lmg-anon's picture
Upload folder using huggingface_hub
0ebac04 verified
raw
history blame contribute delete
No virus
52.6 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.39082855653986454,
"eval_steps": 500,
"global_step": 300,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0013027618551328818,
"grad_norm": 14.949752807617188,
"learning_rate": 4.0625e-06,
"loss": 1.3439,
"step": 1
},
{
"epoch": 0.0026055237102657635,
"grad_norm": 15.725403785705566,
"learning_rate": 8.125e-06,
"loss": 1.3354,
"step": 2
},
{
"epoch": 0.0039082855653986455,
"grad_norm": 12.41278076171875,
"learning_rate": 1.2187499999999998e-05,
"loss": 1.3017,
"step": 3
},
{
"epoch": 0.005211047420531527,
"grad_norm": 6.782334327697754,
"learning_rate": 1.625e-05,
"loss": 1.2634,
"step": 4
},
{
"epoch": 0.006513809275664409,
"grad_norm": 3.730219841003418,
"learning_rate": 2.03125e-05,
"loss": 1.1704,
"step": 5
},
{
"epoch": 0.007816571130797291,
"grad_norm": 2.547757863998413,
"learning_rate": 2.4374999999999996e-05,
"loss": 1.165,
"step": 6
},
{
"epoch": 0.009119332985930172,
"grad_norm": 2.4353976249694824,
"learning_rate": 2.8437499999999997e-05,
"loss": 1.1567,
"step": 7
},
{
"epoch": 0.010422094841063054,
"grad_norm": 2.218895673751831,
"learning_rate": 3.25e-05,
"loss": 1.0846,
"step": 8
},
{
"epoch": 0.011724856696195935,
"grad_norm": 1.7433233261108398,
"learning_rate": 3.6562499999999994e-05,
"loss": 0.9816,
"step": 9
},
{
"epoch": 0.013027618551328817,
"grad_norm": 1.6148579120635986,
"learning_rate": 4.0625e-05,
"loss": 1.0869,
"step": 10
},
{
"epoch": 0.0143303804064617,
"grad_norm": 1.5065230131149292,
"learning_rate": 4.4687499999999995e-05,
"loss": 0.9625,
"step": 11
},
{
"epoch": 0.015633142261594582,
"grad_norm": 1.4786823987960815,
"learning_rate": 4.874999999999999e-05,
"loss": 1.0489,
"step": 12
},
{
"epoch": 0.016935904116727463,
"grad_norm": 1.2991230487823486,
"learning_rate": 5.2812499999999996e-05,
"loss": 0.9935,
"step": 13
},
{
"epoch": 0.018238665971860343,
"grad_norm": 1.3199464082717896,
"learning_rate": 5.687499999999999e-05,
"loss": 1.0188,
"step": 14
},
{
"epoch": 0.019541427826993224,
"grad_norm": 1.2793257236480713,
"learning_rate": 6.09375e-05,
"loss": 1.0265,
"step": 15
},
{
"epoch": 0.020844189682126108,
"grad_norm": 1.2210655212402344,
"learning_rate": 6.5e-05,
"loss": 0.9019,
"step": 16
},
{
"epoch": 0.02214695153725899,
"grad_norm": 1.3837445974349976,
"learning_rate": 6.499971563731688e-05,
"loss": 0.9487,
"step": 17
},
{
"epoch": 0.02344971339239187,
"grad_norm": 1.2946529388427734,
"learning_rate": 6.499886255424368e-05,
"loss": 1.0031,
"step": 18
},
{
"epoch": 0.024752475247524754,
"grad_norm": 1.3142110109329224,
"learning_rate": 6.499744076570868e-05,
"loss": 0.9941,
"step": 19
},
{
"epoch": 0.026055237102657634,
"grad_norm": 1.8206599950790405,
"learning_rate": 6.499545029659211e-05,
"loss": 0.9767,
"step": 20
},
{
"epoch": 0.027357998957790515,
"grad_norm": 1.2736281156539917,
"learning_rate": 6.499289118172568e-05,
"loss": 0.9194,
"step": 21
},
{
"epoch": 0.0286607608129234,
"grad_norm": 1.8105640411376953,
"learning_rate": 6.498976346589196e-05,
"loss": 0.9853,
"step": 22
},
{
"epoch": 0.02996352266805628,
"grad_norm": 1.4621827602386475,
"learning_rate": 6.49860672038236e-05,
"loss": 0.9896,
"step": 23
},
{
"epoch": 0.031266284523189164,
"grad_norm": 1.212299108505249,
"learning_rate": 6.498180246020238e-05,
"loss": 0.9684,
"step": 24
},
{
"epoch": 0.032569046378322045,
"grad_norm": 1.2274290323257446,
"learning_rate": 6.49769693096581e-05,
"loss": 0.9831,
"step": 25
},
{
"epoch": 0.033871808233454925,
"grad_norm": 1.2909352779388428,
"learning_rate": 6.49715678367672e-05,
"loss": 0.9971,
"step": 26
},
{
"epoch": 0.035174570088587806,
"grad_norm": 1.2330715656280518,
"learning_rate": 6.49655981360514e-05,
"loss": 0.9923,
"step": 27
},
{
"epoch": 0.03647733194372069,
"grad_norm": 1.2433793544769287,
"learning_rate": 6.495906031197591e-05,
"loss": 0.9121,
"step": 28
},
{
"epoch": 0.03778009379885357,
"grad_norm": 1.2374027967453003,
"learning_rate": 6.49519544789477e-05,
"loss": 0.9224,
"step": 29
},
{
"epoch": 0.03908285565398645,
"grad_norm": 1.2204357385635376,
"learning_rate": 6.494428076131346e-05,
"loss": 0.919,
"step": 30
},
{
"epoch": 0.040385617509119336,
"grad_norm": 1.1927292346954346,
"learning_rate": 6.493603929335744e-05,
"loss": 0.9589,
"step": 31
},
{
"epoch": 0.041688379364252216,
"grad_norm": 1.1353641748428345,
"learning_rate": 6.492723021929908e-05,
"loss": 0.8869,
"step": 32
},
{
"epoch": 0.0429911412193851,
"grad_norm": 1.212263822555542,
"learning_rate": 6.49178536932905e-05,
"loss": 0.8934,
"step": 33
},
{
"epoch": 0.04429390307451798,
"grad_norm": 1.0893241167068481,
"learning_rate": 6.49079098794138e-05,
"loss": 0.8718,
"step": 34
},
{
"epoch": 0.04559666492965086,
"grad_norm": 1.1801607608795166,
"learning_rate": 6.489739895167818e-05,
"loss": 0.9774,
"step": 35
},
{
"epoch": 0.04689942678478374,
"grad_norm": 1.1302330493927002,
"learning_rate": 6.488632109401691e-05,
"loss": 0.9584,
"step": 36
},
{
"epoch": 0.04820218863991663,
"grad_norm": 1.1432682275772095,
