diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,11207 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.99608457321848,
+  "eval_steps": 500,
+  "global_step": 1595,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0031323414252153485,
+      "grad_norm": 11.050841614106629,
+      "learning_rate": 1.2500000000000002e-07,
+      "loss": 2.1157,
+      "step": 1
+    },
+    {
+      "epoch": 0.006264682850430697,
+      "grad_norm": 10.770240996509447,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 2.1788,
+      "step": 2
+    },
+    {
+      "epoch": 0.009397024275646046,
+      "grad_norm": 10.853179484768518,
+      "learning_rate": 3.75e-07,
+      "loss": 2.109,
+      "step": 3
+    },
+    {
+      "epoch": 0.012529365700861394,
+      "grad_norm": 10.639536499246885,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 2.1164,
+      "step": 4
+    },
+    {
+      "epoch": 0.015661707126076743,
+      "grad_norm": 10.748209885589805,
+      "learning_rate": 6.25e-07,
+      "loss": 2.1148,
+      "step": 5
+    },
+    {
+      "epoch": 0.018794048551292093,
+      "grad_norm": 10.463406145556146,
+      "learning_rate": 7.5e-07,
+      "loss": 2.1268,
+      "step": 6
+    },
+    {
+      "epoch": 0.02192638997650744,
+      "grad_norm": 10.729571035762252,
+      "learning_rate": 8.75e-07,
+      "loss": 2.1396,
+      "step": 7
+    },
+    {
+      "epoch": 0.025058731401722788,
+      "grad_norm": 9.834267944614176,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.1216,
+      "step": 8
+    },
+    {
+      "epoch": 0.028191072826938137,
+      "grad_norm": 9.6285022315209,
+      "learning_rate": 1.125e-06,
+      "loss": 2.0811,
+      "step": 9
+    },
+    {
+      "epoch": 0.031323414252153486,
+      "grad_norm": 8.068778598644403,
+      "learning_rate": 1.25e-06,
+      "loss": 2.0954,
+      "step": 10
+    },
+    {
+      "epoch": 0.03445575567736883,
+      "grad_norm": 7.900246512788569,
+      "learning_rate": 1.3750000000000002e-06,
+      "loss": 2.093,
+      "step": 11
+    },
+    {
+      "epoch": 0.037588097102584185,
+      "grad_norm": 7.672686278535105,
+      "learning_rate": 1.5e-06,
+      "loss": 2.035,
+      "step": 12
+    },
+    {
+      "epoch": 0.04072043852779953,
+      "grad_norm": 4.005341384803414,
+      "learning_rate": 1.6250000000000001e-06,
+      "loss": 1.9538,
+      "step": 13
+    },
+    {
+      "epoch": 0.04385277995301488,
+      "grad_norm": 4.033643898372103,
+      "learning_rate": 1.75e-06,
+      "loss": 1.9371,
+      "step": 14
+    },
+    {
+      "epoch": 0.04698512137823023,
+      "grad_norm": 3.9058441402584543,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 1.9222,
+      "step": 15
+    },
+    {
+      "epoch": 0.050117462803445575,
+      "grad_norm": 3.93510615716937,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.9425,
+      "step": 16
+    },
+    {
+      "epoch": 0.05324980422866092,
+      "grad_norm": 3.9972887378242508,
+      "learning_rate": 2.125e-06,
+      "loss": 1.871,
+      "step": 17
+    },
+    {
+      "epoch": 0.056382145653876274,
+      "grad_norm": 4.637696367766985,
+      "learning_rate": 2.25e-06,
+      "loss": 1.704,
+      "step": 18
+    },
+    {
+      "epoch": 0.05951448707909162,
+      "grad_norm": 4.895075769171093,
+      "learning_rate": 2.375e-06,
+      "loss": 1.7081,
+      "step": 19
+    },
+    {
+      "epoch": 0.06264682850430697,
+      "grad_norm": 4.887182700750148,
+      "learning_rate": 2.5e-06,
+      "loss": 1.6922,
+      "step": 20
+    },
+    {
+      "epoch": 0.06577916992952232,
+      "grad_norm": 4.387829116707637,
+      "learning_rate": 2.6250000000000003e-06,
+      "loss": 1.6223,
+      "step": 21
+    },
+    {
+      "epoch": 0.06891151135473766,
+      "grad_norm": 3.9332482151095056,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 1.6429,
+      "step": 22
+    },
+    {
+      "epoch": 0.07204385277995301,
+      "grad_norm": 3.621520125738942,
+      "learning_rate": 2.875e-06,
+      "loss": 1.5692,
+      "step": 23
+    },
+    {
+      "epoch": 0.07517619420516837,
+      "grad_norm": 4.869727787056477,
+      "learning_rate": 3e-06,
+      "loss": 1.4277,
+      "step": 24
+    },
+    {
+      "epoch": 0.07830853563038372,
+      "grad_norm": 2.760787151921623,
+      "learning_rate": 3.125e-06,
+      "loss": 1.343,
+      "step": 25
+    },
+    {
+      "epoch": 0.08144087705559906,
+      "grad_norm": 2.158543950171073,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 1.3304,
+      "step": 26
+    },
+    {
+      "epoch": 0.08457321848081441,
+      "grad_norm": 1.8629759121807041,
+      "learning_rate": 3.3750000000000003e-06,
+      "loss": 1.3279,
+      "step": 27
+    },
+    {
+      "epoch": 0.08770555990602975,
+      "grad_norm": 1.4660115041698683,
+      "learning_rate": 3.5e-06,
+      "loss": 1.3604,
+      "step": 28
+    },
+    {
+      "epoch": 0.0908379013312451,
+      "grad_norm": 1.2902107389850677,
+      "learning_rate": 3.625e-06,
+      "loss": 1.2715,
+      "step": 29
+    },
+    {
+      "epoch": 0.09397024275646046,
+      "grad_norm": 1.201092304355642,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.3227,
+      "step": 30
+    },
+    {
+      "epoch": 0.0971025841816758,
+      "grad_norm": 1.1572726408932534,
+      "learning_rate": 3.875e-06,
+      "loss": 1.3139,
+      "step": 31
+    },
+    {
+      "epoch": 0.10023492560689115,
+      "grad_norm": 1.0860599652926863,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.277,
+      "step": 32
+    },
+    {
+      "epoch": 0.1033672670321065,
+      "grad_norm": 1.1601900235559528,
+      "learning_rate": 4.125e-06,
+      "loss": 1.3422,
+      "step": 33
+    },
+    {
+      "epoch": 0.10649960845732184,
+      "grad_norm": 1.0702922916707374,
+      "learning_rate": 4.25e-06,
+      "loss": 1.2635,
+      "step": 34
+    },
+    {
+      "epoch": 0.1096319498825372,
+      "grad_norm": 1.1224186551004272,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 1.3686,
+      "step": 35
+    },
+    {
+      "epoch": 0.11276429130775255,
+      "grad_norm": 1.0316309615389836,
+      "learning_rate": 4.5e-06,
+      "loss": 1.2546,
+      "step": 36
+    },
+    {
+      "epoch": 0.1158966327329679,
+      "grad_norm": 1.0350470961145155,
+      "learning_rate": 4.625000000000001e-06,
+      "loss": 1.2617,
+      "step": 37
+    },
+    {
+      "epoch": 0.11902897415818324,
+      "grad_norm": 0.9499273751228338,
+      "learning_rate": 4.75e-06,
+      "loss": 1.2218,
+      "step": 38
+    },
+    {
+      "epoch": 0.12216131558339859,
+      "grad_norm": 0.9633898650879267,
+      "learning_rate": 4.875e-06,
+      "loss": 1.2513,
+      "step": 39
+    },
+    {
+      "epoch": 0.12529365700861395,
+      "grad_norm": 0.9820874085749698,
+      "learning_rate": 5e-06,
+      "loss": 1.2438,
+      "step": 40
+    },
+    {
+      "epoch": 0.1284259984338293,
+      "grad_norm": 0.9949392511683506,
+      "learning_rate": 5.125e-06,
+      "loss": 1.2127,
+      "step": 41
+    },
+    {
+      "epoch": 0.13155833985904464,
+      "grad_norm": 0.963927255316921,
+      "learning_rate": 5.2500000000000006e-06,
+      "loss": 1.1986,
+      "step": 42
+    },
+    {
+      "epoch": 0.13469068128425998,
+      "grad_norm": 0.9265554217258913,
+      "learning_rate": 5.375e-06,
+      "loss": 1.2378,
+      "step": 43
+    },
+    {
+      "epoch": 0.13782302270947533,
+      "grad_norm": 0.895459875699844,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 1.2106,
+      "step": 44
+    },
+    {
+      "epoch": 0.14095536413469067,
+      "grad_norm": 0.8697327947413821,
+      "learning_rate": 5.625e-06,
+      "loss": 1.1968,
+      "step": 45
+    },
+    {
+      "epoch": 0.14408770555990602,
+      "grad_norm": 0.8663662696743084,
+      "learning_rate": 5.75e-06,
+      "loss": 1.1794,
+      "step": 46
+    },
+    {
+      "epoch": 0.14722004698512137,
+      "grad_norm": 0.925686203130037,
+      "learning_rate": 5.8750000000000005e-06,
+      "loss": 1.2847,
+      "step": 47
+    },
+    {
+      "epoch": 0.15035238841033674,
+      "grad_norm": 0.9075811490967844,
+      "learning_rate": 6e-06,
+      "loss": 1.2584,
+      "step": 48
+    },
+    {
+      "epoch": 0.1534847298355521,
+      "grad_norm": 0.8432956780425832,
+      "learning_rate": 6.125000000000001e-06,
+      "loss": 1.1543,
+      "step": 49
+    },
+    {
+      "epoch": 0.15661707126076743,
+      "grad_norm": 0.8713896231009577,
+      "learning_rate": 6.25e-06,
+      "loss": 1.2038,
+      "step": 50
+    },
+    {
+      "epoch": 0.15974941268598278,
+      "grad_norm": 0.7897618524028729,
+      "learning_rate": 6.375e-06,
+      "loss": 1.1681,
+      "step": 51
+    },
+    {
+      "epoch": 0.16288175411119812,
+      "grad_norm": 0.8144877727194363,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 1.1969,
+      "step": 52
+    },
+    {
+      "epoch": 0.16601409553641347,
+      "grad_norm": 0.8337995472721724,
+      "learning_rate": 6.625e-06,
+      "loss": 1.1631,
+      "step": 53
+    },
+    {
+      "epoch": 0.16914643696162882,
+      "grad_norm": 0.8286193073403529,
+      "learning_rate": 6.750000000000001e-06,
+      "loss": 1.231,
+      "step": 54
+    },
+    {
+      "epoch": 0.17227877838684416,
+      "grad_norm": 0.8272909936450883,
+      "learning_rate": 6.875e-06,
+      "loss": 1.1908,
+      "step": 55
+    },
+    {
+      "epoch": 0.1754111198120595,
+      "grad_norm": 0.8159834359002948,
+      "learning_rate": 7e-06,
+      "loss": 1.1909,
+      "step": 56
+    },
+    {
+      "epoch": 0.17854346123727485,
+      "grad_norm": 0.8327400175086643,
+      "learning_rate": 7.125e-06,
+      "loss": 1.2279,
+      "step": 57
+    },
+    {
+      "epoch": 0.1816758026624902,
+      "grad_norm": 0.8345916890457833,
+      "learning_rate": 7.25e-06,
+      "loss": 1.2322,
+      "step": 58
+    },
+    {
+      "epoch": 0.18480814408770557,
+      "grad_norm": 0.8009822359933073,
+      "learning_rate": 7.375000000000001e-06,
+      "loss": 1.1961,
+      "step": 59
+    },
+    {
+      "epoch": 0.18794048551292092,
+      "grad_norm": 0.8174246707753029,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.159,
+      "step": 60
+    },
+    {
+      "epoch": 0.19107282693813626,
+      "grad_norm": 0.8166829885787309,
+      "learning_rate": 7.625e-06,
+      "loss": 1.2641,
+      "step": 61
+    },
+    {
+      "epoch": 0.1942051683633516,
+      "grad_norm": 0.8047242090989581,
+      "learning_rate": 7.75e-06,
+      "loss": 1.1689,
+      "step": 62
+    },
+    {
+      "epoch": 0.19733750978856696,
+      "grad_norm": 0.7730776885197735,
+      "learning_rate": 7.875e-06,
+      "loss": 1.1269,
+      "step": 63
+    },
+    {
+      "epoch": 0.2004698512137823,
+      "grad_norm": 0.7779957947469543,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.1762,
+      "step": 64
+    },
+    {
+      "epoch": 0.20360219263899765,
+      "grad_norm": 0.8000417296639418,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 1.2025,
+      "step": 65
+    },
+    {
+      "epoch": 0.206734534064213,
+      "grad_norm": 0.8222150496696322,
+      "learning_rate": 8.25e-06,
+      "loss": 1.2,
+      "step": 66
+    },
+    {
+      "epoch": 0.20986687548942834,
+      "grad_norm": 0.7895026739737903,
+      "learning_rate": 8.375e-06,
+      "loss": 1.1623,
+      "step": 67
+    },
+    {
+      "epoch": 0.21299921691464369,
+      "grad_norm": 0.7784224076989291,
+      "learning_rate": 8.5e-06,
+      "loss": 1.1554,
+      "step": 68
+    },
+    {
+      "epoch": 0.21613155833985903,
+      "grad_norm": 0.798211358388717,
+      "learning_rate": 8.625000000000001e-06,
+      "loss": 1.1389,
+      "step": 69
+    },
+    {
+      "epoch": 0.2192638997650744,
+      "grad_norm": 0.8027997517397614,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 1.1578,
+      "step": 70
+    },
+    {
+      "epoch": 0.22239624119028975,
+      "grad_norm": 0.7884275020306974,
+      "learning_rate": 8.875e-06,
+      "loss": 1.1476,
+      "step": 71
+    },
+    {
+      "epoch": 0.2255285826155051,
+      "grad_norm": 0.8162843570408501,
+      "learning_rate": 9e-06,
+      "loss": 1.197,
+      "step": 72
+    },
+    {
+      "epoch": 0.22866092404072044,
+      "grad_norm": 0.8195467948994536,
+      "learning_rate": 9.125e-06,
+      "loss": 1.1463,
+      "step": 73
+    },
+    {
+      "epoch": 0.2317932654659358,
+      "grad_norm": 0.785823621627204,
+      "learning_rate": 9.250000000000001e-06,
+      "loss": 1.13,
+      "step": 74
+    },
+    {
+      "epoch": 0.23492560689115113,
+      "grad_norm": 0.7858506830531828,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 1.1392,
+      "step": 75
+    },
+    {
+      "epoch": 0.23805794831636648,
+      "grad_norm": 0.8126411047255567,
+      "learning_rate": 9.5e-06,
+      "loss": 1.174,
+      "step": 76
+    },
+    {
+      "epoch": 0.24119028974158183,
+      "grad_norm": 0.8134240118545614,
+      "learning_rate": 9.625e-06,
+      "loss": 1.149,
+      "step": 77
+    },
+    {
+      "epoch": 0.24432263116679717,
+      "grad_norm": 2.0900195366866186,
+      "learning_rate": 9.75e-06,
+      "loss": 1.1669,
+      "step": 78
+    },
+    {
+      "epoch": 0.24745497259201252,
+      "grad_norm": 0.8049430353655749,
+      "learning_rate": 9.875000000000001e-06,
+      "loss": 1.1632,
+      "step": 79
+    },
+    {
+      "epoch": 0.2505873140172279,
+      "grad_norm": 0.8154348013283664,
+      "learning_rate": 1e-05,
+      "loss": 1.1752,
+      "step": 80
+    },
+    {
+      "epoch": 0.25371965544244324,
+      "grad_norm": 0.7766532231446668,
+      "learning_rate": 1.0125e-05,
+      "loss": 1.1562,
+      "step": 81
+    },
+    {
+      "epoch": 0.2568519968676586,
+      "grad_norm": 0.805338167968417,
+      "learning_rate": 1.025e-05,
+      "loss": 1.1875,
+      "step": 82
+    },
+    {
+      "epoch": 0.25998433829287393,
+      "grad_norm": 0.7634601102161442,
+      "learning_rate": 1.0375000000000001e-05,
+      "loss": 1.1216,
+      "step": 83
+    },
+    {
+      "epoch": 0.2631166797180893,
+      "grad_norm": 0.7954986672515693,
+      "learning_rate": 1.0500000000000001e-05,
+      "loss": 1.1681,
+      "step": 84
+    },
+    {
+      "epoch": 0.2662490211433046,
+      "grad_norm": 0.7666715336890474,
+      "learning_rate": 1.0625e-05,
+      "loss": 1.1383,
+      "step": 85
+    },
+    {
+      "epoch": 0.26938136256851997,
+      "grad_norm": 0.7841383740610983,
+      "learning_rate": 1.075e-05,
+      "loss": 1.1956,
+      "step": 86
+    },
+    {
+      "epoch": 0.2725137039937353,
+      "grad_norm": 0.7593684758700783,
+      "learning_rate": 1.0875e-05,
+      "loss": 1.1199,
+      "step": 87
+    },
+    {
+      "epoch": 0.27564604541895066,
+      "grad_norm": 0.7913739392990039,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 1.1665,
+      "step": 88
+    },
+    {
+      "epoch": 0.278778386844166,
+      "grad_norm": 0.8205958812550826,
+      "learning_rate": 1.1125000000000001e-05,
+      "loss": 1.1756,
+      "step": 89
+    },
+    {
+      "epoch": 0.28191072826938135,
+      "grad_norm": 0.7967279563962926,
+      "learning_rate": 1.125e-05,
+      "loss": 1.1559,
+      "step": 90
+    },
+    {
+      "epoch": 0.2850430696945967,
+      "grad_norm": 0.7673412238486342,
+      "learning_rate": 1.1375e-05,
+      "loss": 1.1744,
+      "step": 91
+    },
+    {
+      "epoch": 0.28817541111981204,
+      "grad_norm": 0.805175551828789,
+      "learning_rate": 1.15e-05,
+      "loss": 1.1629,
+      "step": 92
+    },
+    {
+      "epoch": 0.2913077525450274,
+      "grad_norm": 0.7505564301768121,
+      "learning_rate": 1.1625000000000001e-05,
+      "loss": 1.1384,
+      "step": 93
+    },
+    {
+      "epoch": 0.29444009397024273,
+      "grad_norm": 0.8039055214719127,
+      "learning_rate": 1.1750000000000001e-05,
+      "loss": 1.146,
+      "step": 94
+    },
+    {
+      "epoch": 0.2975724353954581,
+      "grad_norm": 0.7825064856728625,
+      "learning_rate": 1.1875e-05,
+      "loss": 1.1644,
+      "step": 95
+    },
+    {
+      "epoch": 0.3007047768206735,
+      "grad_norm": 0.7935646900032081,
+      "learning_rate": 1.2e-05,
+      "loss": 1.1469,
+      "step": 96
+    },
+    {
+      "epoch": 0.3038371182458888,
+      "grad_norm": 0.8024905927017434,
+      "learning_rate": 1.2125e-05,
+      "loss": 1.1402,
+      "step": 97
+    },
+    {
+      "epoch": 0.3069694596711042,
+      "grad_norm": 0.7830126081823029,
+      "learning_rate": 1.2250000000000001e-05,
+      "loss": 1.1733,
+      "step": 98
+    },
+    {
+      "epoch": 0.3101018010963195,
+      "grad_norm": 0.7997702404780472,
+      "learning_rate": 1.2375000000000001e-05,
+      "loss": 1.1991,
+      "step": 99
+    },
+    {
+      "epoch": 0.31323414252153486,
+      "grad_norm": 0.7557112262293625,
+      "learning_rate": 1.25e-05,
+      "loss": 1.1463,
+      "step": 100
+    },
+    {
+      "epoch": 0.3163664839467502,
+      "grad_norm": 0.7794955385488898,
+      "learning_rate": 1.2625e-05,
+      "loss": 1.192,
+      "step": 101
+    },
+    {
+      "epoch": 0.31949882537196556,
+      "grad_norm": 0.8098258584079975,
+      "learning_rate": 1.275e-05,
+      "loss": 1.213,
+      "step": 102
+    },
+    {
+      "epoch": 0.3226311667971809,
+      "grad_norm": 0.7578592334982148,
+      "learning_rate": 1.2875000000000001e-05,
+      "loss": 1.1538,
+      "step": 103
+    },
+    {
+      "epoch": 0.32576350822239625,
+      "grad_norm": 0.7801947192543522,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 1.1362,
+      "step": 104
+    },
+    {
+      "epoch": 0.3288958496476116,
+      "grad_norm": 0.803547746608655,
+      "learning_rate": 1.3125e-05,
+      "loss": 1.1337,
+      "step": 105
+    },
+    {
+      "epoch": 0.33202819107282694,
+      "grad_norm": 0.749613358405691,
+      "learning_rate": 1.325e-05,
+      "loss": 1.1038,
+      "step": 106
+    },
+    {
+      "epoch": 0.3351605324980423,
+      "grad_norm": 0.8427226009611682,
+      "learning_rate": 1.3375e-05,
+      "loss": 1.1611,
+      "step": 107
+    },
+    {
+      "epoch": 0.33829287392325763,
+      "grad_norm": 0.8147379441469031,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 1.158,
+      "step": 108
+    },
+    {
+      "epoch": 0.341425215348473,
+      "grad_norm": 0.7724455205425795,
+      "learning_rate": 1.3625e-05,
+      "loss": 1.1123,
+      "step": 109
+    },
+    {
+      "epoch": 0.3445575567736883,
+      "grad_norm": 0.7752938561359628,
+      "learning_rate": 1.375e-05,
+      "loss": 1.1516,
+      "step": 110
+    },
+    {
+      "epoch": 0.34768989819890367,
+      "grad_norm": 0.7714492124885363,
+      "learning_rate": 1.3875e-05,
+      "loss": 1.1251,
+      "step": 111
+    },
+    {
+      "epoch": 0.350822239624119,
+      "grad_norm": 0.7951581981638454,
+      "learning_rate": 1.4e-05,
+      "loss": 1.1362,
+      "step": 112
+    },
+    {
+      "epoch": 0.35395458104933436,
+      "grad_norm": 0.7654142695855861,
+      "learning_rate": 1.4125000000000003e-05,
+      "loss": 1.1494,
+      "step": 113
+    },
+    {
+      "epoch": 0.3570869224745497,
+      "grad_norm": 0.7809679002598582,
+      "learning_rate": 1.425e-05,
+      "loss": 1.1533,
+      "step": 114
+    },
+    {
+      "epoch": 0.36021926389976505,
+      "grad_norm": 0.8081986701754312,
+      "learning_rate": 1.4375e-05,
+      "loss": 1.1654,
+      "step": 115
+    },
+    {
+      "epoch": 0.3633516053249804,
+      "grad_norm": 0.8085004016002348,
+      "learning_rate": 1.45e-05,
+      "loss": 1.1343,
+      "step": 116
+    },
+    {
+      "epoch": 0.36648394675019574,
+      "grad_norm": 0.8329554845008073,
+      "learning_rate": 1.4625e-05,
+      "loss": 1.1569,
+      "step": 117
+    },
+    {
+      "epoch": 0.36961628817541115,
+      "grad_norm": 0.783780585930527,
+      "learning_rate": 1.4750000000000003e-05,
+      "loss": 1.125,
+      "step": 118
+    },
+    {
+      "epoch": 0.3727486296006265,
+      "grad_norm": 0.7912033183545758,
+      "learning_rate": 1.4875000000000002e-05,
+      "loss": 1.1073,
+      "step": 119
+    },
+    {
+      "epoch": 0.37588097102584184,
+      "grad_norm": 0.821268337393993,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.1386,
+      "step": 120
+    },
+    {
+      "epoch": 0.3790133124510572,
+      "grad_norm": 0.832261258255235,
+      "learning_rate": 1.5125e-05,
+      "loss": 1.1683,
+      "step": 121
+    },
+    {
+      "epoch": 0.38214565387627253,
+      "grad_norm": 0.8094345539354894,
+      "learning_rate": 1.525e-05,
+      "loss": 1.1508,
+      "step": 122
+    },
+    {
+      "epoch": 0.3852779953014879,
+      "grad_norm": 0.8008883805537554,
+      "learning_rate": 1.5375e-05,
+      "loss": 1.1111,
+      "step": 123
+    },
+    {
+      "epoch": 0.3884103367267032,
+      "grad_norm": 0.7915578589687696,
+      "learning_rate": 1.55e-05,
+      "loss": 1.1479,
+      "step": 124
+    },
+    {
+      "epoch": 0.39154267815191857,
+      "grad_norm": 0.8327731000166747,
+      "learning_rate": 1.5625e-05,
+      "loss": 1.1777,
+      "step": 125
+    },
+    {
+      "epoch": 0.3946750195771339,
+      "grad_norm": 0.8189737080837218,
+      "learning_rate": 1.575e-05,
+      "loss": 1.1558,
+      "step": 126
+    },
+    {
+      "epoch": 0.39780736100234926,
+      "grad_norm": 0.7783531314158876,
+      "learning_rate": 1.5875e-05,
+      "loss": 1.1095,
+      "step": 127
+    },
+    {
+      "epoch": 0.4009397024275646,
+      "grad_norm": 0.8543936473595438,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.1427,
+      "step": 128
+    },
+    {
+      "epoch": 0.40407204385277995,
+      "grad_norm": 0.8243727971086049,
+      "learning_rate": 1.6125000000000002e-05,
+      "loss": 1.1636,
+      "step": 129
+    },
+    {
+      "epoch": 0.4072043852779953,
+      "grad_norm": 0.7919846838112398,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 1.1753,
+      "step": 130
+    },
+    {
+      "epoch": 0.41033672670321064,
+      "grad_norm": 0.7953800117937856,
+      "learning_rate": 1.6375e-05,
+      "loss": 1.1121,
+      "step": 131
+    },
+    {
+      "epoch": 0.413469068128426,
+      "grad_norm": 0.7941547055326952,
+      "learning_rate": 1.65e-05,
+      "loss": 1.1759,
+      "step": 132
+    },
+    {
+      "epoch": 0.41660140955364133,
+      "grad_norm": 0.751465418050413,
+      "learning_rate": 1.6625e-05,
+      "loss": 1.1131,
+      "step": 133
+    },
+    {
+      "epoch": 0.4197337509788567,
+      "grad_norm": 0.782469109387437,
+      "learning_rate": 1.675e-05,
+      "loss": 1.1319,
+      "step": 134
+    },
+    {
+      "epoch": 0.422866092404072,
+      "grad_norm": 0.7537743415586728,
+      "learning_rate": 1.6875e-05,
+      "loss": 1.123,
+      "step": 135
+    },
+    {
+      "epoch": 0.42599843382928737,
+      "grad_norm": 0.8266848129941319,
+      "learning_rate": 1.7e-05,
+      "loss": 1.2021,
+      "step": 136
+    },
+    {
+      "epoch": 0.4291307752545027,
+      "grad_norm": 0.7907070577194092,
+      "learning_rate": 1.7125e-05,
+      "loss": 1.1311,
+      "step": 137
+    },
+    {
+      "epoch": 0.43226311667971806,
+      "grad_norm": 0.770956600368574,
+      "learning_rate": 1.7250000000000003e-05,
+      "loss": 1.1406,
+      "step": 138
+    },
+    {
+      "epoch": 0.43539545810493346,
+      "grad_norm": 0.7942699632538319,
+      "learning_rate": 1.7375000000000002e-05,
+      "loss": 1.121,
+      "step": 139
+    },
+    {
+      "epoch": 0.4385277995301488,
+      "grad_norm": 0.7669673284070234,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 1.1206,
+      "step": 140
+    },
+    {
+      "epoch": 0.44166014095536416,
+      "grad_norm": 0.7910094531836759,
+      "learning_rate": 1.7625e-05,
+      "loss": 1.1069,
+      "step": 141
+    },
+    {
+      "epoch": 0.4447924823805795,
+      "grad_norm": 0.7711860139608495,
+      "learning_rate": 1.775e-05,
+      "loss": 1.1422,
+      "step": 142
+    },
+    {
+      "epoch": 0.44792482380579485,
+      "grad_norm": 0.7957935969277278,
+      "learning_rate": 1.7875e-05,
+      "loss": 1.1059,
+      "step": 143
+    },
+    {
+      "epoch": 0.4510571652310102,
+      "grad_norm": 0.7632895910705135,
+      "learning_rate": 1.8e-05,
+      "loss": 1.1327,
+      "step": 144
+    },
+    {
+      "epoch": 0.45418950665622554,
+      "grad_norm": 0.7972939663996078,
+      "learning_rate": 1.8125e-05,
+      "loss": 1.1898,
+      "step": 145
+    },
+    {
+      "epoch": 0.4573218480814409,
+      "grad_norm": 0.768609653869839,
+      "learning_rate": 1.825e-05,
+      "loss": 1.1361,
+      "step": 146
+    },
+    {
+      "epoch": 0.46045418950665623,
+      "grad_norm": 0.74095309811706,
+      "learning_rate": 1.8375e-05,
+      "loss": 1.0726,
+      "step": 147
+    },
+    {
+      "epoch": 0.4635865309318716,
+      "grad_norm": 0.7937353734822679,
+      "learning_rate": 1.8500000000000002e-05,
+      "loss": 1.1412,
+      "step": 148
+    },
+    {
+      "epoch": 0.4667188723570869,
+      "grad_norm": 0.8231685857372454,
+      "learning_rate": 1.8625000000000002e-05,
+      "loss": 1.1928,
+      "step": 149
+    },
+    {
+      "epoch": 0.46985121378230227,
+      "grad_norm": 0.7582039516040869,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 1.1157,
+      "step": 150
+    },
+    {
+      "epoch": 0.4729835552075176,
+      "grad_norm": 0.7945667336329015,
+      "learning_rate": 1.8875e-05,
+      "loss": 1.1339,
+      "step": 151
+    },
+    {
+      "epoch": 0.47611589663273296,
+      "grad_norm": 0.7511917754873464,
+      "learning_rate": 1.9e-05,
+      "loss": 1.115,
+      "step": 152
+    },
+    {
+      "epoch": 0.4792482380579483,
+      "grad_norm": 0.7567481788127869,
+      "learning_rate": 1.9125000000000004e-05,
+      "loss": 1.1438,
+      "step": 153
+    },
+    {
+      "epoch": 0.48238057948316365,
+      "grad_norm": 0.8035923511844805,
+      "learning_rate": 1.925e-05,
+      "loss": 1.1488,
+      "step": 154
+    },
+    {
+      "epoch": 0.485512920908379,
+      "grad_norm": 0.7832113472335267,
+      "learning_rate": 1.9375e-05,
+      "loss": 1.1473,
+      "step": 155
+    },
+    {
+      "epoch": 0.48864526233359434,
+      "grad_norm": 0.7835793180590945,
+      "learning_rate": 1.95e-05,
+      "loss": 1.2008,
+      "step": 156
+    },
+    {
+      "epoch": 0.4917776037588097,
+      "grad_norm": 0.8058256022750308,
+      "learning_rate": 1.9625e-05,
+      "loss": 1.148,
+      "step": 157
+    },
+    {
+      "epoch": 0.49490994518402504,
+      "grad_norm": 0.7847452922181013,
+      "learning_rate": 1.9750000000000002e-05,
+      "loss": 1.1196,
+      "step": 158
+    },
+    {
+      "epoch": 0.4980422866092404,
+      "grad_norm": 0.7812512360657365,
+      "learning_rate": 1.9875000000000002e-05,
+      "loss": 1.1419,
+      "step": 159
+    },
+    {
+      "epoch": 0.5011746280344558,
+      "grad_norm": 0.8824042145521714,
+      "learning_rate": 2e-05,
+      "loss": 1.0868,
+      "step": 160
+    },
+    {
+      "epoch": 0.5043069694596711,
+      "grad_norm": 0.7567570291946419,
+      "learning_rate": 1.9999976035643366e-05,
+      "loss": 1.1777,
+      "step": 161
+    },
+    {
+      "epoch": 0.5074393108848865,
+      "grad_norm": 0.7952722559164349,
+      "learning_rate": 1.9999904142688313e-05,
+      "loss": 1.1345,
+      "step": 162
+    },
+    {
+      "epoch": 0.5105716523101018,
+      "grad_norm": 0.7743964880190665,
+      "learning_rate": 1.9999784321479416e-05,
+      "loss": 1.1041,
+      "step": 163
+    },
+    {
+      "epoch": 0.5137039937353172,
+      "grad_norm": 0.8498468306677794,
+      "learning_rate": 1.999961657259097e-05,
+      "loss": 1.1567,
+      "step": 164
+    },
+    {
+      "epoch": 0.5168363351605325,
+      "grad_norm": 0.7947824846601378,
+      "learning_rate": 1.9999400896826965e-05,
+      "loss": 1.1733,
+      "step": 165
+    },
+    {
+      "epoch": 0.5199686765857479,
+      "grad_norm": 0.7487074186272767,
+      "learning_rate": 1.999913729522111e-05,
+      "loss": 1.161,
+      "step": 166
+    },
+    {
+      "epoch": 0.5231010180109632,
+      "grad_norm": 0.8037325407031123,
+      "learning_rate": 1.999882576903682e-05,
+      "loss": 1.1646,
+      "step": 167
+    },
+    {
+      "epoch": 0.5262333594361785,
+      "grad_norm": 0.7799604031117069,
+      "learning_rate": 1.9998466319767188e-05,
+      "loss": 1.153,
+      "step": 168
+    },
+    {
+      "epoch": 0.5293657008613939,
+      "grad_norm": 0.8012827006886666,
+      "learning_rate": 1.9998058949135016e-05,
+      "loss": 1.2302,
+      "step": 169
+    },
+    {
+      "epoch": 0.5324980422866092,
+      "grad_norm": 0.7752213646522256,
+      "learning_rate": 1.9997603659092773e-05,
+      "loss": 1.1444,
+      "step": 170
+    },
+    {
+      "epoch": 0.5356303837118246,
+      "grad_norm": 0.7880510647431511,
+      "learning_rate": 1.9997100451822616e-05,
+      "loss": 1.1522,
+      "step": 171
+    },
+    {
+      "epoch": 0.5387627251370399,
+      "grad_norm": 0.802804888286548,
+      "learning_rate": 1.9996549329736345e-05,
+      "loss": 1.1926,
+      "step": 172
+    },
+    {
+      "epoch": 0.5418950665622553,
+      "grad_norm": 0.7425430954083608,
+      "learning_rate": 1.9995950295475422e-05,
+      "loss": 1.1118,
+      "step": 173
+    },
+    {
+      "epoch": 0.5450274079874706,
+      "grad_norm": 0.794380669252506,
+      "learning_rate": 1.9995303351910934e-05,
+      "loss": 1.1748,
+      "step": 174
+    },
+    {
+      "epoch": 0.548159749412686,
+      "grad_norm": 0.7533240888688524,
+      "learning_rate": 1.9994608502143603e-05,
+      "loss": 1.1243,
+      "step": 175
+    },
+    {
+      "epoch": 0.5512920908379013,
+      "grad_norm": 0.7646481885397147,
+      "learning_rate": 1.9993865749503754e-05,
+      "loss": 1.1415,
+      "step": 176
+    },
+    {
+      "epoch": 0.5544244322631167,
+      "grad_norm": 0.7915751390688428,
+      "learning_rate": 1.9993075097551302e-05,
+      "loss": 1.1005,
+      "step": 177
+    },
+    {
+      "epoch": 0.557556773688332,
+      "grad_norm": 0.7921308350778985,
+      "learning_rate": 1.9992236550075746e-05,
+      "loss": 1.1215,
+      "step": 178
+    },
+    {
+      "epoch": 0.5606891151135474,
+      "grad_norm": 0.7469409103210944,
+      "learning_rate": 1.9991350111096133e-05,
+      "loss": 1.1023,
+      "step": 179
+    },
+    {
+      "epoch": 0.5638214565387627,
+      "grad_norm": 0.7382348346776821,
+      "learning_rate": 1.999041578486105e-05,
+      "loss": 1.1578,
+      "step": 180
+    },
+    {
+      "epoch": 0.566953797963978,
+      "grad_norm": 0.7926745472354245,
+      "learning_rate": 1.99894335758486e-05,
+      "loss": 1.1692,
+      "step": 181
+    },
+    {
+      "epoch": 0.5700861393891934,
+      "grad_norm": 0.7569926636991706,
+      "learning_rate": 1.998840348876639e-05,
+      "loss": 1.1385,
+      "step": 182
+    },
+    {
+      "epoch": 0.5732184808144087,
+      "grad_norm": 0.7458272413839812,
+      "learning_rate": 1.9987325528551492e-05,
+      "loss": 1.1332,
+      "step": 183
+    },
+    {
+      "epoch": 0.5763508222396241,
+      "grad_norm": 0.7842596742794642,
+      "learning_rate": 1.9986199700370435e-05,
+      "loss": 1.1599,
+      "step": 184
+    },
+    {
+      "epoch": 0.5794831636648394,
+      "grad_norm": 0.7560805393380029,
+      "learning_rate": 1.998502600961916e-05,
+      "loss": 1.0842,
+      "step": 185
+    },
+    {
+      "epoch": 0.5826155050900548,
+      "grad_norm": 0.7498681844673027,
+      "learning_rate": 1.9983804461923022e-05,
+      "loss": 1.1539,
+      "step": 186
+    },
+    {
+      "epoch": 0.5857478465152701,
+      "grad_norm": 0.7698674252021521,
+      "learning_rate": 1.998253506313674e-05,
+      "loss": 1.1244,
+      "step": 187
+    },
+    {
+      "epoch": 0.5888801879404855,
+      "grad_norm": 0.7668778266672499,
+      "learning_rate": 1.998121781934438e-05,
+      "loss": 1.1268,
+      "step": 188
+    },
+    {
+      "epoch": 0.5920125293657008,
+      "grad_norm": 0.7473755501240149,
+      "learning_rate": 1.997985273685932e-05,
+      "loss": 1.0964,
+      "step": 189
+    },
+    {
+      "epoch": 0.5951448707909162,
+      "grad_norm": 0.7456163683549248,
+      "learning_rate": 1.9978439822224228e-05,
+      "loss": 1.1468,
+      "step": 190
+    },
+    {
+      "epoch": 0.5982772122161315,
+      "grad_norm": 0.7586612489506767,
+      "learning_rate": 1.997697908221102e-05,
+      "loss": 1.1148,
+      "step": 191
+    },
+    {
+      "epoch": 0.601409553641347,
+      "grad_norm": 0.7526265106209291,
+      "learning_rate": 1.9975470523820835e-05,
+      "loss": 1.1596,
+      "step": 192
+    },
+    {
+      "epoch": 0.6045418950665623,
+      "grad_norm": 0.7515061295922136,
+      "learning_rate": 1.9973914154284e-05,
+      "loss": 1.132,
+      "step": 193
+    },
+    {
+      "epoch": 0.6076742364917777,
+      "grad_norm": 0.7628083735034853,
+      "learning_rate": 1.997230998105999e-05,
+      "loss": 1.0992,
+      "step": 194
+    },
+    {
+      "epoch": 0.610806577916993,
+      "grad_norm": 0.7372667581302207,
+      "learning_rate": 1.9970658011837404e-05,
+      "loss": 1.1549,
+      "step": 195
+    },
+    {
+      "epoch": 0.6139389193422083,
+      "grad_norm": 0.7570962352925633,
+      "learning_rate": 1.9968958254533922e-05,
+      "loss": 1.1437,
+      "step": 196
+    },
+    {
+      "epoch": 0.6170712607674237,
+      "grad_norm": 0.7760513349389108,
+      "learning_rate": 1.9967210717296256e-05,
+      "loss": 1.1143,
+      "step": 197
+    },
+    {
+      "epoch": 0.620203602192639,
+      "grad_norm": 0.7710793602174225,
+      "learning_rate": 1.996541540850013e-05,
+      "loss": 1.1611,
+      "step": 198
+    },
+    {
+      "epoch": 0.6233359436178544,
+      "grad_norm": 0.7497240836215765,
+      "learning_rate": 1.9963572336750227e-05,
+      "loss": 1.0894,
+      "step": 199
+    },
+    {
+      "epoch": 0.6264682850430697,
+      "grad_norm": 0.769485721788448,
+      "learning_rate": 1.996168151088015e-05,
+      "loss": 1.1326,
+      "step": 200
+    },
+    {
+      "epoch": 0.6296006264682851,
+      "grad_norm": 0.7914744857793413,
+      "learning_rate": 1.9959742939952393e-05,
+      "loss": 1.1642,
+      "step": 201
+    },
+    {
+      "epoch": 0.6327329678935004,
+      "grad_norm": 0.7749252132709005,
+      "learning_rate": 1.9957756633258264e-05,
+      "loss": 1.1097,
+      "step": 202
+    },
+    {
+      "epoch": 0.6358653093187158,
+      "grad_norm": 0.7248448375095949,
+      "learning_rate": 1.9955722600317887e-05,
+      "loss": 1.1149,
+      "step": 203
+    },
+    {
+      "epoch": 0.6389976507439311,
+      "grad_norm": 0.7893182002778344,
+      "learning_rate": 1.9953640850880118e-05,
+      "loss": 1.1343,
+      "step": 204
+    },
+    {
+      "epoch": 0.6421299921691465,
+      "grad_norm": 0.752880032513866,
+      "learning_rate": 1.995151139492251e-05,
+      "loss": 1.1393,
+      "step": 205
+    },
+    {
+      "epoch": 0.6452623335943618,
+      "grad_norm": 0.7715420215410178,
+      "learning_rate": 1.994933424265127e-05,
+      "loss": 1.1498,
+      "step": 206
+    },
+    {
