sparky-decoder / checkpoint-1000 /trainer_state.json
Dohahemdann's picture
Checkpoint checkpoint-1000
f8ebf87 verified
Raw
History Blame Contribute Delete
173 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.7614696363982486,
"eval_steps": 500,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0007614696363982486,
"grad_norm": 1301.5599365234375,
"learning_rate": 0.0,
"loss": 1.7023,
"step": 1
},
{
"epoch": 0.0015229392727964973,
"grad_norm": 1499.36865234375,
"learning_rate": 2.0000000000000002e-07,
"loss": 1.7043,
"step": 2
},
{
"epoch": 0.002284408909194746,
"grad_norm": 1399.9569091796875,
"learning_rate": 4.0000000000000003e-07,
"loss": 1.6019,
"step": 3
},
{
"epoch": 0.0030458785455929945,
"grad_norm": 1188.724365234375,
"learning_rate": 6.000000000000001e-07,
"loss": 1.7423,
"step": 4
},
{
"epoch": 0.003807348181991243,
"grad_norm": 969.7083129882812,
"learning_rate": 8.000000000000001e-07,
"loss": 1.6516,
"step": 5
},
{
"epoch": 0.004568817818389492,
"grad_norm": 962.1920776367188,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.2925,
"step": 6
},
{
"epoch": 0.00533028745478774,
"grad_norm": 1861.9527587890625,
"learning_rate": 1.2000000000000002e-06,
"loss": 1.4859,
"step": 7
},
{
"epoch": 0.006091757091185989,
"grad_norm": 1748.2479248046875,
"learning_rate": 1.4000000000000001e-06,
"loss": 1.8132,
"step": 8
},
{
"epoch": 0.006853226727584237,
"grad_norm": 1467.362060546875,
"learning_rate": 1.6000000000000001e-06,
"loss": 1.5848,
"step": 9
},
{
"epoch": 0.007614696363982486,
"grad_norm": 1578.7880859375,
"learning_rate": 1.8000000000000001e-06,
"loss": 1.6334,
"step": 10
},
{
"epoch": 0.008376166000380734,
"grad_norm": 1163.3890380859375,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.4836,
"step": 11
},
{
"epoch": 0.009137635636778984,
"grad_norm": 1121.4964599609375,
"learning_rate": 2.2e-06,
"loss": 1.7535,
"step": 12
},
{
"epoch": 0.009899105273177233,
"grad_norm": 1086.3167724609375,
"learning_rate": 2.4000000000000003e-06,
"loss": 1.56,
"step": 13
},
{
"epoch": 0.01066057490957548,
"grad_norm": 876.811767578125,
"learning_rate": 2.6e-06,
"loss": 1.2632,
"step": 14
},
{
"epoch": 0.011422044545973729,
"grad_norm": 1044.1190185546875,
"learning_rate": 2.8000000000000003e-06,
"loss": 1.3597,
"step": 15
},
{
"epoch": 0.012183514182371978,
"grad_norm": 1221.429443359375,
"learning_rate": 3e-06,
"loss": 1.7357,
"step": 16
},
{
"epoch": 0.012944983818770227,
"grad_norm": 1007.3869018554688,
"learning_rate": 3.2000000000000003e-06,
"loss": 1.5886,
"step": 17
},
{
"epoch": 0.013706453455168474,
"grad_norm": 1217.14404296875,
"learning_rate": 3.4000000000000005e-06,
"loss": 1.386,
"step": 18
},
{
"epoch": 0.014467923091566724,
"grad_norm": 1351.4454345703125,
"learning_rate": 3.6000000000000003e-06,
"loss": 1.6928,
"step": 19
},
{
"epoch": 0.015229392727964973,
"grad_norm": 1504.17041015625,
"learning_rate": 3.8000000000000005e-06,
"loss": 1.4281,
"step": 20
},
{
"epoch": 0.01599086236436322,
"grad_norm": 1153.16162109375,
"learning_rate": 4.000000000000001e-06,
"loss": 1.1809,
"step": 21
},
{
"epoch": 0.01675233200076147,
"grad_norm": 1124.1148681640625,
"learning_rate": 4.2000000000000004e-06,
"loss": 1.7594,
"step": 22
},
{
"epoch": 0.01751380163715972,
"grad_norm": 1091.021484375,
"learning_rate": 4.4e-06,
"loss": 1.4819,
"step": 23
},
{
"epoch": 0.018275271273557967,
"grad_norm": 802.41259765625,
"learning_rate": 4.600000000000001e-06,
"loss": 1.3459,
"step": 24
},
{
"epoch": 0.019036740909956214,
"grad_norm": 1319.851318359375,
"learning_rate": 4.800000000000001e-06,
"loss": 1.6257,
"step": 25
},
{
"epoch": 0.019798210546354465,
"grad_norm": 977.6575927734375,
"learning_rate": 5e-06,
"loss": 1.5093,
"step": 26
},
{
"epoch": 0.020559680182752713,
"grad_norm": 774.690673828125,
"learning_rate": 5.2e-06,
"loss": 1.4782,
"step": 27
},
{
"epoch": 0.02132114981915096,
"grad_norm": 937.8590087890625,
"learning_rate": 5.400000000000001e-06,
"loss": 1.6011,
"step": 28
},
{
"epoch": 0.02208261945554921,
"grad_norm": 756.9816284179688,
"learning_rate": 5.600000000000001e-06,
"loss": 1.6296,
"step": 29
},
{
"epoch": 0.022844089091947458,
"grad_norm": 1100.748291015625,
"learning_rate": 5.8e-06,
"loss": 1.6768,
"step": 30
},
{
"epoch": 0.02360555872834571,
"grad_norm": 976.7691650390625,
"learning_rate": 6e-06,
"loss": 1.423,
"step": 31
},
{
"epoch": 0.024367028364743956,
"grad_norm": 871.838134765625,
"learning_rate": 6.200000000000001e-06,
"loss": 1.3862,
"step": 32
},
{
"epoch": 0.025128498001142203,
"grad_norm": 807.175537109375,
"learning_rate": 6.4000000000000006e-06,
"loss": 1.5093,
"step": 33
},
{
"epoch": 0.025889967637540454,
"grad_norm": 964.3636474609375,
"learning_rate": 6.600000000000001e-06,
"loss": 1.7866,
"step": 34
},
{
"epoch": 0.0266514372739387,
"grad_norm": 953.1375122070312,
"learning_rate": 6.800000000000001e-06,
"loss": 1.4513,
"step": 35
},
{
"epoch": 0.02741290691033695,
"grad_norm": 893.89111328125,
"learning_rate": 7e-06,
"loss": 1.8488,
"step": 36
},
{
"epoch": 0.0281743765467352,
"grad_norm": 833.2611694335938,
"learning_rate": 7.2000000000000005e-06,
"loss": 1.3522,
"step": 37
},
{
"epoch": 0.028935846183133447,
"grad_norm": 837.7489013671875,
"learning_rate": 7.4e-06,
"loss": 1.5152,
"step": 38
},
{
"epoch": 0.029697315819531698,
"grad_norm": 814.172607421875,
"learning_rate": 7.600000000000001e-06,
"loss": 1.2972,
"step": 39
},
{
"epoch": 0.030458785455929945,
"grad_norm": 810.1613159179688,
"learning_rate": 7.800000000000002e-06,
"loss": 1.5865,
"step": 40
},
{
"epoch": 0.031220255092328193,
"grad_norm": 496.3629150390625,
"learning_rate": 8.000000000000001e-06,
"loss": 1.4072,
"step": 41
},
{
"epoch": 0.03198172472872644,
"grad_norm": 620.7055053710938,
"learning_rate": 8.2e-06,
"loss": 1.732,
"step": 42
},
{
"epoch": 0.03274319436512469,
"grad_norm": 1089.77294921875,
"learning_rate": 8.400000000000001e-06,
"loss": 1.4347,
"step": 43
},
{
"epoch": 0.03350466400152294,
"grad_norm": 829.736328125,
"learning_rate": 8.6e-06,
"loss": 1.4516,
"step": 44
},
{
"epoch": 0.034266133637921185,
"grad_norm": 1636.154052734375,
"learning_rate": 8.8e-06,
"loss": 1.8134,
"step": 45
},
{
"epoch": 0.03502760327431944,
"grad_norm": 1268.646728515625,
"learning_rate": 9e-06,
"loss": 1.4501,
"step": 46
},
{
"epoch": 0.03578907291071769,
"grad_norm": 590.24755859375,
"learning_rate": 9.200000000000002e-06,
"loss": 1.4093,
"step": 47
},
{
"epoch": 0.036550542547115934,
"grad_norm": 687.4384765625,
"learning_rate": 9.4e-06,
"loss": 1.5449,
"step": 48
},
{
"epoch": 0.03731201218351418,
"grad_norm": 532.1568603515625,
"learning_rate": 9.600000000000001e-06,
"loss": 1.463,
"step": 49
},
{
"epoch": 0.03807348181991243,
"grad_norm": 606.5445556640625,
"learning_rate": 9.800000000000001e-06,
"loss": 1.7299,
"step": 50
},
{
"epoch": 0.038834951456310676,
"grad_norm": 494.4815368652344,
"learning_rate": 1e-05,
"loss": 1.6656,
"step": 51
},
{
"epoch": 0.03959642109270893,
"grad_norm": 478.6385192871094,
"learning_rate": 9.99999628743664e-06,
"loss": 1.6203,
"step": 52
},
{
"epoch": 0.04035789072910718,
"grad_norm": 495.6545104980469,
"learning_rate": 9.999985149752072e-06,
"loss": 1.4333,
"step": 53
},
{
"epoch": 0.041119360365505425,
"grad_norm": 489.4484558105469,
"learning_rate": 9.999966586962836e-06,
"loss": 1.3458,
"step": 54
},
{
"epoch": 0.04188083000190367,
"grad_norm": 470.57647705078125,
"learning_rate": 9.999940599096498e-06,
"loss": 1.563,
"step": 55
},
{
"epoch": 0.04264229963830192,
"grad_norm": 402.5634460449219,
"learning_rate": 9.999907186191652e-06,
"loss": 1.3983,
"step": 56
},
{
"epoch": 0.043403769274700174,
"grad_norm": 365.5474548339844,
"learning_rate": 9.999866348297913e-06,
"loss": 1.5199,
"step": 57
},
{
"epoch": 0.04416523891109842,
"grad_norm": 389.02984619140625,
"learning_rate": 9.99981808547593e-06,
"loss": 1.3679,
"step": 58
},
{
"epoch": 0.04492670854749667,
"grad_norm": 384.76470947265625,
"learning_rate": 9.999762397797375e-06,
"loss": 1.3927,
"step": 59
},
{
"epoch": 0.045688178183894916,
"grad_norm": 380.3616638183594,
"learning_rate": 9.999699285344943e-06,
"loss": 1.4385,
"step": 60
},
{
"epoch": 0.04644964782029316,
"grad_norm": 451.0272216796875,
"learning_rate": 9.99962874821236e-06,
"loss": 1.7561,
"step": 61
},
{
"epoch": 0.04721111745669142,
"grad_norm": 546.6317138671875,
"learning_rate": 9.999550786504374e-06,
"loss": 1.6178,
"step": 62
},
{
"epoch": 0.047972587093089665,
"grad_norm": 302.81549072265625,
"learning_rate": 9.99946540033676e-06,
"loss": 1.2085,
"step": 63
},
{
"epoch": 0.04873405672948791,
"grad_norm": 344.7528076171875,
"learning_rate": 9.999372589836319e-06,
"loss": 1.2769,
"step": 64
},
{
"epoch": 0.04949552636588616,
"grad_norm": 418.7420654296875,
"learning_rate": 9.999272355140878e-06,
"loss": 1.4762,
"step": 65
},
{
"epoch": 0.05025699600228441,
"grad_norm": 323.90228271484375,
"learning_rate": 9.999164696399287e-06,
"loss": 1.2773,
"step": 66
},
{
"epoch": 0.051018465638682654,
"grad_norm": 257.0769958496094,
"learning_rate": 9.999049613771421e-06,
"loss": 1.2942,
"step": 67
},
{
"epoch": 0.05177993527508091,
"grad_norm": 366.9392395019531,
"learning_rate": 9.998927107428181e-06,
"loss": 1.3762,
"step": 68
},
{
"epoch": 0.052541404911479156,
"grad_norm": 319.0862731933594,
"learning_rate": 9.998797177551494e-06,
"loss": 1.4763,
"step": 69
},
{
"epoch": 0.0533028745478774,
"grad_norm": 309.5851135253906,
"learning_rate": 9.998659824334307e-06,
"loss": 1.2058,
"step": 70
},
{
"epoch": 0.05406434418427565,
"grad_norm": 331.6963195800781,
"learning_rate": 9.998515047980595e-06,
"loss": 1.4926,
"step": 71
},
{
"epoch": 0.0548258138206739,
"grad_norm": 267.27130126953125,
"learning_rate": 9.998362848705351e-06,
"loss": 1.3501,
"step": 72
},
{
"epoch": 0.05558728345707215,
"grad_norm": 342.3760681152344,
"learning_rate": 9.9982032267346e-06,
"loss": 1.4401,
"step": 73
},
{
"epoch": 0.0563487530934704,
"grad_norm": 256.15966796875,
"learning_rate": 9.99803618230538e-06,
"loss": 1.0673,
"step": 74
},
{
"epoch": 0.05711022272986865,
"grad_norm": 294.26556396484375,
"learning_rate": 9.997861715665758e-06,
"loss": 1.4545,
"step": 75
},
{
"epoch": 0.057871692366266894,
"grad_norm": 309.1824035644531,
"learning_rate": 9.997679827074822e-06,
"loss": 1.3322,
"step": 76
},
{
"epoch": 0.05863316200266514,
"grad_norm": 281.6103515625,
"learning_rate": 9.997490516802682e-06,
"loss": 1.1179,
"step": 77
},
{
"epoch": 0.059394631639063396,
"grad_norm": 259.3569641113281,
"learning_rate": 9.997293785130464e-06,
"loss": 1.413,
"step": 78
},
{
"epoch": 0.06015610127546164,
"grad_norm": 312.6566467285156,
"learning_rate": 9.997089632350325e-06,
"loss": 1.4359,
"step": 79
},
{
"epoch": 0.06091757091185989,
"grad_norm": 202.5186767578125,
"learning_rate": 9.996878058765434e-06,
"loss": 1.2213,
"step": 80
},
{
"epoch": 0.06167904054825814,
"grad_norm": 235.61985778808594,
"learning_rate": 9.996659064689985e-06,
"loss": 1.089,
"step": 81
},
{
"epoch": 0.062440510184656385,
"grad_norm": 343.52587890625,
"learning_rate": 9.996432650449186e-06,
"loss": 1.5558,
"step": 82
},
{
"epoch": 0.06320197982105463,
"grad_norm": 438.2514343261719,
"learning_rate": 9.996198816379273e-06,
"loss": 1.4458,
"step": 83
},
{
"epoch": 0.06396344945745289,
"grad_norm": 333.7451477050781,
"learning_rate": 9.99595756282749e-06,
"loss": 1.2887,
"step": 84
},
{
"epoch": 0.06472491909385113,
"grad_norm": 232.31222534179688,
"learning_rate": 9.99570889015211e-06,
"loss": 1.2663,
"step": 85
},
{
"epoch": 0.06548638873024938,
"grad_norm": 215.1619415283203,
"learning_rate": 9.995452798722416e-06,
"loss": 1.0614,
"step": 86
},
{
"epoch": 0.06624785836664764,
"grad_norm": 255.4922332763672,
"learning_rate": 9.99518928891871e-06,
"loss": 1.4337,
"step": 87
},
{
"epoch": 0.06700932800304588,
"grad_norm": 550.9943237304688,
"learning_rate": 9.99491836113231e-06,
"loss": 1.48,
"step": 88
},
{
"epoch": 0.06777079763944413,
"grad_norm": 288.2671203613281,
"learning_rate": 9.994640015765552e-06,
"loss": 1.2056,
"step": 89
},
{
"epoch": 0.06853226727584237,
"grad_norm": 231.81983947753906,
"learning_rate": 9.994354253231783e-06,
"loss": 1.0956,
"step": 90
},
{
"epoch": 0.06929373691224062,
"grad_norm": 417.0189514160156,
"learning_rate": 9.994061073955375e-06,
"loss": 1.0774,
"step": 91
},
{
"epoch": 0.07005520654863888,
"grad_norm": 219.20413208007812,
"learning_rate": 9.993760478371697e-06,
"loss": 1.3335,
"step": 92
},
{
"epoch": 0.07081667618503712,
"grad_norm": 282.10443115234375,
"learning_rate": 9.99345246692715e-06,
"loss": 1.3221,
"step": 93
},
{
"epoch": 0.07157814582143537,
"grad_norm": 213.92098999023438,
"learning_rate": 9.993137040079131e-06,
"loss": 1.3243,
"step": 94
},
{
"epoch": 0.07233961545783361,
"grad_norm": 158.23342895507812,
"learning_rate": 9.992814198296061e-06,
"loss": 1.1767,
"step": 95
},
{
"epoch": 0.07310108509423187,
"grad_norm": 177.14779663085938,
"learning_rate": 9.992483942057369e-06,
"loss": 1.2758,
"step": 96
},
{
"epoch": 0.07386255473063012,
"grad_norm": 254.54237365722656,
"learning_rate": 9.992146271853491e-06,
"loss": 1.2782,
"step": 97
},
{
"epoch": 0.07462402436702836,
"grad_norm": 202.73121643066406,
"learning_rate": 9.991801188185878e-06,
"loss": 1.2328,
"step": 98
},
{
"epoch": 0.07538549400342662,
"grad_norm": 198.03977966308594,
"learning_rate": 9.991448691566989e-06,
"loss": 0.9788,
"step": 99
},
{
"epoch": 0.07614696363982486,
"grad_norm": 203.24838256835938,
"learning_rate": 9.991088782520286e-06,
"loss": 1.4558,
"step": 100
},
{
"epoch": 0.07690843327622311,
"grad_norm": 160.49710083007812,
"learning_rate": 9.990721461580247e-06,
"loss": 0.9447,
"step": 101
},
{
"epoch": 0.07766990291262135,
"grad_norm": 162.8723907470703,
"learning_rate": 9.990346729292353e-06,
"loss": 1.0596,
"step": 102
},
{
"epoch": 0.0784313725490196,
"grad_norm": 218.8311767578125,
"learning_rate": 9.989964586213087e-06,
"loss": 1.4439,
"step": 103
},
{
"epoch": 0.07919284218541786,
"grad_norm": 243.22412109375,
"learning_rate": 9.989575032909946e-06,
"loss": 1.2728,
"step": 104
},
{
"epoch": 0.0799543118218161,
"grad_norm": 235.75650024414062,
"learning_rate": 9.989178069961422e-06,
"loss": 1.1584,
"step": 105
},
{
"epoch": 0.08071578145821436,
"grad_norm": 214.09156799316406,
"learning_rate": 9.98877369795702e-06,
"loss": 1.1591,
"step": 106
},
{
"epoch": 0.0814772510946126,
"grad_norm": 205.96751403808594,
"learning_rate": 9.988361917497238e-06,
"loss": 1.114,
"step": 107
},
{
"epoch": 0.08223872073101085,
"grad_norm": 230.2100830078125,
"learning_rate": 9.987942729193584e-06,
"loss": 1.4029,
"step": 108
},
{
"epoch": 0.0830001903674091,
"grad_norm": 246.6984405517578,
"learning_rate": 9.987516133668561e-06,
"loss": 1.5481,
"step": 109
},
{
"epoch": 0.08376166000380734,
"grad_norm": 313.53570556640625,
"learning_rate": 9.987082131555674e-06,
"loss": 0.9881,
"step": 110
},
{
"epoch": 0.0845231296402056,
"grad_norm": 190.1482391357422,
"learning_rate": 9.98664072349943e-06,
"loss": 1.0438,
"step": 111
},
{
"epoch": 0.08528459927660384,
"grad_norm": 230.99050903320312,
"learning_rate": 9.986191910155328e-06,
"loss": 1.3698,
"step": 112
},
{
"epoch": 0.0860460689130021,
"grad_norm": 224.8727569580078,
"learning_rate": 9.985735692189869e-06,
"loss": 1.21,
"step": 113
},
{
"epoch": 0.08680753854940035,
"grad_norm": 193.0661163330078,
"learning_rate": 9.985272070280545e-06,
"loss": 1.3077,
"step": 114
},
{
"epoch": 0.08756900818579859,
"grad_norm": 156.9528045654297,
"learning_rate": 9.984801045115852e-06,
"loss": 0.9841,
"step": 115
},
{
"epoch": 0.08833047782219684,
"grad_norm": 245.15037536621094,
"learning_rate": 9.98432261739527e-06,
"loss": 1.2538,
"step": 116
},
{
"epoch": 0.08909194745859508,
"grad_norm": 278.6109313964844,
"learning_rate": 9.983836787829277e-06,
"loss": 1.0621,
"step": 117
},
{
"epoch": 0.08985341709499334,
"grad_norm": 313.7957763671875,
"learning_rate": 9.983343557139344e-06,
"loss": 1.4762,
"step": 118
},
{
"epoch": 0.09061488673139159,
"grad_norm": 160.8702392578125,
"learning_rate": 9.98284292605793e-06,
"loss": 1.1753,
"step": 119
},
{
"epoch": 0.09137635636778983,
"grad_norm": 222.89654541015625,
"learning_rate": 9.982334895328483e-06,
"loss": 1.4998,
"step": 120
},
{
"epoch": 0.09213782600418809,
"grad_norm": 214.5303955078125,
"learning_rate": 9.981819465705443e-06,
"loss": 1.2822,
"step": 121
},
{
"epoch": 0.09289929564058633,
"grad_norm": 175.49874877929688,
"learning_rate": 9.981296637954238e-06,
"loss": 1.2615,
"step": 122
},
{
"epoch": 0.09366076527698458,
"grad_norm": 167.6100616455078,
"learning_rate": 9.980766412851277e-06,
"loss": 1.4586,
"step": 123
},
{
"epoch": 0.09442223491338284,
"grad_norm": 125.47764587402344,
"learning_rate": 9.98022879118396e-06,
"loss": 0.9281,
"step": 124
},
{
"epoch": 0.09518370454978108,
"grad_norm": 166.0843505859375,
"learning_rate": 9.979683773750668e-06,
"loss": 1.0866,