"learning_rate": 6.487467650028411e-05,
"loss": 0.9452,
"step": 37
},
{
"epoch": 0.04950495049504951,
"grad_norm": 1.1399481296539307,
"learning_rate": 6.486246537425133e-05,
"loss": 0.9017,
"step": 38
},
{
"epoch": 0.05080771235018239,
"grad_norm": 1.0911235809326172,
"learning_rate": 6.484968792960403e-05,
"loss": 0.9036,
"step": 39
},
{
"epoch": 0.05211047420531527,
"grad_norm": 1.071078896522522,
"learning_rate": 6.48363443899378e-05,
"loss": 0.9273,
"step": 40
},
{
"epoch": 0.05341323606044815,
"grad_norm": 1.053589105606079,
"learning_rate": 6.482243498875449e-05,
"loss": 0.8847,
"step": 41
},
{
"epoch": 0.05471599791558103,
"grad_norm": 1.109771490097046,
"learning_rate": 6.480795996945806e-05,
"loss": 0.9364,
"step": 42
},
{
"epoch": 0.05601875977071391,
"grad_norm": 1.1093286275863647,
"learning_rate": 6.479291958535037e-05,
"loss": 0.9088,
"step": 43
},
{
"epoch": 0.0573215216258468,
"grad_norm": 1.0716255903244019,
"learning_rate": 6.477731409962675e-05,
"loss": 0.8885,
"step": 44
},
{
"epoch": 0.05862428348097968,
"grad_norm": 1.1552095413208008,
"learning_rate": 6.476114378537138e-05,
"loss": 0.9736,
"step": 45
},
{
"epoch": 0.05992704533611256,
"grad_norm": 1.1647394895553589,
"learning_rate": 6.474440892555249e-05,
"loss": 0.9138,
"step": 46
},
{
"epoch": 0.06122980719124544,
"grad_norm": 1.1463412046432495,
"learning_rate": 6.472710981301745e-05,
"loss": 0.9204,
"step": 47
},
{
"epoch": 0.06253256904637833,
"grad_norm": 1.175421118736267,
"learning_rate": 6.470924675048762e-05,
"loss": 0.9119,
"step": 48
},
{
"epoch": 0.06383533090151121,
"grad_norm": 1.9249529838562012,
"learning_rate": 6.469082005055304e-05,
"loss": 0.8639,
"step": 49
},
{
"epoch": 0.06513809275664409,
"grad_norm": 1.1802655458450317,
"learning_rate": 6.467183003566701e-05,
"loss": 0.9352,
"step": 50
},
{
"epoch": 0.06644085461177697,
"grad_norm": 1.1035569906234741,
"learning_rate": 6.465227703814038e-05,
"loss": 0.9342,
"step": 51
},
{
"epoch": 0.06774361646690985,
"grad_norm": 1.1226942539215088,
"learning_rate": 6.463216140013581e-05,
"loss": 0.8948,
"step": 52
},
{
"epoch": 0.06904637832204273,
"grad_norm": 1.1143085956573486,
"learning_rate": 6.46114834736617e-05,
"loss": 0.9577,
"step": 53
},
{
"epoch": 0.07034914017717561,
"grad_norm": 1.1214665174484253,
"learning_rate": 6.459024362056608e-05,
"loss": 0.9106,
"step": 54
},
{
"epoch": 0.07165190203230849,
"grad_norm": 1.1105942726135254,
"learning_rate": 6.45684422125303e-05,
"loss": 0.8706,
"step": 55
},
{
"epoch": 0.07295466388744137,
"grad_norm": 1.0562191009521484,
"learning_rate": 6.454607963106247e-05,
"loss": 0.8278,
"step": 56
},
{
"epoch": 0.07425742574257425,
"grad_norm": 1.1298907995224,
"learning_rate": 6.452315626749081e-05,
"loss": 0.8883,
"step": 57
},
{
"epoch": 0.07556018759770713,
"grad_norm": 1.133426547050476,
"learning_rate": 6.449967252295682e-05,
"loss": 0.8868,
"step": 58
},
{
"epoch": 0.07686294945284002,
"grad_norm": 1.09175443649292,
"learning_rate": 6.447562880840822e-05,
"loss": 0.8486,
"step": 59
},
{
"epoch": 0.0781657113079729,
"grad_norm": 1.1153368949890137,
"learning_rate": 6.44510255445918e-05,
"loss": 0.9278,
"step": 60
},
{
"epoch": 0.07946847316310579,
"grad_norm": 1.102013349533081,
"learning_rate": 6.442586316204601e-05,
"loss": 0.8362,
"step": 61
},
{
"epoch": 0.08077123501823867,
"grad_norm": 1.1041877269744873,
"learning_rate": 6.44001421010935e-05,
"loss": 0.8746,
"step": 62
},
{
"epoch": 0.08207399687337155,
"grad_norm": 1.075576901435852,
"learning_rate": 6.437386281183331e-05,
"loss": 0.8567,
"step": 63
},
{
"epoch": 0.08337675872850443,
"grad_norm": 1.1472443342208862,
"learning_rate": 6.43470257541331e-05,
"loss": 0.9023,
"step": 64
},
{
"epoch": 0.08467952058363731,
"grad_norm": 1.1024972200393677,
"learning_rate": 6.431963139762104e-05,
"loss": 0.8982,
"step": 65
},
{
"epoch": 0.0859822824387702,
"grad_norm": 1.1128768920898438,
"learning_rate": 6.42916802216776e-05,
"loss": 0.883,
"step": 66
},
{
"epoch": 0.08728504429390307,
"grad_norm": 1.058384895324707,
"learning_rate": 6.426317271542717e-05,
"loss": 0.8428,
"step": 67
},
{
"epoch": 0.08858780614903596,
"grad_norm": 1.0728991031646729,
"learning_rate": 6.423410937772951e-05,
"loss": 0.8553,
"step": 68
},
{
"epoch": 0.08989056800416884,
"grad_norm": 1.0956963300704956,
"learning_rate": 6.4204490717171e-05,
"loss": 0.8894,
"step": 69
},
{
"epoch": 0.09119332985930172,
"grad_norm": 1.0787287950515747,
"learning_rate": 6.417431725205575e-05,
"loss": 0.9086,
"step": 70
},
{
"epoch": 0.0924960917144346,
"grad_norm": 1.1435580253601074,
"learning_rate": 6.414358951039651e-05,
"loss": 0.9074,
"step": 71
},
{
"epoch": 0.09379885356956748,
"grad_norm": 1.0396053791046143,
"learning_rate": 6.41123080299055e-05,
"loss": 0.8553,
"step": 72
},
{
"epoch": 0.09510161542470036,
"grad_norm": 1.0929105281829834,
"learning_rate": 6.40804733579849e-05,
"loss": 0.8504,
"step": 73
},
{
"epoch": 0.09640437727983325,
"grad_norm": 1.1395248174667358,
"learning_rate": 6.404808605171734e-05,
"loss": 0.9222,
"step": 74
},
{
"epoch": 0.09770713913496613,
"grad_norm": 1.0281647443771362,
