TQ-2.5-72b-RP-Ink-ep1-adpt / trainer_state.json
Fizzarolli's picture
Upload folder using huggingface_hub
155275a verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0040160642570282,
"eval_steps": 500,
"global_step": 250,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004016064257028112,
"grad_norm": 0.7356438636779785,
"learning_rate": 2.4000000000000003e-06,
"loss": 2.1205,
"step": 1
},
{
"epoch": 0.008032128514056224,
"grad_norm": 0.6866766810417175,
"learning_rate": 4.800000000000001e-06,
"loss": 2.122,
"step": 2
},
{
"epoch": 0.012048192771084338,
"grad_norm": 0.7540773153305054,
"learning_rate": 7.2e-06,
"loss": 2.326,
"step": 3
},
{
"epoch": 0.01606425702811245,
"grad_norm": 0.8841753005981445,
"learning_rate": 9.600000000000001e-06,
"loss": 1.9947,
"step": 4
},
{
"epoch": 0.020080321285140562,
"grad_norm": 0.7485767006874084,
"learning_rate": 1.2e-05,
"loss": 1.9986,
"step": 5
},
{
"epoch": 0.024096385542168676,
"grad_norm": 0.5316860675811768,
"learning_rate": 1.44e-05,
"loss": 1.9405,
"step": 6
},
{
"epoch": 0.028112449799196786,
"grad_norm": 0.2008056938648224,
"learning_rate": 1.6800000000000002e-05,
"loss": 1.7299,
"step": 7
},
{
"epoch": 0.0321285140562249,
"grad_norm": 0.08951256424188614,
"learning_rate": 1.9200000000000003e-05,
"loss": 1.8174,
"step": 8
},
{
"epoch": 0.03614457831325301,
"grad_norm": 0.19312520325183868,
"learning_rate": 2.16e-05,
"loss": 1.8569,
"step": 9
},
{
"epoch": 0.040160642570281124,
"grad_norm": 0.2667342722415924,
"learning_rate": 2.4e-05,
"loss": 1.8051,
"step": 10
},
{
"epoch": 0.04417670682730924,
"grad_norm": 0.3407232165336609,
"learning_rate": 2.64e-05,
"loss": 1.8647,
"step": 11
},
{
"epoch": 0.04819277108433735,
"grad_norm": 0.4737996459007263,
"learning_rate": 2.88e-05,
"loss": 1.7965,
"step": 12
},
{
"epoch": 0.05220883534136546,
"grad_norm": 0.41904422640800476,
"learning_rate": 3.12e-05,
"loss": 1.8721,
"step": 13
},
{
"epoch": 0.05622489959839357,
"grad_norm": 0.23236039280891418,
"learning_rate": 3.3600000000000004e-05,
"loss": 2.0646,
"step": 14
},
{
"epoch": 0.060240963855421686,
"grad_norm": 0.26083624362945557,
"learning_rate": 3.6e-05,
"loss": 1.7836,
"step": 15
},
{
"epoch": 0.0642570281124498,
"grad_norm": 0.35279905796051025,
"learning_rate": 3.8400000000000005e-05,
"loss": 1.8611,
"step": 16
},
{
"epoch": 0.06827309236947791,
"grad_norm": 0.30636098980903625,
"learning_rate": 4.08e-05,
"loss": 1.9236,
"step": 17
},
{
"epoch": 0.07228915662650602,
"grad_norm": 0.2782573997974396,
"learning_rate": 4.32e-05,
"loss": 1.6871,
"step": 18
},
{
"epoch": 0.07630522088353414,
"grad_norm": 0.16966235637664795,
"learning_rate": 4.5600000000000004e-05,
"loss": 1.7482,
"step": 19
},
{
"epoch": 0.08032128514056225,
"grad_norm": 0.20790238678455353,
"learning_rate": 4.8e-05,
"loss": 1.7494,
"step": 20
},
{
"epoch": 0.08433734939759036,
"grad_norm": 0.1863771229982376,
"learning_rate": 5.04e-05,
"loss": 1.7022,
"step": 21
},
{
"epoch": 0.08835341365461848,
"grad_norm": 0.1544342190027237,
"learning_rate": 5.28e-05,
"loss": 1.9192,
"step": 22
},
{
"epoch": 0.09236947791164658,
"grad_norm": 0.1717195063829422,
"learning_rate": 5.520000000000001e-05,
"loss": 2.0797,
"step": 23
},
{
"epoch": 0.0963855421686747,
"grad_norm": 0.13888715207576752,
"learning_rate": 5.76e-05,
"loss": 1.8933,
"step": 24
},
{
"epoch": 0.10040160642570281,
"grad_norm": 0.12953609228134155,
"learning_rate": 6e-05,
"loss": 1.8774,
"step": 25
},
{
"epoch": 0.10441767068273092,
"grad_norm": 0.1392662227153778,
"learning_rate": 5.999933829087074e-05,
"loss": 1.9031,
"step": 26
},
{
"epoch": 0.10843373493975904,
"grad_norm": 0.17272761464118958,
"learning_rate": 5.999735319267354e-05,
"loss": 1.8345,
"step": 27
},
{
"epoch": 0.11244979919678715,
"grad_norm": 0.17943339049816132,
"learning_rate": 5.999404479297892e-05,
"loss": 1.7537,
"step": 28
},
{
"epoch": 0.11646586345381527,
"grad_norm": 0.14608968794345856,
"learning_rate": 5.998941323773343e-05,
"loss": 1.8518,
"step": 29
},
{
"epoch": 0.12048192771084337,
"grad_norm": 0.12942039966583252,
"learning_rate": 5.9983458731253224e-05,
"loss": 1.8527,
"step": 30
},
{
"epoch": 0.12449799196787148,
"grad_norm": 0.1064087525010109,
"learning_rate": 5.9976181536215066e-05,
"loss": 1.7588,
"step": 31
},
{
"epoch": 0.1285140562248996,