"step": 125
},
{
"epoch": 0.09594517418617933,
"grad_norm": 155.88449096679688,
"learning_rate": 9.979131361360766e-06,
"loss": 1.1467,
"step": 126
},
{
"epoch": 0.09670664382257757,
"grad_norm": 163.07427978515625,
"learning_rate": 9.9785715548346e-06,
"loss": 1.0676,
"step": 127
},
{
"epoch": 0.09746811345897582,
"grad_norm": 156.6132049560547,
"learning_rate": 9.978004355003498e-06,
"loss": 1.2959,
"step": 128
},
{
"epoch": 0.09822958309537407,
"grad_norm": 193.43089294433594,
"learning_rate": 9.977429762709765e-06,
"loss": 1.1902,
"step": 129
},
{
"epoch": 0.09899105273177232,
"grad_norm": 142.8735809326172,
"learning_rate": 9.976847778806684e-06,
"loss": 1.0323,
"step": 130
},
{
"epoch": 0.09975252236817057,
"grad_norm": 154.01123046875,
"learning_rate": 9.976258404158518e-06,
"loss": 0.9512,
"step": 131
},
{
"epoch": 0.10051399200456881,
"grad_norm": 183.3456573486328,
"learning_rate": 9.975661639640503e-06,
"loss": 1.0477,
"step": 132
},
{
"epoch": 0.10127546164096707,
"grad_norm": 191.70790100097656,
"learning_rate": 9.975057486138849e-06,
"loss": 1.2902,
"step": 133
},
{
"epoch": 0.10203693127736531,
"grad_norm": 156.32797241210938,
"learning_rate": 9.97444594455074e-06,
"loss": 1.1891,
"step": 134
},
{
"epoch": 0.10279840091376356,
"grad_norm": 117.63006591796875,
"learning_rate": 9.973827015784328e-06,
"loss": 0.8506,
"step": 135
},
{
"epoch": 0.10355987055016182,
"grad_norm": 125.6243667602539,
"learning_rate": 9.973200700758741e-06,
"loss": 1.1157,
"step": 136
},
{
"epoch": 0.10432134018656006,
"grad_norm": 143.13687133789062,
"learning_rate": 9.972567000404072e-06,
"loss": 1.3107,
"step": 137
},
{
"epoch": 0.10508280982295831,
"grad_norm": 136.7623291015625,
"learning_rate": 9.97192591566138e-06,
"loss": 0.7335,
"step": 138
},
{
"epoch": 0.10584427945935655,
"grad_norm": 169.33311462402344,
"learning_rate": 9.971277447482696e-06,
"loss": 1.3181,
"step": 139
},
{
"epoch": 0.1066057490957548,
"grad_norm": 143.11961364746094,
"learning_rate": 9.970621596831007e-06,
"loss": 0.6787,
"step": 140
},
{
"epoch": 0.10736721873215306,
"grad_norm": 122.05813598632812,
"learning_rate": 9.969958364680272e-06,
"loss": 1.0809,
"step": 141
},
{
"epoch": 0.1081286883685513,
"grad_norm": 120.5428695678711,
"learning_rate": 9.969287752015406e-06,
"loss": 1.0224,
"step": 142
},
{
"epoch": 0.10889015800494956,
"grad_norm": 105.74716186523438,
"learning_rate": 9.968609759832284e-06,
"loss": 0.9593,
"step": 143
},
{
"epoch": 0.1096516276413478,
"grad_norm": 199.15542602539062,
"learning_rate": 9.967924389137745e-06,
"loss": 1.0812,
"step": 144
},
{
"epoch": 0.11041309727774605,
"grad_norm": 143.60145568847656,
"learning_rate": 9.967231640949579e-06,
"loss": 1.0824,
"step": 145
},
{
"epoch": 0.1111745669141443,
"grad_norm": 117.3849868774414,
"learning_rate": 9.966531516296536e-06,
"loss": 1.0903,
"step": 146
},
{
"epoch": 0.11193603655054254,
"grad_norm": 136.92623901367188,
"learning_rate": 9.965824016218317e-06,
"loss": 1.1075,
"step": 147
},
{
"epoch": 0.1126975061869408,
"grad_norm": 125.6374740600586,
"learning_rate": 9.96510914176558e-06,
"loss": 0.9472,
"step": 148
},
{
"epoch": 0.11345897582333904,
"grad_norm": 116.42830657958984,
"learning_rate": 9.96438689399993e-06,
"loss": 1.1378,
"step": 149
},
{
"epoch": 0.1142204454597373,
"grad_norm": 99.17476654052734,
"learning_rate": 9.963657273993928e-06,
"loss": 0.9665,
"step": 150
},
{
"epoch": 0.11498191509613555,
"grad_norm": 105.80054473876953,
"learning_rate": 9.96292028283107e-06,
"loss": 1.0024,
"step": 151
},
{
"epoch": 0.11574338473253379,
"grad_norm": 114.62339782714844,
"learning_rate": 9.962175921605812e-06,
"loss": 1.1685,
"step": 152
},
{
"epoch": 0.11650485436893204,
"grad_norm": 125.38658142089844,
"learning_rate": 9.961424191423547e-06,
"loss": 1.0197,
"step": 153
},
{
"epoch": 0.11726632400533028,
"grad_norm": 129.48753356933594,
"learning_rate": 9.960665093400616e-06,
"loss": 1.1579,
"step": 154
},
{
"epoch": 0.11802779364172854,
"grad_norm": 117.82019805908203,
"learning_rate": 9.959898628664298e-06,
"loss": 1.0146,
"step": 155
},
{
"epoch": 0.11878926327812679,
"grad_norm": 151.73013305664062,
"learning_rate": 9.95912479835281e-06,
"loss": 1.3081,
"step": 156
},
{
"epoch": 0.11955073291452503,
"grad_norm": 141.16104125976562,
"learning_rate": 9.958343603615311e-06,
"loss": 1.2722,
"step": 157
},
{
"epoch": 0.12031220255092329,
"grad_norm": 94.05651092529297,
"learning_rate": 9.957555045611897e-06,
"loss": 0.8151,
"step": 158
},
{
"epoch": 0.12107367218732153,
"grad_norm": 118.12832641601562,
"learning_rate": 9.956759125513593e-06,
"loss": 1.1914,
"step": 159
},
{
"epoch": 0.12183514182371978,
"grad_norm": 155.52194213867188,
"learning_rate": 9.955955844502365e-06,
"loss": 1.2095,
"step": 160
},
{
"epoch": 0.12259661146011802,
"grad_norm": 120.84684753417969,
"learning_rate": 9.955145203771101e-06,
"loss": 1.034,
"step": 161
},
{
"epoch": 0.12335808109651628,
"grad_norm": 200.08302307128906,
"learning_rate": 9.954327204523626e-06,
"loss": 1.0109,
"step": 162
},
{
"epoch": 0.12411955073291453,
"grad_norm": 91.24543762207031,
"learning_rate": 9.953501847974689e-06,
"loss": 0.889,
"step": 163
},
{
"epoch": 0.12488102036931277,
"grad_norm": 122.83014678955078,
"learning_rate": 9.952669135349963e-06,
"loss": 0.9426,
"step": 164
},
{
"epoch": 0.125642490005711,
"grad_norm": 116.14713287353516,
"learning_rate": 9.951829067886053e-06,
"loss": 0.9366,
"step": 165
},
{
"epoch": 0.12640395964210926,
"grad_norm": 100.40796661376953,
"learning_rate": 9.950981646830476e-06,
"loss": 0.9302,
"step": 166
},
{
"epoch": 0.12716542927850752,
"grad_norm": 123.98023223876953,
"learning_rate": 9.950126873441674e-06,
"loss": 1.2609,
"step": 167
},
{
"epoch": 0.12792689891490577,
"grad_norm": 90.9245834350586,
"learning_rate": 9.949264748989009e-06,
"loss": 0.6413,
"step": 168
},
{
"epoch": 0.12868836855130403,
"grad_norm": 121.35209655761719,
"learning_rate": 9.948395274752754e-06,
"loss": 0.9599,
"step": 169
},
{
"epoch": 0.12944983818770225,
"grad_norm": 95.39146423339844,
"learning_rate": 9.947518452024105e-06,
"loss": 1.0403,
"step": 170
},
{
"epoch": 0.1302113078241005,
"grad_norm": 112.46849060058594,
"learning_rate": 9.946634282105164e-06,
"loss": 1.1076,
"step": 171
},
{
"epoch": 0.13097277746049876,
"grad_norm": 119.87206268310547,
"learning_rate": 9.945742766308946e-06,
"loss": 0.7471,
"step": 172
},
{
"epoch": 0.13173424709689702,
"grad_norm": 159.3794708251953,
"learning_rate": 9.944843905959372e-06,
"loss": 1.0218,
"step": 173
},
{
"epoch": 0.13249571673329527,
"grad_norm": 143.2695770263672,
"learning_rate": 9.943937702391275e-06,
"loss": 1.1849,
"step": 174
},
{
"epoch": 0.1332571863696935,
"grad_norm": 129.3036346435547,
"learning_rate": 9.94302415695039e-06,
"loss": 1.0408,
"step": 175
},
{
"epoch": 0.13401865600609175,
"grad_norm": 112.15624237060547,
"learning_rate": 9.942103270993356e-06,
"loss": 0.9284,
"step": 176
},
{
"epoch": 0.13478012564249,
"grad_norm": 351.1947937011719,
"learning_rate": 9.94117504588771e-06,
"loss": 1.0657,
"step": 177
},
{
"epoch": 0.13554159527888826,
"grad_norm": 198.4166259765625,
"learning_rate": 9.940239483011892e-06,
"loss": 0.8037,
"step": 178
},
{
"epoch": 0.13630306491528651,
"grad_norm": 111.71124267578125,
"learning_rate": 9.939296583755233e-06,
"loss": 0.9574,
"step": 179
},
{
"epoch": 0.13706453455168474,
"grad_norm": 103.34873962402344,
"learning_rate": 9.938346349517966e-06,
"loss": 1.1653,
"step": 180
},
{
"epoch": 0.137826004188083,
"grad_norm": 118.38048553466797,
"learning_rate": 9.937388781711213e-06,
"loss": 1.0831,
"step": 181
},
{
"epoch": 0.13858747382448125,
"grad_norm": 362.5590515136719,
"learning_rate": 9.936423881756983e-06,
"loss": 1.0866,
"step": 182
},
{
"epoch": 0.1393489434608795,
"grad_norm": 262.0923156738281,
"learning_rate": 9.93545165108818e-06,
"loss": 0.896,
"step": 183
},
{
"epoch": 0.14011041309727776,
"grad_norm": 146.92994689941406,
"learning_rate": 9.934472091148588e-06,
"loss": 1.0639,
"step": 184
},
{
"epoch": 0.14087188273367598,
"grad_norm": 94.61048126220703,
"learning_rate": 9.933485203392881e-06,
"loss": 0.9745,
"step": 185
},
{
"epoch": 0.14163335237007424,
"grad_norm": 109.75193786621094,
"learning_rate": 9.932490989286614e-06,
"loss": 1.0192,
"step": 186
},
{
"epoch": 0.1423948220064725,
"grad_norm": 100.21412658691406,
"learning_rate": 9.931489450306216e-06,
"loss": 1.1705,
"step": 187
},
{
"epoch": 0.14315629164287075,
"grad_norm": 103.8467788696289,
"learning_rate": 9.930480587939e-06,
"loss": 0.8971,
"step": 188
},
{
"epoch": 0.143917761279269,
"grad_norm": 138.01211547851562,
"learning_rate": 9.929464403683152e-06,
"loss": 0.9786,
"step": 189
},
{
"epoch": 0.14467923091566723,
"grad_norm": 94.98514556884766,
"learning_rate": 9.92844089904773e-06,
"loss": 0.908,
"step": 190
},
{
"epoch": 0.14544070055206548,
"grad_norm": 77.64883422851562,
"learning_rate": 9.927410075552666e-06,
"loss": 0.8565,
"step": 191
},
{
"epoch": 0.14620217018846374,
"grad_norm": 87.6455078125,
"learning_rate": 9.92637193472876e-06,
"loss": 0.978,
"step": 192
},
{
"epoch": 0.146963639824862,
"grad_norm": 75.11997985839844,
"learning_rate": 9.925326478117673e-06,
"loss": 0.699,
"step": 193
},
{
"epoch": 0.14772510946126025,
"grad_norm": 83.82418823242188,
"learning_rate": 9.92427370727194e-06,
"loss": 0.8723,
"step": 194
},
{
"epoch": 0.14848657909765847,
"grad_norm": 118.23309326171875,
"learning_rate": 9.92321362375495e-06,
"loss": 1.1355,
"step": 195
},
{
"epoch": 0.14924804873405673,
"grad_norm": 80.46004486083984,
"learning_rate": 9.922146229140952e-06,
"loss": 0.9206,
"step": 196
},
{
"epoch": 0.15000951837045498,
"grad_norm": 114.65190887451172,
"learning_rate": 9.921071525015057e-06,
"loss": 1.0244,
"step": 197
},
{
"epoch": 0.15077098800685323,
"grad_norm": 98.35903930664062,
"learning_rate": 9.919989512973227e-06,
"loss": 1.2852,
"step": 198
},
{
"epoch": 0.1515324576432515,
"grad_norm": 97.59054565429688,
"learning_rate": 9.918900194622277e-06,
"loss": 0.9721,
"step": 199
},
{
"epoch": 0.15229392727964972,
"grad_norm": 109.65848541259766,
"learning_rate": 9.91780357157987e-06,
"loss": 1.1749,
"step": 200
},
{
"epoch": 0.15305539691604797,
"grad_norm": 85.1619644165039,
"learning_rate": 9.916699645474525e-06,
"loss": 0.9321,
"step": 201
},
{
"epoch": 0.15381686655244622,
"grad_norm": 95.05500030517578,
"learning_rate": 9.915588417945595e-06,
"loss": 0.9065,
"step": 202
},
{
"epoch": 0.15457833618884448,
"grad_norm": 100.80270385742188,
"learning_rate": 9.914469890643282e-06,
"loss": 1.1461,
"step": 203
},
{
"epoch": 0.1553398058252427,
"grad_norm": 92.37889862060547,
"learning_rate": 9.913344065228629e-06,
"loss": 0.9546,
"step": 204
},
{
"epoch": 0.15610127546164096,
"grad_norm": 103.6626968383789,
"learning_rate": 9.912210943373515e-06,
"loss": 1.0895,
"step": 205
},
{
"epoch": 0.1568627450980392,
"grad_norm": 102.84590148925781,
"learning_rate": 9.911070526760652e-06,
"loss": 1.1842,
"step": 206
},
{
"epoch": 0.15762421473443747,
"grad_norm": 85.6246566772461,
"learning_rate": 9.90992281708359e-06,
"loss": 0.8207,
"step": 207
},
{
"epoch": 0.15838568437083572,
"grad_norm": 100.24021911621094,
"learning_rate": 9.908767816046709e-06,
"loss": 1.0015,
"step": 208
},
{
"epoch": 0.15914715400723395,
"grad_norm": 61.58012008666992,
"learning_rate": 9.90760552536521e-06,
"loss": 0.8035,
"step": 209
},
{
"epoch": 0.1599086236436322,
"grad_norm": 71.4411392211914,
"learning_rate": 9.906435946765127e-06,
"loss": 0.9014,
"step": 210
},
{
"epoch": 0.16067009328003046,
"grad_norm": 115.7383041381836,
"learning_rate": 9.905259081983314e-06,
"loss": 1.0404,
"step": 211
},
{
"epoch": 0.1614315629164287,
"grad_norm": 76.10044860839844,
"learning_rate": 9.904074932767443e-06,
"loss": 0.9077,
"step": 212
},
{
"epoch": 0.16219303255282697,
"grad_norm": 86.42494201660156,
"learning_rate": 9.902883500876008e-06,
"loss": 1.0405,
"step": 213
},
{
"epoch": 0.1629545021892252,
"grad_norm": 69.16837310791016,
"learning_rate": 9.901684788078314e-06,
"loss": 0.7598,
"step": 214
},
{
"epoch": 0.16371597182562345,
"grad_norm": 72.36968994140625,
"learning_rate": 9.90047879615448e-06,
"loss": 0.9537,
"step": 215
},
{
"epoch": 0.1644774414620217,
"grad_norm": 63.72938919067383,
"learning_rate": 9.899265526895437e-06,
"loss": 0.6775,
"step": 216
},
{
"epoch": 0.16523891109841995,
"grad_norm": 112.95844268798828,
"learning_rate": 9.898044982102916e-06,
"loss": 1.349,
"step": 217
},
{
"epoch": 0.1660003807348182,
"grad_norm": 112.42012786865234,
"learning_rate": 9.896817163589462e-06,
"loss": 1.3499,
"step": 218
},
{
"epoch": 0.16676185037121644,
"grad_norm": 58.411903381347656,
"learning_rate": 9.895582073178412e-06,
"loss": 0.7196,
"step": 219
},
{
"epoch": 0.1675233200076147,
"grad_norm": 78.74752044677734,
"learning_rate": 9.894339712703911e-06,
"loss": 1.0564,
"step": 220
},
{
"epoch": 0.16828478964401294,
"grad_norm": 68.431884765625,
"learning_rate": 9.89309008401089e-06,
"loss": 0.8719,
"step": 221
},
{
"epoch": 0.1690462592804112,
"grad_norm": 68.58363342285156,
"learning_rate": 9.891833188955086e-06,
"loss": 0.9706,
"step": 222
},
{
"epoch": 0.16980772891680945,
"grad_norm": 73.43843841552734,
"learning_rate": 9.890569029403016e-06,
"loss": 0.8411,
"step": 223
},
{
"epoch": 0.17056919855320768,
"grad_norm": 79.31640625,
"learning_rate": 9.88929760723199e-06,
"loss": 0.9633,
"step": 224
},
{
"epoch": 0.17133066818960593,
"grad_norm": 66.7998275756836,
"learning_rate": 9.888018924330101e-06,
"loss": 0.8297,
"step": 225
},
{
"epoch": 0.1720921378260042,
"grad_norm": 110.90506744384766,
"learning_rate": 9.886732982596226e-06,
"loss": 0.9995,
"step": 226
},
{
"epoch": 0.17285360746240244,
"grad_norm": 66.6268310546875,
"learning_rate": 9.885439783940023e-06,
"loss": 0.83,
"step": 227
},
{
"epoch": 0.1736150770988007,
"grad_norm": 76.32046508789062,
"learning_rate": 9.884139330281921e-06,
"loss": 1.0362,
"step": 228
},
{
"epoch": 0.17437654673519892,
"grad_norm": 76.37635040283203,
"learning_rate": 9.882831623553133e-06,
"loss": 1.087,
"step": 229
},
{
"epoch": 0.17513801637159718,
"grad_norm": 57.767845153808594,
"learning_rate": 9.881516665695628e-06,
"loss": 0.8016,
"step": 230
},
{
"epoch": 0.17589948600799543,
"grad_norm": 78.45353698730469,
"learning_rate": 9.880194458662159e-06,
"loss": 0.8224,
"step": 231
},
{
"epoch": 0.17666095564439369,
"grad_norm": 73.52896118164062,
"learning_rate": 9.878865004416234e-06,
"loss": 0.8957,
"step": 232
},
{
"epoch": 0.17742242528079194,
"grad_norm": 83.69878387451172,
"learning_rate": 9.877528304932125e-06,
"loss": 1.1008,
"step": 233
},
{
"epoch": 0.17818389491719017,
"grad_norm": 86.45205688476562,
"learning_rate": 9.876184362194868e-06,
"loss": 0.8303,
"step": 234
},
{
"epoch": 0.17894536455358842,
"grad_norm": 99.48762512207031,
"learning_rate": 9.87483317820025e-06,
"loss": 1.2673,
"step": 235
},
{
"epoch": 0.17970683418998668,
"grad_norm": 73.71369171142578,
"learning_rate": 9.873474754954813e-06,
"loss": 1.0541,
"step": 236
},
{
"epoch": 0.18046830382638493,
"grad_norm": 60.825538635253906,
"learning_rate": 9.872109094475849e-06,
"loss": 0.7781,
"step": 237
},
{
"epoch": 0.18122977346278318,
"grad_norm": 93.30781555175781,
"learning_rate": 9.870736198791401e-06,
"loss": 1.0317,
"step": 238
},
{
"epoch": 0.1819912430991814,
"grad_norm": 86.9531478881836,
"learning_rate": 9.869356069940255e-06,
"loss": 1.0528,
"step": 239
},
{
"epoch": 0.18275271273557966,
"grad_norm": 67.83726501464844,
"learning_rate": 9.867968709971932e-06,
"loss": 0.8781,
"step": 240
},
{
"epoch": 0.18351418237197792,
"grad_norm": 72.33470153808594,
"learning_rate": 9.866574120946701e-06,
"loss": 1.0262,
"step": 241
},
{
"epoch": 0.18427565200837617,
"grad_norm": 72.22496795654297,
"learning_rate": 9.865172304935563e-06,
"loss": 1.0208,
"step": 242
},
{
"epoch": 0.18503712164477443,
"grad_norm": 59.84956359863281,
"learning_rate": 9.863763264020245e-06,
"loss": 0.6551,
"step": 243
},
{
"epoch": 0.18579859128117265,
"grad_norm": 65.48404693603516,
"learning_rate": 9.862347000293215e-06,
"loss": 0.7664,
"step": 244
},
{
"epoch": 0.1865600609175709,
"grad_norm": 85.38325500488281,
"learning_rate": 9.860923515857654e-06,
"loss": 0.8044,
"step": 245
},
{
"epoch": 0.18732153055396916,
"grad_norm": 72.45954132080078,
"learning_rate": 9.859492812827477e-06,
"loss": 0.986,
"step": 246
},
{
"epoch": 0.18808300019036742,
"grad_norm": 75.24809265136719,
"learning_rate": 9.858054893327313e-06,
"loss": 0.8028,
"step": 247
},
{
"epoch": 0.18884446982676567,
"grad_norm": 79.74099731445312,
"learning_rate": 9.85660975949251e-06,
"loss": 1.0041,
"step": 248
},
{
"epoch": 0.1896059394631639,
"grad_norm": 66.91632843017578,
"learning_rate": 9.855157413469127e-06,
"loss": 0.851,
"step": 249
},
{
"epoch": 0.19036740909956215,
"grad_norm": 66.16661834716797,
"learning_rate": 9.853697857413933e-06,
