EtashGuha's picture
Upload folder using huggingface_hub
06c865f 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": 6.939575672149658,
"learning_rate": 3.125e-07,
"loss": 1.092,
"step": 1
},
{
"epoch": 0.019169329073482427,
"grad_norm": 6.784217834472656,
"learning_rate": 6.25e-07,
"loss": 1.077,
"step": 2
},
{
"epoch": 0.02875399361022364,
"grad_norm": 6.724907398223877,
"learning_rate": 9.375000000000001e-07,
"loss": 1.0719,
"step": 3
},
{
"epoch": 0.038338658146964855,
"grad_norm": 6.608644962310791,
"learning_rate": 1.25e-06,
"loss": 1.0854,
"step": 4
},
{
"epoch": 0.04792332268370607,
"grad_norm": 6.478516578674316,
"learning_rate": 1.5625e-06,
"loss": 1.0673,
"step": 5
},
{
"epoch": 0.05750798722044728,
"grad_norm": 6.00268030166626,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.073,
"step": 6
},
{
"epoch": 0.0670926517571885,
"grad_norm": 5.025429725646973,
"learning_rate": 2.1875000000000002e-06,
"loss": 1.0296,
"step": 7
},
{
"epoch": 0.07667731629392971,
"grad_norm": 4.760328769683838,
"learning_rate": 2.5e-06,
"loss": 1.0146,
"step": 8
},
{
"epoch": 0.08626198083067092,
"grad_norm": 2.4226059913635254,
"learning_rate": 2.8125e-06,
"loss": 0.9903,
"step": 9
},
{
"epoch": 0.09584664536741214,
"grad_norm": 2.346510648727417,
"learning_rate": 3.125e-06,
"loss": 0.9845,
"step": 10
},
{
"epoch": 0.10543130990415335,
"grad_norm": 2.1026291847229004,
"learning_rate": 3.4375e-06,
"loss": 0.9115,
"step": 11
},
{
"epoch": 0.11501597444089456,
"grad_norm": 3.6948068141937256,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.9408,
"step": 12
},
{
"epoch": 0.12460063897763578,
"grad_norm": 3.8251421451568604,
"learning_rate": 4.0625000000000005e-06,
"loss": 0.9102,
"step": 13
},
{
"epoch": 0.134185303514377,
"grad_norm": 3.8101155757904053,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.9022,
"step": 14
},
{
"epoch": 0.14376996805111822,
"grad_norm": 3.657567262649536,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.9069,
"step": 15
},
{
"epoch": 0.15335463258785942,
"grad_norm": 2.8502235412597656,
"learning_rate": 5e-06,
"loss": 0.9133,
"step": 16
},
{
"epoch": 0.16293929712460065,
"grad_norm": 2.3882975578308105,
"learning_rate": 5.3125e-06,
"loss": 0.8877,
"step": 17
},
{
"epoch": 0.17252396166134185,
"grad_norm": 2.225501775741577,
"learning_rate": 5.625e-06,
"loss": 0.8885,
"step": 18
},
{
"epoch": 0.18210862619808307,
"grad_norm": 1.9925498962402344,
"learning_rate": 5.9375e-06,
"loss": 0.8578,
"step": 19
},
{
"epoch": 0.19169329073482427,
"grad_norm": 1.6763207912445068,
"learning_rate": 6.25e-06,
"loss": 0.8741,
"step": 20
},
{
"epoch": 0.2012779552715655,
"grad_norm": 1.6249445676803589,
"learning_rate": 6.5625e-06,
"loss": 0.8348,
"step": 21
},
{
"epoch": 0.2108626198083067,
"grad_norm": 1.5608230829238892,
"learning_rate": 6.875e-06,
"loss": 0.8248,
"step": 22
},
{
"epoch": 0.22044728434504793,
"grad_norm": 1.5464677810668945,
"learning_rate": 7.1875e-06,
"loss": 0.8526,
"step": 23
},
{
"epoch": 0.23003194888178913,
"grad_norm": 1.2905473709106445,
"learning_rate": 7.500000000000001e-06,
"loss": 0.8346,
"step": 24
},
{
"epoch": 0.23961661341853036,
"grad_norm": 1.2712877988815308,
"learning_rate": 7.8125e-06,
"loss": 0.815,
"step": 25
},
{
"epoch": 0.24920127795527156,
"grad_norm": 1.2141337394714355,
"learning_rate": 8.125000000000001e-06,
"loss": 0.801,
"step": 26
},
{
"epoch": 0.25878594249201275,
"grad_norm": 1.2987741231918335,
"learning_rate": 8.4375e-06,
"loss": 0.7947,
"step": 27
},
{
"epoch": 0.268370607028754,
