diff --git "a/checkpoint-1800/trainer_state.json" "b/checkpoint-1800/trainer_state.json"
deleted file mode 100644--- "a/checkpoint-1800/trainer_state.json"
+++ /dev/null
@@ -1,14514 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 0.12608684231264283,
-  "eval_steps": 200,
-  "global_step": 1800,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 7.004824572924602e-05,
-      "grad_norm": 6.222772121429443,
-      "learning_rate": 9.99930017513135e-05,
-      "loss": 1.1076,
-      "num_input_tokens_seen": 16384,
-      "step": 1
-    },
-    {
-      "epoch": 0.00014009649145849205,
-      "grad_norm": 6.042057037353516,
-      "learning_rate": 9.998600350262697e-05,
-      "loss": 1.1086,
-      "num_input_tokens_seen": 32768,
-      "step": 2
-    },
-    {
-      "epoch": 0.00021014473718773804,
-      "grad_norm": 7.119229316711426,
-      "learning_rate": 9.997900525394046e-05,
-      "loss": 1.4047,
-      "num_input_tokens_seen": 49152,
-      "step": 3
-    },
-    {
-      "epoch": 0.0002801929829169841,
-      "grad_norm": 7.133191108703613,
-      "learning_rate": 9.997200700525395e-05,
-      "loss": 1.3921,
-      "num_input_tokens_seen": 65536,
-      "step": 4
-    },
-    {
-      "epoch": 0.0003502412286462301,
-      "grad_norm": 6.1078338623046875,
-      "learning_rate": 9.996500875656743e-05,
-      "loss": 1.3171,
-      "num_input_tokens_seen": 81920,
-      "step": 5
-    },
-    {
-      "epoch": 0.0004202894743754761,
-      "grad_norm": 6.466420650482178,
-      "learning_rate": 9.995801050788092e-05,
-      "loss": 1.0732,
-      "num_input_tokens_seen": 97344,
-      "step": 6
-    },
-    {
-      "epoch": 0.0004903377201047221,
-      "grad_norm": 5.578189849853516,
-      "learning_rate": 9.99510122591944e-05,
-      "loss": 0.9929,
-      "num_input_tokens_seen": 113728,
-      "step": 7
-    },
-    {
-      "epoch": 0.0005603859658339682,
-      "grad_norm": 7.197720527648926,
-      "learning_rate": 9.994401401050789e-05,
-      "loss": 1.2512,
-      "num_input_tokens_seen": 129528,
-      "step": 8
-    },
-    {
-      "epoch": 0.0006304342115632141,
-      "grad_norm": 6.618913650512695,
-      "learning_rate": 9.993701576182136e-05,
-      "loss": 1.3495,
-      "num_input_tokens_seen": 145704,
-      "step": 9
-    },
-    {
-      "epoch": 0.0007004824572924602,
-      "grad_norm": 6.955508232116699,
-      "learning_rate": 9.993001751313485e-05,
-      "loss": 1.1823,
-      "num_input_tokens_seen": 161664,
-      "step": 10
-    },
-    {
-      "epoch": 0.0007705307030217062,
-      "grad_norm": 6.6807074546813965,
-      "learning_rate": 9.992301926444835e-05,
-      "loss": 1.1693,
-      "num_input_tokens_seen": 177960,
-      "step": 11
-    },
-    {
-      "epoch": 0.0008405789487509522,
-      "grad_norm": 6.784447193145752,
-      "learning_rate": 9.991602101576183e-05,
-      "loss": 1.3744,
-      "num_input_tokens_seen": 194344,
-      "step": 12
-    },
-    {
-      "epoch": 0.0009106271944801982,
-      "grad_norm": 6.7418437004089355,
-      "learning_rate": 9.990902276707532e-05,
-      "loss": 1.22,
-      "num_input_tokens_seen": 210728,
-      "step": 13
-    },
-    {
-      "epoch": 0.0009806754402094443,
-      "grad_norm": 6.43395471572876,
-      "learning_rate": 9.990202451838879e-05,
-      "loss": 1.1772,
-      "num_input_tokens_seen": 227112,
-      "step": 14
-    },
-    {
-      "epoch": 0.0010507236859386903,
-      "grad_norm": 6.09422492980957,
-      "learning_rate": 9.989502626970228e-05,
-      "loss": 1.195,
-      "num_input_tokens_seen": 243496,
-      "step": 15
-    },
-    {
-      "epoch": 0.0011207719316679364,
-      "grad_norm": 6.238271236419678,
-      "learning_rate": 9.988802802101577e-05,
-      "loss": 1.2623,
-      "num_input_tokens_seen": 259744,
-      "step": 16
-    },
-    {
-      "epoch": 0.0011908201773971822,
-      "grad_norm": 6.56187629699707,
-      "learning_rate": 9.988102977232926e-05,
-      "loss": 1.2721,
-      "num_input_tokens_seen": 276128,
-      "step": 17
-    },
-    {
-      "epoch": 0.0012608684231264283,
-      "grad_norm": 6.818358898162842,
-      "learning_rate": 9.987403152364275e-05,
-      "loss": 1.2649,
-      "num_input_tokens_seen": 292512,
-      "step": 18
-    },
-    {
-      "epoch": 0.0013309166688556743,
-      "grad_norm": 5.950352191925049,
-      "learning_rate": 9.986703327495622e-05,
-      "loss": 1.0024,
-      "num_input_tokens_seen": 308632,
-      "step": 19
-    },
-    {
-      "epoch": 0.0014009649145849204,
-      "grad_norm": 6.387479305267334,
-      "learning_rate": 9.986003502626971e-05,
-      "loss": 1.2783,
-      "num_input_tokens_seen": 325016,
-      "step": 20
-    },
-    {
-      "epoch": 0.0014710131603141664,
-      "grad_norm": 6.187346458435059,
-      "learning_rate": 9.985303677758318e-05,
-      "loss": 1.1701,
-      "num_input_tokens_seen": 341384,
-      "step": 21
-    },
-    {
-      "epoch": 0.0015410614060434125,
-      "grad_norm": 5.371951103210449,
-      "learning_rate": 9.984603852889667e-05,
-      "loss": 1.0483,
-      "num_input_tokens_seen": 357768,
-      "step": 22
-    },
-    {
-      "epoch": 0.0016111096517726585,
-      "grad_norm": 6.2206807136535645,
-      "learning_rate": 9.983904028021016e-05,
-      "loss": 1.2516,
-      "num_input_tokens_seen": 374152,
-      "step": 23
-    },
-    {
-      "epoch": 0.0016811578975019044,
-      "grad_norm": 6.121264457702637,
-      "learning_rate": 9.983204203152365e-05,
-      "loss": 1.1506,
-      "num_input_tokens_seen": 390536,
-      "step": 24
-    },
-    {
-      "epoch": 0.0017512061432311504,
-      "grad_norm": 6.353756904602051,
-      "learning_rate": 9.982504378283714e-05,
-      "loss": 1.3118,
-      "num_input_tokens_seen": 406920,
-      "step": 25
-    },
-    {
-      "epoch": 0.0018212543889603965,
-      "grad_norm": 6.270686149597168,
-      "learning_rate": 9.981804553415061e-05,
-      "loss": 1.0883,
-      "num_input_tokens_seen": 422728,
-      "step": 26
-    },
-    {
-      "epoch": 0.0018913026346896425,
-      "grad_norm": 6.117632865905762,
-      "learning_rate": 9.98110472854641e-05,
-      "loss": 1.3346,
-      "num_input_tokens_seen": 439112,
-      "step": 27
-    },
-    {
-      "epoch": 0.0019613508804188886,
-      "grad_norm": 6.429015159606934,
-      "learning_rate": 9.980404903677759e-05,
-      "loss": 1.2494,
-      "num_input_tokens_seen": 455144,
-      "step": 28
-    },
-    {
-      "epoch": 0.0020313991261481346,
-      "grad_norm": 6.4467620849609375,
-      "learning_rate": 9.979705078809107e-05,
-      "loss": 1.3335,
-      "num_input_tokens_seen": 470360,
-      "step": 29
-    },
-    {
-      "epoch": 0.0021014473718773807,
-      "grad_norm": 6.57926082611084,
-      "learning_rate": 9.979005253940455e-05,
-      "loss": 1.2126,
-      "num_input_tokens_seen": 486120,
-      "step": 30
-    },
-    {
-      "epoch": 0.0021714956176066267,
-      "grad_norm": 5.650569915771484,
-      "learning_rate": 9.978305429071804e-05,
-      "loss": 1.1363,
-      "num_input_tokens_seen": 501896,
-      "step": 31
-    },
-    {
-      "epoch": 0.0022415438633358728,
-      "grad_norm": 6.380292892456055,
-      "learning_rate": 9.977605604203153e-05,
-      "loss": 1.2251,
-      "num_input_tokens_seen": 517752,
-      "step": 32
-    },
-    {
-      "epoch": 0.002311592109065119,
-      "grad_norm": 5.704173564910889,
-      "learning_rate": 9.976905779334502e-05,
-      "loss": 1.1685,
-      "num_input_tokens_seen": 534136,
-      "step": 33
-    },
-    {
-      "epoch": 0.0023816403547943644,
-      "grad_norm": 5.342978000640869,
-      "learning_rate": 9.97620595446585e-05,
-      "loss": 1.2012,
-      "num_input_tokens_seen": 550216,
-      "step": 34
-    },
-    {
-      "epoch": 0.0024516886005236105,
-      "grad_norm": 5.7014241218566895,
-      "learning_rate": 9.975506129597198e-05,
-      "loss": 1.2342,
-      "num_input_tokens_seen": 566600,
-      "step": 35
-    },
-    {
-      "epoch": 0.0025217368462528565,
-      "grad_norm": 6.26229190826416,
-      "learning_rate": 9.974806304728546e-05,
-      "loss": 1.2041,
-      "num_input_tokens_seen": 582984,
-      "step": 36
-    },
-    {
-      "epoch": 0.0025917850919821026,
-      "grad_norm": 6.583463191986084,
-      "learning_rate": 9.974106479859896e-05,
-      "loss": 1.3021,
-      "num_input_tokens_seen": 598968,
-      "step": 37
-    },
-    {
-      "epoch": 0.0026618333377113486,
-      "grad_norm": 5.58498477935791,
-      "learning_rate": 9.973406654991245e-05,
-      "loss": 1.1622,
-      "num_input_tokens_seen": 614840,
-      "step": 38
-    },
-    {
-      "epoch": 0.0027318815834405947,
-      "grad_norm": 5.906906604766846,
-      "learning_rate": 9.972706830122592e-05,
-      "loss": 1.1971,
-      "num_input_tokens_seen": 631224,
-      "step": 39
-    },
-    {
-      "epoch": 0.0028019298291698407,
-      "grad_norm": 5.962359428405762,
-      "learning_rate": 9.972007005253941e-05,
-      "loss": 1.1326,
-      "num_input_tokens_seen": 647000,
-      "step": 40
-    },
-    {
-      "epoch": 0.002871978074899087,
-      "grad_norm": 6.447500705718994,
-      "learning_rate": 9.971307180385289e-05,
-      "loss": 1.0905,
-      "num_input_tokens_seen": 662480,
-      "step": 41
-    },
-    {
-      "epoch": 0.002942026320628333,
-      "grad_norm": 5.7290520668029785,
-      "learning_rate": 9.970607355516638e-05,
-      "loss": 1.3585,
-      "num_input_tokens_seen": 678480,
-      "step": 42
-    },
-    {
-      "epoch": 0.003012074566357579,
-      "grad_norm": 6.063445568084717,
-      "learning_rate": 9.969907530647987e-05,
-      "loss": 1.2841,
-      "num_input_tokens_seen": 694256,
-      "step": 43
-    },
-    {
-      "epoch": 0.003082122812086825,
-      "grad_norm": 5.302809238433838,
-      "learning_rate": 9.969207705779335e-05,
-      "loss": 1.1168,
-      "num_input_tokens_seen": 710152,
-      "step": 44
-    },
-    {
-      "epoch": 0.003152171057816071,
-      "grad_norm": 5.634128093719482,
-      "learning_rate": 9.968507880910684e-05,
-      "loss": 1.0609,
-      "num_input_tokens_seen": 726184,
-      "step": 45
-    },
-    {
-      "epoch": 0.003222219303545317,
-      "grad_norm": 5.652642726898193,
-      "learning_rate": 9.967808056042032e-05,
-      "loss": 1.2228,
-      "num_input_tokens_seen": 742520,
-      "step": 46
-    },
-    {
-      "epoch": 0.0032922675492745627,
-      "grad_norm": 5.340751647949219,
-      "learning_rate": 9.96710823117338e-05,
-      "loss": 1.0595,
-      "num_input_tokens_seen": 758904,
-      "step": 47
-    },
-    {
-      "epoch": 0.0033623157950038087,
-      "grad_norm": 5.422239780426025,
-      "learning_rate": 9.966408406304728e-05,
-      "loss": 1.1161,
-      "num_input_tokens_seen": 775040,
-      "step": 48
-    },
-    {
-      "epoch": 0.0034323640407330548,
-      "grad_norm": 5.29241418838501,
-      "learning_rate": 9.965708581436077e-05,
-      "loss": 1.0255,
-      "num_input_tokens_seen": 790856,
-      "step": 49
-    },
-    {
-      "epoch": 0.003502412286462301,
-      "grad_norm": 5.146270275115967,
-      "learning_rate": 9.965008756567426e-05,
-      "loss": 0.9762,
-      "num_input_tokens_seen": 807064,
-      "step": 50
-    },
-    {
-      "epoch": 0.003572460532191547,
-      "grad_norm": 5.825758457183838,
-      "learning_rate": 9.964308931698775e-05,
-      "loss": 1.2108,
-      "num_input_tokens_seen": 823448,
-      "step": 51
-    },
-    {
-      "epoch": 0.003642508777920793,
-      "grad_norm": 6.179538726806641,
-      "learning_rate": 9.963609106830124e-05,
-      "loss": 1.322,
-      "num_input_tokens_seen": 838888,
-      "step": 52
-    },
-    {
-      "epoch": 0.003712557023650039,
-      "grad_norm": 6.464454174041748,
-      "learning_rate": 9.962909281961471e-05,
-      "loss": 1.5077,
-      "num_input_tokens_seen": 855272,
-      "step": 53
-    },
-    {
-      "epoch": 0.003782605269379285,
-      "grad_norm": 5.4227294921875,
-      "learning_rate": 9.96220945709282e-05,
-      "loss": 1.2679,
-      "num_input_tokens_seen": 871656,
-      "step": 54
-    },
-    {
-      "epoch": 0.003852653515108531,
-      "grad_norm": 5.949041366577148,
-      "learning_rate": 9.961509632224169e-05,
-      "loss": 1.3618,
-      "num_input_tokens_seen": 888040,
-      "step": 55
-    },
-    {
-      "epoch": 0.003922701760837777,
-      "grad_norm": 6.050904750823975,
-      "learning_rate": 9.960809807355516e-05,
-      "loss": 1.3155,
-      "num_input_tokens_seen": 904400,
-      "step": 56
-    },
-    {
-      "epoch": 0.003992750006567023,
-      "grad_norm": 6.048308849334717,
-      "learning_rate": 9.960109982486866e-05,
-      "loss": 1.3131,
-      "num_input_tokens_seen": 919952,
-      "step": 57
-    },
-    {
-      "epoch": 0.004062798252296269,
-      "grad_norm": 5.683863162994385,
-      "learning_rate": 9.959410157618214e-05,
-      "loss": 1.1692,
-      "num_input_tokens_seen": 936336,
-      "step": 58
-    },
-    {
-      "epoch": 0.004132846498025515,
-      "grad_norm": 5.449287414550781,
-      "learning_rate": 9.958710332749563e-05,
-      "loss": 1.0613,
-      "num_input_tokens_seen": 952152,
-      "step": 59
-    },
-    {
-      "epoch": 0.004202894743754761,
-      "grad_norm": 5.31496524810791,
-      "learning_rate": 9.958010507880912e-05,
-      "loss": 0.9605,
-      "num_input_tokens_seen": 967824,
-      "step": 60
-    },
-    {
-      "epoch": 0.004272942989484007,
-      "grad_norm": 5.57105016708374,
-      "learning_rate": 9.957310683012259e-05,
-      "loss": 1.1701,
-      "num_input_tokens_seen": 983864,
-      "step": 61
-    },
-    {
-      "epoch": 0.004342991235213253,
-      "grad_norm": 5.3456830978393555,
-      "learning_rate": 9.956610858143608e-05,
-      "loss": 1.0995,
-      "num_input_tokens_seen": 1000248,
-      "step": 62
-    },
-    {
-      "epoch": 0.004413039480942499,
-      "grad_norm": 5.453295707702637,
-      "learning_rate": 9.955911033274957e-05,
-      "loss": 1.2413,
-      "num_input_tokens_seen": 1016632,
-      "step": 63
-    },
-    {
-      "epoch": 0.0044830877266717455,
-      "grad_norm": 4.975449562072754,
-      "learning_rate": 9.955211208406306e-05,
-      "loss": 1.0961,
-      "num_input_tokens_seen": 1033016,
-      "step": 64
-    },
-    {
-      "epoch": 0.004553135972400991,
-      "grad_norm": 5.542137145996094,
-      "learning_rate": 9.954511383537655e-05,
-      "loss": 1.1171,
-      "num_input_tokens_seen": 1049400,
-      "step": 65
-    },
-    {
-      "epoch": 0.004623184218130238,
-      "grad_norm": 5.213950157165527,
-      "learning_rate": 9.953811558669002e-05,
-      "loss": 1.2228,
-      "num_input_tokens_seen": 1065784,
-      "step": 66
-    },
-    {
-      "epoch": 0.004693232463859483,
-      "grad_norm": 5.496099948883057,
-      "learning_rate": 9.953111733800351e-05,
-      "loss": 1.1529,
-      "num_input_tokens_seen": 1082168,
-      "step": 67
-    },
-    {
-      "epoch": 0.004763280709588729,
-      "grad_norm": 5.64145565032959,
-      "learning_rate": 9.952411908931698e-05,
-      "loss": 1.2301,
-      "num_input_tokens_seen": 1098024,
-      "step": 68
-    },
-    {
-      "epoch": 0.004833328955317975,
-      "grad_norm": 5.566709995269775,
-      "learning_rate": 9.951712084063047e-05,
-      "loss": 1.2679,
-      "num_input_tokens_seen": 1114408,
-      "step": 69
-    },
-    {
-      "epoch": 0.004903377201047221,
-      "grad_norm": 6.443673133850098,
-      "learning_rate": 9.951012259194396e-05,
-      "loss": 1.2313,
-      "num_input_tokens_seen": 1130792,
-      "step": 70
-    },
-    {
-      "epoch": 0.0049734254467764675,
-      "grad_norm": 5.882962226867676,
-      "learning_rate": 9.950312434325745e-05,
-      "loss": 1.4304,
-      "num_input_tokens_seen": 1147176,
-      "step": 71
-    },
-    {
-      "epoch": 0.005043473692505713,
-      "grad_norm": 6.0052666664123535,
-      "learning_rate": 9.949612609457094e-05,
-      "loss": 1.3027,
-      "num_input_tokens_seen": 1160968,
-      "step": 72
-    },
-    {
-      "epoch": 0.0051135219382349596,
-      "grad_norm": 5.260256767272949,
-      "learning_rate": 9.948912784588441e-05,
-      "loss": 1.1526,
-      "num_input_tokens_seen": 1177352,
-      "step": 73
-    },
-    {
-      "epoch": 0.005183570183964205,
-      "grad_norm": 5.641814708709717,
-      "learning_rate": 9.94821295971979e-05,
-      "loss": 1.0666,
-      "num_input_tokens_seen": 1193032,
-      "step": 74
-    },
-    {
-      "epoch": 0.005253618429693452,
-      "grad_norm": 5.121115207672119,
-      "learning_rate": 9.947513134851138e-05,
-      "loss": 1.2404,
-      "num_input_tokens_seen": 1208952,
-      "step": 75
-    },
-    {
-      "epoch": 0.005323666675422697,
-      "grad_norm": 5.63930082321167,
-      "learning_rate": 9.946813309982487e-05,
-      "loss": 1.5127,
-      "num_input_tokens_seen": 1225000,
-      "step": 76
-    },
-    {
-      "epoch": 0.005393714921151944,
-      "grad_norm": 4.880716800689697,
-      "learning_rate": 9.946113485113837e-05,
-      "loss": 1.1484,
-      "num_input_tokens_seen": 1241384,
-      "step": 77
-    },
-    {
-      "epoch": 0.005463763166881189,
-      "grad_norm": 5.59611177444458,
-      "learning_rate": 9.945413660245184e-05,
-      "loss": 1.1678,
-      "num_input_tokens_seen": 1257680,
-      "step": 78
-    },
-    {
-      "epoch": 0.005533811412610436,
-      "grad_norm": 5.052026271820068,
-      "learning_rate": 9.944713835376533e-05,
-      "loss": 1.2207,
-      "num_input_tokens_seen": 1274064,
-      "step": 79
-    },
-    {
-      "epoch": 0.0056038596583396815,
-      "grad_norm": 5.285096168518066,
-      "learning_rate": 9.944014010507881e-05,
-      "loss": 1.1457,
-      "num_input_tokens_seen": 1290448,
-      "step": 80
-    },
-    {
-      "epoch": 0.005673907904068927,
-      "grad_norm": 5.4286580085754395,
-      "learning_rate": 9.94331418563923e-05,
-      "loss": 1.3047,
-      "num_input_tokens_seen": 1306832,
-      "step": 81
-    },
-    {
-      "epoch": 0.005743956149798174,
-      "grad_norm": 5.937953472137451,
-      "learning_rate": 9.942614360770578e-05,
-      "loss": 1.4353,
-      "num_input_tokens_seen": 1323216,
-      "step": 82
-    },
-    {
-      "epoch": 0.005814004395527419,
-      "grad_norm": 5.129006385803223,
-      "learning_rate": 9.941914535901927e-05,
-      "loss": 1.1434,
-      "num_input_tokens_seen": 1339408,
-      "step": 83
-    },
-    {
-      "epoch": 0.005884052641256666,
-      "grad_norm": 5.179675102233887,
-      "learning_rate": 9.941214711033276e-05,
-      "loss": 1.2452,
-      "num_input_tokens_seen": 1355792,
-      "step": 84
-    },
-    {
-      "epoch": 0.005954100886985911,
-      "grad_norm": 4.912832736968994,
-      "learning_rate": 9.940514886164624e-05,
-      "loss": 1.1255,
-      "num_input_tokens_seen": 1372176,
-      "step": 85
-    },
-    {
-      "epoch": 0.006024149132715158,
-      "grad_norm": 5.190899848937988,
-      "learning_rate": 9.939815061295973e-05,
-      "loss": 1.2543,
-      "num_input_tokens_seen": 1388560,
-      "step": 86
-    },
-    {
-      "epoch": 0.006094197378444403,
-      "grad_norm": 5.1751275062561035,
-      "learning_rate": 9.939115236427321e-05,
-      "loss": 1.3145,
-      "num_input_tokens_seen": 1404944,
-      "step": 87
-    },
-    {
-      "epoch": 0.00616424562417365,
-      "grad_norm": 5.450705528259277,
-      "learning_rate": 9.938415411558669e-05,
-      "loss": 1.2844,
-      "num_input_tokens_seen": 1421328,
-      "step": 88
-    },
-    {
-      "epoch": 0.0062342938699028955,
-      "grad_norm": 5.593935012817383,
-      "learning_rate": 9.937715586690018e-05,
-      "loss": 1.3284,
-      "num_input_tokens_seen": 1437464,
-      "step": 89
-    },
-    {
-      "epoch": 0.006304342115632142,
-      "grad_norm": 5.156428813934326,
-      "learning_rate": 9.937015761821367e-05,
-      "loss": 1.1682,
-      "num_input_tokens_seen": 1452952,
-      "step": 90
-    },
-    {
-      "epoch": 0.006374390361361388,
-      "grad_norm": 4.673638820648193,
-      "learning_rate": 9.936315936952715e-05,
-      "loss": 1.004,
-      "num_input_tokens_seen": 1469336,
-      "step": 91
-    },
-    {
-      "epoch": 0.006444438607090634,
-      "grad_norm": 4.996700763702393,
-      "learning_rate": 9.935616112084064e-05,
-      "loss": 1.087,
-      "num_input_tokens_seen": 1485448,
-      "step": 92
-    },
-    {
-      "epoch": 0.00651448685281988,
-      "grad_norm": 4.817474365234375,
-      "learning_rate": 9.934916287215412e-05,
-      "loss": 1.151,
-      "num_input_tokens_seen": 1501472,
-      "step": 93
-    },
-    {
-      "epoch": 0.006584535098549125,
-      "grad_norm": 5.400479316711426,
-      "learning_rate": 9.934216462346761e-05,
-      "loss": 1.3144,
-      "num_input_tokens_seen": 1516424,
-      "step": 94
-    },
-    {
-      "epoch": 0.006654583344278372,
-      "grad_norm": 5.232216835021973,
-      "learning_rate": 9.933516637478108e-05,
-      "loss": 1.0019,
-      "num_input_tokens_seen": 1532792,
-      "step": 95
-    },
-    {
-      "epoch": 0.006724631590007617,
-      "grad_norm": 5.392521381378174,
-      "learning_rate": 9.932816812609457e-05,
-      "loss": 1.3195,
-      "num_input_tokens_seen": 1548600,
-      "step": 96
-    },
-    {
-      "epoch": 0.006794679835736864,
-      "grad_norm": 5.5280866622924805,
-      "learning_rate": 9.932116987740806e-05,
-      "loss": 1.283,
-      "num_input_tokens_seen": 1564088,
-      "step": 97
-    },
-    {
-      "epoch": 0.0068647280814661095,
-      "grad_norm": 4.963179588317871,
-      "learning_rate": 9.931417162872155e-05,
-      "loss": 1.2716,
-      "num_input_tokens_seen": 1580040,
-      "step": 98
-    },
-    {
-      "epoch": 0.006934776327195356,
-      "grad_norm": 4.920302391052246,
-      "learning_rate": 9.930717338003504e-05,
-      "loss": 1.088,
-      "num_input_tokens_seen": 1595880,
-      "step": 99
-    },
-    {
-      "epoch": 0.007004824572924602,
-      "grad_norm": 4.935486793518066,
-      "learning_rate": 9.930017513134851e-05,
-      "loss": 1.0122,
-      "num_input_tokens_seen": 1611864,
-      "step": 100
-    },
-    {
-      "epoch": 0.007074872818653848,
-      "grad_norm": 5.099087238311768,
-      "learning_rate": 9.9293176882662e-05,
-      "loss": 1.1605,
-      "num_input_tokens_seen": 1627472,
-      "step": 101
-    },
-    {
-      "epoch": 0.007144921064383094,
-      "grad_norm": 5.3764328956604,
-      "learning_rate": 9.928617863397548e-05,
-      "loss": 1.2225,
-      "num_input_tokens_seen": 1643856,
-      "step": 102
-    },
-    {
-      "epoch": 0.00721496931011234,
-      "grad_norm": 5.281564712524414,
-      "learning_rate": 9.927918038528898e-05,
-      "loss": 1.1483,
-      "num_input_tokens_seen": 1660240,
-      "step": 103
-    },
-    {
-      "epoch": 0.007285017555841586,
-      "grad_norm": 5.395167827606201,
-      "learning_rate": 9.927218213660247e-05,
-      "loss": 1.6014,
-      "num_input_tokens_seen": 1676624,
-      "step": 104
-    },
-    {
-      "epoch": 0.007355065801570832,
-      "grad_norm": 5.322319507598877,
-      "learning_rate": 9.926518388791594e-05,
-      "loss": 1.0933,
-      "num_input_tokens_seen": 1693008,
-      "step": 105
-    },
-    {
-      "epoch": 0.007425114047300078,
-      "grad_norm": 5.301229953765869,
-      "learning_rate": 9.925818563922943e-05,
-      "loss": 1.1998,
-      "num_input_tokens_seen": 1708424,
-      "step": 106
-    },
-    {
-      "epoch": 0.0074951622930293236,
-      "grad_norm": 4.958597183227539,
-      "learning_rate": 9.92511873905429e-05,
-      "loss": 1.3285,
-      "num_input_tokens_seen": 1724808,
-      "step": 107
-    },
-    {
-      "epoch": 0.00756521053875857,
-      "grad_norm": 4.3913960456848145,
-      "learning_rate": 9.924418914185639e-05,
-      "loss": 0.9017,
-      "num_input_tokens_seen": 1740752,
-      "step": 108
-    },
-    {
-      "epoch": 0.007635258784487816,
-      "grad_norm": 5.401021480560303,
-      "learning_rate": 9.923719089316988e-05,
-      "loss": 1.3646,
-      "num_input_tokens_seen": 1755176,
-      "step": 109
-    },
-    {
-      "epoch": 0.007705307030217062,
-      "grad_norm": 4.894444942474365,
-      "learning_rate": 9.923019264448337e-05,
-      "loss": 0.9955,
-      "num_input_tokens_seen": 1771560,
-      "step": 110
-    },
-    {
-      "epoch": 0.007775355275946308,
-      "grad_norm": 4.878688335418701,
-      "learning_rate": 9.922319439579686e-05,
-      "loss": 1.1766,
-      "num_input_tokens_seen": 1787944,
-      "step": 111
-    },
-    {
-      "epoch": 0.007845403521675554,
-      "grad_norm": 4.9379777908325195,
-      "learning_rate": 9.921619614711033e-05,
-      "loss": 1.1631,
-      "num_input_tokens_seen": 1803568,
-      "step": 112
-    },
-    {
-      "epoch": 0.0079154517674048,
-      "grad_norm": 5.101811408996582,
-      "learning_rate": 9.920919789842382e-05,
-      "loss": 1.2165,
-      "num_input_tokens_seen": 1819952,
-      "step": 113
-    },
-    {
-      "epoch": 0.007985500013134045,
-      "grad_norm": 5.32574987411499,
-      "learning_rate": 9.920219964973731e-05,
-      "loss": 1.3012,
-      "num_input_tokens_seen": 1835296,
-      "step": 114
-    },
-    {
-      "epoch": 0.008055548258863293,
-      "grad_norm": 5.2391180992126465,
-      "learning_rate": 9.919520140105079e-05,
-      "loss": 1.2451,
-      "num_input_tokens_seen": 1851224,
-      "step": 115
-    },
-    {
-      "epoch": 0.008125596504592538,
-      "grad_norm": 4.865017890930176,
-      "learning_rate": 9.918820315236427e-05,
-      "loss": 1.1683,
-      "num_input_tokens_seen": 1867608,
-      "step": 116
-    },
-    {
-      "epoch": 0.008195644750321784,
-      "grad_norm": 4.943136215209961,
-      "learning_rate": 9.918120490367776e-05,
-      "loss": 1.31,
-      "num_input_tokens_seen": 1883696,
-      "step": 117
-    },
-    {
-      "epoch": 0.00826569299605103,
-      "grad_norm": 4.769871711730957,
-      "learning_rate": 9.917420665499125e-05,
-      "loss": 1.1212,
-      "num_input_tokens_seen": 1900080,
-      "step": 118
-    },
-    {
-      "epoch": 0.008335741241780275,
-      "grad_norm": 4.785780429840088,
-      "learning_rate": 9.916720840630474e-05,
-      "loss": 1.2415,
-      "num_input_tokens_seen": 1916464,
-      "step": 119
-    },
-    {
-      "epoch": 0.008405789487509523,
-      "grad_norm": 4.802333831787109,
-      "learning_rate": 9.916021015761822e-05,
-      "loss": 1.0513,
-      "num_input_tokens_seen": 1932848,
-      "step": 120
-    },
-    {
-      "epoch": 0.008475837733238768,
-      "grad_norm": 5.22212553024292,
-      "learning_rate": 9.91532119089317e-05,
-      "loss": 1.2574,
-      "num_input_tokens_seen": 1949232,
-      "step": 121
-    },
-    {
-      "epoch": 0.008545885978968014,
-      "grad_norm": 5.104204177856445,
-      "learning_rate": 9.914621366024518e-05,
-      "loss": 1.0436,
-      "num_input_tokens_seen": 1964184,
-      "step": 122
-    },
-    {
-      "epoch": 0.00861593422469726,
-      "grad_norm": 5.11055326461792,
-      "learning_rate": 9.913921541155868e-05,
-      "loss": 1.1939,
-      "num_input_tokens_seen": 1980568,
-      "step": 123
-    },
-    {
-      "epoch": 0.008685982470426507,
-      "grad_norm": 4.784866809844971,
-      "learning_rate": 9.913221716287216e-05,
-      "loss": 1.2056,
-      "num_input_tokens_seen": 1996952,
-      "step": 124
-    },
-    {
-      "epoch": 0.008756030716155752,
-      "grad_norm": 4.763037204742432,
-      "learning_rate": 9.912521891418564e-05,
-      "loss": 1.1403,
-      "num_input_tokens_seen": 2013336,
-      "step": 125
-    },
-    {
-      "epoch": 0.008826078961884998,
-      "grad_norm": 4.813408851623535,
-      "learning_rate": 9.911822066549913e-05,
-      "loss": 1.1897,
-      "num_input_tokens_seen": 2029720,
-      "step": 126
-    },
-    {
-      "epoch": 0.008896127207614244,
-      "grad_norm": 4.79008674621582,
-      "learning_rate": 9.911122241681261e-05,
-      "loss": 1.2315,
-      "num_input_tokens_seen": 2046104,
-      "step": 127
-    },
-    {
-      "epoch": 0.008966175453343491,
-      "grad_norm": 4.843508720397949,
-      "learning_rate": 9.91042241681261e-05,
-      "loss": 1.0883,
-      "num_input_tokens_seen": 2061592,
-      "step": 128
-    },
-    {
-      "epoch": 0.009036223699072737,
-      "grad_norm": 4.917592525482178,
-      "learning_rate": 9.909722591943959e-05,
-      "loss": 1.2512,
-      "num_input_tokens_seen": 2077792,
-      "step": 129
-    },
-    {
-      "epoch": 0.009106271944801982,
-      "grad_norm": 4.9154133796691895,
-      "learning_rate": 9.909022767075307e-05,
-      "loss": 1.3284,
-      "num_input_tokens_seen": 2094176,
-      "step": 130
-    },
-    {
-      "epoch": 0.009176320190531228,
-      "grad_norm": 5.2125420570373535,
-      "learning_rate": 9.908322942206656e-05,
-      "loss": 1.3469,
-      "num_input_tokens_seen": 2110480,
-      "step": 131
-    },
-    {
-      "epoch": 0.009246368436260475,
-      "grad_norm": 4.715712547302246,
-      "learning_rate": 9.907623117338004e-05,
-      "loss": 1.0844,
-      "num_input_tokens_seen": 2126864,
-      "step": 132
-    },
-    {
-      "epoch": 0.009316416681989721,
-      "grad_norm": 4.805694580078125,
-      "learning_rate": 9.906923292469353e-05,
-      "loss": 1.069,
-      "num_input_tokens_seen": 2142848,
-      "step": 133
-    },
-    {
-      "epoch": 0.009386464927718966,
-      "grad_norm": 4.961355209350586,
-      "learning_rate": 9.9062234676007e-05,
-      "loss": 1.3387,
-      "num_input_tokens_seen": 2159232,
-      "step": 134
-    },
-    {
-      "epoch": 0.009456513173448212,
-      "grad_norm": 4.582219123840332,
-      "learning_rate": 9.905523642732049e-05,
-      "loss": 1.2013,
-      "num_input_tokens_seen": 2175616,
-      "step": 135
-    },
-    {
-      "epoch": 0.009526561419177458,
-      "grad_norm": 5.195998191833496,
-      "learning_rate": 9.904823817863398e-05,
-      "loss": 1.2552,
-      "num_input_tokens_seen": 2191872,
-      "step": 136
-    },
-    {
-      "epoch": 0.009596609664906705,
-      "grad_norm": 4.934189319610596,
-      "learning_rate": 9.904123992994747e-05,
-      "loss": 1.2961,
-      "num_input_tokens_seen": 2208208,
-      "step": 137
-    },
-    {
-      "epoch": 0.00966665791063595,
-      "grad_norm": 4.981037616729736,
-      "learning_rate": 9.903424168126096e-05,
-      "loss": 1.1546,
-      "num_input_tokens_seen": 2224592,
-      "step": 138
-    },
-    {
-      "epoch": 0.009736706156365196,
-      "grad_norm": 5.469496250152588,
-      "learning_rate": 9.902724343257443e-05,
-      "loss": 1.3833,
-      "num_input_tokens_seen": 2240976,
-      "step": 139
-    },
-    {
-      "epoch": 0.009806754402094442,
-      "grad_norm": 4.889583587646484,
-      "learning_rate": 9.902024518388792e-05,
-      "loss": 1.2095,
-      "num_input_tokens_seen": 2257360,
-      "step": 140
-    },
-    {
-      "epoch": 0.00987680264782369,
-      "grad_norm": 4.532052516937256,
-      "learning_rate": 9.901324693520141e-05,
-      "loss": 1.143,
-      "num_input_tokens_seen": 2272848,
-      "step": 141
-    },
-    {
-      "epoch": 0.009946850893552935,
-      "grad_norm": 5.278079032897949,
-      "learning_rate": 9.900624868651488e-05,
-      "loss": 1.2849,
-      "num_input_tokens_seen": 2289232,
-      "step": 142
-    },
-    {
-      "epoch": 0.01001689913928218,
-      "grad_norm": 4.549891948699951,
-      "learning_rate": 9.899925043782839e-05,
-      "loss": 1.0482,
-      "num_input_tokens_seen": 2305424,
-      "step": 143
-    },
-    {
-      "epoch": 0.010086947385011426,
-      "grad_norm": 4.7777180671691895,
-      "learning_rate": 9.899225218914186e-05,
-      "loss": 1.1926,
-      "num_input_tokens_seen": 2320968,
-      "step": 144
-    },
-    {
-      "epoch": 0.010156995630740673,
-      "grad_norm": 4.320313453674316,
-      "learning_rate": 9.898525394045535e-05,
-      "loss": 1.0468,
-      "num_input_tokens_seen": 2337352,
-      "step": 145
-    },
-    {
-      "epoch": 0.010227043876469919,
-      "grad_norm": 4.915202617645264,
-      "learning_rate": 9.897825569176882e-05,
-      "loss": 1.1326,
-      "num_input_tokens_seen": 2353064,
-      "step": 146
-    },
-    {
-      "epoch": 0.010297092122199165,
-      "grad_norm": 4.569783687591553,
-      "learning_rate": 9.897125744308231e-05,
-      "loss": 0.8586,
-      "num_input_tokens_seen": 2369128,
-      "step": 147
-    },
-    {
-      "epoch": 0.01036714036792841,
-      "grad_norm": 4.591664791107178,
-      "learning_rate": 9.89642591943958e-05,
-      "loss": 1.1369,
-      "num_input_tokens_seen": 2385512,
-      "step": 148
-    },
-    {
-      "epoch": 0.010437188613657656,
-      "grad_norm": 4.913016319274902,
-      "learning_rate": 9.895726094570929e-05,
-      "loss": 1.1564,
-      "num_input_tokens_seen": 2401208,
-      "step": 149
-    },
-    {
-      "epoch": 0.010507236859386903,
-      "grad_norm": 4.908018112182617,
-      "learning_rate": 9.895026269702278e-05,
-      "loss": 1.1247,
-      "num_input_tokens_seen": 2417592,
-      "step": 150
-    },
-    {
-      "epoch": 0.010577285105116149,
-      "grad_norm": 4.536910057067871,
-      "learning_rate": 9.894326444833625e-05,
-      "loss": 1.014,
-      "num_input_tokens_seen": 2433976,
-      "step": 151
-    },
-    {
-      "epoch": 0.010647333350845395,
-      "grad_norm": 4.899227142333984,
-      "learning_rate": 9.893626619964974e-05,
-      "loss": 1.0418,
-      "num_input_tokens_seen": 2448072,
-      "step": 152
-    },
-    {
-      "epoch": 0.01071738159657464,
-      "grad_norm": 4.600861072540283,
-      "learning_rate": 9.892926795096323e-05,
-      "loss": 1.0459,
-      "num_input_tokens_seen": 2464240,
-      "step": 153
-    },
-    {
-      "epoch": 0.010787429842303888,
-      "grad_norm": 4.707681179046631,
-      "learning_rate": 9.89222697022767e-05,
-      "loss": 1.0859,
-      "num_input_tokens_seen": 2480624,
-      "step": 154
-    },
-    {
-      "epoch": 0.010857478088033133,
-      "grad_norm": 4.748518466949463,
-      "learning_rate": 9.89152714535902e-05,
-      "loss": 1.0608,
-      "num_input_tokens_seen": 2497008,
-      "step": 155
-    },
-    {
-      "epoch": 0.010927526333762379,
-      "grad_norm": 4.794179439544678,
-      "learning_rate": 9.890827320490368e-05,
-      "loss": 1.2243,
-      "num_input_tokens_seen": 2513392,
-      "step": 156
-    },
-    {
-      "epoch": 0.010997574579491624,
-      "grad_norm": 4.593925476074219,
-      "learning_rate": 9.890127495621717e-05,
-      "loss": 1.1002,
-      "num_input_tokens_seen": 2529776,
-      "step": 157
-    },
-    {
-      "epoch": 0.011067622825220872,
-      "grad_norm": 4.318257808685303,
-      "learning_rate": 9.889427670753066e-05,
-      "loss": 0.9561,
-      "num_input_tokens_seen": 2546160,
-      "step": 158
-    },
-    {
-      "epoch": 0.011137671070950117,
-      "grad_norm": 4.631777286529541,
-      "learning_rate": 9.888727845884414e-05,
-      "loss": 1.1553,
-      "num_input_tokens_seen": 2562544,
-      "step": 159
-    },
-    {
-      "epoch": 0.011207719316679363,
-      "grad_norm": 4.896609783172607,
-      "learning_rate": 9.888028021015762e-05,
-      "loss": 1.1779,
-      "num_input_tokens_seen": 2578088,
-      "step": 160
-    },
-    {
-      "epoch": 0.011277767562408609,
-      "grad_norm": 4.3978681564331055,
-      "learning_rate": 9.88732819614711e-05,
-      "loss": 1.1778,
-      "num_input_tokens_seen": 2594416,
-      "step": 161
-    },
-    {
-      "epoch": 0.011347815808137854,
-      "grad_norm": 4.82927942276001,
-      "learning_rate": 9.886628371278459e-05,
-      "loss": 1.0339,
-      "num_input_tokens_seen": 2609776,
-      "step": 162
-    },
-    {
-      "epoch": 0.011417864053867102,
-      "grad_norm": 4.413319110870361,
-      "learning_rate": 9.885928546409809e-05,
-      "loss": 1.0992,
-      "num_input_tokens_seen": 2626160,
-      "step": 163
-    },
-    {
-      "epoch": 0.011487912299596347,
-      "grad_norm": 4.626354694366455,
-      "learning_rate": 9.885228721541156e-05,
-      "loss": 1.1948,
-      "num_input_tokens_seen": 2642464,
-      "step": 164
-    },
-    {
-      "epoch": 0.011557960545325593,
-      "grad_norm": 4.328434467315674,
-      "learning_rate": 9.884528896672505e-05,
-      "loss": 1.1493,
-      "num_input_tokens_seen": 2658528,
-      "step": 165
-    },
-    {
-      "epoch": 0.011628008791054838,
-      "grad_norm": 4.57839822769165,
-      "learning_rate": 9.883829071803853e-05,
-      "loss": 1.0775,
-      "num_input_tokens_seen": 2674912,
-      "step": 166
-    },
-    {
-      "epoch": 0.011698057036784086,
-      "grad_norm": 5.103973865509033,
-      "learning_rate": 9.883129246935202e-05,
-      "loss": 1.2458,
-      "num_input_tokens_seen": 2690792,
-      "step": 167
-    },
-    {
-      "epoch": 0.011768105282513331,
-      "grad_norm": 4.558016300201416,
-      "learning_rate": 9.88242942206655e-05,
-      "loss": 1.0122,
-      "num_input_tokens_seen": 2705616,
-      "step": 168
-    },
-    {
-      "epoch": 0.011838153528242577,
-      "grad_norm": 4.811260223388672,
-      "learning_rate": 9.8817295971979e-05,
-      "loss": 1.2989,
-      "num_input_tokens_seen": 2721704,
-      "step": 169
-    },
-    {
-      "epoch": 0.011908201773971823,
-      "grad_norm": 4.726966857910156,
-      "learning_rate": 9.881029772329248e-05,
-      "loss": 1.176,
-      "num_input_tokens_seen": 2738088,
-      "step": 170
-    },
-    {
-      "epoch": 0.01197825001970107,
-      "grad_norm": 4.874902725219727,
-      "learning_rate": 9.880329947460596e-05,
-      "loss": 1.2586,
-      "num_input_tokens_seen": 2754040,
-      "step": 171
-    },
-    {
-      "epoch": 0.012048298265430316,
-      "grad_norm": 4.379549980163574,
-      "learning_rate": 9.879630122591945e-05,
-      "loss": 1.1771,
-      "num_input_tokens_seen": 2770424,
-      "step": 172
-    },
-    {
-      "epoch": 0.012118346511159561,
-      "grad_norm": 4.455331802368164,
-      "learning_rate": 9.878930297723292e-05,
-      "loss": 1.0714,
-      "num_input_tokens_seen": 2786808,
-      "step": 173
-    },
-    {
-      "epoch": 0.012188394756888807,
-      "grad_norm": 4.42273473739624,
-      "learning_rate": 9.878230472854641e-05,
-      "loss": 1.1798,
-      "num_input_tokens_seen": 2803176,
-      "step": 174
-    },
-    {
-      "epoch": 0.012258443002618052,
-      "grad_norm": 4.4078874588012695,
-      "learning_rate": 9.87753064798599e-05,
-      "loss": 1.1672,
-      "num_input_tokens_seen": 2819448,
-      "step": 175
-    },
-    {
-      "epoch": 0.0123284912483473,
-      "grad_norm": 4.79048490524292,
-      "learning_rate": 9.876830823117339e-05,
-      "loss": 1.3331,
-      "num_input_tokens_seen": 2835832,
-      "step": 176
-    },
-    {
-      "epoch": 0.012398539494076545,
-      "grad_norm": 4.212133884429932,
-      "learning_rate": 9.876130998248688e-05,
-      "loss": 1.0007,
-      "num_input_tokens_seen": 2851776,
-      "step": 177
-    },
-    {
-      "epoch": 0.012468587739805791,
-      "grad_norm": 5.7587738037109375,
-      "learning_rate": 9.875431173380035e-05,
-      "loss": 1.4729,
-      "num_input_tokens_seen": 2867896,
-      "step": 178
-    },
-    {
-      "epoch": 0.012538635985535037,
-      "grad_norm": 4.3469462394714355,
-      "learning_rate": 9.874731348511384e-05,
-      "loss": 0.957,
-      "num_input_tokens_seen": 2884280,
-      "step": 179
-    },
-    {
-      "epoch": 0.012608684231264284,
-      "grad_norm": 4.584625244140625,
-      "learning_rate": 9.874031523642733e-05,
-      "loss": 1.0753,
-      "num_input_tokens_seen": 2899208,
-      "step": 180
-    },
-    {
-      "epoch": 0.01267873247699353,
-      "grad_norm": 4.544627666473389,
-      "learning_rate": 9.87333169877408e-05,
-      "loss": 1.1706,
-      "num_input_tokens_seen": 2915416,
-      "step": 181
-    },
-    {
-      "epoch": 0.012748780722722775,
-      "grad_norm": 4.8749237060546875,
-      "learning_rate": 9.872631873905429e-05,
-      "loss": 1.3382,
-      "num_input_tokens_seen": 2931360,
-      "step": 182
-    },
-    {
-      "epoch": 0.01281882896845202,
-      "grad_norm": 4.593903541564941,
-      "learning_rate": 9.871932049036778e-05,
-      "loss": 1.1588,
-      "num_input_tokens_seen": 2947744,
-      "step": 183
-    },
-    {
-      "epoch": 0.012888877214181268,
-      "grad_norm": 4.478219509124756,
-      "learning_rate": 9.871232224168127e-05,
-      "loss": 1.1013,
-      "num_input_tokens_seen": 2963664,
-      "step": 184
-    },
-    {
-      "epoch": 0.012958925459910514,
-      "grad_norm": 5.028106212615967,
-      "learning_rate": 9.870532399299476e-05,
-      "loss": 1.3223,
-      "num_input_tokens_seen": 2980048,
-      "step": 185
-    },
-    {
-      "epoch": 0.01302897370563976,
-      "grad_norm": 4.866946697235107,
-      "learning_rate": 9.869832574430823e-05,
-      "loss": 1.2376,
-      "num_input_tokens_seen": 2995992,
-      "step": 186
-    },
-    {
-      "epoch": 0.013099021951369005,
-      "grad_norm": 4.421341419219971,
-      "learning_rate": 9.869132749562172e-05,
-      "loss": 1.2252,
-      "num_input_tokens_seen": 3012000,
-      "step": 187
-    },
-    {
-      "epoch": 0.01316907019709825,
-      "grad_norm": 4.88083028793335,
-      "learning_rate": 9.86843292469352e-05,
-      "loss": 1.2951,
-      "num_input_tokens_seen": 3028384,
-      "step": 188
-    },
-    {
-      "epoch": 0.013239118442827498,
-      "grad_norm": 4.654318809509277,
-      "learning_rate": 9.86773309982487e-05,
-      "loss": 1.2839,
-      "num_input_tokens_seen": 3044768,
-      "step": 189
-    },
-    {
-      "epoch": 0.013309166688556744,
-      "grad_norm": 4.626763820648193,
-      "learning_rate": 9.867033274956219e-05,
-      "loss": 1.2389,
-      "num_input_tokens_seen": 3061152,
-      "step": 190
-    },
-    {
-      "epoch": 0.01337921493428599,
-      "grad_norm": 4.178484916687012,
-      "learning_rate": 9.866333450087566e-05,
-      "loss": 1.1186,
-      "num_input_tokens_seen": 3077056,
-      "step": 191
-    },
-    {
-      "epoch": 0.013449263180015235,
-      "grad_norm": 4.755034923553467,
-      "learning_rate": 9.865633625218915e-05,
-      "loss": 1.0594,
-      "num_input_tokens_seen": 3093400,
-      "step": 192
-    },
-    {
-      "epoch": 0.013519311425744482,
-      "grad_norm": 4.437506198883057,
-      "learning_rate": 9.864933800350263e-05,
-      "loss": 1.2078,
-      "num_input_tokens_seen": 3109784,
-      "step": 193
-    },
-    {
-      "epoch": 0.013589359671473728,
-      "grad_norm": 5.140488624572754,
-      "learning_rate": 9.864233975481611e-05,
-      "loss": 1.4312,
-      "num_input_tokens_seen": 3124976,
-      "step": 194
-    },
-    {
-      "epoch": 0.013659407917202973,
-      "grad_norm": 4.72155237197876,
-      "learning_rate": 9.86353415061296e-05,
-      "loss": 1.1752,
-      "num_input_tokens_seen": 3140632,
-      "step": 195
-    },
-    {
-      "epoch": 0.013729456162932219,
-      "grad_norm": 4.914645671844482,
-      "learning_rate": 9.862834325744309e-05,
-      "loss": 1.2464,
-      "num_input_tokens_seen": 3156616,
-      "step": 196
-    },
-    {
-      "epoch": 0.013799504408661466,
-      "grad_norm": 4.23387336730957,
-      "learning_rate": 9.862134500875658e-05,
-      "loss": 0.9722,
-      "num_input_tokens_seen": 3172840,
-      "step": 197
-    },
-    {
-      "epoch": 0.013869552654390712,
-      "grad_norm": 4.659370422363281,
-      "learning_rate": 9.861434676007005e-05,
-      "loss": 1.1981,
-      "num_input_tokens_seen": 3188584,
-      "step": 198
-    },
-    {
-      "epoch": 0.013939600900119958,
-      "grad_norm": 4.580902576446533,
-      "learning_rate": 9.860734851138354e-05,
-      "loss": 1.1913,
-      "num_input_tokens_seen": 3204432,
-      "step": 199
-    },
-    {
-      "epoch": 0.014009649145849203,
-      "grad_norm": 4.208237648010254,
-      "learning_rate": 9.860035026269702e-05,
-      "loss": 1.2056,
-      "num_input_tokens_seen": 3220816,
-      "step": 200
-    },
-    {
-      "epoch": 0.014009649145849203,
-      "eval_loss": 1.2226407527923584,
-      "eval_runtime": 0.3992,
-      "eval_samples_per_second": 2.505,
-      "eval_steps_per_second": 2.505,
-      "num_input_tokens_seen": 3220816,
-      "step": 200
-    },
-    {
-      "epoch": 0.014079697391578449,
-      "grad_norm": 4.526260852813721,
-      "learning_rate": 9.85933520140105e-05,
-      "loss": 1.0488,
-      "num_input_tokens_seen": 3237200,
-      "step": 201
-    },
-    {
-      "epoch": 0.014149745637307696,
-      "grad_norm": 4.46895170211792,
-      "learning_rate": 9.8586353765324e-05,
-      "loss": 1.1101,
-      "num_input_tokens_seen": 3253336,
-      "step": 202
-    },
-    {
-      "epoch": 0.014219793883036942,
-      "grad_norm": 4.367347717285156,
-      "learning_rate": 9.857935551663748e-05,
-      "loss": 1.0425,
-      "num_input_tokens_seen": 3269632,
-      "step": 203
-    },
-    {
-      "epoch": 0.014289842128766187,
-      "grad_norm": 4.860860347747803,
-      "learning_rate": 9.857235726795097e-05,
-      "loss": 1.4068,
-      "num_input_tokens_seen": 3285432,
-      "step": 204
-    },
-    {
-      "epoch": 0.014359890374495433,
-      "grad_norm": 4.336480617523193,
-      "learning_rate": 9.856535901926445e-05,
-      "loss": 1.2579,
-      "num_input_tokens_seen": 3301632,
-      "step": 205
-    },
-    {
-      "epoch": 0.01442993862022468,
-      "grad_norm": 4.587873458862305,
-      "learning_rate": 9.855836077057794e-05,
-      "loss": 1.1508,
-      "num_input_tokens_seen": 3318016,
-      "step": 206
-    },
-    {
-      "epoch": 0.014499986865953926,
-      "grad_norm": 4.719262599945068,
-      "learning_rate": 9.855136252189142e-05,
-      "loss": 1.0208,
-      "num_input_tokens_seen": 3333168,
-      "step": 207
-    },
-    {
-      "epoch": 0.014570035111683172,
-      "grad_norm": 4.419138431549072,
-      "learning_rate": 9.85443642732049e-05,
-      "loss": 1.2576,
-      "num_input_tokens_seen": 3349384,
-      "step": 208
-    },
-    {
-      "epoch": 0.014640083357412417,
-      "grad_norm": 4.3150835037231445,
-      "learning_rate": 9.85373660245184e-05,
-      "loss": 1.1786,
-      "num_input_tokens_seen": 3365768,
-      "step": 209
-    },
-    {
-      "epoch": 0.014710131603141665,
-      "grad_norm": 4.5917649269104,
-      "learning_rate": 9.853036777583188e-05,
-      "loss": 1.2821,
-      "num_input_tokens_seen": 3382152,
-      "step": 210
-    },
-    {
-      "epoch": 0.01478017984887091,
-      "grad_norm": 4.9094343185424805,
-      "learning_rate": 9.852336952714537e-05,
-      "loss": 1.2415,
-      "num_input_tokens_seen": 3397896,
-      "step": 211
-    },
-    {
-      "epoch": 0.014850228094600156,
-      "grad_norm": 4.394861698150635,
-      "learning_rate": 9.851637127845885e-05,
-      "loss": 1.1776,
-      "num_input_tokens_seen": 3414280,
-      "step": 212
-    },
-    {
-      "epoch": 0.014920276340329401,
-      "grad_norm": 4.196374416351318,
-      "learning_rate": 9.850937302977233e-05,
-      "loss": 1.065,
-      "num_input_tokens_seen": 3430584,
-      "step": 213
-    },
-    {
-      "epoch": 0.014990324586058647,
-      "grad_norm": 4.728682518005371,
-      "learning_rate": 9.850237478108582e-05,
-      "loss": 1.2686,
-      "num_input_tokens_seen": 3446968,
-      "step": 214
-    },
-    {
-      "epoch": 0.015060372831787894,
-      "grad_norm": 4.291411876678467,
-      "learning_rate": 9.84953765323993e-05,
-      "loss": 1.1877,
-      "num_input_tokens_seen": 3462568,
-      "step": 215
-    },
-    {
-      "epoch": 0.01513042107751714,
-      "grad_norm": 4.405060768127441,
-      "learning_rate": 9.84883782837128e-05,
-      "loss": 1.2873,
-      "num_input_tokens_seen": 3478952,
-      "step": 216
-    },
-    {
-      "epoch": 0.015200469323246386,
-      "grad_norm": 4.254365921020508,
-      "learning_rate": 9.848138003502628e-05,
-      "loss": 1.1062,
-      "num_input_tokens_seen": 3495304,
-      "step": 217
-    },
-    {
-      "epoch": 0.015270517568975631,
-      "grad_norm": 4.741672039031982,
-      "learning_rate": 9.847438178633976e-05,
-      "loss": 1.1983,
-      "num_input_tokens_seen": 3511688,
-      "step": 218
-    },
-    {
-      "epoch": 0.015340565814704879,
-      "grad_norm": 4.352742671966553,
-      "learning_rate": 9.846738353765325e-05,
-      "loss": 1.2028,
-      "num_input_tokens_seen": 3528072,
-      "step": 219
-    },
-    {
-      "epoch": 0.015410614060434124,
-      "grad_norm": 4.996603488922119,
-      "learning_rate": 9.846038528896672e-05,
-      "loss": 1.1561,
-      "num_input_tokens_seen": 3542904,
-      "step": 220
-    },
-    {
-      "epoch": 0.01548066230616337,
-      "grad_norm": 4.911815166473389,
-      "learning_rate": 9.845338704028021e-05,
-      "loss": 1.3375,
-      "num_input_tokens_seen": 3558352,
-      "step": 221
-    },
-    {
-      "epoch": 0.015550710551892616,
-      "grad_norm": 4.638419151306152,
-      "learning_rate": 9.84463887915937e-05,
-      "loss": 1.1963,
-      "num_input_tokens_seen": 3574736,
-      "step": 222
-    },
-    {
-      "epoch": 0.015620758797621863,
-      "grad_norm": 4.323521614074707,
-      "learning_rate": 9.843939054290719e-05,
-      "loss": 1.1224,
-      "num_input_tokens_seen": 3591120,
-      "step": 223
-    },
-    {
-      "epoch": 0.01569080704335111,
-      "grad_norm": 4.466544151306152,
-      "learning_rate": 9.843239229422068e-05,
-      "loss": 1.3988,
-      "num_input_tokens_seen": 3607392,
-      "step": 224
-    },
-    {
-      "epoch": 0.015760855289080354,
-      "grad_norm": 4.476973533630371,
-      "learning_rate": 9.842539404553415e-05,
-      "loss": 1.184,
-      "num_input_tokens_seen": 3623776,
-      "step": 225
-    },
-    {
-      "epoch": 0.0158309035348096,
