diff --git "a/checkpoint-1250/trainer_state.json" "b/checkpoint-1250/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1250/trainer_state.json"
@@ -0,0 +1,4408 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0032,
+      "grad_norm": 60.66320037841797,
+      "learning_rate": 3.1746031746031746e-06,
+      "loss": 6.1881,
+      "step": 2
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 42.748294830322266,
+      "learning_rate": 6.349206349206349e-06,
+      "loss": 5.7668,
+      "step": 4
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 13.887235641479492,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 5.0841,
+      "step": 6
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 12.488709449768066,
+      "learning_rate": 1.2698412698412699e-05,
+      "loss": 4.7192,
+      "step": 8
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 7.208029270172119,
+      "learning_rate": 1.5873015873015872e-05,
+      "loss": 4.5679,
+      "step": 10
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 7.34755277633667,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 4.368,
+      "step": 12
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 7.489256858825684,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 4.1398,
+      "step": 14
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 4.802424430847168,
+      "learning_rate": 2.5396825396825397e-05,
+      "loss": 4.0532,
+      "step": 16
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 4.138615131378174,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 3.7559,
+      "step": 18
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 5.362161636352539,
+      "learning_rate": 3.1746031746031745e-05,
+      "loss": 3.4854,
+      "step": 20
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 3.878138303756714,
+      "learning_rate": 3.492063492063492e-05,
+      "loss": 3.5869,
+      "step": 22
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 3.480282783508301,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 3.2734,
+      "step": 24
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 3.527137279510498,
+      "learning_rate": 4.126984126984127e-05,
+      "loss": 3.1484,
+      "step": 26
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 3.094705820083618,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 3.0356,
+      "step": 28
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 4.064608573913574,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 3.0157,
+      "step": 30
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 3.166187047958374,
+      "learning_rate": 5.0793650793650794e-05,
+      "loss": 2.9229,
+      "step": 32
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 2.909856081008911,
+      "learning_rate": 5.396825396825397e-05,
+      "loss": 2.8386,
+      "step": 34
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 3.2195472717285156,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 2.7819,
+      "step": 36
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 2.603515625,
+      "learning_rate": 6.0317460317460316e-05,
+      "loss": 2.7116,
+      "step": 38
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 3.353161096572876,
+      "learning_rate": 6.349206349206349e-05,
+      "loss": 2.7137,
+      "step": 40
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 2.7278943061828613,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 2.6213,
+      "step": 42
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 3.275580883026123,
+      "learning_rate": 6.984126984126984e-05,
+      "loss": 2.5585,
+      "step": 44
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 2.934300422668457,
+      "learning_rate": 7.301587301587302e-05,
+      "loss": 2.6126,
+      "step": 46
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 2.483461380004883,
+      "learning_rate": 7.619047619047618e-05,
+      "loss": 2.4616,
+      "step": 48
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5167787075042725,
+      "learning_rate": 7.936507936507937e-05,
+      "loss": 2.4316,
+      "step": 50
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 2.211185932159424,
+      "learning_rate": 8.253968253968255e-05,
+      "loss": 2.63,
+      "step": 52
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 3.2666075229644775,
+      "learning_rate": 8.571428571428571e-05,
+      "loss": 2.4892,
+      "step": 54
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 2.849605083465576,
+      "learning_rate": 8.888888888888889e-05,
+      "loss": 2.4584,
+      "step": 56
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 3.0103588104248047,
+      "learning_rate": 9.206349206349206e-05,
+      "loss": 2.2766,
+      "step": 58
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 2.306534767150879,
+      "learning_rate": 9.523809523809524e-05,
+      "loss": 2.4357,
+      "step": 60
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 2.3400485515594482,
+      "learning_rate": 9.841269841269841e-05,
+      "loss": 2.4118,
+      "step": 62
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 2.583407163619995,
+      "learning_rate": 9.99998248790669e-05,
+      "loss": 2.2326,
+      "step": 64
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 2.57265567779541,
+      "learning_rate": 9.999842391896222e-05,
+      "loss": 2.2923,
+      "step": 66
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 2.307471990585327,
+      "learning_rate": 9.999562203800676e-05,
+      "loss": 2.2235,
+      "step": 68
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 2.357084035873413,
+      "learning_rate": 9.999141931470729e-05,
+      "loss": 2.2896,
+      "step": 70
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 2.6181581020355225,
+      "learning_rate": 9.998581586682116e-05,
+      "loss": 2.3015,
+      "step": 72
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 2.253117799758911,
+      "learning_rate": 9.997881185135307e-05,
+      "loss": 2.1824,
+      "step": 74
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 2.546729326248169,
+      "learning_rate": 9.997040746455062e-05,
+      "loss": 2.1502,
+      "step": 76
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 2.4144699573516846,
+      "learning_rate": 9.996060294189887e-05,
+      "loss": 2.3715,
+      "step": 78
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 2.3093016147613525,
+      "learning_rate": 9.994939855811362e-05,
+      "loss": 2.2753,
+      "step": 80
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 2.5628409385681152,
+      "learning_rate": 9.993679462713395e-05,
+      "loss": 2.3152,
+      "step": 82
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 2.549136161804199,
+      "learning_rate": 9.992279150211314e-05,
+      "loss": 2.1171,
+      "step": 84
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 2.4570610523223877,
+      "learning_rate": 9.990738957540896e-05,
+      "loss": 2.2414,
+      "step": 86
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 2.256564140319824,
+      "learning_rate": 9.989058927857263e-05,
+      "loss": 2.1324,
+      "step": 88
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 2.818751811981201,
+      "learning_rate": 9.987239108233668e-05,
+      "loss": 2.184,
+      "step": 90
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 2.432871103286743,
+      "learning_rate": 9.985279549660185e-05,
+      "loss": 2.1899,
+      "step": 92
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 2.1021323204040527,
+      "learning_rate": 9.983180307042274e-05,
+      "loss": 2.1064,
+      "step": 94
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 2.7487058639526367,
+      "learning_rate": 9.980941439199246e-05,
+      "loss": 2.2197,
+      "step": 96
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 2.82835054397583,
+      "learning_rate": 9.97856300886261e-05,
+      "loss": 2.2048,
+      "step": 98
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.25872802734375,
+      "learning_rate": 9.976045082674319e-05,
+      "loss": 2.1002,
+      "step": 100
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 2.040614366531372,
+      "learning_rate": 9.973387731184902e-05,
+      "loss": 2.1031,
+      "step": 102
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 2.437248706817627,
+      "learning_rate": 9.97059102885149e-05,
+      "loss": 2.1416,
+      "step": 104
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 2.0928878784179688,
+      "learning_rate": 9.967655054035727e-05,
+      "loss": 2.1576,
+      "step": 106
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 2.2243545055389404,
+      "learning_rate": 9.964579889001569e-05,
+      "loss": 1.9863,
+      "step": 108
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 2.1860439777374268,
+      "learning_rate": 9.961365619912989e-05,
+      "loss": 2.0016,
+      "step": 110
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 2.527122735977173,
+      "learning_rate": 9.95801233683156e-05,
+      "loss": 2.1272,
+      "step": 112
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 1.8613876104354858,
+      "learning_rate": 9.954520133713924e-05,
+      "loss": 2.2001,
+      "step": 114
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 2.115910530090332,
+      "learning_rate": 9.950889108409172e-05,
+      "loss": 2.0871,
+      "step": 116
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 2.361309051513672,
+      "learning_rate": 9.947119362656092e-05,
+      "loss": 2.017,
+      "step": 118
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 2.09470272064209,
+      "learning_rate": 9.94321100208032e-05,
+      "loss": 2.1847,
+      "step": 120
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 1.9747451543807983,
+      "learning_rate": 9.939164136191384e-05,
+      "loss": 2.324,
+      "step": 122
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 1.8229223489761353,
+      "learning_rate": 9.934978878379636e-05,
+      "loss": 2.1454,
+      "step": 124
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 1.9113378524780273,
+      "learning_rate": 9.930655345913071e-05,
+      "loss": 2.0096,
+      "step": 126
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 2.385289192199707,
+      "learning_rate": 9.926193659934043e-05,
+      "loss": 2.1029,
+      "step": 128
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 2.101463794708252,
+      "learning_rate": 9.921593945455869e-05,
+      "loss": 2.0172,
+      "step": 130
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 2.2676024436950684,
+      "learning_rate": 9.916856331359335e-05,
+      "loss": 1.9966,
+      "step": 132
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 2.0168168544769287,
+      "learning_rate": 9.911980950389067e-05,
+      "loss": 2.1807,
+      "step": 134
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 2.1054186820983887,
+      "learning_rate": 9.906967939149831e-05,
+      "loss": 1.9759,
+      "step": 136
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 2.3354573249816895,
+      "learning_rate": 9.901817438102695e-05,
+      "loss": 1.995,
+      "step": 138
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 2.2721822261810303,
+      "learning_rate": 9.896529591561093e-05,
+      "loss": 2.2239,
+      "step": 140
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 1.9209738969802856,
+      "learning_rate": 9.891104547686782e-05,
+      "loss": 2.0051,
+      "step": 142
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 1.978259801864624,
+      "learning_rate": 9.8855424584857e-05,
+      "loss": 2.0367,
+      "step": 144
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 1.9169765710830688,
+      "learning_rate": 9.879843479803691e-05,
+      "loss": 2.1009,
+      "step": 146
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 1.8380109071731567,
+      "learning_rate": 9.874007771322151e-05,
+      "loss": 2.1456,
+      "step": 148
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1143693923950195,
+      "learning_rate": 9.868035496553546e-05,
+      "loss": 1.925,
+      "step": 150
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 1.8774141073226929,
+      "learning_rate": 9.86192682283684e-05,
+      "loss": 1.9616,
+      "step": 152
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 2.3532581329345703,
+      "learning_rate": 9.855681921332793e-05,
+      "loss": 2.0289,
+      "step": 154
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 2.1421797275543213,
+      "learning_rate": 9.849300967019175e-05,
+      "loss": 2.0153,
+      "step": 156
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 2.0029852390289307,
+      "learning_rate": 9.84278413868586e-05,
+      "loss": 2.0726,
+      "step": 158
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 2.0344998836517334,
+      "learning_rate": 9.836131618929819e-05,
+      "loss": 2.0215,
+      "step": 160
