{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9997152349311818,
  "eval_steps": 500,
  "global_step": 2633,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00037968675842429994,
      "grad_norm": 4.039186105207878,
      "learning_rate": 3.7878787878787882e-06,
      "loss": 3.1037,
      "step": 1
    },
    {
      "epoch": 0.0018984337921214998,
      "grad_norm": 3.978803446290656,
      "learning_rate": 1.893939393939394e-05,
      "loss": 3.0676,
      "step": 5
    },
    {
      "epoch": 0.0037968675842429997,
      "grad_norm": 3.4024459371591336,
      "learning_rate": 3.787878787878788e-05,
      "loss": 3.0058,
      "step": 10
    },
    {
      "epoch": 0.005695301376364499,
      "grad_norm": 0.769937135761119,
      "learning_rate": 5.681818181818182e-05,
      "loss": 2.7987,
      "step": 15
    },
    {
      "epoch": 0.007593735168485999,
      "grad_norm": 0.7081139470071718,
      "learning_rate": 7.575757575757576e-05,
      "loss": 2.6742,
      "step": 20
    },
    {
      "epoch": 0.009492168960607499,
      "grad_norm": 0.5663597400653296,
      "learning_rate": 9.46969696969697e-05,
      "loss": 2.5936,
      "step": 25
    },
    {
      "epoch": 0.011390602752728999,
      "grad_norm": 0.3713010333300208,
      "learning_rate": 0.00011363636363636364,
      "loss": 2.5012,
      "step": 30
    },
    {
      "epoch": 0.013289036544850499,
      "grad_norm": 0.4205375561732573,
      "learning_rate": 0.00013257575757575756,
      "loss": 2.433,
      "step": 35
    },
    {
      "epoch": 0.015187470336971999,
      "grad_norm": 0.587958832384919,
      "learning_rate": 0.00015151515151515152,
      "loss": 2.4168,
      "step": 40
    },
    {
      "epoch": 0.0170859041290935,
      "grad_norm": 0.6677663506298054,
      "learning_rate": 0.00017045454545454544,
      "loss": 2.3662,
      "step": 45
    },
    {
      "epoch": 0.018984337921214997,
      "grad_norm": 0.7064561950273625,
      "learning_rate": 0.0001893939393939394,
      "loss": 2.359,
      "step": 50
    },
    {
      "epoch": 0.0208827717133365,
      "grad_norm": 0.35917369578708264,
      "learning_rate": 0.00020833333333333335,
      "loss": 2.3354,
      "step": 55
    },
    {
      "epoch": 0.022781205505457997,
      "grad_norm": 0.39187189416502427,
      "learning_rate": 0.00022727272727272727,
      "loss": 2.2868,
      "step": 60
    },
    {
      "epoch": 0.024679639297579496,
      "grad_norm": 0.3470459858223155,
      "learning_rate": 0.0002462121212121212,
      "loss": 2.2955,
      "step": 65
    },
    {
      "epoch": 0.026578073089700997,
      "grad_norm": 0.6714289041162775,
      "learning_rate": 0.0002651515151515151,
      "loss": 2.2607,
      "step": 70
    },
    {
      "epoch": 0.028476506881822496,
      "grad_norm": 0.43889084397376665,
      "learning_rate": 0.00028409090909090913,
      "loss": 2.2774,
      "step": 75
    },
    {
      "epoch": 0.030374940673943997,
      "grad_norm": 0.35987531792307215,
      "learning_rate": 0.00030303030303030303,
      "loss": 2.2579,
      "step": 80
    },
    {
      "epoch": 0.0322733744660655,
      "grad_norm": 0.45583566123152425,
      "learning_rate": 0.000321969696969697,
      "loss": 2.2278,
      "step": 85
    },
    {
      "epoch": 0.034171808258187,
      "grad_norm": 0.5495955157286524,
      "learning_rate": 0.0003409090909090909,
      "loss": 2.2217,
      "step": 90
    },
    {
      "epoch": 0.036070242050308496,
      "grad_norm": 0.5181443647177656,
      "learning_rate": 0.0003598484848484849,
      "loss": 2.2183,
      "step": 95
    },
    {
      "epoch": 0.037968675842429994,
      "grad_norm": 0.969821145035109,
      "learning_rate": 0.0003787878787878788,
      "loss": 2.2112,
      "step": 100
    },
    {
      "epoch": 0.03986710963455149,
      "grad_norm": 0.2635501239040389,
      "learning_rate": 0.00039772727272727274,
      "loss": 2.2081,
      "step": 105
    },
    {
      "epoch": 0.041765543426673,
      "grad_norm": 0.5715349391832327,
      "learning_rate": 0.0004166666666666667,
      "loss": 2.1948,
      "step": 110
    },
    {
      "epoch": 0.043663977218794496,
      "grad_norm": 0.5248994310808021,
      "learning_rate": 0.0004356060606060606,
      "loss": 2.1855,
      "step": 115
    },
    {
      "epoch": 0.045562411010915994,
      "grad_norm": 0.5608654765973496,
      "learning_rate": 0.00045454545454545455,
      "loss": 2.1841,
      "step": 120
    },
    {
      "epoch": 0.04746084480303749,
      "grad_norm": 0.379235271981335,
      "learning_rate": 0.0004734848484848485,
      "loss": 2.1668,
      "step": 125
    },
    {
      "epoch": 0.04935927859515899,
      "grad_norm": 0.6934486639382003,
      "learning_rate": 0.0004924242424242425,
      "loss": 2.1799,
      "step": 130
    },
    {
      "epoch": 0.051257712387280496,
      "grad_norm": 0.32726936547789276,
      "learning_rate": 0.0005113636363636364,
      "loss": 2.1539,
      "step": 135
    },
    {
      "epoch": 0.053156146179401995,
      "grad_norm": 0.4103864066901133,
      "learning_rate": 0.0005303030303030302,
      "loss": 2.1559,
      "step": 140
    },
    {
      "epoch": 0.05505457997152349,
      "grad_norm": 0.6157950630213448,
      "learning_rate": 0.0005492424242424242,
      "loss": 2.1297,
      "step": 145
    },
    {
      "epoch": 0.05695301376364499,
      "grad_norm": 1.4548659360346616,
      "learning_rate": 0.0005681818181818183,
      "loss": 2.161,
      "step": 150
    },
    {
      "epoch": 0.05885144755576649,
      "grad_norm": 0.6872944738590898,
      "learning_rate": 0.0005871212121212122,
      "loss": 2.1462,
      "step": 155
    },
    {
      "epoch": 0.060749881347887995,
      "grad_norm": 1.1748112715233312,
      "learning_rate": 0.0006060606060606061,
      "loss": 2.1558,
      "step": 160
    },
    {
      "epoch": 0.06264831514000949,
      "grad_norm": 0.3342527623973018,
      "learning_rate": 0.000625,
      "loss": 2.1381,
      "step": 165
    },
    {
      "epoch": 0.064546748932131,
      "grad_norm": 0.3090958929703956,
      "learning_rate": 0.000643939393939394,
      "loss": 2.1311,
      "step": 170
    },
    {
      "epoch": 0.0664451827242525,
      "grad_norm": 0.4445838850369956,
      "learning_rate": 0.0006628787878787878,
      "loss": 2.109,
      "step": 175
    },
    {
      "epoch": 0.068343616516374,
      "grad_norm": 0.40447800943933176,
      "learning_rate": 0.0006818181818181818,
      "loss": 2.1338,
      "step": 180
    },
    {
      "epoch": 0.0702420503084955,
      "grad_norm": 0.5270438533785088,
      "learning_rate": 0.0007007575757575758,
      "loss": 2.1221,
      "step": 185
    },
    {
      "epoch": 0.07214048410061699,
      "grad_norm": 1.1988668583819744,
      "learning_rate": 0.0007196969696969698,
      "loss": 2.0986,
      "step": 190
    },
    {
      "epoch": 0.07403891789273849,
      "grad_norm": 0.5173087158849319,
      "learning_rate": 0.0007386363636363636,
      "loss": 2.097,
      "step": 195
    },
    {
      "epoch": 0.07593735168485999,
      "grad_norm": 0.8713149911927134,
      "learning_rate": 0.0007575757575757576,
      "loss": 2.1171,
      "step": 200
    },
    {
      "epoch": 0.07783578547698149,
      "grad_norm": 0.43763842661820435,
      "learning_rate": 0.0007765151515151515,
      "loss": 2.1231,
      "step": 205
    },
    {
      "epoch": 0.07973421926910298,
      "grad_norm": 0.37755079956108445,
      "learning_rate": 0.0007954545454545455,
      "loss": 2.0974,
      "step": 210
    },
    {
      "epoch": 0.08163265306122448,
      "grad_norm": 0.3202651052626521,
      "learning_rate": 0.0008143939393939394,
      "loss": 2.1061,
      "step": 215
    },
    {
      "epoch": 0.083531086853346,
      "grad_norm": 1.0697394978550765,
      "learning_rate": 0.0008333333333333334,
      "loss": 2.1058,
      "step": 220
    },
    {
      "epoch": 0.0854295206454675,
      "grad_norm": 0.9419134947997803,
      "learning_rate": 0.0008522727272727273,
      "loss": 2.0832,
      "step": 225
    },
    {
      "epoch": 0.08732795443758899,
      "grad_norm": 0.3177328584436459,
      "learning_rate": 0.0008712121212121212,
      "loss": 2.0936,
      "step": 230
    },
    {
      "epoch": 0.08922638822971049,
      "grad_norm": 0.4260910025738285,
      "learning_rate": 0.0008901515151515151,
      "loss": 2.0836,
      "step": 235
    },
    {
      "epoch": 0.09112482202183199,
      "grad_norm": 0.37197183427696595,
      "learning_rate": 0.0009090909090909091,
      "loss": 2.0705,
      "step": 240
    },
    {
      "epoch": 0.09302325581395349,
      "grad_norm": 0.3469428530219821,
      "learning_rate": 0.000928030303030303,
      "loss": 2.0851,
      "step": 245
    },
    {
      "epoch": 0.09492168960607499,
      "grad_norm": 0.5432864309293935,
      "learning_rate": 0.000946969696969697,
      "loss": 2.0601,
      "step": 250
    },
    {
      "epoch": 0.09682012339819648,
      "grad_norm": 0.30271756878451744,
      "learning_rate": 0.000965909090909091,
      "loss": 2.068,
      "step": 255
    },
    {
      "epoch": 0.09871855719031798,
      "grad_norm": 0.613099552832842,
      "learning_rate": 0.000984848484848485,
      "loss": 2.0889,
      "step": 260
    },
    {
      "epoch": 0.1006169909824395,
      "grad_norm": 0.2879329961644459,
      "learning_rate": 0.000999999560347478,
      "loss": 2.0728,
      "step": 265
    },
    {
      "epoch": 0.10251542477456099,
      "grad_norm": 0.3522360779077239,
      "learning_rate": 0.000999984172590384,
      "loss": 2.0778,
      "step": 270
    },
    {
      "epoch": 0.10441385856668249,
      "grad_norm": 0.8444049085073484,
      "learning_rate": 0.0009999468029803513,
      "loss": 2.0645,
      "step": 275
    },
    {
      "epoch": 0.10631229235880399,
      "grad_norm": 0.2340268385510596,
      "learning_rate": 0.0009998874531603381,
      "loss": 2.0424,
      "step": 280
    },
    {
      "epoch": 0.10821072615092549,
      "grad_norm": 0.3788090273684815,
      "learning_rate": 0.0009998061257396652,
      "loss": 2.0373,
      "step": 285
    },
    {
      "epoch": 0.11010915994304699,
      "grad_norm": 0.4712223089490736,
      "learning_rate": 0.0009997028242939002,
      "loss": 2.0535,
      "step": 290
    },
    {
      "epoch": 0.11200759373516848,
      "grad_norm": 0.5222779081868865,
      "learning_rate": 0.0009995775533647015,
      "loss": 2.0383,
      "step": 295
    },
    {
      "epoch": 0.11390602752728998,
      "grad_norm": 0.38807754757014135,
      "learning_rate": 0.0009994303184596178,
      "loss": 2.0536,
      "step": 300
    },
    {
      "epoch": 0.11580446131941148,
      "grad_norm": 0.5558959816060527,
      "learning_rate": 0.0009992611260518462,
      "loss": 2.0476,
      "step": 305
    },
    {
      "epoch": 0.11770289511153298,
      "grad_norm": 0.5297156243713772,
      "learning_rate": 0.0009990699835799469,
      "loss": 2.0182,
      "step": 310
    },
    {
      "epoch": 0.11960132890365449,
      "grad_norm": 0.8463883226148833,
      "learning_rate": 0.0009988568994475178,
      "loss": 2.0527,
      "step": 315
    },
    {
      "epoch": 0.12149976269577599,
      "grad_norm": 0.3181378787047677,
      "learning_rate": 0.0009986218830228234,
      "loss": 2.0253,
      "step": 320
    },
    {
      "epoch": 0.12339819648789749,
      "grad_norm": 0.3005739806261844,
      "learning_rate": 0.0009983649446383836,
      "loss": 2.023,
