AutoSDT-Coder-7B / trainer_state.json
flyhero's picture
Upload folder using huggingface_hub
4a7c9df verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9992229992229992,
"eval_steps": 500,
"global_step": 643,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.001554001554001554,
"grad_norm": 1.287060567597129,
"learning_rate": 1e-05,
"loss": 0.7171,
"step": 1
},
{
"epoch": 0.003108003108003108,
"grad_norm": 1.290398939330793,
"learning_rate": 9.999940321631158e-06,
"loss": 0.6721,
"step": 2
},
{
"epoch": 0.004662004662004662,
"grad_norm": 0.9929109717377168,
"learning_rate": 9.999761287949237e-06,
"loss": 0.4993,
"step": 3
},
{
"epoch": 0.006216006216006216,
"grad_norm": 0.9031927531720302,
"learning_rate": 9.99946290322801e-06,
"loss": 0.7566,
"step": 4
},
{
"epoch": 0.00777000777000777,
"grad_norm": 0.8214431188948244,
"learning_rate": 9.999045174590324e-06,
"loss": 0.6818,
"step": 5
},
{
"epoch": 0.009324009324009324,
"grad_norm": 0.9593999539651701,
"learning_rate": 9.998508112007925e-06,
"loss": 0.6547,
"step": 6
},
{
"epoch": 0.010878010878010878,
"grad_norm": 1.1099888789676076,
"learning_rate": 9.997851728301219e-06,
"loss": 0.6804,
"step": 7
},
{
"epoch": 0.012432012432012432,
"grad_norm": 0.8564192310514048,
"learning_rate": 9.99707603913897e-06,
"loss": 0.5633,
"step": 8
},
{
"epoch": 0.013986013986013986,
"grad_norm": 1.063236098586671,
"learning_rate": 9.996181063037924e-06,
"loss": 0.6966,
"step": 9
},
{
"epoch": 0.01554001554001554,
"grad_norm": 0.9541051202078349,
"learning_rate": 9.995166821362368e-06,
"loss": 0.7143,
"step": 10
},
{
"epoch": 0.017094017094017096,
"grad_norm": 0.9386326913828528,
"learning_rate": 9.994033338323612e-06,
"loss": 0.6527,
"step": 11
},
{
"epoch": 0.018648018648018648,
"grad_norm": 0.9850267342487482,
"learning_rate": 9.99278064097943e-06,
"loss": 0.5105,
"step": 12
},
{
"epoch": 0.020202020202020204,
"grad_norm": 0.9003120220786494,
"learning_rate": 9.991408759233394e-06,
"loss": 0.6594,
"step": 13
},
{
"epoch": 0.021756021756021756,
"grad_norm": 0.9818892990706043,
"learning_rate": 9.989917725834166e-06,
"loss": 0.6786,
"step": 14
},
{
"epoch": 0.023310023310023312,
"grad_norm": 1.3235201527859266,
"learning_rate": 9.988307576374727e-06,
"loss": 0.6224,
"step": 15
},
{
"epoch": 0.024864024864024864,
"grad_norm": 0.9490358455901691,
"learning_rate": 9.986578349291514e-06,
"loss": 0.7105,
"step": 16
},
{
"epoch": 0.02641802641802642,
"grad_norm": 1.0959823860120133,
"learning_rate": 9.984730085863504e-06,
"loss": 0.5076,
"step": 17
},
{
"epoch": 0.027972027972027972,
"grad_norm": 1.1760066598209744,
"learning_rate": 9.982762830211239e-06,
"loss": 0.5839,
"step": 18
},
{
"epoch": 0.029526029526029528,
"grad_norm": 0.9624079621780423,
"learning_rate": 9.980676629295763e-06,
"loss": 0.5875,
"step": 19
},
{
"epoch": 0.03108003108003108,
"grad_norm": 1.1138209649690243,
"learning_rate": 9.9784715329175e-06,
"loss": 0.5983,
"step": 20
},
{
"epoch": 0.03263403263403263,
"grad_norm": 0.7824885823731469,
"learning_rate": 9.976147593715074e-06,
"loss": 0.5242,
"step": 21
},
{
"epoch": 0.03418803418803419,
"grad_norm": 1.0508935007482303,
"learning_rate": 9.973704867164044e-06,
"loss": 0.6223,
"step": 22
},
{
"epoch": 0.035742035742035744,
"grad_norm": 0.8235018550902833,
"learning_rate": 9.971143411575585e-06,
"loss": 0.4938,
"step": 23
},
{
"epoch": 0.037296037296037296,
"grad_norm": 1.2898619091079182,
"learning_rate": 9.968463288095096e-06,
"loss": 0.4851,
"step": 24
},
{
"epoch": 0.03885003885003885,
"grad_norm": 0.9265755568894043,
"learning_rate": 9.965664560700734e-06,
"loss": 0.6658,
"step": 25
},
{
"epoch": 0.04040404040404041,
"grad_norm": 1.33414254076587,
"learning_rate": 9.962747296201891e-06,
"loss": 0.5924,
"step": 26
},
{
"epoch": 0.04195804195804196,
"grad_norm": 0.762599048893836,
"learning_rate": 9.959711564237603e-06,
"loss": 0.5684,
"step": 27
},
{
"epoch": 0.04351204351204351,
"grad_norm": 1.0642367716149972,
"learning_rate": 9.956557437274887e-06,
"loss": 0.8413,
"step": 28
},
{
"epoch": 0.045066045066045064,
"grad_norm": 0.9613084594171063,
"learning_rate": 9.953284990607e-06,
"loss": 0.6419,
"step": 29
},
{
"epoch": 0.046620046620046623,
"grad_norm": 1.0408055633101505,
"learning_rate": 9.949894302351653e-06,
"loss": 0.7669,
"step": 30
},
{
"epoch": 0.048174048174048176,
"grad_norm": 1.129431352855463,
"learning_rate": 9.946385453449145e-06,
"loss": 0.65,
"step": 31
},
{
"epoch": 0.04972804972804973,
"grad_norm": 1.040149278900537,
"learning_rate": 9.942758527660429e-06,
"loss": 0.5691,
"step": 32
},
{
"epoch": 0.05128205128205128,
"grad_norm": 1.185572202243125,
"learning_rate": 9.93901361156511e-06,
"loss": 0.671,
"step": 33
},
{
"epoch": 0.05283605283605284,
"grad_norm": 1.279058377235717,
"learning_rate": 9.935150794559379e-06,
"loss": 0.7019,
"step": 34
},
{
"epoch": 0.05439005439005439,
"grad_norm": 0.865984848234437,
"learning_rate": 9.931170168853886e-06,
"loss": 0.6164,
"step": 35
},
{
"epoch": 0.055944055944055944,
"grad_norm": 1.8291545977352945,
"learning_rate": 9.927071829471531e-06,
"loss": 0.5272,
"step": 36
},
{
"epoch": 0.057498057498057496,
"grad_norm": 0.8851885541597515,
"learning_rate": 9.922855874245197e-06,
"loss": 0.6185,
"step": 37
},
{
"epoch": 0.059052059052059055,
"grad_norm": 0.9256948505546105,
"learning_rate": 9.918522403815414e-06,
"loss": 0.5428,
"step": 38
},
{
"epoch": 0.06060606060606061,
"grad_norm": 0.9865980759954314,
"learning_rate": 9.914071521627964e-06,
"loss": 0.6704,
"step": 39
},
{
"epoch": 0.06216006216006216,
"grad_norm": 1.0375856097276237,
"learning_rate": 9.909503333931402e-06,
"loss": 0.562,
"step": 40
},
{
"epoch": 0.06371406371406371,
"grad_norm": 1.0277215273777662,
"learning_rate": 9.904817949774524e-06,
"loss": 0.5346,
"step": 41
},
{
"epoch": 0.06526806526806526,
"grad_norm": 1.0255486122101665,
"learning_rate": 9.900015481003762e-06,
"loss": 0.5937,
"step": 42
},
{
"epoch": 0.06682206682206682,
"grad_norm": 0.9991894136698592,
"learning_rate": 9.895096042260517e-06,
"loss": 0.6547,
"step": 43
},
{
"epoch": 0.06837606837606838,
"grad_norm": 0.9734302555719793,
"learning_rate": 9.890059750978425e-06,
"loss": 0.64,
"step": 44
},
{
"epoch": 0.06993006993006994,
"grad_norm": 1.0702916548112473,
"learning_rate": 9.88490672738054e-06,
"loss": 0.7681,
"step": 45
},
{
"epoch": 0.07148407148407149,
"grad_norm": 0.7279123663050403,
"learning_rate": 9.879637094476482e-06,
"loss": 0.5674,
"step": 46
},
{
"epoch": 0.07303807303807304,
"grad_norm": 1.0078086114529654,
"learning_rate": 9.874250978059489e-06,
"loss": 0.5571,
"step": 47
},
{
"epoch": 0.07459207459207459,
"grad_norm": 0.8826041722896284,
"learning_rate": 9.86874850670342e-06,
"loss": 0.6377,
"step": 48
},
{
"epoch": 0.07614607614607614,
"grad_norm": 0.8063692525626558,
"learning_rate": 9.863129811759678e-06,
"loss": 0.6031,
"step": 49
},
{
"epoch": 0.0777000777000777,
"grad_norm": 0.9178388913448965,
"learning_rate": 9.857395027354085e-06,
"loss": 0.5568,
"step": 50
},
{
"epoch": 0.07925407925407925,
"grad_norm": 1.0644585188170625,
"learning_rate": 9.85154429038367e-06,
"loss": 0.69,
"step": 51
},
{
"epoch": 0.08080808080808081,
"grad_norm": 0.7927465990485753,
"learning_rate": 9.845577740513409e-06,
"loss": 0.5939,
"step": 52
},
{
"epoch": 0.08236208236208237,
"grad_norm": 0.9039359445089957,
"learning_rate": 9.83949552017289e-06,
"loss": 0.5956,
"step": 53
},
{
"epoch": 0.08391608391608392,
"grad_norm": 0.8329996086579473,
"learning_rate": 9.833297774552905e-06,
"loss": 0.5418,
"step": 54
},
{
"epoch": 0.08547008547008547,
"grad_norm": 0.8481919356532074,
"learning_rate": 9.826984651601998e-06,
"loss": 0.5471,
"step": 55
},
{
"epoch": 0.08702408702408702,
"grad_norm": 0.8509131135593907,
"learning_rate": 9.820556302022916e-06,
"loss": 0.5421,
"step": 56
},
{
"epoch": 0.08857808857808858,
"grad_norm": 0.8554172361714503,
"learning_rate": 9.814012879269031e-06,
"loss": 0.6966,
"step": 57
},
{
"epoch": 0.09013209013209013,
"grad_norm": 0.9899411098677792,
"learning_rate": 9.80735453954066e-06,
"loss": 0.524,
"step": 58
},
{
"epoch": 0.09168609168609168,
"grad_norm": 0.7650857989975672,
"learning_rate": 9.800581441781342e-06,
"loss": 0.5388,
"step": 59
},
{
"epoch": 0.09324009324009325,
"grad_norm": 0.8850444609719069,
"learning_rate": 9.79369374767405e-06,
"loss": 0.7299,
"step": 60
},
{
"epoch": 0.0947940947940948,
"grad_norm": 0.8836895106385606,
"learning_rate": 9.786691621637322e-06,
"loss": 0.5961,
"step": 61
},
{
"epoch": 0.09634809634809635,
"grad_norm": 0.9121728587330553,
"learning_rate": 9.779575230821344e-06,
"loss": 0.5426,
"step": 62
},
{
"epoch": 0.0979020979020979,
"grad_norm": 0.7042151959264945,
"learning_rate": 9.772344745103955e-06,
"loss": 0.5818,
"step": 63
},
{
"epoch": 0.09945609945609946,
"grad_norm": 0.9852547742106776,
"learning_rate": 9.76500033708659e-06,
"loss": 0.5852,
"step": 64
},
{
"epoch": 0.10101010101010101,
"grad_norm": 0.9664556160856704,
"learning_rate": 9.757542182090165e-06,
"loss": 0.609,
"step": 65
},
{
"epoch": 0.10256410256410256,
"grad_norm": 0.853406476434564,
"learning_rate": 9.749970458150893e-06,
"loss": 0.6294,
"step": 66
},
{
"epoch": 0.10411810411810411,
"grad_norm": 0.7958625946346,
"learning_rate": 9.742285346016024e-06,
"loss": 0.566,
"step": 67
},
{
"epoch": 0.10567210567210568,
"grad_norm": 0.9121076353990936,
"learning_rate": 9.734487029139544e-06,
"loss": 0.5667,
"step": 68
},
{
"epoch": 0.10722610722610723,
"grad_norm": 1.0944470614224682,
"learning_rate": 9.726575693677782e-06,
"loss": 0.6562,
"step": 69
},
{
"epoch": 0.10878010878010878,
"grad_norm": 0.9473072471718225,
"learning_rate": 9.718551528484979e-06,
"loss": 0.6156,
"step": 70
},
{
"epoch": 0.11033411033411034,
"grad_norm": 0.9165875064014658,
"learning_rate": 9.710414725108771e-06,
"loss": 0.4945,
"step": 71
},
{
"epoch": 0.11188811188811189,
"grad_norm": 0.7947044061935308,
"learning_rate": 9.702165477785618e-06,
"loss": 0.566,
"step": 72
},
{
"epoch": 0.11344211344211344,
"grad_norm": 0.8217207231050921,
"learning_rate": 9.69380398343617e-06,
"loss": 0.5382,
"step": 73
},
{
"epoch": 0.11499611499611499,
"grad_norm": 0.7072030756519719,
"learning_rate": 9.685330441660564e-06,
"loss": 0.6005,
"step": 74
},
{
"epoch": 0.11655011655011654,
"grad_norm": 1.127735633700878,
"learning_rate": 9.676745054733661e-06,
"loss": 0.686,
"step": 75
},
{
"epoch": 0.11810411810411811,
"grad_norm": 0.7567722308435004,
"learning_rate": 9.668048027600217e-06,
"loss": 0.5949,
"step": 76
},
{
"epoch": 0.11965811965811966,
"grad_norm": 0.940904721274449,
"learning_rate": 9.659239567869989e-06,
"loss": 0.7208,
"step": 77
},
{
"epoch": 0.12121212121212122,
"grad_norm": 0.8589933167401518,
"learning_rate": 9.650319885812777e-06,
"loss": 0.6428,
"step": 78
},
{
"epoch": 0.12276612276612277,
"grad_norm": 0.7460567300212997,
"learning_rate": 9.641289194353418e-06,
"loss": 0.4566,
"step": 79
},
{
"epoch": 0.12432012432012432,
