diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6827 @@
+{
+  "best_metric": 1.2795084714889526,
+  "best_model_checkpoint": "saved_model/c2s_jun2024/checkpoint-9692",
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 9692,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": NaN,
+      "learning_rate": 0.0,
+      "loss": 72.6113,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.23219108581543,
+      "learning_rate": 2.5e-06,
+      "loss": 74.5495,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.582162857055664,
+      "learning_rate": 7.5e-06,
+      "loss": 73.7367,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.804972648620605,
+      "learning_rate": 1.2e-05,
+      "loss": 72.8853,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.634269714355469,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 70.9592,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.762855529785156,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 66.9603,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.27646827697754,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 61.4318,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.16312026977539,
+      "learning_rate": 3.15e-05,
+      "loss": 53.3651,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 28.43309783935547,
+      "learning_rate": 3.65e-05,
+      "loss": 33.9745,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.292057991027832,
+      "learning_rate": 4.15e-05,
+      "loss": 13.4627,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.148832321166992,
+      "learning_rate": 4.6500000000000005e-05,
+      "loss": 6.8387,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.579999923706055,
+      "learning_rate": 5.1500000000000005e-05,
+      "loss": 4.7847,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.650771141052246,
+      "learning_rate": 5.65e-05,
+      "loss": 4.1684,
+      "step": 120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.5379791259765625,
+      "learning_rate": 6.15e-05,
+      "loss": 3.8221,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.095062732696533,
+      "learning_rate": 6.65e-05,
+      "loss": 3.5635,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.0580973625183105,
+      "learning_rate": 7.15e-05,
+      "loss": 3.4446,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.517209053039551,
+      "learning_rate": 7.65e-05,
+      "loss": 3.2972,
+      "step": 160
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.954787731170654,
+      "learning_rate": 8.15e-05,
+      "loss": 3.2621,
+      "step": 170
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.085761547088623,
+      "learning_rate": 8.65e-05,
+      "loss": 3.2072,
+      "step": 180
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.4346442222595215,
+      "learning_rate": 9.15e-05,
+      "loss": 3.0868,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.535578727722168,
+      "learning_rate": 9.65e-05,
+      "loss": 3.0201,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.239222526550293,
+      "learning_rate": 9.999378367177788e-05,
+      "loss": 2.9792,
+      "step": 210
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.576033592224121,
+      "learning_rate": 9.997306257770411e-05,
+      "loss": 3.0079,
+      "step": 220
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.455887794494629,
+      "learning_rate": 9.995234148363033e-05,
+      "loss": 2.8296,
+      "step": 230
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.566660404205322,
+      "learning_rate": 9.993162038955657e-05,
+      "loss": 2.7655,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.954742908477783,
+      "learning_rate": 9.99108992954828e-05,
+      "loss": 2.5655,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5510752201080322,
+      "learning_rate": 9.989017820140904e-05,
+      "loss": 2.4527,
+      "step": 260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.358351230621338,
+      "learning_rate": 9.986945710733528e-05,
+      "loss": 2.2679,
+      "step": 270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9349524974823,
+      "learning_rate": 9.98487360132615e-05,
+      "loss": 2.1456,
+      "step": 280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3249402046203613,
+      "learning_rate": 9.982801491918775e-05,
+      "loss": 2.0943,
+      "step": 290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.92372989654541,
+      "learning_rate": 9.980729382511397e-05,
+      "loss": 2.0194,
+      "step": 300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.124359130859375,
+      "learning_rate": 9.97865727310402e-05,
+      "loss": 1.9523,
+      "step": 310
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.372561454772949,
+      "learning_rate": 9.976585163696644e-05,
+      "loss": 1.905,
+      "step": 320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5799174308776855,
+      "learning_rate": 9.974513054289267e-05,
+      "loss": 1.9159,
+      "step": 330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1826956272125244,
+      "learning_rate": 9.97244094488189e-05,
+      "loss": 1.8362,
+      "step": 340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1002371311187744,
+      "learning_rate": 9.970368835474514e-05,
+      "loss": 1.844,
+      "step": 350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1345527172088623,
+      "learning_rate": 9.968296726067136e-05,
+      "loss": 1.8084,
+      "step": 360
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9457321166992188,
+      "learning_rate": 9.96622461665976e-05,
+      "loss": 1.7775,
+      "step": 370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9511795043945312,
+      "learning_rate": 9.964152507252383e-05,
+      "loss": 1.7872,
+      "step": 380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9775121212005615,
+      "learning_rate": 9.962080397845007e-05,
+      "loss": 1.7665,
+      "step": 390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9561394453048706,
+      "learning_rate": 9.96000828843763e-05,
+      "loss": 1.7664,
+      "step": 400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7436013221740723,
+      "learning_rate": 9.957936179030253e-05,
+      "loss": 1.7016,
+      "step": 410
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6739649772644043,
+      "learning_rate": 9.955864069622876e-05,
+      "loss": 1.7219,
+      "step": 420
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.940246343612671,
+      "learning_rate": 9.9537919602155e-05,
+      "loss": 1.7174,
+      "step": 430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8286395072937012,
+      "learning_rate": 9.951719850808123e-05,
+      "loss": 1.6698,
+      "step": 440
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0042293071746826,
+      "learning_rate": 9.949647741400747e-05,
+      "loss": 1.6908,
+      "step": 450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6445887088775635,
+      "learning_rate": 9.94757563199337e-05,
+      "loss": 1.6796,
+      "step": 460
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.068713903427124,
+      "learning_rate": 9.945503522585992e-05,
+      "loss": 1.6685,
+      "step": 470
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8053257465362549,
+      "learning_rate": 9.943431413178617e-05,
+      "loss": 1.6522,
+      "step": 480
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.580461859703064,
+      "learning_rate": 9.94135930377124e-05,
+      "loss": 1.6425,
+      "step": 490
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.607007384300232,
+      "learning_rate": 9.939287194363863e-05,
+      "loss": 1.632,
+      "step": 500
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.649885654449463,
+      "learning_rate": 9.937215084956486e-05,
+      "loss": 1.5966,
+      "step": 510
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7667235136032104,
+      "learning_rate": 9.93514297554911e-05,
+      "loss": 1.5942,
+      "step": 520
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.595691442489624,
+      "learning_rate": 9.933070866141732e-05,
+      "loss": 1.609,
+      "step": 530
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5232254266738892,
+      "learning_rate": 9.930998756734357e-05,
+      "loss": 1.5614,
+      "step": 540
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4872910976409912,
+      "learning_rate": 9.928926647326979e-05,
+      "loss": 1.5657,
+      "step": 550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.609491229057312,
+      "learning_rate": 9.926854537919603e-05,
+      "loss": 1.5935,
+      "step": 560
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6403166055679321,
+      "learning_rate": 9.924782428512226e-05,
+      "loss": 1.6159,
+      "step": 570
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6648396253585815,
+      "learning_rate": 9.922710319104848e-05,
+      "loss": 1.6012,
+      "step": 580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6322458982467651,
+      "learning_rate": 9.920638209697473e-05,
+      "loss": 1.5541,
+      "step": 590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5503164529800415,
+      "learning_rate": 9.918566100290095e-05,
+      "loss": 1.5733,
+      "step": 600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6093209981918335,
+      "learning_rate": 9.916493990882719e-05,
+      "loss": 1.5144,
+      "step": 610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6871626377105713,
+      "learning_rate": 9.914421881475342e-05,
+      "loss": 1.573,
+      "step": 620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7600977420806885,
+      "learning_rate": 9.912349772067966e-05,
+      "loss": 1.5577,
+      "step": 630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4892425537109375,
+      "learning_rate": 9.910277662660588e-05,
+      "loss": 1.5751,
+      "step": 640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5667476654052734,
+      "learning_rate": 9.908205553253213e-05,
+      "loss": 1.5298,
+      "step": 650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3411659002304077,
+      "learning_rate": 9.906133443845835e-05,
+      "loss": 1.5409,
+      "step": 660
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5329233407974243,
+      "learning_rate": 9.904061334438459e-05,
+      "loss": 1.5165,
+      "step": 670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5168925523757935,
+      "learning_rate": 9.901989225031082e-05,
+      "loss": 1.5222,
+      "step": 680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6860578060150146,
+      "learning_rate": 9.899917115623706e-05,
+      "loss": 1.5179,
+      "step": 690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4629698991775513,
+      "learning_rate": 9.897845006216329e-05,
+      "loss": 1.5593,
+      "step": 700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3701924085617065,
+      "learning_rate": 9.895772896808953e-05,
+      "loss": 1.52,
+      "step": 710
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4276106357574463,
+      "learning_rate": 9.893700787401575e-05,
+      "loss": 1.5546,
+      "step": 720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5609627962112427,
+      "learning_rate": 9.8916286779942e-05,
+      "loss": 1.5071,
+      "step": 730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9602493047714233,
+      "learning_rate": 9.889556568586822e-05,
+      "loss": 1.5192,
+      "step": 740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4681726694107056,
+      "learning_rate": 9.887484459179444e-05,
+      "loss": 1.5065,
+      "step": 750
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.547143816947937,
+      "learning_rate": 9.885412349772069e-05,
+      "loss": 1.5303,
+      "step": 760
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7585084438323975,
+      "learning_rate": 9.883340240364691e-05,
+      "loss": 1.5412,
+      "step": 770
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4589301347732544,
+      "learning_rate": 9.881268130957315e-05,
+      "loss": 1.5008,
+      "step": 780
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5748664140701294,
+      "learning_rate": 9.879196021549938e-05,
+      "loss": 1.4856,
+      "step": 790
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4392333030700684,
+      "learning_rate": 9.877123912142562e-05,
+      "loss": 1.4593,
+      "step": 800
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.439276933670044,
+      "learning_rate": 9.875051802735185e-05,
+      "loss": 1.4565,
+      "step": 810
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5028575658798218,
+      "learning_rate": 9.872979693327809e-05,
+      "loss": 1.5106,
+      "step": 820
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5902388095855713,
+      "learning_rate": 9.870907583920431e-05,
+      "loss": 1.459,
+      "step": 830
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5270620584487915,
+      "learning_rate": 9.868835474513056e-05,
+      "loss": 1.4705,
+      "step": 840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.354683518409729,
+      "learning_rate": 9.866763365105678e-05,
+      "loss": 1.4468,
+      "step": 850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3697203397750854,
+      "learning_rate": 9.864691255698301e-05,
+      "loss": 1.4669,
+      "step": 860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5006585121154785,
+      "learning_rate": 9.862619146290925e-05,
+      "loss": 1.4641,
+      "step": 870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3566001653671265,
+      "learning_rate": 9.860547036883548e-05,
+      "loss": 1.4545,
+      "step": 880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3500274419784546,
+      "learning_rate": 9.85847492747617e-05,
+      "loss": 1.477,
+      "step": 890
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3306142091751099,
+      "learning_rate": 9.856402818068794e-05,
+      "loss": 1.4469,
+      "step": 900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2983628511428833,
+      "learning_rate": 9.854330708661418e-05,
+      "loss": 1.4603,
+      "step": 910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3828344345092773,
+      "learning_rate": 9.852258599254041e-05,
+      "loss": 1.4686,
+      "step": 920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.427741527557373,
+      "learning_rate": 9.850186489846665e-05,
+      "loss": 1.4756,
+      "step": 930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.267404556274414,
+      "learning_rate": 9.848114380439287e-05,
+      "loss": 1.4777,
+      "step": 940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3213374614715576,
+      "learning_rate": 9.846042271031912e-05,
+      "loss": 1.4526,
+      "step": 950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6813840866088867,
+      "learning_rate": 9.843970161624534e-05,
+      "loss": 1.49,
+      "step": 960
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2110322713851929,
+      "learning_rate": 9.841898052217157e-05,
+      "loss": 1.4796,
+      "step": 970
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3316526412963867,
+      "learning_rate": 9.839825942809781e-05,
+      "loss": 1.4523,
+      "step": 980
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.313766598701477,
+      "learning_rate": 9.837753833402404e-05,
+      "loss": 1.4195,
+      "step": 990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4528905153274536,
+      "learning_rate": 9.835681723995028e-05,
+      "loss": 1.4433,
+      "step": 1000
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3782751560211182,
+      "learning_rate": 9.833609614587651e-05,
+      "loss": 1.4673,
+      "step": 1010
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5674275159835815,
+      "learning_rate": 9.831537505180273e-05,
+      "loss": 1.4296,
+      "step": 1020
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3901402950286865,
+      "learning_rate": 9.829465395772898e-05,
+      "loss": 1.4516,
+      "step": 1030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1594748497009277,
+      "learning_rate": 9.82739328636552e-05,
+      "loss": 1.4225,
+      "step": 1040
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2297048568725586,
+      "learning_rate": 9.825321176958144e-05,
+      "loss": 1.4416,
+      "step": 1050
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1866023540496826,
+      "learning_rate": 9.823249067550768e-05,
+      "loss": 1.444,
+      "step": 1060
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.407461404800415,
+      "learning_rate": 9.82117695814339e-05,
+      "loss": 1.4415,
+      "step": 1070
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.302164912223816,
+      "learning_rate": 9.819104848736013e-05,
+      "loss": 1.4405,
+      "step": 1080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3704490661621094,
+      "learning_rate": 9.817032739328637e-05,
+      "loss": 1.4408,
+      "step": 1090
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2673710584640503,
+      "learning_rate": 9.81496062992126e-05,
+      "loss": 1.4221,
+      "step": 1100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3337206840515137,
+      "learning_rate": 9.812888520513884e-05,
+      "loss": 1.4193,
+      "step": 1110
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3280502557754517,
+      "learning_rate": 9.810816411106507e-05,
+      "loss": 1.4736,
+      "step": 1120
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2532864809036255,
+      "learning_rate": 9.80874430169913e-05,
+      "loss": 1.4665,
+      "step": 1130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2475242614746094,
+      "learning_rate": 9.806672192291754e-05,
+      "loss": 1.426,
+      "step": 1140
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7034567594528198,
+      "learning_rate": 9.804600082884376e-05,
+      "loss": 1.4473,
+      "step": 1150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3586080074310303,
+      "learning_rate": 9.802527973477e-05,
+      "loss": 1.3959,
+      "step": 1160
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2611415386199951,
+      "learning_rate": 9.800455864069623e-05,
+      "loss": 1.4401,
+      "step": 1170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3101681470870972,
+      "learning_rate": 9.798383754662247e-05,
+      "loss": 1.4431,
+      "step": 1180
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1770988702774048,
+      "learning_rate": 9.796311645254869e-05,
+      "loss": 1.4108,
+      "step": 1190
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2325702905654907,
+      "learning_rate": 9.794239535847494e-05,
+      "loss": 1.4141,
+      "step": 1200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2543164491653442,
+      "learning_rate": 9.792167426440116e-05,
+      "loss": 1.4133,
+      "step": 1210
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1258199214935303,
+      "learning_rate": 9.79009531703274e-05,
+      "loss": 1.4041,
+      "step": 1220
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3423455953598022,
+      "learning_rate": 9.788023207625363e-05,
+      "loss": 1.4144,
+      "step": 1230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.248947024345398,
