PEFT
Safetensors
q2 / trainer_state.json
chen
Upload folder using huggingface_hub
9e4eed7 verified
Raw
History Blame Contribute Delete
53.2 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 304,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006600660066006601,
"grad_norm": 3.269663095474243,
"learning_rate": 0.0,
"loss": 2.0173,
"step": 1
},
{
"epoch": 0.013201320132013201,
"grad_norm": 3.4657950401306152,
"learning_rate": 4.347826086956522e-06,
"loss": 1.8773,
"step": 2
},
{
"epoch": 0.019801980198019802,
"grad_norm": 3.480679512023926,
"learning_rate": 8.695652173913044e-06,
"loss": 2.2683,
"step": 3
},
{
"epoch": 0.026402640264026403,
"grad_norm": 3.3081204891204834,
"learning_rate": 1.3043478260869566e-05,
"loss": 1.5357,
"step": 4
},
{
"epoch": 0.033003300330033,
"grad_norm": 2.9084601402282715,
"learning_rate": 1.739130434782609e-05,
"loss": 2.3349,
"step": 5
},
{
"epoch": 0.039603960396039604,
"grad_norm": 4.738473892211914,
"learning_rate": 2.173913043478261e-05,
"loss": 2.0195,
"step": 6
},
{
"epoch": 0.0462046204620462,
"grad_norm": 3.0607786178588867,
"learning_rate": 2.608695652173913e-05,
"loss": 1.7303,
"step": 7
},
{
"epoch": 0.052805280528052806,
"grad_norm": 3.43908429145813,
"learning_rate": 3.0434782608695656e-05,
"loss": 1.5714,
"step": 8
},
{
"epoch": 0.0594059405940594,
"grad_norm": 3.3616912364959717,
"learning_rate": 3.478260869565218e-05,
"loss": 2.1655,
"step": 9
},
{
"epoch": 0.066006600660066,
"grad_norm": 2.8642077445983887,
"learning_rate": 3.91304347826087e-05,
"loss": 2.2225,
"step": 10
},
{
"epoch": 0.07260726072607261,
"grad_norm": 1.762931227684021,
"learning_rate": 4.347826086956522e-05,
"loss": 4.3722,
"step": 11
},
{
"epoch": 0.07920792079207921,
"grad_norm": 2.3837015628814697,
"learning_rate": 4.782608695652174e-05,
"loss": 1.134,
"step": 12
},
{
"epoch": 0.0858085808580858,
"grad_norm": 3.5527191162109375,
"learning_rate": 5.217391304347826e-05,
"loss": 1.5813,
"step": 13
},
{
"epoch": 0.0924092409240924,
"grad_norm": 4.409274578094482,
"learning_rate": 5.652173913043478e-05,
"loss": 1.8402,
"step": 14
},
{
"epoch": 0.09900990099009901,
"grad_norm": 3.351221799850464,
"learning_rate": 6.086956521739131e-05,
"loss": 1.6236,
"step": 15
},
{
"epoch": 0.10561056105610561,
"grad_norm": 3.299344301223755,
"learning_rate": 6.521739130434783e-05,
"loss": 1.4048,
"step": 16
},
{
"epoch": 0.11221122112211221,
"grad_norm": 2.9774715900421143,
"learning_rate": 6.956521739130436e-05,
"loss": 1.5322,
"step": 17
},
{
"epoch": 0.1188118811881188,
"grad_norm": 3.526827096939087,
"learning_rate": 7.391304347826086e-05,
"loss": 1.2128,
"step": 18
},
{
"epoch": 0.1254125412541254,
"grad_norm": 1.8688900470733643,
"learning_rate": 7.82608695652174e-05,
"loss": 1.1241,
"step": 19
},
{
"epoch": 0.132013201320132,
"grad_norm": 2.7325925827026367,
"learning_rate": 8.260869565217392e-05,
"loss": 1.4779,
"step": 20
},
{
"epoch": 0.13861386138613863,
"grad_norm": 2.606200695037842,
"learning_rate": 8.695652173913044e-05,
"loss": 1.2094,
"step": 21
},
{
"epoch": 0.14521452145214522,
"grad_norm": 3.906005859375,
"learning_rate": 9.130434782608696e-05,
"loss": 1.9388,
"step": 22
},
{
"epoch": 0.15181518151815182,
"grad_norm": 3.1924426555633545,
"learning_rate": 9.565217391304348e-05,
"loss": 1.6678,
"step": 23
},
{
"epoch": 0.15841584158415842,
"grad_norm": 3.2093305587768555,
"learning_rate": 0.0001,
"loss": 1.8155,
"step": 24
},
{
"epoch": 0.16501650165016502,
"grad_norm": 2.6762614250183105,
"learning_rate": 0.00010434782608695653,
"loss": 1.8512,
"step": 25
},
{
"epoch": 0.1716171617161716,
"grad_norm": 2.4602980613708496,
"learning_rate": 0.00010869565217391305,
"loss": 1.4496,
"step": 26
},
{
"epoch": 0.1782178217821782,
"grad_norm": 2.621077299118042,
"learning_rate": 0.00011304347826086956,
"loss": 1.7004,
"step": 27
},
{
"epoch": 0.1848184818481848,
"grad_norm": 3.040069580078125,
"learning_rate": 0.0001173913043478261,
"loss": 1.3742,
"step": 28
},
{
"epoch": 0.19141914191419143,
"grad_norm": 2.6296839714050293,
"learning_rate": 0.00012173913043478263,
"loss": 1.9487,
"step": 29
},
{
"epoch": 0.19801980198019803,
"grad_norm": 2.5183770656585693,
"learning_rate": 0.00012608695652173915,
"loss": 1.4286,
"step": 30
},
{
"epoch": 0.20462046204620463,
"grad_norm": 2.7780044078826904,
"learning_rate": 0.00013043478260869567,
"loss": 1.6399,
"step": 31
},
{
"epoch": 0.21122112211221122,
"grad_norm": 4.337399959564209,
"learning_rate": 0.0001347826086956522,
"loss": 1.8436,
"step": 32
},
{
"epoch": 0.21782178217821782,
"grad_norm": 2.1603963375091553,
"learning_rate": 0.0001391304347826087,
"loss": 1.3298,
"step": 33
},
{
"epoch": 0.22442244224422442,
"grad_norm": 2.3995110988616943,
"learning_rate": 0.0001434782608695652,
"loss": 1.4131,
"step": 34
},
{
"epoch": 0.23102310231023102,
"grad_norm": 2.0356786251068115,
"learning_rate": 0.00014782608695652173,
"loss": 1.2884,
"step": 35
},
{
"epoch": 0.2376237623762376,
"grad_norm": 2.234151840209961,
"learning_rate": 0.00015217391304347827,
"loss": 1.3331,
"step": 36
},
{
"epoch": 0.24422442244224424,
"grad_norm": 2.045015573501587,
"learning_rate": 0.0001565217391304348,
"loss": 1.1223,
"step": 37