+      "epoch": 0.6483946750195771,
+      "grad_norm": 0.7359564724287257,
+      "learning_rate": 1.9947109404501216e-05,
+      "loss": 1.1758,
+      "step": 207
+    },
+    {
+      "epoch": 0.6515270164447925,
+      "grad_norm": 0.7587433580271375,
+      "learning_rate": 1.99448368911357e-05,
+      "loss": 1.1346,
+      "step": 208
+    },
+    {
+      "epoch": 0.6546593578700078,
+      "grad_norm": 0.7720483403367886,
+      "learning_rate": 1.99425167134466e-05,
+      "loss": 1.1527,
+      "step": 209
+    },
+    {
+      "epoch": 0.6577916992952232,
+      "grad_norm": 0.7504516082730889,
+      "learning_rate": 1.9940148882554223e-05,
+      "loss": 1.0908,
+      "step": 210
+    },
+    {
+      "epoch": 0.6609240407204385,
+      "grad_norm": 0.7583232566276451,
+      "learning_rate": 1.9937733409807273e-05,
+      "loss": 1.1387,
+      "step": 211
+    },
+    {
+      "epoch": 0.6640563821456539,
+      "grad_norm": 0.7275565863134074,
+      "learning_rate": 1.9935270306782804e-05,
+      "loss": 1.1061,
+      "step": 212
+    },
+    {
+      "epoch": 0.6671887235708692,
+      "grad_norm": 0.7308212360196517,
+      "learning_rate": 1.9932759585286154e-05,
+      "loss": 1.14,
+      "step": 213
+    },
+    {
+      "epoch": 0.6703210649960846,
+      "grad_norm": 0.7395935935887239,
+      "learning_rate": 1.9930201257350883e-05,
+      "loss": 1.1356,
+      "step": 214
+    },
+    {
+      "epoch": 0.6734534064212999,
+      "grad_norm": 0.7318764377322893,
+      "learning_rate": 1.9927595335238736e-05,
+      "loss": 1.1468,
+      "step": 215
+    },
+    {
+      "epoch": 0.6765857478465153,
+      "grad_norm": 0.716132066894738,
+      "learning_rate": 1.992494183143955e-05,
+      "loss": 1.1329,
+      "step": 216
+    },
+    {
+      "epoch": 0.6797180892717306,
+      "grad_norm": 0.783963003986274,
+      "learning_rate": 1.992224075867124e-05,
+      "loss": 1.134,
+      "step": 217
+    },
+    {
+      "epoch": 0.682850430696946,
+      "grad_norm": 0.7765840844001187,
+      "learning_rate": 1.991949212987969e-05,
+      "loss": 1.1525,
+      "step": 218
+    },
+    {
+      "epoch": 0.6859827721221613,
+      "grad_norm": 0.7664402025464453,
+      "learning_rate": 1.9916695958238736e-05,
+      "loss": 1.1363,
+      "step": 219
+    },
+    {
+      "epoch": 0.6891151135473766,
+      "grad_norm": 0.7619234106278223,
+      "learning_rate": 1.9913852257150053e-05,
+      "loss": 1.1043,
+      "step": 220
+    },
+    {
+      "epoch": 0.692247454972592,
+      "grad_norm": 0.7503302852552577,
+      "learning_rate": 1.9910961040243148e-05,
+      "loss": 1.1503,
+      "step": 221
+    },
+    {
+      "epoch": 0.6953797963978073,
+      "grad_norm": 0.7386647741445785,
+      "learning_rate": 1.9908022321375244e-05,
+      "loss": 1.126,
+      "step": 222
+    },
+    {
+      "epoch": 0.6985121378230227,
+      "grad_norm": 0.772006501836167,
+      "learning_rate": 1.9905036114631247e-05,
+      "loss": 1.1324,
+      "step": 223
+    },
+    {
+      "epoch": 0.701644479248238,
+      "grad_norm": 0.7498396218310553,
+      "learning_rate": 1.9902002434323658e-05,
+      "loss": 1.13,
+      "step": 224
+    },
+    {
+      "epoch": 0.7047768206734534,
+      "grad_norm": 0.7499406276159518,
+      "learning_rate": 1.9898921294992516e-05,
+      "loss": 1.136,
+      "step": 225
+    },
+    {
+      "epoch": 0.7079091620986687,
+      "grad_norm": 0.8050655298488321,
+      "learning_rate": 1.9895792711405332e-05,
+      "loss": 1.1725,
+      "step": 226
+    },
+    {
+      "epoch": 0.7110415035238841,
+      "grad_norm": 0.785084787806491,
+      "learning_rate": 1.9892616698556996e-05,
+      "loss": 1.1521,
+      "step": 227
+    },
+    {
+      "epoch": 0.7141738449490994,
+      "grad_norm": 0.7548865828004917,
+      "learning_rate": 1.9889393271669735e-05,
+      "loss": 1.1182,
+      "step": 228
+    },
+    {
+      "epoch": 0.7173061863743148,
+      "grad_norm": 0.7609216107863833,
+      "learning_rate": 1.9886122446193015e-05,
+      "loss": 1.1345,
+      "step": 229
+    },
+    {
+      "epoch": 0.7204385277995301,
+      "grad_norm": 0.7547249839650533,
+      "learning_rate": 1.9882804237803487e-05,
+      "loss": 1.1427,
+      "step": 230
+    },
+    {
+      "epoch": 0.7235708692247454,
+      "grad_norm": 0.7632257000147761,
+      "learning_rate": 1.9879438662404893e-05,
+      "loss": 1.1235,
+      "step": 231
+    },
+    {
+      "epoch": 0.7267032106499608,
+      "grad_norm": 0.7568637017446223,
+      "learning_rate": 1.9876025736128e-05,
+      "loss": 1.1301,
+      "step": 232
+    },
+    {
+      "epoch": 0.7298355520751761,
+      "grad_norm": 0.7420518311600437,
+      "learning_rate": 1.987256547533053e-05,
+      "loss": 1.1342,
+      "step": 233
+    },
+    {
+      "epoch": 0.7329678935003915,
+      "grad_norm": 0.7646439659344628,
+      "learning_rate": 1.9869057896597068e-05,
+      "loss": 1.1352,
+      "step": 234
+    },
+    {
+      "epoch": 0.7361002349256069,
+      "grad_norm": 0.7312821097164371,
+      "learning_rate": 1.9865503016738983e-05,
+      "loss": 1.0321,
+      "step": 235
+    },
+    {
+      "epoch": 0.7392325763508223,
+      "grad_norm": 0.7546957399103286,
+      "learning_rate": 1.9861900852794357e-05,
+      "loss": 1.1377,
+      "step": 236
+    },
+    {
+      "epoch": 0.7423649177760376,
+      "grad_norm": 0.7194994232421758,
+      "learning_rate": 1.9858251422027903e-05,
+      "loss": 1.1237,
+      "step": 237
+    },
+    {
+      "epoch": 0.745497259201253,
+      "grad_norm": 0.7785862815076194,
+      "learning_rate": 1.985455474193087e-05,
+      "loss": 1.1609,
+      "step": 238
+    },
+    {
+      "epoch": 0.7486296006264683,
+      "grad_norm": 0.740903933948647,
+      "learning_rate": 1.985081083022097e-05,
+      "loss": 1.0945,
+      "step": 239
+    },
+    {
+      "epoch": 0.7517619420516837,
+      "grad_norm": 0.7291780203060889,
+      "learning_rate": 1.984701970484229e-05,
+      "loss": 1.0856,
+      "step": 240
+    },
+    {
+      "epoch": 0.754894283476899,
+      "grad_norm": 0.7587452855951344,
+      "learning_rate": 1.9843181383965207e-05,
+      "loss": 1.1665,
+      "step": 241
+    },
+    {
+      "epoch": 0.7580266249021144,
+      "grad_norm": 0.7407242126385574,
+      "learning_rate": 1.98392958859863e-05,
+      "loss": 1.1194,
+      "step": 242
+    },
+    {
+      "epoch": 0.7611589663273297,
+      "grad_norm": 0.7518927137010917,
+      "learning_rate": 1.9835363229528255e-05,
+      "loss": 1.1213,
+      "step": 243
+    },
+    {
+      "epoch": 0.7642913077525451,
+      "grad_norm": 0.749707981917893,
+      "learning_rate": 1.9831383433439798e-05,
+      "loss": 1.1007,
+      "step": 244
+    },
+    {
+      "epoch": 0.7674236491777604,
+      "grad_norm": 0.7695157385802572,
+      "learning_rate": 1.982735651679557e-05,
+      "loss": 1.1203,
+      "step": 245
+    },
+    {
+      "epoch": 0.7705559906029757,
+      "grad_norm": 0.7438784077225071,
+      "learning_rate": 1.982328249889607e-05,
+      "loss": 1.135,
+      "step": 246
+    },
+    {
+      "epoch": 0.7736883320281911,
+      "grad_norm": 0.7415443518523247,
+      "learning_rate": 1.9819161399267545e-05,
+      "loss": 1.076,
+      "step": 247
+    },
+    {
+      "epoch": 0.7768206734534064,
+      "grad_norm": 0.7482623409973128,
+      "learning_rate": 1.9814993237661887e-05,
+      "loss": 1.1561,
+      "step": 248
+    },
+    {
+      "epoch": 0.7799530148786218,
+      "grad_norm": 0.7460111329694793,
+      "learning_rate": 1.9810778034056562e-05,
+      "loss": 1.1132,
+      "step": 249
+    },
+    {
+      "epoch": 0.7830853563038371,
+      "grad_norm": 0.7288330170842255,
+      "learning_rate": 1.98065158086545e-05,
+      "loss": 1.0535,
+      "step": 250
+    },
+    {
+      "epoch": 0.7862176977290525,
+      "grad_norm": 0.7314199802562354,
+      "learning_rate": 1.9802206581883992e-05,
+      "loss": 1.1316,
+      "step": 251
+    },
+    {
+      "epoch": 0.7893500391542678,
+      "grad_norm": 0.7804910252834903,
+      "learning_rate": 1.9797850374398617e-05,
+      "loss": 1.1219,
+      "step": 252
+    },
+    {
+      "epoch": 0.7924823805794832,
+      "grad_norm": 0.7298549749672437,
+      "learning_rate": 1.9793447207077114e-05,
+      "loss": 1.1124,
+      "step": 253
+    },
+    {
+      "epoch": 0.7956147220046985,
+      "grad_norm": 0.7551562547765204,
+      "learning_rate": 1.9788997101023294e-05,
+      "loss": 1.1458,
+      "step": 254
+    },
+    {
+      "epoch": 0.7987470634299139,
+      "grad_norm": 0.7641456778799793,
+      "learning_rate": 1.9784500077565944e-05,
+      "loss": 1.1743,
+      "step": 255
+    },
+    {
+      "epoch": 0.8018794048551292,
+      "grad_norm": 0.7350148397112547,
+      "learning_rate": 1.977995615825872e-05,
+      "loss": 1.137,
+      "step": 256
+    },
+    {
+      "epoch": 0.8050117462803446,
+      "grad_norm": 0.742949716719734,
+      "learning_rate": 1.9775365364880043e-05,
+      "loss": 1.1294,
+      "step": 257
+    },
+    {
+      "epoch": 0.8081440877055599,
+      "grad_norm": 0.7336184274603342,
+      "learning_rate": 1.9770727719432994e-05,
+      "loss": 1.13,
+      "step": 258
+    },
+    {
+      "epoch": 0.8112764291307752,
+      "grad_norm": 0.7150049627370941,
+      "learning_rate": 1.976604324414521e-05,
+      "loss": 1.079,
+      "step": 259
+    },
+    {
+      "epoch": 0.8144087705559906,
+      "grad_norm": 0.729004714169207,
+      "learning_rate": 1.9761311961468782e-05,
+      "loss": 1.0903,
+      "step": 260
+    },
+    {
+      "epoch": 0.8175411119812059,
+      "grad_norm": 0.7432345451419001,
+      "learning_rate": 1.9756533894080132e-05,
+      "loss": 1.1718,
+      "step": 261
+    },
+    {
+      "epoch": 0.8206734534064213,
+      "grad_norm": 0.7485136400635154,
+      "learning_rate": 1.975170906487993e-05,
+      "loss": 1.123,
+      "step": 262
+    },
+    {
+      "epoch": 0.8238057948316366,
+      "grad_norm": 0.7792494517105039,
+      "learning_rate": 1.9746837496992953e-05,
+      "loss": 1.1018,
+      "step": 263
+    },
+    {
+      "epoch": 0.826938136256852,
+      "grad_norm": 0.7534542131696997,
+      "learning_rate": 1.974191921376801e-05,
+      "loss": 1.1376,
+      "step": 264
+    },
+    {
+      "epoch": 0.8300704776820673,
+      "grad_norm": 0.7486965183472643,
+      "learning_rate": 1.9736954238777793e-05,
+      "loss": 1.1112,
+      "step": 265
+    },
+    {
+      "epoch": 0.8332028191072827,
+      "grad_norm": 0.7675798633230227,
+      "learning_rate": 1.9731942595818786e-05,
+      "loss": 1.1506,
+      "step": 266
+    },
+    {
+      "epoch": 0.836335160532498,
+      "grad_norm": 0.7361146624834083,
+      "learning_rate": 1.9726884308911153e-05,
+      "loss": 1.1072,
+      "step": 267
+    },
+    {
+      "epoch": 0.8394675019577134,
+      "grad_norm": 0.7546626067336956,
+      "learning_rate": 1.9721779402298614e-05,
+      "loss": 1.105,
+      "step": 268
+    },
+    {
+      "epoch": 0.8425998433829287,
+      "grad_norm": 0.7592163829713124,
+      "learning_rate": 1.9716627900448323e-05,
+      "loss": 1.1045,
+      "step": 269
+    },
+    {
+      "epoch": 0.845732184808144,
+      "grad_norm": 0.7263452707909194,
+      "learning_rate": 1.971142982805077e-05,
+      "loss": 1.1131,
+      "step": 270
+    },
+    {
+      "epoch": 0.8488645262333594,
+      "grad_norm": 0.7354286459707577,
+      "learning_rate": 1.9706185210019648e-05,
+      "loss": 1.1676,
+      "step": 271
+    },
+    {
+      "epoch": 0.8519968676585747,
+      "grad_norm": 0.7423168880868267,
+      "learning_rate": 1.9700894071491736e-05,
+      "loss": 1.1529,
+      "step": 272
+    },
+    {
+      "epoch": 0.8551292090837901,
+      "grad_norm": 0.7524433622381994,
+      "learning_rate": 1.969555643782677e-05,
+      "loss": 1.11,
+      "step": 273
+    },
+    {
+      "epoch": 0.8582615505090054,
+      "grad_norm": 0.7553141589784657,
+      "learning_rate": 1.969017233460736e-05,
+      "loss": 1.1777,
+      "step": 274
+    },
+    {
+      "epoch": 0.8613938919342208,
+      "grad_norm": 0.7431732388608101,
+      "learning_rate": 1.9684741787638808e-05,
+      "loss": 1.1588,
+      "step": 275
+    },
+    {
+      "epoch": 0.8645262333594361,
+      "grad_norm": 0.7563523934795757,
+      "learning_rate": 1.9679264822949027e-05,
+      "loss": 1.1277,
+      "step": 276
+    },
+    {
+      "epoch": 0.8676585747846516,
+      "grad_norm": 0.743913029928575,
+      "learning_rate": 1.967374146678841e-05,
+      "loss": 1.0578,
+      "step": 277
+    },
+    {
+      "epoch": 0.8707909162098669,
+      "grad_norm": 0.76235114176879,
+      "learning_rate": 1.966817174562968e-05,
+      "loss": 1.1283,
+      "step": 278
+    },
+    {
+      "epoch": 0.8739232576350823,
+      "grad_norm": 0.7537117820426102,
+      "learning_rate": 1.9662555686167808e-05,
+      "loss": 1.1312,
+      "step": 279
+    },
+    {
+      "epoch": 0.8770555990602976,
+      "grad_norm": 0.7425818024035084,
+      "learning_rate": 1.9656893315319838e-05,
+      "loss": 1.1127,
+      "step": 280
+    },
+    {
+      "epoch": 0.880187940485513,
+      "grad_norm": 0.7902486001191769,
+      "learning_rate": 1.9651184660224786e-05,
+      "loss": 1.1369,
+      "step": 281
+    },
+    {
+      "epoch": 0.8833202819107283,
+      "grad_norm": 0.7580680784170756,
+      "learning_rate": 1.96454297482435e-05,
+      "loss": 1.0847,
+      "step": 282
+    },
+    {
+      "epoch": 0.8864526233359437,
+      "grad_norm": 0.7813627798317231,
+      "learning_rate": 1.9639628606958535e-05,
+      "loss": 1.0851,
+      "step": 283
+    },
+    {
+      "epoch": 0.889584964761159,
+      "grad_norm": 0.7613328270635236,
+      "learning_rate": 1.963378126417401e-05,
+      "loss": 1.077,
+      "step": 284
+    },
+    {
+      "epoch": 0.8927173061863743,
+      "grad_norm": 0.7773885553361843,
+      "learning_rate": 1.9627887747915496e-05,
+      "loss": 1.0678,
+      "step": 285
+    },
+    {
+      "epoch": 0.8958496476115897,
+      "grad_norm": 0.7650287423748288,
+      "learning_rate": 1.9621948086429847e-05,
+      "loss": 1.1636,
+      "step": 286
+    },
+    {
+      "epoch": 0.898981989036805,
+      "grad_norm": 0.7817877274905041,
+      "learning_rate": 1.96159623081851e-05,
+      "loss": 1.1262,
+      "step": 287
+    },
+    {
+      "epoch": 0.9021143304620204,
+      "grad_norm": 0.7527999272006599,
+      "learning_rate": 1.960993044187032e-05,
+      "loss": 1.1082,
+      "step": 288
+    },
+    {
+      "epoch": 0.9052466718872357,
+      "grad_norm": 0.7350684686021072,
+      "learning_rate": 1.9603852516395472e-05,
+      "loss": 1.1019,
+      "step": 289
+    },
+    {
+      "epoch": 0.9083790133124511,
+      "grad_norm": 0.7222753898804246,
+      "learning_rate": 1.9597728560891266e-05,
+      "loss": 1.1204,
+      "step": 290
+    },
+    {
+      "epoch": 0.9115113547376664,
+      "grad_norm": 0.7894532542241856,
+      "learning_rate": 1.9591558604709028e-05,
+      "loss": 1.1405,
+      "step": 291
+    },
+    {
+      "epoch": 0.9146436961628818,
+      "grad_norm": 0.7634432547529683,
+      "learning_rate": 1.958534267742057e-05,
+      "loss": 1.1368,
+      "step": 292
+    },
+    {
+      "epoch": 0.9177760375880971,
+      "grad_norm": 0.763149271096417,
+      "learning_rate": 1.9579080808818035e-05,
+      "loss": 1.122,
+      "step": 293
+    },
+    {
+      "epoch": 0.9209083790133125,
+      "grad_norm": 0.7380685951966892,
+      "learning_rate": 1.9572773028913744e-05,
+      "loss": 1.1044,
+      "step": 294
+    },
+    {
+      "epoch": 0.9240407204385278,
+      "grad_norm": 0.7455823416370709,
+      "learning_rate": 1.956641936794008e-05,
+      "loss": 1.1205,
+      "step": 295
+    },
+    {
+      "epoch": 0.9271730618637432,
+      "grad_norm": 0.7364688923639685,
+      "learning_rate": 1.956001985634932e-05,
+      "loss": 1.1512,
+      "step": 296
+    },
+    {
+      "epoch": 0.9303054032889585,
+      "grad_norm": 0.7134786654834753,
+      "learning_rate": 1.9553574524813507e-05,
+      "loss": 1.1056,
+      "step": 297
+    },
+    {
+      "epoch": 0.9334377447141738,
+      "grad_norm": 0.7170844865094715,
+      "learning_rate": 1.9547083404224273e-05,
+      "loss": 1.1325,
+      "step": 298
+    },
+    {
+      "epoch": 0.9365700861393892,
+      "grad_norm": 0.6823076058596949,
+      "learning_rate": 1.9540546525692736e-05,
+      "loss": 1.0178,
+      "step": 299
+    },
+    {
+      "epoch": 0.9397024275646045,
+      "grad_norm": 0.7416982640172877,
+      "learning_rate": 1.9533963920549307e-05,
+      "loss": 1.1293,
+      "step": 300
+    },
+    {
+      "epoch": 0.9428347689898199,
+      "grad_norm": 0.6949473720891305,
+      "learning_rate": 1.9527335620343567e-05,
+      "loss": 1.1152,
+      "step": 301
+    },
+    {
+      "epoch": 0.9459671104150352,
+      "grad_norm": 0.7300442646550352,
+      "learning_rate": 1.9520661656844107e-05,
+      "loss": 1.1324,
+      "step": 302
+    },
+    {
+      "epoch": 0.9490994518402506,
+      "grad_norm": 0.7453171429773002,
+      "learning_rate": 1.9513942062038375e-05,
+      "loss": 1.1833,
+      "step": 303
+    },
+    {
+      "epoch": 0.9522317932654659,
+      "grad_norm": 0.7387019005056833,
+      "learning_rate": 1.9507176868132517e-05,
+      "loss": 1.101,
+      "step": 304
+    },
+    {
+      "epoch": 0.9553641346906813,
+      "grad_norm": 0.7116175190537836,
+      "learning_rate": 1.950036610755125e-05,
+      "loss": 1.1204,
+      "step": 305
+    },
+    {
+      "epoch": 0.9584964761158966,
+      "grad_norm": 0.7466335049949244,
+      "learning_rate": 1.9493509812937666e-05,
+      "loss": 1.1376,
+      "step": 306
+    },
+    {
+      "epoch": 0.961628817541112,
+      "grad_norm": 0.7380075174756867,
+      "learning_rate": 1.94866080171531e-05,
+      "loss": 1.1225,
+      "step": 307
+    },
+    {
+      "epoch": 0.9647611589663273,
+      "grad_norm": 0.747788910002534,
+      "learning_rate": 1.9479660753276978e-05,
+      "loss": 1.1651,
+      "step": 308
+    },
+    {
+      "epoch": 0.9678935003915426,
+      "grad_norm": 0.7277939144026558,
+      "learning_rate": 1.947266805460664e-05,
+      "loss": 1.1005,
+      "step": 309
+    },
+    {
+      "epoch": 0.971025841816758,
+      "grad_norm": 0.7594540196531739,
+      "learning_rate": 1.9465629954657185e-05,
+      "loss": 1.131,
+      "step": 310
+    },
+    {
+      "epoch": 0.9741581832419733,
+      "grad_norm": 0.7685734409142018,
+      "learning_rate": 1.9458546487161326e-05,
+      "loss": 1.1645,
+      "step": 311
+    },
+    {
+      "epoch": 0.9772905246671887,
+      "grad_norm": 0.7221986487220532,
+      "learning_rate": 1.945141768606921e-05,
+      "loss": 1.0698,
+      "step": 312
+    },
+    {
+      "epoch": 0.980422866092404,
+      "grad_norm": 0.7136839321857378,
+      "learning_rate": 1.9444243585548266e-05,
+      "loss": 1.1085,
+      "step": 313
+    },
+    {
+      "epoch": 0.9835552075176194,
+      "grad_norm": 0.7257358526029265,
+      "learning_rate": 1.943702421998303e-05,
+      "loss": 1.088,
+      "step": 314
+    },
+    {
+      "epoch": 0.9866875489428347,
+      "grad_norm": 0.7283314263257034,
+      "learning_rate": 1.9429759623974992e-05,
+      "loss": 1.1275,
+      "step": 315
+    },
+    {
+      "epoch": 0.9898198903680501,
+      "grad_norm": 0.7150867610674063,
+      "learning_rate": 1.942244983234243e-05,
+      "loss": 1.1191,
+      "step": 316
+    },
+    {
+      "epoch": 0.9929522317932654,
+      "grad_norm": 0.7256772289345883,
+      "learning_rate": 1.9415094880120236e-05,
+      "loss": 1.1267,
+      "step": 317
+    },
+    {
+      "epoch": 0.9960845732184808,
+      "grad_norm": 0.7160733246362201,
+      "learning_rate": 1.940769480255974e-05,
+      "loss": 1.0924,
+      "step": 318
+    },
+    {
+      "epoch": 0.9992169146436961,
+      "grad_norm": 0.6968857896809572,
+      "learning_rate": 1.9400249635128575e-05,
+      "loss": 1.0343,
+      "step": 319
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6968857896809572,
+      "learning_rate": 1.939275941351046e-05,
+      "loss": 1.1068,
+      "step": 320
+    },
+    {
+      "epoch": 1.0031323414252153,
+      "grad_norm": 1.6156520613544338,
+      "learning_rate": 1.9385224173605072e-05,
+      "loss": 1.0187,
+      "step": 321
+    },
+    {
+      "epoch": 1.0062646828504307,
+      "grad_norm": 0.8615572642140844,
+      "learning_rate": 1.9377643951527834e-05,
+      "loss": 0.9501,
+      "step": 322
+    },
+    {
+      "epoch": 1.009397024275646,
+      "grad_norm": 0.8236605914660184,
+      "learning_rate": 1.937001878360979e-05,
+      "loss": 0.9672,
+      "step": 323
+    },
+    {
+      "epoch": 1.0125293657008614,
+      "grad_norm": 0.81342630340967,
+      "learning_rate": 1.9362348706397374e-05,
+      "loss": 0.9559,
+      "step": 324
+    },
+    {
+      "epoch": 1.0156617071260767,
+      "grad_norm": 0.8581319434910718,
+      "learning_rate": 1.935463375665229e-05,
+      "loss": 0.9808,
+      "step": 325
+    },
+    {
+      "epoch": 1.018794048551292,
+      "grad_norm": 0.8129587194790275,
+      "learning_rate": 1.934687397135129e-05,
+      "loss": 0.9512,
+      "step": 326
+    },
+    {
+      "epoch": 1.0219263899765074,
+      "grad_norm": 0.8327468346046575,
+      "learning_rate": 1.933906938768604e-05,
+      "loss": 0.9776,
+      "step": 327
+    },
+    {
+      "epoch": 1.0250587314017228,
+      "grad_norm": 0.8873213443616459,
+      "learning_rate": 1.9331220043062894e-05,
+      "loss": 0.9568,
+      "step": 328
+    },
+    {
+      "epoch": 1.0281910728269381,
+      "grad_norm": 0.8763873438765251,
+      "learning_rate": 1.9323325975102757e-05,
+      "loss": 0.9854,
+      "step": 329
+    },
+    {
+      "epoch": 1.0313234142521535,
+      "grad_norm": 0.8737645235450247,
+      "learning_rate": 1.9315387221640874e-05,
+      "loss": 0.9766,
+      "step": 330
+    },
+    {
+      "epoch": 1.0344557556773688,
+      "grad_norm": 0.8226677263540446,
+      "learning_rate": 1.9307403820726675e-05,
+      "loss": 0.9439,
+      "step": 331
+    },
+    {
+      "epoch": 1.0375880971025842,
+      "grad_norm": 0.8561637749159361,
+      "learning_rate": 1.929937581062357e-05,
+      "loss": 0.962,
+      "step": 332
+    },
+    {
+      "epoch": 1.0407204385277995,
+      "grad_norm": 0.8096315231114,
+      "learning_rate": 1.929130322980878e-05,
+      "loss": 0.9454,
+      "step": 333
+    },
+    {
+      "epoch": 1.0438527799530148,
+      "grad_norm": 0.8289922628570354,
+      "learning_rate": 1.9283186116973146e-05,
+      "loss": 0.9844,
+      "step": 334
+    },
+    {
+      "epoch": 1.0469851213782302,
+      "grad_norm": 0.7901830667535128,
+      "learning_rate": 1.927502451102095e-05,
+      "loss": 0.9493,
+      "step": 335
+    },
+    {
+      "epoch": 1.0501174628034455,
+      "grad_norm": 0.7761466037885246,
+      "learning_rate": 1.926681845106971e-05,
+      "loss": 0.9465,
+      "step": 336
+    },
+    {
+      "epoch": 1.0532498042286609,
+      "grad_norm": 0.8141355109509392,
+      "learning_rate": 1.925856797645002e-05,
+      "loss": 0.95,
+      "step": 337
+    },
+    {
+      "epoch": 1.0563821456538762,
+      "grad_norm": 0.8074412075844187,
+      "learning_rate": 1.925027312670534e-05,
+      "loss": 0.9436,
+      "step": 338
+    },
+    {
+      "epoch": 1.0595144870790916,
+      "grad_norm": 0.8272827849228525,
+      "learning_rate": 1.9241933941591826e-05,
+      "loss": 0.9703,
+      "step": 339
+    },
+    {
+      "epoch": 1.062646828504307,
+      "grad_norm": 0.7914982733889313,
+      "learning_rate": 1.9233550461078114e-05,
+      "loss": 0.9512,
+      "step": 340
+    },
+    {
+      "epoch": 1.0657791699295223,
+      "grad_norm": 0.7994475682723842,
+      "learning_rate": 1.9225122725345145e-05,
+      "loss": 0.9456,
+      "step": 341
+    },
+    {
+      "epoch": 1.0689115113547376,
+      "grad_norm": 0.8465580424768789,
+      "learning_rate": 1.9216650774785975e-05,
+      "loss": 0.9432,
+      "step": 342
+    },
+    {
+      "epoch": 1.072043852779953,
+      "grad_norm": 0.8466406131696012,
+      "learning_rate": 1.920813465000557e-05,
+      "loss": 0.9697,
+      "step": 343
+    },
+    {
+      "epoch": 1.0751761942051683,
+      "grad_norm": 0.8033646805720518,
+      "learning_rate": 1.9199574391820627e-05,
+      "loss": 0.9541,
+      "step": 344
+    },
+    {
+      "epoch": 1.0783085356303836,
+      "grad_norm": 0.8224052298742338,
+      "learning_rate": 1.9190970041259352e-05,
+      "loss": 1.0003,
+      "step": 345
+    },
+    {
+      "epoch": 1.081440877055599,
+      "grad_norm": 0.7997901412752633,
+      "learning_rate": 1.91823216395613e-05,
+      "loss": 0.9903,
+      "step": 346
+    },
+    {
+      "epoch": 1.0845732184808143,
+      "grad_norm": 0.8140939297756549,
+      "learning_rate": 1.917362922817714e-05,
+      "loss": 0.9549,
+      "step": 347
+    },
+    {
+      "epoch": 1.0877055599060297,
+      "grad_norm": 0.7824285972233995,
+      "learning_rate": 1.9164892848768485e-05,
+      "loss": 0.9593,
+      "step": 348
+    },
+    {
+      "epoch": 1.090837901331245,
+      "grad_norm": 0.7836791598591205,
+      "learning_rate": 1.9156112543207674e-05,
+      "loss": 0.9582,
+      "step": 349
+    },
+    {
+      "epoch": 1.0939702427564604,
+      "grad_norm": 0.7911753479916747,
+      "learning_rate": 1.914728835357759e-05,
+      "loss": 0.9348,
+      "step": 350
+    },
+    {
+      "epoch": 1.0971025841816757,
+      "grad_norm": 0.7782991822882203,
+      "learning_rate": 1.913842032217143e-05,
+      "loss": 0.9631,
+      "step": 351
+    },
+    {
+      "epoch": 1.100234925606891,
+      "grad_norm": 0.8014360461302352,
+      "learning_rate": 1.912950849149253e-05,
+      "loss": 0.9351,
+      "step": 352
+    },
+    {
+      "epoch": 1.1033672670321064,
+      "grad_norm": 0.8312030203793829,
+      "learning_rate": 1.912055290425415e-05,
+      "loss": 0.9668,
+      "step": 353
+    },
+    {
+      "epoch": 1.1064996084573218,
+      "grad_norm": 0.7891597416989514,
+      "learning_rate": 1.911155360337926e-05,
+      "loss": 0.9432,
+      "step": 354
+    },
+    {
+      "epoch": 1.109631949882537,
+      "grad_norm": 0.7878291420752258,
+      "learning_rate": 1.9102510632000365e-05,
+      "loss": 0.9521,
+      "step": 355
+    },
+    {
+      "epoch": 1.1127642913077525,
+      "grad_norm": 0.8120607285651226,
+      "learning_rate": 1.909342403345925e-05,
+      "loss": 0.9425,
+      "step": 356
+    },
+    {
+      "epoch": 1.1158966327329678,
+      "grad_norm": 0.7856758097045018,
+      "learning_rate": 1.9084293851306815e-05,
+      "loss": 0.9542,
+      "step": 357
+    },
+    {
+      "epoch": 1.1190289741581831,
+      "grad_norm": 0.7684160480129255,
+      "learning_rate": 1.9075120129302857e-05,
+      "loss": 0.9454,
+      "step": 358
+    },
+    {
+      "epoch": 1.1221613155833985,
+      "grad_norm": 0.8316731012269218,
+      "learning_rate": 1.9065902911415836e-05,
+      "loss": 0.9919,
+      "step": 359
+    },
+    {
+      "epoch": 1.1252936570086138,
+      "grad_norm": 0.822006226986475,
+      "learning_rate": 1.905664224182269e-05,
+      "loss": 0.9751,
+      "step": 360
+    },
+    {
+      "epoch": 1.1284259984338294,
+      "grad_norm": 0.7696798368031216,
+      "learning_rate": 1.9047338164908628e-05,
+      "loss": 0.9717,
+      "step": 361
+    },
+    {
+      "epoch": 1.1315583398590445,
+      "grad_norm": 0.8046009384769924,
+      "learning_rate": 1.903799072526688e-05,
+      "loss": 0.8908,
+      "step": 362
+    },
+    {
+      "epoch": 1.13469068128426,
+      "grad_norm": 0.7979550376850155,
+      "learning_rate": 1.9028599967698533e-05,
+      "loss": 0.9904,
+      "step": 363
+    },
+    {
+      "epoch": 1.1378230227094752,
+      "grad_norm": 0.8107407192621531,
+      "learning_rate": 1.9019165937212273e-05,
+      "loss": 0.9528,
+      "step": 364
+    },
+    {
+      "epoch": 1.1409553641346908,
+      "grad_norm": 0.7767024196508194,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.9159,
+      "step": 365
+    },
+    {
+      "epoch": 1.144087705559906,
+      "grad_norm": 0.7927579094268248,
+      "learning_rate": 1.9000168238557574e-05,
+      "loss": 0.9252,
+      "step": 366
+    },
+    {
+      "epoch": 1.1472200469851215,
+      "grad_norm": 0.8295313527055139,
+      "learning_rate": 1.899060466144267e-05,
+      "loss": 0.941,
+      "step": 367
+    },
+    {
+      "epoch": 1.1503523884103368,
+      "grad_norm": 0.7746260340598791,
+      "learning_rate": 1.8980997993516462e-05,
+      "loss": 0.9237,
+      "step": 368
+    },
+    {
+      "epoch": 1.1534847298355522,
+      "grad_norm": 0.7971878626092886,
+      "learning_rate": 1.897134828082248e-05,
+      "loss": 0.9387,
+      "step": 369
+    },
+    {
+      "epoch": 1.1566170712607675,
+      "grad_norm": 0.807852831490928,
+      "learning_rate": 1.8961655569610557e-05,
+      "loss": 0.9958,
+      "step": 370
+    },
+    {
+      "epoch": 1.1597494126859829,
+      "grad_norm": 0.8147989237592862,
+      "learning_rate": 1.895191990633661e-05,
+      "loss": 0.953,
+      "step": 371
+    },
+    {
+      "epoch": 1.1628817541111982,
+      "grad_norm": 0.7959883702024442,
+      "learning_rate": 1.8942141337662415e-05,
+      "loss": 0.9419,
+      "step": 372
+    },
+    {
+      "epoch": 1.1660140955364136,
+      "grad_norm": 0.7895917093808684,
+      "learning_rate": 1.89323199104554e-05,
+      "loss": 1.0,
+      "step": 373
+    },
+    {
+      "epoch": 1.169146436961629,
+      "grad_norm": 0.7788706927325837,
+      "learning_rate": 1.89224556717884e-05,
+      "loss": 0.9641,
+      "step": 374
+    },
+    {
+      "epoch": 1.1722787783868442,
+      "grad_norm": 0.8275735256549743,
+      "learning_rate": 1.891254866893944e-05,
+      "loss": 0.9596,
+      "step": 375
+    },
+    {
+      "epoch": 1.1754111198120596,
+      "grad_norm": 0.7901905557676397,
+      "learning_rate": 1.8902598949391507e-05,
+      "loss": 0.9881,
+      "step": 376
+    },
+    {
+      "epoch": 1.178543461237275,
+      "grad_norm": 0.7938940344571249,
+      "learning_rate": 1.8892606560832335e-05,
+      "loss": 0.9056,
+      "step": 377
+    },
+    {
+      "epoch": 1.1816758026624903,
+      "grad_norm": 0.7774862380646365,
+      "learning_rate": 1.8882571551154152e-05,
+      "loss": 0.9408,
+      "step": 378
+    },
+    {
+      "epoch": 1.1848081440877056,
+      "grad_norm": 0.8020328348235701,
+      "learning_rate": 1.8872493968453467e-05,
+      "loss": 0.9251,
+      "step": 379
+    },
+    {
+      "epoch": 1.187940485512921,
+      "grad_norm": 0.806279534293864,
+      "learning_rate": 1.8862373861030838e-05,
+      "loss": 0.948,
+      "step": 380
+    },
+    {
+      "epoch": 1.1910728269381363,
+      "grad_norm": 0.8025985732944699,
+      "learning_rate": 1.885221127739064e-05,
+      "loss": 0.9588,
+      "step": 381
+    },
+    {
+      "epoch": 1.1942051683633517,
+      "grad_norm": 0.7681611064314322,
+      "learning_rate": 1.8842006266240826e-05,
+      "loss": 0.9286,
+      "step": 382
+    },
+    {
+      "epoch": 1.197337509788567,
+      "grad_norm": 0.7830003805481669,
+      "learning_rate": 1.88317588764927e-05,
+      "loss": 0.9641,
+      "step": 383
+    },
+    {
+      "epoch": 1.2004698512137824,
+      "grad_norm": 0.7682889170068867,
+      "learning_rate": 1.8821469157260687e-05,
+      "loss": 0.9293,
+      "step": 384
+    },
+    {
+      "epoch": 1.2036021926389977,
+      "grad_norm": 0.8006318522209638,
+      "learning_rate": 1.8811137157862084e-05,
+      "loss": 0.9283,
+      "step": 385
+    },
+    {
+      "epoch": 1.206734534064213,
+      "grad_norm": 0.8064409365222602,
+      "learning_rate": 1.8800762927816835e-05,
+      "loss": 0.9974,
+      "step": 386
+    },
+    {
+      "epoch": 1.2098668754894284,
+      "grad_norm": 0.7771414813466695,
+      "learning_rate": 1.8790346516847288e-05,
+      "loss": 0.9414,
+      "step": 387
+    },
+    {
+      "epoch": 1.2129992169146437,
+      "grad_norm": 0.8516980588228736,
+      "learning_rate": 1.8779887974877965e-05,
+      "loss": 0.9708,
+      "step": 388
+    },
+    {
+      "epoch": 1.216131558339859,
+      "grad_norm": 0.8233081545979598,
+      "learning_rate": 1.876938735203531e-05,
+      "loss": 0.9856,
+      "step": 389
+    },
+    {
+      "epoch": 1.2192638997650744,
+      "grad_norm": 0.8148185121214713,
+      "learning_rate": 1.8758844698647457e-05,
+      "loss": 0.9441,
+      "step": 390
+    },
+    {
+      "epoch": 1.2223962411902898,
+      "grad_norm": 0.8121759062847519,
+      "learning_rate": 1.8748260065243985e-05,
+      "loss": 0.9865,
+      "step": 391
+    },
+    {
+      "epoch": 1.2255285826155051,
+      "grad_norm": 0.8027197651573186,
+      "learning_rate": 1.8737633502555685e-05,
+      "loss": 0.9809,
+      "step": 392
+    },
+    {
+      "epoch": 1.2286609240407205,
+      "grad_norm": 0.8058527578911673,
+      "learning_rate": 1.8726965061514297e-05,
+      "loss": 0.9611,
+      "step": 393
+    },
+    {
+      "epoch": 1.2317932654659358,
+      "grad_norm": 0.7952817742989116,
+      "learning_rate": 1.8716254793252294e-05,
+      "loss": 0.9786,
+      "step": 394
+    },
+    {
+      "epoch": 1.2349256068911512,
+      "grad_norm": 0.7839895990296804,
+      "learning_rate": 1.870550274910261e-05,
+      "loss": 0.9601,
+      "step": 395
+    },
+    {
+      "epoch": 1.2380579483163665,
+      "grad_norm": 0.8549762083059709,
+      "learning_rate": 1.869470898059841e-05,
+      "loss": 1.0219,
+      "step": 396
+    },
+    {
+      "epoch": 1.2411902897415819,
+      "grad_norm": 0.788953768471007,
+      "learning_rate": 1.868387353947284e-05,
+      "loss": 1.0143,
+      "step": 397
+    },
+    {
+      "epoch": 1.2443226311667972,
+      "grad_norm": 0.8217786523602275,
+      "learning_rate": 1.8672996477658767e-05,
+      "loss": 0.9523,
+      "step": 398
+    },
+    {
+      "epoch": 1.2474549725920125,
+      "grad_norm": 0.7704666004345789,
+      "learning_rate": 1.8662077847288554e-05,
+      "loss": 0.9207,
+      "step": 399
+    },
+    {
+      "epoch": 1.250587314017228,
+      "grad_norm": 0.7525516825767109,
+      "learning_rate": 1.8651117700693794e-05,
+      "loss": 0.9477,
+      "step": 400
+    },
+    {
+      "epoch": 1.2537196554424432,
+      "grad_norm": 0.8253629375930066,
+      "learning_rate": 1.8640116090405055e-05,
+      "loss": 0.9731,
+      "step": 401
+    },
+    {
+      "epoch": 1.2568519968676586,
+      "grad_norm": 0.7915558764489208,
+      "learning_rate": 1.8629073069151638e-05,
+      "loss": 0.9199,
+      "step": 402
+    },
+    {
+      "epoch": 1.259984338292874,
+      "grad_norm": 0.7782487189596384,
+      "learning_rate": 1.8617988689861333e-05,
+      "loss": 0.9199,
+      "step": 403
+    },
+    {
+      "epoch": 1.2631166797180893,
+      "grad_norm": 0.7980723005204162,
+      "learning_rate": 1.8606863005660133e-05,
+      "loss": 0.9212,
+      "step": 404
+    },
+    {
+      "epoch": 1.2662490211433046,
+      "grad_norm": 0.8048484955000166,
+      "learning_rate": 1.8595696069872013e-05,
