diff --git "a/checkpoint-1500/trainer_state.json" "b/checkpoint-1500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1500/trainer_state.json"
@@ -0,0 +1,10533 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.29277813923227064,
+  "eval_steps": 500,
+  "global_step": 1500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00048796356538711777,
+      "grad_norm": 0.45235496759414673,
+      "learning_rate": 0.0001,
+      "loss": 1.9001,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009759271307742355,
+      "grad_norm": 0.43486157059669495,
+      "learning_rate": 0.0001,
+      "loss": 2.1464,
+      "step": 2
+    },
+    {
+      "epoch": 0.0014638906961613532,
+      "grad_norm": 0.25276443362236023,
+      "learning_rate": 0.0001,
+      "loss": 1.8925,
+      "step": 3
+    },
+    {
+      "epoch": 0.001951854261548471,
+      "grad_norm": 0.3145711421966553,
+      "learning_rate": 0.0001,
+      "loss": 1.9831,
+      "step": 4
+    },
+    {
+      "epoch": 0.002439817826935589,
+      "grad_norm": 0.26927781105041504,
+      "learning_rate": 0.0001,
+      "loss": 1.8612,
+      "step": 5
+    },
+    {
+      "epoch": 0.0029277813923227064,
+      "grad_norm": 0.23938129842281342,
+      "learning_rate": 0.0001,
+      "loss": 1.8369,
+      "step": 6
+    },
+    {
+      "epoch": 0.0034157449577098243,
+      "grad_norm": 0.24115915596485138,
+      "learning_rate": 0.0001,
+      "loss": 1.8733,
+      "step": 7
+    },
+    {
+      "epoch": 0.003903708523096942,
+      "grad_norm": 0.23560288548469543,
+      "learning_rate": 0.0001,
+      "loss": 1.8033,
+      "step": 8
+    },
+    {
+      "epoch": 0.00439167208848406,
+      "grad_norm": 0.20887281000614166,
+      "learning_rate": 0.0001,
+      "loss": 1.8077,
+      "step": 9
+    },
+    {
+      "epoch": 0.004879635653871178,
+      "grad_norm": 0.18585632741451263,
+      "learning_rate": 0.0001,
+      "loss": 1.7501,
+      "step": 10
+    },
+    {
+      "epoch": 0.005367599219258295,
+      "grad_norm": 0.20972006022930145,
+      "learning_rate": 0.0001,
+      "loss": 1.903,
+      "step": 11
+    },
+    {
+      "epoch": 0.005855562784645413,
+      "grad_norm": 0.20946276187896729,
+      "learning_rate": 0.0001,
+      "loss": 1.8127,
+      "step": 12
+    },
+    {
+      "epoch": 0.006343526350032531,
+      "grad_norm": 0.21361751854419708,
+      "learning_rate": 0.0001,
+      "loss": 1.8491,
+      "step": 13
+    },
+    {
+      "epoch": 0.0068314899154196486,
+      "grad_norm": 0.189361110329628,
+      "learning_rate": 0.0001,
+      "loss": 1.7515,
+      "step": 14
+    },
+    {
+      "epoch": 0.007319453480806766,
+      "grad_norm": 0.2290583848953247,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 15
+    },
+    {
+      "epoch": 0.007807417046193884,
+      "grad_norm": 0.19767743349075317,
+      "learning_rate": 0.0001,
+      "loss": 1.7545,
+      "step": 16
+    },
+    {
+      "epoch": 0.008295380611581003,
+      "grad_norm": 0.17406775057315826,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 17
+    },
+    {
+      "epoch": 0.00878334417696812,
+      "grad_norm": 0.19132326543331146,
+      "learning_rate": 0.0001,
+      "loss": 1.9467,
+      "step": 18
+    },
+    {
+      "epoch": 0.009271307742355238,
+      "grad_norm": 0.17514926195144653,
+      "learning_rate": 0.0001,
+      "loss": 1.7956,
+      "step": 19
+    },
+    {
+      "epoch": 0.009759271307742356,
+      "grad_norm": 0.17001329362392426,
+      "learning_rate": 0.0001,
+      "loss": 1.7781,
+      "step": 20
+    },
+    {
+      "epoch": 0.010247234873129472,
+      "grad_norm": 0.1791444718837738,
+      "learning_rate": 0.0001,
+      "loss": 1.78,
+      "step": 21
+    },
+    {
+      "epoch": 0.01073519843851659,
+      "grad_norm": 0.16620689630508423,
+      "learning_rate": 0.0001,
+      "loss": 1.7415,
+      "step": 22
+    },
+    {
+      "epoch": 0.011223162003903709,
+      "grad_norm": 0.17979761958122253,
+      "learning_rate": 0.0001,
+      "loss": 1.9014,
+      "step": 23
+    },
+    {
+      "epoch": 0.011711125569290826,
+      "grad_norm": 0.1886436641216278,
+      "learning_rate": 0.0001,
+      "loss": 1.7947,
+      "step": 24
+    },
+    {
+      "epoch": 0.012199089134677944,
+      "grad_norm": 0.17228351533412933,
+      "learning_rate": 0.0001,
+      "loss": 1.802,
+      "step": 25
+    },
+    {
+      "epoch": 0.012687052700065062,
+      "grad_norm": 0.16874398291110992,
+      "learning_rate": 0.0001,
+      "loss": 1.7005,
+      "step": 26
+    },
+    {
+      "epoch": 0.013175016265452179,
+      "grad_norm": 0.16742444038391113,
+      "learning_rate": 0.0001,
+      "loss": 1.7115,
+      "step": 27
+    },
+    {
+      "epoch": 0.013662979830839297,
+      "grad_norm": 0.17064233124256134,
+      "learning_rate": 0.0001,
+      "loss": 1.7646,
+      "step": 28
+    },
+    {
+      "epoch": 0.014150943396226415,
+      "grad_norm": 0.18888847529888153,
+      "learning_rate": 0.0001,
+      "loss": 1.8769,
+      "step": 29
+    },
+    {
+      "epoch": 0.014638906961613532,
+      "grad_norm": 0.1690024435520172,
+      "learning_rate": 0.0001,
+      "loss": 1.8754,
+      "step": 30
+    },
+    {
+      "epoch": 0.01512687052700065,
+      "grad_norm": 0.167978897690773,
+      "learning_rate": 0.0001,
+      "loss": 1.8164,
+      "step": 31
+    },
+    {
+      "epoch": 0.015614834092387769,
+      "grad_norm": 0.16921842098236084,
+      "learning_rate": 0.0001,
+      "loss": 1.8118,
+      "step": 32
+    },
+    {
+      "epoch": 0.016102797657774885,
+      "grad_norm": 0.17340773344039917,
+      "learning_rate": 0.0001,
+      "loss": 1.9057,
+      "step": 33
+    },
+    {
+      "epoch": 0.016590761223162005,
+      "grad_norm": 0.17796508967876434,
+      "learning_rate": 0.0001,
+      "loss": 1.9025,
+      "step": 34
+    },
+    {
+      "epoch": 0.017078724788549122,
+      "grad_norm": 0.17664046585559845,
+      "learning_rate": 0.0001,
+      "loss": 1.7674,
+      "step": 35
+    },
+    {
+      "epoch": 0.01756668835393624,
+      "grad_norm": 0.1791142076253891,
+      "learning_rate": 0.0001,
+      "loss": 1.8321,
+      "step": 36
+    },
+    {
+      "epoch": 0.01805465191932336,
+      "grad_norm": 0.1765233427286148,
+      "learning_rate": 0.0001,
+      "loss": 1.816,
+      "step": 37
+    },
+    {
+      "epoch": 0.018542615484710475,
+      "grad_norm": 0.18928393721580505,
+      "learning_rate": 0.0001,
+      "loss": 1.8007,
+      "step": 38
+    },
+    {
+      "epoch": 0.01903057905009759,
+      "grad_norm": 0.18933214247226715,
+      "learning_rate": 0.0001,
+      "loss": 1.8042,
+      "step": 39
+    },
+    {
+      "epoch": 0.01951854261548471,
+      "grad_norm": 0.17741172015666962,
+      "learning_rate": 0.0001,
+      "loss": 1.7273,
+      "step": 40
+    },
+    {
+      "epoch": 0.020006506180871828,
+      "grad_norm": 0.1879342496395111,
+      "learning_rate": 0.0001,
+      "loss": 1.9977,
+      "step": 41
+    },
+    {
+      "epoch": 0.020494469746258945,
+      "grad_norm": 0.16950125992298126,
+      "learning_rate": 0.0001,
+      "loss": 1.7911,
+      "step": 42
+    },
+    {
+      "epoch": 0.020982433311646065,
+      "grad_norm": 0.1855982393026352,
+      "learning_rate": 0.0001,
+      "loss": 1.9702,
+      "step": 43
+    },
+    {
+      "epoch": 0.02147039687703318,
+      "grad_norm": 0.1658235490322113,
+      "learning_rate": 0.0001,
+      "loss": 1.925,
+      "step": 44
+    },
+    {
+      "epoch": 0.021958360442420298,
+      "grad_norm": 0.15398763120174408,
+      "learning_rate": 0.0001,
+      "loss": 1.6805,
+      "step": 45
+    },
+    {
+      "epoch": 0.022446324007807418,
+      "grad_norm": 0.18229977786540985,
+      "learning_rate": 0.0001,
+      "loss": 1.8264,
+      "step": 46
+    },
+    {
+      "epoch": 0.022934287573194535,
+      "grad_norm": 0.17499183118343353,
+      "learning_rate": 0.0001,
+      "loss": 1.7677,
+      "step": 47
+    },
+    {
+      "epoch": 0.02342225113858165,
+      "grad_norm": 0.18704566359519958,
+      "learning_rate": 0.0001,
+      "loss": 1.77,
+      "step": 48
+    },
+    {
+      "epoch": 0.02391021470396877,
+      "grad_norm": 0.17721927165985107,
+      "learning_rate": 0.0001,
+      "loss": 1.7035,
+      "step": 49
+    },
+    {
+      "epoch": 0.024398178269355888,
+      "grad_norm": 0.16900165379047394,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 50
+    },
+    {
+      "epoch": 0.024886141834743004,
+      "grad_norm": 0.17378348112106323,
+      "learning_rate": 0.0001,
+      "loss": 1.8528,
+      "step": 51
+    },
+    {
+      "epoch": 0.025374105400130124,
+      "grad_norm": 0.17355363070964813,
+      "learning_rate": 0.0001,
+      "loss": 1.7932,
+      "step": 52
+    },
+    {
+      "epoch": 0.02586206896551724,
+      "grad_norm": 0.1674293577671051,
+      "learning_rate": 0.0001,
+      "loss": 1.7743,
+      "step": 53
+    },
+    {
+      "epoch": 0.026350032530904358,
+      "grad_norm": 0.17658691108226776,
+      "learning_rate": 0.0001,
+      "loss": 1.8141,
+      "step": 54
+    },
+    {
+      "epoch": 0.026837996096291478,
+      "grad_norm": 0.1653955727815628,
+      "learning_rate": 0.0001,
+      "loss": 1.8187,
+      "step": 55
+    },
+    {
+      "epoch": 0.027325959661678594,
+      "grad_norm": 0.17145146429538727,
+      "learning_rate": 0.0001,
+      "loss": 1.7122,
+      "step": 56
+    },
+    {
+      "epoch": 0.02781392322706571,
+      "grad_norm": 0.16360482573509216,
+      "learning_rate": 0.0001,
+      "loss": 1.6993,
+      "step": 57
+    },
+    {
+      "epoch": 0.02830188679245283,
+      "grad_norm": 0.19677112996578217,
+      "learning_rate": 0.0001,
+      "loss": 1.8165,
+      "step": 58
+    },
+    {
+      "epoch": 0.028789850357839947,
+      "grad_norm": 0.18894536793231964,
+      "learning_rate": 0.0001,
+      "loss": 1.9155,
+      "step": 59
+    },
+    {
+      "epoch": 0.029277813923227064,
+      "grad_norm": 0.17368637025356293,
+      "learning_rate": 0.0001,
+      "loss": 1.7435,
+      "step": 60
+    },
+    {
+      "epoch": 0.029765777488614184,
+      "grad_norm": 0.23283590376377106,
+      "learning_rate": 0.0001,
+      "loss": 1.7954,
+      "step": 61
+    },
+    {
+      "epoch": 0.0302537410540013,
+      "grad_norm": 0.18432758748531342,
+      "learning_rate": 0.0001,
+      "loss": 1.8217,
+      "step": 62
+    },
+    {
+      "epoch": 0.03074170461938842,
+      "grad_norm": 0.21392586827278137,
+      "learning_rate": 0.0001,
+      "loss": 1.8528,
+      "step": 63
+    },
+    {
+      "epoch": 0.031229668184775537,
+      "grad_norm": 0.16611824929714203,
+      "learning_rate": 0.0001,
+      "loss": 1.7672,
+      "step": 64
+    },
+    {
+      "epoch": 0.03171763175016266,
+      "grad_norm": 0.19056276977062225,
+      "learning_rate": 0.0001,
+      "loss": 1.829,
+      "step": 65
+    },
+    {
+      "epoch": 0.03220559531554977,
+      "grad_norm": 0.19547034800052643,
+      "learning_rate": 0.0001,
+      "loss": 1.7662,
+      "step": 66
+    },
+    {
+      "epoch": 0.03269355888093689,
+      "grad_norm": 0.16837890446186066,
+      "learning_rate": 0.0001,
+      "loss": 1.8238,
+      "step": 67
+    },
+    {
+      "epoch": 0.03318152244632401,
+      "grad_norm": 0.18697386980056763,
+      "learning_rate": 0.0001,
+      "loss": 1.7575,
+      "step": 68
+    },
+    {
+      "epoch": 0.033669486011711124,
+      "grad_norm": 0.1844586879014969,
+      "learning_rate": 0.0001,
+      "loss": 1.7476,
+      "step": 69
+    },
+    {
+      "epoch": 0.034157449577098244,
+      "grad_norm": 0.168130025267601,
+      "learning_rate": 0.0001,
+      "loss": 1.7118,
+      "step": 70
+    },
+    {
+      "epoch": 0.034645413142485364,
+      "grad_norm": 0.19640620052814484,
+      "learning_rate": 0.0001,
+      "loss": 1.8309,
+      "step": 71
+    },
+    {
+      "epoch": 0.03513337670787248,
+      "grad_norm": 0.16793334484100342,
+      "learning_rate": 0.0001,
+      "loss": 1.7582,
+      "step": 72
+    },
+    {
+      "epoch": 0.0356213402732596,
+      "grad_norm": 0.18016165494918823,
+      "learning_rate": 0.0001,
+      "loss": 1.9166,
+      "step": 73
+    },
+    {
+      "epoch": 0.03610930383864672,
+      "grad_norm": 0.18318477272987366,
+      "learning_rate": 0.0001,
+      "loss": 1.8624,
+      "step": 74
+    },
+    {
+      "epoch": 0.03659726740403383,
+      "grad_norm": 0.1656666249036789,
+      "learning_rate": 0.0001,
+      "loss": 1.7258,
+      "step": 75
+    },
+    {
+      "epoch": 0.03708523096942095,
+      "grad_norm": 0.18281778693199158,
+      "learning_rate": 0.0001,
+      "loss": 1.826,
+      "step": 76
+    },
+    {
+      "epoch": 0.03757319453480807,
+      "grad_norm": 0.1636461615562439,
+      "learning_rate": 0.0001,
+      "loss": 1.6797,
+      "step": 77
+    },
+    {
+      "epoch": 0.03806115810019518,
+      "grad_norm": 0.16847971081733704,
+      "learning_rate": 0.0001,
+      "loss": 1.7408,
+      "step": 78
+    },
+    {
+      "epoch": 0.0385491216655823,
+      "grad_norm": 0.17901544272899628,
+      "learning_rate": 0.0001,
+      "loss": 1.7973,
+      "step": 79
+    },
+    {
+      "epoch": 0.03903708523096942,
+      "grad_norm": 0.16618984937667847,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 80
+    },
+    {
+      "epoch": 0.039525048796356536,
+      "grad_norm": 0.1680297702550888,
+      "learning_rate": 0.0001,
+      "loss": 1.6893,
+      "step": 81
+    },
+    {
+      "epoch": 0.040013012361743656,
+      "grad_norm": 0.17380650341510773,
+      "learning_rate": 0.0001,
+      "loss": 1.7069,
+      "step": 82
+    },
+    {
+      "epoch": 0.040500975927130776,
+      "grad_norm": 0.1793372482061386,
+      "learning_rate": 0.0001,
+      "loss": 1.8204,
+      "step": 83
+    },
+    {
+      "epoch": 0.04098893949251789,
+      "grad_norm": 0.19319359958171844,
+      "learning_rate": 0.0001,
+      "loss": 1.7433,
+      "step": 84
+    },
+    {
+      "epoch": 0.04147690305790501,
+      "grad_norm": 0.19157706201076508,
+      "learning_rate": 0.0001,
+      "loss": 1.8326,
+      "step": 85
+    },
+    {
+      "epoch": 0.04196486662329213,
+      "grad_norm": 0.17349423468112946,
+      "learning_rate": 0.0001,
+      "loss": 1.8403,
+      "step": 86
+    },
+    {
+      "epoch": 0.04245283018867924,
+      "grad_norm": 0.1907309740781784,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 87
+    },
+    {
+      "epoch": 0.04294079375406636,
+      "grad_norm": 0.19307556748390198,
+      "learning_rate": 0.0001,
+      "loss": 1.7551,
+      "step": 88
+    },
+    {
+      "epoch": 0.04342875731945348,
+      "grad_norm": 0.1732914298772812,
+      "learning_rate": 0.0001,
+      "loss": 1.6941,
+      "step": 89
+    },
+    {
+      "epoch": 0.043916720884840596,
+      "grad_norm": 0.188307523727417,
+      "learning_rate": 0.0001,
+      "loss": 1.8148,
+      "step": 90
+    },
+    {
+      "epoch": 0.044404684450227716,
+      "grad_norm": 0.1689937710762024,
+      "learning_rate": 0.0001,
+      "loss": 1.7763,
+      "step": 91
+    },
+    {
+      "epoch": 0.044892648015614836,
+      "grad_norm": 0.17527559399604797,
+      "learning_rate": 0.0001,
+      "loss": 1.7495,
+      "step": 92
+    },
+    {
+      "epoch": 0.04538061158100195,
+      "grad_norm": 0.18800237774848938,
+      "learning_rate": 0.0001,
+      "loss": 1.8781,
+      "step": 93
+    },
+    {
+      "epoch": 0.04586857514638907,
+      "grad_norm": 0.15984949469566345,
+      "learning_rate": 0.0001,
+      "loss": 1.7044,
+      "step": 94
+    },
+    {
+      "epoch": 0.04635653871177619,
+      "grad_norm": 0.18472951650619507,
+      "learning_rate": 0.0001,
+      "loss": 1.746,
+      "step": 95
+    },
+    {
+      "epoch": 0.0468445022771633,
+      "grad_norm": 0.17268826067447662,
+      "learning_rate": 0.0001,
+      "loss": 1.7826,
+      "step": 96
+    },
+    {
+      "epoch": 0.04733246584255042,
+      "grad_norm": 0.1662369668483734,
+      "learning_rate": 0.0001,
+      "loss": 1.6508,
+      "step": 97
+    },
+    {
+      "epoch": 0.04782042940793754,
+      "grad_norm": 0.19034849107265472,
+      "learning_rate": 0.0001,
+      "loss": 1.8245,
+      "step": 98
+    },
+    {
+      "epoch": 0.048308392973324656,
+      "grad_norm": 0.1605503410100937,
+      "learning_rate": 0.0001,
+      "loss": 1.7494,
+      "step": 99
+    },
+    {
+      "epoch": 0.048796356538711776,
+      "grad_norm": 0.16321751475334167,
+      "learning_rate": 0.0001,
+      "loss": 1.7425,
+      "step": 100
+    },
+    {
+      "epoch": 0.049284320104098896,
+      "grad_norm": 0.1676385998725891,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 101
+    },
+    {
+      "epoch": 0.04977228366948601,
+      "grad_norm": 0.17387284338474274,
+      "learning_rate": 0.0001,
+      "loss": 1.7871,
+      "step": 102
+    },
+    {
+      "epoch": 0.05026024723487313,
+      "grad_norm": 0.17714586853981018,
+      "learning_rate": 0.0001,
+      "loss": 1.8214,
+      "step": 103
+    },
+    {
+      "epoch": 0.05074821080026025,
+      "grad_norm": 0.17361727356910706,
+      "learning_rate": 0.0001,
+      "loss": 1.7973,
+      "step": 104
+    },
+    {
+      "epoch": 0.05123617436564736,
+      "grad_norm": 0.1830386072397232,
+      "learning_rate": 0.0001,
+      "loss": 1.7905,
+      "step": 105
+    },
+    {
+      "epoch": 0.05172413793103448,
+      "grad_norm": 0.17435048520565033,
+      "learning_rate": 0.0001,
+      "loss": 1.7082,
+      "step": 106
+    },
+    {
+      "epoch": 0.0522121014964216,
+      "grad_norm": 0.16035646200180054,
+      "learning_rate": 0.0001,
+      "loss": 1.6853,
+      "step": 107
+    },
+    {
+      "epoch": 0.052700065061808715,
+      "grad_norm": 0.16368399560451508,
+      "learning_rate": 0.0001,
+      "loss": 1.6984,
+      "step": 108
+    },
+    {
+      "epoch": 0.053188028627195835,
+      "grad_norm": 0.17323166131973267,
+      "learning_rate": 0.0001,
+      "loss": 1.8575,
+      "step": 109
+    },
+    {
+      "epoch": 0.053675992192582955,
+      "grad_norm": 0.17869392037391663,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 110
+    },
+    {
+      "epoch": 0.05416395575797007,
+      "grad_norm": 0.1777440756559372,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 111
+    },
+    {
+      "epoch": 0.05465191932335719,
+      "grad_norm": 0.1671936959028244,
+      "learning_rate": 0.0001,
+      "loss": 1.7118,
+      "step": 112
+    },
+    {
+      "epoch": 0.05513988288874431,
+      "grad_norm": 0.1724967062473297,
+      "learning_rate": 0.0001,
+      "loss": 1.7542,
+      "step": 113
+    },
+    {
+      "epoch": 0.05562784645413142,
+      "grad_norm": 0.17642249166965485,
+      "learning_rate": 0.0001,
+      "loss": 1.7675,
+      "step": 114
+    },
+    {
+      "epoch": 0.05611581001951854,
+      "grad_norm": 0.16632404923439026,
+      "learning_rate": 0.0001,
+      "loss": 1.7547,
+      "step": 115
+    },
+    {
+      "epoch": 0.05660377358490566,
+      "grad_norm": 0.17003463208675385,
+      "learning_rate": 0.0001,
+      "loss": 1.8396,
+      "step": 116
+    },
+    {
+      "epoch": 0.057091737150292775,
+      "grad_norm": 0.16343915462493896,
+      "learning_rate": 0.0001,
+      "loss": 1.7101,
+      "step": 117
+    },
+    {
+      "epoch": 0.057579700715679895,
+      "grad_norm": 0.1717432588338852,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 118
+    },
+    {
+      "epoch": 0.058067664281067015,
+      "grad_norm": 0.1686941534280777,
+      "learning_rate": 0.0001,
+      "loss": 1.8002,
+      "step": 119
+    },
+    {
+      "epoch": 0.05855562784645413,
+      "grad_norm": 0.17584019899368286,
+      "learning_rate": 0.0001,
+      "loss": 1.7712,
+      "step": 120
+    },
+    {
+      "epoch": 0.05904359141184125,
+      "grad_norm": 0.17191188037395477,
+      "learning_rate": 0.0001,
+      "loss": 1.8542,
+      "step": 121
+    },
+    {
+      "epoch": 0.05953155497722837,
+      "grad_norm": 0.1637864112854004,
+      "learning_rate": 0.0001,
+      "loss": 1.7662,
+      "step": 122
+    },
+    {
+      "epoch": 0.06001951854261549,
+      "grad_norm": 0.16196034848690033,
+      "learning_rate": 0.0001,
+      "loss": 1.7542,
+      "step": 123
+    },
+    {
+      "epoch": 0.0605074821080026,
+      "grad_norm": 0.1562560349702835,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 124
+    },
+    {
+      "epoch": 0.06099544567338972,
+      "grad_norm": 0.1768195629119873,
+      "learning_rate": 0.0001,
+      "loss": 1.8651,
+      "step": 125
+    },
+    {
+      "epoch": 0.06148340923877684,
+      "grad_norm": 0.16431696712970734,
+      "learning_rate": 0.0001,
+      "loss": 1.7415,
+      "step": 126
+    },
+    {
+      "epoch": 0.061971372804163954,
+      "grad_norm": 0.16407223045825958,
+      "learning_rate": 0.0001,
+      "loss": 1.7264,
+      "step": 127
+    },
+    {
+      "epoch": 0.062459336369551074,
+      "grad_norm": 0.18114937841892242,
+      "learning_rate": 0.0001,
+      "loss": 1.6696,
+      "step": 128
+    },
+    {
+      "epoch": 0.0629472999349382,
+      "grad_norm": 0.173524409532547,
+      "learning_rate": 0.0001,
+      "loss": 1.7922,
+      "step": 129
+    },
+    {
+      "epoch": 0.06343526350032531,
+      "grad_norm": 0.16617029905319214,
+      "learning_rate": 0.0001,
+      "loss": 1.7631,
+      "step": 130
+    },
+    {
+      "epoch": 0.06392322706571242,
+      "grad_norm": 0.17115814983844757,
+      "learning_rate": 0.0001,
+      "loss": 1.7364,
+      "step": 131
+    },
+    {
+      "epoch": 0.06441119063109954,
+      "grad_norm": 0.16786643862724304,
+      "learning_rate": 0.0001,
+      "loss": 1.8388,
+      "step": 132
+    },
+    {
+      "epoch": 0.06489915419648666,
+      "grad_norm": 0.16136440634727478,
+      "learning_rate": 0.0001,
+      "loss": 1.6948,
+      "step": 133
+    },
+    {
+      "epoch": 0.06538711776187378,
+      "grad_norm": 0.17984013259410858,
+      "learning_rate": 0.0001,
+      "loss": 1.7368,
+      "step": 134
+    },
+    {
+      "epoch": 0.0658750813272609,
+      "grad_norm": 0.16125819087028503,
+      "learning_rate": 0.0001,
+      "loss": 1.641,
+      "step": 135
+    },
+    {
+      "epoch": 0.06636304489264802,
+      "grad_norm": 0.17826807498931885,
+      "learning_rate": 0.0001,
+      "loss": 1.7775,
+      "step": 136
+    },
+    {
+      "epoch": 0.06685100845803513,
+      "grad_norm": 0.1661001592874527,
+      "learning_rate": 0.0001,
+      "loss": 1.7159,
+      "step": 137
+    },
+    {
+      "epoch": 0.06733897202342225,
+      "grad_norm": 0.17393222451210022,
+      "learning_rate": 0.0001,
+      "loss": 1.7477,
+      "step": 138
+    },
+    {
+      "epoch": 0.06782693558880937,
+      "grad_norm": 0.16058094799518585,
+      "learning_rate": 0.0001,
+      "loss": 1.721,
+      "step": 139
+    },
+    {
+      "epoch": 0.06831489915419649,
+      "grad_norm": 0.1659715622663498,
+      "learning_rate": 0.0001,
+      "loss": 1.7529,
+      "step": 140
+    },
+    {
+      "epoch": 0.06880286271958361,
+      "grad_norm": 0.1750044971704483,
+      "learning_rate": 0.0001,
+      "loss": 1.7368,
+      "step": 141
+    },
+    {
+      "epoch": 0.06929082628497073,
+      "grad_norm": 0.1732410192489624,
+      "learning_rate": 0.0001,
+      "loss": 1.8097,
+      "step": 142
+    },
+    {
+      "epoch": 0.06977878985035783,
+      "grad_norm": 0.1686231940984726,
+      "learning_rate": 0.0001,
+      "loss": 1.7157,
+      "step": 143
+    },
+    {
+      "epoch": 0.07026675341574495,
+      "grad_norm": 0.18054793775081635,
+      "learning_rate": 0.0001,
+      "loss": 1.7897,
+      "step": 144
+    },
+    {
+      "epoch": 0.07075471698113207,
+      "grad_norm": 0.1711822897195816,
+      "learning_rate": 0.0001,
+      "loss": 1.8566,
+      "step": 145
+    },
+    {
+      "epoch": 0.0712426805465192,
+      "grad_norm": 0.16948488354682922,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 146
+    },
+    {
+      "epoch": 0.07173064411190631,
+      "grad_norm": 0.16647808253765106,
+      "learning_rate": 0.0001,
+      "loss": 1.7269,
+      "step": 147
+    },
+    {
+      "epoch": 0.07221860767729343,
+      "grad_norm": 0.17416086792945862,
+      "learning_rate": 0.0001,
+      "loss": 1.7582,
+      "step": 148
+    },
+    {
+      "epoch": 0.07270657124268054,
+      "grad_norm": 0.16461306810379028,
+      "learning_rate": 0.0001,
+      "loss": 1.7288,
+      "step": 149
+    },
+    {
+      "epoch": 0.07319453480806766,
+      "grad_norm": 0.16426099836826324,
+      "learning_rate": 0.0001,
+      "loss": 1.7417,
+      "step": 150
+    },
+    {
+      "epoch": 0.07368249837345478,
+      "grad_norm": 0.16929030418395996,
+      "learning_rate": 0.0001,
+      "loss": 1.7364,
+      "step": 151
+    },
+    {
+      "epoch": 0.0741704619388419,
+      "grad_norm": 0.15853451192378998,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 152
+    },
+    {
+      "epoch": 0.07465842550422902,
+      "grad_norm": 0.1639445722103119,
+      "learning_rate": 0.0001,
+      "loss": 1.5698,
+      "step": 153
+    },
+    {
+      "epoch": 0.07514638906961614,
+      "grad_norm": 0.17728741466999054,
+      "learning_rate": 0.0001,
+      "loss": 1.7382,
+      "step": 154
+    },
+    {
+      "epoch": 0.07563435263500325,
+      "grad_norm": 0.16459015011787415,
+      "learning_rate": 0.0001,
+      "loss": 1.752,
+      "step": 155
+    },
+    {
+      "epoch": 0.07612231620039037,
+      "grad_norm": 0.19319362938404083,
+      "learning_rate": 0.0001,
+      "loss": 1.6967,
+      "step": 156
+    },
+    {
+      "epoch": 0.07661027976577749,
+      "grad_norm": 0.18388009071350098,
+      "learning_rate": 0.0001,
+      "loss": 1.826,
+      "step": 157
+    },
+    {
+      "epoch": 0.0770982433311646,
+      "grad_norm": 0.1937394142150879,
+      "learning_rate": 0.0001,
+      "loss": 1.8212,
+      "step": 158
+    },
+    {
+      "epoch": 0.07758620689655173,
+      "grad_norm": 0.19462037086486816,
+      "learning_rate": 0.0001,
+      "loss": 1.8562,
+      "step": 159
+    },
+    {
+      "epoch": 0.07807417046193885,
+      "grad_norm": 0.19148805737495422,
+      "learning_rate": 0.0001,
+      "loss": 1.7581,
+      "step": 160
+    },
+    {
+      "epoch": 0.07856213402732595,
+      "grad_norm": 0.20127439498901367,
+      "learning_rate": 0.0001,
+      "loss": 1.7166,
+      "step": 161
+    },
+    {
+      "epoch": 0.07905009759271307,
+      "grad_norm": 0.17111310362815857,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 162
+    },
+    {
+      "epoch": 0.07953806115810019,
+      "grad_norm": 0.17740145325660706,
+      "learning_rate": 0.0001,
+      "loss": 1.7717,
+      "step": 163
+    },
+    {
+      "epoch": 0.08002602472348731,
+      "grad_norm": 0.17634375393390656,
+      "learning_rate": 0.0001,
+      "loss": 1.8395,
+      "step": 164
+    },
+    {
+      "epoch": 0.08051398828887443,
+      "grad_norm": 0.16785071790218353,
+      "learning_rate": 0.0001,
+      "loss": 1.6695,
+      "step": 165
+    },
+    {
+      "epoch": 0.08100195185426155,
+      "grad_norm": 0.18514229357242584,
+      "learning_rate": 0.0001,
+      "loss": 1.8272,
+      "step": 166
+    },
+    {
+      "epoch": 0.08148991541964867,
+      "grad_norm": 0.16504287719726562,
+      "learning_rate": 0.0001,
+      "loss": 1.7281,
+      "step": 167
+    },
+    {
+      "epoch": 0.08197787898503578,
+      "grad_norm": 0.18087442219257355,
+      "learning_rate": 0.0001,
+      "loss": 1.6966,
+      "step": 168
+    },
+    {
+      "epoch": 0.0824658425504229,
+      "grad_norm": 0.1800006479024887,
+      "learning_rate": 0.0001,
+      "loss": 1.6952,
+      "step": 169
+    },
+    {
+      "epoch": 0.08295380611581002,
+      "grad_norm": 0.17561759054660797,
+      "learning_rate": 0.0001,
+      "loss": 1.702,
+      "step": 170
+    },
+    {
+      "epoch": 0.08344176968119714,
+      "grad_norm": 0.19057367742061615,
+      "learning_rate": 0.0001,
+      "loss": 1.7236,
+      "step": 171
+    },
+    {
+      "epoch": 0.08392973324658426,
+      "grad_norm": 0.18084438145160675,
+      "learning_rate": 0.0001,
+      "loss": 1.7954,
+      "step": 172
+    },
+    {
+      "epoch": 0.08441769681197138,
+      "grad_norm": 0.18055742979049683,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 173
+    },
+    {
+      "epoch": 0.08490566037735849,
+      "grad_norm": 0.17763681709766388,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 174
+    },
+    {
+      "epoch": 0.0853936239427456,
+      "grad_norm": 0.17945432662963867,
+      "learning_rate": 0.0001,
+      "loss": 1.7571,
+      "step": 175
+    },
+    {
+      "epoch": 0.08588158750813273,
+      "grad_norm": 0.1694047898054123,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 176
+    },
+    {
+      "epoch": 0.08636955107351985,
+      "grad_norm": 0.18137672543525696,
+      "learning_rate": 0.0001,
+      "loss": 1.8289,
+      "step": 177
+    },
+    {
+      "epoch": 0.08685751463890697,
+      "grad_norm": 0.16401486098766327,
+      "learning_rate": 0.0001,
+      "loss": 1.679,
+      "step": 178
+    },
+    {
+      "epoch": 0.08734547820429409,
+      "grad_norm": 0.1720065027475357,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 179
+    },
+    {
+      "epoch": 0.08783344176968119,
+      "grad_norm": 0.17486561834812164,
+      "learning_rate": 0.0001,
+      "loss": 1.7911,
+      "step": 180
