jeqcho's picture
Upload folder using huggingface_hub
2406fb4 verified
Raw
History Blame Contribute Delete
60.4 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.2703296703296703,
"eval_steps": 500,
"global_step": 345,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006593406593406593,
"grad_norm": 2.303626298904419,
"learning_rate": 0.0,
"loss": 3.6085,
"step": 1
},
{
"epoch": 0.013186813186813187,
"grad_norm": 1.439454436302185,
"learning_rate": 4e-05,
"loss": 2.4913,
"step": 2
},
{
"epoch": 0.01978021978021978,
"grad_norm": 1.7588273286819458,
"learning_rate": 8e-05,
"loss": 2.6433,
"step": 3
},
{
"epoch": 0.026373626373626374,
"grad_norm": 1.5198580026626587,
"learning_rate": 0.00012,
"loss": 2.3145,
"step": 4
},
{
"epoch": 0.03296703296703297,
"grad_norm": 1.8622243404388428,
"learning_rate": 0.00016,
"loss": 2.4423,
"step": 5
},
{
"epoch": 0.03956043956043956,
"grad_norm": 2.6957828998565674,
"learning_rate": 0.0002,
"loss": 2.1699,
"step": 6
},
{
"epoch": 0.046153846153846156,
"grad_norm": 2.3653955459594727,
"learning_rate": 0.00019955654101995565,
"loss": 1.5512,
"step": 7
},
{
"epoch": 0.05274725274725275,
"grad_norm": 1.867249846458435,
"learning_rate": 0.00019911308203991133,
"loss": 1.4124,
"step": 8
},
{
"epoch": 0.05934065934065934,
"grad_norm": 0.9426868557929993,
"learning_rate": 0.00019866962305986697,
"loss": 1.4026,
"step": 9
},
{
"epoch": 0.06593406593406594,
"grad_norm": 1.105395793914795,
"learning_rate": 0.00019822616407982261,
"loss": 1.2543,
"step": 10
},
{
"epoch": 0.07252747252747253,
"grad_norm": 0.788772702217102,
"learning_rate": 0.00019778270509977829,
"loss": 1.2154,
"step": 11
},
{
"epoch": 0.07912087912087912,
"grad_norm": 0.53281569480896,
"learning_rate": 0.00019733924611973393,
"loss": 1.2039,
"step": 12
},
{
"epoch": 0.08571428571428572,
"grad_norm": 0.40713539719581604,
"learning_rate": 0.0001968957871396896,
"loss": 1.0836,
"step": 13
},
{
"epoch": 0.09230769230769231,
"grad_norm": 0.5409032702445984,
"learning_rate": 0.00019645232815964525,
"loss": 1.1098,
"step": 14
},
{
"epoch": 0.0989010989010989,
"grad_norm": 0.47973766922950745,
"learning_rate": 0.00019600886917960092,
"loss": 1.19,
"step": 15
},
{
"epoch": 0.1054945054945055,
"grad_norm": 0.5039457678794861,
"learning_rate": 0.00019556541019955653,
"loss": 1.1809,
"step": 16
},
{
"epoch": 0.11208791208791209,
"grad_norm": 0.432468056678772,
"learning_rate": 0.0001951219512195122,
"loss": 1.1579,
"step": 17
},
{
"epoch": 0.11868131868131868,
"grad_norm": 0.49248069524765015,
"learning_rate": 0.00019467849223946785,
"loss": 1.0721,
"step": 18
},
{
"epoch": 0.12527472527472527,
"grad_norm": 0.506483256816864,
"learning_rate": 0.00019423503325942352,
"loss": 1.2076,
"step": 19
},
{
"epoch": 0.13186813186813187,
"grad_norm": 0.3699564039707184,
"learning_rate": 0.00019379157427937917,
"loss": 0.8558,
"step": 20
},
{
"epoch": 0.13846153846153847,
"grad_norm": 0.40447086095809937,
"learning_rate": 0.00019334811529933484,
"loss": 0.9758,
"step": 21
},
{
"epoch": 0.14505494505494507,
"grad_norm": 0.42658472061157227,
"learning_rate": 0.00019290465631929045,
"loss": 1.062,
"step": 22
},
{
"epoch": 0.15164835164835164,
"grad_norm": 0.3948553204536438,
"learning_rate": 0.00019246119733924613,
"loss": 0.9631,
"step": 23
},
{
"epoch": 0.15824175824175823,
"grad_norm": 0.32725754380226135,
"learning_rate": 0.00019201773835920177,
"loss": 1.0747,
"step": 24
},
{
"epoch": 0.16483516483516483,
"grad_norm": 0.33828970789909363,
"learning_rate": 0.00019157427937915744,
"loss": 1.0653,
"step": 25
},
{
"epoch": 0.17142857142857143,
"grad_norm": 0.27442067861557007,
"learning_rate": 0.00019113082039911309,
"loss": 0.9254,
"step": 26
},
{
"epoch": 0.17802197802197803,
"grad_norm": 0.27329978346824646,
"learning_rate": 0.00019068736141906876,
"loss": 0.9075,
"step": 27
},
{
"epoch": 0.18461538461538463,
"grad_norm": 0.41785359382629395,
"learning_rate": 0.0001902439024390244,
"loss": 1.0079,
"step": 28
},
{
"epoch": 0.1912087912087912,
"grad_norm": 0.282429963350296,
"learning_rate": 0.00018980044345898005,
"loss": 0.8602,
"step": 29
},
{
"epoch": 0.1978021978021978,
"grad_norm": 0.37518325448036194,
"learning_rate": 0.00018935698447893572,
"loss": 1.0907,
"step": 30
},
{
"epoch": 0.2043956043956044,
"grad_norm": 0.2990609407424927,
"learning_rate": 0.00018891352549889136,
"loss": 0.8994,
"step": 31
},
{
"epoch": 0.210989010989011,
"grad_norm": 0.33431681990623474,
"learning_rate": 0.00018847006651884703,
"loss": 0.9115,
"step": 32
},
{
"epoch": 0.2175824175824176,
"grad_norm": 0.31373974680900574,
"learning_rate": 0.00018802660753880268,
"loss": 0.9089,
"step": 33
},
{
"epoch": 0.22417582417582418,
"grad_norm": 0.2787388563156128,
"learning_rate": 0.00018758314855875832,
"loss": 1.0808,
"step": 34
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.3206472396850586,
"learning_rate": 0.00018713968957871397,
"loss": 0.9184,
"step": 35
},
{
"epoch": 0.23736263736263735,
"grad_norm": 0.2402617186307907,
"learning_rate": 0.00018669623059866964,
"loss": 0.7528,
"step": 36
},
{
"epoch": 0.24395604395604395,
"grad_norm": 0.34411153197288513,
"learning_rate": 0.00018625277161862528,
"loss": 0.9755,
"step": 37
},
{
"epoch": 0.25054945054945055,
"grad_norm": 0.34733089804649353,
"learning_rate": 0.00018580931263858095,
"loss": 1.0181,
"step": 38
},
{
"epoch": 0.2571428571428571,
"grad_norm": 0.3262622356414795,
"learning_rate": 0.0001853658536585366,
"loss": 1.0464,
"step": 39
},
{
"epoch": 0.26373626373626374,
"grad_norm": 0.24500015377998352,
"learning_rate": 0.00018492239467849224,
"loss": 0.9446,
"step": 40
},
{
"epoch": 0.2703296703296703,
"grad_norm": 0.2835443913936615,
"learning_rate": 0.0001844789356984479,
"loss": 0.9856,
"step": 41
},
{
"epoch": 0.27692307692307694,
"grad_norm": 0.27747032046318054,
"learning_rate": 0.00018403547671840356,
"loss": 0.9622,
"step": 42
},
{
"epoch": 0.2835164835164835,