"grad_norm": 0.123220294713974,
"learning_rate": 5.99675819736447e-05,
"loss": 1.7637,
"step": 32
},
{
"epoch": 0.13253012048192772,
"grad_norm": 0.1729809194803238,
"learning_rate": 5.9957660422902735e-05,
"loss": 1.9683,
"step": 33
},
{
"epoch": 0.13654618473895583,
"grad_norm": 0.12830182909965515,
"learning_rate": 5.9946417321667896e-05,
"loss": 1.8534,
"step": 34
},
{
"epoch": 0.14056224899598393,
"grad_norm": 0.10919933766126633,
"learning_rate": 5.993385316591769e-05,
"loss": 1.801,
"step": 35
},
{
"epoch": 0.14457831325301204,
"grad_norm": 0.10252447426319122,
"learning_rate": 5.991996850990655e-05,
"loss": 1.7462,
"step": 36
},
{
"epoch": 0.14859437751004015,
"grad_norm": 0.10130741447210312,
"learning_rate": 5.9904763966141394e-05,
"loss": 1.7917,
"step": 37
},
{
"epoch": 0.15261044176706828,
"grad_norm": 0.14359734952449799,
"learning_rate": 5.9888240205354576e-05,
"loss": 1.5508,
"step": 38
},
{
"epoch": 0.1566265060240964,
"grad_norm": 0.13955214619636536,
"learning_rate": 5.987039795647432e-05,
"loss": 1.8419,
"step": 39
},
{
"epoch": 0.1606425702811245,
"grad_norm": 0.1250331997871399,
"learning_rate": 5.985123800659256e-05,
"loss": 1.8584,
"step": 40
},
{
"epoch": 0.1646586345381526,
"grad_norm": 0.09694834798574448,
"learning_rate": 5.983076120093022e-05,
"loss": 1.7475,
"step": 41
},
{
"epoch": 0.1686746987951807,
"grad_norm": 0.10690533369779587,
"learning_rate": 5.98089684427999e-05,
"loss": 1.881,
"step": 42
},
{
"epoch": 0.17269076305220885,
"grad_norm": 0.10619474202394485,
"learning_rate": 5.978586069356608e-05,
"loss": 1.7596,
"step": 43
},
{
"epoch": 0.17670682730923695,
"grad_norm": 0.1079355850815773,
"learning_rate": 5.9761438972602665e-05,
"loss": 1.8796,
"step": 44
},
{
"epoch": 0.18072289156626506,
"grad_norm": 0.1343262791633606,
"learning_rate": 5.973570435724803e-05,
"loss": 1.8608,
"step": 45
},
{
"epoch": 0.18473895582329317,
"grad_norm": 0.08072712272405624,
"learning_rate": 5.970865798275751e-05,
"loss": 1.8684,
"step": 46
},
{
"epoch": 0.18875502008032127,
"grad_norm": 0.08719664812088013,
"learning_rate": 5.968030104225331e-05,
"loss": 1.7377,
"step": 47
},
{
"epoch": 0.1927710843373494,
"grad_norm": 0.1065979152917862,
"learning_rate": 5.9650634786671835e-05,
"loss": 1.9423,
"step": 48
},
{
"epoch": 0.19678714859437751,
"grad_norm": 0.10489759594202042,
"learning_rate": 5.961966052470857e-05,
"loss": 1.7712,
"step": 49
},
{
"epoch": 0.20080321285140562,
"grad_norm": 0.08399316668510437,
"learning_rate": 5.9587379622760314e-05,
"loss": 1.9296,
"step": 50
},
{
"epoch": 0.20481927710843373,
"grad_norm": 0.08468279987573624,
"learning_rate": 5.95537935048649e-05,
"loss": 1.8012,
"step": 51
},
{
"epoch": 0.20883534136546184,
"grad_norm": 0.08074142038822174,
"learning_rate": 5.9518903652638376e-05,
"loss": 1.8227,
"step": 52
},
{
"epoch": 0.21285140562248997,
"grad_norm": 0.13968394696712494,
"learning_rate": 5.948271160520967e-05,
"loss": 1.8852,
"step": 53
},
{
"epoch": 0.21686746987951808,
"grad_norm": 0.14500737190246582,
"learning_rate": 5.944521895915265e-05,
"loss": 1.664,
"step": 54
},
{
"epoch": 0.22088353413654618,
"grad_norm": 0.15516269207000732,
"learning_rate": 5.9406427368415735e-05,
"loss": 1.8514,
"step": 55
},
{
"epoch": 0.2248995983935743,
"grad_norm": 0.07931994646787643,
"learning_rate": 5.93663385442489e-05,
"loss": 2.0106,
"step": 56
},
{
"epoch": 0.2289156626506024,
"grad_norm": 0.07893547415733337,
"learning_rate": 5.932495425512823e-05,
"loss": 1.7823,
"step": 57
},
{
"epoch": 0.23293172690763053,
"grad_norm": 0.10270941257476807,
"learning_rate": 5.928227632667782e-05,
"loss": 1.7455,
"step": 58
},
{
"epoch": 0.23694779116465864,
"grad_norm": 0.0862252414226532,
"learning_rate": 5.923830664158934e-05,
"loss": 1.8221,
"step": 59
},
{
"epoch": 0.24096385542168675,
"grad_norm": 0.09398238360881805,
"learning_rate": 5.919304713953893e-05,
"loss": 1.8726,
"step": 60
},
{
"epoch": 0.24497991967871485,
"grad_norm": 0.12090852856636047,
"learning_rate": 5.914649981710164e-05,
"loss": 1.9723,
"step": 61
},
{
"epoch": 0.24899598393574296,
"grad_norm": 0.10450229793787003,
"learning_rate": 5.909866672766334e-05,
"loss": 1.9845,
"step": 62
},
{
"epoch": 0.25301204819277107,
"grad_norm": 0.10860798507928848,
"learning_rate": 5.904954998133017e-05,