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 1.8781356811523438,
+      "learning_rate": 9.82934359415e-05,
+      "loss": 2.0622,
+      "step": 162
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 1.9795514345169067,
+      "learning_rate": 9.822420254542108e-05,
+      "loss": 2.0249,
+      "step": 164
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 2.012881278991699,
+      "learning_rate": 9.815361794093272e-05,
+      "loss": 1.9815,
+      "step": 166
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 2.264941453933716,
+      "learning_rate": 9.808168410576617e-05,
+      "loss": 2.0232,
+      "step": 168
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 2.4006729125976562,
+      "learning_rate": 9.800840305545715e-05,
+      "loss": 2.0844,
+      "step": 170
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 2.0443308353424072,
+      "learning_rate": 9.793377684328939e-05,
+      "loss": 2.2302,
+      "step": 172
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 2.164515972137451,
+      "learning_rate": 9.785780756023714e-05,
+      "loss": 1.9808,
+      "step": 174
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 1.9512875080108643,
+      "learning_rate": 9.778049733490655e-05,
+      "loss": 2.0968,
+      "step": 176
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 1.9964834451675415,
+      "learning_rate": 9.770184833347606e-05,
+      "loss": 1.9889,
+      "step": 178
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 1.9380826950073242,
+      "learning_rate": 9.762186275963563e-05,
+      "loss": 1.9766,
+      "step": 180
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 1.943260669708252,
+      "learning_rate": 9.754054285452506e-05,
+      "loss": 1.9298,
+      "step": 182
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 2.1821844577789307,
+      "learning_rate": 9.745789089667121e-05,
+      "loss": 2.1202,
+      "step": 184
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 1.7526299953460693,
+      "learning_rate": 9.737390920192408e-05,
+      "loss": 2.0635,
+      "step": 186
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 2.229520082473755,
+      "learning_rate": 9.7288600123392e-05,
+      "loss": 1.9582,
+      "step": 188
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 2.3614768981933594,
+      "learning_rate": 9.720196605137565e-05,
+      "loss": 2.0278,
+      "step": 190
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 2.1270534992218018,
+      "learning_rate": 9.71140094133011e-05,
+      "loss": 2.1036,
+      "step": 192
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 2.2983131408691406,
+      "learning_rate": 9.702473267365182e-05,
+      "loss": 2.0558,
+      "step": 194
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 1.9561504125595093,
+      "learning_rate": 9.693413833389956e-05,
+      "loss": 1.9173,
+      "step": 196
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 2.234160900115967,
+      "learning_rate": 9.684222893243431e-05,
+      "loss": 2.1188,
+      "step": 198
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.883965015411377,
+      "learning_rate": 9.674900704449324e-05,
+      "loss": 1.9584,
+      "step": 200
+    },
+    {
+      "epoch": 0.3232,
+      "grad_norm": 1.7237235307693481,
+      "learning_rate": 9.665447528208836e-05,
+      "loss": 1.9351,
+      "step": 202
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 2.0437498092651367,
+      "learning_rate": 9.655863629393351e-05,
+      "loss": 1.9079,
+      "step": 204
+    },
+    {
+      "epoch": 0.3296,
+      "grad_norm": 2.014540195465088,
+      "learning_rate": 9.64614927653701e-05,
+      "loss": 1.8612,
+      "step": 206
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 2.379439115524292,
+      "learning_rate": 9.636304741829181e-05,
+      "loss": 1.9976,
+      "step": 208
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 1.962538242340088,
+      "learning_rate": 9.626330301106837e-05,
+      "loss": 1.932,
+      "step": 210
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 1.862244725227356,
+      "learning_rate": 9.616226233846828e-05,
+      "loss": 1.8992,
+      "step": 212
+    },
+    {
+      "epoch": 0.3424,
+      "grad_norm": 1.7304776906967163,
+      "learning_rate": 9.605992823158046e-05,
+      "loss": 2.0777,
+      "step": 214
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 2.2403054237365723,
+      "learning_rate": 9.595630355773501e-05,
+      "loss": 1.8658,
+      "step": 216
+    },
+    {
+      "epoch": 0.3488,
+      "grad_norm": 3.3899903297424316,
+      "learning_rate": 9.585139122042274e-05,
+      "loss": 1.9963,
+      "step": 218
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 2.261810064315796,
+      "learning_rate": 9.574519415921396e-05,
+      "loss": 1.947,
+      "step": 220
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 2.2053134441375732,
+      "learning_rate": 9.5637715349676e-05,
+      "loss": 2.0544,
+      "step": 222
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 1.871773362159729,
+      "learning_rate": 9.552895780328987e-05,
+      "loss": 1.8976,
+      "step": 224
+    },
+    {
+      "epoch": 0.3616,
+      "grad_norm": 1.6700202226638794,
+      "learning_rate": 9.541892456736595e-05,
+      "loss": 2.1166,
+      "step": 226
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 1.9986639022827148,
+      "learning_rate": 9.530761872495849e-05,
+      "loss": 1.9311,
+      "step": 228
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 2.288973331451416,
+      "learning_rate": 9.519504339477932e-05,
+      "loss": 1.98,
+      "step": 230
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 2.177896738052368,
+      "learning_rate": 9.508120173111039e-05,
+      "loss": 1.862,
+      "step": 232
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 1.9860484600067139,
+      "learning_rate": 9.496609692371548e-05,
+      "loss": 1.9192,
+      "step": 234
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 1.924127221107483,
+      "learning_rate": 9.484973219775074e-05,
+      "loss": 1.871,
+      "step": 236
+    },
+    {
+      "epoch": 0.3808,
+      "grad_norm": 1.9022867679595947,
+      "learning_rate": 9.473211081367436e-05,
+      "loss": 1.9067,
+      "step": 238
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 1.7447446584701538,
+      "learning_rate": 9.46132360671552e-05,
+      "loss": 1.8984,
+      "step": 240
+    },
+    {
+      "epoch": 0.3872,
+      "grad_norm": 2.809067487716675,
+      "learning_rate": 9.449311128898049e-05,
+      "loss": 1.8327,
+      "step": 242
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 1.9946494102478027,
+      "learning_rate": 9.437173984496246e-05,
+      "loss": 1.9735,
+      "step": 244
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 1.8834348917007446,
+      "learning_rate": 9.424912513584401e-05,
+      "loss": 2.0294,
+      "step": 246
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 1.9426389932632446,
+      "learning_rate": 9.412527059720352e-05,
+      "loss": 1.9919,
+      "step": 248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.823935627937317,
+      "learning_rate": 9.400017969935848e-05,
+      "loss": 1.8907,
+      "step": 250
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 2.1048786640167236,
+      "learning_rate": 9.387385594726829e-05,
+      "loss": 1.8855,
+      "step": 252
+    },
+    {
+      "epoch": 0.4064,
+      "grad_norm": 1.9253580570220947,
+      "learning_rate": 9.374630288043614e-05,
+      "loss": 2.0577,
+      "step": 254
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 1.785396695137024,
+      "learning_rate": 9.361752407280965e-05,
+      "loss": 1.9675,
+      "step": 256
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 1.9203846454620361,
+      "learning_rate": 9.348752313268093e-05,
+      "loss": 1.8934,
+      "step": 258
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 1.986392855644226,
+      "learning_rate": 9.335630370258533e-05,
+      "loss": 1.9838,
+      "step": 260
+    },
+    {
+      "epoch": 0.4192,
+      "grad_norm": 1.953905463218689,
+      "learning_rate": 9.322386945919946e-05,
+      "loss": 1.7604,
+      "step": 262
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 1.7314627170562744,
+      "learning_rate": 9.309022411323816e-05,
+      "loss": 2.0328,
+      "step": 264
+    },
+    {
+      "epoch": 0.4256,
+      "grad_norm": 1.6745048761367798,
+      "learning_rate": 9.295537140935049e-05,
+      "loss": 1.9734,
+      "step": 266
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 1.8622961044311523,
+      "learning_rate": 9.281931512601485e-05,
+      "loss": 1.9509,
+      "step": 268
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 2.014514684677124,
+      "learning_rate": 9.26820590754331e-05,
+      "loss": 1.8272,
+      "step": 270
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 2.118647336959839,
+      "learning_rate": 9.254360710342371e-05,
+      "loss": 1.8347,
+      "step": 272
+    },
+    {
+      "epoch": 0.4384,
+      "grad_norm": 2.04239821434021,
+      "learning_rate": 9.240396308931407e-05,
+      "loss": 1.8675,
+      "step": 274
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 1.951341152191162,
+      "learning_rate": 9.226313094583173e-05,
+      "loss": 1.9559,
+      "step": 276
+    },
+    {
+      "epoch": 0.4448,
+      "grad_norm": 1.7053275108337402,
+      "learning_rate": 9.212111461899479e-05,
+      "loss": 2.0715,
+      "step": 278
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 1.7789607048034668,
+      "learning_rate": 9.197791808800135e-05,
+      "loss": 1.89,
+      "step": 280
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 1.8625364303588867,
+      "learning_rate": 9.183354536511803e-05,
+      "loss": 1.9809,
+      "step": 282
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 1.6965309381484985,
+      "learning_rate": 9.168800049556747e-05,
+      "loss": 1.8365,
+      "step": 284
+    },
+    {
+      "epoch": 0.4576,
+      "grad_norm": 2.1207497119903564,
+      "learning_rate": 9.154128755741509e-05,
+      "loss": 1.8314,
+      "step": 286
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 1.8182010650634766,
+      "learning_rate": 9.139341066145472e-05,
+      "loss": 1.8906,
+      "step": 288
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 1.977777361869812,
+      "learning_rate": 9.124437395109353e-05,
+      "loss": 1.8562,
+      "step": 290
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 1.9953404664993286,
+      "learning_rate": 9.109418160223585e-05,
+      "loss": 1.8364,
+      "step": 292
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 1.9941433668136597,
+      "learning_rate": 9.094283782316619e-05,
+      "loss": 1.7585,
+      "step": 294
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 1.9799609184265137,
+      "learning_rate": 9.079034685443133e-05,
+      "loss": 1.8669,
+      "step": 296
+    },
+    {
+      "epoch": 0.4768,
+      "grad_norm": 1.755238652229309,
+      "learning_rate": 9.063671296872149e-05,
+      "loss": 1.8001,
+      "step": 298
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.059305429458618,
+      "learning_rate": 9.048194047075069e-05,
+      "loss": 1.9259,
+      "step": 300
+    },
+    {
+      "epoch": 0.4832,
+      "grad_norm": 1.7116378545761108,
+      "learning_rate": 9.032603369713596e-05,
+      "loss": 1.6954,
+      "step": 302
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 2.472815990447998,
+      "learning_rate": 9.016899701627604e-05,
+      "loss": 1.8413,
+      "step": 304
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 1.8934400081634521,
+      "learning_rate": 9.00108348282288e-05,
+      "loss": 1.9545,
+      "step": 306
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 2.147753953933716,
+      "learning_rate": 8.985155156458811e-05,
+      "loss": 1.7679,
+      "step": 308
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 2.2302675247192383,
+      "learning_rate": 8.969115168835954e-05,
+      "loss": 1.8257,
+      "step": 310
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 1.6578640937805176,
+      "learning_rate": 8.952963969383538e-05,
+      "loss": 1.7151,
+      "step": 312
+    },
+    {
+      "epoch": 0.5024,
+      "grad_norm": 1.754835844039917,
+      "learning_rate": 8.93670201064687e-05,
+      "loss": 2.0074,
+      "step": 314
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 2.130150556564331,
+      "learning_rate": 8.920329748274649e-05,
+      "loss": 1.8657,
+      "step": 316
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 1.7068381309509277,
+      "learning_rate": 8.903847641006218e-05,
+      "loss": 1.8955,
+      "step": 318
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 2.0879528522491455,
+      "learning_rate": 8.887256150658684e-05,
