EtashGuha's picture
Upload folder using huggingface_hub
f56ce1c verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9904153354632586,
"eval_steps": 500,
"global_step": 312,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.009584664536741214,
"grad_norm": 7.631757736206055,
"learning_rate": 3.125e-07,
"loss": 1.1437,
"step": 1
},
{
"epoch": 0.019169329073482427,
"grad_norm": 7.207672119140625,
"learning_rate": 6.25e-07,
"loss": 1.0804,
"step": 2
},
{
"epoch": 0.02875399361022364,
"grad_norm": 8.071743965148926,
"learning_rate": 9.375000000000001e-07,
"loss": 1.155,
"step": 3
},
{
"epoch": 0.038338658146964855,
"grad_norm": 7.237469673156738,
"learning_rate": 1.25e-06,
"loss": 1.0944,
"step": 4
},
{
"epoch": 0.04792332268370607,
"grad_norm": 7.192005157470703,
"learning_rate": 1.5625e-06,
"loss": 1.1085,
"step": 5
},
{
"epoch": 0.05750798722044728,
"grad_norm": 7.114604473114014,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.099,
"step": 6
},
{
"epoch": 0.0670926517571885,
"grad_norm": 5.225025177001953,
"learning_rate": 2.1875000000000002e-06,
"loss": 1.0407,
"step": 7
},
{
"epoch": 0.07667731629392971,
"grad_norm": 4.8877787590026855,
"learning_rate": 2.5e-06,
"loss": 1.0256,
"step": 8
},
{
"epoch": 0.08626198083067092,
"grad_norm": 3.1853082180023193,
"learning_rate": 2.8125e-06,
"loss": 0.9822,
"step": 9
},
{
"epoch": 0.09584664536741214,
"grad_norm": 2.9565036296844482,
"learning_rate": 3.125e-06,
"loss": 1.0148,
"step": 10
},
{
"epoch": 0.10543130990415335,
"grad_norm": 2.6891329288482666,
"learning_rate": 3.4375e-06,
"loss": 0.9717,
"step": 11
},
{
"epoch": 0.11501597444089456,
"grad_norm": 3.217716693878174,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.9585,
"step": 12
},
{
"epoch": 0.12460063897763578,
"grad_norm": 3.6357855796813965,
"learning_rate": 4.0625000000000005e-06,
"loss": 0.9466,
"step": 13
},
{
"epoch": 0.134185303514377,
"grad_norm": 3.2886316776275635,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.9149,
"step": 14
},
{
"epoch": 0.14376996805111822,
"grad_norm": 2.728691339492798,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.9003,
"step": 15
},
{
"epoch": 0.15335463258785942,
"grad_norm": 2.0728046894073486,
"learning_rate": 5e-06,
"loss": 0.8858,
"step": 16
},
{
"epoch": 0.16293929712460065,
"grad_norm": 2.3851478099823,
"learning_rate": 5.3125e-06,
"loss": 0.8923,
"step": 17
},
{
"epoch": 0.17252396166134185,
"grad_norm": 2.1146421432495117,
"learning_rate": 5.625e-06,
"loss": 0.8444,
"step": 18
},
{
"epoch": 0.18210862619808307,
"grad_norm": 1.742587685585022,
"learning_rate": 5.9375e-06,
"loss": 0.846,
"step": 19
},
{
"epoch": 0.19169329073482427,
"grad_norm": 1.4841606616973877,
"learning_rate": 6.25e-06,
"loss": 0.8413,
"step": 20
},
{
"epoch": 0.2012779552715655,
"grad_norm": 1.2848784923553467,
"learning_rate": 6.5625e-06,
"loss": 0.7976,
"step": 21
},
{
"epoch": 0.2108626198083067,
"grad_norm": 1.739992380142212,
"learning_rate": 6.875e-06,
"loss": 0.7873,
"step": 22
},
{
"epoch": 0.22044728434504793,
"grad_norm": 1.633833408355713,
"learning_rate": 7.1875e-06,
"loss": 0.7909,
"step": 23
},
{
"epoch": 0.23003194888178913,
"grad_norm": 1.2238140106201172,
"learning_rate": 7.500000000000001e-06,
"loss": 0.789,
"step": 24
},
{
"epoch": 0.23961661341853036,
"grad_norm": 1.1153886318206787,
"learning_rate": 7.8125e-06,
"loss": 0.7688,
"step": 25
},
{
"epoch": 0.24920127795527156,
"grad_norm": 1.1999198198318481,
"learning_rate": 8.125000000000001e-06,
"loss": 0.7509,
"step": 26
},
{
"epoch": 0.25878594249201275,
"grad_norm": 1.0238111019134521,
"learning_rate": 8.4375e-06,
"loss": 0.7451,
"step": 27
},
{
"epoch": 0.268370607028754,
"grad_norm": 0.9762786626815796,
"learning_rate": 8.750000000000001e-06,
"loss": 0.7739,
"step": 28
},
{
"epoch": 0.2779552715654952,
"grad_norm": 0.931849479675293,
"learning_rate": 9.0625e-06,
"loss": 0.7406,
"step": 29
},
{
"epoch": 0.28753993610223644,
"grad_norm": 0.8621435165405273,
"learning_rate": 9.375000000000001e-06,
"loss": 0.7302,
"step": 30
},
{
"epoch": 0.2971246006389776,
"grad_norm": 0.9465786218643188,
"learning_rate": 9.6875e-06,
"loss": 0.7396,
"step": 31
},
{
"epoch": 0.30670926517571884,
"grad_norm": 0.9201045632362366,
"learning_rate": 1e-05,
"loss": 0.7593,
"step": 32
},
{
"epoch": 0.31629392971246006,
"grad_norm": 0.7977403402328491,
"learning_rate": 9.999685283773504e-06,
"loss": 0.7255,
"step": 33
},
{
"epoch": 0.3258785942492013,
"grad_norm": 0.8451075553894043,
"learning_rate": 9.998741174712534e-06,
"loss": 0.7172,
"step": 34
},
{
"epoch": 0.3354632587859425,
"grad_norm": 0.9500023126602173,
"learning_rate": 9.997167791667668e-06,
"loss": 0.7309,
"step": 35
},
{
"epoch": 0.3450479233226837,
"grad_norm": 0.8545159101486206,
"learning_rate": 9.994965332706574e-06,
"loss": 0.7316,
"step": 36
},
{
"epoch": 0.3546325878594249,
"grad_norm": 0.7496842741966248,
"learning_rate": 9.992134075089085e-06,
"loss": 0.7276,
"step": 37
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.9326169490814209,
"learning_rate": 9.98867437523228e-06,
"loss": 0.7029,
"step": 38
},
{
"epoch": 0.3738019169329074,
"grad_norm": 0.8307452201843262,
"learning_rate": 9.984586668665641e-06,
"loss": 0.7111,
"step": 39
},
{
"epoch": 0.38338658146964855,
"grad_norm": 0.8053420782089233,
"learning_rate": 9.979871469976197e-06,
"loss": 0.733,
"step": 40
},
{
"epoch": 0.3929712460063898,
