diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14532 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 2070,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014492753623188406,
+      "grad_norm": 48.2703755078483,
+      "learning_rate": 2.4154589371980677e-07,
+      "loss": 11.412,
+      "step": 1
+    },
+    {
+      "epoch": 0.002898550724637681,
+      "grad_norm": 47.846609744441345,
+      "learning_rate": 4.830917874396135e-07,
+      "loss": 11.3179,
+      "step": 2
+    },
+    {
+      "epoch": 0.004347826086956522,
+      "grad_norm": 43.327843265354325,
+      "learning_rate": 7.246376811594203e-07,
+      "loss": 11.7074,
+      "step": 3
+    },
+    {
+      "epoch": 0.005797101449275362,
+      "grad_norm": 45.996221072245085,
+      "learning_rate": 9.66183574879227e-07,
+      "loss": 11.5688,
+      "step": 4
+    },
+    {
+      "epoch": 0.007246376811594203,
+      "grad_norm": 44.80245082664995,
+      "learning_rate": 1.2077294685990338e-06,
+      "loss": 11.5525,
+      "step": 5
+    },
+    {
+      "epoch": 0.008695652173913044,
+      "grad_norm": 43.7912268141257,
+      "learning_rate": 1.4492753623188406e-06,
+      "loss": 11.6803,
+      "step": 6
+    },
+    {
+      "epoch": 0.010144927536231883,
+      "grad_norm": 46.19726966065695,
+      "learning_rate": 1.6908212560386474e-06,
+      "loss": 11.4856,
+      "step": 7
+    },
+    {
+      "epoch": 0.011594202898550725,
+      "grad_norm": 49.683280956555635,
+      "learning_rate": 1.932367149758454e-06,
+      "loss": 11.2487,
+      "step": 8
+    },
+    {
+      "epoch": 0.013043478260869565,
+      "grad_norm": 49.5432312831397,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 11.3125,
+      "step": 9
+    },
+    {
+      "epoch": 0.014492753623188406,
+      "grad_norm": 170.48068034517684,
+      "learning_rate": 2.4154589371980677e-06,
+      "loss": 10.4208,
+      "step": 10
+    },
+    {
+      "epoch": 0.015942028985507246,
+      "grad_norm": 74.68814324825067,
+      "learning_rate": 2.6570048309178746e-06,
+      "loss": 10.1322,
+      "step": 11
+    },
+    {
+      "epoch": 0.017391304347826087,
+      "grad_norm": 76.52431188574639,
+      "learning_rate": 2.898550724637681e-06,
+      "loss": 9.677,
+      "step": 12
+    },
+    {
+      "epoch": 0.01884057971014493,
+      "grad_norm": 96.8305482953425,
+      "learning_rate": 3.140096618357488e-06,
+      "loss": 6.0848,
+      "step": 13
+    },
+    {
+      "epoch": 0.020289855072463767,
+      "grad_norm": 79.49651285333789,
+      "learning_rate": 3.3816425120772947e-06,
+      "loss": 5.2455,
+      "step": 14
+    },
+    {
+      "epoch": 0.021739130434782608,
+      "grad_norm": 70.92706207793557,
+      "learning_rate": 3.6231884057971017e-06,
+      "loss": 4.555,
+      "step": 15
+    },
+    {
+      "epoch": 0.02318840579710145,
+      "grad_norm": 52.332573285132604,
+      "learning_rate": 3.864734299516908e-06,
+      "loss": 3.6889,
+      "step": 16
+    },
+    {
+      "epoch": 0.02463768115942029,
+      "grad_norm": 41.41102395269734,
+      "learning_rate": 4.106280193236716e-06,
+      "loss": 3.0651,
+      "step": 17
+    },
+    {
+      "epoch": 0.02608695652173913,
+      "grad_norm": 7.477039232121631,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 1.6705,
+      "step": 18
+    },
+    {
+      "epoch": 0.02753623188405797,
+      "grad_norm": 6.077802039052816,
+      "learning_rate": 4.589371980676329e-06,
+      "loss": 1.7725,
+      "step": 19
+    },
+    {
+      "epoch": 0.028985507246376812,
+      "grad_norm": 5.2869318809884875,
+      "learning_rate": 4.830917874396135e-06,
+      "loss": 1.6962,
+      "step": 20
+    },
+    {
+      "epoch": 0.030434782608695653,
+      "grad_norm": 4.168623701247975,
+      "learning_rate": 5.072463768115943e-06,
+      "loss": 1.4515,
+      "step": 21
+    },
+    {
+      "epoch": 0.03188405797101449,
+      "grad_norm": 3.1618053994702673,
+      "learning_rate": 5.314009661835749e-06,
+      "loss": 1.4379,
+      "step": 22
+    },
+    {
+      "epoch": 0.03333333333333333,
+      "grad_norm": 2.8417104654582968,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 1.6453,
+      "step": 23
+    },
+    {
+      "epoch": 0.034782608695652174,
+      "grad_norm": 2.2491818177379757,
+      "learning_rate": 5.797101449275362e-06,
+      "loss": 1.6312,
+      "step": 24
+    },
+    {
+      "epoch": 0.036231884057971016,
+      "grad_norm": 1.5951285511686666,
+      "learning_rate": 6.038647342995169e-06,
+      "loss": 1.2319,
+      "step": 25
+    },
+    {
+      "epoch": 0.03768115942028986,
+      "grad_norm": 1.7838476788934974,
+      "learning_rate": 6.280193236714976e-06,
+      "loss": 1.4211,
+      "step": 26
+    },
+    {
+      "epoch": 0.0391304347826087,
+      "grad_norm": 4.11606765432239,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 1.2726,
+      "step": 27
+    },
+    {
+      "epoch": 0.04057971014492753,
+      "grad_norm": 1.1922376154128103,
+      "learning_rate": 6.7632850241545894e-06,
+      "loss": 1.3099,
+      "step": 28
+    },
+    {
+      "epoch": 0.042028985507246375,
+      "grad_norm": 0.8946299762295795,
+      "learning_rate": 7.004830917874397e-06,
+      "loss": 1.0997,
+      "step": 29
+    },
+    {
+      "epoch": 0.043478260869565216,
+      "grad_norm": 0.9128998822679193,
+      "learning_rate": 7.246376811594203e-06,
+      "loss": 1.1738,
+      "step": 30
+    },
+    {
+      "epoch": 0.04492753623188406,
+      "grad_norm": 0.8261962791615375,
+      "learning_rate": 7.48792270531401e-06,
+      "loss": 1.1265,
+      "step": 31
+    },
+    {
+      "epoch": 0.0463768115942029,
+      "grad_norm": 0.7259010338295714,
+      "learning_rate": 7.729468599033817e-06,
+      "loss": 1.0761,
+      "step": 32
+    },
+    {
+      "epoch": 0.04782608695652174,
+      "grad_norm": 0.8144150263048314,
+      "learning_rate": 7.971014492753623e-06,
+      "loss": 1.1418,
+      "step": 33
+    },
+    {
+      "epoch": 0.04927536231884058,
+      "grad_norm": 0.6225984168130394,
+      "learning_rate": 8.212560386473431e-06,
+      "loss": 1.0675,
+      "step": 34
+    },
+    {
+      "epoch": 0.050724637681159424,
+      "grad_norm": 0.6957019400861381,
+      "learning_rate": 8.454106280193238e-06,
+      "loss": 1.1917,
+      "step": 35
+    },
+    {
+      "epoch": 0.05217391304347826,
+      "grad_norm": 0.6473409624233152,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 0.9325,
+      "step": 36
+    },
+    {
+      "epoch": 0.0536231884057971,
+      "grad_norm": 0.6816748318384368,
+      "learning_rate": 8.93719806763285e-06,
+      "loss": 1.2146,
+      "step": 37
+    },
+    {
+      "epoch": 0.05507246376811594,
+      "grad_norm": 0.6277804431834189,
+      "learning_rate": 9.178743961352658e-06,
+      "loss": 1.0025,
+      "step": 38
+    },
+    {
+      "epoch": 0.05652173913043478,
+      "grad_norm": 0.5421011053554157,
+      "learning_rate": 9.420289855072464e-06,
+      "loss": 1.0601,
+      "step": 39
+    },
+    {
+      "epoch": 0.057971014492753624,
+      "grad_norm": 0.5106889084966296,
+      "learning_rate": 9.66183574879227e-06,
+      "loss": 0.9464,
+      "step": 40
+    },
+    {
+      "epoch": 0.059420289855072465,
+      "grad_norm": 0.5149567843509089,
+      "learning_rate": 9.903381642512077e-06,
+      "loss": 1.0016,
+      "step": 41
+    },
+    {
+      "epoch": 0.06086956521739131,
+      "grad_norm": 0.5161093222444725,
+      "learning_rate": 1.0144927536231885e-05,
+      "loss": 0.9742,
+      "step": 42
+    },
+    {
+      "epoch": 0.06231884057971015,
+      "grad_norm": 0.4872914232682372,
+      "learning_rate": 1.0386473429951692e-05,
+      "loss": 0.9586,
+      "step": 43
+    },
+    {
+      "epoch": 0.06376811594202898,
+      "grad_norm": 0.5184224626160103,
+      "learning_rate": 1.0628019323671499e-05,
+      "loss": 0.9966,
+      "step": 44
+    },
+    {
+      "epoch": 0.06521739130434782,
+      "grad_norm": 0.43452221873592267,
+      "learning_rate": 1.0869565217391305e-05,
+      "loss": 1.0099,
+      "step": 45
+    },
+    {
+      "epoch": 0.06666666666666667,
+      "grad_norm": 0.48744214274446906,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.9639,
+      "step": 46
+    },
+    {
+      "epoch": 0.06811594202898551,
+      "grad_norm": 7.312818671660048,
+      "learning_rate": 1.1352657004830918e-05,
+      "loss": 0.995,
+      "step": 47
+    },
+    {
+      "epoch": 0.06956521739130435,
+      "grad_norm": 0.541581425101709,
+      "learning_rate": 1.1594202898550725e-05,
+      "loss": 0.9325,
+      "step": 48
+    },
+    {
+      "epoch": 0.07101449275362319,
+      "grad_norm": 0.4503734798990069,
+      "learning_rate": 1.1835748792270531e-05,
+      "loss": 0.868,
+      "step": 49
+    },
+    {
+      "epoch": 0.07246376811594203,
+      "grad_norm": 0.4886820292319067,
+      "learning_rate": 1.2077294685990338e-05,
+      "loss": 0.9599,
+      "step": 50
+    },
+    {
+      "epoch": 0.07391304347826087,
+      "grad_norm": 0.42811723185783795,
+      "learning_rate": 1.2318840579710146e-05,
+      "loss": 0.8892,
+      "step": 51
+    },
+    {
+      "epoch": 0.07536231884057971,
+      "grad_norm": 0.42226359715459455,
+      "learning_rate": 1.2560386473429953e-05,
+      "loss": 0.9332,
+      "step": 52
+    },
+    {
+      "epoch": 0.07681159420289856,
+      "grad_norm": 0.41341607893203236,
+      "learning_rate": 1.2801932367149761e-05,
+      "loss": 0.8111,
+      "step": 53
+    },
+    {
+      "epoch": 0.0782608695652174,
+      "grad_norm": 0.407434689365239,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 0.9552,
+      "step": 54
+    },
+    {
+      "epoch": 0.07971014492753623,
+      "grad_norm": 0.3510103557036314,
+      "learning_rate": 1.3285024154589374e-05,
+      "loss": 0.9455,
+      "step": 55
+    },
+    {
+      "epoch": 0.08115942028985507,
+      "grad_norm": 0.4560728778977622,
+      "learning_rate": 1.3526570048309179e-05,
+      "loss": 0.8894,
+      "step": 56
+    },
+    {
+      "epoch": 0.08260869565217391,
+      "grad_norm": 0.36250775032726457,
+      "learning_rate": 1.3768115942028985e-05,
+      "loss": 0.7497,
+      "step": 57
+    },
+    {
+      "epoch": 0.08405797101449275,
+      "grad_norm": 0.3762659087593294,
+      "learning_rate": 1.4009661835748794e-05,
+      "loss": 0.8474,
+      "step": 58
+    },
+    {
+      "epoch": 0.08550724637681159,
+      "grad_norm": 0.36500198106850623,
+      "learning_rate": 1.4251207729468599e-05,
+      "loss": 0.8414,
+      "step": 59
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 0.3293408454439321,
+      "learning_rate": 1.4492753623188407e-05,
+      "loss": 0.7616,
+      "step": 60
+    },
+    {
+      "epoch": 0.08840579710144927,
+      "grad_norm": 0.3510265716577906,
+      "learning_rate": 1.4734299516908212e-05,
+      "loss": 0.847,
+      "step": 61
+    },
+    {
+      "epoch": 0.08985507246376812,
+      "grad_norm": 0.3348155244843317,
+      "learning_rate": 1.497584541062802e-05,
+      "loss": 0.7797,
+      "step": 62
+    },
+    {
+      "epoch": 0.09130434782608696,
+      "grad_norm": 0.35285060969891996,
+      "learning_rate": 1.5217391304347828e-05,
+      "loss": 0.9457,
+      "step": 63
+    },
+    {
+      "epoch": 0.0927536231884058,
+      "grad_norm": 0.34433082535838777,
+      "learning_rate": 1.5458937198067633e-05,
+      "loss": 0.9697,
+      "step": 64
+    },
+    {
+      "epoch": 0.09420289855072464,
+      "grad_norm": 0.9324600519344617,
+      "learning_rate": 1.570048309178744e-05,
+      "loss": 0.9696,
+      "step": 65
+    },
+    {
+      "epoch": 0.09565217391304348,
+      "grad_norm": 0.331145083306618,
+      "learning_rate": 1.5942028985507246e-05,
+      "loss": 0.8468,
+      "step": 66
+    },
+    {
+      "epoch": 0.09710144927536232,
+      "grad_norm": 0.3959920805542888,
+      "learning_rate": 1.6183574879227054e-05,
+      "loss": 0.9872,
+      "step": 67
+    },
+    {
+      "epoch": 0.09855072463768116,
+      "grad_norm": 0.3398180883831577,
+      "learning_rate": 1.6425120772946863e-05,
+      "loss": 0.9103,
+      "step": 68
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3507626031123879,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.9097,
+      "step": 69
+    },
+    {
+      "epoch": 0.10144927536231885,
+      "grad_norm": 0.31255794956434924,
+      "learning_rate": 1.6908212560386476e-05,
+      "loss": 0.7564,
+      "step": 70
+    },
+    {
+      "epoch": 0.10289855072463767,
+      "grad_norm": 0.37074114591156665,
+      "learning_rate": 1.714975845410628e-05,
+      "loss": 0.8267,
+      "step": 71
+    },
+    {
+      "epoch": 0.10434782608695652,
+      "grad_norm": 0.3656485623827102,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 0.8924,
+      "step": 72
+    },
+    {
+      "epoch": 0.10579710144927536,
+      "grad_norm": 0.3226469968742567,
+      "learning_rate": 1.7632850241545894e-05,
+      "loss": 0.6622,
+      "step": 73
+    },
+    {
+      "epoch": 0.1072463768115942,
+      "grad_norm": 0.33712867700768034,
+      "learning_rate": 1.78743961352657e-05,
+      "loss": 0.7811,
+      "step": 74
+    },
+    {
+      "epoch": 0.10869565217391304,
+      "grad_norm": 0.34247268410396536,
+      "learning_rate": 1.8115942028985507e-05,
+      "loss": 0.7057,
+      "step": 75
+    },
+    {
+      "epoch": 0.11014492753623188,
+      "grad_norm": 0.3354659334989842,
+      "learning_rate": 1.8357487922705315e-05,
+      "loss": 0.8827,
+      "step": 76
+    },
+    {
+      "epoch": 0.11159420289855072,
+      "grad_norm": 0.3070922646052218,
+      "learning_rate": 1.859903381642512e-05,
+      "loss": 0.7294,
+      "step": 77
+    },
+    {
+      "epoch": 0.11304347826086956,
+      "grad_norm": 0.3138658882523287,
+      "learning_rate": 1.8840579710144928e-05,
+      "loss": 0.6904,
+      "step": 78
+    },
+    {
+      "epoch": 0.1144927536231884,
+      "grad_norm": 1.5216350994227146,
+      "learning_rate": 1.9082125603864733e-05,
+      "loss": 0.803,
+      "step": 79
+    },
+    {
+      "epoch": 0.11594202898550725,
+      "grad_norm": 0.3382431707970964,
+      "learning_rate": 1.932367149758454e-05,
+      "loss": 0.7899,
+      "step": 80
+    },
+    {
+      "epoch": 0.11739130434782609,
+      "grad_norm": 0.296008590542212,
+      "learning_rate": 1.956521739130435e-05,
+      "loss": 0.8161,
+      "step": 81
+    },
+    {
+      "epoch": 0.11884057971014493,
+      "grad_norm": 0.3689599195710936,
+      "learning_rate": 1.9806763285024154e-05,
+      "loss": 0.873,
+      "step": 82
+    },
+    {
+      "epoch": 0.12028985507246377,
+      "grad_norm": 0.3121552852325846,
+      "learning_rate": 2.0048309178743963e-05,
+      "loss": 0.7462,
+      "step": 83
+    },
+    {
+      "epoch": 0.12173913043478261,
+      "grad_norm": 0.3548758204484305,
+      "learning_rate": 2.028985507246377e-05,
+      "loss": 0.9198,
+      "step": 84
+    },
+    {
+      "epoch": 0.12318840579710146,
+      "grad_norm": 0.30541661182285873,
+      "learning_rate": 2.0531400966183576e-05,
+      "loss": 0.8544,
+      "step": 85
+    },
+    {
+      "epoch": 0.1246376811594203,
+      "grad_norm": 0.29774802882201346,
+      "learning_rate": 2.0772946859903384e-05,
+      "loss": 0.7522,
+      "step": 86
+    },
+    {
+      "epoch": 0.12608695652173912,
+      "grad_norm": 0.33320845974248064,
+      "learning_rate": 2.101449275362319e-05,
+      "loss": 0.7726,
+      "step": 87
+    },
+    {
+      "epoch": 0.12753623188405797,
+      "grad_norm": 0.305867939123411,
+      "learning_rate": 2.1256038647342997e-05,
+      "loss": 0.6971,
+      "step": 88
+    },
+    {
+      "epoch": 0.1289855072463768,
+      "grad_norm": 0.3687626183800535,
+      "learning_rate": 2.1497584541062805e-05,
+      "loss": 0.7586,
+      "step": 89
+    },
+    {
+      "epoch": 0.13043478260869565,
+      "grad_norm": 0.40655909646647576,
+      "learning_rate": 2.173913043478261e-05,
+      "loss": 0.86,
+      "step": 90
+    },
+    {
+      "epoch": 0.1318840579710145,
+      "grad_norm": 0.3184064183198976,
+      "learning_rate": 2.198067632850242e-05,
+      "loss": 0.7743,
+      "step": 91
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 4.585823826697241,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.8626,
+      "step": 92
+    },
+    {
+      "epoch": 0.13478260869565217,
+      "grad_norm": 0.3487306040411324,
+      "learning_rate": 2.246376811594203e-05,
+      "loss": 0.8454,
+      "step": 93
+    },
+    {
+      "epoch": 0.13623188405797101,
+      "grad_norm": 0.37267592174262365,
+      "learning_rate": 2.2705314009661836e-05,
+      "loss": 0.8255,
+      "step": 94
+    },
+    {
+      "epoch": 0.13768115942028986,
+      "grad_norm": 0.32986815030036337,
+      "learning_rate": 2.294685990338164e-05,
+      "loss": 0.7607,
+      "step": 95
+    },
+    {
+      "epoch": 0.1391304347826087,
+      "grad_norm": 0.36689670642754957,
+      "learning_rate": 2.318840579710145e-05,
+      "loss": 0.7595,
+      "step": 96
+    },
+    {
+      "epoch": 0.14057971014492754,
+      "grad_norm": 0.3924500092697602,
+      "learning_rate": 2.3429951690821258e-05,
+      "loss": 0.7641,
+      "step": 97
+    },
+    {
+      "epoch": 0.14202898550724638,
+      "grad_norm": 0.3678298491507766,
+      "learning_rate": 2.3671497584541063e-05,
+      "loss": 0.8308,
+      "step": 98
+    },
+    {
+      "epoch": 0.14347826086956522,
+      "grad_norm": 0.4714024681413323,
+      "learning_rate": 2.391304347826087e-05,
+      "loss": 0.6769,
+      "step": 99
+    },
+    {
+      "epoch": 0.14492753623188406,
+      "grad_norm": 0.4412843647168353,
+      "learning_rate": 2.4154589371980676e-05,
+      "loss": 0.9089,
+      "step": 100
+    },
+    {
+      "epoch": 0.1463768115942029,
+      "grad_norm": 0.3625778040519065,
+      "learning_rate": 2.4396135265700484e-05,
+      "loss": 0.7376,
+      "step": 101
+    },
+    {
+      "epoch": 0.14782608695652175,
+      "grad_norm": 0.4200692042032697,
+      "learning_rate": 2.4637681159420292e-05,
+      "loss": 0.7893,
+      "step": 102
+    },
+    {
+      "epoch": 0.1492753623188406,
+      "grad_norm": 5.300388535004034,
+      "learning_rate": 2.4879227053140097e-05,
+      "loss": 0.7625,
+      "step": 103
+    },
+    {
+      "epoch": 0.15072463768115943,
+      "grad_norm": 0.5161561357287162,
+      "learning_rate": 2.5120772946859905e-05,
+      "loss": 0.8625,
+      "step": 104
+    },
+    {
+      "epoch": 0.15217391304347827,
+      "grad_norm": 0.35899298752423286,
+      "learning_rate": 2.5362318840579714e-05,
+      "loss": 0.7075,
+      "step": 105
+    },
+    {
+      "epoch": 0.1536231884057971,
+      "grad_norm": 0.436644448478288,
+      "learning_rate": 2.5603864734299522e-05,
+      "loss": 0.6964,
+      "step": 106
+    },
+    {
+      "epoch": 0.15507246376811595,
+      "grad_norm": 0.35851332585975537,
+      "learning_rate": 2.5845410628019323e-05,
+      "loss": 0.7348,
+      "step": 107
+    },
+    {
+      "epoch": 0.1565217391304348,
+      "grad_norm": 0.3838152624994046,
+      "learning_rate": 2.608695652173913e-05,
+      "loss": 0.8987,
+      "step": 108
+    },
+    {
+      "epoch": 0.15797101449275364,
+      "grad_norm": 0.3343603849395053,
+      "learning_rate": 2.632850241545894e-05,
+      "loss": 0.6829,
+      "step": 109
+    },
+    {
+      "epoch": 0.15942028985507245,
+      "grad_norm": 0.35962622420671014,
+      "learning_rate": 2.6570048309178748e-05,
+      "loss": 0.7414,
+      "step": 110
+    },
+    {
+      "epoch": 0.1608695652173913,
+      "grad_norm": 0.3757160134118775,
+      "learning_rate": 2.6811594202898553e-05,
+      "loss": 0.7767,
+      "step": 111
+    },
+    {
+      "epoch": 0.16231884057971013,
+      "grad_norm": 0.3469599966973019,
+      "learning_rate": 2.7053140096618358e-05,
+      "loss": 0.7929,
+      "step": 112
+    },
+    {
+      "epoch": 0.16376811594202897,
+      "grad_norm": 0.43902244288916725,
+      "learning_rate": 2.7294685990338166e-05,
+      "loss": 0.6906,
+      "step": 113
+    },
+    {
+      "epoch": 0.16521739130434782,
+      "grad_norm": 0.36775564724113863,
+      "learning_rate": 2.753623188405797e-05,
+      "loss": 0.8779,
+      "step": 114
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.4188102948097823,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.6696,
+      "step": 115
+    },
+    {
+      "epoch": 0.1681159420289855,
+      "grad_norm": 0.3387805292980607,
+      "learning_rate": 2.8019323671497587e-05,
+      "loss": 0.7831,
+      "step": 116
+    },
+    {
+      "epoch": 0.16956521739130434,
+      "grad_norm": 0.3598515255578289,
+      "learning_rate": 2.826086956521739e-05,
+      "loss": 0.7654,
+      "step": 117
+    },
+    {
+      "epoch": 0.17101449275362318,
+      "grad_norm": 0.39027009772473525,
+      "learning_rate": 2.8502415458937197e-05,
+      "loss": 0.6693,
+      "step": 118
+    },
+    {
+      "epoch": 0.17246376811594202,
+      "grad_norm": 0.3094539572612201,
+      "learning_rate": 2.8743961352657005e-05,
+      "loss": 0.6016,
+      "step": 119
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 0.42058909226444235,
+      "learning_rate": 2.8985507246376814e-05,
+      "loss": 0.6962,
+      "step": 120
+    },
+    {
+      "epoch": 0.1753623188405797,
+      "grad_norm": 0.34853589375868377,
+      "learning_rate": 2.9227053140096622e-05,
+      "loss": 0.7216,
+      "step": 121
+    },
+    {
+      "epoch": 0.17681159420289855,
+      "grad_norm": 0.4868735037868742,
+      "learning_rate": 2.9468599033816423e-05,
+      "loss": 0.7466,
+      "step": 122
+    },
+    {
+      "epoch": 0.1782608695652174,
+      "grad_norm": 0.4185621046394828,
+      "learning_rate": 2.971014492753623e-05,
+      "loss": 0.701,
+      "step": 123
+    },
+    {
+      "epoch": 0.17971014492753623,
+      "grad_norm": 0.4197568084080103,
+      "learning_rate": 2.995169082125604e-05,
+      "loss": 0.7263,
+      "step": 124
+    },
+    {
+      "epoch": 0.18115942028985507,
+      "grad_norm": 0.4657689848243366,
+      "learning_rate": 3.0193236714975848e-05,
+      "loss": 0.654,
+      "step": 125
+    },
+    {
+      "epoch": 0.1826086956521739,
+      "grad_norm": 0.37793581378328184,
+      "learning_rate": 3.0434782608695656e-05,
+      "loss": 0.6564,
+      "step": 126
+    },
+    {
+      "epoch": 0.18405797101449275,
+      "grad_norm": 0.434674876337351,
+      "learning_rate": 3.067632850241546e-05,
+      "loss": 0.7689,
+      "step": 127
+    },
+    {
+      "epoch": 0.1855072463768116,
+      "grad_norm": 0.48900991299205354,
+      "learning_rate": 3.0917874396135266e-05,
+      "loss": 0.7707,
+      "step": 128
+    },
+    {
+      "epoch": 0.18695652173913044,
+      "grad_norm": 0.5447607203343757,
+      "learning_rate": 3.1159420289855074e-05,
+      "loss": 0.6964,
+      "step": 129
+    },
+    {
+      "epoch": 0.18840579710144928,
+      "grad_norm": 0.4881199429761279,
+      "learning_rate": 3.140096618357488e-05,
+      "loss": 0.6979,
+      "step": 130
+    },
+    {
+      "epoch": 0.18985507246376812,
+      "grad_norm": 3.6888624601856974,
+      "learning_rate": 3.164251207729469e-05,
+      "loss": 0.7116,
+      "step": 131
+    },
+    {
+      "epoch": 0.19130434782608696,
+      "grad_norm": 0.8067506542863169,
+      "learning_rate": 3.188405797101449e-05,
+      "loss": 0.684,
+      "step": 132
+    },
+    {
+      "epoch": 0.1927536231884058,
+      "grad_norm": 0.4346182081032241,
+      "learning_rate": 3.21256038647343e-05,
+      "loss": 0.7532,
+      "step": 133
+    },
+    {
+      "epoch": 0.19420289855072465,
+      "grad_norm": 0.6270064201343746,
+      "learning_rate": 3.236714975845411e-05,
+      "loss": 0.6986,
+      "step": 134
+    },
+    {
+      "epoch": 0.1956521739130435,
+      "grad_norm": 0.4928401828984155,
+      "learning_rate": 3.260869565217392e-05,
+      "loss": 0.7408,
+      "step": 135
+    },
+    {
+      "epoch": 0.19710144927536233,
+      "grad_norm": 0.47674959216585067,
+      "learning_rate": 3.2850241545893725e-05,
+      "loss": 0.6962,
+      "step": 136
+    },
+    {
+      "epoch": 0.19855072463768117,
+      "grad_norm": 0.5262194433847822,
+      "learning_rate": 3.3091787439613533e-05,
+      "loss": 0.6421,
+      "step": 137
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5736351844030817,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.7716,
+      "step": 138
+    },
+    {
+      "epoch": 0.20144927536231885,
+      "grad_norm": 0.5025434605529145,
+      "learning_rate": 3.357487922705314e-05,
+      "loss": 0.7522,
+      "step": 139
+    },
+    {
+      "epoch": 0.2028985507246377,
+      "grad_norm": 0.5074291606372959,
+      "learning_rate": 3.381642512077295e-05,
+      "loss": 0.8226,
+      "step": 140
+    },
+    {
+      "epoch": 0.20434782608695654,
+      "grad_norm": 0.3974703015679785,
+      "learning_rate": 3.405797101449276e-05,
+      "loss": 0.7535,
+      "step": 141
+    },
+    {
+      "epoch": 0.20579710144927535,
+      "grad_norm": 0.45467619929714587,
+      "learning_rate": 3.429951690821256e-05,
+      "loss": 0.7729,
+      "step": 142
+    },
+    {
+      "epoch": 0.2072463768115942,
+      "grad_norm": 0.4084485817949974,
+      "learning_rate": 3.454106280193237e-05,
+      "loss": 0.7303,
+      "step": 143
+    },
+    {
+      "epoch": 0.20869565217391303,
+      "grad_norm": 0.41159007002019965,
+      "learning_rate": 3.478260869565218e-05,
+      "loss": 0.7145,
+      "step": 144
+    },
+    {
+      "epoch": 0.21014492753623187,
+      "grad_norm": 0.4973545025240137,
+      "learning_rate": 3.502415458937198e-05,
+      "loss": 0.7747,
+      "step": 145
+    },
+    {
+      "epoch": 0.21159420289855072,
+      "grad_norm": 0.3734469069038325,
+      "learning_rate": 3.526570048309179e-05,
+      "loss": 0.6809,
+      "step": 146
+    },
+    {
+      "epoch": 0.21304347826086956,
+      "grad_norm": 0.4058026991714768,
+      "learning_rate": 3.5507246376811596e-05,
+      "loss": 0.662,
+      "step": 147
+    },
+    {
+      "epoch": 0.2144927536231884,
+      "grad_norm": 1.0006350138668159,
+      "learning_rate": 3.57487922705314e-05,
+      "loss": 0.7111,
+      "step": 148
+    },
+    {
+      "epoch": 0.21594202898550724,
+      "grad_norm": 0.36130905180297707,
+      "learning_rate": 3.5990338164251205e-05,
+      "loss": 0.8164,
+      "step": 149
+    },
+    {
+      "epoch": 0.21739130434782608,
+      "grad_norm": 0.4391335573453258,
+      "learning_rate": 3.6231884057971014e-05,
+      "loss": 0.6806,
+      "step": 150
+    },
+    {
+      "epoch": 0.21884057971014492,
+      "grad_norm": 0.36378332102331984,
+      "learning_rate": 3.647342995169082e-05,
+      "loss": 0.672,
+      "step": 151
+    },
+    {
+      "epoch": 0.22028985507246376,
+      "grad_norm": 0.36408409425201227,
+      "learning_rate": 3.671497584541063e-05,
+      "loss": 0.6756,
+      "step": 152
+    },
+    {
+      "epoch": 0.2217391304347826,
+      "grad_norm": 0.3645988836273659,
+      "learning_rate": 3.695652173913043e-05,
+      "loss": 0.5685,
+      "step": 153
+    },
+    {
+      "epoch": 0.22318840579710145,
+      "grad_norm": 0.42301410248495724,
+      "learning_rate": 3.719806763285024e-05,
+      "loss": 0.6425,
+      "step": 154
+    },
+    {
+      "epoch": 0.2246376811594203,
+      "grad_norm": 0.363740011337443,
+      "learning_rate": 3.743961352657005e-05,
+      "loss": 0.6702,
+      "step": 155
+    },
+    {
+      "epoch": 0.22608695652173913,
+      "grad_norm": 0.4045601971692223,
+      "learning_rate": 3.7681159420289856e-05,
+      "loss": 0.6256,
+      "step": 156
+    },
+    {
+      "epoch": 0.22753623188405797,
+      "grad_norm": 0.41143862371080964,
+      "learning_rate": 3.7922705314009665e-05,
+      "loss": 0.7695,
+      "step": 157
+    },
+    {
+      "epoch": 0.2289855072463768,
+      "grad_norm": 0.3951097674703999,
+      "learning_rate": 3.8164251207729466e-05,
+      "loss": 0.7124,
+      "step": 158
+    },
+    {
+      "epoch": 0.23043478260869565,
+      "grad_norm": 0.3831302126355593,
+      "learning_rate": 3.8405797101449274e-05,
+      "loss": 0.746,
+      "step": 159
+    },
+    {
+      "epoch": 0.2318840579710145,
+      "grad_norm": 0.37142247727064054,
+      "learning_rate": 3.864734299516908e-05,
+      "loss": 0.726,
+      "step": 160
+    },
+    {
+      "epoch": 0.23333333333333334,
+      "grad_norm": 0.35813121793484276,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.597,
+      "step": 161
+    },
+    {
+      "epoch": 0.23478260869565218,
+      "grad_norm": 0.44255790588313515,
+      "learning_rate": 3.91304347826087e-05,
+      "loss": 0.6954,
+      "step": 162
+    },
+    {
+      "epoch": 0.23623188405797102,
+      "grad_norm": 0.3727301373909844,
+      "learning_rate": 3.937198067632851e-05,
+      "loss": 0.689,
+      "step": 163
+    },
+    {
+      "epoch": 0.23768115942028986,
+      "grad_norm": 0.48116777299429975,
+      "learning_rate": 3.961352657004831e-05,
+      "loss": 0.7812,
+      "step": 164
+    },
+    {
+      "epoch": 0.2391304347826087,
+      "grad_norm": 0.3455365306611197,
+      "learning_rate": 3.985507246376812e-05,
+      "loss": 0.7201,
+      "step": 165
+    },
+    {
+      "epoch": 0.24057971014492754,
+      "grad_norm": 0.48059498778631815,
+      "learning_rate": 4.0096618357487925e-05,
+      "loss": 0.7336,
+      "step": 166
+    },
+    {
+      "epoch": 0.24202898550724639,
+      "grad_norm": 0.3664497120347036,
+      "learning_rate": 4.0338164251207733e-05,
+      "loss": 0.6416,
+      "step": 167
+    },
+    {
+      "epoch": 0.24347826086956523,
+      "grad_norm": 0.5063263979771799,
+      "learning_rate": 4.057971014492754e-05,
+      "loss": 0.6918,
+      "step": 168
+    },
+    {
+      "epoch": 0.24492753623188407,
+      "grad_norm": 0.4636598172748528,
+      "learning_rate": 4.082125603864734e-05,
+      "loss": 0.7861,
+      "step": 169
+    },
+    {
+      "epoch": 0.2463768115942029,
+      "grad_norm": 0.5939118920382572,
+      "learning_rate": 4.106280193236715e-05,
+      "loss": 0.731,
+      "step": 170
+    },
+    {
+      "epoch": 0.24782608695652175,
+      "grad_norm": 0.70082387818902,
+      "learning_rate": 4.130434782608696e-05,
+      "loss": 0.7278,
+      "step": 171
+    },
+    {
+      "epoch": 0.2492753623188406,
+      "grad_norm": 0.4638036883744111,
+      "learning_rate": 4.154589371980677e-05,
+      "loss": 0.6726,
+      "step": 172
+    },
+    {
+      "epoch": 0.25072463768115943,
+      "grad_norm": 0.48095533230230175,
+      "learning_rate": 4.1787439613526576e-05,
+      "loss": 0.6673,
+      "step": 173
+    },
+    {
+      "epoch": 0.25217391304347825,
+      "grad_norm": 0.4690837413533762,
+      "learning_rate": 4.202898550724638e-05,
+      "loss": 0.653,
+      "step": 174
+    },
+    {
+      "epoch": 0.2536231884057971,
+      "grad_norm": 0.4458627462261377,
+      "learning_rate": 4.2270531400966186e-05,
+      "loss": 0.7793,
+      "step": 175
+    },
+    {
+      "epoch": 0.25507246376811593,
+      "grad_norm": 0.4189884078798217,
+      "learning_rate": 4.2512077294685994e-05,
+      "loss": 0.7243,
+      "step": 176
+    },
+    {
+      "epoch": 0.2565217391304348,
+      "grad_norm": 0.4651076969193504,
+      "learning_rate": 4.27536231884058e-05,
+      "loss": 0.6708,
+      "step": 177
+    },
+    {
+      "epoch": 0.2579710144927536,
+      "grad_norm": 0.42064317748362867,
+      "learning_rate": 4.299516908212561e-05,
+      "loss": 0.6885,
+      "step": 178
+    },
+    {
+      "epoch": 0.2594202898550725,
+      "grad_norm": 0.57387087269488,
+      "learning_rate": 4.323671497584541e-05,
+      "loss": 0.7926,
+      "step": 179
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 0.5215829127359812,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 0.7087,
+      "step": 180
+    },
+    {
+      "epoch": 0.26231884057971017,
+      "grad_norm": 0.5244092130860797,
+      "learning_rate": 4.371980676328503e-05,
+      "loss": 0.6009,
+      "step": 181
+    },
+    {
+      "epoch": 0.263768115942029,
+      "grad_norm": 0.40034582164953714,
+      "learning_rate": 4.396135265700484e-05,
+      "loss": 0.7362,
+      "step": 182
+    },
+    {
+      "epoch": 0.26521739130434785,
+      "grad_norm": 0.5219605123880174,
+      "learning_rate": 4.4202898550724645e-05,
+      "loss": 0.6995,
+      "step": 183
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 0.47609730352611157,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.7462,
+      "step": 184
+    },
+    {
+      "epoch": 0.26811594202898553,
+      "grad_norm": 0.4763490286301335,
+      "learning_rate": 4.4685990338164255e-05,
+      "loss": 0.7037,
+      "step": 185
+    },
+    {
+      "epoch": 0.26956521739130435,
+      "grad_norm": 0.4654352184388157,
+      "learning_rate": 4.492753623188406e-05,
+      "loss": 0.7397,
+      "step": 186
+    },
+    {
+      "epoch": 0.2710144927536232,
+      "grad_norm": 0.5309678745932885,
+      "learning_rate": 4.5169082125603865e-05,
+      "loss": 0.7948,
+      "step": 187
+    },
+    {
+      "epoch": 0.27246376811594203,
+      "grad_norm": 0.4357391853306629,
+      "learning_rate": 4.541062801932367e-05,
+      "loss": 0.7201,
+      "step": 188
+    },
+    {
+      "epoch": 0.27391304347826084,
+      "grad_norm": 0.495806282088296,
+      "learning_rate": 4.565217391304348e-05,
+      "loss": 0.7081,
+      "step": 189
+    },
+    {
+      "epoch": 0.2753623188405797,
+      "grad_norm": 0.4734024179586802,
+      "learning_rate": 4.589371980676328e-05,
+      "loss": 0.5961,
+      "step": 190
+    },
+    {
+      "epoch": 0.2768115942028985,
+      "grad_norm": 0.5213770810315379,
+      "learning_rate": 4.613526570048309e-05,
+      "loss": 0.6825,
+      "step": 191
+    },
+    {
+      "epoch": 0.2782608695652174,
+      "grad_norm": 0.45246037733085614,
+      "learning_rate": 4.63768115942029e-05,
+      "loss": 0.6037,
+      "step": 192
+    },
+    {
+      "epoch": 0.2797101449275362,
+      "grad_norm": 0.6271967081898506,
+      "learning_rate": 4.661835748792271e-05,
+      "loss": 0.7358,
+      "step": 193
+    },
+    {
+      "epoch": 0.2811594202898551,
+      "grad_norm": 0.857140335440062,
+      "learning_rate": 4.6859903381642516e-05,
+      "loss": 0.7949,
+      "step": 194
+    },
+    {
+      "epoch": 0.2826086956521739,
+      "grad_norm": 0.9127639840782331,
+      "learning_rate": 4.710144927536232e-05,
+      "loss": 0.6545,
+      "step": 195
+    },
+    {
+      "epoch": 0.28405797101449276,
+      "grad_norm": 0.4959655662040491,
+      "learning_rate": 4.7342995169082125e-05,
+      "loss": 0.6034,
+      "step": 196
+    },
+    {
+      "epoch": 0.2855072463768116,
+      "grad_norm": 0.5178854616939212,
+      "learning_rate": 4.7584541062801933e-05,
+      "loss": 0.6657,
+      "step": 197
+    },
+    {
+      "epoch": 0.28695652173913044,
+      "grad_norm": 0.5052219461873171,
+      "learning_rate": 4.782608695652174e-05,
+      "loss": 0.6443,
+      "step": 198
+    },
+    {
+      "epoch": 0.28840579710144926,
+      "grad_norm": 0.4614518423922908,
+      "learning_rate": 4.806763285024155e-05,
+      "loss": 0.7269,
+      "step": 199
+    },
+    {
+      "epoch": 0.2898550724637681,
+      "grad_norm": 0.4586802080175379,
+      "learning_rate": 4.830917874396135e-05,
+      "loss": 0.7065,
+      "step": 200
+    },
+    {
+      "epoch": 0.29130434782608694,
+      "grad_norm": 0.4666949971627673,
+      "learning_rate": 4.855072463768116e-05,
+      "loss": 0.6757,
+      "step": 201
+    },
+    {
+      "epoch": 0.2927536231884058,
+      "grad_norm": 0.622849855674404,
+      "learning_rate": 4.879227053140097e-05,
+      "loss": 0.7404,
+      "step": 202
+    },
+    {
+      "epoch": 0.2942028985507246,
+      "grad_norm": 0.621911112221847,
+      "learning_rate": 4.9033816425120776e-05,
+      "loss": 0.7203,
+      "step": 203
+    },
+    {
+      "epoch": 0.2956521739130435,
+      "grad_norm": 0.42462020266152434,
+      "learning_rate": 4.9275362318840584e-05,
+      "loss": 0.7221,
+      "step": 204
+    },
+    {
+      "epoch": 0.2971014492753623,
+      "grad_norm": 0.7157909547004069,
+      "learning_rate": 4.9516908212560386e-05,
+      "loss": 0.7178,
+      "step": 205
+    },
+    {
+      "epoch": 0.2985507246376812,
+      "grad_norm": 0.47540747142242634,
+      "learning_rate": 4.9758454106280194e-05,
+      "loss": 0.687,
+      "step": 206
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5356933660212242,
+      "learning_rate": 5e-05,
+      "loss": 0.5966,
+      "step": 207
+    },
+    {
+      "epoch": 0.30144927536231886,
+      "grad_norm": 3.875707624149591,
+      "learning_rate": 4.9973161567364465e-05,
+      "loss": 0.7651,
+      "step": 208
+    },
+    {
+      "epoch": 0.30289855072463767,
+      "grad_norm": 0.7348013333692176,
+      "learning_rate": 4.9946323134728935e-05,
+      "loss": 0.7256,
+      "step": 209
+    },
+    {
+      "epoch": 0.30434782608695654,
+      "grad_norm": 0.552654765255837,
+      "learning_rate": 4.99194847020934e-05,
+      "loss": 0.6954,
+      "step": 210
+    },
+    {
+      "epoch": 0.30579710144927535,
+      "grad_norm": 0.7049833226099168,
+      "learning_rate": 4.989264626945787e-05,
+      "loss": 0.7656,
+      "step": 211
+    },
+    {
+      "epoch": 0.3072463768115942,
+      "grad_norm": 0.5935672598341638,
+      "learning_rate": 4.986580783682233e-05,
+      "loss": 0.6958,
+      "step": 212
+    },
+    {
+      "epoch": 0.30869565217391304,
+      "grad_norm": 0.6495195288245664,
+      "learning_rate": 4.98389694041868e-05,
+      "loss": 0.6997,
+      "step": 213
+    },
+    {
+      "epoch": 0.3101449275362319,
+      "grad_norm": 0.7912017825608941,
+      "learning_rate": 4.981213097155126e-05,
+      "loss": 0.7651,
+      "step": 214
+    },
+    {
+      "epoch": 0.3115942028985507,
+      "grad_norm": 0.6081697492207544,
+      "learning_rate": 4.978529253891573e-05,
+      "loss": 0.741,
+      "step": 215
+    },
+    {
+      "epoch": 0.3130434782608696,
+      "grad_norm": 0.5237190794202718,
+      "learning_rate": 4.9758454106280194e-05,
+      "loss": 0.6351,
+      "step": 216
+    },
+    {
+      "epoch": 0.3144927536231884,
+      "grad_norm": 0.6124946170096812,
+      "learning_rate": 4.9731615673644664e-05,
+      "loss": 0.6809,
+      "step": 217
+    },
+    {
+      "epoch": 0.3159420289855073,
+      "grad_norm": 0.5039329545767305,
+      "learning_rate": 4.9704777241009126e-05,
+      "loss": 0.6485,
+      "step": 218
+    },
+    {
+      "epoch": 0.3173913043478261,
+      "grad_norm": 0.54854349900959,
+      "learning_rate": 4.967793880837359e-05,
+      "loss": 0.7133,
+      "step": 219
+    },
+    {
+      "epoch": 0.3188405797101449,
+      "grad_norm": 0.5217856945206872,
+      "learning_rate": 4.965110037573806e-05,
+      "loss": 0.78,
+      "step": 220
+    },
+    {
+      "epoch": 0.32028985507246377,
+      "grad_norm": 0.5466245482758063,
+      "learning_rate": 4.962426194310252e-05,
+      "loss": 0.7367,
+      "step": 221
+    },
+    {
+      "epoch": 0.3217391304347826,
+      "grad_norm": 0.7282345105417176,
+      "learning_rate": 4.959742351046699e-05,
+      "loss": 0.6954,
+      "step": 222
+    },
+    {
+      "epoch": 0.32318840579710145,
+      "grad_norm": 0.516419359601005,
+      "learning_rate": 4.9570585077831454e-05,
+      "loss": 0.7519,
+      "step": 223
+    },
+    {
+      "epoch": 0.32463768115942027,
+      "grad_norm": 0.5211427796582248,
+      "learning_rate": 4.954374664519592e-05,
+      "loss": 0.6564,
+      "step": 224
+    },
+    {
+      "epoch": 0.32608695652173914,
+      "grad_norm": 0.42978384225909405,
+      "learning_rate": 4.9516908212560386e-05,
+      "loss": 0.687,
+      "step": 225
+    },
+    {
+      "epoch": 0.32753623188405795,
+      "grad_norm": 0.44939955255847236,
+      "learning_rate": 4.9490069779924855e-05,
+      "loss": 0.6175,
+      "step": 226
+    },
+    {
+      "epoch": 0.3289855072463768,
+      "grad_norm": 0.4805257834794986,
+      "learning_rate": 4.946323134728932e-05,
+      "loss": 0.5891,
+      "step": 227
+    },
+    {
+      "epoch": 0.33043478260869563,
+      "grad_norm": 0.5560744932815476,
+      "learning_rate": 4.943639291465378e-05,
+      "loss": 0.7353,
+      "step": 228
+    },
+    {
+      "epoch": 0.3318840579710145,
+      "grad_norm": 0.3961756192417976,
+      "learning_rate": 4.940955448201825e-05,
+      "loss": 0.6504,
+      "step": 229
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.5300190615241616,
+      "learning_rate": 4.938271604938271e-05,
+      "loss": 0.6147,
+      "step": 230
+    },
+    {
+      "epoch": 0.3347826086956522,
+      "grad_norm": 0.4281542999977893,
+      "learning_rate": 4.935587761674719e-05,
+      "loss": 0.7576,
+      "step": 231
+    },
+    {
+      "epoch": 0.336231884057971,
+      "grad_norm": 0.6674575417873713,
+      "learning_rate": 4.932903918411165e-05,
+      "loss": 0.7837,
+      "step": 232
+    },
+    {
+      "epoch": 0.33768115942028987,
+      "grad_norm": 0.5981661539007347,
+      "learning_rate": 4.930220075147612e-05,
+      "loss": 0.6746,
+      "step": 233
+    },
+    {
+      "epoch": 0.3391304347826087,
+      "grad_norm": 0.41904669721076937,
+      "learning_rate": 4.9275362318840584e-05,
+      "loss": 0.6213,
+      "step": 234
+    },
+    {
+      "epoch": 0.34057971014492755,
+      "grad_norm": 0.5140483469439551,
+      "learning_rate": 4.924852388620505e-05,
+      "loss": 0.7331,
+      "step": 235
+    },
+    {
+      "epoch": 0.34202898550724636,
+      "grad_norm": 0.4612364066752335,
+      "learning_rate": 4.922168545356952e-05,
+      "loss": 0.6362,
+      "step": 236
+    },
+    {
+      "epoch": 0.34347826086956523,
+      "grad_norm": 0.4498508963425885,
+      "learning_rate": 4.919484702093398e-05,
+      "loss": 0.6164,
+      "step": 237
+    },
+    {
+      "epoch": 0.34492753623188405,
+      "grad_norm": 0.5586129985408278,
+      "learning_rate": 4.916800858829845e-05,
+      "loss": 0.7438,
+      "step": 238
+    },
+    {
+      "epoch": 0.3463768115942029,
+      "grad_norm": 0.5030472932879021,
+      "learning_rate": 4.914117015566291e-05,
+      "loss": 0.7163,
+      "step": 239
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.4559732003400084,
+      "learning_rate": 4.911433172302738e-05,
+      "loss": 0.5819,
+      "step": 240
+    },
+    {
+      "epoch": 0.3492753623188406,
+      "grad_norm": 0.4878474979693718,
+      "learning_rate": 4.9087493290391844e-05,
+      "loss": 0.6758,
+      "step": 241
+    },
+    {
+      "epoch": 0.3507246376811594,
+      "grad_norm": 0.6615032146067258,
+      "learning_rate": 4.9060654857756313e-05,
+      "loss": 0.7171,
+      "step": 242
+    },
+    {
+      "epoch": 0.3521739130434783,
+      "grad_norm": 0.5930208248990055,
+      "learning_rate": 4.9033816425120776e-05,
+      "loss": 0.7232,
+      "step": 243
+    },
+    {
+      "epoch": 0.3536231884057971,
+      "grad_norm": 0.48395298139388776,
+      "learning_rate": 4.9006977992485246e-05,
+      "loss": 0.769,
+      "step": 244
+    },
+    {
+      "epoch": 0.35507246376811596,
+      "grad_norm": 0.5502774529332332,
+      "learning_rate": 4.898013955984971e-05,
+      "loss": 0.5951,
+      "step": 245
+    },
+    {
+      "epoch": 0.3565217391304348,
+      "grad_norm": 0.6227102239291648,
+      "learning_rate": 4.895330112721417e-05,
+      "loss": 0.6301,
+      "step": 246
+    },
+    {
+      "epoch": 0.35797101449275365,
+      "grad_norm": 0.4496872881250766,
+      "learning_rate": 4.892646269457864e-05,
+      "loss": 0.7381,
+      "step": 247
+    },
+    {
+      "epoch": 0.35942028985507246,
+      "grad_norm": 0.3902252305394972,
+      "learning_rate": 4.8899624261943103e-05,
+      "loss": 0.6779,
+      "step": 248
+    },
+    {
+      "epoch": 0.36086956521739133,
+      "grad_norm": 0.4953935450077986,
+      "learning_rate": 4.887278582930757e-05,
+      "loss": 0.6582,
+      "step": 249
+    },
+    {
+      "epoch": 0.36231884057971014,
+      "grad_norm": 0.44386606193131384,
+      "learning_rate": 4.8845947396672036e-05,
+      "loss": 0.6119,
+      "step": 250
+    },
+    {
+      "epoch": 0.36376811594202896,
+      "grad_norm": 1.957189039544279,
+      "learning_rate": 4.8819108964036505e-05,
+      "loss": 0.6398,
+      "step": 251
+    },
+    {
+      "epoch": 0.3652173913043478,
+      "grad_norm": 0.47286371385458287,
+      "learning_rate": 4.879227053140097e-05,
+      "loss": 0.6352,
+      "step": 252
+    },
+    {
+      "epoch": 0.36666666666666664,
+      "grad_norm": 0.49049957412021783,
+      "learning_rate": 4.876543209876544e-05,
+      "loss": 0.6178,
+      "step": 253
+    },
+    {
+      "epoch": 0.3681159420289855,
+      "grad_norm": 0.6559768674111185,
+      "learning_rate": 4.87385936661299e-05,
+      "loss": 0.6945,
+      "step": 254
+    },
+    {
+      "epoch": 0.3695652173913043,
+      "grad_norm": 0.6203105736124627,
+      "learning_rate": 4.871175523349436e-05,
+      "loss": 0.6242,
+      "step": 255
+    },
+    {
+      "epoch": 0.3710144927536232,
+      "grad_norm": 0.5057308068799966,
+      "learning_rate": 4.868491680085883e-05,
+      "loss": 0.608,
+      "step": 256
+    },
+    {
+      "epoch": 0.372463768115942,
+      "grad_norm": 0.4602451484685541,
+      "learning_rate": 4.8658078368223295e-05,
+      "loss": 0.6752,
+      "step": 257
+    },
+    {
+      "epoch": 0.3739130434782609,
+      "grad_norm": 0.5132163941533325,
+      "learning_rate": 4.8631239935587765e-05,
+      "loss": 0.6914,
+      "step": 258
+    },
+    {
+      "epoch": 0.3753623188405797,
+      "grad_norm": 0.5450742277436593,
+      "learning_rate": 4.860440150295223e-05,
+      "loss": 0.65,
+      "step": 259
+    },
+    {
+      "epoch": 0.37681159420289856,
+      "grad_norm": 0.47203214991964443,
+      "learning_rate": 4.85775630703167e-05,
