diff --git "a/checkpoint-1695/trainer_state.json" "b/checkpoint-1695/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1695/trainer_state.json"
@@ -0,0 +1,11982 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9986731534719153,
+  "eval_steps": 142,
+  "global_step": 1695,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.6921162605285645,
+      "learning_rate": 5e-05,
+      "loss": 3.3182,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 3.3362529277801514,
+      "eval_runtime": 14.4412,
+      "eval_samples_per_second": 33.031,
+      "eval_steps_per_second": 8.31,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.620742321014404,
+      "learning_rate": 0.0001,
+      "loss": 3.2788,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.650161266326904,
+      "learning_rate": 0.00015,
+      "loss": 3.2271,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.024933815002441,
+      "learning_rate": 0.0002,
+      "loss": 2.402,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.751981496810913,
+      "learning_rate": 0.00025,
+      "loss": 1.0544,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4766970872879028,
+      "learning_rate": 0.0003,
+      "loss": 0.3549,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8064658641815186,
+      "learning_rate": 0.00035,
+      "loss": 0.1533,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.232205390930176,
+      "learning_rate": 0.0004,
+      "loss": 0.31,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1898847818374634,
+      "learning_rate": 0.00045000000000000004,
+      "loss": 0.1818,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7394833564758301,
+      "learning_rate": 0.0005,
+      "loss": 0.1751,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.16317571699619293,
+      "learning_rate": 0.0004999995654799487,
+      "loss": 0.1411,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.10235322266817093,
+      "learning_rate": 0.0004999982619213052,
+      "loss": 0.1363,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.19907887279987335,
+      "learning_rate": 0.0004999960893286008,
+      "loss": 0.128,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6823816299438477,
+      "learning_rate": 0.0004999930477093878,
+      "loss": 0.143,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.10187644511461258,
+      "learning_rate": 0.0004999891370742394,
+      "loss": 0.1322,
+      "step": 15
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.09401004016399384,
+      "learning_rate": 0.0004999843574367498,
+      "loss": 0.1361,
+      "step": 16
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17946797609329224,
+      "learning_rate": 0.0004999787088135334,
+      "loss": 0.1412,
+      "step": 17
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.890545666217804,
+      "learning_rate": 0.0004999721912242259,
+      "loss": 0.1593,
+      "step": 18
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.434042751789093,
+      "learning_rate": 0.0004999648046914836,
+      "loss": 0.1548,
+      "step": 19
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28103551268577576,
+      "learning_rate": 0.0004999565492409831,
+      "loss": 0.1459,
+      "step": 20
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2690610885620117,
+      "learning_rate": 0.0004999474249014217,
+      "loss": 0.1248,
+      "step": 21
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.37668731808662415,
+      "learning_rate": 0.0004999374317045172,
+      "loss": 0.1481,
+      "step": 22
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23762015998363495,
+      "learning_rate": 0.0004999265696850074,
+      "loss": 0.1407,
+      "step": 23
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1988176554441452,
+      "learning_rate": 0.0004999148388806504,
+      "loss": 0.1398,
+      "step": 24
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3805619776248932,
+      "learning_rate": 0.0004999022393322246,
+      "loss": 0.1474,
+      "step": 25
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5069771409034729,
+      "learning_rate": 0.0004998887710835278,
+      "loss": 0.1509,
+      "step": 26
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42066043615341187,
+      "learning_rate": 0.0004998744341813779,
+      "loss": 0.1341,
+      "step": 27
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0944904088973999,
+      "learning_rate": 0.0004998592286756123,
+      "loss": 0.1233,
+      "step": 28
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.849244236946106,
+      "learning_rate": 0.0004998431546190875,
+      "loss": 0.1999,
+      "step": 29
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.09785456210374832,
+      "learning_rate": 0.00049982621206768,
+      "loss": 0.1272,
+      "step": 30
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38225457072257996,
+      "learning_rate": 0.0004998084010802845,
+      "loss": 0.1634,
+      "step": 31
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.08135183900594711,
+      "learning_rate": 0.0004997897217188149,
+      "loss": 0.1383,
+      "step": 32
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.17299437522888184,
+      "learning_rate": 0.0004997701740482036,
+      "loss": 0.1427,
+      "step": 33
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.11125747114419937,
+      "learning_rate": 0.0004997497581364015,
+      "loss": 0.1379,
+      "step": 34
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.08914893865585327,
+      "learning_rate": 0.0004997284740543776,
+      "loss": 0.1388,
+      "step": 35
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.034590039402246475,
+      "learning_rate": 0.0004997063218761188,
+      "loss": 0.1387,
+      "step": 36
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.08675777167081833,
+      "learning_rate": 0.0004996833016786296,
+      "loss": 0.1384,
+      "step": 37
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4440009295940399,
+      "learning_rate": 0.0004996594135419318,
+      "loss": 0.152,
+      "step": 38
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0814109519124031,
+      "learning_rate": 0.0004996346575490646,
+      "loss": 0.1373,
+      "step": 39
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.37724560499191284,
+      "learning_rate": 0.0004996090337860836,
+      "loss": 0.1362,
+      "step": 40
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.21177273988723755,
+      "learning_rate": 0.0004995825423420613,
+      "loss": 0.1423,
+      "step": 41
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12168041616678238,
+      "learning_rate": 0.000499555183309086,
+      "loss": 0.1381,
+      "step": 42
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.21096466481685638,
+      "learning_rate": 0.0004995269567822623,
+      "loss": 0.139,
+      "step": 43
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.49395841360092163,
+      "learning_rate": 0.0004994978628597099,
+      "loss": 0.1016,
+      "step": 44
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1108216792345047,
+      "learning_rate": 0.0004994679016425642,
+      "loss": 0.1334,
+      "step": 45
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5518127679824829,
+      "learning_rate": 0.000499437073234975,
+      "loss": 0.1568,
+      "step": 46
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4762812852859497,
+      "learning_rate": 0.0004994053777441069,
+      "loss": 0.1543,
+      "step": 47
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3477722108364105,
+      "learning_rate": 0.0004993728152801384,
+      "loss": 0.1464,
+      "step": 48
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4996407628059387,
+      "learning_rate": 0.0004993393859562621,
+      "loss": 0.154,
+      "step": 49
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.20425601303577423,
+      "learning_rate": 0.0004993050898886833,
+      "loss": 0.1372,
+      "step": 50
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.034631408751010895,
+      "learning_rate": 0.000499269927196621,
+      "loss": 0.137,
+      "step": 51
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.24027873575687408,
+      "learning_rate": 0.0004992338980023062,
+      "loss": 0.1468,
+      "step": 52
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.22242723405361176,
+      "learning_rate": 0.000499197002430982,
+      "loss": 0.1418,
+      "step": 53
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6540514826774597,
+      "learning_rate": 0.0004991592406109036,
+      "loss": 0.1564,
+      "step": 54
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.030118577182292938,
+      "learning_rate": 0.000499120612673337,
+      "loss": 0.1365,
+      "step": 55
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.07544097304344177,
+      "learning_rate": 0.0004990811187525592,
+      "loss": 0.1334,
+      "step": 56
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37415480613708496,
+      "learning_rate": 0.0004990407589858572,
+      "loss": 0.155,
+      "step": 57
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.557809054851532,
+      "learning_rate": 0.0004989995335135282,
+      "loss": 0.1603,
+      "step": 58
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14802873134613037,
+      "learning_rate": 0.0004989574424788787,
+      "loss": 0.1387,
+      "step": 59
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3581993281841278,
+      "learning_rate": 0.0004989144860282236,
+      "loss": 0.1475,
+      "step": 60
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.04818522185087204,
+      "learning_rate": 0.0004988706643108864,
+      "loss": 0.1362,
+      "step": 61
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.21908174455165863,
+      "learning_rate": 0.0004988259774791987,
+      "loss": 0.1386,
+      "step": 62
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1852695643901825,
+      "learning_rate": 0.0004987804256884988,
+      "loss": 0.1387,
+      "step": 63
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.025747304782271385,
+      "learning_rate": 0.0004987340090971323,
+      "loss": 0.1393,
+      "step": 64
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.045346710830926895,
+      "learning_rate": 0.0004986867278664504,
+      "loss": 0.1354,
+      "step": 65
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.34946465492248535,
+      "learning_rate": 0.0004986385821608106,
+      "loss": 0.152,
+      "step": 66
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2552882432937622,
+      "learning_rate": 0.0004985895721475748,
+      "loss": 0.1463,
+      "step": 67
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0560542456805706,
+      "learning_rate": 0.0004985396979971099,
+      "loss": 0.1391,
+      "step": 68
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.14347511529922485,
+      "learning_rate": 0.0004984889598827863,
+      "loss": 0.1353,
+      "step": 69
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.12386342883110046,
+      "learning_rate": 0.0004984373579809778,
+      "loss": 0.1343,
+      "step": 70
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.03070697747170925,
+      "learning_rate": 0.000498384892471061,
+      "loss": 0.1356,
+      "step": 71
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0531514473259449,
+      "learning_rate": 0.0004983315635354144,
+      "loss": 0.1346,
+      "step": 72
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24197503924369812,
+      "learning_rate": 0.0004982773713594178,
+      "loss": 0.1217,
+      "step": 73
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.08417380601167679,
+      "learning_rate": 0.0004982223161314522,
+      "loss": 0.1223,
+      "step": 74
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.40045711398124695,
+      "learning_rate": 0.000498166398042898,
+      "loss": 0.1513,
+      "step": 75
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.12452740222215652,
+      "learning_rate": 0.0004981096172881358,
+      "loss": 0.1296,
+      "step": 76
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.21590262651443481,
+      "learning_rate": 0.0004980519740645444,
+      "loss": 0.1375,
+      "step": 77
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.07704459875822067,
+      "learning_rate": 0.0004979934685725011,
+      "loss": 0.1299,
+      "step": 78
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28334081172943115,
+      "learning_rate": 0.0004979341010153801,
+      "loss": 0.1387,
+      "step": 79
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.12374007701873779,
+      "learning_rate": 0.0004978738715995526,
+      "loss": 0.1383,
+      "step": 80
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.040613267570734024,
+      "learning_rate": 0.000497812780534386,
+      "loss": 0.1367,
+      "step": 81
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.09974126517772675,
+      "learning_rate": 0.0004977508280322423,
+      "loss": 0.1248,
+      "step": 82
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2616259753704071,
+      "learning_rate": 0.0004976880143084786,
+      "loss": 0.1311,
+      "step": 83
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15635579824447632,
+      "learning_rate": 0.0004976243395814452,
+      "loss": 0.1189,
+      "step": 84
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.259250670671463,
+      "learning_rate": 0.000497559804072486,
+      "loss": 0.1099,
+      "step": 85
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2778699398040771,
+      "learning_rate": 0.0004974944080059365,
+      "loss": 0.1416,
+      "step": 86
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2155281901359558,
+      "learning_rate": 0.000497428151609124,
+      "loss": 0.1253,
+      "step": 87
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.17533721029758453,
+      "learning_rate": 0.0004973610351123664,
+      "loss": 0.1446,
+      "step": 88
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.07620590180158615,
+      "learning_rate": 0.0004972930587489715,
+      "loss": 0.1309,
+      "step": 89
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2370779663324356,
+      "learning_rate": 0.0004972242227552358,
+      "loss": 0.149,
+      "step": 90
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.06374065577983856,
+      "learning_rate": 0.0004971545273704444,
+      "loss": 0.1307,
+      "step": 91
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22728750109672546,
+      "learning_rate": 0.0004970839728368697,
+      "loss": 0.1438,
+      "step": 92
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16872233152389526,
+      "learning_rate": 0.0004970125593997706,
+      "loss": 0.1364,
+      "step": 93
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.18773947656154633,
+      "learning_rate": 0.0004969402873073914,
+      "loss": 0.146,
+      "step": 94
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1468167006969452,
+      "learning_rate": 0.0004968671568109616,
+      "loss": 0.1401,
+      "step": 95
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.09030504524707794,
+      "learning_rate": 0.0004967931681646948,
+      "loss": 0.1318,
+      "step": 96
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.061796192079782486,
+      "learning_rate": 0.000496718321625787,
+      "loss": 0.1244,
+      "step": 97
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.045495226979255676,
+      "learning_rate": 0.0004966426174544171,
+      "loss": 0.1265,
+      "step": 98
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.08449025452136993,
+      "learning_rate": 0.0004965660559137448,
+      "loss": 0.1276,
+      "step": 99
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09982559829950333,
+      "learning_rate": 0.0004964886372699102,
+      "loss": 0.1253,
+      "step": 100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.05831208825111389,
+      "learning_rate": 0.0004964103617920332,
+      "loss": 0.1271,
+      "step": 101
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.20548835396766663,
+      "learning_rate": 0.0004963312297522116,
+      "loss": 0.1415,
+      "step": 102
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.09664470702409744,
+      "learning_rate": 0.0004962512414255214,
+      "loss": 0.1083,
+      "step": 103
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16931602358818054,
+      "learning_rate": 0.0004961703970900145,
+      "loss": 0.1431,
+      "step": 104
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.10859667509794235,
+      "learning_rate": 0.000496088697026719,
+      "loss": 0.12,
+      "step": 105
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.21958191692829132,
+      "learning_rate": 0.0004960061415196374,
+      "loss": 0.1492,
+      "step": 106
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.06437578052282333,
+      "learning_rate": 0.0004959227308557459,
+      "loss": 0.1083,
+      "step": 107
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.14975550770759583,
+      "learning_rate": 0.0004958384653249932,
+      "loss": 0.1155,
+      "step": 108
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.11868852376937866,
+      "learning_rate": 0.0004957533452203,
+      "loss": 0.1237,
+      "step": 109
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2610260546207428,
+      "learning_rate": 0.0004956673708375574,
+      "loss": 0.1264,
+      "step": 110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.378467321395874,
+      "learning_rate": 0.000495580542475626,
+      "loss": 0.1222,
+      "step": 111
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0926096960902214,
+      "learning_rate": 0.0004954928604363352,
+      "loss": 0.1096,
+      "step": 112
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.06858692318201065,
+      "learning_rate": 0.0004954043250244819,
+      "loss": 0.1144,
+      "step": 113
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3068992495536804,
+      "learning_rate": 0.0004953149365478293,
+      "loss": 0.1563,
+      "step": 114
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15458936989307404,
+      "learning_rate": 0.0004952246953171061,
+      "loss": 0.1216,
+      "step": 115
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.10287577658891678,
+      "learning_rate": 0.0004951336016460053,
+      "loss": 0.0893,
+      "step": 116
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1215134710073471,
+      "learning_rate": 0.0004950416558511833,
+      "loss": 0.1016,
+      "step": 117
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1392650604248047,
+      "learning_rate": 0.000494948858252258,
+      "loss": 0.1111,
+      "step": 118
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4350431263446808,
+      "learning_rate": 0.0004948552091718092,
+      "loss": 0.1192,
+      "step": 119
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.21448662877082825,
+      "learning_rate": 0.0004947607089353758,
+      "loss": 0.07,
+      "step": 120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6086686849594116,
+      "learning_rate": 0.0004946653578714559,
+      "loss": 0.1352,
+      "step": 121
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25963085889816284,
+      "learning_rate": 0.0004945691563115051,
+      "loss": 0.1447,
+      "step": 122
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.11575956642627716,
+      "learning_rate": 0.0004944721045899356,
+      "loss": 0.1055,
+      "step": 123
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.11230157315731049,
+      "learning_rate": 0.0004943742030441145,
+      "loss": 0.0917,
+      "step": 124
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3376341760158539,
+      "learning_rate": 0.0004942754520143634,
+      "loss": 0.1364,
+      "step": 125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2757412791252136,
+      "learning_rate": 0.0004941758518439566,
+      "loss": 0.1418,
+      "step": 126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1438644975423813,
+      "learning_rate": 0.0004940754028791205,
+      "loss": 0.1162,
+      "step": 127
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.14210884273052216,
+      "learning_rate": 0.0004939741054690316,
+      "loss": 0.1312,
+      "step": 128
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1861649751663208,
+      "learning_rate": 0.0004938719599658162,
+      "loss": 0.1447,
+      "step": 129
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.19665485620498657,
+      "learning_rate": 0.0004937689667245481,
+      "loss": 0.1439,
+      "step": 130
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.22447055578231812,
+      "learning_rate": 0.0004936651261032486,
+      "loss": 0.1568,
+      "step": 131
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.10008269548416138,
+      "learning_rate": 0.0004935604384628843,
+      "loss": 0.1081,
+      "step": 132
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0549234002828598,
+      "learning_rate": 0.0004934549041673661,
+      "loss": 0.1216,
+      "step": 133
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.11616212874650955,
+      "learning_rate": 0.0004933485235835483,
+      "loss": 0.1108,
+      "step": 134
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08554813265800476,
+      "learning_rate": 0.0004932412970812269,
+      "loss": 0.135,
+      "step": 135
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.08642842620611191,
+      "learning_rate": 0.0004931332250331382,
+      "loss": 0.1205,
+      "step": 136
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.20417262613773346,
+      "learning_rate": 0.0004930243078149582,
+      "loss": 0.1169,
+      "step": 137
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.11088764667510986,
+      "learning_rate": 0.0004929145458053005,
+      "loss": 0.1014,
+      "step": 138
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3510516881942749,
+      "learning_rate": 0.0004928039393857155,
+      "loss": 0.0967,
+      "step": 139
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2401883453130722,
+      "learning_rate": 0.0004926924889406888,
+      "loss": 0.106,
+      "step": 140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28403300046920776,
+      "learning_rate": 0.0004925801948576402,
+      "loss": 0.079,
+      "step": 141
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.46027252078056335,
+      "learning_rate": 0.0004924670575269217,
+      "loss": 0.0899,
+      "step": 142
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.09421269595623016,
+      "eval_runtime": 14.7696,
+      "eval_samples_per_second": 32.296,
+      "eval_steps_per_second": 8.125,
+      "step": 142
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29767730832099915,
+      "learning_rate": 0.0004923530773418169,
+      "loss": 0.1265,
+      "step": 143
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37391072511672974,
+      "learning_rate": 0.0004922382546985394,
+      "loss": 0.1244,
+      "step": 144
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8874172568321228,
+      "learning_rate": 0.0004921225899962308,
+      "loss": 0.1796,
+      "step": 145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2554258704185486,
+      "learning_rate": 0.0004920060836369603,
+      "loss": 0.0528,
+      "step": 146
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1981816440820694,
+      "learning_rate": 0.0004918887360257228,
+      "loss": 0.1159,
+      "step": 147
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.14500874280929565,
+      "learning_rate": 0.0004917705475704373,
+      "loss": 0.0992,
+      "step": 148
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1315220594406128,
+      "learning_rate": 0.000491651518681946,
+      "loss": 0.1248,
+      "step": 149
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0798826813697815,
+      "learning_rate": 0.0004915316497740121,
+      "loss": 0.1151,
+      "step": 150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.10213784873485565,
+      "learning_rate": 0.0004914109412633194,
+      "loss": 0.1098,
+      "step": 151
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23167072236537933,
+      "learning_rate": 0.00049128939356947,
+      "loss": 0.1236,
+      "step": 152
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.173340305685997,
+      "learning_rate": 0.0004911670071149831,
+      "loss": 0.1098,
+      "step": 153
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1079009547829628,
+      "learning_rate": 0.0004910437823252937,
+      "loss": 0.1014,
+      "step": 154
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.320765882730484,
+      "learning_rate": 0.0004909197196287509,
+      "loss": 0.1285,
+      "step": 155
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40041017532348633,
+      "learning_rate": 0.0004907948194566166,
+      "loss": 0.1421,
+      "step": 156
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4091287851333618,
+      "learning_rate": 0.0004906690822430638,
+      "loss": 0.1451,
+      "step": 157
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39893922209739685,
+      "learning_rate": 0.0004905425084251753,
+      "loss": 0.1289,
+      "step": 158
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.14173893630504608,
+      "learning_rate": 0.0004904150984429419,
+      "loss": 0.0712,
+      "step": 159
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4723054766654968,
+      "learning_rate": 0.0004902868527392611,
+      "loss": 0.2141,
+      "step": 160
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.13493523001670837,
+      "learning_rate": 0.0004901577717599355,
+      "loss": 0.0881,
+      "step": 161
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.10770414024591446,
+      "learning_rate": 0.0004900278559536716,
+      "loss": 0.0746,
+      "step": 162
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5121994614601135,
+      "learning_rate": 0.0004898971057720773,
+      "loss": 0.1705,
+      "step": 163
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.09419309347867966,
+      "learning_rate": 0.0004897655216696612,
+      "loss": 0.1085,
+      "step": 164
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3557867407798767,
+      "learning_rate": 0.0004896331041038309,
+      "loss": 0.1027,
+      "step": 165
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.082126185297966,
+      "learning_rate": 0.000489499853534891,
+      "loss": 0.1113,
+      "step": 166
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8520584106445312,
+      "learning_rate": 0.0004893657704260419,
+      "loss": 0.1291,
+      "step": 167
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4607222080230713,
+      "learning_rate": 0.000489230855243378,
+      "loss": 0.1241,
+      "step": 168
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5181136727333069,
+      "learning_rate": 0.0004890951084558859,
+      "loss": 0.0957,
+      "step": 169
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42894089221954346,
+      "learning_rate": 0.0004889585305354435,
+      "loss": 0.0895,
+      "step": 170
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14509521424770355,
+      "learning_rate": 0.0004888211219568175,
+      "loss": 0.0732,
+      "step": 171
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.24262909591197968,
+      "learning_rate": 0.0004886828831976621,
+      "loss": 0.0917,
+      "step": 172
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.44387635588645935,
+      "learning_rate": 0.0004885438147385175,
+      "loss": 0.0636,
+      "step": 173
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1804012507200241,
+      "learning_rate": 0.0004884039170628077,
+      "loss": 0.0295,
+      "step": 174
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5566735863685608,
+      "learning_rate": 0.0004882631906568398,
+      "loss": 0.1104,
+      "step": 175
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9653083682060242,
+      "learning_rate": 0.0004881216360098012,
+      "loss": 0.2236,
+      "step": 176
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27046507596969604,
+      "learning_rate": 0.0004879792536137585,
+      "loss": 0.1082,
+      "step": 177
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.47974228858947754,
+      "learning_rate": 0.00048783604396365586,
+      "loss": 0.0884,
+      "step": 178
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.23638087511062622,
+      "learning_rate": 0.0004876920075573129,
+      "loss": 0.0968,
+      "step": 179
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.12476328015327454,
+      "learning_rate": 0.0004875471448954234,
+      "loss": 0.1078,
+      "step": 180
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2455732375383377,
+      "learning_rate": 0.00048740145648155307,
+      "loss": 0.1124,
+      "step": 181
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2744804620742798,
+      "learning_rate": 0.0004872549428221384,
+      "loss": 0.0797,
+      "step": 182
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.19536937773227692,
+      "learning_rate": 0.00048710760442648415,
+      "loss": 0.1091,
+      "step": 183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5277348160743713,
+      "learning_rate": 0.0004869594418067623,
+      "loss": 0.1261,
+      "step": 184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.13960392773151398,
+      "learning_rate": 0.00048681045547801003,
+      "loss": 0.0879,
+      "step": 185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2567049562931061,
+      "learning_rate": 0.00048666064595812746,
+      "loss": 0.083,
+      "step": 186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3075740337371826,
+      "learning_rate": 0.00048651001376787676,
+      "loss": 0.1167,
+      "step": 187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5257586240768433,
+      "learning_rate": 0.0004863585594308794,
+      "loss": 0.1019,
+      "step": 188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.41611766815185547,
+      "learning_rate": 0.00048620628347361496,
+      "loss": 0.1392,
+      "step": 189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30399614572525024,
+      "learning_rate": 0.00048605318642541917,
+      "loss": 0.1339,
+      "step": 190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41276878118515015,
+      "learning_rate": 0.00048589926881848194,
+      "loss": 0.1028,
+      "step": 191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.19717253744602203,
+      "learning_rate": 0.0004857445311878456,
+      "loss": 0.1032,
+      "step": 192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3766873776912689,
+      "learning_rate": 0.0004855889740714028,
+      "loss": 0.1486,
+      "step": 193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17443525791168213,
+      "learning_rate": 0.0004854325980098951,
+      "loss": 0.096,
+      "step": 194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1278471201658249,
+      "learning_rate": 0.0004852754035469109,
+      "loss": 0.0746,
+      "step": 195
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.14356929063796997,
+      "learning_rate": 0.0004851173912288833,
+      "loss": 0.0857,
+      "step": 196
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.20514866709709167,
+      "learning_rate": 0.0004849585616050884,
+      "loss": 0.0833,
+      "step": 197
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4683605134487152,
+      "learning_rate": 0.0004847989152276435,
+      "loss": 0.1538,
+      "step": 198
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29194721579551697,
+      "learning_rate": 0.00048463845265150495,
+      "loss": 0.1035,
+      "step": 199
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.22838515043258667,
+      "learning_rate": 0.0004844771744344666,
+      "loss": 0.0762,
+      "step": 200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3635599911212921,
+      "learning_rate": 0.0004843150811371572,
+      "loss": 0.1165,
+      "step": 201
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2508073151111603,
+      "learning_rate": 0.0004841521733230391,
+      "loss": 0.0736,
+      "step": 202
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24161550402641296,
+      "learning_rate": 0.000483988451558406,
+      "loss": 0.1309,
+      "step": 203
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4697308838367462,
+      "learning_rate": 0.0004838239164123811,
+      "loss": 0.1731,
+      "step": 204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17773008346557617,
+      "learning_rate": 0.0004836585684569148,
+      "loss": 0.1158,
+      "step": 205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.21285519003868103,
+      "learning_rate": 0.0004834924082667833,
+      "loss": 0.0949,
+      "step": 206
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2403111308813095,
+      "learning_rate": 0.0004833254364195859,
+      "loss": 0.0801,
+      "step": 207
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2033465951681137,
+      "learning_rate": 0.0004831576534957437,
+      "loss": 0.069,
+      "step": 208
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5510303378105164,
+      "learning_rate": 0.000482989060078497,
+      "loss": 0.1766,
+      "step": 209
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32342344522476196,
+      "learning_rate": 0.0004828196567539034,
+      "loss": 0.1229,
+      "step": 210
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3102104663848877,
+      "learning_rate": 0.00048264944411083625,
+      "loss": 0.1297,
+      "step": 211
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32639122009277344,
+      "learning_rate": 0.00048247842274098187,
+      "loss": 0.1011,
+      "step": 212
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4720034897327423,
+      "learning_rate": 0.00048230659323883804,
+      "loss": 0.1282,
+      "step": 213
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5249712467193604,
+      "learning_rate": 0.00048213395620171166,
+      "loss": 0.1376,
+      "step": 214
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3953443467617035,
