EtashGuha's picture
Upload folder using huggingface_hub
23ea381 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.980891719745223,
"eval_steps": 500,
"global_step": 156,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01910828025477707,
"grad_norm": 7.283066272735596,
"learning_rate": 6.25e-07,
"loss": 1.1498,
"step": 1
},
{
"epoch": 0.03821656050955414,
"grad_norm": 7.387139797210693,
"learning_rate": 1.25e-06,
"loss": 1.1469,
"step": 2
},
{
"epoch": 0.05732484076433121,
"grad_norm": 7.194793224334717,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.1608,
"step": 3
},
{
"epoch": 0.07643312101910828,
"grad_norm": 7.103536128997803,
"learning_rate": 2.5e-06,
"loss": 1.1289,
"step": 4
},
{
"epoch": 0.09554140127388536,
"grad_norm": 6.497398853302002,
"learning_rate": 3.125e-06,
"loss": 1.1272,
"step": 5
},
{
"epoch": 0.11464968152866242,
"grad_norm": 4.629354000091553,
"learning_rate": 3.7500000000000005e-06,
"loss": 1.0227,
"step": 6
},
{
"epoch": 0.1337579617834395,
"grad_norm": 2.5543441772460938,
"learning_rate": 4.3750000000000005e-06,
"loss": 1.0106,
"step": 7
},
{
"epoch": 0.15286624203821655,
"grad_norm": 2.3123137950897217,
"learning_rate": 5e-06,
"loss": 0.9752,
"step": 8
},
{
"epoch": 0.17197452229299362,
"grad_norm": 4.898059368133545,
"learning_rate": 5.625e-06,
"loss": 0.9848,
"step": 9
},
{
"epoch": 0.1910828025477707,
"grad_norm": 5.3169264793396,
"learning_rate": 6.25e-06,
"loss": 0.9968,
"step": 10
},
{
"epoch": 0.21019108280254778,
"grad_norm": 5.212041854858398,
"learning_rate": 6.875e-06,
"loss": 0.9595,
"step": 11
},
{
"epoch": 0.22929936305732485,
"grad_norm": 3.953948974609375,
"learning_rate": 7.500000000000001e-06,
"loss": 0.9641,
"step": 12
},
{
"epoch": 0.2484076433121019,
"grad_norm": 3.704425096511841,
"learning_rate": 8.125000000000001e-06,
"loss": 0.8944,
"step": 13
},
{
"epoch": 0.267515923566879,
"grad_norm": 3.1059200763702393,
"learning_rate": 8.750000000000001e-06,
"loss": 0.8578,
"step": 14
},
{
"epoch": 0.28662420382165604,
"grad_norm": 2.5509462356567383,
"learning_rate": 9.375000000000001e-06,
"loss": 0.8804,
"step": 15
},
{
"epoch": 0.3057324840764331,
"grad_norm": 2.304457426071167,
"learning_rate": 1e-05,
"loss": 0.8432,
"step": 16
},
{
"epoch": 0.3248407643312102,
"grad_norm": 2.506471633911133,
"learning_rate": 9.998741174712534e-06,
"loss": 0.8422,
"step": 17
},
{
"epoch": 0.34394904458598724,
"grad_norm": 2.3640732765197754,
"learning_rate": 9.994965332706574e-06,
"loss": 0.8418,
"step": 18
},
{
"epoch": 0.3630573248407643,
"grad_norm": 2.196169853210449,
"learning_rate": 9.98867437523228e-06,
"loss": 0.8514,
"step": 19
},
{
"epoch": 0.3821656050955414,
"grad_norm": 1.7853914499282837,
"learning_rate": 9.979871469976197e-06,
"loss": 0.8462,
"step": 20
},
{
"epoch": 0.4012738853503185,
"grad_norm": 1.5007359981536865,
"learning_rate": 9.968561049466214e-06,
"loss": 0.8171,
"step": 21
},
{
"epoch": 0.42038216560509556,
"grad_norm": 1.2658205032348633,
"learning_rate": 9.954748808839675e-06,
"loss": 0.8074,
"step": 22
},
{
"epoch": 0.4394904458598726,
"grad_norm": 1.3059478998184204,
"learning_rate": 9.938441702975689e-06,
"loss": 0.8377,
"step": 23
},
{
"epoch": 0.4585987261146497,
"grad_norm": 1.0882668495178223,
"learning_rate": 9.91964794299315e-06,
"loss": 0.8195,
"step": 24
},
{
"epoch": 0.47770700636942676,
"grad_norm": 1.0192159414291382,
"learning_rate": 9.898376992116179e-06,
"loss": 0.8041,
"step": 25
},
{
"epoch": 0.4968152866242038,
"grad_norm": 0.9797454476356506,
"learning_rate": 9.874639560909118e-06,
"loss": 0.844,
"step": 26
},
{