+      "loss": 0.9581,
+      "step": 405
+    },
+    {
+      "epoch": 1.26938136256852,
+      "grad_norm": 0.8032453880110384,
+      "learning_rate": 1.8584487936018663e-05,
+      "loss": 0.9821,
+      "step": 406
+    },
+    {
+      "epoch": 1.2725137039937353,
+      "grad_norm": 0.8005507805303662,
+      "learning_rate": 1.8573238657819225e-05,
+      "loss": 0.9411,
+      "step": 407
+    },
+    {
+      "epoch": 1.2756460454189507,
+      "grad_norm": 0.7916521575263914,
+      "learning_rate": 1.8561948289190035e-05,
+      "loss": 0.9357,
+      "step": 408
+    },
+    {
+      "epoch": 1.278778386844166,
+      "grad_norm": 0.7979831477850623,
+      "learning_rate": 1.855061688424439e-05,
+      "loss": 0.9679,
+      "step": 409
+    },
+    {
+      "epoch": 1.2819107282693813,
+      "grad_norm": 0.7843000585547435,
+      "learning_rate": 1.8539244497292248e-05,
+      "loss": 0.9636,
+      "step": 410
+    },
+    {
+      "epoch": 1.2850430696945967,
+      "grad_norm": 0.784278644818112,
+      "learning_rate": 1.8527831182839997e-05,
+      "loss": 1.0007,
+      "step": 411
+    },
+    {
+      "epoch": 1.288175411119812,
+      "grad_norm": 0.7946831238148018,
+      "learning_rate": 1.8516376995590185e-05,
+      "loss": 0.9813,
+      "step": 412
+    },
+    {
+      "epoch": 1.2913077525450274,
+      "grad_norm": 0.8109210495548596,
+      "learning_rate": 1.850488199044126e-05,
+      "loss": 0.9349,
+      "step": 413
+    },
+    {
+      "epoch": 1.2944400939702427,
+      "grad_norm": 0.7934617847842895,
+      "learning_rate": 1.84933462224873e-05,
+      "loss": 0.9416,
+      "step": 414
+    },
+    {
+      "epoch": 1.297572435395458,
+      "grad_norm": 0.7902841843838354,
+      "learning_rate": 1.848176974701775e-05,
+      "loss": 0.9346,
+      "step": 415
+    },
+    {
+      "epoch": 1.3007047768206734,
+      "grad_norm": 0.779580351091375,
+      "learning_rate": 1.847015261951718e-05,
+      "loss": 0.9739,
+      "step": 416
+    },
+    {
+      "epoch": 1.3038371182458888,
+      "grad_norm": 0.7785529627916624,
+      "learning_rate": 1.845849489566498e-05,
+      "loss": 0.9583,
+      "step": 417
+    },
+    {
+      "epoch": 1.3069694596711041,
+      "grad_norm": 0.8436224516793075,
+      "learning_rate": 1.8446796631335123e-05,
+      "loss": 0.9681,
+      "step": 418
+    },
+    {
+      "epoch": 1.3101018010963195,
+      "grad_norm": 0.8155304400749016,
+      "learning_rate": 1.8435057882595885e-05,
+      "loss": 0.9455,
+      "step": 419
+    },
+    {
+      "epoch": 1.3132341425215348,
+      "grad_norm": 0.8154776896881121,
+      "learning_rate": 1.8423278705709573e-05,
+      "loss": 0.9319,
+      "step": 420
+    },
+    {
+      "epoch": 1.3163664839467502,
+      "grad_norm": 0.7875316068008983,
+      "learning_rate": 1.841145915713227e-05,
+      "loss": 0.9839,
+      "step": 421
+    },
+    {
+      "epoch": 1.3194988253719655,
+      "grad_norm": 0.7878642203922693,
+      "learning_rate": 1.8399599293513555e-05,
+      "loss": 0.9611,
+      "step": 422
+    },
+    {
+      "epoch": 1.3226311667971808,
+      "grad_norm": 0.8272412655909572,
+      "learning_rate": 1.8387699171696222e-05,
+      "loss": 0.9676,
+      "step": 423
+    },
+    {
+      "epoch": 1.3257635082223962,
+      "grad_norm": 0.8036668057730373,
+      "learning_rate": 1.837575884871603e-05,
+      "loss": 0.9585,
+      "step": 424
+    },
+    {
+      "epoch": 1.3288958496476115,
+      "grad_norm": 0.7745104838154117,
+      "learning_rate": 1.8363778381801404e-05,
+      "loss": 0.9586,
+      "step": 425
+    },
+    {
+      "epoch": 1.3320281910728269,
+      "grad_norm": 0.8044310075899959,
+      "learning_rate": 1.8351757828373183e-05,
+      "loss": 0.9773,
+      "step": 426
+    },
+    {
+      "epoch": 1.3351605324980422,
+      "grad_norm": 0.7816652761458432,
+      "learning_rate": 1.8339697246044333e-05,
+      "loss": 0.9483,
+      "step": 427
+    },
+    {
+      "epoch": 1.3382928739232576,
+      "grad_norm": 0.805576362415665,
+      "learning_rate": 1.8327596692619674e-05,
+      "loss": 1.0151,
+      "step": 428
+    },
+    {
+      "epoch": 1.341425215348473,
+      "grad_norm": 0.7789246864462362,
+      "learning_rate": 1.83154562260956e-05,
+      "loss": 0.9395,
+      "step": 429
+    },
+    {
+      "epoch": 1.3445575567736883,
+      "grad_norm": 0.7843379086528441,
+      "learning_rate": 1.8303275904659807e-05,
+      "loss": 0.9632,
+      "step": 430
+    },
+    {
+      "epoch": 1.3476898981989036,
+      "grad_norm": 0.7738505946833321,
+      "learning_rate": 1.829105578669101e-05,
+      "loss": 0.898,
+      "step": 431
+    },
+    {
+      "epoch": 1.350822239624119,
+      "grad_norm": 0.773409651313221,
+      "learning_rate": 1.8278795930758656e-05,
+      "loss": 0.925,
+      "step": 432
+    },
+    {
+      "epoch": 1.3539545810493343,
+      "grad_norm": 0.775638063691069,
+      "learning_rate": 1.826649639562266e-05,
+      "loss": 0.9296,
+      "step": 433
+    },
+    {
+      "epoch": 1.3570869224745497,
+      "grad_norm": 0.8255837346883753,
+      "learning_rate": 1.8254157240233114e-05,
+      "loss": 0.9578,
+      "step": 434
+    },
+    {
+      "epoch": 1.360219263899765,
+      "grad_norm": 0.8225702778393079,
+      "learning_rate": 1.8241778523729997e-05,
+      "loss": 0.9938,
+      "step": 435
+    },
+    {
+      "epoch": 1.3633516053249803,
+      "grad_norm": 0.8024681843740595,
+      "learning_rate": 1.8229360305442906e-05,
+      "loss": 0.9898,
+      "step": 436
+    },
+    {
+      "epoch": 1.3664839467501957,
+      "grad_norm": 0.7659268418287953,
+      "learning_rate": 1.8216902644890768e-05,
+      "loss": 0.9582,
+      "step": 437
+    },
+    {
+      "epoch": 1.3696162881754113,
+      "grad_norm": 0.7979578032515822,
+      "learning_rate": 1.8204405601781542e-05,
+      "loss": 0.9539,
+      "step": 438
+    },
+    {
+      "epoch": 1.3727486296006264,
+      "grad_norm": 0.7840678597712154,
+      "learning_rate": 1.819186923601195e-05,
+      "loss": 0.9662,
+      "step": 439
+    },
+    {
+      "epoch": 1.375880971025842,
+      "grad_norm": 0.7951677564440199,
+      "learning_rate": 1.8179293607667177e-05,
+      "loss": 0.963,
+      "step": 440
+    },
+    {
+      "epoch": 1.379013312451057,
+      "grad_norm": 0.7824747662278654,
+      "learning_rate": 1.8166678777020595e-05,
+      "loss": 0.9646,
+      "step": 441
+    },
+    {
+      "epoch": 1.3821456538762726,
+      "grad_norm": 0.7889899716482371,
+      "learning_rate": 1.8154024804533464e-05,
+      "loss": 0.9949,
+      "step": 442
+    },
+    {
+      "epoch": 1.3852779953014878,
+      "grad_norm": 0.8294307839008276,
+      "learning_rate": 1.8141331750854647e-05,
+      "loss": 0.9745,
+      "step": 443
+    },
+    {
+      "epoch": 1.3884103367267033,
+      "grad_norm": 0.8026550217216318,
+      "learning_rate": 1.8128599676820306e-05,
+      "loss": 0.9605,
+      "step": 444
+    },
+    {
+      "epoch": 1.3915426781519185,
+      "grad_norm": 0.8043494358409277,
+      "learning_rate": 1.8115828643453647e-05,
+      "loss": 0.9419,
+      "step": 445
+    },
+    {
+      "epoch": 1.394675019577134,
+      "grad_norm": 0.7904914993084787,
+      "learning_rate": 1.8103018711964588e-05,
+      "loss": 0.9636,
+      "step": 446
+    },
+    {
+      "epoch": 1.3978073610023491,
+      "grad_norm": 0.833878643616376,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.9799,
+      "step": 447
+    },
+    {
+      "epoch": 1.4009397024275647,
+      "grad_norm": 0.8351491373093172,
+      "learning_rate": 1.807728240039081e-05,
+      "loss": 0.954,
+      "step": 448
+    },
+    {
+      "epoch": 1.4040720438527798,
+      "grad_norm": 0.8413539861179071,
+      "learning_rate": 1.806435614365692e-05,
+      "loss": 0.9939,
+      "step": 449
+    },
+    {
+      "epoch": 1.4072043852779954,
+      "grad_norm": 0.8339283054897446,
+      "learning_rate": 1.8051391235501695e-05,
+      "loss": 0.9699,
+      "step": 450
+    },
+    {
+      "epoch": 1.4103367267032105,
+      "grad_norm": 0.8013532736907198,
+      "learning_rate": 1.8038387738064274e-05,
+      "loss": 0.9581,
+      "step": 451
+    },
+    {
+      "epoch": 1.413469068128426,
+      "grad_norm": 0.8255816811624092,
+      "learning_rate": 1.8025345713668746e-05,
+      "loss": 1.0,
+      "step": 452
+    },
+    {
+      "epoch": 1.4166014095536412,
+      "grad_norm": 0.8242603985956736,
+      "learning_rate": 1.8012265224823853e-05,
+      "loss": 0.9364,
+      "step": 453
+    },
+    {
+      "epoch": 1.4197337509788568,
+      "grad_norm": 0.8322278182809092,
+      "learning_rate": 1.79991463342227e-05,
+      "loss": 0.9851,
+      "step": 454
+    },
+    {
+      "epoch": 1.422866092404072,
+      "grad_norm": 0.8163638205578437,
+      "learning_rate": 1.7985989104742434e-05,
+      "loss": 0.9499,
+      "step": 455
+    },
+    {
+      "epoch": 1.4259984338292875,
+      "grad_norm": 0.8442359156970612,
+      "learning_rate": 1.7972793599443966e-05,
+      "loss": 0.9491,
+      "step": 456
+    },
+    {
+      "epoch": 1.4291307752545026,
+      "grad_norm": 0.7893355911335334,
+      "learning_rate": 1.795955988157166e-05,
+      "loss": 0.9816,
+      "step": 457
+    },
+    {
+      "epoch": 1.4322631166797182,
+      "grad_norm": 0.806471339752882,
+      "learning_rate": 1.7946288014553017e-05,
+      "loss": 0.9549,
+      "step": 458
+    },
+    {
+      "epoch": 1.4353954581049335,
+      "grad_norm": 0.8364165434031879,
+      "learning_rate": 1.7932978061998393e-05,
+      "loss": 0.9836,
+      "step": 459
+    },
+    {
+      "epoch": 1.4385277995301489,
+      "grad_norm": 0.7859146658803116,
+      "learning_rate": 1.7919630087700672e-05,
+      "loss": 0.9842,
+      "step": 460
+    },
+    {
+      "epoch": 1.4416601409553642,
+      "grad_norm": 0.7690910792708224,
+      "learning_rate": 1.790624415563498e-05,
+      "loss": 0.9396,
+      "step": 461
+    },
+    {
+      "epoch": 1.4447924823805796,
+      "grad_norm": 0.8496502770784555,
+      "learning_rate": 1.7892820329958368e-05,
+      "loss": 0.9579,
+      "step": 462
+    },
+    {
+      "epoch": 1.447924823805795,
+      "grad_norm": 0.8471130092608972,
+      "learning_rate": 1.7879358675009506e-05,
+      "loss": 0.9973,
+      "step": 463
+    },
+    {
+      "epoch": 1.4510571652310102,
+      "grad_norm": 0.784501551931219,
+      "learning_rate": 1.786585925530837e-05,
+      "loss": 0.9641,
+      "step": 464
+    },
+    {
+      "epoch": 1.4541895066562256,
+      "grad_norm": 0.836729539160783,
+      "learning_rate": 1.7852322135555946e-05,
+      "loss": 0.9319,
+      "step": 465
+    },
+    {
+      "epoch": 1.457321848081441,
+      "grad_norm": 0.8356458620469223,
+      "learning_rate": 1.7838747380633904e-05,
+      "loss": 0.9571,
+      "step": 466
+    },
+    {
+      "epoch": 1.4604541895066563,
+      "grad_norm": 0.8216191110677196,
+      "learning_rate": 1.78251350556043e-05,
+      "loss": 0.9553,
+      "step": 467
+    },
+    {
+      "epoch": 1.4635865309318716,
+      "grad_norm": 0.7966731084754197,
+      "learning_rate": 1.7811485225709255e-05,
+      "loss": 0.9519,
+      "step": 468
+    },
+    {
+      "epoch": 1.466718872357087,
+      "grad_norm": 0.869796047610942,
+      "learning_rate": 1.7797797956370646e-05,
+      "loss": 0.9836,
+      "step": 469
+    },
+    {
+      "epoch": 1.4698512137823023,
+      "grad_norm": 0.8527775209514664,
+      "learning_rate": 1.7784073313189796e-05,
+      "loss": 0.9559,
+      "step": 470
+    },
+    {
+      "epoch": 1.4729835552075177,
+      "grad_norm": 0.7722201957501883,
+      "learning_rate": 1.7770311361947155e-05,
+      "loss": 0.9424,
+      "step": 471
+    },
+    {
+      "epoch": 1.476115896632733,
+      "grad_norm": 0.8231174830960291,
+      "learning_rate": 1.775651216860198e-05,
+      "loss": 1.0208,
+      "step": 472
+    },
+    {
+      "epoch": 1.4792482380579484,
+      "grad_norm": 0.8323084233244761,
+      "learning_rate": 1.7742675799292033e-05,
+      "loss": 1.0034,
+      "step": 473
+    },
+    {
+      "epoch": 1.4823805794831637,
+      "grad_norm": 0.8153252414548278,
+      "learning_rate": 1.772880232033325e-05,
+      "loss": 0.9844,
+      "step": 474
+    },
+    {
+      "epoch": 1.485512920908379,
+      "grad_norm": 0.8189526410041873,
+      "learning_rate": 1.7714891798219432e-05,
+      "loss": 0.9666,
+      "step": 475
+    },
+    {
+      "epoch": 1.4886452623335944,
+      "grad_norm": 0.8078003198916532,
+      "learning_rate": 1.770094429962192e-05,
+      "loss": 0.944,
+      "step": 476
+    },
+    {
+      "epoch": 1.4917776037588097,
+      "grad_norm": 0.8110193552079895,
+      "learning_rate": 1.768695989138928e-05,
+      "loss": 0.9544,
+      "step": 477
+    },
+    {
+      "epoch": 1.494909945184025,
+      "grad_norm": 0.8312565557198122,
+      "learning_rate": 1.767293864054698e-05,
+      "loss": 0.9701,
+      "step": 478
+    },
+    {
+      "epoch": 1.4980422866092404,
+      "grad_norm": 0.8061090156479217,
+      "learning_rate": 1.7658880614297075e-05,
+      "loss": 0.9687,
+      "step": 479
+    },
+    {
+      "epoch": 1.5011746280344558,
+      "grad_norm": 0.823951800509433,
+      "learning_rate": 1.7644785880017874e-05,
+      "loss": 0.9723,
+      "step": 480
+    },
+    {
+      "epoch": 1.5043069694596711,
+      "grad_norm": 0.7965679907922073,
+      "learning_rate": 1.7630654505263628e-05,
+      "loss": 0.9073,
+      "step": 481
+    },
+    {
+      "epoch": 1.5074393108848865,
+      "grad_norm": 0.9569281276943931,
+      "learning_rate": 1.7616486557764187e-05,
+      "loss": 0.9808,
+      "step": 482
+    },
+    {
+      "epoch": 1.5105716523101018,
+      "grad_norm": 0.8085640633910255,
+      "learning_rate": 1.7602282105424712e-05,
+      "loss": 0.9732,
+      "step": 483
+    },
+    {
+      "epoch": 1.5137039937353172,
+      "grad_norm": 0.8532004513694563,
+      "learning_rate": 1.7588041216325314e-05,
+      "loss": 0.9507,
+      "step": 484
+    },
+    {
+      "epoch": 1.5168363351605325,
+      "grad_norm": 0.7833644189128819,
+      "learning_rate": 1.7573763958720736e-05,
+      "loss": 0.9411,
+      "step": 485
+    },
+    {
+      "epoch": 1.5199686765857479,
+      "grad_norm": 0.8189040727283652,
+      "learning_rate": 1.755945040104004e-05,
+      "loss": 0.9655,
+      "step": 486
+    },
+    {
+      "epoch": 1.5231010180109632,
+      "grad_norm": 0.7722121152681558,
+      "learning_rate": 1.754510061188627e-05,
+      "loss": 0.9185,
+      "step": 487
+    },
+    {
+      "epoch": 1.5262333594361785,
+      "grad_norm": 0.8037152368297265,
+      "learning_rate": 1.7530714660036112e-05,
+      "loss": 0.9541,
+      "step": 488
+    },
+    {
+      "epoch": 1.529365700861394,
+      "grad_norm": 0.8119393404590528,
+      "learning_rate": 1.7516292614439586e-05,
+      "loss": 0.9657,
+      "step": 489
+    },
+    {
+      "epoch": 1.5324980422866092,
+      "grad_norm": 0.805216081294044,
+      "learning_rate": 1.7501834544219697e-05,
+      "loss": 0.9448,
+      "step": 490
+    },
+    {
+      "epoch": 1.5356303837118246,
+      "grad_norm": 0.7949712346927176,
+      "learning_rate": 1.748734051867212e-05,
+      "loss": 0.9463,
+      "step": 491
+    },
+    {
+      "epoch": 1.53876272513704,
+      "grad_norm": 0.77990875986472,
+      "learning_rate": 1.7472810607264853e-05,
+      "loss": 0.9422,
+      "step": 492
+    },
+    {
+      "epoch": 1.5418950665622553,
+      "grad_norm": 0.7886283708612627,
+      "learning_rate": 1.7458244879637897e-05,
+      "loss": 0.9487,
+      "step": 493
+    },
+    {
+      "epoch": 1.5450274079874706,
+      "grad_norm": 0.7958418765522267,
+      "learning_rate": 1.74436434056029e-05,
+      "loss": 0.9714,
+      "step": 494
+    },
+    {
+      "epoch": 1.548159749412686,
+      "grad_norm": 0.8069518461832199,
+      "learning_rate": 1.742900625514285e-05,
+      "loss": 0.9829,
+      "step": 495
+    },
+    {
+      "epoch": 1.5512920908379013,
+      "grad_norm": 0.8073763682254401,
+      "learning_rate": 1.7414333498411734e-05,
+      "loss": 0.9569,
+      "step": 496
+    },
+    {
+      "epoch": 1.5544244322631167,
+      "grad_norm": 0.8342340049061451,
+      "learning_rate": 1.739962520573418e-05,
+      "loss": 1.0084,
+      "step": 497
+    },
+    {
+      "epoch": 1.557556773688332,
+      "grad_norm": 0.8120689308519743,
+      "learning_rate": 1.7384881447605144e-05,
+      "loss": 1.0443,
+      "step": 498
+    },
+    {
+      "epoch": 1.5606891151135474,
+      "grad_norm": 0.8010972622074474,
+      "learning_rate": 1.7370102294689563e-05,
+      "loss": 0.959,
+      "step": 499
+    },
+    {
+      "epoch": 1.5638214565387627,
+      "grad_norm": 0.7709293012877854,
+      "learning_rate": 1.7355287817822014e-05,
+      "loss": 0.9864,
+      "step": 500
+    },
+    {
+      "epoch": 1.566953797963978,
+      "grad_norm": 0.799753436392405,
+      "learning_rate": 1.7340438088006376e-05,
+      "loss": 0.9981,
+      "step": 501
+    },
+    {
+      "epoch": 1.5700861393891934,
+      "grad_norm": 0.7779275987751509,
+      "learning_rate": 1.7325553176415496e-05,
+      "loss": 0.9644,
+      "step": 502
+    },
+    {
+      "epoch": 1.5732184808144087,
+      "grad_norm": 0.8008846365371122,
+      "learning_rate": 1.731063315439084e-05,
+      "loss": 0.9368,
+      "step": 503
+    },
+    {
+      "epoch": 1.576350822239624,
+      "grad_norm": 0.774005109549041,
+      "learning_rate": 1.7295678093442155e-05,
+      "loss": 0.9435,
+      "step": 504
+    },
+    {
+      "epoch": 1.5794831636648394,
+      "grad_norm": 0.795159470143028,
+      "learning_rate": 1.728068806524712e-05,
+      "loss": 1.0071,
+      "step": 505
+    },
+    {
+      "epoch": 1.5826155050900548,
+      "grad_norm": 0.7963785520950486,
+      "learning_rate": 1.7265663141651016e-05,
+      "loss": 0.9922,
+      "step": 506
+    },
+    {
+      "epoch": 1.5857478465152701,
+      "grad_norm": 0.7877046535940556,
+      "learning_rate": 1.7250603394666364e-05,
+      "loss": 0.9514,
+      "step": 507
+    },
+    {
+      "epoch": 1.5888801879404855,
+      "grad_norm": 0.7715293903451117,
+      "learning_rate": 1.7235508896472593e-05,
+      "loss": 0.933,
+      "step": 508
+    },
+    {
+      "epoch": 1.5920125293657008,
+      "grad_norm": 0.7934117596728014,
+      "learning_rate": 1.72203797194157e-05,
+      "loss": 0.973,
+      "step": 509
+    },
+    {
+      "epoch": 1.5951448707909162,
+      "grad_norm": 0.8181710440792368,
+      "learning_rate": 1.720521593600787e-05,
+      "loss": 0.986,
+      "step": 510
+    },
+    {
+      "epoch": 1.5982772122161315,
+      "grad_norm": 0.7837650575509727,
+      "learning_rate": 1.7190017618927177e-05,
+      "loss": 0.9612,
+      "step": 511
+    },
+    {
+      "epoch": 1.6014095536413469,
+      "grad_norm": 0.8412641440444018,
+      "learning_rate": 1.7174784841017197e-05,
+      "loss": 0.9745,
+      "step": 512
+    },
+    {
+      "epoch": 1.6045418950665624,
+      "grad_norm": 0.7881270804166439,
+      "learning_rate": 1.7159517675286667e-05,
+      "loss": 0.9644,
+      "step": 513
+    },
+    {
+      "epoch": 1.6076742364917775,
+      "grad_norm": 0.8119820827446707,
+      "learning_rate": 1.7144216194909157e-05,
+      "loss": 0.9886,
+      "step": 514
+    },
+    {
+      "epoch": 1.610806577916993,
+      "grad_norm": 0.7998253925780444,
+      "learning_rate": 1.7128880473222688e-05,
+      "loss": 0.9815,
+      "step": 515
+    },
+    {
+      "epoch": 1.6139389193422082,
+      "grad_norm": 0.7739056976249491,
+      "learning_rate": 1.7113510583729403e-05,
+      "loss": 0.9766,
+      "step": 516
+    },
+    {
+      "epoch": 1.6170712607674238,
+      "grad_norm": 0.8069370636653986,
+      "learning_rate": 1.7098106600095204e-05,
+      "loss": 0.9496,
+      "step": 517
+    },
+    {
+      "epoch": 1.620203602192639,
+      "grad_norm": 0.8037770500974222,
+      "learning_rate": 1.7082668596149402e-05,
+      "loss": 0.9646,
+      "step": 518
+    },
+    {
+      "epoch": 1.6233359436178545,
+      "grad_norm": 0.7878273006407442,
+      "learning_rate": 1.706719664588437e-05,
+      "loss": 0.9569,
+      "step": 519
+    },
+    {
+      "epoch": 1.6264682850430696,
+      "grad_norm": 0.7827154823147211,
+      "learning_rate": 1.7051690823455162e-05,
+      "loss": 0.9405,
+      "step": 520
+    },
+    {
+      "epoch": 1.6296006264682852,
+      "grad_norm": 0.8114142283057644,
+      "learning_rate": 1.70361512031792e-05,
+      "loss": 0.9804,
+      "step": 521
+    },
+    {
+      "epoch": 1.6327329678935003,
+      "grad_norm": 0.7889387459740889,
+      "learning_rate": 1.702057785953588e-05,
+      "loss": 0.9405,
+      "step": 522
+    },
+    {
+      "epoch": 1.6358653093187159,
+      "grad_norm": 0.8311030879338528,
+      "learning_rate": 1.7004970867166238e-05,
+      "loss": 0.9593,
+      "step": 523
+    },
+    {
+      "epoch": 1.638997650743931,
+      "grad_norm": 0.7984284134204775,
+      "learning_rate": 1.6989330300872576e-05,
+      "loss": 0.951,
+      "step": 524
+    },
+    {
+      "epoch": 1.6421299921691466,
+      "grad_norm": 0.7997161389779351,
+      "learning_rate": 1.6973656235618114e-05,
+      "loss": 0.9379,
+      "step": 525
+    },
+    {
+      "epoch": 1.6452623335943617,
+      "grad_norm": 0.7928450659072758,
+      "learning_rate": 1.6957948746526637e-05,
+      "loss": 0.9796,
+      "step": 526
+    },
+    {
+      "epoch": 1.6483946750195773,
+      "grad_norm": 0.7864284356410199,
+      "learning_rate": 1.6942207908882115e-05,
+      "loss": 0.9595,
+      "step": 527
+    },
+    {
+      "epoch": 1.6515270164447924,
+      "grad_norm": 0.7748245207452046,
+      "learning_rate": 1.692643379812836e-05,
+      "loss": 0.9363,
+      "step": 528
+    },
+    {
+      "epoch": 1.654659357870008,
+      "grad_norm": 0.7904035921159812,
+      "learning_rate": 1.691062648986865e-05,
+      "loss": 0.9698,
+      "step": 529
+    },
+    {
+      "epoch": 1.657791699295223,
+      "grad_norm": 0.7699380547476845,
+      "learning_rate": 1.6894786059865383e-05,
+      "loss": 0.9229,
+      "step": 530
+    },
+    {
+      "epoch": 1.6609240407204386,
+      "grad_norm": 0.756885007762418,
+      "learning_rate": 1.6878912584039698e-05,
+      "loss": 0.9664,
+      "step": 531
+    },
+    {
+      "epoch": 1.6640563821456538,
+      "grad_norm": 0.8314070737417165,
+      "learning_rate": 1.686300613847113e-05,
+      "loss": 0.9685,
+      "step": 532
+    },
+    {
+      "epoch": 1.6671887235708693,
+      "grad_norm": 0.8339161427130307,
+      "learning_rate": 1.6847066799397224e-05,
+      "loss": 0.9272,
+      "step": 533
+    },
+    {
+      "epoch": 1.6703210649960845,
+      "grad_norm": 0.8349954422927233,
+      "learning_rate": 1.683109464321317e-05,
+      "loss": 0.9764,
+      "step": 534
+    },
+    {
+      "epoch": 1.6734534064213,
+      "grad_norm": 0.7964469038876641,
+      "learning_rate": 1.6815089746471472e-05,
+      "loss": 1.0066,
+      "step": 535
+    },
+    {
+      "epoch": 1.6765857478465152,
+      "grad_norm": 0.7854281643639673,
+      "learning_rate": 1.6799052185881535e-05,
+      "loss": 0.9901,
+      "step": 536
+    },
+    {
+      "epoch": 1.6797180892717307,
+      "grad_norm": 0.7974377816831979,
+      "learning_rate": 1.678298203830932e-05,
+      "loss": 0.969,
+      "step": 537
+    },
+    {
+      "epoch": 1.6828504306969458,
+      "grad_norm": 0.7734247111191668,
+      "learning_rate": 1.6766879380776983e-05,
+      "loss": 0.9183,
+      "step": 538
+    },
+    {
+      "epoch": 1.6859827721221614,
+      "grad_norm": 0.7724972901824376,
+      "learning_rate": 1.6750744290462484e-05,
+      "loss": 0.9186,
+      "step": 539
+    },
+    {
+      "epoch": 1.6891151135473765,
+      "grad_norm": 0.7968886343195833,
+      "learning_rate": 1.6734576844699234e-05,
+      "loss": 0.9409,
+      "step": 540
+    },
+    {
+      "epoch": 1.692247454972592,
+      "grad_norm": 0.8009596190830679,
+      "learning_rate": 1.671837712097573e-05,
+      "loss": 0.9639,
+      "step": 541
+    },
+    {
+      "epoch": 1.6953797963978072,
+      "grad_norm": 0.8367539665803249,
+      "learning_rate": 1.670214519693515e-05,
+      "loss": 0.9655,
+      "step": 542
+    },
+    {
+      "epoch": 1.6985121378230228,
+      "grad_norm": 0.8071935125435302,
+      "learning_rate": 1.6685881150375027e-05,
+      "loss": 0.9635,
+      "step": 543
+    },
+    {
+      "epoch": 1.701644479248238,
+      "grad_norm": 0.8293405021567984,
+      "learning_rate": 1.6669585059246837e-05,
+      "loss": 0.9681,
+      "step": 544
+    },
+    {
+      "epoch": 1.7047768206734535,
+      "grad_norm": 0.8081858624977237,
+      "learning_rate": 1.6653257001655652e-05,
+      "loss": 0.9918,
+      "step": 545
+    },
+    {
+      "epoch": 1.7079091620986686,
+      "grad_norm": 0.8057738404404446,
+      "learning_rate": 1.6636897055859754e-05,
+      "loss": 0.9727,
+      "step": 546
+    },
+    {
+      "epoch": 1.7110415035238842,
+      "grad_norm": 0.7493081094786209,
+      "learning_rate": 1.6620505300270254e-05,
+      "loss": 0.923,
+      "step": 547
+    },
+    {
+      "epoch": 1.7141738449490993,
+      "grad_norm": 0.7989097981932453,
+      "learning_rate": 1.6604081813450724e-05,
+      "loss": 0.9529,
+      "step": 548
+    },
+    {
+      "epoch": 1.7173061863743149,
+      "grad_norm": 0.7713974329610848,
+      "learning_rate": 1.6587626674116827e-05,
+      "loss": 0.962,
+      "step": 549
+    },
+    {
+      "epoch": 1.72043852779953,
+      "grad_norm": 0.8015438391238155,
+      "learning_rate": 1.657113996113593e-05,
+      "loss": 0.9886,
+      "step": 550
+    },
+    {
+      "epoch": 1.7235708692247456,
+      "grad_norm": 0.8065266518656921,
+      "learning_rate": 1.6554621753526724e-05,
+      "loss": 0.9868,
+      "step": 551
+    },
+    {
+      "epoch": 1.7267032106499607,
+      "grad_norm": 0.8310150242959837,
+      "learning_rate": 1.6538072130458853e-05,
+      "loss": 0.9989,
+      "step": 552
+    },
+    {
+      "epoch": 1.7298355520751763,
+      "grad_norm": 0.7832374226495794,
+      "learning_rate": 1.652149117125253e-05,
+      "loss": 0.9472,
+      "step": 553
+    },
+    {
+      "epoch": 1.7329678935003914,
+      "grad_norm": 0.7885410215851165,
+      "learning_rate": 1.6504878955378166e-05,
+      "loss": 0.9987,
+      "step": 554
+    },
+    {
+      "epoch": 1.736100234925607,
+      "grad_norm": 0.8206660199757921,
+      "learning_rate": 1.6488235562455964e-05,
+      "loss": 0.9761,
+      "step": 555
+    },
+    {
+      "epoch": 1.7392325763508223,
+      "grad_norm": 0.7997233241816074,
+      "learning_rate": 1.6471561072255576e-05,
+      "loss": 0.9768,
+      "step": 556
+    },
+    {
+      "epoch": 1.7423649177760376,
+      "grad_norm": 0.7936840487628978,
+      "learning_rate": 1.6454855564695674e-05,
+      "loss": 0.9434,
+      "step": 557
+    },
+    {
+      "epoch": 1.745497259201253,
+      "grad_norm": 0.7829379543649524,
+      "learning_rate": 1.643811911984362e-05,
+      "loss": 0.951,
+      "step": 558
+    },
+    {
+      "epoch": 1.7486296006264683,
+      "grad_norm": 0.7854616494314223,
+      "learning_rate": 1.6421351817915025e-05,
+      "loss": 0.9417,
+      "step": 559
+    },
+    {
+      "epoch": 1.7517619420516837,
+      "grad_norm": 0.7885222818513928,
+      "learning_rate": 1.6404553739273426e-05,
+      "loss": 0.9678,
+      "step": 560
+    },
+    {
+      "epoch": 1.754894283476899,
+      "grad_norm": 0.7691482528488672,
+      "learning_rate": 1.6387724964429845e-05,
+      "loss": 0.9235,
+      "step": 561
+    },
+    {
+      "epoch": 1.7580266249021144,
+      "grad_norm": 0.7730839050185739,
+      "learning_rate": 1.6370865574042435e-05,
+      "loss": 0.9821,
+      "step": 562
+    },
+    {
+      "epoch": 1.7611589663273297,
+      "grad_norm": 0.7879677074806722,
+      "learning_rate": 1.6353975648916086e-05,
+      "loss": 0.9699,
+      "step": 563
+    },
+    {
+      "epoch": 1.764291307752545,
+      "grad_norm": 0.7858596514917898,
+      "learning_rate": 1.633705527000203e-05,
+      "loss": 0.994,
+      "step": 564
+    },
+    {
+      "epoch": 1.7674236491777604,
+      "grad_norm": 0.8012498378641922,
+      "learning_rate": 1.6320104518397473e-05,
+      "loss": 0.967,
+      "step": 565
+    },
+    {
+      "epoch": 1.7705559906029757,
+      "grad_norm": 0.8188666888354701,
+      "learning_rate": 1.6303123475345182e-05,
+      "loss": 0.9785,
+      "step": 566
+    },
+    {
+      "epoch": 1.773688332028191,
+      "grad_norm": 0.7981343235293162,
+      "learning_rate": 1.6286112222233113e-05,
+      "loss": 0.975,
+      "step": 567
+    },
+    {
+      "epoch": 1.7768206734534064,
+      "grad_norm": 0.7838444823903047,
+      "learning_rate": 1.6269070840594017e-05,
+      "loss": 0.9458,
+      "step": 568
+    },
+    {
+      "epoch": 1.7799530148786218,
+      "grad_norm": 0.7846905745122678,
+      "learning_rate": 1.6251999412105036e-05,
+      "loss": 0.9298,
+      "step": 569
+    },
+    {
+      "epoch": 1.7830853563038371,
+      "grad_norm": 0.8149058395623431,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.9436,
+      "step": 570
+    },
+    {
+      "epoch": 1.7862176977290525,
+      "grad_norm": 0.7941672166648026,
+      "learning_rate": 1.6217766742005693e-05,
+      "loss": 0.9985,
+      "step": 571
+    },
+    {
+      "epoch": 1.7893500391542678,
+      "grad_norm": 0.7710930788046828,
+      "learning_rate": 1.6200605664468113e-05,
+      "loss": 0.97,
+      "step": 572
+    },
+    {
+      "epoch": 1.7924823805794832,
+      "grad_norm": 0.7957133349582347,
+      "learning_rate": 1.6183414868225434e-05,
+      "loss": 0.9959,
+      "step": 573
+    },
+    {
+      "epoch": 1.7956147220046985,
+      "grad_norm": 0.79291459979055,
+      "learning_rate": 1.616619443567092e-05,
+      "loss": 0.9442,
+      "step": 574
+    },
+    {
+      "epoch": 1.7987470634299139,
+      "grad_norm": 0.785169568406983,
+      "learning_rate": 1.6148944449339904e-05,
+      "loss": 0.9639,
+      "step": 575
+    },
+    {
+      "epoch": 1.8018794048551292,
+      "grad_norm": 0.7997066561831516,
+      "learning_rate": 1.6131664991909335e-05,
+      "loss": 1.0127,
+      "step": 576
+    },
+    {
+      "epoch": 1.8050117462803446,
+      "grad_norm": 0.8141399391865273,
+      "learning_rate": 1.6114356146197445e-05,
+      "loss": 0.9386,
+      "step": 577
+    },
+    {
+      "epoch": 1.80814408770556,
+      "grad_norm": 0.8097117540881008,
+      "learning_rate": 1.6097017995163294e-05,
+      "loss": 0.9774,
+      "step": 578
+    },
+    {
+      "epoch": 1.8112764291307752,
+      "grad_norm": 0.7730781490152179,
+      "learning_rate": 1.607965062190641e-05,
+      "loss": 0.9453,
+      "step": 579
+    },
+    {
+      "epoch": 1.8144087705559906,
+      "grad_norm": 0.7838923521873067,
+      "learning_rate": 1.6062254109666383e-05,
+      "loss": 0.9628,
+      "step": 580
+    },
+    {
+      "epoch": 1.817541111981206,
+      "grad_norm": 0.790301862199868,
+      "learning_rate": 1.604482854182245e-05,
+      "loss": 0.9068,
+      "step": 581
+    },
+    {
+      "epoch": 1.8206734534064213,
+      "grad_norm": 0.7750930497911785,
+      "learning_rate": 1.6027374001893123e-05,
+      "loss": 0.9927,
+      "step": 582
+    },
+    {
+      "epoch": 1.8238057948316366,
+      "grad_norm": 0.7779386940305499,
+      "learning_rate": 1.6009890573535763e-05,
+      "loss": 0.919,
+      "step": 583
+    },
+    {
+      "epoch": 1.826938136256852,
+      "grad_norm": 0.7992055706023535,
+      "learning_rate": 1.5992378340546194e-05,
+      "loss": 0.9624,
+      "step": 584
+    },
+    {
+      "epoch": 1.8300704776820673,
+      "grad_norm": 0.8153176494752826,
+      "learning_rate": 1.597483738685829e-05,
+      "loss": 1.0051,
+      "step": 585
+    },
+    {
+      "epoch": 1.8332028191072827,
+      "grad_norm": 0.7936500253751112,
+      "learning_rate": 1.595726779654359e-05,
+      "loss": 0.9692,
+      "step": 586
+    },
+    {
+      "epoch": 1.836335160532498,
+      "grad_norm": 0.8140127238131111,
+      "learning_rate": 1.5939669653810882e-05,
+      "loss": 0.9763,
+      "step": 587
+    },
+    {
+      "epoch": 1.8394675019577134,
+      "grad_norm": 0.8084565243659966,
+      "learning_rate": 1.592204304300579e-05,
+      "loss": 0.9402,
+      "step": 588
+    },
+    {
+      "epoch": 1.8425998433829287,
+      "grad_norm": 0.7527144155069814,
+      "learning_rate": 1.59043880486104e-05,
+      "loss": 0.9255,
+      "step": 589
+    },
+    {
+      "epoch": 1.845732184808144,
+      "grad_norm": 0.7985587310153884,
+      "learning_rate": 1.588670475524283e-05,
+      "loss": 0.9465,
+      "step": 590
+    },
+    {
+      "epoch": 1.8488645262333594,
+      "grad_norm": 0.7524132760020612,
+      "learning_rate": 1.5868993247656822e-05,
+      "loss": 0.9865,
+      "step": 591
+    },
+    {
+      "epoch": 1.8519968676585747,
+      "grad_norm": 0.7823183770981433,
+      "learning_rate": 1.5851253610741357e-05,
+      "loss": 0.8974,
+      "step": 592
+    },
+    {
+      "epoch": 1.85512920908379,
+      "grad_norm": 0.769457387825749,
+      "learning_rate": 1.583348592952023e-05,
+      "loss": 0.9776,
+      "step": 593
+    },
+    {
+      "epoch": 1.8582615505090054,
+      "grad_norm": 0.7714656018935003,
+      "learning_rate": 1.581569028915166e-05,
+      "loss": 0.9399,
+      "step": 594
+    },
+    {
+      "epoch": 1.8613938919342208,
+      "grad_norm": 0.7784139001699675,
+      "learning_rate": 1.5797866774927848e-05,
+      "loss": 0.9699,
+      "step": 595
+    },
+    {
+      "epoch": 1.8645262333594361,
+      "grad_norm": 0.7797773255228968,
+      "learning_rate": 1.5780015472274613e-05,
+      "loss": 0.9404,
+      "step": 596
+    },
+    {
+      "epoch": 1.8676585747846515,
+      "grad_norm": 0.7758341272190327,
+      "learning_rate": 1.5762136466750947e-05,
+      "loss": 0.9719,
+      "step": 597
+    },
+    {
+      "epoch": 1.870790916209867,
+      "grad_norm": 0.7786162669677469,
+      "learning_rate": 1.5744229844048627e-05,
+      "loss": 0.9436,
+      "step": 598
+    },
+    {
+      "epoch": 1.8739232576350822,
+      "grad_norm": 0.7543252031375369,
+      "learning_rate": 1.5726295689991787e-05,
+      "loss": 0.9784,
+      "step": 599
+    },
+    {
+      "epoch": 1.8770555990602977,
+      "grad_norm": 0.7757844187122288,
+      "learning_rate": 1.570833409053653e-05,
+      "loss": 0.9713,
+      "step": 600
+    },
+    {
+      "epoch": 1.8801879404855129,
+      "grad_norm": 0.7653394710213172,
+      "learning_rate": 1.5690345131770474e-05,
+      "loss": 0.8965,
+      "step": 601
+    },
+    {
+      "epoch": 1.8833202819107284,
+      "grad_norm": 0.7896033443700016,
+      "learning_rate": 1.5672328899912397e-05,
+      "loss": 1.0193,
+      "step": 602
+    },
+    {
+      "epoch": 1.8864526233359435,
+      "grad_norm": 0.7846649604330512,
+      "learning_rate": 1.5654285481311777e-05,
+      "loss": 0.9474,
+      "step": 603
+    },
+    {
+      "epoch": 1.8895849647611591,
+      "grad_norm": 0.8135389235047997,