"grad_norm": 0.81968092918396,
"learning_rate": 9.974529372743762e-06,
"loss": 0.7054,
"step": 41
},
{
"epoch": 0.402555910543131,
"grad_norm": 0.8351101875305176,
"learning_rate": 9.968561049466214e-06,
"loss": 0.6961,
"step": 42
},
{
"epoch": 0.41214057507987223,
"grad_norm": 0.8805105686187744,
"learning_rate": 9.961967251474823e-06,
"loss": 0.7093,
"step": 43
},
{
"epoch": 0.4217252396166134,
"grad_norm": 0.7041685581207275,
"learning_rate": 9.954748808839675e-06,
"loss": 0.6822,
"step": 44
},
{
"epoch": 0.43130990415335463,
"grad_norm": 0.7762527465820312,
"learning_rate": 9.946906630265184e-06,
"loss": 0.693,
"step": 45
},
{
"epoch": 0.44089456869009586,
"grad_norm": 0.7489679455757141,
"learning_rate": 9.938441702975689e-06,
"loss": 0.7289,
"step": 46
},
{
"epoch": 0.4504792332268371,
"grad_norm": 0.8490743041038513,
"learning_rate": 9.92935509259118e-06,
"loss": 0.6953,
"step": 47
},
{
"epoch": 0.46006389776357826,
"grad_norm": 0.7434765696525574,
"learning_rate": 9.91964794299315e-06,
"loss": 0.6735,
"step": 48
},
{
"epoch": 0.4696485623003195,
"grad_norm": 0.7000515460968018,
"learning_rate": 9.909321476180594e-06,
"loss": 0.6932,
"step": 49
},
{
"epoch": 0.4792332268370607,
"grad_norm": 0.891094446182251,
"learning_rate": 9.898376992116179e-06,
"loss": 0.6902,
"step": 50
},
{
"epoch": 0.48881789137380194,
"grad_norm": 0.711914598941803,
"learning_rate": 9.886815868562596e-06,
"loss": 0.6869,
"step": 51
},
{
"epoch": 0.4984025559105431,
"grad_norm": 0.6698287725448608,
"learning_rate": 9.874639560909118e-06,
"loss": 0.6825,
"step": 52
},
{
"epoch": 0.5079872204472844,
"grad_norm": 0.6114389300346375,
"learning_rate": 9.861849601988384e-06,
"loss": 0.6854,
"step": 53
},
{
"epoch": 0.5175718849840255,
"grad_norm": 0.7109593152999878,
"learning_rate": 9.848447601883436e-06,
"loss": 0.6904,
"step": 54
},
{
"epoch": 0.5271565495207667,
"grad_norm": 0.748586893081665,
"learning_rate": 9.834435247725032e-06,
"loss": 0.684,
"step": 55
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.7264263033866882,
"learning_rate": 9.819814303479268e-06,
"loss": 0.6821,
"step": 56
},
{
"epoch": 0.5463258785942492,
"grad_norm": 0.7709238529205322,
"learning_rate": 9.804586609725499e-06,
"loss": 0.7041,
"step": 57
},
{
"epoch": 0.5559105431309904,
"grad_norm": 0.7797627449035645,
"learning_rate": 9.788754083424654e-06,
"loss": 0.6591,
"step": 58
},
{
"epoch": 0.5654952076677316,
"grad_norm": 0.6770584583282471,
"learning_rate": 9.772318717677905e-06,
"loss": 0.668,
"step": 59
},
{
"epoch": 0.5750798722044729,
"grad_norm": 0.7253358364105225,
"learning_rate": 9.755282581475769e-06,
"loss": 0.6889,
"step": 60
},
{
"epoch": 0.5846645367412141,
"grad_norm": 0.8044716715812683,
"learning_rate": 9.737647819437645e-06,
"loss": 0.6824,
"step": 61
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.7814176082611084,
"learning_rate": 9.719416651541839e-06,
"loss": 0.6652,
"step": 62
},
{
"epoch": 0.6038338658146964,
"grad_norm": 0.8672887086868286,
"learning_rate": 9.700591372846096e-06,
"loss": 0.6681,
"step": 63
},
{
"epoch": 0.6134185303514377,
"grad_norm": 0.686853289604187,
"learning_rate": 9.681174353198687e-06,
"loss": 0.6741,
"step": 64
},
{
"epoch": 0.6230031948881789,
"grad_norm": 0.7145909667015076,
"learning_rate": 9.661168036940071e-06,
"loss": 0.6763,
"step": 65
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.7959132790565491,
"learning_rate": 9.640574942595195e-06,
"loss": 0.6863,
"step": 66
},
{
"epoch": 0.6421725239616614,
"grad_norm": 0.6892458200454712,
"learning_rate": 9.619397662556434e-06,
"loss": 0.6772,
"step": 67
},
{
"epoch": 0.6517571884984026,
"grad_norm": 0.7975493669509888,
"learning_rate": 9.597638862757255e-06,
"loss": 0.6726,
"step": 68
},
{
"epoch": 0.6613418530351438,
"grad_norm": 0.7013779282569885,
"learning_rate": 9.5753012823366e-06,
"loss": 0.6603,
"step": 69
},
{
"epoch": 0.670926517571885,
"grad_norm": 0.7355247139930725,
"learning_rate": 9.552387733294081e-06,
"loss": 0.6855,
"step": 70
},
{
"epoch": 0.6805111821086262,
"grad_norm": 0.7918930649757385,
"learning_rate": 9.528901100135971e-06,
"loss": 0.6717,
"step": 71
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.7762782573699951,
"learning_rate": 9.504844339512096e-06,
"loss": 0.6842,
"step": 72
},
{
"epoch": 0.6996805111821086,
"grad_norm": 0.7748196721076965,
"learning_rate": 9.480220479843627e-06,
"loss": 0.666,
"step": 73
},
{
"epoch": 0.7092651757188498,
"grad_norm": 0.7611200213432312,
"learning_rate": 9.45503262094184e-06,
"loss": 0.6856,
"step": 74
},
{
"epoch": 0.7188498402555911,
"grad_norm": 0.8682624697685242,
"learning_rate": 9.4292839336179e-06,
"loss": 0.6581,
"step": 75
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.7679495811462402,
"learning_rate": 9.40297765928369e-06,
"loss": 0.6734,
"step": 76
},
{
"epoch": 0.7380191693290735,
"grad_norm": 0.7918921709060669,
"learning_rate": 9.376117109543769e-06,
"loss": 0.6694,
"step": 77
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.9087303280830383,
"learning_rate": 9.348705665778479e-06,
"loss": 0.6755,
"step": 78
},
{
"epoch": 0.7571884984025559,
"grad_norm": 0.6955181360244751,
"learning_rate": 9.320746778718274e-06,
"loss": 0.6469,
"step": 79
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.7844282388687134,
"learning_rate": 9.292243968009332e-06,
"loss": 0.6723,
"step": 80
},
{
"epoch": 0.7763578274760383,
"grad_norm": 0.7338053584098816,