},
{
"epoch": 0.2508250825082508,
"grad_norm": 2.8687806129455566,
"learning_rate": 0.00016086956521739132,
"loss": 1.3582,
"step": 38
},
{
"epoch": 0.25742574257425743,
"grad_norm": 2.717107057571411,
"learning_rate": 0.00016521739130434784,
"loss": 1.8503,
"step": 39
},
{
"epoch": 0.264026402640264,
"grad_norm": 2.2588422298431396,
"learning_rate": 0.00016956521739130436,
"loss": 1.4314,
"step": 40
},
{
"epoch": 0.2706270627062706,
"grad_norm": 2.3839213848114014,
"learning_rate": 0.00017391304347826088,
"loss": 1.6702,
"step": 41
},
{
"epoch": 0.27722772277227725,
"grad_norm": 2.2554593086242676,
"learning_rate": 0.0001782608695652174,
"loss": 1.6114,
"step": 42
},
{
"epoch": 0.2838283828382838,
"grad_norm": 3.0558981895446777,
"learning_rate": 0.00018260869565217392,
"loss": 1.7576,
"step": 43
},
{
"epoch": 0.29042904290429045,
"grad_norm": 1.9171123504638672,
"learning_rate": 0.00018695652173913045,
"loss": 1.0488,
"step": 44
},
{
"epoch": 0.297029702970297,
"grad_norm": 2.573887825012207,
"learning_rate": 0.00019130434782608697,
"loss": 1.5062,
"step": 45
},
{
"epoch": 0.30363036303630364,
"grad_norm": 2.058779716491699,
"learning_rate": 0.0001956521739130435,
"loss": 1.1712,
"step": 46
},
{
"epoch": 0.3102310231023102,
"grad_norm": 4.420455455780029,
"learning_rate": 0.0002,
"loss": 2.2398,
"step": 47
},
{
"epoch": 0.31683168316831684,
"grad_norm": 2.39105224609375,
"learning_rate": 0.00019999702094326033,
"loss": 1.3762,
"step": 48
},
{
"epoch": 0.3234323432343234,
"grad_norm": 1.9990428686141968,
"learning_rate": 0.00019998808395053688,
"loss": 1.8914,
"step": 49
},
{
"epoch": 0.33003300330033003,
"grad_norm": 1.8856112957000732,
"learning_rate": 0.0001999731895543058,
"loss": 1.239,
"step": 50
},
{
"epoch": 0.33663366336633666,
"grad_norm": 1.7601982355117798,
"learning_rate": 0.00019995233864199215,
"loss": 1.31,
"step": 51
},
{
"epoch": 0.3432343234323432,
"grad_norm": 2.37358021736145,
"learning_rate": 0.00019992553245591694,
"loss": 1.2271,
"step": 52
},
{
"epoch": 0.34983498349834985,
"grad_norm": 1.8610317707061768,
"learning_rate": 0.00019989277259322314,
"loss": 1.222,
"step": 53
},
{
"epoch": 0.3564356435643564,
"grad_norm": 2.4520764350891113,
"learning_rate": 0.0001998540610057806,
"loss": 1.8557,
"step": 54
},
{
"epoch": 0.36303630363036304,
"grad_norm": 1.9622912406921387,
"learning_rate": 0.00019980940000006954,
"loss": 1.3294,
"step": 55
},
{
"epoch": 0.3696369636963696,
"grad_norm": 2.875389337539673,
"learning_rate": 0.0001997587922370434,
"loss": 1.0667,
"step": 56
},
{
"epoch": 0.37623762376237624,
"grad_norm": 2.358046531677246,
"learning_rate": 0.00019970224073197017,
"loss": 2.0223,
"step": 57
},
{
"epoch": 0.38283828382838286,
"grad_norm": 2.127352237701416,
"learning_rate": 0.00019963974885425266,
"loss": 1.5623,
"step": 58
},
{
"epoch": 0.38943894389438943,
"grad_norm": 2.165083885192871,
"learning_rate": 0.00019957132032722785,
"loss": 1.6115,
"step": 59
},
{
"epoch": 0.39603960396039606,
"grad_norm": 2.4137611389160156,
"learning_rate": 0.00019949695922794506,
"loss": 1.5646,
"step": 60
},
{
"epoch": 0.40264026402640263,
"grad_norm": 2.246424913406372,
"learning_rate": 0.00019941666998692296,
"loss": 1.7207,
"step": 61
},
{
"epoch": 0.40924092409240925,
"grad_norm": 2.1227519512176514,
"learning_rate": 0.00019933045738788563,
"loss": 1.6483,
"step": 62
},
{
"epoch": 0.4158415841584158,
"grad_norm": 2.097881317138672,
"learning_rate": 0.0001992383265674776,
"loss": 1.2604,
"step": 63
},
{
"epoch": 0.42244224422442245,
"grad_norm": 2.4792444705963135,
"learning_rate": 0.0001991402830149576,
"loss": 1.3448,
"step": 64
},
{
"epoch": 0.429042904290429,
"grad_norm": 2.302548408508301,
"learning_rate": 0.00019903633257187185,
"loss": 1.5288,
"step": 65
},
{
"epoch": 0.43564356435643564,
"grad_norm": 1.9828969240188599,
"learning_rate": 0.00019892648143170566,
"loss": 1.1426,
"step": 66
},
{
"epoch": 0.44224422442244227,
"grad_norm": 2.2925431728363037,
"learning_rate": 0.0001988107361395146,
"loss": 1.3227,
"step": 67
},
{
"epoch": 0.44884488448844884,
"grad_norm": 3.0168843269348145,
"learning_rate": 0.0001986891035915346,
"loss": 2.0713,
"step": 68
},
{
"epoch": 0.45544554455445546,
"grad_norm": 2.2189230918884277,
"learning_rate": 0.00019856159103477086,
"loss": 1.2472,
"step": 69
},
{
"epoch": 0.46204620462046203,
"grad_norm": 1.9616774320602417,
"learning_rate": 0.0001984282060665662,
"loss": 1.4617,
"step": 70
},
{
"epoch": 0.46864686468646866,
"grad_norm": 3.0113203525543213,
"learning_rate": 0.00019828895663414836,
"loss": 1.5551,
"step": 71
},
{
"epoch": 0.4752475247524752,
"grad_norm": 2.4521045684814453,
"learning_rate": 0.00019814385103415663,
"loss": 1.405,
"step": 72
},
{
"epoch": 0.48184818481848185,
"grad_norm": 2.3250396251678467,
"learning_rate": 0.00019799289791214725,
"loss": 1.8011,
"step": 73
},
{
"epoch": 0.4884488448844885,
"grad_norm": 2.3330326080322266,
"learning_rate": 0.00019783610626207854,
"loss": 1.6456,
"step": 74
},
{
"epoch": 0.49504950495049505,
"grad_norm": 2.1660053730010986,
"learning_rate": 0.00019767348542577495,
"loss": 1.5428,
"step": 75
},
{
"epoch": 0.5016501650165016,
"grad_norm": 1.8240512609481812,
