Files
qwen2.5-1.5b-slips-immune-u…/lora_adapter/checkpoint-267/trainer_state.json
ModelHub XC a1080453d5 初始化项目,由ModelHub XC社区提供模型
Model: stratosphere/qwen2.5-1.5b-slips-immune-unified
Source: Original Platform
2026-07-26 13:32:07 +08:00

1952 lines
50 KiB
JSON

{
"best_global_step": 250,
"best_metric": 0.7293056845664978,
"best_model_checkpoint": "./qwen_unified_finetuned/checkpoint-250",
"epoch": 3.0,
"eval_steps": 50,
"global_step": 267,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0113555713271824,
"grad_norm": 3.916877031326294,
"learning_rate": 0.0,
"loss": 1.048294186592102,
"step": 1
},
{
"epoch": 0.0227111426543648,
"grad_norm": 4.441281318664551,
"learning_rate": 6.666666666666667e-07,
"loss": 0.8231856822967529,
"step": 2
},
{
"epoch": 0.034066713981547196,
"grad_norm": 4.950929641723633,
"learning_rate": 1.3333333333333334e-06,
"loss": 1.31972074508667,
"step": 3
},
{
"epoch": 0.0454222853087296,
"grad_norm": 4.647597789764404,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.1833959817886353,
"step": 4
},
{
"epoch": 0.05677785663591199,
"grad_norm": 4.787125110626221,
"learning_rate": 2.666666666666667e-06,
"loss": 1.5079869031906128,
"step": 5
},
{
"epoch": 0.06813342796309439,
"grad_norm": 4.424062252044678,
"learning_rate": 3.3333333333333333e-06,
"loss": 1.0498075485229492,
"step": 6
},
{
"epoch": 0.0794889992902768,
"grad_norm": 4.251436710357666,
"learning_rate": 4.000000000000001e-06,
"loss": 1.0172538757324219,
"step": 7
},
{
"epoch": 0.0908445706174592,
"grad_norm": 3.5152761936187744,
"learning_rate": 4.666666666666667e-06,
"loss": 1.1070594787597656,
"step": 8
},
{
"epoch": 0.1022001419446416,
"grad_norm": 3.2286136150360107,
"learning_rate": 5.333333333333334e-06,
"loss": 0.8639000058174133,
"step": 9
},
{
"epoch": 0.11355571327182398,
"grad_norm": 2.9262876510620117,
"learning_rate": 6e-06,
"loss": 0.927219033241272,
"step": 10
},
{
"epoch": 0.12491128459900638,
"grad_norm": 3.439363956451416,
"learning_rate": 6.666666666666667e-06,
"loss": 1.2999275922775269,
"step": 11
},
{
"epoch": 0.13626685592618878,
"grad_norm": 2.8811707496643066,
"learning_rate": 7.333333333333333e-06,
"loss": 1.0791280269622803,
"step": 12
},
{
"epoch": 0.14762242725337119,
"grad_norm": 2.5855603218078613,
"learning_rate": 8.000000000000001e-06,
"loss": 1.0025461912155151,
"step": 13
},
{
"epoch": 0.1589779985805536,
"grad_norm": 2.93908953666687,
"learning_rate": 8.666666666666668e-06,
"loss": 1.2033611536026,
"step": 14
},
{
"epoch": 0.170333569907736,
"grad_norm": 2.1815803050994873,
"learning_rate": 9.333333333333334e-06,
"loss": 0.865071177482605,
"step": 15
},
{
"epoch": 0.1816891412349184,
"grad_norm": 2.5811753273010254,
"learning_rate": 1e-05,
"loss": 0.9821955561637878,
"step": 16
},
{
"epoch": 0.1930447125621008,
"grad_norm": 2.2700631618499756,
"learning_rate": 1.0666666666666667e-05,
"loss": 0.943819522857666,
"step": 17
},
{
"epoch": 0.2044002838892832,
"grad_norm": 2.234851360321045,
"learning_rate": 1.1333333333333334e-05,
"loss": 0.8833191990852356,
"step": 18
},
{
"epoch": 0.2157558552164656,
"grad_norm": 2.1426126956939697,
"learning_rate": 1.2e-05,
"loss": 0.8689224720001221,
"step": 19
},
{
"epoch": 0.22711142654364797,
"grad_norm": 2.0993337631225586,
"learning_rate": 1.2666666666666667e-05,
"loss": 0.8528774380683899,
"step": 20
},
{
"epoch": 0.23846699787083037,
"grad_norm": 1.9766696691513062,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.889632523059845,
"step": 21
},
{
"epoch": 0.24982256919801277,
"grad_norm": 2.3761353492736816,
"learning_rate": 1.4e-05,
"loss": 0.9641993045806885,
"step": 22
},
{
"epoch": 0.26117814052519517,
"grad_norm": 2.226921558380127,
"learning_rate": 1.4666666666666666e-05,
"loss": 1.0349667072296143,
"step": 23
},
{
"epoch": 0.27253371185237757,
"grad_norm": 1.9597960710525513,
"learning_rate": 1.5333333333333334e-05,
"loss": 0.8142374157905579,
"step": 24
},
{
"epoch": 0.28388928317955997,
"grad_norm": 1.769181728363037,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.6785438656806946,
"step": 25
},
{
"epoch": 0.29524485450674237,
"grad_norm": 2.046727418899536,
"learning_rate": 1.6666666666666667e-05,
"loss": 0.9442430734634399,
"step": 26
},
{
"epoch": 0.3066004258339248,
"grad_norm": 2.2059895992279053,
"learning_rate": 1.7333333333333336e-05,
"loss": 1.0432536602020264,
"step": 27
},
{
"epoch": 0.3179559971611072,
"grad_norm": 2.3628499507904053,
"learning_rate": 1.8e-05,
"loss": 1.0692076683044434,
"step": 28
},
{
"epoch": 0.3293115684882896,
"grad_norm": 2.2492411136627197,
"learning_rate": 1.866666666666667e-05,
"loss": 1.1659228801727295,
"step": 29
},
{
"epoch": 0.340667139815472,
"grad_norm": 2.217177152633667,
"learning_rate": 1.9333333333333333e-05,
"loss": 0.706802487373352,
"step": 30
},
{
"epoch": 0.3520227111426544,
"grad_norm": 1.6823749542236328,
"learning_rate": 2e-05,
"loss": 0.7545626163482666,
"step": 31
},
{
"epoch": 0.3633782824698368,
"grad_norm": 2.0650510787963867,
"learning_rate": 1.9999121449564347e-05,
"loss": 0.8590264320373535,
"step": 32
},
{
"epoch": 0.3747338537970192,