"grad_norm": 0.9462477579247586,
"learning_rate": 9.632147709066682e-06,
"loss": 0.6661,
"step": 80
},
{
"epoch": 0.1258741258741259,
"grad_norm": 1.0265019580385855,
"learning_rate": 9.622895648172141e-06,
"loss": 0.5686,
"step": 81
},
{
"epoch": 0.12742812742812742,
"grad_norm": 0.8465826530732603,
"learning_rate": 9.613533232528956e-06,
"loss": 0.58,
"step": 82
},
{
"epoch": 0.128982128982129,
"grad_norm": 1.1107578812568586,
"learning_rate": 9.604060685630608e-06,
"loss": 0.6289,
"step": 83
},
{
"epoch": 0.13053613053613053,
"grad_norm": 0.9988960651039751,
"learning_rate": 9.594478233599551e-06,
"loss": 0.5435,
"step": 84
},
{
"epoch": 0.1320901320901321,
"grad_norm": 0.9725142325555755,
"learning_rate": 9.584786105181831e-06,
"loss": 0.5997,
"step": 85
},
{
"epoch": 0.13364413364413363,
"grad_norm": 0.9253748447907283,
"learning_rate": 9.574984531741613e-06,
"loss": 0.5761,
"step": 86
},
{
"epoch": 0.1351981351981352,
"grad_norm": 0.7533627407586,
"learning_rate": 9.565073747255665e-06,
"loss": 0.4554,
"step": 87
},
{
"epoch": 0.13675213675213677,
"grad_norm": 0.8354149018093497,
"learning_rate": 9.555053988307764e-06,
"loss": 0.6392,
"step": 88
},
{
"epoch": 0.1383061383061383,
"grad_norm": 0.7922119412325215,
"learning_rate": 9.544925494083062e-06,
"loss": 0.5969,
"step": 89
},
{
"epoch": 0.13986013986013987,
"grad_norm": 1.098491709946335,
"learning_rate": 9.53468850636236e-06,
"loss": 0.6687,
"step": 90
},
{
"epoch": 0.1414141414141414,
"grad_norm": 1.1478661593087778,
"learning_rate": 9.524343269516354e-06,
"loss": 0.5643,
"step": 91
},
{
"epoch": 0.14296814296814297,
"grad_norm": 1.1224670962595298,
"learning_rate": 9.513890030499786e-06,
"loss": 0.5974,
"step": 92
},
{
"epoch": 0.1445221445221445,
"grad_norm": 0.809091939178162,
"learning_rate": 9.503329038845556e-06,
"loss": 0.5372,
"step": 93
},
{
"epoch": 0.14607614607614608,
"grad_norm": 0.8198348581856645,
"learning_rate": 9.492660546658771e-06,
"loss": 0.4845,
"step": 94
},
{
"epoch": 0.14763014763014762,
"grad_norm": 1.057619334118549,
"learning_rate": 9.481884808610712e-06,
"loss": 0.6607,
"step": 95
},
{
"epoch": 0.14918414918414918,
"grad_norm": 0.908346525172346,
"learning_rate": 9.471002081932767e-06,
"loss": 0.6958,
"step": 96
},
{
"epoch": 0.15073815073815075,
"grad_norm": 0.9282968819634952,
"learning_rate": 9.460012626410286e-06,
"loss": 0.7207,
"step": 97
},
{
"epoch": 0.1522921522921523,
"grad_norm": 1.0629870155581145,
"learning_rate": 9.448916704376384e-06,
"loss": 0.5156,
"step": 98
},
{
"epoch": 0.15384615384615385,
"grad_norm": 1.0854387649323378,
"learning_rate": 9.437714580705671e-06,
"loss": 0.6926,
"step": 99
},
{
"epoch": 0.1554001554001554,
"grad_norm": 1.0036938568797955,
"learning_rate": 9.426406522807932e-06,
"loss": 0.7549,
"step": 100
},
{
"epoch": 0.15695415695415696,
"grad_norm": 0.8940797944953299,
"learning_rate": 9.414992800621749e-06,
"loss": 0.48,
"step": 101
},
{
"epoch": 0.1585081585081585,
"grad_norm": 0.9000106502572475,
"learning_rate": 9.40347368660805e-06,
"loss": 0.5819,
"step": 102
},
{
"epoch": 0.16006216006216006,
"grad_norm": 1.1002281313656057,
"learning_rate": 9.39184945574361e-06,
"loss": 0.5819,
"step": 103
},
{
"epoch": 0.16161616161616163,
"grad_norm": 1.0635278127825991,
"learning_rate": 9.380120385514484e-06,
"loss": 0.5034,
"step": 104
},
{
"epoch": 0.16317016317016317,
"grad_norm": 0.8048515611001529,
"learning_rate": 9.368286755909383e-06,
"loss": 0.5868,
"step": 105
},
{
"epoch": 0.16472416472416473,
"grad_norm": 0.7793016258380376,
"learning_rate": 9.356348849412991e-06,
"loss": 0.4159,
"step": 106
},
{
"epoch": 0.16627816627816627,
"grad_norm": 0.8749226384769527,
"learning_rate": 9.344306950999226e-06,
"loss": 0.5689,
"step": 107
},
{
"epoch": 0.16783216783216784,
"grad_norm": 0.7914152537538985,
"learning_rate": 9.332161348124426e-06,
"loss": 0.5214,
"step": 108
},
{
"epoch": 0.16938616938616938,
"grad_norm": 0.8511015877493499,
"learning_rate": 9.319912330720502e-06,
"loss": 0.5067,
"step": 109
},
{
"epoch": 0.17094017094017094,
"grad_norm": 0.7604252629336278,
"learning_rate": 9.307560191188e-06,
"loss": 0.6011,
"step": 110
},
{
"epoch": 0.17249417249417248,
"grad_norm": 0.739722467231117,
"learning_rate": 9.295105224389144e-06,
"loss": 0.58,
"step": 111
},
{
"epoch": 0.17404817404817405,
"grad_norm": 0.9930912809185233,
"learning_rate": 9.282547727640767e-06,
"loss": 0.5622,
"step": 112
},
{
"epoch": 0.17560217560217561,
"grad_norm": 1.0011373833431905,
"learning_rate": 9.269888000707243e-06,
"loss": 0.4572,
"step": 113
},
{
"epoch": 0.17715617715617715,
"grad_norm": 1.07232267999874,
"learning_rate": 9.25712634579331e-06,
"loss": 0.6382,
"step": 114
},
{
"epoch": 0.17871017871017872,
"grad_norm": 0.989519175726022,
"learning_rate": 9.244263067536872e-06,
"loss": 0.5169,
"step": 115
},
{
"epoch": 0.18026418026418026,
"grad_norm": 0.9692529023759858,
"learning_rate": 9.23129847300171e-06,
"loss": 0.5671,
"step": 116
},
{
"epoch": 0.18181818181818182,
"grad_norm": 0.9086519385832078,
"learning_rate": 9.218232871670168e-06,
"loss": 0.4454,
"step": 117
},
{
"epoch": 0.18337218337218336,
"grad_norm": 1.1029703568444662,
"learning_rate": 9.205066575435754e-06,
"loss": 0.5983,
"step": 118
},
{
"epoch": 0.18492618492618493,
"grad_norm": 1.0557315125722626,
"learning_rate": 9.191799898595706e-06,
"loss": 0.5914,
"step": 119
},
{
"epoch": 0.1864801864801865,
"grad_norm": 0.7020511039508586,
"learning_rate": 9.178433157843474e-06,
"loss": 0.5783,
"step": 120
},
{
"epoch": 0.18803418803418803,
"grad_norm": 1.030598572867442,
"learning_rate": 9.164966672261171e-06,
"loss": 0.5806,
"step": 121
},
{
"epoch": 0.1895881895881896,
"grad_norm": 1.0483043804107857,
"learning_rate": 9.151400763311958e-06,
"loss": 0.5605,
"step": 122
},
{
"epoch": 0.19114219114219114,
"grad_norm": 0.8689865881103244,
"learning_rate": 9.13773575483236e-06,
"loss": 0.5631,
"step": 123
},
{
"epoch": 0.1926961926961927,
"grad_norm": 0.7359655114483693,
"learning_rate": 9.123971973024543e-06,
"loss": 0.5697,
"step": 124
},
{
"epoch": 0.19425019425019424,
"grad_norm": 0.831516688258424,
"learning_rate": 9.110109746448527e-06,
"loss": 0.5017,
"step": 125
},
{
"epoch": 0.1958041958041958,
"grad_norm": 1.2080050795469695,
"learning_rate": 9.09614940601434e-06,
"loss": 0.5961,
"step": 126
},
{
"epoch": 0.19735819735819735,
"grad_norm": 0.9332888068852607,
"learning_rate": 9.08209128497412e-06,
"loss": 0.6506,
"step": 127
},
{
"epoch": 0.1989121989121989,
"grad_norm": 0.698648999430127,
"learning_rate": 9.06793571891416e-06,
"loss": 0.4766,
"step": 128
},
{
"epoch": 0.20046620046620048,
"grad_norm": 1.0441318680416376,
"learning_rate": 9.053683045746897e-06,
"loss": 0.5247,
"step": 129
},
{
"epoch": 0.20202020202020202,
"grad_norm": 0.8981679337660563,
"learning_rate": 9.039333605702844e-06,
"loss": 0.6338,
"step": 130
},
{
"epoch": 0.20357420357420358,
"grad_norm": 0.8484981101489031,
"learning_rate": 9.024887741322475e-06,
"loss": 0.5684,
"step": 131
},
{
"epoch": 0.20512820512820512,
"grad_norm": 1.0445078112703394,
"learning_rate": 9.010345797448037e-06,
"loss": 0.6309,
"step": 132
},
{
"epoch": 0.2066822066822067,
"grad_norm": 0.6975235761912434,
"learning_rate": 8.995708121215325e-06,
"loss": 0.5138,
"step": 133
},
{
"epoch": 0.20823620823620823,
"grad_norm": 0.7711316132241796,
"learning_rate": 8.980975062045398e-06,
"loss": 0.5182,
"step": 134
},
{
"epoch": 0.2097902097902098,
"grad_norm": 1.010608943282339,
"learning_rate": 8.96614697163623e-06,
"loss": 0.684,
"step": 135
},
{
"epoch": 0.21134421134421136,
"grad_norm": 0.9424855629079735,
"learning_rate": 8.95122420395432e-06,
"loss": 0.5894,
"step": 136
},
{
"epoch": 0.2128982128982129,
"grad_norm": 0.8196370668961501,
"learning_rate": 8.936207115226242e-06,
"loss": 0.5738,
"step": 137
},
{
"epoch": 0.21445221445221446,
"grad_norm": 0.7724030975377769,
"learning_rate": 8.921096063930141e-06,
"loss": 0.6508,
"step": 138
},
{
"epoch": 0.216006216006216,
"grad_norm": 0.9383967814046783,
"learning_rate": 8.905891410787174e-06,
"loss": 0.5184,
"step": 139
},
{
"epoch": 0.21756021756021757,
"grad_norm": 0.9234435984562406,
"learning_rate": 8.8905935187529e-06,
"loss": 0.6006,
"step": 140
},
{
"epoch": 0.2191142191142191,
"grad_norm": 1.3623850981895826,
"learning_rate": 8.875202753008614e-06,
"loss": 0.5531,
"step": 141
},
{
"epoch": 0.22066822066822067,
"grad_norm": 1.2508758031618654,
"learning_rate": 8.859719480952637e-06,
"loss": 0.7084,
"step": 142
},
{
"epoch": 0.2222222222222222,
"grad_norm": 0.8432095945353565,
"learning_rate": 8.844144072191537e-06,
"loss": 0.5614,
"step": 143
},
{
"epoch": 0.22377622377622378,
"grad_norm": 0.869685982949141,
"learning_rate": 8.828476898531308e-06,
"loss": 0.5273,
"step": 144
},
{
"epoch": 0.22533022533022534,
"grad_norm": 1.0535269875683682,
"learning_rate": 8.812718333968498e-06,
"loss": 0.5677,
"step": 145
},
{
"epoch": 0.22688422688422688,
"grad_norm": 0.7840965616384874,
"learning_rate": 8.79686875468128e-06,
"loss": 0.5804,
"step": 146
},
{
"epoch": 0.22843822843822845,
"grad_norm": 0.8922758139793021,
"learning_rate": 8.780928539020467e-06,
"loss": 0.4879,
"step": 147
},
{
"epoch": 0.22999222999222999,
"grad_norm": 1.0663700028063874,
"learning_rate": 8.764898067500488e-06,
"loss": 0.5255,
"step": 148
},
{
"epoch": 0.23154623154623155,
"grad_norm": 1.1052154796655376,
"learning_rate": 8.7487777227903e-06,
"loss": 0.5823,
"step": 149
},
{
"epoch": 0.2331002331002331,
"grad_norm": 0.8372838204749399,
"learning_rate": 8.732567889704253e-06,
"loss": 0.552,
"step": 150
},
{
"epoch": 0.23465423465423466,
"grad_norm": 1.0383914168078336,
"learning_rate": 8.716268955192908e-06,
"loss": 0.4969,
"step": 151
},
{
"epoch": 0.23620823620823622,
"grad_norm": 0.9310356859635174,
"learning_rate": 8.699881308333794e-06,
"loss": 0.6232,
"step": 152
},
{
"epoch": 0.23776223776223776,
"grad_norm": 0.9542737630454637,
"learning_rate": 8.683405340322123e-06,
"loss": 0.647,
"step": 153
},
{
"epoch": 0.23931623931623933,
"grad_norm": 0.7235983273059299,
"learning_rate": 8.666841444461456e-06,
"loss": 0.4978,
"step": 154
},
{
"epoch": 0.24087024087024086,
"grad_norm": 0.9214338180679594,
"learning_rate": 8.650190016154307e-06,
"loss": 0.5112,
"step": 155
},
{
"epoch": 0.24242424242424243,
"grad_norm": 0.7765319698353046,
"learning_rate": 8.633451452892707e-06,
"loss": 0.4189,
"step": 156
},
{
"epoch": 0.24397824397824397,
"grad_norm": 0.8361437850631817,
"learning_rate": 8.616626154248717e-06,
"loss": 0.6564,
"step": 157
},
{
"epoch": 0.24553224553224554,
"grad_norm": 1.1201137563120651,
"learning_rate": 8.59971452186489e-06,
"loss": 0.5598,
"step": 158
},
{
"epoch": 0.24708624708624707,
"grad_norm": 0.8323491927700207,
"learning_rate": 8.582716959444679e-06,
"loss": 0.5835,
"step": 159
},
{
"epoch": 0.24864024864024864,
"grad_norm": 0.9138271140965946,
"learning_rate": 8.565633872742803e-06,
"loss": 0.4858,
"step": 160
},
{
"epoch": 0.2501942501942502,