+      "learning_rate": 0.00048196051222971673,
+      "loss": 0.1186,
+      "step": 215
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15697738528251648,
+      "learning_rate": 0.0004817862619257723,
+      "loss": 0.1079,
+      "step": 216
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32511651515960693,
+      "learning_rate": 0.0004816112058956005,
+      "loss": 0.1052,
+      "step": 217
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1850031018257141,
+      "learning_rate": 0.00048143534474772397,
+      "loss": 0.1236,
+      "step": 218
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.10901057720184326,
+      "learning_rate": 0.0004812586790934645,
+      "loss": 0.1094,
+      "step": 219
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.23395784199237823,
+      "learning_rate": 0.00048108120954694014,
+      "loss": 0.0556,
+      "step": 220
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.21469372510910034,
+      "learning_rate": 0.00048090293672506347,
+      "loss": 0.0594,
+      "step": 221
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17289988696575165,
+      "learning_rate": 0.00048072386124753944,
+      "loss": 0.0219,
+      "step": 222
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.21490757167339325,
+      "learning_rate": 0.0004805439837368631,
+      "loss": 0.0203,
+      "step": 223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1259506940841675,
+      "learning_rate": 0.0004803633048183176,
+      "loss": 0.1576,
+      "step": 224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2934038639068604,
+      "learning_rate": 0.00048018182511997185,
+      "loss": 0.1233,
+      "step": 225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4250846207141876,
+      "learning_rate": 0.0004799995452726783,
+      "loss": 0.1023,
+      "step": 226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4675579071044922,
+      "learning_rate": 0.000479816465910071,
+      "loss": 0.1242,
+      "step": 227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7030429840087891,
+      "learning_rate": 0.0004796325876685632,
+      "loss": 0.0514,
+      "step": 228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5683910846710205,
+      "learning_rate": 0.00047944791118734517,
+      "loss": 0.0923,
+      "step": 229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8425244092941284,
+      "learning_rate": 0.0004792624371083819,
+      "loss": 0.0976,
+      "step": 230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.21189981698989868,
+      "learning_rate": 0.00047907616607641113,
+      "loss": 0.1016,
+      "step": 231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.36100390553474426,
+      "learning_rate": 0.0004788890987389408,
+      "loss": 0.1015,
+      "step": 232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.42600420117378235,
+      "learning_rate": 0.000478701235746247,
+      "loss": 0.1401,
+      "step": 233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.649318516254425,
+      "learning_rate": 0.0004785125777513716,
+      "loss": 0.1012,
+      "step": 234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3490477204322815,
+      "learning_rate": 0.00047832312541012007,
+      "loss": 0.1015,
+      "step": 235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6937799453735352,
+      "learning_rate": 0.0004781328793810592,
+      "loss": 0.1188,
+      "step": 236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0924077033996582,
+      "learning_rate": 0.0004779418403255146,
+      "loss": 0.1093,
+      "step": 237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.36075183749198914,
+      "learning_rate": 0.0004777500089075687,
+      "loss": 0.0971,
+      "step": 238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41673243045806885,
+      "learning_rate": 0.00047755738579405836,
+      "loss": 0.0953,
+      "step": 239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.13159583508968353,
+      "learning_rate": 0.0004773639716545723,
+      "loss": 0.0571,
+      "step": 240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9338862895965576,
+      "learning_rate": 0.00047716976716144917,
+      "loss": 0.202,
+      "step": 241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3190581798553467,
+      "learning_rate": 0.0004769747729897749,
+      "loss": 0.1071,
+      "step": 242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23796042799949646,
+      "learning_rate": 0.0004767789898173806,
+      "loss": 0.0659,
+      "step": 243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19194231927394867,
+      "learning_rate": 0.0004765824183248399,
+      "loss": 0.0611,
+      "step": 244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16703608632087708,
+      "learning_rate": 0.0004763850591954668,
+      "loss": 0.0855,
+      "step": 245
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3395439684391022,
+      "learning_rate": 0.0004761869131153135,
+      "loss": 0.0926,
+      "step": 246
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2820179760456085,
+      "learning_rate": 0.0004759879807731673,
+      "loss": 0.0508,
+      "step": 247
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.20656561851501465,
+      "learning_rate": 0.00047578826286054897,
+      "loss": 0.068,
+      "step": 248
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4477837383747101,
+      "learning_rate": 0.00047558776007171024,
+      "loss": 0.0918,
+      "step": 249
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18997950851917267,
+      "learning_rate": 0.0004753864731036307,
+      "loss": 0.0734,
+      "step": 250
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2841518521308899,
+      "learning_rate": 0.0004751844026560163,
+      "loss": 0.1194,
+      "step": 251
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29770052433013916,
+      "learning_rate": 0.0004749815494312963,
+      "loss": 0.0996,
+      "step": 252
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2982254922389984,
+      "learning_rate": 0.00047477791413462104,
+      "loss": 0.0945,
+      "step": 253
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4625980854034424,
+      "learning_rate": 0.00047457349747385936,
+      "loss": 0.131,
+      "step": 254
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29756709933280945,
+      "learning_rate": 0.00047436830015959653,
+      "loss": 0.1057,
+      "step": 255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.19971434772014618,
+      "learning_rate": 0.00047416232290513127,
+      "loss": 0.0794,
+      "step": 256
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.12171836197376251,
+      "learning_rate": 0.0004739555664264736,
+      "loss": 0.0527,
+      "step": 257
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.23848529160022736,
+      "learning_rate": 0.00047374803144234213,
+      "loss": 0.134,
+      "step": 258
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.12673752009868622,
+      "learning_rate": 0.0004735397186741618,
+      "loss": 0.0774,
+      "step": 259
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.11961629241704941,
+      "learning_rate": 0.00047333062884606116,
+      "loss": 0.0661,
+      "step": 260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18004140257835388,
+      "learning_rate": 0.00047312076268487,
+      "loss": 0.1132,
+      "step": 261
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1698005348443985,
+      "learning_rate": 0.00047291012092011685,
+      "loss": 0.057,
+      "step": 262
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1949334442615509,
+      "learning_rate": 0.0004726987042840263,
+      "loss": 0.0703,
+      "step": 263
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4016534686088562,
+      "learning_rate": 0.0004724865135115163,
+      "loss": 0.1178,
+      "step": 264
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.36885496973991394,
+      "learning_rate": 0.00047227354934019605,
+      "loss": 0.1303,
+      "step": 265
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3214585483074188,
+      "learning_rate": 0.00047205981251036334,
+      "loss": 0.1019,
+      "step": 266
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.15313082933425903,
+      "learning_rate": 0.0004718453037650016,
+      "loss": 0.0581,
+      "step": 267
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3251878321170807,
+      "learning_rate": 0.0004716300238497775,
+      "loss": 0.099,
+      "step": 268
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.20356950163841248,
+      "learning_rate": 0.0004714139735130388,
+      "loss": 0.0767,
+      "step": 269
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2644464373588562,
+      "learning_rate": 0.00047119715350581095,
+      "loss": 0.1003,
+      "step": 270
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.22035302221775055,
+      "learning_rate": 0.000470979564581795,
+      "loss": 0.0722,
+      "step": 271
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5284466743469238,
+      "learning_rate": 0.0004707612074973653,
+      "loss": 0.1282,
+      "step": 272
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.34032565355300903,
+      "learning_rate": 0.0004705420830115658,
+      "loss": 0.099,
+      "step": 273
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26527565717697144,
+      "learning_rate": 0.00047032219188610836,
+      "loss": 0.0911,
+      "step": 274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2254990190267563,
+      "learning_rate": 0.0004701015348853699,
+      "loss": 0.0667,
+      "step": 275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.21334387362003326,
+      "learning_rate": 0.0004698801127763895,
+      "loss": 0.0659,
+      "step": 276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2917044758796692,
+      "learning_rate": 0.0004696579263288661,
+      "loss": 0.1159,
+      "step": 277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.14027804136276245,
+      "learning_rate": 0.00046943497631515526,
+      "loss": 0.0323,
+      "step": 278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3988366425037384,
+      "learning_rate": 0.00046921126351026697,
+      "loss": 0.0887,
+      "step": 279
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36629319190979004,
+      "learning_rate": 0.00046898678869186297,
+      "loss": 0.1079,
+      "step": 280
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.35548141598701477,
+      "learning_rate": 0.0004687615526402536,
+      "loss": 0.1056,
+      "step": 281
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.21030637621879578,
+      "learning_rate": 0.0004685355561383956,
+      "loss": 0.0717,
+      "step": 282
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24192889034748077,
+      "learning_rate": 0.000468308799971889,
+      "loss": 0.1047,
+      "step": 283
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.16289295256137848,
+      "learning_rate": 0.00046808128492897464,
+      "loss": 0.0519,
+      "step": 284
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.08938124030828476,
+      "eval_runtime": 14.7518,
+      "eval_samples_per_second": 32.335,
+      "eval_steps_per_second": 8.135,
+      "step": 284
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.23021583259105682,
+      "learning_rate": 0.00046785301180053126,
+      "loss": 0.1161,
+      "step": 285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3577558398246765,
+      "learning_rate": 0.0004676239813800729,
+      "loss": 0.1239,
+      "step": 286
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.15293735265731812,
+      "learning_rate": 0.0004673941944637461,
+      "loss": 0.0401,
+      "step": 287
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.342631459236145,
+      "learning_rate": 0.00046716365185032696,
+      "loss": 0.1358,
+      "step": 288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4987104833126068,
+      "learning_rate": 0.0004669323543412186,
+      "loss": 0.1312,
+      "step": 289
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.21678434312343597,
+      "learning_rate": 0.0004667003027404483,
+      "loss": 0.0791,
+      "step": 290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2781723141670227,
+      "learning_rate": 0.00046646749785466464,
+      "loss": 0.0809,
+      "step": 291
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3997693359851837,
+      "learning_rate": 0.00046623394049313474,
+      "loss": 0.0938,
+      "step": 292
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2478984147310257,
+      "learning_rate": 0.00046599963146774136,
+      "loss": 0.0671,
+      "step": 293
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35655421018600464,
+      "learning_rate": 0.0004657645715929805,
+      "loss": 0.107,
+      "step": 294
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.41986069083213806,
+      "learning_rate": 0.0004655287616859577,
+      "loss": 0.1381,
+      "step": 295
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2831580340862274,
+      "learning_rate": 0.00046529220256638626,
+      "loss": 0.1012,
+      "step": 296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2183172106742859,
+      "learning_rate": 0.0004650548950565835,
+      "loss": 0.0883,
+      "step": 297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1485687792301178,
+      "learning_rate": 0.0004648168399814684,
+      "loss": 0.094,
+      "step": 298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3192533552646637,
+      "learning_rate": 0.0004645780381685586,
+      "loss": 0.1144,
+      "step": 299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.20768460631370544,
+      "learning_rate": 0.0004643384904479675,
+      "loss": 0.1119,
+      "step": 300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.16704390943050385,
+      "learning_rate": 0.00046409819765240147,
+      "loss": 0.0852,
+      "step": 301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33123648166656494,
+      "learning_rate": 0.0004638571606171567,
+      "loss": 0.1608,
+      "step": 302
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.408978134393692,
+      "learning_rate": 0.0004636153801801167,
+      "loss": 0.0906,
+      "step": 303
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29201096296310425,
+      "learning_rate": 0.00046337285718174896,
+      "loss": 0.1237,
+      "step": 304
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.45836058259010315,
+      "learning_rate": 0.00046312959246510237,
+      "loss": 0.0926,
+      "step": 305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5405777096748352,
+      "learning_rate": 0.0004628855868758041,
+      "loss": 0.0727,
+      "step": 306
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3068138062953949,
+      "learning_rate": 0.00046264084126205676,
+      "loss": 0.1006,
+      "step": 307
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2990975081920624,
+      "learning_rate": 0.00046239535647463534,
+      "loss": 0.1033,
+      "step": 308
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2938540279865265,
+      "learning_rate": 0.00046214913336688424,
+      "loss": 0.1084,
+      "step": 309
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49840983748435974,
+      "learning_rate": 0.00046190217279471466,
+      "loss": 0.1066,
+      "step": 310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4558626711368561,
+      "learning_rate": 0.000461654475616601,
+      "loss": 0.121,
+      "step": 311
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.20964759588241577,
+      "learning_rate": 0.0004614060426935786,
+      "loss": 0.0843,
+      "step": 312
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2254151701927185,
+      "learning_rate": 0.00046115687488923983,
+      "loss": 0.0781,
+      "step": 313
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6117066740989685,
+      "learning_rate": 0.0004609069730697322,
+      "loss": 0.1208,
+      "step": 314
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5352897644042969,
+      "learning_rate": 0.0004606563381037544,
+      "loss": 0.1056,
+      "step": 315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.31001269817352295,
+      "learning_rate": 0.00046040497086255385,
+      "loss": 0.1213,
+      "step": 316
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.22637054324150085,
+      "learning_rate": 0.0004601528722199234,
+      "loss": 0.105,
+      "step": 317
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.20077432692050934,
+      "learning_rate": 0.0004599000430521984,
+      "loss": 0.0837,
+      "step": 318
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24702684581279755,
+      "learning_rate": 0.0004596464842382534,
+      "loss": 0.0695,
+      "step": 319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.253387987613678,
+      "learning_rate": 0.0004593921966594997,
+      "loss": 0.1184,
+      "step": 320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2703799605369568,
+      "learning_rate": 0.0004591371811998817,
+      "loss": 0.117,
+      "step": 321
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23513701558113098,
+      "learning_rate": 0.00045888143874587396,
+      "loss": 0.1359,
+      "step": 322
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.25604313611984253,
+      "learning_rate": 0.00045862497018647833,
+      "loss": 0.1018,
+      "step": 323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.16947636008262634,
+      "learning_rate": 0.0004583677764132207,
+      "loss": 0.0958,
+      "step": 324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.20054908096790314,
+      "learning_rate": 0.0004581098583201478,
+      "loss": 0.0803,
+      "step": 325
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.12656472623348236,
+      "learning_rate": 0.00045785121680382436,
+      "loss": 0.0679,
+      "step": 326
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1423499882221222,
+      "learning_rate": 0.0004575918527633297,
+      "loss": 0.0959,
+      "step": 327
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36370569467544556,
+      "learning_rate": 0.0004573317671002549,
+      "loss": 0.1088,
+      "step": 328
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18775340914726257,
+      "learning_rate": 0.0004570709607186995,
+      "loss": 0.0905,
+      "step": 329
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.14833571016788483,
+      "learning_rate": 0.0004568094345252681,
+      "loss": 0.0661,
+      "step": 330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2987270653247833,
+      "learning_rate": 0.00045654718942906794,
+      "loss": 0.0872,
+      "step": 331
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.21985827386379242,
+      "learning_rate": 0.000456284226341705,
+      "loss": 0.0882,
+      "step": 332
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2726268470287323,
+      "learning_rate": 0.00045602054617728093,
+      "loss": 0.0864,
+      "step": 333
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2882244884967804,
+      "learning_rate": 0.00045575614985239057,
+      "loss": 0.1032,
+      "step": 334
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.427500456571579,
+      "learning_rate": 0.0004554910382861178,
+      "loss": 0.1309,
+      "step": 335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.43029338121414185,
+      "learning_rate": 0.000455225212400033,
+      "loss": 0.1071,
+      "step": 336
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2297673523426056,
+      "learning_rate": 0.0004549586731181896,
+      "loss": 0.0526,
+      "step": 337
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4533613920211792,
+      "learning_rate": 0.0004546914213671209,
+      "loss": 0.1154,
+      "step": 338
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3973630666732788,
+      "learning_rate": 0.0004544234580758367,
+      "loss": 0.0707,
+      "step": 339
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.40036290884017944,
+      "learning_rate": 0.0004541547841758207,
+      "loss": 0.0932,
+      "step": 340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4273395240306854,
+      "learning_rate": 0.0004538854006010262,
+      "loss": 0.1112,
+      "step": 341
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28109779953956604,
+      "learning_rate": 0.0004536153082878738,
+      "loss": 0.1003,
+      "step": 342
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.21950216591358185,
+      "learning_rate": 0.00045334450817524776,
+      "loss": 0.0538,
+      "step": 343
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2968471646308899,
+      "learning_rate": 0.00045307300120449263,
+      "loss": 0.0775,
+      "step": 344
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1488364040851593,
+      "learning_rate": 0.00045280078831941024,
+      "loss": 0.0513,
+      "step": 345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.22750218212604523,
+      "learning_rate": 0.00045252787046625624,
+      "loss": 0.0943,
+      "step": 346
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3048767149448395,
+      "learning_rate": 0.0004522542485937369,
+      "loss": 0.079,
+      "step": 347
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33520030975341797,
+      "learning_rate": 0.0004519799236530057,
+      "loss": 0.1584,
+      "step": 348
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.20777581632137299,
+      "learning_rate": 0.00045170489659766003,
+      "loss": 0.0903,
+      "step": 349
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1602245271205902,
+      "learning_rate": 0.00045142916838373826,
+      "loss": 0.0446,
+      "step": 350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2512218952178955,
+      "learning_rate": 0.0004511527399697158,
+      "loss": 0.069,
+      "step": 351
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17349962890148163,
+      "learning_rate": 0.0004508756123165021,
+      "loss": 0.0765,
+      "step": 352
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26563215255737305,
+      "learning_rate": 0.00045059778638743744,
+      "loss": 0.0966,
+      "step": 353
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.23987066745758057,
+      "learning_rate": 0.00045031926314828926,
+      "loss": 0.0702,
+      "step": 354
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21901372075080872,
+      "learning_rate": 0.000450040043567249,
+      "loss": 0.0457,
+      "step": 355
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24179872870445251,
+      "learning_rate": 0.00044976012861492877,
+      "loss": 0.0651,
+      "step": 356
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3544818162918091,
+      "learning_rate": 0.0004494795192643578,
+      "loss": 0.0622,
+      "step": 357
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4363332986831665,
+      "learning_rate": 0.00044919821649097916,
+      "loss": 0.0972,
+      "step": 358
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.43788430094718933,
+      "learning_rate": 0.0004489162212726465,
+      "loss": 0.0843,
+      "step": 359
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5084832906723022,
+      "learning_rate": 0.00044863353458962044,
+      "loss": 0.0888,
+      "step": 360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.44660842418670654,
+      "learning_rate": 0.0004483501574245652,
+      "loss": 0.113,
+      "step": 361
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7528813481330872,
+      "learning_rate": 0.0004480660907625452,
+      "loss": 0.0512,
+      "step": 362
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9723535776138306,
+      "learning_rate": 0.0004477813355910219,
+      "loss": 0.1154,
+      "step": 363
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2641480565071106,
+      "learning_rate": 0.0004474958928998498,
+      "loss": 0.0575,
+      "step": 364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.12234170734882355,
+      "learning_rate": 0.00044720976368127355,
+      "loss": 0.0441,
+      "step": 365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26976636052131653,
+      "learning_rate": 0.00044692294892992416,
+      "loss": 0.0676,
+      "step": 366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.22729526460170746,
+      "learning_rate": 0.00044663544964281573,
+      "loss": 0.098,
+      "step": 367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2270442545413971,
+      "learning_rate": 0.0004463472668193419,
+      "loss": 0.0842,
+      "step": 368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19249562919139862,
+      "learning_rate": 0.0004460584014612724,
+      "loss": 0.0537,
+      "step": 369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.22312623262405396,
+      "learning_rate": 0.0004457688545727496,
+      "loss": 0.0547,
+      "step": 370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.281658411026001,
+      "learning_rate": 0.0004454786271602849,
+      "loss": 0.089,
+      "step": 371
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49952250719070435,
+      "learning_rate": 0.00044518772023275526,
+      "loss": 0.1298,
+      "step": 372
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.186232328414917,
+      "learning_rate": 0.0004448961348013999,
+      "loss": 0.0628,
+      "step": 373
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2980823814868927,
+      "learning_rate": 0.0004446038718798166,
+      "loss": 0.0828,
+      "step": 374
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3794187605381012,
+      "learning_rate": 0.00044431093248395806,
+      "loss": 0.0776,
+      "step": 375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29262277483940125,
+      "learning_rate": 0.0004440173176321287,
+      "loss": 0.0924,
+      "step": 376
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30543988943099976,
+      "learning_rate": 0.0004437230283449808,
+      "loss": 0.1264,
+      "step": 377
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3436485826969147,
+      "learning_rate": 0.0004434280656455111,
+      "loss": 0.1066,
+      "step": 378
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.23679965734481812,
+      "learning_rate": 0.0004431324305590572,
+      "loss": 0.075,
+      "step": 379
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4399561882019043,
+      "learning_rate": 0.0004428361241132943,
+      "loss": 0.1445,
+      "step": 380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39203163981437683,
+      "learning_rate": 0.0004425391473382309,
+      "loss": 0.0995,
+      "step": 381
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4687665104866028,
+      "learning_rate": 0.0004422415012662061,
+      "loss": 0.1489,
+      "step": 382
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2634904086589813,
+      "learning_rate": 0.00044194318693188526,
+      "loss": 0.1164,
+      "step": 383
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.23031170666217804,
+      "learning_rate": 0.0004416442053722569,
+      "loss": 0.0742,
+      "step": 384
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30467960238456726,
+      "learning_rate": 0.00044134455762662894,
+      "loss": 0.0984,
+      "step": 385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.16692829132080078,
+      "learning_rate": 0.0004410442447366249,
+      "loss": 0.0732,
+      "step": 386
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.20100833475589752,
+      "learning_rate": 0.00044074326774618065,
+      "loss": 0.1082,
+      "step": 387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29799607396125793,
+      "learning_rate": 0.0004404416277015404,
+      "loss": 0.0761,
+      "step": 388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.647639274597168,
+      "learning_rate": 0.0004401393256512534,
+      "loss": 0.1218,
+      "step": 389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2610540986061096,
+      "learning_rate": 0.00043983636264617013,
+      "loss": 0.0923,
+      "step": 390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4049086570739746,
+      "learning_rate": 0.0004395327397394384,
+      "loss": 0.1091,
+      "step": 391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.36092105507850647,
+      "learning_rate": 0.00043922845798650034,
+      "loss": 0.0927,
+      "step": 392
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.542421281337738,
+      "learning_rate": 0.00043892351844508805,
+      "loss": 0.1014,
+      "step": 393
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.291595995426178,
+      "learning_rate": 0.0004386179221752202,
+      "loss": 0.0902,
+      "step": 394
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17152707278728485,
+      "learning_rate": 0.0004383116702391987,
+      "loss": 0.0651,
+      "step": 395
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.16654878854751587,
+      "learning_rate": 0.00043800476370160416,
+      "loss": 0.0824,
+      "step": 396
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18530108034610748,
+      "learning_rate": 0.000437697203629293,
+      "loss": 0.0549,
+      "step": 397
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5760988593101501,
+      "learning_rate": 0.0004373889910913934,
+      "loss": 0.0803,
+      "step": 398
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4253963232040405,
+      "learning_rate": 0.00043708012715930154,
+      "loss": 0.0728,
+      "step": 399
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7932385206222534,
+      "learning_rate": 0.00043677061290667805,
+      "loss": 0.1442,
+      "step": 400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5904709696769714,
+      "learning_rate": 0.00043646044940944407,
+      "loss": 0.0999,
+      "step": 401
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9570127129554749,
+      "learning_rate": 0.0004361496377457777,
+      "loss": 0.1298,
+      "step": 402
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5049470663070679,
+      "learning_rate": 0.00043583817899611017,
+      "loss": 0.0263,
+      "step": 403
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.589408814907074,
+      "learning_rate": 0.00043552607424312195,
+      "loss": 0.1051,
+      "step": 404
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.43722283840179443,
+      "learning_rate": 0.0004352133245717393,
+      "loss": 0.0715,
+      "step": 405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3537758588790894,
+      "learning_rate": 0.00043489993106913036,
+      "loss": 0.0322,
+      "step": 406
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3382836580276489,
+      "learning_rate": 0.000434585894824701,
+      "loss": 0.0818,
+      "step": 407
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9946733713150024,
+      "learning_rate": 0.00043427121693009164,
+      "loss": 0.1536,
+      "step": 408
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9138526320457458,
+      "learning_rate": 0.0004339558984791732,
+      "loss": 0.1299,
+      "step": 409
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35993850231170654,
+      "learning_rate": 0.0004336399405680432,
+      "loss": 0.0654,
+      "step": 410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.30418309569358826,
+      "learning_rate": 0.0004333233442950219,
+      "loss": 0.1026,
+      "step": 411
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.256728857755661,
+      "learning_rate": 0.00043300611076064886,
+      "loss": 0.083,
+      "step": 412
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3472314774990082,
+      "learning_rate": 0.00043268824106767865,
+      "loss": 0.0637,
+      "step": 413
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5615578293800354,
+      "learning_rate": 0.00043236973632107735,
+      "loss": 0.1028,
+      "step": 414
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.35775747895240784,
+      "learning_rate": 0.00043205059762801854,
+      "loss": 0.0829,
+      "step": 415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.287270724773407,
+      "learning_rate": 0.0004317308260978795,
+      "loss": 0.0718,
+      "step": 416
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3237059414386749,
+      "learning_rate": 0.00043141042284223737,
+      "loss": 0.0797,
+      "step": 417
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2153153419494629,
+      "learning_rate": 0.0004310893889748653,
+      "loss": 0.0778,
+      "step": 418
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33600860834121704,
+      "learning_rate": 0.00043076772561172845,
+      "loss": 0.0594,
+      "step": 419
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8778895139694214,
+      "learning_rate": 0.00043044543387098027,
+      "loss": 0.1722,
+      "step": 420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.503434419631958,
+      "learning_rate": 0.0004301225148729586,
+      "loss": 0.1228,
+      "step": 421
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3694842457771301,
+      "learning_rate": 0.00042979896974018166,
+      "loss": 0.1033,
+      "step": 422
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.329818457365036,
+      "learning_rate": 0.00042947479959734423,
+      "loss": 0.0471,
+      "step": 423
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19436267018318176,
+      "learning_rate": 0.0004291500055713138,
+      "loss": 0.0409,
+      "step": 424
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3269858658313751,
+      "learning_rate": 0.0004288245887911263,
+      "loss": 0.096,
+      "step": 425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3542991578578949,
+      "learning_rate": 0.00042849855038798283,
+      "loss": 0.0986,
+      "step": 426
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.07792978733778,
+      "eval_runtime": 14.8115,
+      "eval_samples_per_second": 32.205,
+      "eval_steps_per_second": 8.102,
+      "step": 426
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29685521125793457,
+      "learning_rate": 0.00042817189149524517,
+      "loss": 0.11,
+      "step": 427
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2116887867450714,