-      "grad_norm": 4.648625373840332,
-      "learning_rate": 9.841839579684764e-05,
-      "loss": 1.1768,
-      "num_input_tokens_seen": 3640008,
-      "step": 226
-    },
-    {
-      "epoch": 0.015900951780538845,
-      "grad_norm": 4.364476203918457,
-      "learning_rate": 9.841139754816112e-05,
-      "loss": 1.0208,
-      "num_input_tokens_seen": 3656392,
-      "step": 227
-    },
-    {
-      "epoch": 0.01597100002626809,
-      "grad_norm": 4.3054633140563965,
-      "learning_rate": 9.84043992994746e-05,
-      "loss": 1.1215,
-      "num_input_tokens_seen": 3672392,
-      "step": 228
-    },
-    {
-      "epoch": 0.016041048271997337,
-      "grad_norm": 4.83436918258667,
-      "learning_rate": 9.83974010507881e-05,
-      "loss": 1.2284,
-      "num_input_tokens_seen": 3688776,
-      "step": 229
-    },
-    {
-      "epoch": 0.016111096517726586,
-      "grad_norm": 4.447519779205322,
-      "learning_rate": 9.839040280210158e-05,
-      "loss": 1.1765,
-      "num_input_tokens_seen": 3705080,
-      "step": 230
-    },
-    {
-      "epoch": 0.01618114476345583,
-      "grad_norm": 4.269217491149902,
-      "learning_rate": 9.838340455341507e-05,
-      "loss": 1.0466,
-      "num_input_tokens_seen": 3721464,
-      "step": 231
-    },
-    {
-      "epoch": 0.016251193009185077,
-      "grad_norm": 4.41223669052124,
-      "learning_rate": 9.837640630472854e-05,
-      "loss": 1.2098,
-      "num_input_tokens_seen": 3737184,
-      "step": 232
-    },
-    {
-      "epoch": 0.016321241254914323,
-      "grad_norm": 4.632737159729004,
-      "learning_rate": 9.836940805604203e-05,
-      "loss": 1.1562,
-      "num_input_tokens_seen": 3753192,
-      "step": 233
-    },
-    {
-      "epoch": 0.016391289500643568,
-      "grad_norm": 4.379425525665283,
-      "learning_rate": 9.836240980735552e-05,
-      "loss": 1.1219,
-      "num_input_tokens_seen": 3767976,
-      "step": 234
-    },
-    {
-      "epoch": 0.016461337746372814,
-      "grad_norm": 4.28551721572876,
-      "learning_rate": 9.835541155866901e-05,
-      "loss": 1.0259,
-      "num_input_tokens_seen": 3784008,
-      "step": 235
-    },
-    {
-      "epoch": 0.01653138599210206,
-      "grad_norm": 4.642453670501709,
-      "learning_rate": 9.83484133099825e-05,
-      "loss": 1.1684,
-      "num_input_tokens_seen": 3800000,
-      "step": 236
-    },
-    {
-      "epoch": 0.016601434237831305,
-      "grad_norm": 4.367178440093994,
-      "learning_rate": 9.834141506129597e-05,
-      "loss": 1.2877,
-      "num_input_tokens_seen": 3816384,
-      "step": 237
-    },
-    {
-      "epoch": 0.01667148248356055,
-      "grad_norm": 4.5724005699157715,
-      "learning_rate": 9.833441681260946e-05,
-      "loss": 1.1814,
-      "num_input_tokens_seen": 3830328,
-      "step": 238
-    },
-    {
-      "epoch": 0.0167415307292898,
-      "grad_norm": 4.318159580230713,
-      "learning_rate": 9.832741856392295e-05,
-      "loss": 1.1143,
-      "num_input_tokens_seen": 3846712,
-      "step": 239
-    },
-    {
-      "epoch": 0.016811578975019045,
-      "grad_norm": 4.408501625061035,
-      "learning_rate": 9.832042031523643e-05,
-      "loss": 1.1508,
-      "num_input_tokens_seen": 3861776,
-      "step": 240
-    },
-    {
-      "epoch": 0.01688162722074829,
-      "grad_norm": 4.20060920715332,
-      "learning_rate": 9.831342206654991e-05,
-      "loss": 1.209,
-      "num_input_tokens_seen": 3877736,
-      "step": 241
-    },
-    {
-      "epoch": 0.016951675466477537,
-      "grad_norm": 4.431649208068848,
-      "learning_rate": 9.83064238178634e-05,
-      "loss": 1.2458,
-      "num_input_tokens_seen": 3893320,
-      "step": 242
-    },
-    {
-      "epoch": 0.017021723712206782,
-      "grad_norm": 4.000490188598633,
-      "learning_rate": 9.829942556917689e-05,
-      "loss": 1.0274,
-      "num_input_tokens_seen": 3909704,
-      "step": 243
-    },
-    {
-      "epoch": 0.017091771957936028,
-      "grad_norm": 4.703495025634766,
-      "learning_rate": 9.829242732049038e-05,
-      "loss": 1.1711,
-      "num_input_tokens_seen": 3925808,
-      "step": 244
-    },
-    {
-      "epoch": 0.017161820203665273,
-      "grad_norm": 4.639338970184326,
-      "learning_rate": 9.828542907180386e-05,
-      "loss": 1.3046,
-      "num_input_tokens_seen": 3942192,
-      "step": 245
-    },
-    {
-      "epoch": 0.01723186844939452,
-      "grad_norm": 4.414276599884033,
-      "learning_rate": 9.827843082311734e-05,
-      "loss": 1.271,
-      "num_input_tokens_seen": 3958528,
-      "step": 246
-    },
-    {
-      "epoch": 0.017301916695123768,
-      "grad_norm": 4.404853820800781,
-      "learning_rate": 9.827143257443082e-05,
-      "loss": 1.0693,
-      "num_input_tokens_seen": 3974912,
-      "step": 247
-    },
-    {
-      "epoch": 0.017371964940853014,
-      "grad_norm": 4.519491195678711,
-      "learning_rate": 9.826443432574431e-05,
-      "loss": 1.2894,
-      "num_input_tokens_seen": 3991296,
-      "step": 248
-    },
-    {
-      "epoch": 0.01744201318658226,
-      "grad_norm": 4.261727809906006,
-      "learning_rate": 9.825743607705781e-05,
-      "loss": 1.2059,
-      "num_input_tokens_seen": 4006544,
-      "step": 249
-    },
-    {
-      "epoch": 0.017512061432311505,
-      "grad_norm": 4.102485656738281,
-      "learning_rate": 9.825043782837129e-05,
-      "loss": 0.9365,
-      "num_input_tokens_seen": 4022320,
-      "step": 250
-    },
-    {
-      "epoch": 0.01758210967804075,
-      "grad_norm": 4.804764270782471,
-      "learning_rate": 9.824343957968477e-05,
-      "loss": 1.3344,
-      "num_input_tokens_seen": 4037048,
-      "step": 251
-    },
-    {
-      "epoch": 0.017652157923769996,
-      "grad_norm": 4.130600452423096,
-      "learning_rate": 9.823644133099825e-05,
-      "loss": 1.2349,
-      "num_input_tokens_seen": 4053432,
-      "step": 252
-    },
-    {
-      "epoch": 0.017722206169499242,
-      "grad_norm": 4.234742641448975,
-      "learning_rate": 9.822944308231174e-05,
-      "loss": 1.1371,
-      "num_input_tokens_seen": 4069816,
-      "step": 253
-    },
-    {
-      "epoch": 0.017792254415228487,
-      "grad_norm": 4.754928112030029,
-      "learning_rate": 9.822244483362521e-05,
-      "loss": 1.5168,
-      "num_input_tokens_seen": 4085864,
-      "step": 254
-    },
-    {
-      "epoch": 0.017862302660957733,
-      "grad_norm": 4.542768478393555,
-      "learning_rate": 9.821544658493871e-05,
-      "loss": 1.1943,
-      "num_input_tokens_seen": 4102240,
-      "step": 255
-    },
-    {
-      "epoch": 0.017932350906686982,
-      "grad_norm": 4.411310195922852,
-      "learning_rate": 9.82084483362522e-05,
-      "loss": 1.2694,
-      "num_input_tokens_seen": 4118544,
-      "step": 256
-    },
-    {
-      "epoch": 0.018002399152416228,
-      "grad_norm": 4.205377101898193,
-      "learning_rate": 9.820145008756568e-05,
-      "loss": 1.1581,
-      "num_input_tokens_seen": 4134928,
-      "step": 257
-    },
-    {
-      "epoch": 0.018072447398145473,
-      "grad_norm": 4.451165199279785,
-      "learning_rate": 9.819445183887917e-05,
-      "loss": 1.089,
-      "num_input_tokens_seen": 4150848,
-      "step": 258
-    },
-    {
-      "epoch": 0.01814249564387472,
-      "grad_norm": 4.366336822509766,
-      "learning_rate": 9.818745359019264e-05,
-      "loss": 1.1767,
-      "num_input_tokens_seen": 4167184,
-      "step": 259
-    },
-    {
-      "epoch": 0.018212543889603965,
-      "grad_norm": 4.394649982452393,
-      "learning_rate": 9.818045534150613e-05,
-      "loss": 1.0741,
-      "num_input_tokens_seen": 4183376,
-      "step": 260
-    },
-    {
-      "epoch": 0.01828259213533321,
-      "grad_norm": 4.344518184661865,
-      "learning_rate": 9.817345709281962e-05,
-      "loss": 1.2282,
-      "num_input_tokens_seen": 4199760,
-      "step": 261
-    },
-    {
-      "epoch": 0.018352640381062456,
-      "grad_norm": 4.403041362762451,
-      "learning_rate": 9.816645884413311e-05,
-      "loss": 1.2317,
-      "num_input_tokens_seen": 4215816,
-      "step": 262
-    },
-    {
-      "epoch": 0.0184226886267917,
-      "grad_norm": 4.715320110321045,
-      "learning_rate": 9.81594605954466e-05,
-      "loss": 1.3074,
-      "num_input_tokens_seen": 4231504,
-      "step": 263
-    },
-    {
-      "epoch": 0.01849273687252095,
-      "grad_norm": 4.5754265785217285,
-      "learning_rate": 9.815246234676007e-05,
-      "loss": 1.253,
-      "num_input_tokens_seen": 4247888,
-      "step": 264
-    },
-    {
-      "epoch": 0.018562785118250196,
-      "grad_norm": 4.2346930503845215,
-      "learning_rate": 9.814546409807356e-05,
-      "loss": 1.1727,
-      "num_input_tokens_seen": 4264248,
-      "step": 265
-    },
-    {
-      "epoch": 0.018632833363979442,
-      "grad_norm": 4.186713218688965,
-      "learning_rate": 9.813846584938705e-05,
-      "loss": 1.2693,
-      "num_input_tokens_seen": 4280632,
-      "step": 266
-    },
-    {
-      "epoch": 0.018702881609708687,
-      "grad_norm": 4.6356706619262695,
-      "learning_rate": 9.813146760070052e-05,
-      "loss": 1.3755,
-      "num_input_tokens_seen": 4296648,
-      "step": 267
-    },
-    {
-      "epoch": 0.018772929855437933,
-      "grad_norm": 4.466466903686523,
-      "learning_rate": 9.812446935201401e-05,
-      "loss": 1.283,
-      "num_input_tokens_seen": 4311408,
-      "step": 268
-    },
-    {
-      "epoch": 0.01884297810116718,
-      "grad_norm": 4.3369140625,
-      "learning_rate": 9.81174711033275e-05,
-      "loss": 1.1555,
-      "num_input_tokens_seen": 4326736,
-      "step": 269
-    },
-    {
-      "epoch": 0.018913026346896424,
-      "grad_norm": 4.434782028198242,
-      "learning_rate": 9.811047285464099e-05,
-      "loss": 1.2859,
-      "num_input_tokens_seen": 4343120,
-      "step": 270
-    },
-    {
-      "epoch": 0.01898307459262567,
-      "grad_norm": 4.346708297729492,
-      "learning_rate": 9.810347460595448e-05,
-      "loss": 1.1421,
-      "num_input_tokens_seen": 4359504,
-      "step": 271
-    },
-    {
-      "epoch": 0.019053122838354915,
-      "grad_norm": 4.529878616333008,
-      "learning_rate": 9.809647635726795e-05,
-      "loss": 1.2654,
-      "num_input_tokens_seen": 4375888,
-      "step": 272
-    },
-    {
-      "epoch": 0.019123171084084165,
-      "grad_norm": 4.051745891571045,
-      "learning_rate": 9.808947810858144e-05,
-      "loss": 1.1469,
-      "num_input_tokens_seen": 4392224,
-      "step": 273
-    },
-    {
-      "epoch": 0.01919321932981341,
-      "grad_norm": 4.403522491455078,
-      "learning_rate": 9.808247985989492e-05,
-      "loss": 1.233,
-      "num_input_tokens_seen": 4408608,
-      "step": 274
-    },
-    {
-      "epoch": 0.019263267575542656,
-      "grad_norm": 4.166261196136475,
-      "learning_rate": 9.807548161120842e-05,
-      "loss": 1.1697,
-      "num_input_tokens_seen": 4424992,
-      "step": 275
-    },
-    {
-      "epoch": 0.0193333158212719,
-      "grad_norm": 4.29187536239624,
-      "learning_rate": 9.806848336252191e-05,
-      "loss": 1.0503,
-      "num_input_tokens_seen": 4441376,
-      "step": 276
-    },
-    {
-      "epoch": 0.019403364067001147,
-      "grad_norm": 4.4056172370910645,
-      "learning_rate": 9.806148511383538e-05,
-      "loss": 1.1965,
-      "num_input_tokens_seen": 4457760,
-      "step": 277
-    },
-    {
-      "epoch": 0.019473412312730393,
-      "grad_norm": 4.355875015258789,
-      "learning_rate": 9.805448686514887e-05,
-      "loss": 1.1024,
-      "num_input_tokens_seen": 4474144,
-      "step": 278
-    },
-    {
-      "epoch": 0.019543460558459638,
-      "grad_norm": 4.46420955657959,
-      "learning_rate": 9.804748861646235e-05,
-      "loss": 1.203,
-      "num_input_tokens_seen": 4488912,
-      "step": 279
-    },
-    {
-      "epoch": 0.019613508804188884,
-      "grad_norm": 4.48052453994751,
-      "learning_rate": 9.804049036777583e-05,
-      "loss": 1.2089,
-      "num_input_tokens_seen": 4505296,
-      "step": 280
-    },
-    {
-      "epoch": 0.01968355704991813,
-      "grad_norm": 4.458749294281006,
-      "learning_rate": 9.803349211908932e-05,
-      "loss": 1.1557,
-      "num_input_tokens_seen": 4520576,
-      "step": 281
-    },
-    {
-      "epoch": 0.01975360529564738,
-      "grad_norm": 4.551771640777588,
-      "learning_rate": 9.802649387040281e-05,
-      "loss": 1.1671,
-      "num_input_tokens_seen": 4536960,
-      "step": 282
-    },
-    {
-      "epoch": 0.019823653541376624,
-      "grad_norm": 4.038064956665039,
-      "learning_rate": 9.80194956217163e-05,
-      "loss": 1.1562,
-      "num_input_tokens_seen": 4553344,
-      "step": 283
-    },
-    {
-      "epoch": 0.01989370178710587,
-      "grad_norm": 4.647075653076172,
-      "learning_rate": 9.801249737302978e-05,
-      "loss": 1.3069,
-      "num_input_tokens_seen": 4568928,
-      "step": 284
-    },
-    {
-      "epoch": 0.019963750032835115,
-      "grad_norm": 4.258941650390625,
-      "learning_rate": 9.800549912434326e-05,
-      "loss": 1.0349,
-      "num_input_tokens_seen": 4585312,
-      "step": 285
-    },
-    {
-      "epoch": 0.02003379827856436,
-      "grad_norm": 4.348769664764404,
-      "learning_rate": 9.799850087565674e-05,
-      "loss": 1.1163,
-      "num_input_tokens_seen": 4601696,
-      "step": 286
-    },
-    {
-      "epoch": 0.020103846524293607,
-      "grad_norm": 4.105901718139648,
-      "learning_rate": 9.799150262697023e-05,
-      "loss": 1.0313,
-      "num_input_tokens_seen": 4617312,
-      "step": 287
-    },
-    {
-      "epoch": 0.020173894770022852,
-      "grad_norm": 4.079495429992676,
-      "learning_rate": 9.798450437828372e-05,
-      "loss": 1.0828,
-      "num_input_tokens_seen": 4633696,
-      "step": 288
-    },
-    {
-      "epoch": 0.020243943015752098,
-      "grad_norm": 4.03472375869751,
-      "learning_rate": 9.79775061295972e-05,
-      "loss": 0.9475,
-      "num_input_tokens_seen": 4650080,
-      "step": 289
-    },
-    {
-      "epoch": 0.020313991261481347,
-      "grad_norm": 4.077049732208252,
-      "learning_rate": 9.797050788091069e-05,
-      "loss": 1.1323,
-      "num_input_tokens_seen": 4666328,
-      "step": 290
-    },
-    {
-      "epoch": 0.020384039507210593,
-      "grad_norm": 4.086606025695801,
-      "learning_rate": 9.796350963222417e-05,
-      "loss": 1.1218,
-      "num_input_tokens_seen": 4682256,
-      "step": 291
-    },
-    {
-      "epoch": 0.020454087752939838,
-      "grad_norm": 4.296900749206543,
-      "learning_rate": 9.795651138353766e-05,
-      "loss": 1.2964,
-      "num_input_tokens_seen": 4698640,
-      "step": 292
-    },
-    {
-      "epoch": 0.020524135998669084,
-      "grad_norm": 4.040759086608887,
-      "learning_rate": 9.794951313485115e-05,
-      "loss": 1.1077,
-      "num_input_tokens_seen": 4714928,
-      "step": 293
-    },
-    {
-      "epoch": 0.02059418424439833,
-      "grad_norm": 3.8260273933410645,
-      "learning_rate": 9.794251488616462e-05,
-      "loss": 0.9667,
-      "num_input_tokens_seen": 4731312,
-      "step": 294
-    },
-    {
-      "epoch": 0.020664232490127575,
-      "grad_norm": 4.294517993927002,
-      "learning_rate": 9.793551663747811e-05,
-      "loss": 1.2704,
-      "num_input_tokens_seen": 4747544,
-      "step": 295
-    },
-    {
-      "epoch": 0.02073428073585682,
-      "grad_norm": 4.206037521362305,
-      "learning_rate": 9.79285183887916e-05,
-      "loss": 1.1593,
-      "num_input_tokens_seen": 4763928,
-      "step": 296
-    },
-    {
-      "epoch": 0.020804328981586066,
-      "grad_norm": 4.147867202758789,
-      "learning_rate": 9.792152014010509e-05,
-      "loss": 1.1256,
-      "num_input_tokens_seen": 4780312,
-      "step": 297
-    },
-    {
-      "epoch": 0.020874377227315312,
-      "grad_norm": 4.23718786239624,
-      "learning_rate": 9.791452189141857e-05,
-      "loss": 1.2353,
-      "num_input_tokens_seen": 4796384,
-      "step": 298
-    },
-    {
-      "epoch": 0.02094442547304456,
-      "grad_norm": 4.172685146331787,
-      "learning_rate": 9.790752364273205e-05,
-      "loss": 1.1868,
-      "num_input_tokens_seen": 4812768,
-      "step": 299
-    },
-    {
-      "epoch": 0.021014473718773807,
-      "grad_norm": 4.167289733886719,
-      "learning_rate": 9.790052539404554e-05,
-      "loss": 1.0606,
-      "num_input_tokens_seen": 4829152,
-      "step": 300
-    },
-    {
-      "epoch": 0.021084521964503052,
-      "grad_norm": 4.096963882446289,
-      "learning_rate": 9.789352714535903e-05,
-      "loss": 1.0557,
-      "num_input_tokens_seen": 4845384,
-      "step": 301
-    },
-    {
-      "epoch": 0.021154570210232298,
-      "grad_norm": 4.223779678344727,
-      "learning_rate": 9.788652889667252e-05,
-      "loss": 1.1485,
-      "num_input_tokens_seen": 4861768,
-      "step": 302
-    },
-    {
-      "epoch": 0.021224618455961543,
-      "grad_norm": 3.8243472576141357,
-      "learning_rate": 9.7879530647986e-05,
-      "loss": 1.004,
-      "num_input_tokens_seen": 4878152,
-      "step": 303
-    },
-    {
-      "epoch": 0.02129466670169079,
-      "grad_norm": 4.092590808868408,
-      "learning_rate": 9.787253239929948e-05,
-      "loss": 1.0211,
-      "num_input_tokens_seen": 4894536,
-      "step": 304
-    },
-    {
-      "epoch": 0.021364714947420035,
-      "grad_norm": 4.42412805557251,
-      "learning_rate": 9.786553415061297e-05,
-      "loss": 0.9915,
-      "num_input_tokens_seen": 4910320,
-      "step": 305
-    },
-    {
-      "epoch": 0.02143476319314928,
-      "grad_norm": 4.488316535949707,
-      "learning_rate": 9.785853590192644e-05,
-      "loss": 1.1782,
-      "num_input_tokens_seen": 4926704,
-      "step": 306
-    },
-    {
-      "epoch": 0.021504811438878526,
-      "grad_norm": 4.110256195068359,
-      "learning_rate": 9.785153765323993e-05,
-      "loss": 1.102,
-      "num_input_tokens_seen": 4943088,
-      "step": 307
-    },
-    {
-      "epoch": 0.021574859684607775,
-      "grad_norm": 4.246950149536133,
-      "learning_rate": 9.784453940455342e-05,
-      "loss": 1.067,
-      "num_input_tokens_seen": 4958736,
-      "step": 308
-    },
-    {
-      "epoch": 0.02164490793033702,
-      "grad_norm": 4.175214767456055,
-      "learning_rate": 9.783754115586691e-05,
-      "loss": 1.0638,
-      "num_input_tokens_seen": 4975120,
-      "step": 309
-    },
-    {
-      "epoch": 0.021714956176066266,
-      "grad_norm": 4.427795886993408,
-      "learning_rate": 9.78305429071804e-05,
-      "loss": 1.1347,
-      "num_input_tokens_seen": 4991504,
-      "step": 310
-    },
-    {
-      "epoch": 0.021785004421795512,
-      "grad_norm": 4.158191204071045,
-      "learning_rate": 9.782354465849387e-05,
-      "loss": 1.1662,
-      "num_input_tokens_seen": 5007152,
-      "step": 311
-    },
-    {
-      "epoch": 0.021855052667524758,
-      "grad_norm": 4.184347629547119,
-      "learning_rate": 9.781654640980736e-05,
-      "loss": 1.0791,
-      "num_input_tokens_seen": 5023536,
-      "step": 312
-    },
-    {
-      "epoch": 0.021925100913254003,
-      "grad_norm": 3.8506295680999756,
-      "learning_rate": 9.780954816112084e-05,
-      "loss": 1.0615,
-      "num_input_tokens_seen": 5039728,
-      "step": 313
-    },
-    {
-      "epoch": 0.02199514915898325,
-      "grad_norm": 4.310062408447266,
-      "learning_rate": 9.780254991243432e-05,
-      "loss": 1.1363,
-      "num_input_tokens_seen": 5056008,
-      "step": 314
-    },
-    {
-      "epoch": 0.022065197404712494,
-      "grad_norm": 4.215006351470947,
-      "learning_rate": 9.779555166374781e-05,
-      "loss": 1.1715,
-      "num_input_tokens_seen": 5072096,
-      "step": 315
-    },
-    {
-      "epoch": 0.022135245650441743,
-      "grad_norm": 4.219073295593262,
-      "learning_rate": 9.77885534150613e-05,
-      "loss": 1.219,
-      "num_input_tokens_seen": 5088432,
-      "step": 316
-    },
-    {
-      "epoch": 0.02220529389617099,
-      "grad_norm": 4.319522857666016,
-      "learning_rate": 9.778155516637479e-05,
-      "loss": 1.3085,
-      "num_input_tokens_seen": 5104240,
-      "step": 317
-    },
-    {
-      "epoch": 0.022275342141900235,
-      "grad_norm": 4.118961334228516,
-      "learning_rate": 9.777455691768827e-05,
-      "loss": 1.0926,
-      "num_input_tokens_seen": 5120624,
-      "step": 318
-    },
-    {
-      "epoch": 0.02234539038762948,
-      "grad_norm": 4.195051193237305,
-      "learning_rate": 9.776755866900175e-05,
-      "loss": 1.0894,
-      "num_input_tokens_seen": 5137008,
-      "step": 319
-    },
-    {
-      "epoch": 0.022415438633358726,
-      "grad_norm": 4.114197254180908,
-      "learning_rate": 9.776056042031524e-05,
-      "loss": 1.1897,
-      "num_input_tokens_seen": 5153272,
-      "step": 320
-    },
-    {
-      "epoch": 0.02248548687908797,
-      "grad_norm": 4.014908313751221,
-      "learning_rate": 9.775356217162872e-05,
-      "loss": 1.0932,
-      "num_input_tokens_seen": 5169472,
-      "step": 321
-    },
-    {
-      "epoch": 0.022555535124817217,
-      "grad_norm": 4.190642356872559,
-      "learning_rate": 9.774656392294222e-05,
-      "loss": 1.1413,
-      "num_input_tokens_seen": 5185856,
-      "step": 322
-    },
-    {
-      "epoch": 0.022625583370546463,
-      "grad_norm": 4.562993049621582,
-      "learning_rate": 9.77395656742557e-05,
-      "loss": 1.2865,
-      "num_input_tokens_seen": 5202240,
-      "step": 323
-    },
-    {
-      "epoch": 0.02269563161627571,
-      "grad_norm": 4.607022762298584,
-      "learning_rate": 9.773256742556918e-05,
-      "loss": 1.1465,
-      "num_input_tokens_seen": 5218168,
-      "step": 324
-    },
-    {
-      "epoch": 0.022765679862004957,
-      "grad_norm": 3.956439256668091,
-      "learning_rate": 9.772556917688267e-05,
-      "loss": 1.028,
-      "num_input_tokens_seen": 5234368,
-      "step": 325
-    },
-    {
-      "epoch": 0.022835728107734203,
-      "grad_norm": 4.20713472366333,
-      "learning_rate": 9.771857092819615e-05,
-      "loss": 1.2332,
-      "num_input_tokens_seen": 5249808,
-      "step": 326
-    },
-    {
-      "epoch": 0.02290577635346345,
-      "grad_norm": 4.4092864990234375,
-      "learning_rate": 9.771157267950964e-05,
-      "loss": 1.104,
-      "num_input_tokens_seen": 5266120,
-      "step": 327
-    },
-    {
-      "epoch": 0.022975824599192694,
-      "grad_norm": 4.529845237731934,
-      "learning_rate": 9.770457443082312e-05,
-      "loss": 1.3475,
-      "num_input_tokens_seen": 5282504,
-      "step": 328
-    },
-    {
-      "epoch": 0.02304587284492194,
-      "grad_norm": 4.221986293792725,
-      "learning_rate": 9.769757618213661e-05,
-      "loss": 1.4115,
-      "num_input_tokens_seen": 5298344,
-      "step": 329
-    },
-    {
-      "epoch": 0.023115921090651186,
-      "grad_norm": 4.29000186920166,
-      "learning_rate": 9.76905779334501e-05,
-      "loss": 1.2855,
-      "num_input_tokens_seen": 5314728,
-      "step": 330
-    },
-    {
-      "epoch": 0.02318596933638043,
-      "grad_norm": 4.426812648773193,
-      "learning_rate": 9.768357968476358e-05,
-      "loss": 1.514,
-      "num_input_tokens_seen": 5330816,
-      "step": 331
-    },
-    {
-      "epoch": 0.023256017582109677,
-      "grad_norm": 4.210752964019775,
-      "learning_rate": 9.767658143607706e-05,
-      "loss": 1.0854,
-      "num_input_tokens_seen": 5346552,
-      "step": 332
-    },
-    {
-      "epoch": 0.023326065827838922,
-      "grad_norm": 4.216427326202393,
-      "learning_rate": 9.766958318739054e-05,
-      "loss": 1.1573,
-      "num_input_tokens_seen": 5362936,
-      "step": 333
-    },
-    {
-      "epoch": 0.02339611407356817,
-      "grad_norm": 4.132325649261475,
-      "learning_rate": 9.766258493870403e-05,
-      "loss": 1.0942,
-      "num_input_tokens_seen": 5379320,
-      "step": 334
-    },
-    {
-      "epoch": 0.023466162319297417,
-      "grad_norm": 4.277027130126953,
-      "learning_rate": 9.765558669001752e-05,
-      "loss": 1.1227,
-      "num_input_tokens_seen": 5395704,
-      "step": 335
-    },
-    {
-      "epoch": 0.023536210565026663,
-      "grad_norm": 4.228096961975098,
-      "learning_rate": 9.7648588441331e-05,
-      "loss": 1.1094,
-      "num_input_tokens_seen": 5412088,
-      "step": 336
-    },
-    {
-      "epoch": 0.02360625881075591,
-      "grad_norm": 4.194522380828857,
-      "learning_rate": 9.76415901926445e-05,
-      "loss": 1.2066,
-      "num_input_tokens_seen": 5428472,
-      "step": 337
-    },
-    {
-      "epoch": 0.023676307056485154,
-      "grad_norm": 4.336326599121094,
-      "learning_rate": 9.763459194395797e-05,
-      "loss": 1.2251,
-      "num_input_tokens_seen": 5444856,
-      "step": 338
-    },
-    {
-      "epoch": 0.0237463553022144,
-      "grad_norm": 4.2723307609558105,
-      "learning_rate": 9.762759369527146e-05,
-      "loss": 1.0927,
-      "num_input_tokens_seen": 5460304,
-      "step": 339
-    },
-    {
-      "epoch": 0.023816403547943645,
-      "grad_norm": 4.190036773681641,
-      "learning_rate": 9.762059544658493e-05,
-      "loss": 1.2036,
-      "num_input_tokens_seen": 5476688,
-      "step": 340
-    },
-    {
-      "epoch": 0.02388645179367289,
-      "grad_norm": 4.477560043334961,
-      "learning_rate": 9.761359719789842e-05,
-      "loss": 1.362,
-      "num_input_tokens_seen": 5493072,
-      "step": 341
-    },
-    {
-      "epoch": 0.02395650003940214,
-      "grad_norm": 4.160232067108154,
-      "learning_rate": 9.760659894921192e-05,
-      "loss": 1.1602,
-      "num_input_tokens_seen": 5509456,
-      "step": 342
-    },
-    {
-      "epoch": 0.024026548285131386,
-      "grad_norm": 3.857335090637207,
-      "learning_rate": 9.75996007005254e-05,
-      "loss": 1.0963,
-      "num_input_tokens_seen": 5525840,
-      "step": 343
-    },
-    {
-      "epoch": 0.02409659653086063,
-      "grad_norm": 4.141246318817139,
-      "learning_rate": 9.759260245183889e-05,
-      "loss": 1.2009,
-      "num_input_tokens_seen": 5541888,
-      "step": 344
-    },
-    {
-      "epoch": 0.024166644776589877,
-      "grad_norm": 4.50364875793457,
-      "learning_rate": 9.758560420315236e-05,
-      "loss": 1.1483,
-      "num_input_tokens_seen": 5557848,
-      "step": 345
-    },
-    {
-      "epoch": 0.024236693022319122,
-      "grad_norm": 4.3343353271484375,
-      "learning_rate": 9.757860595446585e-05,
-      "loss": 1.3594,
-      "num_input_tokens_seen": 5573504,
-      "step": 346
-    },
-    {
-      "epoch": 0.024306741268048368,
-      "grad_norm": 4.050408363342285,
-      "learning_rate": 9.757160770577934e-05,
-      "loss": 1.0563,
-      "num_input_tokens_seen": 5589544,
-      "step": 347
-    },
-    {
-      "epoch": 0.024376789513777614,
-      "grad_norm": 4.051811695098877,
-      "learning_rate": 9.756460945709283e-05,
-      "loss": 1.0288,
-      "num_input_tokens_seen": 5605368,
-      "step": 348
-    },
-    {
-      "epoch": 0.02444683775950686,
-      "grad_norm": 4.365113258361816,
-      "learning_rate": 9.755761120840632e-05,
-      "loss": 1.3054,
-      "num_input_tokens_seen": 5621752,
-      "step": 349
-    },
-    {
-      "epoch": 0.024516886005236105,
-      "grad_norm": 4.0057501792907715,
-      "learning_rate": 9.755061295971979e-05,
-      "loss": 1.1302,
-      "num_input_tokens_seen": 5638136,
-      "step": 350
-    },
-    {
-      "epoch": 0.024586934250965354,
-      "grad_norm": 4.254896640777588,
-      "learning_rate": 9.754361471103328e-05,
-      "loss": 1.0495,
-      "num_input_tokens_seen": 5653168,
-      "step": 351
-    },
-    {
-      "epoch": 0.0246569824966946,
-      "grad_norm": 3.8119771480560303,
-      "learning_rate": 9.753661646234677e-05,
-      "loss": 1.0349,
-      "num_input_tokens_seen": 5669504,
-      "step": 352
-    },
-    {
-      "epoch": 0.024727030742423845,
-      "grad_norm": 4.5082621574401855,
-      "learning_rate": 9.752961821366024e-05,
-      "loss": 1.2537,
-      "num_input_tokens_seen": 5685168,
-      "step": 353
-    },
-    {
-      "epoch": 0.02479707898815309,
-      "grad_norm": 4.392731189727783,
-      "learning_rate": 9.752261996497373e-05,
-      "loss": 1.2534,
-      "num_input_tokens_seen": 5701240,
-      "step": 354
-    },
-    {
-      "epoch": 0.024867127233882336,
-      "grad_norm": 4.293395519256592,
-      "learning_rate": 9.751562171628722e-05,
-      "loss": 1.2774,
-      "num_input_tokens_seen": 5717624,
-      "step": 355
-    },
-    {
-      "epoch": 0.024937175479611582,
-      "grad_norm": 4.64813756942749,
-      "learning_rate": 9.750862346760071e-05,
-      "loss": 1.2795,
-      "num_input_tokens_seen": 5733104,
-      "step": 356
-    },
-    {
-      "epoch": 0.025007223725340828,
-      "grad_norm": 4.5166778564453125,
-      "learning_rate": 9.75016252189142e-05,
-      "loss": 1.1301,
-      "num_input_tokens_seen": 5749488,
-      "step": 357
-    },
-    {
-      "epoch": 0.025077271971070073,
-      "grad_norm": 3.894291400909424,
-      "learning_rate": 9.749462697022767e-05,
-      "loss": 0.901,
-      "num_input_tokens_seen": 5765872,
-      "step": 358
-    },
-    {
-      "epoch": 0.02514732021679932,
-      "grad_norm": 4.10056209564209,
-      "learning_rate": 9.748762872154116e-05,
-      "loss": 1.0529,
-      "num_input_tokens_seen": 5780856,
-      "step": 359
-    },
-    {
-      "epoch": 0.025217368462528568,
-      "grad_norm": 4.6277666091918945,
-      "learning_rate": 9.748063047285464e-05,
-      "loss": 1.3649,
-      "num_input_tokens_seen": 5796856,
-      "step": 360
-    },
-    {
-      "epoch": 0.025287416708257814,
-      "grad_norm": 4.029720306396484,
-      "learning_rate": 9.747363222416813e-05,
-      "loss": 0.8863,
-      "num_input_tokens_seen": 5812176,
-      "step": 361
-    },
-    {
-      "epoch": 0.02535746495398706,
-      "grad_norm": 3.7772202491760254,
-      "learning_rate": 9.746663397548161e-05,
-      "loss": 1.0448,
-      "num_input_tokens_seen": 5828064,
-      "step": 362
-    },
-    {
-      "epoch": 0.025427513199716305,
-      "grad_norm": 4.379861354827881,
-      "learning_rate": 9.74596357267951e-05,
-      "loss": 1.3274,
-      "num_input_tokens_seen": 5843680,
-      "step": 363
-    },
-    {
-      "epoch": 0.02549756144544555,
-      "grad_norm": 4.254587173461914,
-      "learning_rate": 9.745263747810859e-05,
-      "loss": 1.1502,
-      "num_input_tokens_seen": 5859024,
-      "step": 364
-    },
-    {
-      "epoch": 0.025567609691174796,
-      "grad_norm": 4.271276473999023,
-      "learning_rate": 9.744563922942207e-05,
-      "loss": 1.2785,
-      "num_input_tokens_seen": 5874320,
-      "step": 365
-    },
-    {
-      "epoch": 0.02563765793690404,
-      "grad_norm": 4.224324703216553,
-      "learning_rate": 9.743864098073555e-05,
-      "loss": 1.0926,
-      "num_input_tokens_seen": 5890704,
-      "step": 366
-    },
-    {
-      "epoch": 0.025707706182633287,
-      "grad_norm": 4.289444446563721,
-      "learning_rate": 9.743164273204903e-05,
-      "loss": 1.1913,
-      "num_input_tokens_seen": 5906016,
-      "step": 367
-    },
-    {
-      "epoch": 0.025777754428362536,
-      "grad_norm": 4.280707359313965,
-      "learning_rate": 9.742464448336253e-05,
-      "loss": 1.2238,
-      "num_input_tokens_seen": 5921784,
-      "step": 368
-    },
-    {
-      "epoch": 0.025847802674091782,
-      "grad_norm": 4.554803848266602,
-      "learning_rate": 9.741764623467602e-05,
-      "loss": 1.2491,
-      "num_input_tokens_seen": 5938072,
-      "step": 369
-    },
-    {
-      "epoch": 0.025917850919821028,
-      "grad_norm": 4.677784442901611,
-      "learning_rate": 9.74106479859895e-05,
-      "loss": 1.2387,
-      "num_input_tokens_seen": 5954456,
-      "step": 370
-    },
-    {
-      "epoch": 0.025987899165550273,
-      "grad_norm": 4.268225193023682,
-      "learning_rate": 9.740364973730298e-05,
-      "loss": 1.2983,
-      "num_input_tokens_seen": 5970664,
-      "step": 371
-    },
-    {
-      "epoch": 0.02605794741127952,
-      "grad_norm": 4.361818790435791,
-      "learning_rate": 9.739665148861646e-05,
-      "loss": 1.199,
-      "num_input_tokens_seen": 5987048,
-      "step": 372
-    },
-    {
-      "epoch": 0.026127995657008764,
-      "grad_norm": 3.9990735054016113,
-      "learning_rate": 9.738965323992995e-05,
-      "loss": 1.0777,
-      "num_input_tokens_seen": 6003432,
-      "step": 373
-    },
-    {
-      "epoch": 0.02619804390273801,
-      "grad_norm": 3.992142915725708,
-      "learning_rate": 9.738265499124344e-05,
-      "loss": 1.0443,
-      "num_input_tokens_seen": 6019816,
-      "step": 374
-    },
-    {
-      "epoch": 0.026268092148467256,
-      "grad_norm": 4.270167827606201,
-      "learning_rate": 9.737565674255693e-05,
-      "loss": 1.1764,
-      "num_input_tokens_seen": 6036200,
-      "step": 375
-    },
-    {
-      "epoch": 0.0263381403941965,
-      "grad_norm": 4.362086296081543,
-      "learning_rate": 9.736865849387041e-05,
-      "loss": 1.2735,
-      "num_input_tokens_seen": 6052120,
-      "step": 376
-    },
-    {
-      "epoch": 0.02640818863992575,
-      "grad_norm": 3.6900475025177,
-      "learning_rate": 9.736166024518389e-05,
-      "loss": 0.8729,
-      "num_input_tokens_seen": 6068264,
-      "step": 377
-    },
-    {
-      "epoch": 0.026478236885654996,
-      "grad_norm": 3.8281285762786865,
-      "learning_rate": 9.735466199649738e-05,
-      "loss": 1.1096,
-      "num_input_tokens_seen": 6084504,
-      "step": 378
-    },
-    {
-      "epoch": 0.02654828513138424,
-      "grad_norm": 3.9335553646087646,
-      "learning_rate": 9.734766374781087e-05,
-      "loss": 1.0763,
-      "num_input_tokens_seen": 6100592,
-      "step": 379
-    },
-    {
-      "epoch": 0.026618333377113487,
-      "grad_norm": 4.332645416259766,
-      "learning_rate": 9.734066549912434e-05,
-      "loss": 1.1751,
-      "num_input_tokens_seen": 6116976,
-      "step": 380
-    },
-    {
-      "epoch": 0.026688381622842733,
-      "grad_norm": 4.160863399505615,
-      "learning_rate": 9.733366725043783e-05,
-      "loss": 1.0778,
-      "num_input_tokens_seen": 6133360,
-      "step": 381
-    },
-    {
-      "epoch": 0.02675842986857198,
-      "grad_norm": 4.388178825378418,
-      "learning_rate": 9.732666900175132e-05,
-      "loss": 1.2214,
-      "num_input_tokens_seen": 6149744,
-      "step": 382
-    },
-    {
-      "epoch": 0.026828478114301224,
-      "grad_norm": 4.354910373687744,
-      "learning_rate": 9.73196707530648e-05,
-      "loss": 1.4115,
-      "num_input_tokens_seen": 6166048,
-      "step": 383
-    },
-    {
-      "epoch": 0.02689852636003047,
-      "grad_norm": 4.058071613311768,
-      "learning_rate": 9.73126725043783e-05,
-      "loss": 1.0934,
-      "num_input_tokens_seen": 6181840,
-      "step": 384
-    },
-    {
-      "epoch": 0.026968574605759715,
-      "grad_norm": 4.060855865478516,
-      "learning_rate": 9.730567425569177e-05,
-      "loss": 1.1395,
-      "num_input_tokens_seen": 6198224,
-      "step": 385
-    },
-    {
-      "epoch": 0.027038622851488964,
-      "grad_norm": 4.316681385040283,
-      "learning_rate": 9.729867600700526e-05,
-      "loss": 1.1052,
-      "num_input_tokens_seen": 6214608,
-      "step": 386
-    },
-    {
-      "epoch": 0.02710867109721821,
-      "grad_norm": 4.322516918182373,
-      "learning_rate": 9.729167775831873e-05,
-      "loss": 1.2512,
-      "num_input_tokens_seen": 6230992,
-      "step": 387
-    },
-    {
-      "epoch": 0.027178719342947456,
-      "grad_norm": 4.090857028961182,
-      "learning_rate": 9.728467950963224e-05,
-      "loss": 1.0772,
-      "num_input_tokens_seen": 6246760,
-      "step": 388
-    },
-    {
-      "epoch": 0.0272487675886767,
-      "grad_norm": 4.0143961906433105,
-      "learning_rate": 9.727768126094571e-05,
-      "loss": 1.0578,
-      "num_input_tokens_seen": 6261968,
-      "step": 389
-    },
-    {
-      "epoch": 0.027318815834405947,
-      "grad_norm": 4.911194324493408,
-      "learning_rate": 9.72706830122592e-05,
-      "loss": 1.3016,
-      "num_input_tokens_seen": 6276664,
-      "step": 390
-    },
-    {
-      "epoch": 0.027388864080135192,
-      "grad_norm": 4.057498931884766,
-      "learning_rate": 9.726368476357269e-05,
-      "loss": 1.026,
-      "num_input_tokens_seen": 6293048,
-      "step": 391
-    },
-    {
-      "epoch": 0.027458912325864438,
-      "grad_norm": 3.9827401638031006,
-      "learning_rate": 9.725668651488616e-05,
-      "loss": 1.136,
-      "num_input_tokens_seen": 6309432,
-      "step": 392
-    },
-    {
-      "epoch": 0.027528960571593684,
-      "grad_norm": 4.640822887420654,
-      "learning_rate": 9.724968826619965e-05,
-      "loss": 1.2823,
-      "num_input_tokens_seen": 6325568,
-      "step": 393
-    },
-    {
-      "epoch": 0.027599008817322933,
-      "grad_norm": 4.372538089752197,
-      "learning_rate": 9.724269001751314e-05,
-      "loss": 1.0354,
-      "num_input_tokens_seen": 6341952,
-      "step": 394
-    },
-    {
-      "epoch": 0.02766905706305218,
-      "grad_norm": 4.018289566040039,
-      "learning_rate": 9.723569176882663e-05,
-      "loss": 1.029,
-      "num_input_tokens_seen": 6358336,
-      "step": 395
-    },
-    {
-      "epoch": 0.027739105308781424,
-      "grad_norm": 4.440858364105225,
-      "learning_rate": 9.722869352014012e-05,
-      "loss": 1.2272,
-      "num_input_tokens_seen": 6374680,
-      "step": 396
-    },
-    {
-      "epoch": 0.02780915355451067,
-      "grad_norm": 4.246788024902344,
-      "learning_rate": 9.722169527145359e-05,
-      "loss": 1.0161,
-      "num_input_tokens_seen": 6390672,
-      "step": 397
-    },
-    {
-      "epoch": 0.027879201800239915,
-      "grad_norm": 4.27274751663208,
-      "learning_rate": 9.721469702276708e-05,
-      "loss": 1.293,
-      "num_input_tokens_seen": 6407056,
-      "step": 398
-    },
-    {
-      "epoch": 0.02794925004596916,
-      "grad_norm": 4.171760559082031,
-      "learning_rate": 9.720769877408056e-05,
-      "loss": 1.2766,
-      "num_input_tokens_seen": 6423440,
-      "step": 399
-    },
-    {
-      "epoch": 0.028019298291698407,
-      "grad_norm": 4.174622535705566,
-      "learning_rate": 9.720070052539405e-05,
-      "loss": 1.049,
-      "num_input_tokens_seen": 6439824,
-      "step": 400
-    },
-    {
-      "epoch": 0.028019298291698407,
-      "eval_loss": 1.1994441747665405,
-      "eval_runtime": 0.2131,
-      "eval_samples_per_second": 4.693,
-      "eval_steps_per_second": 4.693,
-      "num_input_tokens_seen": 6439824,
-      "step": 400
-    },
-    {
-      "epoch": 0.028089346537427652,
-      "grad_norm": 4.199150562286377,
-      "learning_rate": 9.719370227670753e-05,
-      "loss": 1.3432,
-      "num_input_tokens_seen": 6456208,
-      "step": 401
-    },
-    {
-      "epoch": 0.028159394783156898,
-      "grad_norm": 3.9011733531951904,
-      "learning_rate": 9.718670402802102e-05,
-      "loss": 1.0895,
-      "num_input_tokens_seen": 6472592,
-      "step": 402
-    },
-    {
-      "epoch": 0.028229443028886147,
-      "grad_norm": 4.142306327819824,
-      "learning_rate": 9.717970577933451e-05,
-      "loss": 0.9031,
-      "num_input_tokens_seen": 6488976,
-      "step": 403
-    },
-    {
-      "epoch": 0.028299491274615392,
-      "grad_norm": 3.9745633602142334,
-      "learning_rate": 9.717270753064799e-05,
-      "loss": 0.9951,
-      "num_input_tokens_seen": 6505360,
-      "step": 404
-    },
-    {
-      "epoch": 0.028369539520344638,
-      "grad_norm": 3.838865280151367,
-      "learning_rate": 9.716570928196147e-05,
-      "loss": 0.809,
-      "num_input_tokens_seen": 6521744,
-      "step": 405
-    },
-    {
-      "epoch": 0.028439587766073884,
-      "grad_norm": 4.48146390914917,
-      "learning_rate": 9.715871103327496e-05,
-      "loss": 1.4985,
-      "num_input_tokens_seen": 6538128,
-      "step": 406
-    },
-    {
-      "epoch": 0.02850963601180313,
-      "grad_norm": 4.393556594848633,
-      "learning_rate": 9.715171278458844e-05,
-      "loss": 1.2355,
-      "num_input_tokens_seen": 6554512,
-      "step": 407
-    },
-    {
-      "epoch": 0.028579684257532375,
-      "grad_norm": 3.970860004425049,
-      "learning_rate": 9.714471453590194e-05,
-      "loss": 1.1513,
-      "num_input_tokens_seen": 6570896,
-      "step": 408
-    },
-    {
-      "epoch": 0.02864973250326162,
-      "grad_norm": 4.166610240936279,
-      "learning_rate": 9.713771628721542e-05,
-      "loss": 1.108,
-      "num_input_tokens_seen": 6587216,
-      "step": 409
-    },
-    {
-      "epoch": 0.028719780748990866,
-      "grad_norm": 3.9887096881866455,
-      "learning_rate": 9.71307180385289e-05,
-      "loss": 1.1639,
-      "num_input_tokens_seen": 6603600,
-      "step": 410
-    },
-    {
-      "epoch": 0.028789828994720112,
-      "grad_norm": 4.195802211761475,
-      "learning_rate": 9.712371978984239e-05,
-      "loss": 1.1478,
-      "num_input_tokens_seen": 6619984,
-      "step": 411
-    },
-    {
-      "epoch": 0.02885987724044936,
-      "grad_norm": 4.011331081390381,
-      "learning_rate": 9.711672154115587e-05,
-      "loss": 0.9554,
-      "num_input_tokens_seen": 6635904,
-      "step": 412
-    },
-    {
-      "epoch": 0.028929925486178606,
-      "grad_norm": 4.4170026779174805,
-      "learning_rate": 9.710972329246936e-05,
-      "loss": 1.1452,
-      "num_input_tokens_seen": 6651944,
-      "step": 413
-    },
-    {
-      "epoch": 0.028999973731907852,
-      "grad_norm": 4.073450088500977,
-      "learning_rate": 9.710272504378284e-05,
-      "loss": 1.1187,
-      "num_input_tokens_seen": 6668096,
-      "step": 414
-    },
-    {
-      "epoch": 0.029070021977637098,
-      "grad_norm": 4.161722183227539,
-      "learning_rate": 9.709572679509633e-05,
-      "loss": 1.1603,
-      "num_input_tokens_seen": 6684480,
-      "step": 415
-    },
-    {
-      "epoch": 0.029140070223366343,
-      "grad_norm": 4.540097713470459,
-      "learning_rate": 9.708872854640981e-05,
-      "loss": 1.2143,
-      "num_input_tokens_seen": 6700536,
-      "step": 416
-    },
-    {
-      "epoch": 0.02921011846909559,
-      "grad_norm": 4.030871868133545,
-      "learning_rate": 9.70817302977233e-05,
-      "loss": 0.9791,
-      "num_input_tokens_seen": 6716920,
-      "step": 417
-    },
-    {
-      "epoch": 0.029280166714824835,
-      "grad_norm": 4.1743268966674805,
-      "learning_rate": 9.707473204903679e-05,
-      "loss": 0.9818,
-      "num_input_tokens_seen": 6733304,
-      "step": 418
-    },
-    {
-      "epoch": 0.02935021496055408,
-      "grad_norm": 4.227272987365723,
-      "learning_rate": 9.706773380035026e-05,
-      "loss": 1.0945,
-      "num_input_tokens_seen": 6749688,
-      "step": 419
-    },
-    {
-      "epoch": 0.02942026320628333,
-      "grad_norm": 4.406428813934326,
-      "learning_rate": 9.706073555166375e-05,
-      "loss": 1.0302,
-      "num_input_tokens_seen": 6766072,
-      "step": 420
-    },
-    {
-      "epoch": 0.029490311452012575,
-      "grad_norm": 4.17899227142334,
-      "learning_rate": 9.705373730297724e-05,
-      "loss": 1.1048,
-      "num_input_tokens_seen": 6782456,
-      "step": 421
-    },
-    {
-      "epoch": 0.02956035969774182,
-      "grad_norm": 4.034752368927002,
-      "learning_rate": 9.704673905429073e-05,
-      "loss": 1.2639,
-      "num_input_tokens_seen": 6798840,
-      "step": 422
-    },
-    {
-      "epoch": 0.029630407943471066,
-      "grad_norm": 4.795727729797363,
-      "learning_rate": 9.703974080560421e-05,
-      "loss": 1.2448,
-      "num_input_tokens_seen": 6814912,
-      "step": 423
-    },
-    {
-      "epoch": 0.029700456189200312,
-      "grad_norm": 4.509056568145752,
-      "learning_rate": 9.703274255691769e-05,
-      "loss": 1.2157,
-      "num_input_tokens_seen": 6830720,
-      "step": 424
-    },
-    {
-      "epoch": 0.029770504434929557,
-      "grad_norm": 4.064620494842529,
-      "learning_rate": 9.702574430823118e-05,
-      "loss": 1.2042,
-      "num_input_tokens_seen": 6847104,
-      "step": 425
-    },
-    {
-      "epoch": 0.029840552680658803,
-      "grad_norm": 3.9060182571411133,
-      "learning_rate": 9.701874605954465e-05,
-      "loss": 0.9116,
-      "num_input_tokens_seen": 6862952,
-      "step": 426
-    },
-    {
-      "epoch": 0.02991060092638805,
-      "grad_norm": 3.9900951385498047,
-      "learning_rate": 9.701174781085814e-05,
-      "loss": 1.1621,
-      "num_input_tokens_seen": 6879336,
-      "step": 427
-    },
-    {
-      "epoch": 0.029980649172117294,
-      "grad_norm": 4.371436595916748,
-      "learning_rate": 9.700474956217164e-05,
-      "loss": 1.2731,
-      "num_input_tokens_seen": 6895720,
-      "step": 428
-    },
-    {
-      "epoch": 0.030050697417846543,
-      "grad_norm": 3.9422085285186768,
-      "learning_rate": 9.699775131348512e-05,
-      "loss": 0.9636,
-      "num_input_tokens_seen": 6912104,
-      "step": 429
-    },
-    {
-      "epoch": 0.03012074566357579,
-      "grad_norm": 4.080913543701172,
-      "learning_rate": 9.699075306479861e-05,
-      "loss": 1.1507,
-      "num_input_tokens_seen": 6928488,
-      "step": 430
-    },
-    {
-      "epoch": 0.030190793909305035,
-      "grad_norm": 4.493942737579346,
-      "learning_rate": 9.698375481611208e-05,
-      "loss": 1.2274,
-      "num_input_tokens_seen": 6944664,
-      "step": 431
-    },
-    {
-      "epoch": 0.03026084215503428,
-      "grad_norm": 4.073723793029785,
-      "learning_rate": 9.697675656742557e-05,
-      "loss": 1.0498,
-      "num_input_tokens_seen": 6960344,
-      "step": 432
-    },
-    {
-      "epoch": 0.030330890400763526,
-      "grad_norm": 3.9672274589538574,
-      "learning_rate": 9.696975831873906e-05,
-      "loss": 1.007,
-      "num_input_tokens_seen": 6976720,
-      "step": 433
-    },
-    {
-      "epoch": 0.03040093864649277,
-      "grad_norm": 4.497872829437256,
-      "learning_rate": 9.696276007005255e-05,
-      "loss": 1.1339,
-      "num_input_tokens_seen": 6992552,
-      "step": 434
-    },
-    {
-      "epoch": 0.030470986892222017,
-      "grad_norm": 4.422168731689453,
-      "learning_rate": 9.695576182136604e-05,
-      "loss": 1.34,
-      "num_input_tokens_seen": 7008936,
-      "step": 435
-    },
-    {
-      "epoch": 0.030541035137951263,
-      "grad_norm": 4.3009138107299805,
-      "learning_rate": 9.694876357267951e-05,
-      "loss": 1.2479,
-      "num_input_tokens_seen": 7024512,
-      "step": 436
-    },
-    {
-      "epoch": 0.030611083383680508,
-      "grad_norm": 4.04030704498291,
-      "learning_rate": 9.6941765323993e-05,
-      "loss": 1.097,
-      "num_input_tokens_seen": 7040896,
-      "step": 437
-    },
-    {
-      "epoch": 0.030681131629409757,
-      "grad_norm": 3.877417802810669,
-      "learning_rate": 9.693476707530649e-05,
-      "loss": 1.1363,
-      "num_input_tokens_seen": 7057280,
-      "step": 438
-    },
-    {
-      "epoch": 0.030751179875139003,
-      "grad_norm": 3.8185505867004395,
-      "learning_rate": 9.692776882661996e-05,
-      "loss": 0.9067,
-      "num_input_tokens_seen": 7072544,
-      "step": 439
-    },
-    {
-      "epoch": 0.03082122812086825,
-      "grad_norm": 4.028950214385986,
-      "learning_rate": 9.692077057793345e-05,
-      "loss": 1.1195,
-      "num_input_tokens_seen": 7088928,
-      "step": 440
-    },
-    {
-      "epoch": 0.030891276366597494,
-      "grad_norm": 4.2786431312561035,
-      "learning_rate": 9.691377232924694e-05,
-      "loss": 1.1199,
-      "num_input_tokens_seen": 7105248,
-      "step": 441
-    },
-    {
-      "epoch": 0.03096132461232674,
-      "grad_norm": 4.193462371826172,
-      "learning_rate": 9.690677408056043e-05,
-      "loss": 1.1812,
-      "num_input_tokens_seen": 7121008,
-      "step": 442
-    },
-    {
-      "epoch": 0.031031372858055985,
-      "grad_norm": 3.93597412109375,
-      "learning_rate": 9.68997758318739e-05,
-      "loss": 1.0677,
-      "num_input_tokens_seen": 7136944,
-      "step": 443
-    },
-    {
-      "epoch": 0.03110142110378523,
-      "grad_norm": 4.3208537101745605,
-      "learning_rate": 9.68927775831874e-05,
-      "loss": 1.1358,
-      "num_input_tokens_seen": 7152928,
-      "step": 444
-    },
-    {
-      "epoch": 0.031171469349514477,
-      "grad_norm": 3.9743378162384033,
-      "learning_rate": 9.688577933450088e-05,
-      "loss": 1.094,
-      "num_input_tokens_seen": 7169312,
-      "step": 445
-    },
-    {
-      "epoch": 0.031241517595243726,
-      "grad_norm": 4.226114273071289,
-      "learning_rate": 9.687878108581436e-05,
-      "loss": 1.1752,
-      "num_input_tokens_seen": 7185696,
-      "step": 446
-    },
-    {
-      "epoch": 0.03131156584097297,
-      "grad_norm": 4.210222244262695,
-      "learning_rate": 9.687178283712785e-05,
-      "loss": 1.1262,
-      "num_input_tokens_seen": 7201784,
-      "step": 447
-    },
-    {
-      "epoch": 0.03138161408670222,
-      "grad_norm": 4.311635971069336,
-      "learning_rate": 9.686478458844133e-05,
-      "loss": 1.2491,
-      "num_input_tokens_seen": 7218168,
-      "step": 448
-    },
-    {
-      "epoch": 0.03145166233243146,
-      "grad_norm": 4.56603479385376,
-      "learning_rate": 9.685778633975482e-05,
-      "loss": 1.3512,
-      "num_input_tokens_seen": 7233360,
-      "step": 449
-    },
-    {
-      "epoch": 0.03152171057816071,
-      "grad_norm": 4.232856750488281,
-      "learning_rate": 9.685078809106831e-05,
-      "loss": 0.9387,
-      "num_input_tokens_seen": 7248280,
-      "step": 450
-    },
-    {
-      "epoch": 0.031591758823889954,
-      "grad_norm": 4.512947082519531,
-      "learning_rate": 9.684378984238179e-05,
-      "loss": 1.1988,
-      "num_input_tokens_seen": 7264664,
-      "step": 451
-    },
-    {
-      "epoch": 0.0316618070696192,
-      "grad_norm": 4.273897171020508,