"grad_norm": 0.7984227928741979,
"learning_rate": 8.548465669555564e-06,
"loss": 0.6801,
"step": 161
},
{
"epoch": 0.2517482517482518,
"grad_norm": 1.0387416165235834,
"learning_rate": 8.531212759711103e-06,
"loss": 0.6349,
"step": 162
},
{
"epoch": 0.2533022533022533,
"grad_norm": 0.8098250618756626,
"learning_rate": 8.51387555505963e-06,
"loss": 0.5217,
"step": 163
},
{
"epoch": 0.25485625485625485,
"grad_norm": 0.7604279630540275,
"learning_rate": 8.496454469463583e-06,
"loss": 0.4321,
"step": 164
},
{
"epoch": 0.2564102564102564,
"grad_norm": 0.7467153785853686,
"learning_rate": 8.478949918787746e-06,
"loss": 0.7456,
"step": 165
},
{
"epoch": 0.257964257964258,
"grad_norm": 0.9217242408114555,
"learning_rate": 8.461362320889338e-06,
"loss": 0.4174,
"step": 166
},
{
"epoch": 0.2595182595182595,
"grad_norm": 1.0701886438461339,
"learning_rate": 8.443692095608019e-06,
"loss": 0.7204,
"step": 167
},
{
"epoch": 0.26107226107226106,
"grad_norm": 0.9940719591500394,
"learning_rate": 8.425939664755874e-06,
"loss": 0.5567,
"step": 168
},
{
"epoch": 0.26262626262626265,
"grad_norm": 0.8026878231443845,
"learning_rate": 8.408105452107353e-06,
"loss": 0.4792,
"step": 169
},
{
"epoch": 0.2641802641802642,
"grad_norm": 1.22440273921866,
"learning_rate": 8.390189883389143e-06,
"loss": 0.4696,
"step": 170
},
{
"epoch": 0.26573426573426573,
"grad_norm": 0.8474717813756039,
"learning_rate": 8.37219338627001e-06,
"loss": 0.4837,
"step": 171
},
{
"epoch": 0.26728826728826727,
"grad_norm": 0.9271171444493542,
"learning_rate": 8.354116390350594e-06,
"loss": 0.6195,
"step": 172
},
{
"epoch": 0.26884226884226886,
"grad_norm": 0.9175517581827086,
"learning_rate": 8.335959327153148e-06,
"loss": 0.6171,
"step": 173
},
{
"epoch": 0.2703962703962704,
"grad_norm": 1.1214384534629507,
"learning_rate": 8.317722630111233e-06,
"loss": 0.6122,
"step": 174
},
{
"epoch": 0.27195027195027194,
"grad_norm": 1.0369916538716633,
"learning_rate": 8.299406734559385e-06,
"loss": 0.4738,
"step": 175
},
{
"epoch": 0.27350427350427353,
"grad_norm": 0.9517993003576442,
"learning_rate": 8.281012077722712e-06,
"loss": 0.5938,
"step": 176
},
{
"epoch": 0.27505827505827507,
"grad_norm": 0.7651148829258754,
"learning_rate": 8.26253909870646e-06,
"loss": 0.6886,
"step": 177
},
{
"epoch": 0.2766122766122766,
"grad_norm": 0.9253215200171306,
"learning_rate": 8.24398823848553e-06,
"loss": 0.4333,
"step": 178
},
{
"epoch": 0.27816627816627815,
"grad_norm": 1.527596309810129,
"learning_rate": 8.225359939893954e-06,
"loss": 0.6215,
"step": 179
},
{
"epoch": 0.27972027972027974,
"grad_norm": 0.9172593952457445,
"learning_rate": 8.206654647614323e-06,
"loss": 0.5789,
"step": 180
},
{
"epoch": 0.2812742812742813,
"grad_norm": 0.8527814414594325,
"learning_rate": 8.18787280816717e-06,
"loss": 0.6184,
"step": 181
},
{
"epoch": 0.2828282828282828,
"grad_norm": 0.8865469839921362,
"learning_rate": 8.169014869900308e-06,
"loss": 0.6295,
"step": 182
},
{
"epoch": 0.28438228438228436,
"grad_norm": 0.8517309651275252,
"learning_rate": 8.150081282978139e-06,
"loss": 0.5331,
"step": 183
},
{
"epoch": 0.28593628593628595,
"grad_norm": 0.8315076118590468,
"learning_rate": 8.131072499370897e-06,
"loss": 0.5789,
"step": 184
},
{
"epoch": 0.2874902874902875,
"grad_norm": 0.7203294922723644,
"learning_rate": 8.111988972843859e-06,
"loss": 0.4577,
"step": 185
},
{
"epoch": 0.289044289044289,
"grad_norm": 0.8794205779087169,
"learning_rate": 8.09283115894652e-06,
"loss": 0.6125,
"step": 186
},
{
"epoch": 0.2905982905982906,
"grad_norm": 0.9889975335181714,
"learning_rate": 8.073599515001713e-06,
"loss": 0.6116,
"step": 187
},
{
"epoch": 0.29215229215229216,
"grad_norm": 0.9167264940797374,
"learning_rate": 8.054294500094697e-06,
"loss": 0.6634,
"step": 188
},
{
"epoch": 0.2937062937062937,
"grad_norm": 1.1038199923144063,
"learning_rate": 8.034916575062188e-06,
"loss": 0.495,
"step": 189
},
{
"epoch": 0.29526029526029524,
"grad_norm": 1.1290155229402656,
"learning_rate": 8.015466202481371e-06,
"loss": 0.6783,
"step": 190
},
{
"epoch": 0.29681429681429683,
"grad_norm": 1.0055630496859111,
"learning_rate": 7.995943846658852e-06,
"loss": 0.5865,
"step": 191
},
{
"epoch": 0.29836829836829837,
"grad_norm": 1.0372133424369208,
"learning_rate": 7.976349973619567e-06,
"loss": 0.5451,
"step": 192
},
{
"epoch": 0.2999222999222999,
"grad_norm": 0.8040563490915533,
"learning_rate": 7.956685051095672e-06,
"loss": 0.5964,
"step": 193
},
{
"epoch": 0.3014763014763015,
"grad_norm": 1.057546655101924,
"learning_rate": 7.936949548515364e-06,
"loss": 0.5565,
"step": 194
},
{
"epoch": 0.30303030303030304,
"grad_norm": 0.8821680821336786,
"learning_rate": 7.917143936991688e-06,
"loss": 0.5606,
"step": 195
},
{
"epoch": 0.3045843045843046,
"grad_norm": 0.8674907577741584,
"learning_rate": 7.897268689311278e-06,
"loss": 0.5862,
"step": 196
},
{
"epoch": 0.3061383061383061,
"grad_norm": 0.8706444651270692,
"learning_rate": 7.877324279923078e-06,
"loss": 0.6849,
"step": 197
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.4485468544220046,
"learning_rate": 7.857311184927015e-06,
"loss": 0.6806,
"step": 198
},
{
"epoch": 0.30924630924630925,
"grad_norm": 1.2122402985477576,
"learning_rate": 7.837229882062638e-06,
"loss": 0.6029,
"step": 199
},
{
"epoch": 0.3108003108003108,
"grad_norm": 0.6888280161460392,
"learning_rate": 7.817080850697705e-06,
"loss": 0.4127,
"step": 200
},
{
"epoch": 0.3123543123543124,
"grad_norm": 0.82470002461982,
"learning_rate": 7.796864571816745e-06,
"loss": 0.5841,
"step": 201
},
{
"epoch": 0.3139083139083139,
"grad_norm": 0.8090245690817658,
"learning_rate": 7.77658152800958e-06,
"loss": 0.5473,
"step": 202
},
{
"epoch": 0.31546231546231546,
"grad_norm": 0.7971008205776199,
"learning_rate": 7.756232203459794e-06,
"loss": 0.5453,
"step": 203
},
{
"epoch": 0.317016317016317,
"grad_norm": 0.8668563067845961,
"learning_rate": 7.735817083933189e-06,
"loss": 0.6337,
"step": 204
},
{
"epoch": 0.3185703185703186,
"grad_norm": 0.998359039754979,
"learning_rate": 7.715336656766176e-06,
"loss": 0.4682,
"step": 205
},
{
"epoch": 0.3201243201243201,
"grad_norm": 1.0072022163437613,
"learning_rate": 7.69479141085415e-06,
"loss": 0.6439,
"step": 206
},
{
"epoch": 0.32167832167832167,
"grad_norm": 1.0428130086675327,
"learning_rate": 7.674181836639819e-06,
"loss": 0.5053,
"step": 207
},
{
"epoch": 0.32323232323232326,
"grad_norm": 0.8220071235320535,
"learning_rate": 7.653508426101488e-06,
"loss": 0.5921,
"step": 208
},
{
"epoch": 0.3247863247863248,
"grad_norm": 1.403320779877978,
"learning_rate": 7.632771672741326e-06,
"loss": 0.6189,
"step": 209
},
{
"epoch": 0.32634032634032634,
"grad_norm": 1.0783911186573163,
"learning_rate": 7.611972071573579e-06,
"loss": 0.5385,
"step": 210
},
{
"epoch": 0.3278943278943279,
"grad_norm": 0.6792518204386103,
"learning_rate": 7.591110119112757e-06,
"loss": 0.5257,
"step": 211
},
{
"epoch": 0.32944832944832947,
"grad_norm": 0.9842199821582761,
"learning_rate": 7.5701863133617735e-06,
"loss": 0.5607,
"step": 212
},
{
"epoch": 0.331002331002331,
"grad_norm": 0.8239855954978503,
"learning_rate": 7.549201153800073e-06,
"loss": 0.5418,
"step": 213
},
{
"epoch": 0.33255633255633255,
"grad_norm": 0.7559206402903527,
"learning_rate": 7.528155141371688e-06,
"loss": 0.5555,
"step": 214
},
{
"epoch": 0.3341103341103341,
"grad_norm": 0.809815004677253,
"learning_rate": 7.507048778473296e-06,
"loss": 0.55,
"step": 215
},
{
"epoch": 0.3356643356643357,
"grad_norm": 0.8754773402756791,
"learning_rate": 7.485882568942222e-06,
"loss": 0.5637,
"step": 216
},
{
"epoch": 0.3372183372183372,
"grad_norm": 0.6877512908870741,
"learning_rate": 7.464657018044411e-06,
"loss": 0.3471,
"step": 217
},
{
"epoch": 0.33877233877233875,
"grad_norm": 1.0337878547530994,
"learning_rate": 7.443372632462363e-06,
"loss": 0.57,
"step": 218
},
{
"epoch": 0.34032634032634035,
"grad_norm": 1.2498880965990447,
"learning_rate": 7.422029920283044e-06,
"loss": 0.5174,
"step": 219
},
{
"epoch": 0.3418803418803419,
"grad_norm": 0.9357263067576875,
"learning_rate": 7.400629390985753e-06,
"loss": 0.5369,
"step": 220
},
{
"epoch": 0.3434343434343434,
"grad_norm": 0.7849539728009569,
"learning_rate": 7.379171555429965e-06,
"loss": 0.3983,
"step": 221
},
{
"epoch": 0.34498834498834496,
"grad_norm": 0.8659243256930758,
"learning_rate": 7.357656925843125e-06,
"loss": 0.4341,
"step": 222
},
{
"epoch": 0.34654234654234656,
"grad_norm": 0.6240344673857208,
"learning_rate": 7.336086015808439e-06,
"loss": 0.5405,
"step": 223
},
{
"epoch": 0.3480963480963481,
"grad_norm": 0.8177759212146966,
"learning_rate": 7.314459340252593e-06,
"loss": 0.5676,
"step": 224
},
{
"epoch": 0.34965034965034963,
"grad_norm": 0.9093224937417566,
"learning_rate": 7.2927774154334765e-06,
"loss": 0.5168,
"step": 225
},
{
"epoch": 0.35120435120435123,
"grad_norm": 1.0717234885263538,
"learning_rate": 7.271040758927852e-06,
"loss": 0.6916,
"step": 226
},
{
"epoch": 0.35275835275835277,
"grad_norm": 0.9075665009935251,
"learning_rate": 7.2492498896190015e-06,
"loss": 0.6452,
"step": 227
},
{
"epoch": 0.3543123543123543,
"grad_norm": 0.8136392828852748,
"learning_rate": 7.227405327684339e-06,
"loss": 0.4489,
"step": 228
},
{
"epoch": 0.35586635586635584,
"grad_norm": 0.8666252959292928,
"learning_rate": 7.205507594582994e-06,
"loss": 0.4924,
"step": 229
},
{
"epoch": 0.35742035742035744,
"grad_norm": 0.8422180634854972,
"learning_rate": 7.183557213043365e-06,
"loss": 0.5054,
"step": 230
},
{
"epoch": 0.358974358974359,
"grad_norm": 1.0386621335508512,
"learning_rate": 7.161554707050637e-06,
"loss": 0.6398,
"step": 231
},
{
"epoch": 0.3605283605283605,
"grad_norm": 0.8954415489915444,
"learning_rate": 7.1395006018342774e-06,
"loss": 0.4605,
"step": 232
},
{
"epoch": 0.3620823620823621,
"grad_norm": 0.8414488989233131,
"learning_rate": 7.117395423855496e-06,
"loss": 0.6271,
"step": 233
},
{
"epoch": 0.36363636363636365,
"grad_norm": 0.8451743034151997,
"learning_rate": 7.09523970079468e-06,
"loss": 0.4989,
"step": 234
},
{
"epoch": 0.3651903651903652,
"grad_norm": 0.9767891822448627,
"learning_rate": 7.073033961538793e-06,
"loss": 0.5808,
"step": 235
},
{
"epoch": 0.3667443667443667,
"grad_norm": 0.8524578887655285,
"learning_rate": 7.050778736168757e-06,
"loss": 0.5286,
"step": 236
},
{
"epoch": 0.3682983682983683,
"grad_norm": 0.8483585297089656,
"learning_rate": 7.028474555946787e-06,
"loss": 0.5654,
"step": 237
},
{
"epoch": 0.36985236985236986,
"grad_norm": 0.9267082807993728,
"learning_rate": 7.006121953303724e-06,
"loss": 0.4236,
"step": 238
},
{
"epoch": 0.3714063714063714,
"grad_norm": 0.8583744764085101,
"learning_rate": 6.983721461826312e-06,
"loss": 0.5422,
"step": 239
},
{
"epoch": 0.372960372960373,
"grad_norm": 0.9308888061772138,
"learning_rate": 6.9612736162444695e-06,
"loss": 0.647,
"step": 240
},
{
"epoch": 0.3745143745143745,
"grad_norm": 0.8721712511886593,
"learning_rate": 6.938778952418519e-06,
"loss": 0.5396,
