min_subtitle04_abl / trainer_state.json
Minchael's picture
Upload folder using huggingface_hub
40a473f verified
Raw
History Blame Contribute Delete
30.7 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.997134670487106,
"eval_steps": 500,
"global_step": 174,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0057306590257879654,
"grad_norm": 38.03752121795943,
"learning_rate": 1.6666666666666667e-06,
"loss": 1.2484,
"step": 1
},
{
"epoch": 0.011461318051575931,
"grad_norm": 43.55384744422783,
"learning_rate": 3.3333333333333333e-06,
"loss": 1.1422,
"step": 2
},
{
"epoch": 0.017191977077363897,
"grad_norm": 29.24111522147067,
"learning_rate": 5e-06,
"loss": 1.3608,
"step": 3
},
{
"epoch": 0.022922636103151862,
"grad_norm": 35.99663748603026,
"learning_rate": 6.666666666666667e-06,
"loss": 1.1796,
"step": 4
},
{
"epoch": 0.02865329512893983,
"grad_norm": 9.33116465380011,
"learning_rate": 8.333333333333334e-06,
"loss": 1.0545,
"step": 5
},
{
"epoch": 0.034383954154727794,
"grad_norm": 3.118664242502025,
"learning_rate": 1e-05,
"loss": 0.9573,
"step": 6
},
{
"epoch": 0.04011461318051576,
"grad_norm": 2.079257864867788,
"learning_rate": 9.999125804563732e-06,
"loss": 0.7901,
"step": 7
},
{
"epoch": 0.045845272206303724,
"grad_norm": 1.9082425579894249,
"learning_rate": 9.996503523941994e-06,
"loss": 0.9245,
"step": 8
},
{
"epoch": 0.05157593123209169,
"grad_norm": 1.461948067755706,
"learning_rate": 9.992134075089085e-06,
"loss": 0.6894,
"step": 9
},
{
"epoch": 0.05730659025787966,
"grad_norm": 1.6313747953040199,
"learning_rate": 9.986018985905901e-06,
"loss": 0.9604,
"step": 10
},
{
"epoch": 0.06303724928366762,
"grad_norm": 1.5471034718114944,
"learning_rate": 9.978160394705669e-06,
"loss": 0.9204,
"step": 11
},
{
"epoch": 0.06876790830945559,
"grad_norm": 1.8154906830085713,
"learning_rate": 9.968561049466214e-06,
"loss": 0.8748,
"step": 12
},
{
"epoch": 0.07449856733524356,
"grad_norm": 2.2845328410002694,
"learning_rate": 9.957224306869053e-06,
"loss": 0.7594,
"step": 13
},
{
"epoch": 0.08022922636103152,
"grad_norm": 1.9407783525220312,
"learning_rate": 9.944154131125643e-06,
"loss": 0.7763,
"step": 14
},
{
"epoch": 0.08595988538681948,
"grad_norm": 1.9081880770841557,
"learning_rate": 9.92935509259118e-06,
"loss": 0.734,
"step": 15
},
{
"epoch": 0.09169054441260745,
"grad_norm": 1.7583121217226352,
"learning_rate": 9.912832366166443e-06,
"loss": 0.9036,
"step": 16
},
{
"epoch": 0.09742120343839542,
"grad_norm": 1.4914308782737136,
"learning_rate": 9.894591729488243e-06,
"loss": 0.6995,
"step": 17
},
{
"epoch": 0.10315186246418338,
"grad_norm": 1.2991959516663745,
"learning_rate": 9.874639560909118e-06,
"loss": 0.6394,
"step": 18
},
{
"epoch": 0.10888252148997135,
"grad_norm": 1.5862185105332742,
"learning_rate": 9.852982837266955e-06,
"loss": 1.3399,
"step": 19
},
{
"epoch": 0.11461318051575932,
"grad_norm": 2.084964490526454,
"learning_rate": 9.829629131445342e-06,
"loss": 1.0038,
"step": 20
},
{
"epoch": 0.12034383954154727,