+      "learning_rate": 9.785951098217985e-05,
+      "loss": 1.4043,
+      "step": 1240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.129650354385376,
+      "learning_rate": 9.78387898881061e-05,
+      "loss": 1.4216,
+      "step": 1250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2218910455703735,
+      "learning_rate": 9.781806879403232e-05,
+      "loss": 1.3976,
+      "step": 1260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.153981328010559,
+      "learning_rate": 9.779734769995856e-05,
+      "loss": 1.4304,
+      "step": 1270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1724766492843628,
+      "learning_rate": 9.77766266058848e-05,
+      "loss": 1.43,
+      "step": 1280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2830730676651,
+      "learning_rate": 9.775590551181103e-05,
+      "loss": 1.4429,
+      "step": 1290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2320913076400757,
+      "learning_rate": 9.773518441773725e-05,
+      "loss": 1.3898,
+      "step": 1300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2313491106033325,
+      "learning_rate": 9.77144633236635e-05,
+      "loss": 1.4273,
+      "step": 1310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1946086883544922,
+      "learning_rate": 9.769374222958972e-05,
+      "loss": 1.4234,
+      "step": 1320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.127300500869751,
+      "learning_rate": 9.767302113551596e-05,
+      "loss": 1.4144,
+      "step": 1330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4888228178024292,
+      "learning_rate": 9.765230004144219e-05,
+      "loss": 1.4092,
+      "step": 1340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3795928955078125,
+      "learning_rate": 9.763157894736843e-05,
+      "loss": 1.3647,
+      "step": 1350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1433610916137695,
+      "learning_rate": 9.761085785329466e-05,
+      "loss": 1.415,
+      "step": 1360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.040281891822815,
+      "learning_rate": 9.75901367592209e-05,
+      "loss": 1.4244,
+      "step": 1370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1311726570129395,
+      "learning_rate": 9.756941566514712e-05,
+      "loss": 1.3852,
+      "step": 1380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2847346067428589,
+      "learning_rate": 9.754869457107337e-05,
+      "loss": 1.4225,
+      "step": 1390
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2235894203186035,
+      "learning_rate": 9.752797347699959e-05,
+      "loss": 1.3973,
+      "step": 1400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1802481412887573,
+      "learning_rate": 9.750725238292582e-05,
+      "loss": 1.3923,
+      "step": 1410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.141739010810852,
+      "learning_rate": 9.748653128885206e-05,
+      "loss": 1.4049,
+      "step": 1420
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2155243158340454,
+      "learning_rate": 9.746581019477828e-05,
+      "loss": 1.3866,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4717819690704346,
+      "learning_rate": 9.744508910070453e-05,
+      "loss": 1.4264,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1440094709396362,
+      "learning_rate": 9.742436800663075e-05,
+      "loss": 1.4291,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3254936933517456,
+      "learning_rate": 9.740364691255699e-05,
+      "loss": 1.3973,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2041431665420532,
+      "learning_rate": 9.738292581848322e-05,
+      "loss": 1.3779,
+      "step": 1470
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1422394514083862,
+      "learning_rate": 9.736220472440946e-05,
+      "loss": 1.3918,
+      "step": 1480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2341557741165161,
+      "learning_rate": 9.734148363033568e-05,
+      "loss": 1.4065,
+      "step": 1490
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1723967790603638,
+      "learning_rate": 9.732076253626193e-05,
+      "loss": 1.4003,
+      "step": 1500
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.279010534286499,
+      "learning_rate": 9.730004144218815e-05,
+      "loss": 1.3762,
+      "step": 1510
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2639541625976562,
+      "learning_rate": 9.727932034811438e-05,
+      "loss": 1.3932,
+      "step": 1520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1406500339508057,
+      "learning_rate": 9.725859925404062e-05,
+      "loss": 1.4318,
+      "step": 1530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1991297006607056,
+      "learning_rate": 9.723787815996685e-05,
+      "loss": 1.3742,
+      "step": 1540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1058017015457153,
+      "learning_rate": 9.721715706589309e-05,
+      "loss": 1.3975,
+      "step": 1550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3658838272094727,
+      "learning_rate": 9.719643597181932e-05,
+      "loss": 1.4245,
+      "step": 1560
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0663561820983887,
+      "learning_rate": 9.717571487774555e-05,
+      "loss": 1.3779,
+      "step": 1570
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1523654460906982,
+      "learning_rate": 9.715499378367178e-05,
+      "loss": 1.4306,
+      "step": 1580
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.223913311958313,
+      "learning_rate": 9.713427268959802e-05,
+      "loss": 1.3748,
+      "step": 1590
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0876872539520264,
+      "learning_rate": 9.711355159552424e-05,
+      "loss": 1.3806,
+      "step": 1600
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3317033052444458,
+      "learning_rate": 9.709283050145049e-05,
+      "loss": 1.3586,
+      "step": 1610
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2402222156524658,
+      "learning_rate": 9.707210940737671e-05,
+      "loss": 1.3886,
+      "step": 1620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1467841863632202,
+      "learning_rate": 9.705138831330294e-05,
+      "loss": 1.3634,
+      "step": 1630
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1589218378067017,
+      "learning_rate": 9.703066721922918e-05,
+      "loss": 1.3466,
+      "step": 1640
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9369345307350159,
+      "learning_rate": 9.700994612515541e-05,
+      "loss": 1.3819,
+      "step": 1650
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0450528860092163,
+      "learning_rate": 9.698922503108165e-05,
+      "loss": 1.3482,
+      "step": 1660
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0236886739730835,
+      "learning_rate": 9.696850393700788e-05,
+      "loss": 1.3468,
+      "step": 1670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0324066877365112,
+      "learning_rate": 9.69477828429341e-05,
+      "loss": 1.3926,
+      "step": 1680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1705087423324585,
+      "learning_rate": 9.692706174886035e-05,
+      "loss": 1.3547,
+      "step": 1690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1479854583740234,
+      "learning_rate": 9.690634065478658e-05,
+      "loss": 1.3517,
+      "step": 1700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1700282096862793,
+      "learning_rate": 9.688561956071281e-05,
+      "loss": 1.3635,
+      "step": 1710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.079822301864624,
+      "learning_rate": 9.686489846663905e-05,
+      "loss": 1.3878,
+      "step": 1720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.188466191291809,
+      "learning_rate": 9.684417737256528e-05,
+      "loss": 1.36,
+      "step": 1730
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1050995588302612,
+      "learning_rate": 9.68234562784915e-05,
+      "loss": 1.3513,
+      "step": 1740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2480050325393677,
+      "learning_rate": 9.680273518441774e-05,
+      "loss": 1.362,
+      "step": 1750
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1782851219177246,
+      "learning_rate": 9.678201409034397e-05,
+      "loss": 1.378,
+      "step": 1760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1327308416366577,
+      "learning_rate": 9.676129299627021e-05,
+      "loss": 1.3836,
+      "step": 1770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0974417924880981,
+      "learning_rate": 9.674057190219644e-05,
+      "loss": 1.3589,
+      "step": 1780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1006550788879395,
+      "learning_rate": 9.671985080812266e-05,
+      "loss": 1.3734,
+      "step": 1790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0745372772216797,
+      "learning_rate": 9.669912971404891e-05,
+      "loss": 1.4078,
+      "step": 1800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2572031021118164,
+      "learning_rate": 9.667840861997513e-05,
+      "loss": 1.3535,
+      "step": 1810
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.065767526626587,
+      "learning_rate": 9.665768752590137e-05,
+      "loss": 1.3657,
+      "step": 1820
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2773245573043823,
+      "learning_rate": 9.66369664318276e-05,
+      "loss": 1.3813,
+      "step": 1830
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0642096996307373,
+      "learning_rate": 9.661624533775384e-05,
+      "loss": 1.3829,
+      "step": 1840
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1348739862442017,
+      "learning_rate": 9.659552424368008e-05,
+      "loss": 1.3864,
+      "step": 1850
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.136107087135315,
+      "learning_rate": 9.657480314960631e-05,
+      "loss": 1.3523,
+      "step": 1860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1533474922180176,
+      "learning_rate": 9.655408205553253e-05,
+      "loss": 1.3669,
+      "step": 1870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1027289628982544,
+      "learning_rate": 9.653336096145878e-05,
+      "loss": 1.3256,
+      "step": 1880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9988449811935425,
+      "learning_rate": 9.6512639867385e-05,
+      "loss": 1.4024,
+      "step": 1890
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2975176572799683,
+      "learning_rate": 9.649191877331124e-05,
+      "loss": 1.3751,
+      "step": 1900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2186543941497803,
+      "learning_rate": 9.647119767923747e-05,
+      "loss": 1.3444,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1342490911483765,
+      "learning_rate": 9.64504765851637e-05,
+      "loss": 1.3449,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.148695707321167,
+      "learning_rate": 9.642975549108993e-05,
+      "loss": 1.3325,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9545331001281738,
+      "learning_rate": 9.640903439701616e-05,
+      "loss": 1.3375,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0941437482833862,
+      "learning_rate": 9.63883133029424e-05,
+      "loss": 1.3671,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0803030729293823,
+      "learning_rate": 9.636759220886863e-05,
+      "loss": 1.3648,
+      "step": 1960
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0937373638153076,
+      "learning_rate": 9.634687111479487e-05,
+      "loss": 1.3518,
+      "step": 1970
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1884483098983765,
+      "learning_rate": 9.632615002072109e-05,
+      "loss": 1.3461,
+      "step": 1980
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1179327964782715,
+      "learning_rate": 9.630542892664734e-05,
+      "loss": 1.3765,
+      "step": 1990
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9843894839286804,
+      "learning_rate": 9.628470783257356e-05,
+      "loss": 1.3379,
+      "step": 2000
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0279515981674194,
+      "learning_rate": 9.62639867384998e-05,
+      "loss": 1.3389,
+      "step": 2010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0797231197357178,
+      "learning_rate": 9.624326564442603e-05,
+      "loss": 1.346,
+      "step": 2020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1976298093795776,
+      "learning_rate": 9.622254455035227e-05,
+      "loss": 1.3366,
+      "step": 2030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.057880163192749,
+      "learning_rate": 9.620182345627849e-05,
+      "loss": 1.3264,
+      "step": 2040
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1059492826461792,
+      "learning_rate": 9.618110236220474e-05,
+      "loss": 1.3446,
+      "step": 2050
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0970298051834106,
+      "learning_rate": 9.616038126813096e-05,
+      "loss": 1.3521,
+      "step": 2060
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0951462984085083,
+      "learning_rate": 9.61396601740572e-05,
+      "loss": 1.3669,
+      "step": 2070
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0926049947738647,
+      "learning_rate": 9.611893907998343e-05,
+      "loss": 1.354,
+      "step": 2080
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0136979818344116,
+      "learning_rate": 9.609821798590965e-05,
+      "loss": 1.3321,
+      "step": 2090
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.129214882850647,
+      "learning_rate": 9.60774968918359e-05,
+      "loss": 1.382,
+      "step": 2100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1166954040527344,
+      "learning_rate": 9.605677579776212e-05,
+      "loss": 1.3337,
+      "step": 2110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.204231858253479,
+      "learning_rate": 9.603605470368836e-05,
+      "loss": 1.3642,
+      "step": 2120
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0265048742294312,
+      "learning_rate": 9.601533360961459e-05,
+      "loss": 1.3662,
+      "step": 2130
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0513389110565186,
+      "learning_rate": 9.599461251554083e-05,
+      "loss": 1.3395,
+      "step": 2140
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.17727792263031,
+      "learning_rate": 9.597389142146705e-05,
+      "loss": 1.3738,
+      "step": 2150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0676214694976807,
+      "learning_rate": 9.59531703273933e-05,
+      "loss": 1.3383,
+      "step": 2160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9273681640625,
+      "learning_rate": 9.593244923331952e-05,
+      "loss": 1.367,
+      "step": 2170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0774747133255005,
+      "learning_rate": 9.591172813924575e-05,
+      "loss": 1.3369,
+      "step": 2180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.131264090538025,
+      "learning_rate": 9.589100704517199e-05,
+      "loss": 1.3457,
+      "step": 2190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.106242060661316,
+      "learning_rate": 9.587028595109822e-05,
+      "loss": 1.321,
+      "step": 2200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.054598331451416,
+      "learning_rate": 9.584956485702446e-05,
+      "loss": 1.3424,
+      "step": 2210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0380080938339233,
+      "learning_rate": 9.58288437629507e-05,
+      "loss": 1.3425,
+      "step": 2220
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1068315505981445,
+      "learning_rate": 9.580812266887692e-05,
+      "loss": 1.321,
+      "step": 2230
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1228212118148804,
+      "learning_rate": 9.578740157480316e-05,
+      "loss": 1.3301,
+      "step": 2240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9643247127532959,
+      "learning_rate": 9.576668048072939e-05,
+      "loss": 1.3403,
+      "step": 2250
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9587458372116089,
+      "learning_rate": 9.574595938665562e-05,
+      "loss": 1.3402,
+      "step": 2260
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0192015171051025,
+      "learning_rate": 9.572523829258186e-05,
+      "loss": 1.3595,
+      "step": 2270
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1033486127853394,
+      "learning_rate": 9.570451719850808e-05,
+      "loss": 1.3515,
+      "step": 2280
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.244828462600708,
+      "learning_rate": 9.568379610443431e-05,
+      "loss": 1.3148,
+      "step": 2290
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.031778335571289,
+      "learning_rate": 9.566307501036055e-05,
+      "loss": 1.3343,
+      "step": 2300
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0581692457199097,
+      "learning_rate": 9.564235391628678e-05,
+      "loss": 1.3352,
+      "step": 2310
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9989519119262695,
+      "learning_rate": 9.562163282221302e-05,
+      "loss": 1.3206,
+      "step": 2320
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1149669885635376,
+      "learning_rate": 9.560091172813925e-05,
+      "loss": 1.3355,
+      "step": 2330
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1359626054763794,
+      "learning_rate": 9.558019063406548e-05,
+      "loss": 1.3233,
+      "step": 2340
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1091575622558594,
+      "learning_rate": 9.555946953999172e-05,
+      "loss": 1.3678,
+      "step": 2350
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0405771732330322,
+      "learning_rate": 9.553874844591795e-05,
+      "loss": 1.3555,
+      "step": 2360
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.068385124206543,
+      "learning_rate": 9.551802735184418e-05,
+      "loss": 1.346,
+      "step": 2370
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0115128755569458,
+      "learning_rate": 9.549730625777042e-05,
+      "loss": 1.3448,
+      "step": 2380
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.026138424873352,
+      "learning_rate": 9.547658516369665e-05,
+      "loss": 1.3286,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.375127911567688,
+      "learning_rate": 9.545586406962289e-05,
+      "loss": 1.3931,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2297391891479492,
+      "learning_rate": 9.543514297554912e-05,
+      "loss": 1.3223,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.143249750137329,
+      "learning_rate": 9.541442188147534e-05,
+      "loss": 1.3142,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1182348728179932,
+      "learning_rate": 9.539370078740158e-05,
+      "loss": 1.3414,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0450687408447266,
+      "learning_rate": 9.537297969332781e-05,
+      "loss": 1.3119,
+      "step": 2440
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.105624794960022,
+      "learning_rate": 9.535225859925403e-05,
+      "loss": 1.3275,
+      "step": 2450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1117305755615234,
+      "learning_rate": 9.533153750518028e-05,
+      "loss": 1.3384,
+      "step": 2460
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.122660756111145,