"epoch": 0.5159235668789809,
"grad_norm": 0.9786008596420288,
"learning_rate": 9.848447601883436e-06,
"loss": 0.7693,
"step": 27
},
{
"epoch": 0.535031847133758,
"grad_norm": 0.9011598229408264,
"learning_rate": 9.819814303479268e-06,
"loss": 0.807,
"step": 28
},
{
"epoch": 0.554140127388535,
"grad_norm": 0.826195240020752,
"learning_rate": 9.788754083424654e-06,
"loss": 0.7848,
"step": 29
},
{
"epoch": 0.5732484076433121,
"grad_norm": 0.9426838755607605,
"learning_rate": 9.755282581475769e-06,
"loss": 0.8061,
"step": 30
},
{
"epoch": 0.5923566878980892,
"grad_norm": 0.9041593074798584,
"learning_rate": 9.719416651541839e-06,
"loss": 0.7731,
"step": 31
},
{
"epoch": 0.6114649681528662,
"grad_norm": 0.9566060304641724,
"learning_rate": 9.681174353198687e-06,
"loss": 0.7955,
"step": 32
},
{
"epoch": 0.6305732484076433,
"grad_norm": 0.763576328754425,
"learning_rate": 9.640574942595195e-06,
"loss": 0.7941,
"step": 33
},
{
"epoch": 0.6496815286624203,
"grad_norm": 1.0218584537506104,
"learning_rate": 9.597638862757255e-06,
"loss": 0.7971,
"step": 34
},
{
"epoch": 0.6687898089171974,
"grad_norm": 1.0102124214172363,
"learning_rate": 9.552387733294081e-06,
"loss": 0.738,
"step": 35
},
{
"epoch": 0.6878980891719745,
"grad_norm": 0.7672615051269531,
"learning_rate": 9.504844339512096e-06,
"loss": 0.7804,
"step": 36
},
{
"epoch": 0.7070063694267515,
"grad_norm": 1.0719943046569824,
"learning_rate": 9.45503262094184e-06,
"loss": 0.7807,
"step": 37
},
{
"epoch": 0.7261146496815286,
"grad_norm": 0.9412849545478821,
"learning_rate": 9.40297765928369e-06,
"loss": 0.8137,
"step": 38
},
{
"epoch": 0.7452229299363057,
"grad_norm": 0.8677616715431213,
"learning_rate": 9.348705665778479e-06,
"loss": 0.7743,
"step": 39
},
{
"epoch": 0.7643312101910829,
"grad_norm": 0.9233507513999939,
"learning_rate": 9.292243968009332e-06,
"loss": 0.7574,
"step": 40
},
{
"epoch": 0.7834394904458599,
"grad_norm": 1.9202914237976074,
"learning_rate": 9.233620996141421e-06,
"loss": 0.8454,
"step": 41
},
{
"epoch": 0.802547770700637,
"grad_norm": 0.8707855343818665,
"learning_rate": 9.172866268606514e-06,
"loss": 0.7502,
"step": 42
},
{
"epoch": 0.821656050955414,
"grad_norm": 0.8542220592498779,
"learning_rate": 9.110010377239552e-06,
"loss": 0.8205,
"step": 43
},
{
"epoch": 0.8407643312101911,
"grad_norm": 0.7179370522499084,
"learning_rate": 9.045084971874738e-06,
"loss": 0.7447,
"step": 44
},
{
"epoch": 0.8598726114649682,
"grad_norm": 0.7615282535552979,
"learning_rate": 8.978122744408905e-06,
"loss": 0.7994,
"step": 45
},
{
"epoch": 0.8789808917197452,
"grad_norm": 0.6987951397895813,
"learning_rate": 8.90915741234015e-06,
"loss": 0.7636,
"step": 46
},
{
"epoch": 0.8980891719745223,
"grad_norm": 0.8840264678001404,
"learning_rate": 8.838223701790057e-06,
"loss": 0.7604,
"step": 47
},
{
"epoch": 0.9171974522292994,
"grad_norm": 0.7975447773933411,
"learning_rate": 8.765357330018056e-06,
"loss": 0.7555,
"step": 48
},
{
"epoch": 0.9363057324840764,
"grad_norm": 0.834041953086853,
"learning_rate": 8.690594987436705e-06,
"loss": 0.7581,
"step": 49
},
{
"epoch": 0.9554140127388535,
"grad_norm": 0.6846311092376709,
"learning_rate": 8.613974319136959e-06,
"loss": 0.7254,
"step": 50
},
{
"epoch": 0.9745222929936306,
"grad_norm": 0.7656162977218628,
"learning_rate": 8.535533905932739e-06,
"loss": 0.7786,
"step": 51
},
{
"epoch": 0.9936305732484076,
"grad_norm": 0.6645858287811279,
"learning_rate": 8.455313244934324e-06,
"loss": 0.7752,
"step": 52
},
{
"epoch": 1.0127388535031847,
"grad_norm": 0.6466763019561768,
"learning_rate": 8.373352729660373e-06,
"loss": 0.697,