"step": 241
},
{
"epoch": 0.37606837606837606,
"grad_norm": 1.0461384114911212,
"learning_rate": 6.916238007326399e-06,
"loss": 0.5289,
"step": 242
},
{
"epoch": 0.3776223776223776,
"grad_norm": 0.9117669935532693,
"learning_rate": 6.893651319050842e-06,
"loss": 0.5766,
"step": 243
},
{
"epoch": 0.3791763791763792,
"grad_norm": 0.8570178043159496,
"learning_rate": 6.871019426766537e-06,
"loss": 0.5391,
"step": 244
},
{
"epoch": 0.38073038073038074,
"grad_norm": 0.8073022614374971,
"learning_rate": 6.8483428707272456e-06,
"loss": 0.5389,
"step": 245
},
{
"epoch": 0.3822843822843823,
"grad_norm": 1.1407010013632595,
"learning_rate": 6.825622192252922e-06,
"loss": 0.3993,
"step": 246
},
{
"epoch": 0.3838383838383838,
"grad_norm": 0.7096109551356327,
"learning_rate": 6.802857933716774e-06,
"loss": 0.5799,
"step": 247
},
{
"epoch": 0.3853923853923854,
"grad_norm": 1.1253967474034594,
"learning_rate": 6.7800506385323335e-06,
"loss": 0.5872,
"step": 248
},
{
"epoch": 0.38694638694638694,
"grad_norm": 0.9340957791894847,
"learning_rate": 6.757200851140468e-06,
"loss": 0.5624,
"step": 249
},
{
"epoch": 0.3885003885003885,
"grad_norm": 1.1014437410465232,
"learning_rate": 6.734309116996392e-06,
"loss": 0.7582,
"step": 250
},
{
"epoch": 0.3900543900543901,
"grad_norm": 0.8699905581797581,
"learning_rate": 6.711375982556648e-06,
"loss": 0.4406,
"step": 251
},
{
"epoch": 0.3916083916083916,
"grad_norm": 1.173270780538771,
"learning_rate": 6.688401995266061e-06,
"loss": 0.6138,
"step": 252
},
{
"epoch": 0.39316239316239315,
"grad_norm": 1.0081768329864775,
"learning_rate": 6.665387703544661e-06,
"loss": 0.6273,
"step": 253
},
{
"epoch": 0.3947163947163947,
"grad_norm": 0.8819787015794277,
"learning_rate": 6.642333656774607e-06,
"loss": 0.4788,
"step": 254
},
{
"epoch": 0.3962703962703963,
"grad_norm": 0.9122685441057363,
"learning_rate": 6.61924040528706e-06,
"loss": 0.7072,
"step": 255
},
{
"epoch": 0.3978243978243978,
"grad_norm": 0.8791881328338158,
"learning_rate": 6.596108500349054e-06,
"loss": 0.6548,
"step": 256
},
{
"epoch": 0.39937839937839936,
"grad_norm": 0.784551616184446,
"learning_rate": 6.572938494150332e-06,
"loss": 0.5124,
"step": 257
},
{
"epoch": 0.40093240093240096,
"grad_norm": 1.0525063796221164,
"learning_rate": 6.54973093979016e-06,
"loss": 0.5484,
"step": 258
},
{
"epoch": 0.4024864024864025,
"grad_norm": 0.835003041239398,
"learning_rate": 6.526486391264137e-06,
"loss": 0.5737,
"step": 259
},
{
"epoch": 0.40404040404040403,
"grad_norm": 0.8834963101663964,
"learning_rate": 6.503205403450957e-06,
"loss": 0.5344,
"step": 260
},
{
"epoch": 0.40559440559440557,
"grad_norm": 0.9417120521658039,
"learning_rate": 6.479888532099175e-06,
"loss": 0.6799,
"step": 261
},
{
"epoch": 0.40714840714840717,
"grad_norm": 0.8617906130788676,
"learning_rate": 6.4565363338139245e-06,
"loss": 0.6621,
"step": 262
},
{
"epoch": 0.4087024087024087,
"grad_norm": 1.0509044094888915,
"learning_rate": 6.433149366043652e-06,
"loss": 0.6012,
"step": 263
},
{
"epoch": 0.41025641025641024,
"grad_norm": 1.1435254339427408,
"learning_rate": 6.409728187066789e-06,
"loss": 0.5096,
"step": 264
},
{
"epoch": 0.41181041181041184,
"grad_norm": 1.1457940839749883,
"learning_rate": 6.386273355978442e-06,
"loss": 0.5401,
"step": 265
},
{
"epoch": 0.4133644133644134,
"grad_norm": 2.0120271092743196,
"learning_rate": 6.3627854326770326e-06,
"loss": 0.5305,
"step": 266
},
{
"epoch": 0.4149184149184149,
"grad_norm": 0.8333683535180947,
"learning_rate": 6.339264977850943e-06,
"loss": 0.4673,
"step": 267
},
{
"epoch": 0.41647241647241645,
"grad_norm": 0.903528371130852,
"learning_rate": 6.3157125529651205e-06,
"loss": 0.5881,
"step": 268
},
{
"epoch": 0.41802641802641805,
"grad_norm": 0.8406474914833879,
"learning_rate": 6.292128720247692e-06,
"loss": 0.5169,
"step": 269
},
{
"epoch": 0.4195804195804196,
"grad_norm": 1.0599578458757946,
"learning_rate": 6.268514042676519e-06,
"loss": 0.7998,
"step": 270
},
{
"epoch": 0.4211344211344211,
"grad_norm": 0.8286229383890666,
"learning_rate": 6.244869083965777e-06,
"loss": 0.5183,
"step": 271
},
{
"epoch": 0.4226884226884227,
"grad_norm": 0.8288746254893705,
"learning_rate": 6.221194408552494e-06,
"loss": 0.5148,
"step": 272
},
{
"epoch": 0.42424242424242425,
"grad_norm": 0.8954699370556662,
"learning_rate": 6.197490581583078e-06,
"loss": 0.6395,
"step": 273
},
{
"epoch": 0.4257964257964258,
"grad_norm": 0.9215984961269827,
"learning_rate": 6.173758168899814e-06,
"loss": 0.5935,
"step": 274
},
{
"epoch": 0.42735042735042733,
"grad_norm": 0.9225474961665954,
"learning_rate": 6.149997737027377e-06,
"loss": 0.5906,
"step": 275
},
{
"epoch": 0.4289044289044289,
"grad_norm": 1.2221285719562045,
"learning_rate": 6.126209853159293e-06,
"loss": 0.8451,
"step": 276
},
{
"epoch": 0.43045843045843046,
"grad_norm": 0.7995085563421066,
"learning_rate": 6.102395085144406e-06,
"loss": 0.5054,
"step": 277
},
{
"epoch": 0.432012432012432,
"grad_norm": 1.0301315463512406,
"learning_rate": 6.078554001473317e-06,
"loss": 0.7042,
"step": 278
},
{
"epoch": 0.43356643356643354,
"grad_norm": 0.6966173508409831,
"learning_rate": 6.054687171264822e-06,
"loss": 0.5011,
"step": 279
},
{
"epoch": 0.43512043512043513,
"grad_norm": 0.8623621684914001,
"learning_rate": 6.030795164252321e-06,
"loss": 0.5034,
"step": 280
},
{
"epoch": 0.4366744366744367,
"grad_norm": 0.998337745266573,
"learning_rate": 6.006878550770213e-06,
"loss": 0.6222,
"step": 281
},
{
"epoch": 0.4382284382284382,
"grad_norm": 0.7872161913664657,
"learning_rate": 5.982937901740296e-06,
"loss": 0.6075,
"step": 282
},
{
"epoch": 0.4397824397824398,
"grad_norm": 1.1000342755543842,
"learning_rate": 5.958973788658115e-06,
"loss": 0.5029,
"step": 283
},
{
"epoch": 0.44133644133644134,
"grad_norm": 1.0312528408288124,
"learning_rate": 5.934986783579349e-06,
"loss": 0.5106,
"step": 284
},
{
"epoch": 0.4428904428904429,
"grad_norm": 0.9770392362683729,
"learning_rate": 5.91097745910613e-06,
"loss": 0.4993,
"step": 285
},
{
"epoch": 0.4444444444444444,
"grad_norm": 0.9659615323292092,
"learning_rate": 5.886946388373387e-06,
"loss": 0.5144,
"step": 286
},
{
"epoch": 0.445998445998446,
"grad_norm": 1.0973377248982952,
"learning_rate": 5.862894145035158e-06,
"loss": 0.5517,
"step": 287
},
{
"epoch": 0.44755244755244755,
"grad_norm": 1.0588242348941252,
"learning_rate": 5.83882130325091e-06,
"loss": 0.6267,
"step": 288
},
{
"epoch": 0.4491064491064491,
"grad_norm": 0.8267393911257279,
"learning_rate": 5.814728437671808e-06,
"loss": 0.5095,
"step": 289
},
{
"epoch": 0.4506604506604507,
"grad_norm": 0.9546909177431534,
"learning_rate": 5.7906161234270234e-06,
"loss": 0.5733,
"step": 290
},
{
"epoch": 0.4522144522144522,
"grad_norm": 0.8620531179333135,
"learning_rate": 5.766484936109987e-06,
"loss": 0.6206,
"step": 291
},
{
"epoch": 0.45376845376845376,
"grad_norm": 1.0724392623059507,
"learning_rate": 5.7423354517646616e-06,
"loss": 0.5701,
"step": 292
},
{
"epoch": 0.4553224553224553,
"grad_norm": 1.103019835893421,
"learning_rate": 5.718168246871775e-06,
"loss": 0.5365,
"step": 293
},
{
"epoch": 0.4568764568764569,
"grad_norm": 0.8804429160562605,
"learning_rate": 5.69398389833508e-06,
"loss": 0.4656,
"step": 294
},
{
"epoch": 0.45843045843045843,
"grad_norm": 0.728501083025643,
"learning_rate": 5.669782983467562e-06,
"loss": 0.4538,
"step": 295
},
{
"epoch": 0.45998445998445997,
"grad_norm": 0.89679619861348,
"learning_rate": 5.645566079977673e-06,
"loss": 0.5794,
"step": 296
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.8540763123009817,
"learning_rate": 5.621333765955529e-06,
"loss": 0.6715,
"step": 297
},
{
"epoch": 0.4630924630924631,
"grad_norm": 0.9215429357687969,
"learning_rate": 5.5970866198591235e-06,
"loss": 0.4903,
"step": 298
},
{
"epoch": 0.46464646464646464,
"grad_norm": 0.971111182369051,
"learning_rate": 5.572825220500505e-06,
"loss": 0.5717,
"step": 299
},
{
"epoch": 0.4662004662004662,
"grad_norm": 0.7525898375498171,
"learning_rate": 5.548550147031971e-06,
"loss": 0.5359,
"step": 300
},
{
"epoch": 0.4677544677544678,
"grad_norm": 0.956681429994755,
"learning_rate": 5.524261978932234e-06,
"loss": 0.5076,
"step": 301
},
{
"epoch": 0.4693084693084693,
"grad_norm": 0.8724302136744727,
"learning_rate": 5.4999612959925995e-06,
"loss": 0.5158,
"step": 302
},
{
"epoch": 0.47086247086247085,
"grad_norm": 0.7690139881658553,
"learning_rate": 5.475648678303112e-06,
"loss": 0.6411,
"step": 303
},
{
"epoch": 0.47241647241647244,
"grad_norm": 1.044841230716274,
"learning_rate": 5.451324706238721e-06,
"loss": 0.515,
"step": 304
},
{
"epoch": 0.473970473970474,
"grad_norm": 0.7250530480039049,
"learning_rate": 5.426989960445415e-06,
"loss": 0.486,
"step": 305
},
{
"epoch": 0.4755244755244755,
"grad_norm": 0.9555846587094475,
"learning_rate": 5.402645021826367e-06,
"loss": 0.4873,
"step": 306
},
{
"epoch": 0.47707847707847706,
"grad_norm": 0.8949848630568665,
"learning_rate": 5.3782904715280705e-06,
"loss": 0.4881,
"step": 307
},
{
"epoch": 0.47863247863247865,
"grad_norm": 0.9734141387829424,
"learning_rate": 5.35392689092646e-06,
"loss": 0.6362,
"step": 308
},
{
"epoch": 0.4801864801864802,
"grad_norm": 0.9432639944738404,
"learning_rate": 5.329554861613031e-06,
"loss": 0.558,
"step": 309
},
{
"epoch": 0.48174048174048173,
"grad_norm": 0.926110226229468,
"learning_rate": 5.3051749653809685e-06,
"loss": 0.6836,
"step": 310
},
{
"epoch": 0.48329448329448327,
"grad_norm": 0.8306639970751976,
"learning_rate": 5.2807877842112475e-06,
"loss": 0.5398,
"step": 311
},
{
"epoch": 0.48484848484848486,
"grad_norm": 0.8451961871886501,
"learning_rate": 5.256393900258747e-06,
"loss": 0.5038,
"step": 312
},
{
"epoch": 0.4864024864024864,
"grad_norm": 0.92675087789733,
"learning_rate": 5.231993895838348e-06,
"loss": 0.5405,
"step": 313
},
{
"epoch": 0.48795648795648794,
"grad_norm": 0.897235258445696,
"learning_rate": 5.207588353411032e-06,
"loss": 0.5597,
"step": 314
},
{
"epoch": 0.48951048951048953,
"grad_norm": 0.8645857836074616,
"learning_rate": 5.183177855569989e-06,
"loss": 0.605,
"step": 315
},
{
"epoch": 0.49106449106449107,
"grad_norm": 0.8652623075328669,
"learning_rate": 5.158762985026694e-06,
"loss": 0.6347,
"step": 316
},
{
"epoch": 0.4926184926184926,
"grad_norm": 0.974238117866376,
"learning_rate": 5.1343443245970095e-06,
"loss": 0.5848,
"step": 317
},
{
"epoch": 0.49417249417249415,
"grad_norm": 1.0702212280059262,
"learning_rate": 5.10992245718726e-06,
"loss": 0.7456,
"step": 318
},
{
"epoch": 0.49572649572649574,
"grad_norm": 0.9332991045536498,
"learning_rate": 5.085497965780335e-06,
"loss": 0.5094,
"step": 319
},
{
"epoch": 0.4972804972804973,
"grad_norm": 0.7666822911499076,
"learning_rate": 5.061071433421754e-06,
"loss": 0.4988,
"step": 320
},
{
"epoch": 0.4988344988344988,
"grad_norm": 1.0306498703517015,
"learning_rate": 5.0366434432057624e-06,
"loss": 0.5804,
"step": 321
},
{
"epoch": 0.5003885003885004,
"grad_norm": 0.8175426737536865,