-      "learning_rate": 9.683679159369528e-05,
-      "loss": 1.2523,
-      "num_input_tokens_seen": 7281048,
-      "step": 452
-    },
-    {
-      "epoch": 0.031731855315348445,
-      "grad_norm": 4.288438320159912,
-      "learning_rate": 9.682979334500875e-05,
-      "loss": 1.1692,
-      "num_input_tokens_seen": 7297424,
-      "step": 453
-    },
-    {
-      "epoch": 0.03180190356107769,
-      "grad_norm": 4.27367639541626,
-      "learning_rate": 9.682279509632225e-05,
-      "loss": 1.1868,
-      "num_input_tokens_seen": 7312792,
-      "step": 454
-    },
-    {
-      "epoch": 0.031871951806806936,
-      "grad_norm": 3.978926181793213,
-      "learning_rate": 9.681579684763574e-05,
-      "loss": 1.0382,
-      "num_input_tokens_seen": 7329176,
-      "step": 455
-    },
-    {
-      "epoch": 0.03194200005253618,
-      "grad_norm": 4.4399919509887695,
-      "learning_rate": 9.680879859894922e-05,
-      "loss": 1.2072,
-      "num_input_tokens_seen": 7345560,
-      "step": 456
-    },
-    {
-      "epoch": 0.03201204829826543,
-      "grad_norm": 3.9786529541015625,
-      "learning_rate": 9.68018003502627e-05,
-      "loss": 1.1704,
-      "num_input_tokens_seen": 7361944,
-      "step": 457
-    },
-    {
-      "epoch": 0.03208209654399467,
-      "grad_norm": 4.171195030212402,
-      "learning_rate": 9.679480210157618e-05,
-      "loss": 1.1307,
-      "num_input_tokens_seen": 7378328,
-      "step": 458
-    },
-    {
-      "epoch": 0.032152144789723926,
-      "grad_norm": 3.9415268898010254,
-      "learning_rate": 9.678780385288967e-05,
-      "loss": 0.9971,
-      "num_input_tokens_seen": 7394208,
-      "step": 459
-    },
-    {
-      "epoch": 0.03222219303545317,
-      "grad_norm": 4.066036224365234,
-      "learning_rate": 9.678080560420316e-05,
-      "loss": 1.1227,
-      "num_input_tokens_seen": 7410328,
-      "step": 460
-    },
-    {
-      "epoch": 0.03229224128118242,
-      "grad_norm": 4.22513484954834,
-      "learning_rate": 9.677380735551665e-05,
-      "loss": 1.0883,
-      "num_input_tokens_seen": 7426712,
-      "step": 461
-    },
-    {
-      "epoch": 0.03236228952691166,
-      "grad_norm": 4.310954570770264,
-      "learning_rate": 9.676680910683013e-05,
-      "loss": 1.1695,
-      "num_input_tokens_seen": 7442736,
-      "step": 462
-    },
-    {
-      "epoch": 0.03243233777264091,
-      "grad_norm": 4.2868828773498535,
-      "learning_rate": 9.675981085814361e-05,
-      "loss": 1.0594,
-      "num_input_tokens_seen": 7458560,
-      "step": 463
-    },
-    {
-      "epoch": 0.032502386018370154,
-      "grad_norm": 4.318186283111572,
-      "learning_rate": 9.67528126094571e-05,
-      "loss": 1.1791,
-      "num_input_tokens_seen": 7474944,
-      "step": 464
-    },
-    {
-      "epoch": 0.0325724342640994,
-      "grad_norm": 4.040421009063721,
-      "learning_rate": 9.674581436077059e-05,
-      "loss": 1.0649,
-      "num_input_tokens_seen": 7490344,
-      "step": 465
-    },
-    {
-      "epoch": 0.032642482509828645,
-      "grad_norm": 3.914815902709961,
-      "learning_rate": 9.673881611208406e-05,
-      "loss": 1.1381,
-      "num_input_tokens_seen": 7506728,
-      "step": 466
-    },
-    {
-      "epoch": 0.03271253075555789,
-      "grad_norm": 4.054527282714844,
-      "learning_rate": 9.673181786339755e-05,
-      "loss": 1.2264,
-      "num_input_tokens_seen": 7522912,
-      "step": 467
-    },
-    {
-      "epoch": 0.032782579001287136,
-      "grad_norm": 4.295147895812988,
-      "learning_rate": 9.672481961471104e-05,
-      "loss": 1.1369,
-      "num_input_tokens_seen": 7539040,
-      "step": 468
-    },
-    {
-      "epoch": 0.03285262724701638,
-      "grad_norm": 4.109183311462402,
-      "learning_rate": 9.671782136602453e-05,
-      "loss": 1.1676,
-      "num_input_tokens_seen": 7555424,
-      "step": 469
-    },
-    {
-      "epoch": 0.03292267549274563,
-      "grad_norm": 4.131369590759277,
-      "learning_rate": 9.6710823117338e-05,
-      "loss": 1.1188,
-      "num_input_tokens_seen": 7571808,
-      "step": 470
-    },
-    {
-      "epoch": 0.03299272373847487,
-      "grad_norm": 3.998414993286133,
-      "learning_rate": 9.670382486865149e-05,
-      "loss": 1.0201,
-      "num_input_tokens_seen": 7587528,
-      "step": 471
-    },
-    {
-      "epoch": 0.03306277198420412,
-      "grad_norm": 4.1235551834106445,
-      "learning_rate": 9.669682661996498e-05,
-      "loss": 1.1265,
-      "num_input_tokens_seen": 7603912,
-      "step": 472
-    },
-    {
-      "epoch": 0.033132820229933364,
-      "grad_norm": 4.800798416137695,
-      "learning_rate": 9.668982837127845e-05,
-      "loss": 1.3634,
-      "num_input_tokens_seen": 7617512,
-      "step": 473
-    },
-    {
-      "epoch": 0.03320286847566261,
-      "grad_norm": 4.068000316619873,
-      "learning_rate": 9.668283012259196e-05,
-      "loss": 1.1427,
-      "num_input_tokens_seen": 7633040,
-      "step": 474
-    },
-    {
-      "epoch": 0.033272916721391856,
-      "grad_norm": 4.0715484619140625,
-      "learning_rate": 9.667583187390543e-05,
-      "loss": 1.0633,
-      "num_input_tokens_seen": 7648416,
-      "step": 475
-    },
-    {
-      "epoch": 0.0333429649671211,
-      "grad_norm": 3.937807321548462,
-      "learning_rate": 9.666883362521892e-05,
-      "loss": 1.1393,
-      "num_input_tokens_seen": 7664624,
-      "step": 476
-    },
-    {
-      "epoch": 0.033413013212850354,
-      "grad_norm": 4.195656776428223,
-      "learning_rate": 9.666183537653241e-05,
-      "loss": 1.1801,
-      "num_input_tokens_seen": 7680480,
-      "step": 477
-    },
-    {
-      "epoch": 0.0334830614585796,
-      "grad_norm": 4.227575778961182,
-      "learning_rate": 9.665483712784588e-05,
-      "loss": 1.0453,
-      "num_input_tokens_seen": 7696632,
-      "step": 478
-    },
-    {
-      "epoch": 0.033553109704308845,
-      "grad_norm": 4.328822135925293,
-      "learning_rate": 9.664783887915937e-05,
-      "loss": 1.221,
-      "num_input_tokens_seen": 7713016,
-      "step": 479
-    },
-    {
-      "epoch": 0.03362315795003809,
-      "grad_norm": 4.086736679077148,
-      "learning_rate": 9.664084063047286e-05,
-      "loss": 1.2817,
-      "num_input_tokens_seen": 7729400,
-      "step": 480
-    },
-    {
-      "epoch": 0.033693206195767336,
-      "grad_norm": 4.555233955383301,
-      "learning_rate": 9.663384238178635e-05,
-      "loss": 1.483,
-      "num_input_tokens_seen": 7745784,
-      "step": 481
-    },
-    {
-      "epoch": 0.03376325444149658,
-      "grad_norm": 4.118983745574951,
-      "learning_rate": 9.662684413309984e-05,
-      "loss": 0.9139,
-      "num_input_tokens_seen": 7762168,
-      "step": 482
-    },
-    {
-      "epoch": 0.03383330268722583,
-      "grad_norm": 4.232059001922607,
-      "learning_rate": 9.661984588441331e-05,
-      "loss": 1.1269,
-      "num_input_tokens_seen": 7777920,
-      "step": 483
-    },
-    {
-      "epoch": 0.03390335093295507,
-      "grad_norm": 6.288865089416504,
-      "learning_rate": 9.66128476357268e-05,
-      "loss": 1.0642,
-      "num_input_tokens_seen": 7794304,
-      "step": 484
-    },
-    {
-      "epoch": 0.03397339917868432,
-      "grad_norm": 4.133046627044678,
-      "learning_rate": 9.660584938704028e-05,
-      "loss": 1.2067,
-      "num_input_tokens_seen": 7810200,
-      "step": 485
-    },
-    {
-      "epoch": 0.034043447424413564,
-      "grad_norm": 4.147965431213379,
-      "learning_rate": 9.659885113835377e-05,
-      "loss": 1.0367,
-      "num_input_tokens_seen": 7826384,
-      "step": 486
-    },
-    {
-      "epoch": 0.03411349567014281,
-      "grad_norm": 4.1191020011901855,
-      "learning_rate": 9.659185288966725e-05,
-      "loss": 1.0972,
-      "num_input_tokens_seen": 7841704,
-      "step": 487
-    },
-    {
-      "epoch": 0.034183543915872056,
-      "grad_norm": 4.518441677093506,
-      "learning_rate": 9.658485464098074e-05,
-      "loss": 1.263,
-      "num_input_tokens_seen": 7858088,
-      "step": 488
-    },
-    {
-      "epoch": 0.0342535921616013,
-      "grad_norm": 4.321181297302246,
-      "learning_rate": 9.657785639229423e-05,
-      "loss": 1.1378,
-      "num_input_tokens_seen": 7874472,
-      "step": 489
-    },
-    {
-      "epoch": 0.03432364040733055,
-      "grad_norm": 4.366185665130615,
-      "learning_rate": 9.65708581436077e-05,
-      "loss": 1.1636,
-      "num_input_tokens_seen": 7890856,
-      "step": 490
-    },
-    {
-      "epoch": 0.03439368865305979,
-      "grad_norm": 4.042731761932373,
-      "learning_rate": 9.65638598949212e-05,
-      "loss": 1.0601,
-      "num_input_tokens_seen": 7906776,
-      "step": 491
-    },
-    {
-      "epoch": 0.03446373689878904,
-      "grad_norm": 3.743668556213379,
-      "learning_rate": 9.655686164623468e-05,
-      "loss": 1.0441,
-      "num_input_tokens_seen": 7923160,
-      "step": 492
-    },
-    {
-      "epoch": 0.034533785144518284,
-      "grad_norm": 3.8547139167785645,
-      "learning_rate": 9.654986339754816e-05,
-      "loss": 1.0842,
-      "num_input_tokens_seen": 7939296,
-      "step": 493
-    },
-    {
-      "epoch": 0.034603833390247536,
-      "grad_norm": 4.238414287567139,
-      "learning_rate": 9.654286514886166e-05,
-      "loss": 1.2498,
-      "num_input_tokens_seen": 7955504,
-      "step": 494
-    },
-    {
-      "epoch": 0.03467388163597678,
-      "grad_norm": 4.134857177734375,
-      "learning_rate": 9.653586690017514e-05,
-      "loss": 1.1241,
-      "num_input_tokens_seen": 7971888,
-      "step": 495
-    },
-    {
-      "epoch": 0.03474392988170603,
-      "grad_norm": 4.2501983642578125,
-      "learning_rate": 9.652886865148862e-05,
-      "loss": 1.1829,
-      "num_input_tokens_seen": 7988272,
-      "step": 496
-    },
-    {
-      "epoch": 0.03481397812743527,
-      "grad_norm": 7.4397053718566895,
-      "learning_rate": 9.65218704028021e-05,
-      "loss": 0.9952,
-      "num_input_tokens_seen": 8003744,
-      "step": 497
-    },
-    {
-      "epoch": 0.03488402637316452,
-      "grad_norm": 4.2750959396362305,
-      "learning_rate": 9.651487215411559e-05,
-      "loss": 1.2387,
-      "num_input_tokens_seen": 8019184,
-      "step": 498
-    },
-    {
-      "epoch": 0.034954074618893764,
-      "grad_norm": 4.156162261962891,
-      "learning_rate": 9.650787390542908e-05,
-      "loss": 1.1201,
-      "num_input_tokens_seen": 8035176,
-      "step": 499
-    },
-    {
-      "epoch": 0.03502412286462301,
-      "grad_norm": 4.178225040435791,
-      "learning_rate": 9.650087565674257e-05,
-      "loss": 1.2026,
-      "num_input_tokens_seen": 8051560,
-      "step": 500
-    },
-    {
-      "epoch": 0.035094171110352256,
-      "grad_norm": 4.147096157073975,
-      "learning_rate": 9.649387740805605e-05,
-      "loss": 1.2465,
-      "num_input_tokens_seen": 8067944,
-      "step": 501
-    },
-    {
-      "epoch": 0.0351642193560815,
-      "grad_norm": 4.329249858856201,
-      "learning_rate": 9.648687915936953e-05,
-      "loss": 1.2742,
-      "num_input_tokens_seen": 8083824,
-      "step": 502
-    },
-    {
-      "epoch": 0.03523426760181075,
-      "grad_norm": 4.404232978820801,
-      "learning_rate": 9.647988091068302e-05,
-      "loss": 1.1511,
-      "num_input_tokens_seen": 8100208,
-      "step": 503
-    },
-    {
-      "epoch": 0.03530431584753999,
-      "grad_norm": 4.190586090087891,
-      "learning_rate": 9.64728826619965e-05,
-      "loss": 0.9884,
-      "num_input_tokens_seen": 8116048,
-      "step": 504
-    },
-    {
-      "epoch": 0.03537436409326924,
-      "grad_norm": 4.262845516204834,
-      "learning_rate": 9.646588441330998e-05,
-      "loss": 1.1321,
-      "num_input_tokens_seen": 8132432,
-      "step": 505
-    },
-    {
-      "epoch": 0.035444412338998484,
-      "grad_norm": 4.452746391296387,
-      "learning_rate": 9.645888616462347e-05,
-      "loss": 1.1667,
-      "num_input_tokens_seen": 8148816,
-      "step": 506
-    },
-    {
-      "epoch": 0.03551446058472773,
-      "grad_norm": 4.111443042755127,
-      "learning_rate": 9.645188791593696e-05,
-      "loss": 1.0049,
-      "num_input_tokens_seen": 8164856,
-      "step": 507
-    },
-    {
-      "epoch": 0.035584508830456975,
-      "grad_norm": 4.292227268218994,
-      "learning_rate": 9.644488966725045e-05,
-      "loss": 1.1535,
-      "num_input_tokens_seen": 8181240,
-      "step": 508
-    },
-    {
-      "epoch": 0.03565455707618622,
-      "grad_norm": 4.295238971710205,
-      "learning_rate": 9.643789141856394e-05,
-      "loss": 1.236,
-      "num_input_tokens_seen": 8197624,
-      "step": 509
-    },
-    {
-      "epoch": 0.035724605321915466,
-      "grad_norm": 3.930659294128418,
-      "learning_rate": 9.643089316987741e-05,
-      "loss": 0.9195,
-      "num_input_tokens_seen": 8213816,
-      "step": 510
-    },
-    {
-      "epoch": 0.03579465356764472,
-      "grad_norm": 4.092316150665283,
-      "learning_rate": 9.64238949211909e-05,
-      "loss": 1.0799,
-      "num_input_tokens_seen": 8229632,
-      "step": 511
-    },
-    {
-      "epoch": 0.035864701813373964,
-      "grad_norm": 4.2939252853393555,
-      "learning_rate": 9.641689667250437e-05,
-      "loss": 1.111,
-      "num_input_tokens_seen": 8245232,
-      "step": 512
-    },
-    {
-      "epoch": 0.03593475005910321,
-      "grad_norm": 4.191503524780273,
-      "learning_rate": 9.640989842381786e-05,
-      "loss": 0.9399,
-      "num_input_tokens_seen": 8260912,
-      "step": 513
-    },
-    {
-      "epoch": 0.036004798304832455,
-      "grad_norm": 4.141485214233398,
-      "learning_rate": 9.640290017513136e-05,
-      "loss": 1.1334,
-      "num_input_tokens_seen": 8276864,
-      "step": 514
-    },
-    {
-      "epoch": 0.0360748465505617,
-      "grad_norm": 3.890547752380371,
-      "learning_rate": 9.639590192644484e-05,
-      "loss": 1.0055,
-      "num_input_tokens_seen": 8292720,
-      "step": 515
-    },
-    {
-      "epoch": 0.03614489479629095,
-      "grad_norm": 4.405922889709473,
-      "learning_rate": 9.638890367775833e-05,
-      "loss": 1.2238,
-      "num_input_tokens_seen": 8309104,
-      "step": 516
-    },
-    {
-      "epoch": 0.03621494304202019,
-      "grad_norm": 4.207942485809326,
-      "learning_rate": 9.63819054290718e-05,
-      "loss": 1.0688,
-      "num_input_tokens_seen": 8325304,
-      "step": 517
-    },
-    {
-      "epoch": 0.03628499128774944,
-      "grad_norm": 4.174366474151611,
-      "learning_rate": 9.637490718038529e-05,
-      "loss": 1.2303,
-      "num_input_tokens_seen": 8341688,
-      "step": 518
-    },
-    {
-      "epoch": 0.036355039533478684,
-      "grad_norm": 3.9641714096069336,
-      "learning_rate": 9.636790893169878e-05,
-      "loss": 1.2244,
-      "num_input_tokens_seen": 8357760,
-      "step": 519
-    },
-    {
-      "epoch": 0.03642508777920793,
-      "grad_norm": 5.832678318023682,
-      "learning_rate": 9.636091068301227e-05,
-      "loss": 1.0645,
-      "num_input_tokens_seen": 8372712,
-      "step": 520
-    },
-    {
-      "epoch": 0.036495136024937175,
-      "grad_norm": 3.7905161380767822,
-      "learning_rate": 9.635391243432576e-05,
-      "loss": 1.0551,
-      "num_input_tokens_seen": 8389096,
-      "step": 521
-    },
-    {
-      "epoch": 0.03656518427066642,
-      "grad_norm": 3.6744072437286377,
-      "learning_rate": 9.634691418563923e-05,
-      "loss": 1.0687,
-      "num_input_tokens_seen": 8405216,
-      "step": 522
-    },
-    {
-      "epoch": 0.036635232516395666,
-      "grad_norm": 4.897486209869385,
-      "learning_rate": 9.633991593695272e-05,
-      "loss": 1.1968,
-      "num_input_tokens_seen": 8421600,
-      "step": 523
-    },
-    {
-      "epoch": 0.03670528076212491,
-      "grad_norm": 3.821457862854004,
-      "learning_rate": 9.63329176882662e-05,
-      "loss": 1.0473,
-      "num_input_tokens_seen": 8437984,
-      "step": 524
-    },
-    {
-      "epoch": 0.03677532900785416,
-      "grad_norm": 3.873832941055298,
-      "learning_rate": 9.632591943957969e-05,
-      "loss": 0.9656,
-      "num_input_tokens_seen": 8453760,
-      "step": 525
-    },
-    {
-      "epoch": 0.0368453772535834,
-      "grad_norm": 4.139901161193848,
-      "learning_rate": 9.631892119089317e-05,
-      "loss": 1.0881,
-      "num_input_tokens_seen": 8470144,
-      "step": 526
-    },
-    {
-      "epoch": 0.03691542549931265,
-      "grad_norm": 3.9512782096862793,
-      "learning_rate": 9.631192294220666e-05,
-      "loss": 1.1093,
-      "num_input_tokens_seen": 8486528,
-      "step": 527
-    },
-    {
-      "epoch": 0.0369854737450419,
-      "grad_norm": 3.8937103748321533,
-      "learning_rate": 9.630492469352015e-05,
-      "loss": 0.9722,
-      "num_input_tokens_seen": 8502912,
-      "step": 528
-    },
-    {
-      "epoch": 0.03705552199077115,
-      "grad_norm": 4.482640743255615,
-      "learning_rate": 9.629792644483363e-05,
-      "loss": 1.056,
-      "num_input_tokens_seen": 8519296,
-      "step": 529
-    },
-    {
-      "epoch": 0.03712557023650039,
-      "grad_norm": 4.127941131591797,
-      "learning_rate": 9.629092819614711e-05,
-      "loss": 1.0285,
-      "num_input_tokens_seen": 8535160,
-      "step": 530
-    },
-    {
-      "epoch": 0.03719561848222964,
-      "grad_norm": 3.973585844039917,
-      "learning_rate": 9.62839299474606e-05,
-      "loss": 1.0356,
-      "num_input_tokens_seen": 8551256,
-      "step": 531
-    },
-    {
-      "epoch": 0.037265666727958884,
-      "grad_norm": 4.22855281829834,
-      "learning_rate": 9.627693169877408e-05,
-      "loss": 1.134,
-      "num_input_tokens_seen": 8567640,
-      "step": 532
-    },
-    {
-      "epoch": 0.03733571497368813,
-      "grad_norm": 4.144021511077881,
-      "learning_rate": 9.626993345008757e-05,
-      "loss": 1.0963,
-      "num_input_tokens_seen": 8583504,
-      "step": 533
-    },
-    {
-      "epoch": 0.037405763219417375,
-      "grad_norm": 3.8666226863861084,
-      "learning_rate": 9.626293520140106e-05,
-      "loss": 0.912,
-      "num_input_tokens_seen": 8599888,
-      "step": 534
-    },
-    {
-      "epoch": 0.03747581146514662,
-      "grad_norm": 4.215412616729736,
-      "learning_rate": 9.625593695271454e-05,
-      "loss": 1.1055,
-      "num_input_tokens_seen": 8616256,
-      "step": 535
-    },
-    {
-      "epoch": 0.037545859710875866,
-      "grad_norm": 4.353022575378418,
-      "learning_rate": 9.624893870402803e-05,
-      "loss": 1.0379,
-      "num_input_tokens_seen": 8632640,
-      "step": 536
-    },
-    {
-      "epoch": 0.03761590795660511,
-      "grad_norm": 3.778947591781616,
-      "learning_rate": 9.624194045534151e-05,
-      "loss": 1.0547,
-      "num_input_tokens_seen": 8648624,
-      "step": 537
-    },
-    {
-      "epoch": 0.03768595620233436,
-      "grad_norm": 4.481568336486816,
-      "learning_rate": 9.6234942206655e-05,
-      "loss": 1.3407,
-      "num_input_tokens_seen": 8664200,
-      "step": 538
-    },
-    {
-      "epoch": 0.0377560044480636,
-      "grad_norm": 4.066302299499512,
-      "learning_rate": 9.622794395796847e-05,
-      "loss": 0.995,
-      "num_input_tokens_seen": 8680584,
-      "step": 539
-    },
-    {
-      "epoch": 0.03782605269379285,
-      "grad_norm": 4.262768268585205,
-      "learning_rate": 9.622094570928197e-05,
-      "loss": 1.3054,
-      "num_input_tokens_seen": 8696968,
-      "step": 540
-    },
-    {
-      "epoch": 0.037896100939522094,
-      "grad_norm": 3.777597665786743,
-      "learning_rate": 9.621394746059546e-05,
-      "loss": 0.9831,
-      "num_input_tokens_seen": 8713352,
-      "step": 541
-    },
-    {
-      "epoch": 0.03796614918525134,
-      "grad_norm": 3.9732742309570312,
-      "learning_rate": 9.620694921190894e-05,
-      "loss": 1.0699,
-      "num_input_tokens_seen": 8729048,
-      "step": 542
-    },
-    {
-      "epoch": 0.038036197430980585,
-      "grad_norm": 4.543329238891602,
-      "learning_rate": 9.619995096322243e-05,
-      "loss": 1.1546,
-      "num_input_tokens_seen": 8745432,
-      "step": 543
-    },
-    {
-      "epoch": 0.03810624567670983,
-      "grad_norm": 4.903865814208984,
-      "learning_rate": 9.61929527145359e-05,
-      "loss": 1.1548,
-      "num_input_tokens_seen": 8760296,
-      "step": 544
-    },
-    {
-      "epoch": 0.03817629392243908,
-      "grad_norm": 4.197691917419434,
-      "learning_rate": 9.618595446584939e-05,
-      "loss": 1.1616,
-      "num_input_tokens_seen": 8776680,
-      "step": 545
-    },
-    {
-      "epoch": 0.03824634216816833,
-      "grad_norm": 3.912689208984375,
-      "learning_rate": 9.617895621716288e-05,
-      "loss": 0.9926,
-      "num_input_tokens_seen": 8793064,
-      "step": 546
-    },
-    {
-      "epoch": 0.038316390413897575,
-      "grad_norm": 4.291840076446533,
-      "learning_rate": 9.617195796847637e-05,
-      "loss": 1.1943,
-      "num_input_tokens_seen": 8809448,
-      "step": 547
-    },
-    {
-      "epoch": 0.03838643865962682,
-      "grad_norm": 3.9053072929382324,
-      "learning_rate": 9.616495971978985e-05,
-      "loss": 1.2437,
-      "num_input_tokens_seen": 8825536,
-      "step": 548
-    },
-    {
-      "epoch": 0.038456486905356066,
-      "grad_norm": 4.860696315765381,
-      "learning_rate": 9.615796147110333e-05,
-      "loss": 1.3045,
-      "num_input_tokens_seen": 8841920,
-      "step": 549
-    },
-    {
-      "epoch": 0.03852653515108531,
-      "grad_norm": 3.9394373893737793,
-      "learning_rate": 9.615096322241682e-05,
-      "loss": 1.1367,
-      "num_input_tokens_seen": 8858304,
-      "step": 550
-    },
-    {
-      "epoch": 0.03859658339681456,
-      "grad_norm": 3.8160409927368164,
-      "learning_rate": 9.61439649737303e-05,
-      "loss": 1.0864,
-      "num_input_tokens_seen": 8874688,
-      "step": 551
-    },
-    {
-      "epoch": 0.0386666316425438,
-      "grad_norm": 4.3792805671691895,
-      "learning_rate": 9.613696672504378e-05,
-      "loss": 1.2516,
-      "num_input_tokens_seen": 8891072,
-      "step": 552
-    },
-    {
-      "epoch": 0.03873667988827305,
-      "grad_norm": 4.103452682495117,
-      "learning_rate": 9.612996847635727e-05,
-      "loss": 0.9737,
-      "num_input_tokens_seen": 8907456,
-      "step": 553
-    },
-    {
-      "epoch": 0.038806728134002294,
-      "grad_norm": 4.117603302001953,
-      "learning_rate": 9.612297022767076e-05,
-      "loss": 1.096,
-      "num_input_tokens_seen": 8923816,
-      "step": 554
-    },
-    {
-      "epoch": 0.03887677637973154,
-      "grad_norm": 4.272468566894531,
-      "learning_rate": 9.611597197898425e-05,
-      "loss": 1.161,
-      "num_input_tokens_seen": 8939344,
-      "step": 555
-    },
-    {
-      "epoch": 0.038946824625460785,
-      "grad_norm": 4.323635578155518,
-      "learning_rate": 9.610897373029772e-05,
-      "loss": 1.1922,
-      "num_input_tokens_seen": 8954920,
-      "step": 556
-    },
-    {
-      "epoch": 0.03901687287119003,
-      "grad_norm": 3.783510684967041,
-      "learning_rate": 9.610197548161121e-05,
-      "loss": 1.0658,
-      "num_input_tokens_seen": 8971304,
-      "step": 557
-    },
-    {
-      "epoch": 0.039086921116919277,
-      "grad_norm": 4.3757548332214355,
-      "learning_rate": 9.60949772329247e-05,
-      "loss": 1.3186,
-      "num_input_tokens_seen": 8987672,
-      "step": 558
-    },
-    {
-      "epoch": 0.03915696936264852,
-      "grad_norm": 4.048824787139893,
-      "learning_rate": 9.608797898423818e-05,
-      "loss": 1.1452,
-      "num_input_tokens_seen": 9003896,
-      "step": 559
-    },
-    {
-      "epoch": 0.03922701760837777,
-      "grad_norm": 4.06865930557251,
-      "learning_rate": 9.608098073555168e-05,
-      "loss": 0.9861,
-      "num_input_tokens_seen": 9020280,
-      "step": 560
-    },
-    {
-      "epoch": 0.03929706585410701,
-      "grad_norm": 3.966737747192383,
-      "learning_rate": 9.607398248686515e-05,
-      "loss": 1.0323,
-      "num_input_tokens_seen": 9036280,
-      "step": 561
-    },
-    {
-      "epoch": 0.03936711409983626,
-      "grad_norm": 4.466656684875488,
-      "learning_rate": 9.606698423817864e-05,
-      "loss": 1.2462,
-      "num_input_tokens_seen": 9052664,
-      "step": 562
-    },
-    {
-      "epoch": 0.03943716234556551,
-      "grad_norm": 4.312132358551025,
-      "learning_rate": 9.605998598949213e-05,
-      "loss": 1.2133,
-      "num_input_tokens_seen": 9068832,
-      "step": 563
-    },
-    {
-      "epoch": 0.03950721059129476,
-      "grad_norm": 3.9202895164489746,
-      "learning_rate": 9.60529877408056e-05,
-      "loss": 1.0723,
-      "num_input_tokens_seen": 9084680,
-      "step": 564
-    },
-    {
-      "epoch": 0.039577258837024,
-      "grad_norm": 5.139899730682373,
-      "learning_rate": 9.604598949211909e-05,
-      "loss": 1.1165,
-      "num_input_tokens_seen": 9099792,
-      "step": 565
-    },
-    {
-      "epoch": 0.03964730708275325,
-      "grad_norm": 4.398557186126709,
-      "learning_rate": 9.603899124343258e-05,
-      "loss": 1.1737,
-      "num_input_tokens_seen": 9116136,
-      "step": 566
-    },
-    {
-      "epoch": 0.039717355328482494,
-      "grad_norm": 4.350982666015625,
-      "learning_rate": 9.603199299474607e-05,
-      "loss": 1.2174,
-      "num_input_tokens_seen": 9132520,
-      "step": 567
-    },
-    {
-      "epoch": 0.03978740357421174,
-      "grad_norm": 3.787644386291504,
-      "learning_rate": 9.602499474605956e-05,
-      "loss": 0.9914,
-      "num_input_tokens_seen": 9148856,
-      "step": 568
-    },
-    {
-      "epoch": 0.039857451819940985,
-      "grad_norm": 4.630245685577393,
-      "learning_rate": 9.601799649737303e-05,
-      "loss": 1.4135,
-      "num_input_tokens_seen": 9164888,
-      "step": 569
-    },
-    {
-      "epoch": 0.03992750006567023,
-      "grad_norm": 4.063969135284424,
-      "learning_rate": 9.601099824868652e-05,
-      "loss": 1.1312,
-      "num_input_tokens_seen": 9181272,
-      "step": 570
-    },
-    {
-      "epoch": 0.039997548311399476,
-      "grad_norm": 4.2443413734436035,
-      "learning_rate": 9.6004e-05,
-      "loss": 1.1627,
-      "num_input_tokens_seen": 9197344,
-      "step": 571
-    },
-    {
-      "epoch": 0.04006759655712872,
-      "grad_norm": 4.396352767944336,
-      "learning_rate": 9.599700175131349e-05,
-      "loss": 1.1222,
-      "num_input_tokens_seen": 9212312,
-      "step": 572
-    },
-    {
-      "epoch": 0.04013764480285797,
-      "grad_norm": 4.364585876464844,
-      "learning_rate": 9.599000350262697e-05,
-      "loss": 1.0522,
-      "num_input_tokens_seen": 9228696,
-      "step": 573
-    },
-    {
-      "epoch": 0.04020769304858721,
-      "grad_norm": 3.9348409175872803,
-      "learning_rate": 9.598300525394046e-05,
-      "loss": 1.1375,
-      "num_input_tokens_seen": 9245080,
-      "step": 574
-    },
-    {
-      "epoch": 0.04027774129431646,
-      "grad_norm": 4.051416873931885,
-      "learning_rate": 9.597600700525395e-05,
-      "loss": 1.0265,
-      "num_input_tokens_seen": 9260752,
-      "step": 575
-    },
-    {
-      "epoch": 0.040347789540045705,
-      "grad_norm": 4.661770820617676,
-      "learning_rate": 9.596900875656743e-05,
-      "loss": 1.192,
-      "num_input_tokens_seen": 9276792,
-      "step": 576
-    },
-    {
-      "epoch": 0.04041783778577495,
-      "grad_norm": 4.378422260284424,
-      "learning_rate": 9.596201050788092e-05,
-      "loss": 1.0497,
-      "num_input_tokens_seen": 9292768,
-      "step": 577
-    },
-    {
-      "epoch": 0.040487886031504196,
-      "grad_norm": 4.4690399169921875,
-      "learning_rate": 9.595501225919439e-05,
-      "loss": 1.2398,
-      "num_input_tokens_seen": 9309152,
-      "step": 578
-    },
-    {
-      "epoch": 0.04055793427723344,
-      "grad_norm": 4.1711273193359375,
-      "learning_rate": 9.594801401050788e-05,
-      "loss": 1.097,
-      "num_input_tokens_seen": 9325536,
-      "step": 579
-    },
-    {
-      "epoch": 0.040627982522962694,
-      "grad_norm": 3.8115949630737305,
-      "learning_rate": 9.594101576182137e-05,
-      "loss": 1.0317,
-      "num_input_tokens_seen": 9341920,
-      "step": 580
-    },
-    {
-      "epoch": 0.04069803076869194,
-      "grad_norm": 4.072190284729004,
-      "learning_rate": 9.593401751313486e-05,
-      "loss": 1.0649,
-      "num_input_tokens_seen": 9357904,
-      "step": 581
-    },
-    {
-      "epoch": 0.040768079014421185,
-      "grad_norm": 3.895766258239746,
-      "learning_rate": 9.592701926444835e-05,
-      "loss": 1.1906,
-      "num_input_tokens_seen": 9373496,
-      "step": 582
-    },
-    {
-      "epoch": 0.04083812726015043,
-      "grad_norm": 4.026490688323975,
-      "learning_rate": 9.592002101576182e-05,
-      "loss": 0.9913,
-      "num_input_tokens_seen": 9389824,
-      "step": 583
-    },
-    {
-      "epoch": 0.040908175505879676,
-      "grad_norm": 3.612987518310547,
-      "learning_rate": 9.591302276707531e-05,
-      "loss": 0.9376,
-      "num_input_tokens_seen": 9406208,
-      "step": 584
-    },
-    {
-      "epoch": 0.04097822375160892,
-      "grad_norm": 4.4619646072387695,
-      "learning_rate": 9.59060245183888e-05,
-      "loss": 1.2198,
-      "num_input_tokens_seen": 9422592,
-      "step": 585
-    },
-    {
-      "epoch": 0.04104827199733817,
-      "grad_norm": 3.990372896194458,
-      "learning_rate": 9.589902626970229e-05,
-      "loss": 1.082,
-      "num_input_tokens_seen": 9438816,
-      "step": 586
-    },
-    {
-      "epoch": 0.04111832024306741,
-      "grad_norm": 3.7697947025299072,
-      "learning_rate": 9.589202802101577e-05,
-      "loss": 1.0173,
-      "num_input_tokens_seen": 9455200,
-      "step": 587
-    },
-    {
-      "epoch": 0.04118836848879666,
-      "grad_norm": 4.066056728363037,
-      "learning_rate": 9.588502977232925e-05,
-      "loss": 1.124,
-      "num_input_tokens_seen": 9471320,
-      "step": 588
-    },
-    {
-      "epoch": 0.041258416734525905,
-      "grad_norm": 3.913506507873535,
-      "learning_rate": 9.587803152364274e-05,
-      "loss": 1.0501,
-      "num_input_tokens_seen": 9487304,
-      "step": 589
-    },
-    {
-      "epoch": 0.04132846498025515,
-      "grad_norm": 3.9049429893493652,
-      "learning_rate": 9.587103327495623e-05,
-      "loss": 1.0563,
-      "num_input_tokens_seen": 9503688,
-      "step": 590
-    },
-    {
-      "epoch": 0.041398513225984396,
-      "grad_norm": 4.316978454589844,
-      "learning_rate": 9.58640350262697e-05,
-      "loss": 1.1333,
-      "num_input_tokens_seen": 9519488,
-      "step": 591
-    },
-    {
-      "epoch": 0.04146856147171364,
-      "grad_norm": 3.7818517684936523,
-      "learning_rate": 9.585703677758319e-05,
-      "loss": 1.0537,
-      "num_input_tokens_seen": 9535872,
-      "step": 592
-    },
-    {
-      "epoch": 0.04153860971744289,
-      "grad_norm": 3.8751401901245117,
-      "learning_rate": 9.585003852889668e-05,
-      "loss": 1.1745,
-      "num_input_tokens_seen": 9551928,
-      "step": 593
-    },
-    {
-      "epoch": 0.04160865796317213,
-      "grad_norm": 4.357265949249268,
-      "learning_rate": 9.584304028021017e-05,
-      "loss": 1.1154,
-      "num_input_tokens_seen": 9568312,
-      "step": 594
-    },
-    {
-      "epoch": 0.04167870620890138,
-      "grad_norm": 4.184159755706787,
-      "learning_rate": 9.583604203152366e-05,
-      "loss": 1.125,
-      "num_input_tokens_seen": 9583968,
-      "step": 595
-    },
-    {
-      "epoch": 0.041748754454630624,
-      "grad_norm": 3.9540369510650635,
-      "learning_rate": 9.582904378283713e-05,
-      "loss": 1.2032,
-      "num_input_tokens_seen": 9600152,
-      "step": 596
-    },
-    {
-      "epoch": 0.04181880270035987,
-      "grad_norm": 4.401122093200684,
-      "learning_rate": 9.582204553415062e-05,
-      "loss": 1.4808,
-      "num_input_tokens_seen": 9615632,
-      "step": 597
-    },
-    {
-      "epoch": 0.04188885094608912,
-      "grad_norm": 4.418131351470947,
-      "learning_rate": 9.58150472854641e-05,
-      "loss": 1.0077,
-      "num_input_tokens_seen": 9631712,
-      "step": 598
-    },
-    {
-      "epoch": 0.04195889919181837,
-      "grad_norm": 4.362226963043213,
-      "learning_rate": 9.580804903677758e-05,
-      "loss": 1.1614,
-      "num_input_tokens_seen": 9648096,
-      "step": 599
-    },
-    {
-      "epoch": 0.04202894743754761,
-      "grad_norm": 4.051177024841309,
-      "learning_rate": 9.580105078809107e-05,
-      "loss": 1.0718,
-      "num_input_tokens_seen": 9663792,
-      "step": 600
-    },
-    {
-      "epoch": 0.04202894743754761,
-      "eval_loss": 1.1809133291244507,
-      "eval_runtime": 0.2062,
-      "eval_samples_per_second": 4.849,
-      "eval_steps_per_second": 4.849,
-      "num_input_tokens_seen": 9663792,
-      "step": 600
-    },
-    {
-      "epoch": 0.04209899568327686,
-      "grad_norm": 4.478739261627197,
-      "learning_rate": 9.579405253940456e-05,
-      "loss": 1.1963,
-      "num_input_tokens_seen": 9680176,
-      "step": 601
-    },
-    {
-      "epoch": 0.042169043929006104,
-      "grad_norm": 4.05004358291626,
-      "learning_rate": 9.578705429071805e-05,
-      "loss": 1.1005,
-      "num_input_tokens_seen": 9696560,
-      "step": 602
-    },
-    {
-      "epoch": 0.04223909217473535,
-      "grad_norm": 4.092396259307861,
-      "learning_rate": 9.578005604203152e-05,
-      "loss": 1.1796,
-      "num_input_tokens_seen": 9712944,
-      "step": 603
-    },
-    {
-      "epoch": 0.042309140420464596,
-      "grad_norm": 4.428014278411865,
-      "learning_rate": 9.577305779334501e-05,
-      "loss": 0.9734,
-      "num_input_tokens_seen": 9729096,
-      "step": 604
-    },
-    {
-      "epoch": 0.04237918866619384,
-      "grad_norm": 4.202315807342529,
-      "learning_rate": 9.576605954465849e-05,
-      "loss": 1.0502,
-      "num_input_tokens_seen": 9745480,
-      "step": 605
-    },
-    {
-      "epoch": 0.04244923691192309,
-      "grad_norm": 3.7633514404296875,
-      "learning_rate": 9.575906129597198e-05,
-      "loss": 0.9218,
-      "num_input_tokens_seen": 9761272,
-      "step": 606
-    },
-    {
-      "epoch": 0.04251928515765233,
-      "grad_norm": 4.170671463012695,
-      "learning_rate": 9.575206304728548e-05,
-      "loss": 1.1196,
-      "num_input_tokens_seen": 9777656,
-      "step": 607
-    },
-    {
-      "epoch": 0.04258933340338158,
-      "grad_norm": 4.20021915435791,
-      "learning_rate": 9.574506479859895e-05,
-      "loss": 1.1146,
-      "num_input_tokens_seen": 9794032,
-      "step": 608
-    },
-    {
-      "epoch": 0.042659381649110824,
-      "grad_norm": 4.437755107879639,
-      "learning_rate": 9.573806654991244e-05,
-      "loss": 1.0911,
-      "num_input_tokens_seen": 9809936,
-      "step": 609
-    },
-    {
-      "epoch": 0.04272942989484007,
-      "grad_norm": 4.417452335357666,
-      "learning_rate": 9.573106830122592e-05,
-      "loss": 1.2079,
-      "num_input_tokens_seen": 9825232,
-      "step": 610
-    },
-    {
-      "epoch": 0.042799478140569315,
-      "grad_norm": 4.144030570983887,
-      "learning_rate": 9.57240700525394e-05,
-      "loss": 1.1229,
-      "num_input_tokens_seen": 9840648,
-      "step": 611
-    },
-    {
-      "epoch": 0.04286952638629856,
-      "grad_norm": 3.991605043411255,
-      "learning_rate": 9.57170718038529e-05,
-      "loss": 1.0762,
-      "num_input_tokens_seen": 9857032,
-      "step": 612
-    },
-    {
-      "epoch": 0.042939574632027806,
-      "grad_norm": 4.516556262969971,
-      "learning_rate": 9.571007355516638e-05,
-      "loss": 1.3056,
-      "num_input_tokens_seen": 9872328,
-      "step": 613
-    },
-    {
-      "epoch": 0.04300962287775705,
-      "grad_norm": 4.030200481414795,
-      "learning_rate": 9.570307530647987e-05,
-      "loss": 0.9493,
-      "num_input_tokens_seen": 9887832,
-      "step": 614
-    },
-    {
-      "epoch": 0.043079671123486304,
-      "grad_norm": 4.345893859863281,
-      "learning_rate": 9.569607705779335e-05,
-      "loss": 1.2707,
-      "num_input_tokens_seen": 9904216,
-      "step": 615
-    },
-    {
-      "epoch": 0.04314971936921555,
-      "grad_norm": 4.158145427703857,
-      "learning_rate": 9.568907880910684e-05,
-      "loss": 1.0377,
-      "num_input_tokens_seen": 9920072,
-      "step": 616
-    },
-    {
-      "epoch": 0.043219767614944796,
-      "grad_norm": 4.155702590942383,
-      "learning_rate": 9.568208056042032e-05,
-      "loss": 1.091,
-      "num_input_tokens_seen": 9936416,
-      "step": 617
-    },
-    {
-      "epoch": 0.04328981586067404,
-      "grad_norm": 3.76328444480896,
-      "learning_rate": 9.56750823117338e-05,
-      "loss": 1.1011,
-      "num_input_tokens_seen": 9952456,
-      "step": 618
-    },
-    {
-      "epoch": 0.04335986410640329,
-      "grad_norm": 4.252495765686035,
-      "learning_rate": 9.566808406304729e-05,
-      "loss": 1.0616,
-      "num_input_tokens_seen": 9968608,
-      "step": 619
-    },
-    {
-      "epoch": 0.04342991235213253,
-      "grad_norm": 9.254091262817383,
-      "learning_rate": 9.566108581436078e-05,
-      "loss": 1.0315,
-      "num_input_tokens_seen": 9983016,
-      "step": 620
-    },
-    {
-      "epoch": 0.04349996059786178,
-      "grad_norm": 4.028343200683594,
-      "learning_rate": 9.565408756567426e-05,
-      "loss": 1.0667,
-      "num_input_tokens_seen": 9999400,
-      "step": 621
-    },
-    {
-      "epoch": 0.043570008843591024,
-      "grad_norm": 4.051328659057617,
-      "learning_rate": 9.564708931698775e-05,
-      "loss": 1.1375,
-      "num_input_tokens_seen": 10015384,
-      "step": 622
-    },
-    {
-      "epoch": 0.04364005708932027,
-      "grad_norm": 4.495016098022461,
-      "learning_rate": 9.564009106830123e-05,
-      "loss": 1.0691,
-      "num_input_tokens_seen": 10031152,
-      "step": 623
-    },
-    {
-      "epoch": 0.043710105335049515,
-      "grad_norm": 4.876840114593506,
-      "learning_rate": 9.563309281961472e-05,
-      "loss": 1.17,
-      "num_input_tokens_seen": 10047536,
-      "step": 624
-    },
-    {
-      "epoch": 0.04378015358077876,
-      "grad_norm": 4.407329559326172,
-      "learning_rate": 9.562609457092819e-05,
-      "loss": 1.2381,
-      "num_input_tokens_seen": 10063920,
-      "step": 625
-    },
-    {
-      "epoch": 0.043850201826508006,
-      "grad_norm": 4.161394119262695,
-      "learning_rate": 9.561909632224168e-05,
-      "loss": 1.0903,
-      "num_input_tokens_seen": 10079024,
-      "step": 626
-    },
-    {
-      "epoch": 0.04392025007223725,
-      "grad_norm": 4.382974624633789,
-      "learning_rate": 9.561209807355518e-05,
-      "loss": 1.3156,
-      "num_input_tokens_seen": 10095408,
-      "step": 627
-    },
-    {
-      "epoch": 0.0439902983179665,
-      "grad_norm": 4.004157543182373,
-      "learning_rate": 9.560509982486866e-05,
-      "loss": 1.1333,
-      "num_input_tokens_seen": 10111792,
-      "step": 628
-    },
-    {
-      "epoch": 0.04406034656369574,
-      "grad_norm": 3.9019265174865723,
-      "learning_rate": 9.559810157618215e-05,
-      "loss": 1.0948,
-      "num_input_tokens_seen": 10128144,
-      "step": 629
-    },
-    {
-      "epoch": 0.04413039480942499,
-      "grad_norm": 4.410470485687256,
-      "learning_rate": 9.559110332749562e-05,
-      "loss": 1.3219,
-      "num_input_tokens_seen": 10144288,
-      "step": 630
-    },
-    {
-      "epoch": 0.044200443055154234,
-      "grad_norm": 4.233544826507568,
-      "learning_rate": 9.558410507880911e-05,
-      "loss": 0.999,
-      "num_input_tokens_seen": 10160296,
-      "step": 631
-    },
-    {
-      "epoch": 0.04427049130088349,
-      "grad_norm": 4.120091438293457,
-      "learning_rate": 9.557710683012258e-05,
-      "loss": 1.0166,
-      "num_input_tokens_seen": 10176680,
-      "step": 632
-    },
-    {
-      "epoch": 0.04434053954661273,
-      "grad_norm": 5.061972618103027,
-      "learning_rate": 9.557010858143609e-05,
-      "loss": 1.251,
-      "num_input_tokens_seen": 10192088,
-      "step": 633
-    },
-    {
-      "epoch": 0.04441058779234198,
-      "grad_norm": 4.3690948486328125,
-      "learning_rate": 9.556311033274958e-05,
-      "loss": 1.2113,
-      "num_input_tokens_seen": 10208472,
-      "step": 634
-    },
-    {
-      "epoch": 0.044480636038071224,
-      "grad_norm": 3.798710346221924,
-      "learning_rate": 9.555611208406305e-05,
-      "loss": 1.0286,
-      "num_input_tokens_seen": 10224856,
-      "step": 635
-    },
-    {
-      "epoch": 0.04455068428380047,
-      "grad_norm": 4.41818380355835,
-      "learning_rate": 9.554911383537654e-05,
-      "loss": 1.14,
-      "num_input_tokens_seen": 10241200,
-      "step": 636
-    },
-    {
-      "epoch": 0.044620732529529715,
-      "grad_norm": 4.256262302398682,
-      "learning_rate": 9.554211558669001e-05,
-      "loss": 1.3103,
-      "num_input_tokens_seen": 10257584,
-      "step": 637
-    },
-    {
-      "epoch": 0.04469078077525896,
-      "grad_norm": 4.176064968109131,
-      "learning_rate": 9.55351173380035e-05,
-      "loss": 1.1985,
-      "num_input_tokens_seen": 10273760,
-      "step": 638
-    },
-    {
-      "epoch": 0.044760829020988206,
-      "grad_norm": 3.9971530437469482,
-      "learning_rate": 9.552811908931699e-05,
-      "loss": 1.1579,
-      "num_input_tokens_seen": 10290144,
-      "step": 639
-    },
-    {
-      "epoch": 0.04483087726671745,
-      "grad_norm": 4.150514602661133,
-      "learning_rate": 9.552112084063048e-05,
-      "loss": 1.1144,
-      "num_input_tokens_seen": 10306528,
-      "step": 640
-    },
-    {
-      "epoch": 0.0449009255124467,
-      "grad_norm": 4.1868367195129395,
-      "learning_rate": 9.551412259194397e-05,
-      "loss": 1.0099,
-      "num_input_tokens_seen": 10322480,
-      "step": 641
-    },
-    {
-      "epoch": 0.04497097375817594,
-      "grad_norm": 4.409821510314941,
-      "learning_rate": 9.550712434325744e-05,
-      "loss": 1.2574,
-      "num_input_tokens_seen": 10338864,
-      "step": 642
-    },
-    {
-      "epoch": 0.04504102200390519,
-      "grad_norm": 4.500023365020752,
-      "learning_rate": 9.550012609457093e-05,
-      "loss": 1.35,
-      "num_input_tokens_seen": 10355072,
-      "step": 643
-    },
-    {
-      "epoch": 0.045111070249634434,
-      "grad_norm": 10.278129577636719,
-      "learning_rate": 9.549312784588442e-05,
-      "loss": 1.0618,
-      "num_input_tokens_seen": 10371456,
-      "step": 644
-    },
-    {
-      "epoch": 0.04518111849536368,
-      "grad_norm": 3.9800543785095215,
-      "learning_rate": 9.54861295971979e-05,
-      "loss": 1.0341,
-      "num_input_tokens_seen": 10387720,
-      "step": 645
-    },
-    {
-      "epoch": 0.045251166741092926,
-      "grad_norm": 3.855720281600952,
-      "learning_rate": 9.547913134851138e-05,
-      "loss": 1.1323,
-      "num_input_tokens_seen": 10403936,
-      "step": 646
-    },
-    {
-      "epoch": 0.04532121498682217,
-      "grad_norm": 4.719264984130859,
-      "learning_rate": 9.547213309982487e-05,
-      "loss": 1.1407,
-      "num_input_tokens_seen": 10420320,
-      "step": 647
-    },
-    {
-      "epoch": 0.04539126323255142,
-      "grad_norm": 4.6528167724609375,
-      "learning_rate": 9.546513485113836e-05,
-      "loss": 1.1014,
-      "num_input_tokens_seen": 10436704,
-      "step": 648
-    },
-    {
-      "epoch": 0.04546131147828066,
-      "grad_norm": 4.0597028732299805,
-      "learning_rate": 9.545813660245185e-05,
-      "loss": 1.116,
-      "num_input_tokens_seen": 10452592,
-      "step": 649
-    },
-    {
-      "epoch": 0.045531359724009915,
-      "grad_norm": 4.161896705627441,
-      "learning_rate": 9.545113835376533e-05,
-      "loss": 1.1373,
-      "num_input_tokens_seen": 10468976,
-      "step": 650
-    },
-    {
-      "epoch": 0.04560140796973916,
-      "grad_norm": 4.125041961669922,
-      "learning_rate": 9.544414010507881e-05,
-      "loss": 1.0947,
-      "num_input_tokens_seen": 10484584,
-      "step": 651
-    },
-    {
-      "epoch": 0.045671456215468406,
-      "grad_norm": 4.278462886810303,
-      "learning_rate": 9.543714185639229e-05,
-      "loss": 1.1369,
-      "num_input_tokens_seen": 10500504,
-      "step": 652
-    },
-    {
-      "epoch": 0.04574150446119765,
-      "grad_norm": 4.766538619995117,
-      "learning_rate": 9.543014360770579e-05,
-      "loss": 1.1876,
-      "num_input_tokens_seen": 10516472,
-      "step": 653
-    },
-    {
-      "epoch": 0.0458115527069269,
-      "grad_norm": 4.457921504974365,
-      "learning_rate": 9.542314535901928e-05,
-      "loss": 1.0788,
-      "num_input_tokens_seen": 10532272,
-      "step": 654
-    },
-    {
-      "epoch": 0.04588160095265614,
-      "grad_norm": 5.021823883056641,
-      "learning_rate": 9.541614711033275e-05,
-      "loss": 1.1152,
-      "num_input_tokens_seen": 10547696,
-      "step": 655
-    },
-    {
-      "epoch": 0.04595164919838539,
-      "grad_norm": 4.407228469848633,
-      "learning_rate": 9.540914886164624e-05,
-      "loss": 1.0863,
-      "num_input_tokens_seen": 10564080,
-      "step": 656
-    },
-    {
-      "epoch": 0.046021697444114634,
-      "grad_norm": 3.9986062049865723,
-      "learning_rate": 9.540215061295972e-05,
-      "loss": 1.1624,
-      "num_input_tokens_seen": 10580464,
-      "step": 657
-    },
-    {
-      "epoch": 0.04609174568984388,
-      "grad_norm": 7.9165191650390625,
-      "learning_rate": 9.539515236427321e-05,
-      "loss": 1.0809,
-      "num_input_tokens_seen": 10595336,
-      "step": 658
-    },
-    {
-      "epoch": 0.046161793935573125,
-      "grad_norm": 4.357856273651123,
-      "learning_rate": 9.53881541155867e-05,
-      "loss": 1.0324,
-      "num_input_tokens_seen": 10611720,
-      "step": 659
-    },
-    {
-      "epoch": 0.04623184218130237,
-      "grad_norm": 3.8115761280059814,
-      "learning_rate": 9.538115586690018e-05,
-      "loss": 1.1499,
-      "num_input_tokens_seen": 10628104,
-      "step": 660
-    },
-    {
-      "epoch": 0.04630189042703162,
-      "grad_norm": 3.879671096801758,
-      "learning_rate": 9.537415761821367e-05,
-      "loss": 1.0474,
-      "num_input_tokens_seen": 10644096,
-      "step": 661
-    },
-    {
-      "epoch": 0.04637193867276086,
-      "grad_norm": 4.324586391448975,
-      "learning_rate": 9.536715936952715e-05,
-      "loss": 1.1904,
-      "num_input_tokens_seen": 10659408,
-      "step": 662
-    },
-    {
-      "epoch": 0.04644198691849011,
-      "grad_norm": 4.020029067993164,
-      "learning_rate": 9.536016112084064e-05,
-      "loss": 1.0848,
-      "num_input_tokens_seen": 10675792,
-      "step": 663
-    },
-    {
-      "epoch": 0.046512035164219354,
-      "grad_norm": 4.563455581665039,
-      "learning_rate": 9.535316287215411e-05,
-      "loss": 1.1735,
-      "num_input_tokens_seen": 10691632,
-      "step": 664
-    },
-    {
-      "epoch": 0.0465820834099486,
-      "grad_norm": 4.444424629211426,
-      "learning_rate": 9.53461646234676e-05,
-      "loss": 1.258,
-      "num_input_tokens_seen": 10708016,
-      "step": 665
-    },
-    {
-      "epoch": 0.046652131655677845,
-      "grad_norm": 3.9864089488983154,
-      "learning_rate": 9.533916637478109e-05,
-      "loss": 1.1315,
-      "num_input_tokens_seen": 10724176,
-      "step": 666
-    },
-    {
-      "epoch": 0.0467221799014071,
-      "grad_norm": 4.860849857330322,
-      "learning_rate": 9.533216812609458e-05,
-      "loss": 1.2276,
-      "num_input_tokens_seen": 10740560,
-      "step": 667
-    },
-    {
-      "epoch": 0.04679222814713634,
-      "grad_norm": 3.9701120853424072,
-      "learning_rate": 9.532516987740807e-05,
-      "loss": 1.1406,
-      "num_input_tokens_seen": 10756864,
-      "step": 668
-    },
-    {
-      "epoch": 0.04686227639286559,
-      "grad_norm": 3.660257577896118,
-      "learning_rate": 9.531817162872154e-05,
-      "loss": 1.0182,
-      "num_input_tokens_seen": 10773248,
-      "step": 669
-    },
-    {
-      "epoch": 0.046932324638594834,
-      "grad_norm": 3.888510227203369,
-      "learning_rate": 9.531117338003503e-05,
-      "loss": 1.0223,
-      "num_input_tokens_seen": 10789632,
-      "step": 670
-    },
-    {
-      "epoch": 0.04700237288432408,
-      "grad_norm": 4.794105052947998,
-      "learning_rate": 9.530417513134852e-05,
-      "loss": 1.0565,
-      "num_input_tokens_seen": 10804496,
-      "step": 671
-    },
-    {
-      "epoch": 0.047072421130053325,
-      "grad_norm": 4.293116092681885,
-      "learning_rate": 9.529717688266199e-05,
-      "loss": 1.2509,
-      "num_input_tokens_seen": 10819976,
-      "step": 672
-    },
-    {
-      "epoch": 0.04714246937578257,
-      "grad_norm": 5.112069129943848,
-      "learning_rate": 9.52901786339755e-05,
-      "loss": 1.0964,
-      "num_input_tokens_seen": 10836360,
-      "step": 673
-    },
-    {
-      "epoch": 0.04721251762151182,
-      "grad_norm": 3.9091360569000244,
-      "learning_rate": 9.528318038528897e-05,
-      "loss": 1.0647,
-      "num_input_tokens_seen": 10852744,
-      "step": 674
-    },
-    {
-      "epoch": 0.04728256586724106,
-      "grad_norm": 4.032161235809326,
-      "learning_rate": 9.527618213660246e-05,
-      "loss": 1.2362,
-      "num_input_tokens_seen": 10868928,
-      "step": 675
-    },
-    {
-      "epoch": 0.04735261411297031,
-      "grad_norm": 3.931156635284424,
-      "learning_rate": 9.526918388791595e-05,
-      "loss": 1.0571,
-      "num_input_tokens_seen": 10884776,
-      "step": 676
-    },
-    {
-      "epoch": 0.047422662358699554,
-      "grad_norm": 3.9511048793792725,
-      "learning_rate": 9.526218563922942e-05,
-      "loss": 1.0249,
-      "num_input_tokens_seen": 10901160,
-      "step": 677
-    },
-    {
-      "epoch": 0.0474927106044288,
-      "grad_norm": 4.199029445648193,
-      "learning_rate": 9.525518739054291e-05,
-      "loss": 1.2813,
-      "num_input_tokens_seen": 10917544,