"grad_norm": 1.5294224427764915,
"learning_rate": 9.804586609725499e-06,
"loss": 0.8236,
"step": 21
},
{
"epoch": 0.12607449856733524,
"grad_norm": 1.7784781345976681,
"learning_rate": 9.777864028930705e-06,
"loss": 0.8731,
"step": 22
},
{
"epoch": 0.1318051575931232,
"grad_norm": 1.4236810234014226,
"learning_rate": 9.749470733364231e-06,
"loss": 0.6408,
"step": 23
},
{
"epoch": 0.13753581661891118,
"grad_norm": 1.5852008363923855,
"learning_rate": 9.719416651541839e-06,
"loss": 0.8506,
"step": 24
},
{
"epoch": 0.14326647564469913,
"grad_norm": 2.120960168553231,
"learning_rate": 9.687712292719997e-06,
"loss": 0.9738,
"step": 25
},
{
"epoch": 0.1489971346704871,
"grad_norm": 1.7958951444738618,
"learning_rate": 9.654368743221022e-06,
"loss": 0.7693,
"step": 26
},
{
"epoch": 0.15472779369627507,
"grad_norm": 1.9090771777185072,
"learning_rate": 9.619397662556434e-06,
"loss": 1.0445,
"step": 27
},
{
"epoch": 0.16045845272206305,
"grad_norm": 1.3292868621801857,
"learning_rate": 9.582811279349881e-06,
"loss": 0.7637,
"step": 28
},
{
"epoch": 0.166189111747851,
"grad_norm": 1.9325339894925524,
"learning_rate": 9.544622387061055e-06,
"loss": 0.7206,
"step": 29
},
{
"epoch": 0.17191977077363896,
"grad_norm": 2.662679695094159,
"learning_rate": 9.504844339512096e-06,
"loss": 0.6481,
"step": 30
},
{
"epoch": 0.17765042979942694,
"grad_norm": 1.6364790170233385,
"learning_rate": 9.463491046218058e-06,
"loss": 0.786,
"step": 31
},
{
"epoch": 0.1833810888252149,
"grad_norm": 1.6139862561363207,
"learning_rate": 9.420576967523049e-06,
"loss": 0.7789,
"step": 32
},
{
"epoch": 0.18911174785100288,
"grad_norm": 2.0390396224600402,
"learning_rate": 9.376117109543769e-06,
"loss": 1.0139,
"step": 33
},
{
"epoch": 0.19484240687679083,
"grad_norm": 1.6131277961753778,
"learning_rate": 9.330127018922195e-06,
"loss": 0.8731,
"step": 34
},
{
"epoch": 0.20057306590257878,
"grad_norm": 1.7267415549062777,
"learning_rate": 9.282622777389258e-06,
"loss": 0.803,
"step": 35
},
{
"epoch": 0.20630372492836677,
"grad_norm": 2.167605301555236,
"learning_rate": 9.233620996141421e-06,
"loss": 1.0428,
"step": 36
},
{
"epoch": 0.21203438395415472,
"grad_norm": 1.4740350318953115,
"learning_rate": 9.1831388100321e-06,
"loss": 0.8932,
"step": 37
},
{
"epoch": 0.2177650429799427,
"grad_norm": 1.5921364354065235,
"learning_rate": 9.131193871579975e-06,
"loss": 0.7661,
"step": 38
},
{
"epoch": 0.22349570200573066,
"grad_norm": 1.8455623997880948,
"learning_rate": 9.077804344796302e-06,
"loss": 0.5623,
"step": 39
},
{
"epoch": 0.22922636103151864,
"grad_norm": 1.9418952446519602,
"learning_rate": 9.022988898833342e-06,
"loss": 0.7408,
"step": 40
},
{
"epoch": 0.2349570200573066,
"grad_norm": 1.8091822954567782,
"learning_rate": 8.966766701456177e-06,
"loss": 1.0054,
"step": 41
},
{
"epoch": 0.24068767908309455,
"grad_norm": 1.7853338969203865,
"learning_rate": 8.90915741234015e-06,
"loss": 0.8288,
"step": 42
},
{
"epoch": 0.24641833810888253,
"grad_norm": 1.7444662612755464,
"learning_rate": 8.850181176196316e-06,