"learning_rate": 0.00019750504509237046,
"loss": 1.4666,
"step": 76
},
{
"epoch": 0.5082508250825083,
"grad_norm": 2.2784554958343506,
"learning_rate": 0.00019733079529773125,
"loss": 1.5065,
"step": 77
},
{
"epoch": 0.5148514851485149,
"grad_norm": 1.769106149673462,
"learning_rate": 0.00019715074642385786,
"loss": 1.0998,
"step": 78
},
{
"epoch": 0.5214521452145214,
"grad_norm": 1.737535834312439,
"learning_rate": 0.00019696490919826647,
"loss": 1.3843,
"step": 79
},
{
"epoch": 0.528052805280528,
"grad_norm": 1.737778663635254,
"learning_rate": 0.0001967732946933499,
"loss": 1.5803,
"step": 80
},
{
"epoch": 0.5346534653465347,
"grad_norm": 1.9771817922592163,
"learning_rate": 0.00019657591432571778,
"loss": 1.7888,
"step": 81
},
{
"epoch": 0.5412541254125413,
"grad_norm": 1.922725796699524,
"learning_rate": 0.0001963727798555164,
"loss": 1.096,
"step": 82
},
{
"epoch": 0.5478547854785478,
"grad_norm": 2.42258620262146,
"learning_rate": 0.00019616390338572803,
"loss": 1.4231,
"step": 83
},
{
"epoch": 0.5544554455445545,
"grad_norm": 2.0959455966949463,
"learning_rate": 0.00019594929736144976,
"loss": 1.3786,
"step": 84
},
{
"epoch": 0.5610561056105611,
"grad_norm": 2.2483432292938232,
"learning_rate": 0.00019572897456915202,
"loss": 1.5456,
"step": 85
},
{
"epoch": 0.5676567656765676,
"grad_norm": 2.2256951332092285,
"learning_rate": 0.00019550294813591682,
"loss": 1.3946,
"step": 86
},
{
"epoch": 0.5742574257425742,
"grad_norm": 1.9461501836776733,
"learning_rate": 0.0001952712315286556,
"loss": 1.3132,
"step": 87
},
{
"epoch": 0.5808580858085809,
"grad_norm": 1.8409348726272583,
"learning_rate": 0.00019503383855330668,
"loss": 1.1527,
"step": 88
},
{
"epoch": 0.5874587458745875,
"grad_norm": 2.124138832092285,
"learning_rate": 0.00019479078335401297,
"loss": 1.0499,
"step": 89
},
{
"epoch": 0.594059405940594,
"grad_norm": 2.380868673324585,
"learning_rate": 0.00019454208041227907,
"loss": 1.7297,
"step": 90
},
{
"epoch": 0.6006600660066007,
"grad_norm": 2.093554973602295,
"learning_rate": 0.00019428774454610843,
"loss": 1.3311,
"step": 91
},
{
"epoch": 0.6072607260726073,
"grad_norm": 2.391789436340332,
"learning_rate": 0.0001940277909091206,
"loss": 1.4725,
"step": 92
},
{
"epoch": 0.6138613861386139,
"grad_norm": 2.362126350402832,
"learning_rate": 0.0001937622349896483,
"loss": 1.4505,
"step": 93
},
{
"epoch": 0.6204620462046204,
"grad_norm": 2.4472386837005615,
"learning_rate": 0.00019349109260981452,
"loss": 1.4643,
"step": 94
},
{
"epoch": 0.6270627062706271,
"grad_norm": 2.7160544395446777,
"learning_rate": 0.00019321437992458995,
"loss": 1.6129,
"step": 95
},
{
"epoch": 0.6336633663366337,
"grad_norm": 1.9917261600494385,
"learning_rate": 0.0001929321134208304,
"loss": 1.4359,
"step": 96
},
{
"epoch": 0.6402640264026402,
"grad_norm": 1.98976731300354,
"learning_rate": 0.00019264430991629446,
"loss": 1.4551,
"step": 97
},
{
"epoch": 0.6468646864686468,
"grad_norm": 2.612914562225342,
"learning_rate": 0.00019235098655864157,
"loss": 1.4066,
"step": 98
},
{
"epoch": 0.6534653465346535,
"grad_norm": 1.5987509489059448,
"learning_rate": 0.00019205216082441017,
"loss": 1.4239,
"step": 99
},
{
"epoch": 0.6600660066006601,
"grad_norm": 2.1327996253967285,
"learning_rate": 0.00019174785051797666,
"loss": 1.1858,
"step": 100
},
{
"epoch": 0.6666666666666666,
"grad_norm": 3.5038344860076904,
"learning_rate": 0.0001914380737704944,
"loss": 1.5437,
"step": 101
},
{
"epoch": 0.6732673267326733,
"grad_norm": 2.0759711265563965,
"learning_rate": 0.0001911228490388136,
"loss": 1.9746,
"step": 102
},
{
"epoch": 0.6798679867986799,
"grad_norm": 1.7349865436553955,
"learning_rate": 0.0001908021951043814,
"loss": 1.3087,
"step": 103
},
{
"epoch": 0.6864686468646864,
"grad_norm": 1.8272370100021362,
"learning_rate": 0.00019047613107212312,
"loss": 1.5015,
"step": 104
},
{
"epoch": 0.693069306930693,
"grad_norm": 2.6153130531311035,
"learning_rate": 0.00019014467636930385,
"loss": 1.6917,
"step": 105
},
{
"epoch": 0.6996699669966997,
"grad_norm": 2.020885705947876,
"learning_rate": 0.00018980785074437093,
"loss": 1.3401,
"step": 106
},
{
"epoch": 0.7062706270627063,
"grad_norm": 2.322054147720337,
"learning_rate": 0.00018946567426577723,
"loss": 2.2148,
"step": 107
},
{
"epoch": 0.7128712871287128,
"grad_norm": 2.2210605144500732,
"learning_rate": 0.00018911816732078575,
"loss": 1.6508,
"step": 108
},
{
"epoch": 0.7194719471947195,
"grad_norm": 2.340949058532715,
"learning_rate": 0.00018876535061425453,
"loss": 1.8924,
"step": 109
},
{
"epoch": 0.7260726072607261,
"grad_norm": 2.295544385910034,
"learning_rate": 0.0001884072451674034,
"loss": 1.621,
"step": 110
},
{
"epoch": 0.7326732673267327,
"grad_norm": 1.9485543966293335,
"learning_rate": 0.00018804387231656118,
"loss": 1.3108,
"step": 111
},
{
"epoch": 0.7392739273927392,
"grad_norm": 1.9073518514633179,
"learning_rate": 0.00018767525371189475,
"loss": 1.1203,
"step": 112
},
{
"epoch": 0.7458745874587459,
"grad_norm": 1.9978621006011963,
"learning_rate": 0.00018730141131611882,
"loss": 1.5631,
"step": 113
},
{
"epoch": 0.7524752475247525,
"grad_norm": 2.061638355255127,
"learning_rate": 0.0001869223674031876,
"loss": 1.4612,