+      "learning_rate": 1.5636214962448396e-05,
+      "loss": 0.9618,
+      "step": 604
+    },
+    {
+      "epoch": 1.8927173061863742,
+      "grad_norm": 0.7824047803609824,
+      "learning_rate": 1.5618117429931923e-05,
+      "loss": 0.9408,
+      "step": 605
+    },
+    {
+      "epoch": 1.8958496476115898,
+      "grad_norm": 0.8254165694685803,
+      "learning_rate": 1.5599992970501512e-05,
+      "loss": 0.9767,
+      "step": 606
+    },
+    {
+      "epoch": 1.898981989036805,
+      "grad_norm": 0.7942279881179067,
+      "learning_rate": 1.5581841671025358e-05,
+      "loss": 0.9657,
+      "step": 607
+    },
+    {
+      "epoch": 1.9021143304620205,
+      "grad_norm": 0.7607717329156005,
+      "learning_rate": 1.55636636185003e-05,
+      "loss": 0.9641,
+      "step": 608
+    },
+    {
+      "epoch": 1.9052466718872356,
+      "grad_norm": 0.7992715938950239,
+      "learning_rate": 1.5545458900051417e-05,
+      "loss": 0.9466,
+      "step": 609
+    },
+    {
+      "epoch": 1.9083790133124512,
+      "grad_norm": 0.7710025331360174,
+      "learning_rate": 1.552722760293157e-05,
+      "loss": 0.9349,
+      "step": 610
+    },
+    {
+      "epoch": 1.9115113547376663,
+      "grad_norm": 0.7995999712279581,
+      "learning_rate": 1.5508969814521026e-05,
+      "loss": 0.9784,
+      "step": 611
+    },
+    {
+      "epoch": 1.9146436961628819,
+      "grad_norm": 0.7972698916237193,
+      "learning_rate": 1.5490685622327015e-05,
+      "loss": 0.9954,
+      "step": 612
+    },
+    {
+      "epoch": 1.917776037588097,
+      "grad_norm": 0.7658223645888151,
+      "learning_rate": 1.547237511398332e-05,
+      "loss": 0.931,
+      "step": 613
+    },
+    {
+      "epoch": 1.9209083790133126,
+      "grad_norm": 0.7891075170093855,
+      "learning_rate": 1.5454038377249845e-05,
+      "loss": 0.9559,
+      "step": 614
+    },
+    {
+      "epoch": 1.9240407204385277,
+      "grad_norm": 0.8134205306266102,
+      "learning_rate": 1.5435675500012212e-05,
+      "loss": 0.9907,
+      "step": 615
+    },
+    {
+      "epoch": 1.9271730618637433,
+      "grad_norm": 0.803815630288296,
+      "learning_rate": 1.5417286570281334e-05,
+      "loss": 0.9834,
+      "step": 616
+    },
+    {
+      "epoch": 1.9303054032889584,
+      "grad_norm": 0.771918493512692,
+      "learning_rate": 1.5398871676192978e-05,
+      "loss": 1.0094,
+      "step": 617
+    },
+    {
+      "epoch": 1.933437744714174,
+      "grad_norm": 0.793008120980106,
+      "learning_rate": 1.538043090600737e-05,
+      "loss": 0.9266,
+      "step": 618
+    },
+    {
+      "epoch": 1.936570086139389,
+      "grad_norm": 0.7778816113808421,
+      "learning_rate": 1.5361964348108736e-05,
+      "loss": 0.9665,
+      "step": 619
+    },
+    {
+      "epoch": 1.9397024275646046,
+      "grad_norm": 0.7989827850535249,
+      "learning_rate": 1.5343472091004925e-05,
+      "loss": 0.9524,
+      "step": 620
+    },
+    {
+      "epoch": 1.9428347689898198,
+      "grad_norm": 0.813278631018117,
+      "learning_rate": 1.532495422332694e-05,
+      "loss": 1.0166,
+      "step": 621
+    },
+    {
+      "epoch": 1.9459671104150353,
+      "grad_norm": 0.7855425028524953,
+      "learning_rate": 1.5306410833828534e-05,
+      "loss": 0.9816,
+      "step": 622
+    },
+    {
+      "epoch": 1.9490994518402505,
+      "grad_norm": 0.7943986605440319,
+      "learning_rate": 1.5287842011385798e-05,
+      "loss": 0.9852,
+      "step": 623
+    },
+    {
+      "epoch": 1.952231793265466,
+      "grad_norm": 0.8316767537841315,
+      "learning_rate": 1.52692478449967e-05,
+      "loss": 0.9811,
+      "step": 624
+    },
+    {
+      "epoch": 1.9553641346906812,
+      "grad_norm": 0.8079701202216203,
+      "learning_rate": 1.5250628423780684e-05,
+      "loss": 0.989,
+      "step": 625
+    },
+    {
+      "epoch": 1.9584964761158967,
+      "grad_norm": 0.8001611635685686,
+      "learning_rate": 1.523198383697825e-05,
+      "loss": 0.9859,
+      "step": 626
+    },
+    {
+      "epoch": 1.9616288175411118,
+      "grad_norm": 0.7922734841320317,
+      "learning_rate": 1.5213314173950498e-05,
+      "loss": 0.9785,
+      "step": 627
+    },
+    {
+      "epoch": 1.9647611589663274,
+      "grad_norm": 0.7833384810882262,
+      "learning_rate": 1.5194619524178718e-05,
+      "loss": 0.9407,
+      "step": 628
+    },
+    {
+      "epoch": 1.9678935003915425,
+      "grad_norm": 0.8098242578519478,
+      "learning_rate": 1.5175899977263963e-05,
+      "loss": 0.9224,
+      "step": 629
+    },
+    {
+      "epoch": 1.971025841816758,
+      "grad_norm": 0.8182996603065131,
+      "learning_rate": 1.515715562292662e-05,
+      "loss": 0.9622,
+      "step": 630
+    },
+    {
+      "epoch": 1.9741581832419732,
+      "grad_norm": 0.7901742990362269,
+      "learning_rate": 1.513838655100596e-05,
+      "loss": 0.9337,
+      "step": 631
+    },
+    {
+      "epoch": 1.9772905246671888,
+      "grad_norm": 0.7615944653532644,
+      "learning_rate": 1.5119592851459726e-05,
+      "loss": 0.9414,
+      "step": 632
+    },
+    {
+      "epoch": 1.980422866092404,
+      "grad_norm": 0.7946500220579569,
+      "learning_rate": 1.5100774614363708e-05,
+      "loss": 0.9512,
+      "step": 633
+    },
+    {
+      "epoch": 1.9835552075176195,
+      "grad_norm": 0.8076792545370913,
+      "learning_rate": 1.5081931929911298e-05,
+      "loss": 0.973,
+      "step": 634
+    },
+    {
+      "epoch": 1.9866875489428346,
+      "grad_norm": 0.7861723773682193,
+      "learning_rate": 1.5063064888413048e-05,
+      "loss": 0.9842,
+      "step": 635
+    },
+    {
+      "epoch": 1.9898198903680502,
+      "grad_norm": 0.7955732049686428,
+      "learning_rate": 1.5044173580296267e-05,
+      "loss": 0.9272,
+      "step": 636
+    },
+    {
+      "epoch": 1.9929522317932653,
+      "grad_norm": 0.8037074259104876,
+      "learning_rate": 1.5025258096104563e-05,
+      "loss": 0.9697,
+      "step": 637
+    },
+    {
+      "epoch": 1.9960845732184809,
+      "grad_norm": 0.8103505085995998,
+      "learning_rate": 1.500631852649742e-05,
+      "loss": 0.948,
+      "step": 638
+    },
+    {
+      "epoch": 1.999216914643696,
+      "grad_norm": 0.7802553935331468,
+      "learning_rate": 1.4987354962249749e-05,
+      "loss": 0.9517,
+      "step": 639
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7802553935331468,
+      "learning_rate": 1.4968367494251486e-05,
+      "loss": 0.9289,
+      "step": 640
+    },
+    {
+      "epoch": 2.0093970242756463,
+      "grad_norm": 1.3210123581817999,
+      "learning_rate": 1.4949356213507113e-05,
+      "loss": 0.7754,
+      "step": 641
+    },
+    {
+      "epoch": 2.0125293657008614,
+      "grad_norm": 1.3238346751721688,
+      "learning_rate": 1.4930321211135258e-05,
+      "loss": 0.7855,
+      "step": 642
+    },
+    {
+      "epoch": 2.015661707126077,
+      "grad_norm": 1.1616452207714454,
+      "learning_rate": 1.4911262578368233e-05,
+      "loss": 0.7959,
+      "step": 643
+    },
+    {
+      "epoch": 2.018794048551292,
+      "grad_norm": 0.9412974130528684,
+      "learning_rate": 1.4892180406551618e-05,
+      "loss": 0.7173,
+      "step": 644
+    },
+    {
+      "epoch": 2.0219263899765076,
+      "grad_norm": 1.3175139050237945,
+      "learning_rate": 1.48730747871438e-05,
+      "loss": 0.7316,
+      "step": 645
+    },
+    {
+      "epoch": 2.0250587314017228,
+      "grad_norm": 1.5954314046677502,
+      "learning_rate": 1.4853945811715563e-05,
+      "loss": 0.7542,
+      "step": 646
+    },
+    {
+      "epoch": 2.0281910728269383,
+      "grad_norm": 1.37008550444395,
+      "learning_rate": 1.4834793571949617e-05,
+      "loss": 0.7587,
+      "step": 647
+    },
+    {
+      "epoch": 2.0313234142521535,
+      "grad_norm": 1.169303068811387,
+      "learning_rate": 1.4815618159640188e-05,
+      "loss": 0.7195,
+      "step": 648
+    },
+    {
+      "epoch": 2.034455755677369,
+      "grad_norm": 1.120757117831837,
+      "learning_rate": 1.4796419666692555e-05,
+      "loss": 0.6884,
+      "step": 649
+    },
+    {
+      "epoch": 2.037588097102584,
+      "grad_norm": 1.1374034296234556,
+      "learning_rate": 1.477719818512263e-05,
+      "loss": 0.7608,
+      "step": 650
+    },
+    {
+      "epoch": 2.0407204385277997,
+      "grad_norm": 1.018857278719166,
+      "learning_rate": 1.47579538070565e-05,
+      "loss": 0.7136,
+      "step": 651
+    },
+    {
+      "epoch": 2.043852779953015,
+      "grad_norm": 0.991608010520256,
+      "learning_rate": 1.4738686624729987e-05,
+      "loss": 0.7105,
+      "step": 652
+    },
+    {
+      "epoch": 2.0469851213782304,
+      "grad_norm": 1.0388281087329523,
+      "learning_rate": 1.4719396730488225e-05,
+      "loss": 0.7348,
+      "step": 653
+    },
+    {
+      "epoch": 2.0501174628034455,
+      "grad_norm": 1.0171773999697262,
+      "learning_rate": 1.4700084216785189e-05,
+      "loss": 0.7212,
+      "step": 654
+    },
+    {
+      "epoch": 2.053249804228661,
+      "grad_norm": 0.9712484652548604,
+      "learning_rate": 1.4680749176183275e-05,
+      "loss": 0.7152,
+      "step": 655
+    },
+    {
+      "epoch": 2.0563821456538762,
+      "grad_norm": 0.9445446614613634,
+      "learning_rate": 1.4661391701352844e-05,
+      "loss": 0.7144,
+      "step": 656
+    },
+    {
+      "epoch": 2.059514487079092,
+      "grad_norm": 0.9945223666447938,
+      "learning_rate": 1.464201188507178e-05,
+      "loss": 0.7119,
+      "step": 657
+    },
+    {
+      "epoch": 2.062646828504307,
+      "grad_norm": 0.9677117020420452,
+      "learning_rate": 1.4622609820225052e-05,
+      "loss": 0.7148,
+      "step": 658
+    },
+    {
+      "epoch": 2.0657791699295225,
+      "grad_norm": 1.0137018942940528,
+      "learning_rate": 1.4603185599804259e-05,
+      "loss": 0.7435,
+      "step": 659
+    },
+    {
+      "epoch": 2.0689115113547376,
+      "grad_norm": 1.02511003538228,
+      "learning_rate": 1.4583739316907188e-05,
+      "loss": 0.6999,
+      "step": 660
+    },
+    {
+      "epoch": 2.072043852779953,
+      "grad_norm": 0.9904700091988614,
+      "learning_rate": 1.4564271064737376e-05,
+      "loss": 0.7449,
+      "step": 661
+    },
+    {
+      "epoch": 2.0751761942051683,
+      "grad_norm": 0.9279680631763316,
+      "learning_rate": 1.4544780936603644e-05,
+      "loss": 0.7023,
+      "step": 662
+    },
+    {
+      "epoch": 2.078308535630384,
+      "grad_norm": 0.9287114111608203,
+      "learning_rate": 1.4525269025919677e-05,
+      "loss": 0.7163,
+      "step": 663
+    },
+    {
+      "epoch": 2.081440877055599,
+      "grad_norm": 0.943750195528307,
+      "learning_rate": 1.4505735426203545e-05,
+      "loss": 0.7377,
+      "step": 664
+    },
+    {
+      "epoch": 2.0845732184808146,
+      "grad_norm": 0.9053784915303666,
+      "learning_rate": 1.4486180231077278e-05,
+      "loss": 0.7282,
+      "step": 665
+    },
+    {
+      "epoch": 2.0877055599060297,
+      "grad_norm": 0.955797509257669,
+      "learning_rate": 1.446660353426641e-05,
+      "loss": 0.7668,
+      "step": 666
+    },
+    {
+      "epoch": 2.0908379013312453,
+      "grad_norm": 0.9130054155780589,
+      "learning_rate": 1.4447005429599537e-05,
+      "loss": 0.695,
+      "step": 667
+    },
+    {
+      "epoch": 2.0939702427564604,
+      "grad_norm": 0.9516666064010463,
+      "learning_rate": 1.4427386011007847e-05,
+      "loss": 0.7255,
+      "step": 668
+    },
+    {
+      "epoch": 2.097102584181676,
+      "grad_norm": 0.9546422449504612,
+      "learning_rate": 1.4407745372524686e-05,
+      "loss": 0.7224,
+      "step": 669
+    },
+    {
+      "epoch": 2.100234925606891,
+      "grad_norm": 0.9448475040935651,
+      "learning_rate": 1.4388083608285112e-05,
+      "loss": 0.7222,
+      "step": 670
+    },
+    {
+      "epoch": 2.1033672670321066,
+      "grad_norm": 0.9222655058113182,
+      "learning_rate": 1.4368400812525434e-05,
+      "loss": 0.7251,
+      "step": 671
+    },
+    {
+      "epoch": 2.1064996084573218,
+      "grad_norm": 0.9777941043274174,
+      "learning_rate": 1.4348697079582752e-05,
+      "loss": 0.7703,
+      "step": 672
+    },
+    {
+      "epoch": 2.1096319498825373,
+      "grad_norm": 0.9749998119435571,
+      "learning_rate": 1.4328972503894525e-05,
+      "loss": 0.7574,
+      "step": 673
+    },
+    {
+      "epoch": 2.1127642913077525,
+      "grad_norm": 0.9756161103337109,
+      "learning_rate": 1.4309227179998109e-05,
+      "loss": 0.7646,
+      "step": 674
+    },
+    {
+      "epoch": 2.115896632732968,
+      "grad_norm": 0.9561100629580807,
+      "learning_rate": 1.4289461202530297e-05,
+      "loss": 0.7504,
+      "step": 675
+    },
+    {
+      "epoch": 2.119028974158183,
+      "grad_norm": 0.9458044599682323,
+      "learning_rate": 1.4269674666226879e-05,
+      "loss": 0.7406,
+      "step": 676
+    },
+    {
+      "epoch": 2.1221613155833987,
+      "grad_norm": 0.9193297731491735,
+      "learning_rate": 1.4249867665922175e-05,
+      "loss": 0.6953,
+      "step": 677
+    },
+    {
+      "epoch": 2.125293657008614,
+      "grad_norm": 0.938231383030172,
+      "learning_rate": 1.4230040296548588e-05,
+      "loss": 0.7742,
+      "step": 678
+    },
+    {
+      "epoch": 2.1284259984338294,
+      "grad_norm": 0.9608218431580744,
+      "learning_rate": 1.4210192653136151e-05,
+      "loss": 0.745,
+      "step": 679
+    },
+    {
+      "epoch": 2.1315583398590445,
+      "grad_norm": 0.9131562406248528,
+      "learning_rate": 1.4190324830812067e-05,
+      "loss": 0.706,
+      "step": 680
+    },
+    {
+      "epoch": 2.13469068128426,
+      "grad_norm": 0.9330499851170364,
+      "learning_rate": 1.4170436924800251e-05,
+      "loss": 0.7198,
+      "step": 681
+    },
+    {
+      "epoch": 2.137823022709475,
+      "grad_norm": 0.9744409362566172,
+      "learning_rate": 1.415052903042087e-05,
+      "loss": 0.748,
+      "step": 682
+    },
+    {
+      "epoch": 2.140955364134691,
+      "grad_norm": 0.956855447487281,
+      "learning_rate": 1.4130601243089905e-05,
+      "loss": 0.772,
+      "step": 683
+    },
+    {
+      "epoch": 2.144087705559906,
+      "grad_norm": 0.9105854308119294,
+      "learning_rate": 1.4110653658318682e-05,
+      "loss": 0.7066,
+      "step": 684
+    },
+    {
+      "epoch": 2.1472200469851215,
+      "grad_norm": 0.9437505928196139,
+      "learning_rate": 1.4090686371713403e-05,
+      "loss": 0.7511,
+      "step": 685
+    },
+    {
+      "epoch": 2.1503523884103366,
+      "grad_norm": 0.9754488006918184,
+      "learning_rate": 1.4070699478974698e-05,
+      "loss": 0.7273,
+      "step": 686
+    },
+    {
+      "epoch": 2.153484729835552,
+      "grad_norm": 0.9660297851499952,
+      "learning_rate": 1.4050693075897181e-05,
+      "loss": 0.7423,
+      "step": 687
+    },
+    {
+      "epoch": 2.1566170712607673,
+      "grad_norm": 0.9710350268049464,
+      "learning_rate": 1.4030667258368968e-05,
+      "loss": 0.7325,
+      "step": 688
+    },
+    {
+      "epoch": 2.159749412685983,
+      "grad_norm": 0.9168340841543151,
+      "learning_rate": 1.4010622122371219e-05,
+      "loss": 0.7178,
+      "step": 689
+    },
+    {
+      "epoch": 2.162881754111198,
+      "grad_norm": 0.967756101002905,
+      "learning_rate": 1.3990557763977694e-05,
+      "loss": 0.7614,
+      "step": 690
+    },
+    {
+      "epoch": 2.1660140955364136,
+      "grad_norm": 0.9340155531023159,
+      "learning_rate": 1.3970474279354283e-05,
+      "loss": 0.7106,
+      "step": 691
+    },
+    {
+      "epoch": 2.1691464369616287,
+      "grad_norm": 0.9505636046987749,
+      "learning_rate": 1.3950371764758543e-05,
+      "loss": 0.7243,
+      "step": 692
+    },
+    {
+      "epoch": 2.1722787783868442,
+      "grad_norm": 0.9197842895474289,
+      "learning_rate": 1.3930250316539237e-05,
+      "loss": 0.7501,
+      "step": 693
+    },
+    {
+      "epoch": 2.1754111198120594,
+      "grad_norm": 0.9424997582167158,
+      "learning_rate": 1.3910110031135884e-05,
+      "loss": 0.7472,
+      "step": 694
+    },
+    {
+      "epoch": 2.178543461237275,
+      "grad_norm": 0.9606400745011864,
+      "learning_rate": 1.3889951005078271e-05,
+      "loss": 0.7475,
+      "step": 695
+    },
+    {
+      "epoch": 2.18167580266249,
+      "grad_norm": 0.9659045083450463,
+      "learning_rate": 1.3869773334986023e-05,
+      "loss": 0.7488,
+      "step": 696
+    },
+    {
+      "epoch": 2.1848081440877056,
+      "grad_norm": 0.984473222173499,
+      "learning_rate": 1.3849577117568115e-05,
+      "loss": 0.7502,
+      "step": 697
+    },
+    {
+      "epoch": 2.1879404855129208,
+      "grad_norm": 0.9738775239386929,
+      "learning_rate": 1.3829362449622416e-05,
+      "loss": 0.7482,
+      "step": 698
+    },
+    {
+      "epoch": 2.1910728269381363,
+      "grad_norm": 0.9315292619814982,
+      "learning_rate": 1.3809129428035229e-05,
+      "loss": 0.7511,
+      "step": 699
+    },
+    {
+      "epoch": 2.1942051683633514,
+      "grad_norm": 0.9347672091944302,
+      "learning_rate": 1.3788878149780827e-05,
+      "loss": 0.7436,
+      "step": 700
+    },
+    {
+      "epoch": 2.197337509788567,
+      "grad_norm": 0.9359481225994939,
+      "learning_rate": 1.376860871192098e-05,
+      "loss": 0.7509,
+      "step": 701
+    },
+    {
+      "epoch": 2.200469851213782,
+      "grad_norm": 0.9219014695571008,
+      "learning_rate": 1.3748321211604488e-05,
+      "loss": 0.7389,
+      "step": 702
+    },
+    {
+      "epoch": 2.2036021926389977,
+      "grad_norm": 0.9880848258280983,
+      "learning_rate": 1.3728015746066736e-05,
+      "loss": 0.7475,
+      "step": 703
+    },
+    {
+      "epoch": 2.206734534064213,
+      "grad_norm": 0.9408764680190602,
+      "learning_rate": 1.370769241262921e-05,
+      "loss": 0.7402,
+      "step": 704
+    },
+    {
+      "epoch": 2.2098668754894284,
+      "grad_norm": 0.9301989809594925,
+      "learning_rate": 1.3687351308699026e-05,
+      "loss": 0.716,
+      "step": 705
+    },
+    {
+      "epoch": 2.2129992169146435,
+      "grad_norm": 0.9179501547426313,
+      "learning_rate": 1.3666992531768482e-05,
+      "loss": 0.7059,
+      "step": 706
+    },
+    {
+      "epoch": 2.216131558339859,
+      "grad_norm": 0.9666221682985012,
+      "learning_rate": 1.3646616179414573e-05,
+      "loss": 0.7599,
+      "step": 707
+    },
+    {
+      "epoch": 2.219263899765074,
+      "grad_norm": 0.9675432440063244,
+      "learning_rate": 1.3626222349298539e-05,
+      "loss": 0.7518,
+      "step": 708
+    },
+    {
+      "epoch": 2.22239624119029,
+      "grad_norm": 0.9390656597784647,
+      "learning_rate": 1.360581113916538e-05,
+      "loss": 0.7252,
+      "step": 709
+    },
+    {
+      "epoch": 2.225528582615505,
+      "grad_norm": 0.9581962913041326,
+      "learning_rate": 1.3585382646843396e-05,
+      "loss": 0.7605,
+      "step": 710
+    },
+    {
+      "epoch": 2.2286609240407205,
+      "grad_norm": 0.9610602302318673,
+      "learning_rate": 1.356493697024373e-05,
+      "loss": 0.7811,
+      "step": 711
+    },
+    {
+      "epoch": 2.2317932654659356,
+      "grad_norm": 0.9867556442165821,
+      "learning_rate": 1.3544474207359875e-05,
+      "loss": 0.7349,
+      "step": 712
+    },
+    {
+      "epoch": 2.234925606891151,
+      "grad_norm": 0.9425764813872476,
+      "learning_rate": 1.352399445626722e-05,
+      "loss": 0.7226,
+      "step": 713
+    },
+    {
+      "epoch": 2.2380579483163663,
+      "grad_norm": 1.0102139516905289,
+      "learning_rate": 1.3503497815122577e-05,
+      "loss": 0.7671,
+      "step": 714
+    },
+    {
+      "epoch": 2.241190289741582,
+      "grad_norm": 0.965501835494664,
+      "learning_rate": 1.3482984382163713e-05,
+      "loss": 0.7449,
+      "step": 715
+    },
+    {
+      "epoch": 2.244322631166797,
+      "grad_norm": 0.9131072345452055,
+      "learning_rate": 1.3462454255708868e-05,
+      "loss": 0.714,
+      "step": 716
+    },
+    {
+      "epoch": 2.2474549725920125,
+      "grad_norm": 0.9498450701506534,
+      "learning_rate": 1.3441907534156298e-05,
+      "loss": 0.7484,
+      "step": 717
+    },
+    {
+      "epoch": 2.2505873140172277,
+      "grad_norm": 0.9689310041592919,
+      "learning_rate": 1.34213443159838e-05,
+      "loss": 0.7772,
+      "step": 718
+    },
+    {
+      "epoch": 2.2537196554424432,
+      "grad_norm": 0.9414801485286473,
+      "learning_rate": 1.3400764699748223e-05,
+      "loss": 0.7307,
+      "step": 719
+    },
+    {
+      "epoch": 2.256851996867659,
+      "grad_norm": 0.918278021033119,
+      "learning_rate": 1.3380168784085028e-05,
+      "loss": 0.7179,
+      "step": 720
+    },
+    {
+      "epoch": 2.259984338292874,
+      "grad_norm": 0.9166637511796559,
+      "learning_rate": 1.3359556667707786e-05,
+      "loss": 0.7142,
+      "step": 721
+    },
+    {
+      "epoch": 2.263116679718089,
+      "grad_norm": 0.9495008728633978,
+      "learning_rate": 1.3338928449407721e-05,
+      "loss": 0.7138,
+      "step": 722
+    },
+    {
+      "epoch": 2.2662490211433046,
+      "grad_norm": 0.9787409627000769,
+      "learning_rate": 1.3318284228053225e-05,
+      "loss": 0.7126,
+      "step": 723
+    },
+    {
+      "epoch": 2.26938136256852,
+      "grad_norm": 0.9627840952028726,
+      "learning_rate": 1.3297624102589395e-05,
+      "loss": 0.7219,
+      "step": 724
+    },
+    {
+      "epoch": 2.2725137039937353,
+      "grad_norm": 0.9476059282284458,
+      "learning_rate": 1.3276948172037556e-05,
+      "loss": 0.7392,
+      "step": 725
+    },
+    {
+      "epoch": 2.2756460454189504,
+      "grad_norm": 0.9906364986347657,
+      "learning_rate": 1.3256256535494783e-05,
+      "loss": 0.7349,
+      "step": 726
+    },
+    {
+      "epoch": 2.278778386844166,
+      "grad_norm": 0.9675118908082809,
+      "learning_rate": 1.3235549292133425e-05,
+      "loss": 0.7364,
+      "step": 727
+    },
+    {
+      "epoch": 2.2819107282693816,
+      "grad_norm": 0.9429167511984878,
+      "learning_rate": 1.3214826541200638e-05,
+      "loss": 0.7514,
+      "step": 728
+    },
+    {
+      "epoch": 2.2850430696945967,
+      "grad_norm": 0.9847232006632042,
+      "learning_rate": 1.3194088382017897e-05,
+      "loss": 0.7606,
+      "step": 729
+    },
+    {
+      "epoch": 2.288175411119812,
+      "grad_norm": 0.9563800977188294,
+      "learning_rate": 1.3173334913980535e-05,
+      "loss": 0.7388,
+      "step": 730
+    },
+    {
+      "epoch": 2.2913077525450274,
+      "grad_norm": 0.9735616571758612,
+      "learning_rate": 1.3152566236557251e-05,
+      "loss": 0.7312,
+      "step": 731
+    },
+    {
+      "epoch": 2.294440093970243,
+      "grad_norm": 0.9639178348678931,
+      "learning_rate": 1.3131782449289646e-05,
+      "loss": 0.7385,
+      "step": 732
+    },
+    {
+      "epoch": 2.297572435395458,
+      "grad_norm": 0.9529934639226041,
+      "learning_rate": 1.3110983651791731e-05,
+      "loss": 0.7194,
+      "step": 733
+    },
+    {
+      "epoch": 2.3007047768206736,
+      "grad_norm": 0.9678440459923399,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.7589,
+      "step": 734
+    },
+    {
+      "epoch": 2.3038371182458888,
+      "grad_norm": 0.9540843435038134,
+      "learning_rate": 1.3069341424920301e-05,
+      "loss": 0.7811,
+      "step": 735
+    },
+    {
+      "epoch": 2.3069694596711043,
+      "grad_norm": 1.0181162510512147,
+      "learning_rate": 1.3048498195132613e-05,
+      "loss": 0.7492,
+      "step": 736
+    },
+    {
+      "epoch": 2.3101018010963195,
+      "grad_norm": 0.9572210519063549,
+      "learning_rate": 1.3027640354285333e-05,
+      "loss": 0.788,
+      "step": 737
+    },
+    {
+      "epoch": 2.313234142521535,
+      "grad_norm": 0.9538184834756422,
+      "learning_rate": 1.3006768002347414e-05,
+      "loss": 0.7343,
+      "step": 738
+    },
+    {
+      "epoch": 2.31636648394675,
+      "grad_norm": 0.9765399045822022,
+      "learning_rate": 1.298588123935735e-05,
+      "loss": 0.7252,
+      "step": 739
+    },
+    {
+      "epoch": 2.3194988253719657,
+      "grad_norm": 0.9618045560134357,
+      "learning_rate": 1.2964980165422701e-05,
+      "loss": 0.7373,
+      "step": 740
+    },
+    {
+      "epoch": 2.322631166797181,
+      "grad_norm": 0.9565307835074357,
+      "learning_rate": 1.2944064880719634e-05,
+      "loss": 0.7335,
+      "step": 741
+    },
+    {
+      "epoch": 2.3257635082223964,
+      "grad_norm": 0.967282888806741,
+      "learning_rate": 1.2923135485492413e-05,
+      "loss": 0.7845,
+      "step": 742
+    },
+    {
+      "epoch": 2.3288958496476115,
+      "grad_norm": 0.9686861328866166,
+      "learning_rate": 1.290219208005294e-05,
+      "loss": 0.7312,
+      "step": 743
+    },
+    {
+      "epoch": 2.332028191072827,
+      "grad_norm": 0.9726533064758265,
+      "learning_rate": 1.2881234764780257e-05,
+      "loss": 0.7484,
+      "step": 744
+    },
+    {
+      "epoch": 2.3351605324980422,
+      "grad_norm": 0.957502804625888,
+      "learning_rate": 1.2860263640120086e-05,
+      "loss": 0.7295,
+      "step": 745
+    },
+    {
+      "epoch": 2.338292873923258,
+      "grad_norm": 1.0246439398068528,
+      "learning_rate": 1.2839278806584323e-05,
+      "loss": 0.7583,
+      "step": 746
+    },
+    {
+      "epoch": 2.341425215348473,
+      "grad_norm": 0.9881765469071092,
+      "learning_rate": 1.2818280364750577e-05,
+      "loss": 0.7605,
+      "step": 747
+    },
+    {
+      "epoch": 2.3445575567736885,
+      "grad_norm": 0.9226451453922399,
+      "learning_rate": 1.2797268415261681e-05,
+      "loss": 0.7449,
+      "step": 748
+    },
+    {
+      "epoch": 2.3476898981989036,
+      "grad_norm": 0.9389466675244833,
+      "learning_rate": 1.2776243058825199e-05,
+      "loss": 0.7508,
+      "step": 749
+    },
+    {
+      "epoch": 2.350822239624119,
+      "grad_norm": 0.9680649310436178,
+      "learning_rate": 1.2755204396212965e-05,
+      "loss": 0.7514,
+      "step": 750
+    },
+    {
+      "epoch": 2.3539545810493343,
+      "grad_norm": 0.9508971737498871,
+      "learning_rate": 1.273415252826058e-05,
+      "loss": 0.7163,
+      "step": 751
+    },
+    {
+      "epoch": 2.35708692247455,
+      "grad_norm": 1.0037413765304937,
+      "learning_rate": 1.2713087555866937e-05,
+      "loss": 0.7523,
+      "step": 752
+    },
+    {
+      "epoch": 2.360219263899765,
+      "grad_norm": 0.9675958858875301,
+      "learning_rate": 1.2692009579993736e-05,
+      "loss": 0.752,
+      "step": 753
+    },
+    {
+      "epoch": 2.3633516053249806,
+      "grad_norm": 0.9426804383890924,
+      "learning_rate": 1.2670918701665007e-05,
+      "loss": 0.6878,
+      "step": 754
+    },
+    {
+      "epoch": 2.3664839467501957,
+      "grad_norm": 0.9505901023922874,
+      "learning_rate": 1.264981502196662e-05,
+      "loss": 0.73,
+      "step": 755
+    },
+    {
+      "epoch": 2.3696162881754113,
+      "grad_norm": 0.982586546277607,
+      "learning_rate": 1.2628698642045786e-05,
+      "loss": 0.7187,
+      "step": 756
+    },
+    {
+      "epoch": 2.3727486296006264,
+      "grad_norm": 1.0016723707151867,
+      "learning_rate": 1.2607569663110603e-05,
+      "loss": 0.7703,
+      "step": 757
+    },
+    {
+      "epoch": 2.375880971025842,
+      "grad_norm": 0.9650854481892198,
+      "learning_rate": 1.258642818642955e-05,
+      "loss": 0.722,
+      "step": 758
+    },
+    {
+      "epoch": 2.379013312451057,
+      "grad_norm": 0.9838012057082094,
+      "learning_rate": 1.2565274313331001e-05,
+      "loss": 0.7776,
+      "step": 759
+    },
+    {
+      "epoch": 2.3821456538762726,
+      "grad_norm": 0.9546713861914818,
+      "learning_rate": 1.2544108145202748e-05,
+      "loss": 0.7277,
+      "step": 760
+    },
+    {
+      "epoch": 2.3852779953014878,
+      "grad_norm": 0.9520183658938295,
+      "learning_rate": 1.2522929783491508e-05,
+      "loss": 0.7334,
+      "step": 761
+    },
+    {
+      "epoch": 2.3884103367267033,
+      "grad_norm": 0.9540342124185963,
+      "learning_rate": 1.2501739329702453e-05,
+      "loss": 0.7291,
+      "step": 762
+    },
+    {
+      "epoch": 2.3915426781519185,
+      "grad_norm": 0.9549925390587612,
+      "learning_rate": 1.2480536885398697e-05,
+      "loss": 0.7329,
+      "step": 763
+    },
+    {
+      "epoch": 2.394675019577134,
+      "grad_norm": 0.9728937913697199,
+      "learning_rate": 1.2459322552200824e-05,
+      "loss": 0.7802,
+      "step": 764
+    },
+    {
+      "epoch": 2.397807361002349,
+      "grad_norm": 0.9640580751608233,
+      "learning_rate": 1.2438096431786408e-05,
+      "loss": 0.7606,
+      "step": 765
+    },
+    {
+      "epoch": 2.4009397024275647,
+      "grad_norm": 0.9304598478726142,
+      "learning_rate": 1.2416858625889507e-05,
+      "loss": 0.7361,
+      "step": 766
+    },
+    {
+      "epoch": 2.40407204385278,
+      "grad_norm": 0.9806923344541586,
+      "learning_rate": 1.2395609236300198e-05,
+      "loss": 0.7155,
+      "step": 767
+    },
+    {
+      "epoch": 2.4072043852779954,
+      "grad_norm": 0.969342428109374,
+      "learning_rate": 1.2374348364864068e-05,
+      "loss": 0.7559,
+      "step": 768
+    },
+    {
+      "epoch": 2.4103367267032105,
+      "grad_norm": 0.9350655072927523,
+      "learning_rate": 1.2353076113481742e-05,
+      "loss": 0.7544,
+      "step": 769
+    },
+    {
+      "epoch": 2.413469068128426,
+      "grad_norm": 0.9447261809188683,
+      "learning_rate": 1.2331792584108375e-05,
+      "loss": 0.7191,
+      "step": 770
+    },
+    {
+      "epoch": 2.4166014095536412,
+      "grad_norm": 0.9674909744485723,
+      "learning_rate": 1.2310497878753196e-05,
+      "loss": 0.7686,
+      "step": 771
+    },
+    {
+      "epoch": 2.419733750978857,
+      "grad_norm": 0.9854774668922955,
+      "learning_rate": 1.2289192099478979e-05,
+      "loss": 0.7595,
+      "step": 772
+    },
+    {
+      "epoch": 2.422866092404072,
+      "grad_norm": 0.9841981999228102,
+      "learning_rate": 1.2267875348401585e-05,
+      "loss": 0.7653,
+      "step": 773
+    },
+    {
+      "epoch": 2.4259984338292875,
+      "grad_norm": 0.9641773998632913,
+      "learning_rate": 1.224654772768946e-05,
+      "loss": 0.7338,
+      "step": 774
+    },
+    {
+      "epoch": 2.4291307752545026,
+      "grad_norm": 0.947511728550051,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.7583,
+      "step": 775
+    },
+    {
+      "epoch": 2.432263116679718,
+      "grad_norm": 0.9870470819669465,
+      "learning_rate": 1.220386028629479e-05,
+      "loss": 0.781,
+      "step": 776
+    },
+    {
+      "epoch": 2.4353954581049333,
+      "grad_norm": 0.9540467312167563,
+      "learning_rate": 1.218250067020766e-05,
+      "loss": 0.732,
+      "step": 777
+    },
+    {
+      "epoch": 2.438527799530149,
+      "grad_norm": 0.9742439189723893,
+      "learning_rate": 1.2161130593675643e-05,
+      "loss": 0.7654,
+      "step": 778
+    },
+    {
+      "epoch": 2.441660140955364,
+      "grad_norm": 0.947883079443643,
+      "learning_rate": 1.2139750159122771e-05,
+      "loss": 0.7291,
+      "step": 779
+    },
+    {
+      "epoch": 2.4447924823805796,
+      "grad_norm": 1.0108972783151455,
+      "learning_rate": 1.2118359469022714e-05,
+      "loss": 0.7343,
+      "step": 780
+    },
+    {
+      "epoch": 2.4479248238057947,
+      "grad_norm": 0.9909962900496487,
+      "learning_rate": 1.2096958625898294e-05,
+      "loss": 0.75,
+      "step": 781
+    },
+    {
+      "epoch": 2.4510571652310102,
+      "grad_norm": 0.979708350323752,
+      "learning_rate": 1.2075547732321005e-05,
+      "loss": 0.7502,
+      "step": 782
+    },
+    {
+      "epoch": 2.4541895066562254,
+      "grad_norm": 0.9605132744547762,
+      "learning_rate": 1.2054126890910499e-05,
+      "loss": 0.7519,
+      "step": 783
+    },
+    {
+      "epoch": 2.457321848081441,
+      "grad_norm": 0.9692514022395361,
+      "learning_rate": 1.2032696204334115e-05,
+      "loss": 0.7273,
+      "step": 784
+    },
+    {
+      "epoch": 2.460454189506656,
+      "grad_norm": 0.9449589861483914,
+      "learning_rate": 1.2011255775306378e-05,
+      "loss": 0.7511,
+      "step": 785
+    },
+    {
+      "epoch": 2.4635865309318716,
+      "grad_norm": 0.9471704960405938,
+      "learning_rate": 1.1989805706588502e-05,
+      "loss": 0.7347,
+      "step": 786
+    },
+    {
+      "epoch": 2.4667188723570868,
+      "grad_norm": 0.9320311390343631,
+      "learning_rate": 1.1968346100987906e-05,
+      "loss": 0.7082,
+      "step": 787
+    },
+    {
+      "epoch": 2.4698512137823023,
+      "grad_norm": 0.9162015711675013,
+      "learning_rate": 1.1946877061357723e-05,
+      "loss": 0.7287,
+      "step": 788
+    },
+    {
+      "epoch": 2.4729835552075174,
+      "grad_norm": 0.9676114383404737,
+      "learning_rate": 1.1925398690596293e-05,
+      "loss": 0.753,
+      "step": 789
+    },
+    {
+      "epoch": 2.476115896632733,
+      "grad_norm": 0.9877343226967886,
+      "learning_rate": 1.1903911091646684e-05,
+      "loss": 0.7758,
+      "step": 790
+    },
+    {
+      "epoch": 2.479248238057948,
+      "grad_norm": 1.023377404363016,
+      "learning_rate": 1.1882414367496196e-05,
+      "loss": 0.7392,
+      "step": 791
+    },
+    {
+      "epoch": 2.4823805794831637,
+      "grad_norm": 0.9945895799174997,
+      "learning_rate": 1.186090862117586e-05,
+      "loss": 0.7502,
+      "step": 792
+    },
+    {
+      "epoch": 2.485512920908379,
+      "grad_norm": 1.0038795381065821,
+      "learning_rate": 1.183939395575995e-05,
+      "loss": 0.806,
+      "step": 793
+    },
+    {
+      "epoch": 2.4886452623335944,
+      "grad_norm": 0.9926041291737507,
+      "learning_rate": 1.181787047436549e-05,
+      "loss": 0.7617,
+      "step": 794
+    },
+    {
+      "epoch": 2.4917776037588095,
+      "grad_norm": 0.9747916863077687,
+      "learning_rate": 1.1796338280151756e-05,
+      "loss": 0.7506,
+      "step": 795
+    },
+    {
+      "epoch": 2.494909945184025,
+      "grad_norm": 0.928076659108732,
+      "learning_rate": 1.1774797476319785e-05,
+      "loss": 0.7306,
+      "step": 796
+    },
+    {
+      "epoch": 2.49804228660924,
+      "grad_norm": 1.0261524870519174,
+      "learning_rate": 1.175324816611188e-05,
+      "loss": 0.7463,
+      "step": 797
+    },
+    {
+      "epoch": 2.501174628034456,
+      "grad_norm": 0.9847678627771624,
+      "learning_rate": 1.1731690452811107e-05,
+      "loss": 0.7281,
+      "step": 798
+    },
+    {
+      "epoch": 2.5043069694596714,
+      "grad_norm": 0.9643463190600754,
+      "learning_rate": 1.1710124439740816e-05,
+      "loss": 0.76,
+      "step": 799
+    },
+    {
+      "epoch": 2.5074393108848865,
+      "grad_norm": 0.969773920531062,
+      "learning_rate": 1.1688550230264128e-05,
+      "loss": 0.7658,
+      "step": 800
+    },
+    {
+      "epoch": 2.5105716523101016,
+      "grad_norm": 0.9626642855194483,
+      "learning_rate": 1.1666967927783461e-05,
+      "loss": 0.7478,
+      "step": 801
+    },
+    {
+      "epoch": 2.513703993735317,
+      "grad_norm": 0.9884939163865005,
+      "learning_rate": 1.164537763574001e-05,
+      "loss": 0.7153,
+      "step": 802
+    },
+    {
+      "epoch": 2.5168363351605327,
