EtashGuha's picture
Upload folder using huggingface_hub
ffdeb89 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": 7.708484172821045,
"learning_rate": 6.25e-07,
"loss": 1.2073,
"step": 1
},
{
"epoch": 0.03821656050955414,
"grad_norm": 7.611099720001221,
"learning_rate": 1.25e-06,
"loss": 1.2027,
"step": 2
},
{
"epoch": 0.05732484076433121,
"grad_norm": 7.214547157287598,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.1863,
"step": 3
},
{
"epoch": 0.07643312101910828,
"grad_norm": 7.0485100746154785,
"learning_rate": 2.5e-06,
"loss": 1.1672,
"step": 4
},
{
"epoch": 0.09554140127388536,
"grad_norm": 6.79018497467041,
"learning_rate": 3.125e-06,
"loss": 1.1641,
"step": 5
},
{
"epoch": 0.11464968152866242,
"grad_norm": 5.033609390258789,
"learning_rate": 3.7500000000000005e-06,
"loss": 1.1304,
"step": 6
},
{
"epoch": 0.1337579617834395,
"grad_norm": 2.866844892501831,
"learning_rate": 4.3750000000000005e-06,
"loss": 1.0871,
"step": 7
},
{
"epoch": 0.15286624203821655,
"grad_norm": 2.427063226699829,
"learning_rate": 5e-06,
"loss": 1.0407,
"step": 8
},
{
"epoch": 0.17197452229299362,
"grad_norm": 5.020963191986084,
"learning_rate": 5.625e-06,
"loss": 1.053,
"step": 9
},
{
"epoch": 0.1910828025477707,
"grad_norm": 5.467742443084717,
"learning_rate": 6.25e-06,
"loss": 1.0493,
"step": 10
},
{
"epoch": 0.21019108280254778,
"grad_norm": 5.2093048095703125,
"learning_rate": 6.875e-06,
"loss": 1.0052,
"step": 11
},
{
"epoch": 0.22929936305732485,
"grad_norm": 3.968076467514038,
"learning_rate": 7.500000000000001e-06,
"loss": 1.0001,
"step": 12
},
{
"epoch": 0.2484076433121019,
"grad_norm": 4.139124870300293,
"learning_rate": 8.125000000000001e-06,
"loss": 1.0206,
"step": 13
},
{
"epoch": 0.267515923566879,
"grad_norm": 3.207618236541748,
"learning_rate": 8.750000000000001e-06,
"loss": 0.97,
"step": 14
},
{
"epoch": 0.28662420382165604,
"grad_norm": 2.585268259048462,
"learning_rate": 9.375000000000001e-06,
"loss": 0.9518,
"step": 15
},
{
"epoch": 0.3057324840764331,
"grad_norm": 2.2902705669403076,
"learning_rate": 1e-05,
"loss": 0.8687,
"step": 16
},
{
"epoch": 0.3248407643312102,
"grad_norm": 2.4253506660461426,
"learning_rate": 9.998741174712534e-06,
"loss": 0.9033,
"step": 17
},
{
"epoch": 0.34394904458598724,
"grad_norm": 2.1728992462158203,
"learning_rate": 9.994965332706574e-06,
"loss": 0.9332,
"step": 18
},
{
"epoch": 0.3630573248407643,
"grad_norm": 1.7330336570739746,
"learning_rate": 9.98867437523228e-06,
"loss": 0.8931,
"step": 19
},
{
"epoch": 0.3821656050955414,
"grad_norm": 1.7003523111343384,
"learning_rate": 9.979871469976197e-06,
"loss": 0.8825,
"step": 20
},
{
"epoch": 0.4012738853503185,
"grad_norm": 1.7072489261627197,
"learning_rate": 9.968561049466214e-06,
"loss": 0.8748,
"step": 21
},
{
"epoch": 0.42038216560509556,
"grad_norm": 1.5173128843307495,
"learning_rate": 9.954748808839675e-06,
"loss": 0.9187,
"step": 22
},
{
"epoch": 0.4394904458598726,
"grad_norm": 1.184166431427002,
"learning_rate": 9.938441702975689e-06,
"loss": 0.8753,
"step": 23
},
{