+      "learning_rate": 9.53108164111065e-05,
+      "loss": 1.3509,
+      "step": 2470
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.994361937046051,
+      "learning_rate": 9.529009531703274e-05,
+      "loss": 1.3638,
+      "step": 2480
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1339287757873535,
+      "learning_rate": 9.526937422295898e-05,
+      "loss": 1.3282,
+      "step": 2490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0273313522338867,
+      "learning_rate": 9.524865312888521e-05,
+      "loss": 1.3261,
+      "step": 2500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.067122220993042,
+      "learning_rate": 9.522793203481145e-05,
+      "loss": 1.3502,
+      "step": 2510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9780186414718628,
+      "learning_rate": 9.520721094073768e-05,
+      "loss": 1.3209,
+      "step": 2520
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0634074211120605,
+      "learning_rate": 9.51864898466639e-05,
+      "loss": 1.3508,
+      "step": 2530
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0088226795196533,
+      "learning_rate": 9.516576875259015e-05,
+      "loss": 1.2848,
+      "step": 2540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.156569242477417,
+      "learning_rate": 9.514504765851637e-05,
+      "loss": 1.3336,
+      "step": 2550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9981438517570496,
+      "learning_rate": 9.512432656444261e-05,
+      "loss": 1.3237,
+      "step": 2560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0465401411056519,
+      "learning_rate": 9.510360547036884e-05,
+      "loss": 1.3347,
+      "step": 2570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.282174825668335,
+      "learning_rate": 9.508288437629508e-05,
+      "loss": 1.3234,
+      "step": 2580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0925480127334595,
+      "learning_rate": 9.50621632822213e-05,
+      "loss": 1.3604,
+      "step": 2590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0559757947921753,
+      "learning_rate": 9.504144218814753e-05,
+      "loss": 1.3411,
+      "step": 2600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0160987377166748,
+      "learning_rate": 9.502072109407377e-05,
+      "loss": 1.3299,
+      "step": 2610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0814076662063599,
+      "learning_rate": 9.5e-05,
+      "loss": 1.3053,
+      "step": 2620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1541906595230103,
+      "learning_rate": 9.497927890592624e-05,
+      "loss": 1.3368,
+      "step": 2630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0476430654525757,
+      "learning_rate": 9.495855781185246e-05,
+      "loss": 1.3266,
+      "step": 2640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0859614610671997,
+      "learning_rate": 9.493783671777871e-05,
+      "loss": 1.3077,
+      "step": 2650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.047561526298523,
+      "learning_rate": 9.491711562370493e-05,
+      "loss": 1.301,
+      "step": 2660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1071749925613403,
+      "learning_rate": 9.489639452963117e-05,
+      "loss": 1.3069,
+      "step": 2670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0207133293151855,
+      "learning_rate": 9.48756734355574e-05,
+      "loss": 1.3128,
+      "step": 2680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1883114576339722,
+      "learning_rate": 9.485495234148364e-05,
+      "loss": 1.2987,
+      "step": 2690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1708128452301025,
+      "learning_rate": 9.483423124740986e-05,
+      "loss": 1.3386,
+      "step": 2700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0731940269470215,
+      "learning_rate": 9.481351015333611e-05,
+      "loss": 1.3165,
+      "step": 2710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.02231764793396,
+      "learning_rate": 9.479278905926233e-05,
+      "loss": 1.3364,
+      "step": 2720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9825921654701233,
+      "learning_rate": 9.477206796518856e-05,
+      "loss": 1.3078,
+      "step": 2730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1280665397644043,
+      "learning_rate": 9.47513468711148e-05,
+      "loss": 1.3359,
+      "step": 2740
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9910861253738403,
+      "learning_rate": 9.473062577704103e-05,
+      "loss": 1.3361,
+      "step": 2750
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0153850317001343,
+      "learning_rate": 9.470990468296727e-05,
+      "loss": 1.3059,
+      "step": 2760
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.01111900806427,
+      "learning_rate": 9.468918358889349e-05,
+      "loss": 1.3226,
+      "step": 2770
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0714573860168457,
+      "learning_rate": 9.466846249481973e-05,
+      "loss": 1.3195,
+      "step": 2780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0012733936309814,
+      "learning_rate": 9.464774140074596e-05,
+      "loss": 1.2948,
+      "step": 2790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9637882709503174,
+      "learning_rate": 9.46270203066722e-05,
+      "loss": 1.3208,
+      "step": 2800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0453296899795532,
+      "learning_rate": 9.460629921259843e-05,
+      "loss": 1.3095,
+      "step": 2810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0107698440551758,
+      "learning_rate": 9.458557811852467e-05,
+      "loss": 1.3164,
+      "step": 2820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1132638454437256,
+      "learning_rate": 9.456485702445089e-05,
+      "loss": 1.3162,
+      "step": 2830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0389189720153809,
+      "learning_rate": 9.454413593037714e-05,
+      "loss": 1.317,
+      "step": 2840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0654906034469604,
+      "learning_rate": 9.452341483630336e-05,
+      "loss": 1.305,
+      "step": 2850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2564867734909058,
+      "learning_rate": 9.45026937422296e-05,
+      "loss": 1.3301,
+      "step": 2860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0308964252471924,
+      "learning_rate": 9.448197264815583e-05,
+      "loss": 1.334,
+      "step": 2870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0542854070663452,
+      "learning_rate": 9.446125155408206e-05,
+      "loss": 1.3001,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2161365747451782,
+      "learning_rate": 9.444053046000829e-05,
+      "loss": 1.2985,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.206581473350525,
+      "learning_rate": 9.441980936593454e-05,
+      "loss": 1.3177,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1631922721862793,
+      "learning_rate": 9.439908827186076e-05,
+      "loss": 1.3269,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9827607274055481,
+      "learning_rate": 9.437836717778699e-05,
+      "loss": 1.3228,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0078628063201904,
+      "learning_rate": 9.435764608371323e-05,
+      "loss": 1.3047,
+      "step": 2930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1704260110855103,
+      "learning_rate": 9.433692498963945e-05,
+      "loss": 1.3074,
+      "step": 2940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.075964093208313,
+      "learning_rate": 9.43162038955657e-05,
+      "loss": 1.3252,
+      "step": 2950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9463378190994263,
+      "learning_rate": 9.429548280149192e-05,
+      "loss": 1.3201,
+      "step": 2960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.01523756980896,
+      "learning_rate": 9.427476170741815e-05,
+      "loss": 1.3127,
+      "step": 2970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9392449259757996,
+      "learning_rate": 9.425404061334439e-05,
+      "loss": 1.3254,
+      "step": 2980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.01919424533844,
+      "learning_rate": 9.423331951927062e-05,
+      "loss": 1.3021,
+      "step": 2990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1243764162063599,
+      "learning_rate": 9.421259842519685e-05,
+      "loss": 1.3112,
+      "step": 3000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0084974765777588,
+      "learning_rate": 9.41918773311231e-05,
+      "loss": 1.3173,
+      "step": 3010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9945486783981323,
+      "learning_rate": 9.417115623704932e-05,
+      "loss": 1.3114,
+      "step": 3020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1148301362991333,
+      "learning_rate": 9.415043514297555e-05,
+      "loss": 1.3275,
+      "step": 3030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2701823711395264,
+      "learning_rate": 9.412971404890179e-05,
+      "loss": 1.3094,
+      "step": 3040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1923747062683105,
+      "learning_rate": 9.410899295482802e-05,
+      "loss": 1.2812,
+      "step": 3050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2106274366378784,
+      "learning_rate": 9.408827186075426e-05,
+      "loss": 1.3011,
+      "step": 3060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0127681493759155,
+      "learning_rate": 9.406755076668049e-05,
+      "loss": 1.3059,
+      "step": 3070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.042222499847412,
+      "learning_rate": 9.404682967260671e-05,
+      "loss": 1.2961,
+      "step": 3080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9650092124938965,
+      "learning_rate": 9.402610857853296e-05,
+      "loss": 1.3264,
+      "step": 3090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0504155158996582,
+      "learning_rate": 9.400538748445918e-05,
+      "loss": 1.2853,
+      "step": 3100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0501419305801392,
+      "learning_rate": 9.39846663903854e-05,
+      "loss": 1.3179,
+      "step": 3110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.056299090385437,
+      "learning_rate": 9.396394529631165e-05,
+      "loss": 1.2962,
+      "step": 3120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0278836488723755,
+      "learning_rate": 9.394322420223788e-05,
+      "loss": 1.2828,
+      "step": 3130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9813990592956543,
+      "learning_rate": 9.392250310816411e-05,
+      "loss": 1.2986,
+      "step": 3140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0665332078933716,
+      "learning_rate": 9.390178201409035e-05,
+      "loss": 1.2891,
+      "step": 3150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0281347036361694,
+      "learning_rate": 9.388106092001658e-05,
+      "loss": 1.299,
+      "step": 3160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0530226230621338,
+      "learning_rate": 9.386033982594282e-05,
+      "loss": 1.2887,
+      "step": 3170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0053261518478394,
+      "learning_rate": 9.383961873186905e-05,
+      "loss": 1.327,
+      "step": 3180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1362097263336182,
+      "learning_rate": 9.381889763779527e-05,
+      "loss": 1.3001,
+      "step": 3190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0610814094543457,
+      "learning_rate": 9.379817654372152e-05,
+      "loss": 1.2535,
+      "step": 3200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9906120300292969,
+      "learning_rate": 9.377745544964774e-05,
+      "loss": 1.291,
+      "step": 3210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0676803588867188,
+      "learning_rate": 9.375673435557398e-05,
+      "loss": 1.3032,
+      "step": 3220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.056851863861084,
+      "learning_rate": 9.373601326150021e-05,
+      "loss": 1.2879,
+      "step": 3230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.048841118812561,
+      "learning_rate": 9.371529216742645e-05,
+      "loss": 1.2798,
+      "step": 3240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.047361969947815,
+      "learning_rate": 9.369457107335268e-05,
+      "loss": 1.3195,
+      "step": 3250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.076904296875,
+      "learning_rate": 9.367384997927892e-05,
+      "loss": 1.3013,
+      "step": 3260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0863533020019531,
+      "learning_rate": 9.365312888520514e-05,
+      "loss": 1.2971,
+      "step": 3270
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0460786819458008,
+      "learning_rate": 9.363240779113138e-05,
+      "loss": 1.3023,
+      "step": 3280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.906493604183197,
+      "learning_rate": 9.361168669705761e-05,
+      "loss": 1.3053,
+      "step": 3290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1181541681289673,
+      "learning_rate": 9.359096560298383e-05,
+      "loss": 1.3142,
+      "step": 3300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0198432207107544,
+      "learning_rate": 9.357024450891008e-05,
+      "loss": 1.293,
+      "step": 3310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0075292587280273,
+      "learning_rate": 9.35495234148363e-05,
+      "loss": 1.299,
+      "step": 3320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.024592399597168,
+      "learning_rate": 9.352880232076254e-05,
+      "loss": 1.2983,
+      "step": 3330
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9931455254554749,
+      "learning_rate": 9.350808122668877e-05,
+      "loss": 1.279,
+      "step": 3340
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0673152208328247,
+      "learning_rate": 9.348736013261501e-05,
+      "loss": 1.2816,
+      "step": 3350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.068587303161621,
+      "learning_rate": 9.346663903854124e-05,
+      "loss": 1.2934,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9838789701461792,
+      "learning_rate": 9.344591794446748e-05,
+      "loss": 1.2917,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0613404512405396,
+      "learning_rate": 9.34251968503937e-05,
+      "loss": 1.2879,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0173070430755615,
+      "learning_rate": 9.340447575631995e-05,
+      "loss": 1.2966,
+      "step": 3390
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1227622032165527,
+      "learning_rate": 9.338375466224617e-05,
+      "loss": 1.2554,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.007338523864746,
+      "learning_rate": 9.33630335681724e-05,
+      "loss": 1.3115,
+      "step": 3410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0479813814163208,
+      "learning_rate": 9.334231247409864e-05,
+      "loss": 1.3048,
+      "step": 3420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0560479164123535,
+      "learning_rate": 9.332159138002486e-05,
+      "loss": 1.2919,
+      "step": 3430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1081204414367676,
+      "learning_rate": 9.33008702859511e-05,
+      "loss": 1.2967,
+      "step": 3440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0260145664215088,
+      "learning_rate": 9.328014919187733e-05,
+      "loss": 1.3195,
+      "step": 3450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.057966947555542,
+      "learning_rate": 9.325942809780357e-05,
+      "loss": 1.2896,
+      "step": 3460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0711556673049927,
+      "learning_rate": 9.32387070037298e-05,
+      "loss": 1.2817,
+      "step": 3470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0118924379348755,
+      "learning_rate": 9.321798590965604e-05,
+      "loss": 1.3052,
+      "step": 3480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0227614641189575,
+      "learning_rate": 9.319726481558226e-05,
+      "loss": 1.3186,
+      "step": 3490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0655134916305542,
+      "learning_rate": 9.317654372150851e-05,
+      "loss": 1.3087,
+      "step": 3500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1255359649658203,
+      "learning_rate": 9.315582262743473e-05,
+      "loss": 1.2749,
+      "step": 3510
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0832923650741577,
+      "learning_rate": 9.313510153336096e-05,
+      "loss": 1.2892,
+      "step": 3520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.067236304283142,
+      "learning_rate": 9.31143804392872e-05,
+      "loss": 1.284,
+      "step": 3530
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1556322574615479,
+      "learning_rate": 9.309365934521344e-05,
+      "loss": 1.2604,
+      "step": 3540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.151723861694336,
+      "learning_rate": 9.307293825113966e-05,
+      "loss": 1.3045,
+      "step": 3550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0258938074111938,
+      "learning_rate": 9.30522171570659e-05,
+      "loss": 1.2859,
+      "step": 3560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0165237188339233,
+      "learning_rate": 9.303149606299213e-05,
+      "loss": 1.3212,
+      "step": 3570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9969586133956909,
+      "learning_rate": 9.301077496891836e-05,
+      "loss": 1.3038,
+      "step": 3580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1335457563400269,
+      "learning_rate": 9.29900538748446e-05,
+      "loss": 1.2747,
+      "step": 3590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0744903087615967,
+      "learning_rate": 9.296933278077082e-05,
+      "loss": 1.3078,
+      "step": 3600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2294646501541138,
+      "learning_rate": 9.294861168669707e-05,
+      "loss": 1.2631,
+      "step": 3610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0542582273483276,
+      "learning_rate": 9.292789059262329e-05,
+      "loss": 1.2778,
+      "step": 3620
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0787122249603271,
+      "learning_rate": 9.290716949854952e-05,
+      "loss": 1.2952,
+      "step": 3630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.182387351989746,
+      "learning_rate": 9.288644840447576e-05,
+      "loss": 1.2955,
+      "step": 3640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0466411113739014,
+      "learning_rate": 9.2865727310402e-05,
+      "loss": 1.3085,
+      "step": 3650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0271363258361816,
+      "learning_rate": 9.284500621632823e-05,
+      "loss": 1.2881,
+      "step": 3660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1320871114730835,
+      "learning_rate": 9.282428512225446e-05,
+      "loss": 1.2671,
+      "step": 3670
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.1176432371139526,
+      "learning_rate": 9.280356402818069e-05,
+      "loss": 1.299,
+      "step": 3680
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9895033240318298,
+      "learning_rate": 9.278284293410694e-05,
+      "loss": 1.2984,
+      "step": 3690
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.191007137298584,