"loss": 0.9538,
"step": 250
},
{
"epoch": 0.1911288787359604,
"grad_norm": 70.9784927368164,
"learning_rate": 9.85223109349441e-06,
"loss": 1.0546,
"step": 251
},
{
"epoch": 0.19189034837235866,
"grad_norm": 80.85004425048828,
"learning_rate": 9.850757123888736e-06,
"loss": 1.0673,
"step": 252
},
{
"epoch": 0.19265181800875691,
"grad_norm": 70.5213394165039,
"learning_rate": 9.849275950785795e-06,
"loss": 0.7249,
"step": 253
},
{
"epoch": 0.19341328764515514,
"grad_norm": 70.91667938232422,
"learning_rate": 9.847787576385165e-06,
"loss": 1.0218,
"step": 254
},
{
"epoch": 0.1941747572815534,
"grad_norm": 80.40878295898438,
"learning_rate": 9.846292002897123e-06,
"loss": 0.885,
"step": 255
},
{
"epoch": 0.19493622691795165,
"grad_norm": 64.33355712890625,
"learning_rate": 9.844789232542629e-06,
"loss": 0.8721,
"step": 256
},
{
"epoch": 0.1956976965543499,
"grad_norm": 64.21064758300781,
"learning_rate": 9.843279267553338e-06,
"loss": 0.8722,
"step": 257
},
{
"epoch": 0.19645916619074813,
"grad_norm": 228.57568359375,
"learning_rate": 9.841762110171584e-06,
"loss": 1.4434,
"step": 258
},
{
"epoch": 0.19722063582714638,
"grad_norm": 56.00699234008789,
"learning_rate": 9.840237762650389e-06,
"loss": 0.8607,
"step": 259
},
{
"epoch": 0.19798210546354464,
"grad_norm": 50.16679763793945,
"learning_rate": 9.838706227253443e-06,
"loss": 0.749,
"step": 260
},
{
"epoch": 0.1987435750999429,
"grad_norm": 80.30974578857422,
"learning_rate": 9.837167506255115e-06,
"loss": 0.9376,
"step": 261
},
{
"epoch": 0.19950504473634115,
"grad_norm": 67.49705505371094,
"learning_rate": 9.835621601940449e-06,
"loss": 0.996,
"step": 262
},
{
"epoch": 0.20026651437273937,
"grad_norm": 87.47010040283203,
"learning_rate": 9.834068516605146e-06,
"loss": 1.2215,
"step": 263
},
{
"epoch": 0.20102798400913763,
"grad_norm": 76.85598754882812,
"learning_rate": 9.832508252555582e-06,
"loss": 0.9047,
"step": 264
},
{
"epoch": 0.20178945364553588,
"grad_norm": 72.64045715332031,
"learning_rate": 9.830940812108787e-06,
"loss": 0.9827,
"step": 265
},
{
"epoch": 0.20255092328193414,
"grad_norm": 59.62710189819336,
"learning_rate": 9.829366197592449e-06,
"loss": 0.8007,
"step": 266
},
{
"epoch": 0.2033123929183324,
"grad_norm": 71.46318817138672,
"learning_rate": 9.827784411344912e-06,
"loss": 0.8264,
"step": 267
},
{
"epoch": 0.20407386255473062,
"grad_norm": 91.32540130615234,
"learning_rate": 9.826195455715166e-06,
"loss": 0.8318,
"step": 268
},
{
"epoch": 0.20483533219112887,
"grad_norm": 107.7885513305664,
"learning_rate": 9.824599333062855e-06,
"loss": 1.2069,
"step": 269
},
{
"epoch": 0.20559680182752713,
"grad_norm": 78.61428833007812,
"learning_rate": 9.822996045758256e-06,
"loss": 0.84,
"step": 270
},
{
"epoch": 0.20635827146392538,
"grad_norm": 83.16557312011719,
"learning_rate": 9.821385596182296e-06,
"loss": 1.0659,
"step": 271
},
{
"epoch": 0.20711974110032363,
"grad_norm": 69.2863540649414,
"learning_rate": 9.819767986726529e-06,
"loss": 0.802,
"step": 272
},
{
"epoch": 0.20788121073672186,
"grad_norm": 72.9874038696289,
"learning_rate": 9.818143219793152e-06,
"loss": 1.0752,
"step": 273
},
{
"epoch": 0.20864268037312012,
"grad_norm": 59.5941162109375,
"learning_rate": 9.81651129779498e-06,
"loss": 0.7396,
"step": 274
},
{
"epoch": 0.20940415000951837,
"grad_norm": 63.05308151245117,
"learning_rate": 9.81487222315546e-06,
"loss": 0.8281,
"step": 275
},
{
"epoch": 0.21016561964591662,
"grad_norm": 72.18460083007812,
"learning_rate": 9.813225998308659e-06,
"loss": 0.8861,
"step": 276
},
{
"epoch": 0.21092708928231488,
"grad_norm": 66.82262420654297,
"learning_rate": 9.811572625699263e-06,
"loss": 0.7964,
"step": 277
},
{
"epoch": 0.2116885589187131,
"grad_norm": 64.8218002319336,
"learning_rate": 9.809912107782573e-06,
"loss": 1.0174,
"step": 278
},
{
"epoch": 0.21245002855511136,
"grad_norm": 73.7430648803711,
"learning_rate": 9.8082444470245e-06,
"loss": 1.1713,
"step": 279
},
{
"epoch": 0.2132114981915096,
"grad_norm": 67.32575988769531,
"learning_rate": 9.806569645901561e-06,
"loss": 1.0453,
"step": 280
},
{
"epoch": 0.21397296782790787,
"grad_norm": 58.80308532714844,
"learning_rate": 9.804887706900881e-06,
"loss": 0.8187,
"step": 281
},
{
"epoch": 0.21473443746430612,
"grad_norm": 100.3160171508789,
"learning_rate": 9.803198632520179e-06,
"loss": 0.9593,
"step": 282
},
{
"epoch": 0.21549590710070435,
"grad_norm": 64.89568328857422,
"learning_rate": 9.801502425267773e-06,
"loss": 0.7491,
"step": 283
},
{
"epoch": 0.2162573767371026,
"grad_norm": 67.52566528320312,
"learning_rate": 9.799799087662577e-06,
"loss": 0.8042,
"step": 284
},
{
"epoch": 0.21701884637350086,
"grad_norm": 51.88862228393555,
"learning_rate": 9.798088622234088e-06,
"loss": 0.8807,
"step": 285
},
{
"epoch": 0.2177803160098991,
"grad_norm": 70.29974365234375,
"learning_rate": 9.796371031522391e-06,
"loss": 1.0593,
"step": 286
},
{
"epoch": 0.21854178564629737,
"grad_norm": 60.2655143737793,
"learning_rate": 9.794646318078153e-06,
"loss": 0.8488,
"step": 287
},
{
"epoch": 0.2193032552826956,
"grad_norm": 78.09272766113281,
"learning_rate": 9.792914484462618e-06,
"loss": 0.9485,
"step": 288
},
{
"epoch": 0.22006472491909385,
"grad_norm": 55.70231628417969,
"learning_rate": 9.791175533247597e-06,
"loss": 0.8375,
"step": 289
},
{
"epoch": 0.2208261945554921,
"grad_norm": 47.878211975097656,
"learning_rate": 9.789429467015484e-06,
"loss": 0.8201,
"step": 290
},
{
"epoch": 0.22158766419189035,
"grad_norm": 59.19154357910156,
"learning_rate": 9.787676288359225e-06,
"loss": 0.9049,
"step": 291
},
{
"epoch": 0.2223491338282886,
"grad_norm": 79.76387786865234,
"learning_rate": 9.78591599988234e-06,
"loss": 0.8522,
"step": 292
},
{
"epoch": 0.22311060346468684,
"grad_norm": 58.5650749206543,
"learning_rate": 9.7841486041989e-06,
"loss": 0.8223,
"step": 293
},
{
"epoch": 0.2238720731010851,
"grad_norm": 54.914093017578125,
"learning_rate": 9.782374103933531e-06,
"loss": 0.7369,
"step": 294
},
{
"epoch": 0.22463354273748334,
"grad_norm": 56.13349151611328,
"learning_rate": 9.780592501721414e-06,
"loss": 0.9265,
"step": 295
},
{
"epoch": 0.2253950123738816,
"grad_norm": 52.732635498046875,
"learning_rate": 9.77880380020827e-06,
"loss": 0.7705,
"step": 296
},
{
"epoch": 0.22615648201027985,
"grad_norm": 54.85570526123047,
"learning_rate": 9.777008002050367e-06,
"loss": 1.1053,
"step": 297
},
{
"epoch": 0.22691795164667808,
"grad_norm": 48.77083206176758,
"learning_rate": 9.775205109914513e-06,
"loss": 0.8562,
"step": 298
},
{
"epoch": 0.22767942128307633,
"grad_norm": 48.549400329589844,
"learning_rate": 9.773395126478047e-06,
"loss": 0.7249,
"step": 299
},
{
"epoch": 0.2284408909194746,
"grad_norm": 60.0245361328125,
"learning_rate": 9.771578054428841e-06,
"loss": 0.7139,
"step": 300
},
{
"epoch": 0.22920236055587284,
"grad_norm": 55.8681755065918,
"learning_rate": 9.769753896465291e-06,
"loss": 0.7965,
"step": 301
},
{
"epoch": 0.2299638301922711,
"grad_norm": 63.89664840698242,
"learning_rate": 9.76792265529632e-06,
"loss": 0.8464,
"step": 302
},
{
"epoch": 0.23072529982866932,
"grad_norm": 71.57247161865234,
"learning_rate": 9.766084333641366e-06,
"loss": 1.0882,
"step": 303
},
{
"epoch": 0.23148676946506758,
"grad_norm": 79.77525329589844,
"learning_rate": 9.764238934230384e-06,
"loss": 1.0725,
"step": 304
},
{
"epoch": 0.23224823910146583,
"grad_norm": 67.99383544921875,
"learning_rate": 9.76238645980384e-06,
"loss": 0.8869,
"step": 305
},
{
"epoch": 0.23300970873786409,
"grad_norm": 65.60368347167969,
"learning_rate": 9.760526913112702e-06,
"loss": 1.055,
"step": 306
},
{
"epoch": 0.23377117837426234,
"grad_norm": 65.8590087890625,
"learning_rate": 9.758660296918447e-06,
"loss": 0.8672,
"step": 307
},
{
"epoch": 0.23453264801066057,
"grad_norm": 59.94193649291992,
"learning_rate": 9.756786613993048e-06,
"loss": 0.8927,
"step": 308
},
{
"epoch": 0.23529411764705882,
"grad_norm": 59.51659393310547,
"learning_rate": 9.754905867118969e-06,
"loss": 0.9401,
"step": 309
},
{
"epoch": 0.23605558728345707,
"grad_norm": 58.14861297607422,
"learning_rate": 9.753018059089168e-06,
"loss": 0.795,
"step": 310
},
{
"epoch": 0.23681705691985533,
"grad_norm": 54.271827697753906,
"learning_rate": 9.751123192707089e-06,
"loss": 0.8585,
"step": 311
},
{
"epoch": 0.23757852655625358,
"grad_norm": 63.290191650390625,
"learning_rate": 9.749221270786656e-06,
"loss": 0.9526,
"step": 312
},
{
"epoch": 0.2383399961926518,
"grad_norm": 48.89083480834961,
"learning_rate": 9.747312296152268e-06,
"loss": 0.8029,
"step": 313
},
{
"epoch": 0.23910146582905006,
"grad_norm": 53.1309928894043,
"learning_rate": 9.745396271638805e-06,
"loss": 0.8306,
"step": 314
},
{
"epoch": 0.23986293546544832,
"grad_norm": 53.48817443847656,
"learning_rate": 9.74347320009161e-06,
"loss": 0.8496,
"step": 315
},
{
"epoch": 0.24062440510184657,
"grad_norm": 60.37258529663086,
"learning_rate": 9.741543084366493e-06,
"loss": 0.9068,
"step": 316
},
{
"epoch": 0.2413858747382448,
"grad_norm": 68.24382781982422,
"learning_rate": 9.739605927329725e-06,
"loss": 1.0776,
"step": 317
},
{
"epoch": 0.24214734437464305,
"grad_norm": 44.461395263671875,
"learning_rate": 9.737661731858035e-06,
"loss": 0.7072,
"step": 318
},
{
"epoch": 0.2429088140110413,
"grad_norm": 50.15453338623047,
"learning_rate": 9.7357105008386e-06,
"loss": 0.7752,
"step": 319
},
{
"epoch": 0.24367028364743956,
"grad_norm": 57.21116256713867,
"learning_rate": 9.733752237169048e-06,
"loss": 0.7657,
"step": 320
},
{
"epoch": 0.24443175328383782,
"grad_norm": 53.94674301147461,
"learning_rate": 9.731786943757451e-06,
"loss": 0.8257,
"step": 321
},
{
"epoch": 0.24519322292023604,
"grad_norm": 67.50530242919922,
"learning_rate": 9.729814623522318e-06,
"loss": 1.0285,
"step": 322
},
{
"epoch": 0.2459546925566343,
"grad_norm": 61.62517166137695,
"learning_rate": 9.727835279392597e-06,
"loss": 0.8321,
"step": 323
},
{
"epoch": 0.24671616219303255,
"grad_norm": 76.12077331542969,
"learning_rate": 9.725848914307662e-06,
"loss": 1.0304,
"step": 324
},
{
"epoch": 0.2474776318294308,
"grad_norm": 58.61418151855469,
"learning_rate": 9.723855531217316e-06,
"loss": 0.9892,
"step": 325
},
{
"epoch": 0.24823910146582906,
"grad_norm": 51.80055618286133,
"learning_rate": 9.721855133081786e-06,
"loss": 0.7943,
"step": 326
},
{
"epoch": 0.24900057110222729,
"grad_norm": 50.69115447998047,
"learning_rate": 9.71984772287171e-06,
"loss": 0.8635,
"step": 327
},
{
"epoch": 0.24976204073862554,
"grad_norm": 65.22866821289062,
"learning_rate": 9.717833303568145e-06,
"loss": 1.1264,
"step": 328
},
{
"epoch": 0.2505235103750238,
"grad_norm": 64.84595489501953,
"learning_rate": 9.715811878162556e-06,
"loss": 1.1264,
"step": 329
},
{
"epoch": 0.251284980011422,
"grad_norm": 47.65447235107422,
"learning_rate": 9.713783449656807e-06,
"loss": 0.823,
"step": 330
},
{
"epoch": 0.2520464496478203,
"grad_norm": 46.5490608215332,
"learning_rate": 9.71174802106317e-06,
"loss": 0.8889,
"step": 331
},
{
"epoch": 0.25280791928421853,
"grad_norm": 45.66133499145508,
"learning_rate": 9.709705595404304e-06,
"loss": 0.726,
"step": 332
},
{
"epoch": 0.2535693889206168,
"grad_norm": 58.741180419921875,
"learning_rate": 9.707656175713266e-06,
"loss": 0.9001,
"step": 333
},
{
"epoch": 0.25433085855701504,
"grad_norm": 56.66584396362305,
"learning_rate": 9.705599765033498e-06,
"loss": 0.8316,
"step": 334
},
{
"epoch": 0.25509232819341326,
"grad_norm": 47.76698303222656,
"learning_rate": 9.703536366418816e-06,
"loss": 0.9308,
"step": 335
},
{
"epoch": 0.25585379782981155,
"grad_norm": 52.65715789794922,
"learning_rate": 9.701465982933423e-06,
"loss": 0.9306,
"step": 336
},
{
"epoch": 0.2566152674662098,
"grad_norm": 57.91140365600586,
"learning_rate": 9.699388617651891e-06,
"loss": 0.9152,
"step": 337
},
{
"epoch": 0.25737673710260806,
"grad_norm": 46.12555694580078,
"learning_rate": 9.697304273659161e-06,
"loss": 0.8603,
"step": 338
},
{
"epoch": 0.2581382067390063,
"grad_norm": 51.05317687988281,
"learning_rate": 9.695212954050532e-06,
"loss": 1.0602,
"step": 339
},
{
"epoch": 0.2588996763754045,
"grad_norm": 45.56175994873047,
"learning_rate": 9.693114661931673e-06,
"loss": 0.7266,
"step": 340
},
{
"epoch": 0.2596611460118028,
"grad_norm": 149.18678283691406,
"learning_rate": 9.691009400418595e-06,
"loss": 0.9806,
"step": 341
},
{
"epoch": 0.260422615648201,
"grad_norm": 45.493690490722656,
"learning_rate": 9.68889717263767e-06,
"loss": 0.9679,
"step": 342
},
{
"epoch": 0.2611840852845993,
"grad_norm": 66.79802703857422,
"learning_rate": 9.686777981725606e-06,
"loss": 0.9454,
"step": 343
},
{
"epoch": 0.2619455549209975,
"grad_norm": 75.52727508544922,
"learning_rate": 9.684651830829456e-06,
"loss": 0.9213,
"step": 344
},
{
"epoch": 0.26270702455739575,
"grad_norm": 54.42271423339844,
"learning_rate": 9.682518723106608e-06,
"loss": 0.9061,
"step": 345
},
{
"epoch": 0.26346849419379403,
"grad_norm": 84.50782775878906,
"learning_rate": 9.680378661724782e-06,
"loss": 0.7799,
"step": 346
},
{
"epoch": 0.26422996383019226,
"grad_norm": 56.192962646484375,
"learning_rate": 9.678231649862024e-06,
"loss": 0.7537,
"step": 347
},
{
"epoch": 0.26499143346659054,
"grad_norm": 62.09931182861328,
"learning_rate": 9.6760776907067e-06,
"loss": 1.0033,
"step": 348
},
{
"epoch": 0.26575290310298877,
"grad_norm": 48.45965576171875,
"learning_rate": 9.673916787457492e-06,
"loss": 0.845,
"step": 349
},
{
"epoch": 0.266514372739387,
"grad_norm": 54.92905044555664,
"learning_rate": 9.6717489433234e-06,
"loss": 0.7935,
"step": 350
},
{
"epoch": 0.2672758423757853,
"grad_norm": 44.289329528808594,
"learning_rate": 9.669574161523723e-06,
"loss": 0.7688,
"step": 351
},
{
"epoch": 0.2680373120121835,
"grad_norm": 44.33442306518555,
"learning_rate": 9.667392445288071e-06,
"loss": 0.7193,
"step": 352
},
{
"epoch": 0.2687987816485818,
"grad_norm": 51.590057373046875,
"learning_rate": 9.665203797856346e-06,
"loss": 0.7921,
"step": 353
},
{
"epoch": 0.26956025128498,
"grad_norm": 45.111812591552734,
"learning_rate": 9.663008222478747e-06,
"loss": 0.6665,
"step": 354
},
{
"epoch": 0.27032172092137824,
"grad_norm": 57.12602233886719,
"learning_rate": 9.660805722415755e-06,
"loss": 0.8825,
"step": 355
},
{
"epoch": 0.2710831905577765,
"grad_norm": 50.83316421508789,
"learning_rate": 9.658596300938143e-06,
"loss": 0.881,
"step": 356
},
{
"epoch": 0.27184466019417475,
"grad_norm": 52.06378936767578,
"learning_rate": 9.656379961326954e-06,
"loss": 0.9237,
"step": 357
},
{
"epoch": 0.27260612983057303,
"grad_norm": 46.128719329833984,
"learning_rate": 9.65415670687351e-06,
"loss": 0.7108,
"step": 358
},
{
"epoch": 0.27336759946697126,
"grad_norm": 40.89448547363281,
"learning_rate": 9.651926540879401e-06,
"loss": 0.7963,
"step": 359
},
{
"epoch": 0.2741290691033695,
"grad_norm": 43.682315826416016,
"learning_rate": 9.64968946665648e-06,
"loss": 0.7105,
"step": 360
},
{
"epoch": 0.27489053873976776,
"grad_norm": 45.00116729736328,
"learning_rate": 9.647445487526857e-06,
"loss": 0.676,
"step": 361
},
{
"epoch": 0.275652008376166,
"grad_norm": 60.66935348510742,
"learning_rate": 9.645194606822902e-06,
"loss": 0.9782,
"step": 362
},
{
"epoch": 0.2764134780125643,
"grad_norm": 62.74099349975586,
"learning_rate": 9.642936827887223e-06,
"loss": 0.8989,
"step": 363
},
{
"epoch": 0.2771749476489625,
"grad_norm": 49.292667388916016,
"learning_rate": 9.640672154072685e-06,
"loss": 0.8945,
"step": 364
},
{
"epoch": 0.2779364172853607,
"grad_norm": 45.407901763916016,
"learning_rate": 9.638400588742385e-06,
"loss": 0.7232,
"step": 365
},
{
"epoch": 0.278697886921759,
"grad_norm": 92.1402816772461,
"learning_rate": 9.636122135269655e-06,
"loss": 0.6411,
"step": 366
},
{
"epoch": 0.27945935655815723,
"grad_norm": 52.646549224853516,
"learning_rate": 9.633836797038052e-06,
"loss": 0.7967,
"step": 367
},
{
"epoch": 0.2802208261945555,
"grad_norm": 51.09405517578125,
"learning_rate": 9.631544577441367e-06,
"loss": 0.8869,
"step": 368
},
{
"epoch": 0.28098229583095374,
"grad_norm": 44.59373092651367,
"learning_rate": 9.6292454798836e-06,
"loss": 0.842,
"step": 369
},
{
"epoch": 0.28174376546735197,
"grad_norm": 62.891151428222656,
"learning_rate": 9.626939507778972e-06,
"loss": 1.2426,
"step": 370
},
{
"epoch": 0.28250523510375025,
"grad_norm": 47.443321228027344,
"learning_rate": 9.624626664551908e-06,
"loss": 0.8091,
"step": 371
},
{
"epoch": 0.2832667047401485,
"grad_norm": 66.12074279785156,
"learning_rate": 9.622306953637042e-06,
"loss": 0.9351,
"step": 372
},
{
"epoch": 0.28402817437654676,
"grad_norm": 63.217369079589844,
"learning_rate": 9.6199803784792e-06,
"loss": 0.8927,
"step": 373
},
{
"epoch": 0.284789644012945,
"grad_norm": 49.39320373535156,
"learning_rate": 9.617646942533403e-06,
"loss": 0.7507,
"step": 374
},
{
"epoch": 0.2855511136493432,
"grad_norm": 49.84722137451172,
"learning_rate": 9.61530664926487e-06,
"loss": 0.834,