"learning_rate": 9.263200821770462e-06,
"loss": 0.6757,
"step": 81
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.7353216409683228,
"learning_rate": 9.233620996141421e-06,
"loss": 0.6727,
"step": 82
},
{
"epoch": 0.7955271565495208,
"grad_norm": 0.7343015670776367,
"learning_rate": 9.203508214822652e-06,
"loss": 0.6635,
"step": 83
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.792455792427063,
"learning_rate": 9.172866268606514e-06,
"loss": 0.6389,
"step": 84
},
{
"epoch": 0.8146964856230032,
"grad_norm": 0.7722944617271423,
"learning_rate": 9.141699014900084e-06,
"loss": 0.6848,
"step": 85
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.7590317130088806,
"learning_rate": 9.110010377239552e-06,
"loss": 0.6687,
"step": 86
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.6646780371665955,
"learning_rate": 9.077804344796302e-06,
"loss": 0.6772,
"step": 87
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.7406274676322937,
"learning_rate": 9.045084971874738e-06,
"loss": 0.649,
"step": 88
},
{
"epoch": 0.853035143769968,
"grad_norm": 0.7770557999610901,
"learning_rate": 9.011856377401891e-06,
"loss": 0.6667,
"step": 89
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.7461346387863159,
"learning_rate": 8.978122744408905e-06,
"loss": 0.6846,
"step": 90
},
{
"epoch": 0.8722044728434505,
"grad_norm": 0.8803783655166626,
"learning_rate": 8.943888319504456e-06,
"loss": 0.6694,
"step": 91
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.7138260006904602,
"learning_rate": 8.90915741234015e-06,
"loss": 0.6731,
"step": 92
},
{
"epoch": 0.8913738019169329,
"grad_norm": 0.7463406324386597,
"learning_rate": 8.873934395068006e-06,
"loss": 0.6445,
"step": 93
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.7805307507514954,
"learning_rate": 8.838223701790057e-06,
"loss": 0.6833,
"step": 94
},
{
"epoch": 0.9105431309904153,
"grad_norm": 0.7624490857124329,
"learning_rate": 8.802029828000157e-06,
"loss": 0.6291,
"step": 95
},
{
"epoch": 0.9201277955271565,
"grad_norm": 0.7260795831680298,
"learning_rate": 8.765357330018056e-06,
"loss": 0.6729,
"step": 96
},
{
"epoch": 0.9297124600638977,
"grad_norm": 0.8511263132095337,
"learning_rate": 8.728210824415829e-06,
"loss": 0.6617,
"step": 97
},
{
"epoch": 0.939297124600639,
"grad_norm": 0.7045103907585144,
"learning_rate": 8.690594987436705e-06,
"loss": 0.6557,
"step": 98
},
{
"epoch": 0.9488817891373802,
"grad_norm": 0.7786133885383606,
"learning_rate": 8.652514554406388e-06,
"loss": 0.6525,
"step": 99
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.9434441924095154,
"learning_rate": 8.613974319136959e-06,
"loss": 0.6618,
"step": 100
},
{
"epoch": 0.9680511182108626,
"grad_norm": 0.6732690930366516,
"learning_rate": 8.574979133323378e-06,
"loss": 0.6403,
"step": 101
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.8073576092720032,
"learning_rate": 8.535533905932739e-06,
"loss": 0.6523,
"step": 102
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.7025145292282104,
"learning_rate": 8.495643602586287e-06,
"loss": 0.6454,
"step": 103
},
{
"epoch": 0.9968051118210862,
"grad_norm": 0.7717574238777161,
"learning_rate": 8.455313244934324e-06,
"loss": 0.6529,
"step": 104
},
{
"epoch": 1.0063897763578276,
"grad_norm": 0.6937882900238037,
"learning_rate": 8.414547910024035e-06,
"loss": 0.6219,
"step": 105
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.6880476474761963,
"learning_rate": 8.373352729660373e-06,
"loss": 0.5887,
"step": 106
},
{
"epoch": 1.0255591054313098,
"grad_norm": 0.7143478393554688,
"learning_rate": 8.331732889760021e-06,
"loss": 0.5787,
"step": 107
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.7447161674499512,
"learning_rate": 8.289693629698564e-06,
"loss": 0.5986,
"step": 108
},
{
"epoch": 1.0447284345047922,
"grad_norm": 0.655197262763977,
"learning_rate": 8.247240241650918e-06,
"loss": 0.5626,
"step": 109
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.7953078746795654,
"learning_rate": 8.204378069925121e-06,
"loss": 0.5935,
"step": 110
},
{
"epoch": 1.0638977635782747,
"grad_norm": 0.712658166885376,
"learning_rate": 8.16111251028955e-06,
"loss": 0.5891,
"step": 111
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.6255548000335693,
"learning_rate": 8.117449009293668e-06,
"loss": 0.5729,
"step": 112
},
{
"epoch": 1.0830670926517572,
"grad_norm": 0.6540143489837646,
"learning_rate": 8.073393063582386e-06,
"loss": 0.5781,
"step": 113
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.6459587812423706,
"learning_rate": 8.0289502192041e-06,
"loss": 0.5843,
"step": 114
},
{
"epoch": 1.1022364217252396,
"grad_norm": 0.7313612103462219,
"learning_rate": 7.984126070912519e-06,
"loss": 0.5648,
"step": 115
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.6702539324760437,
"learning_rate": 7.938926261462366e-06,
"loss": 0.5617,
"step": 116
},
{
"epoch": 1.121405750798722,
"grad_norm": 0.6335576176643372,
"learning_rate": 7.89335648089903e-06,
"loss": 0.5983,
"step": 117
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.7993137836456299,
"learning_rate": 7.84742246584226e-06,
"loss": 0.6143,
"step": 118
},
{
"epoch": 1.1405750798722045,
"grad_norm": 0.7164519429206848,
"learning_rate": 7.801129998764014e-06,
"loss": 0.5664,
"step": 119
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.7602349519729614,
"learning_rate": 7.754484907260513e-06,
