Files
mlfoundations-dev_code-stra…/trainer_state.json

1135 lines
27 KiB
JSON
Raw Permalink Normal View History

{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 156,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.019230769230769232,
"grad_norm": 6.756826400756836,
"learning_rate": 6.25e-07,
"loss": 1.0467,
"step": 1
},
{
"epoch": 0.038461538461538464,
"grad_norm": 6.748652935028076,
"learning_rate": 1.25e-06,
"loss": 1.035,
"step": 2
},
{
"epoch": 0.057692307692307696,
"grad_norm": 6.735743999481201,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.0379,
"step": 3
},
{
"epoch": 0.07692307692307693,
"grad_norm": 6.658365726470947,
"learning_rate": 2.5e-06,
"loss": 1.0497,
"step": 4
},
{
"epoch": 0.09615384615384616,
"grad_norm": 6.092031955718994,
"learning_rate": 3.125e-06,
"loss": 1.0217,
"step": 5
},
{
"epoch": 0.11538461538461539,
"grad_norm": 4.476426124572754,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.9747,
"step": 6
},
{
"epoch": 0.1346153846153846,
"grad_norm": 2.8898186683654785,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.9507,
"step": 7
},
{
"epoch": 0.15384615384615385,
"grad_norm": 2.6818649768829346,
"learning_rate": 5e-06,
"loss": 0.9473,
"step": 8
},
{
"epoch": 0.17307692307692307,
"grad_norm": 4.384618759155273,
"learning_rate": 5.625e-06,
"loss": 0.9369,
"step": 9
},
{
"epoch": 0.19230769230769232,
"grad_norm": 4.5409979820251465,
"learning_rate": 6.25e-06,
"loss": 0.9425,
"step": 10
},
{
"epoch": 0.21153846153846154,
"grad_norm": 4.295501232147217,
"learning_rate": 6.875e-06,
"loss": 0.9044,
"step": 11
},
{
"epoch": 0.23076923076923078,
"grad_norm": 3.13700795173645,
"learning_rate": 7.500000000000001e-06,
"loss": 0.8622,
"step": 12
},
{
"epoch": 0.25,
"grad_norm": 2.9523777961730957,
"learning_rate": 8.125000000000001e-06,
"loss": 0.858,
"step": 13
},
{
"epoch": 0.2692307692307692,
"grad_norm": 2.254493474960327,
"learning_rate": 8.750000000000001e-06,
"loss": 0.8479,
"step": 14
},
{
"epoch": 0.28846153846153844,
"grad_norm": 1.5103687047958374,
"learning_rate": 9.375000000000001e-06,
"loss": 0.8085,
"step": 15
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.515587329864502,
"learning_rate": 1e-05,
"loss": 0.8253,
"step": 16
},
{
"epoch": 0.3269230769230769,
"grad_norm": 1.5592089891433716,
"learning_rate": 9.998741174712534e-06,
"loss": 0.7737,
"step": 17
},
{
"epoch": 0.34615384615384615,
"grad_norm": 1.36715567111969,
"learning_rate": 9.994965332706574e-06,
"loss": 0.7872,
"step": 18
},
{
"epoch": 0.36538461538461536,
"grad_norm": 1.0765823125839233,
"learning_rate": 9.98867437523228e-06,
"loss": 0.7565,
"step": 19
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.1925386190414429,
"learning_rate": 9.979871469976197e-06,
"loss": 0.7616,
"step": 20
},
{
"epoch": 0.40384615384615385,
"grad_norm": 1.0839797258377075,
"learning_rate": 9.968561049466214e-06,
"loss": 0.7438,
"step": 21
},
{
"epoch": 0.4230769230769231,
"grad_norm": 0.8592272996902466,
"learning_rate": 9.954748808839675e-06,
"loss": 0.747,
"step": 22
},
{
"epoch": 0.4423076923076923,
"grad_norm": 0.8572734594345093,
"learning_rate": 9.938441702975689e-06,
"loss": 0.7425,
"step": 23
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.0244359970092773,
"learning_rate": 9.91964794299315e-06,
"loss": 0.7223,
"step": 24
},
{
"epoch": 0.4807692307692308,
"grad_norm": 0.9166272878646851,
"learning_rate": 9.898376992116179e-06,
"loss": 0.7254,
"step": 25
},
{
"epoch": 0.5,
"grad_norm": 0.8023123741149902,
"learning_rate": 9.874639560909118e-06,
"loss": 0.705,
"step": 26
},
{