+      "learning_rate": 9.276212184003316e-05,
+      "loss": 1.2808,
+      "step": 3700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0878729820251465,
+      "learning_rate": 9.274140074595939e-05,
+      "loss": 1.2864,
+      "step": 3710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.1144053936004639,
+      "learning_rate": 9.272067965188563e-05,
+      "loss": 1.3175,
+      "step": 3720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.128405213356018,
+      "learning_rate": 9.269995855781186e-05,
+      "loss": 1.3147,
+      "step": 3730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0539438724517822,
+      "learning_rate": 9.267923746373808e-05,
+      "loss": 1.2927,
+      "step": 3740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0515836477279663,
+      "learning_rate": 9.265851636966433e-05,
+      "loss": 1.2743,
+      "step": 3750
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1526955366134644,
+      "learning_rate": 9.263779527559055e-05,
+      "loss": 1.2911,
+      "step": 3760
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.010903000831604,
+      "learning_rate": 9.261707418151679e-05,
+      "loss": 1.2735,
+      "step": 3770
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1373246908187866,
+      "learning_rate": 9.259635308744302e-05,
+      "loss": 1.2952,
+      "step": 3780
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9458179473876953,
+      "learning_rate": 9.257563199336925e-05,
+      "loss": 1.2936,
+      "step": 3790
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.949252724647522,
+      "learning_rate": 9.25549108992955e-05,
+      "loss": 1.287,
+      "step": 3800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1074215173721313,
+      "learning_rate": 9.253418980522172e-05,
+      "loss": 1.2937,
+      "step": 3810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.995959460735321,
+      "learning_rate": 9.251346871114795e-05,
+      "loss": 1.265,
+      "step": 3820
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0461138486862183,
+      "learning_rate": 9.249274761707419e-05,
+      "loss": 1.2822,
+      "step": 3830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0449700355529785,
+      "learning_rate": 9.247202652300042e-05,
+      "loss": 1.2896,
+      "step": 3840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0590384006500244,
+      "learning_rate": 9.245130542892664e-05,
+      "loss": 1.2923,
+      "step": 3850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.178272008895874,
+      "learning_rate": 9.243058433485289e-05,
+      "loss": 1.2797,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0651668310165405,
+      "learning_rate": 9.240986324077911e-05,
+      "loss": 1.2632,
+      "step": 3870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0944633483886719,
+      "learning_rate": 9.238914214670535e-05,
+      "loss": 1.2853,
+      "step": 3880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.042576551437378,
+      "learning_rate": 9.236842105263158e-05,
+      "loss": 1.2884,
+      "step": 3890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1282780170440674,
+      "learning_rate": 9.234769995855782e-05,
+      "loss": 1.2937,
+      "step": 3900
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9996076822280884,
+      "learning_rate": 9.232697886448405e-05,
+      "loss": 1.2657,
+      "step": 3910
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9630957245826721,
+      "learning_rate": 9.230625777041029e-05,
+      "loss": 1.2679,
+      "step": 3920
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0428036451339722,
+      "learning_rate": 9.228553667633651e-05,
+      "loss": 1.2921,
+      "step": 3930
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1940115690231323,
+      "learning_rate": 9.226481558226275e-05,
+      "loss": 1.2759,
+      "step": 3940
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1081199645996094,
+      "learning_rate": 9.224409448818898e-05,
+      "loss": 1.2636,
+      "step": 3950
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.989032506942749,
+      "learning_rate": 9.22233733941152e-05,
+      "loss": 1.2489,
+      "step": 3960
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0575727224349976,
+      "learning_rate": 9.220265230004145e-05,
+      "loss": 1.2777,
+      "step": 3970
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0007938146591187,
+      "learning_rate": 9.218193120596767e-05,
+      "loss": 1.286,
+      "step": 3980
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0560977458953857,
+      "learning_rate": 9.216121011189391e-05,
+      "loss": 1.2939,
+      "step": 3990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.308013916015625,
+      "learning_rate": 9.214048901782014e-05,
+      "loss": 1.2728,
+      "step": 4000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2011935710906982,
+      "learning_rate": 9.211976792374638e-05,
+      "loss": 1.2853,
+      "step": 4010
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0514850616455078,
+      "learning_rate": 9.209904682967261e-05,
+      "loss": 1.3102,
+      "step": 4020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.0865683555603027,
+      "learning_rate": 9.207832573559885e-05,
+      "loss": 1.2835,
+      "step": 4030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2012451887130737,
+      "learning_rate": 9.205760464152507e-05,
+      "loss": 1.2801,
+      "step": 4040
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0710102319717407,
+      "learning_rate": 9.203688354745132e-05,
+      "loss": 1.2745,
+      "step": 4050
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0107301473617554,
+      "learning_rate": 9.201616245337754e-05,
+      "loss": 1.2928,
+      "step": 4060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.0026335716247559,
+      "learning_rate": 9.199544135930378e-05,
+      "loss": 1.2916,
+      "step": 4070
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9443692564964294,
+      "learning_rate": 9.197472026523001e-05,
+      "loss": 1.2956,
+      "step": 4080
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9472268223762512,
+      "learning_rate": 9.195399917115625e-05,
+      "loss": 1.2875,
+      "step": 4090
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0817506313323975,
+      "learning_rate": 9.193327807708247e-05,
+      "loss": 1.2779,
+      "step": 4100
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0539813041687012,
+      "learning_rate": 9.19125569830087e-05,
+      "loss": 1.2661,
+      "step": 4110
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9975004196166992,
+      "learning_rate": 9.189183588893494e-05,
+      "loss": 1.2499,
+      "step": 4120
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0313421487808228,
+      "learning_rate": 9.187111479486117e-05,
+      "loss": 1.2907,
+      "step": 4130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0273147821426392,
+      "learning_rate": 9.185039370078741e-05,
+      "loss": 1.2929,
+      "step": 4140
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9810879230499268,
+      "learning_rate": 9.182967260671363e-05,
+      "loss": 1.2974,
+      "step": 4150
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0243279933929443,
+      "learning_rate": 9.180895151263988e-05,
+      "loss": 1.2406,
+      "step": 4160
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0115349292755127,
+      "learning_rate": 9.17882304185661e-05,
+      "loss": 1.2765,
+      "step": 4170
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1206727027893066,
+      "learning_rate": 9.176750932449234e-05,
+      "loss": 1.2956,
+      "step": 4180
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.99837327003479,
+      "learning_rate": 9.174678823041857e-05,
+      "loss": 1.2614,
+      "step": 4190
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0117827653884888,
+      "learning_rate": 9.17260671363448e-05,
+      "loss": 1.2611,
+      "step": 4200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1119413375854492,
+      "learning_rate": 9.170534604227104e-05,
+      "loss": 1.3006,
+      "step": 4210
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0567435026168823,
+      "learning_rate": 9.168462494819728e-05,
+      "loss": 1.2544,
+      "step": 4220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1326267719268799,
+      "learning_rate": 9.16639038541235e-05,
+      "loss": 1.2895,
+      "step": 4230
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.0858250856399536,
+      "learning_rate": 9.164318276004975e-05,
+      "loss": 1.2672,
+      "step": 4240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1111207008361816,
+      "learning_rate": 9.162246166597597e-05,
+      "loss": 1.255,
+      "step": 4250
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0791577100753784,
+      "learning_rate": 9.16017405719022e-05,
+      "loss": 1.2865,
+      "step": 4260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1741734743118286,
+      "learning_rate": 9.158101947782844e-05,
+      "loss": 1.2898,
+      "step": 4270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9900088310241699,
+      "learning_rate": 9.156029838375466e-05,
+      "loss": 1.26,
+      "step": 4280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2227307558059692,
+      "learning_rate": 9.15395772896809e-05,
+      "loss": 1.2818,
+      "step": 4290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0696605443954468,
+      "learning_rate": 9.151885619560713e-05,
+      "loss": 1.2739,
+      "step": 4300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.145063042640686,
+      "learning_rate": 9.149813510153336e-05,
+      "loss": 1.2886,
+      "step": 4310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0415133237838745,
+      "learning_rate": 9.14774140074596e-05,
+      "loss": 1.2999,
+      "step": 4320
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9151254296302795,
+      "learning_rate": 9.145669291338584e-05,
+      "loss": 1.2883,
+      "step": 4330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1220918893814087,
+      "learning_rate": 9.143597181931206e-05,
+      "loss": 1.2751,
+      "step": 4340
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0417348146438599,
+      "learning_rate": 9.14152507252383e-05,
+      "loss": 1.2675,
+      "step": 4350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1090322732925415,
+      "learning_rate": 9.139452963116453e-05,
+      "loss": 1.2521,
+      "step": 4360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.003110408782959,
+      "learning_rate": 9.137380853709076e-05,
+      "loss": 1.2461,
+      "step": 4370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0214240550994873,
+      "learning_rate": 9.1353087443017e-05,
+      "loss": 1.3003,
+      "step": 4380
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0389635562896729,
+      "learning_rate": 9.133236634894323e-05,
+      "loss": 1.2642,
+      "step": 4390
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1644842624664307,
+      "learning_rate": 9.131164525486945e-05,
+      "loss": 1.247,
+      "step": 4400
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0494420528411865,
+      "learning_rate": 9.12909241607957e-05,
+      "loss": 1.2727,
+      "step": 4410
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.1759871244430542,
+      "learning_rate": 9.127020306672192e-05,
+      "loss": 1.2569,
+      "step": 4420
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0006252527236938,
+      "learning_rate": 9.124948197264816e-05,
+      "loss": 1.2761,
+      "step": 4430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0942422151565552,
+      "learning_rate": 9.12287608785744e-05,
+      "loss": 1.2778,
+      "step": 4440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1273877620697021,
+      "learning_rate": 9.120803978450062e-05,
+      "loss": 1.2738,
+      "step": 4450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0103455781936646,
+      "learning_rate": 9.118731869042686e-05,
+      "loss": 1.2559,
+      "step": 4460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0412319898605347,
+      "learning_rate": 9.116659759635309e-05,
+      "loss": 1.2839,
+      "step": 4470
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1623831987380981,
+      "learning_rate": 9.114587650227932e-05,
+      "loss": 1.2684,
+      "step": 4480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.213977336883545,
+      "learning_rate": 9.112515540820556e-05,
+      "loss": 1.2587,
+      "step": 4490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1630234718322754,
+      "learning_rate": 9.110443431413179e-05,
+      "loss": 1.2557,
+      "step": 4500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0047425031661987,
+      "learning_rate": 9.108371322005801e-05,
+      "loss": 1.2785,
+      "step": 4510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0434467792510986,
+      "learning_rate": 9.106299212598426e-05,
+      "loss": 1.2916,
+      "step": 4520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0278736352920532,
+      "learning_rate": 9.104227103191048e-05,
+      "loss": 1.2494,
+      "step": 4530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9865803122520447,
+      "learning_rate": 9.102154993783672e-05,
+      "loss": 1.2487,
+      "step": 4540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0419522523880005,
+      "learning_rate": 9.100082884376295e-05,
+      "loss": 1.2618,
+      "step": 4550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1537505388259888,
+      "learning_rate": 9.098010774968919e-05,
+      "loss": 1.2613,
+      "step": 4560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0565922260284424,
+      "learning_rate": 9.095938665561542e-05,
+      "loss": 1.2592,
+      "step": 4570
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.106313705444336,
+      "learning_rate": 9.093866556154166e-05,
+      "loss": 1.2585,
+      "step": 4580
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0392132997512817,
+      "learning_rate": 9.091794446746788e-05,
+      "loss": 1.2626,
+      "step": 4590
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1017506122589111,
+      "learning_rate": 9.089722337339413e-05,
+      "loss": 1.2696,
+      "step": 4600
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.134838581085205,
+      "learning_rate": 9.087650227932035e-05,
+      "loss": 1.2626,
+      "step": 4610
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3448967933654785,
+      "learning_rate": 9.085578118524659e-05,
+      "loss": 1.2728,
+      "step": 4620
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9399920701980591,
+      "learning_rate": 9.083506009117282e-05,
+      "loss": 1.253,
+      "step": 4630
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0530354976654053,
+      "learning_rate": 9.081433899709904e-05,
+      "loss": 1.2501,
+      "step": 4640
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0609047412872314,
+      "learning_rate": 9.079361790302529e-05,
+      "loss": 1.2599,
+      "step": 4650
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.120044469833374,
+      "learning_rate": 9.077289680895151e-05,
+      "loss": 1.2511,
+      "step": 4660
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.068034291267395,
+      "learning_rate": 9.075217571487775e-05,
+      "loss": 1.2623,
+      "step": 4670
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9914749264717102,
+      "learning_rate": 9.073145462080398e-05,
+      "loss": 1.2754,
+      "step": 4680
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.084227204322815,
+      "learning_rate": 9.071073352673022e-05,
+      "loss": 1.2861,
+      "step": 4690
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9811500906944275,
+      "learning_rate": 9.069001243265644e-05,
+      "loss": 1.2361,
+      "step": 4700
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0811760425567627,
+      "learning_rate": 9.066929133858269e-05,
+      "loss": 1.247,
+      "step": 4710
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0787228345870972,
+      "learning_rate": 9.064857024450891e-05,
+      "loss": 1.2609,
+      "step": 4720
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0164875984191895,
+      "learning_rate": 9.062784915043515e-05,
+      "loss": 1.289,
+      "step": 4730
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1449209451675415,
+      "learning_rate": 9.060712805636138e-05,
+      "loss": 1.269,
+      "step": 4740
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2805284261703491,
+      "learning_rate": 9.058640696228762e-05,
+      "loss": 1.2803,
+      "step": 4750
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.03864586353302,
+      "learning_rate": 9.056568586821385e-05,
+      "loss": 1.2635,
+      "step": 4760
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9715671539306641,
+      "learning_rate": 9.054496477414009e-05,
+      "loss": 1.2232,
+      "step": 4770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1221860647201538,
+      "learning_rate": 9.052424368006631e-05,
+      "loss": 1.2536,
+      "step": 4780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9467464089393616,
+      "learning_rate": 9.050352258599254e-05,
+      "loss": 1.2529,
+      "step": 4790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1306428909301758,
+      "learning_rate": 9.048280149191878e-05,
+      "loss": 1.2621,
+      "step": 4800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.038275122642517,
+      "learning_rate": 9.0462080397845e-05,
+      "loss": 1.2785,
+      "step": 4810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1947646141052246,
+      "learning_rate": 9.044135930377125e-05,
+      "loss": 1.2643,
+      "step": 4820
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0585174560546875,
+      "learning_rate": 9.042063820969747e-05,
+      "loss": 1.248,
+      "step": 4830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1931108236312866,
+      "learning_rate": 9.03999171156237e-05,
+      "loss": 1.2884,
+      "step": 4840
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.280229926109314,
+      "eval_runtime": 1606.712,
+      "eval_samples_per_second": 262.53,
+      "eval_steps_per_second": 4.102,
+      "step": 4846
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0185681581497192,
+      "learning_rate": 9.037919602154994e-05,
+      "loss": 1.2701,
+      "step": 4850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2571942806243896,
+      "learning_rate": 9.035847492747618e-05,
+      "loss": 1.2309,
+      "step": 4860
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1584192514419556,
+      "learning_rate": 9.033775383340241e-05,
+      "loss": 1.2859,
+      "step": 4870
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.05573570728302,
+      "learning_rate": 9.031703273932865e-05,
+      "loss": 1.2502,
+      "step": 4880
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.0500950813293457,
+      "learning_rate": 9.029631164525487e-05,
+      "loss": 1.2288,
+      "step": 4890
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.0603646039962769,
+      "learning_rate": 9.027559055118112e-05,
