sedrickkeh's picture
End of training
0fbde69 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": 9.227949858397265,
"learning_rate": 6.25e-07,
"loss": 1.3425,
"step": 1
},
{
"epoch": 0.03821656050955414,
"grad_norm": 9.471934233459372,
"learning_rate": 1.25e-06,
"loss": 1.3611,
"step": 2
},
{
"epoch": 0.05732484076433121,
"grad_norm": 8.828033648177298,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.296,
"step": 3
},
{
"epoch": 0.07643312101910828,
"grad_norm": 8.797599540805507,
"learning_rate": 2.5e-06,
"loss": 1.3064,
"step": 4
},
{
"epoch": 0.09554140127388536,
"grad_norm": 8.390617215219814,
"learning_rate": 3.125e-06,
"loss": 1.3085,
"step": 5
},
{
"epoch": 0.11464968152866242,
"grad_norm": 6.299739956541256,
"learning_rate": 3.7500000000000005e-06,
"loss": 1.2614,
"step": 6
},
{
"epoch": 0.1337579617834395,
"grad_norm": 3.917583004029473,
"learning_rate": 4.3750000000000005e-06,
"loss": 1.1772,
"step": 7
},
{
"epoch": 0.15286624203821655,
"grad_norm": 3.4404044686632766,
"learning_rate": 5e-06,
"loss": 1.2101,
"step": 8
},
{
"epoch": 0.17197452229299362,
"grad_norm": 4.044671679258301,
"learning_rate": 5.625e-06,
"loss": 1.123,
"step": 9
},
{
"epoch": 0.1910828025477707,
"grad_norm": 4.245004667468501,
"learning_rate": 6.25e-06,
"loss": 1.0775,
"step": 10
},
{
"epoch": 0.21019108280254778,
"grad_norm": 4.076495305896842,
"learning_rate": 6.875e-06,
"loss": 1.076,
"step": 11
},
{
"epoch": 0.22929936305732485,
"grad_norm": 3.445367429322745,
"learning_rate": 7.500000000000001e-06,
"loss": 1.0192,
"step": 12
},
{
"epoch": 0.2484076433121019,
"grad_norm": 3.8248767958253618,
"learning_rate": 8.125000000000001e-06,
"loss": 1.0478,
"step": 13
},
{
"epoch": 0.267515923566879,
"grad_norm": 3.0594705464578977,
"learning_rate": 8.750000000000001e-06,
"loss": 1.0032,
"step": 14
},
{
"epoch": 0.28662420382165604,
"grad_norm": 2.1936462165160884,
"learning_rate": 9.375000000000001e-06,
"loss": 0.9644,
"step": 15
},
{
"epoch": 0.3057324840764331,
"grad_norm": 2.054476213120208,
"learning_rate": 1e-05,
"loss": 0.958,
"step": 16
},
{
"epoch": 0.3248407643312102,
"grad_norm": 2.527571848112667,
"learning_rate": 9.998741174712534e-06,
"loss": 0.9511,
"step": 17
},
{
"epoch": 0.34394904458598724,
"grad_norm": 2.389506532277148,
"learning_rate": 9.994965332706574e-06,
"loss": 0.9528,
"step": 18
},
{
"epoch": 0.3630573248407643,
"grad_norm": 1.8241708807097363,
"learning_rate": 9.98867437523228e-06,
"loss": 0.9513,
"step": 19
},
{
"epoch": 0.3821656050955414,
"grad_norm": 1.3972460574685222,
"learning_rate": 9.979871469976197e-06,
"loss": 0.9267,
"step": 20
},
{
"epoch": 0.4012738853503185,
"grad_norm": 1.427862950834237,
"learning_rate": 9.968561049466214e-06,
"loss": 0.9383,
"step": 21
},
{
"epoch": 0.42038216560509556,
"grad_norm": 1.4581543509507238,
"learning_rate": 9.954748808839675e-06,
"loss": 0.895,
"step": 22
},
{
"epoch": 0.4394904458598726,
"grad_norm": 1.1560231142924084,
"learning_rate": 9.938441702975689e-06,
"loss": 0.8935,
"step": 23
},
{
"epoch": 0.4585987261146497,
"grad_norm": 0.9991839455389308,
"learning_rate": 9.91964794299315e-06,
"loss": 0.9139,
"step": 24
},
{
"epoch": 0.47770700636942676,
"grad_norm": 1.1127830004999821,
"learning_rate": 9.898376992116179e-06,
"loss": 0.8875,
"step": 25
},
{
"epoch": 0.4968152866242038,
"grad_norm": 1.0986880545284423,
"learning_rate": 9.874639560909118e-06,
"loss": 0.8913,
"step": 26
},
{
"epoch": 0.5159235668789809,
"grad_norm": 1.1342712925409288,
"learning_rate": 9.848447601883436e-06,
"loss": 0.9178,
"step": 27