"grad_norm": 0.34073948860168457,
"learning_rate": 0.0001835920177383592,
"loss": 0.9465,
"step": 43
},
{
"epoch": 0.29010989010989013,
"grad_norm": 0.26712504029273987,
"learning_rate": 0.00018314855875831487,
"loss": 1.0602,
"step": 44
},
{
"epoch": 0.2967032967032967,
"grad_norm": 0.37089070677757263,
"learning_rate": 0.00018270509977827052,
"loss": 1.0997,
"step": 45
},
{
"epoch": 0.3032967032967033,
"grad_norm": 0.30469098687171936,
"learning_rate": 0.00018226164079822616,
"loss": 0.9438,
"step": 46
},
{
"epoch": 0.3098901098901099,
"grad_norm": 0.31314560770988464,
"learning_rate": 0.00018181818181818183,
"loss": 0.8367,
"step": 47
},
{
"epoch": 0.31648351648351647,
"grad_norm": 0.33173537254333496,
"learning_rate": 0.00018137472283813748,
"loss": 1.0277,
"step": 48
},
{
"epoch": 0.3230769230769231,
"grad_norm": 0.2833123803138733,
"learning_rate": 0.00018093126385809312,
"loss": 0.7779,
"step": 49
},
{
"epoch": 0.32967032967032966,
"grad_norm": 0.28676313161849976,
"learning_rate": 0.0001804878048780488,
"loss": 1.0172,
"step": 50
},
{
"epoch": 0.3362637362637363,
"grad_norm": 0.25923117995262146,
"learning_rate": 0.00018004434589800444,
"loss": 0.7737,
"step": 51
},
{
"epoch": 0.34285714285714286,
"grad_norm": 0.2750239074230194,
"learning_rate": 0.00017960088691796008,
"loss": 0.8973,
"step": 52
},
{
"epoch": 0.34945054945054943,
"grad_norm": 0.30605781078338623,
"learning_rate": 0.00017915742793791575,
"loss": 0.9545,
"step": 53
},
{
"epoch": 0.35604395604395606,
"grad_norm": 0.2281476855278015,
"learning_rate": 0.0001787139689578714,
"loss": 0.7999,
"step": 54
},
{
"epoch": 0.3626373626373626,
"grad_norm": 0.3172890245914459,
"learning_rate": 0.00017827050997782707,
"loss": 0.893,
"step": 55
},
{
"epoch": 0.36923076923076925,
"grad_norm": 0.38381415605545044,
"learning_rate": 0.00017782705099778271,
"loss": 1.0448,
"step": 56
},
{
"epoch": 0.3758241758241758,
"grad_norm": 0.3915952146053314,
"learning_rate": 0.00017738359201773839,
"loss": 1.0564,
"step": 57
},
{
"epoch": 0.3824175824175824,
"grad_norm": 0.33569052815437317,
"learning_rate": 0.000176940133037694,
"loss": 1.0455,
"step": 58
},
{
"epoch": 0.389010989010989,
"grad_norm": 0.4101047217845917,
"learning_rate": 0.00017649667405764967,
"loss": 1.0923,
"step": 59
},
{
"epoch": 0.3956043956043956,
"grad_norm": 0.42315059900283813,
"learning_rate": 0.00017605321507760532,
"loss": 1.0163,
"step": 60
},
{
"epoch": 0.4021978021978022,
"grad_norm": 0.2661890387535095,
"learning_rate": 0.000175609756097561,
"loss": 0.7853,
"step": 61
},
{
"epoch": 0.4087912087912088,
"grad_norm": 0.3342932462692261,
"learning_rate": 0.00017516629711751663,
"loss": 0.9406,
"step": 62
},
{
"epoch": 0.4153846153846154,
"grad_norm": 0.3210691809654236,
"learning_rate": 0.0001747228381374723,
"loss": 0.8858,
"step": 63
},
{
"epoch": 0.421978021978022,
"grad_norm": 0.2718408703804016,
"learning_rate": 0.00017427937915742792,
"loss": 0.9452,
"step": 64
},
{
"epoch": 0.42857142857142855,
"grad_norm": 0.36029088497161865,
"learning_rate": 0.0001738359201773836,
"loss": 0.8777,
"step": 65
},
{
"epoch": 0.4351648351648352,
"grad_norm": 0.35389444231987,
"learning_rate": 0.00017339246119733924,
"loss": 1.0064,
"step": 66
},
{
"epoch": 0.44175824175824174,
"grad_norm": 0.3141503930091858,
"learning_rate": 0.0001729490022172949,
"loss": 0.8785,
"step": 67
},
{
"epoch": 0.44835164835164837,
"grad_norm": 0.27303633093833923,
"learning_rate": 0.00017250554323725056,
"loss": 0.8559,
"step": 68
},
{
"epoch": 0.45494505494505494,
"grad_norm": 0.3359520137310028,
"learning_rate": 0.00017206208425720623,
"loss": 1.096,
"step": 69
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.30675244331359863,
"learning_rate": 0.00017161862527716187,
"loss": 1.0009,
"step": 70
},
{
"epoch": 0.46813186813186813,
"grad_norm": 0.3307923972606659,
"learning_rate": 0.00017117516629711752,
"loss": 1.0056,
"step": 71
},
{
"epoch": 0.4747252747252747,
"grad_norm": 0.24969322979450226,
"learning_rate": 0.0001707317073170732,
"loss": 0.8563,
"step": 72
},
{
"epoch": 0.48131868131868133,
"grad_norm": 0.24619445204734802,
"learning_rate": 0.00017028824833702883,
"loss": 0.8386,
"step": 73
},
{
"epoch": 0.4879120879120879,
"grad_norm": 0.3848130702972412,
"learning_rate": 0.0001698447893569845,
"loss": 1.2539,
"step": 74
},
{
"epoch": 0.4945054945054945,
"grad_norm": 0.35728219151496887,
"learning_rate": 0.00016940133037694015,
"loss": 0.9608,
"step": 75
},
{
"epoch": 0.5010989010989011,
"grad_norm": 0.41609638929367065,
"learning_rate": 0.00016895787139689582,
"loss": 1.1981,
"step": 76
},
{
"epoch": 0.5076923076923077,
"grad_norm": 0.3290999233722687,
"learning_rate": 0.00016851441241685144,
"loss": 1.0885,
"step": 77
},
{
"epoch": 0.5142857142857142,
"grad_norm": 0.3357957899570465,
"learning_rate": 0.0001680709534368071,
"loss": 0.9842,
"step": 78
},
{
"epoch": 0.5208791208791209,
"grad_norm": 0.37297511100769043,
"learning_rate": 0.00016762749445676275,
"loss": 1.0532,
"step": 79
},
{
"epoch": 0.5274725274725275,
"grad_norm": 0.333650678396225,
"learning_rate": 0.00016718403547671842,
"loss": 1.0956,
"step": 80
},
{
"epoch": 0.5340659340659341,
"grad_norm": 0.3169361650943756,
"learning_rate": 0.00016674057649667407,
"loss": 0.9588,
"step": 81
},
{
"epoch": 0.5406593406593406,
"grad_norm": 0.28924304246902466,
"learning_rate": 0.00016629711751662974,
"loss": 0.9038,
"step": 82
},
{
"epoch": 0.5472527472527473,
"grad_norm": 0.2601667046546936,
"learning_rate": 0.00016585365853658536,
"loss": 0.7903,
"step": 83
},
{
"epoch": 0.5538461538461539,
"grad_norm": 0.28349757194519043,
"learning_rate": 0.00016541019955654103,
"loss": 0.8989,
"step": 84
},
{
"epoch": 0.5604395604395604,
"grad_norm": 0.37565159797668457,
"learning_rate": 0.00016496674057649667,
"loss": 1.1042,
"step": 85
},
{
"epoch": 0.567032967032967,
"grad_norm": 0.38468965888023376,