"epoch": 0.4585987261146497,
"grad_norm": 1.209220290184021,
"learning_rate": 9.91964794299315e-06,
"loss": 0.8688,
"step": 24
},
{
"epoch": 0.47770700636942676,
"grad_norm": 1.1812776327133179,
"learning_rate": 9.898376992116179e-06,
"loss": 0.8692,
"step": 25
},
{
"epoch": 0.4968152866242038,
"grad_norm": 1.200333595275879,
"learning_rate": 9.874639560909118e-06,
"loss": 0.9032,
"step": 26
},
{
"epoch": 0.5159235668789809,
"grad_norm": 1.0679785013198853,
"learning_rate": 9.848447601883436e-06,
"loss": 0.843,
"step": 27
},
{
"epoch": 0.535031847133758,
"grad_norm": 0.872157633304596,
"learning_rate": 9.819814303479268e-06,
"loss": 0.8743,
"step": 28
},
{
"epoch": 0.554140127388535,
"grad_norm": 0.9527489542961121,
"learning_rate": 9.788754083424654e-06,
"loss": 0.9002,
"step": 29
},
{
"epoch": 0.5732484076433121,
"grad_norm": 0.9146716594696045,
"learning_rate": 9.755282581475769e-06,
"loss": 0.8339,
"step": 30
},
{
"epoch": 0.5923566878980892,
"grad_norm": 0.8611162900924683,
"learning_rate": 9.719416651541839e-06,
"loss": 0.8442,
"step": 31
},
{
"epoch": 0.6114649681528662,
"grad_norm": 0.7984014749526978,
"learning_rate": 9.681174353198687e-06,
"loss": 0.8327,
"step": 32
},
{
"epoch": 0.6305732484076433,
"grad_norm": 1.1210668087005615,
"learning_rate": 9.640574942595195e-06,
"loss": 0.8135,
"step": 33
},
{
"epoch": 0.6496815286624203,
"grad_norm": 0.8967783451080322,
"learning_rate": 9.597638862757255e-06,
"loss": 0.8697,
"step": 34
},
{
"epoch": 0.6687898089171974,
"grad_norm": 0.8040908575057983,
"learning_rate": 9.552387733294081e-06,
"loss": 0.8529,
"step": 35
},
{
"epoch": 0.6878980891719745,
"grad_norm": 1.0382592678070068,
"learning_rate": 9.504844339512096e-06,
"loss": 0.8301,
"step": 36
},
{
"epoch": 0.7070063694267515,
"grad_norm": 0.951885461807251,
"learning_rate": 9.45503262094184e-06,
"loss": 0.8276,
"step": 37
},
{
"epoch": 0.7261146496815286,
"grad_norm": 0.8443611264228821,
"learning_rate": 9.40297765928369e-06,
"loss": 0.8255,
"step": 38
},
{
"epoch": 0.7452229299363057,
"grad_norm": 0.8591488003730774,
"learning_rate": 9.348705665778479e-06,
"loss": 0.8254,
"step": 39
},
{
"epoch": 0.7643312101910829,
"grad_norm": 0.7851564288139343,
"learning_rate": 9.292243968009332e-06,
"loss": 0.8579,
"step": 40
},
{
"epoch": 0.7834394904458599,
"grad_norm": 0.8532727956771851,
"learning_rate": 9.233620996141421e-06,
"loss": 0.8215,
"step": 41
},
{
"epoch": 0.802547770700637,
"grad_norm": 0.7524774074554443,
"learning_rate": 9.172866268606514e-06,
"loss": 0.8329,
"step": 42
},
{
"epoch": 0.821656050955414,
"grad_norm": 0.8203862309455872,
"learning_rate": 9.110010377239552e-06,
"loss": 0.8479,
"step": 43
},
{
"epoch": 0.8407643312101911,
"grad_norm": 0.7861639261245728,
"learning_rate": 9.045084971874738e-06,
"loss": 0.8653,
"step": 44
},
{
"epoch": 0.8598726114649682,
"grad_norm": 0.7861957550048828,
"learning_rate": 8.978122744408905e-06,
"loss": 0.8374,
"step": 45
},
{
"epoch": 0.8789808917197452,
"grad_norm": 0.7732595205307007,
"learning_rate": 8.90915741234015e-06,