-      "step": 678
-    },
-    {
-      "epoch": 0.047562758850158045,
-      "grad_norm": 3.8590247631073,
-      "learning_rate": 9.52481891418564e-05,
-      "loss": 1.02,
-      "num_input_tokens_seen": 10933928,
-      "step": 679
-    },
-    {
-      "epoch": 0.04763280709588729,
-      "grad_norm": 5.530341625213623,
-      "learning_rate": 9.524119089316989e-05,
-      "loss": 1.2316,
-      "num_input_tokens_seen": 10949600,
-      "step": 680
-    },
-    {
-      "epoch": 0.047702855341616536,
-      "grad_norm": 4.17647123336792,
-      "learning_rate": 9.523419264448338e-05,
-      "loss": 1.2985,
-      "num_input_tokens_seen": 10965984,
-      "step": 681
-    },
-    {
-      "epoch": 0.04777290358734578,
-      "grad_norm": 4.250451564788818,
-      "learning_rate": 9.522719439579685e-05,
-      "loss": 1.1638,
-      "num_input_tokens_seen": 10982368,
-      "step": 682
-    },
-    {
-      "epoch": 0.04784295183307503,
-      "grad_norm": 4.132594108581543,
-      "learning_rate": 9.522019614711034e-05,
-      "loss": 0.9638,
-      "num_input_tokens_seen": 10998752,
-      "step": 683
-    },
-    {
-      "epoch": 0.04791300007880428,
-      "grad_norm": 5.863363265991211,
-      "learning_rate": 9.521319789842382e-05,
-      "loss": 1.0736,
-      "num_input_tokens_seen": 11014376,
-      "step": 684
-    },
-    {
-      "epoch": 0.047983048324533525,
-      "grad_norm": 3.740323543548584,
-      "learning_rate": 9.52061996497373e-05,
-      "loss": 0.9958,
-      "num_input_tokens_seen": 11030440,
-      "step": 685
-    },
-    {
-      "epoch": 0.04805309657026277,
-      "grad_norm": 4.927120685577393,
-      "learning_rate": 9.519920140105079e-05,
-      "loss": 1.156,
-      "num_input_tokens_seen": 11046824,
-      "step": 686
-    },
-    {
-      "epoch": 0.04812314481599202,
-      "grad_norm": 4.708818435668945,
-      "learning_rate": 9.519220315236428e-05,
-      "loss": 1.2139,
-      "num_input_tokens_seen": 11063208,
-      "step": 687
-    },
-    {
-      "epoch": 0.04819319306172126,
-      "grad_norm": 3.7547767162323,
-      "learning_rate": 9.518520490367777e-05,
-      "loss": 0.9557,
-      "num_input_tokens_seen": 11079592,
-      "step": 688
-    },
-    {
-      "epoch": 0.04826324130745051,
-      "grad_norm": 4.038534641265869,
-      "learning_rate": 9.517820665499124e-05,
-      "loss": 1.1124,
-      "num_input_tokens_seen": 11095976,
-      "step": 689
-    },
-    {
-      "epoch": 0.048333289553179754,
-      "grad_norm": 4.159554481506348,
-      "learning_rate": 9.517120840630473e-05,
-      "loss": 1.0043,
-      "num_input_tokens_seen": 11112360,
-      "step": 690
-    },
-    {
-      "epoch": 0.048403337798909,
-      "grad_norm": 7.104836463928223,
-      "learning_rate": 9.516421015761821e-05,
-      "loss": 0.9736,
-      "num_input_tokens_seen": 11127800,
-      "step": 691
-    },
-    {
-      "epoch": 0.048473386044638245,
-      "grad_norm": 4.073885917663574,
-      "learning_rate": 9.51572119089317e-05,
-      "loss": 1.1249,
-      "num_input_tokens_seen": 11144184,
-      "step": 692
-    },
-    {
-      "epoch": 0.04854343429036749,
-      "grad_norm": 3.7190351486206055,
-      "learning_rate": 9.51502136602452e-05,
-      "loss": 1.1035,
-      "num_input_tokens_seen": 11160568,
-      "step": 693
-    },
-    {
-      "epoch": 0.048613482536096736,
-      "grad_norm": 4.252142429351807,
-      "learning_rate": 9.514321541155867e-05,
-      "loss": 1.1588,
-      "num_input_tokens_seen": 11176952,
-      "step": 694
-    },
-    {
-      "epoch": 0.04868353078182598,
-      "grad_norm": 4.418105125427246,
-      "learning_rate": 9.513621716287216e-05,
-      "loss": 1.2496,
-      "num_input_tokens_seen": 11193336,
-      "step": 695
-    },
-    {
-      "epoch": 0.04875357902755523,
-      "grad_norm": 4.195918560028076,
-      "learning_rate": 9.512921891418564e-05,
-      "loss": 1.0193,
-      "num_input_tokens_seen": 11209720,
-      "step": 696
-    },
-    {
-      "epoch": 0.04882362727328447,
-      "grad_norm": 5.138080596923828,
-      "learning_rate": 9.512222066549913e-05,
-      "loss": 1.1861,
-      "num_input_tokens_seen": 11225888,
-      "step": 697
-    },
-    {
-      "epoch": 0.04889367551901372,
-      "grad_norm": 4.489223003387451,
-      "learning_rate": 9.511522241681261e-05,
-      "loss": 1.1497,
-      "num_input_tokens_seen": 11241744,
-      "step": 698
-    },
-    {
-      "epoch": 0.048963723764742964,
-      "grad_norm": 3.972590208053589,
-      "learning_rate": 9.51082241681261e-05,
-      "loss": 1.2765,
-      "num_input_tokens_seen": 11257768,
-      "step": 699
-    },
-    {
-      "epoch": 0.04903377201047221,
-      "grad_norm": 13.274886131286621,
-      "learning_rate": 9.510122591943959e-05,
-      "loss": 1.1124,
-      "num_input_tokens_seen": 11273216,
-      "step": 700
-    },
-    {
-      "epoch": 0.049103820256201455,
-      "grad_norm": 3.7899255752563477,
-      "learning_rate": 9.509422767075307e-05,
-      "loss": 1.0445,
-      "num_input_tokens_seen": 11289600,
-      "step": 701
-    },
-    {
-      "epoch": 0.04917386850193071,
-      "grad_norm": 4.226947784423828,
-      "learning_rate": 9.508722942206656e-05,
-      "loss": 1.4313,
-      "num_input_tokens_seen": 11305920,
-      "step": 702
-    },
-    {
-      "epoch": 0.049243916747659953,
-      "grad_norm": 4.098162651062012,
-      "learning_rate": 9.508023117338003e-05,
-      "loss": 0.952,
-      "num_input_tokens_seen": 11322304,
-      "step": 703
-    },
-    {
-      "epoch": 0.0493139649933892,
-      "grad_norm": 3.9205965995788574,
-      "learning_rate": 9.507323292469352e-05,
-      "loss": 1.1648,
-      "num_input_tokens_seen": 11338688,
-      "step": 704
-    },
-    {
-      "epoch": 0.049384013239118445,
-      "grad_norm": 4.06537389755249,
-      "learning_rate": 9.506623467600701e-05,
-      "loss": 1.1295,
-      "num_input_tokens_seen": 11353544,
-      "step": 705
-    },
-    {
-      "epoch": 0.04945406148484769,
-      "grad_norm": 4.309032440185547,
-      "learning_rate": 9.50592364273205e-05,
-      "loss": 1.1475,
-      "num_input_tokens_seen": 11369928,
-      "step": 706
-    },
-    {
-      "epoch": 0.049524109730576936,
-      "grad_norm": 4.320526599884033,
-      "learning_rate": 9.505223817863399e-05,
-      "loss": 1.0102,
-      "num_input_tokens_seen": 11386312,
-      "step": 707
-    },
-    {
-      "epoch": 0.04959415797630618,
-      "grad_norm": 5.025510787963867,
-      "learning_rate": 9.504523992994747e-05,
-      "loss": 1.1182,
-      "num_input_tokens_seen": 11402696,
-      "step": 708
-    },
-    {
-      "epoch": 0.04966420622203543,
-      "grad_norm": 3.9406464099884033,
-      "learning_rate": 9.503824168126095e-05,
-      "loss": 1.068,
-      "num_input_tokens_seen": 11419080,
-      "step": 709
-    },
-    {
-      "epoch": 0.04973425446776467,
-      "grad_norm": 3.9148502349853516,
-      "learning_rate": 9.503124343257444e-05,
-      "loss": 1.1062,
-      "num_input_tokens_seen": 11435464,
-      "step": 710
-    },
-    {
-      "epoch": 0.04980430271349392,
-      "grad_norm": 3.9386026859283447,
-      "learning_rate": 9.502424518388791e-05,
-      "loss": 0.9516,
-      "num_input_tokens_seen": 11451848,
-      "step": 711
-    },
-    {
-      "epoch": 0.049874350959223164,
-      "grad_norm": 3.9537665843963623,
-      "learning_rate": 9.50172469352014e-05,
-      "loss": 1.1372,
-      "num_input_tokens_seen": 11468216,
-      "step": 712
-    },
-    {
-      "epoch": 0.04994439920495241,
-      "grad_norm": 3.97929310798645,
-      "learning_rate": 9.501024868651489e-05,
-      "loss": 1.0705,
-      "num_input_tokens_seen": 11484192,
-      "step": 713
-    },
-    {
-      "epoch": 0.050014447450681655,
-      "grad_norm": 3.9326419830322266,
-      "learning_rate": 9.500325043782838e-05,
-      "loss": 1.0986,
-      "num_input_tokens_seen": 11500576,
-      "step": 714
-    },
-    {
-      "epoch": 0.0500844956964109,
-      "grad_norm": 3.769347667694092,
-      "learning_rate": 9.499625218914187e-05,
-      "loss": 0.9265,
-      "num_input_tokens_seen": 11516960,
-      "step": 715
-    },
-    {
-      "epoch": 0.050154543942140146,
-      "grad_norm": 4.264547348022461,
-      "learning_rate": 9.498925394045534e-05,
-      "loss": 1.3166,
-      "num_input_tokens_seen": 11532616,
-      "step": 716
-    },
-    {
-      "epoch": 0.05022459218786939,
-      "grad_norm": 4.885791778564453,
-      "learning_rate": 9.498225569176883e-05,
-      "loss": 1.0669,
-      "num_input_tokens_seen": 11548552,
-      "step": 717
-    },
-    {
-      "epoch": 0.05029464043359864,
-      "grad_norm": 5.4089741706848145,
-      "learning_rate": 9.49752574430823e-05,
-      "loss": 1.3986,
-      "num_input_tokens_seen": 11564936,
-      "step": 718
-    },
-    {
-      "epoch": 0.05036468867932789,
-      "grad_norm": 4.503393173217773,
-      "learning_rate": 9.496825919439581e-05,
-      "loss": 0.9947,
-      "num_input_tokens_seen": 11580720,
-      "step": 719
-    },
-    {
-      "epoch": 0.050434736925057136,
-      "grad_norm": 4.364518165588379,
-      "learning_rate": 9.49612609457093e-05,
-      "loss": 1.12,
-      "num_input_tokens_seen": 11597104,
-      "step": 720
-    },
-    {
-      "epoch": 0.05050478517078638,
-      "grad_norm": 4.229926109313965,
-      "learning_rate": 9.495426269702277e-05,
-      "loss": 1.098,
-      "num_input_tokens_seen": 11612120,
-      "step": 721
-    },
-    {
-      "epoch": 0.05057483341651563,
-      "grad_norm": 4.477171897888184,
-      "learning_rate": 9.494726444833626e-05,
-      "loss": 1.1565,
-      "num_input_tokens_seen": 11627000,
-      "step": 722
-    },
-    {
-      "epoch": 0.05064488166224487,
-      "grad_norm": 4.071736812591553,
-      "learning_rate": 9.494026619964973e-05,
-      "loss": 1.2951,
-      "num_input_tokens_seen": 11643256,
-      "step": 723
-    },
-    {
-      "epoch": 0.05071492990797412,
-      "grad_norm": 4.219758033752441,
-      "learning_rate": 9.493326795096322e-05,
-      "loss": 1.1408,
-      "num_input_tokens_seen": 11659424,
-      "step": 724
-    },
-    {
-      "epoch": 0.050784978153703364,
-      "grad_norm": 4.108195781707764,
-      "learning_rate": 9.492626970227671e-05,
-      "loss": 0.9847,
-      "num_input_tokens_seen": 11675808,
-      "step": 725
-    },
-    {
-      "epoch": 0.05085502639943261,
-      "grad_norm": 3.964359760284424,
-      "learning_rate": 9.49192714535902e-05,
-      "loss": 1.0935,
-      "num_input_tokens_seen": 11691760,
-      "step": 726
-    },
-    {
-      "epoch": 0.050925074645161855,
-      "grad_norm": 4.585779190063477,
-      "learning_rate": 9.491227320490369e-05,
-      "loss": 1.1561,
-      "num_input_tokens_seen": 11706600,
-      "step": 727
-    },
-    {
-      "epoch": 0.0509951228908911,
-      "grad_norm": 3.8540141582489014,
-      "learning_rate": 9.490527495621716e-05,
-      "loss": 1.0163,
-      "num_input_tokens_seen": 11722984,
-      "step": 728
-    },
-    {
-      "epoch": 0.051065171136620346,
-      "grad_norm": 4.138955593109131,
-      "learning_rate": 9.489827670753065e-05,
-      "loss": 1.2842,
-      "num_input_tokens_seen": 11738968,
-      "step": 729
-    },
-    {
-      "epoch": 0.05113521938234959,
-      "grad_norm": 4.138274192810059,
-      "learning_rate": 9.489127845884413e-05,
-      "loss": 1.1452,
-      "num_input_tokens_seen": 11754952,
-      "step": 730
-    },
-    {
-      "epoch": 0.05120526762807884,
-      "grad_norm": 4.374305248260498,
-      "learning_rate": 9.488428021015762e-05,
-      "loss": 1.3622,
-      "num_input_tokens_seen": 11770832,
-      "step": 731
-    },
-    {
-      "epoch": 0.05127531587380808,
-      "grad_norm": 4.242674350738525,
-      "learning_rate": 9.48772819614711e-05,
-      "loss": 1.1914,
-      "num_input_tokens_seen": 11786872,
-      "step": 732
-    },
-    {
-      "epoch": 0.05134536411953733,
-      "grad_norm": 4.173389911651611,
-      "learning_rate": 9.48702837127846e-05,
-      "loss": 1.1853,
-      "num_input_tokens_seen": 11803256,
-      "step": 733
-    },
-    {
-      "epoch": 0.051415412365266575,
-      "grad_norm": 4.014588356018066,
-      "learning_rate": 9.486328546409808e-05,
-      "loss": 1.0436,
-      "num_input_tokens_seen": 11819608,
-      "step": 734
-    },
-    {
-      "epoch": 0.05148546061099582,
-      "grad_norm": 4.759418964385986,
-      "learning_rate": 9.485628721541157e-05,
-      "loss": 1.1605,
-      "num_input_tokens_seen": 11834296,
-      "step": 735
-    },
-    {
-      "epoch": 0.05155550885672507,
-      "grad_norm": 4.258687973022461,
-      "learning_rate": 9.484928896672505e-05,
-      "loss": 1.2993,
-      "num_input_tokens_seen": 11849728,
-      "step": 736
-    },
-    {
-      "epoch": 0.05162555710245432,
-      "grad_norm": 4.690395832061768,
-      "learning_rate": 9.484229071803853e-05,
-      "loss": 1.0655,
-      "num_input_tokens_seen": 11866112,
-      "step": 737
-    },
-    {
-      "epoch": 0.051695605348183564,
-      "grad_norm": 4.373327255249023,
-      "learning_rate": 9.483529246935201e-05,
-      "loss": 1.1364,
-      "num_input_tokens_seen": 11881960,
-      "step": 738
-    },
-    {
-      "epoch": 0.05176565359391281,
-      "grad_norm": 4.008789539337158,
-      "learning_rate": 9.482829422066551e-05,
-      "loss": 1.1174,
-      "num_input_tokens_seen": 11897936,
-      "step": 739
-    },
-    {
-      "epoch": 0.051835701839642055,
-      "grad_norm": 4.391345977783203,
-      "learning_rate": 9.482129597197899e-05,
-      "loss": 1.2045,
-      "num_input_tokens_seen": 11914320,
-      "step": 740
-    },
-    {
-      "epoch": 0.0519057500853713,
-      "grad_norm": 4.119503021240234,
-      "learning_rate": 9.481429772329248e-05,
-      "loss": 0.927,
-      "num_input_tokens_seen": 11930440,
-      "step": 741
-    },
-    {
-      "epoch": 0.051975798331100546,
-      "grad_norm": 4.186014175415039,
-      "learning_rate": 9.480729947460596e-05,
-      "loss": 1.1583,
-      "num_input_tokens_seen": 11946720,
-      "step": 742
-    },
-    {
-      "epoch": 0.05204584657682979,
-      "grad_norm": 4.119131088256836,
-      "learning_rate": 9.480030122591944e-05,
-      "loss": 1.0792,
-      "num_input_tokens_seen": 11962360,
-      "step": 743
-    },
-    {
-      "epoch": 0.05211589482255904,
-      "grad_norm": 3.921030044555664,
-      "learning_rate": 9.479330297723293e-05,
-      "loss": 0.9966,
-      "num_input_tokens_seen": 11978744,
-      "step": 744
-    },
-    {
-      "epoch": 0.05218594306828828,
-      "grad_norm": 3.806251049041748,
-      "learning_rate": 9.478630472854642e-05,
-      "loss": 1.1207,
-      "num_input_tokens_seen": 11994912,
-      "step": 745
-    },
-    {
-      "epoch": 0.05225599131401753,
-      "grad_norm": 4.508687973022461,
-      "learning_rate": 9.47793064798599e-05,
-      "loss": 1.1038,
-      "num_input_tokens_seen": 12011296,
-      "step": 746
-    },
-    {
-      "epoch": 0.052326039559746775,
-      "grad_norm": 4.458346843719482,
-      "learning_rate": 9.47723082311734e-05,
-      "loss": 1.2878,
-      "num_input_tokens_seen": 12027408,
-      "step": 747
-    },
-    {
-      "epoch": 0.05239608780547602,
-      "grad_norm": 5.779678821563721,
-      "learning_rate": 9.476530998248687e-05,
-      "loss": 1.2722,
-      "num_input_tokens_seen": 12043792,
-      "step": 748
-    },
-    {
-      "epoch": 0.052466136051205266,
-      "grad_norm": 4.621145725250244,
-      "learning_rate": 9.475831173380036e-05,
-      "loss": 1.2636,
-      "num_input_tokens_seen": 12059856,
-      "step": 749
-    },
-    {
-      "epoch": 0.05253618429693451,
-      "grad_norm": 4.276626110076904,
-      "learning_rate": 9.475131348511383e-05,
-      "loss": 1.3378,
-      "num_input_tokens_seen": 12076240,
-      "step": 750
-    },
-    {
-      "epoch": 0.05260623254266376,
-      "grad_norm": 4.533468246459961,
-      "learning_rate": 9.474431523642732e-05,
-      "loss": 0.921,
-      "num_input_tokens_seen": 12092416,
-      "step": 751
-    },
-    {
-      "epoch": 0.052676280788393,
-      "grad_norm": 4.626596927642822,
-      "learning_rate": 9.473731698774081e-05,
-      "loss": 1.2807,
-      "num_input_tokens_seen": 12108664,
-      "step": 752
-    },
-    {
-      "epoch": 0.052746329034122255,
-      "grad_norm": 4.3372907638549805,
-      "learning_rate": 9.47303187390543e-05,
-      "loss": 1.2754,
-      "num_input_tokens_seen": 12125048,
-      "step": 753
-    },
-    {
-      "epoch": 0.0528163772798515,
-      "grad_norm": 3.6576266288757324,
-      "learning_rate": 9.472332049036779e-05,
-      "loss": 0.8487,
-      "num_input_tokens_seen": 12141296,
-      "step": 754
-    },
-    {
-      "epoch": 0.052886425525580746,
-      "grad_norm": 3.8973164558410645,
-      "learning_rate": 9.471632224168126e-05,
-      "loss": 1.1211,
-      "num_input_tokens_seen": 12157544,
-      "step": 755
-    },
-    {
-      "epoch": 0.05295647377130999,
-      "grad_norm": 3.9059019088745117,
-      "learning_rate": 9.470932399299475e-05,
-      "loss": 1.2484,
-      "num_input_tokens_seen": 12173928,
-      "step": 756
-    },
-    {
-      "epoch": 0.05302652201703924,
-      "grad_norm": 4.133029937744141,
-      "learning_rate": 9.470232574430822e-05,
-      "loss": 1.0762,
-      "num_input_tokens_seen": 12189864,
-      "step": 757
-    },
-    {
-      "epoch": 0.05309657026276848,
-      "grad_norm": 3.8380961418151855,
-      "learning_rate": 9.469532749562171e-05,
-      "loss": 0.9938,
-      "num_input_tokens_seen": 12206248,
-      "step": 758
-    },
-    {
-      "epoch": 0.05316661850849773,
-      "grad_norm": 4.753637790679932,
-      "learning_rate": 9.468832924693522e-05,
-      "loss": 1.1272,
-      "num_input_tokens_seen": 12222632,
-      "step": 759
-    },
-    {
-      "epoch": 0.053236666754226974,
-      "grad_norm": 4.704193592071533,
-      "learning_rate": 9.468133099824869e-05,
-      "loss": 1.2276,
-      "num_input_tokens_seen": 12239016,
-      "step": 760
-    },
-    {
-      "epoch": 0.05330671499995622,
-      "grad_norm": 3.870870351791382,
-      "learning_rate": 9.467433274956218e-05,
-      "loss": 0.916,
-      "num_input_tokens_seen": 12254784,
-      "step": 761
-    },
-    {
-      "epoch": 0.053376763245685466,
-      "grad_norm": 3.8597328662872314,
-      "learning_rate": 9.466733450087567e-05,
-      "loss": 0.9871,
-      "num_input_tokens_seen": 12271160,
-      "step": 762
-    },
-    {
-      "epoch": 0.05344681149141471,
-      "grad_norm": 3.7109553813934326,
-      "learning_rate": 9.466033625218914e-05,
-      "loss": 1.1248,
-      "num_input_tokens_seen": 12286944,
-      "step": 763
-    },
-    {
-      "epoch": 0.05351685973714396,
-      "grad_norm": 3.985595464706421,
-      "learning_rate": 9.465333800350263e-05,
-      "loss": 1.0524,
-      "num_input_tokens_seen": 12303312,
-      "step": 764
-    },
-    {
-      "epoch": 0.0535869079828732,
-      "grad_norm": 3.797247886657715,
-      "learning_rate": 9.464633975481612e-05,
-      "loss": 1.0799,
-      "num_input_tokens_seen": 12319696,
-      "step": 765
-    },
-    {
-      "epoch": 0.05365695622860245,
-      "grad_norm": 4.88303279876709,
-      "learning_rate": 9.463934150612961e-05,
-      "loss": 1.2865,
-      "num_input_tokens_seen": 12335448,
-      "step": 766
-    },
-    {
-      "epoch": 0.053727004474331694,
-      "grad_norm": 4.273831367492676,
-      "learning_rate": 9.463234325744308e-05,
-      "loss": 1.1724,
-      "num_input_tokens_seen": 12351720,
-      "step": 767
-    },
-    {
-      "epoch": 0.05379705272006094,
-      "grad_norm": 3.9505984783172607,
-      "learning_rate": 9.462534500875657e-05,
-      "loss": 1.1478,
-      "num_input_tokens_seen": 12368104,
-      "step": 768
-    },
-    {
-      "epoch": 0.053867100965790185,
-      "grad_norm": 4.20963191986084,
-      "learning_rate": 9.461834676007006e-05,
-      "loss": 1.1018,
-      "num_input_tokens_seen": 12384488,
-      "step": 769
-    },
-    {
-      "epoch": 0.05393714921151943,
-      "grad_norm": 4.106869220733643,
-      "learning_rate": 9.461134851138354e-05,
-      "loss": 1.1097,
-      "num_input_tokens_seen": 12400128,
-      "step": 770
-    },
-    {
-      "epoch": 0.05400719745724868,
-      "grad_norm": 4.28592586517334,
-      "learning_rate": 9.460435026269702e-05,
-      "loss": 1.036,
-      "num_input_tokens_seen": 12416512,
-      "step": 771
-    },
-    {
-      "epoch": 0.05407724570297793,
-      "grad_norm": 3.821927070617676,
-      "learning_rate": 9.459735201401051e-05,
-      "loss": 1.1215,
-      "num_input_tokens_seen": 12432896,
-      "step": 772
-    },
-    {
-      "epoch": 0.054147293948707174,
-      "grad_norm": 4.14424467086792,
-      "learning_rate": 9.4590353765324e-05,
-      "loss": 1.0092,
-      "num_input_tokens_seen": 12449208,
-      "step": 773
-    },
-    {
-      "epoch": 0.05421734219443642,
-      "grad_norm": 4.610694885253906,
-      "learning_rate": 9.458335551663749e-05,
-      "loss": 1.2265,
-      "num_input_tokens_seen": 12464128,
-      "step": 774
-    },
-    {
-      "epoch": 0.054287390440165666,
-      "grad_norm": 4.410182952880859,
-      "learning_rate": 9.457635726795097e-05,
-      "loss": 1.1904,
-      "num_input_tokens_seen": 12479728,
-      "step": 775
-    },
-    {
-      "epoch": 0.05435743868589491,
-      "grad_norm": 4.096780300140381,
-      "learning_rate": 9.456935901926445e-05,
-      "loss": 1.2317,
-      "num_input_tokens_seen": 12495720,
-      "step": 776
-    },
-    {
-      "epoch": 0.05442748693162416,
-      "grad_norm": 4.028350830078125,
-      "learning_rate": 9.456236077057793e-05,
-      "loss": 1.1825,
-      "num_input_tokens_seen": 12511480,
-      "step": 777
-    },
-    {
-      "epoch": 0.0544975351773534,
-      "grad_norm": 5.264276504516602,
-      "learning_rate": 9.455536252189142e-05,
-      "loss": 1.057,
-      "num_input_tokens_seen": 12527864,
-      "step": 778
-    },
-    {
-      "epoch": 0.05456758342308265,
-      "grad_norm": 4.371725082397461,
-      "learning_rate": 9.454836427320492e-05,
-      "loss": 1.1625,
-      "num_input_tokens_seen": 12544168,
-      "step": 779
-    },
-    {
-      "epoch": 0.054637631668811894,
-      "grad_norm": 4.692862510681152,
-      "learning_rate": 9.45413660245184e-05,
-      "loss": 1.2211,
-      "num_input_tokens_seen": 12560552,
-      "step": 780
-    },
-    {
-      "epoch": 0.05470767991454114,
-      "grad_norm": 3.7462823390960693,
-      "learning_rate": 9.453436777583188e-05,
-      "loss": 1.0815,
-      "num_input_tokens_seen": 12576936,
-      "step": 781
-    },
-    {
-      "epoch": 0.054777728160270385,
-      "grad_norm": 4.161571025848389,
-      "learning_rate": 9.452736952714536e-05,
-      "loss": 0.9788,
-      "num_input_tokens_seen": 12593040,
-      "step": 782
-    },
-    {
-      "epoch": 0.05484777640599963,
-      "grad_norm": 3.96793532371521,
-      "learning_rate": 9.452037127845885e-05,
-      "loss": 1.1396,
-      "num_input_tokens_seen": 12609424,
-      "step": 783
-    },
-    {
-      "epoch": 0.054917824651728876,
-      "grad_norm": 4.183755874633789,
-      "learning_rate": 9.451337302977232e-05,
-      "loss": 1.0868,
-      "num_input_tokens_seen": 12625312,
-      "step": 784
-    },
-    {
-      "epoch": 0.05498787289745812,
-      "grad_norm": 4.506673336029053,
-      "learning_rate": 9.450637478108582e-05,
-      "loss": 1.1112,
-      "num_input_tokens_seen": 12641696,
-      "step": 785
-    },
-    {
-      "epoch": 0.05505792114318737,
-      "grad_norm": 3.8601651191711426,
-      "learning_rate": 9.449937653239931e-05,
-      "loss": 1.2149,
-      "num_input_tokens_seen": 12658080,
-      "step": 786
-    },
-    {
-      "epoch": 0.05512796938891661,
-      "grad_norm": 5.190856456756592,
-      "learning_rate": 9.449237828371279e-05,
-      "loss": 1.2661,
-      "num_input_tokens_seen": 12673032,
-      "step": 787
-    },
-    {
-      "epoch": 0.055198017634645866,
-      "grad_norm": 4.323099136352539,
-      "learning_rate": 9.448538003502628e-05,
-      "loss": 1.139,
-      "num_input_tokens_seen": 12689064,
-      "step": 788
-    },
-    {
-      "epoch": 0.05526806588037511,
-      "grad_norm": 4.271193981170654,
-      "learning_rate": 9.447838178633976e-05,
-      "loss": 1.037,
-      "num_input_tokens_seen": 12705448,
-      "step": 789
-    },
-    {
-      "epoch": 0.05533811412610436,
-      "grad_norm": 3.793525218963623,
-      "learning_rate": 9.447138353765324e-05,
-      "loss": 1.0265,
-      "num_input_tokens_seen": 12721832,
-      "step": 790
-    },
-    {
-      "epoch": 0.0554081623718336,
-      "grad_norm": 3.747575283050537,
-      "learning_rate": 9.446438528896673e-05,
-      "loss": 0.9567,
-      "num_input_tokens_seen": 12738216,
-      "step": 791
-    },
-    {
-      "epoch": 0.05547821061756285,
-      "grad_norm": 4.222849369049072,
-      "learning_rate": 9.445738704028022e-05,
-      "loss": 1.1859,
-      "num_input_tokens_seen": 12754600,
-      "step": 792
-    },
-    {
-      "epoch": 0.055548258863292094,
-      "grad_norm": 9.102783203125,
-      "learning_rate": 9.44503887915937e-05,
-      "loss": 1.0361,
-      "num_input_tokens_seen": 12770568,
-      "step": 793
-    },
-    {
-      "epoch": 0.05561830710902134,
-      "grad_norm": 4.4447808265686035,
-      "learning_rate": 9.444339054290718e-05,
-      "loss": 1.2908,
-      "num_input_tokens_seen": 12785768,
-      "step": 794
-    },
-    {
-      "epoch": 0.055688355354750585,
-      "grad_norm": 4.038604259490967,
-      "learning_rate": 9.443639229422067e-05,
-      "loss": 0.9294,
-      "num_input_tokens_seen": 12801704,
-      "step": 795
-    },
-    {
-      "epoch": 0.05575840360047983,
-      "grad_norm": 4.492194652557373,
-      "learning_rate": 9.442939404553416e-05,
-      "loss": 1.0466,
-      "num_input_tokens_seen": 12818088,
-      "step": 796
-    },
-    {
-      "epoch": 0.055828451846209076,
-      "grad_norm": 3.978029489517212,
-      "learning_rate": 9.442239579684763e-05,
-      "loss": 1.1719,
-      "num_input_tokens_seen": 12834432,
-      "step": 797
-    },
-    {
-      "epoch": 0.05589850009193832,
-      "grad_norm": 4.014431476593018,
-      "learning_rate": 9.441539754816112e-05,
-      "loss": 1.1222,
-      "num_input_tokens_seen": 12850816,
-      "step": 798
-    },
-    {
-      "epoch": 0.05596854833766757,
-      "grad_norm": 4.0948638916015625,
-      "learning_rate": 9.440839929947461e-05,
-      "loss": 1.2013,
-      "num_input_tokens_seen": 12867200,
-      "step": 799
-    },
-    {
-      "epoch": 0.05603859658339681,
-      "grad_norm": 4.18120813369751,
-      "learning_rate": 9.44014010507881e-05,
-      "loss": 0.9403,
-      "num_input_tokens_seen": 12883072,
-      "step": 800
-    },
-    {
-      "epoch": 0.05603859658339681,
-      "eval_loss": 1.1718552112579346,
-      "eval_runtime": 0.2039,
-      "eval_samples_per_second": 4.905,
-      "eval_steps_per_second": 4.905,
-      "num_input_tokens_seen": 12883072,
-      "step": 800
-    },
-    {
-      "epoch": 0.05610864482912606,
-      "grad_norm": 4.425891399383545,
-      "learning_rate": 9.439440280210159e-05,
-      "loss": 1.0435,
-      "num_input_tokens_seen": 12899456,
-      "step": 801
-    },
-    {
-      "epoch": 0.056178693074855304,
-      "grad_norm": 4.319190979003906,
-      "learning_rate": 9.438740455341506e-05,
-      "loss": 1.2612,
-      "num_input_tokens_seen": 12915840,
-      "step": 802
-    },
-    {
-      "epoch": 0.05624874132058455,
-      "grad_norm": 4.28010892868042,
-      "learning_rate": 9.438040630472855e-05,
-      "loss": 1.0853,
-      "num_input_tokens_seen": 12932096,
-      "step": 803
-    },
-    {
-      "epoch": 0.056318789566313796,
-      "grad_norm": 3.9454870223999023,
-      "learning_rate": 9.437340805604203e-05,
-      "loss": 1.055,
-      "num_input_tokens_seen": 12948208,
-      "step": 804
-    },
-    {
-      "epoch": 0.05638883781204305,
-      "grad_norm": 4.009400367736816,
-      "learning_rate": 9.436640980735553e-05,
-      "loss": 1.0681,
-      "num_input_tokens_seen": 12964096,
-      "step": 805
-    },
-    {
-      "epoch": 0.056458886057772294,
-      "grad_norm": 3.7949161529541016,
-      "learning_rate": 9.435941155866902e-05,
-      "loss": 1.0787,
-      "num_input_tokens_seen": 12980480,
-      "step": 806
-    },
-    {
-      "epoch": 0.05652893430350154,
-      "grad_norm": 3.910456418991089,
-      "learning_rate": 9.435241330998249e-05,
-      "loss": 0.9212,
-      "num_input_tokens_seen": 12996864,
-      "step": 807
-    },
-    {
-      "epoch": 0.056598982549230785,
-      "grad_norm": 4.744706630706787,
-      "learning_rate": 9.434541506129598e-05,
-      "loss": 1.0582,
-      "num_input_tokens_seen": 13013248,
-      "step": 808
-    },
-    {
-      "epoch": 0.05666903079496003,
-      "grad_norm": 4.4282732009887695,
-      "learning_rate": 9.433841681260946e-05,
-      "loss": 1.1353,
-      "num_input_tokens_seen": 13029632,
-      "step": 809
-    },
-    {
-      "epoch": 0.056739079040689276,
-      "grad_norm": 3.8422467708587646,
-      "learning_rate": 9.433141856392294e-05,
-      "loss": 0.9881,
-      "num_input_tokens_seen": 13046016,
-      "step": 810
-    },
-    {
-      "epoch": 0.05680912728641852,
-      "grad_norm": 4.1764445304870605,
-      "learning_rate": 9.432442031523643e-05,
-      "loss": 1.183,
-      "num_input_tokens_seen": 13062400,
-      "step": 811
-    },
-    {
-      "epoch": 0.05687917553214777,
-      "grad_norm": 4.713895320892334,
-      "learning_rate": 9.431742206654992e-05,
-      "loss": 1.0752,
-      "num_input_tokens_seen": 13078584,
-      "step": 812
-    },
-    {
-      "epoch": 0.05694922377787701,
-      "grad_norm": 4.265610694885254,
-      "learning_rate": 9.431042381786341e-05,
-      "loss": 0.9469,
-      "num_input_tokens_seen": 13094968,
-      "step": 813
-    },
-    {
-      "epoch": 0.05701927202360626,
-      "grad_norm": 3.9274330139160156,
-      "learning_rate": 9.430342556917688e-05,
-      "loss": 1.1765,
-      "num_input_tokens_seen": 13111304,
-      "step": 814
-    },
-    {
-      "epoch": 0.057089320269335504,
-      "grad_norm": 4.44935941696167,
-      "learning_rate": 9.429642732049037e-05,
-      "loss": 1.1014,
-      "num_input_tokens_seen": 13127304,
-      "step": 815
-    },
-    {
-      "epoch": 0.05715936851506475,
-      "grad_norm": 5.019375801086426,
-      "learning_rate": 9.428942907180386e-05,
-      "loss": 1.0535,
-      "num_input_tokens_seen": 13143688,
-      "step": 816
-    },
-    {
-      "epoch": 0.057229416760793995,
-      "grad_norm": 4.743424892425537,
-      "learning_rate": 9.428243082311734e-05,
-      "loss": 1.3912,
-      "num_input_tokens_seen": 13160072,
-      "step": 817
-    },
-    {
-      "epoch": 0.05729946500652324,
-      "grad_norm": 3.921475887298584,
-      "learning_rate": 9.427543257443083e-05,
-      "loss": 1.1116,
-      "num_input_tokens_seen": 13176456,
-      "step": 818
-    },
-    {
-      "epoch": 0.05736951325225249,
-      "grad_norm": 4.106019020080566,
-      "learning_rate": 9.426843432574431e-05,
-      "loss": 0.9,
-      "num_input_tokens_seen": 13192840,
-      "step": 819
-    },
-    {
-      "epoch": 0.05743956149798173,
-      "grad_norm": 4.298704147338867,
-      "learning_rate": 9.42614360770578e-05,
-      "loss": 1.281,
-      "num_input_tokens_seen": 13209144,
-      "step": 820
-    },
-    {
-      "epoch": 0.05750960974371098,
-      "grad_norm": 4.29774284362793,
-      "learning_rate": 9.425443782837128e-05,
-      "loss": 1.2703,
-      "num_input_tokens_seen": 13224752,
-      "step": 821
-    },
-    {
-      "epoch": 0.057579657989440224,
-      "grad_norm": 4.6176838874816895,
-      "learning_rate": 9.424743957968477e-05,
-      "loss": 1.232,
-      "num_input_tokens_seen": 13240856,
-      "step": 822
-    },
-    {
-      "epoch": 0.057649706235169476,
-      "grad_norm": 4.450786590576172,
-      "learning_rate": 9.424044133099826e-05,
-      "loss": 1.1369,
-      "num_input_tokens_seen": 13256800,
-      "step": 823
-    },
-    {
-      "epoch": 0.05771975448089872,
-      "grad_norm": 3.8302414417266846,
-      "learning_rate": 9.423344308231173e-05,
-      "loss": 0.9985,
-      "num_input_tokens_seen": 13273032,
-      "step": 824
-    },
-    {
-      "epoch": 0.05778980272662797,
-      "grad_norm": 4.641941070556641,
-      "learning_rate": 9.422644483362523e-05,
-      "loss": 1.2238,
-      "num_input_tokens_seen": 13289104,
-      "step": 825
-    },
-    {
-      "epoch": 0.05785985097235721,
-      "grad_norm": 4.369805335998535,
-      "learning_rate": 9.421944658493871e-05,
-      "loss": 1.2047,
-      "num_input_tokens_seen": 13304752,
-      "step": 826
-    },
-    {
-      "epoch": 0.05792989921808646,
-      "grad_norm": 3.863507032394409,
-      "learning_rate": 9.42124483362522e-05,
-      "loss": 1.1098,
-      "num_input_tokens_seen": 13321088,
-      "step": 827
-    },
-    {
-      "epoch": 0.057999947463815704,
-      "grad_norm": 5.323369979858398,
-      "learning_rate": 9.420545008756568e-05,
-      "loss": 1.1722,
-      "num_input_tokens_seen": 13336912,
-      "step": 828
-    },
-    {
-      "epoch": 0.05806999570954495,
-      "grad_norm": 4.006597995758057,
-      "learning_rate": 9.419845183887916e-05,
-      "loss": 1.0382,
-      "num_input_tokens_seen": 13353280,
-      "step": 829
-    },
-    {
-      "epoch": 0.058140043955274195,
-      "grad_norm": 4.1039886474609375,
-      "learning_rate": 9.419145359019265e-05,
-      "loss": 1.2037,
-      "num_input_tokens_seen": 13369664,
-      "step": 830
-    },
-    {
-      "epoch": 0.05821009220100344,
-      "grad_norm": 3.903517007827759,
-      "learning_rate": 9.418445534150614e-05,
-      "loss": 1.2185,
-      "num_input_tokens_seen": 13386048,
-      "step": 831
-    },
-    {
-      "epoch": 0.05828014044673269,
-      "grad_norm": 4.434885025024414,
-      "learning_rate": 9.417745709281963e-05,
-      "loss": 1.2444,
-      "num_input_tokens_seen": 13402432,
-      "step": 832
-    },
-    {
-      "epoch": 0.05835018869246193,
-      "grad_norm": 4.6121296882629395,
-      "learning_rate": 9.417045884413311e-05,
-      "loss": 1.2831,
-      "num_input_tokens_seen": 13418816,
-      "step": 833
-    },
-    {
-      "epoch": 0.05842023693819118,
-      "grad_norm": 3.6966841220855713,
-      "learning_rate": 9.416346059544659e-05,
-      "loss": 1.0751,
-      "num_input_tokens_seen": 13435200,
-      "step": 834
-    },
-    {
-      "epoch": 0.058490285183920424,
-      "grad_norm": 4.292221546173096,
-      "learning_rate": 9.415646234676008e-05,
-      "loss": 1.2068,
-      "num_input_tokens_seen": 13451584,
-      "step": 835
-    },
-    {
-      "epoch": 0.05856033342964967,
-      "grad_norm": 4.053999900817871,
-      "learning_rate": 9.414946409807355e-05,
-      "loss": 1.1735,
-      "num_input_tokens_seen": 13467824,
-      "step": 836
-    },
-    {
-      "epoch": 0.058630381675378915,
-      "grad_norm": 4.4411234855651855,
-      "learning_rate": 9.414246584938704e-05,
-      "loss": 1.0647,
-      "num_input_tokens_seen": 13483200,
-      "step": 837
-    },
-    {
-      "epoch": 0.05870042992110816,
-      "grad_norm": 3.956787347793579,
-      "learning_rate": 9.413546760070053e-05,
-      "loss": 0.9813,
-      "num_input_tokens_seen": 13499584,
-      "step": 838
-    },
-    {
-      "epoch": 0.058770478166837406,
-      "grad_norm": 5.050291061401367,
-      "learning_rate": 9.412846935201402e-05,
-      "loss": 1.1193,
-      "num_input_tokens_seen": 13515448,
-      "step": 839
-    },
-    {
-      "epoch": 0.05884052641256666,
-      "grad_norm": 3.8736393451690674,
-      "learning_rate": 9.412147110332751e-05,
-      "loss": 1.0294,
-      "num_input_tokens_seen": 13531200,
-      "step": 840
-    },
-    {
-      "epoch": 0.058910574658295904,
-      "grad_norm": 6.07747745513916,
-      "learning_rate": 9.411447285464098e-05,
-      "loss": 0.9684,
-      "num_input_tokens_seen": 13547584,
-      "step": 841
-    },
-    {
-      "epoch": 0.05898062290402515,
-      "grad_norm": 4.606445789337158,
-      "learning_rate": 9.410747460595447e-05,
-      "loss": 1.2119,
-      "num_input_tokens_seen": 13563528,
-      "step": 842
-    },
-    {
-      "epoch": 0.059050671149754395,
-      "grad_norm": 4.3981709480285645,
-      "learning_rate": 9.410047635726796e-05,
-      "loss": 1.3313,
-      "num_input_tokens_seen": 13579912,
-      "step": 843
-    },
-    {
-      "epoch": 0.05912071939548364,
-      "grad_norm": 3.64546799659729,
-      "learning_rate": 9.409347810858143e-05,
-      "loss": 0.8892,
-      "num_input_tokens_seen": 13596296,
-      "step": 844
-    },
-    {
-      "epoch": 0.05919076764121289,
-      "grad_norm": 4.15845251083374,
-      "learning_rate": 9.408647985989494e-05,
-      "loss": 1.1464,
-      "num_input_tokens_seen": 13612680,
-      "step": 845
-    },
-    {
-      "epoch": 0.05926081588694213,
-      "grad_norm": 6.049203872680664,
-      "learning_rate": 9.407948161120841e-05,
-      "loss": 1.1907,
-      "num_input_tokens_seen": 13627832,
-      "step": 846
-    },
-    {
-      "epoch": 0.05933086413267138,
-      "grad_norm": 3.7192461490631104,
-      "learning_rate": 9.40724833625219e-05,
-      "loss": 1.165,
-      "num_input_tokens_seen": 13643824,
-      "step": 847
-    },
-    {
-      "epoch": 0.059400912378400623,
-      "grad_norm": 4.183239936828613,
-      "learning_rate": 9.406548511383537e-05,
-      "loss": 1.1697,
-      "num_input_tokens_seen": 13660208,
-      "step": 848
-    },
-    {
-      "epoch": 0.05947096062412987,
-      "grad_norm": 4.126212120056152,
-      "learning_rate": 9.405848686514886e-05,
-      "loss": 1.0532,
-      "num_input_tokens_seen": 13676592,
-      "step": 849
-    },
-    {
-      "epoch": 0.059541008869859115,
-      "grad_norm": 4.033525466918945,
-      "learning_rate": 9.405148861646235e-05,
-      "loss": 1.1497,
-      "num_input_tokens_seen": 13692600,
-      "step": 850
-    },
-    {
-      "epoch": 0.05961105711558836,
-      "grad_norm": 4.162797451019287,
-      "learning_rate": 9.404449036777584e-05,
-      "loss": 1.162,
-      "num_input_tokens_seen": 13708984,
-      "step": 851
-    },
-    {
-      "epoch": 0.059681105361317606,
-      "grad_norm": 4.057224750518799,
-      "learning_rate": 9.403749211908933e-05,
-      "loss": 1.2166,
-      "num_input_tokens_seen": 13724656,
-      "step": 852
-    },
-    {
-      "epoch": 0.05975115360704685,
-      "grad_norm": 4.201955318450928,
-      "learning_rate": 9.40304938704028e-05,
-      "loss": 1.2195,
-      "num_input_tokens_seen": 13741040,
-      "step": 853
-    },
-    {
-      "epoch": 0.0598212018527761,
-      "grad_norm": 3.8704352378845215,
-      "learning_rate": 9.402349562171629e-05,
-      "loss": 0.8946,
-      "num_input_tokens_seen": 13757424,
-      "step": 854
-    },
-    {
-      "epoch": 0.05989125009850534,
-      "grad_norm": 6.010958671569824,
-      "learning_rate": 9.401649737302978e-05,
-      "loss": 1.2095,
-      "num_input_tokens_seen": 13773808,
-      "step": 855
-    },
-    {
-      "epoch": 0.05996129834423459,
-      "grad_norm": 4.975742816925049,
-      "learning_rate": 9.400949912434326e-05,
-      "loss": 1.1064,
-      "num_input_tokens_seen": 13789704,
-      "step": 856
-    },
-    {
-      "epoch": 0.06003134658996384,
-      "grad_norm": 4.021739959716797,
-      "learning_rate": 9.400250087565675e-05,
-      "loss": 1.2036,
-      "num_input_tokens_seen": 13806088,
-      "step": 857
-    },
-    {
-      "epoch": 0.06010139483569309,
-      "grad_norm": 4.262394905090332,
-      "learning_rate": 9.399550262697023e-05,
-      "loss": 1.1053,
-      "num_input_tokens_seen": 13821928,
-      "step": 858
-    },
-    {
-      "epoch": 0.06017144308142233,
-      "grad_norm": 4.3033671379089355,
-      "learning_rate": 9.398850437828372e-05,
-      "loss": 1.0213,
-      "num_input_tokens_seen": 13838232,
-      "step": 859
-    },
-    {
-      "epoch": 0.06024149132715158,
-      "grad_norm": 4.066610336303711,
-      "learning_rate": 9.398150612959721e-05,
-      "loss": 1.0579,
-      "num_input_tokens_seen": 13853912,
-      "step": 860
-    },
-    {
-      "epoch": 0.06031153957288082,
-      "grad_norm": 4.308155059814453,
-      "learning_rate": 9.397450788091069e-05,
-      "loss": 1.3624,
-      "num_input_tokens_seen": 13870224,
-      "step": 861
-    },
-    {
-      "epoch": 0.06038158781861007,
-      "grad_norm": 4.307553291320801,
-      "learning_rate": 9.396750963222417e-05,
-      "loss": 1.0942,
-      "num_input_tokens_seen": 13886608,
-      "step": 862
-    },
-    {
-      "epoch": 0.060451636064339315,
-      "grad_norm": 3.8107142448425293,
-      "learning_rate": 9.396051138353765e-05,
-      "loss": 1.1285,
-      "num_input_tokens_seen": 13902992,
-      "step": 863
-    },
-    {
-      "epoch": 0.06052168431006856,
-      "grad_norm": 4.530765533447266,
-      "learning_rate": 9.395351313485114e-05,
-      "loss": 1.2028,
-      "num_input_tokens_seen": 13919376,
-      "step": 864
-    },
-    {
-      "epoch": 0.060591732555797806,
-      "grad_norm": 4.035069465637207,
-      "learning_rate": 9.394651488616463e-05,
-      "loss": 1.0291,
-      "num_input_tokens_seen": 13935664,
-      "step": 865
-    },
-    {
-      "epoch": 0.06066178080152705,
-      "grad_norm": 4.028316497802734,
-      "learning_rate": 9.393951663747812e-05,
-      "loss": 1.21,
-      "num_input_tokens_seen": 13951096,
-      "step": 866
-    },
-    {
-      "epoch": 0.0607318290472563,
-      "grad_norm": 4.039167881011963,
-      "learning_rate": 9.39325183887916e-05,
-      "loss": 0.929,
-      "num_input_tokens_seen": 13966272,
-      "step": 867
-    },
-    {
-      "epoch": 0.06080187729298554,
-      "grad_norm": 4.139703273773193,
-      "learning_rate": 9.392552014010508e-05,
-      "loss": 1.2575,
-      "num_input_tokens_seen": 13981848,
-      "step": 868
-    },
-    {
-      "epoch": 0.06087192553871479,
-      "grad_norm": 4.222180366516113,
-      "learning_rate": 9.391852189141857e-05,
-      "loss": 1.2067,
-      "num_input_tokens_seen": 13997920,
-      "step": 869
-    },
-    {
-      "epoch": 0.060941973784444034,
-      "grad_norm": 3.7993030548095703,
-      "learning_rate": 9.391152364273206e-05,
-      "loss": 1.0865,
-      "num_input_tokens_seen": 14014304,
-      "step": 870
-    },
-    {
-      "epoch": 0.06101202203017328,
-      "grad_norm": 4.811493396759033,
-      "learning_rate": 9.390452539404554e-05,
-      "loss": 1.1331,
-      "num_input_tokens_seen": 14030688,
-      "step": 871
-    },
-    {
-      "epoch": 0.061082070275902525,
-      "grad_norm": 13.88792610168457,
-      "learning_rate": 9.389752714535903e-05,
-      "loss": 1.1368,
-      "num_input_tokens_seen": 14045584,
-      "step": 872
-    },
-    {
-      "epoch": 0.06115211852163177,
-      "grad_norm": 3.7678709030151367,
-      "learning_rate": 9.389052889667251e-05,
-      "loss": 1.1012,
-      "num_input_tokens_seen": 14061968,
-      "step": 873
-    },
-    {
-      "epoch": 0.061222166767361016,
-      "grad_norm": 4.252075672149658,
-      "learning_rate": 9.3883530647986e-05,
-      "loss": 1.0472,
-      "num_input_tokens_seen": 14077584,
-      "step": 874
-    },
-    {
-      "epoch": 0.06129221501309027,
-      "grad_norm": 3.555629253387451,
-      "learning_rate": 9.387653239929947e-05,
-      "loss": 0.8653,
-      "num_input_tokens_seen": 14093704,
-      "step": 875
-    },
-    {
-      "epoch": 0.061362263258819515,
-      "grad_norm": 4.122331619262695,
-      "learning_rate": 9.386953415061296e-05,
-      "loss": 1.0395,
-      "num_input_tokens_seen": 14109624,
-      "step": 876
-    },
-    {
-      "epoch": 0.06143231150454876,
-      "grad_norm": 3.6772518157958984,
-      "learning_rate": 9.386253590192645e-05,
-      "loss": 0.8842,
-      "num_input_tokens_seen": 14126008,
-      "step": 877
-    },
-    {
-      "epoch": 0.061502359750278006,
-      "grad_norm": 3.791351079940796,
-      "learning_rate": 9.385553765323994e-05,
-      "loss": 1.1118,
-      "num_input_tokens_seen": 14142392,
-      "step": 878
-    },
-    {
-      "epoch": 0.06157240799600725,
-      "grad_norm": 3.781759738922119,
-      "learning_rate": 9.384853940455343e-05,
-      "loss": 1.0577,
-      "num_input_tokens_seen": 14158776,
-      "step": 879
-    },
-    {
-      "epoch": 0.0616424562417365,
-      "grad_norm": 4.2420830726623535,
-      "learning_rate": 9.38415411558669e-05,
-      "loss": 1.268,
-      "num_input_tokens_seen": 14173920,
-      "step": 880
-    },
-    {
-      "epoch": 0.06171250448746574,
-      "grad_norm": 4.000860214233398,
-      "learning_rate": 9.383454290718039e-05,
-      "loss": 1.1626,
-      "num_input_tokens_seen": 14190032,
-      "step": 881
-    },
-    {
-      "epoch": 0.06178255273319499,
-      "grad_norm": 3.760969877243042,
-      "learning_rate": 9.382754465849388e-05,
-      "loss": 0.9684,
-      "num_input_tokens_seen": 14206416,
-      "step": 882
-    },
-    {
-      "epoch": 0.061852600978924234,
-      "grad_norm": 4.81919002532959,
-      "learning_rate": 9.382054640980735e-05,
-      "loss": 1.1056,
-      "num_input_tokens_seen": 14222408,
-      "step": 883
-    },
-    {
-      "epoch": 0.06192264922465348,
-      "grad_norm": 4.951950550079346,
-      "learning_rate": 9.381354816112084e-05,
-      "loss": 1.0334,
-      "num_input_tokens_seen": 14238616,
-      "step": 884
-    },
-    {
-      "epoch": 0.061992697470382725,
-      "grad_norm": 4.15132999420166,
-      "learning_rate": 9.380654991243433e-05,
-      "loss": 1.3171,
-      "num_input_tokens_seen": 14254968,
-      "step": 885
-    },
-    {
-      "epoch": 0.06206274571611197,
-      "grad_norm": 5.100244998931885,
-      "learning_rate": 9.379955166374782e-05,
-      "loss": 1.1684,
-      "num_input_tokens_seen": 14271352,
-      "step": 886
-    },
-    {
-      "epoch": 0.062132793961841216,
-      "grad_norm": 5.999105453491211,
-      "learning_rate": 9.379255341506131e-05,
-      "loss": 0.9824,
-      "num_input_tokens_seen": 14287496,
-      "step": 887
-    },
-    {
-      "epoch": 0.06220284220757046,
-      "grad_norm": 3.8826348781585693,
-      "learning_rate": 9.378555516637478e-05,
-      "loss": 1.0829,
-      "num_input_tokens_seen": 14303880,
-      "step": 888
-    },
-    {
-      "epoch": 0.06227289045329971,
-      "grad_norm": 5.308819770812988,
-      "learning_rate": 9.377855691768827e-05,
-      "loss": 1.1377,
-      "num_input_tokens_seen": 14320264,
-      "step": 889
-    },
-    {
-      "epoch": 0.06234293869902895,
-      "grad_norm": 4.383331775665283,
-      "learning_rate": 9.377155866900175e-05,
-      "loss": 1.0147,
-      "num_input_tokens_seen": 14336232,
-      "step": 890
-    },
-    {
-      "epoch": 0.0624129869447582,
-      "grad_norm": 4.335045337677002,
-      "learning_rate": 9.376456042031524e-05,
-      "loss": 0.9807,
-      "num_input_tokens_seen": 14351704,
-      "step": 891
-    },
-    {
-      "epoch": 0.06248303519048745,
-      "grad_norm": 3.6901326179504395,
-      "learning_rate": 9.375756217162872e-05,
-      "loss": 1.0494,
-      "num_input_tokens_seen": 14368088,
-      "step": 892
-    },
-    {
-      "epoch": 0.0625530834362167,
-      "grad_norm": 3.912727117538452,
-      "learning_rate": 9.375056392294221e-05,
-      "loss": 1.1191,
-      "num_input_tokens_seen": 14383904,
-      "step": 893
-    },
-    {
-      "epoch": 0.06262313168194594,
-      "grad_norm": 3.5688252449035645,
-      "learning_rate": 9.37435656742557e-05,
-      "loss": 0.833,
-      "num_input_tokens_seen": 14399648,
-      "step": 894
-    },
-    {
-      "epoch": 0.06269317992767519,
-      "grad_norm": 4.6460137367248535,
-      "learning_rate": 9.373656742556918e-05,
-      "loss": 1.2523,
-      "num_input_tokens_seen": 14415640,
-      "step": 895
-    },
-    {
-      "epoch": 0.06276322817340443,
-      "grad_norm": 3.8113012313842773,
-      "learning_rate": 9.372956917688266e-05,
-      "loss": 1.1789,
-      "num_input_tokens_seen": 14432024,
-      "step": 896
-    },
-    {
-      "epoch": 0.06283327641913368,
-      "grad_norm": 3.8755953311920166,
-      "learning_rate": 9.372257092819615e-05,
-      "loss": 1.1506,
-      "num_input_tokens_seen": 14448152,
-      "step": 897
-    },
-    {
-      "epoch": 0.06290332466486293,
-      "grad_norm": 4.225901126861572,
-      "learning_rate": 9.371557267950964e-05,
-      "loss": 1.0754,
-      "num_input_tokens_seen": 14464536,
-      "step": 898
-    },
-    {
-      "epoch": 0.06297337291059217,
-      "grad_norm": 3.9437992572784424,
-      "learning_rate": 9.370857443082313e-05,
-      "loss": 1.049,
-      "num_input_tokens_seen": 14480072,
-      "step": 899
-    },
-    {
-      "epoch": 0.06304342115632142,
-      "grad_norm": 3.8961846828460693,
-      "learning_rate": 9.37015761821366e-05,
-      "loss": 1.1925,
-      "num_input_tokens_seen": 14496456,
-      "step": 900
-    },
-    {
-      "epoch": 0.06311346940205066,
-      "grad_norm": 4.844581604003906,
-      "learning_rate": 9.36945779334501e-05,
-      "loss": 1.0867,
-      "num_input_tokens_seen": 14512520,
-      "step": 901
-    },
-    {
-      "epoch": 0.06318351764777991,
-      "grad_norm": 4.89027214050293,
-      "learning_rate": 9.368757968476357e-05,
-      "loss": 1.0997,
-      "num_input_tokens_seen": 14528904,
-      "step": 902
-    },
-    {
-      "epoch": 0.06325356589350915,
-      "grad_norm": 4.303073883056641,
-      "learning_rate": 9.368058143607706e-05,
-      "loss": 1.0626,
-      "num_input_tokens_seen": 14545288,
-      "step": 903
-    },
-    {
-      "epoch": 0.0633236141392384,
-      "grad_norm": 5.145171165466309,