+    },
+    {
+      "epoch": 0.08832140533506831,
+      "grad_norm": 0.16348989307880402,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 181
+    },
+    {
+      "epoch": 0.08880936890045543,
+      "grad_norm": 0.18959268927574158,
+      "learning_rate": 0.0001,
+      "loss": 1.8392,
+      "step": 182
+    },
+    {
+      "epoch": 0.08929733246584255,
+      "grad_norm": 0.1623651087284088,
+      "learning_rate": 0.0001,
+      "loss": 1.6907,
+      "step": 183
+    },
+    {
+      "epoch": 0.08978529603122967,
+      "grad_norm": 0.17227298021316528,
+      "learning_rate": 0.0001,
+      "loss": 1.6641,
+      "step": 184
+    },
+    {
+      "epoch": 0.09027325959661679,
+      "grad_norm": 0.19647656381130219,
+      "learning_rate": 0.0001,
+      "loss": 1.8747,
+      "step": 185
+    },
+    {
+      "epoch": 0.0907612231620039,
+      "grad_norm": 0.16988711059093475,
+      "learning_rate": 0.0001,
+      "loss": 1.7653,
+      "step": 186
+    },
+    {
+      "epoch": 0.09124918672739102,
+      "grad_norm": 0.16556090116500854,
+      "learning_rate": 0.0001,
+      "loss": 1.8197,
+      "step": 187
+    },
+    {
+      "epoch": 0.09173715029277814,
+      "grad_norm": 0.1782325655221939,
+      "learning_rate": 0.0001,
+      "loss": 1.7989,
+      "step": 188
+    },
+    {
+      "epoch": 0.09222511385816526,
+      "grad_norm": 0.16507327556610107,
+      "learning_rate": 0.0001,
+      "loss": 1.688,
+      "step": 189
+    },
+    {
+      "epoch": 0.09271307742355238,
+      "grad_norm": 0.17170821130275726,
+      "learning_rate": 0.0001,
+      "loss": 1.7644,
+      "step": 190
+    },
+    {
+      "epoch": 0.0932010409889395,
+      "grad_norm": 0.16646890342235565,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 191
+    },
+    {
+      "epoch": 0.0936890045543266,
+      "grad_norm": 0.18012554943561554,
+      "learning_rate": 0.0001,
+      "loss": 1.7296,
+      "step": 192
+    },
+    {
+      "epoch": 0.09417696811971372,
+      "grad_norm": 0.16549071669578552,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 193
+    },
+    {
+      "epoch": 0.09466493168510084,
+      "grad_norm": 0.17764067649841309,
+      "learning_rate": 0.0001,
+      "loss": 1.6848,
+      "step": 194
+    },
+    {
+      "epoch": 0.09515289525048796,
+      "grad_norm": 0.16311874985694885,
+      "learning_rate": 0.0001,
+      "loss": 1.669,
+      "step": 195
+    },
+    {
+      "epoch": 0.09564085881587508,
+      "grad_norm": 0.16409951448440552,
+      "learning_rate": 0.0001,
+      "loss": 1.7287,
+      "step": 196
+    },
+    {
+      "epoch": 0.0961288223812622,
+      "grad_norm": 0.16553544998168945,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 197
+    },
+    {
+      "epoch": 0.09661678594664931,
+      "grad_norm": 0.17524249851703644,
+      "learning_rate": 0.0001,
+      "loss": 1.7125,
+      "step": 198
+    },
+    {
+      "epoch": 0.09710474951203643,
+      "grad_norm": 0.17164002358913422,
+      "learning_rate": 0.0001,
+      "loss": 1.8127,
+      "step": 199
+    },
+    {
+      "epoch": 0.09759271307742355,
+      "grad_norm": 0.17365583777427673,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 200
+    },
+    {
+      "epoch": 0.09808067664281067,
+      "grad_norm": 0.17036426067352295,
+      "learning_rate": 0.0001,
+      "loss": 1.7064,
+      "step": 201
+    },
+    {
+      "epoch": 0.09856864020819779,
+      "grad_norm": 0.17621922492980957,
+      "learning_rate": 0.0001,
+      "loss": 1.804,
+      "step": 202
+    },
+    {
+      "epoch": 0.09905660377358491,
+      "grad_norm": 0.1819635033607483,
+      "learning_rate": 0.0001,
+      "loss": 1.8078,
+      "step": 203
+    },
+    {
+      "epoch": 0.09954456733897202,
+      "grad_norm": 0.17199762165546417,
+      "learning_rate": 0.0001,
+      "loss": 1.75,
+      "step": 204
+    },
+    {
+      "epoch": 0.10003253090435914,
+      "grad_norm": 0.17945396900177002,
+      "learning_rate": 0.0001,
+      "loss": 1.771,
+      "step": 205
+    },
+    {
+      "epoch": 0.10052049446974626,
+      "grad_norm": 0.16921235620975494,
+      "learning_rate": 0.0001,
+      "loss": 1.7998,
+      "step": 206
+    },
+    {
+      "epoch": 0.10100845803513338,
+      "grad_norm": 0.1957091987133026,
+      "learning_rate": 0.0001,
+      "loss": 1.766,
+      "step": 207
+    },
+    {
+      "epoch": 0.1014964216005205,
+      "grad_norm": 0.15746654570102692,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 208
+    },
+    {
+      "epoch": 0.10198438516590762,
+      "grad_norm": 0.17212480306625366,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 209
+    },
+    {
+      "epoch": 0.10247234873129472,
+      "grad_norm": 0.17027129232883453,
+      "learning_rate": 0.0001,
+      "loss": 1.7245,
+      "step": 210
+    },
+    {
+      "epoch": 0.10296031229668184,
+      "grad_norm": 0.16660825908184052,
+      "learning_rate": 0.0001,
+      "loss": 1.7172,
+      "step": 211
+    },
+    {
+      "epoch": 0.10344827586206896,
+      "grad_norm": 0.17144639790058136,
+      "learning_rate": 0.0001,
+      "loss": 1.6866,
+      "step": 212
+    },
+    {
+      "epoch": 0.10393623942745608,
+      "grad_norm": 0.16561684012413025,
+      "learning_rate": 0.0001,
+      "loss": 1.7672,
+      "step": 213
+    },
+    {
+      "epoch": 0.1044242029928432,
+      "grad_norm": 0.19971047341823578,
+      "learning_rate": 0.0001,
+      "loss": 1.9155,
+      "step": 214
+    },
+    {
+      "epoch": 0.10491216655823032,
+      "grad_norm": 0.16283684968948364,
+      "learning_rate": 0.0001,
+      "loss": 1.6483,
+      "step": 215
+    },
+    {
+      "epoch": 0.10540013012361743,
+      "grad_norm": 0.16953429579734802,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 216
+    },
+    {
+      "epoch": 0.10588809368900455,
+      "grad_norm": 0.16256603598594666,
+      "learning_rate": 0.0001,
+      "loss": 1.6921,
+      "step": 217
+    },
+    {
+      "epoch": 0.10637605725439167,
+      "grad_norm": 0.1671057939529419,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 218
+    },
+    {
+      "epoch": 0.10686402081977879,
+      "grad_norm": 0.17248627543449402,
+      "learning_rate": 0.0001,
+      "loss": 1.7459,
+      "step": 219
+    },
+    {
+      "epoch": 0.10735198438516591,
+      "grad_norm": 0.1770409643650055,
+      "learning_rate": 0.0001,
+      "loss": 1.7203,
+      "step": 220
+    },
+    {
+      "epoch": 0.10783994795055303,
+      "grad_norm": 0.18381726741790771,
+      "learning_rate": 0.0001,
+      "loss": 1.6943,
+      "step": 221
+    },
+    {
+      "epoch": 0.10832791151594014,
+      "grad_norm": 0.17488622665405273,
+      "learning_rate": 0.0001,
+      "loss": 1.7708,
+      "step": 222
+    },
+    {
+      "epoch": 0.10881587508132726,
+      "grad_norm": 0.1846553534269333,
+      "learning_rate": 0.0001,
+      "loss": 1.7445,
+      "step": 223
+    },
+    {
+      "epoch": 0.10930383864671438,
+      "grad_norm": 0.17383840680122375,
+      "learning_rate": 0.0001,
+      "loss": 1.7281,
+      "step": 224
+    },
+    {
+      "epoch": 0.1097918022121015,
+      "grad_norm": 0.18164974451065063,
+      "learning_rate": 0.0001,
+      "loss": 1.7933,
+      "step": 225
+    },
+    {
+      "epoch": 0.11027976577748862,
+      "grad_norm": 0.17281731963157654,
+      "learning_rate": 0.0001,
+      "loss": 1.8216,
+      "step": 226
+    },
+    {
+      "epoch": 0.11076772934287574,
+      "grad_norm": 0.1618274599313736,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 227
+    },
+    {
+      "epoch": 0.11125569290826284,
+      "grad_norm": 0.18077383935451508,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 228
+    },
+    {
+      "epoch": 0.11174365647364996,
+      "grad_norm": 0.16627463698387146,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 229
+    },
+    {
+      "epoch": 0.11223162003903708,
+      "grad_norm": 0.18027973175048828,
+      "learning_rate": 0.0001,
+      "loss": 1.784,
+      "step": 230
+    },
+    {
+      "epoch": 0.1127195836044242,
+      "grad_norm": 0.16198568046092987,
+      "learning_rate": 0.0001,
+      "loss": 1.6985,
+      "step": 231
+    },
+    {
+      "epoch": 0.11320754716981132,
+      "grad_norm": 0.16335777938365936,
+      "learning_rate": 0.0001,
+      "loss": 1.7173,
+      "step": 232
+    },
+    {
+      "epoch": 0.11369551073519844,
+      "grad_norm": 0.1790529191493988,
+      "learning_rate": 0.0001,
+      "loss": 1.7109,
+      "step": 233
+    },
+    {
+      "epoch": 0.11418347430058555,
+      "grad_norm": 0.18429918587207794,
+      "learning_rate": 0.0001,
+      "loss": 1.889,
+      "step": 234
+    },
+    {
+      "epoch": 0.11467143786597267,
+      "grad_norm": 0.17136062681674957,
+      "learning_rate": 0.0001,
+      "loss": 1.7453,
+      "step": 235
+    },
+    {
+      "epoch": 0.11515940143135979,
+      "grad_norm": 0.17297938466072083,
+      "learning_rate": 0.0001,
+      "loss": 1.7841,
+      "step": 236
+    },
+    {
+      "epoch": 0.11564736499674691,
+      "grad_norm": 0.17786864936351776,
+      "learning_rate": 0.0001,
+      "loss": 1.8118,
+      "step": 237
+    },
+    {
+      "epoch": 0.11613532856213403,
+      "grad_norm": 0.16814367473125458,
+      "learning_rate": 0.0001,
+      "loss": 1.7541,
+      "step": 238
+    },
+    {
+      "epoch": 0.11662329212752115,
+      "grad_norm": 0.16605667769908905,
+      "learning_rate": 0.0001,
+      "loss": 1.5556,
+      "step": 239
+    },
+    {
+      "epoch": 0.11711125569290826,
+      "grad_norm": 0.1906554251909256,
+      "learning_rate": 0.0001,
+      "loss": 1.7896,
+      "step": 240
+    },
+    {
+      "epoch": 0.11759921925829538,
+      "grad_norm": 0.1695682555437088,
+      "learning_rate": 0.0001,
+      "loss": 1.6678,
+      "step": 241
+    },
+    {
+      "epoch": 0.1180871828236825,
+      "grad_norm": 0.16519296169281006,
+      "learning_rate": 0.0001,
+      "loss": 1.6991,
+      "step": 242
+    },
+    {
+      "epoch": 0.11857514638906962,
+      "grad_norm": 0.1673249900341034,
+      "learning_rate": 0.0001,
+      "loss": 1.5972,
+      "step": 243
+    },
+    {
+      "epoch": 0.11906310995445674,
+      "grad_norm": 0.1892324984073639,
+      "learning_rate": 0.0001,
+      "loss": 1.7928,
+      "step": 244
+    },
+    {
+      "epoch": 0.11955107351984386,
+      "grad_norm": 0.1725611835718155,
+      "learning_rate": 0.0001,
+      "loss": 1.7457,
+      "step": 245
+    },
+    {
+      "epoch": 0.12003903708523098,
+      "grad_norm": 0.17512311041355133,
+      "learning_rate": 0.0001,
+      "loss": 1.7149,
+      "step": 246
+    },
+    {
+      "epoch": 0.12052700065061808,
+      "grad_norm": 0.17550413310527802,
+      "learning_rate": 0.0001,
+      "loss": 1.6261,
+      "step": 247
+    },
+    {
+      "epoch": 0.1210149642160052,
+      "grad_norm": 0.17173801362514496,
+      "learning_rate": 0.0001,
+      "loss": 1.746,
+      "step": 248
+    },
+    {
+      "epoch": 0.12150292778139232,
+      "grad_norm": 0.20147019624710083,
+      "learning_rate": 0.0001,
+      "loss": 1.7803,
+      "step": 249
+    },
+    {
+      "epoch": 0.12199089134677944,
+      "grad_norm": 0.17563247680664062,
+      "learning_rate": 0.0001,
+      "loss": 1.8358,
+      "step": 250
+    },
+    {
+      "epoch": 0.12247885491216656,
+      "grad_norm": 0.1900017112493515,
+      "learning_rate": 0.0001,
+      "loss": 1.7524,
+      "step": 251
+    },
+    {
+      "epoch": 0.12296681847755368,
+      "grad_norm": 0.17182764410972595,
+      "learning_rate": 0.0001,
+      "loss": 1.6817,
+      "step": 252
+    },
+    {
+      "epoch": 0.12345478204294079,
+      "grad_norm": 0.178592249751091,
+      "learning_rate": 0.0001,
+      "loss": 1.8205,
+      "step": 253
+    },
+    {
+      "epoch": 0.12394274560832791,
+      "grad_norm": 0.1863623857498169,
+      "learning_rate": 0.0001,
+      "loss": 1.7028,
+      "step": 254
+    },
+    {
+      "epoch": 0.12443070917371503,
+      "grad_norm": 0.1741155982017517,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 255
+    },
+    {
+      "epoch": 0.12491867273910215,
+      "grad_norm": 0.17678064107894897,
+      "learning_rate": 0.0001,
+      "loss": 1.7339,
+      "step": 256
+    },
+    {
+      "epoch": 0.12540663630448926,
+      "grad_norm": 0.18696452677249908,
+      "learning_rate": 0.0001,
+      "loss": 1.7339,
+      "step": 257
+    },
+    {
+      "epoch": 0.1258945998698764,
+      "grad_norm": 0.17113931477069855,
+      "learning_rate": 0.0001,
+      "loss": 1.6598,
+      "step": 258
+    },
+    {
+      "epoch": 0.1263825634352635,
+      "grad_norm": 0.19014360010623932,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 259
+    },
+    {
+      "epoch": 0.12687052700065063,
+      "grad_norm": 0.18524205684661865,
+      "learning_rate": 0.0001,
+      "loss": 1.7354,
+      "step": 260
+    },
+    {
+      "epoch": 0.12735849056603774,
+      "grad_norm": 0.17393754422664642,
+      "learning_rate": 0.0001,
+      "loss": 1.7243,
+      "step": 261
+    },
+    {
+      "epoch": 0.12784645413142484,
+      "grad_norm": 0.19686400890350342,
+      "learning_rate": 0.0001,
+      "loss": 1.744,
+      "step": 262
+    },
+    {
+      "epoch": 0.12833441769681198,
+      "grad_norm": 0.17392823100090027,
+      "learning_rate": 0.0001,
+      "loss": 1.8156,
+      "step": 263
+    },
+    {
+      "epoch": 0.12882238126219908,
+      "grad_norm": 0.1770251989364624,
+      "learning_rate": 0.0001,
+      "loss": 1.7658,
+      "step": 264
+    },
+    {
+      "epoch": 0.12931034482758622,
+      "grad_norm": 0.186493381857872,
+      "learning_rate": 0.0001,
+      "loss": 1.7858,
+      "step": 265
+    },
+    {
+      "epoch": 0.12979830839297332,
+      "grad_norm": 0.1572084277868271,
+      "learning_rate": 0.0001,
+      "loss": 1.6333,
+      "step": 266
+    },
+    {
+      "epoch": 0.13028627195836046,
+      "grad_norm": 0.17564861476421356,
+      "learning_rate": 0.0001,
+      "loss": 1.7765,
+      "step": 267
+    },
+    {
+      "epoch": 0.13077423552374756,
+      "grad_norm": 0.17942026257514954,
+      "learning_rate": 0.0001,
+      "loss": 1.7647,
+      "step": 268
+    },
+    {
+      "epoch": 0.13126219908913467,
+      "grad_norm": 0.19546084105968475,
+      "learning_rate": 0.0001,
+      "loss": 1.6937,
+      "step": 269
+    },
+    {
+      "epoch": 0.1317501626545218,
+      "grad_norm": 0.16391515731811523,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 270
+    },
+    {
+      "epoch": 0.1322381262199089,
+      "grad_norm": 0.1976781040430069,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 271
+    },
+    {
+      "epoch": 0.13272608978529604,
+      "grad_norm": 0.17148089408874512,
+      "learning_rate": 0.0001,
+      "loss": 1.7375,
+      "step": 272
+    },
+    {
+      "epoch": 0.13321405335068315,
+      "grad_norm": 0.18054482340812683,
+      "learning_rate": 0.0001,
+      "loss": 1.6969,
+      "step": 273
+    },
+    {
+      "epoch": 0.13370201691607025,
+      "grad_norm": 0.17732256650924683,
+      "learning_rate": 0.0001,
+      "loss": 1.7741,
+      "step": 274
+    },
+    {
+      "epoch": 0.1341899804814574,
+      "grad_norm": 0.19430918991565704,
+      "learning_rate": 0.0001,
+      "loss": 1.742,
+      "step": 275
+    },
+    {
+      "epoch": 0.1346779440468445,
+      "grad_norm": 0.17336079478263855,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 276
+    },
+    {
+      "epoch": 0.13516590761223163,
+      "grad_norm": 0.1854289472103119,
+      "learning_rate": 0.0001,
+      "loss": 1.7894,
+      "step": 277
+    },
+    {
+      "epoch": 0.13565387117761873,
+      "grad_norm": 0.19312992691993713,
+      "learning_rate": 0.0001,
+      "loss": 1.5598,
+      "step": 278
+    },
+    {
+      "epoch": 0.13614183474300587,
+      "grad_norm": 0.17979946732521057,
+      "learning_rate": 0.0001,
+      "loss": 1.7307,
+      "step": 279
+    },
+    {
+      "epoch": 0.13662979830839297,
+      "grad_norm": 0.19239358603954315,
+      "learning_rate": 0.0001,
+      "loss": 1.7109,
+      "step": 280
+    },
+    {
+      "epoch": 0.13711776187378008,
+      "grad_norm": 0.19618631899356842,
+      "learning_rate": 0.0001,
+      "loss": 1.6726,
+      "step": 281
+    },
+    {
+      "epoch": 0.13760572543916721,
+      "grad_norm": 0.16837313771247864,
+      "learning_rate": 0.0001,
+      "loss": 1.6965,
+      "step": 282
+    },
+    {
+      "epoch": 0.13809368900455432,
+      "grad_norm": 0.1867402195930481,
+      "learning_rate": 0.0001,
+      "loss": 1.7647,
+      "step": 283
+    },
+    {
+      "epoch": 0.13858165256994145,
+      "grad_norm": 0.17642104625701904,
+      "learning_rate": 0.0001,
+      "loss": 1.6309,
+      "step": 284
+    },
+    {
+      "epoch": 0.13906961613532856,
+      "grad_norm": 0.1965477168560028,
+      "learning_rate": 0.0001,
+      "loss": 1.7407,
+      "step": 285
+    },
+    {
+      "epoch": 0.13955757970071567,
+      "grad_norm": 0.16644583642482758,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 286
+    },
+    {
+      "epoch": 0.1400455432661028,
+      "grad_norm": 0.19915439188480377,
+      "learning_rate": 0.0001,
+      "loss": 1.9156,
+      "step": 287
+    },
+    {
+      "epoch": 0.1405335068314899,
+      "grad_norm": 0.18034371733665466,
+      "learning_rate": 0.0001,
+      "loss": 1.8389,
+      "step": 288
+    },
+    {
+      "epoch": 0.14102147039687704,
+      "grad_norm": 0.16677184402942657,
+      "learning_rate": 0.0001,
+      "loss": 1.6491,
+      "step": 289
+    },
+    {
+      "epoch": 0.14150943396226415,
+      "grad_norm": 0.16290052235126495,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 290
+    },
+    {
+      "epoch": 0.14199739752765128,
+      "grad_norm": 0.1820719987154007,
+      "learning_rate": 0.0001,
+      "loss": 1.8679,
+      "step": 291
+    },
+    {
+      "epoch": 0.1424853610930384,
+      "grad_norm": 0.17475441098213196,
+      "learning_rate": 0.0001,
+      "loss": 1.8023,
+      "step": 292
+    },
+    {
+      "epoch": 0.1429733246584255,
+      "grad_norm": 0.17056342959403992,
+      "learning_rate": 0.0001,
+      "loss": 1.8024,
+      "step": 293
+    },
+    {
+      "epoch": 0.14346128822381263,
+      "grad_norm": 0.17104357481002808,
+      "learning_rate": 0.0001,
+      "loss": 1.7974,
+      "step": 294
+    },
+    {
+      "epoch": 0.14394925178919973,
+      "grad_norm": 0.1817089319229126,
+      "learning_rate": 0.0001,
+      "loss": 1.737,
+      "step": 295
+    },
+    {
+      "epoch": 0.14443721535458687,
+      "grad_norm": 0.17631934583187103,
+      "learning_rate": 0.0001,
+      "loss": 1.7832,
+      "step": 296
+    },
+    {
+      "epoch": 0.14492517891997397,
+      "grad_norm": 0.17967306077480316,
+      "learning_rate": 0.0001,
+      "loss": 1.6744,
+      "step": 297
+    },
+    {
+      "epoch": 0.14541314248536108,
+      "grad_norm": 0.16978050768375397,
+      "learning_rate": 0.0001,
+      "loss": 1.702,
+      "step": 298
+    },
+    {
+      "epoch": 0.1459011060507482,
+      "grad_norm": 0.17197778820991516,
+      "learning_rate": 0.0001,
+      "loss": 1.7813,
+      "step": 299
+    },
+    {
+      "epoch": 0.14638906961613532,
+      "grad_norm": 0.18907850980758667,
+      "learning_rate": 0.0001,
+      "loss": 1.7179,
+      "step": 300
+    },
+    {
+      "epoch": 0.14687703318152245,
+      "grad_norm": 0.17906910181045532,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 301
+    },
+    {
+      "epoch": 0.14736499674690956,
+      "grad_norm": 0.1662023514509201,
+      "learning_rate": 0.0001,
+      "loss": 1.6932,
+      "step": 302
+    },
+    {
+      "epoch": 0.1478529603122967,
+      "grad_norm": 0.17590110003948212,
+      "learning_rate": 0.0001,
+      "loss": 1.7014,
+      "step": 303
+    },
+    {
+      "epoch": 0.1483409238776838,
+      "grad_norm": 0.16377031803131104,
+      "learning_rate": 0.0001,
+      "loss": 1.742,
+      "step": 304
+    },
+    {
+      "epoch": 0.1488288874430709,
+      "grad_norm": 0.16067785024642944,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 305
+    },
+    {
+      "epoch": 0.14931685100845804,
+      "grad_norm": 0.1768130213022232,
+      "learning_rate": 0.0001,
+      "loss": 1.7215,
+      "step": 306
+    },
+    {
+      "epoch": 0.14980481457384515,
+      "grad_norm": 0.1801900565624237,
+      "learning_rate": 0.0001,
+      "loss": 1.7875,
+      "step": 307
+    },
+    {
+      "epoch": 0.15029277813923228,
+      "grad_norm": 0.1643214225769043,
+      "learning_rate": 0.0001,
+      "loss": 1.6349,
+      "step": 308
+    },
+    {
+      "epoch": 0.1507807417046194,
+      "grad_norm": 0.1663602888584137,
+      "learning_rate": 0.0001,
+      "loss": 1.6794,
+      "step": 309
+    },
+    {
+      "epoch": 0.1512687052700065,
+      "grad_norm": 0.16909120976924896,
+      "learning_rate": 0.0001,
+      "loss": 1.7565,
+      "step": 310
+    },
+    {
+      "epoch": 0.15175666883539363,
+      "grad_norm": 0.18539294600486755,
+      "learning_rate": 0.0001,
+      "loss": 1.8234,
+      "step": 311
+    },
+    {
+      "epoch": 0.15224463240078073,
+      "grad_norm": 0.16808299720287323,
+      "learning_rate": 0.0001,
+      "loss": 1.7665,
+      "step": 312
+    },
+    {
+      "epoch": 0.15273259596616787,
+      "grad_norm": 0.1774633526802063,
+      "learning_rate": 0.0001,
+      "loss": 1.7847,
+      "step": 313
+    },
+    {
+      "epoch": 0.15322055953155497,
+      "grad_norm": 0.16781893372535706,
+      "learning_rate": 0.0001,
+      "loss": 1.7441,
+      "step": 314
+    },
+    {
+      "epoch": 0.1537085230969421,
+      "grad_norm": 0.17828769981861115,
+      "learning_rate": 0.0001,
+      "loss": 1.8414,
+      "step": 315
+    },
+    {
+      "epoch": 0.1541964866623292,
+      "grad_norm": 0.19906285405158997,
+      "learning_rate": 0.0001,
+      "loss": 1.8278,
+      "step": 316
+    },
+    {
+      "epoch": 0.15468445022771632,
+      "grad_norm": 0.16389663517475128,
+      "learning_rate": 0.0001,
+      "loss": 1.722,
+      "step": 317
+    },
+    {
+      "epoch": 0.15517241379310345,
+      "grad_norm": 0.1807340681552887,
+      "learning_rate": 0.0001,
+      "loss": 1.8251,
+      "step": 318
+    },
+    {
+      "epoch": 0.15566037735849056,
+      "grad_norm": 0.18094702064990997,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 319
+    },
+    {
+      "epoch": 0.1561483409238777,
+      "grad_norm": 0.17396771907806396,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 320
+    },
+    {
+      "epoch": 0.1566363044892648,
+      "grad_norm": 0.16685691475868225,
+      "learning_rate": 0.0001,
+      "loss": 1.7451,
+      "step": 321
+    },
+    {
+      "epoch": 0.1571242680546519,
+      "grad_norm": 0.1724182814359665,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 322
+    },
+    {
+      "epoch": 0.15761223162003904,
+      "grad_norm": 0.1878998726606369,
+      "learning_rate": 0.0001,
+      "loss": 1.842,
+      "step": 323
+    },
+    {
+      "epoch": 0.15810019518542615,
+      "grad_norm": 0.1650148630142212,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 324
+    },
+    {
+      "epoch": 0.15858815875081328,
+      "grad_norm": 0.18328998982906342,
+      "learning_rate": 0.0001,
+      "loss": 1.731,
+      "step": 325
+    },
+    {
+      "epoch": 0.15907612231620039,
+      "grad_norm": 0.16306239366531372,
+      "learning_rate": 0.0001,
+      "loss": 1.7239,
+      "step": 326
+    },
+    {
+      "epoch": 0.15956408588158752,
+      "grad_norm": 0.18774734437465668,
+      "learning_rate": 0.0001,
+      "loss": 1.8776,
+      "step": 327
+    },
+    {
+      "epoch": 0.16005204944697463,
+      "grad_norm": 0.17184047400951385,
+      "learning_rate": 0.0001,
+      "loss": 1.7488,
+      "step": 328
+    },
+    {
+      "epoch": 0.16054001301236173,
+      "grad_norm": 0.16623416543006897,
+      "learning_rate": 0.0001,
+      "loss": 1.7079,
+      "step": 329
+    },
+    {
+      "epoch": 0.16102797657774887,
+      "grad_norm": 0.1680457592010498,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 330
+    },
+    {
+      "epoch": 0.16151594014313597,
+      "grad_norm": 0.1677110642194748,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 331
+    },
+    {
+      "epoch": 0.1620039037085231,
+      "grad_norm": 0.17092961072921753,
+      "learning_rate": 0.0001,
+      "loss": 1.711,
+      "step": 332
+    },
+    {
+      "epoch": 0.1624918672739102,
+      "grad_norm": 0.17093709111213684,
+      "learning_rate": 0.0001,
+      "loss": 1.6921,
+      "step": 333
+    },
+    {
+      "epoch": 0.16297983083929735,
+      "grad_norm": 0.16150657832622528,
+      "learning_rate": 0.0001,
+      "loss": 1.6694,
+      "step": 334
+    },
+    {
+      "epoch": 0.16346779440468445,
+      "grad_norm": 0.16476838290691376,
+      "learning_rate": 0.0001,
+      "loss": 1.7551,
+      "step": 335
+    },
+    {
+      "epoch": 0.16395575797007156,
+      "grad_norm": 0.16959933936595917,
+      "learning_rate": 0.0001,
+      "loss": 1.7384,
+      "step": 336
+    },
+    {
+      "epoch": 0.1644437215354587,
+      "grad_norm": 0.16465017199516296,
+      "learning_rate": 0.0001,
+      "loss": 1.7103,
+      "step": 337
+    },
+    {
+      "epoch": 0.1649316851008458,
+      "grad_norm": 0.17374570667743683,
+      "learning_rate": 0.0001,
+      "loss": 1.7194,
+      "step": 338
+    },
+    {
+      "epoch": 0.16541964866623293,
+      "grad_norm": 0.1738986223936081,
+      "learning_rate": 0.0001,
+      "loss": 1.8012,
+      "step": 339
+    },
+    {
+      "epoch": 0.16590761223162004,
+      "grad_norm": 0.18269729614257812,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 340
+    },
+    {
+      "epoch": 0.16639557579700714,
+      "grad_norm": 0.1626196950674057,
+      "learning_rate": 0.0001,
+      "loss": 1.6541,
+      "step": 341
+    },
+    {
+      "epoch": 0.16688353936239428,
+      "grad_norm": 0.1709810197353363,
+      "learning_rate": 0.0001,
+      "loss": 1.7523,
+      "step": 342
+    },
+    {
+      "epoch": 0.16737150292778138,
+      "grad_norm": 0.19227533042430878,
+      "learning_rate": 0.0001,
+      "loss": 1.7016,
+      "step": 343
+    },
+    {
+      "epoch": 0.16785946649316852,
+      "grad_norm": 0.16071516275405884,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 344
+    },
+    {
+      "epoch": 0.16834743005855562,
+      "grad_norm": 0.1687503308057785,
+      "learning_rate": 0.0001,
+      "loss": 1.7709,
+      "step": 345
+    },
+    {
+      "epoch": 0.16883539362394276,
+      "grad_norm": 0.1793443262577057,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 346
+    },
+    {
+      "epoch": 0.16932335718932987,
+      "grad_norm": 0.1658993363380432,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 347
+    },
+    {
+      "epoch": 0.16981132075471697,
+      "grad_norm": 0.1823633462190628,
+      "learning_rate": 0.0001,
+      "loss": 1.6642,
+      "step": 348
+    },
+    {
+      "epoch": 0.1702992843201041,
+      "grad_norm": 0.1775803565979004,
+      "learning_rate": 0.0001,
+      "loss": 1.8153,
+      "step": 349
+    },
+    {
+      "epoch": 0.1707872478854912,
+      "grad_norm": 0.17859414219856262,
+      "learning_rate": 0.0001,
+      "loss": 1.7871,
+      "step": 350
+    },
+    {
+      "epoch": 0.17127521145087835,
+      "grad_norm": 0.17732694745063782,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 351
+    },
+    {
+      "epoch": 0.17176317501626545,
+      "grad_norm": 0.16252091526985168,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 352
+    },
+    {
+      "epoch": 0.17225113858165256,
+      "grad_norm": 0.1633855104446411,
+      "learning_rate": 0.0001,
+      "loss": 1.7048,
+      "step": 353
+    },
+    {
+      "epoch": 0.1727391021470397,
+      "grad_norm": 0.17340290546417236,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 354
+    },
+    {
+      "epoch": 0.1732270657124268,
+      "grad_norm": 0.1728828251361847,
+      "learning_rate": 0.0001,
+      "loss": 1.7594,
+      "step": 355
+    },
+    {
+      "epoch": 0.17371502927781393,
+      "grad_norm": 0.16987989842891693,
+      "learning_rate": 0.0001,
+      "loss": 1.695,
+      "step": 356
+    },
+    {
+      "epoch": 0.17420299284320104,
+      "grad_norm": 0.16485123336315155,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 357
+    },
+    {
+      "epoch": 0.17469095640858817,
+      "grad_norm": 0.17824482917785645,
+      "learning_rate": 0.0001,
+      "loss": 1.7995,
+      "step": 358
+    },
+    {
+      "epoch": 0.17517891997397528,
+      "grad_norm": 0.16260479390621185,
+      "learning_rate": 0.0001,
+      "loss": 1.6892,
+      "step": 359
+    },
+    {
+      "epoch": 0.17566688353936238,
+      "grad_norm": 0.17875920236110687,
+      "learning_rate": 0.0001,
+      "loss": 1.7326,
+      "step": 360
+    },
+    {
+      "epoch": 0.17615484710474952,
+      "grad_norm": 0.18595051765441895,
+      "learning_rate": 0.0001,
+      "loss": 1.8231,
+      "step": 361
+    },
+    {
+      "epoch": 0.17664281067013662,
+      "grad_norm": 0.16535578668117523,
+      "learning_rate": 0.0001,
+      "loss": 1.5903,
+      "step": 362
+    },
+    {
+      "epoch": 0.17713077423552376,
+      "grad_norm": 0.16219404339790344,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 363
+    },
+    {
+      "epoch": 0.17761873780091086,
+      "grad_norm": 0.16999295353889465,