"step": 375
},
{
"epoch": 0.2863125832857415,
"grad_norm": 46.706214904785156,
"learning_rate": 9.612959502148989e-06,
"loss": 0.7922,
"step": 376
},
{
"epoch": 0.2870740529221397,
"grad_norm": 46.903907775878906,
"learning_rate": 9.610605504671338e-06,
"loss": 0.9275,
"step": 377
},
{
"epoch": 0.287835522558538,
"grad_norm": 42.682334899902344,
"learning_rate": 9.608244660327657e-06,
"loss": 0.8216,
"step": 378
},
{
"epoch": 0.28859699219493623,
"grad_norm": 46.28004837036133,
"learning_rate": 9.605876972623866e-06,
"loss": 1.0026,
"step": 379
},
{
"epoch": 0.28935846183133446,
"grad_norm": 46.75234603881836,
"learning_rate": 9.603502445076038e-06,
"loss": 0.7993,
"step": 380
},
{
"epoch": 0.29011993146773274,
"grad_norm": 47.5782470703125,
"learning_rate": 9.601121081210405e-06,
"loss": 0.8264,
"step": 381
},
{
"epoch": 0.29088140110413097,
"grad_norm": 66.08570861816406,
"learning_rate": 9.598732884563357e-06,
"loss": 1.1345,
"step": 382
},
{
"epoch": 0.29164287074052925,
"grad_norm": 54.552268981933594,
"learning_rate": 9.596337858681423e-06,
"loss": 0.8592,
"step": 383
},
{
"epoch": 0.2924043403769275,
"grad_norm": 44.23638916015625,
"learning_rate": 9.593936007121279e-06,
"loss": 0.7281,
"step": 384
},
{
"epoch": 0.2931658100133257,
"grad_norm": 64.91778564453125,
"learning_rate": 9.591527333449734e-06,
"loss": 1.0959,
"step": 385
},
{
"epoch": 0.293927279649724,
"grad_norm": 50.69198989868164,
"learning_rate": 9.58911184124373e-06,
"loss": 0.8607,
"step": 386
},
{
"epoch": 0.2946887492861222,
"grad_norm": 44.16769790649414,
"learning_rate": 9.586689534090336e-06,
"loss": 0.8788,
"step": 387
},
{
"epoch": 0.2954502189225205,
"grad_norm": 46.0577507019043,
"learning_rate": 9.584260415586737e-06,
"loss": 0.7654,
"step": 388
},
{
"epoch": 0.2962116885589187,
"grad_norm": 53.36354446411133,
"learning_rate": 9.581824489340235e-06,
"loss": 0.6452,
"step": 389
},
{
"epoch": 0.29697315819531694,
"grad_norm": 50.50668716430664,
"learning_rate": 9.579381758968244e-06,
"loss": 0.961,
"step": 390
},
{
"epoch": 0.2977346278317152,
"grad_norm": 51.48319625854492,
"learning_rate": 9.576932228098282e-06,
"loss": 0.8882,
"step": 391
},
{
"epoch": 0.29849609746811345,
"grad_norm": 45.250450134277344,
"learning_rate": 9.574475900367962e-06,
"loss": 0.7654,
"step": 392
},
{
"epoch": 0.29925756710451173,
"grad_norm": 52.613136291503906,
"learning_rate": 9.572012779424994e-06,
"loss": 0.8605,
"step": 393
},
{
"epoch": 0.30001903674090996,
"grad_norm": 53.87302017211914,
"learning_rate": 9.569542868927174e-06,
"loss": 0.8858,
"step": 394
},
{
"epoch": 0.3007805063773082,
"grad_norm": 48.812095642089844,
"learning_rate": 9.56706617254238e-06,
"loss": 0.7843,
"step": 395
},
{
"epoch": 0.30154197601370647,
"grad_norm": 45.33589553833008,
"learning_rate": 9.564582693948575e-06,
"loss": 0.8224,
"step": 396
},
{
"epoch": 0.3023034456501047,
"grad_norm": 45.24856185913086,
"learning_rate": 9.56209243683378e-06,
"loss": 0.7123,
"step": 397
},
{
"epoch": 0.303064915286503,
"grad_norm": 50.84047317504883,
"learning_rate": 9.559595404896096e-06,
"loss": 0.8103,
"step": 398
},
{
"epoch": 0.3038263849229012,
"grad_norm": 52.10234832763672,
"learning_rate": 9.557091601843676e-06,
"loss": 0.9453,
"step": 399
},
{
"epoch": 0.30458785455929943,
"grad_norm": 44.31291961669922,
"learning_rate": 9.554581031394732e-06,
"loss": 0.89,
"step": 400
},
{
"epoch": 0.3053493241956977,
"grad_norm": 58.56898880004883,
"learning_rate": 9.552063697277522e-06,
"loss": 0.7098,
"step": 401
},
{
"epoch": 0.30611079383209594,
"grad_norm": 66.1932373046875,
"learning_rate": 9.549539603230355e-06,
"loss": 1.0265,
"step": 402
},
{
"epoch": 0.30687226346849417,
"grad_norm": 41.02724075317383,
"learning_rate": 9.547008753001572e-06,
"loss": 0.8262,
"step": 403
},
{
"epoch": 0.30763373310489245,
"grad_norm": 43.16440200805664,
"learning_rate": 9.544471150349551e-06,
"loss": 0.8475,
"step": 404
},
{
"epoch": 0.3083952027412907,
"grad_norm": 77.460693359375,
"learning_rate": 9.541926799042695e-06,
"loss": 1.1474,
"step": 405
},
{
"epoch": 0.30915667237768896,
"grad_norm": 53.00200271606445,
"learning_rate": 9.539375702859432e-06,
"loss": 0.6984,
"step": 406
},
{
"epoch": 0.3099181420140872,
"grad_norm": 80.07589721679688,
"learning_rate": 9.536817865588203e-06,
"loss": 0.7818,
"step": 407
},
{
"epoch": 0.3106796116504854,
"grad_norm": 61.049251556396484,
"learning_rate": 9.53425329102746e-06,
"loss": 0.9013,
"step": 408
},
{
"epoch": 0.3114410812868837,
"grad_norm": 59.325626373291016,
"learning_rate": 9.531681982985666e-06,
"loss": 0.9,
"step": 409
},
{
"epoch": 0.3122025509232819,
"grad_norm": 50.168907165527344,
"learning_rate": 9.529103945281273e-06,
"loss": 0.696,
"step": 410
},
{
"epoch": 0.3129640205596802,
"grad_norm": 98.40620422363281,
"learning_rate": 9.526519181742735e-06,
"loss": 0.7809,
"step": 411
},
{
"epoch": 0.3137254901960784,
"grad_norm": 39.94493103027344,
"learning_rate": 9.523927696208491e-06,
"loss": 0.8265,
"step": 412
},
{
"epoch": 0.31448695983247665,
"grad_norm": 37.19593048095703,
"learning_rate": 9.521329492526964e-06,
"loss": 0.5947,
"step": 413
},
{
"epoch": 0.31524842946887494,
"grad_norm": 52.14117431640625,
"learning_rate": 9.518724574556551e-06,
"loss": 0.9694,
"step": 414
},
{
"epoch": 0.31600989910527316,
"grad_norm": 45.72605514526367,
"learning_rate": 9.516112946165619e-06,
"loss": 0.752,
"step": 415
},
{
"epoch": 0.31677136874167144,
"grad_norm": 55.8684196472168,
"learning_rate": 9.513494611232507e-06,
"loss": 0.9199,
"step": 416
},
{
"epoch": 0.31753283837806967,
"grad_norm": 51.279266357421875,
"learning_rate": 9.510869573645506e-06,
"loss": 0.8778,
"step": 417
},
{
"epoch": 0.3182943080144679,
"grad_norm": 49.07886505126953,
"learning_rate": 9.508237837302862e-06,
"loss": 0.9251,
"step": 418
},
{
"epoch": 0.3190557776508662,
"grad_norm": 47.7920036315918,
"learning_rate": 9.505599406112774e-06,
"loss": 0.86,
"step": 419
},
{
"epoch": 0.3198172472872644,
"grad_norm": 46.17924880981445,
"learning_rate": 9.502954283993377e-06,
"loss": 0.874,
"step": 420
},
{
"epoch": 0.3205787169236627,
"grad_norm": 43.47404861450195,
"learning_rate": 9.500302474872742e-06,
"loss": 0.7282,
"step": 421
},
{
"epoch": 0.3213401865600609,
"grad_norm": 41.86683654785156,
"learning_rate": 9.497643982688878e-06,
"loss": 0.7389,
"step": 422
},
{
"epoch": 0.32210165619645914,
"grad_norm": 47.54405212402344,
"learning_rate": 9.49497881138971e-06,
"loss": 0.8291,
"step": 423
},
{
"epoch": 0.3228631258328574,
"grad_norm": 47.38523483276367,
"learning_rate": 9.492306964933085e-06,
"loss": 0.7763,
"step": 424
},
{
"epoch": 0.32362459546925565,
"grad_norm": 41.70780944824219,
"learning_rate": 9.489628447286762e-06,
"loss": 0.7601,
"step": 425
},
{
"epoch": 0.32438606510565393,
"grad_norm": 44.05202102661133,
"learning_rate": 9.48694326242841e-06,
"loss": 0.9465,
"step": 426
},
{
"epoch": 0.32514753474205216,
"grad_norm": 48.388675689697266,
"learning_rate": 9.484251414345595e-06,
"loss": 0.8215,
"step": 427
},
{
"epoch": 0.3259090043784504,
"grad_norm": 44.47600555419922,
"learning_rate": 9.48155290703578e-06,
"loss": 0.9731,
"step": 428
},
{
"epoch": 0.32667047401484867,
"grad_norm": 44.480506896972656,
"learning_rate": 9.478847744506318e-06,
"loss": 0.9429,
"step": 429
},
{
"epoch": 0.3274319436512469,
"grad_norm": 56.323486328125,
"learning_rate": 9.47613593077444e-06,
"loss": 1.0475,
"step": 430
},
{
"epoch": 0.3281934132876452,
"grad_norm": 54.73971939086914,
"learning_rate": 9.473417469867263e-06,
"loss": 0.973,
"step": 431
},
{
"epoch": 0.3289548829240434,
"grad_norm": 41.304107666015625,
"learning_rate": 9.470692365821767e-06,
"loss": 0.775,
"step": 432
},
{
"epoch": 0.32971635256044163,
"grad_norm": 43.46769714355469,
"learning_rate": 9.467960622684803e-06,
"loss": 0.8275,
"step": 433
},
{
"epoch": 0.3304778221968399,
"grad_norm": 39.45466232299805,
"learning_rate": 9.465222244513074e-06,
"loss": 0.973,
"step": 434
},
{
"epoch": 0.33123929183323814,
"grad_norm": 41.127464294433594,
"learning_rate": 9.462477235373148e-06,
"loss": 0.9163,
"step": 435
},
{
"epoch": 0.3320007614696364,
"grad_norm": 43.90326690673828,
"learning_rate": 9.459725599341427e-06,
"loss": 0.8066,
"step": 436
},
{
"epoch": 0.33276223110603464,
"grad_norm": 36.721168518066406,
"learning_rate": 9.456967340504164e-06,
"loss": 0.768,
"step": 437
},
{
"epoch": 0.33352370074243287,
"grad_norm": 40.407039642333984,
"learning_rate": 9.454202462957442e-06,
"loss": 0.79,
"step": 438
},
{
"epoch": 0.33428517037883115,
"grad_norm": 36.014339447021484,
"learning_rate": 9.451430970807174e-06,
"loss": 0.7859,
"step": 439
},
{
"epoch": 0.3350466400152294,
"grad_norm": 36.85624694824219,
"learning_rate": 9.448652868169098e-06,
"loss": 0.8155,
"step": 440
},
{
"epoch": 0.33580810965162766,
"grad_norm": 46.37972640991211,
"learning_rate": 9.445868159168763e-06,
"loss": 0.9041,
"step": 441
},
{
"epoch": 0.3365695792880259,
"grad_norm": 43.00399398803711,
"learning_rate": 9.443076847941535e-06,
"loss": 0.9208,
"step": 442
},
{
"epoch": 0.3373310489244241,
"grad_norm": 46.653106689453125,
"learning_rate": 9.440278938632583e-06,
"loss": 0.8929,
"step": 443
},
{
"epoch": 0.3380925185608224,
"grad_norm": 40.778255462646484,
"learning_rate": 9.43747443539687e-06,
"loss": 0.7525,
"step": 444
},
{
"epoch": 0.3388539881972206,
"grad_norm": 31.992992401123047,
"learning_rate": 9.434663342399157e-06,
"loss": 0.7598,
"step": 445
},
{
"epoch": 0.3396154578336189,
"grad_norm": 46.3204231262207,
"learning_rate": 9.431845663813988e-06,
"loss": 0.9856,
"step": 446
},
{
"epoch": 0.34037692747001713,
"grad_norm": 59.671539306640625,
"learning_rate": 9.429021403825685e-06,
"loss": 1.079,
"step": 447
},
{
"epoch": 0.34113839710641536,
"grad_norm": 65.77830505371094,
"learning_rate": 9.426190566628349e-06,
"loss": 0.8555,
"step": 448
},
{
"epoch": 0.34189986674281364,
"grad_norm": 48.69049835205078,
"learning_rate": 9.423353156425843e-06,
"loss": 0.7956,
"step": 449
},
{
"epoch": 0.34266133637921187,
"grad_norm": 44.31893539428711,
"learning_rate": 9.420509177431793e-06,
"loss": 0.828,
"step": 450
},
{
"epoch": 0.34342280601561015,
"grad_norm": 36.260643005371094,
"learning_rate": 9.417658633869579e-06,
"loss": 0.8325,
"step": 451
},
{
"epoch": 0.3441842756520084,
"grad_norm": 41.35783767700195,
"learning_rate": 9.414801529972333e-06,
"loss": 0.9002,
"step": 452
},
{
"epoch": 0.3449457452884066,
"grad_norm": 43.58637619018555,
"learning_rate": 9.411937869982925e-06,
"loss": 0.8771,
"step": 453
},
{
"epoch": 0.3457072149248049,
"grad_norm": 38.9900016784668,
"learning_rate": 9.409067658153964e-06,
"loss": 0.8082,
"step": 454
},
{
"epoch": 0.3464686845612031,
"grad_norm": 48.0144157409668,
"learning_rate": 9.406190898747786e-06,
"loss": 0.9711,
"step": 455
},
{
"epoch": 0.3472301541976014,
"grad_norm": 43.04066848754883,
"learning_rate": 9.40330759603645e-06,
"loss": 0.971,
"step": 456
},
{
"epoch": 0.3479916238339996,
"grad_norm": 44.24787902832031,
"learning_rate": 9.400417754301739e-06,
"loss": 0.9687,
"step": 457
},
{
"epoch": 0.34875309347039785,
"grad_norm": 45.766780853271484,
"learning_rate": 9.397521377835135e-06,
"loss": 0.9491,
"step": 458
},
{
"epoch": 0.34951456310679613,
"grad_norm": 34.62493896484375,
"learning_rate": 9.394618470937836e-06,
"loss": 0.7288,
"step": 459
},
{
"epoch": 0.35027603274319435,
"grad_norm": 49.47513961791992,
"learning_rate": 9.391709037920726e-06,
"loss": 0.8846,
"step": 460
},
{
"epoch": 0.35103750237959264,
"grad_norm": 40.401512145996094,
"learning_rate": 9.388793083104393e-06,
"loss": 0.8286,
"step": 461
},
{
"epoch": 0.35179897201599086,
"grad_norm": 44.801109313964844,
"learning_rate": 9.3858706108191e-06,
"loss": 0.8382,
"step": 462
},
{
"epoch": 0.3525604416523891,
"grad_norm": 36.950382232666016,
"learning_rate": 9.382941625404793e-06,
"loss": 0.7518,
"step": 463
},
{
"epoch": 0.35332191128878737,
"grad_norm": 46.91257858276367,
"learning_rate": 9.380006131211092e-06,
"loss": 0.9471,
"step": 464
},
{
"epoch": 0.3540833809251856,
"grad_norm": 43.36894989013672,
"learning_rate": 9.377064132597277e-06,
"loss": 0.8558,
"step": 465
},
{
"epoch": 0.3548448505615839,
"grad_norm": 53.6937370300293,
"learning_rate": 9.374115633932292e-06,
"loss": 0.9714,
"step": 466
},
{
"epoch": 0.3556063201979821,
"grad_norm": 49.51094055175781,
"learning_rate": 9.371160639594733e-06,
"loss": 0.9696,
"step": 467
},
{
"epoch": 0.35636778983438033,
"grad_norm": 70.25513458251953,
"learning_rate": 9.36819915397284e-06,
"loss": 1.4756,
"step": 468
},
{
"epoch": 0.3571292594707786,
"grad_norm": 41.462982177734375,
"learning_rate": 9.365231181464494e-06,
"loss": 0.7885,
"step": 469
},
{
"epoch": 0.35789072910717684,
"grad_norm": 52.57938766479492,
"learning_rate": 9.36225672647721e-06,
"loss": 1.0787,
"step": 470
},
{
"epoch": 0.3586521987435751,
"grad_norm": 31.510698318481445,
"learning_rate": 9.35927579342813e-06,
"loss": 0.8847,
"step": 471
},
{
"epoch": 0.35941366837997335,
"grad_norm": 44.495147705078125,
"learning_rate": 9.356288386744015e-06,
"loss": 0.8841,
"step": 472
},
{
"epoch": 0.3601751380163716,
"grad_norm": 35.463748931884766,
"learning_rate": 9.35329451086124e-06,
"loss": 0.9095,
"step": 473
},
{
"epoch": 0.36093660765276986,
"grad_norm": 40.00856399536133,
"learning_rate": 9.350294170225782e-06,
"loss": 0.7667,
"step": 474
},
{
"epoch": 0.3616980772891681,
"grad_norm": 46.01627731323242,
"learning_rate": 9.34728736929323e-06,
"loss": 0.922,
"step": 475
},
{
"epoch": 0.36245954692556637,
"grad_norm": 39.330169677734375,
"learning_rate": 9.344274112528754e-06,
"loss": 0.7242,
"step": 476
},
{
"epoch": 0.3632210165619646,
"grad_norm": 41.52873611450195,
"learning_rate": 9.34125440440712e-06,
"loss": 0.7228,
"step": 477
},
{
"epoch": 0.3639824861983628,
"grad_norm": 33.01096725463867,
"learning_rate": 9.338228249412668e-06,
"loss": 0.8528,
"step": 478
},
{
"epoch": 0.3647439558347611,
"grad_norm": 36.77853012084961,
"learning_rate": 9.33519565203932e-06,
"loss": 0.869,
"step": 479
},
{
"epoch": 0.36550542547115933,
"grad_norm": 44.876426696777344,
"learning_rate": 9.332156616790552e-06,
"loss": 1.0064,
"step": 480
},
{
"epoch": 0.3662668951075576,
"grad_norm": 45.61384963989258,
"learning_rate": 9.329111148179417e-06,
"loss": 0.8066,
"step": 481
},
{
"epoch": 0.36702836474395584,
"grad_norm": 42.6773796081543,
"learning_rate": 9.326059250728508e-06,
"loss": 0.839,
"step": 482
},
{
"epoch": 0.36778983438035406,
"grad_norm": 42.04776382446289,
"learning_rate": 9.32300092896997e-06,
"loss": 0.7902,
"step": 483
},
{
"epoch": 0.36855130401675235,
"grad_norm": 44.73598861694336,
"learning_rate": 9.319936187445491e-06,
"loss": 0.7914,
"step": 484
},
{
"epoch": 0.3693127736531506,
"grad_norm": 28.97296714782715,
"learning_rate": 9.316865030706288e-06,
"loss": 0.7196,
"step": 485
},
{
"epoch": 0.37007424328954885,
"grad_norm": 39.61836624145508,
"learning_rate": 9.313787463313107e-06,
"loss": 0.7217,
"step": 486
},
{
"epoch": 0.3708357129259471,
"grad_norm": 40.51829147338867,
"learning_rate": 9.310703489836211e-06,
"loss": 0.8674,
"step": 487
},
{
"epoch": 0.3715971825623453,
"grad_norm": 29.400951385498047,
"learning_rate": 9.307613114855383e-06,
"loss": 0.7207,
"step": 488
},
{
"epoch": 0.3723586521987436,
"grad_norm": 29.25424575805664,
"learning_rate": 9.304516342959908e-06,
"loss": 0.6184,
"step": 489
},
{
"epoch": 0.3731201218351418,
"grad_norm": 48.31173324584961,
"learning_rate": 9.301413178748568e-06,
"loss": 0.6037,
"step": 490
},
{
"epoch": 0.3738815914715401,
"grad_norm": 40.35416793823242,
"learning_rate": 9.298303626829641e-06,
"loss": 0.9935,
"step": 491
},
{
"epoch": 0.3746430611079383,
"grad_norm": 33.47250747680664,
"learning_rate": 9.29518769182089e-06,
"loss": 0.6797,
"step": 492
},
{
"epoch": 0.37540453074433655,
"grad_norm": 40.839599609375,
"learning_rate": 9.292065378349561e-06,
"loss": 0.8908,
"step": 493
},
{
"epoch": 0.37616600038073483,
"grad_norm": 99.43079376220703,
"learning_rate": 9.288936691052366e-06,
"loss": 0.9635,
"step": 494
},
{
"epoch": 0.37692747001713306,
"grad_norm": 37.33384704589844,
"learning_rate": 9.285801634575484e-06,
"loss": 0.9248,
"step": 495
},
{
"epoch": 0.37768893965353134,
"grad_norm": 43.419593811035156,
"learning_rate": 9.282660213574554e-06,
"loss": 0.9641,
"step": 496
},
{
"epoch": 0.37845040928992957,
"grad_norm": 37.27656555175781,
"learning_rate": 9.279512432714668e-06,
"loss": 0.826,
"step": 497
},
{
"epoch": 0.3792118789263278,
"grad_norm": 33.71250534057617,
"learning_rate": 9.276358296670357e-06,
"loss": 0.6497,
"step": 498
},
{
"epoch": 0.3799733485627261,
"grad_norm": 35.924400329589844,
"learning_rate": 9.273197810125595e-06,
"loss": 0.6516,
"step": 499
},
{
"epoch": 0.3807348181991243,
"grad_norm": 46.43340301513672,
"learning_rate": 9.270030977773785e-06,
"loss": 0.7873,
"step": 500
},
{