"learning_rate": 6.401514667785613e-05,
"loss": 0.868,
"step": 75
},
{
"epoch": 0.09900990099009901,
"grad_norm": 1.0565054416656494,
"learning_rate": 6.398165581281536e-05,
"loss": 0.8302,
"step": 76
},
{
"epoch": 0.1003126628452319,
"grad_norm": 1.0529290437698364,
"learning_rate": 6.394761404265977e-05,
"loss": 0.8669,
"step": 77
},
{
"epoch": 0.10161542470036478,
"grad_norm": 1.0579882860183716,
"learning_rate": 6.391302196309452e-05,
"loss": 0.9204,
"step": 78
},
{
"epoch": 0.10291818655549766,
"grad_norm": 1.0531302690505981,
"learning_rate": 6.387788017945482e-05,
"loss": 0.9202,
"step": 79
},
{
"epoch": 0.10422094841063054,
"grad_norm": 1.0358260869979858,
"learning_rate": 6.384218930669521e-05,
"loss": 0.8647,
"step": 80
},
{
"epoch": 0.10552371026576342,
"grad_norm": 1.0599216222763062,
"learning_rate": 6.380594996937895e-05,
"loss": 0.9093,
"step": 81
},
{
"epoch": 0.1068264721208963,
"grad_norm": 1.2026344537734985,
"learning_rate": 6.376916280166696e-05,
"loss": 0.8731,
"step": 82
},
{
"epoch": 0.10812923397602918,
"grad_norm": 1.1054397821426392,
"learning_rate": 6.373182844730677e-05,
"loss": 0.8835,
"step": 83
},
{
"epoch": 0.10943199583116206,
"grad_norm": 1.030562162399292,
"learning_rate": 6.369394755962129e-05,
"loss": 0.8432,
"step": 84
},
{
"epoch": 0.11073475768629494,
"grad_norm": 1.0522128343582153,
"learning_rate": 6.365552080149737e-05,
"loss": 0.8598,
"step": 85
},
{
"epoch": 0.11203751954142782,
"grad_norm": 1.0692416429519653,
"learning_rate": 6.361654884537413e-05,
"loss": 0.8812,
"step": 86
},
{
"epoch": 0.1133402813965607,
"grad_norm": 1.1009373664855957,
"learning_rate": 6.357703237323127e-05,
"loss": 0.8635,
"step": 87
},
{
"epoch": 0.1146430432516936,
"grad_norm": 1.1628373861312866,
"learning_rate": 6.353697207657708e-05,
"loss": 0.8907,
"step": 88
},
{
"epoch": 0.11594580510682648,
"grad_norm": 1.1273458003997803,
"learning_rate": 6.349636865643642e-05,
"loss": 0.8445,
"step": 89
},
{
"epoch": 0.11724856696195936,
"grad_norm": 1.0971312522888184,
"learning_rate": 6.345522282333835e-05,
"loss": 0.8543,
"step": 90
},
{
"epoch": 0.11855132881709224,
"grad_norm": 1.0826351642608643,
"learning_rate": 6.341353529730374e-05,
"loss": 0.8605,
"step": 91
},
{
"epoch": 0.11985409067222512,
"grad_norm": 1.0686413049697876,
"learning_rate": 6.337130680783273e-05,
"loss": 0.8814,
"step": 92
},
{
"epoch": 0.121156852527358,
"grad_norm": 1.21683931350708,
"learning_rate": 6.332853809389189e-05,
"loss": 0.8979,
"step": 93
},
{
"epoch": 0.12245961438249088,
"grad_norm": 1.0884461402893066,
"learning_rate": 6.328522990390126e-05,
"loss": 0.8577,
"step": 94
},
{
"epoch": 0.12376237623762376,
"grad_norm": 1.047310709953308,
"learning_rate": 6.324138299572135e-05,
"loss": 0.8174,
"step": 95
},
{
"epoch": 0.12506513809275666,
"grad_norm": 1.087296485900879,
"learning_rate": 6.319699813663982e-05,
"loss": 0.8566,
"step": 96
},
{
"epoch": 0.12636789994788952,
"grad_norm": 1.070327639579773,
"learning_rate": 6.315207610335808e-05,
"loss": 0.8615,
"step": 97
},
{
"epoch": 0.12767066180302242,
"grad_norm": 1.0299792289733887,
"learning_rate": 6.310661768197764e-05,
"loss": 0.866,
"step": 98
},
{
"epoch": 0.12897342365815528,
"grad_norm": 1.1011384725570679,
"learning_rate": 6.306062366798642e-05,
"loss": 0.8775,
"step": 99
},
{
"epoch": 0.13027618551328818,
"grad_norm": 1.0193467140197754,
"learning_rate": 6.301409486624481e-05,
"loss": 0.8243,
"step": 100
},
{
"epoch": 0.13157894736842105,
"grad_norm": 1.0899205207824707,
"learning_rate": 6.296703209097159e-05,
"loss": 0.8849,
"step": 101
},
{
"epoch": 0.13288170922355394,
"grad_norm": 1.0473856925964355,
"learning_rate": 6.291943616572962e-05,
"loss": 0.8847,
"step": 102
},
{
"epoch": 0.1341844710786868,
"grad_norm": 0.9922148585319519,
"learning_rate": 6.287130792341154e-05,
"loss": 0.8157,
"step": 103
},
{
"epoch": 0.1354872329338197,
"grad_norm": 1.0680872201919556,
"learning_rate": 6.282264820622511e-05,
"loss": 0.8802,
"step": 104
},
{
"epoch": 0.13678999478895257,
"grad_norm": 1.034562587738037,
"learning_rate": 6.27734578656785e-05,
"loss": 0.8826,
"step": 105
},
{
"epoch": 0.13809275664408546,
"grad_norm": 1.0247395038604736,
"learning_rate": 6.272373776256536e-05,
"loss": 0.8535,
"step": 106
},
{
"epoch": 0.13939551849921833,
"grad_norm": 1.099061131477356,
"learning_rate": 6.267348876694983e-05,
"loss": 0.8583,
"step": 107
},
{
"epoch": 0.14069828035435122,
"grad_norm": 1.0081039667129517,
"learning_rate": 6.262271175815126e-05,
"loss": 0.7988,
"step": 108
},
{
"epoch": 0.14200104220948412,
"grad_norm": 1.1140209436416626,
"learning_rate": 6.257140762472879e-05,
"loss": 0.8426,
"step": 109
},
{
"epoch": 0.14330380406461699,
"grad_norm": 1.0654839277267456,
"learning_rate": 6.251957726446589e-05,
"loss": 0.8682,
"step": 110
},
{
"epoch": 0.14460656591974988,
"grad_norm": 1.0632318258285522,
"learning_rate": 6.246722158435455e-05,
"loss": 0.8379,
"step": 111
},
{
"epoch": 0.14590932777488275,
"grad_norm": 1.0444918870925903,
"learning_rate": 6.241434150057953e-05,
"loss": 0.8525,
"step": 112
},
{
"epoch": 0.14721208963001564,
"grad_norm": 1.0542877912521362,