"step": 114
},
{
"epoch": 0.759075907590759,
"grad_norm": 1.9964576959609985,
"learning_rate": 0.00018653814455696756,
"loss": 1.4753,
"step": 115
},
{
"epoch": 0.7656765676567657,
"grad_norm": 2.142472505569458,
"learning_rate": 0.0001861487656698919,
"loss": 1.4195,
"step": 116
},
{
"epoch": 0.7722772277227723,
"grad_norm": 2.622685194015503,
"learning_rate": 0.00018575425394159653,
"loss": 1.8376,
"step": 117
},
{
"epoch": 0.7788778877887789,
"grad_norm": 1.8622280359268188,
"learning_rate": 0.00018535463287753796,
"loss": 1.3869,
"step": 118
},
{
"epoch": 0.7854785478547854,
"grad_norm": 1.8239712715148926,
"learning_rate": 0.00018494992628759267,
"loss": 1.3901,
"step": 119
},
{
"epoch": 0.7920792079207921,
"grad_norm": 1.7509407997131348,
"learning_rate": 0.0001845401582846385,
"loss": 1.1086,
"step": 120
},
{
"epoch": 0.7986798679867987,
"grad_norm": 2.12126088142395,
"learning_rate": 0.00018412535328311814,
"loss": 1.4786,
"step": 121
},
{
"epoch": 0.8052805280528053,
"grad_norm": 1.9442270994186401,
"learning_rate": 0.00018370553599758422,
"loss": 1.2655,
"step": 122
},
{
"epoch": 0.8118811881188119,
"grad_norm": 1.8618887662887573,
"learning_rate": 0.00018328073144122708,
"loss": 1.1703,
"step": 123
},
{
"epoch": 0.8184818481848185,
"grad_norm": 1.9630706310272217,
"learning_rate": 0.00018285096492438424,
"loss": 1.2309,
"step": 124
},
{
"epoch": 0.8250825082508251,
"grad_norm": 2.227321147918701,
"learning_rate": 0.00018241626205303243,
"loss": 1.6171,
"step": 125
},
{
"epoch": 0.8316831683168316,
"grad_norm": 1.6737253665924072,
"learning_rate": 0.00018197664872726205,
"loss": 1.0966,
"step": 126
},
{
"epoch": 0.8382838283828383,
"grad_norm": 1.8567694425582886,
"learning_rate": 0.00018153215113973396,
"loss": 1.262,
"step": 127
},
{
"epoch": 0.8448844884488449,
"grad_norm": 2.6924386024475098,
"learning_rate": 0.0001810827957741188,
"loss": 1.3436,
"step": 128
},
{
"epoch": 0.8514851485148515,
"grad_norm": 2.324448823928833,
"learning_rate": 0.00018062860940351916,
"loss": 1.3375,
"step": 129
},
{
"epoch": 0.858085808580858,
"grad_norm": 2.095839738845825,
"learning_rate": 0.00018016961908887442,
"loss": 1.4675,
"step": 130
},
{
"epoch": 0.8646864686468647,
"grad_norm": 1.6898919343948364,
"learning_rate": 0.00017970585217734844,
"loss": 1.0081,
"step": 131
},
{
"epoch": 0.8712871287128713,
"grad_norm": 3.500580072402954,
"learning_rate": 0.00017923733630070002,
"loss": 1.9837,
"step": 132
},
{
"epoch": 0.8778877887788779,
"grad_norm": 2.213259696960449,
"learning_rate": 0.00017876409937363677,
"loss": 1.3967,
"step": 133
},
{
"epoch": 0.8844884488448845,
"grad_norm": 1.764636516571045,
"learning_rate": 0.00017828616959215184,
"loss": 1.4183,
"step": 134
},
{
"epoch": 0.8910891089108911,
"grad_norm": 1.8215689659118652,
"learning_rate": 0.00017780357543184397,
"loss": 1.5714,
"step": 135
},
{
"epoch": 0.8976897689768977,
"grad_norm": 1.726089358329773,
"learning_rate": 0.00017731634564622085,
"loss": 1.4959,
"step": 136
},
{
"epoch": 0.9042904290429042,
"grad_norm": 1.7792177200317383,
"learning_rate": 0.00017682450926498607,
"loss": 1.1855,
"step": 137
},
{
"epoch": 0.9108910891089109,
"grad_norm": 1.6838842630386353,
"learning_rate": 0.0001763280955923093,
"loss": 1.3365,
"step": 138
},
{
"epoch": 0.9174917491749175,
"grad_norm": 1.692114233970642,
"learning_rate": 0.00017582713420508052,
"loss": 1.354,
"step": 139
},
{
"epoch": 0.9240924092409241,
"grad_norm": 1.8941043615341187,
"learning_rate": 0.00017532165495114766,
"loss": 1.3819,
"step": 140
},
{
"epoch": 0.9306930693069307,
"grad_norm": 1.8271002769470215,
"learning_rate": 0.0001748116879475383,
"loss": 1.3082,
"step": 141
},
{
"epoch": 0.9372937293729373,
"grad_norm": 2.272860288619995,
"learning_rate": 0.00017429726357866515,
"loss": 1.6585,
"step": 142
},
{
"epoch": 0.9438943894389439,
"grad_norm": 1.9342212677001953,
"learning_rate": 0.00017377841249451595,
"loss": 1.492,
"step": 143
},
{
"epoch": 0.9504950495049505,
"grad_norm": 2.028965950012207,
"learning_rate": 0.00017325516560882703,
"loss": 1.4987,
"step": 144
},
{
"epoch": 0.9570957095709571,
"grad_norm": 2.096977710723877,
"learning_rate": 0.00017272755409724168,
"loss": 1.489,
"step": 145
},
{
"epoch": 0.9636963696369637,
"grad_norm": 2.228801727294922,
"learning_rate": 0.00017219560939545246,
"loss": 1.4862,
"step": 146
},
{
"epoch": 0.9702970297029703,
"grad_norm": 1.7269052267074585,
"learning_rate": 0.0001716593631973283,
"loss": 1.2565,
"step": 147
},
{
"epoch": 0.976897689768977,
"grad_norm": 1.8832287788391113,
"learning_rate": 0.00017111884745302632,
"loss": 1.3296,
"step": 148
},
{
"epoch": 0.9834983498349835,
"grad_norm": 1.8937143087387085,
"learning_rate": 0.00017057409436708782,
"loss": 1.2752,
"step": 149
},
{
"epoch": 0.9900990099009901,
"grad_norm": 2.1348111629486084,
"learning_rate": 0.00017002513639651992,
"loss": 1.2174,
"step": 150
},
{
"epoch": 0.9966996699669967,
"grad_norm": 1.7981044054031372,
"learning_rate": 0.00016947200624886145,
"loss": 1.3272,
"step": 151
},
{
"epoch": 1.0,
"grad_norm": 3.076122999191284,
"learning_rate": 0.00016891473688023426,
"loss": 1.3359,
"step": 152
},
{
"epoch": 1.0066006600660067,