+      "learning_rate": 0.00042784461324843194,
+      "loss": 0.0686,
+      "step": 428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4866883456707001,
+      "learning_rate": 0.00042751671678521486,
+      "loss": 0.0824,
+      "step": 429
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1293468475341797,
+      "learning_rate": 0.00042718820324541475,
+      "loss": 0.0464,
+      "step": 430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3253125250339508,
+      "learning_rate": 0.0004268590737709972,
+      "loss": 0.0996,
+      "step": 431
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25559771060943604,
+      "learning_rate": 0.00042652932950606917,
+      "loss": 0.0545,
+      "step": 432
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2788093686103821,
+      "learning_rate": 0.0004261989715968746,
+      "loss": 0.0502,
+      "step": 433
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6902124285697937,
+      "learning_rate": 0.00042586800119179046,
+      "loss": 0.1598,
+      "step": 434
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4788605570793152,
+      "learning_rate": 0.00042553641944132316,
+      "loss": 0.1552,
+      "step": 435
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.43495067954063416,
+      "learning_rate": 0.00042520422749810395,
+      "loss": 0.0907,
+      "step": 436
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3549440801143646,
+      "learning_rate": 0.0004248714265168853,
+      "loss": 0.1152,
+      "step": 437
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7210204601287842,
+      "learning_rate": 0.00042453801765453687,
+      "loss": 0.1891,
+      "step": 438
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4578750729560852,
+      "learning_rate": 0.00042420400207004126,
+      "loss": 0.1383,
+      "step": 439
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3323976993560791,
+      "learning_rate": 0.00042386938092449036,
+      "loss": 0.0936,
+      "step": 440
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.15145371854305267,
+      "learning_rate": 0.00042353415538108076,
+      "loss": 0.0608,
+      "step": 441
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.10744435340166092,
+      "learning_rate": 0.00042319832660511037,
+      "loss": 0.0865,
+      "step": 442
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.13599476218223572,
+      "learning_rate": 0.0004228618957639738,
+      "loss": 0.0763,
+      "step": 443
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18250028789043427,
+      "learning_rate": 0.00042252486402715865,
+      "loss": 0.0813,
+      "step": 444
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5180188417434692,
+      "learning_rate": 0.00042218723256624136,
+      "loss": 0.1603,
+      "step": 445
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2943187355995178,
+      "learning_rate": 0.000421849002554883,
+      "loss": 0.1031,
+      "step": 446
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.14898087084293365,
+      "learning_rate": 0.0004215101751688253,
+      "loss": 0.071,
+      "step": 447
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2951905131340027,
+      "learning_rate": 0.00042117075158588663,
+      "loss": 0.0772,
+      "step": 448
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.39807453751564026,
+      "learning_rate": 0.00042083073298595787,
+      "loss": 0.0561,
+      "step": 449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.45217999815940857,
+      "learning_rate": 0.0004204901205509981,
+      "loss": 0.1076,
+      "step": 450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.24114732444286346,
+      "learning_rate": 0.000420148915465031,
+      "loss": 0.1169,
+      "step": 451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6120204329490662,
+      "learning_rate": 0.00041980711891413994,
+      "loss": 0.1144,
+      "step": 452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3900619447231293,
+      "learning_rate": 0.0004194647320864646,
+      "loss": 0.0806,
+      "step": 453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3331635296344757,
+      "learning_rate": 0.0004191217561721967,
+      "loss": 0.0655,
+      "step": 454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29893186688423157,
+      "learning_rate": 0.0004187781923635753,
+      "loss": 0.0482,
+      "step": 455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.20024164021015167,
+      "learning_rate": 0.00041843404185488346,
+      "loss": 0.0773,
+      "step": 456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3644329905509949,
+      "learning_rate": 0.0004180893058424435,
+      "loss": 0.1062,
+      "step": 457
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5457159280776978,
+      "learning_rate": 0.0004177439855246132,
+      "loss": 0.1901,
+      "step": 458
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.282032310962677,
+      "learning_rate": 0.0004173980821017812,
+      "loss": 0.0656,
+      "step": 459
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1957680881023407,
+      "learning_rate": 0.00041705159677636334,
+      "loss": 0.0725,
+      "step": 460
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2736223042011261,
+      "learning_rate": 0.00041670453075279827,
+      "loss": 0.0897,
+      "step": 461
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2145017832517624,
+      "learning_rate": 0.0004163568852375431,
+      "loss": 0.046,
+      "step": 462
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1434750258922577,
+      "learning_rate": 0.00041600866143906947,
+      "loss": 0.0483,
+      "step": 463
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2438279092311859,
+      "learning_rate": 0.000415659860567859,
+      "loss": 0.0935,
+      "step": 464
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24830487370491028,
+      "learning_rate": 0.00041531048383639966,
+      "loss": 0.1061,
+      "step": 465
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.25185227394104004,
+      "learning_rate": 0.000414960532459181,
+      "loss": 0.082,
+      "step": 466
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.391631156206131,
+      "learning_rate": 0.00041461000765269,
+      "loss": 0.1274,
+      "step": 467
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30484774708747864,
+      "learning_rate": 0.0004142589106354071,
+      "loss": 0.0672,
+      "step": 468
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2584599554538727,
+      "learning_rate": 0.0004139072426278021,
+      "loss": 0.0863,
+      "step": 469
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27182772755622864,
+      "learning_rate": 0.0004135550048523292,
+      "loss": 0.0996,
+      "step": 470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2670001685619354,
+      "learning_rate": 0.00041320219853342347,
+      "loss": 0.0592,
+      "step": 471
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19571639597415924,
+      "learning_rate": 0.0004128488248974962,
+      "loss": 0.0618,
+      "step": 472
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.436814546585083,
+      "learning_rate": 0.00041249488517293095,
+      "loss": 0.1131,
+      "step": 473
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.21684250235557556,
+      "learning_rate": 0.0004121403805900789,
+      "loss": 0.0759,
+      "step": 474
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.39313605427742004,
+      "learning_rate": 0.0004117853123812549,
+      "loss": 0.0992,
+      "step": 475
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3653202950954437,
+      "learning_rate": 0.00041142968178073294,
+      "loss": 0.099,
+      "step": 476
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.36615628004074097,
+      "learning_rate": 0.00041107349002474206,
+      "loss": 0.06,
+      "step": 477
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2431243658065796,
+      "learning_rate": 0.00041071673835146194,
+      "loss": 0.0689,
+      "step": 478
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7869367599487305,
+      "learning_rate": 0.00041035942800101864,
+      "loss": 0.1308,
+      "step": 479
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2831230163574219,
+      "learning_rate": 0.0004100015602154802,
+      "loss": 0.087,
+      "step": 480
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3709629774093628,
+      "learning_rate": 0.0004096431362388525,
+      "loss": 0.0822,
+      "step": 481
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4082586467266083,
+      "learning_rate": 0.0004092841573170748,
+      "loss": 0.1114,
+      "step": 482
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2919554114341736,
+      "learning_rate": 0.0004089246246980154,
+      "loss": 0.1059,
+      "step": 483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3750731945037842,
+      "learning_rate": 0.0004085645396314673,
+      "loss": 0.082,
+      "step": 484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.21013659238815308,
+      "learning_rate": 0.000408203903369144,
+      "loss": 0.0819,
+      "step": 485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.20771674811840057,
+      "learning_rate": 0.00040784271716467503,
+      "loss": 0.0687,
+      "step": 486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.157434344291687,
+      "learning_rate": 0.00040748098227360154,
+      "loss": 0.0826,
+      "step": 487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.40467727184295654,
+      "learning_rate": 0.000407118699953372,
+      "loss": 0.1131,
+      "step": 488
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.17521728575229645,
+      "learning_rate": 0.0004067558714633378,
+      "loss": 0.116,
+      "step": 489
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2975709140300751,
+      "learning_rate": 0.0004063924980647492,
+      "loss": 0.0787,
+      "step": 490
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.22513332962989807,
+      "learning_rate": 0.0004060285810207503,
+      "loss": 0.0754,
+      "step": 491
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2939409613609314,
+      "learning_rate": 0.00040566412159637514,
+      "loss": 0.0505,
+      "step": 492
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.21415212750434875,
+      "learning_rate": 0.000405299121058543,
+      "loss": 0.0486,
+      "step": 493
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.24846945703029633,
+      "learning_rate": 0.00040493358067605445,
+      "loss": 0.0645,
+      "step": 494
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.42928287386894226,
+      "learning_rate": 0.00040456750171958655,
+      "loss": 0.1455,
+      "step": 495
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30920714139938354,
+      "learning_rate": 0.0004042008854616883,
+      "loss": 0.0743,
+      "step": 496
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.43211719393730164,
+      "learning_rate": 0.00040383373317677687,
+      "loss": 0.1037,
+      "step": 497
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49942275881767273,
+      "learning_rate": 0.00040346604614113215,
+      "loss": 0.123,
+      "step": 498
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18615621328353882,
+      "learning_rate": 0.00040309782563289353,
+      "loss": 0.0783,
+      "step": 499
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22238926589488983,
+      "learning_rate": 0.0004027290729320545,
+      "loss": 0.0698,
+      "step": 500
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31746548414230347,
+      "learning_rate": 0.0004023597893204586,
+      "loss": 0.1682,
+      "step": 501
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19328100979328156,
+      "learning_rate": 0.00040198997608179477,
+      "loss": 0.1028,
+      "step": 502
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.15466806292533875,
+      "learning_rate": 0.00040161963450159333,
+      "loss": 0.065,
+      "step": 503
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3000398874282837,
+      "learning_rate": 0.00040124876586722103,
+      "loss": 0.1071,
+      "step": 504
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.16753748059272766,
+      "learning_rate": 0.00040087737146787654,
+      "loss": 0.056,
+      "step": 505
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.17570586502552032,
+      "learning_rate": 0.00040050545259458654,
+      "loss": 0.0732,
+      "step": 506
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19240190088748932,
+      "learning_rate": 0.00040013301054020055,
+      "loss": 0.0444,
+      "step": 507
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23935984075069427,
+      "learning_rate": 0.00039976004659938714,
+      "loss": 0.0583,
+      "step": 508
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22633028030395508,
+      "learning_rate": 0.00039938656206862857,
+      "loss": 0.065,
+      "step": 509
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18621531128883362,
+      "learning_rate": 0.000399012558246217,
+      "loss": 0.0489,
+      "step": 510
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.37711310386657715,
+      "learning_rate": 0.0003986380364322498,
+      "loss": 0.1367,
+      "step": 511
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26448771357536316,
+      "learning_rate": 0.00039826299792862475,
+      "loss": 0.076,
+      "step": 512
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.22461633384227753,
+      "learning_rate": 0.00039788744403903604,
+      "loss": 0.0734,
+      "step": 513
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.23908165097236633,
+      "learning_rate": 0.00039751137606896907,
+      "loss": 0.0718,
+      "step": 514
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.37807080149650574,
+      "learning_rate": 0.00039713479532569646,
+      "loss": 0.1495,
+      "step": 515
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.16840259730815887,
+      "learning_rate": 0.00039675770311827337,
+      "loss": 0.0491,
+      "step": 516
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35179728269577026,
+      "learning_rate": 0.00039638010075753274,
+      "loss": 0.0839,
+      "step": 517
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3631207048892975,
+      "learning_rate": 0.00039600198955608084,
+      "loss": 0.1348,
+      "step": 518
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.38650691509246826,
+      "learning_rate": 0.00039562337082829304,
+      "loss": 0.15,
+      "step": 519
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2523843050003052,
+      "learning_rate": 0.00039524424589030866,
+      "loss": 0.1172,
+      "step": 520
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2690166234970093,
+      "learning_rate": 0.00039486461606002686,
+      "loss": 0.0619,
+      "step": 521
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.31193405389785767,
+      "learning_rate": 0.0003944844826571018,
+      "loss": 0.0834,
+      "step": 522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.21751855313777924,
+      "learning_rate": 0.00039410384700293814,
+      "loss": 0.068,
+      "step": 523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34191232919692993,
+      "learning_rate": 0.0003937227104206865,
+      "loss": 0.1337,
+      "step": 524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.34457269310951233,
+      "learning_rate": 0.0003933410742352388,
+      "loss": 0.0929,
+      "step": 525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.22599942982196808,
+      "learning_rate": 0.0003929589397732236,
+      "loss": 0.0899,
+      "step": 526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.23162932693958282,
+      "learning_rate": 0.0003925763083630017,
+      "loss": 0.0869,
+      "step": 527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19502510130405426,
+      "learning_rate": 0.00039219318133466104,
+      "loss": 0.0834,
+      "step": 528
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2539670169353485,
+      "learning_rate": 0.0003918095600200128,
+      "loss": 0.0589,
+      "step": 529
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.15578749775886536,
+      "learning_rate": 0.00039142544575258614,
+      "loss": 0.0471,
+      "step": 530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.41006144881248474,
+      "learning_rate": 0.00039104083986762396,
+      "loss": 0.1215,
+      "step": 531
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3161672055721283,
+      "learning_rate": 0.00039065574370207785,
+      "loss": 0.0599,
+      "step": 532
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2556127607822418,
+      "learning_rate": 0.00039027015859460394,
+      "loss": 0.0882,
+      "step": 533
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5484500527381897,
+      "learning_rate": 0.000389884085885558,
+      "loss": 0.1342,
+      "step": 534
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3688224256038666,
+      "learning_rate": 0.0003894975269169906,
+      "loss": 0.062,
+      "step": 535
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6328185796737671,
+      "learning_rate": 0.0003891104830326427,
+      "loss": 0.1068,
+      "step": 536
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5094593167304993,
+      "learning_rate": 0.00038872295557794103,
+      "loss": 0.0593,
+      "step": 537
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.44920942187309265,
+      "learning_rate": 0.0003883349458999931,
+      "loss": 0.1134,
+      "step": 538
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.25559201836586,
+      "learning_rate": 0.0003879464553475828,
+      "loss": 0.0842,
+      "step": 539
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.24992522597312927,
+      "learning_rate": 0.0003875574852711656,
+      "loss": 0.0684,
+      "step": 540
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7482407093048096,
+      "learning_rate": 0.0003871680370228639,
+      "loss": 0.1698,
+      "step": 541
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.42716777324676514,
+      "learning_rate": 0.00038677811195646233,
+      "loss": 0.1335,
+      "step": 542
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5867021083831787,
+      "learning_rate": 0.0003863877114274029,
+      "loss": 0.153,
+      "step": 543
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.14882822334766388,
+      "learning_rate": 0.0003859968367927805,
+      "loss": 0.0548,
+      "step": 544
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.16213174164295197,
+      "learning_rate": 0.0003856054894113381,
+      "loss": 0.0859,
+      "step": 545
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.13216906785964966,
+      "learning_rate": 0.0003852136706434619,
+      "loss": 0.0837,
+      "step": 546
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28230682015419006,
+      "learning_rate": 0.00038482138185117685,
+      "loss": 0.0746,
+      "step": 547
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.15776745975017548,
+      "learning_rate": 0.0003844286243981417,
+      "loss": 0.0758,
+      "step": 548
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.38748612999916077,
+      "learning_rate": 0.0003840353996496444,
+      "loss": 0.0946,
+      "step": 549
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4377779960632324,
+      "learning_rate": 0.0003836417089725971,
+      "loss": 0.078,
+      "step": 550
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4776962101459503,
+      "learning_rate": 0.0003832475537355319,
+      "loss": 0.0996,
+      "step": 551
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.16078083217144012,
+      "learning_rate": 0.00038285293530859553,
+      "loss": 0.0813,
+      "step": 552
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19620949029922485,
+      "learning_rate": 0.00038245785506354514,
+      "loss": 0.0716,
+      "step": 553
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.23539945483207703,
+      "learning_rate": 0.0003820623143737427,
+      "loss": 0.0727,
+      "step": 554
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2797366678714752,
+      "learning_rate": 0.0003816663146141514,
+      "loss": 0.0307,
+      "step": 555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31704849004745483,
+      "learning_rate": 0.00038126985716132976,
+      "loss": 0.0522,
+      "step": 556
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.038294792175293,
+      "learning_rate": 0.00038087294339342765,
+      "loss": 0.1602,
+      "step": 557
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.39535316824913025,
+      "learning_rate": 0.00038047557469018077,
+      "loss": 0.0672,
+      "step": 558
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5337291359901428,
+      "learning_rate": 0.00038007775243290666,
+      "loss": 0.238,
+      "step": 559
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7618711590766907,
+      "learning_rate": 0.0003796794780044992,
+      "loss": 0.0741,
+      "step": 560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3507292568683624,
+      "learning_rate": 0.0003792807527894242,
+      "loss": 0.1035,
+      "step": 561
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29699352383613586,
+      "learning_rate": 0.00037888157817371455,
+      "loss": 0.0732,
+      "step": 562
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1690889596939087,
+      "learning_rate": 0.0003784819555449651,
+      "loss": 0.0625,
+      "step": 563
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28516581654548645,
+      "learning_rate": 0.0003780818862923284,
+      "loss": 0.0705,
+      "step": 564
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3408360481262207,
+      "learning_rate": 0.00037768137180650913,
+      "loss": 0.1025,
+      "step": 565
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28147757053375244,
+      "learning_rate": 0.00037728041347976005,
+      "loss": 0.0495,
+      "step": 566
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31090235710144043,
+      "learning_rate": 0.00037687901270587655,
+      "loss": 0.0874,
+      "step": 567
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.29111558198928833,
+      "learning_rate": 0.00037647717088019217,
+      "loss": 0.0589,
+      "step": 568
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.08437130600214005,
+      "eval_runtime": 14.722,
+      "eval_samples_per_second": 32.4,
+      "eval_steps_per_second": 8.151,
+      "step": 568
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.14942067861557007,
+      "learning_rate": 0.0003760748893995736,
+      "loss": 0.0391,
+      "step": 569
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.28915029764175415,
+      "learning_rate": 0.0003756721696624156,
+      "loss": 0.0522,
+      "step": 570
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.1540856510400772,
+      "learning_rate": 0.0003752690130686367,
+      "loss": 0.0473,
+      "step": 571
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.31572067737579346,
+      "learning_rate": 0.0003748654210196739,
+      "loss": 0.058,
+      "step": 572
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4497004449367523,
+      "learning_rate": 0.0003744613949184779,
+      "loss": 0.0937,
+      "step": 573
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.48680734634399414,
+      "learning_rate": 0.0003740569361695082,
+      "loss": 0.0925,
+      "step": 574
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3604874610900879,
+      "learning_rate": 0.00037365204617872836,
+      "loss": 0.0273,
+      "step": 575
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.31378456950187683,
+      "learning_rate": 0.0003732467263536008,
+      "loss": 0.048,
+      "step": 576
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.37197205424308777,
+      "learning_rate": 0.0003728409781030824,
+      "loss": 0.0445,
+      "step": 577
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.09396038949489594,
+      "learning_rate": 0.00037243480283761913,
+      "loss": 0.0102,
+      "step": 578
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.4994851052761078,
+      "learning_rate": 0.00037202820196914133,
+      "loss": 0.074,
+      "step": 579
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.20099425315856934,
+      "learning_rate": 0.0003716211769110589,
+      "loss": 0.0239,
+      "step": 580
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.33086463809013367,
+      "learning_rate": 0.0003712137290782561,
+      "loss": 0.0305,
+      "step": 581
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.3871704041957855,
+      "learning_rate": 0.0003708058598870871,
+      "loss": 0.0309,
+      "step": 582
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.514127790927887,
+      "learning_rate": 0.0003703975707553706,
+      "loss": 0.0639,
+      "step": 583
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.29386666417121887,
+      "learning_rate": 0.000369988863102385,
+      "loss": 0.0778,
+      "step": 584
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.4717571437358856,
+      "learning_rate": 0.0003695797383488638,
+      "loss": 0.0414,
+      "step": 585
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.61000657081604,
+      "learning_rate": 0.0003691701979169903,
+      "loss": 0.0687,
+      "step": 586
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.3639252483844757,
+      "learning_rate": 0.0003687602432303926,
+      "loss": 0.0337,
+      "step": 587
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.20936115086078644,
+      "learning_rate": 0.0003683498757141391,
+      "loss": 0.0232,
+      "step": 588
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5917474031448364,
+      "learning_rate": 0.00036793909679473294,
+      "loss": 0.0564,
+      "step": 589
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.23065000772476196,
+      "learning_rate": 0.00036752790790010767,
+      "loss": 0.0246,
+      "step": 590
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.1876888871192932,
+      "learning_rate": 0.00036711631045962173,
+      "loss": 0.0351,
+      "step": 591
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6483283042907715,
+      "learning_rate": 0.000366704305904054,
+      "loss": 0.0628,
+      "step": 592
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8450531363487244,
+      "learning_rate": 0.0003662918956655983,
+      "loss": 0.0922,
+      "step": 593
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5190649628639221,
+      "learning_rate": 0.00036587908117785887,
+      "loss": 0.0715,
+      "step": 594
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.597562849521637,
+      "learning_rate": 0.000365465863875845,
+      "loss": 0.0728,
+      "step": 595
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5079246759414673,
+      "learning_rate": 0.0003650522451959663,
+      "loss": 0.1145,
+      "step": 596
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.4016817808151245,
+      "learning_rate": 0.0003646382265760276,
+      "loss": 0.1373,
+      "step": 597
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2420119345188141,
+      "learning_rate": 0.00036422380945522426,
+      "loss": 0.0428,
+      "step": 598
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2923775017261505,
+      "learning_rate": 0.00036380899527413646,
+      "loss": 0.0407,
+      "step": 599
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.2968994379043579,
+      "learning_rate": 0.00036339378547472497,
+      "loss": 0.039,
+      "step": 600
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3694530129432678,
+      "learning_rate": 0.0003629781815003256,
+      "loss": 0.0765,
+      "step": 601
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.19854502379894257,
+      "learning_rate": 0.0003625621847956443,
+      "loss": 0.0283,
+      "step": 602
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.16821172833442688,
+      "learning_rate": 0.0003621457968067526,
+      "loss": 0.0551,
+      "step": 603
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5689147114753723,
+      "learning_rate": 0.00036172901898108177,
+      "loss": 0.0818,
+      "step": 604
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.16156058013439178,
+      "learning_rate": 0.0003613118527674185,
+      "loss": 0.0368,
+      "step": 605
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.21022377908229828,
+      "learning_rate": 0.00036089429961589926,
+      "loss": 0.0614,
+      "step": 606
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.3432522714138031,
+      "learning_rate": 0.00036047636097800593,
+      "loss": 0.0393,
+      "step": 607
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2537219524383545,
+      "learning_rate": 0.00036005803830656036,
+      "loss": 0.0852,
+      "step": 608
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.209491565823555,
+      "learning_rate": 0.00035963933305571916,
+      "loss": 0.0476,
+      "step": 609
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.2286662459373474,
+      "learning_rate": 0.00035922024668096883,
+      "loss": 0.0614,
+      "step": 610
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5772972106933594,
+      "learning_rate": 0.00035880078063912105,
+      "loss": 0.0546,
+      "step": 611
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.37829965353012085,
+      "learning_rate": 0.0003583809363883069,
+      "loss": 0.0526,
+      "step": 612
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.1876819133758545,
+      "learning_rate": 0.0003579607153879724,
+      "loss": 0.0339,
+      "step": 613
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.42904049158096313,
+      "learning_rate": 0.0003575401190988732,
+      "loss": 0.0705,
+      "step": 614
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2780819833278656,
+      "learning_rate": 0.0003571191489830693,
+      "loss": 0.0425,
+      "step": 615
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.3338189721107483,
+      "learning_rate": 0.00035669780650392056,
+      "loss": 0.0713,
+      "step": 616
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.2791332006454468,
+      "learning_rate": 0.000356276093126081,
+      "loss": 0.0392,
+      "step": 617
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.4691467881202698,
+      "learning_rate": 0.0003558540103154939,
+      "loss": 0.0756,
+      "step": 618
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.34194234013557434,
+      "learning_rate": 0.00035543155953938674,
+      "loss": 0.057,
+      "step": 619
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.43898969888687134,
+      "learning_rate": 0.00035500874226626633,
+      "loss": 0.1484,
+      "step": 620
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.3562189042568207,
+      "learning_rate": 0.00035458555996591325,
+      "loss": 0.0801,
+      "step": 621
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.2978869080543518,
+      "learning_rate": 0.0003541620141093771,
+      "loss": 0.0422,
+      "step": 622
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.415714293718338,
+      "learning_rate": 0.00035373810616897116,
+      "loss": 0.042,
+      "step": 623
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.28547269105911255,
+      "learning_rate": 0.00035331383761826756,
+      "loss": 0.0722,
+      "step": 624
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.2831112742424011,
+      "learning_rate": 0.00035288920993209173,
+      "loss": 0.0339,
+      "step": 625
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.372010201215744,
+      "learning_rate": 0.00035246422458651766,
+      "loss": 0.0573,
+      "step": 626
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.07014724612236023,
+      "learning_rate": 0.0003520388830588625,
+      "loss": 0.0108,
+      "step": 627
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.5464847087860107,
+      "learning_rate": 0.0003516131868276817,
+      "loss": 0.0871,
+      "step": 628
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.118097685277462,
+      "learning_rate": 0.00035118713737276376,
+      "loss": 0.0176,
+      "step": 629
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.43580326437950134,
+      "learning_rate": 0.00035076073617512475,
+      "loss": 0.0817,
+      "step": 630
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.42866209149360657,
+      "learning_rate": 0.00035033398471700367,
+      "loss": 0.1195,
+      "step": 631
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.42996305227279663,
+      "learning_rate": 0.0003499068844818571,
+      "loss": 0.12,
+      "step": 632
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.4283413290977478,
+      "learning_rate": 0.0003494794369543539,
+      "loss": 0.085,
+      "step": 633
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.693706214427948,
+      "learning_rate": 0.0003490516436203703,
+      "loss": 0.126,
+      "step": 634
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.264961838722229,
+      "learning_rate": 0.00034862350596698456,
+      "loss": 0.0556,
+      "step": 635
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.2530398368835449,
+      "learning_rate": 0.00034819502548247175,
+      "loss": 0.0514,
+      "step": 636
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.18521098792552948,
+      "learning_rate": 0.0003477662036562989,
+      "loss": 0.0387,
+      "step": 637
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.34398314356803894,
+      "learning_rate": 0.00034733704197911937,
+      "loss": 0.1047,
+      "step": 638
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.16019423305988312,
+      "learning_rate": 0.000346907541942768,
+      "loss": 0.0299,
+      "step": 639
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.3269899785518646,
+      "learning_rate": 0.00034647770504025587,
+      "loss": 0.0405,
+      "step": 640
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.46410876512527466,
+      "learning_rate": 0.00034604753276576487,
+      "loss": 0.0855,