"loss": 0.8295,
"step": 46
},
{
"epoch": 0.8980891719745223,
"grad_norm": 0.7838217616081238,
"learning_rate": 8.838223701790057e-06,
"loss": 0.8284,
"step": 47
},
{
"epoch": 0.9171974522292994,
"grad_norm": 0.757358193397522,
"learning_rate": 8.765357330018056e-06,
"loss": 0.8344,
"step": 48
},
{
"epoch": 0.9363057324840764,
"grad_norm": 0.6782490611076355,
"learning_rate": 8.690594987436705e-06,
"loss": 0.8132,
"step": 49
},
{
"epoch": 0.9554140127388535,
"grad_norm": 0.7422513365745544,
"learning_rate": 8.613974319136959e-06,
"loss": 0.8021,
"step": 50
},
{
"epoch": 0.9745222929936306,
"grad_norm": 0.8458060622215271,
"learning_rate": 8.535533905932739e-06,
"loss": 0.797,
"step": 51
},
{
"epoch": 0.9936305732484076,
"grad_norm": 0.7343083024024963,
"learning_rate": 8.455313244934324e-06,
"loss": 0.819,
"step": 52
},
{
"epoch": 1.0127388535031847,
"grad_norm": 0.8479852080345154,
"learning_rate": 8.373352729660373e-06,
"loss": 0.8041,
"step": 53
},
{
"epoch": 1.0318471337579618,
"grad_norm": 0.7780753970146179,
"learning_rate": 8.289693629698564e-06,
"loss": 0.7353,
"step": 54
},
{
"epoch": 1.0509554140127388,
"grad_norm": 0.7262213826179504,
"learning_rate": 8.204378069925121e-06,
"loss": 0.7098,
"step": 55
},
{
"epoch": 1.070063694267516,
"grad_norm": 0.7861199378967285,
"learning_rate": 8.117449009293668e-06,
"loss": 0.7316,
"step": 56
},
{
"epoch": 1.089171974522293,
"grad_norm": 0.7583940625190735,
"learning_rate": 8.0289502192041e-06,
"loss": 0.7164,
"step": 57
},
{
"epoch": 1.10828025477707,
"grad_norm": 0.8067767024040222,
"learning_rate": 7.938926261462366e-06,
"loss": 0.7395,
"step": 58
},
{
"epoch": 1.127388535031847,
"grad_norm": 0.8200020790100098,
"learning_rate": 7.84742246584226e-06,
"loss": 0.7217,
"step": 59
},
{
"epoch": 1.1464968152866242,
"grad_norm": 0.7458541989326477,
"learning_rate": 7.754484907260513e-06,
"loss": 0.7164,
"step": 60
},
{
"epoch": 1.1656050955414012,
"grad_norm": 0.7313222289085388,
"learning_rate": 7.660160382576683e-06,
"loss": 0.7025,
"step": 61
},
{
"epoch": 1.1847133757961783,
"grad_norm": 0.8323162198066711,
"learning_rate": 7.564496387029532e-06,
"loss": 0.708,
"step": 62
},
{
"epoch": 1.2038216560509554,
"grad_norm": 0.7040552496910095,
"learning_rate": 7.467541090321735e-06,
"loss": 0.716,
"step": 63
},
{
"epoch": 1.2229299363057324,
"grad_norm": 0.6890729665756226,
"learning_rate": 7.369343312364994e-06,
"loss": 0.7143,
"step": 64
},
{
"epoch": 1.2420382165605095,
"grad_norm": 0.7383683323860168,
"learning_rate": 7.269952498697734e-06,
"loss": 0.6923,
"step": 65
},
{
"epoch": 1.2611464968152866,
"grad_norm": 0.7806842923164368,
"learning_rate": 7.169418695587791e-06,
"loss": 0.7027,
"step": 66
},
{
"epoch": 1.2802547770700636,
"grad_norm": 0.7316763997077942,
"learning_rate": 7.067792524832604e-06,
"loss": 0.7037,
"step": 67
},
{
"epoch": 1.2993630573248407,
"grad_norm": 0.7495648264884949,
"learning_rate": 6.965125158269619e-06,
"loss": 0.7166,
"step": 68
},
{
"epoch": 1.3184713375796178,