},
{
"epoch": 0.535031847133758,
"grad_norm": 1.025595920196455,
"learning_rate": 9.819814303479268e-06,
"loss": 0.9223,
"step": 28
},
{
"epoch": 0.554140127388535,
"grad_norm": 1.0257193695613922,
"learning_rate": 9.788754083424654e-06,
"loss": 0.865,
"step": 29
},
{
"epoch": 0.5732484076433121,
"grad_norm": 0.9378584210808578,
"learning_rate": 9.755282581475769e-06,
"loss": 0.873,
"step": 30
},
{
"epoch": 0.5923566878980892,
"grad_norm": 0.9315166489723248,
"learning_rate": 9.719416651541839e-06,
"loss": 0.8962,
"step": 31
},
{
"epoch": 0.6114649681528662,
"grad_norm": 1.010241661377712,
"learning_rate": 9.681174353198687e-06,
"loss": 0.8958,
"step": 32
},
{
"epoch": 0.6305732484076433,
"grad_norm": 0.9956990903313642,
"learning_rate": 9.640574942595195e-06,
"loss": 0.8806,
"step": 33
},
{
"epoch": 0.6496815286624203,
"grad_norm": 0.9717212983720861,
"learning_rate": 9.597638862757255e-06,
"loss": 0.8396,
"step": 34
},
{
"epoch": 0.6687898089171974,
"grad_norm": 0.964287184538465,
"learning_rate": 9.552387733294081e-06,
"loss": 0.8645,
"step": 35
},
{
"epoch": 0.6878980891719745,
"grad_norm": 1.0098642536535931,
"learning_rate": 9.504844339512096e-06,
"loss": 0.8791,
"step": 36
},
{
"epoch": 0.7070063694267515,
"grad_norm": 17.86081132556133,
"learning_rate": 9.45503262094184e-06,
"loss": 0.8776,
"step": 37
},
{
"epoch": 0.7261146496815286,
"grad_norm": 1.3803947675394497,
"learning_rate": 9.40297765928369e-06,
"loss": 0.8929,
"step": 38
},
{
"epoch": 0.7452229299363057,
"grad_norm": 0.9995531007445483,
"learning_rate": 9.348705665778479e-06,
"loss": 0.8563,
"step": 39
},
{
"epoch": 0.7643312101910829,
"grad_norm": 0.8668420690311325,
"learning_rate": 9.292243968009332e-06,
"loss": 0.8045,
"step": 40
},
{
"epoch": 0.7834394904458599,
"grad_norm": 0.9553752052160583,
"learning_rate": 9.233620996141421e-06,
"loss": 0.8777,
"step": 41
},
{
"epoch": 0.802547770700637,
"grad_norm": 0.9112554256937969,
"learning_rate": 9.172866268606514e-06,
"loss": 0.8534,
"step": 42
},
{
"epoch": 0.821656050955414,
"grad_norm": 0.983718380950511,
"learning_rate": 9.110010377239552e-06,
"loss": 0.8874,
"step": 43
},
{
"epoch": 0.8407643312101911,
"grad_norm": 0.9632568835338066,
"learning_rate": 9.045084971874738e-06,
"loss": 0.8453,
"step": 44
},
{
"epoch": 0.8598726114649682,
"grad_norm": 0.9214136807888968,
"learning_rate": 8.978122744408905e-06,
"loss": 0.8592,
"step": 45
},
{
"epoch": 0.8789808917197452,
"grad_norm": 0.9312268909033795,
"learning_rate": 8.90915741234015e-06,
"loss": 0.8298,
"step": 46
},
{
"epoch": 0.8980891719745223,
"grad_norm": 0.8363488314493323,
"learning_rate": 8.838223701790057e-06,
"loss": 0.826,
"step": 47
},
{
"epoch": 0.9171974522292994,
"grad_norm": 1.0227610666623002,
"learning_rate": 8.765357330018056e-06,
"loss": 0.8233,
"step": 48
},
{
"epoch": 0.9363057324840764,
"grad_norm": 0.9744389036399573,
"learning_rate": 8.690594987436705e-06,
"loss": 0.8539,
"step": 49
},
{
"epoch": 0.9554140127388535,
"grad_norm": 0.9461089488478255,
"learning_rate": 8.613974319136959e-06,
"loss": 0.8481,
"step": 50
},
{
"epoch": 0.9745222929936306,
"grad_norm": 1.027801852354281,
"learning_rate": 8.535533905932739e-06,
"loss": 0.8713,
"step": 51
},
{
"epoch": 0.9936305732484076,
"grad_norm": 0.9246906760443965,
"learning_rate": 8.455313244934324e-06,
"loss": 0.8501,
"step": 52
},
{
"epoch": 1.0127388535031847,
"grad_norm": 1.8002608609858541,
"learning_rate": 8.373352729660373e-06,
"loss": 1.279,
"step": 53
},
{
"epoch": 1.0318471337579618,
"grad_norm": 1.104425214684876,