"grad_norm": 1.5830563306808472,
"learning_rate": 0.00016835336149337974,
"loss": 0.9539,
"step": 153
},
{
"epoch": 1.0132013201320131,
"grad_norm": 1.7160429954528809,
"learning_rate": 0.0001677879135356805,
"loss": 0.7741,
"step": 154
},
{
"epoch": 1.0198019801980198,
"grad_norm": 1.4492639303207397,
"learning_rate": 0.00016721842669716752,
"loss": 0.8243,
"step": 155
},
{
"epoch": 1.0264026402640265,
"grad_norm": 1.7320116758346558,
"learning_rate": 0.00016664493490851288,
"loss": 0.7563,
"step": 156
},
{
"epoch": 1.033003300330033,
"grad_norm": 1.8974308967590332,
"learning_rate": 0.00016606747233900815,
"loss": 0.8068,
"step": 157
},
{
"epoch": 1.0396039603960396,
"grad_norm": 1.688904881477356,
"learning_rate": 0.00016548607339452853,
"loss": 0.8673,
"step": 158
},
{
"epoch": 1.046204620462046,
"grad_norm": 1.5127521753311157,
"learning_rate": 0.00016490077271548285,
"loss": 0.6085,
"step": 159
},
{
"epoch": 1.0528052805280528,
"grad_norm": 1.6967127323150635,
"learning_rate": 0.00016431160517474985,
"loss": 0.6155,
"step": 160
},
{
"epoch": 1.0594059405940595,
"grad_norm": 2.6661343574523926,
"learning_rate": 0.00016371860587560008,
"loss": 0.8918,
"step": 161
},
{
"epoch": 1.066006600660066,
"grad_norm": 1.9106711149215698,
"learning_rate": 0.0001631218101496048,
"loss": 0.7613,
"step": 162
},
{
"epoch": 1.0726072607260726,
"grad_norm": 2.284550428390503,
"learning_rate": 0.0001625212535545306,
"loss": 0.95,
"step": 163
},
{
"epoch": 1.0792079207920793,
"grad_norm": 2.4170072078704834,
"learning_rate": 0.00016191697187222087,
"loss": 0.9478,
"step": 164
},
{
"epoch": 1.0858085808580857,
"grad_norm": 2.5793893337249756,
"learning_rate": 0.00016130900110646407,
"loss": 0.9573,
"step": 165
},
{
"epoch": 1.0924092409240924,
"grad_norm": 2.0098459720611572,
"learning_rate": 0.00016069737748084824,
"loss": 0.7022,
"step": 166
},
{
"epoch": 1.099009900990099,
"grad_norm": 1.7491934299468994,
"learning_rate": 0.0001600821374366031,
"loss": 0.727,
"step": 167
},
{
"epoch": 1.1056105610561056,
"grad_norm": 2.729426860809326,
"learning_rate": 0.00015946331763042867,
"loss": 0.7939,
"step": 168
},
{
"epoch": 1.1122112211221122,
"grad_norm": 2.475818634033203,
"learning_rate": 0.00015884095493231125,
"loss": 1.3139,
"step": 169
},
{
"epoch": 1.118811881188119,
"grad_norm": 2.3645308017730713,
"learning_rate": 0.00015821508642332657,
"loss": 0.6167,
"step": 170
},
{
"epoch": 1.1254125412541254,
"grad_norm": 1.958868384361267,
"learning_rate": 0.00015758574939343074,
"loss": 1.1419,
"step": 171
},
{
"epoch": 1.132013201320132,
"grad_norm": 1.6585627794265747,
"learning_rate": 0.00015695298133923809,
"loss": 0.6212,
"step": 172
},
{
"epoch": 1.1386138613861387,
"grad_norm": 1.9600220918655396,
"learning_rate": 0.00015631681996178736,
"loss": 0.626,
"step": 173
},
{
"epoch": 1.1452145214521452,
"grad_norm": 1.6481058597564697,
"learning_rate": 0.00015567730316429535,
"loss": 0.7769,
"step": 174
},
{
"epoch": 1.1518151815181519,
"grad_norm": 2.365338087081909,
"learning_rate": 0.00015503446904989855,
"loss": 1.0907,
"step": 175
},
{
"epoch": 1.1584158415841583,
"grad_norm": 1.9022090435028076,
"learning_rate": 0.000154388355919383,
"loss": 0.7317,
"step": 176
},
{
"epoch": 1.165016501650165,
"grad_norm": 1.998331904411316,
"learning_rate": 0.0001537390022689022,
"loss": 0.7538,
"step": 177
},
{
"epoch": 1.1716171617161717,
"grad_norm": 1.7269340753555298,
"learning_rate": 0.00015308644678768357,
"loss": 0.8061,
"step": 178
},
{
"epoch": 1.1782178217821782,
"grad_norm": 1.9528826475143433,
"learning_rate": 0.00015243072835572318,
"loss": 0.6424,
"step": 179
},
{
"epoch": 1.1848184818481848,
"grad_norm": 1.9622853994369507,
"learning_rate": 0.0001517718860414693,
"loss": 0.8983,
"step": 180
},
{
"epoch": 1.1914191419141915,
"grad_norm": 1.8552788496017456,
"learning_rate": 0.00015110995909949464,
"loss": 0.5563,
"step": 181
},
{
"epoch": 1.198019801980198,
"grad_norm": 2.066662073135376,
"learning_rate": 0.00015044498696815758,
"loss": 0.7457,
"step": 182
},
{
"epoch": 1.2046204620462047,
"grad_norm": 1.888692021369934,
"learning_rate": 0.0001497770092672523,
"loss": 0.6713,
"step": 183
},
{
"epoch": 1.2112211221122111,
"grad_norm": 2.0183756351470947,
"learning_rate": 0.00014910606579564829,
"loss": 0.912,
"step": 184
},
{
"epoch": 1.2178217821782178,
"grad_norm": 2.3537251949310303,
"learning_rate": 0.0001484321965289189,
"loss": 1.0088,
"step": 185
},
{
"epoch": 1.2244224422442245,
"grad_norm": 2.7736098766326904,
"learning_rate": 0.00014775544161695975,
"loss": 1.2947,
"step": 186
},
{
"epoch": 1.231023102310231,
"grad_norm": 1.98241126537323,
"learning_rate": 0.0001470758413815965,
"loss": 0.5495,
"step": 187
},
{
"epoch": 1.2376237623762376,
"grad_norm": 1.9791995286941528,
"learning_rate": 0.00014639343631418238,
"loss": 0.5694,
"step": 188
},
{
"epoch": 1.2442244224422443,
"grad_norm": 1.8967821598052979,
"learning_rate": 0.0001457082670731857,
"loss": 0.7433,
"step": 189
},
{
"epoch": 1.2508250825082508,
"grad_norm": 2.317504644393921,
"learning_rate": 0.00014502037448176734,
"loss": 0.9576,
"step": 190
},
{
"epoch": 1.2574257425742574,
"grad_norm": 1.922574520111084,