"step": 53
},
{
"epoch": 1.0318471337579618,
"grad_norm": 0.6180476546287537,
"learning_rate": 8.289693629698564e-06,
"loss": 0.6652,
"step": 54
},
{
"epoch": 1.0509554140127388,
"grad_norm": 0.6721435785293579,
"learning_rate": 8.204378069925121e-06,
"loss": 0.6916,
"step": 55
},
{
"epoch": 1.070063694267516,
"grad_norm": 0.7600250840187073,
"learning_rate": 8.117449009293668e-06,
"loss": 0.6793,
"step": 56
},
{
"epoch": 1.089171974522293,
"grad_norm": 0.6684707403182983,
"learning_rate": 8.0289502192041e-06,
"loss": 0.6743,
"step": 57
},
{
"epoch": 1.10828025477707,
"grad_norm": 0.7330480217933655,
"learning_rate": 7.938926261462366e-06,
"loss": 0.6754,
"step": 58
},
{
"epoch": 1.127388535031847,
"grad_norm": 0.6713646650314331,
"learning_rate": 7.84742246584226e-06,
"loss": 0.66,
"step": 59
},
{
"epoch": 1.1464968152866242,
"grad_norm": 0.6231556534767151,
"learning_rate": 7.754484907260513e-06,
"loss": 0.6564,
"step": 60
},
{
"epoch": 1.1656050955414012,
"grad_norm": 0.7284997701644897,
"learning_rate": 7.660160382576683e-06,
"loss": 0.6834,
"step": 61
},
{
"epoch": 1.1847133757961783,
"grad_norm": 0.6809297204017639,
"learning_rate": 7.564496387029532e-06,
"loss": 0.6453,
"step": 62
},
{
"epoch": 1.2038216560509554,
"grad_norm": 0.6243814826011658,
"learning_rate": 7.467541090321735e-06,
"loss": 0.6666,
"step": 63
},
{
"epoch": 1.2229299363057324,
"grad_norm": 0.6623536944389343,
"learning_rate": 7.369343312364994e-06,
"loss": 0.6858,
"step": 64
},
{
"epoch": 1.2420382165605095,
"grad_norm": 0.6827297806739807,
"learning_rate": 7.269952498697734e-06,
"loss": 0.6577,
"step": 65
},
{
"epoch": 1.2611464968152866,
"grad_norm": 0.6385012269020081,
"learning_rate": 7.169418695587791e-06,
"loss": 0.6833,
"step": 66
},
{
"epoch": 1.2802547770700636,
"grad_norm": 0.7074577212333679,
"learning_rate": 7.067792524832604e-06,
"loss": 0.6732,
"step": 67
},
{
"epoch": 1.2993630573248407,
"grad_norm": 0.6454325318336487,
"learning_rate": 6.965125158269619e-06,
"loss": 0.6671,
"step": 68
},
{
"epoch": 1.3184713375796178,
"grad_norm": 0.6871979832649231,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.6744,
"step": 69
},
{
"epoch": 1.3375796178343948,
"grad_norm": 0.6541727185249329,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.6529,
"step": 70
},
{
"epoch": 1.356687898089172,
"grad_norm": 0.6993772983551025,
"learning_rate": 6.651395309775837e-06,
"loss": 0.6548,
"step": 71
},
{
"epoch": 1.3757961783439492,
"grad_norm": 0.7177384495735168,
"learning_rate": 6.545084971874738e-06,
"loss": 0.6556,
"step": 72
},
{
"epoch": 1.394904458598726,
"grad_norm": 0.6087607741355896,
"learning_rate": 6.437996637160086e-06,
"loss": 0.6261,
"step": 73
},
{
"epoch": 1.4140127388535033,
"grad_norm": 0.6564837694168091,
"learning_rate": 6.330184227833376e-06,
"loss": 0.6632,
"step": 74
},
{
"epoch": 1.4331210191082802,
"grad_norm": 0.6872571110725403,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.6684,
"step": 75
},
{
"epoch": 1.4522292993630574,
"grad_norm": 0.6152129173278809,
"learning_rate": 6.112604669781572e-06,
"loss": 0.6491,
"step": 76
},
{
"epoch": 1.4713375796178343,
"grad_norm": 0.6309750080108643,
"learning_rate": 6.002947078916365e-06,
"loss": 0.6533,
"step": 77
},
{
"epoch": 1.4904458598726116,
"grad_norm": 0.6075384616851807,
"learning_rate": 5.892784473993184e-06,
"loss": 0.6631,
"step": 78
},
{
"epoch": 1.5095541401273884,
"grad_norm": 0.6240233778953552,
"learning_rate": 5.782172325201155e-06,
"loss": 0.6312,
"step": 79
},
{
"epoch": 1.5286624203821657,
"grad_norm": 0.6493943929672241,