"epoch": 0.3814962878355226,
"grad_norm": 35.527164459228516,
"learning_rate": 9.26685780431775e-06,
"loss": 0.8364,
"step": 501
},
{
"epoch": 0.3822577574719208,
"grad_norm": 38.08017349243164,
"learning_rate": 9.263678294469738e-06,
"loss": 0.8295,
"step": 502
},
{
"epoch": 0.38301922710831904,
"grad_norm": 44.70781326293945,
"learning_rate": 9.260492452951399e-06,
"loss": 0.7772,
"step": 503
},
{
"epoch": 0.3837806967447173,
"grad_norm": 40.65570831298828,
"learning_rate": 9.257300284493788e-06,
"loss": 0.8622,
"step": 504
},
{
"epoch": 0.38454216638111555,
"grad_norm": 46.29487228393555,
"learning_rate": 9.254101793837357e-06,
"loss": 0.9253,
"step": 505
},
{
"epoch": 0.38530363601751383,
"grad_norm": 43.05247497558594,
"learning_rate": 9.250896985731943e-06,
"loss": 0.7856,
"step": 506
},
{
"epoch": 0.38606510565391206,
"grad_norm": 45.552330017089844,
"learning_rate": 9.247685864936773e-06,
"loss": 0.9366,
"step": 507
},
{
"epoch": 0.3868265752903103,
"grad_norm": 40.42953872680664,
"learning_rate": 9.244468436220437e-06,
"loss": 0.8822,
"step": 508
},
{
"epoch": 0.38758804492670856,
"grad_norm": 41.48854446411133,
"learning_rate": 9.241244704360902e-06,
"loss": 0.9342,
"step": 509
},
{
"epoch": 0.3883495145631068,
"grad_norm": 47.127227783203125,
"learning_rate": 9.23801467414549e-06,
"loss": 0.8699,
"step": 510
},
{
"epoch": 0.38911098419950507,
"grad_norm": 38.96549606323242,
"learning_rate": 9.234778350370877e-06,
"loss": 0.8706,
"step": 511
},
{
"epoch": 0.3898724538359033,
"grad_norm": 36.132545471191406,
"learning_rate": 9.231535737843087e-06,
"loss": 0.6892,
"step": 512
},
{
"epoch": 0.3906339234723015,
"grad_norm": 42.765647888183594,
"learning_rate": 9.22828684137748e-06,
"loss": 0.8688,
"step": 513
},
{
"epoch": 0.3913953931086998,
"grad_norm": 48.11317825317383,
"learning_rate": 9.225031665798754e-06,
"loss": 0.9007,
"step": 514
},
{
"epoch": 0.39215686274509803,
"grad_norm": 40.7263298034668,
"learning_rate": 9.221770215940921e-06,
"loss": 0.8925,
"step": 515
},
{
"epoch": 0.39291833238149626,
"grad_norm": 38.51908493041992,
"learning_rate": 9.218502496647323e-06,
"loss": 0.8315,
"step": 516
},
{
"epoch": 0.39367980201789454,
"grad_norm": 42.210105895996094,
"learning_rate": 9.215228512770601e-06,
"loss": 0.8786,
"step": 517
},
{
"epoch": 0.39444127165429277,
"grad_norm": 38.10415267944336,
"learning_rate": 9.211948269172705e-06,
"loss": 0.8275,
"step": 518
},
{
"epoch": 0.39520274129069105,
"grad_norm": 40.787353515625,
"learning_rate": 9.208661770724884e-06,
"loss": 0.7724,
"step": 519
},
{
"epoch": 0.3959642109270893,
"grad_norm": 33.491241455078125,
"learning_rate": 9.205369022307667e-06,
"loss": 0.7493,
"step": 520
},
{
"epoch": 0.3967256805634875,
"grad_norm": 34.938289642333984,
"learning_rate": 9.202070028810868e-06,
"loss": 0.6538,
"step": 521
},
{
"epoch": 0.3974871501998858,
"grad_norm": 34.13314437866211,
"learning_rate": 9.198764795133579e-06,
"loss": 0.6973,
"step": 522
},
{
"epoch": 0.398248619836284,
"grad_norm": 47.31309127807617,
"learning_rate": 9.195453326184157e-06,
"loss": 0.9191,
"step": 523
},
{
"epoch": 0.3990100894726823,
"grad_norm": 31.99027442932129,
"learning_rate": 9.192135626880213e-06,
"loss": 0.6122,
"step": 524
},
{
"epoch": 0.3997715591090805,
"grad_norm": 35.60274124145508,
"learning_rate": 9.188811702148618e-06,
"loss": 0.7511,
"step": 525
},
{
"epoch": 0.40053302874547875,
"grad_norm": 38.91366195678711,
"learning_rate": 9.185481556925483e-06,
"loss": 0.6532,
"step": 526
},
{
"epoch": 0.40129449838187703,
"grad_norm": 39.65351104736328,
"learning_rate": 9.182145196156159e-06,
"loss": 0.8747,
"step": 527
},
{
"epoch": 0.40205596801827526,
"grad_norm": 39.787052154541016,
"learning_rate": 9.178802624795222e-06,
"loss": 0.9385,
"step": 528
},
{
"epoch": 0.40281743765467354,
"grad_norm": 40.41245651245117,
"learning_rate": 9.175453847806482e-06,
"loss": 0.8325,
"step": 529
},
{
"epoch": 0.40357890729107176,
"grad_norm": 33.6597900390625,
"learning_rate": 9.172098870162953e-06,
"loss": 0.7872,
"step": 530
},
{
"epoch": 0.40434037692747,
"grad_norm": 45.033607482910156,
"learning_rate": 9.168737696846866e-06,
"loss": 1.0956,
"step": 531
},
{
"epoch": 0.4051018465638683,
"grad_norm": 50.52238464355469,
"learning_rate": 9.165370332849642e-06,
"loss": 0.5881,
"step": 532
},
{
"epoch": 0.4058633162002665,
"grad_norm": 35.1140251159668,
"learning_rate": 9.161996783171908e-06,
"loss": 0.6781,
"step": 533
},
{
"epoch": 0.4066247858366648,
"grad_norm": 32.561767578125,
"learning_rate": 9.158617052823469e-06,
"loss": 0.6794,
"step": 534
},
{
"epoch": 0.407386255473063,
"grad_norm": 32.99089050292969,
"learning_rate": 9.155231146823309e-06,
"loss": 0.6522,
"step": 535
},
{
"epoch": 0.40814772510946123,
"grad_norm": 32.225341796875,
"learning_rate": 9.151839070199585e-06,
"loss": 0.875,
"step": 536
},
{
"epoch": 0.4089091947458595,
"grad_norm": 31.30025291442871,
"learning_rate": 9.148440827989615e-06,
"loss": 0.6423,
"step": 537
},
{
"epoch": 0.40967066438225774,
"grad_norm": 32.41795349121094,
"learning_rate": 9.145036425239877e-06,
"loss": 0.6811,
"step": 538
},
{
"epoch": 0.410432134018656,
"grad_norm": 48.22577667236328,
"learning_rate": 9.141625867005997e-06,
"loss": 0.7582,
"step": 539
},
{
"epoch": 0.41119360365505425,
"grad_norm": 42.18339920043945,
"learning_rate": 9.138209158352736e-06,
"loss": 0.8656,
"step": 540
},
{
"epoch": 0.4119550732914525,
"grad_norm": 36.78019332885742,
"learning_rate": 9.134786304353998e-06,
"loss": 0.9369,
"step": 541
},
{
"epoch": 0.41271654292785076,
"grad_norm": 29.20226287841797,
"learning_rate": 9.131357310092803e-06,
"loss": 0.7197,
"step": 542
},
{
"epoch": 0.413478012564249,
"grad_norm": 35.52668762207031,
"learning_rate": 9.127922180661297e-06,
"loss": 0.7683,
"step": 543
},
{
"epoch": 0.41423948220064727,
"grad_norm": 38.33584976196289,
"learning_rate": 9.124480921160735e-06,
"loss": 0.8918,
"step": 544
},
{
"epoch": 0.4150009518370455,
"grad_norm": 33.05094528198242,
"learning_rate": 9.121033536701471e-06,
"loss": 0.7041,
"step": 545
},
{
"epoch": 0.4157624214734437,
"grad_norm": 58.042510986328125,
"learning_rate": 9.117580032402962e-06,
"loss": 0.6833,
"step": 546
},
{
"epoch": 0.416523891109842,
"grad_norm": 32.62454605102539,
"learning_rate": 9.114120413393748e-06,
"loss": 0.7837,
"step": 547
},
{
"epoch": 0.41728536074624023,
"grad_norm": 37.49154281616211,
"learning_rate": 9.11065468481145e-06,
"loss": 0.7867,
"step": 548
},
{
"epoch": 0.4180468303826385,
"grad_norm": 40.316165924072266,
"learning_rate": 9.107182851802765e-06,
"loss": 1.0259,
"step": 549
},
{
"epoch": 0.41880830001903674,
"grad_norm": 40.33757781982422,
"learning_rate": 9.103704919523453e-06,
"loss": 0.7735,
"step": 550
},
{
"epoch": 0.41956976965543497,
"grad_norm": 64.47599792480469,
"learning_rate": 9.100220893138327e-06,
"loss": 0.8922,
"step": 551
},
{
"epoch": 0.42033123929183325,
"grad_norm": 67.62235260009766,
"learning_rate": 9.096730777821261e-06,
"loss": 0.941,
"step": 552
},
{
"epoch": 0.4210927089282315,
"grad_norm": 32.08481979370117,
"learning_rate": 9.09323457875516e-06,
"loss": 0.7371,
"step": 553
},
{
"epoch": 0.42185417856462976,
"grad_norm": 42.49589538574219,
"learning_rate": 9.089732301131971e-06,
"loss": 0.8789,
"step": 554
},
{
"epoch": 0.422615648201028,
"grad_norm": 42.37563705444336,
"learning_rate": 9.086223950152663e-06,
"loss": 0.8759,
"step": 555
},
{
"epoch": 0.4233771178374262,
"grad_norm": 33.896888732910156,
"learning_rate": 9.082709531027228e-06,
"loss": 0.9402,
"step": 556
},
{
"epoch": 0.4241385874738245,
"grad_norm": 40.62684631347656,
"learning_rate": 9.079189048974663e-06,
"loss": 0.6872,
"step": 557
},
{
"epoch": 0.4249000571102227,
"grad_norm": 40.33720016479492,
"learning_rate": 9.07566250922298e-06,
"loss": 0.879,
"step": 558
},
{
"epoch": 0.425661526746621,
"grad_norm": 33.24203109741211,
"learning_rate": 9.072129917009175e-06,
"loss": 0.6897,
"step": 559
},
{
"epoch": 0.4264229963830192,
"grad_norm": 47.51837921142578,
"learning_rate": 9.068591277579238e-06,
"loss": 0.723,
"step": 560
},
{
"epoch": 0.42718446601941745,
"grad_norm": 35.74695587158203,
"learning_rate": 9.06504659618814e-06,
"loss": 0.7811,
"step": 561
},
{
"epoch": 0.42794593565581573,
"grad_norm": 32.98100280761719,
"learning_rate": 9.06149587809982e-06,
"loss": 0.9302,
"step": 562
},
{
"epoch": 0.42870740529221396,
"grad_norm": 30.54879379272461,
"learning_rate": 9.057939128587186e-06,
"loss": 0.6805,
"step": 563
},
{
"epoch": 0.42946887492861224,
"grad_norm": 58.388309478759766,
"learning_rate": 9.054376352932102e-06,
"loss": 1.2049,
"step": 564
},
{
"epoch": 0.43023034456501047,
"grad_norm": 39.34674072265625,
"learning_rate": 9.050807556425376e-06,
"loss": 0.7732,
"step": 565
},
{
"epoch": 0.4309918142014087,
"grad_norm": 68.59402465820312,
"learning_rate": 9.047232744366767e-06,
"loss": 0.9679,
"step": 566
},
{
"epoch": 0.431753283837807,
"grad_norm": 38.54568099975586,
"learning_rate": 9.043651922064957e-06,
"loss": 0.7851,
"step": 567
},
{
"epoch": 0.4325147534742052,
"grad_norm": 42.376461029052734,
"learning_rate": 9.04006509483756e-06,
"loss": 0.8066,
"step": 568
},
{
"epoch": 0.4332762231106035,
"grad_norm": 27.669086456298828,
"learning_rate": 9.036472268011107e-06,
"loss": 0.5173,
"step": 569
},
{
"epoch": 0.4340376927470017,
"grad_norm": 39.31504821777344,
"learning_rate": 9.032873446921031e-06,
"loss": 0.8415,
"step": 570
},
{
"epoch": 0.43479916238339994,
"grad_norm": 32.41474533081055,
"learning_rate": 9.02926863691168e-06,
"loss": 0.7695,
"step": 571
},
{
"epoch": 0.4355606320197982,
"grad_norm": 41.595359802246094,
"learning_rate": 9.025657843336283e-06,
"loss": 0.8801,
"step": 572
},
{
"epoch": 0.43632210165619645,
"grad_norm": 42.80996322631836,
"learning_rate": 9.022041071556962e-06,
"loss": 0.9462,
"step": 573
},
{
"epoch": 0.43708357129259473,
"grad_norm": 45.28131103515625,
"learning_rate": 9.018418326944713e-06,
"loss": 0.6996,
"step": 574
},
{
"epoch": 0.43784504092899296,
"grad_norm": 43.1144905090332,
"learning_rate": 9.014789614879406e-06,
"loss": 0.9213,
"step": 575
},
{
"epoch": 0.4386065105653912,
"grad_norm": 41.01248550415039,
"learning_rate": 9.011154940749769e-06,
"loss": 0.788,
"step": 576
},
{
"epoch": 0.43936798020178947,
"grad_norm": 30.024755477905273,
"learning_rate": 9.007514309953387e-06,
"loss": 0.8227,
"step": 577
},
{
"epoch": 0.4401294498381877,
"grad_norm": 36.76607894897461,
"learning_rate": 9.003867727896684e-06,
"loss": 0.7102,
"step": 578
},
{
"epoch": 0.440890919474586,
"grad_norm": 39.9501953125,
"learning_rate": 9.000215199994933e-06,
"loss": 0.9365,
"step": 579
},
{
"epoch": 0.4416523891109842,
"grad_norm": 49.102298736572266,
"learning_rate": 8.996556731672228e-06,
"loss": 1.0171,
"step": 580
},
{
"epoch": 0.4424138587473824,
"grad_norm": 39.701072692871094,
"learning_rate": 8.992892328361485e-06,
"loss": 0.7041,
"step": 581
},
{
"epoch": 0.4431753283837807,
"grad_norm": 42.883201599121094,
"learning_rate": 8.989221995504437e-06,
"loss": 0.8881,
"step": 582
},
{
"epoch": 0.44393679802017894,
"grad_norm": 35.93183135986328,
"learning_rate": 8.985545738551625e-06,
"loss": 0.849,
"step": 583
},
{
"epoch": 0.4446982676565772,
"grad_norm": 33.20316696166992,
"learning_rate": 8.98186356296238e-06,
"loss": 0.8081,
"step": 584
},
{
"epoch": 0.44545973729297544,
"grad_norm": 34.129783630371094,
"learning_rate": 8.978175474204827e-06,
"loss": 0.6737,
"step": 585
},
{
"epoch": 0.44622120692937367,
"grad_norm": 35.47514343261719,
"learning_rate": 8.97448147775587e-06,
"loss": 0.8261,
"step": 586
},
{
"epoch": 0.44698267656577195,
"grad_norm": 34.96803665161133,
"learning_rate": 8.970781579101188e-06,
"loss": 0.877,
"step": 587
},
{
"epoch": 0.4477441462021702,
"grad_norm": 36.19900894165039,
"learning_rate": 8.967075783735226e-06,
"loss": 0.5846,
"step": 588
},
{
"epoch": 0.44850561583856846,
"grad_norm": 35.287315368652344,
"learning_rate": 8.963364097161182e-06,
"loss": 0.783,
"step": 589
},
{
"epoch": 0.4492670854749667,
"grad_norm": 37.13393783569336,
"learning_rate": 8.959646524891005e-06,
"loss": 0.6996,
"step": 590
},
{
"epoch": 0.4500285551113649,
"grad_norm": 35.42992401123047,
"learning_rate": 8.955923072445386e-06,
"loss": 0.7812,
"step": 591
},
{
"epoch": 0.4507900247477632,
"grad_norm": 34.43177032470703,
"learning_rate": 8.952193745353743e-06,
"loss": 0.6311,
"step": 592
},
{
"epoch": 0.4515514943841614,
"grad_norm": 29.86054229736328,
"learning_rate": 8.948458549154224e-06,
"loss": 0.5255,
"step": 593
},
{
"epoch": 0.4523129640205597,
"grad_norm": 40.26968765258789,
"learning_rate": 8.94471748939369e-06,
"loss": 0.8848,
"step": 594
},
{
"epoch": 0.45307443365695793,
"grad_norm": 35.20211410522461,
"learning_rate": 8.940970571627707e-06,
"loss": 0.6966,
"step": 595
},
{
"epoch": 0.45383590329335616,
"grad_norm": 27.7930850982666,
"learning_rate": 8.937217801420544e-06,
"loss": 0.7314,
"step": 596
},
{
"epoch": 0.45459737292975444,
"grad_norm": 34.17549514770508,
"learning_rate": 8.933459184345162e-06,
"loss": 0.8493,
"step": 597
},
{
"epoch": 0.45535884256615267,
"grad_norm": 32.22350311279297,
"learning_rate": 8.9296947259832e-06,
"loss": 0.8395,
"step": 598
},
{
"epoch": 0.45612031220255095,
"grad_norm": 26.227863311767578,
"learning_rate": 8.925924431924974e-06,
"loss": 0.4723,
"step": 599
},
{
"epoch": 0.4568817818389492,
"grad_norm": 32.933223724365234,
"learning_rate": 8.922148307769469e-06,
"loss": 0.7971,
"step": 600
},
{
"epoch": 0.4576432514753474,
"grad_norm": 43.23099899291992,
"learning_rate": 8.918366359124322e-06,
"loss": 0.8534,
"step": 601
},
{
"epoch": 0.4584047211117457,
"grad_norm": 49.473995208740234,
"learning_rate": 8.914578591605824e-06,
"loss": 0.9335,
"step": 602
},
{
"epoch": 0.4591661907481439,
"grad_norm": 31.68277359008789,
"learning_rate": 8.910785010838905e-06,
"loss": 0.8195,
"step": 603
},
{
"epoch": 0.4599276603845422,
"grad_norm": 38.720611572265625,
"learning_rate": 8.906985622457129e-06,
"loss": 1.0383,
"step": 604
},
{
"epoch": 0.4606891300209404,
"grad_norm": 32.30598831176758,
"learning_rate": 8.903180432102686e-06,
"loss": 0.8484,
"step": 605
},
{
"epoch": 0.46145059965733864,
"grad_norm": 35.373565673828125,
"learning_rate": 8.899369445426377e-06,
"loss": 0.7706,
"step": 606
},
{
"epoch": 0.4622120692937369,
"grad_norm": 36.211456298828125,
"learning_rate": 8.895552668087615e-06,
"loss": 0.7762,
"step": 607
},
{
"epoch": 0.46297353893013515,
"grad_norm": 31.659029006958008,
"learning_rate": 8.891730105754412e-06,
"loss": 0.6671,
"step": 608
},
{
"epoch": 0.46373500856653344,
"grad_norm": 34.945106506347656,
"learning_rate": 8.887901764103368e-06,
"loss": 0.9202,
"step": 609
},
{
"epoch": 0.46449647820293166,
"grad_norm": 31.101940155029297,
"learning_rate": 8.88406764881967e-06,
"loss": 0.7485,
"step": 610
},
{
"epoch": 0.4652579478393299,
"grad_norm": 39.410953521728516,
"learning_rate": 8.880227765597074e-06,
"loss": 1.0098,
"step": 611
},
{
"epoch": 0.46601941747572817,
"grad_norm": 30.449037551879883,
"learning_rate": 8.876382120137906e-06,
"loss": 0.7136,
"step": 612
},
{
"epoch": 0.4667808871121264,
"grad_norm": 42.712135314941406,
"learning_rate": 8.872530718153044e-06,
"loss": 0.9093,
"step": 613
},
{
"epoch": 0.4675423567485247,
"grad_norm": 30.30866050720215,
"learning_rate": 8.86867356536192e-06,
"loss": 0.6193,
"step": 614
},
{
"epoch": 0.4683038263849229,
"grad_norm": 38.80727005004883,
"learning_rate": 8.864810667492502e-06,
"loss": 0.496,
"step": 615
},
{
"epoch": 0.46906529602132113,
"grad_norm": 37.91781234741211,
"learning_rate": 8.860942030281294e-06,
"loss": 0.8192,
"step": 616
},
{
"epoch": 0.4698267656577194,
"grad_norm": 35.720909118652344,
"learning_rate": 8.857067659473319e-06,
"loss": 0.897,
"step": 617
},
{
"epoch": 0.47058823529411764,
"grad_norm": 40.22643280029297,
"learning_rate": 8.853187560822113e-06,
"loss": 0.7797,
"step": 618
},
{
"epoch": 0.4713497049305159,
"grad_norm": 36.7399787902832,
"learning_rate": 8.849301740089725e-06,
"loss": 0.8,
"step": 619
},
{
"epoch": 0.47211117456691415,
"grad_norm": 34.08172607421875,
"learning_rate": 8.845410203046695e-06,
"loss": 0.8473,
"step": 620
},
{
"epoch": 0.4728726442033124,
"grad_norm": 45.90078353881836,
"learning_rate": 8.841512955472054e-06,
"loss": 1.0885,
"step": 621
},
{
"epoch": 0.47363411383971066,
"grad_norm": 40.18413543701172,
"learning_rate": 8.837610003153316e-06,
"loss": 0.7083,
"step": 622
},
{
"epoch": 0.4743955834761089,
"grad_norm": 43.39189910888672,
"learning_rate": 8.83370135188646e-06,
"loss": 0.963,
"step": 623
},
{
"epoch": 0.47515705311250717,
"grad_norm": 27.63300895690918,
"learning_rate": 8.829787007475936e-06,
"loss": 0.774,
"step": 624
},
{
"epoch": 0.4759185227489054,
"grad_norm": 29.548891067504883,
"learning_rate": 8.825866975734642e-06,
"loss": 0.5855,
"step": 625
},
{
"epoch": 0.4766799923853036,