"learning_rate": 0.00016452328159645234,
"loss": 1.1407,
"step": 86
},
{
"epoch": 0.5736263736263736,
"grad_norm": 0.28541499376296997,
"learning_rate": 0.000164079822616408,
"loss": 0.8125,
"step": 87
},
{
"epoch": 0.5802197802197803,
"grad_norm": 0.2994617223739624,
"learning_rate": 0.00016363636363636366,
"loss": 0.9195,
"step": 88
},
{
"epoch": 0.5868131868131868,
"grad_norm": 0.3976408839225769,
"learning_rate": 0.0001631929046563193,
"loss": 1.0156,
"step": 89
},
{
"epoch": 0.5934065934065934,
"grad_norm": 0.30622562766075134,
"learning_rate": 0.00016274944567627495,
"loss": 0.9982,
"step": 90
},
{
"epoch": 0.6,
"grad_norm": 0.28931549191474915,
"learning_rate": 0.0001623059866962306,
"loss": 0.928,
"step": 91
},
{
"epoch": 0.6065934065934065,
"grad_norm": 0.31948480010032654,
"learning_rate": 0.00016186252771618626,
"loss": 1.0779,
"step": 92
},
{
"epoch": 0.6131868131868132,
"grad_norm": 0.33104556798934937,
"learning_rate": 0.0001614190687361419,
"loss": 1.0611,
"step": 93
},
{
"epoch": 0.6197802197802198,
"grad_norm": 0.43478959798812866,
"learning_rate": 0.00016097560975609758,
"loss": 0.9161,
"step": 94
},
{
"epoch": 0.6263736263736264,
"grad_norm": 0.32590118050575256,
"learning_rate": 0.00016053215077605322,
"loss": 1.1696,
"step": 95
},
{
"epoch": 0.6329670329670329,
"grad_norm": 0.25838395953178406,
"learning_rate": 0.00016008869179600887,
"loss": 0.7856,
"step": 96
},
{
"epoch": 0.6395604395604395,
"grad_norm": 0.3152015507221222,
"learning_rate": 0.00015964523281596454,
"loss": 0.7904,
"step": 97
},
{
"epoch": 0.6461538461538462,
"grad_norm": 0.34387698769569397,
"learning_rate": 0.00015920177383592018,
"loss": 0.991,
"step": 98
},
{
"epoch": 0.6527472527472528,
"grad_norm": 0.2822190821170807,
"learning_rate": 0.00015875831485587586,
"loss": 0.9494,
"step": 99
},
{
"epoch": 0.6593406593406593,
"grad_norm": 0.26281842589378357,
"learning_rate": 0.0001583148558758315,
"loss": 0.9059,
"step": 100
},
{
"epoch": 0.6659340659340659,
"grad_norm": 0.42032739520072937,
"learning_rate": 0.00015787139689578714,
"loss": 1.0635,
"step": 101
},
{
"epoch": 0.6725274725274726,
"grad_norm": 0.2608965337276459,
"learning_rate": 0.0001574279379157428,
"loss": 0.6873,
"step": 102
},
{
"epoch": 0.6791208791208792,
"grad_norm": 0.25404682755470276,
"learning_rate": 0.00015698447893569846,
"loss": 0.9051,
"step": 103
},
{
"epoch": 0.6857142857142857,
"grad_norm": 0.34183400869369507,
"learning_rate": 0.0001565410199556541,
"loss": 0.9823,
"step": 104
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.29854243993759155,
"learning_rate": 0.00015609756097560978,
"loss": 0.9067,
"step": 105
},
{
"epoch": 0.6989010989010989,
"grad_norm": 0.3323337435722351,
"learning_rate": 0.00015565410199556542,
"loss": 0.8381,
"step": 106
},
{
"epoch": 0.7054945054945055,
"grad_norm": 0.30771464109420776,
"learning_rate": 0.00015521064301552106,
"loss": 0.8658,
"step": 107
},
{
"epoch": 0.7120879120879121,
"grad_norm": 0.26583606004714966,
"learning_rate": 0.0001547671840354767,
"loss": 0.8458,
"step": 108
},
{
"epoch": 0.7186813186813187,
"grad_norm": 0.2765386402606964,
"learning_rate": 0.00015432372505543238,
"loss": 0.7864,
"step": 109
},
{
"epoch": 0.7252747252747253,
"grad_norm": 0.33523914217948914,
"learning_rate": 0.00015388026607538802,
"loss": 0.9053,
"step": 110
},
{
"epoch": 0.7318681318681318,
"grad_norm": 0.3297010660171509,
"learning_rate": 0.0001534368070953437,
"loss": 0.8264,
"step": 111
},
{
"epoch": 0.7384615384615385,
"grad_norm": 0.2933356761932373,
"learning_rate": 0.00015299334811529934,
"loss": 0.8524,
"step": 112
},
{
"epoch": 0.7450549450549451,
"grad_norm": 0.31128862500190735,
"learning_rate": 0.00015254988913525498,
"loss": 0.891,
"step": 113
},
{
"epoch": 0.7516483516483516,
"grad_norm": 0.37845343351364136,
"learning_rate": 0.00015210643015521066,
"loss": 1.2966,
"step": 114
},
{
"epoch": 0.7582417582417582,
"grad_norm": 0.3646884560585022,
"learning_rate": 0.0001516629711751663,
"loss": 1.0198,
"step": 115
},
{
"epoch": 0.7648351648351648,
"grad_norm": 0.3193872570991516,
"learning_rate": 0.00015121951219512197,
"loss": 1.1952,
"step": 116
},
{
"epoch": 0.7714285714285715,
"grad_norm": 0.25547274947166443,
"learning_rate": 0.00015077605321507762,
"loss": 0.9888,
"step": 117
},
{
"epoch": 0.778021978021978,
"grad_norm": 0.2840863764286041,
"learning_rate": 0.0001503325942350333,
"loss": 0.8643,
"step": 118
},
{
"epoch": 0.7846153846153846,
"grad_norm": 0.349303662776947,
"learning_rate": 0.0001498891352549889,
"loss": 0.7961,
"step": 119
},
{
"epoch": 0.7912087912087912,
"grad_norm": 0.25125086307525635,
"learning_rate": 0.00014944567627494458,
"loss": 0.8055,
"step": 120
},
{
"epoch": 0.7978021978021979,
"grad_norm": 0.2910859286785126,
"learning_rate": 0.00014900221729490022,
"loss": 0.6735,
"step": 121
},
{
"epoch": 0.8043956043956044,
"grad_norm": 0.28555774688720703,
"learning_rate": 0.0001485587583148559,
"loss": 0.864,
"step": 122
},
{
"epoch": 0.810989010989011,
"grad_norm": 0.28217050433158875,
"learning_rate": 0.00014811529933481154,
"loss": 0.8938,
"step": 123
},
{
"epoch": 0.8175824175824176,
"grad_norm": 0.33504718542099,
"learning_rate": 0.0001476718403547672,
"loss": 0.9377,
"step": 124
},
{
"epoch": 0.8241758241758241,
"grad_norm": 0.2877736985683441,
"learning_rate": 0.00014722838137472282,
"loss": 0.9173,
"step": 125
},
{
"epoch": 0.8307692307692308,
"grad_norm": 0.3107459247112274,
"learning_rate": 0.0001467849223946785,
"loss": 0.828,
"step": 126
},
{
"epoch": 0.8373626373626374,
"grad_norm": 0.33110880851745605,
"learning_rate": 0.00014634146341463414,
"loss": 0.8686,
"step": 127
},
{
"epoch": 0.843956043956044,
"grad_norm": 0.30587294697761536,
"learning_rate": 0.0001458980044345898,
"loss": 1.0425,
"step": 128
},
{
"epoch": 0.8505494505494505,
"grad_norm": 0.3741580843925476,
"learning_rate": 0.00014545454545454546,
"loss": 1.0715,
"step": 129
},