"loss": 1.9028,
"step": 63
},
{
"epoch": 0.2570281124497992,
"grad_norm": 0.13051600754261017,
"learning_rate": 5.8999151744835415e-05,
"loss": 1.7577,
"step": 64
},
{
"epoch": 0.26104417670682734,
"grad_norm": 0.15481741726398468,
"learning_rate": 5.894747424144397e-05,
"loss": 1.6495,
"step": 65
},
{
"epoch": 0.26506024096385544,
"grad_norm": 0.10475904494524002,
"learning_rate": 5.8894519750854215e-05,
"loss": 1.8233,
"step": 66
},
{
"epoch": 0.26907630522088355,
"grad_norm": 0.08438512682914734,
"learning_rate": 5.8840290609097454e-05,
"loss": 1.9323,
"step": 67
},
{
"epoch": 0.27309236947791166,
"grad_norm": 0.08830106258392334,
"learning_rate": 5.878478920843492e-05,
"loss": 1.8682,
"step": 68
},
{
"epoch": 0.27710843373493976,
"grad_norm": 0.09730786830186844,
"learning_rate": 5.872801799725218e-05,
"loss": 1.8755,
"step": 69
},
{
"epoch": 0.28112449799196787,
"grad_norm": 0.09512544423341751,
"learning_rate": 5.866997947995114e-05,
"loss": 1.6028,
"step": 70
},
{
"epoch": 0.285140562248996,
"grad_norm": 0.11627932637929916,
"learning_rate": 5.8610676216839586e-05,
"loss": 1.9151,
"step": 71
},
{
"epoch": 0.2891566265060241,
"grad_norm": 0.13668064773082733,
"learning_rate": 5.855011082401822e-05,
"loss": 1.664,
"step": 72
},
{
"epoch": 0.2931726907630522,
"grad_norm": 0.10782669484615326,
"learning_rate": 5.848828597326528e-05,
"loss": 1.7598,
"step": 73
},
{
"epoch": 0.2971887550200803,
"grad_norm": 0.09803726524114609,
"learning_rate": 5.842520439191862e-05,
"loss": 1.5882,
"step": 74
},
{
"epoch": 0.30120481927710846,
"grad_norm": 0.09667789936065674,
"learning_rate": 5.836086886275548e-05,
"loss": 2.0459,
"step": 75
},
{
"epoch": 0.30522088353413657,
"grad_norm": 0.10181078314781189,
"learning_rate": 5.829528222386966e-05,
"loss": 1.771,
"step": 76
},
{
"epoch": 0.3092369477911647,
"grad_norm": 0.08946649730205536,
"learning_rate": 5.8228447368546314e-05,
"loss": 1.7914,
"step": 77
},
{
"epoch": 0.3132530120481928,
"grad_norm": 0.0814245194196701,
"learning_rate": 5.816036724513439e-05,
"loss": 1.764,
"step": 78
},
{
"epoch": 0.3172690763052209,
"grad_norm": 0.07315659523010254,
"learning_rate": 5.809104485691649e-05,
"loss": 1.8567,
"step": 79
},
{
"epoch": 0.321285140562249,
"grad_norm": 0.07435451447963715,
"learning_rate": 5.802048326197644e-05,
"loss": 1.8489,
"step": 80
},
{
"epoch": 0.3253012048192771,
"grad_norm": 0.13348247110843658,
"learning_rate": 5.794868557306433e-05,
"loss": 1.6286,
"step": 81
},
{
"epoch": 0.3293172690763052,
"grad_norm": 0.0857180655002594,
"learning_rate": 5.787565495745924e-05,
"loss": 1.8996,
"step": 82
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.11295263469219208,
"learning_rate": 5.780139463682952e-05,
"loss": 1.5915,
"step": 83
},
{
"epoch": 0.3373493975903614,
"grad_norm": 0.08790762722492218,
"learning_rate": 5.7725907887090634e-05,
"loss": 1.7598,
"step": 84
},
{
"epoch": 0.3413654618473896,
"grad_norm": 0.08483127504587173,
"learning_rate": 5.764919803826068e-05,
"loss": 1.862,
"step": 85
},
{
"epoch": 0.3453815261044177,
"grad_norm": 0.08203373849391937,
"learning_rate": 5.757126847431347e-05,
"loss": 1.8408,
"step": 86
},
{
"epoch": 0.3493975903614458,
"grad_norm": 0.0757412239909172,
"learning_rate": 5.749212263302927e-05,
"loss": 1.8627,
"step": 87
},
{
"epoch": 0.3534136546184739,
"grad_norm": 0.09138002246618271,
"learning_rate": 5.741176400584313e-05,
"loss": 1.6849,
"step": 88
},
{
"epoch": 0.357429718875502,
"grad_norm": 0.1019127294421196,
"learning_rate": 5.733019613769086e-05,
"loss": 1.7678,
"step": 89
},
{
"epoch": 0.3614457831325301,
"grad_norm": 0.09306428581476212,
"learning_rate": 5.7247422626852666e-05,
"loss": 1.7371,
"step": 90
},
{
"epoch": 0.3654618473895582,
"grad_norm": 0.10577791184186935,
"learning_rate": 5.7163447124794405e-05,
"loss": 1.6762,
"step": 91
},
{
"epoch": 0.36947791164658633,
"grad_norm": 0.12335649132728577,
"learning_rate": 5.707827333600648e-05,
"loss": 1.7191,
"step": 92
},
{
"epoch": 0.37349397590361444,
"grad_norm": 0.10361435264348984,
"learning_rate": 5.699190501784049e-05,
"loss": 1.8953,
"step": 93
},
{
"epoch": 0.37751004016064255,
"grad_norm": 0.0855787992477417,
"learning_rate": 5.690434598034338e-05,
"loss": 1.884,
"step": 94
},
{
"epoch": 0.3815261044176707,