+      "loss": 1.2609,
+      "step": 4900
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.085644245147705,
+      "learning_rate": 9.025486945710734e-05,
+      "loss": 1.2392,
+      "step": 4910
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.0088363885879517,
+      "learning_rate": 9.023414836303357e-05,
+      "loss": 1.2361,
+      "step": 4920
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.0100585222244263,
+      "learning_rate": 9.021342726895981e-05,
+      "loss": 1.2382,
+      "step": 4930
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.067159652709961,
+      "learning_rate": 9.019270617488604e-05,
+      "loss": 1.2433,
+      "step": 4940
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.0431448221206665,
+      "learning_rate": 9.017198508081226e-05,
+      "loss": 1.2461,
+      "step": 4950
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.3945410251617432,
+      "learning_rate": 9.01512639867385e-05,
+      "loss": 1.2406,
+      "step": 4960
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.0631392002105713,
+      "learning_rate": 9.013054289266474e-05,
+      "loss": 1.2265,
+      "step": 4970
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.0823907852172852,
+      "learning_rate": 9.010982179859097e-05,
+      "loss": 1.2388,
+      "step": 4980
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.1125047206878662,
+      "learning_rate": 9.00891007045172e-05,
+      "loss": 1.2553,
+      "step": 4990
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.0436159372329712,
+      "learning_rate": 9.006837961044343e-05,
+      "loss": 1.2362,
+      "step": 5000
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.2972134351730347,
+      "learning_rate": 9.004765851636968e-05,
+      "loss": 1.2387,
+      "step": 5010
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.0587375164031982,
+      "learning_rate": 9.00269374222959e-05,
+      "loss": 1.2348,
+      "step": 5020
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.0204670429229736,
+      "learning_rate": 9.000621632822213e-05,
+      "loss": 1.2391,
+      "step": 5030
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.0182541608810425,
+      "learning_rate": 8.998549523414837e-05,
+      "loss": 1.2337,
+      "step": 5040
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9534347057342529,
+      "learning_rate": 8.99647741400746e-05,
+      "loss": 1.2403,
+      "step": 5050
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.1534489393234253,
+      "learning_rate": 8.994405304600084e-05,
+      "loss": 1.251,
+      "step": 5060
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.0630913972854614,
+      "learning_rate": 8.992333195192707e-05,
+      "loss": 1.2741,
+      "step": 5070
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.0464857816696167,
+      "learning_rate": 8.99026108578533e-05,
+      "loss": 1.27,
+      "step": 5080
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.1541072130203247,
+      "learning_rate": 8.988188976377954e-05,
+      "loss": 1.2183,
+      "step": 5090
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.007450819015503,
+      "learning_rate": 8.986116866970576e-05,
+      "loss": 1.2345,
+      "step": 5100
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.984767496585846,
+      "learning_rate": 8.9840447575632e-05,
+      "loss": 1.2661,
+      "step": 5110
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.2972489595413208,
+      "learning_rate": 8.981972648155824e-05,
+      "loss": 1.2577,
+      "step": 5120
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.1882805824279785,
+      "learning_rate": 8.979900538748446e-05,
+      "loss": 1.2685,
+      "step": 5130
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.1580913066864014,
+      "learning_rate": 8.977828429341069e-05,
+      "loss": 1.2269,
+      "step": 5140
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.0898735523223877,
+      "learning_rate": 8.975756319933693e-05,
+      "loss": 1.2315,
+      "step": 5150
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.1261813640594482,
+      "learning_rate": 8.973684210526316e-05,
+      "loss": 1.2617,
+      "step": 5160
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0266163349151611,
+      "learning_rate": 8.97161210111894e-05,
+      "loss": 1.2379,
+      "step": 5170
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0973056554794312,
+      "learning_rate": 8.969539991711563e-05,
+      "loss": 1.2527,
+      "step": 5180
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0927435159683228,
+      "learning_rate": 8.967467882304185e-05,
+      "loss": 1.2433,
+      "step": 5190
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.1209070682525635,
+      "learning_rate": 8.96539577289681e-05,
+      "loss": 1.2512,
+      "step": 5200
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.222163200378418,
+      "learning_rate": 8.963323663489432e-05,
+      "loss": 1.2377,
+      "step": 5210
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.1234538555145264,
+      "learning_rate": 8.961251554082056e-05,
+      "loss": 1.2409,
+      "step": 5220
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.1121318340301514,
+      "learning_rate": 8.95917944467468e-05,
+      "loss": 1.2402,
+      "step": 5230
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.0124129056930542,
+      "learning_rate": 8.957107335267303e-05,
+      "loss": 1.2609,
+      "step": 5240
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.0647163391113281,
+      "learning_rate": 8.955035225859925e-05,
+      "loss": 1.2144,
+      "step": 5250
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.0653977394104004,
+      "learning_rate": 8.95296311645255e-05,
+      "loss": 1.2337,
+      "step": 5260
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.0555377006530762,
+      "learning_rate": 8.950891007045172e-05,
+      "loss": 1.2121,
+      "step": 5270
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.0859911441802979,
+      "learning_rate": 8.948818897637796e-05,
+      "loss": 1.232,
+      "step": 5280
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.0167226791381836,
+      "learning_rate": 8.946746788230419e-05,
+      "loss": 1.2252,
+      "step": 5290
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.0780484676361084,
+      "learning_rate": 8.944674678823041e-05,
+      "loss": 1.2537,
+      "step": 5300
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.0674623250961304,
+      "learning_rate": 8.942602569415666e-05,
+      "loss": 1.2511,
+      "step": 5310
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.1867153644561768,
+      "learning_rate": 8.940530460008288e-05,
+      "loss": 1.2406,
+      "step": 5320
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.0617753267288208,
+      "learning_rate": 8.938458350600912e-05,
+      "loss": 1.2648,
+      "step": 5330
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.1268622875213623,
+      "learning_rate": 8.936386241193535e-05,
+      "loss": 1.2461,
+      "step": 5340
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.06646728515625,
+      "learning_rate": 8.934314131786159e-05,
+      "loss": 1.235,
+      "step": 5350
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.0061641931533813,
+      "learning_rate": 8.932242022378781e-05,
+      "loss": 1.2256,
+      "step": 5360
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.1114619970321655,
+      "learning_rate": 8.930169912971406e-05,
+      "loss": 1.2481,
+      "step": 5370
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.121626615524292,
+      "learning_rate": 8.928097803564028e-05,
+      "loss": 1.2592,
+      "step": 5380
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.0986332893371582,
+      "learning_rate": 8.926025694156652e-05,
+      "loss": 1.2397,
+      "step": 5390
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.055759072303772,
+      "learning_rate": 8.923953584749275e-05,
+      "loss": 1.2468,
+      "step": 5400
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.035019874572754,
+      "learning_rate": 8.921881475341899e-05,
+      "loss": 1.2455,
+      "step": 5410
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.0355908870697021,
+      "learning_rate": 8.919809365934522e-05,
+      "loss": 1.2584,
+      "step": 5420
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.0616044998168945,
+      "learning_rate": 8.917737256527146e-05,
+      "loss": 1.2472,
+      "step": 5430
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.0461037158966064,
+      "learning_rate": 8.915665147119768e-05,
+      "loss": 1.2531,
+      "step": 5440
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.2218214273452759,
+      "learning_rate": 8.913593037712393e-05,
+      "loss": 1.2486,
+      "step": 5450
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0007187128067017,
+      "learning_rate": 8.911520928305015e-05,
+      "loss": 1.2464,
+      "step": 5460
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.056276798248291,
+      "learning_rate": 8.909448818897638e-05,
+      "loss": 1.2375,
+      "step": 5470
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0595409870147705,
+      "learning_rate": 8.907376709490262e-05,
+      "loss": 1.2645,
+      "step": 5480
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.1598786115646362,
+      "learning_rate": 8.905304600082884e-05,
+      "loss": 1.2502,
+      "step": 5490
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.0664838552474976,
+      "learning_rate": 8.903232490675508e-05,
+      "loss": 1.2712,
+      "step": 5500
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0016567707061768,
+      "learning_rate": 8.901160381268131e-05,
+      "loss": 1.2386,
+      "step": 5510
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0359711647033691,
+      "learning_rate": 8.899088271860755e-05,
+      "loss": 1.238,
+      "step": 5520
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.1633154153823853,
+      "learning_rate": 8.897016162453378e-05,
+      "loss": 1.2139,
+      "step": 5530
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.085390567779541,
+      "learning_rate": 8.894944053046002e-05,
+      "loss": 1.2564,
+      "step": 5540
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.041222333908081,
+      "learning_rate": 8.892871943638624e-05,
+      "loss": 1.2361,
+      "step": 5550
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.0352637767791748,
+      "learning_rate": 8.890799834231249e-05,
+      "loss": 1.2104,
+      "step": 5560
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.0879154205322266,
+      "learning_rate": 8.888727724823871e-05,
+      "loss": 1.2408,
+      "step": 5570
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.177937388420105,
+      "learning_rate": 8.886655615416494e-05,
+      "loss": 1.221,
+      "step": 5580
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.0638147592544556,
+      "learning_rate": 8.884583506009118e-05,
+      "loss": 1.2309,
+      "step": 5590
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.0461276769638062,
+      "learning_rate": 8.882511396601741e-05,
+      "loss": 1.2498,
+      "step": 5600
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9356592297554016,
+      "learning_rate": 8.880439287194365e-05,
+      "loss": 1.2584,
+      "step": 5610
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9423561096191406,
+      "learning_rate": 8.878367177786988e-05,
+      "loss": 1.2094,
+      "step": 5620
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.1970158815383911,
+      "learning_rate": 8.87629506837961e-05,
+      "loss": 1.2134,
+      "step": 5630
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.1081819534301758,
+      "learning_rate": 8.874222958972234e-05,
+      "loss": 1.2134,
+      "step": 5640
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.1248120069503784,
+      "learning_rate": 8.872150849564858e-05,
+      "loss": 1.2288,
+      "step": 5650
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0750600099563599,
+      "learning_rate": 8.87007874015748e-05,
+      "loss": 1.2191,
+      "step": 5660
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.058366060256958,
+      "learning_rate": 8.868006630750105e-05,
+      "loss": 1.2274,
+      "step": 5670
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9436173439025879,
+      "learning_rate": 8.865934521342727e-05,
+      "loss": 1.2556,
+      "step": 5680
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0624874830245972,
+      "learning_rate": 8.86386241193535e-05,
+      "loss": 1.2383,
+      "step": 5690
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.0870168209075928,
+      "learning_rate": 8.861790302527974e-05,
+      "loss": 1.2432,
+      "step": 5700
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.0561186075210571,
+      "learning_rate": 8.859718193120597e-05,
+      "loss": 1.2329,
+      "step": 5710
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.2139157056808472,
+      "learning_rate": 8.857646083713221e-05,
+      "loss": 1.2183,
+      "step": 5720
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.0662713050842285,
+      "learning_rate": 8.855573974305844e-05,
+      "loss": 1.2371,
+      "step": 5730
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.1198335886001587,
+      "learning_rate": 8.853501864898466e-05,
+      "loss": 1.2396,
+      "step": 5740
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.2191115617752075,
+      "learning_rate": 8.851429755491091e-05,
+      "loss": 1.2355,
+      "step": 5750
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.039201259613037,
+      "learning_rate": 8.849357646083714e-05,
+      "loss": 1.2251,
+      "step": 5760
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.9540196061134338,
+      "learning_rate": 8.847285536676337e-05,
+      "loss": 1.2424,
+      "step": 5770
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.0640240907669067,
+      "learning_rate": 8.84521342726896e-05,
+      "loss": 1.2414,
+      "step": 5780
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.0465424060821533,
+      "learning_rate": 8.843141317861584e-05,
+      "loss": 1.2241,
+      "step": 5790
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.1881465911865234,
+      "learning_rate": 8.841069208454206e-05,
+      "loss": 1.2182,
+      "step": 5800
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0851510763168335,
+      "learning_rate": 8.83899709904683e-05,
+      "loss": 1.2115,
+      "step": 5810
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0743972063064575,
+      "learning_rate": 8.836924989639453e-05,
+      "loss": 1.2211,
+      "step": 5820
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.049249529838562,
+      "learning_rate": 8.834852880232077e-05,
+      "loss": 1.2452,
+      "step": 5830
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.0910190343856812,
+      "learning_rate": 8.8327807708247e-05,
+      "loss": 1.249,
+      "step": 5840
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.0976841449737549,
+      "learning_rate": 8.830708661417322e-05,
+      "loss": 1.2123,
+      "step": 5850
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.3569914102554321,
+      "learning_rate": 8.828636552009947e-05,
+      "loss": 1.2338,
+      "step": 5860
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.068427324295044,
+      "learning_rate": 8.82656444260257e-05,
+      "loss": 1.2351,
+      "step": 5870
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.1309690475463867,
+      "learning_rate": 8.824492333195193e-05,
+      "loss": 1.2367,
+      "step": 5880
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.1933242082595825,
+      "learning_rate": 8.822420223787816e-05,
+      "loss": 1.2261,
+      "step": 5890
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.029557466506958,
+      "learning_rate": 8.82034811438044e-05,
+      "loss": 1.2499,
+      "step": 5900
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.106086015701294,
+      "learning_rate": 8.818276004973062e-05,
+      "loss": 1.2236,
+      "step": 5910
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.0114145278930664,
+      "learning_rate": 8.816203895565687e-05,
+      "loss": 1.2445,
+      "step": 5920
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.1829569339752197,
+      "learning_rate": 8.814131786158309e-05,
+      "loss": 1.2334,
+      "step": 5930
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.1952738761901855,
+      "learning_rate": 8.812059676750933e-05,
+      "loss": 1.2402,
+      "step": 5940
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0816442966461182,
+      "learning_rate": 8.809987567343556e-05,
+      "loss": 1.252,
+      "step": 5950
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.1453193426132202,
+      "learning_rate": 8.80791545793618e-05,
+      "loss": 1.233,
+      "step": 5960
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.045602798461914,
+      "learning_rate": 8.805843348528803e-05,
+      "loss": 1.2195,
+      "step": 5970
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0934393405914307,
+      "learning_rate": 8.803771239121425e-05,
+      "loss": 1.2177,
+      "step": 5980
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.0120600461959839,
+      "learning_rate": 8.801699129714049e-05,
+      "loss": 1.2329,
+      "step": 5990
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.103251576423645,
+      "learning_rate": 8.799627020306672e-05,
+      "loss": 1.2379,
+      "step": 6000
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.0402165651321411,
+      "learning_rate": 8.797554910899296e-05,
+      "loss": 1.2256,
+      "step": 6010
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.0983202457427979,
+      "learning_rate": 8.79548280149192e-05,
+      "loss": 1.2296,
+      "step": 6020
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.0412105321884155,
+      "learning_rate": 8.793410692084543e-05,
+      "loss": 1.2228,
+      "step": 6030
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.0473746061325073,
+      "learning_rate": 8.791338582677165e-05,
+      "loss": 1.2228,
+      "step": 6040
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.004840612411499,
+      "learning_rate": 8.78926647326979e-05,
+      "loss": 1.2108,
+      "step": 6050
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.1168406009674072,
+      "learning_rate": 8.787194363862412e-05,
+      "loss": 1.2104,
+      "step": 6060
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.1124876737594604,
+      "learning_rate": 8.785122254455036e-05,
+      "loss": 1.2336,
+      "step": 6070
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.059415340423584,
+      "learning_rate": 8.783050145047659e-05,
+      "loss": 1.2438,
+      "step": 6080
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.063344955444336,
+      "learning_rate": 8.780978035640283e-05,
+      "loss": 1.2352,
+      "step": 6090
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.0552620887756348,
+      "learning_rate": 8.778905926232905e-05,
+      "loss": 1.2131,
+      "step": 6100