{
"epoch": 0.8571428571428571,
"grad_norm": 0.3586806654930115,
"learning_rate": 0.00014501108647450113,
"loss": 1.1221,
"step": 130
},
{
"epoch": 0.8637362637362638,
"grad_norm": 0.36812883615493774,
"learning_rate": 0.00014456762749445675,
"loss": 1.0742,
"step": 131
},
{
"epoch": 0.8703296703296703,
"grad_norm": 0.32173845171928406,
"learning_rate": 0.00014412416851441242,
"loss": 1.0443,
"step": 132
},
{
"epoch": 0.8769230769230769,
"grad_norm": 0.38413581252098083,
"learning_rate": 0.00014368070953436806,
"loss": 1.0525,
"step": 133
},
{
"epoch": 0.8835164835164835,
"grad_norm": 0.46863028407096863,
"learning_rate": 0.00014323725055432373,
"loss": 0.8269,
"step": 134
},
{
"epoch": 0.8901098901098901,
"grad_norm": 0.3614848256111145,
"learning_rate": 0.00014279379157427938,
"loss": 1.0206,
"step": 135
},
{
"epoch": 0.8967032967032967,
"grad_norm": 0.2927537262439728,
"learning_rate": 0.00014235033259423505,
"loss": 0.9561,
"step": 136
},
{
"epoch": 0.9032967032967033,
"grad_norm": 0.42427268624305725,
"learning_rate": 0.0001419068736141907,
"loss": 1.0297,
"step": 137
},
{
"epoch": 0.9098901098901099,
"grad_norm": 0.3389011025428772,
"learning_rate": 0.00014146341463414634,
"loss": 0.8741,
"step": 138
},
{
"epoch": 0.9164835164835164,
"grad_norm": 0.3461213707923889,
"learning_rate": 0.000141019955654102,
"loss": 0.9779,
"step": 139
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.4119606912136078,
"learning_rate": 0.00014057649667405765,
"loss": 1.0071,
"step": 140
},
{
"epoch": 0.9296703296703297,
"grad_norm": 0.27475905418395996,
"learning_rate": 0.00014013303769401332,
"loss": 0.8762,
"step": 141
},
{
"epoch": 0.9362637362637363,
"grad_norm": 0.36642834544181824,
"learning_rate": 0.00013968957871396897,
"loss": 0.9655,
"step": 142
},
{
"epoch": 0.9428571428571428,
"grad_norm": 0.3168472945690155,
"learning_rate": 0.00013924611973392464,
"loss": 0.8674,
"step": 143
},
{
"epoch": 0.9494505494505494,
"grad_norm": 0.5073333382606506,
"learning_rate": 0.00013880266075388026,
"loss": 0.7491,
"step": 144
},
{
"epoch": 0.9560439560439561,
"grad_norm": 0.2707488536834717,
"learning_rate": 0.00013835920177383593,
"loss": 1.0252,
"step": 145
},
{
"epoch": 0.9626373626373627,
"grad_norm": 0.3995589315891266,
"learning_rate": 0.00013791574279379157,
"loss": 0.93,
"step": 146
},
{
"epoch": 0.9692307692307692,
"grad_norm": 0.3536117672920227,
"learning_rate": 0.00013747228381374724,
"loss": 0.9061,
"step": 147
},
{
"epoch": 0.9758241758241758,
"grad_norm": 0.34937381744384766,
"learning_rate": 0.0001370288248337029,
"loss": 0.9136,
"step": 148
},
{
"epoch": 0.9824175824175824,
"grad_norm": 0.29199913144111633,
"learning_rate": 0.00013658536585365856,
"loss": 0.9812,
"step": 149
},
{
"epoch": 0.989010989010989,
"grad_norm": 0.37444406747817993,
"learning_rate": 0.00013614190687361418,
"loss": 0.9576,
"step": 150
},
{
"epoch": 0.9956043956043956,
"grad_norm": 0.3482721149921417,
"learning_rate": 0.00013569844789356985,
"loss": 0.9318,
"step": 151
},
{
"epoch": 1.0,
"grad_norm": 0.3411265015602112,
"learning_rate": 0.0001352549889135255,
"loss": 0.7835,
"step": 152
},
{
"epoch": 1.0065934065934066,
"grad_norm": 0.3751468062400818,
"learning_rate": 0.00013481152993348116,
"loss": 0.9466,
"step": 153
},
{
"epoch": 1.0131868131868131,
"grad_norm": 0.2971190810203552,
"learning_rate": 0.0001343680709534368,
"loss": 0.7348,
"step": 154
},
{
"epoch": 1.0197802197802197,
"grad_norm": 0.3237416744232178,
"learning_rate": 0.00013392461197339248,
"loss": 0.8655,
"step": 155
},
{
"epoch": 1.0263736263736263,
"grad_norm": 0.29400381445884705,
"learning_rate": 0.00013348115299334812,
"loss": 1.0433,
"step": 156
},
{
"epoch": 1.032967032967033,
"grad_norm": 0.31784510612487793,
"learning_rate": 0.00013303769401330377,
"loss": 0.9537,
"step": 157
},
{
"epoch": 1.0395604395604396,
"grad_norm": 0.29349464178085327,
"learning_rate": 0.00013259423503325944,
"loss": 0.8429,
"step": 158
},
{
"epoch": 1.0461538461538462,
"grad_norm": 0.28283601999282837,
"learning_rate": 0.00013215077605321509,
"loss": 0.7495,
"step": 159
},
{
"epoch": 1.0527472527472528,
"grad_norm": 0.3360508680343628,
"learning_rate": 0.00013170731707317076,
"loss": 0.8179,
"step": 160
},
{
"epoch": 1.0593406593406594,
"grad_norm": 0.3243735432624817,
"learning_rate": 0.0001312638580931264,
"loss": 0.8491,
"step": 161
},
{
"epoch": 1.065934065934066,
"grad_norm": 0.2543736696243286,
"learning_rate": 0.00013082039911308205,
"loss": 0.8473,
"step": 162
},
{
"epoch": 1.0725274725274725,
"grad_norm": 0.41662824153900146,
"learning_rate": 0.0001303769401330377,
"loss": 1.0799,
"step": 163
},
{
"epoch": 1.079120879120879,
"grad_norm": 0.36721354722976685,
"learning_rate": 0.00012993348115299336,
"loss": 0.9888,
"step": 164
},
{
"epoch": 1.0857142857142856,
"grad_norm": 0.38682231307029724,
"learning_rate": 0.000129490022172949,
"loss": 0.8444,
"step": 165
},
{
"epoch": 1.0923076923076924,
"grad_norm": 0.3763047754764557,
"learning_rate": 0.00012904656319290468,
"loss": 0.7915,
"step": 166
},
{
"epoch": 1.098901098901099,
"grad_norm": 0.3762548565864563,
"learning_rate": 0.00012860310421286032,
"loss": 0.8605,
"step": 167
},
{
"epoch": 1.1054945054945056,
"grad_norm": 0.34774428606033325,
"learning_rate": 0.00012815964523281597,
"loss": 0.8596,
"step": 168
},
{
"epoch": 1.1120879120879121,
"grad_norm": 0.33537760376930237,
"learning_rate": 0.0001277161862527716,
"loss": 1.0133,
"step": 169
},
{
"epoch": 1.1186813186813187,
"grad_norm": 0.37760505080223083,
"learning_rate": 0.00012727272727272728,
"loss": 0.8969,
"step": 170
},
{
"epoch": 1.1252747252747253,
"grad_norm": 0.40093833208084106,
"learning_rate": 0.00012682926829268293,
"loss": 0.7781,
"step": 171
},
{
"epoch": 1.1318681318681318,
"grad_norm": 0.37234511971473694,
"learning_rate": 0.0001263858093126386,
"loss": 0.8151,
"step": 172
},
{
"epoch": 1.1384615384615384,
