training_baseline / trainer_state.json
sedrickkeh's picture
End of training
fc8b3a7 verified
raw
history blame
27.7 kB
{
"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": 9.2407865524292,
"learning_rate": 6.25e-07,
"loss": 1.3429,
"step": 1
},
{
"epoch": 0.03821656050955414,
"grad_norm": 9.478723526000977,
"learning_rate": 1.25e-06,
"loss": 1.3616,
"step": 2
},
{
"epoch": 0.05732484076433121,
"grad_norm": 8.822339057922363,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.2959,
"step": 3
},
{
"epoch": 0.07643312101910828,
"grad_norm": 8.794820785522461,
"learning_rate": 2.5e-06,
"loss": 1.3065,
"step": 4
},
{
"epoch": 0.09554140127388536,
"grad_norm": 8.383583068847656,
"learning_rate": 3.125e-06,
"loss": 1.3084,
"step": 5
},
{
"epoch": 0.11464968152866242,
"grad_norm": 6.294646263122559,
"learning_rate": 3.7500000000000005e-06,
"loss": 1.2614,
"step": 6
},
{
"epoch": 0.1337579617834395,
"grad_norm": 3.9098663330078125,
"learning_rate": 4.3750000000000005e-06,
"loss": 1.177,
"step": 7
},
{
"epoch": 0.15286624203821655,
"grad_norm": 3.4435949325561523,
"learning_rate": 5e-06,
"loss": 1.2101,
"step": 8
},
{
"epoch": 0.17197452229299362,
"grad_norm": 4.044913291931152,
"learning_rate": 5.625e-06,
"loss": 1.1228,
"step": 9
},
{
"epoch": 0.1910828025477707,
"grad_norm": 4.243114948272705,
"learning_rate": 6.25e-06,
"loss": 1.0775,
"step": 10
},
{
"epoch": 0.21019108280254778,
"grad_norm": 4.074731349945068,
"learning_rate": 6.875e-06,
"loss": 1.0759,
"step": 11
},
{
"epoch": 0.22929936305732485,
"grad_norm": 3.4349539279937744,
"learning_rate": 7.500000000000001e-06,
"loss": 1.0192,
"step": 12
},
{
"epoch": 0.2484076433121019,
"grad_norm": 3.8267672061920166,
"learning_rate": 8.125000000000001e-06,
"loss": 1.0476,
"step": 13
},
{
"epoch": 0.267515923566879,
"grad_norm": 3.0685765743255615,
"learning_rate": 8.750000000000001e-06,
"loss": 1.0031,
"step": 14
},
{
"epoch": 0.28662420382165604,
"grad_norm": 2.2036056518554688,
"learning_rate": 9.375000000000001e-06,
"loss": 0.9644,
"step": 15
},
{
"epoch": 0.3057324840764331,
"grad_norm": 2.0634169578552246,
"learning_rate": 1e-05,
"loss": 0.958,
"step": 16
},
{
"epoch": 0.3248407643312102,
"grad_norm": 2.5293338298797607,
"learning_rate": 9.998741174712534e-06,
"loss": 0.9512,
"step": 17
},
{
"epoch": 0.34394904458598724,
"grad_norm": 2.4025402069091797,
"learning_rate": 9.994965332706574e-06,
"loss": 0.9528,
"step": 18
},
{
"epoch": 0.3630573248407643,
"grad_norm": 1.8339530229568481,
"learning_rate": 9.98867437523228e-06,
"loss": 0.9513,
"step": 19
},
{
"epoch": 0.3821656050955414,
"grad_norm": 1.4116219282150269,
"learning_rate": 9.979871469976197e-06,
"loss": 0.9269,
"step": 20
},
{
"epoch": 0.4012738853503185,
"grad_norm": 1.4383982419967651,
"learning_rate": 9.968561049466214e-06,
"loss": 0.9384,
"step": 21
},
{
"epoch": 0.42038216560509556,
"grad_norm": 1.4755058288574219,
"learning_rate": 9.954748808839675e-06,
"loss": 0.8952,
"step": 22
},
{
"epoch": 0.4394904458598726,
"grad_norm": 1.1699796915054321,
"learning_rate": 9.938441702975689e-06,
"loss": 0.8937,
"step": 23
},
{
"epoch": 0.4585987261146497,
"grad_norm": 1.0133947134017944,
"learning_rate": 9.91964794299315e-06,
"loss": 0.9142,
"step": 24
},
{
"epoch": 0.47770700636942676,
"grad_norm": 1.1321440935134888,
"learning_rate": 9.898376992116179e-06,
"loss": 0.8876,
"step": 25
},
{
"epoch": 0.4968152866242038,
"grad_norm": 1.1128759384155273,
"learning_rate": 9.874639560909118e-06,
"loss": 0.8914,
"step": 26
},
{
"epoch": 0.5159235668789809,
"grad_norm": 1.152822732925415,