      "step": 325
    },
    {
      "epoch": 0.12529663028001897,
      "grad_norm": 0.7577732908772633,
      "learning_rate": 0.00099808609559052,
      "loss": 2.0081,
      "step": 330
    },
    {
      "epoch": 0.12719506407214048,
      "grad_norm": 0.3189810102860473,
      "learning_rate": 0.0009977853481388575,
      "loss": 2.0103,
      "step": 335
    },
    {
      "epoch": 0.129093497864262,
      "grad_norm": 0.6410939817239918,
      "learning_rate": 0.0009974627155057878,
      "loss": 2.0029,
      "step": 340
    },
    {
      "epoch": 0.13099193165638348,
      "grad_norm": 0.606178387164357,
      "learning_rate": 0.000997118211875886,
      "loss": 1.9762,
      "step": 345
    },
    {
      "epoch": 0.132890365448505,
      "grad_norm": 0.3588634601930927,
      "learning_rate": 0.0009967518523952875,
      "loss": 1.9911,
      "step": 350
    },
    {
      "epoch": 0.13478879924062648,
      "grad_norm": 0.28091052944488715,
      "learning_rate": 0.0009963636531710228,
      "loss": 1.9776,
      "step": 355
    },
    {
      "epoch": 0.136687233032748,
      "grad_norm": 0.38172615236997626,
      "learning_rate": 0.0009959536312703085,
      "loss": 1.9787,
      "step": 360
    },
    {
      "epoch": 0.13858566682486947,
      "grad_norm": 0.4360166845102846,
      "learning_rate": 0.0009955218047197978,
      "loss": 1.9909,
      "step": 365
    },
    {
      "epoch": 0.140484100616991,
      "grad_norm": 0.35275497258123356,
      "learning_rate": 0.000995068192504787,
      "loss": 1.9889,
      "step": 370
    },
    {
      "epoch": 0.14238253440911247,
      "grad_norm": 0.6317060282966646,
      "learning_rate": 0.0009945928145683814,
      "loss": 1.9551,
      "step": 375
    },
    {
      "epoch": 0.14428096820123398,
      "grad_norm": 0.3850672622738674,
      "learning_rate": 0.0009940956918106183,
      "loss": 1.9585,
      "step": 380
    },
    {
      "epoch": 0.1461794019933555,
      "grad_norm": 0.4220457511407704,
      "learning_rate": 0.0009935768460875483,
      "loss": 1.951,
      "step": 385
    },
    {
      "epoch": 0.14807783578547698,
      "grad_norm": 0.7783987596579233,
      "learning_rate": 0.0009930363002102743,
      "loss": 1.9724,
      "step": 390
    },
    {
      "epoch": 0.1499762695775985,
      "grad_norm": 0.32401219109209967,
      "learning_rate": 0.0009924740779439483,
      "loss": 1.9667,
      "step": 395
    },
    {
      "epoch": 0.15187470336971998,
      "grad_norm": 0.2857979164581651,
      "learning_rate": 0.0009918902040067276,
      "loss": 1.9462,
      "step": 400
    },
    {
      "epoch": 0.1537731371618415,
      "grad_norm": 0.6210777091831657,
      "learning_rate": 0.000991284704068686,
      "loss": 1.9618,
      "step": 405
    },
    {
      "epoch": 0.15567157095396297,
      "grad_norm": 0.4226037865904592,
      "learning_rate": 0.0009906576047506884,
      "loss": 1.9602,
      "step": 410
    },
    {
      "epoch": 0.15757000474608449,
      "grad_norm": 0.2508966539644492,
      "learning_rate": 0.0009900089336232166,
      "loss": 1.9518,
      "step": 415
    },
    {
      "epoch": 0.15946843853820597,
      "grad_norm": 0.28625033438978126,
      "learning_rate": 0.0009893387192051607,
      "loss": 1.937,
      "step": 420
    },
    {
      "epoch": 0.16136687233032748,
      "grad_norm": 0.5695158778386894,
      "learning_rate": 0.0009886469909625624,
      "loss": 1.9356,
      "step": 425
    },
    {
      "epoch": 0.16326530612244897,
      "grad_norm": 0.3444114501804719,
      "learning_rate": 0.0009879337793073219,
      "loss": 1.9292,
      "step": 430
    },
    {
      "epoch": 0.16516373991457048,
      "grad_norm": 0.3562145151770506,
      "learning_rate": 0.000987199115595859,
      "loss": 1.9446,
      "step": 435
    },
    {
      "epoch": 0.167062173706692,
      "grad_norm": 0.6662044045300556,
      "learning_rate": 0.0009864430321277354,
      "loss": 1.9123,
      "step": 440
    },
    {
      "epoch": 0.16896060749881348,
      "grad_norm": 0.4008170455675707,
      "learning_rate": 0.0009856655621442347,
      "loss": 1.9153,
      "step": 445
    },
    {
      "epoch": 0.170859041290935,
      "grad_norm": 0.5501520320546401,
      "learning_rate": 0.0009848667398269005,
      "loss": 1.9134,
      "step": 450
    },
    {
      "epoch": 0.17275747508305647,
      "grad_norm": 0.6229887015853178,
      "learning_rate": 0.000984046600296034,
      "loss": 1.9161,
      "step": 455
    },
    {
      "epoch": 0.17465590887517798,
      "grad_norm": 0.3263959879918121,
      "learning_rate": 0.0009832051796091496,
      "loss": 1.8977,
      "step": 460
    },
    {
      "epoch": 0.17655434266729947,
      "grad_norm": 0.32030075439432676,
      "learning_rate": 0.00098234251475939,
      "loss": 1.9385,
      "step": 465
    },
    {
      "epoch": 0.17845277645942098,
      "grad_norm": 0.3241713798628184,
      "learning_rate": 0.0009814586436738997,
      "loss": 1.8839,
      "step": 470
    },
    {
      "epoch": 0.18035121025154247,
      "grad_norm": 0.710599959000179,
      "learning_rate": 0.0009805536052121568,
      "loss": 1.8957,
      "step": 475
    },
    {
      "epoch": 0.18224964404366398,
      "grad_norm": 0.5685009329132018,
      "learning_rate": 0.000979627439164266,
      "loss": 1.9039,
      "step": 480
    },
    {
      "epoch": 0.1841480778357855,
      "grad_norm": 0.2883177054081686,
      "learning_rate": 0.0009786801862492075,
      "loss": 1.8939,
      "step": 485
    },
    {
      "epoch": 0.18604651162790697,
      "grad_norm": 0.3306718558813855,
      "learning_rate": 0.0009777118881130484,
      "loss": 1.8871,
      "step": 490
    },
    {
      "epoch": 0.1879449454200285,
      "grad_norm": 0.27771057670095894,
      "learning_rate": 0.000976722587327111,
      "loss": 1.8709,
      "step": 495
    },
    {
      "epoch": 0.18984337921214997,
      "grad_norm": 0.39767226694729324,
      "learning_rate": 0.0009757123273861006,
      "loss": 1.8707,
      "step": 500
    },
    {
      "epoch": 0.19174181300427148,
      "grad_norm": 0.3486146634838498,
      "learning_rate": 0.0009746811527061942,
      "loss": 1.854,
      "step": 505
    },
    {
      "epoch": 0.19364024679639297,
      "grad_norm": 0.4080559390296307,
      "learning_rate": 0.000973629108623087,
      "loss": 1.8698,
      "step": 510
    },
    {
      "epoch": 0.19553868058851448,
      "grad_norm": 0.33081602292257706,
      "learning_rate": 0.0009725562413900002,
      "loss": 1.8464,
      "step": 515
    },
    {
      "epoch": 0.19743711438063596,
      "grad_norm": 0.6037306322904349,
      "learning_rate": 0.000971462598175646,
      "loss": 1.8439,
      "step": 520
    },
    {
      "epoch": 0.19933554817275748,
      "grad_norm": 0.2879376497544211,
      "learning_rate": 0.0009703482270621553,
      "loss": 1.8713,
      "step": 525
    },
    {
      "epoch": 0.201233981964879,
      "grad_norm": 0.6788567524588369,
      "learning_rate": 0.0009692131770429629,
      "loss": 1.8453,
      "step": 530
    },
    {
      "epoch": 0.20313241575700047,
      "grad_norm": 0.3161414574102799,
      "learning_rate": 0.0009680574980206534,
      "loss": 1.8594,
      "step": 535
    },
    {
      "epoch": 0.20503084954912199,
      "grad_norm": 0.5645515053171333,
      "learning_rate": 0.0009668812408047678,
      "loss": 1.8492,
      "step": 540
    },
    {
      "epoch": 0.20692928334124347,
      "grad_norm": 0.5534701462686209,
      "learning_rate": 0.0009656844571095696,
      "loss": 1.8465,
      "step": 545
    },
    {
      "epoch": 0.20882771713336498,
      "grad_norm": 0.2769884694724677,
      "learning_rate": 0.0009644671995517705,
      "loss": 1.8512,
      "step": 550
    },
    {
      "epoch": 0.21072615092548647,
      "grad_norm": 0.3389677879460361,
      "learning_rate": 0.0009632295216482181,
      "loss": 1.8338,
      "step": 555
    },
    {
      "epoch": 0.21262458471760798,
      "grad_norm": 0.3665481797512662,
      "learning_rate": 0.000961971477813542,
      "loss": 1.818,
      "step": 560
    },
    {
      "epoch": 0.21452301850972946,
      "grad_norm": 0.5102461937389484,
      "learning_rate": 0.0009606931233577622,
      "loss": 1.8304,
      "step": 565
    },
    {
      "epoch": 0.21642145230185098,
      "grad_norm": 0.31696862342234344,
      "learning_rate": 0.0009593945144838571,
      "loss": 1.8491,
      "step": 570
    },
    {
      "epoch": 0.21831988609397246,
      "grad_norm": 0.474051209621524,
      "learning_rate": 0.0009580757082852929,
      "loss": 1.7842,
      "step": 575
    },
    {
      "epoch": 0.22021831988609397,
      "grad_norm": 0.28232130793234095,
      "learning_rate": 0.0009567367627435121,
      "loss": 1.8143,
      "step": 580
    },
    {
      "epoch": 0.22211675367821548,
      "grad_norm": 0.3429346941858972,
      "learning_rate": 0.0009553777367253867,
      "loss": 1.8144,
      "step": 585
    },
    {
      "epoch": 0.22401518747033697,
      "grad_norm": 0.4168698468178576,
      "learning_rate": 0.0009539986899806281,
      "loss": 1.8069,
      "step": 590
    },
    {
      "epoch": 0.22591362126245848,
      "grad_norm": 0.2584426546217752,
      "learning_rate": 0.0009525996831391607,
      "loss": 1.8121,
      "step": 595
    },
    {
      "epoch": 0.22781205505457996,
      "grad_norm": 0.27734758082061345,
      "learning_rate": 0.0009511807777084571,
      "loss": 1.8144,
      "step": 600
    },
    {
      "epoch": 0.22971048884670148,
      "grad_norm": 0.29149160986727496,
      "learning_rate": 0.0009497420360708331,
      "loss": 1.8072,
      "step": 605
    },
    {
      "epoch": 0.23160892263882296,
      "grad_norm": 0.27795468706904236,
      "learning_rate": 0.0009482835214807049,
      "loss": 1.7843,
      "step": 610
    },
    {
      "epoch": 0.23350735643094447,
      "grad_norm": 0.37668313657202124,
      "learning_rate": 0.0009468052980618091,
      "loss": 1.7858,
      "step": 615
    },
    {
      "epoch": 0.23540579022306596,
      "grad_norm": 0.3635355026922671,
      "learning_rate": 0.0009453074308043822,
      "loss": 1.7828,
      "step": 620
    },
    {
      "epoch": 0.23730422401518747,
      "grad_norm": 0.5492319693073675,
      "learning_rate": 0.0009437899855623046,
      "loss": 1.7805,
      "step": 625
    },
    {
      "epoch": 0.23920265780730898,
      "grad_norm": 0.5045520053744543,
      "learning_rate": 0.0009422530290502045,
      "loss": 1.784,
      "step": 630
    },
    {
      "epoch": 0.24110109159943047,
      "grad_norm": 0.29163131272581627,
      "learning_rate": 0.0009406966288405248,
      "loss": 1.7719,
      "step": 635
    },
    {
      "epoch": 0.24299952539155198,
      "grad_norm": 0.48505663312724284,
      "learning_rate": 0.0009391208533605527,
      "loss": 1.7866,
      "step": 640
    },
    {
      "epoch": 0.24489795918367346,
      "grad_norm": 0.5368154913313934,
      "learning_rate": 0.0009375257718894107,
      "loss": 1.7909,
      "step": 645
    },
    {
      "epoch": 0.24679639297579498,
      "grad_norm": 0.35856621494537494,
      "learning_rate": 0.0009359114545550116,
      "loss": 1.7504,
      "step": 650
    },
    {
      "epoch": 0.24869482676791646,
      "grad_norm": 0.3659739590782703,
      "learning_rate": 0.0009342779723309745,
      "loss": 1.7605,
      "step": 655
    },
    {
      "epoch": 0.25059326056003794,
      "grad_norm": 0.511785026274882,