"loss": 0.8842,
"step": 43
},
{
"epoch": 0.2521489971346705,
"grad_norm": 2.3276243649871766,
"learning_rate": 8.789858615727266e-06,
"loss": 0.9948,
"step": 44
},
{
"epoch": 0.25787965616045844,
"grad_norm": 1.7683268045352898,
"learning_rate": 8.728210824415829e-06,
"loss": 0.9773,
"step": 45
},
{
"epoch": 0.2636103151862464,
"grad_norm": 1.7417916601245387,
"learning_rate": 8.665259359149132e-06,
"loss": 0.9921,
"step": 46
},
{
"epoch": 0.2693409742120344,
"grad_norm": 2.1041118445695695,
"learning_rate": 8.601026232680634e-06,
"loss": 0.9106,
"step": 47
},
{
"epoch": 0.27507163323782235,
"grad_norm": 1.4986328536717937,
"learning_rate": 8.535533905932739e-06,
"loss": 0.7517,
"step": 48
},
{
"epoch": 0.2808022922636103,
"grad_norm": 1.695401904621283,
"learning_rate": 8.46880528014271e-06,
"loss": 0.9634,
"step": 49
},
{
"epoch": 0.28653295128939826,
"grad_norm": 1.4981900842940585,
"learning_rate": 8.400863688854598e-06,
"loss": 0.7284,
"step": 50
},
{
"epoch": 0.2922636103151863,
"grad_norm": 1.941412669111069,
"learning_rate": 8.331732889760021e-06,
"loss": 0.9765,
"step": 51
},
{
"epoch": 0.2979942693409742,
"grad_norm": 2.1167996720159787,
"learning_rate": 8.261437056390607e-06,
"loss": 1.1364,
"step": 52
},
{
"epoch": 0.3037249283667622,
"grad_norm": 1.4795496943145472,
"learning_rate": 8.190000769665044e-06,
"loss": 0.7857,
"step": 53
},
{
"epoch": 0.30945558739255014,
"grad_norm": 1.4415096002348486,
"learning_rate": 8.117449009293668e-06,
"loss": 0.8041,
"step": 54
},
{
"epoch": 0.3151862464183381,
"grad_norm": 1.8065267349997336,
"learning_rate": 8.043807145043604e-06,
"loss": 0.9391,
"step": 55
},
{
"epoch": 0.3209169054441261,
"grad_norm": 1.6729056321775089,
"learning_rate": 7.969100927867508e-06,
"loss": 0.778,
"step": 56
},
{
"epoch": 0.32664756446991405,
"grad_norm": 1.6718821440398033,
"learning_rate": 7.89335648089903e-06,
"loss": 0.9816,
"step": 57
},
{
"epoch": 0.332378223495702,
"grad_norm": 1.8680743777502735,
"learning_rate": 7.81660029031811e-06,
"loss": 1.0359,
"step": 58
},
{
"epoch": 0.33810888252148996,
"grad_norm": 2.909307657272174,
"learning_rate": 7.738859196089358e-06,
"loss": 0.9696,
"step": 59
},
{
"epoch": 0.3438395415472779,
"grad_norm": 1.9661456939437394,
"learning_rate": 7.660160382576683e-06,
"loss": 0.8973,
"step": 60
},
{
"epoch": 0.3495702005730659,
"grad_norm": 1.917643955552544,
"learning_rate": 7.580531369037534e-06,
"loss": 0.6643,
"step": 61
},
{
"epoch": 0.3553008595988539,
"grad_norm": 2.1717988514180284,
"learning_rate": 7.500000000000001e-06,
"loss": 0.9754,
"step": 62
},
{
"epoch": 0.36103151862464183,
"grad_norm": 1.5568620128805357,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.7688,
"step": 63
},
{
"epoch": 0.3667621776504298,
"grad_norm": 2.0110349134389565,
"learning_rate": 7.336343141365311e-06,
"loss": 0.8847,
"step": 64
},
{
"epoch": 0.37249283667621774,
"grad_norm": 2.1139452865088897,
"learning_rate": 7.253274878999728e-06,
"loss": 0.7686,
"step": 65
},
{
"epoch": 0.37822349570200575,