"grad_norm": 0.6818484663963318,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.7045,
"step": 69
},
{
"epoch": 1.3375796178343948,
"grad_norm": 0.6459638476371765,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.6958,
"step": 70
},
{
"epoch": 1.356687898089172,
"grad_norm": 0.759671151638031,
"learning_rate": 6.651395309775837e-06,
"loss": 0.6958,
"step": 71
},
{
"epoch": 1.3757961783439492,
"grad_norm": 0.8332314491271973,
"learning_rate": 6.545084971874738e-06,
"loss": 0.7271,
"step": 72
},
{
"epoch": 1.394904458598726,
"grad_norm": 0.7383757829666138,
"learning_rate": 6.437996637160086e-06,
"loss": 0.7338,
"step": 73
},
{
"epoch": 1.4140127388535033,
"grad_norm": 0.7254230380058289,
"learning_rate": 6.330184227833376e-06,
"loss": 0.7093,
"step": 74
},
{
"epoch": 1.4331210191082802,
"grad_norm": 0.8567808866500854,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.7191,
"step": 75
},
{
"epoch": 1.4522292993630574,
"grad_norm": 0.6672337055206299,
"learning_rate": 6.112604669781572e-06,
"loss": 0.6588,
"step": 76
},
{
"epoch": 1.4713375796178343,
"grad_norm": 0.6806163191795349,
"learning_rate": 6.002947078916365e-06,
"loss": 0.695,
"step": 77
},
{
"epoch": 1.4904458598726116,
"grad_norm": 0.7049647569656372,
"learning_rate": 5.892784473993184e-06,
"loss": 0.6863,
"step": 78
},
{
"epoch": 1.5095541401273884,
"grad_norm": 0.7396883368492126,
"learning_rate": 5.782172325201155e-06,
"loss": 0.7092,
"step": 79
},
{
"epoch": 1.5286624203821657,
"grad_norm": 0.6506292223930359,
"learning_rate": 5.671166329088278e-06,
"loss": 0.6712,
"step": 80
},
{
"epoch": 1.5477707006369426,
"grad_norm": 0.6761842370033264,
"learning_rate": 5.559822380516539e-06,
"loss": 0.7179,
"step": 81
},
{
"epoch": 1.5668789808917198,
"grad_norm": 0.7145752310752869,
"learning_rate": 5.448196544517168e-06,
"loss": 0.7166,
"step": 82
},
{
"epoch": 1.5859872611464967,
"grad_norm": 0.763866126537323,
"learning_rate": 5.336345028060199e-06,
"loss": 0.7323,
"step": 83
},
{
"epoch": 1.605095541401274,
"grad_norm": 0.6711301207542419,
"learning_rate": 5.224324151752575e-06,
"loss": 0.7167,
"step": 84
},
{
"epoch": 1.6242038216560508,
"grad_norm": 0.6363449692726135,
"learning_rate": 5.112190321479026e-06,
"loss": 0.7016,
"step": 85
},
{
"epoch": 1.643312101910828,
"grad_norm": 0.7524890303611755,
"learning_rate": 5e-06,
"loss": 0.7403,
"step": 86
},
{
"epoch": 1.662420382165605,
"grad_norm": 0.6629307270050049,
"learning_rate": 4.887809678520976e-06,
"loss": 0.6969,
"step": 87
},
{
"epoch": 1.6815286624203822,
"grad_norm": 0.6388441920280457,
"learning_rate": 4.775675848247427e-06,
"loss": 0.6936,
"step": 88
},
{
"epoch": 1.700636942675159,
"grad_norm": 0.7082536816596985,
"learning_rate": 4.663654971939802e-06,
"loss": 0.7065,
"step": 89
},
{
"epoch": 1.7197452229299364,
"grad_norm": 0.5878234505653381,
"learning_rate": 4.551803455482833e-06,
"loss": 0.6884,
"step": 90
},
{
"epoch": 1.7388535031847132,
"grad_norm": 0.6400618553161621,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.6928,
"step": 91
},
{
"epoch": 1.7579617834394905,