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.1673498153686523,
+      "learning_rate": 8.77683381682553e-05,
+      "loss": 1.213,
+      "step": 6110
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.1740206480026245,
+      "learning_rate": 8.774761707418152e-05,
+      "loss": 1.2162,
+      "step": 6120
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.0944995880126953,
+      "learning_rate": 8.772689598010775e-05,
+      "loss": 1.2005,
+      "step": 6130
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.0152305364608765,
+      "learning_rate": 8.770617488603399e-05,
+      "loss": 1.2198,
+      "step": 6140
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.1265654563903809,
+      "learning_rate": 8.768545379196021e-05,
+      "loss": 1.2362,
+      "step": 6150
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.151207447052002,
+      "learning_rate": 8.766473269788646e-05,
+      "loss": 1.2561,
+      "step": 6160
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.035855770111084,
+      "learning_rate": 8.764401160381268e-05,
+      "loss": 1.2391,
+      "step": 6170
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.091009497642517,
+      "learning_rate": 8.762329050973892e-05,
+      "loss": 1.2342,
+      "step": 6180
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.057400107383728,
+      "learning_rate": 8.760256941566515e-05,
+      "loss": 1.2216,
+      "step": 6190
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.9303261041641235,
+      "learning_rate": 8.758184832159139e-05,
+      "loss": 1.2352,
+      "step": 6200
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.2061206102371216,
+      "learning_rate": 8.756112722751761e-05,
+      "loss": 1.215,
+      "step": 6210
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.0886818170547485,
+      "learning_rate": 8.754040613344386e-05,
+      "loss": 1.2266,
+      "step": 6220
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.279563546180725,
+      "learning_rate": 8.751968503937008e-05,
+      "loss": 1.237,
+      "step": 6230
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.0063716173171997,
+      "learning_rate": 8.749896394529631e-05,
+      "loss": 1.2119,
+      "step": 6240
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.0414812564849854,
+      "learning_rate": 8.747824285122255e-05,
+      "loss": 1.2177,
+      "step": 6250
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.0915932655334473,
+      "learning_rate": 8.745752175714878e-05,
+      "loss": 1.2422,
+      "step": 6260
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.0544465780258179,
+      "learning_rate": 8.743680066307502e-05,
+      "loss": 1.234,
+      "step": 6270
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.1502125263214111,
+      "learning_rate": 8.741607956900125e-05,
+      "loss": 1.2183,
+      "step": 6280
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.3113386631011963,
+      "learning_rate": 8.739535847492748e-05,
+      "loss": 1.202,
+      "step": 6290
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.087583303451538,
+      "learning_rate": 8.737463738085371e-05,
+      "loss": 1.241,
+      "step": 6300
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.135565161705017,
+      "learning_rate": 8.735391628677995e-05,
+      "loss": 1.2411,
+      "step": 6310
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.0986896753311157,
+      "learning_rate": 8.733319519270617e-05,
+      "loss": 1.2178,
+      "step": 6320
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6357513666152954,
+      "learning_rate": 8.731247409863242e-05,
+      "loss": 1.2102,
+      "step": 6330
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.0731899738311768,
+      "learning_rate": 8.729175300455864e-05,
+      "loss": 1.213,
+      "step": 6340
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.1432324647903442,
+      "learning_rate": 8.727103191048487e-05,
+      "loss": 1.2395,
+      "step": 6350
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.0547071695327759,
+      "learning_rate": 8.725031081641111e-05,
+      "loss": 1.2121,
+      "step": 6360
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.1022225618362427,
+      "learning_rate": 8.722958972233734e-05,
+      "loss": 1.2274,
+      "step": 6370
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0772980451583862,
+      "learning_rate": 8.720886862826358e-05,
+      "loss": 1.2244,
+      "step": 6380
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.073470115661621,
+      "learning_rate": 8.718814753418981e-05,
+      "loss": 1.2243,
+      "step": 6390
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.1750410795211792,
+      "learning_rate": 8.716742644011604e-05,
+      "loss": 1.2191,
+      "step": 6400
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.206298828125,
+      "learning_rate": 8.714670534604228e-05,
+      "loss": 1.2284,
+      "step": 6410
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.1253398656845093,
+      "learning_rate": 8.71259842519685e-05,
+      "loss": 1.2222,
+      "step": 6420
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.1828629970550537,
+      "learning_rate": 8.710526315789474e-05,
+      "loss": 1.225,
+      "step": 6430
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.0462387800216675,
+      "learning_rate": 8.708454206382098e-05,
+      "loss": 1.2156,
+      "step": 6440
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.3468637466430664,
+      "learning_rate": 8.706382096974721e-05,
+      "loss": 1.2111,
+      "step": 6450
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.1359398365020752,
+      "learning_rate": 8.704309987567345e-05,
+      "loss": 1.2022,
+      "step": 6460
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.0748237371444702,
+      "learning_rate": 8.702237878159967e-05,
+      "loss": 1.2227,
+      "step": 6470
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.0240176916122437,
+      "learning_rate": 8.70016576875259e-05,
+      "loss": 1.2042,
+      "step": 6480
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.0575100183486938,
+      "learning_rate": 8.698093659345214e-05,
+      "loss": 1.2166,
+      "step": 6490
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.0095828771591187,
+      "learning_rate": 8.696021549937837e-05,
+      "loss": 1.2352,
+      "step": 6500
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.140643835067749,
+      "learning_rate": 8.69394944053046e-05,
+      "loss": 1.2143,
+      "step": 6510
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.0403310060501099,
+      "learning_rate": 8.691877331123084e-05,
+      "loss": 1.2268,
+      "step": 6520
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.1176542043685913,
+      "learning_rate": 8.689805221715706e-05,
+      "loss": 1.2123,
+      "step": 6530
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.041100025177002,
+      "learning_rate": 8.68773311230833e-05,
+      "loss": 1.2356,
+      "step": 6540
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9852685928344727,
+      "learning_rate": 8.685661002900954e-05,
+      "loss": 1.2043,
+      "step": 6550
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.206588864326477,
+      "learning_rate": 8.683588893493577e-05,
+      "loss": 1.2132,
+      "step": 6560
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.1841477155685425,
+      "learning_rate": 8.6815167840862e-05,
+      "loss": 1.241,
+      "step": 6570
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.1257410049438477,
+      "learning_rate": 8.679444674678824e-05,
+      "loss": 1.2241,
+      "step": 6580
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.053404688835144,
+      "learning_rate": 8.677372565271446e-05,
+      "loss": 1.2409,
+      "step": 6590
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.109271764755249,
+      "learning_rate": 8.675300455864071e-05,
+      "loss": 1.2036,
+      "step": 6600
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.0103741884231567,
+      "learning_rate": 8.673228346456693e-05,
+      "loss": 1.2128,
+      "step": 6610
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.149715542793274,
+      "learning_rate": 8.671156237049317e-05,
+      "loss": 1.208,
+      "step": 6620
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.0523351430892944,
+      "learning_rate": 8.66908412764194e-05,
+      "loss": 1.2146,
+      "step": 6630
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.2250432968139648,
+      "learning_rate": 8.667012018234562e-05,
+      "loss": 1.2454,
+      "step": 6640
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9975298643112183,
+      "learning_rate": 8.664939908827186e-05,
+      "loss": 1.228,
+      "step": 6650
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.0323069095611572,
+      "learning_rate": 8.66286779941981e-05,
+      "loss": 1.2223,
+      "step": 6660
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.0941072702407837,
+      "learning_rate": 8.660795690012433e-05,
+      "loss": 1.2214,
+      "step": 6670
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.2442706823349,
+      "learning_rate": 8.658723580605056e-05,
+      "loss": 1.2316,
+      "step": 6680
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.1723049879074097,
+      "learning_rate": 8.65665147119768e-05,
+      "loss": 1.2184,
+      "step": 6690
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.1901644468307495,
+      "learning_rate": 8.654579361790302e-05,
+      "loss": 1.2382,
+      "step": 6700
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.1003718376159668,
+      "learning_rate": 8.652507252382927e-05,
+      "loss": 1.2272,
+      "step": 6710
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.1174200773239136,
+      "learning_rate": 8.650435142975549e-05,
+      "loss": 1.2248,
+      "step": 6720
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.2489265203475952,
+      "learning_rate": 8.648363033568173e-05,
+      "loss": 1.2252,
+      "step": 6730
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.2472732067108154,
+      "learning_rate": 8.646290924160796e-05,
+      "loss": 1.227,
+      "step": 6740
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.0011487007141113,
+      "learning_rate": 8.64421881475342e-05,
+      "loss": 1.2267,
+      "step": 6750
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.060117244720459,
+      "learning_rate": 8.642146705346042e-05,
+      "loss": 1.2263,
+      "step": 6760
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.102454423904419,
+      "learning_rate": 8.640074595938667e-05,
+      "loss": 1.2205,
+      "step": 6770
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0496042966842651,
+      "learning_rate": 8.638002486531289e-05,
+      "loss": 1.2207,
+      "step": 6780
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.005552053451538,
+      "learning_rate": 8.635930377123912e-05,
+      "loss": 1.1992,
+      "step": 6790
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0828580856323242,
+      "learning_rate": 8.633858267716536e-05,
+      "loss": 1.2164,
+      "step": 6800
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.0597617626190186,
+      "learning_rate": 8.631786158309158e-05,
+      "loss": 1.2081,
+      "step": 6810
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1591441631317139,
+      "learning_rate": 8.629714048901783e-05,
+      "loss": 1.2315,
+      "step": 6820
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1969908475875854,
+      "learning_rate": 8.627641939494405e-05,
+      "loss": 1.2307,
+      "step": 6830
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1163939237594604,
+      "learning_rate": 8.625569830087029e-05,
+      "loss": 1.2272,
+      "step": 6840
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.047241449356079,
+      "learning_rate": 8.623497720679652e-05,
+      "loss": 1.1991,
+      "step": 6850
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.0410964488983154,
+      "learning_rate": 8.621425611272276e-05,
+      "loss": 1.211,
+      "step": 6860
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.0674628019332886,
+      "learning_rate": 8.619353501864899e-05,
+      "loss": 1.2195,
+      "step": 6870
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9892363548278809,
+      "learning_rate": 8.617281392457523e-05,
+      "loss": 1.2344,
+      "step": 6880
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.08130943775177,
+      "learning_rate": 8.615209283050145e-05,
+      "loss": 1.2299,
+      "step": 6890
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.0285786390304565,
+      "learning_rate": 8.613137173642768e-05,
+      "loss": 1.2166,
+      "step": 6900
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.0203038454055786,
+      "learning_rate": 8.611065064235392e-05,
+      "loss": 1.2043,
+      "step": 6910
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.17252516746521,
+      "learning_rate": 8.608992954828015e-05,
+      "loss": 1.2278,
+      "step": 6920
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.233001708984375,
+      "learning_rate": 8.606920845420639e-05,
+      "loss": 1.233,
+      "step": 6930
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.0858713388442993,
+      "learning_rate": 8.604848736013262e-05,
+      "loss": 1.2174,
+      "step": 6940
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.069405198097229,
+      "learning_rate": 8.602776626605885e-05,
+      "loss": 1.2176,
+      "step": 6950
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.0860111713409424,
+      "learning_rate": 8.60070451719851e-05,
+      "loss": 1.2179,
+      "step": 6960
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.111624836921692,
+      "learning_rate": 8.598632407791132e-05,
+      "loss": 1.2167,
+      "step": 6970
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.092925786972046,
+      "learning_rate": 8.596560298383755e-05,
+      "loss": 1.2154,
+      "step": 6980
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.115460753440857,
+      "learning_rate": 8.594488188976379e-05,
+      "loss": 1.2052,
+      "step": 6990
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.229777455329895,
+      "learning_rate": 8.592416079569001e-05,
+      "loss": 1.1769,
+      "step": 7000
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.1182063817977905,
+      "learning_rate": 8.590343970161626e-05,
+      "loss": 1.2035,
+      "step": 7010
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.0315207242965698,
+      "learning_rate": 8.588271860754248e-05,
+      "loss": 1.2158,
+      "step": 7020
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.1051239967346191,
+      "learning_rate": 8.586199751346871e-05,
+      "loss": 1.1874,
+      "step": 7030
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.0437036752700806,
+      "learning_rate": 8.584127641939495e-05,
+      "loss": 1.2421,
+      "step": 7040
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.1060231924057007,
+      "learning_rate": 8.582055532532118e-05,
+      "loss": 1.2392,
+      "step": 7050
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.0951759815216064,
+      "learning_rate": 8.57998342312474e-05,
+      "loss": 1.2348,
+      "step": 7060
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7351630926132202,
+      "learning_rate": 8.577911313717365e-05,
+      "loss": 1.2198,
+      "step": 7070
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.3410096168518066,
+      "learning_rate": 8.575839204309988e-05,
+      "loss": 1.2069,
+      "step": 7080
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.1723905801773071,
+      "learning_rate": 8.573767094902611e-05,
+      "loss": 1.2264,
+      "step": 7090
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.0400230884552002,
+      "learning_rate": 8.571694985495235e-05,
+      "loss": 1.2285,
+      "step": 7100
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.1555556058883667,
+      "learning_rate": 8.569622876087858e-05,
+      "loss": 1.1982,
+      "step": 7110
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.0779943466186523,
+      "learning_rate": 8.567550766680482e-05,
+      "loss": 1.2241,
+      "step": 7120
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.011435627937317,
+      "learning_rate": 8.565478657273105e-05,
+      "loss": 1.1919,
+      "step": 7130
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.1413905620574951,
+      "learning_rate": 8.563406547865727e-05,
+      "loss": 1.2026,
+      "step": 7140
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.305766224861145,
+      "learning_rate": 8.561334438458351e-05,
+      "loss": 1.2243,
+      "step": 7150
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.1241309642791748,
+      "learning_rate": 8.559262329050974e-05,
+      "loss": 1.2305,
+      "step": 7160
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.1354045867919922,
+      "learning_rate": 8.557190219643596e-05,
+      "loss": 1.2203,
+      "step": 7170
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0294325351715088,
+      "learning_rate": 8.555118110236221e-05,
+      "loss": 1.2356,
+      "step": 7180
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.1068978309631348,
+      "learning_rate": 8.553046000828844e-05,
+      "loss": 1.2191,
+      "step": 7190
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.0707268714904785,
+      "learning_rate": 8.550973891421467e-05,
+      "loss": 1.2168,
+      "step": 7200
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.158420443534851,
+      "learning_rate": 8.54890178201409e-05,
+      "loss": 1.1924,
+      "step": 7210
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.018847107887268,
+      "learning_rate": 8.546829672606714e-05,
+      "loss": 1.2326,
+      "step": 7220
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.1226823329925537,
+      "learning_rate": 8.544757563199338e-05,
+      "loss": 1.2029,
+      "step": 7230
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.1094051599502563,
+      "learning_rate": 8.542685453791961e-05,
+      "loss": 1.2005,
+      "step": 7240
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.1425297260284424,
+      "learning_rate": 8.540613344384583e-05,
+      "loss": 1.2109,
+      "step": 7250
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.0846836566925049,
+      "learning_rate": 8.538541234977208e-05,
+      "loss": 1.2098,
+      "step": 7260
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.0861196517944336,
+      "learning_rate": 8.53646912556983e-05,
+      "loss": 1.2201,
+      "step": 7270
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.0006380081176758,
+      "learning_rate": 8.534397016162454e-05,
+      "loss": 1.2457,
+      "step": 7280
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.0799134969711304,
+      "learning_rate": 8.532324906755077e-05,
+      "loss": 1.2117,
+      "step": 7290
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.1802451610565186,
+      "learning_rate": 8.530252797347701e-05,
+      "loss": 1.2277,
+      "step": 7300
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.0235154628753662,
+      "learning_rate": 8.528180687940323e-05,
