etbb12b / trainer_state.json
fnruha0921's picture
Upload folder using huggingface_hub
9fdfd38 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 202,
"global_step": 202,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.009900990099009901,
"grad_norm": 152.50006103515625,
"learning_rate": 0.0,
"loss": 15.6875,
"step": 1
},
{
"epoch": 0.019801980198019802,
"grad_norm": 155.11602783203125,
"learning_rate": 1.8181818181818182e-05,
"loss": 15.8984,
"step": 2
},
{
"epoch": 0.0297029702970297,
"grad_norm": 39.620399475097656,
"learning_rate": 3.6363636363636364e-05,
"loss": 11.7891,
"step": 3
},
{
"epoch": 0.039603960396039604,
"grad_norm": 481.17474365234375,
"learning_rate": 5.4545454545454546e-05,
"loss": 12.0391,
"step": 4
},
{
"epoch": 0.04950495049504951,
"grad_norm": 45.559226989746094,
"learning_rate": 7.272727272727273e-05,
"loss": 13.8828,
"step": 5
},
{
"epoch": 0.0594059405940594,
"grad_norm": 65.59961700439453,
"learning_rate": 9.090909090909092e-05,
"loss": 16.6641,
"step": 6
},
{
"epoch": 0.06930693069306931,
"grad_norm": 33.685943603515625,
"learning_rate": 0.00010909090909090909,
"loss": 15.5469,
"step": 7
},
{
"epoch": 0.07920792079207921,
"grad_norm": 73.23568725585938,
"learning_rate": 0.00012727272727272728,
"loss": 19.2344,
"step": 8
},
{
"epoch": 0.0891089108910891,
"grad_norm": 26.491865158081055,
"learning_rate": 0.00014545454545454546,
"loss": 14.6875,
"step": 9
},
{
"epoch": 0.09900990099009901,
"grad_norm": 29.270771026611328,
"learning_rate": 0.00016363636363636366,
"loss": 13.5625,
"step": 10
},
{
"epoch": 0.10891089108910891,
"grad_norm": 22.94576644897461,
"learning_rate": 0.00018181818181818183,
"loss": 13.8359,
"step": 11
},
{
"epoch": 0.1188118811881188,
"grad_norm": 19.714340209960938,
"learning_rate": 0.0002,
"loss": 13.4453,
"step": 12
},
{
"epoch": 0.12871287128712872,
"grad_norm": 29.34218406677246,
"learning_rate": 0.00019998782593171394,
"loss": 14.1328,
"step": 13
},
{
"epoch": 0.13861386138613863,
"grad_norm": 15.325265884399414,
"learning_rate": 0.0001999513070203655,
"loss": 14.1484,
"step": 14
},
{
"epoch": 0.1485148514851485,
"grad_norm": 15.70521354675293,
"learning_rate": 0.00019989045314559295,
"loss": 13.1875,
"step": 15
},
{
"epoch": 0.15841584158415842,
"grad_norm": 13.77431869506836,
"learning_rate": 0.00019980528077049017,
"loss": 12.2109,
"step": 16
},
{
"epoch": 0.16831683168316833,
"grad_norm": 48.380741119384766,
"learning_rate": 0.00019969581293715296,
"loss": 12.6484,
"step": 17
},
{
"epoch": 0.1782178217821782,
"grad_norm": 13.91443157196045,
"learning_rate": 0.00019956207926044532,
"loss": 12.6172,
"step": 18
},
{
"epoch": 0.18811881188118812,
"grad_norm": 9.085782051086426,
"learning_rate": 0.00019940411591998748,
"loss": 11.7734,
"step": 19
},
{
"epoch": 0.19801980198019803,
"grad_norm": 8.792189598083496,
"learning_rate": 0.0001992219656503682,
"loss": 11.2812,
"step": 20
},
{
"epoch": 0.2079207920792079,
"grad_norm": 9.451997756958008,
"learning_rate": 0.0001990156777295835,
"loss": 10.9844,
"step": 21
},
{
"epoch": 0.21782178217821782,
"grad_norm": 8.6239652633667,
"learning_rate": 0.00019878530796570528,
"loss": 10.3516,
"step": 22
},
{
"epoch": 0.22772277227722773,
"grad_norm": 9.932726860046387,
"learning_rate": 0.00019853091868178316,
"loss": 10.2031,
"step": 23
},
{
"epoch": 0.2376237623762376,
"grad_norm": 5.901266098022461,
"learning_rate": 0.00019825257869898418,
"loss": 9.9297,
"step": 24
},
{
"epoch": 0.24752475247524752,