"grad_norm": 0.09389682114124298,
"learning_rate": 5.6815600086089486e-05,
"loss": 1.5828,
"step": 95
},
{
"epoch": 0.3855421686746988,
"grad_norm": 0.08702408522367477,
"learning_rate": 5.672567125001e-05,
"loss": 1.8452,
"step": 96
},
{
"epoch": 0.3895582329317269,
"grad_norm": 0.10916967689990997,
"learning_rate": 5.663456343922039e-05,
"loss": 1.6201,
"step": 97
},
{
"epoch": 0.39357429718875503,
"grad_norm": 0.081611767411232,
"learning_rate": 5.654228067284534e-05,
"loss": 1.9651,
"step": 98
},
{
"epoch": 0.39759036144578314,
"grad_norm": 0.12812922894954681,
"learning_rate": 5.6448827021841444e-05,
"loss": 1.7838,
"step": 99
},
{
"epoch": 0.40160642570281124,
"grad_norm": 0.08412948250770569,
"learning_rate": 5.635420660881763e-05,
"loss": 1.737,
"step": 100
},
{
"epoch": 0.40562248995983935,
"grad_norm": 0.09557431936264038,
"learning_rate": 5.6258423607853306e-05,
"loss": 1.8245,
"step": 101
},
{
"epoch": 0.40963855421686746,
"grad_norm": 0.10354951024055481,
"learning_rate": 5.616148224431423e-05,
"loss": 1.8694,
"step": 102
},
{
"epoch": 0.41365461847389556,
"grad_norm": 0.08835772424936295,
"learning_rate": 5.6063386794666075e-05,
"loss": 1.7883,
"step": 103
},
{
"epoch": 0.41767068273092367,
"grad_norm": 0.10105304419994354,
"learning_rate": 5.596414158628581e-05,
"loss": 1.6835,
"step": 104
},
{
"epoch": 0.42168674698795183,
"grad_norm": 0.13019445538520813,
"learning_rate": 5.586375099727081e-05,
"loss": 1.8853,
"step": 105
},
{
"epoch": 0.42570281124497994,
"grad_norm": 0.12298959493637085,
"learning_rate": 5.5762219456245676e-05,
"loss": 1.7237,
"step": 106
},
{
"epoch": 0.42971887550200805,
"grad_norm": 0.10004926472902298,
"learning_rate": 5.5659551442166924e-05,
"loss": 1.6178,
"step": 107
},
{
"epoch": 0.43373493975903615,
"grad_norm": 0.11601317673921585,
"learning_rate": 5.555575148412536e-05,
"loss": 1.6619,
"step": 108
},
{
"epoch": 0.43775100401606426,
"grad_norm": 0.09062661230564117,
"learning_rate": 5.545082416114632e-05,
"loss": 1.7387,
"step": 109
},
{
"epoch": 0.44176706827309237,
"grad_norm": 0.096687912940979,
"learning_rate": 5.534477410198763e-05,
"loss": 1.8114,
"step": 110
},
{
"epoch": 0.4457831325301205,
"grad_norm": 0.09439574927091599,
"learning_rate": 5.5237605984935435e-05,
"loss": 1.8035,
"step": 111
},
{
"epoch": 0.4497991967871486,
"grad_norm": 0.1305311769247055,
"learning_rate": 5.5129324537597846e-05,
"loss": 1.9107,
"step": 112
},
{
"epoch": 0.4538152610441767,
"grad_norm": 0.09376931935548782,
"learning_rate": 5.5019934536696336e-05,
"loss": 1.8721,
"step": 113
},
{
"epoch": 0.4578313253012048,
"grad_norm": 0.10925185680389404,
"learning_rate": 5.490944080785507e-05,
"loss": 1.8682,
"step": 114
},
{
"epoch": 0.46184738955823296,
"grad_norm": 0.1156877651810646,
"learning_rate": 5.479784822538796e-05,
"loss": 1.5833,
"step": 115
},
{
"epoch": 0.46586345381526106,
"grad_norm": 0.1127604991197586,
"learning_rate": 5.468516171208373e-05,
"loss": 1.6511,
"step": 116
},
{
"epoch": 0.46987951807228917,
"grad_norm": 0.08820759505033493,
"learning_rate": 5.4571386238988685e-05,
"loss": 1.569,
"step": 117
},
{
"epoch": 0.4738955823293173,
"grad_norm": 0.11805479973554611,
"learning_rate": 5.445652682518744e-05,
"loss": 1.7633,
"step": 118
},
{
"epoch": 0.4779116465863454,
"grad_norm": 0.10260842740535736,
"learning_rate": 5.434058853758151e-05,
"loss": 1.5363,
"step": 119
},
{
"epoch": 0.4819277108433735,
"grad_norm": 0.09707831591367722,
"learning_rate": 5.422357649066577e-05,
"loss": 1.8102,
"step": 120
},
{
"epoch": 0.4859437751004016,
"grad_norm": 0.1374276876449585,
"learning_rate": 5.4105495846302885e-05,
"loss": 1.6286,
"step": 121
},
{
"epoch": 0.4899598393574297,
"grad_norm": 0.1383811980485916,
"learning_rate": 5.398635181349553e-05,
"loss": 1.8758,
"step": 122
},
{
"epoch": 0.4939759036144578,
"grad_norm": 0.1613747775554657,
"learning_rate": 5.386614964815666e-05,
"loss": 1.8786,
"step": 123
},
{
"epoch": 0.4979919678714859,
"grad_norm": 0.12664808332920074,
"learning_rate": 5.3744894652877625e-05,
"loss": 2.0399,
"step": 124
},
{
"epoch": 0.5020080321285141,
"grad_norm": 0.10178882628679276,
"learning_rate": 5.362259217669424e-05,
"loss": 1.9754,
"step": 125
},
{
"epoch": 0.5060240963855421,
"grad_norm": 0.12219224870204926,
"learning_rate": 5.349924761485084e-05,