+      "loss": 1.204,
+      "step": 7310
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.1273279190063477,
+      "learning_rate": 8.526108578532946e-05,
+      "loss": 1.2179,
+      "step": 7320
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.9829218983650208,
+      "learning_rate": 8.52403646912557e-05,
+      "loss": 1.1913,
+      "step": 7330
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.1805267333984375,
+      "learning_rate": 8.521964359718194e-05,
+      "loss": 1.1822,
+      "step": 7340
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.026557207107544,
+      "learning_rate": 8.519892250310817e-05,
+      "loss": 1.1912,
+      "step": 7350
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.04379141330719,
+      "learning_rate": 8.517820140903439e-05,
+      "loss": 1.2093,
+      "step": 7360
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.0553088188171387,
+      "learning_rate": 8.515748031496064e-05,
+      "loss": 1.2213,
+      "step": 7370
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8822417855262756,
+      "learning_rate": 8.513675922088686e-05,
+      "loss": 1.2223,
+      "step": 7380
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9738419055938721,
+      "learning_rate": 8.51160381268131e-05,
+      "loss": 1.2011,
+      "step": 7390
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.0899471044540405,
+      "learning_rate": 8.509738914214672e-05,
+      "loss": 1.2406,
+      "step": 7400
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.1138182878494263,
+      "learning_rate": 8.507666804807294e-05,
+      "loss": 1.1963,
+      "step": 7410
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.1311627626419067,
+      "learning_rate": 8.505594695399917e-05,
+      "loss": 1.1977,
+      "step": 7420
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.138723134994507,
+      "learning_rate": 8.503522585992541e-05,
+      "loss": 1.2104,
+      "step": 7430
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.0873496532440186,
+      "learning_rate": 8.501450476585164e-05,
+      "loss": 1.2113,
+      "step": 7440
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.9628106355667114,
+      "learning_rate": 8.499378367177787e-05,
+      "loss": 1.2258,
+      "step": 7450
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.1409735679626465,
+      "learning_rate": 8.497306257770411e-05,
+      "loss": 1.1964,
+      "step": 7460
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.111512541770935,
+      "learning_rate": 8.495234148363034e-05,
+      "loss": 1.2061,
+      "step": 7470
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.0905687808990479,
+      "learning_rate": 8.493162038955657e-05,
+      "loss": 1.2025,
+      "step": 7480
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.1417665481567383,
+      "learning_rate": 8.49108992954828e-05,
+      "loss": 1.1861,
+      "step": 7490
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.024594783782959,
+      "learning_rate": 8.489017820140904e-05,
+      "loss": 1.1893,
+      "step": 7500
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.112746238708496,
+      "learning_rate": 8.486945710733528e-05,
+      "loss": 1.2196,
+      "step": 7510
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.1404857635498047,
+      "learning_rate": 8.484873601326151e-05,
+      "loss": 1.2131,
+      "step": 7520
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.1401077508926392,
+      "learning_rate": 8.482801491918773e-05,
+      "loss": 1.2107,
+      "step": 7530
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.2235480546951294,
+      "learning_rate": 8.480729382511397e-05,
+      "loss": 1.2363,
+      "step": 7540
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.0837748050689697,
+      "learning_rate": 8.47865727310402e-05,
+      "loss": 1.2139,
+      "step": 7550
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.22895348072052,
+      "learning_rate": 8.476585163696643e-05,
+      "loss": 1.1998,
+      "step": 7560
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.105093240737915,
+      "learning_rate": 8.474513054289267e-05,
+      "loss": 1.2212,
+      "step": 7570
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.222209095954895,
+      "learning_rate": 8.47244094488189e-05,
+      "loss": 1.2453,
+      "step": 7580
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.1390637159347534,
+      "learning_rate": 8.470368835474513e-05,
+      "loss": 1.2071,
+      "step": 7590
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0356426239013672,
+      "learning_rate": 8.468296726067137e-05,
+      "loss": 1.2157,
+      "step": 7600
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0771561861038208,
+      "learning_rate": 8.46622461665976e-05,
+      "loss": 1.212,
+      "step": 7610
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.3607767820358276,
+      "learning_rate": 8.464152507252384e-05,
+      "loss": 1.2253,
+      "step": 7620
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0732100009918213,
+      "learning_rate": 8.462080397845007e-05,
+      "loss": 1.1938,
+      "step": 7630
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.0551025867462158,
+      "learning_rate": 8.460008288437629e-05,
+      "loss": 1.2163,
+      "step": 7640
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.127244472503662,
+      "learning_rate": 8.457936179030254e-05,
+      "loss": 1.1956,
+      "step": 7650
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.0608160495758057,
+      "learning_rate": 8.455864069622876e-05,
+      "loss": 1.191,
+      "step": 7660
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.0632834434509277,
+      "learning_rate": 8.4537919602155e-05,
+      "loss": 1.1971,
+      "step": 7670
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.1187163591384888,
+      "learning_rate": 8.451719850808123e-05,
+      "loss": 1.2156,
+      "step": 7680
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.135420799255371,
+      "learning_rate": 8.449647741400747e-05,
+      "loss": 1.1953,
+      "step": 7690
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.1564319133758545,
+      "learning_rate": 8.447575631993369e-05,
+      "loss": 1.2169,
+      "step": 7700
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.1321407556533813,
+      "learning_rate": 8.445503522585993e-05,
+      "loss": 1.1869,
+      "step": 7710
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.097676157951355,
+      "learning_rate": 8.443431413178616e-05,
+      "loss": 1.2039,
+      "step": 7720
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9873398542404175,
+      "learning_rate": 8.44135930377124e-05,
+      "loss": 1.2115,
+      "step": 7730
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.0774983167648315,
+      "learning_rate": 8.439287194363863e-05,
+      "loss": 1.2089,
+      "step": 7740
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.199475884437561,
+      "learning_rate": 8.437215084956485e-05,
+      "loss": 1.1913,
+      "step": 7750
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.2517530918121338,
+      "learning_rate": 8.43514297554911e-05,
+      "loss": 1.24,
+      "step": 7760
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.1117113828659058,
+      "learning_rate": 8.433070866141732e-05,
+      "loss": 1.1907,
+      "step": 7770
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.1518152952194214,
+      "learning_rate": 8.430998756734356e-05,
+      "loss": 1.2194,
+      "step": 7780
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.0633752346038818,
+      "learning_rate": 8.428926647326979e-05,
+      "loss": 1.2173,
+      "step": 7790
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.1065930128097534,
+      "learning_rate": 8.426854537919603e-05,
+      "loss": 1.1803,
+      "step": 7800
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.03267240524292,
+      "learning_rate": 8.424782428512226e-05,
+      "loss": 1.2178,
+      "step": 7810
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9949610233306885,
+      "learning_rate": 8.42271031910485e-05,
+      "loss": 1.1999,
+      "step": 7820
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.0859277248382568,
+      "learning_rate": 8.420638209697472e-05,
+      "loss": 1.2056,
+      "step": 7830
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.0535070896148682,
+      "learning_rate": 8.418566100290096e-05,
+      "loss": 1.2004,
+      "step": 7840
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.1210662126541138,
+      "learning_rate": 8.416493990882719e-05,
+      "loss": 1.2223,
+      "step": 7850
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.0601907968521118,
+      "learning_rate": 8.414421881475343e-05,
+      "loss": 1.1985,
+      "step": 7860
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.0329095125198364,
+      "learning_rate": 8.412349772067966e-05,
+      "loss": 1.202,
+      "step": 7870
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.1331712007522583,
+      "learning_rate": 8.410277662660588e-05,
+      "loss": 1.1923,
+      "step": 7880
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.1752616167068481,
+      "learning_rate": 8.408205553253212e-05,
+      "loss": 1.195,
+      "step": 7890
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.1065553426742554,
+      "learning_rate": 8.406133443845835e-05,
+      "loss": 1.2173,
+      "step": 7900
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.1917232275009155,
+      "learning_rate": 8.404061334438459e-05,
+      "loss": 1.1926,
+      "step": 7910
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.1937179565429688,
+      "learning_rate": 8.401989225031082e-05,
+      "loss": 1.18,
+      "step": 7920
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.1231236457824707,
+      "learning_rate": 8.399917115623706e-05,
+      "loss": 1.2006,
+      "step": 7930
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.2102551460266113,
+      "learning_rate": 8.397845006216328e-05,
+      "loss": 1.209,
+      "step": 7940
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.0514934062957764,
+      "learning_rate": 8.395772896808953e-05,
+      "loss": 1.1959,
+      "step": 7950
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.0771079063415527,
+      "learning_rate": 8.393700787401575e-05,
+      "loss": 1.2303,
+      "step": 7960
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.1473889350891113,
+      "learning_rate": 8.391628677994198e-05,
+      "loss": 1.189,
+      "step": 7970
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.101943850517273,
+      "learning_rate": 8.389556568586822e-05,
+      "loss": 1.1792,
+      "step": 7980
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.217653751373291,
+      "learning_rate": 8.387484459179446e-05,
+      "loss": 1.2032,
+      "step": 7990
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.1558287143707275,
+      "learning_rate": 8.385412349772068e-05,
+      "loss": 1.2048,
+      "step": 8000
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.1772674322128296,
+      "learning_rate": 8.383340240364693e-05,
+      "loss": 1.208,
+      "step": 8010
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.1206742525100708,
+      "learning_rate": 8.381268130957315e-05,
+      "loss": 1.2158,
+      "step": 8020
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.0629864931106567,
+      "learning_rate": 8.379196021549938e-05,
+      "loss": 1.1989,
+      "step": 8030
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.0601340532302856,
+      "learning_rate": 8.377123912142562e-05,
+      "loss": 1.2149,
+      "step": 8040
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.2142128944396973,
+      "learning_rate": 8.375051802735184e-05,
+      "loss": 1.2045,
+      "step": 8050
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.015764832496643,
+      "learning_rate": 8.372979693327809e-05,
+      "loss": 1.2177,
+      "step": 8060
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.1525570154190063,
+      "learning_rate": 8.370907583920431e-05,
+      "loss": 1.2203,
+      "step": 8070
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.0959409475326538,
+      "learning_rate": 8.368835474513054e-05,
+      "loss": 1.2022,
+      "step": 8080
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.073423981666565,
+      "learning_rate": 8.366763365105678e-05,
+      "loss": 1.2225,
+      "step": 8090
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.1058803796768188,
+      "learning_rate": 8.364691255698301e-05,
+      "loss": 1.202,
+      "step": 8100
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.067132592201233,
+      "learning_rate": 8.362619146290924e-05,
+      "loss": 1.2215,
+      "step": 8110
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.1303818225860596,
+      "learning_rate": 8.360547036883548e-05,
+      "loss": 1.2049,
+      "step": 8120
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.1500287055969238,
+      "learning_rate": 8.35847492747617e-05,
+      "loss": 1.2241,
+      "step": 8130
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9954307675361633,
+      "learning_rate": 8.356402818068794e-05,
+      "loss": 1.1806,
+      "step": 8140
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.0905861854553223,
+      "learning_rate": 8.354330708661418e-05,
+      "loss": 1.2108,
+      "step": 8150
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.1111913919448853,
+      "learning_rate": 8.352258599254041e-05,
+      "loss": 1.2026,
+      "step": 8160
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.9808979034423828,
+      "learning_rate": 8.350186489846665e-05,
+      "loss": 1.2101,
+      "step": 8170
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.127181053161621,
+      "learning_rate": 8.348114380439288e-05,
+      "loss": 1.1955,
+      "step": 8180
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.0933669805526733,
+      "learning_rate": 8.34604227103191e-05,
+      "loss": 1.1904,
+      "step": 8190
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.2010207176208496,
+      "learning_rate": 8.343970161624535e-05,
+      "loss": 1.2018,
+      "step": 8200
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.019642949104309,
+      "learning_rate": 8.341898052217157e-05,
+      "loss": 1.221,
+      "step": 8210
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.115064024925232,
+      "learning_rate": 8.339825942809781e-05,
+      "loss": 1.193,
+      "step": 8220
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.008520245552063,
+      "learning_rate": 8.337753833402404e-05,
+      "loss": 1.222,
+      "step": 8230
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.0627065896987915,
+      "learning_rate": 8.335681723995027e-05,
+      "loss": 1.1968,
+      "step": 8240
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.2253535985946655,
+      "learning_rate": 8.33360961458765e-05,
+      "loss": 1.2135,
+      "step": 8250
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.0848592519760132,
+      "learning_rate": 8.331537505180274e-05,
+      "loss": 1.2087,
+      "step": 8260
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.1441541910171509,
+      "learning_rate": 8.329465395772897e-05,
+      "loss": 1.1789,
+      "step": 8270
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.1108758449554443,
+      "learning_rate": 8.32739328636552e-05,
+      "loss": 1.1778,
+      "step": 8280
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.1133983135223389,
+      "learning_rate": 8.325321176958144e-05,
+      "loss": 1.2062,
+      "step": 8290
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.1412779092788696,
+      "learning_rate": 8.323249067550766e-05,
+      "loss": 1.2129,
+      "step": 8300
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.0338503122329712,
+      "learning_rate": 8.321176958143391e-05,
+      "loss": 1.1727,
+      "step": 8310
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.1191462278366089,
+      "learning_rate": 8.319104848736013e-05,
+      "loss": 1.1997,
+      "step": 8320
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.0909239053726196,
+      "learning_rate": 8.317032739328637e-05,
+      "loss": 1.2144,
+      "step": 8330
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.1151865720748901,
+      "learning_rate": 8.31496062992126e-05,
+      "loss": 1.2033,
+      "step": 8340
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.143604040145874,
+      "learning_rate": 8.312888520513884e-05,
+      "loss": 1.2088,
+      "step": 8350
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.2317973375320435,
+      "learning_rate": 8.310816411106507e-05,
+      "loss": 1.2179,
+      "step": 8360
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.1043517589569092,
+      "learning_rate": 8.308744301699131e-05,
+      "loss": 1.1913,
+      "step": 8370
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.149396300315857,
+      "learning_rate": 8.306672192291753e-05,
+      "loss": 1.2066,
+      "step": 8380
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.0468456745147705,
+      "learning_rate": 8.304600082884377e-05,
+      "loss": 1.1823,
+      "step": 8390
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.0730814933776855,
+      "learning_rate": 8.302527973477e-05,
+      "loss": 1.2053,
+      "step": 8400
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.2069722414016724,
+      "learning_rate": 8.300455864069622e-05,
+      "loss": 1.1999,
+      "step": 8410
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.0964728593826294,
+      "learning_rate": 8.298383754662247e-05,
+      "loss": 1.1927,
+      "step": 8420
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.0547986030578613,
+      "learning_rate": 8.296311645254869e-05,
+      "loss": 1.2198,
+      "step": 8430
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.9480800628662109,
+      "learning_rate": 8.294239535847493e-05,
+      "loss": 1.1673,
+      "step": 8440
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.042935848236084,
+      "learning_rate": 8.292167426440116e-05,
+      "loss": 1.211,
+      "step": 8450
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.1812864542007446,
+      "learning_rate": 8.29009531703274e-05,
+      "loss": 1.1775,
+      "step": 8460
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.0850603580474854,
+      "learning_rate": 8.288023207625363e-05,
+      "loss": 1.2023,
+      "step": 8470
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.0494657754898071,
+      "learning_rate": 8.285951098217987e-05,
+      "loss": 1.1897,
+      "step": 8480
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.2138115167617798,
+      "learning_rate": 8.283878988810609e-05,
+      "loss": 1.1691,
+      "step": 8490
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.0943918228149414,
+      "learning_rate": 8.281806879403234e-05,
+      "loss": 1.1972,
+      "step": 8500
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0463635921478271,
+      "learning_rate": 8.279734769995856e-05,
+      "loss": 1.2312,
+      "step": 8510