"grad_norm": 1.6296216249465942,
"learning_rate": 1.9996485952627554e-05,
"loss": 0.7331938743591309,
"step": 33
},
{
"epoch": 0.3860894251242016,
"grad_norm": 1.7522505521774292,
"learning_rate": 1.999209397227302e-05,
"loss": 0.7665808796882629,
"step": 34
},
{
"epoch": 0.397444996451384,
"grad_norm": 2.035996198654175,
"learning_rate": 1.9985946280215996e-05,
"loss": 0.9990826845169067,
"step": 35
},
{
"epoch": 0.4088005677785664,
"grad_norm": 1.993740200996399,
"learning_rate": 1.997804395666799e-05,
"loss": 0.8873282074928284,
"step": 36
},
{
"epoch": 0.4201561391057488,
"grad_norm": 1.9349602460861206,
"learning_rate": 1.996838839014696e-05,
"loss": 0.8252230286598206,
"step": 37
},
{
"epoch": 0.4315117104329312,
"grad_norm": 2.1690878868103027,
"learning_rate": 1.9956981277233342e-05,
"loss": 1.0567874908447266,
"step": 38
},
{
"epoch": 0.44286728176011353,
"grad_norm": 2.005312442779541,
"learning_rate": 1.9943824622271934e-05,
"loss": 0.980779767036438,
"step": 39
},
{
"epoch": 0.45422285308729593,
"grad_norm": 2.0727038383483887,
"learning_rate": 1.9928920737019735e-05,
"loss": 0.8422978520393372,
"step": 40
},
{
"epoch": 0.46557842441447833,
"grad_norm": 2.4477312564849854,
"learning_rate": 1.9912272240239715e-05,
"loss": 1.239004373550415,
"step": 41
},
{
"epoch": 0.47693399574166073,
"grad_norm": 2.073049306869507,
"learning_rate": 1.9893882057240698e-05,
"loss": 1.0170605182647705,
"step": 42
},
{
"epoch": 0.48828956706884313,
"grad_norm": 1.9096711874008179,
"learning_rate": 1.9873753419363336e-05,
"loss": 0.7338559031486511,
"step": 43
},
{
"epoch": 0.49964513839602553,
"grad_norm": 1.514130711555481,
"learning_rate": 1.9851889863412347e-05,
"loss": 0.5721548795700073,
"step": 44
},
{
"epoch": 0.511000709723208,
"grad_norm": 1.8837509155273438,
"learning_rate": 1.9828295231035054e-05,
"loss": 0.7944353222846985,
"step": 45
},
{
"epoch": 0.5223562810503903,
"grad_norm": 1.6668717861175537,
"learning_rate": 1.9802973668046364e-05,
"loss": 0.703384518623352,
"step": 46
},
{
"epoch": 0.5337118523775728,
"grad_norm": 1.619913935661316,
"learning_rate": 1.9775929623700318e-05,
"loss": 0.7028679847717285,
"step": 47
},
{
"epoch": 0.5450674237047551,
"grad_norm": 1.797451138496399,
"learning_rate": 1.9747167849908305e-05,
"loss": 0.8451980948448181,
"step": 48
},
{
"epoch": 0.5564229950319376,
"grad_norm": 1.6591897010803223,
"learning_rate": 1.97166934004041e-05,
"loss": 0.6273907423019409,
"step": 49
},
{
"epoch": 0.5677785663591199,
"grad_norm": 1.6978833675384521,
"learning_rate": 1.968451162985589e-05,
"loss": 0.6259630918502808,
"step": 50
},
{
"epoch": 0.5677785663591199,
"eval_loss": 0.8047120571136475,
"eval_runtime": 27.6692,
"eval_samples_per_second": 5.168,
"eval_steps_per_second": 1.301,
"step": 50
},
{
"epoch": 0.5791341376863024,
"grad_norm": 2.0382018089294434,
"learning_rate": 1.9650628192925372e-05,
"loss": 0.8624173402786255,
"step": 51
},
{
"epoch": 0.5904897090134847,
"grad_norm": 2.2218334674835205,
"learning_rate": 1.9615049043274207e-05,
"loss": 1.0386168956756592,
"step": 52
},
{
"epoch": 0.6018452803406671,
"grad_norm": 1.7063782215118408,
"learning_rate": 1.957778043251788e-05,
"loss": 0.7194206118583679,
"step": 53
},
{
"epoch": 0.6132008516678495,
"grad_norm": 1.621178150177002,
"learning_rate": 1.953882890912723e-05,
"loss": 0.804326593875885,
"step": 54
},
{
"epoch": 0.6245564229950319,
"grad_norm": 1.5988848209381104,
"learning_rate": 1.949820131727783e-05,
"loss": 0.6942477822303772,
"step": 55
},
{
"epoch": 0.6359119943222143,
"grad_norm": 1.7638858556747437,
"learning_rate": 1.945590479564738e-05,
"loss": 0.688396692276001,
"step": 56
},
{
"epoch": 0.6472675656493967,
"grad_norm": 1.8592664003372192,
"learning_rate": 1.9411946776161388e-05,
"loss": 0.7517380118370056,
"step": 57
},
{
"epoch": 0.6586231369765791,
"grad_norm": 2.1580257415771484,
"learning_rate": 1.936633498268728e-05,
"loss": 1.1234959363937378,
"step": 58
},
{
"epoch": 0.6699787083037615,
"grad_norm": 2.0941476821899414,
"learning_rate": 1.931907742967727e-05,
"loss": 0.7969536185264587,
"step": 59
},
{
"epoch": 0.681334279630944,
"grad_norm": 1.8826228380203247,
"learning_rate": 1.9270182420760104e-05,
"loss": 0.7057633996009827,
"step": 60
},
{
"epoch": 0.6926898509581263,
"grad_norm": 1.8172886371612549,
"learning_rate": 1.921965854728207e-05,
"loss": 0.8432231545448303,
"step": 61
},
{
"epoch": 0.7040454222853088,
"grad_norm": 2.468703508377075,
"learning_rate": 1.916751468679737e-05,
"loss": 0.9716429710388184,
"step": 62
},
{
"epoch": 0.7154009936124911,
"grad_norm": 1.5727559328079224,
"learning_rate": 1.911376000150828e-05,
"loss": 0.6017226576805115,
"step": 63
},
{
"epoch": 0.7267565649396736,
"grad_norm": 1.9279130697250366,
"learning_rate": 1.9058403936655235e-05,
"loss": 0.945792019367218,
"step": 64
},
{
"epoch": 0.7381121362668559,
"grad_norm": 1.459025502204895,
"learning_rate": 1.9001456218857207e-05,
"loss": 0.601871907711029,
"step": 65
},
{
"epoch": 0.7494677075940384,
"grad_norm": 1.5340495109558105,
"learning_rate": 1.894292685440266e-05,
"loss": 0.6935733556747437,