"loss": 0.5801,
"step": 120
},
{
"epoch": 1.159744408945687,
"grad_norm": 0.7163518071174622,
"learning_rate": 7.70749306331863e-06,
"loss": 0.5573,
"step": 121
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.6692156195640564,
"learning_rate": 7.660160382576683e-06,
"loss": 0.575,
"step": 122
},
{
"epoch": 1.1789137380191694,
"grad_norm": 0.8085399866104126,
"learning_rate": 7.612492823579744e-06,
"loss": 0.5878,
"step": 123
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.6267865896224976,
"learning_rate": 7.564496387029532e-06,
"loss": 0.5486,
"step": 124
},
{
"epoch": 1.1980830670926517,
"grad_norm": 0.6363678574562073,
"learning_rate": 7.516177115029002e-06,
"loss": 0.5884,
"step": 125
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.7670935392379761,
"learning_rate": 7.467541090321735e-06,
"loss": 0.5565,
"step": 126
},
{
"epoch": 1.2172523961661341,
"grad_norm": 0.640592634677887,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.5943,
"step": 127
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.7349753379821777,
"learning_rate": 7.369343312364994e-06,
"loss": 0.5965,
"step": 128
},
{
"epoch": 1.2364217252396166,
"grad_norm": 0.7965679168701172,
"learning_rate": 7.319793920889171e-06,
"loss": 0.5895,
"step": 129
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.5832921862602234,
"learning_rate": 7.269952498697734e-06,
"loss": 0.5785,
"step": 130
},
{
"epoch": 1.255591054313099,
"grad_norm": 0.7845919132232666,
"learning_rate": 7.219825320152411e-06,
"loss": 0.5873,
"step": 131
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.6836994886398315,
"learning_rate": 7.169418695587791e-06,
"loss": 0.5811,
"step": 132
},
{
"epoch": 1.2747603833865815,
"grad_norm": 0.5873911380767822,
"learning_rate": 7.118738970516944e-06,
"loss": 0.5854,
"step": 133
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.7451477646827698,
"learning_rate": 7.067792524832604e-06,
"loss": 0.5859,
"step": 134
},
{
"epoch": 1.293929712460064,
"grad_norm": 0.6415753364562988,
"learning_rate": 7.016585772004026e-06,
"loss": 0.5812,
"step": 135
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.6458070874214172,
"learning_rate": 6.965125158269619e-06,
"loss": 0.5712,
"step": 136
},
{
"epoch": 1.3130990415335464,
"grad_norm": 0.7372552752494812,
"learning_rate": 6.913417161825449e-06,
"loss": 0.5692,
"step": 137
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.6712161302566528,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.5681,
"step": 138
},
{
"epoch": 1.3322683706070286,
"grad_norm": 0.6904409527778625,
"learning_rate": 6.809285088483361e-06,
"loss": 0.576,
"step": 139
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.636581540107727,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.5779,
"step": 140
},
{
"epoch": 1.351437699680511,
"grad_norm": 0.6586715579032898,
"learning_rate": 6.704241985612625e-06,
"loss": 0.5499,
"step": 141
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.6217679381370544,
"learning_rate": 6.651395309775837e-06,
"loss": 0.5603,
"step": 142
},
{
"epoch": 1.3706070287539935,
"grad_norm": 0.7175468802452087,
"learning_rate": 6.598340745578908e-06,
"loss": 0.5949,
"step": 143
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.6404513716697693,
"learning_rate": 6.545084971874738e-06,
"loss": 0.5864,
"step": 144
},
{
"epoch": 1.389776357827476,
"grad_norm": 0.6180385947227478,
"learning_rate": 6.491634692845781e-06,
"loss": 0.5977,
"step": 145
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.6496903896331787,
"learning_rate": 6.437996637160086e-06,
"loss": 0.5824,
"step": 146
},
{
"epoch": 1.4089456869009584,
"grad_norm": 0.6618816256523132,
"learning_rate": 6.384177557124247e-06,
"loss": 0.5648,
"step": 147
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.6388576626777649,
"learning_rate": 6.330184227833376e-06,
"loss": 0.5641,
"step": 148
},
{
"epoch": 1.428115015974441,
"grad_norm": 0.6645902991294861,
"learning_rate": 6.276023446318214e-06,
"loss": 0.5642,
"step": 149
},
{
"epoch": 1.4376996805111821,
"grad_norm": 0.6369603872299194,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.5672,
"step": 150
},
{
"epoch": 1.4472843450479234,
"grad_norm": 0.7125694751739502,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.5738,
"step": 151
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.6303924918174744,
"learning_rate": 6.112604669781572e-06,
"loss": 0.574,
"step": 152
},
{
"epoch": 1.4664536741214058,
"grad_norm": 0.5970083475112915,
"learning_rate": 6.057842458386315e-06,
"loss": 0.5747,
"step": 153
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.6371821761131287,
"learning_rate": 6.002947078916365e-06,
"loss": 0.5653,
"step": 154
},
{
"epoch": 1.4856230031948883,
"grad_norm": 0.7160565257072449,
"learning_rate": 5.947925441958393e-06,
"loss": 0.5493,
"step": 155
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.6290352940559387,
"learning_rate": 5.892784473993184e-06,
"loss": 0.5723,
"step": 156
},
{
"epoch": 1.5047923322683707,
"grad_norm": 0.5443238615989685,
"learning_rate": 5.837531116523683e-06,
"loss": 0.5914,
"step": 157
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.6566057205200195,
"learning_rate": 5.782172325201155e-06,
"loss": 0.5577,
"step": 158
},
{
"epoch": 1.5239616613418532,
"grad_norm": 0.6568493247032166,
"learning_rate": 5.726715068949564e-06,
"loss": 0.5672,
"step": 159