+      "grad_norm": 0.9485195981984704,
+      "learning_rate": 1.1623779457613263e-05,
+      "loss": 0.7493,
+      "step": 803
+    },
+    {
+      "epoch": 2.519968676585748,
+      "grad_norm": 0.9438738641259264,
+      "learning_rate": 1.160217349692051e-05,
+      "loss": 0.7592,
+      "step": 804
+    },
+    {
+      "epoch": 2.523101018010963,
+      "grad_norm": 0.9775231067597289,
+      "learning_rate": 1.1580559857216347e-05,
+      "loss": 0.7669,
+      "step": 805
+    },
+    {
+      "epoch": 2.5262333594361785,
+      "grad_norm": 0.9964813304861763,
+      "learning_rate": 1.1558938642092161e-05,
+      "loss": 0.75,
+      "step": 806
+    },
+    {
+      "epoch": 2.529365700861394,
+      "grad_norm": 0.9604452115462686,
+      "learning_rate": 1.1537309955175655e-05,
+      "loss": 0.7367,
+      "step": 807
+    },
+    {
+      "epoch": 2.5324980422866092,
+      "grad_norm": 0.9243296564169551,
+      "learning_rate": 1.1515673900130344e-05,
+      "loss": 0.7135,
+      "step": 808
+    },
+    {
+      "epoch": 2.5356303837118244,
+      "grad_norm": 0.9840647742781437,
+      "learning_rate": 1.149403058065506e-05,
+      "loss": 0.7744,
+      "step": 809
+    },
+    {
+      "epoch": 2.53876272513704,
+      "grad_norm": 0.9363389375104039,
+      "learning_rate": 1.1472380100483438e-05,
+      "loss": 0.7333,
+      "step": 810
+    },
+    {
+      "epoch": 2.5418950665622555,
+      "grad_norm": 1.005204535054278,
+      "learning_rate": 1.145072256338345e-05,
+      "loss": 0.7535,
+      "step": 811
+    },
+    {
+      "epoch": 2.5450274079874706,
+      "grad_norm": 0.9828566960512756,
+      "learning_rate": 1.1429058073156886e-05,
+      "loss": 0.8036,
+      "step": 812
+    },
+    {
+      "epoch": 2.5481597494126857,
+      "grad_norm": 0.9671196176463093,
+      "learning_rate": 1.140738673363886e-05,
+      "loss": 0.7781,
+      "step": 813
+    },
+    {
+      "epoch": 2.5512920908379013,
+      "grad_norm": 0.9532345512189722,
+      "learning_rate": 1.1385708648697306e-05,
+      "loss": 0.7126,
+      "step": 814
+    },
+    {
+      "epoch": 2.554424432263117,
+      "grad_norm": 0.9251640219717523,
+      "learning_rate": 1.1364023922232503e-05,
+      "loss": 0.7222,
+      "step": 815
+    },
+    {
+      "epoch": 2.557556773688332,
+      "grad_norm": 0.9869679713832883,
+      "learning_rate": 1.1342332658176556e-05,
+      "loss": 0.7058,
+      "step": 816
+    },
+    {
+      "epoch": 2.560689115113547,
+      "grad_norm": 0.9930580206564761,
+      "learning_rate": 1.13206349604929e-05,
+      "loss": 0.7729,
+      "step": 817
+    },
+    {
+      "epoch": 2.5638214565387627,
+      "grad_norm": 0.9420081126831853,
+      "learning_rate": 1.1298930933175805e-05,
+      "loss": 0.7284,
+      "step": 818
+    },
+    {
+      "epoch": 2.5669537979639783,
+      "grad_norm": 1.0000782701863635,
+      "learning_rate": 1.1277220680249887e-05,
+      "loss": 0.7776,
+      "step": 819
+    },
+    {
+      "epoch": 2.5700861393891934,
+      "grad_norm": 0.9515958392786626,
+      "learning_rate": 1.1255504305769589e-05,
+      "loss": 0.7402,
+      "step": 820
+    },
+    {
+      "epoch": 2.5732184808144085,
+      "grad_norm": 0.9398121571554576,
+      "learning_rate": 1.1233781913818707e-05,
+      "loss": 0.7255,
+      "step": 821
+    },
+    {
+      "epoch": 2.576350822239624,
+      "grad_norm": 0.9937093752648678,
+      "learning_rate": 1.1212053608509862e-05,
+      "loss": 0.7627,
+      "step": 822
+    },
+    {
+      "epoch": 2.5794831636648397,
+      "grad_norm": 0.9791666530858968,
+      "learning_rate": 1.119031949398403e-05,
+      "loss": 0.7689,
+      "step": 823
+    },
+    {
+      "epoch": 2.5826155050900548,
+      "grad_norm": 0.9496103182684476,
+      "learning_rate": 1.1168579674410026e-05,
+      "loss": 0.7458,
+      "step": 824
+    },
+    {
+      "epoch": 2.58574784651527,
+      "grad_norm": 0.9935430896103922,
+      "learning_rate": 1.1146834253984008e-05,
+      "loss": 0.7673,
+      "step": 825
+    },
+    {
+      "epoch": 2.5888801879404855,
+      "grad_norm": 0.9901446257978567,
+      "learning_rate": 1.1125083336928974e-05,
+      "loss": 0.767,
+      "step": 826
+    },
+    {
+      "epoch": 2.592012529365701,
+      "grad_norm": 0.9886430099783694,
+      "learning_rate": 1.1103327027494272e-05,
+      "loss": 0.7649,
+      "step": 827
+    },
+    {
+      "epoch": 2.595144870790916,
+      "grad_norm": 0.957433731138305,
+      "learning_rate": 1.10815654299551e-05,
+      "loss": 0.7496,
+      "step": 828
+    },
+    {
+      "epoch": 2.5982772122161313,
+      "grad_norm": 0.9549512909328034,
+      "learning_rate": 1.1059798648611988e-05,
+      "loss": 0.7479,
+      "step": 829
+    },
+    {
+      "epoch": 2.601409553641347,
+      "grad_norm": 0.9729410513377862,
+      "learning_rate": 1.103802678779032e-05,
+      "loss": 0.7215,
+      "step": 830
+    },
+    {
+      "epoch": 2.6045418950665624,
+      "grad_norm": 0.9739903952440084,
+      "learning_rate": 1.1016249951839824e-05,
+      "loss": 0.7481,
+      "step": 831
+    },
+    {
+      "epoch": 2.6076742364917775,
+      "grad_norm": 0.9413669343604363,
+      "learning_rate": 1.0994468245134071e-05,
+      "loss": 0.7567,
+      "step": 832
+    },
+    {
+      "epoch": 2.610806577916993,
+      "grad_norm": 0.9742157973187829,
+      "learning_rate": 1.0972681772069981e-05,
+      "loss": 0.7463,
+      "step": 833
+    },
+    {
+      "epoch": 2.6139389193422082,
+      "grad_norm": 0.9778627829815928,
+      "learning_rate": 1.0950890637067317e-05,
+      "loss": 0.7633,
+      "step": 834
+    },
+    {
+      "epoch": 2.617071260767424,
+      "grad_norm": 0.9429091429491205,
+      "learning_rate": 1.0929094944568182e-05,
+      "loss": 0.7321,
+      "step": 835
+    },
+    {
+      "epoch": 2.620203602192639,
+      "grad_norm": 0.9807225580513256,
+      "learning_rate": 1.0907294799036525e-05,
+      "loss": 0.7599,
+      "step": 836
+    },
+    {
+      "epoch": 2.6233359436178545,
+      "grad_norm": 0.9466168414336543,
+      "learning_rate": 1.0885490304957642e-05,
+      "loss": 0.7111,
+      "step": 837
+    },
+    {
+      "epoch": 2.6264682850430696,
+      "grad_norm": 0.9473666583781692,
+      "learning_rate": 1.0863681566837666e-05,
+      "loss": 0.7333,
+      "step": 838
+    },
+    {
+      "epoch": 2.629600626468285,
+      "grad_norm": 0.9524909535219792,
+      "learning_rate": 1.0841868689203072e-05,
+      "loss": 0.7625,
+      "step": 839
+    },
+    {
+      "epoch": 2.6327329678935003,
+      "grad_norm": 0.9432869054961515,
+      "learning_rate": 1.0820051776600175e-05,
+      "loss": 0.745,
+      "step": 840
+    },
+    {
+      "epoch": 2.635865309318716,
+      "grad_norm": 0.940106657073856,
+      "learning_rate": 1.0798230933594629e-05,
+      "loss": 0.7438,
+      "step": 841
+    },
+    {
+      "epoch": 2.638997650743931,
+      "grad_norm": 0.959855959982091,
+      "learning_rate": 1.0776406264770933e-05,
+      "loss": 0.7728,
+      "step": 842
+    },
+    {
+      "epoch": 2.6421299921691466,
+      "grad_norm": 1.0045635045124952,
+      "learning_rate": 1.0754577874731911e-05,
+      "loss": 0.774,
+      "step": 843
+    },
+    {
+      "epoch": 2.6452623335943617,
+      "grad_norm": 0.9757139770765741,
+      "learning_rate": 1.073274586809823e-05,
+      "loss": 0.7312,
+      "step": 844
+    },
+    {
+      "epoch": 2.6483946750195773,
+      "grad_norm": 0.9438533061715757,
+      "learning_rate": 1.0710910349507881e-05,
+      "loss": 0.7237,
+      "step": 845
+    },
+    {
+      "epoch": 2.6515270164447924,
+      "grad_norm": 0.9531758260863951,
+      "learning_rate": 1.0689071423615708e-05,
+      "loss": 0.7348,
+      "step": 846
+    },
+    {
+      "epoch": 2.654659357870008,
+      "grad_norm": 0.9446644812284712,
+      "learning_rate": 1.0667229195092868e-05,
+      "loss": 0.7602,
+      "step": 847
+    },
+    {
+      "epoch": 2.657791699295223,
+      "grad_norm": 0.9782412701275756,
+      "learning_rate": 1.0645383768626345e-05,
+      "loss": 0.7765,
+      "step": 848
+    },
+    {
+      "epoch": 2.6609240407204386,
+      "grad_norm": 1.007139575996801,
+      "learning_rate": 1.0623535248918463e-05,
+      "loss": 0.7751,
+      "step": 849
+    },
+    {
+      "epoch": 2.6640563821456538,
+      "grad_norm": 0.972578870009894,
+      "learning_rate": 1.0601683740686366e-05,
+      "loss": 0.8046,
+      "step": 850
+    },
+    {
+      "epoch": 2.6671887235708693,
+      "grad_norm": 0.9533200080779424,
+      "learning_rate": 1.0579829348661522e-05,
+      "loss": 0.7268,
+      "step": 851
+    },
+    {
+      "epoch": 2.6703210649960845,
+      "grad_norm": 0.9569784524332555,
+      "learning_rate": 1.0557972177589214e-05,
+      "loss": 0.7914,
+      "step": 852
+    },
+    {
+      "epoch": 2.6734534064213,
+      "grad_norm": 0.9554744612301984,
+      "learning_rate": 1.0536112332228057e-05,
+      "loss": 0.7656,
+      "step": 853
+    },
+    {
+      "epoch": 2.676585747846515,
+      "grad_norm": 0.9379633252744265,
+      "learning_rate": 1.0514249917349473e-05,
+      "loss": 0.7174,
+      "step": 854
+    },
+    {
+      "epoch": 2.6797180892717307,
+      "grad_norm": 0.9498518801324679,
+      "learning_rate": 1.0492385037737207e-05,
+      "loss": 0.7359,
+      "step": 855
+    },
+    {
+      "epoch": 2.682850430696946,
+      "grad_norm": 0.9323391831700777,
+      "learning_rate": 1.0470517798186813e-05,
+      "loss": 0.7516,
+      "step": 856
+    },
+    {
+      "epoch": 2.6859827721221614,
+      "grad_norm": 0.977808841738496,
+      "learning_rate": 1.044864830350515e-05,
+      "loss": 0.7588,
+      "step": 857
+    },
+    {
+      "epoch": 2.6891151135473765,
+      "grad_norm": 0.9607140331334884,
+      "learning_rate": 1.0426776658509899e-05,
+      "loss": 0.7297,
+      "step": 858
+    },
+    {
+      "epoch": 2.692247454972592,
+      "grad_norm": 0.9644638205133018,
+      "learning_rate": 1.040490296802904e-05,
+      "loss": 0.7602,
+      "step": 859
+    },
+    {
+      "epoch": 2.6953797963978072,
+      "grad_norm": 0.9541062791921975,
+      "learning_rate": 1.0383027336900356e-05,
+      "loss": 0.747,
+      "step": 860
+    },
+    {
+      "epoch": 2.698512137823023,
+      "grad_norm": 0.9713788632902975,
+      "learning_rate": 1.0361149869970929e-05,
+      "loss": 0.7486,
+      "step": 861
+    },
+    {
+      "epoch": 2.701644479248238,
+      "grad_norm": 0.9599164742932305,
+      "learning_rate": 1.0339270672096645e-05,
+      "loss": 0.7677,
+      "step": 862
+    },
+    {
+      "epoch": 2.7047768206734535,
+      "grad_norm": 0.9851639258139112,
+      "learning_rate": 1.0317389848141685e-05,
+      "loss": 0.735,
+      "step": 863
+    },
+    {
+      "epoch": 2.7079091620986686,
+      "grad_norm": 0.9627844834333039,
+      "learning_rate": 1.0295507502978024e-05,
+      "loss": 0.7423,
+      "step": 864
+    },
+    {
+      "epoch": 2.711041503523884,
+      "grad_norm": 0.9528844028329699,
+      "learning_rate": 1.0273623741484924e-05,
+      "loss": 0.7384,
+      "step": 865
+    },
+    {
+      "epoch": 2.7141738449490993,
+      "grad_norm": 1.0340380298089817,
+      "learning_rate": 1.0251738668548439e-05,
+      "loss": 0.8031,
+      "step": 866
+    },
+    {
+      "epoch": 2.717306186374315,
+      "grad_norm": 0.9538815854838193,
+      "learning_rate": 1.0229852389060905e-05,
+      "loss": 0.7339,
+      "step": 867
+    },
+    {
+      "epoch": 2.72043852779953,
+      "grad_norm": 0.9723625755825231,
+      "learning_rate": 1.0207965007920448e-05,
+      "loss": 0.7837,
+      "step": 868
+    },
+    {
+      "epoch": 2.7235708692247456,
+      "grad_norm": 0.967760456332842,
+      "learning_rate": 1.0186076630030469e-05,
+      "loss": 0.7522,
+      "step": 869
+    },
+    {
+      "epoch": 2.7267032106499607,
+      "grad_norm": 0.956046387699816,
+      "learning_rate": 1.0164187360299143e-05,
+      "loss": 0.7055,
+      "step": 870
+    },
+    {
+      "epoch": 2.7298355520751763,
+      "grad_norm": 0.9568472870023063,
+      "learning_rate": 1.0142297303638926e-05,
+      "loss": 0.7284,
+      "step": 871
+    },
+    {
+      "epoch": 2.7329678935003914,
+      "grad_norm": 0.988027374434136,
+      "learning_rate": 1.0120406564966043e-05,
+      "loss": 0.7714,
+      "step": 872
+    },
+    {
+      "epoch": 2.736100234925607,
+      "grad_norm": 1.028366948188634,
+      "learning_rate": 1.0098515249199988e-05,
+      "loss": 0.8163,
+      "step": 873
+    },
+    {
+      "epoch": 2.7392325763508225,
+      "grad_norm": 1.4617689871519928,
+      "learning_rate": 1.0076623461263017e-05,
+      "loss": 0.7347,
+      "step": 874
+    },
+    {
+      "epoch": 2.7423649177760376,
+      "grad_norm": 0.9546506131293968,
+      "learning_rate": 1.0054731306079657e-05,
+      "loss": 0.7131,
+      "step": 875
+    },
+    {
+      "epoch": 2.7454972592012528,
+      "grad_norm": 0.9591903904399064,
+      "learning_rate": 1.0032838888576194e-05,
+      "loss": 0.72,
+      "step": 876
+    },
+    {
+      "epoch": 2.7486296006264683,
+      "grad_norm": 0.9704760247984783,
+      "learning_rate": 1.0010946313680156e-05,
+      "loss": 0.7519,
+      "step": 877
+    },
+    {
+      "epoch": 2.751761942051684,
+      "grad_norm": 0.9496820302125051,
+      "learning_rate": 9.989053686319849e-06,
+      "loss": 0.7435,
+      "step": 878
+    },
+    {
+      "epoch": 2.754894283476899,
+      "grad_norm": 0.9773806934541943,
+      "learning_rate": 9.967161111423813e-06,
+      "loss": 0.7795,
+      "step": 879
+    },
+    {
+      "epoch": 2.758026624902114,
+      "grad_norm": 0.9202467047384671,
+      "learning_rate": 9.945268693920346e-06,
+      "loss": 0.7292,
+      "step": 880
+    },
+    {
+      "epoch": 2.7611589663273297,
+      "grad_norm": 0.974687628354556,
+      "learning_rate": 9.923376538736985e-06,
+      "loss": 0.7761,
+      "step": 881
+    },
+    {
+      "epoch": 2.7642913077525453,
+      "grad_norm": 0.9294513579702289,
+      "learning_rate": 9.901484750800014e-06,
+      "loss": 0.7235,
+      "step": 882
+    },
+    {
+      "epoch": 2.7674236491777604,
+      "grad_norm": 0.9850151969699499,
+      "learning_rate": 9.879593435033957e-06,
+      "loss": 0.7845,
+      "step": 883
+    },
+    {
+      "epoch": 2.7705559906029755,
+      "grad_norm": 0.929292922655392,
+      "learning_rate": 9.857702696361076e-06,
+      "loss": 0.75,
+      "step": 884
+    },
+    {
+      "epoch": 2.773688332028191,
+      "grad_norm": 0.9256442001669153,
+      "learning_rate": 9.835812639700862e-06,
+      "loss": 0.7215,
+      "step": 885
+    },
+    {
+      "epoch": 2.7768206734534067,
+      "grad_norm": 0.9481844945658222,
+      "learning_rate": 9.813923369969535e-06,
+      "loss": 0.7381,
+      "step": 886
+    },
+    {
+      "epoch": 2.779953014878622,
+      "grad_norm": 0.9481037841878706,
+      "learning_rate": 9.792034992079554e-06,
+      "loss": 0.7556,
+      "step": 887
+    },
+    {
+      "epoch": 2.783085356303837,
+      "grad_norm": 0.952957986686458,
+      "learning_rate": 9.770147610939098e-06,
+      "loss": 0.7339,
+      "step": 888
+    },
+    {
+      "epoch": 2.7862176977290525,
+      "grad_norm": 0.9549009956792547,
+      "learning_rate": 9.748261331451566e-06,
+      "loss": 0.7476,
+      "step": 889
+    },
+    {
+      "epoch": 2.789350039154268,
+      "grad_norm": 0.9782654041320651,
+      "learning_rate": 9.726376258515077e-06,
+      "loss": 0.7421,
+      "step": 890
+    },
+    {
+      "epoch": 2.792482380579483,
+      "grad_norm": 0.9900817275321431,
+      "learning_rate": 9.70449249702198e-06,
+      "loss": 0.7703,
+      "step": 891
+    },
+    {
+      "epoch": 2.7956147220046983,
+      "grad_norm": 0.9684826883994128,
+      "learning_rate": 9.682610151858317e-06,
+      "loss": 0.7874,
+      "step": 892
+    },
+    {
+      "epoch": 2.798747063429914,
+      "grad_norm": 0.9509743876454371,
+      "learning_rate": 9.660729327903359e-06,
+      "loss": 0.7554,
+      "step": 893
+    },
+    {
+      "epoch": 2.8018794048551294,
+      "grad_norm": 0.9651752336989294,
+      "learning_rate": 9.638850130029075e-06,
+      "loss": 0.7081,
+      "step": 894
+    },
+    {
+      "epoch": 2.8050117462803446,
+      "grad_norm": 0.9947227809634293,
+      "learning_rate": 9.616972663099648e-06,
+      "loss": 0.7533,
+      "step": 895
+    },
+    {
+      "epoch": 2.8081440877055597,
+      "grad_norm": 0.9575600590215773,
+      "learning_rate": 9.595097031970963e-06,
+      "loss": 0.7567,
+      "step": 896
+    },
+    {
+      "epoch": 2.8112764291307752,
+      "grad_norm": 0.9658030860447785,
+      "learning_rate": 9.5732233414901e-06,
+      "loss": 0.747,
+      "step": 897
+    },
+    {
+      "epoch": 2.814408770555991,
+      "grad_norm": 0.995325139613679,
+      "learning_rate": 9.551351696494854e-06,
+      "loss": 0.7438,
+      "step": 898
+    },
+    {
+      "epoch": 2.817541111981206,
+      "grad_norm": 0.9712772975214642,
+      "learning_rate": 9.529482201813192e-06,
+      "loss": 0.7293,
+      "step": 899
+    },
+    {
+      "epoch": 2.820673453406421,
+      "grad_norm": 0.9546909848683623,
+      "learning_rate": 9.507614962262796e-06,
+      "loss": 0.768,
+      "step": 900
+    },
+    {
+      "epoch": 2.8238057948316366,
+      "grad_norm": 0.9329589715311891,
+      "learning_rate": 9.48575008265053e-06,
+      "loss": 0.7341,
+      "step": 901
+    },
+    {
+      "epoch": 2.826938136256852,
+      "grad_norm": 0.9357071849591595,
+      "learning_rate": 9.463887667771946e-06,
+      "loss": 0.7195,
+      "step": 902
+    },
+    {
+      "epoch": 2.8300704776820673,
+      "grad_norm": 0.933792474330246,
+      "learning_rate": 9.442027822410786e-06,
+      "loss": 0.7563,
+      "step": 903
+    },
+    {
+      "epoch": 2.8332028191072824,
+      "grad_norm": 0.9936060316623524,
+      "learning_rate": 9.420170651338482e-06,
+      "loss": 0.7546,
+      "step": 904
+    },
+    {
+      "epoch": 2.836335160532498,
+      "grad_norm": 0.9522662188147043,
+      "learning_rate": 9.398316259313636e-06,
+      "loss": 0.744,
+      "step": 905
+    },
+    {
+      "epoch": 2.8394675019577136,
+      "grad_norm": 0.9768599306563925,
+      "learning_rate": 9.376464751081542e-06,
+      "loss": 0.7547,
+      "step": 906
+    },
+    {
+      "epoch": 2.8425998433829287,
+      "grad_norm": 0.9612851410888434,
+      "learning_rate": 9.354616231373659e-06,
+      "loss": 0.7662,
+      "step": 907
+    },
+    {
+      "epoch": 2.845732184808144,
+      "grad_norm": 0.9419299539508802,
+      "learning_rate": 9.332770804907137e-06,
+      "loss": 0.7293,
+      "step": 908
+    },
+    {
+      "epoch": 2.8488645262333594,
+      "grad_norm": 0.9437811849610814,
+      "learning_rate": 9.310928576384293e-06,
+      "loss": 0.7258,
+      "step": 909
+    },
+    {
+      "epoch": 2.851996867658575,
+      "grad_norm": 0.9714173836845524,
+      "learning_rate": 9.289089650492119e-06,
+      "loss": 0.732,
+      "step": 910
+    },
+    {
+      "epoch": 2.85512920908379,
+      "grad_norm": 0.9482886998971121,
+      "learning_rate": 9.267254131901772e-06,
+      "loss": 0.7273,
+      "step": 911
+    },
+    {
+      "epoch": 2.858261550509005,
+      "grad_norm": 0.9868925213214547,
+      "learning_rate": 9.245422125268094e-06,
+      "loss": 0.7215,
+      "step": 912
+    },
+    {
+      "epoch": 2.861393891934221,
+      "grad_norm": 0.9529681945509453,
+      "learning_rate": 9.223593735229069e-06,
+      "loss": 0.7498,
+      "step": 913
+    },
+    {
+      "epoch": 2.8645262333594363,
+      "grad_norm": 0.9383207113099425,
+      "learning_rate": 9.201769066405373e-06,
+      "loss": 0.7302,
+      "step": 914
+    },
+    {
+      "epoch": 2.8676585747846515,
+      "grad_norm": 0.9565533325468113,
+      "learning_rate": 9.179948223399828e-06,
+      "loss": 0.7352,
+      "step": 915
+    },
+    {
+      "epoch": 2.870790916209867,
+      "grad_norm": 0.973007632374376,
+      "learning_rate": 9.15813131079693e-06,
+      "loss": 0.7646,
+      "step": 916
+    },
+    {
+      "epoch": 2.873923257635082,
+      "grad_norm": 0.9966309662857158,
+      "learning_rate": 9.136318433162336e-06,
+      "loss": 0.724,
+      "step": 917
+    },
+    {
+      "epoch": 2.8770555990602977,
+      "grad_norm": 0.9853350271329723,
+      "learning_rate": 9.114509695042362e-06,
+      "loss": 0.7615,
+      "step": 918
+    },
+    {
+      "epoch": 2.880187940485513,
+      "grad_norm": 0.9504300421132944,
+      "learning_rate": 9.092705200963478e-06,
+      "loss": 0.7663,
+      "step": 919
+    },
+    {
+      "epoch": 2.8833202819107284,
+      "grad_norm": 0.9477723894786849,
+      "learning_rate": 9.070905055431822e-06,
+      "loss": 0.7372,
+      "step": 920
+    },
+    {
+      "epoch": 2.8864526233359435,
+      "grad_norm": 0.9495506324641605,
+      "learning_rate": 9.049109362932685e-06,
+      "loss": 0.7323,
+      "step": 921
+    },
+    {
+      "epoch": 2.889584964761159,
+      "grad_norm": 0.9315436534258359,
+      "learning_rate": 9.02731822793002e-06,
+      "loss": 0.7298,
+      "step": 922
+    },
+    {
+      "epoch": 2.8927173061863742,
+      "grad_norm": 0.9633745464994017,
+      "learning_rate": 9.005531754865929e-06,
+      "loss": 0.7329,
+      "step": 923
+    },
+    {
+      "epoch": 2.89584964761159,
+      "grad_norm": 0.9520011329637497,
+      "learning_rate": 8.983750048160178e-06,
+      "loss": 0.7461,
+      "step": 924
+    },
+    {
+      "epoch": 2.898981989036805,
+      "grad_norm": 0.947202489605334,
+      "learning_rate": 8.961973212209684e-06,
+      "loss": 0.7353,
+      "step": 925
+    },
+    {
+      "epoch": 2.9021143304620205,
+      "grad_norm": 0.9564588694025306,
+      "learning_rate": 8.940201351388016e-06,
+      "loss": 0.7499,
+      "step": 926
+    },
+    {
+      "epoch": 2.9052466718872356,
+      "grad_norm": 0.9660352117956935,
+      "learning_rate": 8.918434570044904e-06,
+      "loss": 0.774,
+      "step": 927
+    },
+    {
+      "epoch": 2.908379013312451,
+      "grad_norm": 0.9516602740353381,
+      "learning_rate": 8.89667297250573e-06,
+      "loss": 0.7196,
+      "step": 928
+    },
+    {
+      "epoch": 2.9115113547376663,
+      "grad_norm": 0.9503514513189356,
+      "learning_rate": 8.874916663071027e-06,
+      "loss": 0.7621,
+      "step": 929
+    },
+    {
+      "epoch": 2.914643696162882,
+      "grad_norm": 0.9732615080165525,
+      "learning_rate": 8.853165746015997e-06,
+      "loss": 0.7596,
+      "step": 930
+    },
+    {
+      "epoch": 2.917776037588097,
+      "grad_norm": 1.010101952673623,
+      "learning_rate": 8.831420325589976e-06,
+      "loss": 0.7654,
+      "step": 931
+    },
+    {
+      "epoch": 2.9209083790133126,
+      "grad_norm": 0.9740311574523944,
+      "learning_rate": 8.809680506015973e-06,
+      "loss": 0.7513,
+      "step": 932
+    },
+    {
+      "epoch": 2.9240407204385277,
+      "grad_norm": 0.9960714830033841,
+      "learning_rate": 8.787946391490141e-06,
+      "loss": 0.7656,
+      "step": 933
+    },
+    {
+      "epoch": 2.9271730618637433,
+      "grad_norm": 0.9638028878471048,
+      "learning_rate": 8.766218086181298e-06,
+      "loss": 0.7776,
+      "step": 934
+    },
+    {
+      "epoch": 2.9303054032889584,
+      "grad_norm": 0.9722845022040784,
+      "learning_rate": 8.744495694230413e-06,
+      "loss": 0.7301,
+      "step": 935
+    },
+    {
+      "epoch": 2.933437744714174,
+      "grad_norm": 0.9628137063571912,
+      "learning_rate": 8.722779319750116e-06,
+      "loss": 0.7206,
+      "step": 936
+    },
+    {
+      "epoch": 2.936570086139389,
+      "grad_norm": 0.9465035800698545,
+      "learning_rate": 8.701069066824195e-06,
+      "loss": 0.722,
+      "step": 937
+    },
+    {
+      "epoch": 2.9397024275646046,
+      "grad_norm": 0.9902093031760724,
+      "learning_rate": 8.679365039507104e-06,
+      "loss": 0.7857,
+      "step": 938
+    },
+    {
+      "epoch": 2.9428347689898198,
+      "grad_norm": 0.948550028383559,
+      "learning_rate": 8.657667341823449e-06,
+      "loss": 0.7501,
+      "step": 939
+    },
+    {
+      "epoch": 2.9459671104150353,
+      "grad_norm": 0.9610537515494351,
+      "learning_rate": 8.6359760777675e-06,
+      "loss": 0.7242,
+      "step": 940
+    },
+    {
+      "epoch": 2.9490994518402505,
+      "grad_norm": 0.9553995440705682,
+      "learning_rate": 8.614291351302697e-06,
+      "loss": 0.7344,
+      "step": 941
+    },
+    {
+      "epoch": 2.952231793265466,
+      "grad_norm": 0.9521585819489644,
+      "learning_rate": 8.592613266361145e-06,
+      "loss": 0.7399,
+      "step": 942
+    },
+    {
+      "epoch": 2.955364134690681,
+      "grad_norm": 0.9675254410966843,
+      "learning_rate": 8.570941926843117e-06,
+      "loss": 0.7474,
+      "step": 943
+    },
+    {
+      "epoch": 2.9584964761158967,
+      "grad_norm": 0.949925314091798,
+      "learning_rate": 8.549277436616551e-06,
+      "loss": 0.7398,
+      "step": 944
+    },
+    {
+      "epoch": 2.961628817541112,
+      "grad_norm": 0.9743585973394909,
+      "learning_rate": 8.527619899516567e-06,
+      "loss": 0.7281,
+      "step": 945
+    },
+    {
+      "epoch": 2.9647611589663274,
+      "grad_norm": 0.9721394272700253,
+      "learning_rate": 8.505969419344946e-06,
+      "loss": 0.7691,
+      "step": 946
+    },
+    {
+      "epoch": 2.9678935003915425,
+      "grad_norm": 0.9756436120696973,
+      "learning_rate": 8.484326099869659e-06,
+      "loss": 0.7346,
+      "step": 947
+    },
+    {
+      "epoch": 2.971025841816758,
+      "grad_norm": 0.963968460621726,
+      "learning_rate": 8.462690044824347e-06,
+      "loss": 0.7632,
+      "step": 948
+    },
+    {
+      "epoch": 2.9741581832419732,
+      "grad_norm": 0.9886664730071104,
+      "learning_rate": 8.441061357907842e-06,
+      "loss": 0.7795,
+      "step": 949
+    },
+    {
+      "epoch": 2.977290524667189,
+      "grad_norm": 0.9402625444423229,
+      "learning_rate": 8.419440142783653e-06,
+      "loss": 0.7237,
+      "step": 950
+    },
+    {
+      "epoch": 2.980422866092404,
+      "grad_norm": 0.9602243212134389,
+      "learning_rate": 8.397826503079489e-06,
+      "loss": 0.7561,
+      "step": 951
+    },
+    {
+      "epoch": 2.9835552075176195,
+      "grad_norm": 0.9617618663543123,
+      "learning_rate": 8.376220542386742e-06,
+      "loss": 0.7799,
+      "step": 952
+    },
+    {
+      "epoch": 2.9866875489428346,
+      "grad_norm": 0.9190755212609129,
+      "learning_rate": 8.354622364259995e-06,
+      "loss": 0.7042,
+      "step": 953
+    },
+    {
+      "epoch": 2.98981989036805,
+      "grad_norm": 0.9586872985486566,
+      "learning_rate": 8.333032072216542e-06,
+      "loss": 0.7411,
+      "step": 954
+    },
+    {
+      "epoch": 2.9929522317932653,
+      "grad_norm": 0.9490934425198838,
+      "learning_rate": 8.311449769735874e-06,
+      "loss": 0.7699,
+      "step": 955
+    },
+    {
+      "epoch": 2.996084573218481,
+      "grad_norm": 0.9483587215466358,
+      "learning_rate": 8.289875560259189e-06,
+      "loss": 0.729,
+      "step": 956
+    },
+    {
+      "epoch": 2.999216914643696,
+      "grad_norm": 0.9426645402788387,
+      "learning_rate": 8.268309547188894e-06,
+      "loss": 0.7605,
+      "step": 957
+    },
+    {
+      "epoch": 3.0031323414252156,
+      "grad_norm": 2.4449431031791447,
+      "learning_rate": 8.246751833888122e-06,
+      "loss": 1.363,
+      "step": 958
+    },
+    {
+      "epoch": 3.0062646828504307,
+      "grad_norm": 1.6451995720214123,
+      "learning_rate": 8.225202523680218e-06,
+      "loss": 0.5471,
+      "step": 959
+    },
+    {
+      "epoch": 3.0093970242756463,
+      "grad_norm": 1.6727678935052752,
+      "learning_rate": 8.203661719848249e-06,
+      "loss": 0.5595,
+      "step": 960
+    },
+    {
+      "epoch": 3.0125293657008614,
+      "grad_norm": 1.5460240913854986,
+      "learning_rate": 8.182129525634513e-06,
+      "loss": 0.5325,
+      "step": 961
+    },
+    {
+      "epoch": 3.015661707126077,
+      "grad_norm": 1.2460927908669523,
+      "learning_rate": 8.160606044240052e-06,
+      "loss": 0.5299,
+      "step": 962
+    },
+    {
+      "epoch": 3.018794048551292,
+      "grad_norm": 1.2859700150226807,
+      "learning_rate": 8.139091378824144e-06,
+      "loss": 0.526,
+      "step": 963
+    },
+    {
+      "epoch": 3.0219263899765076,
+      "grad_norm": 1.7164212978411557,
+      "learning_rate": 8.117585632503805e-06,
+      "loss": 0.5641,
+      "step": 964
+    },
+    {
+      "epoch": 3.0250587314017228,
+      "grad_norm": 1.7851156197762952,
+      "learning_rate": 8.096088908353316e-06,
+      "loss": 0.5354,
+      "step": 965
+    },
+    {
+      "epoch": 3.0281910728269383,
+      "grad_norm": 1.7956222015211165,
+      "learning_rate": 8.07460130940371e-06,
+      "loss": 0.5468,
+      "step": 966
+    },
+    {
+      "epoch": 3.0313234142521535,
+      "grad_norm": 1.7931934712777515,
+      "learning_rate": 8.05312293864228e-06,
+      "loss": 0.5571,
+      "step": 967
+    },
+    {
+      "epoch": 3.034455755677369,
+      "grad_norm": 1.6976805325976927,
+      "learning_rate": 8.031653899012098e-06,
+      "loss": 0.5492,
+      "step": 968
+    },
+    {
+      "epoch": 3.037588097102584,
+      "grad_norm": 1.596398489281078,
+      "learning_rate": 8.010194293411502e-06,
+      "loss": 0.4879,
+      "step": 969
+    },
+    {
+      "epoch": 3.0407204385277997,
+      "grad_norm": 1.4311365997826093,
+      "learning_rate": 7.988744224693625e-06,
+      "loss": 0.5483,
+      "step": 970
+    },
+    {
+      "epoch": 3.043852779953015,
+      "grad_norm": 1.3005396695284464,
+      "learning_rate": 7.967303795665887e-06,
+      "loss": 0.5155,
+      "step": 971
+    },
+    {
+      "epoch": 3.0469851213782304,
+      "grad_norm": 1.1670816271780478,
+      "learning_rate": 7.945873109089503e-06,
+      "loss": 0.5483,
+      "step": 972
+    },
+    {
+      "epoch": 3.0501174628034455,
+      "grad_norm": 1.2526413892546973,
+      "learning_rate": 7.924452267678998e-06,
+      "loss": 0.5208,
+      "step": 973
+    },
+    {
+      "epoch": 3.053249804228661,
+      "grad_norm": 1.3258191230420822,
+      "learning_rate": 7.903041374101707e-06,
+      "loss": 0.5597,
+      "step": 974
+    },
+    {
+      "epoch": 3.0563821456538762,
+      "grad_norm": 1.3179145976075899,
+      "learning_rate": 7.881640530977287e-06,
+      "loss": 0.5544,
+      "step": 975
+    },
+    {
+      "epoch": 3.059514487079092,
+      "grad_norm": 1.2661427204898215,
+      "learning_rate": 7.860249840877232e-06,
+      "loss": 0.5306,
+      "step": 976
+    },
+    {
+      "epoch": 3.062646828504307,
+      "grad_norm": 1.20947370272958,
+      "learning_rate": 7.838869406324358e-06,
+      "loss": 0.529,
+      "step": 977
+    },
+    {
+      "epoch": 3.0657791699295225,
+      "grad_norm": 1.1968126239778358,
+      "learning_rate": 7.817499329792342e-06,
+      "loss": 0.5685,
+      "step": 978
+    },
+    {
+      "epoch": 3.0689115113547376,
+      "grad_norm": 1.1994215893244842,
+      "learning_rate": 7.796139713705214e-06,
+      "loss": 0.5157,
+      "step": 979
+    },
+    {
+      "epoch": 3.072043852779953,
+      "grad_norm": 1.231635060309257,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.5195,
+      "step": 980
+    },
+    {
+      "epoch": 3.0751761942051683,
+      "grad_norm": 1.2984303455103845,
+      "learning_rate": 7.753452272310544e-06,
+      "loss": 0.5169,
+      "step": 981
+    },
+    {
+      "epoch": 3.078308535630384,
+      "grad_norm": 1.2495741983021167,
+      "learning_rate": 7.732124651598418e-06,
+      "loss": 0.5286,
+      "step": 982
+    },
+    {
+      "epoch": 3.081440877055599,
+      "grad_norm": 1.2371927309933792,
+      "learning_rate": 7.710807900521023e-06,
+      "loss": 0.5285,
+      "step": 983
+    },
+    {
+      "epoch": 3.0845732184808146,
+      "grad_norm": 1.1598033338515672,
+      "learning_rate": 7.689502121246809e-06,
+      "loss": 0.5285,
+      "step": 984
+    },
+    {
+      "epoch": 3.0877055599060297,
+      "grad_norm": 1.137936793677818,
+      "learning_rate": 7.668207415891625e-06,
+      "loss": 0.5257,
+      "step": 985
+    },
+    {
+      "epoch": 3.0908379013312453,
+      "grad_norm": 1.2251396369541545,
+      "learning_rate": 7.646923886518263e-06,
+      "loss": 0.5563,
+      "step": 986
+    },
+    {
+      "epoch": 3.0939702427564604,
+      "grad_norm": 1.1991699852871776,
+      "learning_rate": 7.625651635135933e-06,
+      "loss": 0.527,
+      "step": 987
+    },
+    {
+      "epoch": 3.097102584181676,
+      "grad_norm": 1.1439621826805295,
+      "learning_rate": 7.6043907636998026e-06,
+      "loss": 0.5163,
+      "step": 988
+    },
+    {
+      "epoch": 3.100234925606891,
+      "grad_norm": 1.123851307419761,
+      "learning_rate": 7.583141374110495e-06,
+      "loss": 0.5206,
+      "step": 989
+    },
+    {
+      "epoch": 3.1033672670321066,
+      "grad_norm": 1.1309411775644675,
+      "learning_rate": 7.561903568213595e-06,
+      "loss": 0.5416,
+      "step": 990
+    },
+    {
+      "epoch": 3.1064996084573218,
+      "grad_norm": 1.164015600425183,
+      "learning_rate": 7.540677447799176e-06,
+      "loss": 0.5062,
+      "step": 991
+    },
+    {
+      "epoch": 3.1096319498825373,
+      "grad_norm": 1.1836169096484468,
+      "learning_rate": 7.519463114601304e-06,
+      "loss": 0.5476,
+      "step": 992
+    },
+    {
+      "epoch": 3.1127642913077525,
+      "grad_norm": 1.142816670241574,
+      "learning_rate": 7.4982606702975505e-06,
+      "loss": 0.5342,
+      "step": 993
+    },
+    {
+      "epoch": 3.115896632732968,
+      "grad_norm": 1.1503408738938847,
+      "learning_rate": 7.477070216508493e-06,
+      "loss": 0.5646,
+      "step": 994
+    },
+    {
+      "epoch": 3.119028974158183,
+      "grad_norm": 1.150741615449422,
+      "learning_rate": 7.455891854797256e-06,
+      "loss": 0.5175,
+      "step": 995
+    },
+    {
+      "epoch": 3.1221613155833987,
+      "grad_norm": 1.2034319285665727,
+      "learning_rate": 7.434725686669002e-06,
+      "loss": 0.5765,
+      "step": 996
+    },
+    {
+      "epoch": 3.125293657008614,
+      "grad_norm": 1.1915855507050621,
+      "learning_rate": 7.413571813570453e-06,
+      "loss": 0.5323,
+      "step": 997
+    },
+    {
+      "epoch": 3.1284259984338294,
+      "grad_norm": 1.2081976282063787,
+      "learning_rate": 7.392430336889398e-06,
+      "loss": 0.5266,
+      "step": 998
+    },
+    {
+      "epoch": 3.1315583398590445,
+      "grad_norm": 1.1322504322898508,
+      "learning_rate": 7.371301357954219e-06,
+      "loss": 0.5129,
+      "step": 999
+    },
+    {
+      "epoch": 3.13469068128426,
+      "grad_norm": 1.1548883914063983,
+      "learning_rate": 7.350184978033386e-06,
+      "loss": 0.5193,
+      "step": 1000
+    },
+    {
+      "epoch": 3.137823022709475,
+      "grad_norm": 1.1526932585853438,
+      "learning_rate": 7.329081298334994e-06,
+      "loss": 0.5102,
+      "step": 1001
+    },
+    {
+      "epoch": 3.140955364134691,
+      "grad_norm": 1.1697102123089387,