+      "loss": 0.6754,
+      "step": 260
+    },
+    {
+      "epoch": 0.3782608695652174,
+      "grad_norm": 0.5761359721857935,
+      "learning_rate": 4.855072463768116e-05,
+      "loss": 0.7498,
+      "step": 261
+    },
+    {
+      "epoch": 0.37971014492753624,
+      "grad_norm": 0.417069826454048,
+      "learning_rate": 4.852388620504563e-05,
+      "loss": 0.7171,
+      "step": 262
+    },
+    {
+      "epoch": 0.38115942028985506,
+      "grad_norm": 0.35694630092374025,
+      "learning_rate": 4.849704777241009e-05,
+      "loss": 0.6517,
+      "step": 263
+    },
+    {
+      "epoch": 0.3826086956521739,
+      "grad_norm": 0.4651244997425789,
+      "learning_rate": 4.847020933977456e-05,
+      "loss": 0.6868,
+      "step": 264
+    },
+    {
+      "epoch": 0.38405797101449274,
+      "grad_norm": 0.4028313222745874,
+      "learning_rate": 4.8443370907139024e-05,
+      "loss": 0.7097,
+      "step": 265
+    },
+    {
+      "epoch": 0.3855072463768116,
+      "grad_norm": 0.4437827058159613,
+      "learning_rate": 4.841653247450349e-05,
+      "loss": 0.6755,
+      "step": 266
+    },
+    {
+      "epoch": 0.3869565217391304,
+      "grad_norm": 0.4244846215615247,
+      "learning_rate": 4.8389694041867956e-05,
+      "loss": 0.6864,
+      "step": 267
+    },
+    {
+      "epoch": 0.3884057971014493,
+      "grad_norm": 0.44310012910623525,
+      "learning_rate": 4.836285560923242e-05,
+      "loss": 0.6454,
+      "step": 268
+    },
+    {
+      "epoch": 0.3898550724637681,
+      "grad_norm": 0.4757642543300501,
+      "learning_rate": 4.833601717659689e-05,
+      "loss": 0.6639,
+      "step": 269
+    },
+    {
+      "epoch": 0.391304347826087,
+      "grad_norm": 0.565885386171182,
+      "learning_rate": 4.830917874396135e-05,
+      "loss": 0.799,
+      "step": 270
+    },
+    {
+      "epoch": 0.3927536231884058,
+      "grad_norm": 0.4100126299263926,
+      "learning_rate": 4.828234031132582e-05,
+      "loss": 0.6586,
+      "step": 271
+    },
+    {
+      "epoch": 0.39420289855072466,
+      "grad_norm": 0.4565019540369247,
+      "learning_rate": 4.8255501878690284e-05,
+      "loss": 0.6809,
+      "step": 272
+    },
+    {
+      "epoch": 0.39565217391304347,
+      "grad_norm": 0.4554139091690649,
+      "learning_rate": 4.822866344605475e-05,
+      "loss": 0.7114,
+      "step": 273
+    },
+    {
+      "epoch": 0.39710144927536234,
+      "grad_norm": 0.4724351999778902,
+      "learning_rate": 4.8201825013419216e-05,
+      "loss": 0.611,
+      "step": 274
+    },
+    {
+      "epoch": 0.39855072463768115,
+      "grad_norm": 0.5207385550926253,
+      "learning_rate": 4.817498658078368e-05,
+      "loss": 0.721,
+      "step": 275
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43019224883693835,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 0.7154,
+      "step": 276
+    },
+    {
+      "epoch": 0.40144927536231884,
+      "grad_norm": 0.522469033072573,
+      "learning_rate": 4.812130971551262e-05,
+      "loss": 0.7264,
+      "step": 277
+    },
+    {
+      "epoch": 0.4028985507246377,
+      "grad_norm": 0.4695809052920177,
+      "learning_rate": 4.809447128287709e-05,
+      "loss": 0.5917,
+      "step": 278
+    },
+    {
+      "epoch": 0.4043478260869565,
+      "grad_norm": 0.5312579249251256,
+      "learning_rate": 4.806763285024155e-05,
+      "loss": 0.6322,
+      "step": 279
+    },
+    {
+      "epoch": 0.4057971014492754,
+      "grad_norm": 0.6103259840908285,
+      "learning_rate": 4.804079441760602e-05,
+      "loss": 0.7255,
+      "step": 280
+    },
+    {
+      "epoch": 0.4072463768115942,
+      "grad_norm": 0.5509964882780453,
+      "learning_rate": 4.801395598497048e-05,
+      "loss": 0.6679,
+      "step": 281
+    },
+    {
+      "epoch": 0.40869565217391307,
+      "grad_norm": 0.6573044315013673,
+      "learning_rate": 4.7987117552334945e-05,
+      "loss": 0.6357,
+      "step": 282
+    },
+    {
+      "epoch": 0.4101449275362319,
+      "grad_norm": 0.5910903249453899,
+      "learning_rate": 4.7960279119699414e-05,
+      "loss": 0.5394,
+      "step": 283
+    },
+    {
+      "epoch": 0.4115942028985507,
+      "grad_norm": 0.47832362737303147,
+      "learning_rate": 4.793344068706388e-05,
+      "loss": 0.6773,
+      "step": 284
+    },
+    {
+      "epoch": 0.41304347826086957,
+      "grad_norm": 0.5548655349336106,
+      "learning_rate": 4.790660225442835e-05,
+      "loss": 0.6962,
+      "step": 285
+    },
+    {
+      "epoch": 0.4144927536231884,
+      "grad_norm": 0.4534853918571052,
+      "learning_rate": 4.787976382179281e-05,
+      "loss": 0.7101,
+      "step": 286
+    },
+    {
+      "epoch": 0.41594202898550725,
+      "grad_norm": 0.42141712239602674,
+      "learning_rate": 4.785292538915728e-05,
+      "loss": 0.6164,
+      "step": 287
+    },
+    {
+      "epoch": 0.41739130434782606,
+      "grad_norm": 0.5303006892740134,
+      "learning_rate": 4.782608695652174e-05,
+      "loss": 0.6017,
+      "step": 288
+    },
+    {
+      "epoch": 0.41884057971014493,
+      "grad_norm": 0.46211348435989263,
+      "learning_rate": 4.779924852388621e-05,
+      "loss": 0.6643,
+      "step": 289
+    },
+    {
+      "epoch": 0.42028985507246375,
+      "grad_norm": 0.5422397830559342,
+      "learning_rate": 4.7772410091250674e-05,
+      "loss": 0.6748,
+      "step": 290
+    },
+    {
+      "epoch": 0.4217391304347826,
+      "grad_norm": 0.45566020860996564,
+      "learning_rate": 4.7745571658615143e-05,
+      "loss": 0.5797,
+      "step": 291
+    },
+    {
+      "epoch": 0.42318840579710143,
+      "grad_norm": 0.7046915013882183,
+      "learning_rate": 4.7718733225979606e-05,
+      "loss": 0.6902,
+      "step": 292
+    },
+    {
+      "epoch": 0.4246376811594203,
+      "grad_norm": 0.5594416692480499,
+      "learning_rate": 4.769189479334407e-05,
+      "loss": 0.7175,
+      "step": 293
+    },
+    {
+      "epoch": 0.4260869565217391,
+      "grad_norm": 0.4916116646480955,
+      "learning_rate": 4.766505636070854e-05,
+      "loss": 0.5814,
+      "step": 294
+    },
+    {
+      "epoch": 0.427536231884058,
+      "grad_norm": 0.45352777639895914,
+      "learning_rate": 4.7638217928073e-05,
+      "loss": 0.6896,
+      "step": 295
+    },
+    {
+      "epoch": 0.4289855072463768,
+      "grad_norm": 0.5149754279803409,
+      "learning_rate": 4.761137949543747e-05,
+      "loss": 0.6643,
+      "step": 296
+    },
+    {
+      "epoch": 0.43043478260869567,
+      "grad_norm": 0.43351576199759795,
+      "learning_rate": 4.7584541062801933e-05,
+      "loss": 0.6787,
+      "step": 297
+    },
+    {
+      "epoch": 0.4318840579710145,
+      "grad_norm": 0.41889843156070783,
+      "learning_rate": 4.75577026301664e-05,
+      "loss": 0.5754,
+      "step": 298
+    },
+    {
+      "epoch": 0.43333333333333335,
+      "grad_norm": 0.5025185648234128,
+      "learning_rate": 4.7530864197530866e-05,
+      "loss": 0.7761,
+      "step": 299
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 0.4910170844080093,
+      "learning_rate": 4.7504025764895335e-05,
+      "loss": 0.6271,
+      "step": 300
+    },
+    {
+      "epoch": 0.43623188405797103,
+      "grad_norm": 0.4459249234150752,
+      "learning_rate": 4.74771873322598e-05,
+      "loss": 0.5621,
+      "step": 301
+    },
+    {
+      "epoch": 0.43768115942028984,
+      "grad_norm": 0.4130240042045532,
+      "learning_rate": 4.745034889962426e-05,
+      "loss": 0.6075,
+      "step": 302
+    },
+    {
+      "epoch": 0.4391304347826087,
+      "grad_norm": 0.5339425697493924,
+      "learning_rate": 4.742351046698873e-05,
+      "loss": 0.5985,
+      "step": 303
+    },
+    {
+      "epoch": 0.4405797101449275,
+      "grad_norm": 0.5662260384248383,
+      "learning_rate": 4.739667203435319e-05,
+      "loss": 0.7069,
+      "step": 304
+    },
+    {
+      "epoch": 0.4420289855072464,
+      "grad_norm": 0.5289530183900609,
+      "learning_rate": 4.736983360171766e-05,
+      "loss": 0.6578,
+      "step": 305
+    },
+    {
+      "epoch": 0.4434782608695652,
+      "grad_norm": 0.5621056697561709,
+      "learning_rate": 4.7342995169082125e-05,
+      "loss": 0.5806,
+      "step": 306
+    },
+    {
+      "epoch": 0.4449275362318841,
+      "grad_norm": 0.5028453643236157,
+      "learning_rate": 4.7316156736446595e-05,
+      "loss": 0.5989,
+      "step": 307
+    },
+    {
+      "epoch": 0.4463768115942029,
+      "grad_norm": 0.48129891568938116,
+      "learning_rate": 4.728931830381106e-05,
+      "loss": 0.6438,
+      "step": 308
+    },
+    {
+      "epoch": 0.44782608695652176,
+      "grad_norm": 0.5199119699799553,
+      "learning_rate": 4.726247987117553e-05,
+      "loss": 0.6594,
+      "step": 309
+    },
+    {
+      "epoch": 0.4492753623188406,
+      "grad_norm": 0.463370177698594,
+      "learning_rate": 4.723564143853999e-05,
+      "loss": 0.7021,
+      "step": 310
+    },
+    {
+      "epoch": 0.45072463768115945,
+      "grad_norm": 0.41751731613436227,
+      "learning_rate": 4.720880300590446e-05,
+      "loss": 0.6408,
+      "step": 311
+    },
+    {
+      "epoch": 0.45217391304347826,
+      "grad_norm": 0.41222115997598185,
+      "learning_rate": 4.718196457326892e-05,
+      "loss": 0.6486,
+      "step": 312
+    },
+    {
+      "epoch": 0.45362318840579713,
+      "grad_norm": 0.41762118003102466,
+      "learning_rate": 4.7155126140633385e-05,
+      "loss": 0.7012,
+      "step": 313
+    },
+    {
+      "epoch": 0.45507246376811594,
+      "grad_norm": 0.47083847264116285,
+      "learning_rate": 4.7128287707997854e-05,
+      "loss": 0.5933,
+      "step": 314
+    },
+    {
+      "epoch": 0.45652173913043476,
+      "grad_norm": 0.4385295480312083,
+      "learning_rate": 4.710144927536232e-05,
+      "loss": 0.5925,
+      "step": 315
+    },
+    {
+      "epoch": 0.4579710144927536,
+      "grad_norm": 0.47338682594708603,
+      "learning_rate": 4.7074610842726786e-05,
+      "loss": 0.6403,
+      "step": 316
+    },
+    {
+      "epoch": 0.45942028985507244,
+      "grad_norm": 0.449155602405838,
+      "learning_rate": 4.704777241009125e-05,
+      "loss": 0.6025,
+      "step": 317
+    },
+    {
+      "epoch": 0.4608695652173913,
+      "grad_norm": 0.4516855897509766,
+      "learning_rate": 4.702093397745572e-05,
+      "loss": 0.5849,
+      "step": 318
+    },
+    {
+      "epoch": 0.4623188405797101,
+      "grad_norm": 0.5606454482640691,
+      "learning_rate": 4.699409554482018e-05,
+      "loss": 0.6841,
+      "step": 319
+    },
+    {
+      "epoch": 0.463768115942029,
+      "grad_norm": 0.5070294712602395,
+      "learning_rate": 4.696725711218465e-05,
+      "loss": 0.6293,
+      "step": 320
+    },
+    {
+      "epoch": 0.4652173913043478,
+      "grad_norm": 0.37187122216983104,
+      "learning_rate": 4.6940418679549114e-05,
+      "loss": 0.595,
+      "step": 321
+    },
+    {
+      "epoch": 0.4666666666666667,
+      "grad_norm": 0.683077057258226,
+      "learning_rate": 4.691358024691358e-05,
+      "loss": 0.5657,
+      "step": 322
+    },
+    {
+      "epoch": 0.4681159420289855,
+      "grad_norm": 0.4701558290739009,
+      "learning_rate": 4.6886741814278046e-05,
+      "loss": 0.6412,
+      "step": 323
+    },
+    {
+      "epoch": 0.46956521739130436,
+      "grad_norm": 0.4085454812990897,
+      "learning_rate": 4.6859903381642516e-05,
+      "loss": 0.6615,
+      "step": 324
+    },
+    {
+      "epoch": 0.47101449275362317,
+      "grad_norm": 0.5311673769504186,
+      "learning_rate": 4.6833064949006985e-05,
+      "loss": 0.6267,
+      "step": 325
+    },
+    {
+      "epoch": 0.47246376811594204,
+      "grad_norm": 0.423024811889357,
+      "learning_rate": 4.680622651637145e-05,
+      "loss": 0.6453,
+      "step": 326
+    },
+    {
+      "epoch": 0.47391304347826085,
+      "grad_norm": 0.4560564760332426,
+      "learning_rate": 4.677938808373592e-05,
+      "loss": 0.6633,
+      "step": 327
+    },
+    {
+      "epoch": 0.4753623188405797,
+      "grad_norm": 0.36439217196706425,
+      "learning_rate": 4.675254965110038e-05,
+      "loss": 0.6128,
+      "step": 328
+    },
+    {
+      "epoch": 0.47681159420289854,
+      "grad_norm": 0.3919705007269534,
+      "learning_rate": 4.672571121846484e-05,
+      "loss": 0.6057,
+      "step": 329
+    },
+    {
+      "epoch": 0.4782608695652174,
+      "grad_norm": 0.3876474762734424,
+      "learning_rate": 4.669887278582931e-05,
+      "loss": 0.5812,
+      "step": 330
+    },
+    {
+      "epoch": 0.4797101449275362,
+      "grad_norm": 0.8537368376487238,
+      "learning_rate": 4.6672034353193775e-05,
+      "loss": 0.6831,
+      "step": 331
+    },
+    {
+      "epoch": 0.4811594202898551,
+      "grad_norm": 0.4126095686701784,
+      "learning_rate": 4.6645195920558245e-05,
+      "loss": 0.683,
+      "step": 332
+    },
+    {
+      "epoch": 0.4826086956521739,
+      "grad_norm": 0.4837042105681288,
+      "learning_rate": 4.661835748792271e-05,
+      "loss": 0.551,
+      "step": 333
+    },
+    {
+      "epoch": 0.48405797101449277,
+      "grad_norm": 0.43451962804846805,
+      "learning_rate": 4.659151905528718e-05,
+      "loss": 0.5953,
+      "step": 334
+    },
+    {
+      "epoch": 0.4855072463768116,
+      "grad_norm": 0.4254152211545657,
+      "learning_rate": 4.656468062265164e-05,
+      "loss": 0.6145,
+      "step": 335
+    },
+    {
+      "epoch": 0.48695652173913045,
+      "grad_norm": 0.5081602183421405,
+      "learning_rate": 4.653784219001611e-05,
+      "loss": 0.6101,
+      "step": 336
+    },
+    {
+      "epoch": 0.48840579710144927,
+      "grad_norm": 0.3677242041378604,
+      "learning_rate": 4.651100375738057e-05,
+      "loss": 0.628,
+      "step": 337
+    },
+    {
+      "epoch": 0.48985507246376814,
+      "grad_norm": 0.5736410512077793,
+      "learning_rate": 4.648416532474504e-05,
+      "loss": 0.7567,
+      "step": 338
+    },
+    {
+      "epoch": 0.49130434782608695,
+      "grad_norm": 0.3961758433255439,
+      "learning_rate": 4.6457326892109504e-05,
+      "loss": 0.5232,
+      "step": 339
+    },
+    {
+      "epoch": 0.4927536231884058,
+      "grad_norm": 0.6078832397392898,
+      "learning_rate": 4.643048845947397e-05,
+      "loss": 0.7044,
+      "step": 340
+    },
+    {
+      "epoch": 0.49420289855072463,
+      "grad_norm": 0.39627760424098896,
+      "learning_rate": 4.6403650026838436e-05,
+      "loss": 0.6417,
+      "step": 341
+    },
+    {
+      "epoch": 0.4956521739130435,
+      "grad_norm": 0.4734334069120371,
+      "learning_rate": 4.63768115942029e-05,
+      "loss": 0.6287,
+      "step": 342
+    },
+    {
+      "epoch": 0.4971014492753623,
+      "grad_norm": 0.46243156968178983,
+      "learning_rate": 4.634997316156737e-05,
+      "loss": 0.6103,
+      "step": 343
+    },
+    {
+      "epoch": 0.4985507246376812,
+      "grad_norm": 0.5433190757897186,
+      "learning_rate": 4.632313472893183e-05,
+      "loss": 0.5938,
+      "step": 344
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4559700829884716,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 0.679,
+      "step": 345
+    },
+    {
+      "epoch": 0.5014492753623189,
+      "grad_norm": 0.49482650146755597,
+      "learning_rate": 4.6269457863660764e-05,
+      "loss": 0.5917,
+      "step": 346
+    },
+    {
+      "epoch": 0.5028985507246376,
+      "grad_norm": 0.5029576885090551,
+      "learning_rate": 4.624261943102523e-05,
+      "loss": 0.5109,
+      "step": 347
+    },
+    {
+      "epoch": 0.5043478260869565,
+      "grad_norm": 0.3794583756550058,
+      "learning_rate": 4.6215780998389696e-05,
+      "loss": 0.6882,
+      "step": 348
+    },
+    {
+      "epoch": 0.5057971014492754,
+      "grad_norm": 0.49466579544121436,
+      "learning_rate": 4.6188942565754165e-05,
+      "loss": 0.5764,
+      "step": 349
+    },
+    {
+      "epoch": 0.5072463768115942,
+      "grad_norm": 0.4141689512026932,
+      "learning_rate": 4.616210413311863e-05,
+      "loss": 0.6417,
+      "step": 350
+    },
+    {
+      "epoch": 0.508695652173913,
+      "grad_norm": 0.4572258866434217,
+      "learning_rate": 4.613526570048309e-05,
+      "loss": 0.6427,
+      "step": 351
+    },
+    {
+      "epoch": 0.5101449275362319,
+      "grad_norm": 0.5119906485408763,
+      "learning_rate": 4.610842726784756e-05,
+      "loss": 0.6396,
+      "step": 352
+    },
+    {
+      "epoch": 0.5115942028985507,
+      "grad_norm": 0.43916535973559384,
+      "learning_rate": 4.608158883521202e-05,
+      "loss": 0.5667,
+      "step": 353
+    },
+    {
+      "epoch": 0.5130434782608696,
+      "grad_norm": 0.5789964597820626,
+      "learning_rate": 4.605475040257649e-05,
+      "loss": 0.5616,
+      "step": 354
+    },
+    {
+      "epoch": 0.5144927536231884,
+      "grad_norm": 0.3933092714425148,
+      "learning_rate": 4.6027911969940955e-05,
+      "loss": 0.603,
+      "step": 355
+    },
+    {
+      "epoch": 0.5159420289855072,
+      "grad_norm": 0.5437635404178742,
+      "learning_rate": 4.6001073537305425e-05,
+      "loss": 0.6244,
+      "step": 356
+    },
+    {
+      "epoch": 0.5173913043478261,
+      "grad_norm": 0.4887636002020938,
+      "learning_rate": 4.597423510466989e-05,
+      "loss": 0.6635,
+      "step": 357
+    },
+    {
+      "epoch": 0.518840579710145,
+      "grad_norm": 0.4363998278608295,
+      "learning_rate": 4.594739667203436e-05,
+      "loss": 0.5896,
+      "step": 358
+    },
+    {
+      "epoch": 0.5202898550724637,
+      "grad_norm": 0.4598576444127417,
+      "learning_rate": 4.592055823939882e-05,
+      "loss": 0.6361,
+      "step": 359
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 0.34761717274590787,
+      "learning_rate": 4.589371980676328e-05,
+      "loss": 0.5961,
+      "step": 360
+    },
+    {
+      "epoch": 0.5231884057971015,
+      "grad_norm": 0.4251983395550634,
+      "learning_rate": 4.586688137412775e-05,
+      "loss": 0.6047,
+      "step": 361
+    },
+    {
+      "epoch": 0.5246376811594203,
+      "grad_norm": 0.38041849327413135,
+      "learning_rate": 4.5840042941492215e-05,
+      "loss": 0.581,
+      "step": 362
+    },
+    {
+      "epoch": 0.5260869565217391,
+      "grad_norm": 0.3769564394943478,
+      "learning_rate": 4.5813204508856684e-05,
+      "loss": 0.5916,
+      "step": 363
+    },
+    {
+      "epoch": 0.527536231884058,
+      "grad_norm": 0.4060225292587437,
+      "learning_rate": 4.578636607622115e-05,
+      "loss": 0.6497,
+      "step": 364
+    },
+    {
+      "epoch": 0.5289855072463768,
+      "grad_norm": 0.3752767903004976,
+      "learning_rate": 4.5759527643585617e-05,
+      "loss": 0.6158,
+      "step": 365
+    },
+    {
+      "epoch": 0.5304347826086957,
+      "grad_norm": 0.39868767186559706,
+      "learning_rate": 4.573268921095008e-05,
+      "loss": 0.637,
+      "step": 366
+    },
+    {
+      "epoch": 0.5318840579710145,
+      "grad_norm": 0.4305347328957555,
+      "learning_rate": 4.570585077831455e-05,
+      "loss": 0.6199,
+      "step": 367
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.39277659041458757,
+      "learning_rate": 4.567901234567901e-05,
+      "loss": 0.6041,
+      "step": 368
+    },
+    {
+      "epoch": 0.5347826086956522,
+      "grad_norm": 0.4749667444026662,
+      "learning_rate": 4.565217391304348e-05,
+      "loss": 0.6974,
+      "step": 369
+    },
+    {
+      "epoch": 0.5362318840579711,
+      "grad_norm": 0.4391867067280921,
+      "learning_rate": 4.5625335480407944e-05,
+      "loss": 0.5768,
+      "step": 370
+    },
+    {
+      "epoch": 0.5376811594202898,
+      "grad_norm": 0.5090096020968206,
+      "learning_rate": 4.559849704777241e-05,
+      "loss": 0.5445,
+      "step": 371
+    },
+    {
+      "epoch": 0.5391304347826087,
+      "grad_norm": 0.4011524373138279,
+      "learning_rate": 4.557165861513688e-05,
+      "loss": 0.5444,
+      "step": 372
+    },
+    {
+      "epoch": 0.5405797101449276,
+      "grad_norm": 0.4558341949312885,
+      "learning_rate": 4.5544820182501346e-05,
+      "loss": 0.6478,
+      "step": 373
+    },
+    {
+      "epoch": 0.5420289855072464,
+      "grad_norm": 0.4568966746754348,
+      "learning_rate": 4.5517981749865815e-05,
+      "loss": 0.614,
+      "step": 374
+    },
+    {
+      "epoch": 0.5434782608695652,
+      "grad_norm": 0.4700753115968392,
+      "learning_rate": 4.549114331723028e-05,
+      "loss": 0.759,
+      "step": 375
+    },
+    {
+      "epoch": 0.5449275362318841,
+      "grad_norm": 0.3998862234622315,
+      "learning_rate": 4.546430488459474e-05,
+      "loss": 0.6338,
+      "step": 376
+    },
+    {
+      "epoch": 0.5463768115942029,
+      "grad_norm": 0.43440146434893584,
+      "learning_rate": 4.543746645195921e-05,
+      "loss": 0.5947,
+      "step": 377
+    },
+    {
+      "epoch": 0.5478260869565217,
+      "grad_norm": 0.3648968901901745,
+      "learning_rate": 4.541062801932367e-05,
+      "loss": 0.5506,
+      "step": 378
+    },
+    {
+      "epoch": 0.5492753623188406,
+      "grad_norm": 0.48579400196764083,
+      "learning_rate": 4.538378958668814e-05,
+      "loss": 0.6168,
+      "step": 379
+    },
+    {
+      "epoch": 0.5507246376811594,
+      "grad_norm": 0.7342657082099426,
+      "learning_rate": 4.5356951154052605e-05,
+      "loss": 0.6539,
+      "step": 380
+    },
+    {
+      "epoch": 0.5521739130434783,
+      "grad_norm": 0.45241959724408615,
+      "learning_rate": 4.5330112721417075e-05,
+      "loss": 0.5567,
+      "step": 381
+    },
+    {
+      "epoch": 0.553623188405797,
+      "grad_norm": 0.3986083042401516,
+      "learning_rate": 4.530327428878154e-05,
+      "loss": 0.6844,
+      "step": 382
+    },
+    {
+      "epoch": 0.5550724637681159,
+      "grad_norm": 0.6363551916780863,
+      "learning_rate": 4.527643585614601e-05,
+      "loss": 0.6956,
+      "step": 383
+    },
+    {
+      "epoch": 0.5565217391304348,
+      "grad_norm": 0.48150733103051285,
+      "learning_rate": 4.524959742351047e-05,
+      "loss": 0.6324,
+      "step": 384
+    },
+    {
+      "epoch": 0.5579710144927537,
+      "grad_norm": 0.49675319166001186,
+      "learning_rate": 4.522275899087494e-05,
+      "loss": 0.6708,
+      "step": 385
+    },
+    {
+      "epoch": 0.5594202898550724,
+      "grad_norm": 0.47708659349752586,
+      "learning_rate": 4.51959205582394e-05,
+      "loss": 0.685,
+      "step": 386
+    },
+    {
+      "epoch": 0.5608695652173913,
+      "grad_norm": 0.5000220688999433,
+      "learning_rate": 4.5169082125603865e-05,
+      "loss": 0.6275,
+      "step": 387
+    },
+    {
+      "epoch": 0.5623188405797102,
+      "grad_norm": 0.4368459518298929,
+      "learning_rate": 4.5142243692968334e-05,
+      "loss": 0.6844,
+      "step": 388
+    },
+    {
+      "epoch": 0.563768115942029,
+      "grad_norm": 0.4096537771062687,
+      "learning_rate": 4.51154052603328e-05,
+      "loss": 0.6328,
+      "step": 389
+    },
+    {
+      "epoch": 0.5652173913043478,
+      "grad_norm": 0.3334448687900693,
+      "learning_rate": 4.5088566827697266e-05,
+      "loss": 0.5817,
+      "step": 390
+    },
+    {
+      "epoch": 0.5666666666666667,
+      "grad_norm": 0.4819861650908369,
+      "learning_rate": 4.506172839506173e-05,
+      "loss": 0.6051,
+      "step": 391
+    },
+    {
+      "epoch": 0.5681159420289855,
+      "grad_norm": 0.43386350661453743,
+      "learning_rate": 4.50348899624262e-05,
+      "loss": 0.6212,
+      "step": 392
+    },
+    {
+      "epoch": 0.5695652173913044,
+      "grad_norm": 0.4542325204418208,
+      "learning_rate": 4.500805152979066e-05,
+      "loss": 0.598,
+      "step": 393
+    },
+    {
+      "epoch": 0.5710144927536231,
+      "grad_norm": 0.36353694572274375,
+      "learning_rate": 4.498121309715513e-05,
+      "loss": 0.6753,
+      "step": 394
+    },
+    {
+      "epoch": 0.572463768115942,
+      "grad_norm": 0.5541625024013656,
+      "learning_rate": 4.4954374664519594e-05,
+      "loss": 0.6362,
+      "step": 395
+    },
+    {
+      "epoch": 0.5739130434782609,
+      "grad_norm": 0.3919828315297418,
+      "learning_rate": 4.492753623188406e-05,
+      "loss": 0.625,
+      "step": 396
+    },
+    {
+      "epoch": 0.5753623188405798,
+      "grad_norm": 1.499546878942088,
+      "learning_rate": 4.4900697799248526e-05,
+      "loss": 0.5579,
+      "step": 397
+    },
+    {
+      "epoch": 0.5768115942028985,
+      "grad_norm": 0.556911484391141,
+      "learning_rate": 4.487385936661299e-05,
+      "loss": 0.6347,
+      "step": 398
+    },
+    {
+      "epoch": 0.5782608695652174,
+      "grad_norm": 0.45584289957280666,
+      "learning_rate": 4.484702093397746e-05,
+      "loss": 0.6767,
+      "step": 399
+    },
+    {
+      "epoch": 0.5797101449275363,
+      "grad_norm": 0.4956675178012224,
+      "learning_rate": 4.482018250134192e-05,
+      "loss": 0.5632,
+      "step": 400
+    },
+    {
+      "epoch": 0.5811594202898551,
+      "grad_norm": 0.44540912712042036,
+      "learning_rate": 4.479334406870639e-05,
+      "loss": 0.599,
+      "step": 401
+    },
+    {
+      "epoch": 0.5826086956521739,
+      "grad_norm": 0.42044316632261336,
+      "learning_rate": 4.476650563607085e-05,
+      "loss": 0.5734,
+      "step": 402
+    },
+    {
+      "epoch": 0.5840579710144927,
+      "grad_norm": 0.5107808787097907,
+      "learning_rate": 4.473966720343532e-05,
+      "loss": 0.7368,
+      "step": 403
+    },
+    {
+      "epoch": 0.5855072463768116,
+      "grad_norm": 0.4990397560807668,
+      "learning_rate": 4.4712828770799785e-05,
+      "loss": 0.6155,
+      "step": 404
+    },
+    {
+      "epoch": 0.5869565217391305,
+      "grad_norm": 0.4370235307040917,
+      "learning_rate": 4.4685990338164255e-05,
+      "loss": 0.6169,
+      "step": 405
+    },
+    {
+      "epoch": 0.5884057971014492,
+      "grad_norm": 0.37505407563553644,
+      "learning_rate": 4.465915190552872e-05,
+      "loss": 0.583,
+      "step": 406
+    },
+    {
+      "epoch": 0.5898550724637681,
+      "grad_norm": 0.4330379855434459,
+      "learning_rate": 4.463231347289318e-05,
+      "loss": 0.5817,
+      "step": 407
+    },
+    {
+      "epoch": 0.591304347826087,
+      "grad_norm": 0.4575635042170628,
+      "learning_rate": 4.460547504025765e-05,
+      "loss": 0.6586,
+      "step": 408
+    },
+    {
+      "epoch": 0.5927536231884057,
+      "grad_norm": 0.47572457302172977,
+      "learning_rate": 4.457863660762211e-05,
+      "loss": 0.6085,
+      "step": 409
+    },
+    {
+      "epoch": 0.5942028985507246,
+      "grad_norm": 0.3574119907944081,
+      "learning_rate": 4.455179817498658e-05,
+      "loss": 0.5559,
+      "step": 410
+    },
+    {
+      "epoch": 0.5956521739130435,
+      "grad_norm": 0.4178835165927953,
+      "learning_rate": 4.4524959742351045e-05,
+      "loss": 0.6125,
+      "step": 411
+    },
+    {
+      "epoch": 0.5971014492753624,
+      "grad_norm": 0.37512079629567197,
+      "learning_rate": 4.4498121309715514e-05,
+      "loss": 0.6084,
+      "step": 412
+    },
+    {
+      "epoch": 0.5985507246376811,
+      "grad_norm": 0.48633931038992023,
+      "learning_rate": 4.447128287707998e-05,
+      "loss": 0.5896,
+      "step": 413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4796673589113237,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 0.5768,
+      "step": 414
+    },
+    {
+      "epoch": 0.6014492753623188,
+      "grad_norm": 0.43446070994232633,
+      "learning_rate": 4.441760601180891e-05,
+      "loss": 0.5937,
+      "step": 415
+    },
+    {
+      "epoch": 0.6028985507246377,
+      "grad_norm": 0.40889150420962617,
+      "learning_rate": 4.439076757917338e-05,
+      "loss": 0.7413,
+      "step": 416
+    },
+    {
+      "epoch": 0.6043478260869565,
+      "grad_norm": 0.4408375168183745,
+      "learning_rate": 4.436392914653785e-05,
+      "loss": 0.6439,
+      "step": 417
+    },
+    {
+      "epoch": 0.6057971014492753,
+      "grad_norm": 0.4246703790309954,
+      "learning_rate": 4.433709071390231e-05,
+      "loss": 0.6989,
+      "step": 418
+    },
+    {
+      "epoch": 0.6072463768115942,
+      "grad_norm": 0.38710833762232205,
+      "learning_rate": 4.431025228126678e-05,
+      "loss": 0.5806,
+      "step": 419
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 0.43313968649036916,
+      "learning_rate": 4.428341384863124e-05,
+      "loss": 0.7066,
+      "step": 420
+    },
+    {
+      "epoch": 0.6101449275362318,
+      "grad_norm": 0.35913268364756945,
+      "learning_rate": 4.425657541599571e-05,
+      "loss": 0.6513,
+      "step": 421
+    },
+    {
+      "epoch": 0.6115942028985507,
+      "grad_norm": 0.47580547381182897,
+      "learning_rate": 4.4229736983360176e-05,
+      "loss": 0.6257,
+      "step": 422
+    },
+    {
+      "epoch": 0.6130434782608696,
+      "grad_norm": 0.37911697302582864,
+      "learning_rate": 4.4202898550724645e-05,
+      "loss": 0.5946,
+      "step": 423
+    },
+    {
+      "epoch": 0.6144927536231884,
+      "grad_norm": 0.44074764204295713,
+      "learning_rate": 4.417606011808911e-05,
+      "loss": 0.55,
+      "step": 424
+    },
+    {
+      "epoch": 0.6159420289855072,
+      "grad_norm": 0.3937375526845581,
+      "learning_rate": 4.414922168545357e-05,
+      "loss": 0.5371,
+      "step": 425
+    },
+    {
+      "epoch": 0.6173913043478261,
+      "grad_norm": 0.4573471158606008,
+      "learning_rate": 4.412238325281804e-05,
+      "loss": 0.6158,
+      "step": 426
+    },
+    {
+      "epoch": 0.618840579710145,
+      "grad_norm": 0.3840697276185448,
+      "learning_rate": 4.40955448201825e-05,
+      "loss": 0.6754,
+      "step": 427
+    },
+    {
+      "epoch": 0.6202898550724638,
+      "grad_norm": 0.5317382053849552,
+      "learning_rate": 4.406870638754697e-05,
+      "loss": 0.647,
+      "step": 428
+    },
+    {
+      "epoch": 0.6217391304347826,
+      "grad_norm": 0.34527497173414967,
+      "learning_rate": 4.4041867954911435e-05,
+      "loss": 0.5108,
+      "step": 429
+    },
+    {
+      "epoch": 0.6231884057971014,
+      "grad_norm": 0.44422906839362747,
+      "learning_rate": 4.4015029522275905e-05,
+      "loss": 0.6502,
+      "step": 430
+    },
+    {
+      "epoch": 0.6246376811594203,
+      "grad_norm": 0.473283146231613,
+      "learning_rate": 4.398819108964037e-05,
+      "loss": 0.63,
+      "step": 431
+    },
+    {
+      "epoch": 0.6260869565217392,
+      "grad_norm": 0.40163835413996013,
+      "learning_rate": 4.396135265700484e-05,
+      "loss": 0.6318,
+      "step": 432
+    },
+    {
+      "epoch": 0.6275362318840579,
+      "grad_norm": 0.39716935347056315,
+      "learning_rate": 4.39345142243693e-05,
+      "loss": 0.5803,
+      "step": 433
+    },
+    {
+      "epoch": 0.6289855072463768,
+      "grad_norm": 0.357753745770442,
+      "learning_rate": 4.390767579173376e-05,
+      "loss": 0.595,
+      "step": 434
+    },
+    {
+      "epoch": 0.6304347826086957,
+      "grad_norm": 0.3687517288111165,
+      "learning_rate": 4.388083735909823e-05,
+      "loss": 0.5644,
+      "step": 435
+    },
+    {
+      "epoch": 0.6318840579710145,
+      "grad_norm": 0.38181865324249625,
+      "learning_rate": 4.3853998926462695e-05,
+      "loss": 0.6464,
+      "step": 436
+    },
+    {
+      "epoch": 0.6333333333333333,
+      "grad_norm": 0.44474753275399415,
+      "learning_rate": 4.3827160493827164e-05,
+      "loss": 0.5878,
+      "step": 437
+    },
+    {
+      "epoch": 0.6347826086956522,
+      "grad_norm": 0.3352537728421649,
+      "learning_rate": 4.380032206119163e-05,
+      "loss": 0.5174,
+      "step": 438
+    },
+    {
+      "epoch": 0.636231884057971,
+      "grad_norm": 0.5132005847948536,
+      "learning_rate": 4.3773483628556096e-05,
+      "loss": 0.6436,
+      "step": 439
+    },
+    {
+      "epoch": 0.6376811594202898,
+      "grad_norm": 0.44956523639329127,
+      "learning_rate": 4.374664519592056e-05,
+      "loss": 0.6523,
+      "step": 440
+    },
+    {
+      "epoch": 0.6391304347826087,
+      "grad_norm": 0.3264095375727053,
+      "learning_rate": 4.371980676328503e-05,
+      "loss": 0.5701,
+      "step": 441
+    },
+    {
+      "epoch": 0.6405797101449275,
+      "grad_norm": 0.47943855397136076,
+      "learning_rate": 4.369296833064949e-05,
+      "loss": 0.6899,
+      "step": 442
+    },
+    {
+      "epoch": 0.6420289855072464,
+      "grad_norm": 0.37295675960299046,
+      "learning_rate": 4.366612989801396e-05,
+      "loss": 0.5943,
+      "step": 443
+    },
+    {
+      "epoch": 0.6434782608695652,
+      "grad_norm": 0.37607090500273194,
+      "learning_rate": 4.3639291465378424e-05,
+      "loss": 0.6024,
+      "step": 444
+    },
+    {
+      "epoch": 0.644927536231884,
+      "grad_norm": 0.3709434406889319,
+      "learning_rate": 4.3612453032742886e-05,
+      "loss": 0.5975,
+      "step": 445
+    },
+    {
+      "epoch": 0.6463768115942029,
+      "grad_norm": 0.4001494650314734,
+      "learning_rate": 4.3585614600107356e-05,
+      "loss": 0.6961,
+      "step": 446
+    },
+    {
+      "epoch": 0.6478260869565218,
+      "grad_norm": 0.3723471807287628,
+      "learning_rate": 4.355877616747182e-05,
+      "loss": 0.6208,
+      "step": 447
+    },
+    {
+      "epoch": 0.6492753623188405,
+      "grad_norm": 0.41592983850578813,
+      "learning_rate": 4.353193773483629e-05,
+      "loss": 0.6253,
+      "step": 448
+    },
+    {
+      "epoch": 0.6507246376811594,
+      "grad_norm": 0.39096102588060283,
+      "learning_rate": 4.350509930220075e-05,
+      "loss": 0.6357,
+      "step": 449
+    },
+    {
+      "epoch": 0.6521739130434783,
+      "grad_norm": 0.323828875516656,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 0.5617,
+      "step": 450
+    },
+    {
+      "epoch": 0.6536231884057971,
+      "grad_norm": 0.36464426117660076,
+      "learning_rate": 4.345142243692968e-05,
+      "loss": 0.6906,
+      "step": 451
+    },
+    {
+      "epoch": 0.6550724637681159,
+      "grad_norm": 0.3273186362011285,
+      "learning_rate": 4.342458400429415e-05,
+      "loss": 0.5552,
+      "step": 452
+    },
+    {
+      "epoch": 0.6565217391304348,
+      "grad_norm": 0.38576665695385565,
+      "learning_rate": 4.3397745571658615e-05,
+      "loss": 0.5915,
+      "step": 453
+    },
+    {
+      "epoch": 0.6579710144927536,
+      "grad_norm": 0.4312121223746291,
+      "learning_rate": 4.337090713902308e-05,
+      "loss": 0.622,
+      "step": 454
+    },
+    {
+      "epoch": 0.6594202898550725,
+      "grad_norm": 0.35218029913587057,
+      "learning_rate": 4.334406870638755e-05,
+      "loss": 0.6485,
+      "step": 455
+    },
+    {
+      "epoch": 0.6608695652173913,
+      "grad_norm": 0.4214000155582382,
+      "learning_rate": 4.331723027375201e-05,
+      "loss": 0.5555,
+      "step": 456
+    },
+    {
+      "epoch": 0.6623188405797101,
+      "grad_norm": 0.36980449511799784,
+      "learning_rate": 4.329039184111648e-05,
+      "loss": 0.7021,
+      "step": 457
+    },
+    {
+      "epoch": 0.663768115942029,
+      "grad_norm": 0.3833858984633924,
+      "learning_rate": 4.326355340848094e-05,
+      "loss": 0.5455,
+      "step": 458
+    },
+    {
+      "epoch": 0.6652173913043479,
+      "grad_norm": 0.3621748626108088,
+      "learning_rate": 4.323671497584541e-05,
+      "loss": 0.6451,
+      "step": 459
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.44501860079574185,
+      "learning_rate": 4.3209876543209875e-05,
+      "loss": 0.6063,
+      "step": 460
+    },
+    {
+      "epoch": 0.6681159420289855,
+      "grad_norm": 0.40740730053072266,
+      "learning_rate": 4.3183038110574344e-05,
+      "loss": 0.6471,
+      "step": 461
+    },
+    {
+      "epoch": 0.6695652173913044,
+      "grad_norm": 0.42454592710002687,
+      "learning_rate": 4.315619967793881e-05,
+      "loss": 0.6733,
+      "step": 462
+    },
+    {
+      "epoch": 0.6710144927536232,
+      "grad_norm": 0.4413724435680676,
+      "learning_rate": 4.312936124530328e-05,
+      "loss": 0.5883,
+      "step": 463
+    },
+    {
+      "epoch": 0.672463768115942,
+      "grad_norm": 0.38871349519437576,
+      "learning_rate": 4.3102522812667746e-05,
+      "loss": 0.7363,
+      "step": 464
+    },
+    {
+      "epoch": 0.6739130434782609,
+      "grad_norm": 0.5312564883836602,
+      "learning_rate": 4.307568438003221e-05,
+      "loss": 0.5356,
+      "step": 465
+    },
+    {
+      "epoch": 0.6753623188405797,
+      "grad_norm": 0.5250394278357895,
+      "learning_rate": 4.304884594739668e-05,
+      "loss": 0.6368,
+      "step": 466
+    },
+    {
+      "epoch": 0.6768115942028986,
+      "grad_norm": 0.5002775405148884,
+      "learning_rate": 4.302200751476114e-05,
+      "loss": 0.6285,
+      "step": 467
+    },
+    {
+      "epoch": 0.6782608695652174,
+      "grad_norm": 0.6010481334870542,
+      "learning_rate": 4.299516908212561e-05,
+      "loss": 0.5958,
+      "step": 468
+    },
+    {
+      "epoch": 0.6797101449275362,
+      "grad_norm": 0.5137320966075457,
+      "learning_rate": 4.296833064949007e-05,
+      "loss": 0.6357,
+      "step": 469
+    },
+    {
+      "epoch": 0.6811594202898551,
+      "grad_norm": 0.7634800458708868,
+      "learning_rate": 4.294149221685454e-05,
+      "loss": 0.6266,
+      "step": 470
+    },
+    {
+      "epoch": 0.6826086956521739,
+      "grad_norm": 0.5412119314366018,
+      "learning_rate": 4.2914653784219006e-05,
+      "loss": 0.6257,
+      "step": 471
+    },
+    {
+      "epoch": 0.6840579710144927,
+      "grad_norm": 0.46005094137484503,
+      "learning_rate": 4.288781535158347e-05,
+      "loss": 0.6269,
+      "step": 472
+    },
+    {
+      "epoch": 0.6855072463768116,
+      "grad_norm": 0.5821703799591006,
+      "learning_rate": 4.286097691894794e-05,
+      "loss": 0.6145,
+      "step": 473
+    },
+    {
+      "epoch": 0.6869565217391305,
+      "grad_norm": 0.4076874787060111,
+      "learning_rate": 4.28341384863124e-05,
+      "loss": 0.6048,
+      "step": 474
+    },
+    {
+      "epoch": 0.6884057971014492,
+      "grad_norm": 0.5026991240254921,
+      "learning_rate": 4.280730005367687e-05,
+      "loss": 0.6283,
+      "step": 475
+    },
+    {
+      "epoch": 0.6898550724637681,
+      "grad_norm": 0.4994831943753748,
+      "learning_rate": 4.278046162104133e-05,
+      "loss": 0.5844,
+      "step": 476
+    },
+    {
+      "epoch": 0.691304347826087,
+      "grad_norm": 0.5923231155149884,
+      "learning_rate": 4.27536231884058e-05,
+      "loss": 0.6238,
+      "step": 477
+    },
+    {
+      "epoch": 0.6927536231884058,
+      "grad_norm": 4.1329790925392755,
+      "learning_rate": 4.2726784755770265e-05,
+      "loss": 0.6822,
+      "step": 478
+    },
+    {
+      "epoch": 0.6942028985507246,
+      "grad_norm": 1.232119030556872,
+      "learning_rate": 4.2699946323134735e-05,
+      "loss": 0.6798,
+      "step": 479
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 0.75778686644872,
+      "learning_rate": 4.26731078904992e-05,
+      "loss": 0.5747,
+      "step": 480
+    },
+    {
+      "epoch": 0.6971014492753623,
+      "grad_norm": 0.5303268834658762,
+      "learning_rate": 4.264626945786366e-05,
+      "loss": 0.4669,
+      "step": 481
+    },
+    {
+      "epoch": 0.6985507246376812,
+      "grad_norm": 0.6978339100881147,
+      "learning_rate": 4.261943102522813e-05,
+      "loss": 0.6481,
+      "step": 482
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5390133158454951,
+      "learning_rate": 4.259259259259259e-05,
+      "loss": 0.5633,
+      "step": 483
+    },
+    {
+      "epoch": 0.7014492753623188,
+      "grad_norm": 0.6534947509239906,
+      "learning_rate": 4.256575415995706e-05,
+      "loss": 0.5855,
+      "step": 484
+    },
+    {
+      "epoch": 0.7028985507246377,
+      "grad_norm": 0.6128904403486618,
+      "learning_rate": 4.2538915727321525e-05,
+      "loss": 0.65,
+      "step": 485
+    },
+    {
+      "epoch": 0.7043478260869566,
+      "grad_norm": 0.6224056264214514,
+      "learning_rate": 4.2512077294685994e-05,
+      "loss": 0.6994,
+      "step": 486
+    },
+    {
+      "epoch": 0.7057971014492753,
+      "grad_norm": 2.921316276290605,
+      "learning_rate": 4.248523886205046e-05,
+      "loss": 0.6553,
+      "step": 487
+    },
+    {
+      "epoch": 0.7072463768115942,
+      "grad_norm": 1.094977952559785,
+      "learning_rate": 4.2458400429414926e-05,
+      "loss": 0.6761,
+      "step": 488
+    },
+    {
+      "epoch": 0.7086956521739131,
+      "grad_norm": 0.8497515292463685,
+      "learning_rate": 4.243156199677939e-05,
+      "loss": 0.5919,
+      "step": 489
+    },
+    {
+      "epoch": 0.7101449275362319,
+      "grad_norm": 0.7893564837809413,
+      "learning_rate": 4.240472356414386e-05,
+      "loss": 0.6577,
+      "step": 490
+    },
+    {
+      "epoch": 0.7115942028985507,
+      "grad_norm": 1.1274908966110668,
+      "learning_rate": 4.237788513150832e-05,
+      "loss": 0.7219,
+      "step": 491
+    },
+    {
+      "epoch": 0.7130434782608696,
+      "grad_norm": 0.5753539741053254,
+      "learning_rate": 4.2351046698872784e-05,
+      "loss": 0.5748,
+      "step": 492
+    },
+    {
+      "epoch": 0.7144927536231884,
+      "grad_norm": 0.976724089496432,
+      "learning_rate": 4.2324208266237254e-05,
+      "loss": 0.5696,
+      "step": 493
+    },
+    {
+      "epoch": 0.7159420289855073,
+      "grad_norm": 0.6633654676572153,
+      "learning_rate": 4.2297369833601716e-05,
+      "loss": 0.6226,
+      "step": 494
+    },
+    {
+      "epoch": 0.717391304347826,
+      "grad_norm": 0.9829938269432279,
+      "learning_rate": 4.2270531400966186e-05,
+      "loss": 0.6031,
+      "step": 495
+    },
+    {
+      "epoch": 0.7188405797101449,
+      "grad_norm": 0.5633879991151081,
+      "learning_rate": 4.224369296833065e-05,
+      "loss": 0.6086,
+      "step": 496
+    },
+    {
+      "epoch": 0.7202898550724638,
+      "grad_norm": 0.9430497432801044,
+      "learning_rate": 4.221685453569512e-05,
+      "loss": 0.5859,
+      "step": 497
+    },
+    {
+      "epoch": 0.7217391304347827,
+      "grad_norm": 0.46009916937328854,
+      "learning_rate": 4.219001610305958e-05,
+      "loss": 0.6064,
+      "step": 498
+    },
+    {
+      "epoch": 0.7231884057971014,
+      "grad_norm": 0.7216715180796771,
+      "learning_rate": 4.216317767042405e-05,
+      "loss": 0.5473,
+      "step": 499
+    },
+    {
+      "epoch": 0.7246376811594203,
+      "grad_norm": 0.4509522203864577,
+      "learning_rate": 4.213633923778851e-05,
+      "loss": 0.5709,
+      "step": 500
+    },
+    {
+      "epoch": 0.7260869565217392,
+      "grad_norm": 0.523588042133622,
+      "learning_rate": 4.210950080515298e-05,
+      "loss": 0.5996,
+      "step": 501
+    },
+    {
+      "epoch": 0.7275362318840579,
+      "grad_norm": 0.4624176684479565,
+      "learning_rate": 4.2082662372517445e-05,
+      "loss": 0.5835,
+      "step": 502
+    },
+    {
+      "epoch": 0.7289855072463768,
+      "grad_norm": 0.43623626036112917,
+      "learning_rate": 4.205582393988191e-05,
+      "loss": 0.6786,
+      "step": 503
+    },
+    {
+      "epoch": 0.7304347826086957,
+      "grad_norm": 0.4542767125120218,
+      "learning_rate": 4.202898550724638e-05,
+      "loss": 0.6512,
+      "step": 504
+    },
+    {
+      "epoch": 0.7318840579710145,
+      "grad_norm": 14.137605164626205,
+      "learning_rate": 4.200214707461084e-05,
+      "loss": 0.7138,
+      "step": 505
+    },
+    {
+      "epoch": 0.7333333333333333,
+      "grad_norm": 0.5978196809747577,
+      "learning_rate": 4.197530864197531e-05,
+      "loss": 0.6954,
+      "step": 506
+    },
+    {
+      "epoch": 0.7347826086956522,
+      "grad_norm": 0.47590719779152063,
+      "learning_rate": 4.194847020933977e-05,
+      "loss": 0.5772,
+      "step": 507
+    },
+    {
+      "epoch": 0.736231884057971,
+      "grad_norm": 0.48674586940054604,
+      "learning_rate": 4.192163177670424e-05,
+      "loss": 0.6152,
+      "step": 508
+    },
+    {
+      "epoch": 0.7376811594202899,
+      "grad_norm": 0.5344336837088249,
+      "learning_rate": 4.1894793344068705e-05,
+      "loss": 0.5918,
+      "step": 509
+    },
+    {
+      "epoch": 0.7391304347826086,
+      "grad_norm": 0.4642887974191174,
+      "learning_rate": 4.1867954911433174e-05,
+      "loss": 0.6326,
+      "step": 510
+    },
+    {
+      "epoch": 0.7405797101449275,
+      "grad_norm": 0.5347402718091168,
+      "learning_rate": 4.1841116478797644e-05,
+      "loss": 0.5764,
+      "step": 511
+    },
+    {
+      "epoch": 0.7420289855072464,
+      "grad_norm": 0.5438038315928879,
+      "learning_rate": 4.181427804616211e-05,
+      "loss": 0.6002,
+      "step": 512
+    },
+    {
+      "epoch": 0.7434782608695653,
+      "grad_norm": 0.4825629049018045,
+      "learning_rate": 4.1787439613526576e-05,
+      "loss": 0.5937,
+      "step": 513
+    },
+    {
+      "epoch": 0.744927536231884,
+      "grad_norm": 0.48979088485780425,
+      "learning_rate": 4.176060118089104e-05,
+      "loss": 0.6404,
+      "step": 514
+    },
+    {
+      "epoch": 0.7463768115942029,
+      "grad_norm": 0.4620698383348423,
+      "learning_rate": 4.173376274825551e-05,
+      "loss": 0.5397,
+      "step": 515
+    },
+    {
+      "epoch": 0.7478260869565218,
+      "grad_norm": 0.44176580357824313,
+      "learning_rate": 4.170692431561997e-05,
+      "loss": 0.6207,
+      "step": 516
+    },
+    {
+      "epoch": 0.7492753623188406,
+      "grad_norm": 0.41871753355296315,
+      "learning_rate": 4.168008588298444e-05,
+      "loss": 0.6503,
+      "step": 517
+    },
+    {
+      "epoch": 0.7507246376811594,
+      "grad_norm": 0.47948444268186163,
+      "learning_rate": 4.1653247450348903e-05,
+      "loss": 0.563,
+      "step": 518
+    },
+    {
+      "epoch": 0.7521739130434782,
+      "grad_norm": 0.3822045549344485,
+      "learning_rate": 4.1626409017713366e-05,
+      "loss": 0.5774,
+      "step": 519
+    },
+    {
+      "epoch": 0.7536231884057971,
+      "grad_norm": 0.538431027647131,
+      "learning_rate": 4.1599570585077836e-05,
+      "loss": 0.6085,
+      "step": 520