"step": 66
},
{
"epoch": 0.7608232789212207,
"grad_norm": 1.6071991920471191,
"learning_rate": 1.888282612749132e-05,
"loss": 0.7782894968986511,
"step": 67
},
{
"epoch": 0.7721788502484032,
"grad_norm": 1.6310977935791016,
"learning_rate": 1.8821164598427148e-05,
"loss": 0.7309603095054626,
"step": 68
},
{
"epoch": 0.7835344215755855,
"grad_norm": 1.817807912826538,
"learning_rate": 1.8757953101762786e-05,
"loss": 0.790179967880249,
"step": 69
},
{
"epoch": 0.794889992902768,
"grad_norm": 1.3777976036071777,
"learning_rate": 1.869320274439583e-05,
"loss": 0.5821739435195923,
"step": 70
},
{
"epoch": 0.8062455642299503,
"grad_norm": 1.3368778228759766,
"learning_rate": 1.86269249036172e-05,
"loss": 0.5191158652305603,
"step": 71
},
{
"epoch": 0.8176011355571328,
"grad_norm": 1.7497543096542358,
"learning_rate": 1.8559131225112085e-05,
"loss": 0.7746997475624084,
"step": 72
},
{
"epoch": 0.8289567068843151,
"grad_norm": 1.6474581956863403,
"learning_rate": 1.8489833620913644e-05,
"loss": 0.5990749001502991,
"step": 73
},
{
"epoch": 0.8403122782114976,
"grad_norm": 1.425794005393982,
"learning_rate": 1.841904426730994e-05,
"loss": 0.6433415412902832,
"step": 74
},
{
"epoch": 0.8516678495386799,
"grad_norm": 1.8670772314071655,
"learning_rate": 1.8346775602704464e-05,
"loss": 0.840431809425354,
"step": 75
},
{
"epoch": 0.8630234208658624,
"grad_norm": 1.6109037399291992,
"learning_rate": 1.8273040325430575e-05,
"loss": 0.7096354961395264,
"step": 76
},
{
"epoch": 0.8743789921930447,
"grad_norm": 1.964331865310669,
"learning_rate": 1.8197851391520265e-05,
"loss": 0.7453011274337769,
"step": 77
},
{
"epoch": 0.8857345635202271,
"grad_norm": 1.7134088277816772,
"learning_rate": 1.8121222012427666e-05,
"loss": 0.7802327871322632,
"step": 78
},
{
"epoch": 0.8970901348474095,
"grad_norm": 1.964368224143982,
"learning_rate": 1.804316565270765e-05,
"loss": 0.8367159366607666,
"step": 79
},
{
"epoch": 0.9084457061745919,
"grad_norm": 1.9737969636917114,
"learning_rate": 1.796369602764999e-05,
"loss": 0.8365910053253174,
"step": 80
},
{
"epoch": 0.9198012775017743,
"grad_norm": 1.963098168373108,
"learning_rate": 1.788282710086942e-05,
"loss": 0.9032438397407532,
"step": 81
},
{
"epoch": 0.9311568488289567,
"grad_norm": 1.771173119544983,
"learning_rate": 1.7800573081852124e-05,
"loss": 0.7875261902809143,
"step": 82
},
{
"epoch": 0.9425124201561391,
"grad_norm": 1.9102442264556885,
"learning_rate": 1.771694842345894e-05,
"loss": 0.76175856590271,
"step": 83
},
{
"epoch": 0.9538679914833215,
"grad_norm": 2.03609299659729,
"learning_rate": 1.7631967819385883e-05,
"loss": 0.8382056951522827,
"step": 84
},
{
"epoch": 0.9652235628105039,
"grad_norm": 1.864434003829956,
"learning_rate": 1.7545646201582304e-05,
"loss": 0.785956084728241,
"step": 85
},
{
"epoch": 0.9765791341376863,
"grad_norm": 2.0304460525512695,
"learning_rate": 1.7457998737627183e-05,
"loss": 0.9167435169219971,
"step": 86
},
{
"epoch": 0.9879347054648687,
"grad_norm": 1.691648006439209,
"learning_rate": 1.7369040828064046e-05,
"loss": 0.7328086495399475,
"step": 87
},
{
"epoch": 0.9992902767920511,
"grad_norm": 2.0275533199310303,
"learning_rate": 1.7278788103694944e-05,
"loss": 0.9172717928886414,
"step": 88
},
{
"epoch": 1.0,
"grad_norm": 2.6658289432525635,
"learning_rate": 1.7187256422833928e-05,
"loss": 0.11026950925588608,
"step": 89
},
{
"epoch": 1.0113555713271825,
"grad_norm": 1.5766886472702026,
"learning_rate": 1.7094461868520625e-05,
"loss": 0.7454363107681274,
"step": 90
},
{
"epoch": 1.0227111426543647,
"grad_norm": 1.934645652770996,
"learning_rate": 1.7000420745694256e-05,
"loss": 0.9695757627487183,
"step": 91
},
{
"epoch": 1.0340667139815471,
"grad_norm": 1.7055552005767822,
"learning_rate": 1.6905149578328705e-05,
"loss": 0.7050992846488953,
"step": 92
},
{
"epoch": 1.0454222853087296,
"grad_norm": 1.5468671321868896,
"learning_rate": 1.6808665106529096e-05,
"loss": 0.6628597974777222,
"step": 93
},
{
"epoch": 1.056777856635912,
"grad_norm": 1.80152428150177,
"learning_rate": 1.671098428359037e-05,
"loss": 0.826714277267456,
"step": 94
},
{
"epoch": 1.0681334279630943,
"grad_norm": 1.9063488245010376,
"learning_rate": 1.661212427301844e-05,
"loss": 0.9870727062225342,
"step": 95
},
{
"epoch": 1.0794889992902768,
"grad_norm": 1.7208439111709595,
"learning_rate": 1.6512102445514376e-05,
"loss": 0.754402220249176,
"step": 96
},
{
"epoch": 1.0908445706174592,
"grad_norm": 1.688193440437317,
"learning_rate": 1.64109363759222e-05,
"loss": 0.6640122532844543,
"step": 97
},
{
"epoch": 1.1022001419446417,
"grad_norm": 1.8576338291168213,
"learning_rate": 1.630864384014083e-05,
"loss": 0.8575501441955566,
"step": 98
},
{
"epoch": 1.113555713271824,
"grad_norm": 1.7428895235061646,
"learning_rate": 1.620524281200062e-05,
"loss": 0.5853228569030762,
"step": 99
},
{
"epoch": 1.1249112845990064,
"grad_norm": 1.5940732955932617,
"learning_rate": 1.6100751460105244e-05,
"loss": 0.6318332552909851,
"step": 100
},
{
"epoch": 1.1249112845990064,
"eval_loss": 0.7593621015548706,
"eval_runtime": 27.6141,
"eval_samples_per_second": 5.179,