+      "loss": 1.7092,
+      "step": 320
+    },
+    {
+      "epoch": 0.5152,
+      "grad_norm": 1.8985047340393066,
+      "learning_rate": 8.870555742113998e-05,
+      "loss": 1.8091,
+      "step": 322
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 1.7577992677688599,
+      "learning_rate": 8.85374688330592e-05,
+      "loss": 1.8895,
+      "step": 324
+    },
+    {
+      "epoch": 0.5216,
+      "grad_norm": 1.8277013301849365,
+      "learning_rate": 8.836830045206911e-05,
+      "loss": 1.8192,
+      "step": 326
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 1.8492199182510376,
+      "learning_rate": 8.81980570181494e-05,
+      "loss": 2.0282,
+      "step": 328
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 1.8850246667861938,
+      "learning_rate": 8.802674330140192e-05,
+      "loss": 1.7955,
+      "step": 330
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 1.7965402603149414,
+      "learning_rate": 8.785436410191714e-05,
+      "loss": 1.8271,
+      "step": 332
+    },
+    {
+      "epoch": 0.5344,
+      "grad_norm": 2.0495541095733643,
+      "learning_rate": 8.76809242496396e-05,
+      "loss": 1.9308,
+      "step": 334
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 1.8388515710830688,
+      "learning_rate": 8.750642860423262e-05,
+      "loss": 1.8831,
+      "step": 336
+    },
+    {
+      "epoch": 0.5408,
+      "grad_norm": 2.2101669311523438,
+      "learning_rate": 8.733088205494205e-05,
+      "loss": 1.9837,
+      "step": 338
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 1.7564021348953247,
+      "learning_rate": 8.715428952045936e-05,
+      "loss": 2.0114,
+      "step": 340
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 2.0515785217285156,
+      "learning_rate": 8.697665594878382e-05,
+      "loss": 1.7574,
+      "step": 342
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 2.1503772735595703,
+      "learning_rate": 8.679798631708375e-05,
+      "loss": 1.9549,
+      "step": 344
+    },
+    {
+      "epoch": 0.5536,
+      "grad_norm": 1.6707327365875244,
+      "learning_rate": 8.661828563155727e-05,
+      "loss": 1.9318,
+      "step": 346
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 1.9014642238616943,
+      "learning_rate": 8.643755892729179e-05,
+      "loss": 1.9853,
+      "step": 348
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9820547103881836,
+      "learning_rate": 8.625581126812312e-05,
+      "loss": 1.8178,
+      "step": 350
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 2.810029983520508,
+      "learning_rate": 8.607304774649349e-05,
+      "loss": 2.0081,
+      "step": 352
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 1.8511972427368164,
+      "learning_rate": 8.588927348330887e-05,
+      "loss": 1.7794,
+      "step": 354
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 1.954455852508545,
+      "learning_rate": 8.57044936277955e-05,
+      "loss": 1.9215,
+      "step": 356
+    },
+    {
+      "epoch": 0.5728,
+      "grad_norm": 1.8836822509765625,
+      "learning_rate": 8.551871335735565e-05,
+      "loss": 1.7449,
+      "step": 358
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 1.8966975212097168,
+      "learning_rate": 8.533193787742251e-05,
+      "loss": 1.7689,
+      "step": 360
+    },
+    {
+      "epoch": 0.5792,
+      "grad_norm": 1.7771093845367432,
+      "learning_rate": 8.51441724213143e-05,
+      "loss": 1.8151,
+      "step": 362
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 1.880419135093689,
+      "learning_rate": 8.495542225008771e-05,
+      "loss": 1.805,
+      "step": 364
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 1.820349097251892,
+      "learning_rate": 8.476569265239046e-05,
+      "loss": 1.758,
+      "step": 366
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 1.984392523765564,
+      "learning_rate": 8.457498894431311e-05,
+      "loss": 1.7321,
+      "step": 368
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 1.710229516029358,
+      "learning_rate": 8.438331646924013e-05,
+      "loss": 1.7819,
+      "step": 370
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 1.736141324043274,
+      "learning_rate": 8.419068059770011e-05,
+      "loss": 1.8351,
+      "step": 372
+    },
+    {
+      "epoch": 0.5984,
+      "grad_norm": 1.6661279201507568,
+      "learning_rate": 8.399708672721539e-05,
+      "loss": 1.803,
+      "step": 374
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 4.828789710998535,
+      "learning_rate": 8.380254028215076e-05,
+      "loss": 1.8539,
+      "step": 376
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 2.078886032104492,
+      "learning_rate": 8.360704671356145e-05,
+      "loss": 1.7976,
+      "step": 378
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 1.720009684562683,
+      "learning_rate": 8.341061149904045e-05,
+      "loss": 1.9524,
+      "step": 380
+    },
+    {
+      "epoch": 0.6112,
+      "grad_norm": 1.935594081878662,
+      "learning_rate": 8.321324014256504e-05,
+      "loss": 1.8671,
+      "step": 382
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 1.868320345878601,
+      "learning_rate": 8.30149381743425e-05,
+      "loss": 1.8896,
+      "step": 384
+    },
+    {
+      "epoch": 0.6176,
+      "grad_norm": 2.0732314586639404,
+      "learning_rate": 8.28157111506552e-05,
+      "loss": 1.8446,
+      "step": 386
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 1.5798280239105225,
+      "learning_rate": 8.261556465370493e-05,
+      "loss": 1.9207,
+      "step": 388
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 1.6934467554092407,
+      "learning_rate": 8.24145042914565e-05,
+      "loss": 1.7548,
+      "step": 390
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 1.7732023000717163,
+      "learning_rate": 8.221253569748055e-05,
+      "loss": 1.7041,
+      "step": 392
+    },
+    {
+      "epoch": 0.6304,
+      "grad_norm": 1.9565222263336182,
+      "learning_rate": 8.200966453079575e-05,
+      "loss": 1.8865,
+      "step": 394
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 1.7031235694885254,
+      "learning_rate": 8.180589647571023e-05,
+      "loss": 2.0219,
+      "step": 396
+    },
+    {
+      "epoch": 0.6368,
+      "grad_norm": 1.8705931901931763,
+      "learning_rate": 8.16012372416623e-05,
+      "loss": 1.7774,
+      "step": 398
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7355400323867798,
+      "learning_rate": 8.13956925630605e-05,
+      "loss": 1.7273,
+      "step": 400
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 1.7146542072296143,
+      "learning_rate": 8.118926819912287e-05,
+      "loss": 1.8275,
+      "step": 402
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 1.8502819538116455,
+      "learning_rate": 8.098196993371565e-05,
+      "loss": 1.856,
+      "step": 404
+    },
+    {
+      "epoch": 0.6496,
+      "grad_norm": 1.6460517644882202,
+      "learning_rate": 8.077380357519115e-05,
+      "loss": 1.7826,
+      "step": 406
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 1.6977733373641968,
+      "learning_rate": 8.056477495622511e-05,
+      "loss": 2.0396,
+      "step": 408
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 2.395606756210327,
+      "learning_rate": 8.035488993365312e-05,
+      "loss": 1.755,
+      "step": 410
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 1.6800931692123413,
+      "learning_rate": 8.014415438830667e-05,
+      "loss": 1.9174,
+      "step": 412
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 1.940741777420044,
+      "learning_rate": 7.993257422484826e-05,
+      "loss": 1.7259,
+      "step": 414
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 1.6088985204696655,
+      "learning_rate": 7.972015537160602e-05,
+      "loss": 1.9236,
+      "step": 416
+    },
+    {
+      "epoch": 0.6688,
+      "grad_norm": 1.77496337890625,
+      "learning_rate": 7.950690378040758e-05,
+      "loss": 1.9956,
+      "step": 418
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 2.08013653755188,
+      "learning_rate": 7.929282542641325e-05,
+      "loss": 1.71,
+      "step": 420
+    },
+    {
+      "epoch": 0.6752,
+      "grad_norm": 1.9645555019378662,
+      "learning_rate": 7.907792630794876e-05,
+      "loss": 1.6586,
+      "step": 422
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 2.035111904144287,
+      "learning_rate": 7.886221244633703e-05,
+      "loss": 1.8481,
+      "step": 424
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 1.617519736289978,
+      "learning_rate": 7.864568988572947e-05,
+      "loss": 1.8787,
+      "step": 426
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 1.9266173839569092,
+      "learning_rate": 7.842836469293673e-05,
+      "loss": 1.7332,
+      "step": 428
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 1.6716456413269043,
+      "learning_rate": 7.821024295725865e-05,
+      "loss": 1.8147,
+      "step": 430
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 1.9675475358963013,
+      "learning_rate": 7.79913307903136e-05,
+      "loss": 1.77,
+      "step": 432
+    },
+    {
+      "epoch": 0.6944,
+      "grad_norm": 2.048152208328247,
+      "learning_rate": 7.777163432586734e-05,
+      "loss": 1.7438,
+      "step": 434
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 1.7210822105407715,
+      "learning_rate": 7.755115971966104e-05,
+      "loss": 1.7988,
+      "step": 436
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 2.126711845397949,
+      "learning_rate": 7.732991314923891e-05,
+      "loss": 1.7376,
+      "step": 438
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 1.7960891723632812,
+      "learning_rate": 7.710790081377502e-05,
+      "loss": 1.7875,
+      "step": 440
+    },
+    {
+      "epoch": 0.7072,
+      "grad_norm": 1.6610071659088135,
+      "learning_rate": 7.688512893389964e-05,
+      "loss": 1.7334,
+      "step": 442
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 1.6998896598815918,
+      "learning_rate": 7.666160375152496e-05,
+      "loss": 1.886,
+      "step": 444
+    },
+    {
+      "epoch": 0.7136,
+      "grad_norm": 1.6629440784454346,
+      "learning_rate": 7.643733152967019e-05,
+      "loss": 1.786,
+      "step": 446
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 1.6910452842712402,
+      "learning_rate": 7.621231855228604e-05,
+      "loss": 2.0343,
+      "step": 448
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9952099323272705,
+      "learning_rate": 7.598657112407865e-05,
+      "loss": 1.7571,
+      "step": 450
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 1.7345885038375854,
+      "learning_rate": 7.576009557033304e-05,
+      "loss": 2.0908,
+      "step": 452
+    },
+    {
+      "epoch": 0.7264,
+      "grad_norm": 1.6344877481460571,
+      "learning_rate": 7.553289823673568e-05,
+      "loss": 1.8395,
+      "step": 454
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 2.138115406036377,
+      "learning_rate": 7.530498548919693e-05,
+      "loss": 1.7072,
+      "step": 456
+    },
+    {
+      "epoch": 0.7328,
+      "grad_norm": 1.9216474294662476,
+      "learning_rate": 7.507636371367246e-05,
+      "loss": 1.6516,
+      "step": 458
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 1.4932810068130493,
+      "learning_rate": 7.484703931598445e-05,
+      "loss": 1.9351,
+      "step": 460
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 1.8183472156524658,
+      "learning_rate": 7.461701872164204e-05,
+      "loss": 1.8441,
+      "step": 462
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 1.5970336198806763,
+      "learning_rate": 7.438630837566133e-05,
+      "loss": 1.8145,
+      "step": 464
+    },
+    {
+      "epoch": 0.7456,
+      "grad_norm": 1.7351387739181519,
+      "learning_rate": 7.415491474238475e-05,
+      "loss": 1.8858,
+      "step": 466
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 1.6989448070526123,
+      "learning_rate": 7.39228443053e-05,
+      "loss": 1.8566,
+      "step": 468
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 1.8217098712921143,
+      "learning_rate": 7.369010356685833e-05,
+      "loss": 1.692,
+      "step": 470
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 1.7833845615386963,
+      "learning_rate": 7.345669904829237e-05,
+      "loss": 1.8145,
+      "step": 472
+    },
+    {
+      "epoch": 0.7584,
+      "grad_norm": 1.7113256454467773,
+      "learning_rate": 7.32226372894334e-05,
+      "loss": 1.907,
+      "step": 474
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 1.66838800907135,
+      "learning_rate": 7.298792484852808e-05,
+      "loss": 1.8243,
+      "step": 476
+    },
+    {
+      "epoch": 0.7648,
+      "grad_norm": 1.8057668209075928,
+      "learning_rate": 7.27525683020548e-05,
+      "loss": 1.6788,
+      "step": 478
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 1.7563303709030151,