+    },
+    {
+      "epoch": 0.755072463768116,
+      "grad_norm": 0.35528969092125745,
+      "learning_rate": 4.15727321524423e-05,
+      "loss": 0.6009,
+      "step": 521
+    },
+    {
+      "epoch": 0.7565217391304347,
+      "grad_norm": 0.4515347734731531,
+      "learning_rate": 4.154589371980677e-05,
+      "loss": 0.5579,
+      "step": 522
+    },
+    {
+      "epoch": 0.7579710144927536,
+      "grad_norm": 0.40842988456854906,
+      "learning_rate": 4.151905528717123e-05,
+      "loss": 0.6111,
+      "step": 523
+    },
+    {
+      "epoch": 0.7594202898550725,
+      "grad_norm": 0.4091827511039516,
+      "learning_rate": 4.14922168545357e-05,
+      "loss": 0.6287,
+      "step": 524
+    },
+    {
+      "epoch": 0.7608695652173914,
+      "grad_norm": 0.49858433261609175,
+      "learning_rate": 4.146537842190016e-05,
+      "loss": 0.6857,
+      "step": 525
+    },
+    {
+      "epoch": 0.7623188405797101,
+      "grad_norm": 0.4983638734374692,
+      "learning_rate": 4.143853998926463e-05,
+      "loss": 0.6053,
+      "step": 526
+    },
+    {
+      "epoch": 0.763768115942029,
+      "grad_norm": 0.46242329165053936,
+      "learning_rate": 4.1411701556629095e-05,
+      "loss": 0.5745,
+      "step": 527
+    },
+    {
+      "epoch": 0.7652173913043478,
+      "grad_norm": 0.42992145313121877,
+      "learning_rate": 4.1384863123993565e-05,
+      "loss": 0.6056,
+      "step": 528
+    },
+    {
+      "epoch": 0.7666666666666667,
+      "grad_norm": 0.5733079560388592,
+      "learning_rate": 4.135802469135803e-05,
+      "loss": 0.6339,
+      "step": 529
+    },
+    {
+      "epoch": 0.7681159420289855,
+      "grad_norm": 0.3467646266827582,
+      "learning_rate": 4.133118625872249e-05,
+      "loss": 0.6057,
+      "step": 530
+    },
+    {
+      "epoch": 0.7695652173913043,
+      "grad_norm": 0.4876355404655841,
+      "learning_rate": 4.130434782608696e-05,
+      "loss": 0.5334,
+      "step": 531
+    },
+    {
+      "epoch": 0.7710144927536232,
+      "grad_norm": 0.3993002820075125,
+      "learning_rate": 4.127750939345142e-05,
+      "loss": 0.6431,
+      "step": 532
+    },
+    {
+      "epoch": 0.7724637681159421,
+      "grad_norm": 0.5531641534870261,
+      "learning_rate": 4.125067096081589e-05,
+      "loss": 0.5591,
+      "step": 533
+    },
+    {
+      "epoch": 0.7739130434782608,
+      "grad_norm": 0.44858435855948825,
+      "learning_rate": 4.1223832528180355e-05,
+      "loss": 0.5929,
+      "step": 534
+    },
+    {
+      "epoch": 0.7753623188405797,
+      "grad_norm": 0.6282196147018511,
+      "learning_rate": 4.1196994095544824e-05,
+      "loss": 0.5181,
+      "step": 535
+    },
+    {
+      "epoch": 0.7768115942028986,
+      "grad_norm": 0.3967674108479093,
+      "learning_rate": 4.117015566290929e-05,
+      "loss": 0.59,
+      "step": 536
+    },
+    {
+      "epoch": 0.7782608695652173,
+      "grad_norm": 0.5204142812329497,
+      "learning_rate": 4.1143317230273756e-05,
+      "loss": 0.6374,
+      "step": 537
+    },
+    {
+      "epoch": 0.7797101449275362,
+      "grad_norm": 0.494954976619306,
+      "learning_rate": 4.111647879763822e-05,
+      "loss": 0.6296,
+      "step": 538
+    },
+    {
+      "epoch": 0.7811594202898551,
+      "grad_norm": 0.4706706689193786,
+      "learning_rate": 4.108964036500268e-05,
+      "loss": 0.5833,
+      "step": 539
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 1.512496589807772,
+      "learning_rate": 4.106280193236715e-05,
+      "loss": 0.5858,
+      "step": 540
+    },
+    {
+      "epoch": 0.7840579710144927,
+      "grad_norm": 0.6986263549595149,
+      "learning_rate": 4.1035963499731614e-05,
+      "loss": 0.5599,
+      "step": 541
+    },
+    {
+      "epoch": 0.7855072463768116,
+      "grad_norm": 0.8573296822090767,
+      "learning_rate": 4.1009125067096084e-05,
+      "loss": 0.6397,
+      "step": 542
+    },
+    {
+      "epoch": 0.7869565217391304,
+      "grad_norm": 0.5033595984865323,
+      "learning_rate": 4.0982286634460546e-05,
+      "loss": 0.5046,
+      "step": 543
+    },
+    {
+      "epoch": 0.7884057971014493,
+      "grad_norm": 0.713429803498217,
+      "learning_rate": 4.0955448201825016e-05,
+      "loss": 0.491,
+      "step": 544
+    },
+    {
+      "epoch": 0.7898550724637681,
+      "grad_norm": 0.7848459230690726,
+      "learning_rate": 4.092860976918948e-05,
+      "loss": 0.618,
+      "step": 545
+    },
+    {
+      "epoch": 0.7913043478260869,
+      "grad_norm": 0.844801020655998,
+      "learning_rate": 4.090177133655395e-05,
+      "loss": 0.6478,
+      "step": 546
+    },
+    {
+      "epoch": 0.7927536231884058,
+      "grad_norm": 0.6611299388643612,
+      "learning_rate": 4.087493290391841e-05,
+      "loss": 0.5311,
+      "step": 547
+    },
+    {
+      "epoch": 0.7942028985507247,
+      "grad_norm": 0.6499654815412405,
+      "learning_rate": 4.084809447128288e-05,
+      "loss": 0.5961,
+      "step": 548
+    },
+    {
+      "epoch": 0.7956521739130434,
+      "grad_norm": 0.5252072917572203,
+      "learning_rate": 4.082125603864734e-05,
+      "loss": 0.5862,
+      "step": 549
+    },
+    {
+      "epoch": 0.7971014492753623,
+      "grad_norm": 0.5207296431736637,
+      "learning_rate": 4.0794417606011806e-05,
+      "loss": 0.5172,
+      "step": 550
+    },
+    {
+      "epoch": 0.7985507246376812,
+      "grad_norm": 0.4689572540878984,
+      "learning_rate": 4.0767579173376275e-05,
+      "loss": 0.5726,
+      "step": 551
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5458681731232032,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 0.599,
+      "step": 552
+    },
+    {
+      "epoch": 0.8014492753623188,
+      "grad_norm": 0.38372406832517136,
+      "learning_rate": 4.071390230810521e-05,
+      "loss": 0.5514,
+      "step": 553
+    },
+    {
+      "epoch": 0.8028985507246377,
+      "grad_norm": 0.5094489821646551,
+      "learning_rate": 4.068706387546967e-05,
+      "loss": 0.5899,
+      "step": 554
+    },
+    {
+      "epoch": 0.8043478260869565,
+      "grad_norm": 2.1680852632462835,
+      "learning_rate": 4.066022544283414e-05,
+      "loss": 0.6013,
+      "step": 555
+    },
+    {
+      "epoch": 0.8057971014492754,
+      "grad_norm": 0.5692937335137704,
+      "learning_rate": 4.06333870101986e-05,
+      "loss": 0.6152,
+      "step": 556
+    },
+    {
+      "epoch": 0.8072463768115942,
+      "grad_norm": 0.4703136901533402,
+      "learning_rate": 4.060654857756307e-05,
+      "loss": 0.6275,
+      "step": 557
+    },
+    {
+      "epoch": 0.808695652173913,
+      "grad_norm": 0.5971159611839674,
+      "learning_rate": 4.057971014492754e-05,
+      "loss": 0.5077,
+      "step": 558
+    },
+    {
+      "epoch": 0.8101449275362319,
+      "grad_norm": 1.8068906932840791,
+      "learning_rate": 4.0552871712292004e-05,
+      "loss": 0.6552,
+      "step": 559
+    },
+    {
+      "epoch": 0.8115942028985508,
+      "grad_norm": 0.5449109715967542,
+      "learning_rate": 4.0526033279656474e-05,
+      "loss": 0.6109,
+      "step": 560
+    },
+    {
+      "epoch": 0.8130434782608695,
+      "grad_norm": 0.43388002270079523,
+      "learning_rate": 4.049919484702094e-05,
+      "loss": 0.6131,
+      "step": 561
+    },
+    {
+      "epoch": 0.8144927536231884,
+      "grad_norm": 0.4211682095005935,
+      "learning_rate": 4.0472356414385406e-05,
+      "loss": 0.5582,
+      "step": 562
+    },
+    {
+      "epoch": 0.8159420289855073,
+      "grad_norm": 0.49058073662614093,
+      "learning_rate": 4.044551798174987e-05,
+      "loss": 0.5524,
+      "step": 563
+    },
+    {
+      "epoch": 0.8173913043478261,
+      "grad_norm": 0.4473245845954027,
+      "learning_rate": 4.041867954911434e-05,
+      "loss": 0.5604,
+      "step": 564
+    },
+    {
+      "epoch": 0.8188405797101449,
+      "grad_norm": 0.4356196966368169,
+      "learning_rate": 4.03918411164788e-05,
+      "loss": 0.5579,
+      "step": 565
+    },
+    {
+      "epoch": 0.8202898550724638,
+      "grad_norm": 0.3670371817158889,
+      "learning_rate": 4.0365002683843264e-05,
+      "loss": 0.6055,
+      "step": 566
+    },
+    {
+      "epoch": 0.8217391304347826,
+      "grad_norm": 0.5140778620459774,
+      "learning_rate": 4.0338164251207733e-05,
+      "loss": 0.6115,
+      "step": 567
+    },
+    {
+      "epoch": 0.8231884057971014,
+      "grad_norm": 1.346065214744829,
+      "learning_rate": 4.0311325818572196e-05,
+      "loss": 0.6003,
+      "step": 568
+    },
+    {
+      "epoch": 0.8246376811594203,
+      "grad_norm": 0.8248383138671131,
+      "learning_rate": 4.0284487385936666e-05,
+      "loss": 0.7642,
+      "step": 569
+    },
+    {
+      "epoch": 0.8260869565217391,
+      "grad_norm": 0.5120287581760506,
+      "learning_rate": 4.025764895330113e-05,
+      "loss": 0.6164,
+      "step": 570
+    },
+    {
+      "epoch": 0.827536231884058,
+      "grad_norm": 0.4743646973120087,
+      "learning_rate": 4.02308105206656e-05,
+      "loss": 0.549,
+      "step": 571
+    },
+    {
+      "epoch": 0.8289855072463768,
+      "grad_norm": 3.3210188043275375,
+      "learning_rate": 4.020397208803006e-05,
+      "loss": 0.667,
+      "step": 572
+    },
+    {
+      "epoch": 0.8304347826086956,
+      "grad_norm": 0.6837248648294643,
+      "learning_rate": 4.017713365539453e-05,
+      "loss": 0.533,
+      "step": 573
+    },
+    {
+      "epoch": 0.8318840579710145,
+      "grad_norm": 0.516129522264584,
+      "learning_rate": 4.015029522275899e-05,
+      "loss": 0.6341,
+      "step": 574
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 2.7438937595276647,
+      "learning_rate": 4.012345679012346e-05,
+      "loss": 0.6521,
+      "step": 575
+    },
+    {
+      "epoch": 0.8347826086956521,
+      "grad_norm": 0.8592649758656578,
+      "learning_rate": 4.0096618357487925e-05,
+      "loss": 0.5506,
+      "step": 576
+    },
+    {
+      "epoch": 0.836231884057971,
+      "grad_norm": 1.5555303566525747,
+      "learning_rate": 4.006977992485239e-05,
+      "loss": 0.5788,
+      "step": 577
+    },
+    {
+      "epoch": 0.8376811594202899,
+      "grad_norm": 0.6899087860610227,
+      "learning_rate": 4.004294149221686e-05,
+      "loss": 0.5876,
+      "step": 578
+    },
+    {
+      "epoch": 0.8391304347826087,
+      "grad_norm": 0.7046190555629226,
+      "learning_rate": 4.001610305958132e-05,
+      "loss": 0.6044,
+      "step": 579
+    },
+    {
+      "epoch": 0.8405797101449275,
+      "grad_norm": 0.5740571870652859,
+      "learning_rate": 3.998926462694579e-05,
+      "loss": 0.5523,
+      "step": 580
+    },
+    {
+      "epoch": 0.8420289855072464,
+      "grad_norm": 3.0856739801665896,
+      "learning_rate": 3.996242619431025e-05,
+      "loss": 0.6992,
+      "step": 581
+    },
+    {
+      "epoch": 0.8434782608695652,
+      "grad_norm": 0.6217031879385219,
+      "learning_rate": 3.993558776167472e-05,
+      "loss": 0.5843,
+      "step": 582
+    },
+    {
+      "epoch": 0.8449275362318841,
+      "grad_norm": 0.5855584870859855,
+      "learning_rate": 3.9908749329039185e-05,
+      "loss": 0.804,
+      "step": 583
+    },
+    {
+      "epoch": 0.8463768115942029,
+      "grad_norm": 0.5127730494036674,
+      "learning_rate": 3.9881910896403654e-05,
+      "loss": 0.5784,
+      "step": 584
+    },
+    {
+      "epoch": 0.8478260869565217,
+      "grad_norm": 0.48011376646056936,
+      "learning_rate": 3.985507246376812e-05,
+      "loss": 0.5245,
+      "step": 585
+    },
+    {
+      "epoch": 0.8492753623188406,
+      "grad_norm": 0.5894489075240219,
+      "learning_rate": 3.982823403113258e-05,
+      "loss": 0.6293,
+      "step": 586
+    },
+    {
+      "epoch": 0.8507246376811595,
+      "grad_norm": 0.5378776256572029,
+      "learning_rate": 3.980139559849705e-05,
+      "loss": 0.6049,
+      "step": 587
+    },
+    {
+      "epoch": 0.8521739130434782,
+      "grad_norm": 0.7083934176689143,
+      "learning_rate": 3.977455716586151e-05,
+      "loss": 0.6418,
+      "step": 588
+    },
+    {
+      "epoch": 0.8536231884057971,
+      "grad_norm": 0.6190476116945333,
+      "learning_rate": 3.974771873322598e-05,
+      "loss": 0.5592,
+      "step": 589
+    },
+    {
+      "epoch": 0.855072463768116,
+      "grad_norm": 0.4120642741902257,
+      "learning_rate": 3.9720880300590444e-05,
+      "loss": 0.5752,
+      "step": 590
+    },
+    {
+      "epoch": 0.8565217391304348,
+      "grad_norm": 0.6013798835291985,
+      "learning_rate": 3.9694041867954914e-05,
+      "loss": 0.6497,
+      "step": 591
+    },
+    {
+      "epoch": 0.8579710144927536,
+      "grad_norm": 0.44171429178629873,
+      "learning_rate": 3.9667203435319376e-05,
+      "loss": 0.6097,
+      "step": 592
+    },
+    {
+      "epoch": 0.8594202898550725,
+      "grad_norm": 0.53452983650377,
+      "learning_rate": 3.9640365002683846e-05,
+      "loss": 0.6558,
+      "step": 593
+    },
+    {
+      "epoch": 0.8608695652173913,
+      "grad_norm": 0.4186097201166055,
+      "learning_rate": 3.961352657004831e-05,
+      "loss": 0.7236,
+      "step": 594
+    },
+    {
+      "epoch": 0.8623188405797102,
+      "grad_norm": 0.4645596168717277,
+      "learning_rate": 3.958668813741278e-05,
+      "loss": 0.5854,
+      "step": 595
+    },
+    {
+      "epoch": 0.863768115942029,
+      "grad_norm": 0.43029743789021974,
+      "learning_rate": 3.955984970477724e-05,
+      "loss": 0.6233,
+      "step": 596
+    },
+    {
+      "epoch": 0.8652173913043478,
+      "grad_norm": 0.43292776931294014,
+      "learning_rate": 3.9533011272141704e-05,
+      "loss": 0.6439,
+      "step": 597
+    },
+    {
+      "epoch": 0.8666666666666667,
+      "grad_norm": 0.43423115706910775,
+      "learning_rate": 3.950617283950617e-05,
+      "loss": 0.589,
+      "step": 598
+    },
+    {
+      "epoch": 0.8681159420289855,
+      "grad_norm": 0.4755946805708221,
+      "learning_rate": 3.9479334406870636e-05,
+      "loss": 0.586,
+      "step": 599
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 0.4341006298363921,
+      "learning_rate": 3.9452495974235105e-05,
+      "loss": 0.5903,
+      "step": 600
+    },
+    {
+      "epoch": 0.8710144927536232,
+      "grad_norm": 0.4718786944262586,
+      "learning_rate": 3.942565754159957e-05,
+      "loss": 0.5135,
+      "step": 601
+    },
+    {
+      "epoch": 0.8724637681159421,
+      "grad_norm": 0.40791241360653085,
+      "learning_rate": 3.939881910896404e-05,
+      "loss": 0.6038,
+      "step": 602
+    },
+    {
+      "epoch": 0.8739130434782608,
+      "grad_norm": 0.5059495818189851,
+      "learning_rate": 3.937198067632851e-05,
+      "loss": 0.6354,
+      "step": 603
+    },
+    {
+      "epoch": 0.8753623188405797,
+      "grad_norm": 0.431048802834448,
+      "learning_rate": 3.934514224369297e-05,
+      "loss": 0.6642,
+      "step": 604
+    },
+    {
+      "epoch": 0.8768115942028986,
+      "grad_norm": 0.4965530971692437,
+      "learning_rate": 3.931830381105744e-05,
+      "loss": 0.627,
+      "step": 605
+    },
+    {
+      "epoch": 0.8782608695652174,
+      "grad_norm": 0.46139203603720497,
+      "learning_rate": 3.92914653784219e-05,
+      "loss": 0.6329,
+      "step": 606
+    },
+    {
+      "epoch": 0.8797101449275362,
+      "grad_norm": 0.42125965185438957,
+      "learning_rate": 3.926462694578637e-05,
+      "loss": 0.5932,
+      "step": 607
+    },
+    {
+      "epoch": 0.881159420289855,
+      "grad_norm": 0.367329901744191,
+      "learning_rate": 3.9237788513150834e-05,
+      "loss": 0.6201,
+      "step": 608
+    },
+    {
+      "epoch": 0.8826086956521739,
+      "grad_norm": 0.4524070054726798,
+      "learning_rate": 3.9210950080515304e-05,
+      "loss": 0.5822,
+      "step": 609
+    },
+    {
+      "epoch": 0.8840579710144928,
+      "grad_norm": 0.467261697364059,
+      "learning_rate": 3.918411164787977e-05,
+      "loss": 0.4836,
+      "step": 610
+    },
+    {
+      "epoch": 0.8855072463768116,
+      "grad_norm": 0.4463128896050495,
+      "learning_rate": 3.9157273215244236e-05,
+      "loss": 0.4819,
+      "step": 611
+    },
+    {
+      "epoch": 0.8869565217391304,
+      "grad_norm": 0.40526007809503917,
+      "learning_rate": 3.91304347826087e-05,
+      "loss": 0.551,
+      "step": 612
+    },
+    {
+      "epoch": 0.8884057971014493,
+      "grad_norm": 0.41385125651458887,
+      "learning_rate": 3.910359634997316e-05,
+      "loss": 0.6159,
+      "step": 613
+    },
+    {
+      "epoch": 0.8898550724637682,
+      "grad_norm": 0.41673986104362226,
+      "learning_rate": 3.907675791733763e-05,
+      "loss": 0.6247,
+      "step": 614
+    },
+    {
+      "epoch": 0.8913043478260869,
+      "grad_norm": 0.36308482982577767,
+      "learning_rate": 3.9049919484702094e-05,
+      "loss": 0.6022,
+      "step": 615
+    },
+    {
+      "epoch": 0.8927536231884058,
+      "grad_norm": 0.38126692153672176,
+      "learning_rate": 3.9023081052066564e-05,
+      "loss": 0.6281,
+      "step": 616
+    },
+    {
+      "epoch": 0.8942028985507247,
+      "grad_norm": 0.33298531744437676,
+      "learning_rate": 3.8996242619431026e-05,
+      "loss": 0.5867,
+      "step": 617
+    },
+    {
+      "epoch": 0.8956521739130435,
+      "grad_norm": 0.3352484634893385,
+      "learning_rate": 3.8969404186795496e-05,
+      "loss": 0.5519,
+      "step": 618
+    },
+    {
+      "epoch": 0.8971014492753623,
+      "grad_norm": 0.37639114214176045,
+      "learning_rate": 3.894256575415996e-05,
+      "loss": 0.5195,
+      "step": 619
+    },
+    {
+      "epoch": 0.8985507246376812,
+      "grad_norm": 0.34048760325095795,
+      "learning_rate": 3.891572732152443e-05,
+      "loss": 0.5787,
+      "step": 620
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3832821048268026,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 0.5247,
+      "step": 621
+    },
+    {
+      "epoch": 0.9014492753623189,
+      "grad_norm": 0.40064194244810836,
+      "learning_rate": 3.886205045625336e-05,
+      "loss": 0.6286,
+      "step": 622
+    },
+    {
+      "epoch": 0.9028985507246376,
+      "grad_norm": 0.40965106020853814,
+      "learning_rate": 3.883521202361782e-05,
+      "loss": 0.5611,
+      "step": 623
+    },
+    {
+      "epoch": 0.9043478260869565,
+      "grad_norm": 0.4952521908568183,
+      "learning_rate": 3.8808373590982286e-05,
+      "loss": 0.6228,
+      "step": 624
+    },
+    {
+      "epoch": 0.9057971014492754,
+      "grad_norm": 0.3165074112635924,
+      "learning_rate": 3.8781535158346755e-05,
+      "loss": 0.5923,
+      "step": 625
+    },
+    {
+      "epoch": 0.9072463768115943,
+      "grad_norm": 0.3593150540525319,
+      "learning_rate": 3.875469672571122e-05,
+      "loss": 0.5492,
+      "step": 626
+    },
+    {
+      "epoch": 0.908695652173913,
+      "grad_norm": 0.30592808729109405,
+      "learning_rate": 3.872785829307569e-05,
+      "loss": 0.5315,
+      "step": 627
+    },
+    {
+      "epoch": 0.9101449275362319,
+      "grad_norm": 0.34550819407901817,
+      "learning_rate": 3.870101986044015e-05,
+      "loss": 0.5522,
+      "step": 628
+    },
+    {
+      "epoch": 0.9115942028985508,
+      "grad_norm": 0.45826140995448045,
+      "learning_rate": 3.867418142780462e-05,
+      "loss": 0.5927,
+      "step": 629
+    },
+    {
+      "epoch": 0.9130434782608695,
+      "grad_norm": 0.4563119771926975,
+      "learning_rate": 3.864734299516908e-05,
+      "loss": 0.6512,
+      "step": 630
+    },
+    {
+      "epoch": 0.9144927536231884,
+      "grad_norm": 0.38702857798816853,
+      "learning_rate": 3.862050456253355e-05,
+      "loss": 0.6182,
+      "step": 631
+    },
+    {
+      "epoch": 0.9159420289855073,
+      "grad_norm": 0.3900901657498249,
+      "learning_rate": 3.8593666129898015e-05,
+      "loss": 0.5362,
+      "step": 632
+    },
+    {
+      "epoch": 0.9173913043478261,
+      "grad_norm": 0.44337248837753757,
+      "learning_rate": 3.8566827697262484e-05,
+      "loss": 0.6322,
+      "step": 633
+    },
+    {
+      "epoch": 0.9188405797101449,
+      "grad_norm": 0.35254524944898624,
+      "learning_rate": 3.853998926462695e-05,
+      "loss": 0.5645,
+      "step": 634
+    },
+    {
+      "epoch": 0.9202898550724637,
+      "grad_norm": 0.42395525642297494,
+      "learning_rate": 3.851315083199141e-05,
+      "loss": 0.6362,
+      "step": 635
+    },
+    {
+      "epoch": 0.9217391304347826,
+      "grad_norm": 0.4361979901795581,
+      "learning_rate": 3.848631239935588e-05,
+      "loss": 0.6492,
+      "step": 636
+    },
+    {
+      "epoch": 0.9231884057971015,
+      "grad_norm": 0.46478955582935616,
+      "learning_rate": 3.845947396672034e-05,
+      "loss": 0.5464,
+      "step": 637
+    },
+    {
+      "epoch": 0.9246376811594202,
+      "grad_norm": 0.4138224106120635,
+      "learning_rate": 3.843263553408481e-05,
+      "loss": 0.6005,
+      "step": 638
+    },
+    {
+      "epoch": 0.9260869565217391,
+      "grad_norm": 0.455808647015972,
+      "learning_rate": 3.8405797101449274e-05,
+      "loss": 0.6613,
+      "step": 639
+    },
+    {
+      "epoch": 0.927536231884058,
+      "grad_norm": 0.427802788147173,
+      "learning_rate": 3.8378958668813744e-05,
+      "loss": 0.6763,
+      "step": 640
+    },
+    {
+      "epoch": 0.9289855072463769,
+      "grad_norm": 0.3811292517396575,
+      "learning_rate": 3.8352120236178207e-05,
+      "loss": 0.5262,
+      "step": 641
+    },
+    {
+      "epoch": 0.9304347826086956,
+      "grad_norm": 0.43191325483841736,
+      "learning_rate": 3.8325281803542676e-05,
+      "loss": 0.6173,
+      "step": 642
+    },
+    {
+      "epoch": 0.9318840579710145,
+      "grad_norm": 0.42403510011211526,
+      "learning_rate": 3.829844337090714e-05,
+      "loss": 0.6037,
+      "step": 643
+    },
+    {
+      "epoch": 0.9333333333333333,
+      "grad_norm": 0.41707900019612226,
+      "learning_rate": 3.82716049382716e-05,
+      "loss": 0.6085,
+      "step": 644
+    },
+    {
+      "epoch": 0.9347826086956522,
+      "grad_norm": 0.3505630178456613,
+      "learning_rate": 3.824476650563607e-05,
+      "loss": 0.5158,
+      "step": 645
+    },
+    {
+      "epoch": 0.936231884057971,
+      "grad_norm": 0.3960236953328549,
+      "learning_rate": 3.8217928073000534e-05,
+      "loss": 0.5711,
+      "step": 646
+    },
+    {
+      "epoch": 0.9376811594202898,
+      "grad_norm": 0.4105260806742507,
+      "learning_rate": 3.8191089640365e-05,
+      "loss": 0.5858,
+      "step": 647
+    },
+    {
+      "epoch": 0.9391304347826087,
+      "grad_norm": 0.33329816188212635,
+      "learning_rate": 3.8164251207729466e-05,
+      "loss": 0.6197,
+      "step": 648
+    },
+    {
+      "epoch": 0.9405797101449276,
+      "grad_norm": 0.44958808906007697,
+      "learning_rate": 3.8137412775093936e-05,
+      "loss": 0.6251,
+      "step": 649
+    },
+    {
+      "epoch": 0.9420289855072463,
+      "grad_norm": 0.3560872934225823,
+      "learning_rate": 3.8110574342458405e-05,
+      "loss": 0.5913,
+      "step": 650
+    },
+    {
+      "epoch": 0.9434782608695652,
+      "grad_norm": 0.3269136684995382,
+      "learning_rate": 3.808373590982287e-05,
+      "loss": 0.5154,
+      "step": 651
+    },
+    {
+      "epoch": 0.9449275362318841,
+      "grad_norm": 0.447727731046666,
+      "learning_rate": 3.805689747718734e-05,
+      "loss": 0.6371,
+      "step": 652
+    },
+    {
+      "epoch": 0.946376811594203,
+      "grad_norm": 0.37864338247741586,
+      "learning_rate": 3.80300590445518e-05,
+      "loss": 0.5815,
+      "step": 653
+    },
+    {
+      "epoch": 0.9478260869565217,
+      "grad_norm": 0.37617318149727297,
+      "learning_rate": 3.800322061191627e-05,
+      "loss": 0.5994,
+      "step": 654
+    },
+    {
+      "epoch": 0.9492753623188406,
+      "grad_norm": 0.37249479290176424,
+      "learning_rate": 3.797638217928073e-05,
+      "loss": 0.5855,
+      "step": 655
+    },
+    {
+      "epoch": 0.9507246376811594,
+      "grad_norm": 0.37349529397279085,
+      "learning_rate": 3.79495437466452e-05,
+      "loss": 0.6366,
+      "step": 656
+    },
+    {
+      "epoch": 0.9521739130434783,
+      "grad_norm": 0.4064309078465553,
+      "learning_rate": 3.7922705314009665e-05,
+      "loss": 0.564,
+      "step": 657
+    },
+    {
+      "epoch": 0.9536231884057971,
+      "grad_norm": 3.2564378472711915,
+      "learning_rate": 3.7895866881374134e-05,
+      "loss": 0.6199,
+      "step": 658
+    },
+    {
+      "epoch": 0.9550724637681159,
+      "grad_norm": 0.4257622394894502,
+      "learning_rate": 3.78690284487386e-05,
+      "loss": 0.6179,
+      "step": 659
+    },
+    {
+      "epoch": 0.9565217391304348,
+      "grad_norm": 0.36079814194969073,
+      "learning_rate": 3.784219001610306e-05,
+      "loss": 0.6637,
+      "step": 660
+    },
+    {
+      "epoch": 0.9579710144927536,
+      "grad_norm": 0.3546407933331195,
+      "learning_rate": 3.781535158346753e-05,
+      "loss": 0.562,
+      "step": 661
+    },
+    {
+      "epoch": 0.9594202898550724,
+      "grad_norm": 0.3380010456467773,
+      "learning_rate": 3.778851315083199e-05,
+      "loss": 0.6106,
+      "step": 662
+    },
+    {
+      "epoch": 0.9608695652173913,
+      "grad_norm": 0.4450413707456908,
+      "learning_rate": 3.776167471819646e-05,
+      "loss": 0.5588,
+      "step": 663
+    },
+    {
+      "epoch": 0.9623188405797102,
+      "grad_norm": 0.3503137981645195,
+      "learning_rate": 3.7734836285560924e-05,
+      "loss": 0.5943,
+      "step": 664
+    },
+    {
+      "epoch": 0.9637681159420289,
+      "grad_norm": 0.4620932728470443,
+      "learning_rate": 3.7707997852925394e-05,
+      "loss": 0.665,
+      "step": 665
+    },
+    {
+      "epoch": 0.9652173913043478,
+      "grad_norm": 0.35511456322514035,
+      "learning_rate": 3.7681159420289856e-05,
+      "loss": 0.5885,
+      "step": 666
+    },
+    {
+      "epoch": 0.9666666666666667,
+      "grad_norm": 0.3381162980326822,
+      "learning_rate": 3.7654320987654326e-05,
+      "loss": 0.5543,
+      "step": 667
+    },
+    {
+      "epoch": 0.9681159420289855,
+      "grad_norm": 0.40527670866170007,
+      "learning_rate": 3.762748255501879e-05,
+      "loss": 0.5315,
+      "step": 668
+    },
+    {
+      "epoch": 0.9695652173913043,
+      "grad_norm": 0.3436291560239095,
+      "learning_rate": 3.760064412238326e-05,
+      "loss": 0.5304,
+      "step": 669
+    },
+    {
+      "epoch": 0.9710144927536232,
+      "grad_norm": 0.44490914407508536,
+      "learning_rate": 3.757380568974772e-05,
+      "loss": 0.6451,
+      "step": 670
+    },
+    {
+      "epoch": 0.972463768115942,
+      "grad_norm": 0.34071363921885245,
+      "learning_rate": 3.7546967257112184e-05,
+      "loss": 0.6656,
+      "step": 671
+    },
+    {
+      "epoch": 0.9739130434782609,
+      "grad_norm": 0.4274060634917464,
+      "learning_rate": 3.752012882447665e-05,
+      "loss": 0.5736,
+      "step": 672
+    },
+    {
+      "epoch": 0.9753623188405797,
+      "grad_norm": 0.44530097858565637,
+      "learning_rate": 3.7493290391841116e-05,
+      "loss": 0.6671,
+      "step": 673
+    },
+    {
+      "epoch": 0.9768115942028985,
+      "grad_norm": 0.4800429391051743,
+      "learning_rate": 3.7466451959205585e-05,
+      "loss": 0.665,
+      "step": 674
+    },
+    {
+      "epoch": 0.9782608695652174,
+      "grad_norm": 0.5048795469607198,
+      "learning_rate": 3.743961352657005e-05,
+      "loss": 0.664,
+      "step": 675
+    },
+    {
+      "epoch": 0.9797101449275363,
+      "grad_norm": 0.41025966609715236,
+      "learning_rate": 3.741277509393452e-05,
+      "loss": 0.5653,
+      "step": 676
+    },
+    {
+      "epoch": 0.981159420289855,
+      "grad_norm": 0.3704821518151503,
+      "learning_rate": 3.738593666129898e-05,
+      "loss": 0.5432,
+      "step": 677
+    },
+    {
+      "epoch": 0.9826086956521739,
+      "grad_norm": 0.33328023269603463,
+      "learning_rate": 3.735909822866345e-05,
+      "loss": 0.5077,
+      "step": 678
+    },
+    {
+      "epoch": 0.9840579710144928,
+      "grad_norm": 0.36267745822386244,
+      "learning_rate": 3.733225979602791e-05,
+      "loss": 0.5957,
+      "step": 679
+    },
+    {
+      "epoch": 0.9855072463768116,
+      "grad_norm": 0.34543971218031955,
+      "learning_rate": 3.730542136339238e-05,
+      "loss": 0.5911,
+      "step": 680
+    },
+    {
+      "epoch": 0.9869565217391304,
+      "grad_norm": 2.5749761718864033,
+      "learning_rate": 3.7278582930756845e-05,
+      "loss": 0.5624,
+      "step": 681
+    },
+    {
+      "epoch": 0.9884057971014493,
+      "grad_norm": 0.4213680538219344,
+      "learning_rate": 3.725174449812131e-05,
+      "loss": 0.6305,
+      "step": 682
+    },
+    {
+      "epoch": 0.9898550724637681,
+      "grad_norm": 0.3503305518340307,
+      "learning_rate": 3.722490606548578e-05,
+      "loss": 0.5726,
+      "step": 683
+    },
+    {
+      "epoch": 0.991304347826087,
+      "grad_norm": 0.47858361426280377,
+      "learning_rate": 3.719806763285024e-05,
+      "loss": 0.6518,
+      "step": 684
+    },
+    {
+      "epoch": 0.9927536231884058,
+      "grad_norm": 0.40841795810020365,
+      "learning_rate": 3.717122920021471e-05,
+      "loss": 0.6048,
+      "step": 685
+    },
+    {
+      "epoch": 0.9942028985507246,
+      "grad_norm": 0.4159168319432099,
+      "learning_rate": 3.714439076757917e-05,
+      "loss": 0.6226,
+      "step": 686
+    },
+    {
+      "epoch": 0.9956521739130435,
+      "grad_norm": 0.353982673397565,
+      "learning_rate": 3.711755233494364e-05,
+      "loss": 0.5372,
+      "step": 687
+    },
+    {
+      "epoch": 0.9971014492753624,
+      "grad_norm": 0.4708236125419223,
+      "learning_rate": 3.7090713902308104e-05,
+      "loss": 0.5306,
+      "step": 688
+    },
+    {
+      "epoch": 0.9985507246376811,
+      "grad_norm": 0.5050034586864125,
+      "learning_rate": 3.7063875469672574e-05,
+      "loss": 0.5568,
+      "step": 689
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.42561847995684515,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.7395,
+      "step": 690
+    },
+    {
+      "epoch": 1.0014492753623188,
+      "grad_norm": 0.4141984981902665,
+      "learning_rate": 3.70101986044015e-05,
+      "loss": 0.5274,
+      "step": 691
+    },
+    {
+      "epoch": 1.0028985507246377,
+      "grad_norm": 0.40244657538127293,
+      "learning_rate": 3.698336017176597e-05,
+      "loss": 0.5858,
+      "step": 692
+    },
+    {
+      "epoch": 1.0043478260869565,
+      "grad_norm": 0.37737332184052724,
+      "learning_rate": 3.695652173913043e-05,
+      "loss": 0.5323,
+      "step": 693
+    },
+    {
+      "epoch": 1.0057971014492753,
+      "grad_norm": 0.4104832641924855,
+      "learning_rate": 3.69296833064949e-05,
+      "loss": 0.511,
+      "step": 694
+    },
+    {
+      "epoch": 1.0072463768115942,
+      "grad_norm": 0.43539872914512057,
+      "learning_rate": 3.6902844873859364e-05,
+      "loss": 0.5521,
+      "step": 695
+    },
+    {
+      "epoch": 1.008695652173913,
+      "grad_norm": 0.5216741728865025,
+      "learning_rate": 3.687600644122383e-05,
+      "loss": 0.4813,
+      "step": 696
+    },
+    {
+      "epoch": 1.010144927536232,
+      "grad_norm": 0.4751854447776757,
+      "learning_rate": 3.68491680085883e-05,
+      "loss": 0.4749,
+      "step": 697
+    },
+    {
+      "epoch": 1.0115942028985507,
+      "grad_norm": 0.34316199861048846,
+      "learning_rate": 3.6822329575952766e-05,
+      "loss": 0.5377,
+      "step": 698
+    },
+    {
+      "epoch": 1.0130434782608695,
+      "grad_norm": 0.4058139437725442,
+      "learning_rate": 3.6795491143317235e-05,
+      "loss": 0.5336,
+      "step": 699
+    },
+    {
+      "epoch": 1.0144927536231885,
+      "grad_norm": 0.3171721049552687,
+      "learning_rate": 3.67686527106817e-05,
+      "loss": 0.4144,
+      "step": 700
+    },
+    {
+      "epoch": 1.0159420289855072,
+      "grad_norm": 0.3869454431410363,
+      "learning_rate": 3.674181427804617e-05,
+      "loss": 0.5649,
+      "step": 701
+    },
+    {
+      "epoch": 1.017391304347826,
+      "grad_norm": 0.32055242749315693,
+      "learning_rate": 3.671497584541063e-05,
+      "loss": 0.4494,
+      "step": 702
+    },
+    {
+      "epoch": 1.018840579710145,
+      "grad_norm": 0.3525840569410084,
+      "learning_rate": 3.66881374127751e-05,
+      "loss": 0.5125,
+      "step": 703
+    },
+    {
+      "epoch": 1.0202898550724637,
+      "grad_norm": 0.5341701745257826,
+      "learning_rate": 3.666129898013956e-05,
+      "loss": 0.5073,
+      "step": 704
+    },
+    {
+      "epoch": 1.0217391304347827,
+      "grad_norm": 0.29988900486330294,
+      "learning_rate": 3.663446054750403e-05,
+      "loss": 0.4828,
+      "step": 705
+    },
+    {
+      "epoch": 1.0231884057971015,
+      "grad_norm": 0.33163899918073203,
+      "learning_rate": 3.6607622114868495e-05,
+      "loss": 0.5149,
+      "step": 706
+    },
+    {
+      "epoch": 1.0246376811594202,
+      "grad_norm": 0.2749607995982067,
+      "learning_rate": 3.6580783682232964e-05,
+      "loss": 0.4467,
+      "step": 707
+    },
+    {
+      "epoch": 1.0260869565217392,
+      "grad_norm": 0.335134524186387,
+      "learning_rate": 3.655394524959743e-05,
+      "loss": 0.5448,
+      "step": 708
+    },
+    {
+      "epoch": 1.027536231884058,
+      "grad_norm": 0.336245663515532,
+      "learning_rate": 3.652710681696189e-05,
+      "loss": 0.5312,
+      "step": 709
+    },
+    {
+      "epoch": 1.0289855072463767,
+      "grad_norm": 0.3041401875794213,
+      "learning_rate": 3.650026838432636e-05,
+      "loss": 0.456,
+      "step": 710
+    },
+    {
+      "epoch": 1.0304347826086957,
+      "grad_norm": 0.33112756836747553,
+      "learning_rate": 3.647342995169082e-05,
+      "loss": 0.5265,
+      "step": 711
+    },
+    {
+      "epoch": 1.0318840579710145,
+      "grad_norm": 0.32377857452632586,
+      "learning_rate": 3.644659151905529e-05,
+      "loss": 0.5049,
+      "step": 712
+    },
+    {
+      "epoch": 1.0333333333333334,
+      "grad_norm": 0.35237156346129317,
+      "learning_rate": 3.6419753086419754e-05,
+      "loss": 0.509,
+      "step": 713
+    },
+    {
+      "epoch": 1.0347826086956522,
+      "grad_norm": 0.30061366226048825,
+      "learning_rate": 3.6392914653784224e-05,
+      "loss": 0.469,
+      "step": 714
+    },
+    {
+      "epoch": 1.036231884057971,
+      "grad_norm": 0.35291453847506943,
+      "learning_rate": 3.6366076221148686e-05,
+      "loss": 0.5389,
+      "step": 715
+    },
+    {
+      "epoch": 1.03768115942029,
+      "grad_norm": 0.31133681533501706,
+      "learning_rate": 3.6339237788513156e-05,
+      "loss": 0.4586,
+      "step": 716
+    },
+    {
+      "epoch": 1.0391304347826087,
+      "grad_norm": 0.38873796715326586,
+      "learning_rate": 3.631239935587762e-05,
+      "loss": 0.4938,
+      "step": 717
+    },
+    {
+      "epoch": 1.0405797101449274,
+      "grad_norm": 0.3347900167399755,
+      "learning_rate": 3.628556092324208e-05,
+      "loss": 0.4765,
+      "step": 718
+    },
+    {
+      "epoch": 1.0420289855072464,
+      "grad_norm": 16.79798030704283,
+      "learning_rate": 3.625872249060655e-05,
+      "loss": 0.6782,
+      "step": 719
+    },
+    {
+      "epoch": 1.0434782608695652,
+      "grad_norm": 0.3588584614784706,
+      "learning_rate": 3.6231884057971014e-05,
+      "loss": 0.42,
+      "step": 720
+    },
+    {
+      "epoch": 1.0449275362318842,
+      "grad_norm": 0.3683402353340812,
+      "learning_rate": 3.620504562533548e-05,
+      "loss": 0.4768,
+      "step": 721
+    },
+    {
+      "epoch": 1.046376811594203,
+      "grad_norm": 0.4050960641583972,
+      "learning_rate": 3.6178207192699946e-05,
+      "loss": 0.4417,
+      "step": 722
+    },
+    {
+      "epoch": 1.0478260869565217,
+      "grad_norm": 0.3745285639337781,
+      "learning_rate": 3.6151368760064415e-05,
+      "loss": 0.4808,
+      "step": 723
+    },
+    {
+      "epoch": 1.0492753623188407,
+      "grad_norm": 0.440811714792068,
+      "learning_rate": 3.612453032742888e-05,
+      "loss": 0.4576,
+      "step": 724
+    },
+    {
+      "epoch": 1.0507246376811594,
+      "grad_norm": 0.3644702486058993,
+      "learning_rate": 3.609769189479335e-05,
+      "loss": 0.4801,
+      "step": 725
+    },
+    {
+      "epoch": 1.0521739130434782,
+      "grad_norm": 0.32619631976812075,
+      "learning_rate": 3.607085346215781e-05,
+      "loss": 0.4738,
+      "step": 726
+    },
+    {
+      "epoch": 1.0536231884057972,
+      "grad_norm": 0.4032452494515482,
+      "learning_rate": 3.604401502952228e-05,
+      "loss": 0.4632,
+      "step": 727
+    },
+    {
+      "epoch": 1.055072463768116,
+      "grad_norm": 0.33186477576484597,
+      "learning_rate": 3.601717659688674e-05,
+      "loss": 0.473,
+      "step": 728
+    },
+    {
+      "epoch": 1.0565217391304347,
+      "grad_norm": 0.41265306224613973,
+      "learning_rate": 3.5990338164251205e-05,
+      "loss": 0.5297,
+      "step": 729
+    },
+    {
+      "epoch": 1.0579710144927537,
+      "grad_norm": 0.357811054298638,
+      "learning_rate": 3.5963499731615675e-05,
+      "loss": 0.4358,
+      "step": 730
+    },
+    {
+      "epoch": 1.0594202898550724,
+      "grad_norm": 1.6014513679266071,
+      "learning_rate": 3.593666129898014e-05,
+      "loss": 0.4755,
+      "step": 731
+    },
+    {
+      "epoch": 1.0608695652173914,
+      "grad_norm": 0.386928608403775,
+      "learning_rate": 3.590982286634461e-05,
+      "loss": 0.431,
+      "step": 732
+    },
+    {
+      "epoch": 1.0623188405797102,
+      "grad_norm": 0.3333018164985276,
+      "learning_rate": 3.588298443370907e-05,
+      "loss": 0.4761,
+      "step": 733
+    },
+    {
+      "epoch": 1.063768115942029,
+      "grad_norm": 0.3828708305599461,
+      "learning_rate": 3.585614600107354e-05,
+      "loss": 0.5168,
+      "step": 734
+    },
+    {
+      "epoch": 1.065217391304348,
+      "grad_norm": 0.3330827819992235,
+      "learning_rate": 3.5829307568438e-05,
+      "loss": 0.4949,
+      "step": 735
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 0.3413451946582709,
+      "learning_rate": 3.580246913580247e-05,
+      "loss": 0.4477,
+      "step": 736
+    },
+    {
+      "epoch": 1.0681159420289854,
+      "grad_norm": 0.32611495494902515,
+      "learning_rate": 3.5775630703166934e-05,
+      "loss": 0.5093,
+      "step": 737
+    },
+    {
+      "epoch": 1.0695652173913044,
+      "grad_norm": 0.35968381975521796,
+      "learning_rate": 3.57487922705314e-05,
+      "loss": 0.4419,
+      "step": 738
+    },
+    {
+      "epoch": 1.0710144927536231,
+      "grad_norm": 0.331806860557936,
+      "learning_rate": 3.5721953837895867e-05,
+      "loss": 0.447,
+      "step": 739
+    },
+    {
+      "epoch": 1.0724637681159421,
+      "grad_norm": 0.3202770251387416,
+      "learning_rate": 3.569511540526033e-05,
+      "loss": 0.4719,
+      "step": 740
+    },
+    {
+      "epoch": 1.0739130434782609,
+      "grad_norm": 0.37406348075552714,
+      "learning_rate": 3.56682769726248e-05,
+      "loss": 0.524,
+      "step": 741
+    },
+    {
+      "epoch": 1.0753623188405796,
+      "grad_norm": 0.3333246479182366,
+      "learning_rate": 3.564143853998926e-05,
+      "loss": 0.446,
+      "step": 742
+    },
+    {
+      "epoch": 1.0768115942028986,
+      "grad_norm": 0.34231725200893803,
+      "learning_rate": 3.561460010735374e-05,
+      "loss": 0.4321,
+      "step": 743
+    },
+    {
+      "epoch": 1.0782608695652174,
+      "grad_norm": 0.3914086039213534,
+      "learning_rate": 3.55877616747182e-05,
+      "loss": 0.5154,
+      "step": 744
+    },
+    {
+      "epoch": 1.0797101449275361,
+      "grad_norm": 0.33722219668147707,
+      "learning_rate": 3.556092324208266e-05,
+      "loss": 0.5284,
+      "step": 745
+    },
+    {
+      "epoch": 1.0811594202898551,
+      "grad_norm": 0.31575080298428554,
+      "learning_rate": 3.553408480944713e-05,
+      "loss": 0.4803,
+      "step": 746
+    },
+    {
+      "epoch": 1.0826086956521739,
+      "grad_norm": 0.31987366418089547,
+      "learning_rate": 3.5507246376811596e-05,
+      "loss": 0.4612,
+      "step": 747
+    },
+    {
+      "epoch": 1.0840579710144929,
+      "grad_norm": 0.379863200048738,
+      "learning_rate": 3.5480407944176065e-05,
+      "loss": 0.489,
+      "step": 748
+    },
+    {
+      "epoch": 1.0855072463768116,
+      "grad_norm": 0.2766306075403479,
+      "learning_rate": 3.545356951154053e-05,
+      "loss": 0.4586,
+      "step": 749
+    },
+    {
+      "epoch": 1.0869565217391304,
+      "grad_norm": 0.34487833074900065,
+      "learning_rate": 3.5426731078905e-05,
+      "loss": 0.496,
+      "step": 750
+    },
+    {
+      "epoch": 1.0884057971014494,
+      "grad_norm": 4.906831253016445,
+      "learning_rate": 3.539989264626946e-05,
+      "loss": 0.5816,
+      "step": 751
+    },
+    {
+      "epoch": 1.0898550724637681,
+      "grad_norm": 0.353885455422411,
+      "learning_rate": 3.537305421363393e-05,
+      "loss": 0.5382,
+      "step": 752
+    },
+    {
+      "epoch": 1.0913043478260869,
+      "grad_norm": 0.29502103994366735,
+      "learning_rate": 3.534621578099839e-05,
+      "loss": 0.4972,
+      "step": 753
+    },
+    {
+      "epoch": 1.0927536231884059,
+      "grad_norm": 0.35290823324954934,
+      "learning_rate": 3.531937734836286e-05,
+      "loss": 0.5024,
+      "step": 754
+    },
+    {
+      "epoch": 1.0942028985507246,
+      "grad_norm": 0.2860262112670786,
+      "learning_rate": 3.5292538915727325e-05,
+      "loss": 0.4789,
+      "step": 755
+    },
+    {
+      "epoch": 1.0956521739130434,
+      "grad_norm": 0.3877390248809208,
+      "learning_rate": 3.526570048309179e-05,
+      "loss": 0.4886,
+      "step": 756
+    },
+    {
+      "epoch": 1.0971014492753624,
+      "grad_norm": 0.28709320063852295,
+      "learning_rate": 3.523886205045626e-05,
+      "loss": 0.3936,
+      "step": 757
+    },
+    {
+      "epoch": 1.098550724637681,
+      "grad_norm": 0.34519804599228754,
+      "learning_rate": 3.521202361782072e-05,
+      "loss": 0.4594,
+      "step": 758
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.30444354880038155,
+      "learning_rate": 3.518518518518519e-05,
+      "loss": 0.4198,
+      "step": 759
+    },
+    {
+      "epoch": 1.1014492753623188,
+      "grad_norm": 0.29409450262132447,
+      "learning_rate": 3.515834675254965e-05,
+      "loss": 0.5075,
+      "step": 760
+    },
+    {
+      "epoch": 1.1028985507246376,
+      "grad_norm": 0.30054572231790977,
+      "learning_rate": 3.513150831991412e-05,
+      "loss": 0.4649,
+      "step": 761
+    },
+    {
+      "epoch": 1.1043478260869566,
+      "grad_norm": 0.2989613936621628,
+      "learning_rate": 3.5104669887278584e-05,
+      "loss": 0.4449,
+      "step": 762
+    },
+    {
+      "epoch": 1.1057971014492753,
+      "grad_norm": 0.3037114090980125,
+      "learning_rate": 3.5077831454643054e-05,
+      "loss": 0.4839,
+      "step": 763
+    },
+    {
+      "epoch": 1.107246376811594,
+      "grad_norm": 0.3031693362845878,
+      "learning_rate": 3.5050993022007516e-05,
+      "loss": 0.4157,
+      "step": 764
+    },
+    {
+      "epoch": 1.108695652173913,
+      "grad_norm": 0.3137122464324766,
+      "learning_rate": 3.502415458937198e-05,
+      "loss": 0.4999,
+      "step": 765
+    },
+    {
+      "epoch": 1.1101449275362318,
+      "grad_norm": 0.35389682558480495,
+      "learning_rate": 3.499731615673645e-05,
+      "loss": 0.5107,
+      "step": 766
+    },
+    {
+      "epoch": 1.1115942028985508,
+      "grad_norm": 0.2895422259138419,
+      "learning_rate": 3.497047772410091e-05,
+      "loss": 0.4807,
+      "step": 767
+    },
+    {
+      "epoch": 1.1130434782608696,
+      "grad_norm": 0.35009638294139317,
+      "learning_rate": 3.494363929146538e-05,
+      "loss": 0.4727,
+      "step": 768
+    },
+    {
+      "epoch": 1.1144927536231883,
+      "grad_norm": 0.3708206656399047,
+      "learning_rate": 3.4916800858829844e-05,
+      "loss": 0.5179,
+      "step": 769
+    },
+    {
+      "epoch": 1.1159420289855073,
+      "grad_norm": 0.28903378564992177,
+      "learning_rate": 3.488996242619431e-05,
+      "loss": 0.529,
+      "step": 770
+    },
+    {
+      "epoch": 1.117391304347826,
+      "grad_norm": 0.39287961390615433,
+      "learning_rate": 3.4863123993558776e-05,
+      "loss": 0.468,
+      "step": 771
+    },
+    {
+      "epoch": 1.1188405797101448,
+      "grad_norm": 0.3796057467781331,
+      "learning_rate": 3.4836285560923245e-05,
+      "loss": 0.5136,
+      "step": 772
+    },
+    {
+      "epoch": 1.1202898550724638,
+      "grad_norm": 0.34554797079886895,
+      "learning_rate": 3.480944712828771e-05,
+      "loss": 0.5179,
+      "step": 773
+    },
+    {
+      "epoch": 1.1217391304347826,
+      "grad_norm": 0.3964474111821127,
+      "learning_rate": 3.478260869565218e-05,
+      "loss": 0.421,
+      "step": 774
+    },
+    {
+      "epoch": 1.1231884057971016,
+      "grad_norm": 0.42365345484348044,
+      "learning_rate": 3.475577026301664e-05,
+      "loss": 0.4967,
+      "step": 775
+    },
+    {
+      "epoch": 1.1246376811594203,
+      "grad_norm": 0.31444964103858697,
+      "learning_rate": 3.47289318303811e-05,
+      "loss": 0.5367,
+      "step": 776
+    },
+    {
+      "epoch": 1.126086956521739,
+      "grad_norm": 0.40977846149532443,
+      "learning_rate": 3.470209339774557e-05,
+      "loss": 0.5085,
+      "step": 777
+    },
+    {
+      "epoch": 1.127536231884058,
+      "grad_norm": 0.4040084864158796,
+      "learning_rate": 3.4675254965110035e-05,
+      "loss": 0.5166,
+      "step": 778
+    },
+    {
+      "epoch": 1.1289855072463768,
+      "grad_norm": 0.33605865897707826,
+      "learning_rate": 3.4648416532474505e-05,
+      "loss": 0.5246,
+      "step": 779
+    },
+    {
+      "epoch": 1.1304347826086956,
+      "grad_norm": 0.3220052982912894,
+      "learning_rate": 3.462157809983897e-05,
+      "loss": 0.4491,
+      "step": 780
+    },
+    {