"grad_norm": 2.0313551495556696,
"learning_rate": 7.169418695587791e-06,
"loss": 0.9931,
"step": 66
},
{
"epoch": 0.3839541547277937,
"grad_norm": 2.013802481518114,
"learning_rate": 7.084803913806642e-06,
"loss": 1.1539,
"step": 67
},
{
"epoch": 0.38968481375358166,
"grad_norm": 2.8357707792181333,
"learning_rate": 6.999460121598704e-06,
"loss": 1.0698,
"step": 68
},
{
"epoch": 0.3954154727793696,
"grad_norm": 2.5594895546826906,
"learning_rate": 6.913417161825449e-06,
"loss": 1.072,
"step": 69
},
{
"epoch": 0.40114613180515757,
"grad_norm": 2.1168800221761424,
"learning_rate": 6.8267051218319766e-06,
"loss": 1.0611,
"step": 70
},
{
"epoch": 0.4068767908309456,
"grad_norm": 1.818456858275615,
"learning_rate": 6.739354322926136e-06,
"loss": 0.7296,
"step": 71
},
{
"epoch": 0.41260744985673353,
"grad_norm": 1.7413527942618008,
"learning_rate": 6.651395309775837e-06,
"loss": 0.8589,
"step": 72
},
{
"epoch": 0.4183381088825215,
"grad_norm": 2.489444621758966,
"learning_rate": 6.562858839728224e-06,
"loss": 1.0043,
"step": 73
},
{
"epoch": 0.42406876790830944,
"grad_norm": 1.805567469756333,
"learning_rate": 6.473775872054522e-06,
"loss": 0.8318,
"step": 74
},
{
"epoch": 0.4297994269340974,
"grad_norm": 2.037134319241965,
"learning_rate": 6.384177557124247e-06,
"loss": 0.9895,
"step": 75
},
{
"epoch": 0.4355300859598854,
"grad_norm": 1.5387041191948552,
"learning_rate": 6.294095225512604e-06,
"loss": 0.4761,
"step": 76
},
{
"epoch": 0.44126074498567336,
"grad_norm": 2.409739385797833,
"learning_rate": 6.2035603770448664e-06,
"loss": 1.0034,
"step": 77
},
{
"epoch": 0.4469914040114613,
"grad_norm": 2.350675411482951,
"learning_rate": 6.112604669781572e-06,
"loss": 0.9924,
"step": 78
},
{
"epoch": 0.45272206303724927,
"grad_norm": 2.3453619396682095,
"learning_rate": 6.0212599089484026e-06,
"loss": 0.7836,
"step": 79
},
{
"epoch": 0.4584527220630373,
"grad_norm": 2.2813781169102856,
"learning_rate": 5.929558035814574e-06,
"loss": 0.6844,
"step": 80
},
{
"epoch": 0.46418338108882523,
"grad_norm": 1.6084425933725497,
"learning_rate": 5.837531116523683e-06,
"loss": 0.8947,
"step": 81
},
{
"epoch": 0.4699140401146132,
"grad_norm": 1.8548251851951525,
"learning_rate": 5.745211330880872e-06,
"loss": 0.7958,
"step": 82
},
{
"epoch": 0.47564469914040114,
"grad_norm": 2.188261595014796,
"learning_rate": 5.65263096110026e-06,
"loss": 0.9301,
"step": 83
},
{
"epoch": 0.4813753581661891,
"grad_norm": 2.2451224706760575,
"learning_rate": 5.559822380516539e-06,
"loss": 0.801,
"step": 84
},
{
"epoch": 0.4871060171919771,
"grad_norm": 2.7957055278500045,
"learning_rate": 5.466818042264754e-06,
"loss": 0.8237,
"step": 85
},
{
"epoch": 0.49283667621776506,
"grad_norm": 1.9897372441808463,
"learning_rate": 5.373650467932122e-06,
"loss": 0.9102,
"step": 86
},
{
"epoch": 0.498567335243553,
"grad_norm": 1.5985950087252014,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.7559,
"step": 87
},
{
"epoch": 0.504297994269341,
"grad_norm": 2.154880601955351,