"grad_norm": 0.383432000875473,
"learning_rate": 0.00012594235033259424,
"loss": 0.8534,
"step": 173
},
{
"epoch": 1.145054945054945,
"grad_norm": 0.4024796783924103,
"learning_rate": 0.00012549889135254989,
"loss": 0.8456,
"step": 174
},
{
"epoch": 1.1516483516483516,
"grad_norm": 0.42926251888275146,
"learning_rate": 0.00012505543237250553,
"loss": 0.7341,
"step": 175
},
{
"epoch": 1.1582417582417581,
"grad_norm": 0.38404402136802673,
"learning_rate": 0.0001246119733924612,
"loss": 0.9299,
"step": 176
},
{
"epoch": 1.164835164835165,
"grad_norm": 0.3822978436946869,
"learning_rate": 0.00012416851441241685,
"loss": 0.8499,
"step": 177
},
{
"epoch": 1.1714285714285715,
"grad_norm": 0.3729138672351837,
"learning_rate": 0.00012372505543237252,
"loss": 0.8227,
"step": 178
},
{
"epoch": 1.178021978021978,
"grad_norm": 0.3332406282424927,
"learning_rate": 0.00012328159645232816,
"loss": 0.8247,
"step": 179
},
{
"epoch": 1.1846153846153846,
"grad_norm": 0.38172462582588196,
"learning_rate": 0.0001228381374722838,
"loss": 0.9185,
"step": 180
},
{
"epoch": 1.1912087912087912,
"grad_norm": 0.40163788199424744,
"learning_rate": 0.00012239467849223948,
"loss": 0.8539,
"step": 181
},
{
"epoch": 1.1978021978021978,
"grad_norm": 0.36429041624069214,
"learning_rate": 0.00012195121951219512,
"loss": 0.9113,
"step": 182
},
{
"epoch": 1.2043956043956043,
"grad_norm": 0.413580983877182,
"learning_rate": 0.00012150776053215078,
"loss": 0.819,
"step": 183
},
{
"epoch": 1.210989010989011,
"grad_norm": 0.4143903851509094,
"learning_rate": 0.00012106430155210644,
"loss": 0.8512,
"step": 184
},
{
"epoch": 1.2175824175824177,
"grad_norm": 0.4125838875770569,
"learning_rate": 0.0001206208425720621,
"loss": 0.9585,
"step": 185
},
{
"epoch": 1.2241758241758243,
"grad_norm": 0.3720792233943939,
"learning_rate": 0.00012017738359201774,
"loss": 0.7881,
"step": 186
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.49924013018608093,
"learning_rate": 0.0001197339246119734,
"loss": 0.9763,
"step": 187
},
{
"epoch": 1.2373626373626374,
"grad_norm": 0.39625343680381775,
"learning_rate": 0.00011929046563192906,
"loss": 0.7851,
"step": 188
},
{
"epoch": 1.243956043956044,
"grad_norm": 0.33806920051574707,
"learning_rate": 0.00011884700665188471,
"loss": 0.8916,
"step": 189
},
{
"epoch": 1.2505494505494505,
"grad_norm": 0.43014171719551086,
"learning_rate": 0.00011840354767184036,
"loss": 0.7832,
"step": 190
},
{
"epoch": 1.2571428571428571,
"grad_norm": 0.37463000416755676,
"learning_rate": 0.00011796008869179602,
"loss": 0.9113,
"step": 191
},
{
"epoch": 1.2637362637362637,
"grad_norm": 0.34547415375709534,
"learning_rate": 0.00011751662971175166,
"loss": 0.9936,
"step": 192
},
{
"epoch": 1.2703296703296703,
"grad_norm": 0.4263266324996948,
"learning_rate": 0.00011707317073170732,
"loss": 0.8082,
"step": 193
},
{
"epoch": 1.2769230769230768,
"grad_norm": 0.4370916783809662,
"learning_rate": 0.00011662971175166298,
"loss": 0.7088,
"step": 194
},
{
"epoch": 1.2835164835164834,
"grad_norm": 0.48510512709617615,
"learning_rate": 0.00011618625277161863,
"loss": 0.873,
"step": 195
},
{
"epoch": 1.2901098901098902,
"grad_norm": 0.43364349007606506,
"learning_rate": 0.00011574279379157429,
"loss": 0.9497,
"step": 196
},
{
"epoch": 1.2967032967032968,
"grad_norm": 0.4251789450645447,
"learning_rate": 0.00011529933481152995,
"loss": 0.8802,
"step": 197
},
{
"epoch": 1.3032967032967033,
"grad_norm": 0.42555809020996094,
"learning_rate": 0.00011485587583148558,
"loss": 0.8027,
"step": 198
},
{
"epoch": 1.30989010989011,
"grad_norm": 0.4067121148109436,
"learning_rate": 0.00011441241685144124,
"loss": 0.8642,
"step": 199
},
{
"epoch": 1.3164835164835165,
"grad_norm": 0.42258256673812866,
"learning_rate": 0.0001139689578713969,
"loss": 0.9318,
"step": 200
},
{
"epoch": 1.323076923076923,
"grad_norm": 0.4275679886341095,
"learning_rate": 0.00011352549889135255,
"loss": 0.8968,
"step": 201
},
{
"epoch": 1.3296703296703296,
"grad_norm": 0.49644041061401367,
"learning_rate": 0.00011308203991130821,
"loss": 0.7946,
"step": 202
},
{
"epoch": 1.3362637362637364,
"grad_norm": 0.4212213456630707,
"learning_rate": 0.00011263858093126387,
"loss": 0.9027,
"step": 203
},
{
"epoch": 1.342857142857143,
"grad_norm": 0.3859935402870178,
"learning_rate": 0.00011219512195121953,
"loss": 0.7549,
"step": 204
},
{
"epoch": 1.3494505494505495,
"grad_norm": 0.42257729172706604,
"learning_rate": 0.00011175166297117516,
"loss": 0.9584,
"step": 205
},
{
"epoch": 1.3560439560439561,
"grad_norm": 0.4249301850795746,
"learning_rate": 0.00011130820399113082,
"loss": 0.977,
"step": 206
},
{
"epoch": 1.3626373626373627,
"grad_norm": 0.4894404709339142,
"learning_rate": 0.00011086474501108647,
"loss": 1.0573,
"step": 207
},
{
"epoch": 1.3692307692307693,
"grad_norm": 0.44024428725242615,
"learning_rate": 0.00011042128603104213,
"loss": 0.938,
"step": 208
},
{
"epoch": 1.3758241758241758,
"grad_norm": 0.568278431892395,
"learning_rate": 0.00010997782705099779,
"loss": 0.7926,
"step": 209
},
{
"epoch": 1.3824175824175824,
"grad_norm": 0.45056864619255066,
"learning_rate": 0.00010953436807095345,
"loss": 0.6928,
"step": 210
},
{
"epoch": 1.389010989010989,
"grad_norm": 0.5637768507003784,
"learning_rate": 0.00010909090909090909,
"loss": 0.9798,
"step": 211
},
{
"epoch": 1.3956043956043955,
"grad_norm": 0.48961734771728516,
"learning_rate": 0.00010864745011086475,
"loss": 0.8129,
"step": 212
},
{
"epoch": 1.402197802197802,
"grad_norm": 0.39597514271736145,
"learning_rate": 0.00010820399113082041,
"loss": 0.6318,
"step": 213
},
{
"epoch": 1.4087912087912087,
"grad_norm": 0.4896124005317688,
"learning_rate": 0.00010776053215077607,
"loss": 0.8,
"step": 214
},
{
"epoch": 1.4153846153846155,
"grad_norm": 0.409181147813797,
"learning_rate": 0.00010731707317073172,
"loss": 0.742,
"step": 215
},
{
"epoch": 1.421978021978022,
"grad_norm": 0.5370419025421143,