      "learning_rate": 0.0009326253970335046,
      "loss": 1.7158,
      "step": 660
    },
    {
      "epoch": 0.25249169435215946,
      "grad_norm": 0.30511971139113303,
      "learning_rate": 0.0009309538013182364,
      "loss": 1.7637,
      "step": 665
    },
    {
      "epoch": 0.25439012814428097,
      "grad_norm": 0.28292845629768315,
      "learning_rate": 0.0009292632586770384,
      "loss": 1.7601,
      "step": 670
    },
    {
      "epoch": 0.2562885619364025,
      "grad_norm": 0.24747199414214469,
      "learning_rate": 0.000927553843434783,
      "loss": 1.7477,
      "step": 675
    },
    {
      "epoch": 0.258186995728524,
      "grad_norm": 0.30515982384727836,
      "learning_rate": 0.0009258256307460781,
      "loss": 1.757,
      "step": 680
    },
    {
      "epoch": 0.26008542952064545,
      "grad_norm": 0.4086074740784753,
      "learning_rate": 0.000924078696591963,
      "loss": 1.7437,
      "step": 685
    },
    {
      "epoch": 0.26198386331276696,
      "grad_norm": 0.3606712756700905,
      "learning_rate": 0.0009223131177765685,
      "loss": 1.7156,
      "step": 690
    },
    {
      "epoch": 0.2638822971048885,
      "grad_norm": 0.31644420804415135,
      "learning_rate": 0.0009205289719237393,
      "loss": 1.738,
      "step": 695
    },
    {
      "epoch": 0.26578073089701,
      "grad_norm": 0.47740932792504026,
      "learning_rate": 0.0009187263374736221,
      "loss": 1.7548,
      "step": 700
    },
    {
      "epoch": 0.26767916468913144,
      "grad_norm": 0.2710760174578327,
      "learning_rate": 0.0009169052936792164,
      "loss": 1.7425,
      "step": 705
    },
    {
      "epoch": 0.26957759848125296,
      "grad_norm": 0.27636232385404463,
      "learning_rate": 0.0009150659206028904,
      "loss": 1.7267,
      "step": 710
    },
    {
      "epoch": 0.27147603227337447,
      "grad_norm": 0.4625279056619356,
      "learning_rate": 0.0009132082991128607,
      "loss": 1.7387,
      "step": 715
    },
    {
      "epoch": 0.273374466065496,
      "grad_norm": 4.000898556896865,
      "learning_rate": 0.0009113325108796374,
      "loss": 1.745,
      "step": 720
    },
    {
      "epoch": 0.2752728998576175,
      "grad_norm": 0.38141390080270937,
      "learning_rate": 0.0009094386383724332,
      "loss": 1.7522,
      "step": 725
    },
    {
      "epoch": 0.27717133364973895,
      "grad_norm": 0.35177413794126794,
      "learning_rate": 0.0009075267648555378,
      "loss": 1.6868,
      "step": 730
    },
    {
      "epoch": 0.27906976744186046,
      "grad_norm": 0.8220268286150999,
      "learning_rate": 0.000905596974384657,
      "loss": 1.8246,
      "step": 735
    },
    {
      "epoch": 0.280968201233982,
      "grad_norm": 0.8599631648767855,
      "learning_rate": 0.0009036493518032172,
      "loss": 1.738,
      "step": 740
    },
    {
      "epoch": 0.2828666350261035,
      "grad_norm": 1.0678364091016483,
      "learning_rate": 0.0009016839827386351,
      "loss": 1.7384,
      "step": 745
    },
    {
      "epoch": 0.28476506881822494,
      "grad_norm": 1.8720283645771043,
      "learning_rate": 0.0008997009535985534,
      "loss": 1.8852,
      "step": 750
    },
    {
      "epoch": 0.28666350261034645,
      "grad_norm": 1.1071369931471309,
      "learning_rate": 0.0008977003515670418,
      "loss": 1.7642,
      "step": 755
    },
    {
      "epoch": 0.28856193640246797,
      "grad_norm": 0.9209513939367386,
      "learning_rate": 0.0008956822646007639,
      "loss": 1.8114,
      "step": 760
    },
    {
      "epoch": 0.2904603701945895,
      "grad_norm": 0.4031852696574249,
      "learning_rate": 0.0008936467814251102,
      "loss": 1.749,
      "step": 765
    },
    {
      "epoch": 0.292358803986711,
      "grad_norm": 0.482604929605048,
      "learning_rate": 0.0008915939915302969,
      "loss": 1.6956,
      "step": 770
    },
    {
      "epoch": 0.29425723777883245,
      "grad_norm": 0.31688589446500226,
      "learning_rate": 0.000889523985167432,
      "loss": 1.7137,
      "step": 775
    },
    {
      "epoch": 0.29615567157095396,
      "grad_norm": 0.2662426574337696,
      "learning_rate": 0.0008874368533445476,
      "loss": 1.716,
      "step": 780
    },
    {
      "epoch": 0.29805410536307547,
      "grad_norm": 0.30925480824258833,
      "learning_rate": 0.0008853326878225978,
      "loss": 1.6941,
      "step": 785
    },
    {
      "epoch": 0.299952539155197,
      "grad_norm": 0.3332600934262706,
      "learning_rate": 0.000883211581111425,
      "loss": 1.673,
      "step": 790
    },
    {
      "epoch": 0.30185097294731844,
      "grad_norm": 0.4173857531935143,
      "learning_rate": 0.0008810736264656929,
      "loss": 1.677,
      "step": 795
    },
    {
      "epoch": 0.30374940673943995,
      "grad_norm": 0.3528547476107368,
      "learning_rate": 0.0008789189178807862,
      "loss": 1.6878,
      "step": 800
    },
    {
      "epoch": 0.30564784053156147,
      "grad_norm": 0.35241925687665077,
      "learning_rate": 0.0008767475500886777,
      "loss": 1.6895,
      "step": 805
    },
    {
      "epoch": 0.307546274323683,
      "grad_norm": 0.3235829351373288,
      "learning_rate": 0.0008745596185537648,
      "loss": 1.6903,
      "step": 810
    },
    {
      "epoch": 0.30944470811580443,
      "grad_norm": 0.2948128495705965,
      "learning_rate": 0.0008723552194686705,
      "loss": 1.6609,
      "step": 815
    },
    {
      "epoch": 0.31134314190792595,
      "grad_norm": 0.36422304971890607,
      "learning_rate": 0.0008701344497500159,
      "loss": 1.6614,
      "step": 820
    },
    {
      "epoch": 0.31324157570004746,
      "grad_norm": 0.2825304340578671,
      "learning_rate": 0.0008678974070341584,
      "loss": 1.6769,
      "step": 825
    },
    {
      "epoch": 0.31514000949216897,
      "grad_norm": 0.26538936909594424,
      "learning_rate": 0.0008656441896728994,
      "loss": 1.6485,
      "step": 830
    },
    {
      "epoch": 0.3170384432842905,
      "grad_norm": 0.31905462352993175,
      "learning_rate": 0.0008633748967291598,
      "loss": 1.6482,
      "step": 835
    },
    {
      "epoch": 0.31893687707641194,
      "grad_norm": 0.39448662981546256,
      "learning_rate": 0.0008610896279726255,
      "loss": 1.6259,
      "step": 840
    },
    {
      "epoch": 0.32083531086853345,
      "grad_norm": 0.31299399145300977,
      "learning_rate": 0.0008587884838753603,
      "loss": 1.6398,
      "step": 845
    },
    {
      "epoch": 0.32273374466065496,
      "grad_norm": 0.3247022426656648,
      "learning_rate": 0.0008564715656073893,
      "loss": 1.6501,
      "step": 850
    },
    {
      "epoch": 0.3246321784527765,
      "grad_norm": 0.2859709456607949,
      "learning_rate": 0.0008541389750322498,
      "loss": 1.6268,
      "step": 855
    },
    {
      "epoch": 0.32653061224489793,
      "grad_norm": 0.28129496647987356,
      "learning_rate": 0.0008517908147025143,
      "loss": 1.6085,
      "step": 860
    },
    {
      "epoch": 0.32842904603701945,
      "grad_norm": 0.3456433785224482,
      "learning_rate": 0.0008494271878552808,
      "loss": 1.6285,
      "step": 865
    },
    {
      "epoch": 0.33032747982914096,
      "grad_norm": 0.49566297259294595,
      "learning_rate": 0.0008470481984076344,
      "loss": 1.6269,
      "step": 870
    },
    {
      "epoch": 0.33222591362126247,
      "grad_norm": 0.34226796976284596,
      "learning_rate": 0.000844653950952078,
      "loss": 1.5934,
      "step": 875
    },
    {
      "epoch": 0.334124347413384,
      "grad_norm": 0.2874518927936025,
      "learning_rate": 0.000842244550751935,
      "loss": 1.6233,
      "step": 880
    },
    {
      "epoch": 0.33602278120550544,
      "grad_norm": 0.27909933392041014,
      "learning_rate": 0.0008398201037367201,
      "loss": 1.5986,
      "step": 885
    },
    {
      "epoch": 0.33792121499762695,
      "grad_norm": 0.2755722277474001,
      "learning_rate": 0.0008373807164974832,
      "loss": 1.5831,
      "step": 890
    },
    {
      "epoch": 0.33981964878974846,
      "grad_norm": 0.2647419405762948,
      "learning_rate": 0.0008349264962821219,
      "loss": 1.5928,
      "step": 895
    },
    {
      "epoch": 0.34171808258187,
      "grad_norm": 0.34636953235762286,
      "learning_rate": 0.0008324575509906677,
      "loss": 1.5908,
      "step": 900
    },
    {
      "epoch": 0.34361651637399143,
      "grad_norm": 0.3152387214237848,
      "learning_rate": 0.0008299739891705413,
      "loss": 1.5952,
      "step": 905
    },
    {
      "epoch": 0.34551495016611294,
      "grad_norm": 0.2654394015701188,
      "learning_rate": 0.0008274759200117803,
      "loss": 1.5985,
      "step": 910
    },
    {
      "epoch": 0.34741338395823446,
      "grad_norm": 0.6103144176291811,
      "learning_rate": 0.0008249634533422392,
      "loss": 1.5877,
      "step": 915
    },
    {
      "epoch": 0.34931181775035597,
      "grad_norm": 0.3224755104739263,
      "learning_rate": 0.0008224366996227604,
      "loss": 1.5771,
      "step": 920
    },
    {
      "epoch": 0.3512102515424775,
      "grad_norm": 0.5837396991238525,
      "learning_rate": 0.0008198957699423175,
      "loss": 1.5843,
      "step": 925
    },
    {
      "epoch": 0.35310868533459894,
      "grad_norm": 0.32548897062944837,
      "learning_rate": 0.000817340776013132,
      "loss": 1.5851,
      "step": 930
    },
    {
      "epoch": 0.35500711912672045,
      "grad_norm": 0.31829180968967913,
      "learning_rate": 0.0008147718301657612,
      "loss": 1.5942,
      "step": 935
    },
    {
      "epoch": 0.35690555291884196,
      "grad_norm": 0.2731850877543467,
      "learning_rate": 0.0008121890453441602,
      "loss": 1.5694,
      "step": 940
    },
    {
      "epoch": 0.3588039867109635,
      "grad_norm": 0.24810235775271594,
      "learning_rate": 0.0008095925351007156,
      "loss": 1.559,
      "step": 945
    },
    {
      "epoch": 0.36070242050308493,
      "grad_norm": 0.28169507454726866,
      "learning_rate": 0.0008069824135912536,
      "loss": 1.5638,
      "step": 950
    },
    {
      "epoch": 0.36260085429520644,
      "grad_norm": 0.24752038600316323,
      "learning_rate": 0.0008043587955700211,
      "loss": 1.5718,
      "step": 955
    },
    {
      "epoch": 0.36449928808732796,
      "grad_norm": 0.27876047342351895,
      "learning_rate": 0.0008017217963846403,
      "loss": 1.5501,
      "step": 960
    },
    {
      "epoch": 0.36639772187944947,
      "grad_norm": 0.40897383234845613,
      "learning_rate": 0.0007990715319710381,
      "loss": 1.5306,
      "step": 965
    },
    {
      "epoch": 0.368296155671571,
      "grad_norm": 0.2858525738953562,
      "learning_rate": 0.0007964081188483476,
      "loss": 1.5283,
      "step": 970
    },
    {
      "epoch": 0.37019458946369244,
      "grad_norm": 0.26144274136184015,
      "learning_rate": 0.0007937316741137871,
      "loss": 1.5446,
      "step": 975
    },
    {
      "epoch": 0.37209302325581395,
      "grad_norm": 0.32273246789256727,
      "learning_rate": 0.0007910423154375101,
      "loss": 1.5672,
      "step": 980
    },
    {
      "epoch": 0.37399145704793546,
      "grad_norm": 0.3490759385530748,
      "learning_rate": 0.0007883401610574337,
      "loss": 1.5401,
      "step": 985
    },
    {
      "epoch": 0.375889890840057,
      "grad_norm": 0.28213557009744067,
      "learning_rate": 0.0007856253297740383,
      "loss": 1.5533,
      "step": 990
    },
    {
      "epoch": 0.37778832463217843,