+      "step": 641
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.33000048995018005,
+      "learning_rate": 0.000345617026614643,
+      "loss": 0.0759,
+      "step": 642
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.31162315607070923,
+      "learning_rate": 0.0003451861880833986,
+      "loss": 0.0558,
+      "step": 643
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.42918407917022705,
+      "learning_rate": 0.0003447550186696956,
+      "loss": 0.0365,
+      "step": 644
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.2732023298740387,
+      "learning_rate": 0.00034432351987234786,
+      "loss": 0.0616,
+      "step": 645
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.1899593621492386,
+      "learning_rate": 0.00034389169319131476,
+      "loss": 0.0286,
+      "step": 646
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.447968065738678,
+      "learning_rate": 0.0003434595401276947,
+      "loss": 0.0701,
+      "step": 647
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.15938018262386322,
+      "learning_rate": 0.0003430270621837213,
+      "loss": 0.026,
+      "step": 648
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.17606832087039948,
+      "learning_rate": 0.0003425942608627572,
+      "loss": 0.0245,
+      "step": 649
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.49266988039016724,
+      "learning_rate": 0.0003421611376692892,
+      "loss": 0.0823,
+      "step": 650
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3935730755329132,
+      "learning_rate": 0.0003417276941089232,
+      "loss": 0.0426,
+      "step": 651
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5984533429145813,
+      "learning_rate": 0.0003412939316883782,
+      "loss": 0.0833,
+      "step": 652
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.3196690082550049,
+      "learning_rate": 0.00034085985191548217,
+      "loss": 0.0337,
+      "step": 653
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.39022788405418396,
+      "learning_rate": 0.000340425456299166,
+      "loss": 0.0235,
+      "step": 654
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.29681891202926636,
+      "learning_rate": 0.00033999074634945856,
+      "loss": 0.0155,
+      "step": 655
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5547076463699341,
+      "learning_rate": 0.0003395557235774813,
+      "loss": 0.0942,
+      "step": 656
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9071078300476074,
+      "learning_rate": 0.00033912038949544316,
+      "loss": 0.1004,
+      "step": 657
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5410562753677368,
+      "learning_rate": 0.00033868474561663534,
+      "loss": 0.0743,
+      "step": 658
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5785720348358154,
+      "learning_rate": 0.0003382487934554257,
+      "loss": 0.1017,
+      "step": 659
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.60345858335495,
+      "learning_rate": 0.0003378125345272539,
+      "loss": 0.117,
+      "step": 660
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.23607215285301208,
+      "learning_rate": 0.0003373759703486262,
+      "loss": 0.0149,
+      "step": 661
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.3551620543003082,
+      "learning_rate": 0.0003369391024371093,
+      "loss": 0.0435,
+      "step": 662
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.4280162453651428,
+      "learning_rate": 0.00033650193231132657,
+      "loss": 0.1019,
+      "step": 663
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.12040708214044571,
+      "learning_rate": 0.0003360644614909512,
+      "loss": 0.0165,
+      "step": 664
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.6838027238845825,
+      "learning_rate": 0.00033562669149670213,
+      "loss": 0.0909,
+      "step": 665
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.2861780524253845,
+      "learning_rate": 0.00033518862385033786,
+      "loss": 0.0719,
+      "step": 666
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.24380998313426971,
+      "learning_rate": 0.00033475026007465184,
+      "loss": 0.0388,
+      "step": 667
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.41201332211494446,
+      "learning_rate": 0.00033431160169346714,
+      "loss": 0.0442,
+      "step": 668
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3734363615512848,
+      "learning_rate": 0.0003338726502316304,
+      "loss": 0.0687,
+      "step": 669
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.21814176440238953,
+      "learning_rate": 0.00033343340721500743,
+      "loss": 0.0743,
+      "step": 670
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.17123498022556305,
+      "learning_rate": 0.00033299387417047723,
+      "loss": 0.0446,
+      "step": 671
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.3857256770133972,
+      "learning_rate": 0.0003325540526259275,
+      "loss": 0.0524,
+      "step": 672
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7980711460113525,
+      "learning_rate": 0.00033211394411024813,
+      "loss": 0.0786,
+      "step": 673
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.31176111102104187,
+      "learning_rate": 0.00033167355015332713,
+      "loss": 0.0499,
+      "step": 674
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6255938410758972,
+      "learning_rate": 0.0003312328722860445,
+      "loss": 0.0664,
+      "step": 675
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3685753047466278,
+      "learning_rate": 0.00033079191204026713,
+      "loss": 0.0495,
+      "step": 676
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.4045291841030121,
+      "learning_rate": 0.00033035067094884366,
+      "loss": 0.0697,
+      "step": 677
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6035248637199402,
+      "learning_rate": 0.0003299091505455989,
+      "loss": 0.1206,
+      "step": 678
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3399547338485718,
+      "learning_rate": 0.00032946735236532855,
+      "loss": 0.035,
+      "step": 679
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3604506552219391,
+      "learning_rate": 0.0003290252779437939,
+      "loss": 0.1087,
+      "step": 680
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.28665006160736084,
+      "learning_rate": 0.0003285829288177167,
+      "loss": 0.0858,
+      "step": 681
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.41967740654945374,
+      "learning_rate": 0.0003281403065247733,
+      "loss": 0.0851,
+      "step": 682
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.43989259004592896,
+      "learning_rate": 0.00032769741260358997,
+      "loss": 0.0793,
+      "step": 683
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.29274123907089233,
+      "learning_rate": 0.00032725424859373687,
+      "loss": 0.0538,
+      "step": 684
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.27231287956237793,
+      "learning_rate": 0.0003268108160357233,
+      "loss": 0.0692,
+      "step": 685
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.3030160963535309,
+      "learning_rate": 0.0003263671164709918,
+      "loss": 0.0786,
+      "step": 686
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.19824832677841187,
+      "learning_rate": 0.0003259231514419135,
+      "loss": 0.0699,
+      "step": 687
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.23121508955955505,
+      "learning_rate": 0.0003254789224917818,
+      "loss": 0.0499,
+      "step": 688
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.15240328013896942,
+      "learning_rate": 0.0003250344311648079,
+      "loss": 0.0431,
+      "step": 689
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.16165931522846222,
+      "learning_rate": 0.000324589679006115,
+      "loss": 0.0554,
+      "step": 690
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.29693150520324707,
+      "learning_rate": 0.0003241446675617329,
+      "loss": 0.0554,
+      "step": 691
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7295424938201904,
+      "learning_rate": 0.00032369939837859275,
+      "loss": 0.1232,
+      "step": 692
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.43246909976005554,
+      "learning_rate": 0.0003232538730045215,
+      "loss": 0.0598,
+      "step": 693
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.18855467438697815,
+      "learning_rate": 0.00032280809298823723,
+      "loss": 0.0252,
+      "step": 694
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.26345816254615784,
+      "learning_rate": 0.00032236205987934234,
+      "loss": 0.0809,
+      "step": 695
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.497403085231781,
+      "learning_rate": 0.00032191577522831984,
+      "loss": 0.0482,
+      "step": 696
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.2640454173088074,
+      "learning_rate": 0.0003214692405865264,
+      "loss": 0.0538,
+      "step": 697
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.335443377494812,
+      "learning_rate": 0.00032102245750618833,
+      "loss": 0.1,
+      "step": 698
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.37383145093917847,
+      "learning_rate": 0.00032057542754039526,
+      "loss": 0.0767,
+      "step": 699
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3101638853549957,
+      "learning_rate": 0.00032012815224309496,
+      "loss": 0.0499,
+      "step": 700
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.2282581478357315,
+      "learning_rate": 0.00031968063316908815,
+      "loss": 0.0424,
+      "step": 701
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.1553642451763153,
+      "learning_rate": 0.00031923287187402287,
+      "loss": 0.0446,
+      "step": 702
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.1549568623304367,
+      "learning_rate": 0.0003187848699143894,
+      "loss": 0.0252,
+      "step": 703
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.4515601694583893,
+      "learning_rate": 0.00031833662884751416,
+      "loss": 0.0852,
+      "step": 704
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.23497383296489716,
+      "learning_rate": 0.0003178881502315552,
+      "loss": 0.0291,
+      "step": 705
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5249956846237183,
+      "learning_rate": 0.000317439435625496,
+      "loss": 0.0478,
+      "step": 706
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7284122705459595,
+      "learning_rate": 0.0003169904865891405,
+      "loss": 0.0584,
+      "step": 707
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.34631770849227905,
+      "learning_rate": 0.00031654130468310784,
+      "loss": 0.092,
+      "step": 708
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.43634921312332153,
+      "learning_rate": 0.000316091891468826,
+      "loss": 0.0758,
+      "step": 709
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.2862977683544159,
+      "learning_rate": 0.00031564224850852754,
+      "loss": 0.057,
+      "step": 710
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 0.08585863560438156,
+      "eval_runtime": 14.7073,
+      "eval_samples_per_second": 32.433,
+      "eval_steps_per_second": 8.159,
+      "step": 710
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.4469147324562073,
+      "learning_rate": 0.0003151923773652436,
+      "loss": 0.0807,
+      "step": 711
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.25653207302093506,
+      "learning_rate": 0.00031474227960279834,
+      "loss": 0.0618,
+      "step": 712
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.4219815731048584,
+      "learning_rate": 0.0003142919567858039,
+      "loss": 0.0592,
+      "step": 713
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8190480470657349,
+      "learning_rate": 0.0003138414104796545,
+      "loss": 0.103,
+      "step": 714
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.3400764465332031,
+      "learning_rate": 0.0003133906422505215,
+      "loss": 0.0998,
+      "step": 715
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.21949267387390137,
+      "learning_rate": 0.0003129396536653474,
+      "loss": 0.0395,
+      "step": 716
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.30582305788993835,
+      "learning_rate": 0.0003124884462918411,
+      "loss": 0.0835,
+      "step": 717
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.11762493848800659,
+      "learning_rate": 0.0003120370216984716,
+      "loss": 0.026,
+      "step": 718
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.1867324858903885,
+      "learning_rate": 0.00031158538145446314,
+      "loss": 0.0544,
+      "step": 719
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.15806153416633606,
+      "learning_rate": 0.00031113352712978996,
+      "loss": 0.0406,
+      "step": 720
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2605026662349701,
+      "learning_rate": 0.00031068146029516997,
+      "loss": 0.0431,
+      "step": 721
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.27978816628456116,
+      "learning_rate": 0.00031022918252206005,
+      "loss": 0.0948,
+      "step": 722
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.19412288069725037,
+      "learning_rate": 0.00030977669538265017,
+      "loss": 0.0305,
+      "step": 723
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.42198699712753296,
+      "learning_rate": 0.0003093240004498585,
+      "loss": 0.1205,
+      "step": 724
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.22501109540462494,
+      "learning_rate": 0.0003088710992973249,
+      "loss": 0.0368,
+      "step": 725
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.4651114046573639,
+      "learning_rate": 0.00030841799349940667,
+      "loss": 0.1044,
+      "step": 726
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.4760609567165375,
+      "learning_rate": 0.00030796468463117216,
+      "loss": 0.0829,
+      "step": 727
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.308748722076416,
+      "learning_rate": 0.0003075111742683957,
+      "loss": 0.0382,
+      "step": 728
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.21451212465763092,
+      "learning_rate": 0.0003070574639875521,
+      "loss": 0.0441,
+      "step": 729
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.1944577842950821,
+      "learning_rate": 0.00030660355536581103,
+      "loss": 0.0326,
+      "step": 730
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5249868035316467,
+      "learning_rate": 0.0003061494499810317,
+      "loss": 0.0857,
+      "step": 731
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2004554718732834,
+      "learning_rate": 0.00030569514941175725,
+      "loss": 0.0533,
+      "step": 732
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.21708889305591583,
+      "learning_rate": 0.00030524065523720935,
+      "loss": 0.0562,
+      "step": 733
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.36287248134613037,
+      "learning_rate": 0.00030478596903728267,
+      "loss": 0.1,
+      "step": 734
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.4405117928981781,
+      "learning_rate": 0.0003043310923925394,
+      "loss": 0.0929,
+      "step": 735
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.3343874514102936,
+      "learning_rate": 0.0003038760268842036,
+      "loss": 0.0549,
+      "step": 736
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.2254178822040558,
+      "learning_rate": 0.00030342077409415606,
+      "loss": 0.0495,
+      "step": 737
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19972631335258484,
+      "learning_rate": 0.00030296533560492854,
+      "loss": 0.0301,
+      "step": 738
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.19470427930355072,
+      "learning_rate": 0.0003025097129996983,
+      "loss": 0.0485,
+      "step": 739
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.34024572372436523,
+      "learning_rate": 0.0003020539078622824,
+      "loss": 0.0509,
+      "step": 740
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.26424598693847656,
+      "learning_rate": 0.00030159792177713294,
+      "loss": 0.0293,
+      "step": 741
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.3307158946990967,
+      "learning_rate": 0.00030114175632933043,
+      "loss": 0.0302,
+      "step": 742
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.448448121547699,
+      "learning_rate": 0.0003006854131045793,
+      "loss": 0.0683,
+      "step": 743
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.4010024070739746,
+      "learning_rate": 0.0003002288936892017,
+      "loss": 0.078,
+      "step": 744
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.22475454211235046,
+      "learning_rate": 0.0002997721996701324,
+      "loss": 0.0303,
+      "step": 745
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.31867673993110657,
+      "learning_rate": 0.000299315332634913,
+      "loss": 0.0521,
+      "step": 746
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.2040076106786728,
+      "learning_rate": 0.0002988582941716867,
+      "loss": 0.0225,
+      "step": 747
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.1286782026290894,
+      "learning_rate": 0.00029840108586919246,
+      "loss": 0.0833,
+      "step": 748
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6526787877082825,
+      "learning_rate": 0.00029794370931675963,
+      "loss": 0.1085,
+      "step": 749
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.3272201418876648,
+      "learning_rate": 0.00029748616610430264,
+      "loss": 0.0213,
+      "step": 750
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5573351383209229,
+      "learning_rate": 0.0002970284578223149,
+      "loss": 0.0478,
+      "step": 751
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.31984782218933105,
+      "learning_rate": 0.00029657058606186393,
+      "loss": 0.0353,
+      "step": 752
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.712653398513794,
+      "learning_rate": 0.00029611255241458533,
+      "loss": 0.0787,
+      "step": 753
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5753667950630188,
+      "learning_rate": 0.00029565435847267766,
+      "loss": 0.1024,
+      "step": 754
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.44150909781455994,
+      "learning_rate": 0.00029519600582889657,
+      "loss": 0.0261,
+      "step": 755
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2625367343425751,
+      "learning_rate": 0.00029473749607654914,
+      "loss": 0.0685,
+      "step": 756
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.48232585191726685,
+      "learning_rate": 0.00029427883080948905,
+      "loss": 0.0299,
+      "step": 757
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2749970853328705,
+      "learning_rate": 0.00029382001162211026,
+      "loss": 0.022,
+      "step": 758
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.2759183645248413,
+      "learning_rate": 0.00029336104010934186,
+      "loss": 0.0417,
+      "step": 759
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.11506503075361252,
+      "learning_rate": 0.0002929019178666425,
+      "loss": 0.0114,
+      "step": 760
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.10981517285108566,
+      "learning_rate": 0.0002924426464899947,
+      "loss": 0.0132,
+      "step": 761
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2663099467754364,
+      "learning_rate": 0.0002919832275758994,
+      "loss": 0.0628,
+      "step": 762
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6475871205329895,
+      "learning_rate": 0.0002915236627213705,
+      "loss": 0.0819,
+      "step": 763
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.3743927478790283,
+      "learning_rate": 0.00029106395352392913,
+      "loss": 0.0526,
+      "step": 764
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.2879592776298523,
+      "learning_rate": 0.0002906041015815983,
+      "loss": 0.0441,
+      "step": 765
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.5118088126182556,
+      "learning_rate": 0.0002901441084928969,
+      "loss": 0.0304,
+      "step": 766
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8219065070152283,
+      "learning_rate": 0.000289683975856835,
+      "loss": 0.0682,
+      "step": 767
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.45978665351867676,
+      "learning_rate": 0.00028922370527290715,
+      "loss": 0.0385,
+      "step": 768
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.3295181095600128,
+      "learning_rate": 0.000288763298341088,
+      "loss": 0.0219,
+      "step": 769
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.9101231694221497,
+      "learning_rate": 0.00028830275666182564,
+      "loss": 0.1396,
+      "step": 770
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.4265996217727661,
+      "learning_rate": 0.000287842081836037,
+      "loss": 0.0514,
+      "step": 771
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6179928183555603,
+      "learning_rate": 0.00028738127546510165,
+      "loss": 0.0615,
+      "step": 772
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.39278754591941833,
+      "learning_rate": 0.00028692033915085635,
+      "loss": 0.0422,
+      "step": 773
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.4660128653049469,
+      "learning_rate": 0.00028645927449558986,
+      "loss": 0.1055,
+      "step": 774
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6555572152137756,
+      "learning_rate": 0.0002859980831020366,
+      "loss": 0.0934,
+      "step": 775
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.4676145613193512,
+      "learning_rate": 0.0002855367665733722,
+      "loss": 0.0624,
+      "step": 776
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6434176564216614,
+      "learning_rate": 0.0002850753265132066,
+      "loss": 0.0415,
+      "step": 777
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2195662558078766,
+      "learning_rate": 0.0002846137645255796,
+      "loss": 0.0502,
+      "step": 778
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.33307918906211853,
+      "learning_rate": 0.00028415208221495465,
+      "loss": 0.0692,
+      "step": 779
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.16945867240428925,
+      "learning_rate": 0.0002836902811862136,
+      "loss": 0.0296,
+      "step": 780
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3181239366531372,
+      "learning_rate": 0.00028322836304465093,
+      "loss": 0.056,
+      "step": 781
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.2984309494495392,
+      "learning_rate": 0.000282766329395968,
+      "loss": 0.0603,
+      "step": 782
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.19178460538387299,
+      "learning_rate": 0.0002823041818462681,
+      "loss": 0.0315,
+      "step": 783
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.3132456839084625,
+      "learning_rate": 0.0002818419220020502,
+      "loss": 0.0421,
+      "step": 784
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.4850837290287018,
+      "learning_rate": 0.00028137955147020355,
+      "loss": 0.0835,
+      "step": 785
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2325926274061203,
+      "learning_rate": 0.00028091707185800245,
+      "loss": 0.033,
+      "step": 786
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5203066468238831,
+      "learning_rate": 0.0002804544847731001,
+      "loss": 0.0766,
+      "step": 787
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.2967028021812439,
+      "learning_rate": 0.00027999179182352347,
+      "loss": 0.0287,
+      "step": 788
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.42808797955513,
+      "learning_rate": 0.0002795289946176674,
+      "loss": 0.039,
+      "step": 789
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.36871954798698425,
+      "learning_rate": 0.00027906609476428937,
+      "loss": 0.0388,
+      "step": 790
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.31911173462867737,
+      "learning_rate": 0.0002786030938725034,
+      "loss": 0.0575,
+      "step": 791
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.2864239513874054,
+      "learning_rate": 0.00027813999355177476,
+      "loss": 0.0711,
+      "step": 792
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.11181977391242981,
+      "learning_rate": 0.0002776767954119147,
+      "loss": 0.0126,
+      "step": 793
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.32950931787490845,
+      "learning_rate": 0.0002772135010630741,
+      "loss": 0.025,
+      "step": 794
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.3144387900829315,
+      "learning_rate": 0.0002767501121157386,
+      "loss": 0.0244,
+      "step": 795
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7986451387405396,
+      "learning_rate": 0.0002762866301807222,
+      "loss": 0.013,
+      "step": 796
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.11357055604457855,
+      "learning_rate": 0.0002758230568691627,
+      "loss": 0.0105,
+      "step": 797
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8880018591880798,
+      "learning_rate": 0.00027535939379251523,
+      "loss": 0.1036,
+      "step": 798
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5905174016952515,
+      "learning_rate": 0.000274895642562547,
+      "loss": 0.1287,
+      "step": 799
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.1973883956670761,
+      "learning_rate": 0.0002744318047913318,
+      "loss": 0.0161,
+      "step": 800
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.38918519020080566,
+      "learning_rate": 0.00027396788209124387,
+      "loss": 0.0428,
+      "step": 801
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.2719264328479767,
+      "learning_rate": 0.0002735038760749531,
+      "loss": 0.0496,
+      "step": 802
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3014307916164398,
+      "learning_rate": 0.0002730397883554189,
+      "loss": 0.0241,
+      "step": 803
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.25686872005462646,
+      "learning_rate": 0.00027257562054588453,
+      "loss": 0.0672,
+      "step": 804
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.30520740151405334,
+      "learning_rate": 0.00027211137425987175,
+      "loss": 0.0376,
+      "step": 805
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3014354407787323,
+      "learning_rate": 0.00027164705111117516,
+      "loss": 0.0201,
+      "step": 806
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.1687713861465454,
+      "learning_rate": 0.0002711826527138565,
+      "loss": 0.0328,
+      "step": 807
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.31807759404182434,
+      "learning_rate": 0.00027071818068223906,
+      "loss": 0.121,
+      "step": 808
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6311066150665283,
+      "learning_rate": 0.00027025363663090216,
+      "loss": 0.0745,
+      "step": 809
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.12313732504844666,
+      "learning_rate": 0.0002697890221746754,
+      "loss": 0.0103,
+      "step": 810
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7472290396690369,
+      "learning_rate": 0.00026932433892863324,
+      "loss": 0.0935,
+      "step": 811
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.3236583173274994,
+      "learning_rate": 0.00026885958850808914,
+      "loss": 0.0592,
+      "step": 812
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6342505216598511,
+      "learning_rate": 0.00026839477252859007,
+      "loss": 0.0919,
+      "step": 813
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.37357425689697266,
+      "learning_rate": 0.0002679298926059109,
+      "loss": 0.0426,
+      "step": 814
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6023468375205994,
+      "learning_rate": 0.000267464950356049,
+      "loss": 0.0756,
+      "step": 815
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.35651543736457825,
+      "learning_rate": 0.0002669999473952181,
+      "loss": 0.0323,
+      "step": 816
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.4318545162677765,
+      "learning_rate": 0.00026653488533984307,
+      "loss": 0.1178,
+      "step": 817
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.31720227003097534,
+      "learning_rate": 0.00026606976580655415,
+      "loss": 0.0974,
+      "step": 818
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5617119669914246,
+      "learning_rate": 0.00026560459041218156,
+      "loss": 0.1098,
+      "step": 819
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5732232332229614,
+      "learning_rate": 0.00026513936077374954,
+      "loss": 0.0949,
+      "step": 820
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.33141466975212097,
+      "learning_rate": 0.00026467407850847105,
+      "loss": 0.0417,
+      "step": 821
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.41724076867103577,
+      "learning_rate": 0.00026420874523374173,
+      "loss": 0.0466,
+      "step": 822
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.24849863350391388,
+      "learning_rate": 0.0002637433625671347,
+      "loss": 0.0536,
+      "step": 823
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.3150210380554199,
+      "learning_rate": 0.00026327793212639486,
+      "loss": 0.0806,
+      "step": 824
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2865166664123535,
+      "learning_rate": 0.00026281245552943293,
+      "loss": 0.0533,
+      "step": 825
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.2756612300872803,
+      "learning_rate": 0.00026234693439432043,
+      "loss": 0.0504,
+      "step": 826
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.9047459363937378,
+      "learning_rate": 0.0002618813703392833,
+      "loss": 0.109,
+      "step": 827
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.10576523840427399,
+      "learning_rate": 0.00026141576498269706,
+      "loss": 0.0231,
+      "step": 828
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.4203824996948242,
+      "learning_rate": 0.00026095011994308056,
+      "loss": 0.0727,
+      "step": 829
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.4648183584213257,
+      "learning_rate": 0.0002604844368390905,
+      "loss": 0.1066,
+      "step": 830
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.2482835054397583,
+      "learning_rate": 0.00026001871728951624,
+      "loss": 0.0237,
+      "step": 831
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.4374096393585205,
+      "learning_rate": 0.00025955296291327356,
+      "loss": 0.0934,
+      "step": 832
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.3488870859146118,
+      "learning_rate": 0.00025908717532939946,
+      "loss": 0.0638,
+      "step": 833
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5614967942237854,
+      "learning_rate": 0.00025862135615704613,
+      "loss": 0.0827,
+      "step": 834
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.30991917848587036,
+      "learning_rate": 0.0002581555070154759,
+      "loss": 0.0438,
+      "step": 835
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.44601985812187195,
+      "learning_rate": 0.00025768962952405503,
+      "loss": 0.0797,
+      "step": 836
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3628086745738983,
+      "learning_rate": 0.00025722372530224844,
+      "loss": 0.0366,
+      "step": 837
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.2644861936569214,
+      "learning_rate": 0.000256757795969614,
+      "loss": 0.0331,
+      "step": 838
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.4585146903991699,
+      "learning_rate": 0.0002562918431457967,
+      "loss": 0.0635,
+      "step": 839
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.4738370478153229,
+      "learning_rate": 0.0002558258684505233,
+      "loss": 0.0599,
+      "step": 840
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6536511182785034,
+      "learning_rate": 0.00025535987350359664,
+      "loss": 0.077,
+      "step": 841
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.43449538946151733,
+      "learning_rate": 0.00025489385992489,
+      "loss": 0.0432,
+      "step": 842
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5419031977653503,
+      "learning_rate": 0.0002544278293343411,
+      "loss": 0.093,
+      "step": 843
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.30555063486099243,
+      "learning_rate": 0.0002539617833519472,
+      "loss": 0.0572,
+      "step": 844
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.18094651401042938,
+      "learning_rate": 0.0002534957235977589,
+      "loss": 0.0353,
+      "step": 845
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.34586626291275024,
+      "learning_rate": 0.00025302965169187467,
+      "loss": 0.0554,
+      "step": 846
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.38212889432907104,
+      "learning_rate": 0.00025256356925443507,
+      "loss": 0.0624,
+      "step": 847
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.5566253066062927,
+      "learning_rate": 0.00025209747790561754,
+      "loss": 0.0603,
+      "step": 848
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2991026043891907,
+      "learning_rate": 0.0002516313792656304,
+      "loss": 0.0374,
+      "step": 849
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6026126146316528,
+      "learning_rate": 0.0002511652749547072,
+      "loss": 0.1283,
+      "step": 850
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.28952938318252563,
+      "learning_rate": 0.0002506991665931013,
+      "loss": 0.0708,
+      "step": 851
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3491526246070862,
+      "learning_rate": 0.00025023305580108027,
+      "loss": 0.0536,
+      "step": 852
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 0.08201431483030319,
+      "eval_runtime": 14.6754,
+      "eval_samples_per_second": 32.503,
+      "eval_steps_per_second": 8.177,
+      "step": 852
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7052826881408691,
+      "learning_rate": 0.00024976694419891974,
+      "loss": 0.059,