+      "learning_rate": 0.0001,
+      "loss": 1.8443,
+      "step": 364
+    },
+    {
+      "epoch": 0.17810670136629797,
+      "grad_norm": 0.18015706539154053,
+      "learning_rate": 0.0001,
+      "loss": 1.7956,
+      "step": 365
+    },
+    {
+      "epoch": 0.1785946649316851,
+      "grad_norm": 0.17357711493968964,
+      "learning_rate": 0.0001,
+      "loss": 1.7314,
+      "step": 366
+    },
+    {
+      "epoch": 0.1790826284970722,
+      "grad_norm": 0.167061910033226,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 367
+    },
+    {
+      "epoch": 0.17957059206245934,
+      "grad_norm": 0.1628636121749878,
+      "learning_rate": 0.0001,
+      "loss": 1.6806,
+      "step": 368
+    },
+    {
+      "epoch": 0.18005855562784645,
+      "grad_norm": 0.1639977991580963,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 369
+    },
+    {
+      "epoch": 0.18054651919323358,
+      "grad_norm": 0.19148628413677216,
+      "learning_rate": 0.0001,
+      "loss": 1.7109,
+      "step": 370
+    },
+    {
+      "epoch": 0.1810344827586207,
+      "grad_norm": 0.1734737604856491,
+      "learning_rate": 0.0001,
+      "loss": 1.7999,
+      "step": 371
+    },
+    {
+      "epoch": 0.1815224463240078,
+      "grad_norm": 0.18014098703861237,
+      "learning_rate": 0.0001,
+      "loss": 1.7271,
+      "step": 372
+    },
+    {
+      "epoch": 0.18201040988939493,
+      "grad_norm": 0.16859233379364014,
+      "learning_rate": 0.0001,
+      "loss": 1.5582,
+      "step": 373
+    },
+    {
+      "epoch": 0.18249837345478204,
+      "grad_norm": 0.17208896577358246,
+      "learning_rate": 0.0001,
+      "loss": 1.6721,
+      "step": 374
+    },
+    {
+      "epoch": 0.18298633702016917,
+      "grad_norm": 0.18183362483978271,
+      "learning_rate": 0.0001,
+      "loss": 1.8046,
+      "step": 375
+    },
+    {
+      "epoch": 0.18347430058555628,
+      "grad_norm": 0.17301778495311737,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 376
+    },
+    {
+      "epoch": 0.18396226415094338,
+      "grad_norm": 0.17814116179943085,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 377
+    },
+    {
+      "epoch": 0.18445022771633052,
+      "grad_norm": 0.16107924282550812,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 378
+    },
+    {
+      "epoch": 0.18493819128171762,
+      "grad_norm": 0.17282907664775848,
+      "learning_rate": 0.0001,
+      "loss": 1.7728,
+      "step": 379
+    },
+    {
+      "epoch": 0.18542615484710476,
+      "grad_norm": 0.1867058277130127,
+      "learning_rate": 0.0001,
+      "loss": 1.7882,
+      "step": 380
+    },
+    {
+      "epoch": 0.18591411841249186,
+      "grad_norm": 0.16243025660514832,
+      "learning_rate": 0.0001,
+      "loss": 1.6204,
+      "step": 381
+    },
+    {
+      "epoch": 0.186402081977879,
+      "grad_norm": 0.1962565928697586,
+      "learning_rate": 0.0001,
+      "loss": 1.7507,
+      "step": 382
+    },
+    {
+      "epoch": 0.1868900455432661,
+      "grad_norm": 0.16898339986801147,
+      "learning_rate": 0.0001,
+      "loss": 1.7031,
+      "step": 383
+    },
+    {
+      "epoch": 0.1873780091086532,
+      "grad_norm": 0.18223068118095398,
+      "learning_rate": 0.0001,
+      "loss": 1.6749,
+      "step": 384
+    },
+    {
+      "epoch": 0.18786597267404034,
+      "grad_norm": 0.17187203466892242,
+      "learning_rate": 0.0001,
+      "loss": 1.6576,
+      "step": 385
+    },
+    {
+      "epoch": 0.18835393623942745,
+      "grad_norm": 0.17551177740097046,
+      "learning_rate": 0.0001,
+      "loss": 1.7675,
+      "step": 386
+    },
+    {
+      "epoch": 0.18884189980481458,
+      "grad_norm": 0.1808999627828598,
+      "learning_rate": 0.0001,
+      "loss": 1.8041,
+      "step": 387
+    },
+    {
+      "epoch": 0.1893298633702017,
+      "grad_norm": 0.1667649745941162,
+      "learning_rate": 0.0001,
+      "loss": 1.6559,
+      "step": 388
+    },
+    {
+      "epoch": 0.1898178269355888,
+      "grad_norm": 0.17563803493976593,
+      "learning_rate": 0.0001,
+      "loss": 1.7628,
+      "step": 389
+    },
+    {
+      "epoch": 0.19030579050097593,
+      "grad_norm": 0.18215936422348022,
+      "learning_rate": 0.0001,
+      "loss": 1.7359,
+      "step": 390
+    },
+    {
+      "epoch": 0.19079375406636304,
+      "grad_norm": 0.17120331525802612,
+      "learning_rate": 0.0001,
+      "loss": 1.7454,
+      "step": 391
+    },
+    {
+      "epoch": 0.19128171763175017,
+      "grad_norm": 0.17285685241222382,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 392
+    },
+    {
+      "epoch": 0.19176968119713728,
+      "grad_norm": 0.1752183735370636,
+      "learning_rate": 0.0001,
+      "loss": 1.8662,
+      "step": 393
+    },
+    {
+      "epoch": 0.1922576447625244,
+      "grad_norm": 0.17126931250095367,
+      "learning_rate": 0.0001,
+      "loss": 1.6906,
+      "step": 394
+    },
+    {
+      "epoch": 0.19274560832791152,
+      "grad_norm": 0.16653628647327423,
+      "learning_rate": 0.0001,
+      "loss": 1.7104,
+      "step": 395
+    },
+    {
+      "epoch": 0.19323357189329862,
+      "grad_norm": 0.17633773386478424,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 396
+    },
+    {
+      "epoch": 0.19372153545868576,
+      "grad_norm": 0.18310055136680603,
+      "learning_rate": 0.0001,
+      "loss": 1.7858,
+      "step": 397
+    },
+    {
+      "epoch": 0.19420949902407286,
+      "grad_norm": 0.164474755525589,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 398
+    },
+    {
+      "epoch": 0.19469746258946,
+      "grad_norm": 0.17584070563316345,
+      "learning_rate": 0.0001,
+      "loss": 1.7491,
+      "step": 399
+    },
+    {
+      "epoch": 0.1951854261548471,
+      "grad_norm": 0.1765829622745514,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 400
+    },
+    {
+      "epoch": 0.1956733897202342,
+      "grad_norm": 0.16353271901607513,
+      "learning_rate": 0.0001,
+      "loss": 1.7347,
+      "step": 401
+    },
+    {
+      "epoch": 0.19616135328562134,
+      "grad_norm": 0.16928355395793915,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 402
+    },
+    {
+      "epoch": 0.19664931685100845,
+      "grad_norm": 0.17108477652072906,
+      "learning_rate": 0.0001,
+      "loss": 1.6907,
+      "step": 403
+    },
+    {
+      "epoch": 0.19713728041639558,
+      "grad_norm": 0.16809748113155365,
+      "learning_rate": 0.0001,
+      "loss": 1.735,
+      "step": 404
+    },
+    {
+      "epoch": 0.1976252439817827,
+      "grad_norm": 0.18148702383041382,
+      "learning_rate": 0.0001,
+      "loss": 1.7983,
+      "step": 405
+    },
+    {
+      "epoch": 0.19811320754716982,
+      "grad_norm": 0.16235138475894928,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 406
+    },
+    {
+      "epoch": 0.19860117111255693,
+      "grad_norm": 0.16349023580551147,
+      "learning_rate": 0.0001,
+      "loss": 1.7225,
+      "step": 407
+    },
+    {
+      "epoch": 0.19908913467794404,
+      "grad_norm": 0.174223393201828,
+      "learning_rate": 0.0001,
+      "loss": 1.7012,
+      "step": 408
+    },
+    {
+      "epoch": 0.19957709824333117,
+      "grad_norm": 0.16040785610675812,
+      "learning_rate": 0.0001,
+      "loss": 1.6721,
+      "step": 409
+    },
+    {
+      "epoch": 0.20006506180871828,
+      "grad_norm": 0.17907334864139557,
+      "learning_rate": 0.0001,
+      "loss": 1.8158,
+      "step": 410
+    },
+    {
+      "epoch": 0.2005530253741054,
+      "grad_norm": 0.17376290261745453,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 411
+    },
+    {
+      "epoch": 0.20104098893949252,
+      "grad_norm": 0.17140352725982666,
+      "learning_rate": 0.0001,
+      "loss": 1.701,
+      "step": 412
+    },
+    {
+      "epoch": 0.20152895250487965,
+      "grad_norm": 0.17986051738262177,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 413
+    },
+    {
+      "epoch": 0.20201691607026676,
+      "grad_norm": 0.17669373750686646,
+      "learning_rate": 0.0001,
+      "loss": 1.7948,
+      "step": 414
+    },
+    {
+      "epoch": 0.20250487963565386,
+      "grad_norm": 0.1779654622077942,
+      "learning_rate": 0.0001,
+      "loss": 1.7014,
+      "step": 415
+    },
+    {
+      "epoch": 0.202992843201041,
+      "grad_norm": 0.17824238538742065,
+      "learning_rate": 0.0001,
+      "loss": 1.7493,
+      "step": 416
+    },
+    {
+      "epoch": 0.2034808067664281,
+      "grad_norm": 0.17220439016819,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 417
+    },
+    {
+      "epoch": 0.20396877033181524,
+      "grad_norm": 0.1646149456501007,
+      "learning_rate": 0.0001,
+      "loss": 1.7508,
+      "step": 418
+    },
+    {
+      "epoch": 0.20445673389720234,
+      "grad_norm": 0.18231230974197388,
+      "learning_rate": 0.0001,
+      "loss": 1.6986,
+      "step": 419
+    },
+    {
+      "epoch": 0.20494469746258945,
+      "grad_norm": 0.18709278106689453,
+      "learning_rate": 0.0001,
+      "loss": 1.7505,
+      "step": 420
+    },
+    {
+      "epoch": 0.20543266102797658,
+      "grad_norm": 0.17162282764911652,
+      "learning_rate": 0.0001,
+      "loss": 1.7488,
+      "step": 421
+    },
+    {
+      "epoch": 0.2059206245933637,
+      "grad_norm": 0.16716767847537994,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 422
+    },
+    {
+      "epoch": 0.20640858815875082,
+      "grad_norm": 0.18930698931217194,
+      "learning_rate": 0.0001,
+      "loss": 1.7693,
+      "step": 423
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 0.18446506559848785,
+      "learning_rate": 0.0001,
+      "loss": 1.7766,
+      "step": 424
+    },
+    {
+      "epoch": 0.20738451528952506,
+      "grad_norm": 0.1983187198638916,
+      "learning_rate": 0.0001,
+      "loss": 1.7192,
+      "step": 425
+    },
+    {
+      "epoch": 0.20787247885491217,
+      "grad_norm": 0.19457009434700012,
+      "learning_rate": 0.0001,
+      "loss": 1.8895,
+      "step": 426
+    },
+    {
+      "epoch": 0.20836044242029927,
+      "grad_norm": 0.18323922157287598,
+      "learning_rate": 0.0001,
+      "loss": 1.7256,
+      "step": 427
+    },
+    {
+      "epoch": 0.2088484059856864,
+      "grad_norm": 0.204954132437706,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 428
+    },
+    {
+      "epoch": 0.20933636955107351,
+      "grad_norm": 0.1658659428358078,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 429
+    },
+    {
+      "epoch": 0.20982433311646065,
+      "grad_norm": 0.19778519868850708,
+      "learning_rate": 0.0001,
+      "loss": 1.7141,
+      "step": 430
+    },
+    {
+      "epoch": 0.21031229668184775,
+      "grad_norm": 0.17738375067710876,
+      "learning_rate": 0.0001,
+      "loss": 1.6848,
+      "step": 431
+    },
+    {
+      "epoch": 0.21080026024723486,
+      "grad_norm": 0.19395996630191803,
+      "learning_rate": 0.0001,
+      "loss": 1.769,
+      "step": 432
+    },
+    {
+      "epoch": 0.211288223812622,
+      "grad_norm": 0.17526568472385406,
+      "learning_rate": 0.0001,
+      "loss": 1.7879,
+      "step": 433
+    },
+    {
+      "epoch": 0.2117761873780091,
+      "grad_norm": 0.18097303807735443,
+      "learning_rate": 0.0001,
+      "loss": 1.6922,
+      "step": 434
+    },
+    {
+      "epoch": 0.21226415094339623,
+      "grad_norm": 0.18338918685913086,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 435
+    },
+    {
+      "epoch": 0.21275211450878334,
+      "grad_norm": 0.1740289032459259,
+      "learning_rate": 0.0001,
+      "loss": 1.7512,
+      "step": 436
+    },
+    {
+      "epoch": 0.21324007807417047,
+      "grad_norm": 0.1841944307088852,
+      "learning_rate": 0.0001,
+      "loss": 1.7865,
+      "step": 437
+    },
+    {
+      "epoch": 0.21372804163955758,
+      "grad_norm": 0.17511843144893646,
+      "learning_rate": 0.0001,
+      "loss": 1.6842,
+      "step": 438
+    },
+    {
+      "epoch": 0.2142160052049447,
+      "grad_norm": 0.17134588956832886,
+      "learning_rate": 0.0001,
+      "loss": 1.6142,
+      "step": 439
+    },
+    {
+      "epoch": 0.21470396877033182,
+      "grad_norm": 0.16989292204380035,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 440
+    },
+    {
+      "epoch": 0.21519193233571893,
+      "grad_norm": 0.1850852519273758,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 441
+    },
+    {
+      "epoch": 0.21567989590110606,
+      "grad_norm": 0.1822405755519867,
+      "learning_rate": 0.0001,
+      "loss": 1.7404,
+      "step": 442
+    },
+    {
+      "epoch": 0.21616785946649317,
+      "grad_norm": 0.17180569469928741,
+      "learning_rate": 0.0001,
+      "loss": 1.7174,
+      "step": 443
+    },
+    {
+      "epoch": 0.21665582303188027,
+      "grad_norm": 0.18243129551410675,
+      "learning_rate": 0.0001,
+      "loss": 1.7224,
+      "step": 444
+    },
+    {
+      "epoch": 0.2171437865972674,
+      "grad_norm": 0.16769753396511078,
+      "learning_rate": 0.0001,
+      "loss": 1.6866,
+      "step": 445
+    },
+    {
+      "epoch": 0.2176317501626545,
+      "grad_norm": 0.16903330385684967,
+      "learning_rate": 0.0001,
+      "loss": 1.8218,
+      "step": 446
+    },
+    {
+      "epoch": 0.21811971372804165,
+      "grad_norm": 0.20899462699890137,
+      "learning_rate": 0.0001,
+      "loss": 1.7798,
+      "step": 447
+    },
+    {
+      "epoch": 0.21860767729342875,
+      "grad_norm": 0.172098308801651,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 448
+    },
+    {
+      "epoch": 0.2190956408588159,
+      "grad_norm": 0.18460461497306824,
+      "learning_rate": 0.0001,
+      "loss": 1.6764,
+      "step": 449
+    },
+    {
+      "epoch": 0.219583604424203,
+      "grad_norm": 0.18254351615905762,
+      "learning_rate": 0.0001,
+      "loss": 1.718,
+      "step": 450
+    },
+    {
+      "epoch": 0.2200715679895901,
+      "grad_norm": 0.17702056467533112,
+      "learning_rate": 0.0001,
+      "loss": 1.6639,
+      "step": 451
+    },
+    {
+      "epoch": 0.22055953155497723,
+      "grad_norm": 0.1730417013168335,
+      "learning_rate": 0.0001,
+      "loss": 1.6448,
+      "step": 452
+    },
+    {
+      "epoch": 0.22104749512036434,
+      "grad_norm": 0.16833186149597168,
+      "learning_rate": 0.0001,
+      "loss": 1.8317,
+      "step": 453
+    },
+    {
+      "epoch": 0.22153545868575147,
+      "grad_norm": 0.17784419655799866,
+      "learning_rate": 0.0001,
+      "loss": 1.7377,
+      "step": 454
+    },
+    {
+      "epoch": 0.22202342225113858,
+      "grad_norm": 0.17616881430149078,
+      "learning_rate": 0.0001,
+      "loss": 1.7849,
+      "step": 455
+    },
+    {
+      "epoch": 0.2225113858165257,
+      "grad_norm": 0.1755736917257309,
+      "learning_rate": 0.0001,
+      "loss": 1.6351,
+      "step": 456
+    },
+    {
+      "epoch": 0.22299934938191282,
+      "grad_norm": 0.17838868498802185,
+      "learning_rate": 0.0001,
+      "loss": 1.7603,
+      "step": 457
+    },
+    {
+      "epoch": 0.22348731294729993,
+      "grad_norm": 0.18568748235702515,
+      "learning_rate": 0.0001,
+      "loss": 1.7737,
+      "step": 458
+    },
+    {
+      "epoch": 0.22397527651268706,
+      "grad_norm": 0.1767285317182541,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 459
+    },
+    {
+      "epoch": 0.22446324007807417,
+      "grad_norm": 0.16574998199939728,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 460
+    },
+    {
+      "epoch": 0.2249512036434613,
+      "grad_norm": 0.17935383319854736,
+      "learning_rate": 0.0001,
+      "loss": 1.7729,
+      "step": 461
+    },
+    {
+      "epoch": 0.2254391672088484,
+      "grad_norm": 0.18277722597122192,
+      "learning_rate": 0.0001,
+      "loss": 1.8316,
+      "step": 462
+    },
+    {
+      "epoch": 0.2259271307742355,
+      "grad_norm": 0.17854078114032745,
+      "learning_rate": 0.0001,
+      "loss": 1.7725,
+      "step": 463
+    },
+    {
+      "epoch": 0.22641509433962265,
+      "grad_norm": 0.19233126938343048,
+      "learning_rate": 0.0001,
+      "loss": 1.7991,
+      "step": 464
+    },
+    {
+      "epoch": 0.22690305790500975,
+      "grad_norm": 0.17671933770179749,
+      "learning_rate": 0.0001,
+      "loss": 1.8009,
+      "step": 465
+    },
+    {
+      "epoch": 0.2273910214703969,
+      "grad_norm": 0.17113515734672546,
+      "learning_rate": 0.0001,
+      "loss": 1.6337,
+      "step": 466
+    },
+    {
+      "epoch": 0.227878985035784,
+      "grad_norm": 0.17494109272956848,
+      "learning_rate": 0.0001,
+      "loss": 1.6502,
+      "step": 467
+    },
+    {
+      "epoch": 0.2283669486011711,
+      "grad_norm": 0.1647273153066635,
+      "learning_rate": 0.0001,
+      "loss": 1.669,
+      "step": 468
+    },
+    {
+      "epoch": 0.22885491216655823,
+      "grad_norm": 0.18797969818115234,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 469
+    },
+    {
+      "epoch": 0.22934287573194534,
+      "grad_norm": 0.173686683177948,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 470
+    },
+    {
+      "epoch": 0.22983083929733247,
+      "grad_norm": 0.188637837767601,
+      "learning_rate": 0.0001,
+      "loss": 1.771,
+      "step": 471
+    },
+    {
+      "epoch": 0.23031880286271958,
+      "grad_norm": 0.17387168109416962,
+      "learning_rate": 0.0001,
+      "loss": 1.8145,
+      "step": 472
+    },
+    {
+      "epoch": 0.2308067664281067,
+      "grad_norm": 0.1740867644548416,
+      "learning_rate": 0.0001,
+      "loss": 1.7191,
+      "step": 473
+    },
+    {
+      "epoch": 0.23129472999349382,
+      "grad_norm": 0.172586128115654,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 474
+    },
+    {
+      "epoch": 0.23178269355888093,
+      "grad_norm": 0.17204353213310242,
+      "learning_rate": 0.0001,
+      "loss": 1.7588,
+      "step": 475
+    },
+    {
+      "epoch": 0.23227065712426806,
+      "grad_norm": 0.17666882276535034,
+      "learning_rate": 0.0001,
+      "loss": 1.7243,
+      "step": 476
+    },
+    {
+      "epoch": 0.23275862068965517,
+      "grad_norm": 0.16115796566009521,
+      "learning_rate": 0.0001,
+      "loss": 1.6218,
+      "step": 477
+    },
+    {
+      "epoch": 0.2332465842550423,
+      "grad_norm": 0.18960176408290863,
+      "learning_rate": 0.0001,
+      "loss": 1.6263,
+      "step": 478
+    },
+    {
+      "epoch": 0.2337345478204294,
+      "grad_norm": 0.17686745524406433,
+      "learning_rate": 0.0001,
+      "loss": 1.6955,
+      "step": 479
+    },
+    {
+      "epoch": 0.2342225113858165,
+      "grad_norm": 0.1706487238407135,
+      "learning_rate": 0.0001,
+      "loss": 1.7308,
+      "step": 480
+    },
+    {
+      "epoch": 0.23471047495120365,
+      "grad_norm": 0.18416084349155426,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 481
+    },
+    {
+      "epoch": 0.23519843851659075,
+      "grad_norm": 0.17233574390411377,
+      "learning_rate": 0.0001,
+      "loss": 1.7669,
+      "step": 482
+    },
+    {
+      "epoch": 0.23568640208197789,
+      "grad_norm": 0.16941383481025696,
+      "learning_rate": 0.0001,
+      "loss": 1.6749,
+      "step": 483
+    },
+    {
+      "epoch": 0.236174365647365,
+      "grad_norm": 0.1856997162103653,
+      "learning_rate": 0.0001,
+      "loss": 1.7764,
+      "step": 484
+    },
+    {
+      "epoch": 0.23666232921275213,
+      "grad_norm": 0.16516490280628204,
+      "learning_rate": 0.0001,
+      "loss": 1.6636,
+      "step": 485
+    },
+    {
+      "epoch": 0.23715029277813923,
+      "grad_norm": 0.1633715182542801,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 486
+    },
+    {
+      "epoch": 0.23763825634352634,
+      "grad_norm": 0.1868508756160736,
+      "learning_rate": 0.0001,
+      "loss": 1.7358,
+      "step": 487
+    },
+    {
+      "epoch": 0.23812621990891347,
+      "grad_norm": 0.16829773783683777,
+      "learning_rate": 0.0001,
+      "loss": 1.6904,
+      "step": 488
+    },
+    {
+      "epoch": 0.23861418347430058,
+      "grad_norm": 0.17292778193950653,
+      "learning_rate": 0.0001,
+      "loss": 1.8213,
+      "step": 489
+    },
+    {
+      "epoch": 0.2391021470396877,
+      "grad_norm": 0.17256690561771393,
+      "learning_rate": 0.0001,
+      "loss": 1.7326,
+      "step": 490
+    },
+    {
+      "epoch": 0.23959011060507482,
+      "grad_norm": 0.17045579850673676,
+      "learning_rate": 0.0001,
+      "loss": 1.7426,
+      "step": 491
+    },
+    {
+      "epoch": 0.24007807417046195,
+      "grad_norm": 0.18865583837032318,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 492
+    },
+    {
+      "epoch": 0.24056603773584906,
+      "grad_norm": 0.17187798023223877,
+      "learning_rate": 0.0001,
+      "loss": 1.8182,
+      "step": 493
+    },
+    {
+      "epoch": 0.24105400130123616,
+      "grad_norm": 0.17912860214710236,
+      "learning_rate": 0.0001,
+      "loss": 1.7686,
+      "step": 494
+    },
+    {
+      "epoch": 0.2415419648666233,
+      "grad_norm": 0.16963303089141846,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 495
+    },
+    {
+      "epoch": 0.2420299284320104,
+      "grad_norm": 0.1751076728105545,
+      "learning_rate": 0.0001,
+      "loss": 1.7598,
+      "step": 496
+    },
+    {
+      "epoch": 0.24251789199739754,
+      "grad_norm": 0.1700344830751419,
+      "learning_rate": 0.0001,
+      "loss": 1.7228,
+      "step": 497
+    },
+    {
+      "epoch": 0.24300585556278465,
+      "grad_norm": 0.16622239351272583,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 498
+    },
+    {
+      "epoch": 0.24349381912817175,
+      "grad_norm": 0.17331449687480927,
+      "learning_rate": 0.0001,
+      "loss": 1.7248,
+      "step": 499
+    },
+    {
+      "epoch": 0.24398178269355889,
+      "grad_norm": 0.17973113059997559,
+      "learning_rate": 0.0001,
+      "loss": 1.6948,
+      "step": 500
+    },
+    {
+      "epoch": 0.244469746258946,
+      "grad_norm": 0.16177998483181,
+      "learning_rate": 0.0001,
+      "loss": 1.6094,
+      "step": 501
+    },
+    {
+      "epoch": 0.24495770982433313,
+      "grad_norm": 0.16940131783485413,
+      "learning_rate": 0.0001,
+      "loss": 1.7415,
+      "step": 502
+    },
+    {
+      "epoch": 0.24544567338972023,
+      "grad_norm": 0.1713533252477646,
+      "learning_rate": 0.0001,
+      "loss": 1.7259,
+      "step": 503
+    },
+    {
+      "epoch": 0.24593363695510737,
+      "grad_norm": 0.16875503957271576,
+      "learning_rate": 0.0001,
+      "loss": 1.7243,
+      "step": 504
+    },
+    {
+      "epoch": 0.24642160052049447,
+      "grad_norm": 0.18620549142360687,
+      "learning_rate": 0.0001,
+      "loss": 1.8485,
+      "step": 505
+    },
+    {
+      "epoch": 0.24690956408588158,
+      "grad_norm": 0.15947122871875763,
+      "learning_rate": 0.0001,
+      "loss": 1.5898,
+      "step": 506
+    },
+    {
+      "epoch": 0.2473975276512687,
+      "grad_norm": 0.1782512217760086,
+      "learning_rate": 0.0001,
+      "loss": 1.7273,
+      "step": 507
+    },
+    {
+      "epoch": 0.24788549121665582,
+      "grad_norm": 0.1739760786294937,
+      "learning_rate": 0.0001,
+      "loss": 1.6825,
+      "step": 508
+    },
+    {
+      "epoch": 0.24837345478204295,
+      "grad_norm": 0.17914053797721863,
+      "learning_rate": 0.0001,
+      "loss": 1.6759,
+      "step": 509
+    },
+    {
+      "epoch": 0.24886141834743006,
+      "grad_norm": 0.17503274977207184,
+      "learning_rate": 0.0001,
+      "loss": 1.6735,
+      "step": 510
+    },
+    {
+      "epoch": 0.24934938191281716,
+      "grad_norm": 0.16352853178977966,
+      "learning_rate": 0.0001,
+      "loss": 1.7075,
+      "step": 511
+    },
+    {
+      "epoch": 0.2498373454782043,
+      "grad_norm": 0.18776653707027435,
+      "learning_rate": 0.0001,
+      "loss": 1.8144,
+      "step": 512
+    },
+    {
+      "epoch": 0.2503253090435914,
+      "grad_norm": 0.18547508120536804,
+      "learning_rate": 0.0001,
+      "loss": 1.7718,
+      "step": 513
+    },
+    {
+      "epoch": 0.2508132726089785,
+      "grad_norm": 0.1719854176044464,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 514
+    },
+    {
+      "epoch": 0.25130123617436567,
+      "grad_norm": 0.16849137842655182,
+      "learning_rate": 0.0001,
+      "loss": 1.6504,
+      "step": 515
+    },
+    {
+      "epoch": 0.2517891997397528,
+      "grad_norm": 0.17803512513637543,
+      "learning_rate": 0.0001,
+      "loss": 1.717,
+      "step": 516
+    },
+    {
+      "epoch": 0.2522771633051399,
+      "grad_norm": 0.1738758683204651,
+      "learning_rate": 0.0001,
+      "loss": 1.5982,
+      "step": 517
+    },
+    {
+      "epoch": 0.252765126870527,
+      "grad_norm": 0.16416452825069427,
+      "learning_rate": 0.0001,
+      "loss": 1.5725,
+      "step": 518
+    },
+    {
+      "epoch": 0.2532530904359141,
+      "grad_norm": 0.18330414593219757,
+      "learning_rate": 0.0001,
+      "loss": 1.7746,
+      "step": 519
+    },
+    {
+      "epoch": 0.25374105400130126,
+      "grad_norm": 0.16441510617733002,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 520
+    },
+    {
+      "epoch": 0.25422901756668836,
+      "grad_norm": 0.16388535499572754,
+      "learning_rate": 0.0001,
+      "loss": 1.5936,
+      "step": 521
+    },
+    {
+      "epoch": 0.25471698113207547,
+      "grad_norm": 0.17173834145069122,
+      "learning_rate": 0.0001,
+      "loss": 1.7217,
+      "step": 522
+    },
+    {
+      "epoch": 0.2552049446974626,
+      "grad_norm": 0.16886378824710846,
+      "learning_rate": 0.0001,
+      "loss": 1.7559,
+      "step": 523
+    },
+    {
+      "epoch": 0.2556929082628497,
+      "grad_norm": 0.17050696909427643,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 524
+    },
+    {
+      "epoch": 0.25618087182823684,
+      "grad_norm": 0.187490314245224,
+      "learning_rate": 0.0001,
+      "loss": 1.6574,
+      "step": 525
+    },
+    {
+      "epoch": 0.25666883539362395,
+      "grad_norm": 0.17663131654262543,
+      "learning_rate": 0.0001,
+      "loss": 1.7724,
+      "step": 526
+    },
+    {
+      "epoch": 0.25715679895901106,
+      "grad_norm": 0.17597696185112,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 527
+    },
+    {
+      "epoch": 0.25764476252439816,
+      "grad_norm": 0.1790188103914261,
+      "learning_rate": 0.0001,
+      "loss": 1.7481,
+      "step": 528
+    },
+    {
+      "epoch": 0.25813272608978527,
+      "grad_norm": 0.17364263534545898,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 529
+    },
+    {
+      "epoch": 0.25862068965517243,
+      "grad_norm": 0.16965626180171967,
+      "learning_rate": 0.0001,
+      "loss": 1.6726,
+      "step": 530
+    },
+    {
+      "epoch": 0.25910865322055954,
+      "grad_norm": 0.19401530921459198,
+      "learning_rate": 0.0001,
+      "loss": 1.7239,
+      "step": 531
+    },
+    {
+      "epoch": 0.25959661678594664,
+      "grad_norm": 0.17073091864585876,
+      "learning_rate": 0.0001,
+      "loss": 1.6637,
+      "step": 532
+    },
+    {
+      "epoch": 0.26008458035133375,
+      "grad_norm": 0.1667754203081131,
+      "learning_rate": 0.0001,
+      "loss": 1.6555,
+      "step": 533
+    },
+    {
+      "epoch": 0.2605725439167209,
+      "grad_norm": 0.20140361785888672,
+      "learning_rate": 0.0001,
+      "loss": 1.7967,
+      "step": 534
+    },
+    {
+      "epoch": 0.261060507482108,
+      "grad_norm": 0.1800750494003296,
+      "learning_rate": 0.0001,
+      "loss": 1.6326,
+      "step": 535
+    },
+    {
+      "epoch": 0.2615484710474951,
+      "grad_norm": 0.2055366188287735,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 536
+    },
+    {
+      "epoch": 0.26203643461288223,
+      "grad_norm": 0.193061962723732,
+      "learning_rate": 0.0001,
+      "loss": 1.7036,
+      "step": 537
+    },
+    {
+      "epoch": 0.26252439817826934,
+      "grad_norm": 0.17845755815505981,
+      "learning_rate": 0.0001,
+      "loss": 1.6967,
+      "step": 538
+    },
+    {
+      "epoch": 0.2630123617436565,
+      "grad_norm": 0.17126064002513885,
+      "learning_rate": 0.0001,
+      "loss": 1.8002,
+      "step": 539
+    },
+    {
+      "epoch": 0.2635003253090436,
+      "grad_norm": 0.1969064474105835,
+      "learning_rate": 0.0001,
+      "loss": 1.7633,
+      "step": 540
+    },
+    {
+      "epoch": 0.2639882888744307,
+      "grad_norm": 0.19908367097377777,
+      "learning_rate": 0.0001,
+      "loss": 1.8194,
+      "step": 541
+    },
+    {
+      "epoch": 0.2644762524398178,
+      "grad_norm": 0.16265065968036652,
+      "learning_rate": 0.0001,
+      "loss": 1.6726,
+      "step": 542
+    },
+    {
+      "epoch": 0.2649642160052049,
+      "grad_norm": 0.19965846836566925,
+      "learning_rate": 0.0001,
+      "loss": 1.8231,
+      "step": 543
+    },
+    {
+      "epoch": 0.2654521795705921,
+      "grad_norm": 0.16638900339603424,
+      "learning_rate": 0.0001,
+      "loss": 1.655,
+      "step": 544
+    },
+    {
+      "epoch": 0.2659401431359792,
+      "grad_norm": 0.1745545119047165,
+      "learning_rate": 0.0001,
+      "loss": 1.5847,
+      "step": 545
+    },
+    {
+      "epoch": 0.2664281067013663,
+      "grad_norm": 0.17707784473896027,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 546
+    },
+    {
+      "epoch": 0.2669160702667534,
+      "grad_norm": 0.17382854223251343,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 547
+    },
+    {
+      "epoch": 0.2674040338321405,
+      "grad_norm": 0.18207457661628723,
+      "learning_rate": 0.0001,
+      "loss": 1.6712,
+      "step": 548