      "grad_norm": 0.33445422182184414,
      "learning_rate": 0.0007828979409451468,
      "loss": 1.5361,
      "step": 995
    },
    {
      "epoch": 0.37968675842429994,
      "grad_norm": 0.3535301287121482,
      "learning_rate": 0.0007801581144806751,
      "loss": 1.5189,
      "step": 1000
    },
    {
      "epoch": 0.38158519221642145,
      "grad_norm": 0.3391675710324959,
      "learning_rate": 0.0007774059708373606,
      "loss": 1.5301,
      "step": 1005
    },
    {
      "epoch": 0.38348362600854297,
      "grad_norm": 0.6452174423374145,
      "learning_rate": 0.0007746416310134679,
      "loss": 1.5214,
      "step": 1010
    },
    {
      "epoch": 0.3853820598006645,
      "grad_norm": 0.43341843816494785,
      "learning_rate": 0.0007718652165434664,
      "loss": 1.5235,
      "step": 1015
    },
    {
      "epoch": 0.38728049359278593,
      "grad_norm": 0.3003373383959457,
      "learning_rate": 0.0007690768494926897,
      "loss": 1.5063,
      "step": 1020
    },
    {
      "epoch": 0.38917892738490745,
      "grad_norm": 0.3078153548021384,
      "learning_rate": 0.000766276652451967,
      "loss": 1.5174,
      "step": 1025
    },
    {
      "epoch": 0.39107736117702896,
      "grad_norm": 0.4798286034200326,
      "learning_rate": 0.0007634647485322347,
      "loss": 1.4958,
      "step": 1030
    },
    {
      "epoch": 0.39297579496915047,
      "grad_norm": 0.2962615700403029,
      "learning_rate": 0.0007606412613591228,
      "loss": 1.5187,
      "step": 1035
    },
    {
      "epoch": 0.39487422876127193,
      "grad_norm": 0.3904932019460365,
      "learning_rate": 0.0007578063150675206,
      "loss": 1.4828,
      "step": 1040
    },
    {
      "epoch": 0.39677266255339344,
      "grad_norm": 0.31811472424437215,
      "learning_rate": 0.0007549600342961183,
      "loss": 1.4752,
      "step": 1045
    },
    {
      "epoch": 0.39867109634551495,
      "grad_norm": 0.4803758991042241,
      "learning_rate": 0.0007521025441819278,
      "loss": 1.4969,
      "step": 1050
    },
    {
      "epoch": 0.40056953013763646,
      "grad_norm": 0.472851525192678,
      "learning_rate": 0.0007492339703547808,
      "loss": 1.492,
      "step": 1055
    },
    {
      "epoch": 0.402467963929758,
      "grad_norm": 0.3908915345807985,
      "learning_rate": 0.0007463544389318058,
      "loss": 1.4682,
      "step": 1060
    },
    {
      "epoch": 0.40436639772187943,
      "grad_norm": 0.3886806997641989,
      "learning_rate": 0.0007434640765118824,
      "loss": 1.4759,
      "step": 1065
    },
    {
      "epoch": 0.40626483151400095,
      "grad_norm": 0.5234433759967211,
      "learning_rate": 0.0007405630101700769,
      "loss": 1.4779,
      "step": 1070
    },
    {
      "epoch": 0.40816326530612246,
      "grad_norm": 0.3607186516700363,
      "learning_rate": 0.0007376513674520537,
      "loss": 1.4894,
      "step": 1075
    },
    {
      "epoch": 0.41006169909824397,
      "grad_norm": 0.3422699133158823,
      "learning_rate": 0.0007347292763684691,
      "loss": 1.4713,
      "step": 1080
    },
    {
      "epoch": 0.4119601328903654,
      "grad_norm": 0.456283403837333,
      "learning_rate": 0.000731796865389343,
      "loss": 1.4786,
      "step": 1085
    },
    {
      "epoch": 0.41385856668248694,
      "grad_norm": 0.29907259245021056,
      "learning_rate": 0.0007288542634384101,
      "loss": 1.4428,
      "step": 1090
    },
    {
      "epoch": 0.41575700047460845,
      "grad_norm": 0.4065968463089126,
      "learning_rate": 0.0007259015998874521,
      "loss": 1.4551,
      "step": 1095
    },
    {
      "epoch": 0.41765543426672996,
      "grad_norm": 0.2851166065497382,
      "learning_rate": 0.0007229390045506107,
      "loss": 1.4726,
      "step": 1100
    },
    {
      "epoch": 0.4195538680588514,
      "grad_norm": 0.34495921420404607,
      "learning_rate": 0.0007199666076786786,
      "loss": 1.4491,
      "step": 1105
    },
    {
      "epoch": 0.42145230185097293,
      "grad_norm": 0.2657939111249429,
      "learning_rate": 0.0007169845399533742,
      "loss": 1.4362,
      "step": 1110
    },
    {
      "epoch": 0.42335073564309444,
      "grad_norm": 0.4937795186011948,
      "learning_rate": 0.0007139929324815965,
      "loss": 1.4407,
      "step": 1115
    },
    {
      "epoch": 0.42524916943521596,
      "grad_norm": 0.4407673687752531,
      "learning_rate": 0.0007109919167896597,
      "loss": 1.4319,
      "step": 1120
    },
    {
      "epoch": 0.42714760322733747,
      "grad_norm": 0.5300468580153148,
      "learning_rate": 0.0007079816248175114,
      "loss": 1.4275,
      "step": 1125
    },
    {
      "epoch": 0.4290460370194589,
      "grad_norm": 0.3241335090242318,
      "learning_rate": 0.000704962188912932,
      "loss": 1.4426,
      "step": 1130
    },
    {
      "epoch": 0.43094447081158044,
      "grad_norm": 0.3968224596845394,
      "learning_rate": 0.0007019337418257159,
      "loss": 1.4398,
      "step": 1135
    },
    {
      "epoch": 0.43284290460370195,
      "grad_norm": 0.3781038122854968,
      "learning_rate": 0.0006988964167018346,
      "loss": 1.4298,
      "step": 1140
    },
    {
      "epoch": 0.43474133839582346,
      "grad_norm": 0.2842174258844352,
      "learning_rate": 0.0006958503470775836,
      "loss": 1.4358,
      "step": 1145
    },
    {
      "epoch": 0.4366397721879449,
      "grad_norm": 0.5316938383381906,
      "learning_rate": 0.0006927956668737115,
      "loss": 1.4152,
      "step": 1150
    },
    {
      "epoch": 0.43853820598006643,
      "grad_norm": 0.26129078645372833,
      "learning_rate": 0.000689732510389531,
      "loss": 1.4117,
      "step": 1155
    },
    {
      "epoch": 0.44043663977218794,
      "grad_norm": 0.45097710133639385,
      "learning_rate": 0.0006866610122970162,
      "loss": 1.3814,
      "step": 1160
    },
    {
      "epoch": 0.44233507356430946,
      "grad_norm": 0.34768384716485046,
      "learning_rate": 0.0006835813076348805,
      "loss": 1.409,
      "step": 1165
    },
    {
      "epoch": 0.44423350735643097,
      "grad_norm": 0.6973633265305714,
      "learning_rate": 0.0006804935318026396,
      "loss": 1.4211,
      "step": 1170
    },
    {
      "epoch": 0.4461319411485524,
      "grad_norm": 0.4342853868362056,
      "learning_rate": 0.0006773978205546597,
      "loss": 1.4037,
      "step": 1175
    },
    {
      "epoch": 0.44803037494067394,
      "grad_norm": 0.46738592482568664,
      "learning_rate": 0.0006742943099941876,
      "loss": 1.4309,
      "step": 1180
    },
    {
      "epoch": 0.44992880873279545,
      "grad_norm": 0.25387391273809895,
      "learning_rate": 0.000671183136567368,
      "loss": 1.3923,
      "step": 1185
    },
    {
      "epoch": 0.45182724252491696,
      "grad_norm": 0.46112835773656796,
      "learning_rate": 0.0006680644370572444,
      "loss": 1.4028,
      "step": 1190
    },
    {
      "epoch": 0.4537256763170384,
      "grad_norm": 0.2551607349467472,
      "learning_rate": 0.0006649383485777449,
      "loss": 1.3815,
      "step": 1195
    },
    {
      "epoch": 0.45562411010915993,
      "grad_norm": 0.37310127138077576,
      "learning_rate": 0.0006618050085676546,
      "loss": 1.3911,
      "step": 1200
    },
    {
      "epoch": 0.45752254390128144,
      "grad_norm": 0.4254847053830724,
      "learning_rate": 0.0006586645547845729,
      "loss": 1.3929,
      "step": 1205
    },
    {
      "epoch": 0.45942097769340295,
      "grad_norm": 0.7687175874517974,
      "learning_rate": 0.0006555171252988568,
      "loss": 1.3733,
      "step": 1210
    },
    {
      "epoch": 0.46131941148552447,
      "grad_norm": 0.3146929888691999,
      "learning_rate": 0.0006523628584875507,
      "loss": 1.3711,
      "step": 1215
    },
    {
      "epoch": 0.4632178452776459,
      "grad_norm": 0.27183821399250013,
      "learning_rate": 0.0006492018930283026,
      "loss": 1.3455,
      "step": 1220
    },
    {
      "epoch": 0.46511627906976744,
      "grad_norm": 0.31617364418439614,
      "learning_rate": 0.000646034367893268,
      "loss": 1.3619,
      "step": 1225
    },
    {
      "epoch": 0.46701471286188895,
      "grad_norm": 0.3833518263216662,
      "learning_rate": 0.0006428604223429979,
      "loss": 1.3452,
      "step": 1230
    },
    {
      "epoch": 0.46891314665401046,
      "grad_norm": 0.3267810331058041,
      "learning_rate": 0.0006396801959203186,
      "loss": 1.3627,
      "step": 1235
    },
    {
      "epoch": 0.4708115804461319,
      "grad_norm": 0.34041494057204974,
      "learning_rate": 0.0006364938284441949,
      "loss": 1.3418,
      "step": 1240
    },
    {
      "epoch": 0.47271001423825343,
      "grad_norm": 0.42237475709823624,
      "learning_rate": 0.0006333014600035838,
      "loss": 1.3594,
      "step": 1245
    },
    {
      "epoch": 0.47460844803037494,
      "grad_norm": 0.4449414736784289,
      "learning_rate": 0.0006301032309512754,
      "loss": 1.3293,
      "step": 1250
    },
    {
      "epoch": 0.47650688182249645,
      "grad_norm": 0.3046879446109181,
      "learning_rate": 0.0006268992818977221,
      "loss": 1.3438,
      "step": 1255
    },
    {
      "epoch": 0.47840531561461797,
      "grad_norm": 0.3963369420982114,
      "learning_rate": 0.0006236897537048566,
      "loss": 1.3418,
      "step": 1260
    },
    {
      "epoch": 0.4803037494067394,
      "grad_norm": 0.449251264603866,
      "learning_rate": 0.0006204747874798993,
      "loss": 1.3443,
      "step": 1265
    },
    {
      "epoch": 0.48220218319886093,
      "grad_norm": 0.30313300002195165,
      "learning_rate": 0.0006172545245691538,
      "loss": 1.3414,
      "step": 1270
    },
    {
      "epoch": 0.48410061699098245,
      "grad_norm": 0.33417145989869673,
      "learning_rate": 0.0006140291065517931,
      "loss": 1.3179,
      "step": 1275
    },
    {
      "epoch": 0.48599905078310396,
      "grad_norm": 0.3904784303862518,
      "learning_rate": 0.0006107986752336357,
      "loss": 1.3411,
      "step": 1280
    },
    {
      "epoch": 0.4878974845752254,
      "grad_norm": 0.3222639679317221,
      "learning_rate": 0.0006075633726409091,
      "loss": 1.3086,
      "step": 1285
    },
    {
      "epoch": 0.4897959183673469,
      "grad_norm": 0.3288048359935024,
      "learning_rate": 0.0006043233410140076,
      "loss": 1.3305,
      "step": 1290
    },
    {
      "epoch": 0.49169435215946844,
      "grad_norm": 0.2937886282400547,
      "learning_rate": 0.0006010787228012384,
      "loss": 1.3029,
      "step": 1295
    },
    {
      "epoch": 0.49359278595158995,
      "grad_norm": 0.29213405752561167,
      "learning_rate": 0.0005978296606525572,
      "loss": 1.3161,
      "step": 1300
    },
    {
      "epoch": 0.49549121974371146,
      "grad_norm": 0.34322670276805756,
      "learning_rate": 0.0005945762974132986,
      "loss": 1.3281,
      "step": 1305
    },
    {
      "epoch": 0.4973896535358329,
      "grad_norm": 0.28223010113720515,
      "learning_rate": 0.0005913187761178951,
      "loss": 1.2943,
      "step": 1310
    },
    {
      "epoch": 0.49928808732795443,
      "grad_norm": 0.29364231136684454,
      "learning_rate": 0.0005880572399835881,
      "loss": 1.2988,
      "step": 1315
    },
    {
      "epoch": 0.5011865211200759,
      "grad_norm": 0.3396537753038387,
      "learning_rate": 0.0005847918324041324,
      "loss": 1.3053,
      "step": 1320
    },
    {
      "epoch": 0.5030849549121974,
      "grad_norm": 0.3515005890654536,