-      "learning_rate": 9.367358318739055e-05,
-      "loss": 1.3597,
-      "num_input_tokens_seen": 14561672,
-      "step": 904
-    },
-    {
-      "epoch": 0.06339366238496764,
-      "grad_norm": 5.7905964851379395,
-      "learning_rate": 9.366658493870403e-05,
-      "loss": 1.1075,
-      "num_input_tokens_seen": 14575896,
-      "step": 905
-    },
-    {
-      "epoch": 0.06346371063069689,
-      "grad_norm": 3.7394728660583496,
-      "learning_rate": 9.365958669001752e-05,
-      "loss": 0.9347,
-      "num_input_tokens_seen": 14592280,
-      "step": 906
-    },
-    {
-      "epoch": 0.06353375887642614,
-      "grad_norm": 3.916626453399658,
-      "learning_rate": 9.3652588441331e-05,
-      "loss": 1.0793,
-      "num_input_tokens_seen": 14608072,
-      "step": 907
-    },
-    {
-      "epoch": 0.06360380712215538,
-      "grad_norm": 5.088227272033691,
-      "learning_rate": 9.364559019264449e-05,
-      "loss": 1.158,
-      "num_input_tokens_seen": 14624360,
-      "step": 908
-    },
-    {
-      "epoch": 0.06367385536788463,
-      "grad_norm": 3.8519606590270996,
-      "learning_rate": 9.363859194395798e-05,
-      "loss": 1.1235,
-      "num_input_tokens_seen": 14640744,
-      "step": 909
-    },
-    {
-      "epoch": 0.06374390361361387,
-      "grad_norm": 4.450200080871582,
-      "learning_rate": 9.363159369527145e-05,
-      "loss": 1.0145,
-      "num_input_tokens_seen": 14657128,
-      "step": 910
-    },
-    {
-      "epoch": 0.06381395185934312,
-      "grad_norm": 4.188115119934082,
-      "learning_rate": 9.362459544658494e-05,
-      "loss": 1.1457,
-      "num_input_tokens_seen": 14673128,
-      "step": 911
-    },
-    {
-      "epoch": 0.06388400010507236,
-      "grad_norm": 4.67346715927124,
-      "learning_rate": 9.361759719789843e-05,
-      "loss": 1.2841,
-      "num_input_tokens_seen": 14689512,
-      "step": 912
-    },
-    {
-      "epoch": 0.06395404835080161,
-      "grad_norm": 3.737790822982788,
-      "learning_rate": 9.361059894921192e-05,
-      "loss": 1.0114,
-      "num_input_tokens_seen": 14705872,
-      "step": 913
-    },
-    {
-      "epoch": 0.06402409659653086,
-      "grad_norm": 4.2486653327941895,
-      "learning_rate": 9.36036007005254e-05,
-      "loss": 1.1526,
-      "num_input_tokens_seen": 14721816,
-      "step": 914
-    },
-    {
-      "epoch": 0.0640941448422601,
-      "grad_norm": 4.120566368103027,
-      "learning_rate": 9.359660245183888e-05,
-      "loss": 1.1045,
-      "num_input_tokens_seen": 14738200,
-      "step": 915
-    },
-    {
-      "epoch": 0.06416419308798935,
-      "grad_norm": 5.259902477264404,
-      "learning_rate": 9.358960420315237e-05,
-      "loss": 1.3544,
-      "num_input_tokens_seen": 14753920,
-      "step": 916
-    },
-    {
-      "epoch": 0.06423424133371859,
-      "grad_norm": 3.900827646255493,
-      "learning_rate": 9.358260595446584e-05,
-      "loss": 1.1079,
-      "num_input_tokens_seen": 14769640,
-      "step": 917
-    },
-    {
-      "epoch": 0.06430428957944785,
-      "grad_norm": 4.103065490722656,
-      "learning_rate": 9.357560770577935e-05,
-      "loss": 0.963,
-      "num_input_tokens_seen": 14786024,
-      "step": 918
-    },
-    {
-      "epoch": 0.0643743378251771,
-      "grad_norm": 3.9913623332977295,
-      "learning_rate": 9.356860945709282e-05,
-      "loss": 1.0959,
-      "num_input_tokens_seen": 14802408,
-      "step": 919
-    },
-    {
-      "epoch": 0.06444438607090634,
-      "grad_norm": 3.7369885444641113,
-      "learning_rate": 9.356161120840631e-05,
-      "loss": 1.131,
-      "num_input_tokens_seen": 14818792,
-      "step": 920
-    },
-    {
-      "epoch": 0.06451443431663559,
-      "grad_norm": 4.029351711273193,
-      "learning_rate": 9.35546129597198e-05,
-      "loss": 1.0378,
-      "num_input_tokens_seen": 14833792,
-      "step": 921
-    },
-    {
-      "epoch": 0.06458448256236483,
-      "grad_norm": 4.043665885925293,
-      "learning_rate": 9.354761471103327e-05,
-      "loss": 1.179,
-      "num_input_tokens_seen": 14850176,
-      "step": 922
-    },
-    {
-      "epoch": 0.06465453080809408,
-      "grad_norm": 3.7803280353546143,
-      "learning_rate": 9.354061646234676e-05,
-      "loss": 0.9886,
-      "num_input_tokens_seen": 14866096,
-      "step": 923
-    },
-    {
-      "epoch": 0.06472457905382333,
-      "grad_norm": 5.537375450134277,
-      "learning_rate": 9.353361821366025e-05,
-      "loss": 1.2519,
-      "num_input_tokens_seen": 14882480,
-      "step": 924
-    },
-    {
-      "epoch": 0.06479462729955257,
-      "grad_norm": 4.944652557373047,
-      "learning_rate": 9.352661996497374e-05,
-      "loss": 1.1963,
-      "num_input_tokens_seen": 14898864,
-      "step": 925
-    },
-    {
-      "epoch": 0.06486467554528182,
-      "grad_norm": 4.3231611251831055,
-      "learning_rate": 9.351962171628723e-05,
-      "loss": 1.1858,
-      "num_input_tokens_seen": 14913856,
-      "step": 926
-    },
-    {
-      "epoch": 0.06493472379101106,
-      "grad_norm": 4.386692523956299,
-      "learning_rate": 9.35126234676007e-05,
-      "loss": 1.0464,
-      "num_input_tokens_seen": 14929816,
-      "step": 927
-    },
-    {
-      "epoch": 0.06500477203674031,
-      "grad_norm": 4.607088088989258,
-      "learning_rate": 9.350562521891419e-05,
-      "loss": 1.2197,
-      "num_input_tokens_seen": 14946200,
-      "step": 928
-    },
-    {
-      "epoch": 0.06507482028246955,
-      "grad_norm": 4.7108001708984375,
-      "learning_rate": 9.349862697022767e-05,
-      "loss": 1.2335,
-      "num_input_tokens_seen": 14961816,
-      "step": 929
-    },
-    {
-      "epoch": 0.0651448685281988,
-      "grad_norm": 3.844571352005005,
-      "learning_rate": 9.349162872154115e-05,
-      "loss": 1.2745,
-      "num_input_tokens_seen": 14978200,
-      "step": 930
-    },
-    {
-      "epoch": 0.06521491677392804,
-      "grad_norm": 4.078561782836914,
-      "learning_rate": 9.348463047285464e-05,
-      "loss": 1.1737,
-      "num_input_tokens_seen": 14994440,
-      "step": 931
-    },
-    {
-      "epoch": 0.06528496501965729,
-      "grad_norm": 4.317986011505127,
-      "learning_rate": 9.347763222416813e-05,
-      "loss": 1.3046,
-      "num_input_tokens_seen": 15010824,
-      "step": 932
-    },
-    {
-      "epoch": 0.06535501326538654,
-      "grad_norm": 4.459141254425049,
-      "learning_rate": 9.347063397548162e-05,
-      "loss": 1.2893,
-      "num_input_tokens_seen": 15026608,
-      "step": 933
-    },
-    {
-      "epoch": 0.06542506151111578,
-      "grad_norm": 4.251399993896484,
-      "learning_rate": 9.34636357267951e-05,
-      "loss": 1.2346,
-      "num_input_tokens_seen": 15042328,
-      "step": 934
-    },
-    {
-      "epoch": 0.06549510975684503,
-      "grad_norm": 4.568341255187988,
-      "learning_rate": 9.345663747810858e-05,
-      "loss": 1.4343,
-      "num_input_tokens_seen": 15058712,
-      "step": 935
-    },
-    {
-      "epoch": 0.06556515800257427,
-      "grad_norm": 4.7616424560546875,
-      "learning_rate": 9.344963922942207e-05,
-      "loss": 1.0925,
-      "num_input_tokens_seen": 15075096,
-      "step": 936
-    },
-    {
-      "epoch": 0.06563520624830352,
-      "grad_norm": 3.8224191665649414,
-      "learning_rate": 9.344264098073555e-05,
-      "loss": 1.0958,
-      "num_input_tokens_seen": 15091480,
-      "step": 937
-    },
-    {
-      "epoch": 0.06570525449403276,
-      "grad_norm": 4.985624313354492,
-      "learning_rate": 9.343564273204905e-05,
-      "loss": 1.233,
-      "num_input_tokens_seen": 15107864,
-      "step": 938
-    },
-    {
-      "epoch": 0.06577530273976201,
-      "grad_norm": 4.3780975341796875,
-      "learning_rate": 9.342864448336252e-05,
-      "loss": 1.1819,
-      "num_input_tokens_seen": 15123656,
-      "step": 939
-    },
-    {
-      "epoch": 0.06584535098549125,
-      "grad_norm": 4.435183525085449,
-      "learning_rate": 9.342164623467601e-05,
-      "loss": 1.1107,
-      "num_input_tokens_seen": 15140040,
-      "step": 940
-    },
-    {
-      "epoch": 0.0659153992312205,
-      "grad_norm": 4.560804843902588,
-      "learning_rate": 9.34146479859895e-05,
-      "loss": 1.1274,
-      "num_input_tokens_seen": 15156424,
-      "step": 941
-    },
-    {
-      "epoch": 0.06598544747694975,
-      "grad_norm": 5.184841156005859,
-      "learning_rate": 9.340764973730298e-05,
-      "loss": 1.3124,
-      "num_input_tokens_seen": 15172504,
-      "step": 942
-    },
-    {
-      "epoch": 0.06605549572267899,
-      "grad_norm": 3.5243096351623535,
-      "learning_rate": 9.340065148861647e-05,
-      "loss": 0.8203,
-      "num_input_tokens_seen": 15188888,
-      "step": 943
-    },
-    {
-      "epoch": 0.06612554396840824,
-      "grad_norm": 4.041544437408447,
-      "learning_rate": 9.339365323992995e-05,
-      "loss": 1.0602,
-      "num_input_tokens_seen": 15204672,
-      "step": 944
-    },
-    {
-      "epoch": 0.06619559221413748,
-      "grad_norm": 3.720906972885132,
-      "learning_rate": 9.338665499124344e-05,
-      "loss": 1.0722,
-      "num_input_tokens_seen": 15220688,
-      "step": 945
-    },
-    {
-      "epoch": 0.06626564045986673,
-      "grad_norm": 3.9778380393981934,
-      "learning_rate": 9.337965674255692e-05,
-      "loss": 1.2653,
-      "num_input_tokens_seen": 15236856,
-      "step": 946
-    },
-    {
-      "epoch": 0.06633568870559597,
-      "grad_norm": 4.486488342285156,
-      "learning_rate": 9.33726584938704e-05,
-      "loss": 1.2408,
-      "num_input_tokens_seen": 15253240,
-      "step": 947
-    },
-    {
-      "epoch": 0.06640573695132522,
-      "grad_norm": 8.369994163513184,
-      "learning_rate": 9.33656602451839e-05,
-      "loss": 1.4841,
-      "num_input_tokens_seen": 15267728,
-      "step": 948
-    },
-    {
-      "epoch": 0.06647578519705447,
-      "grad_norm": 4.2056732177734375,
-      "learning_rate": 9.335866199649737e-05,
-      "loss": 1.4258,
-      "num_input_tokens_seen": 15284112,
-      "step": 949
-    },
-    {
-      "epoch": 0.06654583344278371,
-      "grad_norm": 4.396723747253418,
-      "learning_rate": 9.335166374781086e-05,
-      "loss": 1.1578,
-      "num_input_tokens_seen": 15300496,
-      "step": 950
-    },
-    {
-      "epoch": 0.06661588168851296,
-      "grad_norm": 3.7177491188049316,
-      "learning_rate": 9.334466549912435e-05,
-      "loss": 1.0664,
-      "num_input_tokens_seen": 15316608,
-      "step": 951
-    },
-    {
-      "epoch": 0.0666859299342422,
-      "grad_norm": 4.080933094024658,
-      "learning_rate": 9.333766725043784e-05,
-      "loss": 1.1282,
-      "num_input_tokens_seen": 15332976,
-      "step": 952
-    },
-    {
-      "epoch": 0.06675597817997146,
-      "grad_norm": 5.188856601715088,
-      "learning_rate": 9.333066900175132e-05,
-      "loss": 1.2079,
-      "num_input_tokens_seen": 15349080,
-      "step": 953
-    },
-    {
-      "epoch": 0.06682602642570071,
-      "grad_norm": 4.583539962768555,
-      "learning_rate": 9.33236707530648e-05,
-      "loss": 0.9047,
-      "num_input_tokens_seen": 15365256,
-      "step": 954
-    },
-    {
-      "epoch": 0.06689607467142995,
-      "grad_norm": 3.873830795288086,
-      "learning_rate": 9.331667250437829e-05,
-      "loss": 1.159,
-      "num_input_tokens_seen": 15381640,
-      "step": 955
-    },
-    {
-      "epoch": 0.0669661229171592,
-      "grad_norm": 3.9574460983276367,
-      "learning_rate": 9.330967425569176e-05,
-      "loss": 1.0696,
-      "num_input_tokens_seen": 15397800,
-      "step": 956
-    },
-    {
-      "epoch": 0.06703617116288844,
-      "grad_norm": 3.8933448791503906,
-      "learning_rate": 9.330267600700525e-05,
-      "loss": 0.9844,
-      "num_input_tokens_seen": 15414112,
-      "step": 957
-    },
-    {
-      "epoch": 0.06710621940861769,
-      "grad_norm": 4.748478412628174,
-      "learning_rate": 9.329567775831875e-05,
-      "loss": 1.1308,
-      "num_input_tokens_seen": 15430496,
-      "step": 958
-    },
-    {
-      "epoch": 0.06717626765434694,
-      "grad_norm": 6.755379676818848,
-      "learning_rate": 9.328867950963223e-05,
-      "loss": 1.206,
-      "num_input_tokens_seen": 15445072,
-      "step": 959
-    },
-    {
-      "epoch": 0.06724631590007618,
-      "grad_norm": 4.382065773010254,
-      "learning_rate": 9.328168126094572e-05,
-      "loss": 1.0753,
-      "num_input_tokens_seen": 15460336,
-      "step": 960
-    },
-    {
-      "epoch": 0.06731636414580543,
-      "grad_norm": 5.037116527557373,
-      "learning_rate": 9.327468301225919e-05,
-      "loss": 1.0562,
-      "num_input_tokens_seen": 15474752,
-      "step": 961
-    },
-    {
-      "epoch": 0.06738641239153467,
-      "grad_norm": 5.838945388793945,
-      "learning_rate": 9.326768476357268e-05,
-      "loss": 1.314,
-      "num_input_tokens_seen": 15491136,
-      "step": 962
-    },
-    {
-      "epoch": 0.06745646063726392,
-      "grad_norm": 3.690436840057373,
-      "learning_rate": 9.326068651488617e-05,
-      "loss": 0.996,
-      "num_input_tokens_seen": 15507520,
-      "step": 963
-    },
-    {
-      "epoch": 0.06752650888299316,
-      "grad_norm": 4.1123247146606445,
-      "learning_rate": 9.325368826619966e-05,
-      "loss": 1.2031,
-      "num_input_tokens_seen": 15523904,
-      "step": 964
-    },
-    {
-      "epoch": 0.06759655712872241,
-      "grad_norm": 4.120308876037598,
-      "learning_rate": 9.324669001751315e-05,
-      "loss": 0.9671,
-      "num_input_tokens_seen": 15540136,
-      "step": 965
-    },
-    {
-      "epoch": 0.06766660537445165,
-      "grad_norm": 3.9849514961242676,
-      "learning_rate": 9.323969176882662e-05,
-      "loss": 1.1669,
-      "num_input_tokens_seen": 15556312,
-      "step": 966
-    },
-    {
-      "epoch": 0.0677366536201809,
-      "grad_norm": 3.9164884090423584,
-      "learning_rate": 9.323269352014011e-05,
-      "loss": 1.0883,
-      "num_input_tokens_seen": 15571864,
-      "step": 967
-    },
-    {
-      "epoch": 0.06780670186591015,
-      "grad_norm": 4.282434940338135,
-      "learning_rate": 9.32256952714536e-05,
-      "loss": 1.241,
-      "num_input_tokens_seen": 15587800,
-      "step": 968
-    },
-    {
-      "epoch": 0.06787675011163939,
-      "grad_norm": 4.118724346160889,
-      "learning_rate": 9.321869702276707e-05,
-      "loss": 1.0905,
-      "num_input_tokens_seen": 15603128,
-      "step": 969
-    },
-    {
-      "epoch": 0.06794679835736864,
-      "grad_norm": 4.233770847320557,
-      "learning_rate": 9.321169877408056e-05,
-      "loss": 1.0618,
-      "num_input_tokens_seen": 15617864,
-      "step": 970
-    },
-    {
-      "epoch": 0.06801684660309788,
-      "grad_norm": 3.933587074279785,
-      "learning_rate": 9.320470052539405e-05,
-      "loss": 0.982,
-      "num_input_tokens_seen": 15634248,
-      "step": 971
-    },
-    {
-      "epoch": 0.06808689484882713,
-      "grad_norm": 4.641788482666016,
-      "learning_rate": 9.319770227670754e-05,
-      "loss": 0.9793,
-      "num_input_tokens_seen": 15650304,
-      "step": 972
-    },
-    {
-      "epoch": 0.06815694309455637,
-      "grad_norm": 4.138880729675293,
-      "learning_rate": 9.319070402802102e-05,
-      "loss": 1.1991,
-      "num_input_tokens_seen": 15666688,
-      "step": 973
-    },
-    {
-      "epoch": 0.06822699134028562,
-      "grad_norm": 4.823685169219971,
-      "learning_rate": 9.31837057793345e-05,
-      "loss": 0.9162,
-      "num_input_tokens_seen": 15682936,
-      "step": 974
-    },
-    {
-      "epoch": 0.06829703958601487,
-      "grad_norm": 4.432481288909912,
-      "learning_rate": 9.317670753064799e-05,
-      "loss": 0.9626,
-      "num_input_tokens_seen": 15699320,
-      "step": 975
-    },
-    {
-      "epoch": 0.06836708783174411,
-      "grad_norm": 4.115868091583252,
-      "learning_rate": 9.316970928196147e-05,
-      "loss": 1.105,
-      "num_input_tokens_seen": 15715296,
-      "step": 976
-    },
-    {
-      "epoch": 0.06843713607747336,
-      "grad_norm": 3.964905023574829,
-      "learning_rate": 9.316271103327496e-05,
-      "loss": 1.0064,
-      "num_input_tokens_seen": 15731680,
-      "step": 977
-    },
-    {
-      "epoch": 0.0685071843232026,
-      "grad_norm": 3.686522960662842,
-      "learning_rate": 9.315571278458846e-05,
-      "loss": 0.9924,
-      "num_input_tokens_seen": 15747808,
-      "step": 978
-    },
-    {
-      "epoch": 0.06857723256893185,
-      "grad_norm": 4.0614423751831055,
-      "learning_rate": 9.314871453590193e-05,
-      "loss": 1.0425,
-      "num_input_tokens_seen": 15764168,
-      "step": 979
-    },
-    {
-      "epoch": 0.0686472808146611,
-      "grad_norm": 3.756350517272949,
-      "learning_rate": 9.314171628721542e-05,
-      "loss": 1.0757,
-      "num_input_tokens_seen": 15780176,
-      "step": 980
-    },
-    {
-      "epoch": 0.06871732906039034,
-      "grad_norm": 4.30344820022583,
-      "learning_rate": 9.31347180385289e-05,
-      "loss": 0.9496,
-      "num_input_tokens_seen": 15795720,
-      "step": 981
-    },
-    {
-      "epoch": 0.06878737730611958,
-      "grad_norm": 4.055768013000488,
-      "learning_rate": 9.312771978984239e-05,
-      "loss": 1.0189,
-      "num_input_tokens_seen": 15811528,
-      "step": 982
-    },
-    {
-      "epoch": 0.06885742555184883,
-      "grad_norm": 3.8779115676879883,
-      "learning_rate": 9.312072154115586e-05,
-      "loss": 1.0516,
-      "num_input_tokens_seen": 15827392,
-      "step": 983
-    },
-    {
-      "epoch": 0.06892747379757808,
-      "grad_norm": 5.014206886291504,
-      "learning_rate": 9.311372329246936e-05,
-      "loss": 1.3421,
-      "num_input_tokens_seen": 15843776,
-      "step": 984
-    },
-    {
-      "epoch": 0.06899752204330732,
-      "grad_norm": 4.548489570617676,
-      "learning_rate": 9.310672504378285e-05,
-      "loss": 1.1652,
-      "num_input_tokens_seen": 15858880,
-      "step": 985
-    },
-    {
-      "epoch": 0.06906757028903657,
-      "grad_norm": 4.312918186187744,
-      "learning_rate": 9.309972679509633e-05,
-      "loss": 1.2728,
-      "num_input_tokens_seen": 15874840,
-      "step": 986
-    },
-    {
-      "epoch": 0.06913761853476583,
-      "grad_norm": 3.9783735275268555,
-      "learning_rate": 9.309272854640981e-05,
-      "loss": 0.9377,
-      "num_input_tokens_seen": 15890568,
-      "step": 987
-    },
-    {
-      "epoch": 0.06920766678049507,
-      "grad_norm": 4.155986309051514,
-      "learning_rate": 9.308573029772329e-05,
-      "loss": 1.0278,
-      "num_input_tokens_seen": 15906952,
-      "step": 988
-    },
-    {
-      "epoch": 0.06927771502622432,
-      "grad_norm": 3.633018732070923,
-      "learning_rate": 9.307873204903678e-05,
-      "loss": 1.1276,
-      "num_input_tokens_seen": 15923336,
-      "step": 989
-    },
-    {
-      "epoch": 0.06934776327195356,
-      "grad_norm": 3.9513449668884277,
-      "learning_rate": 9.307173380035027e-05,
-      "loss": 0.9076,
-      "num_input_tokens_seen": 15939720,
-      "step": 990
-    },
-    {
-      "epoch": 0.06941781151768281,
-      "grad_norm": 4.296191692352295,
-      "learning_rate": 9.306473555166376e-05,
-      "loss": 1.0375,
-      "num_input_tokens_seen": 15956104,
-      "step": 991
-    },
-    {
-      "epoch": 0.06948785976341205,
-      "grad_norm": 5.266847133636475,
-      "learning_rate": 9.305773730297724e-05,
-      "loss": 1.1645,
-      "num_input_tokens_seen": 15972488,
-      "step": 992
-    },
-    {
-      "epoch": 0.0695579080091413,
-      "grad_norm": 4.321287155151367,
-      "learning_rate": 9.305073905429072e-05,
-      "loss": 1.046,
-      "num_input_tokens_seen": 15988408,
-      "step": 993
-    },
-    {
-      "epoch": 0.06962795625487055,
-      "grad_norm": 4.1421613693237305,
-      "learning_rate": 9.304374080560421e-05,
-      "loss": 1.0639,
-      "num_input_tokens_seen": 16002904,
-      "step": 994
-    },
-    {
-      "epoch": 0.06969800450059979,
-      "grad_norm": 6.811270713806152,
-      "learning_rate": 9.30367425569177e-05,
-      "loss": 1.1012,
-      "num_input_tokens_seen": 16017424,
-      "step": 995
-    },
-    {
-      "epoch": 0.06976805274632904,
-      "grad_norm": 4.968684196472168,
-      "learning_rate": 9.302974430823117e-05,
-      "loss": 1.0935,
-      "num_input_tokens_seen": 16033808,
-      "step": 996
-    },
-    {
-      "epoch": 0.06983810099205828,
-      "grad_norm": 4.592737197875977,
-      "learning_rate": 9.302274605954466e-05,
-      "loss": 0.9698,
-      "num_input_tokens_seen": 16050192,
-      "step": 997
-    },
-    {
-      "epoch": 0.06990814923778753,
-      "grad_norm": 3.7984917163848877,
-      "learning_rate": 9.301574781085815e-05,
-      "loss": 1.0976,
-      "num_input_tokens_seen": 16066192,
-      "step": 998
-    },
-    {
-      "epoch": 0.06997819748351677,
-      "grad_norm": 4.594212055206299,
-      "learning_rate": 9.300874956217164e-05,
-      "loss": 1.3718,
-      "num_input_tokens_seen": 16082576,
-      "step": 999
-    },
-    {
-      "epoch": 0.07004824572924602,
-      "grad_norm": 5.062666893005371,
-      "learning_rate": 9.300175131348511e-05,
-      "loss": 1.3139,
-      "num_input_tokens_seen": 16098960,
-      "step": 1000
-    },
-    {
-      "epoch": 0.07004824572924602,
-      "eval_loss": 1.1650840044021606,
-      "eval_runtime": 0.192,
-      "eval_samples_per_second": 5.208,
-      "eval_steps_per_second": 5.208,
-      "num_input_tokens_seen": 16098960,
-      "step": 1000
-    },
-    {
-      "epoch": 0.07011829397497527,
-      "grad_norm": 4.100902557373047,
-      "learning_rate": 9.29947530647986e-05,
-      "loss": 1.2711,
-      "num_input_tokens_seen": 16115216,
-      "step": 1001
-    },
-    {
-      "epoch": 0.07018834222070451,
-      "grad_norm": 4.24728536605835,
-      "learning_rate": 9.298775481611209e-05,
-      "loss": 0.9946,
-      "num_input_tokens_seen": 16130080,
-      "step": 1002
-    },
-    {
-      "epoch": 0.07025839046643376,
-      "grad_norm": 3.4653356075286865,
-      "learning_rate": 9.298075656742556e-05,
-      "loss": 0.8736,
-      "num_input_tokens_seen": 16146400,
-      "step": 1003
-    },
-    {
-      "epoch": 0.070328438712163,
-      "grad_norm": 5.548775672912598,
-      "learning_rate": 9.297375831873907e-05,
-      "loss": 0.9841,
-      "num_input_tokens_seen": 16162784,
-      "step": 1004
-    },
-    {
-      "epoch": 0.07039848695789225,
-      "grad_norm": 4.11661958694458,
-      "learning_rate": 9.296676007005256e-05,
-      "loss": 0.9857,
-      "num_input_tokens_seen": 16179024,
-      "step": 1005
-    },
-    {
-      "epoch": 0.0704685352036215,
-      "grad_norm": 4.006300449371338,
-      "learning_rate": 9.295976182136603e-05,
-      "loss": 1.0587,
-      "num_input_tokens_seen": 16195408,
-      "step": 1006
-    },
-    {
-      "epoch": 0.07053858344935074,
-      "grad_norm": 4.418802261352539,
-      "learning_rate": 9.295276357267952e-05,
-      "loss": 1.3845,
-      "num_input_tokens_seen": 16211792,
-      "step": 1007
-    },
-    {
-      "epoch": 0.07060863169507998,
-      "grad_norm": 5.625720024108887,
-      "learning_rate": 9.2945765323993e-05,
-      "loss": 1.2198,
-      "num_input_tokens_seen": 16226584,
-      "step": 1008
-    },
-    {
-      "epoch": 0.07067867994080923,
-      "grad_norm": 4.209630489349365,
-      "learning_rate": 9.293876707530648e-05,
-      "loss": 0.9387,
-      "num_input_tokens_seen": 16242256,
-      "step": 1009
-    },
-    {
-      "epoch": 0.07074872818653848,
-      "grad_norm": 4.0324788093566895,
-      "learning_rate": 9.293176882661997e-05,
-      "loss": 1.0713,
-      "num_input_tokens_seen": 16258640,
-      "step": 1010
-    },
-    {
-      "epoch": 0.07081877643226772,
-      "grad_norm": 4.0557684898376465,
-      "learning_rate": 9.292477057793346e-05,
-      "loss": 1.2831,
-      "num_input_tokens_seen": 16275024,
-      "step": 1011
-    },
-    {
-      "epoch": 0.07088882467799697,
-      "grad_norm": 4.511384010314941,
-      "learning_rate": 9.291777232924695e-05,
-      "loss": 1.1949,
-      "num_input_tokens_seen": 16291112,
-      "step": 1012
-    },
-    {
-      "epoch": 0.07095887292372621,
-      "grad_norm": 3.8120172023773193,
-      "learning_rate": 9.291077408056042e-05,
-      "loss": 1.013,
-      "num_input_tokens_seen": 16307496,
-      "step": 1013
-    },
-    {
-      "epoch": 0.07102892116945546,
-      "grad_norm": 4.039558410644531,
-      "learning_rate": 9.290377583187391e-05,
-      "loss": 1.1575,
-      "num_input_tokens_seen": 16323880,
-      "step": 1014
-    },
-    {
-      "epoch": 0.0710989694151847,
-      "grad_norm": 3.9076366424560547,
-      "learning_rate": 9.289677758318739e-05,
-      "loss": 1.1776,
-      "num_input_tokens_seen": 16339624,
-      "step": 1015
-    },
-    {
-      "epoch": 0.07116901766091395,
-      "grad_norm": 3.8083527088165283,
-      "learning_rate": 9.288977933450088e-05,
-      "loss": 0.965,
-      "num_input_tokens_seen": 16356008,
-      "step": 1016
-    },
-    {
-      "epoch": 0.0712390659066432,
-      "grad_norm": 4.5387282371521,
-      "learning_rate": 9.288278108581436e-05,
-      "loss": 1.1113,
-      "num_input_tokens_seen": 16372392,
-      "step": 1017
-    },
-    {
-      "epoch": 0.07130911415237244,
-      "grad_norm": 3.9228522777557373,
-      "learning_rate": 9.287578283712785e-05,
-      "loss": 1.1609,
-      "num_input_tokens_seen": 16388776,
-      "step": 1018
-    },
-    {
-      "epoch": 0.07137916239810169,
-      "grad_norm": 4.170912742614746,
-      "learning_rate": 9.286878458844134e-05,
-      "loss": 1.1324,
-      "num_input_tokens_seen": 16405160,
-      "step": 1019
-    },
-    {
-      "epoch": 0.07144921064383093,
-      "grad_norm": 4.426759719848633,
-      "learning_rate": 9.286178633975482e-05,
-      "loss": 1.2825,
-      "num_input_tokens_seen": 16421544,
-      "step": 1020
-    },
-    {
-      "epoch": 0.07151925888956018,
-      "grad_norm": 3.8606133460998535,
-      "learning_rate": 9.28547880910683e-05,
-      "loss": 1.1734,
-      "num_input_tokens_seen": 16437736,
-      "step": 1021
-    },
-    {
-      "epoch": 0.07158930713528944,
-      "grad_norm": 4.040006637573242,
-      "learning_rate": 9.28477898423818e-05,
-      "loss": 1.0824,
-      "num_input_tokens_seen": 16453776,
-      "step": 1022
-    },
-    {
-      "epoch": 0.07165935538101868,
-      "grad_norm": 3.7698042392730713,
-      "learning_rate": 9.284079159369527e-05,
-      "loss": 1.0951,
-      "num_input_tokens_seen": 16470160,
-      "step": 1023
-    },
-    {
-      "epoch": 0.07172940362674793,
-      "grad_norm": 4.180328369140625,
-      "learning_rate": 9.283379334500877e-05,
-      "loss": 1.0087,
-      "num_input_tokens_seen": 16486280,
-      "step": 1024
-    },
-    {
-      "epoch": 0.07179945187247717,
-      "grad_norm": 6.02299690246582,
-      "learning_rate": 9.282679509632225e-05,
-      "loss": 0.9788,
-      "num_input_tokens_seen": 16501784,
-      "step": 1025
-    },
-    {
-      "epoch": 0.07186950011820642,
-      "grad_norm": 4.239454746246338,
-      "learning_rate": 9.281979684763573e-05,
-      "loss": 1.3031,
-      "num_input_tokens_seen": 16518096,
-      "step": 1026
-    },
-    {
-      "epoch": 0.07193954836393567,
-      "grad_norm": 3.446030616760254,
-      "learning_rate": 9.281279859894921e-05,
-      "loss": 0.9523,
-      "num_input_tokens_seen": 16534480,
-      "step": 1027
-    },
-    {
-      "epoch": 0.07200959660966491,
-      "grad_norm": 4.2813568115234375,
-      "learning_rate": 9.28058003502627e-05,
-      "loss": 1.1041,
-      "num_input_tokens_seen": 16550864,
-      "step": 1028
-    },
-    {
-      "epoch": 0.07207964485539416,
-      "grad_norm": 5.289443016052246,
-      "learning_rate": 9.279880210157619e-05,
-      "loss": 1.3036,
-      "num_input_tokens_seen": 16567248,
-      "step": 1029
-    },
-    {
-      "epoch": 0.0721496931011234,
-      "grad_norm": 3.680283308029175,
-      "learning_rate": 9.279180385288967e-05,
-      "loss": 1.1434,
-      "num_input_tokens_seen": 16583632,
-      "step": 1030
-    },
-    {
-      "epoch": 0.07221974134685265,
-      "grad_norm": 4.283925533294678,
-      "learning_rate": 9.278480560420316e-05,
-      "loss": 1.1569,
-      "num_input_tokens_seen": 16600016,
-      "step": 1031
-    },
-    {
-      "epoch": 0.0722897895925819,
-      "grad_norm": 4.913532733917236,
-      "learning_rate": 9.277780735551665e-05,
-      "loss": 1.218,
-      "num_input_tokens_seen": 16616400,
-      "step": 1032
-    },
-    {
-      "epoch": 0.07235983783831114,
-      "grad_norm": 4.344277381896973,
-      "learning_rate": 9.277080910683013e-05,
-      "loss": 1.1495,
-      "num_input_tokens_seen": 16632024,
-      "step": 1033
-    },
-    {
-      "epoch": 0.07242988608404038,
-      "grad_norm": 3.9231889247894287,
-      "learning_rate": 9.276381085814362e-05,
-      "loss": 1.0492,
-      "num_input_tokens_seen": 16648408,
-      "step": 1034
-    },
-    {
-      "epoch": 0.07249993432976963,
-      "grad_norm": 4.062288284301758,
-      "learning_rate": 9.275681260945709e-05,
-      "loss": 0.927,
-      "num_input_tokens_seen": 16664792,
-      "step": 1035
-    },
-    {
-      "epoch": 0.07256998257549888,
-      "grad_norm": 4.163131237030029,
-      "learning_rate": 9.274981436077058e-05,
-      "loss": 1.0782,
-      "num_input_tokens_seen": 16680216,
-      "step": 1036
-    },
-    {
-      "epoch": 0.07264003082122812,
-      "grad_norm": 5.220231056213379,
-      "learning_rate": 9.274281611208407e-05,
-      "loss": 1.125,
-      "num_input_tokens_seen": 16696160,
-      "step": 1037
-    },
-    {
-      "epoch": 0.07271007906695737,
-      "grad_norm": 3.63785457611084,
-      "learning_rate": 9.273581786339756e-05,
-      "loss": 1.0229,
-      "num_input_tokens_seen": 16712544,
-      "step": 1038
-    },
-    {
-      "epoch": 0.07278012731268661,
-      "grad_norm": 4.612295627593994,
-      "learning_rate": 9.272881961471105e-05,
-      "loss": 1.3076,
-      "num_input_tokens_seen": 16728928,
-      "step": 1039
-    },
-    {
-      "epoch": 0.07285017555841586,
-      "grad_norm": 5.278262615203857,
-      "learning_rate": 9.272182136602452e-05,
-      "loss": 1.2682,
-      "num_input_tokens_seen": 16744184,
-      "step": 1040
-    },
-    {
-      "epoch": 0.0729202238041451,
-      "grad_norm": 4.3274455070495605,
-      "learning_rate": 9.271482311733801e-05,
-      "loss": 1.3517,
-      "num_input_tokens_seen": 16760056,
-      "step": 1041
-    },
-    {
-      "epoch": 0.07299027204987435,
-      "grad_norm": 4.1077375411987305,
-      "learning_rate": 9.270782486865148e-05,
-      "loss": 1.175,
-      "num_input_tokens_seen": 16776280,
-      "step": 1042
-    },
-    {
-      "epoch": 0.0730603202956036,
-      "grad_norm": 3.954604148864746,
-      "learning_rate": 9.270082661996497e-05,
-      "loss": 1.189,
-      "num_input_tokens_seen": 16792456,
-      "step": 1043
-    },
-    {
-      "epoch": 0.07313036854133284,
-      "grad_norm": 4.111297607421875,
-      "learning_rate": 9.269382837127847e-05,
-      "loss": 1.0265,
-      "num_input_tokens_seen": 16808840,
-      "step": 1044
-    },
-    {
-      "epoch": 0.07320041678706209,
-      "grad_norm": 3.56953501701355,
-      "learning_rate": 9.268683012259195e-05,
-      "loss": 1.0114,
-      "num_input_tokens_seen": 16824720,
-      "step": 1045
-    },
-    {
-      "epoch": 0.07327046503279133,
-      "grad_norm": 4.962648868560791,
-      "learning_rate": 9.267983187390544e-05,
-      "loss": 1.1714,
-      "num_input_tokens_seen": 16841104,
-      "step": 1046
-    },
-    {
-      "epoch": 0.07334051327852058,
-      "grad_norm": 3.7930710315704346,
-      "learning_rate": 9.267283362521891e-05,
-      "loss": 1.0903,
-      "num_input_tokens_seen": 16857488,
-      "step": 1047
-    },
-    {
-      "epoch": 0.07341056152424982,
-      "grad_norm": 4.158027172088623,
-      "learning_rate": 9.26658353765324e-05,
-      "loss": 1.1823,
-      "num_input_tokens_seen": 16873856,
-      "step": 1048
-    },
-    {
-      "epoch": 0.07348060976997907,
-      "grad_norm": 4.1571197509765625,
-      "learning_rate": 9.265883712784589e-05,
-      "loss": 1.2572,
-      "num_input_tokens_seen": 16890240,
-      "step": 1049
-    },
-    {
-      "epoch": 0.07355065801570831,
-      "grad_norm": 4.330874443054199,
-      "learning_rate": 9.265183887915938e-05,
-      "loss": 1.194,
-      "num_input_tokens_seen": 16906624,
-      "step": 1050
-    },
-    {
-      "epoch": 0.07362070626143756,
-      "grad_norm": 6.105716705322266,
-      "learning_rate": 9.264484063047287e-05,
-      "loss": 1.0685,
-      "num_input_tokens_seen": 16922864,
-      "step": 1051
-    },
-    {
-      "epoch": 0.0736907545071668,
-      "grad_norm": 4.8344407081604,
-      "learning_rate": 9.263784238178634e-05,
-      "loss": 1.1992,
-      "num_input_tokens_seen": 16939200,
-      "step": 1052
-    },
-    {
-      "epoch": 0.07376080275289605,
-      "grad_norm": 3.553568124771118,
-      "learning_rate": 9.263084413309983e-05,
-      "loss": 0.7907,
-      "num_input_tokens_seen": 16955584,
-      "step": 1053
-    },
-    {
-      "epoch": 0.0738308509986253,
-      "grad_norm": 3.8178694248199463,
-      "learning_rate": 9.26238458844133e-05,
-      "loss": 1.2031,
-      "num_input_tokens_seen": 16971968,
-      "step": 1054
-    },
-    {
-      "epoch": 0.07390089924435454,
-      "grad_norm": 3.5509321689605713,
-      "learning_rate": 9.26168476357268e-05,
-      "loss": 1.1189,
-      "num_input_tokens_seen": 16988352,
-      "step": 1055
-    },
-    {
-      "epoch": 0.0739709474900838,
-      "grad_norm": 3.870811939239502,
-      "learning_rate": 9.260984938704028e-05,
-      "loss": 1.0205,
-      "num_input_tokens_seen": 17004736,
-      "step": 1056
-    },
-    {
-      "epoch": 0.07404099573581305,
-      "grad_norm": 11.86201286315918,
-      "learning_rate": 9.260285113835377e-05,
-      "loss": 1.037,
-      "num_input_tokens_seen": 17020544,
-      "step": 1057
-    },
-    {
-      "epoch": 0.0741110439815423,
-      "grad_norm": 5.2176127433776855,
-      "learning_rate": 9.259585288966726e-05,
-      "loss": 1.0797,
-      "num_input_tokens_seen": 17036472,
-      "step": 1058
-    },
-    {
-      "epoch": 0.07418109222727154,
-      "grad_norm": 3.72566819190979,
-      "learning_rate": 9.258885464098075e-05,
-      "loss": 0.9307,
-      "num_input_tokens_seen": 17052360,
-      "step": 1059
-    },
-    {
-      "epoch": 0.07425114047300078,
-      "grad_norm": 4.323361396789551,
-      "learning_rate": 9.258185639229422e-05,
-      "loss": 1.0783,
-      "num_input_tokens_seen": 17067672,
-      "step": 1060
-    },
-    {
-      "epoch": 0.07432118871873003,
-      "grad_norm": 4.01705265045166,
-      "learning_rate": 9.257485814360771e-05,
-      "loss": 1.0402,
-      "num_input_tokens_seen": 17084056,
-      "step": 1061
-    },
-    {
-      "epoch": 0.07439123696445928,
-      "grad_norm": 4.4460039138793945,
-      "learning_rate": 9.256785989492119e-05,
-      "loss": 1.2294,
-      "num_input_tokens_seen": 17100096,
-      "step": 1062
-    },
-    {
-      "epoch": 0.07446128521018852,
-      "grad_norm": 4.634500503540039,
-      "learning_rate": 9.256086164623468e-05,
-      "loss": 1.1479,
-      "num_input_tokens_seen": 17116440,
-      "step": 1063
-    },
-    {
-      "epoch": 0.07453133345591777,
-      "grad_norm": 4.146971702575684,
-      "learning_rate": 9.255386339754817e-05,
-      "loss": 0.9052,
-      "num_input_tokens_seen": 17132592,
-      "step": 1064
-    },
-    {
-      "epoch": 0.07460138170164701,
-      "grad_norm": 6.171874523162842,
-      "learning_rate": 9.254686514886165e-05,
-      "loss": 1.1135,
-      "num_input_tokens_seen": 17148704,
-      "step": 1065
-    },
-    {
-      "epoch": 0.07467142994737626,
-      "grad_norm": 6.25461483001709,
-      "learning_rate": 9.253986690017514e-05,
-      "loss": 1.0003,
-      "num_input_tokens_seen": 17164920,
-      "step": 1066
-    },
-    {
-      "epoch": 0.0747414781931055,
-      "grad_norm": 3.886582851409912,
-      "learning_rate": 9.253286865148862e-05,
-      "loss": 1.1917,
-      "num_input_tokens_seen": 17181304,
-      "step": 1067
-    },
-    {
-      "epoch": 0.07481152643883475,
-      "grad_norm": 5.067885398864746,
-      "learning_rate": 9.25258704028021e-05,
-      "loss": 1.4475,
-      "num_input_tokens_seen": 17197208,
-      "step": 1068
-    },
-    {
-      "epoch": 0.074881574684564,
-      "grad_norm": 4.186190128326416,
-      "learning_rate": 9.251887215411558e-05,
-      "loss": 1.1255,
-      "num_input_tokens_seen": 17212680,
-      "step": 1069
-    },
-    {
-      "epoch": 0.07495162293029324,
-      "grad_norm": 4.059047698974609,
-      "learning_rate": 9.251187390542908e-05,
-      "loss": 1.1467,
-      "num_input_tokens_seen": 17229064,
-      "step": 1070
-    },
-    {
-      "epoch": 0.07502167117602249,
-      "grad_norm": 4.154530048370361,
-      "learning_rate": 9.250487565674257e-05,
-      "loss": 1.0811,
-      "num_input_tokens_seen": 17245448,
-      "step": 1071
-    },
-    {
-      "epoch": 0.07509171942175173,
-      "grad_norm": 3.760453701019287,
-      "learning_rate": 9.249787740805605e-05,
-      "loss": 1.1493,
-      "num_input_tokens_seen": 17261832,
-      "step": 1072
-    },
-    {
-      "epoch": 0.07516176766748098,
-      "grad_norm": 3.8155417442321777,
-      "learning_rate": 9.249087915936954e-05,
-      "loss": 1.0934,
-      "num_input_tokens_seen": 17278216,
-      "step": 1073
-    },
-    {
-      "epoch": 0.07523181591321022,
-      "grad_norm": 4.807973384857178,
-      "learning_rate": 9.248388091068301e-05,
-      "loss": 1.0704,
-      "num_input_tokens_seen": 17294600,
-      "step": 1074
-    },
-    {
-      "epoch": 0.07530186415893947,
-      "grad_norm": 11.421661376953125,
-      "learning_rate": 9.24768826619965e-05,
-      "loss": 0.9472,
-      "num_input_tokens_seen": 17308960,
-      "step": 1075
-    },
-    {
-      "epoch": 0.07537191240466871,
-      "grad_norm": 3.7491819858551025,
-      "learning_rate": 9.246988441330999e-05,
-      "loss": 1.1395,
-      "num_input_tokens_seen": 17324536,
-      "step": 1076
-    },
-    {
-      "epoch": 0.07544196065039796,
-      "grad_norm": 3.6289992332458496,
-      "learning_rate": 9.246288616462348e-05,
-      "loss": 0.9375,
-      "num_input_tokens_seen": 17340920,
-      "step": 1077
-    },
-    {
-      "epoch": 0.0755120088961272,
-      "grad_norm": 5.741896629333496,
-      "learning_rate": 9.245588791593696e-05,
-      "loss": 1.1656,
-      "num_input_tokens_seen": 17357304,
-      "step": 1078
-    },
-    {
-      "epoch": 0.07558205714185645,
-      "grad_norm": 3.5879697799682617,
-      "learning_rate": 9.244888966725044e-05,
-      "loss": 0.9421,
-      "num_input_tokens_seen": 17373592,
-      "step": 1079
-    },
-    {
-      "epoch": 0.0756521053875857,
-      "grad_norm": 7.3384504318237305,
-      "learning_rate": 9.244189141856393e-05,
-      "loss": 1.1358,
-      "num_input_tokens_seen": 17387872,
-      "step": 1080
-    },
-    {
-      "epoch": 0.07572215363331494,
-      "grad_norm": 3.6677255630493164,
-      "learning_rate": 9.24348931698774e-05,
-      "loss": 0.892,
-      "num_input_tokens_seen": 17403088,
-      "step": 1081
-    },
-    {
-      "epoch": 0.07579220187904419,
-      "grad_norm": 3.953216075897217,
-      "learning_rate": 9.242789492119089e-05,
-      "loss": 0.9757,
-      "num_input_tokens_seen": 17419392,
-      "step": 1082
-    },
-    {
-      "epoch": 0.07586225012477343,
-      "grad_norm": 4.827987194061279,
-      "learning_rate": 9.242089667250438e-05,
-      "loss": 1.1493,
-      "num_input_tokens_seen": 17435776,
-      "step": 1083
-    },
-    {
-      "epoch": 0.07593229837050268,
-      "grad_norm": 4.416223526000977,
-      "learning_rate": 9.241389842381787e-05,
-      "loss": 0.9913,
-      "num_input_tokens_seen": 17452080,
-      "step": 1084
-    },
-    {
-      "epoch": 0.07600234661623193,
-      "grad_norm": 3.7776753902435303,
-      "learning_rate": 9.240690017513136e-05,
-      "loss": 1.0589,
-      "num_input_tokens_seen": 17468160,
-      "step": 1085
-    },
-    {
-      "epoch": 0.07607239486196117,
-      "grad_norm": 4.139477252960205,
-      "learning_rate": 9.239990192644485e-05,
-      "loss": 0.9475,
-      "num_input_tokens_seen": 17484544,
-      "step": 1086
-    },
-    {
-      "epoch": 0.07614244310769042,
-      "grad_norm": 5.218942642211914,
-      "learning_rate": 9.239290367775832e-05,
-      "loss": 1.1626,
-      "num_input_tokens_seen": 17500928,
-      "step": 1087
-    },
-    {
-      "epoch": 0.07621249135341966,
-      "grad_norm": 4.773080348968506,
-      "learning_rate": 9.238590542907181e-05,
-      "loss": 1.154,
-      "num_input_tokens_seen": 17517312,
-      "step": 1088
-    },
-    {
-      "epoch": 0.07628253959914891,
-      "grad_norm": 3.840151309967041,
-      "learning_rate": 9.237890718038528e-05,
-      "loss": 1.0862,
-      "num_input_tokens_seen": 17533696,
-      "step": 1089
-    },
-    {
-      "epoch": 0.07635258784487815,
-      "grad_norm": 4.201962471008301,
-      "learning_rate": 9.237190893169879e-05,
-      "loss": 1.0945,
-      "num_input_tokens_seen": 17549512,
-      "step": 1090
-    },
-    {
-      "epoch": 0.07642263609060741,
-      "grad_norm": 4.4583001136779785,
-      "learning_rate": 9.236491068301226e-05,
-      "loss": 1.074,
-      "num_input_tokens_seen": 17565896,
-      "step": 1091
-    },
-    {
-      "epoch": 0.07649268433633666,
-      "grad_norm": 4.013672351837158,
-      "learning_rate": 9.235791243432575e-05,
-      "loss": 1.2545,
-      "num_input_tokens_seen": 17582264,
-      "step": 1092
-    },
-    {
-      "epoch": 0.0765627325820659,
-      "grad_norm": 3.69555926322937,
-      "learning_rate": 9.235091418563924e-05,
-      "loss": 1.1615,
-      "num_input_tokens_seen": 17597888,
-      "step": 1093
-    },
-    {
-      "epoch": 0.07663278082779515,
-      "grad_norm": 4.341784954071045,
-      "learning_rate": 9.234391593695271e-05,
-      "loss": 1.0369,
-      "num_input_tokens_seen": 17613392,
-      "step": 1094
-    },
-    {
-      "epoch": 0.0767028290735244,
-      "grad_norm": 4.043522357940674,
-      "learning_rate": 9.23369176882662e-05,
-      "loss": 1.0509,
-      "num_input_tokens_seen": 17629216,
-      "step": 1095
-    },
-    {
-      "epoch": 0.07677287731925364,
-      "grad_norm": 4.330739498138428,
-      "learning_rate": 9.232991943957969e-05,
-      "loss": 1.2208,
-      "num_input_tokens_seen": 17645600,
-      "step": 1096
-    },
-    {
-      "epoch": 0.07684292556498289,
-      "grad_norm": 4.8433122634887695,
-      "learning_rate": 9.232292119089318e-05,
-      "loss": 0.9492,
-      "num_input_tokens_seen": 17660952,
-      "step": 1097
-    },
-    {
-      "epoch": 0.07691297381071213,
-      "grad_norm": 3.9039859771728516,
-      "learning_rate": 9.231592294220667e-05,
-      "loss": 1.0601,
-      "num_input_tokens_seen": 17677336,
-      "step": 1098
-    },
-    {
-      "epoch": 0.07698302205644138,
-      "grad_norm": 3.814103126525879,
-      "learning_rate": 9.230892469352014e-05,
-      "loss": 0.9902,
-      "num_input_tokens_seen": 17693720,
-      "step": 1099
-    },
-    {
-      "epoch": 0.07705307030217062,
-      "grad_norm": 3.9864039421081543,
-      "learning_rate": 9.230192644483363e-05,
-      "loss": 1.1622,
-      "num_input_tokens_seen": 17710104,
-      "step": 1100
-    },
-    {
-      "epoch": 0.07712311854789987,
-      "grad_norm": 4.469820499420166,
-      "learning_rate": 9.229492819614711e-05,
-      "loss": 1.044,
-      "num_input_tokens_seen": 17726488,
-      "step": 1101
-    },
-    {
-      "epoch": 0.07719316679362911,
-      "grad_norm": 3.8044216632843018,
-      "learning_rate": 9.22879299474606e-05,
-      "loss": 1.1283,
-      "num_input_tokens_seen": 17742648,
-      "step": 1102
-    },
-    {
-      "epoch": 0.07726321503935836,
-      "grad_norm": 4.859435558319092,
-      "learning_rate": 9.228093169877408e-05,
-      "loss": 1.0995,
-      "num_input_tokens_seen": 17759032,
-      "step": 1103
-    },
-    {
-      "epoch": 0.0773332632850876,
-      "grad_norm": 3.830214023590088,
-      "learning_rate": 9.227393345008757e-05,
-      "loss": 1.1731,
-      "num_input_tokens_seen": 17774872,
-      "step": 1104
-    },
-    {
-      "epoch": 0.07740331153081685,
-      "grad_norm": 4.196676254272461,
-      "learning_rate": 9.226693520140106e-05,
-      "loss": 1.2055,
-      "num_input_tokens_seen": 17790832,
-      "step": 1105
-    },
-    {
-      "epoch": 0.0774733597765461,
-      "grad_norm": 4.50007438659668,
-      "learning_rate": 9.225993695271454e-05,
-      "loss": 0.952,
-      "num_input_tokens_seen": 17805024,
-      "step": 1106
-    },
-    {
-      "epoch": 0.07754340802227534,
-      "grad_norm": 4.392070293426514,
-      "learning_rate": 9.225293870402803e-05,
-      "loss": 1.1548,
-      "num_input_tokens_seen": 17820008,
-      "step": 1107
-    },
-    {
-      "epoch": 0.07761345626800459,
-      "grad_norm": 4.09447717666626,
-      "learning_rate": 9.22459404553415e-05,
-      "loss": 1.1233,
-      "num_input_tokens_seen": 17836392,
-      "step": 1108
-    },
-    {
-      "epoch": 0.07768350451373383,
-      "grad_norm": 4.591554641723633,
-      "learning_rate": 9.223894220665499e-05,
-      "loss": 1.2772,
-      "num_input_tokens_seen": 17852776,
-      "step": 1109
-    },
-    {
-      "epoch": 0.07775355275946308,
-      "grad_norm": 5.629931926727295,
-      "learning_rate": 9.223194395796849e-05,
-      "loss": 1.1453,
-      "num_input_tokens_seen": 17869160,
-      "step": 1110
-    },
-    {
-      "epoch": 0.07782360100519232,
-      "grad_norm": 4.307553768157959,
-      "learning_rate": 9.222494570928197e-05,
-      "loss": 1.1479,
-      "num_input_tokens_seen": 17885544,
-      "step": 1111
-    },
-    {
-      "epoch": 0.07789364925092157,
-      "grad_norm": 4.599300384521484,
-      "learning_rate": 9.221794746059545e-05,
-      "loss": 1.1304,
-      "num_input_tokens_seen": 17901848,
-      "step": 1112
-    },
-    {
-      "epoch": 0.07796369749665082,
-      "grad_norm": 4.217408657073975,
-      "learning_rate": 9.221094921190894e-05,
-      "loss": 1.1611,
-      "num_input_tokens_seen": 17918232,
-      "step": 1113
-    },
-    {
-      "epoch": 0.07803374574238006,
-      "grad_norm": 3.885847568511963,
-      "learning_rate": 9.220395096322242e-05,
-      "loss": 0.968,
-      "num_input_tokens_seen": 17934504,
-      "step": 1114
-    },
-    {
-      "epoch": 0.07810379398810931,
-      "grad_norm": 4.280134677886963,
-      "learning_rate": 9.219695271453591e-05,
-      "loss": 1.0944,
-      "num_input_tokens_seen": 17950888,
-      "step": 1115
-    },
-    {
-      "epoch": 0.07817384223383855,
-      "grad_norm": 4.081259727478027,
-      "learning_rate": 9.21899544658494e-05,
-      "loss": 1.0872,
-      "num_input_tokens_seen": 17967088,
-      "step": 1116
-    },
-    {
-      "epoch": 0.0782438904795678,
-      "grad_norm": 4.206293106079102,
-      "learning_rate": 9.218295621716288e-05,
-      "loss": 1.2013,
-      "num_input_tokens_seen": 17983312,
-      "step": 1117
-    },
-    {
-      "epoch": 0.07831393872529704,
-      "grad_norm": 4.837226390838623,
-      "learning_rate": 9.217595796847636e-05,
-      "loss": 1.2628,
-      "num_input_tokens_seen": 17998768,
-      "step": 1118
-    },
-    {
-      "epoch": 0.07838398697102629,
-      "grad_norm": 4.344440460205078,
-      "learning_rate": 9.216895971978985e-05,
-      "loss": 1.0389,
-      "num_input_tokens_seen": 18014840,
-      "step": 1119
-    },
-    {
-      "epoch": 0.07845403521675554,
-      "grad_norm": 4.357896327972412,
-      "learning_rate": 9.216196147110334e-05,
-      "loss": 1.2444,
-      "num_input_tokens_seen": 18030696,
-      "step": 1120
-    },
-    {
-      "epoch": 0.07852408346248478,
-      "grad_norm": 3.6449878215789795,
-      "learning_rate": 9.215496322241681e-05,
-      "loss": 1.0622,
-      "num_input_tokens_seen": 18047024,
-      "step": 1121
-    },
-    {
-      "epoch": 0.07859413170821403,
-      "grad_norm": 4.154385566711426,
-      "learning_rate": 9.21479649737303e-05,
-      "loss": 1.1551,
-      "num_input_tokens_seen": 18063408,
-      "step": 1122
-    },
-    {
-      "epoch": 0.07866417995394327,
-      "grad_norm": 3.5929031372070312,
-      "learning_rate": 9.214096672504379e-05,
-      "loss": 0.9682,
-      "num_input_tokens_seen": 18079280,
-      "step": 1123
-    },
-    {
-      "epoch": 0.07873422819967252,
-      "grad_norm": 3.5724170207977295,
-      "learning_rate": 9.213396847635728e-05,
-      "loss": 0.8952,
-      "num_input_tokens_seen": 18094488,
-      "step": 1124
-    },
-    {
-      "epoch": 0.07880427644540176,
-      "grad_norm": 4.100067615509033,
-      "learning_rate": 9.212697022767077e-05,
-      "loss": 0.9066,
-      "num_input_tokens_seen": 18110872,
-      "step": 1125
-    },
-    {
-      "epoch": 0.07887432469113102,
-      "grad_norm": 4.431338787078857,
-      "learning_rate": 9.211997197898424e-05,
-      "loss": 1.0116,
-      "num_input_tokens_seen": 18127256,
-      "step": 1126
-    },
-    {
-      "epoch": 0.07894437293686027,
-      "grad_norm": 3.9577043056488037,
-      "learning_rate": 9.211297373029773e-05,
-      "loss": 1.1299,
-      "num_input_tokens_seen": 18143208,
-      "step": 1127
-    },
-    {
-      "epoch": 0.07901442118258951,
-      "grad_norm": 4.753921985626221,
-      "learning_rate": 9.21059754816112e-05,
-      "loss": 1.0686,
-      "num_input_tokens_seen": 18158888,
-      "step": 1128
-    },
-    {
-      "epoch": 0.07908446942831876,
-      "grad_norm": 3.763982057571411,
-      "learning_rate": 9.209897723292469e-05,
-      "loss": 1.0467,