"learning_rate": 0.00010687361419068738,
"loss": 0.7985,
"step": 216
},
{
"epoch": 1.4285714285714286,
"grad_norm": 0.5486623644828796,
"learning_rate": 0.00010643015521064301,
"loss": 0.844,
"step": 217
},
{
"epoch": 1.4351648351648352,
"grad_norm": 0.4042065143585205,
"learning_rate": 0.00010598669623059867,
"loss": 0.8765,
"step": 218
},
{
"epoch": 1.4417582417582417,
"grad_norm": 0.4377235770225525,
"learning_rate": 0.00010554323725055433,
"loss": 0.8348,
"step": 219
},
{
"epoch": 1.4483516483516483,
"grad_norm": 0.4218975305557251,
"learning_rate": 0.00010509977827050999,
"loss": 0.8182,
"step": 220
},
{
"epoch": 1.4549450549450549,
"grad_norm": 0.4103962779045105,
"learning_rate": 0.00010465631929046564,
"loss": 0.817,
"step": 221
},
{
"epoch": 1.4615384615384617,
"grad_norm": 0.5105700492858887,
"learning_rate": 0.0001042128603104213,
"loss": 0.9333,
"step": 222
},
{
"epoch": 1.4681318681318682,
"grad_norm": 0.3896787762641907,
"learning_rate": 0.00010376940133037693,
"loss": 0.8385,
"step": 223
},
{
"epoch": 1.4747252747252748,
"grad_norm": 0.5424078106880188,
"learning_rate": 0.00010332594235033259,
"loss": 0.8077,
"step": 224
},
{
"epoch": 1.4813186813186814,
"grad_norm": 0.505621075630188,
"learning_rate": 0.00010288248337028825,
"loss": 0.8905,
"step": 225
},
{
"epoch": 1.487912087912088,
"grad_norm": 0.3873915672302246,
"learning_rate": 0.0001024390243902439,
"loss": 0.675,
"step": 226
},
{
"epoch": 1.4945054945054945,
"grad_norm": 0.4680170714855194,
"learning_rate": 0.00010199556541019956,
"loss": 0.6547,
"step": 227
},
{
"epoch": 1.501098901098901,
"grad_norm": 0.4344763159751892,
"learning_rate": 0.00010155210643015522,
"loss": 0.9168,
"step": 228
},
{
"epoch": 1.5076923076923077,
"grad_norm": 0.5052294731140137,
"learning_rate": 0.00010110864745011087,
"loss": 0.8161,
"step": 229
},
{
"epoch": 1.5142857142857142,
"grad_norm": 0.46243816614151,
"learning_rate": 0.00010066518847006652,
"loss": 0.9792,
"step": 230
},
{
"epoch": 1.5208791208791208,
"grad_norm": 0.4567134976387024,
"learning_rate": 0.00010022172949002218,
"loss": 0.7665,
"step": 231
},
{
"epoch": 1.5274725274725274,
"grad_norm": 0.4654875099658966,
"learning_rate": 9.977827050997783e-05,
"loss": 0.9581,
"step": 232
},
{
"epoch": 1.534065934065934,
"grad_norm": 0.4650971591472626,
"learning_rate": 9.933481152993348e-05,
"loss": 0.7643,
"step": 233
},
{
"epoch": 1.5406593406593405,
"grad_norm": 0.590111255645752,
"learning_rate": 9.889135254988914e-05,
"loss": 1.0478,
"step": 234
},
{
"epoch": 1.5472527472527473,
"grad_norm": 0.521954357624054,
"learning_rate": 9.84478935698448e-05,
"loss": 0.884,
"step": 235
},
{
"epoch": 1.5538461538461539,
"grad_norm": 0.48372364044189453,
"learning_rate": 9.800443458980046e-05,
"loss": 0.8072,
"step": 236
},
{
"epoch": 1.5604395604395604,
"grad_norm": 0.5271714925765991,
"learning_rate": 9.75609756097561e-05,
"loss": 0.7694,
"step": 237
},
{
"epoch": 1.567032967032967,
"grad_norm": 0.5260690450668335,
"learning_rate": 9.711751662971176e-05,
"loss": 0.9671,
"step": 238
},
{
"epoch": 1.5736263736263736,
"grad_norm": 0.3899517357349396,
"learning_rate": 9.667405764966742e-05,
"loss": 0.7383,
"step": 239
},
{
"epoch": 1.5802197802197804,
"grad_norm": 0.4058227241039276,
"learning_rate": 9.623059866962306e-05,
"loss": 0.7223,
"step": 240
},
{
"epoch": 1.586813186813187,
"grad_norm": 0.5025250315666199,
"learning_rate": 9.578713968957872e-05,
"loss": 0.7403,
"step": 241
},
{
"epoch": 1.5934065934065935,
"grad_norm": 0.4921959936618805,
"learning_rate": 9.534368070953438e-05,
"loss": 0.8244,
"step": 242
},
{
"epoch": 1.6,
"grad_norm": 0.4537874460220337,
"learning_rate": 9.490022172949002e-05,
"loss": 0.7974,
"step": 243
},
{
"epoch": 1.6065934065934067,
"grad_norm": 0.5014695525169373,
"learning_rate": 9.445676274944568e-05,
"loss": 0.8144,
"step": 244
},
{
"epoch": 1.6131868131868132,
"grad_norm": 0.4026423692703247,
"learning_rate": 9.401330376940134e-05,
"loss": 0.7425,
"step": 245
},
{
"epoch": 1.6197802197802198,
"grad_norm": 0.3835386037826538,
"learning_rate": 9.356984478935698e-05,
"loss": 0.8068,
"step": 246
},
{
"epoch": 1.6263736263736264,
"grad_norm": 0.47940903902053833,
"learning_rate": 9.312638580931264e-05,
"loss": 0.8321,
"step": 247
},
{
"epoch": 1.632967032967033,
"grad_norm": 0.49348318576812744,
"learning_rate": 9.26829268292683e-05,
"loss": 0.9327,
"step": 248
},
{
"epoch": 1.6395604395604395,
"grad_norm": 0.4817999601364136,
"learning_rate": 9.223946784922394e-05,
"loss": 0.8583,
"step": 249
},
{
"epoch": 1.646153846153846,
"grad_norm": 0.5370239615440369,
"learning_rate": 9.17960088691796e-05,
"loss": 0.9774,
"step": 250
},
{
"epoch": 1.6527472527472526,
"grad_norm": 0.5423967242240906,
"learning_rate": 9.135254988913526e-05,
"loss": 0.767,
"step": 251
},
{
"epoch": 1.6593406593406592,
"grad_norm": 0.4797903299331665,
"learning_rate": 9.090909090909092e-05,
"loss": 0.7052,
"step": 252
},
{
"epoch": 1.6659340659340658,
"grad_norm": 0.5578131675720215,
"learning_rate": 9.046563192904656e-05,
"loss": 0.7422,
"step": 253
},
{
"epoch": 1.6725274725274726,
"grad_norm": 0.5849676132202148,
"learning_rate": 9.002217294900222e-05,
"loss": 0.7496,
"step": 254
},
{
"epoch": 1.6791208791208792,
"grad_norm": 0.598434567451477,
"learning_rate": 8.957871396895788e-05,
"loss": 0.8455,
"step": 255
},
{
"epoch": 1.6857142857142857,
"grad_norm": 0.5673317313194275,
"learning_rate": 8.913525498891354e-05,
"loss": 0.7943,
"step": 256
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.5396888256072998,
"learning_rate": 8.869179600886919e-05,
"loss": 1.0394,
"step": 257
},
{
"epoch": 1.6989010989010989,
"grad_norm": 0.4502924084663391,
"learning_rate": 8.824833702882484e-05,
"loss": 0.7011,
"step": 258
},
{
"epoch": 1.7054945054945057,
"grad_norm": 0.43324655294418335,
"learning_rate": 8.78048780487805e-05,
"loss": 0.762,
"step": 259
},
{
"epoch": 1.7120879120879122,