"grad_norm": 1.0082370042800903,
"learning_rate": 8.750000000000001e-06,
"loss": 0.7854,
"step": 28
},
{
"epoch": 0.2779552715654952,
"grad_norm": 0.8668848872184753,
"learning_rate": 9.0625e-06,
"loss": 0.7644,
"step": 29
},
{
"epoch": 0.28753993610223644,
"grad_norm": 1.0266895294189453,
"learning_rate": 9.375000000000001e-06,
"loss": 0.7381,
"step": 30
},
{
"epoch": 0.2971246006389776,
"grad_norm": 1.1279035806655884,
"learning_rate": 9.6875e-06,
"loss": 0.7835,
"step": 31
},
{
"epoch": 0.30670926517571884,
"grad_norm": 0.9569628834724426,
"learning_rate": 1e-05,
"loss": 0.7318,
"step": 32
},
{
"epoch": 0.31629392971246006,
"grad_norm": 0.8320262432098389,
"learning_rate": 9.999685283773504e-06,
"loss": 0.7448,
"step": 33
},
{
"epoch": 0.3258785942492013,
"grad_norm": 1.1430109739303589,
"learning_rate": 9.998741174712534e-06,
"loss": 0.7641,
"step": 34
},
{
"epoch": 0.3354632587859425,
"grad_norm": 0.8715132474899292,
"learning_rate": 9.997167791667668e-06,
"loss": 0.757,
"step": 35
},
{
"epoch": 0.3450479233226837,
"grad_norm": 0.8928182125091553,
"learning_rate": 9.994965332706574e-06,
"loss": 0.7382,
"step": 36
},
{
"epoch": 0.3546325878594249,
"grad_norm": 0.9497056603431702,
"learning_rate": 9.992134075089085e-06,
"loss": 0.7911,
"step": 37
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.8989669680595398,
"learning_rate": 9.98867437523228e-06,
"loss": 0.7649,
"step": 38
},
{
"epoch": 0.3738019169329074,
"grad_norm": 0.8037053942680359,
"learning_rate": 9.984586668665641e-06,
"loss": 0.7579,
"step": 39
},
{
"epoch": 0.38338658146964855,
"grad_norm": 0.7577980160713196,
"learning_rate": 9.979871469976197e-06,
"loss": 0.7541,
"step": 40
},
{
"epoch": 0.3929712460063898,
"grad_norm": 0.8141373991966248,
"learning_rate": 9.974529372743762e-06,
"loss": 0.7524,
"step": 41
},
{
"epoch": 0.402555910543131,
"grad_norm": 0.8163743615150452,
"learning_rate": 9.968561049466214e-06,
"loss": 0.7752,
"step": 42
},
{
"epoch": 0.41214057507987223,
"grad_norm": 0.7045100331306458,
"learning_rate": 9.961967251474823e-06,
"loss": 0.7566,
"step": 43
},
{
"epoch": 0.4217252396166134,
"grad_norm": 0.7221251130104065,
"learning_rate": 9.954748808839675e-06,
"loss": 0.7619,
"step": 44
},
{
"epoch": 0.43130990415335463,
"grad_norm": 0.8507441282272339,
"learning_rate": 9.946906630265184e-06,
"loss": 0.7439,
"step": 45
},
{
"epoch": 0.44089456869009586,
"grad_norm": 0.750127911567688,
"learning_rate": 9.938441702975689e-06,
"loss": 0.7372,
"step": 46
},
{
"epoch": 0.4504792332268371,
"grad_norm": 0.6966735124588013,
"learning_rate": 9.92935509259118e-06,
"loss": 0.7383,
"step": 47
},
{
"epoch": 0.46006389776357826,
"grad_norm": 0.6586233377456665,
"learning_rate": 9.91964794299315e-06,
"loss": 0.7211,
"step": 48
},
{
"epoch": 0.4696485623003195,
"grad_norm": 0.6894088983535767,
"learning_rate": 9.909321476180594e-06,
"loss": 0.7245,
"step": 49
},
{
"epoch": 0.4792332268370607,
"grad_norm": 0.7224133610725403,
"learning_rate": 9.898376992116179e-06,
"loss": 0.7468,
"step": 50
},
{
"epoch": 0.48881789137380194,
"grad_norm": 0.7163975834846497,
"learning_rate": 9.886815868562596e-06,
"loss": 0.7039,
"step": 51
},
{
"epoch": 0.4984025559105431,
"grad_norm": 0.677842915058136,
"learning_rate": 9.874639560909118e-06,
"loss": 0.7176,
"step": 52
},
{
"epoch": 0.5079872204472844,
"grad_norm": 0.693691611289978,
"learning_rate": 9.861849601988384e-06,
"loss": 0.7442,
"step": 53
},
{
"epoch": 0.5175718849840255,
"grad_norm": 0.7434005737304688,
"learning_rate": 9.848447601883436e-06,
"loss": 0.7622,
"step": 54