"learning_rate": 9.848447601883436e-06,
"loss": 0.918,
"step": 27
},
{
"epoch": 0.535031847133758,
"grad_norm": 1.0430890321731567,
"learning_rate": 9.819814303479268e-06,
"loss": 0.9226,
"step": 28
},
{
"epoch": 0.554140127388535,
"grad_norm": 1.0511760711669922,
"learning_rate": 9.788754083424654e-06,
"loss": 0.8652,
"step": 29
},
{
"epoch": 0.5732484076433121,
"grad_norm": 0.9650465250015259,
"learning_rate": 9.755282581475769e-06,
"loss": 0.8734,
"step": 30
},
{
"epoch": 0.5923566878980892,
"grad_norm": 0.9453793168067932,
"learning_rate": 9.719416651541839e-06,
"loss": 0.8965,
"step": 31
},
{
"epoch": 0.6114649681528662,
"grad_norm": 1.0314666032791138,
"learning_rate": 9.681174353198687e-06,
"loss": 0.896,
"step": 32
},
{
"epoch": 0.6305732484076433,
"grad_norm": 1.0188624858856201,
"learning_rate": 9.640574942595195e-06,
"loss": 0.8808,
"step": 33
},
{
"epoch": 0.6496815286624203,
"grad_norm": 0.9808051586151123,
"learning_rate": 9.597638862757255e-06,
"loss": 0.8398,
"step": 34
},
{
"epoch": 0.6687898089171974,
"grad_norm": 6.07795524597168,
"learning_rate": 9.552387733294081e-06,
"loss": 0.8645,
"step": 35
},
{
"epoch": 0.6878980891719745,
"grad_norm": 1.1291970014572144,
"learning_rate": 9.504844339512096e-06,
"loss": 0.8813,
"step": 36
},
{
"epoch": 0.7070063694267515,
"grad_norm": 1.2054293155670166,
"learning_rate": 9.45503262094184e-06,
"loss": 0.8801,
"step": 37
},
{
"epoch": 0.7261146496815286,
"grad_norm": 1.0582106113433838,
"learning_rate": 9.40297765928369e-06,
"loss": 0.8923,
"step": 38
},
{
"epoch": 0.7452229299363057,
"grad_norm": 0.9821107983589172,
"learning_rate": 9.348705665778479e-06,
"loss": 0.856,
"step": 39
},
{
"epoch": 0.7643312101910829,
"grad_norm": 0.9655761122703552,
"learning_rate": 9.292243968009332e-06,
"loss": 0.8046,
"step": 40
},
{
"epoch": 0.7834394904458599,
"grad_norm": 1.0355433225631714,
"learning_rate": 9.233620996141421e-06,
"loss": 0.8774,
"step": 41
},
{
"epoch": 0.802547770700637,
"grad_norm": 0.9365308880805969,
"learning_rate": 9.172866268606514e-06,
"loss": 0.8529,
"step": 42
},
{
"epoch": 0.821656050955414,
"grad_norm": 1.033125638961792,
"learning_rate": 9.110010377239552e-06,
"loss": 0.8878,
"step": 43
},
{
"epoch": 0.8407643312101911,
"grad_norm": 1.0032130479812622,
"learning_rate": 9.045084971874738e-06,
"loss": 0.8456,
"step": 44
},
{
"epoch": 0.8598726114649682,
"grad_norm": 0.9883443117141724,
"learning_rate": 8.978122744408905e-06,
"loss": 0.8591,
"step": 45
},
{
"epoch": 0.8789808917197452,
"grad_norm": 1.064292073249817,
"learning_rate": 8.90915741234015e-06,
"loss": 0.8302,
"step": 46
},
{
"epoch": 0.8980891719745223,
"grad_norm": 0.9541292786598206,
"learning_rate": 8.838223701790057e-06,
"loss": 0.8265,
"step": 47
},
{
"epoch": 0.9171974522292994,
"grad_norm": 1.0820716619491577,
"learning_rate": 8.765357330018056e-06,
"loss": 0.8239,
"step": 48
},
{
"epoch": 0.9363057324840764,
"grad_norm": 1.0467376708984375,
"learning_rate": 8.690594987436705e-06,
"loss": 0.8543,
"step": 49
},
{
"epoch": 0.9554140127388535,
"grad_norm": 0.9927631616592407,
"learning_rate": 8.613974319136959e-06,
"loss": 0.8486,
"step": 50
},
{
"epoch": 0.9745222929936306,
"grad_norm": 1.1420801877975464,
"learning_rate": 8.535533905932739e-06,
"loss": 0.8717,
"step": 51
},
{
"epoch": 0.9936305732484076,
"grad_norm": 0.9995855093002319,
"learning_rate": 8.455313244934324e-06,
"loss": 0.851,
"step": 52
},
{
"epoch": 1.0127388535031847,
"grad_norm": 1.773233413696289,
"learning_rate": 8.373352729660373e-06,
"loss": 1.2804,
"step": 53
},
{