"learning_rate": 5.18695597138163e-06,
"loss": 1.0098,
"step": 88
},
{
"epoch": 0.5100286532951289,
"grad_norm": 2.4244395714313036,
"learning_rate": 5.093494332154511e-06,
"loss": 0.7553,
"step": 89
},
{
"epoch": 0.5157593123209169,
"grad_norm": 1.901544938574404,
"learning_rate": 5e-06,
"loss": 0.6786,
"step": 90
},
{
"epoch": 0.5214899713467048,
"grad_norm": 2.062093571576684,
"learning_rate": 4.90650566784549e-06,
"loss": 0.96,
"step": 91
},
{
"epoch": 0.5272206303724928,
"grad_norm": 1.9839628958084239,
"learning_rate": 4.8130440286183725e-06,
"loss": 0.7209,
"step": 92
},
{
"epoch": 0.5329512893982808,
"grad_norm": 1.9337403320473419,
"learning_rate": 4.719647763814041e-06,
"loss": 0.7648,
"step": 93
},
{
"epoch": 0.5386819484240688,
"grad_norm": 1.7178327854322626,
"learning_rate": 4.626349532067879e-06,
"loss": 0.5766,
"step": 94
},
{
"epoch": 0.5444126074498568,
"grad_norm": 2.141559081402764,
"learning_rate": 4.533181957735247e-06,
"loss": 0.9044,
"step": 95
},
{
"epoch": 0.5501432664756447,
"grad_norm": 2.0742645082632682,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.9375,
"step": 96
},
{
"epoch": 0.5558739255014327,
"grad_norm": 2.1449342114938723,
"learning_rate": 4.347369038899744e-06,
"loss": 1.1029,
"step": 97
},
{
"epoch": 0.5616045845272206,
"grad_norm": 2.075710676407305,
"learning_rate": 4.254788669119127e-06,
"loss": 0.9975,
"step": 98
},
{
"epoch": 0.5673352435530086,
"grad_norm": 2.2079443398543632,
"learning_rate": 4.162468883476319e-06,
"loss": 0.7212,
"step": 99
},
{
"epoch": 0.5730659025787965,
"grad_norm": 1.7864455905947398,
"learning_rate": 4.070441964185428e-06,
"loss": 0.8496,
"step": 100
},
{
"epoch": 0.5787965616045845,
"grad_norm": 2.2762228357583387,
"learning_rate": 3.978740091051599e-06,
"loss": 0.5553,
"step": 101
},
{
"epoch": 0.5845272206303725,
"grad_norm": 2.50549167423401,
"learning_rate": 3.887395330218429e-06,
"loss": 0.6498,
"step": 102
},
{
"epoch": 0.5902578796561605,
"grad_norm": 1.871828185565768,
"learning_rate": 3.7964396229551365e-06,
"loss": 0.8946,
"step": 103
},
{
"epoch": 0.5959885386819485,
"grad_norm": 1.7434442626884263,
"learning_rate": 3.705904774487396e-06,
"loss": 0.904,
"step": 104
},
{
"epoch": 0.6017191977077364,
"grad_norm": 1.9446757874680454,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.7417,
"step": 105
},
{
"epoch": 0.6074498567335244,
"grad_norm": 1.8527724775679515,
"learning_rate": 3.526224127945479e-06,
"loss": 0.8268,
"step": 106
},
{
"epoch": 0.6131805157593123,
"grad_norm": 2.2386917522783176,
"learning_rate": 3.4371411602717785e-06,
"loss": 0.7951,
"step": 107
},
{
"epoch": 0.6189111747851003,
"grad_norm": 1.9420983215674614,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.86,
"step": 108
},
{
"epoch": 0.6246418338108882,
"grad_norm": 1.6230246232627938,
"learning_rate": 3.2606456770738636e-06,
"loss": 0.7867,
"step": 109
},
{
"epoch": 0.6303724928366762,
"grad_norm": 2.306590350544272,
"learning_rate": 3.173294878168025e-06,
"loss": 0.7447,
"step": 110
},
{