+      "learning_rate": 7.307990420006266e-06,
+      "loss": 0.5326,
+      "step": 1002
+    },
+    {
+      "epoch": 3.144087705559906,
+      "grad_norm": 1.1265238542719669,
+      "learning_rate": 7.286912444133067e-06,
+      "loss": 0.5299,
+      "step": 1003
+    },
+    {
+      "epoch": 3.1472200469851215,
+      "grad_norm": 1.1537612326667075,
+      "learning_rate": 7.265847471739422e-06,
+      "loss": 0.5151,
+      "step": 1004
+    },
+    {
+      "epoch": 3.1503523884103366,
+      "grad_norm": 1.1284758438997318,
+      "learning_rate": 7.244795603787036e-06,
+      "loss": 0.5215,
+      "step": 1005
+    },
+    {
+      "epoch": 3.153484729835552,
+      "grad_norm": 1.1977561671812984,
+      "learning_rate": 7.2237569411748045e-06,
+      "loss": 0.5126,
+      "step": 1006
+    },
+    {
+      "epoch": 3.1566170712607673,
+      "grad_norm": 1.1649308019431355,
+      "learning_rate": 7.202731584738323e-06,
+      "loss": 0.5289,
+      "step": 1007
+    },
+    {
+      "epoch": 3.159749412685983,
+      "grad_norm": 1.1377687577701208,
+      "learning_rate": 7.1817196352494245e-06,
+      "loss": 0.5191,
+      "step": 1008
+    },
+    {
+      "epoch": 3.162881754111198,
+      "grad_norm": 1.1332419084126149,
+      "learning_rate": 7.160721193415681e-06,
+      "loss": 0.4928,
+      "step": 1009
+    },
+    {
+      "epoch": 3.1660140955364136,
+      "grad_norm": 1.1255882422146672,
+      "learning_rate": 7.139736359879916e-06,
+      "loss": 0.505,
+      "step": 1010
+    },
+    {
+      "epoch": 3.1691464369616287,
+      "grad_norm": 1.1396125393838465,
+      "learning_rate": 7.118765235219742e-06,
+      "loss": 0.5262,
+      "step": 1011
+    },
+    {
+      "epoch": 3.1722787783868442,
+      "grad_norm": 1.221304685970849,
+      "learning_rate": 7.0978079199470596e-06,
+      "loss": 0.5129,
+      "step": 1012
+    },
+    {
+      "epoch": 3.1754111198120594,
+      "grad_norm": 1.1646706730711807,
+      "learning_rate": 7.0768645145075885e-06,
+      "loss": 0.5509,
+      "step": 1013
+    },
+    {
+      "epoch": 3.178543461237275,
+      "grad_norm": 1.1909834940309783,
+      "learning_rate": 7.055935119280369e-06,
+      "loss": 0.5352,
+      "step": 1014
+    },
+    {
+      "epoch": 3.18167580266249,
+      "grad_norm": 1.1275894474381307,
+      "learning_rate": 7.035019834577301e-06,
+      "loss": 0.4987,
+      "step": 1015
+    },
+    {
+      "epoch": 3.1848081440877056,
+      "grad_norm": 1.1098330354725712,
+      "learning_rate": 7.014118760642654e-06,
+      "loss": 0.5429,
+      "step": 1016
+    },
+    {
+      "epoch": 3.1879404855129208,
+      "grad_norm": 1.1380777641842976,
+      "learning_rate": 6.993231997652587e-06,
+      "loss": 0.516,
+      "step": 1017
+    },
+    {
+      "epoch": 3.1910728269381363,
+      "grad_norm": 1.1693763120706757,
+      "learning_rate": 6.972359645714666e-06,
+      "loss": 0.5276,
+      "step": 1018
+    },
+    {
+      "epoch": 3.1942051683633514,
+      "grad_norm": 1.0981642819607513,
+      "learning_rate": 6.951501804867392e-06,
+      "loss": 0.5042,
+      "step": 1019
+    },
+    {
+      "epoch": 3.197337509788567,
+      "grad_norm": 1.1293301389682295,
+      "learning_rate": 6.930658575079705e-06,
+      "loss": 0.5456,
+      "step": 1020
+    },
+    {
+      "epoch": 3.200469851213782,
+      "grad_norm": 1.0968445808904248,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.5079,
+      "step": 1021
+    },
+    {
+      "epoch": 3.2036021926389977,
+      "grad_norm": 1.1042880780334279,
+      "learning_rate": 6.889016348208272e-06,
+      "loss": 0.5247,
+      "step": 1022
+    },
+    {
+      "epoch": 3.206734534064213,
+      "grad_norm": 1.1473219744960321,
+      "learning_rate": 6.868217550710359e-06,
+      "loss": 0.529,
+      "step": 1023
+    },
+    {
+      "epoch": 3.2098668754894284,
+      "grad_norm": 1.1820536568504343,
+      "learning_rate": 6.8474337634427505e-06,
+      "loss": 0.5653,
+      "step": 1024
+    },
+    {
+      "epoch": 3.2129992169146435,
+      "grad_norm": 1.1788578418720204,
+      "learning_rate": 6.826665086019467e-06,
+      "loss": 0.556,
+      "step": 1025
+    },
+    {
+      "epoch": 3.216131558339859,
+      "grad_norm": 1.1275364008712732,
+      "learning_rate": 6.805911617982105e-06,
+      "loss": 0.5211,
+      "step": 1026
+    },
+    {
+      "epoch": 3.219263899765074,
+      "grad_norm": 1.1788948899624612,
+      "learning_rate": 6.785173458799365e-06,
+      "loss": 0.5527,
+      "step": 1027
+    },
+    {
+      "epoch": 3.22239624119029,
+      "grad_norm": 1.190831430517554,
+      "learning_rate": 6.764450707866577e-06,
+      "loss": 0.5569,
+      "step": 1028
+    },
+    {
+      "epoch": 3.225528582615505,
+      "grad_norm": 1.1631264177590235,
+      "learning_rate": 6.743743464505219e-06,
+      "loss": 0.5389,
+      "step": 1029
+    },
+    {
+      "epoch": 3.2286609240407205,
+      "grad_norm": 1.141781399383516,
+      "learning_rate": 6.723051827962445e-06,
+      "loss": 0.5309,
+      "step": 1030
+    },
+    {
+      "epoch": 3.2317932654659356,
+      "grad_norm": 1.179512415938292,
+      "learning_rate": 6.7023758974106065e-06,
+      "loss": 0.525,
+      "step": 1031
+    },
+    {
+      "epoch": 3.234925606891151,
+      "grad_norm": 1.1944956246302587,
+      "learning_rate": 6.681715771946777e-06,
+      "loss": 0.5541,
+      "step": 1032
+    },
+    {
+      "epoch": 3.2380579483163663,
+      "grad_norm": 1.1591633344391639,
+      "learning_rate": 6.661071550592284e-06,
+      "loss": 0.5347,
+      "step": 1033
+    },
+    {
+      "epoch": 3.241190289741582,
+      "grad_norm": 1.12772338651315,
+      "learning_rate": 6.6404433322922146e-06,
+      "loss": 0.5155,
+      "step": 1034
+    },
+    {
+      "epoch": 3.244322631166797,
+      "grad_norm": 1.165589037264499,
+      "learning_rate": 6.619831215914974e-06,
+      "loss": 0.5346,
+      "step": 1035
+    },
+    {
+      "epoch": 3.2474549725920125,
+      "grad_norm": 1.103553776475765,
+      "learning_rate": 6.599235300251778e-06,
+      "loss": 0.5066,
+      "step": 1036
+    },
+    {
+      "epoch": 3.2505873140172277,
+      "grad_norm": 1.151353363986568,
+      "learning_rate": 6.578655684016204e-06,
+      "loss": 0.5456,
+      "step": 1037
+    },
+    {
+      "epoch": 3.2537196554424432,
+      "grad_norm": 1.159828047686325,
+      "learning_rate": 6.558092465843704e-06,
+      "loss": 0.5341,
+      "step": 1038
+    },
+    {
+      "epoch": 3.256851996867659,
+      "grad_norm": 1.1724554056458623,
+      "learning_rate": 6.537545744291134e-06,
+      "loss": 0.5598,
+      "step": 1039
+    },
+    {
+      "epoch": 3.259984338292874,
+      "grad_norm": 1.1868205375108176,
+      "learning_rate": 6.517015617836292e-06,
+      "loss": 0.5404,
+      "step": 1040
+    },
+    {
+      "epoch": 3.263116679718089,
+      "grad_norm": 1.180384610418905,
+      "learning_rate": 6.496502184877426e-06,
+      "loss": 0.5215,
+      "step": 1041
+    },
+    {
+      "epoch": 3.2662490211433046,
+      "grad_norm": 1.205146545691501,
+      "learning_rate": 6.476005543732783e-06,
+      "loss": 0.5314,
+      "step": 1042
+    },
+    {
+      "epoch": 3.26938136256852,
+      "grad_norm": 1.1244830342386452,
+      "learning_rate": 6.45552579264013e-06,
+      "loss": 0.5061,
+      "step": 1043
+    },
+    {
+      "epoch": 3.2725137039937353,
+      "grad_norm": 1.1324469436671378,
+      "learning_rate": 6.4350630297562726e-06,
+      "loss": 0.5411,
+      "step": 1044
+    },
+    {
+      "epoch": 3.2756460454189504,
+      "grad_norm": 1.170084873645457,
+      "learning_rate": 6.414617353156605e-06,
+      "loss": 0.5256,
+      "step": 1045
+    },
+    {
+      "epoch": 3.278778386844166,
+      "grad_norm": 1.154539238112036,
+      "learning_rate": 6.394188860834626e-06,
+      "loss": 0.5353,
+      "step": 1046
+    },
+    {
+      "epoch": 3.2819107282693816,
+      "grad_norm": 1.145934214062239,
+      "learning_rate": 6.373777650701465e-06,
+      "loss": 0.4949,
+      "step": 1047
+    },
+    {
+      "epoch": 3.2850430696945967,
+      "grad_norm": 1.1758118112093967,
+      "learning_rate": 6.35338382058543e-06,
+      "loss": 0.5388,
+      "step": 1048
+    },
+    {
+      "epoch": 3.288175411119812,
+      "grad_norm": 1.1930083102260012,
+      "learning_rate": 6.333007468231521e-06,
+      "loss": 0.5357,
+      "step": 1049
+    },
+    {
+      "epoch": 3.2913077525450274,
+      "grad_norm": 1.2215776268642229,
+      "learning_rate": 6.312648691300975e-06,
+      "loss": 0.5159,
+      "step": 1050
+    },
+    {
+      "epoch": 3.294440093970243,
+      "grad_norm": 1.149266078166264,
+      "learning_rate": 6.292307587370794e-06,
+      "loss": 0.5329,
+      "step": 1051
+    },
+    {
+      "epoch": 3.297572435395458,
+      "grad_norm": 1.188650823733469,
+      "learning_rate": 6.271984253933266e-06,
+      "loss": 0.5607,
+      "step": 1052
+    },
+    {
+      "epoch": 3.3007047768206736,
+      "grad_norm": 1.1743831994156586,
+      "learning_rate": 6.2516787883955165e-06,
+      "loss": 0.5574,
+      "step": 1053
+    },
+    {
+      "epoch": 3.3038371182458888,
+      "grad_norm": 1.1970079452073716,
+      "learning_rate": 6.231391288079027e-06,
+      "loss": 0.5173,
+      "step": 1054
+    },
+    {
+      "epoch": 3.3069694596711043,
+      "grad_norm": 1.122279685545005,
+      "learning_rate": 6.211121850219175e-06,
+      "loss": 0.5361,
+      "step": 1055
+    },
+    {
+      "epoch": 3.3101018010963195,
+      "grad_norm": 1.225293554009653,
+      "learning_rate": 6.1908705719647735e-06,
+      "loss": 0.555,
+      "step": 1056
+    },
+    {
+      "epoch": 3.313234142521535,
+      "grad_norm": 1.1655884451182597,
+      "learning_rate": 6.170637550377587e-06,
+      "loss": 0.5321,
+      "step": 1057
+    },
+    {
+      "epoch": 3.31636648394675,
+      "grad_norm": 1.1850726593613186,
+      "learning_rate": 6.150422882431888e-06,
+      "loss": 0.5158,
+      "step": 1058
+    },
+    {
+      "epoch": 3.3194988253719657,
+      "grad_norm": 1.1741932650006341,
+      "learning_rate": 6.130226665013979e-06,
+      "loss": 0.5278,
+      "step": 1059
+    },
+    {
+      "epoch": 3.322631166797181,
+      "grad_norm": 1.1637151007879059,
+      "learning_rate": 6.110048994921735e-06,
+      "loss": 0.5147,
+      "step": 1060
+    },
+    {
+      "epoch": 3.3257635082223964,
+      "grad_norm": 1.19810414141727,
+      "learning_rate": 6.089889968864122e-06,
+      "loss": 0.5523,
+      "step": 1061
+    },
+    {
+      "epoch": 3.3288958496476115,
+      "grad_norm": 1.1747671263259907,
+      "learning_rate": 6.069749683460765e-06,
+      "loss": 0.5259,
+      "step": 1062
+    },
+    {
+      "epoch": 3.332028191072827,
+      "grad_norm": 1.189733533665569,
+      "learning_rate": 6.049628235241459e-06,
+      "loss": 0.5727,
+      "step": 1063
+    },
+    {
+      "epoch": 3.3351605324980422,
+      "grad_norm": 1.1907858250963501,
+      "learning_rate": 6.029525720645719e-06,
+      "loss": 0.5581,
+      "step": 1064
+    },
+    {
+      "epoch": 3.338292873923258,
+      "grad_norm": 1.1727785316101889,
+      "learning_rate": 6.009442236022307e-06,
+      "loss": 0.5527,
+      "step": 1065
+    },
+    {
+      "epoch": 3.341425215348473,
+      "grad_norm": 1.1473257646366246,
+      "learning_rate": 5.989377877628783e-06,
+      "loss": 0.5354,
+      "step": 1066
+    },
+    {
+      "epoch": 3.3445575567736885,
+      "grad_norm": 1.1904303579624806,
+      "learning_rate": 5.969332741631037e-06,
+      "loss": 0.5299,
+      "step": 1067
+    },
+    {
+      "epoch": 3.3476898981989036,
+      "grad_norm": 1.1467680391867652,
+      "learning_rate": 5.949306924102822e-06,
+      "loss": 0.5347,
+      "step": 1068
+    },
+    {
+      "epoch": 3.350822239624119,
+      "grad_norm": 1.164592432138905,
+      "learning_rate": 5.929300521025304e-06,
+      "loss": 0.5295,
+      "step": 1069
+    },
+    {
+      "epoch": 3.3539545810493343,
+      "grad_norm": 1.183763669072413,
+      "learning_rate": 5.9093136282866014e-06,
+      "loss": 0.5309,
+      "step": 1070
+    },
+    {
+      "epoch": 3.35708692247455,
+      "grad_norm": 1.1341809428622889,
+      "learning_rate": 5.889346341681319e-06,
+      "loss": 0.5057,
+      "step": 1071
+    },
+    {
+      "epoch": 3.360219263899765,
+      "grad_norm": 1.2114438092248463,
+      "learning_rate": 5.869398756910095e-06,
+      "loss": 0.5285,
+      "step": 1072
+    },
+    {
+      "epoch": 3.3633516053249806,
+      "grad_norm": 1.1899159606732028,
+      "learning_rate": 5.8494709695791364e-06,
+      "loss": 0.515,
+      "step": 1073
+    },
+    {
+      "epoch": 3.3664839467501957,
+      "grad_norm": 1.1585091379070802,
+      "learning_rate": 5.829563075199753e-06,
+      "loss": 0.5215,
+      "step": 1074
+    },
+    {
+      "epoch": 3.3696162881754113,
+      "grad_norm": 1.1554751170108555,
+      "learning_rate": 5.809675169187936e-06,
+      "loss": 0.5276,
+      "step": 1075
+    },
+    {
+      "epoch": 3.3727486296006264,
+      "grad_norm": 1.1531221192759264,
+      "learning_rate": 5.789807346863849e-06,
+      "loss": 0.5343,
+      "step": 1076
+    },
+    {
+      "epoch": 3.375880971025842,
+      "grad_norm": 1.2264179028180235,
+      "learning_rate": 5.769959703451413e-06,
+      "loss": 0.5331,
+      "step": 1077
+    },
+    {
+      "epoch": 3.379013312451057,
+      "grad_norm": 1.1909513224731456,
+      "learning_rate": 5.750132334077827e-06,
+      "loss": 0.5221,
+      "step": 1078
+    },
+    {
+      "epoch": 3.3821456538762726,
+      "grad_norm": 1.1410796381664094,
+      "learning_rate": 5.730325333773122e-06,
+      "loss": 0.5236,
+      "step": 1079
+    },
+    {
+      "epoch": 3.3852779953014878,
+      "grad_norm": 1.165393643113191,
+      "learning_rate": 5.710538797469707e-06,
+      "loss": 0.5363,
+      "step": 1080
+    },
+    {
+      "epoch": 3.3884103367267033,
+      "grad_norm": 1.109034591095622,
+      "learning_rate": 5.690772820001894e-06,
+      "loss": 0.5129,
+      "step": 1081
+    },
+    {
+      "epoch": 3.3915426781519185,
+      "grad_norm": 1.1479442023962463,
+      "learning_rate": 5.671027496105477e-06,
+      "loss": 0.5767,
+      "step": 1082
+    },
+    {
+      "epoch": 3.394675019577134,
+      "grad_norm": 1.1983194565401882,
+      "learning_rate": 5.651302920417252e-06,
+      "loss": 0.5233,
+      "step": 1083
+    },
+    {
+      "epoch": 3.397807361002349,
+      "grad_norm": 1.1247461153791098,
+      "learning_rate": 5.63159918747457e-06,
+      "loss": 0.531,
+      "step": 1084
+    },
+    {
+      "epoch": 3.4009397024275647,
+      "grad_norm": 1.1645414490351773,
+      "learning_rate": 5.611916391714887e-06,
+      "loss": 0.5356,
+      "step": 1085
+    },
+    {
+      "epoch": 3.40407204385278,
+      "grad_norm": 1.1955131839630717,
+      "learning_rate": 5.592254627475313e-06,
+      "loss": 0.5181,
+      "step": 1086
+    },
+    {
+      "epoch": 3.4072043852779954,
+      "grad_norm": 1.1669277156801512,
+      "learning_rate": 5.572613988992157e-06,
+      "loss": 0.5505,
+      "step": 1087
+    },
+    {
+      "epoch": 3.4103367267032105,
+      "grad_norm": 1.1969803225344355,
+      "learning_rate": 5.552994570400465e-06,
+      "loss": 0.5279,
+      "step": 1088
+    },
+    {
+      "epoch": 3.413469068128426,
+      "grad_norm": 1.149976867281222,
+      "learning_rate": 5.5333964657335894e-06,
+      "loss": 0.5176,
+      "step": 1089
+    },
+    {
+      "epoch": 3.4166014095536412,
+      "grad_norm": 1.1530538912791253,
+      "learning_rate": 5.513819768922723e-06,
+      "loss": 0.529,
+      "step": 1090
+    },
+    {
+      "epoch": 3.419733750978857,
+      "grad_norm": 1.1611367505386145,
+      "learning_rate": 5.49426457379646e-06,
+      "loss": 0.5045,
+      "step": 1091
+    },
+    {
+      "epoch": 3.422866092404072,
+      "grad_norm": 1.157629907018517,
+      "learning_rate": 5.474730974080327e-06,
+      "loss": 0.5384,
+      "step": 1092
+    },
+    {
+      "epoch": 3.4259984338292875,
+      "grad_norm": 1.1436433927685985,
+      "learning_rate": 5.455219063396356e-06,
+      "loss": 0.5164,
+      "step": 1093
+    },
+    {
+      "epoch": 3.4291307752545026,
+      "grad_norm": 1.1624946647803467,
+      "learning_rate": 5.4357289352626284e-06,
+      "loss": 0.518,
+      "step": 1094
+    },
+    {
+      "epoch": 3.432263116679718,
+      "grad_norm": 1.2059544843365895,
+      "learning_rate": 5.416260683092814e-06,
+      "loss": 0.5413,
+      "step": 1095
+    },
+    {
+      "epoch": 3.4353954581049333,
+      "grad_norm": 1.1848557963338198,
+      "learning_rate": 5.3968144001957445e-06,
+      "loss": 0.556,
+      "step": 1096
+    },
+    {
+      "epoch": 3.438527799530149,
+      "grad_norm": 1.2267248653375662,
+      "learning_rate": 5.377390179774951e-06,
+      "loss": 0.5735,
+      "step": 1097
+    },
+    {
+      "epoch": 3.441660140955364,
+      "grad_norm": 1.219382842483644,
+      "learning_rate": 5.357988114928221e-06,
+      "loss": 0.5224,
+      "step": 1098
+    },
+    {
+      "epoch": 3.4447924823805796,
+      "grad_norm": 1.2019925545507635,
+      "learning_rate": 5.338608298647161e-06,
+      "loss": 0.5188,
+      "step": 1099
+    },
+    {
+      "epoch": 3.4479248238057947,
+      "grad_norm": 1.2054098999261758,
+      "learning_rate": 5.319250823816732e-06,
+      "loss": 0.5603,
+      "step": 1100
+    },
+    {
+      "epoch": 3.4510571652310102,
+      "grad_norm": 1.149167269524483,
+      "learning_rate": 5.299915783214816e-06,
+      "loss": 0.5103,
+      "step": 1101
+    },
+    {
+      "epoch": 3.4541895066562254,
+      "grad_norm": 1.190224838500585,
+      "learning_rate": 5.28060326951178e-06,
+      "loss": 0.5445,
+      "step": 1102
+    },
+    {
+      "epoch": 3.457321848081441,
+      "grad_norm": 1.145339136200069,
+      "learning_rate": 5.2613133752700145e-06,
+      "loss": 0.5432,
+      "step": 1103
+    },
+    {
+      "epoch": 3.460454189506656,
+      "grad_norm": 1.1801121893769682,
+      "learning_rate": 5.242046192943503e-06,
+      "loss": 0.5321,
+      "step": 1104
+    },
+    {
+      "epoch": 3.4635865309318716,
+      "grad_norm": 1.196242124148136,
+      "learning_rate": 5.22280181487737e-06,
+      "loss": 0.5097,
+      "step": 1105
+    },
+    {
+      "epoch": 3.4667188723570868,
+      "grad_norm": 1.1558254840616595,
+      "learning_rate": 5.2035803333074435e-06,
+      "loss": 0.515,
+      "step": 1106
+    },
+    {
+      "epoch": 3.4698512137823023,
+      "grad_norm": 1.1436458411967738,
+      "learning_rate": 5.184381840359815e-06,
+      "loss": 0.5004,
+      "step": 1107
+    },
+    {
+      "epoch": 3.4729835552075174,
+      "grad_norm": 1.137825928298088,
+      "learning_rate": 5.1652064280503884e-06,
+      "loss": 0.5164,
+      "step": 1108
+    },
+    {
+      "epoch": 3.476115896632733,
+      "grad_norm": 1.1431285887634195,
+      "learning_rate": 5.146054188284443e-06,
+      "loss": 0.5324,
+      "step": 1109
+    },
+    {
+      "epoch": 3.479248238057948,
+      "grad_norm": 1.1848905862756827,
+      "learning_rate": 5.126925212856202e-06,
+      "loss": 0.541,
+      "step": 1110
+    },
+    {
+      "epoch": 3.4823805794831637,
+      "grad_norm": 1.15107617387345,
+      "learning_rate": 5.107819593448385e-06,
+      "loss": 0.5049,
+      "step": 1111
+    },
+    {
+      "epoch": 3.485512920908379,
+      "grad_norm": 1.1324253201763683,
+      "learning_rate": 5.088737421631767e-06,
+      "loss": 0.5263,
+      "step": 1112
+    },
+    {
+      "epoch": 3.4886452623335944,
+      "grad_norm": 1.1791937767344005,
+      "learning_rate": 5.069678788864742e-06,
+      "loss": 0.5301,
+      "step": 1113
+    },
+    {
+      "epoch": 3.4917776037588095,
+      "grad_norm": 1.177125639354513,
+      "learning_rate": 5.050643786492888e-06,
+      "loss": 0.5299,
+      "step": 1114
+    },
+    {
+      "epoch": 3.494909945184025,
+      "grad_norm": 1.12975187321893,
+      "learning_rate": 5.031632505748516e-06,
+      "loss": 0.5191,
+      "step": 1115
+    },
+    {
+      "epoch": 3.49804228660924,
+      "grad_norm": 1.1702301585282073,
+      "learning_rate": 5.012645037750254e-06,
+      "loss": 0.4911,
+      "step": 1116
+    },
+    {
+      "epoch": 3.501174628034456,
+      "grad_norm": 1.1898809996788224,
+      "learning_rate": 4.993681473502586e-06,
+      "loss": 0.5846,
+      "step": 1117
+    },
+    {
+      "epoch": 3.5043069694596714,
+      "grad_norm": 1.1772803798080353,
+      "learning_rate": 4.97474190389544e-06,
+      "loss": 0.5449,
+      "step": 1118
+    },
+    {
+      "epoch": 3.5074393108848865,
+      "grad_norm": 1.1458679793457238,
+      "learning_rate": 4.955826419703736e-06,
+      "loss": 0.527,
+      "step": 1119
+    },
+    {
+      "epoch": 3.5105716523101016,
+      "grad_norm": 1.197898331863354,
+      "learning_rate": 4.936935111586953e-06,
+      "loss": 0.5279,
+      "step": 1120
+    },
+    {
+      "epoch": 3.513703993735317,
+      "grad_norm": 1.1490696601995432,
+      "learning_rate": 4.918068070088708e-06,
+      "loss": 0.5479,
+      "step": 1121
+    },
+    {
+      "epoch": 3.5168363351605327,
+      "grad_norm": 1.1338863485542614,
+      "learning_rate": 4.899225385636293e-06,
+      "loss": 0.5143,
+      "step": 1122
+    },
+    {
+      "epoch": 3.519968676585748,
+      "grad_norm": 1.1180638940506065,
+      "learning_rate": 4.880407148540275e-06,
+      "loss": 0.5018,
+      "step": 1123
+    },
+    {
+      "epoch": 3.523101018010963,
+      "grad_norm": 1.1649197234392008,
+      "learning_rate": 4.8616134489940455e-06,
+      "loss": 0.5086,
+      "step": 1124
+    },
+    {
+      "epoch": 3.5262333594361785,
+      "grad_norm": 1.175823733278329,
+      "learning_rate": 4.8428443770733835e-06,
+      "loss": 0.5128,
+      "step": 1125
+    },
+    {
+      "epoch": 3.529365700861394,
+      "grad_norm": 1.1338414552899265,
+      "learning_rate": 4.824100022736036e-06,
+      "loss": 0.4854,
+      "step": 1126
+    },
+    {
+      "epoch": 3.5324980422866092,
+      "grad_norm": 1.208715108777811,
+      "learning_rate": 4.805380475821287e-06,
+      "loss": 0.562,
+      "step": 1127
+    },
+    {
+      "epoch": 3.5356303837118244,
+      "grad_norm": 1.2832853162474576,
+      "learning_rate": 4.786685826049508e-06,
+      "loss": 0.5227,
+      "step": 1128
+    },
+    {
+      "epoch": 3.53876272513704,
+      "grad_norm": 1.20463188289643,
+      "learning_rate": 4.768016163021754e-06,
+      "loss": 0.5343,
+      "step": 1129
+    },
+    {
+      "epoch": 3.5418950665622555,
+      "grad_norm": 1.1932857110540964,
+      "learning_rate": 4.749371576219317e-06,
+      "loss": 0.4933,
+      "step": 1130
+    },
+    {
+      "epoch": 3.5450274079874706,
+      "grad_norm": 1.1785435772368695,
+      "learning_rate": 4.730752155003303e-06,
+      "loss": 0.5262,
+      "step": 1131
+    },
+    {
+      "epoch": 3.5481597494126857,
+      "grad_norm": 1.1881352282816962,
+      "learning_rate": 4.7121579886142065e-06,
+      "loss": 0.5564,
+      "step": 1132
+    },
+    {
+      "epoch": 3.5512920908379013,
+      "grad_norm": 1.1608651788367013,
+      "learning_rate": 4.693589166171466e-06,
+      "loss": 0.5287,
+      "step": 1133
+    },
+    {
+      "epoch": 3.554424432263117,
+      "grad_norm": 1.2094527888449869,
+      "learning_rate": 4.675045776673067e-06,
+      "loss": 0.5463,
+      "step": 1134
+    },
+    {
+      "epoch": 3.557556773688332,
+      "grad_norm": 1.1971780880370575,
+      "learning_rate": 4.65652790899508e-06,
+      "loss": 0.5117,
+      "step": 1135
+    },
+    {
+      "epoch": 3.560689115113547,
+      "grad_norm": 1.1896379190718653,
+      "learning_rate": 4.638035651891266e-06,
+      "loss": 0.543,
+      "step": 1136
+    },
+    {
+      "epoch": 3.5638214565387627,
+      "grad_norm": 1.123251870010005,
+      "learning_rate": 4.619569093992636e-06,
+      "loss": 0.5263,
+      "step": 1137
+    },
+    {
+      "epoch": 3.5669537979639783,
+      "grad_norm": 1.1685045536154566,
+      "learning_rate": 4.601128323807023e-06,
+      "loss": 0.5341,
+      "step": 1138
+    },
+    {
+      "epoch": 3.5700861393891934,
+      "grad_norm": 1.1479704004071765,
+      "learning_rate": 4.582713429718667e-06,
+      "loss": 0.5503,
+      "step": 1139
+    },
+    {
+      "epoch": 3.5732184808144085,
+      "grad_norm": 1.138427168032927,
+      "learning_rate": 4.56432449998779e-06,
+      "loss": 0.492,
+      "step": 1140
+    },
+    {
+      "epoch": 3.576350822239624,
+      "grad_norm": 1.2372766348911877,
+      "learning_rate": 4.545961622750158e-06,
+      "loss": 0.5512,
+      "step": 1141
+    },
+    {
+      "epoch": 3.5794831636648397,
+      "grad_norm": 1.1774919589788904,
+      "learning_rate": 4.527624886016686e-06,
+      "loss": 0.5398,
+      "step": 1142
+    },
+    {
+      "epoch": 3.5826155050900548,
+      "grad_norm": 1.1693438413445578,
+      "learning_rate": 4.509314377672987e-06,
+      "loss": 0.5444,
+      "step": 1143
+    },
+    {
+      "epoch": 3.58574784651527,
+      "grad_norm": 1.206456317247258,
+      "learning_rate": 4.491030185478976e-06,
+      "loss": 0.5362,
+      "step": 1144
+    },
+    {
+      "epoch": 3.5888801879404855,
+      "grad_norm": 1.1483671630028152,
+      "learning_rate": 4.472772397068431e-06,
+      "loss": 0.5115,
+      "step": 1145
+    },
+    {
+      "epoch": 3.592012529365701,
+      "grad_norm": 1.2246520853364462,
+      "learning_rate": 4.454541099948585e-06,
+      "loss": 0.5305,
+      "step": 1146
+    },
+    {
+      "epoch": 3.595144870790916,
+      "grad_norm": 1.1831900976113294,
+      "learning_rate": 4.436336381499701e-06,
+      "loss": 0.5413,
+      "step": 1147
+    },
+    {
+      "epoch": 3.5982772122161313,
+      "grad_norm": 1.2043230578018518,
+      "learning_rate": 4.418158328974647e-06,
+      "loss": 0.5196,
+      "step": 1148
+    },
+    {
+      "epoch": 3.601409553641347,
+      "grad_norm": 1.1773163602914556,
+      "learning_rate": 4.40000702949849e-06,
+      "loss": 0.5001,
+      "step": 1149
+    },
+    {
+      "epoch": 3.6045418950665624,
+      "grad_norm": 1.1478622164586656,
+      "learning_rate": 4.381882570068079e-06,
+      "loss": 0.5303,
+      "step": 1150
+    },
+    {
+      "epoch": 3.6076742364917775,
+      "grad_norm": 1.191685305824786,
+      "learning_rate": 4.3637850375516085e-06,
+      "loss": 0.5137,
+      "step": 1151
+    },
+    {
+      "epoch": 3.610806577916993,
+      "grad_norm": 1.181732648263047,
+      "learning_rate": 4.345714518688226e-06,
+      "loss": 0.5246,
+      "step": 1152
+    },
+    {
+      "epoch": 3.6139389193422082,
+      "grad_norm": 1.1923764986284755,
+      "learning_rate": 4.327671100087605e-06,
+      "loss": 0.5327,
+      "step": 1153
+    },
+    {
+      "epoch": 3.617071260767424,
+      "grad_norm": 1.20809690015103,
+      "learning_rate": 4.3096548682295304e-06,
+      "loss": 0.5498,
+      "step": 1154
+    },
+    {
+      "epoch": 3.620203602192639,
+      "grad_norm": 1.1491784685338666,
+      "learning_rate": 4.291665909463477e-06,
+      "loss": 0.5041,
+      "step": 1155
+    },
+    {
+      "epoch": 3.6233359436178545,
+      "grad_norm": 1.1751280305249536,
+      "learning_rate": 4.273704310008213e-06,
+      "loss": 0.5229,
+      "step": 1156
+    },
+    {
+      "epoch": 3.6264682850430696,
+      "grad_norm": 1.1175615334495055,
+      "learning_rate": 4.255770155951374e-06,
+      "loss": 0.5215,
+      "step": 1157
+    },
+    {
+      "epoch": 3.629600626468285,
+      "grad_norm": 1.155032398758409,
+      "learning_rate": 4.237863533249057e-06,
+      "loss": 0.5344,
+      "step": 1158
+    },
+    {
+      "epoch": 3.6327329678935003,
+      "grad_norm": 1.1988336359099494,
+      "learning_rate": 4.21998452772539e-06,
+      "loss": 0.5579,
+      "step": 1159
+    },
+    {
+      "epoch": 3.635865309318716,
+      "grad_norm": 1.1698366279218717,
+      "learning_rate": 4.202133225072153e-06,
+      "loss": 0.5288,
+      "step": 1160
+    },
+    {
+      "epoch": 3.638997650743931,
+      "grad_norm": 1.1286869148862564,
+      "learning_rate": 4.1843097108483465e-06,
+      "loss": 0.5245,
+      "step": 1161
+    },
+    {
+      "epoch": 3.6421299921691466,
+      "grad_norm": 1.1861981743925876,
+      "learning_rate": 4.16651407047977e-06,
+      "loss": 0.5391,
+      "step": 1162
+    },
+    {
+      "epoch": 3.6452623335943617,
+      "grad_norm": 1.2269024042403185,
+      "learning_rate": 4.148746389258645e-06,
+      "loss": 0.5417,
+      "step": 1163
+    },
+    {
+      "epoch": 3.6483946750195773,
+      "grad_norm": 1.2098949537700534,
+      "learning_rate": 4.1310067523431795e-06,
+      "loss": 0.5187,
+      "step": 1164
+    },
+    {
+      "epoch": 3.6515270164447924,
+      "grad_norm": 1.2311237257384957,
+      "learning_rate": 4.113295244757171e-06,
+      "loss": 0.5112,
+      "step": 1165
+    },
+    {
+      "epoch": 3.654659357870008,
+      "grad_norm": 1.1381539159403147,
+      "learning_rate": 4.0956119513896005e-06,
+      "loss": 0.5065,
+      "step": 1166
+    },
+    {
+      "epoch": 3.657791699295223,
+      "grad_norm": 1.1621884247268668,
+      "learning_rate": 4.077956956994212e-06,
+      "loss": 0.511,
+      "step": 1167
+    },
+    {
+      "epoch": 3.6609240407204386,
+      "grad_norm": 1.168551137549956,
+      "learning_rate": 4.060330346189125e-06,
+      "loss": 0.5177,
+      "step": 1168
+    },
+    {
+      "epoch": 3.6640563821456538,
+      "grad_norm": 1.210641000943774,
+      "learning_rate": 4.042732203456413e-06,
+      "loss": 0.5197,
+      "step": 1169
+    },
+    {
+      "epoch": 3.6671887235708693,
+      "grad_norm": 1.1705201776756187,
+      "learning_rate": 4.0251626131417135e-06,
+      "loss": 0.5467,
+      "step": 1170
+    },
+    {
+      "epoch": 3.6703210649960845,
+      "grad_norm": 1.1690543008503043,
+      "learning_rate": 4.00762165945381e-06,
+      "loss": 0.5404,
+      "step": 1171
+    },
+    {
+      "epoch": 3.6734534064213,
+      "grad_norm": 1.1598971397306825,
+      "learning_rate": 3.990109426464237e-06,
+      "loss": 0.5178,
+      "step": 1172
+    },
+    {
+      "epoch": 3.676585747846515,
+      "grad_norm": 1.1616293085138196,
+      "learning_rate": 3.972625998106876e-06,
+      "loss": 0.5435,
+      "step": 1173
+    },
+    {
+      "epoch": 3.6797180892717307,
+      "grad_norm": 1.1407718103960551,
+      "learning_rate": 3.955171458177551e-06,
+      "loss": 0.5406,
+      "step": 1174
+    },
+    {
+      "epoch": 3.682850430696946,
+      "grad_norm": 1.1669175281894106,
+      "learning_rate": 3.937745890333623e-06,
+      "loss": 0.5251,
+      "step": 1175
+    },
+    {
+      "epoch": 3.6859827721221614,
+      "grad_norm": 1.1257651461718745,
+      "learning_rate": 3.9203493780935935e-06,
+      "loss": 0.5345,
+      "step": 1176
+    },
+    {
+      "epoch": 3.6891151135473765,
+      "grad_norm": 1.1540314221528836,
+      "learning_rate": 3.902982004836709e-06,
+      "loss": 0.5351,
+      "step": 1177
+    },
+    {
+      "epoch": 3.692247454972592,
+      "grad_norm": 1.1524838421819237,
+      "learning_rate": 3.885643853802557e-06,
+      "loss": 0.5488,
+      "step": 1178
+    },
+    {
+      "epoch": 3.6953797963978072,
+      "grad_norm": 1.188466802503505,
+      "learning_rate": 3.868335008090664e-06,
+      "loss": 0.5142,
+      "step": 1179
+    },
+    {
+      "epoch": 3.698512137823023,
+      "grad_norm": 1.138912921077403,
+      "learning_rate": 3.851055550660098e-06,
+      "loss": 0.5269,
+      "step": 1180
+    },
+    {
+      "epoch": 3.701644479248238,
+      "grad_norm": 1.1870550813670917,
+      "learning_rate": 3.8338055643290815e-06,
+      "loss": 0.527,
+      "step": 1181
+    },
+    {
+      "epoch": 3.7047768206734535,
+      "grad_norm": 1.1829270162712773,
+      "learning_rate": 3.8165851317745705e-06,
+      "loss": 0.526,
+      "step": 1182
+    },
+    {
+      "epoch": 3.7079091620986686,
+      "grad_norm": 1.2162157673701168,
+      "learning_rate": 3.799394335531891e-06,
+      "loss": 0.5644,
+      "step": 1183
+    },
+    {
+      "epoch": 3.711041503523884,
+      "grad_norm": 1.1789615458756362,
+      "learning_rate": 3.782233257994309e-06,
+      "loss": 0.5369,
+      "step": 1184
+    },
+    {
+      "epoch": 3.7141738449490993,
+      "grad_norm": 1.1676281601812897,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.5243,
+      "step": 1185
+    },
+    {
+      "epoch": 3.717306186374315,
+      "grad_norm": 1.1571405119696987,
+      "learning_rate": 3.7480005878949645e-06,
+      "loss": 0.5584,
+      "step": 1186
+    },
+    {
+      "epoch": 3.72043852779953,
+      "grad_norm": 1.1512138914116103,
+      "learning_rate": 3.730929159405984e-06,
+      "loss": 0.5272,
+      "step": 1187
+    },
+    {
+      "epoch": 3.7235708692247456,
+      "grad_norm": 1.1290612876998436,
+      "learning_rate": 3.713887777766888e-06,
+      "loss": 0.5212,
+      "step": 1188
+    },
+    {
+      "epoch": 3.7267032106499607,
+      "grad_norm": 1.1615587465761752,
+      "learning_rate": 3.69687652465482e-06,
+      "loss": 0.5271,
+      "step": 1189
+    },
+    {
+      "epoch": 3.7298355520751763,
+      "grad_norm": 1.165644217808434,
+      "learning_rate": 3.679895481602529e-06,
+      "loss": 0.5268,
+      "step": 1190
+    },
+    {
+      "epoch": 3.7329678935003914,
+      "grad_norm": 1.1460726933346908,
+      "learning_rate": 3.6629447299979747e-06,
+      "loss": 0.5345,
+      "step": 1191
+    },
+    {
+      "epoch": 3.736100234925607,
+      "grad_norm": 1.1701380925076978,
+      "learning_rate": 3.64602435108392e-06,
+      "loss": 0.5482,
+      "step": 1192
+    },
+    {
+      "epoch": 3.7392325763508225,
+      "grad_norm": 1.1069644314839941,
+      "learning_rate": 3.629134425957567e-06,
+      "loss": 0.4975,
+      "step": 1193
+    },
+    {
+      "epoch": 3.7423649177760376,
+      "grad_norm": 1.1742604377155452,
+      "learning_rate": 3.612275035570155e-06,
+      "loss": 0.5382,
+      "step": 1194
+    },
+    {
+      "epoch": 3.7454972592012528,
+      "grad_norm": 1.1594884695567513,
+      "learning_rate": 3.595446260726576e-06,
+      "loss": 0.5276,
+      "step": 1195
+    },
+    {
+      "epoch": 3.7486296006264683,
+      "grad_norm": 1.2009203643761532,
+      "learning_rate": 3.578648182084975e-06,
+      "loss": 0.5377,
+      "step": 1196
+    },
+    {
+      "epoch": 3.751761942051684,
+      "grad_norm": 1.170884036670357,
+      "learning_rate": 3.561880880156384e-06,
+      "loss": 0.5576,
+      "step": 1197
+    },
+    {
+      "epoch": 3.754894283476899,
+      "grad_norm": 1.1959487905145205,
+      "learning_rate": 3.5451444353043263e-06,
+      "loss": 0.5373,
+      "step": 1198
+    },
+    {
+      "epoch": 3.758026624902114,
+      "grad_norm": 1.1796481257439184,
+      "learning_rate": 3.5284389277444296e-06,
+      "loss": 0.5313,
+      "step": 1199
+    },
+    {
+      "epoch": 3.7611589663273297,
+      "grad_norm": 1.2094185606216385,
+      "learning_rate": 3.511764437544036e-06,
+      "loss": 0.5162,
+      "step": 1200
+    },
+    {