"epoch": 1.0318471337579618,
"grad_norm": 1.146483302116394,
"learning_rate": 8.289693629698564e-06,
"loss": 0.736,
"step": 54
},
{
"epoch": 1.0509554140127388,
"grad_norm": 0.9704205989837646,
"learning_rate": 8.204378069925121e-06,
"loss": 0.7509,
"step": 55
},
{
"epoch": 1.070063694267516,
"grad_norm": 0.9973506331443787,
"learning_rate": 8.117449009293668e-06,
"loss": 0.7378,
"step": 56
},
{
"epoch": 1.089171974522293,
"grad_norm": 0.9104877710342407,
"learning_rate": 8.0289502192041e-06,
"loss": 0.696,
"step": 57
},
{
"epoch": 1.10828025477707,
"grad_norm": 1.128894567489624,
"learning_rate": 7.938926261462366e-06,
"loss": 0.7265,
"step": 58
},
{
"epoch": 1.127388535031847,
"grad_norm": 0.9157301783561707,
"learning_rate": 7.84742246584226e-06,
"loss": 0.7366,
"step": 59
},
{
"epoch": 1.1464968152866242,
"grad_norm": 0.9085549116134644,
"learning_rate": 7.754484907260513e-06,
"loss": 0.6838,
"step": 60
},
{
"epoch": 1.1656050955414012,
"grad_norm": 0.9201124906539917,
"learning_rate": 7.660160382576683e-06,
"loss": 0.7569,
"step": 61
},
{
"epoch": 1.1847133757961783,
"grad_norm": 0.9689205288887024,
"learning_rate": 7.564496387029532e-06,
"loss": 0.6864,
"step": 62
},
{
"epoch": 1.2038216560509554,
"grad_norm": 0.9296672344207764,
"learning_rate": 7.467541090321735e-06,
"loss": 0.7402,
"step": 63
},
{
"epoch": 1.2229299363057324,
"grad_norm": 0.965122640132904,
"learning_rate": 7.369343312364994e-06,
"loss": 0.6695,
"step": 64
},
{
"epoch": 1.2420382165605095,
"grad_norm": 0.9942020177841187,
"learning_rate": 7.269952498697734e-06,
"loss": 0.7511,
"step": 65
},
{
"epoch": 1.2611464968152866,
"grad_norm": 0.9647576808929443,
"learning_rate": 7.169418695587791e-06,
"loss": 0.7272,
"step": 66
},
{
"epoch": 1.2802547770700636,
"grad_norm": 0.8429343104362488,
"learning_rate": 7.067792524832604e-06,
"loss": 0.743,
"step": 67
},
{
"epoch": 1.2993630573248407,
"grad_norm": 0.8730988502502441,
"learning_rate": 6.965125158269619e-06,
"loss": 0.689,
"step": 68
},
{
"epoch": 1.3184713375796178,
"grad_norm": 0.95030677318573,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.7515,
"step": 69
},
{
"epoch": 1.3375796178343948,
"grad_norm": 0.8007903099060059,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.6767,
"step": 70
},
{
"epoch": 1.356687898089172,
"grad_norm": 0.9224691390991211,
"learning_rate": 6.651395309775837e-06,
"loss": 0.7164,
"step": 71
},
{
"epoch": 1.3757961783439492,
"grad_norm": 0.7532793283462524,
"learning_rate": 6.545084971874738e-06,
"loss": 0.6844,
"step": 72
},
{
"epoch": 1.394904458598726,
"grad_norm": 0.8250970244407654,
"learning_rate": 6.437996637160086e-06,
"loss": 0.7026,
"step": 73
},
{
"epoch": 1.4140127388535033,
"grad_norm": 0.8471970558166504,
"learning_rate": 6.330184227833376e-06,
"loss": 0.6725,
"step": 74
},
{
"epoch": 1.4331210191082802,
"grad_norm": 0.8257316946983337,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.6793,
"step": 75
},
{
"epoch": 1.4522292993630574,
"grad_norm": 0.7594157457351685,
"learning_rate": 6.112604669781572e-06,
"loss": 0.71,
"step": 76
},
{
"epoch": 1.4713375796178343,
"grad_norm": 0.8231765627861023,
"learning_rate": 6.002947078916365e-06,
"loss": 0.7449,
"step": 77
},
{
"epoch": 1.4904458598726116,
"grad_norm": 0.7706438899040222,
"learning_rate": 5.892784473993184e-06,
"loss": 0.6683,
"step": 78
},
{
"epoch": 1.5095541401273884,
"grad_norm": 0.7732676863670349,
"learning_rate": 5.782172325201155e-06,
"loss": 0.6617,
"step": 79
},
{
"epoch": 1.5286624203821657,
"grad_norm": 0.7970235347747803,
"learning_rate": 5.671166329088278e-06,