"loss": 1.6695,
"step": 126
},
{
"epoch": 0.5100401606425703,
"grad_norm": 0.11439970135688782,
"learning_rate": 5.3374866408562266e-05,
"loss": 1.9119,
"step": 127
},
{
"epoch": 0.5140562248995983,
"grad_norm": 0.09417600184679031,
"learning_rate": 5.324945404477383e-05,
"loss": 1.7473,
"step": 128
},
{
"epoch": 0.5180722891566265,
"grad_norm": 0.09558118879795074,
"learning_rate": 5.3123016055919275e-05,
"loss": 1.4051,
"step": 129
},
{
"epoch": 0.5220883534136547,
"grad_norm": 0.10001164674758911,
"learning_rate": 5.29955580196767e-05,
"loss": 1.7789,
"step": 130
},
{
"epoch": 0.5261044176706827,
"grad_norm": 0.09575448930263519,
"learning_rate": 5.2867085558722515e-05,
"loss": 1.7881,
"step": 131
},
{
"epoch": 0.5301204819277109,
"grad_norm": 0.09888429939746857,
"learning_rate": 5.27376043404834e-05,
"loss": 1.7188,
"step": 132
},
{
"epoch": 0.5341365461847389,
"grad_norm": 0.11489002406597137,
"learning_rate": 5.260712007688631e-05,
"loss": 1.5928,
"step": 133
},
{
"epoch": 0.5381526104417671,
"grad_norm": 0.10772331058979034,
"learning_rate": 5.247563852410647e-05,
"loss": 1.7395,
"step": 134
},
{
"epoch": 0.5421686746987951,
"grad_norm": 0.12452363222837448,
"learning_rate": 5.234316548231346e-05,
"loss": 1.7145,
"step": 135
},
{
"epoch": 0.5461847389558233,
"grad_norm": 0.162751242518425,
"learning_rate": 5.220970679541537e-05,
"loss": 1.93,
"step": 136
},
{
"epoch": 0.5502008032128514,
"grad_norm": 0.09236788004636765,
"learning_rate": 5.207526835080096e-05,
"loss": 1.7755,
"step": 137
},
{
"epoch": 0.5542168674698795,
"grad_norm": 0.09892769902944565,
"learning_rate": 5.193985607907996e-05,
"loss": 1.8125,
"step": 138
},
{
"epoch": 0.5582329317269076,
"grad_norm": 0.10891012102365494,
"learning_rate": 5.180347595382148e-05,
"loss": 1.7344,
"step": 139
},
{
"epoch": 0.5622489959839357,
"grad_norm": 0.10631144046783447,
"learning_rate": 5.166613399129045e-05,
"loss": 1.8008,
"step": 140
},
{
"epoch": 0.5662650602409639,
"grad_norm": 0.10682453960180283,
"learning_rate": 5.1527836250182216e-05,
"loss": 1.8495,
"step": 141
},
{
"epoch": 0.570281124497992,
"grad_norm": 0.10644228011369705,
"learning_rate": 5.138858883135532e-05,
"loss": 1.7763,
"step": 142
},
{
"epoch": 0.5742971887550201,
"grad_norm": 0.10813590884208679,
"learning_rate": 5.1248397877562306e-05,
"loss": 1.7689,
"step": 143
},
{
"epoch": 0.5783132530120482,
"grad_norm": 0.1288304626941681,
"learning_rate": 5.1107269573178765e-05,
"loss": 1.5501,
"step": 144
},
{
"epoch": 0.5823293172690763,
"grad_norm": 0.09014229476451874,
"learning_rate": 5.096521014393053e-05,
"loss": 2.0159,
"step": 145
},
{
"epoch": 0.5863453815261044,
"grad_norm": 0.11336657404899597,
"learning_rate": 5.082222585661902e-05,
"loss": 1.6987,
"step": 146
},
{
"epoch": 0.5903614457831325,
"grad_norm": 0.10714301466941833,
"learning_rate": 5.0678323018844784e-05,
"loss": 1.6675,
"step": 147
},
{
"epoch": 0.5943775100401606,
"grad_norm": 0.1200934424996376,
"learning_rate": 5.0533507978729236e-05,
"loss": 1.7993,
"step": 148
},
{
"epoch": 0.5983935742971888,
"grad_norm": 0.09276578575372696,
"learning_rate": 5.0387787124634674e-05,
"loss": 1.7931,
"step": 149
},
{
"epoch": 0.6024096385542169,
"grad_norm": 0.12235195189714432,
"learning_rate": 5.0241166884882394e-05,
"loss": 1.8777,
"step": 150
},
{
"epoch": 0.606425702811245,
"grad_norm": 0.10112816095352173,
"learning_rate": 5.0093653727469127e-05,
"loss": 1.6646,
"step": 151
},
{
"epoch": 0.6104417670682731,
"grad_norm": 0.10711057484149933,
"learning_rate": 4.994525415978175e-05,
"loss": 1.6652,
"step": 152
},
{
"epoch": 0.6144578313253012,
"grad_norm": 0.12113311886787415,
"learning_rate": 4.979597472831016e-05,
"loss": 1.8628,
"step": 153
},
{
"epoch": 0.6184738955823293,
"grad_norm": 0.10867751389741898,
"learning_rate": 4.964582201835856e-05,
"loss": 1.8879,
"step": 154
},
{
"epoch": 0.6224899598393574,
"grad_norm": 0.1424121856689453,
"learning_rate": 4.9494802653754846e-05,
"loss": 1.6689,
"step": 155
},
{
"epoch": 0.6265060240963856,
"grad_norm": 0.11042691767215729,
"learning_rate": 4.934292329655852e-05,
"loss": 1.5745,
"step": 156
},
{
"epoch": 0.6305220883534136,
"grad_norm": 0.11223538219928741,
"learning_rate": 4.919019064676674e-05,
"loss": 1.7756,
"step": 157
},
{
"epoch": 0.6345381526104418,