      "learning_rate": 0.0005815226969434903,
      "loss": 1.2881,
      "step": 1325
    },
    {
      "epoch": 0.5049833887043189,
      "grad_norm": 0.32308285051579544,
      "learning_rate": 0.0005782499773295219,
      "loss": 1.31,
      "step": 1330
    },
    {
      "epoch": 0.5068818224964404,
      "grad_norm": 0.32615560955136413,
      "learning_rate": 0.0005749738174476639,
      "loss": 1.2659,
      "step": 1335
    },
    {
      "epoch": 0.5087802562885619,
      "grad_norm": 0.3566565149320192,
      "learning_rate": 0.0005716943613346059,
      "loss": 1.2746,
      "step": 1340
    },
    {
      "epoch": 0.5106786900806835,
      "grad_norm": 0.4244945485487363,
      "learning_rate": 0.0005684117531719551,
      "loss": 1.2741,
      "step": 1345
    },
    {
      "epoch": 0.512577123872805,
      "grad_norm": 0.31519325513997737,
      "learning_rate": 0.0005651261372799002,
      "loss": 1.2773,
      "step": 1350
    },
    {
      "epoch": 0.5144755576649265,
      "grad_norm": 0.4339192027972204,
      "learning_rate": 0.0005618376581108647,
      "loss": 1.2669,
      "step": 1355
    },
    {
      "epoch": 0.516373991457048,
      "grad_norm": 0.3419108710774503,
      "learning_rate": 0.0005585464602431556,
      "loss": 1.2508,
      "step": 1360
    },
    {
      "epoch": 0.5182724252491694,
      "grad_norm": 0.38978151919816467,
      "learning_rate": 0.0005552526883746087,
      "loss": 1.2745,
      "step": 1365
    },
    {
      "epoch": 0.5201708590412909,
      "grad_norm": 0.4895534717498465,
      "learning_rate": 0.0005519564873162257,
      "loss": 1.2461,
      "step": 1370
    },
    {
      "epoch": 0.5220692928334124,
      "grad_norm": 0.4461111757267344,
      "learning_rate": 0.0005486580019858075,
      "loss": 1.2704,
      "step": 1375
    },
    {
      "epoch": 0.5239677266255339,
      "grad_norm": 0.32031503363419567,
      "learning_rate": 0.0005453573774015837,
      "loss": 1.2638,
      "step": 1380
    },
    {
      "epoch": 0.5258661604176554,
      "grad_norm": 0.3643504769224511,
      "learning_rate": 0.0005420547586758364,
      "loss": 1.2595,
      "step": 1385
    },
    {
      "epoch": 0.527764594209777,
      "grad_norm": 0.32810368902662324,
      "learning_rate": 0.0005387502910085201,
      "loss": 1.2326,
      "step": 1390
    },
    {
      "epoch": 0.5296630280018985,
      "grad_norm": 0.3336242397751466,
      "learning_rate": 0.0005354441196808778,
      "loss": 1.2553,
      "step": 1395
    },
    {
      "epoch": 0.53156146179402,
      "grad_norm": 0.4348846619107913,
      "learning_rate": 0.000532136390049055,
      "loss": 1.2457,
      "step": 1400
    },
    {
      "epoch": 0.5334598955861415,
      "grad_norm": 0.3940477704170248,
      "learning_rate": 0.0005288272475377078,
      "loss": 1.2286,
      "step": 1405
    },
    {
      "epoch": 0.5353583293782629,
      "grad_norm": 0.3034212811156197,
      "learning_rate": 0.0005255168376336094,
      "loss": 1.209,
      "step": 1410
    },
    {
      "epoch": 0.5372567631703844,
      "grad_norm": 0.4307385167019054,
      "learning_rate": 0.0005222053058792543,
      "loss": 1.2476,
      "step": 1415
    },
    {
      "epoch": 0.5391551969625059,
      "grad_norm": 0.34543357053441875,
      "learning_rate": 0.0005188927978664594,
      "loss": 1.2149,
      "step": 1420
    },
    {
      "epoch": 0.5410536307546274,
      "grad_norm": 0.3611941956078568,
      "learning_rate": 0.0005155794592299626,
      "loss": 1.2407,
      "step": 1425
    },
    {
      "epoch": 0.5429520645467489,
      "grad_norm": 0.3375414745847685,
      "learning_rate": 0.0005122654356410205,
      "loss": 1.2412,
      "step": 1430
    },
    {
      "epoch": 0.5448504983388704,
      "grad_norm": 0.286479882843699,
      "learning_rate": 0.0005089508728010033,
      "loss": 1.2167,
      "step": 1435
    },
    {
      "epoch": 0.546748932130992,
      "grad_norm": 0.3385247369838371,
      "learning_rate": 0.0005056359164349902,
      "loss": 1.231,
      "step": 1440
    },
    {
      "epoch": 0.5486473659231135,
      "grad_norm": 0.30665827143092833,
      "learning_rate": 0.000502320712285361,
      "loss": 1.1893,
      "step": 1445
    },
    {
      "epoch": 0.550545799715235,
      "grad_norm": 0.3668246619215131,
      "learning_rate": 0.0004990054061053896,
      "loss": 1.1876,
      "step": 1450
    },
    {
      "epoch": 0.5524442335073564,
      "grad_norm": 0.3886963786863651,
      "learning_rate": 0.0004956901436528358,
      "loss": 1.1897,
      "step": 1455
    },
    {
      "epoch": 0.5543426672994779,
      "grad_norm": 0.33560559480266355,
      "learning_rate": 0.0004923750706835371,
      "loss": 1.2048,
      "step": 1460
    },
    {
      "epoch": 0.5562411010915994,
      "grad_norm": 0.35331444970790565,
      "learning_rate": 0.0004890603329449997,
      "loss": 1.1893,
      "step": 1465
    },
    {
      "epoch": 0.5581395348837209,
      "grad_norm": 0.31797300187844246,
      "learning_rate": 0.000485746076169992,
      "loss": 1.1653,
      "step": 1470
    },
    {
      "epoch": 0.5600379686758424,
      "grad_norm": 0.3697991594252187,
      "learning_rate": 0.00048243244607013654,
      "loss": 1.2023,
      "step": 1475
    },
    {
      "epoch": 0.561936402467964,
      "grad_norm": 0.5326439145387593,
      "learning_rate": 0.0004791195883295036,
      "loss": 1.1813,
      "step": 1480
    },
    {
      "epoch": 0.5638348362600855,
      "grad_norm": 0.371212386610569,
      "learning_rate": 0.0004758076485982076,
      "loss": 1.1702,
      "step": 1485
    },
    {
      "epoch": 0.565733270052207,
      "grad_norm": 0.3968881977650288,
      "learning_rate": 0.00047249677248600185,
      "loss": 1.1673,
      "step": 1490
    },
    {
      "epoch": 0.5676317038443285,
      "grad_norm": 2.9338387571685405,
      "learning_rate": 0.0004691871055558776,
      "loss": 1.1908,
      "step": 1495
    },
    {
      "epoch": 0.5695301376364499,
      "grad_norm": 0.49847319164775866,
      "learning_rate": 0.00046587879331766457,
      "loss": 1.1864,
      "step": 1500
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.4078724706306019,
      "learning_rate": 0.000462571981221633,
      "loss": 1.1663,
      "step": 1505
    },
    {
      "epoch": 0.5733270052206929,
      "grad_norm": 0.3228249575712461,
      "learning_rate": 0.0004592668146520994,
      "loss": 1.1702,
      "step": 1510
    },
    {
      "epoch": 0.5752254390128144,
      "grad_norm": 0.37770036266775847,
      "learning_rate": 0.00045596343892103443,
      "loss": 1.176,
      "step": 1515
    },
    {
      "epoch": 0.5771238728049359,
      "grad_norm": 0.3148160726043947,
      "learning_rate": 0.00045266199926167485,
      "loss": 1.1747,
      "step": 1520
    },
    {
      "epoch": 0.5790223065970574,
      "grad_norm": 0.3626958842571658,
      "learning_rate": 0.00044936264082213724,
      "loss": 1.143,
      "step": 1525
    },
    {
      "epoch": 0.580920740389179,
      "grad_norm": 0.3613417080127485,
      "learning_rate": 0.00044606550865903725,
      "loss": 1.1523,
      "step": 1530
    },
    {
      "epoch": 0.5828191741813005,
      "grad_norm": 0.36620946050669057,
      "learning_rate": 0.0004427707477311123,
      "loss": 1.1732,
      "step": 1535
    },
    {
      "epoch": 0.584717607973422,
      "grad_norm": 0.3418418756481974,
      "learning_rate": 0.000439478502892848,
      "loss": 1.1616,
      "step": 1540
    },
    {
      "epoch": 0.5866160417655434,
      "grad_norm": 0.33600686205594626,
      "learning_rate": 0.0004361889188881102,
      "loss": 1.1158,
      "step": 1545
    },
    {
      "epoch": 0.5885144755576649,
      "grad_norm": 0.3764722521246889,
      "learning_rate": 0.0004329021403437802,
      "loss": 1.154,
      "step": 1550
    },
    {
      "epoch": 0.5904129093497864,
      "grad_norm": 0.34240802002457055,
      "learning_rate": 0.000429618311763398,
      "loss": 1.139,
      "step": 1555
    },
    {
      "epoch": 0.5923113431419079,
      "grad_norm": 0.2902338020123924,
      "learning_rate": 0.00042633757752080727,
      "loss": 1.134,
      "step": 1560
    },
    {
      "epoch": 0.5942097769340294,
      "grad_norm": 0.3752689277444726,
      "learning_rate": 0.00042306008185380927,
      "loss": 1.1343,
      "step": 1565
    },
    {
      "epoch": 0.5961082107261509,
      "grad_norm": 0.31329902926972597,
      "learning_rate": 0.0004197859688578207,
      "loss": 1.1016,
      "step": 1570
    },
    {
      "epoch": 0.5980066445182725,
      "grad_norm": 0.3402332223232388,
      "learning_rate": 0.00041651538247953904,
      "loss": 1.1052,
      "step": 1575
    },
    {
      "epoch": 0.599905078310394,
      "grad_norm": 0.3781151786788464,
      "learning_rate": 0.0004132484665106135,
      "loss": 1.1191,
      "step": 1580
    },
    {
      "epoch": 0.6018035121025154,
      "grad_norm": 0.34137800960299325,
      "learning_rate": 0.0004099853645813235,
      "loss": 1.0947,
      "step": 1585
    },
    {
      "epoch": 0.6037019458946369,
      "grad_norm": 0.3136113067101357,
      "learning_rate": 0.00040672622015426363,
      "loss": 1.0856,
      "step": 1590
    },
    {
      "epoch": 0.6056003796867584,
      "grad_norm": 0.3358747164703956,
      "learning_rate": 0.00040347117651803703,
      "loss": 1.0948,
      "step": 1595
    },
    {
      "epoch": 0.6074988134788799,
      "grad_norm": 0.3020729530283223,
      "learning_rate": 0.00040022037678095454,
      "loss": 1.0802,
      "step": 1600
    },
    {
      "epoch": 0.6093972472710014,
      "grad_norm": 0.3904330186630717,
      "learning_rate": 0.00039697396386474394,
      "loss": 1.127,
      "step": 1605
    },
    {
      "epoch": 0.6112956810631229,
      "grad_norm": 0.4027533805204724,
      "learning_rate": 0.0003937320804982659,
      "loss": 1.1042,
      "step": 1610
    },
    {
      "epoch": 0.6131941148552444,
      "grad_norm": 0.3743040482483265,
      "learning_rate": 0.00039049486921123876,
      "loss": 1.0907,
      "step": 1615
    },
    {
      "epoch": 0.615092548647366,
      "grad_norm": 0.38060393763386724,
      "learning_rate": 0.000387262472327973,
      "loss": 1.1051,
      "step": 1620
    },
    {
      "epoch": 0.6169909824394875,
      "grad_norm": 0.34005085980267563,
      "learning_rate": 0.00038403503196111265,
      "loss": 1.0647,
      "step": 1625
    },
    {
      "epoch": 0.6188894162316089,
      "grad_norm": 0.30724882728663455,
      "learning_rate": 0.0003808126900053887,
      "loss": 1.0771,
      "step": 1630
    },
    {
      "epoch": 0.6207878500237304,
      "grad_norm": 0.306237705852109,
      "learning_rate": 0.0003775955881313797,
      "loss": 1.0831,
      "step": 1635
    },
    {
      "epoch": 0.6226862838158519,
      "grad_norm": 0.4486821484171895,
      "learning_rate": 0.0003743838677792833,
      "loss": 1.0767,
      "step": 1640
    },
    {
      "epoch": 0.6245847176079734,
      "grad_norm": 0.35675874358918835,
      "learning_rate": 0.0003711776701526982,
      "loss": 1.0841,
      "step": 1645
    },
    {
      "epoch": 0.6264831514000949,
      "grad_norm": 0.31338528181698055,
      "learning_rate": 0.00036797713621241615,
      "loss": 1.0851,
      "step": 1650
    },
    {
      "epoch": 0.6283815851922164,
      "grad_norm": 0.3687470434111214,
      "learning_rate": 0.000364782406670224,
      "loss": 1.0707,
      "step": 1655