"loss": 0.7013,
"step": 80
},
{
"epoch": 1.5477707006369426,
"grad_norm": 0.7982433438301086,
"learning_rate": 5.559822380516539e-06,
"loss": 0.6928,
"step": 81
},
{
"epoch": 1.5668789808917198,
"grad_norm": 0.7948117852210999,
"learning_rate": 5.448196544517168e-06,
"loss": 0.7225,
"step": 82
},
{
"epoch": 1.5859872611464967,
"grad_norm": 0.8307470083236694,
"learning_rate": 5.336345028060199e-06,
"loss": 0.7372,
"step": 83
},
{
"epoch": 1.605095541401274,
"grad_norm": 0.7786919474601746,
"learning_rate": 5.224324151752575e-06,
"loss": 0.6999,
"step": 84
},
{
"epoch": 1.6242038216560508,
"grad_norm": 0.8192756772041321,
"learning_rate": 5.112190321479026e-06,
"loss": 0.7043,
"step": 85
},
{
"epoch": 1.643312101910828,
"grad_norm": 0.7583406567573547,
"learning_rate": 5e-06,
"loss": 0.6697,
"step": 86
},
{
"epoch": 1.662420382165605,
"grad_norm": 0.8197351694107056,
"learning_rate": 4.887809678520976e-06,
"loss": 0.7563,
"step": 87
},
{
"epoch": 1.6815286624203822,
"grad_norm": 0.8537828922271729,
"learning_rate": 4.775675848247427e-06,
"loss": 0.714,
"step": 88
},
{
"epoch": 1.700636942675159,
"grad_norm": 0.8561527729034424,
"learning_rate": 4.663654971939802e-06,
"loss": 0.6812,
"step": 89
},
{
"epoch": 1.7197452229299364,
"grad_norm": 0.7625755667686462,
"learning_rate": 4.551803455482833e-06,
"loss": 0.7148,
"step": 90
},
{
"epoch": 1.7388535031847132,
"grad_norm": 0.7977814078330994,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.7023,
"step": 91
},
{
"epoch": 1.7579617834394905,
"grad_norm": 0.8178286552429199,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.7438,
"step": 92
},
{
"epoch": 1.7770700636942676,
"grad_norm": 0.8001686334609985,
"learning_rate": 4.217827674798845e-06,
"loss": 0.6842,
"step": 93
},
{
"epoch": 1.7961783439490446,
"grad_norm": 0.754313051700592,
"learning_rate": 4.107215526006818e-06,
"loss": 0.6574,
"step": 94
},
{
"epoch": 1.8152866242038217,
"grad_norm": 0.7473746538162231,
"learning_rate": 3.997052921083637e-06,
"loss": 0.7421,
"step": 95
},
{
"epoch": 1.8343949044585988,
"grad_norm": 0.7881169319152832,
"learning_rate": 3.887395330218429e-06,
"loss": 0.7061,
"step": 96
},
{
"epoch": 1.8535031847133758,
"grad_norm": 0.7832265496253967,
"learning_rate": 3.778297969310529e-06,
"loss": 0.6998,
"step": 97
},
{
"epoch": 1.872611464968153,
"grad_norm": 0.8894858956336975,
"learning_rate": 3.669815772166625e-06,
"loss": 0.7187,
"step": 98
},
{
"epoch": 1.89171974522293,
"grad_norm": 0.7932900190353394,
"learning_rate": 3.562003362839914e-06,
"loss": 0.6835,
"step": 99
},
{
"epoch": 1.910828025477707,
"grad_norm": 0.8342806696891785,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.6395,
"step": 100
},
{
"epoch": 1.929936305732484,
"grad_norm": 0.8320187926292419,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.7582,
"step": 101
},
{
"epoch": 1.9490445859872612,
"grad_norm": 0.7808068990707397,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.6748,
"step": 102
},
{
"epoch": 1.9681528662420382,
"grad_norm": 0.6990087032318115,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.6935,
"step": 103
},
{
"epoch": 1.9872611464968153,
"grad_norm": 0.8083034753799438,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.6842,
"step": 104
},
{
"epoch": 2.0063694267515926,
"grad_norm": 1.6020653247833252,
"learning_rate": 2.932207475167398e-06,
"loss": 1.1038,
"step": 105
},
{
"epoch": 2.0254777070063694,
"grad_norm": 0.8427936434745789,
"learning_rate": 2.83058130441221e-06,
"loss": 0.6547,
"step": 106
},
{
"epoch": 2.0445859872611467,