"learning_rate": 5.012214578261402e-06,
"loss": 0.5078,
"step": 322
},
{
"epoch": 0.5019425019425019,
"grad_norm": 1.1437884320473477,
"learning_rate": 4.987785421738599e-06,
"loss": 0.5265,
"step": 323
},
{
"epoch": 0.5034965034965035,
"grad_norm": 0.9884279600512011,
"learning_rate": 4.963356556794238e-06,
"loss": 0.7306,
"step": 324
},
{
"epoch": 0.5050505050505051,
"grad_norm": 1.0273572672750226,
"learning_rate": 4.938928566578247e-06,
"loss": 0.5403,
"step": 325
},
{
"epoch": 0.5066045066045066,
"grad_norm": 0.9211672996812349,
"learning_rate": 4.914502034219667e-06,
"loss": 0.6839,
"step": 326
},
{
"epoch": 0.5081585081585082,
"grad_norm": 0.8357279664928383,
"learning_rate": 4.890077542812742e-06,
"loss": 0.4687,
"step": 327
},
{
"epoch": 0.5097125097125097,
"grad_norm": 1.01851904270436,
"learning_rate": 4.865655675402993e-06,
"loss": 0.5688,
"step": 328
},
{
"epoch": 0.5112665112665112,
"grad_norm": 0.9530277794856284,
"learning_rate": 4.841237014973305e-06,
"loss": 0.6302,
"step": 329
},
{
"epoch": 0.5128205128205128,
"grad_norm": 0.8945625499429839,
"learning_rate": 4.8168221444300124e-06,
"loss": 0.5748,
"step": 330
},
{
"epoch": 0.5143745143745144,
"grad_norm": 0.8213408851658214,
"learning_rate": 4.7924116465889684e-06,
"loss": 0.4851,
"step": 331
},
{
"epoch": 0.515928515928516,
"grad_norm": 0.8909942791600822,
"learning_rate": 4.768006104161655e-06,
"loss": 0.6238,
"step": 332
},
{
"epoch": 0.5174825174825175,
"grad_norm": 0.8715309540453933,
"learning_rate": 4.743606099741255e-06,
"loss": 0.4499,
"step": 333
},
{
"epoch": 0.519036519036519,
"grad_norm": 0.9267575769108474,
"learning_rate": 4.719212215788753e-06,
"loss": 0.5537,
"step": 334
},
{
"epoch": 0.5205905205905206,
"grad_norm": 1.0242507045067903,
"learning_rate": 4.6948250346190315e-06,
"loss": 0.5579,
"step": 335
},
{
"epoch": 0.5221445221445221,
"grad_norm": 0.8168789723510115,
"learning_rate": 4.670445138386971e-06,
"loss": 0.6091,
"step": 336
},
{
"epoch": 0.5236985236985237,
"grad_norm": 0.9545835186230072,
"learning_rate": 4.646073109073542e-06,
"loss": 0.5246,
"step": 337
},
{
"epoch": 0.5252525252525253,
"grad_norm": 1.072717728215519,
"learning_rate": 4.621709528471931e-06,
"loss": 0.4918,
"step": 338
},
{
"epoch": 0.5268065268065268,
"grad_norm": 0.8183513806382199,
"learning_rate": 4.5973549781736335e-06,
"loss": 0.491,
"step": 339
},
{
"epoch": 0.5283605283605284,
"grad_norm": 0.9683292063294122,
"learning_rate": 4.573010039554587e-06,
"loss": 0.6359,
"step": 340
},
{
"epoch": 0.5299145299145299,
"grad_norm": 1.0264839081727954,
"learning_rate": 4.548675293761281e-06,
"loss": 0.655,
"step": 341
},
{
"epoch": 0.5314685314685315,
"grad_norm": 0.7361166824699995,
"learning_rate": 4.524351321696889e-06,
"loss": 0.4948,
"step": 342
},
{
"epoch": 0.533022533022533,
"grad_norm": 1.1294660407885464,
"learning_rate": 4.500038704007402e-06,
"loss": 0.5751,
"step": 343
},
{
"epoch": 0.5345765345765345,
"grad_norm": 0.934407670174655,
"learning_rate": 4.475738021067768e-06,
"loss": 0.5434,
"step": 344
},
{
"epoch": 0.5361305361305362,
"grad_norm": 0.9178261235495052,
"learning_rate": 4.451449852968031e-06,
"loss": 0.5195,
"step": 345
},
{
"epoch": 0.5376845376845377,
"grad_norm": 0.9435990884434239,
"learning_rate": 4.427174779499498e-06,
"loss": 0.7334,
"step": 346
},
{
"epoch": 0.5392385392385393,
"grad_norm": 0.7911445940026536,
"learning_rate": 4.402913380140878e-06,
"loss": 0.5774,
"step": 347
},
{
"epoch": 0.5407925407925408,
"grad_norm": 1.0017027537663632,
"learning_rate": 4.378666234044471e-06,
"loss": 0.5454,
"step": 348
},
{
"epoch": 0.5423465423465423,
"grad_norm": 1.063814328831088,
"learning_rate": 4.354433920022328e-06,
"loss": 0.5098,
"step": 349
},
{
"epoch": 0.5439005439005439,
"grad_norm": 0.8748299244500228,
"learning_rate": 4.3302170165324385e-06,
"loss": 0.5449,
"step": 350
},
{
"epoch": 0.5454545454545454,
"grad_norm": 0.8369440203959997,
"learning_rate": 4.306016101664921e-06,
"loss": 0.5319,
"step": 351
},
{
"epoch": 0.5470085470085471,
"grad_norm": 1.255802499656197,
"learning_rate": 4.281831753128226e-06,
"loss": 0.4056,
"step": 352
},
{
"epoch": 0.5485625485625486,
"grad_norm": 1.4515615573152536,
"learning_rate": 4.257664548235341e-06,
"loss": 0.7249,
"step": 353
},
{
"epoch": 0.5501165501165501,
"grad_norm": 0.8907594786465234,
"learning_rate": 4.233515063890013e-06,
"loss": 0.6057,
"step": 354
},
{
"epoch": 0.5516705516705517,
"grad_norm": 1.1763649838296555,
"learning_rate": 4.209383876572977e-06,
"loss": 0.5774,
"step": 355
},
{
"epoch": 0.5532245532245532,
"grad_norm": 0.9284584032037477,
"learning_rate": 4.1852715623281934e-06,
"loss": 0.4249,
"step": 356
},
{
"epoch": 0.5547785547785548,
"grad_norm": 1.038926008258523,
"learning_rate": 4.161178696749092e-06,
"loss": 0.5849,
"step": 357
},
{
"epoch": 0.5563325563325563,
"grad_norm": 0.9584747895306955,
"learning_rate": 4.1371058549648425e-06,
"loss": 0.5939,
"step": 358
},
{
"epoch": 0.5578865578865578,
"grad_norm": 0.8432409707491679,
"learning_rate": 4.1130536116266155e-06,
"loss": 0.5545,
"step": 359
},
{
"epoch": 0.5594405594405595,
"grad_norm": 0.8374098243309068,
"learning_rate": 4.089022540893871e-06,
"loss": 0.5547,
"step": 360
},
{
"epoch": 0.560994560994561,
"grad_norm": 0.9847581566777266,
"learning_rate": 4.0650132164206515e-06,
"loss": 0.5344,
"step": 361
},
{
"epoch": 0.5625485625485626,
"grad_norm": 0.7534590052741679,
"learning_rate": 4.041026211341886e-06,
"loss": 0.3433,
"step": 362
},
{
"epoch": 0.5641025641025641,
"grad_norm": 0.8196558006004733,
"learning_rate": 4.017062098259707e-06,
"loss": 0.4477,
"step": 363
},
{
"epoch": 0.5656565656565656,
"grad_norm": 0.8447608563383764,
"learning_rate": 3.9931214492297875e-06,
"loss": 0.6335,
"step": 364
},
{
"epoch": 0.5672105672105672,
"grad_norm": 0.754934359719861,
"learning_rate": 3.969204835747681e-06,
"loss": 0.5477,
"step": 365
},
{
"epoch": 0.5687645687645687,
"grad_norm": 0.8638053344127238,
"learning_rate": 3.945312828735179e-06,
"loss": 0.6333,
"step": 366
},
{
"epoch": 0.5703185703185704,
"grad_norm": 0.8784017928990301,
"learning_rate": 3.921445998526684e-06,
"loss": 0.5041,
"step": 367
},
{
"epoch": 0.5718725718725719,
"grad_norm": 0.9029724643591418,
"learning_rate": 3.897604914855596e-06,
"loss": 0.5241,
"step": 368
},
{
"epoch": 0.5734265734265734,
"grad_norm": 0.9069716405204118,
"learning_rate": 3.873790146840709e-06,
"loss": 0.501,
"step": 369
},
{
"epoch": 0.574980574980575,
"grad_norm": 0.9342368491052493,
"learning_rate": 3.8500022629726246e-06,
"loss": 0.3958,
"step": 370
},
{
"epoch": 0.5765345765345765,
"grad_norm": 0.896012165148463,
"learning_rate": 3.8262418311001884e-06,
"loss": 0.6129,
"step": 371
},
{
"epoch": 0.578088578088578,
"grad_norm": 1.0159020474194678,
"learning_rate": 3.8025094184169254e-06,
"loss": 0.5708,
"step": 372
},
{
"epoch": 0.5796425796425796,
"grad_norm": 0.8566891987710636,
"learning_rate": 3.778805591447505e-06,
"loss": 0.424,
"step": 373
},
{
"epoch": 0.5811965811965812,
"grad_norm": 0.8831453832251531,
"learning_rate": 3.7551309160342233e-06,
"loss": 0.4455,
"step": 374
},
{
"epoch": 0.5827505827505828,
"grad_norm": 0.8421527973531183,
"learning_rate": 3.731485957323483e-06,
"loss": 0.5675,
"step": 375
},
{
"epoch": 0.5843045843045843,
"grad_norm": 1.2979562207506874,
"learning_rate": 3.707871279752309e-06,
"loss": 0.6143,
"step": 376
},
{
"epoch": 0.5858585858585859,
"grad_norm": 0.995197386244874,
"learning_rate": 3.68428744703488e-06,
"loss": 0.4698,
"step": 377
},
{
"epoch": 0.5874125874125874,
"grad_norm": 1.1717778711933917,
"learning_rate": 3.6607350221490593e-06,
"loss": 0.6312,
"step": 378
},
{
"epoch": 0.5889665889665889,
"grad_norm": 0.8374964354419717,
"learning_rate": 3.6372145673229683e-06,
"loss": 0.5194,
"step": 379
},
{
"epoch": 0.5905205905205905,
"grad_norm": 0.7942047183235935,
"learning_rate": 3.613726644021559e-06,
"loss": 0.4238,
"step": 380
},
{
"epoch": 0.5920745920745921,
"grad_norm": 0.773794273550258,
"learning_rate": 3.590271812933212e-06,
"loss": 0.5011,
"step": 381
},
{
"epoch": 0.5936285936285937,
"grad_norm": 0.8912405571204569,
"learning_rate": 3.5668506339563502e-06,
"loss": 0.4961,
"step": 382
},
{
"epoch": 0.5951825951825952,
"grad_norm": 2.0686952933355705,
"learning_rate": 3.5434636661860776e-06,
"loss": 0.4724,
"step": 383
},
{
"epoch": 0.5967365967365967,
"grad_norm": 1.0721561025968982,
"learning_rate": 3.5201114679008286e-06,
"loss": 0.643,
"step": 384
},
{
"epoch": 0.5982905982905983,
"grad_norm": 0.9615898602828296,
"learning_rate": 3.4967945965490434e-06,
"loss": 0.4899,
"step": 385
},
{
"epoch": 0.5998445998445998,
"grad_norm": 0.8443768517564723,
"learning_rate": 3.4735136087358646e-06,
"loss": 0.5816,
"step": 386
},
{
"epoch": 0.6013986013986014,
"grad_norm": 0.8687572364511918,
"learning_rate": 3.450269060209841e-06,
"loss": 0.6585,
"step": 387
},
{
"epoch": 0.602952602952603,
"grad_norm": 0.7411207969719316,
"learning_rate": 3.42706150584967e-06,
"loss": 0.5498,
"step": 388
},
{
"epoch": 0.6045066045066045,
"grad_norm": 0.8947980443263261,
"learning_rate": 3.4038914996509464e-06,
"loss": 0.6219,
"step": 389
},
{
"epoch": 0.6060606060606061,
"grad_norm": 0.9457854814078657,
"learning_rate": 3.3807595947129405e-06,
"loss": 0.4819,
"step": 390
},
{
"epoch": 0.6076146076146076,
"grad_norm": 0.8076353901243662,
"learning_rate": 3.357666343225396e-06,
"loss": 0.4917,
"step": 391
},
{
"epoch": 0.6091686091686092,
"grad_norm": 0.9568999807001125,
"learning_rate": 3.3346122964553407e-06,
"loss": 0.3623,
"step": 392
},
{
"epoch": 0.6107226107226107,
"grad_norm": 0.9749020198956558,
"learning_rate": 3.3115980047339415e-06,
"loss": 0.6806,
"step": 393
},
{
"epoch": 0.6122766122766122,
"grad_norm": 0.9565776234695726,
"learning_rate": 3.288624017443353e-06,
"loss": 0.5648,
"step": 394
},
{
"epoch": 0.6138306138306139,
"grad_norm": 0.8372606567529206,
"learning_rate": 3.265690883003609e-06,
"loss": 0.6041,
"step": 395
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.9657923180658315,
"learning_rate": 3.2427991488595334e-06,
"loss": 0.5534,
"step": 396
},
{
"epoch": 0.616938616938617,
"grad_norm": 0.779796777254401,
"learning_rate": 3.219949361467668e-06,
"loss": 0.4895,
"step": 397
},
{
"epoch": 0.6184926184926185,
"grad_norm": 1.0023947035662037,
"learning_rate": 3.197142066283225e-06,
"loss": 0.5017,
"step": 398
},
{
"epoch": 0.62004662004662,
"grad_norm": 0.968677826181818,
"learning_rate": 3.174377807747079e-06,
"loss": 0.6008,
"step": 399
},
{
"epoch": 0.6216006216006216,
"grad_norm": 1.1511399987075677,
"learning_rate": 3.1516571292727553e-06,
"loss": 0.5573,
"step": 400
},
{
"epoch": 0.6231546231546231,
"grad_norm": 0.9562397655581797,
"learning_rate": 3.128980573233465e-06,
"loss": 0.678,
"step": 401
},
{
"epoch": 0.6247086247086248,
"grad_norm": 1.0314415028331958,
"learning_rate": 3.1063486809491595e-06,
"loss": 0.5402,
"step": 402
},
{
"epoch": 0.6262626262626263,
"grad_norm": 0.8869287498009347,