+      "epoch": 1.1318840579710145,
+      "grad_norm": 0.4366726878890496,
+      "learning_rate": 3.459473966720344e-05,
+      "loss": 0.523,
+      "step": 781
+    },
+    {
+      "epoch": 1.1333333333333333,
+      "grad_norm": 0.26621474023466923,
+      "learning_rate": 3.45679012345679e-05,
+      "loss": 0.4173,
+      "step": 782
+    },
+    {
+      "epoch": 1.134782608695652,
+      "grad_norm": 0.39572440477831416,
+      "learning_rate": 3.454106280193237e-05,
+      "loss": 0.5299,
+      "step": 783
+    },
+    {
+      "epoch": 1.136231884057971,
+      "grad_norm": 0.4040603491168308,
+      "learning_rate": 3.451422436929683e-05,
+      "loss": 0.4567,
+      "step": 784
+    },
+    {
+      "epoch": 1.1376811594202898,
+      "grad_norm": 0.32904119897841605,
+      "learning_rate": 3.44873859366613e-05,
+      "loss": 0.5497,
+      "step": 785
+    },
+    {
+      "epoch": 1.1391304347826088,
+      "grad_norm": 0.46278497219882125,
+      "learning_rate": 3.4460547504025764e-05,
+      "loss": 0.5304,
+      "step": 786
+    },
+    {
+      "epoch": 1.1405797101449275,
+      "grad_norm": 3.964868351721275,
+      "learning_rate": 3.443370907139023e-05,
+      "loss": 0.6874,
+      "step": 787
+    },
+    {
+      "epoch": 1.1420289855072463,
+      "grad_norm": 0.4333381760274589,
+      "learning_rate": 3.44068706387547e-05,
+      "loss": 0.5262,
+      "step": 788
+    },
+    {
+      "epoch": 1.1434782608695653,
+      "grad_norm": 0.40342770956293456,
+      "learning_rate": 3.438003220611916e-05,
+      "loss": 0.476,
+      "step": 789
+    },
+    {
+      "epoch": 1.144927536231884,
+      "grad_norm": 0.450118522092737,
+      "learning_rate": 3.4353193773483636e-05,
+      "loss": 0.5051,
+      "step": 790
+    },
+    {
+      "epoch": 1.146376811594203,
+      "grad_norm": 0.38593792951653244,
+      "learning_rate": 3.43263553408481e-05,
+      "loss": 0.5204,
+      "step": 791
+    },
+    {
+      "epoch": 1.1478260869565218,
+      "grad_norm": 0.3198244248233121,
+      "learning_rate": 3.429951690821256e-05,
+      "loss": 0.4774,
+      "step": 792
+    },
+    {
+      "epoch": 1.1492753623188405,
+      "grad_norm": 0.37117111558693916,
+      "learning_rate": 3.427267847557703e-05,
+      "loss": 0.4587,
+      "step": 793
+    },
+    {
+      "epoch": 1.1507246376811595,
+      "grad_norm": 0.3304999838282135,
+      "learning_rate": 3.4245840042941493e-05,
+      "loss": 0.5017,
+      "step": 794
+    },
+    {
+      "epoch": 1.1521739130434783,
+      "grad_norm": 0.32790729982023115,
+      "learning_rate": 3.421900161030596e-05,
+      "loss": 0.4553,
+      "step": 795
+    },
+    {
+      "epoch": 1.153623188405797,
+      "grad_norm": 0.3058444493138007,
+      "learning_rate": 3.4192163177670426e-05,
+      "loss": 0.4992,
+      "step": 796
+    },
+    {
+      "epoch": 1.155072463768116,
+      "grad_norm": 0.3137149776147036,
+      "learning_rate": 3.4165324745034895e-05,
+      "loss": 0.4673,
+      "step": 797
+    },
+    {
+      "epoch": 1.1565217391304348,
+      "grad_norm": 0.3055969447342787,
+      "learning_rate": 3.413848631239936e-05,
+      "loss": 0.4489,
+      "step": 798
+    },
+    {
+      "epoch": 1.1579710144927535,
+      "grad_norm": 0.6552379281280202,
+      "learning_rate": 3.411164787976383e-05,
+      "loss": 0.4605,
+      "step": 799
+    },
+    {
+      "epoch": 1.1594202898550725,
+      "grad_norm": 0.35827545936038846,
+      "learning_rate": 3.408480944712829e-05,
+      "loss": 0.5115,
+      "step": 800
+    },
+    {
+      "epoch": 1.1608695652173913,
+      "grad_norm": 0.3096119862536391,
+      "learning_rate": 3.405797101449276e-05,
+      "loss": 0.5188,
+      "step": 801
+    },
+    {
+      "epoch": 1.1623188405797102,
+      "grad_norm": 0.3746648851546505,
+      "learning_rate": 3.403113258185722e-05,
+      "loss": 0.4364,
+      "step": 802
+    },
+    {
+      "epoch": 1.163768115942029,
+      "grad_norm": 0.37513208247393715,
+      "learning_rate": 3.4004294149221685e-05,
+      "loss": 0.4751,
+      "step": 803
+    },
+    {
+      "epoch": 1.1652173913043478,
+      "grad_norm": 0.4093812220396632,
+      "learning_rate": 3.3977455716586155e-05,
+      "loss": 0.4712,
+      "step": 804
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 0.2968811522800294,
+      "learning_rate": 3.395061728395062e-05,
+      "loss": 0.4673,
+      "step": 805
+    },
+    {
+      "epoch": 1.1681159420289855,
+      "grad_norm": 0.37760080915918015,
+      "learning_rate": 3.392377885131509e-05,
+      "loss": 0.4461,
+      "step": 806
+    },
+    {
+      "epoch": 1.1695652173913043,
+      "grad_norm": 2.3847172796970564,
+      "learning_rate": 3.389694041867955e-05,
+      "loss": 0.531,
+      "step": 807
+    },
+    {
+      "epoch": 1.1710144927536232,
+      "grad_norm": 0.36180593225677865,
+      "learning_rate": 3.387010198604402e-05,
+      "loss": 0.4838,
+      "step": 808
+    },
+    {
+      "epoch": 1.172463768115942,
+      "grad_norm": 1.5785551257268962,
+      "learning_rate": 3.384326355340848e-05,
+      "loss": 0.51,
+      "step": 809
+    },
+    {
+      "epoch": 1.1739130434782608,
+      "grad_norm": 0.48512277259322567,
+      "learning_rate": 3.381642512077295e-05,
+      "loss": 0.5021,
+      "step": 810
+    },
+    {
+      "epoch": 1.1753623188405797,
+      "grad_norm": 0.3689744132091387,
+      "learning_rate": 3.3789586688137414e-05,
+      "loss": 0.496,
+      "step": 811
+    },
+    {
+      "epoch": 1.1768115942028985,
+      "grad_norm": 0.5468114360082217,
+      "learning_rate": 3.3762748255501884e-05,
+      "loss": 0.5209,
+      "step": 812
+    },
+    {
+      "epoch": 1.1782608695652175,
+      "grad_norm": 0.3789941620658084,
+      "learning_rate": 3.3735909822866346e-05,
+      "loss": 0.4426,
+      "step": 813
+    },
+    {
+      "epoch": 1.1797101449275362,
+      "grad_norm": 0.4720650796187916,
+      "learning_rate": 3.370907139023081e-05,
+      "loss": 0.4939,
+      "step": 814
+    },
+    {
+      "epoch": 1.181159420289855,
+      "grad_norm": 0.4519968405566694,
+      "learning_rate": 3.368223295759528e-05,
+      "loss": 0.4975,
+      "step": 815
+    },
+    {
+      "epoch": 1.182608695652174,
+      "grad_norm": 0.5104332648013641,
+      "learning_rate": 3.365539452495974e-05,
+      "loss": 0.491,
+      "step": 816
+    },
+    {
+      "epoch": 1.1840579710144927,
+      "grad_norm": 0.49738354731794426,
+      "learning_rate": 3.362855609232421e-05,
+      "loss": 0.4647,
+      "step": 817
+    },
+    {
+      "epoch": 1.1855072463768117,
+      "grad_norm": 0.3769921876268249,
+      "learning_rate": 3.3601717659688674e-05,
+      "loss": 0.4579,
+      "step": 818
+    },
+    {
+      "epoch": 1.1869565217391305,
+      "grad_norm": 0.4663245979257391,
+      "learning_rate": 3.357487922705314e-05,
+      "loss": 0.5632,
+      "step": 819
+    },
+    {
+      "epoch": 1.1884057971014492,
+      "grad_norm": 0.37499049303519677,
+      "learning_rate": 3.3548040794417606e-05,
+      "loss": 0.549,
+      "step": 820
+    },
+    {
+      "epoch": 1.1898550724637682,
+      "grad_norm": 0.4473411807255098,
+      "learning_rate": 3.3521202361782075e-05,
+      "loss": 0.4996,
+      "step": 821
+    },
+    {
+      "epoch": 1.191304347826087,
+      "grad_norm": 0.3830440398346099,
+      "learning_rate": 3.349436392914654e-05,
+      "loss": 0.4959,
+      "step": 822
+    },
+    {
+      "epoch": 1.1927536231884057,
+      "grad_norm": 0.4104640920005607,
+      "learning_rate": 3.3467525496511e-05,
+      "loss": 0.4994,
+      "step": 823
+    },
+    {
+      "epoch": 1.1942028985507247,
+      "grad_norm": 0.40985809678787355,
+      "learning_rate": 3.344068706387547e-05,
+      "loss": 0.5009,
+      "step": 824
+    },
+    {
+      "epoch": 1.1956521739130435,
+      "grad_norm": 0.3289639690915926,
+      "learning_rate": 3.341384863123993e-05,
+      "loss": 0.4977,
+      "step": 825
+    },
+    {
+      "epoch": 1.1971014492753622,
+      "grad_norm": 0.3527697395053973,
+      "learning_rate": 3.33870101986044e-05,
+      "loss": 0.5708,
+      "step": 826
+    },
+    {
+      "epoch": 1.1985507246376812,
+      "grad_norm": 0.43443492772244724,
+      "learning_rate": 3.3360171765968865e-05,
+      "loss": 0.4473,
+      "step": 827
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.337378493207393,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.5505,
+      "step": 828
+    },
+    {
+      "epoch": 1.201449275362319,
+      "grad_norm": 0.3865226415547385,
+      "learning_rate": 3.33064949006978e-05,
+      "loss": 0.5119,
+      "step": 829
+    },
+    {
+      "epoch": 1.2028985507246377,
+      "grad_norm": 0.34911981266613895,
+      "learning_rate": 3.327965646806227e-05,
+      "loss": 0.4573,
+      "step": 830
+    },
+    {
+      "epoch": 1.2043478260869565,
+      "grad_norm": 0.34586488602301124,
+      "learning_rate": 3.325281803542673e-05,
+      "loss": 0.51,
+      "step": 831
+    },
+    {
+      "epoch": 1.2057971014492754,
+      "grad_norm": 0.42107082760580405,
+      "learning_rate": 3.32259796027912e-05,
+      "loss": 0.4662,
+      "step": 832
+    },
+    {
+      "epoch": 1.2072463768115942,
+      "grad_norm": 0.43549995035939054,
+      "learning_rate": 3.319914117015566e-05,
+      "loss": 0.5384,
+      "step": 833
+    },
+    {
+      "epoch": 1.208695652173913,
+      "grad_norm": 0.3862371394781864,
+      "learning_rate": 3.3172302737520125e-05,
+      "loss": 0.5404,
+      "step": 834
+    },
+    {
+      "epoch": 1.210144927536232,
+      "grad_norm": 0.34958162952048627,
+      "learning_rate": 3.3145464304884594e-05,
+      "loss": 0.4774,
+      "step": 835
+    },
+    {
+      "epoch": 1.2115942028985507,
+      "grad_norm": 0.3821812720253012,
+      "learning_rate": 3.3118625872249064e-05,
+      "loss": 0.441,
+      "step": 836
+    },
+    {
+      "epoch": 1.2130434782608694,
+      "grad_norm": 0.4000142422819388,
+      "learning_rate": 3.3091787439613533e-05,
+      "loss": 0.4964,
+      "step": 837
+    },
+    {
+      "epoch": 1.2144927536231884,
+      "grad_norm": 0.406579430351433,
+      "learning_rate": 3.3064949006977996e-05,
+      "loss": 0.4323,
+      "step": 838
+    },
+    {
+      "epoch": 1.2159420289855072,
+      "grad_norm": 0.3732779725444165,
+      "learning_rate": 3.3038110574342466e-05,
+      "loss": 0.4795,
+      "step": 839
+    },
+    {
+      "epoch": 1.2173913043478262,
+      "grad_norm": 0.3856736552896055,
+      "learning_rate": 3.301127214170693e-05,
+      "loss": 0.4587,
+      "step": 840
+    },
+    {
+      "epoch": 1.218840579710145,
+      "grad_norm": 0.9264139639240327,
+      "learning_rate": 3.298443370907139e-05,
+      "loss": 0.5313,
+      "step": 841
+    },
+    {
+      "epoch": 1.2202898550724637,
+      "grad_norm": 0.41170275194820466,
+      "learning_rate": 3.295759527643586e-05,
+      "loss": 0.4814,
+      "step": 842
+    },
+    {
+      "epoch": 1.2217391304347827,
+      "grad_norm": 0.3640744642554216,
+      "learning_rate": 3.2930756843800323e-05,
+      "loss": 0.5029,
+      "step": 843
+    },
+    {
+      "epoch": 1.2231884057971014,
+      "grad_norm": 0.38012885622626436,
+      "learning_rate": 3.290391841116479e-05,
+      "loss": 0.4661,
+      "step": 844
+    },
+    {
+      "epoch": 1.2246376811594204,
+      "grad_norm": 0.3239294935147369,
+      "learning_rate": 3.2877079978529256e-05,
+      "loss": 0.435,
+      "step": 845
+    },
+    {
+      "epoch": 1.2260869565217392,
+      "grad_norm": 0.34770674570393884,
+      "learning_rate": 3.2850241545893725e-05,
+      "loss": 0.4938,
+      "step": 846
+    },
+    {
+      "epoch": 1.227536231884058,
+      "grad_norm": 0.3313586231488532,
+      "learning_rate": 3.282340311325819e-05,
+      "loss": 0.4983,
+      "step": 847
+    },
+    {
+      "epoch": 1.228985507246377,
+      "grad_norm": 0.3752653873420586,
+      "learning_rate": 3.279656468062266e-05,
+      "loss": 0.4987,
+      "step": 848
+    },
+    {
+      "epoch": 1.2304347826086957,
+      "grad_norm": 0.3241261208367448,
+      "learning_rate": 3.276972624798712e-05,
+      "loss": 0.4745,
+      "step": 849
+    },
+    {
+      "epoch": 1.2318840579710144,
+      "grad_norm": 0.3393488135230325,
+      "learning_rate": 3.274288781535158e-05,
+      "loss": 0.433,
+      "step": 850
+    },
+    {
+      "epoch": 1.2333333333333334,
+      "grad_norm": 0.3503274304144233,
+      "learning_rate": 3.271604938271605e-05,
+      "loss": 0.4736,
+      "step": 851
+    },
+    {
+      "epoch": 1.2347826086956522,
+      "grad_norm": 0.3418505543344525,
+      "learning_rate": 3.2689210950080515e-05,
+      "loss": 0.471,
+      "step": 852
+    },
+    {
+      "epoch": 1.236231884057971,
+      "grad_norm": 0.34202854605765615,
+      "learning_rate": 3.2662372517444985e-05,
+      "loss": 0.5153,
+      "step": 853
+    },
+    {
+      "epoch": 1.23768115942029,
+      "grad_norm": 0.3096582468616831,
+      "learning_rate": 3.263553408480945e-05,
+      "loss": 0.4471,
+      "step": 854
+    },
+    {
+      "epoch": 1.2391304347826086,
+      "grad_norm": 0.36906290878465053,
+      "learning_rate": 3.260869565217392e-05,
+      "loss": 0.5084,
+      "step": 855
+    },
+    {
+      "epoch": 1.2405797101449276,
+      "grad_norm": 0.3236250876491843,
+      "learning_rate": 3.258185721953838e-05,
+      "loss": 0.4589,
+      "step": 856
+    },
+    {
+      "epoch": 1.2420289855072464,
+      "grad_norm": 0.3834374189338398,
+      "learning_rate": 3.255501878690285e-05,
+      "loss": 0.5156,
+      "step": 857
+    },
+    {
+      "epoch": 1.2434782608695651,
+      "grad_norm": 0.3019495775859321,
+      "learning_rate": 3.252818035426731e-05,
+      "loss": 0.4715,
+      "step": 858
+    },
+    {
+      "epoch": 1.2449275362318841,
+      "grad_norm": 0.37003140617652347,
+      "learning_rate": 3.250134192163178e-05,
+      "loss": 0.4597,
+      "step": 859
+    },
+    {
+      "epoch": 1.2463768115942029,
+      "grad_norm": 0.32403513297994646,
+      "learning_rate": 3.2474503488996244e-05,
+      "loss": 0.5453,
+      "step": 860
+    },
+    {
+      "epoch": 1.2478260869565219,
+      "grad_norm": 0.31644292072769686,
+      "learning_rate": 3.244766505636071e-05,
+      "loss": 0.485,
+      "step": 861
+    },
+    {
+      "epoch": 1.2492753623188406,
+      "grad_norm": 0.3193997232817535,
+      "learning_rate": 3.2420826623725176e-05,
+      "loss": 0.4498,
+      "step": 862
+    },
+    {
+      "epoch": 1.2507246376811594,
+      "grad_norm": 0.3288694911299711,
+      "learning_rate": 3.239398819108964e-05,
+      "loss": 0.4515,
+      "step": 863
+    },
+    {
+      "epoch": 1.2521739130434781,
+      "grad_norm": 0.3722851249796339,
+      "learning_rate": 3.236714975845411e-05,
+      "loss": 0.4434,
+      "step": 864
+    },
+    {
+      "epoch": 1.2536231884057971,
+      "grad_norm": 0.35640927389422644,
+      "learning_rate": 3.234031132581857e-05,
+      "loss": 0.4788,
+      "step": 865
+    },
+    {
+      "epoch": 1.2550724637681159,
+      "grad_norm": 0.34416013778138244,
+      "learning_rate": 3.231347289318304e-05,
+      "loss": 0.4682,
+      "step": 866
+    },
+    {
+      "epoch": 1.2565217391304349,
+      "grad_norm": 0.35307468327505803,
+      "learning_rate": 3.2286634460547504e-05,
+      "loss": 0.5182,
+      "step": 867
+    },
+    {
+      "epoch": 1.2579710144927536,
+      "grad_norm": 0.3427936110711446,
+      "learning_rate": 3.225979602791197e-05,
+      "loss": 0.4552,
+      "step": 868
+    },
+    {
+      "epoch": 1.2594202898550724,
+      "grad_norm": 0.2763448846412855,
+      "learning_rate": 3.2232957595276436e-05,
+      "loss": 0.4304,
+      "step": 869
+    },
+    {
+      "epoch": 1.2608695652173914,
+      "grad_norm": 0.379596405325515,
+      "learning_rate": 3.22061191626409e-05,
+      "loss": 0.4684,
+      "step": 870
+    },
+    {
+      "epoch": 1.26231884057971,
+      "grad_norm": 0.3940790425475433,
+      "learning_rate": 3.217928073000537e-05,
+      "loss": 0.5502,
+      "step": 871
+    },
+    {
+      "epoch": 1.263768115942029,
+      "grad_norm": 0.36709229703125856,
+      "learning_rate": 3.215244229736983e-05,
+      "loss": 0.4895,
+      "step": 872
+    },
+    {
+      "epoch": 1.2652173913043478,
+      "grad_norm": 0.3765475900776119,
+      "learning_rate": 3.21256038647343e-05,
+      "loss": 0.5219,
+      "step": 873
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 0.3015095446447829,
+      "learning_rate": 3.209876543209876e-05,
+      "loss": 0.4674,
+      "step": 874
+    },
+    {
+      "epoch": 1.2681159420289856,
+      "grad_norm": 0.3754372901666103,
+      "learning_rate": 3.207192699946323e-05,
+      "loss": 0.4681,
+      "step": 875
+    },
+    {
+      "epoch": 1.2695652173913043,
+      "grad_norm": 0.28275837312472013,
+      "learning_rate": 3.2045088566827695e-05,
+      "loss": 0.4487,
+      "step": 876
+    },
+    {
+      "epoch": 1.2710144927536233,
+      "grad_norm": 0.3935736907949941,
+      "learning_rate": 3.2018250134192165e-05,
+      "loss": 0.5029,
+      "step": 877
+    },
+    {
+      "epoch": 1.272463768115942,
+      "grad_norm": 0.39538277413686385,
+      "learning_rate": 3.199141170155663e-05,
+      "loss": 0.4619,
+      "step": 878
+    },
+    {
+      "epoch": 1.2739130434782608,
+      "grad_norm": 0.3270500133449961,
+      "learning_rate": 3.19645732689211e-05,
+      "loss": 0.521,
+      "step": 879
+    },
+    {
+      "epoch": 1.2753623188405796,
+      "grad_norm": 0.35142103803291763,
+      "learning_rate": 3.193773483628556e-05,
+      "loss": 0.4879,
+      "step": 880
+    },
+    {
+      "epoch": 1.2768115942028986,
+      "grad_norm": 0.3393109748778095,
+      "learning_rate": 3.191089640365002e-05,
+      "loss": 0.491,
+      "step": 881
+    },
+    {
+      "epoch": 1.2782608695652173,
+      "grad_norm": 0.4071844260390824,
+      "learning_rate": 3.188405797101449e-05,
+      "loss": 0.4283,
+      "step": 882
+    },
+    {
+      "epoch": 1.2797101449275363,
+      "grad_norm": 0.3601811827305163,
+      "learning_rate": 3.185721953837896e-05,
+      "loss": 0.4632,
+      "step": 883
+    },
+    {
+      "epoch": 1.281159420289855,
+      "grad_norm": 0.34393463950589037,
+      "learning_rate": 3.183038110574343e-05,
+      "loss": 0.5411,
+      "step": 884
+    },
+    {
+      "epoch": 1.2826086956521738,
+      "grad_norm": 0.3019143069007389,
+      "learning_rate": 3.1803542673107894e-05,
+      "loss": 0.4914,
+      "step": 885
+    },
+    {
+      "epoch": 1.2840579710144928,
+      "grad_norm": 0.31666199554410895,
+      "learning_rate": 3.1776704240472364e-05,
+      "loss": 0.407,
+      "step": 886
+    },
+    {
+      "epoch": 1.2855072463768116,
+      "grad_norm": 0.32452996393481603,
+      "learning_rate": 3.1749865807836826e-05,
+      "loss": 0.5102,
+      "step": 887
+    },
+    {
+      "epoch": 1.2869565217391306,
+      "grad_norm": 0.3820308585272605,
+      "learning_rate": 3.172302737520129e-05,
+      "loss": 0.4429,
+      "step": 888
+    },
+    {
+      "epoch": 1.2884057971014493,
+      "grad_norm": 0.5195137701326371,
+      "learning_rate": 3.169618894256576e-05,
+      "loss": 0.4732,
+      "step": 889
+    },
+    {
+      "epoch": 1.289855072463768,
+      "grad_norm": 0.2936225835669792,
+      "learning_rate": 3.166935050993022e-05,
+      "loss": 0.4657,
+      "step": 890
+    },
+    {
+      "epoch": 1.2913043478260868,
+      "grad_norm": 0.42350866865971254,
+      "learning_rate": 3.164251207729469e-05,
+      "loss": 0.486,
+      "step": 891
+    },
+    {
+      "epoch": 1.2927536231884058,
+      "grad_norm": 0.3370830587110153,
+      "learning_rate": 3.1615673644659153e-05,
+      "loss": 0.4732,
+      "step": 892
+    },
+    {
+      "epoch": 1.2942028985507246,
+      "grad_norm": 0.34306276198036184,
+      "learning_rate": 3.158883521202362e-05,
+      "loss": 0.5006,
+      "step": 893
+    },
+    {
+      "epoch": 1.2956521739130435,
+      "grad_norm": 0.4307823243834926,
+      "learning_rate": 3.1561996779388086e-05,
+      "loss": 0.5748,
+      "step": 894
+    },
+    {
+      "epoch": 1.2971014492753623,
+      "grad_norm": 0.3421827711913848,
+      "learning_rate": 3.1535158346752555e-05,
+      "loss": 0.5452,
+      "step": 895
+    },
+    {
+      "epoch": 1.298550724637681,
+      "grad_norm": 0.35986972590664257,
+      "learning_rate": 3.150831991411702e-05,
+      "loss": 0.4928,
+      "step": 896
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.4028315305194365,
+      "learning_rate": 3.148148148148148e-05,
+      "loss": 0.4893,
+      "step": 897
+    },
+    {
+      "epoch": 1.3014492753623188,
+      "grad_norm": 0.3198518521418353,
+      "learning_rate": 3.145464304884595e-05,
+      "loss": 0.4719,
+      "step": 898
+    },
+    {
+      "epoch": 1.3028985507246378,
+      "grad_norm": 0.3048101730507823,
+      "learning_rate": 3.142780461621041e-05,
+      "loss": 0.497,
+      "step": 899
+    },
+    {
+      "epoch": 1.3043478260869565,
+      "grad_norm": 0.4497587869956628,
+      "learning_rate": 3.140096618357488e-05,
+      "loss": 0.5056,
+      "step": 900
+    },
+    {
+      "epoch": 1.3057971014492753,
+      "grad_norm": 0.3584956584198498,
+      "learning_rate": 3.1374127750939345e-05,
+      "loss": 0.5536,
+      "step": 901
+    },
+    {
+      "epoch": 1.3072463768115943,
+      "grad_norm": 0.3461500766700808,
+      "learning_rate": 3.1347289318303815e-05,
+      "loss": 0.4703,
+      "step": 902
+    },
+    {
+      "epoch": 1.308695652173913,
+      "grad_norm": 0.8694177600711719,
+      "learning_rate": 3.132045088566828e-05,
+      "loss": 0.4411,
+      "step": 903
+    },
+    {
+      "epoch": 1.310144927536232,
+      "grad_norm": 0.42333283493724055,
+      "learning_rate": 3.129361245303275e-05,
+      "loss": 0.4677,
+      "step": 904
+    },
+    {
+      "epoch": 1.3115942028985508,
+      "grad_norm": 0.41077456941737023,
+      "learning_rate": 3.126677402039721e-05,
+      "loss": 0.4789,
+      "step": 905
+    },
+    {
+      "epoch": 1.3130434782608695,
+      "grad_norm": 0.49701901911869834,
+      "learning_rate": 3.123993558776168e-05,
+      "loss": 0.4627,
+      "step": 906
+    },
+    {
+      "epoch": 1.3144927536231883,
+      "grad_norm": 0.3737725318170062,
+      "learning_rate": 3.121309715512614e-05,
+      "loss": 0.5316,
+      "step": 907
+    },
+    {
+      "epoch": 1.3159420289855073,
+      "grad_norm": 0.37226289942626994,
+      "learning_rate": 3.1186258722490605e-05,
+      "loss": 0.5297,
+      "step": 908
+    },
+    {
+      "epoch": 1.317391304347826,
+      "grad_norm": 0.4429119113697492,
+      "learning_rate": 3.1159420289855074e-05,
+      "loss": 0.5067,
+      "step": 909
+    },
+    {
+      "epoch": 1.318840579710145,
+      "grad_norm": 0.39281951947534194,
+      "learning_rate": 3.113258185721954e-05,
+      "loss": 0.5163,
+      "step": 910
+    },
+    {
+      "epoch": 1.3202898550724638,
+      "grad_norm": 0.36347806131070287,
+      "learning_rate": 3.1105743424584007e-05,
+      "loss": 0.5138,
+      "step": 911
+    },
+    {
+      "epoch": 1.3217391304347825,
+      "grad_norm": 0.37090038432978484,
+      "learning_rate": 3.107890499194847e-05,
+      "loss": 0.5067,
+      "step": 912
+    },
+    {
+      "epoch": 1.3231884057971015,
+      "grad_norm": 0.3399730639972717,
+      "learning_rate": 3.105206655931294e-05,
+      "loss": 0.4666,
+      "step": 913
+    },
+    {
+      "epoch": 1.3246376811594203,
+      "grad_norm": 0.3615568168196394,
+      "learning_rate": 3.10252281266774e-05,
+      "loss": 0.5118,
+      "step": 914
+    },
+    {
+      "epoch": 1.3260869565217392,
+      "grad_norm": 0.3614274613790376,
+      "learning_rate": 3.099838969404187e-05,
+      "loss": 0.5104,
+      "step": 915
+    },
+    {
+      "epoch": 1.327536231884058,
+      "grad_norm": 0.3147554804974595,
+      "learning_rate": 3.0971551261406334e-05,
+      "loss": 0.4469,
+      "step": 916
+    },
+    {
+      "epoch": 1.3289855072463768,
+      "grad_norm": 0.4522346661612457,
+      "learning_rate": 3.09447128287708e-05,
+      "loss": 0.4577,
+      "step": 917
+    },
+    {
+      "epoch": 1.3304347826086955,
+      "grad_norm": 0.32263332874891165,
+      "learning_rate": 3.0917874396135266e-05,
+      "loss": 0.4621,
+      "step": 918
+    },
+    {
+      "epoch": 1.3318840579710145,
+      "grad_norm": 0.2883562120924791,
+      "learning_rate": 3.089103596349973e-05,
+      "loss": 0.5105,
+      "step": 919
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.3179678130564038,
+      "learning_rate": 3.08641975308642e-05,
+      "loss": 0.481,
+      "step": 920
+    },
+    {
+      "epoch": 1.3347826086956522,
+      "grad_norm": 0.28861961579910617,
+      "learning_rate": 3.083735909822866e-05,
+      "loss": 0.4943,
+      "step": 921
+    },
+    {
+      "epoch": 1.336231884057971,
+      "grad_norm": 0.3363186614291933,
+      "learning_rate": 3.081052066559313e-05,
+      "loss": 0.4577,
+      "step": 922
+    },
+    {
+      "epoch": 1.3376811594202898,
+      "grad_norm": 0.3053941389822725,
+      "learning_rate": 3.078368223295759e-05,
+      "loss": 0.4812,
+      "step": 923
+    },
+    {
+      "epoch": 1.3391304347826087,
+      "grad_norm": 0.34395797623076607,
+      "learning_rate": 3.075684380032206e-05,
+      "loss": 0.4901,
+      "step": 924
+    },
+    {
+      "epoch": 1.3405797101449275,
+      "grad_norm": 0.3574862058583271,
+      "learning_rate": 3.0730005367686526e-05,
+      "loss": 0.5004,
+      "step": 925
+    },
+    {
+      "epoch": 1.3420289855072465,
+      "grad_norm": 0.36766498861515995,
+      "learning_rate": 3.0703166935050995e-05,
+      "loss": 0.4846,
+      "step": 926
+    },
+    {
+      "epoch": 1.3434782608695652,
+      "grad_norm": 0.3064419453723753,
+      "learning_rate": 3.067632850241546e-05,
+      "loss": 0.4842,
+      "step": 927
+    },
+    {
+      "epoch": 1.344927536231884,
+      "grad_norm": 0.3364090622983266,
+      "learning_rate": 3.064949006977992e-05,
+      "loss": 0.5311,
+      "step": 928
+    },
+    {
+      "epoch": 1.346376811594203,
+      "grad_norm": 0.3274279374759603,
+      "learning_rate": 3.06226516371444e-05,
+      "loss": 0.4851,
+      "step": 929
+    },
+    {
+      "epoch": 1.3478260869565217,
+      "grad_norm": 0.2826344391252545,
+      "learning_rate": 3.059581320450886e-05,
+      "loss": 0.4574,
+      "step": 930
+    },
+    {
+      "epoch": 1.3492753623188407,
+      "grad_norm": 0.29856410659478516,
+      "learning_rate": 3.056897477187333e-05,
+      "loss": 0.4674,
+      "step": 931
+    },
+    {
+      "epoch": 1.3507246376811595,
+      "grad_norm": 0.3031177440918603,
+      "learning_rate": 3.054213633923779e-05,
+      "loss": 0.5107,
+      "step": 932
+    },
+    {
+      "epoch": 1.3521739130434782,
+      "grad_norm": 0.3190321958011399,
+      "learning_rate": 3.0515297906602258e-05,
+      "loss": 0.5476,
+      "step": 933
+    },
+    {
+      "epoch": 1.353623188405797,
+      "grad_norm": 0.28726674696872556,
+      "learning_rate": 3.0488459473966724e-05,
+      "loss": 0.4484,
+      "step": 934
+    },
+    {
+      "epoch": 1.355072463768116,
+      "grad_norm": 0.34401661423155566,
+      "learning_rate": 3.046162104133119e-05,
+      "loss": 0.512,
+      "step": 935
+    },
+    {
+      "epoch": 1.3565217391304347,
+      "grad_norm": 0.3264713371237452,
+      "learning_rate": 3.0434782608695656e-05,
+      "loss": 0.4517,
+      "step": 936
+    },
+    {
+      "epoch": 1.3579710144927537,
+      "grad_norm": 0.30036423040371935,
+      "learning_rate": 3.0407944176060122e-05,
+      "loss": 0.4729,
+      "step": 937
+    },
+    {
+      "epoch": 1.3594202898550725,
+      "grad_norm": 0.2836159289751039,
+      "learning_rate": 3.038110574342459e-05,
+      "loss": 0.3927,
+      "step": 938
+    },
+    {
+      "epoch": 1.3608695652173912,
+      "grad_norm": 0.35656364715244465,
+      "learning_rate": 3.035426731078905e-05,
+      "loss": 0.4858,
+      "step": 939
+    },
+    {
+      "epoch": 1.3623188405797102,
+      "grad_norm": 0.2611982183438624,
+      "learning_rate": 3.0327428878153517e-05,
+      "loss": 0.4202,
+      "step": 940
+    },
+    {
+      "epoch": 1.363768115942029,
+      "grad_norm": 0.297442808481956,
+      "learning_rate": 3.0300590445517984e-05,
+      "loss": 0.5323,
+      "step": 941
+    },
+    {
+      "epoch": 1.365217391304348,
+      "grad_norm": 0.30599446236692146,
+      "learning_rate": 3.027375201288245e-05,
+      "loss": 0.4312,
+      "step": 942
+    },
+    {
+      "epoch": 1.3666666666666667,
+      "grad_norm": 0.37120178748535404,
+      "learning_rate": 3.0246913580246916e-05,
+      "loss": 0.5141,
+      "step": 943
+    },
+    {
+      "epoch": 1.3681159420289855,
+      "grad_norm": 0.3109213272274342,
+      "learning_rate": 3.0220075147611382e-05,
+      "loss": 0.4831,
+      "step": 944
+    },
+    {
+      "epoch": 1.3695652173913042,
+      "grad_norm": 0.7134990744412109,
+      "learning_rate": 3.0193236714975848e-05,
+      "loss": 0.4434,
+      "step": 945
+    },
+    {
+      "epoch": 1.3710144927536232,
+      "grad_norm": 0.3556483056405202,
+      "learning_rate": 3.0166398282340314e-05,
+      "loss": 0.4893,
+      "step": 946
+    },
+    {
+      "epoch": 1.372463768115942,
+      "grad_norm": 0.3827510724428325,
+      "learning_rate": 3.013955984970478e-05,
+      "loss": 0.4244,
+      "step": 947
+    },
+    {
+      "epoch": 1.373913043478261,
+      "grad_norm": 0.33618272241328867,
+      "learning_rate": 3.0112721417069246e-05,
+      "loss": 0.4893,
+      "step": 948
+    },
+    {
+      "epoch": 1.3753623188405797,
+      "grad_norm": 0.3440259382494713,
+      "learning_rate": 3.0085882984433713e-05,
+      "loss": 0.5287,
+      "step": 949
+    },
+    {
+      "epoch": 1.3768115942028984,
+      "grad_norm": 0.3714305425319723,
+      "learning_rate": 3.0059044551798175e-05,
+      "loss": 0.4722,
+      "step": 950
+    },
+    {
+      "epoch": 1.3782608695652174,
+      "grad_norm": 0.3365930331575947,
+      "learning_rate": 3.003220611916264e-05,
+      "loss": 0.4906,
+      "step": 951
+    },
+    {
+      "epoch": 1.3797101449275362,
+      "grad_norm": 0.3448515268305264,
+      "learning_rate": 3.0005367686527108e-05,
+      "loss": 0.4747,
+      "step": 952
+    },
+    {
+      "epoch": 1.3811594202898552,
+      "grad_norm": 0.3191880755633343,
+      "learning_rate": 2.9978529253891574e-05,
+      "loss": 0.4512,
+      "step": 953
+    },
+    {
+      "epoch": 1.382608695652174,
+      "grad_norm": 0.30747956275576543,
+      "learning_rate": 2.995169082125604e-05,
+      "loss": 0.4136,
+      "step": 954
+    },
+    {
+      "epoch": 1.3840579710144927,
+      "grad_norm": 0.31307014206977685,
+      "learning_rate": 2.9924852388620506e-05,
+      "loss": 0.519,
+      "step": 955
+    },
+    {
+      "epoch": 1.3855072463768117,
+      "grad_norm": 0.3223807954231026,
+      "learning_rate": 2.9898013955984972e-05,
+      "loss": 0.4719,
+      "step": 956
+    },
+    {
+      "epoch": 1.3869565217391304,
+      "grad_norm": 0.30410535043204734,
+      "learning_rate": 2.9871175523349438e-05,
+      "loss": 0.607,
+      "step": 957
+    },
+    {
+      "epoch": 1.3884057971014494,
+      "grad_norm": 0.29856137689060125,
+      "learning_rate": 2.9844337090713904e-05,
+      "loss": 0.505,
+      "step": 958
+    },
+    {
+      "epoch": 1.3898550724637682,
+      "grad_norm": 0.34325249374708355,
+      "learning_rate": 2.981749865807837e-05,
+      "loss": 0.5084,
+      "step": 959
+    },
+    {
+      "epoch": 1.391304347826087,
+      "grad_norm": 0.2846508750212212,
+      "learning_rate": 2.9790660225442833e-05,
+      "loss": 0.4798,
+      "step": 960
+    },
+    {
+      "epoch": 1.3927536231884057,
+      "grad_norm": 0.29384666806977755,
+      "learning_rate": 2.97638217928073e-05,
+      "loss": 0.4357,
+      "step": 961
+    },
+    {
+      "epoch": 1.3942028985507247,
+      "grad_norm": 0.3098793105792094,
+      "learning_rate": 2.9736983360171765e-05,
+      "loss": 0.4663,
+      "step": 962
+    },
+    {
+      "epoch": 1.3956521739130434,
+      "grad_norm": 0.280703970437938,
+      "learning_rate": 2.971014492753623e-05,
+      "loss": 0.4517,
+      "step": 963
+    },
+    {
+      "epoch": 1.3971014492753624,
+      "grad_norm": 0.295897243223529,
+      "learning_rate": 2.9683306494900698e-05,
+      "loss": 0.4953,
+      "step": 964
+    },
+    {
+      "epoch": 1.3985507246376812,
+      "grad_norm": 0.2990637911720898,
+      "learning_rate": 2.9656468062265164e-05,
+      "loss": 0.4454,
+      "step": 965
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2841004825626051,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.5554,
+      "step": 966
+    },
+    {
+      "epoch": 1.401449275362319,
+      "grad_norm": 0.3382291317076574,
+      "learning_rate": 2.9602791196994096e-05,
+      "loss": 0.5134,
+      "step": 967
+    },
+    {
+      "epoch": 1.4028985507246376,
+      "grad_norm": 0.2848473660753329,
+      "learning_rate": 2.9575952764358562e-05,
+      "loss": 0.4276,
+      "step": 968
+    },
+    {
+      "epoch": 1.4043478260869566,
+      "grad_norm": 0.30949147729278975,
+      "learning_rate": 2.9549114331723028e-05,
+      "loss": 0.4868,
+      "step": 969
+    },
+    {
+      "epoch": 1.4057971014492754,
+      "grad_norm": 1.375106211694375,
+      "learning_rate": 2.952227589908749e-05,
+      "loss": 0.4457,
+      "step": 970
+    },
+    {
+      "epoch": 1.4072463768115941,
+      "grad_norm": 0.3161686918576888,
+      "learning_rate": 2.9495437466451957e-05,
+      "loss": 0.5123,
+      "step": 971
+    },
+    {
+      "epoch": 1.4086956521739131,
+      "grad_norm": 0.3289921703244623,
+      "learning_rate": 2.9468599033816423e-05,
+      "loss": 0.4721,
+      "step": 972
+    },
+    {
+      "epoch": 1.4101449275362319,
+      "grad_norm": 0.32444685871548057,
+      "learning_rate": 2.944176060118089e-05,
+      "loss": 0.5155,
+      "step": 973
+    },
+    {
+      "epoch": 1.4115942028985506,
+      "grad_norm": 0.34878175924902355,
+      "learning_rate": 2.9414922168545356e-05,
+      "loss": 0.4148,
+      "step": 974
+    },
+    {
+      "epoch": 1.4130434782608696,
+      "grad_norm": 0.6529055435476424,
+      "learning_rate": 2.938808373590982e-05,
+      "loss": 0.5577,
+      "step": 975
+    },
+    {
+      "epoch": 1.4144927536231884,
+      "grad_norm": 0.29726742021900276,
+      "learning_rate": 2.9361245303274295e-05,
+      "loss": 0.4579,
+      "step": 976
+    },
+    {
+      "epoch": 1.4159420289855071,
+      "grad_norm": 0.370721498030071,
+      "learning_rate": 2.9334406870638757e-05,
+      "loss": 0.5317,
+      "step": 977
+    },
+    {
+      "epoch": 1.4173913043478261,
+      "grad_norm": 0.356650197544395,
+      "learning_rate": 2.9307568438003223e-05,
+      "loss": 0.446,
+      "step": 978
+    },
+    {
+      "epoch": 1.4188405797101449,
+      "grad_norm": 0.3052312311447589,
+      "learning_rate": 2.928073000536769e-05,
+      "loss": 0.5167,
+      "step": 979
+    },
+    {
+      "epoch": 1.4202898550724639,
+      "grad_norm": 0.31821019277596285,
+      "learning_rate": 2.9253891572732156e-05,
+      "loss": 0.4607,
+      "step": 980
+    },
+    {
+      "epoch": 1.4217391304347826,
+      "grad_norm": 0.31432739981673774,
+      "learning_rate": 2.9227053140096622e-05,
+      "loss": 0.456,
+      "step": 981
+    },
+    {
+      "epoch": 1.4231884057971014,
+      "grad_norm": 0.3031707155403672,
+      "learning_rate": 2.9200214707461088e-05,
+      "loss": 0.5148,
+      "step": 982
+    },
+    {
+      "epoch": 1.4246376811594204,
+      "grad_norm": 0.2855523267853004,
+      "learning_rate": 2.9173376274825554e-05,
+      "loss": 0.5317,
+      "step": 983
+    },
+    {
+      "epoch": 1.4260869565217391,
+      "grad_norm": 0.2901230981280946,
+      "learning_rate": 2.914653784219002e-05,
+      "loss": 0.4524,
+      "step": 984
+    },
+    {
+      "epoch": 1.427536231884058,
+      "grad_norm": 0.3247998883320533,
+      "learning_rate": 2.9119699409554486e-05,
+      "loss": 0.5147,
+      "step": 985
+    },
+    {
+      "epoch": 1.4289855072463769,
+      "grad_norm": 0.26818811417448163,
+      "learning_rate": 2.9092860976918952e-05,
+      "loss": 0.4362,
+      "step": 986
+    },
+    {
+      "epoch": 1.4304347826086956,
+      "grad_norm": 0.2940552568752322,
+      "learning_rate": 2.9066022544283415e-05,
+      "loss": 0.491,
+      "step": 987
+    },
+    {
+      "epoch": 1.4318840579710144,
+      "grad_norm": 0.3457522998826719,
+      "learning_rate": 2.903918411164788e-05,
+      "loss": 0.489,
+      "step": 988
+    },
+    {
+      "epoch": 1.4333333333333333,
+      "grad_norm": 0.33644944078569905,
+      "learning_rate": 2.9012345679012347e-05,
+      "loss": 0.5334,
+      "step": 989
+    },
+    {
+      "epoch": 1.434782608695652,
+      "grad_norm": 0.34951348063221277,
+      "learning_rate": 2.8985507246376814e-05,
+      "loss": 0.5584,
+      "step": 990
+    },
+    {
+      "epoch": 1.436231884057971,
+      "grad_norm": 0.3619439249364829,
+      "learning_rate": 2.895866881374128e-05,
+      "loss": 0.5196,
+      "step": 991
+    },
+    {
+      "epoch": 1.4376811594202898,
+      "grad_norm": 0.2909575341845204,
+      "learning_rate": 2.8931830381105746e-05,
+      "loss": 0.4389,
+      "step": 992
+    },
+    {
+      "epoch": 1.4391304347826086,
+      "grad_norm": 0.38572710386421416,
+      "learning_rate": 2.8904991948470212e-05,
+      "loss": 0.5097,
+      "step": 993
+    },
+    {
+      "epoch": 1.4405797101449276,
+      "grad_norm": 0.30304027869204375,
+      "learning_rate": 2.8878153515834678e-05,
+      "loss": 0.4838,
+      "step": 994
+    },
+    {
+      "epoch": 1.4420289855072463,
+      "grad_norm": 0.30382150828149235,
+      "learning_rate": 2.8851315083199144e-05,
+      "loss": 0.4964,
+      "step": 995
+    },
+    {
+      "epoch": 1.4434782608695653,
+      "grad_norm": 0.32588305845948534,
+      "learning_rate": 2.882447665056361e-05,
+      "loss": 0.5379,
+      "step": 996
+    },
+    {
+      "epoch": 1.444927536231884,
+      "grad_norm": 0.3469605041772776,
+      "learning_rate": 2.8797638217928073e-05,
+      "loss": 0.4791,
+      "step": 997
+    },
+    {
+      "epoch": 1.4463768115942028,
+      "grad_norm": 0.27111724716871016,
+      "learning_rate": 2.877079978529254e-05,
+      "loss": 0.3857,
+      "step": 998
+    },
+    {
+      "epoch": 1.4478260869565218,
+      "grad_norm": 0.3233826007150532,
+      "learning_rate": 2.8743961352657005e-05,
+      "loss": 0.4538,
+      "step": 999
+    },
+    {
+      "epoch": 1.4492753623188406,
+      "grad_norm": 0.28402937137023476,
+      "learning_rate": 2.871712292002147e-05,
+      "loss": 0.3899,
+      "step": 1000
+    },
+    {
+      "epoch": 1.4507246376811596,
+      "grad_norm": 0.31413076614809343,
+      "learning_rate": 2.8690284487385938e-05,
+      "loss": 0.5691,
+      "step": 1001
+    },
+    {
+      "epoch": 1.4521739130434783,
+      "grad_norm": 0.2943139035596239,
+      "learning_rate": 2.8663446054750404e-05,
+      "loss": 0.4505,
+      "step": 1002
+    },
+    {
+      "epoch": 1.453623188405797,
+      "grad_norm": 0.35202564226634103,
+      "learning_rate": 2.863660762211487e-05,
+      "loss": 0.4805,
+      "step": 1003
+    },
+    {
+      "epoch": 1.4550724637681158,
+      "grad_norm": 0.3473503644422686,
+      "learning_rate": 2.8609769189479336e-05,
+      "loss": 0.5192,
+      "step": 1004
+    },
+    {
+      "epoch": 1.4565217391304348,
+      "grad_norm": 0.3321358690145911,
+      "learning_rate": 2.8582930756843802e-05,
+      "loss": 0.498,
+      "step": 1005
+    },
+    {
+      "epoch": 1.4579710144927536,
+      "grad_norm": 0.3963590731326883,
+      "learning_rate": 2.8556092324208268e-05,
+      "loss": 0.4544,
+      "step": 1006
+    },
+    {
+      "epoch": 1.4594202898550726,
+      "grad_norm": 0.3160377865145725,
+      "learning_rate": 2.852925389157273e-05,
+      "loss": 0.48,
+      "step": 1007
+    },
+    {
+      "epoch": 1.4608695652173913,
+      "grad_norm": 0.3023702273452484,
+      "learning_rate": 2.8502415458937197e-05,
+      "loss": 0.4614,
+      "step": 1008
+    },
+    {
+      "epoch": 1.46231884057971,
+      "grad_norm": 0.32055644945325834,
+      "learning_rate": 2.8475577026301663e-05,
+      "loss": 0.4739,
+      "step": 1009
+    },
+    {
+      "epoch": 1.463768115942029,
+      "grad_norm": 0.363135383170579,
+      "learning_rate": 2.844873859366613e-05,
+      "loss": 0.4962,
+      "step": 1010
+    },
+    {
+      "epoch": 1.4652173913043478,
+      "grad_norm": 0.2860447211182234,
+      "learning_rate": 2.8421900161030595e-05,
+      "loss": 0.4877,
+      "step": 1011
+    },
+    {
+      "epoch": 1.4666666666666668,
+      "grad_norm": 0.3349245845217755,
+      "learning_rate": 2.839506172839506e-05,
+      "loss": 0.4809,
+      "step": 1012
+    },
+    {
+      "epoch": 1.4681159420289855,
+      "grad_norm": 0.34913542414472193,
+      "learning_rate": 2.8368223295759528e-05,
+      "loss": 0.4457,
+      "step": 1013
+    },
+    {
+      "epoch": 1.4695652173913043,
+      "grad_norm": 0.2910127806792282,
+      "learning_rate": 2.8341384863123994e-05,
+      "loss": 0.5023,
+      "step": 1014
+    },
+    {
+      "epoch": 1.471014492753623,
+      "grad_norm": 0.2720985633608247,
+      "learning_rate": 2.831454643048846e-05,
+      "loss": 0.4375,
+      "step": 1015
+    },
+    {
+      "epoch": 1.472463768115942,
+      "grad_norm": 0.30354616941545454,
+      "learning_rate": 2.8287707997852926e-05,
+      "loss": 0.4965,
+      "step": 1016
+    },
+    {
+      "epoch": 1.4739130434782608,
+      "grad_norm": 0.3694880111899378,
+      "learning_rate": 2.826086956521739e-05,
+      "loss": 0.4917,
+      "step": 1017
+    },
+    {
+      "epoch": 1.4753623188405798,
+      "grad_norm": 0.3274882343659536,
+      "learning_rate": 2.8234031132581855e-05,
+      "loss": 0.4815,
+      "step": 1018
+    },
+    {
+      "epoch": 1.4768115942028985,
+      "grad_norm": 0.3224830151935816,
+      "learning_rate": 2.820719269994632e-05,
+      "loss": 0.5586,
+      "step": 1019
+    },
+    {
+      "epoch": 1.4782608695652173,
+      "grad_norm": 0.2980296932580383,
+      "learning_rate": 2.8180354267310787e-05,
+      "loss": 0.5271,
+      "step": 1020
+    },
+    {
+      "epoch": 1.4797101449275363,
+      "grad_norm": 0.34507202615871363,
+      "learning_rate": 2.8153515834675253e-05,
+      "loss": 0.4796,
+      "step": 1021
+    },
+    {
+      "epoch": 1.481159420289855,
+      "grad_norm": 0.3101443752712901,
+      "learning_rate": 2.812667740203972e-05,
+      "loss": 0.4396,
+      "step": 1022
+    },
+    {
+      "epoch": 1.482608695652174,
+      "grad_norm": 0.2685688911780887,
+      "learning_rate": 2.8099838969404192e-05,
+      "loss": 0.4331,
+      "step": 1023
+    },
+    {
+      "epoch": 1.4840579710144928,
+      "grad_norm": 0.2898255728895406,
+      "learning_rate": 2.8073000536768655e-05,
+      "loss": 0.4414,
+      "step": 1024
+    },
+    {
+      "epoch": 1.4855072463768115,
+      "grad_norm": 0.3005447183802071,
+      "learning_rate": 2.804616210413312e-05,
+      "loss": 0.4675,
+      "step": 1025
+    },
+    {
+      "epoch": 1.4869565217391305,
+      "grad_norm": 0.2951594502573726,
+      "learning_rate": 2.8019323671497587e-05,
+      "loss": 0.5167,
+      "step": 1026
+    },
+    {
+      "epoch": 1.4884057971014493,
+      "grad_norm": 0.34141270846717114,
+      "learning_rate": 2.7992485238862053e-05,
+      "loss": 0.5213,
+      "step": 1027
+    },
+    {
+      "epoch": 1.4898550724637682,
+      "grad_norm": 0.2961121782295929,
+      "learning_rate": 2.796564680622652e-05,
+      "loss": 0.5084,
+      "step": 1028
+    },
+    {
+      "epoch": 1.491304347826087,
+      "grad_norm": 0.30238719507345607,
+      "learning_rate": 2.7938808373590986e-05,
+      "loss": 0.4607,
+      "step": 1029
+    },
+    {
+      "epoch": 1.4927536231884058,
+      "grad_norm": 0.4133514663264453,
+      "learning_rate": 2.7911969940955452e-05,
+      "loss": 0.4958,
+      "step": 1030
+    },
+    {
+      "epoch": 1.4942028985507245,
+      "grad_norm": 0.2776853903753614,
+      "learning_rate": 2.7885131508319918e-05,
+      "loss": 0.4545,
+      "step": 1031
+    },
+    {
+      "epoch": 1.4956521739130435,
+      "grad_norm": 0.3355782381599797,
+      "learning_rate": 2.7858293075684384e-05,
+      "loss": 0.5098,
+      "step": 1032
+    },
+    {
+      "epoch": 1.4971014492753623,
+      "grad_norm": 0.3375973164794734,
+      "learning_rate": 2.783145464304885e-05,
+      "loss": 0.454,
+      "step": 1033
+    },
+    {
+      "epoch": 1.4985507246376812,
+      "grad_norm": 0.33212025767660713,
+      "learning_rate": 2.7804616210413313e-05,
+      "loss": 0.474,
+      "step": 1034
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2933684635973571,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 0.5098,
+      "step": 1035
+    },
+    {
+      "epoch": 1.5014492753623188,
+      "grad_norm": 0.3001437074236019,
+      "learning_rate": 2.7750939345142245e-05,
+      "loss": 0.5155,
+      "step": 1036
+    },
+    {
+      "epoch": 1.5028985507246375,
+      "grad_norm": 0.31042500910696075,
+      "learning_rate": 2.772410091250671e-05,
+      "loss": 0.4794,
+      "step": 1037
+    },
+    {
+      "epoch": 1.5043478260869565,
+      "grad_norm": 0.29702687455774746,
+      "learning_rate": 2.7697262479871177e-05,
+      "loss": 0.4966,
+      "step": 1038
+    },
+    {
+      "epoch": 1.5057971014492755,
+      "grad_norm": 0.27626007286792775,
+      "learning_rate": 2.7670424047235644e-05,
+      "loss": 0.4312,
+      "step": 1039
+    },
+    {
+      "epoch": 1.5072463768115942,
+      "grad_norm": 0.30248674822166244,
+      "learning_rate": 2.764358561460011e-05,