+      "step": 853
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.43753165006637573,
+      "learning_rate": 0.0002493008334068987,
+      "loss": 0.0751,
+      "step": 854
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.1675650030374527,
+      "learning_rate": 0.00024883472504529287,
+      "loss": 0.0224,
+      "step": 855
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.3790821433067322,
+      "learning_rate": 0.00024836862073436967,
+      "loss": 0.0707,
+      "step": 856
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.3162197470664978,
+      "learning_rate": 0.0002479025220943825,
+      "loss": 0.0375,
+      "step": 857
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5708606243133545,
+      "learning_rate": 0.00024743643074556494,
+      "loss": 0.0632,
+      "step": 858
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.24075205624103546,
+      "learning_rate": 0.00024697034830812535,
+      "loss": 0.0452,
+      "step": 859
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.371979683637619,
+      "learning_rate": 0.00024650427640224114,
+      "loss": 0.0676,
+      "step": 860
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.2892770767211914,
+      "learning_rate": 0.00024603821664805276,
+      "loss": 0.0592,
+      "step": 861
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.14655162394046783,
+      "learning_rate": 0.00024557217066565896,
+      "loss": 0.0161,
+      "step": 862
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.22819265723228455,
+      "learning_rate": 0.0002451061400751101,
+      "loss": 0.0418,
+      "step": 863
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.4637010097503662,
+      "learning_rate": 0.0002446401264964034,
+      "loss": 0.0555,
+      "step": 864
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.30727583169937134,
+      "learning_rate": 0.00024417413154947677,
+      "loss": 0.0258,
+      "step": 865
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.26563793420791626,
+      "learning_rate": 0.00024370815685420338,
+      "loss": 0.0528,
+      "step": 866
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.18691249191761017,
+      "learning_rate": 0.00024324220403038613,
+      "loss": 0.0432,
+      "step": 867
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3845004439353943,
+      "learning_rate": 0.00024277627469775163,
+      "loss": 0.08,
+      "step": 868
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.36308273673057556,
+      "learning_rate": 0.00024231037047594495,
+      "loss": 0.0569,
+      "step": 869
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.43758854269981384,
+      "learning_rate": 0.00024184449298452414,
+      "loss": 0.0451,
+      "step": 870
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.30163124203681946,
+      "learning_rate": 0.00024137864384295388,
+      "loss": 0.0219,
+      "step": 871
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.20218737423419952,
+      "learning_rate": 0.00024091282467060055,
+      "loss": 0.0277,
+      "step": 872
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.2397354692220688,
+      "learning_rate": 0.00024044703708672648,
+      "loss": 0.0307,
+      "step": 873
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.31938987970352173,
+      "learning_rate": 0.00023998128271048374,
+      "loss": 0.0453,
+      "step": 874
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5195713043212891,
+      "learning_rate": 0.00023951556316090952,
+      "loss": 0.0852,
+      "step": 875
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5428398251533508,
+      "learning_rate": 0.00023904988005691953,
+      "loss": 0.0941,
+      "step": 876
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.9265273213386536,
+      "learning_rate": 0.00023858423501730295,
+      "loss": 0.0489,
+      "step": 877
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5896188020706177,
+      "learning_rate": 0.00023811862966071674,
+      "loss": 0.0601,
+      "step": 878
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6030554175376892,
+      "learning_rate": 0.0002376530656056796,
+      "loss": 0.0958,
+      "step": 879
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3220241665840149,
+      "learning_rate": 0.00023718754447056708,
+      "loss": 0.0487,
+      "step": 880
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.49902886152267456,
+      "learning_rate": 0.00023672206787360523,
+      "loss": 0.0457,
+      "step": 881
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3744886517524719,
+      "learning_rate": 0.00023625663743286534,
+      "loss": 0.0771,
+      "step": 882
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.39943140745162964,
+      "learning_rate": 0.0002357912547662584,
+      "loss": 0.0389,
+      "step": 883
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.34382057189941406,
+      "learning_rate": 0.00023532592149152898,
+      "loss": 0.0405,
+      "step": 884
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.22058314085006714,
+      "learning_rate": 0.00023486063922625042,
+      "loss": 0.032,
+      "step": 885
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.32711130380630493,
+      "learning_rate": 0.00023439540958781848,
+      "loss": 0.0589,
+      "step": 886
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.42972657084465027,
+      "learning_rate": 0.0002339302341934459,
+      "loss": 0.0601,
+      "step": 887
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.31368395686149597,
+      "learning_rate": 0.00023346511466015708,
+      "loss": 0.0345,
+      "step": 888
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.28611400723457336,
+      "learning_rate": 0.00023300005260478194,
+      "loss": 0.0432,
+      "step": 889
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5313751101493835,
+      "learning_rate": 0.00023253504964395097,
+      "loss": 0.0427,
+      "step": 890
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.2192300707101822,
+      "learning_rate": 0.00023207010739408908,
+      "loss": 0.0392,
+      "step": 891
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7893845438957214,
+      "learning_rate": 0.00023160522747141,
+      "loss": 0.1338,
+      "step": 892
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5475191473960876,
+      "learning_rate": 0.00023114041149191098,
+      "loss": 0.1458,
+      "step": 893
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.4575919806957245,
+      "learning_rate": 0.00023067566107136685,
+      "loss": 0.0593,
+      "step": 894
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.47522222995758057,
+      "learning_rate": 0.00023021097782532457,
+      "loss": 0.0744,
+      "step": 895
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3471393883228302,
+      "learning_rate": 0.0002297463633690979,
+      "loss": 0.0795,
+      "step": 896
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6821273565292358,
+      "learning_rate": 0.00022928181931776098,
+      "loss": 0.0692,
+      "step": 897
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.4375980496406555,
+      "learning_rate": 0.00022881734728614347,
+      "loss": 0.0704,
+      "step": 898
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.596495509147644,
+      "learning_rate": 0.0002283529488888249,
+      "loss": 0.0744,
+      "step": 899
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.36588358879089355,
+      "learning_rate": 0.00022788862574012824,
+      "loss": 0.0576,
+      "step": 900
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.49629420042037964,
+      "learning_rate": 0.0002274243794541155,
+      "loss": 0.0847,
+      "step": 901
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8968174457550049,
+      "learning_rate": 0.0002269602116445811,
+      "loss": 0.0521,
+      "step": 902
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.260775089263916,
+      "learning_rate": 0.00022649612392504687,
+      "loss": 0.0345,
+      "step": 903
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.33473265171051025,
+      "learning_rate": 0.00022603211790875622,
+      "loss": 0.0483,
+      "step": 904
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.451668918132782,
+      "learning_rate": 0.0002255681952086683,
+      "loss": 0.0862,
+      "step": 905
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6166467070579529,
+      "learning_rate": 0.00022510435743745304,
+      "loss": 0.1038,
+      "step": 906
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.361858069896698,
+      "learning_rate": 0.0002246406062074848,
+      "loss": 0.073,
+      "step": 907
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.42583227157592773,
+      "learning_rate": 0.00022417694313083735,
+      "loss": 0.0923,
+      "step": 908
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.2566489279270172,
+      "learning_rate": 0.00022371336981927788,
+      "loss": 0.0358,
+      "step": 909
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.3075582683086395,
+      "learning_rate": 0.0002232498878842615,
+      "loss": 0.0912,
+      "step": 910
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.1581839770078659,
+      "learning_rate": 0.00022278649893692584,
+      "loss": 0.0309,
+      "step": 911
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.37088707089424133,
+      "learning_rate": 0.00022232320458808532,
+      "loss": 0.074,
+      "step": 912
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.40888333320617676,
+      "learning_rate": 0.00022186000644822522,
+      "loss": 0.0618,
+      "step": 913
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2161816507577896,
+      "learning_rate": 0.00022139690612749672,
+      "loss": 0.0355,
+      "step": 914
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3577941954135895,
+      "learning_rate": 0.00022093390523571067,
+      "loss": 0.0499,
+      "step": 915
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3455984592437744,
+      "learning_rate": 0.0002204710053823326,
+      "loss": 0.0712,
+      "step": 916
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.2223758101463318,
+      "learning_rate": 0.0002200082081764766,
+      "loss": 0.0362,
+      "step": 917
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3027271032333374,
+      "learning_rate": 0.00021954551522689993,
+      "loss": 0.0579,
+      "step": 918
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.12046008557081223,
+      "learning_rate": 0.00021908292814199764,
+      "loss": 0.0227,
+      "step": 919
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.3685935437679291,
+      "learning_rate": 0.00021862044852979652,
+      "loss": 0.0797,
+      "step": 920
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.26438644528388977,
+      "learning_rate": 0.00021815807799794982,
+      "loss": 0.0567,
+      "step": 921
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.44811537861824036,
+      "learning_rate": 0.00021769581815373192,
+      "loss": 0.0623,
+      "step": 922
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.36297371983528137,
+      "learning_rate": 0.000217233670604032,
+      "loss": 0.0512,
+      "step": 923
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.30395954847335815,
+      "learning_rate": 0.00021677163695534913,
+      "loss": 0.0423,
+      "step": 924
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.26092368364334106,
+      "learning_rate": 0.00021630971881378644,
+      "loss": 0.0463,
+      "step": 925
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.639478325843811,
+      "learning_rate": 0.0002158479177850453,
+      "loss": 0.0564,
+      "step": 926
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.25447505712509155,
+      "learning_rate": 0.00021538623547442045,
+      "loss": 0.0317,
+      "step": 927
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.24460607767105103,
+      "learning_rate": 0.00021492467348679345,
+      "loss": 0.0375,
+      "step": 928
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.43972596526145935,
+      "learning_rate": 0.00021446323342662785,
+      "loss": 0.0587,
+      "step": 929
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.08744898438453674,
+      "learning_rate": 0.00021400191689796338,
+      "loss": 0.0074,
+      "step": 930
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.29619458317756653,
+      "learning_rate": 0.00021354072550441018,
+      "loss": 0.0422,
+      "step": 931
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.3273064196109772,
+      "learning_rate": 0.00021307966084914372,
+      "loss": 0.0362,
+      "step": 932
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.2981872260570526,
+      "learning_rate": 0.00021261872453489842,
+      "loss": 0.0246,
+      "step": 933
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7154226899147034,
+      "learning_rate": 0.00021215791816396303,
+      "loss": 0.0856,
+      "step": 934
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.4782339930534363,
+      "learning_rate": 0.00021169724333817443,
+      "loss": 0.0482,
+      "step": 935
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5048168897628784,
+      "learning_rate": 0.00021123670165891208,
+      "loss": 0.0405,
+      "step": 936
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.22893092036247253,
+      "learning_rate": 0.0002107762947270928,
+      "loss": 0.0181,
+      "step": 937
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6863519549369812,
+      "learning_rate": 0.00021031602414316506,
+      "loss": 0.0643,
+      "step": 938
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5500178337097168,
+      "learning_rate": 0.0002098558915071031,
+      "loss": 0.0599,
+      "step": 939
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7170897126197815,
+      "learning_rate": 0.0002093958984184018,
+      "loss": 0.1167,
+      "step": 940
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.48540323972702026,
+      "learning_rate": 0.00020893604647607088,
+      "loss": 0.0443,
+      "step": 941
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.09318219870328903,
+      "learning_rate": 0.0002084763372786295,
+      "loss": 0.0092,
+      "step": 942
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7748222351074219,
+      "learning_rate": 0.00020801677242410067,
+      "loss": 0.0912,
+      "step": 943
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2857852280139923,
+      "learning_rate": 0.00020755735351000537,
+      "loss": 0.0313,
+      "step": 944
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.483059287071228,
+      "learning_rate": 0.00020709808213335758,
+      "loss": 0.0768,
+      "step": 945
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.2753046751022339,
+      "learning_rate": 0.0002066389598906582,
+      "loss": 0.0339,
+      "step": 946
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.29203662276268005,
+      "learning_rate": 0.00020617998837788975,
+      "loss": 0.0185,
+      "step": 947
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.3008381426334381,
+      "learning_rate": 0.00020572116919051098,
+      "loss": 0.0574,
+      "step": 948
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6928682923316956,
+      "learning_rate": 0.0002052625039234509,
+      "loss": 0.051,
+      "step": 949
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.27521032094955444,
+      "learning_rate": 0.00020480399417110352,
+      "loss": 0.05,
+      "step": 950
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.1467350274324417,
+      "learning_rate": 0.00020434564152732238,
+      "loss": 0.0254,
+      "step": 951
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.2940123379230499,
+      "learning_rate": 0.00020388744758541462,
+      "loss": 0.0521,
+      "step": 952
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2931191325187683,
+      "learning_rate": 0.00020342941393813613,
+      "loss": 0.0394,
+      "step": 953
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.2518831491470337,
+      "learning_rate": 0.00020297154217768513,
+      "loss": 0.041,
+      "step": 954
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.48200809955596924,
+      "learning_rate": 0.00020251383389569743,
+      "loss": 0.0693,
+      "step": 955
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.4188168942928314,
+      "learning_rate": 0.0002020562906832404,
+      "loss": 0.0594,
+      "step": 956
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.4201320707798004,
+      "learning_rate": 0.00020159891413080755,
+      "loss": 0.0737,
+      "step": 957
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.4236721396446228,
+      "learning_rate": 0.00020114170582831342,
+      "loss": 0.0443,
+      "step": 958
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.39196375012397766,
+      "learning_rate": 0.00020068466736508704,
+      "loss": 0.0728,
+      "step": 959
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5320992469787598,
+      "learning_rate": 0.00020022780032986765,
+      "loss": 0.0416,
+      "step": 960
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.12672173976898193,
+      "learning_rate": 0.00019977110631079836,
+      "loss": 0.0191,
+      "step": 961
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6431661248207092,
+      "learning_rate": 0.0001993145868954207,
+      "loss": 0.038,
+      "step": 962
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.29868853092193604,
+      "learning_rate": 0.00019885824367066955,
+      "loss": 0.0245,
+      "step": 963
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5283737778663635,
+      "learning_rate": 0.0001984020782228671,
+      "loss": 0.0722,
+      "step": 964
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5567461252212524,
+      "learning_rate": 0.00019794609213771755,
+      "loss": 0.1026,
+      "step": 965
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.9617827534675598,
+      "learning_rate": 0.00019749028700030181,
+      "loss": 0.078,
+      "step": 966
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5466052889823914,
+      "learning_rate": 0.0001970346643950715,
+      "loss": 0.075,
+      "step": 967
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.18108782172203064,
+      "learning_rate": 0.00019657922590584392,
+      "loss": 0.024,
+      "step": 968
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.4150354862213135,
+      "learning_rate": 0.00019612397311579647,
+      "loss": 0.0656,
+      "step": 969
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.21237897872924805,
+      "learning_rate": 0.0001956689076074607,
+      "loss": 0.0378,
+      "step": 970
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.21254923939704895,
+      "learning_rate": 0.0001952140309627174,
+      "loss": 0.0173,
+      "step": 971
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2641647756099701,
+      "learning_rate": 0.0001947593447627907,
+      "loss": 0.049,
+      "step": 972
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.3682314455509186,
+      "learning_rate": 0.00019430485058824276,
+      "loss": 0.0485,
+      "step": 973
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.2566399574279785,
+      "learning_rate": 0.00019385055001896835,
+      "loss": 0.0388,
+      "step": 974
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.20328454673290253,
+      "learning_rate": 0.000193396444634189,
+      "loss": 0.0259,
+      "step": 975
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5327407717704773,
+      "learning_rate": 0.00019294253601244792,
+      "loss": 0.085,
+      "step": 976
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6960484385490417,
+      "learning_rate": 0.00019248882573160437,
+      "loss": 0.1077,
+      "step": 977
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5338547229766846,
+      "learning_rate": 0.00019203531536882785,
+      "loss": 0.0421,
+      "step": 978
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.1970924586057663,
+      "learning_rate": 0.00019158200650059337,
+      "loss": 0.024,
+      "step": 979
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.4665428698062897,
+      "learning_rate": 0.0001911289007026751,
+      "loss": 0.0549,
+      "step": 980
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.4011171758174896,
+      "learning_rate": 0.00019067599955014156,
+      "loss": 0.0482,
+      "step": 981
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.25179675221443176,
+      "learning_rate": 0.00019022330461734982,
+      "loss": 0.0327,
+      "step": 982
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.2733090817928314,
+      "learning_rate": 0.00018977081747794,
+      "loss": 0.0271,
+      "step": 983
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5745018124580383,
+      "learning_rate": 0.00018931853970483012,
+      "loss": 0.0344,
+      "step": 984
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.12801390886306763,
+      "learning_rate": 0.00018886647287021007,
+      "loss": 0.0144,
+      "step": 985
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.1290263682603836,
+      "learning_rate": 0.00018841461854553681,
+      "loss": 0.0132,
+      "step": 986
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.863158643245697,
+      "learning_rate": 0.00018796297830152853,
+      "loss": 0.1274,
+      "step": 987
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3602030277252197,
+      "learning_rate": 0.00018751155370815895,
+      "loss": 0.0549,
+      "step": 988
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.379295289516449,
+      "learning_rate": 0.00018706034633465257,
+      "loss": 0.0266,
+      "step": 989
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.43907806277275085,
+      "learning_rate": 0.00018660935774947858,
+      "loss": 0.0499,
+      "step": 990
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.822163462638855,
+      "learning_rate": 0.00018615858952034548,
+      "loss": 0.1464,
+      "step": 991
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3563006520271301,
+      "learning_rate": 0.00018570804321419614,
+      "loss": 0.0499,
+      "step": 992
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.082757368683815,
+      "learning_rate": 0.00018525772039720167,
+      "loss": 0.0088,
+      "step": 993
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.349202424287796,
+      "learning_rate": 0.00018480762263475638,
+      "loss": 0.0325,
+      "step": 994
+    },
+    {
+      "epoch": 1.76,
+      "eval_loss": 0.0834248885512352,
+      "eval_runtime": 14.6855,
+      "eval_samples_per_second": 32.481,
+      "eval_steps_per_second": 8.171,
+      "step": 994
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.2513701319694519,
+      "learning_rate": 0.0001843577514914725,
+      "loss": 0.0186,
+      "step": 995
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.4057576656341553,
+      "learning_rate": 0.00018390810853117408,
+      "loss": 0.0348,
+      "step": 996
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.46003130078315735,
+      "learning_rate": 0.0001834586953168923,
+      "loss": 0.0689,
+      "step": 997
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.41909146308898926,
+      "learning_rate": 0.00018300951341085946,
+      "loss": 0.0298,
+      "step": 998
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9829010367393494,
+      "learning_rate": 0.00018256056437450399,
+      "loss": 0.2026,
+      "step": 999
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.31356698274612427,
+      "learning_rate": 0.00018211184976844487,
+      "loss": 0.0263,
+      "step": 1000
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.4269973337650299,
+      "learning_rate": 0.00018166337115248585,
+      "loss": 0.1063,
+      "step": 1001
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.3558803200721741,
+      "learning_rate": 0.00018121513008561064,
+      "loss": 0.0389,
+      "step": 1002
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5086562633514404,
+      "learning_rate": 0.0001807671281259771,
+      "loss": 0.0593,
+      "step": 1003
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.1954115778207779,
+      "learning_rate": 0.00018031936683091186,
+      "loss": 0.0327,
+      "step": 1004
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.1789095103740692,
+      "learning_rate": 0.0001798718477569051,
+      "loss": 0.024,
+      "step": 1005
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.38798651099205017,
+      "learning_rate": 0.0001794245724596048,
+      "loss": 0.0791,
+      "step": 1006
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.34710198640823364,
+      "learning_rate": 0.00017897754249381165,
+      "loss": 0.0571,
+      "step": 1007
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.2781204283237457,
+      "learning_rate": 0.00017853075941347363,
+      "loss": 0.0521,
+      "step": 1008
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.2825307548046112,
+      "learning_rate": 0.00017808422477168023,
+      "loss": 0.0474,
+      "step": 1009
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.23594889044761658,
+      "learning_rate": 0.0001776379401206577,
+      "loss": 0.0295,
+      "step": 1010
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.37222880125045776,
+      "learning_rate": 0.00017719190701176286,
+      "loss": 0.0411,
+      "step": 1011
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.25766775012016296,
+      "learning_rate": 0.00017674612699547846,
+      "loss": 0.0419,
+      "step": 1012
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.27667155861854553,
+      "learning_rate": 0.00017630060162140737,
+      "loss": 0.0325,
+      "step": 1013
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.49651435017585754,
+      "learning_rate": 0.00017585533243826712,
+      "loss": 0.0435,
+      "step": 1014
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7008858323097229,
+      "learning_rate": 0.00017541032099388499,
+      "loss": 0.1405,
+      "step": 1015
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.17448720335960388,
+      "learning_rate": 0.0001749655688351921,
+      "loss": 0.0269,
+      "step": 1016
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2893378734588623,
+      "learning_rate": 0.0001745210775082182,
+      "loss": 0.0417,
+      "step": 1017
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.18504270911216736,
+      "learning_rate": 0.0001740768485580866,
+      "loss": 0.0302,
+      "step": 1018
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.2060771882534027,
+      "learning_rate": 0.00017363288352900818,
+      "loss": 0.047,
+      "step": 1019
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8185610771179199,
+      "learning_rate": 0.00017318918396427675,
+      "loss": 0.1398,
+      "step": 1020
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.46132713556289673,
+      "learning_rate": 0.00017274575140626317,
+      "loss": 0.0776,
+      "step": 1021
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.16016420722007751,
+      "learning_rate": 0.0001723025873964101,
+      "loss": 0.0161,
+      "step": 1022
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6459915041923523,
+      "learning_rate": 0.00017185969347522674,
+      "loss": 0.0711,
+      "step": 1023
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.40434324741363525,
+      "learning_rate": 0.0001714170711822834,
+      "loss": 0.0571,
+      "step": 1024
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.5824777483940125,
+      "learning_rate": 0.00017097472205620607,
+      "loss": 0.1141,
+      "step": 1025
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.4143454134464264,
+      "learning_rate": 0.00017053264763467152,
+      "loss": 0.0558,
+      "step": 1026
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.36720553040504456,
+      "learning_rate": 0.00017009084945440113,
+      "loss": 0.0376,
+      "step": 1027
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.27180641889572144,
+      "learning_rate": 0.00016964932905115632,
+      "loss": 0.054,
+      "step": 1028
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.43961653113365173,
+      "learning_rate": 0.0001692080879597329,
+      "loss": 0.0773,
+      "step": 1029
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.2728005647659302,
+      "learning_rate": 0.00016876712771395552,
+      "loss": 0.0142,
+      "step": 1030
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.5099291205406189,
+      "learning_rate": 0.0001683264498466729,
+      "loss": 0.0404,
+      "step": 1031
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.3162379562854767,
+      "learning_rate": 0.00016788605588975193,
+      "loss": 0.0332,
+      "step": 1032
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.4152194857597351,
+      "learning_rate": 0.0001674459473740726,
+      "loss": 0.0352,
+      "step": 1033
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.3174980878829956,
+      "learning_rate": 0.00016700612582952278,
+      "loss": 0.0777,
+      "step": 1034
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.6996863484382629,
+      "learning_rate": 0.0001665665927849926,
+      "loss": 0.1145,
+      "step": 1035
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.2766638398170471,
+      "learning_rate": 0.0001661273497683697,
+      "loss": 0.0179,
+      "step": 1036
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.45079368352890015,
+      "learning_rate": 0.00016568839830653287,
+      "loss": 0.1081,
+      "step": 1037
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.44944706559181213,
+      "learning_rate": 0.0001652497399253481,
+      "loss": 0.0964,
+      "step": 1038
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.5892651081085205,
+      "learning_rate": 0.00016481137614966223,
+      "loss": 0.1138,
+      "step": 1039
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.29900479316711426,
+      "learning_rate": 0.00016437330850329793,
+      "loss": 0.0429,
+      "step": 1040
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.3094378411769867,
+      "learning_rate": 0.00016393553850904878,
+      "loss": 0.0577,
+      "step": 1041
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.23039738833904266,
+      "learning_rate": 0.00016349806768867345,
+      "loss": 0.026,
+      "step": 1042
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.3328697979450226,
+      "learning_rate": 0.00016306089756289063,
+      "loss": 0.0542,
+      "step": 1043
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.3017619252204895,
+      "learning_rate": 0.0001626240296513739,
+      "loss": 0.0363,
+      "step": 1044
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.15930373966693878,
+      "learning_rate": 0.0001621874654727461,
+      "loss": 0.02,
+      "step": 1045
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.40952980518341064,
+      "learning_rate": 0.00016175120654457432,
+      "loss": 0.0523,
+      "step": 1046
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6540464162826538,
+      "learning_rate": 0.00016131525438336475,
+      "loss": 0.0744,
+      "step": 1047
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.3518769443035126,
+      "learning_rate": 0.00016087961050455685,
+      "loss": 0.05,
+      "step": 1048
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.4166756570339203,
+      "learning_rate": 0.0001604442764225188,
+      "loss": 0.0681,
+      "step": 1049
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.39616283774375916,
+      "learning_rate": 0.00016000925365054154,
+      "loss": 0.0416,
+      "step": 1050
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.20040427148342133,
+      "learning_rate": 0.00015957454370083398,
+      "loss": 0.0284,
+      "step": 1051
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7230433821678162,
+      "learning_rate": 0.00015914014808451784,
+      "loss": 0.1035,
+      "step": 1052
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.1968737691640854,
+      "learning_rate": 0.00015870606831162182,
+      "loss": 0.0281,
+      "step": 1053
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.2677242159843445,
+      "learning_rate": 0.0001582723058910769,
+      "loss": 0.0566,
+      "step": 1054
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.12256369739770889,
+      "learning_rate": 0.00015783886233071076,
+      "loss": 0.0192,
+      "step": 1055
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.311192125082016,
+      "learning_rate": 0.00015740573913724276,
+      "loss": 0.035,
+      "step": 1056
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.36169809103012085,
+      "learning_rate": 0.00015697293781627878,
+      "loss": 0.0755,
+      "step": 1057
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8104953765869141,
+      "learning_rate": 0.00015654045987230532,
+      "loss": 0.0418,
+      "step": 1058
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.35273879766464233,
+      "learning_rate": 0.00015610830680868533,
+      "loss": 0.0266,
+      "step": 1059
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.18313364684581757,
+      "learning_rate": 0.00015567648012765212,
+      "loss": 0.0538,
+      "step": 1060
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.40563294291496277,
+      "learning_rate": 0.0001552449813303044,
+      "loss": 0.046,
+      "step": 1061
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.44426023960113525,
+      "learning_rate": 0.00015481381191660143,
+      "loss": 0.0938,
+      "step": 1062
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.4189196228981018,
+      "learning_rate": 0.00015438297338535702,
+      "loss": 0.0344,
+      "step": 1063
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6641749143600464,
+      "learning_rate": 0.0001539524672342351,
+      "loss": 0.0729,
+      "step": 1064
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.2397107034921646,
+      "learning_rate": 0.00015352229495974422,
+      "loss": 0.0493,
+      "step": 1065
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.17326873540878296,
