
+      "grad_norm": 0.5919095894222228,
+      "learning_rate": 3.4607080337083927e-06,
+      "loss": 0.369,
+      "step": 2615
+    },
+    {
+      "epoch": 1.217877094972067,
+      "grad_norm": 0.6384152062485043,
+      "learning_rate": 3.459523310641919e-06,
+      "loss": 0.3989,
+      "step": 2616
+    },
+    {
+      "epoch": 1.218342644320298,
+      "grad_norm": 0.6068066653688865,
+      "learning_rate": 3.4583383348141887e-06,
+      "loss": 0.3784,
+      "step": 2617
+    },
+    {
+      "epoch": 1.2188081936685289,
+      "grad_norm": 0.5852343485812597,
+      "learning_rate": 3.4571531065373514e-06,
+      "loss": 0.3914,
+      "step": 2618
+    },
+    {
+      "epoch": 1.2192737430167597,
+      "grad_norm": 0.6135261106449019,
+      "learning_rate": 3.4559676261236263e-06,
+      "loss": 0.3752,
+      "step": 2619
+    },
+    {
+      "epoch": 1.2197392923649906,
+      "grad_norm": 0.7062566095631803,
+      "learning_rate": 3.4547818938852945e-06,
+      "loss": 0.3805,
+      "step": 2620
+    },
+    {
+      "epoch": 1.2202048417132216,
+      "grad_norm": 0.6186864837613978,
+      "learning_rate": 3.4535959101347084e-06,
+      "loss": 0.3902,
+      "step": 2621
+    },
+    {
+      "epoch": 1.2206703910614525,
+      "grad_norm": 0.6097330458593744,
+      "learning_rate": 3.452409675184283e-06,
+      "loss": 0.3657,
+      "step": 2622
+    },
+    {
+      "epoch": 1.2211359404096833,
+      "grad_norm": 0.6112995571181632,
+      "learning_rate": 3.451223189346501e-06,
+      "loss": 0.3864,
+      "step": 2623
+    },
+    {
+      "epoch": 1.2216014897579144,
+      "grad_norm": 0.6086252353039308,
+      "learning_rate": 3.4500364529339107e-06,
+      "loss": 0.3873,
+      "step": 2624
+    },
+    {
+      "epoch": 1.2220670391061452,
+      "grad_norm": 0.6080712649490733,
+      "learning_rate": 3.4488494662591277e-06,
+      "loss": 0.3927,
+      "step": 2625
+    },
+    {
+      "epoch": 1.222532588454376,
+      "grad_norm": 0.6039338127665118,
+      "learning_rate": 3.4476622296348323e-06,
+      "loss": 0.3919,
+      "step": 2626
+    },
+    {
+      "epoch": 1.2229981378026071,
+      "grad_norm": 0.6044341756408255,
+      "learning_rate": 3.44647474337377e-06,
+      "loss": 0.3849,
+      "step": 2627
+    },
+    {
+      "epoch": 1.223463687150838,
+      "grad_norm": 0.5853993703842448,
+      "learning_rate": 3.4452870077887535e-06,
+      "loss": 0.3783,
+      "step": 2628
+    },
+    {
+      "epoch": 1.2239292364990688,
+      "grad_norm": 0.5992329031679262,
+      "learning_rate": 3.4440990231926608e-06,
+      "loss": 0.3953,
+      "step": 2629
+    },
+    {
+      "epoch": 1.2243947858472999,
+      "grad_norm": 0.6022648870764112,
+      "learning_rate": 3.4429107898984345e-06,
+      "loss": 0.3619,
+      "step": 2630
+    },
+    {
+      "epoch": 1.2248603351955307,
+      "grad_norm": 0.648607700013412,
+      "learning_rate": 3.4417223082190843e-06,
+      "loss": 0.4053,
+      "step": 2631
+    },
+    {
+      "epoch": 1.2253258845437616,
+      "grad_norm": 0.649315734398025,
+      "learning_rate": 3.440533578467684e-06,
+      "loss": 0.393,
+      "step": 2632
+    },
+    {
+      "epoch": 1.2257914338919926,
+      "grad_norm": 0.6114805139922289,
+      "learning_rate": 3.439344600957374e-06,
+      "loss": 0.3853,
+      "step": 2633
+    },
+    {
+      "epoch": 1.2262569832402235,
+      "grad_norm": 0.6042087858820708,
+      "learning_rate": 3.4381553760013585e-06,
+      "loss": 0.3999,
+      "step": 2634
+    },
+    {
+      "epoch": 1.2267225325884543,
+      "grad_norm": 0.6135488117010734,
+      "learning_rate": 3.436965903912908e-06,
+      "loss": 0.4006,
+      "step": 2635
+    },
+    {
+      "epoch": 1.2271880819366854,
+      "grad_norm": 0.5564806651472145,
+      "learning_rate": 3.435776185005357e-06,
+      "loss": 0.4036,
+      "step": 2636
+    },
+    {
+      "epoch": 1.2276536312849162,
+      "grad_norm": 0.619902382039694,
+      "learning_rate": 3.4345862195921076e-06,
+      "loss": 0.3832,
+      "step": 2637
+    },
+    {
+      "epoch": 1.228119180633147,
+      "grad_norm": 0.611953668565239,
+      "learning_rate": 3.433396007986623e-06,
+      "loss": 0.3932,
+      "step": 2638
+    },
+    {
+      "epoch": 1.2285847299813781,
+      "grad_norm": 0.5725492036085453,
+      "learning_rate": 3.432205550502435e-06,
+      "loss": 0.3757,
+      "step": 2639
+    },
+    {
+      "epoch": 1.229050279329609,
+      "grad_norm": 0.5948885897316669,
+      "learning_rate": 3.4310148474531355e-06,
+      "loss": 0.3748,
+      "step": 2640
+    },
+    {
+      "epoch": 1.2295158286778398,
+      "grad_norm": 0.6204212158398673,
+      "learning_rate": 3.429823899152388e-06,
+      "loss": 0.3908,
+      "step": 2641
+    },
+    {
+      "epoch": 1.2299813780260709,
+      "grad_norm": 0.5579269149571481,
+      "learning_rate": 3.4286327059139134e-06,
+      "loss": 0.3981,
+      "step": 2642
+    },
+    {
+      "epoch": 1.2304469273743017,
+      "grad_norm": 0.5952254880610155,
+      "learning_rate": 3.427441268051502e-06,
+      "loss": 0.3795,
+      "step": 2643
+    },
+    {
+      "epoch": 1.2309124767225326,
+      "grad_norm": 0.6143529519906572,
+      "learning_rate": 3.4262495858790056e-06,
+      "loss": 0.3909,
+      "step": 2644
+    },
+    {
+      "epoch": 1.2313780260707634,
+      "grad_norm": 0.5925851764308231,
+      "learning_rate": 3.425057659710343e-06,
+      "loss": 0.3701,
+      "step": 2645
+    },
+    {
+      "epoch": 1.2318435754189945,
+      "grad_norm": 0.6319619318576636,
+      "learning_rate": 3.423865489859495e-06,
+      "loss": 0.3909,
+      "step": 2646
+    },
+    {
+      "epoch": 1.2323091247672253,
+      "grad_norm": 0.6136293964976675,
+      "learning_rate": 3.422673076640508e-06,
+      "loss": 0.3964,
+      "step": 2647
+    },
+    {
+      "epoch": 1.2327746741154562,
+      "grad_norm": 0.5909915671277984,
+      "learning_rate": 3.421480420367492e-06,
+      "loss": 0.3847,
+      "step": 2648
+    },
+    {
+      "epoch": 1.2332402234636872,
+      "grad_norm": 0.6061880633698677,
+      "learning_rate": 3.42028752135462e-06,
+      "loss": 0.3917,
+      "step": 2649
+    },
+    {
+      "epoch": 1.233705772811918,
+      "grad_norm": 0.6581751383732604,
+      "learning_rate": 3.41909437991613e-06,
+      "loss": 0.4036,
+      "step": 2650
+    },
+    {
+      "epoch": 1.234171322160149,
+      "grad_norm": 0.6224731580162748,
+      "learning_rate": 3.4179009963663262e-06,
+      "loss": 0.3828,
+      "step": 2651
+    },
+    {
+      "epoch": 1.23463687150838,
+      "grad_norm": 0.6227354796959975,
+      "learning_rate": 3.416707371019572e-06,
+      "loss": 0.3993,
+      "step": 2652
+    },
+    {
+      "epoch": 1.2351024208566108,
+      "grad_norm": 0.5752332063790211,
+      "learning_rate": 3.4155135041902974e-06,
+      "loss": 0.3886,
+      "step": 2653
+    },
+    {
+      "epoch": 1.2355679702048417,
+      "grad_norm": 0.6055086783442865,
+      "learning_rate": 3.4143193961929945e-06,
+      "loss": 0.3843,
+      "step": 2654
+    },
+    {
+      "epoch": 1.2360335195530725,
+      "grad_norm": 0.5951214313970281,
+      "learning_rate": 3.4131250473422215e-06,
+      "loss": 0.3915,
+      "step": 2655
+    },
+    {
+      "epoch": 1.2364990689013036,
+      "grad_norm": 0.5661599170290511,
+      "learning_rate": 3.411930457952596e-06,
+      "loss": 0.3812,
+      "step": 2656
+    },
+    {
+      "epoch": 1.2369646182495344,
+      "grad_norm": 0.5966052370230033,
+      "learning_rate": 3.4107356283388025e-06,
+      "loss": 0.3895,
+      "step": 2657
+    },
+    {
+      "epoch": 1.2374301675977653,
+      "grad_norm": 0.6203785438751089,
+      "learning_rate": 3.409540558815589e-06,
+      "loss": 0.4067,
+      "step": 2658
+    },
+    {
+      "epoch": 1.2378957169459963,
+      "grad_norm": 0.5918373180718479,
+      "learning_rate": 3.408345249697762e-06,
+      "loss": 0.3538,
+      "step": 2659
+    },
+    {
+      "epoch": 1.2383612662942272,
+      "grad_norm": 0.6003690013563958,
+      "learning_rate": 3.407149701300197e-06,
+      "loss": 0.401,
+      "step": 2660
+    },
+    {
+      "epoch": 1.238826815642458,
+      "grad_norm": 0.6266347059205281,
+      "learning_rate": 3.405953913937829e-06,
+      "loss": 0.3924,
+      "step": 2661
+    },
+    {
+      "epoch": 1.239292364990689,
+      "grad_norm": 0.5964388708042708,
+      "learning_rate": 3.4047578879256565e-06,
+      "loss": 0.395,
+      "step": 2662
+    },
+    {
+      "epoch": 1.23975791433892,
+      "grad_norm": 0.6250290917333674,
+      "learning_rate": 3.403561623578741e-06,
+      "loss": 0.3994,
+      "step": 2663
+    },
+    {
+      "epoch": 1.2402234636871508,
+      "grad_norm": 0.6344418437446055,
+      "learning_rate": 3.402365121212208e-06,
+      "loss": 0.394,
+      "step": 2664
+    },
+    {
+      "epoch": 1.2406890130353818,
+      "grad_norm": 0.6333388590622999,
+      "learning_rate": 3.4011683811412446e-06,
+      "loss": 0.3962,
+      "step": 2665
+    },
+    {
+      "epoch": 1.2411545623836127,
+      "grad_norm": 0.6233503349699804,
+      "learning_rate": 3.3999714036810994e-06,
+      "loss": 0.3826,
+      "step": 2666
+    },
+    {
+      "epoch": 1.2416201117318435,
+      "grad_norm": 0.6481151686636853,
+      "learning_rate": 3.398774189147085e-06,
+      "loss": 0.4132,
+      "step": 2667
+    },
+    {
+      "epoch": 1.2420856610800746,
+      "grad_norm": 0.6801040634232306,
+      "learning_rate": 3.3975767378545776e-06,
+      "loss": 0.3736,
+      "step": 2668
+    },
+    {
+      "epoch": 1.2425512104283054,
+      "grad_norm": 0.5846627373935913,
+      "learning_rate": 3.3963790501190126e-06,
+      "loss": 0.4021,
+      "step": 2669
+    },
+    {
+      "epoch": 1.2430167597765363,
+      "grad_norm": 0.620634855501039,
+      "learning_rate": 3.395181126255891e-06,
+      "loss": 0.3956,
+      "step": 2670
+    },
+    {
+      "epoch": 1.2434823091247673,
+      "grad_norm": 0.6132324566150464,
+      "learning_rate": 3.3939829665807727e-06,
+      "loss": 0.3933,
+      "step": 2671
+    },
+    {
+      "epoch": 1.2439478584729982,
+      "grad_norm": 0.6261710837880852,
+      "learning_rate": 3.3927845714092833e-06,
+      "loss": 0.3853,
+      "step": 2672
+    },
+    {
+      "epoch": 1.244413407821229,
+      "grad_norm": 0.6060683158724338,
+      "learning_rate": 3.391585941057106e-06,
+      "loss": 0.3982,
+      "step": 2673
+    },
+    {
+      "epoch": 1.24487895716946,
+      "grad_norm": 0.6218591607384291,
+      "learning_rate": 3.3903870758399916e-06,
+      "loss": 0.3827,
+      "step": 2674
+    },
+    {
+      "epoch": 1.245344506517691,
+      "grad_norm": 0.5604863556323065,
+      "learning_rate": 3.3891879760737483e-06,
+      "loss": 0.3832,
+      "step": 2675
+    },
+    {
+      "epoch": 1.2458100558659218,
+      "grad_norm": 0.6290185594471721,
+      "learning_rate": 3.3879886420742476e-06,
+      "loss": 0.4094,
+      "step": 2676
+    },
+    {
+      "epoch": 1.2462756052141528,
+      "grad_norm": 0.5562190279075199,
+      "learning_rate": 3.386789074157422e-06,
+      "loss": 0.3835,
+      "step": 2677
+    },
+    {
+      "epoch": 1.2467411545623837,
+      "grad_norm": 0.6286524888999662,
+      "learning_rate": 3.385589272639267e-06,
+      "loss": 0.3885,
+      "step": 2678
+    },
+    {
+      "epoch": 1.2472067039106145,
+      "grad_norm": 0.5801384984602453,
+      "learning_rate": 3.3843892378358377e-06,
+      "loss": 0.3987,
+      "step": 2679
+    },
+    {
+      "epoch": 1.2476722532588453,
+      "grad_norm": 0.578056328595549,
+      "learning_rate": 3.3831889700632535e-06,
+      "loss": 0.3753,
+      "step": 2680
+    },
+    {
+      "epoch": 1.2481378026070764,
+      "grad_norm": 0.6397007774262833,
+      "learning_rate": 3.3819884696376925e-06,
+      "loss": 0.4108,
+      "step": 2681
+    },
+    {
+      "epoch": 1.2486033519553073,
+      "grad_norm": 0.6586791975645327,
+      "learning_rate": 3.380787736875395e-06,
+      "loss": 0.4009,
+      "step": 2682
+    },
+    {
+      "epoch": 1.249068901303538,
+      "grad_norm": 0.5994069101587544,
+      "learning_rate": 3.379586772092662e-06,
+      "loss": 0.3829,
+      "step": 2683
+    },
+    {
+      "epoch": 1.2495344506517692,
+      "grad_norm": 0.622031850716187,
+      "learning_rate": 3.378385575605858e-06,
+      "loss": 0.3961,
+      "step": 2684
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6002394282878988,
+      "learning_rate": 3.377184147731406e-06,
+      "loss": 0.3871,
+      "step": 2685
+    },
+    {
+      "epoch": 1.2504655493482308,
+      "grad_norm": 0.6431855718770174,
+      "learning_rate": 3.37598248878579e-06,
+      "loss": 0.3929,
+      "step": 2686
+    },
+    {
+      "epoch": 1.250931098696462,
+      "grad_norm": 0.6022561117638771,
+      "learning_rate": 3.3747805990855553e-06,
+      "loss": 0.3944,
+      "step": 2687
+    },
+    {
+      "epoch": 1.2513966480446927,
+      "grad_norm": 0.6008871953897591,
+      "learning_rate": 3.373578478947309e-06,
+      "loss": 0.3967,
+      "step": 2688
+    },
+    {
+      "epoch": 1.2518621973929236,
+      "grad_norm": 0.6217992133844792,
+      "learning_rate": 3.3723761286877182e-06,
+      "loss": 0.3674,
+      "step": 2689
+    },
+    {
+      "epoch": 1.2523277467411544,
+      "grad_norm": 0.5688634503523363,
+      "learning_rate": 3.3711735486235096e-06,
+      "loss": 0.3837,
+      "step": 2690
+    },
+    {
+      "epoch": 1.2527932960893855,
+      "grad_norm": 0.5927872631548742,
+      "learning_rate": 3.3699707390714732e-06,
+      "loss": 0.3884,
+      "step": 2691
+    },
+    {
+      "epoch": 1.2532588454376163,
+      "grad_norm": 0.563988867581215,
+      "learning_rate": 3.3687677003484558e-06,
+      "loss": 0.3759,
+      "step": 2692
+    },
+    {
+      "epoch": 1.2537243947858472,
+      "grad_norm": 0.624203193899171,
+      "learning_rate": 3.367564432771367e-06,
+      "loss": 0.3758,
+      "step": 2693
+    },
+    {
+      "epoch": 1.2541899441340782,
+      "grad_norm": 0.5975567231253438,
+      "learning_rate": 3.366360936657177e-06,
+      "loss": 0.3884,
+      "step": 2694
+    },
+    {
+      "epoch": 1.254655493482309,
+      "grad_norm": 0.5889888357782916,
+      "learning_rate": 3.3651572123229133e-06,
+      "loss": 0.3946,
+      "step": 2695
+    },
+    {
+      "epoch": 1.25512104283054,
+      "grad_norm": 0.5555553139180643,
+      "learning_rate": 3.3639532600856674e-06,
+      "loss": 0.3669,
+      "step": 2696
+    },
+    {
+      "epoch": 1.255586592178771,
+      "grad_norm": 0.5778746609017216,
+      "learning_rate": 3.3627490802625884e-06,
+      "loss": 0.3922,
+      "step": 2697
+    },
+    {
+      "epoch": 1.2560521415270018,
+      "grad_norm": 0.5854738631045114,
+      "learning_rate": 3.3615446731708856e-06,
+      "loss": 0.3784,
+      "step": 2698
+    },
+    {
+      "epoch": 1.2565176908752327,
+      "grad_norm": 0.5897340983434048,
+      "learning_rate": 3.3603400391278278e-06,
+      "loss": 0.3823,
+      "step": 2699
+    },
+    {
+      "epoch": 1.2569832402234637,
+      "grad_norm": 0.5986305554202064,
+      "learning_rate": 3.3591351784507457e-06,
+      "loss": 0.3812,
+      "step": 2700
+    },
+    {
+      "epoch": 1.2574487895716946,
+      "grad_norm": 0.609496966501631,
+      "learning_rate": 3.3579300914570277e-06,
+      "loss": 0.3929,
+      "step": 2701
+    },
+    {
+      "epoch": 1.2579143389199254,
+      "grad_norm": 0.5929823787300966,
+      "learning_rate": 3.3567247784641223e-06,
+      "loss": 0.3807,
+      "step": 2702
+    },
+    {
+      "epoch": 1.2583798882681565,
+      "grad_norm": 0.6374118945742637,
+      "learning_rate": 3.3555192397895375e-06,
+      "loss": 0.392,
+      "step": 2703
+    },
+    {
+      "epoch": 1.2588454376163873,
+      "grad_norm": 0.6217737082573123,
+      "learning_rate": 3.354313475750841e-06,
+      "loss": 0.3907,
+      "step": 2704
+    },
+    {
+      "epoch": 1.2593109869646182,
+      "grad_norm": 0.5832076652930075,
+      "learning_rate": 3.35310748666566e-06,
+      "loss": 0.4086,
+      "step": 2705
+    },
+    {
+      "epoch": 1.2597765363128492,
+      "grad_norm": 0.6082503562244301,
+      "learning_rate": 3.3519012728516798e-06,
+      "loss": 0.3857,
+      "step": 2706
+    },
+    {
+      "epoch": 1.26024208566108,
+      "grad_norm": 0.6466843614112937,
+      "learning_rate": 3.3506948346266473e-06,
+      "loss": 0.381,
+      "step": 2707
+    },
+    {
+      "epoch": 1.260707635009311,
+      "grad_norm": 0.61714198059755,
+      "learning_rate": 3.349488172308365e-06,
+      "loss": 0.3856,
+      "step": 2708
+    },
+    {
+      "epoch": 1.261173184357542,
+      "grad_norm": 0.5924492220283866,
+      "learning_rate": 3.3482812862146986e-06,
+      "loss": 0.3793,
+      "step": 2709
+    },
+    {
+      "epoch": 1.2616387337057728,
+      "grad_norm": 0.6150458358370479,
+      "learning_rate": 3.347074176663569e-06,
+      "loss": 0.3852,
+      "step": 2710
+    },
+    {
+      "epoch": 1.2621042830540037,
+      "grad_norm": 0.5837941002796377,
+      "learning_rate": 3.345866843972958e-06,
+      "loss": 0.3808,
+      "step": 2711
+    },
+    {
+      "epoch": 1.2625698324022347,
+      "grad_norm": 0.5859149040583357,
+      "learning_rate": 3.3446592884609054e-06,
+      "loss": 0.3799,
+      "step": 2712
+    },
+    {
+      "epoch": 1.2630353817504656,
+      "grad_norm": 0.5779563749969755,
+      "learning_rate": 3.3434515104455113e-06,
+      "loss": 0.3768,
+      "step": 2713
+    },
+    {
+      "epoch": 1.2635009310986964,
+      "grad_norm": 0.5832318032331311,
+      "learning_rate": 3.3422435102449323e-06,
+      "loss": 0.3892,
+      "step": 2714
+    },
+    {
+      "epoch": 1.2639664804469275,
+      "grad_norm": 0.5976919257124871,
+      "learning_rate": 3.3410352881773843e-06,
+      "loss": 0.4025,
+      "step": 2715
+    },
+    {
+      "epoch": 1.2644320297951583,
+      "grad_norm": 0.618237263387303,
+      "learning_rate": 3.3398268445611413e-06,
+      "loss": 0.3737,
+      "step": 2716
+    },
+    {
+      "epoch": 1.2648975791433892,
+      "grad_norm": 0.610745004813332,
+      "learning_rate": 3.338618179714538e-06,
+      "loss": 0.3758,
+      "step": 2717
+    },
+    {
+      "epoch": 1.2653631284916202,
+      "grad_norm": 0.6308374031769395,
+      "learning_rate": 3.337409293955962e-06,
+      "loss": 0.4128,
+      "step": 2718
+    },
+    {
+      "epoch": 1.265828677839851,
+      "grad_norm": 0.6130564275951952,
+      "learning_rate": 3.336200187603866e-06,
+      "loss": 0.3821,
+      "step": 2719
+    },
+    {
+      "epoch": 1.266294227188082,
+      "grad_norm": 0.6085850339131786,
+      "learning_rate": 3.334990860976756e-06,
+      "loss": 0.4093,
+      "step": 2720
+    },
+    {
+      "epoch": 1.2667597765363128,
+      "grad_norm": 0.6182014214747857,
+      "learning_rate": 3.3337813143931973e-06,
+      "loss": 0.3738,
+      "step": 2721
+    },
+    {
+      "epoch": 1.2672253258845438,
+      "grad_norm": 0.6180118264782883,
+      "learning_rate": 3.332571548171813e-06,
+      "loss": 0.3882,
+      "step": 2722
+    },
+    {
+      "epoch": 1.2676908752327747,
+      "grad_norm": 0.5670481278340116,
+      "learning_rate": 3.3313615626312844e-06,
+      "loss": 0.3932,
+      "step": 2723
+    },
+    {
+      "epoch": 1.2681564245810055,
+      "grad_norm": 0.6143464219970393,
+      "learning_rate": 3.3301513580903516e-06,
+      "loss": 0.3712,
+      "step": 2724
+    },
+    {
+      "epoch": 1.2686219739292364,
+      "grad_norm": 0.587557915647359,
+      "learning_rate": 3.32894093486781e-06,
+      "loss": 0.4047,
+      "step": 2725
+    },
+    {
+      "epoch": 1.2690875232774674,
+      "grad_norm": 0.6383590203958871,
+      "learning_rate": 3.3277302932825135e-06,
+      "loss": 0.3788,
+      "step": 2726
+    },
+    {
+      "epoch": 1.2695530726256983,
+      "grad_norm": 0.5754238489692727,
+      "learning_rate": 3.3265194336533757e-06,
+      "loss": 0.3738,
+      "step": 2727
+    },
+    {
+      "epoch": 1.2700186219739291,
+      "grad_norm": 0.594192984660223,
+      "learning_rate": 3.325308356299364e-06,
+      "loss": 0.3753,
+      "step": 2728
+    },
+    {
+      "epoch": 1.2704841713221602,
+      "grad_norm": 0.6392203751975706,
+      "learning_rate": 3.3240970615395057e-06,
+      "loss": 0.4034,
+      "step": 2729
+    },
+    {
+      "epoch": 1.270949720670391,
+      "grad_norm": 0.6144303155357209,
+      "learning_rate": 3.3228855496928848e-06,
+      "loss": 0.4026,
+      "step": 2730
+    },
+    {
+      "epoch": 1.2714152700186219,
+      "grad_norm": 0.6111147241573422,
+      "learning_rate": 3.321673821078642e-06,
+      "loss": 0.3837,
+      "step": 2731
+    },
+    {
+      "epoch": 1.271880819366853,
+      "grad_norm": 0.6042024315859773,
+      "learning_rate": 3.320461876015976e-06,
+      "loss": 0.37,
+      "step": 2732
+    },
+    {
+      "epoch": 1.2723463687150838,
+      "grad_norm": 0.6226534147892261,
+      "learning_rate": 3.319249714824141e-06,
+      "loss": 0.3665,
+      "step": 2733
+    },
+    {
+      "epoch": 1.2728119180633146,
+      "grad_norm": 0.5685716084353138,
+      "learning_rate": 3.3180373378224496e-06,
+      "loss": 0.4013,
+      "step": 2734
+    },
+    {
+      "epoch": 1.2732774674115457,
+      "grad_norm": 0.6018653997145305,
+      "learning_rate": 3.316824745330271e-06,
+      "loss": 0.3953,
+      "step": 2735
+    },
+    {
+      "epoch": 1.2737430167597765,
+      "grad_norm": 0.5913601657531428,
+      "learning_rate": 3.315611937667031e-06,
+      "loss": 0.3931,
+      "step": 2736
+    },
+    {
+      "epoch": 1.2742085661080074,
+      "grad_norm": 0.6404423261841744,
+      "learning_rate": 3.314398915152211e-06,
+      "loss": 0.3829,
+      "step": 2737
+    },
+    {
+      "epoch": 1.2746741154562384,
+      "grad_norm": 0.5639943220824,
+      "learning_rate": 3.313185678105351e-06,
+      "loss": 0.3927,
+      "step": 2738
+    },
+    {
+      "epoch": 1.2751396648044693,
+      "grad_norm": 0.6188266664742522,
+      "learning_rate": 3.3119722268460454e-06,
+      "loss": 0.3865,
+      "step": 2739
+    },
+    {
+      "epoch": 1.2756052141527001,
+      "grad_norm": 0.59867795587338,
+      "learning_rate": 3.3107585616939475e-06,
+      "loss": 0.3669,
+      "step": 2740
+    },
+    {
+      "epoch": 1.2760707635009312,
+      "grad_norm": 0.6049799249690471,
+      "learning_rate": 3.309544682968765e-06,
+      "loss": 0.3793,
+      "step": 2741
+    },
+    {
+      "epoch": 1.276536312849162,
+      "grad_norm": 0.614905639524562,
+      "learning_rate": 3.3083305909902615e-06,
+      "loss": 0.4067,
+      "step": 2742
+    },
+    {
+      "epoch": 1.2770018621973929,
+      "grad_norm": 0.5973026155438891,
+      "learning_rate": 3.307116286078259e-06,
+      "loss": 0.3833,
+      "step": 2743
+    },
+    {
+      "epoch": 1.277467411545624,
+      "grad_norm": 0.5999198300306324,
+      "learning_rate": 3.3059017685526336e-06,
+      "loss": 0.3842,
+      "step": 2744
+    },
+    {
+      "epoch": 1.2779329608938548,
+      "grad_norm": 0.5964382713710946,
+      "learning_rate": 3.304687038733318e-06,
+      "loss": 0.3921,
+      "step": 2745
+    },
+    {
+      "epoch": 1.2783985102420856,
+      "grad_norm": 0.5920338390554366,
+      "learning_rate": 3.303472096940302e-06,
+      "loss": 0.3925,
+      "step": 2746
+    },
+    {
+      "epoch": 1.2788640595903167,
+      "grad_norm": 0.5737378553316487,
+      "learning_rate": 3.3022569434936287e-06,
+      "loss": 0.3734,
+      "step": 2747
+    },
+    {
+      "epoch": 1.2793296089385475,
+      "grad_norm": 0.5800647078912308,
+      "learning_rate": 3.301041578713399e-06,
+      "loss": 0.4028,
+      "step": 2748
+    },
+    {
+      "epoch": 1.2797951582867784,
+      "grad_norm": 0.5665796885851472,
+      "learning_rate": 3.29982600291977e-06,
+      "loss": 0.3811,
+      "step": 2749
+    },
+    {
+      "epoch": 1.2802607076350094,
+      "grad_norm": 0.602569263304737,
+      "learning_rate": 3.2986102164329514e-06,
+      "loss": 0.4047,
+      "step": 2750
+    },
+    {
+      "epoch": 1.2807262569832403,
+      "grad_norm": 0.6026389764912421,
+      "learning_rate": 3.2973942195732108e-06,
+      "loss": 0.3943,
+      "step": 2751
+    },
+    {
+      "epoch": 1.2811918063314711,
+      "grad_norm": 0.587384162728069,
+      "learning_rate": 3.296178012660871e-06,
+      "loss": 0.385,
+      "step": 2752
+    },
+    {
+      "epoch": 1.2816573556797022,
+      "grad_norm": 0.587517252120419,
+      "learning_rate": 3.2949615960163107e-06,
+      "loss": 0.3946,
+      "step": 2753
+    },
+    {
+      "epoch": 1.282122905027933,
+      "grad_norm": 0.5995803305807124,
+      "learning_rate": 3.293744969959961e-06,
+      "loss": 0.3758,
+      "step": 2754
+    },
+    {
+      "epoch": 1.2825884543761639,
+      "grad_norm": 0.6109023609452379,
+      "learning_rate": 3.292528134812312e-06,
+      "loss": 0.383,
+      "step": 2755
+    },
+    {
+      "epoch": 1.2830540037243947,
+      "grad_norm": 0.5810722103377486,
+      "learning_rate": 3.291311090893906e-06,
+      "loss": 0.3928,
+      "step": 2756
+    },
+    {
+      "epoch": 1.2835195530726258,
+      "grad_norm": 0.6097815022375382,
+      "learning_rate": 3.290093838525341e-06,
+      "loss": 0.3944,
+      "step": 2757
+    },
+    {
+      "epoch": 1.2839851024208566,
+      "grad_norm": 0.6283903060206513,
+      "learning_rate": 3.2888763780272715e-06,
+      "loss": 0.3847,
+      "step": 2758
+    },
+    {
+      "epoch": 1.2844506517690875,
+      "grad_norm": 0.5977212615922658,
+      "learning_rate": 3.2876587097204044e-06,
+      "loss": 0.4055,
+      "step": 2759
+    },
+    {
+      "epoch": 1.2849162011173183,
+      "grad_norm": 0.630429946085765,
+      "learning_rate": 3.286440833925503e-06,
+      "loss": 0.3802,
+      "step": 2760
+    },
+    {
+      "epoch": 1.2853817504655494,
+      "grad_norm": 0.5973582480758943,
+      "learning_rate": 3.2852227509633844e-06,
+      "loss": 0.3921,
+      "step": 2761
+    },
+    {
+      "epoch": 1.2858472998137802,
+      "grad_norm": 0.608209005509099,
+      "learning_rate": 3.284004461154921e-06,
+      "loss": 0.3916,
+      "step": 2762
+    },
+    {
+      "epoch": 1.286312849162011,
+      "grad_norm": 0.5780874219286285,
+      "learning_rate": 3.2827859648210393e-06,
+      "loss": 0.3852,
+      "step": 2763
+    },
+    {
+      "epoch": 1.2867783985102421,
+      "grad_norm": 0.6157625474295784,
+      "learning_rate": 3.28156726228272e-06,
+      "loss": 0.3826,
+      "step": 2764
+    },
+    {
+      "epoch": 1.287243947858473,
+      "grad_norm": 0.5772075562041177,
+      "learning_rate": 3.280348353860998e-06,
+      "loss": 0.3857,
+      "step": 2765
+    },
+    {
+      "epoch": 1.2877094972067038,
+      "grad_norm": 0.5913285211270716,
+      "learning_rate": 3.279129239876964e-06,
+      "loss": 0.3906,
+      "step": 2766
+    },
+    {
+      "epoch": 1.2881750465549349,
+      "grad_norm": 0.6158916919508686,
+      "learning_rate": 3.2779099206517596e-06,
+      "loss": 0.3849,
+      "step": 2767
+    },
+    {
+      "epoch": 1.2886405959031657,
+      "grad_norm": 0.6043912432788762,
+      "learning_rate": 3.2766903965065844e-06,
+      "loss": 0.3932,
+      "step": 2768
+    },
+    {
+      "epoch": 1.2891061452513966,
+      "grad_norm": 0.636770779399659,
+      "learning_rate": 3.2754706677626895e-06,
+      "loss": 0.3744,
+      "step": 2769
+    },
+    {
+      "epoch": 1.2895716945996276,
+      "grad_norm": 0.6070224525623219,
+      "learning_rate": 3.27425073474138e-06,
+      "loss": 0.3992,
+      "step": 2770
+    },
+    {
+      "epoch": 1.2900372439478585,
+      "grad_norm": 0.5966532543166191,
+      "learning_rate": 3.2730305977640154e-06,
+      "loss": 0.39,
+      "step": 2771
+    },
+    {
+      "epoch": 1.2905027932960893,
+      "grad_norm": 0.5984290031632549,
+      "learning_rate": 3.2718102571520094e-06,
+      "loss": 0.4066,
+      "step": 2772
+    },
+    {
+      "epoch": 1.2909683426443204,
+      "grad_norm": 0.5993631047397973,
+      "learning_rate": 3.2705897132268273e-06,
+      "loss": 0.3817,
+      "step": 2773
+    },
+    {
+      "epoch": 1.2914338919925512,
+      "grad_norm": 0.620428450862728,
+      "learning_rate": 3.2693689663099915e-06,
+      "loss": 0.3854,
+      "step": 2774
+    },
+    {
+      "epoch": 1.291899441340782,
+      "grad_norm": 0.6076375976656929,
+      "learning_rate": 3.268148016723074e-06,
+      "loss": 0.3817,
+      "step": 2775
+    },
+    {
+      "epoch": 1.2923649906890131,
+      "grad_norm": 0.6276308757925999,
+      "learning_rate": 3.266926864787703e-06,
+      "loss": 0.3852,
+      "step": 2776
+    },
+    {
+      "epoch": 1.292830540037244,
+      "grad_norm": 0.597882991085975,
+      "learning_rate": 3.2657055108255575e-06,
+      "loss": 0.3843,
+      "step": 2777
+    },
+    {
+      "epoch": 1.2932960893854748,
+      "grad_norm": 0.6027724006276316,
+      "learning_rate": 3.264483955158373e-06,
+      "loss": 0.3914,
+      "step": 2778
+    },
+    {
+      "epoch": 1.2937616387337059,
+      "grad_norm": 0.6540576886836498,
+      "learning_rate": 3.263262198107935e-06,
+      "loss": 0.4062,
+      "step": 2779
+    },
+    {
+      "epoch": 1.2942271880819367,
+      "grad_norm": 0.6034201739735687,
+      "learning_rate": 3.2620402399960847e-06,
+      "loss": 0.4051,
+      "step": 2780
+    },
+    {
+      "epoch": 1.2946927374301676,
+      "grad_norm": 0.576120334006128,
+      "learning_rate": 3.260818081144713e-06,
+      "loss": 0.3632,
+      "step": 2781
+    },
+    {
+      "epoch": 1.2951582867783986,
+      "grad_norm": 0.7739008409900663,
+      "learning_rate": 3.259595721875768e-06,
+      "loss": 0.3779,
+      "step": 2782
+    },
+    {
+      "epoch": 1.2956238361266295,
+      "grad_norm": 0.5782595599416133,
+      "learning_rate": 3.2583731625112456e-06,
+      "loss": 0.3697,
+      "step": 2783
+    },
+    {
+      "epoch": 1.2960893854748603,
+      "grad_norm": 0.6103610662570266,
+      "learning_rate": 3.2571504033731994e-06,
+      "loss": 0.3963,
+      "step": 2784
+    },
+    {
+      "epoch": 1.2965549348230914,
+      "grad_norm": 0.5957888684277644,
+      "learning_rate": 3.255927444783732e-06,
+      "loss": 0.3917,
+      "step": 2785
+    },
+    {
+      "epoch": 1.2970204841713222,
+      "grad_norm": 0.5646769421832881,
+      "learning_rate": 3.2547042870649995e-06,
+      "loss": 0.3842,
+      "step": 2786
+    },
+    {
+      "epoch": 1.297486033519553,
+      "grad_norm": 0.5768644814716753,
+      "learning_rate": 3.2534809305392114e-06,
+      "loss": 0.3836,
+      "step": 2787
+    },
+    {
+      "epoch": 1.2979515828677841,
+      "grad_norm": 0.6163637209665602,
+      "learning_rate": 3.2522573755286284e-06,
+      "loss": 0.3766,
+      "step": 2788
+    },
+    {
+      "epoch": 1.298417132216015,
+      "grad_norm": 0.5818223999026891,
+      "learning_rate": 3.2510336223555646e-06,
+      "loss": 0.3862,
+      "step": 2789
+    },
+    {
+      "epoch": 1.2988826815642458,
+      "grad_norm": 0.549303174234814,
+      "learning_rate": 3.249809671342385e-06,
+      "loss": 0.3804,
+      "step": 2790
+    },
+    {
+      "epoch": 1.2993482309124766,
+      "grad_norm": 0.5943923383610612,
+      "learning_rate": 3.2485855228115074e-06,
+      "loss": 0.3949,
+      "step": 2791
+    },
+    {
+      "epoch": 1.2998137802607077,
+      "grad_norm": 0.5808119614997754,
+      "learning_rate": 3.247361177085403e-06,
+      "loss": 0.3856,
+      "step": 2792
+    },
+    {
+      "epoch": 1.3002793296089385,
+      "grad_norm": 0.5926653362466339,
+      "learning_rate": 3.246136634486592e-06,
+      "loss": 0.3982,
+      "step": 2793
+    },
+    {
+      "epoch": 1.3007448789571694,
+      "grad_norm": 0.5722020276409179,
+      "learning_rate": 3.2449118953376478e-06,
+      "loss": 0.3659,
+      "step": 2794
+    },
+    {
+      "epoch": 1.3012104283054002,
+      "grad_norm": 0.594085917595174,
+      "learning_rate": 3.243686959961198e-06,
+      "loss": 0.3703,
+      "step": 2795
+    },
+    {
+      "epoch": 1.3016759776536313,
+      "grad_norm": 0.5669395135272777,
+      "learning_rate": 3.2424618286799175e-06,
+      "loss": 0.3791,
+      "step": 2796
+    },
+    {
+      "epoch": 1.3021415270018621,
+      "grad_norm": 0.6277107868737829,
+      "learning_rate": 3.241236501816536e-06,
+      "loss": 0.3907,
+      "step": 2797
+    },
+    {
+      "epoch": 1.302607076350093,
+      "grad_norm": 0.5831432351006482,
+      "learning_rate": 3.240010979693834e-06,
+      "loss": 0.38,
+      "step": 2798
+    },
+    {
+      "epoch": 1.303072625698324,
+      "grad_norm": 0.5592311905000711,
+      "learning_rate": 3.2387852626346423e-06,
+      "loss": 0.3857,
+      "step": 2799
+    },
+    {
+      "epoch": 1.303538175046555,
+      "grad_norm": 0.6113388226824089,
+      "learning_rate": 3.2375593509618435e-06,
+      "loss": 0.3859,
+      "step": 2800
+    },
+    {
+      "epoch": 1.3040037243947857,
+      "grad_norm": 0.6081440922983709,
+      "learning_rate": 3.2363332449983737e-06,
+      "loss": 0.3927,
+      "step": 2801
+    },
+    {
+      "epoch": 1.3044692737430168,
+      "grad_norm": 0.6070864436636033,
+      "learning_rate": 3.2351069450672174e-06,
+      "loss": 0.3902,
+      "step": 2802
+    },
+    {
+      "epoch": 1.3049348230912476,
+      "grad_norm": 0.5807738314138874,
+      "learning_rate": 3.2338804514914112e-06,
+      "loss": 0.3723,
+      "step": 2803
+    },
+    {
+      "epoch": 1.3054003724394785,
+      "grad_norm": 0.6340816208644574,
+      "learning_rate": 3.2326537645940415e-06,
+      "loss": 0.3782,
+      "step": 2804
+    },
+    {
+      "epoch": 1.3058659217877095,
+      "grad_norm": 0.6011837559015297,
+      "learning_rate": 3.23142688469825e-06,
+      "loss": 0.3726,
+      "step": 2805
+    },
+    {
+      "epoch": 1.3063314711359404,
+      "grad_norm": 0.645246039310271,
+      "learning_rate": 3.230199812127222e-06,
+      "loss": 0.3976,
+      "step": 2806
+    },
+    {
+      "epoch": 1.3067970204841712,
+      "grad_norm": 0.5827129536495755,
+      "learning_rate": 3.2289725472042005e-06,
+      "loss": 0.4011,
+      "step": 2807
+    },
+    {
+      "epoch": 1.3072625698324023,
+      "grad_norm": 0.6193356762683037,
+      "learning_rate": 3.2277450902524754e-06,
+      "loss": 0.3833,
+      "step": 2808
+    },
+    {
+      "epoch": 1.3077281191806331,
+      "grad_norm": 0.6217010323898046,
+      "learning_rate": 3.2265174415953875e-06,
+      "loss": 0.3827,
+      "step": 2809
+    },
+    {
+      "epoch": 1.308193668528864,
+      "grad_norm": 0.6116018914983711,
+      "learning_rate": 3.2252896015563297e-06,
+      "loss": 0.3829,
+      "step": 2810
+    },
+    {
+      "epoch": 1.308659217877095,
+      "grad_norm": 0.5896632095506863,
+      "learning_rate": 3.2240615704587436e-06,
+      "loss": 0.3969,
+      "step": 2811
+    },
+    {
+      "epoch": 1.309124767225326,
+      "grad_norm": 0.5981056597048651,
+      "learning_rate": 3.222833348626121e-06,
+      "loss": 0.3996,
+      "step": 2812
+    },
+    {
+      "epoch": 1.3095903165735567,
+      "grad_norm": 0.621224466536991,
+      "learning_rate": 3.2216049363820067e-06,
+      "loss": 0.3825,
+      "step": 2813
+    },
+    {
+      "epoch": 1.3100558659217878,
+      "grad_norm": 0.5960952231124167,
+      "learning_rate": 3.2203763340499923e-06,
+      "loss": 0.3707,
+      "step": 2814
+    },
+    {
+      "epoch": 1.3105214152700186,
+      "grad_norm": 0.627199999707428,
+      "learning_rate": 3.219147541953721e-06,
+      "loss": 0.3803,
+      "step": 2815
+    },
+    {
+      "epoch": 1.3109869646182495,
+      "grad_norm": 0.5980471580513307,
+      "learning_rate": 3.2179185604168856e-06,
+      "loss": 0.3842,
+      "step": 2816
+    },
+    {
+      "epoch": 1.3114525139664805,
+      "grad_norm": 0.6481504757267045,
+      "learning_rate": 3.21668938976323e-06,
+      "loss": 0.3889,
+      "step": 2817
+    },
+    {
+      "epoch": 1.3119180633147114,
+      "grad_norm": 0.61624054492369,
+      "learning_rate": 3.215460030316547e-06,
+      "loss": 0.3903,
+      "step": 2818
+    },
+    {
+      "epoch": 1.3123836126629422,
+      "grad_norm": 0.597984042697172,
+      "learning_rate": 3.2142304824006782e-06,
+      "loss": 0.3666,
+      "step": 2819
+    },
+    {
+      "epoch": 1.3128491620111733,
+      "grad_norm": 0.6294519845625969,
+      "learning_rate": 3.2130007463395153e-06,
+      "loss": 0.3813,
+      "step": 2820
+    },
+    {
+      "epoch": 1.3133147113594041,
+      "grad_norm": 0.5862669690765078,
+      "learning_rate": 3.211770822457002e-06,
+      "loss": 0.3668,
+      "step": 2821
+    },
+    {
+      "epoch": 1.313780260707635,
+      "grad_norm": 0.6147830157101715,
+      "learning_rate": 3.2105407110771283e-06,
+      "loss": 0.3997,
+      "step": 2822
+    },
+    {
+      "epoch": 1.314245810055866,
+      "grad_norm": 0.5482511738967286,
+      "learning_rate": 3.2093104125239343e-06,
+      "loss": 0.3944,
+      "step": 2823
+    },
+    {
+      "epoch": 1.314711359404097,
+      "grad_norm": 0.6147111519153922,
+      "learning_rate": 3.208079927121511e-06,
+      "loss": 0.3697,
+      "step": 2824
+    },
+    {
+      "epoch": 1.3151769087523277,
+      "grad_norm": 0.591248767839899,
+      "learning_rate": 3.206849255193997e-06,
+      "loss": 0.3814,
+      "step": 2825
+    },
+    {
+      "epoch": 1.3156424581005586,
+      "grad_norm": 0.6048969806424164,
+      "learning_rate": 3.20561839706558e-06,
+      "loss": 0.3819,
+      "step": 2826
+    },
+    {
+      "epoch": 1.3161080074487896,
+      "grad_norm": 0.5911584876941681,
+      "learning_rate": 3.204387353060498e-06,
+      "loss": 0.3937,
+      "step": 2827
+    },
+    {
+      "epoch": 1.3165735567970205,
+      "grad_norm": 0.5911324342033678,
+      "learning_rate": 3.2031561235030375e-06,
+      "loss": 0.3899,
+      "step": 2828
+    },
+    {
+      "epoch": 1.3170391061452513,
+      "grad_norm": 0.6018686784743512,
+      "learning_rate": 3.2019247087175327e-06,
+      "loss": 0.3801,
+      "step": 2829
+    },
+    {
+      "epoch": 1.3175046554934824,
+      "grad_norm": 0.5961049369544374,
+      "learning_rate": 3.2006931090283676e-06,
+      "loss": 0.3898,
+      "step": 2830
+    },
+    {
+      "epoch": 1.3179702048417132,
+      "grad_norm": 0.6044397201689383,
+      "learning_rate": 3.199461324759977e-06,
+      "loss": 0.3866,
+      "step": 2831
+    },
+    {
+      "epoch": 1.318435754189944,
+      "grad_norm": 0.6078614587832593,
+      "learning_rate": 3.198229356236839e-06,
+      "loss": 0.3718,
+      "step": 2832
+    },
+    {
+      "epoch": 1.318901303538175,
+      "grad_norm": 0.6066199166634714,
+      "learning_rate": 3.196997203783485e-06,
+      "loss": 0.3891,
+      "step": 2833
+    },
+    {
+      "epoch": 1.319366852886406,
+      "grad_norm": 0.5984155802071578,
+      "learning_rate": 3.195764867724494e-06,
+      "loss": 0.3842,
+      "step": 2834
+    },
+    {
+      "epoch": 1.3198324022346368,
+      "grad_norm": 0.5883670254458923,
+      "learning_rate": 3.194532348384491e-06,
+      "loss": 0.3825,
+      "step": 2835
+    },
+    {
+      "epoch": 1.3202979515828677,
+      "grad_norm": 0.6574278615815664,
+      "learning_rate": 3.1932996460881518e-06,
+      "loss": 0.3979,
+      "step": 2836
+    },
+    {
+      "epoch": 1.3207635009310987,
+      "grad_norm": 0.6030018998058222,
+      "learning_rate": 3.1920667611601997e-06,
+      "loss": 0.3866,
+      "step": 2837
+    },
+    {
+      "epoch": 1.3212290502793296,
+      "grad_norm": 0.5632779458564731,
+      "learning_rate": 3.1908336939254057e-06,
+      "loss": 0.3642,
+      "step": 2838
+    },
+    {
+      "epoch": 1.3216945996275604,
+      "grad_norm": 0.5483229925402593,
+      "learning_rate": 3.1896004447085882e-06,
+      "loss": 0.3755,
+      "step": 2839
+    },
+    {
+      "epoch": 1.3221601489757915,
+      "grad_norm": 0.6269579380438165,
+      "learning_rate": 3.1883670138346158e-06,
+      "loss": 0.387,
+      "step": 2840
+    },
+    {
+      "epoch": 1.3226256983240223,
+      "grad_norm": 0.5927341697333683,
+      "learning_rate": 3.187133401628402e-06,
+      "loss": 0.3776,
+      "step": 2841
+    },
+    {
+      "epoch": 1.3230912476722532,
+      "grad_norm": 0.5868845849802973,
+      "learning_rate": 3.185899608414911e-06,
+      "loss": 0.3943,
+      "step": 2842
+    },
+    {
+      "epoch": 1.3235567970204842,
+      "grad_norm": 0.6219703484280656,
+      "learning_rate": 3.1846656345191523e-06,
+      "loss": 0.3932,
+      "step": 2843
+    },
+    {
+      "epoch": 1.324022346368715,
+      "grad_norm": 0.648673977213176,
+      "learning_rate": 3.1834314802661853e-06,
+      "loss": 0.3862,
+      "step": 2844
+    },
+    {
+      "epoch": 1.324487895716946,
+      "grad_norm": 0.6137020663910937,
+      "learning_rate": 3.1821971459811135e-06,
+      "loss": 0.388,
+      "step": 2845
+    },
+    {
+      "epoch": 1.324953445065177,
+      "grad_norm": 0.603972951630654,
+      "learning_rate": 3.180962631989092e-06,
+      "loss": 0.3903,
+      "step": 2846
+    },
+    {
+      "epoch": 1.3254189944134078,
+      "grad_norm": 0.5902753675746861,
+      "learning_rate": 3.179727938615319e-06,
+      "loss": 0.3801,
+      "step": 2847
+    },
+    {
+      "epoch": 1.3258845437616387,
+      "grad_norm": 0.5861876072775375,
+      "learning_rate": 3.1784930661850443e-06,
+      "loss": 0.3751,
+      "step": 2848
+    },
+    {
+      "epoch": 1.3263500931098697,
+      "grad_norm": 0.5699364963681363,
+      "learning_rate": 3.1772580150235605e-06,
+      "loss": 0.3867,
+      "step": 2849
+    },
+    {
+      "epoch": 1.3268156424581006,
+      "grad_norm": 0.6703393191494698,
+      "learning_rate": 3.176022785456212e-06,
+      "loss": 0.4,
+      "step": 2850
+    },
+    {
+      "epoch": 1.3272811918063314,
+      "grad_norm": 0.5888768957145152,
+      "learning_rate": 3.174787377808385e-06,
+      "loss": 0.3968,
+      "step": 2851
+    },
+    {
+      "epoch": 1.3277467411545625,
+      "grad_norm": 0.6758984887493156,
+      "learning_rate": 3.173551792405517e-06,
+      "loss": 0.3907,
+      "step": 2852
+    },
+    {
+      "epoch": 1.3282122905027933,
+      "grad_norm": 0.6126802408510653,
+      "learning_rate": 3.17231602957309e-06,
+      "loss": 0.3943,
+      "step": 2853
+    },
+    {
+      "epoch": 1.3286778398510242,
+      "grad_norm": 0.6287268753797101,
+      "learning_rate": 3.1710800896366335e-06,
+      "loss": 0.4106,
+      "step": 2854
+    },
+    {
+      "epoch": 1.3291433891992552,
+      "grad_norm": 0.6022482531332417,
+      "learning_rate": 3.169843972921722e-06,
+      "loss": 0.3849,
+      "step": 2855
+    },
+    {
+      "epoch": 1.329608938547486,
+      "grad_norm": 0.6015090268349514,
+      "learning_rate": 3.168607679753981e-06,
+      "loss": 0.3915,
+      "step": 2856
+    },
+    {
+      "epoch": 1.330074487895717,
+      "grad_norm": 0.5779461230006787,
+      "learning_rate": 3.167371210459077e-06,
+      "loss": 0.3872,
+      "step": 2857
+    },
+    {
+      "epoch": 1.330540037243948,
+      "grad_norm": 0.6525322225178479,
+      "learning_rate": 3.1661345653627266e-06,
+      "loss": 0.3891,
+      "step": 2858
+    },
+    {
+      "epoch": 1.3310055865921788,
+      "grad_norm": 0.6218059009365772,
+      "learning_rate": 3.1648977447906907e-06,
+      "loss": 0.3981,
+      "step": 2859
+    },
+    {
+      "epoch": 1.3314711359404097,
+      "grad_norm": 0.5861644284995633,
+      "learning_rate": 3.163660749068779e-06,
+      "loss": 0.3819,
+      "step": 2860
+    },
+    {
+      "epoch": 1.3319366852886407,
+      "grad_norm": 0.5560009496382524,
+      "learning_rate": 3.162423578522843e-06,
+      "loss": 0.4094,
+      "step": 2861
+    },
+    {
+      "epoch": 1.3324022346368716,
+      "grad_norm": 0.5994974078622146,
+      "learning_rate": 3.161186233478785e-06,
+      "loss": 0.3799,
+      "step": 2862
+    },
+    {
+      "epoch": 1.3328677839851024,
+      "grad_norm": 0.5942185654594027,
+      "learning_rate": 3.1599487142625505e-06,
+      "loss": 0.3791,
+      "step": 2863
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.634802682543955,
+      "learning_rate": 3.158711021200131e-06,
+      "loss": 0.3934,
+      "step": 2864
+    },
+    {
+      "epoch": 1.3337988826815643,
+      "grad_norm": 0.6617384206445434,
+      "learning_rate": 3.157473154617564e-06,
+      "loss": 0.4061,
+      "step": 2865
+    },
+    {
+      "epoch": 1.3342644320297952,
+      "grad_norm": 0.6138586792295327,
+      "learning_rate": 3.1562351148409344e-06,
+      "loss": 0.3994,
+      "step": 2866
+    },
+    {
+      "epoch": 1.334729981378026,
+      "grad_norm": 0.5952412902114084,
+      "learning_rate": 3.15499690219637e-06,
+      "loss": 0.3999,
+      "step": 2867
+    },
+    {
+      "epoch": 1.3351955307262569,
+      "grad_norm": 0.5714859587938834,
+      "learning_rate": 3.153758517010046e-06,
+      "loss": 0.3843,
+      "step": 2868
+    },
+    {
+      "epoch": 1.335661080074488,
+      "grad_norm": 0.621622292533005,
+      "learning_rate": 3.152519959608182e-06,
+      "loss": 0.389,
+      "step": 2869
+    },
+    {
+      "epoch": 1.3361266294227188,
+      "grad_norm": 0.5909029080302632,
+      "learning_rate": 3.151281230317045e-06,
+      "loss": 0.391,
+      "step": 2870
+    },
+    {
+      "epoch": 1.3365921787709496,
+      "grad_norm": 0.640996662091606,
+      "learning_rate": 3.1500423294629434e-06,
+      "loss": 0.3847,
+      "step": 2871
+    },
+    {
+      "epoch": 1.3370577281191807,
+      "grad_norm": 0.639045030944844,
+      "learning_rate": 3.1488032573722355e-06,
+      "loss": 0.3964,
+      "step": 2872
+    },
+    {
+      "epoch": 1.3375232774674115,
+      "grad_norm": 0.576680191908568,
+      "learning_rate": 3.1475640143713205e-06,
+      "loss": 0.4056,
+      "step": 2873
+    },
+    {
+      "epoch": 1.3379888268156424,
+      "grad_norm": 0.6171386925646785,
+      "learning_rate": 3.146324600786646e-06,
+      "loss": 0.3841,
+      "step": 2874
+    },
+    {
+      "epoch": 1.3384543761638734,
+      "grad_norm": 0.5763982072145861,
+      "learning_rate": 3.145085016944701e-06,
+      "loss": 0.3962,
+      "step": 2875
+    },
+    {
+      "epoch": 1.3389199255121043,
+      "grad_norm": 0.5977379941094381,
+      "learning_rate": 3.143845263172024e-06,
+      "loss": 0.3867,
+      "step": 2876
+    },
+    {
+      "epoch": 1.339385474860335,
+      "grad_norm": 0.5934105104135936,
+      "learning_rate": 3.142605339795194e-06,
+      "loss": 0.384,
+      "step": 2877
+    },
+    {
+      "epoch": 1.3398510242085662,
+      "grad_norm": 0.6160963305211676,
+      "learning_rate": 3.1413652471408357e-06,
+      "loss": 0.4017,
+      "step": 2878
+    },
+    {
+      "epoch": 1.340316573556797,
+      "grad_norm": 0.5919869153037327,
+      "learning_rate": 3.14012498553562e-06,
+      "loss": 0.3665,
+      "step": 2879
+    },
+    {
+      "epoch": 1.3407821229050279,
+      "grad_norm": 0.5794279089661936,
+      "learning_rate": 3.138884555306262e-06,
+      "loss": 0.3974,
+      "step": 2880
+    },
+    {
+      "epoch": 1.341247672253259,
+      "grad_norm": 0.5938732594575226,
+      "learning_rate": 3.1376439567795194e-06,
+      "loss": 0.3881,
+      "step": 2881
+    },
+    {
+      "epoch": 1.3417132216014898,
+      "grad_norm": 0.6250586579873053,
+      "learning_rate": 3.1364031902821946e-06,
+      "loss": 0.3942,
+      "step": 2882
+    },
+    {
+      "epoch": 1.3421787709497206,
+      "grad_norm": 0.6301788205293991,
+      "learning_rate": 3.1351622561411377e-06,
+      "loss": 0.3664,
+      "step": 2883
+    },
+    {
+      "epoch": 1.3426443202979517,
+      "grad_norm": 0.5694459215963666,
+      "learning_rate": 3.133921154683237e-06,
+      "loss": 0.3983,
+      "step": 2884
+    },
+    {
+      "epoch": 1.3431098696461825,
+      "grad_norm": 0.5718933625244931,
+      "learning_rate": 3.132679886235431e-06,
+      "loss": 0.3789,
+      "step": 2885
+    },
+    {
+      "epoch": 1.3435754189944134,
+      "grad_norm": 0.6085949892943974,
+      "learning_rate": 3.131438451124697e-06,
+      "loss": 0.4044,
+      "step": 2886
+    },
+    {
+      "epoch": 1.3440409683426444,
+      "grad_norm": 0.5930409739645944,
+      "learning_rate": 3.13019684967806e-06,
+      "loss": 0.3851,
+      "step": 2887
+    },
+    {
+      "epoch": 1.3445065176908753,
+      "grad_norm": 0.5570346015312471,
+      "learning_rate": 3.1289550822225866e-06,
+      "loss": 0.3939,
+      "step": 2888
+    },
+    {
+      "epoch": 1.344972067039106,
+      "grad_norm": 0.6151468134975377,
+      "learning_rate": 3.127713149085389e-06,
+      "loss": 0.3822,
+      "step": 2889
+    },
+    {
+      "epoch": 1.3454376163873372,
+      "grad_norm": 0.5910472907491371,
+      "learning_rate": 3.12647105059362e-06,
+      "loss": 0.3782,
+      "step": 2890
+    },
+    {
+      "epoch": 1.345903165735568,
+      "grad_norm": 0.6581984215715726,
+      "learning_rate": 3.12522878707448e-06,
+      "loss": 0.3971,
+      "step": 2891
+    },
+    {
+      "epoch": 1.3463687150837989,
+      "grad_norm": 0.5730267518190494,
+      "learning_rate": 3.1239863588552093e-06,
+      "loss": 0.3801,
+      "step": 2892
+    },
+    {
+      "epoch": 1.34683426443203,
+      "grad_norm": 0.601568400045848,
+      "learning_rate": 3.122743766263094e-06,
+      "loss": 0.3851,
+      "step": 2893
+    },
+    {
+      "epoch": 1.3472998137802608,
+      "grad_norm": 0.606395476013816,
+      "learning_rate": 3.1215010096254604e-06,
+      "loss": 0.4007,
+      "step": 2894
+    },
+    {
+      "epoch": 1.3477653631284916,
+      "grad_norm": 0.5990769630856514,
+      "learning_rate": 3.1202580892696833e-06,
+      "loss": 0.3914,
+      "step": 2895
+    },
+    {
+      "epoch": 1.3482309124767227,
+      "grad_norm": 0.5956486413032627,
+      "learning_rate": 3.1190150055231756e-06,
+      "loss": 0.3823,
+      "step": 2896
+    },
+    {
+      "epoch": 1.3486964618249535,
+      "grad_norm": 0.6333576798686493,
+      "learning_rate": 3.1177717587133956e-06,
+      "loss": 0.3953,
+      "step": 2897
+    },
+    {
+      "epoch": 1.3491620111731844,
+      "grad_norm": 0.5980625624768259,
+      "learning_rate": 3.1165283491678434e-06,
+      "loss": 0.3902,
+      "step": 2898
+    },
+    {
+      "epoch": 1.3496275605214152,
+      "grad_norm": 0.6215305248585461,
+      "learning_rate": 3.1152847772140636e-06,
+      "loss": 0.3653,
+      "step": 2899
+    },
+    {
+      "epoch": 1.3500931098696463,
+      "grad_norm": 0.5980403687741186,
+      "learning_rate": 3.114041043179642e-06,
+      "loss": 0.3769,
+      "step": 2900
+    },
+    {
+      "epoch": 1.350558659217877,
+      "grad_norm": 0.5926777589225187,
+      "learning_rate": 3.112797147392208e-06,
+      "loss": 0.3999,
+      "step": 2901
+    },
+    {
+      "epoch": 1.351024208566108,
+      "grad_norm": 0.647627151321544,
+      "learning_rate": 3.111553090179433e-06,
+      "loss": 0.3981,
+      "step": 2902
+    },
+    {
+      "epoch": 1.3514897579143388,
+      "grad_norm": 0.5899453826928691,
+      "learning_rate": 3.1103088718690323e-06,
+      "loss": 0.3837,
+      "step": 2903
+    },
+    {
+      "epoch": 1.3519553072625698,
+      "grad_norm": 0.579717910082826,
+      "learning_rate": 3.109064492788761e-06,
+      "loss": 0.3756,
+      "step": 2904
+    },
+    {
+      "epoch": 1.3524208566108007,
+      "grad_norm": 0.600340215052312,
+      "learning_rate": 3.107819953266419e-06,
+      "loss": 0.3923,
+      "step": 2905
+    },
+    {
+      "epoch": 1.3528864059590315,
+      "grad_norm": 0.644283007482066,
+      "learning_rate": 3.1065752536298477e-06,
+      "loss": 0.3823,
+      "step": 2906
+    },
+    {
+      "epoch": 1.3533519553072626,
+      "grad_norm": 0.595306202630443,
+      "learning_rate": 3.1053303942069307e-06,
+      "loss": 0.372,
+      "step": 2907
+    },
+    {
+      "epoch": 1.3538175046554934,
+      "grad_norm": 0.5879014766002494,
+      "learning_rate": 3.1040853753255927e-06,
+      "loss": 0.3714,
+      "step": 2908
+    },
+    {
+      "epoch": 1.3542830540037243,
+      "grad_norm": 0.6360398261367629,
+      "learning_rate": 3.1028401973138026e-06,
+      "loss": 0.3897,
+      "step": 2909
+    },
+    {
+      "epoch": 1.3547486033519553,
+      "grad_norm": 0.6244986977435711,
+      "learning_rate": 3.1015948604995682e-06,
+      "loss": 0.3846,
+      "step": 2910
+    },
+    {
+      "epoch": 1.3552141527001862,
+      "grad_norm": 0.6202293985032781,
+      "learning_rate": 3.100349365210943e-06,
+      "loss": 0.3902,
+      "step": 2911
+    },
+    {
+      "epoch": 1.355679702048417,
+      "grad_norm": 0.5907566150646826,
+      "learning_rate": 3.0991037117760183e-06,
+      "loss": 0.3904,
+      "step": 2912
+    },
+    {
+      "epoch": 1.356145251396648,
+      "grad_norm": 0.6296206980196691,
+      "learning_rate": 3.097857900522929e-06,
+      "loss": 0.3977,
+      "step": 2913
+    },
+    {
+      "epoch": 1.356610800744879,
+      "grad_norm": 0.7568710973470776,
+      "learning_rate": 3.096611931779852e-06,
+      "loss": 0.3703,
+      "step": 2914
+    },
+    {
+      "epoch": 1.3570763500931098,
+      "grad_norm": 0.6112395748478829,
+      "learning_rate": 3.0953658058750058e-06,
+      "loss": 0.3924,
+      "step": 2915
+    },
+    {
+      "epoch": 1.3575418994413408,
+      "grad_norm": 0.5897268597897314,
+      "learning_rate": 3.094119523136648e-06,
+      "loss": 0.3892,
+      "step": 2916
+    },
+    {
+      "epoch": 1.3580074487895717,
+      "grad_norm": 0.5998098396322566,
+      "learning_rate": 3.0928730838930797e-06,
+      "loss": 0.3941,
+      "step": 2917
+    },
+    {
+      "epoch": 1.3584729981378025,
+      "grad_norm": 0.6248965418577956,
+      "learning_rate": 3.091626488472643e-06,
+      "loss": 0.3812,
+      "step": 2918
+    },
+    {
+      "epoch": 1.3589385474860336,
+      "grad_norm": 0.58083453965697,
+      "learning_rate": 3.090379737203721e-06,
+      "loss": 0.3909,
+      "step": 2919
+    },
+    {
+      "epoch": 1.3594040968342644,
+      "grad_norm": 0.5930833357654126,
+      "learning_rate": 3.089132830414736e-06,
+      "loss": 0.3844,
+      "step": 2920
+    },
+    {
+      "epoch": 1.3598696461824953,
+      "grad_norm": 0.5899153324319117,
+      "learning_rate": 3.087885768434155e-06,
+      "loss": 0.3721,
+      "step": 2921
+    },
+    {
+      "epoch": 1.3603351955307263,
+      "grad_norm": 0.6004578378877354,
+      "learning_rate": 3.0866385515904828e-06,
+      "loss": 0.3995,
+      "step": 2922
+    },
+    {
+      "epoch": 1.3608007448789572,
+      "grad_norm": 0.6226692662717587,
+      "learning_rate": 3.0853911802122655e-06,
+      "loss": 0.3928,
+      "step": 2923
+    },
+    {
+      "epoch": 1.361266294227188,
+      "grad_norm": 0.6023372565389926,
+      "learning_rate": 3.084143654628091e-06,
+      "loss": 0.3754,
+      "step": 2924
+    },
+    {
+      "epoch": 1.361731843575419,
+      "grad_norm": 0.5761532742382595,
+      "learning_rate": 3.0828959751665877e-06,
+      "loss": 0.3698,
+      "step": 2925
+    },
+    {
+      "epoch": 1.36219739292365,
+      "grad_norm": 0.6478407099273158,
+      "learning_rate": 3.081648142156423e-06,
+      "loss": 0.3812,
+      "step": 2926
+    },
+    {
+      "epoch": 1.3626629422718808,
+      "grad_norm": 0.6259707412527665,
+      "learning_rate": 3.0804001559263063e-06,
+      "loss": 0.398,
+      "step": 2927
+    },
+    {
+      "epoch": 1.3631284916201118,
+      "grad_norm": 0.6004904090514128,
+      "learning_rate": 3.0791520168049876e-06,
+      "loss": 0.3769,
+      "step": 2928
+    },
+    {
+      "epoch": 1.3635940409683427,
+      "grad_norm": 0.626028567616798,
+      "learning_rate": 3.0779037251212545e-06,
+      "loss": 0.4029,
+      "step": 2929
+    },
+    {
+      "epoch": 1.3640595903165735,
+      "grad_norm": 0.6185811462098526,
+      "learning_rate": 3.076655281203938e-06,
+      "loss": 0.3816,
+      "step": 2930
+    },
+    {
+      "epoch": 1.3645251396648046,
+      "grad_norm": 0.58833006825041,
+      "learning_rate": 3.075406685381908e-06,
+      "loss": 0.3797,
+      "step": 2931
+    },
+    {
+      "epoch": 1.3649906890130354,
+      "grad_norm": 0.6045510817750065,
+      "learning_rate": 3.0741579379840748e-06,
+      "loss": 0.3813,
+      "step": 2932
+    },
+    {
+      "epoch": 1.3654562383612663,
+      "grad_norm": 0.602009433123053,
+      "learning_rate": 3.0729090393393855e-06,
+      "loss": 0.3915,
+      "step": 2933
+    },
+    {
+      "epoch": 1.3659217877094971,
+      "grad_norm": 0.6171035157839259,
+      "learning_rate": 3.0716599897768336e-06,
+      "loss": 0.3764,
+      "step": 2934
+    },
+    {
+      "epoch": 1.3663873370577282,
+      "grad_norm": 0.6005107882319578,
+      "learning_rate": 3.0704107896254452e-06,
+      "loss": 0.3903,
+      "step": 2935
+    },
+    {
+      "epoch": 1.366852886405959,
+      "grad_norm": 0.553458075967475,
+      "learning_rate": 3.0691614392142915e-06,
+      "loss": 0.3705,
+      "step": 2936
+    },
+    {
+      "epoch": 1.3673184357541899,
+      "grad_norm": 0.5945997974789071,
+      "learning_rate": 3.0679119388724797e-06,
+      "loss": 0.3942,
+      "step": 2937
+    },
+    {
+      "epoch": 1.3677839851024207,
+      "grad_norm": 0.6051810149107306,
+      "learning_rate": 3.0666622889291594e-06,
+      "loss": 0.4051,
+      "step": 2938
+    },
+    {
+      "epoch": 1.3682495344506518,
+      "grad_norm": 0.65622891929995,
+      "learning_rate": 3.0654124897135158e-06,
+      "loss": 0.3789,
+      "step": 2939
+    },
+    {
+      "epoch": 1.3687150837988826,
+      "grad_norm": 0.6292691873022757,
+      "learning_rate": 3.0641625415547776e-06,
+      "loss": 0.3821,
+      "step": 2940
+    },
+    {
+      "epoch": 1.3691806331471135,
+      "grad_norm": 0.6562096792499063,
+      "learning_rate": 3.0629124447822105e-06,
+      "loss": 0.3848,
+      "step": 2941
+    },
+    {
+      "epoch": 1.3696461824953445,
+      "grad_norm": 0.5890595948565808,
+      "learning_rate": 3.0616621997251195e-06,
+      "loss": 0.39,
+      "step": 2942
+    },
+    {
+      "epoch": 1.3701117318435754,
+      "grad_norm": 0.5668612138947984,
+      "learning_rate": 3.0604118067128487e-06,
+      "loss": 0.3697,
+      "step": 2943
+    },
+    {
+      "epoch": 1.3705772811918062,
+      "grad_norm": 0.586192686166284,
+      "learning_rate": 3.0591612660747818e-06,
+      "loss": 0.3894,
+      "step": 2944
+    },
+    {
+      "epoch": 1.3710428305400373,
+      "grad_norm": 0.6008099397628004,
+      "learning_rate": 3.0579105781403407e-06,
+      "loss": 0.3927,
+      "step": 2945
+    },
+    {
+      "epoch": 1.3715083798882681,
+      "grad_norm": 0.6948062424257231,
+      "learning_rate": 3.0566597432389866e-06,
+      "loss": 0.3984,
+      "step": 2946
+    },
+    {
+      "epoch": 1.371973929236499,
+      "grad_norm": 0.604067227434425,
+      "learning_rate": 3.0554087617002183e-06,
+      "loss": 0.3967,
+      "step": 2947
+    },
+    {
+      "epoch": 1.37243947858473,
+      "grad_norm": 0.6465513678766612,
+      "learning_rate": 3.0541576338535757e-06,
+      "loss": 0.3684,
+      "step": 2948
+    },
+    {
+      "epoch": 1.3729050279329609,
+      "grad_norm": 0.6137410827960412,
+      "learning_rate": 3.0529063600286333e-06,
+      "loss": 0.3967,
+      "step": 2949
+    },
+    {
+      "epoch": 1.3733705772811917,
+      "grad_norm": 0.6232105308423384,
+      "learning_rate": 3.051654940555009e-06,
+      "loss": 0.3707,
+      "step": 2950
+    },
+    {
+      "epoch": 1.3738361266294228,
+      "grad_norm": 0.6152824133602466,
+      "learning_rate": 3.0504033757623548e-06,
+      "loss": 0.3897,
+      "step": 2951
+    },
+    {
+      "epoch": 1.3743016759776536,
+      "grad_norm": 0.5747902239296994,
+      "learning_rate": 3.0491516659803634e-06,
+      "loss": 0.3902,
+      "step": 2952
+    },
+    {
+      "epoch": 1.3747672253258845,
+      "grad_norm": 0.5627723335341648,
+      "learning_rate": 3.0478998115387633e-06,
+      "loss": 0.383,
+      "step": 2953
+    },
+    {
+      "epoch": 1.3752327746741155,
+      "grad_norm": 0.6214512131453283,
+      "learning_rate": 3.046647812767325e-06,
+      "loss": 0.391,
+      "step": 2954
+    },
+    {
+      "epoch": 1.3756983240223464,
+      "grad_norm": 0.5947946591006014,
+      "learning_rate": 3.045395669995853e-06,
+      "loss": 0.3804,
+      "step": 2955
+    },
+    {
+      "epoch": 1.3761638733705772,
+      "grad_norm": 0.628172650040406,
+      "learning_rate": 3.0441433835541926e-06,
+      "loss": 0.3961,
+      "step": 2956
+    },
+    {
+      "epoch": 1.3766294227188083,
+      "grad_norm": 0.5968718813293417,
+      "learning_rate": 3.042890953772225e-06,
+      "loss": 0.3905,
+      "step": 2957
+    },
+    {
+      "epoch": 1.3770949720670391,
+      "grad_norm": 0.5898866560402247,
+      "learning_rate": 3.0416383809798716e-06,
+      "loss": 0.3824,
+      "step": 2958
+    },
+    {
+      "epoch": 1.37756052141527,
+      "grad_norm": 0.5753484107424,
+      "learning_rate": 3.0403856655070874e-06,
+      "loss": 0.381,
+      "step": 2959
+    },
+    {
+      "epoch": 1.378026070763501,
+      "grad_norm": 0.6039690218732342,
+      "learning_rate": 3.0391328076838694e-06,
+      "loss": 0.3819,
+      "step": 2960
+    },
+    {
+      "epoch": 1.3784916201117319,
+      "grad_norm": 0.6260574251701141,
+      "learning_rate": 3.0378798078402494e-06,
+      "loss": 0.3929,
+      "step": 2961
+    },
+    {
+      "epoch": 1.3789571694599627,
+      "grad_norm": 0.6047231222196139,
+      "learning_rate": 3.036626666306297e-06,
+      "loss": 0.3737,
+      "step": 2962
+    },
+    {
+      "epoch": 1.3794227188081938,
+      "grad_norm": 0.6086544264371089,
+      "learning_rate": 3.0353733834121195e-06,
+      "loss": 0.3912,
+      "step": 2963
+    },
+    {
+      "epoch": 1.3798882681564246,
+      "grad_norm": 0.5668923921699344,
+      "learning_rate": 3.034119959487863e-06,
+      "loss": 0.3772,
+      "step": 2964
+    },
+    {
+      "epoch": 1.3803538175046555,
+      "grad_norm": 0.6749530701557129,
+      "learning_rate": 3.0328663948637073e-06,
+      "loss": 0.4076,
+      "step": 2965
+    },
+    {
+      "epoch": 1.3808193668528865,
+      "grad_norm": 0.6490328138953604,
+      "learning_rate": 3.031612689869871e-06,
+      "loss": 0.41,
+      "step": 2966
+    },
+    {
+      "epoch": 1.3812849162011174,
+      "grad_norm": 0.5947365112819977,
+      "learning_rate": 3.0303588448366124e-06,
+      "loss": 0.3825,
+      "step": 2967
+    },
+    {
+      "epoch": 1.3817504655493482,
+      "grad_norm": 0.5764577512227144,
+      "learning_rate": 3.0291048600942208e-06,
+      "loss": 0.3837,
+      "step": 2968
+    },
+    {
+      "epoch": 1.3822160148975793,
+      "grad_norm": 0.6386390821040331,
+      "learning_rate": 3.0278507359730282e-06,
+      "loss": 0.4025,
+      "step": 2969
+    },
+    {
+      "epoch": 1.3826815642458101,
+      "grad_norm": 0.6193941949787135,
+      "learning_rate": 3.0265964728033988e-06,
+      "loss": 0.3864,
+      "step": 2970
+    },
+    {
+      "epoch": 1.383147113594041,
+      "grad_norm": 0.6431020963671064,
+      "learning_rate": 3.0253420709157366e-06,
+      "loss": 0.392,
+      "step": 2971
+    },
+    {
+      "epoch": 1.3836126629422718,
+      "grad_norm": 0.6242867861798113,
+      "learning_rate": 3.02408753064048e-06,
+      "loss": 0.3841,
+      "step": 2972
+    },
+    {
+      "epoch": 1.3840782122905029,
+      "grad_norm": 0.6208996151756712,
+      "learning_rate": 3.022832852308105e-06,
+      "loss": 0.3841,
+      "step": 2973
+    },
+    {
+      "epoch": 1.3845437616387337,
+      "grad_norm": 0.6233862717875297,
+      "learning_rate": 3.0215780362491246e-06,
+      "loss": 0.4089,
+      "step": 2974
+    },
+    {
+      "epoch": 1.3850093109869646,
+      "grad_norm": 0.6020773140570329,
+      "learning_rate": 3.0203230827940864e-06,
+      "loss": 0.3865,
+      "step": 2975
+    },
+    {
+      "epoch": 1.3854748603351954,
+      "grad_norm": 0.6377427951695842,
+      "learning_rate": 3.0190679922735742e-06,
+      "loss": 0.4164,
+      "step": 2976
+    },
+    {
+      "epoch": 1.3859404096834265,
+      "grad_norm": 0.6053245805232772,
+      "learning_rate": 3.017812765018211e-06,
+      "loss": 0.3485,
+      "step": 2977
+    },
+    {
+      "epoch": 1.3864059590316573,
+      "grad_norm": 0.5980782620363462,
+      "learning_rate": 3.016557401358651e-06,
+      "loss": 0.3986,
+      "step": 2978
+    },
+    {
+      "epoch": 1.3868715083798882,
+      "grad_norm": 0.6013934711707926,
+      "learning_rate": 3.0153019016255886e-06,
+      "loss": 0.3859,
+      "step": 2979
+    },
+    {
+      "epoch": 1.3873370577281192,
+      "grad_norm": 0.6036882934037096,
+      "learning_rate": 3.014046266149752e-06,
+      "loss": 0.3721,
+      "step": 2980
+    },
+    {
+      "epoch": 1.38780260707635,
+      "grad_norm": 0.6108381597198363,
+      "learning_rate": 3.0127904952619036e-06,
+      "loss": 0.3889,
+      "step": 2981
+    },
+    {
+      "epoch": 1.388268156424581,
+      "grad_norm": 0.6346742285447692,
+      "learning_rate": 3.011534589292845e-06,
+      "loss": 0.3894,
+      "step": 2982
+    },
+    {
+      "epoch": 1.388733705772812,
+      "grad_norm": 0.5964407300573925,
+      "learning_rate": 3.010278548573412e-06,
+      "loss": 0.4026,
+      "step": 2983
+    },
+    {
+      "epoch": 1.3891992551210428,
+      "grad_norm": 0.6087824757859458,
+      "learning_rate": 3.0090223734344755e-06,
+      "loss": 0.3919,
+      "step": 2984
+    },
+    {
+      "epoch": 1.3896648044692737,
+      "grad_norm": 0.6198467108814845,
+      "learning_rate": 3.0077660642069407e-06,
+      "loss": 0.3865,
+      "step": 2985
+    },
+    {
+      "epoch": 1.3901303538175047,
+      "grad_norm": 0.5960348582770119,
+      "learning_rate": 3.00650962122175e-06,
+      "loss": 0.3871,
+      "step": 2986
+    },
+    {
+      "epoch": 1.3905959031657356,
+      "grad_norm": 0.5999942735066773,
+      "learning_rate": 3.005253044809881e-06,
+      "loss": 0.3882,
+      "step": 2987
+    },
+    {
+      "epoch": 1.3910614525139664,
+      "grad_norm": 0.58299259771738,
+      "learning_rate": 3.0039963353023445e-06,
+      "loss": 0.3841,
+      "step": 2988
+    },
+    {
+      "epoch": 1.3915270018621975,
+      "grad_norm": 0.5949817545887606,
+      "learning_rate": 3.002739493030189e-06,
+      "loss": 0.4033,
+      "step": 2989
+    },
+    {
+      "epoch": 1.3919925512104283,
+      "grad_norm": 0.5843780185765143,
+      "learning_rate": 3.001482518324496e-06,
+      "loss": 0.3992,
+      "step": 2990
+    },
+    {
+      "epoch": 1.3924581005586592,
+      "grad_norm": 0.5807133372117187,
+      "learning_rate": 3.0002254115163827e-06,
+      "loss": 0.3881,
+      "step": 2991
+    },
+    {
+      "epoch": 1.3929236499068902,
+      "grad_norm": 0.5833710838980651,
+      "learning_rate": 2.9989681729369994e-06,
+      "loss": 0.3799,
+      "step": 2992
+    },
+    {
+      "epoch": 1.393389199255121,
+      "grad_norm": 0.5882142065587177,
+      "learning_rate": 2.9977108029175354e-06,
+      "loss": 0.3735,
+      "step": 2993
+    },
+    {
+      "epoch": 1.393854748603352,
+      "grad_norm": 0.6178268153659667,
+      "learning_rate": 2.9964533017892106e-06,
+      "loss": 0.3877,
+      "step": 2994
+    },
+    {
+      "epoch": 1.394320297951583,
+      "grad_norm": 0.6393304670100091,
+      "learning_rate": 2.995195669883281e-06,
+      "loss": 0.3867,
+      "step": 2995
+    },
+    {
+      "epoch": 1.3947858472998138,
+      "grad_norm": 0.5776068249794458,
+      "learning_rate": 2.993937907531035e-06,
+      "loss": 0.4012,
+      "step": 2996
+    },
+    {
+      "epoch": 1.3952513966480447,
+      "grad_norm": 0.6284298494851791,
+      "learning_rate": 2.9926800150637996e-06,
+      "loss": 0.3851,
+      "step": 2997
+    },
+    {
+      "epoch": 1.3957169459962757,
+      "grad_norm": 0.5936640263534851,
+      "learning_rate": 2.9914219928129314e-06,
+      "loss": 0.3739,
+      "step": 2998
+    },
+    {
+      "epoch": 1.3961824953445066,
+      "grad_norm": 0.5937222693651365,
+      "learning_rate": 2.990163841109825e-06,
+      "loss": 0.3695,
+      "step": 2999
+    },
+    {
+      "epoch": 1.3966480446927374,
+      "grad_norm": 0.5916074787671464,
+      "learning_rate": 2.9889055602859072e-06,
+      "loss": 0.3741,
+      "step": 3000
+    },
+    {
+      "epoch": 1.3971135940409685,
+      "grad_norm": 0.5999459336256319,
+      "learning_rate": 2.9876471506726383e-06,
+      "loss": 0.3977,
+      "step": 3001
+    },
+    {
+      "epoch": 1.3975791433891993,
+      "grad_norm": 0.587820479997653,
+      "learning_rate": 2.986388612601514e-06,
+      "loss": 0.362,
+      "step": 3002
+    },
+    {
+      "epoch": 1.3980446927374302,
+      "grad_norm": 0.5769256641451316,
+      "learning_rate": 2.9851299464040627e-06,
+      "loss": 0.3911,
+      "step": 3003
+    },
+    {
+      "epoch": 1.3985102420856612,
+      "grad_norm": 0.6209602426672423,
+      "learning_rate": 2.983871152411848e-06,
+      "loss": 0.3977,
+      "step": 3004
+    },
+    {
+      "epoch": 1.398975791433892,
+      "grad_norm": 0.5899894934836312,
+      "learning_rate": 2.9826122309564647e-06,
+      "loss": 0.3914,
+      "step": 3005
+    },
+    {
+      "epoch": 1.399441340782123,
+      "grad_norm": 0.5880690801582216,
+      "learning_rate": 2.981353182369544e-06,
+      "loss": 0.4021,
+      "step": 3006
+    },
+    {
+      "epoch": 1.3999068901303537,
+      "grad_norm": 0.5830136283585784,
+      "learning_rate": 2.980094006982749e-06,
+      "loss": 0.3829,
+      "step": 3007
+    },
+    {
+      "epoch": 1.4003724394785848,
+      "grad_norm": 0.586601795361042,
+      "learning_rate": 2.9788347051277743e-06,
+      "loss": 0.382,
+      "step": 3008
+    },
+    {
+      "epoch": 1.4008379888268156,
+      "grad_norm": 0.5857217804853697,
+      "learning_rate": 2.977575277136353e-06,
+      "loss": 0.3841,
+      "step": 3009
+    },
+    {
+      "epoch": 1.4013035381750465,
+      "grad_norm": 0.6359351399660145,
+      "learning_rate": 2.9763157233402474e-06,
+      "loss": 0.3829,
+      "step": 3010
+    },
+    {
+      "epoch": 1.4017690875232773,
+      "grad_norm": 0.6488681713554728,
+      "learning_rate": 2.9750560440712527e-06,
+      "loss": 0.3748,
+      "step": 3011
+    },
+    {
+      "epoch": 1.4022346368715084,
+      "grad_norm": 0.6057205895869912,
+      "learning_rate": 2.9737962396611997e-06,
+      "loss": 0.3524,
+      "step": 3012
+    },
+    {
+      "epoch": 1.4027001862197392,
+      "grad_norm": 0.6111511687010024,
+      "learning_rate": 2.9725363104419506e-06,
+      "loss": 0.3862,
+      "step": 3013
+    },
+    {
+      "epoch": 1.40316573556797,
+      "grad_norm": 0.6459656562051522,
+      "learning_rate": 2.9712762567454005e-06,
+      "loss": 0.3971,
+      "step": 3014
+    },
+    {
+      "epoch": 1.4036312849162011,
+      "grad_norm": 0.5733490707859357,
+      "learning_rate": 2.9700160789034776e-06,
+      "loss": 0.3725,
+      "step": 3015
+    },
+    {
+      "epoch": 1.404096834264432,
+      "grad_norm": 0.6386242456732323,
+      "learning_rate": 2.9687557772481436e-06,
+      "loss": 0.383,
+      "step": 3016
+    },
+    {
+      "epoch": 1.4045623836126628,
+      "grad_norm": 0.6272842172003553,
+      "learning_rate": 2.967495352111389e-06,
+      "loss": 0.3619,
+      "step": 3017
+    },
+    {
+      "epoch": 1.405027932960894,
+      "grad_norm": 0.6120092982766008,
+      "learning_rate": 2.9662348038252432e-06,
+      "loss": 0.3939,
+      "step": 3018
+    },
+    {
+      "epoch": 1.4054934823091247,
+      "grad_norm": 0.5995289599127573,
+      "learning_rate": 2.964974132721763e-06,
+      "loss": 0.3862,
+      "step": 3019
+    },
+    {
+      "epoch": 1.4059590316573556,
+      "grad_norm": 0.6151996653372488,
+      "learning_rate": 2.963713339133039e-06,
+      "loss": 0.37,
+      "step": 3020
+    },
+    {
+      "epoch": 1.4064245810055866,
+      "grad_norm": 0.6324095703359754,
+      "learning_rate": 2.9624524233911944e-06,
+      "loss": 0.3932,
+      "step": 3021
+    },
+    {
+      "epoch": 1.4068901303538175,
+      "grad_norm": 0.5972912471809244,
+      "learning_rate": 2.9611913858283848e-06,
+      "loss": 0.3878,
+      "step": 3022
+    },
+    {
+      "epoch": 1.4073556797020483,
+      "grad_norm": 0.6195711608283835,
+      "learning_rate": 2.959930226776797e-06,
+      "loss": 0.3979,
+      "step": 3023
+    },
+    {
+      "epoch": 1.4078212290502794,
+      "grad_norm": 0.5896298829542523,
+      "learning_rate": 2.958668946568651e-06,
+      "loss": 0.3813,
+      "step": 3024
+    },
+    {
+      "epoch": 1.4082867783985102,
+      "grad_norm": 0.5849756305416696,
+      "learning_rate": 2.957407545536196e-06,
+      "loss": 0.3854,
+      "step": 3025
+    },
+    {
+      "epoch": 1.408752327746741,
+      "grad_norm": 0.615520976769167,
+      "learning_rate": 2.9561460240117183e-06,
+      "loss": 0.3652,
+      "step": 3026
+    },
+    {
+      "epoch": 1.4092178770949721,
+      "grad_norm": 0.6490105804688729,
+      "learning_rate": 2.9548843823275296e-06,
+      "loss": 0.3883,
+      "step": 3027
+    },
+    {
+      "epoch": 1.409683426443203,
+      "grad_norm": 0.5705129311098465,
+      "learning_rate": 2.9536226208159785e-06,
+      "loss": 0.3883,
+      "step": 3028
+    },
+    {
+      "epoch": 1.4101489757914338,
+      "grad_norm": 0.6217457231208676,
+      "learning_rate": 2.9523607398094415e-06,
+      "loss": 0.3771,
+      "step": 3029
+    },
+    {
+      "epoch": 1.410614525139665,
+      "grad_norm": 0.5946504503430932,
+      "learning_rate": 2.951098739640329e-06,
+      "loss": 0.3803,
+      "step": 3030
+    },
+    {
+      "epoch": 1.4110800744878957,
+      "grad_norm": 0.5989449630203497,
+      "learning_rate": 2.949836620641081e-06,
+      "loss": 0.3768,
+      "step": 3031
+    },
+    {
+      "epoch": 1.4115456238361266,
+      "grad_norm": 0.6091416579909754,
+      "learning_rate": 2.9485743831441705e-06,
+      "loss": 0.3887,
+      "step": 3032
+    },
+    {
+      "epoch": 1.4120111731843576,
+      "grad_norm": 0.6015949575794285,
+      "learning_rate": 2.9473120274821005e-06,
+      "loss": 0.3806,
+      "step": 3033
+    },
+    {
+      "epoch": 1.4124767225325885,
+      "grad_norm": 0.5892290815282336,
+      "learning_rate": 2.9460495539874057e-06,
+      "loss": 0.3947,
+      "step": 3034
+    },
+    {
+      "epoch": 1.4129422718808193,
+      "grad_norm": 0.6185633055087628,
+      "learning_rate": 2.9447869629926507e-06,
+      "loss": 0.3914,
+      "step": 3035
+    },
+    {
+      "epoch": 1.4134078212290504,
+      "grad_norm": 0.5716399452357254,
+      "learning_rate": 2.9435242548304336e-06,
+      "loss": 0.3863,
+      "step": 3036
+    },
+    {
+      "epoch": 1.4138733705772812,
+      "grad_norm": 0.5976386141144825,
+      "learning_rate": 2.942261429833381e-06,
+      "loss": 0.3702,
+      "step": 3037
+    },
+    {
+      "epoch": 1.414338919925512,
+      "grad_norm": 0.5825238945412856,
+      "learning_rate": 2.94099848833415e-06,
+      "loss": 0.3985,
+      "step": 3038
+    },
+    {
+      "epoch": 1.4148044692737431,
+      "grad_norm": 0.6069644420532623,
+      "learning_rate": 2.939735430665431e-06,
+      "loss": 0.4033,
+      "step": 3039
+    },
+    {
+      "epoch": 1.415270018621974,
+      "grad_norm": 0.5935056672417579,
+      "learning_rate": 2.9384722571599435e-06,
+      "loss": 0.3788,
+      "step": 3040
+    },
+    {
+      "epoch": 1.4157355679702048,
+      "grad_norm": 0.6010029480116569,
+      "learning_rate": 2.937208968150435e-06,
+      "loss": 0.3976,
+      "step": 3041
+    },
+    {
+      "epoch": 1.4162011173184357,
+      "grad_norm": 0.5908781883533135,
+      "learning_rate": 2.9359455639696887e-06,
+      "loss": 0.3677,
+      "step": 3042
+    },
+    {
+      "epoch": 1.4166666666666667,
+      "grad_norm": 0.5965697157171944,
+      "learning_rate": 2.9346820449505142e-06,
+      "loss": 0.3748,
+      "step": 3043
+    },
+    {
+      "epoch": 1.4171322160148976,
+      "grad_norm": 0.5932379519074764,
+      "learning_rate": 2.9334184114257523e-06,
+      "loss": 0.384,
+      "step": 3044
+    },
+    {
+      "epoch": 1.4175977653631284,
+      "grad_norm": 0.626113851568326,
+      "learning_rate": 2.932154663728274e-06,
+      "loss": 0.4003,
+      "step": 3045
+    },
+    {
+      "epoch": 1.4180633147113593,
+      "grad_norm": 0.5899017706726124,
+      "learning_rate": 2.9308908021909814e-06,
+      "loss": 0.3879,
+      "step": 3046
+    },
+    {
+      "epoch": 1.4185288640595903,
+      "grad_norm": 0.5932353788715778,
+      "learning_rate": 2.9296268271468042e-06,
+      "loss": 0.3719,
+      "step": 3047
+    },
+    {
+      "epoch": 1.4189944134078212,
+      "grad_norm": 0.6071779626763455,
+      "learning_rate": 2.928362738928705e-06,
+      "loss": 0.3871,
+      "step": 3048
+    },
+    {
+      "epoch": 1.419459962756052,
+      "grad_norm": 0.5886019027692799,
+      "learning_rate": 2.927098537869674e-06,
+      "loss": 0.3831,
+      "step": 3049
+    },
+    {
+      "epoch": 1.419925512104283,
+      "grad_norm": 0.5833560646971547,
+      "learning_rate": 2.9258342243027318e-06,
+      "loss": 0.3883,
+      "step": 3050
+    },
+    {
+      "epoch": 1.420391061452514,
+      "grad_norm": 0.6066682463803755,
+      "learning_rate": 2.9245697985609283e-06,
+      "loss": 0.3947,
+      "step": 3051
+    },
+    {
+      "epoch": 1.4208566108007448,
+      "grad_norm": 0.6178349196550498,
+      "learning_rate": 2.9233052609773448e-06,
+      "loss": 0.3654,
+      "step": 3052
+    },
+    {
+      "epoch": 1.4213221601489758,
+      "grad_norm": 0.5809585939727105,
+      "learning_rate": 2.922040611885088e-06,
+      "loss": 0.3825,
+      "step": 3053
+    },
+    {
+      "epoch": 1.4217877094972067,
+      "grad_norm": 0.5835466155155531,
+      "learning_rate": 2.9207758516172984e-06,
+      "loss": 0.3818,
+      "step": 3054
+    },
+    {
+      "epoch": 1.4222532588454375,
+      "grad_norm": 0.5912467208988756,
+      "learning_rate": 2.9195109805071446e-06,
+      "loss": 0.3887,
+      "step": 3055
+    },
+    {
+      "epoch": 1.4227188081936686,
+      "grad_norm": 0.5967237763527877,
+      "learning_rate": 2.9182459988878215e-06,
+      "loss": 0.3934,
+      "step": 3056
+    },
+    {
+      "epoch": 1.4231843575418994,
+      "grad_norm": 0.6156575605988892,
+      "learning_rate": 2.916980907092557e-06,
+      "loss": 0.3808,
+      "step": 3057
+    },
+    {
+      "epoch": 1.4236499068901303,
+      "grad_norm": 0.5887164751433397,
+      "learning_rate": 2.9157157054546064e-06,
+      "loss": 0.3729,
+      "step": 3058
+    },
+    {
+      "epoch": 1.4241154562383613,
+      "grad_norm": 0.6014745829121734,
+      "learning_rate": 2.914450394307253e-06,
+      "loss": 0.4088,
+      "step": 3059
+    },
+    {
+      "epoch": 1.4245810055865922,
+      "grad_norm": 0.5874718192162397,
+      "learning_rate": 2.9131849739838096e-06,
+      "loss": 0.3704,
+      "step": 3060
+    },
+    {
+      "epoch": 1.425046554934823,
+      "grad_norm": 0.6177323105002356,
+      "learning_rate": 2.9119194448176196e-06,
+      "loss": 0.3722,
+      "step": 3061
+    },
+    {
+      "epoch": 1.425512104283054,
+      "grad_norm": 0.5456186082451866,
+      "learning_rate": 2.910653807142052e-06,
+      "loss": 0.3745,
+      "step": 3062
+    },
+    {
+      "epoch": 1.425977653631285,
+      "grad_norm": 0.60210037707685,
+      "learning_rate": 2.9093880612905067e-06,
+      "loss": 0.4038,
+      "step": 3063
+    },
+    {
+      "epoch": 1.4264432029795158,
+      "grad_norm": 0.6256269148331136,
+      "learning_rate": 2.9081222075964107e-06,
+      "loss": 0.3925,
+      "step": 3064
+    },
+    {
+      "epoch": 1.4269087523277468,
+      "grad_norm": 0.6826243384086451,
+      "learning_rate": 2.90685624639322e-06,
+      "loss": 0.3884,
+      "step": 3065
+    },
+    {
+      "epoch": 1.4273743016759777,
+      "grad_norm": 0.5789224963524368,
+      "learning_rate": 2.90559017801442e-06,
+      "loss": 0.3839,
+      "step": 3066
+    },
+    {
+      "epoch": 1.4278398510242085,
+      "grad_norm": 0.5928509590737158,
+      "learning_rate": 2.9043240027935215e-06,
+      "loss": 0.3863,
+      "step": 3067
+    },
+    {
+      "epoch": 1.4283054003724396,
+      "grad_norm": 0.5986808197800385,
+      "learning_rate": 2.903057721064066e-06,
+      "loss": 0.3865,
+      "step": 3068
+    },
+    {
+      "epoch": 1.4287709497206704,
+      "grad_norm": 0.6212409818009271,
+      "learning_rate": 2.9017913331596226e-06,
+      "loss": 0.378,
+      "step": 3069
+    },
+    {
+      "epoch": 1.4292364990689013,
+      "grad_norm": 0.5903236530444862,
+      "learning_rate": 2.900524839413787e-06,
+      "loss": 0.3964,
+      "step": 3070
+    },
+    {
+      "epoch": 1.4297020484171323,
+      "grad_norm": 0.5829721060959666,
+      "learning_rate": 2.899258240160184e-06,
+      "loss": 0.3864,
+      "step": 3071
+    },
+    {
+      "epoch": 1.4301675977653632,
+      "grad_norm": 0.5976046837843693,
+      "learning_rate": 2.8979915357324674e-06,
+      "loss": 0.3874,
+      "step": 3072
+    },
+    {
+      "epoch": 1.430633147113594,
+      "grad_norm": 0.6170960347525252,
+      "learning_rate": 2.896724726464316e-06,
+      "loss": 0.3983,
+      "step": 3073
+    },
+    {
+      "epoch": 1.431098696461825,
+      "grad_norm": 0.5814671162450225,
+      "learning_rate": 2.895457812689437e-06,
+      "loss": 0.396,
+      "step": 3074
+    },
+    {
+      "epoch": 1.431564245810056,
+      "grad_norm": 0.5836537051596054,
+      "learning_rate": 2.894190794741567e-06,
+      "loss": 0.3735,
+      "step": 3075
+    },
+    {
+      "epoch": 1.4320297951582868,
+      "grad_norm": 0.6062115154404314,
+      "learning_rate": 2.8929236729544674e-06,
+      "loss": 0.3838,
+      "step": 3076
+    },
+    {
+      "epoch": 1.4324953445065176,
+      "grad_norm": 0.5357905449471299,
+      "learning_rate": 2.891656447661929e-06,
+      "loss": 0.391,
+      "step": 3077
+    },
+    {
+      "epoch": 1.4329608938547487,
+      "grad_norm": 0.6199124909117877,
+      "learning_rate": 2.8903891191977686e-06,
+      "loss": 0.4054,
+      "step": 3078
+    },
+    {
+      "epoch": 1.4334264432029795,
+      "grad_norm": 0.6017520427282644,
+      "learning_rate": 2.889121687895831e-06,
+      "loss": 0.3824,
+      "step": 3079
+    },
+    {
+      "epoch": 1.4338919925512104,
+      "grad_norm": 0.562046387956965,
+      "learning_rate": 2.8878541540899873e-06,
+      "loss": 0.381,
+      "step": 3080
+    },
+    {
+      "epoch": 1.4343575418994414,
+      "grad_norm": 0.586631401335633,
+      "learning_rate": 2.886586518114137e-06,
+      "loss": 0.3751,
+      "step": 3081
+    },
+    {
+      "epoch": 1.4348230912476723,
+      "grad_norm": 0.5924496059571155,
+      "learning_rate": 2.885318780302205e-06,
+      "loss": 0.3693,
+      "step": 3082
+    },
+    {
+      "epoch": 1.435288640595903,
+      "grad_norm": 0.5842482784806153,
+      "learning_rate": 2.8840509409881427e-06,
+      "loss": 0.3759,
+      "step": 3083
+    },
+    {
+      "epoch": 1.435754189944134,
+      "grad_norm": 0.5981026013673245,
+      "learning_rate": 2.8827830005059297e-06,
+      "loss": 0.3995,
+      "step": 3084
+    },
+    {
+      "epoch": 1.436219739292365,
+      "grad_norm": 0.6444384237563603,
+      "learning_rate": 2.8815149591895736e-06,
+      "loss": 0.3934,
+      "step": 3085
+    },
+    {
+      "epoch": 1.4366852886405959,
+      "grad_norm": 0.6077030315048633,
+      "learning_rate": 2.8802468173731035e-06,
+      "loss": 0.3902,
+      "step": 3086
+    },
+    {
+      "epoch": 1.4371508379888267,
+      "grad_norm": 0.5879339398837726,
+      "learning_rate": 2.87897857539058e-06,
+      "loss": 0.3861,
+      "step": 3087
+    },
+    {
+      "epoch": 1.4376163873370578,
+      "grad_norm": 0.5987899341162742,
+      "learning_rate": 2.8777102335760876e-06,
+      "loss": 0.3815,
+      "step": 3088
+    },
+    {
+      "epoch": 1.4380819366852886,
+      "grad_norm": 0.5929136666678108,
+      "learning_rate": 2.876441792263738e-06,
+      "loss": 0.3861,
+      "step": 3089
+    },
+    {
+      "epoch": 1.4385474860335195,
+      "grad_norm": 0.5869231601915448,
+      "learning_rate": 2.8751732517876683e-06,
+      "loss": 0.3737,
+      "step": 3090
+    },
+    {
+      "epoch": 1.4390130353817505,
+      "grad_norm": 0.6171075845516047,
+      "learning_rate": 2.8739046124820434e-06,
+      "loss": 0.3901,
+      "step": 3091
+    },
+    {
+      "epoch": 1.4394785847299814,
+      "grad_norm": 0.5872730927774541,
+      "learning_rate": 2.872635874681051e-06,
+      "loss": 0.38,
+      "step": 3092
+    },
+    {
+      "epoch": 1.4399441340782122,
+      "grad_norm": 0.6008915345940883,
+      "learning_rate": 2.8713670387189093e-06,
+      "loss": 0.3675,