-      "num_input_tokens_seen": 18175192,
-      "step": 1129
-    },
-    {
-      "epoch": 0.079154517674048,
-      "grad_norm": 3.729553699493408,
-      "learning_rate": 9.20919789842382e-05,
-      "loss": 1.1152,
-      "num_input_tokens_seen": 18191384,
-      "step": 1130
-    },
-    {
-      "epoch": 0.07922456591977725,
-      "grad_norm": 3.7760956287384033,
-      "learning_rate": 9.208498073555167e-05,
-      "loss": 1.0994,
-      "num_input_tokens_seen": 18207768,
-      "step": 1131
-    },
-    {
-      "epoch": 0.0792946141655065,
-      "grad_norm": 4.64035177230835,
-      "learning_rate": 9.207798248686516e-05,
-      "loss": 1.1037,
-      "num_input_tokens_seen": 18224152,
-      "step": 1132
-    },
-    {
-      "epoch": 0.07936466241123574,
-      "grad_norm": 4.1443352699279785,
-      "learning_rate": 9.207098423817863e-05,
-      "loss": 1.2329,
-      "num_input_tokens_seen": 18240536,
-      "step": 1133
-    },
-    {
-      "epoch": 0.07943471065696499,
-      "grad_norm": 5.332706451416016,
-      "learning_rate": 9.206398598949212e-05,
-      "loss": 1.1303,
-      "num_input_tokens_seen": 18255528,
-      "step": 1134
-    },
-    {
-      "epoch": 0.07950475890269423,
-      "grad_norm": 3.914705514907837,
-      "learning_rate": 9.20569877408056e-05,
-      "loss": 1.1182,
-      "num_input_tokens_seen": 18271768,
-      "step": 1135
-    },
-    {
-      "epoch": 0.07957480714842348,
-      "grad_norm": 4.994162559509277,
-      "learning_rate": 9.20499894921191e-05,
-      "loss": 1.175,
-      "num_input_tokens_seen": 18288152,
-      "step": 1136
-    },
-    {
-      "epoch": 0.07964485539415272,
-      "grad_norm": 4.132298946380615,
-      "learning_rate": 9.204299124343259e-05,
-      "loss": 0.9402,
-      "num_input_tokens_seen": 18303784,
-      "step": 1137
-    },
-    {
-      "epoch": 0.07971490363988197,
-      "grad_norm": 3.9048449993133545,
-      "learning_rate": 9.203599299474606e-05,
-      "loss": 1.1283,
-      "num_input_tokens_seen": 18319968,
-      "step": 1138
-    },
-    {
-      "epoch": 0.07978495188561122,
-      "grad_norm": 3.981844425201416,
-      "learning_rate": 9.202899474605955e-05,
-      "loss": 1.0472,
-      "num_input_tokens_seen": 18335976,
-      "step": 1139
-    },
-    {
-      "epoch": 0.07985500013134046,
-      "grad_norm": 4.491240501403809,
-      "learning_rate": 9.202199649737304e-05,
-      "loss": 1.1022,
-      "num_input_tokens_seen": 18352360,
-      "step": 1140
-    },
-    {
-      "epoch": 0.07992504837706971,
-      "grad_norm": 4.152430534362793,
-      "learning_rate": 9.201499824868652e-05,
-      "loss": 1.0688,
-      "num_input_tokens_seen": 18368736,
-      "step": 1141
-    },
-    {
-      "epoch": 0.07999509662279895,
-      "grad_norm": 4.337832450866699,
-      "learning_rate": 9.2008e-05,
-      "loss": 1.0397,
-      "num_input_tokens_seen": 18385120,
-      "step": 1142
-    },
-    {
-      "epoch": 0.0800651448685282,
-      "grad_norm": 4.865042209625244,
-      "learning_rate": 9.200100175131349e-05,
-      "loss": 0.9616,
-      "num_input_tokens_seen": 18401504,
-      "step": 1143
-    },
-    {
-      "epoch": 0.08013519311425744,
-      "grad_norm": 3.783113479614258,
-      "learning_rate": 9.199400350262698e-05,
-      "loss": 1.0001,
-      "num_input_tokens_seen": 18417176,
-      "step": 1144
-    },
-    {
-      "epoch": 0.08020524135998669,
-      "grad_norm": 4.98455286026001,
-      "learning_rate": 9.198700525394046e-05,
-      "loss": 1.2139,
-      "num_input_tokens_seen": 18432584,
-      "step": 1145
-    },
-    {
-      "epoch": 0.08027528960571594,
-      "grad_norm": 4.1859517097473145,
-      "learning_rate": 9.198000700525394e-05,
-      "loss": 1.1333,
-      "num_input_tokens_seen": 18448968,
-      "step": 1146
-    },
-    {
-      "epoch": 0.08034533785144518,
-      "grad_norm": 3.7193386554718018,
-      "learning_rate": 9.197300875656743e-05,
-      "loss": 1.0055,
-      "num_input_tokens_seen": 18465352,
-      "step": 1147
-    },
-    {
-      "epoch": 0.08041538609717443,
-      "grad_norm": 4.280893325805664,
-      "learning_rate": 9.196601050788091e-05,
-      "loss": 1.1261,
-      "num_input_tokens_seen": 18481736,
-      "step": 1148
-    },
-    {
-      "epoch": 0.08048543434290367,
-      "grad_norm": 3.9979352951049805,
-      "learning_rate": 9.19590122591944e-05,
-      "loss": 1.025,
-      "num_input_tokens_seen": 18498120,
-      "step": 1149
-    },
-    {
-      "epoch": 0.08055548258863292,
-      "grad_norm": 5.594225883483887,
-      "learning_rate": 9.195201401050789e-05,
-      "loss": 1.0527,
-      "num_input_tokens_seen": 18513944,
-      "step": 1150
-    },
-    {
-      "epoch": 0.08062553083436216,
-      "grad_norm": 4.758842468261719,
-      "learning_rate": 9.194501576182137e-05,
-      "loss": 1.0915,
-      "num_input_tokens_seen": 18530328,
-      "step": 1151
-    },
-    {
-      "epoch": 0.08069557908009141,
-      "grad_norm": 5.597489356994629,
-      "learning_rate": 9.193801751313486e-05,
-      "loss": 1.0673,
-      "num_input_tokens_seen": 18546632,
-      "step": 1152
-    },
-    {
-      "epoch": 0.08076562732582065,
-      "grad_norm": 5.279472827911377,
-      "learning_rate": 9.193101926444834e-05,
-      "loss": 1.2897,
-      "num_input_tokens_seen": 18561856,
-      "step": 1153
-    },
-    {
-      "epoch": 0.0808356755715499,
-      "grad_norm": 4.672069072723389,
-      "learning_rate": 9.192402101576183e-05,
-      "loss": 1.0298,
-      "num_input_tokens_seen": 18577944,
-      "step": 1154
-    },
-    {
-      "epoch": 0.08090572381727915,
-      "grad_norm": 3.65533447265625,
-      "learning_rate": 9.19170227670753e-05,
-      "loss": 0.933,
-      "num_input_tokens_seen": 18593720,
-      "step": 1155
-    },
-    {
-      "epoch": 0.08097577206300839,
-      "grad_norm": 4.212414741516113,
-      "learning_rate": 9.19100245183888e-05,
-      "loss": 1.0496,
-      "num_input_tokens_seen": 18609864,
-      "step": 1156
-    },
-    {
-      "epoch": 0.08104582030873764,
-      "grad_norm": 4.471503734588623,
-      "learning_rate": 9.190302626970229e-05,
-      "loss": 1.2261,
-      "num_input_tokens_seen": 18626248,
-      "step": 1157
-    },
-    {
-      "epoch": 0.08111586855446688,
-      "grad_norm": 4.952723979949951,
-      "learning_rate": 9.189602802101577e-05,
-      "loss": 1.056,
-      "num_input_tokens_seen": 18642632,
-      "step": 1158
-    },
-    {
-      "epoch": 0.08118591680019613,
-      "grad_norm": 3.921449661254883,
-      "learning_rate": 9.188902977232926e-05,
-      "loss": 1.1617,
-      "num_input_tokens_seen": 18659016,
-      "step": 1159
-    },
-    {
-      "epoch": 0.08125596504592539,
-      "grad_norm": 3.728752374649048,
-      "learning_rate": 9.188203152364273e-05,
-      "loss": 1.1217,
-      "num_input_tokens_seen": 18675400,
-      "step": 1160
-    },
-    {
-      "epoch": 0.08132601329165463,
-      "grad_norm": 3.8742613792419434,
-      "learning_rate": 9.187503327495622e-05,
-      "loss": 1.1538,
-      "num_input_tokens_seen": 18691232,
-      "step": 1161
-    },
-    {
-      "epoch": 0.08139606153738388,
-      "grad_norm": 3.827157735824585,
-      "learning_rate": 9.186803502626971e-05,
-      "loss": 1.1457,
-      "num_input_tokens_seen": 18707616,
-      "step": 1162
-    },
-    {
-      "epoch": 0.08146610978311312,
-      "grad_norm": 3.8507778644561768,
-      "learning_rate": 9.18610367775832e-05,
-      "loss": 1.0317,
-      "num_input_tokens_seen": 18724000,
-      "step": 1163
-    },
-    {
-      "epoch": 0.08153615802884237,
-      "grad_norm": 5.328095436096191,
-      "learning_rate": 9.185403852889669e-05,
-      "loss": 1.0921,
-      "num_input_tokens_seen": 18740384,
-      "step": 1164
-    },
-    {
-      "epoch": 0.08160620627457162,
-      "grad_norm": 4.8900322914123535,
-      "learning_rate": 9.184704028021016e-05,
-      "loss": 1.1308,
-      "num_input_tokens_seen": 18756768,
-      "step": 1165
-    },
-    {
-      "epoch": 0.08167625452030086,
-      "grad_norm": 3.810084104537964,
-      "learning_rate": 9.184004203152365e-05,
-      "loss": 1.1244,
-      "num_input_tokens_seen": 18772632,
-      "step": 1166
-    },
-    {
-      "epoch": 0.08174630276603011,
-      "grad_norm": 4.318419456481934,
-      "learning_rate": 9.183304378283714e-05,
-      "loss": 1.0372,
-      "num_input_tokens_seen": 18788272,
-      "step": 1167
-    },
-    {
-      "epoch": 0.08181635101175935,
-      "grad_norm": 4.093379020690918,
-      "learning_rate": 9.182604553415061e-05,
-      "loss": 1.18,
-      "num_input_tokens_seen": 18803672,
-      "step": 1168
-    },
-    {
-      "epoch": 0.0818863992574886,
-      "grad_norm": 4.630450248718262,
-      "learning_rate": 9.18190472854641e-05,
-      "loss": 1.1439,
-      "num_input_tokens_seen": 18820056,
-      "step": 1169
-    },
-    {
-      "epoch": 0.08195644750321784,
-      "grad_norm": 4.388457775115967,
-      "learning_rate": 9.181204903677759e-05,
-      "loss": 1.0971,
-      "num_input_tokens_seen": 18836440,
-      "step": 1170
-    },
-    {
-      "epoch": 0.08202649574894709,
-      "grad_norm": 3.6942262649536133,
-      "learning_rate": 9.180505078809108e-05,
-      "loss": 1.1594,
-      "num_input_tokens_seen": 18852824,
-      "step": 1171
-    },
-    {
-      "epoch": 0.08209654399467634,
-      "grad_norm": 3.937696933746338,
-      "learning_rate": 9.179805253940455e-05,
-      "loss": 1.1841,
-      "num_input_tokens_seen": 18869208,
-      "step": 1172
-    },
-    {
-      "epoch": 0.08216659224040558,
-      "grad_norm": 4.062703609466553,
-      "learning_rate": 9.179105429071804e-05,
-      "loss": 1.083,
-      "num_input_tokens_seen": 18885320,
-      "step": 1173
-    },
-    {
-      "epoch": 0.08223664048613483,
-      "grad_norm": 7.794081211090088,
-      "learning_rate": 9.178405604203153e-05,
-      "loss": 1.2287,
-      "num_input_tokens_seen": 18900224,
-      "step": 1174
-    },
-    {
-      "epoch": 0.08230668873186407,
-      "grad_norm": 4.429391860961914,
-      "learning_rate": 9.1777057793345e-05,
-      "loss": 1.0504,
-      "num_input_tokens_seen": 18916456,
-      "step": 1175
-    },
-    {
-      "epoch": 0.08237673697759332,
-      "grad_norm": 3.954869508743286,
-      "learning_rate": 9.17700595446585e-05,
-      "loss": 1.1558,
-      "num_input_tokens_seen": 18932840,
-      "step": 1176
-    },
-    {
-      "epoch": 0.08244678522332256,
-      "grad_norm": 5.555337429046631,
-      "learning_rate": 9.176306129597198e-05,
-      "loss": 1.3628,
-      "num_input_tokens_seen": 18949224,
-      "step": 1177
-    },
-    {
-      "epoch": 0.08251683346905181,
-      "grad_norm": 3.575295925140381,
-      "learning_rate": 9.175606304728547e-05,
-      "loss": 1.0651,
-      "num_input_tokens_seen": 18965552,
-      "step": 1178
-    },
-    {
-      "epoch": 0.08258688171478105,
-      "grad_norm": 5.927703380584717,
-      "learning_rate": 9.174906479859896e-05,
-      "loss": 1.0582,
-      "num_input_tokens_seen": 18981496,
-      "step": 1179
-    },
-    {
-      "epoch": 0.0826569299605103,
-      "grad_norm": 6.553986549377441,
-      "learning_rate": 9.174206654991243e-05,
-      "loss": 1.4058,
-      "num_input_tokens_seen": 18996808,
-      "step": 1180
-    },
-    {
-      "epoch": 0.08272697820623955,
-      "grad_norm": 4.315832138061523,
-      "learning_rate": 9.173506830122592e-05,
-      "loss": 1.1166,
-      "num_input_tokens_seen": 19013192,
-      "step": 1181
-    },
-    {
-      "epoch": 0.08279702645196879,
-      "grad_norm": 3.818033218383789,
-      "learning_rate": 9.172807005253941e-05,
-      "loss": 1.0744,
-      "num_input_tokens_seen": 19029464,
-      "step": 1182
-    },
-    {
-      "epoch": 0.08286707469769804,
-      "grad_norm": 3.4207711219787598,
-      "learning_rate": 9.17210718038529e-05,
-      "loss": 0.8952,
-      "num_input_tokens_seen": 19045592,
-      "step": 1183
-    },
-    {
-      "epoch": 0.08293712294342728,
-      "grad_norm": 4.3305864334106445,
-      "learning_rate": 9.171407355516639e-05,
-      "loss": 0.9617,
-      "num_input_tokens_seen": 19061864,
-      "step": 1184
-    },
-    {
-      "epoch": 0.08300717118915653,
-      "grad_norm": 5.365218162536621,
-      "learning_rate": 9.170707530647986e-05,
-      "loss": 1.1669,
-      "num_input_tokens_seen": 19075448,
-      "step": 1185
-    },
-    {
-      "epoch": 0.08307721943488577,
-      "grad_norm": 3.9939708709716797,
-      "learning_rate": 9.170007705779335e-05,
-      "loss": 1.1325,
-      "num_input_tokens_seen": 19091832,
-      "step": 1186
-    },
-    {
-      "epoch": 0.08314726768061502,
-      "grad_norm": 3.8088884353637695,
-      "learning_rate": 9.169307880910683e-05,
-      "loss": 1.0132,
-      "num_input_tokens_seen": 19107920,
-      "step": 1187
-    },
-    {
-      "epoch": 0.08321731592634427,
-      "grad_norm": 3.858799457550049,
-      "learning_rate": 9.168608056042032e-05,
-      "loss": 0.9805,
-      "num_input_tokens_seen": 19123776,
-      "step": 1188
-    },
-    {
-      "epoch": 0.08328736417207351,
-      "grad_norm": 4.042770862579346,
-      "learning_rate": 9.16790823117338e-05,
-      "loss": 1.1668,
-      "num_input_tokens_seen": 19139752,
-      "step": 1189
-    },
-    {
-      "epoch": 0.08335741241780276,
-      "grad_norm": 4.2054762840271,
-      "learning_rate": 9.16720840630473e-05,
-      "loss": 1.0702,
-      "num_input_tokens_seen": 19156136,
-      "step": 1190
-    },
-    {
-      "epoch": 0.083427460663532,
-      "grad_norm": 4.450238227844238,
-      "learning_rate": 9.166508581436078e-05,
-      "loss": 1.0751,
-      "num_input_tokens_seen": 19172240,
-      "step": 1191
-    },
-    {
-      "epoch": 0.08349750890926125,
-      "grad_norm": 4.126129627227783,
-      "learning_rate": 9.165808756567426e-05,
-      "loss": 0.9957,
-      "num_input_tokens_seen": 19188624,
-      "step": 1192
-    },
-    {
-      "epoch": 0.0835675571549905,
-      "grad_norm": 4.131893157958984,
-      "learning_rate": 9.165108931698775e-05,
-      "loss": 1.2004,
-      "num_input_tokens_seen": 19205008,
-      "step": 1193
-    },
-    {
-      "epoch": 0.08363760540071974,
-      "grad_norm": 4.25187873840332,
-      "learning_rate": 9.164409106830123e-05,
-      "loss": 1.3571,
-      "num_input_tokens_seen": 19220856,
-      "step": 1194
-    },
-    {
-      "epoch": 0.083707653646449,
-      "grad_norm": 3.842498302459717,
-      "learning_rate": 9.163709281961471e-05,
-      "loss": 1.0963,
-      "num_input_tokens_seen": 19237208,
-      "step": 1195
-    },
-    {
-      "epoch": 0.08377770189217824,
-      "grad_norm": 3.694279432296753,
-      "learning_rate": 9.16300945709282e-05,
-      "loss": 1.1177,
-      "num_input_tokens_seen": 19253592,
-      "step": 1196
-    },
-    {
-      "epoch": 0.08384775013790749,
-      "grad_norm": 4.382254123687744,
-      "learning_rate": 9.162309632224169e-05,
-      "loss": 1.0344,
-      "num_input_tokens_seen": 19269976,
-      "step": 1197
-    },
-    {
-      "epoch": 0.08391779838363674,
-      "grad_norm": 4.267289161682129,
-      "learning_rate": 9.161609807355518e-05,
-      "loss": 1.1211,
-      "num_input_tokens_seen": 19286360,
-      "step": 1198
-    },
-    {
-      "epoch": 0.08398784662936598,
-      "grad_norm": 5.554534435272217,
-      "learning_rate": 9.160909982486865e-05,
-      "loss": 0.9674,
-      "num_input_tokens_seen": 19301800,
-      "step": 1199
-    },
-    {
-      "epoch": 0.08405789487509523,
-      "grad_norm": 4.1479668617248535,
-      "learning_rate": 9.160210157618214e-05,
-      "loss": 1.2334,
-      "num_input_tokens_seen": 19317392,
-      "step": 1200
-    },
-    {
-      "epoch": 0.08405789487509523,
-      "eval_loss": 1.1600490808486938,
-      "eval_runtime": 0.2015,
-      "eval_samples_per_second": 4.962,
-      "eval_steps_per_second": 4.962,
-      "num_input_tokens_seen": 19317392,
-      "step": 1200
-    },
-    {
-      "epoch": 0.08412794312082447,
-      "grad_norm": 4.1876349449157715,
-      "learning_rate": 9.159510332749563e-05,
-      "loss": 1.2036,
-      "num_input_tokens_seen": 19333776,
-      "step": 1201
-    },
-    {
-      "epoch": 0.08419799136655372,
-      "grad_norm": 4.031203746795654,
-      "learning_rate": 9.15881050788091e-05,
-      "loss": 1.2127,
-      "num_input_tokens_seen": 19349616,
-      "step": 1202
-    },
-    {
-      "epoch": 0.08426803961228296,
-      "grad_norm": 4.013350963592529,
-      "learning_rate": 9.15811068301226e-05,
-      "loss": 1.2147,
-      "num_input_tokens_seen": 19366000,
-      "step": 1203
-    },
-    {
-      "epoch": 0.08433808785801221,
-      "grad_norm": 4.509790897369385,
-      "learning_rate": 9.157410858143608e-05,
-      "loss": 1.3484,
-      "num_input_tokens_seen": 19381904,
-      "step": 1204
-    },
-    {
-      "epoch": 0.08440813610374145,
-      "grad_norm": 4.630336761474609,
-      "learning_rate": 9.156711033274957e-05,
-      "loss": 1.0246,
-      "num_input_tokens_seen": 19398288,
-      "step": 1205
-    },
-    {
-      "epoch": 0.0844781843494707,
-      "grad_norm": 3.819884777069092,
-      "learning_rate": 9.156011208406304e-05,
-      "loss": 1.1242,
-      "num_input_tokens_seen": 19414248,
-      "step": 1206
-    },
-    {
-      "epoch": 0.08454823259519995,
-      "grad_norm": 3.7933132648468018,
-      "learning_rate": 9.155311383537653e-05,
-      "loss": 1.0766,
-      "num_input_tokens_seen": 19430632,
-      "step": 1207
-    },
-    {
-      "epoch": 0.08461828084092919,
-      "grad_norm": 5.7384934425354,
-      "learning_rate": 9.154611558669002e-05,
-      "loss": 1.0691,
-      "num_input_tokens_seen": 19446248,
-      "step": 1208
-    },
-    {
-      "epoch": 0.08468832908665844,
-      "grad_norm": 3.9594175815582275,
-      "learning_rate": 9.153911733800351e-05,
-      "loss": 1.2029,
-      "num_input_tokens_seen": 19462632,
-      "step": 1209
-    },
-    {
-      "epoch": 0.08475837733238768,
-      "grad_norm": 3.8251891136169434,
-      "learning_rate": 9.1532119089317e-05,
-      "loss": 0.9994,
-      "num_input_tokens_seen": 19479016,
-      "step": 1210
-    },
-    {
-      "epoch": 0.08482842557811693,
-      "grad_norm": 3.9750332832336426,
-      "learning_rate": 9.152512084063049e-05,
-      "loss": 1.1737,
-      "num_input_tokens_seen": 19495112,
-      "step": 1211
-    },
-    {
-      "epoch": 0.08489847382384617,
-      "grad_norm": 3.986170530319214,
-      "learning_rate": 9.151812259194396e-05,
-      "loss": 1.1441,
-      "num_input_tokens_seen": 19511216,
-      "step": 1212
-    },
-    {
-      "epoch": 0.08496852206957542,
-      "grad_norm": 3.914065361022949,
-      "learning_rate": 9.151112434325745e-05,
-      "loss": 1.2233,
-      "num_input_tokens_seen": 19527600,
-      "step": 1213
-    },
-    {
-      "epoch": 0.08503857031530467,
-      "grad_norm": 4.328094482421875,
-      "learning_rate": 9.150412609457093e-05,
-      "loss": 1.2076,
-      "num_input_tokens_seen": 19543984,
-      "step": 1214
-    },
-    {
-      "epoch": 0.08510861856103391,
-      "grad_norm": 4.112467288970947,
-      "learning_rate": 9.149712784588441e-05,
-      "loss": 1.1732,
-      "num_input_tokens_seen": 19560368,
-      "step": 1215
-    },
-    {
-      "epoch": 0.08517866680676316,
-      "grad_norm": 4.680009365081787,
-      "learning_rate": 9.14901295971979e-05,
-      "loss": 0.985,
-      "num_input_tokens_seen": 19575616,
-      "step": 1216
-    },
-    {
-      "epoch": 0.0852487150524924,
-      "grad_norm": 4.4872660636901855,
-      "learning_rate": 9.148313134851139e-05,
-      "loss": 1.1799,
-      "num_input_tokens_seen": 19592000,
-      "step": 1217
-    },
-    {
-      "epoch": 0.08531876329822165,
-      "grad_norm": 3.7546637058258057,
-      "learning_rate": 9.147613309982488e-05,
-      "loss": 1.1989,
-      "num_input_tokens_seen": 19608384,
-      "step": 1218
-    },
-    {
-      "epoch": 0.0853888115439509,
-      "grad_norm": 5.590888500213623,
-      "learning_rate": 9.146913485113835e-05,
-      "loss": 1.1411,
-      "num_input_tokens_seen": 19624768,
-      "step": 1219
-    },
-    {
-      "epoch": 0.08545885978968014,
-      "grad_norm": 3.958021640777588,
-      "learning_rate": 9.146213660245184e-05,
-      "loss": 0.9309,
-      "num_input_tokens_seen": 19641152,
-      "step": 1220
-    },
-    {
-      "epoch": 0.08552890803540938,
-      "grad_norm": 3.7641196250915527,
-      "learning_rate": 9.145513835376533e-05,
-      "loss": 1.0299,
-      "num_input_tokens_seen": 19657536,
-      "step": 1221
-    },
-    {
-      "epoch": 0.08559895628113863,
-      "grad_norm": 4.395461559295654,
-      "learning_rate": 9.14481401050788e-05,
-      "loss": 1.1404,
-      "num_input_tokens_seen": 19673712,
-      "step": 1222
-    },
-    {
-      "epoch": 0.08566900452686788,
-      "grad_norm": 3.8162319660186768,
-      "learning_rate": 9.144114185639231e-05,
-      "loss": 1.1638,
-      "num_input_tokens_seen": 19689336,
-      "step": 1223
-    },
-    {
-      "epoch": 0.08573905277259712,
-      "grad_norm": 3.7025444507598877,
-      "learning_rate": 9.143414360770578e-05,
-      "loss": 0.9995,
-      "num_input_tokens_seen": 19705464,
-      "step": 1224
-    },
-    {
-      "epoch": 0.08580910101832637,
-      "grad_norm": 3.8621439933776855,
-      "learning_rate": 9.142714535901927e-05,
-      "loss": 1.1639,
-      "num_input_tokens_seen": 19721848,
-      "step": 1225
-    },
-    {
-      "epoch": 0.08587914926405561,
-      "grad_norm": 4.243250846862793,
-      "learning_rate": 9.142014711033275e-05,
-      "loss": 1.0104,
-      "num_input_tokens_seen": 19738072,
-      "step": 1226
-    },
-    {
-      "epoch": 0.08594919750978486,
-      "grad_norm": 4.05800724029541,
-      "learning_rate": 9.141314886164624e-05,
-      "loss": 1.0257,
-      "num_input_tokens_seen": 19754456,
-      "step": 1227
-    },
-    {
-      "epoch": 0.0860192457555141,
-      "grad_norm": 4.0894455909729,
-      "learning_rate": 9.140615061295972e-05,
-      "loss": 1.254,
-      "num_input_tokens_seen": 19770840,
-      "step": 1228
-    },
-    {
-      "epoch": 0.08608929400124336,
-      "grad_norm": 4.296894073486328,
-      "learning_rate": 9.139915236427321e-05,
-      "loss": 1.1298,
-      "num_input_tokens_seen": 19786864,
-      "step": 1229
-    },
-    {
-      "epoch": 0.08615934224697261,
-      "grad_norm": 4.0352888107299805,
-      "learning_rate": 9.13921541155867e-05,
-      "loss": 1.0611,
-      "num_input_tokens_seen": 19801800,
-      "step": 1230
-    },
-    {
-      "epoch": 0.08622939049270185,
-      "grad_norm": 4.087375640869141,
-      "learning_rate": 9.138515586690018e-05,
-      "loss": 0.9686,
-      "num_input_tokens_seen": 19818184,
-      "step": 1231
-    },
-    {
-      "epoch": 0.0862994387384311,
-      "grad_norm": 4.045078754425049,
-      "learning_rate": 9.137815761821367e-05,
-      "loss": 1.0915,
-      "num_input_tokens_seen": 19833016,
-      "step": 1232
-    },
-    {
-      "epoch": 0.08636948698416035,
-      "grad_norm": 4.399363040924072,
-      "learning_rate": 9.137115936952714e-05,
-      "loss": 1.1875,
-      "num_input_tokens_seen": 19848912,
-      "step": 1233
-    },
-    {
-      "epoch": 0.08643953522988959,
-      "grad_norm": 4.420406818389893,
-      "learning_rate": 9.136416112084063e-05,
-      "loss": 1.0534,
-      "num_input_tokens_seen": 19865296,
-      "step": 1234
-    },
-    {
-      "epoch": 0.08650958347561884,
-      "grad_norm": 4.131808280944824,
-      "learning_rate": 9.135716287215412e-05,
-      "loss": 1.1865,
-      "num_input_tokens_seen": 19881376,
-      "step": 1235
-    },
-    {
-      "epoch": 0.08657963172134808,
-      "grad_norm": 3.8256850242614746,
-      "learning_rate": 9.13501646234676e-05,
-      "loss": 1.0539,
-      "num_input_tokens_seen": 19897704,
-      "step": 1236
-    },
-    {
-      "epoch": 0.08664967996707733,
-      "grad_norm": 4.3497233390808105,
-      "learning_rate": 9.13431663747811e-05,
-      "loss": 1.191,
-      "num_input_tokens_seen": 19914088,
-      "step": 1237
-    },
-    {
-      "epoch": 0.08671972821280657,
-      "grad_norm": 4.18136739730835,
-      "learning_rate": 9.133616812609458e-05,
-      "loss": 1.0539,
-      "num_input_tokens_seen": 19930128,
-      "step": 1238
-    },
-    {
-      "epoch": 0.08678977645853582,
-      "grad_norm": 4.782970905303955,
-      "learning_rate": 9.132916987740806e-05,
-      "loss": 1.1992,
-      "num_input_tokens_seen": 19946512,
-      "step": 1239
-    },
-    {
-      "epoch": 0.08685982470426507,
-      "grad_norm": 4.16589879989624,
-      "learning_rate": 9.132217162872155e-05,
-      "loss": 1.1463,
-      "num_input_tokens_seen": 19962488,
-      "step": 1240
-    },
-    {
-      "epoch": 0.08692987294999431,
-      "grad_norm": 3.73541522026062,
-      "learning_rate": 9.131517338003502e-05,
-      "loss": 1.0272,
-      "num_input_tokens_seen": 19978584,
-      "step": 1241
-    },
-    {
-      "epoch": 0.08699992119572356,
-      "grad_norm": 4.225815773010254,
-      "learning_rate": 9.130817513134851e-05,
-      "loss": 1.177,
-      "num_input_tokens_seen": 19994816,
-      "step": 1242
-    },
-    {
-      "epoch": 0.0870699694414528,
-      "grad_norm": 7.807470321655273,
-      "learning_rate": 9.1301176882662e-05,
-      "loss": 1.1635,
-      "num_input_tokens_seen": 20010576,
-      "step": 1243
-    },
-    {
-      "epoch": 0.08714001768718205,
-      "grad_norm": 4.818174839019775,
-      "learning_rate": 9.129417863397549e-05,
-      "loss": 1.1892,
-      "num_input_tokens_seen": 20025712,
-      "step": 1244
-    },
-    {
-      "epoch": 0.0872100659329113,
-      "grad_norm": 3.8367979526519775,
-      "learning_rate": 9.128718038528898e-05,
-      "loss": 1.0096,
-      "num_input_tokens_seen": 20041904,
-      "step": 1245
-    },
-    {
-      "epoch": 0.08728011417864054,
-      "grad_norm": 3.9912586212158203,
-      "learning_rate": 9.128018213660245e-05,
-      "loss": 1.097,
-      "num_input_tokens_seen": 20058288,
-      "step": 1246
-    },
-    {
-      "epoch": 0.08735016242436978,
-      "grad_norm": 4.842557907104492,
-      "learning_rate": 9.127318388791594e-05,
-      "loss": 1.2012,
-      "num_input_tokens_seen": 20074672,
-      "step": 1247
-    },
-    {
-      "epoch": 0.08742021067009903,
-      "grad_norm": 3.816938877105713,
-      "learning_rate": 9.126618563922943e-05,
-      "loss": 1.1683,
-      "num_input_tokens_seen": 20090664,
-      "step": 1248
-    },
-    {
-      "epoch": 0.08749025891582828,
-      "grad_norm": 3.712480306625366,
-      "learning_rate": 9.125918739054292e-05,
-      "loss": 1.1978,
-      "num_input_tokens_seen": 20107048,
-      "step": 1249
-    },
-    {
-      "epoch": 0.08756030716155752,
-      "grad_norm": 4.185492515563965,
-      "learning_rate": 9.12521891418564e-05,
-      "loss": 1.2042,
-      "num_input_tokens_seen": 20123432,
-      "step": 1250
-    },
-    {
-      "epoch": 0.08763035540728677,
-      "grad_norm": 5.510714530944824,
-      "learning_rate": 9.124519089316988e-05,
-      "loss": 0.9757,
-      "num_input_tokens_seen": 20139112,
-      "step": 1251
-    },
-    {
-      "epoch": 0.08770040365301601,
-      "grad_norm": 3.9170289039611816,
-      "learning_rate": 9.123819264448337e-05,
-      "loss": 1.0213,
-      "num_input_tokens_seen": 20155496,
-      "step": 1252
-    },
-    {
-      "epoch": 0.08777045189874526,
-      "grad_norm": 3.738008975982666,
-      "learning_rate": 9.123119439579684e-05,
-      "loss": 0.9446,
-      "num_input_tokens_seen": 20171760,
-      "step": 1253
-    },
-    {
-      "epoch": 0.0878405001444745,
-      "grad_norm": 4.845873832702637,
-      "learning_rate": 9.122419614711033e-05,
-      "loss": 1.2135,
-      "num_input_tokens_seen": 20188056,
-      "step": 1254
-    },
-    {
-      "epoch": 0.08791054839020375,
-      "grad_norm": 4.166906356811523,
-      "learning_rate": 9.121719789842382e-05,
-      "loss": 1.1558,
-      "num_input_tokens_seen": 20204440,
-      "step": 1255
-    },
-    {
-      "epoch": 0.087980596635933,
-      "grad_norm": 4.039194107055664,
-      "learning_rate": 9.121019964973731e-05,
-      "loss": 1.0297,
-      "num_input_tokens_seen": 20220824,
-      "step": 1256
-    },
-    {
-      "epoch": 0.08805064488166224,
-      "grad_norm": 3.545482635498047,
-      "learning_rate": 9.12032014010508e-05,
-      "loss": 0.9757,
-      "num_input_tokens_seen": 20236888,
-      "step": 1257
-    },
-    {
-      "epoch": 0.08812069312739149,
-      "grad_norm": 3.82114839553833,
-      "learning_rate": 9.119620315236427e-05,
-      "loss": 1.1637,
-      "num_input_tokens_seen": 20253272,
-      "step": 1258
-    },
-    {
-      "epoch": 0.08819074137312073,
-      "grad_norm": 4.770678997039795,
-      "learning_rate": 9.118920490367776e-05,
-      "loss": 1.1421,
-      "num_input_tokens_seen": 20269656,
-      "step": 1259
-    },
-    {
-      "epoch": 0.08826078961884998,
-      "grad_norm": 4.4319539070129395,
-      "learning_rate": 9.118220665499124e-05,
-      "loss": 1.1565,
-      "num_input_tokens_seen": 20285456,
-      "step": 1260
-    },
-    {
-      "epoch": 0.08833083786457922,
-      "grad_norm": 4.0923357009887695,
-      "learning_rate": 9.117520840630473e-05,
-      "loss": 1.2328,
-      "num_input_tokens_seen": 20301232,
-      "step": 1261
-    },
-    {
-      "epoch": 0.08840088611030847,
-      "grad_norm": 5.8347344398498535,
-      "learning_rate": 9.116821015761821e-05,
-      "loss": 0.8824,
-      "num_input_tokens_seen": 20317224,
-      "step": 1262
-    },
-    {
-      "epoch": 0.08847093435603771,
-      "grad_norm": 4.525367259979248,
-      "learning_rate": 9.11612119089317e-05,
-      "loss": 1.1554,
-      "num_input_tokens_seen": 20332616,
-      "step": 1263
-    },
-    {
-      "epoch": 0.08854098260176697,
-      "grad_norm": 3.9754436016082764,
-      "learning_rate": 9.115421366024519e-05,
-      "loss": 1.0423,
-      "num_input_tokens_seen": 20348336,
-      "step": 1264
-    },
-    {
-      "epoch": 0.08861103084749622,
-      "grad_norm": 4.40745735168457,
-      "learning_rate": 9.114721541155868e-05,
-      "loss": 1.0485,
-      "num_input_tokens_seen": 20364312,
-      "step": 1265
-    },
-    {
-      "epoch": 0.08868107909322547,
-      "grad_norm": 7.126221179962158,
-      "learning_rate": 9.114021716287216e-05,
-      "loss": 1.2035,
-      "num_input_tokens_seen": 20380696,
-      "step": 1266
-    },
-    {
-      "epoch": 0.08875112733895471,
-      "grad_norm": 4.306386947631836,
-      "learning_rate": 9.113321891418564e-05,
-      "loss": 1.0399,
-      "num_input_tokens_seen": 20397080,
-      "step": 1267
-    },
-    {
-      "epoch": 0.08882117558468396,
-      "grad_norm": 3.566943407058716,
-      "learning_rate": 9.112622066549912e-05,
-      "loss": 1.0463,
-      "num_input_tokens_seen": 20413464,
-      "step": 1268
-    },
-    {
-      "epoch": 0.0888912238304132,
-      "grad_norm": 3.975228786468506,
-      "learning_rate": 9.111922241681262e-05,
-      "loss": 1.2576,
-      "num_input_tokens_seen": 20429848,
-      "step": 1269
-    },
-    {
-      "epoch": 0.08896127207614245,
-      "grad_norm": 4.928854465484619,
-      "learning_rate": 9.11122241681261e-05,
-      "loss": 1.1555,
-      "num_input_tokens_seen": 20446192,
-      "step": 1270
-    },
-    {
-      "epoch": 0.0890313203218717,
-      "grad_norm": 4.288821697235107,
-      "learning_rate": 9.110522591943958e-05,
-      "loss": 1.2559,
-      "num_input_tokens_seen": 20462576,
-      "step": 1271
-    },
-    {
-      "epoch": 0.08910136856760094,
-      "grad_norm": 3.9346396923065186,
-      "learning_rate": 9.109822767075307e-05,
-      "loss": 1.1479,
-      "num_input_tokens_seen": 20478520,
-      "step": 1272
-    },
-    {
-      "epoch": 0.08917141681333018,
-      "grad_norm": 3.7976620197296143,
-      "learning_rate": 9.109122942206655e-05,
-      "loss": 0.9903,
-      "num_input_tokens_seen": 20494408,
-      "step": 1273
-    },
-    {
-      "epoch": 0.08924146505905943,
-      "grad_norm": 5.373577117919922,
-      "learning_rate": 9.108423117338004e-05,
-      "loss": 0.8863,
-      "num_input_tokens_seen": 20510792,
-      "step": 1274
-    },
-    {
-      "epoch": 0.08931151330478868,
-      "grad_norm": 4.248324394226074,
-      "learning_rate": 9.107723292469353e-05,
-      "loss": 1.3492,
-      "num_input_tokens_seen": 20527064,
-      "step": 1275
-    },
-    {
-      "epoch": 0.08938156155051792,
-      "grad_norm": 4.453672885894775,
-      "learning_rate": 9.107023467600701e-05,
-      "loss": 0.9763,
-      "num_input_tokens_seen": 20543448,
-      "step": 1276
-    },
-    {
-      "epoch": 0.08945160979624717,
-      "grad_norm": 4.8721184730529785,
-      "learning_rate": 9.10632364273205e-05,
-      "loss": 0.9455,
-      "num_input_tokens_seen": 20559832,
-      "step": 1277
-    },
-    {
-      "epoch": 0.08952165804197641,
-      "grad_norm": 5.0173540115356445,
-      "learning_rate": 9.105623817863398e-05,
-      "loss": 1.0303,
-      "num_input_tokens_seen": 20576216,
-      "step": 1278
-    },
-    {
-      "epoch": 0.08959170628770566,
-      "grad_norm": 5.00100040435791,
-      "learning_rate": 9.104923992994747e-05,
-      "loss": 1.0393,
-      "num_input_tokens_seen": 20592600,
-      "step": 1279
-    },
-    {
-      "epoch": 0.0896617545334349,
-      "grad_norm": 4.271099090576172,
-      "learning_rate": 9.104224168126094e-05,
-      "loss": 1.2307,
-      "num_input_tokens_seen": 20608632,
-      "step": 1280
-    },
-    {
-      "epoch": 0.08973180277916415,
-      "grad_norm": 4.246976852416992,
-      "learning_rate": 9.103524343257443e-05,
-      "loss": 1.1405,
-      "num_input_tokens_seen": 20625016,
-      "step": 1281
-    },
-    {
-      "epoch": 0.0898018510248934,
-      "grad_norm": 5.033923149108887,
-      "learning_rate": 9.102824518388792e-05,
-      "loss": 1.0849,
-      "num_input_tokens_seen": 20641400,
-      "step": 1282
-    },
-    {
-      "epoch": 0.08987189927062264,
-      "grad_norm": 4.4118571281433105,
-      "learning_rate": 9.102124693520141e-05,
-      "loss": 1.118,
-      "num_input_tokens_seen": 20657448,
-      "step": 1283
-    },
-    {
-      "epoch": 0.08994194751635189,
-      "grad_norm": 4.150144577026367,
-      "learning_rate": 9.10142486865149e-05,
-      "loss": 1.0676,
-      "num_input_tokens_seen": 20673080,
-      "step": 1284
-    },
-    {
-      "epoch": 0.09001199576208113,
-      "grad_norm": 3.767683744430542,
-      "learning_rate": 9.100725043782837e-05,
-      "loss": 0.8968,
-      "num_input_tokens_seen": 20689464,
-      "step": 1285
-    },
-    {
-      "epoch": 0.09008204400781038,
-      "grad_norm": 4.816582202911377,
-      "learning_rate": 9.100025218914186e-05,
-      "loss": 1.0039,
-      "num_input_tokens_seen": 20703896,
-      "step": 1286
-    },
-    {
-      "epoch": 0.09015209225353962,
-      "grad_norm": 3.8913414478302,
-      "learning_rate": 9.099325394045533e-05,
-      "loss": 1.0077,
-      "num_input_tokens_seen": 20720280,
-      "step": 1287
-    },
-    {
-      "epoch": 0.09022214049926887,
-      "grad_norm": 4.305298328399658,
-      "learning_rate": 9.098625569176882e-05,
-      "loss": 1.1555,
-      "num_input_tokens_seen": 20735944,
-      "step": 1288
-    },
-    {
-      "epoch": 0.09029218874499811,
-      "grad_norm": 3.3120992183685303,
-      "learning_rate": 9.097925744308233e-05,
-      "loss": 0.8591,
-      "num_input_tokens_seen": 20752128,
-      "step": 1289
-    },
-    {
-      "epoch": 0.09036223699072736,
-      "grad_norm": 4.705013751983643,
-      "learning_rate": 9.09722591943958e-05,
-      "loss": 1.4579,
-      "num_input_tokens_seen": 20768512,
-      "step": 1290
-    },
-    {
-      "epoch": 0.0904322852364566,
-      "grad_norm": 5.08630895614624,
-      "learning_rate": 9.096526094570929e-05,
-      "loss": 1.1049,
-      "num_input_tokens_seen": 20783976,
-      "step": 1291
-    },
-    {
-      "epoch": 0.09050233348218585,
-      "grad_norm": 3.634686231613159,
-      "learning_rate": 9.095826269702278e-05,
-      "loss": 1.0344,
-      "num_input_tokens_seen": 20800360,
-      "step": 1292
-    },
-    {
-      "epoch": 0.0905723817279151,
-      "grad_norm": 4.220744609832764,
-      "learning_rate": 9.095126444833625e-05,
-      "loss": 1.1843,
-      "num_input_tokens_seen": 20816744,
-      "step": 1293
-    },
-    {
-      "epoch": 0.09064242997364434,
-      "grad_norm": 4.724472522735596,
-      "learning_rate": 9.094426619964974e-05,
-      "loss": 1.1365,
-      "num_input_tokens_seen": 20833128,
-      "step": 1294
-    },
-    {
-      "epoch": 0.09071247821937359,
-      "grad_norm": 3.9398090839385986,
-      "learning_rate": 9.093726795096323e-05,
-      "loss": 1.0703,
-      "num_input_tokens_seen": 20849448,
-      "step": 1295
-    },
-    {
-      "epoch": 0.09078252646510283,
-      "grad_norm": 4.260062217712402,
-      "learning_rate": 9.093026970227672e-05,
-      "loss": 1.0968,
-      "num_input_tokens_seen": 20865832,
-      "step": 1296
-    },
-    {
-      "epoch": 0.09085257471083208,
-      "grad_norm": 4.383310317993164,
-      "learning_rate": 9.09232714535902e-05,
-      "loss": 1.2542,
-      "num_input_tokens_seen": 20881288,
-      "step": 1297
-    },
-    {
-      "epoch": 0.09092262295656132,
-      "grad_norm": 4.479433059692383,
-      "learning_rate": 9.091627320490368e-05,
-      "loss": 0.9533,
-      "num_input_tokens_seen": 20897328,
-      "step": 1298
-    },
-    {
-      "epoch": 0.09099267120229058,
-      "grad_norm": 4.911858081817627,
-      "learning_rate": 9.090927495621717e-05,
-      "loss": 1.3399,
-      "num_input_tokens_seen": 20913712,
-      "step": 1299
-    },
-    {
-      "epoch": 0.09106271944801983,
-      "grad_norm": 4.015485763549805,
-      "learning_rate": 9.090227670753065e-05,
-      "loss": 1.1156,
-      "num_input_tokens_seen": 20929984,
-      "step": 1300
-    },
-    {
-      "epoch": 0.09113276769374908,
-      "grad_norm": 3.8690338134765625,
-      "learning_rate": 9.089527845884413e-05,
-      "loss": 1.0634,
-      "num_input_tokens_seen": 20946368,
-      "step": 1301
-    },
-    {
-      "epoch": 0.09120281593947832,
-      "grad_norm": 5.142012596130371,
-      "learning_rate": 9.088828021015762e-05,
-      "loss": 1.0579,
-      "num_input_tokens_seen": 20962752,
-      "step": 1302
-    },
-    {
-      "epoch": 0.09127286418520757,
-      "grad_norm": 3.954049587249756,
-      "learning_rate": 9.088128196147111e-05,
-      "loss": 1.0862,
-      "num_input_tokens_seen": 20979136,
-      "step": 1303
-    },
-    {
-      "epoch": 0.09134291243093681,
-      "grad_norm": 4.13312292098999,
-      "learning_rate": 9.08742837127846e-05,
-      "loss": 1.0548,
-      "num_input_tokens_seen": 20995520,
-      "step": 1304
-    },
-    {
-      "epoch": 0.09141296067666606,
-      "grad_norm": 4.24699592590332,
-      "learning_rate": 9.086728546409808e-05,
-      "loss": 1.0126,
-      "num_input_tokens_seen": 21011904,
-      "step": 1305
-    },
-    {
-      "epoch": 0.0914830089223953,
-      "grad_norm": 4.847048759460449,
-      "learning_rate": 9.086028721541156e-05,
-      "loss": 0.9973,
-      "num_input_tokens_seen": 21027664,
-      "step": 1306
-    },
-    {
-      "epoch": 0.09155305716812455,
-      "grad_norm": 4.573661804199219,
-      "learning_rate": 9.085328896672504e-05,
-      "loss": 1.005,
-      "num_input_tokens_seen": 21044048,
-      "step": 1307
-    },
-    {
-      "epoch": 0.0916231054138538,
-      "grad_norm": 4.13530158996582,
-      "learning_rate": 9.084629071803853e-05,
-      "loss": 1.1033,
-      "num_input_tokens_seen": 21060432,
-      "step": 1308
-    },
-    {
-      "epoch": 0.09169315365958304,
-      "grad_norm": 4.017937183380127,
-      "learning_rate": 9.083929246935203e-05,
-      "loss": 1.1971,
-      "num_input_tokens_seen": 21076816,
-      "step": 1309
-    },
-    {
-      "epoch": 0.09176320190531229,
-      "grad_norm": 5.928586483001709,
-      "learning_rate": 9.08322942206655e-05,
-      "loss": 1.0547,
-      "num_input_tokens_seen": 21093200,
-      "step": 1310
-    },
-    {
-      "epoch": 0.09183325015104153,
-      "grad_norm": 4.2442169189453125,
-      "learning_rate": 9.082529597197899e-05,
-      "loss": 1.2794,
-      "num_input_tokens_seen": 21109256,
-      "step": 1311
-    },
-    {
-      "epoch": 0.09190329839677078,
-      "grad_norm": 4.891444683074951,
-      "learning_rate": 9.081829772329247e-05,
-      "loss": 1.1833,
-      "num_input_tokens_seen": 21124848,
-      "step": 1312
-    },
-    {
-      "epoch": 0.09197334664250002,
-      "grad_norm": 4.323850154876709,
-      "learning_rate": 9.081129947460596e-05,
-      "loss": 1.1683,
-      "num_input_tokens_seen": 21141176,
-      "step": 1313
-    },
-    {
-      "epoch": 0.09204339488822927,
-      "grad_norm": 4.239765644073486,
-      "learning_rate": 9.080430122591943e-05,
-      "loss": 1.1073,
-      "num_input_tokens_seen": 21157240,
-      "step": 1314
-    },
-    {
-      "epoch": 0.09211344313395851,
-      "grad_norm": 4.12881326675415,
-      "learning_rate": 9.079730297723293e-05,
-      "loss": 1.2522,
-      "num_input_tokens_seen": 21173624,
-      "step": 1315
-    },
-    {
-      "epoch": 0.09218349137968776,
-      "grad_norm": 4.238161087036133,
-      "learning_rate": 9.079030472854642e-05,
-      "loss": 1.1828,
-      "num_input_tokens_seen": 21190008,
-      "step": 1316
-    },
-    {
-      "epoch": 0.092253539625417,
-      "grad_norm": 4.124176502227783,
-      "learning_rate": 9.07833064798599e-05,
-      "loss": 1.1388,
-      "num_input_tokens_seen": 21206392,
-      "step": 1317
-    },
-    {
-      "epoch": 0.09232358787114625,
-      "grad_norm": 3.772136926651001,
-      "learning_rate": 9.077630823117339e-05,
-      "loss": 1.068,
-      "num_input_tokens_seen": 21222776,
-      "step": 1318
-    },
-    {
-      "epoch": 0.0923936361168755,
-      "grad_norm": 4.628321170806885,
-      "learning_rate": 9.076930998248687e-05,
-      "loss": 1.2363,
-      "num_input_tokens_seen": 21239160,
-      "step": 1319
-    },
-    {
-      "epoch": 0.09246368436260474,
-      "grad_norm": 5.3034348487854,
-      "learning_rate": 9.076231173380035e-05,
-      "loss": 1.0638,
-      "num_input_tokens_seen": 21255544,
-      "step": 1320
-    },
-    {
-      "epoch": 0.09253373260833399,
-      "grad_norm": 3.6543760299682617,
-      "learning_rate": 9.075531348511384e-05,
-      "loss": 1.0071,
-      "num_input_tokens_seen": 21271928,
-      "step": 1321
-    },
-    {
-      "epoch": 0.09260378085406323,
-      "grad_norm": 4.1335062980651855,
-      "learning_rate": 9.074831523642733e-05,
-      "loss": 1.084,
-      "num_input_tokens_seen": 21288312,
-      "step": 1322
-    },
-    {
-      "epoch": 0.09267382909979248,
-      "grad_norm": 3.6392204761505127,
-      "learning_rate": 9.074131698774082e-05,
-      "loss": 1.1146,
-      "num_input_tokens_seen": 21304696,
-      "step": 1323
-    },
-    {
-      "epoch": 0.09274387734552172,
-      "grad_norm": 4.035269737243652,
-      "learning_rate": 9.073431873905429e-05,
-      "loss": 0.9578,
-      "num_input_tokens_seen": 21321080,
-      "step": 1324
-    },
-    {
-      "epoch": 0.09281392559125097,
-      "grad_norm": 4.650269508361816,
-      "learning_rate": 9.072732049036778e-05,
-      "loss": 1.0242,
-      "num_input_tokens_seen": 21337464,
-      "step": 1325
-    },
-    {
-      "epoch": 0.09288397383698022,
-      "grad_norm": 5.850543022155762,
-      "learning_rate": 9.072032224168127e-05,
-      "loss": 1.1196,
-      "num_input_tokens_seen": 21352968,
-      "step": 1326
-    },
-    {
-      "epoch": 0.09295402208270946,
-      "grad_norm": 4.177901744842529,
-      "learning_rate": 9.071332399299474e-05,
-      "loss": 1.1351,
-      "num_input_tokens_seen": 21368968,
-      "step": 1327
-    },
-    {
-      "epoch": 0.09302407032843871,
-      "grad_norm": 4.582173824310303,
-      "learning_rate": 9.070632574430823e-05,
-      "loss": 0.9115,
-      "num_input_tokens_seen": 21385352,
-      "step": 1328
-    },
-    {
-      "epoch": 0.09309411857416795,
-      "grad_norm": 4.7911787033081055,
-      "learning_rate": 9.069932749562173e-05,
-      "loss": 1.0413,
-      "num_input_tokens_seen": 21401144,
-      "step": 1329
-    },
-    {
-      "epoch": 0.0931641668198972,
-      "grad_norm": 4.058457374572754,
-      "learning_rate": 9.069232924693521e-05,
-      "loss": 1.0611,
-      "num_input_tokens_seen": 21416640,
-      "step": 1330
-    },
-    {
-      "epoch": 0.09323421506562644,
-      "grad_norm": 4.972208499908447,
-      "learning_rate": 9.06853309982487e-05,
-      "loss": 1.016,
-      "num_input_tokens_seen": 21433024,
-      "step": 1331
-    },
-    {
-      "epoch": 0.09330426331135569,
-      "grad_norm": 4.0875091552734375,
-      "learning_rate": 9.067833274956217e-05,
-      "loss": 1.089,
-      "num_input_tokens_seen": 21448888,
-      "step": 1332
-    },
-    {
-      "epoch": 0.09337431155708495,
-      "grad_norm": 3.923112154006958,
-      "learning_rate": 9.067133450087566e-05,
-      "loss": 0.9824,
-      "num_input_tokens_seen": 21465272,
-      "step": 1333
-    },
-    {
-      "epoch": 0.0934443598028142,
-      "grad_norm": 4.067697525024414,
-      "learning_rate": 9.066433625218914e-05,
-      "loss": 1.0492,
-      "num_input_tokens_seen": 21481656,
-      "step": 1334
-    },
-    {
-      "epoch": 0.09351440804854344,
-      "grad_norm": 4.185417652130127,
-      "learning_rate": 9.065733800350264e-05,
-      "loss": 1.1073,
-      "num_input_tokens_seen": 21498040,
-      "step": 1335
-    },
-    {
-      "epoch": 0.09358445629427269,
-      "grad_norm": 7.31542444229126,
-      "learning_rate": 9.065033975481613e-05,
-      "loss": 1.4322,
-      "num_input_tokens_seen": 21514088,
-      "step": 1336
-    },
-    {
-      "epoch": 0.09365450454000193,
-      "grad_norm": 4.754745006561279,
-      "learning_rate": 9.06433415061296e-05,
-      "loss": 0.9953,
-      "num_input_tokens_seen": 21530472,
-      "step": 1337
-    },
-    {
-      "epoch": 0.09372455278573118,
-      "grad_norm": 5.81265926361084,
-      "learning_rate": 9.063634325744309e-05,
-      "loss": 1.1434,
-      "num_input_tokens_seen": 21545728,
-      "step": 1338
-    },
-    {
-      "epoch": 0.09379460103146042,
-      "grad_norm": 5.586238861083984,
-      "learning_rate": 9.062934500875657e-05,
-      "loss": 0.9818,
-      "num_input_tokens_seen": 21562112,
-      "step": 1339
-    },
-    {
-      "epoch": 0.09386464927718967,
-      "grad_norm": 4.096534729003906,
-      "learning_rate": 9.062234676007005e-05,
-      "loss": 1.1856,
-      "num_input_tokens_seen": 21578496,
-      "step": 1340
-    },
-    {
-      "epoch": 0.09393469752291891,
-      "grad_norm": 4.913814544677734,
-      "learning_rate": 9.061534851138354e-05,
-      "loss": 1.041,
-      "num_input_tokens_seen": 21594792,
-      "step": 1341
-    },
-    {
-      "epoch": 0.09400474576864816,
-      "grad_norm": 3.8853912353515625,
-      "learning_rate": 9.060835026269703e-05,
-      "loss": 1.1651,
-      "num_input_tokens_seen": 21611176,
-      "step": 1342
-    },
-    {
-      "epoch": 0.0940747940143774,
-      "grad_norm": 4.187959671020508,
-      "learning_rate": 9.060135201401052e-05,
-      "loss": 1.1757,
-      "num_input_tokens_seen": 21627560,
-      "step": 1343
-    },
-    {
-      "epoch": 0.09414484226010665,
-      "grad_norm": 4.128627777099609,
-      "learning_rate": 9.0594353765324e-05,
-      "loss": 0.9243,
-      "num_input_tokens_seen": 21643576,
-      "step": 1344
-    },
-    {
-      "epoch": 0.0942148905058359,
-      "grad_norm": 4.7016825675964355,
-      "learning_rate": 9.058735551663748e-05,
-      "loss": 1.2425,
-      "num_input_tokens_seen": 21658600,
-      "step": 1345
-    },
-    {
-      "epoch": 0.09428493875156514,
-      "grad_norm": 3.970548391342163,
-      "learning_rate": 9.058035726795097e-05,
-      "loss": 1.0495,
-      "num_input_tokens_seen": 21674264,
-      "step": 1346
-    },
-    {
-      "epoch": 0.09435498699729439,
-      "grad_norm": 3.812196731567383,
-      "learning_rate": 9.057335901926445e-05,
-      "loss": 0.9558,
-      "num_input_tokens_seen": 21690112,
-      "step": 1347
-    },
-    {
-      "epoch": 0.09442503524302363,
-      "grad_norm": 3.6845176219940186,
-      "learning_rate": 9.056636077057794e-05,
-      "loss": 0.9758,
-      "num_input_tokens_seen": 21705744,
-      "step": 1348
-    },
-    {
-      "epoch": 0.09449508348875288,
-      "grad_norm": 4.119202136993408,
-      "learning_rate": 9.055936252189142e-05,
-      "loss": 1.0948,
-      "num_input_tokens_seen": 21721776,
-      "step": 1349
-    },
-    {
-      "epoch": 0.09456513173448212,
-      "grad_norm": 4.176985740661621,
-      "learning_rate": 9.055236427320491e-05,
-      "loss": 0.9475,
-      "num_input_tokens_seen": 21737912,
-      "step": 1350
-    },
-    {
-      "epoch": 0.09463517998021137,
-      "grad_norm": 4.057264804840088,
-      "learning_rate": 9.054536602451839e-05,
-      "loss": 1.1746,
-      "num_input_tokens_seen": 21754296,
-      "step": 1351
-    },
-    {
-      "epoch": 0.09470522822594062,
-      "grad_norm": 4.5631914138793945,
-      "learning_rate": 9.053836777583188e-05,
-      "loss": 1.0894,
-      "num_input_tokens_seen": 21770680,
-      "step": 1352
-    },
-    {
-      "epoch": 0.09477527647166986,
-      "grad_norm": 4.854849815368652,
-      "learning_rate": 9.053136952714536e-05,
-      "loss": 1.0686,
-      "num_input_tokens_seen": 21787064,
-      "step": 1353
-    },
-    {
-      "epoch": 0.09484532471739911,
-      "grad_norm": 5.326946258544922,
-      "learning_rate": 9.052437127845884e-05,
-      "loss": 0.872,
-      "num_input_tokens_seen": 21803448,