+      "learning_rate": 0.00015309245805723205,
+      "loss": 0.0131,
+      "step": 1066
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.69275963306427,
+      "learning_rate": 0.00015266295802088064,
+      "loss": 0.1512,
+      "step": 1067
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.3260841369628906,
+      "learning_rate": 0.00015223379634370115,
+      "loss": 0.0602,
+      "step": 1068
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.45368266105651855,
+      "learning_rate": 0.00015180497451752826,
+      "loss": 0.0593,
+      "step": 1069
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.5664640069007874,
+      "learning_rate": 0.0001513764940330155,
+      "loss": 0.0651,
+      "step": 1070
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.21212846040725708,
+      "learning_rate": 0.00015094835637962975,
+      "loss": 0.0232,
+      "step": 1071
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.364945650100708,
+      "learning_rate": 0.0001505205630456461,
+      "loss": 0.0436,
+      "step": 1072
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.48835766315460205,
+      "learning_rate": 0.00015009311551814297,
+      "loss": 0.0885,
+      "step": 1073
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.22198058664798737,
+      "learning_rate": 0.00014966601528299637,
+      "loss": 0.026,
+      "step": 1074
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.2598209083080292,
+      "learning_rate": 0.00014923926382487534,
+      "loss": 0.0306,
+      "step": 1075
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.22863651812076569,
+      "learning_rate": 0.0001488128626272363,
+      "loss": 0.0476,
+      "step": 1076
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.4222748875617981,
+      "learning_rate": 0.00014838681317231822,
+      "loss": 0.0837,
+      "step": 1077
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.5555634498596191,
+      "learning_rate": 0.00014796111694113752,
+      "loss": 0.0747,
+      "step": 1078
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.28704702854156494,
+      "learning_rate": 0.0001475357754134824,
+      "loss": 0.0388,
+      "step": 1079
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.3526531457901001,
+      "learning_rate": 0.00014711079006790828,
+      "loss": 0.0396,
+      "step": 1080
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.41639888286590576,
+      "learning_rate": 0.0001466861623817325,
+      "loss": 0.0954,
+      "step": 1081
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.288824200630188,
+      "learning_rate": 0.0001462618938310288,
+      "loss": 0.0355,
+      "step": 1082
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.257003515958786,
+      "learning_rate": 0.00014583798589062292,
+      "loss": 0.0257,
+      "step": 1083
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.23509138822555542,
+      "learning_rate": 0.00014541444003408682,
+      "loss": 0.0548,
+      "step": 1084
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.3425995707511902,
+      "learning_rate": 0.0001449912577337337,
+      "loss": 0.0691,
+      "step": 1085
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.2606826722621918,
+      "learning_rate": 0.00014456844046061332,
+      "loss": 0.029,
+      "step": 1086
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.10555114597082138,
+      "learning_rate": 0.00014414598968450615,
+      "loss": 0.0166,
+      "step": 1087
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.47334909439086914,
+      "learning_rate": 0.00014372390687391906,
+      "loss": 0.0438,
+      "step": 1088
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.36925116181373596,
+      "learning_rate": 0.00014330219349607947,
+      "loss": 0.0163,
+      "step": 1089
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.2707056999206543,
+      "learning_rate": 0.0001428808510169307,
+      "loss": 0.0709,
+      "step": 1090
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.20645679533481598,
+      "learning_rate": 0.00014245988090112694,
+      "loss": 0.0351,
+      "step": 1091
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.1839297115802765,
+      "learning_rate": 0.00014203928461202763,
+      "loss": 0.025,
+      "step": 1092
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6433751583099365,
+      "learning_rate": 0.0001416190636116932,
+      "loss": 0.0693,
+      "step": 1093
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.44755876064300537,
+      "learning_rate": 0.00014119921936087907,
+      "loss": 0.0788,
+      "step": 1094
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.3716716766357422,
+      "learning_rate": 0.00014077975331903118,
+      "loss": 0.0429,
+      "step": 1095
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.7285858392715454,
+      "learning_rate": 0.00014036066694428096,
+      "loss": 0.035,
+      "step": 1096
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.22279293835163116,
+      "learning_rate": 0.00013994196169343963,
+      "loss": 0.012,
+      "step": 1097
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.17774836719036102,
+      "learning_rate": 0.00013952363902199405,
+      "loss": 0.0238,
+      "step": 1098
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.2312237024307251,
+      "learning_rate": 0.0001391057003841008,
+      "loss": 0.0546,
+      "step": 1099
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6004791855812073,
+      "learning_rate": 0.0001386881472325816,
+      "loss": 0.0625,
+      "step": 1100
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.21409562230110168,
+      "learning_rate": 0.0001382709810189183,
+      "loss": 0.034,
+      "step": 1101
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.295493483543396,
+      "learning_rate": 0.00013785420319324744,
+      "loss": 0.0332,
+      "step": 1102
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.5428887009620667,
+      "learning_rate": 0.00013743781520435573,
+      "loss": 0.0649,
+      "step": 1103
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.331990122795105,
+      "learning_rate": 0.00013702181849967453,
+      "loss": 0.046,
+      "step": 1104
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.4544171392917633,
+      "learning_rate": 0.00013660621452527504,
+      "loss": 0.0563,
+      "step": 1105
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.35486406087875366,
+      "learning_rate": 0.0001361910047258635,
+      "loss": 0.0583,
+      "step": 1106
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.24665361642837524,
+      "learning_rate": 0.00013577619054477575,
+      "loss": 0.0267,
+      "step": 1107
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.07276459783315659,
+      "learning_rate": 0.00013536177342397243,
+      "loss": 0.0064,
+      "step": 1108
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.4690609872341156,
+      "learning_rate": 0.00013494775480403384,
+      "loss": 0.0553,
+      "step": 1109
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.4010032117366791,
+      "learning_rate": 0.00013453413612415512,
+      "loss": 0.0514,
+      "step": 1110
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.3563205301761627,
+      "learning_rate": 0.00013412091882214112,
+      "loss": 0.0553,
+      "step": 1111
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6027369499206543,
+      "learning_rate": 0.00013370810433440167,
+      "loss": 0.0677,
+      "step": 1112
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.5082702040672302,
+      "learning_rate": 0.00013329569409594605,
+      "loss": 0.1265,
+      "step": 1113
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.5313373804092407,
+      "learning_rate": 0.00013288368954037834,
+      "loss": 0.0234,
+      "step": 1114
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.33385488390922546,
+      "learning_rate": 0.00013247209209989242,
+      "loss": 0.0252,
+      "step": 1115
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.36459195613861084,
+      "learning_rate": 0.00013206090320526704,
+      "loss": 0.0211,
+      "step": 1116
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.5477709770202637,
+      "learning_rate": 0.00013165012428586096,
+      "loss": 0.0416,
+      "step": 1117
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.3133089542388916,
+      "learning_rate": 0.0001312397567696074,
+      "loss": 0.036,
+      "step": 1118
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.284045934677124,
+      "learning_rate": 0.00013082980208300971,
+      "loss": 0.0249,
+      "step": 1119
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.3401576578617096,
+      "learning_rate": 0.00013042026165113618,
+      "loss": 0.0281,
+      "step": 1120
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.21280981600284576,
+      "learning_rate": 0.00013001113689761496,
+      "loss": 0.0186,
+      "step": 1121
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.1294207125902176,
+      "learning_rate": 0.00012960242924462957,
+      "loss": 0.0156,
+      "step": 1122
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.11151756346225739,
+      "learning_rate": 0.00012919414011291298,
+      "loss": 0.0111,
+      "step": 1123
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7448397874832153,
+      "learning_rate": 0.0001287862709217439,
+      "loss": 0.0898,
+      "step": 1124
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.4080904424190521,
+      "learning_rate": 0.00012837882308894117,
+      "loss": 0.0323,
+      "step": 1125
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.33506283164024353,
+      "learning_rate": 0.00012797179803085862,
+      "loss": 0.0309,
+      "step": 1126
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.32063427567481995,
+      "learning_rate": 0.00012756519716238096,
+      "loss": 0.0978,
+      "step": 1127
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7773024439811707,
+      "learning_rate": 0.0001271590218969176,
+      "loss": 0.1017,
+      "step": 1128
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2934909164905548,
+      "learning_rate": 0.00012675327364639917,
+      "loss": 0.0192,
+      "step": 1129
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6137431859970093,
+      "learning_rate": 0.0001263479538212717,
+      "loss": 0.1013,
+      "step": 1130
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2201017141342163,
+      "learning_rate": 0.00012594306383049186,
+      "loss": 0.0129,
+      "step": 1131
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.04556626081466675,
+      "learning_rate": 0.00012553860508152212,
+      "loss": 0.0066,
+      "step": 1132
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.17257572710514069,
+      "learning_rate": 0.00012513457898032616,
+      "loss": 0.0133,
+      "step": 1133
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.29365074634552,
+      "learning_rate": 0.0001247309869313633,
+      "loss": 0.0194,
+      "step": 1134
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.032978832721710205,
+      "learning_rate": 0.00012432783033758447,
+      "loss": 0.0047,
+      "step": 1135
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.05853278562426567,
+      "learning_rate": 0.0001239251106004265,
+      "loss": 0.0062,
+      "step": 1136
+    },
+    {
+      "epoch": 2.01,
+      "eval_loss": 0.08742678165435791,
+      "eval_runtime": 14.7019,
+      "eval_samples_per_second": 32.445,
+      "eval_steps_per_second": 8.162,
+      "step": 1136
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.1357499063014984,
+      "learning_rate": 0.00012352282911980782,
+      "loss": 0.0081,
+      "step": 1137
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.17858386039733887,
+      "learning_rate": 0.00012312098729412346,
+      "loss": 0.0148,
+      "step": 1138
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.12391608208417892,
+      "learning_rate": 0.00012271958652023993,
+      "loss": 0.0082,
+      "step": 1139
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.10268434137105942,
+      "learning_rate": 0.0001223186281934909,
+      "loss": 0.0068,
+      "step": 1140
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.26439937949180603,
+      "learning_rate": 0.00012191811370767172,
+      "loss": 0.0149,
+      "step": 1141
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.051827941089868546,
+      "learning_rate": 0.00012151804445503492,
+      "loss": 0.0066,
+      "step": 1142
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.354404091835022,
+      "learning_rate": 0.00012111842182628555,
+      "loss": 0.0539,
+      "step": 1143
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.036191366612911224,
+      "learning_rate": 0.00012071924721057579,
+      "loss": 0.0034,
+      "step": 1144
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.10083664208650589,
+      "learning_rate": 0.00012032052199550083,
+      "loss": 0.0096,
+      "step": 1145
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.1553259789943695,
+      "learning_rate": 0.00011992224756709343,
+      "loss": 0.0077,
+      "step": 1146
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.3661736845970154,
+      "learning_rate": 0.00011952442530981921,
+      "loss": 0.027,
+      "step": 1147
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.25327032804489136,
+      "learning_rate": 0.00011912705660657244,
+      "loss": 0.0212,
+      "step": 1148
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.30773070454597473,
+      "learning_rate": 0.0001187301428386702,
+      "loss": 0.0175,
+      "step": 1149
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.056706108152866364,
+      "learning_rate": 0.00011833368538584863,
+      "loss": 0.0056,
+      "step": 1150
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.24290138483047485,
+      "learning_rate": 0.00011793768562625734,
+      "loss": 0.0165,
+      "step": 1151
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.10750260949134827,
+      "learning_rate": 0.00011754214493645493,
+      "loss": 0.0075,
+      "step": 1152
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.07075980305671692,
+      "learning_rate": 0.00011714706469140449,
+      "loss": 0.0052,
+      "step": 1153
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.03554920107126236,
+      "learning_rate": 0.0001167524462644681,
+      "loss": 0.0043,
+      "step": 1154
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.10947668552398682,
+      "learning_rate": 0.00011635829102740294,
+      "loss": 0.0077,
+      "step": 1155
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.036122776567935944,
+      "learning_rate": 0.00011596460035035572,
+      "loss": 0.0018,
+      "step": 1156
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.036581866443157196,
+      "learning_rate": 0.00011557137560185829,
+      "loss": 0.0033,
+      "step": 1157
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.3237001597881317,
+      "learning_rate": 0.00011517861814882308,
+      "loss": 0.0203,
+      "step": 1158
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.3374278247356415,
+      "learning_rate": 0.00011478632935653805,
+      "loss": 0.0076,
+      "step": 1159
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.3109148144721985,
+      "learning_rate": 0.0001143945105886619,
+      "loss": 0.0252,
+      "step": 1160
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.1825418621301651,
+      "learning_rate": 0.00011400316320721951,
+      "loss": 0.0095,
+      "step": 1161
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.056046485900878906,
+      "learning_rate": 0.00011361228857259709,
+      "loss": 0.0036,
+      "step": 1162
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.10362865775823593,
+      "learning_rate": 0.00011322188804353761,
+      "loss": 0.0048,
+      "step": 1163
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.09634903073310852,
+      "learning_rate": 0.00011283196297713608,
+      "loss": 0.0056,
+      "step": 1164
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.0892496109008789,
+      "learning_rate": 0.00011244251472883446,
+      "loss": 0.0051,
+      "step": 1165
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.9342542886734009,
+      "learning_rate": 0.00011205354465241732,
+      "loss": 0.0462,
+      "step": 1166
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.08937977254390717,
+      "learning_rate": 0.00011166505410000697,
+      "loss": 0.0055,
+      "step": 1167
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.032398249953985214,
+      "learning_rate": 0.00011127704442205897,
+      "loss": 0.0027,
+      "step": 1168
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.20088808238506317,
+      "learning_rate": 0.0001108895169673573,
+      "loss": 0.0328,
+      "step": 1169
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.0885910764336586,
+      "learning_rate": 0.00011050247308300945,
+      "loss": 0.005,
+      "step": 1170
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.009513720870018005,
+      "learning_rate": 0.00011011591411444199,
+      "loss": 0.0009,
+      "step": 1171
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.04859737306833267,
+      "learning_rate": 0.00010972984140539605,
+      "loss": 0.0024,
+      "step": 1172
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.08774285018444061,
+      "learning_rate": 0.00010934425629792214,
+      "loss": 0.0048,
+      "step": 1173
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.09334032982587814,
+      "learning_rate": 0.00010895916013237619,
+      "loss": 0.0022,
+      "step": 1174
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.016344185918569565,
+      "learning_rate": 0.00010857455424741388,
+      "loss": 0.0014,
+      "step": 1175
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.06272843480110168,
+      "learning_rate": 0.00010819043997998721,
+      "loss": 0.0018,
+      "step": 1176
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.42132094502449036,
+      "learning_rate": 0.00010780681866533897,
+      "loss": 0.0221,
+      "step": 1177
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.2525283992290497,
+      "learning_rate": 0.00010742369163699841,
+      "loss": 0.0114,
+      "step": 1178
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.9688707590103149,
+      "learning_rate": 0.00010704106022677645,
+      "loss": 0.0166,
+      "step": 1179
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.14904755353927612,
+      "learning_rate": 0.00010665892576476122,
+      "loss": 0.0044,
+      "step": 1180
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.020122570917010307,
+      "learning_rate": 0.00010627728957931346,
+      "loss": 0.0017,
+      "step": 1181
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.44014814496040344,
+      "learning_rate": 0.00010589615299706187,
+      "loss": 0.0078,
+      "step": 1182
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.6460086107254028,
+      "learning_rate": 0.00010551551734289827,
+      "loss": 0.0398,
+      "step": 1183
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.1972631961107254,
+      "learning_rate": 0.00010513538393997316,
+      "loss": 0.0286,
+      "step": 1184
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.04835314676165581,
+      "learning_rate": 0.00010475575410969138,
+      "loss": 0.0026,
+      "step": 1185
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.4275042712688446,
+      "learning_rate": 0.00010437662917170695,
+      "loss": 0.0136,
+      "step": 1186
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.010206771083176136,
+      "learning_rate": 0.00010399801044391918,
+      "loss": 0.0007,
+      "step": 1187
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.01756274327635765,
+      "learning_rate": 0.00010361989924246737,
+      "loss": 0.0011,
+      "step": 1188
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6749821901321411,
+      "learning_rate": 0.00010324229688172665,
+      "loss": 0.0332,
+      "step": 1189
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.1459697037935257,
+      "learning_rate": 0.00010286520467430357,
+      "loss": 0.0052,
+      "step": 1190
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.0659116879105568,
+      "learning_rate": 0.00010248862393103092,
+      "loss": 0.0037,
+      "step": 1191
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.11409203708171844,
+      "learning_rate": 0.000102112555960964,
+      "loss": 0.0046,
+      "step": 1192
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.19234444200992584,
+      "learning_rate": 0.00010173700207137529,
+      "loss": 0.0084,
+      "step": 1193
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.01539614424109459,
+      "learning_rate": 0.00010136196356775024,
+      "loss": 0.001,
+      "step": 1194
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.006582081783562899,
+      "learning_rate": 0.00010098744175378308,
+      "loss": 0.0005,
+      "step": 1195
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.5718114972114563,
+      "learning_rate": 0.00010061343793137149,
+      "loss": 0.0232,
+      "step": 1196
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.4466152787208557,
+      "learning_rate": 0.00010023995340061292,
+      "loss": 0.0192,
+      "step": 1197
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.3126049041748047,
+      "learning_rate": 9.986698945979946e-05,
+      "loss": 0.0057,
+      "step": 1198
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.7232750058174133,
+      "learning_rate": 9.94945474054135e-05,
+      "loss": 0.0372,
+      "step": 1199
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.00877452827990055,
+      "learning_rate": 9.91226285321235e-05,
+      "loss": 0.0005,
+      "step": 1200
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.023644492030143738,
+      "learning_rate": 9.8751234132779e-05,
+      "loss": 0.0012,
+      "step": 1201
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.008336883969604969,
+      "learning_rate": 9.838036549840668e-05,
+      "loss": 0.0005,
+      "step": 1202
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.1168123185634613,
+      "learning_rate": 9.801002391820527e-05,
+      "loss": 0.0036,
+      "step": 1203
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.4848583936691284,
+      "learning_rate": 9.764021067954146e-05,
+      "loss": 0.0177,
+      "step": 1204
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.32454514503479004,
+      "learning_rate": 9.727092706794555e-05,
+      "loss": 0.0056,
+      "step": 1205
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.1333390176296234,
+      "learning_rate": 9.690217436710646e-05,
+      "loss": 0.0026,
+      "step": 1206
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.09174130856990814,
+      "learning_rate": 9.653395385886787e-05,
+      "loss": 0.0062,
+      "step": 1207
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.01976648159325123,
+      "learning_rate": 9.616626682322327e-05,
+      "loss": 0.0008,
+      "step": 1208
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2334078848361969,
+      "learning_rate": 9.579911453831166e-05,
+      "loss": 0.0069,
+      "step": 1209
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.010424863547086716,
+      "learning_rate": 9.543249828041342e-05,
+      "loss": 0.0007,
+      "step": 1210
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.06431838870048523,
+      "learning_rate": 9.506641932394552e-05,
+      "loss": 0.0021,
+      "step": 1211
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.2083590179681778,
+      "learning_rate": 9.470087894145704e-05,
+      "loss": 0.0092,
+      "step": 1212
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.03698310628533363,
+      "learning_rate": 9.433587840362501e-05,
+      "loss": 0.0018,
+      "step": 1213
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.836845874786377,
+      "learning_rate": 9.397141897924974e-05,
+      "loss": 0.0274,
+      "step": 1214
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.007941615767776966,
+      "learning_rate": 9.360750193525076e-05,
+      "loss": 0.0004,
+      "step": 1215
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.03961505368351936,
+      "learning_rate": 9.324412853666217e-05,
+      "loss": 0.0018,
+      "step": 1216
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.040892381221055984,
+      "learning_rate": 9.28813000466281e-05,
+      "loss": 0.0015,
+      "step": 1217
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.05060145631432533,
+      "learning_rate": 9.25190177263986e-05,
+      "loss": 0.0025,
+      "step": 1218
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.5681605935096741,
+      "learning_rate": 9.215728283532502e-05,
+      "loss": 0.0695,
+      "step": 1219
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.012955489568412304,
+      "learning_rate": 9.179609663085595e-05,
+      "loss": 0.0006,
+      "step": 1220
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6021496653556824,
+      "learning_rate": 9.143546036853279e-05,
+      "loss": 0.0213,
+      "step": 1221
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.011097903363406658,
+      "learning_rate": 9.107537530198464e-05,
+      "loss": 0.0006,
+      "step": 1222
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.009882250800728798,
+      "learning_rate": 9.071584268292515e-05,
+      "loss": 0.0006,
+      "step": 1223
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.020904725417494774,
+      "learning_rate": 9.035686376114749e-05,
+      "loss": 0.0009,
+      "step": 1224
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.1304357051849365,
+      "learning_rate": 8.999843978451977e-05,
+      "loss": 0.0159,
+      "step": 1225
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.019047705456614494,
+      "learning_rate": 8.964057199898148e-05,
+      "loss": 0.0008,
+      "step": 1226
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.05119523033499718,
+      "learning_rate": 8.928326164853811e-05,
+      "loss": 0.0018,
+      "step": 1227
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.01547545101493597,
+      "learning_rate": 8.892650997525794e-05,
+      "loss": 0.0008,
+      "step": 1228
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7677561044692993,
+      "learning_rate": 8.857031821926711e-05,
+      "loss": 0.0276,
+      "step": 1229
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.07431495934724808,
+      "learning_rate": 8.821468761874518e-05,
+      "loss": 0.0029,
+      "step": 1230
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.018387913703918457,
+      "learning_rate": 8.785961940992118e-05,
+      "loss": 0.0008,
+      "step": 1231
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.34750181436538696,
+      "learning_rate": 8.75051148270691e-05,
+      "loss": 0.009,
+      "step": 1232
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.36946621537208557,
+      "learning_rate": 8.715117510250378e-05,
+      "loss": 0.0281,
+      "step": 1233
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.03397729992866516,
+      "learning_rate": 8.67978014665766e-05,
+      "loss": 0.0012,
+      "step": 1234
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.12867552042007446,
+      "learning_rate": 8.644499514767088e-05,
+      "loss": 0.0062,
+      "step": 1235
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.010095584206283092,
+      "learning_rate": 8.609275737219793e-05,
+      "loss": 0.0006,
+      "step": 1236
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.028368016704916954,
+      "learning_rate": 8.57410893645929e-05,
+      "loss": 0.0014,
+      "step": 1237
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.12146025896072388,
+      "learning_rate": 8.538999234731004e-05,
+      "loss": 0.005,
+      "step": 1238
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.0617150254547596,
+      "learning_rate": 8.50394675408191e-05,
+      "loss": 0.0031,
+      "step": 1239
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.005746606737375259,
+      "learning_rate": 8.468951616360038e-05,
+      "loss": 0.0003,
+      "step": 1240
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.4718758761882782,
+      "learning_rate": 8.434013943214097e-05,
+      "loss": 0.0496,
+      "step": 1241
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.008204291574656963,
+      "learning_rate": 8.399133856093061e-05,
+      "loss": 0.0004,
+      "step": 1242
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.5281336307525635,
+      "learning_rate": 8.36431147624569e-05,
+      "loss": 0.0375,
+      "step": 1243
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.006096679251641035,
+      "learning_rate": 8.329546924720177e-05,
+      "loss": 0.0004,
+      "step": 1244
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.040774155408144,
+      "learning_rate": 8.294840322363672e-05,
+      "loss": 0.0015,
+      "step": 1245
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.060770515352487564,
+      "learning_rate": 8.260191789821884e-05,
+      "loss": 0.0016,
+      "step": 1246
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.2820521295070648,
+      "learning_rate": 8.225601447538689e-05,
+      "loss": 0.0085,
+      "step": 1247
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.04571341350674629,
+      "learning_rate": 8.191069415755645e-05,
+      "loss": 0.0018,
+      "step": 1248
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.015350689180195332,
+      "learning_rate": 8.156595814511655e-05,
+      "loss": 0.0011,
+      "step": 1249
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.40809088945388794,
+      "learning_rate": 8.122180763642475e-05,
+      "loss": 0.0129,
+      "step": 1250
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.030470022931694984,
+      "learning_rate": 8.087824382780335e-05,
+      "loss": 0.0014,
+      "step": 1251
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.11672550439834595,
+      "learning_rate": 8.05352679135354e-05,
+      "loss": 0.0031,
+      "step": 1252
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.009694907814264297,
+      "learning_rate": 8.01928810858601e-05,
+      "loss": 0.0006,
+      "step": 1253
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.03160417824983597,
+      "learning_rate": 7.985108453496909e-05,
+      "loss": 0.0015,
+      "step": 1254
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.3607199490070343,
+      "learning_rate": 7.950987944900193e-05,
+      "loss": 0.0141,
+      "step": 1255
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.12325584143400192,
+      "learning_rate": 7.916926701404217e-05,
+      "loss": 0.0052,
+      "step": 1256
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.03867779299616814,
+      "learning_rate": 7.882924841411343e-05,
+      "loss": 0.0016,
+      "step": 1257
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.008120791055262089,
+      "learning_rate": 7.848982483117473e-05,
+      "loss": 0.0004,
+      "step": 1258
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.12468260526657104,
+      "learning_rate": 7.815099744511708e-05,
+      "loss": 0.0053,
+      "step": 1259
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.006734863854944706,
+      "learning_rate": 7.78127674337587e-05,
+      "loss": 0.0005,
+      "step": 1260
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.11184979975223541,
+      "learning_rate": 7.747513597284134e-05,
+      "loss": 0.0023,
+      "step": 1261
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.030304264277219772,
+      "learning_rate": 7.713810423602619e-05,
+      "loss": 0.0018,
+      "step": 1262
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.4860706925392151,
+      "learning_rate": 7.680167339488967e-05,
+      "loss": 0.0358,
+      "step": 1263
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.08765775710344315,
+      "learning_rate": 7.646584461891929e-05,
+      "loss": 0.0031,
+      "step": 1264
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.06726156920194626,
+      "learning_rate": 7.613061907550975e-05,
+      "loss": 0.003,
+      "step": 1265
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.04926925525069237,
+      "learning_rate": 7.579599792995872e-05,
+      "loss": 0.0024,
+      "step": 1266
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.31866374611854553,
+      "learning_rate": 7.546198234546309e-05,
+      "loss": 0.0263,
+      "step": 1267
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.016148075461387634,
+      "learning_rate": 7.512857348311466e-05,
+      "loss": 0.0006,
+      "step": 1268
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.013526340946555138,
+      "learning_rate": 7.479577250189606e-05,
+      "loss": 0.0005,
+      "step": 1269
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.28299248218536377,
+      "learning_rate": 7.446358055867688e-05,
+      "loss": 0.0102,
+      "step": 1270
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.34623104333877563,
+      "learning_rate": 7.413199880820953e-05,
+      "loss": 0.0105,
+      "step": 1271
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.471078485250473,
+      "learning_rate": 7.380102840312541e-05,
+      "loss": 0.0406,
+      "step": 1272
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.015806253999471664,
+      "learning_rate": 7.347067049393091e-05,
+      "loss": 0.0008,
+      "step": 1273
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.2348398119211197,
+      "learning_rate": 7.314092622900285e-05,
+      "loss": 0.0119,
+      "step": 1274
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.018902868032455444,
+      "learning_rate": 7.281179675458527e-05,
+      "loss": 0.0008,
+      "step": 1275
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.07969076931476593,
+      "learning_rate": 7.248328321478512e-05,
+      "loss": 0.0026,
+      "step": 1276
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.09565503150224686,