"grad_norm": 0.5309177041053772,
"learning_rate": 8.736141906873615e-05,
"loss": 0.7526,
"step": 260
},
{
"epoch": 1.7186813186813188,
"grad_norm": 0.486122727394104,
"learning_rate": 8.69179600886918e-05,
"loss": 0.8821,
"step": 261
},
{
"epoch": 1.7252747252747254,
"grad_norm": 0.5427271723747253,
"learning_rate": 8.647450110864746e-05,
"loss": 0.7716,
"step": 262
},
{
"epoch": 1.731868131868132,
"grad_norm": 0.552329957485199,
"learning_rate": 8.603104212860311e-05,
"loss": 0.9007,
"step": 263
},
{
"epoch": 1.7384615384615385,
"grad_norm": 0.5551875829696655,
"learning_rate": 8.558758314855876e-05,
"loss": 0.8845,
"step": 264
},
{
"epoch": 1.745054945054945,
"grad_norm": 0.5471997261047363,
"learning_rate": 8.514412416851442e-05,
"loss": 0.9556,
"step": 265
},
{
"epoch": 1.7516483516483516,
"grad_norm": 0.5877672433853149,
"learning_rate": 8.470066518847007e-05,
"loss": 0.9253,
"step": 266
},
{
"epoch": 1.7582417582417582,
"grad_norm": 0.5753462910652161,
"learning_rate": 8.425720620842572e-05,
"loss": 0.865,
"step": 267
},
{
"epoch": 1.7648351648351648,
"grad_norm": 0.5661343336105347,
"learning_rate": 8.381374722838138e-05,
"loss": 1.1195,
"step": 268
},
{
"epoch": 1.7714285714285714,
"grad_norm": 0.48300594091415405,
"learning_rate": 8.337028824833703e-05,
"loss": 0.8047,
"step": 269
},
{
"epoch": 1.778021978021978,
"grad_norm": 0.5009127855300903,
"learning_rate": 8.292682926829268e-05,
"loss": 0.8013,
"step": 270
},
{
"epoch": 1.7846153846153845,
"grad_norm": 0.448288232088089,
"learning_rate": 8.248337028824834e-05,
"loss": 0.628,
"step": 271
},
{
"epoch": 1.791208791208791,
"grad_norm": 0.4730840027332306,
"learning_rate": 8.2039911308204e-05,
"loss": 0.804,
"step": 272
},
{
"epoch": 1.7978021978021979,
"grad_norm": 0.4517776668071747,
"learning_rate": 8.159645232815965e-05,
"loss": 0.8543,
"step": 273
},
{
"epoch": 1.8043956043956044,
"grad_norm": 0.5334697961807251,
"learning_rate": 8.11529933481153e-05,
"loss": 0.7027,
"step": 274
},
{
"epoch": 1.810989010989011,
"grad_norm": 0.5420060157775879,
"learning_rate": 8.070953436807095e-05,
"loss": 0.885,
"step": 275
},
{
"epoch": 1.8175824175824176,
"grad_norm": 0.5378316044807434,
"learning_rate": 8.026607538802661e-05,
"loss": 0.7681,
"step": 276
},
{
"epoch": 1.8241758241758241,
"grad_norm": 0.5504192113876343,
"learning_rate": 7.982261640798227e-05,
"loss": 1.0043,
"step": 277
},
{
"epoch": 1.830769230769231,
"grad_norm": 0.5274063348770142,
"learning_rate": 7.937915742793793e-05,
"loss": 1.0655,
"step": 278
},
{
"epoch": 1.8373626373626375,
"grad_norm": 0.49303489923477173,
"learning_rate": 7.893569844789357e-05,
"loss": 0.8298,
"step": 279
},
{
"epoch": 1.843956043956044,
"grad_norm": 0.5413931608200073,
"learning_rate": 7.849223946784923e-05,
"loss": 0.7224,
"step": 280
},
{
"epoch": 1.8505494505494506,
"grad_norm": 0.5850419402122498,
"learning_rate": 7.804878048780489e-05,
"loss": 0.8905,
"step": 281
},
{
"epoch": 1.8571428571428572,
"grad_norm": 0.5188719630241394,
"learning_rate": 7.760532150776053e-05,
"loss": 0.894,
"step": 282
},
{
"epoch": 1.8637362637362638,
"grad_norm": 0.5849002599716187,
"learning_rate": 7.716186252771619e-05,
"loss": 0.8756,
"step": 283
},
{
"epoch": 1.8703296703296703,
"grad_norm": 0.5260016918182373,
"learning_rate": 7.671840354767185e-05,
"loss": 0.8052,
"step": 284
},
{
"epoch": 1.876923076923077,
"grad_norm": 0.5511124730110168,
"learning_rate": 7.627494456762749e-05,
"loss": 0.8885,
"step": 285
},
{
"epoch": 1.8835164835164835,
"grad_norm": 0.4948745667934418,
"learning_rate": 7.583148558758315e-05,
"loss": 0.7462,
"step": 286
},
{
"epoch": 1.89010989010989,
"grad_norm": 0.4871424436569214,
"learning_rate": 7.538802660753881e-05,
"loss": 0.8983,
"step": 287
},
{
"epoch": 1.8967032967032966,
"grad_norm": 0.4769727289676666,
"learning_rate": 7.494456762749445e-05,
"loss": 0.9099,
"step": 288
},
{
"epoch": 1.9032967032967032,
"grad_norm": 0.5380547642707825,
"learning_rate": 7.450110864745011e-05,
"loss": 0.7948,
"step": 289
},
{
"epoch": 1.9098901098901098,
"grad_norm": 0.608492910861969,
"learning_rate": 7.405764966740577e-05,
"loss": 0.8169,
"step": 290
},
{
"epoch": 1.9164835164835163,
"grad_norm": 0.4403230547904968,
"learning_rate": 7.361419068736141e-05,
"loss": 0.7714,
"step": 291
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.5714691877365112,
"learning_rate": 7.317073170731707e-05,
"loss": 0.8704,
"step": 292
},
{
"epoch": 1.9296703296703297,
"grad_norm": 0.43401604890823364,
"learning_rate": 7.272727272727273e-05,
"loss": 0.8245,
"step": 293
},
{
"epoch": 1.9362637362637363,
"grad_norm": 0.5744357109069824,
"learning_rate": 7.228381374722837e-05,
"loss": 0.9585,
"step": 294
},
{
"epoch": 1.9428571428571428,
"grad_norm": 0.48480620980262756,
"learning_rate": 7.184035476718403e-05,
"loss": 0.848,
"step": 295
},
{
"epoch": 1.9494505494505494,
"grad_norm": 0.5181294083595276,
"learning_rate": 7.139689578713969e-05,
"loss": 0.7215,
"step": 296
},
{
"epoch": 1.9560439560439562,
"grad_norm": 0.4262259602546692,
"learning_rate": 7.095343680709535e-05,
"loss": 0.6381,
"step": 297
},
{
"epoch": 1.9626373626373628,
"grad_norm": 0.5429973006248474,
"learning_rate": 7.0509977827051e-05,
"loss": 0.7133,
"step": 298
},
{
"epoch": 1.9692307692307693,
"grad_norm": 0.5489481091499329,
"learning_rate": 7.006651884700666e-05,
"loss": 0.8847,
"step": 299
},
{
"epoch": 1.975824175824176,
"grad_norm": 0.49627819657325745,
"learning_rate": 6.962305986696232e-05,
"loss": 0.7998,
"step": 300
},
{
"epoch": 1.9824175824175825,
"grad_norm": 0.5082553625106812,
"learning_rate": 6.917960088691796e-05,
"loss": 0.7329,
"step": 301
},
{
"epoch": 1.989010989010989,
"grad_norm": 0.6370390057563782,
"learning_rate": 6.873614190687362e-05,
"loss": 0.8852,
"step": 302
},
{
"epoch": 1.9956043956043956,
"grad_norm": 0.5251163244247437,
"learning_rate": 6.829268292682928e-05,
"loss": 0.7608,
"step": 303