"epoch": 0.5192307692307693,
"grad_norm": 0.7520763874053955,
"learning_rate": 9.848447601883436e-06,
"loss": 0.7097,
"step": 27
},
{
"epoch": 0.5384615384615384,
"grad_norm": 0.7471436262130737,
"learning_rate": 9.819814303479268e-06,
"loss": 0.7087,
"step": 28
},
{
"epoch": 0.5576923076923077,
"grad_norm": 0.8560906648635864,
"learning_rate": 9.788754083424654e-06,
"loss": 0.7106,
"step": 29
},
{
"epoch": 0.5769230769230769,
"grad_norm": 0.7833404541015625,
"learning_rate": 9.755282581475769e-06,
"loss": 0.7004,
"step": 30
},
{
"epoch": 0.5961538461538461,
"grad_norm": 0.6619976758956909,
"learning_rate": 9.719416651541839e-06,
"loss": 0.7029,
"step": 31
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.7908540368080139,
"learning_rate": 9.681174353198687e-06,
"loss": 0.702,
"step": 32
},
{
"epoch": 0.6346153846153846,
"grad_norm": 0.7704586386680603,
"learning_rate": 9.640574942595195e-06,
"loss": 0.7126,
"step": 33
},
{
"epoch": 0.6538461538461539,
"grad_norm": 0.6045657992362976,
"learning_rate": 9.597638862757255e-06,
"loss": 0.6879,
"step": 34
},
{
"epoch": 0.6730769230769231,
"grad_norm": 0.6769744157791138,
"learning_rate": 9.552387733294081e-06,
"loss": 0.7112,
"step": 35
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.70807945728302,
"learning_rate": 9.504844339512096e-06,
"loss": 0.6787,
"step": 36
},
{
"epoch": 0.7115384615384616,
"grad_norm": 0.5988538861274719,
"learning_rate": 9.45503262094184e-06,
"loss": 0.6876,
"step": 37
},
{
"epoch": 0.7307692307692307,
"grad_norm": 0.6042438745498657,
"learning_rate": 9.40297765928369e-06,
"loss": 0.6943,
"step": 38
},
{
"epoch": 0.75,
"grad_norm": 0.6781715154647827,
"learning_rate": 9.348705665778479e-06,
"loss": 0.6842,
"step": 39
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.6857075691223145,
"learning_rate": 9.292243968009332e-06,
"loss": 0.6676,
"step": 40
},
{
"epoch": 0.7884615384615384,
"grad_norm": 0.5267153978347778,
"learning_rate": 9.233620996141421e-06,
"loss": 0.6693,
"step": 41
},
{
"epoch": 0.8076923076923077,
"grad_norm": 0.5316743850708008,
"learning_rate": 9.172866268606514e-06,
"loss": 0.6685,
"step": 42
},
{
"epoch": 0.8269230769230769,
"grad_norm": 0.7179816961288452,
"learning_rate": 9.110010377239552e-06,
"loss": 0.6766,
"step": 43
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.6035124659538269,
"learning_rate": 9.045084971874738e-06,
"loss": 0.6821,
"step": 44
},
{
"epoch": 0.8653846153846154,
"grad_norm": 0.7270795702934265,
"learning_rate": 8.978122744408905e-06,
"loss": 0.6683,
"step": 45
},
{
"epoch": 0.8846153846153846,
"grad_norm": 0.5073479413986206,
"learning_rate": 8.90915741234015e-06,
"loss": 0.6788,
"step": 46
},
{
"epoch": 0.9038461538461539,
"grad_norm": 0.5434288382530212,
"learning_rate": 8.838223701790057e-06,
"loss": 0.66,
"step": 47
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.6094857454299927,
"learning_rate": 8.765357330018056e-06,
"loss": 0.6581,
"step": 48
},
{
"epoch": 0.9423076923076923,
"grad_norm": 0.46337971091270447,
"learning_rate": 8.690594987436705e-06,
"loss": 0.6627,
"step": 49
},
{
"epoch": 0.9615384615384616,
"grad_norm": 0.6698839068412781,
"learning_rate": 8.613974319136959e-06,
"loss": 0.6598,
"step": 50
},
{
"epoch": 0.9807692307692307,
"grad_norm": 0.4814399480819702,
"learning_rate": 8.535533905932739e-06,
"loss": 0.665,
"step": 51
},
{
"epoch": 1.0,
"grad_norm": 0.5604783296585083,
"learning_rate": 8.455313244934324e-06,
"loss": 0.6737,
"step": 52
},
{
"epoch": 1.0192307692307692,
"grad_norm": 0.5326887369155884,
"learning_rate": 8.373352729660373e-06,