"learning_rate": 6.236093793850217e-05,
"loss": 0.834,
"step": 113
},
{
"epoch": 0.1485148514851485,
"grad_norm": 0.9843441247940063,
"learning_rate": 6.230701183264436e-05,
"loss": 0.7711,
"step": 114
},
{
"epoch": 0.1498176133402814,
"grad_norm": 1.0625543594360352,
"learning_rate": 6.225256412667207e-05,
"loss": 0.8169,
"step": 115
},
{
"epoch": 0.15112037519541427,
"grad_norm": 1.0951343774795532,
"learning_rate": 6.219759577337887e-05,
"loss": 0.9114,
"step": 116
},
{
"epoch": 0.15242313705054716,
"grad_norm": 1.0442739725112915,
"learning_rate": 6.214210773466932e-05,
"loss": 0.8038,
"step": 117
},
{
"epoch": 0.15372589890568003,
"grad_norm": 1.0549882650375366,
"learning_rate": 6.2086100981542e-05,
"loss": 0.8677,
"step": 118
},
{
"epoch": 0.15502866076081293,
"grad_norm": 1.0807327032089233,
"learning_rate": 6.202957649407264e-05,
"loss": 0.8276,
"step": 119
},
{
"epoch": 0.1563314226159458,
"grad_norm": 1.0784152746200562,
"learning_rate": 6.197253526139696e-05,
"loss": 0.8823,
"step": 120
},
{
"epoch": 0.1576341844710787,
"grad_norm": 1.076708436012268,
"learning_rate": 6.191497828169325e-05,
"loss": 0.842,
"step": 121
},
{
"epoch": 0.15893694632621158,
"grad_norm": 1.0658226013183594,
"learning_rate": 6.185690656216507e-05,
"loss": 0.8626,
"step": 122
},
{
"epoch": 0.16023970818134445,
"grad_norm": 1.064942717552185,
"learning_rate": 6.179832111902348e-05,
"loss": 0.8735,
"step": 123
},
{
"epoch": 0.16154247003647734,
"grad_norm": 1.0782710313796997,
"learning_rate": 6.17392229774693e-05,
"loss": 0.828,
"step": 124
},
{
"epoch": 0.1628452318916102,
"grad_norm": 1.0276188850402832,
"learning_rate": 6.167961317167528e-05,
"loss": 0.8475,
"step": 125
},
{
"epoch": 0.1641479937467431,
"grad_norm": 1.1249077320098877,
"learning_rate": 6.161949274476778e-05,
"loss": 0.9029,
"step": 126
},
{
"epoch": 0.16545075560187597,
"grad_norm": 1.027971863746643,
"learning_rate": 6.155886274880874e-05,
"loss": 0.8057,
"step": 127
},
{
"epoch": 0.16675351745700887,
"grad_norm": 1.0177087783813477,
"learning_rate": 6.14977242447771e-05,
"loss": 0.7928,
"step": 128
},
{
"epoch": 0.16805627931214173,
"grad_norm": 1.0410820245742798,
"learning_rate": 6.143607830255037e-05,
"loss": 0.8222,
"step": 129
},
{
"epoch": 0.16935904116727463,
"grad_norm": 1.086966872215271,
"learning_rate": 6.137392600088579e-05,
"loss": 0.837,
"step": 130
},
{
"epoch": 0.1706618030224075,
"grad_norm": 1.099645972251892,
"learning_rate": 6.131126842740155e-05,
"loss": 0.8604,
"step": 131
},
{
"epoch": 0.1719645648775404,
"grad_norm": 1.0512402057647705,
"learning_rate": 6.124810667855768e-05,
"loss": 0.8524,
"step": 132
},
{
"epoch": 0.17326732673267325,
"grad_norm": 1.0146111249923706,
"learning_rate": 6.118444185963691e-05,
"loss": 0.7948,
"step": 133
},
{
"epoch": 0.17457008858780615,
"grad_norm": 1.0564138889312744,
"learning_rate": 6.112027508472533e-05,
"loss": 0.8393,
"step": 134
},
{
"epoch": 0.17587285044293904,
"grad_norm": 1.1982516050338745,
"learning_rate": 6.105560747669284e-05,
"loss": 0.8326,
"step": 135
},
{
"epoch": 0.1771756122980719,
"grad_norm": 1.0336800813674927,
"learning_rate": 6.0990440167173595e-05,
"loss": 0.8456,
"step": 136
},
{
"epoch": 0.1784783741532048,
"grad_norm": 1.0412505865097046,
"learning_rate": 6.092477429654611e-05,
"loss": 0.8026,
"step": 137
},
{
"epoch": 0.17978113600833767,
"grad_norm": 1.0531303882598877,
"learning_rate": 6.085861101391333e-05,
"loss": 0.8675,
"step": 138
},
{
"epoch": 0.18108389786347057,
"grad_norm": 1.0367439985275269,
"learning_rate": 6.0791951477082555e-05,
"loss": 0.8328,
"step": 139
},
{
"epoch": 0.18238665971860343,
"grad_norm": 1.057989478111267,
"learning_rate": 6.072479685254518e-05,
"loss": 0.8129,
"step": 140
},
{
"epoch": 0.18368942157373633,
"grad_norm": 1.1223191022872925,
"learning_rate": 6.065714831545621e-05,
"loss": 0.8519,
"step": 141
},
{
"epoch": 0.1849921834288692,
"grad_norm": 1.0480860471725464,
"learning_rate": 6.058900704961377e-05,
"loss": 0.8478,
"step": 142
},
{
"epoch": 0.1862949452840021,
"grad_norm": 1.0725680589675903,
"learning_rate": 6.0520374247438375e-05,
"loss": 0.8612,
"step": 143
},
{
"epoch": 0.18759770713913496,
"grad_norm": 1.0441476106643677,
"learning_rate": 6.045125110995204e-05,
"loss": 0.8521,
"step": 144
},
{
"epoch": 0.18890046899426785,
"grad_norm": 1.0792895555496216,
"learning_rate": 6.0381638846757264e-05,
"loss": 0.8364,
"step": 145
},
{
"epoch": 0.19020323084940072,
"grad_norm": 1.0203429460525513,
"learning_rate": 6.03115386760159e-05,
"loss": 0.7973,
"step": 146
},
{
"epoch": 0.1915059927045336,
"grad_norm": 1.0836308002471924,
"learning_rate": 6.024095182442782e-05,
"loss": 0.8719,
"step": 147
},
{
"epoch": 0.1928087545596665,
"grad_norm": 1.0394123792648315,
"learning_rate": 6.016987952720939e-05,
"loss": 0.8809,
"step": 148
},
{
"epoch": 0.19411151641479937,
"grad_norm": 1.0890783071517944,
"learning_rate": 6.0098323028071956e-05,
"loss": 0.8996,
"step": 149
},
{
"epoch": 0.19541427826993227,
"grad_norm": 1.0037873983383179,
"learning_rate": 6.002628357920002e-05,
"loss": 0.8424,
"step": 150
},
{
"epoch": 0.19671704012506513,