+    },
+    {
+      "epoch": 0.26789199739752767,
+      "grad_norm": 0.18442131578922272,
+      "learning_rate": 0.0001,
+      "loss": 1.7411,
+      "step": 549
+    },
+    {
+      "epoch": 0.2683799609629148,
+      "grad_norm": 0.17174646258354187,
+      "learning_rate": 0.0001,
+      "loss": 1.6931,
+      "step": 550
+    },
+    {
+      "epoch": 0.2688679245283019,
+      "grad_norm": 0.19472530484199524,
+      "learning_rate": 0.0001,
+      "loss": 1.6551,
+      "step": 551
+    },
+    {
+      "epoch": 0.269355888093689,
+      "grad_norm": 0.1676761358976364,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 552
+    },
+    {
+      "epoch": 0.2698438516590761,
+      "grad_norm": 0.18984854221343994,
+      "learning_rate": 0.0001,
+      "loss": 1.8853,
+      "step": 553
+    },
+    {
+      "epoch": 0.27033181522446326,
+      "grad_norm": 0.18168702721595764,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 554
+    },
+    {
+      "epoch": 0.27081977878985036,
+      "grad_norm": 0.17709583044052124,
+      "learning_rate": 0.0001,
+      "loss": 1.7217,
+      "step": 555
+    },
+    {
+      "epoch": 0.27130774235523747,
+      "grad_norm": 0.18289346992969513,
+      "learning_rate": 0.0001,
+      "loss": 1.7239,
+      "step": 556
+    },
+    {
+      "epoch": 0.2717957059206246,
+      "grad_norm": 0.2056882381439209,
+      "learning_rate": 0.0001,
+      "loss": 1.6916,
+      "step": 557
+    },
+    {
+      "epoch": 0.27228366948601174,
+      "grad_norm": 0.22222530841827393,
+      "learning_rate": 0.0001,
+      "loss": 1.9045,
+      "step": 558
+    },
+    {
+      "epoch": 0.27277163305139884,
+      "grad_norm": 0.1603502780199051,
+      "learning_rate": 0.0001,
+      "loss": 1.5649,
+      "step": 559
+    },
+    {
+      "epoch": 0.27325959661678595,
+      "grad_norm": 0.1759396493434906,
+      "learning_rate": 0.0001,
+      "loss": 1.6566,
+      "step": 560
+    },
+    {
+      "epoch": 0.27374756018217306,
+      "grad_norm": 0.17986807227134705,
+      "learning_rate": 0.0001,
+      "loss": 1.8009,
+      "step": 561
+    },
+    {
+      "epoch": 0.27423552374756016,
+      "grad_norm": 0.17451557517051697,
+      "learning_rate": 0.0001,
+      "loss": 1.7191,
+      "step": 562
+    },
+    {
+      "epoch": 0.2747234873129473,
+      "grad_norm": 0.17183034121990204,
+      "learning_rate": 0.0001,
+      "loss": 1.7648,
+      "step": 563
+    },
+    {
+      "epoch": 0.27521145087833443,
+      "grad_norm": 0.170223668217659,
+      "learning_rate": 0.0001,
+      "loss": 1.7731,
+      "step": 564
+    },
+    {
+      "epoch": 0.27569941444372154,
+      "grad_norm": 0.19552116096019745,
+      "learning_rate": 0.0001,
+      "loss": 1.9056,
+      "step": 565
+    },
+    {
+      "epoch": 0.27618737800910864,
+      "grad_norm": 0.16772179305553436,
+      "learning_rate": 0.0001,
+      "loss": 1.6326,
+      "step": 566
+    },
+    {
+      "epoch": 0.27667534157449575,
+      "grad_norm": 0.18195265531539917,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 567
+    },
+    {
+      "epoch": 0.2771633051398829,
+      "grad_norm": 0.1670740246772766,
+      "learning_rate": 0.0001,
+      "loss": 1.6312,
+      "step": 568
+    },
+    {
+      "epoch": 0.27765126870527,
+      "grad_norm": 0.1744481474161148,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 569
+    },
+    {
+      "epoch": 0.2781392322706571,
+      "grad_norm": 0.16244308650493622,
+      "learning_rate": 0.0001,
+      "loss": 1.6302,
+      "step": 570
+    },
+    {
+      "epoch": 0.27862719583604423,
+      "grad_norm": 0.1672458052635193,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 571
+    },
+    {
+      "epoch": 0.27911515940143133,
+      "grad_norm": 0.17869865894317627,
+      "learning_rate": 0.0001,
+      "loss": 1.8339,
+      "step": 572
+    },
+    {
+      "epoch": 0.2796031229668185,
+      "grad_norm": 0.16552932560443878,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 573
+    },
+    {
+      "epoch": 0.2800910865322056,
+      "grad_norm": 0.17194536328315735,
+      "learning_rate": 0.0001,
+      "loss": 1.7957,
+      "step": 574
+    },
+    {
+      "epoch": 0.2805790500975927,
+      "grad_norm": 0.16269123554229736,
+      "learning_rate": 0.0001,
+      "loss": 1.6373,
+      "step": 575
+    },
+    {
+      "epoch": 0.2810670136629798,
+      "grad_norm": 0.17862126231193542,
+      "learning_rate": 0.0001,
+      "loss": 1.658,
+      "step": 576
+    },
+    {
+      "epoch": 0.281554977228367,
+      "grad_norm": 0.16371159255504608,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 577
+    },
+    {
+      "epoch": 0.2820429407937541,
+      "grad_norm": 0.17550936341285706,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 578
+    },
+    {
+      "epoch": 0.2825309043591412,
+      "grad_norm": 0.1784610152244568,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 579
+    },
+    {
+      "epoch": 0.2830188679245283,
+      "grad_norm": 0.1721472591161728,
+      "learning_rate": 0.0001,
+      "loss": 1.7146,
+      "step": 580
+    },
+    {
+      "epoch": 0.2835068314899154,
+      "grad_norm": 0.1666477918624878,
+      "learning_rate": 0.0001,
+      "loss": 1.6365,
+      "step": 581
+    },
+    {
+      "epoch": 0.28399479505530256,
+      "grad_norm": 0.17378363013267517,
+      "learning_rate": 0.0001,
+      "loss": 1.5784,
+      "step": 582
+    },
+    {
+      "epoch": 0.28448275862068967,
+      "grad_norm": 0.17766043543815613,
+      "learning_rate": 0.0001,
+      "loss": 1.7673,
+      "step": 583
+    },
+    {
+      "epoch": 0.2849707221860768,
+      "grad_norm": 0.17343150079250336,
+      "learning_rate": 0.0001,
+      "loss": 1.7419,
+      "step": 584
+    },
+    {
+      "epoch": 0.2854586857514639,
+      "grad_norm": 0.18182925879955292,
+      "learning_rate": 0.0001,
+      "loss": 1.7579,
+      "step": 585
+    },
+    {
+      "epoch": 0.285946649316851,
+      "grad_norm": 0.17902551591396332,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 586
+    },
+    {
+      "epoch": 0.28643461288223815,
+      "grad_norm": 0.18300925195217133,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 587
+    },
+    {
+      "epoch": 0.28692257644762525,
+      "grad_norm": 0.1679466813802719,
+      "learning_rate": 0.0001,
+      "loss": 1.6519,
+      "step": 588
+    },
+    {
+      "epoch": 0.28741054001301236,
+      "grad_norm": 0.17271865904331207,
+      "learning_rate": 0.0001,
+      "loss": 1.7603,
+      "step": 589
+    },
+    {
+      "epoch": 0.28789850357839947,
+      "grad_norm": 0.1763513684272766,
+      "learning_rate": 0.0001,
+      "loss": 1.6604,
+      "step": 590
+    },
+    {
+      "epoch": 0.2883864671437866,
+      "grad_norm": 0.1821538656949997,
+      "learning_rate": 0.0001,
+      "loss": 1.8052,
+      "step": 591
+    },
+    {
+      "epoch": 0.28887443070917374,
+      "grad_norm": 0.18797627091407776,
+      "learning_rate": 0.0001,
+      "loss": 1.8535,
+      "step": 592
+    },
+    {
+      "epoch": 0.28936239427456084,
+      "grad_norm": 0.19498631358146667,
+      "learning_rate": 0.0001,
+      "loss": 1.8813,
+      "step": 593
+    },
+    {
+      "epoch": 0.28985035783994795,
+      "grad_norm": 0.18044626712799072,
+      "learning_rate": 0.0001,
+      "loss": 1.7757,
+      "step": 594
+    },
+    {
+      "epoch": 0.29033832140533505,
+      "grad_norm": 0.16611842811107635,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 595
+    },
+    {
+      "epoch": 0.29082628497072216,
+      "grad_norm": 0.17572735249996185,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 596
+    },
+    {
+      "epoch": 0.2913142485361093,
+      "grad_norm": 0.17393283545970917,
+      "learning_rate": 0.0001,
+      "loss": 1.7883,
+      "step": 597
+    },
+    {
+      "epoch": 0.2918022121014964,
+      "grad_norm": 0.17930260300636292,
+      "learning_rate": 0.0001,
+      "loss": 1.7834,
+      "step": 598
+    },
+    {
+      "epoch": 0.29229017566688353,
+      "grad_norm": 0.16885998845100403,
+      "learning_rate": 0.0001,
+      "loss": 1.6765,
+      "step": 599
+    },
+    {
+      "epoch": 0.29277813923227064,
+      "grad_norm": 0.17516076564788818,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 600
+    },
+    {
+      "epoch": 0.2932661027976578,
+      "grad_norm": 0.1720881313085556,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 601
+    },
+    {
+      "epoch": 0.2937540663630449,
+      "grad_norm": 0.16959141194820404,
+      "learning_rate": 0.0001,
+      "loss": 1.7788,
+      "step": 602
+    },
+    {
+      "epoch": 0.294242029928432,
+      "grad_norm": 0.1675613969564438,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 603
+    },
+    {
+      "epoch": 0.2947299934938191,
+      "grad_norm": 0.16708707809448242,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 604
+    },
+    {
+      "epoch": 0.2952179570592062,
+      "grad_norm": 0.1769132763147354,
+      "learning_rate": 0.0001,
+      "loss": 1.7038,
+      "step": 605
+    },
+    {
+      "epoch": 0.2957059206245934,
+      "grad_norm": 0.17345738410949707,
+      "learning_rate": 0.0001,
+      "loss": 1.6945,
+      "step": 606
+    },
+    {
+      "epoch": 0.2961938841899805,
+      "grad_norm": 0.17599478363990784,
+      "learning_rate": 0.0001,
+      "loss": 1.7294,
+      "step": 607
+    },
+    {
+      "epoch": 0.2966818477553676,
+      "grad_norm": 0.17908738553524017,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 608
+    },
+    {
+      "epoch": 0.2971698113207547,
+      "grad_norm": 0.16808348894119263,
+      "learning_rate": 0.0001,
+      "loss": 1.6358,
+      "step": 609
+    },
+    {
+      "epoch": 0.2976577748861418,
+      "grad_norm": 0.1823492795228958,
+      "learning_rate": 0.0001,
+      "loss": 1.7362,
+      "step": 610
+    },
+    {
+      "epoch": 0.298145738451529,
+      "grad_norm": 0.17771977186203003,
+      "learning_rate": 0.0001,
+      "loss": 1.6266,
+      "step": 611
+    },
+    {
+      "epoch": 0.2986337020169161,
+      "grad_norm": 0.17940644919872284,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 612
+    },
+    {
+      "epoch": 0.2991216655823032,
+      "grad_norm": 0.18967844545841217,
+      "learning_rate": 0.0001,
+      "loss": 1.6844,
+      "step": 613
+    },
+    {
+      "epoch": 0.2996096291476903,
+      "grad_norm": 0.1788313239812851,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 614
+    },
+    {
+      "epoch": 0.3000975927130774,
+      "grad_norm": 0.19167670607566833,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 615
+    },
+    {
+      "epoch": 0.30058555627846456,
+      "grad_norm": 0.18530650436878204,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 616
+    },
+    {
+      "epoch": 0.30107351984385167,
+      "grad_norm": 0.17250166833400726,
+      "learning_rate": 0.0001,
+      "loss": 1.7999,
+      "step": 617
+    },
+    {
+      "epoch": 0.3015614834092388,
+      "grad_norm": 0.17163299024105072,
+      "learning_rate": 0.0001,
+      "loss": 1.6412,
+      "step": 618
+    },
+    {
+      "epoch": 0.3020494469746259,
+      "grad_norm": 0.17998485267162323,
+      "learning_rate": 0.0001,
+      "loss": 1.7421,
+      "step": 619
+    },
+    {
+      "epoch": 0.302537410540013,
+      "grad_norm": 0.1786029040813446,
+      "learning_rate": 0.0001,
+      "loss": 1.7859,
+      "step": 620
+    },
+    {
+      "epoch": 0.30302537410540015,
+      "grad_norm": 0.1714048534631729,
+      "learning_rate": 0.0001,
+      "loss": 1.6867,
+      "step": 621
+    },
+    {
+      "epoch": 0.30351333767078725,
+      "grad_norm": 0.16846685111522675,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 622
+    },
+    {
+      "epoch": 0.30400130123617436,
+      "grad_norm": 0.16885484755039215,
+      "learning_rate": 0.0001,
+      "loss": 1.6127,
+      "step": 623
+    },
+    {
+      "epoch": 0.30448926480156147,
+      "grad_norm": 0.1702856868505478,
+      "learning_rate": 0.0001,
+      "loss": 1.6751,
+      "step": 624
+    },
+    {
+      "epoch": 0.3049772283669486,
+      "grad_norm": 0.16784608364105225,
+      "learning_rate": 0.0001,
+      "loss": 1.6461,
+      "step": 625
+    },
+    {
+      "epoch": 0.30546519193233573,
+      "grad_norm": 0.17238928377628326,
+      "learning_rate": 0.0001,
+      "loss": 1.7196,
+      "step": 626
+    },
+    {
+      "epoch": 0.30595315549772284,
+      "grad_norm": 0.16515295207500458,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 627
+    },
+    {
+      "epoch": 0.30644111906310995,
+      "grad_norm": 0.1686062067747116,
+      "learning_rate": 0.0001,
+      "loss": 1.7103,
+      "step": 628
+    },
+    {
+      "epoch": 0.30692908262849705,
+      "grad_norm": 0.18762940168380737,
+      "learning_rate": 0.0001,
+      "loss": 1.7952,
+      "step": 629
+    },
+    {
+      "epoch": 0.3074170461938842,
+      "grad_norm": 0.17039833962917328,
+      "learning_rate": 0.0001,
+      "loss": 1.5479,
+      "step": 630
+    },
+    {
+      "epoch": 0.3079050097592713,
+      "grad_norm": 0.18100206553936005,
+      "learning_rate": 0.0001,
+      "loss": 1.9089,
+      "step": 631
+    },
+    {
+      "epoch": 0.3083929733246584,
+      "grad_norm": 0.16924722492694855,
+      "learning_rate": 0.0001,
+      "loss": 1.6359,
+      "step": 632
+    },
+    {
+      "epoch": 0.30888093689004553,
+      "grad_norm": 0.1761018931865692,
+      "learning_rate": 0.0001,
+      "loss": 1.7348,
+      "step": 633
+    },
+    {
+      "epoch": 0.30936890045543264,
+      "grad_norm": 0.16776175796985626,
+      "learning_rate": 0.0001,
+      "loss": 1.6608,
+      "step": 634
+    },
+    {
+      "epoch": 0.3098568640208198,
+      "grad_norm": 0.16289463639259338,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 635
+    },
+    {
+      "epoch": 0.3103448275862069,
+      "grad_norm": 0.16914112865924835,
+      "learning_rate": 0.0001,
+      "loss": 1.7026,
+      "step": 636
+    },
+    {
+      "epoch": 0.310832791151594,
+      "grad_norm": 0.16428320109844208,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 637
+    },
+    {
+      "epoch": 0.3113207547169811,
+      "grad_norm": 0.19079333543777466,
+      "learning_rate": 0.0001,
+      "loss": 1.9146,
+      "step": 638
+    },
+    {
+      "epoch": 0.3118087182823682,
+      "grad_norm": 0.17098505795001984,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 639
+    },
+    {
+      "epoch": 0.3122966818477554,
+      "grad_norm": 0.1739153414964676,
+      "learning_rate": 0.0001,
+      "loss": 1.7048,
+      "step": 640
+    },
+    {
+      "epoch": 0.3127846454131425,
+      "grad_norm": 0.18081118166446686,
+      "learning_rate": 0.0001,
+      "loss": 1.7641,
+      "step": 641
+    },
+    {
+      "epoch": 0.3132726089785296,
+      "grad_norm": 0.16817376017570496,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 642
+    },
+    {
+      "epoch": 0.3137605725439167,
+      "grad_norm": 0.17938318848609924,
+      "learning_rate": 0.0001,
+      "loss": 1.6034,
+      "step": 643
+    },
+    {
+      "epoch": 0.3142485361093038,
+      "grad_norm": 0.16863268613815308,
+      "learning_rate": 0.0001,
+      "loss": 1.6662,
+      "step": 644
+    },
+    {
+      "epoch": 0.314736499674691,
+      "grad_norm": 0.16806314885616302,
+      "learning_rate": 0.0001,
+      "loss": 1.6439,
+      "step": 645
+    },
+    {
+      "epoch": 0.3152244632400781,
+      "grad_norm": 0.1747475564479828,
+      "learning_rate": 0.0001,
+      "loss": 1.737,
+      "step": 646
+    },
+    {
+      "epoch": 0.3157124268054652,
+      "grad_norm": 0.16692426800727844,
+      "learning_rate": 0.0001,
+      "loss": 1.698,
+      "step": 647
+    },
+    {
+      "epoch": 0.3162003903708523,
+      "grad_norm": 0.1674102246761322,
+      "learning_rate": 0.0001,
+      "loss": 1.6721,
+      "step": 648
+    },
+    {
+      "epoch": 0.31668835393623945,
+      "grad_norm": 0.16820529103279114,
+      "learning_rate": 0.0001,
+      "loss": 1.7698,
+      "step": 649
+    },
+    {
+      "epoch": 0.31717631750162656,
+      "grad_norm": 0.16092821955680847,
+      "learning_rate": 0.0001,
+      "loss": 1.5796,
+      "step": 650
+    },
+    {
+      "epoch": 0.31766428106701367,
+      "grad_norm": 0.17172005772590637,
+      "learning_rate": 0.0001,
+      "loss": 1.7628,
+      "step": 651
+    },
+    {
+      "epoch": 0.31815224463240077,
+      "grad_norm": 0.1805141717195511,
+      "learning_rate": 0.0001,
+      "loss": 1.8697,
+      "step": 652
+    },
+    {
+      "epoch": 0.3186402081977879,
+      "grad_norm": 0.16913415491580963,
+      "learning_rate": 0.0001,
+      "loss": 1.7005,
+      "step": 653
+    },
+    {
+      "epoch": 0.31912817176317504,
+      "grad_norm": 0.17932672798633575,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 654
+    },
+    {
+      "epoch": 0.31961613532856215,
+      "grad_norm": 0.16596753895282745,
+      "learning_rate": 0.0001,
+      "loss": 1.7051,
+      "step": 655
+    },
+    {
+      "epoch": 0.32010409889394925,
+      "grad_norm": 0.17863227427005768,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 656
+    },
+    {
+      "epoch": 0.32059206245933636,
+      "grad_norm": 0.17783784866333008,
+      "learning_rate": 0.0001,
+      "loss": 1.7435,
+      "step": 657
+    },
+    {
+      "epoch": 0.32108002602472346,
+      "grad_norm": 0.16385318338871002,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 658
+    },
+    {
+      "epoch": 0.3215679895901106,
+      "grad_norm": 0.19614166021347046,
+      "learning_rate": 0.0001,
+      "loss": 1.7854,
+      "step": 659
+    },
+    {
+      "epoch": 0.32205595315549773,
+      "grad_norm": 0.16909103095531464,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 660
+    },
+    {
+      "epoch": 0.32254391672088484,
+      "grad_norm": 0.21125972270965576,
+      "learning_rate": 0.0001,
+      "loss": 1.8397,
+      "step": 661
+    },
+    {
+      "epoch": 0.32303188028627194,
+      "grad_norm": 0.16905754804611206,
+      "learning_rate": 0.0001,
+      "loss": 1.6874,
+      "step": 662
+    },
+    {
+      "epoch": 0.32351984385165905,
+      "grad_norm": 0.18877671658992767,
+      "learning_rate": 0.0001,
+      "loss": 1.7631,
+      "step": 663
+    },
+    {
+      "epoch": 0.3240078074170462,
+      "grad_norm": 0.17639201879501343,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 664
+    },
+    {
+      "epoch": 0.3244957709824333,
+      "grad_norm": 0.17496541142463684,
+      "learning_rate": 0.0001,
+      "loss": 1.6925,
+      "step": 665
+    },
+    {
+      "epoch": 0.3249837345478204,
+      "grad_norm": 0.17060521245002747,
+      "learning_rate": 0.0001,
+      "loss": 1.6821,
+      "step": 666
+    },
+    {
+      "epoch": 0.32547169811320753,
+      "grad_norm": 0.178414449095726,
+      "learning_rate": 0.0001,
+      "loss": 1.7472,
+      "step": 667
+    },
+    {
+      "epoch": 0.3259596616785947,
+      "grad_norm": 0.17526184022426605,
+      "learning_rate": 0.0001,
+      "loss": 1.7548,
+      "step": 668
+    },
+    {
+      "epoch": 0.3264476252439818,
+      "grad_norm": 0.17008668184280396,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 669
+    },
+    {
+      "epoch": 0.3269355888093689,
+      "grad_norm": 0.18747012317180634,
+      "learning_rate": 0.0001,
+      "loss": 1.7441,
+      "step": 670
+    },
+    {
+      "epoch": 0.327423552374756,
+      "grad_norm": 0.19508373737335205,
+      "learning_rate": 0.0001,
+      "loss": 1.8438,
+      "step": 671
+    },
+    {
+      "epoch": 0.3279115159401431,
+      "grad_norm": 0.17060911655426025,
+      "learning_rate": 0.0001,
+      "loss": 1.7048,
+      "step": 672
+    },
+    {
+      "epoch": 0.3283994795055303,
+      "grad_norm": 0.17706140875816345,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 673
+    },
+    {
+      "epoch": 0.3288874430709174,
+      "grad_norm": 0.19284893572330475,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 674
+    },
+    {
+      "epoch": 0.3293754066363045,
+      "grad_norm": 0.1685590147972107,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 675
+    },
+    {
+      "epoch": 0.3298633702016916,
+      "grad_norm": 0.16774006187915802,
+      "learning_rate": 0.0001,
+      "loss": 1.6817,
+      "step": 676
+    },
+    {
+      "epoch": 0.3303513337670787,
+      "grad_norm": 0.16591045260429382,
+      "learning_rate": 0.0001,
+      "loss": 1.6687,
+      "step": 677
+    },
+    {
+      "epoch": 0.33083929733246586,
+      "grad_norm": 0.17572779953479767,
+      "learning_rate": 0.0001,
+      "loss": 1.6891,
+      "step": 678
+    },
+    {
+      "epoch": 0.33132726089785297,
+      "grad_norm": 0.16528433561325073,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 679
+    },
+    {
+      "epoch": 0.3318152244632401,
+      "grad_norm": 0.17171725630760193,
+      "learning_rate": 0.0001,
+      "loss": 1.6874,
+      "step": 680
+    },
+    {
+      "epoch": 0.3323031880286272,
+      "grad_norm": 0.19209344685077667,
+      "learning_rate": 0.0001,
+      "loss": 1.8317,
+      "step": 681
+    },
+    {
+      "epoch": 0.3327911515940143,
+      "grad_norm": 0.1659175455570221,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 682
+    },
+    {
+      "epoch": 0.33327911515940145,
+      "grad_norm": 0.1880204826593399,
+      "learning_rate": 0.0001,
+      "loss": 1.8037,
+      "step": 683
+    },
+    {
+      "epoch": 0.33376707872478856,
+      "grad_norm": 0.18484802544116974,
+      "learning_rate": 0.0001,
+      "loss": 1.7906,
+      "step": 684
+    },
+    {
+      "epoch": 0.33425504229017566,
+      "grad_norm": 0.19149412214756012,
+      "learning_rate": 0.0001,
+      "loss": 1.7662,
+      "step": 685
+    },
+    {
+      "epoch": 0.33474300585556277,
+      "grad_norm": 0.16718704998493195,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 686
+    },
+    {
+      "epoch": 0.3352309694209499,
+      "grad_norm": 0.18291272222995758,
+      "learning_rate": 0.0001,
+      "loss": 1.756,
+      "step": 687
+    },
+    {
+      "epoch": 0.33571893298633704,
+      "grad_norm": 0.1927364319562912,
+      "learning_rate": 0.0001,
+      "loss": 1.7726,
+      "step": 688
+    },
+    {
+      "epoch": 0.33620689655172414,
+      "grad_norm": 0.19132055342197418,
+      "learning_rate": 0.0001,
+      "loss": 1.7634,
+      "step": 689
+    },
+    {
+      "epoch": 0.33669486011711125,
+      "grad_norm": 0.1806240826845169,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 690
+    },
+    {
+      "epoch": 0.33718282368249836,
+      "grad_norm": 0.1724700927734375,
+      "learning_rate": 0.0001,
+      "loss": 1.6393,
+      "step": 691
+    },
+    {
+      "epoch": 0.3376707872478855,
+      "grad_norm": 0.18998382985591888,
+      "learning_rate": 0.0001,
+      "loss": 1.7594,
+      "step": 692
+    },
+    {
+      "epoch": 0.3381587508132726,
+      "grad_norm": 0.17564703524112701,
+      "learning_rate": 0.0001,
+      "loss": 1.6145,
+      "step": 693
+    },
+    {
+      "epoch": 0.33864671437865973,
+      "grad_norm": 0.17720158398151398,
+      "learning_rate": 0.0001,
+      "loss": 1.7299,
+      "step": 694
+    },
+    {
+      "epoch": 0.33913467794404684,
+      "grad_norm": 0.172332763671875,
+      "learning_rate": 0.0001,
+      "loss": 1.7088,
+      "step": 695
+    },
+    {
+      "epoch": 0.33962264150943394,
+      "grad_norm": 0.1811947077512741,
+      "learning_rate": 0.0001,
+      "loss": 1.8222,
+      "step": 696
+    },
+    {
+      "epoch": 0.3401106050748211,
+      "grad_norm": 0.17585624754428864,
+      "learning_rate": 0.0001,
+      "loss": 1.7157,
+      "step": 697
+    },
+    {
+      "epoch": 0.3405985686402082,
+      "grad_norm": 0.1790352463722229,
+      "learning_rate": 0.0001,
+      "loss": 1.827,
+      "step": 698
+    },
+    {
+      "epoch": 0.3410865322055953,
+      "grad_norm": 0.17808064818382263,
+      "learning_rate": 0.0001,
+      "loss": 1.6412,
+      "step": 699
+    },
+    {
+      "epoch": 0.3415744957709824,
+      "grad_norm": 0.17093808948993683,
+      "learning_rate": 0.0001,
+      "loss": 1.7082,
+      "step": 700
+    },
+    {
+      "epoch": 0.34206245933636953,
+      "grad_norm": 0.1723935753107071,
+      "learning_rate": 0.0001,
+      "loss": 1.7075,
+      "step": 701
+    },
+    {
+      "epoch": 0.3425504229017567,
+      "grad_norm": 0.17479616403579712,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 702
+    },
+    {
+      "epoch": 0.3430383864671438,
+      "grad_norm": 0.1855219006538391,
+      "learning_rate": 0.0001,
+      "loss": 1.8153,
+      "step": 703
+    },
+    {
+      "epoch": 0.3435263500325309,
+      "grad_norm": 0.17290082573890686,
+      "learning_rate": 0.0001,
+      "loss": 1.7041,
+      "step": 704
+    },
+    {
+      "epoch": 0.344014313597918,
+      "grad_norm": 0.1722654551267624,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 705
+    },
+    {
+      "epoch": 0.3445022771633051,
+      "grad_norm": 0.17546425759792328,
+      "learning_rate": 0.0001,
+      "loss": 1.7616,
+      "step": 706
+    },
+    {
+      "epoch": 0.3449902407286923,
+      "grad_norm": 0.17708982527256012,
+      "learning_rate": 0.0001,
+      "loss": 1.7198,
+      "step": 707
+    },
+    {
+      "epoch": 0.3454782042940794,
+      "grad_norm": 0.17742423713207245,
+      "learning_rate": 0.0001,
+      "loss": 1.7147,
+      "step": 708
+    },
+    {
+      "epoch": 0.3459661678594665,
+      "grad_norm": 0.1823413074016571,
+      "learning_rate": 0.0001,
+      "loss": 1.8066,
+      "step": 709
+    },
+    {
+      "epoch": 0.3464541314248536,
+      "grad_norm": 0.16385461390018463,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 710
+    },
+    {
+      "epoch": 0.3469420949902407,
+      "grad_norm": 0.17154817283153534,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 711
+    },
+    {
+      "epoch": 0.34743005855562786,
+      "grad_norm": 0.1650257259607315,
+      "learning_rate": 0.0001,
+      "loss": 1.7626,
+      "step": 712
+    },
+    {
+      "epoch": 0.34791802212101497,
+      "grad_norm": 0.16265569627285004,
+      "learning_rate": 0.0001,
+      "loss": 1.5434,
+      "step": 713
+    },
+    {
+      "epoch": 0.3484059856864021,
+      "grad_norm": 0.16514728963375092,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 714
+    },
+    {
+      "epoch": 0.3488939492517892,
+      "grad_norm": 0.17206943035125732,
+      "learning_rate": 0.0001,
+      "loss": 1.777,
+      "step": 715
+    },
+    {
+      "epoch": 0.34938191281717634,
+      "grad_norm": 0.1647469699382782,
+      "learning_rate": 0.0001,
+      "loss": 1.6269,
+      "step": 716
+    },
+    {
+      "epoch": 0.34986987638256345,
+      "grad_norm": 0.17735053598880768,
+      "learning_rate": 0.0001,
+      "loss": 1.6997,
+      "step": 717
+    },
+    {
+      "epoch": 0.35035783994795056,
+      "grad_norm": 0.16874809563159943,
+      "learning_rate": 0.0001,
+      "loss": 1.6991,
+      "step": 718
+    },
+    {
+      "epoch": 0.35084580351333766,
+      "grad_norm": 0.18406035006046295,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 719
+    },
+    {
+      "epoch": 0.35133376707872477,
+      "grad_norm": 0.17759063839912415,
+      "learning_rate": 0.0001,
+      "loss": 1.7019,
+      "step": 720
+    },
+    {
+      "epoch": 0.35182173064411193,
+      "grad_norm": 0.18785350024700165,
+      "learning_rate": 0.0001,
+      "loss": 1.7654,
+      "step": 721
+    },
+    {
+      "epoch": 0.35230969420949904,
+      "grad_norm": 0.1746816635131836,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 722
+    },
+    {
+      "epoch": 0.35279765777488614,
+      "grad_norm": 0.1674381047487259,
+      "learning_rate": 0.0001,
+      "loss": 1.5949,
+      "step": 723
+    },
+    {
+      "epoch": 0.35328562134027325,
+      "grad_norm": 0.17563143372535706,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 724
+    },
+    {
+      "epoch": 0.35377358490566035,
+      "grad_norm": 0.18411913514137268,
+      "learning_rate": 0.0001,
+      "loss": 1.7224,
+      "step": 725
+    },
+    {
+      "epoch": 0.3542615484710475,
+      "grad_norm": 0.17446587979793549,
+      "learning_rate": 0.0001,
+      "loss": 1.7638,
+      "step": 726
+    },
+    {
+      "epoch": 0.3547495120364346,
+      "grad_norm": 0.17961637675762177,
+      "learning_rate": 0.0001,
+      "loss": 1.7185,
+      "step": 727
+    },
+    {
+      "epoch": 0.35523747560182173,
+      "grad_norm": 0.19032423198223114,
+      "learning_rate": 0.0001,
+      "loss": 1.7122,
+      "step": 728
+    },
+    {
+      "epoch": 0.35572543916720883,
+      "grad_norm": 0.17081114649772644,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 729
+    },
+    {
+      "epoch": 0.35621340273259594,
+      "grad_norm": 0.20095732808113098,
+      "learning_rate": 0.0001,
+      "loss": 1.773,
+      "step": 730
+    },
+    {
+      "epoch": 0.3567013662979831,
+      "grad_norm": 0.16805239021778107,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 731
+    },
+    {
+      "epoch": 0.3571893298633702,
+      "grad_norm": 0.17981626093387604,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 732
+    },
+    {
+      "epoch": 0.3576772934287573,
+      "grad_norm": 0.1920524388551712,
+      "learning_rate": 0.0001,
+      "loss": 1.7259,
+      "step": 733
+    },
+    {
+      "epoch": 0.3581652569941444,
+      "grad_norm": 0.18607400357723236,