"loss": 0.6544,
"step": 53
},
{
"epoch": 1.0384615384615385,
"grad_norm": 0.518214762210846,
"learning_rate": 8.289693629698564e-06,
"loss": 0.631,
"step": 54
},
{
"epoch": 1.0576923076923077,
"grad_norm": 0.539520263671875,
"learning_rate": 8.204378069925121e-06,
"loss": 0.6275,
"step": 55
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.5140302181243896,
"learning_rate": 8.117449009293668e-06,
"loss": 0.6379,
"step": 56
},
{
"epoch": 1.0961538461538463,
"grad_norm": 0.7126110792160034,
"learning_rate": 8.0289502192041e-06,
"loss": 0.6209,
"step": 57
},
{
"epoch": 1.1153846153846154,
"grad_norm": 0.5747066140174866,
"learning_rate": 7.938926261462366e-06,
"loss": 0.6324,
"step": 58
},
{
"epoch": 1.1346153846153846,
"grad_norm": 0.6441423296928406,
"learning_rate": 7.84742246584226e-06,
"loss": 0.6463,
"step": 59
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.5559582710266113,
"learning_rate": 7.754484907260513e-06,
"loss": 0.6315,
"step": 60
},
{
"epoch": 1.1730769230769231,
"grad_norm": 0.5032030344009399,
"learning_rate": 7.660160382576683e-06,
"loss": 0.641,
"step": 61
},
{
"epoch": 1.1923076923076923,
"grad_norm": 0.5290240049362183,
"learning_rate": 7.564496387029532e-06,
"loss": 0.6246,
"step": 62
},
{
"epoch": 1.2115384615384615,
"grad_norm": 0.47469562292099,
"learning_rate": 7.467541090321735e-06,
"loss": 0.6314,
"step": 63
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.6516867876052856,
"learning_rate": 7.369343312364994e-06,
"loss": 0.6186,
"step": 64
},
{
"epoch": 1.25,
"grad_norm": 0.4732189476490021,
"learning_rate": 7.269952498697734e-06,
"loss": 0.6357,
"step": 65
},
{
"epoch": 1.2692307692307692,
"grad_norm": 0.5501453280448914,
"learning_rate": 7.169418695587791e-06,
"loss": 0.6415,
"step": 66
},
{
"epoch": 1.2884615384615383,
"grad_norm": 0.49621427059173584,
"learning_rate": 7.067792524832604e-06,
"loss": 0.6248,
"step": 67
},
{
"epoch": 1.3076923076923077,
"grad_norm": 0.4925438463687897,
"learning_rate": 6.965125158269619e-06,
"loss": 0.6218,
"step": 68
},
{
"epoch": 1.3269230769230769,
"grad_norm": 0.5866847038269043,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.629,
"step": 69
},
{
"epoch": 1.3461538461538463,
"grad_norm": 0.5412906408309937,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.6283,
"step": 70
},
{
"epoch": 1.3653846153846154,
"grad_norm": 0.5679116249084473,
"learning_rate": 6.651395309775837e-06,
"loss": 0.6251,
"step": 71
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.46475809812545776,
"learning_rate": 6.545084971874738e-06,
"loss": 0.6194,
"step": 72
},
{
"epoch": 1.4038461538461537,
"grad_norm": 0.6032177805900574,
"learning_rate": 6.437996637160086e-06,
"loss": 0.624,
"step": 73
},
{
"epoch": 1.4230769230769231,
"grad_norm": 0.4972044825553894,
"learning_rate": 6.330184227833376e-06,
"loss": 0.6245,
"step": 74
},
{
"epoch": 1.4423076923076923,
"grad_norm": 0.49976637959480286,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.6153,
"step": 75
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.5510290861129761,
"learning_rate": 6.112604669781572e-06,
"loss": 0.6387,
"step": 76
},
{
"epoch": 1.4807692307692308,
"grad_norm": 0.5240339636802673,
"learning_rate": 6.002947078916365e-06,
"loss": 0.6281,
"step": 77
},
{
"epoch": 1.5,
"grad_norm": 0.5085740089416504,
"learning_rate": 5.892784473993184e-06,
"loss": 0.6193,
"step": 78
},
{
"epoch": 1.5192307692307692,
"grad_norm": 0.47661781311035156,
"learning_rate": 5.782172325201155e-06,
"loss": 0.6061,
"step": 79
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.4540369510650635,