"grad_norm": 1.0470969676971436,
"learning_rate": 5.995376244122935e-05,
"loss": 0.8019,
"step": 151
},
{
"epoch": 0.19801980198019803,
"grad_norm": 1.0230716466903687,
"learning_rate": 5.988076088322486e-05,
"loss": 0.7927,
"step": 152
},
{
"epoch": 0.1993225638353309,
"grad_norm": 1.0318727493286133,
"learning_rate": 5.980728018265849e-05,
"loss": 0.803,
"step": 153
},
{
"epoch": 0.2006253256904638,
"grad_norm": 1.0479062795639038,
"learning_rate": 5.9733321625386824e-05,
"loss": 0.8597,
"step": 154
},
{
"epoch": 0.20192808754559666,
"grad_norm": 1.025315761566162,
"learning_rate": 5.965888650562853e-05,
"loss": 0.7937,
"step": 155
},
{
"epoch": 0.20323084940072955,
"grad_norm": 1.0599974393844604,
"learning_rate": 5.958397612594182e-05,
"loss": 0.8628,
"step": 156
},
{
"epoch": 0.20453361125586242,
"grad_norm": 1.0344469547271729,
"learning_rate": 5.950859179720152e-05,
"loss": 0.8436,
"step": 157
},
{
"epoch": 0.2058363731109953,
"grad_norm": 1.1034519672393799,
"learning_rate": 5.943273483857627e-05,
"loss": 0.8124,
"step": 158
},
{
"epoch": 0.20713913496612818,
"grad_norm": 1.0487233400344849,
"learning_rate": 5.935640657750534e-05,
"loss": 0.8337,
"step": 159
},
{
"epoch": 0.20844189682126107,
"grad_norm": 1.0436427593231201,
"learning_rate": 5.9279608349675444e-05,
"loss": 0.8438,
"step": 160
},
{
"epoch": 0.20974465867639394,
"grad_norm": 1.0612242221832275,
"learning_rate": 5.920234149899736e-05,
"loss": 0.8735,
"step": 161
},
{
"epoch": 0.21104742053152684,
"grad_norm": 1.168532371520996,
"learning_rate": 5.912460737758241e-05,
"loss": 0.8422,
"step": 162
},
{
"epoch": 0.21235018238665973,
"grad_norm": 1.0675426721572876,
"learning_rate": 5.9046407345718805e-05,
"loss": 0.8631,
"step": 163
},
{
"epoch": 0.2136529442417926,
"grad_norm": 1.048614501953125,
"learning_rate": 5.8967742771847826e-05,
"loss": 0.8473,
"step": 164
},
{
"epoch": 0.2149557060969255,
"grad_norm": 1.0362776517868042,
"learning_rate": 5.888861503253989e-05,
"loss": 0.8888,
"step": 165
},
{
"epoch": 0.21625846795205836,
"grad_norm": 0.9976462721824646,
"learning_rate": 5.8809025512470445e-05,
"loss": 0.8079,
"step": 166
},
{
"epoch": 0.21756122980719125,
"grad_norm": 1.0061395168304443,
"learning_rate": 5.87289756043958e-05,
"loss": 0.7954,
"step": 167
},
{
"epoch": 0.21886399166232412,
"grad_norm": 1.0533777475357056,
"learning_rate": 5.864846670912864e-05,
"loss": 0.8089,
"step": 168
},
{
"epoch": 0.22016675351745701,
"grad_norm": 0.9939700365066528,
"learning_rate": 5.856750023551361e-05,
"loss": 0.8176,
"step": 169
},
{
"epoch": 0.22146951537258988,
"grad_norm": 1.0513004064559937,
"learning_rate": 5.848607760040267e-05,
"loss": 0.8602,
"step": 170
},
{
"epoch": 0.22277227722772278,
"grad_norm": 1.055511713027954,
"learning_rate": 5.840420022863016e-05,
"loss": 0.8511,
"step": 171
},
{
"epoch": 0.22407503908285564,
"grad_norm": 1.0215332508087158,
"learning_rate": 5.832186955298808e-05,
"loss": 0.8233,
"step": 172
},
{
"epoch": 0.22537780093798854,
"grad_norm": 1.018233299255371,
"learning_rate": 5.8239087014200806e-05,
"loss": 0.8354,
"step": 173
},
{
"epoch": 0.2266805627931214,
"grad_norm": 1.0055863857269287,
"learning_rate": 5.815585406090005e-05,
"loss": 0.8488,
"step": 174
},
{
"epoch": 0.2279833246482543,
"grad_norm": 1.008152961730957,
"learning_rate": 5.8072172149599394e-05,
"loss": 0.7623,
"step": 175
},
{
"epoch": 0.2292860865033872,
"grad_norm": 1.0418919324874878,
"learning_rate": 5.7988042744668856e-05,
"loss": 0.8109,
"step": 176
},
{
"epoch": 0.23058884835852006,
"grad_norm": 1.0832551717758179,
"learning_rate": 5.790346731830927e-05,
"loss": 0.8951,
"step": 177
},
{
"epoch": 0.23189161021365295,
"grad_norm": 1.0513347387313843,
"learning_rate": 5.7818447350526476e-05,
"loss": 0.8598,
"step": 178
},
{
"epoch": 0.23319437206878582,
"grad_norm": 0.9888846278190613,
"learning_rate": 5.7732984329105484e-05,
"loss": 0.746,
"step": 179
},
{
"epoch": 0.23449713392391872,
"grad_norm": 1.0668085813522339,
"learning_rate": 5.764707974958437e-05,
"loss": 0.8039,
"step": 180
},
{
"epoch": 0.23579989577905158,
"grad_norm": 1.0392318964004517,
"learning_rate": 5.7560735115228195e-05,
"loss": 0.8155,
"step": 181
},
{
"epoch": 0.23710265763418448,
"grad_norm": 1.052474021911621,
"learning_rate": 5.7473951937002576e-05,
"loss": 0.8215,
"step": 182
},
{
"epoch": 0.23840541948931734,
"grad_norm": 1.067699670791626,
"learning_rate": 5.738673173354739e-05,
"loss": 0.8259,
"step": 183
},
{
"epoch": 0.23970818134445024,
"grad_norm": 1.042398452758789,
"learning_rate": 5.729907603115005e-05,
"loss": 0.8086,
"step": 184
},
{
"epoch": 0.2410109431995831,
"grad_norm": 1.0346574783325195,
"learning_rate": 5.721098636371894e-05,
"loss": 0.8184,
"step": 185
},
{
"epoch": 0.242313705054716,
"grad_norm": 1.0383342504501343,
"learning_rate": 5.712246427275646e-05,
"loss": 0.8543,
"step": 186
},
{
"epoch": 0.24361646690984887,
"grad_norm": 1.0456079244613647,
"learning_rate": 5.703351130733211e-05,
"loss": 0.8244,
"step": 187
},
{
"epoch": 0.24491922876498176,
"grad_norm": 0.9899833798408508,
"learning_rate": 5.694412902405535e-05,
"loss": 0.7916,
"step": 188
},