+      "step": 3093
+    },
+    {
+      "epoch": 1.4404096834264433,
+      "grad_norm": 0.6365492209651624,
+      "learning_rate": 2.870098104929858e-06,
+      "loss": 0.3865,
+      "step": 3094
+    },
+    {
+      "epoch": 1.440875232774674,
+      "grad_norm": 0.6459004224371382,
+      "learning_rate": 2.8688290736481656e-06,
+      "loss": 0.3831,
+      "step": 3095
+    },
+    {
+      "epoch": 1.441340782122905,
+      "grad_norm": 0.6032818882845395,
+      "learning_rate": 2.8675599452081233e-06,
+      "loss": 0.3878,
+      "step": 3096
+    },
+    {
+      "epoch": 1.441806331471136,
+      "grad_norm": 0.5792795903604387,
+      "learning_rate": 2.866290719944052e-06,
+      "loss": 0.3796,
+      "step": 3097
+    },
+    {
+      "epoch": 1.4422718808193669,
+      "grad_norm": 0.6048330724068606,
+      "learning_rate": 2.865021398190295e-06,
+      "loss": 0.3916,
+      "step": 3098
+    },
+    {
+      "epoch": 1.4427374301675977,
+      "grad_norm": 0.60727979104018,
+      "learning_rate": 2.8637519802812203e-06,
+      "loss": 0.4101,
+      "step": 3099
+    },
+    {
+      "epoch": 1.4432029795158288,
+      "grad_norm": 0.577746801377434,
+      "learning_rate": 2.862482466551225e-06,
+      "loss": 0.3864,
+      "step": 3100
+    },
+    {
+      "epoch": 1.4436685288640596,
+      "grad_norm": 0.6259427754423115,
+      "learning_rate": 2.861212857334728e-06,
+      "loss": 0.3923,
+      "step": 3101
+    },
+    {
+      "epoch": 1.4441340782122905,
+      "grad_norm": 0.6046051226076687,
+      "learning_rate": 2.8599431529661755e-06,
+      "loss": 0.3988,
+      "step": 3102
+    },
+    {
+      "epoch": 1.4445996275605215,
+      "grad_norm": 0.6301718375015092,
+      "learning_rate": 2.858673353780036e-06,
+      "loss": 0.3884,
+      "step": 3103
+    },
+    {
+      "epoch": 1.4450651769087524,
+      "grad_norm": 0.6228961566043505,
+      "learning_rate": 2.8574034601108074e-06,
+      "loss": 0.387,
+      "step": 3104
+    },
+    {
+      "epoch": 1.4455307262569832,
+      "grad_norm": 0.591237104542675,
+      "learning_rate": 2.8561334722930073e-06,
+      "loss": 0.3823,
+      "step": 3105
+    },
+    {
+      "epoch": 1.4459962756052143,
+      "grad_norm": 0.5974153778488044,
+      "learning_rate": 2.854863390661183e-06,
+      "loss": 0.3768,
+      "step": 3106
+    },
+    {
+      "epoch": 1.446461824953445,
+      "grad_norm": 0.6092815257686041,
+      "learning_rate": 2.8535932155499024e-06,
+      "loss": 0.3802,
+      "step": 3107
+    },
+    {
+      "epoch": 1.446927374301676,
+      "grad_norm": 0.6252007918005614,
+      "learning_rate": 2.85232294729376e-06,
+      "loss": 0.3851,
+      "step": 3108
+    },
+    {
+      "epoch": 1.447392923649907,
+      "grad_norm": 0.6106631319659247,
+      "learning_rate": 2.8510525862273753e-06,
+      "loss": 0.3836,
+      "step": 3109
+    },
+    {
+      "epoch": 1.4478584729981379,
+      "grad_norm": 0.6438255848309441,
+      "learning_rate": 2.8497821326853917e-06,
+      "loss": 0.4063,
+      "step": 3110
+    },
+    {
+      "epoch": 1.4483240223463687,
+      "grad_norm": 0.6059890528467136,
+      "learning_rate": 2.8485115870024773e-06,
+      "loss": 0.3778,
+      "step": 3111
+    },
+    {
+      "epoch": 1.4487895716945998,
+      "grad_norm": 0.603087122445023,
+      "learning_rate": 2.847240949513323e-06,
+      "loss": 0.3864,
+      "step": 3112
+    },
+    {
+      "epoch": 1.4492551210428306,
+      "grad_norm": 0.6060619014741072,
+      "learning_rate": 2.8459702205526444e-06,
+      "loss": 0.4008,
+      "step": 3113
+    },
+    {
+      "epoch": 1.4497206703910615,
+      "grad_norm": 0.5880178821507493,
+      "learning_rate": 2.844699400455184e-06,
+      "loss": 0.3908,
+      "step": 3114
+    },
+    {
+      "epoch": 1.4501862197392923,
+      "grad_norm": 0.6069701158103327,
+      "learning_rate": 2.843428489555704e-06,
+      "loss": 0.3934,
+      "step": 3115
+    },
+    {
+      "epoch": 1.4506517690875234,
+      "grad_norm": 0.6416856503488512,
+      "learning_rate": 2.8421574881889934e-06,
+      "loss": 0.3911,
+      "step": 3116
+    },
+    {
+      "epoch": 1.4511173184357542,
+      "grad_norm": 0.6134141206694268,
+      "learning_rate": 2.840886396689864e-06,
+      "loss": 0.3858,
+      "step": 3117
+    },
+    {
+      "epoch": 1.451582867783985,
+      "grad_norm": 0.5992635834577625,
+      "learning_rate": 2.8396152153931513e-06,
+      "loss": 0.3784,
+      "step": 3118
+    },
+    {
+      "epoch": 1.4520484171322159,
+      "grad_norm": 0.5817059898686173,
+      "learning_rate": 2.8383439446337146e-06,
+      "loss": 0.3672,
+      "step": 3119
+    },
+    {
+      "epoch": 1.452513966480447,
+      "grad_norm": 0.5820660350458602,
+      "learning_rate": 2.837072584746438e-06,
+      "loss": 0.3704,
+      "step": 3120
+    },
+    {
+      "epoch": 1.4529795158286778,
+      "grad_norm": 0.6104736263755396,
+      "learning_rate": 2.835801136066227e-06,
+      "loss": 0.401,
+      "step": 3121
+    },
+    {
+      "epoch": 1.4534450651769086,
+      "grad_norm": 0.6263329326370641,
+      "learning_rate": 2.8345295989280107e-06,
+      "loss": 0.3737,
+      "step": 3122
+    },
+    {
+      "epoch": 1.4539106145251397,
+      "grad_norm": 0.6060334320317553,
+      "learning_rate": 2.8332579736667436e-06,
+      "loss": 0.4011,
+      "step": 3123
+    },
+    {
+      "epoch": 1.4543761638733705,
+      "grad_norm": 0.5674660091866903,
+      "learning_rate": 2.831986260617402e-06,
+      "loss": 0.3596,
+      "step": 3124
+    },
+    {
+      "epoch": 1.4548417132216014,
+      "grad_norm": 0.6423333925486415,
+      "learning_rate": 2.8307144601149837e-06,
+      "loss": 0.3635,
+      "step": 3125
+    },
+    {
+      "epoch": 1.4553072625698324,
+      "grad_norm": 0.6308380767184012,
+      "learning_rate": 2.829442572494513e-06,
+      "loss": 0.3895,
+      "step": 3126
+    },
+    {
+      "epoch": 1.4557728119180633,
+      "grad_norm": 0.5800486520969838,
+      "learning_rate": 2.8281705980910346e-06,
+      "loss": 0.3769,
+      "step": 3127
+    },
+    {
+      "epoch": 1.4562383612662941,
+      "grad_norm": 0.6243486655924543,
+      "learning_rate": 2.8268985372396172e-06,
+      "loss": 0.4018,
+      "step": 3128
+    },
+    {
+      "epoch": 1.4567039106145252,
+      "grad_norm": 0.5686002437541676,
+      "learning_rate": 2.8256263902753504e-06,
+      "loss": 0.3971,
+      "step": 3129
+    },
+    {
+      "epoch": 1.457169459962756,
+      "grad_norm": 0.5981904445986358,
+      "learning_rate": 2.8243541575333504e-06,
+      "loss": 0.3936,
+      "step": 3130
+    },
+    {
+      "epoch": 1.4576350093109869,
+      "grad_norm": 0.624620087784625,
+      "learning_rate": 2.8230818393487514e-06,
+      "loss": 0.3837,
+      "step": 3131
+    },
+    {
+      "epoch": 1.458100558659218,
+      "grad_norm": 0.6541262493140401,
+      "learning_rate": 2.8218094360567137e-06,
+      "loss": 0.3869,
+      "step": 3132
+    },
+    {
+      "epoch": 1.4585661080074488,
+      "grad_norm": 0.6346488238670464,
+      "learning_rate": 2.820536947992418e-06,
+      "loss": 0.4007,
+      "step": 3133
+    },
+    {
+      "epoch": 1.4590316573556796,
+      "grad_norm": 0.5976747149536233,
+      "learning_rate": 2.8192643754910677e-06,
+      "loss": 0.409,
+      "step": 3134
+    },
+    {
+      "epoch": 1.4594972067039107,
+      "grad_norm": 0.5754647460234958,
+      "learning_rate": 2.8179917188878885e-06,
+      "loss": 0.3713,
+      "step": 3135
+    },
+    {
+      "epoch": 1.4599627560521415,
+      "grad_norm": 0.5803176846663001,
+      "learning_rate": 2.8167189785181293e-06,
+      "loss": 0.3887,
+      "step": 3136
+    },
+    {
+      "epoch": 1.4604283054003724,
+      "grad_norm": 0.6041644846942446,
+      "learning_rate": 2.8154461547170597e-06,
+      "loss": 0.4045,
+      "step": 3137
+    },
+    {
+      "epoch": 1.4608938547486034,
+      "grad_norm": 0.588853352376265,
+      "learning_rate": 2.8141732478199717e-06,
+      "loss": 0.3907,
+      "step": 3138
+    },
+    {
+      "epoch": 1.4613594040968343,
+      "grad_norm": 0.5848891152880928,
+      "learning_rate": 2.812900258162179e-06,
+      "loss": 0.383,
+      "step": 3139
+    },
+    {
+      "epoch": 1.4618249534450651,
+      "grad_norm": 0.6230832927425018,
+      "learning_rate": 2.8116271860790185e-06,
+      "loss": 0.3717,
+      "step": 3140
+    },
+    {
+      "epoch": 1.4622905027932962,
+      "grad_norm": 0.5964211094275874,
+      "learning_rate": 2.8103540319058454e-06,
+      "loss": 0.3663,
+      "step": 3141
+    },
+    {
+      "epoch": 1.462756052141527,
+      "grad_norm": 0.59196983917654,
+      "learning_rate": 2.8090807959780404e-06,
+      "loss": 0.3919,
+      "step": 3142
+    },
+    {
+      "epoch": 1.4632216014897579,
+      "grad_norm": 0.5763069021521199,
+      "learning_rate": 2.807807478631005e-06,
+      "loss": 0.375,
+      "step": 3143
+    },
+    {
+      "epoch": 1.463687150837989,
+      "grad_norm": 0.6218354544441625,
+      "learning_rate": 2.8065340802001595e-06,
+      "loss": 0.3741,
+      "step": 3144
+    },
+    {
+      "epoch": 1.4641527001862198,
+      "grad_norm": 0.5828810613000497,
+      "learning_rate": 2.8052606010209477e-06,
+      "loss": 0.3897,
+      "step": 3145
+    },
+    {
+      "epoch": 1.4646182495344506,
+      "grad_norm": 0.6153774867997187,
+      "learning_rate": 2.803987041428835e-06,
+      "loss": 0.3965,
+      "step": 3146
+    },
+    {
+      "epoch": 1.4650837988826817,
+      "grad_norm": 0.6009920525480784,
+      "learning_rate": 2.802713401759307e-06,
+      "loss": 0.3894,
+      "step": 3147
+    },
+    {
+      "epoch": 1.4655493482309125,
+      "grad_norm": 0.6002218383824046,
+      "learning_rate": 2.8014396823478702e-06,
+      "loss": 0.394,
+      "step": 3148
+    },
+    {
+      "epoch": 1.4660148975791434,
+      "grad_norm": 0.6053457819988954,
+      "learning_rate": 2.8001658835300532e-06,
+      "loss": 0.3814,
+      "step": 3149
+    },
+    {
+      "epoch": 1.4664804469273742,
+      "grad_norm": 0.5861361643991172,
+      "learning_rate": 2.7988920056414044e-06,
+      "loss": 0.3721,
+      "step": 3150
+    },
+    {
+      "epoch": 1.4669459962756053,
+      "grad_norm": 0.6036691986641777,
+      "learning_rate": 2.797618049017494e-06,
+      "loss": 0.3871,
+      "step": 3151
+    },
+    {
+      "epoch": 1.4674115456238361,
+      "grad_norm": 0.6379919357864589,
+      "learning_rate": 2.796344013993912e-06,
+      "loss": 0.3769,
+      "step": 3152
+    },
+    {
+      "epoch": 1.467877094972067,
+      "grad_norm": 0.6302335386727993,
+      "learning_rate": 2.79506990090627e-06,
+      "loss": 0.391,
+      "step": 3153
+    },
+    {
+      "epoch": 1.4683426443202978,
+      "grad_norm": 0.5949838883039362,
+      "learning_rate": 2.793795710090199e-06,
+      "loss": 0.387,
+      "step": 3154
+    },
+    {
+      "epoch": 1.4688081936685289,
+      "grad_norm": 0.5884642196453863,
+      "learning_rate": 2.792521441881352e-06,
+      "loss": 0.4003,
+      "step": 3155
+    },
+    {
+      "epoch": 1.4692737430167597,
+      "grad_norm": 0.6023203227692469,
+      "learning_rate": 2.791247096615401e-06,
+      "loss": 0.3787,
+      "step": 3156
+    },
+    {
+      "epoch": 1.4697392923649906,
+      "grad_norm": 0.6018834582943798,
+      "learning_rate": 2.789972674628039e-06,
+      "loss": 0.4135,
+      "step": 3157
+    },
+    {
+      "epoch": 1.4702048417132216,
+      "grad_norm": 0.648314665367624,
+      "learning_rate": 2.7886981762549788e-06,
+      "loss": 0.3841,
+      "step": 3158
+    },
+    {
+      "epoch": 1.4706703910614525,
+      "grad_norm": 0.5735917983089897,
+      "learning_rate": 2.787423601831954e-06,
+      "loss": 0.3979,
+      "step": 3159
+    },
+    {
+      "epoch": 1.4711359404096833,
+      "grad_norm": 0.6032828944827814,
+      "learning_rate": 2.7861489516947175e-06,
+      "loss": 0.3743,
+      "step": 3160
+    },
+    {
+      "epoch": 1.4716014897579144,
+      "grad_norm": 0.6427705184691626,
+      "learning_rate": 2.784874226179043e-06,
+      "loss": 0.3958,
+      "step": 3161
+    },
+    {
+      "epoch": 1.4720670391061452,
+      "grad_norm": 0.618184574457425,
+      "learning_rate": 2.7835994256207225e-06,
+      "loss": 0.3902,
+      "step": 3162
+    },
+    {
+      "epoch": 1.472532588454376,
+      "grad_norm": 0.5801153328855957,
+      "learning_rate": 2.782324550355571e-06,
+      "loss": 0.3946,
+      "step": 3163
+    },
+    {
+      "epoch": 1.4729981378026071,
+      "grad_norm": 0.5998306788342941,
+      "learning_rate": 2.781049600719417e-06,
+      "loss": 0.387,
+      "step": 3164
+    },
+    {
+      "epoch": 1.473463687150838,
+      "grad_norm": 0.577026886233205,
+      "learning_rate": 2.779774577048116e-06,
+      "loss": 0.399,
+      "step": 3165
+    },
+    {
+      "epoch": 1.4739292364990688,
+      "grad_norm": 0.5855247659764102,
+      "learning_rate": 2.7784994796775388e-06,
+      "loss": 0.3956,
+      "step": 3166
+    },
+    {
+      "epoch": 1.4743947858472999,
+      "grad_norm": 0.6367503220174511,
+      "learning_rate": 2.7772243089435756e-06,
+      "loss": 0.3812,
+      "step": 3167
+    },
+    {
+      "epoch": 1.4748603351955307,
+      "grad_norm": 0.6334878693029589,
+      "learning_rate": 2.775949065182137e-06,
+      "loss": 0.3909,
+      "step": 3168
+    },
+    {
+      "epoch": 1.4753258845437616,
+      "grad_norm": 0.5950256089562118,
+      "learning_rate": 2.7746737487291534e-06,
+      "loss": 0.3992,
+      "step": 3169
+    },
+    {
+      "epoch": 1.4757914338919926,
+      "grad_norm": 0.5755793940259635,
+      "learning_rate": 2.7733983599205716e-06,
+      "loss": 0.3894,
+      "step": 3170
+    },
+    {
+      "epoch": 1.4762569832402235,
+      "grad_norm": 0.6146226223733681,
+      "learning_rate": 2.7721228990923616e-06,
+      "loss": 0.3979,
+      "step": 3171
+    },
+    {
+      "epoch": 1.4767225325884543,
+      "grad_norm": 0.609801429276278,
+      "learning_rate": 2.770847366580508e-06,
+      "loss": 0.4,
+      "step": 3172
+    },
+    {
+      "epoch": 1.4771880819366854,
+      "grad_norm": 0.5916033546154249,
+      "learning_rate": 2.7695717627210186e-06,
+      "loss": 0.3778,
+      "step": 3173
+    },
+    {
+      "epoch": 1.4776536312849162,
+      "grad_norm": 0.587455764753547,
+      "learning_rate": 2.7682960878499147e-06,
+      "loss": 0.3778,
+      "step": 3174
+    },
+    {
+      "epoch": 1.478119180633147,
+      "grad_norm": 0.5830765126567564,
+      "learning_rate": 2.7670203423032427e-06,
+      "loss": 0.3917,
+      "step": 3175
+    },
+    {
+      "epoch": 1.4785847299813781,
+      "grad_norm": 0.592742325483436,
+      "learning_rate": 2.7657445264170625e-06,
+      "loss": 0.4059,
+      "step": 3176
+    },
+    {
+      "epoch": 1.479050279329609,
+      "grad_norm": 0.5837535895015101,
+      "learning_rate": 2.764468640527455e-06,
+      "loss": 0.3808,
+      "step": 3177
+    },
+    {
+      "epoch": 1.4795158286778398,
+      "grad_norm": 0.6053632289623074,
+      "learning_rate": 2.7631926849705174e-06,
+      "loss": 0.3861,
+      "step": 3178
+    },
+    {
+      "epoch": 1.4799813780260709,
+      "grad_norm": 0.6235355255322398,
+      "learning_rate": 2.7619166600823698e-06,
+      "loss": 0.3764,
+      "step": 3179
+    },
+    {
+      "epoch": 1.4804469273743017,
+      "grad_norm": 0.6119781799308613,
+      "learning_rate": 2.760640566199144e-06,
+      "loss": 0.4066,
+      "step": 3180
+    },
+    {
+      "epoch": 1.4809124767225326,
+      "grad_norm": 0.6211230263401921,
+      "learning_rate": 2.759364403656996e-06,
+      "loss": 0.3841,
+      "step": 3181
+    },
+    {
+      "epoch": 1.4813780260707636,
+      "grad_norm": 0.6103833267841344,
+      "learning_rate": 2.7580881727920967e-06,
+      "loss": 0.4081,
+      "step": 3182
+    },
+    {
+      "epoch": 1.4818435754189945,
+      "grad_norm": 0.5916065383173005,
+      "learning_rate": 2.7568118739406354e-06,
+      "loss": 0.3899,
+      "step": 3183
+    },
+    {
+      "epoch": 1.4823091247672253,
+      "grad_norm": 0.5865557580433853,
+      "learning_rate": 2.7555355074388184e-06,
+      "loss": 0.3891,
+      "step": 3184
+    },
+    {
+      "epoch": 1.4827746741154562,
+      "grad_norm": 0.5999998020774977,
+      "learning_rate": 2.754259073622873e-06,
+      "loss": 0.387,
+      "step": 3185
+    },
+    {
+      "epoch": 1.4832402234636872,
+      "grad_norm": 0.5620757186865595,
+      "learning_rate": 2.7529825728290416e-06,
+      "loss": 0.38,
+      "step": 3186
+    },
+    {
+      "epoch": 1.483705772811918,
+      "grad_norm": 0.6393265298569518,
+      "learning_rate": 2.751706005393584e-06,
+      "loss": 0.4026,
+      "step": 3187
+    },
+    {
+      "epoch": 1.484171322160149,
+      "grad_norm": 0.638249302625534,
+      "learning_rate": 2.75042937165278e-06,
+      "loss": 0.4003,
+      "step": 3188
+    },
+    {
+      "epoch": 1.4846368715083798,
+      "grad_norm": 0.59743402897093,
+      "learning_rate": 2.749152671942924e-06,
+      "loss": 0.385,
+      "step": 3189
+    },
+    {
+      "epoch": 1.4851024208566108,
+      "grad_norm": 0.5811459462911565,
+      "learning_rate": 2.747875906600329e-06,
+      "loss": 0.3859,
+      "step": 3190
+    },
+    {
+      "epoch": 1.4855679702048417,
+      "grad_norm": 0.6005043017297664,
+      "learning_rate": 2.746599075961326e-06,
+      "loss": 0.3919,
+      "step": 3191
+    },
+    {
+      "epoch": 1.4860335195530725,
+      "grad_norm": 0.6043077602567216,
+      "learning_rate": 2.745322180362263e-06,
+      "loss": 0.3771,
+      "step": 3192
+    },
+    {
+      "epoch": 1.4864990689013036,
+      "grad_norm": 0.5953089099713708,
+      "learning_rate": 2.7440452201395034e-06,
+      "loss": 0.4076,
+      "step": 3193
+    },
+    {
+      "epoch": 1.4869646182495344,
+      "grad_norm": 0.5951887848357723,
+      "learning_rate": 2.7427681956294295e-06,
+      "loss": 0.397,
+      "step": 3194
+    },
+    {
+      "epoch": 1.4874301675977653,
+      "grad_norm": 0.6000682721900397,
+      "learning_rate": 2.74149110716844e-06,
+      "loss": 0.3746,
+      "step": 3195
+    },
+    {
+      "epoch": 1.4878957169459963,
+      "grad_norm": 0.635998102524262,
+      "learning_rate": 2.740213955092951e-06,
+      "loss": 0.3899,
+      "step": 3196
+    },
+    {
+      "epoch": 1.4883612662942272,
+      "grad_norm": 0.6031952370536073,
+      "learning_rate": 2.7389367397393925e-06,
+      "loss": 0.382,
+      "step": 3197
+    },
+    {
+      "epoch": 1.488826815642458,
+      "grad_norm": 0.5697584261948143,
+      "learning_rate": 2.7376594614442157e-06,
+      "loss": 0.379,
+      "step": 3198
+    },
+    {
+      "epoch": 1.489292364990689,
+      "grad_norm": 0.5518914949635604,
+      "learning_rate": 2.7363821205438855e-06,
+      "loss": 0.3646,
+      "step": 3199
+    },
+    {
+      "epoch": 1.48975791433892,
+      "grad_norm": 0.5844045593837968,
+      "learning_rate": 2.735104717374884e-06,
+      "loss": 0.4017,
+      "step": 3200
+    },
+    {
+      "epoch": 1.4902234636871508,
+      "grad_norm": 0.5651695575310406,
+      "learning_rate": 2.733827252273708e-06,
+      "loss": 0.3665,
+      "step": 3201
+    },
+    {
+      "epoch": 1.4906890130353818,
+      "grad_norm": 0.5753952793482473,
+      "learning_rate": 2.732549725576875e-06,
+      "loss": 0.3901,
+      "step": 3202
+    },
+    {
+      "epoch": 1.4911545623836127,
+      "grad_norm": 0.6110439103851705,
+      "learning_rate": 2.731272137620913e-06,
+      "loss": 0.3737,
+      "step": 3203
+    },
+    {
+      "epoch": 1.4916201117318435,
+      "grad_norm": 0.5772329117474351,
+      "learning_rate": 2.7299944887423706e-06,
+      "loss": 0.3694,
+      "step": 3204
+    },
+    {
+      "epoch": 1.4920856610800746,
+      "grad_norm": 0.5580980273188293,
+      "learning_rate": 2.7287167792778114e-06,
+      "loss": 0.3925,
+      "step": 3205
+    },
+    {
+      "epoch": 1.4925512104283054,
+      "grad_norm": 0.5754482643793617,
+      "learning_rate": 2.727439009563813e-06,
+      "loss": 0.398,
+      "step": 3206
+    },
+    {
+      "epoch": 1.4930167597765363,
+      "grad_norm": 0.6219610628333831,
+      "learning_rate": 2.7261611799369706e-06,
+      "loss": 0.398,
+      "step": 3207
+    },
+    {
+      "epoch": 1.4934823091247673,
+      "grad_norm": 0.568795295912185,
+      "learning_rate": 2.7248832907338974e-06,
+      "loss": 0.3959,
+      "step": 3208
+    },
+    {
+      "epoch": 1.4939478584729982,
+      "grad_norm": 0.5973698829076461,
+      "learning_rate": 2.7236053422912156e-06,
+      "loss": 0.3851,
+      "step": 3209
+    },
+    {
+      "epoch": 1.494413407821229,
+      "grad_norm": 0.608342986640216,
+      "learning_rate": 2.7223273349455704e-06,
+      "loss": 0.384,
+      "step": 3210
+    },
+    {
+      "epoch": 1.49487895716946,
+      "grad_norm": 0.6096801525591259,
+      "learning_rate": 2.721049269033618e-06,
+      "loss": 0.3943,
+      "step": 3211
+    },
+    {
+      "epoch": 1.495344506517691,
+      "grad_norm": 0.5869808062509579,
+      "learning_rate": 2.7197711448920317e-06,
+      "loss": 0.3775,
+      "step": 3212
+    },
+    {
+      "epoch": 1.4958100558659218,
+      "grad_norm": 0.6007463729720943,
+      "learning_rate": 2.7184929628574997e-06,
+      "loss": 0.4012,
+      "step": 3213
+    },
+    {
+      "epoch": 1.4962756052141528,
+      "grad_norm": 0.5983857554964923,
+      "learning_rate": 2.7172147232667257e-06,
+      "loss": 0.3775,
+      "step": 3214
+    },
+    {
+      "epoch": 1.4967411545623837,
+      "grad_norm": 0.5826723307061211,
+      "learning_rate": 2.7159364264564283e-06,
+      "loss": 0.3966,
+      "step": 3215
+    },
+    {
+      "epoch": 1.4972067039106145,
+      "grad_norm": 0.5857358464150311,
+      "learning_rate": 2.7146580727633414e-06,
+      "loss": 0.3899,
+      "step": 3216
+    },
+    {
+      "epoch": 1.4976722532588456,
+      "grad_norm": 0.5862516122326722,
+      "learning_rate": 2.7133796625242133e-06,
+      "loss": 0.3983,
+      "step": 3217
+    },
+    {
+      "epoch": 1.4981378026070764,
+      "grad_norm": 0.5792995391324233,
+      "learning_rate": 2.7121011960758085e-06,
+      "loss": 0.3838,
+      "step": 3218
+    },
+    {
+      "epoch": 1.4986033519553073,
+      "grad_norm": 0.5899081782274064,
+      "learning_rate": 2.710822673754904e-06,
+      "loss": 0.3901,
+      "step": 3219
+    },
+    {
+      "epoch": 1.499068901303538,
+      "grad_norm": 0.5680053116811028,
+      "learning_rate": 2.709544095898295e-06,
+      "loss": 0.3767,
+      "step": 3220
+    },
+    {
+      "epoch": 1.4995344506517692,
+      "grad_norm": 0.5996649006724615,
+      "learning_rate": 2.708265462842788e-06,
+      "loss": 0.4011,
+      "step": 3221
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6320197882709836,
+      "learning_rate": 2.706986774925205e-06,
+      "loss": 0.3959,
+      "step": 3222
+    },
+    {
+      "epoch": 1.500465549348231,
+      "grad_norm": 0.5958210122947873,
+      "learning_rate": 2.7057080324823837e-06,
+      "loss": 0.3835,
+      "step": 3223
+    },
+    {
+      "epoch": 1.5009310986964617,
+      "grad_norm": 0.6296680769026699,
+      "learning_rate": 2.7044292358511747e-06,
+      "loss": 0.372,
+      "step": 3224
+    },
+    {
+      "epoch": 1.5013966480446927,
+      "grad_norm": 0.5998504369705278,
+      "learning_rate": 2.7031503853684442e-06,
+      "loss": 0.3935,
+      "step": 3225
+    },
+    {
+      "epoch": 1.5018621973929238,
+      "grad_norm": 0.6187678220810785,
+      "learning_rate": 2.7018714813710717e-06,
+      "loss": 0.396,
+      "step": 3226
+    },
+    {
+      "epoch": 1.5023277467411544,
+      "grad_norm": 0.6123553904626551,
+      "learning_rate": 2.7005925241959497e-06,
+      "loss": 0.3834,
+      "step": 3227
+    },
+    {
+      "epoch": 1.5027932960893855,
+      "grad_norm": 0.603376504765102,
+      "learning_rate": 2.6993135141799876e-06,
+      "loss": 0.382,
+      "step": 3228
+    },
+    {
+      "epoch": 1.5032588454376163,
+      "grad_norm": 0.5910869692981449,
+      "learning_rate": 2.6980344516601055e-06,
+      "loss": 0.3853,
+      "step": 3229
+    },
+    {
+      "epoch": 1.5037243947858472,
+      "grad_norm": 0.6316304656761642,
+      "learning_rate": 2.6967553369732395e-06,
+      "loss": 0.4108,
+      "step": 3230
+    },
+    {
+      "epoch": 1.5041899441340782,
+      "grad_norm": 0.6164471800486808,
+      "learning_rate": 2.69547617045634e-06,
+      "loss": 0.3896,
+      "step": 3231
+    },
+    {
+      "epoch": 1.504655493482309,
+      "grad_norm": 0.6185783435952891,
+      "learning_rate": 2.694196952446368e-06,
+      "loss": 0.3811,
+      "step": 3232
+    },
+    {
+      "epoch": 1.50512104283054,
+      "grad_norm": 0.6465811700338265,
+      "learning_rate": 2.692917683280301e-06,
+      "loss": 0.3958,
+      "step": 3233
+    },
+    {
+      "epoch": 1.505586592178771,
+      "grad_norm": 0.6252961974043207,
+      "learning_rate": 2.691638363295129e-06,
+      "loss": 0.3926,
+      "step": 3234
+    },
+    {
+      "epoch": 1.5060521415270018,
+      "grad_norm": 0.6264515919095875,
+      "learning_rate": 2.6903589928278544e-06,
+      "loss": 0.3918,
+      "step": 3235
+    },
+    {
+      "epoch": 1.5065176908752327,
+      "grad_norm": 0.6065266839656921,
+      "learning_rate": 2.689079572215495e-06,
+      "loss": 0.4024,
+      "step": 3236
+    },
+    {
+      "epoch": 1.5069832402234637,
+      "grad_norm": 0.5914068484332357,
+      "learning_rate": 2.6878001017950795e-06,
+      "loss": 0.3888,
+      "step": 3237
+    },
+    {
+      "epoch": 1.5074487895716946,
+      "grad_norm": 0.583332303647008,
+      "learning_rate": 2.6865205819036516e-06,
+      "loss": 0.3915,
+      "step": 3238
+    },
+    {
+      "epoch": 1.5079143389199254,
+      "grad_norm": 0.6250129780981643,
+      "learning_rate": 2.6852410128782675e-06,
+      "loss": 0.4103,
+      "step": 3239
+    },
+    {
+      "epoch": 1.5083798882681565,
+      "grad_norm": 0.5540988846662058,
+      "learning_rate": 2.6839613950559947e-06,
+      "loss": 0.3755,
+      "step": 3240
+    },
+    {
+      "epoch": 1.5088454376163873,
+      "grad_norm": 0.5726726532404707,
+      "learning_rate": 2.682681728773917e-06,
+      "loss": 0.3905,
+      "step": 3241
+    },
+    {
+      "epoch": 1.5093109869646182,
+      "grad_norm": 0.6010063762325887,
+      "learning_rate": 2.6814020143691265e-06,
+      "loss": 0.3854,
+      "step": 3242
+    },
+    {
+      "epoch": 1.5097765363128492,
+      "grad_norm": 0.5736936594580525,
+      "learning_rate": 2.680122252178733e-06,
+      "loss": 0.3852,
+      "step": 3243
+    },
+    {
+      "epoch": 1.51024208566108,
+      "grad_norm": 0.5994618843313786,
+      "learning_rate": 2.6788424425398544e-06,
+      "loss": 0.3905,
+      "step": 3244
+    },
+    {
+      "epoch": 1.510707635009311,
+      "grad_norm": 0.629101580237994,
+      "learning_rate": 2.677562585789624e-06,
+      "loss": 0.3937,
+      "step": 3245
+    },
+    {
+      "epoch": 1.511173184357542,
+      "grad_norm": 0.6110852469687497,
+      "learning_rate": 2.6762826822651854e-06,
+      "loss": 0.3897,
+      "step": 3246
+    },
+    {
+      "epoch": 1.5116387337057728,
+      "grad_norm": 0.6284404725974208,
+      "learning_rate": 2.6750027323036977e-06,
+      "loss": 0.4057,
+      "step": 3247
+    },
+    {
+      "epoch": 1.5121042830540037,
+      "grad_norm": 0.6493926764372369,
+      "learning_rate": 2.6737227362423275e-06,
+      "loss": 0.391,
+      "step": 3248
+    },
+    {
+      "epoch": 1.5125698324022347,
+      "grad_norm": 0.5961133835447927,
+      "learning_rate": 2.6724426944182574e-06,
+      "loss": 0.3785,
+      "step": 3249
+    },
+    {
+      "epoch": 1.5130353817504656,
+      "grad_norm": 0.5787059563195643,
+      "learning_rate": 2.671162607168682e-06,
+      "loss": 0.3767,
+      "step": 3250
+    },
+    {
+      "epoch": 1.5135009310986964,
+      "grad_norm": 0.6037204003785424,
+      "learning_rate": 2.669882474830805e-06,
+      "loss": 0.3936,
+      "step": 3251
+    },
+    {
+      "epoch": 1.5139664804469275,
+      "grad_norm": 0.6087830465886284,
+      "learning_rate": 2.668602297741844e-06,
+      "loss": 0.3771,
+      "step": 3252
+    },
+    {
+      "epoch": 1.5144320297951583,
+      "grad_norm": 0.5900331519910005,
+      "learning_rate": 2.6673220762390288e-06,
+      "loss": 0.3846,
+      "step": 3253
+    },
+    {
+      "epoch": 1.5148975791433892,
+      "grad_norm": 0.6186215385408113,
+      "learning_rate": 2.6660418106596005e-06,
+      "loss": 0.3787,
+      "step": 3254
+    },
+    {
+      "epoch": 1.5153631284916202,
+      "grad_norm": 0.6047159640894036,
+      "learning_rate": 2.6647615013408108e-06,
+      "loss": 0.3812,
+      "step": 3255
+    },
+    {
+      "epoch": 1.5158286778398509,
+      "grad_norm": 0.6282644547763089,
+      "learning_rate": 2.6634811486199225e-06,
+      "loss": 0.387,
+      "step": 3256
+    },
+    {
+      "epoch": 1.516294227188082,
+      "grad_norm": 0.5967136685597547,
+      "learning_rate": 2.662200752834214e-06,
+      "loss": 0.3788,
+      "step": 3257
+    },
+    {
+      "epoch": 1.516759776536313,
+      "grad_norm": 0.5932188648671977,
+      "learning_rate": 2.660920314320968e-06,
+      "loss": 0.3808,
+      "step": 3258
+    },
+    {
+      "epoch": 1.5172253258845436,
+      "grad_norm": 0.6231033405993077,
+      "learning_rate": 2.6596398334174855e-06,
+      "loss": 0.3778,
+      "step": 3259
+    },
+    {
+      "epoch": 1.5176908752327747,
+      "grad_norm": 0.5989866921994427,
+      "learning_rate": 2.658359310461075e-06,
+      "loss": 0.3863,
+      "step": 3260
+    },
+    {
+      "epoch": 1.5181564245810057,
+      "grad_norm": 0.5826683932735297,
+      "learning_rate": 2.6570787457890563e-06,
+      "loss": 0.4021,
+      "step": 3261
+    },
+    {
+      "epoch": 1.5186219739292364,
+      "grad_norm": 0.6025644455219235,
+      "learning_rate": 2.65579813973876e-06,
+      "loss": 0.3725,
+      "step": 3262
+    },
+    {
+      "epoch": 1.5190875232774674,
+      "grad_norm": 0.6055814022064189,
+      "learning_rate": 2.6545174926475292e-06,
+      "loss": 0.3963,
+      "step": 3263
+    },
+    {
+      "epoch": 1.5195530726256983,
+      "grad_norm": 0.6043767834298154,
+      "learning_rate": 2.653236804852717e-06,
+      "loss": 0.3696,
+      "step": 3264
+    },
+    {
+      "epoch": 1.5200186219739291,
+      "grad_norm": 0.5682549401010817,
+      "learning_rate": 2.6519560766916864e-06,
+      "loss": 0.3711,
+      "step": 3265
+    },
+    {
+      "epoch": 1.5204841713221602,
+      "grad_norm": 0.5785844988553747,
+      "learning_rate": 2.6506753085018104e-06,
+      "loss": 0.3859,
+      "step": 3266
+    },
+    {
+      "epoch": 1.520949720670391,
+      "grad_norm": 0.5996899270773979,
+      "learning_rate": 2.6493945006204774e-06,
+      "loss": 0.3801,
+      "step": 3267
+    },
+    {
+      "epoch": 1.5214152700186219,
+      "grad_norm": 0.5724213989974685,
+      "learning_rate": 2.648113653385078e-06,
+      "loss": 0.3805,
+      "step": 3268
+    },
+    {
+      "epoch": 1.521880819366853,
+      "grad_norm": 0.5813393802744342,
+      "learning_rate": 2.646832767133022e-06,
+      "loss": 0.3578,
+      "step": 3269
+    },
+    {
+      "epoch": 1.5223463687150838,
+      "grad_norm": 0.6062139610588523,
+      "learning_rate": 2.6455518422017227e-06,
+      "loss": 0.3884,
+      "step": 3270
+    },
+    {
+      "epoch": 1.5228119180633146,
+      "grad_norm": 0.6579080795890827,
+      "learning_rate": 2.6442708789286076e-06,
+      "loss": 0.3928,
+      "step": 3271
+    },
+    {
+      "epoch": 1.5232774674115457,
+      "grad_norm": 0.5939001381871145,
+      "learning_rate": 2.642989877651111e-06,
+      "loss": 0.3734,
+      "step": 3272
+    },
+    {
+      "epoch": 1.5237430167597765,
+      "grad_norm": 0.5911599454175619,
+      "learning_rate": 2.6417088387066814e-06,
+      "loss": 0.406,
+      "step": 3273
+    },
+    {
+      "epoch": 1.5242085661080074,
+      "grad_norm": 0.6027762698350625,
+      "learning_rate": 2.640427762432774e-06,
+      "loss": 0.394,
+      "step": 3274
+    },
+    {
+      "epoch": 1.5246741154562384,
+      "grad_norm": 0.5817370797840627,
+      "learning_rate": 2.6391466491668536e-06,
+      "loss": 0.3847,
+      "step": 3275
+    },
+    {
+      "epoch": 1.5251396648044693,
+      "grad_norm": 0.5854382888317751,
+      "learning_rate": 2.637865499246398e-06,
+      "loss": 0.3918,
+      "step": 3276
+    },
+    {
+      "epoch": 1.5256052141527001,
+      "grad_norm": 0.6042861723788826,
+      "learning_rate": 2.6365843130088913e-06,
+      "loss": 0.4018,
+      "step": 3277
+    },
+    {
+      "epoch": 1.5260707635009312,
+      "grad_norm": 0.6129188857200822,
+      "learning_rate": 2.635303090791829e-06,
+      "loss": 0.3855,
+      "step": 3278
+    },
+    {
+      "epoch": 1.526536312849162,
+      "grad_norm": 0.592562703905814,
+      "learning_rate": 2.634021832932714e-06,
+      "loss": 0.3641,
+      "step": 3279
+    },
+    {
+      "epoch": 1.5270018621973929,
+      "grad_norm": 0.5636709626115174,
+      "learning_rate": 2.632740539769062e-06,
+      "loss": 0.3939,
+      "step": 3280
+    },
+    {
+      "epoch": 1.527467411545624,
+      "grad_norm": 0.6202742974447036,
+      "learning_rate": 2.631459211638395e-06,
+      "loss": 0.3845,
+      "step": 3281
+    },
+    {
+      "epoch": 1.5279329608938548,
+      "grad_norm": 0.5970908433521444,
+      "learning_rate": 2.630177848878245e-06,
+      "loss": 0.3922,
+      "step": 3282
+    },
+    {
+      "epoch": 1.5283985102420856,
+      "grad_norm": 0.5740423166335445,
+      "learning_rate": 2.6288964518261556e-06,
+      "loss": 0.369,
+      "step": 3283
+    },
+    {
+      "epoch": 1.5288640595903167,
+      "grad_norm": 0.5888219156955908,
+      "learning_rate": 2.627615020819675e-06,
+      "loss": 0.3737,
+      "step": 3284
+    },
+    {
+      "epoch": 1.5293296089385475,
+      "grad_norm": 0.5873484286456847,
+      "learning_rate": 2.626333556196362e-06,
+      "loss": 0.3872,
+      "step": 3285
+    },
+    {
+      "epoch": 1.5297951582867784,
+      "grad_norm": 0.5779078781044142,
+      "learning_rate": 2.6250520582937883e-06,
+      "loss": 0.3808,
+      "step": 3286
+    },
+    {
+      "epoch": 1.5302607076350094,
+      "grad_norm": 0.5799279961121901,
+      "learning_rate": 2.6237705274495273e-06,
+      "loss": 0.3731,
+      "step": 3287
+    },
+    {
+      "epoch": 1.5307262569832403,
+      "grad_norm": 0.6527669498342549,
+      "learning_rate": 2.6224889640011664e-06,
+      "loss": 0.3863,
+      "step": 3288
+    },
+    {
+      "epoch": 1.5311918063314711,
+      "grad_norm": 0.5972454389514757,
+      "learning_rate": 2.6212073682862997e-06,
+      "loss": 0.3776,
+      "step": 3289
+    },
+    {
+      "epoch": 1.5316573556797022,
+      "grad_norm": 0.6057653652956326,
+      "learning_rate": 2.61992574064253e-06,
+      "loss": 0.3775,
+      "step": 3290
+    },
+    {
+      "epoch": 1.5321229050279328,
+      "grad_norm": 0.5927839334735739,
+      "learning_rate": 2.618644081407468e-06,
+      "loss": 0.3893,
+      "step": 3291
+    },
+    {
+      "epoch": 1.5325884543761639,
+      "grad_norm": 0.5908888705043666,
+      "learning_rate": 2.617362390918734e-06,
+      "loss": 0.3716,
+      "step": 3292
+    },
+    {
+      "epoch": 1.533054003724395,
+      "grad_norm": 0.6030747215961937,
+      "learning_rate": 2.6160806695139558e-06,
+      "loss": 0.3952,
+      "step": 3293
+    },
+    {
+      "epoch": 1.5335195530726256,
+      "grad_norm": 0.6442085836465814,
+      "learning_rate": 2.6147989175307688e-06,
+      "loss": 0.3887,
+      "step": 3294
+    },
+    {
+      "epoch": 1.5339851024208566,
+      "grad_norm": 0.6212947608388515,
+      "learning_rate": 2.6135171353068157e-06,
+      "loss": 0.3906,
+      "step": 3295
+    },
+    {
+      "epoch": 1.5344506517690877,
+      "grad_norm": 0.6106357375679292,
+      "learning_rate": 2.6122353231797515e-06,
+      "loss": 0.374,
+      "step": 3296
+    },
+    {
+      "epoch": 1.5349162011173183,
+      "grad_norm": 0.6286447498004839,
+      "learning_rate": 2.6109534814872335e-06,
+      "loss": 0.3838,
+      "step": 3297
+    },
+    {
+      "epoch": 1.5353817504655494,
+      "grad_norm": 0.5926284665783996,
+      "learning_rate": 2.60967161056693e-06,
+      "loss": 0.3845,
+      "step": 3298
+    },
+    {
+      "epoch": 1.5358472998137802,
+      "grad_norm": 0.6127283402992972,
+      "learning_rate": 2.6083897107565165e-06,
+      "loss": 0.376,
+      "step": 3299
+    },
+    {
+      "epoch": 1.536312849162011,
+      "grad_norm": 0.6047587414646877,
+      "learning_rate": 2.607107782393675e-06,
+      "loss": 0.3753,
+      "step": 3300
+    },
+    {
+      "epoch": 1.5367783985102421,
+      "grad_norm": 0.6168290047799648,
+      "learning_rate": 2.6058258258160973e-06,
+      "loss": 0.3722,
+      "step": 3301
+    },
+    {
+      "epoch": 1.537243947858473,
+      "grad_norm": 0.5813128285828917,
+      "learning_rate": 2.6045438413614805e-06,
+      "loss": 0.3724,
+      "step": 3302
+    },
+    {
+      "epoch": 1.5377094972067038,
+      "grad_norm": 0.6237237715596297,
+      "learning_rate": 2.6032618293675298e-06,
+      "loss": 0.3803,
+      "step": 3303
+    },
+    {
+      "epoch": 1.5381750465549349,
+      "grad_norm": 0.5995867153202119,
+      "learning_rate": 2.601979790171958e-06,
+      "loss": 0.4088,
+      "step": 3304
+    },
+    {
+      "epoch": 1.5386405959031657,
+      "grad_norm": 0.5764055051559703,
+      "learning_rate": 2.6006977241124836e-06,
+      "loss": 0.3782,
+      "step": 3305
+    },
+    {
+      "epoch": 1.5391061452513966,
+      "grad_norm": 0.6002140188021787,
+      "learning_rate": 2.5994156315268355e-06,
+      "loss": 0.3823,
+      "step": 3306
+    },
+    {
+      "epoch": 1.5395716945996276,
+      "grad_norm": 0.6046366677856064,
+      "learning_rate": 2.598133512752745e-06,
+      "loss": 0.3818,
+      "step": 3307
+    },
+    {
+      "epoch": 1.5400372439478585,
+      "grad_norm": 0.6198535763266937,
+      "learning_rate": 2.596851368127954e-06,
+      "loss": 0.3641,
+      "step": 3308
+    },
+    {
+      "epoch": 1.5405027932960893,
+      "grad_norm": 0.6005824206468133,
+      "learning_rate": 2.595569197990209e-06,
+      "loss": 0.3762,
+      "step": 3309
+    },
+    {
+      "epoch": 1.5409683426443204,
+      "grad_norm": 0.6213426685446012,
+      "learning_rate": 2.5942870026772655e-06,
+      "loss": 0.3778,
+      "step": 3310
+    },
+    {
+      "epoch": 1.5414338919925512,
+      "grad_norm": 0.5776100034944857,
+      "learning_rate": 2.593004782526882e-06,
+      "loss": 0.3716,
+      "step": 3311
+    },
+    {
+      "epoch": 1.541899441340782,
+      "grad_norm": 0.597014789302986,
+      "learning_rate": 2.5917225378768286e-06,
+      "loss": 0.3892,
+      "step": 3312
+    },
+    {
+      "epoch": 1.5423649906890131,
+      "grad_norm": 0.5994317845474398,
+      "learning_rate": 2.5904402690648773e-06,
+      "loss": 0.4206,
+      "step": 3313
+    },
+    {
+      "epoch": 1.542830540037244,
+      "grad_norm": 0.6042019291633062,
+      "learning_rate": 2.5891579764288085e-06,
+      "loss": 0.3882,
+      "step": 3314
+    },
+    {
+      "epoch": 1.5432960893854748,
+      "grad_norm": 0.629799046784112,
+      "learning_rate": 2.5878756603064074e-06,
+      "loss": 0.3727,
+      "step": 3315
+    },
+    {
+      "epoch": 1.5437616387337059,
+      "grad_norm": 0.6357261632187915,
+      "learning_rate": 2.5865933210354692e-06,
+      "loss": 0.4002,
+      "step": 3316
+    },
+    {
+      "epoch": 1.5442271880819367,
+      "grad_norm": 0.6569656296481413,
+      "learning_rate": 2.5853109589537902e-06,
+      "loss": 0.3839,
+      "step": 3317
+    },
+    {
+      "epoch": 1.5446927374301676,
+      "grad_norm": 0.6077308007893025,
+      "learning_rate": 2.584028574399176e-06,
+      "loss": 0.3828,
+      "step": 3318
+    },
+    {
+      "epoch": 1.5451582867783986,
+      "grad_norm": 0.5944671212569109,
+      "learning_rate": 2.5827461677094377e-06,
+      "loss": 0.3646,
+      "step": 3319
+    },
+    {
+      "epoch": 1.5456238361266295,
+      "grad_norm": 0.6194077427134649,
+      "learning_rate": 2.5814637392223907e-06,
+      "loss": 0.3849,
+      "step": 3320
+    },
+    {
+      "epoch": 1.5460893854748603,
+      "grad_norm": 0.6082857459024981,
+      "learning_rate": 2.5801812892758575e-06,
+      "loss": 0.4122,
+      "step": 3321
+    },
+    {
+      "epoch": 1.5465549348230914,
+      "grad_norm": 0.6301373483198731,
+      "learning_rate": 2.5788988182076664e-06,
+      "loss": 0.3912,
+      "step": 3322
+    },
+    {
+      "epoch": 1.5470204841713222,
+      "grad_norm": 0.6371348626018006,
+      "learning_rate": 2.5776163263556503e-06,
+      "loss": 0.3842,
+      "step": 3323
+    },
+    {
+      "epoch": 1.547486033519553,
+      "grad_norm": 0.6572654481729093,
+      "learning_rate": 2.5763338140576476e-06,
+      "loss": 0.3894,
+      "step": 3324
+    },
+    {
+      "epoch": 1.5479515828677841,
+      "grad_norm": 0.5818711716706618,
+      "learning_rate": 2.5750512816515046e-06,
+      "loss": 0.4046,
+      "step": 3325
+    },
+    {
+      "epoch": 1.5484171322160147,
+      "grad_norm": 0.5776301538301815,
+      "learning_rate": 2.5737687294750675e-06,
+      "loss": 0.385,
+      "step": 3326
+    },
+    {
+      "epoch": 1.5488826815642458,
+      "grad_norm": 0.5674433342717917,
+      "learning_rate": 2.572486157866194e-06,
+      "loss": 0.3873,
+      "step": 3327
+    },
+    {
+      "epoch": 1.5493482309124769,
+      "grad_norm": 0.602603811588578,
+      "learning_rate": 2.571203567162742e-06,
+      "loss": 0.3812,
+      "step": 3328
+    },
+    {
+      "epoch": 1.5498137802607075,
+      "grad_norm": 0.586562626363641,
+      "learning_rate": 2.569920957702578e-06,
+      "loss": 0.377,
+      "step": 3329
+    },
+    {
+      "epoch": 1.5502793296089385,
+      "grad_norm": 0.6138601814147604,
+      "learning_rate": 2.568638329823571e-06,
+      "loss": 0.3976,
+      "step": 3330
+    },
+    {
+      "epoch": 1.5507448789571696,
+      "grad_norm": 0.6061298674201911,
+      "learning_rate": 2.567355683863595e-06,
+      "loss": 0.3836,
+      "step": 3331
+    },
+    {
+      "epoch": 1.5512104283054002,
+      "grad_norm": 0.6454159317682211,
+      "learning_rate": 2.5660730201605304e-06,
+      "loss": 0.3736,
+      "step": 3332
+    },
+    {
+      "epoch": 1.5516759776536313,
+      "grad_norm": 0.5956705633423122,
+      "learning_rate": 2.5647903390522617e-06,
+      "loss": 0.3823,
+      "step": 3333
+    },
+    {
+      "epoch": 1.5521415270018621,
+      "grad_norm": 0.6351917331435705,
+      "learning_rate": 2.5635076408766764e-06,
+      "loss": 0.3904,
+      "step": 3334
+    },
+    {
+      "epoch": 1.552607076350093,
+      "grad_norm": 0.5952061423824508,
+      "learning_rate": 2.562224925971669e-06,
+      "loss": 0.3905,
+      "step": 3335
+    },
+    {
+      "epoch": 1.553072625698324,
+      "grad_norm": 0.626545820280038,
+      "learning_rate": 2.5609421946751355e-06,
+      "loss": 0.3916,
+      "step": 3336
+    },
+    {
+      "epoch": 1.553538175046555,
+      "grad_norm": 0.6098190012007897,
+      "learning_rate": 2.5596594473249792e-06,
+      "loss": 0.4068,
+      "step": 3337
+    },
+    {
+      "epoch": 1.5540037243947857,
+      "grad_norm": 0.5604144382126361,
+      "learning_rate": 2.5583766842591057e-06,
+      "loss": 0.3694,
+      "step": 3338
+    },
+    {
+      "epoch": 1.5544692737430168,
+      "grad_norm": 0.5941295318127869,
+      "learning_rate": 2.5570939058154258e-06,
+      "loss": 0.3786,
+      "step": 3339
+    },
+    {
+      "epoch": 1.5549348230912476,
+      "grad_norm": 0.6123174414240019,
+      "learning_rate": 2.5558111123318525e-06,
+      "loss": 0.3863,
+      "step": 3340
+    },
+    {
+      "epoch": 1.5554003724394785,
+      "grad_norm": 0.6438944957460344,
+      "learning_rate": 2.554528304146306e-06,
+      "loss": 0.3822,
+      "step": 3341
+    },
+    {
+      "epoch": 1.5558659217877095,
+      "grad_norm": 0.5894055285602462,
+      "learning_rate": 2.553245481596707e-06,
+      "loss": 0.4036,
+      "step": 3342
+    },
+    {
+      "epoch": 1.5563314711359404,
+      "grad_norm": 0.6222529274299436,
+      "learning_rate": 2.551962645020982e-06,
+      "loss": 0.3794,
+      "step": 3343
+    },
+    {
+      "epoch": 1.5567970204841712,
+      "grad_norm": 0.5833369186787518,
+      "learning_rate": 2.5506797947570604e-06,
+      "loss": 0.3728,
+      "step": 3344
+    },
+    {
+      "epoch": 1.5572625698324023,
+      "grad_norm": 0.5834542646945718,
+      "learning_rate": 2.5493969311428768e-06,
+      "loss": 0.3857,
+      "step": 3345
+    },
+    {
+      "epoch": 1.5577281191806331,
+      "grad_norm": 0.6139246644528428,
+      "learning_rate": 2.548114054516366e-06,
+      "loss": 0.394,
+      "step": 3346
+    },
+    {
+      "epoch": 1.558193668528864,
+      "grad_norm": 0.6092689361840927,
+      "learning_rate": 2.5468311652154693e-06,
+      "loss": 0.3735,
+      "step": 3347
+    },
+    {
+      "epoch": 1.558659217877095,
+      "grad_norm": 0.634594162895049,
+      "learning_rate": 2.54554826357813e-06,
+      "loss": 0.3888,
+      "step": 3348
+    },
+    {
+      "epoch": 1.559124767225326,
+      "grad_norm": 0.5702684383274803,
+      "learning_rate": 2.544265349942295e-06,
+      "loss": 0.3749,
+      "step": 3349
+    },
+    {
+      "epoch": 1.5595903165735567,
+      "grad_norm": 0.5684601460379092,
+      "learning_rate": 2.542982424645914e-06,
+      "loss": 0.3955,
+      "step": 3350
+    },
+    {
+      "epoch": 1.5600558659217878,
+      "grad_norm": 0.6072246292943757,
+      "learning_rate": 2.5416994880269403e-06,
+      "loss": 0.3769,
+      "step": 3351
+    },
+    {
+      "epoch": 1.5605214152700186,
+      "grad_norm": 0.5945795153710323,
+      "learning_rate": 2.54041654042333e-06,
+      "loss": 0.3966,
+      "step": 3352
+    },
+    {
+      "epoch": 1.5609869646182495,
+      "grad_norm": 0.5785348981024613,
+      "learning_rate": 2.5391335821730417e-06,
+      "loss": 0.3925,
+      "step": 3353
+    },
+    {
+      "epoch": 1.5614525139664805,
+      "grad_norm": 0.5879952707660179,
+      "learning_rate": 2.5378506136140372e-06,
+      "loss": 0.372,
+      "step": 3354
+    },
+    {
+      "epoch": 1.5619180633147114,
+      "grad_norm": 0.5980483573435685,
+      "learning_rate": 2.5365676350842823e-06,
+      "loss": 0.3784,
+      "step": 3355
+    },
+    {
+      "epoch": 1.5623836126629422,
+      "grad_norm": 0.6003078408669492,
+      "learning_rate": 2.5352846469217405e-06,
+      "loss": 0.3921,
+      "step": 3356
+    },
+    {
+      "epoch": 1.5628491620111733,
+      "grad_norm": 0.6382357252983742,
+      "learning_rate": 2.5340016494643856e-06,
+      "loss": 0.4193,
+      "step": 3357
+    },
+    {
+      "epoch": 1.5633147113594041,
+      "grad_norm": 0.5703725660338612,
+      "learning_rate": 2.5327186430501866e-06,
+      "loss": 0.384,
+      "step": 3358
+    },
+    {
+      "epoch": 1.563780260707635,
+      "grad_norm": 0.5362145544730691,
+      "learning_rate": 2.5314356280171195e-06,
+      "loss": 0.392,
+      "step": 3359
+    },
+    {
+      "epoch": 1.564245810055866,
+      "grad_norm": 0.5905686495989368,
+      "learning_rate": 2.5301526047031593e-06,
+      "loss": 0.3658,
+      "step": 3360
+    },
+    {
+      "epoch": 1.5647113594040967,
+      "grad_norm": 0.550671552563723,
+      "learning_rate": 2.528869573446288e-06,
+      "loss": 0.3953,
+      "step": 3361
+    },
+    {
+      "epoch": 1.5651769087523277,
+      "grad_norm": 0.6483056065503304,
+      "learning_rate": 2.527586534584483e-06,
+      "loss": 0.3807,
+      "step": 3362
+    },
+    {
+      "epoch": 1.5656424581005588,
+      "grad_norm": 0.6038501507730132,
+      "learning_rate": 2.5263034884557285e-06,
+      "loss": 0.3722,
+      "step": 3363
+    },
+    {
+      "epoch": 1.5661080074487894,
+      "grad_norm": 0.6521248932830757,
+      "learning_rate": 2.525020435398011e-06,
+      "loss": 0.3827,
+      "step": 3364
+    },
+    {
+      "epoch": 1.5665735567970205,
+      "grad_norm": 0.5821023625686161,
+      "learning_rate": 2.523737375749315e-06,
+      "loss": 0.3589,
+      "step": 3365
+    },
+    {
+      "epoch": 1.5670391061452515,
+      "grad_norm": 0.6216610362528038,
+      "learning_rate": 2.5224543098476303e-06,
+      "loss": 0.3885,
+      "step": 3366
+    },
+    {
+      "epoch": 1.5675046554934822,
+      "grad_norm": 0.6028299550665558,
+      "learning_rate": 2.521171238030946e-06,
+      "loss": 0.3767,
+      "step": 3367
+    },
+    {
+      "epoch": 1.5679702048417132,
+      "grad_norm": 0.6111125507606293,
+      "learning_rate": 2.5198881606372543e-06,
+      "loss": 0.3903,
+      "step": 3368
+    },
+    {
+      "epoch": 1.5684357541899443,
+      "grad_norm": 0.6242101692366916,
+      "learning_rate": 2.518605078004548e-06,
+      "loss": 0.3753,
+      "step": 3369
+    },
+    {
+      "epoch": 1.568901303538175,
+      "grad_norm": 0.6033022564643482,
+      "learning_rate": 2.517321990470822e-06,
+      "loss": 0.3778,
+      "step": 3370
+    },
+    {
+      "epoch": 1.569366852886406,
+      "grad_norm": 0.6101301926340473,
+      "learning_rate": 2.5160388983740718e-06,
+      "loss": 0.3873,
+      "step": 3371
+    },
+    {
+      "epoch": 1.5698324022346368,
+      "grad_norm": 0.5907255800480928,
+      "learning_rate": 2.5147558020522946e-06,
+      "loss": 0.3768,
+      "step": 3372
+    },
+    {
+      "epoch": 1.5702979515828677,
+      "grad_norm": 0.5994792846030925,
+      "learning_rate": 2.513472701843487e-06,
+      "loss": 0.384,
+      "step": 3373
+    },
+    {
+      "epoch": 1.5707635009310987,
+      "grad_norm": 0.6012634884891042,
+      "learning_rate": 2.5121895980856507e-06,
+      "loss": 0.3708,
+      "step": 3374
+    },
+    {
+      "epoch": 1.5712290502793296,
+      "grad_norm": 0.5908070552046794,
+      "learning_rate": 2.5109064911167837e-06,
+      "loss": 0.3982,
+      "step": 3375
+    },
+    {
+      "epoch": 1.5716945996275604,
+      "grad_norm": 0.5851811111643864,
+      "learning_rate": 2.5096233812748877e-06,
+      "loss": 0.3564,
+      "step": 3376
+    },
+    {
+      "epoch": 1.5721601489757915,
+      "grad_norm": 0.6040525603182767,
+      "learning_rate": 2.5083402688979648e-06,
+      "loss": 0.369,
+      "step": 3377
+    },
+    {
+      "epoch": 1.5726256983240223,
+      "grad_norm": 0.6282042442800078,
+      "learning_rate": 2.5070571543240163e-06,
+      "loss": 0.3819,
+      "step": 3378
+    },
+    {
+      "epoch": 1.5730912476722532,
+      "grad_norm": 0.5918268267081838,
+      "learning_rate": 2.505774037891045e-06,
+      "loss": 0.3733,
+      "step": 3379
+    },
+    {
+      "epoch": 1.5735567970204842,
+      "grad_norm": 0.5894736700040035,
+      "learning_rate": 2.5044909199370553e-06,
+      "loss": 0.3888,
+      "step": 3380
+    },
+    {
+      "epoch": 1.574022346368715,
+      "grad_norm": 0.5814716751849753,
+      "learning_rate": 2.503207800800051e-06,
+      "loss": 0.3786,
+      "step": 3381
+    },
+    {
+      "epoch": 1.574487895716946,
+      "grad_norm": 0.5765604900925964,
+      "learning_rate": 2.5019246808180355e-06,
+      "loss": 0.3657,
+      "step": 3382
+    },
+    {
+      "epoch": 1.574953445065177,
+      "grad_norm": 0.615016726116646,
+      "learning_rate": 2.5006415603290124e-06,
+      "loss": 0.3976,
+      "step": 3383
+    },
+    {
+      "epoch": 1.5754189944134078,
+      "grad_norm": 0.6130376671023051,
+      "learning_rate": 2.4993584396709876e-06,
+      "loss": 0.3786,
+      "step": 3384
+    },
+    {
+      "epoch": 1.5758845437616387,
+      "grad_norm": 0.6081888524205403,
+      "learning_rate": 2.4980753191819653e-06,
+      "loss": 0.3916,
+      "step": 3385
+    },
+    {
+      "epoch": 1.5763500931098697,
+      "grad_norm": 0.6229356037368141,
+      "learning_rate": 2.49679219919995e-06,
+      "loss": 0.3944,
+      "step": 3386
+    },
+    {
+      "epoch": 1.5768156424581006,
+      "grad_norm": 0.5846184375855153,
+      "learning_rate": 2.495509080062945e-06,
+      "loss": 0.3955,
+      "step": 3387
+    },
+    {
+      "epoch": 1.5772811918063314,
+      "grad_norm": 0.5942468861484034,
+      "learning_rate": 2.4942259621089553e-06,
+      "loss": 0.3793,
+      "step": 3388
+    },
+    {
+      "epoch": 1.5777467411545625,
+      "grad_norm": 0.5892248879535344,
+      "learning_rate": 2.4929428456759846e-06,
+      "loss": 0.3998,
+      "step": 3389
+    },
+    {
+      "epoch": 1.5782122905027933,
+      "grad_norm": 0.6652790292246837,
+      "learning_rate": 2.491659731102036e-06,
+      "loss": 0.3928,
+      "step": 3390
+    },
+    {
+      "epoch": 1.5786778398510242,
+      "grad_norm": 0.5822899063781756,
+      "learning_rate": 2.490376618725113e-06,
+      "loss": 0.376,
+      "step": 3391
+    },
+    {
+      "epoch": 1.5791433891992552,
+      "grad_norm": 0.5861301050351946,
+      "learning_rate": 2.4890935088832163e-06,
+      "loss": 0.3948,
+      "step": 3392
+    },
+    {
+      "epoch": 1.579608938547486,
+      "grad_norm": 0.5999265973128673,
+      "learning_rate": 2.4878104019143497e-06,
+      "loss": 0.3681,
+      "step": 3393
+    },
+    {
+      "epoch": 1.580074487895717,
+      "grad_norm": 0.6151338553104807,
+      "learning_rate": 2.4865272981565134e-06,
+      "loss": 0.3631,
+      "step": 3394
+    },
+    {
+      "epoch": 1.580540037243948,
+      "grad_norm": 0.6559352462123799,
+      "learning_rate": 2.485244197947707e-06,
+      "loss": 0.382,
+      "step": 3395
+    },
+    {
+      "epoch": 1.5810055865921788,
+      "grad_norm": 0.6060349635528601,
+      "learning_rate": 2.4839611016259286e-06,
+      "loss": 0.3985,
+      "step": 3396
+    },
+    {
+      "epoch": 1.5814711359404097,
+      "grad_norm": 0.585662554697928,
+      "learning_rate": 2.482678009529179e-06,
+      "loss": 0.3944,
+      "step": 3397
+    },
+    {
+      "epoch": 1.5819366852886407,
+      "grad_norm": 0.6001547033381017,
+      "learning_rate": 2.4813949219954528e-06,
+      "loss": 0.3733,
+      "step": 3398
+    },
+    {
+      "epoch": 1.5824022346368714,
+      "grad_norm": 0.6207069336478601,
+      "learning_rate": 2.480111839362747e-06,
+      "loss": 0.3999,
+      "step": 3399
+    },
+    {
+      "epoch": 1.5828677839851024,
+      "grad_norm": 0.6207517644961045,
+      "learning_rate": 2.4788287619690547e-06,
+      "loss": 0.3852,
+      "step": 3400
+    },
+    {
+      "epoch": 1.5833333333333335,
+      "grad_norm": 0.5907142019889171,
+      "learning_rate": 2.4775456901523706e-06,
+      "loss": 0.363,
+      "step": 3401
+    },
+    {
+      "epoch": 1.583798882681564,
+      "grad_norm": 0.5950046334873272,
+      "learning_rate": 2.4762626242506856e-06,
+      "loss": 0.3914,
+      "step": 3402
+    },
+    {
+      "epoch": 1.5842644320297952,
+      "grad_norm": 0.6117523527627414,
+      "learning_rate": 2.47497956460199e-06,
+      "loss": 0.3951,
+      "step": 3403
+    },
+    {
+      "epoch": 1.5847299813780262,
+      "grad_norm": 0.587388891416984,
+      "learning_rate": 2.4736965115442715e-06,
+      "loss": 0.3708,
+      "step": 3404
+    },
+    {
+      "epoch": 1.5851955307262569,
+      "grad_norm": 0.6166753663857866,
+      "learning_rate": 2.472413465415518e-06,
+      "loss": 0.3679,
+      "step": 3405
+    },
+    {
+      "epoch": 1.585661080074488,
+      "grad_norm": 0.6482010996473748,
+      "learning_rate": 2.4711304265537135e-06,
+      "loss": 0.4036,
+      "step": 3406
+    },
+    {
+      "epoch": 1.5861266294227188,
+      "grad_norm": 0.5894795399821461,
+      "learning_rate": 2.4698473952968403e-06,
+      "loss": 0.3886,
+      "step": 3407
+    },
+    {
+      "epoch": 1.5865921787709496,
+      "grad_norm": 0.5732511989870094,
+      "learning_rate": 2.4685643719828814e-06,
+      "loss": 0.3684,
+      "step": 3408
+    },
+    {
+      "epoch": 1.5870577281191807,
+      "grad_norm": 0.6329808283766557,
+      "learning_rate": 2.4672813569498143e-06,
+      "loss": 0.3926,
+      "step": 3409
+    },
+    {
+      "epoch": 1.5875232774674115,
+      "grad_norm": 0.6369410429477688,
+      "learning_rate": 2.465998350535616e-06,
+      "loss": 0.3967,
+      "step": 3410
+    },
+    {
+      "epoch": 1.5879888268156424,
+      "grad_norm": 0.5922458875341953,
+      "learning_rate": 2.4647153530782595e-06,
+      "loss": 0.3867,
+      "step": 3411
+    },
+    {
+      "epoch": 1.5884543761638734,
+      "grad_norm": 0.5738986447658486,
+      "learning_rate": 2.463432364915719e-06,
+      "loss": 0.3861,
+      "step": 3412
+    },
+    {
+      "epoch": 1.5889199255121043,
+      "grad_norm": 0.6094372276994817,
+      "learning_rate": 2.4621493863859636e-06,
+      "loss": 0.3778,
+      "step": 3413
+    },
+    {
+      "epoch": 1.589385474860335,
+      "grad_norm": 0.5759936677646003,
+      "learning_rate": 2.460866417826959e-06,
+      "loss": 0.3808,
+      "step": 3414
+    },
+    {
+      "epoch": 1.5898510242085662,
+      "grad_norm": 0.5792348595126064,
+      "learning_rate": 2.45958345957667e-06,
+      "loss": 0.3688,
+      "step": 3415
+    },
+    {
+      "epoch": 1.590316573556797,
+      "grad_norm": 0.5820921073340181,
+      "learning_rate": 2.4583005119730605e-06,
+      "loss": 0.3962,
+      "step": 3416
+    },
+    {
+      "epoch": 1.5907821229050279,
+      "grad_norm": 0.5878682438808004,
+      "learning_rate": 2.457017575354087e-06,
+      "loss": 0.3829,
+      "step": 3417
+    },
+    {
+      "epoch": 1.591247672253259,
+      "grad_norm": 0.592006285725705,
+      "learning_rate": 2.455734650057706e-06,
+      "loss": 0.3696,
+      "step": 3418
+    },
+    {
+      "epoch": 1.5917132216014898,
+      "grad_norm": 0.5788581474647592,
+      "learning_rate": 2.4544517364218705e-06,
+      "loss": 0.3932,
+      "step": 3419
+    },
+    {
+      "epoch": 1.5921787709497206,
+      "grad_norm": 0.6108969155585248,
+      "learning_rate": 2.4531688347845315e-06,
+      "loss": 0.3912,
+      "step": 3420
+    },
+    {
+      "epoch": 1.5926443202979517,
+      "grad_norm": 0.6389565134091817,
+      "learning_rate": 2.451885945483635e-06,
+      "loss": 0.3657,
+      "step": 3421
+    },
+    {
+      "epoch": 1.5931098696461825,
+      "grad_norm": 0.6221942987656298,
+      "learning_rate": 2.4506030688571245e-06,
+      "loss": 0.382,
+      "step": 3422
+    },
+    {
+      "epoch": 1.5935754189944134,
+      "grad_norm": 0.5782488430227479,
+      "learning_rate": 2.4493202052429396e-06,
+      "loss": 0.3854,
+      "step": 3423
+    },
+    {
+      "epoch": 1.5940409683426444,
+      "grad_norm": 0.6205642625715355,
+      "learning_rate": 2.4480373549790183e-06,
+      "loss": 0.3756,
+      "step": 3424
+    },
+    {
+      "epoch": 1.5945065176908753,
+      "grad_norm": 0.6188092371115614,
+      "learning_rate": 2.4467545184032936e-06,
+      "loss": 0.3878,
+      "step": 3425
+    },
+    {
+      "epoch": 1.594972067039106,
+      "grad_norm": 0.5880630869085345,
+      "learning_rate": 2.445471695853695e-06,
+      "loss": 0.3673,
+      "step": 3426
+    },
+    {
+      "epoch": 1.5954376163873372,
+      "grad_norm": 0.6062087129484918,
+      "learning_rate": 2.444188887668148e-06,
+      "loss": 0.4147,
+      "step": 3427
+    },
+    {
+      "epoch": 1.595903165735568,
+      "grad_norm": 0.6078079683217542,
+      "learning_rate": 2.442906094184575e-06,
+      "loss": 0.4007,
+      "step": 3428
+    },
+    {
+      "epoch": 1.5963687150837989,
+      "grad_norm": 0.6125997193097813,
+      "learning_rate": 2.4416233157408947e-06,
+      "loss": 0.3921,
+      "step": 3429
+    },
+    {
+      "epoch": 1.59683426443203,
+      "grad_norm": 0.6135192822740414,
+      "learning_rate": 2.440340552675022e-06,
+      "loss": 0.3792,
+      "step": 3430
+    },
+    {
+      "epoch": 1.5972998137802608,
+      "grad_norm": 0.6210371233718099,
+      "learning_rate": 2.439057805324865e-06,
+      "loss": 0.3673,
+      "step": 3431
+    },
+    {
+      "epoch": 1.5977653631284916,
+      "grad_norm": 0.5748726546387346,
+      "learning_rate": 2.4377750740283318e-06,
+      "loss": 0.3994,
+      "step": 3432
+    },
+    {
+      "epoch": 1.5982309124767227,
+      "grad_norm": 0.5611685383481405,
+      "learning_rate": 2.4364923591233244e-06,
+      "loss": 0.3748,
+      "step": 3433
+    },
+    {
+      "epoch": 1.5986964618249533,
+      "grad_norm": 0.6269143497816724,
+      "learning_rate": 2.4352096609477395e-06,
+      "loss": 0.3958,
+      "step": 3434
+    },
+    {
+      "epoch": 1.5991620111731844,
+      "grad_norm": 0.5874919250104647,
+      "learning_rate": 2.433926979839469e-06,
+      "loss": 0.3961,
+      "step": 3435
+    },
+    {
+      "epoch": 1.5996275605214154,
+      "grad_norm": 0.5798940536915418,
+      "learning_rate": 2.432644316136406e-06,
+      "loss": 0.3771,
+      "step": 3436
+    },
+    {
+      "epoch": 1.600093109869646,
+      "grad_norm": 0.571992134993947,
+      "learning_rate": 2.4313616701764304e-06,
+      "loss": 0.3796,
+      "step": 3437
+    },
+    {
+      "epoch": 1.600558659217877,
+      "grad_norm": 0.6258814373778896,
+      "learning_rate": 2.430079042297423e-06,
+      "loss": 0.3929,
+      "step": 3438
+    },
+    {
+      "epoch": 1.6010242085661082,
+      "grad_norm": 0.592111921748307,
+      "learning_rate": 2.4287964328372583e-06,
+      "loss": 0.3905,
+      "step": 3439
+    },
+    {
+      "epoch": 1.6014897579143388,
+      "grad_norm": 1.2594667768426937,
+      "learning_rate": 2.4275138421338067e-06,
+      "loss": 0.3699,
+      "step": 3440
+    },
+    {
+      "epoch": 1.6019553072625698,
+      "grad_norm": 0.5728348067227974,
+      "learning_rate": 2.426231270524933e-06,
+      "loss": 0.3776,
+      "step": 3441
+    },
+    {
+      "epoch": 1.6024208566108007,
+      "grad_norm": 0.632910365589827,
+      "learning_rate": 2.4249487183484966e-06,
+      "loss": 0.382,
+      "step": 3442
+    },
+    {
+      "epoch": 1.6028864059590315,
+      "grad_norm": 0.5870903724788458,
+      "learning_rate": 2.4236661859423523e-06,
+      "loss": 0.3963,
+      "step": 3443
+    },
+    {
+      "epoch": 1.6033519553072626,
+      "grad_norm": 0.6165036216116591,
+      "learning_rate": 2.4223836736443505e-06,
+      "loss": 0.3807,
+      "step": 3444
+    },
+    {
+      "epoch": 1.6038175046554934,
+      "grad_norm": 0.6624139769982373,
+      "learning_rate": 2.421101181792334e-06,
+      "loss": 0.4001,
+      "step": 3445
+    },
+    {
+      "epoch": 1.6042830540037243,
+      "grad_norm": 0.6458183488590868,
+      "learning_rate": 2.4198187107241437e-06,
+      "loss": 0.3806,
+      "step": 3446
+    },
+    {
+      "epoch": 1.6047486033519553,
+      "grad_norm": 0.5875967618915735,
+      "learning_rate": 2.4185362607776097e-06,
+      "loss": 0.3879,
+      "step": 3447
+    },
+    {
+      "epoch": 1.6052141527001862,
+      "grad_norm": 0.5929924570581323,
+      "learning_rate": 2.4172538322905627e-06,
+      "loss": 0.389,
+      "step": 3448
+    },
+    {
+      "epoch": 1.605679702048417,
+      "grad_norm": 0.600035048340961,
+      "learning_rate": 2.415971425600825e-06,
+      "loss": 0.3891,
+      "step": 3449
+    },
+    {
+      "epoch": 1.606145251396648,
+      "grad_norm": 0.5879301548311502,
+      "learning_rate": 2.4146890410462098e-06,
+      "loss": 0.3928,
+      "step": 3450
+    },
+    {
+      "epoch": 1.606610800744879,
+      "grad_norm": 0.6047043091237808,
+      "learning_rate": 2.413406678964531e-06,
+      "loss": 0.3976,
+      "step": 3451
+    },
+    {
+      "epoch": 1.6070763500931098,
+      "grad_norm": 0.5990227766942893,
+      "learning_rate": 2.412124339693593e-06,
+      "loss": 0.3935,
+      "step": 3452
+    },
+    {
+      "epoch": 1.6075418994413408,
+      "grad_norm": 0.5763208635803743,
+      "learning_rate": 2.4108420235711932e-06,
+      "loss": 0.3929,
+      "step": 3453
+    },
+    {
+      "epoch": 1.6080074487895717,
+      "grad_norm": 0.6166884194638825,
+      "learning_rate": 2.409559730935123e-06,
+      "loss": 0.3779,
+      "step": 3454
+    },
+    {
+      "epoch": 1.6084729981378025,
+      "grad_norm": 0.655433212079156,
+      "learning_rate": 2.4082774621231722e-06,
+      "loss": 0.3728,
+      "step": 3455
+    },
+    {
+      "epoch": 1.6089385474860336,
+      "grad_norm": 0.5924757067437042,
+      "learning_rate": 2.4069952174731186e-06,
+      "loss": 0.3728,
+      "step": 3456
+    },
+    {
+      "epoch": 1.6094040968342644,
+      "grad_norm": 0.5595483561202519,
+      "learning_rate": 2.4057129973227358e-06,
+      "loss": 0.3924,
+      "step": 3457
+    },
+    {
+      "epoch": 1.6098696461824953,
+      "grad_norm": 0.6494928758720817,
+      "learning_rate": 2.4044308020097916e-06,
+      "loss": 0.395,
+      "step": 3458
+    },
+    {
+      "epoch": 1.6103351955307263,
+      "grad_norm": 0.5907221207725439,
+      "learning_rate": 2.403148631872047e-06,
+      "loss": 0.3768,
+      "step": 3459
+    },
+    {
+      "epoch": 1.6108007448789572,
+      "grad_norm": 0.5670747623213463,
+      "learning_rate": 2.4018664872472556e-06,
+      "loss": 0.3756,
+      "step": 3460
+    },
+    {
+      "epoch": 1.611266294227188,
+      "grad_norm": 0.5948952816892741,
+      "learning_rate": 2.4005843684731657e-06,
+      "loss": 0.3653,
+      "step": 3461
+    },
+    {
+      "epoch": 1.611731843575419,
+      "grad_norm": 0.5908885528513546,
+      "learning_rate": 2.399302275887516e-06,
+      "loss": 0.3674,
+      "step": 3462
+    },
+    {
+      "epoch": 1.61219739292365,
+      "grad_norm": 0.5760060087513104,
+      "learning_rate": 2.3980202098280426e-06,
+      "loss": 0.401,
+      "step": 3463
+    },
+    {
+      "epoch": 1.6126629422718808,
+      "grad_norm": 0.6027643315388275,
+      "learning_rate": 2.3967381706324706e-06,
+      "loss": 0.3672,
+      "step": 3464
+    },
+    {
+      "epoch": 1.6131284916201118,
+      "grad_norm": 0.5957329066044859,
+      "learning_rate": 2.39545615863852e-06,
+      "loss": 0.3971,
+      "step": 3465
+    },
+    {
+      "epoch": 1.6135940409683427,
+      "grad_norm": 0.6021680030224253,
+      "learning_rate": 2.3941741741839027e-06,
+      "loss": 0.3762,
+      "step": 3466
+    },
+    {
+      "epoch": 1.6140595903165735,
+      "grad_norm": 0.5823567097546977,
+      "learning_rate": 2.3928922176063253e-06,
+      "loss": 0.3583,
+      "step": 3467
+    },
+    {
+      "epoch": 1.6145251396648046,
+      "grad_norm": 0.587406966840796,
+      "learning_rate": 2.3916102892434843e-06,
+      "loss": 0.39,
+      "step": 3468
+    },
+    {
+      "epoch": 1.6149906890130352,
+      "grad_norm": 0.6277784434829247,
+      "learning_rate": 2.3903283894330716e-06,
+      "loss": 0.3806,
+      "step": 3469
+    },
+    {
+      "epoch": 1.6154562383612663,
+      "grad_norm": 0.6227882626499093,
+      "learning_rate": 2.389046518512767e-06,
+      "loss": 0.404,
+      "step": 3470
+    },
+    {
+      "epoch": 1.6159217877094973,
+      "grad_norm": 0.5724536526713287,
+      "learning_rate": 2.3877646768202494e-06,
+      "loss": 0.3716,
+      "step": 3471
+    },
+    {
+      "epoch": 1.616387337057728,
+      "grad_norm": 0.5756488917211782,
+      "learning_rate": 2.3864828646931847e-06,
+      "loss": 0.3682,
+      "step": 3472
+    },
+    {
+      "epoch": 1.616852886405959,
+      "grad_norm": 0.5837825982988918,
+      "learning_rate": 2.385201082469233e-06,
+      "loss": 0.3806,
+      "step": 3473
+    },
+    {
+      "epoch": 1.61731843575419,
+      "grad_norm": 0.6023587358486998,
+      "learning_rate": 2.3839193304860446e-06,
+      "loss": 0.4036,
+      "step": 3474
+    },
+    {
+      "epoch": 1.6177839851024207,
+      "grad_norm": 0.6065301448455783,
+      "learning_rate": 2.3826376090812667e-06,
+      "loss": 0.3881,
+      "step": 3475
+    },
+    {
+      "epoch": 1.6182495344506518,
+      "grad_norm": 0.6951826795520059,
+      "learning_rate": 2.3813559185925327e-06,
+      "loss": 0.3662,
+      "step": 3476
+    },
+    {
+      "epoch": 1.6187150837988828,
+      "grad_norm": 0.6624621310200014,
+      "learning_rate": 2.380074259357471e-06,
+      "loss": 0.395,
+      "step": 3477
+    },
+    {
+      "epoch": 1.6191806331471135,
+      "grad_norm": 0.6107883157667071,
+      "learning_rate": 2.3787926317137007e-06,
+      "loss": 0.3784,
+      "step": 3478
+    },
+    {
+      "epoch": 1.6196461824953445,
+      "grad_norm": 0.5993352020891098,
+      "learning_rate": 2.3775110359988344e-06,
+      "loss": 0.3938,
+      "step": 3479
+    },
+    {
+      "epoch": 1.6201117318435754,
+      "grad_norm": 0.6642211031758445,
+      "learning_rate": 2.3762294725504736e-06,
+      "loss": 0.3772,
+      "step": 3480
+    },
+    {
+      "epoch": 1.6205772811918062,
+      "grad_norm": 0.5922010185721925,
+      "learning_rate": 2.374947941706213e-06,
+      "loss": 0.375,
+      "step": 3481
+    },
+    {
+      "epoch": 1.6210428305400373,
+      "grad_norm": 0.6451914756176987,
+      "learning_rate": 2.373666443803638e-06,
+      "loss": 0.3919,
+      "step": 3482
+    },
+    {
+      "epoch": 1.6215083798882681,
+      "grad_norm": 0.5735267810506091,
+      "learning_rate": 2.372384979180326e-06,
+      "loss": 0.3957,
+      "step": 3483
+    },
+    {
+      "epoch": 1.621973929236499,
+      "grad_norm": 0.6702143117626684,
+      "learning_rate": 2.3711035481738453e-06,
+      "loss": 0.3811,
+      "step": 3484
+    },
+    {
+      "epoch": 1.62243947858473,
+      "grad_norm": 0.6042329996997765,
+      "learning_rate": 2.369822151121755e-06,
+      "loss": 0.3807,
+      "step": 3485
+    },
+    {
+      "epoch": 1.6229050279329609,
+      "grad_norm": 0.6328004899660474,
+      "learning_rate": 2.3685407883616056e-06,
+      "loss": 0.3775,
+      "step": 3486
+    },
+    {
+      "epoch": 1.6233705772811917,
+      "grad_norm": 0.612447408991501,
+      "learning_rate": 2.3672594602309385e-06,
+      "loss": 0.3791,
+      "step": 3487
+    },
+    {
+      "epoch": 1.6238361266294228,
+      "grad_norm": 0.6249055133124229,
+      "learning_rate": 2.365978167067287e-06,
+      "loss": 0.4136,
+      "step": 3488
+    },
+    {
+      "epoch": 1.6243016759776536,
+      "grad_norm": 0.5898356662006469,
+      "learning_rate": 2.364696909208173e-06,
+      "loss": 0.3735,
+      "step": 3489
+    },
+    {
+      "epoch": 1.6247672253258845,
+      "grad_norm": 0.5883403586853114,
+      "learning_rate": 2.363415686991109e-06,
+      "loss": 0.3936,
+      "step": 3490
+    },
+    {
+      "epoch": 1.6252327746741155,
+      "grad_norm": 0.6001955610596604,
+      "learning_rate": 2.3621345007536027e-06,
+      "loss": 0.3956,
+      "step": 3491
+    },
+    {
+      "epoch": 1.6256983240223464,
+      "grad_norm": 0.5657958059964789,
+      "learning_rate": 2.360853350833147e-06,
+      "loss": 0.377,
+      "step": 3492
+    },
+    {
+      "epoch": 1.6261638733705772,
+      "grad_norm": 0.6212810595707668,
+      "learning_rate": 2.3595722375672277e-06,
+      "loss": 0.3689,
+      "step": 3493
+    },
+    {
+      "epoch": 1.6266294227188083,
+      "grad_norm": 0.6114519885780458,
+      "learning_rate": 2.3582911612933194e-06,
+      "loss": 0.4093,
+      "step": 3494
+    },
+    {
+      "epoch": 1.6270949720670391,
+      "grad_norm": 0.5838512720444676,
+      "learning_rate": 2.3570101223488896e-06,
+      "loss": 0.3714,
+      "step": 3495
+    },
+    {
+      "epoch": 1.62756052141527,
+      "grad_norm": 0.6144023303638707,
+      "learning_rate": 2.355729121071394e-06,
+      "loss": 0.381,
+      "step": 3496
+    },
+    {
+      "epoch": 1.628026070763501,
+      "grad_norm": 0.6044159745754771,
+      "learning_rate": 2.3544481577982777e-06,
+      "loss": 0.3733,
+      "step": 3497
+    },
+    {
+      "epoch": 1.6284916201117319,
+      "grad_norm": 0.6304253741257194,
+      "learning_rate": 2.3531672328669788e-06,
+      "loss": 0.3938,
+      "step": 3498
+    },
+    {
+      "epoch": 1.6289571694599627,
+      "grad_norm": 0.5991277617890083,
+      "learning_rate": 2.3518863466149223e-06,
+      "loss": 0.3777,
+      "step": 3499
+    },
+    {
+      "epoch": 1.6294227188081938,
+      "grad_norm": 0.6163747980280212,
+      "learning_rate": 2.3506054993795243e-06,
+      "loss": 0.3796,
+      "step": 3500
+    },
+    {
+      "epoch": 1.6298882681564246,
+      "grad_norm": 0.5799911358552265,
+      "learning_rate": 2.349324691498189e-06,
+      "loss": 0.3696,
+      "step": 3501
+    },
+    {
+      "epoch": 1.6303538175046555,
+      "grad_norm": 0.603381151421239,
+      "learning_rate": 2.3480439233083145e-06,
+      "loss": 0.3805,
+      "step": 3502
+    },
+    {
+      "epoch": 1.6308193668528865,
+      "grad_norm": 0.5939617401650598,
+      "learning_rate": 2.346763195147284e-06,
+      "loss": 0.3817,
+      "step": 3503
+    },
+    {
+      "epoch": 1.6312849162011172,
+      "grad_norm": 0.618033621666375,
+      "learning_rate": 2.345482507352471e-06,
+      "loss": 0.3738,
+      "step": 3504
+    },
+    {
+      "epoch": 1.6317504655493482,
+      "grad_norm": 0.6496364397095213,
+      "learning_rate": 2.34420186026124e-06,
+      "loss": 0.3918,
+      "step": 3505
+    },
+    {
+      "epoch": 1.6322160148975793,
+      "grad_norm": 0.5900854438928237,
+      "learning_rate": 2.3429212542109445e-06,
+      "loss": 0.3781,
+      "step": 3506
+    },
+    {
+      "epoch": 1.63268156424581,
+      "grad_norm": 0.6081903563169392,
+      "learning_rate": 2.341640689538926e-06,
+      "loss": 0.3859,
+      "step": 3507
+    },
+    {
+      "epoch": 1.633147113594041,
+      "grad_norm": 0.5770768604789464,
+      "learning_rate": 2.3403601665825158e-06,
+      "loss": 0.3844,
+      "step": 3508
+    },
+    {
+      "epoch": 1.633612662942272,
+      "grad_norm": 0.5923672030650747,
+      "learning_rate": 2.339079685679032e-06,
+      "loss": 0.3849,
+      "step": 3509
+    },
+    {
+      "epoch": 1.6340782122905027,
+      "grad_norm": 0.6015269532809103,
+      "learning_rate": 2.3377992471657874e-06,
+      "loss": 0.3938,
+      "step": 3510
+    },
+    {
+      "epoch": 1.6345437616387337,
+      "grad_norm": 0.5682659947698702,
+      "learning_rate": 2.3365188513800784e-06,
+      "loss": 0.3721,
+      "step": 3511
+    },
+    {
+      "epoch": 1.6350093109869648,
+      "grad_norm": 0.5979979497816684,
+      "learning_rate": 2.335238498659191e-06,
+      "loss": 0.3879,
+      "step": 3512
+    },
+    {
+      "epoch": 1.6354748603351954,
+      "grad_norm": 0.6037131936994778,
+      "learning_rate": 2.3339581893404e-06,
+      "loss": 0.3791,
+      "step": 3513
+    },
+    {
+      "epoch": 1.6359404096834265,
+      "grad_norm": 0.6111320790814828,
+      "learning_rate": 2.3326779237609716e-06,
+      "loss": 0.3729,
+      "step": 3514
+    },
+    {
+      "epoch": 1.6364059590316573,
+      "grad_norm": 0.6218208239413111,
+      "learning_rate": 2.3313977022581567e-06,
+      "loss": 0.3777,
+      "step": 3515
+    },
+    {
+      "epoch": 1.6368715083798882,
+      "grad_norm": 0.6527099025558417,
+      "learning_rate": 2.3301175251691964e-06,
+      "loss": 0.3705,
+      "step": 3516
+    },
+    {
+      "epoch": 1.6373370577281192,
+      "grad_norm": 0.5829374753808249,
+      "learning_rate": 2.328837392831319e-06,
+      "loss": 0.3677,
+      "step": 3517
+    },
+    {
+      "epoch": 1.63780260707635,
+      "grad_norm": 0.608726300955895,
+      "learning_rate": 2.327557305581743e-06,
+      "loss": 0.3818,
+      "step": 3518
+    },
+    {
+      "epoch": 1.638268156424581,
+      "grad_norm": 0.6301542042050216,
+      "learning_rate": 2.326277263757674e-06,
+      "loss": 0.3908,
+      "step": 3519
+    },
+    {
+      "epoch": 1.638733705772812,
+      "grad_norm": 0.5690432500278211,
+      "learning_rate": 2.324997267696304e-06,
+      "loss": 0.3797,
+      "step": 3520
+    },
+    {
+      "epoch": 1.6391992551210428,
+      "grad_norm": 0.6338118849559748,
+      "learning_rate": 2.323717317734815e-06,
+      "loss": 0.3854,
+      "step": 3521
+    },
+    {
+      "epoch": 1.6396648044692737,
+      "grad_norm": 0.6047906103815329,
+      "learning_rate": 2.3224374142103767e-06,
+      "loss": 0.3752,
+      "step": 3522
+    },
+    {
+      "epoch": 1.6401303538175047,
+      "grad_norm": 0.5999280196402931,
+      "learning_rate": 2.321157557460146e-06,
+      "loss": 0.402,
+      "step": 3523
+    },
+    {
+      "epoch": 1.6405959031657356,
+      "grad_norm": 0.6192093695863823,
+      "learning_rate": 2.3198777478212678e-06,
+      "loss": 0.3854,
+      "step": 3524
+    },
+    {
+      "epoch": 1.6410614525139664,
+      "grad_norm": 0.5833089210781524,
+      "learning_rate": 2.3185979856308734e-06,
+      "loss": 0.3703,
+      "step": 3525
+    },
+    {
+      "epoch": 1.6415270018621975,
+      "grad_norm": 0.5966372359178783,
+      "learning_rate": 2.317318271226084e-06,
+      "loss": 0.3926,
+      "step": 3526
+    },
+    {
+      "epoch": 1.6419925512104283,
+      "grad_norm": 0.5900962368363084,
+      "learning_rate": 2.3160386049440057e-06,
+      "loss": 0.3748,
+      "step": 3527
+    },
+    {
+      "epoch": 1.6424581005586592,
+      "grad_norm": 0.573346456854765,
+      "learning_rate": 2.3147589871217337e-06,
+      "loss": 0.3812,
+      "step": 3528
+    },
+    {
+      "epoch": 1.6429236499068902,
+      "grad_norm": 0.6022876636291857,
+      "learning_rate": 2.3134794180963484e-06,
+      "loss": 0.373,
+      "step": 3529
+    },
+    {
+      "epoch": 1.643389199255121,
+      "grad_norm": 0.6001368771267699,
+      "learning_rate": 2.312199898204921e-06,
+      "loss": 0.3909,
+      "step": 3530
+    },
+    {
+      "epoch": 1.643854748603352,
+      "grad_norm": 0.6186757884047726,
+      "learning_rate": 2.310920427784506e-06,
+      "loss": 0.3774,
+      "step": 3531
+    },
+    {
+      "epoch": 1.644320297951583,
+      "grad_norm": 0.6204724209971851,
+      "learning_rate": 2.3096410071721464e-06,
+      "loss": 0.3894,
+      "step": 3532
+    },
+    {
+      "epoch": 1.6447858472998138,
+      "grad_norm": 0.5941913070437141,
+      "learning_rate": 2.3083616367048716e-06,
+      "loss": 0.3731,
+      "step": 3533
+    },
+    {
+      "epoch": 1.6452513966480447,
+      "grad_norm": 0.5988864936113284,
+      "learning_rate": 2.3070823167196995e-06,
+      "loss": 0.3683,
+      "step": 3534
+    },
+    {
+      "epoch": 1.6457169459962757,
+      "grad_norm": 0.5911280575753998,
+      "learning_rate": 2.3058030475536328e-06,
+      "loss": 0.3912,
+      "step": 3535
+    },
+    {
+      "epoch": 1.6461824953445066,
+      "grad_norm": 0.6028942066044422,
+      "learning_rate": 2.304523829543661e-06,
+      "loss": 0.3854,
+      "step": 3536
+    },
+    {
+      "epoch": 1.6466480446927374,
+      "grad_norm": 0.6001758512196758,
+      "learning_rate": 2.303244663026761e-06,
+      "loss": 0.3889,
+      "step": 3537
+    },
+    {
+      "epoch": 1.6471135940409685,
+      "grad_norm": 0.5929668429708398,
+      "learning_rate": 2.3019655483398954e-06,
+      "loss": 0.3872,
+      "step": 3538
+    },
+    {
+      "epoch": 1.6475791433891993,
+      "grad_norm": 0.5824924374918856,
+      "learning_rate": 2.3006864858200137e-06,
+      "loss": 0.3831,
+      "step": 3539
+    },
+    {
+      "epoch": 1.6480446927374302,
+      "grad_norm": 0.5707431047679731,
+      "learning_rate": 2.2994074758040508e-06,
+      "loss": 0.375,
+      "step": 3540
+    },
+    {
+      "epoch": 1.6485102420856612,
+      "grad_norm": 0.554050988485632,
+      "learning_rate": 2.298128518628929e-06,
+      "loss": 0.3654,
+      "step": 3541
+    },
+    {
+      "epoch": 1.6489757914338918,
+      "grad_norm": 0.6316934620343061,
+      "learning_rate": 2.296849614631556e-06,
+      "loss": 0.4061,
+      "step": 3542
+    },
+    {
+      "epoch": 1.649441340782123,
+      "grad_norm": 0.5824572363095282,
+      "learning_rate": 2.2955707641488257e-06,
+      "loss": 0.4012,
+      "step": 3543
+    },
+    {
+      "epoch": 1.649906890130354,
+      "grad_norm": 0.6157048595429053,
+      "learning_rate": 2.2942919675176167e-06,
+      "loss": 0.3872,
+      "step": 3544
+    },
+    {
+      "epoch": 1.6503724394785846,
+      "grad_norm": 0.6185862425210628,
+      "learning_rate": 2.2930132250747955e-06,
+      "loss": 0.3714,
+      "step": 3545
+    },
+    {
+      "epoch": 1.6508379888268156,
+      "grad_norm": 0.5871240129891143,
+      "learning_rate": 2.291734537157213e-06,
+      "loss": 0.3914,
+      "step": 3546
+    },
+    {
+      "epoch": 1.6513035381750467,
+      "grad_norm": 0.6069548246223249,
+      "learning_rate": 2.2904559041017067e-06,
+      "loss": 0.3732,
+      "step": 3547
+    },
+    {
+      "epoch": 1.6517690875232773,
+      "grad_norm": 0.5481964396705619,
+      "learning_rate": 2.2891773262450963e-06,
+      "loss": 0.3769,
+      "step": 3548
+    },
+    {
+      "epoch": 1.6522346368715084,
+      "grad_norm": 0.5978634690346435,
+      "learning_rate": 2.2878988039241923e-06,
+      "loss": 0.3908,
+      "step": 3549
+    },
+    {
+      "epoch": 1.6527001862197392,
+      "grad_norm": 0.5739458613872462,
+      "learning_rate": 2.286620337475788e-06,
+      "loss": 0.3772,
+      "step": 3550
+    },
+    {
+      "epoch": 1.65316573556797,
+      "grad_norm": 0.5906800957600725,
+      "learning_rate": 2.28534192723666e-06,
+      "loss": 0.3764,
+      "step": 3551
+    },
+    {
+      "epoch": 1.6536312849162011,
+      "grad_norm": 0.6093260823580311,
+      "learning_rate": 2.2840635735435717e-06,
+      "loss": 0.3759,
+      "step": 3552
+    },
+    {
+      "epoch": 1.654096834264432,
+      "grad_norm": 0.5904257684876055,
+      "learning_rate": 2.282785276733275e-06,
+      "loss": 0.3864,
+      "step": 3553
+    },
+    {
+      "epoch": 1.6545623836126628,
+      "grad_norm": 0.5978813765824175,
+      "learning_rate": 2.281507037142501e-06,
+      "loss": 0.3996,
+      "step": 3554
+    },
+    {
+      "epoch": 1.655027932960894,
+      "grad_norm": 0.613174740200068,
+      "learning_rate": 2.280228855107969e-06,
+      "loss": 0.3713,
+      "step": 3555
+    },
+    {
+      "epoch": 1.6554934823091247,
+      "grad_norm": 0.6243086347426036,
+      "learning_rate": 2.2789507309663828e-06,
+      "loss": 0.3788,
+      "step": 3556
+    },
+    {
+      "epoch": 1.6559590316573556,
+      "grad_norm": 0.5664272841683077,
+      "learning_rate": 2.2776726650544304e-06,
+      "loss": 0.3757,
+      "step": 3557
+    },
+    {
+      "epoch": 1.6564245810055866,
+      "grad_norm": 0.585431916342177,
+      "learning_rate": 2.2763946577087852e-06,
+      "loss": 0.375,
+      "step": 3558
+    },
+    {
+      "epoch": 1.6568901303538175,
+      "grad_norm": 0.5933118720858785,
+      "learning_rate": 2.2751167092661043e-06,
+      "loss": 0.3825,
+      "step": 3559
+    },
+    {
+      "epoch": 1.6573556797020483,
+      "grad_norm": 0.5801749013113281,
+      "learning_rate": 2.2738388200630294e-06,
+      "loss": 0.3683,
+      "step": 3560
+    },
+    {
+      "epoch": 1.6578212290502794,
+      "grad_norm": 0.6069584915094919,
+      "learning_rate": 2.2725609904361875e-06,
+      "loss": 0.3989,
+      "step": 3561
+    },
+    {
+      "epoch": 1.6582867783985102,
+      "grad_norm": 0.554208498560856,
+      "learning_rate": 2.2712832207221895e-06,
+      "loss": 0.3668,
+      "step": 3562
+    },
+    {
+      "epoch": 1.658752327746741,
+      "grad_norm": 0.6374571772740507,
+      "learning_rate": 2.2700055112576298e-06,
+      "loss": 0.4012,
+      "step": 3563
+    },
+    {
+      "epoch": 1.6592178770949721,
+      "grad_norm": 0.6856115200024218,
+      "learning_rate": 2.2687278623790875e-06,
+      "loss": 0.381,
+      "step": 3564
+    },
+    {
+      "epoch": 1.659683426443203,
+      "grad_norm": 0.6058988034978298,
+      "learning_rate": 2.2674502744231257e-06,
+      "loss": 0.3775,
+      "step": 3565
+    },
+    {
+      "epoch": 1.6601489757914338,
+      "grad_norm": 0.6080628928185956,
+      "learning_rate": 2.2661727477262926e-06,
+      "loss": 0.3857,
+      "step": 3566
+    },
+    {
+      "epoch": 1.660614525139665,
+      "grad_norm": 0.5818935587884111,
+      "learning_rate": 2.2648952826251177e-06,
+      "loss": 0.3883,
+      "step": 3567
+    },
+    {
+      "epoch": 1.6610800744878957,
+      "grad_norm": 0.5700109945783416,
+      "learning_rate": 2.2636178794561145e-06,
+      "loss": 0.3868,
+      "step": 3568
+    },
+    {
+      "epoch": 1.6615456238361266,
+      "grad_norm": 0.5962004011899051,
+      "learning_rate": 2.2623405385557847e-06,
+      "loss": 0.3729,
+      "step": 3569
+    },
+    {
+      "epoch": 1.6620111731843576,
+      "grad_norm": 0.6258491517766632,
+      "learning_rate": 2.261063260260608e-06,
+      "loss": 0.3553,
+      "step": 3570
+    },
+    {
+      "epoch": 1.6624767225325885,
+      "grad_norm": 0.5663196713612129,
+      "learning_rate": 2.2597860449070505e-06,
+      "loss": 0.3765,
+      "step": 3571
+    },