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.6268693208694458,
"learning_rate": 5.671166329088278e-06,
"loss": 0.5853,
"step": 160
},
{
"epoch": 1.5431309904153354,
"grad_norm": 0.6314179301261902,
"learning_rate": 5.615533098453215e-06,
"loss": 0.5599,
"step": 161
},
{
"epoch": 1.5527156549520766,
"grad_norm": 0.5839089155197144,
"learning_rate": 5.559822380516539e-06,
"loss": 0.5578,
"step": 162
},
{
"epoch": 1.5623003194888179,
"grad_norm": 0.5961830019950867,
"learning_rate": 5.504041188505022e-06,
"loss": 0.5634,
"step": 163
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.6408848166465759,
"learning_rate": 5.448196544517168e-06,
"loss": 0.5677,
"step": 164
},
{
"epoch": 1.5814696485623003,
"grad_norm": 0.6148958802223206,
"learning_rate": 5.392295478639226e-06,
"loss": 0.5647,
"step": 165
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.5794315338134766,
"learning_rate": 5.336345028060199e-06,
"loss": 0.5792,
"step": 166
},
{
"epoch": 1.6006389776357828,
"grad_norm": 0.60219806432724,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.5477,
"step": 167
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.6012296676635742,
"learning_rate": 5.224324151752575e-06,
"loss": 0.583,
"step": 168
},
{
"epoch": 1.619808306709265,
"grad_norm": 0.6214751601219177,
"learning_rate": 5.168267827938971e-06,
"loss": 0.5439,
"step": 169
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.6018662452697754,
"learning_rate": 5.112190321479026e-06,
"loss": 0.5671,
"step": 170
},
{
"epoch": 1.6389776357827475,
"grad_norm": 0.5933774709701538,
"learning_rate": 5.05609869177323e-06,
"loss": 0.5886,
"step": 171
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.5990563035011292,
"learning_rate": 5e-06,
"loss": 0.5653,
"step": 172
},
{
"epoch": 1.65814696485623,
"grad_norm": 0.5904830694198608,
"learning_rate": 4.943901308226771e-06,
"loss": 0.5602,
"step": 173
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.6122473478317261,
"learning_rate": 4.887809678520976e-06,
"loss": 0.5801,
"step": 174
},
{
"epoch": 1.6773162939297124,
"grad_norm": 0.6307440400123596,
"learning_rate": 4.831732172061032e-06,
"loss": 0.5842,
"step": 175
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.6917486190795898,
"learning_rate": 4.775675848247427e-06,
"loss": 0.5595,
"step": 176
},
{
"epoch": 1.6964856230031948,
"grad_norm": 0.6185805797576904,
"learning_rate": 4.719647763814041e-06,
"loss": 0.5727,
"step": 177
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.5742789506912231,
"learning_rate": 4.663654971939802e-06,
"loss": 0.5575,
"step": 178
},
{
"epoch": 1.7156549520766773,
"grad_norm": 0.6561394333839417,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.5706,
"step": 179
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.6904816627502441,
"learning_rate": 4.551803455482833e-06,
"loss": 0.6007,
"step": 180
},
{
"epoch": 1.7348242811501597,
"grad_norm": 0.6116853356361389,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.5686,
"step": 181
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.6065500378608704,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.5829,
"step": 182
},
{
"epoch": 1.7539936102236422,
"grad_norm": 0.6076284050941467,
"learning_rate": 4.384466901546786e-06,
"loss": 0.5745,
"step": 183
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.5575659275054932,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.5708,
"step": 184
},
{
"epoch": 1.7731629392971247,
"grad_norm": 0.5849297046661377,
"learning_rate": 4.273284931050438e-06,
"loss": 0.5532,
"step": 185
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.5678250193595886,
"learning_rate": 4.217827674798845e-06,
"loss": 0.5676,
"step": 186
},
{
"epoch": 1.792332268370607,
"grad_norm": 0.5182955265045166,
"learning_rate": 4.162468883476319e-06,
"loss": 0.5656,
"step": 187
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.6109117269515991,
"learning_rate": 4.107215526006818e-06,
"loss": 0.5734,
"step": 188
},
{
"epoch": 1.8115015974440896,
"grad_norm": 0.6146546006202698,
"learning_rate": 4.052074558041608e-06,
"loss": 0.5891,
"step": 189
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.5946270823478699,
"learning_rate": 3.997052921083637e-06,
"loss": 0.5838,
"step": 190
},
{
"epoch": 1.830670926517572,
"grad_norm": 0.5628454089164734,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.5682,
"step": 191
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.5776045322418213,
"learning_rate": 3.887395330218429e-06,
"loss": 0.5774,
"step": 192
},
{
"epoch": 1.8498402555910545,
"grad_norm": 0.5610283017158508,
"learning_rate": 3.832773180720475e-06,
"loss": 0.5755,
"step": 193
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.5823736190795898,
"learning_rate": 3.778297969310529e-06,
"loss": 0.5894,
"step": 194
},
{
"epoch": 1.8690095846645367,
"grad_norm": 0.5524364709854126,
"learning_rate": 3.723976553681787e-06,
"loss": 0.5881,
"step": 195
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.6131855249404907,
"learning_rate": 3.669815772166625e-06,
"loss": 0.5699,
"step": 196
},
{
"epoch": 1.8881789137380192,
"grad_norm": 0.6087161898612976,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.5747,
"step": 197
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.591026782989502,
"learning_rate": 3.562003362839914e-06,
"loss": 0.5596,
"step": 198
},
{
"epoch": 1.9073482428115016,