"grad_norm": 0.8315389156341553,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.565,
"step": 107
},
{
"epoch": 2.0636942675159236,
"grad_norm": 0.7301693558692932,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5997,
"step": 108
},
{
"epoch": 2.082802547770701,
"grad_norm": 0.6864070892333984,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5757,
"step": 109
},
{
"epoch": 2.1019108280254777,
"grad_norm": 0.7325605750083923,
"learning_rate": 2.43550361297047e-06,
"loss": 0.6274,
"step": 110
},
{
"epoch": 2.121019108280255,
"grad_norm": 0.7794485092163086,
"learning_rate": 2.339839617423318e-06,
"loss": 0.6346,
"step": 111
},
{
"epoch": 2.140127388535032,
"grad_norm": 0.7400060892105103,
"learning_rate": 2.245515092739488e-06,
"loss": 0.6,
"step": 112
},
{
"epoch": 2.159235668789809,
"grad_norm": 0.8290343284606934,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.6427,
"step": 113
},
{
"epoch": 2.178343949044586,
"grad_norm": 0.7199001312255859,
"learning_rate": 2.061073738537635e-06,
"loss": 0.5617,
"step": 114
},
{
"epoch": 2.1974522292993632,
"grad_norm": 0.7009884715080261,
"learning_rate": 1.971049780795901e-06,
"loss": 0.5894,
"step": 115
},
{
"epoch": 2.21656050955414,
"grad_norm": 0.7035958766937256,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.5989,
"step": 116
},
{
"epoch": 2.2356687898089174,
"grad_norm": 0.7354803085327148,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.6264,
"step": 117
},
{
"epoch": 2.254777070063694,
"grad_norm": 0.6822336316108704,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.5717,
"step": 118
},
{
"epoch": 2.2738853503184715,
"grad_norm": 0.7196558117866516,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.608,
"step": 119
},
{
"epoch": 2.2929936305732483,
"grad_norm": 0.7429059743881226,
"learning_rate": 1.544686755065677e-06,
"loss": 0.6242,
"step": 120
},
{
"epoch": 2.3121019108280256,
"grad_norm": 0.7424628138542175,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5927,
"step": 121
},
{
"epoch": 2.3312101910828025,
"grad_norm": 0.7479440569877625,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.5857,
"step": 122
},
{
"epoch": 2.3503184713375798,
"grad_norm": 0.7213776707649231,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.6115,
"step": 123
},
{
"epoch": 2.3694267515923566,
"grad_norm": 0.6591588258743286,
"learning_rate": 1.234642669981946e-06,
"loss": 0.5559,
"step": 124
},
{
"epoch": 2.388535031847134,
"grad_norm": 0.7499435544013977,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.6358,
"step": 125
},
{
"epoch": 2.4076433121019107,
"grad_norm": 0.6854066848754883,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5673,
"step": 126
},
{
"epoch": 2.426751592356688,
"grad_norm": 0.7064222693443298,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.603,
"step": 127
},
{
"epoch": 2.445859872611465,
"grad_norm": 0.7362521290779114,
"learning_rate": 9.549150281252633e-07,
"loss": 0.6113,
"step": 128
},
{
"epoch": 2.464968152866242,
"grad_norm": 0.742847740650177,
"learning_rate": 8.899896227604509e-07,
"loss": 0.601,
"step": 129
},
{
"epoch": 2.484076433121019,
"grad_norm": 0.6763571500778198,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5505,
"step": 130
},
{
"epoch": 2.5031847133757963,
"grad_norm": 0.7390684485435486,
"learning_rate": 7.663790038585794e-07,
"loss": 0.6827,
"step": 131
},
{
"epoch": 2.522292993630573,
"grad_norm": 0.6655154228210449,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5709,
"step": 132
},
{
"epoch": 2.5414012738853504,