+    {
+      "epoch": 1.6629422718808193,
+      "grad_norm": 0.6247498584884338,
+      "learning_rate": 2.2585088928315603e-06,
+      "loss": 0.3877,
+      "step": 3572
+    },
+    {
+      "epoch": 1.6634078212290504,
+      "grad_norm": 0.5990822380256109,
+      "learning_rate": 2.257231804370571e-06,
+      "loss": 0.3934,
+      "step": 3573
+    },
+    {
+      "epoch": 1.6638733705772812,
+      "grad_norm": 0.5950932753179675,
+      "learning_rate": 2.2559547798604974e-06,
+      "loss": 0.3956,
+      "step": 3574
+    },
+    {
+      "epoch": 1.664338919925512,
+      "grad_norm": 0.6350547788110109,
+      "learning_rate": 2.254677819637738e-06,
+      "loss": 0.3736,
+      "step": 3575
+    },
+    {
+      "epoch": 1.6648044692737431,
+      "grad_norm": 0.578744173911571,
+      "learning_rate": 2.253400924038674e-06,
+      "loss": 0.3725,
+      "step": 3576
+    },
+    {
+      "epoch": 1.6652700186219738,
+      "grad_norm": 0.5954401005844683,
+      "learning_rate": 2.2521240933996713e-06,
+      "loss": 0.38,
+      "step": 3577
+    },
+    {
+      "epoch": 1.6657355679702048,
+      "grad_norm": 0.6111647733091992,
+      "learning_rate": 2.2508473280570767e-06,
+      "loss": 0.3656,
+      "step": 3578
+    },
+    {
+      "epoch": 1.666201117318436,
+      "grad_norm": 0.6019218473184902,
+      "learning_rate": 2.249570628347221e-06,
+      "loss": 0.3853,
+      "step": 3579
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.6112854195027974,
+      "learning_rate": 2.248293994606416e-06,
+      "loss": 0.3951,
+      "step": 3580
+    },
+    {
+      "epoch": 1.6671322160148976,
+      "grad_norm": 0.6340316605566847,
+      "learning_rate": 2.2470174271709593e-06,
+      "loss": 0.3905,
+      "step": 3581
+    },
+    {
+      "epoch": 1.6675977653631286,
+      "grad_norm": 0.6090003833081427,
+      "learning_rate": 2.245740926377128e-06,
+      "loss": 0.4015,
+      "step": 3582
+    },
+    {
+      "epoch": 1.6680633147113593,
+      "grad_norm": 0.5780202917122667,
+      "learning_rate": 2.2444644925611816e-06,
+      "loss": 0.3886,
+      "step": 3583
+    },
+    {
+      "epoch": 1.6685288640595903,
+      "grad_norm": 0.6167824452670897,
+      "learning_rate": 2.243188126059366e-06,
+      "loss": 0.3817,
+      "step": 3584
+    },
+    {
+      "epoch": 1.6689944134078212,
+      "grad_norm": 0.6271532740330483,
+      "learning_rate": 2.241911827207904e-06,
+      "loss": 0.3756,
+      "step": 3585
+    },
+    {
+      "epoch": 1.669459962756052,
+      "grad_norm": 0.6336061567481529,
+      "learning_rate": 2.240635596343005e-06,
+      "loss": 0.3906,
+      "step": 3586
+    },
+    {
+      "epoch": 1.669925512104283,
+      "grad_norm": 0.6164919086438051,
+      "learning_rate": 2.239359433800856e-06,
+      "loss": 0.3818,
+      "step": 3587
+    },
+    {
+      "epoch": 1.670391061452514,
+      "grad_norm": 0.5901689571290734,
+      "learning_rate": 2.2380833399176306e-06,
+      "loss": 0.3776,
+      "step": 3588
+    },
+    {
+      "epoch": 1.6708566108007448,
+      "grad_norm": 0.5782529418916865,
+      "learning_rate": 2.236807315029483e-06,
+      "loss": 0.3788,
+      "step": 3589
+    },
+    {
+      "epoch": 1.6713221601489758,
+      "grad_norm": 0.5794034693450809,
+      "learning_rate": 2.2355313594725465e-06,
+      "loss": 0.3826,
+      "step": 3590
+    },
+    {
+      "epoch": 1.6717877094972067,
+      "grad_norm": 0.6432973375057006,
+      "learning_rate": 2.2342554735829375e-06,
+      "loss": 0.3817,
+      "step": 3591
+    },
+    {
+      "epoch": 1.6722532588454375,
+      "grad_norm": 0.5717934028855738,
+      "learning_rate": 2.232979657696758e-06,
+      "loss": 0.3923,
+      "step": 3592
+    },
+    {
+      "epoch": 1.6727188081936686,
+      "grad_norm": 0.604312885120609,
+      "learning_rate": 2.2317039121500857e-06,
+      "loss": 0.3756,
+      "step": 3593
+    },
+    {
+      "epoch": 1.6731843575418994,
+      "grad_norm": 0.60355237411023,
+      "learning_rate": 2.230428237278983e-06,
+      "loss": 0.3833,
+      "step": 3594
+    },
+    {
+      "epoch": 1.6736499068901303,
+      "grad_norm": 0.5871265232684658,
+      "learning_rate": 2.2291526334194924e-06,
+      "loss": 0.3814,
+      "step": 3595
+    },
+    {
+      "epoch": 1.6741154562383613,
+      "grad_norm": 0.6360446213204384,
+      "learning_rate": 2.2278771009076392e-06,
+      "loss": 0.3844,
+      "step": 3596
+    },
+    {
+      "epoch": 1.6745810055865922,
+      "grad_norm": 0.6173040243028332,
+      "learning_rate": 2.226601640079429e-06,
+      "loss": 0.3855,
+      "step": 3597
+    },
+    {
+      "epoch": 1.675046554934823,
+      "grad_norm": 0.5723360154733633,
+      "learning_rate": 2.225326251270848e-06,
+      "loss": 0.37,
+      "step": 3598
+    },
+    {
+      "epoch": 1.675512104283054,
+      "grad_norm": 0.599915820006921,
+      "learning_rate": 2.2240509348178634e-06,
+      "loss": 0.396,
+      "step": 3599
+    },
+    {
+      "epoch": 1.675977653631285,
+      "grad_norm": 0.6081947429017167,
+      "learning_rate": 2.222775691056425e-06,
+      "loss": 0.3832,
+      "step": 3600
+    },
+    {
+      "epoch": 1.6764432029795158,
+      "grad_norm": 0.6254445726084878,
+      "learning_rate": 2.2215005203224616e-06,
+      "loss": 0.3808,
+      "step": 3601
+    },
+    {
+      "epoch": 1.6769087523277468,
+      "grad_norm": 0.6460975742064783,
+      "learning_rate": 2.2202254229518842e-06,
+      "loss": 0.3688,
+      "step": 3602
+    },
+    {
+      "epoch": 1.6773743016759777,
+      "grad_norm": 0.6449080584117269,
+      "learning_rate": 2.218950399280583e-06,
+      "loss": 0.3917,
+      "step": 3603
+    },
+    {
+      "epoch": 1.6778398510242085,
+      "grad_norm": 0.6083486826761622,
+      "learning_rate": 2.2176754496444304e-06,
+      "loss": 0.3856,
+      "step": 3604
+    },
+    {
+      "epoch": 1.6783054003724396,
+      "grad_norm": 0.6389237047591109,
+      "learning_rate": 2.2164005743792783e-06,
+      "loss": 0.4073,
+      "step": 3605
+    },
+    {
+      "epoch": 1.6787709497206704,
+      "grad_norm": 0.5908945268798834,
+      "learning_rate": 2.215125773820958e-06,
+      "loss": 0.3923,
+      "step": 3606
+    },
+    {
+      "epoch": 1.6792364990689013,
+      "grad_norm": 0.5998620571029831,
+      "learning_rate": 2.2138510483052825e-06,
+      "loss": 0.3777,
+      "step": 3607
+    },
+    {
+      "epoch": 1.6797020484171323,
+      "grad_norm": 0.6037525412555245,
+      "learning_rate": 2.2125763981680466e-06,
+      "loss": 0.3806,
+      "step": 3608
+    },
+    {
+      "epoch": 1.6801675977653632,
+      "grad_norm": 0.5904700953844587,
+      "learning_rate": 2.211301823745022e-06,
+      "loss": 0.3523,
+      "step": 3609
+    },
+    {
+      "epoch": 1.680633147113594,
+      "grad_norm": 0.6202029608411114,
+      "learning_rate": 2.210027325371962e-06,
+      "loss": 0.3789,
+      "step": 3610
+    },
+    {
+      "epoch": 1.681098696461825,
+      "grad_norm": 0.5838440471007457,
+      "learning_rate": 2.2087529033846e-06,
+      "loss": 0.3696,
+      "step": 3611
+    },
+    {
+      "epoch": 1.6815642458100557,
+      "grad_norm": 0.6027196798687443,
+      "learning_rate": 2.207478558118649e-06,
+      "loss": 0.3869,
+      "step": 3612
+    },
+    {
+      "epoch": 1.6820297951582868,
+      "grad_norm": 0.5914861504374449,
+      "learning_rate": 2.206204289909802e-06,
+      "loss": 0.3905,
+      "step": 3613
+    },
+    {
+      "epoch": 1.6824953445065178,
+      "grad_norm": 0.629120227385794,
+      "learning_rate": 2.2049300990937313e-06,
+      "loss": 0.3761,
+      "step": 3614
+    },
+    {
+      "epoch": 1.6829608938547485,
+      "grad_norm": 0.6010287512762302,
+      "learning_rate": 2.2036559860060884e-06,
+      "loss": 0.3695,
+      "step": 3615
+    },
+    {
+      "epoch": 1.6834264432029795,
+      "grad_norm": 0.5548854369636239,
+      "learning_rate": 2.202381950982507e-06,
+      "loss": 0.3827,
+      "step": 3616
+    },
+    {
+      "epoch": 1.6838919925512106,
+      "grad_norm": 0.6153415586080355,
+      "learning_rate": 2.201107994358596e-06,
+      "loss": 0.3917,
+      "step": 3617
+    },
+    {
+      "epoch": 1.6843575418994412,
+      "grad_norm": 0.6112705396613325,
+      "learning_rate": 2.1998341164699476e-06,
+      "loss": 0.4028,
+      "step": 3618
+    },
+    {
+      "epoch": 1.6848230912476723,
+      "grad_norm": 0.6033496110831116,
+      "learning_rate": 2.19856031765213e-06,
+      "loss": 0.3763,
+      "step": 3619
+    },
+    {
+      "epoch": 1.6852886405959033,
+      "grad_norm": 0.6293136950021325,
+      "learning_rate": 2.1972865982406933e-06,
+      "loss": 0.3842,
+      "step": 3620
+    },
+    {
+      "epoch": 1.685754189944134,
+      "grad_norm": 0.5860623786449719,
+      "learning_rate": 2.1960129585711653e-06,
+      "loss": 0.3794,
+      "step": 3621
+    },
+    {
+      "epoch": 1.686219739292365,
+      "grad_norm": 0.5646391954917162,
+      "learning_rate": 2.194739398979053e-06,
+      "loss": 0.3678,
+      "step": 3622
+    },
+    {
+      "epoch": 1.6866852886405959,
+      "grad_norm": 0.5763748456311608,
+      "learning_rate": 2.193465919799841e-06,
+      "loss": 0.3788,
+      "step": 3623
+    },
+    {
+      "epoch": 1.6871508379888267,
+      "grad_norm": 0.6616765206607907,
+      "learning_rate": 2.1921925213689953e-06,
+      "loss": 0.4015,
+      "step": 3624
+    },
+    {
+      "epoch": 1.6876163873370578,
+      "grad_norm": 0.6180590263294825,
+      "learning_rate": 2.19091920402196e-06,
+      "loss": 0.3691,
+      "step": 3625
+    },
+    {
+      "epoch": 1.6880819366852886,
+      "grad_norm": 0.586819805481528,
+      "learning_rate": 2.189645968094155e-06,
+      "loss": 0.3796,
+      "step": 3626
+    },
+    {
+      "epoch": 1.6885474860335195,
+      "grad_norm": 0.5928910449311671,
+      "learning_rate": 2.1883728139209824e-06,
+      "loss": 0.3645,
+      "step": 3627
+    },
+    {
+      "epoch": 1.6890130353817505,
+      "grad_norm": 0.6282597342690779,
+      "learning_rate": 2.187099741837822e-06,
+      "loss": 0.3773,
+      "step": 3628
+    },
+    {
+      "epoch": 1.6894785847299814,
+      "grad_norm": 0.5807512974329231,
+      "learning_rate": 2.1858267521800296e-06,
+      "loss": 0.3796,
+      "step": 3629
+    },
+    {
+      "epoch": 1.6899441340782122,
+      "grad_norm": 0.606881132025788,
+      "learning_rate": 2.1845538452829403e-06,
+      "loss": 0.3825,
+      "step": 3630
+    },
+    {
+      "epoch": 1.6904096834264433,
+      "grad_norm": 0.564900351590418,
+      "learning_rate": 2.1832810214818715e-06,
+      "loss": 0.3869,
+      "step": 3631
+    },
+    {
+      "epoch": 1.690875232774674,
+      "grad_norm": 0.5977485835725183,
+      "learning_rate": 2.182008281112112e-06,
+      "loss": 0.3815,
+      "step": 3632
+    },
+    {
+      "epoch": 1.691340782122905,
+      "grad_norm": 0.5795962108233051,
+      "learning_rate": 2.1807356245089336e-06,
+      "loss": 0.3703,
+      "step": 3633
+    },
+    {
+      "epoch": 1.691806331471136,
+      "grad_norm": 0.592452157028812,
+      "learning_rate": 2.1794630520075825e-06,
+      "loss": 0.3754,
+      "step": 3634
+    },
+    {
+      "epoch": 1.6922718808193669,
+      "grad_norm": 0.6022555243024755,
+      "learning_rate": 2.178190563943287e-06,
+      "loss": 0.3666,
+      "step": 3635
+    },
+    {
+      "epoch": 1.6927374301675977,
+      "grad_norm": 0.5906275131442605,
+      "learning_rate": 2.176918160651249e-06,
+      "loss": 0.3875,
+      "step": 3636
+    },
+    {
+      "epoch": 1.6932029795158288,
+      "grad_norm": 0.586872297591545,
+      "learning_rate": 2.1756458424666504e-06,
+      "loss": 0.4024,
+      "step": 3637
+    },
+    {
+      "epoch": 1.6936685288640596,
+      "grad_norm": 0.6098815479359213,
+      "learning_rate": 2.1743736097246496e-06,
+      "loss": 0.3819,
+      "step": 3638
+    },
+    {
+      "epoch": 1.6941340782122905,
+      "grad_norm": 0.582411413376565,
+      "learning_rate": 2.1731014627603836e-06,
+      "loss": 0.3716,
+      "step": 3639
+    },
+    {
+      "epoch": 1.6945996275605215,
+      "grad_norm": 0.6395315737720496,
+      "learning_rate": 2.171829401908966e-06,
+      "loss": 0.3671,
+      "step": 3640
+    },
+    {
+      "epoch": 1.6950651769087524,
+      "grad_norm": 0.6047576865069842,
+      "learning_rate": 2.1705574275054875e-06,
+      "loss": 0.3911,
+      "step": 3641
+    },
+    {
+      "epoch": 1.6955307262569832,
+      "grad_norm": 0.577626197357069,
+      "learning_rate": 2.1692855398850163e-06,
+      "loss": 0.3795,
+      "step": 3642
+    },
+    {
+      "epoch": 1.6959962756052143,
+      "grad_norm": 0.6453581892989976,
+      "learning_rate": 2.1680137393825986e-06,
+      "loss": 0.4158,
+      "step": 3643
+    },
+    {
+      "epoch": 1.696461824953445,
+      "grad_norm": 0.5937217569088595,
+      "learning_rate": 2.166742026333257e-06,
+      "loss": 0.3657,
+      "step": 3644
+    },
+    {
+      "epoch": 1.696927374301676,
+      "grad_norm": 0.6413884675846614,
+      "learning_rate": 2.16547040107199e-06,
+      "loss": 0.3919,
+      "step": 3645
+    },
+    {
+      "epoch": 1.697392923649907,
+      "grad_norm": 0.5786113317565376,
+      "learning_rate": 2.1641988639337735e-06,
+      "loss": 0.3978,
+      "step": 3646
+    },
+    {
+      "epoch": 1.6978584729981379,
+      "grad_norm": 0.5867928737307108,
+      "learning_rate": 2.162927415253563e-06,
+      "loss": 0.3935,
+      "step": 3647
+    },
+    {
+      "epoch": 1.6983240223463687,
+      "grad_norm": 0.6042641657348156,
+      "learning_rate": 2.161656055366286e-06,
+      "loss": 0.393,
+      "step": 3648
+    },
+    {
+      "epoch": 1.6987895716945998,
+      "grad_norm": 0.609087289271517,
+      "learning_rate": 2.16038478460685e-06,
+      "loss": 0.3884,
+      "step": 3649
+    },
+    {
+      "epoch": 1.6992551210428304,
+      "grad_norm": 0.5791276021810634,
+      "learning_rate": 2.1591136033101366e-06,
+      "loss": 0.3964,
+      "step": 3650
+    },
+    {
+      "epoch": 1.6997206703910615,
+      "grad_norm": 0.5889709764249135,
+      "learning_rate": 2.1578425118110074e-06,
+      "loss": 0.3786,
+      "step": 3651
+    },
+    {
+      "epoch": 1.7001862197392925,
+      "grad_norm": 0.6116764814263596,
+      "learning_rate": 2.1565715104442968e-06,
+      "loss": 0.3791,
+      "step": 3652
+    },
+    {
+      "epoch": 1.7006517690875231,
+      "grad_norm": 0.6238513039576928,
+      "learning_rate": 2.155300599544817e-06,
+      "loss": 0.3671,
+      "step": 3653
+    },
+    {
+      "epoch": 1.7011173184357542,
+      "grad_norm": 0.6236482246744344,
+      "learning_rate": 2.1540297794473555e-06,
+      "loss": 0.3841,
+      "step": 3654
+    },
+    {
+      "epoch": 1.7015828677839853,
+      "grad_norm": 0.6092770446157714,
+      "learning_rate": 2.152759050486678e-06,
+      "loss": 0.3876,
+      "step": 3655
+    },
+    {
+      "epoch": 1.7020484171322159,
+      "grad_norm": 0.6168106900376644,
+      "learning_rate": 2.151488412997523e-06,
+      "loss": 0.3832,
+      "step": 3656
+    },
+    {
+      "epoch": 1.702513966480447,
+      "grad_norm": 0.6285283133652455,
+      "learning_rate": 2.1502178673146087e-06,
+      "loss": 0.3896,
+      "step": 3657
+    },
+    {
+      "epoch": 1.7029795158286778,
+      "grad_norm": 0.5821573135350989,
+      "learning_rate": 2.1489474137726247e-06,
+      "loss": 0.3664,
+      "step": 3658
+    },
+    {
+      "epoch": 1.7034450651769086,
+      "grad_norm": 0.5871865251629572,
+      "learning_rate": 2.1476770527062404e-06,
+      "loss": 0.3765,
+      "step": 3659
+    },
+    {
+      "epoch": 1.7039106145251397,
+      "grad_norm": 0.5800351472448743,
+      "learning_rate": 2.1464067844500984e-06,
+      "loss": 0.373,
+      "step": 3660
+    },
+    {
+      "epoch": 1.7043761638733705,
+      "grad_norm": 0.6300776142189923,
+      "learning_rate": 2.145136609338819e-06,
+      "loss": 0.3777,
+      "step": 3661
+    },
+    {
+      "epoch": 1.7048417132216014,
+      "grad_norm": 0.6465212808724523,
+      "learning_rate": 2.1438665277069927e-06,
+      "loss": 0.386,
+      "step": 3662
+    },
+    {
+      "epoch": 1.7053072625698324,
+      "grad_norm": 0.5857013666564918,
+      "learning_rate": 2.142596539889193e-06,
+      "loss": 0.3961,
+      "step": 3663
+    },
+    {
+      "epoch": 1.7057728119180633,
+      "grad_norm": 0.6101315033086253,
+      "learning_rate": 2.1413266462199643e-06,
+      "loss": 0.3759,
+      "step": 3664
+    },
+    {
+      "epoch": 1.7062383612662941,
+      "grad_norm": 0.6115615620512916,
+      "learning_rate": 2.1400568470338257e-06,
+      "loss": 0.4024,
+      "step": 3665
+    },
+    {
+      "epoch": 1.7067039106145252,
+      "grad_norm": 0.6048794228036459,
+      "learning_rate": 2.138787142665272e-06,
+      "loss": 0.3878,
+      "step": 3666
+    },
+    {
+      "epoch": 1.707169459962756,
+      "grad_norm": 0.5713449522771594,
+      "learning_rate": 2.1375175334487757e-06,
+      "loss": 0.3893,
+      "step": 3667
+    },
+    {
+      "epoch": 1.7076350093109869,
+      "grad_norm": 0.6379721302632294,
+      "learning_rate": 2.1362480197187805e-06,
+      "loss": 0.3831,
+      "step": 3668
+    },
+    {
+      "epoch": 1.708100558659218,
+      "grad_norm": 0.5890325004867958,
+      "learning_rate": 2.134978601809707e-06,
+      "loss": 0.3686,
+      "step": 3669
+    },
+    {
+      "epoch": 1.7085661080074488,
+      "grad_norm": 0.5952104480618,
+      "learning_rate": 2.133709280055949e-06,
+      "loss": 0.3807,
+      "step": 3670
+    },
+    {
+      "epoch": 1.7090316573556796,
+      "grad_norm": 0.6000175388958491,
+      "learning_rate": 2.1324400547918775e-06,
+      "loss": 0.3896,
+      "step": 3671
+    },
+    {
+      "epoch": 1.7094972067039107,
+      "grad_norm": 0.6061120274930506,
+      "learning_rate": 2.131170926351836e-06,
+      "loss": 0.3851,
+      "step": 3672
+    },
+    {
+      "epoch": 1.7099627560521415,
+      "grad_norm": 0.5988574504376282,
+      "learning_rate": 2.1299018950701427e-06,
+      "loss": 0.4078,
+      "step": 3673
+    },
+    {
+      "epoch": 1.7104283054003724,
+      "grad_norm": 0.5999788157678305,
+      "learning_rate": 2.128632961281091e-06,
+      "loss": 0.3814,
+      "step": 3674
+    },
+    {
+      "epoch": 1.7108938547486034,
+      "grad_norm": 0.5683804232010347,
+      "learning_rate": 2.1273641253189494e-06,
+      "loss": 0.3696,
+      "step": 3675
+    },
+    {
+      "epoch": 1.7113594040968343,
+      "grad_norm": 0.6221214244592613,
+      "learning_rate": 2.126095387517958e-06,
+      "loss": 0.3926,
+      "step": 3676
+    },
+    {
+      "epoch": 1.7118249534450651,
+      "grad_norm": 0.6198981221259542,
+      "learning_rate": 2.124826748212332e-06,
+      "loss": 0.3884,
+      "step": 3677
+    },
+    {
+      "epoch": 1.7122905027932962,
+      "grad_norm": 0.5921614937451734,
+      "learning_rate": 2.1235582077362626e-06,
+      "loss": 0.3741,
+      "step": 3678
+    },
+    {
+      "epoch": 1.712756052141527,
+      "grad_norm": 0.5634479842799585,
+      "learning_rate": 2.122289766423913e-06,
+      "loss": 0.3827,
+      "step": 3679
+    },
+    {
+      "epoch": 1.7132216014897579,
+      "grad_norm": 0.5707321278901942,
+      "learning_rate": 2.1210214246094206e-06,
+      "loss": 0.3888,
+      "step": 3680
+    },
+    {
+      "epoch": 1.713687150837989,
+      "grad_norm": 0.6170413882372131,
+      "learning_rate": 2.119753182626897e-06,
+      "loss": 0.377,
+      "step": 3681
+    },
+    {
+      "epoch": 1.7141527001862198,
+      "grad_norm": 0.6013865020439517,
+      "learning_rate": 2.1184850408104272e-06,
+      "loss": 0.3572,
+      "step": 3682
+    },
+    {
+      "epoch": 1.7146182495344506,
+      "grad_norm": 0.6090088561745863,
+      "learning_rate": 2.1172169994940707e-06,
+      "loss": 0.3799,
+      "step": 3683
+    },
+    {
+      "epoch": 1.7150837988826817,
+      "grad_norm": 0.5782099674483053,
+      "learning_rate": 2.1159490590118586e-06,
+      "loss": 0.3665,
+      "step": 3684
+    },
+    {
+      "epoch": 1.7155493482309123,
+      "grad_norm": 0.6283632116786227,
+      "learning_rate": 2.1146812196977955e-06,
+      "loss": 0.3839,
+      "step": 3685
+    },
+    {
+      "epoch": 1.7160148975791434,
+      "grad_norm": 0.6076751986260731,
+      "learning_rate": 2.1134134818858635e-06,
+      "loss": 0.4022,
+      "step": 3686
+    },
+    {
+      "epoch": 1.7164804469273744,
+      "grad_norm": 0.6361368987585748,
+      "learning_rate": 2.1121458459100135e-06,
+      "loss": 0.3898,
+      "step": 3687
+    },
+    {
+      "epoch": 1.716945996275605,
+      "grad_norm": 0.5699306710043516,
+      "learning_rate": 2.1108783121041703e-06,
+      "loss": 0.3573,
+      "step": 3688
+    },
+    {
+      "epoch": 1.7174115456238361,
+      "grad_norm": 0.5864252997393117,
+      "learning_rate": 2.109610880802232e-06,
+      "loss": 0.3725,
+      "step": 3689
+    },
+    {
+      "epoch": 1.7178770949720672,
+      "grad_norm": 0.5841673176276372,
+      "learning_rate": 2.108343552338072e-06,
+      "loss": 0.3841,
+      "step": 3690
+    },
+    {
+      "epoch": 1.7183426443202978,
+      "grad_norm": 0.5665622872584516,
+      "learning_rate": 2.1070763270455334e-06,
+      "loss": 0.3874,
+      "step": 3691
+    },
+    {
+      "epoch": 1.7188081936685289,
+      "grad_norm": 0.5923535962855636,
+      "learning_rate": 2.105809205258434e-06,
+      "loss": 0.3869,
+      "step": 3692
+    },
+    {
+      "epoch": 1.7192737430167597,
+      "grad_norm": 0.6224486129205867,
+      "learning_rate": 2.104542187310563e-06,
+      "loss": 0.386,
+      "step": 3693
+    },
+    {
+      "epoch": 1.7197392923649906,
+      "grad_norm": 0.6189117567226928,
+      "learning_rate": 2.1032752735356846e-06,
+      "loss": 0.3977,
+      "step": 3694
+    },
+    {
+      "epoch": 1.7202048417132216,
+      "grad_norm": 0.6237842131486024,
+      "learning_rate": 2.102008464267533e-06,
+      "loss": 0.3978,
+      "step": 3695
+    },
+    {
+      "epoch": 1.7206703910614525,
+      "grad_norm": 0.5954853773776775,
+      "learning_rate": 2.1007417598398163e-06,
+      "loss": 0.3878,
+      "step": 3696
+    },
+    {
+      "epoch": 1.7211359404096833,
+      "grad_norm": 0.5720791057277949,
+      "learning_rate": 2.0994751605862134e-06,
+      "loss": 0.3761,
+      "step": 3697
+    },
+    {
+      "epoch": 1.7216014897579144,
+      "grad_norm": 0.5935860062460919,
+      "learning_rate": 2.098208666840378e-06,
+      "loss": 0.3954,
+      "step": 3698
+    },
+    {
+      "epoch": 1.7220670391061452,
+      "grad_norm": 0.5850586784957205,
+      "learning_rate": 2.0969422789359346e-06,
+      "loss": 0.3717,
+      "step": 3699
+    },
+    {
+      "epoch": 1.722532588454376,
+      "grad_norm": 0.5554546165140247,
+      "learning_rate": 2.09567599720648e-06,
+      "loss": 0.3912,
+      "step": 3700
+    },
+    {
+      "epoch": 1.7229981378026071,
+      "grad_norm": 0.5949915699924089,
+      "learning_rate": 2.0944098219855806e-06,
+      "loss": 0.3755,
+      "step": 3701
+    },
+    {
+      "epoch": 1.723463687150838,
+      "grad_norm": 0.5904549081794311,
+      "learning_rate": 2.0931437536067802e-06,
+      "loss": 0.3967,
+      "step": 3702
+    },
+    {
+      "epoch": 1.7239292364990688,
+      "grad_norm": 0.6316298565072841,
+      "learning_rate": 2.09187779240359e-06,
+      "loss": 0.4037,
+      "step": 3703
+    },
+    {
+      "epoch": 1.7243947858472999,
+      "grad_norm": 0.5983863805535727,
+      "learning_rate": 2.0906119387094945e-06,
+      "loss": 0.3755,
+      "step": 3704
+    },
+    {
+      "epoch": 1.7248603351955307,
+      "grad_norm": 0.6054676227869443,
+      "learning_rate": 2.0893461928579477e-06,
+      "loss": 0.388,
+      "step": 3705
+    },
+    {
+      "epoch": 1.7253258845437616,
+      "grad_norm": 0.6106868646825214,
+      "learning_rate": 2.0880805551823813e-06,
+      "loss": 0.4051,
+      "step": 3706
+    },
+    {
+      "epoch": 1.7257914338919926,
+      "grad_norm": 0.5759970193343704,
+      "learning_rate": 2.0868150260161908e-06,
+      "loss": 0.3669,
+      "step": 3707
+    },
+    {
+      "epoch": 1.7262569832402235,
+      "grad_norm": 0.5655932833507887,
+      "learning_rate": 2.085549605692748e-06,
+      "loss": 0.3896,
+      "step": 3708
+    },
+    {
+      "epoch": 1.7267225325884543,
+      "grad_norm": 0.5764831039184842,
+      "learning_rate": 2.0842842945453944e-06,
+      "loss": 0.3875,
+      "step": 3709
+    },
+    {
+      "epoch": 1.7271880819366854,
+      "grad_norm": 0.587180013010843,
+      "learning_rate": 2.0830190929074434e-06,
+      "loss": 0.3816,
+      "step": 3710
+    },
+    {
+      "epoch": 1.7276536312849162,
+      "grad_norm": 0.5656582350700448,
+      "learning_rate": 2.081754001112179e-06,
+      "loss": 0.3525,
+      "step": 3711
+    },
+    {
+      "epoch": 1.728119180633147,
+      "grad_norm": 0.579157782455035,
+      "learning_rate": 2.0804890194928566e-06,
+      "loss": 0.3851,
+      "step": 3712
+    },
+    {
+      "epoch": 1.7285847299813781,
+      "grad_norm": 0.593832329198073,
+      "learning_rate": 2.0792241483827016e-06,
+      "loss": 0.3934,
+      "step": 3713
+    },
+    {
+      "epoch": 1.729050279329609,
+      "grad_norm": 0.5554084597007477,
+      "learning_rate": 2.0779593881149126e-06,
+      "loss": 0.3935,
+      "step": 3714
+    },
+    {
+      "epoch": 1.7295158286778398,
+      "grad_norm": 0.6078711710318623,
+      "learning_rate": 2.0766947390226565e-06,
+      "loss": 0.3934,
+      "step": 3715
+    },
+    {
+      "epoch": 1.7299813780260709,
+      "grad_norm": 0.6432189952357651,
+      "learning_rate": 2.075430201439072e-06,
+      "loss": 0.3915,
+      "step": 3716
+    },
+    {
+      "epoch": 1.7304469273743017,
+      "grad_norm": 0.5944049166015979,
+      "learning_rate": 2.0741657756972687e-06,
+      "loss": 0.4071,
+      "step": 3717
+    },
+    {
+      "epoch": 1.7309124767225326,
+      "grad_norm": 0.5884929021066863,
+      "learning_rate": 2.0729014621303267e-06,
+      "loss": 0.3907,
+      "step": 3718
+    },
+    {
+      "epoch": 1.7313780260707636,
+      "grad_norm": 0.6094615698711247,
+      "learning_rate": 2.0716372610712958e-06,
+      "loss": 0.3635,
+      "step": 3719
+    },
+    {
+      "epoch": 1.7318435754189943,
+      "grad_norm": 0.5639207621080111,
+      "learning_rate": 2.0703731728531957e-06,
+      "loss": 0.3935,
+      "step": 3720
+    },
+    {
+      "epoch": 1.7323091247672253,
+      "grad_norm": 0.5958519063400614,
+      "learning_rate": 2.069109197809019e-06,
+      "loss": 0.3766,
+      "step": 3721
+    },
+    {
+      "epoch": 1.7327746741154564,
+      "grad_norm": 0.599826761258893,
+      "learning_rate": 2.067845336271727e-06,
+      "loss": 0.3828,
+      "step": 3722
+    },
+    {
+      "epoch": 1.733240223463687,
+      "grad_norm": 0.5898349454828263,
+      "learning_rate": 2.066581588574249e-06,
+      "loss": 0.3843,
+      "step": 3723
+    },
+    {
+      "epoch": 1.733705772811918,
+      "grad_norm": 0.5968161972340951,
+      "learning_rate": 2.065317955049486e-06,
+      "loss": 0.3776,
+      "step": 3724
+    },
+    {
+      "epoch": 1.7341713221601491,
+      "grad_norm": 0.5818745725826934,
+      "learning_rate": 2.064054436030312e-06,
+      "loss": 0.3713,
+      "step": 3725
+    },
+    {
+      "epoch": 1.7346368715083798,
+      "grad_norm": 0.5895200420996903,
+      "learning_rate": 2.0627910318495657e-06,
+      "loss": 0.3755,
+      "step": 3726
+    },
+    {
+      "epoch": 1.7351024208566108,
+      "grad_norm": 0.6611449013712735,
+      "learning_rate": 2.061527742840058e-06,
+      "loss": 0.3694,
+      "step": 3727
+    },
+    {
+      "epoch": 1.7355679702048417,
+      "grad_norm": 0.625360775143754,
+      "learning_rate": 2.0602645693345697e-06,
+      "loss": 0.378,
+      "step": 3728
+    },
+    {
+      "epoch": 1.7360335195530725,
+      "grad_norm": 0.5693803348801276,
+      "learning_rate": 2.0590015116658502e-06,
+      "loss": 0.3924,
+      "step": 3729
+    },
+    {
+      "epoch": 1.7364990689013036,
+      "grad_norm": 0.5893145877301103,
+      "learning_rate": 2.0577385701666204e-06,
+      "loss": 0.3868,
+      "step": 3730
+    },
+    {
+      "epoch": 1.7369646182495344,
+      "grad_norm": 0.6149528012443197,
+      "learning_rate": 2.0564757451695672e-06,
+      "loss": 0.3857,
+      "step": 3731
+    },
+    {
+      "epoch": 1.7374301675977653,
+      "grad_norm": 0.583866242123322,
+      "learning_rate": 2.0552130370073493e-06,
+      "loss": 0.3681,
+      "step": 3732
+    },
+    {
+      "epoch": 1.7378957169459963,
+      "grad_norm": 0.6080039474692868,
+      "learning_rate": 2.053950446012595e-06,
+      "loss": 0.4112,
+      "step": 3733
+    },
+    {
+      "epoch": 1.7383612662942272,
+      "grad_norm": 0.5627537178182078,
+      "learning_rate": 2.0526879725179e-06,
+      "loss": 0.3741,
+      "step": 3734
+    },
+    {
+      "epoch": 1.738826815642458,
+      "grad_norm": 0.5886979540023355,
+      "learning_rate": 2.0514256168558304e-06,
+      "loss": 0.3635,
+      "step": 3735
+    },
+    {
+      "epoch": 1.739292364990689,
+      "grad_norm": 0.5915735076436358,
+      "learning_rate": 2.0501633793589193e-06,
+      "loss": 0.3866,
+      "step": 3736
+    },
+    {
+      "epoch": 1.73975791433892,
+      "grad_norm": 0.5834325076256367,
+      "learning_rate": 2.0489012603596713e-06,
+      "loss": 0.3841,
+      "step": 3737
+    },
+    {
+      "epoch": 1.7402234636871508,
+      "grad_norm": 0.6051525643561297,
+      "learning_rate": 2.047639260190559e-06,
+      "loss": 0.3861,
+      "step": 3738
+    },
+    {
+      "epoch": 1.7406890130353818,
+      "grad_norm": 0.6365207754858467,
+      "learning_rate": 2.0463773791840228e-06,
+      "loss": 0.401,
+      "step": 3739
+    },
+    {
+      "epoch": 1.7411545623836127,
+      "grad_norm": 0.6151043514439775,
+      "learning_rate": 2.04511561767247e-06,
+      "loss": 0.3896,
+      "step": 3740
+    },
+    {
+      "epoch": 1.7416201117318435,
+      "grad_norm": 0.6230914080169249,
+      "learning_rate": 2.043853975988282e-06,
+      "loss": 0.4037,
+      "step": 3741
+    },
+    {
+      "epoch": 1.7420856610800746,
+      "grad_norm": 0.6099658345389445,
+      "learning_rate": 2.0425924544638044e-06,
+      "loss": 0.3883,
+      "step": 3742
+    },
+    {
+      "epoch": 1.7425512104283054,
+      "grad_norm": 0.57665276075851,
+      "learning_rate": 2.0413310534313504e-06,
+      "loss": 0.3594,
+      "step": 3743
+    },
+    {
+      "epoch": 1.7430167597765363,
+      "grad_norm": 0.5774843714507805,
+      "learning_rate": 2.040069773223203e-06,
+      "loss": 0.3701,
+      "step": 3744
+    },
+    {
+      "epoch": 1.7434823091247673,
+      "grad_norm": 0.6070012720260877,
+      "learning_rate": 2.038808614171616e-06,
+      "loss": 0.4046,
+      "step": 3745
+    },
+    {
+      "epoch": 1.7439478584729982,
+      "grad_norm": 0.5984614737711169,
+      "learning_rate": 2.0375475766088064e-06,
+      "loss": 0.3794,
+      "step": 3746
+    },
+    {
+      "epoch": 1.744413407821229,
+      "grad_norm": 0.5927570112746946,
+      "learning_rate": 2.036286660866962e-06,
+      "loss": 0.3944,
+      "step": 3747
+    },
+    {
+      "epoch": 1.74487895716946,
+      "grad_norm": 0.662938202754001,
+      "learning_rate": 2.035025867278238e-06,
+      "loss": 0.4069,
+      "step": 3748
+    },
+    {
+      "epoch": 1.745344506517691,
+      "grad_norm": 0.6114611118591452,
+      "learning_rate": 2.0337651961747576e-06,
+      "loss": 0.3786,
+      "step": 3749
+    },
+    {
+      "epoch": 1.7458100558659218,
+      "grad_norm": 0.5954158767795932,
+      "learning_rate": 2.0325046478886117e-06,
+      "loss": 0.3789,
+      "step": 3750
+    },
+    {
+      "epoch": 1.7462756052141528,
+      "grad_norm": 0.5764372123046948,
+      "learning_rate": 2.031244222751858e-06,
+      "loss": 0.3769,
+      "step": 3751
+    },
+    {
+      "epoch": 1.7467411545623837,
+      "grad_norm": 0.5790506640207855,
+      "learning_rate": 2.029983921096523e-06,
+      "loss": 0.3945,
+      "step": 3752
+    },
+    {
+      "epoch": 1.7472067039106145,
+      "grad_norm": 0.6063592076609099,
+      "learning_rate": 2.0287237432546e-06,
+      "loss": 0.379,
+      "step": 3753
+    },
+    {
+      "epoch": 1.7476722532588456,
+      "grad_norm": 0.5691247184563027,
+      "learning_rate": 2.02746368955805e-06,
+      "loss": 0.3747,
+      "step": 3754
+    },
+    {
+      "epoch": 1.7481378026070762,
+      "grad_norm": 0.6022141513699936,
+      "learning_rate": 2.0262037603388007e-06,
+      "loss": 0.3842,
+      "step": 3755
+    },
+    {
+      "epoch": 1.7486033519553073,
+      "grad_norm": 0.6033974913194454,
+      "learning_rate": 2.0249439559287477e-06,
+      "loss": 0.3829,
+      "step": 3756
+    },
+    {
+      "epoch": 1.7490689013035383,
+      "grad_norm": 0.5988195365385932,
+      "learning_rate": 2.0236842766597535e-06,
+      "loss": 0.3883,
+      "step": 3757
+    },
+    {
+      "epoch": 1.749534450651769,
+      "grad_norm": 0.5842008109320291,
+      "learning_rate": 2.0224247228636475e-06,
+      "loss": 0.378,
+      "step": 3758
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6202893164510039,
+      "learning_rate": 2.0211652948722253e-06,
+      "loss": 0.3792,
+      "step": 3759
+    },
+    {
+      "epoch": 1.750465549348231,
+      "grad_norm": 0.5991652002336227,
+      "learning_rate": 2.019905993017252e-06,
+      "loss": 0.3752,
+      "step": 3760
+    },
+    {
+      "epoch": 1.7509310986964617,
+      "grad_norm": 0.5934860767878555,
+      "learning_rate": 2.0186468176304567e-06,
+      "loss": 0.3504,
+      "step": 3761
+    },
+    {
+      "epoch": 1.7513966480446927,
+      "grad_norm": 0.6075329130383296,
+      "learning_rate": 2.017387769043536e-06,
+      "loss": 0.3939,
+      "step": 3762
+    },
+    {
+      "epoch": 1.7518621973929238,
+      "grad_norm": 0.5941642453900886,
+      "learning_rate": 2.0161288475881523e-06,
+      "loss": 0.3924,
+      "step": 3763
+    },
+    {
+      "epoch": 1.7523277467411544,
+      "grad_norm": 0.6488067245152204,
+      "learning_rate": 2.0148700535959377e-06,
+      "loss": 0.3893,
+      "step": 3764
+    },
+    {
+      "epoch": 1.7527932960893855,
+      "grad_norm": 0.5997723192625141,
+      "learning_rate": 2.013611387398487e-06,
+      "loss": 0.3901,
+      "step": 3765
+    },
+    {
+      "epoch": 1.7532588454376163,
+      "grad_norm": 0.5973216900815483,
+      "learning_rate": 2.012352849327363e-06,
+      "loss": 0.3835,
+      "step": 3766
+    },
+    {
+      "epoch": 1.7537243947858472,
+      "grad_norm": 0.5918954010653268,
+      "learning_rate": 2.0110944397140936e-06,
+      "loss": 0.4065,
+      "step": 3767
+    },
+    {
+      "epoch": 1.7541899441340782,
+      "grad_norm": 0.5963543057310818,
+      "learning_rate": 2.0098361588901756e-06,
+      "loss": 0.3793,
+      "step": 3768
+    },
+    {
+      "epoch": 1.754655493482309,
+      "grad_norm": 0.5987247777645411,
+      "learning_rate": 2.008578007187069e-06,
+      "loss": 0.3763,
+      "step": 3769
+    },
+    {
+      "epoch": 1.75512104283054,
+      "grad_norm": 0.6209124887923186,
+      "learning_rate": 2.0073199849362017e-06,
+      "loss": 0.3964,
+      "step": 3770
+    },
+    {
+      "epoch": 1.755586592178771,
+      "grad_norm": 0.5938722182132969,
+      "learning_rate": 2.0060620924689656e-06,
+      "loss": 0.3692,
+      "step": 3771
+    },
+    {
+      "epoch": 1.7560521415270018,
+      "grad_norm": 0.60197827634131,
+      "learning_rate": 2.0048043301167204e-06,
+      "loss": 0.3777,
+      "step": 3772
+    },
+    {
+      "epoch": 1.7565176908752327,
+      "grad_norm": 0.602178962733013,
+      "learning_rate": 2.00354669821079e-06,
+      "loss": 0.4071,
+      "step": 3773
+    },
+    {
+      "epoch": 1.7569832402234637,
+      "grad_norm": 0.5730378526525891,
+      "learning_rate": 2.002289197082465e-06,
+      "loss": 0.3786,
+      "step": 3774
+    },
+    {
+      "epoch": 1.7574487895716946,
+      "grad_norm": 0.6286965116649483,
+      "learning_rate": 2.001031827063e-06,
+      "loss": 0.3697,
+      "step": 3775
+    },
+    {
+      "epoch": 1.7579143389199254,
+      "grad_norm": 0.5979181866276013,
+      "learning_rate": 1.9997745884836186e-06,
+      "loss": 0.3855,
+      "step": 3776
+    },
+    {
+      "epoch": 1.7583798882681565,
+      "grad_norm": 0.5645209916828006,
+      "learning_rate": 1.9985174816755048e-06,
+      "loss": 0.3687,
+      "step": 3777
+    },
+    {
+      "epoch": 1.7588454376163873,
+      "grad_norm": 0.590505110045678,
+      "learning_rate": 1.997260506969812e-06,
+      "loss": 0.3906,
+      "step": 3778
+    },
+    {
+      "epoch": 1.7593109869646182,
+      "grad_norm": 0.6040149508925605,
+      "learning_rate": 1.996003664697656e-06,
+      "loss": 0.3771,
+      "step": 3779
+    },
+    {
+      "epoch": 1.7597765363128492,
+      "grad_norm": 0.5675637855896988,
+      "learning_rate": 1.99474695519012e-06,
+      "loss": 0.3756,
+      "step": 3780
+    },
+    {
+      "epoch": 1.76024208566108,
+      "grad_norm": 0.5951785771401621,
+      "learning_rate": 1.993490378778251e-06,
+      "loss": 0.3694,
+      "step": 3781
+    },
+    {
+      "epoch": 1.760707635009311,
+      "grad_norm": 0.5957385755539352,
+      "learning_rate": 1.9922339357930606e-06,
+      "loss": 0.3879,
+      "step": 3782
+    },
+    {
+      "epoch": 1.761173184357542,
+      "grad_norm": 0.5493874614355424,
+      "learning_rate": 1.990977626565525e-06,
+      "loss": 0.3763,
+      "step": 3783
+    },
+    {
+      "epoch": 1.7616387337057728,
+      "grad_norm": 0.6237591959530403,
+      "learning_rate": 1.9897214514265884e-06,
+      "loss": 0.3796,
+      "step": 3784
+    },
+    {
+      "epoch": 1.7621042830540037,
+      "grad_norm": 0.611749668571598,
+      "learning_rate": 1.9884654107071554e-06,
+      "loss": 0.3908,
+      "step": 3785
+    },
+    {
+      "epoch": 1.7625698324022347,
+      "grad_norm": 0.5964871349627008,
+      "learning_rate": 1.9872095047380977e-06,
+      "loss": 0.3604,
+      "step": 3786
+    },
+    {
+      "epoch": 1.7630353817504656,
+      "grad_norm": 0.6085160324438352,
+      "learning_rate": 1.9859537338502495e-06,
+      "loss": 0.397,
+      "step": 3787
+    },
+    {
+      "epoch": 1.7635009310986964,
+      "grad_norm": 0.6169773353133288,
+      "learning_rate": 1.984698098374412e-06,
+      "loss": 0.3785,
+      "step": 3788
+    },
+    {
+      "epoch": 1.7639664804469275,
+      "grad_norm": 0.5895032267420106,
+      "learning_rate": 1.9834425986413497e-06,
+      "loss": 0.378,
+      "step": 3789
+    },
+    {
+      "epoch": 1.7644320297951583,
+      "grad_norm": 0.578813789898202,
+      "learning_rate": 1.98218723498179e-06,
+      "loss": 0.386,
+      "step": 3790
+    },
+    {
+      "epoch": 1.7648975791433892,
+      "grad_norm": 0.6007758379532855,
+      "learning_rate": 1.9809320077264253e-06,
+      "loss": 0.3647,
+      "step": 3791
+    },
+    {
+      "epoch": 1.7653631284916202,
+      "grad_norm": 0.6314558457051707,
+      "learning_rate": 1.9796769172059144e-06,
+      "loss": 0.3863,
+      "step": 3792
+    },
+    {
+      "epoch": 1.7658286778398509,
+      "grad_norm": 0.6085143572206659,
+      "learning_rate": 1.978421963750876e-06,
+      "loss": 0.381,
+      "step": 3793
+    },
+    {
+      "epoch": 1.766294227188082,
+      "grad_norm": 0.5770356478065,
+      "learning_rate": 1.9771671476918953e-06,
+      "loss": 0.3842,
+      "step": 3794
+    },
+    {
+      "epoch": 1.766759776536313,
+      "grad_norm": 0.5880684852453649,
+      "learning_rate": 1.9759124693595204e-06,
+      "loss": 0.3716,
+      "step": 3795
+    },
+    {
+      "epoch": 1.7672253258845436,
+      "grad_norm": 0.57361381577612,
+      "learning_rate": 1.974657929084264e-06,
+      "loss": 0.3988,
+      "step": 3796
+    },
+    {
+      "epoch": 1.7676908752327747,
+      "grad_norm": 0.6096245844008259,
+      "learning_rate": 1.973403527196602e-06,
+      "loss": 0.3772,
+      "step": 3797
+    },
+    {
+      "epoch": 1.7681564245810057,
+      "grad_norm": 0.5879533581349331,
+      "learning_rate": 1.9721492640269735e-06,
+      "loss": 0.376,
+      "step": 3798
+    },
+    {
+      "epoch": 1.7686219739292364,
+      "grad_norm": 0.5771105758245507,
+      "learning_rate": 1.9708951399057792e-06,
+      "loss": 0.3747,
+      "step": 3799
+    },
+    {
+      "epoch": 1.7690875232774674,
+      "grad_norm": 0.5780285401209765,
+      "learning_rate": 1.9696411551633885e-06,
+      "loss": 0.3616,
+      "step": 3800
+    },
+    {
+      "epoch": 1.7695530726256983,
+      "grad_norm": 0.5887155500947641,
+      "learning_rate": 1.9683873101301297e-06,
+      "loss": 0.3726,
+      "step": 3801
+    },
+    {
+      "epoch": 1.7700186219739291,
+      "grad_norm": 0.5747720845248138,
+      "learning_rate": 1.967133605136293e-06,
+      "loss": 0.3917,
+      "step": 3802
+    },
+    {
+      "epoch": 1.7704841713221602,
+      "grad_norm": 0.5883312651140105,
+      "learning_rate": 1.9658800405121383e-06,
+      "loss": 0.3839,
+      "step": 3803
+    },
+    {
+      "epoch": 1.770949720670391,
+      "grad_norm": 0.6153590033204398,
+      "learning_rate": 1.964626616587881e-06,
+      "loss": 0.3789,
+      "step": 3804
+    },
+    {
+      "epoch": 1.7714152700186219,
+      "grad_norm": 0.6212537820691151,
+      "learning_rate": 1.963373333693704e-06,
+      "loss": 0.3936,
+      "step": 3805
+    },
+    {
+      "epoch": 1.771880819366853,
+      "grad_norm": 0.5806160974515936,
+      "learning_rate": 1.9621201921597515e-06,
+      "loss": 0.3886,
+      "step": 3806
+    },
+    {
+      "epoch": 1.7723463687150838,
+      "grad_norm": 0.6048583779824376,
+      "learning_rate": 1.9608671923161315e-06,
+      "loss": 0.3859,
+      "step": 3807
+    },
+    {
+      "epoch": 1.7728119180633146,
+      "grad_norm": 0.5983453876312675,
+      "learning_rate": 1.9596143344929134e-06,
+      "loss": 0.3805,
+      "step": 3808
+    },
+    {
+      "epoch": 1.7732774674115457,
+      "grad_norm": 0.5869814076819339,
+      "learning_rate": 1.9583616190201297e-06,
+      "loss": 0.3976,
+      "step": 3809
+    },
+    {
+      "epoch": 1.7737430167597765,
+      "grad_norm": 0.5924044139785919,
+      "learning_rate": 1.957109046227775e-06,
+      "loss": 0.3787,
+      "step": 3810
+    },
+    {
+      "epoch": 1.7742085661080074,
+      "grad_norm": 0.6729668243854295,
+      "learning_rate": 1.9558566164458078e-06,
+      "loss": 0.388,
+      "step": 3811
+    },
+    {
+      "epoch": 1.7746741154562384,
+      "grad_norm": 0.6164339802627019,
+      "learning_rate": 1.9546043300041476e-06,
+      "loss": 0.4116,
+      "step": 3812
+    },
+    {
+      "epoch": 1.7751396648044693,
+      "grad_norm": 0.6138781456016069,
+      "learning_rate": 1.9533521872326762e-06,
+      "loss": 0.3748,
+      "step": 3813
+    },
+    {
+      "epoch": 1.7756052141527001,
+      "grad_norm": 0.617371713574641,
+      "learning_rate": 1.952100188461237e-06,
+      "loss": 0.4026,
+      "step": 3814
+    },
+    {
+      "epoch": 1.7760707635009312,
+      "grad_norm": 0.5741384182242565,
+      "learning_rate": 1.950848334019638e-06,
+      "loss": 0.3861,
+      "step": 3815
+    },
+    {
+      "epoch": 1.776536312849162,
+      "grad_norm": 0.5960114360235533,
+      "learning_rate": 1.9495966242376457e-06,
+      "loss": 0.3866,
+      "step": 3816
+    },
+    {
+      "epoch": 1.7770018621973929,
+      "grad_norm": 0.568370240651758,
+      "learning_rate": 1.9483450594449923e-06,
+      "loss": 0.3905,
+      "step": 3817
+    },
+    {
+      "epoch": 1.777467411545624,
+      "grad_norm": 0.629637605554144,
+      "learning_rate": 1.9470936399713662e-06,
+      "loss": 0.3807,
+      "step": 3818
+    },
+    {
+      "epoch": 1.7779329608938548,
+      "grad_norm": 0.618291685171744,
+      "learning_rate": 1.9458423661464247e-06,
+      "loss": 0.3873,
+      "step": 3819
+    },
+    {
+      "epoch": 1.7783985102420856,
+      "grad_norm": 0.5811042087555964,
+      "learning_rate": 1.9445912382997825e-06,
+      "loss": 0.3878,
+      "step": 3820
+    },
+    {
+      "epoch": 1.7788640595903167,
+      "grad_norm": 0.6019170243695685,
+      "learning_rate": 1.943340256761015e-06,
+      "loss": 0.3679,
+      "step": 3821
+    },
+    {
+      "epoch": 1.7793296089385475,
+      "grad_norm": 0.6023797020659,
+      "learning_rate": 1.9420894218596593e-06,
+      "loss": 0.38,
+      "step": 3822
+    },
+    {
+      "epoch": 1.7797951582867784,
+      "grad_norm": 0.599762098134045,
+      "learning_rate": 1.940838733925219e-06,
+      "loss": 0.3841,
+      "step": 3823
+    },
+    {
+      "epoch": 1.7802607076350094,
+      "grad_norm": 0.6088409829704594,
+      "learning_rate": 1.939588193287152e-06,
+      "loss": 0.3923,
+      "step": 3824
+    },
+    {
+      "epoch": 1.7807262569832403,
+      "grad_norm": 0.6222037333560263,
+      "learning_rate": 1.9383378002748817e-06,
+      "loss": 0.3776,
+      "step": 3825
+    },
+    {
+      "epoch": 1.7811918063314711,
+      "grad_norm": 0.5817649593265566,
+      "learning_rate": 1.9370875552177903e-06,
+      "loss": 0.3807,
+      "step": 3826
+    },
+    {
+      "epoch": 1.7816573556797022,
+      "grad_norm": 0.5607956192430237,
+      "learning_rate": 1.935837458445223e-06,
+      "loss": 0.3882,
+      "step": 3827
+    },
+    {
+      "epoch": 1.7821229050279328,
+      "grad_norm": 0.595806575029281,
+      "learning_rate": 1.934587510286485e-06,
+      "loss": 0.3862,
+      "step": 3828
+    },
+    {
+      "epoch": 1.7825884543761639,
+      "grad_norm": 0.6060114353340733,
+      "learning_rate": 1.933337711070842e-06,
+      "loss": 0.38,
+      "step": 3829
+    },
+    {
+      "epoch": 1.783054003724395,
+      "grad_norm": 0.6101567213844055,
+      "learning_rate": 1.9320880611275207e-06,
+      "loss": 0.3694,
+      "step": 3830
+    },
+    {
+      "epoch": 1.7835195530726256,
+      "grad_norm": 0.6213453590593754,
+      "learning_rate": 1.930838560785709e-06,
+      "loss": 0.3972,
+      "step": 3831
+    },
+    {
+      "epoch": 1.7839851024208566,
+      "grad_norm": 0.5924581556963717,
+      "learning_rate": 1.929589210374555e-06,
+      "loss": 0.3896,
+      "step": 3832
+    },
+    {
+      "epoch": 1.7844506517690877,
+      "grad_norm": 0.6057035508032956,
+      "learning_rate": 1.9283400102231672e-06,
+      "loss": 0.3865,
+      "step": 3833
+    },
+    {
+      "epoch": 1.7849162011173183,
+      "grad_norm": 0.567431103007713,
+      "learning_rate": 1.927090960660614e-06,
+      "loss": 0.3724,
+      "step": 3834
+    },
+    {
+      "epoch": 1.7853817504655494,
+      "grad_norm": 0.5794315483124438,
+      "learning_rate": 1.925842062015926e-06,
+      "loss": 0.3755,
+      "step": 3835
+    },
+    {
+      "epoch": 1.7858472998137802,
+      "grad_norm": 0.5683786825735127,
+      "learning_rate": 1.9245933146180925e-06,
+      "loss": 0.3728,
+      "step": 3836
+    },
+    {
+      "epoch": 1.786312849162011,
+      "grad_norm": 0.6241663928376098,
+      "learning_rate": 1.923344718796063e-06,
+      "loss": 0.375,
+      "step": 3837
+    },
+    {
+      "epoch": 1.7867783985102421,
+      "grad_norm": 0.5697423485146706,
+      "learning_rate": 1.922096274878746e-06,
+      "loss": 0.3883,
+      "step": 3838
+    },
+    {
+      "epoch": 1.787243947858473,
+      "grad_norm": 0.6263629287977779,
+      "learning_rate": 1.9208479831950132e-06,
+      "loss": 0.3912,
+      "step": 3839
+    },
+    {
+      "epoch": 1.7877094972067038,
+      "grad_norm": 0.5812537076823235,
+      "learning_rate": 1.9195998440736945e-06,
+      "loss": 0.3827,
+      "step": 3840
+    },
+    {
+      "epoch": 1.7881750465549349,
+      "grad_norm": 0.5943881898319097,
+      "learning_rate": 1.9183518578435778e-06,
+      "loss": 0.3865,
+      "step": 3841
+    },
+    {
+      "epoch": 1.7886405959031657,
+      "grad_norm": 0.5993025649272914,
+      "learning_rate": 1.917104024833413e-06,
+      "loss": 0.3888,
+      "step": 3842
+    },
+    {
+      "epoch": 1.7891061452513966,
+      "grad_norm": 0.6028082028840628,
+      "learning_rate": 1.9158563453719094e-06,
+      "loss": 0.3719,
+      "step": 3843
+    },
+    {
+      "epoch": 1.7895716945996276,
+      "grad_norm": 0.5803803995016876,
+      "learning_rate": 1.9146088197877353e-06,
+      "loss": 0.3711,
+      "step": 3844
+    },
+    {
+      "epoch": 1.7900372439478585,
+      "grad_norm": 0.6305574861124825,
+      "learning_rate": 1.9133614484095185e-06,
+      "loss": 0.3965,
+      "step": 3845
+    },
+    {
+      "epoch": 1.7905027932960893,
+      "grad_norm": 0.5648483358205459,
+      "learning_rate": 1.9121142315658455e-06,
+      "loss": 0.3834,
+      "step": 3846
+    },
+    {
+      "epoch": 1.7909683426443204,
+      "grad_norm": 0.5582779242392825,
+      "learning_rate": 1.9108671695852646e-06,
+      "loss": 0.3697,
+      "step": 3847
+    },
+    {
+      "epoch": 1.7914338919925512,
+      "grad_norm": 0.5893957859517256,
+      "learning_rate": 1.9096202627962803e-06,
+      "loss": 0.3959,
+      "step": 3848
+    },
+    {
+      "epoch": 1.791899441340782,
+      "grad_norm": 0.595654731586911,
+      "learning_rate": 1.9083735115273576e-06,
+      "loss": 0.3692,
+      "step": 3849
+    },
+    {
+      "epoch": 1.7923649906890131,
+      "grad_norm": 0.5869989568033425,
+      "learning_rate": 1.907126916106921e-06,
+      "loss": 0.3892,
+      "step": 3850
+    },
+    {
+      "epoch": 1.792830540037244,
+      "grad_norm": 0.5786534880154511,
+      "learning_rate": 1.9058804768633527e-06,
+      "loss": 0.3716,
+      "step": 3851
+    },
+    {
+      "epoch": 1.7932960893854748,
+      "grad_norm": 0.5844505950269081,
+      "learning_rate": 1.9046341941249953e-06,
+      "loss": 0.3705,
+      "step": 3852
+    },
+    {
+      "epoch": 1.7937616387337059,
+      "grad_norm": 0.5936895879723694,
+      "learning_rate": 1.9033880682201479e-06,
+      "loss": 0.3875,
+      "step": 3853
+    },
+    {
+      "epoch": 1.7942271880819367,
+      "grad_norm": 0.5769532683211586,
+      "learning_rate": 1.902142099477071e-06,
+      "loss": 0.3838,
+      "step": 3854
+    },
+    {
+      "epoch": 1.7946927374301676,
+      "grad_norm": 0.5567744324032788,
+      "learning_rate": 1.9008962882239824e-06,
+      "loss": 0.3792,
+      "step": 3855
+    },
+    {
+      "epoch": 1.7951582867783986,
+      "grad_norm": 0.6012110463546436,
+      "learning_rate": 1.8996506347890583e-06,
+      "loss": 0.3676,
+      "step": 3856
+    },
+    {
+      "epoch": 1.7956238361266295,
+      "grad_norm": 0.6043005203114423,
+      "learning_rate": 1.8984051395004315e-06,
+      "loss": 0.3824,
+      "step": 3857
+    },
+    {
+      "epoch": 1.7960893854748603,
+      "grad_norm": 0.6068669887279361,
+      "learning_rate": 1.8971598026861978e-06,
+      "loss": 0.3812,
+      "step": 3858
+    },
+    {
+      "epoch": 1.7965549348230914,
+      "grad_norm": 0.5924506417708332,
+      "learning_rate": 1.895914624674408e-06,
+      "loss": 0.3902,
+      "step": 3859
+    },
+    {
+      "epoch": 1.7970204841713222,
+      "grad_norm": 0.6323303404162091,
+      "learning_rate": 1.8946696057930706e-06,
+      "loss": 0.3911,
+      "step": 3860
+    },
+    {
+      "epoch": 1.797486033519553,
+      "grad_norm": 0.579548632925977,
+      "learning_rate": 1.8934247463701525e-06,
+      "loss": 0.3841,
+      "step": 3861
+    },
+    {
+      "epoch": 1.7979515828677841,
+      "grad_norm": 0.6187836529447736,
+      "learning_rate": 1.8921800467335817e-06,
+      "loss": 0.3982,
+      "step": 3862
+    },
+    {
+      "epoch": 1.7984171322160147,
+      "grad_norm": 0.5917070992949576,
+      "learning_rate": 1.89093550721124e-06,
+      "loss": 0.3768,
+      "step": 3863
+    },
+    {
+      "epoch": 1.7988826815642458,
+      "grad_norm": 0.6315471121689229,
+      "learning_rate": 1.889691128130969e-06,
+      "loss": 0.3842,
+      "step": 3864
+    },
+    {
+      "epoch": 1.7993482309124769,
+      "grad_norm": 0.622136718430451,
+      "learning_rate": 1.8884469098205674e-06,
+      "loss": 0.3911,
+      "step": 3865
+    },
+    {
+      "epoch": 1.7998137802607075,
+      "grad_norm": 0.5850700648652273,
+      "learning_rate": 1.8872028526077928e-06,
+      "loss": 0.4065,
+      "step": 3866
+    },
+    {
+      "epoch": 1.8002793296089385,
+      "grad_norm": 0.5974657046394676,
+      "learning_rate": 1.885958956820359e-06,
+      "loss": 0.396,
+      "step": 3867
+    },
+    {
+      "epoch": 1.8007448789571696,
+      "grad_norm": 0.590925084069316,
+      "learning_rate": 1.8847152227859374e-06,
+      "loss": 0.3808,
+      "step": 3868
+    },
+    {
+      "epoch": 1.8012104283054002,
+      "grad_norm": 0.5729755579989204,
+      "learning_rate": 1.883471650832157e-06,
+      "loss": 0.3697,
+      "step": 3869
+    },
+    {
+      "epoch": 1.8016759776536313,
+      "grad_norm": 0.5979223711467743,
+      "learning_rate": 1.8822282412866053e-06,
+      "loss": 0.3909,
+      "step": 3870
+    },
+    {
+      "epoch": 1.8021415270018621,
+      "grad_norm": 0.5795606719865596,
+      "learning_rate": 1.8809849944768252e-06,
+      "loss": 0.3902,
+      "step": 3871
+    },
+    {
+      "epoch": 1.802607076350093,
+      "grad_norm": 0.5985714839402975,
+      "learning_rate": 1.8797419107303175e-06,
+      "loss": 0.3666,
+      "step": 3872
+    },
+    {
+      "epoch": 1.803072625698324,
+      "grad_norm": 0.5807962885723424,
+      "learning_rate": 1.8784989903745396e-06,
+      "loss": 0.3784,
+      "step": 3873
+    },
+    {
+      "epoch": 1.803538175046555,
+      "grad_norm": 0.6228005461211485,
+      "learning_rate": 1.8772562337369072e-06,
+      "loss": 0.3865,
+      "step": 3874
+    },
+    {
+      "epoch": 1.8040037243947857,
+      "grad_norm": 0.5900754532127226,
+      "learning_rate": 1.8760136411447913e-06,
+      "loss": 0.4148,
+      "step": 3875
+    },
+    {
+      "epoch": 1.8044692737430168,
+      "grad_norm": 0.6053727249288743,
+      "learning_rate": 1.8747712129255214e-06,
+      "loss": 0.377,
+      "step": 3876
+    },
+    {
+      "epoch": 1.8049348230912476,
+      "grad_norm": 0.5914500091056365,
+      "learning_rate": 1.87352894940638e-06,
+      "loss": 0.3873,
+      "step": 3877
+    },
+    {
+      "epoch": 1.8054003724394785,
+      "grad_norm": 0.5868529242930925,
+      "learning_rate": 1.8722868509146116e-06,
+      "loss": 0.3776,
+      "step": 3878
+    },
+    {
+      "epoch": 1.8058659217877095,
+      "grad_norm": 0.6281115246022627,
+      "learning_rate": 1.8710449177774142e-06,
+      "loss": 0.3931,
+      "step": 3879
+    },
+    {
+      "epoch": 1.8063314711359404,
+      "grad_norm": 0.6039447946753296,
+      "learning_rate": 1.869803150321941e-06,
+      "loss": 0.4114,
+      "step": 3880
+    },
+    {
+      "epoch": 1.8067970204841712,
+      "grad_norm": 0.5814787272356107,
+      "learning_rate": 1.8685615488753036e-06,
+      "loss": 0.3775,
+      "step": 3881
+    },
+    {
+      "epoch": 1.8072625698324023,
+      "grad_norm": 0.5714782206867072,
+      "learning_rate": 1.8673201137645702e-06,
+      "loss": 0.3667,
+      "step": 3882
+    },
+    {
+      "epoch": 1.8077281191806331,
+      "grad_norm": 0.5903481426214621,
+      "learning_rate": 1.8660788453167637e-06,
+      "loss": 0.3707,
+      "step": 3883
+    },
+    {
+      "epoch": 1.808193668528864,
+      "grad_norm": 0.611099272561059,
+      "learning_rate": 1.8648377438588638e-06,
+      "loss": 0.3862,
+      "step": 3884
+    },
+    {
+      "epoch": 1.808659217877095,
+      "grad_norm": 0.5793766278889244,
+      "learning_rate": 1.8635968097178054e-06,
+      "loss": 0.3784,
+      "step": 3885
+    },
+    {
+      "epoch": 1.809124767225326,
+      "grad_norm": 0.6252943929248876,
+      "learning_rate": 1.8623560432204815e-06,
+      "loss": 0.3863,
+      "step": 3886
+    },
+    {
+      "epoch": 1.8095903165735567,
+      "grad_norm": 0.5913932888810463,
+      "learning_rate": 1.8611154446937388e-06,
+      "loss": 0.3947,
+      "step": 3887
+    },
+    {
+      "epoch": 1.8100558659217878,
+      "grad_norm": 0.6089849046139202,
+      "learning_rate": 1.8598750144643805e-06,
+      "loss": 0.397,
+      "step": 3888
+    },
+    {
+      "epoch": 1.8105214152700186,
+      "grad_norm": 0.58098422611407,
+      "learning_rate": 1.8586347528591647e-06,
+      "loss": 0.3749,
+      "step": 3889
+    },
+    {
+      "epoch": 1.8109869646182495,
+      "grad_norm": 0.6380873001671531,
+      "learning_rate": 1.8573946602048072e-06,
+      "loss": 0.3872,
+      "step": 3890
+    },
+    {
+      "epoch": 1.8114525139664805,
+      "grad_norm": 0.5850170985226855,
+      "learning_rate": 1.856154736827977e-06,
+      "loss": 0.3885,
+      "step": 3891
+    },
+    {
+      "epoch": 1.8119180633147114,
+      "grad_norm": 0.6295982522540455,
+      "learning_rate": 1.8549149830552988e-06,
+      "loss": 0.3731,
+      "step": 3892
+    },
+    {
+      "epoch": 1.8123836126629422,
+      "grad_norm": 0.5457724606128235,
+      "learning_rate": 1.8536753992133548e-06,
+      "loss": 0.3773,
+      "step": 3893
+    },
+    {
+      "epoch": 1.8128491620111733,
+      "grad_norm": 0.6371690729365209,
+      "learning_rate": 1.85243598562868e-06,
+      "loss": 0.3753,
+      "step": 3894
+    },
+    {
+      "epoch": 1.8133147113594041,
+      "grad_norm": 0.6093891732017063,
+      "learning_rate": 1.851196742627766e-06,
+      "loss": 0.3859,
+      "step": 3895
+    },
+    {
+      "epoch": 1.813780260707635,
+      "grad_norm": 0.5936521489217733,
+      "learning_rate": 1.8499576705370566e-06,
+      "loss": 0.3914,
+      "step": 3896
+    },
+    {
+      "epoch": 1.814245810055866,
+      "grad_norm": 0.5840173286282059,
+      "learning_rate": 1.8487187696829559e-06,
+      "loss": 0.3979,
+      "step": 3897
+    },
+    {
+      "epoch": 1.8147113594040967,
+      "grad_norm": 0.6099714158517487,
+      "learning_rate": 1.847480040391819e-06,
+      "loss": 0.383,
+      "step": 3898
+    },
+    {
+      "epoch": 1.8151769087523277,
+      "grad_norm": 0.6086287307772131,
+      "learning_rate": 1.8462414829899555e-06,
+      "loss": 0.3794,
+      "step": 3899
+    },
+    {
+      "epoch": 1.8156424581005588,
+      "grad_norm": 0.5678895491197445,
+      "learning_rate": 1.8450030978036304e-06,
+      "loss": 0.3793,
+      "step": 3900
+    },
+    {
+      "epoch": 1.8161080074487894,
+      "grad_norm": 0.586108741378151,
+      "learning_rate": 1.8437648851590667e-06,
+      "loss": 0.3731,
+      "step": 3901
+    },
+    {
+      "epoch": 1.8165735567970205,
+      "grad_norm": 0.5772358987501989,
+      "learning_rate": 1.8425268453824369e-06,
+      "loss": 0.3918,
+      "step": 3902
+    },
+    {
+      "epoch": 1.8170391061452515,
+      "grad_norm": 0.6060730464359514,
+      "learning_rate": 1.8412889787998705e-06,
+      "loss": 0.3673,
+      "step": 3903
+    },
+    {
+      "epoch": 1.8175046554934822,
+      "grad_norm": 0.5739007348074339,
+      "learning_rate": 1.8400512857374504e-06,
+      "loss": 0.3772,
+      "step": 3904
+    },
+    {
+      "epoch": 1.8179702048417132,
+      "grad_norm": 0.6295299440868951,
+      "learning_rate": 1.8388137665212154e-06,
+      "loss": 0.3841,
+      "step": 3905
+    },
+    {
+      "epoch": 1.8184357541899443,
+      "grad_norm": 0.5808394128482632,
+      "learning_rate": 1.8375764214771575e-06,
+      "loss": 0.3846,
+      "step": 3906
+    },
+    {
+      "epoch": 1.818901303538175,
+      "grad_norm": 0.635727418757039,
+      "learning_rate": 1.8363392509312223e-06,
+      "loss": 0.3801,
+      "step": 3907
+    },
+    {
+      "epoch": 1.819366852886406,
+      "grad_norm": 0.5448104122895449,
+      "learning_rate": 1.8351022552093089e-06,
+      "loss": 0.3763,
+      "step": 3908
+    },
+    {
+      "epoch": 1.8198324022346368,
+      "grad_norm": 0.6326120995714051,
+      "learning_rate": 1.8338654346372736e-06,
+      "loss": 0.3937,
+      "step": 3909
+    },
+    {
+      "epoch": 1.8202979515828677,
+      "grad_norm": 0.6034291085394712,
+      "learning_rate": 1.8326287895409235e-06,
+      "loss": 0.3828,
+      "step": 3910
+    },
+    {
+      "epoch": 1.8207635009310987,
+      "grad_norm": 0.666623069904567,
+      "learning_rate": 1.83139232024602e-06,
+      "loss": 0.3744,
+      "step": 3911
+    },
+    {
+      "epoch": 1.8212290502793296,
+      "grad_norm": 0.5902750086534783,
+      "learning_rate": 1.8301560270782778e-06,
+      "loss": 0.3986,
+      "step": 3912
+    },
+    {
+      "epoch": 1.8216945996275604,
+      "grad_norm": 0.6105983750761604,
+      "learning_rate": 1.8289199103633676e-06,
+      "loss": 0.3907,
+      "step": 3913
+    },
+    {
+      "epoch": 1.8221601489757915,
+      "grad_norm": 0.5994958800692454,
+      "learning_rate": 1.8276839704269105e-06,
+      "loss": 0.3846,
+      "step": 3914
+    },
+    {
+      "epoch": 1.8226256983240223,
+      "grad_norm": 0.5934074448267825,
+      "learning_rate": 1.8264482075944844e-06,
+      "loss": 0.3789,
+      "step": 3915
+    },
+    {
+      "epoch": 1.8230912476722532,
+      "grad_norm": 0.591453363892912,
+      "learning_rate": 1.8252126221916153e-06,
+      "loss": 0.3974,
+      "step": 3916
+    },
+    {
+      "epoch": 1.8235567970204842,
+      "grad_norm": 0.5949750984695455,
+      "learning_rate": 1.8239772145437888e-06,
+      "loss": 0.3874,
+      "step": 3917
+    },
+    {
+      "epoch": 1.824022346368715,
+      "grad_norm": 0.6069242586425747,
+      "learning_rate": 1.8227419849764399e-06,
+      "loss": 0.3812,
+      "step": 3918
+    },
+    {
+      "epoch": 1.824487895716946,
+      "grad_norm": 0.5950049157751205,
+      "learning_rate": 1.8215069338149572e-06,
+      "loss": 0.3768,
+      "step": 3919
+    },
+    {
+      "epoch": 1.824953445065177,
+      "grad_norm": 0.5763371902523845,
+      "learning_rate": 1.8202720613846814e-06,
+      "loss": 0.3764,
+      "step": 3920
+    },
+    {
+      "epoch": 1.8254189944134078,
+      "grad_norm": 0.5914453581668594,
+      "learning_rate": 1.819037368010909e-06,
+      "loss": 0.3822,
+      "step": 3921
+    },
+    {
+      "epoch": 1.8258845437616387,
+      "grad_norm": 0.5707188308626211,
+      "learning_rate": 1.817802854018887e-06,
+      "loss": 0.3884,
+      "step": 3922
+    },
+    {
+      "epoch": 1.8263500931098697,
+      "grad_norm": 0.6052083575639027,
+      "learning_rate": 1.8165685197338157e-06,
+      "loss": 0.3786,
+      "step": 3923
+    },
+    {
+      "epoch": 1.8268156424581006,
+      "grad_norm": 0.5749661706429003,
+      "learning_rate": 1.8153343654808475e-06,
+      "loss": 0.4052,
+      "step": 3924
+    },
+    {
+      "epoch": 1.8272811918063314,
+      "grad_norm": 0.641083212109998,
+      "learning_rate": 1.8141003915850895e-06,
+      "loss": 0.3991,
+      "step": 3925
+    },
+    {
+      "epoch": 1.8277467411545625,
+      "grad_norm": 0.5980697685103125,
+      "learning_rate": 1.8128665983715984e-06,
+      "loss": 0.3782,
+      "step": 3926
+    },
+    {
+      "epoch": 1.8282122905027933,
+      "grad_norm": 0.6245257625998296,
+      "learning_rate": 1.8116329861653853e-06,
+      "loss": 0.3697,
+      "step": 3927
+    },
+    {
+      "epoch": 1.8286778398510242,
+      "grad_norm": 0.6013156699886615,
+      "learning_rate": 1.8103995552914122e-06,
+      "loss": 0.375,
+      "step": 3928
+    },
+    {
+      "epoch": 1.8291433891992552,
+      "grad_norm": 0.5886733721796977,
+      "learning_rate": 1.8091663060745952e-06,
+      "loss": 0.3906,
+      "step": 3929
+    },
+    {
+      "epoch": 1.829608938547486,
+      "grad_norm": 0.6193079557267547,
+      "learning_rate": 1.8079332388398007e-06,
+      "loss": 0.3853,
+      "step": 3930
+    },
+    {
+      "epoch": 1.830074487895717,
+      "grad_norm": 0.5614048320239449,
+      "learning_rate": 1.8067003539118493e-06,
+      "loss": 0.369,
+      "step": 3931
+    },
+    {
+      "epoch": 1.830540037243948,
+      "grad_norm": 0.6183323489400652,
+      "learning_rate": 1.8054676516155096e-06,
+      "loss": 0.3985,
+      "step": 3932
+    },
+    {
+      "epoch": 1.8310055865921788,
+      "grad_norm": 0.5687473815820344,
+      "learning_rate": 1.8042351322755068e-06,
+      "loss": 0.383,
+      "step": 3933
+    },
+    {
+      "epoch": 1.8314711359404097,
+      "grad_norm": 0.5984355149528795,
+      "learning_rate": 1.803002796216516e-06,
+      "loss": 0.3796,
+      "step": 3934
+    },
+    {
+      "epoch": 1.8319366852886407,
+      "grad_norm": 0.6139244877337328,
+      "learning_rate": 1.8017706437631612e-06,
+      "loss": 0.3713,
+      "step": 3935
+    },
+    {
+      "epoch": 1.8324022346368714,
+      "grad_norm": 0.5641271244062374,
+      "learning_rate": 1.800538675240024e-06,
+      "loss": 0.3857,
+      "step": 3936
+    },
+    {
+      "epoch": 1.8328677839851024,
+      "grad_norm": 0.6070319483586712,
+      "learning_rate": 1.799306890971633e-06,
+      "loss": 0.3765,
+      "step": 3937
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 0.6538593317119651,
+      "learning_rate": 1.7980752912824686e-06,
+      "loss": 0.3771,
+      "step": 3938
+    },
+    {
+      "epoch": 1.833798882681564,
+      "grad_norm": 0.6180422087630669,
+      "learning_rate": 1.7968438764969631e-06,
+      "loss": 0.409,
+      "step": 3939
+    },
+    {
+      "epoch": 1.8342644320297952,
+      "grad_norm": 0.598128919134499,
+      "learning_rate": 1.795612646939503e-06,
+      "loss": 0.3914,
+      "step": 3940
+    },
+    {
+      "epoch": 1.8347299813780262,
+      "grad_norm": 0.5602362684111526,
+      "learning_rate": 1.794381602934421e-06,
+      "loss": 0.3776,
+      "step": 3941
+    },
+    {
+      "epoch": 1.8351955307262569,
+      "grad_norm": 0.6201524076093301,
+      "learning_rate": 1.7931507448060042e-06,
+      "loss": 0.4023,
+      "step": 3942
+    },
+    {
+      "epoch": 1.835661080074488,
+      "grad_norm": 0.5897361668551656,
+      "learning_rate": 1.7919200728784898e-06,
+      "loss": 0.3699,
+      "step": 3943
+    },
+    {
+      "epoch": 1.8361266294227188,
+      "grad_norm": 0.6071458923862663,
+      "learning_rate": 1.7906895874760663e-06,
+      "loss": 0.3861,
+      "step": 3944
+    },
+    {
+      "epoch": 1.8365921787709496,
+      "grad_norm": 0.6077314239188907,
+      "learning_rate": 1.7894592889228726e-06,
+      "loss": 0.3879,
+      "step": 3945
+    },
+    {
+      "epoch": 1.8370577281191807,
+      "grad_norm": 0.5963080321722283,
+      "learning_rate": 1.7882291775429989e-06,
+      "loss": 0.3794,
+      "step": 3946
+    },
+    {
+      "epoch": 1.8375232774674115,
+      "grad_norm": 0.6296330373248001,
+      "learning_rate": 1.7869992536604847e-06,
+      "loss": 0.3962,
+      "step": 3947
+    },
+    {
+      "epoch": 1.8379888268156424,
+      "grad_norm": 0.6397390687681084,
+      "learning_rate": 1.7857695175993228e-06,
+      "loss": 0.4041,
+      "step": 3948
+    },
+    {
+      "epoch": 1.8384543761638734,
+      "grad_norm": 0.6091097831834549,
+      "learning_rate": 1.784539969683454e-06,
+      "loss": 0.3776,
+      "step": 3949
+    },
+    {
+      "epoch": 1.8389199255121043,
+      "grad_norm": 0.6024951230301993,
+      "learning_rate": 1.7833106102367704e-06,
+      "loss": 0.376,
+      "step": 3950
+    },
+    {
+      "epoch": 1.839385474860335,
+      "grad_norm": 0.5635084505470148,
+      "learning_rate": 1.7820814395831144e-06,
+      "loss": 0.3501,
+      "step": 3951
+    },
+    {
+      "epoch": 1.8398510242085662,
+      "grad_norm": 0.6137908995639111,
+      "learning_rate": 1.7808524580462796e-06,
+      "loss": 0.3819,
+      "step": 3952
+    },
+    {
+      "epoch": 1.840316573556797,
+      "grad_norm": 0.6221435996637831,
+      "learning_rate": 1.7796236659500087e-06,
+      "loss": 0.3593,
+      "step": 3953
+    },
+    {
+      "epoch": 1.8407821229050279,
+      "grad_norm": 0.5889122992389108,
+      "learning_rate": 1.7783950636179948e-06,
+      "loss": 0.3665,
+      "step": 3954
+    },
+    {
+      "epoch": 1.841247672253259,
+      "grad_norm": 0.60850288857796,
+      "learning_rate": 1.7771666513738788e-06,
+      "loss": 0.3761,
+      "step": 3955
+    },
+    {
+      "epoch": 1.8417132216014898,
+      "grad_norm": 0.5645679764278374,
+      "learning_rate": 1.7759384295412572e-06,
+      "loss": 0.3762,
+      "step": 3956
+    },
+    {
+      "epoch": 1.8421787709497206,
+      "grad_norm": 0.5774615475646533,
+      "learning_rate": 1.7747103984436714e-06,
+      "loss": 0.3773,
+      "step": 3957
+    },
+    {
+      "epoch": 1.8426443202979517,
+      "grad_norm": 0.5665739979930584,
+      "learning_rate": 1.7734825584046133e-06,
+      "loss": 0.3813,
+      "step": 3958
+    },
+    {
+      "epoch": 1.8431098696461825,
+      "grad_norm": 0.5634783407611893,
+      "learning_rate": 1.7722549097475256e-06,
+      "loss": 0.3763,
+      "step": 3959
+    },
+    {
+      "epoch": 1.8435754189944134,
+      "grad_norm": 0.6062857742403994,
+      "learning_rate": 1.7710274527958003e-06,
+      "loss": 0.3977,
+      "step": 3960
+    },
+    {
+      "epoch": 1.8440409683426444,
+      "grad_norm": 0.5933036384068948,
+      "learning_rate": 1.7698001878727786e-06,
+      "loss": 0.3814,
+      "step": 3961
+    },
+    {
+      "epoch": 1.8445065176908753,
+      "grad_norm": 0.5877060666607589,
+      "learning_rate": 1.7685731153017516e-06,
+      "loss": 0.3737,
+      "step": 3962
+    },
+    {
+      "epoch": 1.844972067039106,
+      "grad_norm": 0.5815504775270183,
+      "learning_rate": 1.7673462354059584e-06,
+      "loss": 0.3892,
+      "step": 3963
+    },
+    {
+      "epoch": 1.8454376163873372,
+      "grad_norm": 0.575041305461713,
+      "learning_rate": 1.7661195485085896e-06,
+      "loss": 0.3699,
+      "step": 3964
+    },
+    {
+      "epoch": 1.845903165735568,
+      "grad_norm": 0.594944425729458,
+      "learning_rate": 1.7648930549327834e-06,
+      "loss": 0.3886,
+      "step": 3965
+    },
+    {
+      "epoch": 1.8463687150837989,
+      "grad_norm": 0.6022755829131047,
+      "learning_rate": 1.7636667550016267e-06,
+      "loss": 0.3792,
+      "step": 3966
+    },
+    {
+      "epoch": 1.84683426443203,
+      "grad_norm": 0.5846933410367008,
+      "learning_rate": 1.7624406490381565e-06,
+      "loss": 0.3797,
+      "step": 3967
+    },
+    {
+      "epoch": 1.8472998137802608,
+      "grad_norm": 0.6367098717711257,
+      "learning_rate": 1.7612147373653587e-06,
+      "loss": 0.3955,
+      "step": 3968
+    },
+    {
+      "epoch": 1.8477653631284916,
+      "grad_norm": 0.6105693803944189,
+      "learning_rate": 1.7599890203061671e-06,
+      "loss": 0.3836,
+      "step": 3969
+    },
+    {
+      "epoch": 1.8482309124767227,
+      "grad_norm": 0.5829006814632057,
+      "learning_rate": 1.758763498183465e-06,
+      "loss": 0.3769,
+      "step": 3970
+    },
+    {
+      "epoch": 1.8486964618249533,
+      "grad_norm": 0.5695393351526763,
+      "learning_rate": 1.7575381713200825e-06,
+      "loss": 0.3737,
+      "step": 3971
+    },
+    {
+      "epoch": 1.8491620111731844,
+      "grad_norm": 0.5881633172603666,
+      "learning_rate": 1.7563130400388023e-06,
+      "loss": 0.3812,
+      "step": 3972
+    },
+    {
+      "epoch": 1.8496275605214154,
+      "grad_norm": 0.5697479086572883,
+      "learning_rate": 1.7550881046623524e-06,
+      "loss": 0.391,
+      "step": 3973
+    },
+    {
+      "epoch": 1.850093109869646,
+      "grad_norm": 0.6051239859688664,
+      "learning_rate": 1.7538633655134092e-06,
+      "loss": 0.3866,
+      "step": 3974
+    },
+    {
+      "epoch": 1.850558659217877,
+      "grad_norm": 0.5871395615467054,
+      "learning_rate": 1.7526388229145973e-06,
+      "loss": 0.3731,
+      "step": 3975
+    },
+    {
+      "epoch": 1.8510242085661082,
+      "grad_norm": 0.5654695639843986,
+      "learning_rate": 1.751414477188493e-06,
+      "loss": 0.3815,
+      "step": 3976
+    },
+    {
+      "epoch": 1.8514897579143388,
+      "grad_norm": 0.5997510790617059,
+      "learning_rate": 1.7501903286576161e-06,
+      "loss": 0.3677,
+      "step": 3977
+    },
+    {
+      "epoch": 1.8519553072625698,
+      "grad_norm": 0.586847578410014,
+      "learning_rate": 1.7489663776444354e-06,
+      "loss": 0.3908,
+      "step": 3978
+    },
+    {
+      "epoch": 1.8524208566108007,
+      "grad_norm": 0.6172216533854343,
+      "learning_rate": 1.7477426244713724e-06,
+      "loss": 0.3827,
+      "step": 3979
+    },
+    {
+      "epoch": 1.8528864059590315,
+      "grad_norm": 0.6216079891726326,
+      "learning_rate": 1.7465190694607897e-06,
+      "loss": 0.3754,
+      "step": 3980
+    },
+    {
+      "epoch": 1.8533519553072626,
+      "grad_norm": 0.6096982316738178,
+      "learning_rate": 1.7452957129350015e-06,
+      "loss": 0.397,
+      "step": 3981
+    },
+    {
+      "epoch": 1.8538175046554934,
+      "grad_norm": 0.6243149824297742,
+      "learning_rate": 1.7440725552162687e-06,
+      "loss": 0.3807,
+      "step": 3982
+    },
+    {
+      "epoch": 1.8542830540037243,
+      "grad_norm": 0.5934698470031529,
+      "learning_rate": 1.7428495966268012e-06,
+      "loss": 0.3724,
+      "step": 3983
+    },
+    {
+      "epoch": 1.8547486033519553,
+      "grad_norm": 0.5627331724050036,
+      "learning_rate": 1.7416268374887546e-06,
+      "loss": 0.3788,
+      "step": 3984
+    },
+    {
+      "epoch": 1.8552141527001862,
+      "grad_norm": 0.5952950305424703,
+      "learning_rate": 1.740404278124233e-06,
+      "loss": 0.3972,
+      "step": 3985
+    },
+    {
+      "epoch": 1.855679702048417,
+      "grad_norm": 0.5906576511454605,
+      "learning_rate": 1.7391819188552871e-06,
+      "loss": 0.3607,
+      "step": 3986
+    },
+    {
+      "epoch": 1.856145251396648,
+      "grad_norm": 0.6056940137647876,
+      "learning_rate": 1.7379597600039161e-06,
+      "loss": 0.3963,
+      "step": 3987
+    },
+    {
+      "epoch": 1.856610800744879,
+      "grad_norm": 0.5908250567208375,
+      "learning_rate": 1.7367378018920655e-06,
+      "loss": 0.3793,
+      "step": 3988
+    },
+    {
+      "epoch": 1.8570763500931098,
+      "grad_norm": 0.5999433099656879,
+      "learning_rate": 1.7355160448416282e-06,
+      "loss": 0.3952,
+      "step": 3989
+    },
+    {
+      "epoch": 1.8575418994413408,
+      "grad_norm": 0.636673712161235,
+      "learning_rate": 1.734294489174443e-06,
+      "loss": 0.3572,
+      "step": 3990
+    },
+    {
+      "epoch": 1.8580074487895717,
+      "grad_norm": 0.5583813389735458,
+      "learning_rate": 1.733073135212298e-06,
+      "loss": 0.378,
+      "step": 3991
+    },
+    {
+      "epoch": 1.8584729981378025,
+      "grad_norm": 0.582416283422367,
+      "learning_rate": 1.7318519832769271e-06,
+      "loss": 0.3717,
+      "step": 3992
+    },
+    {
+      "epoch": 1.8589385474860336,
+      "grad_norm": 0.5693060810331709,
+      "learning_rate": 1.73063103369001e-06,
+      "loss": 0.3695,
+      "step": 3993
+    },
+    {
+      "epoch": 1.8594040968342644,
+      "grad_norm": 0.5966651044606225,
+      "learning_rate": 1.7294102867731727e-06,
+      "loss": 0.3978,
+      "step": 3994
+    },
+    {
+      "epoch": 1.8598696461824953,
+      "grad_norm": 0.5757050227696792,
+      "learning_rate": 1.7281897428479912e-06,
+      "loss": 0.3945,
+      "step": 3995
+    },
+    {
+      "epoch": 1.8603351955307263,
+      "grad_norm": 0.554330708691486,
+      "learning_rate": 1.7269694022359856e-06,
+      "loss": 0.384,
+      "step": 3996
+    },
+    {
+      "epoch": 1.8608007448789572,
+      "grad_norm": 0.6056658260222507,
+      "learning_rate": 1.7257492652586211e-06,
+      "loss": 0.3976,
+      "step": 3997
+    },
+    {
+      "epoch": 1.861266294227188,
+      "grad_norm": 0.6209594988651841,
+      "learning_rate": 1.7245293322373114e-06,
+      "loss": 0.3931,
+      "step": 3998
+    },
+    {
+      "epoch": 1.861731843575419,
+      "grad_norm": 0.5970497001402106,
+      "learning_rate": 1.723309603493416e-06,
+      "loss": 0.3808,
+      "step": 3999
+    },
+    {
+      "epoch": 1.86219739292365,
+      "grad_norm": 0.5806593121059189,
+      "learning_rate": 1.722090079348241e-06,
+      "loss": 0.3725,
+      "step": 4000
+    },
+    {
+      "epoch": 1.8626629422718808,
+      "grad_norm": 0.6180192462405613,
+      "learning_rate": 1.7208707601230374e-06,
+      "loss": 0.3997,
+      "step": 4001
+    },
+    {
+      "epoch": 1.8631284916201118,
+      "grad_norm": 0.6522873397364412,
+      "learning_rate": 1.7196516461390023e-06,
+      "loss": 0.3921,
+      "step": 4002
+    },
+    {
+      "epoch": 1.8635940409683427,
+      "grad_norm": 0.6188772451780784,
+      "learning_rate": 1.7184327377172808e-06,
+      "loss": 0.3872,
+      "step": 4003
+    },
+    {
+      "epoch": 1.8640595903165735,
+      "grad_norm": 0.6266692314422723,
+      "learning_rate": 1.7172140351789616e-06,
+      "loss": 0.3746,
+      "step": 4004
+    },
+    {
+      "epoch": 1.8645251396648046,
+      "grad_norm": 0.5995116049014133,
+      "learning_rate": 1.7159955388450798e-06,
+      "loss": 0.37,
+      "step": 4005
+    },
+    {
+      "epoch": 1.8649906890130352,
+      "grad_norm": 0.567510869422854,
+      "learning_rate": 1.714777249036616e-06,
+      "loss": 0.3627,
+      "step": 4006
+    },
+    {
+      "epoch": 1.8654562383612663,
+      "grad_norm": 0.5648197935531025,
+      "learning_rate": 1.7135591660744977e-06,
+      "loss": 0.3944,
+      "step": 4007
+    },
+    {
+      "epoch": 1.8659217877094973,
+      "grad_norm": 0.5907115986371252,
+      "learning_rate": 1.7123412902795964e-06,
+      "loss": 0.3641,
+      "step": 4008
+    },
+    {
+      "epoch": 1.866387337057728,
+      "grad_norm": 0.6260149167977905,
+      "learning_rate": 1.7111236219727301e-06,
+      "loss": 0.3916,
+      "step": 4009
+    },
+    {
+      "epoch": 1.866852886405959,
+      "grad_norm": 0.5827480523422258,
+      "learning_rate": 1.7099061614746592e-06,
+      "loss": 0.3825,
+      "step": 4010
+    },
+    {
+      "epoch": 1.86731843575419,
+      "grad_norm": 0.6121794283885263,
+      "learning_rate": 1.7086889091060948e-06,
+      "loss": 0.3839,
+      "step": 4011
+    },
+    {
+      "epoch": 1.8677839851024207,
+      "grad_norm": 0.6367706862747867,
+      "learning_rate": 1.7074718651876892e-06,
+      "loss": 0.372,
+      "step": 4012
+    },
+    {
+      "epoch": 1.8682495344506518,
+      "grad_norm": 0.5707828915711408,
+      "learning_rate": 1.7062550300400399e-06,
+      "loss": 0.3651,
+      "step": 4013
+    },
+    {
+      "epoch": 1.8687150837988828,
+      "grad_norm": 0.6078395735431122,
+      "learning_rate": 1.7050384039836897e-06,
+      "loss": 0.3805,
+      "step": 4014
+    },
+    {
+      "epoch": 1.8691806331471135,
+      "grad_norm": 0.6106968402682293,
+      "learning_rate": 1.7038219873391295e-06,
+      "loss": 0.3807,
+      "step": 4015
+    },
+    {
+      "epoch": 1.8696461824953445,
+      "grad_norm": 0.5640267737794802,
+      "learning_rate": 1.7026057804267903e-06,
+      "loss": 0.3719,
+      "step": 4016
+    },
+    {
+      "epoch": 1.8701117318435754,
+      "grad_norm": 0.590317131312482,
+      "learning_rate": 1.70138978356705e-06,
+      "loss": 0.3688,
+      "step": 4017
+    },
+    {
+      "epoch": 1.8705772811918062,
+      "grad_norm": 0.6000238760076982,
+      "learning_rate": 1.700173997080231e-06,
+      "loss": 0.3751,
+      "step": 4018
+    },
+    {
+      "epoch": 1.8710428305400373,
+      "grad_norm": 0.5831735369922275,
+      "learning_rate": 1.6989584212866014e-06,
+      "loss": 0.402,
+      "step": 4019
+    },
+    {
+      "epoch": 1.8715083798882681,
+      "grad_norm": 0.5990605804582091,
+      "learning_rate": 1.6977430565063722e-06,
+      "loss": 0.3883,
+      "step": 4020
+    },
+    {
+      "epoch": 1.871973929236499,
+      "grad_norm": 0.5915232499786063,
+      "learning_rate": 1.6965279030596992e-06,
+      "loss": 0.3724,
+      "step": 4021
+    },
+    {
+      "epoch": 1.87243947858473,
+      "grad_norm": 0.5810760945738727,
+      "learning_rate": 1.6953129612666822e-06,
+      "loss": 0.3928,
+      "step": 4022
+    },
+    {
+      "epoch": 1.8729050279329609,
+      "grad_norm": 0.634032915705945,
+      "learning_rate": 1.6940982314473669e-06,
+      "loss": 0.3706,
+      "step": 4023
+    },
+    {
+      "epoch": 1.8733705772811917,
+      "grad_norm": 0.6417271285454472,
+      "learning_rate": 1.692883713921742e-06,
+      "loss": 0.3915,
+      "step": 4024
+    },
+    {
+      "epoch": 1.8738361266294228,
+      "grad_norm": 0.5942491554595456,
+      "learning_rate": 1.6916694090097391e-06,
+      "loss": 0.385,
+      "step": 4025
+    },
+    {
+      "epoch": 1.8743016759776536,
+      "grad_norm": 0.6194728430534115,
+      "learning_rate": 1.690455317031236e-06,
+      "loss": 0.3801,
+      "step": 4026
+    },
+    {
+      "epoch": 1.8747672253258845,
+      "grad_norm": 0.6242833375091912,
+      "learning_rate": 1.6892414383060532e-06,
+      "loss": 0.3827,
+      "step": 4027
+    },
+    {
+      "epoch": 1.8752327746741155,
+      "grad_norm": 0.5908014022619311,
+      "learning_rate": 1.6880277731539553e-06,
+      "loss": 0.3711,
+      "step": 4028
+    },
+    {
+      "epoch": 1.8756983240223464,
+      "grad_norm": 0.610865049275944,
+      "learning_rate": 1.6868143218946498e-06,
+      "loss": 0.3722,
+      "step": 4029
+    },
+    {
+      "epoch": 1.8761638733705772,
+      "grad_norm": 0.570653526645025,
+      "learning_rate": 1.6856010848477897e-06,
+      "loss": 0.3882,
+      "step": 4030
+    },
+    {
+      "epoch": 1.8766294227188083,
+      "grad_norm": 0.5788788118398936,
+      "learning_rate": 1.68438806233297e-06,
+      "loss": 0.3654,
+      "step": 4031
+    },
+    {
+      "epoch": 1.8770949720670391,
+      "grad_norm": 0.5977507339416102,
+      "learning_rate": 1.68317525466973e-06,
+      "loss": 0.4,
+      "step": 4032
+    },
+    {
+      "epoch": 1.87756052141527,
+      "grad_norm": 0.578883132173403,
+      "learning_rate": 1.6819626621775504e-06,
+      "loss": 0.3845,
+      "step": 4033
+    },
+    {
+      "epoch": 1.878026070763501,
+      "grad_norm": 0.6136111939207259,
+      "learning_rate": 1.6807502851758595e-06,
+      "loss": 0.3805,
+      "step": 4034
+    },
+    {
+      "epoch": 1.8784916201117319,
+      "grad_norm": 0.6340377597725104,
+      "learning_rate": 1.6795381239840254e-06,
+      "loss": 0.3771,
+      "step": 4035
+    },
+    {
+      "epoch": 1.8789571694599627,
+      "grad_norm": 0.5925620167544223,
+      "learning_rate": 1.678326178921359e-06,
+      "loss": 0.3936,
+      "step": 4036
+    },
+    {
+      "epoch": 1.8794227188081938,
+      "grad_norm": 0.614601404400024,
+      "learning_rate": 1.6771144503071156e-06,
+      "loss": 0.3929,
+      "step": 4037
+    },
+    {
+      "epoch": 1.8798882681564246,
+      "grad_norm": 0.5966672466746488,
+      "learning_rate": 1.675902938460495e-06,
+      "loss": 0.398,
+      "step": 4038
+    },
+    {
+      "epoch": 1.8803538175046555,
+      "grad_norm": 0.6013430468774709,
+      "learning_rate": 1.674691643700637e-06,
+      "loss": 0.3758,
+      "step": 4039
+    },
+    {
+      "epoch": 1.8808193668528865,
+      "grad_norm": 0.6490617075304898,
+      "learning_rate": 1.6734805663466253e-06,
+      "loss": 0.3879,
+      "step": 4040
+    },
+    {
+      "epoch": 1.8812849162011172,
+      "grad_norm": 0.5859084876499409,
+      "learning_rate": 1.6722697067174867e-06,
+      "loss": 0.37,
+      "step": 4041
+    },
+    {
+      "epoch": 1.8817504655493482,
+      "grad_norm": 0.5584852157721049,
+      "learning_rate": 1.6710590651321905e-06,
+      "loss": 0.3754,
+      "step": 4042
+    },
+    {
+      "epoch": 1.8822160148975793,
+      "grad_norm": 0.594411530618784,
+      "learning_rate": 1.6698486419096493e-06,
+      "loss": 0.3872,
+      "step": 4043
+    },
+    {
+      "epoch": 1.88268156424581,
+      "grad_norm": 0.6205304432690172,
+      "learning_rate": 1.668638437368716e-06,
+      "loss": 0.3762,
+      "step": 4044
+    },
+    {
+      "epoch": 1.883147113594041,
+      "grad_norm": 0.59100086672915,
+      "learning_rate": 1.6674284518281874e-06,
+      "loss": 0.3911,
+      "step": 4045
+    },
+    {
+      "epoch": 1.883612662942272,
+      "grad_norm": 0.6007826920259373,
+      "learning_rate": 1.6662186856068035e-06,
+      "loss": 0.3867,
+      "step": 4046
+    },
+    {
+      "epoch": 1.8840782122905027,
+      "grad_norm": 0.5877567971019327,
+      "learning_rate": 1.6650091390232449e-06,
+      "loss": 0.3611,
+      "step": 4047
+    },
+    {
+      "epoch": 1.8845437616387337,
+      "grad_norm": 0.6234484917356565,
+      "learning_rate": 1.663799812396135e-06,
+      "loss": 0.383,
+      "step": 4048
+    },
+    {
+      "epoch": 1.8850093109869648,
+      "grad_norm": 0.5745435272582603,
+      "learning_rate": 1.662590706044038e-06,
+      "loss": 0.3764,
+      "step": 4049
+    },
+    {
+      "epoch": 1.8854748603351954,
+      "grad_norm": 0.60051245836092,