{
"epoch": 0.24622199062011466,
"grad_norm": 1.1231224536895752,
"learning_rate": 5.685431898704839e-05,
"loss": 0.7904,
"step": 189
},
{
"epoch": 0.24752475247524752,
"grad_norm": 1.0400362014770508,
"learning_rate": 5.676408276791882e-05,
"loss": 0.845,
"step": 190
},
{
"epoch": 0.24882751433038042,
"grad_norm": 1.0910183191299438,
"learning_rate": 5.6673421945732075e-05,
"loss": 0.8293,
"step": 191
},
{
"epoch": 0.2501302761855133,
"grad_norm": 1.0755969285964966,
"learning_rate": 5.658233810698381e-05,
"loss": 0.83,
"step": 192
},
{
"epoch": 0.25143303804064615,
"grad_norm": 1.0231711864471436,
"learning_rate": 5.649083284557219e-05,
"loss": 0.8442,
"step": 193
},
{
"epoch": 0.25273579989577905,
"grad_norm": 1.0368572473526,
"learning_rate": 5.639890776276991e-05,
"loss": 0.8574,
"step": 194
},
{
"epoch": 0.25403856175091194,
"grad_norm": 1.0136909484863281,
"learning_rate": 5.630656446719626e-05,
"loss": 0.8151,
"step": 195
},
{
"epoch": 0.25534132360604483,
"grad_norm": 1.0600650310516357,
"learning_rate": 5.621380457478889e-05,
"loss": 0.7792,
"step": 196
},
{
"epoch": 0.2566440854611777,
"grad_norm": 1.0362857580184937,
"learning_rate": 5.612062970877565e-05,
"loss": 0.8246,
"step": 197
},
{
"epoch": 0.25794684731631057,
"grad_norm": 1.0172935724258423,
"learning_rate": 5.602704149964604e-05,
"loss": 0.8196,
"step": 198
},
{
"epoch": 0.25924960917144346,
"grad_norm": 1.0049176216125488,
"learning_rate": 5.5933041585122805e-05,
"loss": 0.8229,
"step": 199
},
{
"epoch": 0.26055237102657636,
"grad_norm": 1.0751328468322754,
"learning_rate": 5.5838631610133196e-05,
"loss": 0.8456,
"step": 200
},
{
"epoch": 0.2618551328817092,
"grad_norm": 1.0684475898742676,
"learning_rate": 5.57438132267802e-05,
"loss": 0.8607,
"step": 201
},
{
"epoch": 0.2631578947368421,
"grad_norm": 1.029704213142395,
"learning_rate": 5.564858809431367e-05,
"loss": 0.7801,
"step": 202
},
{
"epoch": 0.264460656591975,
"grad_norm": 1.044959306716919,
"learning_rate": 5.5552957879101246e-05,
"loss": 0.8379,
"step": 203
},
{
"epoch": 0.2657634184471079,
"grad_norm": 1.030415654182434,
"learning_rate": 5.545692425459921e-05,
"loss": 0.8705,
"step": 204
},
{
"epoch": 0.2670661803022408,
"grad_norm": 1.0018935203552246,
"learning_rate": 5.5360488901323206e-05,
"loss": 0.8122,
"step": 205
},
{
"epoch": 0.2683689421573736,
"grad_norm": 1.0166568756103516,
"learning_rate": 5.526365350681881e-05,
"loss": 0.8171,
"step": 206
},
{
"epoch": 0.2696717040125065,
"grad_norm": 1.0195856094360352,
"learning_rate": 5.516641976563204e-05,
"loss": 0.8496,
"step": 207
},
{
"epoch": 0.2709744658676394,
"grad_norm": 0.985876739025116,
"learning_rate": 5.506878937927966e-05,
"loss": 0.801,
"step": 208
},
{
"epoch": 0.2722772277227723,
"grad_norm": 1.0150160789489746,
"learning_rate": 5.497076405621944e-05,
"loss": 0.784,
"step": 209
},
{
"epoch": 0.27357998957790514,
"grad_norm": 1.095823884010315,
"learning_rate": 5.487234551182022e-05,
"loss": 0.8957,
"step": 210
},
{
"epoch": 0.27488275143303803,
"grad_norm": 1.033151388168335,
"learning_rate": 5.477353546833193e-05,
"loss": 0.8346,
"step": 211
},
{
"epoch": 0.2761855132881709,
"grad_norm": 0.982230007648468,
"learning_rate": 5.467433565485546e-05,
"loss": 0.7826,
"step": 212
},
{
"epoch": 0.2774882751433038,
"grad_norm": 0.9829743504524231,
"learning_rate": 5.4574747807312326e-05,
"loss": 0.791,
"step": 213
},
{
"epoch": 0.27879103699843666,
"grad_norm": 1.0257138013839722,
"learning_rate": 5.4474773668414395e-05,
"loss": 0.8498,
"step": 214
},
{
"epoch": 0.28009379885356955,
"grad_norm": 1.0054680109024048,
"learning_rate": 5.437441498763332e-05,
"loss": 0.8365,
"step": 215
},
{
"epoch": 0.28139656070870245,
"grad_norm": 1.0021213293075562,
"learning_rate": 5.427367352116993e-05,
"loss": 0.8313,
"step": 216
},
{
"epoch": 0.28269932256383534,
"grad_norm": 1.0124083757400513,
"learning_rate": 5.417255103192355e-05,
"loss": 0.81,
"step": 217
},
{
"epoch": 0.28400208441896824,
"grad_norm": 1.0104458332061768,
"learning_rate": 5.407104928946109e-05,
"loss": 0.7821,
"step": 218
},
{
"epoch": 0.2853048462741011,
"grad_norm": 1.0982369184494019,
"learning_rate": 5.39691700699861e-05,
"loss": 0.8512,
"step": 219
},
{
"epoch": 0.28660760812923397,
"grad_norm": 1.0122724771499634,
"learning_rate": 5.386691515630771e-05,
"loss": 0.8634,
"step": 220
},
{
"epoch": 0.28791036998436687,
"grad_norm": 1.0455682277679443,
"learning_rate": 5.376428633780941e-05,
"loss": 0.8399,
"step": 221
},
{
"epoch": 0.28921313183949976,
"grad_norm": 0.9880922436714172,
"learning_rate": 5.3661285410417705e-05,
"loss": 0.8044,
"step": 222
},
{
"epoch": 0.2905158936946326,
"grad_norm": 1.0006861686706543,
"learning_rate": 5.355791417657078e-05,
"loss": 0.7957,
"step": 223
},
{
"epoch": 0.2918186555497655,
"grad_norm": 1.0367894172668457,
"learning_rate": 5.345417444518687e-05,
"loss": 0.8207,
"step": 224
},
{
"epoch": 0.2931214174048984,
"grad_norm": 1.0520100593566895,
"learning_rate": 5.335006803163265e-05,
"loss": 0.8586,
"step": 225
},
{
"epoch": 0.2944241792600313,
"grad_norm": 1.0345007181167603,
"learning_rate": 5.3245596757691435e-05,
"loss": 0.8362,