"learning_rate": 5.671166329088278e-06,
"loss": 0.603,
"step": 80
},
{
"epoch": 1.5576923076923077,
"grad_norm": 0.48901426792144775,
"learning_rate": 5.559822380516539e-06,
"loss": 0.6384,
"step": 81
},
{
"epoch": 1.5769230769230769,
"grad_norm": 0.49942049384117126,
"learning_rate": 5.448196544517168e-06,
"loss": 0.6201,
"step": 82
},
{
"epoch": 1.5961538461538463,
"grad_norm": 0.4124830365180969,
"learning_rate": 5.336345028060199e-06,
"loss": 0.6219,
"step": 83
},
{
"epoch": 1.6153846153846154,
"grad_norm": 0.44156306982040405,
"learning_rate": 5.224324151752575e-06,
"loss": 0.619,
"step": 84
},
{
"epoch": 1.6346153846153846,
"grad_norm": 0.4886167049407959,
"learning_rate": 5.112190321479026e-06,
"loss": 0.6353,
"step": 85
},
{
"epoch": 1.6538461538461537,
"grad_norm": 0.4457215666770935,
"learning_rate": 5e-06,
"loss": 0.6309,
"step": 86
},
{
"epoch": 1.6730769230769231,
"grad_norm": 0.37990936636924744,
"learning_rate": 4.887809678520976e-06,
"loss": 0.6306,
"step": 87
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.48935168981552124,
"learning_rate": 4.775675848247427e-06,
"loss": 0.6247,
"step": 88
},
{
"epoch": 1.7115384615384617,
"grad_norm": 0.43845734000205994,
"learning_rate": 4.663654971939802e-06,
"loss": 0.6243,
"step": 89
},
{
"epoch": 1.7307692307692308,
"grad_norm": 0.38639092445373535,
"learning_rate": 4.551803455482833e-06,
"loss": 0.6181,
"step": 90
},
{
"epoch": 1.75,
"grad_norm": 0.4700919985771179,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.6434,
"step": 91
},
{
"epoch": 1.7692307692307692,
"grad_norm": 0.4569476544857025,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.6248,
"step": 92
},
{
"epoch": 1.7884615384615383,
"grad_norm": 0.42245760560035706,
"learning_rate": 4.217827674798845e-06,
"loss": 0.6305,
"step": 93
},
{
"epoch": 1.8076923076923077,
"grad_norm": 0.48794662952423096,
"learning_rate": 4.107215526006818e-06,
"loss": 0.6209,
"step": 94
},
{
"epoch": 1.8269230769230769,
"grad_norm": 0.5091656446456909,
"learning_rate": 3.997052921083637e-06,
"loss": 0.6148,
"step": 95
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.46186864376068115,
"learning_rate": 3.887395330218429e-06,
"loss": 0.6293,
"step": 96
},
{
"epoch": 1.8653846153846154,
"grad_norm": 0.409128338098526,
"learning_rate": 3.778297969310529e-06,
"loss": 0.6114,
"step": 97
},
{
"epoch": 1.8846153846153846,
"grad_norm": 0.44377630949020386,
"learning_rate": 3.669815772166625e-06,
"loss": 0.6167,
"step": 98
},
{
"epoch": 1.9038461538461537,
"grad_norm": 0.6437395811080933,
"learning_rate": 3.562003362839914e-06,
"loss": 0.6031,
"step": 99
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.44504550099372864,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.6085,
"step": 100
},
{
"epoch": 1.9423076923076923,
"grad_norm": 0.4889739453792572,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.6071,
"step": 101
},
{
"epoch": 1.9615384615384617,
"grad_norm": 0.6323836445808411,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.6268,
"step": 102
},
{
"epoch": 1.9807692307692308,
"grad_norm": 0.47047120332717896,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.6142,
"step": 103
},
{
"epoch": 2.0,
"grad_norm": 0.385359525680542,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.6343,
"step": 104
},
{
"epoch": 2.019230769230769,
"grad_norm": 0.4017738699913025,
"learning_rate": 2.932207475167398e-06,
"loss": 0.6071,
"step": 105
},
{
"epoch": 2.0384615384615383,
"grad_norm": 0.4930027723312378,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5926,
"step": 106
},
{