},
{
"epoch": 0.5271565495207667,
"grad_norm": 0.6400785446166992,
"learning_rate": 9.834435247725032e-06,
"loss": 0.7183,
"step": 55
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.7107494473457336,
"learning_rate": 9.819814303479268e-06,
"loss": 0.7556,
"step": 56
},
{
"epoch": 0.5463258785942492,
"grad_norm": 0.6146755218505859,
"learning_rate": 9.804586609725499e-06,
"loss": 0.7142,
"step": 57
},
{
"epoch": 0.5559105431309904,
"grad_norm": 0.7014209032058716,
"learning_rate": 9.788754083424654e-06,
"loss": 0.716,
"step": 58
},
{
"epoch": 0.5654952076677316,
"grad_norm": 0.7889743447303772,
"learning_rate": 9.772318717677905e-06,
"loss": 0.7368,
"step": 59
},
{
"epoch": 0.5750798722044729,
"grad_norm": 0.644163191318512,
"learning_rate": 9.755282581475769e-06,
"loss": 0.7456,
"step": 60
},
{
"epoch": 0.5846645367412141,
"grad_norm": 0.6978005766868591,
"learning_rate": 9.737647819437645e-06,
"loss": 0.7367,
"step": 61
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.7659692764282227,
"learning_rate": 9.719416651541839e-06,
"loss": 0.7023,
"step": 62
},
{
"epoch": 0.6038338658146964,
"grad_norm": 0.6721009016036987,
"learning_rate": 9.700591372846096e-06,
"loss": 0.7314,
"step": 63
},
{
"epoch": 0.6134185303514377,
"grad_norm": 0.5567018985748291,
"learning_rate": 9.681174353198687e-06,
"loss": 0.7017,
"step": 64
},
{
"epoch": 0.6230031948881789,
"grad_norm": 0.6971027851104736,
"learning_rate": 9.661168036940071e-06,
"loss": 0.7307,
"step": 65
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.8438630700111389,
"learning_rate": 9.640574942595195e-06,
"loss": 0.7203,
"step": 66
},
{
"epoch": 0.6421725239616614,
"grad_norm": 0.5880740284919739,
"learning_rate": 9.619397662556434e-06,
"loss": 0.7196,
"step": 67
},
{
"epoch": 0.6517571884984026,
"grad_norm": 0.6353415250778198,
"learning_rate": 9.597638862757255e-06,
"loss": 0.7497,
"step": 68
},
{
"epoch": 0.6613418530351438,
"grad_norm": 0.8600002527236938,
"learning_rate": 9.5753012823366e-06,
"loss": 0.722,
"step": 69
},
{
"epoch": 0.670926517571885,
"grad_norm": 0.7225138545036316,
"learning_rate": 9.552387733294081e-06,
"loss": 0.7196,
"step": 70
},
{
"epoch": 0.6805111821086262,
"grad_norm": 0.7042371034622192,
"learning_rate": 9.528901100135971e-06,
"loss": 0.7034,
"step": 71
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.6961320638656616,
"learning_rate": 9.504844339512096e-06,
"loss": 0.7045,
"step": 72
},
{
"epoch": 0.6996805111821086,
"grad_norm": 0.6758981943130493,
"learning_rate": 9.480220479843627e-06,
"loss": 0.7111,
"step": 73
},
{
"epoch": 0.7092651757188498,
"grad_norm": 0.8035303950309753,
"learning_rate": 9.45503262094184e-06,
"loss": 0.7555,
"step": 74
},
{
"epoch": 0.7188498402555911,
"grad_norm": 0.6938453316688538,
"learning_rate": 9.4292839336179e-06,
"loss": 0.7349,
"step": 75
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.6870027184486389,
"learning_rate": 9.40297765928369e-06,
"loss": 0.7501,
"step": 76
},
{
"epoch": 0.7380191693290735,
"grad_norm": 0.7021780610084534,
"learning_rate": 9.376117109543769e-06,
"loss": 0.713,
"step": 77
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.7315818071365356,
"learning_rate": 9.348705665778479e-06,
"loss": 0.7179,
"step": 78
},
{
"epoch": 0.7571884984025559,
"grad_norm": 0.6191219687461853,
"learning_rate": 9.320746778718274e-06,
"loss": 0.6881,
"step": 79
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.6984411478042603,
"learning_rate": 9.292243968009332e-06,
"loss": 0.7276,
"step": 80
},
{
"epoch": 0.7763578274760383,