"epoch": 0.6361031518624641,
"grad_norm": 2.2209079378925516,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.8641,
"step": 111
},
{
"epoch": 0.6418338108882522,
"grad_norm": 2.232201608269791,
"learning_rate": 3.000539878401296e-06,
"loss": 0.6391,
"step": 112
},
{
"epoch": 0.6475644699140402,
"grad_norm": 1.8217544475888263,
"learning_rate": 2.9151960861933616e-06,
"loss": 0.7849,
"step": 113
},
{
"epoch": 0.6532951289398281,
"grad_norm": 2.191945560141893,
"learning_rate": 2.83058130441221e-06,
"loss": 0.595,
"step": 114
},
{
"epoch": 0.6590257879656161,
"grad_norm": 1.965774776845591,
"learning_rate": 2.746725121000273e-06,
"loss": 1.0082,
"step": 115
},
{
"epoch": 0.664756446991404,
"grad_norm": 2.031578262041227,
"learning_rate": 2.66365685863469e-06,
"loss": 1.1375,
"step": 116
},
{
"epoch": 0.670487106017192,
"grad_norm": 2.0306880825854137,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.7931,
"step": 117
},
{
"epoch": 0.6762177650429799,
"grad_norm": 2.200006443142367,
"learning_rate": 2.5000000000000015e-06,
"loss": 0.6923,
"step": 118
},
{
"epoch": 0.6819484240687679,
"grad_norm": 1.64143732674854,
"learning_rate": 2.4194686309624664e-06,
"loss": 0.835,
"step": 119
},
{
"epoch": 0.6876790830945558,
"grad_norm": 2.2062826729434906,
"learning_rate": 2.339839617423318e-06,
"loss": 0.7819,
"step": 120
},
{
"epoch": 0.6934097421203438,
"grad_norm": 1.7555732429143611,
"learning_rate": 2.261140803910644e-06,
"loss": 0.7783,
"step": 121
},
{
"epoch": 0.6991404011461319,
"grad_norm": 2.1444677292321135,
"learning_rate": 2.1833997096818897e-06,
"loss": 0.7391,
"step": 122
},
{
"epoch": 0.7048710601719198,
"grad_norm": 1.8356146316402338,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.9004,
"step": 123
},
{
"epoch": 0.7106017191977078,
"grad_norm": 1.7365309487468923,
"learning_rate": 2.030899072132493e-06,
"loss": 0.6856,
"step": 124
},
{
"epoch": 0.7163323782234957,
"grad_norm": 1.870274141128081,
"learning_rate": 1.956192854956397e-06,
"loss": 0.7922,
"step": 125
},
{
"epoch": 0.7220630372492837,
"grad_norm": 2.6649040334384906,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.9751,
"step": 126
},
{
"epoch": 0.7277936962750716,
"grad_norm": 1.6483033078124796,
"learning_rate": 1.809999230334958e-06,
"loss": 0.7539,
"step": 127
},
{
"epoch": 0.7335243553008596,
"grad_norm": 2.010163376478925,
"learning_rate": 1.7385629436093958e-06,
"loss": 0.8649,
"step": 128
},
{
"epoch": 0.7392550143266475,
"grad_norm": 2.0275309273554636,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.8878,
"step": 129
},
{
"epoch": 0.7449856733524355,
"grad_norm": 2.3490611202071103,
"learning_rate": 1.5991363111454023e-06,
"loss": 0.8193,
"step": 130
},
{
"epoch": 0.7507163323782235,
"grad_norm": 1.900621611609537,
"learning_rate": 1.5311947198572918e-06,
"loss": 0.6636,
"step": 131
},
{
"epoch": 0.7564469914040115,
"grad_norm": 2.007561743166001,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.8756,
"step": 132
},
{
"epoch": 0.7621776504297995,