"grad_norm": 4.901180744171143,
"learning_rate": 0.00019795036331797405,
"loss": 9.5938,
"step": 25
},
{
"epoch": 0.25742574257425743,
"grad_norm": 4.704899311065674,
"learning_rate": 0.00019762435429854577,
"loss": 9.5703,
"step": 26
},
{
"epoch": 0.26732673267326734,
"grad_norm": 4.755985736846924,
"learning_rate": 0.00019727463983750077,
"loss": 9.5312,
"step": 27
},
{
"epoch": 0.27722772277227725,
"grad_norm": 5.313395023345947,
"learning_rate": 0.00019690131454478858,
"loss": 9.3672,
"step": 28
},
{
"epoch": 0.2871287128712871,
"grad_norm": 6.608179092407227,
"learning_rate": 0.00019650447941791155,
"loss": 9.6875,
"step": 29
},
{
"epoch": 0.297029702970297,
"grad_norm": 7.129603385925293,
"learning_rate": 0.0001960842418146016,
"loss": 9.6016,
"step": 30
},
{
"epoch": 0.3069306930693069,
"grad_norm": 1075.1676025390625,
"learning_rate": 0.00019564071542377596,
"loss": 18.0156,
"step": 31
},
{
"epoch": 0.31683168316831684,
"grad_norm": 12.196839332580566,
"learning_rate": 0.00019517402023478062,
"loss": 10.8125,
"step": 32
},
{
"epoch": 0.32673267326732675,
"grad_norm": 6.137757301330566,
"learning_rate": 0.0001946842825049289,
"loss": 10.0625,
"step": 33
},
{
"epoch": 0.33663366336633666,
"grad_norm": 9.860557556152344,
"learning_rate": 0.00019417163472534456,
"loss": 10.1719,
"step": 34
},
{
"epoch": 0.3465346534653465,
"grad_norm": 9.4225435256958,
"learning_rate": 0.00019363621558511805,
"loss": 8.7197,
"step": 35
},
{
"epoch": 0.3564356435643564,
"grad_norm": 4.244920253753662,
"learning_rate": 0.00019307816993378662,
"loss": 9.5391,
"step": 36
},
{
"epoch": 0.36633663366336633,
"grad_norm": 4.227206707000732,
"learning_rate": 0.00019249764874214732,
"loss": 9.3281,
"step": 37
},
{
"epoch": 0.37623762376237624,
"grad_norm": 3.867461681365967,
"learning_rate": 0.00019189480906141413,
"loss": 9.0625,
"step": 38
},
{
"epoch": 0.38613861386138615,
"grad_norm": 5.17470121383667,
"learning_rate": 0.00019126981398073008,
"loss": 9.1094,
"step": 39
},
{
"epoch": 0.39603960396039606,
"grad_norm": 5.798826694488525,
"learning_rate": 0.00019062283258304612,
"loss": 9.3359,
"step": 40
},
{
"epoch": 0.40594059405940597,
"grad_norm": 4.38295841217041,
"learning_rate": 0.0001899540398993781,
"loss": 9.4219,
"step": 41
},
{
"epoch": 0.4158415841584158,
"grad_norm": 4.414400577545166,
"learning_rate": 0.00018926361686145494,
"loss": 9.1016,
"step": 42
},
{
"epoch": 0.42574257425742573,
"grad_norm": 4.6771559715271,
"learning_rate": 0.00018855175025277022,
"loss": 8.9219,
"step": 43
},
{
"epoch": 0.43564356435643564,
"grad_norm": 4.436445713043213,
"learning_rate": 0.00018781863265805065,
"loss": 9.1562,
"step": 44
},
{
"epoch": 0.44554455445544555,
"grad_norm": 3.0445127487182617,
"learning_rate": 0.00018706446241115537,
"loss": 9.0,
"step": 45
},
{
"epoch": 0.45544554455445546,
"grad_norm": 13.921093940734863,
"learning_rate": 0.00018628944354141962,
"loss": 8.6484,
"step": 46
},
{
"epoch": 0.46534653465346537,
"grad_norm": 7.534108638763428,
"learning_rate": 0.00018549378571845767,
"loss": 9.7734,
"step": 47
},
{
"epoch": 0.4752475247524752,
"grad_norm": 3.2327933311462402,
"learning_rate": 0.00018467770419543998,
"loss": 9.375,
"step": 48
},
{
"epoch": 0.48514851485148514,
"grad_norm": 4.523435115814209,
"learning_rate": 0.00018384141975085952,
"loss": 9.1406,
"step": 49
},
{
"epoch": 0.49504950495049505,
"grad_norm": 83.07750701904297,
"learning_rate": 0.00018298515862880387,
"loss": 8.9531,
"step": 50
},
{