-      "step": 1354
-    },
-    {
-      "epoch": 0.09491537296312835,
-      "grad_norm": 4.283742904663086,
-      "learning_rate": 9.051737302977234e-05,
-      "loss": 1.2683,
-      "num_input_tokens_seen": 21819832,
-      "step": 1355
-    },
-    {
-      "epoch": 0.0949854212088576,
-      "grad_norm": 4.165935039520264,
-      "learning_rate": 9.051037478108582e-05,
-      "loss": 0.977,
-      "num_input_tokens_seen": 21836216,
-      "step": 1356
-    },
-    {
-      "epoch": 0.09505546945458684,
-      "grad_norm": 4.502480983734131,
-      "learning_rate": 9.05033765323993e-05,
-      "loss": 1.2854,
-      "num_input_tokens_seen": 21852600,
-      "step": 1357
-    },
-    {
-      "epoch": 0.09512551770031609,
-      "grad_norm": 4.185445308685303,
-      "learning_rate": 9.04963782837128e-05,
-      "loss": 1.2225,
-      "num_input_tokens_seen": 21868984,
-      "step": 1358
-    },
-    {
-      "epoch": 0.09519556594604534,
-      "grad_norm": 7.288909435272217,
-      "learning_rate": 9.048938003502627e-05,
-      "loss": 1.154,
-      "num_input_tokens_seen": 21884648,
-      "step": 1359
-    },
-    {
-      "epoch": 0.09526561419177458,
-      "grad_norm": 4.038896560668945,
-      "learning_rate": 9.048238178633976e-05,
-      "loss": 1.1437,
-      "num_input_tokens_seen": 21900704,
-      "step": 1360
-    },
-    {
-      "epoch": 0.09533566243750383,
-      "grad_norm": 4.216241836547852,
-      "learning_rate": 9.047538353765325e-05,
-      "loss": 1.1379,
-      "num_input_tokens_seen": 21916520,
-      "step": 1361
-    },
-    {
-      "epoch": 0.09540571068323307,
-      "grad_norm": 4.2549147605896,
-      "learning_rate": 9.046838528896673e-05,
-      "loss": 1.2578,
-      "num_input_tokens_seen": 21932904,
-      "step": 1362
-    },
-    {
-      "epoch": 0.09547575892896232,
-      "grad_norm": 3.6919445991516113,
-      "learning_rate": 9.046138704028022e-05,
-      "loss": 0.9876,
-      "num_input_tokens_seen": 21949288,
-      "step": 1363
-    },
-    {
-      "epoch": 0.09554580717469156,
-      "grad_norm": 5.467876434326172,
-      "learning_rate": 9.04543887915937e-05,
-      "loss": 0.9735,
-      "num_input_tokens_seen": 21965672,
-      "step": 1364
-    },
-    {
-      "epoch": 0.09561585542042081,
-      "grad_norm": 4.036736011505127,
-      "learning_rate": 9.044739054290719e-05,
-      "loss": 1.0712,
-      "num_input_tokens_seen": 21980792,
-      "step": 1365
-    },
-    {
-      "epoch": 0.09568590366615005,
-      "grad_norm": 4.083346843719482,
-      "learning_rate": 9.044039229422066e-05,
-      "loss": 1.0883,
-      "num_input_tokens_seen": 21996888,
-      "step": 1366
-    },
-    {
-      "epoch": 0.0957559519118793,
-      "grad_norm": 3.553262948989868,
-      "learning_rate": 9.043339404553415e-05,
-      "loss": 1.0116,
-      "num_input_tokens_seen": 22013160,
-      "step": 1367
-    },
-    {
-      "epoch": 0.09582600015760856,
-      "grad_norm": 4.787721633911133,
-      "learning_rate": 9.042639579684764e-05,
-      "loss": 1.1444,
-      "num_input_tokens_seen": 22029544,
-      "step": 1368
-    },
-    {
-      "epoch": 0.0958960484033378,
-      "grad_norm": 3.8053700923919678,
-      "learning_rate": 9.041939754816113e-05,
-      "loss": 1.1654,
-      "num_input_tokens_seen": 22045888,
-      "step": 1369
-    },
-    {
-      "epoch": 0.09596609664906705,
-      "grad_norm": 3.7679660320281982,
-      "learning_rate": 9.041239929947462e-05,
-      "loss": 1.1753,
-      "num_input_tokens_seen": 22062272,
-      "step": 1370
-    },
-    {
-      "epoch": 0.0960361448947963,
-      "grad_norm": 5.086554527282715,
-      "learning_rate": 9.040540105078809e-05,
-      "loss": 0.9579,
-      "num_input_tokens_seen": 22078080,
-      "step": 1371
-    },
-    {
-      "epoch": 0.09610619314052554,
-      "grad_norm": 4.255527496337891,
-      "learning_rate": 9.039840280210158e-05,
-      "loss": 1.0953,
-      "num_input_tokens_seen": 22093808,
-      "step": 1372
-    },
-    {
-      "epoch": 0.09617624138625479,
-      "grad_norm": 6.081700325012207,
-      "learning_rate": 9.039140455341507e-05,
-      "loss": 1.0363,
-      "num_input_tokens_seen": 22110192,
-      "step": 1373
-    },
-    {
-      "epoch": 0.09624628963198403,
-      "grad_norm": 4.376565456390381,
-      "learning_rate": 9.038440630472854e-05,
-      "loss": 1.1737,
-      "num_input_tokens_seen": 22126576,
-      "step": 1374
-    },
-    {
-      "epoch": 0.09631633787771328,
-      "grad_norm": 4.051114559173584,
-      "learning_rate": 9.037740805604205e-05,
-      "loss": 1.1921,
-      "num_input_tokens_seen": 22142768,
-      "step": 1375
-    },
-    {
-      "epoch": 0.09638638612344252,
-      "grad_norm": 4.46164083480835,
-      "learning_rate": 9.037040980735552e-05,
-      "loss": 1.1541,
-      "num_input_tokens_seen": 22158600,
-      "step": 1376
-    },
-    {
-      "epoch": 0.09645643436917177,
-      "grad_norm": 4.242503643035889,
-      "learning_rate": 9.036341155866901e-05,
-      "loss": 1.1314,
-      "num_input_tokens_seen": 22174984,
-      "step": 1377
-    },
-    {
-      "epoch": 0.09652648261490102,
-      "grad_norm": 3.6338908672332764,
-      "learning_rate": 9.035641330998248e-05,
-      "loss": 0.9257,
-      "num_input_tokens_seen": 22190880,
-      "step": 1378
-    },
-    {
-      "epoch": 0.09659653086063026,
-      "grad_norm": 4.73402738571167,
-      "learning_rate": 9.034941506129597e-05,
-      "loss": 1.1981,
-      "num_input_tokens_seen": 22206632,
-      "step": 1379
-    },
-    {
-      "epoch": 0.09666657910635951,
-      "grad_norm": 4.450289726257324,
-      "learning_rate": 9.034241681260946e-05,
-      "loss": 1.0851,
-      "num_input_tokens_seen": 22222896,
-      "step": 1380
-    },
-    {
-      "epoch": 0.09673662735208875,
-      "grad_norm": 5.578179359436035,
-      "learning_rate": 9.033541856392295e-05,
-      "loss": 1.2856,
-      "num_input_tokens_seen": 22238280,
-      "step": 1381
-    },
-    {
-      "epoch": 0.096806675597818,
-      "grad_norm": 3.8745546340942383,
-      "learning_rate": 9.032842031523644e-05,
-      "loss": 0.9841,
-      "num_input_tokens_seen": 22254664,
-      "step": 1382
-    },
-    {
-      "epoch": 0.09687672384354724,
-      "grad_norm": 5.7268548011779785,
-      "learning_rate": 9.032142206654991e-05,
-      "loss": 1.2024,
-      "num_input_tokens_seen": 22270000,
-      "step": 1383
-    },
-    {
-      "epoch": 0.09694677208927649,
-      "grad_norm": 4.380898952484131,
-      "learning_rate": 9.03144238178634e-05,
-      "loss": 1.0589,
-      "num_input_tokens_seen": 22286384,
-      "step": 1384
-    },
-    {
-      "epoch": 0.09701682033500574,
-      "grad_norm": 5.762500762939453,
-      "learning_rate": 9.030742556917689e-05,
-      "loss": 1.2061,
-      "num_input_tokens_seen": 22302272,
-      "step": 1385
-    },
-    {
-      "epoch": 0.09708686858073498,
-      "grad_norm": 3.739488363265991,
-      "learning_rate": 9.030042732049037e-05,
-      "loss": 0.9867,
-      "num_input_tokens_seen": 22318656,
-      "step": 1386
-    },
-    {
-      "epoch": 0.09715691682646423,
-      "grad_norm": 4.584897994995117,
-      "learning_rate": 9.029342907180385e-05,
-      "loss": 1.1934,
-      "num_input_tokens_seen": 22334704,
-      "step": 1387
-    },
-    {
-      "epoch": 0.09722696507219347,
-      "grad_norm": 4.161139488220215,
-      "learning_rate": 9.028643082311734e-05,
-      "loss": 1.1638,
-      "num_input_tokens_seen": 22349800,
-      "step": 1388
-    },
-    {
-      "epoch": 0.09729701331792272,
-      "grad_norm": 4.115293979644775,
-      "learning_rate": 9.027943257443083e-05,
-      "loss": 1.0181,
-      "num_input_tokens_seen": 22366184,
-      "step": 1389
-    },
-    {
-      "epoch": 0.09736706156365196,
-      "grad_norm": 3.7355988025665283,
-      "learning_rate": 9.027243432574432e-05,
-      "loss": 1.1182,
-      "num_input_tokens_seen": 22382568,
-      "step": 1390
-    },
-    {
-      "epoch": 0.09743710980938121,
-      "grad_norm": 4.15507173538208,
-      "learning_rate": 9.02654360770578e-05,
-      "loss": 1.0272,
-      "num_input_tokens_seen": 22398480,
-      "step": 1391
-    },
-    {
-      "epoch": 0.09750715805511045,
-      "grad_norm": 3.770918607711792,
-      "learning_rate": 9.025843782837128e-05,
-      "loss": 0.9834,
-      "num_input_tokens_seen": 22414864,
-      "step": 1392
-    },
-    {
-      "epoch": 0.0975772063008397,
-      "grad_norm": 4.214321136474609,
-      "learning_rate": 9.025143957968476e-05,
-      "loss": 1.1738,
-      "num_input_tokens_seen": 22429752,
-      "step": 1393
-    },
-    {
-      "epoch": 0.09764725454656895,
-      "grad_norm": 3.9854986667633057,
-      "learning_rate": 9.024444133099825e-05,
-      "loss": 1.2832,
-      "num_input_tokens_seen": 22446136,
-      "step": 1394
-    },
-    {
-      "epoch": 0.09771730279229819,
-      "grad_norm": 4.996057510375977,
-      "learning_rate": 9.023744308231174e-05,
-      "loss": 1.1691,
-      "num_input_tokens_seen": 22461160,
-      "step": 1395
-    },
-    {
-      "epoch": 0.09778735103802744,
-      "grad_norm": 3.682765007019043,
-      "learning_rate": 9.023044483362523e-05,
-      "loss": 0.9548,
-      "num_input_tokens_seen": 22477336,
-      "step": 1396
-    },
-    {
-      "epoch": 0.09785739928375668,
-      "grad_norm": 4.367272853851318,
-      "learning_rate": 9.022344658493871e-05,
-      "loss": 1.0512,
-      "num_input_tokens_seen": 22492952,
-      "step": 1397
-    },
-    {
-      "epoch": 0.09792744752948593,
-      "grad_norm": 3.9716336727142334,
-      "learning_rate": 9.021644833625219e-05,
-      "loss": 1.103,
-      "num_input_tokens_seen": 22509336,
-      "step": 1398
-    },
-    {
-      "epoch": 0.09799749577521517,
-      "grad_norm": 4.043631553649902,
-      "learning_rate": 9.020945008756568e-05,
-      "loss": 1.1439,
-      "num_input_tokens_seen": 22525568,
-      "step": 1399
-    },
-    {
-      "epoch": 0.09806754402094442,
-      "grad_norm": 4.343166351318359,
-      "learning_rate": 9.020245183887917e-05,
-      "loss": 1.1948,
-      "num_input_tokens_seen": 22541328,
-      "step": 1400
-    },
-    {
-      "epoch": 0.09806754402094442,
-      "eval_loss": 1.1561514139175415,
-      "eval_runtime": 0.1977,
-      "eval_samples_per_second": 5.058,
-      "eval_steps_per_second": 5.058,
-      "num_input_tokens_seen": 22541328,
-      "step": 1400
-    },
-    {
-      "epoch": 0.09813759226667366,
-      "grad_norm": 4.709417819976807,
-      "learning_rate": 9.019545359019265e-05,
-      "loss": 1.1398,
-      "num_input_tokens_seen": 22557304,
-      "step": 1401
-    },
-    {
-      "epoch": 0.09820764051240291,
-      "grad_norm": 7.022638320922852,
-      "learning_rate": 9.018845534150614e-05,
-      "loss": 1.0342,
-      "num_input_tokens_seen": 22573688,
-      "step": 1402
-    },
-    {
-      "epoch": 0.09827768875813217,
-      "grad_norm": 3.7976694107055664,
-      "learning_rate": 9.018145709281962e-05,
-      "loss": 0.9829,
-      "num_input_tokens_seen": 22589848,
-      "step": 1403
-    },
-    {
-      "epoch": 0.09834773700386142,
-      "grad_norm": 3.70877742767334,
-      "learning_rate": 9.01744588441331e-05,
-      "loss": 0.9707,
-      "num_input_tokens_seen": 22606232,
-      "step": 1404
-    },
-    {
-      "epoch": 0.09841778524959066,
-      "grad_norm": 7.724960803985596,
-      "learning_rate": 9.016746059544658e-05,
-      "loss": 0.9602,
-      "num_input_tokens_seen": 22621912,
-      "step": 1405
-    },
-    {
-      "epoch": 0.09848783349531991,
-      "grad_norm": 3.9619522094726562,
-      "learning_rate": 9.016046234676007e-05,
-      "loss": 0.998,
-      "num_input_tokens_seen": 22638296,
-      "step": 1406
-    },
-    {
-      "epoch": 0.09855788174104915,
-      "grad_norm": 3.8303041458129883,
-      "learning_rate": 9.015346409807356e-05,
-      "loss": 1.0762,
-      "num_input_tokens_seen": 22654496,
-      "step": 1407
-    },
-    {
-      "epoch": 0.0986279299867784,
-      "grad_norm": 4.029507637023926,
-      "learning_rate": 9.014646584938705e-05,
-      "loss": 1.2072,
-      "num_input_tokens_seen": 22670544,
-      "step": 1408
-    },
-    {
-      "epoch": 0.09869797823250764,
-      "grad_norm": 3.8487346172332764,
-      "learning_rate": 9.013946760070054e-05,
-      "loss": 1.1834,
-      "num_input_tokens_seen": 22686592,
-      "step": 1409
-    },
-    {
-      "epoch": 0.09876802647823689,
-      "grad_norm": 3.700751543045044,
-      "learning_rate": 9.013246935201401e-05,
-      "loss": 0.8698,
-      "num_input_tokens_seen": 22702976,
-      "step": 1410
-    },
-    {
-      "epoch": 0.09883807472396614,
-      "grad_norm": 3.686884641647339,
-      "learning_rate": 9.01254711033275e-05,
-      "loss": 0.9591,
-      "num_input_tokens_seen": 22719360,
-      "step": 1411
-    },
-    {
-      "epoch": 0.09890812296969538,
-      "grad_norm": 4.176409721374512,
-      "learning_rate": 9.011847285464099e-05,
-      "loss": 1.1578,
-      "num_input_tokens_seen": 22735744,
-      "step": 1412
-    },
-    {
-      "epoch": 0.09897817121542463,
-      "grad_norm": 4.331852912902832,
-      "learning_rate": 9.011147460595446e-05,
-      "loss": 0.9769,
-      "num_input_tokens_seen": 22752128,
-      "step": 1413
-    },
-    {
-      "epoch": 0.09904821946115387,
-      "grad_norm": 3.8534255027770996,
-      "learning_rate": 9.010447635726795e-05,
-      "loss": 1.1536,
-      "num_input_tokens_seen": 22768512,
-      "step": 1414
-    },
-    {
-      "epoch": 0.09911826770688312,
-      "grad_norm": 4.066548824310303,
-      "learning_rate": 9.009747810858144e-05,
-      "loss": 1.1199,
-      "num_input_tokens_seen": 22784760,
-      "step": 1415
-    },
-    {
-      "epoch": 0.09918831595261236,
-      "grad_norm": 4.076517581939697,
-      "learning_rate": 9.009047985989493e-05,
-      "loss": 1.1132,
-      "num_input_tokens_seen": 22801144,
-      "step": 1416
-    },
-    {
-      "epoch": 0.09925836419834161,
-      "grad_norm": 3.8858346939086914,
-      "learning_rate": 9.008348161120842e-05,
-      "loss": 0.9509,
-      "num_input_tokens_seen": 22817320,
-      "step": 1417
-    },
-    {
-      "epoch": 0.09932841244407085,
-      "grad_norm": 6.4605584144592285,
-      "learning_rate": 9.007648336252189e-05,
-      "loss": 1.2701,
-      "num_input_tokens_seen": 22833704,
-      "step": 1418
-    },
-    {
-      "epoch": 0.0993984606898001,
-      "grad_norm": 4.157481670379639,
-      "learning_rate": 9.006948511383538e-05,
-      "loss": 1.0169,
-      "num_input_tokens_seen": 22850088,
-      "step": 1419
-    },
-    {
-      "epoch": 0.09946850893552935,
-      "grad_norm": 3.725755214691162,
-      "learning_rate": 9.006248686514886e-05,
-      "loss": 1.0183,
-      "num_input_tokens_seen": 22866472,
-      "step": 1420
-    },
-    {
-      "epoch": 0.09953855718125859,
-      "grad_norm": 4.012838363647461,
-      "learning_rate": 9.005548861646236e-05,
-      "loss": 0.8425,
-      "num_input_tokens_seen": 22882856,
-      "step": 1421
-    },
-    {
-      "epoch": 0.09960860542698784,
-      "grad_norm": 3.8754239082336426,
-      "learning_rate": 9.004849036777583e-05,
-      "loss": 1.1375,
-      "num_input_tokens_seen": 22899240,
-      "step": 1422
-    },
-    {
-      "epoch": 0.09967865367271708,
-      "grad_norm": 3.90873384475708,
-      "learning_rate": 9.004149211908932e-05,
-      "loss": 1.0574,
-      "num_input_tokens_seen": 22915160,
-      "step": 1423
-    },
-    {
-      "epoch": 0.09974870191844633,
-      "grad_norm": 5.698948860168457,
-      "learning_rate": 9.003449387040281e-05,
-      "loss": 1.1338,
-      "num_input_tokens_seen": 22930592,
-      "step": 1424
-    },
-    {
-      "epoch": 0.09981875016417557,
-      "grad_norm": 4.103662014007568,
-      "learning_rate": 9.002749562171629e-05,
-      "loss": 1.2384,
-      "num_input_tokens_seen": 22946976,
-      "step": 1425
-    },
-    {
-      "epoch": 0.09988879840990482,
-      "grad_norm": 4.404048442840576,
-      "learning_rate": 9.002049737302977e-05,
-      "loss": 1.3855,
-      "num_input_tokens_seen": 22963360,
-      "step": 1426
-    },
-    {
-      "epoch": 0.09995884665563406,
-      "grad_norm": 4.043710708618164,
-      "learning_rate": 9.001349912434326e-05,
-      "loss": 1.2713,
-      "num_input_tokens_seen": 22979544,
-      "step": 1427
-    },
-    {
-      "epoch": 0.10002889490136331,
-      "grad_norm": 4.169802188873291,
-      "learning_rate": 9.000650087565675e-05,
-      "loss": 1.0777,
-      "num_input_tokens_seen": 22995072,
-      "step": 1428
-    },
-    {
-      "epoch": 0.10009894314709256,
-      "grad_norm": 4.010350227355957,
-      "learning_rate": 8.999950262697024e-05,
-      "loss": 1.1245,
-      "num_input_tokens_seen": 23010904,
-      "step": 1429
-    },
-    {
-      "epoch": 0.1001689913928218,
-      "grad_norm": 4.496591567993164,
-      "learning_rate": 8.999250437828372e-05,
-      "loss": 1.3372,
-      "num_input_tokens_seen": 23027288,
-      "step": 1430
-    },
-    {
-      "epoch": 0.10023903963855105,
-      "grad_norm": 4.2428765296936035,
-      "learning_rate": 8.99855061295972e-05,
-      "loss": 1.0258,
-      "num_input_tokens_seen": 23043352,
-      "step": 1431
-    },
-    {
-      "epoch": 0.10030908788428029,
-      "grad_norm": 4.083342552185059,
-      "learning_rate": 8.997850788091068e-05,
-      "loss": 1.227,
-      "num_input_tokens_seen": 23059736,
-      "step": 1432
-    },
-    {
-      "epoch": 0.10037913613000954,
-      "grad_norm": 3.860734462738037,
-      "learning_rate": 8.997150963222417e-05,
-      "loss": 1.0791,
-      "num_input_tokens_seen": 23075400,
-      "step": 1433
-    },
-    {
-      "epoch": 0.10044918437573878,
-      "grad_norm": 3.985151767730713,
-      "learning_rate": 8.996451138353766e-05,
-      "loss": 1.0486,
-      "num_input_tokens_seen": 23091704,
-      "step": 1434
-    },
-    {
-      "epoch": 0.10051923262146803,
-      "grad_norm": 4.039731502532959,
-      "learning_rate": 8.995751313485114e-05,
-      "loss": 0.9793,
-      "num_input_tokens_seen": 23108088,
-      "step": 1435
-    },
-    {
-      "epoch": 0.10058928086719728,
-      "grad_norm": 6.1780619621276855,
-      "learning_rate": 8.995051488616463e-05,
-      "loss": 1.0645,
-      "num_input_tokens_seen": 23123128,
-      "step": 1436
-    },
-    {
-      "epoch": 0.10065932911292653,
-      "grad_norm": 4.5783586502075195,
-      "learning_rate": 8.994351663747811e-05,
-      "loss": 1.1634,
-      "num_input_tokens_seen": 23139168,
-      "step": 1437
-    },
-    {
-      "epoch": 0.10072937735865578,
-      "grad_norm": 3.889927864074707,
-      "learning_rate": 8.99365183887916e-05,
-      "loss": 0.97,
-      "num_input_tokens_seen": 23154952,
-      "step": 1438
-    },
-    {
-      "epoch": 0.10079942560438503,
-      "grad_norm": 3.927945852279663,
-      "learning_rate": 8.992952014010509e-05,
-      "loss": 1.2428,
-      "num_input_tokens_seen": 23170688,
-      "step": 1439
-    },
-    {
-      "epoch": 0.10086947385011427,
-      "grad_norm": 3.8991434574127197,
-      "learning_rate": 8.992252189141856e-05,
-      "loss": 0.9519,
-      "num_input_tokens_seen": 23186432,
-      "step": 1440
-    },
-    {
-      "epoch": 0.10093952209584352,
-      "grad_norm": 3.6479310989379883,
-      "learning_rate": 8.991552364273206e-05,
-      "loss": 0.9656,
-      "num_input_tokens_seen": 23202816,
-      "step": 1441
-    },
-    {
-      "epoch": 0.10100957034157276,
-      "grad_norm": 4.637960910797119,
-      "learning_rate": 8.990852539404554e-05,
-      "loss": 1.2853,
-      "num_input_tokens_seen": 23218304,
-      "step": 1442
-    },
-    {
-      "epoch": 0.10107961858730201,
-      "grad_norm": 4.000091552734375,
-      "learning_rate": 8.990152714535903e-05,
-      "loss": 1.0421,
-      "num_input_tokens_seen": 23234688,
-      "step": 1443
-    },
-    {
-      "epoch": 0.10114966683303125,
-      "grad_norm": 4.959738731384277,
-      "learning_rate": 8.989452889667251e-05,
-      "loss": 1.0904,
-      "num_input_tokens_seen": 23250656,
-      "step": 1444
-    },
-    {
-      "epoch": 0.1012197150787605,
-      "grad_norm": 3.9251675605773926,
-      "learning_rate": 8.988753064798599e-05,
-      "loss": 0.9219,
-      "num_input_tokens_seen": 23266984,
-      "step": 1445
-    },
-    {
-      "epoch": 0.10128976332448975,
-      "grad_norm": 4.28665828704834,
-      "learning_rate": 8.988053239929948e-05,
-      "loss": 1.1465,
-      "num_input_tokens_seen": 23283368,
-      "step": 1446
-    },
-    {
-      "epoch": 0.10135981157021899,
-      "grad_norm": 4.421731472015381,
-      "learning_rate": 8.987353415061297e-05,
-      "loss": 1.1098,
-      "num_input_tokens_seen": 23298728,
-      "step": 1447
-    },
-    {
-      "epoch": 0.10142985981594824,
-      "grad_norm": 5.080065727233887,
-      "learning_rate": 8.986653590192646e-05,
-      "loss": 1.1172,
-      "num_input_tokens_seen": 23315112,
-      "step": 1448
-    },
-    {
-      "epoch": 0.10149990806167748,
-      "grad_norm": 5.618803977966309,
-      "learning_rate": 8.985953765323993e-05,
-      "loss": 0.9551,
-      "num_input_tokens_seen": 23329864,
-      "step": 1449
-    },
-    {
-      "epoch": 0.10156995630740673,
-      "grad_norm": 3.756836414337158,
-      "learning_rate": 8.985253940455342e-05,
-      "loss": 1.0981,
-      "num_input_tokens_seen": 23345672,
-      "step": 1450
-    },
-    {
-      "epoch": 0.10164000455313597,
-      "grad_norm": 4.461424827575684,
-      "learning_rate": 8.984554115586691e-05,
-      "loss": 1.1914,
-      "num_input_tokens_seen": 23362056,
-      "step": 1451
-    },
-    {
-      "epoch": 0.10171005279886522,
-      "grad_norm": 5.267919063568115,
-      "learning_rate": 8.983854290718038e-05,
-      "loss": 1.1928,
-      "num_input_tokens_seen": 23378440,
-      "step": 1452
-    },
-    {
-      "epoch": 0.10178010104459446,
-      "grad_norm": 5.513551235198975,
-      "learning_rate": 8.983154465849387e-05,
-      "loss": 1.2405,
-      "num_input_tokens_seen": 23394824,
-      "step": 1453
-    },
-    {
-      "epoch": 0.10185014929032371,
-      "grad_norm": 4.46366548538208,
-      "learning_rate": 8.982454640980736e-05,
-      "loss": 1.1436,
-      "num_input_tokens_seen": 23410568,
-      "step": 1454
-    },
-    {
-      "epoch": 0.10192019753605296,
-      "grad_norm": 5.066692352294922,
-      "learning_rate": 8.981754816112085e-05,
-      "loss": 1.1389,
-      "num_input_tokens_seen": 23426952,
-      "step": 1455
-    },
-    {
-      "epoch": 0.1019902457817822,
-      "grad_norm": 3.980743169784546,
-      "learning_rate": 8.981054991243434e-05,
-      "loss": 1.0623,
-      "num_input_tokens_seen": 23443256,
-      "step": 1456
-    },
-    {
-      "epoch": 0.10206029402751145,
-      "grad_norm": 4.088611125946045,
-      "learning_rate": 8.980355166374781e-05,
-      "loss": 1.0388,
-      "num_input_tokens_seen": 23459640,
-      "step": 1457
-    },
-    {
-      "epoch": 0.10213034227324069,
-      "grad_norm": 3.9585626125335693,
-      "learning_rate": 8.97965534150613e-05,
-      "loss": 1.2051,
-      "num_input_tokens_seen": 23475176,
-      "step": 1458
-    },
-    {
-      "epoch": 0.10220039051896994,
-      "grad_norm": 3.7923290729522705,
-      "learning_rate": 8.978955516637478e-05,
-      "loss": 1.0001,
-      "num_input_tokens_seen": 23490704,
-      "step": 1459
-    },
-    {
-      "epoch": 0.10227043876469918,
-      "grad_norm": 3.9089629650115967,
-      "learning_rate": 8.978255691768826e-05,
-      "loss": 0.9786,
-      "num_input_tokens_seen": 23506168,
-      "step": 1460
-    },
-    {
-      "epoch": 0.10234048701042843,
-      "grad_norm": 6.2259039878845215,
-      "learning_rate": 8.977555866900175e-05,
-      "loss": 1.2854,
-      "num_input_tokens_seen": 23522552,
-      "step": 1461
-    },
-    {
-      "epoch": 0.10241053525615768,
-      "grad_norm": 4.071867942810059,
-      "learning_rate": 8.976856042031524e-05,
-      "loss": 1.0724,
-      "num_input_tokens_seen": 23538936,
-      "step": 1462
-    },
-    {
-      "epoch": 0.10248058350188692,
-      "grad_norm": 4.587897777557373,
-      "learning_rate": 8.976156217162873e-05,
-      "loss": 1.1307,
-      "num_input_tokens_seen": 23554536,
-      "step": 1463
-    },
-    {
-      "epoch": 0.10255063174761617,
-      "grad_norm": 3.944937229156494,
-      "learning_rate": 8.97545639229422e-05,
-      "loss": 1.1503,
-      "num_input_tokens_seen": 23570888,
-      "step": 1464
-    },
-    {
-      "epoch": 0.10262067999334541,
-      "grad_norm": 3.7418766021728516,
-      "learning_rate": 8.97475656742557e-05,
-      "loss": 1.0414,
-      "num_input_tokens_seen": 23587272,
-      "step": 1465
-    },
-    {
-      "epoch": 0.10269072823907466,
-      "grad_norm": 3.9055676460266113,
-      "learning_rate": 8.974056742556918e-05,
-      "loss": 1.2284,
-      "num_input_tokens_seen": 23603640,
-      "step": 1466
-    },
-    {
-      "epoch": 0.1027607764848039,
-      "grad_norm": 3.9338066577911377,
-      "learning_rate": 8.973356917688267e-05,
-      "loss": 1.2389,
-      "num_input_tokens_seen": 23620024,
-      "step": 1467
-    },
-    {
-      "epoch": 0.10283082473053315,
-      "grad_norm": 4.024602890014648,
-      "learning_rate": 8.972657092819616e-05,
-      "loss": 1.1358,
-      "num_input_tokens_seen": 23636408,
-      "step": 1468
-    },
-    {
-      "epoch": 0.1029008729762624,
-      "grad_norm": 4.09812068939209,
-      "learning_rate": 8.971957267950963e-05,
-      "loss": 1.0734,
-      "num_input_tokens_seen": 23652480,
-      "step": 1469
-    },
-    {
-      "epoch": 0.10297092122199164,
-      "grad_norm": 4.7382025718688965,
-      "learning_rate": 8.971257443082312e-05,
-      "loss": 1.4112,
-      "num_input_tokens_seen": 23668424,
-      "step": 1470
-    },
-    {
-      "epoch": 0.10304096946772089,
-      "grad_norm": 4.518669605255127,
-      "learning_rate": 8.970557618213661e-05,
-      "loss": 1.3466,
-      "num_input_tokens_seen": 23684808,
-      "step": 1471
-    },
-    {
-      "epoch": 0.10311101771345015,
-      "grad_norm": 4.023036003112793,
-      "learning_rate": 8.969857793345009e-05,
-      "loss": 1.0246,
-      "num_input_tokens_seen": 23701192,
-      "step": 1472
-    },
-    {
-      "epoch": 0.10318106595917939,
-      "grad_norm": 4.6244215965271,
-      "learning_rate": 8.969157968476358e-05,
-      "loss": 1.2391,
-      "num_input_tokens_seen": 23717576,
-      "step": 1473
-    },
-    {
-      "epoch": 0.10325111420490864,
-      "grad_norm": 4.517683506011963,
-      "learning_rate": 8.968458143607706e-05,
-      "loss": 1.3872,
-      "num_input_tokens_seen": 23733960,
-      "step": 1474
-    },
-    {
-      "epoch": 0.10332116245063788,
-      "grad_norm": 4.048764705657959,
-      "learning_rate": 8.967758318739055e-05,
-      "loss": 1.0453,
-      "num_input_tokens_seen": 23750344,
-      "step": 1475
-    },
-    {
-      "epoch": 0.10339121069636713,
-      "grad_norm": 4.248376369476318,
-      "learning_rate": 8.967058493870403e-05,
-      "loss": 1.176,
-      "num_input_tokens_seen": 23766160,
-      "step": 1476
-    },
-    {
-      "epoch": 0.10346125894209637,
-      "grad_norm": 3.780548095703125,
-      "learning_rate": 8.966358669001752e-05,
-      "loss": 0.9048,
-      "num_input_tokens_seen": 23782544,
-      "step": 1477
-    },
-    {
-      "epoch": 0.10353130718782562,
-      "grad_norm": 4.26375675201416,
-      "learning_rate": 8.9656588441331e-05,
-      "loss": 0.8651,
-      "num_input_tokens_seen": 23798928,
-      "step": 1478
-    },
-    {
-      "epoch": 0.10360135543355486,
-      "grad_norm": 3.9202687740325928,
-      "learning_rate": 8.964959019264448e-05,
-      "loss": 1.1058,
-      "num_input_tokens_seen": 23815312,
-      "step": 1479
-    },
-    {
-      "epoch": 0.10367140367928411,
-      "grad_norm": 3.983797788619995,
-      "learning_rate": 8.964259194395797e-05,
-      "loss": 1.0778,
-      "num_input_tokens_seen": 23831696,
-      "step": 1480
-    },
-    {
-      "epoch": 0.10374145192501336,
-      "grad_norm": 4.471195220947266,
-      "learning_rate": 8.963559369527146e-05,
-      "loss": 1.1858,
-      "num_input_tokens_seen": 23847768,
-      "step": 1481
-    },
-    {
-      "epoch": 0.1038115001707426,
-      "grad_norm": 3.560317039489746,
-      "learning_rate": 8.962859544658495e-05,
-      "loss": 1.0205,
-      "num_input_tokens_seen": 23864152,
-      "step": 1482
-    },
-    {
-      "epoch": 0.10388154841647185,
-      "grad_norm": 3.8699846267700195,
-      "learning_rate": 8.962159719789843e-05,
-      "loss": 1.1438,
-      "num_input_tokens_seen": 23880536,
-      "step": 1483
-    },
-    {
-      "epoch": 0.10395159666220109,
-      "grad_norm": 4.547862529754639,
-      "learning_rate": 8.961459894921191e-05,
-      "loss": 1.0303,
-      "num_input_tokens_seen": 23896704,
-      "step": 1484
-    },
-    {
-      "epoch": 0.10402164490793034,
-      "grad_norm": 4.669456481933594,
-      "learning_rate": 8.96076007005254e-05,
-      "loss": 1.1994,
-      "num_input_tokens_seen": 23913088,
-      "step": 1485
-    },
-    {
-      "epoch": 0.10409169315365958,
-      "grad_norm": 4.346285343170166,
-      "learning_rate": 8.960060245183887e-05,
-      "loss": 1.2677,
-      "num_input_tokens_seen": 23929472,
-      "step": 1486
-    },
-    {
-      "epoch": 0.10416174139938883,
-      "grad_norm": 6.5028581619262695,
-      "learning_rate": 8.959360420315236e-05,
-      "loss": 0.989,
-      "num_input_tokens_seen": 23945216,
-      "step": 1487
-    },
-    {
-      "epoch": 0.10423178964511808,
-      "grad_norm": 3.935488224029541,
-      "learning_rate": 8.958660595446586e-05,
-      "loss": 1.2657,
-      "num_input_tokens_seen": 23961600,
-      "step": 1488
-    },
-    {
-      "epoch": 0.10430183789084732,
-      "grad_norm": 3.772397518157959,
-      "learning_rate": 8.957960770577934e-05,
-      "loss": 1.1038,
-      "num_input_tokens_seen": 23977984,
-      "step": 1489
-    },
-    {
-      "epoch": 0.10437188613657657,
-      "grad_norm": 4.508286476135254,
-      "learning_rate": 8.957260945709283e-05,
-      "loss": 1.2694,
-      "num_input_tokens_seen": 23993752,
-      "step": 1490
-    },
-    {
-      "epoch": 0.10444193438230581,
-      "grad_norm": 4.667380332946777,
-      "learning_rate": 8.95656112084063e-05,
-      "loss": 1.2837,
-      "num_input_tokens_seen": 24009832,
-      "step": 1491
-    },
-    {
-      "epoch": 0.10451198262803506,
-      "grad_norm": 7.675503730773926,
-      "learning_rate": 8.955861295971979e-05,
-      "loss": 1.121,
-      "num_input_tokens_seen": 24025784,
-      "step": 1492
-    },
-    {
-      "epoch": 0.1045820308737643,
-      "grad_norm": 4.427548408508301,
-      "learning_rate": 8.955161471103328e-05,
-      "loss": 0.835,
-      "num_input_tokens_seen": 24041568,
-      "step": 1493
-    },
-    {
-      "epoch": 0.10465207911949355,
-      "grad_norm": 3.9065396785736084,
-      "learning_rate": 8.954461646234677e-05,
-      "loss": 1.1322,
-      "num_input_tokens_seen": 24057952,
-      "step": 1494
-    },
-    {
-      "epoch": 0.1047221273652228,
-      "grad_norm": 4.052605628967285,
-      "learning_rate": 8.953761821366026e-05,
-      "loss": 1.1133,
-      "num_input_tokens_seen": 24074336,
-      "step": 1495
-    },
-    {
-      "epoch": 0.10479217561095204,
-      "grad_norm": 3.758476734161377,
-      "learning_rate": 8.953061996497373e-05,
-      "loss": 1.1302,
-      "num_input_tokens_seen": 24090720,
-      "step": 1496
-    },
-    {
-      "epoch": 0.10486222385668129,
-      "grad_norm": 4.4470014572143555,
-      "learning_rate": 8.952362171628722e-05,
-      "loss": 1.0969,
-      "num_input_tokens_seen": 24107024,
-      "step": 1497
-    },
-    {
-      "epoch": 0.10493227210241053,
-      "grad_norm": 4.222001075744629,
-      "learning_rate": 8.951662346760071e-05,
-      "loss": 1.147,
-      "num_input_tokens_seen": 24123408,
-      "step": 1498
-    },
-    {
-      "epoch": 0.10500232034813978,
-      "grad_norm": 4.72997522354126,
-      "learning_rate": 8.950962521891418e-05,
-      "loss": 1.1086,
-      "num_input_tokens_seen": 24137672,
-      "step": 1499
-    },
-    {
-      "epoch": 0.10507236859386902,
-      "grad_norm": 4.342312812805176,
-      "learning_rate": 8.950262697022767e-05,
-      "loss": 1.2044,
-      "num_input_tokens_seen": 24153248,
-      "step": 1500
-    },
-    {
-      "epoch": 0.10514241683959827,
-      "grad_norm": 4.723706245422363,
-      "learning_rate": 8.949562872154116e-05,
-      "loss": 1.1075,
-      "num_input_tokens_seen": 24169240,
-      "step": 1501
-    },
-    {
-      "epoch": 0.10521246508532751,
-      "grad_norm": 4.244345188140869,
-      "learning_rate": 8.948863047285465e-05,
-      "loss": 1.1839,
-      "num_input_tokens_seen": 24184608,
-      "step": 1502
-    },
-    {
-      "epoch": 0.10528251333105676,
-      "grad_norm": 3.6271615028381348,
-      "learning_rate": 8.948163222416812e-05,
-      "loss": 1.0755,
-      "num_input_tokens_seen": 24200992,
-      "step": 1503
-    },
-    {
-      "epoch": 0.105352561576786,
-      "grad_norm": 3.858696937561035,
-      "learning_rate": 8.947463397548161e-05,
-      "loss": 1.0598,
-      "num_input_tokens_seen": 24217376,
-      "step": 1504
-    },
-    {
-      "epoch": 0.10542260982251525,
-      "grad_norm": 7.14077091217041,
-      "learning_rate": 8.94676357267951e-05,
-      "loss": 1.0362,
-      "num_input_tokens_seen": 24232368,
-      "step": 1505
-    },
-    {
-      "epoch": 0.10549265806824451,
-      "grad_norm": 4.203495979309082,
-      "learning_rate": 8.946063747810858e-05,
-      "loss": 1.2491,
-      "num_input_tokens_seen": 24248520,
-      "step": 1506
-    },
-    {
-      "epoch": 0.10556270631397376,
-      "grad_norm": 4.344188213348389,
-      "learning_rate": 8.945363922942207e-05,
-      "loss": 0.905,
-      "num_input_tokens_seen": 24264824,
-      "step": 1507
-    },
-    {
-      "epoch": 0.105632754559703,
-      "grad_norm": 6.156280517578125,
-      "learning_rate": 8.944664098073557e-05,
-      "loss": 1.3046,
-      "num_input_tokens_seen": 24281208,
-      "step": 1508
-    },
-    {
-      "epoch": 0.10570280280543225,
-      "grad_norm": 4.687212944030762,
-      "learning_rate": 8.943964273204904e-05,
-      "loss": 1.1898,
-      "num_input_tokens_seen": 24297592,
-      "step": 1509
-    },
-    {
-      "epoch": 0.10577285105116149,
-      "grad_norm": 3.9128546714782715,
-      "learning_rate": 8.943264448336253e-05,
-      "loss": 1.0506,
-      "num_input_tokens_seen": 24313976,
-      "step": 1510
-    },
-    {
-      "epoch": 0.10584289929689074,
-      "grad_norm": 5.766979694366455,
-      "learning_rate": 8.9425646234676e-05,
-      "loss": 1.119,
-      "num_input_tokens_seen": 24330296,
-      "step": 1511
-    },
-    {
-      "epoch": 0.10591294754261998,
-      "grad_norm": 3.9610238075256348,
-      "learning_rate": 8.94186479859895e-05,
-      "loss": 1.279,
-      "num_input_tokens_seen": 24346680,
-      "step": 1512
-    },
-    {
-      "epoch": 0.10598299578834923,
-      "grad_norm": 4.262688636779785,
-      "learning_rate": 8.941164973730297e-05,
-      "loss": 1.3217,
-      "num_input_tokens_seen": 24362408,
-      "step": 1513
-    },
-    {
-      "epoch": 0.10605304403407848,
-      "grad_norm": 5.190121173858643,
-      "learning_rate": 8.940465148861647e-05,
-      "loss": 1.0615,
-      "num_input_tokens_seen": 24378248,
-      "step": 1514
-    },
-    {
-      "epoch": 0.10612309227980772,
-      "grad_norm": 4.5859479904174805,
-      "learning_rate": 8.939765323992996e-05,
-      "loss": 1.1377,
-      "num_input_tokens_seen": 24394632,
-      "step": 1515
-    },
-    {
-      "epoch": 0.10619314052553697,
-      "grad_norm": 4.021294593811035,
-      "learning_rate": 8.939065499124344e-05,
-      "loss": 0.9913,
-      "num_input_tokens_seen": 24411016,
-      "step": 1516
-    },
-    {
-      "epoch": 0.10626318877126621,
-      "grad_norm": 4.296265602111816,
-      "learning_rate": 8.938365674255692e-05,
-      "loss": 1.1753,
-      "num_input_tokens_seen": 24426792,
-      "step": 1517
-    },
-    {
-      "epoch": 0.10633323701699546,
-      "grad_norm": 3.4397289752960205,
-      "learning_rate": 8.93766584938704e-05,
-      "loss": 0.8159,
-      "num_input_tokens_seen": 24443176,
-      "step": 1518
-    },
-    {
-      "epoch": 0.1064032852627247,
-      "grad_norm": 4.009952545166016,
-      "learning_rate": 8.936966024518389e-05,
-      "loss": 1.0728,
-      "num_input_tokens_seen": 24459384,
-      "step": 1519
-    },
-    {
-      "epoch": 0.10647333350845395,
-      "grad_norm": 4.786280632019043,
-      "learning_rate": 8.936266199649738e-05,
-      "loss": 1.1303,
-      "num_input_tokens_seen": 24474904,
-      "step": 1520
-    },
-    {
-      "epoch": 0.1065433817541832,
-      "grad_norm": 3.869297981262207,
-      "learning_rate": 8.935566374781087e-05,
-      "loss": 1.0829,
-      "num_input_tokens_seen": 24490456,
-      "step": 1521
-    },
-    {
-      "epoch": 0.10661342999991244,
-      "grad_norm": 3.995553731918335,
-      "learning_rate": 8.934866549912435e-05,
-      "loss": 1.0813,
-      "num_input_tokens_seen": 24506840,
-      "step": 1522
-    },
-    {
-      "epoch": 0.10668347824564169,
-      "grad_norm": 4.195018291473389,
-      "learning_rate": 8.934166725043783e-05,
-      "loss": 1.0585,
-      "num_input_tokens_seen": 24522440,
-      "step": 1523
-    },
-    {
-      "epoch": 0.10675352649137093,
-      "grad_norm": 4.0432515144348145,
-      "learning_rate": 8.933466900175132e-05,
-      "loss": 1.0757,
-      "num_input_tokens_seen": 24538824,
-      "step": 1524
-    },
-    {
-      "epoch": 0.10682357473710018,
-      "grad_norm": 5.120638847351074,
-      "learning_rate": 8.93276707530648e-05,
-      "loss": 1.1328,
-      "num_input_tokens_seen": 24555208,
-      "step": 1525
-    },
-    {
-      "epoch": 0.10689362298282942,
-      "grad_norm": 3.925096035003662,
-      "learning_rate": 8.932067250437828e-05,
-      "loss": 1.1569,
-      "num_input_tokens_seen": 24571544,
-      "step": 1526
-    },
-    {
-      "epoch": 0.10696367122855867,
-      "grad_norm": 3.930328130722046,
-      "learning_rate": 8.931367425569177e-05,
-      "loss": 0.9385,
-      "num_input_tokens_seen": 24587736,
-      "step": 1527
-    },
-    {
-      "epoch": 0.10703371947428791,
-      "grad_norm": 3.7056055068969727,
-      "learning_rate": 8.930667600700526e-05,
-      "loss": 0.8675,
-      "num_input_tokens_seen": 24604120,
-      "step": 1528
-    },
-    {
-      "epoch": 0.10710376772001716,
-      "grad_norm": 5.945568561553955,
-      "learning_rate": 8.929967775831875e-05,
-      "loss": 1.0395,
-      "num_input_tokens_seen": 24620504,
-      "step": 1529
-    },
-    {
-      "epoch": 0.1071738159657464,
-      "grad_norm": 3.7765939235687256,
-      "learning_rate": 8.929267950963222e-05,
-      "loss": 0.8796,
-      "num_input_tokens_seen": 24635440,
-      "step": 1530
-    },
-    {
-      "epoch": 0.10724386421147565,
-      "grad_norm": 4.229284286499023,
-      "learning_rate": 8.928568126094571e-05,
-      "loss": 1.0941,
-      "num_input_tokens_seen": 24651824,
-      "step": 1531
-    },
-    {
-      "epoch": 0.1073139124572049,
-      "grad_norm": 4.198834419250488,
-      "learning_rate": 8.92786830122592e-05,
-      "loss": 1.118,
-      "num_input_tokens_seen": 24668208,
-      "step": 1532
-    },
-    {
-      "epoch": 0.10738396070293414,
-      "grad_norm": 8.091620445251465,
-      "learning_rate": 8.927168476357267e-05,
-      "loss": 1.1515,
-      "num_input_tokens_seen": 24684248,
-      "step": 1533
-    },
-    {
-      "epoch": 0.10745400894866339,
-      "grad_norm": 4.091879844665527,
-      "learning_rate": 8.926468651488618e-05,
-      "loss": 1.1283,
-      "num_input_tokens_seen": 24700632,
-      "step": 1534
-    },
-    {
-      "epoch": 0.10752405719439263,
-      "grad_norm": 3.90326189994812,
-      "learning_rate": 8.925768826619966e-05,
-      "loss": 1.047,
-      "num_input_tokens_seen": 24717016,
-      "step": 1535
-    },
-    {
-      "epoch": 0.10759410544012188,
-      "grad_norm": 4.097111225128174,
-      "learning_rate": 8.925069001751314e-05,
-      "loss": 1.1623,
-      "num_input_tokens_seen": 24732776,
-      "step": 1536
-    },
-    {
-      "epoch": 0.10766415368585112,
-      "grad_norm": 3.5537095069885254,
-      "learning_rate": 8.924369176882663e-05,
-      "loss": 0.989,
-      "num_input_tokens_seen": 24749064,
-      "step": 1537
-    },
-    {
-      "epoch": 0.10773420193158037,
-      "grad_norm": 4.3086256980896,
-      "learning_rate": 8.92366935201401e-05,
-      "loss": 1.0864,
-      "num_input_tokens_seen": 24765448,
-      "step": 1538
-    },
-    {
-      "epoch": 0.10780425017730962,
-      "grad_norm": 4.177425861358643,
-      "learning_rate": 8.922969527145359e-05,
-      "loss": 1.0652,
-      "num_input_tokens_seen": 24780816,
-      "step": 1539
-    },
-    {
-      "epoch": 0.10787429842303886,
-      "grad_norm": 3.6013338565826416,
-      "learning_rate": 8.922269702276708e-05,
-      "loss": 1.1045,
-      "num_input_tokens_seen": 24796600,
-      "step": 1540
-    },
-    {
-      "epoch": 0.10794434666876812,
-      "grad_norm": 4.05686092376709,
-      "learning_rate": 8.921569877408057e-05,
-      "loss": 1.1408,
-      "num_input_tokens_seen": 24812984,
-      "step": 1541
-    },
-    {
-      "epoch": 0.10801439491449737,
-      "grad_norm": 4.245424747467041,
-      "learning_rate": 8.920870052539406e-05,
-      "loss": 1.2634,
-      "num_input_tokens_seen": 24829368,
-      "step": 1542
-    },
-    {
-      "epoch": 0.10808444316022661,
-      "grad_norm": 3.9563350677490234,
-      "learning_rate": 8.920170227670753e-05,
-      "loss": 1.1015,
-      "num_input_tokens_seen": 24845752,
-      "step": 1543
-    },
-    {
-      "epoch": 0.10815449140595586,
-      "grad_norm": 4.209373474121094,
-      "learning_rate": 8.919470402802102e-05,
-      "loss": 1.2394,
-      "num_input_tokens_seen": 24862136,
-      "step": 1544
-    },
-    {
-      "epoch": 0.1082245396516851,
-      "grad_norm": 3.6590163707733154,
-      "learning_rate": 8.91877057793345e-05,
-      "loss": 1.0168,
-      "num_input_tokens_seen": 24878520,
-      "step": 1545
-    },
-    {
-      "epoch": 0.10829458789741435,
-      "grad_norm": 3.937568187713623,
-      "learning_rate": 8.918070753064799e-05,
-      "loss": 1.0999,
-      "num_input_tokens_seen": 24894696,
-      "step": 1546
-    },
-    {
-      "epoch": 0.1083646361431436,
-      "grad_norm": 3.948453426361084,
-      "learning_rate": 8.917370928196147e-05,
-      "loss": 1.0565,
-      "num_input_tokens_seen": 24910208,
-      "step": 1547
-    },
-    {
-      "epoch": 0.10843468438887284,
-      "grad_norm": 3.61549711227417,
-      "learning_rate": 8.916671103327496e-05,
-      "loss": 1.0294,
-      "num_input_tokens_seen": 24926592,
-      "step": 1548
-    },
-    {
-      "epoch": 0.10850473263460209,
-      "grad_norm": 4.091664791107178,
-      "learning_rate": 8.915971278458845e-05,
-      "loss": 1.0596,
-      "num_input_tokens_seen": 24942976,
-      "step": 1549
-    },
-    {
-      "epoch": 0.10857478088033133,
-      "grad_norm": 5.494830131530762,
-      "learning_rate": 8.915271453590193e-05,
-      "loss": 1.1564,
-      "num_input_tokens_seen": 24957984,
-      "step": 1550
-    },
-    {
-      "epoch": 0.10864482912606058,
-      "grad_norm": 4.546476364135742,
-      "learning_rate": 8.914571628721541e-05,
-      "loss": 1.0753,
-      "num_input_tokens_seen": 24974368,
-      "step": 1551
-    },
-    {
-      "epoch": 0.10871487737178982,
-      "grad_norm": 3.775996446609497,
-      "learning_rate": 8.91387180385289e-05,
-      "loss": 1.11,
-      "num_input_tokens_seen": 24990200,
-      "step": 1552
-    },
-    {
-      "epoch": 0.10878492561751907,
-      "grad_norm": 3.9989728927612305,
-      "learning_rate": 8.913171978984238e-05,
-      "loss": 1.0121,
-      "num_input_tokens_seen": 25006584,
-      "step": 1553
-    },
-    {
-      "epoch": 0.10885497386324831,
-      "grad_norm": 4.417224884033203,
-      "learning_rate": 8.912472154115588e-05,
-      "loss": 1.1891,
-      "num_input_tokens_seen": 25022464,
-      "step": 1554
-    },
-    {
-      "epoch": 0.10892502210897756,
-      "grad_norm": 4.604903697967529,
-      "learning_rate": 8.911772329246936e-05,
-      "loss": 0.9414,
-      "num_input_tokens_seen": 25038848,
-      "step": 1555
-    },
-    {
-      "epoch": 0.1089950703547068,
-      "grad_norm": 4.823176860809326,
-      "learning_rate": 8.911072504378284e-05,
-      "loss": 1.1259,
-      "num_input_tokens_seen": 25053776,
-      "step": 1556
-    },
-    {
-      "epoch": 0.10906511860043605,
-      "grad_norm": 3.6778531074523926,
-      "learning_rate": 8.910372679509632e-05,
-      "loss": 0.9995,
-      "num_input_tokens_seen": 25069872,
-      "step": 1557
-    },
-    {
-      "epoch": 0.1091351668461653,
-      "grad_norm": 4.344213485717773,
-      "learning_rate": 8.909672854640981e-05,
-      "loss": 1.1984,
-      "num_input_tokens_seen": 25086256,
-      "step": 1558
-    },
-    {
-      "epoch": 0.10920521509189454,
-      "grad_norm": 4.592464923858643,
-      "learning_rate": 8.90897302977233e-05,
-      "loss": 1.502,
-      "num_input_tokens_seen": 25102640,
-      "step": 1559
-    },
-    {
-      "epoch": 0.10927526333762379,
-      "grad_norm": 4.103248119354248,
-      "learning_rate": 8.908273204903678e-05,
-      "loss": 0.9454,
-      "num_input_tokens_seen": 25118328,
-      "step": 1560
-    },
-    {
-      "epoch": 0.10934531158335303,
-      "grad_norm": 4.637456893920898,
-      "learning_rate": 8.907573380035027e-05,
-      "loss": 1.3611,
-      "num_input_tokens_seen": 25134712,
-      "step": 1561
-    },
-    {
-      "epoch": 0.10941535982908228,
-      "grad_norm": 4.4709930419921875,
-      "learning_rate": 8.906873555166376e-05,
-      "loss": 1.1147,
-      "num_input_tokens_seen": 25149304,
-      "step": 1562
-    },
-    {
-      "epoch": 0.10948540807481152,
-      "grad_norm": 4.154660701751709,
-      "learning_rate": 8.906173730297724e-05,
-      "loss": 1.2855,
-      "num_input_tokens_seen": 25165360,
-      "step": 1563
-    },
-    {
-      "epoch": 0.10955545632054077,
-      "grad_norm": 4.1212334632873535,
-      "learning_rate": 8.905473905429073e-05,
-      "loss": 1.2015,
-      "num_input_tokens_seen": 25181744,
-      "step": 1564
-    },
-    {
-      "epoch": 0.10962550456627002,
-      "grad_norm": 3.8060882091522217,
-      "learning_rate": 8.90477408056042e-05,
-      "loss": 1.0333,
-      "num_input_tokens_seen": 25197800,
-      "step": 1565
-    },
-    {
-      "epoch": 0.10969555281199926,
-      "grad_norm": 3.4948956966400146,
-      "learning_rate": 8.904074255691769e-05,
-      "loss": 0.941,
-      "num_input_tokens_seen": 25214008,
-      "step": 1566
-    },
-    {
-      "epoch": 0.1097656010577285,
-      "grad_norm": 4.181606292724609,
-      "learning_rate": 8.903374430823118e-05,
-      "loss": 1.1185,
-      "num_input_tokens_seen": 25229496,
-      "step": 1567
-    },
-    {
-      "epoch": 0.10983564930345775,
-      "grad_norm": 4.206098556518555,
-      "learning_rate": 8.902674605954467e-05,
-      "loss": 1.0363,
-      "num_input_tokens_seen": 25244864,
-      "step": 1568
-    },
-    {
-      "epoch": 0.109905697549187,
-      "grad_norm": 3.797475576400757,
-      "learning_rate": 8.901974781085815e-05,
-      "loss": 1.0443,
-      "num_input_tokens_seen": 25261248,
-      "step": 1569
-    },
-    {
-      "epoch": 0.10997574579491624,
-      "grad_norm": 4.131814479827881,
-      "learning_rate": 8.901274956217163e-05,
-      "loss": 0.9977,
-      "num_input_tokens_seen": 25277632,
-      "step": 1570
-    },
-    {
-      "epoch": 0.11004579404064549,
-      "grad_norm": 3.9447309970855713,
-      "learning_rate": 8.900575131348512e-05,
-      "loss": 1.0839,
-      "num_input_tokens_seen": 25294016,
-      "step": 1571
-    },
-    {
-      "epoch": 0.11011584228637473,
-      "grad_norm": 3.916949510574341,
-      "learning_rate": 8.89987530647986e-05,
-      "loss": 1.1793,
-      "num_input_tokens_seen": 25309912,
-      "step": 1572
-    },
-    {
-      "epoch": 0.11018589053210398,
-      "grad_norm": 3.7132885456085205,
-      "learning_rate": 8.899175481611208e-05,
-      "loss": 1.081,
-      "num_input_tokens_seen": 25326296,
-      "step": 1573
-    },
-    {
-      "epoch": 0.11025593877783323,
-      "grad_norm": 4.5842390060424805,
-      "learning_rate": 8.898475656742558e-05,
-      "loss": 0.926,
-      "num_input_tokens_seen": 25342328,
-      "step": 1574
-    },
-    {
-      "epoch": 0.11032598702356247,
-      "grad_norm": 3.578962802886963,
-      "learning_rate": 8.897775831873906e-05,
-      "loss": 1.0599,
-      "num_input_tokens_seen": 25357640,
-      "step": 1575
-    },
-    {
-      "epoch": 0.11039603526929173,
-      "grad_norm": 3.5823471546173096,
-      "learning_rate": 8.897076007005255e-05,
-      "loss": 0.9519,
-      "num_input_tokens_seen": 25373424,
-      "step": 1576
-    },
-    {
-      "epoch": 0.11046608351502098,
-      "grad_norm": 3.721482515335083,
-      "learning_rate": 8.896376182136602e-05,
-      "loss": 0.976,
-      "num_input_tokens_seen": 25389808,
-      "step": 1577
-    },
-    {
-      "epoch": 0.11053613176075022,
-      "grad_norm": 4.874295711517334,
-      "learning_rate": 8.895676357267951e-05,
-      "loss": 1.3507,
-      "num_input_tokens_seen": 25406192,
-      "step": 1578
-    },
-    {
-      "epoch": 0.11060618000647947,
-      "grad_norm": 3.8547258377075195,
-      "learning_rate": 8.8949765323993e-05,
-      "loss": 0.9444,