"grad_norm": 0.7262267470359802,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.6991,
"step": 92
},
{
"epoch": 1.7770700636942676,
"grad_norm": 0.6692636013031006,
"learning_rate": 4.217827674798845e-06,
"loss": 0.6845,
"step": 93
},
{
"epoch": 1.7961783439490446,
"grad_norm": 0.6550084352493286,
"learning_rate": 4.107215526006818e-06,
"loss": 0.706,
"step": 94
},
{
"epoch": 1.8152866242038217,
"grad_norm": 0.7849605679512024,
"learning_rate": 3.997052921083637e-06,
"loss": 0.7135,
"step": 95
},
{
"epoch": 1.8343949044585988,
"grad_norm": 0.7244142889976501,
"learning_rate": 3.887395330218429e-06,
"loss": 0.7136,
"step": 96
},
{
"epoch": 1.8535031847133758,
"grad_norm": 0.6480051279067993,
"learning_rate": 3.778297969310529e-06,
"loss": 0.7117,
"step": 97
},
{
"epoch": 1.872611464968153,
"grad_norm": 0.6031900644302368,
"learning_rate": 3.669815772166625e-06,
"loss": 0.6943,
"step": 98
},
{
"epoch": 1.89171974522293,
"grad_norm": 0.6481626629829407,
"learning_rate": 3.562003362839914e-06,
"loss": 0.7036,
"step": 99
},
{
"epoch": 1.910828025477707,
"grad_norm": 0.6028628945350647,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.691,
"step": 100
},
{
"epoch": 1.929936305732484,
"grad_norm": 0.6366122364997864,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.715,
"step": 101
},
{
"epoch": 1.9490445859872612,
"grad_norm": 0.7005493640899658,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.6892,
"step": 102
},
{
"epoch": 1.9681528662420382,
"grad_norm": 0.645294725894928,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.6585,
"step": 103
},
{
"epoch": 1.9872611464968153,
"grad_norm": 0.6466977596282959,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.7003,
"step": 104
},
{
"epoch": 2.0063694267515926,
"grad_norm": 0.6410728096961975,
"learning_rate": 2.932207475167398e-06,
"loss": 0.6645,
"step": 105
},
{
"epoch": 2.0254777070063694,
"grad_norm": 0.6502147912979126,
"learning_rate": 2.83058130441221e-06,
"loss": 0.6352,
"step": 106
},
{
"epoch": 2.0445859872611467,
"grad_norm": 0.690471351146698,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.6252,
"step": 107
},
{
"epoch": 2.0636942675159236,
"grad_norm": 0.6230637431144714,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.6184,
"step": 108
},
{
"epoch": 2.082802547770701,
"grad_norm": 0.6241819262504578,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5941,
"step": 109
},
{
"epoch": 2.1019108280254777,
"grad_norm": 0.6172868013381958,
"learning_rate": 2.43550361297047e-06,
"loss": 0.6202,
"step": 110
},
{
"epoch": 2.121019108280255,
"grad_norm": 0.6882143616676331,
"learning_rate": 2.339839617423318e-06,
"loss": 0.6181,
"step": 111
},
{
"epoch": 2.140127388535032,
"grad_norm": 0.678433358669281,
"learning_rate": 2.245515092739488e-06,
"loss": 0.6061,
"step": 112
},
{
"epoch": 2.159235668789809,
"grad_norm": 0.6652299165725708,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.6266,
"step": 113
},
{
"epoch": 2.178343949044586,
"grad_norm": 0.6305775046348572,
"learning_rate": 2.061073738537635e-06,