"eval_steps_per_second": 1.304,
"step": 100
},
{
"epoch": 1.1362668559261888,
"grad_norm": 1.9089066982269287,
"learning_rate": 1.599518814463925e-05,
"loss": 0.9151409268379211,
"step": 101
},
{
"epoch": 1.1476224272533713,
"grad_norm": 1.905922770500183,
"learning_rate": 1.5888571414141997e-05,
"loss": 0.8580778241157532,
"step": 102
},
{
"epoch": 1.1589779985805535,
"grad_norm": 1.964565634727478,
"learning_rate": 1.5780920002248484e-05,
"loss": 0.8872693777084351,
"step": 103
},
{
"epoch": 1.170333569907736,
"grad_norm": 2.039443254470825,
"learning_rate": 1.5672252824397683e-05,
"loss": 0.7390342950820923,
"step": 104
},
{
"epoch": 1.1816891412349184,
"grad_norm": 1.685909390449524,
"learning_rate": 1.556258897450887e-05,
"loss": 0.6857212781906128,
"step": 105
},
{
"epoch": 1.1930447125621009,
"grad_norm": 1.9805678129196167,
"learning_rate": 1.5451947721626676e-05,
"loss": 0.685741662979126,
"step": 106
},
{
"epoch": 1.204400283889283,
"grad_norm": 1.6393771171569824,
"learning_rate": 1.534034850653528e-05,
"loss": 0.6081647276878357,
"step": 107
},
{
"epoch": 1.2157558552164656,
"grad_norm": 1.371028184890747,
"learning_rate": 1.5227810938342493e-05,
"loss": 0.44291260838508606,
"step": 108
},
{
"epoch": 1.227111426543648,
"grad_norm": 1.8976811170578003,
"learning_rate": 1.5114354791034225e-05,
"loss": 0.82900071144104,
"step": 109
},
{
"epoch": 1.2384669978708303,
"grad_norm": 1.4269609451293945,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.44125357270240784,
"step": 110
},
{
"epoch": 1.2498225691980127,
"grad_norm": 1.9113351106643677,
"learning_rate": 1.4884766658530126e-05,
"loss": 0.7995925545692444,
"step": 111
},
{
"epoch": 1.2611781405251952,
"grad_norm": 1.9662408828735352,
"learning_rate": 1.4768675014285063e-05,
"loss": 0.8006548881530762,
"step": 112
},
{
"epoch": 1.2725337118523776,
"grad_norm": 2.118868350982666,
"learning_rate": 1.465174546573774e-05,
"loss": 1.086897611618042,
"step": 113
},
{
"epoch": 1.28388928317956,
"grad_norm": 1.7964507341384888,
"learning_rate": 1.4533998558589319e-05,
"loss": 0.7104189395904541,
"step": 114
},
{
"epoch": 1.2952448545067423,
"grad_norm": 1.659604549407959,
"learning_rate": 1.4415454982159121e-05,
"loss": 0.7449539303779602,
"step": 115
},
{
"epoch": 1.3066004258339248,
"grad_norm": 1.7559200525283813,
"learning_rate": 1.429613556574928e-05,
"loss": 0.760001540184021,
"step": 116
},
{
"epoch": 1.3179559971611072,
"grad_norm": 1.6712889671325684,
"learning_rate": 1.4176061274984858e-05,
"loss": 0.5844006538391113,
"step": 117
},
{
"epoch": 1.3293115684882895,
"grad_norm": 1.8857864141464233,
"learning_rate": 1.405525320812994e-05,
"loss": 0.799013614654541,
"step": 118
},
{
"epoch": 1.340667139815472,
"grad_norm": 1.9506901502609253,
"learning_rate": 1.3933732592380485e-05,
"loss": 0.8254635334014893,
"step": 119
},
{
"epoch": 1.3520227111426544,
"grad_norm": 1.556196689605713,
"learning_rate": 1.3811520780134471e-05,
"loss": 0.6055241227149963,
"step": 120
},
{
"epoch": 1.3633782824698368,
"grad_norm": 1.5888230800628662,
"learning_rate": 1.3688639245240078e-05,
"loss": 0.641814112663269,
"step": 121
},
{
"epoch": 1.3747338537970193,
"grad_norm": 1.740252137184143,
"learning_rate": 1.3565109579222511e-05,
"loss": 0.746912956237793,
"step": 122
},
{
"epoch": 1.3860894251242015,
"grad_norm": 2.0926947593688965,
"learning_rate": 1.3440953487490145e-05,
"loss": 0.8887531161308289,
"step": 123
},
{
"epoch": 1.397444996451384,
"grad_norm": 1.7420989274978638,
"learning_rate": 1.331619278552068e-05,
"loss": 0.6902853846549988,
"step": 124
},
{
"epoch": 1.4088005677785664,
"grad_norm": 1.8094593286514282,
"learning_rate": 1.3190849395027926e-05,
"loss": 0.7649242877960205,
"step": 125
},
{
"epoch": 1.4201561391057487,
"grad_norm": 1.4437446594238281,
"learning_rate": 1.306494534010995e-05,
"loss": 0.5692201256752014,
"step": 126
},
{
"epoch": 1.4315117104329311,
"grad_norm": 1.3073190450668335,
"learning_rate": 1.2938502743379212e-05,
"loss": 0.4875521957874298,
"step": 127
},
{
"epoch": 1.4428672817601136,
"grad_norm": 1.6500370502471924,
"learning_rate": 1.2811543822075396e-05,
"loss": 0.6286922097206116,
"step": 128
},
{
"epoch": 1.454222853087296,
"grad_norm": 1.4302570819854736,
"learning_rate": 1.2684090884161636e-05,
"loss": 0.48701754212379456,
"step": 129
},
{
"epoch": 1.4655784244144783,
"grad_norm": 1.7269134521484375,
"learning_rate": 1.2556166324404747e-05,
"loss": 0.6981077790260315,
"step": 130
},
{
"epoch": 1.4769339957416607,
"grad_norm": 1.5457602739334106,
"learning_rate": 1.242779262044028e-05,
"loss": 0.48003125190734863,
"step": 131
},
{
"epoch": 1.4882895670688432,
"grad_norm": 1.8600218296051025,
"learning_rate": 1.2298992328822937e-05,
"loss": 0.7610920667648315,
"step": 132
},
{
"epoch": 1.4996451383960254,
"grad_norm": 1.3750872611999512,
"learning_rate": 1.2169788081063181e-05,
"loss": 0.4584071934223175,
"step": 133
},
{
"epoch": 1.5110007097232079,
"grad_norm": 1.5970045328140259,
"learning_rate": 1.2040202579650649e-05,
"loss": 0.6092408895492554,
"step": 134
},
{