"learning_rate": 3.0837619926736027e-06,
"loss": 0.4637,
"step": 403
},
{
"epoch": 0.6278166278166278,
"grad_norm": 0.915327973200595,
"learning_rate": 3.061221047581482e-06,
"loss": 0.4994,
"step": 404
},
{
"epoch": 0.6293706293706294,
"grad_norm": 0.8673891612365153,
"learning_rate": 3.038726383755531e-06,
"loss": 0.5784,
"step": 405
},
{
"epoch": 0.6309246309246309,
"grad_norm": 0.8149059444255452,
"learning_rate": 3.0162785381736893e-06,
"loss": 0.487,
"step": 406
},
{
"epoch": 0.6324786324786325,
"grad_norm": 1.2317767684482286,
"learning_rate": 2.9938780466962768e-06,
"loss": 0.6945,
"step": 407
},
{
"epoch": 0.634032634032634,
"grad_norm": 0.8114884457654089,
"learning_rate": 2.9715254440532147e-06,
"loss": 0.4866,
"step": 408
},
{
"epoch": 0.6355866355866356,
"grad_norm": 1.0025564372683453,
"learning_rate": 2.9492212638312458e-06,
"loss": 0.4607,
"step": 409
},
{
"epoch": 0.6371406371406372,
"grad_norm": 0.7375496261692214,
"learning_rate": 2.9269660384612064e-06,
"loss": 0.5557,
"step": 410
},
{
"epoch": 0.6386946386946387,
"grad_norm": 0.7955765463661515,
"learning_rate": 2.90476029920532e-06,
"loss": 0.4934,
"step": 411
},
{
"epoch": 0.6402486402486403,
"grad_norm": 0.8629641026853778,
"learning_rate": 2.882604576144505e-06,
"loss": 0.6129,
"step": 412
},
{
"epoch": 0.6418026418026418,
"grad_norm": 0.9757114529523541,
"learning_rate": 2.8604993981657247e-06,
"loss": 0.4185,
"step": 413
},
{
"epoch": 0.6433566433566433,
"grad_norm": 0.7834021215202662,
"learning_rate": 2.8384452929493645e-06,
"loss": 0.3618,
"step": 414
},
{
"epoch": 0.6449106449106449,
"grad_norm": 1.0670630280828879,
"learning_rate": 2.8164427869566367e-06,
"loss": 0.5638,
"step": 415
},
{
"epoch": 0.6464646464646465,
"grad_norm": 0.8113716929371272,
"learning_rate": 2.7944924054170087e-06,
"loss": 0.5776,
"step": 416
},
{
"epoch": 0.6480186480186481,
"grad_norm": 0.9957638811754096,
"learning_rate": 2.7725946723156626e-06,
"loss": 0.4072,
"step": 417
},
{
"epoch": 0.6495726495726496,
"grad_norm": 0.8596642587969949,
"learning_rate": 2.750750110381001e-06,
"loss": 0.5096,
"step": 418
},
{
"epoch": 0.6511266511266511,
"grad_norm": 1.0292360670108274,
"learning_rate": 2.728959241072149e-06,
"loss": 0.4778,
"step": 419
},
{
"epoch": 0.6526806526806527,
"grad_norm": 1.2070046089694704,
"learning_rate": 2.7072225845665256e-06,
"loss": 0.5104,
"step": 420
},
{
"epoch": 0.6542346542346542,
"grad_norm": 0.7633836364809529,
"learning_rate": 2.6855406597474098e-06,
"loss": 0.5486,
"step": 421
},
{
"epoch": 0.6557886557886557,
"grad_norm": 1.1041806227920956,
"learning_rate": 2.6639139841915628e-06,
"loss": 0.4444,
"step": 422
},
{
"epoch": 0.6573426573426573,
"grad_norm": 0.9575125133042081,
"learning_rate": 2.6423430741568746e-06,
"loss": 0.4381,
"step": 423
},
{
"epoch": 0.6588966588966589,
"grad_norm": 0.9800662598205424,
"learning_rate": 2.6208284445700373e-06,
"loss": 0.5281,
"step": 424
},
{
"epoch": 0.6604506604506605,
"grad_norm": 0.7475115086795047,
"learning_rate": 2.5993706090142484e-06,
"loss": 0.5738,
"step": 425
},
{
"epoch": 0.662004662004662,
"grad_norm": 0.7957543863519091,
"learning_rate": 2.577970079716959e-06,
"loss": 0.5223,
"step": 426
},
{
"epoch": 0.6635586635586636,
"grad_norm": 0.8150818241751017,
"learning_rate": 2.5566273675376386e-06,
"loss": 0.4616,
"step": 427
},
{
"epoch": 0.6651126651126651,
"grad_norm": 0.867516029458422,
"learning_rate": 2.535342981955591e-06,
"loss": 0.5592,
"step": 428
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.8202323607458766,
"learning_rate": 2.5141174310577774e-06,
"loss": 0.4761,
"step": 429
},
{
"epoch": 0.6682206682206682,
"grad_norm": 1.0531709032808028,
"learning_rate": 2.492951221526705e-06,
"loss": 0.5706,
"step": 430
},
{
"epoch": 0.6697746697746698,
"grad_norm": 0.7925613983131308,
"learning_rate": 2.4718448586283126e-06,
"loss": 0.4943,
"step": 431
},
{
"epoch": 0.6713286713286714,
"grad_norm": 1.015333414269206,
"learning_rate": 2.4507988461999283e-06,
"loss": 0.547,
"step": 432
},
{
"epoch": 0.6728826728826729,
"grad_norm": 1.0021582367745274,
"learning_rate": 2.429813686638227e-06,
"loss": 0.6319,
"step": 433
},
{
"epoch": 0.6744366744366744,
"grad_norm": 1.2720975755074075,
"learning_rate": 2.408889880887246e-06,
"loss": 0.4789,
"step": 434
},
{
"epoch": 0.675990675990676,
"grad_norm": 0.7759116382434137,
"learning_rate": 2.38802792842642e-06,
"loss": 0.5063,
"step": 435
},
{
"epoch": 0.6775446775446775,
"grad_norm": 0.7897990263143547,
"learning_rate": 2.3672283272586745e-06,
"loss": 0.6591,
"step": 436
},
{
"epoch": 0.679098679098679,
"grad_norm": 0.8072958152033197,
"learning_rate": 2.346491573898513e-06,
"loss": 0.4612,
"step": 437
},
{
"epoch": 0.6806526806526807,
"grad_norm": 0.7877836535078193,
"learning_rate": 2.3258181633601836e-06,
"loss": 0.4869,
"step": 438
},
{
"epoch": 0.6822066822066822,
"grad_norm": 0.8902691924995161,
"learning_rate": 2.30520858914585e-06,
"loss": 0.5416,
"step": 439
},
{
"epoch": 0.6837606837606838,
"grad_norm": 0.6369068921434858,
"learning_rate": 2.2846633432338256e-06,
"loss": 0.4979,
"step": 440
},
{
"epoch": 0.6853146853146853,
"grad_norm": 0.8625022953005387,
"learning_rate": 2.2641829160668137e-06,
"loss": 0.5279,
"step": 441
},
{
"epoch": 0.6868686868686869,
"grad_norm": 0.8374815085950136,
"learning_rate": 2.243767796540207e-06,
"loss": 0.5359,
"step": 442
},
{
"epoch": 0.6884226884226884,
"grad_norm": 1.0648834804711098,
"learning_rate": 2.223418471990421e-06,
"loss": 0.6371,
"step": 443
},
{
"epoch": 0.6899766899766899,
"grad_norm": 1.3640624616380017,
"learning_rate": 2.2031354281832555e-06,
"loss": 0.7164,
"step": 444
},
{
"epoch": 0.6915306915306916,
"grad_norm": 0.9650947768114302,
"learning_rate": 2.1829191493022974e-06,
"loss": 0.4751,
"step": 445
},
{
"epoch": 0.6930846930846931,
"grad_norm": 0.9507196966289148,
"learning_rate": 2.1627701179373645e-06,
"loss": 0.5319,
"step": 446
},
{
"epoch": 0.6946386946386947,
"grad_norm": 1.1406920101918319,
"learning_rate": 2.142688815072986e-06,
"loss": 0.6378,
"step": 447
},
{
"epoch": 0.6961926961926962,
"grad_norm": 0.9181843236568763,
"learning_rate": 2.1226757200769225e-06,
"loss": 0.428,
"step": 448
},
{
"epoch": 0.6977466977466977,
"grad_norm": 1.2092134761120403,
"learning_rate": 2.102731310688723e-06,
"loss": 0.5284,
"step": 449
},
{
"epoch": 0.6993006993006993,
"grad_norm": 1.1089425350721833,
"learning_rate": 2.0828560630083127e-06,
"loss": 0.5574,
"step": 450
},
{
"epoch": 0.7008547008547008,
"grad_norm": 0.6997904053981342,
"learning_rate": 2.0630504514846372e-06,
"loss": 0.582,
"step": 451
},
{
"epoch": 0.7024087024087025,
"grad_norm": 0.8680833720150671,
"learning_rate": 2.0433149489043296e-06,
"loss": 0.4877,
"step": 452
},
{
"epoch": 0.703962703962704,
"grad_norm": 1.0412111329827756,
"learning_rate": 2.0236500263804355e-06,
"loss": 0.5206,
"step": 453
},
{
"epoch": 0.7055167055167055,
"grad_norm": 0.9511530379710712,
"learning_rate": 2.0040561533411494e-06,
"loss": 0.5571,
"step": 454
},
{
"epoch": 0.7070707070707071,
"grad_norm": 0.9857137048304959,
"learning_rate": 1.9845337975186297e-06,
"loss": 0.5947,
"step": 455
},
{
"epoch": 0.7086247086247086,
"grad_norm": 0.7358056924894385,
"learning_rate": 1.9650834249378125e-06,
"loss": 0.446,
"step": 456
},
{
"epoch": 0.7101787101787101,
"grad_norm": 0.9283899065701541,
"learning_rate": 1.945705499905305e-06,
"loss": 0.5944,
"step": 457
},
{
"epoch": 0.7117327117327117,
"grad_norm": 0.9985343909554646,
"learning_rate": 1.926400484998289e-06,
"loss": 0.6292,
"step": 458
},
{
"epoch": 0.7132867132867133,
"grad_norm": 0.8149487764948858,
"learning_rate": 1.9071688410534828e-06,
"loss": 0.4852,
"step": 459
},
{
"epoch": 0.7148407148407149,
"grad_norm": 1.0046527442068103,
"learning_rate": 1.8880110271561415e-06,
"loss": 0.4948,
"step": 460
},
{
"epoch": 0.7163947163947164,
"grad_norm": 0.7380450104665942,
"learning_rate": 1.8689275006291035e-06,
"loss": 0.5346,
"step": 461
},
{
"epoch": 0.717948717948718,
"grad_norm": 0.9071832673516289,
"learning_rate": 1.8499187170218614e-06,
"loss": 0.6154,
"step": 462
},
{
"epoch": 0.7195027195027195,
"grad_norm": 0.885237217235856,
"learning_rate": 1.8309851300996934e-06,
"loss": 0.546,
"step": 463
},
{
"epoch": 0.721056721056721,
"grad_norm": 0.8643676814843989,
"learning_rate": 1.8121271918328314e-06,
"loss": 0.5854,
"step": 464
},
{
"epoch": 0.7226107226107226,
"grad_norm": 0.7637623236114972,
"learning_rate": 1.793345352385678e-06,
"loss": 0.5339,
"step": 465
},
{
"epoch": 0.7241647241647242,
"grad_norm": 0.920346480267907,
"learning_rate": 1.7746400601060476e-06,
"loss": 0.6024,
"step": 466
},
{
"epoch": 0.7257187257187258,
"grad_norm": 0.8108469777726085,
"learning_rate": 1.7560117615144717e-06,
"loss": 0.4749,
"step": 467
},
{
"epoch": 0.7272727272727273,
"grad_norm": 0.7229261666322232,
"learning_rate": 1.7374609012935412e-06,
"loss": 0.496,
"step": 468
},
{
"epoch": 0.7288267288267288,
"grad_norm": 0.8134361172252622,
"learning_rate": 1.7189879222772894e-06,
"loss": 0.4782,
"step": 469
},
{
"epoch": 0.7303807303807304,
"grad_norm": 1.1199242105205238,
"learning_rate": 1.7005932654406165e-06,
"loss": 0.5725,
"step": 470
},
{
"epoch": 0.7319347319347319,
"grad_norm": 0.9531044801518151,
"learning_rate": 1.682277369888769e-06,
"loss": 0.3771,
"step": 471
},
{
"epoch": 0.7334887334887334,
"grad_norm": 0.8181386506919823,
"learning_rate": 1.6640406728468534e-06,
"loss": 0.4525,
"step": 472
},
{
"epoch": 0.7350427350427351,
"grad_norm": 0.9822241425856895,
"learning_rate": 1.6458836096494046e-06,
"loss": 0.5829,
"step": 473
},
{
"epoch": 0.7365967365967366,
"grad_norm": 0.9429402946349031,
"learning_rate": 1.6278066137299898e-06,
"loss": 0.6585,
"step": 474
},
{
"epoch": 0.7381507381507382,
"grad_norm": 0.8416072050437153,
"learning_rate": 1.6098101166108593e-06,
"loss": 0.4289,
"step": 475
},
{
"epoch": 0.7397047397047397,
"grad_norm": 0.8273010419688407,
"learning_rate": 1.5918945478926484e-06,
"loss": 0.515,
"step": 476
},
{
"epoch": 0.7412587412587412,
"grad_norm": 0.85924664459265,
"learning_rate": 1.5740603352441281e-06,
"loss": 0.4741,
"step": 477
},
{
"epoch": 0.7428127428127428,
"grad_norm": 0.9930289461569045,
"learning_rate": 1.5563079043919843e-06,
"loss": 0.548,
"step": 478
},
{
"epoch": 0.7443667443667443,
"grad_norm": 0.7836995285858434,
"learning_rate": 1.5386376791106627e-06,
"loss": 0.4101,
"step": 479
},
{
"epoch": 0.745920745920746,
"grad_norm": 0.8342511178476052,
"learning_rate": 1.5210500812122548e-06,
"loss": 0.4793,
"step": 480
},
{
"epoch": 0.7474747474747475,
"grad_norm": 1.0560959113411836,
"learning_rate": 1.5035455305364188e-06,
"loss": 0.5464,
"step": 481
},
{
"epoch": 0.749028749028749,
"grad_norm": 1.039285125963634,
"learning_rate": 1.4861244449403717e-06,
"loss": 0.3743,
"step": 482
},
{
"epoch": 0.7505827505827506,
"grad_norm": 0.7525227170176826,
"learning_rate": 1.4687872402888991e-06,
"loss": 0.6774,
"step": 483