"learning_rate": 5.671166329088278e-06,
"loss": 0.6583,
"step": 80
},
{
"epoch": 1.5477707006369426,
"grad_norm": 0.6722797155380249,
"learning_rate": 5.559822380516539e-06,
"loss": 0.6638,
"step": 81
},
{
"epoch": 1.5668789808917198,
"grad_norm": 0.6353743076324463,
"learning_rate": 5.448196544517168e-06,
"loss": 0.6516,
"step": 82
},
{
"epoch": 1.5859872611464967,
"grad_norm": 0.638333261013031,
"learning_rate": 5.336345028060199e-06,
"loss": 0.6592,
"step": 83
},
{
"epoch": 1.605095541401274,
"grad_norm": 0.7258989810943604,
"learning_rate": 5.224324151752575e-06,
"loss": 0.6816,
"step": 84
},
{
"epoch": 1.6242038216560508,
"grad_norm": 0.6479246616363525,
"learning_rate": 5.112190321479026e-06,
"loss": 0.6678,
"step": 85
},
{
"epoch": 1.643312101910828,
"grad_norm": 0.6187183856964111,
"learning_rate": 5e-06,
"loss": 0.649,
"step": 86
},
{
"epoch": 1.662420382165605,
"grad_norm": 0.6216777563095093,
"learning_rate": 4.887809678520976e-06,
"loss": 0.6587,
"step": 87
},
{
"epoch": 1.6815286624203822,
"grad_norm": 0.6864318251609802,
"learning_rate": 4.775675848247427e-06,
"loss": 0.6515,
"step": 88
},
{
"epoch": 1.700636942675159,
"grad_norm": 0.5659608244895935,
"learning_rate": 4.663654971939802e-06,
"loss": 0.6576,
"step": 89
},
{
"epoch": 1.7197452229299364,
"grad_norm": 0.601165235042572,
"learning_rate": 4.551803455482833e-06,
"loss": 0.6585,
"step": 90
},
{
"epoch": 1.7388535031847132,
"grad_norm": 0.6609579920768738,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.6659,
"step": 91
},
{
"epoch": 1.7579617834394905,
"grad_norm": 0.6505253314971924,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.6703,
"step": 92
},
{
"epoch": 1.7770700636942676,
"grad_norm": 0.6021715402603149,
"learning_rate": 4.217827674798845e-06,
"loss": 0.6813,
"step": 93
},
{
"epoch": 1.7961783439490446,
"grad_norm": 2.0026886463165283,
"learning_rate": 4.107215526006818e-06,
"loss": 0.6986,
"step": 94
},
{
"epoch": 1.8152866242038217,
"grad_norm": 0.583001434803009,
"learning_rate": 3.997052921083637e-06,
"loss": 0.6531,
"step": 95
},
{
"epoch": 1.8343949044585988,
"grad_norm": 0.6410379409790039,
"learning_rate": 3.887395330218429e-06,
"loss": 0.6655,
"step": 96
},
{
"epoch": 1.8535031847133758,
"grad_norm": 0.6351285576820374,
"learning_rate": 3.778297969310529e-06,
"loss": 0.6533,
"step": 97
},
{
"epoch": 1.872611464968153,
"grad_norm": 0.5866528749465942,
"learning_rate": 3.669815772166625e-06,
"loss": 0.6845,
"step": 98
},
{
"epoch": 1.89171974522293,
"grad_norm": 0.569778323173523,
"learning_rate": 3.562003362839914e-06,
"loss": 0.6494,
"step": 99
},
{
"epoch": 1.910828025477707,
"grad_norm": 0.554115891456604,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.6423,
"step": 100
},
{
"epoch": 1.929936305732484,
"grad_norm": 0.5609800815582275,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.6306,
"step": 101
},
{
"epoch": 1.9490445859872612,
"grad_norm": 0.5329970717430115,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.6657,
"step": 102
},
{
"epoch": 1.9681528662420382,
"grad_norm": 0.5660147666931152,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.6599,
"step": 103
},
{
"epoch": 1.9872611464968153,
"grad_norm": 0.5551929473876953,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.6617,
"step": 104
},
{
"epoch": 2.0063694267515926,
"grad_norm": 0.5826137661933899,
"learning_rate": 2.932207475167398e-06,
"loss": 0.6042,
"step": 105
},
{
"epoch": 2.0254777070063694,
"grad_norm": 0.5620737075805664,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5824,
"step": 106
},
{
"epoch": 2.0445859872611467,