"loss": 0.6136,
"step": 114
},
{
"epoch": 2.1974522292993632,
"grad_norm": 0.6421285271644592,
"learning_rate": 1.971049780795901e-06,
"loss": 0.5996,
"step": 115
},
{
"epoch": 2.21656050955414,
"grad_norm": 0.5938356518745422,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.6067,
"step": 116
},
{
"epoch": 2.2356687898089174,
"grad_norm": 0.6172759532928467,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.6304,
"step": 117
},
{
"epoch": 2.254777070063694,
"grad_norm": 0.6095340251922607,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.6251,
"step": 118
},
{
"epoch": 2.2738853503184715,
"grad_norm": 0.6015990376472473,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.6078,
"step": 119
},
{
"epoch": 2.2929936305732483,
"grad_norm": 0.562563419342041,
"learning_rate": 1.544686755065677e-06,
"loss": 0.6215,
"step": 120
},
{
"epoch": 2.3121019108280256,
"grad_norm": 0.6117820739746094,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.6161,
"step": 121
},
{
"epoch": 2.3312101910828025,
"grad_norm": 0.6286083459854126,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.6061,
"step": 122
},
{
"epoch": 2.3503184713375798,
"grad_norm": 0.6039571166038513,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.6138,
"step": 123
},
{
"epoch": 2.3694267515923566,
"grad_norm": 0.5991165637969971,
"learning_rate": 1.234642669981946e-06,
"loss": 0.609,
"step": 124
},
{
"epoch": 2.388535031847134,
"grad_norm": 0.5799904465675354,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.5998,
"step": 125
},
{
"epoch": 2.4076433121019107,
"grad_norm": 0.5716231465339661,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.6081,
"step": 126
},
{
"epoch": 2.426751592356688,
"grad_norm": 0.6321515440940857,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.6126,
"step": 127
},
{
"epoch": 2.445859872611465,
"grad_norm": 0.6207079887390137,
"learning_rate": 9.549150281252633e-07,
"loss": 0.6433,
"step": 128
},
{
"epoch": 2.464968152866242,
"grad_norm": 0.5693238973617554,
"learning_rate": 8.899896227604509e-07,
"loss": 0.6239,
"step": 129
},
{
"epoch": 2.484076433121019,
"grad_norm": 0.6268942952156067,
"learning_rate": 8.271337313934869e-07,
"loss": 0.6018,
"step": 130
},
{
"epoch": 2.5031847133757963,
"grad_norm": 0.5692458152770996,
"learning_rate": 7.663790038585794e-07,
"loss": 0.6089,
"step": 131
},
{
"epoch": 2.522292993630573,
"grad_norm": 0.5981783270835876,
"learning_rate": 7.077560319906696e-07,
"loss": 0.6138,
"step": 132
},
{
"epoch": 2.5414012738853504,
"grad_norm": 0.5806254148483276,
"learning_rate": 6.512943342215234e-07,
"loss": 0.6205,
"step": 133
},
{
"epoch": 2.5605095541401273,
"grad_norm": 0.6364081501960754,
"learning_rate": 5.9702234071631e-07,
"loss": 0.6085,
"step": 134
},
{
"epoch": 2.5796178343949046,
"grad_norm": 0.5379471182823181,
"learning_rate": 5.449673790581611e-07,
"loss": 0.6078,
"step": 135
},
{
"epoch": 2.5987261146496814,
"grad_norm": 0.5562213063240051,
"learning_rate": 4.951556604879049e-07,
"loss": 0.6079,
"step": 136
},
{