+      "loss": 0.5571,
+      "step": 1040
+    },
+    {
+      "epoch": 1.508695652173913,
+      "grad_norm": 0.30886810124630376,
+      "learning_rate": 2.7616747181964576e-05,
+      "loss": 0.5157,
+      "step": 1041
+    },
+    {
+      "epoch": 1.5101449275362318,
+      "grad_norm": 0.2766998203983483,
+      "learning_rate": 2.7589908749329042e-05,
+      "loss": 0.4802,
+      "step": 1042
+    },
+    {
+      "epoch": 1.5115942028985507,
+      "grad_norm": 0.2930564064361767,
+      "learning_rate": 2.7563070316693508e-05,
+      "loss": 0.4346,
+      "step": 1043
+    },
+    {
+      "epoch": 1.5130434782608697,
+      "grad_norm": 0.3021299734706425,
+      "learning_rate": 2.753623188405797e-05,
+      "loss": 0.424,
+      "step": 1044
+    },
+    {
+      "epoch": 1.5144927536231885,
+      "grad_norm": 0.300297304057552,
+      "learning_rate": 2.7509393451422437e-05,
+      "loss": 0.5042,
+      "step": 1045
+    },
+    {
+      "epoch": 1.5159420289855072,
+      "grad_norm": 0.32251092016584043,
+      "learning_rate": 2.7482555018786903e-05,
+      "loss": 0.5228,
+      "step": 1046
+    },
+    {
+      "epoch": 1.517391304347826,
+      "grad_norm": 0.28027744804448423,
+      "learning_rate": 2.745571658615137e-05,
+      "loss": 0.4005,
+      "step": 1047
+    },
+    {
+      "epoch": 1.518840579710145,
+      "grad_norm": 0.35068102203668,
+      "learning_rate": 2.7428878153515835e-05,
+      "loss": 0.4543,
+      "step": 1048
+    },
+    {
+      "epoch": 1.5202898550724637,
+      "grad_norm": 0.3519909596862909,
+      "learning_rate": 2.74020397208803e-05,
+      "loss": 0.4308,
+      "step": 1049
+    },
+    {
+      "epoch": 1.5217391304347827,
+      "grad_norm": 0.3279603569621184,
+      "learning_rate": 2.7375201288244768e-05,
+      "loss": 0.4563,
+      "step": 1050
+    },
+    {
+      "epoch": 1.5231884057971015,
+      "grad_norm": 0.3123333826206131,
+      "learning_rate": 2.7348362855609234e-05,
+      "loss": 0.4796,
+      "step": 1051
+    },
+    {
+      "epoch": 1.5246376811594202,
+      "grad_norm": 0.3786082929605651,
+      "learning_rate": 2.73215244229737e-05,
+      "loss": 0.4699,
+      "step": 1052
+    },
+    {
+      "epoch": 1.526086956521739,
+      "grad_norm": 0.2987854198643322,
+      "learning_rate": 2.7294685990338166e-05,
+      "loss": 0.5033,
+      "step": 1053
+    },
+    {
+      "epoch": 1.527536231884058,
+      "grad_norm": 0.2856444661663017,
+      "learning_rate": 2.7267847557702632e-05,
+      "loss": 0.5121,
+      "step": 1054
+    },
+    {
+      "epoch": 1.528985507246377,
+      "grad_norm": 0.34233136806927067,
+      "learning_rate": 2.7241009125067095e-05,
+      "loss": 0.5266,
+      "step": 1055
+    },
+    {
+      "epoch": 1.5304347826086957,
+      "grad_norm": 0.2925490260180701,
+      "learning_rate": 2.721417069243156e-05,
+      "loss": 0.5282,
+      "step": 1056
+    },
+    {
+      "epoch": 1.5318840579710145,
+      "grad_norm": 0.28624250953418484,
+      "learning_rate": 2.7187332259796027e-05,
+      "loss": 0.4228,
+      "step": 1057
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 0.36892271205960986,
+      "learning_rate": 2.7160493827160493e-05,
+      "loss": 0.5322,
+      "step": 1058
+    },
+    {
+      "epoch": 1.5347826086956522,
+      "grad_norm": 0.31099299091219945,
+      "learning_rate": 2.713365539452496e-05,
+      "loss": 0.5231,
+      "step": 1059
+    },
+    {
+      "epoch": 1.5362318840579712,
+      "grad_norm": 0.3384470762863493,
+      "learning_rate": 2.7106816961889425e-05,
+      "loss": 0.5012,
+      "step": 1060
+    },
+    {
+      "epoch": 1.53768115942029,
+      "grad_norm": 0.30210082438884567,
+      "learning_rate": 2.707997852925389e-05,
+      "loss": 0.4853,
+      "step": 1061
+    },
+    {
+      "epoch": 1.5391304347826087,
+      "grad_norm": 0.3110417076499207,
+      "learning_rate": 2.7053140096618358e-05,
+      "loss": 0.5261,
+      "step": 1062
+    },
+    {
+      "epoch": 1.5405797101449274,
+      "grad_norm": 1.9310222543387976,
+      "learning_rate": 2.7026301663982824e-05,
+      "loss": 0.5916,
+      "step": 1063
+    },
+    {
+      "epoch": 1.5420289855072464,
+      "grad_norm": 0.3433386114056728,
+      "learning_rate": 2.699946323134729e-05,
+      "loss": 0.4571,
+      "step": 1064
+    },
+    {
+      "epoch": 1.5434782608695652,
+      "grad_norm": 0.3309906667592126,
+      "learning_rate": 2.6972624798711753e-05,
+      "loss": 0.4688,
+      "step": 1065
+    },
+    {
+      "epoch": 1.5449275362318842,
+      "grad_norm": 0.29072863860011616,
+      "learning_rate": 2.694578636607622e-05,
+      "loss": 0.4636,
+      "step": 1066
+    },
+    {
+      "epoch": 1.546376811594203,
+      "grad_norm": 0.303646521753606,
+      "learning_rate": 2.6918947933440685e-05,
+      "loss": 0.5115,
+      "step": 1067
+    },
+    {
+      "epoch": 1.5478260869565217,
+      "grad_norm": 0.34580275744019917,
+      "learning_rate": 2.689210950080515e-05,
+      "loss": 0.5092,
+      "step": 1068
+    },
+    {
+      "epoch": 1.5492753623188404,
+      "grad_norm": 0.3385141331273027,
+      "learning_rate": 2.6865271068169624e-05,
+      "loss": 0.4684,
+      "step": 1069
+    },
+    {
+      "epoch": 1.5507246376811594,
+      "grad_norm": 0.35919253427363557,
+      "learning_rate": 2.683843263553409e-05,
+      "loss": 0.525,
+      "step": 1070
+    },
+    {
+      "epoch": 1.5521739130434784,
+      "grad_norm": 0.3464418349850528,
+      "learning_rate": 2.6811594202898553e-05,
+      "loss": 0.5036,
+      "step": 1071
+    },
+    {
+      "epoch": 1.5536231884057972,
+      "grad_norm": 0.3482073902469133,
+      "learning_rate": 2.678475577026302e-05,
+      "loss": 0.4991,
+      "step": 1072
+    },
+    {
+      "epoch": 1.555072463768116,
+      "grad_norm": 0.3343119764801353,
+      "learning_rate": 2.6757917337627485e-05,
+      "loss": 0.5615,
+      "step": 1073
+    },
+    {
+      "epoch": 1.5565217391304347,
+      "grad_norm": 0.29222849822164526,
+      "learning_rate": 2.673107890499195e-05,
+      "loss": 0.4951,
+      "step": 1074
+    },
+    {
+      "epoch": 1.5579710144927537,
+      "grad_norm": 0.31368101687605016,
+      "learning_rate": 2.6704240472356417e-05,
+      "loss": 0.4332,
+      "step": 1075
+    },
+    {
+      "epoch": 1.5594202898550724,
+      "grad_norm": 0.31563321476585815,
+      "learning_rate": 2.6677402039720884e-05,
+      "loss": 0.4807,
+      "step": 1076
+    },
+    {
+      "epoch": 1.5608695652173914,
+      "grad_norm": 0.2771632028980185,
+      "learning_rate": 2.665056360708535e-05,
+      "loss": 0.4293,
+      "step": 1077
+    },
+    {
+      "epoch": 1.5623188405797102,
+      "grad_norm": 0.35885070515840845,
+      "learning_rate": 2.6623725174449816e-05,
+      "loss": 0.491,
+      "step": 1078
+    },
+    {
+      "epoch": 1.563768115942029,
+      "grad_norm": 0.45855129836544456,
+      "learning_rate": 2.6596886741814282e-05,
+      "loss": 0.506,
+      "step": 1079
+    },
+    {
+      "epoch": 1.5652173913043477,
+      "grad_norm": 0.31151851560502486,
+      "learning_rate": 2.6570048309178748e-05,
+      "loss": 0.4129,
+      "step": 1080
+    },
+    {
+      "epoch": 1.5666666666666667,
+      "grad_norm": 0.3940707203914161,
+      "learning_rate": 2.654320987654321e-05,
+      "loss": 0.4713,
+      "step": 1081
+    },
+    {
+      "epoch": 1.5681159420289856,
+      "grad_norm": 0.29630838485277544,
+      "learning_rate": 2.6516371443907677e-05,
+      "loss": 0.4597,
+      "step": 1082
+    },
+    {
+      "epoch": 1.5695652173913044,
+      "grad_norm": 1.3737628784009406,
+      "learning_rate": 2.6489533011272143e-05,
+      "loss": 0.5068,
+      "step": 1083
+    },
+    {
+      "epoch": 1.5710144927536231,
+      "grad_norm": 0.46680947803437023,
+      "learning_rate": 2.646269457863661e-05,
+      "loss": 0.4453,
+      "step": 1084
+    },
+    {
+      "epoch": 1.572463768115942,
+      "grad_norm": 0.32836072314477216,
+      "learning_rate": 2.6435856146001075e-05,
+      "loss": 0.5094,
+      "step": 1085
+    },
+    {
+      "epoch": 1.5739130434782609,
+      "grad_norm": 0.30537238692197544,
+      "learning_rate": 2.640901771336554e-05,
+      "loss": 0.4709,
+      "step": 1086
+    },
+    {
+      "epoch": 1.5753623188405799,
+      "grad_norm": 0.4136853222401632,
+      "learning_rate": 2.6382179280730008e-05,
+      "loss": 0.4336,
+      "step": 1087
+    },
+    {
+      "epoch": 1.5768115942028986,
+      "grad_norm": 0.3865605698775386,
+      "learning_rate": 2.6355340848094474e-05,
+      "loss": 0.4848,
+      "step": 1088
+    },
+    {
+      "epoch": 1.5782608695652174,
+      "grad_norm": 0.36430998727945507,
+      "learning_rate": 2.632850241545894e-05,
+      "loss": 0.5249,
+      "step": 1089
+    },
+    {
+      "epoch": 1.5797101449275361,
+      "grad_norm": 0.47223614909730766,
+      "learning_rate": 2.6301663982823406e-05,
+      "loss": 0.5398,
+      "step": 1090
+    },
+    {
+      "epoch": 1.5811594202898551,
+      "grad_norm": 1.7291963265409254,
+      "learning_rate": 2.6274825550187872e-05,
+      "loss": 0.5247,
+      "step": 1091
+    },
+    {
+      "epoch": 1.5826086956521739,
+      "grad_norm": 0.35052147672518846,
+      "learning_rate": 2.6247987117552335e-05,
+      "loss": 0.47,
+      "step": 1092
+    },
+    {
+      "epoch": 1.5840579710144929,
+      "grad_norm": 0.37445221619061947,
+      "learning_rate": 2.62211486849168e-05,
+      "loss": 0.4582,
+      "step": 1093
+    },
+    {
+      "epoch": 1.5855072463768116,
+      "grad_norm": 0.40176853170877347,
+      "learning_rate": 2.6194310252281267e-05,
+      "loss": 0.5314,
+      "step": 1094
+    },
+    {
+      "epoch": 1.5869565217391304,
+      "grad_norm": 0.3129420729765643,
+      "learning_rate": 2.6167471819645733e-05,
+      "loss": 0.5057,
+      "step": 1095
+    },
+    {
+      "epoch": 1.5884057971014491,
+      "grad_norm": 0.3668781780261487,
+      "learning_rate": 2.61406333870102e-05,
+      "loss": 0.4761,
+      "step": 1096
+    },
+    {
+      "epoch": 1.5898550724637681,
+      "grad_norm": 0.3328390730485344,
+      "learning_rate": 2.6113794954374665e-05,
+      "loss": 0.5399,
+      "step": 1097
+    },
+    {
+      "epoch": 1.591304347826087,
+      "grad_norm": 0.34513824383552366,
+      "learning_rate": 2.608695652173913e-05,
+      "loss": 0.4,
+      "step": 1098
+    },
+    {
+      "epoch": 1.5927536231884059,
+      "grad_norm": 0.32500164182044994,
+      "learning_rate": 2.6060118089103598e-05,
+      "loss": 0.4751,
+      "step": 1099
+    },
+    {
+      "epoch": 1.5942028985507246,
+      "grad_norm": 0.2762870180657687,
+      "learning_rate": 2.6033279656468064e-05,
+      "loss": 0.5027,
+      "step": 1100
+    },
+    {
+      "epoch": 1.5956521739130434,
+      "grad_norm": 0.31612062911505995,
+      "learning_rate": 2.600644122383253e-05,
+      "loss": 0.4754,
+      "step": 1101
+    },
+    {
+      "epoch": 1.5971014492753624,
+      "grad_norm": 0.2938548583813651,
+      "learning_rate": 2.5979602791196993e-05,
+      "loss": 0.4709,
+      "step": 1102
+    },
+    {
+      "epoch": 1.598550724637681,
+      "grad_norm": 0.26715822202118394,
+      "learning_rate": 2.595276435856146e-05,
+      "loss": 0.4444,
+      "step": 1103
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3416266894284156,
+      "learning_rate": 2.5925925925925925e-05,
+      "loss": 0.5503,
+      "step": 1104
+    },
+    {
+      "epoch": 1.6014492753623188,
+      "grad_norm": 0.3216038263370538,
+      "learning_rate": 2.589908749329039e-05,
+      "loss": 0.5112,
+      "step": 1105
+    },
+    {
+      "epoch": 1.6028985507246376,
+      "grad_norm": 0.27503545942301794,
+      "learning_rate": 2.5872249060654857e-05,
+      "loss": 0.462,
+      "step": 1106
+    },
+    {
+      "epoch": 1.6043478260869564,
+      "grad_norm": 0.36952921169438074,
+      "learning_rate": 2.5845410628019323e-05,
+      "loss": 0.5123,
+      "step": 1107
+    },
+    {
+      "epoch": 1.6057971014492753,
+      "grad_norm": 0.3117300813713928,
+      "learning_rate": 2.581857219538379e-05,
+      "loss": 0.5011,
+      "step": 1108
+    },
+    {
+      "epoch": 1.6072463768115943,
+      "grad_norm": 0.2992474583613941,
+      "learning_rate": 2.5791733762748256e-05,
+      "loss": 0.5008,
+      "step": 1109
+    },
+    {
+      "epoch": 1.608695652173913,
+      "grad_norm": 0.3014995756227707,
+      "learning_rate": 2.576489533011272e-05,
+      "loss": 0.4762,
+      "step": 1110
+    },
+    {
+      "epoch": 1.6101449275362318,
+      "grad_norm": 0.35343798929006137,
+      "learning_rate": 2.5738056897477188e-05,
+      "loss": 0.5262,
+      "step": 1111
+    },
+    {
+      "epoch": 1.6115942028985506,
+      "grad_norm": 0.2577183024889789,
+      "learning_rate": 2.571121846484165e-05,
+      "loss": 0.4279,
+      "step": 1112
+    },
+    {
+      "epoch": 1.6130434782608696,
+      "grad_norm": 0.3034987372845954,
+      "learning_rate": 2.5684380032206117e-05,
+      "loss": 0.5026,
+      "step": 1113
+    },
+    {
+      "epoch": 1.6144927536231886,
+      "grad_norm": 0.33673464929484864,
+      "learning_rate": 2.5657541599570583e-05,
+      "loss": 0.4862,
+      "step": 1114
+    },
+    {
+      "epoch": 1.6159420289855073,
+      "grad_norm": 0.3384160427246048,
+      "learning_rate": 2.563070316693505e-05,
+      "loss": 0.5445,
+      "step": 1115
+    },
+    {
+      "epoch": 1.617391304347826,
+      "grad_norm": 0.32413449758544177,
+      "learning_rate": 2.5603864734299522e-05,
+      "loss": 0.4661,
+      "step": 1116
+    },
+    {
+      "epoch": 1.6188405797101448,
+      "grad_norm": 0.3144244703969793,
+      "learning_rate": 2.5577026301663988e-05,
+      "loss": 0.4783,
+      "step": 1117
+    },
+    {
+      "epoch": 1.6202898550724638,
+      "grad_norm": 0.35066050252787756,
+      "learning_rate": 2.5550187869028454e-05,
+      "loss": 0.4951,
+      "step": 1118
+    },
+    {
+      "epoch": 1.6217391304347826,
+      "grad_norm": 0.27519048555942127,
+      "learning_rate": 2.5523349436392917e-05,
+      "loss": 0.4505,
+      "step": 1119
+    },
+    {
+      "epoch": 1.6231884057971016,
+      "grad_norm": 0.3545251856066128,
+      "learning_rate": 2.5496511003757383e-05,
+      "loss": 0.5553,
+      "step": 1120
+    },
+    {
+      "epoch": 1.6246376811594203,
+      "grad_norm": 0.32213559162233424,
+      "learning_rate": 2.546967257112185e-05,
+      "loss": 0.5033,
+      "step": 1121
+    },
+    {
+      "epoch": 1.626086956521739,
+      "grad_norm": 0.32983910484047146,
+      "learning_rate": 2.5442834138486315e-05,
+      "loss": 0.4894,
+      "step": 1122
+    },
+    {
+      "epoch": 1.6275362318840578,
+      "grad_norm": 0.3202291330759991,
+      "learning_rate": 2.541599570585078e-05,
+      "loss": 0.5531,
+      "step": 1123
+    },
+    {
+      "epoch": 1.6289855072463768,
+      "grad_norm": 0.32166747117547634,
+      "learning_rate": 2.5389157273215247e-05,
+      "loss": 0.4613,
+      "step": 1124
+    },
+    {
+      "epoch": 1.6304347826086958,
+      "grad_norm": 0.2886548591988718,
+      "learning_rate": 2.5362318840579714e-05,
+      "loss": 0.4716,
+      "step": 1125
+    },
+    {
+      "epoch": 1.6318840579710145,
+      "grad_norm": 0.40006321648929954,
+      "learning_rate": 2.533548040794418e-05,
+      "loss": 0.5574,
+      "step": 1126
+    },
+    {
+      "epoch": 1.6333333333333333,
+      "grad_norm": 0.3145464195782174,
+      "learning_rate": 2.5308641975308646e-05,
+      "loss": 0.4951,
+      "step": 1127
+    },
+    {
+      "epoch": 1.634782608695652,
+      "grad_norm": 0.3053615646973681,
+      "learning_rate": 2.5281803542673112e-05,
+      "loss": 0.4936,
+      "step": 1128
+    },
+    {
+      "epoch": 1.636231884057971,
+      "grad_norm": 0.333201209211236,
+      "learning_rate": 2.5254965110037575e-05,
+      "loss": 0.4611,
+      "step": 1129
+    },
+    {
+      "epoch": 1.6376811594202898,
+      "grad_norm": 0.2851998465250763,
+      "learning_rate": 2.522812667740204e-05,
+      "loss": 0.4738,
+      "step": 1130
+    },
+    {
+      "epoch": 1.6391304347826088,
+      "grad_norm": 0.32565555228536963,
+      "learning_rate": 2.5201288244766507e-05,
+      "loss": 0.5086,
+      "step": 1131
+    },
+    {
+      "epoch": 1.6405797101449275,
+      "grad_norm": 0.3218307726528686,
+      "learning_rate": 2.5174449812130973e-05,
+      "loss": 0.4572,
+      "step": 1132
+    },
+    {
+      "epoch": 1.6420289855072463,
+      "grad_norm": 0.33430033216292593,
+      "learning_rate": 2.514761137949544e-05,
+      "loss": 0.449,
+      "step": 1133
+    },
+    {
+      "epoch": 1.643478260869565,
+      "grad_norm": 0.27199374077107724,
+      "learning_rate": 2.5120772946859905e-05,
+      "loss": 0.4553,
+      "step": 1134
+    },
+    {
+      "epoch": 1.644927536231884,
+      "grad_norm": 0.3023340056172451,
+      "learning_rate": 2.509393451422437e-05,
+      "loss": 0.544,
+      "step": 1135
+    },
+    {
+      "epoch": 1.646376811594203,
+      "grad_norm": 0.2613098320681714,
+      "learning_rate": 2.5067096081588838e-05,
+      "loss": 0.4587,
+      "step": 1136
+    },
+    {
+      "epoch": 1.6478260869565218,
+      "grad_norm": 0.2974510036323834,
+      "learning_rate": 2.5040257648953304e-05,
+      "loss": 0.5263,
+      "step": 1137
+    },
+    {
+      "epoch": 1.6492753623188405,
+      "grad_norm": 0.2940780318124218,
+      "learning_rate": 2.501341921631777e-05,
+      "loss": 0.5851,
+      "step": 1138
+    },
+    {
+      "epoch": 1.6507246376811593,
+      "grad_norm": 0.27179550390114365,
+      "learning_rate": 2.4986580783682233e-05,
+      "loss": 0.5108,
+      "step": 1139
+    },
+    {
+      "epoch": 1.6521739130434783,
+      "grad_norm": 0.2810015927705504,
+      "learning_rate": 2.49597423510467e-05,
+      "loss": 0.4957,
+      "step": 1140
+    },
+    {
+      "epoch": 1.6536231884057973,
+      "grad_norm": 0.45762102953727496,
+      "learning_rate": 2.4932903918411165e-05,
+      "loss": 0.4951,
+      "step": 1141
+    },
+    {
+      "epoch": 1.655072463768116,
+      "grad_norm": 0.2904087653620613,
+      "learning_rate": 2.490606548577563e-05,
+      "loss": 0.4773,
+      "step": 1142
+    },
+    {
+      "epoch": 1.6565217391304348,
+      "grad_norm": 0.3015137691314487,
+      "learning_rate": 2.4879227053140097e-05,
+      "loss": 0.4079,
+      "step": 1143
+    },
+    {
+      "epoch": 1.6579710144927535,
+      "grad_norm": 0.28317996532363227,
+      "learning_rate": 2.4852388620504563e-05,
+      "loss": 0.5032,
+      "step": 1144
+    },
+    {
+      "epoch": 1.6594202898550725,
+      "grad_norm": 0.2672615381516704,
+      "learning_rate": 2.482555018786903e-05,
+      "loss": 0.4898,
+      "step": 1145
+    },
+    {
+      "epoch": 1.6608695652173913,
+      "grad_norm": 0.2966522019850434,
+      "learning_rate": 2.4798711755233495e-05,
+      "loss": 0.4761,
+      "step": 1146
+    },
+    {
+      "epoch": 1.6623188405797102,
+      "grad_norm": 0.3239107334984002,
+      "learning_rate": 2.477187332259796e-05,
+      "loss": 0.4442,
+      "step": 1147
+    },
+    {
+      "epoch": 1.663768115942029,
+      "grad_norm": 0.31042620564122636,
+      "learning_rate": 2.4745034889962428e-05,
+      "loss": 0.4839,
+      "step": 1148
+    },
+    {
+      "epoch": 1.6652173913043478,
+      "grad_norm": 0.35048458578602115,
+      "learning_rate": 2.471819645732689e-05,
+      "loss": 0.528,
+      "step": 1149
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.2980529297686646,
+      "learning_rate": 2.4691358024691357e-05,
+      "loss": 0.4833,
+      "step": 1150
+    },
+    {
+      "epoch": 1.6681159420289855,
+      "grad_norm": 0.3672928132817581,
+      "learning_rate": 2.4664519592055826e-05,
+      "loss": 0.4872,
+      "step": 1151
+    },
+    {
+      "epoch": 1.6695652173913045,
+      "grad_norm": 0.2820395944630137,
+      "learning_rate": 2.4637681159420292e-05,
+      "loss": 0.4809,
+      "step": 1152
+    },
+    {
+      "epoch": 1.6710144927536232,
+      "grad_norm": 0.31450043773062825,
+      "learning_rate": 2.461084272678476e-05,
+      "loss": 0.4956,
+      "step": 1153
+    },
+    {
+      "epoch": 1.672463768115942,
+      "grad_norm": 0.30192432417594417,
+      "learning_rate": 2.4584004294149224e-05,
+      "loss": 0.4635,
+      "step": 1154
+    },
+    {
+      "epoch": 1.6739130434782608,
+      "grad_norm": 0.27676854093728326,
+      "learning_rate": 2.455716586151369e-05,
+      "loss": 0.4223,
+      "step": 1155
+    },
+    {
+      "epoch": 1.6753623188405797,
+      "grad_norm": 0.30919443229083027,
+      "learning_rate": 2.4530327428878157e-05,
+      "loss": 0.5351,
+      "step": 1156
+    },
+    {
+      "epoch": 1.6768115942028987,
+      "grad_norm": 0.29361155971628206,
+      "learning_rate": 2.4503488996242623e-05,
+      "loss": 0.5295,
+      "step": 1157
+    },
+    {
+      "epoch": 1.6782608695652175,
+      "grad_norm": 0.3047886425873606,
+      "learning_rate": 2.4476650563607086e-05,
+      "loss": 0.4427,
+      "step": 1158
+    },
+    {
+      "epoch": 1.6797101449275362,
+      "grad_norm": 0.28166533462952836,
+      "learning_rate": 2.4449812130971552e-05,
+      "loss": 0.4584,
+      "step": 1159
+    },
+    {
+      "epoch": 1.681159420289855,
+      "grad_norm": 0.3019997528285114,
+      "learning_rate": 2.4422973698336018e-05,
+      "loss": 0.4257,
+      "step": 1160
+    },
+    {
+      "epoch": 1.6826086956521737,
+      "grad_norm": 0.3171044179739617,
+      "learning_rate": 2.4396135265700484e-05,
+      "loss": 0.5039,
+      "step": 1161
+    },
+    {
+      "epoch": 1.6840579710144927,
+      "grad_norm": 0.34308975241506323,
+      "learning_rate": 2.436929683306495e-05,
+      "loss": 0.5566,
+      "step": 1162
+    },
+    {
+      "epoch": 1.6855072463768117,
+      "grad_norm": 3.039656774093585,
+      "learning_rate": 2.4342458400429416e-05,
+      "loss": 0.5861,
+      "step": 1163
+    },
+    {
+      "epoch": 1.6869565217391305,
+      "grad_norm": 0.35515967239526663,
+      "learning_rate": 2.4315619967793882e-05,
+      "loss": 0.477,
+      "step": 1164
+    },
+    {
+      "epoch": 1.6884057971014492,
+      "grad_norm": 0.37102830820163163,
+      "learning_rate": 2.428878153515835e-05,
+      "loss": 0.4585,
+      "step": 1165
+    },
+    {
+      "epoch": 1.689855072463768,
+      "grad_norm": 0.2824586787793108,
+      "learning_rate": 2.4261943102522815e-05,
+      "loss": 0.4651,
+      "step": 1166
+    },
+    {
+      "epoch": 1.691304347826087,
+      "grad_norm": 0.26716569849813837,
+      "learning_rate": 2.423510466988728e-05,
+      "loss": 0.4307,
+      "step": 1167
+    },
+    {
+      "epoch": 1.692753623188406,
+      "grad_norm": 0.30556825406142546,
+      "learning_rate": 2.4208266237251743e-05,
+      "loss": 0.4892,
+      "step": 1168
+    },
+    {
+      "epoch": 1.6942028985507247,
+      "grad_norm": 0.3604146757235219,
+      "learning_rate": 2.418142780461621e-05,
+      "loss": 0.5213,
+      "step": 1169
+    },
+    {
+      "epoch": 1.6956521739130435,
+      "grad_norm": 0.32770296185197473,
+      "learning_rate": 2.4154589371980676e-05,
+      "loss": 0.544,
+      "step": 1170
+    },
+    {
+      "epoch": 1.6971014492753622,
+      "grad_norm": 0.36987261868006993,
+      "learning_rate": 2.4127750939345142e-05,
+      "loss": 0.4964,
+      "step": 1171
+    },
+    {
+      "epoch": 1.6985507246376812,
+      "grad_norm": 0.30640544794163954,
+      "learning_rate": 2.4100912506709608e-05,
+      "loss": 0.4344,
+      "step": 1172
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.32343856057785547,
+      "learning_rate": 2.4074074074074074e-05,
+      "loss": 0.4784,
+      "step": 1173
+    },
+    {
+      "epoch": 1.701449275362319,
+      "grad_norm": 0.3305229567420513,
+      "learning_rate": 2.4047235641438544e-05,
+      "loss": 0.4911,
+      "step": 1174
+    },
+    {
+      "epoch": 1.7028985507246377,
+      "grad_norm": 0.2782883099862748,
+      "learning_rate": 2.402039720880301e-05,
+      "loss": 0.4965,
+      "step": 1175
+    },
+    {
+      "epoch": 1.7043478260869565,
+      "grad_norm": 0.3142883185393042,
+      "learning_rate": 2.3993558776167472e-05,
+      "loss": 0.4349,
+      "step": 1176
+    },
+    {
+      "epoch": 1.7057971014492752,
+      "grad_norm": 0.6268301065499224,
+      "learning_rate": 2.396672034353194e-05,
+      "loss": 0.5338,
+      "step": 1177
+    },
+    {
+      "epoch": 1.7072463768115942,
+      "grad_norm": 0.3213649355828965,
+      "learning_rate": 2.3939881910896405e-05,
+      "loss": 0.5142,
+      "step": 1178
+    },
+    {
+      "epoch": 1.7086956521739132,
+      "grad_norm": 0.3190793133605725,
+      "learning_rate": 2.391304347826087e-05,
+      "loss": 0.438,
+      "step": 1179
+    },
+    {
+      "epoch": 1.710144927536232,
+      "grad_norm": 0.2750310402906081,
+      "learning_rate": 2.3886205045625337e-05,
+      "loss": 0.426,
+      "step": 1180
+    },
+    {
+      "epoch": 1.7115942028985507,
+      "grad_norm": 0.2757694770103066,
+      "learning_rate": 2.3859366612989803e-05,
+      "loss": 0.4465,
+      "step": 1181
+    },
+    {
+      "epoch": 1.7130434782608694,
+      "grad_norm": 0.30493489587722383,
+      "learning_rate": 2.383252818035427e-05,
+      "loss": 0.5027,
+      "step": 1182
+    },
+    {
+      "epoch": 1.7144927536231884,
+      "grad_norm": 0.32344723622363536,
+      "learning_rate": 2.3805689747718735e-05,
+      "loss": 0.4986,
+      "step": 1183
+    },
+    {
+      "epoch": 1.7159420289855074,
+      "grad_norm": 0.26856516264127966,
+      "learning_rate": 2.37788513150832e-05,
+      "loss": 0.4792,
+      "step": 1184
+    },
+    {
+      "epoch": 1.7173913043478262,
+      "grad_norm": 0.3250264038161418,
+      "learning_rate": 2.3752012882447668e-05,
+      "loss": 0.5504,
+      "step": 1185
+    },
+    {
+      "epoch": 1.718840579710145,
+      "grad_norm": 0.34292399150690334,
+      "learning_rate": 2.372517444981213e-05,
+      "loss": 0.4606,
+      "step": 1186
+    },
+    {
+      "epoch": 1.7202898550724637,
+      "grad_norm": 0.2721807363721305,
+      "learning_rate": 2.3698336017176596e-05,
+      "loss": 0.4407,
+      "step": 1187
+    },
+    {
+      "epoch": 1.7217391304347827,
+      "grad_norm": 0.3653483444914663,
+      "learning_rate": 2.3671497584541063e-05,
+      "loss": 0.5293,
+      "step": 1188
+    },
+    {
+      "epoch": 1.7231884057971014,
+      "grad_norm": 0.3310831455581937,
+      "learning_rate": 2.364465915190553e-05,
+      "loss": 0.5508,
+      "step": 1189
+    },
+    {
+      "epoch": 1.7246376811594204,
+      "grad_norm": 0.365199402173693,
+      "learning_rate": 2.3617820719269995e-05,
+      "loss": 0.5591,
+      "step": 1190
+    },
+    {
+      "epoch": 1.7260869565217392,
+      "grad_norm": 0.3178140233613037,
+      "learning_rate": 2.359098228663446e-05,
+      "loss": 0.4831,
+      "step": 1191
+    },
+    {
+      "epoch": 1.727536231884058,
+      "grad_norm": 0.3563520913233769,
+      "learning_rate": 2.3564143853998927e-05,
+      "loss": 0.4858,
+      "step": 1192
+    },
+    {
+      "epoch": 1.7289855072463767,
+      "grad_norm": 0.2976845205796779,
+      "learning_rate": 2.3537305421363393e-05,
+      "loss": 0.476,
+      "step": 1193
+    },
+    {
+      "epoch": 1.7304347826086957,
+      "grad_norm": 0.34241650845225546,
+      "learning_rate": 2.351046698872786e-05,
+      "loss": 0.4882,
+      "step": 1194
+    },
+    {
+      "epoch": 1.7318840579710146,
+      "grad_norm": 0.3005603672201895,
+      "learning_rate": 2.3483628556092325e-05,
+      "loss": 0.4768,
+      "step": 1195
+    },
+    {
+      "epoch": 1.7333333333333334,
+      "grad_norm": 0.3854426600408823,
+      "learning_rate": 2.345679012345679e-05,
+      "loss": 0.5329,
+      "step": 1196
+    },
+    {
+      "epoch": 1.7347826086956522,
+      "grad_norm": 0.3723887263936342,
+      "learning_rate": 2.3429951690821258e-05,
+      "loss": 0.482,
+      "step": 1197
+    },
+    {
+      "epoch": 1.736231884057971,
+      "grad_norm": 0.3529567447157191,
+      "learning_rate": 2.3403113258185724e-05,
+      "loss": 0.5135,
+      "step": 1198
+    },
+    {
+      "epoch": 1.73768115942029,
+      "grad_norm": 0.40448836607297356,
+      "learning_rate": 2.337627482555019e-05,
+      "loss": 0.4882,
+      "step": 1199
+    },
+    {
+      "epoch": 1.7391304347826086,
+      "grad_norm": 0.32133621387232536,
+      "learning_rate": 2.3349436392914656e-05,
+      "loss": 0.473,
+      "step": 1200
+    },
+    {
+      "epoch": 1.7405797101449276,
+      "grad_norm": 1.0852661570450723,
+      "learning_rate": 2.3322597960279122e-05,
+      "loss": 0.4425,
+      "step": 1201
+    },
+    {
+      "epoch": 1.7420289855072464,
+      "grad_norm": 0.3711355431298706,
+      "learning_rate": 2.329575952764359e-05,
+      "loss": 0.4347,
+      "step": 1202
+    },
+    {
+      "epoch": 1.7434782608695651,
+      "grad_norm": 0.3379298215669769,
+      "learning_rate": 2.3268921095008055e-05,
+      "loss": 0.4532,
+      "step": 1203
+    },
+    {
+      "epoch": 1.744927536231884,
+      "grad_norm": 0.3569245007981008,
+      "learning_rate": 2.324208266237252e-05,
+      "loss": 0.5119,
+      "step": 1204
+    },
+    {
+      "epoch": 1.7463768115942029,
+      "grad_norm": 0.2830441058655675,
+      "learning_rate": 2.3215244229736983e-05,
+      "loss": 0.4829,
+      "step": 1205
+    },
+    {
+      "epoch": 1.7478260869565219,
+      "grad_norm": 0.29739221481915823,
+      "learning_rate": 2.318840579710145e-05,
+      "loss": 0.4562,
+      "step": 1206
+    },
+    {
+      "epoch": 1.7492753623188406,
+      "grad_norm": 0.325030040824517,
+      "learning_rate": 2.3161567364465916e-05,
+      "loss": 0.4666,
+      "step": 1207
+    },
+    {
+      "epoch": 1.7507246376811594,
+      "grad_norm": 0.305603115485528,
+      "learning_rate": 2.3134728931830382e-05,
+      "loss": 0.5258,
+      "step": 1208
+    },
+    {
+      "epoch": 1.7521739130434781,
+      "grad_norm": 0.35825304542957837,
+      "learning_rate": 2.3107890499194848e-05,
+      "loss": 0.5211,
+      "step": 1209
+    },
+    {
+      "epoch": 1.7536231884057971,
+      "grad_norm": 0.29913595053269965,
+      "learning_rate": 2.3081052066559314e-05,
+      "loss": 0.5621,
+      "step": 1210
+    },
+    {
+      "epoch": 1.755072463768116,
+      "grad_norm": 0.3233729807843389,
+      "learning_rate": 2.305421363392378e-05,
+      "loss": 0.4016,
+      "step": 1211
+    },
+    {
+      "epoch": 1.7565217391304349,
+      "grad_norm": 0.41568565650293215,
+      "learning_rate": 2.3027375201288246e-05,
+      "loss": 0.5101,
+      "step": 1212
+    },
+    {
+      "epoch": 1.7579710144927536,
+      "grad_norm": 0.2868122902243216,
+      "learning_rate": 2.3000536768652712e-05,
+      "loss": 0.4703,
+      "step": 1213
+    },
+    {
+      "epoch": 1.7594202898550724,
+      "grad_norm": 0.3058263429464893,
+      "learning_rate": 2.297369833601718e-05,
+      "loss": 0.4488,
+      "step": 1214
+    },
+    {
+      "epoch": 1.7608695652173914,
+      "grad_norm": 0.3019552092758227,
+      "learning_rate": 2.294685990338164e-05,
+      "loss": 0.4847,
+      "step": 1215
+    },
+    {
+      "epoch": 1.76231884057971,
+      "grad_norm": 0.26413528924068774,
+      "learning_rate": 2.2920021470746107e-05,
+      "loss": 0.4644,
+      "step": 1216
+    },
+    {
+      "epoch": 1.763768115942029,
+      "grad_norm": 0.36682267539610747,
+      "learning_rate": 2.2893183038110574e-05,
+      "loss": 0.4519,
+      "step": 1217
+    },
+    {
+      "epoch": 1.7652173913043478,
+      "grad_norm": 0.28753522716414265,
+      "learning_rate": 2.286634460547504e-05,
+      "loss": 0.473,
+      "step": 1218
+    },
+    {
+      "epoch": 1.7666666666666666,
+      "grad_norm": 0.3026541886137627,
+      "learning_rate": 2.2839506172839506e-05,
+      "loss": 0.5092,
+      "step": 1219
+    },
+    {
+      "epoch": 1.7681159420289854,
+      "grad_norm": 0.30391904667726166,
+      "learning_rate": 2.2812667740203972e-05,
+      "loss": 0.5232,
+      "step": 1220
+    },
+    {
+      "epoch": 1.7695652173913043,
+      "grad_norm": 0.2852925782044705,
+      "learning_rate": 2.278582930756844e-05,
+      "loss": 0.4252,
+      "step": 1221
+    },
+    {
+      "epoch": 1.7710144927536233,
+      "grad_norm": 0.3000533868403904,
+      "learning_rate": 2.2758990874932908e-05,
+      "loss": 0.5337,
+      "step": 1222
+    },
+    {
+      "epoch": 1.772463768115942,
+      "grad_norm": 0.30059390751594706,
+      "learning_rate": 2.273215244229737e-05,
+      "loss": 0.4557,
+      "step": 1223
+    },
+    {
+      "epoch": 1.7739130434782608,
+      "grad_norm": 0.26736425462212277,
+      "learning_rate": 2.2705314009661836e-05,
+      "loss": 0.5,
+      "step": 1224
+    },
+    {
+      "epoch": 1.7753623188405796,
+      "grad_norm": 0.3070528336704938,
+      "learning_rate": 2.2678475577026303e-05,
+      "loss": 0.5455,
+      "step": 1225
+    },
+    {
+      "epoch": 1.7768115942028986,
+      "grad_norm": 0.3420183725834646,
+      "learning_rate": 2.265163714439077e-05,
+      "loss": 0.4505,
+      "step": 1226
+    },
+    {
+      "epoch": 1.7782608695652173,
+      "grad_norm": 0.32331303715104653,
+      "learning_rate": 2.2624798711755235e-05,
+      "loss": 0.4736,
+      "step": 1227
+    },
+    {
+      "epoch": 1.7797101449275363,
+      "grad_norm": 0.25845807346373795,
+      "learning_rate": 2.25979602791197e-05,
+      "loss": 0.436,
+      "step": 1228
+    },
+    {
+      "epoch": 1.781159420289855,
+      "grad_norm": 0.3116846376099379,
+      "learning_rate": 2.2571121846484167e-05,
+      "loss": 0.4876,
+      "step": 1229
+    },
+    {
+      "epoch": 1.7826086956521738,
+      "grad_norm": 0.2824150019214866,
+      "learning_rate": 2.2544283413848633e-05,
+      "loss": 0.4713,
+      "step": 1230
+    },
+    {
+      "epoch": 1.7840579710144926,
+      "grad_norm": 0.31803201506817763,
+      "learning_rate": 2.25174449812131e-05,
+      "loss": 0.4723,
+      "step": 1231
+    },
+    {
+      "epoch": 1.7855072463768116,
+      "grad_norm": 0.3104614414725165,
+      "learning_rate": 2.2490606548577565e-05,
+      "loss": 0.4612,
+      "step": 1232
+    },
+    {
+      "epoch": 1.7869565217391306,
+      "grad_norm": 0.3292079708354603,
+      "learning_rate": 2.246376811594203e-05,
+      "loss": 0.4648,
+      "step": 1233
+    },
+    {
+      "epoch": 1.7884057971014493,
+      "grad_norm": 0.3072965120142354,
+      "learning_rate": 2.2436929683306494e-05,
+      "loss": 0.5005,
+      "step": 1234
+    },
+    {
+      "epoch": 1.789855072463768,
+      "grad_norm": 0.32424907723585317,
+      "learning_rate": 2.241009125067096e-05,
+      "loss": 0.4711,
+      "step": 1235
+    },
+    {
+      "epoch": 1.7913043478260868,
+      "grad_norm": 0.31647288097845033,
+      "learning_rate": 2.2383252818035427e-05,
+      "loss": 0.4186,
+      "step": 1236
+    },
+    {
+      "epoch": 1.7927536231884058,
+      "grad_norm": 0.3370391739929275,
+      "learning_rate": 2.2356414385399893e-05,
+      "loss": 0.4478,
+      "step": 1237
+    },
+    {
+      "epoch": 1.7942028985507248,
+      "grad_norm": 0.28465711257853077,
+      "learning_rate": 2.232957595276436e-05,
+      "loss": 0.4547,
+      "step": 1238
+    },
+    {
+      "epoch": 1.7956521739130435,
+      "grad_norm": 0.3500145854808722,
+      "learning_rate": 2.2302737520128825e-05,
+      "loss": 0.4533,
+      "step": 1239
+    },
+    {
+      "epoch": 1.7971014492753623,
+      "grad_norm": 0.35577089369408404,
+      "learning_rate": 2.227589908749329e-05,
+      "loss": 0.468,
+      "step": 1240
+    },
+    {
+      "epoch": 1.798550724637681,
+      "grad_norm": 0.27215161526036075,
+      "learning_rate": 2.2249060654857757e-05,
+      "loss": 0.4879,
+      "step": 1241
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.31464918485266263,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.5211,
+      "step": 1242
+    },
+    {
+      "epoch": 1.8014492753623188,
+      "grad_norm": 0.28959716892415166,
+      "learning_rate": 2.219538378958669e-05,
+      "loss": 0.4034,
+      "step": 1243
+    },
+    {
+      "epoch": 1.8028985507246378,
+      "grad_norm": 0.3133095055561112,
+      "learning_rate": 2.2168545356951156e-05,
+      "loss": 0.4886,
+      "step": 1244
+    },
+    {
+      "epoch": 1.8043478260869565,
+      "grad_norm": 0.3035834128041399,
+      "learning_rate": 2.214170692431562e-05,
+      "loss": 0.5295,
+      "step": 1245
+    },
+    {
+      "epoch": 1.8057971014492753,
+      "grad_norm": 0.3041539027975744,
+      "learning_rate": 2.2114868491680088e-05,
+      "loss": 0.5061,
+      "step": 1246
+    },
+    {
+      "epoch": 1.807246376811594,
+      "grad_norm": 0.2966047899734678,
+      "learning_rate": 2.2088030059044554e-05,
+      "loss": 0.4658,
+      "step": 1247
+    },
+    {
+      "epoch": 1.808695652173913,
+      "grad_norm": 0.29187310468867916,
+      "learning_rate": 2.206119162640902e-05,
+      "loss": 0.4448,
+      "step": 1248
+    },
+    {
+      "epoch": 1.810144927536232,
+      "grad_norm": 0.26644541230528856,
+      "learning_rate": 2.2034353193773486e-05,
+      "loss": 0.4782,
+      "step": 1249
+    },
+    {
+      "epoch": 1.8115942028985508,
+      "grad_norm": 0.29148279845891767,
+      "learning_rate": 2.2007514761137952e-05,
+      "loss": 0.4607,
+      "step": 1250
+    },
+    {
+      "epoch": 1.8130434782608695,
+      "grad_norm": 0.27695780648996593,
+      "learning_rate": 2.198067632850242e-05,
+      "loss": 0.431,
+      "step": 1251
+    },
+    {
+      "epoch": 1.8144927536231883,
+      "grad_norm": 0.27383782215227476,
+      "learning_rate": 2.195383789586688e-05,
+      "loss": 0.4782,
+      "step": 1252
+    },
+    {
+      "epoch": 1.8159420289855073,
+      "grad_norm": 0.256922040034402,
+      "learning_rate": 2.1926999463231347e-05,
+      "loss": 0.4699,
+      "step": 1253
+    },
+    {
+      "epoch": 1.8173913043478263,
+      "grad_norm": 0.28676204052163945,
+      "learning_rate": 2.1900161030595813e-05,
+      "loss": 0.4642,
+      "step": 1254
+    },
+    {
+      "epoch": 1.818840579710145,
+      "grad_norm": 0.27090218727496,
+      "learning_rate": 2.187332259796028e-05,
+      "loss": 0.5015,
+      "step": 1255
+    },
+    {
+      "epoch": 1.8202898550724638,
+      "grad_norm": 0.29832305479155297,
+      "learning_rate": 2.1846484165324746e-05,
+      "loss": 0.5476,
+      "step": 1256
+    },
+    {
+      "epoch": 1.8217391304347825,
+      "grad_norm": 0.263630806640635,
+      "learning_rate": 2.1819645732689212e-05,
+      "loss": 0.4291,
+      "step": 1257
+    },
+    {
+      "epoch": 1.8231884057971013,
+      "grad_norm": 0.2851339779626431,
+      "learning_rate": 2.1792807300053678e-05,
+      "loss": 0.5007,
+      "step": 1258
+    },
+    {
+      "epoch": 1.8246376811594203,
+      "grad_norm": 0.28813505570565895,
+      "learning_rate": 2.1765968867418144e-05,
+      "loss": 0.5072,
+      "step": 1259
+    },
+    {
+      "epoch": 1.8260869565217392,
+      "grad_norm": 0.2855298396602445,
+      "learning_rate": 2.173913043478261e-05,
+      "loss": 0.5819,
+      "step": 1260
+    },
+    {
+      "epoch": 1.827536231884058,
+      "grad_norm": 0.29281428576959717,
+      "learning_rate": 2.1712292002147076e-05,
+      "loss": 0.5055,
+      "step": 1261
+    },
+    {
+      "epoch": 1.8289855072463768,
+      "grad_norm": 0.3124420934472098,
+      "learning_rate": 2.168545356951154e-05,
+      "loss": 0.4743,
+      "step": 1262
+    },
+    {
+      "epoch": 1.8304347826086955,
+      "grad_norm": 0.28710424131737194,
+      "learning_rate": 2.1658615136876005e-05,
+      "loss": 0.4775,
+      "step": 1263
+    },
+    {
+      "epoch": 1.8318840579710145,
+      "grad_norm": 0.2827625651958568,
+      "learning_rate": 2.163177670424047e-05,
+      "loss": 0.4919,
+      "step": 1264
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 0.2853579763416708,
+      "learning_rate": 2.1604938271604937e-05,
+      "loss": 0.5057,
+      "step": 1265
+    },
+    {
+      "epoch": 1.8347826086956522,
+      "grad_norm": 0.3012843262516302,
+      "learning_rate": 2.1578099838969404e-05,
+      "loss": 0.5333,
+      "step": 1266
+    },
+    {
+      "epoch": 1.836231884057971,
+      "grad_norm": 0.3065514320478756,
+      "learning_rate": 2.1551261406333873e-05,
+      "loss": 0.5173,
+      "step": 1267
+    },
+    {
+      "epoch": 1.8376811594202898,
+      "grad_norm": 0.26727540883895357,
+      "learning_rate": 2.152442297369834e-05,
+      "loss": 0.4248,
+      "step": 1268
+    },
+    {
+      "epoch": 1.8391304347826087,
+      "grad_norm": 0.29562518431891094,
+      "learning_rate": 2.1497584541062805e-05,
+      "loss": 0.469,
+      "step": 1269
+    },
+    {
+      "epoch": 1.8405797101449275,
+      "grad_norm": 0.298575851524304,
+      "learning_rate": 2.147074610842727e-05,
+      "loss": 0.4467,
+      "step": 1270
+    },
+    {
+      "epoch": 1.8420289855072465,
+      "grad_norm": 1.0364992953210546,
+      "learning_rate": 2.1443907675791734e-05,
+      "loss": 0.5137,
+      "step": 1271
+    },
+    {
+      "epoch": 1.8434782608695652,
+      "grad_norm": 0.31898501485381864,
+      "learning_rate": 2.14170692431562e-05,
+      "loss": 0.4855,
+      "step": 1272
+    },
+    {
+      "epoch": 1.844927536231884,
+      "grad_norm": 0.3045897927125765,
+      "learning_rate": 2.1390230810520666e-05,
+      "loss": 0.4238,
+      "step": 1273
+    },
+    {
+      "epoch": 1.8463768115942027,
+      "grad_norm": 0.3166686935868539,
+      "learning_rate": 2.1363392377885133e-05,
+      "loss": 0.5035,
+      "step": 1274
+    },
+    {
+      "epoch": 1.8478260869565217,
+      "grad_norm": 0.2688013508113102,
+      "learning_rate": 2.13365539452496e-05,
+      "loss": 0.4263,
+      "step": 1275
+    },
+    {
+      "epoch": 1.8492753623188407,
+      "grad_norm": 0.304899379478993,
+      "learning_rate": 2.1309715512614065e-05,
+      "loss": 0.4904,
+      "step": 1276
+    },
+    {
+      "epoch": 1.8507246376811595,
+      "grad_norm": 0.35844167840160257,
+      "learning_rate": 2.128287707997853e-05,
+      "loss": 0.4878,
+      "step": 1277
+    },
+    {
+      "epoch": 1.8521739130434782,
+      "grad_norm": 0.27706401361248995,
+      "learning_rate": 2.1256038647342997e-05,
+      "loss": 0.4274,
+      "step": 1278
+    },
+    {
+      "epoch": 1.853623188405797,
+      "grad_norm": 0.6396040828561423,
+      "learning_rate": 2.1229200214707463e-05,
+      "loss": 0.4881,
+      "step": 1279
+    },
+    {
+      "epoch": 1.855072463768116,
+      "grad_norm": 0.3127894695259788,
+      "learning_rate": 2.120236178207193e-05,
+      "loss": 0.4723,
+      "step": 1280
+    },
+    {
+      "epoch": 1.856521739130435,
+      "grad_norm": 0.29512152030072514,
+      "learning_rate": 2.1175523349436392e-05,
+      "loss": 0.4736,
+      "step": 1281
+    },
+    {
+      "epoch": 1.8579710144927537,
+      "grad_norm": 0.32603319773420636,
+      "learning_rate": 2.1148684916800858e-05,
+      "loss": 0.5158,
+      "step": 1282
+    },
+    {
+      "epoch": 1.8594202898550725,
+      "grad_norm": 0.31299453677117295,
+      "learning_rate": 2.1121846484165324e-05,
+      "loss": 0.5014,
+      "step": 1283
+    },
+    {
+      "epoch": 1.8608695652173912,
+      "grad_norm": 0.3041031362582152,
+      "learning_rate": 2.109500805152979e-05,
+      "loss": 0.4746,
+      "step": 1284
+    },
+    {
+      "epoch": 1.8623188405797102,
+      "grad_norm": 0.2754178261200115,
+      "learning_rate": 2.1068169618894257e-05,
+      "loss": 0.4237,
+      "step": 1285
+    },
+    {
+      "epoch": 1.863768115942029,
+      "grad_norm": 0.292174368122742,
+      "learning_rate": 2.1041331186258723e-05,
+      "loss": 0.5038,
+      "step": 1286
+    },
+    {
+      "epoch": 1.865217391304348,
+      "grad_norm": 0.3138925305688222,
+      "learning_rate": 2.101449275362319e-05,
+      "loss": 0.531,
+      "step": 1287
+    },
+    {
+      "epoch": 1.8666666666666667,
+      "grad_norm": 0.2920449919714151,
+      "learning_rate": 2.0987654320987655e-05,
+      "loss": 0.4655,
+      "step": 1288
+    },
+    {
+      "epoch": 1.8681159420289855,
+      "grad_norm": 0.2774442013795414,
+      "learning_rate": 2.096081588835212e-05,
+      "loss": 0.4613,
+      "step": 1289
+    },
+    {
+      "epoch": 1.8695652173913042,
+      "grad_norm": 0.29089291776029375,
+      "learning_rate": 2.0933977455716587e-05,
+      "loss": 0.4932,
+      "step": 1290
+    },
+    {
+      "epoch": 1.8710144927536232,
+      "grad_norm": 0.3568060057258649,
+      "learning_rate": 2.0907139023081053e-05,
+      "loss": 0.4316,
+      "step": 1291
+    },
+    {
+      "epoch": 1.8724637681159422,
+      "grad_norm": 0.347097941623319,
+      "learning_rate": 2.088030059044552e-05,
+      "loss": 0.4974,
+      "step": 1292
+    },
+    {
+      "epoch": 1.873913043478261,
+      "grad_norm": 0.3433506640776097,
+      "learning_rate": 2.0853462157809986e-05,
+      "loss": 0.5941,
+      "step": 1293
+    },
+    {
+      "epoch": 1.8753623188405797,
+      "grad_norm": 0.36194743289211745,
+      "learning_rate": 2.0826623725174452e-05,
+      "loss": 0.5001,
+      "step": 1294
+    },
+    {
+      "epoch": 1.8768115942028984,
+      "grad_norm": 0.3960985897096472,
+      "learning_rate": 2.0799785292538918e-05,
+      "loss": 0.4855,
+      "step": 1295
+    },
+    {
+      "epoch": 1.8782608695652174,
+      "grad_norm": 0.3526409226448084,
+      "learning_rate": 2.0772946859903384e-05,
+      "loss": 0.5259,
+      "step": 1296
+    },
+    {
+      "epoch": 1.8797101449275362,
+      "grad_norm": 0.32968502854409054,
+      "learning_rate": 2.074610842726785e-05,
+      "loss": 0.4089,
+      "step": 1297
+    },
+    {
+      "epoch": 1.8811594202898552,
+      "grad_norm": 0.31356361732084437,
+      "learning_rate": 2.0719269994632316e-05,