+      "learning_rate": 7.251657424453928e-05,
+      "loss": 2.0148,
+      "step": 480
+    },
+    {
+      "epoch": 0.7712,
+      "grad_norm": 1.75275719165802,
+      "learning_rate": 7.227994928836988e-05,
+      "loss": 1.7584,
+      "step": 482
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 1.6364191770553589,
+      "learning_rate": 7.204270006361228e-05,
+      "loss": 1.9348,
+      "step": 484
+    },
+    {
+      "epoch": 0.7776,
+      "grad_norm": 1.7930974960327148,
+      "learning_rate": 7.180483321782374e-05,
+      "loss": 1.9014,
+      "step": 486
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 1.8914506435394287,
+      "learning_rate": 7.156635541586682e-05,
+      "loss": 1.7977,
+      "step": 488
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 1.7024521827697754,
+      "learning_rate": 7.132727333972265e-05,
+      "loss": 1.6993,
+      "step": 490
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 1.7870112657546997,
+      "learning_rate": 7.108759368830371e-05,
+      "loss": 1.6965,
+      "step": 492
+    },
+    {
+      "epoch": 0.7904,
+      "grad_norm": 1.763691782951355,
+      "learning_rate": 7.084732317726611e-05,
+      "loss": 1.7948,
+      "step": 494
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 1.683468222618103,
+      "learning_rate": 7.060646853882145e-05,
+      "loss": 1.9145,
+      "step": 496
+    },
+    {
+      "epoch": 0.7968,
+      "grad_norm": 1.9888768196105957,
+      "learning_rate": 7.036503652154812e-05,
+      "loss": 1.8192,
+      "step": 498
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5705928802490234,
+      "learning_rate": 7.012303389020234e-05,
+      "loss": 1.7831,
+      "step": 500
+    },
+    {
+      "epoch": 0.8032,
+      "grad_norm": 1.860660433769226,
+      "learning_rate": 6.988046742552845e-05,
+      "loss": 1.7904,
+      "step": 502
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 1.8895405530929565,
+      "learning_rate": 6.963734392406907e-05,
+      "loss": 1.8645,
+      "step": 504
+    },
+    {
+      "epoch": 0.8096,
+      "grad_norm": 1.74190354347229,
+      "learning_rate": 6.93936701979746e-05,
+      "loss": 1.8455,
+      "step": 506
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 1.9230369329452515,
+      "learning_rate": 6.914945307481228e-05,
+      "loss": 1.8388,
+      "step": 508
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 1.5093566179275513,
+      "learning_rate": 6.890469939737506e-05,
+      "loss": 1.752,
+      "step": 510
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 1.5916728973388672,
+      "learning_rate": 6.865941602348966e-05,
+      "loss": 1.7105,
+      "step": 512
+    },
+    {
+      "epoch": 0.8224,
+      "grad_norm": 1.7378982305526733,
+      "learning_rate": 6.841360982582463e-05,
+      "loss": 1.9789,
+      "step": 514
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 1.7520698308944702,
+      "learning_rate": 6.816728769169757e-05,
+      "loss": 1.7566,
+      "step": 516
+    },
+    {
+      "epoch": 0.8288,
+      "grad_norm": 1.8129826784133911,
+      "learning_rate": 6.792045652288234e-05,
+      "loss": 1.8551,
+      "step": 518
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 1.9102818965911865,
+      "learning_rate": 6.767312323541555e-05,
+      "loss": 1.7726,
+      "step": 520
+    },
+    {
+      "epoch": 0.8352,
+      "grad_norm": 1.5088154077529907,
+      "learning_rate": 6.742529475940284e-05,
+      "loss": 1.6381,
+      "step": 522
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 1.7010055780410767,
+      "learning_rate": 6.717697803882467e-05,
+      "loss": 1.8741,
+      "step": 524
+    },
+    {
+      "epoch": 0.8416,
+      "grad_norm": 1.6840184926986694,
+      "learning_rate": 6.692818003134184e-05,
+      "loss": 1.8617,
+      "step": 526
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 1.7205629348754883,
+      "learning_rate": 6.667890770810035e-05,
+      "loss": 1.7349,
+      "step": 528
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 1.520727515220642,
+      "learning_rate": 6.64291680535363e-05,
+      "loss": 1.749,
+      "step": 530
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 1.5941743850708008,
+      "learning_rate": 6.617896806518005e-05,
+      "loss": 1.7076,
+      "step": 532
+    },
+    {
+      "epoch": 0.8544,
+      "grad_norm": 1.7745941877365112,
+      "learning_rate": 6.592831475346018e-05,
+      "loss": 1.792,
+      "step": 534
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 1.5072052478790283,
+      "learning_rate": 6.56772151415071e-05,
+      "loss": 1.6149,
+      "step": 536
+    },
+    {
+      "epoch": 0.8608,
+      "grad_norm": 1.6202104091644287,
+      "learning_rate": 6.542567626495619e-05,
+      "loss": 1.756,
+      "step": 538
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 1.4974113702774048,
+      "learning_rate": 6.517370517175081e-05,
+      "loss": 1.7919,
+      "step": 540
+    },
+    {
+      "epoch": 0.8672,
+      "grad_norm": 1.653824806213379,
+      "learning_rate": 6.492130892194461e-05,
+      "loss": 2.0103,
+      "step": 542
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 1.683524489402771,
+      "learning_rate": 6.466849458750394e-05,
+      "loss": 2.0337,
+      "step": 544
+    },
+    {
+      "epoch": 0.8736,
+      "grad_norm": 1.5982547998428345,
+      "learning_rate": 6.441526925210949e-05,
+      "loss": 1.8919,
+      "step": 546
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 1.838497519493103,
+      "learning_rate": 6.416164001095799e-05,
+      "loss": 1.7648,
+      "step": 548
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.524348258972168,
+      "learning_rate": 6.390761397056328e-05,
+      "loss": 1.6804,
+      "step": 550
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 1.6498512029647827,
+      "learning_rate": 6.365319824855727e-05,
+      "loss": 1.6334,
+      "step": 552
+    },
+    {
+      "epoch": 0.8864,
+      "grad_norm": 1.5689668655395508,
+      "learning_rate": 6.339839997349045e-05,
+      "loss": 1.9048,
+      "step": 554
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 1.7050296068191528,
+      "learning_rate": 6.314322628463219e-05,
+      "loss": 1.6864,
+      "step": 556
+    },
+    {
+      "epoch": 0.8928,
+      "grad_norm": 2.038351535797119,
+      "learning_rate": 6.288768433177068e-05,
+      "loss": 1.7531,
+      "step": 558
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 1.7489795684814453,
+      "learning_rate": 6.26317812750126e-05,
+      "loss": 1.8467,
+      "step": 560
+    },
+    {
+      "epoch": 0.8992,
+      "grad_norm": 1.6068861484527588,
+      "learning_rate": 6.237552428458256e-05,
+      "loss": 1.8459,
+      "step": 562
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 1.616613745689392,
+      "learning_rate": 6.21189205406221e-05,
+      "loss": 1.8173,
+      "step": 564
+    },
+    {
+      "epoch": 0.9056,
+      "grad_norm": 1.6885602474212646,
+      "learning_rate": 6.186197723298855e-05,
+      "loss": 1.8358,
+      "step": 566
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 1.688711404800415,
+      "learning_rate": 6.160470156105362e-05,
+      "loss": 1.6996,
+      "step": 568
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 1.74298894405365,
+      "learning_rate": 6.134710073350156e-05,
+      "loss": 1.722,
+      "step": 570
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 1.6249070167541504,
+      "learning_rate": 6.108918196812734e-05,
+      "loss": 1.7909,
+      "step": 572
+    },
+    {
+      "epoch": 0.9184,
+      "grad_norm": 1.659416675567627,
+      "learning_rate": 6.083095249163424e-05,
+      "loss": 1.6625,
+      "step": 574
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 1.6332143545150757,
+      "learning_rate": 6.057241953943154e-05,
+      "loss": 1.8297,
+      "step": 576
+    },
+    {
+      "epoch": 0.9248,
+      "grad_norm": 1.6717133522033691,
+      "learning_rate": 6.031359035543158e-05,
+      "loss": 2.0601,
+      "step": 578
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 1.8736896514892578,
+      "learning_rate": 6.005447219184702e-05,
+      "loss": 1.8117,
+      "step": 580
+    },
+    {
+      "epoch": 0.9312,
+      "grad_norm": 1.6602182388305664,
+      "learning_rate": 5.9795072308987485e-05,
+      "loss": 1.7275,
+      "step": 582
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 1.6776071786880493,
+      "learning_rate": 5.9535397975056154e-05,
+      "loss": 1.8988,
+      "step": 584
+    },
+    {
+      "epoch": 0.9376,
+      "grad_norm": 1.588109016418457,
+      "learning_rate": 5.927545646594617e-05,
+      "loss": 1.7716,
+      "step": 586
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 1.6331814527511597,
+      "learning_rate": 5.901525506503668e-05,
+      "loss": 1.8081,
+      "step": 588
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 1.7309777736663818,
+      "learning_rate": 5.87548010629889e-05,
+      "loss": 1.8243,
+      "step": 590
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 1.6374008655548096,
+      "learning_rate": 5.8494101757541676e-05,
+      "loss": 1.7065,
+      "step": 592
+    },
+    {
+      "epoch": 0.9504,
+      "grad_norm": 1.6291025876998901,
+      "learning_rate": 5.8233164453307156e-05,
+      "loss": 1.8013,
+      "step": 594
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 1.7339948415756226,
+      "learning_rate": 5.797199646156596e-05,
+      "loss": 1.7998,
+      "step": 596
+    },
+    {
+      "epoch": 0.9568,
+      "grad_norm": 1.581697940826416,
+      "learning_rate": 5.7710605100062485e-05,
+      "loss": 1.645,
+      "step": 598
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.713205337524414,
+      "learning_rate": 5.7448997692799764e-05,
+      "loss": 1.9092,
+      "step": 600
+    },
+    {
+      "epoch": 0.9632,
+      "grad_norm": 1.7023775577545166,
+      "learning_rate": 5.718718156983428e-05,
+      "loss": 1.7403,
+      "step": 602
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 1.606632947921753,
+      "learning_rate": 5.69251640670706e-05,
+      "loss": 1.679,
+      "step": 604
+    },
+    {
+      "epoch": 0.9696,
+      "grad_norm": 1.5328476428985596,
+      "learning_rate": 5.6662952526055793e-05,
+      "loss": 1.7899,
+      "step": 606
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 1.5965962409973145,
+      "learning_rate": 5.6400554293773744e-05,
+      "loss": 1.7776,
+      "step": 608
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 1.5174623727798462,
+      "learning_rate": 5.61379767224393e-05,
+      "loss": 1.6602,
+      "step": 610
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 1.6876877546310425,
+      "learning_rate": 5.587522716929228e-05,
+      "loss": 1.6656,
+      "step": 612
+    },
+    {
+      "epoch": 0.9824,
+      "grad_norm": 1.5483810901641846,
+      "learning_rate": 5.561231299639127e-05,
+      "loss": 1.6531,
+      "step": 614
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 1.464625597000122,
+      "learning_rate": 5.534924157040745e-05,
+      "loss": 1.8967,
+      "step": 616
+    },
+    {
+      "epoch": 0.9888,
+      "grad_norm": 1.7587417364120483,
+      "learning_rate": 5.508602026241807e-05,
+      "loss": 1.6637,
+      "step": 618
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 1.5783720016479492,
+      "learning_rate": 5.482265644769998e-05,
+      "loss": 1.7628,
+      "step": 620
+    },
+    {
+      "epoch": 0.9952,
+      "grad_norm": 1.602127194404602,
+      "learning_rate": 5.4559157505522985e-05,
+      "loss": 1.7458,
+      "step": 622
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 1.8909751176834106,
+      "learning_rate": 5.429553081894304e-05,
+      "loss": 1.6952,
+      "step": 624
+    },
+    {
+      "epoch": 1.0016,
+      "grad_norm": 1.5928688049316406,
+      "learning_rate": 5.4031783774595455e-05,
+      "loss": 1.5991,
+      "step": 626
+    },
+    {
+      "epoch": 1.0048,
+      "grad_norm": 1.487565517425537,
+      "learning_rate": 5.3767923762487824e-05,
+      "loss": 1.4893,
+      "step": 628
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 1.4564417600631714,
+      "learning_rate": 5.3503958175793055e-05,
+      "loss": 1.4741,
+      "step": 630
+    },
+    {
+      "epoch": 1.0112,
+      "grad_norm": 1.4703936576843262,
+      "learning_rate": 5.323989441064216e-05,
+      "loss": 1.5821,
+      "step": 632
+    },
+    {
+      "epoch": 1.0144,
+      "grad_norm": 1.746761441230774,
+      "learning_rate": 5.2975739865917074e-05,
+      "loss": 1.565,
+      "step": 634
+    },
+    {
+      "epoch": 1.0176,
+      "grad_norm": 1.7585957050323486,
+      "learning_rate": 5.271150194304326e-05,
+      "loss": 1.5653,
+      "step": 636
+    },
+    {
+      "epoch": 1.0208,