    },
    {
      "epoch": 0.6302800189843379,
      "grad_norm": 0.31784199572514304,
      "learning_rate": 0.0003615936219827176,
      "loss": 1.0606,
      "step": 1660
    },
    {
      "epoch": 0.6321784527764595,
      "grad_norm": 0.42249881083466695,
      "learning_rate": 0.00035841092234512723,
      "loss": 1.0725,
      "step": 1665
    },
    {
      "epoch": 0.634076886568581,
      "grad_norm": 0.3459092803870884,
      "learning_rate": 0.0003552344476851531,
      "loss": 1.0574,
      "step": 1670
    },
    {
      "epoch": 0.6359753203607024,
      "grad_norm": 0.31341808911866365,
      "learning_rate": 0.00035206433765681334,
      "loss": 1.0399,
      "step": 1675
    },
    {
      "epoch": 0.6378737541528239,
      "grad_norm": 0.3619705130805933,
      "learning_rate": 0.00034890073163430503,
      "loss": 1.0295,
      "step": 1680
    },
    {
      "epoch": 0.6397721879449454,
      "grad_norm": 0.302930290181787,
      "learning_rate": 0.00034574376870587535,
      "loss": 1.0302,
      "step": 1685
    },
    {
      "epoch": 0.6416706217370669,
      "grad_norm": 0.32319117475207404,
      "learning_rate": 0.00034259358766770766,
      "loss": 1.0457,
      "step": 1690
    },
    {
      "epoch": 0.6435690555291884,
      "grad_norm": 0.35131830575911627,
      "learning_rate": 0.0003394503270178185,
      "loss": 1.0527,
      "step": 1695
    },
    {
      "epoch": 0.6454674893213099,
      "grad_norm": 0.35565194703093944,
      "learning_rate": 0.0003363141249499696,
      "loss": 1.0271,
      "step": 1700
    },
    {
      "epoch": 0.6473659231134314,
      "grad_norm": 0.3631695241713009,
      "learning_rate": 0.00033318511934759046,
      "loss": 1.044,
      "step": 1705
    },
    {
      "epoch": 0.649264356905553,
      "grad_norm": 0.3236231297783939,
      "learning_rate": 0.0003300634477777179,
      "loss": 1.0102,
      "step": 1710
    },
    {
      "epoch": 0.6511627906976745,
      "grad_norm": 0.3213681061725248,
      "learning_rate": 0.00032694924748494713,
      "loss": 1.0083,
      "step": 1715
    },
    {
      "epoch": 0.6530612244897959,
      "grad_norm": 0.6068864104166338,
      "learning_rate": 0.00032384265538539783,
      "loss": 0.9907,
      "step": 1720
    },
    {
      "epoch": 0.6549596582819174,
      "grad_norm": 0.41210521486813323,
      "learning_rate": 0.0003207438080606949,
      "loss": 1.0233,
      "step": 1725
    },
    {
      "epoch": 0.6568580920740389,
      "grad_norm": 0.46261139879081853,
      "learning_rate": 0.00031765284175196324,
      "loss": 1.0167,
      "step": 1730
    },
    {
      "epoch": 0.6587565258661604,
      "grad_norm": 0.3436327494287044,
      "learning_rate": 0.0003145698923538384,
      "loss": 1.0282,
      "step": 1735
    },
    {
      "epoch": 0.6606549596582819,
      "grad_norm": 0.35619308018504636,
      "learning_rate": 0.00031149509540849156,
      "loss": 0.99,
      "step": 1740
    },
    {
      "epoch": 0.6625533934504034,
      "grad_norm": 0.34309970855605526,
      "learning_rate": 0.0003084285860996704,
      "loss": 1.015,
      "step": 1745
    },
    {
      "epoch": 0.6644518272425249,
      "grad_norm": 0.393508338674986,
      "learning_rate": 0.0003053704992467558,
      "loss": 1.0116,
      "step": 1750
    },
    {
      "epoch": 0.6663502610346465,
      "grad_norm": 0.3397492909000668,
      "learning_rate": 0.0003023209692988349,
      "loss": 0.9983,
      "step": 1755
    },
    {
      "epoch": 0.668248694826768,
      "grad_norm": 0.30189558040021247,
      "learning_rate": 0.0002992801303287892,
      "loss": 1.0073,
      "step": 1760
    },
    {
      "epoch": 0.6701471286188894,
      "grad_norm": 0.33390373051452854,
      "learning_rate": 0.00029624811602740105,
      "loss": 0.9931,
      "step": 1765
    },
    {
      "epoch": 0.6720455624110109,
      "grad_norm": 0.3565911003751968,
      "learning_rate": 0.0002932250596974747,
      "loss": 0.9794,
      "step": 1770
    },
    {
      "epoch": 0.6739439962031324,
      "grad_norm": 0.30184049203060254,
      "learning_rate": 0.00029021109424797706,
      "loss": 0.9853,
      "step": 1775
    },
    {
      "epoch": 0.6758424299952539,
      "grad_norm": 0.2996692988493312,
      "learning_rate": 0.00028720635218819313,
      "loss": 0.9716,
      "step": 1780
    },
    {
      "epoch": 0.6777408637873754,
      "grad_norm": 0.3158967543323723,
      "learning_rate": 0.00028421096562190087,
      "loss": 0.971,
      "step": 1785
    },
    {
      "epoch": 0.6796392975794969,
      "grad_norm": 0.3259541143986902,
      "learning_rate": 0.00028122506624156287,
      "loss": 0.9667,
      "step": 1790
    },
    {
      "epoch": 0.6815377313716184,
      "grad_norm": 0.3050723057723714,
      "learning_rate": 0.00027824878532253675,
      "loss": 0.9698,
      "step": 1795
    },
    {
      "epoch": 0.68343616516374,
      "grad_norm": 0.3225081325517291,
      "learning_rate": 0.0002752822537173033,
      "loss": 0.9627,
      "step": 1800
    },
    {
      "epoch": 0.6853345989558615,
      "grad_norm": 0.3337812239040389,
      "learning_rate": 0.00027232560184971434,
      "loss": 0.9673,
      "step": 1805
    },
    {
      "epoch": 0.6872330327479829,
      "grad_norm": 0.4504964673114019,
      "learning_rate": 0.00026937895970925794,
      "loss": 0.9504,
      "step": 1810
    },
    {
      "epoch": 0.6891314665401044,
      "grad_norm": 0.6271532351806949,
      "learning_rate": 0.00026644245684534317,
      "loss": 0.9655,
      "step": 1815
    },
    {
      "epoch": 0.6910299003322259,
      "grad_norm": 0.42681861959807327,
      "learning_rate": 0.00026351622236160487,
      "loss": 0.9418,
      "step": 1820
    },
    {
      "epoch": 0.6929283341243474,
      "grad_norm": 0.4353835236159284,
      "learning_rate": 0.00026060038491022787,
      "loss": 0.9777,
      "step": 1825
    },
    {
      "epoch": 0.6948267679164689,
      "grad_norm": 0.3479809339329037,
      "learning_rate": 0.00025769507268628993,
      "loss": 0.9721,
      "step": 1830
    },
    {
      "epoch": 0.6967252017085904,
      "grad_norm": 0.31070434225155574,
      "learning_rate": 0.00025480041342212695,
      "loss": 0.9553,
      "step": 1835
    },
    {
      "epoch": 0.6986236355007119,
      "grad_norm": 0.3849172751179302,
      "learning_rate": 0.00025191653438171545,
      "loss": 0.9461,
      "step": 1840
    },
    {
      "epoch": 0.7005220692928334,
      "grad_norm": 0.4264474066939237,
      "learning_rate": 0.00024904356235507945,
      "loss": 0.9467,
      "step": 1845
    },
    {
      "epoch": 0.702420503084955,
      "grad_norm": 0.3548766629306175,
      "learning_rate": 0.0002461816236527141,
      "loss": 0.947,
      "step": 1850
    },
    {
      "epoch": 0.7043189368770764,
      "grad_norm": 0.3514101652953766,
      "learning_rate": 0.0002433308441000338,
      "loss": 0.9268,
      "step": 1855
    },
    {
      "epoch": 0.7062173706691979,
      "grad_norm": 0.380453955195146,
      "learning_rate": 0.00024049134903183955,
      "loss": 0.9293,
      "step": 1860
    },
    {
      "epoch": 0.7081158044613194,
      "grad_norm": 0.35298900401975053,
      "learning_rate": 0.00023766326328680958,
      "loss": 0.9348,
      "step": 1865
    },
    {
      "epoch": 0.7100142382534409,
      "grad_norm": 0.42441310564082074,
      "learning_rate": 0.00023484671120200935,
      "loss": 0.9276,
      "step": 1870
    },
    {
      "epoch": 0.7119126720455624,
      "grad_norm": 0.319038899787948,
      "learning_rate": 0.00023204181660742602,
      "loss": 0.9431,
      "step": 1875
    },
    {
      "epoch": 0.7138111058376839,
      "grad_norm": 0.4049851574728775,
      "learning_rate": 0.00022924870282052445,
      "loss": 0.906,
      "step": 1880
    },
    {
      "epoch": 0.7157095396298054,
      "grad_norm": 0.3513877016599214,
      "learning_rate": 0.00022646749264082478,
      "loss": 0.9095,
      "step": 1885
    },
    {
      "epoch": 0.717607973421927,
      "grad_norm": 0.4081310968482307,
      "learning_rate": 0.00022369830834450367,
      "loss": 0.9003,
      "step": 1890
    },
    {
      "epoch": 0.7195064072140485,
      "grad_norm": 0.35805106514821416,
      "learning_rate": 0.00022094127167901934,
      "loss": 0.9099,
      "step": 1895
    },
    {
      "epoch": 0.7214048410061699,
      "grad_norm": 0.3631465368774142,
      "learning_rate": 0.0002181965038577577,
      "loss": 0.9092,
      "step": 1900
    },
    {
      "epoch": 0.7233032747982914,
      "grad_norm": 0.3650905218048075,
      "learning_rate": 0.0002154641255547038,
      "loss": 0.9252,
      "step": 1905
    },
    {
      "epoch": 0.7252017085904129,
      "grad_norm": 0.31898731584804263,
      "learning_rate": 0.00021274425689913617,
      "loss": 0.9071,
      "step": 1910
    },
    {
      "epoch": 0.7271001423825344,
      "grad_norm": 0.31603070803205635,
      "learning_rate": 0.00021003701747034616,
      "loss": 0.9165,
      "step": 1915
    },
    {
      "epoch": 0.7289985761746559,
      "grad_norm": 0.30904969427001383,
      "learning_rate": 0.00020734252629237893,
      "loss": 0.8971,
      "step": 1920
    },
    {
      "epoch": 0.7308970099667774,
      "grad_norm": 0.32433517710502985,
      "learning_rate": 0.00020466090182880248,
      "loss": 0.8908,
      "step": 1925
    },
    {
      "epoch": 0.7327954437588989,
      "grad_norm": 0.3168746074632352,
      "learning_rate": 0.00020199226197749792,
      "loss": 0.9011,
      "step": 1930
    },
    {
      "epoch": 0.7346938775510204,
      "grad_norm": 0.31832881987489503,
      "learning_rate": 0.00019933672406547665,
      "loss": 0.9138,
      "step": 1935
    },
    {
      "epoch": 0.736592311343142,
      "grad_norm": 0.301051575919513,
      "learning_rate": 0.00019669440484372213,
      "loss": 0.8897,
      "step": 1940
    },
    {
      "epoch": 0.7384907451352634,
      "grad_norm": 0.33220133814920705,
      "learning_rate": 0.00019406542048205666,
      "loss": 0.8868,
      "step": 1945
    },
    {
      "epoch": 0.7403891789273849,
      "grad_norm": 0.31331199745305405,
      "learning_rate": 0.0001914498865640344,
      "loss": 0.8921,
      "step": 1950
    },
    {
      "epoch": 0.7422876127195064,
      "grad_norm": 0.3150890748673222,
      "learning_rate": 0.00018884791808185947,
      "loss": 0.8925,
      "step": 1955
    },
    {
      "epoch": 0.7441860465116279,
      "grad_norm": 0.31284305447491684,
      "learning_rate": 0.0001862596294313299,
      "loss": 0.8779,
      "step": 1960
    },
    {
      "epoch": 0.7460844803037494,
      "grad_norm": 0.33272829874860427,
      "learning_rate": 0.00018368513440680884,
      "loss": 0.8799,
      "step": 1965
    },
    {
      "epoch": 0.7479829140958709,
      "grad_norm": 0.32686823493529116,
      "learning_rate": 0.0001811245461962212,
      "loss": 0.8837,
      "step": 1970
    },
    {
      "epoch": 0.7498813478879924,
      "grad_norm": 0.32129617206658423,
      "learning_rate": 0.0001785779773760775,
      "loss": 0.891,
      "step": 1975
    },
    {
      "epoch": 0.751779781680114,
      "grad_norm": 0.29507710361264766,
      "learning_rate": 0.0001760455399065246,
      "loss": 0.87,
      "step": 1980
    },
    {
      "epoch": 0.7536782154722355,
      "grad_norm": 0.3392524599510225,
      "learning_rate": 0.00017352734512642276,
      "loss": 0.8614,
      "step": 1985
    },
    {
      "epoch": 0.7555766492643569,