+      "learning_rate": 0.0001,
+      "loss": 1.7359,
+      "step": 734
+    },
+    {
+      "epoch": 0.3586532205595316,
+      "grad_norm": 0.19803552329540253,
+      "learning_rate": 0.0001,
+      "loss": 1.6946,
+      "step": 735
+    },
+    {
+      "epoch": 0.3591411841249187,
+      "grad_norm": 0.1769004613161087,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 736
+    },
+    {
+      "epoch": 0.3596291476903058,
+      "grad_norm": 0.1721493899822235,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 737
+    },
+    {
+      "epoch": 0.3601171112556929,
+      "grad_norm": 0.20942774415016174,
+      "learning_rate": 0.0001,
+      "loss": 1.756,
+      "step": 738
+    },
+    {
+      "epoch": 0.36060507482108,
+      "grad_norm": 0.17248250544071198,
+      "learning_rate": 0.0001,
+      "loss": 1.7233,
+      "step": 739
+    },
+    {
+      "epoch": 0.36109303838646717,
+      "grad_norm": 0.19852924346923828,
+      "learning_rate": 0.0001,
+      "loss": 1.7598,
+      "step": 740
+    },
+    {
+      "epoch": 0.3615810019518543,
+      "grad_norm": 0.18484365940093994,
+      "learning_rate": 0.0001,
+      "loss": 1.6987,
+      "step": 741
+    },
+    {
+      "epoch": 0.3620689655172414,
+      "grad_norm": 0.168433278799057,
+      "learning_rate": 0.0001,
+      "loss": 1.7183,
+      "step": 742
+    },
+    {
+      "epoch": 0.3625569290826285,
+      "grad_norm": 0.20165374875068665,
+      "learning_rate": 0.0001,
+      "loss": 1.7116,
+      "step": 743
+    },
+    {
+      "epoch": 0.3630448926480156,
+      "grad_norm": 0.1742124855518341,
+      "learning_rate": 0.0001,
+      "loss": 1.6578,
+      "step": 744
+    },
+    {
+      "epoch": 0.36353285621340276,
+      "grad_norm": 0.17456160485744476,
+      "learning_rate": 0.0001,
+      "loss": 1.7331,
+      "step": 745
+    },
+    {
+      "epoch": 0.36402081977878986,
+      "grad_norm": 0.19454042613506317,
+      "learning_rate": 0.0001,
+      "loss": 1.7565,
+      "step": 746
+    },
+    {
+      "epoch": 0.36450878334417697,
+      "grad_norm": 0.17689649760723114,
+      "learning_rate": 0.0001,
+      "loss": 1.6662,
+      "step": 747
+    },
+    {
+      "epoch": 0.3649967469095641,
+      "grad_norm": 0.17777618765830994,
+      "learning_rate": 0.0001,
+      "loss": 1.698,
+      "step": 748
+    },
+    {
+      "epoch": 0.3654847104749512,
+      "grad_norm": 0.20195430517196655,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 749
+    },
+    {
+      "epoch": 0.36597267404033834,
+      "grad_norm": 0.1702609807252884,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 750
+    },
+    {
+      "epoch": 0.36646063760572545,
+      "grad_norm": 0.17847135663032532,
+      "learning_rate": 0.0001,
+      "loss": 1.7025,
+      "step": 751
+    },
+    {
+      "epoch": 0.36694860117111255,
+      "grad_norm": 0.20334745943546295,
+      "learning_rate": 0.0001,
+      "loss": 1.7135,
+      "step": 752
+    },
+    {
+      "epoch": 0.36743656473649966,
+      "grad_norm": 0.1718234270811081,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 753
+    },
+    {
+      "epoch": 0.36792452830188677,
+      "grad_norm": 0.1919901967048645,
+      "learning_rate": 0.0001,
+      "loss": 1.7126,
+      "step": 754
+    },
+    {
+      "epoch": 0.36841249186727393,
+      "grad_norm": 0.20147548615932465,
+      "learning_rate": 0.0001,
+      "loss": 1.7644,
+      "step": 755
+    },
+    {
+      "epoch": 0.36890045543266103,
+      "grad_norm": 0.1740388721227646,
+      "learning_rate": 0.0001,
+      "loss": 1.7818,
+      "step": 756
+    },
+    {
+      "epoch": 0.36938841899804814,
+      "grad_norm": 0.19271323084831238,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 757
+    },
+    {
+      "epoch": 0.36987638256343525,
+      "grad_norm": 0.17856548726558685,
+      "learning_rate": 0.0001,
+      "loss": 1.685,
+      "step": 758
+    },
+    {
+      "epoch": 0.3703643461288224,
+      "grad_norm": 0.17026090621948242,
+      "learning_rate": 0.0001,
+      "loss": 1.5788,
+      "step": 759
+    },
+    {
+      "epoch": 0.3708523096942095,
+      "grad_norm": 0.1912890523672104,
+      "learning_rate": 0.0001,
+      "loss": 1.7173,
+      "step": 760
+    },
+    {
+      "epoch": 0.3713402732595966,
+      "grad_norm": 0.1698838770389557,
+      "learning_rate": 0.0001,
+      "loss": 1.7024,
+      "step": 761
+    },
+    {
+      "epoch": 0.3718282368249837,
+      "grad_norm": 0.16909971833229065,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 762
+    },
+    {
+      "epoch": 0.37231620039037083,
+      "grad_norm": 0.18271504342556,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 763
+    },
+    {
+      "epoch": 0.372804163955758,
+      "grad_norm": 0.17217294871807098,
+      "learning_rate": 0.0001,
+      "loss": 1.6615,
+      "step": 764
+    },
+    {
+      "epoch": 0.3732921275211451,
+      "grad_norm": 0.15897059440612793,
+      "learning_rate": 0.0001,
+      "loss": 1.6629,
+      "step": 765
+    },
+    {
+      "epoch": 0.3737800910865322,
+      "grad_norm": 0.19164003431797028,
+      "learning_rate": 0.0001,
+      "loss": 1.655,
+      "step": 766
+    },
+    {
+      "epoch": 0.3742680546519193,
+      "grad_norm": 0.20361578464508057,
+      "learning_rate": 0.0001,
+      "loss": 1.8391,
+      "step": 767
+    },
+    {
+      "epoch": 0.3747560182173064,
+      "grad_norm": 0.17985540628433228,
+      "learning_rate": 0.0001,
+      "loss": 1.7719,
+      "step": 768
+    },
+    {
+      "epoch": 0.3752439817826936,
+      "grad_norm": 0.2011963576078415,
+      "learning_rate": 0.0001,
+      "loss": 1.6659,
+      "step": 769
+    },
+    {
+      "epoch": 0.3757319453480807,
+      "grad_norm": 0.16523408889770508,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 770
+    },
+    {
+      "epoch": 0.3762199089134678,
+      "grad_norm": 0.17251962423324585,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 771
+    },
+    {
+      "epoch": 0.3767078724788549,
+      "grad_norm": 0.1810590624809265,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 772
+    },
+    {
+      "epoch": 0.377195836044242,
+      "grad_norm": 0.17502832412719727,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 773
+    },
+    {
+      "epoch": 0.37768379960962917,
+      "grad_norm": 0.1724274754524231,
+      "learning_rate": 0.0001,
+      "loss": 1.6256,
+      "step": 774
+    },
+    {
+      "epoch": 0.3781717631750163,
+      "grad_norm": 0.17217612266540527,
+      "learning_rate": 0.0001,
+      "loss": 1.6476,
+      "step": 775
+    },
+    {
+      "epoch": 0.3786597267404034,
+      "grad_norm": 0.18240110576152802,
+      "learning_rate": 0.0001,
+      "loss": 1.7655,
+      "step": 776
+    },
+    {
+      "epoch": 0.3791476903057905,
+      "grad_norm": 0.18582911789417267,
+      "learning_rate": 0.0001,
+      "loss": 1.6939,
+      "step": 777
+    },
+    {
+      "epoch": 0.3796356538711776,
+      "grad_norm": 0.19335263967514038,
+      "learning_rate": 0.0001,
+      "loss": 1.7682,
+      "step": 778
+    },
+    {
+      "epoch": 0.38012361743656475,
+      "grad_norm": 0.17951013147830963,
+      "learning_rate": 0.0001,
+      "loss": 1.7146,
+      "step": 779
+    },
+    {
+      "epoch": 0.38061158100195186,
+      "grad_norm": 0.16441676020622253,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 780
+    },
+    {
+      "epoch": 0.38109954456733897,
+      "grad_norm": 0.1792975813150406,
+      "learning_rate": 0.0001,
+      "loss": 1.8009,
+      "step": 781
+    },
+    {
+      "epoch": 0.38158750813272607,
+      "grad_norm": 0.1687975525856018,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 782
+    },
+    {
+      "epoch": 0.38207547169811323,
+      "grad_norm": 0.18872180581092834,
+      "learning_rate": 0.0001,
+      "loss": 1.7103,
+      "step": 783
+    },
+    {
+      "epoch": 0.38256343526350034,
+      "grad_norm": 0.17432475090026855,
+      "learning_rate": 0.0001,
+      "loss": 1.761,
+      "step": 784
+    },
+    {
+      "epoch": 0.38305139882888745,
+      "grad_norm": 0.1807851493358612,
+      "learning_rate": 0.0001,
+      "loss": 1.7931,
+      "step": 785
+    },
+    {
+      "epoch": 0.38353936239427455,
+      "grad_norm": 0.16983912885189056,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 786
+    },
+    {
+      "epoch": 0.38402732595966166,
+      "grad_norm": 0.17184971272945404,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 787
+    },
+    {
+      "epoch": 0.3845152895250488,
+      "grad_norm": 0.17927592992782593,
+      "learning_rate": 0.0001,
+      "loss": 1.7892,
+      "step": 788
+    },
+    {
+      "epoch": 0.3850032530904359,
+      "grad_norm": 0.16639713943004608,
+      "learning_rate": 0.0001,
+      "loss": 1.6786,
+      "step": 789
+    },
+    {
+      "epoch": 0.38549121665582303,
+      "grad_norm": 0.18183507025241852,
+      "learning_rate": 0.0001,
+      "loss": 1.698,
+      "step": 790
+    },
+    {
+      "epoch": 0.38597918022121014,
+      "grad_norm": 0.17353712022304535,
+      "learning_rate": 0.0001,
+      "loss": 1.7949,
+      "step": 791
+    },
+    {
+      "epoch": 0.38646714378659724,
+      "grad_norm": 0.17486366629600525,
+      "learning_rate": 0.0001,
+      "loss": 1.7142,
+      "step": 792
+    },
+    {
+      "epoch": 0.3869551073519844,
+      "grad_norm": 0.1796843558549881,
+      "learning_rate": 0.0001,
+      "loss": 1.7483,
+      "step": 793
+    },
+    {
+      "epoch": 0.3874430709173715,
+      "grad_norm": 0.17541739344596863,
+      "learning_rate": 0.0001,
+      "loss": 1.7432,
+      "step": 794
+    },
+    {
+      "epoch": 0.3879310344827586,
+      "grad_norm": 0.16648191213607788,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 795
+    },
+    {
+      "epoch": 0.3884189980481457,
+      "grad_norm": 0.18359704315662384,
+      "learning_rate": 0.0001,
+      "loss": 1.7265,
+      "step": 796
+    },
+    {
+      "epoch": 0.38890696161353283,
+      "grad_norm": 0.1687793731689453,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 797
+    },
+    {
+      "epoch": 0.38939492517892,
+      "grad_norm": 0.17722128331661224,
+      "learning_rate": 0.0001,
+      "loss": 1.7007,
+      "step": 798
+    },
+    {
+      "epoch": 0.3898828887443071,
+      "grad_norm": 0.1933368444442749,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 799
+    },
+    {
+      "epoch": 0.3903708523096942,
+      "grad_norm": 0.16960865259170532,
+      "learning_rate": 0.0001,
+      "loss": 1.6947,
+      "step": 800
+    },
+    {
+      "epoch": 0.3908588158750813,
+      "grad_norm": 0.1767859160900116,
+      "learning_rate": 0.0001,
+      "loss": 1.7347,
+      "step": 801
+    },
+    {
+      "epoch": 0.3913467794404684,
+      "grad_norm": 0.18150225281715393,
+      "learning_rate": 0.0001,
+      "loss": 1.7162,
+      "step": 802
+    },
+    {
+      "epoch": 0.3918347430058556,
+      "grad_norm": 0.18507182598114014,
+      "learning_rate": 0.0001,
+      "loss": 1.7678,
+      "step": 803
+    },
+    {
+      "epoch": 0.3923227065712427,
+      "grad_norm": 0.20872928202152252,
+      "learning_rate": 0.0001,
+      "loss": 1.8185,
+      "step": 804
+    },
+    {
+      "epoch": 0.3928106701366298,
+      "grad_norm": 0.17374268174171448,
+      "learning_rate": 0.0001,
+      "loss": 1.7479,
+      "step": 805
+    },
+    {
+      "epoch": 0.3932986337020169,
+      "grad_norm": 0.18323782086372375,
+      "learning_rate": 0.0001,
+      "loss": 1.7401,
+      "step": 806
+    },
+    {
+      "epoch": 0.39378659726740406,
+      "grad_norm": 0.17181357741355896,
+      "learning_rate": 0.0001,
+      "loss": 1.6796,
+      "step": 807
+    },
+    {
+      "epoch": 0.39427456083279117,
+      "grad_norm": 0.16647014021873474,
+      "learning_rate": 0.0001,
+      "loss": 1.691,
+      "step": 808
+    },
+    {
+      "epoch": 0.39476252439817827,
+      "grad_norm": 0.17472054064273834,
+      "learning_rate": 0.0001,
+      "loss": 1.7107,
+      "step": 809
+    },
+    {
+      "epoch": 0.3952504879635654,
+      "grad_norm": 0.25710731744766235,
+      "learning_rate": 0.0001,
+      "loss": 1.6238,
+      "step": 810
+    },
+    {
+      "epoch": 0.3957384515289525,
+      "grad_norm": 0.1753971129655838,
+      "learning_rate": 0.0001,
+      "loss": 1.6922,
+      "step": 811
+    },
+    {
+      "epoch": 0.39622641509433965,
+      "grad_norm": 0.17090165615081787,
+      "learning_rate": 0.0001,
+      "loss": 1.7708,
+      "step": 812
+    },
+    {
+      "epoch": 0.39671437865972675,
+      "grad_norm": 0.17664147913455963,
+      "learning_rate": 0.0001,
+      "loss": 1.7075,
+      "step": 813
+    },
+    {
+      "epoch": 0.39720234222511386,
+      "grad_norm": 0.1785152405500412,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 814
+    },
+    {
+      "epoch": 0.39769030579050096,
+      "grad_norm": 0.1745617389678955,
+      "learning_rate": 0.0001,
+      "loss": 1.6949,
+      "step": 815
+    },
+    {
+      "epoch": 0.39817826935588807,
+      "grad_norm": 0.17589262127876282,
+      "learning_rate": 0.0001,
+      "loss": 1.6584,
+      "step": 816
+    },
+    {
+      "epoch": 0.39866623292127523,
+      "grad_norm": 0.18425175547599792,
+      "learning_rate": 0.0001,
+      "loss": 1.7803,
+      "step": 817
+    },
+    {
+      "epoch": 0.39915419648666234,
+      "grad_norm": 0.1793275624513626,
+      "learning_rate": 0.0001,
+      "loss": 1.8079,
+      "step": 818
+    },
+    {
+      "epoch": 0.39964216005204944,
+      "grad_norm": 0.17202021181583405,
+      "learning_rate": 0.0001,
+      "loss": 1.6261,
+      "step": 819
+    },
+    {
+      "epoch": 0.40013012361743655,
+      "grad_norm": 0.1752047836780548,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 820
+    },
+    {
+      "epoch": 0.40061808718282366,
+      "grad_norm": 0.18005704879760742,
+      "learning_rate": 0.0001,
+      "loss": 1.7306,
+      "step": 821
+    },
+    {
+      "epoch": 0.4011060507482108,
+      "grad_norm": 0.17903628945350647,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 822
+    },
+    {
+      "epoch": 0.4015940143135979,
+      "grad_norm": 0.16842132806777954,
+      "learning_rate": 0.0001,
+      "loss": 1.6413,
+      "step": 823
+    },
+    {
+      "epoch": 0.40208197787898503,
+      "grad_norm": 0.18366946280002594,
+      "learning_rate": 0.0001,
+      "loss": 1.6895,
+      "step": 824
+    },
+    {
+      "epoch": 0.40256994144437214,
+      "grad_norm": 0.18511328101158142,
+      "learning_rate": 0.0001,
+      "loss": 1.538,
+      "step": 825
+    },
+    {
+      "epoch": 0.4030579050097593,
+      "grad_norm": 0.1877312809228897,
+      "learning_rate": 0.0001,
+      "loss": 1.7592,
+      "step": 826
+    },
+    {
+      "epoch": 0.4035458685751464,
+      "grad_norm": 0.17125628888607025,
+      "learning_rate": 0.0001,
+      "loss": 1.7051,
+      "step": 827
+    },
+    {
+      "epoch": 0.4040338321405335,
+      "grad_norm": 0.16580699384212494,
+      "learning_rate": 0.0001,
+      "loss": 1.6126,
+      "step": 828
+    },
+    {
+      "epoch": 0.4045217957059206,
+      "grad_norm": 0.1665370762348175,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 829
+    },
+    {
+      "epoch": 0.4050097592713077,
+      "grad_norm": 0.17047293484210968,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 830
+    },
+    {
+      "epoch": 0.4054977228366949,
+      "grad_norm": 0.18139976263046265,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 831
+    },
+    {
+      "epoch": 0.405985686402082,
+      "grad_norm": 0.16962595283985138,
+      "learning_rate": 0.0001,
+      "loss": 1.6657,
+      "step": 832
+    },
+    {
+      "epoch": 0.4064736499674691,
+      "grad_norm": 0.17353881895542145,
+      "learning_rate": 0.0001,
+      "loss": 1.7811,
+      "step": 833
+    },
+    {
+      "epoch": 0.4069616135328562,
+      "grad_norm": 0.16227243840694427,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 834
+    },
+    {
+      "epoch": 0.4074495770982433,
+      "grad_norm": 0.17133179306983948,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 835
+    },
+    {
+      "epoch": 0.40793754066363047,
+      "grad_norm": 0.16746650636196136,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 836
+    },
+    {
+      "epoch": 0.4084255042290176,
+      "grad_norm": 0.1883692890405655,
+      "learning_rate": 0.0001,
+      "loss": 1.7999,
+      "step": 837
+    },
+    {
+      "epoch": 0.4089134677944047,
+      "grad_norm": 0.18568891286849976,
+      "learning_rate": 0.0001,
+      "loss": 1.738,
+      "step": 838
+    },
+    {
+      "epoch": 0.4094014313597918,
+      "grad_norm": 0.17253141105175018,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 839
+    },
+    {
+      "epoch": 0.4098893949251789,
+      "grad_norm": 0.17330259084701538,
+      "learning_rate": 0.0001,
+      "loss": 1.7362,
+      "step": 840
+    },
+    {
+      "epoch": 0.41037735849056606,
+      "grad_norm": 0.16686461865901947,
+      "learning_rate": 0.0001,
+      "loss": 1.5303,
+      "step": 841
+    },
+    {
+      "epoch": 0.41086532205595316,
+      "grad_norm": 0.17059090733528137,
+      "learning_rate": 0.0001,
+      "loss": 1.6638,
+      "step": 842
+    },
+    {
+      "epoch": 0.41135328562134027,
+      "grad_norm": 0.180087611079216,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 843
+    },
+    {
+      "epoch": 0.4118412491867274,
+      "grad_norm": 0.18984520435333252,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 844
+    },
+    {
+      "epoch": 0.4123292127521145,
+      "grad_norm": 0.17229309678077698,
+      "learning_rate": 0.0001,
+      "loss": 1.7177,
+      "step": 845
+    },
+    {
+      "epoch": 0.41281717631750164,
+      "grad_norm": 0.18390652537345886,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 846
+    },
+    {
+      "epoch": 0.41330513988288875,
+      "grad_norm": 0.17387309670448303,
+      "learning_rate": 0.0001,
+      "loss": 1.7527,
+      "step": 847
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 0.18207229673862457,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 848
+    },
+    {
+      "epoch": 0.41428106701366296,
+      "grad_norm": 0.16579647362232208,
+      "learning_rate": 0.0001,
+      "loss": 1.6774,
+      "step": 849
+    },
+    {
+      "epoch": 0.4147690305790501,
+      "grad_norm": 0.17914260923862457,
+      "learning_rate": 0.0001,
+      "loss": 1.6687,
+      "step": 850
+    },
+    {
+      "epoch": 0.41525699414443723,
+      "grad_norm": 0.17592263221740723,
+      "learning_rate": 0.0001,
+      "loss": 1.7046,
+      "step": 851
+    },
+    {
+      "epoch": 0.41574495770982434,
+      "grad_norm": 0.17433282732963562,
+      "learning_rate": 0.0001,
+      "loss": 1.6609,
+      "step": 852
+    },
+    {
+      "epoch": 0.41623292127521144,
+      "grad_norm": 0.17256763577461243,
+      "learning_rate": 0.0001,
+      "loss": 1.7336,
+      "step": 853
+    },
+    {
+      "epoch": 0.41672088484059855,
+      "grad_norm": 0.1765652596950531,
+      "learning_rate": 0.0001,
+      "loss": 1.7198,
+      "step": 854
+    },
+    {
+      "epoch": 0.4172088484059857,
+      "grad_norm": 0.16835534572601318,
+      "learning_rate": 0.0001,
+      "loss": 1.7543,
+      "step": 855
+    },
+    {
+      "epoch": 0.4176968119713728,
+      "grad_norm": 0.1699194461107254,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 856
+    },
+    {
+      "epoch": 0.4181847755367599,
+      "grad_norm": 0.16849122941493988,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 857
+    },
+    {
+      "epoch": 0.41867273910214703,
+      "grad_norm": 0.17524921894073486,
+      "learning_rate": 0.0001,
+      "loss": 1.7692,
+      "step": 858
+    },
+    {
+      "epoch": 0.41916070266753414,
+      "grad_norm": 0.17228025197982788,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 859
+    },
+    {
+      "epoch": 0.4196486662329213,
+      "grad_norm": 0.16608983278274536,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 860
+    },
+    {
+      "epoch": 0.4201366297983084,
+      "grad_norm": 0.17376263439655304,
+      "learning_rate": 0.0001,
+      "loss": 1.636,
+      "step": 861
+    },
+    {
+      "epoch": 0.4206245933636955,
+      "grad_norm": 0.1751623898744583,
+      "learning_rate": 0.0001,
+      "loss": 1.7697,
+      "step": 862
+    },
+    {
+      "epoch": 0.4211125569290826,
+      "grad_norm": 0.17934104800224304,
+      "learning_rate": 0.0001,
+      "loss": 1.732,
+      "step": 863
+    },
+    {
+      "epoch": 0.4216005204944697,
+      "grad_norm": 0.18164652585983276,
+      "learning_rate": 0.0001,
+      "loss": 1.5695,
+      "step": 864
+    },
+    {
+      "epoch": 0.4220884840598569,
+      "grad_norm": 0.18023453652858734,
+      "learning_rate": 0.0001,
+      "loss": 1.6828,
+      "step": 865
+    },
+    {
+      "epoch": 0.422576447625244,
+      "grad_norm": 0.17686617374420166,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 866
+    },
+    {
+      "epoch": 0.4230644111906311,
+      "grad_norm": 0.16582423448562622,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 867
+    },
+    {
+      "epoch": 0.4235523747560182,
+      "grad_norm": 0.17170484364032745,
+      "learning_rate": 0.0001,
+      "loss": 1.6991,
+      "step": 868
+    },
+    {
+      "epoch": 0.4240403383214053,
+      "grad_norm": 0.1795840859413147,
+      "learning_rate": 0.0001,
+      "loss": 1.7438,
+      "step": 869
+    },
+    {
+      "epoch": 0.42452830188679247,
+      "grad_norm": 0.1778561770915985,
+      "learning_rate": 0.0001,
+      "loss": 1.7729,
+      "step": 870
+    },
+    {
+      "epoch": 0.4250162654521796,
+      "grad_norm": 0.1834580898284912,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 871
+    },
+    {
+      "epoch": 0.4255042290175667,
+      "grad_norm": 0.17521165311336517,
+      "learning_rate": 0.0001,
+      "loss": 1.6526,
+      "step": 872
+    },
+    {
+      "epoch": 0.4259921925829538,
+      "grad_norm": 0.16697698831558228,
+      "learning_rate": 0.0001,
+      "loss": 1.721,
+      "step": 873
+    },
+    {
+      "epoch": 0.42648015614834095,
+      "grad_norm": 0.17600980401039124,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 874
+    },
+    {
+      "epoch": 0.42696811971372806,
+      "grad_norm": 0.18683357536792755,
+      "learning_rate": 0.0001,
+      "loss": 1.6892,
+      "step": 875
+    },
+    {
+      "epoch": 0.42745608327911516,
+      "grad_norm": 0.1782413125038147,
+      "learning_rate": 0.0001,
+      "loss": 1.7943,
+      "step": 876
+    },
+    {
+      "epoch": 0.42794404684450227,
+      "grad_norm": 0.17114804685115814,
+      "learning_rate": 0.0001,
+      "loss": 1.6939,
+      "step": 877
+    },
+    {
+      "epoch": 0.4284320104098894,
+      "grad_norm": 0.1772536039352417,
+      "learning_rate": 0.0001,
+      "loss": 1.8404,
+      "step": 878
+    },
+    {
+      "epoch": 0.42891997397527654,
+      "grad_norm": 0.1731899231672287,
+      "learning_rate": 0.0001,
+      "loss": 1.6943,
+      "step": 879
+    },
+    {
+      "epoch": 0.42940793754066364,
+      "grad_norm": 0.18309330940246582,
+      "learning_rate": 0.0001,
+      "loss": 1.7519,
+      "step": 880
+    },
+    {
+      "epoch": 0.42989590110605075,
+      "grad_norm": 0.1829816848039627,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 881
+    },
+    {
+      "epoch": 0.43038386467143785,
+      "grad_norm": 0.17777863144874573,
+      "learning_rate": 0.0001,
+      "loss": 1.7048,
+      "step": 882
+    },
+    {
+      "epoch": 0.43087182823682496,
+      "grad_norm": 0.16940666735172272,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 883
+    },
+    {
+      "epoch": 0.4313597918022121,
+      "grad_norm": 0.18435153365135193,
+      "learning_rate": 0.0001,
+      "loss": 1.7757,
+      "step": 884
+    },
+    {
+      "epoch": 0.43184775536759923,
+      "grad_norm": 0.17758092284202576,
+      "learning_rate": 0.0001,
+      "loss": 1.6591,
+      "step": 885
+    },
+    {
+      "epoch": 0.43233571893298633,
+      "grad_norm": 0.18258903920650482,
+      "learning_rate": 0.0001,
+      "loss": 1.7081,
+      "step": 886
+    },
+    {
+      "epoch": 0.43282368249837344,
+      "grad_norm": 0.1651601642370224,
+      "learning_rate": 0.0001,
+      "loss": 1.5936,
+      "step": 887
+    },
+    {
+      "epoch": 0.43331164606376055,
+      "grad_norm": 0.16870960593223572,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 888
+    },
+    {
+      "epoch": 0.4337996096291477,
+      "grad_norm": 0.17348608374595642,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 889
+    },
+    {
+      "epoch": 0.4342875731945348,
+      "grad_norm": 0.18016240000724792,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 890
+    },
+    {
+      "epoch": 0.4347755367599219,
+      "grad_norm": 0.1777016818523407,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 891
+    },
+    {
+      "epoch": 0.435263500325309,
+      "grad_norm": 0.17680144309997559,
+      "learning_rate": 0.0001,
+      "loss": 1.6833,
+      "step": 892
+    },
+    {
+      "epoch": 0.4357514638906962,
+      "grad_norm": 0.17899180948734283,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 893
+    },
+    {
+      "epoch": 0.4362394274560833,
+      "grad_norm": 0.16206824779510498,
+      "learning_rate": 0.0001,
+      "loss": 1.6499,
+      "step": 894
+    },
+    {
+      "epoch": 0.4367273910214704,
+      "grad_norm": 0.19201014935970306,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 895
+    },
+    {
+      "epoch": 0.4372153545868575,
+      "grad_norm": 0.1658422350883484,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 896
+    },
+    {
+      "epoch": 0.4377033181522446,
+      "grad_norm": 0.20035921037197113,
+      "learning_rate": 0.0001,
+      "loss": 1.8142,
+      "step": 897
+    },
+    {
+      "epoch": 0.4381912817176318,
+      "grad_norm": 0.17217601835727692,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 898
+    },
+    {
+      "epoch": 0.4386792452830189,
+      "grad_norm": 0.19714410603046417,
+      "learning_rate": 0.0001,
+      "loss": 1.7104,
+      "step": 899
+    },
+    {
+      "epoch": 0.439167208848406,
+      "grad_norm": 0.17584411799907684,
+      "learning_rate": 0.0001,
+      "loss": 1.6754,
+      "step": 900
+    },
+    {
+      "epoch": 0.00048796356538711777,
+      "grad_norm": 0.25160425901412964,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 901
+    },
+    {
+      "epoch": 0.0009759271307742355,
+      "grad_norm": 0.22162112593650818,
+      "learning_rate": 0.0001,
+      "loss": 1.8092,
+      "step": 902
+    },
+    {
+      "epoch": 0.0014638906961613532,
+      "grad_norm": 0.1778949797153473,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 903
+    },
+    {
+      "epoch": 0.001951854261548471,
+      "grad_norm": 0.21969731152057648,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 904
+    },
+    {
+      "epoch": 0.002439817826935589,
+      "grad_norm": 0.20550690591335297,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 905
+    },
+    {
+      "epoch": 0.0029277813923227064,
+      "grad_norm": 0.18702560663223267,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 906
+    },
+    {
+      "epoch": 0.0034157449577098243,
+      "grad_norm": 0.19987599551677704,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 907
+    },
+    {
+      "epoch": 0.003903708523096942,
+      "grad_norm": 0.19325925409793854,
+      "learning_rate": 0.0001,
+      "loss": 1.5776,
+      "step": 908
+    },
+    {
+      "epoch": 0.00439167208848406,
+      "grad_norm": 0.17455409467220306,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 909
+    },
+    {
+      "epoch": 0.004879635653871178,
+      "grad_norm": 0.18825402855873108,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 910
+    },
+    {
+      "epoch": 0.005367599219258295,
+      "grad_norm": 0.18928749859333038,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 911
+    },
+    {
+      "epoch": 0.005855562784645413,
+      "grad_norm": 0.17466868460178375,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 912
+    },
+    {
+      "epoch": 0.006343526350032531,
+      "grad_norm": 0.18268266320228577,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 913
+    },
+    {
+      "epoch": 0.0068314899154196486,
+      "grad_norm": 0.1771923452615738,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 914
+    },
+    {
+      "epoch": 0.007319453480806766,
+      "grad_norm": 0.17141546308994293,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 915
+    },
+    {
+      "epoch": 0.007807417046193884,
+      "grad_norm": 0.17356516420841217,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 916
+    },
+    {
+      "epoch": 0.008295380611581003,
+      "grad_norm": 0.16422489285469055,
+      "learning_rate": 0.0001,
+      "loss": 1.4906,
+      "step": 917
+    },
+    {
+      "epoch": 0.00878334417696812,
+      "grad_norm": 0.1787266582250595,
+      "learning_rate": 0.0001,
+      "loss": 1.7343,
+      "step": 918
+    },
+    {
+      "epoch": 0.009271307742355238,
+      "grad_norm": 0.17580311000347137,
+      "learning_rate": 0.0001,
+      "loss": 1.6336,
+      "step": 919
+    },
+    {
+      "epoch": 0.009759271307742356,
+      "grad_norm": 0.1774471253156662,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 920
+    },
+    {
+      "epoch": 0.010247234873129472,
+      "grad_norm": 0.17901279032230377,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 921