+      "grad_norm": 1.5630764961242676,
+      "learning_rate": 5.244718804578246e-05,
+      "loss": 1.496,
+      "step": 638
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 1.6863924264907837,
+      "learning_rate": 5.218280558002506e-05,
+      "loss": 1.4683,
+      "step": 640
+    },
+    {
+      "epoch": 1.0272,
+      "grad_norm": 1.4588528871536255,
+      "learning_rate": 5.191836195358278e-05,
+      "loss": 1.513,
+      "step": 642
+    },
+    {
+      "epoch": 1.0304,
+      "grad_norm": 1.4486134052276611,
+      "learning_rate": 5.165386457598099e-05,
+      "loss": 1.484,
+      "step": 644
+    },
+    {
+      "epoch": 1.0336,
+      "grad_norm": 1.5040372610092163,
+      "learning_rate": 5.13893208582511e-05,
+      "loss": 1.4945,
+      "step": 646
+    },
+    {
+      "epoch": 1.0368,
+      "grad_norm": 1.556773066520691,
+      "learning_rate": 5.1124738212722966e-05,
+      "loss": 1.395,
+      "step": 648
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5296416282653809,
+      "learning_rate": 5.086012405281717e-05,
+      "loss": 1.3816,
+      "step": 650
+    },
+    {
+      "epoch": 1.0432,
+      "grad_norm": 1.5270073413848877,
+      "learning_rate": 5.0595485792837305e-05,
+      "loss": 1.5622,
+      "step": 652
+    },
+    {
+      "epoch": 1.0464,
+      "grad_norm": 1.4216368198394775,
+      "learning_rate": 5.033083084776222e-05,
+      "loss": 1.4719,
+      "step": 654
+    },
+    {
+      "epoch": 1.0496,
+      "grad_norm": 1.6059627532958984,
+      "learning_rate": 5.0066166633038305e-05,
+      "loss": 1.5124,
+      "step": 656
+    },
+    {
+      "epoch": 1.0528,
+      "grad_norm": 1.7300755977630615,
+      "learning_rate": 4.980150056437163e-05,
+      "loss": 1.4899,
+      "step": 658
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 1.4066635370254517,
+      "learning_rate": 4.9536840057520224e-05,
+      "loss": 1.4709,
+      "step": 660
+    },
+    {
+      "epoch": 1.0592,
+      "grad_norm": 1.563008189201355,
+      "learning_rate": 4.927219252808631e-05,
+      "loss": 1.4385,
+      "step": 662
+    },
+    {
+      "epoch": 1.0624,
+      "grad_norm": 1.6098390817642212,
+      "learning_rate": 4.900756539130846e-05,
+      "loss": 1.4528,
+      "step": 664
+    },
+    {
+      "epoch": 1.0656,
+      "grad_norm": 1.5592678785324097,
+      "learning_rate": 4.874296606185387e-05,
+      "loss": 1.5697,
+      "step": 666
+    },
+    {
+      "epoch": 1.0688,
+      "grad_norm": 1.7256783246994019,
+      "learning_rate": 4.847840195361058e-05,
+      "loss": 1.5878,
+      "step": 668
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 3.1409246921539307,
+      "learning_rate": 4.821388047947979e-05,
+      "loss": 1.4804,
+      "step": 670
+    },
+    {
+      "epoch": 1.0752,
+      "grad_norm": 1.415878415107727,
+      "learning_rate": 4.7949409051168085e-05,
+      "loss": 1.5137,
+      "step": 672
+    },
+    {
+      "epoch": 1.0784,
+      "grad_norm": 1.6788074970245361,
+      "learning_rate": 4.768499507897981e-05,
+      "loss": 1.4662,
+      "step": 674
+    },
+    {
+      "epoch": 1.0816,
+      "grad_norm": 1.3528187274932861,
+      "learning_rate": 4.742064597160948e-05,
+      "loss": 1.4879,
+      "step": 676
+    },
+    {
+      "epoch": 1.0848,
+      "grad_norm": 1.5209320783615112,
+      "learning_rate": 4.715636913593404e-05,
+      "loss": 1.5457,
+      "step": 678
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 1.7500444650650024,
+      "learning_rate": 4.689217197680554e-05,
+      "loss": 1.4667,
+      "step": 680
+    },
+    {
+      "epoch": 1.0912,
+      "grad_norm": 1.6211199760437012,
+      "learning_rate": 4.6628061896843474e-05,
+      "loss": 1.5392,
+      "step": 682
+    },
+    {
+      "epoch": 1.0944,
+      "grad_norm": 1.584383249282837,
+      "learning_rate": 4.6364046296227484e-05,
+      "loss": 1.4909,
+      "step": 684
+    },
+    {
+      "epoch": 1.0976,
+      "grad_norm": 1.7815152406692505,
+      "learning_rate": 4.6100132572489915e-05,
+      "loss": 1.654,
+      "step": 686
+    },
+    {
+      "epoch": 1.1008,
+      "grad_norm": 1.3984824419021606,
+      "learning_rate": 4.5836328120308674e-05,
+      "loss": 1.5179,
+      "step": 688
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 2.673772096633911,
+      "learning_rate": 4.5572640331299875e-05,
+      "loss": 1.5183,
+      "step": 690
+    },
+    {
+      "epoch": 1.1072,
+      "grad_norm": 1.3696187734603882,
+      "learning_rate": 4.530907659381086e-05,
+      "loss": 1.4193,
+      "step": 692
+    },
+    {
+      "epoch": 1.1104,
+      "grad_norm": 1.5257173776626587,
+      "learning_rate": 4.504564429271311e-05,
+      "loss": 1.4448,
+      "step": 694
+    },
+    {
+      "epoch": 1.1136,
+      "grad_norm": 1.4060217142105103,
+      "learning_rate": 4.478235080919536e-05,
+      "loss": 1.4528,
+      "step": 696
+    },
+    {
+      "epoch": 1.1168,
+      "grad_norm": 1.7510145902633667,
+      "learning_rate": 4.451920352055678e-05,
+      "loss": 1.4895,
+      "step": 698
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.639930009841919,
+      "learning_rate": 4.425620980000026e-05,
+      "loss": 1.5675,
+      "step": 700
+    },
+    {
+      "epoch": 1.1232,
+      "grad_norm": 1.6045302152633667,
+      "learning_rate": 4.39933770164258e-05,
+      "loss": 1.546,
+      "step": 702
+    },
+    {
+      "epoch": 1.1264,
+      "grad_norm": 1.6283916234970093,
+      "learning_rate": 4.373071253422408e-05,
+      "loss": 1.5913,
+      "step": 704
+    },
+    {
+      "epoch": 1.1296,
+      "grad_norm": 1.594617247581482,
+      "learning_rate": 4.346822371307009e-05,
+      "loss": 1.5827,
+      "step": 706
+    },
+    {
+      "epoch": 1.1328,
+      "grad_norm": 1.6242777109146118,
+      "learning_rate": 4.320591790771691e-05,
+      "loss": 1.493,
+      "step": 708
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 1.445251226425171,
+      "learning_rate": 4.294380246778966e-05,
+      "loss": 1.4685,
+      "step": 710
+    },
+    {
+      "epoch": 1.1392,
+      "grad_norm": 1.4812084436416626,
+      "learning_rate": 4.2681884737579524e-05,
+      "loss": 1.4637,
+      "step": 712
+    },
+    {
+      "epoch": 1.1424,
+      "grad_norm": 1.327941656112671,
+      "learning_rate": 4.242017205583805e-05,
+      "loss": 1.4974,
+      "step": 714
+    },
+    {
+      "epoch": 1.1456,
+      "grad_norm": 1.5381219387054443,
+      "learning_rate": 4.215867175557142e-05,
+      "loss": 1.4419,
+      "step": 716
+    },
+    {
+      "epoch": 1.1488,
+      "grad_norm": 1.447100281715393,
+      "learning_rate": 4.189739116383506e-05,
+      "loss": 1.5027,
+      "step": 718
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 1.7003172636032104,
+      "learning_rate": 4.163633760152834e-05,
+      "loss": 1.4394,
+      "step": 720
+    },
+    {
+      "epoch": 1.1552,
+      "grad_norm": 1.5542171001434326,
+      "learning_rate": 4.137551838318936e-05,
+      "loss": 1.502,
+      "step": 722
+    },
+    {
+      "epoch": 1.1584,
+      "grad_norm": 1.4329487085342407,
+      "learning_rate": 4.1114940816790135e-05,
+      "loss": 1.4303,
+      "step": 724
+    },
+    {
+      "epoch": 1.1616,
+      "grad_norm": 1.5292679071426392,
+      "learning_rate": 4.08546122035317e-05,
+      "loss": 1.4335,
+      "step": 726
+    },
+    {
+      "epoch": 1.1648,
+      "grad_norm": 1.5414072275161743,
+      "learning_rate": 4.059453983763967e-05,
+      "loss": 1.4574,
+      "step": 728
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 1.5525455474853516,
+      "learning_rate": 4.03347310061597e-05,
+      "loss": 1.3842,
+      "step": 730
+    },
+    {
+      "epoch": 1.1712,
+      "grad_norm": 1.444132685661316,
+      "learning_rate": 4.007519298875347e-05,
+      "loss": 1.4233,
+      "step": 732
+    },
+    {
+      "epoch": 1.1743999999999999,
+      "grad_norm": 1.6456663608551025,
+      "learning_rate": 3.98159330574946e-05,
+      "loss": 1.4416,
+      "step": 734
+    },
+    {
+      "epoch": 1.1776,
+      "grad_norm": 1.5266261100769043,
+      "learning_rate": 3.955695847666494e-05,
+      "loss": 1.4292,
+      "step": 736
+    },
+    {
+      "epoch": 1.1808,
+      "grad_norm": 1.6392107009887695,
+      "learning_rate": 3.929827650255104e-05,
+      "loss": 1.5965,
+      "step": 738
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 1.4020483493804932,
+      "learning_rate": 3.903989438324077e-05,
+      "loss": 1.3887,
+      "step": 740
+    },
+    {
+      "epoch": 1.1872,
+      "grad_norm": 1.5396273136138916,
+      "learning_rate": 3.878181935842033e-05,
+      "loss": 1.5077,
+      "step": 742
+    },
+    {
+      "epoch": 1.1904,
+      "grad_norm": 1.5113873481750488,
+      "learning_rate": 3.85240586591713e-05,
+      "loss": 1.4624,
+      "step": 744
+    },
+    {
+      "epoch": 1.1936,
+      "grad_norm": 1.4673871994018555,
+      "learning_rate": 3.8266619507768126e-05,
+      "loss": 1.4441,
+      "step": 746
+    },
+    {
+      "epoch": 1.1968,
+      "grad_norm": 1.387566328048706,
+      "learning_rate": 3.800950911747565e-05,
+      "loss": 1.4103,
+      "step": 748
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.4239872694015503,
+      "learning_rate": 3.775273469234712e-05,
+      "loss": 1.5861,
+      "step": 750
+    },
+    {
+      "epoch": 1.2032,
+      "grad_norm": 1.672402024269104,
+      "learning_rate": 3.749630342702221e-05,
+      "loss": 1.5554,
+      "step": 752
+    },
+    {
+      "epoch": 1.2064,
+      "grad_norm": 1.3757636547088623,
+      "learning_rate": 3.724022250652557e-05,
+      "loss": 1.4742,
+      "step": 754
+    },
+    {
+      "epoch": 1.2096,
+      "grad_norm": 1.5252386331558228,
+      "learning_rate": 3.698449910606536e-05,
+      "loss": 1.4546,
+      "step": 756
+    },
+    {
+      "epoch": 1.2128,
+      "grad_norm": 1.351670503616333,
+      "learning_rate": 3.672914039083233e-05,
+      "loss": 1.4842,
+      "step": 758
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 2.256891965866089,
+      "learning_rate": 3.6474153515799e-05,
+      "loss": 1.4753,
+      "step": 760
+    },
+    {
+      "epoch": 1.2192,
+      "grad_norm": 1.6337233781814575,
+      "learning_rate": 3.6219545625519145e-05,
+      "loss": 1.51,
+      "step": 762
+    },
+    {
+      "epoch": 1.2224,
+      "grad_norm": 1.6732077598571777,
+      "learning_rate": 3.596532385392772e-05,
+      "loss": 1.6077,
+      "step": 764
+    },
+    {
+      "epoch": 1.2256,
+      "grad_norm": 1.410620927810669,
+      "learning_rate": 3.5711495324140845e-05,
+      "loss": 1.4561,
+      "step": 766
+    },
+    {
+      "epoch": 1.2288000000000001,
+      "grad_norm": 1.5092262029647827,
+      "learning_rate": 3.54580671482563e-05,
+      "loss": 1.4808,
+      "step": 768
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 1.5939708948135376,
+      "learning_rate": 3.520504642715424e-05,
+      "loss": 1.568,
+      "step": 770
+    },
+    {
+      "epoch": 1.2352,
+      "grad_norm": 1.4955201148986816,
+      "learning_rate": 3.495244025029822e-05,
+      "loss": 1.4955,
+      "step": 772
+    },
+    {
+      "epoch": 1.2384,
+      "grad_norm": 1.6155612468719482,
+      "learning_rate": 3.470025569553653e-05,
+      "loss": 1.4821,
+      "step": 774
+    },
+    {
+      "epoch": 1.2416,
+      "grad_norm": 1.5372660160064697,
+      "learning_rate": 3.444849982890393e-05,
+      "loss": 1.4093,
+      "step": 776
+    },
+    {
+      "epoch": 1.2448,
+      "grad_norm": 1.7753242254257202,
+      "learning_rate": 3.4197179704423656e-05,
+      "loss": 1.4489,
+      "step": 778
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 1.5369523763656616,
+      "learning_rate": 3.39463023639097e-05,
+      "loss": 1.469,
+      "step": 780
+    },
+    {
+      "epoch": 1.2511999999999999,
+      "grad_norm": 1.5201616287231445,
+      "learning_rate": 3.36958748367696e-05,
+      "loss": 1.4533,
+      "step": 782
+    },
+    {
+      "epoch": 1.2544,
+      "grad_norm": 1.6345765590667725,
+      "learning_rate": 3.3445904139807405e-05,
+      "loss": 1.5058,
+      "step": 784
+    },
+    {
+      "epoch": 1.2576,
+      "grad_norm": 1.4603197574615479,
+      "learning_rate": 3.319639727702716e-05,
+      "loss": 1.4791,
+      "step": 786
+    },
+    {
+      "epoch": 1.2608,
+      "grad_norm": 1.395509958267212,
+      "learning_rate": 3.2947361239436525e-05,
+      "loss": 1.3728,
+      "step": 788
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 1.625909447669983,
+      "learning_rate": 3.2698803004851026e-05,
+      "loss": 1.4991,
+      "step": 790
+    },
+    {
+      "epoch": 1.2671999999999999,
+      "grad_norm": 1.4244285821914673,
+      "learning_rate": 3.245072953769844e-05,
+      "loss": 1.5196,
+      "step": 792
+    },
+    {
+      "epoch": 1.2704,