"epoch": 0.504950495049505,
"grad_norm": 6.1501641273498535,
"learning_rate": 0.00018210915247774784,
"loss": 10.0547,
"step": 51
},
{
"epoch": 0.5148514851485149,
"grad_norm": 3.1539700031280518,
"learning_rate": 0.00018121363828788484,
"loss": 9.3359,
"step": 52
},
{
"epoch": 0.5247524752475248,
"grad_norm": 3.68802809715271,
"learning_rate": 0.0001802988583270126,
"loss": 9.1328,
"step": 53
},
{
"epoch": 0.5346534653465347,
"grad_norm": 3.102386713027954,
"learning_rate": 0.00017936506007499136,
"loss": 8.9609,
"step": 54
},
{
"epoch": 0.5445544554455446,
"grad_norm": 3.1378631591796875,
"learning_rate": 0.00017841249615679192,
"loss": 8.7891,
"step": 55
},
{
"epoch": 0.5544554455445545,
"grad_norm": 3.207996129989624,
"learning_rate": 0.00017744142427415172,
"loss": 8.75,
"step": 56
},
{
"epoch": 0.5643564356435643,
"grad_norm": 4.185523509979248,
"learning_rate": 0.0001764521071358577,
"loss": 8.5781,
"step": 57
},
{
"epoch": 0.5742574257425742,
"grad_norm": 2.66141676902771,
"learning_rate": 0.00017544481238667426,
"loss": 8.6172,
"step": 58
},
{
"epoch": 0.5841584158415841,
"grad_norm": 4.6184258460998535,
"learning_rate": 0.00017441981253493607,
"loss": 8.625,
"step": 59
},
{
"epoch": 0.594059405940594,
"grad_norm": 114.2392578125,
"learning_rate": 0.00017337738487882508,
"loss": 13.5,
"step": 60
},
{
"epoch": 0.6039603960396039,
"grad_norm": 6.742166996002197,
"learning_rate": 0.00017231781143135173,
"loss": 9.3594,
"step": 61
},
{
"epoch": 0.6138613861386139,
"grad_norm": 2.791332483291626,
"learning_rate": 0.00017124137884406054,
"loss": 8.9375,
"step": 62
},
{
"epoch": 0.6237623762376238,
"grad_norm": 7.167698383331299,
"learning_rate": 0.00017014837832948087,
"loss": 8.6094,
"step": 63
},
{
"epoch": 0.6336633663366337,
"grad_norm": 4.072234153747559,
"learning_rate": 0.00016903910558234393,
"loss": 8.9688,
"step": 64
},
{
"epoch": 0.6435643564356436,
"grad_norm": 4.949003219604492,
"learning_rate": 0.0001679138606995868,
"loss": 8.7266,
"step": 65
},
{
"epoch": 0.6534653465346535,
"grad_norm": 70.55978393554688,
"learning_rate": 0.00016677294809916597,
"loss": 8.9453,
"step": 66
},
{
"epoch": 0.6633663366336634,
"grad_norm": 4.001098155975342,
"learning_rate": 0.0001656166764377016,
"loss": 8.9453,
"step": 67
},
{
"epoch": 0.6732673267326733,
"grad_norm": 3.2970945835113525,
"learning_rate": 0.00016444535852697508,
"loss": 8.7578,
"step": 68
},
{
"epoch": 0.6831683168316832,
"grad_norm": 7.960860252380371,
"learning_rate": 0.00016325931124930266,
"loss": 8.4883,
"step": 69
},
{
"epoch": 0.693069306930693,
"grad_norm": 4.774203300476074,
"learning_rate": 0.00016205885547180762,
"loss": 9.0859,
"step": 70
},
{
"epoch": 0.7029702970297029,
"grad_norm": 2.7817211151123047,
"learning_rate": 0.00016084431595961464,
"loss": 9.0234,
"step": 71
},
{
"epoch": 0.7128712871287128,
"grad_norm": 3.5456042289733887,
"learning_rate": 0.0001596160212879894,
"loss": 8.5508,
"step": 72
},
{
"epoch": 0.7227722772277227,
"grad_norm": 2.8811049461364746,
"learning_rate": 0.00015837430375344766,
"loss": 8.3203,
"step": 73
},
{
"epoch": 0.7326732673267327,
"grad_norm": 2.634841203689575,
"learning_rate": 0.00015711949928385742,
"loss": 8.2539,
"step": 74
},
{
"epoch": 0.7425742574257426,
"grad_norm": 2.80352520942688,
"learning_rate": 0.0001558519473475585,
"loss": 8.3594,
"step": 75
},
{
"epoch": 0.7524752475247525,
"grad_norm": 2.983501434326172,
"learning_rate": 0.000154571990861525,
"loss": 8.3555,
"step": 76
},