"epoch": 1.5223562810503903,
"grad_norm": 1.8978732824325562,
"learning_rate": 1.1910258594065079e-05,
"loss": 0.8766818642616272,
"step": 135
},
{
"epoch": 1.5337118523775728,
"grad_norm": 1.7095293998718262,
"learning_rate": 1.1779978956775507e-05,
"loss": 0.629298985004425,
"step": 136
},
{
"epoch": 1.5450674237047552,
"grad_norm": 1.8608379364013672,
"learning_rate": 1.1649386559228342e-05,
"loss": 0.6764330267906189,
"step": 137
},
{
"epoch": 1.5564229950319377,
"grad_norm": 1.760704517364502,
"learning_rate": 1.1518504347825146e-05,
"loss": 0.7110517621040344,
"step": 138
},
{
"epoch": 1.56777856635912,
"grad_norm": 1.937847375869751,
"learning_rate": 1.1387355319890685e-05,
"loss": 0.7304476499557495,
"step": 139
},
{
"epoch": 1.5791341376863024,
"grad_norm": 1.5163425207138062,
"learning_rate": 1.1255962519632082e-05,
"loss": 0.6055701971054077,
"step": 140
},
{
"epoch": 1.5904897090134846,
"grad_norm": 1.6075628995895386,
"learning_rate": 1.1124349034089724e-05,
"loss": 0.5708965063095093,
"step": 141
},
{
"epoch": 1.601845280340667,
"grad_norm": 1.7377432584762573,
"learning_rate": 1.0992537989080618e-05,
"loss": 0.672942042350769,
"step": 142
},
{
"epoch": 1.6132008516678495,
"grad_norm": 2.052077293395996,
"learning_rate": 1.086055254513497e-05,
"loss": 0.7728007435798645,
"step": 143
},
{
"epoch": 1.624556422995032,
"grad_norm": 1.5325052738189697,
"learning_rate": 1.0728415893426636e-05,
"loss": 0.5753393769264221,
"step": 144
},
{
"epoch": 1.6359119943222145,
"grad_norm": 2.0628015995025635,
"learning_rate": 1.05961512516982e-05,
"loss": 0.7965511083602905,
"step": 145
},
{
"epoch": 1.6472675656493967,
"grad_norm": 1.5534552335739136,
"learning_rate": 1.0463781860181385e-05,
"loss": 0.4485720694065094,
"step": 146
},
{
"epoch": 1.6586231369765791,
"grad_norm": 2.0554072856903076,
"learning_rate": 1.033133097751351e-05,
"loss": 0.87266606092453,
"step": 147
},
{
"epoch": 1.6699787083037614,
"grad_norm": 1.702256441116333,
"learning_rate": 1.0198821876650702e-05,
"loss": 0.5705003142356873,
"step": 148
},
{
"epoch": 1.6813342796309438,
"grad_norm": 1.5394976139068604,
"learning_rate": 1.0066277840778626e-05,
"loss": 0.4785357117652893,
"step": 149
},
{
"epoch": 1.6926898509581263,
"grad_norm": 1.7153429985046387,
"learning_rate": 9.933722159221375e-06,
"loss": 0.7122579216957092,
"step": 150
},
{
"epoch": 1.6926898509581263,
"eval_loss": 0.7425792217254639,
"eval_runtime": 27.5719,
"eval_samples_per_second": 5.186,
"eval_steps_per_second": 1.306,
"step": 150
},
{
"epoch": 1.7040454222853088,
"grad_norm": 1.7121310234069824,
"learning_rate": 9.801178123349298e-06,
"loss": 0.6155564188957214,
"step": 151
},
{
"epoch": 1.7154009936124912,
"grad_norm": 1.8881406784057617,
"learning_rate": 9.668669022486495e-06,
"loss": 0.7582035064697266,
"step": 152
},
{
"epoch": 1.7267565649396737,
"grad_norm": 1.9322803020477295,
"learning_rate": 9.536218139818615e-06,
"loss": 0.8061675429344177,
"step": 153
},
{
"epoch": 1.738112136266856,
"grad_norm": 1.5622960329055786,
"learning_rate": 9.403848748301802e-06,
"loss": 0.5297854542732239,
"step": 154
},
{
"epoch": 1.7494677075940384,
"grad_norm": 1.8679121732711792,
"learning_rate": 9.271584106573364e-06,
"loss": 0.6446192264556885,
"step": 155
},
{
"epoch": 1.7608232789212206,
"grad_norm": 1.65900456905365,
"learning_rate": 9.139447454865034e-06,
"loss": 0.6376427412033081,
"step": 156
},
{
"epoch": 1.772178850248403,
"grad_norm": 2.008634328842163,
"learning_rate": 9.007462010919387e-06,
"loss": 0.8298434615135193,
"step": 157
},
{
"epoch": 1.7835344215755855,
"grad_norm": 1.817112684249878,
"learning_rate": 8.87565096591028e-06,
"loss": 0.6945917010307312,
"step": 158
},
{
"epoch": 1.794889992902768,
"grad_norm": 2.0954995155334473,
"learning_rate": 8.744037480367922e-06,
"loss": 0.922713577747345,
"step": 159
},
{
"epoch": 1.8062455642299504,
"grad_norm": 1.8708635568618774,
"learning_rate": 8.61264468010932e-06,
"loss": 0.7695997953414917,
"step": 160
},
{
"epoch": 1.8176011355571329,
"grad_norm": 2.0717625617980957,
"learning_rate": 8.481495652174859e-06,
"loss": 0.9634745121002197,
"step": 161
},
{
"epoch": 1.828956706884315,
"grad_norm": 1.4181921482086182,
"learning_rate": 8.350613440771661e-06,
"loss": 0.41311249136924744,
"step": 162
},
{
"epoch": 1.8403122782114976,
"grad_norm": 1.5503257513046265,
"learning_rate": 8.2200210432245e-06,
"loss": 0.6834385395050049,
"step": 163
},
{
"epoch": 1.8516678495386798,
"grad_norm": 1.767377257347107,
"learning_rate": 8.089741405934923e-06,
"loss": 0.7185457944869995,
"step": 164
},
{
"epoch": 1.8630234208658623,
"grad_norm": 1.8889371156692505,
"learning_rate": 7.959797420349356e-06,
"loss": 0.9002049565315247,
"step": 165
},
{
"epoch": 1.8743789921930447,
"grad_norm": 2.01930832862854,
"learning_rate": 7.83021191893682e-06,
"loss": 0.6417741775512695,
"step": 166
},
{
"epoch": 1.8857345635202272,
"grad_norm": 1.8323336839675903,
"learning_rate": 7.701007671177066e-06,
"loss": 0.8583472371101379,
"step": 167
},
{
"epoch": 1.8970901348474096,
"grad_norm": 1.3641561269760132,
"learning_rate": 7.572207379559722e-06,