+      "grad_norm": 1.4113078117370605,
+      "learning_rate": 3.2203147788823764e-05,
+      "loss": 1.4194,
+      "step": 794
+    },
+    {
+      "epoch": 1.2736,
+      "grad_norm": 1.588931918144226,
+      "learning_rate": 3.1956064695294305e-05,
+      "loss": 1.4783,
+      "step": 796
+    },
+    {
+      "epoch": 1.2768,
+      "grad_norm": 1.6504685878753662,
+      "learning_rate": 3.170948718020546e-05,
+      "loss": 1.4376,
+      "step": 798
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.408836007118225,
+      "learning_rate": 3.1463422152486674e-05,
+      "loss": 1.4438,
+      "step": 800
+    },
+    {
+      "epoch": 1.2832,
+      "grad_norm": 1.6479829549789429,
+      "learning_rate": 3.121787650670781e-05,
+      "loss": 1.5115,
+      "step": 802
+    },
+    {
+      "epoch": 1.2864,
+      "grad_norm": 1.4651159048080444,
+      "learning_rate": 3.097285712288605e-05,
+      "loss": 1.4111,
+      "step": 804
+    },
+    {
+      "epoch": 1.2896,
+      "grad_norm": 1.4425824880599976,
+      "learning_rate": 3.072837086629305e-05,
+      "loss": 1.4359,
+      "step": 806
+    },
+    {
+      "epoch": 1.2928,
+      "grad_norm": 1.4235820770263672,
+      "learning_rate": 3.0484424587262655e-05,
+      "loss": 1.4132,
+      "step": 808
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 1.471458911895752,
+      "learning_rate": 3.024102512099889e-05,
+      "loss": 1.4653,
+      "step": 810
+    },
+    {
+      "epoch": 1.2992,
+      "grad_norm": 1.6857844591140747,
+      "learning_rate": 2.9998179287384485e-05,
+      "loss": 1.5749,
+      "step": 812
+    },
+    {
+      "epoch": 1.3024,
+      "grad_norm": 1.4331309795379639,
+      "learning_rate": 2.975589389078975e-05,
+      "loss": 1.4682,
+      "step": 814
+    },
+    {
+      "epoch": 1.3056,
+      "grad_norm": 1.3838329315185547,
+      "learning_rate": 2.9514175719881966e-05,
+      "loss": 1.4477,
+      "step": 816
+    },
+    {
+      "epoch": 1.3088,
+      "grad_norm": 1.350062370300293,
+      "learning_rate": 2.9273031547435114e-05,
+      "loss": 1.4243,
+      "step": 818
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 1.5023717880249023,
+      "learning_rate": 2.9032468130140168e-05,
+      "loss": 1.4972,
+      "step": 820
+    },
+    {
+      "epoch": 1.3152,
+      "grad_norm": 1.5788859128952026,
+      "learning_rate": 2.8792492208415768e-05,
+      "loss": 1.5411,
+      "step": 822
+    },
+    {
+      "epoch": 1.3184,
+      "grad_norm": 1.4530489444732666,
+      "learning_rate": 2.8553110506219283e-05,
+      "loss": 1.4582,
+      "step": 824
+    },
+    {
+      "epoch": 1.3216,
+      "grad_norm": 1.3994284868240356,
+      "learning_rate": 2.831432973085848e-05,
+      "loss": 1.4381,
+      "step": 826
+    },
+    {
+      "epoch": 1.3248,
+      "grad_norm": 1.51462984085083,
+      "learning_rate": 2.8076156572803635e-05,
+      "loss": 1.4712,
+      "step": 828
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 1.505707859992981,
+      "learning_rate": 2.783859770549996e-05,
+      "loss": 1.4613,
+      "step": 830
+    },
+    {
+      "epoch": 1.3312,
+      "grad_norm": 1.4707581996917725,
+      "learning_rate": 2.760165978518067e-05,
+      "loss": 1.4202,
+      "step": 832
+    },
+    {
+      "epoch": 1.3344,
+      "grad_norm": 1.5572878122329712,
+      "learning_rate": 2.7365349450680466e-05,
+      "loss": 1.4835,
+      "step": 834
+    },
+    {
+      "epoch": 1.3376000000000001,
+      "grad_norm": 1.6469281911849976,
+      "learning_rate": 2.7129673323249604e-05,
+      "loss": 1.4834,
+      "step": 836
+    },
+    {
+      "epoch": 1.3408,
+      "grad_norm": 1.6217268705368042,
+      "learning_rate": 2.689463800636824e-05,
+      "loss": 1.4905,
+      "step": 838
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 1.3879499435424805,
+      "learning_rate": 2.6660250085561457e-05,
+      "loss": 1.4575,
+      "step": 840
+    },
+    {
+      "epoch": 1.3472,
+      "grad_norm": 1.436532735824585,
+      "learning_rate": 2.6426516128214807e-05,
+      "loss": 1.5241,
+      "step": 842
+    },
+    {
+      "epoch": 1.3504,
+      "grad_norm": 1.4625749588012695,
+      "learning_rate": 2.619344268339021e-05,
+      "loss": 1.4646,
+      "step": 844
+    },
+    {
+      "epoch": 1.3536000000000001,
+      "grad_norm": 1.2639025449752808,
+      "learning_rate": 2.5961036281642493e-05,
+      "loss": 1.4009,
+      "step": 846
+    },
+    {
+      "epoch": 1.3568,
+      "grad_norm": 1.3335165977478027,
+      "learning_rate": 2.572930343483637e-05,
+      "loss": 1.3794,
+      "step": 848
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 1.357528805732727,
+      "learning_rate": 2.54982506359641e-05,
+      "loss": 1.4393,
+      "step": 850
+    },
+    {
+      "epoch": 1.3632,
+      "grad_norm": 1.277764916419983,
+      "learning_rate": 2.526788435896339e-05,
+      "loss": 1.4483,
+      "step": 852
+    },
+    {
+      "epoch": 1.3664,
+      "grad_norm": 1.3506453037261963,
+      "learning_rate": 2.5038211058536133e-05,
+      "loss": 1.4466,
+      "step": 854
+    },
+    {
+      "epoch": 1.3696,
+      "grad_norm": 1.5710474252700806,
+      "learning_rate": 2.4809237169967458e-05,
+      "loss": 1.553,
+      "step": 856
+    },
+    {
+      "epoch": 1.3728,
+      "grad_norm": 1.3960719108581543,
+      "learning_rate": 2.4580969108945533e-05,
+      "loss": 1.4599,
+      "step": 858
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 1.4087730646133423,
+      "learning_rate": 2.435341327138168e-05,
+      "loss": 1.4237,
+      "step": 860
+    },
+    {
+      "epoch": 1.3792,
+      "grad_norm": 1.4425933361053467,
+      "learning_rate": 2.4126576033231208e-05,
+      "loss": 1.3405,
+      "step": 862
+    },
+    {
+      "epoch": 1.3824,
+      "grad_norm": 1.4608137607574463,
+      "learning_rate": 2.3900463750314834e-05,
+      "loss": 1.4638,
+      "step": 864
+    },
+    {
+      "epoch": 1.3856,
+      "grad_norm": 1.4107086658477783,
+      "learning_rate": 2.3675082758140475e-05,
+      "loss": 1.5278,
+      "step": 866
+    },
+    {
+      "epoch": 1.3888,
+      "grad_norm": 1.3733059167861938,
+      "learning_rate": 2.3450439371725825e-05,
+      "loss": 1.5054,
+      "step": 868
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 1.3773467540740967,
+      "learning_rate": 2.3226539885421343e-05,
+      "loss": 1.4151,
+      "step": 870
+    },
+    {
+      "epoch": 1.3952,
+      "grad_norm": 1.4656720161437988,
+      "learning_rate": 2.3003390572734006e-05,
+      "loss": 1.4811,
+      "step": 872
+    },
+    {
+      "epoch": 1.3984,
+      "grad_norm": 1.3903800249099731,
+      "learning_rate": 2.2780997686151378e-05,
+      "loss": 1.4274,
+      "step": 874
+    },
+    {
+      "epoch": 1.4016,
+      "grad_norm": 1.3573490381240845,
+      "learning_rate": 2.255936745696652e-05,
+      "loss": 1.4141,
+      "step": 876
+    },
+    {
+      "epoch": 1.4048,
+      "grad_norm": 1.312809944152832,
+      "learning_rate": 2.2338506095103334e-05,
+      "loss": 1.431,
+      "step": 878
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 1.3963521718978882,
+      "learning_rate": 2.2118419788942672e-05,
+      "loss": 1.4246,
+      "step": 880
+    },
+    {
+      "epoch": 1.4112,
+      "grad_norm": 1.5214847326278687,
+      "learning_rate": 2.189911470514881e-05,
+      "loss": 1.4489,
+      "step": 882
+    },
+    {
+      "epoch": 1.4144,
+      "grad_norm": 1.5130972862243652,
+      "learning_rate": 2.1680596988496705e-05,
+      "loss": 1.4718,
+      "step": 884
+    },
+    {
+      "epoch": 1.4176,
+      "grad_norm": 1.4655705690383911,
+      "learning_rate": 2.1462872761699905e-05,
+      "loss": 1.4782,
+      "step": 886
+    },
+    {
+      "epoch": 1.4208,
+      "grad_norm": 1.4640536308288574,
+      "learning_rate": 2.1245948125238867e-05,
+      "loss": 1.4003,
+      "step": 888
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 1.50718092918396,
+      "learning_rate": 2.1029829157190117e-05,
+      "loss": 1.43,
+      "step": 890
+    },
+    {
+      "epoch": 1.4272,
+      "grad_norm": 1.3067643642425537,
+      "learning_rate": 2.081452191305587e-05,
+      "loss": 1.4523,
+      "step": 892
+    },
+    {
+      "epoch": 1.4304000000000001,
+      "grad_norm": 1.4971873760223389,
+      "learning_rate": 2.06000324255945e-05,
+      "loss": 1.4485,
+      "step": 894
+    },
+    {
+      "epoch": 1.4336,
+      "grad_norm": 1.4533696174621582,
+      "learning_rate": 2.0386366704651315e-05,
+      "loss": 1.4637,
+      "step": 896
+    },
+    {
+      "epoch": 1.4368,
+      "grad_norm": 1.3518776893615723,
+      "learning_rate": 2.0173530736990304e-05,
+      "loss": 1.3545,
+      "step": 898
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.395410180091858,
+      "learning_rate": 1.9961530486126327e-05,
+      "loss": 1.383,
+      "step": 900
+    },
+    {
+      "epoch": 1.4432,
+      "grad_norm": 2.61722731590271,
+      "learning_rate": 1.9750371892158103e-05,
+      "loss": 1.6465,
+      "step": 902
+    },
+    {
+      "epoch": 1.4464000000000001,
+      "grad_norm": 1.403191089630127,
+      "learning_rate": 1.9540060871601646e-05,
+      "loss": 1.4416,
+      "step": 904
+    },
+    {
+      "epoch": 1.4496,
+      "grad_norm": 1.3970694541931152,
+      "learning_rate": 1.933060331722457e-05,
+      "loss": 1.5657,
+      "step": 906
+    },
+    {
+      "epoch": 1.4527999999999999,
+      "grad_norm": 1.3750842809677124,
+      "learning_rate": 1.9122005097881014e-05,
+      "loss": 1.5444,
+      "step": 908
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 1.5135127305984497,
+      "learning_rate": 1.8914272058347088e-05,
+      "loss": 1.5246,
+      "step": 910
+    },
+    {
+      "epoch": 1.4592,
+      "grad_norm": 1.4489692449569702,
+      "learning_rate": 1.8707410019157196e-05,
+      "loss": 1.4398,
+      "step": 912
+    },
+    {
+      "epoch": 1.4624,
+      "grad_norm": 1.508467435836792,
+      "learning_rate": 1.8501424776440907e-05,
+      "loss": 1.4185,
+      "step": 914
+    },
+    {
+      "epoch": 1.4656,
+      "grad_norm": 1.3195210695266724,
+      "learning_rate": 1.829632210176061e-05,
+      "loss": 1.4815,
+      "step": 916
+    },
+    {
+      "epoch": 1.4687999999999999,
+      "grad_norm": 1.3782057762145996,
+      "learning_rate": 1.809210774194971e-05,
+      "loss": 1.4895,
+      "step": 918
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 1.4828996658325195,
+      "learning_rate": 1.7888787418951645e-05,
+      "loss": 1.4907,
+      "step": 920
+    },
+    {
+      "epoch": 1.4752,
+      "grad_norm": 1.409317970275879,
+      "learning_rate": 1.7686366829659628e-05,
+      "loss": 1.4705,
+      "step": 922
+    },
+    {
+      "epoch": 1.4784,
+      "grad_norm": 1.3606739044189453,
+      "learning_rate": 1.74848516457569e-05,
+      "loss": 1.4784,
+      "step": 924
+    },
+    {
+      "epoch": 1.4816,
+      "grad_norm": 1.4307705163955688,
+      "learning_rate": 1.72842475135579e-05,
+      "loss": 1.4639,
+      "step": 926
+    },
+    {
+      "epoch": 1.4848,
+      "grad_norm": 1.4825451374053955,
+      "learning_rate": 1.7084560053850024e-05,
+      "loss": 1.4771,
+      "step": 928
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 1.5933488607406616,
+      "learning_rate": 1.6885794861736183e-05,
+      "loss": 1.3967,
+      "step": 930
+    },
+    {
+      "epoch": 1.4912,
+      "grad_norm": 1.4622035026550293,
+      "learning_rate": 1.668795750647796e-05,
+      "loss": 1.3764,
+      "step": 932
+    },
+    {
+      "epoch": 1.4944,
+      "grad_norm": 1.387617588043213,
+      "learning_rate": 1.6491053531339607e-05,
+      "loss": 1.4615,
+      "step": 934
+    },
+    {
+      "epoch": 1.4976,
+      "grad_norm": 1.3928184509277344,
+      "learning_rate": 1.62950884534327e-05,
+      "loss": 1.3788,
+      "step": 936
+    },
+    {
+      "epoch": 1.5008,
+      "grad_norm": 1.4277182817459106,
+      "learning_rate": 1.6100067763561626e-05,
+      "loss": 1.3684,
+      "step": 938
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 1.519103765487671,
+      "learning_rate": 1.5905996926069628e-05,
+      "loss": 1.4659,
+      "step": 940
+    },
+    {
+      "epoch": 1.5072,
+      "grad_norm": 1.3705520629882812,
+      "learning_rate": 1.5712881378685755e-05,
+      "loss": 1.4145,
+      "step": 942
+    },
+    {
+      "epoch": 1.5104,
+      "grad_norm": 1.437618374824524,
+      "learning_rate": 1.5520726532372537e-05,
+      "loss": 1.3951,
+      "step": 944
+    },
+    {
+      "epoch": 1.5135999999999998,
+      "grad_norm": 1.406826376914978,
+      "learning_rate": 1.532953777117429e-05,
+      "loss": 1.509,
+      "step": 946
+    },
+    {
+      "epoch": 1.5168,
+      "grad_norm": 1.4625985622406006,