+      "loss": 0.5163,
+      "step": 1298
+    },
+    {
+      "epoch": 1.882608695652174,
+      "grad_norm": 0.3427992740512394,
+      "learning_rate": 2.0692431561996782e-05,
+      "loss": 0.5196,
+      "step": 1299
+    },
+    {
+      "epoch": 1.8840579710144927,
+      "grad_norm": 0.31325906176651624,
+      "learning_rate": 2.0665593129361245e-05,
+      "loss": 0.4887,
+      "step": 1300
+    },
+    {
+      "epoch": 1.8855072463768114,
+      "grad_norm": 0.2962587958372876,
+      "learning_rate": 2.063875469672571e-05,
+      "loss": 0.4972,
+      "step": 1301
+    },
+    {
+      "epoch": 1.8869565217391304,
+      "grad_norm": 0.32768267756192265,
+      "learning_rate": 2.0611916264090177e-05,
+      "loss": 0.4018,
+      "step": 1302
+    },
+    {
+      "epoch": 1.8884057971014494,
+      "grad_norm": 0.33194347131378105,
+      "learning_rate": 2.0585077831454643e-05,
+      "loss": 0.499,
+      "step": 1303
+    },
+    {
+      "epoch": 1.8898550724637682,
+      "grad_norm": 0.2932119296945969,
+      "learning_rate": 2.055823939881911e-05,
+      "loss": 0.4639,
+      "step": 1304
+    },
+    {
+      "epoch": 1.891304347826087,
+      "grad_norm": 0.3764989655928959,
+      "learning_rate": 2.0531400966183576e-05,
+      "loss": 0.516,
+      "step": 1305
+    },
+    {
+      "epoch": 1.8927536231884057,
+      "grad_norm": 0.3531420090951196,
+      "learning_rate": 2.0504562533548042e-05,
+      "loss": 0.5185,
+      "step": 1306
+    },
+    {
+      "epoch": 1.8942028985507247,
+      "grad_norm": 0.3010580871678021,
+      "learning_rate": 2.0477724100912508e-05,
+      "loss": 0.4991,
+      "step": 1307
+    },
+    {
+      "epoch": 1.8956521739130436,
+      "grad_norm": 0.29344052606590904,
+      "learning_rate": 2.0450885668276974e-05,
+      "loss": 0.4724,
+      "step": 1308
+    },
+    {
+      "epoch": 1.8971014492753624,
+      "grad_norm": 0.34527331703901987,
+      "learning_rate": 2.042404723564144e-05,
+      "loss": 0.5225,
+      "step": 1309
+    },
+    {
+      "epoch": 1.8985507246376812,
+      "grad_norm": 0.3118341745147994,
+      "learning_rate": 2.0397208803005903e-05,
+      "loss": 0.5096,
+      "step": 1310
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.4212037823678205,
+      "learning_rate": 2.037037037037037e-05,
+      "loss": 0.4954,
+      "step": 1311
+    },
+    {
+      "epoch": 1.901449275362319,
+      "grad_norm": 1.186945577141863,
+      "learning_rate": 2.0343531937734835e-05,
+      "loss": 0.525,
+      "step": 1312
+    },
+    {
+      "epoch": 1.9028985507246376,
+      "grad_norm": 0.35995105362938973,
+      "learning_rate": 2.03166935050993e-05,
+      "loss": 0.4655,
+      "step": 1313
+    },
+    {
+      "epoch": 1.9043478260869566,
+      "grad_norm": 0.2803749141962748,
+      "learning_rate": 2.028985507246377e-05,
+      "loss": 0.4443,
+      "step": 1314
+    },
+    {
+      "epoch": 1.9057971014492754,
+      "grad_norm": 0.2810489649699971,
+      "learning_rate": 2.0263016639828237e-05,
+      "loss": 0.4722,
+      "step": 1315
+    },
+    {
+      "epoch": 1.9072463768115941,
+      "grad_norm": 0.30476564064352385,
+      "learning_rate": 2.0236178207192703e-05,
+      "loss": 0.4782,
+      "step": 1316
+    },
+    {
+      "epoch": 1.908695652173913,
+      "grad_norm": 0.26799442712957444,
+      "learning_rate": 2.020933977455717e-05,
+      "loss": 0.3711,
+      "step": 1317
+    },
+    {
+      "epoch": 1.9101449275362319,
+      "grad_norm": 0.30200296290988266,
+      "learning_rate": 2.0182501341921632e-05,
+      "loss": 0.4994,
+      "step": 1318
+    },
+    {
+      "epoch": 1.9115942028985509,
+      "grad_norm": 0.31939099104943003,
+      "learning_rate": 2.0155662909286098e-05,
+      "loss": 0.4946,
+      "step": 1319
+    },
+    {
+      "epoch": 1.9130434782608696,
+      "grad_norm": 0.3068851680782278,
+      "learning_rate": 2.0128824476650564e-05,
+      "loss": 0.4616,
+      "step": 1320
+    },
+    {
+      "epoch": 1.9144927536231884,
+      "grad_norm": 0.3023208269337483,
+      "learning_rate": 2.010198604401503e-05,
+      "loss": 0.4763,
+      "step": 1321
+    },
+    {
+      "epoch": 1.9159420289855071,
+      "grad_norm": 0.3024815227588282,
+      "learning_rate": 2.0075147611379496e-05,
+      "loss": 0.4892,
+      "step": 1322
+    },
+    {
+      "epoch": 1.9173913043478261,
+      "grad_norm": 0.32337045749778953,
+      "learning_rate": 2.0048309178743963e-05,
+      "loss": 0.4847,
+      "step": 1323
+    },
+    {
+      "epoch": 1.9188405797101449,
+      "grad_norm": 0.2775258100528873,
+      "learning_rate": 2.002147074610843e-05,
+      "loss": 0.447,
+      "step": 1324
+    },
+    {
+      "epoch": 1.9202898550724639,
+      "grad_norm": 0.2894689084026751,
+      "learning_rate": 1.9994632313472895e-05,
+      "loss": 0.4561,
+      "step": 1325
+    },
+    {
+      "epoch": 1.9217391304347826,
+      "grad_norm": 0.3136492116234664,
+      "learning_rate": 1.996779388083736e-05,
+      "loss": 0.513,
+      "step": 1326
+    },
+    {
+      "epoch": 1.9231884057971014,
+      "grad_norm": 0.25726562553145293,
+      "learning_rate": 1.9940955448201827e-05,
+      "loss": 0.4732,
+      "step": 1327
+    },
+    {
+      "epoch": 1.9246376811594201,
+      "grad_norm": 0.29313931534891746,
+      "learning_rate": 1.991411701556629e-05,
+      "loss": 0.4824,
+      "step": 1328
+    },
+    {
+      "epoch": 1.9260869565217391,
+      "grad_norm": 0.3245324762668572,
+      "learning_rate": 1.9887278582930756e-05,
+      "loss": 0.4719,
+      "step": 1329
+    },
+    {
+      "epoch": 1.927536231884058,
+      "grad_norm": 0.2825614921402087,
+      "learning_rate": 1.9860440150295222e-05,
+      "loss": 0.4607,
+      "step": 1330
+    },
+    {
+      "epoch": 1.9289855072463769,
+      "grad_norm": 0.26485872541861677,
+      "learning_rate": 1.9833601717659688e-05,
+      "loss": 0.4902,
+      "step": 1331
+    },
+    {
+      "epoch": 1.9304347826086956,
+      "grad_norm": 0.28708767044003036,
+      "learning_rate": 1.9806763285024154e-05,
+      "loss": 0.464,
+      "step": 1332
+    },
+    {
+      "epoch": 1.9318840579710144,
+      "grad_norm": 0.29675243481316566,
+      "learning_rate": 1.977992485238862e-05,
+      "loss": 0.5058,
+      "step": 1333
+    },
+    {
+      "epoch": 1.9333333333333333,
+      "grad_norm": 0.2835769832589822,
+      "learning_rate": 1.9753086419753087e-05,
+      "loss": 0.5023,
+      "step": 1334
+    },
+    {
+      "epoch": 1.9347826086956523,
+      "grad_norm": 0.3843873450170395,
+      "learning_rate": 1.9726247987117553e-05,
+      "loss": 0.4539,
+      "step": 1335
+    },
+    {
+      "epoch": 1.936231884057971,
+      "grad_norm": 0.29259710654659776,
+      "learning_rate": 1.969940955448202e-05,
+      "loss": 0.4876,
+      "step": 1336
+    },
+    {
+      "epoch": 1.9376811594202898,
+      "grad_norm": 0.2901300055132593,
+      "learning_rate": 1.9672571121846485e-05,
+      "loss": 0.5418,
+      "step": 1337
+    },
+    {
+      "epoch": 1.9391304347826086,
+      "grad_norm": 0.2719106341759771,
+      "learning_rate": 1.964573268921095e-05,
+      "loss": 0.4376,
+      "step": 1338
+    },
+    {
+      "epoch": 1.9405797101449276,
+      "grad_norm": 0.2840996716380229,
+      "learning_rate": 1.9618894256575417e-05,
+      "loss": 0.5078,
+      "step": 1339
+    },
+    {
+      "epoch": 1.9420289855072463,
+      "grad_norm": 5.946655975501915,
+      "learning_rate": 1.9592055823939883e-05,
+      "loss": 0.6071,
+      "step": 1340
+    },
+    {
+      "epoch": 1.9434782608695653,
+      "grad_norm": 0.3083209392903689,
+      "learning_rate": 1.956521739130435e-05,
+      "loss": 0.476,
+      "step": 1341
+    },
+    {
+      "epoch": 1.944927536231884,
+      "grad_norm": 0.3189040042298081,
+      "learning_rate": 1.9538378958668816e-05,
+      "loss": 0.4886,
+      "step": 1342
+    },
+    {
+      "epoch": 1.9463768115942028,
+      "grad_norm": 0.2920370744669305,
+      "learning_rate": 1.9511540526033282e-05,
+      "loss": 0.4871,
+      "step": 1343
+    },
+    {
+      "epoch": 1.9478260869565216,
+      "grad_norm": 0.30710542670229024,
+      "learning_rate": 1.9484702093397748e-05,
+      "loss": 0.5296,
+      "step": 1344
+    },
+    {
+      "epoch": 1.9492753623188406,
+      "grad_norm": 0.33290375374662645,
+      "learning_rate": 1.9457863660762214e-05,
+      "loss": 0.5039,
+      "step": 1345
+    },
+    {
+      "epoch": 1.9507246376811596,
+      "grad_norm": 0.3201136657518179,
+      "learning_rate": 1.943102522812668e-05,
+      "loss": 0.5562,
+      "step": 1346
+    },
+    {
+      "epoch": 1.9521739130434783,
+      "grad_norm": 0.2834790046306402,
+      "learning_rate": 1.9404186795491143e-05,
+      "loss": 0.5026,
+      "step": 1347
+    },
+    {
+      "epoch": 1.953623188405797,
+      "grad_norm": 0.3023987772650251,
+      "learning_rate": 1.937734836285561e-05,
+      "loss": 0.4607,
+      "step": 1348
+    },
+    {
+      "epoch": 1.9550724637681158,
+      "grad_norm": 0.61842500520766,
+      "learning_rate": 1.9350509930220075e-05,
+      "loss": 0.5083,
+      "step": 1349
+    },
+    {
+      "epoch": 1.9565217391304348,
+      "grad_norm": 0.28431314772247096,
+      "learning_rate": 1.932367149758454e-05,
+      "loss": 0.4381,
+      "step": 1350
+    },
+    {
+      "epoch": 1.9579710144927536,
+      "grad_norm": 0.28922134943197386,
+      "learning_rate": 1.9296833064949007e-05,
+      "loss": 0.5042,
+      "step": 1351
+    },
+    {
+      "epoch": 1.9594202898550726,
+      "grad_norm": 0.3398952053798585,
+      "learning_rate": 1.9269994632313474e-05,
+      "loss": 0.5823,
+      "step": 1352
+    },
+    {
+      "epoch": 1.9608695652173913,
+      "grad_norm": 0.2933088409635406,
+      "learning_rate": 1.924315619967794e-05,
+      "loss": 0.5117,
+      "step": 1353
+    },
+    {
+      "epoch": 1.96231884057971,
+      "grad_norm": 0.28452221937531197,
+      "learning_rate": 1.9216317767042406e-05,
+      "loss": 0.5014,
+      "step": 1354
+    },
+    {
+      "epoch": 1.9637681159420288,
+      "grad_norm": 0.3431564358613733,
+      "learning_rate": 1.9189479334406872e-05,
+      "loss": 0.5004,
+      "step": 1355
+    },
+    {
+      "epoch": 1.9652173913043478,
+      "grad_norm": 0.29866438230282527,
+      "learning_rate": 1.9162640901771338e-05,
+      "loss": 0.4668,
+      "step": 1356
+    },
+    {
+      "epoch": 1.9666666666666668,
+      "grad_norm": 0.31282041209931233,
+      "learning_rate": 1.91358024691358e-05,
+      "loss": 0.478,
+      "step": 1357
+    },
+    {
+      "epoch": 1.9681159420289855,
+      "grad_norm": 0.27585050965807306,
+      "learning_rate": 1.9108964036500267e-05,
+      "loss": 0.458,
+      "step": 1358
+    },
+    {
+      "epoch": 1.9695652173913043,
+      "grad_norm": 0.29943407731775756,
+      "learning_rate": 1.9082125603864733e-05,
+      "loss": 0.4685,
+      "step": 1359
+    },
+    {
+      "epoch": 1.971014492753623,
+      "grad_norm": 0.36370754002131483,
+      "learning_rate": 1.9055287171229203e-05,
+      "loss": 0.4864,
+      "step": 1360
+    },
+    {
+      "epoch": 1.972463768115942,
+      "grad_norm": 0.2818369497611728,
+      "learning_rate": 1.902844873859367e-05,
+      "loss": 0.3647,
+      "step": 1361
+    },
+    {
+      "epoch": 1.973913043478261,
+      "grad_norm": 0.31444328219509243,
+      "learning_rate": 1.9001610305958135e-05,
+      "loss": 0.5593,
+      "step": 1362
+    },
+    {
+      "epoch": 1.9753623188405798,
+      "grad_norm": 0.29217505932655186,
+      "learning_rate": 1.89747718733226e-05,
+      "loss": 0.4606,
+      "step": 1363
+    },
+    {
+      "epoch": 1.9768115942028985,
+      "grad_norm": 0.2887298578180074,
+      "learning_rate": 1.8947933440687067e-05,
+      "loss": 0.4693,
+      "step": 1364
+    },
+    {
+      "epoch": 1.9782608695652173,
+      "grad_norm": 0.3007350027357869,
+      "learning_rate": 1.892109500805153e-05,
+      "loss": 0.4575,
+      "step": 1365
+    },
+    {
+      "epoch": 1.9797101449275363,
+      "grad_norm": 0.28022918917301265,
+      "learning_rate": 1.8894256575415996e-05,
+      "loss": 0.4412,
+      "step": 1366
+    },
+    {
+      "epoch": 1.981159420289855,
+      "grad_norm": 0.32612237980146064,
+      "learning_rate": 1.8867418142780462e-05,
+      "loss": 0.5018,
+      "step": 1367
+    },
+    {
+      "epoch": 1.982608695652174,
+      "grad_norm": 0.26476557741702605,
+      "learning_rate": 1.8840579710144928e-05,
+      "loss": 0.4294,
+      "step": 1368
+    },
+    {
+      "epoch": 1.9840579710144928,
+      "grad_norm": 0.28174585585869416,
+      "learning_rate": 1.8813741277509394e-05,
+      "loss": 0.4914,
+      "step": 1369
+    },
+    {
+      "epoch": 1.9855072463768115,
+      "grad_norm": 0.30361583383126906,
+      "learning_rate": 1.878690284487386e-05,
+      "loss": 0.4352,
+      "step": 1370
+    },
+    {
+      "epoch": 1.9869565217391303,
+      "grad_norm": 0.2649565863402794,
+      "learning_rate": 1.8760064412238327e-05,
+      "loss": 0.4491,
+      "step": 1371
+    },
+    {
+      "epoch": 1.9884057971014493,
+      "grad_norm": 0.2630797812371338,
+      "learning_rate": 1.8733225979602793e-05,
+      "loss": 0.4775,
+      "step": 1372
+    },
+    {
+      "epoch": 1.9898550724637682,
+      "grad_norm": 0.2858928894983428,
+      "learning_rate": 1.870638754696726e-05,
+      "loss": 0.466,
+      "step": 1373
+    },
+    {
+      "epoch": 1.991304347826087,
+      "grad_norm": 0.3132053561794151,
+      "learning_rate": 1.8679549114331725e-05,
+      "loss": 0.5051,
+      "step": 1374
+    },
+    {
+      "epoch": 1.9927536231884058,
+      "grad_norm": 0.288525312176386,
+      "learning_rate": 1.865271068169619e-05,
+      "loss": 0.5381,
+      "step": 1375
+    },
+    {
+      "epoch": 1.9942028985507245,
+      "grad_norm": 0.27387962716803127,
+      "learning_rate": 1.8625872249060654e-05,
+      "loss": 0.4684,
+      "step": 1376
+    },
+    {
+      "epoch": 1.9956521739130435,
+      "grad_norm": 0.28738312103663016,
+      "learning_rate": 1.859903381642512e-05,
+      "loss": 0.4264,
+      "step": 1377
+    },
+    {
+      "epoch": 1.9971014492753625,
+      "grad_norm": 0.3556560210674581,
+      "learning_rate": 1.8572195383789586e-05,
+      "loss": 0.5116,
+      "step": 1378
+    },
+    {
+      "epoch": 1.9985507246376812,
+      "grad_norm": 0.3239364265662191,
+      "learning_rate": 1.8545356951154052e-05,
+      "loss": 0.524,
+      "step": 1379
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2831017590599986,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.4528,
+      "step": 1380
+    },
+    {
+      "epoch": 2.0014492753623188,
+      "grad_norm": 0.3523786241603801,
+      "learning_rate": 1.8491680085882984e-05,
+      "loss": 0.3702,
+      "step": 1381
+    },
+    {
+      "epoch": 2.0028985507246375,
+      "grad_norm": 0.3030663523734319,
+      "learning_rate": 1.846484165324745e-05,
+      "loss": 0.3364,
+      "step": 1382
+    },
+    {
+      "epoch": 2.0043478260869567,
+      "grad_norm": 0.3140992539468059,
+      "learning_rate": 1.8438003220611917e-05,
+      "loss": 0.3604,
+      "step": 1383
+    },
+    {
+      "epoch": 2.0057971014492755,
+      "grad_norm": 0.33703359223943585,
+      "learning_rate": 1.8411164787976383e-05,
+      "loss": 0.3995,
+      "step": 1384
+    },
+    {
+      "epoch": 2.0072463768115942,
+      "grad_norm": 0.31543966911381965,
+      "learning_rate": 1.838432635534085e-05,
+      "loss": 0.398,
+      "step": 1385
+    },
+    {
+      "epoch": 2.008695652173913,
+      "grad_norm": 0.3162075337562776,
+      "learning_rate": 1.8357487922705315e-05,
+      "loss": 0.3711,
+      "step": 1386
+    },
+    {
+      "epoch": 2.0101449275362318,
+      "grad_norm": 0.3315543853764206,
+      "learning_rate": 1.833064949006978e-05,
+      "loss": 0.3733,
+      "step": 1387
+    },
+    {
+      "epoch": 2.0115942028985505,
+      "grad_norm": 0.3562957815235896,
+      "learning_rate": 1.8303811057434247e-05,
+      "loss": 0.3979,
+      "step": 1388
+    },
+    {
+      "epoch": 2.0130434782608697,
+      "grad_norm": 0.28341374991032275,
+      "learning_rate": 1.8276972624798713e-05,
+      "loss": 0.3344,
+      "step": 1389
+    },
+    {
+      "epoch": 2.0144927536231885,
+      "grad_norm": 0.2802899932244061,
+      "learning_rate": 1.825013419216318e-05,
+      "loss": 0.3477,
+      "step": 1390
+    },
+    {
+      "epoch": 2.0159420289855072,
+      "grad_norm": 0.3424385111997662,
+      "learning_rate": 1.8223295759527646e-05,
+      "loss": 0.404,
+      "step": 1391
+    },
+    {
+      "epoch": 2.017391304347826,
+      "grad_norm": 0.2984055315815846,
+      "learning_rate": 1.8196457326892112e-05,
+      "loss": 0.3632,
+      "step": 1392
+    },
+    {
+      "epoch": 2.0188405797101447,
+      "grad_norm": 0.2983097121531309,
+      "learning_rate": 1.8169618894256578e-05,
+      "loss": 0.3756,
+      "step": 1393
+    },
+    {
+      "epoch": 2.020289855072464,
+      "grad_norm": 0.27573986873525247,
+      "learning_rate": 1.814278046162104e-05,
+      "loss": 0.4272,
+      "step": 1394
+    },
+    {
+      "epoch": 2.0217391304347827,
+      "grad_norm": 0.3037230522518841,
+      "learning_rate": 1.8115942028985507e-05,
+      "loss": 0.3596,
+      "step": 1395
+    },
+    {
+      "epoch": 2.0231884057971015,
+      "grad_norm": 0.26797546209215395,
+      "learning_rate": 1.8089103596349973e-05,
+      "loss": 0.3673,
+      "step": 1396
+    },
+    {
+      "epoch": 2.02463768115942,
+      "grad_norm": 0.33238669799496534,
+      "learning_rate": 1.806226516371444e-05,
+      "loss": 0.3727,
+      "step": 1397
+    },
+    {
+      "epoch": 2.026086956521739,
+      "grad_norm": 0.31405379508560144,
+      "learning_rate": 1.8035426731078905e-05,
+      "loss": 0.4106,
+      "step": 1398
+    },
+    {
+      "epoch": 2.027536231884058,
+      "grad_norm": 0.24931890950961952,
+      "learning_rate": 1.800858829844337e-05,
+      "loss": 0.3413,
+      "step": 1399
+    },
+    {
+      "epoch": 2.028985507246377,
+      "grad_norm": 0.2757289599197962,
+      "learning_rate": 1.7981749865807837e-05,
+      "loss": 0.405,
+      "step": 1400
+    },
+    {
+      "epoch": 2.0304347826086957,
+      "grad_norm": 0.265334942485615,
+      "learning_rate": 1.7954911433172304e-05,
+      "loss": 0.3478,
+      "step": 1401
+    },
+    {
+      "epoch": 2.0318840579710145,
+      "grad_norm": 0.27009414571202384,
+      "learning_rate": 1.792807300053677e-05,
+      "loss": 0.3809,
+      "step": 1402
+    },
+    {
+      "epoch": 2.033333333333333,
+      "grad_norm": 1.055881132874562,
+      "learning_rate": 1.7901234567901236e-05,
+      "loss": 0.3555,
+      "step": 1403
+    },
+    {
+      "epoch": 2.034782608695652,
+      "grad_norm": 0.3399683985781049,
+      "learning_rate": 1.78743961352657e-05,
+      "loss": 0.3936,
+      "step": 1404
+    },
+    {
+      "epoch": 2.036231884057971,
+      "grad_norm": 0.38978587070503584,
+      "learning_rate": 1.7847557702630165e-05,
+      "loss": 0.4131,
+      "step": 1405
+    },
+    {
+      "epoch": 2.03768115942029,
+      "grad_norm": 0.3323719722356293,
+      "learning_rate": 1.782071926999463e-05,
+      "loss": 0.3929,
+      "step": 1406
+    },
+    {
+      "epoch": 2.0391304347826087,
+      "grad_norm": 0.3496319650720124,
+      "learning_rate": 1.77938808373591e-05,
+      "loss": 0.3974,
+      "step": 1407
+    },
+    {
+      "epoch": 2.0405797101449274,
+      "grad_norm": 0.29021047672453953,
+      "learning_rate": 1.7767042404723566e-05,
+      "loss": 0.4147,
+      "step": 1408
+    },
+    {
+      "epoch": 2.042028985507246,
+      "grad_norm": 0.3746794123510468,
+      "learning_rate": 1.7740203972088033e-05,
+      "loss": 0.4319,
+      "step": 1409
+    },
+    {
+      "epoch": 2.0434782608695654,
+      "grad_norm": 0.32112065217197794,
+      "learning_rate": 1.77133655394525e-05,
+      "loss": 0.3461,
+      "step": 1410
+    },
+    {
+      "epoch": 2.044927536231884,
+      "grad_norm": 0.32295608409620047,
+      "learning_rate": 1.7686527106816965e-05,
+      "loss": 0.348,
+      "step": 1411
+    },
+    {
+      "epoch": 2.046376811594203,
+      "grad_norm": 0.2877795513155748,
+      "learning_rate": 1.765968867418143e-05,
+      "loss": 0.3937,
+      "step": 1412
+    },
+    {
+      "epoch": 2.0478260869565217,
+      "grad_norm": 0.3269016501234976,
+      "learning_rate": 1.7632850241545894e-05,
+      "loss": 0.4032,
+      "step": 1413
+    },
+    {
+      "epoch": 2.0492753623188404,
+      "grad_norm": 0.3065874525483955,
+      "learning_rate": 1.760601180891036e-05,
+      "loss": 0.3697,
+      "step": 1414
+    },
+    {
+      "epoch": 2.050724637681159,
+      "grad_norm": 0.30251777186615336,
+      "learning_rate": 1.7579173376274826e-05,
+      "loss": 0.4002,
+      "step": 1415
+    },
+    {
+      "epoch": 2.0521739130434784,
+      "grad_norm": 0.2923862896542669,
+      "learning_rate": 1.7552334943639292e-05,
+      "loss": 0.4159,
+      "step": 1416
+    },
+    {
+      "epoch": 2.053623188405797,
+      "grad_norm": 0.2699902274060441,
+      "learning_rate": 1.7525496511003758e-05,
+      "loss": 0.4173,
+      "step": 1417
+    },
+    {
+      "epoch": 2.055072463768116,
+      "grad_norm": 0.2904906346194216,
+      "learning_rate": 1.7498658078368224e-05,
+      "loss": 0.3799,
+      "step": 1418
+    },
+    {
+      "epoch": 2.0565217391304347,
+      "grad_norm": 0.2648593871130252,
+      "learning_rate": 1.747181964573269e-05,
+      "loss": 0.3918,
+      "step": 1419
+    },
+    {
+      "epoch": 2.0579710144927534,
+      "grad_norm": 0.27163514907233427,
+      "learning_rate": 1.7444981213097157e-05,
+      "loss": 0.37,
+      "step": 1420
+    },
+    {
+      "epoch": 2.0594202898550726,
+      "grad_norm": 0.25859866053020986,
+      "learning_rate": 1.7418142780461623e-05,
+      "loss": 0.4068,
+      "step": 1421
+    },
+    {
+      "epoch": 2.0608695652173914,
+      "grad_norm": 0.26762775078295287,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 0.4238,
+      "step": 1422
+    },
+    {
+      "epoch": 2.06231884057971,
+      "grad_norm": 0.2647847770956608,
+      "learning_rate": 1.736446591519055e-05,
+      "loss": 0.3958,
+      "step": 1423
+    },
+    {
+      "epoch": 2.063768115942029,
+      "grad_norm": 0.24667071694024753,
+      "learning_rate": 1.7337627482555018e-05,
+      "loss": 0.3443,
+      "step": 1424
+    },
+    {
+      "epoch": 2.0652173913043477,
+      "grad_norm": 0.27158923135252777,
+      "learning_rate": 1.7310789049919484e-05,
+      "loss": 0.3871,
+      "step": 1425
+    },
+    {
+      "epoch": 2.066666666666667,
+      "grad_norm": 0.2566345926387272,
+      "learning_rate": 1.728395061728395e-05,
+      "loss": 0.4334,
+      "step": 1426
+    },
+    {
+      "epoch": 2.0681159420289856,
+      "grad_norm": 0.24107253116295427,
+      "learning_rate": 1.7257112184648416e-05,
+      "loss": 0.4235,
+      "step": 1427
+    },
+    {
+      "epoch": 2.0695652173913044,
+      "grad_norm": 0.28099353675057254,
+      "learning_rate": 1.7230273752012882e-05,
+      "loss": 0.3238,
+      "step": 1428
+    },
+    {
+      "epoch": 2.071014492753623,
+      "grad_norm": 0.3208249477068484,
+      "learning_rate": 1.720343531937735e-05,
+      "loss": 0.4569,
+      "step": 1429
+    },
+    {
+      "epoch": 2.072463768115942,
+      "grad_norm": 0.2280711164683154,
+      "learning_rate": 1.7176596886741818e-05,
+      "loss": 0.3728,
+      "step": 1430
+    },
+    {
+      "epoch": 2.0739130434782607,
+      "grad_norm": 0.24172226721506201,
+      "learning_rate": 1.714975845410628e-05,
+      "loss": 0.4002,
+      "step": 1431
+    },
+    {
+      "epoch": 2.07536231884058,
+      "grad_norm": 0.2815390040260254,
+      "learning_rate": 1.7122920021470747e-05,
+      "loss": 0.3797,
+      "step": 1432
+    },
+    {
+      "epoch": 2.0768115942028986,
+      "grad_norm": 0.24636656338577184,
+      "learning_rate": 1.7096081588835213e-05,
+      "loss": 0.3533,
+      "step": 1433
+    },
+    {
+      "epoch": 2.0782608695652174,
+      "grad_norm": 0.2716289213218912,
+      "learning_rate": 1.706924315619968e-05,
+      "loss": 0.3981,
+      "step": 1434
+    },
+    {
+      "epoch": 2.079710144927536,
+      "grad_norm": 0.25829679326857236,
+      "learning_rate": 1.7042404723564145e-05,
+      "loss": 0.4223,
+      "step": 1435
+    },
+    {
+      "epoch": 2.081159420289855,
+      "grad_norm": 0.250177425572541,
+      "learning_rate": 1.701556629092861e-05,
+      "loss": 0.3447,
+      "step": 1436
+    },
+    {
+      "epoch": 2.082608695652174,
+      "grad_norm": 0.2502529045762175,
+      "learning_rate": 1.6988727858293077e-05,
+      "loss": 0.363,
+      "step": 1437
+    },
+    {
+      "epoch": 2.084057971014493,
+      "grad_norm": 0.2447443764539991,
+      "learning_rate": 1.6961889425657543e-05,
+      "loss": 0.3834,
+      "step": 1438
+    },
+    {
+      "epoch": 2.0855072463768116,
+      "grad_norm": 0.24219777526195496,
+      "learning_rate": 1.693505099302201e-05,
+      "loss": 0.3257,
+      "step": 1439
+    },
+    {
+      "epoch": 2.0869565217391304,
+      "grad_norm": 0.2920526142203735,
+      "learning_rate": 1.6908212560386476e-05,
+      "loss": 0.4298,
+      "step": 1440
+    },
+    {
+      "epoch": 2.088405797101449,
+      "grad_norm": 0.2516657259945037,
+      "learning_rate": 1.6881374127750942e-05,
+      "loss": 0.3853,
+      "step": 1441
+    },
+    {
+      "epoch": 2.0898550724637683,
+      "grad_norm": 0.24165452159967626,
+      "learning_rate": 1.6854535695115405e-05,
+      "loss": 0.3824,
+      "step": 1442
+    },
+    {
+      "epoch": 2.091304347826087,
+      "grad_norm": 0.2355200461177853,
+      "learning_rate": 1.682769726247987e-05,
+      "loss": 0.3333,
+      "step": 1443
+    },
+    {
+      "epoch": 2.092753623188406,
+      "grad_norm": 0.24122064994896833,
+      "learning_rate": 1.6800858829844337e-05,
+      "loss": 0.3108,
+      "step": 1444
+    },
+    {
+      "epoch": 2.0942028985507246,
+      "grad_norm": 0.27102253831694195,
+      "learning_rate": 1.6774020397208803e-05,
+      "loss": 0.3869,
+      "step": 1445
+    },
+    {
+      "epoch": 2.0956521739130434,
+      "grad_norm": 0.23602739749485968,
+      "learning_rate": 1.674718196457327e-05,
+      "loss": 0.3993,
+      "step": 1446
+    },
+    {
+      "epoch": 2.097101449275362,
+      "grad_norm": 0.2807521284756467,
+      "learning_rate": 1.6720343531937735e-05,
+      "loss": 0.4124,
+      "step": 1447
+    },
+    {
+      "epoch": 2.0985507246376813,
+      "grad_norm": 0.23997496682279448,
+      "learning_rate": 1.66935050993022e-05,
+      "loss": 0.3634,
+      "step": 1448
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.23149144183564388,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.3466,
+      "step": 1449
+    },
+    {
+      "epoch": 2.101449275362319,
+      "grad_norm": 0.2451881745573423,
+      "learning_rate": 1.6639828234031134e-05,
+      "loss": 0.3046,
+      "step": 1450
+    },
+    {
+      "epoch": 2.1028985507246376,
+      "grad_norm": 0.24181484293881986,
+      "learning_rate": 1.66129898013956e-05,
+      "loss": 0.3777,
+      "step": 1451
+    },
+    {
+      "epoch": 2.1043478260869564,
+      "grad_norm": 0.26395665403756063,
+      "learning_rate": 1.6586151368760062e-05,
+      "loss": 0.3637,
+      "step": 1452
+    },
+    {
+      "epoch": 2.1057971014492756,
+      "grad_norm": 0.25149885689244855,
+      "learning_rate": 1.6559312936124532e-05,
+      "loss": 0.3668,
+      "step": 1453
+    },
+    {
+      "epoch": 2.1072463768115943,
+      "grad_norm": 0.24249920612235193,
+      "learning_rate": 1.6532474503488998e-05,
+      "loss": 0.3903,
+      "step": 1454
+    },
+    {
+      "epoch": 2.108695652173913,
+      "grad_norm": 0.26713497584114004,
+      "learning_rate": 1.6505636070853464e-05,
+      "loss": 0.3562,
+      "step": 1455
+    },
+    {
+      "epoch": 2.110144927536232,
+      "grad_norm": 0.2593799864084221,
+      "learning_rate": 1.647879763821793e-05,
+      "loss": 0.3559,
+      "step": 1456
+    },
+    {
+      "epoch": 2.1115942028985506,
+      "grad_norm": 0.2592694481669805,
+      "learning_rate": 1.6451959205582396e-05,
+      "loss": 0.4043,
+      "step": 1457
+    },
+    {
+      "epoch": 2.1130434782608694,
+      "grad_norm": 0.24758345963815817,
+      "learning_rate": 1.6425120772946863e-05,
+      "loss": 0.3778,
+      "step": 1458
+    },
+    {
+      "epoch": 2.1144927536231886,
+      "grad_norm": 0.26493518217950146,
+      "learning_rate": 1.639828234031133e-05,
+      "loss": 0.3574,
+      "step": 1459
+    },
+    {
+      "epoch": 2.1159420289855073,
+      "grad_norm": 0.25281018813456596,
+      "learning_rate": 1.637144390767579e-05,
+      "loss": 0.3922,
+      "step": 1460
+    },
+    {
+      "epoch": 2.117391304347826,
+      "grad_norm": 0.24534360749969875,
+      "learning_rate": 1.6344605475040258e-05,
+      "loss": 0.3379,
+      "step": 1461
+    },
+    {
+      "epoch": 2.118840579710145,
+      "grad_norm": 0.24638959833813115,
+      "learning_rate": 1.6317767042404724e-05,
+      "loss": 0.3538,
+      "step": 1462
+    },
+    {
+      "epoch": 2.1202898550724636,
+      "grad_norm": 0.24803895065658563,
+      "learning_rate": 1.629092860976919e-05,
+      "loss": 0.4084,
+      "step": 1463
+    },
+    {
+      "epoch": 2.121739130434783,
+      "grad_norm": 0.25908341056136924,
+      "learning_rate": 1.6264090177133656e-05,
+      "loss": 0.4083,
+      "step": 1464
+    },
+    {
+      "epoch": 2.1231884057971016,
+      "grad_norm": 0.23918352019175904,
+      "learning_rate": 1.6237251744498122e-05,
+      "loss": 0.379,
+      "step": 1465
+    },
+    {
+      "epoch": 2.1246376811594203,
+      "grad_norm": 0.2318776018543893,
+      "learning_rate": 1.6210413311862588e-05,
+      "loss": 0.3552,
+      "step": 1466
+    },
+    {
+      "epoch": 2.126086956521739,
+      "grad_norm": 0.46583670802882937,
+      "learning_rate": 1.6183574879227054e-05,
+      "loss": 0.3675,
+      "step": 1467
+    },
+    {
+      "epoch": 2.127536231884058,
+      "grad_norm": 0.2597134207234505,
+      "learning_rate": 1.615673644659152e-05,
+      "loss": 0.3349,
+      "step": 1468
+    },
+    {
+      "epoch": 2.1289855072463766,
+      "grad_norm": 0.24390171120238152,
+      "learning_rate": 1.6129898013955987e-05,
+      "loss": 0.3716,
+      "step": 1469
+    },
+    {
+      "epoch": 2.130434782608696,
+      "grad_norm": 0.24362666570186983,
+      "learning_rate": 1.610305958132045e-05,
+      "loss": 0.3634,
+      "step": 1470
+    },
+    {
+      "epoch": 2.1318840579710145,
+      "grad_norm": 0.2723467761379894,
+      "learning_rate": 1.6076221148684915e-05,
+      "loss": 0.3985,
+      "step": 1471
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 0.2431944553336726,
+      "learning_rate": 1.604938271604938e-05,
+      "loss": 0.3581,
+      "step": 1472
+    },
+    {
+      "epoch": 2.134782608695652,
+      "grad_norm": 0.26094641628999893,
+      "learning_rate": 1.6022544283413848e-05,
+      "loss": 0.372,
+      "step": 1473
+    },
+    {
+      "epoch": 2.136231884057971,
+      "grad_norm": 0.24313430833284463,
+      "learning_rate": 1.5995705850778314e-05,
+      "loss": 0.3398,
+      "step": 1474
+    },
+    {
+      "epoch": 2.13768115942029,
+      "grad_norm": 0.2560446345933897,
+      "learning_rate": 1.596886741814278e-05,
+      "loss": 0.3542,
+      "step": 1475
+    },
+    {
+      "epoch": 2.139130434782609,
+      "grad_norm": 0.24771966399779138,
+      "learning_rate": 1.5942028985507246e-05,
+      "loss": 0.373,
+      "step": 1476
+    },
+    {
+      "epoch": 2.1405797101449275,
+      "grad_norm": 0.6791987543987067,
+      "learning_rate": 1.5915190552871716e-05,
+      "loss": 0.3547,
+      "step": 1477
+    },
+    {
+      "epoch": 2.1420289855072463,
+      "grad_norm": 0.2649813056093659,
+      "learning_rate": 1.5888352120236182e-05,
+      "loss": 0.3889,
+      "step": 1478
+    },
+    {
+      "epoch": 2.143478260869565,
+      "grad_norm": 0.25422139002817895,
+      "learning_rate": 1.5861513687600644e-05,
+      "loss": 0.3968,
+      "step": 1479
+    },
+    {
+      "epoch": 2.1449275362318843,
+      "grad_norm": 0.2576624449637141,
+      "learning_rate": 1.583467525496511e-05,
+      "loss": 0.3593,
+      "step": 1480
+    },
+    {
+      "epoch": 2.146376811594203,
+      "grad_norm": 0.2586375234022925,
+      "learning_rate": 1.5807836822329577e-05,
+      "loss": 0.3671,
+      "step": 1481
+    },
+    {
+      "epoch": 2.1478260869565218,
+      "grad_norm": 0.2558240672741214,
+      "learning_rate": 1.5780998389694043e-05,
+      "loss": 0.4098,
+      "step": 1482
+    },
+    {
+      "epoch": 2.1492753623188405,
+      "grad_norm": 0.268688991057188,
+      "learning_rate": 1.575415995705851e-05,
+      "loss": 0.3933,
+      "step": 1483
+    },
+    {
+      "epoch": 2.1507246376811593,
+      "grad_norm": 0.2532381024335753,
+      "learning_rate": 1.5727321524422975e-05,
+      "loss": 0.3494,
+      "step": 1484
+    },
+    {
+      "epoch": 2.1521739130434785,
+      "grad_norm": 0.26098343354326276,
+      "learning_rate": 1.570048309178744e-05,
+      "loss": 0.3546,
+      "step": 1485
+    },
+    {
+      "epoch": 2.1536231884057973,
+      "grad_norm": 0.2891962630112305,
+      "learning_rate": 1.5673644659151907e-05,
+      "loss": 0.4239,
+      "step": 1486
+    },
+    {
+      "epoch": 2.155072463768116,
+      "grad_norm": 0.26439594268339023,
+      "learning_rate": 1.5646806226516374e-05,
+      "loss": 0.4137,
+      "step": 1487
+    },
+    {
+      "epoch": 2.1565217391304348,
+      "grad_norm": 0.2661719925783797,
+      "learning_rate": 1.561996779388084e-05,
+      "loss": 0.3996,
+      "step": 1488
+    },
+    {
+      "epoch": 2.1579710144927535,
+      "grad_norm": 0.2389028862134638,
+      "learning_rate": 1.5593129361245302e-05,
+      "loss": 0.3406,
+      "step": 1489
+    },
+    {
+      "epoch": 2.1594202898550723,
+      "grad_norm": 0.24546221102827662,
+      "learning_rate": 1.556629092860977e-05,
+      "loss": 0.3792,
+      "step": 1490
+    },
+    {
+      "epoch": 2.1608695652173915,
+      "grad_norm": 0.276109060092543,
+      "learning_rate": 1.5539452495974235e-05,
+      "loss": 0.4163,
+      "step": 1491
+    },
+    {
+      "epoch": 2.1623188405797102,
+      "grad_norm": 0.467588348904684,
+      "learning_rate": 1.55126140633387e-05,
+      "loss": 0.361,
+      "step": 1492
+    },
+    {
+      "epoch": 2.163768115942029,
+      "grad_norm": 0.23622254966412615,
+      "learning_rate": 1.5485775630703167e-05,
+      "loss": 0.3845,
+      "step": 1493
+    },
+    {
+      "epoch": 2.1652173913043478,
+      "grad_norm": 0.24695504994769543,
+      "learning_rate": 1.5458937198067633e-05,
+      "loss": 0.3272,
+      "step": 1494
+    },
+    {
+      "epoch": 2.1666666666666665,
+      "grad_norm": 0.25126118120410157,
+      "learning_rate": 1.54320987654321e-05,
+      "loss": 0.3526,
+      "step": 1495
+    },
+    {
+      "epoch": 2.1681159420289857,
+      "grad_norm": 0.26226328500893514,
+      "learning_rate": 1.5405260332796565e-05,
+      "loss": 0.4089,
+      "step": 1496
+    },
+    {
+      "epoch": 2.1695652173913045,
+      "grad_norm": 0.23389975075817726,
+      "learning_rate": 1.537842190016103e-05,
+      "loss": 0.3409,
+      "step": 1497
+    },
+    {
+      "epoch": 2.1710144927536232,
+      "grad_norm": 0.2615319879343272,
+      "learning_rate": 1.5351583467525498e-05,
+      "loss": 0.3646,
+      "step": 1498
+    },
+    {
+      "epoch": 2.172463768115942,
+      "grad_norm": 0.2708157191632523,
+      "learning_rate": 1.532474503488996e-05,
+      "loss": 0.3632,
+      "step": 1499
+    },
+    {
+      "epoch": 2.1739130434782608,
+      "grad_norm": 0.23612916073428386,
+      "learning_rate": 1.529790660225443e-05,
+      "loss": 0.3333,
+      "step": 1500
+    },
+    {
+      "epoch": 2.1753623188405795,
+      "grad_norm": 0.26253628983361604,
+      "learning_rate": 1.5271068169618896e-05,
+      "loss": 0.3783,
+      "step": 1501
+    },
+    {
+      "epoch": 2.1768115942028987,
+      "grad_norm": 0.2799278780036722,
+      "learning_rate": 1.5244229736983362e-05,
+      "loss": 0.39,
+      "step": 1502
+    },
+    {
+      "epoch": 2.1782608695652175,
+      "grad_norm": 2.923354758482103,
+      "learning_rate": 1.5217391304347828e-05,
+      "loss": 0.3859,
+      "step": 1503
+    },
+    {
+      "epoch": 2.1797101449275362,
+      "grad_norm": 0.26300827777912456,
+      "learning_rate": 1.5190552871712294e-05,
+      "loss": 0.3882,
+      "step": 1504
+    },
+    {
+      "epoch": 2.181159420289855,
+      "grad_norm": 0.24980607270012614,
+      "learning_rate": 1.5163714439076759e-05,
+      "loss": 0.4039,
+      "step": 1505
+    },
+    {
+      "epoch": 2.1826086956521737,
+      "grad_norm": 0.2650949294763254,
+      "learning_rate": 1.5136876006441225e-05,
+      "loss": 0.4225,
+      "step": 1506
+    },
+    {
+      "epoch": 2.184057971014493,
+      "grad_norm": 0.23582919001983096,
+      "learning_rate": 1.5110037573805691e-05,
+      "loss": 0.3915,
+      "step": 1507
+    },
+    {
+      "epoch": 2.1855072463768117,
+      "grad_norm": 0.26149592082357265,
+      "learning_rate": 1.5083199141170157e-05,
+      "loss": 0.4391,
+      "step": 1508
+    },
+    {
+      "epoch": 2.1869565217391305,
+      "grad_norm": 0.24956056460952147,
+      "learning_rate": 1.5056360708534623e-05,
+      "loss": 0.3973,
+      "step": 1509
+    },
+    {
+      "epoch": 2.1884057971014492,
+      "grad_norm": 0.2670527077121037,
+      "learning_rate": 1.5029522275899088e-05,
+      "loss": 0.3859,
+      "step": 1510
+    },
+    {
+      "epoch": 2.189855072463768,
+      "grad_norm": 0.2478647439410972,
+      "learning_rate": 1.5002683843263554e-05,
+      "loss": 0.3954,
+      "step": 1511
+    },
+    {
+      "epoch": 2.1913043478260867,
+      "grad_norm": 0.2683179076496102,
+      "learning_rate": 1.497584541062802e-05,
+      "loss": 0.415,
+      "step": 1512
+    },
+    {
+      "epoch": 2.192753623188406,
+      "grad_norm": 0.2594324410575354,
+      "learning_rate": 1.4949006977992486e-05,
+      "loss": 0.3551,
+      "step": 1513
+    },
+    {
+      "epoch": 2.1942028985507247,
+      "grad_norm": 0.2457851168205183,
+      "learning_rate": 1.4922168545356952e-05,
+      "loss": 0.4231,
+      "step": 1514
+    },
+    {
+      "epoch": 2.1956521739130435,
+      "grad_norm": 0.24177430563161123,
+      "learning_rate": 1.4895330112721417e-05,
+      "loss": 0.3922,
+      "step": 1515
+    },
+    {
+      "epoch": 2.197101449275362,
+      "grad_norm": 0.2572364642689413,
+      "learning_rate": 1.4868491680085883e-05,
+      "loss": 0.3742,
+      "step": 1516
+    },
+    {
+      "epoch": 2.198550724637681,
+      "grad_norm": 0.24100963020540228,
+      "learning_rate": 1.4841653247450349e-05,
+      "loss": 0.3837,
+      "step": 1517
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.119712217144526,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.4164,
+      "step": 1518
+    },
+    {
+      "epoch": 2.201449275362319,
+      "grad_norm": 0.2516637687665249,
+      "learning_rate": 1.4787976382179281e-05,
+      "loss": 0.3679,
+      "step": 1519
+    },
+    {
+      "epoch": 2.2028985507246377,
+      "grad_norm": 0.26976714186644757,
+      "learning_rate": 1.4761137949543746e-05,
+      "loss": 0.3729,
+      "step": 1520
+    },
+    {
+      "epoch": 2.2043478260869565,
+      "grad_norm": 0.26502840804982475,
+      "learning_rate": 1.4734299516908212e-05,
+      "loss": 0.444,
+      "step": 1521
+    },
+    {
+      "epoch": 2.205797101449275,
+      "grad_norm": 0.24414225976585754,
+      "learning_rate": 1.4707461084272678e-05,
+      "loss": 0.3969,
+      "step": 1522
+    },
+    {
+      "epoch": 2.207246376811594,
+      "grad_norm": 0.2337178526893521,
+      "learning_rate": 1.4680622651637147e-05,
+      "loss": 0.3732,
+      "step": 1523
+    },
+    {
+      "epoch": 2.208695652173913,
+      "grad_norm": 0.26253557351427015,
+      "learning_rate": 1.4653784219001612e-05,
+      "loss": 0.3613,
+      "step": 1524
+    },
+    {
+      "epoch": 2.210144927536232,
+      "grad_norm": 0.26389875956191183,
+      "learning_rate": 1.4626945786366078e-05,
+      "loss": 0.3975,
+      "step": 1525
+    },
+    {
+      "epoch": 2.2115942028985507,
+      "grad_norm": 0.28084385613675417,
+      "learning_rate": 1.4600107353730544e-05,
+      "loss": 0.3476,
+      "step": 1526
+    },
+    {
+      "epoch": 2.2130434782608694,
+      "grad_norm": 0.2646052280590228,
+      "learning_rate": 1.457326892109501e-05,
+      "loss": 0.3714,
+      "step": 1527
+    },
+    {
+      "epoch": 2.214492753623188,
+      "grad_norm": 0.2645379652903494,
+      "learning_rate": 1.4546430488459476e-05,
+      "loss": 0.3934,
+      "step": 1528
+    },
+    {
+      "epoch": 2.2159420289855074,
+      "grad_norm": 0.2666041230309001,
+      "learning_rate": 1.451959205582394e-05,
+      "loss": 0.4162,
+      "step": 1529
+    },
+    {
+      "epoch": 2.217391304347826,
+      "grad_norm": 0.25393841436995956,
+      "learning_rate": 1.4492753623188407e-05,
+      "loss": 0.3521,
+      "step": 1530
+    },
+    {
+      "epoch": 2.218840579710145,
+      "grad_norm": 2.5834333622726957,
+      "learning_rate": 1.4465915190552873e-05,
+      "loss": 0.3571,
+      "step": 1531
+    },
+    {
+      "epoch": 2.2202898550724637,
+      "grad_norm": 0.25785009466800646,
+      "learning_rate": 1.4439076757917339e-05,
+      "loss": 0.4149,
+      "step": 1532
+    },
+    {
+      "epoch": 2.2217391304347824,
+      "grad_norm": 0.24957707518993402,
+      "learning_rate": 1.4412238325281805e-05,
+      "loss": 0.374,
+      "step": 1533
+    },
+    {
+      "epoch": 2.2231884057971016,
+      "grad_norm": 0.23262952045687874,
+      "learning_rate": 1.438539989264627e-05,
+      "loss": 0.3882,
+      "step": 1534
+    },
+    {
+      "epoch": 2.2246376811594204,
+      "grad_norm": 0.25960049374696026,
+      "learning_rate": 1.4358561460010736e-05,
+      "loss": 0.3939,
+      "step": 1535
+    },
+    {
+      "epoch": 2.226086956521739,
+      "grad_norm": 0.2491894668049094,
+      "learning_rate": 1.4331723027375202e-05,
+      "loss": 0.3577,
+      "step": 1536
+    },
+    {
+      "epoch": 2.227536231884058,
+      "grad_norm": 0.25138391003811594,
+      "learning_rate": 1.4304884594739668e-05,
+      "loss": 0.3752,
+      "step": 1537
+    },
+    {
+      "epoch": 2.2289855072463767,
+      "grad_norm": 0.214850791994487,
+      "learning_rate": 1.4278046162104134e-05,
+      "loss": 0.3217,
+      "step": 1538
+    },
+    {
+      "epoch": 2.230434782608696,
+      "grad_norm": 0.26161881685821325,
+      "learning_rate": 1.4251207729468599e-05,
+      "loss": 0.3867,
+      "step": 1539
+    },
+    {
+      "epoch": 2.2318840579710146,
+      "grad_norm": 0.28080848285012555,
+      "learning_rate": 1.4224369296833065e-05,
+      "loss": 0.3626,
+      "step": 1540
+    },
+    {
+      "epoch": 2.2333333333333334,
+      "grad_norm": 0.25475653356537886,
+      "learning_rate": 1.419753086419753e-05,
+      "loss": 0.3469,
+      "step": 1541
+    },
+    {
+      "epoch": 2.234782608695652,
+      "grad_norm": 0.3712708204701495,
+      "learning_rate": 1.4170692431561997e-05,
+      "loss": 0.3743,
+      "step": 1542
+    },
+    {
+      "epoch": 2.236231884057971,
+      "grad_norm": 0.8361471019434635,
+      "learning_rate": 1.4143853998926463e-05,
+      "loss": 0.3946,
+      "step": 1543
+    },
+    {
+      "epoch": 2.2376811594202897,
+      "grad_norm": 0.31565988365410325,
+      "learning_rate": 1.4117015566290927e-05,
+      "loss": 0.3839,
+      "step": 1544
+    },
+    {
+      "epoch": 2.239130434782609,
+      "grad_norm": 0.3180964477239743,
+      "learning_rate": 1.4090177133655394e-05,
+      "loss": 0.3998,
+      "step": 1545
+    },
+    {
+      "epoch": 2.2405797101449276,
+      "grad_norm": 0.2595863556849321,
+      "learning_rate": 1.406333870101986e-05,
+      "loss": 0.3842,
+      "step": 1546
+    },
+    {
+      "epoch": 2.2420289855072464,
+      "grad_norm": 0.2300656737294017,
+      "learning_rate": 1.4036500268384328e-05,
+      "loss": 0.3791,
+      "step": 1547
+    },
+    {
+      "epoch": 2.243478260869565,
+      "grad_norm": 0.26630741953742765,
+      "learning_rate": 1.4009661835748794e-05,
+      "loss": 0.3565,
+      "step": 1548
+    },
+    {
+      "epoch": 2.244927536231884,
+      "grad_norm": 0.3449002614397516,
+      "learning_rate": 1.398282340311326e-05,
+      "loss": 0.356,
+      "step": 1549
+    },
+    {
+      "epoch": 2.246376811594203,
+      "grad_norm": 0.278493690085927,
+      "learning_rate": 1.3955984970477726e-05,
+      "loss": 0.3849,
+      "step": 1550
+    },
+    {
+      "epoch": 2.247826086956522,
+      "grad_norm": 0.30149468853130207,
+      "learning_rate": 1.3929146537842192e-05,
+      "loss": 0.4505,
+      "step": 1551
+    },
+    {
+      "epoch": 2.2492753623188406,
+      "grad_norm": 0.27048184170368406,
+      "learning_rate": 1.3902308105206656e-05,
+      "loss": 0.3985,
+      "step": 1552
+    },
+    {
+      "epoch": 2.2507246376811594,
+      "grad_norm": 0.2913064801887286,
+      "learning_rate": 1.3875469672571123e-05,
+      "loss": 0.3717,
+      "step": 1553
+    },
+    {
+      "epoch": 2.252173913043478,
+      "grad_norm": 0.2730993886762121,
+      "learning_rate": 1.3848631239935589e-05,
+      "loss": 0.3484,
+      "step": 1554
+    },
+    {
+      "epoch": 2.253623188405797,
+      "grad_norm": 0.29507804335844673,
+      "learning_rate": 1.3821792807300055e-05,
+      "loss": 0.3219,
+      "step": 1555
+    },
+    {
+      "epoch": 2.255072463768116,
+      "grad_norm": 0.28418547043948034,
+      "learning_rate": 1.3794954374664521e-05,