"grad_norm": 0.6701769232749939,
"learning_rate": 9.263200821770462e-06,
"loss": 0.7063,
"step": 81
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.6397016048431396,
"learning_rate": 9.233620996141421e-06,
"loss": 0.6909,
"step": 82
},
{
"epoch": 0.7955271565495208,
"grad_norm": 0.7517321109771729,
"learning_rate": 9.203508214822652e-06,
"loss": 0.7169,
"step": 83
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.6491408944129944,
"learning_rate": 9.172866268606514e-06,
"loss": 0.7508,
"step": 84
},
{
"epoch": 0.8146964856230032,
"grad_norm": 0.6870864629745483,
"learning_rate": 9.141699014900084e-06,
"loss": 0.7087,
"step": 85
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.7963765263557434,
"learning_rate": 9.110010377239552e-06,
"loss": 0.7227,
"step": 86
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.6234023571014404,
"learning_rate": 9.077804344796302e-06,
"loss": 0.7345,
"step": 87
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.7726128697395325,
"learning_rate": 9.045084971874738e-06,
"loss": 0.7069,
"step": 88
},
{
"epoch": 0.853035143769968,
"grad_norm": 0.6299132108688354,
"learning_rate": 9.011856377401891e-06,
"loss": 0.7078,
"step": 89
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.6428517699241638,
"learning_rate": 8.978122744408905e-06,
"loss": 0.7004,
"step": 90
},
{
"epoch": 0.8722044728434505,
"grad_norm": 0.6417080760002136,
"learning_rate": 8.943888319504456e-06,
"loss": 0.7104,
"step": 91
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.6996341943740845,
"learning_rate": 8.90915741234015e-06,
"loss": 0.725,
"step": 92
},
{
"epoch": 0.8913738019169329,
"grad_norm": 0.7153524160385132,
"learning_rate": 8.873934395068006e-06,
"loss": 0.7068,
"step": 93
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.7380518317222595,
"learning_rate": 8.838223701790057e-06,
"loss": 0.7374,
"step": 94
},
{
"epoch": 0.9105431309904153,
"grad_norm": 0.6673450469970703,
"learning_rate": 8.802029828000157e-06,
"loss": 0.6961,
"step": 95
},
{
"epoch": 0.9201277955271565,
"grad_norm": 0.6847995519638062,
"learning_rate": 8.765357330018056e-06,
"loss": 0.7032,
"step": 96
},
{
"epoch": 0.9297124600638977,
"grad_norm": 0.7103424668312073,
"learning_rate": 8.728210824415829e-06,
"loss": 0.6924,
"step": 97
},
{
"epoch": 0.939297124600639,
"grad_norm": 0.618880569934845,
"learning_rate": 8.690594987436705e-06,
"loss": 0.6999,
"step": 98
},
{
"epoch": 0.9488817891373802,
"grad_norm": 0.7030272483825684,
"learning_rate": 8.652514554406388e-06,
"loss": 0.7118,
"step": 99
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.6193640232086182,
"learning_rate": 8.613974319136959e-06,
"loss": 0.6934,
"step": 100
},
{
"epoch": 0.9680511182108626,
"grad_norm": 0.6055111885070801,
"learning_rate": 8.574979133323378e-06,
"loss": 0.7118,
"step": 101
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.6267082691192627,
"learning_rate": 8.535533905932739e-06,
"loss": 0.6991,
"step": 102
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.6348923444747925,
"learning_rate": 8.495643602586287e-06,
"loss": 0.6876,
"step": 103
},
{
"epoch": 0.9968051118210862,
"grad_norm": 0.6422063112258911,
"learning_rate": 8.455313244934324e-06,
"loss": 0.6937,
"step": 104
},
{
"epoch": 1.0063897763578276,
"grad_norm": 0.6531258821487427,
"learning_rate": 8.414547910024035e-06,
"loss": 0.6681,
"step": 105
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.7352398633956909,
"learning_rate": 8.373352729660373e-06,
"loss": 0.6438,
"step": 106
},
{