"grad_norm": 55.26377487182617,
"learning_rate": 8.821941262483924e-06,
"loss": 1.2597,
"step": 626
},
{
"epoch": 0.4774414620217019,
"grad_norm": 36.47280502319336,
"learning_rate": 8.81800987355357e-06,
"loss": 0.8026,
"step": 627
},
{
"epoch": 0.4782029316581001,
"grad_norm": 31.497495651245117,
"learning_rate": 8.814072814781788e-06,
"loss": 0.7601,
"step": 628
},
{
"epoch": 0.47896440129449835,
"grad_norm": 37.276546478271484,
"learning_rate": 8.810130092015211e-06,
"loss": 0.8585,
"step": 629
},
{
"epoch": 0.47972587093089664,
"grad_norm": 31.686859130859375,
"learning_rate": 8.806181711108883e-06,
"loss": 0.8257,
"step": 630
},
{
"epoch": 0.48048734056729486,
"grad_norm": 45.68613815307617,
"learning_rate": 8.802227677926249e-06,
"loss": 0.774,
"step": 631
},
{
"epoch": 0.48124881020369314,
"grad_norm": 30.245437622070312,
"learning_rate": 8.798267998339149e-06,
"loss": 0.75,
"step": 632
},
{
"epoch": 0.48201027984009137,
"grad_norm": 26.60959815979004,
"learning_rate": 8.794302678227809e-06,
"loss": 0.592,
"step": 633
},
{
"epoch": 0.4827717494764896,
"grad_norm": 32.366424560546875,
"learning_rate": 8.790331723480828e-06,
"loss": 0.6491,
"step": 634
},
{
"epoch": 0.4835332191128879,
"grad_norm": 35.88400650024414,
"learning_rate": 8.786355139995173e-06,
"loss": 0.7825,
"step": 635
},
{
"epoch": 0.4842946887492861,
"grad_norm": 33.99674606323242,
"learning_rate": 8.782372933676174e-06,
"loss": 0.7043,
"step": 636
},
{
"epoch": 0.4850561583856844,
"grad_norm": 30.469955444335938,
"learning_rate": 8.778385110437507e-06,
"loss": 0.8757,
"step": 637
},
{
"epoch": 0.4858176280220826,
"grad_norm": 32.127506256103516,
"learning_rate": 8.77439167620119e-06,
"loss": 0.7006,
"step": 638
},
{
"epoch": 0.48657909765848084,
"grad_norm": 33.425872802734375,
"learning_rate": 8.770392636897576e-06,
"loss": 0.8351,
"step": 639
},
{
"epoch": 0.4873405672948791,
"grad_norm": 38.01253128051758,
"learning_rate": 8.766387998465337e-06,
"loss": 0.962,
"step": 640
},
{
"epoch": 0.48810203693127735,
"grad_norm": 42.143707275390625,
"learning_rate": 8.762377766851467e-06,
"loss": 0.9168,
"step": 641
},
{
"epoch": 0.48886350656767563,
"grad_norm": 32.78673553466797,
"learning_rate": 8.758361948011256e-06,
"loss": 0.7231,
"step": 642
},
{
"epoch": 0.48962497620407386,
"grad_norm": 41.47932815551758,
"learning_rate": 8.754340547908301e-06,
"loss": 0.842,
"step": 643
},
{
"epoch": 0.4903864458404721,
"grad_norm": 30.17685317993164,
"learning_rate": 8.750313572514482e-06,
"loss": 0.7623,
"step": 644
},
{
"epoch": 0.49114791547687037,
"grad_norm": 44.7989616394043,
"learning_rate": 8.746281027809958e-06,
"loss": 0.9316,
"step": 645
},
{
"epoch": 0.4919093851132686,
"grad_norm": 33.90018844604492,
"learning_rate": 8.742242919783161e-06,
"loss": 0.8259,
"step": 646
},
{
"epoch": 0.4926708547496669,
"grad_norm": 28.30022621154785,
"learning_rate": 8.738199254430785e-06,
"loss": 0.747,
"step": 647
},
{
"epoch": 0.4934323243860651,
"grad_norm": 24.467458724975586,
"learning_rate": 8.734150037757774e-06,
"loss": 0.6057,
"step": 648
},
{
"epoch": 0.49419379402246333,
"grad_norm": 38.404075622558594,
"learning_rate": 8.730095275777317e-06,
"loss": 0.8806,
"step": 649
},
{
"epoch": 0.4949552636588616,
"grad_norm": 37.866458892822266,
"learning_rate": 8.72603497451084e-06,
"loss": 0.7884,
"step": 650
},
{
"epoch": 0.49571673329525984,
"grad_norm": 34.43553924560547,
"learning_rate": 8.721969139987994e-06,
"loss": 0.7668,
"step": 651
},
{
"epoch": 0.4964782029316581,
"grad_norm": 30.790082931518555,
"learning_rate": 8.717897778246642e-06,
"loss": 0.8646,
"step": 652
},
{
"epoch": 0.49723967256805635,
"grad_norm": 32.69529724121094,
"learning_rate": 8.713820895332864e-06,
"loss": 0.8441,
"step": 653
},
{
"epoch": 0.49800114220445457,
"grad_norm": 31.344131469726562,
"learning_rate": 8.709738497300932e-06,
"loss": 0.7419,
"step": 654
},
{
"epoch": 0.49876261184085285,
"grad_norm": 34.3707275390625,
"learning_rate": 8.70565059021331e-06,
"loss": 0.9512,
"step": 655
},
{
"epoch": 0.4995240814772511,
"grad_norm": 30.182653427124023,
"learning_rate": 8.701557180140647e-06,
"loss": 0.8681,
"step": 656
},
{
"epoch": 0.5002855511136494,
"grad_norm": 35.306846618652344,
"learning_rate": 8.697458273161757e-06,
"loss": 1.0412,
"step": 657
},
{
"epoch": 0.5010470207500476,
"grad_norm": 31.0938720703125,
"learning_rate": 8.693353875363623e-06,
"loss": 0.692,
"step": 658
},
{
"epoch": 0.5018084903864458,
"grad_norm": 31.533510208129883,
"learning_rate": 8.689243992841379e-06,
"loss": 0.6184,
"step": 659
},
{
"epoch": 0.502569960022844,
"grad_norm": 33.87437438964844,
"learning_rate": 8.685128631698304e-06,
"loss": 0.7929,
"step": 660
},
{
"epoch": 0.5033314296592424,
"grad_norm": 33.65624237060547,
"learning_rate": 8.681007798045814e-06,
"loss": 0.9535,
"step": 661
},
{
"epoch": 0.5040928992956406,
"grad_norm": 33.318153381347656,
"learning_rate": 8.676881498003453e-06,
"loss": 0.8338,
"step": 662
},
{
"epoch": 0.5048543689320388,
"grad_norm": 32.10042953491211,
"learning_rate": 8.672749737698878e-06,
"loss": 0.7802,
"step": 663
},
{
"epoch": 0.5056158385684371,
"grad_norm": 43.71488571166992,
"learning_rate": 8.668612523267862e-06,
"loss": 0.9463,
"step": 664
},
{
"epoch": 0.5063773082048353,
"grad_norm": 41.089439392089844,
"learning_rate": 8.66446986085427e-06,
"loss": 1.0539,
"step": 665
},
{
"epoch": 0.5071387778412336,
"grad_norm": 28.69977378845215,
"learning_rate": 8.66032175661006e-06,
"loss": 0.7834,
"step": 666
},
{
"epoch": 0.5079002474776318,
"grad_norm": 37.99454116821289,
"learning_rate": 8.656168216695276e-06,
"loss": 0.8248,
"step": 667
},
{
"epoch": 0.5086617171140301,
"grad_norm": 35.04216384887695,
"learning_rate": 8.652009247278026e-06,
"loss": 0.7818,
"step": 668
},
{
"epoch": 0.5094231867504283,
"grad_norm": 29.913171768188477,
"learning_rate": 8.647844854534487e-06,
"loss": 0.6774,
"step": 669
},
{
"epoch": 0.5101846563868265,
"grad_norm": 31.227197647094727,
"learning_rate": 8.643675044648885e-06,
"loss": 0.7429,
"step": 670
},
{
"epoch": 0.5109461260232249,
"grad_norm": 33.46771240234375,
"learning_rate": 8.639499823813499e-06,
"loss": 0.8269,
"step": 671
},
{
"epoch": 0.5117075956596231,
"grad_norm": 30.12987518310547,
"learning_rate": 8.635319198228633e-06,
"loss": 0.6747,
"step": 672
},
{
"epoch": 0.5124690652960213,
"grad_norm": 33.5628662109375,
"learning_rate": 8.631133174102623e-06,
"loss": 0.7735,
"step": 673
},
{
"epoch": 0.5132305349324195,
"grad_norm": 30.92879295349121,
"learning_rate": 8.62694175765182e-06,
"loss": 0.7511,
"step": 674
},
{
"epoch": 0.5139920045688178,
"grad_norm": 29.685874938964844,
"learning_rate": 8.622744955100589e-06,
"loss": 0.7242,
"step": 675
},
{
"epoch": 0.5147534742052161,
"grad_norm": 27.040605545043945,
"learning_rate": 8.618542772681282e-06,
"loss": 0.6924,
"step": 676
},
{
"epoch": 0.5155149438416143,
"grad_norm": 36.78512954711914,
"learning_rate": 8.614335216634247e-06,
"loss": 0.7219,
"step": 677
},
{
"epoch": 0.5162764134780126,
"grad_norm": 39.6012077331543,
"learning_rate": 8.610122293207815e-06,
"loss": 0.8196,
"step": 678
},
{
"epoch": 0.5170378831144108,
"grad_norm": 33.704833984375,
"learning_rate": 8.60590400865828e-06,
"loss": 0.9052,
"step": 679
},
{
"epoch": 0.517799352750809,
"grad_norm": 35.91033172607422,
"learning_rate": 8.601680369249905e-06,
"loss": 0.8486,
"step": 680
},
{
"epoch": 0.5185608223872074,
"grad_norm": 34.80812072753906,
"learning_rate": 8.5974513812549e-06,
"loss": 0.7627,
"step": 681
},
{
"epoch": 0.5193222920236056,
"grad_norm": 34.02023696899414,
"learning_rate": 8.593217050953417e-06,
"loss": 0.6358,
"step": 682
},
{
"epoch": 0.5200837616600038,
"grad_norm": 40.504356384277344,
"learning_rate": 8.588977384633549e-06,
"loss": 0.8517,
"step": 683
},
{
"epoch": 0.520845231296402,
"grad_norm": 38.43982696533203,
"learning_rate": 8.584732388591303e-06,
"loss": 0.8482,
"step": 684
},
{
"epoch": 0.5216067009328003,
"grad_norm": 31.85175323486328,
"learning_rate": 8.580482069130609e-06,
"loss": 0.8291,
"step": 685
},
{
"epoch": 0.5223681705691986,
"grad_norm": 43.681121826171875,
"learning_rate": 8.576226432563298e-06,
"loss": 0.9371,
"step": 686
},
{
"epoch": 0.5231296402055968,
"grad_norm": 37.21615982055664,
"learning_rate": 8.571965485209098e-06,
"loss": 0.8825,
"step": 687
},
{
"epoch": 0.523891109841995,
"grad_norm": 34.33724594116211,
"learning_rate": 8.56769923339562e-06,
"loss": 0.8466,
"step": 688
},
{
"epoch": 0.5246525794783933,
"grad_norm": 28.02695655822754,
"learning_rate": 8.563427683458365e-06,
"loss": 0.6877,
"step": 689
},
{
"epoch": 0.5254140491147915,
"grad_norm": 38.37015914916992,
"learning_rate": 8.559150841740686e-06,
"loss": 0.9541,
"step": 690
},
{
"epoch": 0.5261755187511898,
"grad_norm": 28.946571350097656,
"learning_rate": 8.554868714593802e-06,
"loss": 0.6303,
"step": 691
},
{
"epoch": 0.5269369883875881,
"grad_norm": 34.604644775390625,
"learning_rate": 8.550581308376783e-06,
"loss": 0.8105,
"step": 692
},
{
"epoch": 0.5276984580239863,
"grad_norm": 33.502750396728516,
"learning_rate": 8.546288629456535e-06,
"loss": 0.7431,
"step": 693
},
{
"epoch": 0.5284599276603845,
"grad_norm": 33.93081283569336,
"learning_rate": 8.541990684207793e-06,
"loss": 0.8172,
"step": 694
},
{
"epoch": 0.5292213972967827,
"grad_norm": 42.208980560302734,
"learning_rate": 8.537687479013118e-06,
"loss": 0.6655,
"step": 695
},
{
"epoch": 0.5299828669331811,
"grad_norm": 31.920230865478516,
"learning_rate": 8.533379020262874e-06,
"loss": 0.7987,
"step": 696
},
{
"epoch": 0.5307443365695793,
"grad_norm": 36.94475555419922,
"learning_rate": 8.529065314355237e-06,
"loss": 0.9525,
"step": 697
},
{
"epoch": 0.5315058062059775,
"grad_norm": 31.45237159729004,
"learning_rate": 8.524746367696165e-06,
"loss": 0.629,
"step": 698
},
{
"epoch": 0.5322672758423758,
"grad_norm": 27.37586784362793,
"learning_rate": 8.520422186699405e-06,
"loss": 0.7257,
"step": 699
},
{
"epoch": 0.533028745478774,
"grad_norm": 23.65913963317871,
"learning_rate": 8.516092777786475e-06,
"loss": 0.6115,
"step": 700
},
{
"epoch": 0.5337902151151723,
"grad_norm": 29.19643211364746,
"learning_rate": 8.511758147386659e-06,
"loss": 0.8932,
"step": 701
},
{
"epoch": 0.5345516847515706,
"grad_norm": 36.52588653564453,
"learning_rate": 8.50741830193699e-06,
"loss": 0.8328,
"step": 702
},
{
"epoch": 0.5353131543879688,
"grad_norm": 36.54971694946289,
"learning_rate": 8.503073247882252e-06,
"loss": 0.9755,
"step": 703
},
{
"epoch": 0.536074624024367,
"grad_norm": 42.013641357421875,
"learning_rate": 8.498722991674955e-06,
"loss": 1.0506,
"step": 704
},
{
"epoch": 0.5368360936607652,
"grad_norm": 36.055870056152344,
"learning_rate": 8.494367539775345e-06,
"loss": 0.8876,
"step": 705
},
{
"epoch": 0.5375975632971636,
"grad_norm": 30.088356018066406,
"learning_rate": 8.490006898651377e-06,
"loss": 0.7966,
"step": 706
},
{
"epoch": 0.5383590329335618,
"grad_norm": 40.09549331665039,
"learning_rate": 8.485641074778714e-06,
"loss": 0.8852,
"step": 707
},
{
"epoch": 0.53912050256996,
"grad_norm": 31.852556228637695,
"learning_rate": 8.48127007464071e-06,
"loss": 0.7979,
"step": 708
},
{
"epoch": 0.5398819722063583,
"grad_norm": 35.1148567199707,
"learning_rate": 8.47689390472842e-06,
"loss": 0.7933,
"step": 709
},
{
"epoch": 0.5406434418427565,
"grad_norm": 26.733436584472656,
"learning_rate": 8.472512571540558e-06,
"loss": 0.4899,
"step": 710
},
{
"epoch": 0.5414049114791548,
"grad_norm": 44.820186614990234,
"learning_rate": 8.468126081583524e-06,
"loss": 0.7472,
"step": 711
},
{
"epoch": 0.542166381115553,
"grad_norm": 34.0068359375,
"learning_rate": 8.463734441371359e-06,
"loss": 0.7348,
"step": 712
},
{
"epoch": 0.5429278507519513,
"grad_norm": 37.57758712768555,
"learning_rate": 8.459337657425762e-06,
"loss": 0.7294,
"step": 713
},
{
"epoch": 0.5436893203883495,
"grad_norm": 37.07135009765625,
"learning_rate": 8.454935736276071e-06,
"loss": 0.7843,
"step": 714
},
{
"epoch": 0.5444507900247477,
"grad_norm": 37.17467498779297,
"learning_rate": 8.450528684459249e-06,
"loss": 0.7833,
"step": 715
},
{
"epoch": 0.5452122596611461,
"grad_norm": 27.351919174194336,
"learning_rate": 8.44611650851988e-06,
"loss": 0.6807,
"step": 716
},
{
"epoch": 0.5459737292975443,
"grad_norm": 34.79734420776367,
"learning_rate": 8.441699215010157e-06,
"loss": 0.6943,
"step": 717
},
{
"epoch": 0.5467351989339425,
"grad_norm": 34.36366271972656,
"learning_rate": 8.437276810489871e-06,
"loss": 0.7774,
"step": 718
},
{
"epoch": 0.5474966685703407,
"grad_norm": 31.424224853515625,
"learning_rate": 8.432849301526406e-06,
"loss": 0.8368,
"step": 719
},
{
"epoch": 0.548258138206739,
"grad_norm": 33.47733688354492,
"learning_rate": 8.428416694694728e-06,
"loss": 0.7105,
"step": 720
},
{
"epoch": 0.5490196078431373,
"grad_norm": 27.397165298461914,
"learning_rate": 8.423978996577367e-06,
"loss": 0.674,
"step": 721
},
{
"epoch": 0.5497810774795355,
"grad_norm": 30.114990234375,
"learning_rate": 8.419536213764417e-06,
"loss": 0.7754,
"step": 722
},
{
"epoch": 0.5505425471159338,
"grad_norm": 29.79185676574707,
"learning_rate": 8.415088352853525e-06,
"loss": 0.7032,
"step": 723
},
{
"epoch": 0.551304016752332,
"grad_norm": 33.54899597167969,
"learning_rate": 8.410635420449877e-06,
"loss": 0.7938,
"step": 724
},
{
"epoch": 0.5520654863887302,
"grad_norm": 41.448951721191406,
"learning_rate": 8.406177423166188e-06,
"loss": 0.763,
"step": 725
},
{
"epoch": 0.5528269560251285,
"grad_norm": 28.263797760009766,
"learning_rate": 8.4017143676227e-06,
"loss": 0.7605,
"step": 726
},
{
"epoch": 0.5535884256615268,
"grad_norm": 26.69054412841797,
"learning_rate": 8.397246260447162e-06,
"loss": 0.6578,
"step": 727
},
{
"epoch": 0.554349895297925,
"grad_norm": 24.359060287475586,
"learning_rate": 8.392773108274827e-06,
"loss": 0.5786,
"step": 728
},
{
"epoch": 0.5551113649343232,
"grad_norm": 32.5286979675293,
"learning_rate": 8.388294917748439e-06,
"loss": 0.8173,
"step": 729
},
{
"epoch": 0.5558728345707215,
"grad_norm": 27.81627082824707,
"learning_rate": 8.383811695518224e-06,
"loss": 0.7195,
"step": 730
},
{
"epoch": 0.5566343042071198,
"grad_norm": 27.911325454711914,
"learning_rate": 8.379323448241881e-06,
"loss": 0.7935,
"step": 731
},
{
"epoch": 0.557395773843518,
"grad_norm": 28.142335891723633,
"learning_rate": 8.37483018258457e-06,
"loss": 0.8224,
"step": 732
},
{
"epoch": 0.5581572434799162,
"grad_norm": 26.930925369262695,
"learning_rate": 8.37033190521891e-06,
"loss": 0.7377,
"step": 733
},
{
"epoch": 0.5589187131163145,
"grad_norm": 27.190326690673828,
"learning_rate": 8.365828622824948e-06,
"loss": 0.7584,
"step": 734
},
{
"epoch": 0.5596801827527127,
"grad_norm": 34.811767578125,
"learning_rate": 8.361320342090177e-06,
"loss": 0.8059,
"step": 735
},
{
"epoch": 0.560441652389111,
"grad_norm": 27.499380111694336,
"learning_rate": 8.35680706970951e-06,
"loss": 0.7386,
"step": 736
},
{
"epoch": 0.5612031220255093,
"grad_norm": 27.849960327148438,
"learning_rate": 8.352288812385268e-06,
"loss": 0.8627,
"step": 737
},
{
"epoch": 0.5619645916619075,
"grad_norm": 38.18281936645508,
"learning_rate": 8.34776557682718e-06,
"loss": 0.6196,
"step": 738
},
{
"epoch": 0.5627260612983057,
"grad_norm": 42.69196701049805,
"learning_rate": 8.343237369752361e-06,
"loss": 0.9305,
"step": 739
},
{
"epoch": 0.5634875309347039,
"grad_norm": 34.84543228149414,
"learning_rate": 8.338704197885319e-06,
"loss": 0.8611,
"step": 740
},
{
"epoch": 0.5642490005711023,
"grad_norm": 31.243303298950195,
"learning_rate": 8.334166067957925e-06,
"loss": 0.7997,
"step": 741
},
{
"epoch": 0.5650104702075005,
"grad_norm": 39.43867111206055,
"learning_rate": 8.329622986709418e-06,
"loss": 0.9909,
"step": 742
},
{
"epoch": 0.5657719398438987,
"grad_norm": 26.780221939086914,
"learning_rate": 8.32507496088639e-06,
"loss": 0.8134,
"step": 743
},
{
"epoch": 0.566533409480297,
"grad_norm": 26.742109298706055,
"learning_rate": 8.320521997242773e-06,
"loss": 0.6863,
"step": 744
},
{
"epoch": 0.5672948791166952,
"grad_norm": 33.564453125,
"learning_rate": 8.315964102539835e-06,
"loss": 0.8355,
"step": 745
},
{
"epoch": 0.5680563487530935,
"grad_norm": 38.662662506103516,
"learning_rate": 8.311401283546164e-06,
"loss": 0.8981,
"step": 746
},
{
"epoch": 0.5688178183894917,
"grad_norm": 34.034584045410156,
"learning_rate": 8.306833547037662e-06,
"loss": 0.7791,
"step": 747
},
{
"epoch": 0.56957928802589,
"grad_norm": 33.56183624267578,
"learning_rate": 8.302260899797532e-06,
"loss": 0.6985,
"step": 748
},
{
"epoch": 0.5703407576622882,
"grad_norm": 30.30634880065918,
"learning_rate": 8.297683348616274e-06,
"loss": 0.8649,
"step": 749
},
{
"epoch": 0.5711022272986864,
"grad_norm": 30.258676528930664,
"learning_rate": 8.293100900291667e-06,
"loss": 0.6825,
"step": 750
},
{
"epoch": 0.5718636969350848,
"grad_norm": 29.29175567626953,
"learning_rate": 8.288513561628761e-06,