},
{
"epoch": 0.7521367521367521,
"grad_norm": 1.0954674880778155,
"learning_rate": 1.451534330444438e-06,
"loss": 0.5184,
"step": 484
},
{
"epoch": 0.7536907536907537,
"grad_norm": 0.7965815930286931,
"learning_rate": 1.4343661272571967e-06,
"loss": 0.4797,
"step": 485
},
{
"epoch": 0.7552447552447552,
"grad_norm": 0.9801653007138387,
"learning_rate": 1.4172830405553216e-06,
"loss": 0.7127,
"step": 486
},
{
"epoch": 0.7567987567987567,
"grad_norm": 0.7911339013108166,
"learning_rate": 1.4002854781351104e-06,
"loss": 0.6389,
"step": 487
},
{
"epoch": 0.7583527583527584,
"grad_norm": 1.0716997870139735,
"learning_rate": 1.3833738457512842e-06,
"loss": 0.4308,
"step": 488
},
{
"epoch": 0.7599067599067599,
"grad_norm": 0.8538650301172175,
"learning_rate": 1.3665485471072937e-06,
"loss": 0.5437,
"step": 489
},
{
"epoch": 0.7614607614607615,
"grad_norm": 0.9268324110689813,
"learning_rate": 1.3498099838456947e-06,
"loss": 0.5925,
"step": 490
},
{
"epoch": 0.763014763014763,
"grad_norm": 1.0985563955045539,
"learning_rate": 1.3331585555385458e-06,
"loss": 0.5372,
"step": 491
},
{
"epoch": 0.7645687645687645,
"grad_norm": 0.8917025830540084,
"learning_rate": 1.3165946596778773e-06,
"loss": 0.563,
"step": 492
},
{
"epoch": 0.7661227661227661,
"grad_norm": 0.9138600611635691,
"learning_rate": 1.3001186916662066e-06,
"loss": 0.656,
"step": 493
},
{
"epoch": 0.7676767676767676,
"grad_norm": 0.7678531340040682,
"learning_rate": 1.2837310448070929e-06,
"loss": 0.4876,
"step": 494
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.8377909993364742,
"learning_rate": 1.2674321102957476e-06,
"loss": 0.4719,
"step": 495
},
{
"epoch": 0.7707847707847708,
"grad_norm": 0.9620034622741296,
"learning_rate": 1.251222277209702e-06,
"loss": 0.638,
"step": 496
},
{
"epoch": 0.7723387723387724,
"grad_norm": 0.8691503561119966,
"learning_rate": 1.2351019324995128e-06,
"loss": 0.5645,
"step": 497
},
{
"epoch": 0.7738927738927739,
"grad_norm": 1.019242268045496,
"learning_rate": 1.2190714609795334e-06,
"loss": 0.5954,
"step": 498
},
{
"epoch": 0.7754467754467754,
"grad_norm": 0.8600773262545407,
"learning_rate": 1.203131245318721e-06,
"loss": 0.4787,
"step": 499
},
{
"epoch": 0.777000777000777,
"grad_norm": 0.7582051607025506,
"learning_rate": 1.1872816660315029e-06,
"loss": 0.5564,
"step": 500
},
{
"epoch": 0.7785547785547785,
"grad_norm": 1.13872043452736,
"learning_rate": 1.171523101468693e-06,
"loss": 0.5101,
"step": 501
},
{
"epoch": 0.7801087801087802,
"grad_norm": 0.9435747581759769,
"learning_rate": 1.1558559278084647e-06,
"loss": 0.6206,
"step": 502
},
{
"epoch": 0.7816627816627817,
"grad_norm": 0.8633704185786354,
"learning_rate": 1.1402805190473649e-06,
"loss": 0.5582,
"step": 503
},
{
"epoch": 0.7832167832167832,
"grad_norm": 0.7417786554815952,
"learning_rate": 1.124797246991387e-06,
"loss": 0.4371,
"step": 504
},
{
"epoch": 0.7847707847707848,
"grad_norm": 1.0878350516065922,
"learning_rate": 1.1094064812471028e-06,
"loss": 0.495,
"step": 505
},
{
"epoch": 0.7863247863247863,
"grad_norm": 0.9285222085925927,
"learning_rate": 1.0941085892128272e-06,
"loss": 0.5541,
"step": 506
},
{
"epoch": 0.7878787878787878,
"grad_norm": 0.8733376888573161,
"learning_rate": 1.07890393606986e-06,
"loss": 0.5254,
"step": 507
},
{
"epoch": 0.7894327894327894,
"grad_norm": 0.9249119247943376,
"learning_rate": 1.0637928847737594e-06,
"loss": 0.6275,
"step": 508
},
{
"epoch": 0.790986790986791,
"grad_norm": 0.8091969696708994,
"learning_rate": 1.0487757960456812e-06,
"loss": 0.5614,
"step": 509
},
{
"epoch": 0.7925407925407926,
"grad_norm": 0.9574853811974608,
"learning_rate": 1.0338530283637704e-06,
"loss": 0.4403,
"step": 510
},
{
"epoch": 0.7940947940947941,
"grad_norm": 1.2352084520669158,
"learning_rate": 1.0190249379546024e-06,
"loss": 0.6218,
"step": 511
},
{
"epoch": 0.7956487956487956,
"grad_norm": 0.898957730122624,
"learning_rate": 1.0042918787846757e-06,
"loss": 0.5151,
"step": 512
},
{
"epoch": 0.7972027972027972,
"grad_norm": 0.9789627682865999,
"learning_rate": 9.896542025519645e-07,
"loss": 0.5126,
"step": 513
},
{
"epoch": 0.7987567987567987,
"grad_norm": 0.8612205498222681,
"learning_rate": 9.751122586775253e-07,
"loss": 0.5889,
"step": 514
},
{
"epoch": 0.8003108003108003,
"grad_norm": 1.1219328556676227,
"learning_rate": 9.606663942971568e-07,
"loss": 0.5793,
"step": 515
},
{
"epoch": 0.8018648018648019,
"grad_norm": 1.0414597225046132,
"learning_rate": 9.463169542531059e-07,
"loss": 0.5137,
"step": 516
},
{
"epoch": 0.8034188034188035,
"grad_norm": 0.8844131961044368,
"learning_rate": 9.320642810858421e-07,
"loss": 0.5543,
"step": 517
},
{
"epoch": 0.804972804972805,
"grad_norm": 0.8508175460334907,
"learning_rate": 9.179087150258814e-07,
"loss": 0.5992,
"step": 518
},
{
"epoch": 0.8065268065268065,
"grad_norm": 0.9527620756716164,
"learning_rate": 9.038505939856612e-07,
"loss": 0.4986,
"step": 519
},
{
"epoch": 0.8080808080808081,
"grad_norm": 0.8675412414033893,
"learning_rate": 8.898902535514747e-07,
"loss": 0.5465,
"step": 520
},
{
"epoch": 0.8096348096348096,
"grad_norm": 0.9106070035517946,
"learning_rate": 8.76028026975459e-07,
"loss": 0.6284,
"step": 521
},
{
"epoch": 0.8111888111888111,
"grad_norm": 0.9448258212071192,
"learning_rate": 8.62264245167641e-07,
"loss": 0.6337,
"step": 522
},
{
"epoch": 0.8127428127428128,
"grad_norm": 0.8167720543226298,
"learning_rate": 8.485992366880419e-07,
"loss": 0.4287,
"step": 523
},
{
"epoch": 0.8142968142968143,
"grad_norm": 0.8126136928569179,
"learning_rate": 8.35033327738829e-07,
"loss": 0.4002,
"step": 524
},
{
"epoch": 0.8158508158508159,
"grad_norm": 0.9251994602254382,
"learning_rate": 8.215668421565276e-07,
"loss": 0.624,
"step": 525
},
{
"epoch": 0.8174048174048174,
"grad_norm": 1.043989325595444,
"learning_rate": 8.082001014042945e-07,
"loss": 0.5308,
"step": 526
},
{
"epoch": 0.818958818958819,
"grad_norm": 0.8603513049330545,
"learning_rate": 7.949334245642459e-07,
"loss": 0.418,
"step": 527
},
{
"epoch": 0.8205128205128205,
"grad_norm": 0.9200257922531035,
"learning_rate": 7.817671283298345e-07,
"loss": 0.5815,
"step": 528
},
{
"epoch": 0.822066822066822,
"grad_norm": 0.8267010262009848,
"learning_rate": 7.687015269982917e-07,
"loss": 0.5761,
"step": 529
},
{
"epoch": 0.8236208236208237,
"grad_norm": 1.0208487089497174,
"learning_rate": 7.557369324631303e-07,
"loss": 0.5874,
"step": 530
},
{
"epoch": 0.8251748251748252,
"grad_norm": 0.9096017765334373,
"learning_rate": 7.4287365420669e-07,
"loss": 0.7353,
"step": 531
},
{
"epoch": 0.8267288267288267,
"grad_norm": 0.8240298271847039,
"learning_rate": 7.301119992927585e-07,
"loss": 0.4626,
"step": 532
},
{
"epoch": 0.8282828282828283,
"grad_norm": 0.9551447108213159,
"learning_rate": 7.174522723592342e-07,
"loss": 0.5483,
"step": 533
},
{
"epoch": 0.8298368298368298,
"grad_norm": 0.8119880605731007,
"learning_rate": 7.048947756108576e-07,
"loss": 0.538,
"step": 534
},
{
"epoch": 0.8313908313908314,
"grad_norm": 0.696193829551657,
"learning_rate": 6.924398088119988e-07,
"loss": 0.4711,
"step": 535
},
{
"epoch": 0.8329448329448329,
"grad_norm": 0.8690121207713474,
"learning_rate": 6.800876692794994e-07,
"loss": 0.545,
"step": 536
},
{
"epoch": 0.8344988344988346,
"grad_norm": 0.9178082058159108,
"learning_rate": 6.678386518755747e-07,
"loss": 0.6015,
"step": 537
},
{
"epoch": 0.8360528360528361,
"grad_norm": 0.9846180274571776,
"learning_rate": 6.556930490007762e-07,
"loss": 0.5199,
"step": 538
},
{
"epoch": 0.8376068376068376,
"grad_norm": 0.902090010388194,
"learning_rate": 6.436511505870091e-07,
"loss": 0.4475,
"step": 539
},
{
"epoch": 0.8391608391608392,
"grad_norm": 0.9794639484139962,
"learning_rate": 6.317132440906188e-07,
"loss": 0.6774,
"step": 540
},
{
"epoch": 0.8407148407148407,
"grad_norm": 0.959849989606041,
"learning_rate": 6.198796144855168e-07,
"loss": 0.5281,
"step": 541
},
{
"epoch": 0.8422688422688422,
"grad_norm": 0.839879332088564,
"learning_rate": 6.081505442563912e-07,
"loss": 0.4718,
"step": 542
},
{
"epoch": 0.8438228438228438,
"grad_norm": 0.8946560239256621,
"learning_rate": 5.965263133919508e-07,
"loss": 0.5591,
"step": 543
},
{
"epoch": 0.8453768453768454,
"grad_norm": 0.7803446123118281,
"learning_rate": 5.850071993782525e-07,
"loss": 0.487,
"step": 544
},
{
"epoch": 0.846930846930847,
"grad_norm": 0.8557306905816813,
"learning_rate": 5.735934771920704e-07,
"loss": 0.3648,
"step": 545
},
{
"epoch": 0.8484848484848485,
"grad_norm": 1.1143798478654294,
"learning_rate": 5.622854192943317e-07,
"loss": 0.617,
"step": 546
},
{
"epoch": 0.85003885003885,
"grad_norm": 0.8289932147243455,
"learning_rate": 5.510832956236173e-07,
"loss": 0.4645,
"step": 547
},
{
"epoch": 0.8515928515928516,
"grad_norm": 1.1875959846429844,
"learning_rate": 5.399873735897137e-07,
"loss": 0.574,
"step": 548
},
{
"epoch": 0.8531468531468531,
"grad_norm": 1.0110337398247589,
"learning_rate": 5.289979180672344e-07,
"loss": 0.5043,
"step": 549
},
{
"epoch": 0.8547008547008547,
"grad_norm": 1.030393329446215,
"learning_rate": 5.181151913892896e-07,
"loss": 0.6686,
"step": 550
},
{
"epoch": 0.8562548562548562,
"grad_norm": 0.7985911619479463,
"learning_rate": 5.073394533412296e-07,
"loss": 0.4997,
"step": 551
},
{
"epoch": 0.8578088578088578,
"grad_norm": 0.7409828694431111,
"learning_rate": 4.966709611544435e-07,
"loss": 0.4658,
"step": 552
},
{
"epoch": 0.8593628593628594,
"grad_norm": 0.9377326798150777,
"learning_rate": 4.861099695002158e-07,
"loss": 0.5396,
"step": 553
},
{
"epoch": 0.8609168609168609,
"grad_norm": 0.9284319606041459,
"learning_rate": 4.7565673048364735e-07,
"loss": 0.5622,
"step": 554
},
{
"epoch": 0.8624708624708625,
"grad_norm": 1.0825280137566733,
"learning_rate": 4.6531149363764126e-07,
"loss": 0.4918,
"step": 555
},
{
"epoch": 0.864024864024864,
"grad_norm": 0.9050580420904966,
"learning_rate": 4.550745059169398e-07,
"loss": 0.561,
"step": 556
},
{
"epoch": 0.8655788655788655,
"grad_norm": 0.8106816259375483,
"learning_rate": 4.4494601169223715e-07,
"loss": 0.5667,
"step": 557
},
{
"epoch": 0.8671328671328671,
"grad_norm": 0.7324122672367018,
"learning_rate": 4.349262527443371e-07,
"loss": 0.4168,
"step": 558
},
{
"epoch": 0.8686868686868687,
"grad_norm": 0.8563060227977591,
"learning_rate": 4.2501546825838735e-07,
"loss": 0.5519,
"step": 559
},
{
"epoch": 0.8702408702408703,
"grad_norm": 1.0242601352816976,
"learning_rate": 4.152138948181689e-07,
"loss": 0.6415,
"step": 560
},
{
"epoch": 0.8717948717948718,
"grad_norm": 0.9408471315505526,
"learning_rate": 4.0552176640045017e-07,
"loss": 0.6281,
"step": 561
},
{
"epoch": 0.8733488733488733,
"grad_norm": 0.9252425519783338,
"learning_rate": 3.959393143693946e-07,
"loss": 0.5774,
"step": 562
},
{
"epoch": 0.8749028749028749,
"grad_norm": 0.7918385206518544,
"learning_rate": 3.864667674710454e-07,
"loss": 0.6713,
"step": 563
},
{
"epoch": 0.8764568764568764,
"grad_norm": 0.7826963848048722,
"learning_rate": 3.7710435182786053e-07,