{
"epoch": 0.7623762376237624,
"grad_norm": 2.525681972503662,
"learning_rate": 0.00015327997609859386,
"loss": 8.0117,
"step": 77
},
{
"epoch": 0.7722772277227723,
"grad_norm": 2.6573147773742676,
"learning_rate": 0.0001519762525937862,
"loss": 8.1836,
"step": 78
},
{
"epoch": 0.7821782178217822,
"grad_norm": 3.265751361846924,
"learning_rate": 0.0001506611730497459,
"loss": 8.1836,
"step": 79
},
{
"epoch": 0.7920792079207921,
"grad_norm": 3.280087947845459,
"learning_rate": 0.00014933509324132116,
"loss": 8.2617,
"step": 80
},
{
"epoch": 0.801980198019802,
"grad_norm": 2.2106235027313232,
"learning_rate": 0.0001479983719193151,
"loss": 8.0078,
"step": 81
},
{
"epoch": 0.8118811881188119,
"grad_norm": 2.518003225326538,
"learning_rate": 0.0001466513707134312,
"loss": 8.0078,
"step": 82
},
{
"epoch": 0.8217821782178217,
"grad_norm": 38.51225280761719,
"learning_rate": 0.00014529445403443965,
"loss": 7.7637,
"step": 83
},
{
"epoch": 0.8316831683168316,
"grad_norm": 5.821120738983154,
"learning_rate": 0.0001439279889755918,
"loss": 8.375,
"step": 84
},
{
"epoch": 0.8415841584158416,
"grad_norm": 3.276437282562256,
"learning_rate": 0.0001425523452133084,
"loss": 8.7188,
"step": 85
},
{
"epoch": 0.8514851485148515,
"grad_norm": 6.970194339752197,
"learning_rate": 0.00014116789490716954,
"loss": 8.5781,
"step": 86
},
{
"epoch": 0.8613861386138614,
"grad_norm": 94.84854125976562,
"learning_rate": 0.00013977501259923215,
"loss": 23.9531,
"step": 87
},
{
"epoch": 0.8712871287128713,
"grad_norm": 4.508042335510254,
"learning_rate": 0.00013837407511270365,
"loss": 9.0234,
"step": 88
},
{
"epoch": 0.8811881188118812,
"grad_norm": 2.887805223464966,
"learning_rate": 0.00013696546144999786,
"loss": 8.375,
"step": 89
},
{
"epoch": 0.8910891089108911,
"grad_norm": 2.4554858207702637,
"learning_rate": 0.00013554955269020195,
"loss": 8.3242,
"step": 90
},
{
"epoch": 0.900990099009901,
"grad_norm": 2.6604201793670654,
"learning_rate": 0.00013412673188598092,
"loss": 8.207,
"step": 91
},
{
"epoch": 0.9108910891089109,
"grad_norm": 3.3090832233428955,
"learning_rate": 0.00013269738395994883,
"loss": 8.2266,
"step": 92
},
{
"epoch": 0.9207920792079208,
"grad_norm": 2.517388105392456,
"learning_rate": 0.00013126189560053352,
"loss": 8.1797,
"step": 93
},
{
"epoch": 0.9306930693069307,
"grad_norm": 2.1608598232269287,
"learning_rate": 0.00012982065515736418,
"loss": 7.9102,
"step": 94
},
{
"epoch": 0.9405940594059405,
"grad_norm": 2.3422179222106934,
"learning_rate": 0.00012837405253620905,
"loss": 7.6211,
"step": 95
},
{
"epoch": 0.9504950495049505,
"grad_norm": 2.019343852996826,
"learning_rate": 0.00012692247909349243,
"loss": 7.8633,
"step": 96
},
{
"epoch": 0.9603960396039604,
"grad_norm": 2.547525405883789,
"learning_rate": 0.00012546632753041904,
"loss": 7.5391,
"step": 97
},
{
"epoch": 0.9702970297029703,
"grad_norm": 2.123288869857788,
"learning_rate": 0.00012400599178673483,
"loss": 7.6953,
"step": 98
},
{
"epoch": 0.9801980198019802,
"grad_norm": 2.303816080093384,
"learning_rate": 0.00012254186693415222,
"loss": 7.8242,
"step": 99
},
{
"epoch": 0.9900990099009901,
"grad_norm": 2.4239020347595215,
"learning_rate": 0.00012107434906946982,
"loss": 7.5977,
"step": 100
},
{
"epoch": 1.0,
"grad_norm": 2.0720744132995605,
"learning_rate": 0.00011960383520741419,
"loss": 7.3789,
"step": 101
},
{
"checkpoint_runtime": 117.0071
},
{
"epoch": 1.00990099009901,
"grad_norm": 2.3910915851593018,
"learning_rate": 0.0001181307231732336,