+    },
+    {
+      "epoch": 0.01073519843851659,
+      "grad_norm": 0.17782244086265564,
+      "learning_rate": 0.0001,
+      "loss": 1.5776,
+      "step": 922
+    },
+    {
+      "epoch": 0.011223162003903709,
+      "grad_norm": 0.17713989317417145,
+      "learning_rate": 0.0001,
+      "loss": 1.7171,
+      "step": 923
+    },
+    {
+      "epoch": 0.011711125569290826,
+      "grad_norm": 0.17973695695400238,
+      "learning_rate": 0.0001,
+      "loss": 1.6245,
+      "step": 924
+    },
+    {
+      "epoch": 0.012199089134677944,
+      "grad_norm": 0.17270395159721375,
+      "learning_rate": 0.0001,
+      "loss": 1.6232,
+      "step": 925
+    },
+    {
+      "epoch": 0.012687052700065062,
+      "grad_norm": 0.17838077247142792,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 926
+    },
+    {
+      "epoch": 0.013175016265452179,
+      "grad_norm": 0.17474934458732605,
+      "learning_rate": 0.0001,
+      "loss": 1.5513,
+      "step": 927
+    },
+    {
+      "epoch": 0.013662979830839297,
+      "grad_norm": 0.18032978475093842,
+      "learning_rate": 0.0001,
+      "loss": 1.6039,
+      "step": 928
+    },
+    {
+      "epoch": 0.014150943396226415,
+      "grad_norm": 0.19847191870212555,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 929
+    },
+    {
+      "epoch": 0.014638906961613532,
+      "grad_norm": 0.1837671548128128,
+      "learning_rate": 0.0001,
+      "loss": 1.7186,
+      "step": 930
+    },
+    {
+      "epoch": 0.01512687052700065,
+      "grad_norm": 0.1803646832704544,
+      "learning_rate": 0.0001,
+      "loss": 1.6564,
+      "step": 931
+    },
+    {
+      "epoch": 0.015614834092387769,
+      "grad_norm": 0.17980346083641052,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 932
+    },
+    {
+      "epoch": 0.016102797657774885,
+      "grad_norm": 0.18618029356002808,
+      "learning_rate": 0.0001,
+      "loss": 1.7277,
+      "step": 933
+    },
+    {
+      "epoch": 0.016590761223162005,
+      "grad_norm": 0.18398627638816833,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 934
+    },
+    {
+      "epoch": 0.017078724788549122,
+      "grad_norm": 0.185965895652771,
+      "learning_rate": 0.0001,
+      "loss": 1.6037,
+      "step": 935
+    },
+    {
+      "epoch": 0.01756668835393624,
+      "grad_norm": 0.1780605912208557,
+      "learning_rate": 0.0001,
+      "loss": 1.666,
+      "step": 936
+    },
+    {
+      "epoch": 0.01805465191932336,
+      "grad_norm": 0.17847025394439697,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 937
+    },
+    {
+      "epoch": 0.018542615484710475,
+      "grad_norm": 0.17785435914993286,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 938
+    },
+    {
+      "epoch": 0.01903057905009759,
+      "grad_norm": 0.178766131401062,
+      "learning_rate": 0.0001,
+      "loss": 1.6488,
+      "step": 939
+    },
+    {
+      "epoch": 0.01951854261548471,
+      "grad_norm": 0.17145715653896332,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 940
+    },
+    {
+      "epoch": 0.020006506180871828,
+      "grad_norm": 0.19819146394729614,
+      "learning_rate": 0.0001,
+      "loss": 1.8361,
+      "step": 941
+    },
+    {
+      "epoch": 0.020494469746258945,
+      "grad_norm": 0.1808314174413681,
+      "learning_rate": 0.0001,
+      "loss": 1.6356,
+      "step": 942
+    },
+    {
+      "epoch": 0.020982433311646065,
+      "grad_norm": 0.194684699177742,
+      "learning_rate": 0.0001,
+      "loss": 1.775,
+      "step": 943
+    },
+    {
+      "epoch": 0.02147039687703318,
+      "grad_norm": 0.18260495364665985,
+      "learning_rate": 0.0001,
+      "loss": 1.7603,
+      "step": 944
+    },
+    {
+      "epoch": 0.021958360442420298,
+      "grad_norm": 0.17065013945102692,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 945
+    },
+    {
+      "epoch": 0.022446324007807418,
+      "grad_norm": 0.18216338753700256,
+      "learning_rate": 0.0001,
+      "loss": 1.6448,
+      "step": 946
+    },
+    {
+      "epoch": 0.022934287573194535,
+      "grad_norm": 0.19104807078838348,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 947
+    },
+    {
+      "epoch": 0.02342225113858165,
+      "grad_norm": 0.19180017709732056,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 948
+    },
+    {
+      "epoch": 0.02391021470396877,
+      "grad_norm": 0.18234263360500336,
+      "learning_rate": 0.0001,
+      "loss": 1.5557,
+      "step": 949
+    },
+    {
+      "epoch": 0.024398178269355888,
+      "grad_norm": 0.1747346669435501,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 950
+    },
+    {
+      "epoch": 0.024886141834743004,
+      "grad_norm": 0.18235090374946594,
+      "learning_rate": 0.0001,
+      "loss": 1.6798,
+      "step": 951
+    },
+    {
+      "epoch": 0.025374105400130124,
+      "grad_norm": 0.19145113229751587,
+      "learning_rate": 0.0001,
+      "loss": 1.6369,
+      "step": 952
+    },
+    {
+      "epoch": 0.02586206896551724,
+      "grad_norm": 0.17904643714427948,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 953
+    },
+    {
+      "epoch": 0.026350032530904358,
+      "grad_norm": 0.19035224616527557,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 954
+    },
+    {
+      "epoch": 0.026837996096291478,
+      "grad_norm": 0.18414707481861115,
+      "learning_rate": 0.0001,
+      "loss": 1.654,
+      "step": 955
+    },
+    {
+      "epoch": 0.027325959661678594,
+      "grad_norm": 0.17875875532627106,
+      "learning_rate": 0.0001,
+      "loss": 1.565,
+      "step": 956
+    },
+    {
+      "epoch": 0.02781392322706571,
+      "grad_norm": 0.18031494319438934,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 957
+    },
+    {
+      "epoch": 0.02830188679245283,
+      "grad_norm": 0.18969757854938507,
+      "learning_rate": 0.0001,
+      "loss": 1.646,
+      "step": 958
+    },
+    {
+      "epoch": 0.028789850357839947,
+      "grad_norm": 0.20995917916297913,
+      "learning_rate": 0.0001,
+      "loss": 1.7464,
+      "step": 959
+    },
+    {
+      "epoch": 0.029277813923227064,
+      "grad_norm": 0.17736504971981049,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 960
+    },
+    {
+      "epoch": 0.029765777488614184,
+      "grad_norm": 0.20638507604599,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 961
+    },
+    {
+      "epoch": 0.0302537410540013,
+      "grad_norm": 0.19201546907424927,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 962
+    },
+    {
+      "epoch": 0.03074170461938842,
+      "grad_norm": 0.19754734635353088,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 963
+    },
+    {
+      "epoch": 0.031229668184775537,
+      "grad_norm": 0.17794667184352875,
+      "learning_rate": 0.0001,
+      "loss": 1.6228,
+      "step": 964
+    },
+    {
+      "epoch": 0.03171763175016266,
+      "grad_norm": 0.18569526076316833,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 965
+    },
+    {
+      "epoch": 0.03220559531554977,
+      "grad_norm": 0.19523699581623077,
+      "learning_rate": 0.0001,
+      "loss": 1.6172,
+      "step": 966
+    },
+    {
+      "epoch": 0.03269355888093689,
+      "grad_norm": 0.18242095410823822,
+      "learning_rate": 0.0001,
+      "loss": 1.6666,
+      "step": 967
+    },
+    {
+      "epoch": 0.03318152244632401,
+      "grad_norm": 0.18098166584968567,
+      "learning_rate": 0.0001,
+      "loss": 1.6144,
+      "step": 968
+    },
+    {
+      "epoch": 0.033669486011711124,
+      "grad_norm": 0.19416692852973938,
+      "learning_rate": 0.0001,
+      "loss": 1.6068,
+      "step": 969
+    },
+    {
+      "epoch": 0.034157449577098244,
+      "grad_norm": 0.17713020741939545,
+      "learning_rate": 0.0001,
+      "loss": 1.578,
+      "step": 970
+    },
+    {
+      "epoch": 0.034645413142485364,
+      "grad_norm": 0.18263442814350128,
+      "learning_rate": 0.0001,
+      "loss": 1.689,
+      "step": 971
+    },
+    {
+      "epoch": 0.03513337670787248,
+      "grad_norm": 0.18622995913028717,
+      "learning_rate": 0.0001,
+      "loss": 1.6026,
+      "step": 972
+    },
+    {
+      "epoch": 0.0356213402732596,
+      "grad_norm": 0.1996946781873703,
+      "learning_rate": 0.0001,
+      "loss": 1.7636,
+      "step": 973
+    },
+    {
+      "epoch": 0.03610930383864672,
+      "grad_norm": 0.1867581605911255,
+      "learning_rate": 0.0001,
+      "loss": 1.7104,
+      "step": 974
+    },
+    {
+      "epoch": 0.03659726740403383,
+      "grad_norm": 0.1805364042520523,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 975
+    },
+    {
+      "epoch": 0.03708523096942095,
+      "grad_norm": 0.18575693666934967,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 976
+    },
+    {
+      "epoch": 0.03757319453480807,
+      "grad_norm": 0.18839362263679504,
+      "learning_rate": 0.0001,
+      "loss": 1.5385,
+      "step": 977
+    },
+    {
+      "epoch": 0.03806115810019518,
+      "grad_norm": 0.1833166927099228,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 978
+    },
+    {
+      "epoch": 0.0385491216655823,
+      "grad_norm": 0.18456041812896729,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 979
+    },
+    {
+      "epoch": 0.03903708523096942,
+      "grad_norm": 0.169480562210083,
+      "learning_rate": 0.0001,
+      "loss": 1.4622,
+      "step": 980
+    },
+    {
+      "epoch": 0.039525048796356536,
+      "grad_norm": 0.18231213092803955,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 981
+    },
+    {
+      "epoch": 0.040013012361743656,
+      "grad_norm": 0.1789146065711975,
+      "learning_rate": 0.0001,
+      "loss": 1.5611,
+      "step": 982
+    },
+    {
+      "epoch": 0.040500975927130776,
+      "grad_norm": 0.18640698492527008,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 983
+    },
+    {
+      "epoch": 0.04098893949251789,
+      "grad_norm": 0.1967524290084839,
+      "learning_rate": 0.0001,
+      "loss": 1.6045,
+      "step": 984
+    },
+    {
+      "epoch": 0.04147690305790501,
+      "grad_norm": 0.19263187050819397,
+      "learning_rate": 0.0001,
+      "loss": 1.6902,
+      "step": 985
+    },
+    {
+      "epoch": 0.04196486662329213,
+      "grad_norm": 0.18511182069778442,
+      "learning_rate": 0.0001,
+      "loss": 1.679,
+      "step": 986
+    },
+    {
+      "epoch": 0.04245283018867924,
+      "grad_norm": 0.1992368996143341,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 987
+    },
+    {
+      "epoch": 0.04294079375406636,
+      "grad_norm": 0.18218041956424713,
+      "learning_rate": 0.0001,
+      "loss": 1.6078,
+      "step": 988
+    },
+    {
+      "epoch": 0.04342875731945348,
+      "grad_norm": 0.182425856590271,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 989
+    },
+    {
+      "epoch": 0.043916720884840596,
+      "grad_norm": 0.1946290284395218,
+      "learning_rate": 0.0001,
+      "loss": 1.6629,
+      "step": 990
+    },
+    {
+      "epoch": 0.044404684450227716,
+      "grad_norm": 0.18229249119758606,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 991
+    },
+    {
+      "epoch": 0.044892648015614836,
+      "grad_norm": 0.1886509656906128,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 992
+    },
+    {
+      "epoch": 0.04538061158100195,
+      "grad_norm": 0.19646328687667847,
+      "learning_rate": 0.0001,
+      "loss": 1.7327,
+      "step": 993
+    },
+    {
+      "epoch": 0.04586857514638907,
+      "grad_norm": 0.1876036375761032,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 994
+    },
+    {
+      "epoch": 0.04635653871177619,
+      "grad_norm": 0.21887090802192688,
+      "learning_rate": 0.0001,
+      "loss": 1.5949,
+      "step": 995
+    },
+    {
+      "epoch": 0.0468445022771633,
+      "grad_norm": 0.18973636627197266,
+      "learning_rate": 0.0001,
+      "loss": 1.6368,
+      "step": 996
+    },
+    {
+      "epoch": 0.04733246584255042,
+      "grad_norm": 0.1869606077671051,
+      "learning_rate": 0.0001,
+      "loss": 1.5081,
+      "step": 997
+    },
+    {
+      "epoch": 0.04782042940793754,
+      "grad_norm": 0.1914638876914978,
+      "learning_rate": 0.0001,
+      "loss": 1.6842,
+      "step": 998
+    },
+    {
+      "epoch": 0.048308392973324656,
+      "grad_norm": 0.18226292729377747,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 999
+    },
+    {
+      "epoch": 0.048796356538711776,
+      "grad_norm": 0.18389537930488586,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 1000
+    },
+    {
+      "epoch": 0.049284320104098896,
+      "grad_norm": 0.18721693754196167,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 1001
+    },
+    {
+      "epoch": 0.04977228366948601,
+      "grad_norm": 0.18689191341400146,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 1002
+    },
+    {
+      "epoch": 0.05026024723487313,
+      "grad_norm": 0.1945948302745819,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 1003
+    },
+    {
+      "epoch": 0.05074821080026025,
+      "grad_norm": 0.18874327838420868,
+      "learning_rate": 0.0001,
+      "loss": 1.6522,
+      "step": 1004
+    },
+    {
+      "epoch": 0.05123617436564736,
+      "grad_norm": 0.1909698098897934,
+      "learning_rate": 0.0001,
+      "loss": 1.6385,
+      "step": 1005
+    },
+    {
+      "epoch": 0.05172413793103448,
+      "grad_norm": 0.18150854110717773,
+      "learning_rate": 0.0001,
+      "loss": 1.5733,
+      "step": 1006
+    },
+    {
+      "epoch": 0.0522121014964216,
+      "grad_norm": 0.1733761876821518,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 1007
+    },
+    {
+      "epoch": 0.052700065061808715,
+      "grad_norm": 0.17895954847335815,
+      "learning_rate": 0.0001,
+      "loss": 1.5606,
+      "step": 1008
+    },
+    {
+      "epoch": 0.053188028627195835,
+      "grad_norm": 0.19132757186889648,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 1009
+    },
+    {
+      "epoch": 0.053675992192582955,
+      "grad_norm": 0.19307127594947815,
+      "learning_rate": 0.0001,
+      "loss": 1.6352,
+      "step": 1010
+    },
+    {
+      "epoch": 0.05416395575797007,
+      "grad_norm": 0.19357778131961823,
+      "learning_rate": 0.0001,
+      "loss": 1.6289,
+      "step": 1011
+    },
+    {
+      "epoch": 0.05465191932335719,
+      "grad_norm": 0.18533514440059662,
+      "learning_rate": 0.0001,
+      "loss": 1.572,
+      "step": 1012
+    },
+    {
+      "epoch": 0.05513988288874431,
+      "grad_norm": 0.18907758593559265,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 1013
+    },
+    {
+      "epoch": 0.05562784645413142,
+      "grad_norm": 0.19305996596813202,
+      "learning_rate": 0.0001,
+      "loss": 1.6204,
+      "step": 1014
+    },
+    {
+      "epoch": 0.05611581001951854,
+      "grad_norm": 0.18632037937641144,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 1015
+    },
+    {
+      "epoch": 0.05660377358490566,
+      "grad_norm": 0.19198158383369446,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 1016
+    },
+    {
+      "epoch": 0.057091737150292775,
+      "grad_norm": 0.18633070588111877,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 1017
+    },
+    {
+      "epoch": 0.057579700715679895,
+      "grad_norm": 0.19591306149959564,
+      "learning_rate": 0.0001,
+      "loss": 1.4697,
+      "step": 1018
+    },
+    {
+      "epoch": 0.058067664281067015,
+      "grad_norm": 0.18810021877288818,
+      "learning_rate": 0.0001,
+      "loss": 1.6529,
+      "step": 1019
+    },
+    {
+      "epoch": 0.05855562784645413,
+      "grad_norm": 0.2101171314716339,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 1020
+    },
+    {
+      "epoch": 0.05904359141184125,
+      "grad_norm": 0.1966390758752823,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 1021
+    },
+    {
+      "epoch": 0.05953155497722837,
+      "grad_norm": 0.17996057868003845,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 1022
+    },
+    {
+      "epoch": 0.06001951854261549,
+      "grad_norm": 0.18714721500873566,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 1023
+    },
+    {
+      "epoch": 0.0605074821080026,
+      "grad_norm": 0.17736780643463135,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 1024
+    },
+    {
+      "epoch": 0.06099544567338972,
+      "grad_norm": 0.19277577102184296,
+      "learning_rate": 0.0001,
+      "loss": 1.7098,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06148340923877684,
+      "grad_norm": 0.18830280005931854,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 1026
+    },
+    {
+      "epoch": 0.061971372804163954,
+      "grad_norm": 0.1853412538766861,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 1027
+    },
+    {
+      "epoch": 0.062459336369551074,
+      "grad_norm": 0.19096817076206207,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 1028
+    },
+    {
+      "epoch": 0.0629472999349382,
+      "grad_norm": 0.19431835412979126,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 1029
+    },
+    {
+      "epoch": 0.06343526350032531,
+      "grad_norm": 0.19152455031871796,
+      "learning_rate": 0.0001,
+      "loss": 1.6266,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06392322706571242,
+      "grad_norm": 0.1924085170030594,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 1031
+    },
+    {
+      "epoch": 0.06441119063109954,
+      "grad_norm": 0.19816018640995026,
+      "learning_rate": 0.0001,
+      "loss": 1.6888,
+      "step": 1032
+    },
+    {
+      "epoch": 0.06489915419648666,
+      "grad_norm": 0.1799917370080948,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 1033
+    },
+    {
+      "epoch": 0.06538711776187378,
+      "grad_norm": 0.19680505990982056,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 1034
+    },
+    {
+      "epoch": 0.0658750813272609,
+      "grad_norm": 0.18836094439029694,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06636304489264802,
+      "grad_norm": 0.20049379765987396,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 1036
+    },
+    {
+      "epoch": 0.06685100845803513,
+      "grad_norm": 0.19109798967838287,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 1037
+    },
+    {
+      "epoch": 0.06733897202342225,
+      "grad_norm": 0.1971965879201889,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 1038
+    },
+    {
+      "epoch": 0.06782693558880937,
+      "grad_norm": 0.18136058747768402,
+      "learning_rate": 0.0001,
+      "loss": 1.594,
+      "step": 1039
+    },
+    {
+      "epoch": 0.06831489915419649,
+      "grad_norm": 0.18866358697414398,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06880286271958361,
+      "grad_norm": 0.19541703164577484,
+      "learning_rate": 0.0001,
+      "loss": 1.5999,
+      "step": 1041
+    },
+    {
+      "epoch": 0.06929082628497073,
+      "grad_norm": 0.19072206318378448,
+      "learning_rate": 0.0001,
+      "loss": 1.6654,
+      "step": 1042
+    },
+    {
+      "epoch": 0.06977878985035783,
+      "grad_norm": 0.1988363415002823,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 1043
+    },
+    {
+      "epoch": 0.07026675341574495,
+      "grad_norm": 0.20157575607299805,
+      "learning_rate": 0.0001,
+      "loss": 1.6354,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07075471698113207,
+      "grad_norm": 0.19422085583209991,
+      "learning_rate": 0.0001,
+      "loss": 1.7082,
+      "step": 1045
+    },
+    {
+      "epoch": 0.0712426805465192,
+      "grad_norm": 0.19726014137268066,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07173064411190631,
+      "grad_norm": 0.18612346053123474,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 1047
+    },
+    {
+      "epoch": 0.07221860767729343,
+      "grad_norm": 0.19668620824813843,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 1048
+    },
+    {
+      "epoch": 0.07270657124268054,
+      "grad_norm": 0.19139578938484192,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 1049
+    },
+    {
+      "epoch": 0.07319453480806766,
+      "grad_norm": 0.18543675541877747,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07368249837345478,
+      "grad_norm": 0.19100967049598694,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 1051
+    },
+    {
+      "epoch": 0.0741704619388419,
+      "grad_norm": 0.1905674934387207,
+      "learning_rate": 0.0001,
+      "loss": 1.5048,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07465842550422902,
+      "grad_norm": 0.1817818284034729,
+      "learning_rate": 0.0001,
+      "loss": 1.4364,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07514638906961614,
+      "grad_norm": 0.19873027503490448,
+      "learning_rate": 0.0001,
+      "loss": 1.6095,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07563435263500325,
+      "grad_norm": 0.1952141374349594,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07612231620039037,
+      "grad_norm": 0.18775443732738495,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 1056
+    },
+    {
+      "epoch": 0.07661027976577749,
+      "grad_norm": 0.22584345936775208,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 1057
+    },
+    {
+      "epoch": 0.0770982433311646,
+      "grad_norm": 0.19559216499328613,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07758620689655173,
+      "grad_norm": 0.21493011713027954,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07807417046193885,
+      "grad_norm": 0.19376049935817719,
+      "learning_rate": 0.0001,
+      "loss": 1.6154,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07856213402732595,
+      "grad_norm": 0.19737021625041962,
+      "learning_rate": 0.0001,
+      "loss": 1.5796,
+      "step": 1061
+    },
+    {
+      "epoch": 0.07905009759271307,
+      "grad_norm": 0.1924256533384323,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 1062
+    },
+    {
+      "epoch": 0.07953806115810019,
+      "grad_norm": 0.20532450079917908,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 1063
+    },
+    {
+      "epoch": 0.08002602472348731,
+      "grad_norm": 0.2012764811515808,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 1064
+    },
+    {
+      "epoch": 0.08051398828887443,
+      "grad_norm": 0.20070607960224152,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 1065
+    },
+    {
+      "epoch": 0.08100195185426155,
+      "grad_norm": 0.19062010943889618,
+      "learning_rate": 0.0001,
+      "loss": 1.6905,
+      "step": 1066
+    },
+    {
+      "epoch": 0.08148991541964867,
+      "grad_norm": 0.20105774700641632,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 1067
+    },
+    {
+      "epoch": 0.08197787898503578,
+      "grad_norm": 0.19306330382823944,
+      "learning_rate": 0.0001,
+      "loss": 1.5686,
+      "step": 1068
+    },
+    {
+      "epoch": 0.0824658425504229,
+      "grad_norm": 0.18556571006774902,
+      "learning_rate": 0.0001,
+      "loss": 1.5638,
+      "step": 1069
+    },
+    {
+      "epoch": 0.08295380611581002,
+      "grad_norm": 0.19539238512516022,
+      "learning_rate": 0.0001,
+      "loss": 1.5695,
+      "step": 1070
+    },
+    {
+      "epoch": 0.08344176968119714,
+      "grad_norm": 0.19931377470493317,
+      "learning_rate": 0.0001,
+      "loss": 1.5871,
+      "step": 1071
+    },
+    {
+      "epoch": 0.08392973324658426,
+      "grad_norm": 0.19479453563690186,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 1072
+    },
+    {
+      "epoch": 0.08441769681197138,
+      "grad_norm": 0.19778072834014893,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 1073
+    },
+    {
+      "epoch": 0.08490566037735849,
+      "grad_norm": 0.18823745846748352,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 1074
+    },
+    {
+      "epoch": 0.0853936239427456,
+      "grad_norm": 0.19320188462734222,
+      "learning_rate": 0.0001,
+      "loss": 1.614,
+      "step": 1075
+    },
+    {
+      "epoch": 0.08588158750813273,
+      "grad_norm": 0.19224143028259277,
+      "learning_rate": 0.0001,
+      "loss": 1.5417,
+      "step": 1076
+    },
+    {
+      "epoch": 0.08636955107351985,
+      "grad_norm": 0.1908334493637085,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 1077
+    },
+    {
+      "epoch": 0.08685751463890697,
+      "grad_norm": 0.18100884556770325,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 1078
+    },
+    {
+      "epoch": 0.08734547820429409,
+      "grad_norm": 0.1889817714691162,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 1079
+    },
+    {
+      "epoch": 0.08783344176968119,
+      "grad_norm": 0.1986936628818512,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08832140533506831,
+      "grad_norm": 0.1832761913537979,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 1081
+    },
+    {
+      "epoch": 0.08880936890045543,
+      "grad_norm": 0.2123635858297348,
+      "learning_rate": 0.0001,
+      "loss": 1.6991,
+      "step": 1082
+    },
+    {
+      "epoch": 0.08929733246584255,
+      "grad_norm": 0.18817688524723053,
+      "learning_rate": 0.0001,
+      "loss": 1.5592,
+      "step": 1083
+    },
+    {
+      "epoch": 0.08978529603122967,
+      "grad_norm": 0.18663235008716583,
+      "learning_rate": 0.0001,
+      "loss": 1.5272,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09027325959661679,
+      "grad_norm": 0.21672074496746063,
+      "learning_rate": 0.0001,
+      "loss": 1.7311,
+      "step": 1085
+    },
+    {
+      "epoch": 0.0907612231620039,
+      "grad_norm": 0.19105497002601624,
+      "learning_rate": 0.0001,
+      "loss": 1.6297,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09124918672739102,
+      "grad_norm": 0.19626879692077637,
+      "learning_rate": 0.0001,
+      "loss": 1.6911,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09173715029277814,
+      "grad_norm": 0.19393813610076904,
+      "learning_rate": 0.0001,
+      "loss": 1.6525,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09222511385816526,
+      "grad_norm": 0.18055984377861023,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09271307742355238,
+      "grad_norm": 0.19464011490345,
+      "learning_rate": 0.0001,
+      "loss": 1.6359,
+      "step": 1090
+    },
+    {
+      "epoch": 0.0932010409889395,
+      "grad_norm": 0.1953517645597458,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 1091
+    },
+    {
+      "epoch": 0.0936890045543266,
+      "grad_norm": 0.2157415747642517,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09417696811971372,
+      "grad_norm": 0.1933530867099762,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09466493168510084,
+      "grad_norm": 0.19232773780822754,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09515289525048796,
+      "grad_norm": 0.20085276663303375,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09564085881587508,
+      "grad_norm": 0.195418119430542,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 1096
+    },
+    {
+      "epoch": 0.0961288223812622,
+      "grad_norm": 0.18770374357700348,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09661678594664931,
+      "grad_norm": 0.20245583355426788,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09710474951203643,
+      "grad_norm": 0.19320428371429443,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09759271307742355,
+      "grad_norm": 0.18845516443252563,
+      "learning_rate": 0.0001,
+      "loss": 1.525,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09808067664281067,
+      "grad_norm": 0.19028283655643463,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09856864020819779,
+      "grad_norm": 0.20150424540042877,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09905660377358491,
+      "grad_norm": 0.19038866460323334,
+      "learning_rate": 0.0001,
+      "loss": 1.6702,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09954456733897202,
+      "grad_norm": 0.196619912981987,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 1104
+    },
+    {
+      "epoch": 0.10003253090435914,
+      "grad_norm": 0.19666366279125214,
+      "learning_rate": 0.0001,
+      "loss": 1.6388,
+      "step": 1105
+    },
+    {
+      "epoch": 0.10052049446974626,
+      "grad_norm": 0.1931435614824295,
+      "learning_rate": 0.0001,
+      "loss": 1.669,
+      "step": 1106
+    },
+    {
+      "epoch": 0.10100845803513338,
+      "grad_norm": 0.21595527231693268,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 1107
+    },
+    {
+      "epoch": 0.1014964216005205,
+      "grad_norm": 0.1833282709121704,
+      "learning_rate": 0.0001,
+      "loss": 1.5769,
+      "step": 1108
+    },
+    {
+      "epoch": 0.10198438516590762,
+      "grad_norm": 0.18882323801517487,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 1109
+    },
+    {
+      "epoch": 0.10247234873129472,
+      "grad_norm": 0.1929425448179245,
+      "learning_rate": 0.0001,
+      "loss": 1.5915,
+      "step": 1110
+    },
+    {
+      "epoch": 0.10296031229668184,
+      "grad_norm": 0.18974991142749786,
+      "learning_rate": 0.0001,
+      "loss": 1.5802,
+      "step": 1111
+    },
+    {
+      "epoch": 0.10344827586206896,
+      "grad_norm": 0.19779950380325317,
+      "learning_rate": 0.0001,
+      "loss": 1.5605,
+      "step": 1112
+    },
+    {
+      "epoch": 0.10393623942745608,
+      "grad_norm": 0.19960924983024597,
+      "learning_rate": 0.0001,
+      "loss": 1.6344,
+      "step": 1113
+    },
+    {
+      "epoch": 0.1044242029928432,
+      "grad_norm": 0.24380812048912048,
+      "learning_rate": 0.0001,
+      "loss": 1.7643,
+      "step": 1114