+      "learning_rate": 1.6613818202854632e-06,
+      "loss": 0.3919,
+      "step": 4050
+    },
+    {
+      "epoch": 1.8859404096834265,
+      "grad_norm": 0.6249659272646555,
+      "learning_rate": 1.6601731554388596e-06,
+      "loss": 0.3642,
+      "step": 4051
+    },
+    {
+      "epoch": 1.8864059590316573,
+      "grad_norm": 0.61932351027203,
+      "learning_rate": 1.6589647118226171e-06,
+      "loss": 0.3994,
+      "step": 4052
+    },
+    {
+      "epoch": 1.8868715083798882,
+      "grad_norm": 0.603952599618994,
+      "learning_rate": 1.6577564897550679e-06,
+      "loss": 0.3877,
+      "step": 4053
+    },
+    {
+      "epoch": 1.8873370577281192,
+      "grad_norm": 0.6287015650587684,
+      "learning_rate": 1.6565484895544895e-06,
+      "loss": 0.3755,
+      "step": 4054
+    },
+    {
+      "epoch": 1.88780260707635,
+      "grad_norm": 0.6020658191308592,
+      "learning_rate": 1.6553407115390952e-06,
+      "loss": 0.3844,
+      "step": 4055
+    },
+    {
+      "epoch": 1.888268156424581,
+      "grad_norm": 0.6143412600833907,
+      "learning_rate": 1.6541331560270432e-06,
+      "loss": 0.3689,
+      "step": 4056
+    },
+    {
+      "epoch": 1.888733705772812,
+      "grad_norm": 0.6285993879219444,
+      "learning_rate": 1.652925823336432e-06,
+      "loss": 0.3889,
+      "step": 4057
+    },
+    {
+      "epoch": 1.8891992551210428,
+      "grad_norm": 0.6225671317265159,
+      "learning_rate": 1.6517187137853025e-06,
+      "loss": 0.3774,
+      "step": 4058
+    },
+    {
+      "epoch": 1.8896648044692737,
+      "grad_norm": 0.5730941111204662,
+      "learning_rate": 1.6505118276916354e-06,
+      "loss": 0.3845,
+      "step": 4059
+    },
+    {
+      "epoch": 1.8901303538175047,
+      "grad_norm": 0.6188334836854512,
+      "learning_rate": 1.649305165373354e-06,
+      "loss": 0.3894,
+      "step": 4060
+    },
+    {
+      "epoch": 1.8905959031657356,
+      "grad_norm": 0.6082319789714162,
+      "learning_rate": 1.6480987271483204e-06,
+      "loss": 0.378,
+      "step": 4061
+    },
+    {
+      "epoch": 1.8910614525139664,
+      "grad_norm": 0.6324894234295024,
+      "learning_rate": 1.6468925133343405e-06,
+      "loss": 0.3898,
+      "step": 4062
+    },
+    {
+      "epoch": 1.8915270018621975,
+      "grad_norm": 0.6203361344750439,
+      "learning_rate": 1.6456865242491594e-06,
+      "loss": 0.3827,
+      "step": 4063
+    },
+    {
+      "epoch": 1.8919925512104283,
+      "grad_norm": 0.5769501210598239,
+      "learning_rate": 1.644480760210463e-06,
+      "loss": 0.3758,
+      "step": 4064
+    },
+    {
+      "epoch": 1.8924581005586592,
+      "grad_norm": 0.5724163228092232,
+      "learning_rate": 1.643275221535878e-06,
+      "loss": 0.3742,
+      "step": 4065
+    },
+    {
+      "epoch": 1.8929236499068902,
+      "grad_norm": 0.5909344986696276,
+      "learning_rate": 1.6420699085429725e-06,
+      "loss": 0.3837,
+      "step": 4066
+    },
+    {
+      "epoch": 1.893389199255121,
+      "grad_norm": 0.654762881078854,
+      "learning_rate": 1.6408648215492547e-06,
+      "loss": 0.4001,
+      "step": 4067
+    },
+    {
+      "epoch": 1.893854748603352,
+      "grad_norm": 0.6381604802202689,
+      "learning_rate": 1.6396599608721724e-06,
+      "loss": 0.3662,
+      "step": 4068
+    },
+    {
+      "epoch": 1.894320297951583,
+      "grad_norm": 0.5802621990144488,
+      "learning_rate": 1.6384553268291152e-06,
+      "loss": 0.3908,
+      "step": 4069
+    },
+    {
+      "epoch": 1.8947858472998138,
+      "grad_norm": 0.6040399413976217,
+      "learning_rate": 1.6372509197374126e-06,
+      "loss": 0.3919,
+      "step": 4070
+    },
+    {
+      "epoch": 1.8952513966480447,
+      "grad_norm": 0.6059875184253006,
+      "learning_rate": 1.6360467399143337e-06,
+      "loss": 0.3815,
+      "step": 4071
+    },
+    {
+      "epoch": 1.8957169459962757,
+      "grad_norm": 0.5852124641593037,
+      "learning_rate": 1.6348427876770867e-06,
+      "loss": 0.3629,
+      "step": 4072
+    },
+    {
+      "epoch": 1.8961824953445066,
+      "grad_norm": 0.6205073315666811,
+      "learning_rate": 1.633639063342824e-06,
+      "loss": 0.3723,
+      "step": 4073
+    },
+    {
+      "epoch": 1.8966480446927374,
+      "grad_norm": 0.6048830179949973,
+      "learning_rate": 1.6324355672286339e-06,
+      "loss": 0.391,
+      "step": 4074
+    },
+    {
+      "epoch": 1.8971135940409685,
+      "grad_norm": 0.603895577972041,
+      "learning_rate": 1.6312322996515453e-06,
+      "loss": 0.3721,
+      "step": 4075
+    },
+    {
+      "epoch": 1.8975791433891993,
+      "grad_norm": 0.6302497036276025,
+      "learning_rate": 1.630029260928527e-06,
+      "loss": 0.3968,
+      "step": 4076
+    },
+    {
+      "epoch": 1.8980446927374302,
+      "grad_norm": 0.620192271956383,
+      "learning_rate": 1.6288264513764906e-06,
+      "loss": 0.3698,
+      "step": 4077
+    },
+    {
+      "epoch": 1.8985102420856612,
+      "grad_norm": 0.5527251940060828,
+      "learning_rate": 1.6276238713122828e-06,
+      "loss": 0.3815,
+      "step": 4078
+    },
+    {
+      "epoch": 1.8989757914338918,
+      "grad_norm": 0.5551200985073371,
+      "learning_rate": 1.6264215210526918e-06,
+      "loss": 0.3825,
+      "step": 4079
+    },
+    {
+      "epoch": 1.899441340782123,
+      "grad_norm": 0.6071057415354925,
+      "learning_rate": 1.6252194009144451e-06,
+      "loss": 0.3871,
+      "step": 4080
+    },
+    {
+      "epoch": 1.899906890130354,
+      "grad_norm": 0.6303011495245507,
+      "learning_rate": 1.624017511214211e-06,
+      "loss": 0.4087,
+      "step": 4081
+    },
+    {
+      "epoch": 1.9003724394785846,
+      "grad_norm": 0.59628776422951,
+      "learning_rate": 1.6228158522685946e-06,
+      "loss": 0.3729,
+      "step": 4082
+    },
+    {
+      "epoch": 1.9008379888268156,
+      "grad_norm": 0.6581531108788702,
+      "learning_rate": 1.6216144243941423e-06,
+      "loss": 0.3836,
+      "step": 4083
+    },
+    {
+      "epoch": 1.9013035381750467,
+      "grad_norm": 0.5828274033050703,
+      "learning_rate": 1.6204132279073377e-06,
+      "loss": 0.3674,
+      "step": 4084
+    },
+    {
+      "epoch": 1.9017690875232773,
+      "grad_norm": 0.6027643453642733,
+      "learning_rate": 1.6192122631246055e-06,
+      "loss": 0.395,
+      "step": 4085
+    },
+    {
+      "epoch": 1.9022346368715084,
+      "grad_norm": 0.5772851853069649,
+      "learning_rate": 1.6180115303623083e-06,
+      "loss": 0.3742,
+      "step": 4086
+    },
+    {
+      "epoch": 1.9027001862197392,
+      "grad_norm": 0.5854705885005324,
+      "learning_rate": 1.616811029936748e-06,
+      "loss": 0.3779,
+      "step": 4087
+    },
+    {
+      "epoch": 1.90316573556797,
+      "grad_norm": 0.5763811363195254,
+      "learning_rate": 1.6156107621641625e-06,
+      "loss": 0.3788,
+      "step": 4088
+    },
+    {
+      "epoch": 1.9036312849162011,
+      "grad_norm": 0.6378851977491488,
+      "learning_rate": 1.614410727360734e-06,
+      "loss": 0.3742,
+      "step": 4089
+    },
+    {
+      "epoch": 1.904096834264432,
+      "grad_norm": 0.6061970780098137,
+      "learning_rate": 1.6132109258425792e-06,
+      "loss": 0.3855,
+      "step": 4090
+    },
+    {
+      "epoch": 1.9045623836126628,
+      "grad_norm": 0.5861249095018672,
+      "learning_rate": 1.612011357925754e-06,
+      "loss": 0.381,
+      "step": 4091
+    },
+    {
+      "epoch": 1.905027932960894,
+      "grad_norm": 0.597260974772144,
+      "learning_rate": 1.6108120239262521e-06,
+      "loss": 0.3852,
+      "step": 4092
+    },
+    {
+      "epoch": 1.9054934823091247,
+      "grad_norm": 0.6127758652310201,
+      "learning_rate": 1.6096129241600089e-06,
+      "loss": 0.3839,
+      "step": 4093
+    },
+    {
+      "epoch": 1.9059590316573556,
+      "grad_norm": 0.6232125408841421,
+      "learning_rate": 1.6084140589428943e-06,
+      "loss": 0.3906,
+      "step": 4094
+    },
+    {
+      "epoch": 1.9064245810055866,
+      "grad_norm": 0.6028600368784979,
+      "learning_rate": 1.6072154285907181e-06,
+      "loss": 0.3698,
+      "step": 4095
+    },
+    {
+      "epoch": 1.9068901303538175,
+      "grad_norm": 0.5983797707364137,
+      "learning_rate": 1.606017033419228e-06,
+      "loss": 0.4027,
+      "step": 4096
+    },
+    {
+      "epoch": 1.9073556797020483,
+      "grad_norm": 0.6056516106673476,
+      "learning_rate": 1.60481887374411e-06,
+      "loss": 0.375,
+      "step": 4097
+    },
+    {
+      "epoch": 1.9078212290502794,
+      "grad_norm": 0.5828647104461655,
+      "learning_rate": 1.603620949880988e-06,
+      "loss": 0.362,
+      "step": 4098
+    },
+    {
+      "epoch": 1.9082867783985102,
+      "grad_norm": 0.6094189958525131,
+      "learning_rate": 1.6024232621454233e-06,
+      "loss": 0.3941,
+      "step": 4099
+    },
+    {
+      "epoch": 1.908752327746741,
+      "grad_norm": 0.5954455752912717,
+      "learning_rate": 1.6012258108529152e-06,
+      "loss": 0.3758,
+      "step": 4100
+    },
+    {
+      "epoch": 1.9092178770949721,
+      "grad_norm": 0.5697468341325009,
+      "learning_rate": 1.6000285963189017e-06,
+      "loss": 0.3658,
+      "step": 4101
+    },
+    {
+      "epoch": 1.909683426443203,
+      "grad_norm": 0.5975920168017237,
+      "learning_rate": 1.5988316188587567e-06,
+      "loss": 0.3956,
+      "step": 4102
+    },
+    {
+      "epoch": 1.9101489757914338,
+      "grad_norm": 0.5780151593381432,
+      "learning_rate": 1.5976348787877926e-06,
+      "loss": 0.391,
+      "step": 4103
+    },
+    {
+      "epoch": 1.910614525139665,
+      "grad_norm": 0.6153145430484728,
+      "learning_rate": 1.596438376421259e-06,
+      "loss": 0.3763,
+      "step": 4104
+    },
+    {
+      "epoch": 1.9110800744878957,
+      "grad_norm": 0.6066248355548473,
+      "learning_rate": 1.5952421120743442e-06,
+      "loss": 0.3935,
+      "step": 4105
+    },
+    {
+      "epoch": 1.9115456238361266,
+      "grad_norm": 0.589861670698035,
+      "learning_rate": 1.5940460860621718e-06,
+      "loss": 0.3946,
+      "step": 4106
+    },
+    {
+      "epoch": 1.9120111731843576,
+      "grad_norm": 0.6073607652145968,
+      "learning_rate": 1.592850298699804e-06,
+      "loss": 0.3762,
+      "step": 4107
+    },
+    {
+      "epoch": 1.9124767225325885,
+      "grad_norm": 0.604144839580751,
+      "learning_rate": 1.5916547503022378e-06,
+      "loss": 0.3655,
+      "step": 4108
+    },
+    {
+      "epoch": 1.9129422718808193,
+      "grad_norm": 0.5647230491363847,
+      "learning_rate": 1.5904594411844118e-06,
+      "loss": 0.3709,
+      "step": 4109
+    },
+    {
+      "epoch": 1.9134078212290504,
+      "grad_norm": 0.5813409901458911,
+      "learning_rate": 1.5892643716611977e-06,
+      "loss": 0.3775,
+      "step": 4110
+    },
+    {
+      "epoch": 1.9138733705772812,
+      "grad_norm": 0.6106464704356157,
+      "learning_rate": 1.5880695420474041e-06,
+      "loss": 0.385,
+      "step": 4111
+    },
+    {
+      "epoch": 1.914338919925512,
+      "grad_norm": 0.6136912583863607,
+      "learning_rate": 1.5868749526577793e-06,
+      "loss": 0.3895,
+      "step": 4112
+    },
+    {
+      "epoch": 1.9148044692737431,
+      "grad_norm": 0.5931349719433896,
+      "learning_rate": 1.5856806038070062e-06,
+      "loss": 0.3865,
+      "step": 4113
+    },
+    {
+      "epoch": 1.9152700186219738,
+      "grad_norm": 0.5849756214790336,
+      "learning_rate": 1.5844864958097038e-06,
+      "loss": 0.3573,
+      "step": 4114
+    },
+    {
+      "epoch": 1.9157355679702048,
+      "grad_norm": 0.5858909541334124,
+      "learning_rate": 1.5832926289804278e-06,
+      "loss": 0.4006,
+      "step": 4115
+    },
+    {
+      "epoch": 1.916201117318436,
+      "grad_norm": 0.5917346655064935,
+      "learning_rate": 1.5820990036336744e-06,
+      "loss": 0.3915,
+      "step": 4116
+    },
+    {
+      "epoch": 1.9166666666666665,
+      "grad_norm": 0.5686529124969363,
+      "learning_rate": 1.58090562008387e-06,
+      "loss": 0.3635,
+      "step": 4117
+    },
+    {
+      "epoch": 1.9171322160148976,
+      "grad_norm": 0.572793507155979,
+      "learning_rate": 1.5797124786453813e-06,
+      "loss": 0.3837,
+      "step": 4118
+    },
+    {
+      "epoch": 1.9175977653631286,
+      "grad_norm": 0.6052103672784449,
+      "learning_rate": 1.5785195796325091e-06,
+      "loss": 0.3653,
+      "step": 4119
+    },
+    {
+      "epoch": 1.9180633147113593,
+      "grad_norm": 0.6203611677289693,
+      "learning_rate": 1.577326923359493e-06,
+      "loss": 0.3649,
+      "step": 4120
+    },
+    {
+      "epoch": 1.9185288640595903,
+      "grad_norm": 0.5826531481936803,
+      "learning_rate": 1.5761345101405057e-06,
+      "loss": 0.3855,
+      "step": 4121
+    },
+    {
+      "epoch": 1.9189944134078212,
+      "grad_norm": 0.6084261017907526,
+      "learning_rate": 1.574942340289658e-06,
+      "loss": 0.3841,
+      "step": 4122
+    },
+    {
+      "epoch": 1.919459962756052,
+      "grad_norm": 0.615687716236235,
+      "learning_rate": 1.5737504141209946e-06,
+      "loss": 0.3776,
+      "step": 4123
+    },
+    {
+      "epoch": 1.919925512104283,
+      "grad_norm": 0.591471236013041,
+      "learning_rate": 1.5725587319484987e-06,
+      "loss": 0.3836,
+      "step": 4124
+    },
+    {
+      "epoch": 1.920391061452514,
+      "grad_norm": 0.5916473899963225,
+      "learning_rate": 1.5713672940860872e-06,
+      "loss": 0.3749,
+      "step": 4125
+    },
+    {
+      "epoch": 1.9208566108007448,
+      "grad_norm": 0.5788136145662336,
+      "learning_rate": 1.5701761008476135e-06,
+      "loss": 0.3852,
+      "step": 4126
+    },
+    {
+      "epoch": 1.9213221601489758,
+      "grad_norm": 0.5714386181338448,
+      "learning_rate": 1.568985152546864e-06,
+      "loss": 0.3743,
+      "step": 4127
+    },
+    {
+      "epoch": 1.9217877094972067,
+      "grad_norm": 0.6305748474327388,
+      "learning_rate": 1.5677944494975661e-06,
+      "loss": 0.3862,
+      "step": 4128
+    },
+    {
+      "epoch": 1.9222532588454375,
+      "grad_norm": 0.5897186501230324,
+      "learning_rate": 1.566603992013378e-06,
+      "loss": 0.3922,
+      "step": 4129
+    },
+    {
+      "epoch": 1.9227188081936686,
+      "grad_norm": 0.6041815134055851,
+      "learning_rate": 1.5654137804078935e-06,
+      "loss": 0.3909,
+      "step": 4130
+    },
+    {
+      "epoch": 1.9231843575418994,
+      "grad_norm": 0.597963202138042,
+      "learning_rate": 1.5642238149946426e-06,
+      "loss": 0.4038,
+      "step": 4131
+    },
+    {
+      "epoch": 1.9236499068901303,
+      "grad_norm": 0.6079231962519998,
+      "learning_rate": 1.5630340960870927e-06,
+      "loss": 0.3786,
+      "step": 4132
+    },
+    {
+      "epoch": 1.9241154562383613,
+      "grad_norm": 0.6085930800804389,
+      "learning_rate": 1.5618446239986423e-06,
+      "loss": 0.3596,
+      "step": 4133
+    },
+    {
+      "epoch": 1.9245810055865922,
+      "grad_norm": 0.6116909469363038,
+      "learning_rate": 1.5606553990426272e-06,
+      "loss": 0.3777,
+      "step": 4134
+    },
+    {
+      "epoch": 1.925046554934823,
+      "grad_norm": 0.583548525399762,
+      "learning_rate": 1.5594664215323163e-06,
+      "loss": 0.3923,
+      "step": 4135
+    },
+    {
+      "epoch": 1.925512104283054,
+      "grad_norm": 0.5956516228794476,
+      "learning_rate": 1.5582776917809164e-06,
+      "loss": 0.3794,
+      "step": 4136
+    },
+    {
+      "epoch": 1.925977653631285,
+      "grad_norm": 0.5733246897534616,
+      "learning_rate": 1.5570892101015662e-06,
+      "loss": 0.3834,
+      "step": 4137
+    },
+    {
+      "epoch": 1.9264432029795158,
+      "grad_norm": 0.5689134713124583,
+      "learning_rate": 1.5559009768073403e-06,
+      "loss": 0.3975,
+      "step": 4138
+    },
+    {
+      "epoch": 1.9269087523277468,
+      "grad_norm": 0.5793276879934001,
+      "learning_rate": 1.5547129922112469e-06,
+      "loss": 0.3697,
+      "step": 4139
+    },
+    {
+      "epoch": 1.9273743016759777,
+      "grad_norm": 0.6232285646943401,
+      "learning_rate": 1.5535252566262305e-06,
+      "loss": 0.3794,
+      "step": 4140
+    },
+    {
+      "epoch": 1.9278398510242085,
+      "grad_norm": 0.6273886422452172,
+      "learning_rate": 1.5523377703651686e-06,
+      "loss": 0.3835,
+      "step": 4141
+    },
+    {
+      "epoch": 1.9283054003724396,
+      "grad_norm": 0.6135220864326743,
+      "learning_rate": 1.5511505337408727e-06,
+      "loss": 0.3977,
+      "step": 4142
+    },
+    {
+      "epoch": 1.9287709497206704,
+      "grad_norm": 0.5919123912216434,
+      "learning_rate": 1.5499635470660893e-06,
+      "loss": 0.376,
+      "step": 4143
+    },
+    {
+      "epoch": 1.9292364990689013,
+      "grad_norm": 0.5824987223669216,
+      "learning_rate": 1.5487768106534995e-06,
+      "loss": 0.3567,
+      "step": 4144
+    },
+    {
+      "epoch": 1.9297020484171323,
+      "grad_norm": 0.5892711604278601,
+      "learning_rate": 1.5475903248157178e-06,
+      "loss": 0.3918,
+      "step": 4145
+    },
+    {
+      "epoch": 1.9301675977653632,
+      "grad_norm": 0.5969679755625846,
+      "learning_rate": 1.546404089865293e-06,
+      "loss": 0.3982,
+      "step": 4146
+    },
+    {
+      "epoch": 1.930633147113594,
+      "grad_norm": 0.5498585676806925,
+      "learning_rate": 1.5452181061147053e-06,
+      "loss": 0.3707,
+      "step": 4147
+    },
+    {
+      "epoch": 1.931098696461825,
+      "grad_norm": 0.5677750502110738,
+      "learning_rate": 1.5440323738763743e-06,
+      "loss": 0.3897,
+      "step": 4148
+    },
+    {
+      "epoch": 1.9315642458100557,
+      "grad_norm": 0.5657369589766351,
+      "learning_rate": 1.542846893462649e-06,
+      "loss": 0.3829,
+      "step": 4149
+    },
+    {
+      "epoch": 1.9320297951582868,
+      "grad_norm": 0.5763005624105891,
+      "learning_rate": 1.5416616651858123e-06,
+      "loss": 0.3725,
+      "step": 4150
+    },
+    {
+      "epoch": 1.9324953445065178,
+      "grad_norm": 0.6126073462673844,
+      "learning_rate": 1.5404766893580808e-06,
+      "loss": 0.4112,
+      "step": 4151
+    },
+    {
+      "epoch": 1.9329608938547485,
+      "grad_norm": 0.6175040723536787,
+      "learning_rate": 1.5392919662916083e-06,
+      "loss": 0.3869,
+      "step": 4152
+    },
+    {
+      "epoch": 1.9334264432029795,
+      "grad_norm": 0.5954473139572185,
+      "learning_rate": 1.5381074962984766e-06,
+      "loss": 0.3817,
+      "step": 4153
+    },
+    {
+      "epoch": 1.9338919925512106,
+      "grad_norm": 0.5685864970199153,
+      "learning_rate": 1.5369232796907036e-06,
+      "loss": 0.378,
+      "step": 4154
+    },
+    {
+      "epoch": 1.9343575418994412,
+      "grad_norm": 0.6325908622982922,
+      "learning_rate": 1.5357393167802404e-06,
+      "loss": 0.3732,
+      "step": 4155
+    },
+    {
+      "epoch": 1.9348230912476723,
+      "grad_norm": 0.5950760827172615,
+      "learning_rate": 1.5345556078789709e-06,
+      "loss": 0.3835,
+      "step": 4156
+    },
+    {
+      "epoch": 1.9352886405959033,
+      "grad_norm": 0.6309653434963831,
+      "learning_rate": 1.5333721532987126e-06,
+      "loss": 0.3827,
+      "step": 4157
+    },
+    {
+      "epoch": 1.935754189944134,
+      "grad_norm": 0.5838362841306609,
+      "learning_rate": 1.5321889533512146e-06,
+      "loss": 0.3678,
+      "step": 4158
+    },
+    {
+      "epoch": 1.936219739292365,
+      "grad_norm": 0.6086515133788597,
+      "learning_rate": 1.5310060083481608e-06,
+      "loss": 0.4005,
+      "step": 4159
+    },
+    {
+      "epoch": 1.9366852886405959,
+      "grad_norm": 0.5770848423121413,
+      "learning_rate": 1.5298233186011675e-06,
+      "loss": 0.3716,
+      "step": 4160
+    },
+    {
+      "epoch": 1.9371508379888267,
+      "grad_norm": 0.6306632209320181,
+      "learning_rate": 1.5286408844217821e-06,
+      "loss": 0.3698,
+      "step": 4161
+    },
+    {
+      "epoch": 1.9376163873370578,
+      "grad_norm": 0.5922418226290386,
+      "learning_rate": 1.5274587061214858e-06,
+      "loss": 0.3817,
+      "step": 4162
+    },
+    {
+      "epoch": 1.9380819366852886,
+      "grad_norm": 0.6486020840093621,
+      "learning_rate": 1.5262767840116937e-06,
+      "loss": 0.3904,
+      "step": 4163
+    },
+    {
+      "epoch": 1.9385474860335195,
+      "grad_norm": 0.5988548991098822,
+      "learning_rate": 1.5250951184037515e-06,
+      "loss": 0.3768,
+      "step": 4164
+    },
+    {
+      "epoch": 1.9390130353817505,
+      "grad_norm": 0.6082366680238571,
+      "learning_rate": 1.5239137096089386e-06,
+      "loss": 0.382,
+      "step": 4165
+    },
+    {
+      "epoch": 1.9394785847299814,
+      "grad_norm": 0.5732351003183848,
+      "learning_rate": 1.522732557938464e-06,
+      "loss": 0.374,
+      "step": 4166
+    },
+    {
+      "epoch": 1.9399441340782122,
+      "grad_norm": 0.571216536774572,
+      "learning_rate": 1.5215516637034738e-06,
+      "loss": 0.3855,
+      "step": 4167
+    },
+    {
+      "epoch": 1.9404096834264433,
+      "grad_norm": 0.6036760452785647,
+      "learning_rate": 1.520371027215043e-06,
+      "loss": 0.3765,
+      "step": 4168
+    },
+    {
+      "epoch": 1.940875232774674,
+      "grad_norm": 0.5864018780743491,
+      "learning_rate": 1.5191906487841784e-06,
+      "loss": 0.3901,
+      "step": 4169
+    },
+    {
+      "epoch": 1.941340782122905,
+      "grad_norm": 0.5922430835574898,
+      "learning_rate": 1.5180105287218194e-06,
+      "loss": 0.3664,
+      "step": 4170
+    },
+    {
+      "epoch": 1.941806331471136,
+      "grad_norm": 0.5689145810146498,
+      "learning_rate": 1.51683066733884e-06,
+      "loss": 0.3868,
+      "step": 4171
+    },
+    {
+      "epoch": 1.9422718808193669,
+      "grad_norm": 0.6190551591524884,
+      "learning_rate": 1.5156510649460415e-06,
+      "loss": 0.383,
+      "step": 4172
+    },
+    {
+      "epoch": 1.9427374301675977,
+      "grad_norm": 0.5952864957580766,
+      "learning_rate": 1.5144717218541604e-06,
+      "loss": 0.3982,
+      "step": 4173
+    },
+    {
+      "epoch": 1.9432029795158288,
+      "grad_norm": 0.6003868293494042,
+      "learning_rate": 1.5132926383738627e-06,
+      "loss": 0.3754,
+      "step": 4174
+    },
+    {
+      "epoch": 1.9436685288640596,
+      "grad_norm": 0.5495346278728725,
+      "learning_rate": 1.512113814815748e-06,
+      "loss": 0.3647,
+      "step": 4175
+    },
+    {
+      "epoch": 1.9441340782122905,
+      "grad_norm": 0.6267088785929283,
+      "learning_rate": 1.5109352514903464e-06,
+      "loss": 0.3777,
+      "step": 4176
+    },
+    {
+      "epoch": 1.9445996275605215,
+      "grad_norm": 0.5653090698006431,
+      "learning_rate": 1.5097569487081187e-06,
+      "loss": 0.3673,
+      "step": 4177
+    },
+    {
+      "epoch": 1.9450651769087524,
+      "grad_norm": 0.5847277495345116,
+      "learning_rate": 1.5085789067794581e-06,
+      "loss": 0.3831,
+      "step": 4178
+    },
+    {
+      "epoch": 1.9455307262569832,
+      "grad_norm": 0.6231390143248772,
+      "learning_rate": 1.50740112601469e-06,
+      "loss": 0.39,
+      "step": 4179
+    },
+    {
+      "epoch": 1.9459962756052143,
+      "grad_norm": 0.6115062998029059,
+      "learning_rate": 1.5062236067240687e-06,
+      "loss": 0.3726,
+      "step": 4180
+    },
+    {
+      "epoch": 1.946461824953445,
+      "grad_norm": 0.5862806015515976,
+      "learning_rate": 1.5050463492177812e-06,
+      "loss": 0.392,
+      "step": 4181
+    },
+    {
+      "epoch": 1.946927374301676,
+      "grad_norm": 0.5611590903999587,
+      "learning_rate": 1.5038693538059442e-06,
+      "loss": 0.3769,
+      "step": 4182
+    },
+    {
+      "epoch": 1.947392923649907,
+      "grad_norm": 0.6119732756534338,
+      "learning_rate": 1.5026926207986078e-06,
+      "loss": 0.3734,
+      "step": 4183
+    },
+    {
+      "epoch": 1.9478584729981379,
+      "grad_norm": 0.5828036210024661,
+      "learning_rate": 1.5015161505057507e-06,
+      "loss": 0.3851,
+      "step": 4184
+    },
+    {
+      "epoch": 1.9483240223463687,
+      "grad_norm": 0.5612021167851985,
+      "learning_rate": 1.500339943237284e-06,
+      "loss": 0.3755,
+      "step": 4185
+    },
+    {
+      "epoch": 1.9487895716945998,
+      "grad_norm": 0.6240574943433584,
+      "learning_rate": 1.499163999303046e-06,
+      "loss": 0.3827,
+      "step": 4186
+    },
+    {
+      "epoch": 1.9492551210428304,
+      "grad_norm": 0.5444475297733026,
+      "learning_rate": 1.4979883190128111e-06,
+      "loss": 0.3741,
+      "step": 4187
+    },
+    {
+      "epoch": 1.9497206703910615,
+      "grad_norm": 0.5955601721051783,
+      "learning_rate": 1.4968129026762813e-06,
+      "loss": 0.376,
+      "step": 4188
+    },
+    {
+      "epoch": 1.9501862197392925,
+      "grad_norm": 0.5579386835366269,
+      "learning_rate": 1.4956377506030877e-06,
+      "loss": 0.3651,
+      "step": 4189
+    },
+    {
+      "epoch": 1.9506517690875231,
+      "grad_norm": 0.6010195947645213,
+      "learning_rate": 1.494462863102793e-06,
+      "loss": 0.3724,
+      "step": 4190
+    },
+    {
+      "epoch": 1.9511173184357542,
+      "grad_norm": 0.5952774663116154,
+      "learning_rate": 1.493288240484893e-06,
+      "loss": 0.3915,
+      "step": 4191
+    },
+    {
+      "epoch": 1.9515828677839853,
+      "grad_norm": 0.5489979142952202,
+      "learning_rate": 1.4921138830588094e-06,
+      "loss": 0.3676,
+      "step": 4192
+    },
+    {
+      "epoch": 1.9520484171322159,
+      "grad_norm": 0.589914650865932,
+      "learning_rate": 1.4909397911338962e-06,
+      "loss": 0.3821,
+      "step": 4193
+    },
+    {
+      "epoch": 1.952513966480447,
+      "grad_norm": 0.5863475726257217,
+      "learning_rate": 1.4897659650194364e-06,
+      "loss": 0.364,
+      "step": 4194
+    },
+    {
+      "epoch": 1.9529795158286778,
+      "grad_norm": 0.5967426404545532,
+      "learning_rate": 1.4885924050246454e-06,
+      "loss": 0.3846,
+      "step": 4195
+    },
+    {
+      "epoch": 1.9534450651769086,
+      "grad_norm": 0.5874393576627261,
+      "learning_rate": 1.4874191114586655e-06,
+      "loss": 0.3867,
+      "step": 4196
+    },
+    {
+      "epoch": 1.9539106145251397,
+      "grad_norm": 0.5726368649302968,
+      "learning_rate": 1.4862460846305707e-06,
+      "loss": 0.3727,
+      "step": 4197
+    },
+    {
+      "epoch": 1.9543761638733705,
+      "grad_norm": 0.5900288962528988,
+      "learning_rate": 1.4850733248493631e-06,
+      "loss": 0.4,
+      "step": 4198
+    },
+    {
+      "epoch": 1.9548417132216014,
+      "grad_norm": 0.5656949428158995,
+      "learning_rate": 1.4839008324239774e-06,
+      "loss": 0.3642,
+      "step": 4199
+    },
+    {
+      "epoch": 1.9553072625698324,
+      "grad_norm": 0.5487958478200935,
+      "learning_rate": 1.4827286076632746e-06,
+      "loss": 0.3763,
+      "step": 4200
+    },
+    {
+      "epoch": 1.9557728119180633,
+      "grad_norm": 0.605010218008371,
+      "learning_rate": 1.4815566508760466e-06,
+      "loss": 0.3923,
+      "step": 4201
+    },
+    {
+      "epoch": 1.9562383612662941,
+      "grad_norm": 0.56731612242036,
+      "learning_rate": 1.4803849623710154e-06,
+      "loss": 0.3826,
+      "step": 4202
+    },
+    {
+      "epoch": 1.9567039106145252,
+      "grad_norm": 0.6045192578478097,
+      "learning_rate": 1.4792135424568315e-06,
+      "loss": 0.3894,
+      "step": 4203
+    },
+    {
+      "epoch": 1.957169459962756,
+      "grad_norm": 0.6033082713953026,
+      "learning_rate": 1.478042391442075e-06,
+      "loss": 0.374,
+      "step": 4204
+    },
+    {
+      "epoch": 1.9576350093109869,
+      "grad_norm": 0.5971328641511158,
+      "learning_rate": 1.4768715096352526e-06,
+      "loss": 0.3752,
+      "step": 4205
+    },
+    {
+      "epoch": 1.958100558659218,
+      "grad_norm": 0.6104273684990121,
+      "learning_rate": 1.4757008973448054e-06,
+      "loss": 0.3974,
+      "step": 4206
+    },
+    {
+      "epoch": 1.9585661080074488,
+      "grad_norm": 0.5921124417870145,
+      "learning_rate": 1.4745305548791e-06,
+      "loss": 0.3802,
+      "step": 4207
+    },
+    {
+      "epoch": 1.9590316573556796,
+      "grad_norm": 0.5688741347172743,
+      "learning_rate": 1.4733604825464309e-06,
+      "loss": 0.3597,
+      "step": 4208
+    },
+    {
+      "epoch": 1.9594972067039107,
+      "grad_norm": 0.610335066387889,
+      "learning_rate": 1.4721906806550232e-06,
+      "loss": 0.3825,
+      "step": 4209
+    },
+    {
+      "epoch": 1.9599627560521415,
+      "grad_norm": 0.6263331312819799,
+      "learning_rate": 1.4710211495130325e-06,
+      "loss": 0.3964,
+      "step": 4210
+    },
+    {
+      "epoch": 1.9604283054003724,
+      "grad_norm": 0.5791036579993047,
+      "learning_rate": 1.4698518894285396e-06,
+      "loss": 0.3663,
+      "step": 4211
+    },
+    {
+      "epoch": 1.9608938547486034,
+      "grad_norm": 0.6113808819965525,
+      "learning_rate": 1.4686829007095555e-06,
+      "loss": 0.3906,
+      "step": 4212
+    },
+    {
+      "epoch": 1.9613594040968343,
+      "grad_norm": 0.6179334839683084,
+      "learning_rate": 1.4675141836640188e-06,
+      "loss": 0.3756,
+      "step": 4213
+    },
+    {
+      "epoch": 1.9618249534450651,
+      "grad_norm": 0.5940454449888984,
+      "learning_rate": 1.4663457385997992e-06,
+      "loss": 0.3991,
+      "step": 4214
+    },
+    {
+      "epoch": 1.9622905027932962,
+      "grad_norm": 0.6047073855852426,
+      "learning_rate": 1.4651775658246922e-06,
+      "loss": 0.4101,
+      "step": 4215
+    },
+    {
+      "epoch": 1.962756052141527,
+      "grad_norm": 0.6627345341801596,
+      "learning_rate": 1.464009665646422e-06,
+      "loss": 0.3927,
+      "step": 4216
+    },
+    {
+      "epoch": 1.9632216014897579,
+      "grad_norm": 0.5848412521124394,
+      "learning_rate": 1.4628420383726407e-06,
+      "loss": 0.3736,
+      "step": 4217
+    },
+    {
+      "epoch": 1.963687150837989,
+      "grad_norm": 0.6080334783580243,
+      "learning_rate": 1.4616746843109305e-06,
+      "loss": 0.372,
+      "step": 4218
+    },
+    {
+      "epoch": 1.9641527001862198,
+      "grad_norm": 0.5689092040093611,
+      "learning_rate": 1.4605076037687993e-06,
+      "loss": 0.3899,
+      "step": 4219
+    },
+    {
+      "epoch": 1.9646182495344506,
+      "grad_norm": 0.6038705146007132,
+      "learning_rate": 1.4593407970536843e-06,
+      "loss": 0.3637,
+      "step": 4220
+    },
+    {
+      "epoch": 1.9650837988826817,
+      "grad_norm": 0.5717210488747502,
+      "learning_rate": 1.4581742644729494e-06,
+      "loss": 0.3801,
+      "step": 4221
+    },
+    {
+      "epoch": 1.9655493482309123,
+      "grad_norm": 0.60202596355087,
+      "learning_rate": 1.4570080063338885e-06,
+      "loss": 0.3764,
+      "step": 4222
+    },
+    {
+      "epoch": 1.9660148975791434,
+      "grad_norm": 0.57543241685234,
+      "learning_rate": 1.4558420229437197e-06,
+      "loss": 0.3868,
+      "step": 4223
+    },
+    {
+      "epoch": 1.9664804469273744,
+      "grad_norm": 0.6155027364697536,
+      "learning_rate": 1.4546763146095927e-06,
+      "loss": 0.4093,
+      "step": 4224
+    },
+    {
+      "epoch": 1.966945996275605,
+      "grad_norm": 0.5988877992291735,
+      "learning_rate": 1.4535108816385812e-06,
+      "loss": 0.3738,
+      "step": 4225
+    },
+    {
+      "epoch": 1.9674115456238361,
+      "grad_norm": 0.6130903493487578,
+      "learning_rate": 1.4523457243376882e-06,
+      "loss": 0.3828,
+      "step": 4226
+    },
+    {
+      "epoch": 1.9678770949720672,
+      "grad_norm": 0.5792773747544124,
+      "learning_rate": 1.4511808430138459e-06,
+      "loss": 0.3808,
+      "step": 4227
+    },
+    {
+      "epoch": 1.9683426443202978,
+      "grad_norm": 0.609359334178344,
+      "learning_rate": 1.4500162379739085e-06,
+      "loss": 0.3865,
+      "step": 4228
+    },
+    {
+      "epoch": 1.9688081936685289,
+      "grad_norm": 0.6029882042792395,
+      "learning_rate": 1.4488519095246627e-06,
+      "loss": 0.3939,
+      "step": 4229
+    },
+    {
+      "epoch": 1.9692737430167597,
+      "grad_norm": 0.5803204598956835,
+      "learning_rate": 1.447687857972821e-06,
+      "loss": 0.3789,
+      "step": 4230
+    },
+    {
+      "epoch": 1.9697392923649906,
+      "grad_norm": 0.558929983261978,
+      "learning_rate": 1.44652408362502e-06,
+      "loss": 0.363,
+      "step": 4231
+    },
+    {
+      "epoch": 1.9702048417132216,
+      "grad_norm": 0.590641013363504,
+      "learning_rate": 1.4453605867878276e-06,
+      "loss": 0.374,
+      "step": 4232
+    },
+    {
+      "epoch": 1.9706703910614525,
+      "grad_norm": 0.5720569181507261,
+      "learning_rate": 1.4441973677677345e-06,
+      "loss": 0.384,
+      "step": 4233
+    },
+    {
+      "epoch": 1.9711359404096833,
+      "grad_norm": 0.5729245015530351,
+      "learning_rate": 1.4430344268711617e-06,
+      "loss": 0.3813,
+      "step": 4234
+    },
+    {
+      "epoch": 1.9716014897579144,
+      "grad_norm": 0.6295581983271246,
+      "learning_rate": 1.441871764404456e-06,
+      "loss": 0.4,
+      "step": 4235
+    },
+    {
+      "epoch": 1.9720670391061452,
+      "grad_norm": 0.6189962504808142,
+      "learning_rate": 1.4407093806738898e-06,
+      "loss": 0.3858,
+      "step": 4236
+    },
+    {
+      "epoch": 1.972532588454376,
+      "grad_norm": 0.5573010291689459,
+      "learning_rate": 1.4395472759856599e-06,
+      "loss": 0.3601,
+      "step": 4237
+    },
+    {
+      "epoch": 1.9729981378026071,
+      "grad_norm": 0.5900118777804311,
+      "learning_rate": 1.4383854506458972e-06,
+      "loss": 0.3682,
+      "step": 4238
+    },
+    {
+      "epoch": 1.973463687150838,
+      "grad_norm": 0.8109489430967952,
+      "learning_rate": 1.4372239049606513e-06,
+      "loss": 0.4002,
+      "step": 4239
+    },
+    {
+      "epoch": 1.9739292364990688,
+      "grad_norm": 0.6046257488414593,
+      "learning_rate": 1.4360626392359006e-06,
+      "loss": 0.3752,
+      "step": 4240
+    },
+    {
+      "epoch": 1.9743947858472999,
+      "grad_norm": 0.6102004056850058,
+      "learning_rate": 1.4349016537775506e-06,
+      "loss": 0.3808,
+      "step": 4241
+    },
+    {
+      "epoch": 1.9748603351955307,
+      "grad_norm": 0.6156350001062423,
+      "learning_rate": 1.4337409488914338e-06,
+      "loss": 0.3572,
+      "step": 4242
+    },
+    {
+      "epoch": 1.9753258845437616,
+      "grad_norm": 0.5827534380694583,
+      "learning_rate": 1.4325805248833052e-06,
+      "loss": 0.3756,
+      "step": 4243
+    },
+    {
+      "epoch": 1.9757914338919926,
+      "grad_norm": 0.5960429084583675,
+      "learning_rate": 1.4314203820588495e-06,
+      "loss": 0.3838,
+      "step": 4244
+    },
+    {
+      "epoch": 1.9762569832402235,
+      "grad_norm": 0.579268294722258,
+      "learning_rate": 1.4302605207236764e-06,
+      "loss": 0.3801,
+      "step": 4245
+    },
+    {
+      "epoch": 1.9767225325884543,
+      "grad_norm": 0.5811060972935937,
+      "learning_rate": 1.4291009411833195e-06,
+      "loss": 0.3781,
+      "step": 4246
+    },
+    {
+      "epoch": 1.9771880819366854,
+      "grad_norm": 0.5963459158748717,
+      "learning_rate": 1.4279416437432414e-06,
+      "loss": 0.3677,
+      "step": 4247
+    },
+    {
+      "epoch": 1.9776536312849162,
+      "grad_norm": 0.5616423107249866,
+      "learning_rate": 1.426782628708826e-06,
+      "loss": 0.3948,
+      "step": 4248
+    },
+    {
+      "epoch": 1.978119180633147,
+      "grad_norm": 0.5962903785306302,
+      "learning_rate": 1.4256238963853875e-06,
+      "loss": 0.3507,
+      "step": 4249
+    },
+    {
+      "epoch": 1.9785847299813781,
+      "grad_norm": 0.5739217441457962,
+      "learning_rate": 1.4244654470781638e-06,
+      "loss": 0.364,
+      "step": 4250
+    },
+    {
+      "epoch": 1.979050279329609,
+      "grad_norm": 0.5883711384903131,
+      "learning_rate": 1.4233072810923163e-06,
+      "loss": 0.3806,
+      "step": 4251
+    },
+    {
+      "epoch": 1.9795158286778398,
+      "grad_norm": 0.5798122427409176,
+      "learning_rate": 1.4221493987329344e-06,
+      "loss": 0.3811,
+      "step": 4252
+    },
+    {
+      "epoch": 1.9799813780260709,
+      "grad_norm": 0.5879341899185615,
+      "learning_rate": 1.4209918003050328e-06,
+      "loss": 0.3728,
+      "step": 4253
+    },
+    {
+      "epoch": 1.9804469273743017,
+      "grad_norm": 0.5773696015465853,
+      "learning_rate": 1.4198344861135483e-06,
+      "loss": 0.3722,
+      "step": 4254
+    },
+    {
+      "epoch": 1.9809124767225326,
+      "grad_norm": 0.6197193275319087,
+      "learning_rate": 1.4186774564633474e-06,
+      "loss": 0.3964,
+      "step": 4255
+    },
+    {
+      "epoch": 1.9813780260707636,
+      "grad_norm": 0.574736662549977,
+      "learning_rate": 1.4175207116592168e-06,
+      "loss": 0.3872,
+      "step": 4256
+    },
+    {
+      "epoch": 1.9818435754189943,
+      "grad_norm": 0.6657788343473079,
+      "learning_rate": 1.416364252005873e-06,
+      "loss": 0.3708,
+      "step": 4257
+    },
+    {
+      "epoch": 1.9823091247672253,
+      "grad_norm": 0.5977240020445029,
+      "learning_rate": 1.4152080778079525e-06,
+      "loss": 0.3881,
+      "step": 4258
+    },
+    {
+      "epoch": 1.9827746741154564,
+      "grad_norm": 0.6041000946971748,
+      "learning_rate": 1.4140521893700215e-06,
+      "loss": 0.382,
+      "step": 4259
+    },
+    {
+      "epoch": 1.983240223463687,
+      "grad_norm": 0.5777404267129926,
+      "learning_rate": 1.412896586996566e-06,
+      "loss": 0.378,
+      "step": 4260
+    },
+    {
+      "epoch": 1.983705772811918,
+      "grad_norm": 0.5435123888491419,
+      "learning_rate": 1.411741270992001e-06,
+      "loss": 0.3661,
+      "step": 4261
+    },
+    {
+      "epoch": 1.9841713221601491,
+      "grad_norm": 0.6369685313865813,
+      "learning_rate": 1.4105862416606642e-06,
+      "loss": 0.3965,
+      "step": 4262
+    },
+    {
+      "epoch": 1.9846368715083798,
+      "grad_norm": 0.6043207071659663,
+      "learning_rate": 1.4094314993068165e-06,
+      "loss": 0.3798,
+      "step": 4263
+    },
+    {
+      "epoch": 1.9851024208566108,
+      "grad_norm": 0.5917816727060028,
+      "learning_rate": 1.408277044234645e-06,
+      "loss": 0.3845,
+      "step": 4264
+    },
+    {
+      "epoch": 1.9855679702048417,
+      "grad_norm": 0.6168581559240094,
+      "learning_rate": 1.407122876748262e-06,
+      "loss": 0.3786,
+      "step": 4265
+    },
+    {
+      "epoch": 1.9860335195530725,
+      "grad_norm": 0.5814410629376398,
+      "learning_rate": 1.4059689971517007e-06,
+      "loss": 0.3947,
+      "step": 4266
+    },
+    {
+      "epoch": 1.9864990689013036,
+      "grad_norm": 0.5896870776750979,
+      "learning_rate": 1.4048154057489221e-06,
+      "loss": 0.3784,
+      "step": 4267
+    },
+    {
+      "epoch": 1.9869646182495344,
+      "grad_norm": 0.6005525083226447,
+      "learning_rate": 1.403662102843808e-06,
+      "loss": 0.3859,
+      "step": 4268
+    },
+    {
+      "epoch": 1.9874301675977653,
+      "grad_norm": 0.5655341495390451,
+      "learning_rate": 1.4025090887401665e-06,
+      "loss": 0.3756,
+      "step": 4269
+    },
+    {
+      "epoch": 1.9878957169459963,
+      "grad_norm": 0.5909320364393974,
+      "learning_rate": 1.4013563637417298e-06,
+      "loss": 0.3735,
+      "step": 4270
+    },
+    {
+      "epoch": 1.9883612662942272,
+      "grad_norm": 0.5736504227092826,
+      "learning_rate": 1.4002039281521523e-06,
+      "loss": 0.3856,
+      "step": 4271
+    },
+    {
+      "epoch": 1.988826815642458,
+      "grad_norm": 0.573228767560183,
+      "learning_rate": 1.3990517822750114e-06,
+      "loss": 0.3697,
+      "step": 4272
+    },
+    {
+      "epoch": 1.989292364990689,
+      "grad_norm": 0.5871086307201181,
+      "learning_rate": 1.3978999264138126e-06,
+      "loss": 0.3671,
+      "step": 4273
+    },
+    {
+      "epoch": 1.98975791433892,
+      "grad_norm": 0.6137613429234426,
+      "learning_rate": 1.3967483608719812e-06,
+      "loss": 0.3793,
+      "step": 4274
+    },
+    {
+      "epoch": 1.9902234636871508,
+      "grad_norm": 0.5891483970152263,
+      "learning_rate": 1.3955970859528656e-06,
+      "loss": 0.3815,
+      "step": 4275
+    },
+    {
+      "epoch": 1.9906890130353818,
+      "grad_norm": 0.5776923809740643,
+      "learning_rate": 1.3944461019597398e-06,
+      "loss": 0.3906,
+      "step": 4276
+    },
+    {
+      "epoch": 1.9911545623836127,
+      "grad_norm": 0.6082610332890064,
+      "learning_rate": 1.3932954091958012e-06,
+      "loss": 0.3953,
+      "step": 4277
+    },
+    {
+      "epoch": 1.9916201117318435,
+      "grad_norm": 0.5795348405914476,
+      "learning_rate": 1.3921450079641682e-06,
+      "loss": 0.3828,
+      "step": 4278
+    },
+    {
+      "epoch": 1.9920856610800746,
+      "grad_norm": 0.6213502270944079,
+      "learning_rate": 1.3909948985678856e-06,
+      "loss": 0.3724,
+      "step": 4279
+    },
+    {
+      "epoch": 1.9925512104283054,
+      "grad_norm": 0.6758292615940856,
+      "learning_rate": 1.3898450813099173e-06,
+      "loss": 0.3661,
+      "step": 4280
+    },
+    {
+      "epoch": 1.9930167597765363,
+      "grad_norm": 0.6059120729171888,
+      "learning_rate": 1.3886955564931537e-06,
+      "loss": 0.3743,
+      "step": 4281
+    },
+    {
+      "epoch": 1.9934823091247673,
+      "grad_norm": 0.5880724039419701,
+      "learning_rate": 1.3875463244204082e-06,
+      "loss": 0.3537,
+      "step": 4282
+    },
+    {
+      "epoch": 1.9939478584729982,
+      "grad_norm": 0.6288240556214038,
+      "learning_rate": 1.3863973853944134e-06,
+      "loss": 0.384,
+      "step": 4283
+    },
+    {
+      "epoch": 1.994413407821229,
+      "grad_norm": 0.5826376726840172,
+      "learning_rate": 1.3852487397178288e-06,
+      "loss": 0.3758,
+      "step": 4284
+    },
+    {
+      "epoch": 1.99487895716946,
+      "grad_norm": 0.6100620311027719,
+      "learning_rate": 1.384100387693235e-06,
+      "loss": 0.3693,
+      "step": 4285
+    },
+    {
+      "epoch": 1.995344506517691,
+      "grad_norm": 0.5592011678267989,
+      "learning_rate": 1.3829523296231345e-06,
+      "loss": 0.3672,
+      "step": 4286
+    },
+    {
+      "epoch": 1.9958100558659218,
+      "grad_norm": 0.5857893359110665,
+      "learning_rate": 1.3818045658099532e-06,
+      "loss": 0.3701,
+      "step": 4287
+    },
+    {
+      "epoch": 1.9962756052141528,
+      "grad_norm": 0.6192938738843312,
+      "learning_rate": 1.3806570965560403e-06,
+      "loss": 0.3898,
+      "step": 4288
+    },
+    {
+      "epoch": 1.9967411545623837,
+      "grad_norm": 0.5850159566098321,
+      "learning_rate": 1.379509922163665e-06,
+      "loss": 0.3859,
+      "step": 4289
+    },
+    {
+      "epoch": 1.9972067039106145,
+      "grad_norm": 0.6147771646482425,
+      "learning_rate": 1.3783630429350223e-06,
+      "loss": 0.3783,
+      "step": 4290
+    },
+    {
+      "epoch": 1.9976722532588456,
+      "grad_norm": 0.5914223574582773,
+      "learning_rate": 1.377216459172225e-06,
+      "loss": 0.3871,
+      "step": 4291
+    },
+    {
+      "epoch": 1.9981378026070762,
+      "grad_norm": 0.6107891328457378,
+      "learning_rate": 1.3760701711773119e-06,
+      "loss": 0.4087,
+      "step": 4292
+    },
+    {
+      "epoch": 1.9986033519553073,
+      "grad_norm": 0.6058786254149864,
+      "learning_rate": 1.3749241792522433e-06,
+      "loss": 0.3654,
+      "step": 4293
+    },
+    {
+      "epoch": 1.9990689013035383,
+      "grad_norm": 0.5823684468175119,
+      "learning_rate": 1.3737784836988999e-06,
+      "loss": 0.3923,
+      "step": 4294
+    },
+    {
+      "epoch": 1.999534450651769,
+      "grad_norm": 0.5802791191955234,
+      "learning_rate": 1.3726330848190833e-06,
+      "loss": 0.3747,
+      "step": 4295
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7496208415782099,
+      "learning_rate": 1.3714879829145221e-06,
+      "loss": 0.3249,
+      "step": 4296
+    },
+    {
+      "epoch": 2.000465549348231,
+      "grad_norm": 1.0519830011788776,
+      "learning_rate": 1.3703431782868625e-06,
+      "loss": 0.3454,
+      "step": 4297
+    },
+    {
+      "epoch": 2.0009310986964617,
+      "grad_norm": 1.0290149878636687,
+      "learning_rate": 1.3691986712376712e-06,
+      "loss": 0.3421,
+      "step": 4298
+    },
+    {
+      "epoch": 2.0013966480446927,
+      "grad_norm": 0.8902390856567796,
+      "learning_rate": 1.36805446206844e-06,
+      "loss": 0.32,
+      "step": 4299
+    },
+    {
+      "epoch": 2.001862197392924,
+      "grad_norm": 0.7791808329634747,
+      "learning_rate": 1.3669105510805819e-06,
+      "loss": 0.3265,
+      "step": 4300
+    },
+    {
+      "epoch": 2.0023277467411544,
+      "grad_norm": 0.7150779197080821,
+      "learning_rate": 1.3657669385754283e-06,
+      "loss": 0.3176,
+      "step": 4301
+    },
+    {
+      "epoch": 2.0027932960893855,
+      "grad_norm": 0.7211588013591923,
+      "learning_rate": 1.3646236248542354e-06,
+      "loss": 0.3217,
+      "step": 4302
+    },
+    {
+      "epoch": 2.0032588454376166,
+      "grad_norm": 0.7933013573868798,
+      "learning_rate": 1.363480610218178e-06,
+      "loss": 0.3351,
+      "step": 4303
+    },
+    {
+      "epoch": 2.003724394785847,
+      "grad_norm": 0.7865311440515367,
+      "learning_rate": 1.3623378949683538e-06,
+      "loss": 0.3168,
+      "step": 4304
+    },
+    {
+      "epoch": 2.0041899441340782,
+      "grad_norm": 0.8509509345074884,
+      "learning_rate": 1.3611954794057824e-06,
+      "loss": 0.3317,
+      "step": 4305
+    },
+    {
+      "epoch": 2.0046554934823093,
+      "grad_norm": 0.7588583375532485,
+      "learning_rate": 1.3600533638314012e-06,
+      "loss": 0.325,
+      "step": 4306
+    },
+    {
+      "epoch": 2.00512104283054,
+      "grad_norm": 0.6820012028821241,
+      "learning_rate": 1.3589115485460713e-06,
+      "loss": 0.3219,
+      "step": 4307
+    },
+    {
+      "epoch": 2.005586592178771,
+      "grad_norm": 0.6413710476559428,
+      "learning_rate": 1.357770033850575e-06,
+      "loss": 0.3258,
+      "step": 4308
+    },
+    {
+      "epoch": 2.006052141527002,
+      "grad_norm": 0.6192515149487482,
+      "learning_rate": 1.3566288200456131e-06,
+      "loss": 0.3407,
+      "step": 4309
+    },
+    {
+      "epoch": 2.0065176908752327,
+      "grad_norm": 0.64024820015502,
+      "learning_rate": 1.35548790743181e-06,
+      "loss": 0.331,
+      "step": 4310
+    },
+    {
+      "epoch": 2.0069832402234637,
+      "grad_norm": 0.6454290853819612,
+      "learning_rate": 1.3543472963097072e-06,
+      "loss": 0.3301,
+      "step": 4311
+    },
+    {
+      "epoch": 2.007448789571695,
+      "grad_norm": 0.7140895373227563,
+      "learning_rate": 1.3532069869797697e-06,
+      "loss": 0.3276,
+      "step": 4312
+    },
+    {
+      "epoch": 2.0079143389199254,
+      "grad_norm": 0.6340445687117205,
+      "learning_rate": 1.3520669797423832e-06,
+      "loss": 0.3218,
+      "step": 4313
+    },
+    {
+      "epoch": 2.0083798882681565,
+      "grad_norm": 0.6479778257910266,
+      "learning_rate": 1.3509272748978519e-06,
+      "loss": 0.3263,
+      "step": 4314
+    },
+    {
+      "epoch": 2.008845437616387,
+      "grad_norm": 0.6263695862820127,
+      "learning_rate": 1.3497878727463987e-06,
+      "loss": 0.3354,
+      "step": 4315
+    },
+    {
+      "epoch": 2.009310986964618,
+      "grad_norm": 0.6196575147258334,
+      "learning_rate": 1.3486487735881736e-06,
+      "loss": 0.3161,
+      "step": 4316
+    },
+    {
+      "epoch": 2.0097765363128492,
+      "grad_norm": 0.610124270189082,
+      "learning_rate": 1.3475099777232402e-06,
+      "loss": 0.3203,
+      "step": 4317
+    },
+    {
+      "epoch": 2.01024208566108,
+      "grad_norm": 0.6255871639632992,
+      "learning_rate": 1.3463714854515841e-06,
+      "loss": 0.3415,
+      "step": 4318
+    },
+    {
+      "epoch": 2.010707635009311,
+      "grad_norm": 0.608231633779378,
+      "learning_rate": 1.3452332970731113e-06,
+      "loss": 0.3355,
+      "step": 4319
+    },
+    {
+      "epoch": 2.011173184357542,
+      "grad_norm": 0.6136043375232215,
+      "learning_rate": 1.3440954128876489e-06,
+      "loss": 0.3273,
+      "step": 4320
+    },
+    {
+      "epoch": 2.0116387337057726,
+      "grad_norm": 0.5990037846343331,
+      "learning_rate": 1.3429578331949407e-06,
+      "loss": 0.3183,
+      "step": 4321
+    },
+    {
+      "epoch": 2.0121042830540037,
+      "grad_norm": 0.6151510259000459,
+      "learning_rate": 1.3418205582946546e-06,
+      "loss": 0.3147,
+      "step": 4322
+    },
+    {
+      "epoch": 2.0125698324022347,
+      "grad_norm": 0.6117614220929611,
+      "learning_rate": 1.3406835884863736e-06,
+      "loss": 0.3325,
+      "step": 4323
+    },
+    {
+      "epoch": 2.0130353817504654,
+      "grad_norm": 0.6415103722145125,
+      "learning_rate": 1.3395469240696035e-06,
+      "loss": 0.3267,
+      "step": 4324
+    },
+    {
+      "epoch": 2.0135009310986964,
+      "grad_norm": 0.5861321059174398,
+      "learning_rate": 1.3384105653437699e-06,
+      "loss": 0.3276,
+      "step": 4325
+    },
+    {
+      "epoch": 2.0139664804469275,
+      "grad_norm": 0.5918491291258902,
+      "learning_rate": 1.3372745126082143e-06,
+      "loss": 0.3202,
+      "step": 4326
+    },
+    {
+      "epoch": 2.014432029795158,
+      "grad_norm": 0.5948330223787701,
+      "learning_rate": 1.3361387661622015e-06,
+      "loss": 0.342,
+      "step": 4327
+    },
+    {
+      "epoch": 2.014897579143389,
+      "grad_norm": 0.5951123474511156,
+      "learning_rate": 1.335003326304915e-06,
+      "loss": 0.309,
+      "step": 4328
+    },
+    {
+      "epoch": 2.0153631284916202,
+      "grad_norm": 0.6279740318309927,
+      "learning_rate": 1.3338681933354547e-06,
+      "loss": 0.3319,
+      "step": 4329
+    },
+    {
+      "epoch": 2.015828677839851,
+      "grad_norm": 0.5995322216649176,
+      "learning_rate": 1.3327333675528436e-06,
+      "loss": 0.3481,
+      "step": 4330
+    },
+    {
+      "epoch": 2.016294227188082,
+      "grad_norm": 0.6154471054811731,
+      "learning_rate": 1.3315988492560194e-06,
+      "loss": 0.3259,
+      "step": 4331
+    },
+    {
+      "epoch": 2.016759776536313,
+      "grad_norm": 0.6482775266471791,
+      "learning_rate": 1.3304646387438428e-06,
+      "loss": 0.3458,
+      "step": 4332
+    },
+    {
+      "epoch": 2.0172253258845436,
+      "grad_norm": 0.5974054988630519,
+      "learning_rate": 1.3293307363150925e-06,
+      "loss": 0.3336,
+      "step": 4333
+    },
+    {
+      "epoch": 2.0176908752327747,
+      "grad_norm": 0.6608628626216749,
+      "learning_rate": 1.3281971422684631e-06,
+      "loss": 0.3153,
+      "step": 4334
+    },
+    {
+      "epoch": 2.0181564245810057,
+      "grad_norm": 0.5660461414965359,
+      "learning_rate": 1.3270638569025718e-06,
+      "loss": 0.3272,
+      "step": 4335
+    },
+    {
+      "epoch": 2.0186219739292364,
+      "grad_norm": 0.5787709743819552,
+      "learning_rate": 1.3259308805159537e-06,
+      "loss": 0.3301,
+      "step": 4336
+    },
+    {
+      "epoch": 2.0190875232774674,
+      "grad_norm": 0.6034892166250981,
+      "learning_rate": 1.3247982134070605e-06,
+      "loss": 0.3326,
+      "step": 4337
+    },
+    {
+      "epoch": 2.0195530726256985,
+      "grad_norm": 0.6211209935659563,
+      "learning_rate": 1.323665855874263e-06,
+      "loss": 0.3377,
+      "step": 4338
+    },
+    {
+      "epoch": 2.020018621973929,
+      "grad_norm": 0.6219866781656952,
+      "learning_rate": 1.3225338082158518e-06,
+      "loss": 0.3267,
+      "step": 4339
+    },
+    {
+      "epoch": 2.02048417132216,
+      "grad_norm": 0.6073476214269514,
+      "learning_rate": 1.3214020707300362e-06,
+      "loss": 0.3427,
+      "step": 4340
+    },
+    {
+      "epoch": 2.0209497206703912,
+      "grad_norm": 0.6201400059232377,
+      "learning_rate": 1.320270643714941e-06,
+      "loss": 0.3474,
+      "step": 4341
+    },
+    {
+      "epoch": 2.021415270018622,
+      "grad_norm": 0.5914490576296886,
+      "learning_rate": 1.319139527468612e-06,
+      "loss": 0.3189,
+      "step": 4342
+    },
+    {
+      "epoch": 2.021880819366853,
+      "grad_norm": 0.6033597408457637,
+      "learning_rate": 1.3180087222890125e-06,
+      "loss": 0.329,
+      "step": 4343
+    },
+    {
+      "epoch": 2.022346368715084,
+      "grad_norm": 0.6130966319469149,
+      "learning_rate": 1.3168782284740223e-06,
+      "loss": 0.3203,
+      "step": 4344
+    },
+    {
+      "epoch": 2.0228119180633146,
+      "grad_norm": 0.6428740294125902,
+      "learning_rate": 1.315748046321442e-06,
+      "loss": 0.347,
+      "step": 4345
+    },
+    {
+      "epoch": 2.0232774674115457,
+      "grad_norm": 0.5890002183689418,
+      "learning_rate": 1.3146181761289863e-06,
+      "loss": 0.3173,
+      "step": 4346
+    },
+    {
+      "epoch": 2.0237430167597767,
+      "grad_norm": 0.5900849763938111,
+      "learning_rate": 1.3134886181942914e-06,
+      "loss": 0.3131,
+      "step": 4347
+    },
+    {
+      "epoch": 2.0242085661080074,
+      "grad_norm": 0.5932309405171788,
+      "learning_rate": 1.3123593728149098e-06,
+      "loss": 0.308,
+      "step": 4348
+    },
+    {
+      "epoch": 2.0246741154562384,
+      "grad_norm": 0.6069595484536785,
+      "learning_rate": 1.3112304402883113e-06,
+      "loss": 0.3265,
+      "step": 4349
+    },
+    {
+      "epoch": 2.0251396648044695,
+      "grad_norm": 0.5876885096831961,
+      "learning_rate": 1.3101018209118816e-06,
+      "loss": 0.3243,
+      "step": 4350
+    },
+    {
+      "epoch": 2.0256052141527,
+      "grad_norm": 0.5808640391607712,
+      "learning_rate": 1.3089735149829292e-06,
+      "loss": 0.3132,
+      "step": 4351
+    },
+    {
+      "epoch": 2.026070763500931,
+      "grad_norm": 0.5884400897237638,
+      "learning_rate": 1.3078455227986756e-06,
+      "loss": 0.3275,
+      "step": 4352
+    },
+    {
+      "epoch": 2.026536312849162,
+      "grad_norm": 0.5976918844122031,
+      "learning_rate": 1.306717844656259e-06,
+      "loss": 0.313,
+      "step": 4353
+    },
+    {
+      "epoch": 2.027001862197393,
+      "grad_norm": 0.5746110006213196,
+      "learning_rate": 1.3055904808527384e-06,
+      "loss": 0.3196,
+      "step": 4354
+    },
+    {
+      "epoch": 2.027467411545624,
+      "grad_norm": 0.6267802590310583,
+      "learning_rate": 1.3044634316850885e-06,
+      "loss": 0.3244,
+      "step": 4355
+    },
+    {
+      "epoch": 2.0279329608938546,
+      "grad_norm": 0.6039527658081338,
+      "learning_rate": 1.3033366974501994e-06,
+      "loss": 0.3197,
+      "step": 4356
+    },
+    {
+      "epoch": 2.0283985102420856,
+      "grad_norm": 0.6208069025758972,
+      "learning_rate": 1.3022102784448814e-06,
+      "loss": 0.3415,
+      "step": 4357
+    },
+    {
+      "epoch": 2.0288640595903167,
+      "grad_norm": 0.6015013116248445,
+      "learning_rate": 1.3010841749658582e-06,
+      "loss": 0.3242,
+      "step": 4358
+    },
+    {
+      "epoch": 2.0293296089385473,
+      "grad_norm": 0.6194720173528768,
+      "learning_rate": 1.2999583873097733e-06,
+      "loss": 0.3289,
+      "step": 4359
+    },
+    {
+      "epoch": 2.0297951582867784,
+      "grad_norm": 0.5902610346260678,
+      "learning_rate": 1.2988329157731865e-06,
+      "loss": 0.3533,
+      "step": 4360
+    },
+    {
+      "epoch": 2.0302607076350094,
+      "grad_norm": 0.5889672470394793,
+      "learning_rate": 1.2977077606525723e-06,
+      "loss": 0.3379,
+      "step": 4361
+    },
+    {
+      "epoch": 2.03072625698324,
+      "grad_norm": 0.6178203164846521,
+      "learning_rate": 1.2965829222443238e-06,
+      "loss": 0.3365,
+      "step": 4362
+    },
+    {
+      "epoch": 2.031191806331471,
+      "grad_norm": 0.6335165513189629,
+      "learning_rate": 1.2954584008447515e-06,
+      "loss": 0.3268,
+      "step": 4363
+    },
+    {
+      "epoch": 2.031657355679702,
+      "grad_norm": 0.702093905923223,
+      "learning_rate": 1.2943341967500792e-06,
+      "loss": 0.3387,
+      "step": 4364
+    },
+    {
+      "epoch": 2.032122905027933,
+      "grad_norm": 0.6310620311604662,
+      "learning_rate": 1.2932103102564507e-06,
+      "loss": 0.3469,
+      "step": 4365
+    },
+    {
+      "epoch": 2.032588454376164,
+      "grad_norm": 0.6089443448798454,
+      "learning_rate": 1.2920867416599225e-06,
+      "loss": 0.3108,
+      "step": 4366
+    },
+    {
+      "epoch": 2.033054003724395,
+      "grad_norm": 0.5700748189494752,
+      "learning_rate": 1.2909634912564706e-06,
+      "loss": 0.3181,
+      "step": 4367
+    },
+    {
+      "epoch": 2.0335195530726256,
+      "grad_norm": 0.6009739297772652,
+      "learning_rate": 1.2898405593419866e-06,
+      "loss": 0.3203,
+      "step": 4368
+    },
+    {
+      "epoch": 2.0339851024208566,
+      "grad_norm": 0.6805951636114382,
+      "learning_rate": 1.2887179462122761e-06,
+      "loss": 0.3487,
+      "step": 4369
+    },
+    {
+      "epoch": 2.0344506517690877,
+      "grad_norm": 0.5956952013418487,
+      "learning_rate": 1.2875956521630611e-06,
+      "loss": 0.3175,
+      "step": 4370
+    },
+    {
+      "epoch": 2.0349162011173183,
+      "grad_norm": 0.5744458311094073,
+      "learning_rate": 1.2864736774899834e-06,
+      "loss": 0.3185,
+      "step": 4371
+    },
+    {
+      "epoch": 2.0353817504655494,
+      "grad_norm": 0.5970976745876662,
+      "learning_rate": 1.2853520224885966e-06,
+      "loss": 0.321,
+      "step": 4372
+    },
+    {
+      "epoch": 2.0358472998137804,
+      "grad_norm": 0.6130415603537711,
+      "learning_rate": 1.2842306874543698e-06,
+      "loss": 0.3383,
+      "step": 4373
+    },
+    {
+      "epoch": 2.036312849162011,
+      "grad_norm": 0.5993939247361985,
+      "learning_rate": 1.2831096726826907e-06,
+      "loss": 0.3289,
+      "step": 4374
+    },
+    {
+      "epoch": 2.036778398510242,
+      "grad_norm": 0.5734575961993081,
+      "learning_rate": 1.2819889784688622e-06,
+      "loss": 0.3126,
+      "step": 4375
+    },
+    {
+      "epoch": 2.037243947858473,
+      "grad_norm": 0.594955933418023,
+      "learning_rate": 1.2808686051080992e-06,
+      "loss": 0.337,
+      "step": 4376
+    },
+    {
+      "epoch": 2.037709497206704,
+      "grad_norm": 0.5892925915705512,
+      "learning_rate": 1.2797485528955368e-06,
+      "loss": 0.3222,
+      "step": 4377
+    },
+    {
+      "epoch": 2.038175046554935,
+      "grad_norm": 0.6272045583504208,
+      "learning_rate": 1.2786288221262237e-06,
+      "loss": 0.2981,
+      "step": 4378
+    },
+    {
+      "epoch": 2.038640595903166,
+      "grad_norm": 0.6018109996576917,
+      "learning_rate": 1.2775094130951214e-06,
+      "loss": 0.3311,
+      "step": 4379
+    },
+    {
+      "epoch": 2.0391061452513966,
+      "grad_norm": 0.5874363365456656,
+      "learning_rate": 1.276390326097111e-06,
+      "loss": 0.3303,
+      "step": 4380
+    },
+    {
+      "epoch": 2.0395716945996276,
+      "grad_norm": 0.5944450183719021,
+      "learning_rate": 1.275271561426985e-06,
+      "loss": 0.3331,
+      "step": 4381
+    },
+    {
+      "epoch": 2.0400372439478587,
+      "grad_norm": 0.6071755632388289,
+      "learning_rate": 1.2741531193794535e-06,
+      "loss": 0.334,
+      "step": 4382
+    },
+    {
+      "epoch": 2.0405027932960893,
+      "grad_norm": 0.6310286796541141,
+      "learning_rate": 1.2730350002491414e-06,
+      "loss": 0.3108,
+      "step": 4383
+    },
+    {
+      "epoch": 2.0409683426443204,
+      "grad_norm": 0.6139126590840762,
+      "learning_rate": 1.2719172043305862e-06,
+      "loss": 0.3388,
+      "step": 4384
+    },
+    {
+      "epoch": 2.0414338919925514,
+      "grad_norm": 0.5945680694260383,
+      "learning_rate": 1.2707997319182427e-06,
+      "loss": 0.3121,
+      "step": 4385
+    },
+    {
+      "epoch": 2.041899441340782,
+      "grad_norm": 0.5855118069173522,
+      "learning_rate": 1.2696825833064808e-06,
+      "loss": 0.3254,
+      "step": 4386
+    },
+    {
+      "epoch": 2.042364990689013,
+      "grad_norm": 0.5874561091335501,
+      "learning_rate": 1.2685657587895822e-06,
+      "loss": 0.3223,
+      "step": 4387
+    },
+    {
+      "epoch": 2.0428305400372437,
+      "grad_norm": 0.6228682217544979,
+      "learning_rate": 1.2674492586617465e-06,
+      "loss": 0.3281,
+      "step": 4388
+    },
+    {
+      "epoch": 2.043296089385475,
+      "grad_norm": 0.5767631748173887,
+      "learning_rate": 1.2663330832170851e-06,
+      "loss": 0.318,
+      "step": 4389
+    },
+    {
+      "epoch": 2.043761638733706,
+      "grad_norm": 0.5615625612715236,
+      "learning_rate": 1.2652172327496257e-06,
+      "loss": 0.3139,
+      "step": 4390
+    },
+    {
+      "epoch": 2.0442271880819365,
+      "grad_norm": 0.630219666783665,
+      "learning_rate": 1.2641017075533108e-06,
+      "loss": 0.3308,
+      "step": 4391
+    },
+    {
+      "epoch": 2.0446927374301676,
+      "grad_norm": 0.6083382326198874,
+      "learning_rate": 1.2629865079219955e-06,
+      "loss": 0.3192,
+      "step": 4392
+    },
+    {
+      "epoch": 2.0451582867783986,
+      "grad_norm": 0.586286594781609,
+      "learning_rate": 1.2618716341494481e-06,
+      "loss": 0.3155,
+      "step": 4393
+    },
+    {
+      "epoch": 2.0456238361266292,
+      "grad_norm": 0.6317306206475326,
+      "learning_rate": 1.260757086529356e-06,
+      "loss": 0.3021,
+      "step": 4394
+    },
+    {
+      "epoch": 2.0460893854748603,
+      "grad_norm": 0.61783641809781,
+      "learning_rate": 1.2596428653553167e-06,
+      "loss": 0.3223,
+      "step": 4395
+    },
+    {
+      "epoch": 2.0465549348230914,
+      "grad_norm": 0.6091417569577959,
+      "learning_rate": 1.2585289709208404e-06,
+      "loss": 0.3474,
+      "step": 4396
+    },
+    {
+      "epoch": 2.047020484171322,
+      "grad_norm": 0.6300735094663203,
+      "learning_rate": 1.2574154035193554e-06,
+      "loss": 0.3413,
+      "step": 4397
+    },
+    {
+      "epoch": 2.047486033519553,
+      "grad_norm": 0.6011120186555062,
+      "learning_rate": 1.2563021634442021e-06,
+      "loss": 0.3193,
+      "step": 4398
+    },
+    {
+      "epoch": 2.047951582867784,
+      "grad_norm": 0.6518025692341497,
+      "learning_rate": 1.255189250988633e-06,
+      "loss": 0.3255,
+      "step": 4399
+    },
+    {
+      "epoch": 2.0484171322160147,
+      "grad_norm": 0.5888547152314314,
+      "learning_rate": 1.254076666445817e-06,
+      "loss": 0.304,
+      "step": 4400
+    },
+    {
+      "epoch": 2.048882681564246,
+      "grad_norm": 0.5956912652601017,
+      "learning_rate": 1.252964410108834e-06,
+      "loss": 0.3159,
+      "step": 4401
+    },
+    {
+      "epoch": 2.049348230912477,
+      "grad_norm": 0.6363029694372998,
+      "learning_rate": 1.2518524822706796e-06,
+      "loss": 0.3261,
+      "step": 4402
+    },
+    {
+      "epoch": 2.0498137802607075,
+      "grad_norm": 0.6021402637884089,
+      "learning_rate": 1.2507408832242627e-06,
+      "loss": 0.3213,
+      "step": 4403
+    },
+    {
+      "epoch": 2.0502793296089385,
+      "grad_norm": 0.6021387203067593,
+      "learning_rate": 1.2496296132624034e-06,
+      "loss": 0.3177,
+      "step": 4404
+    },
+    {
+      "epoch": 2.0507448789571696,
+      "grad_norm": 0.5870898242266026,
+      "learning_rate": 1.248518672677837e-06,
+      "loss": 0.3071,
+      "step": 4405
+    },
+    {
+      "epoch": 2.0512104283054002,
+      "grad_norm": 0.6409234534905199,
+      "learning_rate": 1.2474080617632134e-06,
+      "loss": 0.3176,
+      "step": 4406
+    },
+    {
+      "epoch": 2.0516759776536313,
+      "grad_norm": 0.613889283394037,
+      "learning_rate": 1.2462977808110913e-06,
+      "loss": 0.3211,
+      "step": 4407
+    },
+    {
+      "epoch": 2.0521415270018624,
+      "grad_norm": 0.6235100200208196,
+      "learning_rate": 1.2451878301139476e-06,
+      "loss": 0.3155,
+      "step": 4408
+    },
+    {
+      "epoch": 2.052607076350093,
+      "grad_norm": 0.6335218995283609,
+      "learning_rate": 1.2440782099641674e-06,
+      "loss": 0.3461,
+      "step": 4409
+    },
+    {
+      "epoch": 2.053072625698324,
+      "grad_norm": 0.6211905171466297,
+      "learning_rate": 1.242968920654052e-06,
+      "loss": 0.3286,
+      "step": 4410
+    },
+    {
+      "epoch": 2.053538175046555,
+      "grad_norm": 0.615399383131524,
+      "learning_rate": 1.241859962475816e-06,
+      "loss": 0.3328,
+      "step": 4411
+    },
+    {
+      "epoch": 2.0540037243947857,
+      "grad_norm": 0.6120642151617234,
+      "learning_rate": 1.240751335721584e-06,
+      "loss": 0.3281,
+      "step": 4412
+    },
+    {
+      "epoch": 2.054469273743017,
+      "grad_norm": 0.5872917308737061,
+      "learning_rate": 1.2396430406833926e-06,
+      "loss": 0.3183,
+      "step": 4413
+    },
+    {
+      "epoch": 2.054934823091248,
+      "grad_norm": 0.601679870652626,
+      "learning_rate": 1.238535077653197e-06,
+      "loss": 0.3188,
+      "step": 4414
+    },
+    {
+      "epoch": 2.0554003724394785,
+      "grad_norm": 0.5923878515942024,
+      "learning_rate": 1.2374274469228596e-06,
+      "loss": 0.3265,
+      "step": 4415
+    },
+    {
+      "epoch": 2.0558659217877095,
+      "grad_norm": 0.6250228921949985,
+      "learning_rate": 1.2363201487841547e-06,
+      "loss": 0.3256,
+      "step": 4416
+    },
+    {
+      "epoch": 2.0563314711359406,
+      "grad_norm": 0.6204617007852004,
+      "learning_rate": 1.2352131835287731e-06,
+      "loss": 0.3316,
+      "step": 4417
+    },
+    {
+      "epoch": 2.0567970204841712,
+      "grad_norm": 0.6274986748849709,
+      "learning_rate": 1.2341065514483158e-06,
+      "loss": 0.3298,
+      "step": 4418
+    },
+    {
+      "epoch": 2.0572625698324023,
+      "grad_norm": 0.600981760701074,
+      "learning_rate": 1.2330002528342946e-06,
+      "loss": 0.3503,
+      "step": 4419
+    },
+    {
+      "epoch": 2.0577281191806334,
+      "grad_norm": 0.6054830430479471,
+      "learning_rate": 1.2318942879781356e-06,
+      "loss": 0.3331,
+      "step": 4420
+    },
+    {
+      "epoch": 2.058193668528864,
+      "grad_norm": 0.6100538298688275,
+      "learning_rate": 1.2307886571711775e-06,
+      "loss": 0.3284,
+      "step": 4421
+    },
+    {
+      "epoch": 2.058659217877095,
+      "grad_norm": 0.6066724586818942,
+      "learning_rate": 1.2296833607046673e-06,
+      "loss": 0.3241,
+      "step": 4422
+    },
+    {
+      "epoch": 2.0591247672253257,
+      "grad_norm": 0.6003054520057474,
+      "learning_rate": 1.2285783988697685e-06,
+      "loss": 0.3312,
+      "step": 4423
+    },
+    {
+      "epoch": 2.0595903165735567,
+      "grad_norm": 0.5911688364646326,
+      "learning_rate": 1.2274737719575527e-06,
+      "loss": 0.328,
+      "step": 4424
+    },
+    {
+      "epoch": 2.060055865921788,
+      "grad_norm": 0.584185269333467,
+      "learning_rate": 1.2263694802590056e-06,
+      "loss": 0.3385,
+      "step": 4425
+    },
+    {
+      "epoch": 2.0605214152700184,
+      "grad_norm": 0.6204731418918048,
+      "learning_rate": 1.2252655240650248e-06,
+      "loss": 0.3169,
+      "step": 4426
+    },
+    {
+      "epoch": 2.0609869646182495,
+      "grad_norm": 0.6027376442059256,
+      "learning_rate": 1.2241619036664178e-06,
+      "loss": 0.326,
+      "step": 4427
+    },
+    {
+      "epoch": 2.0614525139664805,
+      "grad_norm": 0.6218120474092352,
+      "learning_rate": 1.223058619353903e-06,
+      "loss": 0.3401,
+      "step": 4428
+    },
+    {
+      "epoch": 2.061918063314711,
+      "grad_norm": 0.6119182356495577,
+      "learning_rate": 1.2219556714181145e-06,
+      "loss": 0.3243,
+      "step": 4429
+    },
+    {
+      "epoch": 2.0623836126629422,
+      "grad_norm": 0.6053996602938139,
+      "learning_rate": 1.2208530601495934e-06,
+      "loss": 0.3297,
+      "step": 4430
+    },
+    {
+      "epoch": 2.0628491620111733,
+      "grad_norm": 0.6082024485275109,
+      "learning_rate": 1.2197507858387937e-06,
+      "loss": 0.3272,
+      "step": 4431
+    },
+    {
+      "epoch": 2.063314711359404,
+      "grad_norm": 0.6162712219829696,
+      "learning_rate": 1.2186488487760804e-06,
+      "loss": 0.3224,
+      "step": 4432
+    },
+    {
+      "epoch": 2.063780260707635,
+      "grad_norm": 0.6020634589931881,
+      "learning_rate": 1.2175472492517316e-06,
+      "loss": 0.32,
+      "step": 4433
+    },
+    {
+      "epoch": 2.064245810055866,
+      "grad_norm": 0.5913476085447932,
+      "learning_rate": 1.2164459875559329e-06,
+      "loss": 0.3229,
+      "step": 4434
+    },
+    {
+      "epoch": 2.0647113594040967,
+      "grad_norm": 0.5987931809796035,
+      "learning_rate": 1.2153450639787842e-06,
+      "loss": 0.3179,
+      "step": 4435
+    },
+    {
+      "epoch": 2.0651769087523277,
+      "grad_norm": 0.6162103524402949,
+      "learning_rate": 1.2142444788102937e-06,
+      "loss": 0.3223,
+      "step": 4436
+    },
+    {
+      "epoch": 2.065642458100559,
+      "grad_norm": 0.6096128219719594,
+      "learning_rate": 1.213144232340382e-06,
+      "loss": 0.3064,
+      "step": 4437
+    },
+    {
+      "epoch": 2.0661080074487894,
+      "grad_norm": 0.5900092048695056,
+      "learning_rate": 1.2120443248588817e-06,
+      "loss": 0.302,
+      "step": 4438
+    },
+    {
+      "epoch": 2.0665735567970205,
+      "grad_norm": 0.6584214379134559,
+      "learning_rate": 1.2109447566555324e-06,
+      "loss": 0.3148,
+      "step": 4439
+    },
+    {
+      "epoch": 2.0670391061452515,
+      "grad_norm": 0.6185448418362607,
+      "learning_rate": 1.2098455280199875e-06,
+      "loss": 0.3391,
+      "step": 4440
+    },
+    {
+      "epoch": 2.067504655493482,
+      "grad_norm": 0.5991163582411976,
+      "learning_rate": 1.2087466392418112e-06,
+      "loss": 0.3052,
+      "step": 4441
+    },
+    {
+      "epoch": 2.0679702048417132,
+      "grad_norm": 0.8164029155602738,
+      "learning_rate": 1.2076480906104745e-06,
+      "loss": 0.3286,
+      "step": 4442
+    },
+    {
+      "epoch": 2.0684357541899443,
+      "grad_norm": 0.6031308175567504,
+      "learning_rate": 1.2065498824153635e-06,
+      "loss": 0.3316,
+      "step": 4443
+    },
+    {
+      "epoch": 2.068901303538175,
+      "grad_norm": 0.5859314289977592,
+      "learning_rate": 1.2054520149457705e-06,
+      "loss": 0.3133,
+      "step": 4444
+    },
+    {
+      "epoch": 2.069366852886406,
+      "grad_norm": 0.6144371473602246,
+      "learning_rate": 1.2043544884909006e-06,
+      "loss": 0.3116,
+      "step": 4445
+    },
+    {
+      "epoch": 2.069832402234637,
+      "grad_norm": 0.6105700647662647,
+      "learning_rate": 1.2032573033398692e-06,
+      "loss": 0.3266,
+      "step": 4446
+    },
+    {
+      "epoch": 2.0702979515828677,
+      "grad_norm": 0.5778232541811603,
+      "learning_rate": 1.2021604597817004e-06,
+      "loss": 0.3163,
+      "step": 4447
+    },
+    {
+      "epoch": 2.0707635009310987,
+      "grad_norm": 0.5964032490365752,
+      "learning_rate": 1.2010639581053268e-06,
+      "loss": 0.313,
+      "step": 4448
+    },
+    {
+      "epoch": 2.07122905027933,
+      "grad_norm": 0.6160723577132444,
+      "learning_rate": 1.1999677985995969e-06,
+      "loss": 0.3237,
+      "step": 4449
+    },
+    {
+      "epoch": 2.0716945996275604,
+      "grad_norm": 0.6149986866378818,
+      "learning_rate": 1.1988719815532633e-06,
+      "loss": 0.3371,
+      "step": 4450
+    },
+    {
+      "epoch": 2.0721601489757915,
+      "grad_norm": 0.6126377743870793,
+      "learning_rate": 1.1977765072549891e-06,
+      "loss": 0.3226,
+      "step": 4451
+    },
+    {
+      "epoch": 2.0726256983240225,
+      "grad_norm": 0.6078265606386374,
+      "learning_rate": 1.1966813759933494e-06,
+      "loss": 0.3321,
+      "step": 4452
+    },
+    {
+      "epoch": 2.073091247672253,
+      "grad_norm": 0.5995784380163277,
+      "learning_rate": 1.1955865880568288e-06,
+      "loss": 0.3342,
+      "step": 4453
+    },
+    {
+      "epoch": 2.0735567970204842,
+      "grad_norm": 0.5981940874740228,
+      "learning_rate": 1.1944921437338182e-06,
+      "loss": 0.329,
+      "step": 4454
+    },
+    {
+      "epoch": 2.0740223463687153,
+      "grad_norm": 0.6195556361832414,
+      "learning_rate": 1.1933980433126228e-06,
+      "loss": 0.312,
+      "step": 4455
+    },
+    {
+      "epoch": 2.074487895716946,
+      "grad_norm": 0.5912693020082356,
+      "learning_rate": 1.1923042870814522e-06,
+      "loss": 0.3253,
+      "step": 4456
+    },
+    {
+      "epoch": 2.074953445065177,
+      "grad_norm": 0.6364671221894009,
+      "learning_rate": 1.191210875328429e-06,
+      "loss": 0.3163,
+      "step": 4457
+    },
+    {
+      "epoch": 2.0754189944134076,
+      "grad_norm": 0.5812562220571181,
+      "learning_rate": 1.190117808341585e-06,
+      "loss": 0.3169,
+      "step": 4458
+    },
+    {
+      "epoch": 2.0758845437616387,
+      "grad_norm": 0.5894947971108658,
+      "learning_rate": 1.1890250864088576e-06,
+      "loss": 0.3334,
+      "step": 4459
+    },
+    {
+      "epoch": 2.0763500931098697,
+      "grad_norm": 0.5993639984744019,
+      "learning_rate": 1.1879327098180973e-06,
+      "loss": 0.3289,
+      "step": 4460
+    },
+    {
+      "epoch": 2.0768156424581004,
+      "grad_norm": 0.598668109624748,
+      "learning_rate": 1.1868406788570624e-06,
+      "loss": 0.3303,
+      "step": 4461
+    },
+    {
+      "epoch": 2.0772811918063314,
+      "grad_norm": 0.618953042549037,
+      "learning_rate": 1.1857489938134187e-06,
+      "loss": 0.3187,
+      "step": 4462
+    },
+    {
+      "epoch": 2.0777467411545625,
+      "grad_norm": 0.6154559457937165,
+      "learning_rate": 1.1846576549747422e-06,
+      "loss": 0.3331,
+      "step": 4463
+    },
+    {
+      "epoch": 2.078212290502793,
+      "grad_norm": 0.6058357398524367,
+      "learning_rate": 1.183566662628519e-06,
+      "loss": 0.3334,
+      "step": 4464
+    },
+    {
+      "epoch": 2.078677839851024,
+      "grad_norm": 0.6117309283367789,
+      "learning_rate": 1.1824760170621399e-06,
+      "loss": 0.3443,
+      "step": 4465
+    },
+    {
+      "epoch": 2.0791433891992552,
+      "grad_norm": 0.6271949313073754,
+      "learning_rate": 1.1813857185629094e-06,
+      "loss": 0.3229,
+      "step": 4466
+    },
+    {
+      "epoch": 2.079608938547486,
+      "grad_norm": 0.6154178497346753,
+      "learning_rate": 1.180295767418036e-06,
+      "loss": 0.3283,
+      "step": 4467
+    },
+    {
+      "epoch": 2.080074487895717,
+      "grad_norm": 0.5831346067143466,
+      "learning_rate": 1.1792061639146396e-06,
+      "loss": 0.3137,
+      "step": 4468
+    },
+    {
+      "epoch": 2.080540037243948,
+      "grad_norm": 0.5999363573110665,
+      "learning_rate": 1.1781169083397484e-06,
+      "loss": 0.3149,
+      "step": 4469
+    },
+    {
+      "epoch": 2.0810055865921786,
+      "grad_norm": 0.6309266623837652,
+      "learning_rate": 1.1770280009802977e-06,
+      "loss": 0.3445,
+      "step": 4470
+    },
+    {
+      "epoch": 2.0814711359404097,
+      "grad_norm": 0.5998365748426168,
+      "learning_rate": 1.1759394421231295e-06,
+      "loss": 0.3285,
+      "step": 4471
+    },
+    {
+      "epoch": 2.0819366852886407,
+      "grad_norm": 0.577475550905181,
+      "learning_rate": 1.1748512320549998e-06,
+      "loss": 0.3221,
+      "step": 4472
+    },
+    {
+      "epoch": 2.0824022346368714,
+      "grad_norm": 0.5910601956640646,
+      "learning_rate": 1.1737633710625672e-06,
+      "loss": 0.313,
+      "step": 4473
+    },
+    {
+      "epoch": 2.0828677839851024,
+      "grad_norm": 0.5841241232648096,
+      "learning_rate": 1.1726758594323995e-06,
+      "loss": 0.3404,
+      "step": 4474
+    },
+    {
+      "epoch": 2.0833333333333335,
+      "grad_norm": 0.6143470745999623,
+      "learning_rate": 1.1715886974509736e-06,
+      "loss": 0.3156,
+      "step": 4475
+    },
+    {
+      "epoch": 2.083798882681564,
+      "grad_norm": 0.6205229589089738,
+      "learning_rate": 1.1705018854046749e-06,
+      "loss": 0.3229,
+      "step": 4476
+    },
+    {
+      "epoch": 2.084264432029795,
+      "grad_norm": 0.6159241197561796,
+      "learning_rate": 1.1694154235797935e-06,
+      "loss": 0.3214,
+      "step": 4477
+    },
+    {
+      "epoch": 2.0847299813780262,
+      "grad_norm": 0.6353309159697667,
+      "learning_rate": 1.1683293122625316e-06,
+      "loss": 0.3404,
+      "step": 4478
+    },
+    {
+      "epoch": 2.085195530726257,
+      "grad_norm": 0.6026535616160205,
+      "learning_rate": 1.1672435517389947e-06,
+      "loss": 0.3305,
+      "step": 4479
+    },
+    {
+      "epoch": 2.085661080074488,
+      "grad_norm": 0.5889146188227236,
+      "learning_rate": 1.1661581422951988e-06,
+      "loss": 0.3131,
+      "step": 4480
+    },
+    {
+      "epoch": 2.086126629422719,
+      "grad_norm": 0.6140522082544531,
+      "learning_rate": 1.165073084217067e-06,
+      "loss": 0.3303,
+      "step": 4481
+    },
+    {
+      "epoch": 2.0865921787709496,
+      "grad_norm": 0.6016791515716939,
+      "learning_rate": 1.1639883777904288e-06,
+      "loss": 0.3396,
+      "step": 4482
+    },
+    {
+      "epoch": 2.0870577281191807,
+      "grad_norm": 0.6397839830784326,
+      "learning_rate": 1.1629040233010213e-06,
+      "loss": 0.3262,
+      "step": 4483
+    },
+    {
+      "epoch": 2.0875232774674117,
+      "grad_norm": 0.5958615464380626,
+      "learning_rate": 1.1618200210344907e-06,
+      "loss": 0.3296,
+      "step": 4484
+    },
+    {
+      "epoch": 2.0879888268156424,
+      "grad_norm": 0.631201980863247,
+      "learning_rate": 1.1607363712763876e-06,
+      "loss": 0.3337,
+      "step": 4485
+    },
+    {
+      "epoch": 2.0884543761638734,
+      "grad_norm": 0.5736775169386106,
+      "learning_rate": 1.159653074312172e-06,
+      "loss": 0.2975,
+      "step": 4486
+    },
+    {
+      "epoch": 2.0889199255121045,
+      "grad_norm": 0.613099658565082,
+      "learning_rate": 1.158570130427209e-06,
+      "loss": 0.3334,
+      "step": 4487
+    },
+    {
+      "epoch": 2.089385474860335,
+      "grad_norm": 0.6061266137028667,
+      "learning_rate": 1.1574875399067722e-06,
+      "loss": 0.327,
+      "step": 4488
+    },
+    {
+      "epoch": 2.089851024208566,
+      "grad_norm": 0.630163167308759,
+      "learning_rate": 1.156405303036043e-06,
+      "loss": 0.3259,
+      "step": 4489
+    },
+    {
+      "epoch": 2.0903165735567972,
+      "grad_norm": 0.60584769736662,
+      "learning_rate": 1.1553234201001072e-06,
+      "loss": 0.3068,
+      "step": 4490
+    },
+    {
+      "epoch": 2.090782122905028,
+      "grad_norm": 0.6221012465124769,
+      "learning_rate": 1.1542418913839568e-06,
+      "loss": 0.3275,
+      "step": 4491
+    },
+    {
+      "epoch": 2.091247672253259,
+      "grad_norm": 0.5991736453554259,
+      "learning_rate": 1.1531607171724957e-06,
+      "loss": 0.3295,
+      "step": 4492
+    },
+    {
+      "epoch": 2.0917132216014895,
+      "grad_norm": 0.612540351004494,
+      "learning_rate": 1.152079897750529e-06,
+      "loss": 0.3264,
+      "step": 4493
+    },
+    {
+      "epoch": 2.0921787709497206,
+      "grad_norm": 0.61076262204053,
+      "learning_rate": 1.1509994334027697e-06,
+      "loss": 0.335,
+      "step": 4494
+    },
+    {
+      "epoch": 2.0926443202979517,
+      "grad_norm": 0.6052380629866411,
+      "learning_rate": 1.1499193244138383e-06,
+      "loss": 0.3435,
+      "step": 4495
+    },
+    {
+      "epoch": 2.0931098696461823,
+      "grad_norm": 0.6291859707149071,
+      "learning_rate": 1.1488395710682624e-06,
+      "loss": 0.3155,
+      "step": 4496
+    },
+    {
+      "epoch": 2.0935754189944134,
+      "grad_norm": 0.64634579311,
+      "learning_rate": 1.1477601736504723e-06,
+      "loss": 0.3384,
+      "step": 4497
+    },
+    {
+      "epoch": 2.0940409683426444,
+      "grad_norm": 0.6282176266560082,
+      "learning_rate": 1.1466811324448095e-06,
+      "loss": 0.3236,
+      "step": 4498
+    },
+    {
+      "epoch": 2.094506517690875,
+      "grad_norm": 0.5929482342871275,
+      "learning_rate": 1.145602447735517e-06,
+      "loss": 0.3163,
+      "step": 4499
+    },
+    {
+      "epoch": 2.094972067039106,
+      "grad_norm": 0.6152927449487827,
+      "learning_rate": 1.144524119806747e-06,
+      "loss": 0.3276,
+      "step": 4500
+    },
+    {
+      "epoch": 2.095437616387337,
+      "grad_norm": 0.5792846664228144,
+      "learning_rate": 1.1434461489425576e-06,
+      "loss": 0.3397,
+      "step": 4501
+    },
+    {
+      "epoch": 2.095903165735568,
+      "grad_norm": 0.6091120172288307,
+      "learning_rate": 1.1423685354269107e-06,
+      "loss": 0.3141,
+      "step": 4502
+    },
+    {
+      "epoch": 2.096368715083799,
+      "grad_norm": 0.5946355548245865,
+      "learning_rate": 1.141291279543674e-06,
+      "loss": 0.3308,
+      "step": 4503
+    },
+    {
+      "epoch": 2.09683426443203,
+      "grad_norm": 0.5959108888346057,
+      "learning_rate": 1.1402143815766258e-06,
+      "loss": 0.3194,
+      "step": 4504
+    },
+    {
+      "epoch": 2.0972998137802605,
+      "grad_norm": 0.6213865043174734,
+      "learning_rate": 1.1391378418094448e-06,
+      "loss": 0.3306,
+      "step": 4505
+    },
+    {
+      "epoch": 2.0977653631284916,
+      "grad_norm": 0.5947683280248559,
+      "learning_rate": 1.138061660525716e-06,
+      "loss": 0.3225,
+      "step": 4506
+    },
+    {
+      "epoch": 2.0982309124767227,
+      "grad_norm": 0.6307327614815719,
+      "learning_rate": 1.1369858380089323e-06,
+      "loss": 0.3318,
+      "step": 4507
+    },
+    {
+      "epoch": 2.0986964618249533,
+      "grad_norm": 0.6185814815383875,
+      "learning_rate": 1.1359103745424921e-06,
+      "loss": 0.3299,
+      "step": 4508
+    },
+    {
+      "epoch": 2.0991620111731844,
+      "grad_norm": 0.5925252043305308,
+      "learning_rate": 1.1348352704096956e-06,
+      "loss": 0.3222,
+      "step": 4509
+    },
+    {
+      "epoch": 2.0996275605214154,
+      "grad_norm": 0.6052371031712677,
+      "learning_rate": 1.1337605258937526e-06,
+      "loss": 0.3075,
+      "step": 4510
+    },
+    {
+      "epoch": 2.100093109869646,
+      "grad_norm": 0.6105535378296132,
+      "learning_rate": 1.1326861412777762e-06,
+      "loss": 0.3292,
+      "step": 4511
+    },
+    {
+      "epoch": 2.100558659217877,
+      "grad_norm": 0.6178749636052067,
+      "learning_rate": 1.1316121168447839e-06,
+      "loss": 0.3166,
+      "step": 4512
+    },
+    {
+      "epoch": 2.101024208566108,
+      "grad_norm": 0.5991816427986674,
+      "learning_rate": 1.1305384528777006e-06,
+      "loss": 0.3226,
+      "step": 4513
+    },
+    {
+      "epoch": 2.101489757914339,
+      "grad_norm": 0.6159328435824784,
+      "learning_rate": 1.1294651496593534e-06,
+      "loss": 0.333,
+      "step": 4514
+    },
+    {
+      "epoch": 2.10195530726257,
+      "grad_norm": 0.5935514712181771,
+      "learning_rate": 1.1283922074724765e-06,
+      "loss": 0.3167,
+      "step": 4515
+    },
+    {
+      "epoch": 2.102420856610801,
+      "grad_norm": 0.6142350853330774,
+      "learning_rate": 1.1273196265997097e-06,
+      "loss": 0.3417,
+      "step": 4516
+    },
+    {
+      "epoch": 2.1028864059590315,
+      "grad_norm": 0.5997162696239188,
+      "learning_rate": 1.1262474073235941e-06,
+      "loss": 0.3255,
+      "step": 4517
+    },
+    {
+      "epoch": 2.1033519553072626,
+      "grad_norm": 0.5906566421752784,
+      "learning_rate": 1.1251755499265788e-06,
+      "loss": 0.3398,
+      "step": 4518
+    },
+    {
+      "epoch": 2.1038175046554937,
+      "grad_norm": 0.6092322699476675,
+      "learning_rate": 1.1241040546910176e-06,
+      "loss": 0.332,
+      "step": 4519
+    },
+    {
+      "epoch": 2.1042830540037243,
+      "grad_norm": 0.6134947341749856,
+      "learning_rate": 1.1230329218991657e-06,
+      "loss": 0.3485,
+      "step": 4520
+    },
+    {
+      "epoch": 2.1047486033519553,
+      "grad_norm": 0.6163485876706961,
+      "learning_rate": 1.121962151833187e-06,
+      "loss": 0.3268,
+      "step": 4521
+    },
+    {
+      "epoch": 2.1052141527001864,
+      "grad_norm": 0.596884738341314,
+      "learning_rate": 1.1208917447751457e-06,
+      "loss": 0.3343,
+      "step": 4522
+    },
+    {
+      "epoch": 2.105679702048417,
+      "grad_norm": 0.6261024282627176,
+      "learning_rate": 1.119821701007014e-06,
+      "loss": 0.3288,
+      "step": 4523
+    },
+    {
+      "epoch": 2.106145251396648,
+      "grad_norm": 0.6135688823689305,
+      "learning_rate": 1.118752020810667e-06,
+      "loss": 0.33,
+      "step": 4524
+    },
+    {
+      "epoch": 2.106610800744879,
+      "grad_norm": 0.5810223295586471,
+      "learning_rate": 1.1176827044678837e-06,
+      "loss": 0.3121,
+      "step": 4525
+    },
+    {
+      "epoch": 2.10707635009311,
+      "grad_norm": 0.5933562942444525,
+      "learning_rate": 1.1166137522603448e-06,
+      "loss": 0.3244,
+      "step": 4526
+    },
+    {
+      "epoch": 2.107541899441341,
+      "grad_norm": 0.607312683377229,
+      "learning_rate": 1.1155451644696424e-06,
+      "loss": 0.3322,
+      "step": 4527
+    },
+    {
+      "epoch": 2.1080074487895715,
+      "grad_norm": 0.5880972068751772,