"epoch": 2.6178343949044587,
"grad_norm": 0.5902286171913147,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.6007,
"step": 137
},
{
"epoch": 2.6369426751592355,
"grad_norm": 0.6162657141685486,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5939,
"step": 138
},
{
"epoch": 2.656050955414013,
"grad_norm": 0.5354500412940979,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.6016,
"step": 139
},
{
"epoch": 2.6751592356687897,
"grad_norm": 0.5618153214454651,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5971,
"step": 140
},
{
"epoch": 2.694267515923567,
"grad_norm": 0.5426648259162903,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5973,
"step": 141
},
{
"epoch": 2.713375796178344,
"grad_norm": 0.5665692687034607,
"learning_rate": 2.447174185242324e-07,
"loss": 0.6194,
"step": 142
},
{
"epoch": 2.732484076433121,
"grad_norm": 0.556145191192627,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.6287,
"step": 143
},
{
"epoch": 2.7515923566878984,
"grad_norm": 0.5904255509376526,
"learning_rate": 1.801856965207338e-07,
"loss": 0.6378,
"step": 144
},
{
"epoch": 2.770700636942675,
"grad_norm": 0.5670586824417114,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.6061,
"step": 145
},
{
"epoch": 2.789808917197452,
"grad_norm": 0.6461930274963379,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5864,
"step": 146
},
{
"epoch": 2.8089171974522293,
"grad_norm": 0.5549110174179077,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.614,
"step": 147
},
{
"epoch": 2.8280254777070066,
"grad_norm": 0.5617705583572388,
"learning_rate": 8.035205700685167e-08,
"loss": 0.6184,
"step": 148
},
{
"epoch": 2.8471337579617835,
"grad_norm": 0.5576688051223755,
"learning_rate": 6.15582970243117e-08,
"loss": 0.6153,
"step": 149
},
{
"epoch": 2.8662420382165603,
"grad_norm": 0.5535495281219482,
"learning_rate": 4.52511911603265e-08,
"loss": 0.6011,
"step": 150
},
{
"epoch": 2.8853503184713376,
"grad_norm": 0.5694997906684875,
"learning_rate": 3.143895053378698e-08,
"loss": 0.6115,
"step": 151
},
{
"epoch": 2.904458598726115,
"grad_norm": 0.5671418309211731,
"learning_rate": 2.012853002380466e-08,
"loss": 0.5982,
"step": 152
},
{
"epoch": 2.9235668789808917,
"grad_norm": 0.582301676273346,
"learning_rate": 1.132562476771959e-08,
"loss": 0.6132,
"step": 153
},
{
"epoch": 2.9426751592356686,
"grad_norm": 0.5656945705413818,
"learning_rate": 5.034667293427053e-09,
"loss": 0.5871,
"step": 154
},
{
"epoch": 2.961783439490446,
"grad_norm": 0.570439875125885,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.6125,
"step": 155
},
{
"epoch": 2.980891719745223,
"grad_norm": 0.5605185627937317,
"learning_rate": 0.0,
"loss": 0.5954,
"step": 156
},
{
"epoch": 2.980891719745223,
"step": 156,
"total_flos": 1.266291131972649e+17,
"train_loss": 0.7465221996490772,
"train_runtime": 3657.1071,
"train_samples_per_second": 4.102,
"train_steps_per_second": 0.043
}
],
"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": 1.266291131972649e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}