+      "epoch": 3.7642913077525453,
+      "grad_norm": 1.1754904046598893,
+      "learning_rate": 3.4951210446218385e-06,
+      "loss": 0.5529,
+      "step": 1201
+    },
+    {
+      "epoch": 3.7674236491777604,
+      "grad_norm": 1.153840764724347,
+      "learning_rate": 3.478508828747471e-06,
+      "loss": 0.4987,
+      "step": 1202
+    },
+    {
+      "epoch": 3.7705559906029755,
+      "grad_norm": 1.1333187974311192,
+      "learning_rate": 3.4619278695411495e-06,
+      "loss": 0.5005,
+      "step": 1203
+    },
+    {
+      "epoch": 3.773688332028191,
+      "grad_norm": 1.1455993988251625,
+      "learning_rate": 3.445378246473279e-06,
+      "loss": 0.5115,
+      "step": 1204
+    },
+    {
+      "epoch": 3.7768206734534067,
+      "grad_norm": 1.1749646759832453,
+      "learning_rate": 3.4288600388640714e-06,
+      "loss": 0.5241,
+      "step": 1205
+    },
+    {
+      "epoch": 3.779953014878622,
+      "grad_norm": 1.159353327088895,
+      "learning_rate": 3.4123733258831727e-06,
+      "loss": 0.5221,
+      "step": 1206
+    },
+    {
+      "epoch": 3.783085356303837,
+      "grad_norm": 1.1770098803771174,
+      "learning_rate": 3.3959181865492798e-06,
+      "loss": 0.536,
+      "step": 1207
+    },
+    {
+      "epoch": 3.7862176977290525,
+      "grad_norm": 1.173049503757697,
+      "learning_rate": 3.3794946997297527e-06,
+      "loss": 0.5423,
+      "step": 1208
+    },
+    {
+      "epoch": 3.789350039154268,
+      "grad_norm": 1.2771010631863746,
+      "learning_rate": 3.3631029441402497e-06,
+      "loss": 0.5587,
+      "step": 1209
+    },
+    {
+      "epoch": 3.792482380579483,
+      "grad_norm": 1.193386115742984,
+      "learning_rate": 3.3467429983443477e-06,
+      "loss": 0.5277,
+      "step": 1210
+    },
+    {
+      "epoch": 3.7956147220046983,
+      "grad_norm": 1.1915512034829279,
+      "learning_rate": 3.330414940753164e-06,
+      "loss": 0.5583,
+      "step": 1211
+    },
+    {
+      "epoch": 3.798747063429914,
+      "grad_norm": 1.1896725640667438,
+      "learning_rate": 3.3141188496249755e-06,
+      "loss": 0.5141,
+      "step": 1212
+    },
+    {
+      "epoch": 3.8018794048551294,
+      "grad_norm": 1.1687306108915543,
+      "learning_rate": 3.2978548030648504e-06,
+      "loss": 0.5443,
+      "step": 1213
+    },
+    {
+      "epoch": 3.8050117462803446,
+      "grad_norm": 1.1857286795537076,
+      "learning_rate": 3.2816228790242723e-06,
+      "loss": 0.5472,
+      "step": 1214
+    },
+    {
+      "epoch": 3.8081440877055597,
+      "grad_norm": 1.138346768843636,
+      "learning_rate": 3.2654231553007665e-06,
+      "loss": 0.549,
+      "step": 1215
+    },
+    {
+      "epoch": 3.8112764291307752,
+      "grad_norm": 1.1444752892169308,
+      "learning_rate": 3.249255709537521e-06,
+      "loss": 0.522,
+      "step": 1216
+    },
+    {
+      "epoch": 3.814408770555991,
+      "grad_norm": 1.18144246751696,
+      "learning_rate": 3.233120619223021e-06,
+      "loss": 0.5303,
+      "step": 1217
+    },
+    {
+      "epoch": 3.817541111981206,
+      "grad_norm": 1.2096091122117285,
+      "learning_rate": 3.2170179616906818e-06,
+      "loss": 0.5207,
+      "step": 1218
+    },
+    {
+      "epoch": 3.820673453406421,
+      "grad_norm": 1.1530748784567288,
+      "learning_rate": 3.200947814118467e-06,
+      "loss": 0.5376,
+      "step": 1219
+    },
+    {
+      "epoch": 3.8238057948316366,
+      "grad_norm": 1.1215171003312445,
+      "learning_rate": 3.1849102535285283e-06,
+      "loss": 0.5082,
+      "step": 1220
+    },
+    {
+      "epoch": 3.826938136256852,
+      "grad_norm": 1.1703733058989874,
+      "learning_rate": 3.1689053567868278e-06,
+      "loss": 0.5337,
+      "step": 1221
+    },
+    {
+      "epoch": 3.8300704776820673,
+      "grad_norm": 1.130262675345898,
+      "learning_rate": 3.1529332006027813e-06,
+      "loss": 0.5404,
+      "step": 1222
+    },
+    {
+      "epoch": 3.8332028191072824,
+      "grad_norm": 1.1669511760006823,
+      "learning_rate": 3.1369938615288707e-06,
+      "loss": 0.5311,
+      "step": 1223
+    },
+    {
+      "epoch": 3.836335160532498,
+      "grad_norm": 1.1656001584283648,
+      "learning_rate": 3.1210874159603044e-06,
+      "loss": 0.5422,
+      "step": 1224
+    },
+    {
+      "epoch": 3.8394675019577136,
+      "grad_norm": 1.1747394216229974,
+      "learning_rate": 3.1052139401346216e-06,
+      "loss": 0.5132,
+      "step": 1225
+    },
+    {
+      "epoch": 3.8425998433829287,
+      "grad_norm": 1.1495579361244552,
+      "learning_rate": 3.089373510131354e-06,
+      "loss": 0.493,
+      "step": 1226
+    },
+    {
+      "epoch": 3.845732184808144,
+      "grad_norm": 1.180793820104102,
+      "learning_rate": 3.0735662018716437e-06,
+      "loss": 0.5291,
+      "step": 1227
+    },
+    {
+      "epoch": 3.8488645262333594,
+      "grad_norm": 1.124911552092485,
+      "learning_rate": 3.057792091117889e-06,
+      "loss": 0.5189,
+      "step": 1228
+    },
+    {
+      "epoch": 3.851996867658575,
+      "grad_norm": 1.120771756001303,
+      "learning_rate": 3.042051253473366e-06,
+      "loss": 0.4857,
+      "step": 1229
+    },
+    {
+      "epoch": 3.85512920908379,
+      "grad_norm": 1.1265839282142127,
+      "learning_rate": 3.026343764381887e-06,
+      "loss": 0.5188,
+      "step": 1230
+    },
+    {
+      "epoch": 3.858261550509005,
+      "grad_norm": 1.153903476616413,
+      "learning_rate": 3.0106696991274278e-06,
+      "loss": 0.518,
+      "step": 1231
+    },
+    {
+      "epoch": 3.861393891934221,
+      "grad_norm": 1.1653054593485324,
+      "learning_rate": 2.995029132833768e-06,
+      "loss": 0.5413,
+      "step": 1232
+    },
+    {
+      "epoch": 3.8645262333594363,
+      "grad_norm": 1.1219932990648924,
+      "learning_rate": 2.9794221404641233e-06,
+      "loss": 0.5584,
+      "step": 1233
+    },
+    {
+      "epoch": 3.8676585747846515,
+      "grad_norm": 1.1764250161654142,
+      "learning_rate": 2.9638487968208028e-06,
+      "loss": 0.5485,
+      "step": 1234
+    },
+    {
+      "epoch": 3.870790916209867,
+      "grad_norm": 1.1756895255385884,
+      "learning_rate": 2.9483091765448426e-06,
+      "loss": 0.5223,
+      "step": 1235
+    },
+    {
+      "epoch": 3.873923257635082,
+      "grad_norm": 1.1493491848055564,
+      "learning_rate": 2.9328033541156363e-06,
+      "loss": 0.5137,
+      "step": 1236
+    },
+    {
+      "epoch": 3.8770555990602977,
+      "grad_norm": 1.1435017702746313,
+      "learning_rate": 2.9173314038505986e-06,
+      "loss": 0.5404,
+      "step": 1237
+    },
+    {
+      "epoch": 3.880187940485513,
+      "grad_norm": 1.1617452479406913,
+      "learning_rate": 2.901893399904797e-06,
+      "loss": 0.5315,
+      "step": 1238
+    },
+    {
+      "epoch": 3.8833202819107284,
+      "grad_norm": 1.2030832962067164,
+      "learning_rate": 2.886489416270598e-06,
+      "loss": 0.5385,
+      "step": 1239
+    },
+    {
+      "epoch": 3.8864526233359435,
+      "grad_norm": 1.1328735216805788,
+      "learning_rate": 2.871119526777315e-06,
+      "loss": 0.5199,
+      "step": 1240
+    },
+    {
+      "epoch": 3.889584964761159,
+      "grad_norm": 1.1928234546236907,
+      "learning_rate": 2.855783805090846e-06,
+      "loss": 0.5258,
+      "step": 1241
+    },
+    {
+      "epoch": 3.8927173061863742,
+      "grad_norm": 1.208100129151956,
+      "learning_rate": 2.8404823247133373e-06,
+      "loss": 0.5168,
+      "step": 1242
+    },
+    {
+      "epoch": 3.89584964761159,
+      "grad_norm": 1.1414115708828445,
+      "learning_rate": 2.825215158982809e-06,
+      "loss": 0.5429,
+      "step": 1243
+    },
+    {
+      "epoch": 3.898981989036805,
+      "grad_norm": 1.184617005441351,
+      "learning_rate": 2.8099823810728255e-06,
+      "loss": 0.5604,
+      "step": 1244
+    },
+    {
+      "epoch": 3.9021143304620205,
+      "grad_norm": 1.1378204967249268,
+      "learning_rate": 2.7947840639921308e-06,
+      "loss": 0.528,
+      "step": 1245
+    },
+    {
+      "epoch": 3.9052466718872356,
+      "grad_norm": 1.1876757504170918,
+      "learning_rate": 2.779620280584303e-06,
+      "loss": 0.5226,
+      "step": 1246
+    },
+    {
+      "epoch": 3.908379013312451,
+      "grad_norm": 1.1809603854780806,
+      "learning_rate": 2.764491103527406e-06,
+      "loss": 0.5266,
+      "step": 1247
+    },
+    {
+      "epoch": 3.9115113547376663,
+      "grad_norm": 1.1873512461852518,
+      "learning_rate": 2.7493966053336397e-06,
+      "loss": 0.5414,
+      "step": 1248
+    },
+    {
+      "epoch": 3.914643696162882,
+      "grad_norm": 1.1504208913330314,
+      "learning_rate": 2.734336858348987e-06,
+      "loss": 0.5283,
+      "step": 1249
+    },
+    {
+      "epoch": 3.917776037588097,
+      "grad_norm": 1.143048240337088,
+      "learning_rate": 2.7193119347528842e-06,
+      "loss": 0.5337,
+      "step": 1250
+    },
+    {
+      "epoch": 3.9209083790133126,
+      "grad_norm": 1.1689363408230222,
+      "learning_rate": 2.7043219065578497e-06,
+      "loss": 0.557,
+      "step": 1251
+    },
+    {
+      "epoch": 3.9240407204385277,
+      "grad_norm": 1.1555945926450477,
+      "learning_rate": 2.6893668456091627e-06,
+      "loss": 0.5256,
+      "step": 1252
+    },
+    {
+      "epoch": 3.9271730618637433,
+      "grad_norm": 1.158032059169674,
+      "learning_rate": 2.674446823584507e-06,
+      "loss": 0.5011,
+      "step": 1253
+    },
+    {
+      "epoch": 3.9303054032889584,
+      "grad_norm": 1.163791390523606,
+      "learning_rate": 2.6595619119936267e-06,
+      "loss": 0.5324,
+      "step": 1254
+    },
+    {
+      "epoch": 3.933437744714174,
+      "grad_norm": 1.1473109388161766,
+      "learning_rate": 2.644712182177992e-06,
+      "loss": 0.5454,
+      "step": 1255
+    },
+    {
+      "epoch": 3.936570086139389,
+      "grad_norm": 1.1721160935749964,
+      "learning_rate": 2.62989770531044e-06,
+      "loss": 0.5144,
+      "step": 1256
+    },
+    {
+      "epoch": 3.9397024275646046,
+      "grad_norm": 1.2026978621361013,
+      "learning_rate": 2.6151185523948577e-06,
+      "loss": 0.5587,
+      "step": 1257
+    },
+    {
+      "epoch": 3.9428347689898198,
+      "grad_norm": 1.161111467446485,
+      "learning_rate": 2.6003747942658243e-06,
+      "loss": 0.5172,
+      "step": 1258
+    },
+    {
+      "epoch": 3.9459671104150353,
+      "grad_norm": 1.1895161788905066,
+      "learning_rate": 2.5856665015882686e-06,
+      "loss": 0.5647,
+      "step": 1259
+    },
+    {
+      "epoch": 3.9490994518402505,
+      "grad_norm": 1.1667391962309854,
+      "learning_rate": 2.570993744857151e-06,
+      "loss": 0.539,
+      "step": 1260
+    },
+    {
+      "epoch": 3.952231793265466,
+      "grad_norm": 1.1990408248065403,
+      "learning_rate": 2.5563565943971027e-06,
+      "loss": 0.5308,
+      "step": 1261
+    },
+    {
+      "epoch": 3.955364134690681,
+      "grad_norm": 1.177576874471793,
+      "learning_rate": 2.541755120362108e-06,
+      "loss": 0.5302,
+      "step": 1262
+    },
+    {
+      "epoch": 3.9584964761158967,
+      "grad_norm": 1.1329715387380235,
+      "learning_rate": 2.5271893927351464e-06,
+      "loss": 0.529,
+      "step": 1263
+    },
+    {
+      "epoch": 3.961628817541112,
+      "grad_norm": 1.20780770381308,
+      "learning_rate": 2.5126594813278792e-06,
+      "loss": 0.5298,
+      "step": 1264
+    },
+    {
+      "epoch": 3.9647611589663274,
+      "grad_norm": 1.2074087314834707,
+      "learning_rate": 2.4981654557803026e-06,
+      "loss": 0.5553,
+      "step": 1265
+    },
+    {
+      "epoch": 3.9678935003915425,
+      "grad_norm": 1.176718451434302,
+      "learning_rate": 2.4837073855604186e-06,
+      "loss": 0.5376,
+      "step": 1266
+    },
+    {
+      "epoch": 3.971025841816758,
+      "grad_norm": 1.1582592387697839,
+      "learning_rate": 2.469285339963892e-06,
+      "loss": 0.5016,
+      "step": 1267
+    },
+    {
+      "epoch": 3.9741581832419732,
+      "grad_norm": 1.126175949855287,
+      "learning_rate": 2.4548993881137327e-06,
+      "loss": 0.5164,
+      "step": 1268
+    },
+    {
+      "epoch": 3.977290524667189,
+      "grad_norm": 1.182910505553956,
+      "learning_rate": 2.4405495989599627e-06,
+      "loss": 0.5606,
+      "step": 1269
+    },
+    {
+      "epoch": 3.980422866092404,
+      "grad_norm": 1.1661378582361217,
+      "learning_rate": 2.4262360412792663e-06,
+      "loss": 0.552,
+      "step": 1270
+    },
+    {
+      "epoch": 3.9835552075176195,
+      "grad_norm": 1.1721964398123115,
+      "learning_rate": 2.4119587836746885e-06,
+      "loss": 0.5481,
+      "step": 1271
+    },
+    {
+      "epoch": 3.9866875489428346,
+      "grad_norm": 1.1861569431208108,
+      "learning_rate": 2.3977178945752877e-06,
+      "loss": 0.5361,
+      "step": 1272
+    },
+    {
+      "epoch": 3.98981989036805,
+      "grad_norm": 1.142933675959546,
+      "learning_rate": 2.383513442235812e-06,
+      "loss": 0.5216,
+      "step": 1273
+    },
+    {
+      "epoch": 3.9929522317932653,
+      "grad_norm": 1.173762534357803,
+      "learning_rate": 2.3693454947363782e-06,
+      "loss": 0.5545,
+      "step": 1274
+    },
+    {
+      "epoch": 3.996084573218481,
+      "grad_norm": 1.1831215530299448,
+      "learning_rate": 2.35521411998213e-06,
+      "loss": 0.5296,
+      "step": 1275
+    },
+    {
+      "epoch": 3.999216914643696,
+      "grad_norm": 1.1124833568408279,
+      "learning_rate": 2.341119385702929e-06,
+      "loss": 0.5065,
+      "step": 1276
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.1124833568408279,
+      "learning_rate": 2.3270613594530235e-06,
+      "loss": 0.5165,
+      "step": 1277
+    },
+    {
+      "epoch": 4.003132341425215,
+      "grad_norm": 2.6425185615357405,
+      "learning_rate": 2.3130401086107247e-06,
+      "loss": 0.4391,
+      "step": 1278
+    },
+    {
+      "epoch": 4.006264682850431,
+      "grad_norm": 1.8968989656450992,
+      "learning_rate": 2.2990557003780833e-06,
+      "loss": 0.4202,
+      "step": 1279
+    },
+    {
+      "epoch": 4.009397024275646,
+      "grad_norm": 1.8142320294338747,
+      "learning_rate": 2.2851082017805704e-06,
+      "loss": 0.4009,
+      "step": 1280
+    },
+    {
+      "epoch": 4.012529365700861,
+      "grad_norm": 1.7165806912562718,
+      "learning_rate": 2.27119767966675e-06,
+      "loss": 0.405,
+      "step": 1281
+    },
+    {
+      "epoch": 4.0156617071260765,
+      "grad_norm": 1.5924208377321079,
+      "learning_rate": 2.2573242007079697e-06,
+      "loss": 0.421,
+      "step": 1282
+    },
+    {
+      "epoch": 4.0187940485512925,
+      "grad_norm": 1.5038694947047038,
+      "learning_rate": 2.2434878313980235e-06,
+      "loss": 0.4002,
+      "step": 1283
+    },
+    {
+      "epoch": 4.021926389976508,
+      "grad_norm": 1.340573966307308,
+      "learning_rate": 2.2296886380528494e-06,
+      "loss": 0.4272,
+      "step": 1284
+    },
+    {
+      "epoch": 4.025058731401723,
+      "grad_norm": 1.2987422109641342,
+      "learning_rate": 2.215926686810206e-06,
+      "loss": 0.4349,
+      "step": 1285
+    },
+    {
+      "epoch": 4.028191072826938,
+      "grad_norm": 1.252719539838883,
+      "learning_rate": 2.202202043629356e-06,
+      "loss": 0.3898,
+      "step": 1286
+    },
+    {
+      "epoch": 4.031323414252154,
+      "grad_norm": 1.327752575517584,
+      "learning_rate": 2.1885147742907465e-06,
+      "loss": 0.4175,
+      "step": 1287
+    },
+    {
+      "epoch": 4.034455755677369,
+      "grad_norm": 1.4402839116286432,
+      "learning_rate": 2.1748649443957004e-06,
+      "loss": 0.4137,
+      "step": 1288
+    },
+    {
+      "epoch": 4.037588097102584,
+      "grad_norm": 1.5655168444654703,
+      "learning_rate": 2.1612526193660975e-06,
+      "loss": 0.4051,
+      "step": 1289
+    },
+    {
+      "epoch": 4.040720438527799,
+      "grad_norm": 1.6161366724042316,
+      "learning_rate": 2.1476778644440553e-06,
+      "loss": 0.4028,
+      "step": 1290
+    },
+    {
+      "epoch": 4.043852779953015,
+      "grad_norm": 1.5596875540342134,
+      "learning_rate": 2.134140744691633e-06,
+      "loss": 0.3997,
+      "step": 1291
+    },
+    {
+      "epoch": 4.04698512137823,
+      "grad_norm": 1.4460862400789365,
+      "learning_rate": 2.1206413249904967e-06,
+      "loss": 0.4125,
+      "step": 1292
+    },
+    {
+      "epoch": 4.0501174628034455,
+      "grad_norm": 1.3882745890795896,
+      "learning_rate": 2.1071796700416334e-06,
+      "loss": 0.3972,
+      "step": 1293
+    },
+    {
+      "epoch": 4.053249804228661,
+      "grad_norm": 1.4012869387398328,
+      "learning_rate": 2.093755844365021e-06,
+      "loss": 0.4155,
+      "step": 1294
+    },
+    {
+      "epoch": 4.056382145653877,
+      "grad_norm": 1.3519015902940699,
+      "learning_rate": 2.0803699122993293e-06,
+      "loss": 0.4178,
+      "step": 1295
+    },
+    {
+      "epoch": 4.059514487079092,
+      "grad_norm": 1.3602338471577913,
+      "learning_rate": 2.067021938001611e-06,
+      "loss": 0.4193,
+      "step": 1296
+    },
+    {
+      "epoch": 4.062646828504307,
+      "grad_norm": 1.2707959489599356,
+      "learning_rate": 2.0537119854469845e-06,
+      "loss": 0.4235,
+      "step": 1297
+    },
+    {
+      "epoch": 4.065779169929522,
+      "grad_norm": 1.2304315457883275,
+      "learning_rate": 2.0404401184283408e-06,
+      "loss": 0.4281,
+      "step": 1298
+    },
+    {
+      "epoch": 4.068911511354738,
+      "grad_norm": 1.3126596083493967,
+      "learning_rate": 2.0272064005560354e-06,
+      "loss": 0.4116,
+      "step": 1299
+    },
+    {
+      "epoch": 4.072043852779953,
+      "grad_norm": 1.2164141170274532,
+      "learning_rate": 2.01401089525757e-06,
+      "loss": 0.4128,
+      "step": 1300
+    },
+    {
+      "epoch": 4.075176194205168,
+      "grad_norm": 1.2157628287632507,
+      "learning_rate": 2.000853665777305e-06,
+      "loss": 0.3966,
+      "step": 1301
+    },
+    {
+      "epoch": 4.078308535630383,
+      "grad_norm": 1.1768433581291058,
+      "learning_rate": 1.987734775176151e-06,
+      "loss": 0.3773,
+      "step": 1302
+    },
+    {
+      "epoch": 4.081440877055599,
+      "grad_norm": 1.195358771674673,
+      "learning_rate": 1.974654286331258e-06,
+      "loss": 0.4021,
+      "step": 1303
+    },
+    {
+      "epoch": 4.084573218480815,
+      "grad_norm": 1.2360866223981186,
+      "learning_rate": 1.961612261935727e-06,
+      "loss": 0.3914,
+      "step": 1304
+    },
+    {
+      "epoch": 4.08770555990603,
+      "grad_norm": 1.2186686913113904,
+      "learning_rate": 1.9486087644983056e-06,
+      "loss": 0.3896,
+      "step": 1305
+    },
+    {
+      "epoch": 4.090837901331245,
+      "grad_norm": 1.2346935215427521,
+      "learning_rate": 1.9356438563430822e-06,
+      "loss": 0.4038,
+      "step": 1306
+    },
+    {
+      "epoch": 4.093970242756461,
+      "grad_norm": 1.198362834738886,
+      "learning_rate": 1.9227175996091952e-06,
+      "loss": 0.4009,
+      "step": 1307
+    },
+    {
+      "epoch": 4.097102584181676,
+      "grad_norm": 1.1668330606150041,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.3942,
+      "step": 1308
+    },
+    {
+      "epoch": 4.100234925606891,
+      "grad_norm": 1.2063732731647043,
+      "learning_rate": 1.8969812880354167e-06,
+      "loss": 0.394,
+      "step": 1309
+    },
+    {
+      "epoch": 4.103367267032106,
+      "grad_norm": 1.2113132171130907,
+      "learning_rate": 1.8841713565463548e-06,
+      "loss": 0.3945,
+      "step": 1310
+    },
+    {
+      "epoch": 4.106499608457322,
+      "grad_norm": 1.2247544654973348,
+      "learning_rate": 1.871400323179695e-06,
+      "loss": 0.4096,
+      "step": 1311
+    },
+    {
+      "epoch": 4.109631949882537,
+      "grad_norm": 1.2881131176691998,
+      "learning_rate": 1.858668249145359e-06,
+      "loss": 0.4071,
+      "step": 1312
+    },
+    {
+      "epoch": 4.1127642913077525,
+      "grad_norm": 1.2380244490730665,
+      "learning_rate": 1.8459751954665372e-06,
+      "loss": 0.3887,
+      "step": 1313
+    },
+    {
+      "epoch": 4.115896632732968,
+      "grad_norm": 1.2457742852610907,
+      "learning_rate": 1.8333212229794039e-06,
+      "loss": 0.3938,
+      "step": 1314
+    },
+    {
+      "epoch": 4.119028974158184,
+      "grad_norm": 1.299735360583674,
+      "learning_rate": 1.820706392332824e-06,
+      "loss": 0.4278,
+      "step": 1315
+    },
+    {
+      "epoch": 4.122161315583399,
+      "grad_norm": 1.22006539661709,
+      "learning_rate": 1.8081307639880519e-06,
+      "loss": 0.3969,
+      "step": 1316
+    },
+    {
+      "epoch": 4.125293657008614,
+      "grad_norm": 1.2043704461002585,
+      "learning_rate": 1.7955943982184608e-06,
+      "loss": 0.3772,
+      "step": 1317
+    },
+    {
+      "epoch": 4.128425998433829,
+      "grad_norm": 1.1954512153531711,
+      "learning_rate": 1.7830973551092346e-06,
+      "loss": 0.4232,
+      "step": 1318
+    },
+    {
+      "epoch": 4.131558339859045,
+      "grad_norm": 1.2198506026583384,
+      "learning_rate": 1.7706396945570947e-06,
+      "loss": 0.3996,
+      "step": 1319
+    },
+    {
+      "epoch": 4.13469068128426,
+      "grad_norm": 1.2218183019573885,
+      "learning_rate": 1.7582214762700057e-06,
+      "loss": 0.428,
+      "step": 1320
+    },
+    {
+      "epoch": 4.137823022709475,
+      "grad_norm": 1.1976267387815713,
+      "learning_rate": 1.7458427597668893e-06,
+      "loss": 0.4062,
+      "step": 1321
+    },
+    {
+      "epoch": 4.14095536413469,
+      "grad_norm": 1.1697967121996906,
+      "learning_rate": 1.7335036043773402e-06,
+      "loss": 0.3889,
+      "step": 1322
+    },
+    {
+      "epoch": 4.144087705559906,
+      "grad_norm": 1.1924932065029525,
+      "learning_rate": 1.7212040692413468e-06,
+      "loss": 0.4245,
+      "step": 1323
+    },
+    {
+      "epoch": 4.1472200469851215,
+      "grad_norm": 1.2147406012062079,
+      "learning_rate": 1.7089442133089928e-06,
+      "loss": 0.3927,
+      "step": 1324
+    },
+    {
+      "epoch": 4.150352388410337,
+      "grad_norm": 1.1788944142485256,
+      "learning_rate": 1.6967240953401954e-06,
+      "loss": 0.3862,
+      "step": 1325
+    },
+    {
+      "epoch": 4.153484729835552,
+      "grad_norm": 1.1947326262463431,
+      "learning_rate": 1.6845437739044012e-06,
+      "loss": 0.3828,
+      "step": 1326
+    },
+    {
+      "epoch": 4.156617071260768,
+      "grad_norm": 1.1173824303336672,
+      "learning_rate": 1.672403307380328e-06,
+      "loss": 0.3977,
+      "step": 1327
+    },
+    {
+      "epoch": 4.159749412685983,
+      "grad_norm": 1.1753324405533268,
+      "learning_rate": 1.6603027539556682e-06,
+      "loss": 0.3866,
+      "step": 1328
+    },
+    {
+      "epoch": 4.162881754111198,
+      "grad_norm": 1.1729189872697827,
+      "learning_rate": 1.6482421716268215e-06,
+      "loss": 0.421,
+      "step": 1329
+    },
+    {
+      "epoch": 4.166014095536413,
+      "grad_norm": 1.2345796879070534,
+      "learning_rate": 1.6362216181986002e-06,
+      "loss": 0.3713,
+      "step": 1330
+    },
+    {
+      "epoch": 4.169146436961629,
+      "grad_norm": 1.155602810779086,
+      "learning_rate": 1.6242411512839728e-06,
+      "loss": 0.387,
+      "step": 1331
+    },
+    {
+      "epoch": 4.172278778386844,
+      "grad_norm": 1.1762864409104852,
+      "learning_rate": 1.6123008283037778e-06,
+      "loss": 0.3991,
+      "step": 1332
+    },
+    {
+      "epoch": 4.175411119812059,
+      "grad_norm": 1.2221930726179149,
+      "learning_rate": 1.6004007064864468e-06,
+      "loss": 0.3585,
+      "step": 1333
+    },
+    {
+      "epoch": 4.1785434612372745,
+      "grad_norm": 1.1528755586082111,
+      "learning_rate": 1.5885408428677296e-06,
+      "loss": 0.4074,
+      "step": 1334
+    },
+    {
+      "epoch": 4.1816758026624905,
+      "grad_norm": 1.1990119644346569,
+      "learning_rate": 1.5767212942904275e-06,
+      "loss": 0.4252,
+      "step": 1335
+    },
+    {
+      "epoch": 4.184808144087706,
+      "grad_norm": 1.190374191471458,
+      "learning_rate": 1.564942117404119e-06,
+      "loss": 0.4137,
+      "step": 1336
+    },
+    {
+      "epoch": 4.187940485512921,
+      "grad_norm": 1.2490135566571798,
+      "learning_rate": 1.5532033686648785e-06,
+      "loss": 0.3961,
+      "step": 1337
+    },
+    {
+      "epoch": 4.191072826938136,
+      "grad_norm": 1.211936868558367,
+      "learning_rate": 1.5415051043350194e-06,
+      "loss": 0.3968,
+      "step": 1338
+    },
+    {
+      "epoch": 4.194205168363352,
+      "grad_norm": 1.2268059684742754,
+      "learning_rate": 1.5298473804828195e-06,
+      "loss": 0.3992,
+      "step": 1339
+    },
+    {
+      "epoch": 4.197337509788567,
+      "grad_norm": 1.2449496248919925,
+      "learning_rate": 1.518230252982248e-06,
+      "loss": 0.3926,
+      "step": 1340
+    },
+    {
+      "epoch": 4.200469851213782,
+      "grad_norm": 1.192327150884329,
+      "learning_rate": 1.5066537775127044e-06,
+      "loss": 0.3974,
+      "step": 1341
+    },
+    {
+      "epoch": 4.203602192638997,
+      "grad_norm": 1.235268675446175,
+      "learning_rate": 1.4951180095587426e-06,
+      "loss": 0.3963,
+      "step": 1342
+    },
+    {
+      "epoch": 4.206734534064213,
+      "grad_norm": 1.2044322613986833,
+      "learning_rate": 1.4836230044098164e-06,
+      "loss": 0.4214,
+      "step": 1343
+    },
+    {
+      "epoch": 4.209866875489428,
+      "grad_norm": 1.233078464789091,
+      "learning_rate": 1.4721688171600047e-06,
+      "loss": 0.3956,
+      "step": 1344
+    },
+    {
+      "epoch": 4.2129992169146435,
+      "grad_norm": 1.2257391290015276,
+      "learning_rate": 1.4607555027077524e-06,
+      "loss": 0.424,
+      "step": 1345
+    },
+    {
+      "epoch": 4.216131558339859,
+      "grad_norm": 1.239469422660483,
+      "learning_rate": 1.4493831157556094e-06,
+      "loss": 0.3933,
+      "step": 1346
+    },
+    {
+      "epoch": 4.219263899765075,
+      "grad_norm": 1.1537827041020894,
+      "learning_rate": 1.4380517108099623e-06,
+      "loss": 0.37,
+      "step": 1347
+    },
+    {
+      "epoch": 4.22239624119029,
+      "grad_norm": 1.2110428750506566,
+      "learning_rate": 1.426761342180777e-06,
+      "loss": 0.4159,
+      "step": 1348
+    },
+    {
+      "epoch": 4.225528582615505,
+      "grad_norm": 1.2034706739426686,
+      "learning_rate": 1.4155120639813392e-06,
+      "loss": 0.4007,
+      "step": 1349
+    },
+    {
+      "epoch": 4.22866092404072,
+      "grad_norm": 1.2096667555894145,
+      "learning_rate": 1.4043039301279904e-06,
+      "loss": 0.4168,
+      "step": 1350
+    },
+    {
+      "epoch": 4.231793265465936,
+      "grad_norm": 1.2136387061332223,
+      "learning_rate": 1.393136994339872e-06,
+      "loss": 0.3891,
+      "step": 1351
+    },
+    {
+      "epoch": 4.234925606891151,
+      "grad_norm": 1.2279471606708432,
+      "learning_rate": 1.3820113101386712e-06,
+      "loss": 0.398,
+      "step": 1352
+    },
+    {
+      "epoch": 4.238057948316366,
+      "grad_norm": 1.1712817094482313,
+      "learning_rate": 1.3709269308483619e-06,
+      "loss": 0.3974,
+      "step": 1353
+    },
+    {
+      "epoch": 4.241190289741581,
+      "grad_norm": 1.2244805625115447,
+      "learning_rate": 1.359883909594948e-06,
+      "loss": 0.4101,
+      "step": 1354
+    },
+    {
+      "epoch": 4.244322631166797,
+      "grad_norm": 1.2055456362307584,
+      "learning_rate": 1.348882299306209e-06,
+      "loss": 0.4189,
+      "step": 1355
+    },
+    {
+      "epoch": 4.2474549725920125,
+      "grad_norm": 1.2472551970779924,
+      "learning_rate": 1.3379221527114484e-06,
+      "loss": 0.3974,
+      "step": 1356
+    },
+    {
+      "epoch": 4.250587314017228,
+      "grad_norm": 1.1928935135771526,
+      "learning_rate": 1.327003522341237e-06,
+      "loss": 0.4061,
+      "step": 1357
+    },
+    {
+      "epoch": 4.253719655442444,
+      "grad_norm": 1.2297720660775902,
+      "learning_rate": 1.3161264605271662e-06,
+      "loss": 0.4037,
+      "step": 1358
+    },
+    {
+      "epoch": 4.256851996867659,
+      "grad_norm": 1.2100253932220972,
+      "learning_rate": 1.3052910194015922e-06,
+      "loss": 0.3932,
+      "step": 1359
+    },
+    {
+      "epoch": 4.259984338292874,
+      "grad_norm": 1.189046936366124,
+      "learning_rate": 1.2944972508973908e-06,
+      "loss": 0.3878,
+      "step": 1360
+    },
+    {
+      "epoch": 4.263116679718089,
+      "grad_norm": 1.1920673345461743,
+      "learning_rate": 1.2837452067477062e-06,
+      "loss": 0.3915,
+      "step": 1361
+    },
+    {
+      "epoch": 4.266249021143304,
+      "grad_norm": 1.2118917776405405,
+      "learning_rate": 1.273034938485702e-06,
+      "loss": 0.3939,
+      "step": 1362
+    },
+    {
+      "epoch": 4.26938136256852,
+      "grad_norm": 1.2126398847084519,
+      "learning_rate": 1.2623664974443195e-06,
+      "loss": 0.3988,
+      "step": 1363
+    },
+    {
+      "epoch": 4.272513703993735,
+      "grad_norm": 1.1928692981682214,
+      "learning_rate": 1.2517399347560165e-06,
+      "loss": 0.3966,
+      "step": 1364
+    },
+    {
+      "epoch": 4.27564604541895,
+      "grad_norm": 1.1977505153553263,
+      "learning_rate": 1.2411553013525457e-06,
+      "loss": 0.4051,
+      "step": 1365
+    },
+    {
+      "epoch": 4.2787783868441664,
+      "grad_norm": 1.222427847128266,
+      "learning_rate": 1.230612647964694e-06,
+      "loss": 0.4103,
+      "step": 1366
+    },
+    {
+      "epoch": 4.281910728269382,
+      "grad_norm": 1.200297992267844,
+      "learning_rate": 1.2201120251220378e-06,
+      "loss": 0.4016,
+      "step": 1367
+    },
+    {
+      "epoch": 4.285043069694597,
+      "grad_norm": 1.216728301148234,
+      "learning_rate": 1.209653483152714e-06,
+      "loss": 0.4014,
+      "step": 1368
+    },
+    {
+      "epoch": 4.288175411119812,
+      "grad_norm": 1.224540090293941,
+      "learning_rate": 1.1992370721831693e-06,
+      "loss": 0.3974,
+      "step": 1369
+    },
+    {
+      "epoch": 4.291307752545027,
+      "grad_norm": 1.1950041179139188,
+      "learning_rate": 1.1888628421379222e-06,
+      "loss": 0.4002,
+      "step": 1370
+    },
+    {
+      "epoch": 4.294440093970243,
+      "grad_norm": 1.190070377224994,
+      "learning_rate": 1.1785308427393182e-06,
+      "loss": 0.3964,
+      "step": 1371
+    },
+    {
+      "epoch": 4.297572435395458,
+      "grad_norm": 1.2008002010734806,
+      "learning_rate": 1.1682411235073032e-06,
+      "loss": 0.4076,
+      "step": 1372
+    },
+    {
+      "epoch": 4.300704776820673,
+      "grad_norm": 1.2132858157877429,
+      "learning_rate": 1.1579937337591774e-06,
+      "loss": 0.4015,
+      "step": 1373
+    },
+    {
+      "epoch": 4.303837118245889,
+      "grad_norm": 1.1826916321444987,
+      "learning_rate": 1.1477887226093642e-06,
+      "loss": 0.4171,
+      "step": 1374
+    },
+    {
+      "epoch": 4.306969459671104,
+      "grad_norm": 1.2089546478448403,
+      "learning_rate": 1.1376261389691633e-06,
+      "loss": 0.4146,
+      "step": 1375
+    },
+    {
+      "epoch": 4.3101018010963195,
+      "grad_norm": 1.206358523522593,
+      "learning_rate": 1.1275060315465346e-06,
+      "loss": 0.3872,
+      "step": 1376
+    },
+    {
+      "epoch": 4.313234142521535,
+      "grad_norm": 1.179621274726758,
+      "learning_rate": 1.1174284488458509e-06,
+      "loss": 0.4172,
+      "step": 1377
+    },
+    {
+      "epoch": 4.316366483946751,
+      "grad_norm": 1.268118607500806,
+      "learning_rate": 1.1073934391676666e-06,
+      "loss": 0.4213,
+      "step": 1378
+    },
+    {
+      "epoch": 4.319498825371966,
+      "grad_norm": 1.2431330503496503,
+      "learning_rate": 1.097401050608493e-06,
+      "loss": 0.3729,
+      "step": 1379
+    },
+    {
+      "epoch": 4.322631166797181,
+      "grad_norm": 1.2057856092166312,
+      "learning_rate": 1.0874513310605628e-06,
+      "loss": 0.388,
+      "step": 1380
+    },
+    {
+      "epoch": 4.325763508222396,
+      "grad_norm": 1.1984630219610473,
+      "learning_rate": 1.0775443282116017e-06,
+      "loss": 0.3919,
+      "step": 1381
+    },
+    {
+      "epoch": 4.328895849647612,
+      "grad_norm": 1.1721834023779714,
+      "learning_rate": 1.0676800895446016e-06,
+      "loss": 0.4033,
+      "step": 1382
+    },
+    {
+      "epoch": 4.332028191072827,
+      "grad_norm": 1.2483769893731758,
+      "learning_rate": 1.0578586623375863e-06,
+      "loss": 0.3826,
+      "step": 1383
+    },
+    {
+      "epoch": 4.335160532498042,
+      "grad_norm": 1.250417832052342,
+      "learning_rate": 1.048080093663395e-06,
+      "loss": 0.4275,
+      "step": 1384
+    },
+    {
+      "epoch": 4.338292873923257,
+      "grad_norm": 1.2565058613082094,
+      "learning_rate": 1.0383444303894453e-06,
+      "loss": 0.3947,
+      "step": 1385
+    },
+    {
+      "epoch": 4.341425215348473,
+      "grad_norm": 1.2252255070005802,
+      "learning_rate": 1.0286517191775215e-06,
+      "loss": 0.4061,
+      "step": 1386
+    },
+    {
+      "epoch": 4.3445575567736885,
+      "grad_norm": 1.235428710435152,
+      "learning_rate": 1.0190020064835404e-06,
+      "loss": 0.4089,
+      "step": 1387
+    },
+    {
+      "epoch": 4.347689898198904,
+      "grad_norm": 1.2195624521585995,
+      "learning_rate": 1.009395338557333e-06,
+      "loss": 0.3892,
+      "step": 1388
+    },
+    {
+      "epoch": 4.350822239624119,
+      "grad_norm": 1.2421525738212085,
+      "learning_rate": 9.998317614424246e-07,
+      "loss": 0.4151,
+      "step": 1389
+    },
+    {
+      "epoch": 4.353954581049335,
+      "grad_norm": 1.2269216736656088,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.4203,
+      "step": 1390
+    },
+    {
+      "epoch": 4.35708692247455,
+      "grad_norm": 1.2627277845413256,
+      "learning_rate": 9.80834062787731e-07,
+      "loss": 0.4115,
+      "step": 1391
+    },
+    {
+      "epoch": 4.360219263899765,
+      "grad_norm": 1.2175493590777189,
+      "learning_rate": 9.714000323014705e-07,
+      "loss": 0.3879,
+      "step": 1392
+    },
+    {
+      "epoch": 4.36335160532498,
+      "grad_norm": 1.1697012465299366,
+      "learning_rate": 9.62009274733121e-07,
+      "loss": 0.3977,
+      "step": 1393
+    },
+    {
+      "epoch": 4.366483946750196,
+      "grad_norm": 1.217611380413171,
+      "learning_rate": 9.526618350913752e-07,
+      "loss": 0.3774,
+      "step": 1394
+    },
+    {
+      "epoch": 4.369616288175411,
+      "grad_norm": 1.1843080929354863,
+      "learning_rate": 9.43357758177309e-07,
+      "loss": 0.39,
+      "step": 1395
+    },
+    {
+      "epoch": 4.372748629600626,
+      "grad_norm": 1.1973637151971002,
+      "learning_rate": 9.34097088584166e-07,
+      "loss": 0.4149,
+      "step": 1396
+    },
+    {
+      "epoch": 4.3758809710258415,
+      "grad_norm": 1.1880735619823894,
+      "learning_rate": 9.248798706971462e-07,
+      "loss": 0.3826,
+      "step": 1397
+    },
+    {
+      "epoch": 4.3790133124510575,
+      "grad_norm": 1.2411864789666913,
+      "learning_rate": 9.15706148693184e-07,
+      "loss": 0.4115,
+      "step": 1398
+    },
+    {
+      "epoch": 4.382145653876273,
+      "grad_norm": 1.2197612576036618,