"grad_norm": 0.5625490546226501,
"learning_rate": 3.50836530715422e-06,
"loss": 0.5709,
"step": 199
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.5438273549079895,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.5771,
"step": 200
},
{
"epoch": 1.926517571884984,
"grad_norm": 0.5201634764671326,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.5763,
"step": 201
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.6035382151603699,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.5495,
"step": 202
},
{
"epoch": 1.9456869009584663,
"grad_norm": 0.539597749710083,
"learning_rate": 3.295758014387375e-06,
"loss": 0.5836,
"step": 203
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.4887038469314575,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.5865,
"step": 204
},
{
"epoch": 1.9648562300319488,
"grad_norm": 0.5846530795097351,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.5638,
"step": 205
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.5618496537208557,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.5784,
"step": 206
},
{
"epoch": 1.9840255591054312,
"grad_norm": 0.5010558366775513,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.572,
"step": 207
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.5525842308998108,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.5578,
"step": 208
},
{
"epoch": 2.0031948881789137,
"grad_norm": 0.5372030735015869,
"learning_rate": 2.9834142279959754e-06,
"loss": 0.5419,
"step": 209
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.5794517397880554,
"learning_rate": 2.932207475167398e-06,
"loss": 0.5237,
"step": 210
},
{
"epoch": 2.022364217252396,
"grad_norm": 0.5482734441757202,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.523,
"step": 211
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.49512386322021484,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5488,
"step": 212
},
{
"epoch": 2.0415335463258786,
"grad_norm": 0.5444460511207581,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.5362,
"step": 213
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.5378485321998596,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.5117,
"step": 214
},
{
"epoch": 2.060702875399361,
"grad_norm": 0.570003092288971,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.4838,
"step": 215
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.6234847903251648,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5046,
"step": 216
},
{
"epoch": 2.0798722044728435,
"grad_norm": 0.5408498644828796,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.5185,
"step": 217
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.5204379558563232,
"learning_rate": 2.532458909678266e-06,
"loss": 0.496,
"step": 218
},
{
"epoch": 2.099041533546326,
"grad_norm": 0.5795621871948242,
"learning_rate": 2.483822884971e-06,
"loss": 0.5253,
"step": 219
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.5828584432601929,
"learning_rate": 2.43550361297047e-06,
"loss": 0.5122,
"step": 220
},
{
"epoch": 2.1182108626198084,
"grad_norm": 0.6278659105300903,
"learning_rate": 2.387507176420256e-06,
"loss": 0.4836,
"step": 221
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.5321851968765259,
"learning_rate": 2.339839617423318e-06,
"loss": 0.5254,
"step": 222
},
{
"epoch": 2.137380191693291,
"grad_norm": 0.5402694940567017,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.5279,
"step": 223
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.528405487537384,
"learning_rate": 2.245515092739488e-06,
"loss": 0.4915,
"step": 224
},
{
"epoch": 2.1565495207667733,
"grad_norm": 0.5780546069145203,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.5057,
"step": 225
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.5286846160888672,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.4826,
"step": 226
},
{
"epoch": 2.1757188498402558,
"grad_norm": 0.5604357719421387,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.5041,
"step": 227
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.5224547386169434,
"learning_rate": 2.061073738537635e-06,
"loss": 0.504,
"step": 228
},
{
"epoch": 2.194888178913738,
"grad_norm": 0.49711328744888306,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.4867,
"step": 229
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.5324923396110535,
"learning_rate": 1.971049780795901e-06,
"loss": 0.5058,
"step": 230
},
{
"epoch": 2.2140575079872207,
"grad_norm": 0.5620575547218323,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.491,
"step": 231
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.6187978386878967,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.4932,
"step": 232
},
{
"epoch": 2.2332268370607027,
"grad_norm": 0.5564191937446594,
"learning_rate": 1.838887489710452e-06,
"loss": 0.5167,
"step": 233
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.544456422328949,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.5106,
"step": 234
},
{
"epoch": 2.252396166134185,
"grad_norm": 0.49560070037841797,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.4985,
"step": 235
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.5730255246162415,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.4861,
"step": 236
},
{
"epoch": 2.2715654952076676,
"grad_norm": 0.5446303486824036,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.5204,