"loss": 6.1016,
"step": 102
},
{
"epoch": 1.0198019801980198,
"grad_norm": 2.539257764816284,
"learning_rate": 0.00011665541149507239,
"loss": 7.4375,
"step": 103
},
{
"epoch": 1.0297029702970297,
"grad_norm": 2.1717493534088135,
"learning_rate": 0.0001151782992961554,
"loss": 7.2188,
"step": 104
},
{
"epoch": 1.0396039603960396,
"grad_norm": 2.1999051570892334,
"learning_rate": 0.0001136997861868109,
"loss": 7.7773,
"step": 105
},
{
"epoch": 1.0495049504950495,
"grad_norm": 20.015344619750977,
"learning_rate": 0.00011222027215636235,
"loss": 13.418,
"step": 106
},
{
"epoch": 1.0594059405940595,
"grad_norm": 3.3290457725524902,
"learning_rate": 0.00011074015746491722,
"loss": 7.6797,
"step": 107
},
{
"epoch": 1.0693069306930694,
"grad_norm": 2.3182663917541504,
"learning_rate": 0.00010925984253508279,
"loss": 7.4141,
"step": 108
},
{
"epoch": 1.0792079207920793,
"grad_norm": 2.7846322059631348,
"learning_rate": 0.00010777972784363765,
"loss": 7.5781,
"step": 109
},
{
"epoch": 1.0891089108910892,
"grad_norm": 3.0692903995513916,
"learning_rate": 0.00010630021381318915,
"loss": 7.207,
"step": 110
},
{
"epoch": 1.099009900990099,
"grad_norm": 2.5749456882476807,
"learning_rate": 0.0001048217007038446,
"loss": 6.8672,
"step": 111
},
{
"epoch": 1.108910891089109,
"grad_norm": 2.473937511444092,
"learning_rate": 0.00010334458850492763,
"loss": 6.7148,
"step": 112
},
{
"epoch": 1.118811881188119,
"grad_norm": 2.3984756469726562,
"learning_rate": 0.00010186927682676646,
"loss": 6.6914,
"step": 113
},
{
"epoch": 1.1287128712871288,
"grad_norm": 2.423884391784668,
"learning_rate": 0.00010039616479258587,
"loss": 7.0469,
"step": 114
},
{
"epoch": 1.1386138613861387,
"grad_norm": 3.279681921005249,
"learning_rate": 9.892565093053024e-05,
"loss": 6.5078,
"step": 115
},
{
"epoch": 1.1485148514851484,
"grad_norm": 3.0356831550598145,
"learning_rate": 9.745813306584781e-05,
"loss": 6.5273,
"step": 116
},
{
"epoch": 1.1584158415841583,
"grad_norm": 2.747650146484375,
"learning_rate": 9.599400821326521e-05,
"loss": 6.4219,
"step": 117
},
{
"epoch": 1.1683168316831682,
"grad_norm": 2.4899802207946777,
"learning_rate": 9.453367246958095e-05,
"loss": 7.0586,
"step": 118
},
{
"epoch": 1.1782178217821782,
"grad_norm": 2.7331626415252686,
"learning_rate": 9.307752090650759e-05,
"loss": 6.3047,
"step": 119
},
{
"epoch": 1.188118811881188,
"grad_norm": 2.377115488052368,
"learning_rate": 9.162594746379097e-05,
"loss": 5.9258,
"step": 120
},
{
"epoch": 1.198019801980198,
"grad_norm": 2.8060243129730225,
"learning_rate": 9.017934484263584e-05,
"loss": 5.6406,
"step": 121
},
{
"epoch": 1.2079207920792079,
"grad_norm": 2.350632429122925,
"learning_rate": 8.873810439946648e-05,
"loss": 5.5859,
"step": 122
},
{
"epoch": 1.2178217821782178,
"grad_norm": 2.519296169281006,
"learning_rate": 8.73026160400512e-05,
"loss": 5.1758,
"step": 123
},
{
"epoch": 1.2277227722772277,
"grad_norm": 2.0400750637054443,
"learning_rate": 8.58732681140191e-05,
"loss": 5.5195,
"step": 124
},
{
"epoch": 1.2376237623762376,
"grad_norm": 1.9192492961883545,
"learning_rate": 8.445044730979808e-05,
"loss": 4.6875,
"step": 125
},
{
"epoch": 1.2475247524752475,
"grad_norm": 2.3437716960906982,
"learning_rate": 8.303453855000217e-05,
"loss": 4.4492,
"step": 126
},
{
"epoch": 1.2574257425742574,
"grad_norm": 2.027672052383423,
"learning_rate": 8.162592488729637e-05,
"loss": 4.3867,
"step": 127
},
{
"epoch": 1.2673267326732673,
"grad_norm": 2.480762243270874,