+      "learning_rate": 9.065759665407514e-07,
+      "loss": 0.4113,
+      "step": 1399
+    },
+    {
+      "epoch": 4.385277995301488,
+      "grad_norm": 1.255965107907724,
+      "learning_rate": 8.974893679996388e-07,
+      "loss": 0.3768,
+      "step": 1400
+    },
+    {
+      "epoch": 4.388410336726703,
+      "grad_norm": 1.1777066026687946,
+      "learning_rate": 8.884463966207391e-07,
+      "loss": 0.3983,
+      "step": 1401
+    },
+    {
+      "epoch": 4.391542678151919,
+      "grad_norm": 1.1768660233012396,
+      "learning_rate": 8.794470957458523e-07,
+      "loss": 0.403,
+      "step": 1402
+    },
+    {
+      "epoch": 4.394675019577134,
+      "grad_norm": 1.2598498922043773,
+      "learning_rate": 8.704915085074728e-07,
+      "loss": 0.4133,
+      "step": 1403
+    },
+    {
+      "epoch": 4.397807361002349,
+      "grad_norm": 1.1732711385614463,
+      "learning_rate": 8.615796778285735e-07,
+      "loss": 0.3724,
+      "step": 1404
+    },
+    {
+      "epoch": 4.400939702427564,
+      "grad_norm": 1.167025579090733,
+      "learning_rate": 8.527116464224128e-07,
+      "loss": 0.4062,
+      "step": 1405
+    },
+    {
+      "epoch": 4.40407204385278,
+      "grad_norm": 1.2023162019645415,
+      "learning_rate": 8.438874567923261e-07,
+      "loss": 0.3917,
+      "step": 1406
+    },
+    {
+      "epoch": 4.407204385277995,
+      "grad_norm": 1.1810461949196573,
+      "learning_rate": 8.351071512315167e-07,
+      "loss": 0.4052,
+      "step": 1407
+    },
+    {
+      "epoch": 4.4103367267032105,
+      "grad_norm": 1.202253524025422,
+      "learning_rate": 8.263707718228641e-07,
+      "loss": 0.3979,
+      "step": 1408
+    },
+    {
+      "epoch": 4.413469068128426,
+      "grad_norm": 1.1576984628276288,
+      "learning_rate": 8.17678360438704e-07,
+      "loss": 0.401,
+      "step": 1409
+    },
+    {
+      "epoch": 4.416601409553642,
+      "grad_norm": 1.2291103726343342,
+      "learning_rate": 8.090299587406514e-07,
+      "loss": 0.4177,
+      "step": 1410
+    },
+    {
+      "epoch": 4.419733750978857,
+      "grad_norm": 1.2466018705955504,
+      "learning_rate": 8.004256081793782e-07,
+      "loss": 0.4247,
+      "step": 1411
+    },
+    {
+      "epoch": 4.422866092404072,
+      "grad_norm": 1.24245531665182,
+      "learning_rate": 7.918653499944318e-07,
+      "loss": 0.3814,
+      "step": 1412
+    },
+    {
+      "epoch": 4.425998433829287,
+      "grad_norm": 1.2207227340577,
+      "learning_rate": 7.833492252140284e-07,
+      "loss": 0.4179,
+      "step": 1413
+    },
+    {
+      "epoch": 4.429130775254503,
+      "grad_norm": 1.2374658186692722,
+      "learning_rate": 7.748772746548572e-07,
+      "loss": 0.438,
+      "step": 1414
+    },
+    {
+      "epoch": 4.432263116679718,
+      "grad_norm": 1.2445922732186314,
+      "learning_rate": 7.664495389218884e-07,
+      "loss": 0.4164,
+      "step": 1415
+    },
+    {
+      "epoch": 4.435395458104933,
+      "grad_norm": 1.2048318810786247,
+      "learning_rate": 7.580660584081755e-07,
+      "loss": 0.4114,
+      "step": 1416
+    },
+    {
+      "epoch": 4.438527799530148,
+      "grad_norm": 1.2092315947423604,
+      "learning_rate": 7.497268732946605e-07,
+      "loss": 0.3972,
+      "step": 1417
+    },
+    {
+      "epoch": 4.441660140955364,
+      "grad_norm": 1.1684399668507264,
+      "learning_rate": 7.414320235499839e-07,
+      "loss": 0.4007,
+      "step": 1418
+    },
+    {
+      "epoch": 4.44479248238058,
+      "grad_norm": 1.2018031522146513,
+      "learning_rate": 7.331815489302941e-07,
+      "loss": 0.3926,
+      "step": 1419
+    },
+    {
+      "epoch": 4.447924823805795,
+      "grad_norm": 1.2377323452479039,
+      "learning_rate": 7.249754889790539e-07,
+      "loss": 0.4024,
+      "step": 1420
+    },
+    {
+      "epoch": 4.45105716523101,
+      "grad_norm": 1.2372044961613387,
+      "learning_rate": 7.168138830268534e-07,
+      "loss": 0.4153,
+      "step": 1421
+    },
+    {
+      "epoch": 4.454189506656226,
+      "grad_norm": 1.2367482818132942,
+      "learning_rate": 7.086967701912195e-07,
+      "loss": 0.4145,
+      "step": 1422
+    },
+    {
+      "epoch": 4.457321848081441,
+      "grad_norm": 1.1720988594585238,
+      "learning_rate": 7.006241893764298e-07,
+      "loss": 0.3933,
+      "step": 1423
+    },
+    {
+      "epoch": 4.460454189506656,
+      "grad_norm": 1.1906519908299524,
+      "learning_rate": 6.925961792733271e-07,
+      "loss": 0.4213,
+      "step": 1424
+    },
+    {
+      "epoch": 4.463586530931871,
+      "grad_norm": 1.1861914037615011,
+      "learning_rate": 6.846127783591294e-07,
+      "loss": 0.374,
+      "step": 1425
+    },
+    {
+      "epoch": 4.466718872357087,
+      "grad_norm": 1.2151391090817059,
+      "learning_rate": 6.76674024897247e-07,
+      "loss": 0.3958,
+      "step": 1426
+    },
+    {
+      "epoch": 4.469851213782302,
+      "grad_norm": 1.1906210035368663,
+      "learning_rate": 6.687799569371079e-07,
+      "loss": 0.3997,
+      "step": 1427
+    },
+    {
+      "epoch": 4.4729835552075174,
+      "grad_norm": 1.2698252501752594,
+      "learning_rate": 6.609306123139614e-07,
+      "loss": 0.4152,
+      "step": 1428
+    },
+    {
+      "epoch": 4.476115896632733,
+      "grad_norm": 1.239086528624892,
+      "learning_rate": 6.531260286487073e-07,
+      "loss": 0.4244,
+      "step": 1429
+    },
+    {
+      "epoch": 4.479248238057949,
+      "grad_norm": 1.239389095459042,
+      "learning_rate": 6.453662433477137e-07,
+      "loss": 0.3967,
+      "step": 1430
+    },
+    {
+      "epoch": 4.482380579483164,
+      "grad_norm": 1.2078463523567677,
+      "learning_rate": 6.37651293602628e-07,
+      "loss": 0.3998,
+      "step": 1431
+    },
+    {
+      "epoch": 4.485512920908379,
+      "grad_norm": 1.2405207380196077,
+      "learning_rate": 6.299812163902152e-07,
+      "loss": 0.4079,
+      "step": 1432
+    },
+    {
+      "epoch": 4.488645262333594,
+      "grad_norm": 1.1965967124161982,
+      "learning_rate": 6.22356048472168e-07,
+      "loss": 0.3947,
+      "step": 1433
+    },
+    {
+      "epoch": 4.49177760375881,
+      "grad_norm": 1.2365962276051872,
+      "learning_rate": 6.147758263949322e-07,
+      "loss": 0.4145,
+      "step": 1434
+    },
+    {
+      "epoch": 4.494909945184025,
+      "grad_norm": 1.2320724240611538,
+      "learning_rate": 6.072405864895403e-07,
+      "loss": 0.3989,
+      "step": 1435
+    },
+    {
+      "epoch": 4.49804228660924,
+      "grad_norm": 1.2278614714663911,
+      "learning_rate": 5.997503648714254e-07,
+      "loss": 0.4087,
+      "step": 1436
+    },
+    {
+      "epoch": 4.501174628034455,
+      "grad_norm": 1.2459676825726136,
+      "learning_rate": 5.923051974402594e-07,
+      "loss": 0.4076,
+      "step": 1437
+    },
+    {
+      "epoch": 4.504306969459671,
+      "grad_norm": 1.2458668198276721,
+      "learning_rate": 5.849051198797662e-07,
+      "loss": 0.4248,
+      "step": 1438
+    },
+    {
+      "epoch": 4.5074393108848865,
+      "grad_norm": 1.209767689172734,
+      "learning_rate": 5.775501676575702e-07,
+      "loss": 0.4014,
+      "step": 1439
+    },
+    {
+      "epoch": 4.510571652310102,
+      "grad_norm": 1.2285591384985166,
+      "learning_rate": 5.702403760250086e-07,
+      "loss": 0.4147,
+      "step": 1440
+    },
+    {
+      "epoch": 4.513703993735318,
+      "grad_norm": 1.1718999655815265,
+      "learning_rate": 5.629757800169732e-07,
+      "loss": 0.4044,
+      "step": 1441
+    },
+    {
+      "epoch": 4.516836335160533,
+      "grad_norm": 1.217216752100384,
+      "learning_rate": 5.557564144517369e-07,
+      "loss": 0.4091,
+      "step": 1442
+    },
+    {
+      "epoch": 4.519968676585748,
+      "grad_norm": 1.1838737229775769,
+      "learning_rate": 5.485823139307911e-07,
+      "loss": 0.4034,
+      "step": 1443
+    },
+    {
+      "epoch": 4.523101018010963,
+      "grad_norm": 1.1673948755376424,
+      "learning_rate": 5.414535128386766e-07,
+      "loss": 0.3661,
+      "step": 1444
+    },
+    {
+      "epoch": 4.526233359436178,
+      "grad_norm": 1.1757487458448712,
+      "learning_rate": 5.343700453428169e-07,
+      "loss": 0.3971,
+      "step": 1445
+    },
+    {
+      "epoch": 4.529365700861394,
+      "grad_norm": 1.2191061942291306,
+      "learning_rate": 5.273319453933634e-07,
+      "loss": 0.3885,
+      "step": 1446
+    },
+    {
+      "epoch": 4.532498042286609,
+      "grad_norm": 1.1814168865664434,
+      "learning_rate": 5.203392467230229e-07,
+      "loss": 0.3888,
+      "step": 1447
+    },
+    {
+      "epoch": 4.535630383711824,
+      "grad_norm": 1.180338891284529,
+      "learning_rate": 5.133919828468992e-07,
+      "loss": 0.4031,
+      "step": 1448
+    },
+    {
+      "epoch": 4.53876272513704,
+      "grad_norm": 1.2002998136423375,
+      "learning_rate": 5.064901870623362e-07,
+      "loss": 0.4124,
+      "step": 1449
+    },
+    {
+      "epoch": 4.5418950665622555,
+      "grad_norm": 1.225998648068942,
+      "learning_rate": 4.996338924487509e-07,
+      "loss": 0.4079,
+      "step": 1450
+    },
+    {
+      "epoch": 4.545027407987471,
+      "grad_norm": 1.232921807498376,
+      "learning_rate": 4.928231318674836e-07,
+      "loss": 0.4288,
+      "step": 1451
+    },
+    {
+      "epoch": 4.548159749412686,
+      "grad_norm": 1.2108344873262966,
+      "learning_rate": 4.860579379616303e-07,
+      "loss": 0.3752,
+      "step": 1452
+    },
+    {
+      "epoch": 4.551292090837901,
+      "grad_norm": 1.1982395456743722,
+      "learning_rate": 4.793383431558962e-07,
+      "loss": 0.3956,
+      "step": 1453
+    },
+    {
+      "epoch": 4.554424432263117,
+      "grad_norm": 1.2245255923948302,
+      "learning_rate": 4.7266437965643455e-07,
+      "loss": 0.39,
+      "step": 1454
+    },
+    {
+      "epoch": 4.557556773688332,
+      "grad_norm": 1.2237018326288942,
+      "learning_rate": 4.660360794506946e-07,
+      "loss": 0.4064,
+      "step": 1455
+    },
+    {
+      "epoch": 4.560689115113547,
+      "grad_norm": 1.203953237163164,
+      "learning_rate": 4.594534743072654e-07,
+      "loss": 0.4169,
+      "step": 1456
+    },
+    {
+      "epoch": 4.563821456538763,
+      "grad_norm": 1.2087498003259758,
+      "learning_rate": 4.5291659577572753e-07,
+      "loss": 0.3832,
+      "step": 1457
+    },
+    {
+      "epoch": 4.566953797963978,
+      "grad_norm": 1.1942569748553682,
+      "learning_rate": 4.4642547518649824e-07,
+      "loss": 0.3776,
+      "step": 1458
+    },
+    {
+      "epoch": 4.570086139389193,
+      "grad_norm": 1.1879804731198786,
+      "learning_rate": 4.399801436506812e-07,
+      "loss": 0.3873,
+      "step": 1459
+    },
+    {
+      "epoch": 4.5732184808144085,
+      "grad_norm": 1.2110866417839938,
+      "learning_rate": 4.335806320599234e-07,
+      "loss": 0.399,
+      "step": 1460
+    },
+    {
+      "epoch": 4.576350822239624,
+      "grad_norm": 1.2296410077584767,
+      "learning_rate": 4.272269710862587e-07,
+      "loss": 0.3992,
+      "step": 1461
+    },
+    {
+      "epoch": 4.57948316366484,
+      "grad_norm": 1.2328738318463752,
+      "learning_rate": 4.209191911819688e-07,
+      "loss": 0.4196,
+      "step": 1462
+    },
+    {
+      "epoch": 4.582615505090055,
+      "grad_norm": 1.1895802067659933,
+      "learning_rate": 4.1465732257943037e-07,
+      "loss": 0.4496,
+      "step": 1463
+    },
+    {
+      "epoch": 4.58574784651527,
+      "grad_norm": 1.2283635780345825,
+      "learning_rate": 4.084413952909738e-07,
+      "loss": 0.4111,
+      "step": 1464
+    },
+    {
+      "epoch": 4.588880187940486,
+      "grad_norm": 1.2164839835142356,
+      "learning_rate": 4.022714391087379e-07,
+      "loss": 0.381,
+      "step": 1465
+    },
+    {
+      "epoch": 4.592012529365701,
+      "grad_norm": 1.207591372814288,
+      "learning_rate": 3.9614748360452984e-07,
+      "loss": 0.4174,
+      "step": 1466
+    },
+    {
+      "epoch": 4.595144870790916,
+      "grad_norm": 1.2361007397271178,
+      "learning_rate": 3.9006955812967983e-07,
+      "loss": 0.4093,
+      "step": 1467
+    },
+    {
+      "epoch": 4.598277212216131,
+      "grad_norm": 1.2148169973114509,
+      "learning_rate": 3.840376918149025e-07,
+      "loss": 0.4096,
+      "step": 1468
+    },
+    {
+      "epoch": 4.601409553641347,
+      "grad_norm": 1.262330409419549,
+      "learning_rate": 3.780519135701566e-07,
+      "loss": 0.4104,
+      "step": 1469
+    },
+    {
+      "epoch": 4.604541895066562,
+      "grad_norm": 1.2330283137163904,
+      "learning_rate": 3.7211225208450777e-07,
+      "loss": 0.4192,
+      "step": 1470
+    },
+    {
+      "epoch": 4.6076742364917775,
+      "grad_norm": 1.2671239020786587,
+      "learning_rate": 3.6621873582599055e-07,
+      "loss": 0.4054,
+      "step": 1471
+    },
+    {
+      "epoch": 4.610806577916993,
+      "grad_norm": 1.2333107036927184,
+      "learning_rate": 3.603713930414676e-07,
+      "loss": 0.4063,
+      "step": 1472
+    },
+    {
+      "epoch": 4.613938919342209,
+      "grad_norm": 1.202856883096194,
+      "learning_rate": 3.5457025175650084e-07,
+      "loss": 0.4148,
+      "step": 1473
+    },
+    {
+      "epoch": 4.617071260767424,
+      "grad_norm": 1.1747823598797158,
+      "learning_rate": 3.488153397752159e-07,
+      "loss": 0.3891,
+      "step": 1474
+    },
+    {
+      "epoch": 4.620203602192639,
+      "grad_norm": 1.1992072519105093,
+      "learning_rate": 3.431066846801634e-07,
+      "loss": 0.4267,
+      "step": 1475
+    },
+    {
+      "epoch": 4.623335943617854,
+      "grad_norm": 1.273775818849566,
+      "learning_rate": 3.374443138321937e-07,
+      "loss": 0.3985,
+      "step": 1476
+    },
+    {
+      "epoch": 4.62646828504307,
+      "grad_norm": 1.2299835011094196,
+      "learning_rate": 3.3182825437031994e-07,
+      "loss": 0.4089,
+      "step": 1477
+    },
+    {
+      "epoch": 4.629600626468285,
+      "grad_norm": 1.2034138009185376,
+      "learning_rate": 3.262585332115964e-07,
+      "loss": 0.4043,
+      "step": 1478
+    },
+    {
+      "epoch": 4.6327329678935,
+      "grad_norm": 1.209424726509648,
+      "learning_rate": 3.207351770509759e-07,
+      "loss": 0.4147,
+      "step": 1479
+    },
+    {
+      "epoch": 4.635865309318715,
+      "grad_norm": 1.1736004823028015,
+      "learning_rate": 3.152582123611958e-07,
+      "loss": 0.3687,
+      "step": 1480
+    },
+    {
+      "epoch": 4.6389976507439314,
+      "grad_norm": 1.2024178154556568,
+      "learning_rate": 3.098276653926413e-07,
+      "loss": 0.4049,
+      "step": 1481
+    },
+    {
+      "epoch": 4.642129992169147,
+      "grad_norm": 1.235968122606252,
+      "learning_rate": 3.0444356217322893e-07,
+      "loss": 0.4122,
+      "step": 1482
+    },
+    {
+      "epoch": 4.645262333594362,
+      "grad_norm": 1.275106262365568,
+      "learning_rate": 2.9910592850826983e-07,
+      "loss": 0.4086,
+      "step": 1483
+    },
+    {
+      "epoch": 4.648394675019577,
+      "grad_norm": 1.2510666555833756,
+      "learning_rate": 2.9381478998035563e-07,
+      "loss": 0.4055,
+      "step": 1484
+    },
+    {
+      "epoch": 4.651527016444793,
+      "grad_norm": 1.2274467661857713,
+      "learning_rate": 2.8857017194923174e-07,
+      "loss": 0.4034,
+      "step": 1485
+    },
+    {
+      "epoch": 4.654659357870008,
+      "grad_norm": 1.1853349236972364,
+      "learning_rate": 2.833720995516798e-07,
+      "loss": 0.4195,
+      "step": 1486
+    },
+    {
+      "epoch": 4.657791699295223,
+      "grad_norm": 1.1854238027175328,
+      "learning_rate": 2.7822059770138966e-07,
+      "loss": 0.4074,
+      "step": 1487
+    },
+    {
+      "epoch": 4.660924040720438,
+      "grad_norm": 1.2436312130461868,
+      "learning_rate": 2.731156910888477e-07,
+      "loss": 0.4033,
+      "step": 1488
+    },
+    {
+      "epoch": 4.664056382145654,
+      "grad_norm": 1.2428796817271779,
+      "learning_rate": 2.6805740418121453e-07,
+      "loss": 0.4191,
+      "step": 1489
+    },
+    {
+      "epoch": 4.667188723570869,
+      "grad_norm": 1.2427783356951405,
+      "learning_rate": 2.6304576122221035e-07,
+      "loss": 0.3755,
+      "step": 1490
+    },
+    {
+      "epoch": 4.6703210649960845,
+      "grad_norm": 1.1881132036518127,
+      "learning_rate": 2.580807862319912e-07,
+      "loss": 0.3891,
+      "step": 1491
+    },
+    {
+      "epoch": 4.6734534064213,
+      "grad_norm": 1.2442335114182685,
+      "learning_rate": 2.5316250300704726e-07,
+      "loss": 0.4154,
+      "step": 1492
+    },
+    {
+      "epoch": 4.676585747846516,
+      "grad_norm": 1.2053182135162017,
+      "learning_rate": 2.482909351200735e-07,
+      "loss": 0.3979,
+      "step": 1493
+    },
+    {
+      "epoch": 4.679718089271731,
+      "grad_norm": 1.1770238326901394,
+      "learning_rate": 2.434661059198695e-07,
+      "loss": 0.3948,
+      "step": 1494
+    },
+    {
+      "epoch": 4.682850430696946,
+      "grad_norm": 1.218619191692209,
+      "learning_rate": 2.386880385312218e-07,
+      "loss": 0.371,
+      "step": 1495
+    },
+    {
+      "epoch": 4.685982772122161,
+      "grad_norm": 1.1637502072533377,
+      "learning_rate": 2.339567558547906e-07,
+      "loss": 0.3917,
+      "step": 1496
+    },
+    {
+      "epoch": 4.689115113547377,
+      "grad_norm": 1.1735101557117311,
+      "learning_rate": 2.292722805670078e-07,
+      "loss": 0.4077,
+      "step": 1497
+    },
+    {
+      "epoch": 4.692247454972592,
+      "grad_norm": 1.2417399018426394,
+      "learning_rate": 2.2463463511995908e-07,
+      "loss": 0.4048,
+      "step": 1498
+    },
+    {
+      "epoch": 4.695379796397807,
+      "grad_norm": 1.1728907713133705,
+      "learning_rate": 2.200438417412809e-07,
+      "loss": 0.3891,
+      "step": 1499
+    },
+    {
+      "epoch": 4.698512137823022,
+      "grad_norm": 1.2016608119778363,
+      "learning_rate": 2.1549992243405814e-07,
+      "loss": 0.3882,
+      "step": 1500
+    },
+    {
+      "epoch": 4.701644479248238,
+      "grad_norm": 1.206612395808125,
+      "learning_rate": 2.1100289897670877e-07,
+      "loss": 0.4196,
+      "step": 1501
+    },
+    {
+      "epoch": 4.7047768206734535,
+      "grad_norm": 1.2493616742742244,
+      "learning_rate": 2.0655279292288832e-07,
+      "loss": 0.3967,
+      "step": 1502
+    },
+    {
+      "epoch": 4.707909162098669,
+      "grad_norm": 1.2226247159036543,
+      "learning_rate": 2.021496256013833e-07,
+      "loss": 0.3989,
+      "step": 1503
+    },
+    {
+      "epoch": 4.711041503523884,
+      "grad_norm": 1.1794232427283857,
+      "learning_rate": 1.9779341811600795e-07,
+      "loss": 0.3988,
+      "step": 1504
+    },
+    {
+      "epoch": 4.7141738449491,
+      "grad_norm": 1.2359058049155338,
+      "learning_rate": 1.9348419134550322e-07,
+      "loss": 0.4213,
+      "step": 1505
+    },
+    {
+      "epoch": 4.717306186374315,
+      "grad_norm": 1.2239539416305099,
+      "learning_rate": 1.8922196594344023e-07,
+      "loss": 0.3922,
+      "step": 1506
+    },
+    {
+      "epoch": 4.72043852779953,
+      "grad_norm": 1.2127862487844647,
+      "learning_rate": 1.8500676233811466e-07,
+      "loss": 0.3771,
+      "step": 1507
+    },
+    {
+      "epoch": 4.723570869224745,
+      "grad_norm": 1.198843093205888,
+      "learning_rate": 1.8083860073245806e-07,
+      "loss": 0.4181,
+      "step": 1508
+    },
+    {
+      "epoch": 4.726703210649961,
+      "grad_norm": 1.2252062247390472,
+      "learning_rate": 1.7671750110392903e-07,
+      "loss": 0.403,
+      "step": 1509
+    },
+    {
+      "epoch": 4.729835552075176,
+      "grad_norm": 1.23819671302605,
+      "learning_rate": 1.7264348320442992e-07,
+      "loss": 0.4092,
+      "step": 1510
+    },
+    {
+      "epoch": 4.732967893500391,
+      "grad_norm": 1.2113129084206005,
+      "learning_rate": 1.6861656656020464e-07,
+      "loss": 0.3774,
+      "step": 1511
+    },
+    {
+      "epoch": 4.7361002349256065,
+      "grad_norm": 1.1845710084452867,
+      "learning_rate": 1.6463677047174553e-07,
+      "loss": 0.3841,
+      "step": 1512
+    },
+    {
+      "epoch": 4.7392325763508225,
+      "grad_norm": 1.1746238547350345,
+      "learning_rate": 1.6070411401370335e-07,
+      "loss": 0.3939,
+      "step": 1513
+    },
+    {
+      "epoch": 4.742364917776038,
+      "grad_norm": 1.212119626132927,
+      "learning_rate": 1.568186160347951e-07,
+      "loss": 0.3838,
+      "step": 1514
+    },
+    {
+      "epoch": 4.745497259201253,
+      "grad_norm": 1.2178905084917646,
+      "learning_rate": 1.5298029515771195e-07,
+      "loss": 0.4132,
+      "step": 1515
+    },
+    {
+      "epoch": 4.748629600626469,
+      "grad_norm": 1.2150279012900946,
+      "learning_rate": 1.4918916977903264e-07,
+      "loss": 0.3986,
+      "step": 1516
+    },
+    {
+      "epoch": 4.751761942051684,
+      "grad_norm": 1.2367095696943584,
+      "learning_rate": 1.4544525806913234e-07,
+      "loss": 0.3979,
+      "step": 1517
+    },
+    {
+      "epoch": 4.754894283476899,
+      "grad_norm": 1.2377643398696656,
+      "learning_rate": 1.4174857797209951e-07,
+      "loss": 0.4018,
+      "step": 1518
+    },
+    {
+      "epoch": 4.758026624902114,
+      "grad_norm": 1.2167876153278898,
+      "learning_rate": 1.3809914720564478e-07,
+      "loss": 0.3906,
+      "step": 1519
+    },
+    {
+      "epoch": 4.761158966327329,
+      "grad_norm": 1.2446471369617926,
+      "learning_rate": 1.344969832610199e-07,
+      "loss": 0.3756,
+      "step": 1520
+    },
+    {
+      "epoch": 4.764291307752545,
+      "grad_norm": 1.1893090822752832,
+      "learning_rate": 1.3094210340293456e-07,
+      "loss": 0.4165,
+      "step": 1521
+    },
+    {
+      "epoch": 4.76742364917776,
+      "grad_norm": 1.2320387940829778,
+      "learning_rate": 1.274345246694708e-07,
+      "loss": 0.3913,
+      "step": 1522
+    },
+    {
+      "epoch": 4.7705559906029755,
+      "grad_norm": 1.216546020584182,
+      "learning_rate": 1.2397426387200096e-07,
+      "loss": 0.3959,
+      "step": 1523
+    },
+    {
+      "epoch": 4.7736883320281915,
+      "grad_norm": 1.2176186473887705,
+      "learning_rate": 1.205613375951098e-07,
+      "loss": 0.4247,
+      "step": 1524
+    },
+    {
+      "epoch": 4.776820673453407,
+      "grad_norm": 1.2283180997381093,
+      "learning_rate": 1.1719576219651585e-07,
+      "loss": 0.418,
+      "step": 1525
+    },
+    {
+      "epoch": 4.779953014878622,
+      "grad_norm": 1.214550392378903,
+      "learning_rate": 1.1387755380698584e-07,
+      "loss": 0.3819,
+      "step": 1526
+    },
+    {
+      "epoch": 4.783085356303837,
+      "grad_norm": 1.2324975421906008,
+      "learning_rate": 1.1060672833026697e-07,
+      "loss": 0.4278,
+      "step": 1527
+    },
+    {
+      "epoch": 4.786217697729052,
+      "grad_norm": 1.238934876307768,
+      "learning_rate": 1.0738330144300479e-07,
+      "loss": 0.425,
+      "step": 1528
+    },
+    {
+      "epoch": 4.789350039154268,
+      "grad_norm": 1.181886288925798,
+      "learning_rate": 1.0420728859466988e-07,
+      "loss": 0.4209,
+      "step": 1529
+    },
+    {
+      "epoch": 4.792482380579483,
+      "grad_norm": 1.1925253947318342,
+      "learning_rate": 1.0107870500748351e-07,
+      "loss": 0.3773,
+      "step": 1530
+    },
+    {
+      "epoch": 4.795614722004698,
+      "grad_norm": 1.2509864343113049,
+      "learning_rate": 9.799756567634433e-08,
+      "loss": 0.3853,
+      "step": 1531
+    },
+    {
+      "epoch": 4.798747063429914,
+      "grad_norm": 1.184682423863237,
+      "learning_rate": 9.496388536875623e-08,
+      "loss": 0.417,
+      "step": 1532
+    },
+    {
+      "epoch": 4.801879404855129,
+      "grad_norm": 1.2246944998945095,
+      "learning_rate": 9.197767862475837e-08,
+      "loss": 0.3874,
+      "step": 1533
+    },
+    {
+      "epoch": 4.8050117462803446,
+      "grad_norm": 1.214520286942335,
+      "learning_rate": 8.903895975685528e-08,
+      "loss": 0.4037,
+      "step": 1534
+    },
+    {
+      "epoch": 4.80814408770556,
+      "grad_norm": 1.2224353069074514,
+      "learning_rate": 8.614774284994797e-08,
+      "loss": 0.4099,
+      "step": 1535
+    },
+    {
+      "epoch": 4.811276429130775,
+      "grad_norm": 1.2068932961719712,
+      "learning_rate": 8.33040417612685e-08,
+      "loss": 0.3891,
+      "step": 1536
+    },
+    {
+      "epoch": 4.814408770555991,
+      "grad_norm": 1.1952441796807236,
+      "learning_rate": 8.05078701203088e-08,
+      "loss": 0.3899,
+      "step": 1537
+    },
+    {
+      "epoch": 4.817541111981206,
+      "grad_norm": 1.226741893844113,
+      "learning_rate": 7.7759241328762e-08,
+      "loss": 0.4111,
+      "step": 1538
+    },
+    {
+      "epoch": 4.820673453406421,
+      "grad_norm": 1.1808940728419945,
+      "learning_rate": 7.505816856045012e-08,
+      "loss": 0.3943,
+      "step": 1539
+    },
+    {
+      "epoch": 4.823805794831637,
+      "grad_norm": 1.203115331813816,
+      "learning_rate": 7.24046647612675e-08,
+      "loss": 0.4127,
+      "step": 1540
+    },
+    {
+      "epoch": 4.826938136256852,
+      "grad_norm": 1.1865631093267626,
+      "learning_rate": 6.979874264911756e-08,
+      "loss": 0.4058,
+      "step": 1541
+    },
+    {
+      "epoch": 4.830070477682067,
+      "grad_norm": 1.185030215141837,
+      "learning_rate": 6.724041471384835e-08,
+      "loss": 0.4124,
+      "step": 1542
+    },
+    {
+      "epoch": 4.8332028191072824,
+      "grad_norm": 1.204462154369856,
+      "learning_rate": 6.472969321719702e-08,
+      "loss": 0.4326,
+      "step": 1543
+    },
+    {
+      "epoch": 4.836335160532498,
+      "grad_norm": 1.2203256822089201,
+      "learning_rate": 6.226659019272997e-08,
+      "loss": 0.3917,
+      "step": 1544
+    },
+    {
+      "epoch": 4.839467501957714,
+      "grad_norm": 1.2180754919299253,
+      "learning_rate": 5.985111744578165e-08,
+      "loss": 0.4013,
+      "step": 1545
+    },
+    {
+      "epoch": 4.842599843382929,
+      "grad_norm": 1.18445065771052,
+      "learning_rate": 5.748328655340141e-08,
+      "loss": 0.3912,
+      "step": 1546
+    },
+    {
+      "epoch": 4.845732184808144,
+      "grad_norm": 1.1965989755685764,
+      "learning_rate": 5.516310886429899e-08,
+      "loss": 0.4041,
+      "step": 1547
+    },
+    {
+      "epoch": 4.84886452623336,
+      "grad_norm": 1.193838779871222,
+      "learning_rate": 5.2890595498787944e-08,
+      "loss": 0.4006,
+      "step": 1548
+    },
+    {
+      "epoch": 4.851996867658575,
+      "grad_norm": 1.1686189895653558,
+      "learning_rate": 5.0665757348732355e-08,
+      "loss": 0.4082,
+      "step": 1549
+    },
+    {
+      "epoch": 4.85512920908379,
+      "grad_norm": 1.2058648478611114,
+      "learning_rate": 4.848860507749353e-08,
+      "loss": 0.3881,
+      "step": 1550
+    },
+    {
+      "epoch": 4.858261550509005,
+      "grad_norm": 1.1637318481244103,
+      "learning_rate": 4.635914911988448e-08,
+      "loss": 0.3942,
+      "step": 1551
+    },
+    {
+      "epoch": 4.86139389193422,
+      "grad_norm": 1.2165645886224812,
+      "learning_rate": 4.427739968211331e-08,
+      "loss": 0.4035,
+      "step": 1552
+    },
+    {
+      "epoch": 4.864526233359436,
+      "grad_norm": 1.2040433534371484,
+      "learning_rate": 4.2243366741735457e-08,
+      "loss": 0.3868,
+      "step": 1553
+    },
+    {
+      "epoch": 4.8676585747846515,
+      "grad_norm": 1.2369435812923857,
+      "learning_rate": 4.025706004760932e-08,
+      "loss": 0.3988,
+      "step": 1554
+    },
+    {
+      "epoch": 4.870790916209867,
+      "grad_norm": 1.2232422405149723,
+      "learning_rate": 3.831848911984959e-08,
+      "loss": 0.3968,
+      "step": 1555
+    },
+    {
+      "epoch": 4.873923257635083,
+      "grad_norm": 1.1711459811080631,
+      "learning_rate": 3.642766324977509e-08,
+      "loss": 0.3923,
+      "step": 1556
+    },
+    {
+      "epoch": 4.877055599060298,
+      "grad_norm": 1.2174185734188516,
+      "learning_rate": 3.4584591499872146e-08,
+      "loss": 0.4029,
+      "step": 1557
+    },
+    {
+      "epoch": 4.880187940485513,
+      "grad_norm": 1.1614173175371763,
+      "learning_rate": 3.278928270374459e-08,
+      "loss": 0.3949,
+      "step": 1558
+    },
+    {
+      "epoch": 4.883320281910728,
+      "grad_norm": 1.2280723648931007,
+      "learning_rate": 3.104174546607941e-08,
+      "loss": 0.4199,
+      "step": 1559
+    },
+    {
+      "epoch": 4.886452623335944,
+      "grad_norm": 1.209148320604221,
+      "learning_rate": 2.9341988162595593e-08,
+      "loss": 0.4192,
+      "step": 1560
+    },
+    {
+      "epoch": 4.889584964761159,
+      "grad_norm": 1.2065698817517443,
+      "learning_rate": 2.7690018940011996e-08,
+      "loss": 0.3655,
+      "step": 1561
+    },
+    {
+      "epoch": 4.892717306186374,
+      "grad_norm": 1.1832085868374411,
+      "learning_rate": 2.6085845716004034e-08,
+      "loss": 0.4102,
+      "step": 1562
+    },
+    {
+      "epoch": 4.895849647611589,
+      "grad_norm": 1.223809874203652,
+      "learning_rate": 2.452947617916701e-08,
+      "loss": 0.3826,
+      "step": 1563
+    },
+    {
+      "epoch": 4.898981989036805,
+      "grad_norm": 1.1564398514888772,
+      "learning_rate": 2.3020917788981744e-08,
+      "loss": 0.399,
+      "step": 1564
+    },
+    {
+      "epoch": 4.9021143304620205,
+      "grad_norm": 1.2322802342710437,
+      "learning_rate": 2.156017777577346e-08,
+      "loss": 0.4385,
+      "step": 1565
+    },
+    {
+      "epoch": 4.905246671887236,
+      "grad_norm": 1.2327395751465036,
+      "learning_rate": 2.0147263140680717e-08,
+      "loss": 0.3948,
+      "step": 1566
+    },
+    {
+      "epoch": 4.908379013312451,
+      "grad_norm": 1.1755915458959958,
+      "learning_rate": 1.8782180655622096e-08,
+      "loss": 0.3909,
+      "step": 1567
+    },
+    {
+      "epoch": 4.911511354737667,
+      "grad_norm": 1.2040808443639759,
+      "learning_rate": 1.7464936863261785e-08,
+      "loss": 0.4109,
+      "step": 1568
+    },
+    {
+      "epoch": 4.914643696162882,
+      "grad_norm": 1.2313492049462424,
+      "learning_rate": 1.61955380769796e-08,
+      "loss": 0.3865,
+      "step": 1569
+    },
+    {
+      "epoch": 4.917776037588097,
+      "grad_norm": 1.2358379850928518,
+      "learning_rate": 1.497399038084102e-08,
+      "loss": 0.4126,
+      "step": 1570
+    },
+    {
+      "epoch": 4.920908379013312,
+      "grad_norm": 1.2159888247443027,
+      "learning_rate": 1.3800299629568303e-08,
+      "loss": 0.3979,
+      "step": 1571
+    },
+    {
+      "epoch": 4.924040720438528,
+      "grad_norm": 1.2314356631041403,
+      "learning_rate": 1.2674471448508308e-08,
+      "loss": 0.3954,
+      "step": 1572
+    },
+    {
+      "epoch": 4.927173061863743,
+      "grad_norm": 1.1763805345232639,
+      "learning_rate": 1.1596511233610275e-08,
+      "loss": 0.4174,
+      "step": 1573
+    },
+    {
+      "epoch": 4.930305403288958,
+      "grad_norm": 1.21950132315641,
+      "learning_rate": 1.0566424151401412e-08,
+      "loss": 0.3712,
+      "step": 1574
+    },
+    {
+      "epoch": 4.9334377447141735,
+      "grad_norm": 1.1857749532551431,
+      "learning_rate": 9.584215138953579e-09,
+      "loss": 0.4018,
+      "step": 1575
+    },
+    {
+      "epoch": 4.9365700861393895,
+      "grad_norm": 1.1885549206002513,
+      "learning_rate": 8.649888903869973e-09,
+      "loss": 0.4014,
+      "step": 1576
+    },
+    {
+      "epoch": 4.939702427564605,
+      "grad_norm": 1.2601185286531782,
+      "learning_rate": 7.763449924256262e-09,
+      "loss": 0.3956,
+      "step": 1577
+    },
+    {
+      "epoch": 4.94283476898982,
+      "grad_norm": 1.1994949171986244,
+      "learning_rate": 6.924902448699478e-09,
+      "loss": 0.4229,
+      "step": 1578
+    },
+    {
+      "epoch": 4.945967110415035,
+      "grad_norm": 1.231920206414556,
+      "learning_rate": 6.134250496249161e-09,
+      "loss": 0.4123,
+      "step": 1579
+    },
+    {
+      "epoch": 4.949099451840251,
+      "grad_norm": 1.231501336996204,
+      "learning_rate": 5.391497856399585e-09,
+      "loss": 0.4019,
+      "step": 1580
+    },
+    {
+      "epoch": 4.952231793265466,
+      "grad_norm": 1.2025254247538497,
+      "learning_rate": 4.696648089068667e-09,
+      "loss": 0.4047,
+      "step": 1581
+    },
+    {
+      "epoch": 4.955364134690681,
+      "grad_norm": 1.2462699073996386,
+      "learning_rate": 4.0497045245813105e-09,
+      "loss": 0.3905,
+      "step": 1582
+    },
+    {
+      "epoch": 4.958496476115896,
+      "grad_norm": 1.2150582015348728,
+      "learning_rate": 3.450670263654976e-09,
+      "loss": 0.4034,
+      "step": 1583
+    },
+    {
+      "epoch": 4.961628817541112,
+      "grad_norm": 1.1657154816919872,
+      "learning_rate": 2.899548177384137e-09,
+      "loss": 0.3983,
+      "step": 1584
+    },
+    {
+      "epoch": 4.964761158966327,
+      "grad_norm": 1.201657869580201,
+      "learning_rate": 2.396340907225847e-09,
+      "loss": 0.3985,
+      "step": 1585
+    },
+    {
+      "epoch": 4.9678935003915425,
+      "grad_norm": 1.1936954506770183,
+      "learning_rate": 1.9410508649875258e-09,
+      "loss": 0.4038,
+      "step": 1586
+    },
+    {
+      "epoch": 4.971025841816758,
+      "grad_norm": 1.1919200016703124,
+      "learning_rate": 1.5336802328147492e-09,
+      "loss": 0.3716,
+      "step": 1587
+    },
+    {
+      "epoch": 4.974158183241974,
+      "grad_norm": 1.2304919494528042,
+      "learning_rate": 1.1742309631845861e-09,
+      "loss": 0.4267,
+      "step": 1588
+    },
+    {
+      "epoch": 4.977290524667189,
+      "grad_norm": 1.2084987865361176,
+      "learning_rate": 8.627047788911658e-10,
+      "loss": 0.443,
+      "step": 1589
+    },
+    {
+      "epoch": 4.980422866092404,
+      "grad_norm": 1.2473534930975694,
+      "learning_rate": 5.991031730367968e-10,
+      "loss": 0.4292,
+      "step": 1590
+    },
+    {
+      "epoch": 4.983555207517619,
+      "grad_norm": 1.2137391828259563,
+      "learning_rate": 3.8342740903307697e-10,
+      "loss": 0.4086,
+      "step": 1591
+    },
+    {
+      "epoch": 4.986687548942835,
+      "grad_norm": 1.239277828293045,
+      "learning_rate": 2.1567852058423932e-10,
+      "loss": 0.3963,
+      "step": 1592
+    },
+    {
+      "epoch": 4.98981989036805,
+      "grad_norm": 1.198137612800272,
+      "learning_rate": 9.585731168937351e-11,
+      "loss": 0.3968,
+      "step": 1593
+    },
+    {
+      "epoch": 4.992952231793265,
+      "grad_norm": 1.2102944840840637,
+      "learning_rate": 2.396435663687413e-11,
+      "loss": 0.3788,
+      "step": 1594
+    },
+    {
+      "epoch": 4.99608457321848,
+      "grad_norm": 1.2188363908623323,
+      "learning_rate": 0.0,
+      "loss": 0.3945,
+      "step": 1595
+    },
+    {
+      "epoch": 4.99608457321848,
+      "step": 1595,
+      "total_flos": 1.3245374097522688e+18,
+      "train_loss": 0.0,
+      "train_runtime": 1.8638,
+      "train_samples_per_second": 54818.89,
+      "train_steps_per_second": 855.791
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1595,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3245374097522688e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}