+      "loss": 0.3412,
+      "step": 1556
+    },
+    {
+      "epoch": 2.256521739130435,
+      "grad_norm": 0.23898122611394282,
+      "learning_rate": 1.3768115942028985e-05,
+      "loss": 0.3825,
+      "step": 1557
+    },
+    {
+      "epoch": 2.2579710144927536,
+      "grad_norm": 0.2966037675150378,
+      "learning_rate": 1.3741277509393452e-05,
+      "loss": 0.3391,
+      "step": 1558
+    },
+    {
+      "epoch": 2.2594202898550724,
+      "grad_norm": 0.336848326235953,
+      "learning_rate": 1.3714439076757918e-05,
+      "loss": 0.395,
+      "step": 1559
+    },
+    {
+      "epoch": 2.260869565217391,
+      "grad_norm": 0.2722507559170827,
+      "learning_rate": 1.3687600644122384e-05,
+      "loss": 0.3558,
+      "step": 1560
+    },
+    {
+      "epoch": 2.2623188405797103,
+      "grad_norm": 0.2668793089188747,
+      "learning_rate": 1.366076221148685e-05,
+      "loss": 0.3325,
+      "step": 1561
+    },
+    {
+      "epoch": 2.263768115942029,
+      "grad_norm": 0.31039322433546784,
+      "learning_rate": 1.3633923778851316e-05,
+      "loss": 0.3879,
+      "step": 1562
+    },
+    {
+      "epoch": 2.265217391304348,
+      "grad_norm": 0.288457375992248,
+      "learning_rate": 1.360708534621578e-05,
+      "loss": 0.3928,
+      "step": 1563
+    },
+    {
+      "epoch": 2.2666666666666666,
+      "grad_norm": 0.2735574240926222,
+      "learning_rate": 1.3580246913580247e-05,
+      "loss": 0.3802,
+      "step": 1564
+    },
+    {
+      "epoch": 2.2681159420289854,
+      "grad_norm": 0.26280010543390336,
+      "learning_rate": 1.3553408480944713e-05,
+      "loss": 0.3504,
+      "step": 1565
+    },
+    {
+      "epoch": 2.269565217391304,
+      "grad_norm": 0.3080823982213214,
+      "learning_rate": 1.3526570048309179e-05,
+      "loss": 0.3772,
+      "step": 1566
+    },
+    {
+      "epoch": 2.2710144927536233,
+      "grad_norm": 0.277292211551906,
+      "learning_rate": 1.3499731615673645e-05,
+      "loss": 0.4081,
+      "step": 1567
+    },
+    {
+      "epoch": 2.272463768115942,
+      "grad_norm": 0.26869130104232364,
+      "learning_rate": 1.347289318303811e-05,
+      "loss": 0.3448,
+      "step": 1568
+    },
+    {
+      "epoch": 2.273913043478261,
+      "grad_norm": 0.27781485973024744,
+      "learning_rate": 1.3446054750402576e-05,
+      "loss": 0.3928,
+      "step": 1569
+    },
+    {
+      "epoch": 2.2753623188405796,
+      "grad_norm": 0.24601016996456326,
+      "learning_rate": 1.3419216317767045e-05,
+      "loss": 0.3919,
+      "step": 1570
+    },
+    {
+      "epoch": 2.2768115942028984,
+      "grad_norm": 0.2442491835619423,
+      "learning_rate": 1.339237788513151e-05,
+      "loss": 0.3681,
+      "step": 1571
+    },
+    {
+      "epoch": 2.2782608695652176,
+      "grad_norm": 0.29939513174678234,
+      "learning_rate": 1.3365539452495976e-05,
+      "loss": 0.4182,
+      "step": 1572
+    },
+    {
+      "epoch": 2.2797101449275363,
+      "grad_norm": 0.2379937617351913,
+      "learning_rate": 1.3338701019860442e-05,
+      "loss": 0.3469,
+      "step": 1573
+    },
+    {
+      "epoch": 2.281159420289855,
+      "grad_norm": 0.24206655548265582,
+      "learning_rate": 1.3311862587224908e-05,
+      "loss": 0.3823,
+      "step": 1574
+    },
+    {
+      "epoch": 2.282608695652174,
+      "grad_norm": 0.25153264548901644,
+      "learning_rate": 1.3285024154589374e-05,
+      "loss": 0.3633,
+      "step": 1575
+    },
+    {
+      "epoch": 2.2840579710144926,
+      "grad_norm": 0.24838756048823832,
+      "learning_rate": 1.3258185721953838e-05,
+      "loss": 0.3752,
+      "step": 1576
+    },
+    {
+      "epoch": 2.2855072463768114,
+      "grad_norm": 0.24475452285064764,
+      "learning_rate": 1.3231347289318305e-05,
+      "loss": 0.3748,
+      "step": 1577
+    },
+    {
+      "epoch": 2.2869565217391306,
+      "grad_norm": 0.26345551133178485,
+      "learning_rate": 1.320450885668277e-05,
+      "loss": 0.347,
+      "step": 1578
+    },
+    {
+      "epoch": 2.2884057971014493,
+      "grad_norm": 0.24643703206359371,
+      "learning_rate": 1.3177670424047237e-05,
+      "loss": 0.3933,
+      "step": 1579
+    },
+    {
+      "epoch": 2.289855072463768,
+      "grad_norm": 0.2904257571389863,
+      "learning_rate": 1.3150831991411703e-05,
+      "loss": 0.3985,
+      "step": 1580
+    },
+    {
+      "epoch": 2.291304347826087,
+      "grad_norm": 0.23888677788284127,
+      "learning_rate": 1.3123993558776167e-05,
+      "loss": 0.3486,
+      "step": 1581
+    },
+    {
+      "epoch": 2.292753623188406,
+      "grad_norm": 0.25676575141902436,
+      "learning_rate": 1.3097155126140634e-05,
+      "loss": 0.3529,
+      "step": 1582
+    },
+    {
+      "epoch": 2.294202898550725,
+      "grad_norm": 0.2290381327720587,
+      "learning_rate": 1.30703166935051e-05,
+      "loss": 0.3446,
+      "step": 1583
+    },
+    {
+      "epoch": 2.2956521739130435,
+      "grad_norm": 0.2673558483575196,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 0.4187,
+      "step": 1584
+    },
+    {
+      "epoch": 2.2971014492753623,
+      "grad_norm": 0.2619657872949245,
+      "learning_rate": 1.3016639828234032e-05,
+      "loss": 0.4018,
+      "step": 1585
+    },
+    {
+      "epoch": 2.298550724637681,
+      "grad_norm": 0.2471540318096013,
+      "learning_rate": 1.2989801395598496e-05,
+      "loss": 0.3582,
+      "step": 1586
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.24938930140778784,
+      "learning_rate": 1.2962962962962962e-05,
+      "loss": 0.3113,
+      "step": 1587
+    },
+    {
+      "epoch": 2.301449275362319,
+      "grad_norm": 0.226686670452519,
+      "learning_rate": 1.2936124530327429e-05,
+      "loss": 0.3655,
+      "step": 1588
+    },
+    {
+      "epoch": 2.302898550724638,
+      "grad_norm": 0.2448061300560438,
+      "learning_rate": 1.2909286097691895e-05,
+      "loss": 0.3634,
+      "step": 1589
+    },
+    {
+      "epoch": 2.3043478260869565,
+      "grad_norm": 0.23174962185631323,
+      "learning_rate": 1.288244766505636e-05,
+      "loss": 0.3644,
+      "step": 1590
+    },
+    {
+      "epoch": 2.3057971014492753,
+      "grad_norm": 0.26870532656946267,
+      "learning_rate": 1.2855609232420825e-05,
+      "loss": 0.3919,
+      "step": 1591
+    },
+    {
+      "epoch": 2.307246376811594,
+      "grad_norm": 0.295247629024476,
+      "learning_rate": 1.2828770799785291e-05,
+      "loss": 0.3849,
+      "step": 1592
+    },
+    {
+      "epoch": 2.3086956521739133,
+      "grad_norm": 0.25583435866670345,
+      "learning_rate": 1.2801932367149761e-05,
+      "loss": 0.3998,
+      "step": 1593
+    },
+    {
+      "epoch": 2.310144927536232,
+      "grad_norm": 0.25026084065358445,
+      "learning_rate": 1.2775093934514227e-05,
+      "loss": 0.3571,
+      "step": 1594
+    },
+    {
+      "epoch": 2.3115942028985508,
+      "grad_norm": 0.2942700249747664,
+      "learning_rate": 1.2748255501878691e-05,
+      "loss": 0.3749,
+      "step": 1595
+    },
+    {
+      "epoch": 2.3130434782608695,
+      "grad_norm": 0.248980629991386,
+      "learning_rate": 1.2721417069243158e-05,
+      "loss": 0.384,
+      "step": 1596
+    },
+    {
+      "epoch": 2.3144927536231883,
+      "grad_norm": 0.2532058397112117,
+      "learning_rate": 1.2694578636607624e-05,
+      "loss": 0.3709,
+      "step": 1597
+    },
+    {
+      "epoch": 2.315942028985507,
+      "grad_norm": 0.26982970940029316,
+      "learning_rate": 1.266774020397209e-05,
+      "loss": 0.3548,
+      "step": 1598
+    },
+    {
+      "epoch": 2.3173913043478263,
+      "grad_norm": 0.255607765102706,
+      "learning_rate": 1.2640901771336556e-05,
+      "loss": 0.3663,
+      "step": 1599
+    },
+    {
+      "epoch": 2.318840579710145,
+      "grad_norm": 0.25522783943627514,
+      "learning_rate": 1.261406333870102e-05,
+      "loss": 0.373,
+      "step": 1600
+    },
+    {
+      "epoch": 2.3202898550724638,
+      "grad_norm": 0.23165148455281287,
+      "learning_rate": 1.2587224906065487e-05,
+      "loss": 0.3594,
+      "step": 1601
+    },
+    {
+      "epoch": 2.3217391304347825,
+      "grad_norm": 0.2315974386021932,
+      "learning_rate": 1.2560386473429953e-05,
+      "loss": 0.3194,
+      "step": 1602
+    },
+    {
+      "epoch": 2.3231884057971013,
+      "grad_norm": 0.26823600554364063,
+      "learning_rate": 1.2533548040794419e-05,
+      "loss": 0.4055,
+      "step": 1603
+    },
+    {
+      "epoch": 2.3246376811594205,
+      "grad_norm": 0.2769963244395018,
+      "learning_rate": 1.2506709608158885e-05,
+      "loss": 0.3917,
+      "step": 1604
+    },
+    {
+      "epoch": 2.3260869565217392,
+      "grad_norm": 0.2528968534522811,
+      "learning_rate": 1.247987117552335e-05,
+      "loss": 0.3419,
+      "step": 1605
+    },
+    {
+      "epoch": 2.327536231884058,
+      "grad_norm": 0.25160922497130767,
+      "learning_rate": 1.2453032742887815e-05,
+      "loss": 0.3872,
+      "step": 1606
+    },
+    {
+      "epoch": 2.3289855072463768,
+      "grad_norm": 0.247492946747638,
+      "learning_rate": 1.2426194310252282e-05,
+      "loss": 0.3221,
+      "step": 1607
+    },
+    {
+      "epoch": 2.3304347826086955,
+      "grad_norm": 0.24529622457309413,
+      "learning_rate": 1.2399355877616748e-05,
+      "loss": 0.3213,
+      "step": 1608
+    },
+    {
+      "epoch": 2.3318840579710143,
+      "grad_norm": 0.2530193260025237,
+      "learning_rate": 1.2372517444981214e-05,
+      "loss": 0.3504,
+      "step": 1609
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.2697304808458077,
+      "learning_rate": 1.2345679012345678e-05,
+      "loss": 0.3604,
+      "step": 1610
+    },
+    {
+      "epoch": 2.3347826086956522,
+      "grad_norm": 0.25794714226841003,
+      "learning_rate": 1.2318840579710146e-05,
+      "loss": 0.3781,
+      "step": 1611
+    },
+    {
+      "epoch": 2.336231884057971,
+      "grad_norm": 0.27312415594607103,
+      "learning_rate": 1.2292002147074612e-05,
+      "loss": 0.4151,
+      "step": 1612
+    },
+    {
+      "epoch": 2.3376811594202898,
+      "grad_norm": 0.22763034416379813,
+      "learning_rate": 1.2265163714439078e-05,
+      "loss": 0.373,
+      "step": 1613
+    },
+    {
+      "epoch": 2.3391304347826085,
+      "grad_norm": 0.24147613295919393,
+      "learning_rate": 1.2238325281803543e-05,
+      "loss": 0.3416,
+      "step": 1614
+    },
+    {
+      "epoch": 2.3405797101449277,
+      "grad_norm": 0.24221281514548484,
+      "learning_rate": 1.2211486849168009e-05,
+      "loss": 0.347,
+      "step": 1615
+    },
+    {
+      "epoch": 2.3420289855072465,
+      "grad_norm": 0.27036392653264024,
+      "learning_rate": 1.2184648416532475e-05,
+      "loss": 0.3722,
+      "step": 1616
+    },
+    {
+      "epoch": 2.3434782608695652,
+      "grad_norm": 0.28875833209074553,
+      "learning_rate": 1.2157809983896941e-05,
+      "loss": 0.4155,
+      "step": 1617
+    },
+    {
+      "epoch": 2.344927536231884,
+      "grad_norm": 0.2473071236867308,
+      "learning_rate": 1.2130971551261407e-05,
+      "loss": 0.3571,
+      "step": 1618
+    },
+    {
+      "epoch": 2.3463768115942027,
+      "grad_norm": 0.25668025739386363,
+      "learning_rate": 1.2104133118625872e-05,
+      "loss": 0.4249,
+      "step": 1619
+    },
+    {
+      "epoch": 2.3478260869565215,
+      "grad_norm": 0.27416799604861636,
+      "learning_rate": 1.2077294685990338e-05,
+      "loss": 0.3987,
+      "step": 1620
+    },
+    {
+      "epoch": 2.3492753623188407,
+      "grad_norm": 0.2365647100348582,
+      "learning_rate": 1.2050456253354804e-05,
+      "loss": 0.3634,
+      "step": 1621
+    },
+    {
+      "epoch": 2.3507246376811595,
+      "grad_norm": 0.27636558798058686,
+      "learning_rate": 1.2023617820719272e-05,
+      "loss": 0.4219,
+      "step": 1622
+    },
+    {
+      "epoch": 2.3521739130434782,
+      "grad_norm": 0.24753399152318645,
+      "learning_rate": 1.1996779388083736e-05,
+      "loss": 0.378,
+      "step": 1623
+    },
+    {
+      "epoch": 2.353623188405797,
+      "grad_norm": 0.2974985218029262,
+      "learning_rate": 1.1969940955448202e-05,
+      "loss": 0.3691,
+      "step": 1624
+    },
+    {
+      "epoch": 2.355072463768116,
+      "grad_norm": 0.2400268147449765,
+      "learning_rate": 1.1943102522812668e-05,
+      "loss": 0.3604,
+      "step": 1625
+    },
+    {
+      "epoch": 2.356521739130435,
+      "grad_norm": 0.26089767720269386,
+      "learning_rate": 1.1916264090177135e-05,
+      "loss": 0.399,
+      "step": 1626
+    },
+    {
+      "epoch": 2.3579710144927537,
+      "grad_norm": 0.23960649203229525,
+      "learning_rate": 1.18894256575416e-05,
+      "loss": 0.4135,
+      "step": 1627
+    },
+    {
+      "epoch": 2.3594202898550725,
+      "grad_norm": 0.2564677593387533,
+      "learning_rate": 1.1862587224906065e-05,
+      "loss": 0.3766,
+      "step": 1628
+    },
+    {
+      "epoch": 2.360869565217391,
+      "grad_norm": 0.23217865131785306,
+      "learning_rate": 1.1835748792270531e-05,
+      "loss": 0.3706,
+      "step": 1629
+    },
+    {
+      "epoch": 2.36231884057971,
+      "grad_norm": 0.253409443424873,
+      "learning_rate": 1.1808910359634997e-05,
+      "loss": 0.3507,
+      "step": 1630
+    },
+    {
+      "epoch": 2.3637681159420287,
+      "grad_norm": 0.2585803797976821,
+      "learning_rate": 1.1782071926999464e-05,
+      "loss": 0.341,
+      "step": 1631
+    },
+    {
+      "epoch": 2.365217391304348,
+      "grad_norm": 0.2343633560766101,
+      "learning_rate": 1.175523349436393e-05,
+      "loss": 0.3835,
+      "step": 1632
+    },
+    {
+      "epoch": 2.3666666666666667,
+      "grad_norm": 0.22943771254165923,
+      "learning_rate": 1.1728395061728396e-05,
+      "loss": 0.3369,
+      "step": 1633
+    },
+    {
+      "epoch": 2.3681159420289855,
+      "grad_norm": 0.2520076354267771,
+      "learning_rate": 1.1701556629092862e-05,
+      "loss": 0.3721,
+      "step": 1634
+    },
+    {
+      "epoch": 2.369565217391304,
+      "grad_norm": 0.23350354169235513,
+      "learning_rate": 1.1674718196457328e-05,
+      "loss": 0.38,
+      "step": 1635
+    },
+    {
+      "epoch": 2.3710144927536234,
+      "grad_norm": 0.24654601299800227,
+      "learning_rate": 1.1647879763821794e-05,
+      "loss": 0.4016,
+      "step": 1636
+    },
+    {
+      "epoch": 2.372463768115942,
+      "grad_norm": 0.23800732022778495,
+      "learning_rate": 1.162104133118626e-05,
+      "loss": 0.3526,
+      "step": 1637
+    },
+    {
+      "epoch": 2.373913043478261,
+      "grad_norm": 0.24450160756693554,
+      "learning_rate": 1.1594202898550725e-05,
+      "loss": 0.4154,
+      "step": 1638
+    },
+    {
+      "epoch": 2.3753623188405797,
+      "grad_norm": 0.26047505838335094,
+      "learning_rate": 1.1567364465915191e-05,
+      "loss": 0.3873,
+      "step": 1639
+    },
+    {
+      "epoch": 2.3768115942028984,
+      "grad_norm": 0.35237406523845116,
+      "learning_rate": 1.1540526033279657e-05,
+      "loss": 0.3573,
+      "step": 1640
+    },
+    {
+      "epoch": 2.378260869565217,
+      "grad_norm": 0.2464247686725185,
+      "learning_rate": 1.1513687600644123e-05,
+      "loss": 0.3513,
+      "step": 1641
+    },
+    {
+      "epoch": 2.3797101449275364,
+      "grad_norm": 0.24305252893817997,
+      "learning_rate": 1.148684916800859e-05,
+      "loss": 0.3515,
+      "step": 1642
+    },
+    {
+      "epoch": 2.381159420289855,
+      "grad_norm": 0.24473955308504647,
+      "learning_rate": 1.1460010735373054e-05,
+      "loss": 0.4069,
+      "step": 1643
+    },
+    {
+      "epoch": 2.382608695652174,
+      "grad_norm": 0.22287472790277094,
+      "learning_rate": 1.143317230273752e-05,
+      "loss": 0.3401,
+      "step": 1644
+    },
+    {
+      "epoch": 2.3840579710144927,
+      "grad_norm": 0.23881686125956375,
+      "learning_rate": 1.1406333870101986e-05,
+      "loss": 0.4016,
+      "step": 1645
+    },
+    {
+      "epoch": 2.3855072463768114,
+      "grad_norm": 0.23327446919723854,
+      "learning_rate": 1.1379495437466454e-05,
+      "loss": 0.3556,
+      "step": 1646
+    },
+    {
+      "epoch": 2.3869565217391306,
+      "grad_norm": 0.24396771678051765,
+      "learning_rate": 1.1352657004830918e-05,
+      "loss": 0.3918,
+      "step": 1647
+    },
+    {
+      "epoch": 2.3884057971014494,
+      "grad_norm": 0.24450824467838372,
+      "learning_rate": 1.1325818572195384e-05,
+      "loss": 0.3605,
+      "step": 1648
+    },
+    {
+      "epoch": 2.389855072463768,
+      "grad_norm": 0.27919448053191337,
+      "learning_rate": 1.129898013955985e-05,
+      "loss": 0.3786,
+      "step": 1649
+    },
+    {
+      "epoch": 2.391304347826087,
+      "grad_norm": 0.261991121797679,
+      "learning_rate": 1.1272141706924317e-05,
+      "loss": 0.3696,
+      "step": 1650
+    },
+    {
+      "epoch": 2.3927536231884057,
+      "grad_norm": 0.21085700795205384,
+      "learning_rate": 1.1245303274288783e-05,
+      "loss": 0.3033,
+      "step": 1651
+    },
+    {
+      "epoch": 2.3942028985507244,
+      "grad_norm": 0.22978194186144127,
+      "learning_rate": 1.1218464841653247e-05,
+      "loss": 0.3714,
+      "step": 1652
+    },
+    {
+      "epoch": 2.3956521739130436,
+      "grad_norm": 0.25088071978596216,
+      "learning_rate": 1.1191626409017713e-05,
+      "loss": 0.3195,
+      "step": 1653
+    },
+    {
+      "epoch": 2.3971014492753624,
+      "grad_norm": 0.2381310397831137,
+      "learning_rate": 1.116478797638218e-05,
+      "loss": 0.3404,
+      "step": 1654
+    },
+    {
+      "epoch": 2.398550724637681,
+      "grad_norm": 0.24188352968749965,
+      "learning_rate": 1.1137949543746646e-05,
+      "loss": 0.3577,
+      "step": 1655
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.2680659121690275,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 0.3738,
+      "step": 1656
+    },
+    {
+      "epoch": 2.4014492753623187,
+      "grad_norm": 0.24947078992123775,
+      "learning_rate": 1.1084272678475578e-05,
+      "loss": 0.3996,
+      "step": 1657
+    },
+    {
+      "epoch": 2.402898550724638,
+      "grad_norm": 0.26438557516423244,
+      "learning_rate": 1.1057434245840044e-05,
+      "loss": 0.3918,
+      "step": 1658
+    },
+    {
+      "epoch": 2.4043478260869566,
+      "grad_norm": 0.2459562582649405,
+      "learning_rate": 1.103059581320451e-05,
+      "loss": 0.3514,
+      "step": 1659
+    },
+    {
+      "epoch": 2.4057971014492754,
+      "grad_norm": 0.2614107116304787,
+      "learning_rate": 1.1003757380568976e-05,
+      "loss": 0.3452,
+      "step": 1660
+    },
+    {
+      "epoch": 2.407246376811594,
+      "grad_norm": 0.26588141942238935,
+      "learning_rate": 1.097691894793344e-05,
+      "loss": 0.3893,
+      "step": 1661
+    },
+    {
+      "epoch": 2.408695652173913,
+      "grad_norm": 0.2561460931983173,
+      "learning_rate": 1.0950080515297907e-05,
+      "loss": 0.3612,
+      "step": 1662
+    },
+    {
+      "epoch": 2.4101449275362317,
+      "grad_norm": 0.26203192216579646,
+      "learning_rate": 1.0923242082662373e-05,
+      "loss": 0.3527,
+      "step": 1663
+    },
+    {
+      "epoch": 2.411594202898551,
+      "grad_norm": 0.2801535164029354,
+      "learning_rate": 1.0896403650026839e-05,
+      "loss": 0.3742,
+      "step": 1664
+    },
+    {
+      "epoch": 2.4130434782608696,
+      "grad_norm": 0.28338654825522025,
+      "learning_rate": 1.0869565217391305e-05,
+      "loss": 0.4081,
+      "step": 1665
+    },
+    {
+      "epoch": 2.4144927536231884,
+      "grad_norm": 0.22440848405306443,
+      "learning_rate": 1.084272678475577e-05,
+      "loss": 0.3166,
+      "step": 1666
+    },
+    {
+      "epoch": 2.415942028985507,
+      "grad_norm": 0.23489668860066648,
+      "learning_rate": 1.0815888352120236e-05,
+      "loss": 0.3186,
+      "step": 1667
+    },
+    {
+      "epoch": 2.417391304347826,
+      "grad_norm": 0.2386299301125299,
+      "learning_rate": 1.0789049919484702e-05,
+      "loss": 0.3942,
+      "step": 1668
+    },
+    {
+      "epoch": 2.418840579710145,
+      "grad_norm": 0.23436076505372425,
+      "learning_rate": 1.076221148684917e-05,
+      "loss": 0.3699,
+      "step": 1669
+    },
+    {
+      "epoch": 2.420289855072464,
+      "grad_norm": 0.22909140214577542,
+      "learning_rate": 1.0735373054213636e-05,
+      "loss": 0.3476,
+      "step": 1670
+    },
+    {
+      "epoch": 2.4217391304347826,
+      "grad_norm": 0.2634134796188035,
+      "learning_rate": 1.07085346215781e-05,
+      "loss": 0.349,
+      "step": 1671
+    },
+    {
+      "epoch": 2.4231884057971014,
+      "grad_norm": 0.2889603805151576,
+      "learning_rate": 1.0681696188942566e-05,
+      "loss": 0.3477,
+      "step": 1672
+    },
+    {
+      "epoch": 2.42463768115942,
+      "grad_norm": 0.22004010840649466,
+      "learning_rate": 1.0654857756307032e-05,
+      "loss": 0.3377,
+      "step": 1673
+    },
+    {
+      "epoch": 2.426086956521739,
+      "grad_norm": 0.22623503233422046,
+      "learning_rate": 1.0628019323671499e-05,
+      "loss": 0.343,
+      "step": 1674
+    },
+    {
+      "epoch": 2.427536231884058,
+      "grad_norm": 0.23996984275551328,
+      "learning_rate": 1.0601180891035965e-05,
+      "loss": 0.3623,
+      "step": 1675
+    },
+    {
+      "epoch": 2.428985507246377,
+      "grad_norm": 0.25513314558224837,
+      "learning_rate": 1.0574342458400429e-05,
+      "loss": 0.4259,
+      "step": 1676
+    },
+    {
+      "epoch": 2.4304347826086956,
+      "grad_norm": 0.2265078118915947,
+      "learning_rate": 1.0547504025764895e-05,
+      "loss": 0.3691,
+      "step": 1677
+    },
+    {
+      "epoch": 2.4318840579710144,
+      "grad_norm": 0.23586487503109818,
+      "learning_rate": 1.0520665593129361e-05,
+      "loss": 0.3442,
+      "step": 1678
+    },
+    {
+      "epoch": 2.4333333333333336,
+      "grad_norm": 0.24515098029968993,
+      "learning_rate": 1.0493827160493827e-05,
+      "loss": 0.3572,
+      "step": 1679
+    },
+    {
+      "epoch": 2.4347826086956523,
+      "grad_norm": 0.24521992528057063,
+      "learning_rate": 1.0466988727858294e-05,
+      "loss": 0.3703,
+      "step": 1680
+    },
+    {
+      "epoch": 2.436231884057971,
+      "grad_norm": 0.2356508765619123,
+      "learning_rate": 1.044015029522276e-05,
+      "loss": 0.3969,
+      "step": 1681
+    },
+    {
+      "epoch": 2.43768115942029,
+      "grad_norm": 0.25389786816355275,
+      "learning_rate": 1.0413311862587226e-05,
+      "loss": 0.3975,
+      "step": 1682
+    },
+    {
+      "epoch": 2.4391304347826086,
+      "grad_norm": 0.26355397779718365,
+      "learning_rate": 1.0386473429951692e-05,
+      "loss": 0.3739,
+      "step": 1683
+    },
+    {
+      "epoch": 2.4405797101449274,
+      "grad_norm": 0.2551779484837169,
+      "learning_rate": 1.0359634997316158e-05,
+      "loss": 0.3885,
+      "step": 1684
+    },
+    {
+      "epoch": 2.4420289855072466,
+      "grad_norm": 0.23631562097115857,
+      "learning_rate": 1.0332796564680623e-05,
+      "loss": 0.3855,
+      "step": 1685
+    },
+    {
+      "epoch": 2.4434782608695653,
+      "grad_norm": 0.23618850704064212,
+      "learning_rate": 1.0305958132045089e-05,
+      "loss": 0.4032,
+      "step": 1686
+    },
+    {
+      "epoch": 2.444927536231884,
+      "grad_norm": 0.24259183480511126,
+      "learning_rate": 1.0279119699409555e-05,
+      "loss": 0.3573,
+      "step": 1687
+    },
+    {
+      "epoch": 2.446376811594203,
+      "grad_norm": 0.24202233121882677,
+      "learning_rate": 1.0252281266774021e-05,
+      "loss": 0.3871,
+      "step": 1688
+    },
+    {
+      "epoch": 2.4478260869565216,
+      "grad_norm": 0.23825262433919453,
+      "learning_rate": 1.0225442834138487e-05,
+      "loss": 0.3619,
+      "step": 1689
+    },
+    {
+      "epoch": 2.449275362318841,
+      "grad_norm": 0.5378715452062908,
+      "learning_rate": 1.0198604401502951e-05,
+      "loss": 0.417,
+      "step": 1690
+    },
+    {
+      "epoch": 2.4507246376811596,
+      "grad_norm": 0.23590631249131455,
+      "learning_rate": 1.0171765968867418e-05,
+      "loss": 0.3728,
+      "step": 1691
+    },
+    {
+      "epoch": 2.4521739130434783,
+      "grad_norm": 0.22704366611632165,
+      "learning_rate": 1.0144927536231885e-05,
+      "loss": 0.3584,
+      "step": 1692
+    },
+    {
+      "epoch": 2.453623188405797,
+      "grad_norm": 0.24678300364706693,
+      "learning_rate": 1.0118089103596352e-05,
+      "loss": 0.3648,
+      "step": 1693
+    },
+    {
+      "epoch": 2.455072463768116,
+      "grad_norm": 0.24623958958542122,
+      "learning_rate": 1.0091250670960816e-05,
+      "loss": 0.3672,
+      "step": 1694
+    },
+    {
+      "epoch": 2.4565217391304346,
+      "grad_norm": 0.2548241126895578,
+      "learning_rate": 1.0064412238325282e-05,
+      "loss": 0.3891,
+      "step": 1695
+    },
+    {
+      "epoch": 2.457971014492754,
+      "grad_norm": 0.2307626058795168,
+      "learning_rate": 1.0037573805689748e-05,
+      "loss": 0.3588,
+      "step": 1696
+    },
+    {
+      "epoch": 2.4594202898550726,
+      "grad_norm": 0.22542946104824368,
+      "learning_rate": 1.0010735373054214e-05,
+      "loss": 0.3975,
+      "step": 1697
+    },
+    {
+      "epoch": 2.4608695652173913,
+      "grad_norm": 0.24207984238673924,
+      "learning_rate": 9.98389694041868e-06,
+      "loss": 0.3916,
+      "step": 1698
+    },
+    {
+      "epoch": 2.46231884057971,
+      "grad_norm": 0.27096598935315286,
+      "learning_rate": 9.957058507783145e-06,
+      "loss": 0.3776,
+      "step": 1699
+    },
+    {
+      "epoch": 2.463768115942029,
+      "grad_norm": 0.22850501591889746,
+      "learning_rate": 9.930220075147611e-06,
+      "loss": 0.3576,
+      "step": 1700
+    },
+    {
+      "epoch": 2.465217391304348,
+      "grad_norm": 0.22631443374823518,
+      "learning_rate": 9.903381642512077e-06,
+      "loss": 0.3618,
+      "step": 1701
+    },
+    {
+      "epoch": 2.466666666666667,
+      "grad_norm": 0.2321696076731389,
+      "learning_rate": 9.876543209876543e-06,
+      "loss": 0.3665,
+      "step": 1702
+    },
+    {
+      "epoch": 2.4681159420289855,
+      "grad_norm": 0.24687628048049745,
+      "learning_rate": 9.84970477724101e-06,
+      "loss": 0.3586,
+      "step": 1703
+    },
+    {
+      "epoch": 2.4695652173913043,
+      "grad_norm": 0.24373847387295003,
+      "learning_rate": 9.822866344605476e-06,
+      "loss": 0.3652,
+      "step": 1704
+    },
+    {
+      "epoch": 2.471014492753623,
+      "grad_norm": 2.8444772494805672,
+      "learning_rate": 9.796027911969942e-06,
+      "loss": 0.4705,
+      "step": 1705
+    },
+    {
+      "epoch": 2.472463768115942,
+      "grad_norm": 0.2930409317815382,
+      "learning_rate": 9.769189479334408e-06,
+      "loss": 0.3877,
+      "step": 1706
+    },
+    {
+      "epoch": 2.473913043478261,
+      "grad_norm": 0.2601267088028184,
+      "learning_rate": 9.742351046698874e-06,
+      "loss": 0.3644,
+      "step": 1707
+    },
+    {
+      "epoch": 2.47536231884058,
+      "grad_norm": 0.2487197364866856,
+      "learning_rate": 9.71551261406334e-06,
+      "loss": 0.4045,
+      "step": 1708
+    },
+    {
+      "epoch": 2.4768115942028985,
+      "grad_norm": 0.2374071864441594,
+      "learning_rate": 9.688674181427804e-06,
+      "loss": 0.3568,
+      "step": 1709
+    },
+    {
+      "epoch": 2.4782608695652173,
+      "grad_norm": 0.24310394382631415,
+      "learning_rate": 9.66183574879227e-06,
+      "loss": 0.3498,
+      "step": 1710
+    },
+    {
+      "epoch": 2.479710144927536,
+      "grad_norm": 0.23519481570010747,
+      "learning_rate": 9.634997316156737e-06,
+      "loss": 0.3985,
+      "step": 1711
+    },
+    {
+      "epoch": 2.4811594202898553,
+      "grad_norm": 0.24301143269826625,
+      "learning_rate": 9.608158883521203e-06,
+      "loss": 0.3962,
+      "step": 1712
+    },
+    {
+      "epoch": 2.482608695652174,
+      "grad_norm": 0.23603059564584986,
+      "learning_rate": 9.581320450885669e-06,
+      "loss": 0.3232,
+      "step": 1713
+    },
+    {
+      "epoch": 2.4840579710144928,
+      "grad_norm": 0.30965155321248744,
+      "learning_rate": 9.554482018250133e-06,
+      "loss": 0.4068,
+      "step": 1714
+    },
+    {
+      "epoch": 2.4855072463768115,
+      "grad_norm": 0.23388751278509945,
+      "learning_rate": 9.527643585614601e-06,
+      "loss": 0.3458,
+      "step": 1715
+    },
+    {
+      "epoch": 2.4869565217391303,
+      "grad_norm": 0.2686097858868765,
+      "learning_rate": 9.500805152979067e-06,
+      "loss": 0.4051,
+      "step": 1716
+    },
+    {
+      "epoch": 2.488405797101449,
+      "grad_norm": 0.2718429545342115,
+      "learning_rate": 9.473966720343534e-06,
+      "loss": 0.4266,
+      "step": 1717
+    },
+    {
+      "epoch": 2.4898550724637682,
+      "grad_norm": 0.2382473738490544,
+      "learning_rate": 9.447128287707998e-06,
+      "loss": 0.4204,
+      "step": 1718
+    },
+    {
+      "epoch": 2.491304347826087,
+      "grad_norm": 0.23698376840914315,
+      "learning_rate": 9.420289855072464e-06,
+      "loss": 0.3631,
+      "step": 1719
+    },
+    {
+      "epoch": 2.4927536231884058,
+      "grad_norm": 0.23506995248352258,
+      "learning_rate": 9.39345142243693e-06,
+      "loss": 0.3271,
+      "step": 1720
+    },
+    {
+      "epoch": 2.4942028985507245,
+      "grad_norm": 0.2564784422890809,
+      "learning_rate": 9.366612989801396e-06,
+      "loss": 0.405,
+      "step": 1721
+    },
+    {
+      "epoch": 2.4956521739130437,
+      "grad_norm": 0.23915383613372546,
+      "learning_rate": 9.339774557165862e-06,
+      "loss": 0.3521,
+      "step": 1722
+    },
+    {
+      "epoch": 2.4971014492753625,
+      "grad_norm": 0.4073306961152648,
+      "learning_rate": 9.312936124530327e-06,
+      "loss": 0.3665,
+      "step": 1723
+    },
+    {
+      "epoch": 2.4985507246376812,
+      "grad_norm": 0.2285687119732312,
+      "learning_rate": 9.286097691894793e-06,
+      "loss": 0.3588,
+      "step": 1724
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.2458977310163074,
+      "learning_rate": 9.259259259259259e-06,
+      "loss": 0.3362,
+      "step": 1725
+    },
+    {
+      "epoch": 2.5014492753623188,
+      "grad_norm": 0.26512727560130933,
+      "learning_rate": 9.232420826623725e-06,
+      "loss": 0.3928,
+      "step": 1726
+    },
+    {
+      "epoch": 2.5028985507246375,
+      "grad_norm": 0.2457255606384545,
+      "learning_rate": 9.205582393988191e-06,
+      "loss": 0.4004,
+      "step": 1727
+    },
+    {
+      "epoch": 2.5043478260869563,
+      "grad_norm": 0.25178128091977336,
+      "learning_rate": 9.178743961352658e-06,
+      "loss": 0.4259,
+      "step": 1728
+    },
+    {
+      "epoch": 2.5057971014492755,
+      "grad_norm": 0.24079198772295304,
+      "learning_rate": 9.151905528717124e-06,
+      "loss": 0.3565,
+      "step": 1729
+    },
+    {
+      "epoch": 2.5072463768115942,
+      "grad_norm": 0.2931582636874582,
+      "learning_rate": 9.12506709608159e-06,
+      "loss": 0.3875,
+      "step": 1730
+    },
+    {
+      "epoch": 2.508695652173913,
+      "grad_norm": 0.25170263590126585,
+      "learning_rate": 9.098228663446056e-06,
+      "loss": 0.3722,
+      "step": 1731
+    },
+    {
+      "epoch": 2.5101449275362318,
+      "grad_norm": 0.24220178096105427,
+      "learning_rate": 9.07139023081052e-06,
+      "loss": 0.3667,
+      "step": 1732
+    },
+    {
+      "epoch": 2.511594202898551,
+      "grad_norm": 0.22224946420231548,
+      "learning_rate": 9.044551798174986e-06,
+      "loss": 0.3191,
+      "step": 1733
+    },
+    {
+      "epoch": 2.5130434782608697,
+      "grad_norm": 0.24089894753896055,
+      "learning_rate": 9.017713365539453e-06,
+      "loss": 0.362,
+      "step": 1734
+    },
+    {
+      "epoch": 2.5144927536231885,
+      "grad_norm": 0.22841326431279493,
+      "learning_rate": 8.990874932903919e-06,
+      "loss": 0.3453,
+      "step": 1735
+    },
+    {
+      "epoch": 2.5159420289855072,
+      "grad_norm": 0.2393750097499759,
+      "learning_rate": 8.964036500268385e-06,
+      "loss": 0.3583,
+      "step": 1736
+    },
+    {
+      "epoch": 2.517391304347826,
+      "grad_norm": 0.26855067884257117,
+      "learning_rate": 8.93719806763285e-06,
+      "loss": 0.4091,
+      "step": 1737
+    },
+    {
+      "epoch": 2.5188405797101447,
+      "grad_norm": 0.23517858529776384,
+      "learning_rate": 8.910359634997315e-06,
+      "loss": 0.4249,
+      "step": 1738
+    },
+    {
+      "epoch": 2.5202898550724635,
+      "grad_norm": 0.2560093265834709,
+      "learning_rate": 8.883521202361783e-06,
+      "loss": 0.3909,
+      "step": 1739
+    },
+    {
+      "epoch": 2.5217391304347827,
+      "grad_norm": 0.24412506294757666,
+      "learning_rate": 8.85668276972625e-06,
+      "loss": 0.3974,
+      "step": 1740
+    },
+    {
+      "epoch": 2.5231884057971015,
+      "grad_norm": 0.24174238885908517,
+      "learning_rate": 8.829844337090715e-06,
+      "loss": 0.3788,
+      "step": 1741
+    },
+    {
+      "epoch": 2.52463768115942,
+      "grad_norm": 0.24203049886043884,
+      "learning_rate": 8.80300590445518e-06,
+      "loss": 0.3996,
+      "step": 1742
+    },
+    {
+      "epoch": 2.526086956521739,
+      "grad_norm": 0.2331818758188412,
+      "learning_rate": 8.776167471819646e-06,
+      "loss": 0.3727,
+      "step": 1743
+    },
+    {
+      "epoch": 2.527536231884058,
+      "grad_norm": 0.2259773165695989,
+      "learning_rate": 8.749329039184112e-06,
+      "loss": 0.337,
+      "step": 1744
+    },
+    {
+      "epoch": 2.528985507246377,
+      "grad_norm": 0.21970295461037015,
+      "learning_rate": 8.722490606548578e-06,
+      "loss": 0.3076,
+      "step": 1745
+    },
+    {
+      "epoch": 2.5304347826086957,
+      "grad_norm": 0.240548024122901,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 0.3617,
+      "step": 1746
+    },
+    {
+      "epoch": 2.5318840579710145,
+      "grad_norm": 0.22471292230572035,
+      "learning_rate": 8.668813741277509e-06,
+      "loss": 0.3845,
+      "step": 1747
+    },
+    {
+      "epoch": 2.533333333333333,
+      "grad_norm": 0.2324219627223449,
+      "learning_rate": 8.641975308641975e-06,
+      "loss": 0.3577,
+      "step": 1748
+    },
+    {
+      "epoch": 2.534782608695652,
+      "grad_norm": 0.2517119814575534,
+      "learning_rate": 8.615136876006441e-06,
+      "loss": 0.3838,
+      "step": 1749
+    },
+    {
+      "epoch": 2.536231884057971,
+      "grad_norm": 0.23162613939840387,
+      "learning_rate": 8.588298443370909e-06,
+      "loss": 0.3807,
+      "step": 1750
+    },
+    {
+      "epoch": 2.53768115942029,
+      "grad_norm": 0.23210374731896627,
+      "learning_rate": 8.561460010735373e-06,
+      "loss": 0.4135,
+      "step": 1751
+    },
+    {
+      "epoch": 2.5391304347826087,
+      "grad_norm": 0.2506407668081656,
+      "learning_rate": 8.53462157809984e-06,
+      "loss": 0.3798,
+      "step": 1752
+    },
+    {
+      "epoch": 2.5405797101449274,
+      "grad_norm": 0.24427314286408663,
+      "learning_rate": 8.507783145464306e-06,
+      "loss": 0.3534,
+      "step": 1753
+    },
+    {
+      "epoch": 2.5420289855072467,
+      "grad_norm": 0.2521195759976504,
+      "learning_rate": 8.480944712828772e-06,
+      "loss": 0.4297,
+      "step": 1754
+    },
+    {
+      "epoch": 2.5434782608695654,
+      "grad_norm": 0.22423334849213983,
+      "learning_rate": 8.454106280193238e-06,
+      "loss": 0.3649,
+      "step": 1755
+    },
+    {
+      "epoch": 2.544927536231884,
+      "grad_norm": 0.22454826458337243,
+      "learning_rate": 8.427267847557702e-06,
+      "loss": 0.3854,
+      "step": 1756
+    },
+    {
+      "epoch": 2.546376811594203,
+      "grad_norm": 0.2495166610490925,
+      "learning_rate": 8.400429414922168e-06,
+      "loss": 0.3238,
+      "step": 1757
+    },
+    {
+      "epoch": 2.5478260869565217,
+      "grad_norm": 0.2598103637415767,
+      "learning_rate": 8.373590982286635e-06,
+      "loss": 0.424,
+      "step": 1758
+    },
+    {
+      "epoch": 2.5492753623188404,
+      "grad_norm": 0.2489444890547027,
+      "learning_rate": 8.3467525496511e-06,
+      "loss": 0.3986,
+      "step": 1759
+    },
+    {
+      "epoch": 2.550724637681159,
+      "grad_norm": 0.24194370005383292,
+      "learning_rate": 8.319914117015567e-06,
+      "loss": 0.3844,
+      "step": 1760
+    },
+    {
+      "epoch": 2.5521739130434784,
+      "grad_norm": 0.23640849731439492,
+      "learning_rate": 8.293075684380031e-06,
+      "loss": 0.3811,
+      "step": 1761
+    },
+    {
+      "epoch": 2.553623188405797,
+      "grad_norm": 0.2272161818033302,
+      "learning_rate": 8.266237251744499e-06,
+      "loss": 0.3691,
+      "step": 1762
+    },
+    {
+      "epoch": 2.555072463768116,
+      "grad_norm": 0.21753209322286424,
+      "learning_rate": 8.239398819108965e-06,
+      "loss": 0.3484,
+      "step": 1763
+    },
+    {
+      "epoch": 2.5565217391304347,
+      "grad_norm": 0.24397631738740422,
+      "learning_rate": 8.212560386473431e-06,
+      "loss": 0.37,
+      "step": 1764
+    },
+    {
+      "epoch": 2.557971014492754,
+      "grad_norm": 0.2280370924702514,
+      "learning_rate": 8.185721953837896e-06,
+      "loss": 0.3813,
+      "step": 1765
+    },
+    {
+      "epoch": 2.5594202898550726,
+      "grad_norm": 0.25304936700858394,
+      "learning_rate": 8.158883521202362e-06,
+      "loss": 0.3995,
+      "step": 1766
+    },
+    {
+      "epoch": 2.5608695652173914,
+      "grad_norm": 0.2490687010035602,
+      "learning_rate": 8.132045088566828e-06,
+      "loss": 0.411,
+      "step": 1767
+    },
+    {
+      "epoch": 2.56231884057971,
+      "grad_norm": 0.2513432328777938,
+      "learning_rate": 8.105206655931294e-06,
+      "loss": 0.4075,
+      "step": 1768
+    },
+    {
+      "epoch": 2.563768115942029,
+      "grad_norm": 0.23915876131137653,
+      "learning_rate": 8.07836822329576e-06,
+      "loss": 0.4069,
+      "step": 1769
+    },
+    {
+      "epoch": 2.5652173913043477,
+      "grad_norm": 0.23030860469836206,
+      "learning_rate": 8.051529790660225e-06,
+      "loss": 0.4023,
+      "step": 1770
+    },
+    {
+      "epoch": 2.5666666666666664,
+      "grad_norm": 0.22533124000674842,
+      "learning_rate": 8.02469135802469e-06,
+      "loss": 0.353,
+      "step": 1771
+    },
+    {
+      "epoch": 2.5681159420289856,
+      "grad_norm": 0.23594375320707797,
+      "learning_rate": 7.997852925389157e-06,
+      "loss": 0.347,
+      "step": 1772
+    },
+    {
+      "epoch": 2.5695652173913044,
+      "grad_norm": 0.23574195339153892,
+      "learning_rate": 7.971014492753623e-06,
+      "loss": 0.3703,
+      "step": 1773
+    },
+    {
+      "epoch": 2.571014492753623,
+      "grad_norm": 0.2471542224831895,
+      "learning_rate": 7.944176060118091e-06,
+      "loss": 0.3881,
+      "step": 1774
+    },
+    {
+      "epoch": 2.572463768115942,
+      "grad_norm": 0.237479615359418,
+      "learning_rate": 7.917337627482555e-06,
+      "loss": 0.3866,
+      "step": 1775
+    },
+    {
+      "epoch": 2.573913043478261,
+      "grad_norm": 0.2302231762999212,
+      "learning_rate": 7.890499194847021e-06,
+      "loss": 0.399,
+      "step": 1776
+    },
+    {
+      "epoch": 2.57536231884058,
+      "grad_norm": 0.3134114587170842,
+      "learning_rate": 7.863660762211488e-06,
+      "loss": 0.3493,
+      "step": 1777
+    },
+    {
+      "epoch": 2.5768115942028986,
+      "grad_norm": 0.2246109336895405,
+      "learning_rate": 7.836822329575954e-06,
+      "loss": 0.3601,
+      "step": 1778
+    },
+    {
+      "epoch": 2.5782608695652174,
+      "grad_norm": 0.26148405527459617,
+      "learning_rate": 7.80998389694042e-06,
+      "loss": 0.3813,
+      "step": 1779
+    },
+    {
+      "epoch": 2.579710144927536,
+      "grad_norm": 0.22015793799835642,
+      "learning_rate": 7.783145464304884e-06,
+      "loss": 0.3359,
+      "step": 1780
+    },
+    {
+      "epoch": 2.581159420289855,
+      "grad_norm": 0.22781683594318913,
+      "learning_rate": 7.75630703166935e-06,
+      "loss": 0.3932,
+      "step": 1781
+    },
+    {
+      "epoch": 2.5826086956521737,
+      "grad_norm": 0.22622094981865146,
+      "learning_rate": 7.729468599033817e-06,
+      "loss": 0.3422,
+      "step": 1782
+    },
+    {
+      "epoch": 2.584057971014493,
+      "grad_norm": 0.23289969908151067,
+      "learning_rate": 7.702630166398283e-06,
+      "loss": 0.3605,
+      "step": 1783
+    },
+    {
+      "epoch": 2.5855072463768116,
+      "grad_norm": 0.246003880930885,
+      "learning_rate": 7.675791733762749e-06,
+      "loss": 0.4108,
+      "step": 1784
+    },
+    {
+      "epoch": 2.5869565217391304,
+      "grad_norm": 0.22934975828663517,
+      "learning_rate": 7.648953301127215e-06,
+      "loss": 0.3613,
+      "step": 1785
+    },
+    {
+      "epoch": 2.588405797101449,
+      "grad_norm": 0.23776511365549477,
+      "learning_rate": 7.622114868491681e-06,
+      "loss": 0.3452,
+      "step": 1786
+    },
+    {
+      "epoch": 2.5898550724637683,
+      "grad_norm": 0.23173291377330582,
+      "learning_rate": 7.595276435856147e-06,
+      "loss": 0.3636,
+      "step": 1787
+    },
+    {
+      "epoch": 2.591304347826087,
+      "grad_norm": 0.22746789158031808,
+      "learning_rate": 7.568438003220612e-06,
+      "loss": 0.3738,
+      "step": 1788
+    },
+    {
+      "epoch": 2.592753623188406,
+      "grad_norm": 0.23726838497942668,
+      "learning_rate": 7.5415995705850785e-06,
+      "loss": 0.3537,
+      "step": 1789
+    },
+    {
+      "epoch": 2.5942028985507246,
+      "grad_norm": 0.2536496869289765,
+      "learning_rate": 7.514761137949544e-06,
+      "loss": 0.3823,
+      "step": 1790
+    },
+    {
+      "epoch": 2.5956521739130434,
+      "grad_norm": 0.2323632110547211,
+      "learning_rate": 7.48792270531401e-06,
+      "loss": 0.3398,
+      "step": 1791
+    },
+    {
+      "epoch": 2.597101449275362,
+      "grad_norm": 0.24434757356294867,
+      "learning_rate": 7.461084272678476e-06,
+      "loss": 0.3987,
+      "step": 1792
+    },
+    {
+      "epoch": 2.598550724637681,
+      "grad_norm": 0.23017559291290987,
+      "learning_rate": 7.434245840042941e-06,
+      "loss": 0.3727,
+      "step": 1793
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.23150175814809001,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.3801,
+      "step": 1794
+    },
+    {
+      "epoch": 2.601449275362319,
+      "grad_norm": 0.22181897932620395,
+      "learning_rate": 7.380568974771873e-06,
+      "loss": 0.3828,
+      "step": 1795
+    },
+    {
+      "epoch": 2.6028985507246376,
+      "grad_norm": 0.23263891313339105,
+      "learning_rate": 7.353730542136339e-06,
+      "loss": 0.3557,
+      "step": 1796
+    },
+    {
+      "epoch": 2.6043478260869564,
+      "grad_norm": 0.6529574844149014,
+      "learning_rate": 7.326892109500806e-06,
+      "loss": 0.3669,
+      "step": 1797
+    },
+    {
+      "epoch": 2.6057971014492756,
+      "grad_norm": 0.21973835286616067,
+      "learning_rate": 7.300053676865272e-06,
+      "loss": 0.3439,
+      "step": 1798
+    },
+    {
+      "epoch": 2.6072463768115943,
+      "grad_norm": 0.24012497988938392,
+      "learning_rate": 7.273215244229738e-06,
+      "loss": 0.3586,
+      "step": 1799
+    },
+    {
+      "epoch": 2.608695652173913,
+      "grad_norm": 0.21946520371623274,
+      "learning_rate": 7.246376811594203e-06,
+      "loss": 0.3417,
+      "step": 1800
+    },
+    {
+      "epoch": 2.610144927536232,
+      "grad_norm": 0.2093748367861486,
+      "learning_rate": 7.2195383789586695e-06,
+      "loss": 0.2987,
+      "step": 1801
+    },
+    {
+      "epoch": 2.6115942028985506,
+      "grad_norm": 0.23063885100070947,
+      "learning_rate": 7.192699946323135e-06,
+      "loss": 0.376,
+      "step": 1802
+    },
+    {
+      "epoch": 2.6130434782608694,
+      "grad_norm": 0.24566039962287953,
+      "learning_rate": 7.165861513687601e-06,
+      "loss": 0.3889,
+      "step": 1803
+    },
+    {
+      "epoch": 2.6144927536231886,
+      "grad_norm": 0.22630481844434078,
+      "learning_rate": 7.139023081052067e-06,
+      "loss": 0.3805,
+      "step": 1804
+    },
+    {
+      "epoch": 2.6159420289855073,
+      "grad_norm": 0.23513134414589493,
+      "learning_rate": 7.112184648416532e-06,
+      "loss": 0.3703,
+      "step": 1805
+    },
+    {
+      "epoch": 2.617391304347826,
+      "grad_norm": 0.258085032589009,
+      "learning_rate": 7.0853462157809985e-06,
+      "loss": 0.42,
+      "step": 1806
+    },
+    {
+      "epoch": 2.618840579710145,
+      "grad_norm": 0.24787040551614853,
+      "learning_rate": 7.058507783145464e-06,
+      "loss": 0.383,
+      "step": 1807
+    },
+    {
+      "epoch": 2.620289855072464,
+      "grad_norm": 0.23833700860460905,
+      "learning_rate": 7.03166935050993e-06,
+      "loss": 0.3724,
+      "step": 1808
+    },
+    {
+      "epoch": 2.621739130434783,
+      "grad_norm": 0.22142826203858182,
+      "learning_rate": 7.004830917874397e-06,
+      "loss": 0.3417,
+      "step": 1809
+    },
+    {
+      "epoch": 2.6231884057971016,
+      "grad_norm": 0.2207677353089534,
+      "learning_rate": 6.977992485238863e-06,
+      "loss": 0.3337,
+      "step": 1810
+    },
+    {
+      "epoch": 2.6246376811594203,
+      "grad_norm": 0.2411024299096652,
+      "learning_rate": 6.951154052603328e-06,
+      "loss": 0.3628,
+      "step": 1811
+    },
+    {
+      "epoch": 2.626086956521739,
+      "grad_norm": 0.24307442140976468,
+      "learning_rate": 6.924315619967794e-06,
+      "loss": 0.405,
+      "step": 1812
+    },
+    {
+      "epoch": 2.627536231884058,
+      "grad_norm": 0.22134105902238405,
+      "learning_rate": 6.8974771873322605e-06,
+      "loss": 0.3666,
+      "step": 1813
+    },
+    {
+      "epoch": 2.6289855072463766,
+      "grad_norm": 0.24753571131810514,
+      "learning_rate": 6.870638754696726e-06,
+      "loss": 0.3757,
+      "step": 1814