+      "learning_rate": 1.1144769413772653e-06,
+      "loss": 0.3175,
+      "step": 4528
+    },
+    {
+      "epoch": 2.1084729981378025,
+      "grad_norm": 0.5904651110615992,
+      "learning_rate": 1.1134090832646082e-06,
+      "loss": 0.3275,
+      "step": 4529
+    },
+    {
+      "epoch": 2.1089385474860336,
+      "grad_norm": 0.5647430290793389,
+      "learning_rate": 1.1123415904129714e-06,
+      "loss": 0.3187,
+      "step": 4530
+    },
+    {
+      "epoch": 2.109404096834264,
+      "grad_norm": 0.610559374802217,
+      "learning_rate": 1.1112744631035587e-06,
+      "loss": 0.3414,
+      "step": 4531
+    },
+    {
+      "epoch": 2.1098696461824953,
+      "grad_norm": 0.5896475111954144,
+      "learning_rate": 1.110207701617475e-06,
+      "loss": 0.344,
+      "step": 4532
+    },
+    {
+      "epoch": 2.1103351955307263,
+      "grad_norm": 0.5825731506128011,
+      "learning_rate": 1.1091413062357323e-06,
+      "loss": 0.3219,
+      "step": 4533
+    },
+    {
+      "epoch": 2.110800744878957,
+      "grad_norm": 0.6351653697968019,
+      "learning_rate": 1.1080752772392428e-06,
+      "loss": 0.3324,
+      "step": 4534
+    },
+    {
+      "epoch": 2.111266294227188,
+      "grad_norm": 0.6096542725512359,
+      "learning_rate": 1.107009614908825e-06,
+      "loss": 0.3368,
+      "step": 4535
+    },
+    {
+      "epoch": 2.111731843575419,
+      "grad_norm": 0.5852105499389012,
+      "learning_rate": 1.1059443195252005e-06,
+      "loss": 0.3344,
+      "step": 4536
+    },
+    {
+      "epoch": 2.1121973929236497,
+      "grad_norm": 0.6013223063036616,
+      "learning_rate": 1.1048793913689912e-06,
+      "loss": 0.3334,
+      "step": 4537
+    },
+    {
+      "epoch": 2.112662942271881,
+      "grad_norm": 0.6058463877056318,
+      "learning_rate": 1.1038148307207259e-06,
+      "loss": 0.3338,
+      "step": 4538
+    },
+    {
+      "epoch": 2.113128491620112,
+      "grad_norm": 0.6452581282445871,
+      "learning_rate": 1.1027506378608358e-06,
+      "loss": 0.3405,
+      "step": 4539
+    },
+    {
+      "epoch": 2.1135940409683425,
+      "grad_norm": 0.5930564297015346,
+      "learning_rate": 1.101686813069653e-06,
+      "loss": 0.3302,
+      "step": 4540
+    },
+    {
+      "epoch": 2.1140595903165735,
+      "grad_norm": 0.5807613007618287,
+      "learning_rate": 1.100623356627416e-06,
+      "loss": 0.3299,
+      "step": 4541
+    },
+    {
+      "epoch": 2.1145251396648046,
+      "grad_norm": 0.6086716668070695,
+      "learning_rate": 1.0995602688142628e-06,
+      "loss": 0.3379,
+      "step": 4542
+    },
+    {
+      "epoch": 2.114990689013035,
+      "grad_norm": 0.6078896852676955,
+      "learning_rate": 1.098497549910237e-06,
+      "loss": 0.3264,
+      "step": 4543
+    },
+    {
+      "epoch": 2.1154562383612663,
+      "grad_norm": 0.6259553130282746,
+      "learning_rate": 1.0974352001952844e-06,
+      "loss": 0.3317,
+      "step": 4544
+    },
+    {
+      "epoch": 2.1159217877094973,
+      "grad_norm": 0.6156539912257583,
+      "learning_rate": 1.0963732199492528e-06,
+      "loss": 0.3125,
+      "step": 4545
+    },
+    {
+      "epoch": 2.116387337057728,
+      "grad_norm": 0.6296839506670336,
+      "learning_rate": 1.095311609451891e-06,
+      "loss": 0.3297,
+      "step": 4546
+    },
+    {
+      "epoch": 2.116852886405959,
+      "grad_norm": 0.6253396948306207,
+      "learning_rate": 1.094250368982856e-06,
+      "loss": 0.3512,
+      "step": 4547
+    },
+    {
+      "epoch": 2.11731843575419,
+      "grad_norm": 0.6078656742517505,
+      "learning_rate": 1.0931894988217024e-06,
+      "loss": 0.3261,
+      "step": 4548
+    },
+    {
+      "epoch": 2.1177839851024207,
+      "grad_norm": 0.6519527324292417,
+      "learning_rate": 1.092128999247887e-06,
+      "loss": 0.3353,
+      "step": 4549
+    },
+    {
+      "epoch": 2.118249534450652,
+      "grad_norm": 0.6081452042044474,
+      "learning_rate": 1.0910688705407723e-06,
+      "loss": 0.3345,
+      "step": 4550
+    },
+    {
+      "epoch": 2.118715083798883,
+      "grad_norm": 0.6259426221934316,
+      "learning_rate": 1.0900091129796217e-06,
+      "loss": 0.3284,
+      "step": 4551
+    },
+    {
+      "epoch": 2.1191806331471135,
+      "grad_norm": 0.6232038131864173,
+      "learning_rate": 1.0889497268435991e-06,
+      "loss": 0.331,
+      "step": 4552
+    },
+    {
+      "epoch": 2.1196461824953445,
+      "grad_norm": 0.6234075204129463,
+      "learning_rate": 1.087890712411773e-06,
+      "loss": 0.3303,
+      "step": 4553
+    },
+    {
+      "epoch": 2.1201117318435756,
+      "grad_norm": 0.6093752390976931,
+      "learning_rate": 1.0868320699631132e-06,
+      "loss": 0.337,
+      "step": 4554
+    },
+    {
+      "epoch": 2.120577281191806,
+      "grad_norm": 0.6368068109174958,
+      "learning_rate": 1.0857737997764903e-06,
+      "loss": 0.3184,
+      "step": 4555
+    },
+    {
+      "epoch": 2.1210428305400373,
+      "grad_norm": 0.6411326540491776,
+      "learning_rate": 1.0847159021306794e-06,
+      "loss": 0.3227,
+      "step": 4556
+    },
+    {
+      "epoch": 2.1215083798882683,
+      "grad_norm": 0.6170149737359064,
+      "learning_rate": 1.0836583773043538e-06,
+      "loss": 0.3228,
+      "step": 4557
+    },
+    {
+      "epoch": 2.121973929236499,
+      "grad_norm": 0.6099041987632856,
+      "learning_rate": 1.0826012255760918e-06,
+      "loss": 0.3294,
+      "step": 4558
+    },
+    {
+      "epoch": 2.12243947858473,
+      "grad_norm": 0.6460225156526789,
+      "learning_rate": 1.081544447224373e-06,
+      "loss": 0.3412,
+      "step": 4559
+    },
+    {
+      "epoch": 2.122905027932961,
+      "grad_norm": 0.6271557948696591,
+      "learning_rate": 1.0804880425275765e-06,
+      "loss": 0.3208,
+      "step": 4560
+    },
+    {
+      "epoch": 2.1233705772811917,
+      "grad_norm": 0.591514119311708,
+      "learning_rate": 1.0794320117639854e-06,
+      "loss": 0.3253,
+      "step": 4561
+    },
+    {
+      "epoch": 2.123836126629423,
+      "grad_norm": 0.6002795886788707,
+      "learning_rate": 1.0783763552117834e-06,
+      "loss": 0.3313,
+      "step": 4562
+    },
+    {
+      "epoch": 2.1243016759776534,
+      "grad_norm": 0.5961203921969606,
+      "learning_rate": 1.0773210731490547e-06,
+      "loss": 0.3192,
+      "step": 4563
+    },
+    {
+      "epoch": 2.1247672253258845,
+      "grad_norm": 0.5915020993088089,
+      "learning_rate": 1.0762661658537871e-06,
+      "loss": 0.3198,
+      "step": 4564
+    },
+    {
+      "epoch": 2.1252327746741155,
+      "grad_norm": 0.618399421175885,
+      "learning_rate": 1.0752116336038666e-06,
+      "loss": 0.3374,
+      "step": 4565
+    },
+    {
+      "epoch": 2.1256983240223466,
+      "grad_norm": 0.6064756586181629,
+      "learning_rate": 1.0741574766770823e-06,
+      "loss": 0.3066,
+      "step": 4566
+    },
+    {
+      "epoch": 2.126163873370577,
+      "grad_norm": 0.6017071433207956,
+      "learning_rate": 1.0731036953511262e-06,
+      "loss": 0.314,
+      "step": 4567
+    },
+    {
+      "epoch": 2.1266294227188083,
+      "grad_norm": 0.6480918351756455,
+      "learning_rate": 1.0720502899035875e-06,
+      "loss": 0.3242,
+      "step": 4568
+    },
+    {
+      "epoch": 2.127094972067039,
+      "grad_norm": 0.6285112539883325,
+      "learning_rate": 1.070997260611957e-06,
+      "loss": 0.3426,
+      "step": 4569
+    },
+    {
+      "epoch": 2.12756052141527,
+      "grad_norm": 0.607126537439939,
+      "learning_rate": 1.069944607753631e-06,
+      "loss": 0.3322,
+      "step": 4570
+    },
+    {
+      "epoch": 2.128026070763501,
+      "grad_norm": 0.5975525774046184,
+      "learning_rate": 1.0688923316059016e-06,
+      "loss": 0.3325,
+      "step": 4571
+    },
+    {
+      "epoch": 2.1284916201117317,
+      "grad_norm": 0.6052572993262768,
+      "learning_rate": 1.0678404324459627e-06,
+      "loss": 0.3335,
+      "step": 4572
+    },
+    {
+      "epoch": 2.1289571694599627,
+      "grad_norm": 0.5731967084766926,
+      "learning_rate": 1.06678891055091e-06,
+      "loss": 0.3216,
+      "step": 4573
+    },
+    {
+      "epoch": 2.129422718808194,
+      "grad_norm": 0.5775203856352581,
+      "learning_rate": 1.0657377661977403e-06,
+      "loss": 0.3295,
+      "step": 4574
+    },
+    {
+      "epoch": 2.1298882681564244,
+      "grad_norm": 0.605525945941863,
+      "learning_rate": 1.0646869996633489e-06,
+      "loss": 0.3341,
+      "step": 4575
+    },
+    {
+      "epoch": 2.1303538175046555,
+      "grad_norm": 0.5976090520401095,
+      "learning_rate": 1.0636366112245334e-06,
+      "loss": 0.3241,
+      "step": 4576
+    },
+    {
+      "epoch": 2.1308193668528865,
+      "grad_norm": 0.6226991622221757,
+      "learning_rate": 1.0625866011579903e-06,
+      "loss": 0.3461,
+      "step": 4577
+    },
+    {
+      "epoch": 2.131284916201117,
+      "grad_norm": 0.6270832026442019,
+      "learning_rate": 1.0615369697403175e-06,
+      "loss": 0.3358,
+      "step": 4578
+    },
+    {
+      "epoch": 2.131750465549348,
+      "grad_norm": 0.6167116815225445,
+      "learning_rate": 1.0604877172480139e-06,
+      "loss": 0.3241,
+      "step": 4579
+    },
+    {
+      "epoch": 2.1322160148975793,
+      "grad_norm": 0.590519666899913,
+      "learning_rate": 1.0594388439574769e-06,
+      "loss": 0.3235,
+      "step": 4580
+    },
+    {
+      "epoch": 2.13268156424581,
+      "grad_norm": 0.6067474648710808,
+      "learning_rate": 1.0583903501450028e-06,
+      "loss": 0.3289,
+      "step": 4581
+    },
+    {
+      "epoch": 2.133147113594041,
+      "grad_norm": 0.6594968176070554,
+      "learning_rate": 1.057342236086793e-06,
+      "loss": 0.3258,
+      "step": 4582
+    },
+    {
+      "epoch": 2.133612662942272,
+      "grad_norm": 0.5987797734232632,
+      "learning_rate": 1.0562945020589443e-06,
+      "loss": 0.3286,
+      "step": 4583
+    },
+    {
+      "epoch": 2.1340782122905027,
+      "grad_norm": 0.6115712319725372,
+      "learning_rate": 1.0552471483374535e-06,
+      "loss": 0.3273,
+      "step": 4584
+    },
+    {
+      "epoch": 2.1345437616387337,
+      "grad_norm": 0.5957874326403733,
+      "learning_rate": 1.05420017519822e-06,
+      "loss": 0.3255,
+      "step": 4585
+    },
+    {
+      "epoch": 2.135009310986965,
+      "grad_norm": 0.6271418144893873,
+      "learning_rate": 1.0531535829170416e-06,
+      "loss": 0.3355,
+      "step": 4586
+    },
+    {
+      "epoch": 2.1354748603351954,
+      "grad_norm": 0.6135178255296939,
+      "learning_rate": 1.0521073717696141e-06,
+      "loss": 0.3227,
+      "step": 4587
+    },
+    {
+      "epoch": 2.1359404096834265,
+      "grad_norm": 0.5944906569298739,
+      "learning_rate": 1.0510615420315368e-06,
+      "loss": 0.3217,
+      "step": 4588
+    },
+    {
+      "epoch": 2.1364059590316575,
+      "grad_norm": 0.5893800601567561,
+      "learning_rate": 1.050016093978303e-06,
+      "loss": 0.3214,
+      "step": 4589
+    },
+    {
+      "epoch": 2.136871508379888,
+      "grad_norm": 0.600408807876371,
+      "learning_rate": 1.0489710278853107e-06,
+      "loss": 0.3324,
+      "step": 4590
+    },
+    {
+      "epoch": 2.137337057728119,
+      "grad_norm": 0.574911310420192,
+      "learning_rate": 1.0479263440278553e-06,
+      "loss": 0.3042,
+      "step": 4591
+    },
+    {
+      "epoch": 2.1378026070763503,
+      "grad_norm": 0.5961160829490058,
+      "learning_rate": 1.0468820426811299e-06,
+      "loss": 0.3036,
+      "step": 4592
+    },
+    {
+      "epoch": 2.138268156424581,
+      "grad_norm": 0.5981400009387797,
+      "learning_rate": 1.0458381241202292e-06,
+      "loss": 0.3293,
+      "step": 4593
+    },
+    {
+      "epoch": 2.138733705772812,
+      "grad_norm": 0.589997697963117,
+      "learning_rate": 1.0447945886201471e-06,
+      "loss": 0.3232,
+      "step": 4594
+    },
+    {
+      "epoch": 2.139199255121043,
+      "grad_norm": 0.62179855162501,
+      "learning_rate": 1.0437514364557735e-06,
+      "loss": 0.3445,
+      "step": 4595
+    },
+    {
+      "epoch": 2.1396648044692737,
+      "grad_norm": 0.6154258883517898,
+      "learning_rate": 1.042708667901901e-06,
+      "loss": 0.3238,
+      "step": 4596
+    },
+    {
+      "epoch": 2.1401303538175047,
+      "grad_norm": 0.655056182318863,
+      "learning_rate": 1.04166628323322e-06,
+      "loss": 0.3371,
+      "step": 4597
+    },
+    {
+      "epoch": 2.1405959031657353,
+      "grad_norm": 0.6160480956840201,
+      "learning_rate": 1.040624282724318e-06,
+      "loss": 0.3573,
+      "step": 4598
+    },
+    {
+      "epoch": 2.1410614525139664,
+      "grad_norm": 0.6084011128427558,
+      "learning_rate": 1.0395826666496842e-06,
+      "loss": 0.3401,
+      "step": 4599
+    },
+    {
+      "epoch": 2.1415270018621975,
+      "grad_norm": 0.6165773548244435,
+      "learning_rate": 1.0385414352837035e-06,
+      "loss": 0.3354,
+      "step": 4600
+    },
+    {
+      "epoch": 2.1419925512104285,
+      "grad_norm": 0.5969672656824455,
+      "learning_rate": 1.0375005889006614e-06,
+      "loss": 0.3202,
+      "step": 4601
+    },
+    {
+      "epoch": 2.142458100558659,
+      "grad_norm": 0.5863608034146617,
+      "learning_rate": 1.0364601277747432e-06,
+      "loss": 0.332,
+      "step": 4602
+    },
+    {
+      "epoch": 2.14292364990689,
+      "grad_norm": 0.6106105833770633,
+      "learning_rate": 1.0354200521800298e-06,
+      "loss": 0.328,
+      "step": 4603
+    },
+    {
+      "epoch": 2.143389199255121,
+      "grad_norm": 0.6412710530176882,
+      "learning_rate": 1.0343803623905003e-06,
+      "loss": 0.3159,
+      "step": 4604
+    },
+    {
+      "epoch": 2.143854748603352,
+      "grad_norm": 0.6183312433382191,
+      "learning_rate": 1.0333410586800368e-06,
+      "loss": 0.3126,
+      "step": 4605
+    },
+    {
+      "epoch": 2.144320297951583,
+      "grad_norm": 0.5867579573630011,
+      "learning_rate": 1.0323021413224151e-06,
+      "loss": 0.3386,
+      "step": 4606
+    },
+    {
+      "epoch": 2.1447858472998136,
+      "grad_norm": 0.6222559797553252,
+      "learning_rate": 1.0312636105913102e-06,
+      "loss": 0.3386,
+      "step": 4607
+    },
+    {
+      "epoch": 2.1452513966480447,
+      "grad_norm": 0.6229123714446438,
+      "learning_rate": 1.0302254667602963e-06,
+      "loss": 0.3223,
+      "step": 4608
+    },
+    {
+      "epoch": 2.1457169459962757,
+      "grad_norm": 0.613458252631826,
+      "learning_rate": 1.0291877101028458e-06,
+      "loss": 0.3483,
+      "step": 4609
+    },
+    {
+      "epoch": 2.1461824953445063,
+      "grad_norm": 0.6490838895730205,
+      "learning_rate": 1.0281503408923274e-06,
+      "loss": 0.3149,
+      "step": 4610
+    },
+    {
+      "epoch": 2.1466480446927374,
+      "grad_norm": 0.6364253380969328,
+      "learning_rate": 1.02711335940201e-06,
+      "loss": 0.344,
+      "step": 4611
+    },
+    {
+      "epoch": 2.1471135940409685,
+      "grad_norm": 0.6034089504389452,
+      "learning_rate": 1.0260767659050578e-06,
+      "loss": 0.3169,
+      "step": 4612
+    },
+    {
+      "epoch": 2.147579143389199,
+      "grad_norm": 0.6293107423253389,
+      "learning_rate": 1.0250405606745348e-06,
+      "loss": 0.3229,
+      "step": 4613
+    },
+    {
+      "epoch": 2.14804469273743,
+      "grad_norm": 0.5802868284277468,
+      "learning_rate": 1.0240047439834031e-06,
+      "loss": 0.3302,
+      "step": 4614
+    },
+    {
+      "epoch": 2.148510242085661,
+      "grad_norm": 0.5782226043820508,
+      "learning_rate": 1.0229693161045195e-06,
+      "loss": 0.3215,
+      "step": 4615
+    },
+    {
+      "epoch": 2.148975791433892,
+      "grad_norm": 0.6162295015387489,
+      "learning_rate": 1.0219342773106418e-06,
+      "loss": 0.3293,
+      "step": 4616
+    },
+    {
+      "epoch": 2.149441340782123,
+      "grad_norm": 0.6390539288992764,
+      "learning_rate": 1.0208996278744237e-06,
+      "loss": 0.3483,
+      "step": 4617
+    },
+    {
+      "epoch": 2.149906890130354,
+      "grad_norm": 0.6169429563048248,
+      "learning_rate": 1.0198653680684156e-06,
+      "loss": 0.3309,
+      "step": 4618
+    },
+    {
+      "epoch": 2.1503724394785846,
+      "grad_norm": 0.6213149726992132,
+      "learning_rate": 1.0188314981650677e-06,
+      "loss": 0.3429,
+      "step": 4619
+    },
+    {
+      "epoch": 2.1508379888268156,
+      "grad_norm": 0.6055678400043688,
+      "learning_rate": 1.0177980184367237e-06,
+      "loss": 0.322,
+      "step": 4620
+    },
+    {
+      "epoch": 2.1513035381750467,
+      "grad_norm": 0.6014889139194141,
+      "learning_rate": 1.016764929155628e-06,
+      "loss": 0.3168,
+      "step": 4621
+    },
+    {
+      "epoch": 2.1517690875232773,
+      "grad_norm": 0.5984911769820614,
+      "learning_rate": 1.0157322305939218e-06,
+      "loss": 0.3255,
+      "step": 4622
+    },
+    {
+      "epoch": 2.1522346368715084,
+      "grad_norm": 0.6308325904246154,
+      "learning_rate": 1.0146999230236418e-06,
+      "loss": 0.3209,
+      "step": 4623
+    },
+    {
+      "epoch": 2.1527001862197395,
+      "grad_norm": 0.6074746835622545,
+      "learning_rate": 1.0136680067167198e-06,
+      "loss": 0.3286,
+      "step": 4624
+    },
+    {
+      "epoch": 2.15316573556797,
+      "grad_norm": 0.5987111237706473,
+      "learning_rate": 1.0126364819449914e-06,
+      "loss": 0.3165,
+      "step": 4625
+    },
+    {
+      "epoch": 2.153631284916201,
+      "grad_norm": 0.6475974269638144,
+      "learning_rate": 1.0116053489801824e-06,
+      "loss": 0.3445,
+      "step": 4626
+    },
+    {
+      "epoch": 2.154096834264432,
+      "grad_norm": 0.5910393030593971,
+      "learning_rate": 1.0105746080939174e-06,
+      "loss": 0.3269,
+      "step": 4627
+    },
+    {
+      "epoch": 2.154562383612663,
+      "grad_norm": 0.6087566853106711,
+      "learning_rate": 1.0095442595577186e-06,
+      "loss": 0.3295,
+      "step": 4628
+    },
+    {
+      "epoch": 2.155027932960894,
+      "grad_norm": 0.6312098953194628,
+      "learning_rate": 1.0085143036430054e-06,
+      "loss": 0.3458,
+      "step": 4629
+    },
+    {
+      "epoch": 2.155493482309125,
+      "grad_norm": 0.6281184926964535,
+      "learning_rate": 1.0074847406210907e-06,
+      "loss": 0.3174,
+      "step": 4630
+    },
+    {
+      "epoch": 2.1559590316573556,
+      "grad_norm": 0.6294824310533094,
+      "learning_rate": 1.006455570763188e-06,
+      "loss": 0.3315,
+      "step": 4631
+    },
+    {
+      "epoch": 2.1564245810055866,
+      "grad_norm": 0.6061597849086924,
+      "learning_rate": 1.0054267943404028e-06,
+      "loss": 0.3318,
+      "step": 4632
+    },
+    {
+      "epoch": 2.1568901303538173,
+      "grad_norm": 0.6101180988652934,
+      "learning_rate": 1.0043984116237404e-06,
+      "loss": 0.3428,
+      "step": 4633
+    },
+    {
+      "epoch": 2.1573556797020483,
+      "grad_norm": 0.6363954420556613,
+      "learning_rate": 1.0033704228841024e-06,
+      "loss": 0.3237,
+      "step": 4634
+    },
+    {
+      "epoch": 2.1578212290502794,
+      "grad_norm": 0.605668196620827,
+      "learning_rate": 1.0023428283922835e-06,
+      "loss": 0.3203,
+      "step": 4635
+    },
+    {
+      "epoch": 2.1582867783985105,
+      "grad_norm": 0.6067802278178178,
+      "learning_rate": 1.0013156284189774e-06,
+      "loss": 0.3362,
+      "step": 4636
+    },
+    {
+      "epoch": 2.158752327746741,
+      "grad_norm": 0.6099678411835406,
+      "learning_rate": 1.0002888232347738e-06,
+      "loss": 0.331,
+      "step": 4637
+    },
+    {
+      "epoch": 2.159217877094972,
+      "grad_norm": 0.58170884694297,
+      "learning_rate": 9.99262413110156e-07,
+      "loss": 0.3246,
+      "step": 4638
+    },
+    {
+      "epoch": 2.1596834264432028,
+      "grad_norm": 0.6321743012071763,
+      "learning_rate": 9.982363983155057e-07,
+      "loss": 0.3191,
+      "step": 4639
+    },
+    {
+      "epoch": 2.160148975791434,
+      "grad_norm": 0.5754357238006986,
+      "learning_rate": 9.972107791211005e-07,
+      "loss": 0.3368,
+      "step": 4640
+    },
+    {
+      "epoch": 2.160614525139665,
+      "grad_norm": 0.6470535272946952,
+      "learning_rate": 9.961855557971108e-07,
+      "loss": 0.3284,
+      "step": 4641
+    },
+    {
+      "epoch": 2.1610800744878955,
+      "grad_norm": 0.585993358517896,
+      "learning_rate": 9.951607286136073e-07,
+      "loss": 0.3243,
+      "step": 4642
+    },
+    {
+      "epoch": 2.1615456238361266,
+      "grad_norm": 0.5992064165285459,
+      "learning_rate": 9.94136297840551e-07,
+      "loss": 0.3389,
+      "step": 4643
+    },
+    {
+      "epoch": 2.1620111731843576,
+      "grad_norm": 0.5850767939939282,
+      "learning_rate": 9.931122637478032e-07,
+      "loss": 0.3135,
+      "step": 4644
+    },
+    {
+      "epoch": 2.1624767225325883,
+      "grad_norm": 0.6187692666257527,
+      "learning_rate": 9.920886266051193e-07,
+      "loss": 0.3405,
+      "step": 4645
+    },
+    {
+      "epoch": 2.1629422718808193,
+      "grad_norm": 0.5807775978146448,
+      "learning_rate": 9.910653866821488e-07,
+      "loss": 0.3111,
+      "step": 4646
+    },
+    {
+      "epoch": 2.1634078212290504,
+      "grad_norm": 0.9492218713714456,
+      "learning_rate": 9.900425442484356e-07,
+      "loss": 0.3344,
+      "step": 4647
+    },
+    {
+      "epoch": 2.163873370577281,
+      "grad_norm": 0.6202575390158007,
+      "learning_rate": 9.890200995734245e-07,
+      "loss": 0.3391,
+      "step": 4648
+    },
+    {
+      "epoch": 2.164338919925512,
+      "grad_norm": 0.6127621928942273,
+      "learning_rate": 9.879980529264495e-07,
+      "loss": 0.3284,
+      "step": 4649
+    },
+    {
+      "epoch": 2.164804469273743,
+      "grad_norm": 0.6151793181381525,
+      "learning_rate": 9.869764045767416e-07,
+      "loss": 0.3191,
+      "step": 4650
+    },
+    {
+      "epoch": 2.1652700186219738,
+      "grad_norm": 0.5776382097544208,
+      "learning_rate": 9.85955154793428e-07,
+      "loss": 0.3089,
+      "step": 4651
+    },
+    {
+      "epoch": 2.165735567970205,
+      "grad_norm": 0.5921959040507097,
+      "learning_rate": 9.84934303845531e-07,
+      "loss": 0.3268,
+      "step": 4652
+    },
+    {
+      "epoch": 2.166201117318436,
+      "grad_norm": 0.5936704468778364,
+      "learning_rate": 9.839138520019654e-07,
+      "loss": 0.3189,
+      "step": 4653
+    },
+    {
+      "epoch": 2.1666666666666665,
+      "grad_norm": 0.589656547757433,
+      "learning_rate": 9.828937995315439e-07,
+      "loss": 0.3099,
+      "step": 4654
+    },
+    {
+      "epoch": 2.1671322160148976,
+      "grad_norm": 0.6116171946708161,
+      "learning_rate": 9.81874146702971e-07,
+      "loss": 0.3443,
+      "step": 4655
+    },
+    {
+      "epoch": 2.1675977653631286,
+      "grad_norm": 0.5870785233831448,
+      "learning_rate": 9.808548937848486e-07,
+      "loss": 0.3243,
+      "step": 4656
+    },
+    {
+      "epoch": 2.1680633147113593,
+      "grad_norm": 0.6149887723779534,
+      "learning_rate": 9.798360410456728e-07,
+      "loss": 0.3254,
+      "step": 4657
+    },
+    {
+      "epoch": 2.1685288640595903,
+      "grad_norm": 0.6237125237303129,
+      "learning_rate": 9.788175887538325e-07,
+      "loss": 0.3357,
+      "step": 4658
+    },
+    {
+      "epoch": 2.1689944134078214,
+      "grad_norm": 0.585971087716994,
+      "learning_rate": 9.777995371776115e-07,
+      "loss": 0.3285,
+      "step": 4659
+    },
+    {
+      "epoch": 2.169459962756052,
+      "grad_norm": 0.5876658816084821,
+      "learning_rate": 9.76781886585191e-07,
+      "loss": 0.3224,
+      "step": 4660
+    },
+    {
+      "epoch": 2.169925512104283,
+      "grad_norm": 0.6464134315216151,
+      "learning_rate": 9.757646372446436e-07,
+      "loss": 0.3472,
+      "step": 4661
+    },
+    {
+      "epoch": 2.170391061452514,
+      "grad_norm": 0.6302169467733524,
+      "learning_rate": 9.747477894239355e-07,
+      "loss": 0.315,
+      "step": 4662
+    },
+    {
+      "epoch": 2.1708566108007448,
+      "grad_norm": 0.6192219206036329,
+      "learning_rate": 9.737313433909296e-07,
+      "loss": 0.3509,
+      "step": 4663
+    },
+    {
+      "epoch": 2.171322160148976,
+      "grad_norm": 0.6288030152582732,
+      "learning_rate": 9.727152994133831e-07,
+      "loss": 0.3215,
+      "step": 4664
+    },
+    {
+      "epoch": 2.171787709497207,
+      "grad_norm": 0.619424467027356,
+      "learning_rate": 9.716996577589441e-07,
+      "loss": 0.3451,
+      "step": 4665
+    },
+    {
+      "epoch": 2.1722532588454375,
+      "grad_norm": 0.5751810414790727,
+      "learning_rate": 9.70684418695159e-07,
+      "loss": 0.3163,
+      "step": 4666
+    },
+    {
+      "epoch": 2.1727188081936686,
+      "grad_norm": 0.5969947959864776,
+      "learning_rate": 9.696695824894633e-07,
+      "loss": 0.3346,
+      "step": 4667
+    },
+    {
+      "epoch": 2.1731843575418996,
+      "grad_norm": 0.5875117257476964,
+      "learning_rate": 9.686551494091906e-07,
+      "loss": 0.331,
+      "step": 4668
+    },
+    {
+      "epoch": 2.1736499068901303,
+      "grad_norm": 0.600327315336241,
+      "learning_rate": 9.67641119721567e-07,
+      "loss": 0.3269,
+      "step": 4669
+    },
+    {
+      "epoch": 2.1741154562383613,
+      "grad_norm": 0.6071641021680952,
+      "learning_rate": 9.66627493693711e-07,
+      "loss": 0.3163,
+      "step": 4670
+    },
+    {
+      "epoch": 2.1745810055865924,
+      "grad_norm": 0.6136103933055752,
+      "learning_rate": 9.656142715926362e-07,
+      "loss": 0.3416,
+      "step": 4671
+    },
+    {
+      "epoch": 2.175046554934823,
+      "grad_norm": 0.6127958687422665,
+      "learning_rate": 9.646014536852502e-07,
+      "loss": 0.338,
+      "step": 4672
+    },
+    {
+      "epoch": 2.175512104283054,
+      "grad_norm": 0.6173612858203961,
+      "learning_rate": 9.63589040238352e-07,
+      "loss": 0.3274,
+      "step": 4673
+    },
+    {
+      "epoch": 2.1759776536312847,
+      "grad_norm": 0.6294309397123153,
+      "learning_rate": 9.625770315186365e-07,
+      "loss": 0.3225,
+      "step": 4674
+    },
+    {
+      "epoch": 2.1764432029795158,
+      "grad_norm": 0.6259927876232313,
+      "learning_rate": 9.615654277926897e-07,
+      "loss": 0.333,
+      "step": 4675
+    },
+    {
+      "epoch": 2.176908752327747,
+      "grad_norm": 0.6481153940551533,
+      "learning_rate": 9.605542293269927e-07,
+      "loss": 0.3401,
+      "step": 4676
+    },
+    {
+      "epoch": 2.1773743016759775,
+      "grad_norm": 0.6149740417052113,
+      "learning_rate": 9.5954343638792e-07,
+      "loss": 0.3123,
+      "step": 4677
+    },
+    {
+      "epoch": 2.1778398510242085,
+      "grad_norm": 0.5964167852458291,
+      "learning_rate": 9.585330492417376e-07,
+      "loss": 0.3335,
+      "step": 4678
+    },
+    {
+      "epoch": 2.1783054003724396,
+      "grad_norm": 0.6237326001038341,
+      "learning_rate": 9.57523068154604e-07,
+      "loss": 0.3207,
+      "step": 4679
+    },
+    {
+      "epoch": 2.17877094972067,
+      "grad_norm": 0.6377765717501683,
+      "learning_rate": 9.56513493392576e-07,
+      "loss": 0.3206,
+      "step": 4680
+    },
+    {
+      "epoch": 2.1792364990689013,
+      "grad_norm": 0.6312767295026047,
+      "learning_rate": 9.55504325221597e-07,
+      "loss": 0.3247,
+      "step": 4681
+    },
+    {
+      "epoch": 2.1797020484171323,
+      "grad_norm": 0.637763617983705,
+      "learning_rate": 9.544955639075056e-07,
+      "loss": 0.3068,
+      "step": 4682
+    },
+    {
+      "epoch": 2.180167597765363,
+      "grad_norm": 0.6529589285917208,
+      "learning_rate": 9.534872097160345e-07,
+      "loss": 0.338,
+      "step": 4683
+    },
+    {
+      "epoch": 2.180633147113594,
+      "grad_norm": 0.6503794968665997,
+      "learning_rate": 9.524792629128085e-07,
+      "loss": 0.3406,
+      "step": 4684
+    },
+    {
+      "epoch": 2.181098696461825,
+      "grad_norm": 0.6129077047264545,
+      "learning_rate": 9.514717237633436e-07,
+      "loss": 0.33,
+      "step": 4685
+    },
+    {
+      "epoch": 2.1815642458100557,
+      "grad_norm": 0.625409318140244,
+      "learning_rate": 9.504645925330499e-07,
+      "loss": 0.319,
+      "step": 4686
+    },
+    {
+      "epoch": 2.1820297951582868,
+      "grad_norm": 0.5910995573881717,
+      "learning_rate": 9.494578694872312e-07,
+      "loss": 0.3227,
+      "step": 4687
+    },
+    {
+      "epoch": 2.182495344506518,
+      "grad_norm": 0.6438374257012183,
+      "learning_rate": 9.4845155489108e-07,
+      "loss": 0.324,
+      "step": 4688
+    },
+    {
+      "epoch": 2.1829608938547485,
+      "grad_norm": 0.6269945765785363,
+      "learning_rate": 9.474456490096856e-07,
+      "loss": 0.3248,
+      "step": 4689
+    },
+    {
+      "epoch": 2.1834264432029795,
+      "grad_norm": 0.6024490374546188,
+      "learning_rate": 9.464401521080255e-07,
+      "loss": 0.329,
+      "step": 4690
+    },
+    {
+      "epoch": 2.1838919925512106,
+      "grad_norm": 0.6051417607903833,
+      "learning_rate": 9.454350644509724e-07,
+      "loss": 0.3347,
+      "step": 4691
+    },
+    {
+      "epoch": 2.184357541899441,
+      "grad_norm": 0.6351765282215,
+      "learning_rate": 9.444303863032914e-07,
+      "loss": 0.3345,
+      "step": 4692
+    },
+    {
+      "epoch": 2.1848230912476723,
+      "grad_norm": 0.5763026998397851,
+      "learning_rate": 9.43426117929637e-07,
+      "loss": 0.3178,
+      "step": 4693
+    },
+    {
+      "epoch": 2.1852886405959033,
+      "grad_norm": 0.5921411318290103,
+      "learning_rate": 9.424222595945576e-07,
+      "loss": 0.3327,
+      "step": 4694
+    },
+    {
+      "epoch": 2.185754189944134,
+      "grad_norm": 0.6512302744738747,
+      "learning_rate": 9.414188115624947e-07,
+      "loss": 0.3266,
+      "step": 4695
+    },
+    {
+      "epoch": 2.186219739292365,
+      "grad_norm": 0.606851359147315,
+      "learning_rate": 9.404157740977785e-07,
+      "loss": 0.3339,
+      "step": 4696
+    },
+    {
+      "epoch": 2.186685288640596,
+      "grad_norm": 0.635486359053776,
+      "learning_rate": 9.394131474646348e-07,
+      "loss": 0.3245,
+      "step": 4697
+    },
+    {
+      "epoch": 2.1871508379888267,
+      "grad_norm": 0.6283130959873795,
+      "learning_rate": 9.384109319271775e-07,
+      "loss": 0.3279,
+      "step": 4698
+    },
+    {
+      "epoch": 2.1876163873370578,
+      "grad_norm": 0.6061287657937208,
+      "learning_rate": 9.374091277494146e-07,
+      "loss": 0.33,
+      "step": 4699
+    },
+    {
+      "epoch": 2.188081936685289,
+      "grad_norm": 0.5716602017203931,
+      "learning_rate": 9.364077351952463e-07,
+      "loss": 0.3323,
+      "step": 4700
+    },
+    {
+      "epoch": 2.1885474860335195,
+      "grad_norm": 0.6515908831006141,
+      "learning_rate": 9.354067545284626e-07,
+      "loss": 0.3442,
+      "step": 4701
+    },
+    {
+      "epoch": 2.1890130353817505,
+      "grad_norm": 0.599365990351363,
+      "learning_rate": 9.344061860127432e-07,
+      "loss": 0.3274,
+      "step": 4702
+    },
+    {
+      "epoch": 2.1894785847299816,
+      "grad_norm": 0.679943946983557,
+      "learning_rate": 9.334060299116657e-07,
+      "loss": 0.3236,
+      "step": 4703
+    },
+    {
+      "epoch": 2.189944134078212,
+      "grad_norm": 0.595420605485231,
+      "learning_rate": 9.324062864886929e-07,
+      "loss": 0.3439,
+      "step": 4704
+    },
+    {
+      "epoch": 2.1904096834264433,
+      "grad_norm": 0.599230386892895,
+      "learning_rate": 9.314069560071803e-07,
+      "loss": 0.3113,
+      "step": 4705
+    },
+    {
+      "epoch": 2.1908752327746743,
+      "grad_norm": 0.6365972027402346,
+      "learning_rate": 9.304080387303762e-07,
+      "loss": 0.3434,
+      "step": 4706
+    },
+    {
+      "epoch": 2.191340782122905,
+      "grad_norm": 0.570675813477703,
+      "learning_rate": 9.294095349214199e-07,
+      "loss": 0.3179,
+      "step": 4707
+    },
+    {
+      "epoch": 2.191806331471136,
+      "grad_norm": 0.5925358380269561,
+      "learning_rate": 9.284114448433392e-07,
+      "loss": 0.3189,
+      "step": 4708
+    },
+    {
+      "epoch": 2.1922718808193666,
+      "grad_norm": 0.6354166487294922,
+      "learning_rate": 9.27413768759057e-07,
+      "loss": 0.3079,
+      "step": 4709
+    },
+    {
+      "epoch": 2.1927374301675977,
+      "grad_norm": 0.6204112121616582,
+      "learning_rate": 9.264165069313827e-07,
+      "loss": 0.3306,
+      "step": 4710
+    },
+    {
+      "epoch": 2.1932029795158288,
+      "grad_norm": 0.6149736012406125,
+      "learning_rate": 9.2541965962302e-07,
+      "loss": 0.3324,
+      "step": 4711
+    },
+    {
+      "epoch": 2.1936685288640594,
+      "grad_norm": 0.5852598310726783,
+      "learning_rate": 9.244232270965625e-07,
+      "loss": 0.3234,
+      "step": 4712
+    },
+    {
+      "epoch": 2.1941340782122905,
+      "grad_norm": 0.6091255018898093,
+      "learning_rate": 9.23427209614493e-07,
+      "loss": 0.3111,
+      "step": 4713
+    },
+    {
+      "epoch": 2.1945996275605215,
+      "grad_norm": 0.6237035757258301,
+      "learning_rate": 9.22431607439187e-07,
+      "loss": 0.3438,
+      "step": 4714
+    },
+    {
+      "epoch": 2.195065176908752,
+      "grad_norm": 0.5958481149686939,
+      "learning_rate": 9.214364208329104e-07,
+      "loss": 0.3208,
+      "step": 4715
+    },
+    {
+      "epoch": 2.195530726256983,
+      "grad_norm": 0.6117477020406946,
+      "learning_rate": 9.204416500578175e-07,
+      "loss": 0.319,
+      "step": 4716
+    },
+    {
+      "epoch": 2.1959962756052143,
+      "grad_norm": 0.5916399869667567,
+      "learning_rate": 9.194472953759565e-07,
+      "loss": 0.3321,
+      "step": 4717
+    },
+    {
+      "epoch": 2.196461824953445,
+      "grad_norm": 0.6438691522707888,
+      "learning_rate": 9.184533570492618e-07,
+      "loss": 0.334,
+      "step": 4718
+    },
+    {
+      "epoch": 2.196927374301676,
+      "grad_norm": 0.6057095461493256,
+      "learning_rate": 9.174598353395614e-07,
+      "loss": 0.3289,
+      "step": 4719
+    },
+    {
+      "epoch": 2.197392923649907,
+      "grad_norm": 0.5986396917051939,
+      "learning_rate": 9.16466730508574e-07,
+      "loss": 0.3109,
+      "step": 4720
+    },
+    {
+      "epoch": 2.1978584729981376,
+      "grad_norm": 0.626966833972248,
+      "learning_rate": 9.154740428179054e-07,
+      "loss": 0.346,
+      "step": 4721
+    },
+    {
+      "epoch": 2.1983240223463687,
+      "grad_norm": 0.608252747221315,
+      "learning_rate": 9.144817725290516e-07,
+      "loss": 0.3229,
+      "step": 4722
+    },
+    {
+      "epoch": 2.1987895716945998,
+      "grad_norm": 0.6011702380526205,
+      "learning_rate": 9.134899199034039e-07,
+      "loss": 0.3409,
+      "step": 4723
+    },
+    {
+      "epoch": 2.1992551210428304,
+      "grad_norm": 0.5956059574134211,
+      "learning_rate": 9.124984852022381e-07,
+      "loss": 0.3362,
+      "step": 4724
+    },
+    {
+      "epoch": 2.1997206703910615,
+      "grad_norm": 0.5902338482461806,
+      "learning_rate": 9.115074686867206e-07,
+      "loss": 0.321,
+      "step": 4725
+    },
+    {
+      "epoch": 2.2001862197392925,
+      "grad_norm": 0.6203981243410295,
+      "learning_rate": 9.105168706179099e-07,
+      "loss": 0.3224,
+      "step": 4726
+    },
+    {
+      "epoch": 2.200651769087523,
+      "grad_norm": 0.6118359083931103,
+      "learning_rate": 9.095266912567536e-07,
+      "loss": 0.327,
+      "step": 4727
+    },
+    {
+      "epoch": 2.201117318435754,
+      "grad_norm": 0.5929955876955025,
+      "learning_rate": 9.085369308640871e-07,
+      "loss": 0.3028,
+      "step": 4728
+    },
+    {
+      "epoch": 2.2015828677839853,
+      "grad_norm": 0.5908642820798853,
+      "learning_rate": 9.075475897006372e-07,
+      "loss": 0.3156,
+      "step": 4729
+    },
+    {
+      "epoch": 2.202048417132216,
+      "grad_norm": 0.628882958196309,
+      "learning_rate": 9.065586680270213e-07,
+      "loss": 0.3212,
+      "step": 4730
+    },
+    {
+      "epoch": 2.202513966480447,
+      "grad_norm": 0.6163799450952394,
+      "learning_rate": 9.055701661037427e-07,
+      "loss": 0.3232,
+      "step": 4731
+    },
+    {
+      "epoch": 2.202979515828678,
+      "grad_norm": 0.597455505374949,
+      "learning_rate": 9.045820841911984e-07,
+      "loss": 0.3203,
+      "step": 4732
+    },
+    {
+      "epoch": 2.2034450651769086,
+      "grad_norm": 0.6281489379619838,
+      "learning_rate": 9.035944225496709e-07,
+      "loss": 0.3131,
+      "step": 4733
+    },
+    {
+      "epoch": 2.2039106145251397,
+      "grad_norm": 0.6219010249875596,
+      "learning_rate": 9.026071814393344e-07,
+      "loss": 0.3372,
+      "step": 4734
+    },
+    {
+      "epoch": 2.2043761638733708,
+      "grad_norm": 0.6724672798873151,
+      "learning_rate": 9.016203611202526e-07,
+      "loss": 0.3348,
+      "step": 4735
+    },
+    {
+      "epoch": 2.2048417132216014,
+      "grad_norm": 0.619565486313203,
+      "learning_rate": 9.00633961852376e-07,
+      "loss": 0.3407,
+      "step": 4736
+    },
+    {
+      "epoch": 2.2053072625698324,
+      "grad_norm": 0.5779972826797019,
+      "learning_rate": 8.99647983895546e-07,
+      "loss": 0.3184,
+      "step": 4737
+    },
+    {
+      "epoch": 2.2057728119180635,
+      "grad_norm": 0.6155222514247776,
+      "learning_rate": 8.986624275094941e-07,
+      "loss": 0.3186,
+      "step": 4738
+    },
+    {
+      "epoch": 2.206238361266294,
+      "grad_norm": 0.6177938849788249,
+      "learning_rate": 8.976772929538377e-07,
+      "loss": 0.3343,
+      "step": 4739
+    },
+    {
+      "epoch": 2.206703910614525,
+      "grad_norm": 0.5990854385228062,
+      "learning_rate": 8.966925804880846e-07,
+      "loss": 0.3374,
+      "step": 4740
+    },
+    {
+      "epoch": 2.2071694599627563,
+      "grad_norm": 0.595538215927126,
+      "learning_rate": 8.957082903716319e-07,
+      "loss": 0.3163,
+      "step": 4741
+    },
+    {
+      "epoch": 2.207635009310987,
+      "grad_norm": 0.6337877384281239,
+      "learning_rate": 8.947244228637655e-07,
+      "loss": 0.3357,
+      "step": 4742
+    },
+    {
+      "epoch": 2.208100558659218,
+      "grad_norm": 0.6225538747248834,
+      "learning_rate": 8.937409782236586e-07,
+      "loss": 0.3182,
+      "step": 4743
+    },
+    {
+      "epoch": 2.2085661080074486,
+      "grad_norm": 0.5848773988610175,
+      "learning_rate": 8.927579567103753e-07,
+      "loss": 0.3191,
+      "step": 4744
+    },
+    {
+      "epoch": 2.2090316573556796,
+      "grad_norm": 0.601515137356102,
+      "learning_rate": 8.917753585828648e-07,
+      "loss": 0.3189,
+      "step": 4745
+    },
+    {
+      "epoch": 2.2094972067039107,
+      "grad_norm": 0.5974500051575496,
+      "learning_rate": 8.907931840999682e-07,
+      "loss": 0.3317,
+      "step": 4746
+    },
+    {
+      "epoch": 2.2099627560521413,
+      "grad_norm": 0.6035580058751108,
+      "learning_rate": 8.898114335204139e-07,
+      "loss": 0.3401,
+      "step": 4747
+    },
+    {
+      "epoch": 2.2104283054003724,
+      "grad_norm": 0.5975971195270503,
+      "learning_rate": 8.888301071028171e-07,
+      "loss": 0.3231,
+      "step": 4748
+    },
+    {
+      "epoch": 2.2108938547486034,
+      "grad_norm": 0.6056367062902823,
+      "learning_rate": 8.878492051056833e-07,
+      "loss": 0.3267,
+      "step": 4749
+    },
+    {
+      "epoch": 2.211359404096834,
+      "grad_norm": 0.6442341560220645,
+      "learning_rate": 8.86868727787406e-07,
+      "loss": 0.3314,
+      "step": 4750
+    },
+    {
+      "epoch": 2.211824953445065,
+      "grad_norm": 0.6275038521181339,
+      "learning_rate": 8.858886754062648e-07,
+      "loss": 0.3221,
+      "step": 4751
+    },
+    {
+      "epoch": 2.212290502793296,
+      "grad_norm": 0.6012818247588242,
+      "learning_rate": 8.849090482204304e-07,
+      "loss": 0.3225,
+      "step": 4752
+    },
+    {
+      "epoch": 2.212756052141527,
+      "grad_norm": 0.6237757244436609,
+      "learning_rate": 8.839298464879581e-07,
+      "loss": 0.3281,
+      "step": 4753
+    },
+    {
+      "epoch": 2.213221601489758,
+      "grad_norm": 0.6217907792479088,
+      "learning_rate": 8.82951070466794e-07,
+      "loss": 0.3352,
+      "step": 4754
+    },
+    {
+      "epoch": 2.213687150837989,
+      "grad_norm": 0.598231170770901,
+      "learning_rate": 8.819727204147718e-07,
+      "loss": 0.3209,
+      "step": 4755
+    },
+    {
+      "epoch": 2.2141527001862196,
+      "grad_norm": 0.6157641319075552,
+      "learning_rate": 8.809947965896109e-07,
+      "loss": 0.326,
+      "step": 4756
+    },
+    {
+      "epoch": 2.2146182495344506,
+      "grad_norm": 0.6166045824004052,
+      "learning_rate": 8.800172992489186e-07,
+      "loss": 0.3184,
+      "step": 4757
+    },
+    {
+      "epoch": 2.2150837988826817,
+      "grad_norm": 0.6154978202232614,
+      "learning_rate": 8.790402286501939e-07,
+      "loss": 0.3319,
+      "step": 4758
+    },
+    {
+      "epoch": 2.2155493482309123,
+      "grad_norm": 0.6120975263263624,
+      "learning_rate": 8.780635850508191e-07,
+      "loss": 0.3279,
+      "step": 4759
+    },
+    {
+      "epoch": 2.2160148975791434,
+      "grad_norm": 0.6018612926126697,
+      "learning_rate": 8.770873687080641e-07,
+      "loss": 0.3146,
+      "step": 4760
+    },
+    {
+      "epoch": 2.2164804469273744,
+      "grad_norm": 0.6192129753821443,
+      "learning_rate": 8.761115798790886e-07,
+      "loss": 0.3183,
+      "step": 4761
+    },
+    {
+      "epoch": 2.216945996275605,
+      "grad_norm": 0.6055218287431242,
+      "learning_rate": 8.751362188209394e-07,
+      "loss": 0.3219,
+      "step": 4762
+    },
+    {
+      "epoch": 2.217411545623836,
+      "grad_norm": 0.6380814191469081,
+      "learning_rate": 8.741612857905479e-07,
+      "loss": 0.3488,
+      "step": 4763
+    },
+    {
+      "epoch": 2.217877094972067,
+      "grad_norm": 0.5871761410208373,
+      "learning_rate": 8.731867810447366e-07,
+      "loss": 0.3335,
+      "step": 4764
+    },
+    {
+      "epoch": 2.218342644320298,
+      "grad_norm": 0.6086015487906394,
+      "learning_rate": 8.722127048402118e-07,
+      "loss": 0.3199,
+      "step": 4765
+    },
+    {
+      "epoch": 2.218808193668529,
+      "grad_norm": 0.5813334596494988,
+      "learning_rate": 8.712390574335682e-07,
+      "loss": 0.3327,
+      "step": 4766
+    },
+    {
+      "epoch": 2.21927374301676,
+      "grad_norm": 0.6051809029671825,
+      "learning_rate": 8.702658390812896e-07,
+      "loss": 0.3199,
+      "step": 4767
+    },
+    {
+      "epoch": 2.2197392923649906,
+      "grad_norm": 0.6279258629543938,
+      "learning_rate": 8.692930500397425e-07,
+      "loss": 0.3251,
+      "step": 4768
+    },
+    {
+      "epoch": 2.2202048417132216,
+      "grad_norm": 0.6032320696097984,
+      "learning_rate": 8.683206905651839e-07,
+      "loss": 0.3149,
+      "step": 4769
+    },
+    {
+      "epoch": 2.2206703910614527,
+      "grad_norm": 0.6122772213619438,
+      "learning_rate": 8.673487609137568e-07,
+      "loss": 0.3319,
+      "step": 4770
+    },
+    {
+      "epoch": 2.2211359404096833,
+      "grad_norm": 0.593516506387028,
+      "learning_rate": 8.663772613414895e-07,
+      "loss": 0.3329,
+      "step": 4771
+    },
+    {
+      "epoch": 2.2216014897579144,
+      "grad_norm": 0.5975276302296083,
+      "learning_rate": 8.654061921042984e-07,
+      "loss": 0.3396,
+      "step": 4772
+    },
+    {
+      "epoch": 2.2220670391061454,
+      "grad_norm": 0.6077610448186297,
+      "learning_rate": 8.644355534579874e-07,
+      "loss": 0.3384,
+      "step": 4773
+    },
+    {
+      "epoch": 2.222532588454376,
+      "grad_norm": 0.5954971681797537,
+      "learning_rate": 8.634653456582436e-07,
+      "loss": 0.3241,
+      "step": 4774
+    },
+    {
+      "epoch": 2.222998137802607,
+      "grad_norm": 0.5959592174025291,
+      "learning_rate": 8.624955689606454e-07,
+      "loss": 0.3272,
+      "step": 4775
+    },
+    {
+      "epoch": 2.223463687150838,
+      "grad_norm": 0.6081062863777247,
+      "learning_rate": 8.615262236206526e-07,
+      "loss": 0.3234,
+      "step": 4776
+    },
+    {
+      "epoch": 2.223929236499069,
+      "grad_norm": 0.5771487022303117,
+      "learning_rate": 8.605573098936149e-07,
+      "loss": 0.3231,
+      "step": 4777
+    },
+    {
+      "epoch": 2.2243947858473,
+      "grad_norm": 0.612914591012698,
+      "learning_rate": 8.595888280347681e-07,
+      "loss": 0.3068,
+      "step": 4778
+    },
+    {
+      "epoch": 2.2248603351955305,
+      "grad_norm": 0.6369296765261999,
+      "learning_rate": 8.586207782992323e-07,
+      "loss": 0.3461,
+      "step": 4779
+    },
+    {
+      "epoch": 2.2253258845437616,
+      "grad_norm": 0.6169577504734828,
+      "learning_rate": 8.576531609420133e-07,
+      "loss": 0.3294,
+      "step": 4780
+    },
+    {
+      "epoch": 2.2257914338919926,
+      "grad_norm": 0.6234564319722934,
+      "learning_rate": 8.56685976218008e-07,
+      "loss": 0.3269,
+      "step": 4781
+    },
+    {
+      "epoch": 2.2262569832402237,
+      "grad_norm": 0.5753384266114749,
+      "learning_rate": 8.557192243819943e-07,
+      "loss": 0.3179,
+      "step": 4782
+    },
+    {
+      "epoch": 2.2267225325884543,
+      "grad_norm": 0.6333782197963094,
+      "learning_rate": 8.547529056886361e-07,
+      "loss": 0.326,
+      "step": 4783
+    },
+    {
+      "epoch": 2.2271880819366854,
+      "grad_norm": 0.600357669298544,
+      "learning_rate": 8.537870203924861e-07,
+      "loss": 0.3001,
+      "step": 4784
+    },
+    {
+      "epoch": 2.227653631284916,
+      "grad_norm": 0.6303781539269738,
+      "learning_rate": 8.528215687479824e-07,
+      "loss": 0.3198,
+      "step": 4785
+    },
+    {
+      "epoch": 2.228119180633147,
+      "grad_norm": 0.6495044237766744,
+      "learning_rate": 8.518565510094459e-07,
+      "loss": 0.3397,
+      "step": 4786
+    },
+    {
+      "epoch": 2.228584729981378,
+      "grad_norm": 0.589116978162581,
+      "learning_rate": 8.508919674310873e-07,
+      "loss": 0.3155,
+      "step": 4787
+    },
+    {
+      "epoch": 2.2290502793296088,
+      "grad_norm": 0.6277290679539215,
+      "learning_rate": 8.499278182669985e-07,
+      "loss": 0.3386,
+      "step": 4788
+    },
+    {
+      "epoch": 2.22951582867784,
+      "grad_norm": 0.5879744045504882,
+      "learning_rate": 8.489641037711608e-07,
+      "loss": 0.3333,
+      "step": 4789
+    },
+    {
+      "epoch": 2.229981378026071,
+      "grad_norm": 0.6021779190758916,
+      "learning_rate": 8.4800082419744e-07,
+      "loss": 0.3307,
+      "step": 4790
+    },
+    {
+      "epoch": 2.2304469273743015,
+      "grad_norm": 0.5819108232071494,
+      "learning_rate": 8.470379797995854e-07,
+      "loss": 0.321,
+      "step": 4791
+    },
+    {
+      "epoch": 2.2309124767225326,
+      "grad_norm": 0.5977468429089067,
+      "learning_rate": 8.460755708312338e-07,
+      "loss": 0.3022,
+      "step": 4792
+    },
+    {
+      "epoch": 2.2313780260707636,
+      "grad_norm": 0.6044489270740441,
+      "learning_rate": 8.451135975459077e-07,
+      "loss": 0.3207,
+      "step": 4793
+    },
+    {
+      "epoch": 2.2318435754189943,
+      "grad_norm": 0.5998926815393638,
+      "learning_rate": 8.441520601970116e-07,
+      "loss": 0.3335,
+      "step": 4794
+    },
+    {
+      "epoch": 2.2323091247672253,
+      "grad_norm": 0.6082111174779854,
+      "learning_rate": 8.431909590378392e-07,
+      "loss": 0.3367,
+      "step": 4795
+    },
+    {
+      "epoch": 2.2327746741154564,
+      "grad_norm": 0.6081364732532017,
+      "learning_rate": 8.42230294321566e-07,
+      "loss": 0.3293,
+      "step": 4796
+    },
+    {
+      "epoch": 2.233240223463687,
+      "grad_norm": 0.6269603504569491,
+      "learning_rate": 8.412700663012543e-07,
+      "loss": 0.3435,
+      "step": 4797
+    },
+    {
+      "epoch": 2.233705772811918,
+      "grad_norm": 0.595994770732162,
+      "learning_rate": 8.40310275229852e-07,
+      "loss": 0.3246,
+      "step": 4798
+    },
+    {
+      "epoch": 2.234171322160149,
+      "grad_norm": 0.5898676261828308,
+      "learning_rate": 8.393509213601903e-07,
+      "loss": 0.321,
+      "step": 4799
+    },
+    {
+      "epoch": 2.2346368715083798,
+      "grad_norm": 0.6109399010358738,
+      "learning_rate": 8.383920049449839e-07,
+      "loss": 0.3207,
+      "step": 4800
+    },
+    {
+      "epoch": 2.235102420856611,
+      "grad_norm": 0.6237674323361383,
+      "learning_rate": 8.374335262368375e-07,
+      "loss": 0.3211,
+      "step": 4801
+    },
+    {
+      "epoch": 2.235567970204842,
+      "grad_norm": 0.5876193228370749,
+      "learning_rate": 8.364754854882357e-07,
+      "loss": 0.3293,
+      "step": 4802
+    },
+    {
+      "epoch": 2.2360335195530725,
+      "grad_norm": 0.6374253206138037,
+      "learning_rate": 8.355178829515484e-07,
+      "loss": 0.3421,
+      "step": 4803
+    },
+    {
+      "epoch": 2.2364990689013036,
+      "grad_norm": 0.6116484107833943,
+      "learning_rate": 8.345607188790314e-07,
+      "loss": 0.3202,
+      "step": 4804
+    },
+    {
+      "epoch": 2.2369646182495346,
+      "grad_norm": 0.6214763319862308,
+      "learning_rate": 8.336039935228254e-07,
+      "loss": 0.3253,
+      "step": 4805
+    },
+    {
+      "epoch": 2.2374301675977653,
+      "grad_norm": 0.587881861267163,
+      "learning_rate": 8.326477071349534e-07,
+      "loss": 0.3239,
+      "step": 4806
+    },
+    {
+      "epoch": 2.2378957169459963,
+      "grad_norm": 0.6188477273649601,
+      "learning_rate": 8.316918599673249e-07,
+      "loss": 0.3142,
+      "step": 4807
+    },
+    {
+      "epoch": 2.2383612662942274,
+      "grad_norm": 0.6049768566198108,
+      "learning_rate": 8.307364522717318e-07,
+      "loss": 0.3221,
+      "step": 4808
+    },
+    {
+      "epoch": 2.238826815642458,
+      "grad_norm": 0.6223005852541283,
+      "learning_rate": 8.297814842998517e-07,
+      "loss": 0.3251,
+      "step": 4809
+    },
+    {
+      "epoch": 2.239292364990689,
+      "grad_norm": 0.6182332808984516,
+      "learning_rate": 8.288269563032466e-07,
+      "loss": 0.3348,
+      "step": 4810
+    },
+    {
+      "epoch": 2.23975791433892,
+      "grad_norm": 0.6078393519663073,
+      "learning_rate": 8.278728685333603e-07,
+      "loss": 0.3384,
+      "step": 4811
+    },
+    {
+      "epoch": 2.2402234636871508,
+      "grad_norm": 0.5982414934029975,
+      "learning_rate": 8.269192212415233e-07,
+      "loss": 0.3291,
+      "step": 4812
+    },
+    {
+      "epoch": 2.240689013035382,
+      "grad_norm": 0.5905882238074934,
+      "learning_rate": 8.259660146789497e-07,
+      "loss": 0.3288,
+      "step": 4813
+    },
+    {
+      "epoch": 2.2411545623836124,
+      "grad_norm": 0.6165606399081087,
+      "learning_rate": 8.250132490967347e-07,
+      "loss": 0.3371,
+      "step": 4814
+    },
+    {
+      "epoch": 2.2416201117318435,
+      "grad_norm": 0.6254260668830706,
+      "learning_rate": 8.240609247458606e-07,
+      "loss": 0.3402,
+      "step": 4815
+    },
+    {
+      "epoch": 2.2420856610800746,
+      "grad_norm": 0.6338319377519824,
+      "learning_rate": 8.231090418771931e-07,
+      "loss": 0.335,
+      "step": 4816
+    },
+    {
+      "epoch": 2.2425512104283056,
+      "grad_norm": 0.6328495973524867,
+      "learning_rate": 8.2215760074148e-07,
+      "loss": 0.3485,
+      "step": 4817
+    },
+    {
+      "epoch": 2.2430167597765363,
+      "grad_norm": 0.6022920047389779,
+      "learning_rate": 8.212066015893527e-07,
+      "loss": 0.3426,
+      "step": 4818
+    },
+    {
+      "epoch": 2.2434823091247673,
+      "grad_norm": 0.5928502300412649,
+      "learning_rate": 8.202560446713276e-07,
+      "loss": 0.3241,
+      "step": 4819
+    },
+    {
+      "epoch": 2.243947858472998,
+      "grad_norm": 0.6066024361030856,
+      "learning_rate": 8.193059302378048e-07,
+      "loss": 0.3271,
+      "step": 4820
+    },
+    {
+      "epoch": 2.244413407821229,
+      "grad_norm": 0.5847771153103343,
+      "learning_rate": 8.183562585390655e-07,
+      "loss": 0.3372,
+      "step": 4821
+    },
+    {
+      "epoch": 2.24487895716946,
+      "grad_norm": 0.6075398323060317,
+      "learning_rate": 8.174070298252778e-07,
+      "loss": 0.3311,
+      "step": 4822
+    },
+    {
+      "epoch": 2.2453445065176907,
+      "grad_norm": 0.6037715331175912,
+      "learning_rate": 8.164582443464891e-07,
+      "loss": 0.3316,
+      "step": 4823
+    },
+    {
+      "epoch": 2.2458100558659218,
+      "grad_norm": 0.586892244254252,
+      "learning_rate": 8.155099023526328e-07,
+      "loss": 0.3217,
+      "step": 4824
+    },
+    {
+      "epoch": 2.246275605214153,
+      "grad_norm": 0.619757401808421,
+      "learning_rate": 8.145620040935257e-07,
+      "loss": 0.333,
+      "step": 4825
+    },
+    {
+      "epoch": 2.2467411545623834,
+      "grad_norm": 0.612836199546338,
+      "learning_rate": 8.136145498188653e-07,
+      "loss": 0.3252,
+      "step": 4826
+    },
+    {
+      "epoch": 2.2472067039106145,
+      "grad_norm": 0.6355540616349896,
+      "learning_rate": 8.126675397782338e-07,
+      "loss": 0.3215,
+      "step": 4827
+    },
+    {
+      "epoch": 2.2476722532588456,
+      "grad_norm": 0.6209907158507386,
+      "learning_rate": 8.117209742210977e-07,
+      "loss": 0.3358,
+      "step": 4828
+    },
+    {
+      "epoch": 2.248137802607076,
+      "grad_norm": 0.5882037843532294,
+      "learning_rate": 8.107748533968027e-07,
+      "loss": 0.3172,
+      "step": 4829
+    },
+    {
+      "epoch": 2.2486033519553073,
+      "grad_norm": 0.6141529267517146,
+      "learning_rate": 8.098291775545816e-07,
+      "loss": 0.331,
+      "step": 4830
+    },
+    {
+      "epoch": 2.2490689013035383,
+      "grad_norm": 0.5932279417743179,
+      "learning_rate": 8.088839469435458e-07,
+      "loss": 0.3493,
+      "step": 4831
+    },
+    {
+      "epoch": 2.249534450651769,
+      "grad_norm": 0.655788902811834,
+      "learning_rate": 8.079391618126927e-07,
+      "loss": 0.3333,
+      "step": 4832
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6150157845292318,
+      "learning_rate": 8.06994822410902e-07,
+      "loss": 0.3177,
+      "step": 4833
+    },
+    {
+      "epoch": 2.250465549348231,
+      "grad_norm": 0.5984703281076917,
+      "learning_rate": 8.060509289869342e-07,
+      "loss": 0.3371,
+      "step": 4834
+    },
+    {
+      "epoch": 2.2509310986964617,
+      "grad_norm": 0.6205442963849584,
+      "learning_rate": 8.051074817894314e-07,
+      "loss": 0.3171,
+      "step": 4835
+    },
+    {
+      "epoch": 2.2513966480446927,
+      "grad_norm": 0.6104834486065385,
+      "learning_rate": 8.041644810669238e-07,
+      "loss": 0.3243,
+      "step": 4836
+    },
+    {
+      "epoch": 2.251862197392924,
+      "grad_norm": 0.6317624822724809,
+      "learning_rate": 8.032219270678182e-07,
+      "loss": 0.3212,
+      "step": 4837
+    },
+    {
+      "epoch": 2.2523277467411544,
+      "grad_norm": 0.6024251676382852,
+      "learning_rate": 8.022798200404053e-07,
+      "loss": 0.3281,
+      "step": 4838
+    },
+    {
+      "epoch": 2.2527932960893855,
+      "grad_norm": 0.6055486152729338,
+      "learning_rate": 8.013381602328588e-07,
+      "loss": 0.3096,
+      "step": 4839
+    },
+    {
+      "epoch": 2.2532588454376166,
+      "grad_norm": 0.5960739592662807,
+      "learning_rate": 8.003969478932352e-07,
+      "loss": 0.3339,
+      "step": 4840
+    },
+    {
+      "epoch": 2.253724394785847,
+      "grad_norm": 0.5955769412058901,
+      "learning_rate": 7.994561832694712e-07,
+      "loss": 0.3314,
+      "step": 4841
+    },
+    {
+      "epoch": 2.2541899441340782,
+      "grad_norm": 0.6501571874317624,
+      "learning_rate": 7.985158666093873e-07,
+      "loss": 0.341,
+      "step": 4842
+    },
+    {
+      "epoch": 2.2546554934823093,
+      "grad_norm": 0.6139585204819684,
+      "learning_rate": 7.975759981606842e-07,
+      "loss": 0.3278,
+      "step": 4843
+    },
+    {
+      "epoch": 2.25512104283054,
+      "grad_norm": 0.6064318143762688,
+      "learning_rate": 7.966365781709464e-07,
+      "loss": 0.3282,
+      "step": 4844
+    },
+    {
+      "epoch": 2.255586592178771,
+      "grad_norm": 0.648823368315172,
+      "learning_rate": 7.956976068876401e-07,
+      "loss": 0.3376,
+      "step": 4845
+    },
+    {
+      "epoch": 2.256052141527002,
+      "grad_norm": 0.5995343163976833,
+      "learning_rate": 7.947590845581113e-07,
+      "loss": 0.3295,
+      "step": 4846
+    },
+    {
+      "epoch": 2.2565176908752327,
+      "grad_norm": 0.6260392279981402,
+      "learning_rate": 7.938210114295897e-07,
+      "loss": 0.3131,
+      "step": 4847
+    },
+    {
+      "epoch": 2.2569832402234637,
+      "grad_norm": 0.6342550709896897,
+      "learning_rate": 7.928833877491865e-07,
+      "loss": 0.3388,
+      "step": 4848
+    },
+    {
+      "epoch": 2.2574487895716944,
+      "grad_norm": 0.6023746348293277,
+      "learning_rate": 7.919462137638934e-07,
+      "loss": 0.3313,
+      "step": 4849
+    },
+    {
+      "epoch": 2.2579143389199254,
+      "grad_norm": 0.6160759548796946,
+      "learning_rate": 7.91009489720585e-07,
+      "loss": 0.3257,
+      "step": 4850
+    },
+    {
+      "epoch": 2.2583798882681565,
+      "grad_norm": 0.5991515688052288,
+      "learning_rate": 7.900732158660157e-07,
+      "loss": 0.334,
+      "step": 4851
+    },
+    {
+      "epoch": 2.2588454376163876,
+      "grad_norm": 0.5935740569835476,
+      "learning_rate": 7.891373924468226e-07,
+      "loss": 0.3258,
+      "step": 4852
+    },
+    {
+      "epoch": 2.259310986964618,
+      "grad_norm": 0.6088123635862658,
+      "learning_rate": 7.882020197095252e-07,
+      "loss": 0.3301,
+      "step": 4853
+    },
+    {
+      "epoch": 2.2597765363128492,
+      "grad_norm": 0.5952198874619187,
+      "learning_rate": 7.872670979005217e-07,
+      "loss": 0.3308,
+      "step": 4854
+    },
+    {
+      "epoch": 2.26024208566108,
+      "grad_norm": 0.6300988922838163,
+      "learning_rate": 7.863326272660912e-07,
+      "loss": 0.3224,
+      "step": 4855
+    },
+    {
+      "epoch": 2.260707635009311,
+      "grad_norm": 0.5751664431819462,
+      "learning_rate": 7.853986080523987e-07,
+      "loss": 0.3207,
+      "step": 4856
+    },
+    {
+      "epoch": 2.261173184357542,
+      "grad_norm": 0.6008867361049617,
+      "learning_rate": 7.844650405054857e-07,
+      "loss": 0.3175,
+      "step": 4857
+    },
+    {
+      "epoch": 2.2616387337057726,
+      "grad_norm": 0.5903484054516316,
+      "learning_rate": 7.83531924871275e-07,
+      "loss": 0.3286,
+      "step": 4858
+    },
+    {
+      "epoch": 2.2621042830540037,
+      "grad_norm": 0.6272169850433967,
+      "learning_rate": 7.825992613955724e-07,
+      "loss": 0.337,
+      "step": 4859
+    },
+    {
+      "epoch": 2.2625698324022347,
+      "grad_norm": 0.6098575583955017,
+      "learning_rate": 7.816670503240645e-07,
+      "loss": 0.3129,
+      "step": 4860
+    },
+    {
+      "epoch": 2.2630353817504654,
+      "grad_norm": 0.6149844045631253,
+      "learning_rate": 7.807352919023161e-07,
+      "loss": 0.3285,
+      "step": 4861
+    },
+    {
+      "epoch": 2.2635009310986964,
+      "grad_norm": 0.6321585692690022,
+      "learning_rate": 7.798039863757754e-07,
+      "loss": 0.3142,
+      "step": 4862
+    },
+    {
+      "epoch": 2.2639664804469275,
+      "grad_norm": 0.5905470169858057,
+      "learning_rate": 7.788731339897713e-07,
+      "loss": 0.3191,
+      "step": 4863
+    },
+    {
+      "epoch": 2.264432029795158,
+      "grad_norm": 0.5988172132994263,
+      "learning_rate": 7.779427349895105e-07,
+      "loss": 0.3256,
+      "step": 4864
+    },
+    {
+      "epoch": 2.264897579143389,
+      "grad_norm": 0.601203379851939,
+      "learning_rate": 7.770127896200844e-07,
+      "loss": 0.3253,
+      "step": 4865
+    },
+    {
+      "epoch": 2.2653631284916202,
+      "grad_norm": 0.6168290358006119,
+      "learning_rate": 7.760832981264607e-07,
+      "loss": 0.3026,
+      "step": 4866
+    },
+    {
+      "epoch": 2.265828677839851,
+      "grad_norm": 0.6327432241051099,
+      "learning_rate": 7.751542607534903e-07,
+      "loss": 0.3306,
+      "step": 4867
+    },
+    {
+      "epoch": 2.266294227188082,
+      "grad_norm": 0.6406020424891687,
+      "learning_rate": 7.742256777459048e-07,
+      "loss": 0.3424,
+      "step": 4868
+    },
+    {
+      "epoch": 2.266759776536313,
+      "grad_norm": 0.6286121201907415,
+      "learning_rate": 7.732975493483136e-07,
+      "loss": 0.3382,
+      "step": 4869
+    },
+    {
+      "epoch": 2.2672253258845436,
+      "grad_norm": 0.5977446851033588,
+      "learning_rate": 7.723698758052082e-07,
+      "loss": 0.3332,
+      "step": 4870
+    },
+    {
+      "epoch": 2.2676908752327747,
+      "grad_norm": 0.606922549255984,
+      "learning_rate": 7.714426573609606e-07,
+      "loss": 0.3311,
+      "step": 4871
+    },
+    {
+      "epoch": 2.2681564245810057,
+      "grad_norm": 0.6064534262594649,
+      "learning_rate": 7.705158942598209e-07,
+      "loss": 0.3298,
+      "step": 4872
+    },
+    {
+      "epoch": 2.2686219739292364,
+      "grad_norm": 0.5828624573973091,
+      "learning_rate": 7.695895867459221e-07,
+      "loss": 0.3238,
+      "step": 4873
+    },
+    {
+      "epoch": 2.2690875232774674,
+      "grad_norm": 0.608304665496331,
+      "learning_rate": 7.686637350632742e-07,
+      "loss": 0.319,
+      "step": 4874
+    },
+    {
+      "epoch": 2.2695530726256985,
+      "grad_norm": 0.5975644861516702,
+      "learning_rate": 7.677383394557689e-07,
+      "loss": 0.3214,
+      "step": 4875
+    },
+    {
+      "epoch": 2.270018621973929,
+      "grad_norm": 0.5817681967370206,
+      "learning_rate": 7.668134001671784e-07,
+      "loss": 0.311,
+      "step": 4876
+    },
+    {
+      "epoch": 2.27048417132216,
+      "grad_norm": 0.6070090727123633,
+      "learning_rate": 7.658889174411529e-07,
+      "loss": 0.3348,
+      "step": 4877
+    },
+    {
+      "epoch": 2.2709497206703912,
+      "grad_norm": 0.5960541253431433,
+      "learning_rate": 7.649648915212215e-07,
+      "loss": 0.3396,
+      "step": 4878
+    },
+    {
+      "epoch": 2.271415270018622,
+      "grad_norm": 0.6020979114034763,
+      "learning_rate": 7.640413226507982e-07,
+      "loss": 0.3254,
+      "step": 4879
+    },
+    {
+      "epoch": 2.271880819366853,
+      "grad_norm": 0.6170334251836382,
+      "learning_rate": 7.631182110731708e-07,
+      "loss": 0.3331,
+      "step": 4880
+    },
+    {
+      "epoch": 2.272346368715084,
+      "grad_norm": 0.637142051047782,
+      "learning_rate": 7.621955570315084e-07,
+      "loss": 0.3151,
+      "step": 4881
+    },
+    {
+      "epoch": 2.2728119180633146,
+      "grad_norm": 0.617178140703126,
+      "learning_rate": 7.612733607688608e-07,
+      "loss": 0.3337,
+      "step": 4882
+    },
+    {
+      "epoch": 2.2732774674115457,
+      "grad_norm": 0.6074215280908556,
+      "learning_rate": 7.603516225281571e-07,
+      "loss": 0.3265,
+      "step": 4883
+    },
+    {
+      "epoch": 2.2737430167597763,
+      "grad_norm": 0.6007284804489831,
+      "learning_rate": 7.594303425522035e-07,
+      "loss": 0.3159,
+      "step": 4884
+    },
+    {
+      "epoch": 2.2742085661080074,
+      "grad_norm": 0.6014541011033879,
+      "learning_rate": 7.585095210836885e-07,
+      "loss": 0.3302,
+      "step": 4885
+    },
+    {
+      "epoch": 2.2746741154562384,
+      "grad_norm": 0.6038482669394285,
+      "learning_rate": 7.575891583651771e-07,
+      "loss": 0.3171,
+      "step": 4886
+    },
+    {
+      "epoch": 2.2751396648044695,
+      "grad_norm": 0.6974382033710796,
+      "learning_rate": 7.566692546391155e-07,
+      "loss": 0.3268,
+      "step": 4887
+    },
+    {
+      "epoch": 2.2756052141527,
+      "grad_norm": 0.5932890439152789,
+      "learning_rate": 7.557498101478286e-07,
+      "loss": 0.3263,
+      "step": 4888
+    },
+    {
+      "epoch": 2.276070763500931,
+      "grad_norm": 0.6126138368397622,
+      "learning_rate": 7.54830825133519e-07,
+      "loss": 0.3173,
+      "step": 4889
+    },
+    {
+      "epoch": 2.276536312849162,
+      "grad_norm": 0.6421957233456274,
+      "learning_rate": 7.539122998382695e-07,
+      "loss": 0.3233,
+      "step": 4890
+    },
+    {
+      "epoch": 2.277001862197393,
+      "grad_norm": 0.5764164761346097,
+      "learning_rate": 7.529942345040423e-07,
+      "loss": 0.3189,
+      "step": 4891
+    },
+    {
+      "epoch": 2.277467411545624,
+      "grad_norm": 0.6248913284959914,
+      "learning_rate": 7.520766293726761e-07,
+      "loss": 0.3335,
+      "step": 4892
+    },
+    {
+      "epoch": 2.277932960893855,
+      "grad_norm": 0.6033252621011826,
+      "learning_rate": 7.511594846858919e-07,
+      "loss": 0.3349,
+      "step": 4893
+    },
+    {
+      "epoch": 2.2783985102420856,
+      "grad_norm": 0.5902872861230384,
+      "learning_rate": 7.502428006852855e-07,
+      "loss": 0.3341,
+      "step": 4894
+    },
+    {
+      "epoch": 2.2788640595903167,
+      "grad_norm": 0.5707700348070315,
+      "learning_rate": 7.493265776123345e-07,
+      "loss": 0.3244,
+      "step": 4895
+    },
+    {
+      "epoch": 2.2793296089385473,
+      "grad_norm": 0.6253639547515015,
+      "learning_rate": 7.48410815708393e-07,
+      "loss": 0.3344,
+      "step": 4896
+    },
+    {
+      "epoch": 2.2797951582867784,
+      "grad_norm": 0.600912880498408,
+      "learning_rate": 7.474955152146956e-07,
+      "loss": 0.3324,
+      "step": 4897
+    },
+    {
+      "epoch": 2.2802607076350094,
+      "grad_norm": 0.6083952141596127,
+      "learning_rate": 7.465806763723527e-07,
+      "loss": 0.3334,
+      "step": 4898
+    },
+    {
+      "epoch": 2.28072625698324,
+      "grad_norm": 0.6203009934387842,
+      "learning_rate": 7.456662994223555e-07,
+      "loss": 0.3346,
+      "step": 4899
+    },
+    {
+      "epoch": 2.281191806331471,
+      "grad_norm": 0.5895513174585134,
+      "learning_rate": 7.447523846055729e-07,
+      "loss": 0.3394,
+      "step": 4900
+    },
+    {
+      "epoch": 2.281657355679702,
+      "grad_norm": 0.6279067781740296,
+      "learning_rate": 7.438389321627512e-07,
+      "loss": 0.3098,
+      "step": 4901
+    },
+    {
+      "epoch": 2.282122905027933,
+      "grad_norm": 0.6178328435499576,
+      "learning_rate": 7.429259423345153e-07,
+      "loss": 0.3304,
+      "step": 4902
+    },
+    {
+      "epoch": 2.282588454376164,
+      "grad_norm": 0.6053063746481857,
+      "learning_rate": 7.420134153613698e-07,
+      "loss": 0.3394,
+      "step": 4903
+    },
+    {
+      "epoch": 2.283054003724395,
+      "grad_norm": 0.6288771667436471,
+      "learning_rate": 7.411013514836946e-07,
+      "loss": 0.3236,
+      "step": 4904
+    },
+    {
+      "epoch": 2.2835195530726256,
+      "grad_norm": 0.6300098461047636,
+      "learning_rate": 7.401897509417491e-07,
+      "loss": 0.3173,
+      "step": 4905
+    },
+    {
+      "epoch": 2.2839851024208566,
+      "grad_norm": 0.5952877661846168,
+      "learning_rate": 7.392786139756716e-07,
+      "loss": 0.3197,
+      "step": 4906
+    },
+    {
+      "epoch": 2.2844506517690877,
+      "grad_norm": 0.5919012892967439,
+      "learning_rate": 7.383679408254762e-07,
+      "loss": 0.3248,
+      "step": 4907
+    },
+    {
+      "epoch": 2.2849162011173183,
+      "grad_norm": 0.6149992498128315,
+      "learning_rate": 7.374577317310569e-07,
+      "loss": 0.3325,
+      "step": 4908
+    },
+    {
+      "epoch": 2.2853817504655494,
+      "grad_norm": 0.603176206056785,
+      "learning_rate": 7.365479869321829e-07,
+      "loss": 0.3312,
+      "step": 4909
+    },
+    {
+      "epoch": 2.2858472998137804,
+      "grad_norm": 0.6031897935345014,
+      "learning_rate": 7.356387066685034e-07,
+      "loss": 0.3265,
+      "step": 4910
+    },
+    {
+      "epoch": 2.286312849162011,
+      "grad_norm": 0.6135577904972285,
+      "learning_rate": 7.347298911795456e-07,
+      "loss": 0.3296,
+      "step": 4911
+    },
+    {
+      "epoch": 2.286778398510242,
+      "grad_norm": 0.5845714115037669,
+      "learning_rate": 7.338215407047117e-07,
+      "loss": 0.3194,
+      "step": 4912
+    },
+    {
+      "epoch": 2.287243947858473,
+      "grad_norm": 0.6899518291498753,
+      "learning_rate": 7.329136554832819e-07,
+      "loss": 0.3324,
+      "step": 4913
+    },
+    {
+      "epoch": 2.287709497206704,
+      "grad_norm": 0.6022676427231513,
+      "learning_rate": 7.320062357544172e-07,
+      "loss": 0.3282,
+      "step": 4914
+    },
+    {
+      "epoch": 2.288175046554935,
+      "grad_norm": 0.5915539504820428,
+      "learning_rate": 7.310992817571527e-07,
+      "loss": 0.3103,
+      "step": 4915
+    },
+    {
+      "epoch": 2.288640595903166,
+      "grad_norm": 0.6521722858614443,
+      "learning_rate": 7.301927937304005e-07,
+      "loss": 0.335,
+      "step": 4916
+    },
+    {
+      "epoch": 2.2891061452513966,
+      "grad_norm": 0.5842298912438134,
+      "learning_rate": 7.292867719129517e-07,
+      "loss": 0.3116,
+      "step": 4917
+    },
+    {
+      "epoch": 2.2895716945996276,
+      "grad_norm": 0.6740591843321365,
+      "learning_rate": 7.28381216543475e-07,
+      "loss": 0.3326,
+      "step": 4918
+    },
+    {
+      "epoch": 2.2900372439478582,
+      "grad_norm": 0.5952987124892885,
+      "learning_rate": 7.274761278605136e-07,
+      "loss": 0.3111,
+      "step": 4919
+    },
+    {
+      "epoch": 2.2905027932960893,
+      "grad_norm": 0.5944325573012672,
+      "learning_rate": 7.265715061024911e-07,
+      "loss": 0.3215,
+      "step": 4920
+    },
+    {
+      "epoch": 2.2909683426443204,
+      "grad_norm": 0.659892049536303,
+      "learning_rate": 7.256673515077045e-07,
+      "loss": 0.3437,
+      "step": 4921
+    },
+    {
+      "epoch": 2.2914338919925514,
+      "grad_norm": 0.6252673725519756,
+      "learning_rate": 7.247636643143308e-07,
+      "loss": 0.3278,
+      "step": 4922
+    },
+    {
+      "epoch": 2.291899441340782,
+      "grad_norm": 0.5844622467177158,
+      "learning_rate": 7.23860444760423e-07,
+      "loss": 0.3164,
+      "step": 4923
+    },
+    {
+      "epoch": 2.292364990689013,
+      "grad_norm": 0.5710652438058175,
+      "learning_rate": 7.229576930839097e-07,
+      "loss": 0.3202,
+      "step": 4924
+    },
+    {
+      "epoch": 2.2928305400372437,
+      "grad_norm": 0.6356533064831142,
+      "learning_rate": 7.220554095225974e-07,
+      "loss": 0.3472,
+      "step": 4925
+    },
+    {
+      "epoch": 2.293296089385475,
+      "grad_norm": 0.6137098271071187,
+      "learning_rate": 7.211535943141698e-07,
+      "loss": 0.3121,
+      "step": 4926
+    },
+    {
+      "epoch": 2.293761638733706,
+      "grad_norm": 0.600064273765987,
+      "learning_rate": 7.202522476961852e-07,
+      "loss": 0.3183,
+      "step": 4927
+    },
+    {
+      "epoch": 2.294227188081937,
+      "grad_norm": 0.5761405377291697,
+      "learning_rate": 7.193513699060811e-07,
+      "loss": 0.3292,
+      "step": 4928
+    },
+    {
+      "epoch": 2.2946927374301676,
+      "grad_norm": 0.5997284589143109,
+      "learning_rate": 7.184509611811688e-07,
+      "loss": 0.3126,
+      "step": 4929
+    },
+    {
+      "epoch": 2.2951582867783986,
+      "grad_norm": 0.5909914170153734,
+      "learning_rate": 7.175510217586379e-07,
+      "loss": 0.3452,
+      "step": 4930
+    },
+    {
+      "epoch": 2.2956238361266292,
+      "grad_norm": 0.6258175274896018,
+      "learning_rate": 7.166515518755548e-07,
+      "loss": 0.335,
+      "step": 4931
+    },
+    {
+      "epoch": 2.2960893854748603,
+      "grad_norm": 0.6014775359601097,
+      "learning_rate": 7.157525517688604e-07,
+      "loss": 0.3378,
+      "step": 4932
+    },
+    {
+      "epoch": 2.2965549348230914,
+      "grad_norm": 0.5981622713879425,
+      "learning_rate": 7.14854021675371e-07,
+      "loss": 0.3037,
+      "step": 4933
+    },
+    {
+      "epoch": 2.297020484171322,
+      "grad_norm": 0.6223976744258753,
+      "learning_rate": 7.139559618317843e-07,
+      "loss": 0.331,
+      "step": 4934
+    },
+    {
+      "epoch": 2.297486033519553,
+      "grad_norm": 0.6173048769132924,
+      "learning_rate": 7.130583724746687e-07,
+      "loss": 0.3355,
+      "step": 4935
+    },
+    {
+      "epoch": 2.297951582867784,
+      "grad_norm": 0.5984466285990032,
+      "learning_rate": 7.121612538404701e-07,
+      "loss": 0.3347,
+      "step": 4936
+    },
+    {
+      "epoch": 2.2984171322160147,
+      "grad_norm": 0.5962542577415607,
+      "learning_rate": 7.112646061655112e-07,
+      "loss": 0.3337,
+      "step": 4937
+    },
+    {
+      "epoch": 2.298882681564246,
+      "grad_norm": 0.6061530000402422,
+      "learning_rate": 7.103684296859917e-07,
+      "loss": 0.3282,
+      "step": 4938
+    },
+    {
+      "epoch": 2.299348230912477,
+      "grad_norm": 0.6017189339416066,
+      "learning_rate": 7.094727246379837e-07,
+      "loss": 0.3305,
+      "step": 4939
+    },
+    {
+      "epoch": 2.2998137802607075,
+      "grad_norm": 0.6054655598775827,
+      "learning_rate": 7.085774912574389e-07,
+      "loss": 0.3456,
+      "step": 4940
+    },
+    {
+      "epoch": 2.3002793296089385,
+      "grad_norm": 0.6097863525218193,
+      "learning_rate": 7.076827297801817e-07,
+      "loss": 0.3206,
+      "step": 4941
+    },
+    {
+      "epoch": 2.3007448789571696,
+      "grad_norm": 0.630513959411621,
+      "learning_rate": 7.067884404419143e-07,
+      "loss": 0.3424,
+      "step": 4942
+    },
+    {
+      "epoch": 2.3012104283054002,
+      "grad_norm": 0.5972668315700271,
+      "learning_rate": 7.058946234782143e-07,
+      "loss": 0.3113,
+      "step": 4943
+    },
+    {
+      "epoch": 2.3016759776536313,
+      "grad_norm": 0.5891626795551355,
+      "learning_rate": 7.050012791245333e-07,
+      "loss": 0.3371,
+      "step": 4944
+    },
+    {
+      "epoch": 2.3021415270018624,
+      "grad_norm": 0.631396387290339,
+      "learning_rate": 7.041084076161997e-07,
+      "loss": 0.3375,
+      "step": 4945
+    },
+    {
+      "epoch": 2.302607076350093,
+      "grad_norm": 0.6375512139037617,
+      "learning_rate": 7.03216009188418e-07,
+      "loss": 0.3384,
+      "step": 4946
+    },
+    {
+      "epoch": 2.303072625698324,
+      "grad_norm": 0.5706653679257795,
+      "learning_rate": 7.023240840762657e-07,
+      "loss": 0.3265,
+      "step": 4947
+    },
+    {
+      "epoch": 2.303538175046555,
+      "grad_norm": 0.6149328410689094,
+      "learning_rate": 7.014326325146977e-07,
+      "loss": 0.3277,
+      "step": 4948
+    },
+    {
+      "epoch": 2.3040037243947857,
+      "grad_norm": 0.6341842511831409,
+      "learning_rate": 7.005416547385447e-07,
+      "loss": 0.3219,
+      "step": 4949
+    },
+    {
+      "epoch": 2.304469273743017,
+      "grad_norm": 0.6202405236458989,
+      "learning_rate": 6.996511509825094e-07,
+      "loss": 0.3154,
+      "step": 4950
+    },
+    {
+      "epoch": 2.304934823091248,
+      "grad_norm": 0.6168758160503677,
+      "learning_rate": 6.987611214811735e-07,
+      "loss": 0.3342,
+      "step": 4951
+    },
+    {
+      "epoch": 2.3054003724394785,
+      "grad_norm": 0.5955623749807532,
+      "learning_rate": 6.978715664689903e-07,
+      "loss": 0.3284,
+      "step": 4952
+    },
+    {
+      "epoch": 2.3058659217877095,
+      "grad_norm": 0.5879750745823689,
+      "learning_rate": 6.969824861802907e-07,
+      "loss": 0.3114,
+      "step": 4953
+    },
+    {
+      "epoch": 2.30633147113594,
+      "grad_norm": 0.6425221871253277,
+      "learning_rate": 6.9609388084928e-07,
+      "loss": 0.3299,
+      "step": 4954
+    },
+    {
+      "epoch": 2.3067970204841712,
+      "grad_norm": 0.6035205170991506,
+      "learning_rate": 6.952057507100377e-07,
+      "loss": 0.3498,
+      "step": 4955
+    },
+    {
+      "epoch": 2.3072625698324023,
+      "grad_norm": 0.6033031126571419,
+      "learning_rate": 6.943180959965162e-07,
+      "loss": 0.3214,
+      "step": 4956
+    },
+    {
+      "epoch": 2.3077281191806334,
+      "grad_norm": 0.6346994520615494,
+      "learning_rate": 6.934309169425485e-07,
+      "loss": 0.3282,
+      "step": 4957
+    },
+    {
+      "epoch": 2.308193668528864,
+      "grad_norm": 0.606402216765471,
+      "learning_rate": 6.925442137818369e-07,
+      "loss": 0.325,
+      "step": 4958
+    },
+    {
+      "epoch": 2.308659217877095,
+      "grad_norm": 0.6119881744353912,
+      "learning_rate": 6.916579867479595e-07,
+      "loss": 0.3218,
+      "step": 4959
+    },
+    {
+      "epoch": 2.3091247672253257,
+      "grad_norm": 0.6072494931738304,
+      "learning_rate": 6.907722360743699e-07,
+      "loss": 0.3261,
+      "step": 4960
+    },
+    {
+      "epoch": 2.3095903165735567,
+      "grad_norm": 0.6121253525672323,
+      "learning_rate": 6.898869619943971e-07,
+      "loss": 0.3253,
+      "step": 4961
+    },
+    {
+      "epoch": 2.310055865921788,
+      "grad_norm": 0.6565094081332119,
+      "learning_rate": 6.890021647412417e-07,
+      "loss": 0.345,
+      "step": 4962
+    },
+    {
+      "epoch": 2.310521415270019,
+      "grad_norm": 0.6387621489908528,
+      "learning_rate": 6.881178445479819e-07,
+      "loss": 0.3166,
+      "step": 4963
+    },
+    {
+      "epoch": 2.3109869646182495,
+      "grad_norm": 0.6052723652987962,
+      "learning_rate": 6.87234001647567e-07,
+      "loss": 0.3358,
+      "step": 4964
+    },
+    {
+      "epoch": 2.3114525139664805,
+      "grad_norm": 0.6097804874107401,
+      "learning_rate": 6.863506362728232e-07,
+      "loss": 0.3134,
+      "step": 4965
+    },
+    {
+      "epoch": 2.311918063314711,
+      "grad_norm": 0.614679831915768,
+      "learning_rate": 6.854677486564504e-07,
+      "loss": 0.3218,
+      "step": 4966
+    },
+    {
+      "epoch": 2.3123836126629422,
+      "grad_norm": 0.6017118052998394,
+      "learning_rate": 6.845853390310208e-07,
+      "loss": 0.325,
+      "step": 4967
+    },
+    {
+      "epoch": 2.3128491620111733,
+      "grad_norm": 0.642546380404536,
+      "learning_rate": 6.83703407628983e-07,
+      "loss": 0.3295,
+      "step": 4968
+    },
+    {
+      "epoch": 2.313314711359404,
+      "grad_norm": 0.6112717349562332,
+      "learning_rate": 6.828219546826595e-07,
+      "loss": 0.3284,
+      "step": 4969
+    },
+    {
+      "epoch": 2.313780260707635,
+      "grad_norm": 0.5943060134329762,
+      "learning_rate": 6.819409804242442e-07,
+      "loss": 0.3367,
+      "step": 4970
+    },
+    {
+      "epoch": 2.314245810055866,
+      "grad_norm": 0.6472124488799097,
+      "learning_rate": 6.810604850858082e-07,
+      "loss": 0.337,
+      "step": 4971
+    },
+    {
+      "epoch": 2.3147113594040967,
+      "grad_norm": 0.5928616426233901,
+      "learning_rate": 6.801804688992936e-07,
+      "loss": 0.3302,
+      "step": 4972
+    },
+    {
+      "epoch": 2.3151769087523277,
+      "grad_norm": 0.6098173013238398,
+      "learning_rate": 6.793009320965189e-07,
+      "loss": 0.3411,
+      "step": 4973
+    },
+    {
+      "epoch": 2.315642458100559,
+      "grad_norm": 0.6027257399380582,
+      "learning_rate": 6.784218749091737e-07,
+      "loss": 0.3396,
+      "step": 4974
+    },
+    {
+      "epoch": 2.3161080074487894,
+      "grad_norm": 0.5927673409068638,
+      "learning_rate": 6.775432975688237e-07,
+      "loss": 0.3391,
+      "step": 4975
+    },
+    {
+      "epoch": 2.3165735567970205,
+      "grad_norm": 0.6058443457252043,
+      "learning_rate": 6.766652003069061e-07,
+      "loss": 0.3333,
+      "step": 4976
+    },
+    {
+      "epoch": 2.3170391061452515,
+      "grad_norm": 0.6076765695437694,
+      "learning_rate": 6.757875833547329e-07,
+      "loss": 0.3469,
+      "step": 4977
+    },
+    {
+      "epoch": 2.317504655493482,
+      "grad_norm": 0.6342986898981172,
+      "learning_rate": 6.749104469434903e-07,
+      "loss": 0.3266,
+      "step": 4978
+    },
+    {
+      "epoch": 2.3179702048417132,
+      "grad_norm": 0.6144994465687779,
+      "learning_rate": 6.740337913042353e-07,
+      "loss": 0.3243,
+      "step": 4979
+    },
+    {
+      "epoch": 2.3184357541899443,
+      "grad_norm": 0.5652253606910892,
+      "learning_rate": 6.731576166679005e-07,
+      "loss": 0.3068,
+      "step": 4980
+    },
+    {
+      "epoch": 2.318901303538175,
+      "grad_norm": 0.6551628781606844,
+      "learning_rate": 6.722819232652919e-07,
+      "loss": 0.3387,
+      "step": 4981
+    },
+    {
+      "epoch": 2.319366852886406,
+      "grad_norm": 0.6042813080467304,
+      "learning_rate": 6.714067113270867e-07,
+      "loss": 0.3378,
+      "step": 4982
+    },
+    {
+      "epoch": 2.319832402234637,
+      "grad_norm": 0.6142090426504356,
+      "learning_rate": 6.705319810838379e-07,
+      "loss": 0.3304,
+      "step": 4983
+    },
+    {
+      "epoch": 2.3202979515828677,
+      "grad_norm": 0.5914214099704826,
+      "learning_rate": 6.696577327659684e-07,
+      "loss": 0.3194,
+      "step": 4984
+    },
+    {
+      "epoch": 2.3207635009310987,
+      "grad_norm": 0.6188518683028421,
+      "learning_rate": 6.687839666037774e-07,
+      "loss": 0.3177,
+      "step": 4985
+    },
+    {
+      "epoch": 2.32122905027933,
+      "grad_norm": 0.599480091494078,
+      "learning_rate": 6.67910682827436e-07,
+      "loss": 0.3265,
+      "step": 4986
+    },
+    {
+      "epoch": 2.3216945996275604,
+      "grad_norm": 0.6274768139884748,
+      "learning_rate": 6.67037881666987e-07,
+      "loss": 0.3266,
+      "step": 4987
+    },
+    {
+      "epoch": 2.3221601489757915,
+      "grad_norm": 0.6495841762695457,
+      "learning_rate": 6.661655633523459e-07,
+      "loss": 0.3455,
+      "step": 4988
+    },
+    {
+      "epoch": 2.322625698324022,
+      "grad_norm": 0.5964318770604712,
+      "learning_rate": 6.652937281133051e-07,
+      "loss": 0.3199,
+      "step": 4989
+    },
+    {
+      "epoch": 2.323091247672253,
+      "grad_norm": 0.5803726604644048,
+      "learning_rate": 6.644223761795248e-07,
+      "loss": 0.316,
+      "step": 4990
+    },
+    {
+      "epoch": 2.3235567970204842,
+      "grad_norm": 0.6329101101796978,
+      "learning_rate": 6.635515077805388e-07,
+      "loss": 0.3171,
+      "step": 4991
+    },
+    {
+      "epoch": 2.3240223463687153,
+      "grad_norm": 0.6271173365291863,
+      "learning_rate": 6.626811231457573e-07,
+      "loss": 0.3411,
+      "step": 4992
+    },
+    {
+      "epoch": 2.324487895716946,
+      "grad_norm": 0.5961024347010396,
+      "learning_rate": 6.618112225044587e-07,
+      "loss": 0.3158,
+      "step": 4993
+    },
+    {
+      "epoch": 2.324953445065177,
+      "grad_norm": 0.6098007126542002,
+      "learning_rate": 6.60941806085795e-07,
+      "loss": 0.3171,
+      "step": 4994
+    },
+    {
+      "epoch": 2.3254189944134076,
+      "grad_norm": 0.5926694076823563,
+      "learning_rate": 6.600728741187917e-07,
+      "loss": 0.3184,
+      "step": 4995
+    },
+    {
+      "epoch": 2.3258845437616387,
+      "grad_norm": 0.6124779463167369,
+      "learning_rate": 6.592044268323469e-07,
+      "loss": 0.3231,
+      "step": 4996
+    },
+    {
+      "epoch": 2.3263500931098697,
+      "grad_norm": 0.5797324328167943,
+      "learning_rate": 6.58336464455229e-07,
+      "loss": 0.3161,
+      "step": 4997
+    },
+    {
+      "epoch": 2.326815642458101,
+      "grad_norm": 0.6003087896329584,
+      "learning_rate": 6.574689872160814e-07,
+      "loss": 0.3265,
+      "step": 4998
+    },
+    {
+      "epoch": 2.3272811918063314,
+      "grad_norm": 0.6318533199012003,
+      "learning_rate": 6.566019953434166e-07,
+      "loss": 0.3449,
+      "step": 4999
+    },
+    {
+      "epoch": 2.3277467411545625,
+      "grad_norm": 0.577632094995157,
+      "learning_rate": 6.557354890656215e-07,
+      "loss": 0.3188,
+      "step": 5000
+    },
+    {
+      "epoch": 2.328212290502793,
+      "grad_norm": 0.5729781610168442,
+      "learning_rate": 6.548694686109555e-07,
+      "loss": 0.3221,
+      "step": 5001
+    },
+    {
+      "epoch": 2.328677839851024,
+      "grad_norm": 0.5722388015306568,
+      "learning_rate": 6.540039342075474e-07,
+      "loss": 0.3138,
+      "step": 5002
+    },
+    {
+      "epoch": 2.3291433891992552,
+      "grad_norm": 0.6091315001591282,
+      "learning_rate": 6.531388860834001e-07,
+      "loss": 0.3318,
+      "step": 5003
+    },
+    {
+      "epoch": 2.329608938547486,
+      "grad_norm": 0.5609883348314515,
+      "learning_rate": 6.522743244663887e-07,
+      "loss": 0.3107,
+      "step": 5004
+    },
+    {
+      "epoch": 2.330074487895717,
+      "grad_norm": 0.6109099932390909,
+      "learning_rate": 6.514102495842581e-07,
+      "loss": 0.3295,
+      "step": 5005
+    },
+    {
+      "epoch": 2.330540037243948,
+      "grad_norm": 0.6220026121501105,
+      "learning_rate": 6.505466616646272e-07,
+      "loss": 0.3246,
+      "step": 5006
+    },
+    {
+      "epoch": 2.3310055865921786,
+      "grad_norm": 0.6040012576769672,
+      "learning_rate": 6.496835609349842e-07,
+      "loss": 0.3447,
+      "step": 5007
+    },
+    {
+      "epoch": 2.3314711359404097,
+      "grad_norm": 0.6173233821761919,