"grad_norm": 1.7862412355757082,
"learning_rate": 1.3989737673193682e-06,
"loss": 0.766,
"step": 133
},
{
"epoch": 0.7679083094555874,
"grad_norm": 2.2366239499274885,
"learning_rate": 1.3347406408508695e-06,
"loss": 0.9335,
"step": 134
},
{
"epoch": 0.7736389684813754,
"grad_norm": 1.994686690899052,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.8424,
"step": 135
},
{
"epoch": 0.7793696275071633,
"grad_norm": 1.7263909625437526,
"learning_rate": 1.2101413842727345e-06,
"loss": 0.623,
"step": 136
},
{
"epoch": 0.7851002865329513,
"grad_norm": 1.7696809751764415,
"learning_rate": 1.1498188238036862e-06,
"loss": 0.7536,
"step": 137
},
{
"epoch": 0.7908309455587392,
"grad_norm": 2.3063198885718665,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.6875,
"step": 138
},
{
"epoch": 0.7965616045845272,
"grad_norm": 2.75676518789804,
"learning_rate": 1.0332332985438248e-06,
"loss": 0.7452,
"step": 139
},
{
"epoch": 0.8022922636103151,
"grad_norm": 2.720445851218192,
"learning_rate": 9.770111011666582e-07,
"loss": 0.7981,
"step": 140
},
{
"epoch": 0.8080229226361032,
"grad_norm": 2.4485977567557384,
"learning_rate": 9.221956552036992e-07,
"loss": 0.6612,
"step": 141
},
{
"epoch": 0.8137535816618912,
"grad_norm": 2.0067997786953815,
"learning_rate": 8.688061284200266e-07,
"loss": 0.7729,
"step": 142
},
{
"epoch": 0.8194842406876791,
"grad_norm": 1.848960943899034,
"learning_rate": 8.168611899679013e-07,
"loss": 0.7337,
"step": 143
},
{
"epoch": 0.8252148997134671,
"grad_norm": 2.330713084429838,
"learning_rate": 7.663790038585794e-07,
"loss": 0.7533,
"step": 144
},
{
"epoch": 0.830945558739255,
"grad_norm": 2.3887483753031025,
"learning_rate": 7.173772226107434e-07,
"loss": 0.9094,
"step": 145
},
{
"epoch": 0.836676217765043,
"grad_norm": 2.0130432674940746,
"learning_rate": 6.698729810778065e-07,
"loss": 0.9653,
"step": 146
},
{
"epoch": 0.8424068767908309,
"grad_norm": 1.6267387501364545,
"learning_rate": 6.238828904562316e-07,
"loss": 0.9619,
"step": 147
},
{
"epoch": 0.8481375358166189,
"grad_norm": 2.526626827002361,
"learning_rate": 5.794230324769518e-07,
"loss": 0.964,
"step": 148
},
{
"epoch": 0.8538681948424068,
"grad_norm": 1.9896774502272758,
"learning_rate": 5.365089537819435e-07,
"loss": 0.8691,
"step": 149
},
{
"epoch": 0.8595988538681948,
"grad_norm": 2.585629402971992,
"learning_rate": 4.951556604879049e-07,
"loss": 1.0747,
"step": 150
},
{
"epoch": 0.8653295128939829,
"grad_norm": 2.495918637440088,
"learning_rate": 4.5537761293894535e-07,
"loss": 0.8902,
"step": 151
},
{
"epoch": 0.8710601719197708,
"grad_norm": 2.1021328707420746,
"learning_rate": 4.171887206501191e-07,
"loss": 0.7818,
"step": 152
},
{
"epoch": 0.8767908309455588,
"grad_norm": 2.1846625762736616,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.8266,
"step": 153
},
{
"epoch": 0.8825214899713467,
"grad_norm": 2.2179930582893386,
"learning_rate": 3.4563125677897936e-07,
"loss": 0.8551,
"step": 154
},
{
"epoch": 0.8882521489971347,
"grad_norm": 1.7740479889250165,