+      "learning_rate": 7.215538675156804e-05,
+      "loss": 0.0018,
+      "step": 1277
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.10772550851106644,
+      "learning_rate": 7.182810850475494e-05,
+      "loss": 0.0024,
+      "step": 1278
+    },
+    {
+      "epoch": 2.26,
+      "eval_loss": 0.10833004117012024,
+      "eval_runtime": 14.7071,
+      "eval_samples_per_second": 32.433,
+      "eval_steps_per_second": 8.159,
+      "step": 1278
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.16737312078475952,
+      "learning_rate": 7.15014496120172e-05,
+      "loss": 0.005,
+      "step": 1279
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.013761342503130436,
+      "learning_rate": 7.11754112088737e-05,
+      "loss": 0.0006,
+      "step": 1280
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.32918259501457214,
+      "learning_rate": 7.084999442868629e-05,
+      "loss": 0.007,
+      "step": 1281
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.009015708230435848,
+      "learning_rate": 7.052520040265581e-05,
+      "loss": 0.0005,
+      "step": 1282
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.03732029348611832,
+      "learning_rate": 7.020103025981839e-05,
+      "loss": 0.0011,
+      "step": 1283
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.010399947874248028,
+      "learning_rate": 6.987748512704143e-05,
+      "loss": 0.0005,
+      "step": 1284
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.34805890917778015,
+      "learning_rate": 6.955456612901973e-05,
+      "loss": 0.014,
+      "step": 1285
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.03678420931100845,
+      "learning_rate": 6.923227438827159e-05,
+      "loss": 0.0011,
+      "step": 1286
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.09479320049285889,
+      "learning_rate": 6.891061102513479e-05,
+      "loss": 0.002,
+      "step": 1287
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.09461376816034317,
+      "learning_rate": 6.858957715776265e-05,
+      "loss": 0.0032,
+      "step": 1288
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.042890515178442,
+      "learning_rate": 6.826917390212056e-05,
+      "loss": 0.0014,
+      "step": 1289
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.5706468224525452,
+      "learning_rate": 6.79494023719815e-05,
+      "loss": 0.0081,
+      "step": 1290
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.01644587516784668,
+      "learning_rate": 6.763026367892269e-05,
+      "loss": 0.0009,
+      "step": 1291
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.020033186301589012,
+      "learning_rate": 6.731175893232141e-05,
+      "loss": 0.001,
+      "step": 1292
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.13267932832241058,
+      "learning_rate": 6.699388923935118e-05,
+      "loss": 0.0016,
+      "step": 1293
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.01488049328327179,
+      "learning_rate": 6.667665570497813e-05,
+      "loss": 0.0004,
+      "step": 1294
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.07691626995801926,
+      "learning_rate": 6.636005943195683e-05,
+      "loss": 0.0023,
+      "step": 1295
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.36601608991622925,
+      "learning_rate": 6.604410152082683e-05,
+      "loss": 0.0058,
+      "step": 1296
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.053158730268478394,
+      "learning_rate": 6.57287830699084e-05,
+      "loss": 0.0013,
+      "step": 1297
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.0060308403335511684,
+      "learning_rate": 6.541410517529906e-05,
+      "loss": 0.0002,
+      "step": 1298
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.06450387835502625,
+      "learning_rate": 6.510006893086973e-05,
+      "loss": 0.0016,
+      "step": 1299
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.0036570043303072453,
+      "learning_rate": 6.478667542826064e-05,
+      "loss": 0.0002,
+      "step": 1300
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.010067092254757881,
+      "learning_rate": 6.447392575687805e-05,
+      "loss": 0.0003,
+      "step": 1301
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.0020321488846093416,
+      "learning_rate": 6.41618210038899e-05,
+      "loss": 0.0001,
+      "step": 1302
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.07873855531215668,
+      "learning_rate": 6.38503622542223e-05,
+      "loss": 0.0023,
+      "step": 1303
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.021666008979082108,
+      "learning_rate": 6.353955059055597e-05,
+      "loss": 0.0006,
+      "step": 1304
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.1410478949546814,
+      "learning_rate": 6.322938709332196e-05,
+      "loss": 0.0031,
+      "step": 1305
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.01548179890960455,
+      "learning_rate": 6.291987284069849e-05,
+      "loss": 0.0006,
+      "step": 1306
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.10515931993722916,
+      "learning_rate": 6.261100890860668e-05,
+      "loss": 0.0024,
+      "step": 1307
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.05864081159234047,
+      "learning_rate": 6.230279637070704e-05,
+      "loss": 0.0017,
+      "step": 1308
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.007448627147823572,
+      "learning_rate": 6.199523629839591e-05,
+      "loss": 0.0003,
+      "step": 1309
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.1981428563594818,
+      "learning_rate": 6.168832976080133e-05,
+      "loss": 0.0027,
+      "step": 1310
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.002334183780476451,
+      "learning_rate": 6.138207782477976e-05,
+      "loss": 0.0001,
+      "step": 1311
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.12165654450654984,
+      "learning_rate": 6.107648155491202e-05,
+      "loss": 0.0029,
+      "step": 1312
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.038640353828668594,
+      "learning_rate": 6.077154201349966e-05,
+      "loss": 0.0008,
+      "step": 1313
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.004505124408751726,
+      "learning_rate": 6.046726026056154e-05,
+      "loss": 0.0002,
+      "step": 1314
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.3108009696006775,
+      "learning_rate": 6.01636373538299e-05,
+      "loss": 0.0064,
+      "step": 1315
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.003968897275626659,
+      "learning_rate": 5.986067434874662e-05,
+      "loss": 0.0002,
+      "step": 1316
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.026421288028359413,
+      "learning_rate": 5.955837229845965e-05,
+      "loss": 0.0007,
+      "step": 1317
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.007577298209071159,
+      "learning_rate": 5.925673225381939e-05,
+      "loss": 0.0003,
+      "step": 1318
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.009191282093524933,
+      "learning_rate": 5.89557552633751e-05,
+      "loss": 0.0003,
+      "step": 1319
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.0959949716925621,
+      "learning_rate": 5.865544237337117e-05,
+      "loss": 0.0015,
+      "step": 1320
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.014931642450392246,
+      "learning_rate": 5.835579462774312e-05,
+      "loss": 0.0005,
+      "step": 1321
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.023637857288122177,
+      "learning_rate": 5.80568130681148e-05,
+      "loss": 0.0003,
+      "step": 1322
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.07706478983163834,
+      "learning_rate": 5.775849873379393e-05,
+      "loss": 0.0022,
+      "step": 1323
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3625439703464508,
+      "learning_rate": 5.746085266176907e-05,
+      "loss": 0.0146,
+      "step": 1324
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.12187057733535767,
+      "learning_rate": 5.7163875886705824e-05,
+      "loss": 0.0036,
+      "step": 1325
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.0629158765077591,
+      "learning_rate": 5.686756944094282e-05,
+      "loss": 0.0018,
+      "step": 1326
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.0620671845972538,
+      "learning_rate": 5.657193435448896e-05,
+      "loss": 0.0012,
+      "step": 1327
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.00476891128346324,
+      "learning_rate": 5.627697165501927e-05,
+      "loss": 0.0002,
+      "step": 1328
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.019357487559318542,
+      "learning_rate": 5.598268236787138e-05,
+      "loss": 0.0005,
+      "step": 1329
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.21635344624519348,
+      "learning_rate": 5.5689067516041994e-05,
+      "loss": 0.007,
+      "step": 1330
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6375563740730286,
+      "learning_rate": 5.539612812018344e-05,
+      "loss": 0.0377,
+      "step": 1331
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.0075180609710514545,
+      "learning_rate": 5.5103865198600085e-05,
+      "loss": 0.0003,
+      "step": 1332
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.030566420406103134,
+      "learning_rate": 5.481227976724476e-05,
+      "loss": 0.0006,
+      "step": 1333
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.3805221617221832,
+      "learning_rate": 5.45213728397152e-05,
+      "loss": 0.0245,
+      "step": 1334
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.033048298209905624,
+      "learning_rate": 5.423114542725049e-05,
+      "loss": 0.0006,
+      "step": 1335
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.18197427690029144,
+      "learning_rate": 5.3941598538727625e-05,
+      "loss": 0.0034,
+      "step": 1336
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.009043999947607517,
+      "learning_rate": 5.365273318065811e-05,
+      "loss": 0.0004,
+      "step": 1337
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.036380622535943985,
+      "learning_rate": 5.3364550357184325e-05,
+      "loss": 0.0012,
+      "step": 1338
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.02875285968184471,
+      "learning_rate": 5.307705107007593e-05,
+      "loss": 0.0007,
+      "step": 1339
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.006231918465346098,
+      "learning_rate": 5.2790236318726484e-05,
+      "loss": 0.0003,
+      "step": 1340
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.012466797605156898,
+      "learning_rate": 5.2504107100150245e-05,
+      "loss": 0.0006,
+      "step": 1341
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.013808784075081348,
+      "learning_rate": 5.221866440897807e-05,
+      "loss": 0.0004,
+      "step": 1342
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.5687355995178223,
+      "learning_rate": 5.193390923745475e-05,
+      "loss": 0.0199,
+      "step": 1343
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.0935133546590805,
+      "learning_rate": 5.1649842575434844e-05,
+      "loss": 0.0036,
+      "step": 1344
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6903636455535889,
+      "learning_rate": 5.136646541037956e-05,
+      "loss": 0.0154,
+      "step": 1345
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.07615023851394653,
+      "learning_rate": 5.108377872735351e-05,
+      "loss": 0.002,
+      "step": 1346
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.017100483179092407,
+      "learning_rate": 5.0801783509020844e-05,
+      "loss": 0.0005,
+      "step": 1347
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.01030554249882698,
+      "learning_rate": 5.052048073564228e-05,
+      "loss": 0.0004,
+      "step": 1348
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.33771464228630066,
+      "learning_rate": 5.023987138507133e-05,
+      "loss": 0.0053,
+      "step": 1349
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.03562993183732033,
+      "learning_rate": 4.995995643275103e-05,
+      "loss": 0.0012,
+      "step": 1350
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.037580545991659164,
+      "learning_rate": 4.968073685171082e-05,
+      "loss": 0.0009,
+      "step": 1351
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.018147172406315804,
+      "learning_rate": 4.940221361256259e-05,
+      "loss": 0.0004,
+      "step": 1352
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.1366257667541504,
+      "learning_rate": 4.912438768349792e-05,
+      "loss": 0.0039,
+      "step": 1353
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.020096778869628906,
+      "learning_rate": 4.884726003028428e-05,
+      "loss": 0.0004,
+      "step": 1354
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.005000903271138668,
+      "learning_rate": 4.8570831616261745e-05,
+      "loss": 0.0002,
+      "step": 1355
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.02229972742497921,
+      "learning_rate": 4.829510340234e-05,
+      "loss": 0.0008,
+      "step": 1356
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.04677216708660126,
+      "learning_rate": 4.802007634699437e-05,
+      "loss": 0.0008,
+      "step": 1357
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.008060449734330177,
+      "learning_rate": 4.7745751406263163e-05,
+      "loss": 0.0002,
+      "step": 1358
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.004165045917034149,
+      "learning_rate": 4.74721295337438e-05,
+      "loss": 0.0003,
+      "step": 1359
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.2668968439102173,
+      "learning_rate": 4.719921168058977e-05,
+      "loss": 0.0055,
+      "step": 1360
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.33916279673576355,
+      "learning_rate": 4.6926998795507406e-05,
+      "loss": 0.0061,
+      "step": 1361
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.023179393261671066,
+      "learning_rate": 4.6655491824752263e-05,
+      "loss": 0.0006,
+      "step": 1362
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.01882367953658104,
+      "learning_rate": 4.6384691712126225e-05,
+      "loss": 0.0007,
+      "step": 1363
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.013556770980358124,
+      "learning_rate": 4.611459939897386e-05,
+      "loss": 0.0003,
+      "step": 1364
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.215042382478714,
+      "learning_rate": 4.5845215824179335e-05,
+      "loss": 0.0051,
+      "step": 1365
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.02435019426047802,
+      "learning_rate": 4.557654192416319e-05,
+      "loss": 0.0004,
+      "step": 1366
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.017334023490548134,
+      "learning_rate": 4.530857863287913e-05,
+      "loss": 0.0005,
+      "step": 1367
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.20091189444065094,
+      "learning_rate": 4.5041326881810395e-05,
+      "loss": 0.0069,
+      "step": 1368
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.05958046764135361,
+      "learning_rate": 4.4774787599967004e-05,
+      "loss": 0.0009,
+      "step": 1369
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.08757390081882477,
+      "learning_rate": 4.450896171388219e-05,
+      "loss": 0.0025,
+      "step": 1370
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.028023116290569305,
+      "learning_rate": 4.424385014760937e-05,
+      "loss": 0.0008,
+      "step": 1371
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.022646216675639153,
+      "learning_rate": 4.397945382271909e-05,
+      "loss": 0.0004,
+      "step": 1372
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.04051864892244339,
+      "learning_rate": 4.37157736582951e-05,
+      "loss": 0.0009,
+      "step": 1373
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7693873643875122,
+      "learning_rate": 4.3452810570932115e-05,
+      "loss": 0.0696,
+      "step": 1374
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.0007687499164603651,
+      "learning_rate": 4.3190565474731904e-05,
+      "loss": 0.0,
+      "step": 1375
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.00293734110891819,
+      "learning_rate": 4.292903928130054e-05,
+      "loss": 0.0001,
+      "step": 1376
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6718714237213135,
+      "learning_rate": 4.266823289974517e-05,
+      "loss": 0.0105,
+      "step": 1377
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.003744264366105199,
+      "learning_rate": 4.240814723667033e-05,
+      "loss": 0.0001,
+      "step": 1378
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.002130451612174511,
+      "learning_rate": 4.214878319617568e-05,
+      "loss": 0.0001,
+      "step": 1379
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.02480950951576233,
+      "learning_rate": 4.189014167985225e-05,
+      "loss": 0.0004,
+      "step": 1380
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.00624463614076376,
+      "learning_rate": 4.163222358677937e-05,
+      "loss": 0.0002,
+      "step": 1381
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.0020641738083213568,
+      "learning_rate": 4.137502981352173e-05,
+      "loss": 0.0001,
+      "step": 1382
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.14240607619285583,
+      "learning_rate": 4.111856125412608e-05,
+      "loss": 0.0036,
+      "step": 1383
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.006745543330907822,
+      "learning_rate": 4.086281880011833e-05,
+      "loss": 0.0002,
+      "step": 1384
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.015076296404004097,
+      "learning_rate": 4.060780334050032e-05,
+      "loss": 0.0005,
+      "step": 1385
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.008114277385175228,
+      "learning_rate": 4.035351576174667e-05,
+      "loss": 0.0003,
+      "step": 1386
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.014272456988692284,
+      "learning_rate": 4.0099956947801745e-05,
+      "loss": 0.0003,
+      "step": 1387
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.2593598961830139,
+      "learning_rate": 3.9847127780076626e-05,
+      "loss": 0.0062,
+      "step": 1388
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.011931393295526505,
+      "learning_rate": 3.959502913744614e-05,
+      "loss": 0.0004,
+      "step": 1389
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.37758180499076843,
+      "learning_rate": 3.934366189624561e-05,
+      "loss": 0.0124,
+      "step": 1390
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7305591702461243,
+      "learning_rate": 3.9093026930267864e-05,
+      "loss": 0.0191,
+      "step": 1391
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.3506815433502197,
+      "learning_rate": 3.8843125110760186e-05,
+      "loss": 0.004,
+      "step": 1392
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.1179872751235962,
+      "learning_rate": 3.859395730642151e-05,
+      "loss": 0.1287,
+      "step": 1393
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.07131043821573257,
+      "learning_rate": 3.8345524383398974e-05,
+      "loss": 0.0015,
+      "step": 1394
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.10362354665994644,
+      "learning_rate": 3.8097827205285374e-05,
+      "loss": 0.002,
+      "step": 1395
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.035806868225336075,
+      "learning_rate": 3.7850866633115767e-05,
+      "loss": 0.0006,
+      "step": 1396
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.09917337447404861,
+      "learning_rate": 3.760464352536469e-05,
+      "loss": 0.0013,
+      "step": 1397
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.013792157173156738,
+      "learning_rate": 3.735915873794327e-05,
+      "loss": 0.0005,
+      "step": 1398
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.11405621469020844,
+      "learning_rate": 3.711441312419589e-05,
+      "loss": 0.0017,
+      "step": 1399
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.37025147676467896,
+      "learning_rate": 3.687040753489765e-05,
+      "loss": 0.0066,
+      "step": 1400
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.005007702391594648,
+      "learning_rate": 3.662714281825111e-05,
+      "loss": 0.0001,
+      "step": 1401
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.035408273339271545,
+      "learning_rate": 3.6384619819883335e-05,
+      "loss": 0.0005,
+      "step": 1402
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.066094160079956,
+      "learning_rate": 3.614283938284332e-05,
+      "loss": 0.0229,
+      "step": 1403
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.06345271319150925,
+      "learning_rate": 3.590180234759857e-05,
+      "loss": 0.0012,
+      "step": 1404
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.13740839064121246,
+      "learning_rate": 3.566150955203251e-05,
+      "loss": 0.003,
+      "step": 1405
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.18154090642929077,
+      "learning_rate": 3.542196183144148e-05,
+      "loss": 0.0045,
+      "step": 1406
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.66633141040802,
+      "learning_rate": 3.518316001853164e-05,
+      "loss": 0.0242,
+      "step": 1407
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.054164353758096695,
+      "learning_rate": 3.494510494341657e-05,
+      "loss": 0.0022,
+      "step": 1408
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.011136863380670547,
+      "learning_rate": 3.470779743361374e-05,
+      "loss": 0.0005,
+      "step": 1409
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.011947316117584705,
+      "learning_rate": 3.447123831404228e-05,
+      "loss": 0.0005,
+      "step": 1410
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.17515721917152405,
+      "learning_rate": 3.423542840701957e-05,
+      "loss": 0.0035,
+      "step": 1411
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.014279712922871113,
+      "learning_rate": 3.4000368532258604e-05,
+      "loss": 0.0006,
+      "step": 1412
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.004492651205509901,
+      "learning_rate": 3.376605950686532e-05,
+      "loss": 0.0003,
+      "step": 1413
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.012061775662004948,
+      "learning_rate": 3.3532502145335373e-05,
+      "loss": 0.0004,
+      "step": 1414
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.019419865682721138,
+      "learning_rate": 3.3299697259551706e-05,
+      "loss": 0.0007,
+      "step": 1415
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.18258291482925415,
+      "learning_rate": 3.3067645658781425e-05,
+      "loss": 0.004,
+      "step": 1416
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.2847168445587158,
+      "learning_rate": 3.283634814967307e-05,
+      "loss": 0.0092,
+      "step": 1417
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.4876096248626709,
+      "learning_rate": 3.260580553625389e-05,
+      "loss": 0.0131,
+      "step": 1418
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.6045449376106262,
+      "learning_rate": 3.237601861992709e-05,
+      "loss": 0.0193,
+      "step": 1419
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.1586480587720871,
+      "learning_rate": 3.214698819946879e-05,
+      "loss": 0.0026,
+      "step": 1420
+    },
+    {
+      "epoch": 2.51,
+      "eval_loss": 0.12486789375543594,
+      "eval_runtime": 14.7046,
+      "eval_samples_per_second": 32.439,
+      "eval_steps_per_second": 8.161,
+      "step": 1420
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.12747687101364136,
+      "learning_rate": 3.191871507102545e-05,
+      "loss": 0.0017,
+      "step": 1421
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.04522531479597092,
+      "learning_rate": 3.1691200028111044e-05,
+      "loss": 0.0009,
+      "step": 1422
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.2810252904891968,
+      "learning_rate": 3.146444386160441e-05,
+      "loss": 0.0041,
+      "step": 1423
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.06912285834550858,
+      "learning_rate": 3.123844735974646e-05,
+      "loss": 0.0008,
+      "step": 1424
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.3842872679233551,
+      "learning_rate": 3.1013211308137054e-05,
+      "loss": 0.0075,
+      "step": 1425
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.018251223489642143,
+      "learning_rate": 3.078873648973304e-05,
+      "loss": 0.0003,
+      "step": 1426
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.005700815003365278,
+      "learning_rate": 3.0565023684844765e-05,
+      "loss": 0.0002,
+      "step": 1427
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.1778111457824707,
+      "learning_rate": 3.034207367113387e-05,
+      "loss": 0.0241,
+      "step": 1428
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.5692623853683472,
+      "learning_rate": 3.0119887223610475e-05,
+      "loss": 0.0074,
+      "step": 1429
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.007092227227985859,
+      "learning_rate": 2.9898465114630123e-05,
+      "loss": 0.0003,
+      "step": 1430
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.006958132144063711,
+      "learning_rate": 2.9677808113891675e-05,
+      "loss": 0.0003,
+      "step": 1431
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.2849915027618408,
+      "learning_rate": 2.945791698843431e-05,
+      "loss": 0.0114,
+      "step": 1432
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.3223225176334381,
+      "learning_rate": 2.9238792502634782e-05,
+      "loss": 0.0105,
+      "step": 1433
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.3097008466720581,
+      "learning_rate": 2.902043541820501e-05,
+      "loss": 0.0067,
+      "step": 1434
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.08397488296031952,
+      "learning_rate": 2.880284649418913e-05,
+      "loss": 0.0029,
+      "step": 1435
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.029154475778341293,
+      "learning_rate": 2.8586026486961235e-05,
+      "loss": 0.0008,
+      "step": 1436
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.02804793231189251,
+      "learning_rate": 2.836997615022249e-05,
+      "loss": 0.0007,
+      "step": 1437
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.008161618374288082,
+      "learning_rate": 2.8154696234998472e-05,
+      "loss": 0.0003,
+      "step": 1438
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.055503591895103455,
+      "learning_rate": 2.7940187489636697e-05,
+      "loss": 0.0015,
+      "step": 1439
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.02855825237929821,
+      "learning_rate": 2.7726450659803947e-05,
+      "loss": 0.0008,
+      "step": 1440
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.01375576015561819,
+      "learning_rate": 2.751348648848373e-05,
+      "loss": 0.0004,
+      "step": 1441
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.0038817201275378466,
+      "learning_rate": 2.7301295715973757e-05,
+      "loss": 0.0001,
+      "step": 1442
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.023767409846186638,
+      "learning_rate": 2.708987907988314e-05,
+      "loss": 0.0007,
+      "step": 1443
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.06856090575456619,
+      "learning_rate": 2.687923731512995e-05,
+      "loss": 0.0016,
+      "step": 1444
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.001726289396174252,
+      "learning_rate": 2.666937115393886e-05,
+      "loss": 0.0001,
+      "step": 1445
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.011897512711584568,
+      "learning_rate": 2.646028132583822e-05,
+      "loss": 0.0003,
+      "step": 1446
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.011256224475800991,
+      "learning_rate": 2.6251968557657908e-05,
+      "loss": 0.0004,
+      "step": 1447
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.09208813309669495,
+      "learning_rate": 2.6044433573526454e-05,
+      "loss": 0.0021,
+      "step": 1448
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.028126435354351997,
+      "learning_rate": 2.58376770948687e-05,
+      "loss": 0.0003,
+      "step": 1449
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.02156115137040615,
+      "learning_rate": 2.5631699840403476e-05,
+      "loss": 0.0007,
+      "step": 1450
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.039257489144802094,
+      "learning_rate": 2.542650252614062e-05,
+      "loss": 0.0013,
+      "step": 1451
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.2106196880340576,
+      "learning_rate": 2.5222085865379023e-05,
+      "loss": 0.0606,
+      "step": 1452
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.011918625794351101,
+      "learning_rate": 2.501845056870375e-05,
+      "loss": 0.0002,
+      "step": 1453
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.35332852602005005,
+      "learning_rate": 2.4815597343983697e-05,
+      "loss": 0.0813,
+      "step": 1454
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.0004311532247811556,
+      "learning_rate": 2.4613526896369308e-05,
+      "loss": 0.0,
+      "step": 1455
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.20872049033641815,
+      "learning_rate": 2.441223992828978e-05,
+      "loss": 0.0047,
+      "step": 1456
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.3243696689605713,
+      "learning_rate": 2.421173713945099e-05,
+      "loss": 0.024,
+      "step": 1457
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.005421569105237722,
+      "learning_rate": 2.4012019226832772e-05,
+      "loss": 0.0002,
+      "step": 1458
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.3085567355155945,
+      "learning_rate": 2.381308688468656e-05,
+      "loss": 0.0047,
+      "step": 1459
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.002236233791336417,
+      "learning_rate": 2.361494080453319e-05,
+      "loss": 0.0001,
+      "step": 1460
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.5045457482337952,
+      "learning_rate": 2.3417581675160088e-05,
+      "loss": 0.0054,
+      "step": 1461
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.0775829628109932,
+      "learning_rate": 2.3221010182619406e-05,
+      "loss": 0.0014,
+      "step": 1462
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.47293341159820557,
+      "learning_rate": 2.302522701022511e-05,
+      "loss": 0.0563,
+      "step": 1463
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.30130696296691895,
+      "learning_rate": 2.2830232838550845e-05,
+      "loss": 0.0057,
+      "step": 1464
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.0052946461364626884,
+      "learning_rate": 2.2636028345427745e-05,
+      "loss": 0.0002,
+      "step": 1465
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.017164453864097595,
+      "learning_rate": 2.244261420594168e-05,
+      "loss": 0.0005,
+      "step": 1466
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.8137333393096924,
+      "learning_rate": 2.224999109243131e-05,
+      "loss": 0.039,
+      "step": 1467
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.19935303926467896,
+      "learning_rate": 2.205815967448546e-05,
+      "loss": 0.0056,
+      "step": 1468
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.3961120545864105,
+      "learning_rate": 2.1867120618940833e-05,
+      "loss": 0.0056,
+      "step": 1469
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.019902199506759644,
+      "learning_rate": 2.167687458987991e-05,
+      "loss": 0.0007,
+      "step": 1470
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.009017308242619038,
+      "learning_rate": 2.1487422248628457e-05,
+      "loss": 0.0004,
+      "step": 1471
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.001593849272467196,
+      "learning_rate": 2.1298764253753044e-05,
+      "loss": 0.0001,
+      "step": 1472
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.023191401734948158,
+      "learning_rate": 2.1110901261059245e-05,
+      "loss": 0.0006,
+      "step": 1473
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.03758575767278671,
+      "learning_rate": 2.0923833923588874e-05,
+      "loss": 0.0011,
+      "step": 1474
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.3719123303890228,
+      "learning_rate": 2.0737562891618074e-05,
+      "loss": 0.0061,
+      "step": 1475
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.004338722676038742,
+      "learning_rate": 2.0552088812654885e-05,
+      "loss": 0.0002,
+      "step": 1476
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.004211151506751776,
+      "learning_rate": 2.0367412331436796e-05,
+      "loss": 0.0002,
+      "step": 1477
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.008191877976059914,
+      "learning_rate": 2.018353408992901e-05,
+      "loss": 0.0003,
+      "step": 1478
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.4456917941570282,
+      "learning_rate": 2.0000454727321703e-05,
+      "loss": 0.0153,
+      "step": 1479
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.03597655147314072,
+      "learning_rate": 1.9818174880028213e-05,
+      "loss": 0.0013,
+      "step": 1480
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.0015153115382418036,
+      "learning_rate": 1.963669518168243e-05,
+      "loss": 0.0001,
+      "step": 1481
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.008952159434556961,
+      "learning_rate": 1.9456016263136923e-05,
+      "loss": 0.0003,
+      "step": 1482
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.00541423074901104,
+      "learning_rate": 1.927613875246059e-05,
+      "loss": 0.0002,
+      "step": 1483
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.09652198851108551,
+      "learning_rate": 1.909706327493657e-05,
+      "loss": 0.0016,
+      "step": 1484
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.22741255164146423,
+      "learning_rate": 1.8918790453059914e-05,
+      "loss": 0.0031,
+      "step": 1485
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.040858399122953415,
+      "learning_rate": 1.8741320906535546e-05,