"epoch": 2.0576923076923075,
"grad_norm": 0.41167354583740234,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.6021,
"step": 107
},
{
"epoch": 2.076923076923077,
"grad_norm": 0.43824106454849243,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5946,
"step": 108
},
{
"epoch": 2.0961538461538463,
"grad_norm": 0.42565321922302246,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5968,
"step": 109
},
{
"epoch": 2.1153846153846154,
"grad_norm": 0.4275932013988495,
"learning_rate": 2.43550361297047e-06,
"loss": 0.586,
"step": 110
},
{
"epoch": 2.1346153846153846,
"grad_norm": 0.40257298946380615,
"learning_rate": 2.339839617423318e-06,
"loss": 0.588,
"step": 111
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.3709084987640381,
"learning_rate": 2.245515092739488e-06,
"loss": 0.5875,
"step": 112
},
{
"epoch": 2.173076923076923,
"grad_norm": 0.34968435764312744,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.5922,
"step": 113
},
{
"epoch": 2.1923076923076925,
"grad_norm": 0.3987502157688141,
"learning_rate": 2.061073738537635e-06,
"loss": 0.6045,
"step": 114
},
{
"epoch": 2.2115384615384617,
"grad_norm": 0.3902794122695923,
"learning_rate": 1.971049780795901e-06,
"loss": 0.5879,
"step": 115
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.39212530851364136,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.5974,
"step": 116
},
{
"epoch": 2.25,
"grad_norm": 0.3752593994140625,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.5905,
"step": 117
},
{
"epoch": 2.269230769230769,
"grad_norm": 0.3571646213531494,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.5813,
"step": 118
},
{
"epoch": 2.2884615384615383,
"grad_norm": 0.3796848952770233,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.5798,
"step": 119
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.38161250948905945,
"learning_rate": 1.544686755065677e-06,
"loss": 0.5973,
"step": 120
},
{
"epoch": 2.326923076923077,
"grad_norm": 0.3628106415271759,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5915,
"step": 121
},
{
"epoch": 2.3461538461538463,
"grad_norm": 0.3747085928916931,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.6004,
"step": 122
},
{
"epoch": 2.3653846153846154,
"grad_norm": 0.3763367831707001,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.6029,
"step": 123
},
{
"epoch": 2.3846153846153846,
"grad_norm": 0.32238879799842834,
"learning_rate": 1.234642669981946e-06,
"loss": 0.6025,
"step": 124
},
{
"epoch": 2.4038461538461537,
"grad_norm": 0.37101826071739197,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.5737,
"step": 125
},
{
"epoch": 2.423076923076923,
"grad_norm": 0.34124448895454407,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5701,
"step": 126
},
{
"epoch": 2.4423076923076925,
"grad_norm": 0.34349343180656433,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.604,
"step": 127
},
{
"epoch": 2.4615384615384617,
"grad_norm": 0.36411768198013306,
"learning_rate": 9.549150281252633e-07,
"loss": 0.6136,
"step": 128
},
{
"epoch": 2.480769230769231,
"grad_norm": 0.3538264036178589,
"learning_rate": 8.899896227604509e-07,
"loss": 0.5856,
"step": 129
},
{
"epoch": 2.5,
"grad_norm": 0.35449934005737305,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5907,
"step": 130
},
{
"epoch": 2.519230769230769,
"grad_norm": 0.31669095158576965,
"learning_rate": 7.663790038585794e-07,
"loss": 0.5904,
"step": 131
},
{
"epoch": 2.5384615384615383,
"grad_norm": 0.3434077501296997,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5775,
"step": 132
},
{
"epoch": 2.5576923076923075,