"loss": 0.5814041495323181,
"step": 168
},
{
"epoch": 1.9084457061745919,
"grad_norm": 1.6025844812393188,
"learning_rate": 7.443833675595254e-06,
"loss": 0.5391830205917358,
"step": 169
},
{
"epoch": 1.9198012775017743,
"grad_norm": 1.523732304573059,
"learning_rate": 7.315909115838367e-06,
"loss": 0.6024059057235718,
"step": 170
},
{
"epoch": 1.9311568488289566,
"grad_norm": 1.6438531875610352,
"learning_rate": 7.1884561779246055e-06,
"loss": 0.5504345297813416,
"step": 171
},
{
"epoch": 1.942512420156139,
"grad_norm": 1.884101152420044,
"learning_rate": 7.061497256620793e-06,
"loss": 0.8089767694473267,
"step": 172
},
{
"epoch": 1.9538679914833215,
"grad_norm": 1.705494999885559,
"learning_rate": 6.935054659890053e-06,
"loss": 0.6709920763969421,
"step": 173
},
{
"epoch": 1.965223562810504,
"grad_norm": 1.8286340236663818,
"learning_rate": 6.809150604972079e-06,
"loss": 0.9714908003807068,
"step": 174
},
{
"epoch": 1.9765791341376864,
"grad_norm": 1.6881917715072632,
"learning_rate": 6.683807214479323e-06,
"loss": 0.6124534606933594,
"step": 175
},
{
"epoch": 1.9879347054648688,
"grad_norm": 2.147641181945801,
"learning_rate": 6.559046512509859e-06,
"loss": 0.9034600257873535,
"step": 176
},
{
"epoch": 1.999290276792051,
"grad_norm": 1.7281945943832397,
"learning_rate": 6.434890420777491e-06,
"loss": 0.7872213125228882,
"step": 177
},
{
"epoch": 2.0,
"grad_norm": 4.5360822677612305,
"learning_rate": 6.311360754759923e-06,
"loss": 0.2226628065109253,
"step": 178
},
{
"epoch": 2.0113555713271825,
"grad_norm": 1.6504145860671997,
"learning_rate": 6.188479219865529e-06,
"loss": 0.6821281313896179,
"step": 179
},
{
"epoch": 2.022711142654365,
"grad_norm": 1.4790889024734497,
"learning_rate": 6.06626740761952e-06,
"loss": 0.5527986288070679,
"step": 180
},
{
"epoch": 2.0340667139815474,
"grad_norm": 1.8767532110214233,
"learning_rate": 5.944746791870062e-06,
"loss": 0.7406971454620361,
"step": 181
},
{
"epoch": 2.0454222853087294,
"grad_norm": 1.6067395210266113,
"learning_rate": 5.823938725015148e-06,
"loss": 0.5897183418273926,
"step": 182
},
{
"epoch": 2.056777856635912,
"grad_norm": 1.6543129682540894,
"learning_rate": 5.703864434250721e-06,
"loss": 0.5693725943565369,
"step": 183
},
{
"epoch": 2.0681334279630943,
"grad_norm": 1.763333797454834,
"learning_rate": 5.584545017840886e-06,
"loss": 0.6083229184150696,
"step": 184
},
{
"epoch": 2.0794889992902768,
"grad_norm": 1.2259186506271362,
"learning_rate": 5.4660014414106825e-06,
"loss": 0.3644542098045349,
"step": 185
},
{
"epoch": 2.090844570617459,
"grad_norm": 1.6024012565612793,
"learning_rate": 5.348254534262262e-06,
"loss": 0.607455313205719,
"step": 186
},
{
"epoch": 2.1022001419446417,
"grad_norm": 1.5699106454849243,
"learning_rate": 5.231324985714942e-06,
"loss": 0.4221411645412445,
"step": 187
},
{
"epoch": 2.113555713271824,
"grad_norm": 1.439180850982666,
"learning_rate": 5.1152333414698774e-06,
"loss": 0.4416266679763794,
"step": 188
},
{
"epoch": 2.1249112845990066,
"grad_norm": 1.8221555948257446,
"learning_rate": 5.000000000000003e-06,
"loss": 0.5936031341552734,
"step": 189
},
{
"epoch": 2.1362668559261886,
"grad_norm": 1.888954520225525,
"learning_rate": 4.885645208965779e-06,
"loss": 0.7918609976768494,
"step": 190
},
{
"epoch": 2.147622427253371,
"grad_norm": 1.6411021947860718,
"learning_rate": 4.772189061657511e-06,
"loss": 0.5752283334732056,
"step": 191
},
{
"epoch": 2.1589779985805535,
"grad_norm": 1.6003109216690063,
"learning_rate": 4.659651493464721e-06,
"loss": 0.6083833575248718,
"step": 192
},
{
"epoch": 2.170333569907736,
"grad_norm": 1.657971739768982,
"learning_rate": 4.548052278373327e-06,
"loss": 0.6448208093643188,
"step": 193
},
{
"epoch": 2.1816891412349184,
"grad_norm": 1.9829140901565552,
"learning_rate": 4.437411025491131e-06,
"loss": 0.6960307359695435,
"step": 194
},
{
"epoch": 2.193044712562101,
"grad_norm": 1.8921784162521362,
"learning_rate": 4.327747175602321e-06,
"loss": 0.731373131275177,
"step": 195
},
{
"epoch": 2.2044002838892833,
"grad_norm": 1.7255470752716064,
"learning_rate": 4.219079997751515e-06,
"loss": 0.5975107550621033,
"step": 196
},
{
"epoch": 2.215755855216466,
"grad_norm": 1.5999464988708496,
"learning_rate": 4.111428585858005e-06,
"loss": 0.5760666131973267,
"step": 197
},
{
"epoch": 2.227111426543648,
"grad_norm": 1.240045428276062,
"learning_rate": 4.0048118553607485e-06,
"loss": 0.39564967155456543,
"step": 198
},
{
"epoch": 2.2384669978708303,
"grad_norm": 1.9376376867294312,
"learning_rate": 3.899248539894756e-06,
"loss": 0.6683239936828613,
"step": 199
},
{
"epoch": 2.2498225691980127,
"grad_norm": 1.8060548305511475,
"learning_rate": 3.794757187999386e-06,
"loss": 0.7160100936889648,
"step": 200
},
{
"epoch": 2.2498225691980127,
"eval_loss": 0.7327238321304321,
"eval_runtime": 27.6122,
"eval_samples_per_second": 5.179,
"eval_steps_per_second": 1.304,
"step": 200
},
{
"epoch": 2.261178140525195,
"grad_norm": 1.5847166776657104,
"learning_rate": 3.6913561598591775e-06,
"loss": 0.5229685306549072,
"step": 201
},
{
"epoch": 2.2725337118523776,