+      "learning_rate": 1.5139320452066313e-05,
+      "loss": 1.4415,
+      "step": 948
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.500941514968872,
+      "learning_rate": 1.4950079904804759e-05,
+      "loss": 1.4751,
+      "step": 950
+    },
+    {
+      "epoch": 1.5232,
+      "grad_norm": 1.4515104293823242,
+      "learning_rate": 1.4761821431777373e-05,
+      "loss": 1.441,
+      "step": 952
+    },
+    {
+      "epoch": 1.5264,
+      "grad_norm": 1.3307709693908691,
+      "learning_rate": 1.4574550307854817e-05,
+      "loss": 1.4566,
+      "step": 954
+    },
+    {
+      "epoch": 1.5295999999999998,
+      "grad_norm": 1.393490195274353,
+      "learning_rate": 1.4388271780242929e-05,
+      "loss": 1.4483,
+      "step": 956
+    },
+    {
+      "epoch": 1.5328,
+      "grad_norm": 1.3786524534225464,
+      "learning_rate": 1.4202991068335697e-05,
+      "loss": 1.4522,
+      "step": 958
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 1.4812992811203003,
+      "learning_rate": 1.4018713363569035e-05,
+      "loss": 1.5197,
+      "step": 960
+    },
+    {
+      "epoch": 1.5392000000000001,
+      "grad_norm": 1.5414302349090576,
+      "learning_rate": 1.3835443829275268e-05,
+      "loss": 1.4215,
+      "step": 962
+    },
+    {
+      "epoch": 1.5424,
+      "grad_norm": 1.420955777168274,
+      "learning_rate": 1.365318760053848e-05,
+      "loss": 1.4357,
+      "step": 964
+    },
+    {
+      "epoch": 1.5455999999999999,
+      "grad_norm": 1.322597861289978,
+      "learning_rate": 1.3471949784050702e-05,
+      "loss": 1.3364,
+      "step": 966
+    },
+    {
+      "epoch": 1.5488,
+      "grad_norm": 1.356268048286438,
+      "learning_rate": 1.3291735457968701e-05,
+      "loss": 1.4199,
+      "step": 968
+    },
+    {
+      "epoch": 1.552,
+      "grad_norm": 1.3691470623016357,
+      "learning_rate": 1.3112549671771796e-05,
+      "loss": 1.425,
+      "step": 970
+    },
+    {
+      "epoch": 1.5552000000000001,
+      "grad_norm": 1.4271376132965088,
+      "learning_rate": 1.2934397446120306e-05,
+      "loss": 1.4198,
+      "step": 972
+    },
+    {
+      "epoch": 1.5584,
+      "grad_norm": 1.37627375125885,
+      "learning_rate": 1.2757283772714957e-05,
+      "loss": 1.395,
+      "step": 974
+    },
+    {
+      "epoch": 1.5615999999999999,
+      "grad_norm": 1.3378015756607056,
+      "learning_rate": 1.2581213614156928e-05,
+      "loss": 1.3637,
+      "step": 976
+    },
+    {
+      "epoch": 1.5648,
+      "grad_norm": 1.5486069917678833,
+      "learning_rate": 1.2406191903808844e-05,
+      "loss": 1.4776,
+      "step": 978
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 1.4717572927474976,
+      "learning_rate": 1.2232223545656552e-05,
+      "loss": 1.4642,
+      "step": 980
+    },
+    {
+      "epoch": 1.5712000000000002,
+      "grad_norm": 1.5726338624954224,
+      "learning_rate": 1.205931341417173e-05,
+      "loss": 1.4629,
+      "step": 982
+    },
+    {
+      "epoch": 1.5744,
+      "grad_norm": 1.359563946723938,
+      "learning_rate": 1.1887466354175253e-05,
+      "loss": 1.4208,
+      "step": 984
+    },
+    {
+      "epoch": 1.5776,
+      "grad_norm": 1.390755295753479,
+      "learning_rate": 1.1716687180701474e-05,
+      "loss": 1.4227,
+      "step": 986
+    },
+    {
+      "epoch": 1.5808,
+      "grad_norm": 1.516890048980713,
+      "learning_rate": 1.1546980678863361e-05,
+      "loss": 1.4247,
+      "step": 988
+    },
+    {
+      "epoch": 1.584,
+      "grad_norm": 1.5359797477722168,
+      "learning_rate": 1.1378351603718312e-05,
+      "loss": 1.3898,
+      "step": 990
+    },
+    {
+      "epoch": 1.5872000000000002,
+      "grad_norm": 1.3802939653396606,
+      "learning_rate": 1.1210804680135022e-05,
+      "loss": 1.3354,
+      "step": 992
+    },
+    {
+      "epoch": 1.5904,
+      "grad_norm": 1.4020977020263672,
+      "learning_rate": 1.1044344602661034e-05,
+      "loss": 1.4493,
+      "step": 994
+    },
+    {
+      "epoch": 1.5936,
+      "grad_norm": 1.373717188835144,
+      "learning_rate": 1.0878976035391252e-05,
+      "loss": 1.3899,
+      "step": 996
+    },
+    {
+      "epoch": 1.5968,
+      "grad_norm": 1.3707786798477173,
+      "learning_rate": 1.0714703611837201e-05,
+      "loss": 1.4054,
+      "step": 998
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.388121485710144,
+      "learning_rate": 1.0551531934797243e-05,
+      "loss": 1.3393,
+      "step": 1000
+    },
+    {
+      "epoch": 1.6032,
+      "grad_norm": 1.5597281455993652,
+      "learning_rate": 1.0389465576227558e-05,
+      "loss": 1.4488,
+      "step": 1002
+    },
+    {
+      "epoch": 1.6064,
+      "grad_norm": 1.2360092401504517,
+      "learning_rate": 1.0228509077114146e-05,
+      "loss": 1.4802,
+      "step": 1004
+    },
+    {
+      "epoch": 1.6096,
+      "grad_norm": 1.3134233951568604,
+      "learning_rate": 1.0068666947345456e-05,
+      "loss": 1.3538,
+      "step": 1006
+    },
+    {
+      "epoch": 1.6128,
+      "grad_norm": 1.334416151046753,
+      "learning_rate": 9.909943665586102e-06,
+      "loss": 1.3917,
+      "step": 1008
+    },
+    {
+      "epoch": 1.616,
+      "grad_norm": 1.3590480089187622,
+      "learning_rate": 9.752343679151399e-06,
+      "loss": 1.4165,
+      "step": 1010
+    },
+    {
+      "epoch": 1.6192,
+      "grad_norm": 1.4533120393753052,
+      "learning_rate": 9.595871403882661e-06,
+      "loss": 1.5106,
+      "step": 1012
+    },
+    {
+      "epoch": 1.6223999999999998,
+      "grad_norm": 1.5530134439468384,
+      "learning_rate": 9.440531224023552e-06,
+      "loss": 1.4217,
+      "step": 1014
+    },
+    {
+      "epoch": 1.6256,
+      "grad_norm": 1.4636127948760986,
+      "learning_rate": 9.286327492097196e-06,
+      "loss": 1.3306,
+      "step": 1016
+    },
+    {
+      "epoch": 1.6288,
+      "grad_norm": 1.4135422706604004,
+      "learning_rate": 9.133264528784274e-06,
+      "loss": 1.425,
+      "step": 1018
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 1.4650404453277588,
+      "learning_rate": 8.981346622801905e-06,
+      "loss": 1.5058,
+      "step": 1020
+    },
+    {
+      "epoch": 1.6352,
+      "grad_norm": 1.4439706802368164,
+      "learning_rate": 8.830578030783493e-06,
+      "loss": 1.4755,
+      "step": 1022
+    },
+    {
+      "epoch": 1.6383999999999999,
+      "grad_norm": 1.7018922567367554,
+      "learning_rate": 8.680962977159502e-06,
+      "loss": 1.457,
+      "step": 1024
+    },
+    {
+      "epoch": 1.6416,
+      "grad_norm": 5.544206619262695,
+      "learning_rate": 8.53250565403903e-06,
+      "loss": 1.5578,
+      "step": 1026
+    },
+    {
+      "epoch": 1.6448,
+      "grad_norm": 1.4300087690353394,
+      "learning_rate": 8.385210221092382e-06,
+      "loss": 1.4479,
+      "step": 1028
+    },
+    {
+      "epoch": 1.6480000000000001,
+      "grad_norm": 1.5086474418640137,
+      "learning_rate": 8.239080805434513e-06,
+      "loss": 1.4618,
+      "step": 1030
+    },
+    {
+      "epoch": 1.6512,
+      "grad_norm": 1.4398646354675293,
+      "learning_rate": 8.094121501509399e-06,
+      "loss": 1.4416,
+      "step": 1032
+    },
+    {
+      "epoch": 1.6543999999999999,
+      "grad_norm": 1.3153340816497803,
+      "learning_rate": 7.950336370975304e-06,
+      "loss": 1.3792,
+      "step": 1034
+    },
+    {
+      "epoch": 1.6576,
+      "grad_norm": 1.351747751235962,
+      "learning_rate": 7.80772944259096e-06,
+      "loss": 1.4545,
+      "step": 1036
+    },
+    {
+      "epoch": 1.6608,
+      "grad_norm": 1.4315369129180908,
+      "learning_rate": 7.666304712102695e-06,
+      "loss": 1.3891,
+      "step": 1038
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 1.2921706438064575,
+      "learning_rate": 7.526066142132521e-06,
+      "loss": 1.4527,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6672,
+      "grad_norm": 1.3649907112121582,
+      "learning_rate": 7.3870176620670194e-06,
+      "loss": 1.3692,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6703999999999999,
+      "grad_norm": 1.3974645137786865,
+      "learning_rate": 7.249163167947287e-06,
+      "loss": 1.3702,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6736,
+      "grad_norm": 1.3394696712493896,
+      "learning_rate": 7.1125065223598076e-06,
+      "loss": 1.3851,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6768,
+      "grad_norm": 1.3612364530563354,
+      "learning_rate": 6.9770515543281455e-06,
+      "loss": 1.4355,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 1.442084789276123,
+      "learning_rate": 6.842802059205727e-06,
+      "loss": 1.4225,
+      "step": 1050
+    },
+    {
+      "epoch": 1.6832,
+      "grad_norm": 1.2821780443191528,
+      "learning_rate": 6.709761798569442e-06,
+      "loss": 1.4098,
+      "step": 1052
+    },
+    {
+      "epoch": 1.6864,
+      "grad_norm": 1.3669836521148682,
+      "learning_rate": 6.577934500114335e-06,
+      "loss": 1.4238,
+      "step": 1054
+    },
+    {
+      "epoch": 1.6896,
+      "grad_norm": 1.3460770845413208,
+      "learning_rate": 6.44732385754902e-06,
+      "loss": 1.4161,
+      "step": 1056
+    },
+    {
+      "epoch": 1.6928,
+      "grad_norm": 1.488351821899414,
+      "learning_rate": 6.3179335304923095e-06,
+      "loss": 1.4265,
+      "step": 1058
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 1.3682546615600586,
+      "learning_rate": 6.189767144370645e-06,
+      "loss": 1.408,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6992,
+      "grad_norm": 1.5109198093414307,
+      "learning_rate": 6.062828290316469e-06,
+      "loss": 1.4179,
+      "step": 1062
+    },
+    {
+      "epoch": 1.7024,
+      "grad_norm": 1.4661554098129272,
+      "learning_rate": 5.937120525067641e-06,
+      "loss": 1.3816,
+      "step": 1064
+    },
+    {
+      "epoch": 1.7056,
+      "grad_norm": 1.3422075510025024,
+      "learning_rate": 5.812647370867763e-06,
+      "loss": 1.4014,
+      "step": 1066
+    },
+    {
+      "epoch": 1.7088,
+      "grad_norm": 1.3991550207138062,
+      "learning_rate": 5.689412315367543e-06,
+      "loss": 1.4443,
+      "step": 1068
+    },
+    {
+      "epoch": 1.712,
+      "grad_norm": 1.3040037155151367,
+      "learning_rate": 5.567418811526981e-06,
+      "loss": 1.3615,
+      "step": 1070
+    },
+    {
+      "epoch": 1.7151999999999998,
+      "grad_norm": 1.3308135271072388,
+      "learning_rate": 5.4466702775186785e-06,
+      "loss": 1.3662,
+      "step": 1072
+    },
+    {
+      "epoch": 1.7184,
+      "grad_norm": 1.3251948356628418,
+      "learning_rate": 5.327170096632089e-06,
+      "loss": 1.4263,
+      "step": 1074
+    },
+    {
+      "epoch": 1.7216,
+      "grad_norm": 1.219191074371338,
+      "learning_rate": 5.208921617178641e-06,
+      "loss": 1.4676,
+      "step": 1076
+    },
+    {
+      "epoch": 1.7248,
+      "grad_norm": 1.364389181137085,
+      "learning_rate": 5.091928152397984e-06,
+      "loss": 1.4124,
+      "step": 1078
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 1.4420322179794312,
+      "learning_rate": 4.976192980365124e-06,
+      "loss": 1.4189,
+      "step": 1080
+    },
+    {
+      "epoch": 1.7311999999999999,
+      "grad_norm": 1.3646230697631836,
+      "learning_rate": 4.861719343898613e-06,
+      "loss": 1.4043,
+      "step": 1082
+    },
+    {
+      "epoch": 1.7344,
+      "grad_norm": 1.4223638772964478,
+      "learning_rate": 4.748510450469623e-06,
+      "loss": 1.4538,
+      "step": 1084
+    },
+    {
+      "epoch": 1.7376,
+      "grad_norm": 1.3192670345306396,
+      "learning_rate": 4.63656947211214e-06,
+      "loss": 1.4187,
+      "step": 1086
+    },
+    {
+      "epoch": 1.7408000000000001,
+      "grad_norm": 1.3329640626907349,
+      "learning_rate": 4.525899545334023e-06,
+      "loss": 1.3402,
+      "step": 1088
+    },
+    {
+      "epoch": 1.744,
+      "grad_norm": 1.5323829650878906,
+      "learning_rate": 4.416503771029201e-06,
+      "loss": 1.6182,
+      "step": 1090
+    },
+    {
+      "epoch": 1.7471999999999999,
+      "grad_norm": 1.3993905782699585,
+      "learning_rate": 4.308385214390709e-06,
+      "loss": 1.4359,
+      "step": 1092
+    },
+    {
+      "epoch": 1.7504,
+      "grad_norm": 1.3519290685653687,
+      "learning_rate": 4.2015469048248375e-06,
+      "loss": 1.4073,
+      "step": 1094
+    },
+    {
+      "epoch": 1.7536,
+      "grad_norm": 1.3193116188049316,
+      "learning_rate": 4.095991835866275e-06,
+      "loss": 1.3293,
+      "step": 1096
+    },
+    {
+      "epoch": 1.7568000000000001,
+      "grad_norm": 1.3401769399642944,
+      "learning_rate": 3.99172296509418e-06,
+      "loss": 1.4281,
+      "step": 1098
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.3036984205245972,
+      "learning_rate": 3.888743214049346e-06,
+      "loss": 1.4516,
+      "step": 1100
+    },