"learning_rate": 8.289693629698564e-06,
"loss": 0.7357,
"step": 54
},
{
"epoch": 1.0509554140127388,
"grad_norm": 0.8755379958278277,
"learning_rate": 8.204378069925121e-06,
"loss": 0.7536,
"step": 55
},
{
"epoch": 1.070063694267516,
"grad_norm": 0.944678195943801,
"learning_rate": 8.117449009293668e-06,
"loss": 0.741,
"step": 56
},
{
"epoch": 1.089171974522293,
"grad_norm": 0.9124994052742826,
"learning_rate": 8.0289502192041e-06,
"loss": 0.6967,
"step": 57
},
{
"epoch": 1.10828025477707,
"grad_norm": 1.0291869185814695,
"learning_rate": 7.938926261462366e-06,
"loss": 0.7194,
"step": 58
},
{
"epoch": 1.127388535031847,
"grad_norm": 0.8698891738699522,
"learning_rate": 7.84742246584226e-06,
"loss": 0.7413,
"step": 59
},
{
"epoch": 1.1464968152866242,
"grad_norm": 0.8481567203427745,
"learning_rate": 7.754484907260513e-06,
"loss": 0.6827,
"step": 60
},
{
"epoch": 1.1656050955414012,
"grad_norm": 0.8829604306329443,
"learning_rate": 7.660160382576683e-06,
"loss": 0.7579,
"step": 61
},
{
"epoch": 1.1847133757961783,
"grad_norm": 0.874735246202617,
"learning_rate": 7.564496387029532e-06,
"loss": 0.6857,
"step": 62
},
{
"epoch": 1.2038216560509554,
"grad_norm": 0.8541570181813871,
"learning_rate": 7.467541090321735e-06,
"loss": 0.7391,
"step": 63
},
{
"epoch": 1.2229299363057324,
"grad_norm": 0.930997953704132,
"learning_rate": 7.369343312364994e-06,
"loss": 0.6692,
"step": 64
},
{
"epoch": 1.2420382165605095,
"grad_norm": 0.9595741684205553,
"learning_rate": 7.269952498697734e-06,
"loss": 0.7451,
"step": 65
},
{
"epoch": 1.2611464968152866,
"grad_norm": 0.902874862862275,
"learning_rate": 7.169418695587791e-06,
"loss": 0.7208,
"step": 66
},
{
"epoch": 1.2802547770700636,
"grad_norm": 0.8653329745133199,
"learning_rate": 7.067792524832604e-06,
"loss": 0.7428,
"step": 67
},
{
"epoch": 1.2993630573248407,
"grad_norm": 0.8665137973627551,
"learning_rate": 6.965125158269619e-06,
"loss": 0.6926,
"step": 68
},
{
"epoch": 1.3184713375796178,
"grad_norm": 0.8972212012244402,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.752,
"step": 69
},
{
"epoch": 1.3375796178343948,
"grad_norm": 0.801712780729876,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.6728,
"step": 70
},
{
"epoch": 1.356687898089172,
"grad_norm": 0.9336216847911322,
"learning_rate": 6.651395309775837e-06,
"loss": 0.7117,
"step": 71
},
{
"epoch": 1.3757961783439492,
"grad_norm": 0.7616924636789059,
"learning_rate": 6.545084971874738e-06,
"loss": 0.687,
"step": 72
},
{
"epoch": 1.394904458598726,
"grad_norm": 0.7910574988690745,
"learning_rate": 6.437996637160086e-06,
"loss": 0.7004,
"step": 73
},
{
"epoch": 1.4140127388535033,
"grad_norm": 0.8126832517528152,
"learning_rate": 6.330184227833376e-06,
"loss": 0.6727,
"step": 74
},
{
"epoch": 1.4331210191082802,
"grad_norm": 0.8246493231827546,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.6771,
"step": 75
},
{
"epoch": 1.4522292993630574,
"grad_norm": 0.7442810518836369,
"learning_rate": 6.112604669781572e-06,
"loss": 0.7058,
"step": 76
},
{
"epoch": 1.4713375796178343,
"grad_norm": 0.805922567682453,
"learning_rate": 6.002947078916365e-06,
"loss": 0.7452,
"step": 77
},
{
"epoch": 1.4904458598726116,
"grad_norm": 0.7774303654316796,
"learning_rate": 5.892784473993184e-06,
"loss": 0.673,
"step": 78
},
{
"epoch": 1.5095541401273884,
"grad_norm": 0.7578613322345805,
"learning_rate": 5.782172325201155e-06,
"loss": 0.6638,
"step": 79
},
{
"epoch": 1.5286624203821657,
"grad_norm": 0.7992920894405023,
"learning_rate": 5.671166329088278e-06,
"loss": 0.7015,