"loss": 0.5259,
"step": 751
},
{
"epoch": 0.572625166571483,
"grad_norm": 26.954097747802734,
"learning_rate": 8.283921339439871e-06,
"loss": 0.6971,
"step": 752
},
{
"epoch": 0.5733866362078812,
"grad_norm": 26.251937866210938,
"learning_rate": 8.279324240544565e-06,
"loss": 0.6413,
"step": 753
},
{
"epoch": 0.5741481058442794,
"grad_norm": 27.7598819732666,
"learning_rate": 8.274722271769648e-06,
"loss": 0.7426,
"step": 754
},
{
"epoch": 0.5749095754806777,
"grad_norm": 38.244510650634766,
"learning_rate": 8.270115439949164e-06,
"loss": 0.82,
"step": 755
},
{
"epoch": 0.575671045117076,
"grad_norm": 32.2285270690918,
"learning_rate": 8.26550375192437e-06,
"loss": 0.8009,
"step": 756
},
{
"epoch": 0.5764325147534742,
"grad_norm": 34.027008056640625,
"learning_rate": 8.260887214543747e-06,
"loss": 0.8165,
"step": 757
},
{
"epoch": 0.5771939843898725,
"grad_norm": 31.854063034057617,
"learning_rate": 8.256265834662965e-06,
"loss": 0.8268,
"step": 758
},
{
"epoch": 0.5779554540262707,
"grad_norm": 34.228004455566406,
"learning_rate": 8.251639619144888e-06,
"loss": 0.7154,
"step": 759
},
{
"epoch": 0.5787169236626689,
"grad_norm": 26.681699752807617,
"learning_rate": 8.24700857485957e-06,
"loss": 0.7981,
"step": 760
},
{
"epoch": 0.5794783932990673,
"grad_norm": 30.64720916748047,
"learning_rate": 8.242372708684223e-06,
"loss": 0.8458,
"step": 761
},
{
"epoch": 0.5802398629354655,
"grad_norm": 30.193696975708008,
"learning_rate": 8.23773202750323e-06,
"loss": 0.8717,
"step": 762
},
{
"epoch": 0.5810013325718637,
"grad_norm": 29.077957153320312,
"learning_rate": 8.233086538208115e-06,
"loss": 0.7841,
"step": 763
},
{
"epoch": 0.5817628022082619,
"grad_norm": 38.71321105957031,
"learning_rate": 8.228436247697556e-06,
"loss": 0.8796,
"step": 764
},
{
"epoch": 0.5825242718446602,
"grad_norm": 31.869609832763672,
"learning_rate": 8.223781162877341e-06,
"loss": 0.8083,
"step": 765
},
{
"epoch": 0.5832857414810585,
"grad_norm": 37.75756072998047,
"learning_rate": 8.2191212906604e-06,
"loss": 1.0693,
"step": 766
},
{
"epoch": 0.5840472111174567,
"grad_norm": 34.645416259765625,
"learning_rate": 8.214456637966754e-06,
"loss": 0.9185,
"step": 767
},
{
"epoch": 0.584808680753855,
"grad_norm": 27.102375030517578,
"learning_rate": 8.209787211723534e-06,
"loss": 0.659,
"step": 768
},
{
"epoch": 0.5855701503902532,
"grad_norm": 27.539791107177734,
"learning_rate": 8.205113018864954e-06,
"loss": 0.6687,
"step": 769
},
{
"epoch": 0.5863316200266514,
"grad_norm": 30.286901473999023,
"learning_rate": 8.20043406633231e-06,
"loss": 0.7485,
"step": 770
},
{
"epoch": 0.5870930896630497,
"grad_norm": 33.38515090942383,
"learning_rate": 8.195750361073966e-06,
"loss": 0.8467,
"step": 771
},
{
"epoch": 0.587854559299448,
"grad_norm": 44.452693939208984,
"learning_rate": 8.191061910045341e-06,
"loss": 0.6534,
"step": 772
},
{
"epoch": 0.5886160289358462,
"grad_norm": 33.47657775878906,
"learning_rate": 8.186368720208906e-06,
"loss": 0.7487,
"step": 773
},
{
"epoch": 0.5893774985722444,
"grad_norm": 29.39069938659668,
"learning_rate": 8.181670798534164e-06,
"loss": 0.8182,
"step": 774
},
{
"epoch": 0.5901389682086426,
"grad_norm": 33.332374572753906,
"learning_rate": 8.17696815199765e-06,
"loss": 0.9018,
"step": 775
},
{
"epoch": 0.590900437845041,
"grad_norm": 27.58820343017578,
"learning_rate": 8.172260787582914e-06,
"loss": 0.7479,
"step": 776
},
{
"epoch": 0.5916619074814392,
"grad_norm": 31.04929542541504,
"learning_rate": 8.167548712280507e-06,
"loss": 0.7872,
"step": 777
},
{
"epoch": 0.5924233771178374,
"grad_norm": 38.24930953979492,
"learning_rate": 8.162831933087985e-06,
"loss": 0.7874,
"step": 778
},
{
"epoch": 0.5931848467542357,
"grad_norm": 36.27316665649414,
"learning_rate": 8.158110457009882e-06,
"loss": 0.8672,
"step": 779
},
{
"epoch": 0.5939463163906339,
"grad_norm": 36.5713996887207,
"learning_rate": 8.15338429105771e-06,
"loss": 0.5997,
"step": 780
},
{
"epoch": 0.5947077860270322,
"grad_norm": 35.257774353027344,
"learning_rate": 8.148653442249949e-06,
"loss": 0.8171,
"step": 781
},
{
"epoch": 0.5954692556634305,
"grad_norm": 36.267547607421875,
"learning_rate": 8.143917917612023e-06,
"loss": 0.8415,
"step": 782
},
{
"epoch": 0.5962307252998287,
"grad_norm": 27.755474090576172,
"learning_rate": 8.13917772417631e-06,
"loss": 0.6817,
"step": 783
},
{
"epoch": 0.5969921949362269,
"grad_norm": 36.79349136352539,
"learning_rate": 8.134432868982118e-06,
"loss": 0.7736,
"step": 784
},
{
"epoch": 0.5977536645726251,
"grad_norm": 32.57246398925781,
"learning_rate": 8.129683359075675e-06,
"loss": 0.7669,
"step": 785
},
{
"epoch": 0.5985151342090235,
"grad_norm": 37.34968185424805,
"learning_rate": 8.124929201510124e-06,
"loss": 0.7779,
"step": 786
},
{
"epoch": 0.5992766038454217,
"grad_norm": 40.94609069824219,
"learning_rate": 8.12017040334551e-06,
"loss": 0.9911,
"step": 787
},
{
"epoch": 0.6000380734818199,
"grad_norm": 28.85545539855957,
"learning_rate": 8.11540697164877e-06,
"loss": 0.6715,
"step": 788
},
{
"epoch": 0.6007995431182181,
"grad_norm": 49.465614318847656,
"learning_rate": 8.110638913493718e-06,
"loss": 0.6255,
"step": 789
},
{
"epoch": 0.6015610127546164,
"grad_norm": 50.699974060058594,
"learning_rate": 8.105866235961045e-06,
"loss": 0.566,
"step": 790
},
{
"epoch": 0.6023224823910147,
"grad_norm": 27.396591186523438,
"learning_rate": 8.101088946138294e-06,
"loss": 0.6544,
"step": 791
},
{
"epoch": 0.6030839520274129,
"grad_norm": 32.85529708862305,
"learning_rate": 8.096307051119863e-06,
"loss": 0.8224,
"step": 792
},
{
"epoch": 0.6038454216638112,
"grad_norm": 26.31869125366211,
"learning_rate": 8.091520558006988e-06,
"loss": 0.7248,
"step": 793
},
{
"epoch": 0.6046068913002094,
"grad_norm": 28.539140701293945,
"learning_rate": 8.086729473907734e-06,
"loss": 0.7012,
"step": 794
},
{
"epoch": 0.6053683609366076,
"grad_norm": 34.67433166503906,
"learning_rate": 8.081933805936977e-06,
"loss": 0.793,
"step": 795
},
{
"epoch": 0.606129830573006,
"grad_norm": 26.926082611083984,
"learning_rate": 8.07713356121641e-06,
"loss": 0.6892,
"step": 796
},
{
"epoch": 0.6068913002094042,
"grad_norm": 37.93033981323242,
"learning_rate": 8.07232874687452e-06,
"loss": 0.7574,
"step": 797
},
{
"epoch": 0.6076527698458024,
"grad_norm": 61.16469192504883,
"learning_rate": 8.067519370046572e-06,
"loss": 1.1853,
"step": 798
},
{
"epoch": 0.6084142394822006,
"grad_norm": 36.617897033691406,
"learning_rate": 8.06270543787462e-06,
"loss": 0.9867,
"step": 799
},
{
"epoch": 0.6091757091185989,
"grad_norm": 41.403324127197266,
"learning_rate": 8.057886957507466e-06,
"loss": 1.0133,
"step": 800
},
{
"epoch": 0.6099371787549971,
"grad_norm": 35.976966857910156,
"learning_rate": 8.053063936100682e-06,
"loss": 0.9601,
"step": 801
},
{
"epoch": 0.6106986483913954,
"grad_norm": 28.360689163208008,
"learning_rate": 8.048236380816576e-06,
"loss": 0.7825,
"step": 802
},
{
"epoch": 0.6114601180277937,
"grad_norm": 28.78184700012207,
"learning_rate": 8.043404298824189e-06,
"loss": 0.7889,
"step": 803
},
{
"epoch": 0.6122215876641919,
"grad_norm": 21.7941951751709,
"learning_rate": 8.038567697299287e-06,
"loss": 0.5789,
"step": 804
},
{
"epoch": 0.6129830573005901,
"grad_norm": 33.62324142456055,
"learning_rate": 8.033726583424341e-06,
"loss": 0.8517,
"step": 805
},
{
"epoch": 0.6137445269369883,
"grad_norm": 33.0076789855957,
"learning_rate": 8.028880964388534e-06,
"loss": 0.8585,
"step": 806
},
{
"epoch": 0.6145059965733867,
"grad_norm": 27.630537033081055,
"learning_rate": 8.024030847387729e-06,
"loss": 0.7323,
"step": 807
},
{
"epoch": 0.6152674662097849,
"grad_norm": 24.592273712158203,
"learning_rate": 8.019176239624476e-06,
"loss": 0.7206,
"step": 808
},
{
"epoch": 0.6160289358461831,
"grad_norm": 26.27126693725586,
"learning_rate": 8.014317148307987e-06,
"loss": 0.7633,
"step": 809
},
{
"epoch": 0.6167904054825813,
"grad_norm": 28.837806701660156,
"learning_rate": 8.009453580654138e-06,
"loss": 0.5625,
"step": 810
},
{
"epoch": 0.6175518751189796,
"grad_norm": 22.987646102905273,
"learning_rate": 8.004585543885448e-06,
"loss": 0.6402,
"step": 811
},
{
"epoch": 0.6183133447553779,
"grad_norm": 26.343568801879883,
"learning_rate": 7.999713045231078e-06,
"loss": 0.6907,
"step": 812
},
{
"epoch": 0.6190748143917761,
"grad_norm": 25.834089279174805,
"learning_rate": 7.99483609192681e-06,
"loss": 0.5879,
"step": 813
},
{
"epoch": 0.6198362840281744,
"grad_norm": 26.641136169433594,
"learning_rate": 7.989954691215044e-06,
"loss": 0.7393,
"step": 814
},
{
"epoch": 0.6205977536645726,
"grad_norm": 31.973413467407227,
"learning_rate": 7.985068850344783e-06,
"loss": 0.8564,
"step": 815
},
{
"epoch": 0.6213592233009708,
"grad_norm": 32.119537353515625,
"learning_rate": 7.980178576571626e-06,
"loss": 0.9437,
"step": 816
},
{
"epoch": 0.6221206929373692,
"grad_norm": 25.91680908203125,
"learning_rate": 7.975283877157752e-06,
"loss": 0.6618,
"step": 817
},
{
"epoch": 0.6228821625737674,
"grad_norm": 26.057126998901367,
"learning_rate": 7.970384759371914e-06,
"loss": 0.7615,
"step": 818
},
{
"epoch": 0.6236436322101656,
"grad_norm": 28.24910545349121,
"learning_rate": 7.965481230489425e-06,
"loss": 0.7863,
"step": 819
},
{
"epoch": 0.6244051018465638,
"grad_norm": 32.42158889770508,
"learning_rate": 7.960573297792153e-06,
"loss": 0.7045,
"step": 820
},
{
"epoch": 0.6251665714829621,
"grad_norm": 33.47333526611328,
"learning_rate": 7.955660968568499e-06,
"loss": 0.8996,
"step": 821
},
{
"epoch": 0.6259280411193604,
"grad_norm": 33.64387512207031,
"learning_rate": 7.950744250113398e-06,
"loss": 0.8674,
"step": 822
},
{
"epoch": 0.6266895107557586,
"grad_norm": 21.23735809326172,
"learning_rate": 7.945823149728304e-06,
"loss": 0.6238,
"step": 823
},
{
"epoch": 0.6274509803921569,
"grad_norm": 33.83268356323242,
"learning_rate": 7.94089767472117e-06,
"loss": 0.7337,
"step": 824
},
{
"epoch": 0.6282124500285551,
"grad_norm": 31.84981918334961,
"learning_rate": 7.935967832406456e-06,
"loss": 0.7967,
"step": 825
},
{
"epoch": 0.6289739196649533,
"grad_norm": 31.125083923339844,
"learning_rate": 7.9310336301051e-06,
"loss": 1.0462,
"step": 826
},
{
"epoch": 0.6297353893013516,
"grad_norm": 26.396533966064453,
"learning_rate": 7.92609507514452e-06,
"loss": 0.6512,
"step": 827
},
{
"epoch": 0.6304968589377499,
"grad_norm": 26.503463745117188,
"learning_rate": 7.921152174858592e-06,
"loss": 0.7569,
"step": 828
},
{
"epoch": 0.6312583285741481,
"grad_norm": 34.26448440551758,
"learning_rate": 7.91620493658765e-06,
"loss": 0.799,
"step": 829
},
{
"epoch": 0.6320197982105463,
"grad_norm": 32.36408233642578,
"learning_rate": 7.911253367678469e-06,
"loss": 0.9017,
"step": 830
},
{
"epoch": 0.6327812678469446,
"grad_norm": 30.752483367919922,
"learning_rate": 7.906297475484255e-06,
"loss": 0.8698,
"step": 831
},
{
"epoch": 0.6335427374833429,
"grad_norm": 27.395137786865234,
"learning_rate": 7.90133726736463e-06,
"loss": 0.6505,
"step": 832
},
{
"epoch": 0.6343042071197411,
"grad_norm": 32.6951789855957,
"learning_rate": 7.896372750685632e-06,
"loss": 0.8879,
"step": 833
},
{
"epoch": 0.6350656767561393,
"grad_norm": 39.803714752197266,
"learning_rate": 7.891403932819692e-06,
"loss": 0.9498,
"step": 834
},
{
"epoch": 0.6358271463925376,
"grad_norm": 34.31433868408203,
"learning_rate": 7.886430821145633e-06,
"loss": 0.9549,
"step": 835
},
{
"epoch": 0.6365886160289358,
"grad_norm": 27.380159378051758,
"learning_rate": 7.881453423048651e-06,
"loss": 0.7375,
"step": 836
},
{
"epoch": 0.6373500856653341,
"grad_norm": 26.90700912475586,
"learning_rate": 7.876471745920304e-06,
"loss": 0.6806,
"step": 837
},
{
"epoch": 0.6381115553017324,
"grad_norm": 25.832687377929688,
"learning_rate": 7.871485797158515e-06,
"loss": 0.8036,
"step": 838
},
{
"epoch": 0.6388730249381306,
"grad_norm": 34.30730056762695,
"learning_rate": 7.866495584167542e-06,
"loss": 0.9942,
"step": 839
},
{
"epoch": 0.6396344945745288,
"grad_norm": 25.98689079284668,
"learning_rate": 7.861501114357976e-06,
"loss": 0.7233,
"step": 840
},
{
"epoch": 0.640395964210927,
"grad_norm": 32.80167007446289,
"learning_rate": 7.856502395146735e-06,
"loss": 0.7946,
"step": 841
},
{
"epoch": 0.6411574338473254,
"grad_norm": 34.85222625732422,
"learning_rate": 7.851499433957039e-06,
"loss": 1.0334,
"step": 842
},
{
"epoch": 0.6419189034837236,
"grad_norm": 36.892765045166016,
"learning_rate": 7.846492238218414e-06,
"loss": 0.8306,
"step": 843
},
{
"epoch": 0.6426803731201218,
"grad_norm": 29.023441314697266,
"learning_rate": 7.841480815366676e-06,
"loss": 0.9039,
"step": 844
},
{
"epoch": 0.64344184275652,
"grad_norm": 30.305448532104492,
"learning_rate": 7.83646517284391e-06,
"loss": 0.6917,
"step": 845
},
{
"epoch": 0.6442033123929183,
"grad_norm": 32.46067428588867,
"learning_rate": 7.831445318098476e-06,
"loss": 0.7484,
"step": 846
},
{
"epoch": 0.6449647820293166,
"grad_norm": 29.135446548461914,
"learning_rate": 7.826421258584984e-06,
"loss": 0.7558,
"step": 847
},
{
"epoch": 0.6457262516657148,
"grad_norm": 27.77723503112793,
"learning_rate": 7.821393001764287e-06,
"loss": 0.7524,
"step": 848
},
{
"epoch": 0.6464877213021131,
"grad_norm": 26.93240737915039,
"learning_rate": 7.81636055510348e-06,
"loss": 0.8165,
"step": 849
},
{
"epoch": 0.6472491909385113,
"grad_norm": 29.845855712890625,
"learning_rate": 7.811323926075867e-06,
"loss": 0.7827,
"step": 850
},
{
"epoch": 0.6480106605749095,
"grad_norm": 28.056304931640625,
"learning_rate": 7.806283122160974e-06,
"loss": 0.7516,
"step": 851
},
{
"epoch": 0.6487721302113079,
"grad_norm": 40.83567428588867,
"learning_rate": 7.801238150844523e-06,
"loss": 0.8165,
"step": 852
},
{
"epoch": 0.6495335998477061,
"grad_norm": 26.75080680847168,
"learning_rate": 7.796189019618421e-06,
"loss": 0.5547,
"step": 853
},
{
"epoch": 0.6502950694841043,
"grad_norm": 35.45816421508789,
"learning_rate": 7.791135735980758e-06,
"loss": 0.9976,
"step": 854
},
{
"epoch": 0.6510565391205025,
"grad_norm": 24.48919105529785,
"learning_rate": 7.786078307435788e-06,
"loss": 0.5694,
"step": 855
},
{
"epoch": 0.6518180087569008,
"grad_norm": 25.51799964904785,
"learning_rate": 7.781016741493922e-06,
"loss": 0.7217,
"step": 856
},
{
"epoch": 0.6525794783932991,
"grad_norm": 52.26631164550781,
"learning_rate": 7.77595104567171e-06,
"loss": 0.7708,
"step": 857
},
{
"epoch": 0.6533409480296973,
"grad_norm": 27.07032012939453,
"learning_rate": 7.770881227491842e-06,
"loss": 0.7683,
"step": 858
},
{
"epoch": 0.6541024176660956,
"grad_norm": 57.096466064453125,
"learning_rate": 7.765807294483125e-06,
"loss": 0.7533,
"step": 859
},
{
"epoch": 0.6548638873024938,
"grad_norm": 30.253202438354492,
"learning_rate": 7.76072925418048e-06,
"loss": 0.8218,
"step": 860
},
{
"epoch": 0.655625356938892,
"grad_norm": 31.545814514160156,
"learning_rate": 7.755647114124922e-06,
"loss": 0.8249,
"step": 861
},
{
"epoch": 0.6563868265752903,
"grad_norm": 28.886499404907227,
"learning_rate": 7.750560881863563e-06,
"loss": 0.8746,
"step": 862
},
{
"epoch": 0.6571482962116886,
"grad_norm": 26.418315887451172,
"learning_rate": 7.74547056494958e-06,
"loss": 0.6539,
"step": 863
},
{
"epoch": 0.6579097658480868,
"grad_norm": 31.141759872436523,
"learning_rate": 7.740376170942227e-06,
"loss": 0.9195,
"step": 864
},
{
"epoch": 0.658671235484485,
"grad_norm": 29.36141014099121,
"learning_rate": 7.73527770740681e-06,
"loss": 0.886,
"step": 865
},
{
"epoch": 0.6594327051208833,
"grad_norm": 27.355432510375977,
"learning_rate": 7.730175181914672e-06,
"loss": 0.7236,
"step": 866
},
{
"epoch": 0.6601941747572816,
"grad_norm": 45.42149353027344,
"learning_rate": 7.725068602043195e-06,
"loss": 1.1618,
"step": 867
},
{
"epoch": 0.6609556443936798,
"grad_norm": 33.360931396484375,
"learning_rate": 7.719957975375779e-06,
"loss": 0.8988,
"step": 868
},
{
"epoch": 0.661717114030078,
"grad_norm": 28.71207046508789,
"learning_rate": 7.714843309501836e-06,
"loss": 0.5699,
"step": 869
},
{
"epoch": 0.6624785836664763,
"grad_norm": 40.582035064697266,
"learning_rate": 7.709724612016772e-06,
"loss": 0.8159,
"step": 870
},
{
"epoch": 0.6632400533028745,
"grad_norm": 40.518150329589844,
"learning_rate": 7.704601890521982e-06,
"loss": 0.8031,
"step": 871
},
{
"epoch": 0.6640015229392728,
"grad_norm": 32.10520553588867,
"learning_rate": 7.699475152624839e-06,
"loss": 0.7844,
"step": 872
},
{
"epoch": 0.6647629925756711,
"grad_norm": 33.51770782470703,
"learning_rate": 7.694344405938679e-06,
"loss": 0.9458,
"step": 873
},
{
"epoch": 0.6655244622120693,
"grad_norm": 28.527631759643555,
"learning_rate": 7.68920965808279e-06,
"loss": 0.8622,
"step": 874
},
{
"epoch": 0.6662859318484675,
"grad_norm": 38.19065856933594,
"learning_rate": 7.684070916682403e-06,
"loss": 0.6396,
"step": 875
},
{
"epoch": 0.6670474014848657,
"grad_norm": 27.93634033203125,
"learning_rate": 7.67892818936868e-06,
"loss": 0.7168,
"step": 876
},
{