+    },
+    {
+      "epoch": 2.630434782608696,
+      "grad_norm": 0.22052117468602914,
+      "learning_rate": 6.843800322061192e-06,
+      "loss": 0.3658,
+      "step": 1815
+    },
+    {
+      "epoch": 2.6318840579710145,
+      "grad_norm": 0.2240473452768075,
+      "learning_rate": 6.816961889425658e-06,
+      "loss": 0.3288,
+      "step": 1816
+    },
+    {
+      "epoch": 2.6333333333333333,
+      "grad_norm": 0.23427589102526614,
+      "learning_rate": 6.790123456790123e-06,
+      "loss": 0.3873,
+      "step": 1817
+    },
+    {
+      "epoch": 2.634782608695652,
+      "grad_norm": 0.2200630110125934,
+      "learning_rate": 6.7632850241545894e-06,
+      "loss": 0.3664,
+      "step": 1818
+    },
+    {
+      "epoch": 2.6362318840579713,
+      "grad_norm": 0.22018626187701326,
+      "learning_rate": 6.736446591519055e-06,
+      "loss": 0.3423,
+      "step": 1819
+    },
+    {
+      "epoch": 2.63768115942029,
+      "grad_norm": 0.2525070837601251,
+      "learning_rate": 6.7096081588835225e-06,
+      "loss": 0.4146,
+      "step": 1820
+    },
+    {
+      "epoch": 2.639130434782609,
+      "grad_norm": 0.22184775100735113,
+      "learning_rate": 6.682769726247988e-06,
+      "loss": 0.3442,
+      "step": 1821
+    },
+    {
+      "epoch": 2.6405797101449275,
+      "grad_norm": 0.23636480564458984,
+      "learning_rate": 6.655931293612454e-06,
+      "loss": 0.3622,
+      "step": 1822
+    },
+    {
+      "epoch": 2.6420289855072463,
+      "grad_norm": 0.22147421135139164,
+      "learning_rate": 6.629092860976919e-06,
+      "loss": 0.3782,
+      "step": 1823
+    },
+    {
+      "epoch": 2.643478260869565,
+      "grad_norm": 0.24603239641639185,
+      "learning_rate": 6.602254428341385e-06,
+      "loss": 0.4063,
+      "step": 1824
+    },
+    {
+      "epoch": 2.644927536231884,
+      "grad_norm": 0.23159270932447462,
+      "learning_rate": 6.5754159957058515e-06,
+      "loss": 0.377,
+      "step": 1825
+    },
+    {
+      "epoch": 2.646376811594203,
+      "grad_norm": 0.2105455774554014,
+      "learning_rate": 6.548577563070317e-06,
+      "loss": 0.327,
+      "step": 1826
+    },
+    {
+      "epoch": 2.6478260869565218,
+      "grad_norm": 0.25620857332710895,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 0.3937,
+      "step": 1827
+    },
+    {
+      "epoch": 2.6492753623188405,
+      "grad_norm": 0.22469586076992126,
+      "learning_rate": 6.494900697799248e-06,
+      "loss": 0.3517,
+      "step": 1828
+    },
+    {
+      "epoch": 2.6507246376811593,
+      "grad_norm": 0.22576026831366242,
+      "learning_rate": 6.468062265163714e-06,
+      "loss": 0.3591,
+      "step": 1829
+    },
+    {
+      "epoch": 2.6521739130434785,
+      "grad_norm": 0.2192462255694706,
+      "learning_rate": 6.44122383252818e-06,
+      "loss": 0.361,
+      "step": 1830
+    },
+    {
+      "epoch": 2.6536231884057973,
+      "grad_norm": 0.23381609847056584,
+      "learning_rate": 6.414385399892646e-06,
+      "loss": 0.4258,
+      "step": 1831
+    },
+    {
+      "epoch": 2.655072463768116,
+      "grad_norm": 0.21088337829414863,
+      "learning_rate": 6.3875469672571135e-06,
+      "loss": 0.3589,
+      "step": 1832
+    },
+    {
+      "epoch": 2.6565217391304348,
+      "grad_norm": 0.2348314510153269,
+      "learning_rate": 6.360708534621579e-06,
+      "loss": 0.3854,
+      "step": 1833
+    },
+    {
+      "epoch": 2.6579710144927535,
+      "grad_norm": 0.22631453842046725,
+      "learning_rate": 6.333870101986045e-06,
+      "loss": 0.3901,
+      "step": 1834
+    },
+    {
+      "epoch": 2.6594202898550723,
+      "grad_norm": 0.22994007869388175,
+      "learning_rate": 6.30703166935051e-06,
+      "loss": 0.3654,
+      "step": 1835
+    },
+    {
+      "epoch": 2.660869565217391,
+      "grad_norm": 0.2203624234609837,
+      "learning_rate": 6.280193236714976e-06,
+      "loss": 0.3927,
+      "step": 1836
+    },
+    {
+      "epoch": 2.6623188405797102,
+      "grad_norm": 0.2195249377140087,
+      "learning_rate": 6.2533548040794425e-06,
+      "loss": 0.3373,
+      "step": 1837
+    },
+    {
+      "epoch": 2.663768115942029,
+      "grad_norm": 0.23181377885580856,
+      "learning_rate": 6.226516371443908e-06,
+      "loss": 0.4036,
+      "step": 1838
+    },
+    {
+      "epoch": 2.6652173913043478,
+      "grad_norm": 0.2219526383481847,
+      "learning_rate": 6.199677938808374e-06,
+      "loss": 0.3534,
+      "step": 1839
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.23362554464225765,
+      "learning_rate": 6.172839506172839e-06,
+      "loss": 0.4265,
+      "step": 1840
+    },
+    {
+      "epoch": 2.6681159420289857,
+      "grad_norm": 0.24196797259877534,
+      "learning_rate": 6.146001073537306e-06,
+      "loss": 0.38,
+      "step": 1841
+    },
+    {
+      "epoch": 2.6695652173913045,
+      "grad_norm": 0.23489975729889526,
+      "learning_rate": 6.119162640901771e-06,
+      "loss": 0.3683,
+      "step": 1842
+    },
+    {
+      "epoch": 2.6710144927536232,
+      "grad_norm": 0.2299695993386685,
+      "learning_rate": 6.0923242082662375e-06,
+      "loss": 0.3546,
+      "step": 1843
+    },
+    {
+      "epoch": 2.672463768115942,
+      "grad_norm": 0.26701383005463475,
+      "learning_rate": 6.065485775630704e-06,
+      "loss": 0.3768,
+      "step": 1844
+    },
+    {
+      "epoch": 2.6739130434782608,
+      "grad_norm": 0.33256383947003426,
+      "learning_rate": 6.038647342995169e-06,
+      "loss": 0.3197,
+      "step": 1845
+    },
+    {
+      "epoch": 2.6753623188405795,
+      "grad_norm": 0.2431108486304523,
+      "learning_rate": 6.011808910359636e-06,
+      "loss": 0.3592,
+      "step": 1846
+    },
+    {
+      "epoch": 2.6768115942028987,
+      "grad_norm": 0.24168905691623235,
+      "learning_rate": 5.984970477724101e-06,
+      "loss": 0.4084,
+      "step": 1847
+    },
+    {
+      "epoch": 2.6782608695652175,
+      "grad_norm": 0.22030359227920543,
+      "learning_rate": 5.958132045088567e-06,
+      "loss": 0.3122,
+      "step": 1848
+    },
+    {
+      "epoch": 2.6797101449275362,
+      "grad_norm": 0.22133704619831146,
+      "learning_rate": 5.931293612453033e-06,
+      "loss": 0.3527,
+      "step": 1849
+    },
+    {
+      "epoch": 2.681159420289855,
+      "grad_norm": 0.23488002154972068,
+      "learning_rate": 5.904455179817499e-06,
+      "loss": 0.3577,
+      "step": 1850
+    },
+    {
+      "epoch": 2.6826086956521737,
+      "grad_norm": 0.23371719237593863,
+      "learning_rate": 5.877616747181965e-06,
+      "loss": 0.3624,
+      "step": 1851
+    },
+    {
+      "epoch": 2.684057971014493,
+      "grad_norm": 0.2390257085571172,
+      "learning_rate": 5.850778314546431e-06,
+      "loss": 0.3692,
+      "step": 1852
+    },
+    {
+      "epoch": 2.6855072463768117,
+      "grad_norm": 0.25246189224755156,
+      "learning_rate": 5.823939881910897e-06,
+      "loss": 0.4123,
+      "step": 1853
+    },
+    {
+      "epoch": 2.6869565217391305,
+      "grad_norm": 0.22169205705749817,
+      "learning_rate": 5.797101449275362e-06,
+      "loss": 0.3683,
+      "step": 1854
+    },
+    {
+      "epoch": 2.6884057971014492,
+      "grad_norm": 0.21896924257649364,
+      "learning_rate": 5.7702630166398285e-06,
+      "loss": 0.366,
+      "step": 1855
+    },
+    {
+      "epoch": 2.689855072463768,
+      "grad_norm": 0.22653260988190463,
+      "learning_rate": 5.743424584004295e-06,
+      "loss": 0.3597,
+      "step": 1856
+    },
+    {
+      "epoch": 2.6913043478260867,
+      "grad_norm": 0.2287381038960718,
+      "learning_rate": 5.71658615136876e-06,
+      "loss": 0.3974,
+      "step": 1857
+    },
+    {
+      "epoch": 2.692753623188406,
+      "grad_norm": 0.23347533317562097,
+      "learning_rate": 5.689747718733227e-06,
+      "loss": 0.3857,
+      "step": 1858
+    },
+    {
+      "epoch": 2.6942028985507247,
+      "grad_norm": 0.22367324176057518,
+      "learning_rate": 5.662909286097692e-06,
+      "loss": 0.3801,
+      "step": 1859
+    },
+    {
+      "epoch": 2.6956521739130435,
+      "grad_norm": 0.23472201362272851,
+      "learning_rate": 5.636070853462158e-06,
+      "loss": 0.4245,
+      "step": 1860
+    },
+    {
+      "epoch": 2.697101449275362,
+      "grad_norm": 0.3157297294018279,
+      "learning_rate": 5.6092324208266236e-06,
+      "loss": 0.3867,
+      "step": 1861
+    },
+    {
+      "epoch": 2.6985507246376814,
+      "grad_norm": 0.2291967553591276,
+      "learning_rate": 5.58239398819109e-06,
+      "loss": 0.368,
+      "step": 1862
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.22354248296465548,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 0.3865,
+      "step": 1863
+    },
+    {
+      "epoch": 2.701449275362319,
+      "grad_norm": 0.22739828851934407,
+      "learning_rate": 5.528717122920022e-06,
+      "loss": 0.4034,
+      "step": 1864
+    },
+    {
+      "epoch": 2.7028985507246377,
+      "grad_norm": 0.2267575561496053,
+      "learning_rate": 5.501878690284488e-06,
+      "loss": 0.3503,
+      "step": 1865
+    },
+    {
+      "epoch": 2.7043478260869565,
+      "grad_norm": 0.24634783666743043,
+      "learning_rate": 5.475040257648953e-06,
+      "loss": 0.3976,
+      "step": 1866
+    },
+    {
+      "epoch": 2.705797101449275,
+      "grad_norm": 0.23064788007901818,
+      "learning_rate": 5.4482018250134195e-06,
+      "loss": 0.3492,
+      "step": 1867
+    },
+    {
+      "epoch": 2.707246376811594,
+      "grad_norm": 0.2691754884654957,
+      "learning_rate": 5.421363392377885e-06,
+      "loss": 0.3994,
+      "step": 1868
+    },
+    {
+      "epoch": 2.708695652173913,
+      "grad_norm": 0.22996872590791068,
+      "learning_rate": 5.394524959742351e-06,
+      "loss": 0.3842,
+      "step": 1869
+    },
+    {
+      "epoch": 2.710144927536232,
+      "grad_norm": 0.2423580047509324,
+      "learning_rate": 5.367686527106818e-06,
+      "loss": 0.3451,
+      "step": 1870
+    },
+    {
+      "epoch": 2.7115942028985507,
+      "grad_norm": 0.23179393107934326,
+      "learning_rate": 5.340848094471283e-06,
+      "loss": 0.3142,
+      "step": 1871
+    },
+    {
+      "epoch": 2.7130434782608694,
+      "grad_norm": 0.24728239436941335,
+      "learning_rate": 5.314009661835749e-06,
+      "loss": 0.4023,
+      "step": 1872
+    },
+    {
+      "epoch": 2.7144927536231886,
+      "grad_norm": 0.25281705888035133,
+      "learning_rate": 5.2871712292002145e-06,
+      "loss": 0.3679,
+      "step": 1873
+    },
+    {
+      "epoch": 2.7159420289855074,
+      "grad_norm": 0.2503489075733349,
+      "learning_rate": 5.260332796564681e-06,
+      "loss": 0.3686,
+      "step": 1874
+    },
+    {
+      "epoch": 2.717391304347826,
+      "grad_norm": 0.2118793346562308,
+      "learning_rate": 5.233494363929147e-06,
+      "loss": 0.3503,
+      "step": 1875
+    },
+    {
+      "epoch": 2.718840579710145,
+      "grad_norm": 0.22433352553126884,
+      "learning_rate": 5.206655931293613e-06,
+      "loss": 0.3819,
+      "step": 1876
+    },
+    {
+      "epoch": 2.7202898550724637,
+      "grad_norm": 0.23356449816008817,
+      "learning_rate": 5.179817498658079e-06,
+      "loss": 0.3704,
+      "step": 1877
+    },
+    {
+      "epoch": 2.7217391304347824,
+      "grad_norm": 0.22883220912475813,
+      "learning_rate": 5.152979066022544e-06,
+      "loss": 0.3812,
+      "step": 1878
+    },
+    {
+      "epoch": 2.723188405797101,
+      "grad_norm": 0.2123879981586008,
+      "learning_rate": 5.1261406333870105e-06,
+      "loss": 0.3564,
+      "step": 1879
+    },
+    {
+      "epoch": 2.7246376811594204,
+      "grad_norm": 0.22948078088179824,
+      "learning_rate": 5.099302200751476e-06,
+      "loss": 0.3846,
+      "step": 1880
+    },
+    {
+      "epoch": 2.726086956521739,
+      "grad_norm": 0.21642023725127324,
+      "learning_rate": 5.072463768115943e-06,
+      "loss": 0.3468,
+      "step": 1881
+    },
+    {
+      "epoch": 2.727536231884058,
+      "grad_norm": 0.22925279104543358,
+      "learning_rate": 5.045625335480408e-06,
+      "loss": 0.373,
+      "step": 1882
+    },
+    {
+      "epoch": 2.7289855072463767,
+      "grad_norm": 0.23657457987131836,
+      "learning_rate": 5.018786902844874e-06,
+      "loss": 0.3966,
+      "step": 1883
+    },
+    {
+      "epoch": 2.730434782608696,
+      "grad_norm": 0.22094146109857993,
+      "learning_rate": 4.99194847020934e-06,
+      "loss": 0.3903,
+      "step": 1884
+    },
+    {
+      "epoch": 2.7318840579710146,
+      "grad_norm": 0.22239394945020372,
+      "learning_rate": 4.9651100375738055e-06,
+      "loss": 0.3746,
+      "step": 1885
+    },
+    {
+      "epoch": 2.7333333333333334,
+      "grad_norm": 0.243518585995688,
+      "learning_rate": 4.938271604938272e-06,
+      "loss": 0.4275,
+      "step": 1886
+    },
+    {
+      "epoch": 2.734782608695652,
+      "grad_norm": 0.25187515639105273,
+      "learning_rate": 4.911433172302738e-06,
+      "loss": 0.3532,
+      "step": 1887
+    },
+    {
+      "epoch": 2.736231884057971,
+      "grad_norm": 0.22632621759318555,
+      "learning_rate": 4.884594739667204e-06,
+      "loss": 0.3769,
+      "step": 1888
+    },
+    {
+      "epoch": 2.7376811594202897,
+      "grad_norm": 0.22478743881018365,
+      "learning_rate": 4.85775630703167e-06,
+      "loss": 0.3925,
+      "step": 1889
+    },
+    {
+      "epoch": 2.7391304347826084,
+      "grad_norm": 0.21774392701540912,
+      "learning_rate": 4.830917874396135e-06,
+      "loss": 0.3422,
+      "step": 1890
+    },
+    {
+      "epoch": 2.7405797101449276,
+      "grad_norm": 0.2154279005368851,
+      "learning_rate": 4.8040794417606014e-06,
+      "loss": 0.3395,
+      "step": 1891
+    },
+    {
+      "epoch": 2.7420289855072464,
+      "grad_norm": 0.250856961061175,
+      "learning_rate": 4.777241009125067e-06,
+      "loss": 0.3844,
+      "step": 1892
+    },
+    {
+      "epoch": 2.743478260869565,
+      "grad_norm": 0.22435761376804456,
+      "learning_rate": 4.750402576489534e-06,
+      "loss": 0.3767,
+      "step": 1893
+    },
+    {
+      "epoch": 2.744927536231884,
+      "grad_norm": 0.21992969440428498,
+      "learning_rate": 4.723564143853999e-06,
+      "loss": 0.3178,
+      "step": 1894
+    },
+    {
+      "epoch": 2.746376811594203,
+      "grad_norm": 0.2334336353725486,
+      "learning_rate": 4.696725711218465e-06,
+      "loss": 0.3919,
+      "step": 1895
+    },
+    {
+      "epoch": 2.747826086956522,
+      "grad_norm": 0.2216197480193231,
+      "learning_rate": 4.669887278582931e-06,
+      "loss": 0.3388,
+      "step": 1896
+    },
+    {
+      "epoch": 2.7492753623188406,
+      "grad_norm": 0.22330046792674918,
+      "learning_rate": 4.6430488459473965e-06,
+      "loss": 0.3627,
+      "step": 1897
+    },
+    {
+      "epoch": 2.7507246376811594,
+      "grad_norm": 0.22226569426691617,
+      "learning_rate": 4.616210413311863e-06,
+      "loss": 0.3657,
+      "step": 1898
+    },
+    {
+      "epoch": 2.752173913043478,
+      "grad_norm": 0.230265527395819,
+      "learning_rate": 4.589371980676329e-06,
+      "loss": 0.4176,
+      "step": 1899
+    },
+    {
+      "epoch": 2.753623188405797,
+      "grad_norm": 0.21860986911594793,
+      "learning_rate": 4.562533548040795e-06,
+      "loss": 0.3597,
+      "step": 1900
+    },
+    {
+      "epoch": 2.755072463768116,
+      "grad_norm": 0.21878736281978017,
+      "learning_rate": 4.53569511540526e-06,
+      "loss": 0.3511,
+      "step": 1901
+    },
+    {
+      "epoch": 2.756521739130435,
+      "grad_norm": 0.2126842228269991,
+      "learning_rate": 4.508856682769726e-06,
+      "loss": 0.3649,
+      "step": 1902
+    },
+    {
+      "epoch": 2.7579710144927536,
+      "grad_norm": 0.2265362594489967,
+      "learning_rate": 4.482018250134192e-06,
+      "loss": 0.3639,
+      "step": 1903
+    },
+    {
+      "epoch": 2.7594202898550724,
+      "grad_norm": 0.24180262661923574,
+      "learning_rate": 4.455179817498658e-06,
+      "loss": 0.3872,
+      "step": 1904
+    },
+    {
+      "epoch": 2.7608695652173916,
+      "grad_norm": 0.21206981799411836,
+      "learning_rate": 4.428341384863125e-06,
+      "loss": 0.319,
+      "step": 1905
+    },
+    {
+      "epoch": 2.7623188405797103,
+      "grad_norm": 2.8559315823110727,
+      "learning_rate": 4.40150295222759e-06,
+      "loss": 0.3828,
+      "step": 1906
+    },
+    {
+      "epoch": 2.763768115942029,
+      "grad_norm": 0.2165891805436599,
+      "learning_rate": 4.374664519592056e-06,
+      "loss": 0.3601,
+      "step": 1907
+    },
+    {
+      "epoch": 2.765217391304348,
+      "grad_norm": 0.2687478416264375,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 0.4265,
+      "step": 1908
+    },
+    {
+      "epoch": 2.7666666666666666,
+      "grad_norm": 0.23162147279056533,
+      "learning_rate": 4.3209876543209875e-06,
+      "loss": 0.3776,
+      "step": 1909
+    },
+    {
+      "epoch": 2.7681159420289854,
+      "grad_norm": 0.24001548884386711,
+      "learning_rate": 4.2941492216854545e-06,
+      "loss": 0.3539,
+      "step": 1910
+    },
+    {
+      "epoch": 2.769565217391304,
+      "grad_norm": 0.21822860633731908,
+      "learning_rate": 4.26731078904992e-06,
+      "loss": 0.3721,
+      "step": 1911
+    },
+    {
+      "epoch": 2.7710144927536233,
+      "grad_norm": 0.22484154136612883,
+      "learning_rate": 4.240472356414386e-06,
+      "loss": 0.3794,
+      "step": 1912
+    },
+    {
+      "epoch": 2.772463768115942,
+      "grad_norm": 0.25675934996166644,
+      "learning_rate": 4.213633923778851e-06,
+      "loss": 0.4349,
+      "step": 1913
+    },
+    {
+      "epoch": 2.773913043478261,
+      "grad_norm": 0.24379391339974027,
+      "learning_rate": 4.186795491143317e-06,
+      "loss": 0.4392,
+      "step": 1914
+    },
+    {
+      "epoch": 2.7753623188405796,
+      "grad_norm": 0.21306475644565454,
+      "learning_rate": 4.159957058507783e-06,
+      "loss": 0.3595,
+      "step": 1915
+    },
+    {
+      "epoch": 2.776811594202899,
+      "grad_norm": 0.24061893214197644,
+      "learning_rate": 4.1331186258722495e-06,
+      "loss": 0.4079,
+      "step": 1916
+    },
+    {
+      "epoch": 2.7782608695652176,
+      "grad_norm": 0.2227123946898031,
+      "learning_rate": 4.106280193236716e-06,
+      "loss": 0.3993,
+      "step": 1917
+    },
+    {
+      "epoch": 2.7797101449275363,
+      "grad_norm": 0.22460600749977255,
+      "learning_rate": 4.079441760601181e-06,
+      "loss": 0.3517,
+      "step": 1918
+    },
+    {
+      "epoch": 2.781159420289855,
+      "grad_norm": 0.22317560971512612,
+      "learning_rate": 4.052603327965647e-06,
+      "loss": 0.3392,
+      "step": 1919
+    },
+    {
+      "epoch": 2.782608695652174,
+      "grad_norm": 0.251545566462596,
+      "learning_rate": 4.025764895330112e-06,
+      "loss": 0.4172,
+      "step": 1920
+    },
+    {
+      "epoch": 2.7840579710144926,
+      "grad_norm": 0.24032524857232762,
+      "learning_rate": 3.9989264626945785e-06,
+      "loss": 0.4041,
+      "step": 1921
+    },
+    {
+      "epoch": 2.7855072463768114,
+      "grad_norm": 0.23355053943345475,
+      "learning_rate": 3.9720880300590454e-06,
+      "loss": 0.399,
+      "step": 1922
+    },
+    {
+      "epoch": 2.7869565217391306,
+      "grad_norm": 0.2650792625387716,
+      "learning_rate": 3.945249597423511e-06,
+      "loss": 0.366,
+      "step": 1923
+    },
+    {
+      "epoch": 2.7884057971014493,
+      "grad_norm": 0.24920047860144612,
+      "learning_rate": 3.918411164787977e-06,
+      "loss": 0.3717,
+      "step": 1924
+    },
+    {
+      "epoch": 2.789855072463768,
+      "grad_norm": 0.22497324802029764,
+      "learning_rate": 3.891572732152442e-06,
+      "loss": 0.3554,
+      "step": 1925
+    },
+    {
+      "epoch": 2.791304347826087,
+      "grad_norm": 0.2430456602220627,
+      "learning_rate": 3.864734299516908e-06,
+      "loss": 0.3837,
+      "step": 1926
+    },
+    {
+      "epoch": 2.792753623188406,
+      "grad_norm": 0.23237653073700315,
+      "learning_rate": 3.837895866881374e-06,
+      "loss": 0.4283,
+      "step": 1927
+    },
+    {
+      "epoch": 2.794202898550725,
+      "grad_norm": 0.2458387819745744,
+      "learning_rate": 3.8110574342458405e-06,
+      "loss": 0.3827,
+      "step": 1928
+    },
+    {
+      "epoch": 2.7956521739130435,
+      "grad_norm": 0.23093469706892966,
+      "learning_rate": 3.784219001610306e-06,
+      "loss": 0.3635,
+      "step": 1929
+    },
+    {
+      "epoch": 2.7971014492753623,
+      "grad_norm": 0.24018081788643378,
+      "learning_rate": 3.757380568974772e-06,
+      "loss": 0.3959,
+      "step": 1930
+    },
+    {
+      "epoch": 2.798550724637681,
+      "grad_norm": 0.21167980466508968,
+      "learning_rate": 3.730542136339238e-06,
+      "loss": 0.3397,
+      "step": 1931
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.21167642108349205,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.2966,
+      "step": 1932
+    },
+    {
+      "epoch": 2.8014492753623186,
+      "grad_norm": 0.22246501796322174,
+      "learning_rate": 3.6768652710681694e-06,
+      "loss": 0.3471,
+      "step": 1933
+    },
+    {
+      "epoch": 2.802898550724638,
+      "grad_norm": 0.21898837354019404,
+      "learning_rate": 3.650026838432636e-06,
+      "loss": 0.3636,
+      "step": 1934
+    },
+    {
+      "epoch": 2.8043478260869565,
+      "grad_norm": 0.22057898413689808,
+      "learning_rate": 3.6231884057971017e-06,
+      "loss": 0.3805,
+      "step": 1935
+    },
+    {
+      "epoch": 2.8057971014492753,
+      "grad_norm": 0.21911240754826977,
+      "learning_rate": 3.5963499731615674e-06,
+      "loss": 0.3804,
+      "step": 1936
+    },
+    {
+      "epoch": 2.807246376811594,
+      "grad_norm": 0.21040512826343807,
+      "learning_rate": 3.5695115405260335e-06,
+      "loss": 0.358,
+      "step": 1937
+    },
+    {
+      "epoch": 2.8086956521739133,
+      "grad_norm": 0.22517831098190927,
+      "learning_rate": 3.5426731078904992e-06,
+      "loss": 0.3814,
+      "step": 1938
+    },
+    {
+      "epoch": 2.810144927536232,
+      "grad_norm": 0.22692037568667617,
+      "learning_rate": 3.515834675254965e-06,
+      "loss": 0.405,
+      "step": 1939
+    },
+    {
+      "epoch": 2.8115942028985508,
+      "grad_norm": 0.21054577410504252,
+      "learning_rate": 3.4889962426194315e-06,
+      "loss": 0.3231,
+      "step": 1940
+    },
+    {
+      "epoch": 2.8130434782608695,
+      "grad_norm": 0.21603644115846818,
+      "learning_rate": 3.462157809983897e-06,
+      "loss": 0.3606,
+      "step": 1941
+    },
+    {
+      "epoch": 2.8144927536231883,
+      "grad_norm": 0.21532363538849883,
+      "learning_rate": 3.435319377348363e-06,
+      "loss": 0.3626,
+      "step": 1942
+    },
+    {
+      "epoch": 2.815942028985507,
+      "grad_norm": 0.22447387294670362,
+      "learning_rate": 3.408480944712829e-06,
+      "loss": 0.3614,
+      "step": 1943
+    },
+    {
+      "epoch": 2.8173913043478263,
+      "grad_norm": 0.24462971782857168,
+      "learning_rate": 3.3816425120772947e-06,
+      "loss": 0.4489,
+      "step": 1944
+    },
+    {
+      "epoch": 2.818840579710145,
+      "grad_norm": 0.2246947387946271,
+      "learning_rate": 3.3548040794417613e-06,
+      "loss": 0.3898,
+      "step": 1945
+    },
+    {
+      "epoch": 2.8202898550724638,
+      "grad_norm": 0.24533916969836606,
+      "learning_rate": 3.327965646806227e-06,
+      "loss": 0.3777,
+      "step": 1946
+    },
+    {
+      "epoch": 2.8217391304347825,
+      "grad_norm": 0.37104571577158546,
+      "learning_rate": 3.3011272141706927e-06,
+      "loss": 0.4285,
+      "step": 1947
+    },
+    {
+      "epoch": 2.8231884057971013,
+      "grad_norm": 0.2252493784676985,
+      "learning_rate": 3.2742887815351584e-06,
+      "loss": 0.4041,
+      "step": 1948
+    },
+    {
+      "epoch": 2.8246376811594205,
+      "grad_norm": 0.23358946546345724,
+      "learning_rate": 3.247450348899624e-06,
+      "loss": 0.4158,
+      "step": 1949
+    },
+    {
+      "epoch": 2.8260869565217392,
+      "grad_norm": 0.21586160621892583,
+      "learning_rate": 3.22061191626409e-06,
+      "loss": 0.3488,
+      "step": 1950
+    },
+    {
+      "epoch": 2.827536231884058,
+      "grad_norm": 0.3055949920007374,
+      "learning_rate": 3.1937734836285568e-06,
+      "loss": 0.3766,
+      "step": 1951
+    },
+    {
+      "epoch": 2.8289855072463768,
+      "grad_norm": 0.2093497712044223,
+      "learning_rate": 3.1669350509930225e-06,
+      "loss": 0.3491,
+      "step": 1952
+    },
+    {
+      "epoch": 2.8304347826086955,
+      "grad_norm": 0.21283625140471957,
+      "learning_rate": 3.140096618357488e-06,
+      "loss": 0.3869,
+      "step": 1953
+    },
+    {
+      "epoch": 2.8318840579710143,
+      "grad_norm": 0.21477929276167043,
+      "learning_rate": 3.113258185721954e-06,
+      "loss": 0.3494,
+      "step": 1954
+    },
+    {
+      "epoch": 2.8333333333333335,
+      "grad_norm": 0.2106104901914822,
+      "learning_rate": 3.0864197530864196e-06,
+      "loss": 0.36,
+      "step": 1955
+    },
+    {
+      "epoch": 2.8347826086956522,
+      "grad_norm": 0.23305285262666858,
+      "learning_rate": 3.0595813204508857e-06,
+      "loss": 0.4305,
+      "step": 1956
+    },
+    {
+      "epoch": 2.836231884057971,
+      "grad_norm": 0.22466858251147637,
+      "learning_rate": 3.032742887815352e-06,
+      "loss": 0.3847,
+      "step": 1957
+    },
+    {
+      "epoch": 2.8376811594202898,
+      "grad_norm": 0.24561966059885118,
+      "learning_rate": 3.005904455179818e-06,
+      "loss": 0.372,
+      "step": 1958
+    },
+    {
+      "epoch": 2.839130434782609,
+      "grad_norm": 0.24884208733974736,
+      "learning_rate": 2.9790660225442837e-06,
+      "loss": 0.3848,
+      "step": 1959
+    },
+    {
+      "epoch": 2.8405797101449277,
+      "grad_norm": 0.23587441969983608,
+      "learning_rate": 2.9522275899087494e-06,
+      "loss": 0.3723,
+      "step": 1960
+    },
+    {
+      "epoch": 2.8420289855072465,
+      "grad_norm": 0.22289741421781245,
+      "learning_rate": 2.9253891572732155e-06,
+      "loss": 0.3639,
+      "step": 1961
+    },
+    {
+      "epoch": 2.8434782608695652,
+      "grad_norm": 0.21516329109133195,
+      "learning_rate": 2.898550724637681e-06,
+      "loss": 0.3471,
+      "step": 1962
+    },
+    {
+      "epoch": 2.844927536231884,
+      "grad_norm": 0.22475075634579428,
+      "learning_rate": 2.8717122920021473e-06,
+      "loss": 0.3948,
+      "step": 1963
+    },
+    {
+      "epoch": 2.8463768115942027,
+      "grad_norm": 0.246650267181553,
+      "learning_rate": 2.8448738593666134e-06,
+      "loss": 0.3866,
+      "step": 1964
+    },
+    {
+      "epoch": 2.8478260869565215,
+      "grad_norm": 0.20439464539156504,
+      "learning_rate": 2.818035426731079e-06,
+      "loss": 0.3293,
+      "step": 1965
+    },
+    {
+      "epoch": 2.8492753623188407,
+      "grad_norm": 0.2219043464799365,
+      "learning_rate": 2.791196994095545e-06,
+      "loss": 0.3424,
+      "step": 1966
+    },
+    {
+      "epoch": 2.8507246376811595,
+      "grad_norm": 0.23028216886417335,
+      "learning_rate": 2.764358561460011e-06,
+      "loss": 0.3736,
+      "step": 1967
+    },
+    {
+      "epoch": 2.8521739130434782,
+      "grad_norm": 0.2692112527395984,
+      "learning_rate": 2.7375201288244767e-06,
+      "loss": 0.3615,
+      "step": 1968
+    },
+    {
+      "epoch": 2.853623188405797,
+      "grad_norm": 0.21852360587413638,
+      "learning_rate": 2.7106816961889424e-06,
+      "loss": 0.352,
+      "step": 1969
+    },
+    {
+      "epoch": 2.855072463768116,
+      "grad_norm": 0.23283633142906562,
+      "learning_rate": 2.683843263553409e-06,
+      "loss": 0.3831,
+      "step": 1970
+    },
+    {
+      "epoch": 2.856521739130435,
+      "grad_norm": 0.22253240700761454,
+      "learning_rate": 2.6570048309178746e-06,
+      "loss": 0.3404,
+      "step": 1971
+    },
+    {
+      "epoch": 2.8579710144927537,
+      "grad_norm": 0.2186544446437513,
+      "learning_rate": 2.6301663982823403e-06,
+      "loss": 0.3482,
+      "step": 1972
+    },
+    {
+      "epoch": 2.8594202898550725,
+      "grad_norm": 0.24369866231612267,
+      "learning_rate": 2.6033279656468065e-06,
+      "loss": 0.4298,
+      "step": 1973
+    },
+    {
+      "epoch": 2.860869565217391,
+      "grad_norm": 0.22628698320755639,
+      "learning_rate": 2.576489533011272e-06,
+      "loss": 0.3456,
+      "step": 1974
+    },
+    {
+      "epoch": 2.86231884057971,
+      "grad_norm": 0.21467256321432143,
+      "learning_rate": 2.549651100375738e-06,
+      "loss": 0.344,
+      "step": 1975
+    },
+    {
+      "epoch": 2.8637681159420287,
+      "grad_norm": 0.22488105416337417,
+      "learning_rate": 2.522812667740204e-06,
+      "loss": 0.3897,
+      "step": 1976
+    },
+    {
+      "epoch": 2.865217391304348,
+      "grad_norm": 0.21172401558442616,
+      "learning_rate": 2.49597423510467e-06,
+      "loss": 0.3561,
+      "step": 1977
+    },
+    {
+      "epoch": 2.8666666666666667,
+      "grad_norm": 0.23817304874870301,
+      "learning_rate": 2.469135802469136e-06,
+      "loss": 0.3469,
+      "step": 1978
+    },
+    {
+      "epoch": 2.8681159420289855,
+      "grad_norm": 0.23063863203760862,
+      "learning_rate": 2.442297369833602e-06,
+      "loss": 0.3744,
+      "step": 1979
+    },
+    {
+      "epoch": 2.869565217391304,
+      "grad_norm": 0.2279855386844957,
+      "learning_rate": 2.4154589371980677e-06,
+      "loss": 0.4174,
+      "step": 1980
+    },
+    {
+      "epoch": 2.8710144927536234,
+      "grad_norm": 0.20446797527153535,
+      "learning_rate": 2.3886205045625334e-06,
+      "loss": 0.3103,
+      "step": 1981
+    },
+    {
+      "epoch": 2.872463768115942,
+      "grad_norm": 0.2313299210534719,
+      "learning_rate": 2.3617820719269995e-06,
+      "loss": 0.3971,
+      "step": 1982
+    },
+    {
+      "epoch": 2.873913043478261,
+      "grad_norm": 0.21186439987907596,
+      "learning_rate": 2.3349436392914656e-06,
+      "loss": 0.3081,
+      "step": 1983
+    },
+    {
+      "epoch": 2.8753623188405797,
+      "grad_norm": 0.21930666784044156,
+      "learning_rate": 2.3081052066559313e-06,
+      "loss": 0.3594,
+      "step": 1984
+    },
+    {
+      "epoch": 2.8768115942028984,
+      "grad_norm": 0.211911521071487,
+      "learning_rate": 2.2812667740203974e-06,
+      "loss": 0.3347,
+      "step": 1985
+    },
+    {
+      "epoch": 2.878260869565217,
+      "grad_norm": 0.6711232292058408,
+      "learning_rate": 2.254428341384863e-06,
+      "loss": 0.4081,
+      "step": 1986
+    },
+    {
+      "epoch": 2.879710144927536,
+      "grad_norm": 0.22267835282728443,
+      "learning_rate": 2.227589908749329e-06,
+      "loss": 0.3665,
+      "step": 1987
+    },
+    {
+      "epoch": 2.881159420289855,
+      "grad_norm": 0.22181600928646944,
+      "learning_rate": 2.200751476113795e-06,
+      "loss": 0.4074,
+      "step": 1988
+    },
+    {
+      "epoch": 2.882608695652174,
+      "grad_norm": 0.23661315520430906,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 0.3856,
+      "step": 1989
+    },
+    {
+      "epoch": 2.8840579710144927,
+      "grad_norm": 0.20941940533698208,
+      "learning_rate": 2.1470746108427272e-06,
+      "loss": 0.3375,
+      "step": 1990
+    },
+    {
+      "epoch": 2.8855072463768114,
+      "grad_norm": 0.21125596048646939,
+      "learning_rate": 2.120236178207193e-06,
+      "loss": 0.3557,
+      "step": 1991
+    },
+    {
+      "epoch": 2.8869565217391306,
+      "grad_norm": 0.24322069203581723,
+      "learning_rate": 2.0933977455716586e-06,
+      "loss": 0.4068,
+      "step": 1992
+    },
+    {
+      "epoch": 2.8884057971014494,
+      "grad_norm": 0.20858954044987255,
+      "learning_rate": 2.0665593129361248e-06,
+      "loss": 0.3168,
+      "step": 1993
+    },
+    {
+      "epoch": 2.889855072463768,
+      "grad_norm": 0.2236295696408854,
+      "learning_rate": 2.0397208803005905e-06,
+      "loss": 0.369,
+      "step": 1994
+    },
+    {
+      "epoch": 2.891304347826087,
+      "grad_norm": 0.22156324434803293,
+      "learning_rate": 2.012882447665056e-06,
+      "loss": 0.3437,
+      "step": 1995
+    },
+    {
+      "epoch": 2.8927536231884057,
+      "grad_norm": 0.21259380447613244,
+      "learning_rate": 1.9860440150295227e-06,
+      "loss": 0.3649,
+      "step": 1996
+    },
+    {
+      "epoch": 2.8942028985507244,
+      "grad_norm": 0.21497355911092556,
+      "learning_rate": 1.9592055823939884e-06,
+      "loss": 0.3651,
+      "step": 1997
+    },
+    {
+      "epoch": 2.8956521739130436,
+      "grad_norm": 0.224066475424279,
+      "learning_rate": 1.932367149758454e-06,
+      "loss": 0.3813,
+      "step": 1998
+    },
+    {
+      "epoch": 2.8971014492753624,
+      "grad_norm": 0.2361099140230203,
+      "learning_rate": 1.9055287171229203e-06,
+      "loss": 0.4119,
+      "step": 1999
+    },
+    {
+      "epoch": 2.898550724637681,
+      "grad_norm": 0.2285383121878008,
+      "learning_rate": 1.878690284487386e-06,
+      "loss": 0.412,
+      "step": 2000
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.22744543201115258,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.3667,
+      "step": 2001
+    },
+    {
+      "epoch": 2.901449275362319,
+      "grad_norm": 0.20438497271453532,
+      "learning_rate": 1.825013419216318e-06,
+      "loss": 0.3392,
+      "step": 2002
+    },
+    {
+      "epoch": 2.902898550724638,
+      "grad_norm": 0.22154818027410342,
+      "learning_rate": 1.7981749865807837e-06,
+      "loss": 0.3657,
+      "step": 2003
+    },
+    {
+      "epoch": 2.9043478260869566,
+      "grad_norm": 0.23655652039164077,
+      "learning_rate": 1.7713365539452496e-06,
+      "loss": 0.3472,
+      "step": 2004
+    },
+    {
+      "epoch": 2.9057971014492754,
+      "grad_norm": 0.21968133490213995,
+      "learning_rate": 1.7444981213097157e-06,
+      "loss": 0.3875,
+      "step": 2005
+    },
+    {
+      "epoch": 2.907246376811594,
+      "grad_norm": 0.20672773395313468,
+      "learning_rate": 1.7176596886741814e-06,
+      "loss": 0.3327,
+      "step": 2006
+    },
+    {
+      "epoch": 2.908695652173913,
+      "grad_norm": 0.2205358363917607,
+      "learning_rate": 1.6908212560386474e-06,
+      "loss": 0.389,
+      "step": 2007
+    },
+    {
+      "epoch": 2.9101449275362317,
+      "grad_norm": 0.23819173128880847,
+      "learning_rate": 1.6639828234031135e-06,
+      "loss": 0.3942,
+      "step": 2008
+    },
+    {
+      "epoch": 2.911594202898551,
+      "grad_norm": 0.2059964263544637,
+      "learning_rate": 1.6371443907675792e-06,
+      "loss": 0.3568,
+      "step": 2009
+    },
+    {
+      "epoch": 2.9130434782608696,
+      "grad_norm": 0.2171138770231751,
+      "learning_rate": 1.610305958132045e-06,
+      "loss": 0.38,
+      "step": 2010
+    },
+    {
+      "epoch": 2.9144927536231884,
+      "grad_norm": 0.23005385529348343,
+      "learning_rate": 1.5834675254965112e-06,
+      "loss": 0.3655,
+      "step": 2011
+    },
+    {
+      "epoch": 2.915942028985507,
+      "grad_norm": 0.21016123356167332,
+      "learning_rate": 1.556629092860977e-06,
+      "loss": 0.3672,
+      "step": 2012
+    },
+    {
+      "epoch": 2.9173913043478263,
+      "grad_norm": 0.22483884303098328,
+      "learning_rate": 1.5297906602254428e-06,
+      "loss": 0.3581,
+      "step": 2013
+    },
+    {
+      "epoch": 2.918840579710145,
+      "grad_norm": 0.22322264421802532,
+      "learning_rate": 1.502952227589909e-06,
+      "loss": 0.3537,
+      "step": 2014
+    },
+    {
+      "epoch": 2.920289855072464,
+      "grad_norm": 0.2311031178966409,
+      "learning_rate": 1.4761137949543747e-06,
+      "loss": 0.405,
+      "step": 2015
+    },
+    {
+      "epoch": 2.9217391304347826,
+      "grad_norm": 0.22217953477714308,
+      "learning_rate": 1.4492753623188406e-06,
+      "loss": 0.4127,
+      "step": 2016
+    },
+    {
+      "epoch": 2.9231884057971014,
+      "grad_norm": 0.22709560866583908,
+      "learning_rate": 1.4224369296833067e-06,
+      "loss": 0.3834,
+      "step": 2017
+    },
+    {
+      "epoch": 2.92463768115942,
+      "grad_norm": 0.22408928030872904,
+      "learning_rate": 1.3955984970477724e-06,
+      "loss": 0.3976,
+      "step": 2018
+    },
+    {
+      "epoch": 2.926086956521739,
+      "grad_norm": 0.2106162699815264,
+      "learning_rate": 1.3687600644122383e-06,
+      "loss": 0.371,
+      "step": 2019
+    },
+    {
+      "epoch": 2.927536231884058,
+      "grad_norm": 0.2018244638135191,
+      "learning_rate": 1.3419216317767045e-06,
+      "loss": 0.3538,
+      "step": 2020
+    },
+    {
+      "epoch": 2.928985507246377,
+      "grad_norm": 0.2347309750501861,
+      "learning_rate": 1.3150831991411702e-06,
+      "loss": 0.4322,
+      "step": 2021
+    },
+    {
+      "epoch": 2.9304347826086956,
+      "grad_norm": 0.23350101438562867,
+      "learning_rate": 1.288244766505636e-06,
+      "loss": 0.3775,
+      "step": 2022
+    },
+    {
+      "epoch": 2.9318840579710144,
+      "grad_norm": 0.8735079283802203,
+      "learning_rate": 1.261406333870102e-06,
+      "loss": 0.4373,
+      "step": 2023
+    },
+    {
+      "epoch": 2.9333333333333336,
+      "grad_norm": 0.8056038885314158,
+      "learning_rate": 1.234567901234568e-06,
+      "loss": 0.3715,
+      "step": 2024
+    },
+    {
+      "epoch": 2.9347826086956523,
+      "grad_norm": 0.21993418679321491,
+      "learning_rate": 1.2077294685990338e-06,
+      "loss": 0.3555,
+      "step": 2025
+    },
+    {
+      "epoch": 2.936231884057971,
+      "grad_norm": 0.23405472869389934,
+      "learning_rate": 1.1808910359634997e-06,
+      "loss": 0.3752,
+      "step": 2026
+    },
+    {
+      "epoch": 2.93768115942029,
+      "grad_norm": 0.2313082257049979,
+      "learning_rate": 1.1540526033279657e-06,
+      "loss": 0.4198,
+      "step": 2027
+    },
+    {
+      "epoch": 2.9391304347826086,
+      "grad_norm": 0.22480212739012523,
+      "learning_rate": 1.1272141706924316e-06,
+      "loss": 0.3798,
+      "step": 2028
+    },
+    {
+      "epoch": 2.9405797101449274,
+      "grad_norm": 0.21838170742723495,
+      "learning_rate": 1.1003757380568975e-06,
+      "loss": 0.3796,
+      "step": 2029
+    },
+    {
+      "epoch": 2.942028985507246,
+      "grad_norm": 0.21405428252058262,
+      "learning_rate": 1.0735373054213636e-06,
+      "loss": 0.3559,
+      "step": 2030
+    },
+    {
+      "epoch": 2.9434782608695653,
+      "grad_norm": 0.21549860413480246,
+      "learning_rate": 1.0466988727858293e-06,
+      "loss": 0.3334,
+      "step": 2031
+    },
+    {
+      "epoch": 2.944927536231884,
+      "grad_norm": 0.28906789347226164,
+      "learning_rate": 1.0198604401502952e-06,
+      "loss": 0.3527,
+      "step": 2032
+    },
+    {
+      "epoch": 2.946376811594203,
+      "grad_norm": 0.2064045807699817,
+      "learning_rate": 9.930220075147614e-07,
+      "loss": 0.3469,
+      "step": 2033
+    },
+    {
+      "epoch": 2.9478260869565216,
+      "grad_norm": 0.22024850212103755,
+      "learning_rate": 9.66183574879227e-07,
+      "loss": 0.3803,
+      "step": 2034
+    },
+    {
+      "epoch": 2.949275362318841,
+      "grad_norm": 0.21800534224417312,
+      "learning_rate": 9.39345142243693e-07,
+      "loss": 0.4064,
+      "step": 2035
+    },
+    {
+      "epoch": 2.9507246376811596,
+      "grad_norm": 0.2201865038559122,
+      "learning_rate": 9.12506709608159e-07,
+      "loss": 0.3785,
+      "step": 2036
+    },
+    {
+      "epoch": 2.9521739130434783,
+      "grad_norm": 0.2210475513425969,
+      "learning_rate": 8.856682769726248e-07,
+      "loss": 0.4036,
+      "step": 2037
+    },
+    {
+      "epoch": 2.953623188405797,
+      "grad_norm": 0.1984962905202788,
+      "learning_rate": 8.588298443370907e-07,
+      "loss": 0.3271,
+      "step": 2038
+    },
+    {
+      "epoch": 2.955072463768116,
+      "grad_norm": 0.24847445353598208,
+      "learning_rate": 8.319914117015567e-07,
+      "loss": 0.3313,
+      "step": 2039
+    },
+    {
+      "epoch": 2.9565217391304346,
+      "grad_norm": 0.2210563477725767,
+      "learning_rate": 8.051529790660226e-07,
+      "loss": 0.3857,
+      "step": 2040
+    },
+    {
+      "epoch": 2.9579710144927533,
+      "grad_norm": 0.21300962762024758,
+      "learning_rate": 7.783145464304885e-07,
+      "loss": 0.3656,
+      "step": 2041
+    },
+    {
+      "epoch": 2.9594202898550726,
+      "grad_norm": 0.20426722668302832,
+      "learning_rate": 7.514761137949545e-07,
+      "loss": 0.3239,
+      "step": 2042
+    },
+    {
+      "epoch": 2.9608695652173913,
+      "grad_norm": 0.21577157388977716,
+      "learning_rate": 7.246376811594203e-07,
+      "loss": 0.3689,
+      "step": 2043
+    },
+    {
+      "epoch": 2.96231884057971,
+      "grad_norm": 0.1989269908554471,
+      "learning_rate": 6.977992485238862e-07,
+      "loss": 0.2938,
+      "step": 2044
+    },
+    {
+      "epoch": 2.963768115942029,
+      "grad_norm": 0.22654103161247222,
+      "learning_rate": 6.709608158883522e-07,
+      "loss": 0.3623,
+      "step": 2045
+    },
+    {
+      "epoch": 2.965217391304348,
+      "grad_norm": 0.22926355644211238,
+      "learning_rate": 6.44122383252818e-07,
+      "loss": 0.3641,
+      "step": 2046
+    },
+    {
+      "epoch": 2.966666666666667,
+      "grad_norm": 0.22858125913553615,
+      "learning_rate": 6.17283950617284e-07,
+      "loss": 0.3596,
+      "step": 2047
+    },
+    {
+      "epoch": 2.9681159420289855,
+      "grad_norm": 0.23631989683689095,
+      "learning_rate": 5.904455179817499e-07,
+      "loss": 0.3811,
+      "step": 2048
+    },
+    {
+      "epoch": 2.9695652173913043,
+      "grad_norm": 0.2191396177656684,
+      "learning_rate": 5.636070853462158e-07,
+      "loss": 0.3804,
+      "step": 2049
+    },
+    {
+      "epoch": 2.971014492753623,
+      "grad_norm": 0.222858361371568,
+      "learning_rate": 5.367686527106818e-07,
+      "loss": 0.4073,
+      "step": 2050
+    },
+    {
+      "epoch": 2.972463768115942,
+      "grad_norm": 0.20318427662477487,
+      "learning_rate": 5.099302200751476e-07,
+      "loss": 0.3697,
+      "step": 2051
+    },
+    {
+      "epoch": 2.973913043478261,
+      "grad_norm": 0.217758620275763,
+      "learning_rate": 4.830917874396135e-07,
+      "loss": 0.4074,
+      "step": 2052
+    },
+    {
+      "epoch": 2.97536231884058,
+      "grad_norm": 0.22085137440848657,
+      "learning_rate": 4.562533548040795e-07,
+      "loss": 0.3512,
+      "step": 2053
+    },
+    {
+      "epoch": 2.9768115942028985,
+      "grad_norm": 0.22012562095031038,
+      "learning_rate": 4.2941492216854536e-07,
+      "loss": 0.3802,
+      "step": 2054
+    },
+    {
+      "epoch": 2.9782608695652173,
+      "grad_norm": 0.22618508012844524,
+      "learning_rate": 4.025764895330113e-07,
+      "loss": 0.374,
+      "step": 2055
+    },
+    {
+      "epoch": 2.9797101449275365,
+      "grad_norm": 0.20800555397382295,
+      "learning_rate": 3.7573805689747724e-07,
+      "loss": 0.33,
+      "step": 2056
+    },
+    {
+      "epoch": 2.9811594202898553,
+      "grad_norm": 0.2132367722323136,
+      "learning_rate": 3.488996242619431e-07,
+      "loss": 0.3509,
+      "step": 2057
+    },
+    {
+      "epoch": 2.982608695652174,
+      "grad_norm": 0.236211820119767,
+      "learning_rate": 3.22061191626409e-07,
+      "loss": 0.4553,
+      "step": 2058
+    },
+    {
+      "epoch": 2.9840579710144928,
+      "grad_norm": 0.21731210967631273,
+      "learning_rate": 2.9522275899087494e-07,
+      "loss": 0.4114,
+      "step": 2059
+    },
+    {
+      "epoch": 2.9855072463768115,
+      "grad_norm": 0.22337130008338704,
+      "learning_rate": 2.683843263553409e-07,
+      "loss": 0.4264,
+      "step": 2060
+    },
+    {
+      "epoch": 2.9869565217391303,
+      "grad_norm": 0.20921981714387702,
+      "learning_rate": 2.4154589371980677e-07,
+      "loss": 0.3334,
+      "step": 2061
+    },
+    {
+      "epoch": 2.988405797101449,
+      "grad_norm": 0.2065612804088296,
+      "learning_rate": 2.1470746108427268e-07,
+      "loss": 0.3262,
+      "step": 2062
+    },
+    {
+      "epoch": 2.9898550724637682,
+      "grad_norm": 0.2154820919743264,
+      "learning_rate": 1.8786902844873862e-07,
+      "loss": 0.3863,
+      "step": 2063
+    },
+    {
+      "epoch": 2.991304347826087,
+      "grad_norm": 4.776049818404092,
+      "learning_rate": 1.610305958132045e-07,
+      "loss": 0.4863,
+      "step": 2064
+    },
+    {
+      "epoch": 2.9927536231884058,
+      "grad_norm": 0.21262659774660408,
+      "learning_rate": 1.3419216317767045e-07,
+      "loss": 0.3271,
+      "step": 2065
+    },
+    {
+      "epoch": 2.9942028985507245,
+      "grad_norm": 0.2166325401832577,
+      "learning_rate": 1.0735373054213634e-07,
+      "loss": 0.3759,
+      "step": 2066
+    },
+    {
+      "epoch": 2.9956521739130437,
+      "grad_norm": 0.22097635962706288,
+      "learning_rate": 8.051529790660226e-08,
+      "loss": 0.369,
+      "step": 2067
+    },
+    {
+      "epoch": 2.9971014492753625,
+      "grad_norm": 0.2236357694002658,
+      "learning_rate": 5.367686527106817e-08,
+      "loss": 0.4169,
+      "step": 2068
+    },
+    {
+      "epoch": 2.9985507246376812,
+      "grad_norm": 0.21261718905247393,
+      "learning_rate": 2.6838432635534085e-08,
+      "loss": 0.3464,
+      "step": 2069
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.23176378642380113,
+      "learning_rate": 0.0,
+      "loss": 0.3583,
+      "step": 2070
+    },
+    {
+      "epoch": 3.0,
+      "step": 2070,
+      "total_flos": 1.7567558587184579e+18,
+      "train_loss": 0.21386527973384672,
+      "train_runtime": 90615.9255,
+      "train_samples_per_second": 0.365,
+      "train_steps_per_second": 0.023
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2070,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.7567558587184579e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}