"learning_rate": 3.122877072800046e-07,
"loss": 0.674,
"step": 155
},
{
"epoch": 0.8939828080229226,
"grad_norm": 1.7703382235738212,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.7312,
"step": 156
},
{
"epoch": 0.8997134670487106,
"grad_norm": 2.2063676324592865,
"learning_rate": 2.5052926663577006e-07,
"loss": 0.9626,
"step": 157
},
{
"epoch": 0.9054441260744985,
"grad_norm": 1.9634406501187076,
"learning_rate": 2.2213597106929608e-07,
"loss": 0.8026,
"step": 158
},
{
"epoch": 0.9111747851002865,
"grad_norm": 1.6585310003133953,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.7607,
"step": 159
},
{
"epoch": 0.9169054441260746,
"grad_norm": 1.5523982860573942,
"learning_rate": 1.7037086855465902e-07,
"loss": 0.6194,
"step": 160
},
{
"epoch": 0.9226361031518625,
"grad_norm": 1.804644365110078,
"learning_rate": 1.4701716273304524e-07,
"loss": 0.5978,
"step": 161
},
{
"epoch": 0.9283667621776505,
"grad_norm": 1.7934562877714892,
"learning_rate": 1.253604390908819e-07,
"loss": 0.6244,
"step": 162
},
{
"epoch": 0.9340974212034384,
"grad_norm": 1.753965314404328,
"learning_rate": 1.0540827051175817e-07,
"loss": 0.8244,
"step": 163
},
{
"epoch": 0.9398280802292264,
"grad_norm": 2.0565182221725222,
"learning_rate": 8.716763383355863e-08,
"loss": 0.8432,
"step": 164
},
{
"epoch": 0.9455587392550143,
"grad_norm": 1.7216420298436368,
"learning_rate": 7.064490740882057e-08,
"loss": 0.8291,
"step": 165
},
{
"epoch": 0.9512893982808023,
"grad_norm": 1.9827806263185936,
"learning_rate": 5.584586887435739e-08,
"loss": 0.6268,
"step": 166
},
{
"epoch": 0.9570200573065902,
"grad_norm": 1.8772494608310286,
"learning_rate": 4.2775693130948094e-08,
"loss": 0.7902,
"step": 167
},
{
"epoch": 0.9627507163323782,
"grad_norm": 1.4819292937442876,
"learning_rate": 3.143895053378698e-08,
"loss": 0.4946,
"step": 168
},
{
"epoch": 0.9684813753581661,
"grad_norm": 2.089816786264,
"learning_rate": 2.1839605294330935e-08,
"loss": 0.7076,
"step": 169
},
{
"epoch": 0.9742120343839542,
"grad_norm": 1.8363884283150356,
"learning_rate": 1.3981014094099354e-08,
"loss": 0.9666,
"step": 170
},
{
"epoch": 0.9799426934097422,
"grad_norm": 2.0579757695634817,
"learning_rate": 7.865924910916977e-09,
"loss": 0.7468,
"step": 171
},
{
"epoch": 0.9856733524355301,
"grad_norm": 2.062073532885379,
"learning_rate": 3.496476058006959e-09,
"loss": 0.8763,
"step": 172
},
{
"epoch": 0.9914040114613181,
"grad_norm": 2.12297815323157,
"learning_rate": 8.741954362678773e-10,
"loss": 0.9101,
"step": 173
},
{
"epoch": 0.997134670487106,
"grad_norm": 2.0438330265052214,
"learning_rate": 0.0,
"loss": 0.9878,
"step": 174
},
{
"epoch": 0.997134670487106,
"step": 174,
"total_flos": 32682297016320.0,
"train_loss": 0.847670377328478,
"train_runtime": 5260.9512,
"train_samples_per_second": 0.531,
"train_steps_per_second": 0.033
}
],
"logging_steps": 1.0,
"max_steps": 174,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"total_flos": 32682297016320.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}