+      "learning_rate": 6.488209476226914e-07,
+      "loss": 0.3248,
+      "step": 5008
+    },
+    {
+      "epoch": 2.3319366852886407,
+      "grad_norm": 0.6108774626255816,
+      "learning_rate": 6.479588219549821e-07,
+      "loss": 0.332,
+      "step": 5009
+    },
+    {
+      "epoch": 2.3324022346368714,
+      "grad_norm": 0.5976801518459264,
+      "learning_rate": 6.470971841589599e-07,
+      "loss": 0.3286,
+      "step": 5010
+    },
+    {
+      "epoch": 2.3328677839851024,
+      "grad_norm": 0.6312291747261409,
+      "learning_rate": 6.462360344615997e-07,
+      "loss": 0.3225,
+      "step": 5011
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.6136601343210781,
+      "learning_rate": 6.453753730897511e-07,
+      "loss": 0.3183,
+      "step": 5012
+    },
+    {
+      "epoch": 2.333798882681564,
+      "grad_norm": 0.6194672169579984,
+      "learning_rate": 6.445152002701318e-07,
+      "loss": 0.3247,
+      "step": 5013
+    },
+    {
+      "epoch": 2.334264432029795,
+      "grad_norm": 0.5891364214735111,
+      "learning_rate": 6.43655516229331e-07,
+      "loss": 0.3318,
+      "step": 5014
+    },
+    {
+      "epoch": 2.3347299813780262,
+      "grad_norm": 0.625267584998024,
+      "learning_rate": 6.427963211938101e-07,
+      "loss": 0.3344,
+      "step": 5015
+    },
+    {
+      "epoch": 2.335195530726257,
+      "grad_norm": 0.653766803255445,
+      "learning_rate": 6.419376153899032e-07,
+      "loss": 0.3453,
+      "step": 5016
+    },
+    {
+      "epoch": 2.335661080074488,
+      "grad_norm": 0.6061094105841327,
+      "learning_rate": 6.410793990438113e-07,
+      "loss": 0.3204,
+      "step": 5017
+    },
+    {
+      "epoch": 2.336126629422719,
+      "grad_norm": 0.5914615222653541,
+      "learning_rate": 6.402216723816113e-07,
+      "loss": 0.3156,
+      "step": 5018
+    },
+    {
+      "epoch": 2.3365921787709496,
+      "grad_norm": 0.6233162022767375,
+      "learning_rate": 6.393644356292469e-07,
+      "loss": 0.3122,
+      "step": 5019
+    },
+    {
+      "epoch": 2.3370577281191807,
+      "grad_norm": 0.6010397899698318,
+      "learning_rate": 6.385076890125355e-07,
+      "loss": 0.3313,
+      "step": 5020
+    },
+    {
+      "epoch": 2.3375232774674117,
+      "grad_norm": 0.5862738094467911,
+      "learning_rate": 6.376514327571653e-07,
+      "loss": 0.3209,
+      "step": 5021
+    },
+    {
+      "epoch": 2.3379888268156424,
+      "grad_norm": 0.6303251738427538,
+      "learning_rate": 6.36795667088693e-07,
+      "loss": 0.3395,
+      "step": 5022
+    },
+    {
+      "epoch": 2.3384543761638734,
+      "grad_norm": 0.6029121744499348,
+      "learning_rate": 6.359403922325482e-07,
+      "loss": 0.3283,
+      "step": 5023
+    },
+    {
+      "epoch": 2.338919925512104,
+      "grad_norm": 0.6138426767773835,
+      "learning_rate": 6.35085608414032e-07,
+      "loss": 0.3266,
+      "step": 5024
+    },
+    {
+      "epoch": 2.339385474860335,
+      "grad_norm": 0.6467852098347975,
+      "learning_rate": 6.342313158583127e-07,
+      "loss": 0.3199,
+      "step": 5025
+    },
+    {
+      "epoch": 2.339851024208566,
+      "grad_norm": 0.6087614504368968,
+      "learning_rate": 6.333775147904328e-07,
+      "loss": 0.3173,
+      "step": 5026
+    },
+    {
+      "epoch": 2.3403165735567972,
+      "grad_norm": 0.6058561056379663,
+      "learning_rate": 6.325242054353025e-07,
+      "loss": 0.3312,
+      "step": 5027
+    },
+    {
+      "epoch": 2.340782122905028,
+      "grad_norm": 0.5910339652116244,
+      "learning_rate": 6.316713880177047e-07,
+      "loss": 0.3221,
+      "step": 5028
+    },
+    {
+      "epoch": 2.341247672253259,
+      "grad_norm": 0.6475521789444902,
+      "learning_rate": 6.308190627622923e-07,
+      "loss": 0.3302,
+      "step": 5029
+    },
+    {
+      "epoch": 2.3417132216014895,
+      "grad_norm": 0.6035919903937653,
+      "learning_rate": 6.299672298935872e-07,
+      "loss": 0.3069,
+      "step": 5030
+    },
+    {
+      "epoch": 2.3421787709497206,
+      "grad_norm": 0.5963484896541501,
+      "learning_rate": 6.291158896359809e-07,
+      "loss": 0.3175,
+      "step": 5031
+    },
+    {
+      "epoch": 2.3426443202979517,
+      "grad_norm": 0.6367730218910094,
+      "learning_rate": 6.282650422137399e-07,
+      "loss": 0.3286,
+      "step": 5032
+    },
+    {
+      "epoch": 2.3431098696461827,
+      "grad_norm": 0.6167281774409779,
+      "learning_rate": 6.274146878509963e-07,
+      "loss": 0.3081,
+      "step": 5033
+    },
+    {
+      "epoch": 2.3435754189944134,
+      "grad_norm": 0.6185053680149519,
+      "learning_rate": 6.265648267717522e-07,
+      "loss": 0.3404,
+      "step": 5034
+    },
+    {
+      "epoch": 2.3440409683426444,
+      "grad_norm": 0.6212472910859391,
+      "learning_rate": 6.257154591998829e-07,
+      "loss": 0.3293,
+      "step": 5035
+    },
+    {
+      "epoch": 2.344506517690875,
+      "grad_norm": 0.6588508932615591,
+      "learning_rate": 6.248665853591321e-07,
+      "loss": 0.3483,
+      "step": 5036
+    },
+    {
+      "epoch": 2.344972067039106,
+      "grad_norm": 0.65345256344401,
+      "learning_rate": 6.24018205473112e-07,
+      "loss": 0.3278,
+      "step": 5037
+    },
+    {
+      "epoch": 2.345437616387337,
+      "grad_norm": 0.5909786736587087,
+      "learning_rate": 6.231703197653069e-07,
+      "loss": 0.3131,
+      "step": 5038
+    },
+    {
+      "epoch": 2.345903165735568,
+      "grad_norm": 0.6303576672016897,
+      "learning_rate": 6.223229284590707e-07,
+      "loss": 0.3371,
+      "step": 5039
+    },
+    {
+      "epoch": 2.346368715083799,
+      "grad_norm": 0.5856852892150086,
+      "learning_rate": 6.21476031777625e-07,
+      "loss": 0.3162,
+      "step": 5040
+    },
+    {
+      "epoch": 2.34683426443203,
+      "grad_norm": 0.6169192298750829,
+      "learning_rate": 6.206296299440639e-07,
+      "loss": 0.3209,
+      "step": 5041
+    },
+    {
+      "epoch": 2.3472998137802605,
+      "grad_norm": 0.6189798290883255,
+      "learning_rate": 6.197837231813483e-07,
+      "loss": 0.3426,
+      "step": 5042
+    },
+    {
+      "epoch": 2.3477653631284916,
+      "grad_norm": 0.5866715001433894,
+      "learning_rate": 6.189383117123113e-07,
+      "loss": 0.3317,
+      "step": 5043
+    },
+    {
+      "epoch": 2.3482309124767227,
+      "grad_norm": 0.6198866755611747,
+      "learning_rate": 6.180933957596546e-07,
+      "loss": 0.3204,
+      "step": 5044
+    },
+    {
+      "epoch": 2.3486964618249533,
+      "grad_norm": 0.6047789867652217,
+      "learning_rate": 6.17248975545948e-07,
+      "loss": 0.323,
+      "step": 5045
+    },
+    {
+      "epoch": 2.3491620111731844,
+      "grad_norm": 0.6185995463755826,
+      "learning_rate": 6.164050512936323e-07,
+      "loss": 0.3377,
+      "step": 5046
+    },
+    {
+      "epoch": 2.3496275605214154,
+      "grad_norm": 0.6440811995832156,
+      "learning_rate": 6.15561623225018e-07,
+      "loss": 0.3333,
+      "step": 5047
+    },
+    {
+      "epoch": 2.350093109869646,
+      "grad_norm": 0.5982729086920664,
+      "learning_rate": 6.14718691562283e-07,
+      "loss": 0.3356,
+      "step": 5048
+    },
+    {
+      "epoch": 2.350558659217877,
+      "grad_norm": 1.2928607203518518,
+      "learning_rate": 6.138762565274767e-07,
+      "loss": 0.297,
+      "step": 5049
+    },
+    {
+      "epoch": 2.351024208566108,
+      "grad_norm": 0.633317398014539,
+      "learning_rate": 6.130343183425152e-07,
+      "loss": 0.3318,
+      "step": 5050
+    },
+    {
+      "epoch": 2.351489757914339,
+      "grad_norm": 0.6095862055004017,
+      "learning_rate": 6.121928772291858e-07,
+      "loss": 0.3432,
+      "step": 5051
+    },
+    {
+      "epoch": 2.35195530726257,
+      "grad_norm": 0.5984833417494677,
+      "learning_rate": 6.113519334091444e-07,
+      "loss": 0.3439,
+      "step": 5052
+    },
+    {
+      "epoch": 2.352420856610801,
+      "grad_norm": 0.6167751066370426,
+      "learning_rate": 6.105114871039156e-07,
+      "loss": 0.3347,
+      "step": 5053
+    },
+    {
+      "epoch": 2.3528864059590315,
+      "grad_norm": 0.6127609738886197,
+      "learning_rate": 6.096715385348914e-07,
+      "loss": 0.33,
+      "step": 5054
+    },
+    {
+      "epoch": 2.3533519553072626,
+      "grad_norm": 0.6290605713323065,
+      "learning_rate": 6.088320879233358e-07,
+      "loss": 0.3337,
+      "step": 5055
+    },
+    {
+      "epoch": 2.3538175046554937,
+      "grad_norm": 0.6056688887449695,
+      "learning_rate": 6.079931354903801e-07,
+      "loss": 0.3255,
+      "step": 5056
+    },
+    {
+      "epoch": 2.3542830540037243,
+      "grad_norm": 0.6005666507176033,
+      "learning_rate": 6.071546814570231e-07,
+      "loss": 0.3158,
+      "step": 5057
+    },
+    {
+      "epoch": 2.3547486033519553,
+      "grad_norm": 0.6120878771309016,
+      "learning_rate": 6.063167260441344e-07,
+      "loss": 0.3389,
+      "step": 5058
+    },
+    {
+      "epoch": 2.355214152700186,
+      "grad_norm": 0.6089562162067219,
+      "learning_rate": 6.054792694724518e-07,
+      "loss": 0.3269,
+      "step": 5059
+    },
+    {
+      "epoch": 2.355679702048417,
+      "grad_norm": 0.6146475126809352,
+      "learning_rate": 6.046423119625803e-07,
+      "loss": 0.3492,
+      "step": 5060
+    },
+    {
+      "epoch": 2.356145251396648,
+      "grad_norm": 0.5761914970243307,
+      "learning_rate": 6.038058537349953e-07,
+      "loss": 0.3192,
+      "step": 5061
+    },
+    {
+      "epoch": 2.356610800744879,
+      "grad_norm": 0.60881836233187,
+      "learning_rate": 6.029698950100388e-07,
+      "loss": 0.3257,
+      "step": 5062
+    },
+    {
+      "epoch": 2.35707635009311,
+      "grad_norm": 0.6066035231404581,
+      "learning_rate": 6.021344360079226e-07,
+      "loss": 0.3235,
+      "step": 5063
+    },
+    {
+      "epoch": 2.357541899441341,
+      "grad_norm": 0.6101903016360782,
+      "learning_rate": 6.012994769487274e-07,
+      "loss": 0.324,
+      "step": 5064
+    },
+    {
+      "epoch": 2.3580074487895715,
+      "grad_norm": 0.6559176821628129,
+      "learning_rate": 6.004650180524002e-07,
+      "loss": 0.3258,
+      "step": 5065
+    },
+    {
+      "epoch": 2.3584729981378025,
+      "grad_norm": 0.6076208809309024,
+      "learning_rate": 5.996310595387564e-07,
+      "loss": 0.3181,
+      "step": 5066
+    },
+    {
+      "epoch": 2.3589385474860336,
+      "grad_norm": 0.5975309070235684,
+      "learning_rate": 5.987976016274832e-07,
+      "loss": 0.3038,
+      "step": 5067
+    },
+    {
+      "epoch": 2.3594040968342647,
+      "grad_norm": 0.6215062960554794,
+      "learning_rate": 5.979646445381318e-07,
+      "loss": 0.3256,
+      "step": 5068
+    },
+    {
+      "epoch": 2.3598696461824953,
+      "grad_norm": 0.6466707997072592,
+      "learning_rate": 5.971321884901218e-07,
+      "loss": 0.3266,
+      "step": 5069
+    },
+    {
+      "epoch": 2.3603351955307263,
+      "grad_norm": 0.5976545936964516,
+      "learning_rate": 5.963002337027432e-07,
+      "loss": 0.3194,
+      "step": 5070
+    },
+    {
+      "epoch": 2.360800744878957,
+      "grad_norm": 0.6053198936808994,
+      "learning_rate": 5.954687803951531e-07,
+      "loss": 0.3391,
+      "step": 5071
+    },
+    {
+      "epoch": 2.361266294227188,
+      "grad_norm": 0.6055196781469252,
+      "learning_rate": 5.946378287863747e-07,
+      "loss": 0.3242,
+      "step": 5072
+    },
+    {
+      "epoch": 2.361731843575419,
+      "grad_norm": 0.6084099541632102,
+      "learning_rate": 5.938073790953017e-07,
+      "loss": 0.324,
+      "step": 5073
+    },
+    {
+      "epoch": 2.3621973929236497,
+      "grad_norm": 0.5928779509888487,
+      "learning_rate": 5.929774315406931e-07,
+      "loss": 0.3451,
+      "step": 5074
+    },
+    {
+      "epoch": 2.362662942271881,
+      "grad_norm": 0.6177968378461445,
+      "learning_rate": 5.921479863411775e-07,
+      "loss": 0.3394,
+      "step": 5075
+    },
+    {
+      "epoch": 2.363128491620112,
+      "grad_norm": 0.5933864896402657,
+      "learning_rate": 5.91319043715251e-07,
+      "loss": 0.3112,
+      "step": 5076
+    },
+    {
+      "epoch": 2.3635940409683425,
+      "grad_norm": 0.6215358945538525,
+      "learning_rate": 5.904906038812757e-07,
+      "loss": 0.3236,
+      "step": 5077
+    },
+    {
+      "epoch": 2.3640595903165735,
+      "grad_norm": 0.6120122309013961,
+      "learning_rate": 5.896626670574826e-07,
+      "loss": 0.3392,
+      "step": 5078
+    },
+    {
+      "epoch": 2.3645251396648046,
+      "grad_norm": 0.6119120813782631,
+      "learning_rate": 5.888352334619712e-07,
+      "loss": 0.321,
+      "step": 5079
+    },
+    {
+      "epoch": 2.364990689013035,
+      "grad_norm": 0.6076877200940213,
+      "learning_rate": 5.880083033127054e-07,
+      "loss": 0.3119,
+      "step": 5080
+    },
+    {
+      "epoch": 2.3654562383612663,
+      "grad_norm": 0.614104273012463,
+      "learning_rate": 5.87181876827519e-07,
+      "loss": 0.3257,
+      "step": 5081
+    },
+    {
+      "epoch": 2.3659217877094973,
+      "grad_norm": 0.578790026189604,
+      "learning_rate": 5.863559542241132e-07,
+      "loss": 0.3338,
+      "step": 5082
+    },
+    {
+      "epoch": 2.366387337057728,
+      "grad_norm": 0.6117681101113478,
+      "learning_rate": 5.855305357200544e-07,
+      "loss": 0.3153,
+      "step": 5083
+    },
+    {
+      "epoch": 2.366852886405959,
+      "grad_norm": 0.6472066724053935,
+      "learning_rate": 5.847056215327785e-07,
+      "loss": 0.3332,
+      "step": 5084
+    },
+    {
+      "epoch": 2.36731843575419,
+      "grad_norm": 0.6257260097840323,
+      "learning_rate": 5.838812118795862e-07,
+      "loss": 0.3385,
+      "step": 5085
+    },
+    {
+      "epoch": 2.3677839851024207,
+      "grad_norm": 0.6167544026974903,
+      "learning_rate": 5.830573069776474e-07,
+      "loss": 0.3123,
+      "step": 5086
+    },
+    {
+      "epoch": 2.368249534450652,
+      "grad_norm": 0.6073943973394468,
+      "learning_rate": 5.822339070439992e-07,
+      "loss": 0.337,
+      "step": 5087
+    },
+    {
+      "epoch": 2.368715083798883,
+      "grad_norm": 0.628366249087836,
+      "learning_rate": 5.814110122955435e-07,
+      "loss": 0.3322,
+      "step": 5088
+    },
+    {
+      "epoch": 2.3691806331471135,
+      "grad_norm": 0.6040645771987048,
+      "learning_rate": 5.805886229490493e-07,
+      "loss": 0.328,
+      "step": 5089
+    },
+    {
+      "epoch": 2.3696461824953445,
+      "grad_norm": 0.5922491201366743,
+      "learning_rate": 5.797667392211565e-07,
+      "loss": 0.326,
+      "step": 5090
+    },
+    {
+      "epoch": 2.3701117318435756,
+      "grad_norm": 0.620502217609778,
+      "learning_rate": 5.789453613283666e-07,
+      "loss": 0.323,
+      "step": 5091
+    },
+    {
+      "epoch": 2.370577281191806,
+      "grad_norm": 0.6171753015338307,
+      "learning_rate": 5.781244894870502e-07,
+      "loss": 0.3366,
+      "step": 5092
+    },
+    {
+      "epoch": 2.3710428305400373,
+      "grad_norm": 0.6118956833446069,
+      "learning_rate": 5.77304123913445e-07,
+      "loss": 0.3102,
+      "step": 5093
+    },
+    {
+      "epoch": 2.3715083798882683,
+      "grad_norm": 0.630336540776192,
+      "learning_rate": 5.764842648236551e-07,
+      "loss": 0.3304,
+      "step": 5094
+    },
+    {
+      "epoch": 2.371973929236499,
+      "grad_norm": 0.5882753234900796,
+      "learning_rate": 5.7566491243365e-07,
+      "loss": 0.3241,
+      "step": 5095
+    },
+    {
+      "epoch": 2.37243947858473,
+      "grad_norm": 0.5963711160037427,
+      "learning_rate": 5.748460669592676e-07,
+      "loss": 0.3382,
+      "step": 5096
+    },
+    {
+      "epoch": 2.372905027932961,
+      "grad_norm": 0.6001508349109954,
+      "learning_rate": 5.740277286162099e-07,
+      "loss": 0.3292,
+      "step": 5097
+    },
+    {
+      "epoch": 2.3733705772811917,
+      "grad_norm": 0.6389547866331289,
+      "learning_rate": 5.732098976200476e-07,
+      "loss": 0.3276,
+      "step": 5098
+    },
+    {
+      "epoch": 2.373836126629423,
+      "grad_norm": 0.5937974944885251,
+      "learning_rate": 5.723925741862177e-07,
+      "loss": 0.3363,
+      "step": 5099
+    },
+    {
+      "epoch": 2.3743016759776534,
+      "grad_norm": 0.6148633315848588,
+      "learning_rate": 5.715757585300205e-07,
+      "loss": 0.3374,
+      "step": 5100
+    },
+    {
+      "epoch": 2.3747672253258845,
+      "grad_norm": 0.6045413902336713,
+      "learning_rate": 5.707594508666262e-07,
+      "loss": 0.3111,
+      "step": 5101
+    },
+    {
+      "epoch": 2.3752327746741155,
+      "grad_norm": 0.5986591191504916,
+      "learning_rate": 5.699436514110698e-07,
+      "loss": 0.3215,
+      "step": 5102
+    },
+    {
+      "epoch": 2.3756983240223466,
+      "grad_norm": 0.6245045432915713,
+      "learning_rate": 5.691283603782513e-07,
+      "loss": 0.3392,
+      "step": 5103
+    },
+    {
+      "epoch": 2.376163873370577,
+      "grad_norm": 0.617465328794659,
+      "learning_rate": 5.683135779829385e-07,
+      "loss": 0.3256,
+      "step": 5104
+    },
+    {
+      "epoch": 2.3766294227188083,
+      "grad_norm": 0.6055784319121285,
+      "learning_rate": 5.674993044397636e-07,
+      "loss": 0.3265,
+      "step": 5105
+    },
+    {
+      "epoch": 2.377094972067039,
+      "grad_norm": 0.6251740822006897,
+      "learning_rate": 5.666855399632265e-07,
+      "loss": 0.3318,
+      "step": 5106
+    },
+    {
+      "epoch": 2.37756052141527,
+      "grad_norm": 0.641848702666422,
+      "learning_rate": 5.65872284767692e-07,
+      "loss": 0.3354,
+      "step": 5107
+    },
+    {
+      "epoch": 2.378026070763501,
+      "grad_norm": 0.6197765996991222,
+      "learning_rate": 5.650595390673911e-07,
+      "loss": 0.322,
+      "step": 5108
+    },
+    {
+      "epoch": 2.3784916201117317,
+      "grad_norm": 0.6025355831475842,
+      "learning_rate": 5.642473030764181e-07,
+      "loss": 0.3178,
+      "step": 5109
+    },
+    {
+      "epoch": 2.3789571694599627,
+      "grad_norm": 0.6210456887111159,
+      "learning_rate": 5.634355770087388e-07,
+      "loss": 0.3336,
+      "step": 5110
+    },
+    {
+      "epoch": 2.379422718808194,
+      "grad_norm": 0.5941051802250787,
+      "learning_rate": 5.626243610781792e-07,
+      "loss": 0.3149,
+      "step": 5111
+    },
+    {
+      "epoch": 2.3798882681564244,
+      "grad_norm": 0.5853739256525874,
+      "learning_rate": 5.618136554984325e-07,
+      "loss": 0.3358,
+      "step": 5112
+    },
+    {
+      "epoch": 2.3803538175046555,
+      "grad_norm": 0.6562081313555406,
+      "learning_rate": 5.610034604830583e-07,
+      "loss": 0.3282,
+      "step": 5113
+    },
+    {
+      "epoch": 2.3808193668528865,
+      "grad_norm": 0.567712897992975,
+      "learning_rate": 5.60193776245482e-07,
+      "loss": 0.3176,
+      "step": 5114
+    },
+    {
+      "epoch": 2.381284916201117,
+      "grad_norm": 0.627566293137629,
+      "learning_rate": 5.593846029989924e-07,
+      "loss": 0.3274,
+      "step": 5115
+    },
+    {
+      "epoch": 2.381750465549348,
+      "grad_norm": 0.6135418249125228,
+      "learning_rate": 5.585759409567463e-07,
+      "loss": 0.327,
+      "step": 5116
+    },
+    {
+      "epoch": 2.3822160148975793,
+      "grad_norm": 0.5987610048093915,
+      "learning_rate": 5.57767790331763e-07,
+      "loss": 0.3268,
+      "step": 5117
+    },
+    {
+      "epoch": 2.38268156424581,
+      "grad_norm": 0.6267787820610079,
+      "learning_rate": 5.569601513369297e-07,
+      "loss": 0.3271,
+      "step": 5118
+    },
+    {
+      "epoch": 2.383147113594041,
+      "grad_norm": 0.580988150485509,
+      "learning_rate": 5.561530241849977e-07,
+      "loss": 0.3359,
+      "step": 5119
+    },
+    {
+      "epoch": 2.383612662942272,
+      "grad_norm": 0.6226828925423965,
+      "learning_rate": 5.553464090885829e-07,
+      "loss": 0.3225,
+      "step": 5120
+    },
+    {
+      "epoch": 2.3840782122905027,
+      "grad_norm": 0.5754818293752393,
+      "learning_rate": 5.545403062601671e-07,
+      "loss": 0.331,
+      "step": 5121
+    },
+    {
+      "epoch": 2.3845437616387337,
+      "grad_norm": 0.5904976972273279,
+      "learning_rate": 5.537347159120976e-07,
+      "loss": 0.3225,
+      "step": 5122
+    },
+    {
+      "epoch": 2.385009310986965,
+      "grad_norm": 0.603767103257359,
+      "learning_rate": 5.529296382565849e-07,
+      "loss": 0.3174,
+      "step": 5123
+    },
+    {
+      "epoch": 2.3854748603351954,
+      "grad_norm": 0.6256307845704491,
+      "learning_rate": 5.521250735057063e-07,
+      "loss": 0.334,
+      "step": 5124
+    },
+    {
+      "epoch": 2.3859404096834265,
+      "grad_norm": 0.5641021039575546,
+      "learning_rate": 5.513210218714038e-07,
+      "loss": 0.3162,
+      "step": 5125
+    },
+    {
+      "epoch": 2.3864059590316575,
+      "grad_norm": 0.6221747763454268,
+      "learning_rate": 5.505174835654822e-07,
+      "loss": 0.3382,
+      "step": 5126
+    },
+    {
+      "epoch": 2.386871508379888,
+      "grad_norm": 0.6064738022630484,
+      "learning_rate": 5.497144587996145e-07,
+      "loss": 0.3214,
+      "step": 5127
+    },
+    {
+      "epoch": 2.387337057728119,
+      "grad_norm": 0.5859291780708608,
+      "learning_rate": 5.489119477853349e-07,
+      "loss": 0.3299,
+      "step": 5128
+    },
+    {
+      "epoch": 2.3878026070763503,
+      "grad_norm": 0.5918933999734741,
+      "learning_rate": 5.481099507340442e-07,
+      "loss": 0.3143,
+      "step": 5129
+    },
+    {
+      "epoch": 2.388268156424581,
+      "grad_norm": 0.6328866262824123,
+      "learning_rate": 5.473084678570089e-07,
+      "loss": 0.3279,
+      "step": 5130
+    },
+    {
+      "epoch": 2.388733705772812,
+      "grad_norm": 0.6298611821880976,
+      "learning_rate": 5.465074993653574e-07,
+      "loss": 0.3382,
+      "step": 5131
+    },
+    {
+      "epoch": 2.389199255121043,
+      "grad_norm": 0.5912984922989637,
+      "learning_rate": 5.457070454700833e-07,
+      "loss": 0.3224,
+      "step": 5132
+    },
+    {
+      "epoch": 2.3896648044692737,
+      "grad_norm": 0.6238457168965731,
+      "learning_rate": 5.449071063820458e-07,
+      "loss": 0.315,
+      "step": 5133
+    },
+    {
+      "epoch": 2.3901303538175047,
+      "grad_norm": 0.5852387129943706,
+      "learning_rate": 5.441076823119687e-07,
+      "loss": 0.3175,
+      "step": 5134
+    },
+    {
+      "epoch": 2.3905959031657353,
+      "grad_norm": 0.5921903549632257,
+      "learning_rate": 5.433087734704379e-07,
+      "loss": 0.3101,
+      "step": 5135
+    },
+    {
+      "epoch": 2.3910614525139664,
+      "grad_norm": 0.5863098398442368,
+      "learning_rate": 5.425103800679057e-07,
+      "loss": 0.3167,
+      "step": 5136
+    },
+    {
+      "epoch": 2.3915270018621975,
+      "grad_norm": 0.620295388985107,
+      "learning_rate": 5.417125023146885e-07,
+      "loss": 0.319,
+      "step": 5137
+    },
+    {
+      "epoch": 2.3919925512104285,
+      "grad_norm": 0.5752799429162937,
+      "learning_rate": 5.409151404209654e-07,
+      "loss": 0.3092,
+      "step": 5138
+    },
+    {
+      "epoch": 2.392458100558659,
+      "grad_norm": 0.6013082275685886,
+      "learning_rate": 5.401182945967809e-07,
+      "loss": 0.3188,
+      "step": 5139
+    },
+    {
+      "epoch": 2.39292364990689,
+      "grad_norm": 0.6585918584067769,
+      "learning_rate": 5.393219650520426e-07,
+      "loss": 0.3318,
+      "step": 5140
+    },
+    {
+      "epoch": 2.393389199255121,
+      "grad_norm": 0.5905328034068071,
+      "learning_rate": 5.385261519965226e-07,
+      "loss": 0.3234,
+      "step": 5141
+    },
+    {
+      "epoch": 2.393854748603352,
+      "grad_norm": 0.5905368997783279,
+      "learning_rate": 5.377308556398586e-07,
+      "loss": 0.3139,
+      "step": 5142
+    },
+    {
+      "epoch": 2.394320297951583,
+      "grad_norm": 0.609772570847165,
+      "learning_rate": 5.36936076191549e-07,
+      "loss": 0.3439,
+      "step": 5143
+    },
+    {
+      "epoch": 2.394785847299814,
+      "grad_norm": 0.640041066880716,
+      "learning_rate": 5.36141813860957e-07,
+      "loss": 0.3476,
+      "step": 5144
+    },
+    {
+      "epoch": 2.3952513966480447,
+      "grad_norm": 0.6108995869431376,
+      "learning_rate": 5.353480688573121e-07,
+      "loss": 0.3349,
+      "step": 5145
+    },
+    {
+      "epoch": 2.3957169459962757,
+      "grad_norm": 0.5846506996476478,
+      "learning_rate": 5.34554841389705e-07,
+      "loss": 0.3359,
+      "step": 5146
+    },
+    {
+      "epoch": 2.3961824953445063,
+      "grad_norm": 0.5850682110970031,
+      "learning_rate": 5.337621316670893e-07,
+      "loss": 0.3184,
+      "step": 5147
+    },
+    {
+      "epoch": 2.3966480446927374,
+      "grad_norm": 0.6121910004900978,
+      "learning_rate": 5.329699398982851e-07,
+      "loss": 0.3408,
+      "step": 5148
+    },
+    {
+      "epoch": 2.3971135940409685,
+      "grad_norm": 0.6016136918599329,
+      "learning_rate": 5.321782662919745e-07,
+      "loss": 0.3268,
+      "step": 5149
+    },
+    {
+      "epoch": 2.397579143389199,
+      "grad_norm": 0.6495402362910808,
+      "learning_rate": 5.31387111056702e-07,
+      "loss": 0.3272,
+      "step": 5150
+    },
+    {
+      "epoch": 2.39804469273743,
+      "grad_norm": 0.61029105896247,
+      "learning_rate": 5.305964744008784e-07,
+      "loss": 0.3119,
+      "step": 5151
+    },
+    {
+      "epoch": 2.398510242085661,
+      "grad_norm": 0.6082091512273291,
+      "learning_rate": 5.298063565327744e-07,
+      "loss": 0.3277,
+      "step": 5152
+    },
+    {
+      "epoch": 2.398975791433892,
+      "grad_norm": 0.6450866170055584,
+      "learning_rate": 5.290167576605265e-07,
+      "loss": 0.342,
+      "step": 5153
+    },
+    {
+      "epoch": 2.399441340782123,
+      "grad_norm": 0.6097969929382525,
+      "learning_rate": 5.28227677992135e-07,
+      "loss": 0.3215,
+      "step": 5154
+    },
+    {
+      "epoch": 2.399906890130354,
+      "grad_norm": 0.6166610265751975,
+      "learning_rate": 5.274391177354604e-07,
+      "loss": 0.3294,
+      "step": 5155
+    },
+    {
+      "epoch": 2.4003724394785846,
+      "grad_norm": 0.6124597410687189,
+      "learning_rate": 5.266510770982292e-07,
+      "loss": 0.3142,
+      "step": 5156
+    },
+    {
+      "epoch": 2.4008379888268156,
+      "grad_norm": 0.6121888063509029,
+      "learning_rate": 5.258635562880304e-07,
+      "loss": 0.3229,
+      "step": 5157
+    },
+    {
+      "epoch": 2.4013035381750467,
+      "grad_norm": 0.5909713888665209,
+      "learning_rate": 5.250765555123147e-07,
+      "loss": 0.3377,
+      "step": 5158
+    },
+    {
+      "epoch": 2.4017690875232773,
+      "grad_norm": 0.6168373319583407,
+      "learning_rate": 5.24290074978398e-07,
+      "loss": 0.3259,
+      "step": 5159
+    },
+    {
+      "epoch": 2.4022346368715084,
+      "grad_norm": 0.634659731981674,
+      "learning_rate": 5.235041148934566e-07,
+      "loss": 0.3291,
+      "step": 5160
+    },
+    {
+      "epoch": 2.4027001862197395,
+      "grad_norm": 0.5883507898052013,
+      "learning_rate": 5.227186754645317e-07,
+      "loss": 0.3209,
+      "step": 5161
+    },
+    {
+      "epoch": 2.40316573556797,
+      "grad_norm": 0.6097665527856083,
+      "learning_rate": 5.219337568985275e-07,
+      "loss": 0.3233,
+      "step": 5162
+    },
+    {
+      "epoch": 2.403631284916201,
+      "grad_norm": 0.6021810110060514,
+      "learning_rate": 5.211493594022097e-07,
+      "loss": 0.3164,
+      "step": 5163
+    },
+    {
+      "epoch": 2.404096834264432,
+      "grad_norm": 0.6133837423564041,
+      "learning_rate": 5.203654831822055e-07,
+      "loss": 0.3233,
+      "step": 5164
+    },
+    {
+      "epoch": 2.404562383612663,
+      "grad_norm": 0.579695272435037,
+      "learning_rate": 5.195821284450098e-07,
+      "loss": 0.334,
+      "step": 5165
+    },
+    {
+      "epoch": 2.405027932960894,
+      "grad_norm": 0.6037828410006971,
+      "learning_rate": 5.18799295396975e-07,
+      "loss": 0.3293,
+      "step": 5166
+    },
+    {
+      "epoch": 2.405493482309125,
+      "grad_norm": 0.6097841557626666,
+      "learning_rate": 5.180169842443166e-07,
+      "loss": 0.3272,
+      "step": 5167
+    },
+    {
+      "epoch": 2.4059590316573556,
+      "grad_norm": 0.6645323654054468,
+      "learning_rate": 5.172351951931173e-07,
+      "loss": 0.3315,
+      "step": 5168
+    },
+    {
+      "epoch": 2.4064245810055866,
+      "grad_norm": 0.6225301951476,
+      "learning_rate": 5.164539284493167e-07,
+      "loss": 0.3267,
+      "step": 5169
+    },
+    {
+      "epoch": 2.4068901303538173,
+      "grad_norm": 0.6254710225106216,
+      "learning_rate": 5.156731842187191e-07,
+      "loss": 0.3078,
+      "step": 5170
+    },
+    {
+      "epoch": 2.4073556797020483,
+      "grad_norm": 0.6291917898274045,
+      "learning_rate": 5.14892962706991e-07,
+      "loss": 0.3606,
+      "step": 5171
+    },
+    {
+      "epoch": 2.4078212290502794,
+      "grad_norm": 0.5849147230535825,
+      "learning_rate": 5.141132641196628e-07,
+      "loss": 0.3148,
+      "step": 5172
+    },
+    {
+      "epoch": 2.4082867783985105,
+      "grad_norm": 0.6086428115927729,
+      "learning_rate": 5.133340886621238e-07,
+      "loss": 0.3074,
+      "step": 5173
+    },
+    {
+      "epoch": 2.408752327746741,
+      "grad_norm": 0.6027153544072473,
+      "learning_rate": 5.125554365396287e-07,
+      "loss": 0.3338,
+      "step": 5174
+    },
+    {
+      "epoch": 2.409217877094972,
+      "grad_norm": 0.6377809406256787,
+      "learning_rate": 5.117773079572918e-07,
+      "loss": 0.3227,
+      "step": 5175
+    },
+    {
+      "epoch": 2.4096834264432028,
+      "grad_norm": 0.6250241466250996,
+      "learning_rate": 5.109997031200916e-07,
+      "loss": 0.3337,
+      "step": 5176
+    },
+    {
+      "epoch": 2.410148975791434,
+      "grad_norm": 0.6042442483123307,
+      "learning_rate": 5.10222622232868e-07,
+      "loss": 0.3163,
+      "step": 5177
+    },
+    {
+      "epoch": 2.410614525139665,
+      "grad_norm": 0.6216313461852793,
+      "learning_rate": 5.094460655003212e-07,
+      "loss": 0.3478,
+      "step": 5178
+    },
+    {
+      "epoch": 2.411080074487896,
+      "grad_norm": 0.6591416495207998,
+      "learning_rate": 5.086700331270153e-07,
+      "loss": 0.3427,
+      "step": 5179
+    },
+    {
+      "epoch": 2.4115456238361266,
+      "grad_norm": 0.5846738067749871,
+      "learning_rate": 5.078945253173767e-07,
+      "loss": 0.3239,
+      "step": 5180
+    },
+    {
+      "epoch": 2.4120111731843576,
+      "grad_norm": 0.6119057555296664,
+      "learning_rate": 5.07119542275691e-07,
+      "loss": 0.326,
+      "step": 5181
+    },
+    {
+      "epoch": 2.4124767225325883,
+      "grad_norm": 0.6031567777599682,
+      "learning_rate": 5.063450842061088e-07,
+      "loss": 0.3212,
+      "step": 5182
+    },
+    {
+      "epoch": 2.4129422718808193,
+      "grad_norm": 0.5874804674220119,
+      "learning_rate": 5.055711513126394e-07,
+      "loss": 0.3271,
+      "step": 5183
+    },
+    {
+      "epoch": 2.4134078212290504,
+      "grad_norm": 0.6112373886521549,
+      "learning_rate": 5.047977437991552e-07,
+      "loss": 0.3252,
+      "step": 5184
+    },
+    {
+      "epoch": 2.413873370577281,
+      "grad_norm": 0.5865688183731336,
+      "learning_rate": 5.040248618693916e-07,
+      "loss": 0.3331,
+      "step": 5185
+    },
+    {
+      "epoch": 2.414338919925512,
+      "grad_norm": 0.5990067057073539,
+      "learning_rate": 5.032525057269427e-07,
+      "loss": 0.317,
+      "step": 5186
+    },
+    {
+      "epoch": 2.414804469273743,
+      "grad_norm": 0.6066361889605406,
+      "learning_rate": 5.024806755752648e-07,
+      "loss": 0.3214,
+      "step": 5187
+    },
+    {
+      "epoch": 2.4152700186219738,
+      "grad_norm": 0.6034462545323916,
+      "learning_rate": 5.017093716176786e-07,
+      "loss": 0.3337,
+      "step": 5188
+    },
+    {
+      "epoch": 2.415735567970205,
+      "grad_norm": 0.6891838920781899,
+      "learning_rate": 5.009385940573627e-07,
+      "loss": 0.3359,
+      "step": 5189
+    },
+    {
+      "epoch": 2.416201117318436,
+      "grad_norm": 0.6091496802942422,
+      "learning_rate": 5.001683430973573e-07,
+      "loss": 0.3358,
+      "step": 5190
+    },
+    {
+      "epoch": 2.4166666666666665,
+      "grad_norm": 0.6316645050307822,
+      "learning_rate": 4.993986189405658e-07,
+      "loss": 0.3425,
+      "step": 5191
+    },
+    {
+      "epoch": 2.4171322160148976,
+      "grad_norm": 0.6032567748330294,
+      "learning_rate": 4.986294217897525e-07,
+      "loss": 0.3273,
+      "step": 5192
+    },
+    {
+      "epoch": 2.4175977653631286,
+      "grad_norm": 0.5943184023440156,
+      "learning_rate": 4.978607518475406e-07,
+      "loss": 0.3335,
+      "step": 5193
+    },
+    {
+      "epoch": 2.4180633147113593,
+      "grad_norm": 0.5941819791908372,
+      "learning_rate": 4.97092609316418e-07,
+      "loss": 0.3326,
+      "step": 5194
+    },
+    {
+      "epoch": 2.4185288640595903,
+      "grad_norm": 0.5889475987012104,
+      "learning_rate": 4.963249943987295e-07,
+      "loss": 0.3146,
+      "step": 5195
+    },
+    {
+      "epoch": 2.4189944134078214,
+      "grad_norm": 0.6058586371573653,
+      "learning_rate": 4.955579072966846e-07,
+      "loss": 0.3159,
+      "step": 5196
+    },
+    {
+      "epoch": 2.419459962756052,
+      "grad_norm": 0.5665767763034407,
+      "learning_rate": 4.947913482123523e-07,
+      "loss": 0.3341,
+      "step": 5197
+    },
+    {
+      "epoch": 2.419925512104283,
+      "grad_norm": 0.591757067775042,
+      "learning_rate": 4.940253173476614e-07,
+      "loss": 0.3204,
+      "step": 5198
+    },
+    {
+      "epoch": 2.420391061452514,
+      "grad_norm": 0.6118700977244542,
+      "learning_rate": 4.932598149044038e-07,
+      "loss": 0.3248,
+      "step": 5199
+    },
+    {
+      "epoch": 2.4208566108007448,
+      "grad_norm": 0.6058923692143664,
+      "learning_rate": 4.924948410842309e-07,
+      "loss": 0.3276,
+      "step": 5200
+    },
+    {
+      "epoch": 2.421322160148976,
+      "grad_norm": 0.5900490921098117,
+      "learning_rate": 4.917303960886541e-07,
+      "loss": 0.3264,
+      "step": 5201
+    },
+    {
+      "epoch": 2.421787709497207,
+      "grad_norm": 0.6207156252282513,
+      "learning_rate": 4.909664801190478e-07,
+      "loss": 0.3263,
+      "step": 5202
+    },
+    {
+      "epoch": 2.4222532588454375,
+      "grad_norm": 0.5825706172571893,
+      "learning_rate": 4.902030933766439e-07,
+      "loss": 0.3209,
+      "step": 5203
+    },
+    {
+      "epoch": 2.4227188081936686,
+      "grad_norm": 0.6007925847453637,
+      "learning_rate": 4.894402360625378e-07,
+      "loss": 0.3335,
+      "step": 5204
+    },
+    {
+      "epoch": 2.423184357541899,
+      "grad_norm": 0.5790223772457959,
+      "learning_rate": 4.886779083776847e-07,
+      "loss": 0.3222,
+      "step": 5205
+    },
+    {
+      "epoch": 2.4236499068901303,
+      "grad_norm": 0.6381412826328753,
+      "learning_rate": 4.879161105228989e-07,
+      "loss": 0.3305,
+      "step": 5206
+    },
+    {
+      "epoch": 2.4241154562383613,
+      "grad_norm": 0.5875806021772944,
+      "learning_rate": 4.871548426988551e-07,
+      "loss": 0.3416,
+      "step": 5207
+    },
+    {
+      "epoch": 2.4245810055865924,
+      "grad_norm": 0.5959673893192334,
+      "learning_rate": 4.86394105106092e-07,
+      "loss": 0.3219,
+      "step": 5208
+    },
+    {
+      "epoch": 2.425046554934823,
+      "grad_norm": 0.6176036517140638,
+      "learning_rate": 4.856338979450043e-07,
+      "loss": 0.3251,
+      "step": 5209
+    },
+    {
+      "epoch": 2.425512104283054,
+      "grad_norm": 0.5767329838832501,
+      "learning_rate": 4.848742214158484e-07,
+      "loss": 0.3321,
+      "step": 5210
+    },
+    {
+      "epoch": 2.4259776536312847,
+      "grad_norm": 0.6565632576349822,
+      "learning_rate": 4.841150757187415e-07,
+      "loss": 0.3151,
+      "step": 5211
+    },
+    {
+      "epoch": 2.4264432029795158,
+      "grad_norm": 0.6080920839201038,
+      "learning_rate": 4.833564610536615e-07,
+      "loss": 0.3178,
+      "step": 5212
+    },
+    {
+      "epoch": 2.426908752327747,
+      "grad_norm": 0.6244262404971238,
+      "learning_rate": 4.825983776204443e-07,
+      "loss": 0.3076,
+      "step": 5213
+    },
+    {
+      "epoch": 2.427374301675978,
+      "grad_norm": 0.5923380580523528,
+      "learning_rate": 4.818408256187873e-07,
+      "loss": 0.3206,
+      "step": 5214
+    },
+    {
+      "epoch": 2.4278398510242085,
+      "grad_norm": 0.6355894307946355,
+      "learning_rate": 4.810838052482489e-07,
+      "loss": 0.3304,
+      "step": 5215
+    },
+    {
+      "epoch": 2.4283054003724396,
+      "grad_norm": 0.6127642297297109,
+      "learning_rate": 4.803273167082448e-07,
+      "loss": 0.3275,
+      "step": 5216
+    },
+    {
+      "epoch": 2.42877094972067,
+      "grad_norm": 0.6245232775787449,
+      "learning_rate": 4.795713601980531e-07,
+      "loss": 0.3337,
+      "step": 5217
+    },
+    {
+      "epoch": 2.4292364990689013,
+      "grad_norm": 0.601383270342634,
+      "learning_rate": 4.788159359168099e-07,
+      "loss": 0.3034,
+      "step": 5218
+    },
+    {
+      "epoch": 2.4297020484171323,
+      "grad_norm": 0.5868483912455316,
+      "learning_rate": 4.780610440635122e-07,
+      "loss": 0.331,
+      "step": 5219
+    },
+    {
+      "epoch": 2.430167597765363,
+      "grad_norm": 0.6209885590204076,
+      "learning_rate": 4.773066848370173e-07,
+      "loss": 0.3197,
+      "step": 5220
+    },
+    {
+      "epoch": 2.430633147113594,
+      "grad_norm": 0.6326635402482824,
+      "learning_rate": 4.7655285843604044e-07,
+      "loss": 0.3208,
+      "step": 5221
+    },
+    {
+      "epoch": 2.431098696461825,
+      "grad_norm": 0.6181811293144004,
+      "learning_rate": 4.757995650591563e-07,
+      "loss": 0.3247,
+      "step": 5222
+    },
+    {
+      "epoch": 2.4315642458100557,
+      "grad_norm": 0.60558924519674,
+      "learning_rate": 4.7504680490480293e-07,
+      "loss": 0.3314,
+      "step": 5223
+    },
+    {
+      "epoch": 2.4320297951582868,
+      "grad_norm": 0.6283351954395042,
+      "learning_rate": 4.7429457817127406e-07,
+      "loss": 0.3193,
+      "step": 5224
+    },
+    {
+      "epoch": 2.432495344506518,
+      "grad_norm": 0.6089107814612084,
+      "learning_rate": 4.7354288505672307e-07,
+      "loss": 0.3282,
+      "step": 5225
+    },
+    {
+      "epoch": 2.4329608938547485,
+      "grad_norm": 0.599335372575662,
+      "learning_rate": 4.727917257591649e-07,
+      "loss": 0.321,
+      "step": 5226
+    },
+    {
+      "epoch": 2.4334264432029795,
+      "grad_norm": 0.6162344908857087,
+      "learning_rate": 4.72041100476473e-07,
+      "loss": 0.3413,
+      "step": 5227
+    },
+    {
+      "epoch": 2.4338919925512106,
+      "grad_norm": 0.6178214702206408,
+      "learning_rate": 4.7129100940637887e-07,
+      "loss": 0.3077,
+      "step": 5228
+    },
+    {
+      "epoch": 2.434357541899441,
+      "grad_norm": 0.6315086552853564,
+      "learning_rate": 4.705414527464758e-07,
+      "loss": 0.3385,
+      "step": 5229
+    },
+    {
+      "epoch": 2.4348230912476723,
+      "grad_norm": 0.5928176413398647,
+      "learning_rate": 4.697924306942131e-07,
+      "loss": 0.3145,
+      "step": 5230
+    },
+    {
+      "epoch": 2.4352886405959033,
+      "grad_norm": 0.6228448208314725,
+      "learning_rate": 4.6904394344690187e-07,
+      "loss": 0.3248,
+      "step": 5231
+    },
+    {
+      "epoch": 2.435754189944134,
+      "grad_norm": 0.5881801478719825,
+      "learning_rate": 4.682959912017121e-07,
+      "loss": 0.3219,
+      "step": 5232
+    },
+    {
+      "epoch": 2.436219739292365,
+      "grad_norm": 0.6139394026008628,
+      "learning_rate": 4.6754857415567067e-07,
+      "loss": 0.3127,
+      "step": 5233
+    },
+    {
+      "epoch": 2.436685288640596,
+      "grad_norm": 0.6176821186207754,
+      "learning_rate": 4.668016925056662e-07,
+      "loss": 0.3191,
+      "step": 5234
+    },
+    {
+      "epoch": 2.4371508379888267,
+      "grad_norm": 0.6210428287417566,
+      "learning_rate": 4.6605534644844517e-07,
+      "loss": 0.3356,
+      "step": 5235
+    },
+    {
+      "epoch": 2.4376163873370578,
+      "grad_norm": 0.60529513597326,
+      "learning_rate": 4.6530953618061203e-07,
+      "loss": 0.3309,
+      "step": 5236
+    },
+    {
+      "epoch": 2.438081936685289,
+      "grad_norm": 0.602125629116997,
+      "learning_rate": 4.645642618986318e-07,
+      "loss": 0.3319,
+      "step": 5237
+    },
+    {
+      "epoch": 2.4385474860335195,
+      "grad_norm": 0.631490485677985,
+      "learning_rate": 4.638195237988266e-07,
+      "loss": 0.3209,
+      "step": 5238
+    },
+    {
+      "epoch": 2.4390130353817505,
+      "grad_norm": 0.6164988792549896,
+      "learning_rate": 4.6307532207737843e-07,
+      "loss": 0.3352,
+      "step": 5239
+    },
+    {
+      "epoch": 2.439478584729981,
+      "grad_norm": 0.5947383922365892,
+      "learning_rate": 4.623316569303285e-07,
+      "loss": 0.3353,
+      "step": 5240
+    },
+    {
+      "epoch": 2.439944134078212,
+      "grad_norm": 0.6183010218280212,
+      "learning_rate": 4.615885285535751e-07,
+      "loss": 0.3303,
+      "step": 5241
+    },
+    {
+      "epoch": 2.4404096834264433,
+      "grad_norm": 0.6263133228837184,
+      "learning_rate": 4.60845937142875e-07,
+      "loss": 0.3312,
+      "step": 5242
+    },
+    {
+      "epoch": 2.4408752327746743,
+      "grad_norm": 0.6127435631529181,
+      "learning_rate": 4.601038828938467e-07,
+      "loss": 0.3416,
+      "step": 5243
+    },
+    {
+      "epoch": 2.441340782122905,
+      "grad_norm": 0.6238715954922375,
+      "learning_rate": 4.5936236600196386e-07,
+      "loss": 0.3344,
+      "step": 5244
+    },
+    {
+      "epoch": 2.441806331471136,
+      "grad_norm": 0.6269946431012574,
+      "learning_rate": 4.5862138666255855e-07,
+      "loss": 0.3278,
+      "step": 5245
+    },
+    {
+      "epoch": 2.4422718808193666,
+      "grad_norm": 0.5822685925963564,
+      "learning_rate": 4.578809450708238e-07,
+      "loss": 0.3238,
+      "step": 5246
+    },
+    {
+      "epoch": 2.4427374301675977,
+      "grad_norm": 0.5977597935276295,
+      "learning_rate": 4.5714104142180943e-07,
+      "loss": 0.3274,
+      "step": 5247
+    },
+    {
+      "epoch": 2.4432029795158288,
+      "grad_norm": 0.6473429781210678,
+      "learning_rate": 4.5640167591042285e-07,
+      "loss": 0.3355,
+      "step": 5248
+    },
+    {
+      "epoch": 2.44366852886406,
+      "grad_norm": 0.6332135092475308,
+      "learning_rate": 4.556628487314313e-07,
+      "loss": 0.3409,
+      "step": 5249
+    },
+    {
+      "epoch": 2.4441340782122905,
+      "grad_norm": 0.6278788294617661,
+      "learning_rate": 4.5492456007945854e-07,
+      "loss": 0.3298,
+      "step": 5250
+    },
+    {
+      "epoch": 2.4445996275605215,
+      "grad_norm": 0.6733095977012497,
+      "learning_rate": 4.5418681014898827e-07,
+      "loss": 0.3197,
+      "step": 5251
+    },
+    {
+      "epoch": 2.445065176908752,
+      "grad_norm": 0.6192507769155967,
+      "learning_rate": 4.5344959913436124e-07,
+      "loss": 0.3221,
+      "step": 5252
+    },
+    {
+      "epoch": 2.445530726256983,
+      "grad_norm": 0.6053452734493137,
+      "learning_rate": 4.527129272297756e-07,
+      "loss": 0.3197,
+      "step": 5253
+    },
+    {
+      "epoch": 2.4459962756052143,
+      "grad_norm": 0.65983447011435,
+      "learning_rate": 4.5197679462928885e-07,
+      "loss": 0.3304,
+      "step": 5254
+    },
+    {
+      "epoch": 2.446461824953445,
+      "grad_norm": 0.6150416838829934,
+      "learning_rate": 4.512412015268164e-07,
+      "loss": 0.3242,
+      "step": 5255
+    },
+    {
+      "epoch": 2.446927374301676,
+      "grad_norm": 0.5805306503670259,
+      "learning_rate": 4.505061481161296e-07,
+      "loss": 0.3291,
+      "step": 5256
+    },
+    {
+      "epoch": 2.447392923649907,
+      "grad_norm": 0.640693634123379,
+      "learning_rate": 4.4977163459085983e-07,
+      "loss": 0.3309,
+      "step": 5257
+    },
+    {
+      "epoch": 2.4478584729981376,
+      "grad_norm": 0.6116352605732303,
+      "learning_rate": 4.4903766114449576e-07,
+      "loss": 0.3441,
+      "step": 5258
+    },
+    {
+      "epoch": 2.4483240223463687,
+      "grad_norm": 0.5850248812424541,
+      "learning_rate": 4.4830422797038227e-07,
+      "loss": 0.3123,
+      "step": 5259
+    },
+    {
+      "epoch": 2.4487895716945998,
+      "grad_norm": 0.6283772361228458,
+      "learning_rate": 4.4757133526172477e-07,
+      "loss": 0.3291,
+      "step": 5260
+    },
+    {
+      "epoch": 2.4492551210428304,
+      "grad_norm": 0.6109342322563184,
+      "learning_rate": 4.468389832115827e-07,
+      "loss": 0.3074,
+      "step": 5261
+    },
+    {
+      "epoch": 2.4497206703910615,
+      "grad_norm": 0.6114971593877635,
+      "learning_rate": 4.4610717201287625e-07,
+      "loss": 0.3433,
+      "step": 5262
+    },
+    {
+      "epoch": 2.4501862197392925,
+      "grad_norm": 0.6246621112235308,
+      "learning_rate": 4.453759018583823e-07,
+      "loss": 0.3421,
+      "step": 5263
+    },
+    {
+      "epoch": 2.450651769087523,
+      "grad_norm": 0.5930930260573664,
+      "learning_rate": 4.446451729407339e-07,
+      "loss": 0.332,
+      "step": 5264
+    },
+    {
+      "epoch": 2.451117318435754,
+      "grad_norm": 0.5816026744212344,
+      "learning_rate": 4.4391498545242135e-07,
+      "loss": 0.3329,
+      "step": 5265
+    },
+    {
+      "epoch": 2.4515828677839853,
+      "grad_norm": 0.6194727583547796,
+      "learning_rate": 4.431853395857963e-07,
+      "loss": 0.3192,
+      "step": 5266
+    },
+    {
+      "epoch": 2.452048417132216,
+      "grad_norm": 0.626051109925295,
+      "learning_rate": 4.424562355330628e-07,
+      "loss": 0.3264,
+      "step": 5267
+    },
+    {
+      "epoch": 2.452513966480447,
+      "grad_norm": 0.60986454020775,
+      "learning_rate": 4.4172767348628437e-07,
+      "loss": 0.3169,
+      "step": 5268
+    },
+    {
+      "epoch": 2.452979515828678,
+      "grad_norm": 0.6013891451247491,
+      "learning_rate": 4.4099965363738144e-07,
+      "loss": 0.3286,
+      "step": 5269
+    },
+    {
+      "epoch": 2.4534450651769086,
+      "grad_norm": 0.6245664786135338,
+      "learning_rate": 4.4027217617813285e-07,
+      "loss": 0.3347,
+      "step": 5270
+    },
+    {
+      "epoch": 2.4539106145251397,
+      "grad_norm": 0.6202114810341978,
+      "learning_rate": 4.3954524130017215e-07,
+      "loss": 0.3417,
+      "step": 5271
+    },
+    {
+      "epoch": 2.4543761638733708,
+      "grad_norm": 0.5930642111862863,
+      "learning_rate": 4.3881884919499266e-07,
+      "loss": 0.3286,
+      "step": 5272
+    },
+    {
+      "epoch": 2.4548417132216014,
+      "grad_norm": 0.6133380092692879,
+      "learning_rate": 4.380930000539418e-07,
+      "loss": 0.3356,
+      "step": 5273
+    },
+    {
+      "epoch": 2.4553072625698324,
+      "grad_norm": 0.6057641751367557,
+      "learning_rate": 4.373676940682262e-07,
+      "loss": 0.3171,
+      "step": 5274
+    },
+    {
+      "epoch": 2.455772811918063,
+      "grad_norm": 0.5993382544605126,
+      "learning_rate": 4.3664293142890944e-07,
+      "loss": 0.3127,
+      "step": 5275
+    },
+    {
+      "epoch": 2.456238361266294,
+      "grad_norm": 0.6000210940553049,
+      "learning_rate": 4.3591871232690987e-07,
+      "loss": 0.3131,
+      "step": 5276
+    },
+    {
+      "epoch": 2.456703910614525,
+      "grad_norm": 0.630406994379559,
+      "learning_rate": 4.3519503695300447e-07,
+      "loss": 0.3151,
+      "step": 5277
+    },
+    {
+      "epoch": 2.4571694599627563,
+      "grad_norm": 0.6136818738247198,
+      "learning_rate": 4.3447190549782776e-07,
+      "loss": 0.3343,
+      "step": 5278
+    },
+    {
+      "epoch": 2.457635009310987,
+      "grad_norm": 0.556393570741047,
+      "learning_rate": 4.3374931815186786e-07,
+      "loss": 0.3097,
+      "step": 5279
+    },
+    {
+      "epoch": 2.458100558659218,
+      "grad_norm": 0.6159436544909277,
+      "learning_rate": 4.330272751054729e-07,
+      "loss": 0.3258,
+      "step": 5280
+    },
+    {
+      "epoch": 2.4585661080074486,
+      "grad_norm": 0.5741817121448937,
+      "learning_rate": 4.3230577654884534e-07,
+      "loss": 0.3132,
+      "step": 5281
+    },
+    {
+      "epoch": 2.4590316573556796,
+      "grad_norm": 0.6090637792011405,
+      "learning_rate": 4.315848226720451e-07,
+      "loss": 0.3453,
+      "step": 5282
+    },
+    {
+      "epoch": 2.4594972067039107,
+      "grad_norm": 0.6059613979319272,
+      "learning_rate": 4.308644136649895e-07,
+      "loss": 0.3308,
+      "step": 5283
+    },
+    {
+      "epoch": 2.4599627560521418,
+      "grad_norm": 0.5922904575118711,
+      "learning_rate": 4.301445497174508e-07,
+      "loss": 0.3206,
+      "step": 5284
+    },
+    {
+      "epoch": 2.4604283054003724,
+      "grad_norm": 0.5999336119065184,
+      "learning_rate": 4.2942523101905693e-07,
+      "loss": 0.3267,
+      "step": 5285
+    },
+    {
+      "epoch": 2.4608938547486034,
+      "grad_norm": 0.6131631287678361,
+      "learning_rate": 4.2870645775929615e-07,
+      "loss": 0.3227,
+      "step": 5286
+    },
+    {
+      "epoch": 2.461359404096834,
+      "grad_norm": 0.6521721389350505,
+      "learning_rate": 4.27988230127509e-07,
+      "loss": 0.3482,
+      "step": 5287
+    },
+    {
+      "epoch": 2.461824953445065,
+      "grad_norm": 0.6042697745030016,
+      "learning_rate": 4.2727054831289335e-07,
+      "loss": 0.3173,
+      "step": 5288
+    },
+    {
+      "epoch": 2.462290502793296,
+      "grad_norm": 0.6070901119608295,
+      "learning_rate": 4.26553412504504e-07,
+      "loss": 0.3298,
+      "step": 5289
+    },
+    {
+      "epoch": 2.462756052141527,
+      "grad_norm": 0.6351745770364868,
+      "learning_rate": 4.258368228912524e-07,
+      "loss": 0.3157,
+      "step": 5290
+    },
+    {
+      "epoch": 2.463221601489758,
+      "grad_norm": 0.6354383455090864,
+      "learning_rate": 4.2512077966190416e-07,
+      "loss": 0.3516,
+      "step": 5291
+    },
+    {
+      "epoch": 2.463687150837989,
+      "grad_norm": 0.5933464029150591,
+      "learning_rate": 4.2440528300508295e-07,
+      "loss": 0.3258,
+      "step": 5292
+    },
+    {
+      "epoch": 2.4641527001862196,
+      "grad_norm": 0.6157689400555938,
+      "learning_rate": 4.236903331092665e-07,
+      "loss": 0.3329,
+      "step": 5293
+    },
+    {
+      "epoch": 2.4646182495344506,
+      "grad_norm": 0.5769114872305043,
+      "learning_rate": 4.229759301627903e-07,
+      "loss": 0.3259,
+      "step": 5294
+    },
+    {
+      "epoch": 2.4650837988826817,
+      "grad_norm": 0.6059987053511899,
+      "learning_rate": 4.222620743538458e-07,
+      "loss": 0.3074,
+      "step": 5295
+    },
+    {
+      "epoch": 2.4655493482309123,
+      "grad_norm": 0.5701728675085149,
+      "learning_rate": 4.2154876587047896e-07,
+      "loss": 0.3136,
+      "step": 5296
+    },
+    {
+      "epoch": 2.4660148975791434,
+      "grad_norm": 0.612489564779676,
+      "learning_rate": 4.208360049005908e-07,
+      "loss": 0.3199,
+      "step": 5297
+    },
+    {
+      "epoch": 2.4664804469273744,
+      "grad_norm": 0.5910737488775027,
+      "learning_rate": 4.20123791631942e-07,
+      "loss": 0.3103,
+      "step": 5298
+    },
+    {
+      "epoch": 2.466945996275605,
+      "grad_norm": 0.63430372402464,
+      "learning_rate": 4.194121262521453e-07,
+      "loss": 0.3317,
+      "step": 5299
+    },
+    {
+      "epoch": 2.467411545623836,
+      "grad_norm": 0.5806395688572442,
+      "learning_rate": 4.18701008948669e-07,
+      "loss": 0.3388,
+      "step": 5300
+    },
+    {
+      "epoch": 2.467877094972067,
+      "grad_norm": 0.6043994056774239,
+      "learning_rate": 4.179904399088411e-07,
+      "loss": 0.3256,
+      "step": 5301
+    },
+    {
+      "epoch": 2.468342644320298,
+      "grad_norm": 0.5829708528096695,
+      "learning_rate": 4.172804193198407e-07,
+      "loss": 0.3207,
+      "step": 5302
+    },
+    {
+      "epoch": 2.468808193668529,
+      "grad_norm": 0.6240212932816056,
+      "learning_rate": 4.165709473687035e-07,
+      "loss": 0.3401,
+      "step": 5303
+    },
+    {
+      "epoch": 2.46927374301676,
+      "grad_norm": 0.5769954377836011,
+      "learning_rate": 4.1586202424232226e-07,
+      "loss": 0.3048,
+      "step": 5304
+    },
+    {
+      "epoch": 2.4697392923649906,
+      "grad_norm": 0.6362728214451562,
+      "learning_rate": 4.151536501274445e-07,
+      "loss": 0.3304,
+      "step": 5305
+    },
+    {
+      "epoch": 2.4702048417132216,
+      "grad_norm": 0.5889091942041778,
+      "learning_rate": 4.1444582521067145e-07,
+      "loss": 0.3123,
+      "step": 5306
+    },
+    {
+      "epoch": 2.4706703910614527,
+      "grad_norm": 0.5837376782026872,
+      "learning_rate": 4.1373854967846294e-07,
+      "loss": 0.3064,
+      "step": 5307
+    },
+    {
+      "epoch": 2.4711359404096833,
+      "grad_norm": 0.6007638624502558,
+      "learning_rate": 4.130318237171302e-07,
+      "loss": 0.3158,
+      "step": 5308
+    },
+    {
+      "epoch": 2.4716014897579144,
+      "grad_norm": 0.5953509983900916,
+      "learning_rate": 4.123256475128426e-07,
+      "loss": 0.3274,
+      "step": 5309
+    },
+    {
+      "epoch": 2.472067039106145,
+      "grad_norm": 0.5719755605356158,
+      "learning_rate": 4.1162002125162416e-07,
+      "loss": 0.3277,
+      "step": 5310
+    },
+    {
+      "epoch": 2.472532588454376,
+      "grad_norm": 0.6184935171804771,
+      "learning_rate": 4.1091494511935283e-07,
+      "loss": 0.3404,
+      "step": 5311
+    },
+    {
+      "epoch": 2.472998137802607,
+      "grad_norm": 0.5918186438548884,
+      "learning_rate": 4.1021041930176253e-07,
+      "loss": 0.3113,
+      "step": 5312
+    },
+    {
+      "epoch": 2.473463687150838,
+      "grad_norm": 0.6510641731797392,
+      "learning_rate": 4.095064439844429e-07,
+      "loss": 0.3372,
+      "step": 5313
+    },
+    {
+      "epoch": 2.473929236499069,
+      "grad_norm": 0.5825947221019623,
+      "learning_rate": 4.08803019352837e-07,
+      "loss": 0.3096,
+      "step": 5314
+    },
+    {
+      "epoch": 2.4743947858473,
+      "grad_norm": 0.6023538734115641,
+      "learning_rate": 4.0810014559224396e-07,
+      "loss": 0.3249,
+      "step": 5315
+    },
+    {
+      "epoch": 2.4748603351955305,
+      "grad_norm": 0.6460980508693118,
+      "learning_rate": 4.0739782288781673e-07,
+      "loss": 0.3438,
+      "step": 5316
+    },
+    {
+      "epoch": 2.4753258845437616,
+      "grad_norm": 0.6103082346950591,
+      "learning_rate": 4.0669605142456465e-07,
+      "loss": 0.3421,
+      "step": 5317
+    },
+    {
+      "epoch": 2.4757914338919926,
+      "grad_norm": 0.6074771383054616,
+      "learning_rate": 4.05994831387351e-07,
+      "loss": 0.3309,
+      "step": 5318
+    },
+    {
+      "epoch": 2.4762569832402237,
+      "grad_norm": 0.5744437153839957,
+      "learning_rate": 4.0529416296089393e-07,
+      "loss": 0.3165,
+      "step": 5319
+    },
+    {
+      "epoch": 2.4767225325884543,
+      "grad_norm": 0.6123492038427699,
+      "learning_rate": 4.045940463297643e-07,
+      "loss": 0.3169,
+      "step": 5320
+    },
+    {
+      "epoch": 2.4771880819366854,
+      "grad_norm": 0.6038991214301495,
+      "learning_rate": 4.038944816783924e-07,
+      "loss": 0.3346,
+      "step": 5321
+    },
+    {
+      "epoch": 2.477653631284916,
+      "grad_norm": 0.5940426217264144,
+      "learning_rate": 4.0319546919105864e-07,
+      "loss": 0.3288,
+      "step": 5322
+    },
+    {
+      "epoch": 2.478119180633147,
+      "grad_norm": 0.6007225452915881,
+      "learning_rate": 4.02497009051899e-07,
+      "loss": 0.3279,
+      "step": 5323
+    },
+    {
+      "epoch": 2.478584729981378,
+      "grad_norm": 0.612372021128869,
+      "learning_rate": 4.017991014449052e-07,
+      "loss": 0.3155,
+      "step": 5324
+    },
+    {
+      "epoch": 2.4790502793296088,
+      "grad_norm": 0.6078518432980216,
+      "learning_rate": 4.011017465539233e-07,
+      "loss": 0.3463,
+      "step": 5325
+    },
+    {
+      "epoch": 2.47951582867784,
+      "grad_norm": 0.6005584087322542,
+      "learning_rate": 4.004049445626518e-07,
+      "loss": 0.3212,
+      "step": 5326
+    },
+    {
+      "epoch": 2.479981378026071,
+      "grad_norm": 0.5812831274472771,
+      "learning_rate": 3.9970869565464614e-07,
+      "loss": 0.3264,
+      "step": 5327
+    },
+    {
+      "epoch": 2.4804469273743015,
+      "grad_norm": 0.6182892160020635,
+      "learning_rate": 3.990130000133138e-07,
+      "loss": 0.3171,
+      "step": 5328
+    },
+    {
+      "epoch": 2.4809124767225326,
+      "grad_norm": 0.6287432042275661,
+      "learning_rate": 3.9831785782191825e-07,
+      "loss": 0.3377,
+      "step": 5329
+    },
+    {
+      "epoch": 2.4813780260707636,
+      "grad_norm": 0.6175837522116955,
+      "learning_rate": 3.9762326926357673e-07,
+      "loss": 0.3183,
+      "step": 5330
+    },
+    {
+      "epoch": 2.4818435754189943,
+      "grad_norm": 0.6216527212451282,
+      "learning_rate": 3.969292345212591e-07,
+      "loss": 0.3492,
+      "step": 5331
+    },
+    {
+      "epoch": 2.4823091247672253,
+      "grad_norm": 0.6181258591955974,
+      "learning_rate": 3.9623575377779193e-07,
+      "loss": 0.3386,
+      "step": 5332
+    },
+    {
+      "epoch": 2.4827746741154564,
+      "grad_norm": 0.6265970211788182,
+      "learning_rate": 3.955428272158543e-07,
+      "loss": 0.3346,
+      "step": 5333
+    },
+    {
+      "epoch": 2.483240223463687,
+      "grad_norm": 0.5982792061004573,
+      "learning_rate": 3.948504550179791e-07,
+      "loss": 0.3083,
+      "step": 5334
+    },
+    {
+      "epoch": 2.483705772811918,
+      "grad_norm": 0.6373004651793105,
+      "learning_rate": 3.94158637366554e-07,
+      "loss": 0.3416,
+      "step": 5335
+    },
+    {
+      "epoch": 2.484171322160149,
+      "grad_norm": 0.5955932589913066,
+      "learning_rate": 3.9346737444381974e-07,
+      "loss": 0.3312,
+      "step": 5336
+    },
+    {
+      "epoch": 2.4846368715083798,
+      "grad_norm": 0.5935438484930112,
+      "learning_rate": 3.927766664318719e-07,
+      "loss": 0.3276,
+      "step": 5337
+    },
+    {
+      "epoch": 2.485102420856611,
+      "grad_norm": 0.5850057143172082,
+      "learning_rate": 3.9208651351265945e-07,
+      "loss": 0.32,
+      "step": 5338
+    },
+    {
+      "epoch": 2.485567970204842,
+      "grad_norm": 0.6173331698546607,
+      "learning_rate": 3.9139691586798505e-07,
+      "loss": 0.3239,
+      "step": 5339
+    },
+    {
+      "epoch": 2.4860335195530725,
+      "grad_norm": 0.6096274575305856,
+      "learning_rate": 3.907078736795039e-07,
+      "loss": 0.3362,
+      "step": 5340
+    },
+    {
+      "epoch": 2.4864990689013036,
+      "grad_norm": 0.6270193303489165,
+      "learning_rate": 3.9001938712872804e-07,
+      "loss": 0.3544,
+      "step": 5341
+    },
+    {
+      "epoch": 2.4869646182495346,
+      "grad_norm": 0.5793319613731045,
+      "learning_rate": 3.893314563970205e-07,
+      "loss": 0.3204,
+      "step": 5342
+    },
+    {
+      "epoch": 2.4874301675977653,
+      "grad_norm": 0.5926805748054752,
+      "learning_rate": 3.886440816655976e-07,
+      "loss": 0.322,
+      "step": 5343
+    },
+    {
+      "epoch": 2.4878957169459963,
+      "grad_norm": 0.6075397789687701,
+      "learning_rate": 3.8795726311553123e-07,
+      "loss": 0.3326,
+      "step": 5344
+    },
+    {
+      "epoch": 2.4883612662942274,
+      "grad_norm": 0.6845922703101732,
+      "learning_rate": 3.8727100092774584e-07,
+      "loss": 0.3161,
+      "step": 5345
+    },
+    {
+      "epoch": 2.488826815642458,
+      "grad_norm": 0.6164755617086349,
+      "learning_rate": 3.8658529528301817e-07,
+      "loss": 0.322,
+      "step": 5346
+    },
+    {
+      "epoch": 2.489292364990689,
+      "grad_norm": 0.6001016954878218,
+      "learning_rate": 3.8590014636198046e-07,
+      "loss": 0.336,
+      "step": 5347
+    },
+    {
+      "epoch": 2.48975791433892,
+      "grad_norm": 0.6582854959352635,
+      "learning_rate": 3.852155543451172e-07,
+      "loss": 0.357,
+      "step": 5348
+    },
+    {
+      "epoch": 2.4902234636871508,
+      "grad_norm": 0.6067440043594794,
+      "learning_rate": 3.845315194127652e-07,
+      "loss": 0.3294,
+      "step": 5349
+    },
+    {
+      "epoch": 2.490689013035382,
+      "grad_norm": 0.613313080234627,
+      "learning_rate": 3.838480417451168e-07,
+      "loss": 0.3316,
+      "step": 5350
+    },
+    {
+      "epoch": 2.4911545623836124,
+      "grad_norm": 0.6347161892780195,
+      "learning_rate": 3.8316512152221486e-07,
+      "loss": 0.3172,
+      "step": 5351
+    },
+    {
+      "epoch": 2.4916201117318435,
+      "grad_norm": 0.6151030550224303,
+      "learning_rate": 3.824827589239577e-07,
+      "loss": 0.348,
+      "step": 5352
+    },
+    {
+      "epoch": 2.4920856610800746,
+      "grad_norm": 0.619262010377189,
+      "learning_rate": 3.8180095413009654e-07,
+      "loss": 0.3367,
+      "step": 5353
+    },
+    {
+      "epoch": 2.4925512104283056,
+      "grad_norm": 0.6026985911556728,
+      "learning_rate": 3.811197073202333e-07,
+      "loss": 0.3362,
+      "step": 5354
+    },
+    {
+      "epoch": 2.4930167597765363,
+      "grad_norm": 0.629456218732948,
+      "learning_rate": 3.8043901867382574e-07,
+      "loss": 0.3283,
+      "step": 5355
+    },
+    {
+      "epoch": 2.4934823091247673,
+      "grad_norm": 0.6074802678086226,
+      "learning_rate": 3.797588883701836e-07,
+      "loss": 0.3186,
+      "step": 5356
+    },
+    {
+      "epoch": 2.493947858472998,
+      "grad_norm": 0.6233572676954271,
+      "learning_rate": 3.790793165884682e-07,
+      "loss": 0.3225,
+      "step": 5357
+    },
+    {
+      "epoch": 2.494413407821229,
+      "grad_norm": 0.5919528886979986,
+      "learning_rate": 3.784003035076966e-07,
+      "loss": 0.3098,
+      "step": 5358
+    },
+    {
+      "epoch": 2.49487895716946,
+      "grad_norm": 0.6059323899861924,
+      "learning_rate": 3.77721849306735e-07,
+      "loss": 0.3468,
+      "step": 5359
+    },
+    {
+      "epoch": 2.4953445065176907,
+      "grad_norm": 0.637515784477195,
+      "learning_rate": 3.770439541643059e-07,
+      "loss": 0.3247,
+      "step": 5360
+    },
+    {
+      "epoch": 2.4958100558659218,
+      "grad_norm": 0.6495491325496149,
+      "learning_rate": 3.763666182589826e-07,
+      "loss": 0.3265,
+      "step": 5361
+    },
+    {
+      "epoch": 2.496275605214153,
+      "grad_norm": 0.6067018921315109,
+      "learning_rate": 3.756898417691918e-07,
+      "loss": 0.322,
+      "step": 5362
+    },
+    {
+      "epoch": 2.4967411545623834,
+      "grad_norm": 0.6340638752498587,
+      "learning_rate": 3.750136248732106e-07,
+      "loss": 0.3381,
+      "step": 5363
+    },
+    {
+      "epoch": 2.4972067039106145,
+      "grad_norm": 0.6020170461289803,
+      "learning_rate": 3.7433796774917347e-07,
+      "loss": 0.3382,
+      "step": 5364
+    },
+    {
+      "epoch": 2.4976722532588456,
+      "grad_norm": 0.6066873971109005,
+      "learning_rate": 3.7366287057506333e-07,
+      "loss": 0.3367,
+      "step": 5365
+    },
+    {
+      "epoch": 2.498137802607076,
+      "grad_norm": 0.6172001222061847,
+      "learning_rate": 3.72988333528716e-07,
+      "loss": 0.3274,
+      "step": 5366
+    },
+    {
+      "epoch": 2.4986033519553073,
+      "grad_norm": 0.6233442791880677,
+      "learning_rate": 3.723143567878212e-07,
+      "loss": 0.3286,
+      "step": 5367
+    },
+    {
+      "epoch": 2.4990689013035383,
+      "grad_norm": 0.6024200844862596,
+      "learning_rate": 3.716409405299212e-07,
+      "loss": 0.3198,
+      "step": 5368
+    },
+    {
+      "epoch": 2.499534450651769,
+      "grad_norm": 0.6165131791347098,
+      "learning_rate": 3.709680849324085e-07,
+      "loss": 0.3218,
+      "step": 5369
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.6148872638383455,
+      "learning_rate": 3.702957901725307e-07,
+      "loss": 0.3394,
+      "step": 5370
+    },
+    {
+      "epoch": 2.500465549348231,
+      "grad_norm": 0.6135567463568582,
+      "learning_rate": 3.696240564273848e-07,
+      "loss": 0.3249,
+      "step": 5371
+    },
+    {
+      "epoch": 2.5009310986964617,
+      "grad_norm": 0.6143369964334751,
+      "learning_rate": 3.68952883873922e-07,
+      "loss": 0.3189,
+      "step": 5372
+    },
+    {
+      "epoch": 2.5013966480446927,
+      "grad_norm": 0.6001038595948466,
+      "learning_rate": 3.6828227268894594e-07,
+      "loss": 0.3355,
+      "step": 5373
+    },
+    {
+      "epoch": 2.501862197392924,
+      "grad_norm": 0.5841342413541823,
+      "learning_rate": 3.676122230491108e-07,
+      "loss": 0.3164,
+      "step": 5374
+    },
+    {
+      "epoch": 2.5023277467411544,
+      "grad_norm": 0.6189585987129558,
+      "learning_rate": 3.6694273513092265e-07,
+      "loss": 0.3425,
+      "step": 5375
+    },
+    {
+      "epoch": 2.5027932960893855,
+      "grad_norm": 0.6067165596622363,
+      "learning_rate": 3.662738091107423e-07,
+      "loss": 0.3352,
+      "step": 5376
+    },
+    {
+      "epoch": 2.5032588454376166,
+      "grad_norm": 0.6328632607321274,
+      "learning_rate": 3.656054451647803e-07,
+      "loss": 0.3478,
+      "step": 5377
+    },
+    {
+      "epoch": 2.503724394785847,
+      "grad_norm": 0.6068821461510381,
+      "learning_rate": 3.649376434690985e-07,
+      "loss": 0.3036,
+      "step": 5378
+    },
+    {
+      "epoch": 2.5041899441340782,
+      "grad_norm": 0.6015663933310095,
+      "learning_rate": 3.642704041996126e-07,
+      "loss": 0.3291,
+      "step": 5379
+    },
+    {
+      "epoch": 2.504655493482309,
+      "grad_norm": 0.5986396037606695,
+      "learning_rate": 3.636037275320897e-07,
+      "loss": 0.3042,
+      "step": 5380
+    },
+    {
+      "epoch": 2.50512104283054,
+      "grad_norm": 0.6618878223583332,
+      "learning_rate": 3.629376136421475e-07,
+      "loss": 0.338,
+      "step": 5381
+    },
+    {
+      "epoch": 2.505586592178771,
+      "grad_norm": 0.6564079386627263,
+      "learning_rate": 3.622720627052567e-07,
+      "loss": 0.3482,
+      "step": 5382
+    },
+    {
+      "epoch": 2.506052141527002,
+      "grad_norm": 0.5889684815302673,
+      "learning_rate": 3.6160707489673874e-07,
+      "loss": 0.3327,
+      "step": 5383
+    },
+    {
+      "epoch": 2.5065176908752327,
+      "grad_norm": 0.6105393159581501,
+      "learning_rate": 3.6094265039176765e-07,
+      "loss": 0.325,
+      "step": 5384
+    },
+    {
+      "epoch": 2.5069832402234637,
+      "grad_norm": 0.6189544220547761,
+      "learning_rate": 3.602787893653692e-07,
+      "loss": 0.347,
+      "step": 5385
+    },
+    {
+      "epoch": 2.5074487895716944,
+      "grad_norm": 0.5993966110481805,
+      "learning_rate": 3.5961549199241884e-07,
+      "loss": 0.3131,
+      "step": 5386
+    },
+    {
+      "epoch": 2.5079143389199254,
+      "grad_norm": 0.6105525120905546,
+      "learning_rate": 3.5895275844764545e-07,
+      "loss": 0.3443,
+      "step": 5387
+    },
+    {
+      "epoch": 2.5083798882681565,
+      "grad_norm": 0.5803038153417353,
+      "learning_rate": 3.5829058890562955e-07,
+      "loss": 0.3257,
+      "step": 5388
+    },
+    {
+      "epoch": 2.5088454376163876,
+      "grad_norm": 0.5953951057914729,
+      "learning_rate": 3.576289835408012e-07,
+      "loss": 0.3192,
+      "step": 5389
+    },
+    {
+      "epoch": 2.509310986964618,
+      "grad_norm": 0.6406901057452745,
+      "learning_rate": 3.569679425274436e-07,
+      "loss": 0.3127,
+      "step": 5390
+    },
+    {
+      "epoch": 2.5097765363128492,
+      "grad_norm": 0.6152350204618136,
+      "learning_rate": 3.563074660396909e-07,
+      "loss": 0.3278,
+      "step": 5391
+    },
+    {
+      "epoch": 2.51024208566108,
+      "grad_norm": 0.6079231171378816,
+      "learning_rate": 3.5564755425152755e-07,
+      "loss": 0.3359,
+      "step": 5392
+    },
+    {
+      "epoch": 2.510707635009311,
+      "grad_norm": 0.6150932976356819,
+      "learning_rate": 3.549882073367911e-07,
+      "loss": 0.3165,
+      "step": 5393
+    },
+    {
+      "epoch": 2.511173184357542,
+      "grad_norm": 0.584668861688503,
+      "learning_rate": 3.543294254691679e-07,
+      "loss": 0.3,
+      "step": 5394
+    },
+    {
+      "epoch": 2.511638733705773,
+      "grad_norm": 0.6031032044600075,
+      "learning_rate": 3.536712088221972e-07,
+      "loss": 0.314,
+      "step": 5395
+    },
+    {
+      "epoch": 2.5121042830540037,
+      "grad_norm": 0.6132592422876905,
+      "learning_rate": 3.5301355756927003e-07,
+      "loss": 0.3256,
+      "step": 5396
+    },
+    {
+      "epoch": 2.5125698324022347,
+      "grad_norm": 0.5937324195608167,
+      "learning_rate": 3.5235647188362617e-07,
+      "loss": 0.3456,
+      "step": 5397
+    },
+    {
+      "epoch": 2.5130353817504654,
+      "grad_norm": 0.6147734040489424,
+      "learning_rate": 3.516999519383571e-07,
+      "loss": 0.3411,
+      "step": 5398
+    },
+    {
+      "epoch": 2.5135009310986964,
+      "grad_norm": 0.615362985917816,
+      "learning_rate": 3.510439979064073e-07,
+      "loss": 0.3346,
+      "step": 5399
+    },
+    {
+      "epoch": 2.5139664804469275,
+      "grad_norm": 0.6008750642150296,
+      "learning_rate": 3.5038860996057024e-07,
+      "loss": 0.3372,
+      "step": 5400
+    },
+    {
+      "epoch": 2.5144320297951586,
+      "grad_norm": 0.6340764846723581,
+      "learning_rate": 3.4973378827348997e-07,
+      "loss": 0.3234,
+      "step": 5401
+    },
+    {
+      "epoch": 2.514897579143389,
+      "grad_norm": 0.5946339845076845,
+      "learning_rate": 3.4907953301766217e-07,
+      "loss": 0.3274,
+      "step": 5402
+    },
+    {
+      "epoch": 2.5153631284916202,
+      "grad_norm": 0.6409452807484592,
+      "learning_rate": 3.4842584436543454e-07,
+      "loss": 0.3326,
+      "step": 5403
+    },
+    {
+      "epoch": 2.515828677839851,
+      "grad_norm": 0.5914784199199098,
+      "learning_rate": 3.4777272248900236e-07,
+      "loss": 0.3195,
+      "step": 5404
+    },
+    {
+      "epoch": 2.516294227188082,
+      "grad_norm": 0.5911010837815637,
+      "learning_rate": 3.4712016756041523e-07,
+      "loss": 0.3331,
+      "step": 5405
+    },
+    {
+      "epoch": 2.516759776536313,
+      "grad_norm": 0.6125183007173084,
+      "learning_rate": 3.464681797515698e-07,
+      "loss": 0.3295,
+      "step": 5406
+    },
+    {
+      "epoch": 2.5172253258845436,
+      "grad_norm": 0.618963528223916,
+      "learning_rate": 3.458167592342165e-07,
+      "loss": 0.3174,
+      "step": 5407
+    },
+    {
+      "epoch": 2.5176908752327747,
+      "grad_norm": 0.6005904848688498,
+      "learning_rate": 3.4516590617995505e-07,
+      "loss": 0.3339,
+      "step": 5408
+    },
+    {
+      "epoch": 2.5181564245810057,
+      "grad_norm": 0.6001265008629275,
+      "learning_rate": 3.4451562076023484e-07,
+      "loss": 0.3211,
+      "step": 5409
+    },
+    {
+      "epoch": 2.5186219739292364,
+      "grad_norm": 0.654857498627313,
+      "learning_rate": 3.438659031463565e-07,
+      "loss": 0.3186,
+      "step": 5410
+    },
+    {
+      "epoch": 2.5190875232774674,
+      "grad_norm": 0.5773540260945633,
+      "learning_rate": 3.432167535094724e-07,
+      "loss": 0.3151,
+      "step": 5411
+    },
+    {
+      "epoch": 2.5195530726256985,
+      "grad_norm": 0.5927924084118499,
+      "learning_rate": 3.425681720205823e-07,
+      "loss": 0.3275,
+      "step": 5412
+    },
+    {
+      "epoch": 2.520018621973929,
+      "grad_norm": 0.6172565438913266,
+      "learning_rate": 3.4192015885053916e-07,
+      "loss": 0.3362,
+      "step": 5413
+    },
+    {
+      "epoch": 2.52048417132216,
+      "grad_norm": 0.6163067806734143,
+      "learning_rate": 3.412727141700445e-07,
+      "loss": 0.3248,
+      "step": 5414
+    },
+    {
+      "epoch": 2.520949720670391,
+      "grad_norm": 0.6329224242605433,
+      "learning_rate": 3.406258381496505e-07,
+      "loss": 0.3229,
+      "step": 5415
+    },
+    {
+      "epoch": 2.521415270018622,
+      "grad_norm": 0.5972086258941033,
+      "learning_rate": 3.399795309597603e-07,
+      "loss": 0.3268,
+      "step": 5416
+    },
+    {
+      "epoch": 2.521880819366853,
+      "grad_norm": 0.6119685864296432,
+      "learning_rate": 3.3933379277062666e-07,
+      "loss": 0.3325,
+      "step": 5417
+    },
+    {
+      "epoch": 2.522346368715084,
+      "grad_norm": 0.6004668209522015,
+      "learning_rate": 3.3868862375235066e-07,
+      "loss": 0.3303,
+      "step": 5418
+    },
+    {
+      "epoch": 2.5228119180633146,
+      "grad_norm": 0.6312766533410668,
+      "learning_rate": 3.380440240748878e-07,
+      "loss": 0.3136,
+      "step": 5419
+    },
+    {
+      "epoch": 2.5232774674115457,
+      "grad_norm": 0.6409297742890119,
+      "learning_rate": 3.373999939080394e-07,
+      "loss": 0.3459,
+      "step": 5420
+    },
+    {
+      "epoch": 2.5237430167597763,
+      "grad_norm": 0.5762816540558687,
+      "learning_rate": 3.3675653342145816e-07,
+      "loss": 0.3202,
+      "step": 5421
+    },
+    {
+      "epoch": 2.5242085661080074,
+      "grad_norm": 0.6154364308249534,
+      "learning_rate": 3.361136427846473e-07,
+      "loss": 0.3183,
+      "step": 5422
+    },
+    {
+      "epoch": 2.5246741154562384,
+      "grad_norm": 0.6361473998258668,
+      "learning_rate": 3.354713221669598e-07,
+      "loss": 0.3345,
+      "step": 5423
+    },
+    {
+      "epoch": 2.5251396648044695,
+      "grad_norm": 0.5969737649089297,
+      "learning_rate": 3.3482957173759737e-07,
+      "loss": 0.3159,
+      "step": 5424
+    },
+    {
+      "epoch": 2.5256052141527,
+      "grad_norm": 0.6118641303514738,
+      "learning_rate": 3.3418839166561357e-07,
+      "loss": 0.3298,
+      "step": 5425
+    },
+    {
+      "epoch": 2.526070763500931,
+      "grad_norm": 0.5936073891561653,
+      "learning_rate": 3.33547782119909e-07,
+      "loss": 0.3215,
+      "step": 5426
+    },
+    {
+      "epoch": 2.526536312849162,
+      "grad_norm": 0.5997299337133074,
+      "learning_rate": 3.3290774326923625e-07,
+      "loss": 0.3174,
+      "step": 5427
+    },
+    {
+      "epoch": 2.527001862197393,
+      "grad_norm": 0.5977249609126214,
+      "learning_rate": 3.3226827528219706e-07,
+      "loss": 0.3335,
+      "step": 5428
+    },
+    {
+      "epoch": 2.527467411545624,
+      "grad_norm": 0.550643050896851,
+      "learning_rate": 3.31629378327242e-07,
+      "loss": 0.3197,
+      "step": 5429
+    },
+    {
+      "epoch": 2.527932960893855,
+      "grad_norm": 0.6031709625670468,
+      "learning_rate": 3.3099105257267144e-07,
+      "loss": 0.323,
+      "step": 5430
+    },
+    {
+      "epoch": 2.5283985102420856,
+      "grad_norm": 0.628702256455163,
+      "learning_rate": 3.303532981866367e-07,
+      "loss": 0.3412,
+      "step": 5431
+    },
+    {
+      "epoch": 2.5288640595903167,
+      "grad_norm": 0.5869524236692155,
+      "learning_rate": 3.2971611533713643e-07,
+      "loss": 0.3414,
+      "step": 5432
+    },
+    {
+      "epoch": 2.5293296089385473,
+      "grad_norm": 0.5783278804624818,
+      "learning_rate": 3.290795041920197e-07,
+      "loss": 0.3185,
+      "step": 5433
+    },
+    {
+      "epoch": 2.5297951582867784,
+      "grad_norm": 0.6114011831199032,
+      "learning_rate": 3.2844346491898603e-07,
+      "loss": 0.3228,
+      "step": 5434
+    },
+    {
+      "epoch": 2.5302607076350094,
+      "grad_norm": 0.6111282709771415,
+      "learning_rate": 3.278079976855822e-07,
+      "loss": 0.3132,
+      "step": 5435
+    },
+    {
+      "epoch": 2.5307262569832405,
+      "grad_norm": 0.6051079586815271,
+      "learning_rate": 3.271731026592062e-07,
+      "loss": 0.3295,
+      "step": 5436
+    },
+    {
+      "epoch": 2.531191806331471,
+      "grad_norm": 0.5992706066281824,
+      "learning_rate": 3.265387800071038e-07,
+      "loss": 0.313,
+      "step": 5437
+    },
+    {
+      "epoch": 2.531657355679702,
+      "grad_norm": 0.6029667303123079,
+      "learning_rate": 3.2590502989637093e-07,
+      "loss": 0.3265,
+      "step": 5438
+    },
+    {
+      "epoch": 2.532122905027933,
+      "grad_norm": 0.5925361174467528,
+      "learning_rate": 3.252718524939527e-07,
+      "loss": 0.318,
+      "step": 5439
+    },
+    {
+      "epoch": 2.532588454376164,
+      "grad_norm": 0.6064595809007656,
+      "learning_rate": 3.2463924796664326e-07,
+      "loss": 0.327,
+      "step": 5440
+    },
+    {
+      "epoch": 2.533054003724395,
+      "grad_norm": 0.6086679196294492,
+      "learning_rate": 3.2400721648108397e-07,
+      "loss": 0.3144,
+      "step": 5441
+    },
+    {
+      "epoch": 2.5335195530726256,
+      "grad_norm": 0.6022353233572794,
+      "learning_rate": 3.2337575820376944e-07,
+      "loss": 0.3165,
+      "step": 5442
+    },
+    {
+      "epoch": 2.5339851024208566,
+      "grad_norm": 0.6145182394939922,
+      "learning_rate": 3.227448733010397e-07,
+      "loss": 0.3282,
+      "step": 5443
+    },
+    {
+      "epoch": 2.5344506517690877,
+      "grad_norm": 0.6131581432134133,
+      "learning_rate": 3.221145619390842e-07,
+      "loss": 0.3317,
+      "step": 5444
+    },
+    {
+      "epoch": 2.5349162011173183,
+      "grad_norm": 0.5933804017412879,
+      "learning_rate": 3.214848242839427e-07,
+      "loss": 0.3321,
+      "step": 5445
+    },
+    {
+      "epoch": 2.5353817504655494,
+      "grad_norm": 0.5998641581633004,
+      "learning_rate": 3.208556605015031e-07,
+      "loss": 0.3173,
+      "step": 5446
+    },
+    {
+      "epoch": 2.5358472998137804,
+      "grad_norm": 0.6248013519985028,
+      "learning_rate": 3.202270707575017e-07,
+      "loss": 0.3304,
+      "step": 5447
+    },
+    {
+      "epoch": 2.536312849162011,
+      "grad_norm": 0.6334357174984995,
+      "learning_rate": 3.195990552175246e-07,
+      "loss": 0.3311,
+      "step": 5448
+    },
+    {
+      "epoch": 2.536778398510242,
+      "grad_norm": 0.6232341068192774,
+      "learning_rate": 3.1897161404700507e-07,
+      "loss": 0.3354,
+      "step": 5449
+    },
+    {
+      "epoch": 2.5372439478584727,
+      "grad_norm": 0.622510457707116,
+      "learning_rate": 3.183447474112264e-07,
+      "loss": 0.3287,
+      "step": 5450
+    },
+    {
+      "epoch": 2.537709497206704,
+      "grad_norm": 0.6052167564054277,
+      "learning_rate": 3.17718455475321e-07,
+      "loss": 0.341,
+      "step": 5451
+    },
+    {
+      "epoch": 2.538175046554935,
+      "grad_norm": 0.6029133786603633,
+      "learning_rate": 3.17092738404268e-07,
+      "loss": 0.3396,
+      "step": 5452
+    },
+    {
+      "epoch": 2.538640595903166,
+      "grad_norm": 0.6023027770555903,
+      "learning_rate": 3.164675963628963e-07,
+      "loss": 0.3247,
+      "step": 5453
+    },
+    {
+      "epoch": 2.5391061452513966,
+      "grad_norm": 0.6096587988266041,
+      "learning_rate": 3.15843029515884e-07,
+      "loss": 0.3215,
+      "step": 5454
+    },
+    {
+      "epoch": 2.5395716945996276,
+      "grad_norm": 0.6273845613502879,
+      "learning_rate": 3.1521903802775647e-07,
+      "loss": 0.3319,
+      "step": 5455
+    },
+    {
+      "epoch": 2.5400372439478582,
+      "grad_norm": 0.616562622800256,
+      "learning_rate": 3.145956220628871e-07,
+      "loss": 0.346,
+      "step": 5456
+    },
+    {
+      "epoch": 2.5405027932960893,
+      "grad_norm": 0.5814088209540305,
+      "learning_rate": 3.1397278178549926e-07,
+      "loss": 0.3134,
+      "step": 5457
+    },
+    {
+      "epoch": 2.5409683426443204,
+      "grad_norm": 0.6095786622522564,
+      "learning_rate": 3.1335051735966407e-07,
+      "loss": 0.3342,
+      "step": 5458
+    },
+    {
+      "epoch": 2.5414338919925514,
+      "grad_norm": 0.6122484451937094,
+      "learning_rate": 3.1272882894929995e-07,
+      "loss": 0.3264,
+      "step": 5459
+    },
+    {
+      "epoch": 2.541899441340782,
+      "grad_norm": 0.615122107724369,
+      "learning_rate": 3.1210771671817545e-07,
+      "loss": 0.3112,
+      "step": 5460
+    },
+    {
+      "epoch": 2.542364990689013,
+      "grad_norm": 0.5961789243105731,
+      "learning_rate": 3.1148718082990555e-07,
+      "loss": 0.325,
+      "step": 5461
+    },
+    {
+      "epoch": 2.5428305400372437,
+      "grad_norm": 0.6236445772527093,
+      "learning_rate": 3.108672214479541e-07,
+      "loss": 0.3468,
+      "step": 5462
+    },
+    {
+      "epoch": 2.543296089385475,
+      "grad_norm": 0.594713594804183,
+      "learning_rate": 3.1024783873563393e-07,
+      "loss": 0.3355,
+      "step": 5463
+    },
+    {
+      "epoch": 2.543761638733706,
+      "grad_norm": 0.6061523206107379,
+      "learning_rate": 3.096290328561041e-07,
+      "loss": 0.3283,
+      "step": 5464
+    },
+    {
+      "epoch": 2.544227188081937,
+      "grad_norm": 0.6004330916119782,
+      "learning_rate": 3.0901080397237327e-07,
+      "loss": 0.3239,
+      "step": 5465
+    },
+    {
+      "epoch": 2.5446927374301676,
+      "grad_norm": 0.6358854409255706,
+      "learning_rate": 3.0839315224729854e-07,
+      "loss": 0.3368,
+      "step": 5466
+    },
+    {
+      "epoch": 2.5451582867783986,
+      "grad_norm": 0.7158643066528086,
+      "learning_rate": 3.077760778435823e-07,
+      "loss": 0.3228,
+      "step": 5467
+    },
+    {
+      "epoch": 2.5456238361266292,
+      "grad_norm": 0.5972989357127529,
+      "learning_rate": 3.0715958092377844e-07,
+      "loss": 0.3206,
+      "step": 5468
+    },
+    {
+      "epoch": 2.5460893854748603,
+      "grad_norm": 0.600769383831201,
+      "learning_rate": 3.065436616502854e-07,
+      "loss": 0.3104,
+      "step": 5469
+    },
+    {
+      "epoch": 2.5465549348230914,
+      "grad_norm": 0.6228021142675033,
+      "learning_rate": 3.0592832018535165e-07,
+      "loss": 0.3185,
+      "step": 5470
+    },
+    {
+      "epoch": 2.5470204841713224,
+      "grad_norm": 0.6034482294898921,
+      "learning_rate": 3.053135566910731e-07,
+      "loss": 0.3315,
+      "step": 5471
+    },
+    {
+      "epoch": 2.547486033519553,
+      "grad_norm": 0.5650302913600528,
+      "learning_rate": 3.0469937132939276e-07,
+      "loss": 0.3243,
+      "step": 5472
+    },
+    {
+      "epoch": 2.547951582867784,
+      "grad_norm": 0.5983454647758107,
+      "learning_rate": 3.040857642621003e-07,
+      "loss": 0.3459,
+      "step": 5473
+    },
+    {
+      "epoch": 2.5484171322160147,
+      "grad_norm": 0.6080559782290484,
+      "learning_rate": 3.034727356508371e-07,
+      "loss": 0.3294,
+      "step": 5474
+    },
+    {
+      "epoch": 2.548882681564246,
+      "grad_norm": 0.6093143943825647,
+      "learning_rate": 3.028602856570878e-07,
+      "loss": 0.3335,
+      "step": 5475
+    },
+    {
+      "epoch": 2.549348230912477,
+      "grad_norm": 0.6129356008187292,
+      "learning_rate": 3.0224841444218557e-07,
+      "loss": 0.3248,
+      "step": 5476
+    },
+    {
+      "epoch": 2.5498137802607075,
+      "grad_norm": 0.58605163983781,
+      "learning_rate": 3.016371221673139e-07,
+      "loss": 0.3355,
+      "step": 5477
+    },
+    {
+      "epoch": 2.5502793296089385,
+      "grad_norm": 0.5942649388135576,
+      "learning_rate": 3.010264089935008e-07,
+      "loss": 0.3218,
+      "step": 5478
+    },
+    {
+      "epoch": 2.5507448789571696,
+      "grad_norm": 0.5935736811442406,
+      "learning_rate": 3.004162750816217e-07,
+      "loss": 0.3168,
+      "step": 5479
+    },
+    {
+      "epoch": 2.5512104283054002,
+      "grad_norm": 0.5778476957637384,
+      "learning_rate": 2.998067205924013e-07,
+      "loss": 0.3152,
+      "step": 5480
+    },
+    {
+      "epoch": 2.5516759776536313,
+      "grad_norm": 0.6224726687595141,
+      "learning_rate": 2.991977456864109e-07,
+      "loss": 0.3023,
+      "step": 5481
+    },
+    {
+      "epoch": 2.5521415270018624,
+      "grad_norm": 0.6098302520883292,
+      "learning_rate": 2.9858935052406834e-07,
+      "loss": 0.329,
+      "step": 5482
+    },
+    {
+      "epoch": 2.552607076350093,
+      "grad_norm": 0.6551400874541043,
+      "learning_rate": 2.9798153526564037e-07,
+      "loss": 0.3454,
+      "step": 5483
+    },
+    {
+      "epoch": 2.553072625698324,
+      "grad_norm": 0.5860621395390316,
+      "learning_rate": 2.973743000712387e-07,
+      "loss": 0.3397,
+      "step": 5484
+    },
+    {
+      "epoch": 2.5535381750465547,
+      "grad_norm": 0.6140868699528707,
+      "learning_rate": 2.967676451008239e-07,
+      "loss": 0.3239,
+      "step": 5485
+    },
+    {
+      "epoch": 2.5540037243947857,
+      "grad_norm": 0.6048348041887376,
+      "learning_rate": 2.961615705142043e-07,
+      "loss": 0.3227,
+      "step": 5486
+    },
+    {
+      "epoch": 2.554469273743017,
+      "grad_norm": 0.6068218247885355,
+      "learning_rate": 2.955560764710333e-07,
+      "loss": 0.322,
+      "step": 5487
+    },
+    {
+      "epoch": 2.554934823091248,
+      "grad_norm": 0.6005958750159981,
+      "learning_rate": 2.9495116313081245e-07,