"grad_norm": 1.6213287115097046,
"learning_rate": 3.589063624077802e-06,
"loss": 0.5970481634140015,
"step": 202
},
{
"epoch": 2.28388928317956,
"grad_norm": 1.818298578262329,
"learning_rate": 3.4878975544856285e-06,
"loss": 0.7519898414611816,
"step": 203
},
{
"epoch": 2.2952448545067425,
"grad_norm": 1.6277714967727661,
"learning_rate": 3.387875726981563e-06,
"loss": 0.5007752180099487,
"step": 204
},
{
"epoch": 2.306600425833925,
"grad_norm": 1.6472209692001343,
"learning_rate": 3.2890157164096315e-06,
"loss": 0.6382125616073608,
"step": 205
},
{
"epoch": 2.317955997161107,
"grad_norm": 1.821578025817871,
"learning_rate": 3.1913348934709076e-06,
"loss": 0.7000253200531006,
"step": 206
},
{
"epoch": 2.3293115684882895,
"grad_norm": 1.8459348678588867,
"learning_rate": 3.094850421671295e-06,
"loss": 0.5165408253669739,
"step": 207
},
{
"epoch": 2.340667139815472,
"grad_norm": 2.36333966255188,
"learning_rate": 2.999579254305748e-06,
"loss": 0.8205851912498474,
"step": 208
},
{
"epoch": 2.3520227111426544,
"grad_norm": 1.6719001531600952,
"learning_rate": 2.905538131479376e-06,
"loss": 0.5079140663146973,
"step": 209
},
{
"epoch": 2.363378282469837,
"grad_norm": 1.5651746988296509,
"learning_rate": 2.812743577166075e-06,
"loss": 0.4731737971305847,
"step": 210
},
{
"epoch": 2.3747338537970193,
"grad_norm": 2.142404317855835,
"learning_rate": 2.721211896305059e-06,
"loss": 0.842810869216919,
"step": 211
},
{
"epoch": 2.3860894251242017,
"grad_norm": 1.796759843826294,
"learning_rate": 2.6309591719359563e-06,
"loss": 0.7187446355819702,
"step": 212
},
{
"epoch": 2.3974449964513838,
"grad_norm": 1.492469310760498,
"learning_rate": 2.542001262372821e-06,
"loss": 0.5724196434020996,
"step": 213
},
{
"epoch": 2.408800567778566,
"grad_norm": 1.905091643333435,
"learning_rate": 2.454353798417698e-06,
"loss": 0.8085651397705078,
"step": 214
},
{
"epoch": 2.4201561391057487,
"grad_norm": 1.7704670429229736,
"learning_rate": 2.3680321806141182e-06,
"loss": 0.6132161021232605,
"step": 215
},
{
"epoch": 2.431511710432931,
"grad_norm": 1.8361713886260986,
"learning_rate": 2.283051576541062e-06,
"loss": 0.7117694616317749,
"step": 216
},
{
"epoch": 2.4428672817601136,
"grad_norm": 1.6990126371383667,
"learning_rate": 2.19942691814788e-06,
"loss": 0.6097843050956726,
"step": 217
},
{
"epoch": 2.454222853087296,
"grad_norm": 1.7435643672943115,
"learning_rate": 2.1171728991305797e-06,
"loss": 0.5934941172599792,
"step": 218
},
{
"epoch": 2.4655784244144785,
"grad_norm": 1.9126805067062378,
"learning_rate": 2.0363039723500155e-06,
"loss": 0.704249918460846,
"step": 219
},
{
"epoch": 2.4769339957416605,
"grad_norm": 1.7842662334442139,
"learning_rate": 1.9568343472923524e-06,
"loss": 0.6474202871322632,
"step": 220
},
{
"epoch": 2.488289567068843,
"grad_norm": 1.7116307020187378,
"learning_rate": 1.8787779875723389e-06,
"loss": 0.5358954668045044,
"step": 221
},
{
"epoch": 2.4996451383960254,
"grad_norm": 1.9682855606079102,
"learning_rate": 1.8021486084797368e-06,
"loss": 0.7930698990821838,
"step": 222
},
{
"epoch": 2.511000709723208,
"grad_norm": 1.4548183679580688,
"learning_rate": 1.7269596745694295e-06,
"loss": 0.4583667814731598,
"step": 223
},
{
"epoch": 2.5223562810503903,
"grad_norm": 2.000056505203247,
"learning_rate": 1.6532243972955397e-06,
"loss": 0.8167339563369751,
"step": 224
},
{
"epoch": 2.533711852377573,
"grad_norm": 2.187757730484009,
"learning_rate": 1.580955732690065e-06,
"loss": 0.8731608390808105,
"step": 225
},
{
"epoch": 2.5450674237047552,
"grad_norm": 2.552974224090576,
"learning_rate": 1.5101663790863597e-06,
"loss": 0.9876814484596252,
"step": 226
},
{
"epoch": 2.5564229950319377,
"grad_norm": 2.047971725463867,
"learning_rate": 1.4408687748879157e-06,
"loss": 0.7313542366027832,
"step": 227
},
{
"epoch": 2.56777856635912,
"grad_norm": 2.023643732070923,
"learning_rate": 1.3730750963828033e-06,
"loss": 0.7768765091896057,
"step": 228
},
{
"epoch": 2.579134137686302,
"grad_norm": 1.7678884267807007,
"learning_rate": 1.3067972556041753e-06,
"loss": 0.5483944416046143,
"step": 229
},
{
"epoch": 2.5904897090134846,
"grad_norm": 1.57295560836792,
"learning_rate": 1.2420468982372158e-06,
"loss": 0.5286450386047363,
"step": 230
},
{
"epoch": 2.601845280340667,
"grad_norm": 2.199338912963867,
"learning_rate": 1.1788354015728543e-06,
"loss": 0.7952181100845337,
"step": 231
},
{
"epoch": 2.6132008516678495,
"grad_norm": 2.037968635559082,
"learning_rate": 1.1171738725086833e-06,
"loss": 0.8746033906936646,
"step": 232
},
{
"epoch": 2.624556422995032,
"grad_norm": 1.6114685535430908,
"learning_rate": 1.0570731455973415e-06,
"loss": 0.5900834202766418,
"step": 233
},
{
"epoch": 2.6359119943222145,
"grad_norm": 1.8988250494003296,
"learning_rate": 9.985437811427934e-07,
"loss": 0.7454937696456909,
"step": 234
},
{
"epoch": 2.647267565649397,
"grad_norm": 1.6500483751296997,
"learning_rate": 9.415960633447674e-07,
"loss": 0.5034666657447815,
"step": 235
},
{
"epoch": 2.658623136976579,
"grad_norm": 1.9361827373504639,
"learning_rate": 8.862399984917214e-07,
"loss": 0.651042103767395,