"grad_norm": 0.6190159916877747,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.576,
"step": 107
},
{
"epoch": 2.0636942675159236,
"grad_norm": 0.5424576997756958,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5832,
"step": 108
},
{
"epoch": 2.082802547770701,
"grad_norm": 0.5414028167724609,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5767,
"step": 109
},
{
"epoch": 2.1019108280254777,
"grad_norm": 0.5423881411552429,
"learning_rate": 2.43550361297047e-06,
"loss": 0.5692,
"step": 110
},
{
"epoch": 2.121019108280255,
"grad_norm": 0.5424202680587769,
"learning_rate": 2.339839617423318e-06,
"loss": 0.5891,
"step": 111
},
{
"epoch": 2.140127388535032,
"grad_norm": 0.6213288307189941,
"learning_rate": 2.245515092739488e-06,
"loss": 0.5799,
"step": 112
},
{
"epoch": 2.159235668789809,
"grad_norm": 0.5852116942405701,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.5949,
"step": 113
},
{
"epoch": 2.178343949044586,
"grad_norm": 0.5753389596939087,
"learning_rate": 2.061073738537635e-06,
"loss": 0.566,
"step": 114
},
{
"epoch": 2.1974522292993632,
"grad_norm": 0.5467677712440491,
"learning_rate": 1.971049780795901e-06,
"loss": 0.603,
"step": 115
},
{
"epoch": 2.21656050955414,
"grad_norm": 0.5652862191200256,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.5849,
"step": 116
},
{
"epoch": 2.2356687898089174,
"grad_norm": 0.5632086992263794,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.6131,
"step": 117
},
{
"epoch": 2.254777070063694,
"grad_norm": 0.5731777548789978,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.5665,
"step": 118
},
{
"epoch": 2.2738853503184715,
"grad_norm": 0.5389222502708435,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.5709,
"step": 119
},
{
"epoch": 2.2929936305732483,
"grad_norm": 0.5164850950241089,
"learning_rate": 1.544686755065677e-06,
"loss": 0.5732,
"step": 120
},
{
"epoch": 2.3121019108280256,
"grad_norm": 0.5392653346061707,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5803,
"step": 121
},
{
"epoch": 2.3312101910828025,
"grad_norm": 0.5912431478500366,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.5934,
"step": 122
},
{
"epoch": 2.3503184713375798,
"grad_norm": 0.5611681938171387,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.5743,
"step": 123
},
{
"epoch": 2.3694267515923566,
"grad_norm": 0.598409116268158,
"learning_rate": 1.234642669981946e-06,
"loss": 0.575,
"step": 124
},
{
"epoch": 2.388535031847134,
"grad_norm": 0.5573514103889465,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.5681,
"step": 125
},
{
"epoch": 2.4076433121019107,
"grad_norm": 0.5197586417198181,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5417,
"step": 126
},
{
"epoch": 2.426751592356688,
"grad_norm": 0.5155230760574341,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.5835,
"step": 127
},
{
"epoch": 2.445859872611465,
"grad_norm": 0.5502220988273621,
"learning_rate": 9.549150281252633e-07,
"loss": 0.5685,
"step": 128
},
{
"epoch": 2.464968152866242,
"grad_norm": 0.5894801020622253,
"learning_rate": 8.899896227604509e-07,
"loss": 0.5897,
"step": 129
},
{
"epoch": 2.484076433121019,
"grad_norm": 1.0258440971374512,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5928,
"step": 130
},
{
"epoch": 2.5031847133757963,
"grad_norm": 0.5264387130737305,
"learning_rate": 7.663790038585794e-07,
"loss": 0.5872,
"step": 131
},
{
"epoch": 2.522292993630573,
"grad_norm": 0.5449411273002625,
"learning_rate": 7.077560319906696e-07,
"loss": 0.588,
"step": 132
},
{
"epoch": 2.5414012738853504,
"grad_norm": 0.5089368224143982,
"learning_rate": 6.512943342215234e-07,