+      "loss": 0.0007,
+      "step": 1486
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.053088657557964325,
+      "learning_rate": 1.856465525227602e-05,
+      "loss": 0.0019,
+      "step": 1487
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.008057578466832638,
+      "learning_rate": 1.8388794104399558e-05,
+      "loss": 0.0003,
+      "step": 1488
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.009569887071847916,
+      "learning_rate": 1.8213738074227742e-05,
+      "loss": 0.0004,
+      "step": 1489
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.00848764181137085,
+      "learning_rate": 1.803948777028336e-05,
+      "loss": 0.0003,
+      "step": 1490
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.6634844541549683,
+      "learning_rate": 1.7866043798288433e-05,
+      "loss": 0.0438,
+      "step": 1491
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.15894575417041779,
+      "learning_rate": 1.7693406761162016e-05,
+      "loss": 0.0055,
+      "step": 1492
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.013144123367965221,
+      "learning_rate": 1.752157725901815e-05,
+      "loss": 0.0005,
+      "step": 1493
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.018960585817694664,
+      "learning_rate": 1.735055588916379e-05,
+      "loss": 0.0003,
+      "step": 1494
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.016689004376530647,
+      "learning_rate": 1.718034324609663e-05,
+      "loss": 0.0006,
+      "step": 1495
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.006429940462112427,
+      "learning_rate": 1.701093992150307e-05,
+      "loss": 0.0003,
+      "step": 1496
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.11746788769960403,
+      "learning_rate": 1.684234650425631e-05,
+      "loss": 0.0032,
+      "step": 1497
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.004738051909953356,
+      "learning_rate": 1.6674563580414053e-05,
+      "loss": 0.0002,
+      "step": 1498
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.004201785195618868,
+      "learning_rate": 1.65075917332167e-05,
+      "loss": 0.0002,
+      "step": 1499
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.07351760566234589,
+      "learning_rate": 1.6341431543085207e-05,
+      "loss": 0.0017,
+      "step": 1500
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.03148205578327179,
+      "learning_rate": 1.6176083587618935e-05,
+      "loss": 0.0011,
+      "step": 1501
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.07184291630983353,
+      "learning_rate": 1.6011548441594e-05,
+      "loss": 0.0018,
+      "step": 1502
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.01941969059407711,
+      "learning_rate": 1.5847826676960914e-05,
+      "loss": 0.0007,
+      "step": 1503
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.004026324022561312,
+      "learning_rate": 1.5684918862842846e-05,
+      "loss": 0.0001,
+      "step": 1504
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.002735046437010169,
+      "learning_rate": 1.5522825565533445e-05,
+      "loss": 0.0002,
+      "step": 1505
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.09817846864461899,
+      "learning_rate": 1.5361547348495013e-05,
+      "loss": 0.0016,
+      "step": 1506
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.022421833127737045,
+      "learning_rate": 1.5201084772356544e-05,
+      "loss": 0.0004,
+      "step": 1507
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.1519351303577423,
+      "learning_rate": 1.5041438394911622e-05,
+      "loss": 0.0019,
+      "step": 1508
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.1200493648648262,
+      "learning_rate": 1.4882608771116719e-05,
+      "loss": 0.002,
+      "step": 1509
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.002202227944508195,
+      "learning_rate": 1.4724596453089101e-05,
+      "loss": 0.0001,
+      "step": 1510
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.030175110325217247,
+      "learning_rate": 1.4567401990104867e-05,
+      "loss": 0.0009,
+      "step": 1511
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.03393395245075226,
+      "learning_rate": 1.441102592859725e-05,
+      "loss": 0.0014,
+      "step": 1512
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7548046112060547,
+      "learning_rate": 1.4255468812154477e-05,
+      "loss": 0.0468,
+      "step": 1513
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.015761546790599823,
+      "learning_rate": 1.4100731181518056e-05,
+      "loss": 0.0006,
+      "step": 1514
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.014691539108753204,
+      "learning_rate": 1.3946813574580858e-05,
+      "loss": 0.0006,
+      "step": 1515
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.06156891956925392,
+      "learning_rate": 1.3793716526385058e-05,
+      "loss": 0.0015,
+      "step": 1516
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.011981514282524586,
+      "learning_rate": 1.36414405691207e-05,
+      "loss": 0.0002,
+      "step": 1517
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.04529397189617157,
+      "learning_rate": 1.3489986232123302e-05,
+      "loss": 0.0017,
+      "step": 1518
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.0034320768900215626,
+      "learning_rate": 1.333935404187253e-05,
+      "loss": 0.0002,
+      "step": 1519
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.5339131951332092,
+      "learning_rate": 1.3189544521990032e-05,
+      "loss": 0.031,
+      "step": 1520
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.21039195358753204,
+      "learning_rate": 1.3040558193237657e-05,
+      "loss": 0.002,
+      "step": 1521
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.0334392786026001,
+      "learning_rate": 1.2892395573515819e-05,
+      "loss": 0.0014,
+      "step": 1522
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.009169626981019974,
+      "learning_rate": 1.2745057177861647e-05,
+      "loss": 0.0003,
+      "step": 1523
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.006436166353523731,
+      "learning_rate": 1.2598543518446887e-05,
+      "loss": 0.0004,
+      "step": 1524
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.004530397243797779,
+      "learning_rate": 1.245285510457661e-05,
+      "loss": 0.0002,
+      "step": 1525
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.013779646717011929,
+      "learning_rate": 1.2307992442687072e-05,
+      "loss": 0.0005,
+      "step": 1526
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.0018188911490142345,
+      "learning_rate": 1.2163956036344153e-05,
+      "loss": 0.0001,
+      "step": 1527
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.15664638578891754,
+      "learning_rate": 1.2020746386241565e-05,
+      "loss": 0.0029,
+      "step": 1528
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.017927415668964386,
+      "learning_rate": 1.1878363990198871e-05,
+      "loss": 0.0004,
+      "step": 1529
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.17458246648311615,
+      "learning_rate": 1.1736809343160237e-05,
+      "loss": 0.0035,
+      "step": 1530
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.016326846554875374,
+      "learning_rate": 1.1596082937192276e-05,
+      "loss": 0.0006,
+      "step": 1531
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.75627201795578,
+      "learning_rate": 1.1456185261482565e-05,
+      "loss": 0.0268,
+      "step": 1532
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.003338505746796727,
+      "learning_rate": 1.1317116802337906e-05,
+      "loss": 0.0002,
+      "step": 1533
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.003949652425944805,
+      "learning_rate": 1.1178878043182462e-05,
+      "loss": 0.0002,
+      "step": 1534
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.005014631897211075,
+      "learning_rate": 1.1041469464556419e-05,
+      "loss": 0.0003,
+      "step": 1535
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.09127765148878098,
+      "learning_rate": 1.090489154411406e-05,
+      "loss": 0.0022,
+      "step": 1536
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.002308650640770793,
+      "learning_rate": 1.0769144756622106e-05,
+      "loss": 0.0001,
+      "step": 1537
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.01227644830942154,
+      "learning_rate": 1.0634229573958155e-05,
+      "loss": 0.0005,
+      "step": 1538
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.07127789407968521,
+      "learning_rate": 1.0500146465108995e-05,
+      "loss": 0.0016,
+      "step": 1539
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.3965778648853302,
+      "learning_rate": 1.0366895896169098e-05,
+      "loss": 0.0119,
+      "step": 1540
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.006936135236173868,
+      "learning_rate": 1.0234478330338775e-05,
+      "loss": 0.0003,
+      "step": 1541
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.1017901748418808,
+      "learning_rate": 1.0102894227922737e-05,
+      "loss": 0.0019,
+      "step": 1542
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.004074485041201115,
+      "learning_rate": 9.972144046328429e-06,
+      "loss": 0.0002,
+      "step": 1543
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.5157943367958069,
+      "learning_rate": 9.842228240064421e-06,
+      "loss": 0.0767,
+      "step": 1544
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.35279086232185364,
+      "learning_rate": 9.713147260738936e-06,
+      "loss": 0.0063,
+      "step": 1545
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.005003898870199919,
+      "learning_rate": 9.584901557058156e-06,
+      "loss": 0.0002,
+      "step": 1546
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.001333480584435165,
+      "learning_rate": 9.457491574824757e-06,
+      "loss": 0.0001,
+      "step": 1547
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.02091939002275467,
+      "learning_rate": 9.330917756936174e-06,
+      "loss": 0.0009,
+      "step": 1548
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.30123287439346313,
+      "learning_rate": 9.2051805433834e-06,
+      "loss": 0.0046,
+      "step": 1549
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.17032013833522797,
+      "learning_rate": 9.080280371249112e-06,
+      "loss": 0.0036,
+      "step": 1550
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.4087766706943512,
+      "learning_rate": 8.956217674706363e-06,
+      "loss": 0.0049,
+      "step": 1551
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.09622213989496231,
+      "learning_rate": 8.832992885016988e-06,
+      "loss": 0.0024,
+      "step": 1552
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.016199596226215363,
+      "learning_rate": 8.710606430530066e-06,
+      "loss": 0.0005,
+      "step": 1553
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.0021451227366924286,
+      "learning_rate": 8.589058736680643e-06,
+      "loss": 0.0001,
+      "step": 1554
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.005553426221013069,
+      "learning_rate": 8.46835022598791e-06,
+      "loss": 0.0002,
+      "step": 1555
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.5673149824142456,
+      "learning_rate": 8.348481318054075e-06,
+      "loss": 0.0614,
+      "step": 1556
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.014100473374128342,
+      "learning_rate": 8.229452429562661e-06,
+      "loss": 0.0004,
+      "step": 1557
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.026370083913207054,
+      "learning_rate": 8.111263974277166e-06,
+      "loss": 0.0006,
+      "step": 1558
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.00458921492099762,
+      "learning_rate": 7.993916363039672e-06,
+      "loss": 0.0002,
+      "step": 1559
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.0019055847078561783,
+      "learning_rate": 7.877410003769236e-06,
+      "loss": 0.0001,
+      "step": 1560
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.010362006723880768,
+      "learning_rate": 7.761745301460676e-06,
+      "loss": 0.0004,
+      "step": 1561
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.06622084230184555,
+      "learning_rate": 7.646922658183092e-06,
+      "loss": 0.0016,
+      "step": 1562
+    },
+    {
+      "epoch": 2.76,
+      "eval_loss": 0.12071493268013,
+      "eval_runtime": 14.7006,
+      "eval_samples_per_second": 32.448,
+      "eval_steps_per_second": 8.163,
+      "step": 1562
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.21637581288814545,
+      "learning_rate": 7.532942473078341e-06,
+      "loss": 0.0077,
+      "step": 1563
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.2466088980436325,
+      "learning_rate": 7.419805142359875e-06,
+      "loss": 0.005,
+      "step": 1564
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.1368948519229889,
+      "learning_rate": 7.307511059311184e-06,
+      "loss": 0.0024,
+      "step": 1565
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.4194111227989197,
+      "learning_rate": 7.196060614284544e-06,
+      "loss": 0.0097,
+      "step": 1566
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.0024943724274635315,
+      "learning_rate": 7.085454194699553e-06,
+      "loss": 0.0001,
+      "step": 1567
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.5519493222236633,
+      "learning_rate": 6.975692185041848e-06,
+      "loss": 0.0128,
+      "step": 1568
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.17635726928710938,
+      "learning_rate": 6.866774966861833e-06,
+      "loss": 0.0047,
+      "step": 1569
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.17873595654964447,
+      "learning_rate": 6.758702918773202e-06,
+      "loss": 0.0041,
+      "step": 1570
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.008024133741855621,
+      "learning_rate": 6.651476416451696e-06,
+      "loss": 0.0004,
+      "step": 1571
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.018708229064941406,
+      "learning_rate": 6.545095832633907e-06,
+      "loss": 0.0008,
+      "step": 1572
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.3967719078063965,
+      "learning_rate": 6.439561537115751e-06,
+      "loss": 0.0771,
+      "step": 1573
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.039222270250320435,
+      "learning_rate": 6.334873896751414e-06,
+      "loss": 0.001,
+      "step": 1574
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.013372889719903469,
+      "learning_rate": 6.231033275451908e-06,
+      "loss": 0.0007,
+      "step": 1575
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.008132797665894032,
+      "learning_rate": 6.12804003418388e-06,
+      "loss": 0.0003,
+      "step": 1576
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.15415289998054504,
+      "learning_rate": 6.0258945309683565e-06,
+      "loss": 0.0044,
+      "step": 1577
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.5614458918571472,
+      "learning_rate": 5.9245971208795045e-06,
+      "loss": 0.0128,
+      "step": 1578
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.0058565386570990086,
+      "learning_rate": 5.824148156043374e-06,
+      "loss": 0.0002,
+      "step": 1579
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.6066858172416687,
+      "learning_rate": 5.724547985636652e-06,
+      "loss": 0.0365,
+      "step": 1580
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.39940959215164185,
+      "learning_rate": 5.625796955885526e-06,
+      "loss": 0.0168,
+      "step": 1581
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.10061442852020264,
+      "learning_rate": 5.527895410064459e-06,
+      "loss": 0.0013,
+      "step": 1582
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.003030581632629037,
+      "learning_rate": 5.430843688494836e-06,
+      "loss": 0.0001,
+      "step": 1583
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.11822108179330826,
+      "learning_rate": 5.3346421285440925e-06,
+      "loss": 0.0033,
+      "step": 1584
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.004489241633564234,
+      "learning_rate": 5.239291064624258e-06,
+      "loss": 0.0003,
+      "step": 1585
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.0766096115112305,
+      "learning_rate": 5.144790828190887e-06,
+      "loss": 0.036,
+      "step": 1586
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.09894857555627823,
+      "learning_rate": 5.051141747741989e-06,
+      "loss": 0.0018,
+      "step": 1587
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.49432384967803955,
+      "learning_rate": 4.958344148816824e-06,
+      "loss": 0.0061,
+      "step": 1588
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.07527617365121841,
+      "learning_rate": 4.8663983539946885e-06,
+      "loss": 0.0017,
+      "step": 1589
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.0023846172261983156,
+      "learning_rate": 4.775304682893944e-06,
+      "loss": 0.0001,
+      "step": 1590
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.01218903437256813,
+      "learning_rate": 4.685063452170735e-06,
+      "loss": 0.0006,
+      "step": 1591
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.029260369017720222,
+      "learning_rate": 4.595674975518133e-06,
+      "loss": 0.0013,
+      "step": 1592
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.00425742520019412,
+      "learning_rate": 4.507139563664802e-06,
+      "loss": 0.0002,
+      "step": 1593
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.003876454196870327,
+      "learning_rate": 4.419457524374032e-06,
+      "loss": 0.0002,
+      "step": 1594
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.1555083692073822,
+      "learning_rate": 4.332629162442675e-06,
+      "loss": 0.0028,
+      "step": 1595
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.0016881643095985055,
+      "learning_rate": 4.246654779699988e-06,
+      "loss": 0.0001,
+      "step": 1596
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.00919495802372694,
+      "learning_rate": 4.161534675006739e-06,
+      "loss": 0.0005,
+      "step": 1597
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.008377696387469769,
+      "learning_rate": 4.077269144254103e-06,
+      "loss": 0.0003,
+      "step": 1598
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.2848001718521118,
+      "learning_rate": 3.993858480362572e-06,
+      "loss": 0.0519,
+      "step": 1599
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.07129717618227005,
+      "learning_rate": 3.9113029732809615e-06,
+      "loss": 0.0011,
+      "step": 1600
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.001333805383183062,
+      "learning_rate": 3.8296029099854635e-06,
+      "loss": 0.0001,
+      "step": 1601
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.017828090116381645,
+      "learning_rate": 3.748758574478622e-06,
+      "loss": 0.0008,
+      "step": 1602
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.010987207293510437,
+      "learning_rate": 3.6687702477883332e-06,
+      "loss": 0.0004,
+      "step": 1603
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.006206910125911236,
+      "learning_rate": 3.5896382079668166e-06,
+      "loss": 0.0003,
+      "step": 1604
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.001939537120051682,
+      "learning_rate": 3.511362730089729e-06,
+      "loss": 0.0001,
+      "step": 1605
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.2372162640094757,
+      "learning_rate": 3.4339440862552194e-06,
+      "loss": 0.0031,
+      "step": 1606
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.0658475011587143,
+      "learning_rate": 3.3573825455829043e-06,
+      "loss": 0.002,
+      "step": 1607
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.012416253797709942,
+      "learning_rate": 3.2816783742129762e-06,
+      "loss": 0.0005,
+      "step": 1608
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.001760126673616469,
+      "learning_rate": 3.206831835305263e-06,
+      "loss": 0.0001,
+      "step": 1609
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.1820244938135147,
+      "learning_rate": 3.132843189038365e-06,
+      "loss": 0.0035,
+      "step": 1610
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.2944984436035156,
+      "learning_rate": 3.059712692608657e-06,
+      "loss": 0.0086,
+      "step": 1611
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.013544095680117607,
+      "learning_rate": 2.9874406002295128e-06,
+      "loss": 0.0006,
+      "step": 1612
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.008425934240221977,
+      "learning_rate": 2.9160271631303025e-06,
+      "loss": 0.0003,
+      "step": 1613
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.0034127351827919483,
+      "learning_rate": 2.84547262955559e-06,
+      "loss": 0.0001,
+      "step": 1614
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.03862582519650459,
+      "learning_rate": 2.775777244764216e-06,
+      "loss": 0.0007,
+      "step": 1615
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.9643748998641968,
+      "learning_rate": 2.7069412510285773e-06,
+      "loss": 0.0381,
+      "step": 1616
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.027254153043031693,
+      "learning_rate": 2.6389648876335716e-06,
+      "loss": 0.0007,
+      "step": 1617
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.14486004412174225,
+      "learning_rate": 2.571848390875986e-06,
+      "loss": 0.0039,
+      "step": 1618
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.002894668374210596,
+      "learning_rate": 2.5055919940635276e-06,
+      "loss": 0.0001,
+      "step": 1619
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.8268805742263794,
+      "learning_rate": 2.440195927514044e-06,
+      "loss": 0.0447,
+      "step": 1620
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.17663922905921936,
+      "learning_rate": 2.375660418554776e-06,
+      "loss": 0.0076,
+      "step": 1621
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.12479265034198761,
+      "learning_rate": 2.3119856915214677e-06,
+      "loss": 0.002,
+      "step": 1622
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.14911742508411407,
+      "learning_rate": 2.249171967757674e-06,
+      "loss": 0.0046,
+      "step": 1623
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.0442122220993042,
+      "learning_rate": 2.1872194656140377e-06,
+      "loss": 0.0154,
+      "step": 1624
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.018565386533737183,
+      "learning_rate": 2.126128400447347e-06,
+      "loss": 0.0007,
+      "step": 1625
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7466828227043152,
+      "learning_rate": 2.0658989846199516e-06,
+      "loss": 0.0113,
+      "step": 1626
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.034939322620630264,
+      "learning_rate": 2.006531427499014e-06,
+      "loss": 0.0008,
+      "step": 1627
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.028358131647109985,
+      "learning_rate": 1.948025935455594e-06,
+      "loss": 0.001,
+      "step": 1628
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.043177518993616104,
+      "learning_rate": 1.8903827118642303e-06,
+      "loss": 0.0008,
+      "step": 1629
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.03869365155696869,
+      "learning_rate": 1.8336019571019712e-06,
+      "loss": 0.0022,
+      "step": 1630
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.0050878459587693214,
+      "learning_rate": 1.7776838685478748e-06,
+      "loss": 0.0002,
+      "step": 1631
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.11395423114299774,
+      "learning_rate": 1.7226286405821746e-06,
+      "loss": 0.0024,
+      "step": 1632
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.5369199514389038,
+      "learning_rate": 1.6684364645856431e-06,
+      "loss": 0.0145,
+      "step": 1633
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.012085442431271076,
+      "learning_rate": 1.6151075289390082e-06,
+      "loss": 0.0003,
+      "step": 1634
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.15687036514282227,
+      "learning_rate": 1.5626420190222034e-06,
+      "loss": 0.0059,
+      "step": 1635
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.02339506521821022,
+      "learning_rate": 1.5110401172137578e-06,
+      "loss": 0.0004,
+      "step": 1636
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.15162977576255798,
+      "learning_rate": 1.4603020028901292e-06,
+      "loss": 0.0026,
+      "step": 1637
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.0018832300556823611,
+      "learning_rate": 1.4104278524251778e-06,
+      "loss": 0.0001,
+      "step": 1638
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.030508514493703842,
+      "learning_rate": 1.3614178391894438e-06,
+      "loss": 0.0013,
+      "step": 1639
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.00550305750221014,
+      "learning_rate": 1.3132721335495645e-06,
+      "loss": 0.0002,
+      "step": 1640
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.01754370890557766,
+      "learning_rate": 1.265990902867803e-06,
+      "loss": 0.0005,
+      "step": 1641
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.00891774520277977,
+      "learning_rate": 1.2195743115012148e-06,
+      "loss": 0.0002,
+      "step": 1642
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.02182842791080475,
+      "learning_rate": 1.1740225208013712e-06,
+      "loss": 0.0007,
+      "step": 1643
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.0056608193553984165,
+      "learning_rate": 1.1293356891136086e-06,
+      "loss": 0.0002,
+      "step": 1644
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.004720357712358236,
+      "learning_rate": 1.0855139717765028e-06,
+      "loss": 0.0003,
+      "step": 1645
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.005244838539510965,
+      "learning_rate": 1.0425575211213956e-06,
+      "loss": 0.0003,
+      "step": 1646
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.0327480174601078,
+      "learning_rate": 1.0004664864717573e-06,
+      "loss": 0.001,
+      "step": 1647
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.8325297832489014,
+      "learning_rate": 9.592410141427977e-07,
+      "loss": 0.0526,
+      "step": 1648
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.0062986682169139385,
+      "learning_rate": 9.188812474408837e-07,
+      "loss": 0.0003,
+      "step": 1649
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.01729346066713333,
+      "learning_rate": 8.793873266630393e-07,
+      "loss": 0.0006,
+      "step": 1650
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.00606880709528923,
+      "learning_rate": 8.407593890964461e-07,
+      "loss": 0.0004,
+      "step": 1651
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.009202031418681145,
+      "learning_rate": 8.029975690179992e-07,
+      "loss": 0.0005,
+      "step": 1652
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.14149095118045807,
+      "learning_rate": 7.661019976939187e-07,
+      "loss": 0.0014,
+      "step": 1653
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.06853083521127701,
+      "learning_rate": 7.300728033790282e-07,
+      "loss": 0.0026,
+      "step": 1654
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.5628806352615356,
+      "learning_rate": 6.949101113166712e-07,
+      "loss": 0.0322,
+      "step": 1655
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.004697425756603479,
+      "learning_rate": 6.606140437379616e-07,
+      "loss": 0.0002,
+      "step": 1656
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.026971176266670227,
+      "learning_rate": 6.271847198615343e-07,
+      "loss": 0.0008,
+      "step": 1657
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.416694313287735,
+      "learning_rate": 5.946222558931014e-07,
+      "loss": 0.0072,
+      "step": 1658
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.16609083116054535,
+      "learning_rate": 5.629267650249792e-07,
+      "loss": 0.0056,
+      "step": 1659
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.020159004256129265,
+      "learning_rate": 5.320983574358118e-07,
+      "loss": 0.0006,
+      "step": 1660
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.020069165155291557,
+      "learning_rate": 5.02137140290071e-07,
+      "loss": 0.0007,
+      "step": 1661
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.04708680510520935,
+      "learning_rate": 4.7304321773775085e-07,
+      "loss": 0.0024,
+      "step": 1662
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.03632630407810211,
+      "learning_rate": 4.4481669091400725e-07,
+      "loss": 0.0015,
+      "step": 1663
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.006391770206391811,
+      "learning_rate": 4.1745765793874126e-07,
+      "loss": 0.0004,
+      "step": 1664
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8153448700904846,
+      "learning_rate": 3.90966213916405e-07,
+      "loss": 0.0167,
+      "step": 1665
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.08031131327152252,
+      "learning_rate": 3.6534245093544636e-07,
+      "loss": 0.0036,
+      "step": 1666
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.015597201883792877,
+      "learning_rate": 3.405864580681983e-07,
+      "loss": 0.0006,
+      "step": 1667
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.024990934878587723,
+      "learning_rate": 3.1669832137046216e-07,
+      "loss": 0.0008,
+      "step": 1668
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.005498305428773165,
+      "learning_rate": 2.9367812388123025e-07,
+      "loss": 0.0002,
+      "step": 1669
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.19190239906311035,
+      "learning_rate": 2.715259456224084e-07,
+      "loss": 0.0051,
+      "step": 1670
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.008078343234956264,
+      "learning_rate": 2.502418635985382e-07,
+      "loss": 0.0004,
+      "step": 1671
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.02244127169251442,
+      "learning_rate": 2.2982595179646404e-07,
+      "loss": 0.0009,
+      "step": 1672
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.027221323922276497,
+      "learning_rate": 2.1027828118519442e-07,
+      "loss": 0.001,
+      "step": 1673
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.05436151847243309,
+      "learning_rate": 1.91598919715541e-07,
+      "loss": 0.0012,
+      "step": 1674
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.751646876335144,
+      "learning_rate": 1.7378793232000755e-07,
+      "loss": 0.0248,
+      "step": 1675
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.04363579303026199,
+      "learning_rate": 1.5684538091240153e-07,
+      "loss": 0.0015,
+      "step": 1676
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.02322092093527317,
+      "learning_rate": 1.407713243877784e-07,
+      "loss": 0.001,
+      "step": 1677
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.023351455107331276,
+      "learning_rate": 1.2556581862213646e-07,
+      "loss": 0.001,
+      "step": 1678
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.02784132957458496,
+      "learning_rate": 1.1122891647222244e-07,
+      "loss": 0.0007,
+      "step": 1679
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.018551329150795937,
+      "learning_rate": 9.776066777542058e-08,
+      "loss": 0.0008,
+      "step": 1680
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.0073452303186059,
+      "learning_rate": 8.516111934955828e-08,
+      "loss": 0.0004,
+      "step": 1681
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.022210106253623962,
+      "learning_rate": 7.343031499262853e-08,
+      "loss": 0.0005,
+      "step": 1682
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.36818450689315796,
+      "learning_rate": 6.256829548284549e-08,
+      "loss": 0.0045,
+      "step": 1683
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.012365012429654598,
+      "learning_rate": 5.257509857828358e-08,
+      "loss": 0.0005,
+      "step": 1684
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.007756817154586315,
+      "learning_rate": 4.345075901693307e-08,
+      "loss": 0.0003,
+      "step": 1685
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.31708094477653503,
+      "learning_rate": 3.51953085164225e-08,
+      "loss": 0.01,
+      "step": 1686
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.026957141235470772,
+      "learning_rate": 2.7808775774074147e-08,
+      "loss": 0.0005,
+      "step": 1687
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7157958149909973,
+      "learning_rate": 2.1291186466626535e-08,
+      "loss": 0.0264,
+      "step": 1688
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.003837962169200182,
+      "learning_rate": 1.5642563250289897e-08,
+      "loss": 0.0002,
+      "step": 1689
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.22489942610263824,
+      "learning_rate": 1.0862925760551923e-08,
+      "loss": 0.0026,
+      "step": 1690
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.4433555603027344,
+      "learning_rate": 6.952290612205481e-09,
+      "loss": 0.037,
+      "step": 1691
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.0032361983321607113,
+      "learning_rate": 3.910671399265376e-09,
+      "loss": 0.0002,
+      "step": 1692
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.13408610224723816,
+      "learning_rate": 1.7380786948850702e-09,
+      "loss": 0.0032,
+      "step": 1693
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.017384354025125504,
+      "learning_rate": 4.345200513289349e-10,
+      "loss": 0.0004,
+      "step": 1694
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0010622803820297122,
+      "learning_rate": 0.0,
+      "loss": 0.0001,
+      "step": 1695
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1695,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 565,
+  "total_flos": 1.549439947809751e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}