},
{
"epoch": 2.0,
"grad_norm": 0.8000172972679138,
"learning_rate": 6.784922394678492e-05,
"loss": 0.5656,
"step": 304
},
{
"epoch": 2.0065934065934066,
"grad_norm": 0.4724150002002716,
"learning_rate": 6.740576496674058e-05,
"loss": 0.7345,
"step": 305
},
{
"epoch": 2.013186813186813,
"grad_norm": 0.5429744720458984,
"learning_rate": 6.696230598669624e-05,
"loss": 0.8151,
"step": 306
},
{
"epoch": 2.0197802197802197,
"grad_norm": 0.5398493409156799,
"learning_rate": 6.651884700665188e-05,
"loss": 0.6771,
"step": 307
},
{
"epoch": 2.0263736263736263,
"grad_norm": 0.42921003699302673,
"learning_rate": 6.607538802660754e-05,
"loss": 0.6704,
"step": 308
},
{
"epoch": 2.032967032967033,
"grad_norm": 0.5017111301422119,
"learning_rate": 6.56319290465632e-05,
"loss": 0.6689,
"step": 309
},
{
"epoch": 2.0395604395604394,
"grad_norm": 0.5576778054237366,
"learning_rate": 6.518847006651884e-05,
"loss": 0.6943,
"step": 310
},
{
"epoch": 2.046153846153846,
"grad_norm": 0.5303216576576233,
"learning_rate": 6.47450110864745e-05,
"loss": 0.643,
"step": 311
},
{
"epoch": 2.0527472527472526,
"grad_norm": 0.6024826765060425,
"learning_rate": 6.430155210643016e-05,
"loss": 0.6094,
"step": 312
},
{
"epoch": 2.059340659340659,
"grad_norm": 0.5809922814369202,
"learning_rate": 6.38580931263858e-05,
"loss": 0.6358,
"step": 313
},
{
"epoch": 2.065934065934066,
"grad_norm": 0.5435502529144287,
"learning_rate": 6.341463414634146e-05,
"loss": 0.67,
"step": 314
},
{
"epoch": 2.0725274725274727,
"grad_norm": 0.5232275724411011,
"learning_rate": 6.297117516629712e-05,
"loss": 0.9177,
"step": 315
},
{
"epoch": 2.0791208791208793,
"grad_norm": 0.6438103318214417,
"learning_rate": 6.252771618625277e-05,
"loss": 0.8123,
"step": 316
},
{
"epoch": 2.085714285714286,
"grad_norm": 0.5559529662132263,
"learning_rate": 6.208425720620842e-05,
"loss": 0.6895,
"step": 317
},
{
"epoch": 2.0923076923076924,
"grad_norm": 0.6401575803756714,
"learning_rate": 6.164079822616408e-05,
"loss": 0.8364,
"step": 318
},
{
"epoch": 2.098901098901099,
"grad_norm": 0.6006264090538025,
"learning_rate": 6.119733924611974e-05,
"loss": 0.7388,
"step": 319
},
{
"epoch": 2.1054945054945056,
"grad_norm": 0.6150893568992615,
"learning_rate": 6.075388026607539e-05,
"loss": 0.8096,
"step": 320
},
{
"epoch": 2.112087912087912,
"grad_norm": 0.5198052525520325,
"learning_rate": 6.031042128603105e-05,
"loss": 0.8271,
"step": 321
},
{
"epoch": 2.1186813186813187,
"grad_norm": 0.6085712909698486,
"learning_rate": 5.98669623059867e-05,
"loss": 0.7654,
"step": 322
},
{
"epoch": 2.1252747252747253,
"grad_norm": 0.6414448022842407,
"learning_rate": 5.942350332594236e-05,
"loss": 0.7289,
"step": 323
},
{
"epoch": 2.131868131868132,
"grad_norm": 0.5917273759841919,
"learning_rate": 5.898004434589801e-05,
"loss": 0.6184,
"step": 324
},
{
"epoch": 2.1384615384615384,
"grad_norm": 0.7415244579315186,
"learning_rate": 5.853658536585366e-05,
"loss": 0.7319,
"step": 325
},
{
"epoch": 2.145054945054945,
"grad_norm": 0.6198651194572449,
"learning_rate": 5.809312638580932e-05,
"loss": 0.7715,
"step": 326
},
{
"epoch": 2.1516483516483516,
"grad_norm": 0.674970269203186,
"learning_rate": 5.7649667405764975e-05,
"loss": 0.7564,
"step": 327
},
{
"epoch": 2.158241758241758,
"grad_norm": 0.565639853477478,
"learning_rate": 5.720620842572062e-05,
"loss": 0.6232,
"step": 328
},
{
"epoch": 2.1648351648351647,
"grad_norm": 0.6033426523208618,
"learning_rate": 5.676274944567628e-05,
"loss": 0.7566,
"step": 329
},
{
"epoch": 2.1714285714285713,
"grad_norm": 0.6068849563598633,
"learning_rate": 5.6319290465631935e-05,
"loss": 0.6902,
"step": 330
},
{
"epoch": 2.178021978021978,
"grad_norm": 0.6270563006401062,
"learning_rate": 5.587583148558758e-05,
"loss": 0.8026,
"step": 331
},
{
"epoch": 2.184615384615385,
"grad_norm": 0.5736610889434814,
"learning_rate": 5.543237250554324e-05,
"loss": 0.5957,
"step": 332
},
{
"epoch": 2.1912087912087914,
"grad_norm": 0.6985477209091187,
"learning_rate": 5.4988913525498895e-05,
"loss": 0.7766,
"step": 333
},
{
"epoch": 2.197802197802198,
"grad_norm": 0.7159436345100403,
"learning_rate": 5.4545454545454546e-05,
"loss": 0.7003,
"step": 334
},
{
"epoch": 2.2043956043956046,
"grad_norm": 0.5028312802314758,
"learning_rate": 5.4101995565410204e-05,
"loss": 0.5197,
"step": 335
},
{
"epoch": 2.210989010989011,
"grad_norm": 0.6573224067687988,
"learning_rate": 5.365853658536586e-05,
"loss": 0.6585,
"step": 336
},
{
"epoch": 2.2175824175824177,
"grad_norm": 0.5629363059997559,
"learning_rate": 5.3215077605321506e-05,
"loss": 0.7434,
"step": 337
},
{
"epoch": 2.2241758241758243,
"grad_norm": 0.6618596911430359,
"learning_rate": 5.2771618625277164e-05,
"loss": 0.6888,
"step": 338
},
{
"epoch": 2.230769230769231,
"grad_norm": 0.5068679451942444,
"learning_rate": 5.232815964523282e-05,
"loss": 0.6502,
"step": 339
},
{
"epoch": 2.2373626373626374,
"grad_norm": 0.6307594776153564,
"learning_rate": 5.1884700665188466e-05,
"loss": 0.7559,
"step": 340
},
{
"epoch": 2.243956043956044,
"grad_norm": 0.6525044441223145,
"learning_rate": 5.1441241685144124e-05,
"loss": 0.8478,
"step": 341
},
{
"epoch": 2.2505494505494505,
"grad_norm": 0.6841545701026917,
"learning_rate": 5.099778270509978e-05,
"loss": 0.6478,
"step": 342
},
{
"epoch": 2.257142857142857,
"grad_norm": 0.7310889363288879,
"learning_rate": 5.0554323725055433e-05,
"loss": 0.8269,
"step": 343
},
{
"epoch": 2.2637362637362637,
"grad_norm": 0.6633881330490112,
"learning_rate": 5.011086474501109e-05,
"loss": 0.7219,
"step": 344
},
{
"epoch": 2.2703296703296703,
"grad_norm": 0.718220055103302,
"learning_rate": 4.966740576496674e-05,
"loss": 0.8397,
"step": 345
}
],
"logging_steps": 1,
"max_steps": 456,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 15,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 5.603461697716224e+16,
"train_batch_size": 22,
"trial_name": null,
"trial_params": null
}