"learning_rate": 8.022498740076788e-05,
"loss": 4.582,
"step": 128
},
{
"epoch": 1.2772277227722773,
"grad_norm": 2.1822168827056885,
"learning_rate": 7.883210509283048e-05,
"loss": 4.6445,
"step": 129
},
{
"epoch": 1.2871287128712872,
"grad_norm": 2.1713345050811768,
"learning_rate": 7.74476547866916e-05,
"loss": 4.8281,
"step": 130
},
{
"epoch": 1.297029702970297,
"grad_norm": 2.949915885925293,
"learning_rate": 7.607201102440822e-05,
"loss": 4.8945,
"step": 131
},
{
"epoch": 1.306930693069307,
"grad_norm": 5.816825866699219,
"learning_rate": 7.470554596556038e-05,
"loss": 6.4536,
"step": 132
},
{
"epoch": 1.316831683168317,
"grad_norm": 2.8560147285461426,
"learning_rate": 7.334862928656882e-05,
"loss": 6.2148,
"step": 133
},
{
"epoch": 1.3267326732673268,
"grad_norm": 1.9923173189163208,
"learning_rate": 7.200162808068492e-05,
"loss": 4.8516,
"step": 134
},
{
"epoch": 1.3366336633663367,
"grad_norm": 3.521530866622925,
"learning_rate": 7.06649067586789e-05,
"loss": 5.7188,
"step": 135
},
{
"epoch": 1.3465346534653464,
"grad_norm": 32.16452407836914,
"learning_rate": 6.933882695025414e-05,
"loss": 6.3516,
"step": 136
},
{
"epoch": 1.3564356435643563,
"grad_norm": 2.6356420516967773,
"learning_rate": 6.802374740621382e-05,
"loss": 4.4414,
"step": 137
},
{
"epoch": 1.3663366336633662,
"grad_norm": 1.9716286659240723,
"learning_rate": 6.672002390140615e-05,
"loss": 4.2656,
"step": 138
},
{
"epoch": 1.3762376237623761,
"grad_norm": 1.7040444612503052,
"learning_rate": 6.542800913847502e-05,
"loss": 3.9785,
"step": 139
},
{
"epoch": 1.386138613861386,
"grad_norm": 2.3597848415374756,
"learning_rate": 6.414805265244153e-05,
"loss": 4.4141,
"step": 140
},
{
"epoch": 1.396039603960396,
"grad_norm": 2.488919734954834,
"learning_rate": 6.288050071614263e-05,
"loss": 4.8203,
"step": 141
},
{
"epoch": 1.4059405940594059,
"grad_norm": 2.223335027694702,
"learning_rate": 6.162569624655235e-05,
"loss": 4.1738,
"step": 142
},
{
"epoch": 1.4158415841584158,
"grad_norm": 1.9648133516311646,
"learning_rate": 6.0383978712010627e-05,
"loss": 4.0176,
"step": 143
},
{
"epoch": 1.4257425742574257,
"grad_norm": 2.9666850566864014,
"learning_rate": 5.9155684040385396e-05,
"loss": 4.4258,
"step": 144
},
{
"epoch": 1.4356435643564356,
"grad_norm": 2.051849603652954,
"learning_rate": 5.794114452819239e-05,
"loss": 4.0508,
"step": 145
},
{
"epoch": 1.4455445544554455,
"grad_norm": 1.7934975624084473,
"learning_rate": 5.674068875069736e-05,
"loss": 3.6016,
"step": 146
},
{
"epoch": 1.4554455445544554,
"grad_norm": 4.365362167358398,
"learning_rate": 5.5554641473024916e-05,
"loss": 6.043,
"step": 147
},
{
"epoch": 1.4653465346534653,
"grad_norm": 2.678313732147217,
"learning_rate": 5.4383323562298405e-05,
"loss": 5.1289,
"step": 148
},
{
"epoch": 1.4752475247524752,
"grad_norm": 1.8244177103042603,
"learning_rate": 5.3227051900834024e-05,
"loss": 3.7969,
"step": 149
},
{
"epoch": 1.4851485148514851,
"grad_norm": 1.8080298900604248,
"learning_rate": 5.208613930041323e-05,
"loss": 4.1211,
"step": 150
},
{
"epoch": 1.495049504950495,
"grad_norm": 50.39912414550781,
"learning_rate": 5.09608944176561e-05,
"loss": 9.9336,
"step": 151
},
{
"epoch": 1.504950495049505,
"grad_norm": 2.4307775497436523,
"learning_rate": 4.985162167051917e-05,
"loss": 4.9141,
"step": 152
},
{
"epoch": 1.5148514851485149,
"grad_norm": 1.858076572418213,
"learning_rate": 4.875862115593951e-05,
"loss": 3.7559,
"step": 153
},
{
"epoch": 1.5247524752475248,