"step": 226
},
{
"epoch": 0.2957269411151641,
"grad_norm": 1.0230737924575806,
"learning_rate": 5.3140762451531344e-05,
"loss": 0.8482,
"step": 227
},
{
"epoch": 0.297029702970297,
"grad_norm": 0.9852018356323242,
"learning_rate": 5.303556694767326e-05,
"loss": 0.8661,
"step": 228
},
{
"epoch": 0.2983324648254299,
"grad_norm": 0.9801431894302368,
"learning_rate": 5.2930012086958783e-05,
"loss": 0.8702,
"step": 229
},
{
"epoch": 0.2996352266805628,
"grad_norm": 1.0442442893981934,
"learning_rate": 5.282409971651797e-05,
"loss": 0.8566,
"step": 230
},
{
"epoch": 0.3009379885356957,
"grad_norm": 1.103188157081604,
"learning_rate": 5.271783168973701e-05,
"loss": 0.8206,
"step": 231
},
{
"epoch": 0.30224075039082854,
"grad_norm": 1.0066590309143066,
"learning_rate": 5.261120986622583e-05,
"loss": 0.8369,
"step": 232
},
{
"epoch": 0.30354351224596143,
"grad_norm": 0.9748640060424805,
"learning_rate": 5.250423611178553e-05,
"loss": 0.7792,
"step": 233
},
{
"epoch": 0.30484627410109433,
"grad_norm": 1.0654217004776,
"learning_rate": 5.239691229837573e-05,
"loss": 0.8242,
"step": 234
},
{
"epoch": 0.3061490359562272,
"grad_norm": 1.0146901607513428,
"learning_rate": 5.228924030408182e-05,
"loss": 0.8166,
"step": 235
},
{
"epoch": 0.30745179781136006,
"grad_norm": 1.0119966268539429,
"learning_rate": 5.218122201308209e-05,
"loss": 0.8074,
"step": 236
},
{
"epoch": 0.30875455966649296,
"grad_norm": 1.0073226690292358,
"learning_rate": 5.207285931561475e-05,
"loss": 0.8292,
"step": 237
},
{
"epoch": 0.31005732152162585,
"grad_norm": 1.0207470655441284,
"learning_rate": 5.1964154107944876e-05,
"loss": 0.8245,
"step": 238
},
{
"epoch": 0.31136008337675875,
"grad_norm": 0.9803332090377808,
"learning_rate": 5.185510829233121e-05,
"loss": 0.774,
"step": 239
},
{
"epoch": 0.3126628452318916,
"grad_norm": 1.0056718587875366,
"learning_rate": 5.174572377699286e-05,
"loss": 0.8308,
"step": 240
},
{
"epoch": 0.3139656070870245,
"grad_norm": 0.9751390218734741,
"learning_rate": 5.163600247607596e-05,
"loss": 0.7648,
"step": 241
},
{
"epoch": 0.3152683689421574,
"grad_norm": 1.0351225137710571,
"learning_rate": 5.152594630962008e-05,
"loss": 0.7695,
"step": 242
},
{
"epoch": 0.31657113079729027,
"grad_norm": 1.0021389722824097,
"learning_rate": 5.1415557203524736e-05,
"loss": 0.8158,
"step": 243
},
{
"epoch": 0.31787389265242316,
"grad_norm": 1.0016628503799438,
"learning_rate": 5.130483708951561e-05,
"loss": 0.793,
"step": 244
},
{
"epoch": 0.319176654507556,
"grad_norm": 1.008516788482666,
"learning_rate": 5.119378790511079e-05,
"loss": 0.8143,
"step": 245
},
{
"epoch": 0.3204794163626889,
"grad_norm": 1.0514389276504517,
"learning_rate": 5.108241159358679e-05,
"loss": 0.7945,
"step": 246
},
{
"epoch": 0.3217821782178218,
"grad_norm": 0.9676159024238586,
"learning_rate": 5.097071010394469e-05,
"loss": 0.7912,
"step": 247
},
{
"epoch": 0.3230849400729547,
"grad_norm": 1.0046639442443848,
"learning_rate": 5.085868539087586e-05,
"loss": 0.8136,
"step": 248
},
{
"epoch": 0.3243877019280875,
"grad_norm": 0.993964433670044,
"learning_rate": 5.074633941472788e-05,
"loss": 0.7893,
"step": 249
},
{
"epoch": 0.3256904637832204,
"grad_norm": 0.9951984286308289,
"learning_rate": 5.0633674141470194e-05,
"loss": 0.7463,
"step": 250
},
{
"epoch": 0.3269932256383533,
"grad_norm": 0.9659764766693115,
"learning_rate": 5.0520691542659655e-05,
"loss": 0.8079,
"step": 251
},
{
"epoch": 0.3282959874934862,
"grad_norm": 1.006102442741394,
"learning_rate": 5.0407393595406125e-05,
"loss": 0.7864,
"step": 252
},
{
"epoch": 0.32959874934861905,
"grad_norm": 0.9752157330513,
"learning_rate": 5.029378228233782e-05,
"loss": 0.7683,
"step": 253
},
{
"epoch": 0.33090151120375194,
"grad_norm": 1.0095077753067017,
"learning_rate": 5.0179859591566575e-05,
"loss": 0.8504,
"step": 254
},
{
"epoch": 0.33220427305888484,
"grad_norm": 1.0161761045455933,
"learning_rate": 5.0065627516653145e-05,
"loss": 0.8308,
"step": 255
},
{
"epoch": 0.33350703491401773,
"grad_norm": 0.9710779190063477,
"learning_rate": 4.995108805657226e-05,
"loss": 0.7761,
"step": 256
},
{
"epoch": 0.3348097967691506,
"grad_norm": 1.0192509889602661,
"learning_rate": 4.983624321567764e-05,
"loss": 0.7955,
"step": 257
},
{
"epoch": 0.33611255862428346,
"grad_norm": 1.0141615867614746,
"learning_rate": 4.972109500366698e-05,
"loss": 0.8243,
"step": 258
},
{
"epoch": 0.33741532047941636,
"grad_norm": 1.0046883821487427,
"learning_rate": 4.9605645435546694e-05,
"loss": 0.8272,
"step": 259
},
{
"epoch": 0.33871808233454925,
"grad_norm": 1.031710147857666,
"learning_rate": 4.948989653159671e-05,
"loss": 0.8187,
"step": 260
},
{
"epoch": 0.34002084418968215,
"grad_norm": 0.9797648787498474,
"learning_rate": 4.937385031733513e-05,
"loss": 0.8062,
"step": 261
},
{
"epoch": 0.341323606044815,
"grad_norm": 0.9559136629104614,
"learning_rate": 4.925750882348274e-05,
"loss": 0.7627,
"step": 262
},
{
"epoch": 0.3426263678999479,
"grad_norm": 1.0041617155075073,
"learning_rate": 4.9140874085927486e-05,
"loss": 0.821,
"step": 263
},
{
"epoch": 0.3439291297550808,
"grad_norm": 1.0307433605194092,
"learning_rate": 4.9023948145688896e-05,