"grad_norm": 0.12498051673173904,
"learning_rate": 4.903661144201873e-05,
"loss": 1.6208,
"step": 158
},
{
"epoch": 0.6385542168674698,
"grad_norm": 0.11375685781240463,
"learning_rate": 4.888219245729863e-05,
"loss": 1.9195,
"step": 159
},
{
"epoch": 0.642570281124498,
"grad_norm": 0.11936648935079575,
"learning_rate": 4.872694050463656e-05,
"loss": 1.7615,
"step": 160
},
{
"epoch": 0.6465863453815262,
"grad_norm": 0.14738889038562775,
"learning_rate": 4.857086243280815e-05,
"loss": 1.4813,
"step": 161
},
{
"epoch": 0.6506024096385542,
"grad_norm": 0.10601120442152023,
"learning_rate": 4.8413965127032396e-05,
"loss": 1.7443,
"step": 162
},
{
"epoch": 0.6546184738955824,
"grad_norm": 0.16903774440288544,
"learning_rate": 4.825625550866795e-05,
"loss": 1.7467,
"step": 163
},
{
"epoch": 0.6586345381526104,
"grad_norm": 0.1252775341272354,
"learning_rate": 4.8097740534907745e-05,
"loss": 1.7344,
"step": 164
},
{
"epoch": 0.6626506024096386,
"grad_norm": 0.09926458448171616,
"learning_rate": 4.793842719847214e-05,
"loss": 1.8299,
"step": 165
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.11151187121868134,
"learning_rate": 4.7778322527300415e-05,
"loss": 1.9818,
"step": 166
},
{
"epoch": 0.6706827309236948,
"grad_norm": 0.09780837595462799,
"learning_rate": 4.761743358424073e-05,
"loss": 1.8249,
"step": 167
},
{
"epoch": 0.6746987951807228,
"grad_norm": 0.11365465819835663,
"learning_rate": 4.745576746673859e-05,
"loss": 1.7066,
"step": 168
},
{
"epoch": 0.678714859437751,
"grad_norm": 0.11123645305633545,
"learning_rate": 4.72933313065237e-05,
"loss": 1.7652,
"step": 169
},
{
"epoch": 0.6827309236947792,
"grad_norm": 0.10402809083461761,
"learning_rate": 4.713013226929543e-05,
"loss": 1.9567,
"step": 170
},
{
"epoch": 0.6867469879518072,
"grad_norm": 0.11607303470373154,
"learning_rate": 4.696617755440661e-05,
"loss": 1.7311,
"step": 171
},
{
"epoch": 0.6907630522088354,
"grad_norm": 0.11645929515361786,
"learning_rate": 4.6801474394546034e-05,
"loss": 1.8036,
"step": 172
},
{
"epoch": 0.6947791164658634,
"grad_norm": 0.1081853061914444,
"learning_rate": 4.663603005541933e-05,
"loss": 1.8126,
"step": 173
},
{
"epoch": 0.6987951807228916,
"grad_norm": 0.10500285774469376,
"learning_rate": 4.646985183542845e-05,
"loss": 1.7044,
"step": 174
},
{
"epoch": 0.7028112449799196,
"grad_norm": 0.11390309780836105,
"learning_rate": 4.630294706534977e-05,
"loss": 1.8124,
"step": 175
},
{
"epoch": 0.7068273092369478,
"grad_norm": 0.1308196634054184,
"learning_rate": 4.613532310801062e-05,
"loss": 1.6631,
"step": 176
},
{
"epoch": 0.7108433734939759,
"grad_norm": 0.11062182486057281,
"learning_rate": 4.5966987357964526e-05,
"loss": 1.6611,
"step": 177
},
{
"epoch": 0.714859437751004,
"grad_norm": 0.10045716166496277,
"learning_rate": 4.5797947241164984e-05,
"loss": 1.8025,
"step": 178
},
{
"epoch": 0.7188755020080321,
"grad_norm": 0.10317470133304596,
"learning_rate": 4.562821021463791e-05,
"loss": 1.7107,
"step": 179
},
{
"epoch": 0.7228915662650602,
"grad_norm": 0.11115267127752304,
"learning_rate": 4.545778376615263e-05,
"loss": 1.7796,
"step": 180
},
{
"epoch": 0.7269076305220884,
"grad_norm": 0.11856064200401306,
"learning_rate": 4.5286675413891584e-05,
"loss": 1.8696,
"step": 181
},
{
"epoch": 0.7309236947791165,
"grad_norm": 0.1215963363647461,
"learning_rate": 4.5114892706118725e-05,
"loss": 1.8084,
"step": 182
},
{
"epoch": 0.7349397590361446,
"grad_norm": 0.09853583574295044,
"learning_rate": 4.4942443220846415e-05,
"loss": 1.8725,
"step": 183
},
{
"epoch": 0.7389558232931727,
"grad_norm": 0.10426657646894455,
"learning_rate": 4.476933456550128e-05,
"loss": 1.6489,
"step": 184
},
{
"epoch": 0.7429718875502008,
"grad_norm": 0.10910302400588989,
"learning_rate": 4.459557437658844e-05,
"loss": 1.9391,
"step": 185
},
{
"epoch": 0.7469879518072289,
"grad_norm": 0.09922551363706589,
"learning_rate": 4.4421170319354825e-05,
"loss": 1.8068,
"step": 186
},
{
"epoch": 0.751004016064257,
"grad_norm": 0.12102405726909637,
"learning_rate": 4.4246130087450885e-05,
"loss": 1.7633,
"step": 187
},
{
"epoch": 0.7550200803212851,
"grad_norm": 0.1103704422712326,
"learning_rate": 4.407046140259123e-05,
"loss": 1.6617,
"step": 188
},
{
"epoch": 0.7590361445783133,
"grad_norm": 0.13233919441699982,