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.1013418436050415,
+      "learning_rate": 8.27766266058848e-05,
+      "loss": 1.2311,
+      "step": 8520
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9996619820594788,
+      "learning_rate": 8.275590551181103e-05,
+      "loss": 1.2129,
+      "step": 8530
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0416114330291748,
+      "learning_rate": 8.273518441773727e-05,
+      "loss": 1.1822,
+      "step": 8540
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0862529277801514,
+      "learning_rate": 8.271446332366349e-05,
+      "loss": 1.2145,
+      "step": 8550
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.1965521574020386,
+      "learning_rate": 8.269374222958972e-05,
+      "loss": 1.1969,
+      "step": 8560
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.1601965427398682,
+      "learning_rate": 8.267302113551596e-05,
+      "loss": 1.2112,
+      "step": 8570
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.0845617055892944,
+      "learning_rate": 8.265230004144219e-05,
+      "loss": 1.1885,
+      "step": 8580
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.0907461643218994,
+      "learning_rate": 8.263157894736843e-05,
+      "loss": 1.2091,
+      "step": 8590
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.027777075767517,
+      "learning_rate": 8.261085785329465e-05,
+      "loss": 1.1919,
+      "step": 8600
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2468430995941162,
+      "learning_rate": 8.25901367592209e-05,
+      "loss": 1.1948,
+      "step": 8610
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.1889891624450684,
+      "learning_rate": 8.256941566514712e-05,
+      "loss": 1.2143,
+      "step": 8620
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2794381380081177,
+      "learning_rate": 8.254869457107336e-05,
+      "loss": 1.2132,
+      "step": 8630
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.0493546724319458,
+      "learning_rate": 8.252797347699959e-05,
+      "loss": 1.1766,
+      "step": 8640
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.001658320426941,
+      "learning_rate": 8.250725238292583e-05,
+      "loss": 1.1715,
+      "step": 8650
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.06058669090271,
+      "learning_rate": 8.248653128885205e-05,
+      "loss": 1.1793,
+      "step": 8660
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.09765625,
+      "learning_rate": 8.24658101947783e-05,
+      "loss": 1.1954,
+      "step": 8670
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.0230962038040161,
+      "learning_rate": 8.244508910070452e-05,
+      "loss": 1.2158,
+      "step": 8680
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.028173804283142,
+      "learning_rate": 8.242436800663075e-05,
+      "loss": 1.1874,
+      "step": 8690
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0620988607406616,
+      "learning_rate": 8.240364691255699e-05,
+      "loss": 1.2101,
+      "step": 8700
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0968023538589478,
+      "learning_rate": 8.238292581848322e-05,
+      "loss": 1.2131,
+      "step": 8710
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.1402392387390137,
+      "learning_rate": 8.236220472440946e-05,
+      "loss": 1.2173,
+      "step": 8720
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.1263937950134277,
+      "learning_rate": 8.234148363033568e-05,
+      "loss": 1.1796,
+      "step": 8730
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.1383157968521118,
+      "learning_rate": 8.232076253626191e-05,
+      "loss": 1.1724,
+      "step": 8740
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.255072832107544,
+      "learning_rate": 8.230004144218815e-05,
+      "loss": 1.2019,
+      "step": 8750
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.4536107778549194,
+      "learning_rate": 8.227932034811438e-05,
+      "loss": 1.189,
+      "step": 8760
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.0804840326309204,
+      "learning_rate": 8.225859925404062e-05,
+      "loss": 1.2009,
+      "step": 8770
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.1545298099517822,
+      "learning_rate": 8.223787815996686e-05,
+      "loss": 1.2047,
+      "step": 8780
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.2308061122894287,
+      "learning_rate": 8.221715706589308e-05,
+      "loss": 1.1958,
+      "step": 8790
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.2053345441818237,
+      "learning_rate": 8.219643597181933e-05,
+      "loss": 1.183,
+      "step": 8800
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.2016477584838867,
+      "learning_rate": 8.217571487774555e-05,
+      "loss": 1.1831,
+      "step": 8810
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.2991312742233276,
+      "learning_rate": 8.215499378367178e-05,
+      "loss": 1.1893,
+      "step": 8820
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.0688191652297974,
+      "learning_rate": 8.213427268959802e-05,
+      "loss": 1.1962,
+      "step": 8830
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.0042345523834229,
+      "learning_rate": 8.211355159552425e-05,
+      "loss": 1.1935,
+      "step": 8840
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.1283092498779297,
+      "learning_rate": 8.209283050145047e-05,
+      "loss": 1.1983,
+      "step": 8850
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.1149276494979858,
+      "learning_rate": 8.207210940737672e-05,
+      "loss": 1.1967,
+      "step": 8860
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.0818403959274292,
+      "learning_rate": 8.205138831330294e-05,
+      "loss": 1.187,
+      "step": 8870
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.1263095140457153,
+      "learning_rate": 8.203066721922918e-05,
+      "loss": 1.1866,
+      "step": 8880
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.0879017114639282,
+      "learning_rate": 8.200994612515541e-05,
+      "loss": 1.2018,
+      "step": 8890
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.1448615789413452,
+      "learning_rate": 8.198922503108164e-05,
+      "loss": 1.1879,
+      "step": 8900
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.9900506138801575,
+      "learning_rate": 8.196850393700788e-05,
+      "loss": 1.1985,
+      "step": 8910
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.0438898801803589,
+      "learning_rate": 8.19477828429341e-05,
+      "loss": 1.1881,
+      "step": 8920
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.1896075010299683,
+      "learning_rate": 8.192706174886034e-05,
+      "loss": 1.2022,
+      "step": 8930
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.152300238609314,
+      "learning_rate": 8.190634065478658e-05,
+      "loss": 1.1828,
+      "step": 8940
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.0616978406906128,
+      "learning_rate": 8.188561956071281e-05,
+      "loss": 1.1788,
+      "step": 8950
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.0310215950012207,
+      "learning_rate": 8.186489846663903e-05,
+      "loss": 1.1898,
+      "step": 8960
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.1227622032165527,
+      "learning_rate": 8.184417737256528e-05,
+      "loss": 1.1936,
+      "step": 8970
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.0663117170333862,
+      "learning_rate": 8.18234562784915e-05,
+      "loss": 1.1857,
+      "step": 8980
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.202330470085144,
+      "learning_rate": 8.180273518441774e-05,
+      "loss": 1.1811,
+      "step": 8990
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.036596417427063,
+      "learning_rate": 8.178201409034397e-05,
+      "loss": 1.2013,
+      "step": 9000
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.0503324270248413,
+      "learning_rate": 8.176129299627021e-05,
+      "loss": 1.1755,
+      "step": 9010
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.0510581731796265,
+      "learning_rate": 8.174057190219644e-05,
+      "loss": 1.1936,
+      "step": 9020
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.0580915212631226,
+      "learning_rate": 8.171985080812268e-05,
+      "loss": 1.2006,
+      "step": 9030
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.0723899602890015,
+      "learning_rate": 8.16991297140489e-05,
+      "loss": 1.1841,
+      "step": 9040
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.0289912223815918,
+      "learning_rate": 8.167840861997515e-05,
+      "loss": 1.1762,
+      "step": 9050
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.2664040327072144,
+      "learning_rate": 8.165768752590137e-05,
+      "loss": 1.1695,
+      "step": 9060
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.0689359903335571,
+      "learning_rate": 8.163696643182759e-05,
+      "loss": 1.1888,
+      "step": 9070
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.131007194519043,
+      "learning_rate": 8.161624533775384e-05,
+      "loss": 1.1856,
+      "step": 9080
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.06068754196167,
+      "learning_rate": 8.159552424368006e-05,
+      "loss": 1.2055,
+      "step": 9090
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.0843102931976318,
+      "learning_rate": 8.15748031496063e-05,
+      "loss": 1.1898,
+      "step": 9100
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.141878366470337,
+      "learning_rate": 8.155408205553253e-05,
+      "loss": 1.1885,
+      "step": 9110
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.1082584857940674,
+      "learning_rate": 8.153336096145877e-05,
+      "loss": 1.1843,
+      "step": 9120
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.1285297870635986,
+      "learning_rate": 8.1512639867385e-05,
+      "loss": 1.2126,
+      "step": 9130
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.2701120376586914,
+      "learning_rate": 8.149191877331124e-05,
+      "loss": 1.1816,
+      "step": 9140
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.1763367652893066,
+      "learning_rate": 8.147119767923746e-05,
+      "loss": 1.2031,
+      "step": 9150
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.0942009687423706,
+      "learning_rate": 8.145047658516371e-05,
+      "loss": 1.1922,
+      "step": 9160
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.1373592615127563,
+      "learning_rate": 8.142975549108993e-05,
+      "loss": 1.1891,
+      "step": 9170
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.0547840595245361,
+      "learning_rate": 8.140903439701617e-05,
+      "loss": 1.2119,
+      "step": 9180
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.973156213760376,
+      "learning_rate": 8.13883133029424e-05,
+      "loss": 1.1929,
+      "step": 9190
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.07832670211792,
+      "learning_rate": 8.136759220886864e-05,
+      "loss": 1.1693,
+      "step": 9200
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.173492670059204,
+      "learning_rate": 8.134687111479487e-05,
+      "loss": 1.1998,
+      "step": 9210
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.161414623260498,
+      "learning_rate": 8.132615002072109e-05,
+      "loss": 1.1769,
+      "step": 9220
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.1352944374084473,
+      "learning_rate": 8.130542892664733e-05,
+      "loss": 1.1837,
+      "step": 9230
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.102401852607727,
+      "learning_rate": 8.128470783257356e-05,
+      "loss": 1.175,
+      "step": 9240
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.2963926792144775,
+      "learning_rate": 8.12639867384998e-05,
+      "loss": 1.1782,
+      "step": 9250
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.0536015033721924,
+      "learning_rate": 8.124326564442602e-05,
+      "loss": 1.1799,
+      "step": 9260
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.1193984746932983,
+      "learning_rate": 8.122254455035227e-05,
+      "loss": 1.1729,
+      "step": 9270
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.1112126111984253,
+      "learning_rate": 8.120182345627849e-05,
+      "loss": 1.168,
+      "step": 9280
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.0647683143615723,
+      "learning_rate": 8.118110236220473e-05,
+      "loss": 1.2057,
+      "step": 9290
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.1709637641906738,
+      "learning_rate": 8.116038126813096e-05,
+      "loss": 1.1757,
+      "step": 9300
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.0774716138839722,
+      "learning_rate": 8.11396601740572e-05,
+      "loss": 1.1784,
+      "step": 9310
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.088477373123169,
+      "learning_rate": 8.111893907998343e-05,
+      "loss": 1.1726,
+      "step": 9320
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.0596317052841187,
+      "learning_rate": 8.109821798590967e-05,
+      "loss": 1.2029,
+      "step": 9330
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1406421661376953,
+      "learning_rate": 8.107749689183589e-05,
+      "loss": 1.2154,
+      "step": 9340
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1345916986465454,
+      "learning_rate": 8.105677579776214e-05,
+      "loss": 1.1684,
+      "step": 9350
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.2776840925216675,
+      "learning_rate": 8.103605470368836e-05,
+      "loss": 1.1915,
+      "step": 9360
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1104062795639038,
+      "learning_rate": 8.101533360961459e-05,
+      "loss": 1.1935,
+      "step": 9370
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.0262198448181152,
+      "learning_rate": 8.099461251554083e-05,
+      "loss": 1.2047,
+      "step": 9380
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.1350001096725464,
+      "learning_rate": 8.097389142146705e-05,
+      "loss": 1.1798,
+      "step": 9390
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.0631601810455322,
+      "learning_rate": 8.095317032739328e-05,
+      "loss": 1.1744,
+      "step": 9400
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.1678036451339722,
+      "learning_rate": 8.093244923331952e-05,
+      "loss": 1.208,
+      "step": 9410
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.0665799379348755,
+      "learning_rate": 8.091172813924576e-05,
+      "loss": 1.1792,
+      "step": 9420
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.1364054679870605,
+      "learning_rate": 8.089100704517199e-05,
+      "loss": 1.2036,
+      "step": 9430
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.0586035251617432,
+      "learning_rate": 8.087028595109823e-05,
+      "loss": 1.201,
+      "step": 9440
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9820613861083984,
+      "learning_rate": 8.084956485702445e-05,
+      "loss": 1.1912,
+      "step": 9450
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.0832290649414062,
+      "learning_rate": 8.08288437629507e-05,
+      "loss": 1.2044,
+      "step": 9460
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.2247551679611206,
+      "learning_rate": 8.080812266887692e-05,
+      "loss": 1.1613,
+      "step": 9470
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.0764803886413574,
+      "learning_rate": 8.078740157480315e-05,
+      "loss": 1.1885,
+      "step": 9480
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1871037483215332,
+      "learning_rate": 8.076668048072939e-05,
+      "loss": 1.2089,
+      "step": 9490
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.0709806680679321,
+      "learning_rate": 8.074595938665562e-05,
+      "loss": 1.1743,
+      "step": 9500
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.2077934741973877,
+      "learning_rate": 8.072523829258184e-05,
+      "loss": 1.192,
+      "step": 9510
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.111323356628418,
+      "learning_rate": 8.070451719850809e-05,
+      "loss": 1.178,
+      "step": 9520
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.1351996660232544,
+      "learning_rate": 8.06858682138417e-05,
+      "loss": 1.1921,
+      "step": 9530
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.1319098472595215,
+      "learning_rate": 8.066514711976792e-05,
+      "loss": 1.2069,
+      "step": 9540
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.080480694770813,
+      "learning_rate": 8.064442602569417e-05,
+      "loss": 1.1949,
+      "step": 9550
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.126483678817749,
+      "learning_rate": 8.062370493162039e-05,
+      "loss": 1.1523,
+      "step": 9560
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.0840978622436523,
+      "learning_rate": 8.060298383754663e-05,
+      "loss": 1.2074,
+      "step": 9570
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.1075447797775269,
+      "learning_rate": 8.058226274347286e-05,
+      "loss": 1.1937,
+      "step": 9580
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.0781664848327637,
+      "learning_rate": 8.05615416493991e-05,
+      "loss": 1.1837,
+      "step": 9590
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.0375909805297852,
+      "learning_rate": 8.054082055532532e-05,
+      "loss": 1.1928,
+      "step": 9600
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.101149320602417,
+      "learning_rate": 8.052009946125155e-05,
+      "loss": 1.2006,
+      "step": 9610
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.9537743926048279,
+      "learning_rate": 8.049937836717779e-05,
+      "loss": 1.1939,
+      "step": 9620
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.1106035709381104,
+      "learning_rate": 8.047865727310402e-05,
+      "loss": 1.1874,
+      "step": 9630
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.1406303644180298,
+      "learning_rate": 8.045793617903026e-05,
+      "loss": 1.1885,
+      "step": 9640
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.0510179996490479,
+      "learning_rate": 8.043721508495648e-05,
+      "loss": 1.1831,
+      "step": 9650
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.1432597637176514,
+      "learning_rate": 8.041649399088273e-05,
+      "loss": 1.1746,
+      "step": 9660
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.0710557699203491,
+      "learning_rate": 8.039577289680895e-05,
+      "loss": 1.1668,
+      "step": 9670
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.067130446434021,
+      "learning_rate": 8.037505180273519e-05,
+      "loss": 1.1895,
+      "step": 9680
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1639565229415894,
+      "learning_rate": 8.035433070866142e-05,
+      "loss": 1.1786,
+      "step": 9690
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.2795084714889526,
+      "eval_runtime": 1604.6697,
+      "eval_samples_per_second": 262.864,
+      "eval_steps_per_second": 4.107,
+      "step": 9692
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 48460,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "total_flos": 4.1335523730815713e+18,
+  "train_batch_size": 6,
+  "trial_name": null,
+  "trial_params": null
+}