"loss": 0.7842,
"step": 264
},
{
"epoch": 0.34523189161021367,
"grad_norm": 0.9956508874893188,
"learning_rate": 4.8906733048882286e-05,
"loss": 0.8067,
"step": 265
},
{
"epoch": 0.3465346534653465,
"grad_norm": 1.0119211673736572,
"learning_rate": 4.878923084668299e-05,
"loss": 0.8037,
"step": 266
},
{
"epoch": 0.3478374153204794,
"grad_norm": 1.353070855140686,
"learning_rate": 4.867144359529053e-05,
"loss": 0.8174,
"step": 267
},
{
"epoch": 0.3491401771756123,
"grad_norm": 1.0314981937408447,
"learning_rate": 4.855337335589247e-05,
"loss": 0.8061,
"step": 268
},
{
"epoch": 0.3504429390307452,
"grad_norm": 1.0623273849487305,
"learning_rate": 4.8435022194628533e-05,
"loss": 0.8216,
"step": 269
},
{
"epoch": 0.3517457008858781,
"grad_norm": 0.9801738858222961,
"learning_rate": 4.831639218255434e-05,
"loss": 0.8048,
"step": 270
},
{
"epoch": 0.3530484627410109,
"grad_norm": 0.9815056324005127,
"learning_rate": 4.8197485395605186e-05,
"loss": 0.8232,
"step": 271
},
{
"epoch": 0.3543512245961438,
"grad_norm": 1.0663788318634033,
"learning_rate": 4.8078303914559706e-05,
"loss": 0.8185,
"step": 272
},
{
"epoch": 0.3556539864512767,
"grad_norm": 0.9978227019309998,
"learning_rate": 4.795884982500349e-05,
"loss": 0.777,
"step": 273
},
{
"epoch": 0.3569567483064096,
"grad_norm": 1.0210098028182983,
"learning_rate": 4.7839125217292556e-05,
"loss": 0.8353,
"step": 274
},
{
"epoch": 0.35825951016154245,
"grad_norm": 0.9370939135551453,
"learning_rate": 4.77191321865168e-05,
"loss": 0.7465,
"step": 275
},
{
"epoch": 0.35956227201667534,
"grad_norm": 1.0085511207580566,
"learning_rate": 4.759887283246331e-05,
"loss": 0.8009,
"step": 276
},
{
"epoch": 0.36086503387180824,
"grad_norm": 0.9499126672744751,
"learning_rate": 4.747834925957963e-05,
"loss": 0.8041,
"step": 277
},
{
"epoch": 0.36216779572694113,
"grad_norm": 0.9528019428253174,
"learning_rate": 4.735756357693692e-05,
"loss": 0.7454,
"step": 278
},
{
"epoch": 0.36347055758207397,
"grad_norm": 1.011826515197754,
"learning_rate": 4.7236517898193094e-05,
"loss": 0.7772,
"step": 279
},
{
"epoch": 0.36477331943720687,
"grad_norm": 0.9649999141693115,
"learning_rate": 4.711521434155578e-05,
"loss": 0.8043,
"step": 280
},
{
"epoch": 0.36607608129233976,
"grad_norm": 0.9927408695220947,
"learning_rate": 4.6993655029745263e-05,
"loss": 0.8378,
"step": 281
},
{
"epoch": 0.36737884314747266,
"grad_norm": 1.0498725175857544,
"learning_rate": 4.6871842089957375e-05,
"loss": 0.8455,
"step": 282
},
{
"epoch": 0.36868160500260555,
"grad_norm": 0.9903897643089294,
"learning_rate": 4.674977765382622e-05,
"loss": 0.7628,
"step": 283
},
{
"epoch": 0.3699843668577384,
"grad_norm": 1.0428470373153687,
"learning_rate": 4.662746385738691e-05,
"loss": 0.8221,
"step": 284
},
{
"epoch": 0.3712871287128713,
"grad_norm": 0.9479437470436096,
"learning_rate": 4.6504902841038185e-05,
"loss": 0.7747,
"step": 285
},
{
"epoch": 0.3725898905680042,
"grad_norm": 0.9896458983421326,
"learning_rate": 4.6382096749504944e-05,
"loss": 0.7987,
"step": 286
},
{
"epoch": 0.3738926524231371,
"grad_norm": 0.9583244919776917,
"learning_rate": 4.6259047731800675e-05,
"loss": 0.7863,
"step": 287
},
{
"epoch": 0.3751954142782699,
"grad_norm": 0.9551007747650146,
"learning_rate": 4.613575794118994e-05,
"loss": 0.7665,
"step": 288
},
{
"epoch": 0.3764981761334028,
"grad_norm": 1.0227469205856323,
"learning_rate": 4.601222953515063e-05,
"loss": 0.8208,
"step": 289
},
{
"epoch": 0.3778009379885357,
"grad_norm": 1.0073151588439941,
"learning_rate": 4.588846467533622e-05,
"loss": 0.8182,
"step": 290
},
{
"epoch": 0.3791036998436686,
"grad_norm": 0.9630952477455139,
"learning_rate": 4.576446552753793e-05,
"loss": 0.7829,
"step": 291
},
{
"epoch": 0.38040646169880143,
"grad_norm": 0.9635125398635864,
"learning_rate": 4.564023426164688e-05,
"loss": 0.8203,
"step": 292
},
{
"epoch": 0.38170922355393433,
"grad_norm": 0.9933289885520935,
"learning_rate": 4.551577305161605e-05,
"loss": 0.7674,
"step": 293
},
{
"epoch": 0.3830119854090672,
"grad_norm": 1.050747036933899,
"learning_rate": 4.539108407542228e-05,
"loss": 0.8082,
"step": 294
},
{
"epoch": 0.3843147472642001,
"grad_norm": 1.0443624258041382,
"learning_rate": 4.526616951502815e-05,
"loss": 0.8243,
"step": 295
},
{
"epoch": 0.385617509119333,
"grad_norm": 1.0552526712417603,
"learning_rate": 4.5141031556343774e-05,
"loss": 0.7807,
"step": 296
},
{
"epoch": 0.38692027097446585,
"grad_norm": 1.0113283395767212,
"learning_rate": 4.501567238918859e-05,
"loss": 0.8494,
"step": 297
},
{
"epoch": 0.38822303282959875,
"grad_norm": 1.0051511526107788,
"learning_rate": 4.4890094207253014e-05,
"loss": 0.7844,
"step": 298
},
{
"epoch": 0.38952579468473164,
"grad_norm": 1.0065789222717285,
"learning_rate": 4.476429920806002e-05,
"loss": 0.7767,
"step": 299
},
{
"epoch": 0.39082855653986454,
"grad_norm": 1.0727571249008179,
"learning_rate": 4.463828959292677e-05,
"loss": 0.8314,
"step": 300
}
],
"logging_steps": 1,
"max_steps": 767,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4.6131456216150835e+17,
"train_batch_size": 6,
"trial_name": null,
"trial_params": null
}