"learning_rate": 0.00014432979952534852,
"loss": 0.9927,
"step": 191
},
{
"epoch": 1.2640264026402641,
"grad_norm": 2.2528586387634277,
"learning_rate": 0.00014363658334916882,
"loss": 0.7291,
"step": 192
},
{
"epoch": 1.2706270627062706,
"grad_norm": 1.7135688066482544,
"learning_rate": 0.00014294076725583464,
"loss": 0.6866,
"step": 193
},
{
"epoch": 1.2772277227722773,
"grad_norm": 2.167564630508423,
"learning_rate": 0.00014224239270285846,
"loss": 0.7121,
"step": 194
},
{
"epoch": 1.283828382838284,
"grad_norm": 1.8123031854629517,
"learning_rate": 0.00014154150130018866,
"loss": 0.6412,
"step": 195
},
{
"epoch": 1.2904290429042904,
"grad_norm": 2.0353360176086426,
"learning_rate": 0.00014083813480773036,
"loss": 1.0845,
"step": 196
},
{
"epoch": 1.297029702970297,
"grad_norm": 1.9678878784179688,
"learning_rate": 0.00014013233513285732,
"loss": 0.7445,
"step": 197
},
{
"epoch": 1.3036303630363038,
"grad_norm": 1.7406024932861328,
"learning_rate": 0.0001394241443279152,
"loss": 0.6806,
"step": 198
},
{
"epoch": 1.3102310231023102,
"grad_norm": 1.7399944067001343,
"learning_rate": 0.00013871360458771575,
"loss": 0.6555,
"step": 199
},
{
"epoch": 1.316831683168317,
"grad_norm": 2.085008144378662,
"learning_rate": 0.000138000758247023,
"loss": 0.5252,
"step": 200
},
{
"epoch": 1.3234323432343233,
"grad_norm": 2.6593925952911377,
"learning_rate": 0.00013728564777803088,
"loss": 0.8745,
"step": 201
},
{
"epoch": 1.33003300330033,
"grad_norm": 1.8092249631881714,
"learning_rate": 0.00013656831578783262,
"loss": 0.6941,
"step": 202
},
{
"epoch": 1.3366336633663367,
"grad_norm": 1.9926271438598633,
"learning_rate": 0.00013584880501588223,
"loss": 0.694,
"step": 203
},
{
"epoch": 1.3432343234323432,
"grad_norm": 2.030097246170044,
"learning_rate": 0.000135127158331448,
"loss": 0.7283,
"step": 204
},
{
"epoch": 1.3498349834983498,
"grad_norm": 2.254887104034424,
"learning_rate": 0.00013440341873105832,
"loss": 0.8754,
"step": 205
},
{
"epoch": 1.3564356435643563,
"grad_norm": 2.724303960800171,
"learning_rate": 0.00013367762933593988,
"loss": 1.1761,
"step": 206
},
{
"epoch": 1.363036303630363,
"grad_norm": 1.1965293884277344,
"learning_rate": 0.00013294983338944842,
"loss": 4.0649,
"step": 207
},
{
"epoch": 1.3696369636963697,
"grad_norm": 2.0384724140167236,
"learning_rate": 0.00013222007425449234,
"loss": 0.8273,
"step": 208
},
{
"epoch": 1.3762376237623761,
"grad_norm": 2.0111265182495117,
"learning_rate": 0.00013148839541094907,
"loss": 0.6496,
"step": 209
},
{
"epoch": 1.3828382838283828,
"grad_norm": 2.6681222915649414,
"learning_rate": 0.00013075484045307442,
"loss": 1.0668,
"step": 210
},
{
"epoch": 1.3894389438943895,
"grad_norm": 2.460010051727295,
"learning_rate": 0.00013001945308690514,
"loss": 0.7771,
"step": 211
},
{
"epoch": 1.396039603960396,
"grad_norm": 1.731806993484497,
"learning_rate": 0.00012928227712765504,
"loss": 0.5023,
"step": 212
},
{
"epoch": 1.4026402640264026,
"grad_norm": 1.795204997062683,
"learning_rate": 0.00012854335649710436,
"loss": 0.547,
"step": 213
},
{
"epoch": 1.4092409240924093,
"grad_norm": 2.8087148666381836,
"learning_rate": 0.00012780273522098275,
"loss": 1.0993,
"step": 214
},
{
"epoch": 1.4158415841584158,
"grad_norm": 2.681370735168457,
"learning_rate": 0.00012706045742634636,
"loss": 0.926,
"step": 215
},
{
"epoch": 1.4224422442244224,
"grad_norm": 1.771410584449768,
"learning_rate": 0.0001263165673389484,
"loss": 0.5915,
"step": 216
},
{
"epoch": 1.4290429042904291,
"grad_norm": 2.15568208694458,
"learning_rate": 0.00012557110928060455,
"loss": 0.7995,
"step": 217
},
{
"epoch": 1.4356435643564356,
"grad_norm": 1.9349448680877686,
"learning_rate": 0.00012482412766655183,
"loss": 0.6768,
"step": 218
},
{
"epoch": 1.4422442244224423,
"grad_norm": 2.04817795753479,
"learning_rate": 0.00012407566700280246,
"loss": 1.0367,
"step": 219
},
{
"epoch": 1.448844884488449,
"grad_norm": 1.6918604373931885,
"learning_rate": 0.00012332577188349216,
"loss": 0.506,
"step": 220
},
{
"epoch": 1.4554455445544554,
"grad_norm": 2.418933868408203,
"learning_rate": 0.00012257448698822313,
"loss": 1.0148,
"step": 221
},
{
"epoch": 1.462046204620462,
"grad_norm": 2.0591979026794434,
"learning_rate": 0.00012182185707940194,
"loss": 0.7387,
"step": 222
},
{
"epoch": 1.4686468646864688,
"grad_norm": 2.7957074642181396,
"learning_rate": 0.00012106792699957263,
"loss": 1.0185,
"step": 223
},
{
"epoch": 1.4752475247524752,
"grad_norm": 1.8848057985305786,
"learning_rate": 0.00012031274166874498,
"loss": 0.8376,
"step": 224
},
{
"epoch": 1.481848184818482,
"grad_norm": 1.6933871507644653,
"learning_rate": 0.00011955634608171791,
"loss": 0.3668,
"step": 225
},
{
"epoch": 1.4884488448844886,
"grad_norm": 3.080806016921997,
"learning_rate": 0.0001187987853053989,
"loss": 1.0281,
"step": 226
},
{
"epoch": 1.495049504950495,
"grad_norm": 2.0880093574523926,
"learning_rate": 0.00011804010447611862,
"loss": 0.5822,
"step": 227
},
{
"epoch": 1.5016501650165015,
"grad_norm": 1.8675711154937744,
"learning_rate": 0.00011728034879694185,
"loss": 0.8522,
"step": 228
},
{
"epoch": 1.5082508250825084,
"grad_norm": 2.2376694679260254,
"learning_rate": 0.00011651956353497418,