+    },
+    {
+      "epoch": 0.10491216655823032,
+      "grad_norm": 0.19412434101104736,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 1115
+    },
+    {
+      "epoch": 0.10540013012361743,
+      "grad_norm": 0.20298714935779572,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 1116
+    },
+    {
+      "epoch": 0.10588809368900455,
+      "grad_norm": 0.19304509460926056,
+      "learning_rate": 0.0001,
+      "loss": 1.5581,
+      "step": 1117
+    },
+    {
+      "epoch": 0.10637605725439167,
+      "grad_norm": 0.20273859798908234,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 1118
+    },
+    {
+      "epoch": 0.10686402081977879,
+      "grad_norm": 0.21369436383247375,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 1119
+    },
+    {
+      "epoch": 0.10735198438516591,
+      "grad_norm": 0.1963755339384079,
+      "learning_rate": 0.0001,
+      "loss": 1.596,
+      "step": 1120
+    },
+    {
+      "epoch": 0.10783994795055303,
+      "grad_norm": 0.21036356687545776,
+      "learning_rate": 0.0001,
+      "loss": 1.5676,
+      "step": 1121
+    },
+    {
+      "epoch": 0.10832791151594014,
+      "grad_norm": 0.1949380338191986,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 1122
+    },
+    {
+      "epoch": 0.10881587508132726,
+      "grad_norm": 0.20131665468215942,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 1123
+    },
+    {
+      "epoch": 0.10930383864671438,
+      "grad_norm": 0.2017805278301239,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 1124
+    },
+    {
+      "epoch": 0.1097918022121015,
+      "grad_norm": 0.1949644386768341,
+      "learning_rate": 0.0001,
+      "loss": 1.6483,
+      "step": 1125
+    },
+    {
+      "epoch": 0.11027976577748862,
+      "grad_norm": 0.21223554015159607,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 1126
+    },
+    {
+      "epoch": 0.11076772934287574,
+      "grad_norm": 0.1937919557094574,
+      "learning_rate": 0.0001,
+      "loss": 1.473,
+      "step": 1127
+    },
+    {
+      "epoch": 0.11125569290826284,
+      "grad_norm": 0.1919855922460556,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 1128
+    },
+    {
+      "epoch": 0.11174365647364996,
+      "grad_norm": 0.19788561761379242,
+      "learning_rate": 0.0001,
+      "loss": 1.5735,
+      "step": 1129
+    },
+    {
+      "epoch": 0.11223162003903708,
+      "grad_norm": 0.2033752202987671,
+      "learning_rate": 0.0001,
+      "loss": 1.6344,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1127195836044242,
+      "grad_norm": 0.18651780486106873,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 1131
+    },
+    {
+      "epoch": 0.11320754716981132,
+      "grad_norm": 0.2055680751800537,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 1132
+    },
+    {
+      "epoch": 0.11369551073519844,
+      "grad_norm": 0.19889622926712036,
+      "learning_rate": 0.0001,
+      "loss": 1.575,
+      "step": 1133
+    },
+    {
+      "epoch": 0.11418347430058555,
+      "grad_norm": 0.21087592840194702,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 1134
+    },
+    {
+      "epoch": 0.11467143786597267,
+      "grad_norm": 0.20450933277606964,
+      "learning_rate": 0.0001,
+      "loss": 1.6181,
+      "step": 1135
+    },
+    {
+      "epoch": 0.11515940143135979,
+      "grad_norm": 0.20027051866054535,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 1136
+    },
+    {
+      "epoch": 0.11564736499674691,
+      "grad_norm": 0.21202796697616577,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 1137
+    },
+    {
+      "epoch": 0.11613532856213403,
+      "grad_norm": 0.2110545039176941,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 1138
+    },
+    {
+      "epoch": 0.11662329212752115,
+      "grad_norm": 0.19653548300266266,
+      "learning_rate": 0.0001,
+      "loss": 1.4339,
+      "step": 1139
+    },
+    {
+      "epoch": 0.11711125569290826,
+      "grad_norm": 0.20760329067707062,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11759921925829538,
+      "grad_norm": 0.18545225262641907,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 1141
+    },
+    {
+      "epoch": 0.1180871828236825,
+      "grad_norm": 0.20842860639095306,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 1142
+    },
+    {
+      "epoch": 0.11857514638906962,
+      "grad_norm": 0.2025844156742096,
+      "learning_rate": 0.0001,
+      "loss": 1.4659,
+      "step": 1143
+    },
+    {
+      "epoch": 0.11906310995445674,
+      "grad_norm": 0.20418566465377808,
+      "learning_rate": 0.0001,
+      "loss": 1.6403,
+      "step": 1144
+    },
+    {
+      "epoch": 0.11955107351984386,
+      "grad_norm": 0.21025940775871277,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 1145
+    },
+    {
+      "epoch": 0.12003903708523098,
+      "grad_norm": 0.18947705626487732,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 1146
+    },
+    {
+      "epoch": 0.12052700065061808,
+      "grad_norm": 0.19618074595928192,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 1147
+    },
+    {
+      "epoch": 0.1210149642160052,
+      "grad_norm": 0.20363377034664154,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 1148
+    },
+    {
+      "epoch": 0.12150292778139232,
+      "grad_norm": 0.20256803929805756,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 1149
+    },
+    {
+      "epoch": 0.12199089134677944,
+      "grad_norm": 0.2074446976184845,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12247885491216656,
+      "grad_norm": 0.20207305252552032,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 1151
+    },
+    {
+      "epoch": 0.12296681847755368,
+      "grad_norm": 0.19513028860092163,
+      "learning_rate": 0.0001,
+      "loss": 1.5529,
+      "step": 1152
+    },
+    {
+      "epoch": 0.12345478204294079,
+      "grad_norm": 0.21784643828868866,
+      "learning_rate": 0.0001,
+      "loss": 1.6741,
+      "step": 1153
+    },
+    {
+      "epoch": 0.12394274560832791,
+      "grad_norm": 0.19317090511322021,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 1154
+    },
+    {
+      "epoch": 0.12443070917371503,
+      "grad_norm": 0.20658369362354279,
+      "learning_rate": 0.0001,
+      "loss": 1.5898,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12491867273910215,
+      "grad_norm": 0.20049959421157837,
+      "learning_rate": 0.0001,
+      "loss": 1.6031,
+      "step": 1156
+    },
+    {
+      "epoch": 0.12540663630448926,
+      "grad_norm": 0.20353883504867554,
+      "learning_rate": 0.0001,
+      "loss": 1.6034,
+      "step": 1157
+    },
+    {
+      "epoch": 0.1258945998698764,
+      "grad_norm": 0.20507696270942688,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1263825634352635,
+      "grad_norm": 0.20546303689479828,
+      "learning_rate": 0.0001,
+      "loss": 1.5722,
+      "step": 1159
+    },
+    {
+      "epoch": 0.12687052700065063,
+      "grad_norm": 0.21547870337963104,
+      "learning_rate": 0.0001,
+      "loss": 1.6045,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12735849056603774,
+      "grad_norm": 0.2026733011007309,
+      "learning_rate": 0.0001,
+      "loss": 1.5867,
+      "step": 1161
+    },
+    {
+      "epoch": 0.12784645413142484,
+      "grad_norm": 0.19540677964687347,
+      "learning_rate": 0.0001,
+      "loss": 1.6144,
+      "step": 1162
+    },
+    {
+      "epoch": 0.12833441769681198,
+      "grad_norm": 0.20927543938159943,
+      "learning_rate": 0.0001,
+      "loss": 1.681,
+      "step": 1163
+    },
+    {
+      "epoch": 0.12882238126219908,
+      "grad_norm": 0.19741636514663696,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 1164
+    },
+    {
+      "epoch": 0.12931034482758622,
+      "grad_norm": 0.2057577520608902,
+      "learning_rate": 0.0001,
+      "loss": 1.6606,
+      "step": 1165
+    },
+    {
+      "epoch": 0.12979830839297332,
+      "grad_norm": 0.1989096701145172,
+      "learning_rate": 0.0001,
+      "loss": 1.5171,
+      "step": 1166
+    },
+    {
+      "epoch": 0.13028627195836046,
+      "grad_norm": 0.20165401697158813,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 1167
+    },
+    {
+      "epoch": 0.13077423552374756,
+      "grad_norm": 0.19830471277236938,
+      "learning_rate": 0.0001,
+      "loss": 1.6272,
+      "step": 1168
+    },
+    {
+      "epoch": 0.13126219908913467,
+      "grad_norm": 0.24735324084758759,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1317501626545218,
+      "grad_norm": 0.20342503488063812,
+      "learning_rate": 0.0001,
+      "loss": 1.5162,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1322381262199089,
+      "grad_norm": 0.20828960835933685,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 1171
+    },
+    {
+      "epoch": 0.13272608978529604,
+      "grad_norm": 0.21619682013988495,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 1172
+    },
+    {
+      "epoch": 0.13321405335068315,
+      "grad_norm": 0.2064361870288849,
+      "learning_rate": 0.0001,
+      "loss": 1.5601,
+      "step": 1173
+    },
+    {
+      "epoch": 0.13370201691607025,
+      "grad_norm": 0.20160190761089325,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 1174
+    },
+    {
+      "epoch": 0.1341899804814574,
+      "grad_norm": 0.19937147200107574,
+      "learning_rate": 0.0001,
+      "loss": 1.6174,
+      "step": 1175
+    },
+    {
+      "epoch": 0.1346779440468445,
+      "grad_norm": 0.19670027494430542,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 1176
+    },
+    {
+      "epoch": 0.13516590761223163,
+      "grad_norm": 0.21226651966571808,
+      "learning_rate": 0.0001,
+      "loss": 1.6624,
+      "step": 1177
+    },
+    {
+      "epoch": 0.13565387117761873,
+      "grad_norm": 0.1816990077495575,
+      "learning_rate": 0.0001,
+      "loss": 1.4428,
+      "step": 1178
+    },
+    {
+      "epoch": 0.13614183474300587,
+      "grad_norm": 0.20409594476222992,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 1179
+    },
+    {
+      "epoch": 0.13662979830839297,
+      "grad_norm": 0.19361446797847748,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 1180
+    },
+    {
+      "epoch": 0.13711776187378008,
+      "grad_norm": 0.1968996822834015,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 1181
+    },
+    {
+      "epoch": 0.13760572543916721,
+      "grad_norm": 0.2013946771621704,
+      "learning_rate": 0.0001,
+      "loss": 1.5662,
+      "step": 1182
+    },
+    {
+      "epoch": 0.13809368900455432,
+      "grad_norm": 0.20192427933216095,
+      "learning_rate": 0.0001,
+      "loss": 1.6427,
+      "step": 1183
+    },
+    {
+      "epoch": 0.13858165256994145,
+      "grad_norm": 0.18497639894485474,
+      "learning_rate": 0.0001,
+      "loss": 1.5073,
+      "step": 1184
+    },
+    {
+      "epoch": 0.13906961613532856,
+      "grad_norm": 0.20845462381839752,
+      "learning_rate": 0.0001,
+      "loss": 1.5992,
+      "step": 1185
+    },
+    {
+      "epoch": 0.13955757970071567,
+      "grad_norm": 0.20056185126304626,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1400455432661028,
+      "grad_norm": 0.21718023717403412,
+      "learning_rate": 0.0001,
+      "loss": 1.7617,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1405335068314899,
+      "grad_norm": 0.20736059546470642,
+      "learning_rate": 0.0001,
+      "loss": 1.7106,
+      "step": 1188
+    },
+    {
+      "epoch": 0.14102147039687704,
+      "grad_norm": 0.19962933659553528,
+      "learning_rate": 0.0001,
+      "loss": 1.5205,
+      "step": 1189
+    },
+    {
+      "epoch": 0.14150943396226415,
+      "grad_norm": 0.1965351700782776,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 1190
+    },
+    {
+      "epoch": 0.14199739752765128,
+      "grad_norm": 0.20114606618881226,
+      "learning_rate": 0.0001,
+      "loss": 1.7303,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1424853610930384,
+      "grad_norm": 0.2035091370344162,
+      "learning_rate": 0.0001,
+      "loss": 1.6627,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1429733246584255,
+      "grad_norm": 0.20694808661937714,
+      "learning_rate": 0.0001,
+      "loss": 1.6631,
+      "step": 1193
+    },
+    {
+      "epoch": 0.14346128822381263,
+      "grad_norm": 0.21123093366622925,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 1194
+    },
+    {
+      "epoch": 0.14394925178919973,
+      "grad_norm": 0.20809170603752136,
+      "learning_rate": 0.0001,
+      "loss": 1.5972,
+      "step": 1195
+    },
+    {
+      "epoch": 0.14443721535458687,
+      "grad_norm": 0.2276293933391571,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 1196
+    },
+    {
+      "epoch": 0.14492517891997397,
+      "grad_norm": 0.19207881391048431,
+      "learning_rate": 0.0001,
+      "loss": 1.5509,
+      "step": 1197
+    },
+    {
+      "epoch": 0.14541314248536108,
+      "grad_norm": 0.21243374049663544,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 1198
+    },
+    {
+      "epoch": 0.1459011060507482,
+      "grad_norm": 0.20206978917121887,
+      "learning_rate": 0.0001,
+      "loss": 1.6466,
+      "step": 1199
+    },
+    {
+      "epoch": 0.14638906961613532,
+      "grad_norm": 0.20090003311634064,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 1200
+    },
+    {
+      "epoch": 0.14687703318152245,
+      "grad_norm": 0.197933629155159,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 1201
+    },
+    {
+      "epoch": 0.14736499674690956,
+      "grad_norm": 0.20552389323711395,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 1202
+    },
+    {
+      "epoch": 0.1478529603122967,
+      "grad_norm": 0.1985408514738083,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1483409238776838,
+      "grad_norm": 0.2063884735107422,
+      "learning_rate": 0.0001,
+      "loss": 1.6156,
+      "step": 1204
+    },
+    {
+      "epoch": 0.1488288874430709,
+      "grad_norm": 0.19716759026050568,
+      "learning_rate": 0.0001,
+      "loss": 1.5355,
+      "step": 1205
+    },
+    {
+      "epoch": 0.14931685100845804,
+      "grad_norm": 0.20738033950328827,
+      "learning_rate": 0.0001,
+      "loss": 1.5793,
+      "step": 1206
+    },
+    {
+      "epoch": 0.14980481457384515,
+      "grad_norm": 0.21956196427345276,
+      "learning_rate": 0.0001,
+      "loss": 1.6484,
+      "step": 1207
+    },
+    {
+      "epoch": 0.15029277813923228,
+      "grad_norm": 0.18943889439105988,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1507807417046194,
+      "grad_norm": 0.19319723546504974,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1512687052700065,
+      "grad_norm": 0.20123296976089478,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 1210
+    },
+    {
+      "epoch": 0.15175666883539363,
+      "grad_norm": 0.21409326791763306,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 1211
+    },
+    {
+      "epoch": 0.15224463240078073,
+      "grad_norm": 0.1999148428440094,
+      "learning_rate": 0.0001,
+      "loss": 1.6435,
+      "step": 1212
+    },
+    {
+      "epoch": 0.15273259596616787,
+      "grad_norm": 0.21267859637737274,
+      "learning_rate": 0.0001,
+      "loss": 1.6582,
+      "step": 1213
+    },
+    {
+      "epoch": 0.15322055953155497,
+      "grad_norm": 0.194093257188797,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1537085230969421,
+      "grad_norm": 0.20447438955307007,
+      "learning_rate": 0.0001,
+      "loss": 1.7055,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1541964866623292,
+      "grad_norm": 0.21048034727573395,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 1216
+    },
+    {
+      "epoch": 0.15468445022771632,
+      "grad_norm": 0.19501455128192902,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 1217
+    },
+    {
+      "epoch": 0.15517241379310345,
+      "grad_norm": 0.2119220793247223,
+      "learning_rate": 0.0001,
+      "loss": 1.676,
+      "step": 1218
+    },
+    {
+      "epoch": 0.15566037735849056,
+      "grad_norm": 0.20096439123153687,
+      "learning_rate": 0.0001,
+      "loss": 1.5335,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1561483409238777,
+      "grad_norm": 0.19789251685142517,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1566363044892648,
+      "grad_norm": 0.19607964158058167,
+      "learning_rate": 0.0001,
+      "loss": 1.6243,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1571242680546519,
+      "grad_norm": 0.2003655731678009,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 1222
+    },
+    {
+      "epoch": 0.15761223162003904,
+      "grad_norm": 0.22465141117572784,
+      "learning_rate": 0.0001,
+      "loss": 1.7123,
+      "step": 1223
+    },
+    {
+      "epoch": 0.15810019518542615,
+      "grad_norm": 0.20125439763069153,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 1224
+    },
+    {
+      "epoch": 0.15858815875081328,
+      "grad_norm": 0.19792059063911438,
+      "learning_rate": 0.0001,
+      "loss": 1.6016,
+      "step": 1225
+    },
+    {
+      "epoch": 0.15907612231620039,
+      "grad_norm": 0.1911260485649109,
+      "learning_rate": 0.0001,
+      "loss": 1.5992,
+      "step": 1226
+    },
+    {
+      "epoch": 0.15956408588158752,
+      "grad_norm": 0.20277781784534454,
+      "learning_rate": 0.0001,
+      "loss": 1.7467,
+      "step": 1227
+    },
+    {
+      "epoch": 0.16005204944697463,
+      "grad_norm": 0.2036551833152771,
+      "learning_rate": 0.0001,
+      "loss": 1.6187,
+      "step": 1228
+    },
+    {
+      "epoch": 0.16054001301236173,
+      "grad_norm": 0.19617246091365814,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 1229
+    },
+    {
+      "epoch": 0.16102797657774887,
+      "grad_norm": 0.20863553881645203,
+      "learning_rate": 0.0001,
+      "loss": 1.5719,
+      "step": 1230
+    },
+    {
+      "epoch": 0.16151594014313597,
+      "grad_norm": 0.19776329398155212,
+      "learning_rate": 0.0001,
+      "loss": 1.5005,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1620039037085231,
+      "grad_norm": 0.20257459580898285,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1624918672739102,
+      "grad_norm": 0.20090225338935852,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 1233
+    },
+    {
+      "epoch": 0.16297983083929735,
+      "grad_norm": 0.1917838156223297,
+      "learning_rate": 0.0001,
+      "loss": 1.5461,
+      "step": 1234
+    },
+    {
+      "epoch": 0.16346779440468445,
+      "grad_norm": 0.1895637810230255,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 1235
+    },
+    {
+      "epoch": 0.16395575797007156,
+      "grad_norm": 0.19651246070861816,
+      "learning_rate": 0.0001,
+      "loss": 1.6139,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1644437215354587,
+      "grad_norm": 0.19278347492218018,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1649316851008458,
+      "grad_norm": 0.1980368047952652,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 1238
+    },
+    {
+      "epoch": 0.16541964866623293,
+      "grad_norm": 0.19496038556098938,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 1239
+    },
+    {
+      "epoch": 0.16590761223162004,
+      "grad_norm": 0.20839765667915344,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 1240
+    },
+    {
+      "epoch": 0.16639557579700714,
+      "grad_norm": 0.19290494918823242,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 1241
+    },
+    {
+      "epoch": 0.16688353936239428,
+      "grad_norm": 0.20755812525749207,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 1242
+    },
+    {
+      "epoch": 0.16737150292778138,
+      "grad_norm": 0.21205563843250275,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 1243
+    },
+    {
+      "epoch": 0.16785946649316852,
+      "grad_norm": 0.19532056152820587,
+      "learning_rate": 0.0001,
+      "loss": 1.5427,
+      "step": 1244
+    },
+    {
+      "epoch": 0.16834743005855562,
+      "grad_norm": 0.20462632179260254,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 1245
+    },
+    {
+      "epoch": 0.16883539362394276,
+      "grad_norm": 0.1949065774679184,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 1246
+    },
+    {
+      "epoch": 0.16932335718932987,
+      "grad_norm": 0.20552422106266022,
+      "learning_rate": 0.0001,
+      "loss": 1.5206,
+      "step": 1247
+    },
+    {
+      "epoch": 0.16981132075471697,
+      "grad_norm": 0.1937859207391739,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1702992843201041,
+      "grad_norm": 0.21231001615524292,
+      "learning_rate": 0.0001,
+      "loss": 1.682,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1707872478854912,
+      "grad_norm": 0.21864455938339233,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 1250
+    },
+    {
+      "epoch": 0.17127521145087835,
+      "grad_norm": 0.20900295674800873,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 1251
+    },
+    {
+      "epoch": 0.17176317501626545,
+      "grad_norm": 0.20172104239463806,
+      "learning_rate": 0.0001,
+      "loss": 1.5454,
+      "step": 1252
+    },
+    {
+      "epoch": 0.17225113858165256,
+      "grad_norm": 0.19377106428146362,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1727391021470397,
+      "grad_norm": 0.20241466164588928,
+      "learning_rate": 0.0001,
+      "loss": 1.4751,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1732270657124268,
+      "grad_norm": 0.2040991485118866,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 1255
+    },
+    {
+      "epoch": 0.17371502927781393,
+      "grad_norm": 0.19520303606987,
+      "learning_rate": 0.0001,
+      "loss": 1.5673,
+      "step": 1256
+    },
+    {
+      "epoch": 0.17420299284320104,
+      "grad_norm": 0.19177937507629395,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 1257
+    },
+    {
+      "epoch": 0.17469095640858817,
+      "grad_norm": 0.19967958331108093,
+      "learning_rate": 0.0001,
+      "loss": 1.6683,
+      "step": 1258
+    },
+    {
+      "epoch": 0.17517891997397528,
+      "grad_norm": 0.19648973643779755,
+      "learning_rate": 0.0001,
+      "loss": 1.5659,
+      "step": 1259
+    },
+    {
+      "epoch": 0.17566688353936238,
+      "grad_norm": 0.21236220002174377,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 1260
+    },
+    {
+      "epoch": 0.17615484710474952,
+      "grad_norm": 0.2165279984474182,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 1261
+    },
+    {
+      "epoch": 0.17664281067013662,
+      "grad_norm": 0.192067489027977,
+      "learning_rate": 0.0001,
+      "loss": 1.4738,
+      "step": 1262
+    },
+    {
+      "epoch": 0.17713077423552376,
+      "grad_norm": 0.18832434713840485,
+      "learning_rate": 0.0001,
+      "loss": 1.4642,
+      "step": 1263
+    },
+    {
+      "epoch": 0.17761873780091086,
+      "grad_norm": 0.20100747048854828,
+      "learning_rate": 0.0001,
+      "loss": 1.7199,
+      "step": 1264
+    },
+    {
+      "epoch": 0.17810670136629797,
+      "grad_norm": 0.2086695283651352,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1785946649316851,
+      "grad_norm": 0.2020365595817566,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1790826284970722,
+      "grad_norm": 0.19790609180927277,
+      "learning_rate": 0.0001,
+      "loss": 1.5258,
+      "step": 1267
+    },
+    {
+      "epoch": 0.17957059206245934,
+      "grad_norm": 0.19993458688259125,
+      "learning_rate": 0.0001,
+      "loss": 1.562,
+      "step": 1268
+    },
+    {
+      "epoch": 0.18005855562784645,
+      "grad_norm": 0.19273720681667328,
+      "learning_rate": 0.0001,
+      "loss": 1.4675,
+      "step": 1269
+    },
+    {
+      "epoch": 0.18054651919323358,
+      "grad_norm": 0.21065354347229004,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1810344827586207,
+      "grad_norm": 0.20578408241271973,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1815224463240078,
+      "grad_norm": 0.2011623978614807,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 1272
+    },
+    {
+      "epoch": 0.18201040988939493,
+      "grad_norm": 0.19478732347488403,
+      "learning_rate": 0.0001,
+      "loss": 1.4385,
+      "step": 1273
+    },
+    {
+      "epoch": 0.18249837345478204,
+      "grad_norm": 0.19872449338436127,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 1274
+    },
+    {
+      "epoch": 0.18298633702016917,
+      "grad_norm": 0.20244348049163818,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 1275
+    },
+    {
+      "epoch": 0.18347430058555628,
+      "grad_norm": 0.20218241214752197,
+      "learning_rate": 0.0001,
+      "loss": 1.5547,
+      "step": 1276
+    },
+    {
+      "epoch": 0.18396226415094338,
+      "grad_norm": 0.2011180967092514,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 1277
+    },
+    {
+      "epoch": 0.18445022771633052,
+      "grad_norm": 0.19339530169963837,
+      "learning_rate": 0.0001,
+      "loss": 1.4543,
+      "step": 1278
+    },
+    {
+      "epoch": 0.18493819128171762,
+      "grad_norm": 0.20447854697704315,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 1279
+    },
+    {
+      "epoch": 0.18542615484710476,
+      "grad_norm": 0.20977169275283813,
+      "learning_rate": 0.0001,
+      "loss": 1.6585,
+      "step": 1280
+    },
+    {
+      "epoch": 0.18591411841249186,
+      "grad_norm": 0.19544698297977448,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 1281
+    },
+    {
+      "epoch": 0.186402081977879,
+      "grad_norm": 0.21273665130138397,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1868900455432661,
+      "grad_norm": 0.20312640070915222,
+      "learning_rate": 0.0001,
+      "loss": 1.5779,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1873780091086532,
+      "grad_norm": 0.20123647153377533,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 1284
+    },
+    {
+      "epoch": 0.18786597267404034,
+      "grad_norm": 0.21253986656665802,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 1285
+    },
+    {
+      "epoch": 0.18835393623942745,
+      "grad_norm": 0.20770984888076782,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 1286
+    },
+    {
+      "epoch": 0.18884189980481458,
+      "grad_norm": 0.21010833978652954,
+      "learning_rate": 0.0001,
+      "loss": 1.6707,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1893298633702017,
+      "grad_norm": 0.20496207475662231,
+      "learning_rate": 0.0001,
+      "loss": 1.5322,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1898178269355888,
+      "grad_norm": 0.20395588874816895,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 1289
+    },
+    {
+      "epoch": 0.19030579050097593,
+      "grad_norm": 0.21773125231266022,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 1290
+    },
+    {
+      "epoch": 0.19079375406636304,
+      "grad_norm": 0.20682354271411896,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 1291
+    },
+    {
+      "epoch": 0.19128171763175017,
+      "grad_norm": 0.19925051927566528,
+      "learning_rate": 0.0001,
+      "loss": 1.5105,
+      "step": 1292
+    },
+    {
+      "epoch": 0.19176968119713728,
+      "grad_norm": 0.21521666646003723,
+      "learning_rate": 0.0001,
+      "loss": 1.7389,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1922576447625244,
+      "grad_norm": 0.1969422698020935,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 1294
+    },
+    {
+      "epoch": 0.19274560832791152,
+      "grad_norm": 0.19946162402629852,
+      "learning_rate": 0.0001,
+      "loss": 1.5796,
+      "step": 1295
+    },
+    {
+      "epoch": 0.19323357189329862,
+      "grad_norm": 0.21713663637638092,
+      "learning_rate": 0.0001,
+      "loss": 1.5857,
+      "step": 1296
+    },
+    {
+      "epoch": 0.19372153545868576,
+      "grad_norm": 0.23006384074687958,
+      "learning_rate": 0.0001,
+      "loss": 1.6505,
+      "step": 1297
+    },
+    {
+      "epoch": 0.19420949902407286,
+      "grad_norm": 0.2125861644744873,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 1298
+    },
+    {
+      "epoch": 0.19469746258946,
+      "grad_norm": 0.20375065505504608,
+      "learning_rate": 0.0001,
+      "loss": 1.6238,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1951854261548471,
+      "grad_norm": 0.21823540329933167,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1956733897202342,
+      "grad_norm": 0.20082317292690277,
+      "learning_rate": 0.0001,
+      "loss": 1.609,
+      "step": 1301
+    },
+    {
+      "epoch": 0.19616135328562134,
+      "grad_norm": 0.2042534351348877,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 1302
+    },
+    {
+      "epoch": 0.19664931685100845,
+      "grad_norm": 0.20837345719337463,
+      "learning_rate": 0.0001,
+      "loss": 1.56,
+      "step": 1303
+    },
+    {
+      "epoch": 0.19713728041639558,
+      "grad_norm": 0.20953041315078735,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 1304
+    },
+    {
+      "epoch": 0.1976252439817827,
+      "grad_norm": 0.2184993326663971,
+      "learning_rate": 0.0001,
+      "loss": 1.6644,
+      "step": 1305
+    },
+    {
+      "epoch": 0.19811320754716982,
+      "grad_norm": 0.21605993807315826,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 1306
+    },
+    {
+      "epoch": 0.19860117111255693,
+      "grad_norm": 0.20878028869628906,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 1307
+    },
+    {
+      "epoch": 0.19908913467794404,
+      "grad_norm": 0.20920151472091675,
+      "learning_rate": 0.0001,
+      "loss": 1.5738,