"grad_norm": 0.3626374900341034,
"learning_rate": 6.512943342215234e-07,
"loss": 0.5878,
"step": 133
},
{
"epoch": 2.5769230769230766,
"grad_norm": 0.3235149681568146,
"learning_rate": 5.9702234071631e-07,
"loss": 0.601,
"step": 134
},
{
"epoch": 2.5961538461538463,
"grad_norm": 0.3342594802379608,
"learning_rate": 5.449673790581611e-07,
"loss": 0.5926,
"step": 135
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.3404471278190613,
"learning_rate": 4.951556604879049e-07,
"loss": 0.5772,
"step": 136
},
{
"epoch": 2.6346153846153846,
"grad_norm": 0.3687044680118561,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5895,
"step": 137
},
{
"epoch": 2.6538461538461537,
"grad_norm": 0.3606056571006775,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5984,
"step": 138
},
{
"epoch": 2.6730769230769234,
"grad_norm": 0.3252229392528534,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5967,
"step": 139
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.3195320963859558,
"learning_rate": 3.18825646801314e-07,
"loss": 0.6131,
"step": 140
},
{
"epoch": 2.7115384615384617,
"grad_norm": 0.33670124411582947,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5952,
"step": 141
},
{
"epoch": 2.730769230769231,
"grad_norm": 0.3292597532272339,
"learning_rate": 2.447174185242324e-07,
"loss": 0.6009,
"step": 142
},
{
"epoch": 2.75,
"grad_norm": 0.31205758452415466,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.6009,
"step": 143
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.3304506540298462,
"learning_rate": 1.801856965207338e-07,
"loss": 0.5979,
"step": 144
},
{
"epoch": 2.7884615384615383,
"grad_norm": 0.33979734778404236,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5747,
"step": 145
},
{
"epoch": 2.8076923076923075,
"grad_norm": 0.37355899810791016,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5994,
"step": 146
},
{
"epoch": 2.8269230769230766,
"grad_norm": 0.3773685693740845,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.5913,
"step": 147
},
{
"epoch": 2.8461538461538463,
"grad_norm": 0.31100109219551086,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5829,
"step": 148
},
{
"epoch": 2.8653846153846154,
"grad_norm": 0.31232011318206787,
"learning_rate": 6.15582970243117e-08,
"loss": 0.5927,
"step": 149
},
{
"epoch": 2.8846153846153846,
"grad_norm": 0.3333931267261505,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5781,
"step": 150
},
{
"epoch": 2.9038461538461537,
"grad_norm": 0.31353145837783813,
"learning_rate": 3.143895053378698e-08,
"loss": 0.6044,
"step": 151
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.33146995306015015,
"learning_rate": 2.012853002380466e-08,
"loss": 0.5993,
"step": 152
},
{
"epoch": 2.9423076923076925,
"grad_norm": 0.31752604246139526,
"learning_rate": 1.132562476771959e-08,
"loss": 0.5933,
"step": 153
},
{
"epoch": 2.9615384615384617,
"grad_norm": 0.32748833298683167,
"learning_rate": 5.034667293427053e-09,
"loss": 0.5813,
"step": 154
},
{
"epoch": 2.980769230769231,
"grad_norm": 0.3221939206123352,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5851,
"step": 155
},
{
"epoch": 3.0,
"grad_norm": 0.34672605991363525,
"learning_rate": 0.0,
"loss": 0.5678,
"step": 156
},
{
"epoch": 3.0,
"step": 156,
"total_flos": 2.3644419613786112e+17,
"train_loss": 0.6642100642888974,
"train_runtime": 3868.2988,
"train_samples_per_second": 3.859,
"train_steps_per_second": 0.04
}
],
"logging_steps": 1,
"max_steps": 156,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.3644419613786112e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}