"grad_norm": 1.654653787612915,
"learning_rate": 4.7682188568648294e-05,
"loss": 3.8965,
"step": 154
},
{
"epoch": 1.5346534653465347,
"grad_norm": 1.9130183458328247,
"learning_rate": 4.662261512117496e-05,
"loss": 3.5098,
"step": 155
},
{
"epoch": 1.5445544554455446,
"grad_norm": 2.3735601902008057,
"learning_rate": 4.558018746506396e-05,
"loss": 3.4219,
"step": 156
},
{
"epoch": 1.5544554455445545,
"grad_norm": 1.6245256662368774,
"learning_rate": 4.455518761332576e-05,
"loss": 3.4375,
"step": 157
},
{
"epoch": 1.5643564356435644,
"grad_norm": 1.7158585786819458,
"learning_rate": 4.35478928641423e-05,
"loss": 4.0195,
"step": 158
},
{
"epoch": 1.5742574257425743,
"grad_norm": 1.537367343902588,
"learning_rate": 4.25585757258483e-05,
"loss": 3.1113,
"step": 159
},
{
"epoch": 1.5841584158415842,
"grad_norm": 2.29418683052063,
"learning_rate": 4.158750384320811e-05,
"loss": 3.9023,
"step": 160
},
{
"epoch": 1.5940594059405941,
"grad_norm": 14.955975532531738,
"learning_rate": 4.063493992500865e-05,
"loss": 7.5469,
"step": 161
},
{
"epoch": 1.603960396039604,
"grad_norm": 2.4816298484802246,
"learning_rate": 3.9701141672987406e-05,
"loss": 4.8594,
"step": 162
},
{
"epoch": 1.613861386138614,
"grad_norm": 1.8471182584762573,
"learning_rate": 3.878636171211518e-05,
"loss": 3.4258,
"step": 163
},
{
"epoch": 1.6237623762376239,
"grad_norm": 1.9999842643737793,
"learning_rate": 3.789084752225219e-05,
"loss": 4.6172,
"step": 164
},
{
"epoch": 1.6336633663366338,
"grad_norm": 1.5966260433197021,
"learning_rate": 3.701484137119613e-05,
"loss": 3.7656,
"step": 165
},
{
"epoch": 1.6435643564356437,
"grad_norm": 1.9665318727493286,
"learning_rate": 3.615858024914048e-05,
"loss": 3.8867,
"step": 166
},
{
"epoch": 1.6534653465346536,
"grad_norm": 39.52935791015625,
"learning_rate": 3.532229580456006e-05,
"loss": 7.5195,
"step": 167
},
{
"epoch": 1.6633663366336635,
"grad_norm": 2.001122236251831,
"learning_rate": 3.450621428154235e-05,
"loss": 4.2148,
"step": 168
},
{
"epoch": 1.6732673267326734,
"grad_norm": 1.9016386270523071,
"learning_rate": 3.371055645858038e-05,
"loss": 3.6035,
"step": 169
},
{
"epoch": 1.6831683168316833,
"grad_norm": 8.8970947265625,
"learning_rate": 3.2935537588844645e-05,
"loss": 5.6367,
"step": 170
},
{
"epoch": 1.693069306930693,
"grad_norm": 1.5331684350967407,
"learning_rate": 3.218136734194936e-05,
"loss": 4.3047,
"step": 171
},
{
"epoch": 1.702970297029703,
"grad_norm": 1.695820927619934,
"learning_rate": 3.144824974722983e-05,
"loss": 3.3164,
"step": 172
},
{
"epoch": 1.7128712871287128,
"grad_norm": 7.624035835266113,
"learning_rate": 3.0736383138545086e-05,
"loss": 3.7305,
"step": 173
},
{
"epoch": 1.7227722772277227,
"grad_norm": 1.5859265327453613,
"learning_rate": 3.0045960100621918e-05,
"loss": 3.2656,
"step": 174
},
{
"epoch": 1.7326732673267327,
"grad_norm": 1.5074691772460938,
"learning_rate": 2.937716741695392e-05,
"loss": 3.1797,
"step": 175
},
{
"epoch": 1.7425742574257426,
"grad_norm": 1.3923929929733276,
"learning_rate": 2.873018601926994e-05,
"loss": 3.1094,
"step": 176
},
{
"epoch": 1.7524752475247525,
"grad_norm": 1.4183502197265625,
"learning_rate": 2.8105190938585885e-05,
"loss": 3.1641,
"step": 177
},
{
"epoch": 1.7623762376237624,
"grad_norm": 1.5268654823303223,
"learning_rate": 2.7502351257852678e-05,
"loss": 2.7832,
"step": 178
},
{
"epoch": 1.7722772277227723,
"grad_norm": 1.5342729091644287,
"learning_rate": 2.692183006621338e-05,
"loss": 2.8652,
"step": 179
},
{