"loss": 0.6734,
"step": 229
},
{
"epoch": 1.5148514851485149,
"grad_norm": 1.8461328744888306,
"learning_rate": 0.00011575779401866475,
"loss": 0.5624,
"step": 230
},
{
"epoch": 1.5214521452145213,
"grad_norm": 2.6068685054779053,
"learning_rate": 0.00011499508563510586,
"loss": 1.1245,
"step": 231
},
{
"epoch": 1.528052805280528,
"grad_norm": 1.9328029155731201,
"learning_rate": 0.00011423148382732853,
"loss": 0.7624,
"step": 232
},
{
"epoch": 1.5346534653465347,
"grad_norm": 2.3885467052459717,
"learning_rate": 0.00011346703409159493,
"loss": 0.6787,
"step": 233
},
{
"epoch": 1.5412541254125411,
"grad_norm": 2.723644495010376,
"learning_rate": 0.00011270178197468789,
"loss": 0.937,
"step": 234
},
{
"epoch": 1.5478547854785478,
"grad_norm": 2.718723773956299,
"learning_rate": 0.00011193577307119687,
"loss": 0.6521,
"step": 235
},
{
"epoch": 1.5544554455445545,
"grad_norm": 2.100010633468628,
"learning_rate": 0.00011116905302080162,
"loss": 0.6487,
"step": 236
},
{
"epoch": 1.561056105610561,
"grad_norm": 1.9502499103546143,
"learning_rate": 0.00011040166750555287,
"loss": 0.7394,
"step": 237
},
{
"epoch": 1.5676567656765676,
"grad_norm": 2.46386456489563,
"learning_rate": 0.00010963366224715035,
"loss": 1.0024,
"step": 238
},
{
"epoch": 1.5742574257425743,
"grad_norm": 1.982226848602295,
"learning_rate": 0.00010886508300421892,
"loss": 0.6429,
"step": 239
},
{
"epoch": 1.5808580858085808,
"grad_norm": 2.1574912071228027,
"learning_rate": 0.00010809597556958208,
"loss": 0.7553,
"step": 240
},
{
"epoch": 1.5874587458745875,
"grad_norm": 2.2293241024017334,
"learning_rate": 0.00010732638576753355,
"loss": 0.7684,
"step": 241
},
{
"epoch": 1.5940594059405941,
"grad_norm": 2.16422438621521,
"learning_rate": 0.00010655635945110705,
"loss": 0.7109,
"step": 242
},
{
"epoch": 1.6006600660066006,
"grad_norm": 1.6767030954360962,
"learning_rate": 0.00010578594249934432,
"loss": 0.5425,
"step": 243
},
{
"epoch": 1.6072607260726073,
"grad_norm": 1.9446439743041992,
"learning_rate": 0.00010501518081456163,
"loss": 0.659,
"step": 244
},
{
"epoch": 1.613861386138614,
"grad_norm": 1.7710820436477661,
"learning_rate": 0.00010424412031961484,
"loss": 0.549,
"step": 245
},
{
"epoch": 1.6204620462046204,
"grad_norm": 1.881514072418213,
"learning_rate": 0.00010347280695516319,
"loss": 0.6608,
"step": 246
},
{
"epoch": 1.627062706270627,
"grad_norm": 2.171924352645874,
"learning_rate": 0.00010270128667693224,
"loss": 0.8295,
"step": 247
},
{
"epoch": 1.6336633663366338,
"grad_norm": 1.8770331144332886,
"learning_rate": 0.00010192960545297567,
"loss": 0.5008,
"step": 248
},
{
"epoch": 1.6402640264026402,
"grad_norm": 2.109541416168213,
"learning_rate": 0.0001011578092609365,
"loss": 0.5212,
"step": 249
},
{
"epoch": 1.6468646864686467,
"grad_norm": 2.7761456966400146,
"learning_rate": 0.00010038594408530769,
"loss": 1.0942,
"step": 250
},
{
"epoch": 1.6534653465346536,
"grad_norm": 2.638009786605835,
"learning_rate": 9.961405591469231e-05,
"loss": 1.0,
"step": 251
},
{
"epoch": 1.66006600660066,
"grad_norm": 1.703428030014038,
"learning_rate": 9.884219073906353e-05,
"loss": 0.4952,
"step": 252
},
{
"epoch": 1.6666666666666665,
"grad_norm": 2.3541927337646484,
"learning_rate": 9.807039454702435e-05,
"loss": 0.83,
"step": 253
},
{
"epoch": 1.6732673267326734,
"grad_norm": 1.742366075515747,
"learning_rate": 9.729871332306776e-05,
"loss": 0.7274,
"step": 254
},
{
"epoch": 1.6798679867986799,
"grad_norm": 2.8046398162841797,
"learning_rate": 9.652719304483682e-05,
"loss": 1.246,
"step": 255
},
{
"epoch": 1.6864686468646863,
"grad_norm": 2.256788730621338,
"learning_rate": 9.57558796803852e-05,
"loss": 0.8288,
"step": 256
},
{
"epoch": 1.693069306930693,
"grad_norm": 1.8939522504806519,
"learning_rate": 9.498481918543835e-05,
"loss": 0.6782,
"step": 257
},
{
"epoch": 1.6996699669966997,
"grad_norm": 2.374213457107544,
"learning_rate": 9.42140575006557e-05,
"loss": 0.8925,
"step": 258
},
{
"epoch": 1.7062706270627062,
"grad_norm": 2.4217422008514404,
"learning_rate": 9.344364054889299e-05,
"loss": 0.5865,
"step": 259
},
{
"epoch": 1.7128712871287128,
"grad_norm": 2.4119760990142822,
"learning_rate": 9.267361423246644e-05,
"loss": 0.8292,
"step": 260
},
{
"epoch": 1.7194719471947195,
"grad_norm": 2.435215473175049,
"learning_rate": 9.190402443041793e-05,
"loss": 0.911,
"step": 261
},
{
"epoch": 1.726072607260726,
"grad_norm": 1.7748898267745972,
"learning_rate": 9.113491699578109e-05,
"loss": 0.5359,
"step": 262
},
{
"epoch": 1.7326732673267327,
"grad_norm": 2.5013463497161865,
"learning_rate": 9.036633775284968e-05,
"loss": 0.9266,
"step": 263
},
{
"epoch": 1.7392739273927393,
"grad_norm": 2.6795082092285156,
"learning_rate": 8.959833249444714e-05,
"loss": 1.1157,
"step": 264
},
{
"epoch": 1.7458745874587458,
"grad_norm": 2.3808956146240234,
"learning_rate": 8.883094697919839e-05,
"loss": 0.7673,
"step": 265
},
{
"epoch": 1.7524752475247525,
"grad_norm": 2.104304075241089,
"learning_rate": 8.806422692880317e-05,
"loss": 1.0859,
"step": 266
},
{
"epoch": 1.7590759075907592,
"grad_norm": 2.0404794216156006,
"learning_rate": 8.729821802531212e-05,
"loss": 0.7336,
"step": 267
},
{