-      "num_input_tokens_seen": 25421632,
-      "step": 1579
-    },
-    {
-      "epoch": 0.11067622825220871,
-      "grad_norm": 4.847586631774902,
-      "learning_rate": 8.894276707530649e-05,
-      "loss": 1.0526,
-      "num_input_tokens_seen": 25438016,
-      "step": 1580
-    },
-    {
-      "epoch": 0.11074627649793796,
-      "grad_norm": 3.950594425201416,
-      "learning_rate": 8.893576882661998e-05,
-      "loss": 1.0688,
-      "num_input_tokens_seen": 25454400,
-      "step": 1581
-    },
-    {
-      "epoch": 0.1108163247436672,
-      "grad_norm": 3.7372758388519287,
-      "learning_rate": 8.892877057793345e-05,
-      "loss": 1.2211,
-      "num_input_tokens_seen": 25470304,
-      "step": 1582
-    },
-    {
-      "epoch": 0.11088637298939645,
-      "grad_norm": 3.8695788383483887,
-      "learning_rate": 8.892177232924694e-05,
-      "loss": 1.1006,
-      "num_input_tokens_seen": 25486688,
-      "step": 1583
-    },
-    {
-      "epoch": 0.1109564212351257,
-      "grad_norm": 4.623810768127441,
-      "learning_rate": 8.891477408056042e-05,
-      "loss": 1.034,
-      "num_input_tokens_seen": 25503072,
-      "step": 1584
-    },
-    {
-      "epoch": 0.11102646948085494,
-      "grad_norm": 4.03538179397583,
-      "learning_rate": 8.89077758318739e-05,
-      "loss": 1.0915,
-      "num_input_tokens_seen": 25519008,
-      "step": 1585
-    },
-    {
-      "epoch": 0.11109651772658419,
-      "grad_norm": 7.486603736877441,
-      "learning_rate": 8.890077758318739e-05,
-      "loss": 1.0137,
-      "num_input_tokens_seen": 25533808,
-      "step": 1586
-    },
-    {
-      "epoch": 0.11116656597231343,
-      "grad_norm": 4.660414218902588,
-      "learning_rate": 8.889377933450088e-05,
-      "loss": 1.0172,
-      "num_input_tokens_seen": 25549784,
-      "step": 1587
-    },
-    {
-      "epoch": 0.11123661421804268,
-      "grad_norm": 3.9375548362731934,
-      "learning_rate": 8.888678108581437e-05,
-      "loss": 0.9843,
-      "num_input_tokens_seen": 25566168,
-      "step": 1588
-    },
-    {
-      "epoch": 0.11130666246377192,
-      "grad_norm": 4.275035858154297,
-      "learning_rate": 8.887978283712786e-05,
-      "loss": 1.1802,
-      "num_input_tokens_seen": 25582552,
-      "step": 1589
-    },
-    {
-      "epoch": 0.11137671070950117,
-      "grad_norm": 4.799124240875244,
-      "learning_rate": 8.887278458844133e-05,
-      "loss": 1.2702,
-      "num_input_tokens_seen": 25598936,
-      "step": 1590
-    },
-    {
-      "epoch": 0.11144675895523042,
-      "grad_norm": 4.143614768981934,
-      "learning_rate": 8.886578633975482e-05,
-      "loss": 1.1797,
-      "num_input_tokens_seen": 25615320,
-      "step": 1591
-    },
-    {
-      "epoch": 0.11151680720095966,
-      "grad_norm": 4.490556716918945,
-      "learning_rate": 8.88587880910683e-05,
-      "loss": 1.1351,
-      "num_input_tokens_seen": 25630624,
-      "step": 1592
-    },
-    {
-      "epoch": 0.1115868554466889,
-      "grad_norm": 6.010688781738281,
-      "learning_rate": 8.885178984238179e-05,
-      "loss": 1.059,
-      "num_input_tokens_seen": 25646048,
-      "step": 1593
-    },
-    {
-      "epoch": 0.11165690369241815,
-      "grad_norm": 3.7447726726531982,
-      "learning_rate": 8.884479159369527e-05,
-      "loss": 0.9902,
-      "num_input_tokens_seen": 25661528,
-      "step": 1594
-    },
-    {
-      "epoch": 0.1117269519381474,
-      "grad_norm": 4.77920389175415,
-      "learning_rate": 8.883779334500876e-05,
-      "loss": 1.1158,
-      "num_input_tokens_seen": 25677912,
-      "step": 1595
-    },
-    {
-      "epoch": 0.11179700018387664,
-      "grad_norm": 3.9812231063842773,
-      "learning_rate": 8.883079509632225e-05,
-      "loss": 1.096,
-      "num_input_tokens_seen": 25694296,
-      "step": 1596
-    },
-    {
-      "epoch": 0.11186704842960589,
-      "grad_norm": 3.7404634952545166,
-      "learning_rate": 8.882379684763573e-05,
-      "loss": 0.9965,
-      "num_input_tokens_seen": 25710448,
-      "step": 1597
-    },
-    {
-      "epoch": 0.11193709667533513,
-      "grad_norm": 4.466211318969727,
-      "learning_rate": 8.881679859894922e-05,
-      "loss": 1.1495,
-      "num_input_tokens_seen": 25726624,
-      "step": 1598
-    },
-    {
-      "epoch": 0.11200714492106438,
-      "grad_norm": 3.6850225925445557,
-      "learning_rate": 8.880980035026269e-05,
-      "loss": 0.9685,
-      "num_input_tokens_seen": 25742456,
-      "step": 1599
-    },
-    {
-      "epoch": 0.11207719316679363,
-      "grad_norm": 4.128363609313965,
-      "learning_rate": 8.880280210157619e-05,
-      "loss": 1.1052,
-      "num_input_tokens_seen": 25758840,
-      "step": 1600
-    },
-    {
-      "epoch": 0.11207719316679363,
-      "eval_loss": 1.1512293815612793,
-      "eval_runtime": 0.1988,
-      "eval_samples_per_second": 5.031,
-      "eval_steps_per_second": 5.031,
-      "num_input_tokens_seen": 25758840,
-      "step": 1600
-    },
-    {
-      "epoch": 0.11214724141252287,
-      "grad_norm": 4.852661609649658,
-      "learning_rate": 8.879580385288968e-05,
-      "loss": 1.0778,
-      "num_input_tokens_seen": 25774312,
-      "step": 1601
-    },
-    {
-      "epoch": 0.11221728965825212,
-      "grad_norm": 4.501857280731201,
-      "learning_rate": 8.878880560420316e-05,
-      "loss": 1.302,
-      "num_input_tokens_seen": 25790696,
-      "step": 1602
-    },
-    {
-      "epoch": 0.11228733790398136,
-      "grad_norm": 4.142490863800049,
-      "learning_rate": 8.878180735551665e-05,
-      "loss": 1.0375,
-      "num_input_tokens_seen": 25807080,
-      "step": 1603
-    },
-    {
-      "epoch": 0.11235738614971061,
-      "grad_norm": 3.606905698776245,
-      "learning_rate": 8.877480910683012e-05,
-      "loss": 0.9254,
-      "num_input_tokens_seen": 25822552,
-      "step": 1604
-    },
-    {
-      "epoch": 0.11242743439543985,
-      "grad_norm": 3.837010145187378,
-      "learning_rate": 8.876781085814361e-05,
-      "loss": 1.1756,
-      "num_input_tokens_seen": 25838088,
-      "step": 1605
-    },
-    {
-      "epoch": 0.1124974826411691,
-      "grad_norm": 3.9082963466644287,
-      "learning_rate": 8.87608126094571e-05,
-      "loss": 1.0201,
-      "num_input_tokens_seen": 25854240,
-      "step": 1606
-    },
-    {
-      "epoch": 0.11256753088689835,
-      "grad_norm": 4.062923908233643,
-      "learning_rate": 8.875381436077059e-05,
-      "loss": 1.1034,
-      "num_input_tokens_seen": 25870624,
-      "step": 1607
-    },
-    {
-      "epoch": 0.11263757913262759,
-      "grad_norm": 4.331594944000244,
-      "learning_rate": 8.874681611208407e-05,
-      "loss": 1.2043,
-      "num_input_tokens_seen": 25886656,
-      "step": 1608
-    },
-    {
-      "epoch": 0.11270762737835684,
-      "grad_norm": 3.77466082572937,
-      "learning_rate": 8.873981786339755e-05,
-      "loss": 0.936,
-      "num_input_tokens_seen": 25902704,
-      "step": 1609
-    },
-    {
-      "epoch": 0.1127776756240861,
-      "grad_norm": 3.3747365474700928,
-      "learning_rate": 8.873281961471104e-05,
-      "loss": 0.9071,
-      "num_input_tokens_seen": 25919088,
-      "step": 1610
-    },
-    {
-      "epoch": 0.11284772386981534,
-      "grad_norm": 5.377493381500244,
-      "learning_rate": 8.872582136602451e-05,
-      "loss": 0.9246,
-      "num_input_tokens_seen": 25935472,
-      "step": 1611
-    },
-    {
-      "epoch": 0.11291777211554459,
-      "grad_norm": 5.506969451904297,
-      "learning_rate": 8.8718823117338e-05,
-      "loss": 0.9211,
-      "num_input_tokens_seen": 25951664,
-      "step": 1612
-    },
-    {
-      "epoch": 0.11298782036127383,
-      "grad_norm": 4.874104976654053,
-      "learning_rate": 8.871182486865149e-05,
-      "loss": 1.1654,
-      "num_input_tokens_seen": 25968048,
-      "step": 1613
-    },
-    {
-      "epoch": 0.11305786860700308,
-      "grad_norm": 4.666824817657471,
-      "learning_rate": 8.870482661996498e-05,
-      "loss": 1.2155,
-      "num_input_tokens_seen": 25983784,
-      "step": 1614
-    },
-    {
-      "epoch": 0.11312791685273232,
-      "grad_norm": 3.949862241744995,
-      "learning_rate": 8.869782837127847e-05,
-      "loss": 1.1243,
-      "num_input_tokens_seen": 26000168,
-      "step": 1615
-    },
-    {
-      "epoch": 0.11319796509846157,
-      "grad_norm": 3.866542339324951,
-      "learning_rate": 8.869083012259196e-05,
-      "loss": 1.1302,
-      "num_input_tokens_seen": 26015456,
-      "step": 1616
-    },
-    {
-      "epoch": 0.11326801334419082,
-      "grad_norm": 3.8679909706115723,
-      "learning_rate": 8.868383187390543e-05,
-      "loss": 1.0886,
-      "num_input_tokens_seen": 26031224,
-      "step": 1617
-    },
-    {
-      "epoch": 0.11333806158992006,
-      "grad_norm": 4.7508087158203125,
-      "learning_rate": 8.867683362521892e-05,
-      "loss": 1.2837,
-      "num_input_tokens_seen": 26046952,
-      "step": 1618
-    },
-    {
-      "epoch": 0.1134081098356493,
-      "grad_norm": 3.878549337387085,
-      "learning_rate": 8.86698353765324e-05,
-      "loss": 0.99,
-      "num_input_tokens_seen": 26063280,
-      "step": 1619
-    },
-    {
-      "epoch": 0.11347815808137855,
-      "grad_norm": 3.8016276359558105,
-      "learning_rate": 8.86628371278459e-05,
-      "loss": 1.1682,
-      "num_input_tokens_seen": 26079616,
-      "step": 1620
-    },
-    {
-      "epoch": 0.1135482063271078,
-      "grad_norm": 4.040102481842041,
-      "learning_rate": 8.865583887915937e-05,
-      "loss": 1.1008,
-      "num_input_tokens_seen": 26095232,
-      "step": 1621
-    },
-    {
-      "epoch": 0.11361825457283704,
-      "grad_norm": 3.932529926300049,
-      "learning_rate": 8.864884063047286e-05,
-      "loss": 1.1663,
-      "num_input_tokens_seen": 26111616,
-      "step": 1622
-    },
-    {
-      "epoch": 0.11368830281856629,
-      "grad_norm": 4.568112373352051,
-      "learning_rate": 8.864184238178635e-05,
-      "loss": 1.1932,
-      "num_input_tokens_seen": 26128000,
-      "step": 1623
-    },
-    {
-      "epoch": 0.11375835106429553,
-      "grad_norm": 4.23036527633667,
-      "learning_rate": 8.863484413309982e-05,
-      "loss": 1.0223,
-      "num_input_tokens_seen": 26144384,
-      "step": 1624
-    },
-    {
-      "epoch": 0.11382839931002478,
-      "grad_norm": 4.209012031555176,
-      "learning_rate": 8.862784588441331e-05,
-      "loss": 1.0992,
-      "num_input_tokens_seen": 26160768,
-      "step": 1625
-    },
-    {
-      "epoch": 0.11389844755575403,
-      "grad_norm": 3.865983724594116,
-      "learning_rate": 8.86208476357268e-05,
-      "loss": 1.1213,
-      "num_input_tokens_seen": 26177152,
-      "step": 1626
-    },
-    {
-      "epoch": 0.11396849580148327,
-      "grad_norm": 3.781083822250366,
-      "learning_rate": 8.861384938704029e-05,
-      "loss": 1.0132,
-      "num_input_tokens_seen": 26193536,
-      "step": 1627
-    },
-    {
-      "epoch": 0.11403854404721252,
-      "grad_norm": 4.330471038818359,
-      "learning_rate": 8.860685113835378e-05,
-      "loss": 0.9749,
-      "num_input_tokens_seen": 26208976,
-      "step": 1628
-    },
-    {
-      "epoch": 0.11410859229294176,
-      "grad_norm": 4.772238254547119,
-      "learning_rate": 8.859985288966725e-05,
-      "loss": 1.2796,
-      "num_input_tokens_seen": 26225360,
-      "step": 1629
-    },
-    {
-      "epoch": 0.11417864053867101,
-      "grad_norm": 4.0468668937683105,
-      "learning_rate": 8.859285464098074e-05,
-      "loss": 1.0056,
-      "num_input_tokens_seen": 26241744,
-      "step": 1630
-    },
-    {
-      "epoch": 0.11424868878440025,
-      "grad_norm": 3.9648735523223877,
-      "learning_rate": 8.858585639229422e-05,
-      "loss": 1.2185,
-      "num_input_tokens_seen": 26258128,
-      "step": 1631
-    },
-    {
-      "epoch": 0.1143187370301295,
-      "grad_norm": 4.7014079093933105,
-      "learning_rate": 8.85788581436077e-05,
-      "loss": 1.1795,
-      "num_input_tokens_seen": 26274512,
-      "step": 1632
-    },
-    {
-      "epoch": 0.11438878527585875,
-      "grad_norm": 4.6375627517700195,
-      "learning_rate": 8.85718598949212e-05,
-      "loss": 1.0074,
-      "num_input_tokens_seen": 26290008,
-      "step": 1633
-    },
-    {
-      "epoch": 0.11445883352158799,
-      "grad_norm": 4.427719593048096,
-      "learning_rate": 8.856486164623468e-05,
-      "loss": 1.2769,
-      "num_input_tokens_seen": 26305512,
-      "step": 1634
-    },
-    {
-      "epoch": 0.11452888176731724,
-      "grad_norm": 6.001821994781494,
-      "learning_rate": 8.855786339754817e-05,
-      "loss": 1.0606,
-      "num_input_tokens_seen": 26319504,
-      "step": 1635
-    },
-    {
-      "epoch": 0.11459893001304648,
-      "grad_norm": 3.970672369003296,
-      "learning_rate": 8.855086514886165e-05,
-      "loss": 1.1944,
-      "num_input_tokens_seen": 26335888,
-      "step": 1636
-    },
-    {
-      "epoch": 0.11466897825877573,
-      "grad_norm": 3.924450635910034,
-      "learning_rate": 8.854386690017514e-05,
-      "loss": 0.9607,
-      "num_input_tokens_seen": 26351536,
-      "step": 1637
-    },
-    {
-      "epoch": 0.11473902650450497,
-      "grad_norm": 4.400977611541748,
-      "learning_rate": 8.853686865148861e-05,
-      "loss": 1.0641,
-      "num_input_tokens_seen": 26367808,
-      "step": 1638
-    },
-    {
-      "epoch": 0.11480907475023422,
-      "grad_norm": 3.9734365940093994,
-      "learning_rate": 8.85298704028021e-05,
-      "loss": 1.2258,
-      "num_input_tokens_seen": 26383864,
-      "step": 1639
-    },
-    {
-      "epoch": 0.11487912299596346,
-      "grad_norm": 3.792949914932251,
-      "learning_rate": 8.85228721541156e-05,
-      "loss": 1.0401,
-      "num_input_tokens_seen": 26400248,
-      "step": 1640
-    },
-    {
-      "epoch": 0.11494917124169271,
-      "grad_norm": 5.14591121673584,
-      "learning_rate": 8.851587390542908e-05,
-      "loss": 1.0484,
-      "num_input_tokens_seen": 26416056,
-      "step": 1641
-    },
-    {
-      "epoch": 0.11501921948742196,
-      "grad_norm": 5.0158162117004395,
-      "learning_rate": 8.850887565674256e-05,
-      "loss": 1.2823,
-      "num_input_tokens_seen": 26431400,
-      "step": 1642
-    },
-    {
-      "epoch": 0.1150892677331512,
-      "grad_norm": 4.459201812744141,
-      "learning_rate": 8.850187740805605e-05,
-      "loss": 1.2371,
-      "num_input_tokens_seen": 26446920,
-      "step": 1643
-    },
-    {
-      "epoch": 0.11515931597888045,
-      "grad_norm": 3.717949867248535,
-      "learning_rate": 8.849487915936953e-05,
-      "loss": 1.1299,
-      "num_input_tokens_seen": 26463304,
-      "step": 1644
-    },
-    {
-      "epoch": 0.1152293642246097,
-      "grad_norm": 3.7555253505706787,
-      "learning_rate": 8.848788091068302e-05,
-      "loss": 1.0835,
-      "num_input_tokens_seen": 26479296,
-      "step": 1645
-    },
-    {
-      "epoch": 0.11529941247033895,
-      "grad_norm": 4.3726325035095215,
-      "learning_rate": 8.84808826619965e-05,
-      "loss": 0.9606,
-      "num_input_tokens_seen": 26495024,
-      "step": 1646
-    },
-    {
-      "epoch": 0.1153694607160682,
-      "grad_norm": 3.728700876235962,
-      "learning_rate": 8.847388441331e-05,
-      "loss": 1.0486,
-      "num_input_tokens_seen": 26511408,
-      "step": 1647
-    },
-    {
-      "epoch": 0.11543950896179744,
-      "grad_norm": 4.276855945587158,
-      "learning_rate": 8.846688616462347e-05,
-      "loss": 0.9869,
-      "num_input_tokens_seen": 26527688,
-      "step": 1648
-    },
-    {
-      "epoch": 0.11550955720752669,
-      "grad_norm": 5.386009693145752,
-      "learning_rate": 8.845988791593696e-05,
-      "loss": 1.0021,
-      "num_input_tokens_seen": 26544072,
-      "step": 1649
-    },
-    {
-      "epoch": 0.11557960545325593,
-      "grad_norm": 4.978610992431641,
-      "learning_rate": 8.845288966725045e-05,
-      "loss": 1.2531,
-      "num_input_tokens_seen": 26560456,
-      "step": 1650
-    },
-    {
-      "epoch": 0.11564965369898518,
-      "grad_norm": 5.325594425201416,
-      "learning_rate": 8.844589141856392e-05,
-      "loss": 0.9983,
-      "num_input_tokens_seen": 26576840,
-      "step": 1651
-    },
-    {
-      "epoch": 0.11571970194471443,
-      "grad_norm": 4.359868049621582,
-      "learning_rate": 8.843889316987741e-05,
-      "loss": 0.9652,
-      "num_input_tokens_seen": 26593224,
-      "step": 1652
-    },
-    {
-      "epoch": 0.11578975019044367,
-      "grad_norm": 7.921500205993652,
-      "learning_rate": 8.84318949211909e-05,
-      "loss": 1.0767,
-      "num_input_tokens_seen": 26607352,
-      "step": 1653
-    },
-    {
-      "epoch": 0.11585979843617292,
-      "grad_norm": 3.51788330078125,
-      "learning_rate": 8.842489667250439e-05,
-      "loss": 1.0677,
-      "num_input_tokens_seen": 26623696,
-      "step": 1654
-    },
-    {
-      "epoch": 0.11592984668190216,
-      "grad_norm": 4.120747089385986,
-      "learning_rate": 8.841789842381788e-05,
-      "loss": 1.2139,
-      "num_input_tokens_seen": 26639832,
-      "step": 1655
-    },
-    {
-      "epoch": 0.11599989492763141,
-      "grad_norm": 4.077361106872559,
-      "learning_rate": 8.841090017513135e-05,
-      "loss": 1.0639,
-      "num_input_tokens_seen": 26655432,
-      "step": 1656
-    },
-    {
-      "epoch": 0.11606994317336065,
-      "grad_norm": 3.9629955291748047,
-      "learning_rate": 8.840390192644484e-05,
-      "loss": 1.0846,
-      "num_input_tokens_seen": 26671816,
-      "step": 1657
-    },
-    {
-      "epoch": 0.1161399914190899,
-      "grad_norm": 3.933544635772705,
-      "learning_rate": 8.839690367775831e-05,
-      "loss": 1.1543,
-      "num_input_tokens_seen": 26688096,
-      "step": 1658
-    },
-    {
-      "epoch": 0.11621003966481915,
-      "grad_norm": 4.702983379364014,
-      "learning_rate": 8.83899054290718e-05,
-      "loss": 1.0699,
-      "num_input_tokens_seen": 26704480,
-      "step": 1659
-    },
-    {
-      "epoch": 0.11628008791054839,
-      "grad_norm": 4.536739826202393,
-      "learning_rate": 8.83829071803853e-05,
-      "loss": 1.149,
-      "num_input_tokens_seen": 26720864,
-      "step": 1660
-    },
-    {
-      "epoch": 0.11635013615627764,
-      "grad_norm": 4.419711589813232,
-      "learning_rate": 8.837590893169878e-05,
-      "loss": 1.1994,
-      "num_input_tokens_seen": 26737248,
-      "step": 1661
-    },
-    {
-      "epoch": 0.11642018440200688,
-      "grad_norm": 4.106175899505615,
-      "learning_rate": 8.836891068301227e-05,
-      "loss": 1.0682,
-      "num_input_tokens_seen": 26753632,
-      "step": 1662
-    },
-    {
-      "epoch": 0.11649023264773613,
-      "grad_norm": 3.469658374786377,
-      "learning_rate": 8.836191243432574e-05,
-      "loss": 1.0356,
-      "num_input_tokens_seen": 26769944,
-      "step": 1663
-    },
-    {
-      "epoch": 0.11656028089346537,
-      "grad_norm": 7.273227691650391,
-      "learning_rate": 8.835491418563923e-05,
-      "loss": 1.1699,
-      "num_input_tokens_seen": 26784520,
-      "step": 1664
-    },
-    {
-      "epoch": 0.11663032913919462,
-      "grad_norm": 3.611165761947632,
-      "learning_rate": 8.834791593695271e-05,
-      "loss": 0.8595,
-      "num_input_tokens_seen": 26800360,
-      "step": 1665
-    },
-    {
-      "epoch": 0.11670037738492386,
-      "grad_norm": 4.405304908752441,
-      "learning_rate": 8.834091768826621e-05,
-      "loss": 1.2055,
-      "num_input_tokens_seen": 26816744,
-      "step": 1666
-    },
-    {
-      "epoch": 0.11677042563065311,
-      "grad_norm": 3.897247791290283,
-      "learning_rate": 8.83339194395797e-05,
-      "loss": 0.9599,
-      "num_input_tokens_seen": 26832520,
-      "step": 1667
-    },
-    {
-      "epoch": 0.11684047387638236,
-      "grad_norm": 3.898019313812256,
-      "learning_rate": 8.832692119089317e-05,
-      "loss": 1.0838,
-      "num_input_tokens_seen": 26848080,
-      "step": 1668
-    },
-    {
-      "epoch": 0.1169105221221116,
-      "grad_norm": 4.6351542472839355,
-      "learning_rate": 8.831992294220666e-05,
-      "loss": 1.2776,
-      "num_input_tokens_seen": 26864464,
-      "step": 1669
-    },
-    {
-      "epoch": 0.11698057036784085,
-      "grad_norm": 4.020237922668457,
-      "learning_rate": 8.831292469352015e-05,
-      "loss": 0.9955,
-      "num_input_tokens_seen": 26880848,
-      "step": 1670
-    },
-    {
-      "epoch": 0.11705061861357009,
-      "grad_norm": 5.813192367553711,
-      "learning_rate": 8.830592644483363e-05,
-      "loss": 1.2867,
-      "num_input_tokens_seen": 26897232,
-      "step": 1671
-    },
-    {
-      "epoch": 0.11712066685929934,
-      "grad_norm": 4.058423042297363,
-      "learning_rate": 8.829892819614711e-05,
-      "loss": 1.0697,
-      "num_input_tokens_seen": 26912872,
-      "step": 1672
-    },
-    {
-      "epoch": 0.11719071510502858,
-      "grad_norm": 4.76987361907959,
-      "learning_rate": 8.82919299474606e-05,
-      "loss": 0.9226,
-      "num_input_tokens_seen": 26929256,
-      "step": 1673
-    },
-    {
-      "epoch": 0.11726076335075783,
-      "grad_norm": 3.8400967121124268,
-      "learning_rate": 8.828493169877409e-05,
-      "loss": 1.0089,
-      "num_input_tokens_seen": 26945624,
-      "step": 1674
-    },
-    {
-      "epoch": 0.11733081159648708,
-      "grad_norm": 4.49709415435791,
-      "learning_rate": 8.827793345008757e-05,
-      "loss": 1.0898,
-      "num_input_tokens_seen": 26961464,
-      "step": 1675
-    },
-    {
-      "epoch": 0.11740085984221632,
-      "grad_norm": 4.143093109130859,
-      "learning_rate": 8.827093520140105e-05,
-      "loss": 1.0493,
-      "num_input_tokens_seen": 26976720,
-      "step": 1676
-    },
-    {
-      "epoch": 0.11747090808794557,
-      "grad_norm": 4.138030529022217,
-      "learning_rate": 8.826393695271454e-05,
-      "loss": 1.1555,
-      "num_input_tokens_seen": 26993056,
-      "step": 1677
-    },
-    {
-      "epoch": 0.11754095633367481,
-      "grad_norm": 3.8191847801208496,
-      "learning_rate": 8.825693870402802e-05,
-      "loss": 1.0993,
-      "num_input_tokens_seen": 27009440,
-      "step": 1678
-    },
-    {
-      "epoch": 0.11761100457940406,
-      "grad_norm": 3.8392176628112793,
-      "learning_rate": 8.824994045534151e-05,
-      "loss": 1.1067,
-      "num_input_tokens_seen": 27024880,
-      "step": 1679
-    },
-    {
-      "epoch": 0.11768105282513332,
-      "grad_norm": 4.468568801879883,
-      "learning_rate": 8.8242942206655e-05,
-      "loss": 1.1424,
-      "num_input_tokens_seen": 27040672,
-      "step": 1680
-    },
-    {
-      "epoch": 0.11775110107086256,
-      "grad_norm": 3.6515510082244873,
-      "learning_rate": 8.823594395796848e-05,
-      "loss": 1.0659,
-      "num_input_tokens_seen": 27057056,
-      "step": 1681
-    },
-    {
-      "epoch": 0.11782114931659181,
-      "grad_norm": 4.479739189147949,
-      "learning_rate": 8.822894570928197e-05,
-      "loss": 1.0399,
-      "num_input_tokens_seen": 27073440,
-      "step": 1682
-    },
-    {
-      "epoch": 0.11789119756232105,
-      "grad_norm": 3.762479782104492,
-      "learning_rate": 8.822194746059545e-05,
-      "loss": 1.1041,
-      "num_input_tokens_seen": 27089824,
-      "step": 1683
-    },
-    {
-      "epoch": 0.1179612458080503,
-      "grad_norm": 4.694389343261719,
-      "learning_rate": 8.821494921190894e-05,
-      "loss": 1.2785,
-      "num_input_tokens_seen": 27106208,
-      "step": 1684
-    },
-    {
-      "epoch": 0.11803129405377955,
-      "grad_norm": 3.738931179046631,
-      "learning_rate": 8.820795096322241e-05,
-      "loss": 0.9039,
-      "num_input_tokens_seen": 27122352,
-      "step": 1685
-    },
-    {
-      "epoch": 0.11810134229950879,
-      "grad_norm": 4.065624237060547,
-      "learning_rate": 8.820095271453591e-05,
-      "loss": 1.0048,
-      "num_input_tokens_seen": 27138160,
-      "step": 1686
-    },
-    {
-      "epoch": 0.11817139054523804,
-      "grad_norm": 3.5373826026916504,
-      "learning_rate": 8.81939544658494e-05,
-      "loss": 0.8786,
-      "num_input_tokens_seen": 27154544,
-      "step": 1687
-    },
-    {
-      "epoch": 0.11824143879096728,
-      "grad_norm": 3.773066282272339,
-      "learning_rate": 8.818695621716288e-05,
-      "loss": 1.0043,
-      "num_input_tokens_seen": 27170928,
-      "step": 1688
-    },
-    {
-      "epoch": 0.11831148703669653,
-      "grad_norm": 3.3876242637634277,
-      "learning_rate": 8.817995796847637e-05,
-      "loss": 0.9909,
-      "num_input_tokens_seen": 27187312,
-      "step": 1689
-    },
-    {
-      "epoch": 0.11838153528242577,
-      "grad_norm": 4.526343822479248,
-      "learning_rate": 8.817295971978984e-05,
-      "loss": 1.0899,
-      "num_input_tokens_seen": 27202208,
-      "step": 1690
-    },
-    {
-      "epoch": 0.11845158352815502,
-      "grad_norm": 4.691114902496338,
-      "learning_rate": 8.816596147110333e-05,
-      "loss": 1.0823,
-      "num_input_tokens_seen": 27218592,
-      "step": 1691
-    },
-    {
-      "epoch": 0.11852163177388426,
-      "grad_norm": 3.90531849861145,
-      "learning_rate": 8.815896322241682e-05,
-      "loss": 1.1438,
-      "num_input_tokens_seen": 27234976,
-      "step": 1692
-    },
-    {
-      "epoch": 0.11859168001961351,
-      "grad_norm": 3.5546317100524902,
-      "learning_rate": 8.81519649737303e-05,
-      "loss": 1.0326,
-      "num_input_tokens_seen": 27251360,
-      "step": 1693
-    },
-    {
-      "epoch": 0.11866172826534276,
-      "grad_norm": 5.117360591888428,
-      "learning_rate": 8.81449667250438e-05,
-      "loss": 1.1921,
-      "num_input_tokens_seen": 27267744,
-      "step": 1694
-    },
-    {
-      "epoch": 0.118731776511072,
-      "grad_norm": 4.055267810821533,
-      "learning_rate": 8.813796847635727e-05,
-      "loss": 1.0607,
-      "num_input_tokens_seen": 27283688,
-      "step": 1695
-    },
-    {
-      "epoch": 0.11880182475680125,
-      "grad_norm": 4.04268741607666,
-      "learning_rate": 8.813097022767076e-05,
-      "loss": 1.1862,
-      "num_input_tokens_seen": 27300072,
-      "step": 1696
-    },
-    {
-      "epoch": 0.11887187300253049,
-      "grad_norm": 4.048800945281982,
-      "learning_rate": 8.812397197898425e-05,
-      "loss": 0.9231,
-      "num_input_tokens_seen": 27316456,
-      "step": 1697
-    },
-    {
-      "epoch": 0.11894192124825974,
-      "grad_norm": 4.445494174957275,
-      "learning_rate": 8.811697373029772e-05,
-      "loss": 1.241,
-      "num_input_tokens_seen": 27332464,
-      "step": 1698
-    },
-    {
-      "epoch": 0.11901196949398898,
-      "grad_norm": 4.522054672241211,
-      "learning_rate": 8.810997548161121e-05,
-      "loss": 1.3945,
-      "num_input_tokens_seen": 27348848,
-      "step": 1699
-    },
-    {
-      "epoch": 0.11908201773971823,
-      "grad_norm": 4.106349468231201,
-      "learning_rate": 8.81029772329247e-05,
-      "loss": 1.1457,
-      "num_input_tokens_seen": 27365232,
-      "step": 1700
-    },
-    {
-      "epoch": 0.11915206598544748,
-      "grad_norm": 6.059356689453125,
-      "learning_rate": 8.809597898423819e-05,
-      "loss": 1.3381,
-      "num_input_tokens_seen": 27380448,
-      "step": 1701
-    },
-    {
-      "epoch": 0.11922211423117672,
-      "grad_norm": 3.8089959621429443,
-      "learning_rate": 8.808898073555166e-05,
-      "loss": 1.0699,
-      "num_input_tokens_seen": 27396832,
-      "step": 1702
-    },
-    {
-      "epoch": 0.11929216247690597,
-      "grad_norm": 4.21024227142334,
-      "learning_rate": 8.808198248686515e-05,
-      "loss": 1.306,
-      "num_input_tokens_seen": 27413096,
-      "step": 1703
-    },
-    {
-      "epoch": 0.11936221072263521,
-      "grad_norm": 4.286004066467285,
-      "learning_rate": 8.807498423817864e-05,
-      "loss": 1.2325,
-      "num_input_tokens_seen": 27429480,
-      "step": 1704
-    },
-    {
-      "epoch": 0.11943225896836446,
-      "grad_norm": 3.512561559677124,
-      "learning_rate": 8.806798598949212e-05,
-      "loss": 0.8804,
-      "num_input_tokens_seen": 27445864,
-      "step": 1705
-    },
-    {
-      "epoch": 0.1195023072140937,
-      "grad_norm": 4.096526145935059,
-      "learning_rate": 8.806098774080562e-05,
-      "loss": 1.0591,
-      "num_input_tokens_seen": 27462248,
-      "step": 1706
-    },
-    {
-      "epoch": 0.11957235545982295,
-      "grad_norm": 5.032350063323975,
-      "learning_rate": 8.805398949211909e-05,
-      "loss": 0.8948,
-      "num_input_tokens_seen": 27478312,
-      "step": 1707
-    },
-    {
-      "epoch": 0.1196424037055522,
-      "grad_norm": 4.756420612335205,
-      "learning_rate": 8.804699124343258e-05,
-      "loss": 1.0584,
-      "num_input_tokens_seen": 27494696,
-      "step": 1708
-    },
-    {
-      "epoch": 0.11971245195128144,
-      "grad_norm": 4.869518756866455,
-      "learning_rate": 8.803999299474607e-05,
-      "loss": 0.9394,
-      "num_input_tokens_seen": 27511080,
-      "step": 1709
-    },
-    {
-      "epoch": 0.11978250019701069,
-      "grad_norm": 3.451759099960327,
-      "learning_rate": 8.803299474605954e-05,
-      "loss": 0.9171,
-      "num_input_tokens_seen": 27527328,
-      "step": 1710
-    },
-    {
-      "epoch": 0.11985254844273993,
-      "grad_norm": 4.247021675109863,
-      "learning_rate": 8.802599649737303e-05,
-      "loss": 1.1204,
-      "num_input_tokens_seen": 27543712,
-      "step": 1711
-    },
-    {
-      "epoch": 0.11992259668846918,
-      "grad_norm": 4.597024917602539,
-      "learning_rate": 8.801899824868652e-05,
-      "loss": 1.196,
-      "num_input_tokens_seen": 27560096,
-      "step": 1712
-    },
-    {
-      "epoch": 0.11999264493419842,
-      "grad_norm": 4.242952823638916,
-      "learning_rate": 8.801200000000001e-05,
-      "loss": 1.1747,
-      "num_input_tokens_seen": 27576320,
-      "step": 1713
-    },
-    {
-      "epoch": 0.12006269317992768,
-      "grad_norm": 5.1166486740112305,
-      "learning_rate": 8.80050017513135e-05,
-      "loss": 1.4222,
-      "num_input_tokens_seen": 27591024,
-      "step": 1714
-    },
-    {
-      "epoch": 0.12013274142565693,
-      "grad_norm": 4.6713714599609375,
-      "learning_rate": 8.799800350262697e-05,
-      "loss": 1.1869,
-      "num_input_tokens_seen": 27606352,
-      "step": 1715
-    },
-    {
-      "epoch": 0.12020278967138617,
-      "grad_norm": 4.62678861618042,
-      "learning_rate": 8.799100525394046e-05,
-      "loss": 1.1524,
-      "num_input_tokens_seen": 27622736,
-      "step": 1716
-    },
-    {
-      "epoch": 0.12027283791711542,
-      "grad_norm": 3.611985206604004,
-      "learning_rate": 8.798400700525394e-05,
-      "loss": 1.1179,
-      "num_input_tokens_seen": 27639120,
-      "step": 1717
-    },
-    {
-      "epoch": 0.12034288616284466,
-      "grad_norm": 4.165099143981934,
-      "learning_rate": 8.797700875656743e-05,
-      "loss": 1.0104,
-      "num_input_tokens_seen": 27654024,
-      "step": 1718
-    },
-    {
-      "epoch": 0.12041293440857391,
-      "grad_norm": 4.532061576843262,
-      "learning_rate": 8.797001050788091e-05,
-      "loss": 1.05,
-      "num_input_tokens_seen": 27670408,
-      "step": 1719
-    },
-    {
-      "epoch": 0.12048298265430316,
-      "grad_norm": 4.880197048187256,
-      "learning_rate": 8.79630122591944e-05,
-      "loss": 1.0321,
-      "num_input_tokens_seen": 27686792,
-      "step": 1720
-    },
-    {
-      "epoch": 0.1205530309000324,
-      "grad_norm": 3.521052360534668,
-      "learning_rate": 8.795601401050789e-05,
-      "loss": 0.9048,
-      "num_input_tokens_seen": 27703176,
-      "step": 1721
-    },
-    {
-      "epoch": 0.12062307914576165,
-      "grad_norm": 3.965725898742676,
-      "learning_rate": 8.794901576182137e-05,
-      "loss": 1.1348,
-      "num_input_tokens_seen": 27719024,
-      "step": 1722
-    },
-    {
-      "epoch": 0.12069312739149089,
-      "grad_norm": 3.936962842941284,
-      "learning_rate": 8.794201751313486e-05,
-      "loss": 1.1531,
-      "num_input_tokens_seen": 27734736,
-      "step": 1723
-    },
-    {
-      "epoch": 0.12076317563722014,
-      "grad_norm": 5.225526332855225,
-      "learning_rate": 8.793501926444834e-05,
-      "loss": 1.2784,
-      "num_input_tokens_seen": 27751120,
-      "step": 1724
-    },
-    {
-      "epoch": 0.12083322388294938,
-      "grad_norm": 4.125289440155029,
-      "learning_rate": 8.792802101576182e-05,
-      "loss": 1.1893,
-      "num_input_tokens_seen": 27767288,
-      "step": 1725
-    },
-    {
-      "epoch": 0.12090327212867863,
-      "grad_norm": 3.9352405071258545,
-      "learning_rate": 8.792102276707532e-05,
-      "loss": 1.1867,
-      "num_input_tokens_seen": 27783672,
-      "step": 1726
-    },
-    {
-      "epoch": 0.12097332037440787,
-      "grad_norm": 3.908578634262085,
-      "learning_rate": 8.79140245183888e-05,
-      "loss": 1.0024,
-      "num_input_tokens_seen": 27799640,
-      "step": 1727
-    },
-    {
-      "epoch": 0.12104336862013712,
-      "grad_norm": 3.694387435913086,
-      "learning_rate": 8.790702626970229e-05,
-      "loss": 1.0652,
-      "num_input_tokens_seen": 27816024,
-      "step": 1728
-    },
-    {
-      "epoch": 0.12111341686586637,
-      "grad_norm": 4.0100016593933105,
-      "learning_rate": 8.790002802101576e-05,
-      "loss": 1.0511,
-      "num_input_tokens_seen": 27832408,
-      "step": 1729
-    },
-    {
-      "epoch": 0.12118346511159561,
-      "grad_norm": 5.454882621765137,
-      "learning_rate": 8.789302977232925e-05,
-      "loss": 1.1096,
-      "num_input_tokens_seen": 27848792,
-      "step": 1730
-    },
-    {
-      "epoch": 0.12125351335732486,
-      "grad_norm": 5.065526485443115,
-      "learning_rate": 8.788603152364274e-05,
-      "loss": 1.0354,
-      "num_input_tokens_seen": 27864688,
-      "step": 1731
-    },
-    {
-      "epoch": 0.1213235616030541,
-      "grad_norm": 3.73103666305542,
-      "learning_rate": 8.787903327495623e-05,
-      "loss": 1.0328,
-      "num_input_tokens_seen": 27881072,
-      "step": 1732
-    },
-    {
-      "epoch": 0.12139360984878335,
-      "grad_norm": 3.971198081970215,
-      "learning_rate": 8.787203502626971e-05,
-      "loss": 1.1908,
-      "num_input_tokens_seen": 27896912,
-      "step": 1733
-    },
-    {
-      "epoch": 0.1214636580945126,
-      "grad_norm": 3.933809518814087,
-      "learning_rate": 8.786503677758319e-05,
-      "loss": 1.1125,
-      "num_input_tokens_seen": 27913104,
-      "step": 1734
-    },
-    {
-      "epoch": 0.12153370634024184,
-      "grad_norm": 3.92167329788208,
-      "learning_rate": 8.785803852889668e-05,
-      "loss": 1.0007,
-      "num_input_tokens_seen": 27929488,
-      "step": 1735
-    },
-    {
-      "epoch": 0.12160375458597109,
-      "grad_norm": 4.441089630126953,
-      "learning_rate": 8.785104028021017e-05,
-      "loss": 0.9748,
-      "num_input_tokens_seen": 27945504,
-      "step": 1736
-    },
-    {
-      "epoch": 0.12167380283170033,
-      "grad_norm": 4.023623466491699,
-      "learning_rate": 8.784404203152364e-05,
-      "loss": 0.8826,
-      "num_input_tokens_seen": 27961888,
-      "step": 1737
-    },
-    {
-      "epoch": 0.12174385107742958,
-      "grad_norm": 4.0328826904296875,
-      "learning_rate": 8.783704378283713e-05,
-      "loss": 1.2769,
-      "num_input_tokens_seen": 27978024,
-      "step": 1738
-    },
-    {
-      "epoch": 0.12181389932315882,
-      "grad_norm": 4.5445733070373535,
-      "learning_rate": 8.783004553415062e-05,
-      "loss": 1.3745,
-      "num_input_tokens_seen": 27993840,
-      "step": 1739
-    },
-    {
-      "epoch": 0.12188394756888807,
-      "grad_norm": 3.609834671020508,
-      "learning_rate": 8.782304728546411e-05,
-      "loss": 0.916,
-      "num_input_tokens_seen": 28010224,
-      "step": 1740
-    },
-    {
-      "epoch": 0.12195399581461731,
-      "grad_norm": 3.849306344985962,
-      "learning_rate": 8.78160490367776e-05,
-      "loss": 1.1135,
-      "num_input_tokens_seen": 28026232,
-      "step": 1741
-    },
-    {
-      "epoch": 0.12202404406034656,
-      "grad_norm": 4.11102294921875,
-      "learning_rate": 8.780905078809107e-05,
-      "loss": 1.2269,
-      "num_input_tokens_seen": 28041880,
-      "step": 1742
-    },
-    {
-      "epoch": 0.1220940923060758,
-      "grad_norm": 4.156986713409424,
-      "learning_rate": 8.780205253940456e-05,
-      "loss": 1.0321,
-      "num_input_tokens_seen": 28058264,
-      "step": 1743
-    },
-    {
-      "epoch": 0.12216414055180505,
-      "grad_norm": 3.9670159816741943,
-      "learning_rate": 8.779505429071803e-05,
-      "loss": 0.9752,
-      "num_input_tokens_seen": 28073168,
-      "step": 1744
-    },
-    {
-      "epoch": 0.1222341887975343,
-      "grad_norm": 5.342650890350342,
-      "learning_rate": 8.778805604203152e-05,
-      "loss": 1.1416,
-      "num_input_tokens_seen": 28089552,
-      "step": 1745
-    },
-    {
-      "epoch": 0.12230423704326354,
-      "grad_norm": 4.031285285949707,
-      "learning_rate": 8.778105779334501e-05,
-      "loss": 1.1134,
-      "num_input_tokens_seen": 28105264,
-      "step": 1746
-    },
-    {
-      "epoch": 0.12237428528899279,
-      "grad_norm": 3.5976450443267822,
-      "learning_rate": 8.77740595446585e-05,
-      "loss": 1.0342,
-      "num_input_tokens_seen": 28121648,
-      "step": 1747
-    },
-    {
-      "epoch": 0.12244433353472203,
-      "grad_norm": 4.947859764099121,
-      "learning_rate": 8.776706129597199e-05,
-      "loss": 1.0809,
-      "num_input_tokens_seen": 28137640,
-      "step": 1748
-    },
-    {
-      "epoch": 0.12251438178045129,
-      "grad_norm": 4.004949569702148,
-      "learning_rate": 8.776006304728546e-05,
-      "loss": 1.0921,
-      "num_input_tokens_seen": 28154024,
-      "step": 1749
-    },
-    {
-      "epoch": 0.12258443002618054,
-      "grad_norm": 3.9022445678710938,
-      "learning_rate": 8.775306479859895e-05,
-      "loss": 1.0844,
-      "num_input_tokens_seen": 28170408,
-      "step": 1750
-    },
-    {
-      "epoch": 0.12265447827190978,
-      "grad_norm": 4.171925067901611,
-      "learning_rate": 8.774606654991244e-05,
-      "loss": 1.1894,
-      "num_input_tokens_seen": 28186792,
-      "step": 1751
-    },
-    {
-      "epoch": 0.12272452651763903,
-      "grad_norm": 3.9387433528900146,
-      "learning_rate": 8.773906830122592e-05,
-      "loss": 1.0303,
-      "num_input_tokens_seen": 28203176,
-      "step": 1752
-    },
-    {
-      "epoch": 0.12279457476336827,
-      "grad_norm": 5.067278861999512,
-      "learning_rate": 8.773207005253942e-05,
-      "loss": 1.1924,
-      "num_input_tokens_seen": 28219192,
-      "step": 1753
-    },
-    {
-      "epoch": 0.12286462300909752,
-      "grad_norm": 3.673807144165039,
-      "learning_rate": 8.77250718038529e-05,
-      "loss": 1.0438,
-      "num_input_tokens_seen": 28235576,
-      "step": 1754
-    },
-    {
-      "epoch": 0.12293467125482677,
-      "grad_norm": 5.303588390350342,
-      "learning_rate": 8.771807355516638e-05,
-      "loss": 1.2601,
-      "num_input_tokens_seen": 28251960,
-      "step": 1755
-    },
-    {
-      "epoch": 0.12300471950055601,
-      "grad_norm": 5.343825340270996,
-      "learning_rate": 8.771107530647986e-05,
-      "loss": 1.1126,
-      "num_input_tokens_seen": 28268344,
-      "step": 1756
-    },
-    {
-      "epoch": 0.12307476774628526,
-      "grad_norm": 4.125874996185303,
-      "learning_rate": 8.770407705779335e-05,
-      "loss": 1.1497,
-      "num_input_tokens_seen": 28284144,
-      "step": 1757
-    },
-    {
-      "epoch": 0.1231448159920145,
-      "grad_norm": 4.628546714782715,
-      "learning_rate": 8.769707880910683e-05,
-      "loss": 1.1757,
-      "num_input_tokens_seen": 28299896,
-      "step": 1758
-    },
-    {
-      "epoch": 0.12321486423774375,
-      "grad_norm": 3.946603775024414,
-      "learning_rate": 8.769008056042032e-05,
-      "loss": 1.2739,
-      "num_input_tokens_seen": 28316280,
-      "step": 1759
-    },
-    {
-      "epoch": 0.123284912483473,
-      "grad_norm": 3.4837770462036133,
-      "learning_rate": 8.768308231173381e-05,
-      "loss": 0.9682,
-      "num_input_tokens_seen": 28332128,
-      "step": 1760
-    },
-    {
-      "epoch": 0.12335496072920224,
-      "grad_norm": 3.9601573944091797,
-      "learning_rate": 8.767608406304729e-05,
-      "loss": 1.2647,
-      "num_input_tokens_seen": 28347488,
-      "step": 1761
-    },
-    {
-      "epoch": 0.12342500897493149,
-      "grad_norm": 4.178001403808594,
-      "learning_rate": 8.766908581436078e-05,
-      "loss": 1.0055,
-      "num_input_tokens_seen": 28363872,
-      "step": 1762
-    },
-    {
-      "epoch": 0.12349505722066073,
-      "grad_norm": 3.9182498455047607,
-      "learning_rate": 8.766208756567426e-05,
-      "loss": 1.1407,
-      "num_input_tokens_seen": 28380208,
-      "step": 1763
-    },
-    {
-      "epoch": 0.12356510546638998,
-      "grad_norm": 4.071939468383789,
-      "learning_rate": 8.765508931698774e-05,
-      "loss": 1.3196,
-      "num_input_tokens_seen": 28396592,
-      "step": 1764
-    },
-    {
-      "epoch": 0.12363515371211922,
-      "grad_norm": 4.657908916473389,
-      "learning_rate": 8.764809106830123e-05,
-      "loss": 1.0739,
-      "num_input_tokens_seen": 28412976,
-      "step": 1765
-    },
-    {
-      "epoch": 0.12370520195784847,
-      "grad_norm": 3.9706201553344727,
-      "learning_rate": 8.764109281961472e-05,
-      "loss": 1.0904,
-      "num_input_tokens_seen": 28429088,
-      "step": 1766
-    },
-    {
-      "epoch": 0.12377525020357771,
-      "grad_norm": 4.571341514587402,
-      "learning_rate": 8.76340945709282e-05,
-      "loss": 1.1314,
-      "num_input_tokens_seen": 28445472,
-      "step": 1767
-    },
-    {
-      "epoch": 0.12384529844930696,
-      "grad_norm": 4.197002410888672,
-      "learning_rate": 8.762709632224169e-05,
-      "loss": 0.8251,
-      "num_input_tokens_seen": 28461656,
-      "step": 1768
-    },
-    {
-      "epoch": 0.1239153466950362,
-      "grad_norm": 5.376040935516357,
-      "learning_rate": 8.762009807355517e-05,
-      "loss": 1.1626,
-      "num_input_tokens_seen": 28477088,
-      "step": 1769
-    },
-    {
-      "epoch": 0.12398539494076545,
-      "grad_norm": 3.987495183944702,
-      "learning_rate": 8.761309982486866e-05,
-      "loss": 1.2449,
-      "num_input_tokens_seen": 28493472,
-      "step": 1770
-    },
-    {
-      "epoch": 0.1240554431864947,
-      "grad_norm": 4.379208564758301,
-      "learning_rate": 8.760610157618213e-05,
-      "loss": 1.2834,
-      "num_input_tokens_seen": 28509856,
-      "step": 1771
-    },
-    {
-      "epoch": 0.12412549143222394,
-      "grad_norm": 3.7258729934692383,
-      "learning_rate": 8.759910332749562e-05,
-      "loss": 1.1115,
-      "num_input_tokens_seen": 28525664,
-      "step": 1772
-    },
-    {
-      "epoch": 0.12419553967795319,
-      "grad_norm": 4.0574774742126465,
-      "learning_rate": 8.759210507880911e-05,
-      "loss": 1.1005,
-      "num_input_tokens_seen": 28541920,
-      "step": 1773
-    },
-    {
-      "epoch": 0.12426558792368243,
-      "grad_norm": 3.8423895835876465,
-      "learning_rate": 8.75851068301226e-05,
-      "loss": 1.1067,
-      "num_input_tokens_seen": 28558216,
-      "step": 1774
-    },
-    {
-      "epoch": 0.12433563616941168,
-      "grad_norm": 3.8898398876190186,
-      "learning_rate": 8.757810858143609e-05,
-      "loss": 1.1963,
-      "num_input_tokens_seen": 28574536,
-      "step": 1775
-    },
-    {
-      "epoch": 0.12440568441514092,
-      "grad_norm": 3.286412000656128,
-      "learning_rate": 8.757111033274956e-05,
-      "loss": 0.9159,
-      "num_input_tokens_seen": 28590920,
-      "step": 1776
-    },
-    {
-      "epoch": 0.12447573266087017,
-      "grad_norm": 3.7219464778900146,
-      "learning_rate": 8.756411208406305e-05,
-      "loss": 1.0883,
-      "num_input_tokens_seen": 28607192,
-      "step": 1777
-    },
-    {
-      "epoch": 0.12454578090659942,
-      "grad_norm": 3.8907012939453125,
-      "learning_rate": 8.755711383537654e-05,
-      "loss": 1.0226,
-      "num_input_tokens_seen": 28623176,
-      "step": 1778
-    },
-    {
-      "epoch": 0.12461582915232866,
-      "grad_norm": 3.8087925910949707,
-      "learning_rate": 8.755011558669003e-05,
-      "loss": 1.0115,
-      "num_input_tokens_seen": 28639528,
-      "step": 1779
-    },
-    {
-      "epoch": 0.1246858773980579,
-      "grad_norm": 4.8956217765808105,
-      "learning_rate": 8.754311733800352e-05,
-      "loss": 1.0108,
-      "num_input_tokens_seen": 28654976,
-      "step": 1780
-    },
-    {
-      "epoch": 0.12475592564378715,
-      "grad_norm": 3.7400572299957275,
-      "learning_rate": 8.753611908931699e-05,
-      "loss": 0.8787,
-      "num_input_tokens_seen": 28671064,
-      "step": 1781
-    },
-    {
-      "epoch": 0.1248259738895164,
-      "grad_norm": 4.689199924468994,
-      "learning_rate": 8.752912084063048e-05,
-      "loss": 1.2326,
-      "num_input_tokens_seen": 28686664,
-      "step": 1782
-    },
-    {
-      "epoch": 0.12489602213524566,
-      "grad_norm": 3.6594929695129395,
-      "learning_rate": 8.752212259194395e-05,
-      "loss": 1.1626,
-      "num_input_tokens_seen": 28703048,
-      "step": 1783
-    },
-    {
-      "epoch": 0.1249660703809749,
-      "grad_norm": 4.6070356369018555,
-      "learning_rate": 8.751512434325744e-05,
-      "loss": 1.358,
-      "num_input_tokens_seen": 28719000,
-      "step": 1784
-    },
-    {
-      "epoch": 0.12503611862670413,
-      "grad_norm": 4.658362865447998,
-      "learning_rate": 8.750812609457093e-05,
-      "loss": 1.2852,
-      "num_input_tokens_seen": 28735384,
-      "step": 1785
-    },
-    {
-      "epoch": 0.1251061668724334,
-      "grad_norm": 3.6963465213775635,
-      "learning_rate": 8.750112784588442e-05,
-      "loss": 1.1068,
-      "num_input_tokens_seen": 28750856,
-      "step": 1786
-    },
-    {
-      "epoch": 0.12517621511816263,
-      "grad_norm": 4.419562816619873,
-      "learning_rate": 8.749412959719791e-05,
-      "loss": 1.1559,
-      "num_input_tokens_seen": 28766824,
-      "step": 1787
-    },
-    {
-      "epoch": 0.12524626336389189,
-      "grad_norm": 4.601676940917969,
-      "learning_rate": 8.74871313485114e-05,
-      "loss": 1.0642,
-      "num_input_tokens_seen": 28783208,
-      "step": 1788
-    },
-    {
-      "epoch": 0.12531631160962112,
-      "grad_norm": 3.8597445487976074,
-      "learning_rate": 8.748013309982487e-05,
-      "loss": 1.1149,
-      "num_input_tokens_seen": 28799160,
-      "step": 1789
-    },
-    {
-      "epoch": 0.12538635985535038,
-      "grad_norm": 3.654649257659912,
-      "learning_rate": 8.747313485113835e-05,
-      "loss": 1.3127,
-      "num_input_tokens_seen": 28815440,
-      "step": 1790
-    },
-    {
-      "epoch": 0.1254564081010796,
-      "grad_norm": 4.043321132659912,
-      "learning_rate": 8.746613660245184e-05,
-      "loss": 1.0844,
-      "num_input_tokens_seen": 28831824,
-      "step": 1791
-    },
-    {
-      "epoch": 0.12552645634680887,
-      "grad_norm": 4.5223894119262695,
-      "learning_rate": 8.745913835376532e-05,
-      "loss": 1.0627,
-      "num_input_tokens_seen": 28846984,
-      "step": 1792
-    },
-    {
-      "epoch": 0.1255965045925381,
-      "grad_norm": 4.074361801147461,
-      "learning_rate": 8.745214010507881e-05,
-      "loss": 0.9772,
-      "num_input_tokens_seen": 28863368,
-      "step": 1793
-    },
-    {
-      "epoch": 0.12566655283826736,
-      "grad_norm": 4.661183834075928,
-      "learning_rate": 8.74451418563923e-05,
-      "loss": 1.152,
-      "num_input_tokens_seen": 28879752,
-      "step": 1794
-    },
-    {
-      "epoch": 0.1257366010839966,
-      "grad_norm": 3.95831561088562,
-      "learning_rate": 8.743814360770579e-05,
-      "loss": 1.117,
-      "num_input_tokens_seen": 28895728,
-      "step": 1795
-    },
-    {
-      "epoch": 0.12580664932972585,
-      "grad_norm": 4.271726131439209,
-      "learning_rate": 8.743114535901927e-05,
-      "loss": 1.0935,
-      "num_input_tokens_seen": 28912112,
-      "step": 1796
-    },
-    {
-      "epoch": 0.12587669757545508,
-      "grad_norm": 4.079075336456299,
-      "learning_rate": 8.742414711033275e-05,
-      "loss": 1.1397,
-      "num_input_tokens_seen": 28928496,
-      "step": 1797
-    },
-    {
-      "epoch": 0.12594674582118434,
-      "grad_norm": 4.030980587005615,
-      "learning_rate": 8.741714886164623e-05,
-      "loss": 0.9405,
-      "num_input_tokens_seen": 28943968,
-      "step": 1798
-    },
-    {
-      "epoch": 0.12601679406691357,
-      "grad_norm": 3.7285454273223877,
-      "learning_rate": 8.741015061295973e-05,
-      "loss": 0.8448,
-      "num_input_tokens_seen": 28959800,
-      "step": 1799
-    },
-    {
-      "epoch": 0.12608684231264283,
-      "grad_norm": 3.964663028717041,
-      "learning_rate": 8.74031523642732e-05,
-      "loss": 1.1614,
-      "num_input_tokens_seen": 28976184,
-      "step": 1800
-    },
-    {
-      "epoch": 0.12608684231264283,
-      "eval_loss": 1.1493111848831177,
-      "eval_runtime": 0.196,
-      "eval_samples_per_second": 5.102,
-      "eval_steps_per_second": 5.102,
-      "num_input_tokens_seen": 28976184,
-      "step": 1800
-    }
-  ],
-  "logging_steps": 1,
-  "max_steps": 14275,
-  "num_input_tokens_seen": 28976184,
-  "num_train_epochs": 1,
-  "save_steps": 200,
-  "stateful_callbacks": {
-    "TrainerControl": {
-      "args": {
-        "should_epoch_stop": false,
-        "should_evaluate": false,
-        "should_log": false,
-        "should_save": true,
-        "should_training_stop": false
-      },
-      "attributes": {}
-    }
-  },
-  "total_flos": 6.222312927938765e+16,
-  "train_batch_size": 1,
-  "trial_name": null,
-  "trial_params": null
-}