"step": 80
},
{
"epoch": 1.5477707006369426,
"grad_norm": 0.7868566279868858,
"learning_rate": 5.559822380516539e-06,
"loss": 0.6952,
"step": 81
},
{
"epoch": 1.5668789808917198,
"grad_norm": 0.7860362301931151,
"learning_rate": 5.448196544517168e-06,
"loss": 0.7219,
"step": 82
},
{
"epoch": 1.5859872611464967,
"grad_norm": 0.8363775013607909,
"learning_rate": 5.336345028060199e-06,
"loss": 0.7374,
"step": 83
},
{
"epoch": 1.605095541401274,
"grad_norm": 0.7776989433227444,
"learning_rate": 5.224324151752575e-06,
"loss": 0.6996,
"step": 84
},
{
"epoch": 1.6242038216560508,
"grad_norm": 0.8143329515718004,
"learning_rate": 5.112190321479026e-06,
"loss": 0.7033,
"step": 85
},
{
"epoch": 1.643312101910828,
"grad_norm": 0.7526134072866455,
"learning_rate": 5e-06,
"loss": 0.6665,
"step": 86
},
{
"epoch": 1.662420382165605,
"grad_norm": 0.8063339804591604,
"learning_rate": 4.887809678520976e-06,
"loss": 0.7533,
"step": 87
},
{
"epoch": 1.6815286624203822,
"grad_norm": 0.8376637502903548,
"learning_rate": 4.775675848247427e-06,
"loss": 0.7148,
"step": 88
},
{
"epoch": 1.700636942675159,
"grad_norm": 0.8424926705426312,
"learning_rate": 4.663654971939802e-06,
"loss": 0.6803,
"step": 89
},
{
"epoch": 1.7197452229299364,
"grad_norm": 0.7492913461032326,
"learning_rate": 4.551803455482833e-06,
"loss": 0.72,
"step": 90
},
{
"epoch": 1.7388535031847132,
"grad_norm": 0.7930358634705931,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.7042,
"step": 91
},
{
"epoch": 1.7579617834394905,
"grad_norm": 0.8091448675999294,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.7448,
"step": 92
},
{
"epoch": 1.7770700636942676,
"grad_norm": 0.7688368111174908,
"learning_rate": 4.217827674798845e-06,
"loss": 0.6808,
"step": 93
},
{
"epoch": 1.7961783439490446,
"grad_norm": 0.74060506091446,
"learning_rate": 4.107215526006818e-06,
"loss": 0.6567,
"step": 94
},
{
"epoch": 1.8152866242038217,
"grad_norm": 0.7489349419875532,
"learning_rate": 3.997052921083637e-06,
"loss": 0.7403,
"step": 95
},
{
"epoch": 1.8343949044585988,
"grad_norm": 0.7844313968278532,
"learning_rate": 3.887395330218429e-06,
"loss": 0.7019,
"step": 96
},
{
"epoch": 1.8535031847133758,
"grad_norm": 0.765432277156269,
"learning_rate": 3.778297969310529e-06,
"loss": 0.7027,
"step": 97
},
{
"epoch": 1.872611464968153,
"grad_norm": 0.8675734268129139,
"learning_rate": 3.669815772166625e-06,
"loss": 0.7224,
"step": 98
},
{
"epoch": 1.89171974522293,
"grad_norm": 0.788236710989505,
"learning_rate": 3.562003362839914e-06,
"loss": 0.6862,
"step": 99
},
{
"epoch": 1.910828025477707,
"grad_norm": 0.8422941453985829,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.6405,
"step": 100
},
{
"epoch": 1.929936305732484,
"grad_norm": 0.8342255786473474,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.7607,
"step": 101
},
{
"epoch": 1.9490445859872612,
"grad_norm": 0.782685386156924,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.6784,
"step": 102
},
{
"epoch": 1.9681528662420382,
"grad_norm": 0.6840466342699705,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.691,
"step": 103
},
{
"epoch": 1.9872611464968153,
"grad_norm": 0.803403451260333,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.6872,
"step": 104
},
{
"epoch": 2.0063694267515926,
"grad_norm": 1.6157144876234706,
"learning_rate": 2.932207475167398e-06,
"loss": 1.0976,
"step": 105
},
{
"epoch": 2.0254777070063694,
"grad_norm": 0.852275820888973,
"learning_rate": 2.83058130441221e-06,
"loss": 0.6557,
"step": 106
},
{
"epoch": 2.0445859872611467,