"learning_rate": 4.389417201421404e-05,
"loss": 1.6365,
"step": 189
},
{
"epoch": 0.7630522088353414,
"grad_norm": 0.1326707899570465,
"learning_rate": 4.3717269699139156e-05,
"loss": 1.7263,
"step": 190
},
{
"epoch": 0.7670682730923695,
"grad_norm": 0.11567298322916031,
"learning_rate": 4.3539762261225044e-05,
"loss": 1.7935,
"step": 191
},
{
"epoch": 0.7710843373493976,
"grad_norm": 0.10680199414491653,
"learning_rate": 4.336165753102451e-05,
"loss": 1.7901,
"step": 192
},
{
"epoch": 0.7751004016064257,
"grad_norm": 0.21424338221549988,
"learning_rate": 4.3182963365439275e-05,
"loss": 1.8038,
"step": 193
},
{
"epoch": 0.7791164658634538,
"grad_norm": 0.11191921681165695,
"learning_rate": 4.3003687647373405e-05,
"loss": 1.6104,
"step": 194
},
{
"epoch": 0.7831325301204819,
"grad_norm": 0.12618285417556763,
"learning_rate": 4.282383828538551e-05,
"loss": 1.7352,
"step": 195
},
{
"epoch": 0.7871485943775101,
"grad_norm": 0.11540846526622772,
"learning_rate": 4.264342321333989e-05,
"loss": 1.5843,
"step": 196
},
{
"epoch": 0.7911646586345381,
"grad_norm": 0.2725273668766022,
"learning_rate": 4.24624503900566e-05,
"loss": 1.7595,
"step": 197
},
{
"epoch": 0.7951807228915663,
"grad_norm": 0.13503190875053406,
"learning_rate": 4.228092779896021e-05,
"loss": 1.7162,
"step": 198
},
{
"epoch": 0.7991967871485943,
"grad_norm": 0.32556626200675964,
"learning_rate": 4.209886344772781e-05,
"loss": 1.7433,
"step": 199
},
{
"epoch": 0.8032128514056225,
"grad_norm": 0.11719845980405807,
"learning_rate": 4.191626536793561e-05,
"loss": 1.7366,
"step": 200
},
{
"epoch": 0.8072289156626506,
"grad_norm": 0.13130998611450195,
"learning_rate": 4.173314161470468e-05,
"loss": 1.8753,
"step": 201
},
{
"epoch": 0.8112449799196787,
"grad_norm": 0.14412151277065277,
"learning_rate": 4.154950026634566e-05,
"loss": 1.7503,
"step": 202
},
{
"epoch": 0.8152610441767069,
"grad_norm": 0.13965652883052826,
"learning_rate": 4.136534942400234e-05,
"loss": 1.6057,
"step": 203
},
{
"epoch": 0.8192771084337349,
"grad_norm": 0.1472691297531128,
"learning_rate": 4.118069721129427e-05,
"loss": 1.9009,
"step": 204
},
{
"epoch": 0.8232931726907631,
"grad_norm": 0.1335720419883728,
"learning_rate": 4.099555177395845e-05,
"loss": 1.7673,
"step": 205
},
{
"epoch": 0.8273092369477911,
"grad_norm": 0.1401488333940506,
"learning_rate": 4.0809921279489954e-05,
"loss": 1.7398,
"step": 206
},
{
"epoch": 0.8313253012048193,
"grad_norm": 0.12071944773197174,
"learning_rate": 4.062381391678163e-05,
"loss": 1.7446,
"step": 207
},
{
"epoch": 0.8353413654618473,
"grad_norm": 0.11511554569005966,
"learning_rate": 4.04372378957629e-05,
"loss": 1.6161,
"step": 208
},
{
"epoch": 0.8393574297188755,
"grad_norm": 0.1713258922100067,
"learning_rate": 4.025020144703749e-05,
"loss": 1.7174,
"step": 209
},
{
"epoch": 0.8433734939759037,
"grad_norm": 0.13804182410240173,
"learning_rate": 4.006271282152048e-05,
"loss": 1.5781,
"step": 210
},
{
"epoch": 0.8473895582329317,
"grad_norm": 0.17464035749435425,
"learning_rate": 3.9874780290074176e-05,
"loss": 1.9369,
"step": 211
},
{
"epoch": 0.8514056224899599,
"grad_norm": 0.11888068914413452,
"learning_rate": 3.9686412143143384e-05,
"loss": 1.8222,
"step": 212
},
{
"epoch": 0.8554216867469879,
"grad_norm": 0.14188675582408905,
"learning_rate": 3.94976166903896e-05,
"loss": 1.7729,
"step": 213
},
{
"epoch": 0.8594377510040161,
"grad_norm": 0.13000579178333282,
"learning_rate": 3.930840226032446e-05,
"loss": 1.7222,
"step": 214
},
{
"epoch": 0.8634538152610441,
"grad_norm": 0.14280284941196442,
"learning_rate": 3.911877719994235e-05,
"loss": 2.0587,
"step": 215
},
{
"epoch": 0.8674698795180723,
"grad_norm": 0.16663293540477753,
"learning_rate": 3.8928749874352185e-05,
"loss": 1.834,
"step": 216
},
{
"epoch": 0.8714859437751004,
"grad_norm": 0.09696398675441742,
"learning_rate": 3.873832866640836e-05,
"loss": 1.7275,
"step": 217
},
{
"epoch": 0.8755020080321285,
"grad_norm": 0.14156584441661835,
"learning_rate": 3.8547521976341004e-05,
"loss": 1.6058,
"step": 218
},
{
"epoch": 0.8795180722891566,
"grad_norm": 0.13290172815322876,
"learning_rate": 3.835633822138535e-05,
"loss": 1.6701,
"step": 219
},
{
"epoch": 0.8835341365461847,
"grad_norm": 0.13043315708637238,
"learning_rate": 3.816478583541048e-05,
"loss": 1.656,
"step": 220
},
{