"step": 237
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.5347279906272888,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.514,
"step": 238
},
{
"epoch": 2.29073482428115,
"grad_norm": 0.5006322860717773,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.5042,
"step": 239
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.5397834777832031,
"learning_rate": 1.544686755065677e-06,
"loss": 0.5044,
"step": 240
},
{
"epoch": 2.3099041533546325,
"grad_norm": 0.5032399892807007,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.5072,
"step": 241
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.47611555457115173,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5302,
"step": 242
},
{
"epoch": 2.329073482428115,
"grad_norm": 0.5055474042892456,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.5042,
"step": 243
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.5255368947982788,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.5111,
"step": 244
},
{
"epoch": 2.3482428115015974,
"grad_norm": 0.5098780393600464,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.515,
"step": 245
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.4925176501274109,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.5114,
"step": 246
},
{
"epoch": 2.36741214057508,
"grad_norm": 0.537067174911499,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.4797,
"step": 247
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.4920751750469208,
"learning_rate": 1.234642669981946e-06,
"loss": 0.4941,
"step": 248
},
{
"epoch": 2.3865814696485623,
"grad_norm": 0.4458199143409729,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.5222,
"step": 249
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.48658084869384766,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.518,
"step": 250
},
{
"epoch": 2.405750798722045,
"grad_norm": 0.5211532115936279,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.4984,
"step": 251
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.5080324411392212,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5229,
"step": 252
},
{
"epoch": 2.4249201277955272,
"grad_norm": 0.5140640735626221,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.5146,
"step": 253
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.48215627670288086,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.5131,
"step": 254
},
{
"epoch": 2.4440894568690097,
"grad_norm": 0.5040873289108276,
"learning_rate": 9.881436225981107e-07,
"loss": 0.4914,
"step": 255
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.5259299874305725,
"learning_rate": 9.549150281252633e-07,
"loss": 0.5022,
"step": 256
},
{
"epoch": 2.463258785942492,
"grad_norm": 0.5273531079292297,
"learning_rate": 9.221956552036992e-07,
"loss": 0.4945,
"step": 257
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.49242982268333435,
"learning_rate": 8.899896227604509e-07,
"loss": 0.4967,
"step": 258
},
{
"epoch": 2.4824281150159746,
"grad_norm": 0.48913198709487915,
"learning_rate": 8.58300985099918e-07,
"loss": 0.4912,
"step": 259
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.4919417202472687,
"learning_rate": 8.271337313934869e-07,
"loss": 0.4789,
"step": 260
},
{
"epoch": 2.501597444089457,
"grad_norm": 0.4618432819843292,
"learning_rate": 7.964917851773496e-07,
"loss": 0.5138,
"step": 261
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.5181034207344055,
"learning_rate": 7.663790038585794e-07,
"loss": 0.4869,
"step": 262
},
{
"epoch": 2.520766773162939,
"grad_norm": 0.4994615912437439,
"learning_rate": 7.367991782295392e-07,
"loss": 0.487,
"step": 263
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.4796070158481598,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5115,
"step": 264
},
{
"epoch": 2.539936102236422,
"grad_norm": 0.4891747832298279,
"learning_rate": 6.792532212817271e-07,
"loss": 0.5304,
"step": 265
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.4947539269924164,
"learning_rate": 6.512943342215234e-07,
"loss": 0.5268,
"step": 266
},
{
"epoch": 2.559105431309904,
"grad_norm": 0.48322218656539917,
"learning_rate": 6.238828904562316e-07,
"loss": 0.4955,
"step": 267
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.4670308232307434,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5186,
"step": 268
},
{
"epoch": 2.5782747603833864,
"grad_norm": 0.5252702236175537,
"learning_rate": 5.707160663821009e-07,
"loss": 0.4998,
"step": 269
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.4638543426990509,
"learning_rate": 5.449673790581611e-07,
"loss": 0.4974,
"step": 270
},
{
"epoch": 2.597444089456869,
"grad_norm": 0.5163476467132568,
"learning_rate": 5.197795201563744e-07,
"loss": 0.4794,
"step": 271
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.4854280948638916,
"learning_rate": 4.951556604879049e-07,
"loss": 0.4974,
"step": 272
},
{
"epoch": 2.6166134185303513,
"grad_norm": 0.4789203107357025,
"learning_rate": 4.710988998640298e-07,
"loss": 0.4988,
"step": 273
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.4644279181957245,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.512,
"step": 274
},
{
"epoch": 2.635782747603834,
"grad_norm": 0.5351918935775757,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.4738,
"step": 275
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.48578253388404846,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5278,