      "grad_norm": 0.34922494756037853,
      "learning_rate": 0.00017102350374845155,
      "loss": 0.8739,
      "step": 1990
    },
    {
      "epoch": 0.7574750830564784,
      "grad_norm": 0.38856477147363533,
      "learning_rate": 0.00016853412585424128,
      "loss": 0.8608,
      "step": 1995
    },
    {
      "epoch": 0.7593735168485999,
      "grad_norm": 0.3667333097382861,
      "learning_rate": 0.00016605932088953397,
      "loss": 0.8524,
      "step": 2000
    },
    {
      "epoch": 0.7612719506407214,
      "grad_norm": 0.4112130915608781,
      "learning_rate": 0.00016359919765937149,
      "loss": 0.851,
      "step": 2005
    },
    {
      "epoch": 0.7631703844328429,
      "grad_norm": 0.35792666839319404,
      "learning_rate": 0.00016115386432331147,
      "loss": 0.8444,
      "step": 2010
    },
    {
      "epoch": 0.7650688182249644,
      "grad_norm": 0.35486084734332507,
      "learning_rate": 0.00015872342839067305,
      "loss": 0.8561,
      "step": 2015
    },
    {
      "epoch": 0.7669672520170859,
      "grad_norm": 0.34325463219420715,
      "learning_rate": 0.0001563079967158088,
      "loss": 0.8549,
      "step": 2020
    },
    {
      "epoch": 0.7688656858092074,
      "grad_norm": 0.33043895406245294,
      "learning_rate": 0.0001539076754934084,
      "loss": 0.8576,
      "step": 2025
    },
    {
      "epoch": 0.770764119601329,
      "grad_norm": 0.30878788182142447,
      "learning_rate": 0.00015152257025382844,
      "loss": 0.8559,
      "step": 2030
    },
    {
      "epoch": 0.7726625533934504,
      "grad_norm": 0.35444741515541456,
      "learning_rate": 0.00014915278585845348,
      "loss": 0.8415,
      "step": 2035
    },
    {
      "epoch": 0.7745609871855719,
      "grad_norm": 0.30847085513086014,
      "learning_rate": 0.00014679842649508568,
      "loss": 0.8259,
      "step": 2040
    },
    {
      "epoch": 0.7764594209776934,
      "grad_norm": 0.31252404771224196,
      "learning_rate": 0.00014445959567336441,
      "loss": 0.859,
      "step": 2045
    },
    {
      "epoch": 0.7783578547698149,
      "grad_norm": 0.3240169921265319,
      "learning_rate": 0.0001421363962202149,
      "loss": 0.8301,
      "step": 2050
    },
    {
      "epoch": 0.7802562885619364,
      "grad_norm": 0.32322492788162155,
      "learning_rate": 0.00013982893027532757,
      "loss": 0.822,
      "step": 2055
    },
    {
      "epoch": 0.7821547223540579,
      "grad_norm": 0.3516161817741868,
      "learning_rate": 0.00013753729928666825,
      "loss": 0.8374,
      "step": 2060
    },
    {
      "epoch": 0.7840531561461794,
      "grad_norm": 0.31124890284462753,
      "learning_rate": 0.00013526160400601682,
      "loss": 0.8293,
      "step": 2065
    },
    {
      "epoch": 0.7859515899383009,
      "grad_norm": 0.307376831052294,
      "learning_rate": 0.00013300194448453818,
      "loss": 0.8314,
      "step": 2070
    },
    {
      "epoch": 0.7878500237304225,
      "grad_norm": 0.33386681520064726,
      "learning_rate": 0.00013075842006838407,
      "loss": 0.8485,
      "step": 2075
    },
    {
      "epoch": 0.7897484575225439,
      "grad_norm": 0.3223752393505699,
      "learning_rate": 0.0001285311293943241,
      "loss": 0.8424,
      "step": 2080
    },
    {
      "epoch": 0.7916468913146654,
      "grad_norm": 0.3575122610804609,
      "learning_rate": 0.00012632017038541026,
      "loss": 0.8326,
      "step": 2085
    },
    {
      "epoch": 0.7935453251067869,
      "grad_norm": 0.32346828366821057,
      "learning_rate": 0.0001241256402466709,
      "loss": 0.8422,
      "step": 2090
    },
    {
      "epoch": 0.7954437588989084,
      "grad_norm": 0.3341501772850559,
      "learning_rate": 0.00012194763546083803,
      "loss": 0.8155,
      "step": 2095
    },
    {
      "epoch": 0.7973421926910299,
      "grad_norm": 0.33689872208599597,
      "learning_rate": 0.00011978625178410434,
      "loss": 0.8179,
      "step": 2100
    },
    {
      "epoch": 0.7992406264831514,
      "grad_norm": 0.3640068937736561,
      "learning_rate": 0.00011764158424191435,
      "loss": 0.8107,
      "step": 2105
    },
    {
      "epoch": 0.8011390602752729,
      "grad_norm": 0.31299920707259116,
      "learning_rate": 0.00011551372712478575,
      "loss": 0.8013,
      "step": 2110
    },
    {
      "epoch": 0.8030374940673944,
      "grad_norm": 0.3184880942321623,
      "learning_rate": 0.0001134027739841642,
      "loss": 0.8012,
      "step": 2115
    },
    {
      "epoch": 0.804935927859516,
      "grad_norm": 0.3162681368592125,
      "learning_rate": 0.00011130881762831069,
      "loss": 0.8156,
      "step": 2120
    },
    {
      "epoch": 0.8068343616516374,
      "grad_norm": 0.3211345805440541,
      "learning_rate": 0.00010923195011822058,
      "loss": 0.8051,
      "step": 2125
    },
    {
      "epoch": 0.8087327954437589,
      "grad_norm": 0.3249567779384397,
      "learning_rate": 0.00010717226276357667,
      "loss": 0.8024,
      "step": 2130
    },
    {
      "epoch": 0.8106312292358804,
      "grad_norm": 0.31173448153562644,
      "learning_rate": 0.00010512984611873466,
      "loss": 0.8028,
      "step": 2135
    },
    {
      "epoch": 0.8125296630280019,
      "grad_norm": 0.34264672110601546,
      "learning_rate": 0.00010310478997874162,
      "loss": 0.7893,
      "step": 2140
    },
    {
      "epoch": 0.8144280968201234,
      "grad_norm": 0.36225713373489543,
      "learning_rate": 0.0001010971833753882,
      "loss": 0.7941,
      "step": 2145
    },
    {
      "epoch": 0.8163265306122449,
      "grad_norm": 0.39829779195110515,
      "learning_rate": 9.910711457329479e-05,
      "loss": 0.784,
      "step": 2150
    },
    {
      "epoch": 0.8182249644043664,
      "grad_norm": 0.4154842554468083,
      "learning_rate": 9.713467106603024e-05,
      "loss": 0.7973,
      "step": 2155
    },
    {
      "epoch": 0.8201233981964879,
      "grad_norm": 0.35334120375262246,
      "learning_rate": 9.517993957226612e-05,
      "loss": 0.7887,
      "step": 2160
    },
    {
      "epoch": 0.8220218319886095,
      "grad_norm": 0.342997739965322,
      "learning_rate": 9.3243006031963e-05,
      "loss": 0.7884,
      "step": 2165
    },
    {
      "epoch": 0.8239202657807309,
      "grad_norm": 0.345714194018433,
      "learning_rate": 9.132395560259337e-05,
      "loss": 0.7894,
      "step": 2170
    },
    {
      "epoch": 0.8258186995728524,
      "grad_norm": 0.34821270209325056,
      "learning_rate": 8.942287265539639e-05,
      "loss": 0.796,
      "step": 2175
    },
    {
      "epoch": 0.8277171333649739,
      "grad_norm": 0.32502745681436357,
      "learning_rate": 8.753984077166937e-05,
      "loss": 0.8015,
      "step": 2180
    },
    {
      "epoch": 0.8296155671570954,
      "grad_norm": 0.388775004922499,
      "learning_rate": 8.567494273909277e-05,
      "loss": 0.8025,
      "step": 2185
    },
    {
      "epoch": 0.8315140009492169,
      "grad_norm": 0.3127036551166177,
      "learning_rate": 8.382826054809079e-05,
      "loss": 0.7895,
      "step": 2190
    },
    {
      "epoch": 0.8334124347413384,
      "grad_norm": 0.29839600378854597,
      "learning_rate": 8.1999875388226e-05,
      "loss": 0.7818,
      "step": 2195
    },
    {
      "epoch": 0.8353108685334599,
      "grad_norm": 0.31004929243565,
      "learning_rate": 8.018986764463032e-05,
      "loss": 0.7881,
      "step": 2200
    },
    {
      "epoch": 0.8372093023255814,
      "grad_norm": 0.32248432401397453,
      "learning_rate": 7.839831689447102e-05,
      "loss": 0.7961,
      "step": 2205
    },
    {
      "epoch": 0.8391077361177028,
      "grad_norm": 0.32300916569046356,
      "learning_rate": 7.662530190345157e-05,
      "loss": 0.7859,
      "step": 2210
    },
    {
      "epoch": 0.8410061699098244,
      "grad_norm": 0.3437848320718501,
      "learning_rate": 7.487090062234898e-05,
      "loss": 0.7706,
      "step": 2215
    },
    {
      "epoch": 0.8429046037019459,
      "grad_norm": 0.3104343754734283,
      "learning_rate": 7.313519018358695e-05,
      "loss": 0.7715,
      "step": 2220
    },
    {
      "epoch": 0.8448030374940674,
      "grad_norm": 0.3171733947505178,
      "learning_rate": 7.141824689784421e-05,
      "loss": 0.7487,
      "step": 2225
    },
    {
      "epoch": 0.8467014712861889,
      "grad_norm": 0.3114906950878784,
      "learning_rate": 6.972014625069984e-05,
      "loss": 0.7557,
      "step": 2230
    },
    {
      "epoch": 0.8485999050783104,
      "grad_norm": 0.3427221492524242,
      "learning_rate": 6.804096289931443e-05,
      "loss": 0.7505,
      "step": 2235
    },
    {
      "epoch": 0.8504983388704319,
      "grad_norm": 0.3257908676508504,
      "learning_rate": 6.638077066914811e-05,
      "loss": 0.7691,
      "step": 2240
    },
    {
      "epoch": 0.8523967726625534,
      "grad_norm": 0.3245771425045967,
      "learning_rate": 6.473964255071418e-05,
      "loss": 0.7712,
      "step": 2245
    },
    {
      "epoch": 0.8542952064546749,
      "grad_norm": 0.30642317659601426,
      "learning_rate": 6.311765069637037e-05,
      "loss": 0.7508,
      "step": 2250
    },
    {
      "epoch": 0.8561936402467963,
      "grad_norm": 0.35717316205581434,
      "learning_rate": 6.151486641714705e-05,
      "loss": 0.7585,
      "step": 2255
    },
    {
      "epoch": 0.8580920740389179,
      "grad_norm": 0.29834051216293683,
      "learning_rate": 5.993136017961143e-05,
      "loss": 0.7564,
      "step": 2260
    },
    {
      "epoch": 0.8599905078310394,
      "grad_norm": 0.31967989226553994,
      "learning_rate": 5.83672016027697e-05,
      "loss": 0.7402,
      "step": 2265
    },
    {
      "epoch": 0.8618889416231609,
      "grad_norm": 0.3316739821901055,
      "learning_rate": 5.6822459455006246e-05,
      "loss": 0.7617,
      "step": 2270
    },
    {
      "epoch": 0.8637873754152824,
      "grad_norm": 0.33111810076533643,
      "learning_rate": 5.529720165106056e-05,
      "loss": 0.7702,
      "step": 2275
    },
    {
      "epoch": 0.8656858092074039,
      "grad_norm": 0.31953471953308643,
      "learning_rate": 5.3791495249040644e-05,
      "loss": 0.7585,
      "step": 2280
    },
    {
      "epoch": 0.8675842429995254,
      "grad_norm": 0.34435784933473257,
      "learning_rate": 5.2305406447475504e-05,
      "loss": 0.7539,
      "step": 2285
    },
    {
      "epoch": 0.8694826767916469,
      "grad_norm": 0.33531663584742627,
      "learning_rate": 5.083900058240437e-05,
      "loss": 0.7574,
      "step": 2290
    },
    {
      "epoch": 0.8713811105837684,
      "grad_norm": 0.32627115970227594,
      "learning_rate": 4.939234212450405e-05,
      "loss": 0.7502,
      "step": 2295
    },
    {
      "epoch": 0.8732795443758898,
      "grad_norm": 0.34710941754901015,
      "learning_rate": 4.796549467625494e-05,
      "loss": 0.7517,
      "step": 2300
    },
    {
      "epoch": 0.8751779781680114,
      "grad_norm": 0.3259737091035039,
      "learning_rate": 4.6558520969144205e-05,
      "loss": 0.7531,
      "step": 2305
    },
    {
      "epoch": 0.8770764119601329,
      "grad_norm": 0.3543527295348622,
      "learning_rate": 4.517148286090822e-05,
      "loss": 0.7395,
      "step": 2310
    },
    {
      "epoch": 0.8789748457522544,
      "grad_norm": 0.3411443809878121,
      "learning_rate": 4.3804441332812915e-05,
      "loss": 0.7493,
      "step": 2315
    },
    {
      "epoch": 0.8808732795443759,
      "grad_norm": 0.323002963765091,