"grad_norm": 0.8343073617133979,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.5636,
"step": 107
},
{
"epoch": 2.0636942675159236,
"grad_norm": 0.7273900448935333,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.6008,
"step": 108
},
{
"epoch": 2.082802547770701,
"grad_norm": 0.6819154605325781,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5766,
"step": 109
},
{
"epoch": 2.1019108280254777,
"grad_norm": 0.7223666158643476,
"learning_rate": 2.43550361297047e-06,
"loss": 0.6301,
"step": 110
},
{
"epoch": 2.121019108280255,
"grad_norm": 0.7725441886194817,
"learning_rate": 2.339839617423318e-06,
"loss": 0.6335,
"step": 111
},
{
"epoch": 2.140127388535032,
"grad_norm": 0.7433133843658496,
"learning_rate": 2.245515092739488e-06,
"loss": 0.6019,
"step": 112
},
{
"epoch": 2.159235668789809,
"grad_norm": 0.8213061651578086,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.6381,
"step": 113
},
{
"epoch": 2.178343949044586,
"grad_norm": 0.7325568761495765,
"learning_rate": 2.061073738537635e-06,
"loss": 0.565,
"step": 114
},
{
"epoch": 2.1974522292993632,
"grad_norm": 0.6846757845178079,
"learning_rate": 1.971049780795901e-06,
"loss": 0.5821,
"step": 115
},
{
"epoch": 2.21656050955414,
"grad_norm": 0.7019485070717555,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.5967,
"step": 116
},
{
"epoch": 2.2356687898089174,
"grad_norm": 0.731886553104978,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.6303,
"step": 117
},
{
"epoch": 2.254777070063694,
"grad_norm": 0.6811732391214775,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.5701,
"step": 118
},
{
"epoch": 2.2738853503184715,
"grad_norm": 0.7183445882743708,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.6089,
"step": 119
},
{
"epoch": 2.2929936305732483,
"grad_norm": 0.7415481076121242,
"learning_rate": 1.544686755065677e-06,
"loss": 0.6253,
"step": 120
},
{
"epoch": 2.3121019108280256,
"grad_norm": 0.7407289705202259,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.591,
"step": 121
},
{
"epoch": 2.3312101910828025,
"grad_norm": 0.7442606057030374,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.5855,
"step": 122
},
{
"epoch": 2.3503184713375798,
"grad_norm": 0.7207012766861567,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.612,
"step": 123
},
{
"epoch": 2.3694267515923566,
"grad_norm": 0.6547390379943093,
"learning_rate": 1.234642669981946e-06,
"loss": 0.5567,
"step": 124
},
{
"epoch": 2.388535031847134,
"grad_norm": 0.7443095002023548,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.6366,
"step": 125
},
{
"epoch": 2.4076433121019107,
"grad_norm": 0.6799775133199653,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5655,
"step": 126
},
{
"epoch": 2.426751592356688,
"grad_norm": 0.7031449550648861,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.6048,
"step": 127
},
{
"epoch": 2.445859872611465,
"grad_norm": 0.733679448941415,
"learning_rate": 9.549150281252633e-07,
"loss": 0.6128,
"step": 128
},
{
"epoch": 2.464968152866242,
"grad_norm": 0.7665247991159213,
"learning_rate": 8.899896227604509e-07,
"loss": 0.6091,
"step": 129
},
{
"epoch": 2.484076433121019,
"grad_norm": 0.6703229720063976,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5487,
"step": 130
},
{
"epoch": 2.5031847133757963,
"grad_norm": 0.7311778842991127,
"learning_rate": 7.663790038585794e-07,
"loss": 0.6829,
"step": 131
},
{
"epoch": 2.522292993630573,
"grad_norm": 0.663709876474419,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5689,
"step": 132
},
{
"epoch": 2.5414012738853504,
"grad_norm": 0.7163372829050498,