+      "loss": 0.3257,
+      "step": 5488
+    },
+    {
+      "epoch": 2.5554003724394785,
+      "grad_norm": 0.5977103921424048,
+      "learning_rate": 2.943468306528913e-07,
+      "loss": 0.31,
+      "step": 5489
+    },
+    {
+      "epoch": 2.5558659217877095,
+      "grad_norm": 0.605728546500175,
+      "learning_rate": 2.9374307919646377e-07,
+      "loss": 0.3185,
+      "step": 5490
+    },
+    {
+      "epoch": 2.55633147113594,
+      "grad_norm": 0.5838161010366467,
+      "learning_rate": 2.931399089205741e-07,
+      "loss": 0.3174,
+      "step": 5491
+    },
+    {
+      "epoch": 2.5567970204841712,
+      "grad_norm": 0.60222110383688,
+      "learning_rate": 2.925373199841106e-07,
+      "loss": 0.3154,
+      "step": 5492
+    },
+    {
+      "epoch": 2.5572625698324023,
+      "grad_norm": 0.5933736638827337,
+      "learning_rate": 2.919353125458094e-07,
+      "loss": 0.3086,
+      "step": 5493
+    },
+    {
+      "epoch": 2.5577281191806334,
+      "grad_norm": 0.6288584596138415,
+      "learning_rate": 2.9133388676425455e-07,
+      "loss": 0.3413,
+      "step": 5494
+    },
+    {
+      "epoch": 2.558193668528864,
+      "grad_norm": 0.6200604713698977,
+      "learning_rate": 2.9073304279787553e-07,
+      "loss": 0.3345,
+      "step": 5495
+    },
+    {
+      "epoch": 2.558659217877095,
+      "grad_norm": 0.6025264149175407,
+      "learning_rate": 2.901327808049476e-07,
+      "loss": 0.3319,
+      "step": 5496
+    },
+    {
+      "epoch": 2.5591247672253257,
+      "grad_norm": 0.6225174023089183,
+      "learning_rate": 2.895331009435962e-07,
+      "loss": 0.3166,
+      "step": 5497
+    },
+    {
+      "epoch": 2.5595903165735567,
+      "grad_norm": 0.6050494051157181,
+      "learning_rate": 2.8893400337178997e-07,
+      "loss": 0.3243,
+      "step": 5498
+    },
+    {
+      "epoch": 2.560055865921788,
+      "grad_norm": 0.5913813191383517,
+      "learning_rate": 2.8833548824734547e-07,
+      "loss": 0.3191,
+      "step": 5499
+    },
+    {
+      "epoch": 2.560521415270019,
+      "grad_norm": 0.6094995880232921,
+      "learning_rate": 2.877375557279258e-07,
+      "loss": 0.323,
+      "step": 5500
+    },
+    {
+      "epoch": 2.5609869646182495,
+      "grad_norm": 0.5953793065643563,
+      "learning_rate": 2.8714020597104155e-07,
+      "loss": 0.3104,
+      "step": 5501
+    },
+    {
+      "epoch": 2.5614525139664805,
+      "grad_norm": 0.6248403587824202,
+      "learning_rate": 2.865434391340477e-07,
+      "loss": 0.3103,
+      "step": 5502
+    },
+    {
+      "epoch": 2.561918063314711,
+      "grad_norm": 0.6429646178102544,
+      "learning_rate": 2.8594725537414767e-07,
+      "loss": 0.3309,
+      "step": 5503
+    },
+    {
+      "epoch": 2.5623836126629422,
+      "grad_norm": 0.6093436351885263,
+      "learning_rate": 2.853516548483898e-07,
+      "loss": 0.3367,
+      "step": 5504
+    },
+    {
+      "epoch": 2.5628491620111733,
+      "grad_norm": 0.6175354836653819,
+      "learning_rate": 2.8475663771366956e-07,
+      "loss": 0.3144,
+      "step": 5505
+    },
+    {
+      "epoch": 2.5633147113594044,
+      "grad_norm": 0.5970558712989233,
+      "learning_rate": 2.841622041267292e-07,
+      "loss": 0.3119,
+      "step": 5506
+    },
+    {
+      "epoch": 2.563780260707635,
+      "grad_norm": 0.6177892429836915,
+      "learning_rate": 2.8356835424415556e-07,
+      "loss": 0.3258,
+      "step": 5507
+    },
+    {
+      "epoch": 2.564245810055866,
+      "grad_norm": 0.6020106910045734,
+      "learning_rate": 2.829750882223836e-07,
+      "loss": 0.315,
+      "step": 5508
+    },
+    {
+      "epoch": 2.5647113594040967,
+      "grad_norm": 0.6053411137478001,
+      "learning_rate": 2.8238240621769415e-07,
+      "loss": 0.3299,
+      "step": 5509
+    },
+    {
+      "epoch": 2.5651769087523277,
+      "grad_norm": 0.6067002546213581,
+      "learning_rate": 2.817903083862125e-07,
+      "loss": 0.3329,
+      "step": 5510
+    },
+    {
+      "epoch": 2.565642458100559,
+      "grad_norm": 0.6390549848790155,
+      "learning_rate": 2.811987948839126e-07,
+      "loss": 0.3373,
+      "step": 5511
+    },
+    {
+      "epoch": 2.5661080074487894,
+      "grad_norm": 0.5974897048718771,
+      "learning_rate": 2.8060786586661196e-07,
+      "loss": 0.3204,
+      "step": 5512
+    },
+    {
+      "epoch": 2.5665735567970205,
+      "grad_norm": 0.5868221332656389,
+      "learning_rate": 2.8001752148997584e-07,
+      "loss": 0.3345,
+      "step": 5513
+    },
+    {
+      "epoch": 2.5670391061452515,
+      "grad_norm": 0.6252978098860659,
+      "learning_rate": 2.7942776190951564e-07,
+      "loss": 0.33,
+      "step": 5514
+    },
+    {
+      "epoch": 2.567504655493482,
+      "grad_norm": 0.5964707944732232,
+      "learning_rate": 2.788385872805874e-07,
+      "loss": 0.3183,
+      "step": 5515
+    },
+    {
+      "epoch": 2.5679702048417132,
+      "grad_norm": 0.600798304655396,
+      "learning_rate": 2.78249997758393e-07,
+      "loss": 0.3189,
+      "step": 5516
+    },
+    {
+      "epoch": 2.5684357541899443,
+      "grad_norm": 0.6227678731355668,
+      "learning_rate": 2.776619934979824e-07,
+      "loss": 0.3362,
+      "step": 5517
+    },
+    {
+      "epoch": 2.568901303538175,
+      "grad_norm": 0.6327510297474038,
+      "learning_rate": 2.770745746542494e-07,
+      "loss": 0.3151,
+      "step": 5518
+    },
+    {
+      "epoch": 2.569366852886406,
+      "grad_norm": 0.611519784482493,
+      "learning_rate": 2.7648774138193337e-07,
+      "loss": 0.3233,
+      "step": 5519
+    },
+    {
+      "epoch": 2.5698324022346366,
+      "grad_norm": 0.6064716800211711,
+      "learning_rate": 2.759014938356208e-07,
+      "loss": 0.3161,
+      "step": 5520
+    },
+    {
+      "epoch": 2.5702979515828677,
+      "grad_norm": 0.6193605983722354,
+      "learning_rate": 2.753158321697436e-07,
+      "loss": 0.3184,
+      "step": 5521
+    },
+    {
+      "epoch": 2.5707635009310987,
+      "grad_norm": 0.6234607335175127,
+      "learning_rate": 2.7473075653857787e-07,
+      "loss": 0.3421,
+      "step": 5522
+    },
+    {
+      "epoch": 2.57122905027933,
+      "grad_norm": 0.6020665527845989,
+      "learning_rate": 2.7414626709624694e-07,
+      "loss": 0.3264,
+      "step": 5523
+    },
+    {
+      "epoch": 2.5716945996275604,
+      "grad_norm": 0.6277215810816873,
+      "learning_rate": 2.7356236399672e-07,
+      "loss": 0.3391,
+      "step": 5524
+    },
+    {
+      "epoch": 2.5721601489757915,
+      "grad_norm": 0.6131808688201946,
+      "learning_rate": 2.7297904739381017e-07,
+      "loss": 0.3183,
+      "step": 5525
+    },
+    {
+      "epoch": 2.572625698324022,
+      "grad_norm": 0.5695353516403446,
+      "learning_rate": 2.723963174411773e-07,
+      "loss": 0.3326,
+      "step": 5526
+    },
+    {
+      "epoch": 2.573091247672253,
+      "grad_norm": 0.6327412547727915,
+      "learning_rate": 2.718141742923255e-07,
+      "loss": 0.327,
+      "step": 5527
+    },
+    {
+      "epoch": 2.5735567970204842,
+      "grad_norm": 0.6047198238958524,
+      "learning_rate": 2.712326181006061e-07,
+      "loss": 0.3241,
+      "step": 5528
+    },
+    {
+      "epoch": 2.5740223463687153,
+      "grad_norm": 0.5820469099837305,
+      "learning_rate": 2.706516490192149e-07,
+      "loss": 0.318,
+      "step": 5529
+    },
+    {
+      "epoch": 2.574487895716946,
+      "grad_norm": 0.5733101127163053,
+      "learning_rate": 2.700712672011921e-07,
+      "loss": 0.3217,
+      "step": 5530
+    },
+    {
+      "epoch": 2.574953445065177,
+      "grad_norm": 0.6167344932654663,
+      "learning_rate": 2.694914727994244e-07,
+      "loss": 0.3321,
+      "step": 5531
+    },
+    {
+      "epoch": 2.5754189944134076,
+      "grad_norm": 0.5968463296854569,
+      "learning_rate": 2.6891226596664406e-07,
+      "loss": 0.3231,
+      "step": 5532
+    },
+    {
+      "epoch": 2.5758845437616387,
+      "grad_norm": 0.6046488582866905,
+      "learning_rate": 2.683336468554273e-07,
+      "loss": 0.3207,
+      "step": 5533
+    },
+    {
+      "epoch": 2.5763500931098697,
+      "grad_norm": 0.61453207358343,
+      "learning_rate": 2.6775561561819597e-07,
+      "loss": 0.3479,
+      "step": 5534
+    },
+    {
+      "epoch": 2.576815642458101,
+      "grad_norm": 0.6475526127832754,
+      "learning_rate": 2.671781724072176e-07,
+      "loss": 0.31,
+      "step": 5535
+    },
+    {
+      "epoch": 2.5772811918063314,
+      "grad_norm": 0.6015057070520273,
+      "learning_rate": 2.6660131737460485e-07,
+      "loss": 0.3325,
+      "step": 5536
+    },
+    {
+      "epoch": 2.5777467411545625,
+      "grad_norm": 0.6067799565533306,
+      "learning_rate": 2.6602505067231414e-07,
+      "loss": 0.3297,
+      "step": 5537
+    },
+    {
+      "epoch": 2.578212290502793,
+      "grad_norm": 0.6142298269769203,
+      "learning_rate": 2.65449372452149e-07,
+      "loss": 0.3188,
+      "step": 5538
+    },
+    {
+      "epoch": 2.578677839851024,
+      "grad_norm": 0.6393190191119946,
+      "learning_rate": 2.648742828657552e-07,
+      "loss": 0.3249,
+      "step": 5539
+    },
+    {
+      "epoch": 2.5791433891992552,
+      "grad_norm": 0.5936142757057987,
+      "learning_rate": 2.642997820646262e-07,
+      "loss": 0.3309,
+      "step": 5540
+    },
+    {
+      "epoch": 2.5796089385474863,
+      "grad_norm": 0.5922014500932228,
+      "learning_rate": 2.6372587020009914e-07,
+      "loss": 0.3198,
+      "step": 5541
+    },
+    {
+      "epoch": 2.580074487895717,
+      "grad_norm": 0.5903746817146399,
+      "learning_rate": 2.631525474233554e-07,
+      "loss": 0.3384,
+      "step": 5542
+    },
+    {
+      "epoch": 2.580540037243948,
+      "grad_norm": 0.5944534701287643,
+      "learning_rate": 2.6257981388542214e-07,
+      "loss": 0.3266,
+      "step": 5543
+    },
+    {
+      "epoch": 2.5810055865921786,
+      "grad_norm": 0.600002887396565,
+      "learning_rate": 2.620076697371718e-07,
+      "loss": 0.3135,
+      "step": 5544
+    },
+    {
+      "epoch": 2.5814711359404097,
+      "grad_norm": 0.6237733754436555,
+      "learning_rate": 2.6143611512931904e-07,
+      "loss": 0.339,
+      "step": 5545
+    },
+    {
+      "epoch": 2.5819366852886407,
+      "grad_norm": 0.5925031307297968,
+      "learning_rate": 2.608651502124268e-07,
+      "loss": 0.3371,
+      "step": 5546
+    },
+    {
+      "epoch": 2.5824022346368714,
+      "grad_norm": 0.6316475695544831,
+      "learning_rate": 2.6029477513689923e-07,
+      "loss": 0.3226,
+      "step": 5547
+    },
+    {
+      "epoch": 2.5828677839851024,
+      "grad_norm": 0.61380849818849,
+      "learning_rate": 2.5972499005298746e-07,
+      "loss": 0.311,
+      "step": 5548
+    },
+    {
+      "epoch": 2.5833333333333335,
+      "grad_norm": 0.6379865655443996,
+      "learning_rate": 2.5915579511078677e-07,
+      "loss": 0.3365,
+      "step": 5549
+    },
+    {
+      "epoch": 2.583798882681564,
+      "grad_norm": 0.5773071031193222,
+      "learning_rate": 2.585871904602363e-07,
+      "loss": 0.3151,
+      "step": 5550
+    },
+    {
+      "epoch": 2.584264432029795,
+      "grad_norm": 0.6166578327881029,
+      "learning_rate": 2.580191762511189e-07,
+      "loss": 0.3168,
+      "step": 5551
+    },
+    {
+      "epoch": 2.5847299813780262,
+      "grad_norm": 0.6090269658387918,
+      "learning_rate": 2.5745175263306525e-07,
+      "loss": 0.3207,
+      "step": 5552
+    },
+    {
+      "epoch": 2.585195530726257,
+      "grad_norm": 0.6150689966410123,
+      "learning_rate": 2.5688491975554673e-07,
+      "loss": 0.3375,
+      "step": 5553
+    },
+    {
+      "epoch": 2.585661080074488,
+      "grad_norm": 0.6192726333046097,
+      "learning_rate": 2.563186777678808e-07,
+      "loss": 0.3266,
+      "step": 5554
+    },
+    {
+      "epoch": 2.5861266294227185,
+      "grad_norm": 0.6216021338021686,
+      "learning_rate": 2.5575302681922905e-07,
+      "loss": 0.3375,
+      "step": 5555
+    },
+    {
+      "epoch": 2.5865921787709496,
+      "grad_norm": 0.610313191056007,
+      "learning_rate": 2.55187967058598e-07,
+      "loss": 0.3215,
+      "step": 5556
+    },
+    {
+      "epoch": 2.5870577281191807,
+      "grad_norm": 0.6176014032321927,
+      "learning_rate": 2.546234986348367e-07,
+      "loss": 0.3213,
+      "step": 5557
+    },
+    {
+      "epoch": 2.5875232774674117,
+      "grad_norm": 0.6012000443342811,
+      "learning_rate": 2.5405962169664097e-07,
+      "loss": 0.3213,
+      "step": 5558
+    },
+    {
+      "epoch": 2.5879888268156424,
+      "grad_norm": 0.6137243859415373,
+      "learning_rate": 2.53496336392548e-07,
+      "loss": 0.3311,
+      "step": 5559
+    },
+    {
+      "epoch": 2.5884543761638734,
+      "grad_norm": 0.599460725317936,
+      "learning_rate": 2.5293364287094115e-07,
+      "loss": 0.3123,
+      "step": 5560
+    },
+    {
+      "epoch": 2.588919925512104,
+      "grad_norm": 0.6137496558249568,
+      "learning_rate": 2.523715412800479e-07,
+      "loss": 0.3354,
+      "step": 5561
+    },
+    {
+      "epoch": 2.589385474860335,
+      "grad_norm": 0.573647806070067,
+      "learning_rate": 2.518100317679381e-07,
+      "loss": 0.3227,
+      "step": 5562
+    },
+    {
+      "epoch": 2.589851024208566,
+      "grad_norm": 0.6023964449354146,
+      "learning_rate": 2.5124911448252713e-07,
+      "loss": 0.3214,
+      "step": 5563
+    },
+    {
+      "epoch": 2.5903165735567972,
+      "grad_norm": 0.5928050034096337,
+      "learning_rate": 2.5068878957157453e-07,
+      "loss": 0.3268,
+      "step": 5564
+    },
+    {
+      "epoch": 2.590782122905028,
+      "grad_norm": 0.6069365152718665,
+      "learning_rate": 2.50129057182682e-07,
+      "loss": 0.3283,
+      "step": 5565
+    },
+    {
+      "epoch": 2.591247672253259,
+      "grad_norm": 0.6091665352796839,
+      "learning_rate": 2.495699174632973e-07,
+      "loss": 0.3371,
+      "step": 5566
+    },
+    {
+      "epoch": 2.5917132216014895,
+      "grad_norm": 0.5912985195850008,
+      "learning_rate": 2.490113705607111e-07,
+      "loss": 0.3193,
+      "step": 5567
+    },
+    {
+      "epoch": 2.5921787709497206,
+      "grad_norm": 0.6102192194502774,
+      "learning_rate": 2.484534166220576e-07,
+      "loss": 0.3248,
+      "step": 5568
+    },
+    {
+      "epoch": 2.5926443202979517,
+      "grad_norm": 0.6109419026516192,
+      "learning_rate": 2.4789605579431523e-07,
+      "loss": 0.3386,
+      "step": 5569
+    },
+    {
+      "epoch": 2.5931098696461827,
+      "grad_norm": 0.6008055078748002,
+      "learning_rate": 2.4733928822430587e-07,
+      "loss": 0.334,
+      "step": 5570
+    },
+    {
+      "epoch": 2.5935754189944134,
+      "grad_norm": 0.621340300202294,
+      "learning_rate": 2.467831140586957e-07,
+      "loss": 0.3267,
+      "step": 5571
+    },
+    {
+      "epoch": 2.5940409683426444,
+      "grad_norm": 0.6111061841835888,
+      "learning_rate": 2.4622753344399424e-07,
+      "loss": 0.3318,
+      "step": 5572
+    },
+    {
+      "epoch": 2.594506517690875,
+      "grad_norm": 0.6249239691954628,
+      "learning_rate": 2.456725465265547e-07,
+      "loss": 0.3256,
+      "step": 5573
+    },
+    {
+      "epoch": 2.594972067039106,
+      "grad_norm": 0.654428458665009,
+      "learning_rate": 2.451181534525726e-07,
+      "loss": 0.3216,
+      "step": 5574
+    },
+    {
+      "epoch": 2.595437616387337,
+      "grad_norm": 0.6193737267770549,
+      "learning_rate": 2.445643543680903e-07,
+      "loss": 0.3301,
+      "step": 5575
+    },
+    {
+      "epoch": 2.5959031657355682,
+      "grad_norm": 0.5945013365629425,
+      "learning_rate": 2.4401114941899063e-07,
+      "loss": 0.3246,
+      "step": 5576
+    },
+    {
+      "epoch": 2.596368715083799,
+      "grad_norm": 0.5951690005137175,
+      "learning_rate": 2.4345853875100024e-07,
+      "loss": 0.3238,
+      "step": 5577
+    },
+    {
+      "epoch": 2.59683426443203,
+      "grad_norm": 0.595416683415301,
+      "learning_rate": 2.429065225096908e-07,
+      "loss": 0.3232,
+      "step": 5578
+    },
+    {
+      "epoch": 2.5972998137802605,
+      "grad_norm": 0.5990754958069938,
+      "learning_rate": 2.4235510084047646e-07,
+      "loss": 0.3137,
+      "step": 5579
+    },
+    {
+      "epoch": 2.5977653631284916,
+      "grad_norm": 0.6283982561871007,
+      "learning_rate": 2.418042738886145e-07,
+      "loss": 0.3356,
+      "step": 5580
+    },
+    {
+      "epoch": 2.5982309124767227,
+      "grad_norm": 0.604532297752921,
+      "learning_rate": 2.4125404179920625e-07,
+      "loss": 0.3504,
+      "step": 5581
+    },
+    {
+      "epoch": 2.5986964618249533,
+      "grad_norm": 0.6091752842224035,
+      "learning_rate": 2.407044047171955e-07,
+      "loss": 0.3041,
+      "step": 5582
+    },
+    {
+      "epoch": 2.5991620111731844,
+      "grad_norm": 0.5926580007127635,
+      "learning_rate": 2.4015536278736955e-07,
+      "loss": 0.3226,
+      "step": 5583
+    },
+    {
+      "epoch": 2.5996275605214154,
+      "grad_norm": 0.6005465849171338,
+      "learning_rate": 2.3960691615436016e-07,
+      "loss": 0.3156,
+      "step": 5584
+    },
+    {
+      "epoch": 2.600093109869646,
+      "grad_norm": 0.591006649320003,
+      "learning_rate": 2.3905906496263983e-07,
+      "loss": 0.325,
+      "step": 5585
+    },
+    {
+      "epoch": 2.600558659217877,
+      "grad_norm": 0.622729434049119,
+      "learning_rate": 2.385118093565264e-07,
+      "loss": 0.317,
+      "step": 5586
+    },
+    {
+      "epoch": 2.601024208566108,
+      "grad_norm": 0.6143098130592252,
+      "learning_rate": 2.3796514948018018e-07,
+      "loss": 0.3337,
+      "step": 5587
+    },
+    {
+      "epoch": 2.601489757914339,
+      "grad_norm": 0.6012378522159194,
+      "learning_rate": 2.3741908547760372e-07,
+      "loss": 0.321,
+      "step": 5588
+    },
+    {
+      "epoch": 2.60195530726257,
+      "grad_norm": 0.6044340365133768,
+      "learning_rate": 2.368736174926442e-07,
+      "loss": 0.3165,
+      "step": 5589
+    },
+    {
+      "epoch": 2.6024208566108005,
+      "grad_norm": 0.606121569941451,
+      "learning_rate": 2.3632874566898972e-07,
+      "loss": 0.3186,
+      "step": 5590
+    },
+    {
+      "epoch": 2.6028864059590315,
+      "grad_norm": 0.5951309709956143,
+      "learning_rate": 2.3578447015017299e-07,
+      "loss": 0.3203,
+      "step": 5591
+    },
+    {
+      "epoch": 2.6033519553072626,
+      "grad_norm": 0.6236874636708629,
+      "learning_rate": 2.3524079107956955e-07,
+      "loss": 0.3419,
+      "step": 5592
+    },
+    {
+      "epoch": 2.6038175046554937,
+      "grad_norm": 0.624521069524866,
+      "learning_rate": 2.346977086003971e-07,
+      "loss": 0.3366,
+      "step": 5593
+    },
+    {
+      "epoch": 2.6042830540037243,
+      "grad_norm": 0.6028398854798026,
+      "learning_rate": 2.3415522285571536e-07,
+      "loss": 0.3348,
+      "step": 5594
+    },
+    {
+      "epoch": 2.6047486033519553,
+      "grad_norm": 0.6126262410348035,
+      "learning_rate": 2.3361333398843e-07,
+      "loss": 0.33,
+      "step": 5595
+    },
+    {
+      "epoch": 2.605214152700186,
+      "grad_norm": 0.5803148132303175,
+      "learning_rate": 2.3307204214128658e-07,
+      "loss": 0.3326,
+      "step": 5596
+    },
+    {
+      "epoch": 2.605679702048417,
+      "grad_norm": 0.6123705733641707,
+      "learning_rate": 2.325313474568733e-07,
+      "loss": 0.327,
+      "step": 5597
+    },
+    {
+      "epoch": 2.606145251396648,
+      "grad_norm": 0.595040725769985,
+      "learning_rate": 2.319912500776231e-07,
+      "loss": 0.3259,
+      "step": 5598
+    },
+    {
+      "epoch": 2.606610800744879,
+      "grad_norm": 0.6003292378811855,
+      "learning_rate": 2.314517501458105e-07,
+      "loss": 0.3166,
+      "step": 5599
+    },
+    {
+      "epoch": 2.60707635009311,
+      "grad_norm": 0.5953070939101932,
+      "learning_rate": 2.3091284780355189e-07,
+      "loss": 0.3241,
+      "step": 5600
+    },
+    {
+      "epoch": 2.607541899441341,
+      "grad_norm": 0.6309329000476577,
+      "learning_rate": 2.3037454319280784e-07,
+      "loss": 0.3311,
+      "step": 5601
+    },
+    {
+      "epoch": 2.6080074487895715,
+      "grad_norm": 0.5979505957846425,
+      "learning_rate": 2.2983683645537935e-07,
+      "loss": 0.3328,
+      "step": 5602
+    },
+    {
+      "epoch": 2.6084729981378025,
+      "grad_norm": 0.5893404258031923,
+      "learning_rate": 2.2929972773291231e-07,
+      "loss": 0.326,
+      "step": 5603
+    },
+    {
+      "epoch": 2.6089385474860336,
+      "grad_norm": 0.6054633669473507,
+      "learning_rate": 2.287632171668938e-07,
+      "loss": 0.3324,
+      "step": 5604
+    },
+    {
+      "epoch": 2.6094040968342647,
+      "grad_norm": 0.6386156519218986,
+      "learning_rate": 2.2822730489865274e-07,
+      "loss": 0.3237,
+      "step": 5605
+    },
+    {
+      "epoch": 2.6098696461824953,
+      "grad_norm": 0.632899024375951,
+      "learning_rate": 2.2769199106936146e-07,
+      "loss": 0.3352,
+      "step": 5606
+    },
+    {
+      "epoch": 2.6103351955307263,
+      "grad_norm": 0.6211905457634838,
+      "learning_rate": 2.2715727582003522e-07,
+      "loss": 0.3179,
+      "step": 5607
+    },
+    {
+      "epoch": 2.610800744878957,
+      "grad_norm": 0.6140621347808394,
+      "learning_rate": 2.266231592915294e-07,
+      "loss": 0.3541,
+      "step": 5608
+    },
+    {
+      "epoch": 2.611266294227188,
+      "grad_norm": 0.5865691760815078,
+      "learning_rate": 2.2608964162454365e-07,
+      "loss": 0.3123,
+      "step": 5609
+    },
+    {
+      "epoch": 2.611731843575419,
+      "grad_norm": 0.6109961904745058,
+      "learning_rate": 2.2555672295961978e-07,
+      "loss": 0.3229,
+      "step": 5610
+    },
+    {
+      "epoch": 2.61219739292365,
+      "grad_norm": 0.6443964111019647,
+      "learning_rate": 2.2502440343714044e-07,
+      "loss": 0.3271,
+      "step": 5611
+    },
+    {
+      "epoch": 2.612662942271881,
+      "grad_norm": 0.6082746991037891,
+      "learning_rate": 2.2449268319733097e-07,
+      "loss": 0.3342,
+      "step": 5612
+    },
+    {
+      "epoch": 2.613128491620112,
+      "grad_norm": 0.6190121048066412,
+      "learning_rate": 2.239615623802599e-07,
+      "loss": 0.3184,
+      "step": 5613
+    },
+    {
+      "epoch": 2.6135940409683425,
+      "grad_norm": 0.6281729938689463,
+      "learning_rate": 2.2343104112583696e-07,
+      "loss": 0.3297,
+      "step": 5614
+    },
+    {
+      "epoch": 2.6140595903165735,
+      "grad_norm": 0.627776301642722,
+      "learning_rate": 2.2290111957381344e-07,
+      "loss": 0.3073,
+      "step": 5615
+    },
+    {
+      "epoch": 2.6145251396648046,
+      "grad_norm": 0.5990714519925605,
+      "learning_rate": 2.2237179786378433e-07,
+      "loss": 0.3273,
+      "step": 5616
+    },
+    {
+      "epoch": 2.614990689013035,
+      "grad_norm": 0.639702436767832,
+      "learning_rate": 2.2184307613518418e-07,
+      "loss": 0.3173,
+      "step": 5617
+    },
+    {
+      "epoch": 2.6154562383612663,
+      "grad_norm": 0.6237481395481986,
+      "learning_rate": 2.213149545272919e-07,
+      "loss": 0.3249,
+      "step": 5618
+    },
+    {
+      "epoch": 2.6159217877094973,
+      "grad_norm": 0.6038874039655311,
+      "learning_rate": 2.2078743317922758e-07,
+      "loss": 0.3068,
+      "step": 5619
+    },
+    {
+      "epoch": 2.616387337057728,
+      "grad_norm": 0.5983725997516735,
+      "learning_rate": 2.202605122299517e-07,
+      "loss": 0.3235,
+      "step": 5620
+    },
+    {
+      "epoch": 2.616852886405959,
+      "grad_norm": 0.6008827309474076,
+      "learning_rate": 2.1973419181826827e-07,
+      "loss": 0.3298,
+      "step": 5621
+    },
+    {
+      "epoch": 2.61731843575419,
+      "grad_norm": 0.5941694138202219,
+      "learning_rate": 2.192084720828233e-07,
+      "loss": 0.3361,
+      "step": 5622
+    },
+    {
+      "epoch": 2.6177839851024207,
+      "grad_norm": 0.6035127863831394,
+      "learning_rate": 2.1868335316210298e-07,
+      "loss": 0.3222,
+      "step": 5623
+    },
+    {
+      "epoch": 2.618249534450652,
+      "grad_norm": 0.6085024148113326,
+      "learning_rate": 2.1815883519443665e-07,
+      "loss": 0.333,
+      "step": 5624
+    },
+    {
+      "epoch": 2.618715083798883,
+      "grad_norm": 0.5914306974414036,
+      "learning_rate": 2.1763491831799432e-07,
+      "loss": 0.3378,
+      "step": 5625
+    },
+    {
+      "epoch": 2.6191806331471135,
+      "grad_norm": 0.5850820193608975,
+      "learning_rate": 2.171116026707884e-07,
+      "loss": 0.3255,
+      "step": 5626
+    },
+    {
+      "epoch": 2.6196461824953445,
+      "grad_norm": 0.5750277117913153,
+      "learning_rate": 2.1658888839067333e-07,
+      "loss": 0.3384,
+      "step": 5627
+    },
+    {
+      "epoch": 2.6201117318435756,
+      "grad_norm": 0.6323772721346533,
+      "learning_rate": 2.1606677561534368e-07,
+      "loss": 0.3397,
+      "step": 5628
+    },
+    {
+      "epoch": 2.620577281191806,
+      "grad_norm": 0.5997636896567717,
+      "learning_rate": 2.1554526448233555e-07,
+      "loss": 0.3277,
+      "step": 5629
+    },
+    {
+      "epoch": 2.6210428305400373,
+      "grad_norm": 0.5866715430367306,
+      "learning_rate": 2.150243551290293e-07,
+      "loss": 0.304,
+      "step": 5630
+    },
+    {
+      "epoch": 2.621508379888268,
+      "grad_norm": 0.5744982362386514,
+      "learning_rate": 2.145040476926441e-07,
+      "loss": 0.3185,
+      "step": 5631
+    },
+    {
+      "epoch": 2.621973929236499,
+      "grad_norm": 0.5919359552656733,
+      "learning_rate": 2.1398434231024058e-07,
+      "loss": 0.3143,
+      "step": 5632
+    },
+    {
+      "epoch": 2.62243947858473,
+      "grad_norm": 0.5949889719417472,
+      "learning_rate": 2.1346523911872197e-07,
+      "loss": 0.3256,
+      "step": 5633
+    },
+    {
+      "epoch": 2.622905027932961,
+      "grad_norm": 0.5853526817562802,
+      "learning_rate": 2.129467382548328e-07,
+      "loss": 0.3217,
+      "step": 5634
+    },
+    {
+      "epoch": 2.6233705772811917,
+      "grad_norm": 0.6067129761270066,
+      "learning_rate": 2.12428839855158e-07,
+      "loss": 0.3183,
+      "step": 5635
+    },
+    {
+      "epoch": 2.623836126629423,
+      "grad_norm": 0.6156366280458206,
+      "learning_rate": 2.1191154405612479e-07,
+      "loss": 0.321,
+      "step": 5636
+    },
+    {
+      "epoch": 2.6243016759776534,
+      "grad_norm": 0.5951536575327437,
+      "learning_rate": 2.1139485099400058e-07,
+      "loss": 0.3287,
+      "step": 5637
+    },
+    {
+      "epoch": 2.6247672253258845,
+      "grad_norm": 0.6107968440946446,
+      "learning_rate": 2.1087876080489483e-07,
+      "loss": 0.3399,
+      "step": 5638
+    },
+    {
+      "epoch": 2.6252327746741155,
+      "grad_norm": 0.5897093045958863,
+      "learning_rate": 2.103632736247585e-07,
+      "loss": 0.315,
+      "step": 5639
+    },
+    {
+      "epoch": 2.6256983240223466,
+      "grad_norm": 0.5951702638823947,
+      "learning_rate": 2.098483895893824e-07,
+      "loss": 0.302,
+      "step": 5640
+    },
+    {
+      "epoch": 2.626163873370577,
+      "grad_norm": 0.5934072355698268,
+      "learning_rate": 2.0933410883439949e-07,
+      "loss": 0.3326,
+      "step": 5641
+    },
+    {
+      "epoch": 2.6266294227188083,
+      "grad_norm": 0.6270524795554961,
+      "learning_rate": 2.0882043149528408e-07,
+      "loss": 0.3232,
+      "step": 5642
+    },
+    {
+      "epoch": 2.627094972067039,
+      "grad_norm": 0.5994993892855633,
+      "learning_rate": 2.083073577073502e-07,
+      "loss": 0.3213,
+      "step": 5643
+    },
+    {
+      "epoch": 2.62756052141527,
+      "grad_norm": 0.5920133194762357,
+      "learning_rate": 2.077948876057545e-07,
+      "loss": 0.3219,
+      "step": 5644
+    },
+    {
+      "epoch": 2.628026070763501,
+      "grad_norm": 0.6057783734935599,
+      "learning_rate": 2.0728302132549283e-07,
+      "loss": 0.3228,
+      "step": 5645
+    },
+    {
+      "epoch": 2.628491620111732,
+      "grad_norm": 0.5846481117465795,
+      "learning_rate": 2.0677175900140345e-07,
+      "loss": 0.3175,
+      "step": 5646
+    },
+    {
+      "epoch": 2.6289571694599627,
+      "grad_norm": 0.6320937599145947,
+      "learning_rate": 2.062611007681653e-07,
+      "loss": 0.3248,
+      "step": 5647
+    },
+    {
+      "epoch": 2.629422718808194,
+      "grad_norm": 0.598803138505597,
+      "learning_rate": 2.0575104676029773e-07,
+      "loss": 0.3309,
+      "step": 5648
+    },
+    {
+      "epoch": 2.6298882681564244,
+      "grad_norm": 0.5918858044704718,
+      "learning_rate": 2.0524159711216019e-07,
+      "loss": 0.3085,
+      "step": 5649
+    },
+    {
+      "epoch": 2.6303538175046555,
+      "grad_norm": 0.6220855629903452,
+      "learning_rate": 2.0473275195795533e-07,
+      "loss": 0.328,
+      "step": 5650
+    },
+    {
+      "epoch": 2.6308193668528865,
+      "grad_norm": 0.5991942891954063,
+      "learning_rate": 2.042245114317243e-07,
+      "loss": 0.3399,
+      "step": 5651
+    },
+    {
+      "epoch": 2.631284916201117,
+      "grad_norm": 0.6107114039379065,
+      "learning_rate": 2.037168756673491e-07,
+      "loss": 0.3429,
+      "step": 5652
+    },
+    {
+      "epoch": 2.631750465549348,
+      "grad_norm": 0.6260672379465982,
+      "learning_rate": 2.032098447985545e-07,
+      "loss": 0.3185,
+      "step": 5653
+    },
+    {
+      "epoch": 2.6322160148975793,
+      "grad_norm": 0.6052567252457929,
+      "learning_rate": 2.027034189589036e-07,
+      "loss": 0.3206,
+      "step": 5654
+    },
+    {
+      "epoch": 2.63268156424581,
+      "grad_norm": 0.588608357828618,
+      "learning_rate": 2.0219759828180087e-07,
+      "loss": 0.3406,
+      "step": 5655
+    },
+    {
+      "epoch": 2.633147113594041,
+      "grad_norm": 0.610461306643181,
+      "learning_rate": 2.0169238290049186e-07,
+      "loss": 0.313,
+      "step": 5656
+    },
+    {
+      "epoch": 2.633612662942272,
+      "grad_norm": 0.583539233695051,
+      "learning_rate": 2.011877729480624e-07,
+      "loss": 0.3185,
+      "step": 5657
+    },
+    {
+      "epoch": 2.6340782122905027,
+      "grad_norm": 0.6325466168886574,
+      "learning_rate": 2.0068376855743837e-07,
+      "loss": 0.3199,
+      "step": 5658
+    },
+    {
+      "epoch": 2.6345437616387337,
+      "grad_norm": 0.6252824470264737,
+      "learning_rate": 2.0018036986138716e-07,
+      "loss": 0.3358,
+      "step": 5659
+    },
+    {
+      "epoch": 2.635009310986965,
+      "grad_norm": 0.6001506000477626,
+      "learning_rate": 1.9967757699251495e-07,
+      "loss": 0.3207,
+      "step": 5660
+    },
+    {
+      "epoch": 2.6354748603351954,
+      "grad_norm": 0.6125466133052153,
+      "learning_rate": 1.9917539008326992e-07,
+      "loss": 0.3241,
+      "step": 5661
+    },
+    {
+      "epoch": 2.6359404096834265,
+      "grad_norm": 0.6080521427114582,
+      "learning_rate": 1.986738092659407e-07,
+      "loss": 0.3347,
+      "step": 5662
+    },
+    {
+      "epoch": 2.6364059590316575,
+      "grad_norm": 0.6434374064351918,
+      "learning_rate": 1.9817283467265442e-07,
+      "loss": 0.3385,
+      "step": 5663
+    },
+    {
+      "epoch": 2.636871508379888,
+      "grad_norm": 0.6006132329944446,
+      "learning_rate": 1.976724664353799e-07,
+      "loss": 0.2985,
+      "step": 5664
+    },
+    {
+      "epoch": 2.637337057728119,
+      "grad_norm": 0.6122791194093903,
+      "learning_rate": 1.9717270468592704e-07,
+      "loss": 0.3161,
+      "step": 5665
+    },
+    {
+      "epoch": 2.63780260707635,
+      "grad_norm": 0.6286680985401097,
+      "learning_rate": 1.966735495559438e-07,
+      "loss": 0.3283,
+      "step": 5666
+    },
+    {
+      "epoch": 2.638268156424581,
+      "grad_norm": 0.6154500994265564,
+      "learning_rate": 1.9617500117692062e-07,
+      "loss": 0.3264,
+      "step": 5667
+    },
+    {
+      "epoch": 2.638733705772812,
+      "grad_norm": 0.6108795990571892,
+      "learning_rate": 1.9567705968018574e-07,
+      "loss": 0.3416,
+      "step": 5668
+    },
+    {
+      "epoch": 2.639199255121043,
+      "grad_norm": 0.6243299033054306,
+      "learning_rate": 1.9517972519690952e-07,
+      "loss": 0.3116,
+      "step": 5669
+    },
+    {
+      "epoch": 2.6396648044692737,
+      "grad_norm": 0.6035270135331039,
+      "learning_rate": 1.946829978581019e-07,
+      "loss": 0.3394,
+      "step": 5670
+    },
+    {
+      "epoch": 2.6401303538175047,
+      "grad_norm": 0.6198722717419539,
+      "learning_rate": 1.9418687779461238e-07,
+      "loss": 0.3302,
+      "step": 5671
+    },
+    {
+      "epoch": 2.6405959031657353,
+      "grad_norm": 0.6056869098077063,
+      "learning_rate": 1.9369136513712999e-07,
+      "loss": 0.3285,
+      "step": 5672
+    },
+    {
+      "epoch": 2.6410614525139664,
+      "grad_norm": 0.6332797151929713,
+      "learning_rate": 1.931964600161859e-07,
+      "loss": 0.3199,
+      "step": 5673
+    },
+    {
+      "epoch": 2.6415270018621975,
+      "grad_norm": 0.6137947069892316,
+      "learning_rate": 1.9270216256214975e-07,
+      "loss": 0.3228,
+      "step": 5674
+    },
+    {
+      "epoch": 2.6419925512104285,
+      "grad_norm": 0.616394040847838,
+      "learning_rate": 1.9220847290523008e-07,
+      "loss": 0.3283,
+      "step": 5675
+    },
+    {
+      "epoch": 2.642458100558659,
+      "grad_norm": 0.6061951168285797,
+      "learning_rate": 1.917153911754774e-07,
+      "loss": 0.296,
+      "step": 5676
+    },
+    {
+      "epoch": 2.64292364990689,
+      "grad_norm": 0.6071316979856622,
+      "learning_rate": 1.9122291750278132e-07,
+      "loss": 0.3155,
+      "step": 5677
+    },
+    {
+      "epoch": 2.643389199255121,
+      "grad_norm": 0.5926656715356454,
+      "learning_rate": 1.9073105201687008e-07,
+      "loss": 0.3207,
+      "step": 5678
+    },
+    {
+      "epoch": 2.643854748603352,
+      "grad_norm": 0.6632199900407713,
+      "learning_rate": 1.902397948473142e-07,
+      "loss": 0.3444,
+      "step": 5679
+    },
+    {
+      "epoch": 2.644320297951583,
+      "grad_norm": 0.6177547589847178,
+      "learning_rate": 1.8974914612352124e-07,
+      "loss": 0.3407,
+      "step": 5680
+    },
+    {
+      "epoch": 2.644785847299814,
+      "grad_norm": 0.593520209254225,
+      "learning_rate": 1.8925910597474008e-07,
+      "loss": 0.3297,
+      "step": 5681
+    },
+    {
+      "epoch": 2.6452513966480447,
+      "grad_norm": 0.6063006671998074,
+      "learning_rate": 1.887696745300596e-07,
+      "loss": 0.3333,
+      "step": 5682
+    },
+    {
+      "epoch": 2.6457169459962757,
+      "grad_norm": 0.5942904659380395,
+      "learning_rate": 1.8828085191840707e-07,
+      "loss": 0.2982,
+      "step": 5683
+    },
+    {
+      "epoch": 2.6461824953445063,
+      "grad_norm": 0.5979212621279267,
+      "learning_rate": 1.8779263826854966e-07,
+      "loss": 0.3138,
+      "step": 5684
+    },
+    {
+      "epoch": 2.6466480446927374,
+      "grad_norm": 0.6018117943949056,
+      "learning_rate": 1.8730503370909565e-07,
+      "loss": 0.3284,
+      "step": 5685
+    },
+    {
+      "epoch": 2.6471135940409685,
+      "grad_norm": 0.5805964985464891,
+      "learning_rate": 1.8681803836849033e-07,
+      "loss": 0.3128,
+      "step": 5686
+    },
+    {
+      "epoch": 2.6475791433891995,
+      "grad_norm": 0.5812759131983285,
+      "learning_rate": 1.8633165237502083e-07,
+      "loss": 0.3248,
+      "step": 5687
+    },
+    {
+      "epoch": 2.64804469273743,
+      "grad_norm": 0.6512388487263517,
+      "learning_rate": 1.8584587585681213e-07,
+      "loss": 0.3593,
+      "step": 5688
+    },
+    {
+      "epoch": 2.648510242085661,
+      "grad_norm": 0.5831872544944072,
+      "learning_rate": 1.8536070894182995e-07,
+      "loss": 0.319,
+      "step": 5689
+    },
+    {
+      "epoch": 2.648975791433892,
+      "grad_norm": 0.6202401042064473,
+      "learning_rate": 1.8487615175787787e-07,
+      "loss": 0.3386,
+      "step": 5690
+    },
+    {
+      "epoch": 2.649441340782123,
+      "grad_norm": 0.592733595451453,
+      "learning_rate": 1.8439220443260103e-07,
+      "loss": 0.3258,
+      "step": 5691
+    },
+    {
+      "epoch": 2.649906890130354,
+      "grad_norm": 0.5880597062200428,
+      "learning_rate": 1.839088670934813e-07,
+      "loss": 0.3181,
+      "step": 5692
+    },
+    {
+      "epoch": 2.6503724394785846,
+      "grad_norm": 0.6393023514051439,
+      "learning_rate": 1.8342613986784218e-07,
+      "loss": 0.3149,
+      "step": 5693
+    },
+    {
+      "epoch": 2.6508379888268156,
+      "grad_norm": 0.5860110787886033,
+      "learning_rate": 1.8294402288284524e-07,
+      "loss": 0.3199,
+      "step": 5694
+    },
+    {
+      "epoch": 2.6513035381750467,
+      "grad_norm": 0.5885079187650952,
+      "learning_rate": 1.824625162654914e-07,
+      "loss": 0.3362,
+      "step": 5695
+    },
+    {
+      "epoch": 2.6517690875232773,
+      "grad_norm": 0.6494628926120786,
+      "learning_rate": 1.819816201426211e-07,
+      "loss": 0.34,
+      "step": 5696
+    },
+    {
+      "epoch": 2.6522346368715084,
+      "grad_norm": 0.6108588572837769,
+      "learning_rate": 1.8150133464091418e-07,
+      "loss": 0.3231,
+      "step": 5697
+    },
+    {
+      "epoch": 2.6527001862197395,
+      "grad_norm": 0.6171120154386441,
+      "learning_rate": 1.8102165988688857e-07,
+      "loss": 0.3087,
+      "step": 5698
+    },
+    {
+      "epoch": 2.65316573556797,
+      "grad_norm": 0.6020768285057659,
+      "learning_rate": 1.805425960069021e-07,
+      "loss": 0.3284,
+      "step": 5699
+    },
+    {
+      "epoch": 2.653631284916201,
+      "grad_norm": 0.5869148204082689,
+      "learning_rate": 1.8006414312715244e-07,
+      "loss": 0.3201,
+      "step": 5700
+    },
+    {
+      "epoch": 2.6540968342644318,
+      "grad_norm": 0.5876900160686561,
+      "learning_rate": 1.7958630137367429e-07,
+      "loss": 0.3348,
+      "step": 5701
+    },
+    {
+      "epoch": 2.654562383612663,
+      "grad_norm": 0.6125387381764154,
+      "learning_rate": 1.7910907087234365e-07,
+      "loss": 0.3511,
+      "step": 5702
+    },
+    {
+      "epoch": 2.655027932960894,
+      "grad_norm": 0.6581176302362701,
+      "learning_rate": 1.786324517488733e-07,
+      "loss": 0.3367,
+      "step": 5703
+    },
+    {
+      "epoch": 2.655493482309125,
+      "grad_norm": 0.5940894795942426,
+      "learning_rate": 1.781564441288164e-07,
+      "loss": 0.3285,
+      "step": 5704
+    },
+    {
+      "epoch": 2.6559590316573556,
+      "grad_norm": 0.5961302861346508,
+      "learning_rate": 1.7768104813756543e-07,
+      "loss": 0.3271,
+      "step": 5705
+    },
+    {
+      "epoch": 2.6564245810055866,
+      "grad_norm": 0.6042459296259636,
+      "learning_rate": 1.7720626390035024e-07,
+      "loss": 0.3246,
+      "step": 5706
+    },
+    {
+      "epoch": 2.6568901303538173,
+      "grad_norm": 0.6029717000768288,
+      "learning_rate": 1.767320915422399e-07,
+      "loss": 0.3345,
+      "step": 5707
+    },
+    {
+      "epoch": 2.6573556797020483,
+      "grad_norm": 0.6486093410455878,
+      "learning_rate": 1.7625853118814374e-07,
+      "loss": 0.3313,
+      "step": 5708
+    },
+    {
+      "epoch": 2.6578212290502794,
+      "grad_norm": 0.5897796566838113,
+      "learning_rate": 1.7578558296280856e-07,
+      "loss": 0.3362,
+      "step": 5709
+    },
+    {
+      "epoch": 2.6582867783985105,
+      "grad_norm": 0.6054066089455657,
+      "learning_rate": 1.7531324699081949e-07,
+      "loss": 0.3261,
+      "step": 5710
+    },
+    {
+      "epoch": 2.658752327746741,
+      "grad_norm": 0.5888042336641004,
+      "learning_rate": 1.7484152339660138e-07,
+      "loss": 0.3207,
+      "step": 5711
+    },
+    {
+      "epoch": 2.659217877094972,
+      "grad_norm": 0.5906833304376098,
+      "learning_rate": 1.7437041230441793e-07,
+      "loss": 0.3249,
+      "step": 5712
+    },
+    {
+      "epoch": 2.6596834264432028,
+      "grad_norm": 0.6021865007221981,
+      "learning_rate": 1.7389991383837012e-07,
+      "loss": 0.327,
+      "step": 5713
+    },
+    {
+      "epoch": 2.660148975791434,
+      "grad_norm": 0.6185883890548549,
+      "learning_rate": 1.7343002812239968e-07,
+      "loss": 0.3145,
+      "step": 5714
+    },
+    {
+      "epoch": 2.660614525139665,
+      "grad_norm": 0.5949243119706747,
+      "learning_rate": 1.7296075528028423e-07,
+      "loss": 0.3126,
+      "step": 5715
+    },
+    {
+      "epoch": 2.661080074487896,
+      "grad_norm": 0.5828797545625177,
+      "learning_rate": 1.724920954356424e-07,
+      "loss": 0.3253,
+      "step": 5716
+    },
+    {
+      "epoch": 2.6615456238361266,
+      "grad_norm": 0.6359258705251964,
+      "learning_rate": 1.7202404871193013e-07,
+      "loss": 0.3259,
+      "step": 5717
+    },
+    {
+      "epoch": 2.6620111731843576,
+      "grad_norm": 0.58435955888646,
+      "learning_rate": 1.715566152324419e-07,
+      "loss": 0.3281,
+      "step": 5718
+    },
+    {
+      "epoch": 2.6624767225325883,
+      "grad_norm": 0.6266293424989876,
+      "learning_rate": 1.7108979512031055e-07,
+      "loss": 0.3261,
+      "step": 5719
+    },
+    {
+      "epoch": 2.6629422718808193,
+      "grad_norm": 0.5853712760678,
+      "learning_rate": 1.7062358849850858e-07,
+      "loss": 0.3214,
+      "step": 5720
+    },
+    {
+      "epoch": 2.6634078212290504,
+      "grad_norm": 0.5946391002450228,
+      "learning_rate": 1.7015799548984467e-07,
+      "loss": 0.3382,
+      "step": 5721
+    },
+    {
+      "epoch": 2.6638733705772815,
+      "grad_norm": 0.6708391350645141,
+      "learning_rate": 1.6969301621696798e-07,
+      "loss": 0.3236,
+      "step": 5722
+    },
+    {
+      "epoch": 2.664338919925512,
+      "grad_norm": 0.588536804496156,
+      "learning_rate": 1.6922865080236468e-07,
+      "loss": 0.3053,
+      "step": 5723
+    },
+    {
+      "epoch": 2.664804469273743,
+      "grad_norm": 0.5893322441723589,
+      "learning_rate": 1.6876489936835967e-07,
+      "loss": 0.323,
+      "step": 5724
+    },
+    {
+      "epoch": 2.6652700186219738,
+      "grad_norm": 0.605884236466675,
+      "learning_rate": 1.6830176203711635e-07,
+      "loss": 0.3328,
+      "step": 5725
+    },
+    {
+      "epoch": 2.665735567970205,
+      "grad_norm": 0.6210066197440725,
+      "learning_rate": 1.678392389306363e-07,
+      "loss": 0.3233,
+      "step": 5726
+    },
+    {
+      "epoch": 2.666201117318436,
+      "grad_norm": 0.6279993987695539,
+      "learning_rate": 1.673773301707579e-07,
+      "loss": 0.3339,
+      "step": 5727
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.621285576077583,
+      "learning_rate": 1.669160358791605e-07,
+      "loss": 0.3369,
+      "step": 5728
+    },
+    {
+      "epoch": 2.6671322160148976,
+      "grad_norm": 0.5849717684113009,
+      "learning_rate": 1.6645535617735937e-07,
+      "loss": 0.3087,
+      "step": 5729
+    },
+    {
+      "epoch": 2.6675977653631286,
+      "grad_norm": 0.6078911299731263,
+      "learning_rate": 1.659952911867077e-07,
+      "loss": 0.3162,
+      "step": 5730
+    },
+    {
+      "epoch": 2.6680633147113593,
+      "grad_norm": 0.591896648808487,
+      "learning_rate": 1.655358410283986e-07,
+      "loss": 0.3122,
+      "step": 5731
+    },
+    {
+      "epoch": 2.6685288640595903,
+      "grad_norm": 0.5964866453986414,
+      "learning_rate": 1.6507700582346185e-07,
+      "loss": 0.3211,
+      "step": 5732
+    },
+    {
+      "epoch": 2.6689944134078214,
+      "grad_norm": 0.5957389351652717,
+      "learning_rate": 1.6461878569276529e-07,
+      "loss": 0.3341,
+      "step": 5733
+    },
+    {
+      "epoch": 2.669459962756052,
+      "grad_norm": 0.6002591229305841,
+      "learning_rate": 1.6416118075701566e-07,
+      "loss": 0.3382,
+      "step": 5734
+    },
+    {
+      "epoch": 2.669925512104283,
+      "grad_norm": 0.6142733250432468,
+      "learning_rate": 1.6370419113675622e-07,
+      "loss": 0.3247,
+      "step": 5735
+    },
+    {
+      "epoch": 2.6703910614525137,
+      "grad_norm": 0.5914869650484056,
+      "learning_rate": 1.6324781695236907e-07,
+      "loss": 0.3146,
+      "step": 5736
+    },
+    {
+      "epoch": 2.6708566108007448,
+      "grad_norm": 0.6094965636349814,
+      "learning_rate": 1.627920583240747e-07,
+      "loss": 0.3184,
+      "step": 5737
+    },
+    {
+      "epoch": 2.671322160148976,
+      "grad_norm": 0.6439413366716291,
+      "learning_rate": 1.6233691537192954e-07,
+      "loss": 0.3378,
+      "step": 5738
+    },
+    {
+      "epoch": 2.671787709497207,
+      "grad_norm": 0.598343465211485,
+      "learning_rate": 1.618823882158302e-07,
+      "loss": 0.3273,
+      "step": 5739
+    },
+    {
+      "epoch": 2.6722532588454375,
+      "grad_norm": 0.5893738585104364,
+      "learning_rate": 1.6142847697550952e-07,
+      "loss": 0.3151,
+      "step": 5740
+    },
+    {
+      "epoch": 2.6727188081936686,
+      "grad_norm": 0.6134963953060955,
+      "learning_rate": 1.6097518177053821e-07,
+      "loss": 0.347,
+      "step": 5741
+    },
+    {
+      "epoch": 2.673184357541899,
+      "grad_norm": 0.6148688576089164,
+      "learning_rate": 1.6052250272032553e-07,
+      "loss": 0.3122,
+      "step": 5742
+    },
+    {
+      "epoch": 2.6736499068901303,
+      "grad_norm": 0.6331550390426439,
+      "learning_rate": 1.6007043994411775e-07,
+      "loss": 0.3314,
+      "step": 5743
+    },
+    {
+      "epoch": 2.6741154562383613,
+      "grad_norm": 0.6126761387006749,
+      "learning_rate": 1.5961899356099876e-07,
+      "loss": 0.3299,
+      "step": 5744
+    },
+    {
+      "epoch": 2.6745810055865924,
+      "grad_norm": 0.5854220288032421,
+      "learning_rate": 1.5916816368989062e-07,
+      "loss": 0.3261,
+      "step": 5745
+    },
+    {
+      "epoch": 2.675046554934823,
+      "grad_norm": 0.6173846536309635,
+      "learning_rate": 1.5871795044955203e-07,
+      "loss": 0.3284,
+      "step": 5746
+    },
+    {
+      "epoch": 2.675512104283054,
+      "grad_norm": 0.6138464554540063,
+      "learning_rate": 1.5826835395858026e-07,
+      "loss": 0.3231,
+      "step": 5747
+    },
+    {
+      "epoch": 2.6759776536312847,
+      "grad_norm": 0.5990682271465689,
+      "learning_rate": 1.5781937433541033e-07,
+      "loss": 0.3164,
+      "step": 5748
+    },
+    {
+      "epoch": 2.6764432029795158,
+      "grad_norm": 0.606124041324639,
+      "learning_rate": 1.573710116983132e-07,
+      "loss": 0.3101,
+      "step": 5749
+    },
+    {
+      "epoch": 2.676908752327747,
+      "grad_norm": 0.6013910963443029,
+      "learning_rate": 1.56923266165398e-07,
+      "loss": 0.33,
+      "step": 5750
+    },
+    {
+      "epoch": 2.677374301675978,
+      "grad_norm": 0.609533748587728,
+      "learning_rate": 1.5647613785461286e-07,
+      "loss": 0.3284,
+      "step": 5751
+    },
+    {
+      "epoch": 2.6778398510242085,
+      "grad_norm": 0.6114840409758413,
+      "learning_rate": 1.5602962688374107e-07,
+      "loss": 0.3235,
+      "step": 5752
+    },
+    {
+      "epoch": 2.6783054003724396,
+      "grad_norm": 0.632467216844771,
+      "learning_rate": 1.555837333704044e-07,
+      "loss": 0.3355,
+      "step": 5753
+    },
+    {
+      "epoch": 2.67877094972067,
+      "grad_norm": 0.6093523438312215,
+      "learning_rate": 1.5513845743206162e-07,
+      "loss": 0.3074,
+      "step": 5754
+    },
+    {
+      "epoch": 2.6792364990689013,
+      "grad_norm": 0.5934296752831227,
+      "learning_rate": 1.5469379918600953e-07,
+      "loss": 0.3161,
+      "step": 5755
+    },
+    {
+      "epoch": 2.6797020484171323,
+      "grad_norm": 0.6018229580746085,
+      "learning_rate": 1.5424975874938132e-07,
+      "loss": 0.3198,
+      "step": 5756
+    },
+    {
+      "epoch": 2.6801675977653634,
+      "grad_norm": 0.6392640641192983,
+      "learning_rate": 1.5380633623914786e-07,
+      "loss": 0.3231,
+      "step": 5757
+    },
+    {
+      "epoch": 2.680633147113594,
+      "grad_norm": 0.6168296823077685,
+      "learning_rate": 1.5336353177211716e-07,
+      "loss": 0.3231,
+      "step": 5758
+    },
+    {
+      "epoch": 2.681098696461825,
+      "grad_norm": 0.5968993511163421,
+      "learning_rate": 1.529213454649342e-07,
+      "loss": 0.3247,
+      "step": 5759
+    },
+    {
+      "epoch": 2.6815642458100557,
+      "grad_norm": 0.59315954107653,
+      "learning_rate": 1.5247977743408221e-07,
+      "loss": 0.3239,
+      "step": 5760
+    },
+    {
+      "epoch": 2.6820297951582868,
+      "grad_norm": 0.58568760594003,
+      "learning_rate": 1.5203882779587982e-07,
+      "loss": 0.3144,
+      "step": 5761
+    },
+    {
+      "epoch": 2.682495344506518,
+      "grad_norm": 0.6007308376451281,
+      "learning_rate": 1.515984966664838e-07,
+      "loss": 0.332,
+      "step": 5762
+    },
+    {
+      "epoch": 2.6829608938547485,
+      "grad_norm": 0.6189579771898427,
+      "learning_rate": 1.5115878416188835e-07,
+      "loss": 0.3286,
+      "step": 5763
+    },
+    {
+      "epoch": 2.6834264432029795,
+      "grad_norm": 0.6335950187875414,
+      "learning_rate": 1.5071969039792356e-07,
+      "loss": 0.3462,
+      "step": 5764
+    },
+    {
+      "epoch": 2.6838919925512106,
+      "grad_norm": 0.6321693622531869,
+      "learning_rate": 1.5028121549025804e-07,
+      "loss": 0.3477,
+      "step": 5765
+    },
+    {
+      "epoch": 2.684357541899441,
+      "grad_norm": 0.6345092085897381,
+      "learning_rate": 1.4984335955439522e-07,
+      "loss": 0.3286,
+      "step": 5766
+    },
+    {
+      "epoch": 2.6848230912476723,
+      "grad_norm": 0.5867365225079442,
+      "learning_rate": 1.4940612270567784e-07,
+      "loss": 0.3259,
+      "step": 5767
+    },
+    {
+      "epoch": 2.6852886405959033,
+      "grad_norm": 0.5967616809298655,
+      "learning_rate": 1.489695050592843e-07,
+      "loss": 0.3229,
+      "step": 5768
+    },
+    {
+      "epoch": 2.685754189944134,
+      "grad_norm": 0.627770698178148,
+      "learning_rate": 1.4853350673022986e-07,
+      "loss": 0.3227,
+      "step": 5769
+    },
+    {
+      "epoch": 2.686219739292365,
+      "grad_norm": 0.5997102047979532,
+      "learning_rate": 1.480981278333665e-07,
+      "loss": 0.3278,
+      "step": 5770
+    },
+    {
+      "epoch": 2.6866852886405956,
+      "grad_norm": 0.5870968603503613,
+      "learning_rate": 1.4766336848338392e-07,
+      "loss": 0.3361,
+      "step": 5771
+    },
+    {
+      "epoch": 2.6871508379888267,
+      "grad_norm": 0.6142001093138524,
+      "learning_rate": 1.472292287948082e-07,
+      "loss": 0.3294,
+      "step": 5772
+    },
+    {
+      "epoch": 2.6876163873370578,
+      "grad_norm": 0.6042935958945134,
+      "learning_rate": 1.4679570888200129e-07,
+      "loss": 0.3171,
+      "step": 5773
+    },
+    {
+      "epoch": 2.688081936685289,
+      "grad_norm": 0.6067643471830363,
+      "learning_rate": 1.4636280885916343e-07,
+      "loss": 0.3272,
+      "step": 5774
+    },
+    {
+      "epoch": 2.6885474860335195,
+      "grad_norm": 0.6291660194909346,
+      "learning_rate": 1.4593052884033066e-07,
+      "loss": 0.3333,
+      "step": 5775
+    },
+    {
+      "epoch": 2.6890130353817505,
+      "grad_norm": 0.6324958197567048,
+      "learning_rate": 1.454988689393752e-07,
+      "loss": 0.334,
+      "step": 5776
+    },
+    {
+      "epoch": 2.689478584729981,
+      "grad_norm": 0.6464334835284813,
+      "learning_rate": 1.4506782927000724e-07,
+      "loss": 0.3392,
+      "step": 5777
+    },
+    {
+      "epoch": 2.689944134078212,
+      "grad_norm": 0.622154413231814,
+      "learning_rate": 1.446374099457723e-07,
+      "loss": 0.3284,
+      "step": 5778
+    },
+    {
+      "epoch": 2.6904096834264433,
+      "grad_norm": 0.6056847913991866,
+      "learning_rate": 1.442076110800536e-07,
+      "loss": 0.3219,
+      "step": 5779
+    },
+    {
+      "epoch": 2.6908752327746743,
+      "grad_norm": 0.609666333669141,
+      "learning_rate": 1.4377843278607028e-07,
+      "loss": 0.3265,
+      "step": 5780
+    },
+    {
+      "epoch": 2.691340782122905,
+      "grad_norm": 0.5769553104353934,
+      "learning_rate": 1.4334987517687805e-07,
+      "loss": 0.3016,
+      "step": 5781
+    },
+    {
+      "epoch": 2.691806331471136,
+      "grad_norm": 0.6403037429396912,
+      "learning_rate": 1.4292193836536848e-07,
+      "loss": 0.3267,
+      "step": 5782
+    },
+    {
+      "epoch": 2.6922718808193666,
+      "grad_norm": 0.5901876807592947,
+      "learning_rate": 1.4249462246427148e-07,
+      "loss": 0.3261,
+      "step": 5783
+    },
+    {
+      "epoch": 2.6927374301675977,
+      "grad_norm": 0.6351345621393601,
+      "learning_rate": 1.4206792758615167e-07,
+      "loss": 0.3188,
+      "step": 5784
+    },
+    {
+      "epoch": 2.6932029795158288,
+      "grad_norm": 0.6090115568598772,
+      "learning_rate": 1.4164185384341002e-07,
+      "loss": 0.3312,
+      "step": 5785
+    },
+    {
+      "epoch": 2.69366852886406,
+      "grad_norm": 0.6177610791213106,
+      "learning_rate": 1.4121640134828585e-07,
+      "loss": 0.321,
+      "step": 5786
+    },
+    {
+      "epoch": 2.6941340782122905,
+      "grad_norm": 0.6051505047599184,
+      "learning_rate": 1.407915702128526e-07,
+      "loss": 0.3262,
+      "step": 5787
+    },
+    {
+      "epoch": 2.6945996275605215,
+      "grad_norm": 0.5981695090316088,
+      "learning_rate": 1.403673605490205e-07,
+      "loss": 0.3358,
+      "step": 5788
+    },
+    {
+      "epoch": 2.695065176908752,
+      "grad_norm": 0.6738392213318226,
+      "learning_rate": 1.399437724685371e-07,
+      "loss": 0.3478,
+      "step": 5789
+    },
+    {
+      "epoch": 2.695530726256983,
+      "grad_norm": 0.5937829199840712,
+      "learning_rate": 1.3952080608298584e-07,
+      "loss": 0.3426,
+      "step": 5790
+    },
+    {
+      "epoch": 2.6959962756052143,
+      "grad_norm": 0.5989717208239846,
+      "learning_rate": 1.3909846150378547e-07,
+      "loss": 0.3114,
+      "step": 5791
+    },
+    {
+      "epoch": 2.6964618249534453,
+      "grad_norm": 0.5877727825000771,
+      "learning_rate": 1.3867673884219196e-07,
+      "loss": 0.3227,
+      "step": 5792
+    },
+    {
+      "epoch": 2.696927374301676,
+      "grad_norm": 0.6225501717850701,
+      "learning_rate": 1.382556382092967e-07,
+      "loss": 0.3311,
+      "step": 5793
+    },
+    {
+      "epoch": 2.697392923649907,
+      "grad_norm": 0.5947983745973153,
+      "learning_rate": 1.3783515971602796e-07,
+      "loss": 0.3291,
+      "step": 5794
+    },
+    {
+      "epoch": 2.6978584729981376,
+      "grad_norm": 0.6204526237214023,
+      "learning_rate": 1.3741530347314986e-07,
+      "loss": 0.331,
+      "step": 5795
+    },
+    {
+      "epoch": 2.6983240223463687,
+      "grad_norm": 0.5987240748456032,
+      "learning_rate": 1.3699606959126198e-07,
+      "loss": 0.3262,
+      "step": 5796
+    },
+    {
+      "epoch": 2.6987895716945998,
+      "grad_norm": 0.580876724190615,
+      "learning_rate": 1.3657745818080105e-07,
+      "loss": 0.3309,
+      "step": 5797
+    },
+    {
+      "epoch": 2.6992551210428304,
+      "grad_norm": 0.6274594673825864,
+      "learning_rate": 1.3615946935203934e-07,
+      "loss": 0.3312,
+      "step": 5798
+    },
+    {
+      "epoch": 2.6997206703910615,
+      "grad_norm": 0.645302207907262,
+      "learning_rate": 1.3574210321508464e-07,
+      "loss": 0.3197,
+      "step": 5799
+    },
+    {
+      "epoch": 2.7001862197392925,
+      "grad_norm": 0.6028894272658154,
+      "learning_rate": 1.353253598798815e-07,
+      "loss": 0.3146,
+      "step": 5800
+    },
+    {
+      "epoch": 2.700651769087523,
+      "grad_norm": 0.643665908268502,
+      "learning_rate": 1.3490923945620933e-07,
+      "loss": 0.3469,
+      "step": 5801
+    },
+    {
+      "epoch": 2.701117318435754,
+      "grad_norm": 0.6379812898457787,
+      "learning_rate": 1.344937420536846e-07,
+      "loss": 0.3172,
+      "step": 5802
+    },
+    {
+      "epoch": 2.7015828677839853,
+      "grad_norm": 0.6315352448510828,
+      "learning_rate": 1.3407886778175978e-07,
+      "loss": 0.3614,
+      "step": 5803
+    },
+    {
+      "epoch": 2.702048417132216,
+      "grad_norm": 0.6263358288891683,
+      "learning_rate": 1.3366461674972187e-07,
+      "loss": 0.3369,
+      "step": 5804
+    },
+    {
+      "epoch": 2.702513966480447,
+      "grad_norm": 0.5912137814867398,
+      "learning_rate": 1.3325098906669386e-07,
+      "loss": 0.3398,
+      "step": 5805
+    },
+    {
+      "epoch": 2.7029795158286776,
+      "grad_norm": 0.5842759637568516,
+      "learning_rate": 1.3283798484163667e-07,
+      "loss": 0.3111,
+      "step": 5806
+    },
+    {
+      "epoch": 2.7034450651769086,
+      "grad_norm": 0.6090755612358697,
+      "learning_rate": 1.3242560418334467e-07,
+      "loss": 0.3288,
+      "step": 5807
+    },
+    {
+      "epoch": 2.7039106145251397,
+      "grad_norm": 0.6115493794710046,
+      "learning_rate": 1.320138472004484e-07,
+      "loss": 0.3452,
+      "step": 5808
+    },
+    {
+      "epoch": 2.7043761638733708,
+      "grad_norm": 0.603219052199392,
+      "learning_rate": 1.3160271400141477e-07,
+      "loss": 0.3112,
+      "step": 5809
+    },
+    {
+      "epoch": 2.7048417132216014,
+      "grad_norm": 0.5969787334601554,
+      "learning_rate": 1.311922046945463e-07,
+      "loss": 0.314,
+      "step": 5810
+    },
+    {
+      "epoch": 2.7053072625698324,
+      "grad_norm": 0.5833616202765521,
+      "learning_rate": 1.3078231938798035e-07,
+      "loss": 0.3063,
+      "step": 5811
+    },
+    {
+      "epoch": 2.705772811918063,
+      "grad_norm": 0.5981459475591916,
+      "learning_rate": 1.3037305818969088e-07,
+      "loss": 0.3199,
+      "step": 5812
+    },
+    {
+      "epoch": 2.706238361266294,
+      "grad_norm": 0.6009828386960842,
+      "learning_rate": 1.299644212074866e-07,
+      "loss": 0.3123,
+      "step": 5813
+    },
+    {
+      "epoch": 2.706703910614525,
+      "grad_norm": 0.589365780836706,
+      "learning_rate": 1.2955640854901253e-07,
+      "loss": 0.3255,
+      "step": 5814
+    },
+    {
+      "epoch": 2.7071694599627563,
+      "grad_norm": 0.6150798511038769,
+      "learning_rate": 1.2914902032174936e-07,
+      "loss": 0.3216,
+      "step": 5815
+    },
+    {
+      "epoch": 2.707635009310987,
+      "grad_norm": 0.6052707232298346,
+      "learning_rate": 1.2874225663301181e-07,
+      "loss": 0.3301,
+      "step": 5816
+    },
+    {
+      "epoch": 2.708100558659218,
+      "grad_norm": 0.6121468900246346,
+      "learning_rate": 1.283361175899517e-07,
+      "loss": 0.3263,
+      "step": 5817
+    },
+    {
+      "epoch": 2.7085661080074486,
+      "grad_norm": 0.7625390883938817,
+      "learning_rate": 1.2793060329955587e-07,
+      "loss": 0.3217,
+      "step": 5818
+    },
+    {
+      "epoch": 2.7090316573556796,
+      "grad_norm": 0.5912630541744706,
+      "learning_rate": 1.2752571386864588e-07,
+      "loss": 0.3242,
+      "step": 5819
+    },
+    {
+      "epoch": 2.7094972067039107,
+      "grad_norm": 0.5901364229610823,
+      "learning_rate": 1.2712144940388e-07,
+      "loss": 0.3125,
+      "step": 5820
+    },
+    {
+      "epoch": 2.7099627560521418,
+      "grad_norm": 0.5803998168480895,
+      "learning_rate": 1.2671781001175026e-07,
+      "loss": 0.3349,
+      "step": 5821
+    },
+    {
+      "epoch": 2.7104283054003724,
+      "grad_norm": 0.5993037804288914,
+      "learning_rate": 1.2631479579858523e-07,
+      "loss": 0.3268,
+      "step": 5822
+    },
+    {
+      "epoch": 2.7108938547486034,
+      "grad_norm": 0.6122043262247423,
+      "learning_rate": 1.259124068705489e-07,
+      "loss": 0.3376,
+      "step": 5823
+    },
+    {
+      "epoch": 2.711359404096834,
+      "grad_norm": 0.617388040066498,
+      "learning_rate": 1.2551064333363976e-07,
+      "loss": 0.3394,
+      "step": 5824
+    },
+    {
+      "epoch": 2.711824953445065,
+      "grad_norm": 0.5978811960818744,
+      "learning_rate": 1.251095052936907e-07,
+      "loss": 0.3478,
+      "step": 5825
+    },
+    {
+      "epoch": 2.712290502793296,
+      "grad_norm": 0.6337983597476811,
+      "learning_rate": 1.247089928563733e-07,
+      "loss": 0.3332,
+      "step": 5826
+    },
+    {
+      "epoch": 2.7127560521415273,
+      "grad_norm": 0.6107016771811655,
+      "learning_rate": 1.243091061271906e-07,
+      "loss": 0.3381,
+      "step": 5827
+    },
+    {
+      "epoch": 2.713221601489758,
+      "grad_norm": 0.5835604554722768,
+      "learning_rate": 1.2390984521148176e-07,
+      "loss": 0.3338,
+      "step": 5828
+    },
+    {
+      "epoch": 2.713687150837989,
+      "grad_norm": 0.6287534840808375,
+      "learning_rate": 1.2351121021442313e-07,
+      "loss": 0.3357,
+      "step": 5829
+    },
+    {
+      "epoch": 2.7141527001862196,
+      "grad_norm": 0.5742734242455008,
+      "learning_rate": 1.231132012410241e-07,
+      "loss": 0.3105,
+      "step": 5830
+    },
+    {
+      "epoch": 2.7146182495344506,
+      "grad_norm": 0.6214697725300509,
+      "learning_rate": 1.2271581839612883e-07,
+      "loss": 0.3256,
+      "step": 5831
+    },
+    {
+      "epoch": 2.7150837988826817,
+      "grad_norm": 0.5862365858561062,
+      "learning_rate": 1.2231906178441832e-07,
+      "loss": 0.3145,
+      "step": 5832
+    },
+    {
+      "epoch": 2.7155493482309123,
+      "grad_norm": 0.5743274888924736,
+      "learning_rate": 1.2192293151040756e-07,
+      "loss": 0.3288,
+      "step": 5833
+    },
+    {
+      "epoch": 2.7160148975791434,
+      "grad_norm": 0.6763230870033794,
+      "learning_rate": 1.2152742767844643e-07,
+      "loss": 0.3093,
+      "step": 5834
+    },
+    {
+      "epoch": 2.7164804469273744,
+      "grad_norm": 0.58400892257606,
+      "learning_rate": 1.2113255039272049e-07,
+      "loss": 0.3211,
+      "step": 5835
+    },
+    {
+      "epoch": 2.716945996275605,
+      "grad_norm": 0.6137723056410048,
+      "learning_rate": 1.20738299757249e-07,
+      "loss": 0.3239,
+      "step": 5836
+    },
+    {
+      "epoch": 2.717411545623836,
+      "grad_norm": 0.5879929442930938,
+      "learning_rate": 1.203446758758875e-07,
+      "loss": 0.3232,
+      "step": 5837
+    },
+    {
+      "epoch": 2.717877094972067,
+      "grad_norm": 0.6236279971917141,
+      "learning_rate": 1.1995167885232638e-07,
+      "loss": 0.3124,
+      "step": 5838
+    },
+    {
+      "epoch": 2.718342644320298,
+      "grad_norm": 0.5873951224710523,
+      "learning_rate": 1.1955930879008925e-07,
+      "loss": 0.3133,
+      "step": 5839
+    },
+    {
+      "epoch": 2.718808193668529,
+      "grad_norm": 0.6073690091882904,
+      "learning_rate": 1.1916756579253647e-07,
+      "loss": 0.322,
+      "step": 5840
+    },
+    {
+      "epoch": 2.7192737430167595,
+      "grad_norm": 0.6380317353414764,
+      "learning_rate": 1.187764499628627e-07,
+      "loss": 0.3303,
+      "step": 5841
+    },
+    {
+      "epoch": 2.7197392923649906,
+      "grad_norm": 0.6423480269798808,
+      "learning_rate": 1.1838596140409637e-07,
+      "loss": 0.3422,
+      "step": 5842
+    },
+    {
+      "epoch": 2.7202048417132216,
+      "grad_norm": 0.5943024994566208,
+      "learning_rate": 1.1799610021910213e-07,
+      "loss": 0.3181,
+      "step": 5843
+    },
+    {
+      "epoch": 2.7206703910614527,
+      "grad_norm": 0.6411730200191215,
+      "learning_rate": 1.1760686651057813e-07,
+      "loss": 0.3287,
+      "step": 5844
+    },
+    {
+      "epoch": 2.7211359404096833,
+      "grad_norm": 0.6313799001246223,
+      "learning_rate": 1.1721826038105815e-07,
+      "loss": 0.3355,
+      "step": 5845
+    },
+    {
+      "epoch": 2.7216014897579144,
+      "grad_norm": 0.5918411658131908,
+      "learning_rate": 1.1683028193291034e-07,
+      "loss": 0.334,
+      "step": 5846
+    },
+    {
+      "epoch": 2.722067039106145,
+      "grad_norm": 0.5827488099259479,
+      "learning_rate": 1.1644293126833738e-07,
+      "loss": 0.3236,
+      "step": 5847
+    },
+    {
+      "epoch": 2.722532588454376,
+      "grad_norm": 0.6041224093211329,
+      "learning_rate": 1.1605620848937627e-07,
+      "loss": 0.3131,
+      "step": 5848
+    },
+    {
+      "epoch": 2.722998137802607,
+      "grad_norm": 0.6288778228121814,
+      "learning_rate": 1.1567011369789938e-07,
+      "loss": 0.334,
+      "step": 5849
+    },
+    {
+      "epoch": 2.723463687150838,
+      "grad_norm": 0.6179898696917154,
+      "learning_rate": 1.1528464699561314e-07,
+      "loss": 0.33,
+      "step": 5850
+    },
+    {
+      "epoch": 2.723929236499069,
+      "grad_norm": 0.6107880659384448,
+      "learning_rate": 1.1489980848405857e-07,
+      "loss": 0.3252,
+      "step": 5851
+    },
+    {
+      "epoch": 2.7243947858473,
+      "grad_norm": 0.6097965321515207,
+      "learning_rate": 1.145155982646115e-07,
+      "loss": 0.326,
+      "step": 5852
+    },
+    {
+      "epoch": 2.7248603351955305,
+      "grad_norm": 0.599156961218786,
+      "learning_rate": 1.141320164384821e-07,
+      "loss": 0.3303,
+      "step": 5853
+    },
+    {
+      "epoch": 2.7253258845437616,
+      "grad_norm": 0.5924098599858253,
+      "learning_rate": 1.1374906310671452e-07,
+      "loss": 0.315,
+      "step": 5854
+    },
+    {
+      "epoch": 2.7257914338919926,
+      "grad_norm": 0.5899890660682731,
+      "learning_rate": 1.1336673837018807e-07,
+      "loss": 0.3289,
+      "step": 5855
+    },
+    {
+      "epoch": 2.7262569832402237,
+      "grad_norm": 0.5887458124273124,
+      "learning_rate": 1.129850423296161e-07,
+      "loss": 0.2979,
+      "step": 5856
+    },
+    {
+      "epoch": 2.7267225325884543,
+      "grad_norm": 0.6072175891279096,
+      "learning_rate": 1.1260397508554649e-07,
+      "loss": 0.3269,
+      "step": 5857
+    },
+    {
+      "epoch": 2.7271880819366854,
+      "grad_norm": 0.6165804224543253,
+      "learning_rate": 1.1222353673836145e-07,
+      "loss": 0.3445,
+      "step": 5858
+    },
+    {
+      "epoch": 2.727653631284916,
+      "grad_norm": 0.6017851147659675,
+      "learning_rate": 1.1184372738827748e-07,
+      "loss": 0.3236,
+      "step": 5859
+    },
+    {
+      "epoch": 2.728119180633147,
+      "grad_norm": 0.6285757589690687,
+      "learning_rate": 1.1146454713534482e-07,
+      "loss": 0.3283,
+      "step": 5860
+    },
+    {
+      "epoch": 2.728584729981378,
+      "grad_norm": 0.6155163695320787,
+      "learning_rate": 1.110859960794497e-07,
+      "loss": 0.3088,
+      "step": 5861
+    },
+    {
+      "epoch": 2.729050279329609,
+      "grad_norm": 0.6233743518520072,
+      "learning_rate": 1.1070807432031066e-07,
+      "loss": 0.3183,
+      "step": 5862
+    },
+    {
+      "epoch": 2.72951582867784,
+      "grad_norm": 0.6098417644968241,
+      "learning_rate": 1.1033078195748142e-07,
+      "loss": 0.3273,
+      "step": 5863
+    },
+    {
+      "epoch": 2.729981378026071,
+      "grad_norm": 0.5979149824576125,
+      "learning_rate": 1.0995411909034942e-07,
+      "loss": 0.3239,
+      "step": 5864
+    },
+    {
+      "epoch": 2.7304469273743015,
+      "grad_norm": 0.605444050505194,
+      "learning_rate": 1.0957808581813778e-07,
+      "loss": 0.311,
+      "step": 5865
+    },
+    {
+      "epoch": 2.7309124767225326,
+      "grad_norm": 0.6126375331455409,
+      "learning_rate": 1.0920268223990116e-07,
+      "loss": 0.3208,
+      "step": 5866
+    },
+    {
+      "epoch": 2.7313780260707636,
+      "grad_norm": 0.6147270426825273,
+      "learning_rate": 1.08827908454531e-07,
+      "loss": 0.3214,
+      "step": 5867
+    },
+    {
+      "epoch": 2.7318435754189943,
+      "grad_norm": 0.6079671137198157,
+      "learning_rate": 1.0845376456075057e-07,
+      "loss": 0.3329,
+      "step": 5868
+    },
+    {
+      "epoch": 2.7323091247672253,
+      "grad_norm": 0.7157432797104291,
+      "learning_rate": 1.080802506571188e-07,
+      "loss": 0.3386,
+      "step": 5869
+    },
+    {
+      "epoch": 2.7327746741154564,
+      "grad_norm": 0.6045332510603376,
+      "learning_rate": 1.0770736684202838e-07,
+      "loss": 0.3221,
+      "step": 5870
+    },
+    {
+      "epoch": 2.733240223463687,
+      "grad_norm": 0.6147751742006097,
+      "learning_rate": 1.0733511321370543e-07,
+      "loss": 0.3279,
+      "step": 5871
+    },
+    {
+      "epoch": 2.733705772811918,
+      "grad_norm": 0.6394086576011224,
+      "learning_rate": 1.0696348987021015e-07,
+      "loss": 0.3259,
+      "step": 5872
+    },
+    {
+      "epoch": 2.734171322160149,
+      "grad_norm": 0.5984773142106975,
+      "learning_rate": 1.0659249690943785e-07,
+      "loss": 0.3431,
+      "step": 5873
+    },
+    {
+      "epoch": 2.7346368715083798,
+      "grad_norm": 0.6922049897900661,
+      "learning_rate": 1.0622213442911588e-07,
+      "loss": 0.3334,
+      "step": 5874
+    },
+    {
+      "epoch": 2.735102420856611,
+      "grad_norm": 0.6049063577571943,
+      "learning_rate": 1.0585240252680734e-07,
+      "loss": 0.3163,
+      "step": 5875
+    },
+    {
+      "epoch": 2.7355679702048414,
+      "grad_norm": 0.6070837310586003,
+      "learning_rate": 1.0548330129990791e-07,
+      "loss": 0.3292,
+      "step": 5876
+    },
+    {
+      "epoch": 2.7360335195530725,
+      "grad_norm": 0.5998030798598749,
+      "learning_rate": 1.0511483084564788e-07,
+      "loss": 0.3226,
+      "step": 5877
+    },
+    {
+      "epoch": 2.7364990689013036,
+      "grad_norm": 0.6180748131552442,
+      "learning_rate": 1.04746991261091e-07,
+      "loss": 0.3347,
+      "step": 5878
+    },
+    {
+      "epoch": 2.7369646182495346,
+      "grad_norm": 0.6378753671335775,
+      "learning_rate": 1.0437978264313475e-07,
+      "loss": 0.3161,
+      "step": 5879
+    },
+    {
+      "epoch": 2.7374301675977653,
+      "grad_norm": 0.5880308128681491,
+      "learning_rate": 1.0401320508851093e-07,
+      "loss": 0.3129,
+      "step": 5880
+    },
+    {
+      "epoch": 2.7378957169459963,
+      "grad_norm": 0.6110140469785954,
+      "learning_rate": 1.0364725869378506e-07,
+      "loss": 0.3235,
+      "step": 5881
+    },
+    {
+      "epoch": 2.738361266294227,
+      "grad_norm": 0.5770825513585421,
+      "learning_rate": 1.0328194355535587e-07,
+      "loss": 0.3199,
+      "step": 5882
+    },
+    {
+      "epoch": 2.738826815642458,
+      "grad_norm": 0.5981491708918717,
+      "learning_rate": 1.02917259769455e-07,
+      "loss": 0.323,
+      "step": 5883
+    },
+    {
+      "epoch": 2.739292364990689,
+      "grad_norm": 0.6174427463753153,
+      "learning_rate": 1.0255320743215059e-07,
+      "loss": 0.337,
+      "step": 5884
+    },
+    {
+      "epoch": 2.73975791433892,
+      "grad_norm": 0.5970684826295677,
+      "learning_rate": 1.0218978663934204e-07,
+      "loss": 0.303,
+      "step": 5885
+    },
+    {
+      "epoch": 2.7402234636871508,
+      "grad_norm": 0.6113750349663768,
+      "learning_rate": 1.0182699748676222e-07,
+      "loss": 0.3214,
+      "step": 5886
+    },
+    {
+      "epoch": 2.740689013035382,
+      "grad_norm": 0.5954144722107543,
+      "learning_rate": 1.0146484006997942e-07,
+      "loss": 0.3406,
+      "step": 5887
+    },
+    {
+      "epoch": 2.7411545623836124,
+      "grad_norm": 0.5960998287526277,
+      "learning_rate": 1.0110331448439426e-07,
+      "loss": 0.3314,
+      "step": 5888
+    },
+    {
+      "epoch": 2.7416201117318435,
+      "grad_norm": 0.618120014442954,
+      "learning_rate": 1.0074242082524083e-07,
+      "loss": 0.323,
+      "step": 5889
+    },
+    {
+      "epoch": 2.7420856610800746,
+      "grad_norm": 0.6081197026438436,
+      "learning_rate": 1.0038215918758753e-07,
+      "loss": 0.3299,
+      "step": 5890
+    },
+    {
+      "epoch": 2.7425512104283056,
+      "grad_norm": 0.5886666987923861,
+      "learning_rate": 1.0002252966633542e-07,
+      "loss": 0.3245,
+      "step": 5891
+    },
+    {
+      "epoch": 2.7430167597765363,
+      "grad_norm": 0.5898773080185125,
+      "learning_rate": 9.966353235621978e-08,
+      "loss": 0.3233,
+      "step": 5892
+    },
+    {
+      "epoch": 2.7434823091247673,
+      "grad_norm": 0.6034675858030902,
+      "learning_rate": 9.930516735180946e-08,
+      "loss": 0.3283,
+      "step": 5893
+    },
+    {
+      "epoch": 2.743947858472998,
+      "grad_norm": 0.6414986716726986,
+      "learning_rate": 9.894743474750528e-08,
+      "loss": 0.3265,
+      "step": 5894
+    },
+    {
+      "epoch": 2.744413407821229,
+      "grad_norm": 0.6072026027445852,
+      "learning_rate": 9.859033463754303e-08,
+      "loss": 0.331,
+      "step": 5895
+    },
+    {
+      "epoch": 2.74487895716946,
+      "grad_norm": 0.6096880675195459,
+      "learning_rate": 9.823386711599187e-08,
+      "loss": 0.3167,
+      "step": 5896
+    },
+    {
+      "epoch": 2.745344506517691,
+      "grad_norm": 0.599336431662216,
+      "learning_rate": 9.78780322767528e-08,
+      "loss": 0.3259,
+      "step": 5897
+    },
+    {
+      "epoch": 2.7458100558659218,
+      "grad_norm": 0.6078994260633472,
+      "learning_rate": 9.752283021356223e-08,
+      "loss": 0.3255,
+      "step": 5898
+    },
+    {
+      "epoch": 2.746275605214153,
+      "grad_norm": 0.6154586788506672,
+      "learning_rate": 9.71682610199881e-08,
+      "loss": 0.3128,
+      "step": 5899
+    },
+    {
+      "epoch": 2.7467411545623834,
+      "grad_norm": 0.6799945015669174,
+      "learning_rate": 9.681432478943265e-08,
+      "loss": 0.3026,
+      "step": 5900
+    },
+    {
+      "epoch": 2.7472067039106145,
+      "grad_norm": 0.6346827723527921,
+      "learning_rate": 9.646102161513099e-08,
+      "loss": 0.31,
+      "step": 5901
+    },
+    {
+      "epoch": 2.7476722532588456,
+      "grad_norm": 0.6103278317208763,
+      "learning_rate": 9.610835159015203e-08,
+      "loss": 0.3332,
+      "step": 5902
+    },
+    {
+      "epoch": 2.748137802607076,
+      "grad_norm": 0.5863798262253149,
+      "learning_rate": 9.575631480739644e-08,
+      "loss": 0.3247,
+      "step": 5903
+    },
+    {
+      "epoch": 2.7486033519553073,
+      "grad_norm": 0.6411038976954885,
+      "learning_rate": 9.540491135960006e-08,
+      "loss": 0.3337,
+      "step": 5904
+    },
+    {
+      "epoch": 2.7490689013035383,
+      "grad_norm": 0.6034329949674115,
+      "learning_rate": 9.505414133933078e-08,
+      "loss": 0.3298,
+      "step": 5905
+    },
+    {
+      "epoch": 2.749534450651769,
+      "grad_norm": 0.6235891693453788,
+      "learning_rate": 9.470400483898911e-08,
+      "loss": 0.3247,
+      "step": 5906
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.6225612526915314,
+      "learning_rate": 9.435450195080964e-08,
+      "loss": 0.3288,
+      "step": 5907
+    },
+    {
+      "epoch": 2.750465549348231,
+      "grad_norm": 0.5935138082112145,
+      "learning_rate": 9.400563276686037e-08,
+      "loss": 0.3412,
+      "step": 5908
+    },
+    {
+      "epoch": 2.7509310986964617,
+      "grad_norm": 0.6034082026890627,
+      "learning_rate": 9.365739737904084e-08,
+      "loss": 0.3316,
+      "step": 5909
+    },
+    {
+      "epoch": 2.7513966480446927,
+      "grad_norm": 0.6002907328487825,
+      "learning_rate": 9.330979587908518e-08,
+      "loss": 0.3213,
+      "step": 5910
+    },
+    {
+      "epoch": 2.751862197392924,
+      "grad_norm": 0.6046870510357157,
+      "learning_rate": 9.29628283585593e-08,
+      "loss": 0.3155,
+      "step": 5911
+    },
+    {
+      "epoch": 2.7523277467411544,
+      "grad_norm": 0.6063660415845514,
+      "learning_rate": 9.261649490886316e-08,
+      "loss": 0.3311,
+      "step": 5912
+    },
+    {
+      "epoch": 2.7527932960893855,
+      "grad_norm": 0.5922715826598132,
+      "learning_rate": 9.227079562122932e-08,
+      "loss": 0.3212,
+      "step": 5913
+    },
+    {
+      "epoch": 2.7532588454376166,
+      "grad_norm": 0.5763854427958689,
+      "learning_rate": 9.192573058672244e-08,
+      "loss": 0.3333,
+      "step": 5914
+    },
+    {
+      "epoch": 2.753724394785847,
+      "grad_norm": 0.5888567922756572,
+      "learning_rate": 9.158129989624175e-08,
+      "loss": 0.3178,
+      "step": 5915
+    },
+    {
+      "epoch": 2.7541899441340782,
+      "grad_norm": 0.6075134176894417,
+      "learning_rate": 9.123750364051803e-08,
+      "loss": 0.3292,
+      "step": 5916
+    },
+    {
+      "epoch": 2.754655493482309,
+      "grad_norm": 0.619017084588481,
+      "learning_rate": 9.089434191011547e-08,
+      "loss": 0.3135,
+      "step": 5917
+    },
+    {
+      "epoch": 2.75512104283054,
+      "grad_norm": 0.6298709507928966,
+      "learning_rate": 9.055181479543096e-08,
+      "loss": 0.3307,
+      "step": 5918
+    },
+    {
+      "epoch": 2.755586592178771,
+      "grad_norm": 0.6305675638148129,
+      "learning_rate": 9.02099223866948e-08,
+      "loss": 0.3234,
+      "step": 5919
+    },
+    {
+      "epoch": 2.756052141527002,
+      "grad_norm": 0.5988073975450157,
+      "learning_rate": 8.986866477396855e-08,
+      "loss": 0.3149,
+      "step": 5920
+    },
+    {
+      "epoch": 2.7565176908752327,
+      "grad_norm": 0.6080110415303656,
+      "learning_rate": 8.952804204714866e-08,
+      "loss": 0.3436,
+      "step": 5921
+    },
+    {
+      "epoch": 2.7569832402234637,
+      "grad_norm": 0.6191965500674832,
+      "learning_rate": 8.91880542959625e-08,
+      "loss": 0.3295,
+      "step": 5922
+    },
+    {
+      "epoch": 2.7574487895716944,
+      "grad_norm": 0.6086607262004464,
+      "learning_rate": 8.884870160997094e-08,
+      "loss": 0.3268,
+      "step": 5923
+    },
+    {
+      "epoch": 2.7579143389199254,
+      "grad_norm": 0.5831869726520879,
+      "learning_rate": 8.850998407856831e-08,
+      "loss": 0.314,
+      "step": 5924
+    },
+    {
+      "epoch": 2.7583798882681565,
+      "grad_norm": 0.6236215047691515,
+      "learning_rate": 8.817190179098045e-08,
+      "loss": 0.3297,
+      "step": 5925
+    },
+    {
+      "epoch": 2.7588454376163876,
+      "grad_norm": 0.5956117792243061,
+      "learning_rate": 8.783445483626557e-08,
+      "loss": 0.3237,
+      "step": 5926
+    },
+    {
+      "epoch": 2.759310986964618,
+      "grad_norm": 0.614964022735397,
+      "learning_rate": 8.749764330331617e-08,
+      "loss": 0.3126,
+      "step": 5927
+    },
+    {
+      "epoch": 2.7597765363128492,
+      "grad_norm": 0.5943241078789745,
+      "learning_rate": 8.7161467280856e-08,
+      "loss": 0.3401,
+      "step": 5928
+    },
+    {
+      "epoch": 2.76024208566108,
+      "grad_norm": 0.5894147864751849,
+      "learning_rate": 8.682592685744201e-08,
+      "loss": 0.3128,
+      "step": 5929
+    },
+    {
+      "epoch": 2.760707635009311,
+      "grad_norm": 0.5921215482568158,
+      "learning_rate": 8.64910221214632e-08,
+      "loss": 0.3232,
+      "step": 5930
+    },
+    {
+      "epoch": 2.761173184357542,
+      "grad_norm": 0.6162998105233549,
+      "learning_rate": 8.615675316114181e-08,
+      "loss": 0.3235,
+      "step": 5931
+    },
+    {
+      "epoch": 2.761638733705773,
+      "grad_norm": 0.6248040270056854,
+      "learning_rate": 8.582312006453208e-08,
+      "loss": 0.3501,
+      "step": 5932
+    },
+    {
+      "epoch": 2.7621042830540037,
+      "grad_norm": 0.6054293732242089,
+      "learning_rate": 8.549012291952097e-08,
+      "loss": 0.3285,
+      "step": 5933
+    },
+    {
+      "epoch": 2.7625698324022347,
+      "grad_norm": 0.6012346172506109,
+      "learning_rate": 8.515776181382745e-08,
+      "loss": 0.3194,
+      "step": 5934
+    },
+    {
+      "epoch": 2.7630353817504654,
+      "grad_norm": 0.6301131207624514,
+      "learning_rate": 8.482603683500374e-08,
+      "loss": 0.3296,
+      "step": 5935
+    },
+    {
+      "epoch": 2.7635009310986964,
+      "grad_norm": 0.6107072057780983,
+      "learning_rate": 8.449494807043435e-08,
+      "loss": 0.3215,
+      "step": 5936
+    },
+    {
+      "epoch": 2.7639664804469275,
+      "grad_norm": 0.586654740018445,
+      "learning_rate": 8.416449560733564e-08,
+      "loss": 0.3197,
+      "step": 5937
+    },
+    {
+      "epoch": 2.7644320297951586,
+      "grad_norm": 0.6029633017667284,
+      "learning_rate": 8.383467953275603e-08,
+      "loss": 0.3184,
+      "step": 5938
+    },
+    {
+      "epoch": 2.764897579143389,
+      "grad_norm": 0.5934008587851529,
+      "learning_rate": 8.350549993357799e-08,
+      "loss": 0.3133,
+      "step": 5939
+    },
+    {
+      "epoch": 2.7653631284916202,
+      "grad_norm": 0.6097445695965537,
+      "learning_rate": 8.317695689651462e-08,
+      "loss": 0.3323,
+      "step": 5940
+    },
+    {
+      "epoch": 2.765828677839851,
+      "grad_norm": 0.6047688029356473,
+      "learning_rate": 8.284905050811198e-08,
+      "loss": 0.3386,
+      "step": 5941
+    },
+    {
+      "epoch": 2.766294227188082,
+      "grad_norm": 0.6209754117582508,
+      "learning_rate": 8.252178085474849e-08,
+      "loss": 0.3148,
+      "step": 5942
+    },
+    {
+      "epoch": 2.766759776536313,
+      "grad_norm": 0.6068884127220863,
+      "learning_rate": 8.219514802263518e-08,
+      "loss": 0.3278,
+      "step": 5943
+    },
+    {
+      "epoch": 2.7672253258845436,
+      "grad_norm": 0.6078393669021663,
+      "learning_rate": 8.186915209781404e-08,
+      "loss": 0.3407,
+      "step": 5944
+    },
+    {
+      "epoch": 2.7676908752327747,
+      "grad_norm": 0.6080810865882171,
+      "learning_rate": 8.154379316616113e-08,
+      "loss": 0.3233,
+      "step": 5945
+    },
+    {
+      "epoch": 2.7681564245810057,
+      "grad_norm": 0.6164222838329129,
+      "learning_rate": 8.121907131338285e-08,
+      "loss": 0.3239,
+      "step": 5946
+    },
+    {
+      "epoch": 2.7686219739292364,
+      "grad_norm": 0.6076815584453561,
+      "learning_rate": 8.089498662501915e-08,
+      "loss": 0.3358,
+      "step": 5947
+    },
+    {
+      "epoch": 2.7690875232774674,
+      "grad_norm": 0.5833610958168249,
+      "learning_rate": 8.057153918644173e-08,
+      "loss": 0.3176,
+      "step": 5948
+    },
+    {
+      "epoch": 2.7695530726256985,
+      "grad_norm": 0.5954072628376339,
+      "learning_rate": 8.02487290828538e-08,
+      "loss": 0.3417,
+      "step": 5949
+    },
+    {
+      "epoch": 2.770018621973929,
+      "grad_norm": 0.6104323395711461,
+      "learning_rate": 7.992655639929153e-08,
+      "loss": 0.3291,
+      "step": 5950
+    },
+    {
+      "epoch": 2.77048417132216,
+      "grad_norm": 0.6178248755358674,
+      "learning_rate": 7.960502122062314e-08,
+      "loss": 0.3095,
+      "step": 5951
+    },
+    {
+      "epoch": 2.770949720670391,
+      "grad_norm": 0.6209761279607184,
+      "learning_rate": 7.928412363154835e-08,
+      "loss": 0.3121,
+      "step": 5952
+    },
+    {
+      "epoch": 2.771415270018622,
+      "grad_norm": 0.6218474110104736,
+      "learning_rate": 7.896386371659931e-08,
+      "loss": 0.3383,
+      "step": 5953
+    },
+    {
+      "epoch": 2.771880819366853,
+      "grad_norm": 0.6013664784242675,
+      "learning_rate": 7.864424156013988e-08,
+      "loss": 0.3226,
+      "step": 5954
+    },
+    {
+      "epoch": 2.772346368715084,
+      "grad_norm": 0.6153785963997678,
+      "learning_rate": 7.832525724636636e-08,
+      "loss": 0.3216,
+      "step": 5955
+    },
+    {
+      "epoch": 2.7728119180633146,
+      "grad_norm": 0.6211236001054429,
+      "learning_rate": 7.800691085930706e-08,
+      "loss": 0.3524,
+      "step": 5956
+    },
+    {
+      "epoch": 2.7732774674115457,
+      "grad_norm": 0.6516244155195013,
+      "learning_rate": 7.768920248282158e-08,
+      "loss": 0.3326,
+      "step": 5957
+    },
+    {
+      "epoch": 2.7737430167597763,
+      "grad_norm": 0.595446002812711,
+      "learning_rate": 7.73721322006013e-08,
+      "loss": 0.3324,
+      "step": 5958
+    },
+    {
+      "epoch": 2.7742085661080074,
+      "grad_norm": 0.593340265278182,
+      "learning_rate": 7.705570009617164e-08,
+      "loss": 0.313,
+      "step": 5959
+    },
+    {
+      "epoch": 2.7746741154562384,
+      "grad_norm": 0.5941456299710235,
+      "learning_rate": 7.673990625288702e-08,
+      "loss": 0.3162,
+      "step": 5960
+    },
+    {
+      "epoch": 2.7751396648044695,
+      "grad_norm": 0.5988724497111197,
+      "learning_rate": 7.642475075393507e-08,
+      "loss": 0.3267,
+      "step": 5961
+    },
+    {
+      "epoch": 2.7756052141527,
+      "grad_norm": 0.6128418738350998,
+      "learning_rate": 7.611023368233633e-08,
+      "loss": 0.3177,
+      "step": 5962
+    },
+    {
+      "epoch": 2.776070763500931,
+      "grad_norm": 0.5762411991611789,
+      "learning_rate": 7.57963551209412e-08,
+      "loss": 0.3299,
+      "step": 5963
+    },
+    {
+      "epoch": 2.776536312849162,
+      "grad_norm": 0.5961669735704458,
+      "learning_rate": 7.548311515243268e-08,
+      "loss": 0.338,
+      "step": 5964
+    },
+    {
+      "epoch": 2.777001862197393,
+      "grad_norm": 0.5919187454850177,
+      "learning_rate": 7.517051385932616e-08,
+      "loss": 0.3167,
+      "step": 5965
+    },
+    {
+      "epoch": 2.777467411545624,
+      "grad_norm": 0.5919077086875811,
+      "learning_rate": 7.485855132396802e-08,
+      "loss": 0.3275,
+      "step": 5966
+    },
+    {
+      "epoch": 2.777932960893855,
+      "grad_norm": 0.6076568336332242,
+      "learning_rate": 7.454722762853612e-08,
+      "loss": 0.3247,
+      "step": 5967