+    {
+      "epoch": 1.7631999999999999,
+      "grad_norm": 1.2876367568969727,
+      "learning_rate": 3.7870554681523287e-06,
+      "loss": 1.3857,
+      "step": 1102
+    },
+    {
+      "epoch": 1.7664,
+      "grad_norm": 1.286058783531189,
+      "learning_rate": 3.6866625766226293e-06,
+      "loss": 1.3421,
+      "step": 1104
+    },
+    {
+      "epoch": 1.7696,
+      "grad_norm": 1.2945473194122314,
+      "learning_rate": 3.587567352398796e-06,
+      "loss": 1.4383,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7728000000000002,
+      "grad_norm": 1.3773088455200195,
+      "learning_rate": 3.489772572059674e-06,
+      "loss": 1.4287,
+      "step": 1108
+    },
+    {
+      "epoch": 1.776,
+      "grad_norm": 1.2338217496871948,
+      "learning_rate": 3.393280975746588e-06,
+      "loss": 1.3083,
+      "step": 1110
+    },
+    {
+      "epoch": 1.7792,
+      "grad_norm": 2.751504421234131,
+      "learning_rate": 3.2980952670865317e-06,
+      "loss": 1.4268,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7824,
+      "grad_norm": 1.335714340209961,
+      "learning_rate": 3.2042181131164528e-06,
+      "loss": 1.4319,
+      "step": 1114
+    },
+    {
+      "epoch": 1.7856,
+      "grad_norm": 1.2379279136657715,
+      "learning_rate": 3.11165214420851e-06,
+      "loss": 1.3696,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7888,
+      "grad_norm": 1.3956067562103271,
+      "learning_rate": 3.020399953996389e-06,
+      "loss": 1.4996,
+      "step": 1118
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 1.3509949445724487,
+      "learning_rate": 2.9304640993025988e-06,
+      "loss": 1.4443,
+      "step": 1120
+    },
+    {
+      "epoch": 1.7952,
+      "grad_norm": 1.2706594467163086,
+      "learning_rate": 2.8418471000668523e-06,
+      "loss": 1.3773,
+      "step": 1122
+    },
+    {
+      "epoch": 1.7984,
+      "grad_norm": 1.4019887447357178,
+      "learning_rate": 2.7545514392754437e-06,
+      "loss": 1.4795,
+      "step": 1124
+    },
+    {
+      "epoch": 1.8016,
+      "grad_norm": 1.3648223876953125,
+      "learning_rate": 2.6685795628917266e-06,
+      "loss": 1.3566,
+      "step": 1126
+    },
+    {
+      "epoch": 1.8048,
+      "grad_norm": 1.5225048065185547,
+      "learning_rate": 2.5839338797875036e-06,
+      "loss": 1.4291,
+      "step": 1128
+    },
+    {
+      "epoch": 1.808,
+      "grad_norm": 1.4786689281463623,
+      "learning_rate": 2.500616761675578e-06,
+      "loss": 1.6036,
+      "step": 1130
+    },
+    {
+      "epoch": 1.8112,
+      "grad_norm": 1.439139723777771,
+      "learning_rate": 2.41863054304331e-06,
+      "loss": 1.445,
+      "step": 1132
+    },
+    {
+      "epoch": 1.8144,
+      "grad_norm": 1.322020411491394,
+      "learning_rate": 2.3379775210871648e-06,
+      "loss": 1.4561,
+      "step": 1134
+    },
+    {
+      "epoch": 1.8176,
+      "grad_norm": 1.3266148567199707,
+      "learning_rate": 2.2586599556483734e-06,
+      "loss": 1.4347,
+      "step": 1136
+    },
+    {
+      "epoch": 1.8208,
+      "grad_norm": 1.3589297533035278,
+      "learning_rate": 2.180680069149621e-06,
+      "loss": 1.3969,
+      "step": 1138
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 1.3703463077545166,
+      "learning_rate": 2.104040046532768e-06,
+      "loss": 1.4253,
+      "step": 1140
+    },
+    {
+      "epoch": 1.8272,
+      "grad_norm": 1.2915222644805908,
+      "learning_rate": 2.0287420351976063e-06,
+      "loss": 1.4272,
+      "step": 1142
+    },
+    {
+      "epoch": 1.8304,
+      "grad_norm": 1.2881819009780884,
+      "learning_rate": 1.954788144941727e-06,
+      "loss": 1.3726,
+      "step": 1144
+    },
+    {
+      "epoch": 1.8336000000000001,
+      "grad_norm": 1.2942179441452026,
+      "learning_rate": 1.8821804479013772e-06,
+      "loss": 1.3269,
+      "step": 1146
+    },
+    {
+      "epoch": 1.8368,
+      "grad_norm": 1.4517358541488647,
+      "learning_rate": 1.81092097849343e-06,
+      "loss": 1.4521,
+      "step": 1148
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 1.3010282516479492,
+      "learning_rate": 1.7410117333583498e-06,
+      "loss": 1.3935,
+      "step": 1150
+    },
+    {
+      "epoch": 1.8432,
+      "grad_norm": 1.5645641088485718,
+      "learning_rate": 1.6724546713042577e-06,
+      "loss": 1.3637,
+      "step": 1152
+    },
+    {
+      "epoch": 1.8464,
+      "grad_norm": 1.4219244718551636,
+      "learning_rate": 1.6052517132520651e-06,
+      "loss": 1.4567,
+      "step": 1154
+    },
+    {
+      "epoch": 1.8496000000000001,
+      "grad_norm": 1.4337780475616455,
+      "learning_rate": 1.5394047421816327e-06,
+      "loss": 1.429,
+      "step": 1156
+    },
+    {
+      "epoch": 1.8528,
+      "grad_norm": 1.561288595199585,
+      "learning_rate": 1.4749156030790024e-06,
+      "loss": 1.4846,
+      "step": 1158
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 1.3074960708618164,
+      "learning_rate": 1.4117861028847267e-06,
+      "loss": 1.3928,
+      "step": 1160
+    },
+    {
+      "epoch": 1.8592,
+      "grad_norm": 1.285032033920288,
+      "learning_rate": 1.3500180104432325e-06,
+      "loss": 1.365,
+      "step": 1162
+    },
+    {
+      "epoch": 1.8624,
+      "grad_norm": 1.2760525941848755,
+      "learning_rate": 1.2896130564532427e-06,
+      "loss": 1.4376,
+      "step": 1164
+    },
+    {
+      "epoch": 1.8656000000000001,
+      "grad_norm": 1.3905929327011108,
+      "learning_rate": 1.2305729334192994e-06,
+      "loss": 1.4799,
+      "step": 1166
+    },
+    {
+      "epoch": 1.8688,
+      "grad_norm": 1.388978362083435,
+      "learning_rate": 1.1728992956043238e-06,
+      "loss": 1.4397,
+      "step": 1168
+    },
+    {
+      "epoch": 1.8719999999999999,
+      "grad_norm": 3.00641131401062,
+      "learning_rate": 1.1165937589833087e-06,
+      "loss": 1.5248,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8752,
+      "grad_norm": 1.3412961959838867,
+      "learning_rate": 1.061657901197971e-06,
+      "loss": 1.4096,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8784,
+      "grad_norm": 1.2458962202072144,
+      "learning_rate": 1.008093261512616e-06,
+      "loss": 1.4463,
+      "step": 1174
+    },
+    {
+      "epoch": 1.8816000000000002,
+      "grad_norm": 1.331132411956787,
+      "learning_rate": 9.559013407709595e-07,
+      "loss": 1.3897,
+      "step": 1176
+    },
+    {
+      "epoch": 1.8848,
+      "grad_norm": 1.3351112604141235,
+      "learning_rate": 9.050836013541009e-07,
+      "loss": 1.4563,
+      "step": 1178
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 1.434446930885315,
+      "learning_rate": 8.55641467139534e-07,
+      "loss": 1.4178,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8912,
+      "grad_norm": 1.281200885772705,
+      "learning_rate": 8.075763234612622e-07,
+      "loss": 1.4287,
+      "step": 1182
+    },
+    {
+      "epoch": 1.8944,
+      "grad_norm": 1.7233554124832153,
+      "learning_rate": 7.60889517070984e-07,
+      "loss": 1.4914,
+      "step": 1184
+    },
+    {
+      "epoch": 1.8976,
+      "grad_norm": 1.3849607706069946,
+      "learning_rate": 7.155823561003361e-07,
+      "loss": 1.3976,
+      "step": 1186
+    },
+    {
+      "epoch": 1.9008,
+      "grad_norm": 1.3818809986114502,
+      "learning_rate": 6.716561100242658e-07,
+      "loss": 1.4095,
+      "step": 1188
+    },
+    {
+      "epoch": 1.904,
+      "grad_norm": 1.2812706232070923,
+      "learning_rate": 6.291120096254433e-07,
+      "loss": 1.3447,
+      "step": 1190
+    },
+    {
+      "epoch": 1.9072,
+      "grad_norm": 1.4153764247894287,
+      "learning_rate": 5.879512469598058e-07,
+      "loss": 1.3582,
+      "step": 1192
+    },
+    {
+      "epoch": 1.9104,
+      "grad_norm": 1.609390377998352,
+      "learning_rate": 5.481749753231124e-07,
+      "loss": 1.3982,
+      "step": 1194
+    },
+    {
+      "epoch": 1.9136,
+      "grad_norm": 1.3431967496871948,
+      "learning_rate": 5.097843092186583e-07,
+      "loss": 1.4785,
+      "step": 1196
+    },
+    {
+      "epoch": 1.9167999999999998,
+      "grad_norm": 1.350616216659546,
+      "learning_rate": 4.7278032432604425e-07,
+      "loss": 1.3739,
+      "step": 1198
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.213797688484192,
+      "learning_rate": 4.371640574710345e-07,
+      "loss": 1.4042,
+      "step": 1200
+    },
+    {
+      "epoch": 1.9232,
+      "grad_norm": 1.3762907981872559,
+      "learning_rate": 4.0293650659650184e-07,
+      "loss": 1.3671,
+      "step": 1202
+    },
+    {
+      "epoch": 1.9264000000000001,
+      "grad_norm": 1.3090656995773315,
+      "learning_rate": 3.7009863073446673e-07,
+      "loss": 1.426,
+      "step": 1204
+    },
+    {
+      "epoch": 1.9296,
+      "grad_norm": 1.3925652503967285,
+      "learning_rate": 3.386513499792354e-07,
+      "loss": 1.4414,
+      "step": 1206
+    },
+    {
+      "epoch": 1.9327999999999999,
+      "grad_norm": 1.2840951681137085,
+      "learning_rate": 3.0859554546160965e-07,
+      "loss": 1.297,
+      "step": 1208
+    },
+    {
+      "epoch": 1.936,
+      "grad_norm": 1.2935904264450073,
+      "learning_rate": 2.7993205932420053e-07,
+      "loss": 1.3802,
+      "step": 1210
+    },
+    {
+      "epoch": 1.9392,
+      "grad_norm": 1.2855424880981445,
+      "learning_rate": 2.5266169469783105e-07,
+      "loss": 1.5116,
+      "step": 1212
+    },
+    {
+      "epoch": 1.9424000000000001,
+      "grad_norm": 1.345445990562439,
+      "learning_rate": 2.2678521567903176e-07,
+      "loss": 1.4102,
+      "step": 1214
+    },
+    {
+      "epoch": 1.9456,
+      "grad_norm": 1.389737606048584,
+      "learning_rate": 2.023033473086411e-07,
+      "loss": 1.4212,
+      "step": 1216
+    },
+    {
+      "epoch": 1.9487999999999999,
+      "grad_norm": 1.4330286979675293,
+      "learning_rate": 1.7921677555147177e-07,
+      "loss": 1.4581,
+      "step": 1218
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 1.428533673286438,
+      "learning_rate": 1.5752614727712057e-07,
+      "loss": 1.3911,
+      "step": 1220
+    },
+    {
+      "epoch": 1.9552,
+      "grad_norm": 1.358577847480774,
+      "learning_rate": 1.3723207024180507e-07,
+      "loss": 1.4441,
+      "step": 1222
+    },
+    {
+      "epoch": 1.9584000000000001,
+      "grad_norm": 1.422744631767273,
+      "learning_rate": 1.1833511307136613e-07,
+      "loss": 1.4378,
+      "step": 1224
+    },
+    {
+      "epoch": 1.9616,
+      "grad_norm": 1.3814959526062012,
+      "learning_rate": 1.0083580524531955e-07,
+      "loss": 1.4551,
+      "step": 1226
+    },
+    {
+      "epoch": 1.9647999999999999,
+      "grad_norm": 1.2662824392318726,
+      "learning_rate": 8.473463708202345e-08,
+      "loss": 1.4284,
+      "step": 1228
+    },
+    {
+      "epoch": 1.968,
+      "grad_norm": 1.2959048748016357,
+      "learning_rate": 7.003205972494486e-08,
+      "loss": 1.3193,
+      "step": 1230
+    },
+    {
+      "epoch": 1.9712,
+      "grad_norm": 1.3156046867370605,
+      "learning_rate": 5.672848513000873e-08,
+      "loss": 1.4021,
+      "step": 1232
+    },
+    {
+      "epoch": 1.9744000000000002,
+      "grad_norm": 1.3085626363754272,
+      "learning_rate": 4.482428605407374e-08,
+      "loss": 1.3919,
+      "step": 1234
+    },
+    {
+      "epoch": 1.9776,
+      "grad_norm": 1.3836755752563477,
+      "learning_rate": 3.431979604445745e-08,
+      "loss": 1.4616,
+      "step": 1236
+    },
+    {
+      "epoch": 1.9808,
+      "grad_norm": 1.3475788831710815,
+      "learning_rate": 2.521530942962702e-08,
+      "loss": 1.4426,
+      "step": 1238
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 1.4258958101272583,
+      "learning_rate": 1.7511081310922495e-08,
+      "loss": 1.3884,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9872,
+      "grad_norm": 1.2867543697357178,
+      "learning_rate": 1.1207327555429192e-08,
+      "loss": 1.3472,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9904,
+      "grad_norm": 1.4145276546478271,
+      "learning_rate": 6.304224789910329e-09,
+      "loss": 1.4185,
+      "step": 1244
+    },
+    {
+      "epoch": 1.9936,
+      "grad_norm": 1.2391281127929688,
+      "learning_rate": 2.801910395877627e-09,
+      "loss": 1.3197,
+      "step": 1246
+    },
+    {
+      "epoch": 1.9968,
+      "grad_norm": 1.4041131734848022,
+      "learning_rate": 7.004825057277398e-10,
+      "loss": 1.4755,
+      "step": 1248
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.3858015537261963,
+      "learning_rate": 0.0,
+      "loss": 1.3312,
+      "step": 1250
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1250,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 625,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.1133643907229286e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}