"learning_rate": 6.512943342215234e-07,
"loss": 0.6079,
"step": 133
},
{
"epoch": 2.5605095541401273,
"grad_norm": 0.6491561575546041,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5572,
"step": 134
},
{
"epoch": 2.5796178343949046,
"grad_norm": 0.6877575239338882,
"learning_rate": 5.449673790581611e-07,
"loss": 0.5882,
"step": 135
},
{
"epoch": 2.5987261146496814,
"grad_norm": 0.6739227673483531,
"learning_rate": 4.951556604879049e-07,
"loss": 0.5862,
"step": 136
},
{
"epoch": 2.6178343949044587,
"grad_norm": 0.697740584151181,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5934,
"step": 137
},
{
"epoch": 2.6369426751592355,
"grad_norm": 0.6925027003503761,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.6019,
"step": 138
},
{
"epoch": 2.656050955414013,
"grad_norm": 0.660260708802265,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5926,
"step": 139
},
{
"epoch": 2.6751592356687897,
"grad_norm": 0.6867716462291468,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5823,
"step": 140
},
{
"epoch": 2.694267515923567,
"grad_norm": 0.6859078885829797,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.6171,
"step": 141
},
{
"epoch": 2.713375796178344,
"grad_norm": 0.6869630276719553,
"learning_rate": 2.447174185242324e-07,
"loss": 0.6246,
"step": 142
},
{
"epoch": 2.732484076433121,
"grad_norm": 0.6841228763448503,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.6079,
"step": 143
},
{
"epoch": 2.7515923566878984,
"grad_norm": 0.6689340577792746,
"learning_rate": 1.801856965207338e-07,
"loss": 0.5467,
"step": 144
},
{
"epoch": 2.770700636942675,
"grad_norm": 0.6604828892682805,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5784,
"step": 145
},
{
"epoch": 2.789808917197452,
"grad_norm": 0.6930116287882491,
"learning_rate": 1.253604390908819e-07,
"loss": 0.6159,
"step": 146
},
{
"epoch": 2.8089171974522293,
"grad_norm": 0.7142794142835408,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.6437,
"step": 147
},
{
"epoch": 2.8280254777070066,
"grad_norm": 0.6456650864170574,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5587,
"step": 148
},
{
"epoch": 2.8471337579617835,
"grad_norm": 0.624531561549103,
"learning_rate": 6.15582970243117e-08,
"loss": 0.5938,
"step": 149
},
{
"epoch": 2.8662420382165603,
"grad_norm": 0.664828417388466,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5849,
"step": 150
},
{
"epoch": 2.8853503184713376,
"grad_norm": 0.6550255176830048,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5988,
"step": 151
},
{
"epoch": 2.904458598726115,
"grad_norm": 0.702301375505477,
"learning_rate": 2.012853002380466e-08,
"loss": 0.573,
"step": 152
},
{
"epoch": 2.9235668789808917,
"grad_norm": 0.7092391869858767,
"learning_rate": 1.132562476771959e-08,
"loss": 0.654,
"step": 153
},
{
"epoch": 2.9426751592356686,
"grad_norm": 0.6600741685715265,
"learning_rate": 5.034667293427053e-09,
"loss": 0.5779,
"step": 154
},
{
"epoch": 2.961783439490446,
"grad_norm": 0.6755862775388255,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5806,
"step": 155
},
{
"epoch": 2.980891719745223,
"grad_norm": 0.6525314928129297,
"learning_rate": 0.0,
"loss": 0.6078,
"step": 156
},
{
"epoch": 2.980891719745223,
"step": 156,
"total_flos": 6.084353864735949e+16,
"train_loss": 0.7641331553459167,
"train_runtime": 1876.8154,
"train_samples_per_second": 7.994,
"train_steps_per_second": 0.083
}
],
"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": 6.084353864735949e+16,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}