"epoch": 1.7656765676567656,
"grad_norm": 2.1330618858337402,
"learning_rate": 8.653296590840509e-05,
"loss": 0.791,
"step": 268
},
{
"epoch": 1.7722772277227723,
"grad_norm": 1.5914537906646729,
"learning_rate": 8.57685161726715e-05,
"loss": 0.4834,
"step": 269
},
{
"epoch": 1.778877887788779,
"grad_norm": 1.9733717441558838,
"learning_rate": 8.500491436489413e-05,
"loss": 0.8296,
"step": 270
},
{
"epoch": 1.7854785478547854,
"grad_norm": 1.7113840579986572,
"learning_rate": 8.424220598133527e-05,
"loss": 0.6665,
"step": 271
},
{
"epoch": 1.7920792079207921,
"grad_norm": 2.625967264175415,
"learning_rate": 8.348043646502589e-05,
"loss": 0.6638,
"step": 272
},
{
"epoch": 1.7986798679867988,
"grad_norm": 2.0700736045837402,
"learning_rate": 8.271965120305814e-05,
"loss": 1.0193,
"step": 273
},
{
"epoch": 1.8052805280528053,
"grad_norm": 1.6580454111099243,
"learning_rate": 8.195989552388139e-05,
"loss": 0.5835,
"step": 274
},
{
"epoch": 1.811881188118812,
"grad_norm": 2.130967378616333,
"learning_rate": 8.120121469460115e-05,
"loss": 0.738,
"step": 275
},
{
"epoch": 1.8184818481848186,
"grad_norm": 2.2612602710723877,
"learning_rate": 8.04436539182821e-05,
"loss": 0.8632,
"step": 276
},
{
"epoch": 1.825082508250825,
"grad_norm": 1.2444508075714111,
"learning_rate": 7.968725833125505e-05,
"loss": 0.4055,
"step": 277
},
{
"epoch": 1.8316831683168315,
"grad_norm": 1.5637201070785522,
"learning_rate": 7.89320730004274e-05,
"loss": 0.6677,
"step": 278
},
{
"epoch": 1.8382838283828384,
"grad_norm": 1.9795053005218506,
"learning_rate": 7.817814292059808e-05,
"loss": 0.7498,
"step": 279
},
{
"epoch": 1.844884488448845,
"grad_norm": 2.1388440132141113,
"learning_rate": 7.742551301177688e-05,
"loss": 0.6413,
"step": 280
},
{
"epoch": 1.8514851485148514,
"grad_norm": 2.597912311553955,
"learning_rate": 7.667422811650785e-05,
"loss": 0.9043,
"step": 281
},
{
"epoch": 1.858085808580858,
"grad_norm": 1.9656403064727783,
"learning_rate": 7.592433299719757e-05,
"loss": 0.638,
"step": 282
},
{
"epoch": 1.8646864686468647,
"grad_norm": 2.2590365409851074,
"learning_rate": 7.517587233344819e-05,
"loss": 0.6159,
"step": 283
},
{
"epoch": 1.8712871287128712,
"grad_norm": 1.8697465658187866,
"learning_rate": 7.442889071939547e-05,
"loss": 0.7238,
"step": 284
},
{
"epoch": 1.8778877887788779,
"grad_norm": 1.8442035913467407,
"learning_rate": 7.368343266105162e-05,
"loss": 0.6116,
"step": 285
},
{
"epoch": 1.8844884488448845,
"grad_norm": 2.237382650375366,
"learning_rate": 7.293954257365368e-05,
"loss": 0.8672,
"step": 286
},
{
"epoch": 1.891089108910891,
"grad_norm": 2.0788445472717285,
"learning_rate": 7.219726477901727e-05,
"loss": 1.2157,
"step": 287
},
{
"epoch": 1.8976897689768977,
"grad_norm": 2.1315560340881348,
"learning_rate": 7.145664350289567e-05,
"loss": 0.6643,
"step": 288
},
{
"epoch": 1.9042904290429044,
"grad_norm": 2.7273383140563965,
"learning_rate": 7.071772287234497e-05,
"loss": 0.9015,
"step": 289
},
{
"epoch": 1.9108910891089108,
"grad_norm": 2.51533842086792,
"learning_rate": 6.99805469130949e-05,
"loss": 0.7459,
"step": 290
},
{
"epoch": 1.9174917491749175,
"grad_norm": 1.1954275369644165,
"learning_rate": 6.924515954692563e-05,
"loss": 0.352,
"step": 291
},
{
"epoch": 1.9240924092409242,
"grad_norm": 2.0475802421569824,
"learning_rate": 6.851160458905092e-05,
"loss": 0.6949,
"step": 292
},
{
"epoch": 1.9306930693069306,
"grad_norm": 1.8832380771636963,
"learning_rate": 6.777992574550767e-05,
"loss": 0.6011,
"step": 293
},
{
"epoch": 1.9372937293729373,
"grad_norm": 1.9153025150299072,
"learning_rate": 6.705016661055161e-05,
"loss": 0.8822,
"step": 294
},
{
"epoch": 1.943894389438944,
"grad_norm": 2.1158411502838135,
"learning_rate": 6.632237066406013e-05,
"loss": 0.5226,
"step": 295
},
{
"epoch": 1.9504950495049505,
"grad_norm": 2.0630874633789062,
"learning_rate": 6.559658126894168e-05,
"loss": 0.6912,
"step": 296
},
{
"epoch": 1.9570957095709571,
"grad_norm": 1.8396505117416382,
"learning_rate": 6.487284166855202e-05,
"loss": 0.6315,
"step": 297
},
{
"epoch": 1.9636963696369638,
"grad_norm": 1.6715048551559448,
"learning_rate": 6.415119498411776e-05,
"loss": 0.5281,
"step": 298
},
{
"epoch": 1.9702970297029703,
"grad_norm": 2.4328067302703857,
"learning_rate": 6.34316842121674e-05,
"loss": 1.0786,
"step": 299
},
{
"epoch": 1.976897689768977,
"grad_norm": 1.847193956375122,
"learning_rate": 6.271435222196916e-05,
"loss": 0.6713,
"step": 300
},
{
"epoch": 1.9834983498349836,
"grad_norm": 1.8145736455917358,
"learning_rate": 6.1999241752977e-05,
"loss": 0.6318,
"step": 301
},
{
"epoch": 1.99009900990099,
"grad_norm": 1.8386404514312744,
"learning_rate": 6.128639541228426e-05,
"loss": 0.8397,
"step": 302
},
{
"epoch": 1.9966996699669965,
"grad_norm": 1.5646839141845703,
"learning_rate": 6.057585567208484e-05,
"loss": 0.4659,
"step": 303
},
{
"epoch": 2.0,
"grad_norm": 5.237025260925293,
"learning_rate": 5.986766486714268e-05,
"loss": 1.2294,
"step": 304
}
],
"logging_steps": 1,
"max_steps": 453,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"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": 3.574914020187644e+16,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}