"epoch": 0.6678088711212641,
"grad_norm": 28.99605941772461,
"learning_rate": 7.673781483778698e-06,
"loss": 0.7734,
"step": 877
},
{
"epoch": 0.6685703407576623,
"grad_norm": 39.4337158203125,
"learning_rate": 7.66863080755545e-06,
"loss": 1.4895,
"step": 878
},
{
"epoch": 0.6693318103940605,
"grad_norm": 28.5806827545166,
"learning_rate": 7.663476168347818e-06,
"loss": 0.7658,
"step": 879
},
{
"epoch": 0.6700932800304588,
"grad_norm": 37.91471481323242,
"learning_rate": 7.658317573810572e-06,
"loss": 0.8842,
"step": 880
},
{
"epoch": 0.670854749666857,
"grad_norm": 31.70551872253418,
"learning_rate": 7.653155031604356e-06,
"loss": 0.7191,
"step": 881
},
{
"epoch": 0.6716162193032553,
"grad_norm": 31.07270050048828,
"learning_rate": 7.647988549395676e-06,
"loss": 0.8618,
"step": 882
},
{
"epoch": 0.6723776889396536,
"grad_norm": 32.019893646240234,
"learning_rate": 7.642818134856888e-06,
"loss": 0.7914,
"step": 883
},
{
"epoch": 0.6731391585760518,
"grad_norm": 30.21106719970703,
"learning_rate": 7.637643795666192e-06,
"loss": 0.6793,
"step": 884
},
{
"epoch": 0.67390062821245,
"grad_norm": 23.416296005249023,
"learning_rate": 7.632465539507608e-06,
"loss": 0.6975,
"step": 885
},
{
"epoch": 0.6746620978488482,
"grad_norm": 28.055620193481445,
"learning_rate": 7.62728337407098e-06,
"loss": 0.827,
"step": 886
},
{
"epoch": 0.6754235674852466,
"grad_norm": 24.000978469848633,
"learning_rate": 7.622097307051957e-06,
"loss": 0.5859,
"step": 887
},
{
"epoch": 0.6761850371216448,
"grad_norm": 29.766515731811523,
"learning_rate": 7.616907346151977e-06,
"loss": 0.9002,
"step": 888
},
{
"epoch": 0.676946506758043,
"grad_norm": 28.786598205566406,
"learning_rate": 7.611713499078265e-06,
"loss": 0.8271,
"step": 889
},
{
"epoch": 0.6777079763944412,
"grad_norm": 30.589488983154297,
"learning_rate": 7.606515773543814e-06,
"loss": 0.6112,
"step": 890
},
{
"epoch": 0.6784694460308395,
"grad_norm": 24.776464462280273,
"learning_rate": 7.601314177267381e-06,
"loss": 0.7205,
"step": 891
},
{
"epoch": 0.6792309156672378,
"grad_norm": 26.0139217376709,
"learning_rate": 7.596108717973465e-06,
"loss": 0.7979,
"step": 892
},
{
"epoch": 0.679992385303636,
"grad_norm": 66.2892074584961,
"learning_rate": 7.590899403392307e-06,
"loss": 0.7666,
"step": 893
},
{
"epoch": 0.6807538549400343,
"grad_norm": 30.251981735229492,
"learning_rate": 7.585686241259871e-06,
"loss": 0.7796,
"step": 894
},
{
"epoch": 0.6815153245764325,
"grad_norm": 30.38053321838379,
"learning_rate": 7.5804692393178334e-06,
"loss": 0.8498,
"step": 895
},
{
"epoch": 0.6822767942128307,
"grad_norm": 34.0738639831543,
"learning_rate": 7.575248405313578e-06,
"loss": 0.9128,
"step": 896
},
{
"epoch": 0.683038263849229,
"grad_norm": 28.353290557861328,
"learning_rate": 7.5700237470001705e-06,
"loss": 0.7225,
"step": 897
},
{
"epoch": 0.6837997334856273,
"grad_norm": 25.537622451782227,
"learning_rate": 7.564795272136365e-06,
"loss": 0.6939,
"step": 898
},
{
"epoch": 0.6845612031220255,
"grad_norm": 28.82963752746582,
"learning_rate": 7.559562988486576e-06,
"loss": 0.6956,
"step": 899
},
{
"epoch": 0.6853226727584237,
"grad_norm": 24.72161865234375,
"learning_rate": 7.554326903820881e-06,
"loss": 0.7923,
"step": 900
},
{
"epoch": 0.686084142394822,
"grad_norm": 33.14993667602539,
"learning_rate": 7.549087025914995e-06,
"loss": 0.8066,
"step": 901
},
{
"epoch": 0.6868456120312203,
"grad_norm": 35.39863204956055,
"learning_rate": 7.5438433625502725e-06,
"loss": 0.9213,
"step": 902
},
{
"epoch": 0.6876070816676185,
"grad_norm": 29.69724464416504,
"learning_rate": 7.538595921513683e-06,
"loss": 0.7398,
"step": 903
},
{
"epoch": 0.6883685513040168,
"grad_norm": 35.563053131103516,
"learning_rate": 7.533344710597814e-06,
"loss": 0.7713,
"step": 904
},
{
"epoch": 0.689130020940415,
"grad_norm": 30.227357864379883,
"learning_rate": 7.5280897376008415e-06,
"loss": 0.7612,
"step": 905
},
{
"epoch": 0.6898914905768132,
"grad_norm": 27.88680076599121,
"learning_rate": 7.522831010326538e-06,
"loss": 0.7491,
"step": 906
},
{
"epoch": 0.6906529602132115,
"grad_norm": 37.14356994628906,
"learning_rate": 7.5175685365842435e-06,
"loss": 0.9955,
"step": 907
},
{
"epoch": 0.6914144298496098,
"grad_norm": 28.307918548583984,
"learning_rate": 7.512302324188867e-06,
"loss": 0.8756,
"step": 908
},
{
"epoch": 0.692175899486008,
"grad_norm": 36.45597457885742,
"learning_rate": 7.507032380960868e-06,
"loss": 0.8675,
"step": 909
},
{
"epoch": 0.6929373691224062,
"grad_norm": 34.10755157470703,
"learning_rate": 7.5017587147262425e-06,
"loss": 1.0245,
"step": 910
},
{
"epoch": 0.6936988387588044,
"grad_norm": 33.00735092163086,
"learning_rate": 7.496481333316522e-06,
"loss": 0.7733,
"step": 911
},
{
"epoch": 0.6944603083952028,
"grad_norm": 36.1781005859375,
"learning_rate": 7.491200244568748e-06,
"loss": 0.9425,
"step": 912
},
{
"epoch": 0.695221778031601,
"grad_norm": 25.176944732666016,
"learning_rate": 7.4859154563254766e-06,
"loss": 0.7413,
"step": 913
},
{
"epoch": 0.6959832476679992,
"grad_norm": 29.096973419189453,
"learning_rate": 7.480626976434746e-06,
"loss": 0.7264,
"step": 914
},
{
"epoch": 0.6967447173043975,
"grad_norm": 28.575490951538086,
"learning_rate": 7.475334812750089e-06,
"loss": 0.7541,
"step": 915
},
{
"epoch": 0.6975061869407957,
"grad_norm": 28.50647735595703,
"learning_rate": 7.470038973130497e-06,
"loss": 0.6062,
"step": 916
},
{
"epoch": 0.698267656577194,
"grad_norm": 30.17618179321289,
"learning_rate": 7.464739465440431e-06,
"loss": 0.8535,
"step": 917
},
{
"epoch": 0.6990291262135923,
"grad_norm": 21.86348533630371,
"learning_rate": 7.4594362975497896e-06,
"loss": 0.5617,
"step": 918
},
{
"epoch": 0.6997905958499905,
"grad_norm": 23.84324073791504,
"learning_rate": 7.454129477333916e-06,
"loss": 0.6717,
"step": 919
},
{
"epoch": 0.7005520654863887,
"grad_norm": 29.712072372436523,
"learning_rate": 7.44881901267357e-06,
"loss": 0.9733,
"step": 920
},
{
"epoch": 0.7013135351227869,
"grad_norm": 27.97547149658203,
"learning_rate": 7.443504911454927e-06,
"loss": 0.7411,
"step": 921
},
{
"epoch": 0.7020750047591853,
"grad_norm": 32.59486770629883,
"learning_rate": 7.43818718156956e-06,
"loss": 0.8966,
"step": 922
},
{
"epoch": 0.7028364743955835,
"grad_norm": 31.34674644470215,
"learning_rate": 7.432865830914435e-06,
"loss": 0.9788,
"step": 923
},
{
"epoch": 0.7035979440319817,
"grad_norm": 18.920564651489258,
"learning_rate": 7.427540867391894e-06,
"loss": 0.6062,
"step": 924
},
{
"epoch": 0.70435941366838,
"grad_norm": 21.33208465576172,
"learning_rate": 7.422212298909638e-06,
"loss": 0.6215,
"step": 925
},
{
"epoch": 0.7051208833047782,
"grad_norm": 32.146976470947266,
"learning_rate": 7.416880133380732e-06,
"loss": 0.8168,
"step": 926
},
{
"epoch": 0.7058823529411765,
"grad_norm": 25.904571533203125,
"learning_rate": 7.411544378723572e-06,
"loss": 0.7482,
"step": 927
},
{
"epoch": 0.7066438225775747,
"grad_norm": 29.501108169555664,
"learning_rate": 7.406205042861892e-06,
"loss": 0.778,
"step": 928
},
{
"epoch": 0.707405292213973,
"grad_norm": 38.986534118652344,
"learning_rate": 7.400862133724739e-06,
"loss": 0.891,
"step": 929
},
{
"epoch": 0.7081667618503712,
"grad_norm": 24.509849548339844,
"learning_rate": 7.39551565924647e-06,
"loss": 0.7243,
"step": 930
},
{
"epoch": 0.7089282314867694,
"grad_norm": 24.55140495300293,
"learning_rate": 7.390165627366735e-06,
"loss": 0.7393,
"step": 931
},
{
"epoch": 0.7096897011231678,
"grad_norm": 28.782194137573242,
"learning_rate": 7.384812046030468e-06,
"loss": 0.7491,
"step": 932
},
{
"epoch": 0.710451170759566,
"grad_norm": 26.1166934967041,
"learning_rate": 7.379454923187868e-06,
"loss": 0.667,
"step": 933
},
{
"epoch": 0.7112126403959642,
"grad_norm": 30.481565475463867,
"learning_rate": 7.374094266794403e-06,
"loss": 0.8279,
"step": 934
},
{
"epoch": 0.7119741100323624,
"grad_norm": 28.972707748413086,
"learning_rate": 7.368730084810782e-06,
"loss": 0.6376,
"step": 935
},
{
"epoch": 0.7127355796687607,
"grad_norm": 26.157737731933594,
"learning_rate": 7.363362385202952e-06,
"loss": 0.8613,
"step": 936
},
{
"epoch": 0.713497049305159,
"grad_norm": 26.83064842224121,
"learning_rate": 7.357991175942083e-06,
"loss": 0.777,
"step": 937
},
{
"epoch": 0.7142585189415572,
"grad_norm": 26.857431411743164,
"learning_rate": 7.352616465004556e-06,
"loss": 0.6033,
"step": 938
},
{
"epoch": 0.7150199885779555,
"grad_norm": 28.428842544555664,
"learning_rate": 7.347238260371954e-06,
"loss": 0.9263,
"step": 939
},
{
"epoch": 0.7157814582143537,
"grad_norm": 28.889423370361328,
"learning_rate": 7.341856570031045e-06,
"loss": 0.6927,
"step": 940
},
{
"epoch": 0.7165429278507519,
"grad_norm": 33.6306266784668,
"learning_rate": 7.336471401973778e-06,
"loss": 0.7787,
"step": 941
},
{
"epoch": 0.7173043974871502,
"grad_norm": 23.571929931640625,
"learning_rate": 7.331082764197262e-06,
"loss": 0.565,
"step": 942
},
{
"epoch": 0.7180658671235485,
"grad_norm": 29.913192749023438,
"learning_rate": 7.325690664703763e-06,
"loss": 0.8598,
"step": 943
},
{
"epoch": 0.7188273367599467,
"grad_norm": 30.621116638183594,
"learning_rate": 7.320295111500683e-06,
"loss": 0.6628,
"step": 944
},
{
"epoch": 0.7195888063963449,
"grad_norm": 35.32845687866211,
"learning_rate": 7.314896112600558e-06,
"loss": 0.9929,
"step": 945
},
{
"epoch": 0.7203502760327432,
"grad_norm": 27.44571304321289,
"learning_rate": 7.309493676021035e-06,
"loss": 0.8326,
"step": 946
},
{
"epoch": 0.7211117456691415,
"grad_norm": 32.142967224121094,
"learning_rate": 7.30408780978487e-06,
"loss": 0.8856,
"step": 947
},
{
"epoch": 0.7218732153055397,
"grad_norm": 30.383384704589844,
"learning_rate": 7.298678521919914e-06,
"loss": 0.7051,
"step": 948
},
{
"epoch": 0.7226346849419379,
"grad_norm": 29.19371795654297,
"learning_rate": 7.293265820459095e-06,
"loss": 0.8118,
"step": 949
},
{
"epoch": 0.7233961545783362,
"grad_norm": 31.852752685546875,
"learning_rate": 7.28784971344041e-06,
"loss": 0.7124,
"step": 950
},
{
"epoch": 0.7241576242147344,
"grad_norm": 26.42278289794922,
"learning_rate": 7.2824302089069184e-06,
"loss": 0.7328,
"step": 951
},
{
"epoch": 0.7249190938511327,
"grad_norm": 31.66285514831543,
"learning_rate": 7.277007314906719e-06,
"loss": 0.814,
"step": 952
},
{
"epoch": 0.725680563487531,
"grad_norm": 26.553312301635742,
"learning_rate": 7.271581039492949e-06,
"loss": 0.7502,
"step": 953
},
{
"epoch": 0.7264420331239292,
"grad_norm": 36.027652740478516,
"learning_rate": 7.2661513907237634e-06,
"loss": 1.0965,
"step": 954
},
{
"epoch": 0.7272035027603274,
"grad_norm": 25.32257652282715,
"learning_rate": 7.260718376662329e-06,
"loss": 0.8641,
"step": 955
},
{
"epoch": 0.7279649723967256,
"grad_norm": 32.486141204833984,
"learning_rate": 7.25528200537681e-06,
"loss": 0.8929,
"step": 956
},
{
"epoch": 0.728726442033124,
"grad_norm": 46.99209976196289,
"learning_rate": 7.249842284940352e-06,
"loss": 0.8834,
"step": 957
},
{
"epoch": 0.7294879116695222,
"grad_norm": 25.190576553344727,
"learning_rate": 7.244399223431082e-06,
"loss": 0.6918,
"step": 958
},
{
"epoch": 0.7302493813059204,
"grad_norm": 28.75440788269043,
"learning_rate": 7.238952828932084e-06,
"loss": 0.8444,
"step": 959
},
{
"epoch": 0.7310108509423187,
"grad_norm": 26.397907257080078,
"learning_rate": 7.233503109531389e-06,
"loss": 0.6744,
"step": 960
},
{
"epoch": 0.7317723205787169,
"grad_norm": 34.44126892089844,
"learning_rate": 7.228050073321971e-06,
"loss": 0.8709,
"step": 961
},
{
"epoch": 0.7325337902151152,
"grad_norm": 28.410762786865234,
"learning_rate": 7.222593728401724e-06,
"loss": 0.8122,
"step": 962
},
{
"epoch": 0.7332952598515134,
"grad_norm": 30.714998245239258,
"learning_rate": 7.2171340828734626e-06,
"loss": 0.7721,
"step": 963
},
{
"epoch": 0.7340567294879117,
"grad_norm": 24.129196166992188,
"learning_rate": 7.211671144844896e-06,
"loss": 0.5979,
"step": 964
},
{
"epoch": 0.7348181991243099,
"grad_norm": 36.182044982910156,
"learning_rate": 7.206204922428625e-06,
"loss": 0.9329,
"step": 965
},
{
"epoch": 0.7355796687607081,
"grad_norm": 28.672395706176758,
"learning_rate": 7.200735423742131e-06,
"loss": 0.7124,
"step": 966
},
{
"epoch": 0.7363411383971065,
"grad_norm": 23.973031997680664,
"learning_rate": 7.195262656907758e-06,
"loss": 0.7238,
"step": 967
},
{
"epoch": 0.7371026080335047,
"grad_norm": 26.83302116394043,
"learning_rate": 7.1897866300527e-06,
"loss": 0.8061,
"step": 968
},
{
"epoch": 0.7378640776699029,
"grad_norm": 24.463834762573242,
"learning_rate": 7.1843073513089985e-06,
"loss": 0.6155,
"step": 969
},
{
"epoch": 0.7386255473063011,
"grad_norm": 27.022062301635742,
"learning_rate": 7.178824828813522e-06,
"loss": 0.6644,
"step": 970
},
{
"epoch": 0.7393870169426994,
"grad_norm": 28.182111740112305,
"learning_rate": 7.173339070707953e-06,
"loss": 0.8427,
"step": 971
},
{
"epoch": 0.7401484865790977,
"grad_norm": 33.68251037597656,
"learning_rate": 7.1678500851387835e-06,
"loss": 0.873,
"step": 972
},
{
"epoch": 0.7409099562154959,
"grad_norm": 28.491044998168945,
"learning_rate": 7.162357880257294e-06,
"loss": 0.8262,
"step": 973
},
{
"epoch": 0.7416714258518942,
"grad_norm": 28.93327522277832,
"learning_rate": 7.1568624642195495e-06,
"loss": 0.7282,
"step": 974
},
{
"epoch": 0.7424328954882924,
"grad_norm": 22.979516983032227,
"learning_rate": 7.151363845186382e-06,
"loss": 0.5553,
"step": 975
},
{
"epoch": 0.7431943651246906,
"grad_norm": 28.500520706176758,
"learning_rate": 7.14586203132338e-06,
"loss": 0.8657,
"step": 976
},
{
"epoch": 0.743955834761089,
"grad_norm": 38.0922966003418,
"learning_rate": 7.140357030800875e-06,
"loss": 0.9531,
"step": 977
},
{
"epoch": 0.7447173043974872,
"grad_norm": 30.303611755371094,
"learning_rate": 7.134848851793934e-06,
"loss": 0.7469,
"step": 978
},
{
"epoch": 0.7454787740338854,
"grad_norm": 34.94667053222656,
"learning_rate": 7.129337502482343e-06,
"loss": 0.9157,
"step": 979
},
{
"epoch": 0.7462402436702836,
"grad_norm": 28.89977264404297,
"learning_rate": 7.123822991050594e-06,
"loss": 0.7208,
"step": 980
},
{
"epoch": 0.7470017133066819,
"grad_norm": 28.439899444580078,
"learning_rate": 7.118305325687876e-06,
"loss": 0.7317,
"step": 981
},
{
"epoch": 0.7477631829430802,
"grad_norm": 26.193099975585938,
"learning_rate": 7.112784514588064e-06,
"loss": 0.6008,
"step": 982
},
{
"epoch": 0.7485246525794784,
"grad_norm": 25.30535125732422,
"learning_rate": 7.1072605659497e-06,
"loss": 0.754,
"step": 983
},
{
"epoch": 0.7492861222158766,
"grad_norm": 23.79256248474121,
"learning_rate": 7.101733487975988e-06,
"loss": 0.6873,
"step": 984
},
{
"epoch": 0.7500475918522749,
"grad_norm": 26.857532501220703,
"learning_rate": 7.096203288874781e-06,
"loss": 0.9162,
"step": 985
},
{
"epoch": 0.7508090614886731,
"grad_norm": 31.37613296508789,
"learning_rate": 7.090669976858563e-06,
"loss": 0.7478,
"step": 986
},
{
"epoch": 0.7515705311250714,
"grad_norm": 26.164493560791016,
"learning_rate": 7.085133560144443e-06,
"loss": 0.7949,
"step": 987
},
{
"epoch": 0.7523320007614697,
"grad_norm": 31.36321449279785,
"learning_rate": 7.07959404695414e-06,
"loss": 0.751,
"step": 988
},
{
"epoch": 0.7530934703978679,
"grad_norm": 27.06949806213379,
"learning_rate": 7.074051445513972e-06,
"loss": 0.7524,
"step": 989
},
{
"epoch": 0.7538549400342661,
"grad_norm": 33.156089782714844,
"learning_rate": 7.0685057640548415e-06,
"loss": 0.838,
"step": 990
},
{
"epoch": 0.7546164096706643,
"grad_norm": 37.368370056152344,
"learning_rate": 7.062957010812227e-06,
"loss": 0.8547,
"step": 991
},
{
"epoch": 0.7553778793070627,
"grad_norm": 32.58074951171875,
"learning_rate": 7.057405194026168e-06,
"loss": 0.6642,
"step": 992
},
{
"epoch": 0.7561393489434609,
"grad_norm": 28.085039138793945,
"learning_rate": 7.051850321941252e-06,
"loss": 0.7924,
"step": 993
},
{
"epoch": 0.7569008185798591,
"grad_norm": 26.330997467041016,
"learning_rate": 7.046292402806605e-06,
"loss": 0.7999,
"step": 994
},
{
"epoch": 0.7576622882162574,
"grad_norm": 38.67786407470703,
"learning_rate": 7.040731444875878e-06,
"loss": 1.0057,
"step": 995
},
{
"epoch": 0.7584237578526556,
"grad_norm": 25.687808990478516,
"learning_rate": 7.035167456407237e-06,
"loss": 0.712,
"step": 996
},
{
"epoch": 0.7591852274890539,
"grad_norm": 30.188947677612305,
"learning_rate": 7.029600445663341e-06,
"loss": 0.7298,
"step": 997
},
{
"epoch": 0.7599466971254522,
"grad_norm": 26.661039352416992,
"learning_rate": 7.024030420911346e-06,
"loss": 0.8296,
"step": 998
},
{
"epoch": 0.7607081667618504,
"grad_norm": 31.66457176208496,
"learning_rate": 7.018457390422878e-06,
"loss": 0.84,
"step": 999
},
{
"epoch": 0.7614696363982486,
"grad_norm": 26.696500778198242,
"learning_rate": 7.012881362474029e-06,
"loss": 0.8177,
"step": 1000
}
],
"logging_steps": 1,
"max_steps": 2628,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 5.505600572322048e+16,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}