"step": 276
},
{
"epoch": 2.6549520766773163,
"grad_norm": 0.4682360291481018,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.5055,
"step": 277
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.49804800748825073,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.4815,
"step": 278
},
{
"epoch": 2.6741214057507987,
"grad_norm": 0.48392346501350403,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.5133,
"step": 279
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.4864685535430908,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5029,
"step": 280
},
{
"epoch": 2.693290734824281,
"grad_norm": 0.5228228569030762,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.4995,
"step": 281
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.4820217192173004,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5045,
"step": 282
},
{
"epoch": 2.7124600638977636,
"grad_norm": 0.4983007311820984,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.49,
"step": 283
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.48497146368026733,
"learning_rate": 2.447174185242324e-07,
"loss": 0.4986,
"step": 284
},
{
"epoch": 2.731629392971246,
"grad_norm": 0.4843176603317261,
"learning_rate": 2.276812823220964e-07,
"loss": 0.4864,
"step": 285
},
{
"epoch": 2.741214057507987,
"grad_norm": 0.4612375795841217,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.5138,
"step": 286
},
{
"epoch": 2.7507987220447285,
"grad_norm": 0.4730239808559418,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.498,
"step": 287
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.47294536232948303,
"learning_rate": 1.801856965207338e-07,
"loss": 0.517,
"step": 288
},
{
"epoch": 2.769968051118211,
"grad_norm": 0.490991473197937,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.5219,
"step": 289
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.5257830619812012,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5048,
"step": 290
},
{
"epoch": 2.7891373801916934,
"grad_norm": 0.46119868755340576,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.5258,
"step": 291
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.5001710653305054,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5056,
"step": 292
},
{
"epoch": 2.8083067092651754,
"grad_norm": 0.45364686846733093,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.525,
"step": 293
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.511743426322937,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.5028,
"step": 294
},
{
"epoch": 2.8274760383386583,
"grad_norm": 0.48278823494911194,
"learning_rate": 9.0678523819408e-08,
"loss": 0.5026,
"step": 295
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.4801782965660095,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5173,
"step": 296
},
{
"epoch": 2.8466453674121404,
"grad_norm": 0.4829420745372772,
"learning_rate": 7.064490740882057e-08,
"loss": 0.5111,
"step": 297
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.49983468651771545,
"learning_rate": 6.15582970243117e-08,
"loss": 0.4989,
"step": 298
},
{
"epoch": 2.8658146964856233,
"grad_norm": 0.4892556667327881,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.5111,
"step": 299
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.5195422172546387,
"learning_rate": 4.52511911603265e-08,
"loss": 0.4896,
"step": 300
},
{
"epoch": 2.8849840255591053,
"grad_norm": 0.4906642436981201,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.5274,
"step": 301
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.499541699886322,
"learning_rate": 3.143895053378698e-08,
"loss": 0.4915,
"step": 302
},
{
"epoch": 2.9041533546325877,
"grad_norm": 0.48538094758987427,
"learning_rate": 2.547062725623828e-08,
"loss": 0.5059,
"step": 303
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.4701797366142273,
"learning_rate": 2.012853002380466e-08,
"loss": 0.4837,
"step": 304
},
{
"epoch": 2.92332268370607,
"grad_norm": 0.5019702315330505,
"learning_rate": 1.541333133436018e-08,
"loss": 0.5137,
"step": 305
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.5122066736221313,
"learning_rate": 1.132562476771959e-08,
"loss": 0.5035,
"step": 306
},
{
"epoch": 2.9424920127795526,
"grad_norm": 0.4683079719543457,
"learning_rate": 7.865924910916977e-09,
"loss": 0.5052,
"step": 307
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.48147568106651306,
"learning_rate": 5.034667293427053e-09,
"loss": 0.4787,
"step": 308
},
{
"epoch": 2.961661341853035,
"grad_norm": 0.4846435487270355,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.498,
"step": 309
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.4767644703388214,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5103,
"step": 310
},
{
"epoch": 2.9808306709265175,
"grad_norm": 0.48420077562332153,
"learning_rate": 3.147162264971471e-10,
"loss": 0.4959,
"step": 311
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.4982874095439911,
"learning_rate": 0.0,
"loss": 0.5048,
"step": 312
},
{
"epoch": 2.9904153354632586,
"step": 312,
"total_flos": 2.995131507754271e+17,
"train_loss": 0.6095600314438343,
"train_runtime": 8590.5636,
"train_samples_per_second": 3.492,
"train_steps_per_second": 0.036
}
],
"logging_steps": 1,
"max_steps": 312,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.995131507754271e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}