+      "step": 1308
+    },
+    {
+      "epoch": 0.19957709824333117,
+      "grad_norm": 0.20488986372947693,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 1309
+    },
+    {
+      "epoch": 0.20006506180871828,
+      "grad_norm": 0.20617173612117767,
+      "learning_rate": 0.0001,
+      "loss": 1.6885,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2005530253741054,
+      "grad_norm": 0.22102181613445282,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 1311
+    },
+    {
+      "epoch": 0.20104098893949252,
+      "grad_norm": 0.21153733134269714,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 1312
+    },
+    {
+      "epoch": 0.20152895250487965,
+      "grad_norm": 0.21832062304019928,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 1313
+    },
+    {
+      "epoch": 0.20201691607026676,
+      "grad_norm": 0.2205532193183899,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 1314
+    },
+    {
+      "epoch": 0.20250487963565386,
+      "grad_norm": 0.21700820326805115,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 1315
+    },
+    {
+      "epoch": 0.202992843201041,
+      "grad_norm": 0.22612810134887695,
+      "learning_rate": 0.0001,
+      "loss": 1.6248,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2034808067664281,
+      "grad_norm": 0.20869220793247223,
+      "learning_rate": 0.0001,
+      "loss": 1.5233,
+      "step": 1317
+    },
+    {
+      "epoch": 0.20396877033181524,
+      "grad_norm": 0.20189987123012543,
+      "learning_rate": 0.0001,
+      "loss": 1.6216,
+      "step": 1318
+    },
+    {
+      "epoch": 0.20445673389720234,
+      "grad_norm": 0.22843563556671143,
+      "learning_rate": 0.0001,
+      "loss": 1.5712,
+      "step": 1319
+    },
+    {
+      "epoch": 0.20494469746258945,
+      "grad_norm": 0.20801427960395813,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 1320
+    },
+    {
+      "epoch": 0.20543266102797658,
+      "grad_norm": 0.21543341875076294,
+      "learning_rate": 0.0001,
+      "loss": 1.617,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2059206245933637,
+      "grad_norm": 0.1988707035779953,
+      "learning_rate": 0.0001,
+      "loss": 1.5099,
+      "step": 1322
+    },
+    {
+      "epoch": 0.20640858815875082,
+      "grad_norm": 0.21976548433303833,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 1323
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 0.2349298596382141,
+      "learning_rate": 0.0001,
+      "loss": 1.6349,
+      "step": 1324
+    },
+    {
+      "epoch": 0.20738451528952506,
+      "grad_norm": 0.2023395299911499,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 1325
+    },
+    {
+      "epoch": 0.20787247885491217,
+      "grad_norm": 0.24304138123989105,
+      "learning_rate": 0.0001,
+      "loss": 1.7519,
+      "step": 1326
+    },
+    {
+      "epoch": 0.20836044242029927,
+      "grad_norm": 0.21027036011219025,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2088484059856864,
+      "grad_norm": 0.20986603200435638,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 1328
+    },
+    {
+      "epoch": 0.20933636955107351,
+      "grad_norm": 0.2174183577299118,
+      "learning_rate": 0.0001,
+      "loss": 1.506,
+      "step": 1329
+    },
+    {
+      "epoch": 0.20982433311646065,
+      "grad_norm": 0.20996814966201782,
+      "learning_rate": 0.0001,
+      "loss": 1.5895,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21031229668184775,
+      "grad_norm": 0.20884643495082855,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 1331
+    },
+    {
+      "epoch": 0.21080026024723486,
+      "grad_norm": 0.2158053070306778,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 1332
+    },
+    {
+      "epoch": 0.211288223812622,
+      "grad_norm": 0.2126191407442093,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2117761873780091,
+      "grad_norm": 0.20810382068157196,
+      "learning_rate": 0.0001,
+      "loss": 1.562,
+      "step": 1334
+    },
+    {
+      "epoch": 0.21226415094339623,
+      "grad_norm": 0.2033027708530426,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 1335
+    },
+    {
+      "epoch": 0.21275211450878334,
+      "grad_norm": 0.21471640467643738,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 1336
+    },
+    {
+      "epoch": 0.21324007807417047,
+      "grad_norm": 0.21316203474998474,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 1337
+    },
+    {
+      "epoch": 0.21372804163955758,
+      "grad_norm": 0.21801623702049255,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2142160052049447,
+      "grad_norm": 0.19841396808624268,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 1339
+    },
+    {
+      "epoch": 0.21470396877033182,
+      "grad_norm": 0.22380690276622772,
+      "learning_rate": 0.0001,
+      "loss": 1.6075,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21519193233571893,
+      "grad_norm": 0.20887979865074158,
+      "learning_rate": 0.0001,
+      "loss": 1.487,
+      "step": 1341
+    },
+    {
+      "epoch": 0.21567989590110606,
+      "grad_norm": 0.21433189511299133,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 1342
+    },
+    {
+      "epoch": 0.21616785946649317,
+      "grad_norm": 0.22923235595226288,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 1343
+    },
+    {
+      "epoch": 0.21665582303188027,
+      "grad_norm": 0.20079989731311798,
+      "learning_rate": 0.0001,
+      "loss": 1.6014,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2171437865972674,
+      "grad_norm": 0.21039649844169617,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2176317501626545,
+      "grad_norm": 0.22379867732524872,
+      "learning_rate": 0.0001,
+      "loss": 1.7028,
+      "step": 1346
+    },
+    {
+      "epoch": 0.21811971372804165,
+      "grad_norm": 0.20918886363506317,
+      "learning_rate": 0.0001,
+      "loss": 1.6503,
+      "step": 1347
+    },
+    {
+      "epoch": 0.21860767729342875,
+      "grad_norm": 0.22096358239650726,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2190956408588159,
+      "grad_norm": 0.20308831334114075,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 1349
+    },
+    {
+      "epoch": 0.219583604424203,
+      "grad_norm": 0.20950287580490112,
+      "learning_rate": 0.0001,
+      "loss": 1.5901,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2200715679895901,
+      "grad_norm": 0.20067515969276428,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 1351
+    },
+    {
+      "epoch": 0.22055953155497723,
+      "grad_norm": 0.2053097039461136,
+      "learning_rate": 0.0001,
+      "loss": 1.525,
+      "step": 1352
+    },
+    {
+      "epoch": 0.22104749512036434,
+      "grad_norm": 0.21317452192306519,
+      "learning_rate": 0.0001,
+      "loss": 1.7089,
+      "step": 1353
+    },
+    {
+      "epoch": 0.22153545868575147,
+      "grad_norm": 0.20587880909442902,
+      "learning_rate": 0.0001,
+      "loss": 1.6195,
+      "step": 1354
+    },
+    {
+      "epoch": 0.22202342225113858,
+      "grad_norm": 0.20507857203483582,
+      "learning_rate": 0.0001,
+      "loss": 1.6527,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2225113858165257,
+      "grad_norm": 0.19736212491989136,
+      "learning_rate": 0.0001,
+      "loss": 1.5156,
+      "step": 1356
+    },
+    {
+      "epoch": 0.22299934938191282,
+      "grad_norm": 0.20208993554115295,
+      "learning_rate": 0.0001,
+      "loss": 1.6376,
+      "step": 1357
+    },
+    {
+      "epoch": 0.22348731294729993,
+      "grad_norm": 0.2129242867231369,
+      "learning_rate": 0.0001,
+      "loss": 1.6449,
+      "step": 1358
+    },
+    {
+      "epoch": 0.22397527651268706,
+      "grad_norm": 0.20393933355808258,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 1359
+    },
+    {
+      "epoch": 0.22446324007807417,
+      "grad_norm": 0.19787192344665527,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2249512036434613,
+      "grad_norm": 0.21041975915431976,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 1361
+    },
+    {
+      "epoch": 0.2254391672088484,
+      "grad_norm": 0.21465621888637543,
+      "learning_rate": 0.0001,
+      "loss": 1.6976,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2259271307742355,
+      "grad_norm": 0.21297186613082886,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 1363
+    },
+    {
+      "epoch": 0.22641509433962265,
+      "grad_norm": 0.212122842669487,
+      "learning_rate": 0.0001,
+      "loss": 1.6634,
+      "step": 1364
+    },
+    {
+      "epoch": 0.22690305790500975,
+      "grad_norm": 0.20742806792259216,
+      "learning_rate": 0.0001,
+      "loss": 1.6743,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2273910214703969,
+      "grad_norm": 0.2097882479429245,
+      "learning_rate": 0.0001,
+      "loss": 1.5079,
+      "step": 1366
+    },
+    {
+      "epoch": 0.227878985035784,
+      "grad_norm": 0.2064836323261261,
+      "learning_rate": 0.0001,
+      "loss": 1.5146,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2283669486011711,
+      "grad_norm": 0.20146428048610687,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 1368
+    },
+    {
+      "epoch": 0.22885491216655823,
+      "grad_norm": 0.22511009871959686,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 1369
+    },
+    {
+      "epoch": 0.22934287573194534,
+      "grad_norm": 0.21153807640075684,
+      "learning_rate": 0.0001,
+      "loss": 1.5621,
+      "step": 1370
+    },
+    {
+      "epoch": 0.22983083929733247,
+      "grad_norm": 0.21294273436069489,
+      "learning_rate": 0.0001,
+      "loss": 1.6372,
+      "step": 1371
+    },
+    {
+      "epoch": 0.23031880286271958,
+      "grad_norm": 0.21079088747501373,
+      "learning_rate": 0.0001,
+      "loss": 1.6835,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2308067664281067,
+      "grad_norm": 0.20809081196784973,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 1373
+    },
+    {
+      "epoch": 0.23129472999349382,
+      "grad_norm": 0.21136967837810516,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 1374
+    },
+    {
+      "epoch": 0.23178269355888093,
+      "grad_norm": 0.21215151250362396,
+      "learning_rate": 0.0001,
+      "loss": 1.6307,
+      "step": 1375
+    },
+    {
+      "epoch": 0.23227065712426806,
+      "grad_norm": 0.2087780237197876,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 1376
+    },
+    {
+      "epoch": 0.23275862068965517,
+      "grad_norm": 0.1958921253681183,
+      "learning_rate": 0.0001,
+      "loss": 1.5113,
+      "step": 1377
+    },
+    {
+      "epoch": 0.2332465842550423,
+      "grad_norm": 0.20907118916511536,
+      "learning_rate": 0.0001,
+      "loss": 1.5028,
+      "step": 1378
+    },
+    {
+      "epoch": 0.2337345478204294,
+      "grad_norm": 0.20895539224147797,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2342225113858165,
+      "grad_norm": 0.21354170143604279,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 1380
+    },
+    {
+      "epoch": 0.23471047495120365,
+      "grad_norm": 0.2014855593442917,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 1381
+    },
+    {
+      "epoch": 0.23519843851659075,
+      "grad_norm": 0.20819585025310516,
+      "learning_rate": 0.0001,
+      "loss": 1.6451,
+      "step": 1382
+    },
+    {
+      "epoch": 0.23568640208197789,
+      "grad_norm": 0.2062918245792389,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 1383
+    },
+    {
+      "epoch": 0.236174365647365,
+      "grad_norm": 0.22420808672904968,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 1384
+    },
+    {
+      "epoch": 0.23666232921275213,
+      "grad_norm": 0.2028312236070633,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 1385
+    },
+    {
+      "epoch": 0.23715029277813923,
+      "grad_norm": 0.1977011263370514,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 1386
+    },
+    {
+      "epoch": 0.23763825634352634,
+      "grad_norm": 0.21594934165477753,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 1387
+    },
+    {
+      "epoch": 0.23812621990891347,
+      "grad_norm": 0.21049955487251282,
+      "learning_rate": 0.0001,
+      "loss": 1.5645,
+      "step": 1388
+    },
+    {
+      "epoch": 0.23861418347430058,
+      "grad_norm": 0.2092577964067459,
+      "learning_rate": 0.0001,
+      "loss": 1.6993,
+      "step": 1389
+    },
+    {
+      "epoch": 0.2391021470396877,
+      "grad_norm": 0.21054156124591827,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 1390
+    },
+    {
+      "epoch": 0.23959011060507482,
+      "grad_norm": 0.20244278013706207,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 1391
+    },
+    {
+      "epoch": 0.24007807417046195,
+      "grad_norm": 0.2174467295408249,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 1392
+    },
+    {
+      "epoch": 0.24056603773584906,
+      "grad_norm": 0.2089550644159317,
+      "learning_rate": 0.0001,
+      "loss": 1.693,
+      "step": 1393
+    },
+    {
+      "epoch": 0.24105400130123616,
+      "grad_norm": 0.21089689433574677,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 1394
+    },
+    {
+      "epoch": 0.2415419648666233,
+      "grad_norm": 0.20715463161468506,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2420299284320104,
+      "grad_norm": 0.21272169053554535,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 1396
+    },
+    {
+      "epoch": 0.24251789199739754,
+      "grad_norm": 0.21278724074363708,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 1397
+    },
+    {
+      "epoch": 0.24300585556278465,
+      "grad_norm": 0.20560777187347412,
+      "learning_rate": 0.0001,
+      "loss": 1.4915,
+      "step": 1398
+    },
+    {
+      "epoch": 0.24349381912817175,
+      "grad_norm": 0.20997489988803864,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 1399
+    },
+    {
+      "epoch": 0.24398178269355889,
+      "grad_norm": 0.22127652168273926,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 1400
+    },
+    {
+      "epoch": 0.244469746258946,
+      "grad_norm": 0.20486050844192505,
+      "learning_rate": 0.0001,
+      "loss": 1.4894,
+      "step": 1401
+    },
+    {
+      "epoch": 0.24495770982433313,
+      "grad_norm": 0.20870132744312286,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 1402
+    },
+    {
+      "epoch": 0.24544567338972023,
+      "grad_norm": 0.2159174531698227,
+      "learning_rate": 0.0001,
+      "loss": 1.5926,
+      "step": 1403
+    },
+    {
+      "epoch": 0.24593363695510737,
+      "grad_norm": 0.20647220313549042,
+      "learning_rate": 0.0001,
+      "loss": 1.6031,
+      "step": 1404
+    },
+    {
+      "epoch": 0.24642160052049447,
+      "grad_norm": 0.22608523070812225,
+      "learning_rate": 0.0001,
+      "loss": 1.717,
+      "step": 1405
+    },
+    {
+      "epoch": 0.24690956408588158,
+      "grad_norm": 0.19434261322021484,
+      "learning_rate": 0.0001,
+      "loss": 1.4748,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2473975276512687,
+      "grad_norm": 0.21329692006111145,
+      "learning_rate": 0.0001,
+      "loss": 1.5945,
+      "step": 1407
+    },
+    {
+      "epoch": 0.24788549121665582,
+      "grad_norm": 0.2082255631685257,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 1408
+    },
+    {
+      "epoch": 0.24837345478204295,
+      "grad_norm": 0.21107988059520721,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 1409
+    },
+    {
+      "epoch": 0.24886141834743006,
+      "grad_norm": 0.20110711455345154,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 1410
+    },
+    {
+      "epoch": 0.24934938191281716,
+      "grad_norm": 0.2062540501356125,
+      "learning_rate": 0.0001,
+      "loss": 1.5938,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2498373454782043,
+      "grad_norm": 0.2092439830303192,
+      "learning_rate": 0.0001,
+      "loss": 1.6857,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2503253090435914,
+      "grad_norm": 0.22201520204544067,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 1413
+    },
+    {
+      "epoch": 0.2508132726089785,
+      "grad_norm": 0.2209090292453766,
+      "learning_rate": 0.0001,
+      "loss": 1.5933,
+      "step": 1414
+    },
+    {
+      "epoch": 0.25130123617436567,
+      "grad_norm": 0.20598077774047852,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2517891997397528,
+      "grad_norm": 0.217951238155365,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2522771633051399,
+      "grad_norm": 0.20061807334423065,
+      "learning_rate": 0.0001,
+      "loss": 1.4733,
+      "step": 1417
+    },
+    {
+      "epoch": 0.252765126870527,
+      "grad_norm": 0.20080770552158356,
+      "learning_rate": 0.0001,
+      "loss": 1.4549,
+      "step": 1418
+    },
+    {
+      "epoch": 0.2532530904359141,
+      "grad_norm": 0.2048519402742386,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 1419
+    },
+    {
+      "epoch": 0.25374105400130126,
+      "grad_norm": 0.20036721229553223,
+      "learning_rate": 0.0001,
+      "loss": 1.5525,
+      "step": 1420
+    },
+    {
+      "epoch": 0.25422901756668836,
+      "grad_norm": 0.19026263058185577,
+      "learning_rate": 0.0001,
+      "loss": 1.4788,
+      "step": 1421
+    },
+    {
+      "epoch": 0.25471698113207547,
+      "grad_norm": 0.21005001664161682,
+      "learning_rate": 0.0001,
+      "loss": 1.5978,
+      "step": 1422
+    },
+    {
+      "epoch": 0.2552049446974626,
+      "grad_norm": 0.20866861939430237,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 1423
+    },
+    {
+      "epoch": 0.2556929082628497,
+      "grad_norm": 0.20086826384067535,
+      "learning_rate": 0.0001,
+      "loss": 1.5717,
+      "step": 1424
+    },
+    {
+      "epoch": 0.25618087182823684,
+      "grad_norm": 0.21193620562553406,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 1425
+    },
+    {
+      "epoch": 0.25666883539362395,
+      "grad_norm": 0.20752951502799988,
+      "learning_rate": 0.0001,
+      "loss": 1.6411,
+      "step": 1426
+    },
+    {
+      "epoch": 0.25715679895901106,
+      "grad_norm": 0.2076740860939026,
+      "learning_rate": 0.0001,
+      "loss": 1.5757,
+      "step": 1427
+    },
+    {
+      "epoch": 0.25764476252439816,
+      "grad_norm": 0.2160855382680893,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 1428
+    },
+    {
+      "epoch": 0.25813272608978527,
+      "grad_norm": 0.20601573586463928,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 1429
+    },
+    {
+      "epoch": 0.25862068965517243,
+      "grad_norm": 0.20943140983581543,
+      "learning_rate": 0.0001,
+      "loss": 1.5442,
+      "step": 1430
+    },
+    {
+      "epoch": 0.25910865322055954,
+      "grad_norm": 0.2184855192899704,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 1431
+    },
+    {
+      "epoch": 0.25959661678594664,
+      "grad_norm": 0.2096775472164154,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 1432
+    },
+    {
+      "epoch": 0.26008458035133375,
+      "grad_norm": 0.20223049819469452,
+      "learning_rate": 0.0001,
+      "loss": 1.5356,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2605725439167209,
+      "grad_norm": 0.2273465394973755,
+      "learning_rate": 0.0001,
+      "loss": 1.6606,
+      "step": 1434
+    },
+    {
+      "epoch": 0.261060507482108,
+      "grad_norm": 0.20657625794410706,
+      "learning_rate": 0.0001,
+      "loss": 1.5155,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2615484710474951,
+      "grad_norm": 0.22336073219776154,
+      "learning_rate": 0.0001,
+      "loss": 1.5677,
+      "step": 1436
+    },
+    {
+      "epoch": 0.26203643461288223,
+      "grad_norm": 0.2139173299074173,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 1437
+    },
+    {
+      "epoch": 0.26252439817826934,
+      "grad_norm": 0.21106921136379242,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 1438
+    },
+    {
+      "epoch": 0.2630123617436565,
+      "grad_norm": 0.2122088372707367,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 1439
+    },
+    {
+      "epoch": 0.2635003253090436,
+      "grad_norm": 0.21802666783332825,
+      "learning_rate": 0.0001,
+      "loss": 1.6336,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2639882888744307,
+      "grad_norm": 0.22520272433757782,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2644762524398178,
+      "grad_norm": 0.20212695002555847,
+      "learning_rate": 0.0001,
+      "loss": 1.5565,
+      "step": 1442
+    },
+    {
+      "epoch": 0.2649642160052049,
+      "grad_norm": 0.21890683472156525,
+      "learning_rate": 0.0001,
+      "loss": 1.6999,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2654521795705921,
+      "grad_norm": 0.20161810517311096,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2659401431359792,
+      "grad_norm": 0.20803825557231903,
+      "learning_rate": 0.0001,
+      "loss": 1.4655,
+      "step": 1445
+    },
+    {
+      "epoch": 0.2664281067013663,
+      "grad_norm": 0.21093057096004486,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2669160702667534,
+      "grad_norm": 0.20793235301971436,
+      "learning_rate": 0.0001,
+      "loss": 1.5948,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2674040338321405,
+      "grad_norm": 0.2096520960330963,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 1448
+    },
+    {
+      "epoch": 0.26789199739752767,
+      "grad_norm": 0.22152283787727356,
+      "learning_rate": 0.0001,
+      "loss": 1.617,
+      "step": 1449
+    },
+    {
+      "epoch": 0.2683799609629148,
+      "grad_norm": 0.2101874053478241,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2688679245283019,
+      "grad_norm": 0.2182386815547943,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 1451
+    },
+    {
+      "epoch": 0.269355888093689,
+      "grad_norm": 0.20778711140155792,
+      "learning_rate": 0.0001,
+      "loss": 1.4838,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2698438516590761,
+      "grad_norm": 0.2238636314868927,
+      "learning_rate": 0.0001,
+      "loss": 1.7494,
+      "step": 1453
+    },
+    {
+      "epoch": 0.27033181522446326,
+      "grad_norm": 0.21300889551639557,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 1454
+    },
+    {
+      "epoch": 0.27081977878985036,
+      "grad_norm": 0.21480794250965118,
+      "learning_rate": 0.0001,
+      "loss": 1.5903,
+      "step": 1455
+    },
+    {
+      "epoch": 0.27130774235523747,
+      "grad_norm": 0.2183651328086853,
+      "learning_rate": 0.0001,
+      "loss": 1.5938,
+      "step": 1456
+    },
+    {
+      "epoch": 0.2717957059206246,
+      "grad_norm": 0.22869636118412018,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 1457
+    },
+    {
+      "epoch": 0.27228366948601174,
+      "grad_norm": 0.2531774044036865,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 1458
+    },
+    {
+      "epoch": 0.27277163305139884,
+      "grad_norm": 0.1954401433467865,
+      "learning_rate": 0.0001,
+      "loss": 1.4522,
+      "step": 1459
+    },
+    {
+      "epoch": 0.27325959661678595,
+      "grad_norm": 0.20861664414405823,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 1460
+    },
+    {
+      "epoch": 0.27374756018217306,
+      "grad_norm": 0.22775766253471375,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 1461
+    },
+    {
+      "epoch": 0.27423552374756016,
+      "grad_norm": 0.21558338403701782,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 1462
+    },
+    {
+      "epoch": 0.2747234873129473,
+      "grad_norm": 0.21214412152767181,
+      "learning_rate": 0.0001,
+      "loss": 1.6374,
+      "step": 1463
+    },
+    {
+      "epoch": 0.27521145087833443,
+      "grad_norm": 0.20641180872917175,
+      "learning_rate": 0.0001,
+      "loss": 1.6487,
+      "step": 1464
+    },
+    {
+      "epoch": 0.27569941444372154,
+      "grad_norm": 0.226535826921463,
+      "learning_rate": 0.0001,
+      "loss": 1.7734,
+      "step": 1465
+    },
+    {
+      "epoch": 0.27618737800910864,
+      "grad_norm": 0.2066386491060257,
+      "learning_rate": 0.0001,
+      "loss": 1.5098,
+      "step": 1466
+    },
+    {
+      "epoch": 0.27667534157449575,
+      "grad_norm": 0.2100495845079422,
+      "learning_rate": 0.0001,
+      "loss": 1.5078,
+      "step": 1467
+    },
+    {
+      "epoch": 0.2771633051398829,
+      "grad_norm": 0.2035837024450302,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 1468
+    },
+    {
+      "epoch": 0.27765126870527,
+      "grad_norm": 0.20833101868629456,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2781392322706571,
+      "grad_norm": 0.20039208233356476,
+      "learning_rate": 0.0001,
+      "loss": 1.5099,
+      "step": 1470
+    },
+    {
+      "epoch": 0.27862719583604423,
+      "grad_norm": 0.2074117809534073,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 1471
+    },
+    {
+      "epoch": 0.27911515940143133,
+      "grad_norm": 0.22486601769924164,
+      "learning_rate": 0.0001,
+      "loss": 1.7084,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2796031229668185,
+      "grad_norm": 0.20245811343193054,
+      "learning_rate": 0.0001,
+      "loss": 1.4773,
+      "step": 1473
+    },
+    {
+      "epoch": 0.2800910865322056,
+      "grad_norm": 0.21456313133239746,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2805790500975927,
+      "grad_norm": 0.19818849861621857,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 1475
+    },
+    {
+      "epoch": 0.2810670136629798,
+      "grad_norm": 0.21809843182563782,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 1476
+    },
+    {
+      "epoch": 0.281554977228367,
+      "grad_norm": 0.20338422060012817,
+      "learning_rate": 0.0001,
+      "loss": 1.5472,
+      "step": 1477
+    },
+    {
+      "epoch": 0.2820429407937541,
+      "grad_norm": 0.2087392956018448,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 1478
+    },
+    {
+      "epoch": 0.2825309043591412,
+      "grad_norm": 0.21048003435134888,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 1479
+    },
+    {
+      "epoch": 0.2830188679245283,
+      "grad_norm": 0.20787788927555084,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2835068314899154,
+      "grad_norm": 0.20244333148002625,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 1481
+    },
+    {
+      "epoch": 0.28399479505530256,
+      "grad_norm": 0.2096267193555832,
+      "learning_rate": 0.0001,
+      "loss": 1.4599,
+      "step": 1482
+    },
+    {
+      "epoch": 0.28448275862068967,
+      "grad_norm": 0.21420978009700775,
+      "learning_rate": 0.0001,
+      "loss": 1.6356,
+      "step": 1483
+    },
+    {
+      "epoch": 0.2849707221860768,
+      "grad_norm": 0.22837483882904053,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 1484
+    },
+    {
+      "epoch": 0.2854586857514639,
+      "grad_norm": 0.21912246942520142,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 1485
+    },
+    {
+      "epoch": 0.285946649316851,
+      "grad_norm": 0.21429242193698883,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 1486
+    },
+    {
+      "epoch": 0.28643461288223815,
+      "grad_norm": 0.21875980496406555,
+      "learning_rate": 0.0001,
+      "loss": 1.5264,
+      "step": 1487
+    },
+    {
+      "epoch": 0.28692257644762525,
+      "grad_norm": 0.20935304462909698,
+      "learning_rate": 0.0001,
+      "loss": 1.5276,
+      "step": 1488
+    },
+    {
+      "epoch": 0.28741054001301236,
+      "grad_norm": 0.21394003927707672,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 1489
+    },
+    {
+      "epoch": 0.28789850357839947,
+      "grad_norm": 0.2134189009666443,
+      "learning_rate": 0.0001,
+      "loss": 1.5238,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2883864671437866,
+      "grad_norm": 0.2189924418926239,
+      "learning_rate": 0.0001,
+      "loss": 1.6752,
+      "step": 1491
+    },
+    {
+      "epoch": 0.28887443070917374,
+      "grad_norm": 0.22713236510753632,
+      "learning_rate": 0.0001,
+      "loss": 1.7328,
+      "step": 1492
+    },
+    {
+      "epoch": 0.28936239427456084,
+      "grad_norm": 0.22816357016563416,
+      "learning_rate": 0.0001,
+      "loss": 1.7468,
+      "step": 1493
+    },
+    {
+      "epoch": 0.28985035783994795,
+      "grad_norm": 0.21150028705596924,
+      "learning_rate": 0.0001,
+      "loss": 1.6527,
+      "step": 1494
+    },
+    {
+      "epoch": 0.29033832140533505,
+      "grad_norm": 0.19985318183898926,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 1495
+    },
+    {
+      "epoch": 0.29082628497072216,
+      "grad_norm": 0.205070361495018,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 1496
+    },
+    {
+      "epoch": 0.2913142485361093,
+      "grad_norm": 0.21083366870880127,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2918022121014964,
+      "grad_norm": 0.21962401270866394,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 1498
+    },
+    {
+      "epoch": 0.29229017566688353,
+      "grad_norm": 0.21149326860904694,
+      "learning_rate": 0.0001,
+      "loss": 1.5554,
+      "step": 1499
+    },
+    {
+      "epoch": 0.29277813923227064,
+      "grad_norm": 0.21829526126384735,
+      "learning_rate": 0.0001,
+      "loss": 1.6243,
+      "step": 1500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2049,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 300,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.936745174106112e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}