"loss": 0.6063,
"step": 133
},
{
"epoch": 2.5605095541401273,
"grad_norm": 0.5194270610809326,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5653,
"step": 134
},
{
"epoch": 2.5796178343949046,
"grad_norm": 0.46488288044929504,
"learning_rate": 5.449673790581611e-07,
"loss": 0.6105,
"step": 135
},
{
"epoch": 2.5987261146496814,
"grad_norm": 0.546769380569458,
"learning_rate": 4.951556604879049e-07,
"loss": 0.5746,
"step": 136
},
{
"epoch": 2.6178343949044587,
"grad_norm": 0.5026277899742126,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5826,
"step": 137
},
{
"epoch": 2.6369426751592355,
"grad_norm": 0.509587287902832,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5774,
"step": 138
},
{
"epoch": 2.656050955414013,
"grad_norm": 0.5122659802436829,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5886,
"step": 139
},
{
"epoch": 2.6751592356687897,
"grad_norm": 0.5195062160491943,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5855,
"step": 140
},
{
"epoch": 2.694267515923567,
"grad_norm": 0.5074093341827393,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5774,
"step": 141
},
{
"epoch": 2.713375796178344,
"grad_norm": 0.5081133246421814,
"learning_rate": 2.447174185242324e-07,
"loss": 0.5752,
"step": 142
},
{
"epoch": 2.732484076433121,
"grad_norm": 0.49944645166397095,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.6027,
"step": 143
},
{
"epoch": 2.7515923566878984,
"grad_norm": 0.5384798645973206,
"learning_rate": 1.801856965207338e-07,
"loss": 0.5705,
"step": 144
},
{
"epoch": 2.770700636942675,
"grad_norm": 0.4824913442134857,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5458,
"step": 145
},
{
"epoch": 2.789808917197452,
"grad_norm": 0.5386421084403992,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5973,
"step": 146
},
{
"epoch": 2.8089171974522293,
"grad_norm": 0.5020357370376587,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.5524,
"step": 147
},
{
"epoch": 2.8280254777070066,
"grad_norm": 0.5273739695549011,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5675,
"step": 148
},
{
"epoch": 2.8471337579617835,
"grad_norm": 0.5206378102302551,
"learning_rate": 6.15582970243117e-08,
"loss": 0.5573,
"step": 149
},
{
"epoch": 2.8662420382165603,
"grad_norm": 0.48928534984588623,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5716,
"step": 150
},
{
"epoch": 2.8853503184713376,
"grad_norm": 0.52506422996521,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5911,
"step": 151
},
{
"epoch": 2.904458598726115,
"grad_norm": 0.4814032018184662,
"learning_rate": 2.012853002380466e-08,
"loss": 0.5804,
"step": 152
},
{
"epoch": 2.9235668789808917,
"grad_norm": 0.4976033866405487,
"learning_rate": 1.132562476771959e-08,
"loss": 0.5846,
"step": 153
},
{
"epoch": 2.9426751592356686,
"grad_norm": 0.5002295970916748,
"learning_rate": 5.034667293427053e-09,
"loss": 0.582,
"step": 154
},
{
"epoch": 2.961783439490446,
"grad_norm": 0.49213045835494995,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5942,
"step": 155
},
{
"epoch": 2.980891719745223,
"grad_norm": 0.5032421946525574,
"learning_rate": 0.0,
"loss": 0.5574,
"step": 156
},
{
"epoch": 2.980891719745223,
"step": 156,
"total_flos": 1.4444846521853542e+17,
"train_loss": 0.7008312650215931,
"train_runtime": 4021.9564,
"train_samples_per_second": 3.73,
"train_steps_per_second": 0.039
}
],
"logging_steps": 1,
"max_steps": 156,
"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": true
},
"attributes": {}
}
},
"total_flos": 1.4444846521853542e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}