      "learning_rate": 4.245745648697241e-05,
      "loss": 0.7562,
      "step": 2320
    },
    {
      "epoch": 0.8827717133364974,
      "grad_norm": 0.3369758289265976,
      "learning_rate": 4.1130587543706796e-05,
      "loss": 0.7466,
      "step": 2325
    },
    {
      "epoch": 0.8846701471286189,
      "grad_norm": 0.3152326006804494,
      "learning_rate": 3.982389283893878e-05,
      "loss": 0.7458,
      "step": 2330
    },
    {
      "epoch": 0.8865685809207404,
      "grad_norm": 0.3504098491203572,
      "learning_rate": 3.853742982162839e-05,
      "loss": 0.7503,
      "step": 2335
    },
    {
      "epoch": 0.8884670147128619,
      "grad_norm": 0.33687850279262355,
      "learning_rate": 3.72712550512479e-05,
      "loss": 0.7651,
      "step": 2340
    },
    {
      "epoch": 0.8903654485049833,
      "grad_norm": 0.3386005467292844,
      "learning_rate": 3.602542419529453e-05,
      "loss": 0.766,
      "step": 2345
    },
    {
      "epoch": 0.8922638822971048,
      "grad_norm": 0.3110137804067388,
      "learning_rate": 3.479999202684353e-05,
      "loss": 0.7397,
      "step": 2350
    },
    {
      "epoch": 0.8941623160892264,
      "grad_norm": 0.32583342601441784,
      "learning_rate": 3.359501242213981e-05,
      "loss": 0.7346,
      "step": 2355
    },
    {
      "epoch": 0.8960607498813479,
      "grad_norm": 0.3187549346009566,
      "learning_rate": 3.24105383582291e-05,
      "loss": 0.7492,
      "step": 2360
    },
    {
      "epoch": 0.8979591836734694,
      "grad_norm": 0.3218971294090315,
      "learning_rate": 3.1246621910629323e-05,
      "loss": 0.7504,
      "step": 2365
    },
    {
      "epoch": 0.8998576174655909,
      "grad_norm": 0.3509832182230347,
      "learning_rate": 3.0103314251040683e-05,
      "loss": 0.7474,
      "step": 2370
    },
    {
      "epoch": 0.9017560512577124,
      "grad_norm": 0.3045155962925135,
      "learning_rate": 2.8980665645095993e-05,
      "loss": 0.7457,
      "step": 2375
    },
    {
      "epoch": 0.9036544850498339,
      "grad_norm": 0.33768892145147017,
      "learning_rate": 2.787872545015069e-05,
      "loss": 0.7369,
      "step": 2380
    },
    {
      "epoch": 0.9055529188419554,
      "grad_norm": 0.30591307319847055,
      "learning_rate": 2.679754211311314e-05,
      "loss": 0.7349,
      "step": 2385
    },
    {
      "epoch": 0.9074513526340768,
      "grad_norm": 0.31814695669489196,
      "learning_rate": 2.5737163168314093e-05,
      "loss": 0.7465,
      "step": 2390
    },
    {
      "epoch": 0.9093497864261983,
      "grad_norm": 0.33557857522275575,
      "learning_rate": 2.4697635235417403e-05,
      "loss": 0.7304,
      "step": 2395
    },
    {
      "epoch": 0.9112482202183199,
      "grad_norm": 0.31077407798760204,
      "learning_rate": 2.3679004017370165e-05,
      "loss": 0.7322,
      "step": 2400
    },
    {
      "epoch": 0.9131466540104414,
      "grad_norm": 0.3696076542535449,
      "learning_rate": 2.2681314298393208e-05,
      "loss": 0.7415,
      "step": 2405
    },
    {
      "epoch": 0.9150450878025629,
      "grad_norm": 0.34756329848102396,
      "learning_rate": 2.1704609942012344e-05,
      "loss": 0.7437,
      "step": 2410
    },
    {
      "epoch": 0.9169435215946844,
      "grad_norm": 0.3105145840944022,
      "learning_rate": 2.074893388912996e-05,
      "loss": 0.7261,
      "step": 2415
    },
    {
      "epoch": 0.9188419553868059,
      "grad_norm": 0.3385109368325919,
      "learning_rate": 1.9814328156136986e-05,
      "loss": 0.7507,
      "step": 2420
    },
    {
      "epoch": 0.9207403891789274,
      "grad_norm": 0.30326901789443217,
      "learning_rate": 1.8900833833065622e-05,
      "loss": 0.7368,
      "step": 2425
    },
    {
      "epoch": 0.9226388229710489,
      "grad_norm": 0.31849955729705076,
      "learning_rate": 1.800849108178304e-05,
      "loss": 0.732,
      "step": 2430
    },
    {
      "epoch": 0.9245372567631703,
      "grad_norm": 0.3023724936604926,
      "learning_rate": 1.7137339134225326e-05,
      "loss": 0.7395,
      "step": 2435
    },
    {
      "epoch": 0.9264356905552918,
      "grad_norm": 0.3045797697728719,
      "learning_rate": 1.628741629067282e-05,
      "loss": 0.7372,
      "step": 2440
    },
    {
      "epoch": 0.9283341243474134,
      "grad_norm": 0.3072440159247055,
      "learning_rate": 1.5458759918066333e-05,
      "loss": 0.7271,
      "step": 2445
    },
    {
      "epoch": 0.9302325581395349,
      "grad_norm": 0.2973843731076206,
      "learning_rate": 1.4651406448364046e-05,
      "loss": 0.7286,
      "step": 2450
    },
    {
      "epoch": 0.9321309919316564,
      "grad_norm": 0.33315733476007137,
      "learning_rate": 1.3865391376940151e-05,
      "loss": 0.7245,
      "step": 2455
    },
    {
      "epoch": 0.9340294257237779,
      "grad_norm": 0.35918744134846453,
      "learning_rate": 1.3100749261024003e-05,
      "loss": 0.7233,
      "step": 2460
    },
    {
      "epoch": 0.9359278595158994,
      "grad_norm": 0.2959597147751244,
      "learning_rate": 1.2357513718180724e-05,
      "loss": 0.7406,
      "step": 2465
    },
    {
      "epoch": 0.9378262933080209,
      "grad_norm": 0.6025608542347619,
      "learning_rate": 1.1635717424833602e-05,
      "loss": 0.7319,
      "step": 2470
    },
    {
      "epoch": 0.9397247271001424,
      "grad_norm": 0.31728158924540384,
      "learning_rate": 1.0935392114827026e-05,
      "loss": 0.7142,
      "step": 2475
    },
    {
      "epoch": 0.9416231608922638,
      "grad_norm": 0.3202471419277355,
      "learning_rate": 1.0256568578031533e-05,
      "loss": 0.7234,
      "step": 2480
    },
    {
      "epoch": 0.9435215946843853,
      "grad_norm": 0.33730636346394754,
      "learning_rate": 9.599276658990353e-06,
      "loss": 0.7266,
      "step": 2485
    },
    {
      "epoch": 0.9454200284765069,
      "grad_norm": 0.3105080016851193,
      "learning_rate": 8.963545255606664e-06,
      "loss": 0.7309,
      "step": 2490
    },
    {
      "epoch": 0.9473184622686284,
      "grad_norm": 0.31936758390953845,
      "learning_rate": 8.349402317873788e-06,
      "loss": 0.7385,
      "step": 2495
    },
    {
      "epoch": 0.9492168960607499,
      "grad_norm": 0.35011965244445487,
      "learning_rate": 7.756874846645834e-06,
      "loss": 0.7298,
      "step": 2500
    },
    {
      "epoch": 0.9511153298528714,
      "grad_norm": 0.3189446744264921,
      "learning_rate": 7.185988892450923e-06,
      "loss": 0.7345,
      "step": 2505
    },
    {
      "epoch": 0.9530137636449929,
      "grad_norm": 0.3463481807220011,
      "learning_rate": 6.636769554345778e-06,
      "loss": 0.7296,
      "step": 2510
    },
    {
      "epoch": 0.9549121974371144,
      "grad_norm": 0.3144784342698344,
      "learning_rate": 6.109240978812047e-06,
      "loss": 0.733,
      "step": 2515
    },
    {
      "epoch": 0.9568106312292359,
      "grad_norm": 0.32358238607927053,
      "learning_rate": 5.603426358695207e-06,
      "loss": 0.7227,
      "step": 2520
    },
    {
      "epoch": 0.9587090650213573,
      "grad_norm": 0.3070502390360057,
      "learning_rate": 5.11934793218427e-06,
      "loss": 0.7379,
      "step": 2525
    },
    {
      "epoch": 0.9606074988134788,
      "grad_norm": 0.30930683987481156,
      "learning_rate": 4.657026981834622e-06,
      "loss": 0.7244,
      "step": 2530
    },
    {
      "epoch": 0.9625059326056004,
      "grad_norm": 0.3184603698723896,
      "learning_rate": 4.216483833631879e-06,
      "loss": 0.7374,
      "step": 2535
    },
    {
      "epoch": 0.9644043663977219,
      "grad_norm": 0.313081890733716,
      "learning_rate": 3.7977378560985487e-06,
      "loss": 0.7253,
      "step": 2540
    },
    {
      "epoch": 0.9663028001898434,
      "grad_norm": 0.33414880766899385,
      "learning_rate": 3.4008074594423233e-06,
      "loss": 0.7246,
      "step": 2545
    },
    {
      "epoch": 0.9682012339819649,
      "grad_norm": 0.30264216877470684,
      "learning_rate": 3.0257100947470027e-06,
      "loss": 0.7274,
      "step": 2550
    },
    {
      "epoch": 0.9700996677740864,
      "grad_norm": 0.34059297981030995,
      "learning_rate": 2.672462253204666e-06,
      "loss": 0.7429,
      "step": 2555
    },
    {
      "epoch": 0.9719981015662079,
      "grad_norm": 0.3255229667279891,
      "learning_rate": 2.3410794653911936e-06,
      "loss": 0.7246,
      "step": 2560
    },
    {
      "epoch": 0.9738965353583294,
      "grad_norm": 0.32454959915539006,
      "learning_rate": 2.03157630058326e-06,
      "loss": 0.7202,
      "step": 2565
    },
    {
      "epoch": 0.9757949691504508,
      "grad_norm": 0.3205716386887886,
      "learning_rate": 1.7439663661176219e-06,
      "loss": 0.7363,
      "step": 2570
    },
    {
      "epoch": 0.9776934029425723,
      "grad_norm": 0.29763916460228407,
      "learning_rate": 1.478262306793099e-06,
      "loss": 0.7319,
      "step": 2575
    },
    {
      "epoch": 0.9795918367346939,
      "grad_norm": 0.3316371045620813,
      "learning_rate": 1.234475804314683e-06,
      "loss": 0.7519,
      "step": 2580
    },
    {
      "epoch": 0.9814902705268154,
      "grad_norm": 0.3105059150408721,
      "learning_rate": 1.012617576779673e-06,
      "loss": 0.7261,
      "step": 2585
    },
    {
      "epoch": 0.9833887043189369,
      "grad_norm": 0.30466660031856074,
      "learning_rate": 8.126973782067171e-07,
      "loss": 0.7321,
      "step": 2590
    },
    {
      "epoch": 0.9852871381110584,
      "grad_norm": 0.3373827673378842,
      "learning_rate": 6.347239981068231e-07,
      "loss": 0.7395,
      "step": 2595
    },
    {
      "epoch": 0.9871855719031799,
      "grad_norm": 0.31614437085864816,
      "learning_rate": 4.787052610970566e-07,
      "loss": 0.7253,
      "step": 2600
    },
    {
      "epoch": 0.9890840056953014,
      "grad_norm": 0.3300895421062348,
      "learning_rate": 3.446480265563712e-07,
      "loss": 0.7329,
      "step": 2605
    },
    {
      "epoch": 0.9909824394874229,
      "grad_norm": 0.33291802253713565,
      "learning_rate": 2.3255818832423892e-07,
      "loss": 0.7245,
      "step": 2610
    },
    {
      "epoch": 0.9928808732795443,
      "grad_norm": 0.32675535698611996,
      "learning_rate": 1.4244067444124652e-07,
      "loss": 0.7218,
      "step": 2615
    },
    {
      "epoch": 0.9947793070716658,
      "grad_norm": 0.32024826764191183,
      "learning_rate": 7.429944693276847e-08,
      "loss": 0.736,
      "step": 2620
    },
    {
      "epoch": 0.9966777408637874,
      "grad_norm": 0.3176106426665133,
      "learning_rate": 2.8137501634439844e-08,
      "loss": 0.7406,
      "step": 2625
    },
    {
      "epoch": 0.9985761746559089,
      "grad_norm": 0.3213640893094606,
      "learning_rate": 3.956868060761565e-09,
      "loss": 0.7448,
      "step": 2630
    },
    {
      "epoch": 0.9997152349311818,
      "eval_loss": 2.2808492183685303,
      "eval_runtime": 8.5401,
      "eval_samples_per_second": 46.018,
      "eval_steps_per_second": 11.592,
      "step": 2633
    },
    {
      "epoch": 0.9997152349311818,
      "step": 2633,
      "total_flos": 32299470028800.0,
      "train_loss": 1.3562805264158209,
      "train_runtime": 3817.2623,
      "train_samples_per_second": 11.039,
      "train_steps_per_second": 0.69
    }
  ],
  "logging_steps": 5,
  "max_steps": 2633,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 32299470028800.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}