"epoch": 0.8875502008032129,
"grad_norm": 0.1180863231420517,
"learning_rate": 3.797287326854723e-05,
"loss": 1.8315,
"step": 221
},
{
"epoch": 0.891566265060241,
"grad_norm": 0.12132333219051361,
"learning_rate": 3.778060898681541e-05,
"loss": 1.8825,
"step": 222
},
{
"epoch": 0.8955823293172691,
"grad_norm": 0.14479801058769226,
"learning_rate": 3.758800147175042e-05,
"loss": 1.6463,
"step": 223
},
{
"epoch": 0.8995983935742972,
"grad_norm": 0.11980578303337097,
"learning_rate": 3.739505922002898e-05,
"loss": 1.7547,
"step": 224
},
{
"epoch": 0.9036144578313253,
"grad_norm": 0.1616433560848236,
"learning_rate": 3.7201790743094374e-05,
"loss": 1.6393,
"step": 225
},
{
"epoch": 0.9076305220883534,
"grad_norm": 0.1393657922744751,
"learning_rate": 3.7008204566781004e-05,
"loss": 1.705,
"step": 226
},
{
"epoch": 0.9116465863453815,
"grad_norm": 0.13711820542812347,
"learning_rate": 3.681430923093818e-05,
"loss": 1.7637,
"step": 227
},
{
"epoch": 0.9156626506024096,
"grad_norm": 0.1211545392870903,
"learning_rate": 3.662011328905351e-05,
"loss": 1.8126,
"step": 228
},
{
"epoch": 0.9196787148594378,
"grad_norm": 0.11105578392744064,
"learning_rate": 3.64256253078755e-05,
"loss": 1.5691,
"step": 229
},
{
"epoch": 0.9236947791164659,
"grad_norm": 0.19085654616355896,
"learning_rate": 3.6230853867035654e-05,
"loss": 1.6751,
"step": 230
},
{
"epoch": 0.927710843373494,
"grad_norm": 0.12571419775485992,
"learning_rate": 3.603580755867001e-05,
"loss": 1.589,
"step": 231
},
{
"epoch": 0.9317269076305221,
"grad_norm": 0.1361204832792282,
"learning_rate": 3.58404949870401e-05,
"loss": 1.8733,
"step": 232
},
{
"epoch": 0.9357429718875502,
"grad_norm": 0.11016932129859924,
"learning_rate": 3.5644924768153364e-05,
"loss": 1.898,
"step": 233
},
{
"epoch": 0.9397590361445783,
"grad_norm": 0.12060891091823578,
"learning_rate": 3.544910552938309e-05,
"loss": 1.7406,
"step": 234
},
{
"epoch": 0.9437751004016064,
"grad_norm": 0.11152440309524536,
"learning_rate": 3.5253045909087813e-05,
"loss": 1.9775,
"step": 235
},
{
"epoch": 0.9477911646586346,
"grad_norm": 0.13362078368663788,
"learning_rate": 3.505675455623023e-05,
"loss": 1.5462,
"step": 236
},
{
"epoch": 0.9518072289156626,
"grad_norm": 0.17355649173259735,
"learning_rate": 3.4860240129995696e-05,
"loss": 1.9621,
"step": 237
},
{
"epoch": 0.9558232931726908,
"grad_norm": 0.10163702070713043,
"learning_rate": 3.4663511299410203e-05,
"loss": 1.9847,
"step": 238
},
{
"epoch": 0.9598393574297188,
"grad_norm": 0.12204550206661224,
"learning_rate": 3.446657674295796e-05,
"loss": 1.8851,
"step": 239
},
{
"epoch": 0.963855421686747,
"grad_norm": 0.14867813885211945,
"learning_rate": 3.426944514819856e-05,
"loss": 1.6835,
"step": 240
},
{
"epoch": 0.9678714859437751,
"grad_norm": 0.12139367312192917,
"learning_rate": 3.407212521138372e-05,
"loss": 1.8269,
"step": 241
},
{
"epoch": 0.9718875502008032,
"grad_norm": 0.2183162420988083,
"learning_rate": 3.3874625637073704e-05,
"loss": 1.6785,
"step": 242
},
{
"epoch": 0.9759036144578314,
"grad_norm": 0.1285049170255661,
"learning_rate": 3.367695513775324e-05,
"loss": 1.6806,
"step": 243
},
{
"epoch": 0.9799196787148594,
"grad_norm": 0.12188960611820221,
"learning_rate": 3.347912243344727e-05,
"loss": 1.7203,
"step": 244
},
{
"epoch": 0.9839357429718876,
"grad_norm": 0.15413041412830353,
"learning_rate": 3.328113625133624e-05,
"loss": 1.6022,
"step": 245
},
{
"epoch": 0.9879518072289156,
"grad_norm": 0.1414673924446106,
"learning_rate": 3.308300532537107e-05,
"loss": 1.741,
"step": 246
},
{
"epoch": 0.9919678714859438,
"grad_norm": 0.1399112194776535,
"learning_rate": 3.288473839588796e-05,
"loss": 1.5319,
"step": 247
},
{
"epoch": 0.9959839357429718,
"grad_norm": 0.13522355258464813,
"learning_rate": 3.26863442092227e-05,
"loss": 1.6901,
"step": 248
},
{
"epoch": 1.0,
"grad_norm": 0.11732550710439682,
"learning_rate": 3.248783151732494e-05,
"loss": 1.8542,
"step": 249
},
{
"epoch": 1.0040160642570282,
"grad_norm": 0.26141083240509033,
"learning_rate": 3.2289209077372035e-05,
"loss": 3.761,
"step": 250
}
],
"logging_steps": 1,
"max_steps": 498,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 25,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.4115757240237425e+19,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}