"step": 236
},
{
"epoch": 2.6699787083037614,
"grad_norm": 1.9240092039108276,
"learning_rate": 8.3248531320263e-07,
"loss": 0.8603060245513916,
"step": 237
},
{
"epoch": 2.681334279630944,
"grad_norm": 1.9688630104064941,
"learning_rate": 7.803414527179343e-07,
"loss": 0.8090041875839233,
"step": 238
},
{
"epoch": 2.6926898509581263,
"grad_norm": 1.7253657579421997,
"learning_rate": 7.298175792398976e-07,
"loss": 0.5564637184143066,
"step": 239
},
{
"epoch": 2.7040454222853088,
"grad_norm": 1.871412754058838,
"learning_rate": 6.809225703227352e-07,
"loss": 0.7740951776504517,
"step": 240
},
{
"epoch": 2.715400993612491,
"grad_norm": 1.8274517059326172,
"learning_rate": 6.336650173127224e-07,
"loss": 0.6926577091217041,
"step": 241
},
{
"epoch": 2.7267565649396737,
"grad_norm": 2.128312349319458,
"learning_rate": 5.880532238386161e-07,
"loss": 0.8623908758163452,
"step": 242
},
{
"epoch": 2.7381121362668557,
"grad_norm": 1.9063048362731934,
"learning_rate": 5.440952043526215e-07,
"loss": 0.642934262752533,
"step": 243
},
{
"epoch": 2.7494677075940386,
"grad_norm": 1.9280798435211182,
"learning_rate": 5.017986827221733e-07,
"loss": 0.5877834558486938,
"step": 244
},
{
"epoch": 2.7608232789212206,
"grad_norm": 1.7953683137893677,
"learning_rate": 4.61171090872774e-07,
"loss": 0.6689532995223999,
"step": 245
},
{
"epoch": 2.772178850248403,
"grad_norm": 2.076279640197754,
"learning_rate": 4.222195674821239e-07,
"loss": 0.8235094547271729,
"step": 246
},
{
"epoch": 2.7835344215755855,
"grad_norm": 2.2308530807495117,
"learning_rate": 3.8495095672579584e-07,
"loss": 0.8819231986999512,
"step": 247
},
{
"epoch": 2.794889992902768,
"grad_norm": 1.5626147985458374,
"learning_rate": 3.493718070746299e-07,
"loss": 0.45997241139411926,
"step": 248
},
{
"epoch": 2.8062455642299504,
"grad_norm": 1.80852472782135,
"learning_rate": 3.154883701441136e-07,
"loss": 0.5900078415870667,
"step": 249
},
{
"epoch": 2.817601135557133,
"grad_norm": 1.591232180595398,
"learning_rate": 2.8330659959589944e-07,
"loss": 0.4933575987815857,
"step": 250
},
{
"epoch": 2.817601135557133,
"eval_loss": 0.7293056845664978,
"eval_runtime": 27.6187,
"eval_samples_per_second": 5.178,
"eval_steps_per_second": 1.303,
"step": 250
},
{
"epoch": 2.8289567068843153,
"grad_norm": 1.8240505456924438,
"learning_rate": 2.528321500916986e-07,
"loss": 0.6979119777679443,
"step": 251
},
{
"epoch": 2.8403122782114973,
"grad_norm": 1.9255794286727905,
"learning_rate": 2.2407037629968431e-07,
"loss": 0.7218936681747437,
"step": 252
},
{
"epoch": 2.85166784953868,
"grad_norm": 1.9031062126159668,
"learning_rate": 1.9702633195363918e-07,
"loss": 0.7347554564476013,
"step": 253
},
{
"epoch": 2.8630234208658623,
"grad_norm": 2.0092129707336426,
"learning_rate": 1.717047689649487e-07,
"loss": 0.8473051190376282,
"step": 254
},
{
"epoch": 2.8743789921930447,
"grad_norm": 2.1279916763305664,
"learning_rate": 1.4811013658765471e-07,
"loss": 0.8153268098831177,
"step": 255
},
{
"epoch": 2.885734563520227,
"grad_norm": 1.5587924718856812,
"learning_rate": 1.262465806366664e-07,
"loss": 0.42626267671585083,
"step": 256
},
{
"epoch": 2.8970901348474096,
"grad_norm": 1.851258635520935,
"learning_rate": 1.0611794275930398e-07,
"loss": 0.8002920150756836,
"step": 257
},
{
"epoch": 2.908445706174592,
"grad_norm": 1.7526359558105469,
"learning_rate": 8.772775976028547e-08,
"loss": 0.8035959005355835,
"step": 258
},
{
"epoch": 2.919801277501774,
"grad_norm": 1.8454657793045044,
"learning_rate": 7.10792629802659e-08,
"loss": 0.6260057687759399,
"step": 259
},
{
"epoch": 2.9311568488289566,
"grad_norm": 1.751435399055481,
"learning_rate": 5.617537772806603e-08,
"loss": 0.6086565256118774,
"step": 260
},
{
"epoch": 2.942512420156139,
"grad_norm": 1.614911437034607,
"learning_rate": 4.3018722766661194e-08,
"loss": 0.5733129382133484,
"step": 261
},
{
"epoch": 2.9538679914833215,
"grad_norm": 1.6565971374511719,
"learning_rate": 3.161160985304168e-08,
"loss": 0.45287761092185974,
"step": 262
},
{
"epoch": 2.965223562810504,
"grad_norm": 1.740827202796936,
"learning_rate": 2.1956043332010957e-08,
"loss": 0.5719068646430969,
"step": 263
},
{
"epoch": 2.9765791341376864,
"grad_norm": 1.4406249523162842,
"learning_rate": 1.405371978400516e-08,
"loss": 0.5211077332496643,
"step": 264
},
{
"epoch": 2.987934705464869,
"grad_norm": 1.6702423095703125,
"learning_rate": 7.906027726981568e-09,
"loss": 0.5729439854621887,
"step": 265
},
{
"epoch": 2.999290276792051,
"grad_norm": 1.812246561050415,
"learning_rate": 3.514047372448204e-09,
"loss": 0.6405225396156311,
"step": 266
},
{
"epoch": 3.0,
"grad_norm": 8.397170066833496,
"learning_rate": 8.785504356556562e-10,
"loss": 1.24993097782135,
"step": 267
},
{
"epoch": 3.0,
"eval_loss": 0.729335606098175,
"eval_runtime": 27.6017,
"eval_samples_per_second": 5.181,
"eval_steps_per_second": 1.304,
"step": 267
}
],
"logging_steps": 1,
"max_steps": 267,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 50,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 7.224059784218112e+16,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}