"epoch": 1.7821782178217822,
"grad_norm": 1.6859102249145508,
"learning_rate": 2.636378441488197e-05,
"loss": 3.1641,
"step": 180
},
{
"epoch": 1.7920792079207921,
"grad_norm": 1.5989638566970825,
"learning_rate": 2.5828365274655457e-05,
"loss": 3.0234,
"step": 181
},
{
"epoch": 1.801980198019802,
"grad_norm": 1.5343683958053589,
"learning_rate": 2.5315717495071078e-05,
"loss": 2.5156,
"step": 182
},
{
"epoch": 1.811881188118812,
"grad_norm": 1.508675456047058,
"learning_rate": 2.4825979765219392e-05,
"loss": 2.6621,
"step": 183
},
{
"epoch": 1.8217821782178216,
"grad_norm": 19.975343704223633,
"learning_rate": 2.4359284576224078e-05,
"loss": 7.1133,
"step": 184
},
{
"epoch": 1.8316831683168315,
"grad_norm": 3.3649535179138184,
"learning_rate": 2.3915758185398423e-05,
"loss": 4.6992,
"step": 185
},
{
"epoch": 1.8415841584158414,
"grad_norm": 2.1585795879364014,
"learning_rate": 2.3495520582088455e-05,
"loss": 3.5879,
"step": 186
},
{
"epoch": 1.8514851485148514,
"grad_norm": 2.5024328231811523,
"learning_rate": 2.3098685455211446e-05,
"loss": 4.875,
"step": 187
},
{
"epoch": 1.8613861386138613,
"grad_norm": 15.675300598144531,
"learning_rate": 2.272536016249926e-05,
"loss": 7.0469,
"step": 188
},
{
"epoch": 1.8712871287128712,
"grad_norm": 2.0795493125915527,
"learning_rate": 2.2375645701454248e-05,
"loss": 4.3086,
"step": 189
},
{
"epoch": 1.881188118811881,
"grad_norm": 1.4968867301940918,
"learning_rate": 2.2049636682025957e-05,
"loss": 3.1797,
"step": 190
},
{
"epoch": 1.891089108910891,
"grad_norm": 1.6626455783843994,
"learning_rate": 2.174742130101581e-05,
"loss": 3.2617,
"step": 191
},
{
"epoch": 1.900990099009901,
"grad_norm": 1.4873976707458496,
"learning_rate": 2.1469081318216833e-05,
"loss": 3.0488,
"step": 192
},
{
"epoch": 1.9108910891089108,
"grad_norm": 1.8592286109924316,
"learning_rate": 2.121469203429473e-05,
"loss": 3.2988,
"step": 193
},
{
"epoch": 1.9207920792079207,
"grad_norm": 1.742360234260559,
"learning_rate": 2.0984322270416487e-05,
"loss": 3.0312,
"step": 194
},
{
"epoch": 1.9306930693069306,
"grad_norm": 1.4714621305465698,
"learning_rate": 2.077803434963179e-05,
"loss": 2.6895,
"step": 195
},
{
"epoch": 1.9405940594059405,
"grad_norm": 1.268083095550537,
"learning_rate": 2.059588408001252e-05,
"loss": 2.5215,
"step": 196
},
{
"epoch": 1.9504950495049505,
"grad_norm": 1.3318591117858887,
"learning_rate": 2.0437920739554686e-05,
"loss": 2.543,
"step": 197
},
{
"epoch": 1.9603960396039604,
"grad_norm": 1.3971670866012573,
"learning_rate": 2.0304187062847038e-05,
"loss": 2.6191,
"step": 198
},
{
"epoch": 1.9702970297029703,
"grad_norm": 1.3031758069992065,
"learning_rate": 2.0194719229509842e-05,
"loss": 2.418,
"step": 199
},
{
"epoch": 1.9801980198019802,
"grad_norm": 1.3859241008758545,
"learning_rate": 2.0109546854407064e-05,
"loss": 2.5918,
"step": 200
},
{
"epoch": 1.99009900990099,
"grad_norm": 1.3760048151016235,
"learning_rate": 2.0048692979634493e-05,
"loss": 2.6914,
"step": 201
},
{
"epoch": 2.0,
"grad_norm": 1.2137415409088135,
"learning_rate": 2.0012174068286065e-05,
"loss": 1.6016,
"step": 202
},
{
"epoch": 2.0,
"eval_loss": 2.703822612762451,
"eval_runtime": 12.5311,
"eval_samples_per_second": 4.309,
"eval_steps_per_second": 0.559,
"step": 202
}
],
"logging_steps": 1.0,
"max_steps": 202,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 101,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.5479892660867564e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}