+    },
+    {
+      "epoch": 2.7783985102420856,
+      "grad_norm": 0.6499355851813138,
+      "learning_rate": 7.423654285504151e-08,
+      "loss": 0.3235,
+      "step": 5968
+    },
+    {
+      "epoch": 2.7788640595903167,
+      "grad_norm": 0.6046860456815658,
+      "learning_rate": 7.392649708532512e-08,
+      "loss": 0.3241,
+      "step": 5969
+    },
+    {
+      "epoch": 2.7793296089385473,
+      "grad_norm": 0.6208319550081947,
+      "learning_rate": 7.361709040106052e-08,
+      "loss": 0.3508,
+      "step": 5970
+    },
+    {
+      "epoch": 2.7797951582867784,
+      "grad_norm": 0.6320419724914424,
+      "learning_rate": 7.330832288375333e-08,
+      "loss": 0.3428,
+      "step": 5971
+    },
+    {
+      "epoch": 2.7802607076350094,
+      "grad_norm": 0.6182749133073878,
+      "learning_rate": 7.300019461473962e-08,
+      "loss": 0.334,
+      "step": 5972
+    },
+    {
+      "epoch": 2.7807262569832405,
+      "grad_norm": 0.6120733960496159,
+      "learning_rate": 7.269270567518777e-08,
+      "loss": 0.346,
+      "step": 5973
+    },
+    {
+      "epoch": 2.781191806331471,
+      "grad_norm": 0.6194107433827439,
+      "learning_rate": 7.238585614609827e-08,
+      "loss": 0.3221,
+      "step": 5974
+    },
+    {
+      "epoch": 2.781657355679702,
+      "grad_norm": 0.5890060643045532,
+      "learning_rate": 7.207964610830176e-08,
+      "loss": 0.3462,
+      "step": 5975
+    },
+    {
+      "epoch": 2.782122905027933,
+      "grad_norm": 0.6286780937609607,
+      "learning_rate": 7.177407564246202e-08,
+      "loss": 0.3334,
+      "step": 5976
+    },
+    {
+      "epoch": 2.782588454376164,
+      "grad_norm": 0.6038836798894625,
+      "learning_rate": 7.146914482907303e-08,
+      "loss": 0.3173,
+      "step": 5977
+    },
+    {
+      "epoch": 2.783054003724395,
+      "grad_norm": 0.6191738136625373,
+      "learning_rate": 7.116485374846077e-08,
+      "loss": 0.3257,
+      "step": 5978
+    },
+    {
+      "epoch": 2.7835195530726256,
+      "grad_norm": 0.6142461316632567,
+      "learning_rate": 7.086120248078337e-08,
+      "loss": 0.3394,
+      "step": 5979
+    },
+    {
+      "epoch": 2.7839851024208566,
+      "grad_norm": 0.5993948510261314,
+      "learning_rate": 7.055819110602963e-08,
+      "loss": 0.3344,
+      "step": 5980
+    },
+    {
+      "epoch": 2.7844506517690877,
+      "grad_norm": 0.5997088401803526,
+      "learning_rate": 7.025581970401874e-08,
+      "loss": 0.313,
+      "step": 5981
+    },
+    {
+      "epoch": 2.7849162011173183,
+      "grad_norm": 0.6111143337515327,
+      "learning_rate": 6.995408835440448e-08,
+      "loss": 0.3172,
+      "step": 5982
+    },
+    {
+      "epoch": 2.7853817504655494,
+      "grad_norm": 0.6021441284441199,
+      "learning_rate": 6.965299713666884e-08,
+      "loss": 0.3293,
+      "step": 5983
+    },
+    {
+      "epoch": 2.7858472998137804,
+      "grad_norm": 0.5944406557502845,
+      "learning_rate": 6.935254613012644e-08,
+      "loss": 0.3074,
+      "step": 5984
+    },
+    {
+      "epoch": 2.786312849162011,
+      "grad_norm": 0.6302889232224453,
+      "learning_rate": 6.905273541392338e-08,
+      "loss": 0.3288,
+      "step": 5985
+    },
+    {
+      "epoch": 2.786778398510242,
+      "grad_norm": 0.576675499144573,
+      "learning_rate": 6.875356506703706e-08,
+      "loss": 0.3276,
+      "step": 5986
+    },
+    {
+      "epoch": 2.7872439478584727,
+      "grad_norm": 0.5878881004677315,
+      "learning_rate": 6.845503516827611e-08,
+      "loss": 0.3128,
+      "step": 5987
+    },
+    {
+      "epoch": 2.787709497206704,
+      "grad_norm": 0.5879247358408541,
+      "learning_rate": 6.815714579628014e-08,
+      "loss": 0.3359,
+      "step": 5988
+    },
+    {
+      "epoch": 2.788175046554935,
+      "grad_norm": 0.6108225284405973,
+      "learning_rate": 6.785989702952051e-08,
+      "loss": 0.3188,
+      "step": 5989
+    },
+    {
+      "epoch": 2.788640595903166,
+      "grad_norm": 0.638745531083409,
+      "learning_rate": 6.756328894629933e-08,
+      "loss": 0.323,
+      "step": 5990
+    },
+    {
+      "epoch": 2.7891061452513966,
+      "grad_norm": 0.5893639128898048,
+      "learning_rate": 6.726732162475075e-08,
+      "loss": 0.339,
+      "step": 5991
+    },
+    {
+      "epoch": 2.7895716945996276,
+      "grad_norm": 0.6007914299491104,
+      "learning_rate": 6.697199514283909e-08,
+      "loss": 0.3251,
+      "step": 5992
+    },
+    {
+      "epoch": 2.7900372439478582,
+      "grad_norm": 0.6109661221227124,
+      "learning_rate": 6.667730957836044e-08,
+      "loss": 0.3165,
+      "step": 5993
+    },
+    {
+      "epoch": 2.7905027932960893,
+      "grad_norm": 0.5884595720774142,
+      "learning_rate": 6.638326500894243e-08,
+      "loss": 0.3344,
+      "step": 5994
+    },
+    {
+      "epoch": 2.7909683426443204,
+      "grad_norm": 0.638151832015529,
+      "learning_rate": 6.608986151204283e-08,
+      "loss": 0.3433,
+      "step": 5995
+    },
+    {
+      "epoch": 2.7914338919925514,
+      "grad_norm": 0.58956006106296,
+      "learning_rate": 6.579709916495147e-08,
+      "loss": 0.3243,
+      "step": 5996
+    },
+    {
+      "epoch": 2.791899441340782,
+      "grad_norm": 0.6104197190068597,
+      "learning_rate": 6.550497804478862e-08,
+      "loss": 0.3201,
+      "step": 5997
+    },
+    {
+      "epoch": 2.792364990689013,
+      "grad_norm": 0.6039444113722613,
+      "learning_rate": 6.521349822850603e-08,
+      "loss": 0.3269,
+      "step": 5998
+    },
+    {
+      "epoch": 2.7928305400372437,
+      "grad_norm": 0.6057940883669567,
+      "learning_rate": 6.492265979288675e-08,
+      "loss": 0.3162,
+      "step": 5999
+    },
+    {
+      "epoch": 2.793296089385475,
+      "grad_norm": 0.6305230362698377,
+      "learning_rate": 6.463246281454422e-08,
+      "loss": 0.3333,
+      "step": 6000
+    },
+    {
+      "epoch": 2.793761638733706,
+      "grad_norm": 0.6136739828366798,
+      "learning_rate": 6.434290736992282e-08,
+      "loss": 0.3074,
+      "step": 6001
+    },
+    {
+      "epoch": 2.794227188081937,
+      "grad_norm": 0.6397329261438596,
+      "learning_rate": 6.405399353529935e-08,
+      "loss": 0.3363,
+      "step": 6002
+    },
+    {
+      "epoch": 2.7946927374301676,
+      "grad_norm": 0.5844095086444984,
+      "learning_rate": 6.376572138677983e-08,
+      "loss": 0.3283,
+      "step": 6003
+    },
+    {
+      "epoch": 2.7951582867783986,
+      "grad_norm": 0.606382666820531,
+      "learning_rate": 6.347809100030217e-08,
+      "loss": 0.3264,
+      "step": 6004
+    },
+    {
+      "epoch": 2.7956238361266292,
+      "grad_norm": 0.5998623495020922,
+      "learning_rate": 6.31911024516349e-08,
+      "loss": 0.3264,
+      "step": 6005
+    },
+    {
+      "epoch": 2.7960893854748603,
+      "grad_norm": 0.6142899587284811,
+      "learning_rate": 6.290475581637811e-08,
+      "loss": 0.3368,
+      "step": 6006
+    },
+    {
+      "epoch": 2.7965549348230914,
+      "grad_norm": 0.6260468379274652,
+      "learning_rate": 6.261905116996175e-08,
+      "loss": 0.3407,
+      "step": 6007
+    },
+    {
+      "epoch": 2.7970204841713224,
+      "grad_norm": 0.5849598267638191,
+      "learning_rate": 6.233398858764756e-08,
+      "loss": 0.3246,
+      "step": 6008
+    },
+    {
+      "epoch": 2.797486033519553,
+      "grad_norm": 0.6129491270806132,
+      "learning_rate": 6.204956814452767e-08,
+      "loss": 0.3206,
+      "step": 6009
+    },
+    {
+      "epoch": 2.797951582867784,
+      "grad_norm": 0.6172363719783726,
+      "learning_rate": 6.176578991552524e-08,
+      "loss": 0.3165,
+      "step": 6010
+    },
+    {
+      "epoch": 2.7984171322160147,
+      "grad_norm": 0.6041697384856116,
+      "learning_rate": 6.148265397539433e-08,
+      "loss": 0.3322,
+      "step": 6011
+    },
+    {
+      "epoch": 2.798882681564246,
+      "grad_norm": 0.6007725051913387,
+      "learning_rate": 6.120016039871946e-08,
+      "loss": 0.3267,
+      "step": 6012
+    },
+    {
+      "epoch": 2.799348230912477,
+      "grad_norm": 0.5954002956939817,
+      "learning_rate": 6.09183092599161e-08,
+      "loss": 0.3211,
+      "step": 6013
+    },
+    {
+      "epoch": 2.7998137802607075,
+      "grad_norm": 0.5942530796523451,
+      "learning_rate": 6.063710063323097e-08,
+      "loss": 0.3187,
+      "step": 6014
+    },
+    {
+      "epoch": 2.8002793296089385,
+      "grad_norm": 0.6092425565831441,
+      "learning_rate": 6.035653459274093e-08,
+      "loss": 0.332,
+      "step": 6015
+    },
+    {
+      "epoch": 2.8007448789571696,
+      "grad_norm": 0.6397860252276714,
+      "learning_rate": 6.007661121235298e-08,
+      "loss": 0.3301,
+      "step": 6016
+    },
+    {
+      "epoch": 2.8012104283054002,
+      "grad_norm": 0.5956553907010829,
+      "learning_rate": 5.979733056580673e-08,
+      "loss": 0.3084,
+      "step": 6017
+    },
+    {
+      "epoch": 2.8016759776536313,
+      "grad_norm": 0.5951531335880497,
+      "learning_rate": 5.951869272667082e-08,
+      "loss": 0.3188,
+      "step": 6018
+    },
+    {
+      "epoch": 2.8021415270018624,
+      "grad_norm": 0.6230227937442158,
+      "learning_rate": 5.924069776834518e-08,
+      "loss": 0.3278,
+      "step": 6019
+    },
+    {
+      "epoch": 2.802607076350093,
+      "grad_norm": 0.5942255004536174,
+      "learning_rate": 5.896334576405982e-08,
+      "loss": 0.319,
+      "step": 6020
+    },
+    {
+      "epoch": 2.803072625698324,
+      "grad_norm": 0.5711056506348215,
+      "learning_rate": 5.8686636786876864e-08,
+      "loss": 0.3203,
+      "step": 6021
+    },
+    {
+      "epoch": 2.8035381750465547,
+      "grad_norm": 0.6068055284378973,
+      "learning_rate": 5.841057090968716e-08,
+      "loss": 0.3134,
+      "step": 6022
+    },
+    {
+      "epoch": 2.8040037243947857,
+      "grad_norm": 0.5984173606082712,
+      "learning_rate": 5.813514820521338e-08,
+      "loss": 0.3165,
+      "step": 6023
+    },
+    {
+      "epoch": 2.804469273743017,
+      "grad_norm": 0.59120560489511,
+      "learning_rate": 5.786036874600831e-08,
+      "loss": 0.3253,
+      "step": 6024
+    },
+    {
+      "epoch": 2.804934823091248,
+      "grad_norm": 0.630082036484834,
+      "learning_rate": 5.758623260445517e-08,
+      "loss": 0.3397,
+      "step": 6025
+    },
+    {
+      "epoch": 2.8054003724394785,
+      "grad_norm": 0.6174900546638644,
+      "learning_rate": 5.7312739852768695e-08,
+      "loss": 0.3205,
+      "step": 6026
+    },
+    {
+      "epoch": 2.8058659217877095,
+      "grad_norm": 0.6005371468744308,
+      "learning_rate": 5.703989056299264e-08,
+      "loss": 0.3258,
+      "step": 6027
+    },
+    {
+      "epoch": 2.80633147113594,
+      "grad_norm": 0.6030574512629766,
+      "learning_rate": 5.676768480700201e-08,
+      "loss": 0.3299,
+      "step": 6028
+    },
+    {
+      "epoch": 2.8067970204841712,
+      "grad_norm": 0.5996814251798464,
+      "learning_rate": 5.649612265650306e-08,
+      "loss": 0.3118,
+      "step": 6029
+    },
+    {
+      "epoch": 2.8072625698324023,
+      "grad_norm": 0.6314778776887471,
+      "learning_rate": 5.622520418303079e-08,
+      "loss": 0.3233,
+      "step": 6030
+    },
+    {
+      "epoch": 2.8077281191806334,
+      "grad_norm": 0.6098226798408849,
+      "learning_rate": 5.5954929457952003e-08,
+      "loss": 0.3112,
+      "step": 6031
+    },
+    {
+      "epoch": 2.808193668528864,
+      "grad_norm": 0.58165517276173,
+      "learning_rate": 5.568529855246335e-08,
+      "loss": 0.3056,
+      "step": 6032
+    },
+    {
+      "epoch": 2.808659217877095,
+      "grad_norm": 0.5995574205556823,
+      "learning_rate": 5.541631153759219e-08,
+      "loss": 0.3041,
+      "step": 6033
+    },
+    {
+      "epoch": 2.8091247672253257,
+      "grad_norm": 0.6027565585167963,
+      "learning_rate": 5.514796848419601e-08,
+      "loss": 0.3111,
+      "step": 6034
+    },
+    {
+      "epoch": 2.8095903165735567,
+      "grad_norm": 0.6195609747743773,
+      "learning_rate": 5.4880269462962985e-08,
+      "loss": 0.3406,
+      "step": 6035
+    },
+    {
+      "epoch": 2.810055865921788,
+      "grad_norm": 0.5963971163755891,
+      "learning_rate": 5.461321454441032e-08,
+      "loss": 0.3335,
+      "step": 6036
+    },
+    {
+      "epoch": 2.810521415270019,
+      "grad_norm": 0.6017640015766141,
+      "learning_rate": 5.4346803798888405e-08,
+      "loss": 0.3273,
+      "step": 6037
+    },
+    {
+      "epoch": 2.8109869646182495,
+      "grad_norm": 0.5777610067617461,
+      "learning_rate": 5.408103729657499e-08,
+      "loss": 0.31,
+      "step": 6038
+    },
+    {
+      "epoch": 2.8114525139664805,
+      "grad_norm": 0.6014273878305331,
+      "learning_rate": 5.3815915107479355e-08,
+      "loss": 0.327,
+      "step": 6039
+    },
+    {
+      "epoch": 2.811918063314711,
+      "grad_norm": 0.6128754531885559,
+      "learning_rate": 5.355143730144119e-08,
+      "loss": 0.3115,
+      "step": 6040
+    },
+    {
+      "epoch": 2.8123836126629422,
+      "grad_norm": 0.6083939062845921,
+      "learning_rate": 5.3287603948130316e-08,
+      "loss": 0.358,
+      "step": 6041
+    },
+    {
+      "epoch": 2.8128491620111733,
+      "grad_norm": 0.6130241632525515,
+      "learning_rate": 5.30244151170467e-08,
+      "loss": 0.3332,
+      "step": 6042
+    },
+    {
+      "epoch": 2.8133147113594044,
+      "grad_norm": 0.5799923994786297,
+      "learning_rate": 5.276187087752044e-08,
+      "loss": 0.3322,
+      "step": 6043
+    },
+    {
+      "epoch": 2.813780260707635,
+      "grad_norm": 0.6091696227902578,
+      "learning_rate": 5.249997129871176e-08,
+      "loss": 0.3401,
+      "step": 6044
+    },
+    {
+      "epoch": 2.814245810055866,
+      "grad_norm": 0.5851771682190731,
+      "learning_rate": 5.223871644961159e-08,
+      "loss": 0.3243,
+      "step": 6045
+    },
+    {
+      "epoch": 2.8147113594040967,
+      "grad_norm": 0.6187544859497014,
+      "learning_rate": 5.197810639904044e-08,
+      "loss": 0.3202,
+      "step": 6046
+    },
+    {
+      "epoch": 2.8151769087523277,
+      "grad_norm": 0.5770851282725672,
+      "learning_rate": 5.171814121564922e-08,
+      "loss": 0.326,
+      "step": 6047
+    },
+    {
+      "epoch": 2.815642458100559,
+      "grad_norm": 0.5963967942891479,
+      "learning_rate": 5.145882096791899e-08,
+      "loss": 0.3112,
+      "step": 6048
+    },
+    {
+      "epoch": 2.8161080074487894,
+      "grad_norm": 0.6169905262725428,
+      "learning_rate": 5.120014572416093e-08,
+      "loss": 0.3298,
+      "step": 6049
+    },
+    {
+      "epoch": 2.8165735567970205,
+      "grad_norm": 0.6113226514735913,
+      "learning_rate": 5.0942115552516105e-08,
+      "loss": 0.3327,
+      "step": 6050
+    },
+    {
+      "epoch": 2.8170391061452515,
+      "grad_norm": 0.6037377355859884,
+      "learning_rate": 5.068473052095596e-08,
+      "loss": 0.3252,
+      "step": 6051
+    },
+    {
+      "epoch": 2.817504655493482,
+      "grad_norm": 0.6267784947394361,
+      "learning_rate": 5.042799069728183e-08,
+      "loss": 0.3239,
+      "step": 6052
+    },
+    {
+      "epoch": 2.8179702048417132,
+      "grad_norm": 0.5915207210426499,
+      "learning_rate": 5.0171896149124876e-08,
+      "loss": 0.3219,
+      "step": 6053
+    },
+    {
+      "epoch": 2.8184357541899443,
+      "grad_norm": 0.6206277634572867,
+      "learning_rate": 4.9916446943946703e-08,
+      "loss": 0.3175,
+      "step": 6054
+    },
+    {
+      "epoch": 2.818901303538175,
+      "grad_norm": 0.632850226804657,
+      "learning_rate": 4.9661643149038485e-08,
+      "loss": 0.3263,
+      "step": 6055
+    },
+    {
+      "epoch": 2.819366852886406,
+      "grad_norm": 0.5968219742416985,
+      "learning_rate": 4.9407484831521804e-08,
+      "loss": 0.3232,
+      "step": 6056
+    },
+    {
+      "epoch": 2.8198324022346366,
+      "grad_norm": 0.5902481176809331,
+      "learning_rate": 4.91539720583481e-08,
+      "loss": 0.3135,
+      "step": 6057
+    },
+    {
+      "epoch": 2.8202979515828677,
+      "grad_norm": 0.6364766404193624,
+      "learning_rate": 4.8901104896298404e-08,
+      "loss": 0.3231,
+      "step": 6058
+    },
+    {
+      "epoch": 2.8207635009310987,
+      "grad_norm": 0.5852320643387401,
+      "learning_rate": 4.864888341198387e-08,
+      "loss": 0.312,
+      "step": 6059
+    },
+    {
+      "epoch": 2.82122905027933,
+      "grad_norm": 0.6234966594736736,
+      "learning_rate": 4.839730767184608e-08,
+      "loss": 0.3244,
+      "step": 6060
+    },
+    {
+      "epoch": 2.8216945996275604,
+      "grad_norm": 0.6044244454725539,
+      "learning_rate": 4.8146377742155635e-08,
+      "loss": 0.3399,
+      "step": 6061
+    },
+    {
+      "epoch": 2.8221601489757915,
+      "grad_norm": 0.610558760333314,
+      "learning_rate": 4.789609368901382e-08,
+      "loss": 0.3269,
+      "step": 6062
+    },
+    {
+      "epoch": 2.822625698324022,
+      "grad_norm": 0.6198936080979023,
+      "learning_rate": 4.764645557835068e-08,
+      "loss": 0.3381,
+      "step": 6063
+    },
+    {
+      "epoch": 2.823091247672253,
+      "grad_norm": 0.6233886159436226,
+      "learning_rate": 4.739746347592805e-08,
+      "loss": 0.3302,
+      "step": 6064
+    },
+    {
+      "epoch": 2.8235567970204842,
+      "grad_norm": 0.6050139893626,
+      "learning_rate": 4.7149117447335135e-08,
+      "loss": 0.3247,
+      "step": 6065
+    },
+    {
+      "epoch": 2.8240223463687153,
+      "grad_norm": 0.6235298982546221,
+      "learning_rate": 4.6901417557992935e-08,
+      "loss": 0.3178,
+      "step": 6066
+    },
+    {
+      "epoch": 2.824487895716946,
+      "grad_norm": 0.6057656300983348,
+      "learning_rate": 4.66543638731512e-08,
+      "loss": 0.3107,
+      "step": 6067
+    },
+    {
+      "epoch": 2.824953445065177,
+      "grad_norm": 0.6105110505275518,
+      "learning_rate": 4.640795645788981e-08,
+      "loss": 0.312,
+      "step": 6068
+    },
+    {
+      "epoch": 2.8254189944134076,
+      "grad_norm": 0.608835325018187,
+      "learning_rate": 4.616219537711852e-08,
+      "loss": 0.3113,
+      "step": 6069
+    },
+    {
+      "epoch": 2.8258845437616387,
+      "grad_norm": 0.5941087584233198,
+      "learning_rate": 4.59170806955761e-08,
+      "loss": 0.3379,
+      "step": 6070
+    },
+    {
+      "epoch": 2.8263500931098697,
+      "grad_norm": 0.6277457317045492,
+      "learning_rate": 4.5672612477832e-08,
+      "loss": 0.3409,
+      "step": 6071
+    },
+    {
+      "epoch": 2.826815642458101,
+      "grad_norm": 0.5943838728390274,
+      "learning_rate": 4.542879078828527e-08,
+      "loss": 0.3174,
+      "step": 6072
+    },
+    {
+      "epoch": 2.8272811918063314,
+      "grad_norm": 0.6096062816909914,
+      "learning_rate": 4.5185615691163696e-08,
+      "loss": 0.3251,
+      "step": 6073
+    },
+    {
+      "epoch": 2.8277467411545625,
+      "grad_norm": 0.6139351748434146,
+      "learning_rate": 4.494308725052604e-08,
+      "loss": 0.3363,
+      "step": 6074
+    },
+    {
+      "epoch": 2.828212290502793,
+      "grad_norm": 0.6043172974312586,
+      "learning_rate": 4.470120553025925e-08,
+      "loss": 0.3186,
+      "step": 6075
+    },
+    {
+      "epoch": 2.828677839851024,
+      "grad_norm": 0.606724323407064,
+      "learning_rate": 4.445997059408125e-08,
+      "loss": 0.3227,
+      "step": 6076
+    },
+    {
+      "epoch": 2.8291433891992552,
+      "grad_norm": 0.6353846592619885,
+      "learning_rate": 4.421938250553898e-08,
+      "loss": 0.3107,
+      "step": 6077
+    },
+    {
+      "epoch": 2.8296089385474863,
+      "grad_norm": 0.6174663471328742,
+      "learning_rate": 4.3979441328009245e-08,
+      "loss": 0.342,
+      "step": 6078
+    },
+    {
+      "epoch": 2.830074487895717,
+      "grad_norm": 0.5785195682027636,
+      "learning_rate": 4.374014712469787e-08,
+      "loss": 0.3218,
+      "step": 6079
+    },
+    {
+      "epoch": 2.830540037243948,
+      "grad_norm": 0.5841755070985241,
+      "learning_rate": 4.350149995864084e-08,
+      "loss": 0.3175,
+      "step": 6080
+    },
+    {
+      "epoch": 2.8310055865921786,
+      "grad_norm": 0.6211383711211296,
+      "learning_rate": 4.326349989270368e-08,
+      "loss": 0.3275,
+      "step": 6081
+    },
+    {
+      "epoch": 2.8314711359404097,
+      "grad_norm": 0.5910164267835459,
+      "learning_rate": 4.302614698958097e-08,
+      "loss": 0.3226,
+      "step": 6082
+    },
+    {
+      "epoch": 2.8319366852886407,
+      "grad_norm": 0.6137214690418995,
+      "learning_rate": 4.278944131179713e-08,
+      "loss": 0.3284,
+      "step": 6083
+    },
+    {
+      "epoch": 2.8324022346368714,
+      "grad_norm": 0.606360128429641,
+      "learning_rate": 4.255338292170619e-08,
+      "loss": 0.3303,
+      "step": 6084
+    },
+    {
+      "epoch": 2.8328677839851024,
+      "grad_norm": 0.6144174593626709,
+      "learning_rate": 4.231797188149145e-08,
+      "loss": 0.3243,
+      "step": 6085
+    },
+    {
+      "epoch": 2.8333333333333335,
+      "grad_norm": 0.6049337579332242,
+      "learning_rate": 4.2083208253166085e-08,
+      "loss": 0.3326,
+      "step": 6086
+    },
+    {
+      "epoch": 2.833798882681564,
+      "grad_norm": 0.588708284706178,
+      "learning_rate": 4.184909209857202e-08,
+      "loss": 0.3231,
+      "step": 6087
+    },
+    {
+      "epoch": 2.834264432029795,
+      "grad_norm": 0.6037408711792357,
+      "learning_rate": 4.161562347938103e-08,
+      "loss": 0.3236,
+      "step": 6088
+    },
+    {
+      "epoch": 2.8347299813780262,
+      "grad_norm": 0.6211785113807068,
+      "learning_rate": 4.138280245709447e-08,
+      "loss": 0.3303,
+      "step": 6089
+    },
+    {
+      "epoch": 2.835195530726257,
+      "grad_norm": 0.6035205473786263,
+      "learning_rate": 4.1150629093043016e-08,
+      "loss": 0.3194,
+      "step": 6090
+    },
+    {
+      "epoch": 2.835661080074488,
+      "grad_norm": 0.6159203834721481,
+      "learning_rate": 4.091910344838634e-08,
+      "loss": 0.3207,
+      "step": 6091
+    },
+    {
+      "epoch": 2.8361266294227185,
+      "grad_norm": 0.6277196376578653,
+      "learning_rate": 4.068822558411428e-08,
+      "loss": 0.3278,
+      "step": 6092
+    },
+    {
+      "epoch": 2.8365921787709496,
+      "grad_norm": 0.5892371289562799,
+      "learning_rate": 4.045799556104513e-08,
+      "loss": 0.3297,
+      "step": 6093
+    },
+    {
+      "epoch": 2.8370577281191807,
+      "grad_norm": 0.5793393375768995,
+      "learning_rate": 4.022841343982675e-08,
+      "loss": 0.3293,
+      "step": 6094
+    },
+    {
+      "epoch": 2.8375232774674117,
+      "grad_norm": 0.6381992097720461,
+      "learning_rate": 3.999947928093717e-08,
+      "loss": 0.3297,
+      "step": 6095
+    },
+    {
+      "epoch": 2.8379888268156424,
+      "grad_norm": 0.5936577919780165,
+      "learning_rate": 3.977119314468314e-08,
+      "loss": 0.322,
+      "step": 6096
+    },
+    {
+      "epoch": 2.8384543761638734,
+      "grad_norm": 0.6155982675101087,
+      "learning_rate": 3.954355509119989e-08,
+      "loss": 0.3396,
+      "step": 6097
+    },
+    {
+      "epoch": 2.838919925512104,
+      "grad_norm": 0.6343764453060312,
+      "learning_rate": 3.931656518045307e-08,
+      "loss": 0.332,
+      "step": 6098
+    },
+    {
+      "epoch": 2.839385474860335,
+      "grad_norm": 0.6040461231310111,
+      "learning_rate": 3.909022347223734e-08,
+      "loss": 0.327,
+      "step": 6099
+    },
+    {
+      "epoch": 2.839851024208566,
+      "grad_norm": 0.6117159933701583,
+      "learning_rate": 3.886453002617641e-08,
+      "loss": 0.3272,
+      "step": 6100
+    },
+    {
+      "epoch": 2.8403165735567972,
+      "grad_norm": 0.6404426172308234,
+      "learning_rate": 3.863948490172354e-08,
+      "loss": 0.3135,
+      "step": 6101
+    },
+    {
+      "epoch": 2.840782122905028,
+      "grad_norm": 0.6144135391929217,
+      "learning_rate": 3.841508815816047e-08,
+      "loss": 0.3384,
+      "step": 6102
+    },
+    {
+      "epoch": 2.841247672253259,
+      "grad_norm": 0.624193862130294,
+      "learning_rate": 3.819133985459883e-08,
+      "loss": 0.3188,
+      "step": 6103
+    },
+    {
+      "epoch": 2.8417132216014895,
+      "grad_norm": 0.6196530687503252,
+      "learning_rate": 3.796824004997951e-08,
+      "loss": 0.3365,
+      "step": 6104
+    },
+    {
+      "epoch": 2.8421787709497206,
+      "grad_norm": 0.6014966955906467,
+      "learning_rate": 3.774578880307217e-08,
+      "loss": 0.3324,
+      "step": 6105
+    },
+    {
+      "epoch": 2.8426443202979517,
+      "grad_norm": 0.6088272636129385,
+      "learning_rate": 3.752398617247549e-08,
+      "loss": 0.3073,
+      "step": 6106
+    },
+    {
+      "epoch": 2.8431098696461827,
+      "grad_norm": 0.6044011072319111,
+      "learning_rate": 3.7302832216618015e-08,
+      "loss": 0.3319,
+      "step": 6107
+    },
+    {
+      "epoch": 2.8435754189944134,
+      "grad_norm": 0.634366002527612,
+      "learning_rate": 3.708232699375647e-08,
+      "loss": 0.3359,
+      "step": 6108
+    },
+    {
+      "epoch": 2.8440409683426444,
+      "grad_norm": 0.5913257091134858,
+      "learning_rate": 3.686247056197745e-08,
+      "loss": 0.3135,
+      "step": 6109
+    },
+    {
+      "epoch": 2.844506517690875,
+      "grad_norm": 0.5919890953686203,
+      "learning_rate": 3.6643262979196296e-08,
+      "loss": 0.325,
+      "step": 6110
+    },
+    {
+      "epoch": 2.844972067039106,
+      "grad_norm": 0.570303653932179,
+      "learning_rate": 3.642470430315764e-08,
+      "loss": 0.3278,
+      "step": 6111
+    },
+    {
+      "epoch": 2.845437616387337,
+      "grad_norm": 0.6018162132155799,
+      "learning_rate": 3.620679459143489e-08,
+      "loss": 0.3252,
+      "step": 6112
+    },
+    {
+      "epoch": 2.8459031657355682,
+      "grad_norm": 0.6903105702469282,
+      "learning_rate": 3.5989533901430727e-08,
+      "loss": 0.307,
+      "step": 6113
+    },
+    {
+      "epoch": 2.846368715083799,
+      "grad_norm": 0.6218551923613157,
+      "learning_rate": 3.577292229037632e-08,
+      "loss": 0.3259,
+      "step": 6114
+    },
+    {
+      "epoch": 2.84683426443203,
+      "grad_norm": 0.6133132002198325,
+      "learning_rate": 3.5556959815333244e-08,
+      "loss": 0.3249,
+      "step": 6115
+    },
+    {
+      "epoch": 2.8472998137802605,
+      "grad_norm": 0.5679354639860859,
+      "learning_rate": 3.534164653319072e-08,
+      "loss": 0.3202,
+      "step": 6116
+    },
+    {
+      "epoch": 2.8477653631284916,
+      "grad_norm": 0.6238852610993194,
+      "learning_rate": 3.512698250066698e-08,
+      "loss": 0.3346,
+      "step": 6117
+    },
+    {
+      "epoch": 2.8482309124767227,
+      "grad_norm": 0.6224113021167592,
+      "learning_rate": 3.491296777430986e-08,
+      "loss": 0.3182,
+      "step": 6118
+    },
+    {
+      "epoch": 2.8486964618249533,
+      "grad_norm": 0.6229891525027814,
+      "learning_rate": 3.469960241049647e-08,
+      "loss": 0.326,
+      "step": 6119
+    },
+    {
+      "epoch": 2.8491620111731844,
+      "grad_norm": 0.6293955243873184,
+      "learning_rate": 3.448688646543158e-08,
+      "loss": 0.3305,
+      "step": 6120
+    },
+    {
+      "epoch": 2.8496275605214154,
+      "grad_norm": 0.6367163407715672,
+      "learning_rate": 3.427481999515009e-08,
+      "loss": 0.3362,
+      "step": 6121
+    },
+    {
+      "epoch": 2.850093109869646,
+      "grad_norm": 0.6322490712558316,
+      "learning_rate": 3.406340305551481e-08,
+      "loss": 0.3092,
+      "step": 6122
+    },
+    {
+      "epoch": 2.850558659217877,
+      "grad_norm": 0.6203579220439956,
+      "learning_rate": 3.3852635702218694e-08,
+      "loss": 0.3132,
+      "step": 6123
+    },
+    {
+      "epoch": 2.851024208566108,
+      "grad_norm": 0.5879721418602295,
+      "learning_rate": 3.364251799078261e-08,
+      "loss": 0.326,
+      "step": 6124
+    },
+    {
+      "epoch": 2.851489757914339,
+      "grad_norm": 0.6010615898795425,
+      "learning_rate": 3.343304997655616e-08,
+      "loss": 0.3212,
+      "step": 6125
+    },
+    {
+      "epoch": 2.85195530726257,
+      "grad_norm": 0.6641561563695308,
+      "learning_rate": 3.3224231714718835e-08,
+      "loss": 0.3492,
+      "step": 6126
+    },
+    {
+      "epoch": 2.8524208566108005,
+      "grad_norm": 0.6003332926056079,
+      "learning_rate": 3.301606326027801e-08,
+      "loss": 0.3257,
+      "step": 6127
+    },
+    {
+      "epoch": 2.8528864059590315,
+      "grad_norm": 0.6127950838879753,
+      "learning_rate": 3.28085446680701e-08,
+      "loss": 0.3141,
+      "step": 6128
+    },
+    {
+      "epoch": 2.8533519553072626,
+      "grad_norm": 0.6792832248386348,
+      "learning_rate": 3.260167599276054e-08,
+      "loss": 0.3067,
+      "step": 6129
+    },
+    {
+      "epoch": 2.8538175046554937,
+      "grad_norm": 0.6245858737908139,
+      "learning_rate": 3.2395457288843526e-08,
+      "loss": 0.3287,
+      "step": 6130
+    },
+    {
+      "epoch": 2.8542830540037243,
+      "grad_norm": 0.6385608160833756,
+      "learning_rate": 3.218988861064198e-08,
+      "loss": 0.3349,
+      "step": 6131
+    },
+    {
+      "epoch": 2.8547486033519553,
+      "grad_norm": 0.5962474613883065,
+      "learning_rate": 3.19849700123076e-08,
+      "loss": 0.3286,
+      "step": 6132
+    },
+    {
+      "epoch": 2.855214152700186,
+      "grad_norm": 0.609185841089511,
+      "learning_rate": 3.1780701547820814e-08,
+      "loss": 0.3131,
+      "step": 6133
+    },
+    {
+      "epoch": 2.855679702048417,
+      "grad_norm": 0.6112244935356198,
+      "learning_rate": 3.157708327099024e-08,
+      "loss": 0.3272,
+      "step": 6134
+    },
+    {
+      "epoch": 2.856145251396648,
+      "grad_norm": 0.5942955063546984,
+      "learning_rate": 3.137411523545492e-08,
+      "loss": 0.3336,
+      "step": 6135
+    },
+    {
+      "epoch": 2.856610800744879,
+      "grad_norm": 0.5948723183142024,
+      "learning_rate": 3.117179749468041e-08,
+      "loss": 0.3294,
+      "step": 6136
+    },
+    {
+      "epoch": 2.85707635009311,
+      "grad_norm": 0.5999480599255367,
+      "learning_rate": 3.0970130101962425e-08,
+      "loss": 0.3307,
+      "step": 6137
+    },
+    {
+      "epoch": 2.857541899441341,
+      "grad_norm": 0.6082950886036119,
+      "learning_rate": 3.076911311042513e-08,
+      "loss": 0.3269,
+      "step": 6138
+    },
+    {
+      "epoch": 2.8580074487895715,
+      "grad_norm": 0.6091526825115503,
+      "learning_rate": 3.056874657302089e-08,
+      "loss": 0.3309,
+      "step": 6139
+    },
+    {
+      "epoch": 2.8584729981378025,
+      "grad_norm": 0.6003316081004969,
+      "learning_rate": 3.036903054253082e-08,
+      "loss": 0.315,
+      "step": 6140
+    },
+    {
+      "epoch": 2.8589385474860336,
+      "grad_norm": 0.6136795247517056,
+      "learning_rate": 3.0169965071565334e-08,
+      "loss": 0.3367,
+      "step": 6141
+    },
+    {
+      "epoch": 2.8594040968342647,
+      "grad_norm": 0.5966303555412558,
+      "learning_rate": 2.997155021256304e-08,
+      "loss": 0.3331,
+      "step": 6142
+    },
+    {
+      "epoch": 2.8598696461824953,
+      "grad_norm": 0.599057798281752,
+      "learning_rate": 2.977378601779074e-08,
+      "loss": 0.3232,
+      "step": 6143
+    },
+    {
+      "epoch": 2.8603351955307263,
+      "grad_norm": 0.6178666077327492,
+      "learning_rate": 2.9576672539344542e-08,
+      "loss": 0.3062,
+      "step": 6144
+    },
+    {
+      "epoch": 2.860800744878957,
+      "grad_norm": 0.6290939637839866,
+      "learning_rate": 2.9380209829148465e-08,
+      "loss": 0.3339,
+      "step": 6145
+    },
+    {
+      "epoch": 2.861266294227188,
+      "grad_norm": 0.6145143468947959,
+      "learning_rate": 2.9184397938955834e-08,
+      "loss": 0.3238,
+      "step": 6146
+    },
+    {
+      "epoch": 2.861731843575419,
+      "grad_norm": 0.6126152951404651,
+      "learning_rate": 2.8989236920348163e-08,
+      "loss": 0.3265,
+      "step": 6147
+    },
+    {
+      "epoch": 2.86219739292365,
+      "grad_norm": 0.5861962793844461,
+      "learning_rate": 2.8794726824735165e-08,
+      "loss": 0.3182,
+      "step": 6148
+    },
+    {
+      "epoch": 2.862662942271881,
+      "grad_norm": 0.6244533475408514,
+      "learning_rate": 2.8600867703355572e-08,
+      "loss": 0.3214,
+      "step": 6149
+    },
+    {
+      "epoch": 2.863128491620112,
+      "grad_norm": 0.6205916134912833,
+      "learning_rate": 2.8407659607276595e-08,
+      "loss": 0.3322,
+      "step": 6150
+    },
+    {
+      "epoch": 2.8635940409683425,
+      "grad_norm": 0.599175655575075,
+      "learning_rate": 2.8215102587393626e-08,
+      "loss": 0.3317,
+      "step": 6151
+    },
+    {
+      "epoch": 2.8640595903165735,
+      "grad_norm": 0.599885666731648,
+      "learning_rate": 2.8023196694431377e-08,
+      "loss": 0.3456,
+      "step": 6152
+    },
+    {
+      "epoch": 2.8645251396648046,
+      "grad_norm": 0.6348470394106197,
+      "learning_rate": 2.7831941978941347e-08,
+      "loss": 0.336,
+      "step": 6153
+    },
+    {
+      "epoch": 2.864990689013035,
+      "grad_norm": 0.6101207087573923,
+      "learning_rate": 2.7641338491305458e-08,
+      "loss": 0.3294,
+      "step": 6154
+    },
+    {
+      "epoch": 2.8654562383612663,
+      "grad_norm": 0.6174022106842832,
+      "learning_rate": 2.7451386281732994e-08,
+      "loss": 0.3395,
+      "step": 6155
+    },
+    {
+      "epoch": 2.8659217877094973,
+      "grad_norm": 0.6407939037141392,
+      "learning_rate": 2.7262085400261985e-08,
+      "loss": 0.3265,
+      "step": 6156
+    },
+    {
+      "epoch": 2.866387337057728,
+      "grad_norm": 0.590438836087675,
+      "learning_rate": 2.7073435896758093e-08,
+      "loss": 0.3233,
+      "step": 6157
+    },
+    {
+      "epoch": 2.866852886405959,
+      "grad_norm": 0.9417740419722547,
+      "learning_rate": 2.6885437820917127e-08,
+      "loss": 0.3351,
+      "step": 6158
+    },
+    {
+      "epoch": 2.86731843575419,
+      "grad_norm": 0.5963366784535717,
+      "learning_rate": 2.6698091222261413e-08,
+      "loss": 0.3077,
+      "step": 6159
+    },
+    {
+      "epoch": 2.8677839851024207,
+      "grad_norm": 0.576473955352587,
+      "learning_rate": 2.6511396150142865e-08,
+      "loss": 0.3265,
+      "step": 6160
+    },
+    {
+      "epoch": 2.868249534450652,
+      "grad_norm": 0.6081079829012429,
+      "learning_rate": 2.6325352653741586e-08,
+      "loss": 0.3161,
+      "step": 6161
+    },
+    {
+      "epoch": 2.868715083798883,
+      "grad_norm": 0.5812725221199195,
+      "learning_rate": 2.613996078206532e-08,
+      "loss": 0.3249,
+      "step": 6162
+    },
+    {
+      "epoch": 2.8691806331471135,
+      "grad_norm": 0.6089975157336062,
+      "learning_rate": 2.5955220583951112e-08,
+      "loss": 0.3366,
+      "step": 6163
+    },
+    {
+      "epoch": 2.8696461824953445,
+      "grad_norm": 0.6081827470798699,
+      "learning_rate": 2.5771132108063923e-08,
+      "loss": 0.3304,
+      "step": 6164
+    },
+    {
+      "epoch": 2.8701117318435756,
+      "grad_norm": 0.5925881809792082,
+      "learning_rate": 2.5587695402896905e-08,
+      "loss": 0.3241,
+      "step": 6165
+    },
+    {
+      "epoch": 2.870577281191806,
+      "grad_norm": 0.5797248439345088,
+      "learning_rate": 2.540491051677141e-08,
+      "loss": 0.3208,
+      "step": 6166
+    },
+    {
+      "epoch": 2.8710428305400373,
+      "grad_norm": 0.6185018522197453,
+      "learning_rate": 2.522277749783808e-08,
+      "loss": 0.3276,
+      "step": 6167
+    },
+    {
+      "epoch": 2.871508379888268,
+      "grad_norm": 0.5916483638349672,
+      "learning_rate": 2.50412963940741e-08,
+      "loss": 0.3171,
+      "step": 6168
+    },
+    {
+      "epoch": 2.871973929236499,
+      "grad_norm": 0.5810735695209202,
+      "learning_rate": 2.48604672532865e-08,
+      "loss": 0.3054,
+      "step": 6169
+    },
+    {
+      "epoch": 2.87243947858473,
+      "grad_norm": 0.591316216971724,
+      "learning_rate": 2.468029012311024e-08,
+      "loss": 0.3193,
+      "step": 6170
+    },
+    {
+      "epoch": 2.872905027932961,
+      "grad_norm": 0.592893274277165,
+      "learning_rate": 2.4500765051007626e-08,
+      "loss": 0.3228,
+      "step": 6171
+    },
+    {
+      "epoch": 2.8733705772811917,
+      "grad_norm": 0.6120659373334629,
+      "learning_rate": 2.4321892084269994e-08,
+      "loss": 0.3228,
+      "step": 6172
+    },
+    {
+      "epoch": 2.873836126629423,
+      "grad_norm": 0.6393932873332807,
+      "learning_rate": 2.4143671270016877e-08,
+      "loss": 0.3306,
+      "step": 6173
+    },
+    {
+      "epoch": 2.8743016759776534,
+      "grad_norm": 0.5858942340494091,
+      "learning_rate": 2.3966102655196e-08,
+      "loss": 0.3319,
+      "step": 6174
+    },
+    {
+      "epoch": 2.8747672253258845,
+      "grad_norm": 0.6003326151824714,
+      "learning_rate": 2.378918628658272e-08,
+      "loss": 0.326,
+      "step": 6175
+    },
+    {
+      "epoch": 2.8752327746741155,
+      "grad_norm": 0.5756111634299933,
+      "learning_rate": 2.3612922210781706e-08,
+      "loss": 0.2975,
+      "step": 6176
+    },
+    {
+      "epoch": 2.8756983240223466,
+      "grad_norm": 0.617392151652724,
+      "learning_rate": 2.3437310474224428e-08,
+      "loss": 0.3393,
+      "step": 6177
+    },
+    {
+      "epoch": 2.876163873370577,
+      "grad_norm": 0.6273413549010448,
+      "learning_rate": 2.3262351123171378e-08,
+      "loss": 0.3252,
+      "step": 6178
+    },
+    {
+      "epoch": 2.8766294227188083,
+      "grad_norm": 0.6155865212663922,
+      "learning_rate": 2.3088044203711247e-08,
+      "loss": 0.3153,
+      "step": 6179
+    },
+    {
+      "epoch": 2.877094972067039,
+      "grad_norm": 0.6568430850354349,
+      "learning_rate": 2.2914389761760358e-08,
+      "loss": 0.3451,
+      "step": 6180
+    },
+    {
+      "epoch": 2.87756052141527,
+      "grad_norm": 0.6131560074093335,
+      "learning_rate": 2.2741387843063512e-08,
+      "loss": 0.3208,
+      "step": 6181
+    },
+    {
+      "epoch": 2.878026070763501,
+      "grad_norm": 0.6001989073305612,
+      "learning_rate": 2.2569038493193696e-08,
+      "loss": 0.3188,
+      "step": 6182
+    },
+    {
+      "epoch": 2.878491620111732,
+      "grad_norm": 0.6644933124091351,
+      "learning_rate": 2.2397341757551538e-08,
+      "loss": 0.3243,
+      "step": 6183
+    },
+    {
+      "epoch": 2.8789571694599627,
+      "grad_norm": 0.6112914655708457,
+      "learning_rate": 2.222629768136614e-08,
+      "loss": 0.3229,
+      "step": 6184
+    },
+    {
+      "epoch": 2.879422718808194,
+      "grad_norm": 0.6552633615449306,
+      "learning_rate": 2.2055906309694786e-08,
+      "loss": 0.3357,
+      "step": 6185
+    },
+    {
+      "epoch": 2.8798882681564244,
+      "grad_norm": 0.6160515144357356,
+      "learning_rate": 2.188616768742241e-08,
+      "loss": 0.3237,
+      "step": 6186
+    },
+    {
+      "epoch": 2.8803538175046555,
+      "grad_norm": 0.6066047374680886,
+      "learning_rate": 2.1717081859262136e-08,
+      "loss": 0.3416,
+      "step": 6187
+    },
+    {
+      "epoch": 2.8808193668528865,
+      "grad_norm": 0.5869091069268242,
+      "learning_rate": 2.154864886975555e-08,
+      "loss": 0.3199,
+      "step": 6188
+    },
+    {
+      "epoch": 2.881284916201117,
+      "grad_norm": 0.5983947753511701,
+      "learning_rate": 2.1380868763271334e-08,
+      "loss": 0.3203,
+      "step": 6189
+    },
+    {
+      "epoch": 2.881750465549348,
+      "grad_norm": 0.6029948412325274,
+      "learning_rate": 2.1213741584007185e-08,
+      "loss": 0.3283,
+      "step": 6190
+    },
+    {
+      "epoch": 2.8822160148975793,
+      "grad_norm": 0.6090301133296191,
+      "learning_rate": 2.1047267375988445e-08,
+      "loss": 0.3188,
+      "step": 6191
+    },
+    {
+      "epoch": 2.88268156424581,
+      "grad_norm": 0.6204773236903836,
+      "learning_rate": 2.088144618306781e-08,
+      "loss": 0.3309,
+      "step": 6192
+    },
+    {
+      "epoch": 2.883147113594041,
+      "grad_norm": 0.6379438379617484,
+      "learning_rate": 2.0716278048927284e-08,
+      "loss": 0.3345,
+      "step": 6193
+    },
+    {
+      "epoch": 2.883612662942272,
+      "grad_norm": 0.6001291562404845,
+      "learning_rate": 2.0551763017075676e-08,
+      "loss": 0.3309,
+      "step": 6194
+    },
+    {
+      "epoch": 2.8840782122905027,
+      "grad_norm": 0.5736321124106047,
+      "learning_rate": 2.0387901130849986e-08,
+      "loss": 0.3171,
+      "step": 6195
+    },
+    {
+      "epoch": 2.8845437616387337,
+      "grad_norm": 0.6381859177605255,
+      "learning_rate": 2.0224692433415683e-08,
+      "loss": 0.3401,
+      "step": 6196
+    },
+    {
+      "epoch": 2.885009310986965,
+      "grad_norm": 0.6202759855009535,
+      "learning_rate": 2.00621369677656e-08,
+      "loss": 0.3261,
+      "step": 6197
+    },
+    {
+      "epoch": 2.8854748603351954,
+      "grad_norm": 0.6311049881351487,
+      "learning_rate": 1.990023477672076e-08,
+      "loss": 0.3255,
+      "step": 6198
+    },
+    {
+      "epoch": 2.8859404096834265,
+      "grad_norm": 0.6015571016423136,
+      "learning_rate": 1.9738985902930386e-08,
+      "loss": 0.3315,
+      "step": 6199
+    },
+    {
+      "epoch": 2.8864059590316575,
+      "grad_norm": 0.6315652584761223,
+      "learning_rate": 1.9578390388870494e-08,
+      "loss": 0.3177,
+      "step": 6200
+    },
+    {
+      "epoch": 2.886871508379888,
+      "grad_norm": 0.5924676731773729,
+      "learning_rate": 1.9418448276846412e-08,
+      "loss": 0.3292,
+      "step": 6201
+    },
+    {
+      "epoch": 2.887337057728119,
+      "grad_norm": 0.6391867931517905,
+      "learning_rate": 1.925915960899083e-08,
+      "loss": 0.3235,
+      "step": 6202
+    },
+    {
+      "epoch": 2.88780260707635,
+      "grad_norm": 0.5873703536415629,
+      "learning_rate": 1.9100524427263513e-08,
+      "loss": 0.321,
+      "step": 6203
+    },
+    {
+      "epoch": 2.888268156424581,
+      "grad_norm": 0.6093540578155664,
+      "learning_rate": 1.8942542773452978e-08,
+      "loss": 0.3159,
+      "step": 6204
+    },
+    {
+      "epoch": 2.888733705772812,
+      "grad_norm": 0.6090367968687317,
+      "learning_rate": 1.8785214689175657e-08,
+      "loss": 0.3313,
+      "step": 6205
+    },
+    {
+      "epoch": 2.889199255121043,
+      "grad_norm": 0.6308464598775337,
+      "learning_rate": 1.8628540215875346e-08,
+      "loss": 0.3482,
+      "step": 6206
+    },
+    {
+      "epoch": 2.8896648044692737,
+      "grad_norm": 0.6066227893950777,
+      "learning_rate": 1.847251939482375e-08,
+      "loss": 0.3083,
+      "step": 6207
+    },
+    {
+      "epoch": 2.8901303538175047,
+      "grad_norm": 0.5810087781537883,
+      "learning_rate": 1.831715226712022e-08,
+      "loss": 0.3343,
+      "step": 6208
+    },
+    {
+      "epoch": 2.8905959031657353,
+      "grad_norm": 0.5894592031263982,
+      "learning_rate": 1.8162438873692566e-08,
+      "loss": 0.3391,
+      "step": 6209
+    },
+    {
+      "epoch": 2.8910614525139664,
+      "grad_norm": 0.6723253462873507,
+      "learning_rate": 1.8008379255295972e-08,
+      "loss": 0.3328,
+      "step": 6210
+    },
+    {
+      "epoch": 2.8915270018621975,
+      "grad_norm": 0.6005955744296563,
+      "learning_rate": 1.785497345251325e-08,
+      "loss": 0.3136,
+      "step": 6211
+    },
+    {
+      "epoch": 2.8919925512104285,
+      "grad_norm": 0.6468037660631027,
+      "learning_rate": 1.7702221505755124e-08,
+      "loss": 0.314,
+      "step": 6212
+    },
+    {
+      "epoch": 2.892458100558659,
+      "grad_norm": 0.6047241269882732,
+      "learning_rate": 1.755012345525997e-08,
+      "loss": 0.323,
+      "step": 6213
+    },
+    {
+      "epoch": 2.89292364990689,
+      "grad_norm": 0.6333657009253152,
+      "learning_rate": 1.7398679341094625e-08,
+      "loss": 0.3199,
+      "step": 6214
+    },
+    {
+      "epoch": 2.893389199255121,
+      "grad_norm": 0.612023907476288,
+      "learning_rate": 1.7247889203152178e-08,
+      "loss": 0.326,
+      "step": 6215
+    },
+    {
+      "epoch": 2.893854748603352,
+      "grad_norm": 0.6253353161437807,
+      "learning_rate": 1.7097753081154745e-08,
+      "loss": 0.3199,
+      "step": 6216
+    },
+    {
+      "epoch": 2.894320297951583,
+      "grad_norm": 0.5777191745384375,
+      "learning_rate": 1.6948271014652073e-08,
+      "loss": 0.3248,
+      "step": 6217
+    },
+    {
+      "epoch": 2.894785847299814,
+      "grad_norm": 0.6223854035143206,
+      "learning_rate": 1.6799443043021e-08,
+      "loss": 0.3231,
+      "step": 6218
+    },
+    {
+      "epoch": 2.8952513966480447,
+      "grad_norm": 0.6248036985131868,
+      "learning_rate": 1.6651269205466282e-08,
+      "loss": 0.339,
+      "step": 6219
+    },
+    {
+      "epoch": 2.8957169459962757,
+      "grad_norm": 0.6339061551119733,
+      "learning_rate": 1.6503749541020576e-08,
+      "loss": 0.3145,
+      "step": 6220
+    },
+    {
+      "epoch": 2.8961824953445063,
+      "grad_norm": 0.578239787381323,
+      "learning_rate": 1.6356884088543634e-08,
+      "loss": 0.3331,
+      "step": 6221
+    },
+    {
+      "epoch": 2.8966480446927374,
+      "grad_norm": 0.609082583845745,
+      "learning_rate": 1.6210672886724233e-08,
+      "loss": 0.3331,
+      "step": 6222
+    },
+    {
+      "epoch": 2.8971135940409685,
+      "grad_norm": 0.5980092824315858,
+      "learning_rate": 1.606511597407684e-08,
+      "loss": 0.3094,
+      "step": 6223
+    },
+    {
+      "epoch": 2.8975791433891995,
+      "grad_norm": 0.6062420563220096,
+      "learning_rate": 1.5920213388945504e-08,
+      "loss": 0.3472,
+      "step": 6224
+    },
+    {
+      "epoch": 2.89804469273743,
+      "grad_norm": 0.5920447394847475,
+      "learning_rate": 1.5775965169500528e-08,
+      "loss": 0.3223,
+      "step": 6225
+    },
+    {
+      "epoch": 2.898510242085661,
+      "grad_norm": 0.6129022304051913,
+      "learning_rate": 1.563237135374013e-08,
+      "loss": 0.3221,
+      "step": 6226
+    },
+    {
+      "epoch": 2.898975791433892,
+      "grad_norm": 0.5949547157737298,
+      "learning_rate": 1.548943197949071e-08,
+      "loss": 0.3132,
+      "step": 6227
+    },
+    {
+      "epoch": 2.899441340782123,
+      "grad_norm": 0.6148641770655932,
+      "learning_rate": 1.5347147084406045e-08,
+      "loss": 0.331,
+      "step": 6228
+    },
+    {
+      "epoch": 2.899906890130354,
+      "grad_norm": 0.5919948378552693,
+      "learning_rate": 1.5205516705966706e-08,
+      "loss": 0.3332,
+      "step": 6229
+    },
+    {
+      "epoch": 2.9003724394785846,
+      "grad_norm": 0.6040637617486767,
+      "learning_rate": 1.5064540881482292e-08,
+      "loss": 0.3257,
+      "step": 6230
+    },
+    {
+      "epoch": 2.9008379888268156,
+      "grad_norm": 0.6131768237281414,
+      "learning_rate": 1.4924219648088657e-08,
+      "loss": 0.3264,
+      "step": 6231
+    },
+    {
+      "epoch": 2.9013035381750467,
+      "grad_norm": 0.6524203098479858,
+      "learning_rate": 1.4784553042749838e-08,
+      "loss": 0.3228,
+      "step": 6232
+    },
+    {
+      "epoch": 2.9017690875232773,
+      "grad_norm": 0.5837510189478903,
+      "learning_rate": 1.4645541102257244e-08,
+      "loss": 0.3225,
+      "step": 6233
+    },
+    {
+      "epoch": 2.9022346368715084,
+      "grad_norm": 0.6008483195471794,
+      "learning_rate": 1.4507183863230468e-08,
+      "loss": 0.337,
+      "step": 6234
+    },
+    {
+      "epoch": 2.9027001862197395,
+      "grad_norm": 0.5903573581410563,
+      "learning_rate": 1.4369481362115355e-08,
+      "loss": 0.3371,
+      "step": 6235
+    },
+    {
+      "epoch": 2.90316573556797,
+      "grad_norm": 0.6048197576945102,
+      "learning_rate": 1.4232433635186505e-08,
+      "loss": 0.3386,
+      "step": 6236
+    },
+    {
+      "epoch": 2.903631284916201,
+      "grad_norm": 0.6297907190915534,
+      "learning_rate": 1.409604071854559e-08,
+      "loss": 0.3524,
+      "step": 6237
+    },
+    {
+      "epoch": 2.9040968342644318,
+      "grad_norm": 0.6452452496470608,
+      "learning_rate": 1.396030264812137e-08,
+      "loss": 0.326,
+      "step": 6238
+    },
+    {
+      "epoch": 2.904562383612663,
+      "grad_norm": 0.5891722803171795,
+      "learning_rate": 1.3825219459670525e-08,
+      "loss": 0.3237,
+      "step": 6239
+    },
+    {
+      "epoch": 2.905027932960894,
+      "grad_norm": 0.5905285938393876,
+      "learning_rate": 1.3690791188777642e-08,
+      "loss": 0.3242,
+      "step": 6240
+    },
+    {
+      "epoch": 2.905493482309125,
+      "grad_norm": 0.604169971222827,
+      "learning_rate": 1.3557017870853839e-08,
+      "loss": 0.3217,
+      "step": 6241
+    },
+    {
+      "epoch": 2.9059590316573556,
+      "grad_norm": 0.6026069904760022,
+      "learning_rate": 1.342389954113843e-08,
+      "loss": 0.3367,
+      "step": 6242
+    },
+    {
+      "epoch": 2.9064245810055866,
+      "grad_norm": 0.5855224825124511,
+      "learning_rate": 1.3291436234697808e-08,
+      "loss": 0.3128,
+      "step": 6243
+    },
+    {
+      "epoch": 2.9068901303538173,
+      "grad_norm": 0.6096987778186526,
+      "learning_rate": 1.3159627986426005e-08,
+      "loss": 0.3255,
+      "step": 6244
+    },
+    {
+      "epoch": 2.9073556797020483,
+      "grad_norm": 0.6013107840522399,
+      "learning_rate": 1.302847483104469e-08,
+      "loss": 0.3049,
+      "step": 6245
+    },
+    {
+      "epoch": 2.9078212290502794,
+      "grad_norm": 0.6187001991804566,
+      "learning_rate": 1.2897976803102341e-08,
+      "loss": 0.3104,
+      "step": 6246
+    },
+    {
+      "epoch": 2.9082867783985105,
+      "grad_norm": 0.6124140099702032,
+      "learning_rate": 1.2768133936975624e-08,
+      "loss": 0.3173,
+      "step": 6247
+    },
+    {
+      "epoch": 2.908752327746741,
+      "grad_norm": 0.6259332067040718,
+      "learning_rate": 1.2638946266867735e-08,
+      "loss": 0.3202,
+      "step": 6248
+    },
+    {
+      "epoch": 2.909217877094972,
+      "grad_norm": 0.616429089128467,
+      "learning_rate": 1.2510413826810342e-08,
+      "loss": 0.3365,
+      "step": 6249
+    },
+    {
+      "epoch": 2.9096834264432028,
+      "grad_norm": 0.6038132516572956,
+      "learning_rate": 1.2382536650661636e-08,
+      "loss": 0.3179,
+      "step": 6250
+    },
+    {
+      "epoch": 2.910148975791434,
+      "grad_norm": 0.5900861579239468,
+      "learning_rate": 1.2255314772107174e-08,
+      "loss": 0.3063,
+      "step": 6251
+    },
+    {
+      "epoch": 2.910614525139665,
+      "grad_norm": 0.6267653790224524,
+      "learning_rate": 1.212874822466098e-08,
+      "loss": 0.3248,
+      "step": 6252
+    },
+    {
+      "epoch": 2.911080074487896,
+      "grad_norm": 0.6013006213600915,
+      "learning_rate": 1.2002837041663052e-08,
+      "loss": 0.3229,
+      "step": 6253
+    },
+    {
+      "epoch": 2.9115456238361266,
+      "grad_norm": 0.629414213878728,
+      "learning_rate": 1.1877581256281856e-08,
+      "loss": 0.3393,
+      "step": 6254
+    },
+    {
+      "epoch": 2.9120111731843576,
+      "grad_norm": 0.6279756891681463,
+      "learning_rate": 1.1752980901512389e-08,
+      "loss": 0.3371,
+      "step": 6255
+    },
+    {
+      "epoch": 2.9124767225325883,
+      "grad_norm": 0.5937039808411767,
+      "learning_rate": 1.1629036010177563e-08,
+      "loss": 0.3333,
+      "step": 6256
+    },
+    {
+      "epoch": 2.9129422718808193,
+      "grad_norm": 0.6069534279097875,
+      "learning_rate": 1.1505746614927372e-08,
+      "loss": 0.3298,
+      "step": 6257
+    },
+    {
+      "epoch": 2.9134078212290504,
+      "grad_norm": 0.6290360127989635,
+      "learning_rate": 1.1383112748239166e-08,
+      "loss": 0.3107,
+      "step": 6258
+    },
+    {
+      "epoch": 2.9138733705772815,
+      "grad_norm": 0.5954994434060333,
+      "learning_rate": 1.1261134442417387e-08,
+      "loss": 0.3208,
+      "step": 6259
+    },
+    {
+      "epoch": 2.914338919925512,
+      "grad_norm": 0.6290235236097937,
+      "learning_rate": 1.1139811729594663e-08,
+      "loss": 0.3188,
+      "step": 6260
+    },
+    {
+      "epoch": 2.914804469273743,
+      "grad_norm": 0.5672271685365449,
+      "learning_rate": 1.1019144641729319e-08,
+      "loss": 0.3173,
+      "step": 6261
+    },
+    {
+      "epoch": 2.9152700186219738,
+      "grad_norm": 0.5963834882277916,
+      "learning_rate": 1.0899133210608704e-08,
+      "loss": 0.3276,
+      "step": 6262
+    },
+    {
+      "epoch": 2.915735567970205,
+      "grad_norm": 0.6446459546230884,
+      "learning_rate": 1.0779777467846419e-08,
+      "loss": 0.3267,
+      "step": 6263
+    },
+    {
+      "epoch": 2.916201117318436,
+      "grad_norm": 0.6140239764030514,
+      "learning_rate": 1.0661077444883705e-08,
+      "loss": 0.327,
+      "step": 6264
+    },
+    {
+      "epoch": 2.9166666666666665,
+      "grad_norm": 0.6571477442299728,
+      "learning_rate": 1.0543033172988881e-08,
+      "loss": 0.3376,
+      "step": 6265
+    },
+    {
+      "epoch": 2.9171322160148976,
+      "grad_norm": 0.6248520799426412,
+      "learning_rate": 1.0425644683257629e-08,
+      "loss": 0.3303,
+      "step": 6266
+    },
+    {
+      "epoch": 2.9175977653631286,
+      "grad_norm": 0.6359065621539344,
+      "learning_rate": 1.0308912006612715e-08,
+      "loss": 0.3185,
+      "step": 6267
+    },
+    {
+      "epoch": 2.9180633147113593,
+      "grad_norm": 0.5947402900067781,
+      "learning_rate": 1.0192835173804538e-08,
+      "loss": 0.3345,
+      "step": 6268
+    },
+    {
+      "epoch": 2.9185288640595903,
+      "grad_norm": 0.6043241267792967,
+      "learning_rate": 1.0077414215410586e-08,
+      "loss": 0.3165,
+      "step": 6269
+    },
+    {
+      "epoch": 2.9189944134078214,
+      "grad_norm": 0.621865224208232,
+      "learning_rate": 9.962649161835147e-09,
+      "loss": 0.3133,
+      "step": 6270
+    },
+    {
+      "epoch": 2.919459962756052,
+      "grad_norm": 0.6102444658155871,
+      "learning_rate": 9.848540043310151e-09,
+      "loss": 0.3248,
+      "step": 6271
+    },
+    {
+      "epoch": 2.919925512104283,
+      "grad_norm": 0.596781037300196,
+      "learning_rate": 9.735086889894884e-09,
+      "loss": 0.318,
+      "step": 6272
+    },
+    {
+      "epoch": 2.9203910614525137,
+      "grad_norm": 0.6219073177176879,
+      "learning_rate": 9.622289731475443e-09,
+      "loss": 0.3348,
+      "step": 6273
+    },
+    {
+      "epoch": 2.9208566108007448,
+      "grad_norm": 0.6080150996247233,
+      "learning_rate": 9.510148597765279e-09,
+      "loss": 0.3258,
+      "step": 6274
+    },
+    {
+      "epoch": 2.921322160148976,
+      "grad_norm": 0.6103185569420561,
+      "learning_rate": 9.398663518304929e-09,
+      "loss": 0.3263,
+      "step": 6275
+    },
+    {
+      "epoch": 2.921787709497207,
+      "grad_norm": 0.6176385933276092,
+      "learning_rate": 9.287834522462292e-09,
+      "loss": 0.3336,
+      "step": 6276
+    },
+    {
+      "epoch": 2.9222532588454375,
+      "grad_norm": 0.5862739945703891,
+      "learning_rate": 9.177661639432622e-09,
+      "loss": 0.3147,
+      "step": 6277
+    },
+    {
+      "epoch": 2.9227188081936686,
+      "grad_norm": 0.626998981043966,
+      "learning_rate": 9.068144898237707e-09,
+      "loss": 0.3264,
+      "step": 6278
+    },
+    {
+      "epoch": 2.923184357541899,
+      "grad_norm": 0.6149966930670874,
+      "learning_rate": 8.959284327727247e-09,
+      "loss": 0.3226,
+      "step": 6279
+    },
+    {
+      "epoch": 2.9236499068901303,
+      "grad_norm": 0.6054449665334783,
+      "learning_rate": 8.851079956577192e-09,
+      "loss": 0.3256,
+      "step": 6280
+    },
+    {
+      "epoch": 2.9241154562383613,
+      "grad_norm": 0.6036402104170521,
+      "learning_rate": 8.743531813291406e-09,
+      "loss": 0.3151,
+      "step": 6281
+    },
+    {
+      "epoch": 2.9245810055865924,
+      "grad_norm": 0.6228051504241633,
+      "learning_rate": 8.636639926200841e-09,
+      "loss": 0.3352,
+      "step": 6282
+    },
+    {
+      "epoch": 2.925046554934823,
+      "grad_norm": 0.6714758603645512,
+      "learning_rate": 8.53040432346297e-09,
+      "loss": 0.3227,
+      "step": 6283
+    },
+    {
+      "epoch": 2.925512104283054,
+      "grad_norm": 0.5811542111967615,
+      "learning_rate": 8.424825033063188e-09,
+      "loss": 0.3301,
+      "step": 6284
+    },
+    {
+      "epoch": 2.9259776536312847,
+      "grad_norm": 0.6510006354129075,
+      "learning_rate": 8.319902082813135e-09,
+      "loss": 0.3233,
+      "step": 6285
+    },
+    {
+      "epoch": 2.9264432029795158,
+      "grad_norm": 0.5834999064377437,
+      "learning_rate": 8.215635500352371e-09,
+      "loss": 0.3151,
+      "step": 6286
+    },
+    {
+      "epoch": 2.926908752327747,
+      "grad_norm": 0.6142333490306449,
+      "learning_rate": 8.112025313146977e-09,
+      "loss": 0.329,
+      "step": 6287
+    },
+    {
+      "epoch": 2.927374301675978,
+      "grad_norm": 0.6171919038103632,
+      "learning_rate": 8.009071548490676e-09,
+      "loss": 0.3159,
+      "step": 6288
+    },
+    {
+      "epoch": 2.9278398510242085,
+      "grad_norm": 0.6162711287931582,
+      "learning_rate": 7.906774233503722e-09,
+      "loss": 0.3342,
+      "step": 6289
+    },
+    {
+      "epoch": 2.9283054003724396,
+      "grad_norm": 0.6147083004824436,
+      "learning_rate": 7.80513339513317e-09,
+      "loss": 0.3475,
+      "step": 6290
+    },
+    {
+      "epoch": 2.92877094972067,
+      "grad_norm": 0.6213827537197353,
+      "learning_rate": 7.704149060154541e-09,
+      "loss": 0.3213,
+      "step": 6291
+    },
+    {
+      "epoch": 2.9292364990689013,
+      "grad_norm": 0.5796458542457161,
+      "learning_rate": 7.603821255168787e-09,
+      "loss": 0.3137,
+      "step": 6292
+    },
+    {
+      "epoch": 2.9297020484171323,
+      "grad_norm": 0.5878936675621818,
+      "learning_rate": 7.504150006605038e-09,
+      "loss": 0.3357,
+      "step": 6293
+    },
+    {
+      "epoch": 2.9301675977653634,
+      "grad_norm": 0.6196937374201812,
+      "learning_rate": 7.405135340718961e-09,
+      "loss": 0.3344,
+      "step": 6294
+    },
+    {
+      "epoch": 2.930633147113594,
+      "grad_norm": 0.6406099995475769,
+      "learning_rate": 7.3067772835933025e-09,
+      "loss": 0.3302,
+      "step": 6295
+    },
+    {
+      "epoch": 2.931098696461825,
+      "grad_norm": 0.5959238868189899,
+      "learning_rate": 7.209075861138171e-09,
+      "loss": 0.3189,
+      "step": 6296
+    },
+    {
+      "epoch": 2.9315642458100557,
+      "grad_norm": 0.6028877391912789,
+      "learning_rate": 7.1120310990902e-09,
+      "loss": 0.3136,
+      "step": 6297
+    },
+    {
+      "epoch": 2.9320297951582868,
+      "grad_norm": 0.5978502088972731,
+      "learning_rate": 7.015643023013108e-09,
+      "loss": 0.3168,
+      "step": 6298
+    },
+    {
+      "epoch": 2.932495344506518,
+      "grad_norm": 0.5952417979334614,
+      "learning_rate": 6.91991165829825e-09,
+      "loss": 0.3097,
+      "step": 6299
+    },
+    {
+      "epoch": 2.9329608938547485,
+      "grad_norm": 0.6015966630139323,
+      "learning_rate": 6.824837030163234e-09,
+      "loss": 0.3295,
+      "step": 6300
+    },
+    {
+      "epoch": 2.9334264432029795,
+      "grad_norm": 0.6379334362915596,
+      "learning_rate": 6.730419163653024e-09,
+      "loss": 0.3258,
+      "step": 6301
+    },
+    {
+      "epoch": 2.9338919925512106,
+      "grad_norm": 0.6109387992023941,
+      "learning_rate": 6.636658083639669e-09,
+      "loss": 0.3421,
+      "step": 6302
+    },
+    {
+      "epoch": 2.934357541899441,
+      "grad_norm": 0.5985634786152769,
+      "learning_rate": 6.543553814821746e-09,
+      "loss": 0.3204,
+      "step": 6303
+    },
+    {
+      "epoch": 2.9348230912476723,
+      "grad_norm": 0.6225417966742165,
+      "learning_rate": 6.4511063817257466e-09,
+      "loss": 0.3233,
+      "step": 6304
+    },
+    {
+      "epoch": 2.9352886405959033,
+      "grad_norm": 0.6057583136449132,
+      "learning_rate": 6.359315808703859e-09,
+      "loss": 0.3348,
+      "step": 6305
+    },
+    {
+      "epoch": 2.935754189944134,
+      "grad_norm": 0.5941430788606544,
+      "learning_rate": 6.268182119936184e-09,
+      "loss": 0.3138,
+      "step": 6306
+    },
+    {
+      "epoch": 2.936219739292365,
+      "grad_norm": 0.5787345014335359,
+      "learning_rate": 6.17770533942963e-09,
+      "loss": 0.3154,
+      "step": 6307
+    },
+    {
+      "epoch": 2.9366852886405956,
+      "grad_norm": 0.5824068889556205,
+      "learning_rate": 6.087885491017908e-09,
+      "loss": 0.3302,
+      "step": 6308
+    },
+    {
+      "epoch": 2.9371508379888267,
+      "grad_norm": 0.6131880359928478,
+      "learning_rate": 5.998722598361539e-09,
+      "loss": 0.3293,
+      "step": 6309
+    },
+    {
+      "epoch": 2.9376163873370578,
+      "grad_norm": 0.6148945913599614,
+      "learning_rate": 5.910216684947845e-09,
+      "loss": 0.354,
+      "step": 6310
+    },
+    {
+      "epoch": 2.938081936685289,
+      "grad_norm": 0.6009071613362356,
+      "learning_rate": 5.8223677740920635e-09,
+      "loss": 0.34,
+      "step": 6311
+    },
+    {
+      "epoch": 2.9385474860335195,
+      "grad_norm": 0.5782246112086374,
+      "learning_rate": 5.735175888935407e-09,
+      "loss": 0.3344,
+      "step": 6312
+    },
+    {
+      "epoch": 2.9390130353817505,
+      "grad_norm": 0.6183187964611667,
+      "learning_rate": 5.648641052446169e-09,
+      "loss": 0.3067,
+      "step": 6313
+    },
+    {
+      "epoch": 2.939478584729981,
+      "grad_norm": 0.5974210306590381,
+      "learning_rate": 5.562763287420003e-09,
+      "loss": 0.342,
+      "step": 6314
+    },
+    {
+      "epoch": 2.939944134078212,
+      "grad_norm": 0.5716017546771096,
+      "learning_rate": 5.477542616478815e-09,
+      "loss": 0.3127,
+      "step": 6315
+    },
+    {
+      "epoch": 2.9404096834264433,
+      "grad_norm": 0.6371306059166789,
+      "learning_rate": 5.392979062071591e-09,
+      "loss": 0.3381,
+      "step": 6316
+    },
+    {
+      "epoch": 2.9408752327746743,
+      "grad_norm": 0.5915971865663107,
+      "learning_rate": 5.3090726464749555e-09,
+      "loss": 0.3206,
+      "step": 6317
+    },
+    {
+      "epoch": 2.941340782122905,
+      "grad_norm": 0.6110107412318451,
+      "learning_rate": 5.225823391791507e-09,
+      "loss": 0.3228,
+      "step": 6318
+    },
+    {
+      "epoch": 2.941806331471136,
+      "grad_norm": 0.577294286484851,
+      "learning_rate": 5.143231319950926e-09,
+      "loss": 0.3322,
+      "step": 6319
+    },
+    {
+      "epoch": 2.9422718808193666,
+      "grad_norm": 0.5883553071363519,
+      "learning_rate": 5.0612964527102515e-09,
+      "loss": 0.3232,
+      "step": 6320
+    },
+    {
+      "epoch": 2.9427374301675977,
+      "grad_norm": 0.6103034546868826,
+      "learning_rate": 4.980018811653053e-09,
+      "loss": 0.322,
+      "step": 6321
+    },
+    {
+      "epoch": 2.9432029795158288,
+      "grad_norm": 0.6169775078641982,
+      "learning_rate": 4.899398418189705e-09,
+      "loss": 0.3225,
+      "step": 6322
+    },
+    {
+      "epoch": 2.94366852886406,
+      "grad_norm": 0.6316243843021199,
+      "learning_rate": 4.819435293557384e-09,
+      "loss": 0.3347,
+      "step": 6323
+    },
+    {
+      "epoch": 2.9441340782122905,
+      "grad_norm": 0.5978526781812734,
+      "learning_rate": 4.74012945882063e-09,
+      "loss": 0.3377,
+      "step": 6324
+    },
+    {
+      "epoch": 2.9445996275605215,
+      "grad_norm": 0.619324575948808,
+      "learning_rate": 4.661480934869955e-09,
+      "loss": 0.3368,
+      "step": 6325
+    },
+    {
+      "epoch": 2.945065176908752,
+      "grad_norm": 0.5862145154491393,
+      "learning_rate": 4.583489742423786e-09,
+      "loss": 0.3096,
+      "step": 6326
+    },
+    {
+      "epoch": 2.945530726256983,
+      "grad_norm": 0.6192741896539704,
+      "learning_rate": 4.5061559020267985e-09,
+      "loss": 0.3196,
+      "step": 6327
+    },
+    {
+      "epoch": 2.9459962756052143,
+      "grad_norm": 0.5942156052915056,
+      "learning_rate": 4.429479434050199e-09,
+      "loss": 0.326,
+      "step": 6328
+    },
+    {
+      "epoch": 2.9464618249534453,
+      "grad_norm": 0.6022798676600029,
+      "learning_rate": 4.353460358692552e-09,
+      "loss": 0.3259,
+      "step": 6329
+    },
+    {
+      "epoch": 2.946927374301676,
+      "grad_norm": 0.6146931033116475,
+      "learning_rate": 4.278098695979227e-09,
+      "loss": 0.3299,
+      "step": 6330
+    },
+    {
+      "epoch": 2.947392923649907,
+      "grad_norm": 0.6228768038133652,
+      "learning_rate": 4.203394465762123e-09,
+      "loss": 0.3412,
+      "step": 6331
+    },
+    {
+      "epoch": 2.9478584729981376,
+      "grad_norm": 0.5978000188308262,
+      "learning_rate": 4.129347687720497e-09,
+      "loss": 0.3137,
+      "step": 6332
+    },
+    {
+      "epoch": 2.9483240223463687,
+      "grad_norm": 0.6012102455795719,
+      "learning_rate": 4.055958381359304e-09,
+      "loss": 0.3273,
+      "step": 6333
+    },
+    {
+      "epoch": 2.9487895716945998,
+      "grad_norm": 0.5695680041601762,
+      "learning_rate": 3.983226566011689e-09,
+      "loss": 0.3195,
+      "step": 6334
+    },
+    {
+      "epoch": 2.9492551210428304,
+      "grad_norm": 0.5950159676230445,
+      "learning_rate": 3.911152260836493e-09,
+      "loss": 0.326,
+      "step": 6335
+    },
+    {
+      "epoch": 2.9497206703910615,
+      "grad_norm": 0.6159913335568062,
+      "learning_rate": 3.839735484820195e-09,
+      "loss": 0.3002,
+      "step": 6336
+    },
+    {
+      "epoch": 2.9501862197392925,
+      "grad_norm": 0.5916464507266811,
+      "learning_rate": 3.768976256775247e-09,
+      "loss": 0.3307,
+      "step": 6337
+    },
+    {
+      "epoch": 2.950651769087523,
+      "grad_norm": 0.6139970989604534,
+      "learning_rate": 3.698874595342017e-09,
+      "loss": 0.3401,
+      "step": 6338
+    },
+    {
+      "epoch": 2.951117318435754,
+      "grad_norm": 0.6201205406276251,
+      "learning_rate": 3.6294305189860104e-09,
+      "loss": 0.3155,
+      "step": 6339
+    },
+    {
+      "epoch": 2.9515828677839853,
+      "grad_norm": 0.6222729820254641,
+      "learning_rate": 3.560644046001205e-09,
+      "loss": 0.3154,
+      "step": 6340
+    },
+    {
+      "epoch": 2.952048417132216,
+      "grad_norm": 0.6006642256872674,
+      "learning_rate": 3.492515194506996e-09,
+      "loss": 0.3306,
+      "step": 6341
+    },
+    {
+      "epoch": 2.952513966480447,
+      "grad_norm": 0.6191479965307279,
+      "learning_rate": 3.4250439824506933e-09,
+      "loss": 0.3164,
+      "step": 6342
+    },
+    {
+      "epoch": 2.9529795158286776,
+      "grad_norm": 0.6304735695223874,
+      "learning_rate": 3.35823042760558e-09,
+      "loss": 0.3405,
+      "step": 6343
+    },
+    {
+      "epoch": 2.9534450651769086,
+      "grad_norm": 0.6139523701678713,
+      "learning_rate": 3.292074547572022e-09,
+      "loss": 0.3163,
+      "step": 6344
+    },
+    {
+      "epoch": 2.9539106145251397,
+      "grad_norm": 0.6094860933997419,
+      "learning_rate": 3.226576359776634e-09,
+      "loss": 0.3141,
+      "step": 6345
+    },
+    {
+      "epoch": 2.9543761638733708,
+      "grad_norm": 0.605070339152517,
+      "learning_rate": 3.1617358814736708e-09,
+      "loss": 0.3276,
+      "step": 6346
+    },
+    {
+      "epoch": 2.9548417132216014,
+      "grad_norm": 0.634761079376964,
+      "learning_rate": 3.0975531297436356e-09,
+      "loss": 0.3259,
+      "step": 6347
+    },
+    {
+      "epoch": 2.9553072625698324,
+      "grad_norm": 0.6065829699426136,
+      "learning_rate": 3.03402812149356e-09,
+      "loss": 0.3305,
+      "step": 6348
+    },
+    {
+      "epoch": 2.955772811918063,
+      "grad_norm": 0.6040784000088028,
+      "learning_rate": 2.971160873457557e-09,
+      "loss": 0.3268,
+      "step": 6349
+    },
+    {
+      "epoch": 2.956238361266294,
+      "grad_norm": 0.6170579174496987,
+      "learning_rate": 2.908951402196547e-09,
+      "loss": 0.3305,
+      "step": 6350
+    },
+    {
+      "epoch": 2.956703910614525,
+      "grad_norm": 0.614656639592586,
+      "learning_rate": 2.8473997240974217e-09,
+      "loss": 0.3135,
+      "step": 6351
+    },
+    {
+      "epoch": 2.9571694599627563,
+      "grad_norm": 0.6043067044081749,
+      "learning_rate": 2.786505855374988e-09,
+      "loss": 0.3343,
+      "step": 6352
+    },
+    {
+      "epoch": 2.957635009310987,
+      "grad_norm": 0.6127863265687097,
+      "learning_rate": 2.7262698120697483e-09,
+      "loss": 0.3099,
+      "step": 6353
+    },
+    {
+      "epoch": 2.958100558659218,
+      "grad_norm": 0.5892052799393268,
+      "learning_rate": 2.6666916100495654e-09,
+      "loss": 0.3158,
+      "step": 6354
+    },
+    {
+      "epoch": 2.9585661080074486,
+      "grad_norm": 0.6158145656161989,
+      "learning_rate": 2.607771265008552e-09,
+      "loss": 0.3217,
+      "step": 6355
+    },
+    {
+      "epoch": 2.9590316573556796,
+      "grad_norm": 0.5822668968700528,
+      "learning_rate": 2.5495087924676254e-09,
+      "loss": 0.3143,
+      "step": 6356
+    },
+    {
+      "epoch": 2.9594972067039107,
+      "grad_norm": 0.60741428844414,
+      "learning_rate": 2.491904207774787e-09,
+      "loss": 0.3185,
+      "step": 6357
+    },
+    {
+      "epoch": 2.9599627560521418,
+      "grad_norm": 0.6068649737270883,
+      "learning_rate": 2.4349575261045645e-09,
+      "loss": 0.3339,
+      "step": 6358
+    },
+    {
+      "epoch": 2.9604283054003724,
+      "grad_norm": 0.608688919810352,
+      "learning_rate": 2.3786687624577363e-09,
+      "loss": 0.3294,
+      "step": 6359
+    },
+    {
+      "epoch": 2.9608938547486034,
+      "grad_norm": 0.6001708025099102,
+      "learning_rate": 2.323037931661887e-09,
+      "loss": 0.3344,
+      "step": 6360
+    },
+    {
+      "epoch": 2.961359404096834,
+      "grad_norm": 0.5992878056834328,
+      "learning_rate": 2.2680650483722364e-09,
+      "loss": 0.3308,
+      "step": 6361
+    },
+    {
+      "epoch": 2.961824953445065,
+      "grad_norm": 0.5944157016569404,
+      "learning_rate": 2.213750127069425e-09,
+      "loss": 0.3275,
+      "step": 6362
+    },
+    {
+      "epoch": 2.962290502793296,
+      "grad_norm": 0.5890227506635592,
+      "learning_rate": 2.160093182061174e-09,
+      "loss": 0.3373,
+      "step": 6363
+    },
+    {
+      "epoch": 2.9627560521415273,
+      "grad_norm": 0.5657645161731798,
+      "learning_rate": 2.1070942274825643e-09,
+      "loss": 0.3241,
+      "step": 6364
+    },
+    {
+      "epoch": 2.963221601489758,
+      "grad_norm": 0.6055081393263424,
+      "learning_rate": 2.054753277294097e-09,
+      "loss": 0.3296,
+      "step": 6365
+    },
+    {
+      "epoch": 2.963687150837989,
+      "grad_norm": 0.6343711743950271,
+      "learning_rate": 2.003070345284186e-09,
+      "loss": 0.3503,
+      "step": 6366
+    },
+    {
+      "epoch": 2.9641527001862196,
+      "grad_norm": 0.6038308809453933,
+      "learning_rate": 1.9520454450672187e-09,
+      "loss": 0.3485,
+      "step": 6367
+    },
+    {
+      "epoch": 2.9646182495344506,
+      "grad_norm": 0.6279287987814717,
+      "learning_rate": 1.9016785900841107e-09,
+      "loss": 0.3405,
+      "step": 6368
+    },
+    {
+      "epoch": 2.9650837988826817,
+      "grad_norm": 0.5738923869955052,
+      "learning_rate": 1.8519697936031366e-09,
+      "loss": 0.3175,
+      "step": 6369
+    },
+    {
+      "epoch": 2.9655493482309123,
+      "grad_norm": 0.617721704098387,
+      "learning_rate": 1.8029190687182673e-09,
+      "loss": 0.3381,
+      "step": 6370
+    },
+    {
+      "epoch": 2.9660148975791434,
+      "grad_norm": 0.6163293206988568,
+      "learning_rate": 1.7545264283508334e-09,
+      "loss": 0.332,
+      "step": 6371
+    },
+    {
+      "epoch": 2.9664804469273744,
+      "grad_norm": 0.5954695009248123,
+      "learning_rate": 1.7067918852486932e-09,
+      "loss": 0.3262,
+      "step": 6372
+    },
+    {
+      "epoch": 2.966945996275605,
+      "grad_norm": 0.6181012115338781,
+      "learning_rate": 1.6597154519862324e-09,
+      "loss": 0.3251,
+      "step": 6373
+    },
+    {
+      "epoch": 2.967411545623836,
+      "grad_norm": 0.6141183985473776,
+      "learning_rate": 1.6132971409643649e-09,
+      "loss": 0.3275,
+      "step": 6374
+    },
+    {
+      "epoch": 2.967877094972067,
+      "grad_norm": 0.6073772666136686,
+      "learning_rate": 1.567536964411087e-09,
+      "loss": 0.3206,
+      "step": 6375
+    },
+    {
+      "epoch": 2.968342644320298,
+      "grad_norm": 0.6435506507765253,
+      "learning_rate": 1.522434934380368e-09,
+      "loss": 0.3387,
+      "step": 6376
+    },
+    {
+      "epoch": 2.968808193668529,
+      "grad_norm": 0.6338778640783921,
+      "learning_rate": 1.477991062753259e-09,
+      "loss": 0.3446,
+      "step": 6377
+    },
+    {
+      "epoch": 2.9692737430167595,
+      "grad_norm": 0.6383160520172536,
+      "learning_rate": 1.4342053612376173e-09,
+      "loss": 0.337,
+      "step": 6378
+    },
+    {
+      "epoch": 2.9697392923649906,
+      "grad_norm": 0.5918040188014165,
+      "learning_rate": 1.391077841367272e-09,
+      "loss": 0.3415,
+      "step": 6379
+    },
+    {
+      "epoch": 2.9702048417132216,
+      "grad_norm": 0.6268540268186651,
+      "learning_rate": 1.3486085145028583e-09,
+      "loss": 0.321,
+      "step": 6380
+    },
+    {
+      "epoch": 2.9706703910614527,
+      "grad_norm": 0.6381924849025404,
+      "learning_rate": 1.3067973918326483e-09,
+      "loss": 0.3324,
+      "step": 6381
+    },
+    {
+      "epoch": 2.9711359404096833,
+      "grad_norm": 0.5793669798322505,
+      "learning_rate": 1.2656444843697769e-09,
+      "loss": 0.3261,
+      "step": 6382
+    },
+    {
+      "epoch": 2.9716014897579144,
+      "grad_norm": 0.6550325044117256,
+      "learning_rate": 1.2251498029552944e-09,
+      "loss": 0.3515,
+      "step": 6383
+    },
+    {
+      "epoch": 2.972067039106145,
+      "grad_norm": 0.6013968744725544,
+      "learning_rate": 1.1853133582565014e-09,
+      "loss": 0.3173,
+      "step": 6384
+    },
+    {
+      "epoch": 2.972532588454376,
+      "grad_norm": 0.6189581805899186,
+      "learning_rate": 1.1461351607672256e-09,
+      "loss": 0.3226,
+      "step": 6385
+    },
+    {
+      "epoch": 2.972998137802607,
+      "grad_norm": 0.5963453929038275,
+      "learning_rate": 1.1076152208081004e-09,
+      "loss": 0.3084,
+      "step": 6386
+    },
+    {
+      "epoch": 2.973463687150838,
+      "grad_norm": 0.6191807680823359,
+      "learning_rate": 1.0697535485257316e-09,
+      "loss": 0.3306,
+      "step": 6387
+    },
+    {
+      "epoch": 2.973929236499069,
+      "grad_norm": 0.5657529165163734,
+      "learning_rate": 1.0325501538940853e-09,
+      "loss": 0.3044,
+      "step": 6388
+    },
+    {
+      "epoch": 2.9743947858473,
+      "grad_norm": 0.6087389645789354,
+      "learning_rate": 9.96005046713655e-10,
+      "loss": 0.3497,
+      "step": 6389
+    },
+    {
+      "epoch": 2.9748603351955305,
+      "grad_norm": 0.607104177674844,
+      "learning_rate": 9.6011823661063e-10,
+      "loss": 0.3238,
+      "step": 6390
+    },
+    {
+      "epoch": 2.9753258845437616,
+      "grad_norm": 0.5940266354319524,
+      "learning_rate": 9.248897330391138e-10,
+      "loss": 0.3235,
+      "step": 6391
+    },
+    {
+      "epoch": 2.9757914338919926,
+      "grad_norm": 0.6002009031790284,
+      "learning_rate": 8.90319545278906e-10,
+      "loss": 0.3115,
+      "step": 6392
+    },
+    {
+      "epoch": 2.9762569832402237,
+      "grad_norm": 0.6056084845332135,
+      "learning_rate": 8.564076824363333e-10,
+      "loss": 0.3325,
+      "step": 6393
+    },
+    {
+      "epoch": 2.9767225325884543,
+      "grad_norm": 0.5741573874289685,
+      "learning_rate": 8.231541534450826e-10,
+      "loss": 0.3314,
+      "step": 6394
+    },
+    {
+      "epoch": 2.9771880819366854,
+      "grad_norm": 0.6276103225039416,
+      "learning_rate": 7.905589670645364e-10,
+      "loss": 0.3275,
+      "step": 6395
+    },
+    {
+      "epoch": 2.977653631284916,
+      "grad_norm": 0.6035812977555048,
+      "learning_rate": 7.586221318811592e-10,
+      "loss": 0.331,
+      "step": 6396
+    },
+    {
+      "epoch": 2.978119180633147,
+      "grad_norm": 0.6031188537899089,
+      "learning_rate": 7.273436563079439e-10,
+      "loss": 0.3273,
+      "step": 6397
+    },
+    {
+      "epoch": 2.978584729981378,
+      "grad_norm": 0.5849511573139736,
+      "learning_rate": 6.967235485844104e-10,
+      "loss": 0.3338,
+      "step": 6398
+    },
+    {
+      "epoch": 2.979050279329609,
+      "grad_norm": 0.6113366517964771,
+      "learning_rate": 6.66761816776329e-10,
+      "loss": 0.3352,
+      "step": 6399
+    },
+    {
+      "epoch": 2.97951582867784,
+      "grad_norm": 0.6437246115019234,
+      "learning_rate": 6.374584687768303e-10,
+      "loss": 0.3275,
+      "step": 6400
+    },
+    {
+      "epoch": 2.979981378026071,
+      "grad_norm": 0.6147829561175902,
+      "learning_rate": 6.088135123047401e-10,
+      "loss": 0.3335,
+      "step": 6401
+    },
+    {
+      "epoch": 2.9804469273743015,
+      "grad_norm": 0.5960892718271646,
+      "learning_rate": 5.808269549056888e-10,
+      "loss": 0.3278,
+      "step": 6402
+    },
+    {
+      "epoch": 2.9809124767225326,
+      "grad_norm": 0.584622864316521,
+      "learning_rate": 5.534988039523903e-10,
+      "loss": 0.3297,
+      "step": 6403
+    },
+    {
+      "epoch": 2.9813780260707636,
+      "grad_norm": 0.6107212180631427,
+      "learning_rate": 5.268290666438081e-10,
+      "loss": 0.3394,
+      "step": 6404
+    },
+    {
+      "epoch": 2.9818435754189943,
+      "grad_norm": 0.6210225546508152,
+      "learning_rate": 5.008177500048783e-10,
+      "loss": 0.329,
+      "step": 6405
+    },
+    {
+      "epoch": 2.9823091247672253,
+      "grad_norm": 0.6127923657789461,
+      "learning_rate": 4.754648608878976e-10,
+      "loss": 0.3401,
+      "step": 6406
+    },
+    {
+      "epoch": 2.9827746741154564,
+      "grad_norm": 0.5942185092275255,
+      "learning_rate": 4.507704059714124e-10,
+      "loss": 0.327,
+      "step": 6407
+    },
+    {
+      "epoch": 2.983240223463687,
+      "grad_norm": 0.6012123585507093,
+      "learning_rate": 4.2673439176049713e-10,
+      "loss": 0.3285,
+      "step": 6408
+    },
+    {
+      "epoch": 2.983705772811918,
+      "grad_norm": 0.5739890228926826,
+      "learning_rate": 4.033568245867536e-10,
+      "loss": 0.305,
+      "step": 6409
+    },
+    {
+      "epoch": 2.984171322160149,
+      "grad_norm": 0.5934931734878648,
+      "learning_rate": 3.8063771060831144e-10,
+      "loss": 0.336,
+      "step": 6410
+    },
+    {
+      "epoch": 2.9846368715083798,
+      "grad_norm": 0.6020717446951814,
+      "learning_rate": 3.5857705581038293e-10,
+      "loss": 0.3184,
+      "step": 6411
+    },
+    {
+      "epoch": 2.985102420856611,
+      "grad_norm": 0.5862604783444043,
+      "learning_rate": 3.371748660038754e-10,
+      "loss": 0.3072,
+      "step": 6412
+    },
+    {
+      "epoch": 2.9855679702048414,
+      "grad_norm": 0.6187771290763098,
+      "learning_rate": 3.1643114682650133e-10,
+      "loss": 0.3364,
+      "step": 6413
+    },
+    {
+      "epoch": 2.9860335195530725,
+      "grad_norm": 0.6179771820549566,
+      "learning_rate": 2.9634590374305607e-10,
+      "loss": 0.3372,
+      "step": 6414
+    },
+    {
+      "epoch": 2.9864990689013036,
+      "grad_norm": 0.586812116195312,
+      "learning_rate": 2.7691914204430736e-10,
+      "loss": 0.316,
+      "step": 6415
+    },
+    {
+      "epoch": 2.9869646182495346,
+      "grad_norm": 0.6006775307549337,
+      "learning_rate": 2.5815086684755077e-10,
+      "loss": 0.3308,
+      "step": 6416
+    },
+    {
+      "epoch": 2.9874301675977653,
+      "grad_norm": 0.6297962876834464,
+      "learning_rate": 2.400410830968869e-10,
+      "loss": 0.3346,
+      "step": 6417
+    },
+    {
+      "epoch": 2.9878957169459963,
+      "grad_norm": 0.6187264341265419,
+      "learning_rate": 2.225897955632217e-10,
+      "loss": 0.3401,
+      "step": 6418
+    },
+    {
+      "epoch": 2.988361266294227,
+      "grad_norm": 0.6192632987587624,
+      "learning_rate": 2.05797008843156e-10,
+      "loss": 0.3251,
+      "step": 6419
+    },
+    {
+      "epoch": 2.988826815642458,
+      "grad_norm": 0.5888697993301997,
+      "learning_rate": 1.896627273603735e-10,
+      "loss": 0.3144,
+      "step": 6420
+    },
+    {
+      "epoch": 2.989292364990689,
+      "grad_norm": 0.5833464067095125,
+      "learning_rate": 1.7418695536508544e-10,
+      "loss": 0.3278,
+      "step": 6421
+    },
+    {
+      "epoch": 2.98975791433892,
+      "grad_norm": 0.5883937381518596,
+      "learning_rate": 1.5936969693430838e-10,
+      "loss": 0.3165,
+      "step": 6422
+    },
+    {
+      "epoch": 2.9902234636871508,
+      "grad_norm": 0.5956579156602947,
+      "learning_rate": 1.4521095597075375e-10,
+      "loss": 0.3359,
+      "step": 6423
+    },
+    {
+      "epoch": 2.990689013035382,
+      "grad_norm": 0.574973815520477,
+      "learning_rate": 1.3171073620449338e-10,
+      "loss": 0.3037,
+      "step": 6424
+    },
+    {
+      "epoch": 2.9911545623836124,
+      "grad_norm": 0.6421420888173192,
+      "learning_rate": 1.1886904119184917e-10,
+      "loss": 0.3189,
+      "step": 6425
+    },
+    {
+      "epoch": 2.9916201117318435,
+      "grad_norm": 0.5984409107263292,
+      "learning_rate": 1.0668587431539313e-10,
+      "loss": 0.3326,
+      "step": 6426
+    },
+    {
+      "epoch": 2.9920856610800746,
+      "grad_norm": 0.6005657130238001,
+      "learning_rate": 9.516123878450245e-11,
+      "loss": 0.3447,
+      "step": 6427
+    },
+    {
+      "epoch": 2.9925512104283056,
+      "grad_norm": 0.5908647919845594,
+      "learning_rate": 8.429513763508202e-11,
+      "loss": 0.3196,
+      "step": 6428
+    },
+    {
+      "epoch": 2.9930167597765363,
+      "grad_norm": 0.5871208899267494,
+      "learning_rate": 7.408757372956432e-11,
+      "loss": 0.3352,
+      "step": 6429
+    },
+    {
+      "epoch": 2.9934823091247673,
+      "grad_norm": 0.585620730834881,
+      "learning_rate": 6.453854975690954e-11,
+      "loss": 0.3247,
+      "step": 6430
+    },
+    {
+      "epoch": 2.993947858472998,
+      "grad_norm": 0.606200963272576,
+      "learning_rate": 5.5648068232327937e-11,
+      "loss": 0.3138,
+      "step": 6431
+    },
+    {
+      "epoch": 2.994413407821229,
+      "grad_norm": 0.6229236537570065,
+      "learning_rate": 4.7416131498112525e-11,
+      "loss": 0.3429,
+      "step": 6432
+    },
+    {
+      "epoch": 2.99487895716946,
+      "grad_norm": 0.586912032812573,
+      "learning_rate": 3.984274172225133e-11,
+      "loss": 0.3127,
+      "step": 6433
+    },
+    {
+      "epoch": 2.995344506517691,
+      "grad_norm": 0.5875955644893867,
+      "learning_rate": 3.292790090037024e-11,
+      "loss": 0.3131,
+      "step": 6434
+    },
+    {
+      "epoch": 2.9958100558659218,
+      "grad_norm": 0.580801473664419,
+      "learning_rate": 2.6671610853790107e-11,
+      "loss": 0.3343,
+      "step": 6435
+    },
+    {
+      "epoch": 2.996275605214153,
+      "grad_norm": 0.6203046475763739,
+      "learning_rate": 2.1073873230359475e-11,
+      "loss": 0.3194,
+      "step": 6436
+    },
+    {
+      "epoch": 2.9967411545623834,
+      "grad_norm": 0.5937826930671019,
+      "learning_rate": 1.613468950473207e-11,
+      "loss": 0.3199,
+      "step": 6437
+    },
+    {
+      "epoch": 2.9972067039106145,
+      "grad_norm": 0.6309592585924947,
+      "learning_rate": 1.1854060978366833e-11,
+      "loss": 0.3493,
+      "step": 6438
+    },
+    {
+      "epoch": 2.9976722532588456,
+      "grad_norm": 0.6043810593755434,
+      "learning_rate": 8.231988778140132e-12,
+      "loss": 0.3473,
+      "step": 6439
+    },
+    {
+      "epoch": 2.998137802607076,
+      "grad_norm": 0.5664950641450852,
+      "learning_rate": 5.26847385884377e-12,
+      "loss": 0.3099,
+      "step": 6440
+    },
+    {
+      "epoch": 2.9986033519553073,
+      "grad_norm": 0.6149076563233128,
+      "learning_rate": 2.963517000964533e-12,
+      "loss": 0.323,
+      "step": 6441
+    },
+    {
+      "epoch": 2.9990689013035383,
+      "grad_norm": 0.6224628483416362,
+      "learning_rate": 1.3171188117944156e-12,
+      "loss": 0.3326,
+      "step": 6442
+    },
+    {
+      "epoch": 2.999534450651769,
+      "grad_norm": 0.625227240738905,
+      "learning_rate": 3.292797245979529e-13,
+      "loss": 0.3266,
+      "step": 6443
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.6996849935524039,
+      "learning_rate": 0.0,
+      "loss": 0.3138,
+      "step": 6444
+    },
+    {
+      "epoch": 3.0,
+      "step": 6444,
+      "total_flos": 4.271915416263066e+16,
+      "train_loss": 0.3954192769085303,
+      "train_runtime": 93126.1025,
+      "train_samples_per_second": 70.83,
+      "train_steps_per_second": 0.069
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 6444,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 250,
+  "total_flos": 4.271915416263066e+16,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}