"grad_norm": 0.7252643704414368,
"learning_rate": 6.512943342215234e-07,
"loss": 0.6091,
"step": 133
},
{
"epoch": 2.5605095541401273,
"grad_norm": 0.6566873788833618,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5609,
"step": 134
},
{
"epoch": 2.5796178343949046,
"grad_norm": 0.6787551641464233,
"learning_rate": 5.449673790581611e-07,
"loss": 0.5865,
"step": 135
},
{
"epoch": 2.5987261146496814,
"grad_norm": 0.6773931980133057,
"learning_rate": 4.951556604879049e-07,
"loss": 0.5847,
"step": 136
},
{
"epoch": 2.6178343949044587,
"grad_norm": 0.7028627395629883,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5941,
"step": 137
},
{
"epoch": 2.6369426751592355,
"grad_norm": 0.6910837888717651,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.6,
"step": 138
},
{
"epoch": 2.656050955414013,
"grad_norm": 0.6671729683876038,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5928,
"step": 139
},
{
"epoch": 2.6751592356687897,
"grad_norm": 0.6896730065345764,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5801,
"step": 140
},
{
"epoch": 2.694267515923567,
"grad_norm": 0.6889122128486633,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.6199,
"step": 141
},
{
"epoch": 2.713375796178344,
"grad_norm": 0.691997230052948,
"learning_rate": 2.447174185242324e-07,
"loss": 0.6229,
"step": 142
},
{
"epoch": 2.732484076433121,
"grad_norm": 0.6809545755386353,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.6053,
"step": 143
},
{
"epoch": 2.7515923566878984,
"grad_norm": 0.670366108417511,
"learning_rate": 1.801856965207338e-07,
"loss": 0.5444,
"step": 144
},
{
"epoch": 2.770700636942675,
"grad_norm": 0.664279580116272,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5756,
"step": 145
},
{
"epoch": 2.789808917197452,
"grad_norm": 0.6993831396102905,
"learning_rate": 1.253604390908819e-07,
"loss": 0.6188,
"step": 146
},
{
"epoch": 2.8089171974522293,
"grad_norm": 0.7156822085380554,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.6433,
"step": 147
},
{
"epoch": 2.8280254777070066,
"grad_norm": 0.6507427096366882,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5587,
"step": 148
},
{
"epoch": 2.8471337579617835,
"grad_norm": 0.624755322933197,
"learning_rate": 6.15582970243117e-08,
"loss": 0.5927,
"step": 149
},
{
"epoch": 2.8662420382165603,
"grad_norm": 0.6648584604263306,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5883,
"step": 150
},
{
"epoch": 2.8853503184713376,
"grad_norm": 0.6581620573997498,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5976,
"step": 151
},
{
"epoch": 2.904458598726115,
"grad_norm": 0.699501633644104,
"learning_rate": 2.012853002380466e-08,
"loss": 0.5732,
"step": 152
},
{
"epoch": 2.9235668789808917,
"grad_norm": 0.7073370814323425,
"learning_rate": 1.132562476771959e-08,
"loss": 0.6537,
"step": 153
},
{
"epoch": 2.9426751592356686,
"grad_norm": 0.6681355834007263,
"learning_rate": 5.034667293427053e-09,
"loss": 0.5774,
"step": 154
},
{
"epoch": 2.961783439490446,
"grad_norm": 0.6787852048873901,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.58,
"step": 155
},
{
"epoch": 2.980891719745223,
"grad_norm": 0.6549314856529236,
"learning_rate": 0.0,
"loss": 0.6071,
"step": 156
},
{
"epoch": 2.980891719745223,
"step": 156,
"total_flos": 55447011295232.0,
"train_loss": 0.7642028572467657,
"train_runtime": 2175.16,
"train_samples_per_second": 6.897,
"train_steps_per_second": 0.072
}
],
"logging_steps": 1.0,
"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": 55447011295232.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}