"loss": 0.4774,
"step": 564
},
{
"epoch": 0.878010878010878,
"grad_norm": 0.7365534429069017,
"learning_rate": 3.6785229093331987e-07,
"loss": 0.3846,
"step": 565
},
{
"epoch": 0.8795648795648796,
"grad_norm": 0.7733142200970872,
"learning_rate": 3.587108056465827e-07,
"loss": 0.403,
"step": 566
},
{
"epoch": 0.8811188811188811,
"grad_norm": 1.022816223688587,
"learning_rate": 3.496801141872225e-07,
"loss": 0.5213,
"step": 567
},
{
"epoch": 0.8826728826728827,
"grad_norm": 0.9427543161563908,
"learning_rate": 3.407604321300123e-07,
"loss": 0.5404,
"step": 568
},
{
"epoch": 0.8842268842268842,
"grad_norm": 1.02774835732094,
"learning_rate": 3.3195197239978384e-07,
"loss": 0.5665,
"step": 569
},
{
"epoch": 0.8857808857808858,
"grad_norm": 1.064922469524254,
"learning_rate": 3.232549452663403e-07,
"loss": 0.5141,
"step": 570
},
{
"epoch": 0.8873348873348873,
"grad_norm": 0.7085691145526173,
"learning_rate": 3.146695583394377e-07,
"loss": 0.54,
"step": 571
},
{
"epoch": 0.8888888888888888,
"grad_norm": 0.7520808912542442,
"learning_rate": 3.061960165638317e-07,
"loss": 0.4426,
"step": 572
},
{
"epoch": 0.8904428904428905,
"grad_norm": 0.8842159851635149,
"learning_rate": 2.9783452221438304e-07,
"loss": 0.5344,
"step": 573
},
{
"epoch": 0.891996891996892,
"grad_norm": 1.2686448157314256,
"learning_rate": 2.895852748912298e-07,
"loss": 0.512,
"step": 574
},
{
"epoch": 0.8935508935508936,
"grad_norm": 1.054717271270882,
"learning_rate": 2.814484715150212e-07,
"loss": 0.6288,
"step": 575
},
{
"epoch": 0.8951048951048951,
"grad_norm": 0.9040626236395383,
"learning_rate": 2.734243063222181e-07,
"loss": 0.4089,
"step": 576
},
{
"epoch": 0.8966588966588966,
"grad_norm": 1.0743427211769383,
"learning_rate": 2.655129708604576e-07,
"loss": 0.6188,
"step": 577
},
{
"epoch": 0.8982128982128982,
"grad_norm": 0.8974286721485075,
"learning_rate": 2.5771465398397757e-07,
"loss": 0.5755,
"step": 578
},
{
"epoch": 0.8997668997668997,
"grad_norm": 1.030531093175199,
"learning_rate": 2.5002954184910887e-07,
"loss": 0.4618,
"step": 579
},
{
"epoch": 0.9013209013209014,
"grad_norm": 1.0117403154653173,
"learning_rate": 2.424578179098358e-07,
"loss": 0.5362,
"step": 580
},
{
"epoch": 0.9028749028749029,
"grad_norm": 0.9476481033300673,
"learning_rate": 2.3499966291341213e-07,
"loss": 0.5312,
"step": 581
},
{
"epoch": 0.9044289044289044,
"grad_norm": 1.0383350051929625,
"learning_rate": 2.2765525489604702e-07,
"loss": 0.5124,
"step": 582
},
{
"epoch": 0.905982905982906,
"grad_norm": 1.202514418358822,
"learning_rate": 2.2042476917865706e-07,
"loss": 0.592,
"step": 583
},
{
"epoch": 0.9075369075369075,
"grad_norm": 0.9288774041490958,
"learning_rate": 2.1330837836267882e-07,
"loss": 0.6253,
"step": 584
},
{
"epoch": 0.9090909090909091,
"grad_norm": 1.0753907470152138,
"learning_rate": 2.0630625232595126e-07,
"loss": 0.5975,
"step": 585
},
{
"epoch": 0.9106449106449106,
"grad_norm": 0.7903008621784656,
"learning_rate": 1.9941855821865918e-07,
"loss": 0.5071,
"step": 586
},
{
"epoch": 0.9121989121989122,
"grad_norm": 0.9003087903430828,
"learning_rate": 1.9264546045934196e-07,
"loss": 0.5412,
"step": 587
},
{
"epoch": 0.9137529137529138,
"grad_norm": 0.855353129364411,
"learning_rate": 1.859871207309688e-07,
"loss": 0.715,
"step": 588
},
{
"epoch": 0.9153069153069153,
"grad_norm": 0.8096629835043319,
"learning_rate": 1.7944369797708362e-07,
"loss": 0.6314,
"step": 589
},
{
"epoch": 0.9168609168609169,
"grad_norm": 0.8646658337676514,
"learning_rate": 1.7301534839800348e-07,
"loss": 0.7883,
"step": 590
},
{
"epoch": 0.9184149184149184,
"grad_norm": 0.8516940037455805,
"learning_rate": 1.6670222544709515e-07,
"loss": 0.4638,
"step": 591
},
{
"epoch": 0.9199689199689199,
"grad_norm": 1.229821089655676,
"learning_rate": 1.6050447982711214e-07,
"loss": 0.5523,
"step": 592
},
{
"epoch": 0.9215229215229215,
"grad_norm": 0.823339154576201,
"learning_rate": 1.5442225948659183e-07,
"loss": 0.4838,
"step": 593
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.9100465786354403,
"learning_rate": 1.4845570961633192e-07,
"loss": 0.4591,
"step": 594
},
{
"epoch": 0.9246309246309247,
"grad_norm": 0.8249462059650916,
"learning_rate": 1.426049726459172e-07,
"loss": 0.6088,
"step": 595
},
{
"epoch": 0.9261849261849262,
"grad_norm": 0.8140830885015211,
"learning_rate": 1.368701882403234e-07,
"loss": 0.5409,
"step": 596
},
{
"epoch": 0.9277389277389277,
"grad_norm": 0.7740424422028419,
"learning_rate": 1.3125149329658083e-07,
"loss": 0.5304,
"step": 597
},
{
"epoch": 0.9292929292929293,
"grad_norm": 0.7720807782608536,
"learning_rate": 1.2574902194050996e-07,
"loss": 0.5871,
"step": 598
},
{
"epoch": 0.9308469308469308,
"grad_norm": 0.8881572533340473,
"learning_rate": 1.2036290552351838e-07,
"loss": 0.5654,
"step": 599
},
{
"epoch": 0.9324009324009324,
"grad_norm": 0.7381756170416801,
"learning_rate": 1.150932726194609e-07,
"loss": 0.5499,
"step": 600
},
{
"epoch": 0.933954933954934,
"grad_norm": 0.7736924336820196,
"learning_rate": 1.0994024902157674e-07,
"loss": 0.5977,
"step": 601
},
{
"epoch": 0.9355089355089355,
"grad_norm": 0.9074736160128457,
"learning_rate": 1.0490395773948336e-07,
"loss": 0.4871,
"step": 602
},
{
"epoch": 0.9370629370629371,
"grad_norm": 0.7621234368730365,
"learning_rate": 9.998451899624007e-08,
"loss": 0.5144,
"step": 603
},
{
"epoch": 0.9386169386169386,
"grad_norm": 1.014770955103902,
"learning_rate": 9.5182050225478e-08,
"loss": 0.6022,
"step": 604
},
{
"epoch": 0.9401709401709402,
"grad_norm": 0.9183809943783089,
"learning_rate": 9.049666606859852e-08,
"loss": 0.5193,
"step": 605
},
{
"epoch": 0.9417249417249417,
"grad_norm": 0.8380871828967922,
"learning_rate": 8.592847837203655e-08,
"loss": 0.5955,
"step": 606
},
{
"epoch": 0.9432789432789432,
"grad_norm": 0.8781219603592406,
"learning_rate": 8.147759618458706e-08,
"loss": 0.6726,
"step": 607
},
{
"epoch": 0.9448329448329449,
"grad_norm": 0.9897864403463899,
"learning_rate": 7.714412575480556e-08,
"loss": 0.641,
"step": 608
},
{
"epoch": 0.9463869463869464,
"grad_norm": 0.9693302536123463,
"learning_rate": 7.292817052847068e-08,
"loss": 0.5614,
"step": 609
},
{
"epoch": 0.947940947940948,
"grad_norm": 0.8842272966323539,
"learning_rate": 6.882983114611497e-08,
"loss": 0.3866,
"step": 610
},
{
"epoch": 0.9494949494949495,
"grad_norm": 0.8928788478713224,
"learning_rate": 6.484920544062245e-08,
"loss": 0.7111,
"step": 611
},
{
"epoch": 0.951048951048951,
"grad_norm": 0.7471058967315293,
"learning_rate": 6.098638843489213e-08,
"loss": 0.4559,
"step": 612
},
{
"epoch": 0.9526029526029526,
"grad_norm": 1.1989009033463545,
"learning_rate": 5.7241472339572e-08,
"loss": 0.6242,
"step": 613
},
{
"epoch": 0.9541569541569541,
"grad_norm": 0.825791557449914,
"learning_rate": 5.361454655085529e-08,
"loss": 0.5102,
"step": 614
},
{
"epoch": 0.9557109557109557,
"grad_norm": 0.9246788337692015,
"learning_rate": 5.0105697648347716e-08,
"loss": 0.5252,
"step": 615
},
{
"epoch": 0.9572649572649573,
"grad_norm": 0.9616840921861742,
"learning_rate": 4.671500939300133e-08,
"loss": 0.5559,
"step": 616
},
{
"epoch": 0.9588189588189588,
"grad_norm": 0.9548923348306078,
"learning_rate": 4.344256272511338e-08,
"loss": 0.7163,
"step": 617
},
{
"epoch": 0.9603729603729604,
"grad_norm": 1.0822537483462542,
"learning_rate": 4.0288435762396164e-08,
"loss": 0.5881,
"step": 618
},
{
"epoch": 0.9619269619269619,
"grad_norm": 0.7725876652440066,
"learning_rate": 3.725270379811019e-08,
"loss": 0.5614,
"step": 619
},
{
"epoch": 0.9634809634809635,
"grad_norm": 0.8565244919322886,
"learning_rate": 3.4335439299268414e-08,
"loss": 0.4397,
"step": 620
},
{
"epoch": 0.965034965034965,
"grad_norm": 0.9606122692265989,
"learning_rate": 3.1536711904904816e-08,
"loss": 0.5277,
"step": 621
},
{
"epoch": 0.9665889665889665,
"grad_norm": 1.138988513753176,
"learning_rate": 2.8856588424414632e-08,
"loss": 0.6066,
"step": 622
},
{
"epoch": 0.9681429681429682,
"grad_norm": 0.7295542248220023,
"learning_rate": 2.6295132835956748e-08,
"loss": 0.5854,
"step": 623
},
{
"epoch": 0.9696969696969697,
"grad_norm": 0.9657806038448509,
"learning_rate": 2.3852406284927687e-08,
"loss": 0.5492,
"step": 624
},
{
"epoch": 0.9712509712509713,
"grad_norm": 0.9874016708725635,
"learning_rate": 2.152846708250167e-08,
"loss": 0.5257,
"step": 625
},
{
"epoch": 0.9728049728049728,
"grad_norm": 0.9153930251144268,
"learning_rate": 1.93233707042384e-08,
"loss": 0.5436,
"step": 626
},
{
"epoch": 0.9743589743589743,
"grad_norm": 0.8803440571286754,
"learning_rate": 1.723716978876133e-08,
"loss": 0.5056,
"step": 627
},
{
"epoch": 0.9759129759129759,
"grad_norm": 0.8302956653359888,
"learning_rate": 1.5269914136497033e-08,
"loss": 0.5629,
"step": 628
},
{
"epoch": 0.9774669774669774,
"grad_norm": 1.1113800628583712,
"learning_rate": 1.3421650708487777e-08,
"loss": 0.6096,
"step": 629
},
{
"epoch": 0.9790209790209791,
"grad_norm": 0.893165695075744,
"learning_rate": 1.1692423625273563e-08,
"loss": 0.4906,
"step": 630
},
{
"epoch": 0.9805749805749806,
"grad_norm": 0.8508627162029742,
"learning_rate": 1.0082274165834627e-08,
"loss": 0.5798,
"step": 631
},
{
"epoch": 0.9821289821289821,
"grad_norm": 0.9254388849186278,
"learning_rate": 8.591240766607779e-09,
"loss": 0.5168,
"step": 632
},
{
"epoch": 0.9836829836829837,
"grad_norm": 0.8978461708685718,
"learning_rate": 7.219359020570471e-09,
"loss": 0.5819,
"step": 633
},
{
"epoch": 0.9852369852369852,
"grad_norm": 0.8755369497191023,
"learning_rate": 5.966661676388152e-09,
"loss": 0.5777,
"step": 634
},
{
"epoch": 0.9867909867909868,
"grad_norm": 0.9185099392157843,
"learning_rate": 4.833178637633773e-09,
"loss": 0.6064,
"step": 635
},
{
"epoch": 0.9883449883449883,
"grad_norm": 0.7565206452951102,
"learning_rate": 3.818936962076136e-09,
"loss": 0.481,
"step": 636
},
{
"epoch": 0.98989898989899,
"grad_norm": 0.6749881277152423,
"learning_rate": 2.9239608610298618e-09,
"loss": 0.5021,
"step": 637
},
{
"epoch": 0.9914529914529915,
"grad_norm": 1.0427691738919331,
"learning_rate": 2.148271698781401e-09,
"loss": 0.5495,
"step": 638
},
{
"epoch": 0.993006993006993,
"grad_norm": 0.6650883713661591,
"learning_rate": 1.4918879920750029e-09,
"loss": 0.4986,
"step": 639
},
{
"epoch": 0.9945609945609946,
"grad_norm": 0.9035648077675859,
"learning_rate": 9.548254096752862e-10,
"loss": 0.6534,
"step": 640
},
{
"epoch": 0.9961149961149961,
"grad_norm": 0.8779326174407937,
"learning_rate": 5.370967719897646e-10,
"loss": 0.5301,
"step": 641
},
{
"epoch": 0.9976689976689976,
"grad_norm": 0.9522652817049506,
"learning_rate": 2.387120507629792e-10,
"loss": 0.516,
"step": 642
},
{
"epoch": 0.9992229992229992,
"grad_norm": 0.8372559748420602,
"learning_rate": 5.967836884168687e-11,
"loss": 0.4753,
"step": 643
}
],
"logging_steps": 1,
"max_steps": 643,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"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": 27996768092160.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}