"epoch": 1.0255591054313098,
"grad_norm": 0.5887332558631897,
"learning_rate": 8.331732889760021e-06,
"loss": 0.6295,
"step": 107
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.663162350654602,
"learning_rate": 8.289693629698564e-06,
"loss": 0.6198,
"step": 108
},
{
"epoch": 1.0447284345047922,
"grad_norm": 0.6586565375328064,
"learning_rate": 8.247240241650918e-06,
"loss": 0.6178,
"step": 109
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.5869852304458618,
"learning_rate": 8.204378069925121e-06,
"loss": 0.6102,
"step": 110
},
{
"epoch": 1.0638977635782747,
"grad_norm": 0.6440607905387878,
"learning_rate": 8.16111251028955e-06,
"loss": 0.6202,
"step": 111
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.7003380060195923,
"learning_rate": 8.117449009293668e-06,
"loss": 0.6194,
"step": 112
},
{
"epoch": 1.0830670926517572,
"grad_norm": 0.626481294631958,
"learning_rate": 8.073393063582386e-06,
"loss": 0.6408,
"step": 113
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.6357825994491577,
"learning_rate": 8.0289502192041e-06,
"loss": 0.5642,
"step": 114
},
{
"epoch": 1.1022364217252396,
"grad_norm": 0.6931595802307129,
"learning_rate": 7.984126070912519e-06,
"loss": 0.6228,
"step": 115
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.7018569111824036,
"learning_rate": 7.938926261462366e-06,
"loss": 0.6297,
"step": 116
},
{
"epoch": 1.121405750798722,
"grad_norm": 0.6461851596832275,
"learning_rate": 7.89335648089903e-06,
"loss": 0.6336,
"step": 117
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.7627151608467102,
"learning_rate": 7.84742246584226e-06,
"loss": 0.6435,
"step": 118
},
{
"epoch": 1.1405750798722045,
"grad_norm": 0.7167479991912842,
"learning_rate": 7.801129998764014e-06,
"loss": 0.6336,
"step": 119
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.5891663432121277,
"learning_rate": 7.754484907260513e-06,
"loss": 0.6154,
"step": 120
},
{
"epoch": 1.159744408945687,
"grad_norm": 0.6229691505432129,
"learning_rate": 7.70749306331863e-06,
"loss": 0.6428,
"step": 121
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.6488680243492126,
"learning_rate": 7.660160382576683e-06,
"loss": 0.6121,
"step": 122
},
{
"epoch": 1.1789137380191694,
"grad_norm": 0.6240501999855042,
"learning_rate": 7.612492823579744e-06,
"loss": 0.6307,
"step": 123
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.5995883345603943,
"learning_rate": 7.564496387029532e-06,
"loss": 0.63,
"step": 124
},
{
"epoch": 1.1980830670926517,
"grad_norm": 0.6613398790359497,
"learning_rate": 7.516177115029002e-06,
"loss": 0.6202,
"step": 125
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.6537646055221558,
"learning_rate": 7.467541090321735e-06,
"loss": 0.6307,
"step": 126
},
{
"epoch": 1.2172523961661341,
"grad_norm": 0.6241499185562134,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.6415,
"step": 127
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.6667526960372925,
"learning_rate": 7.369343312364994e-06,
"loss": 0.6293,
"step": 128
},
{
"epoch": 1.2364217252396166,
"grad_norm": 0.6463608145713806,
"learning_rate": 7.319793920889171e-06,
"loss": 0.6249,
"step": 129
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.7481969594955444,
"learning_rate": 7.269952498697734e-06,
"loss": 0.6202,
"step": 130
},
{
"epoch": 1.255591054313099,
"grad_norm": 0.6490833759307861,
"learning_rate": 7.219825320152411e-06,
"loss": 0.6011,
"step": 131
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.6010299324989319,
"learning_rate": 7.169418695587791e-06,
"loss": 0.64,
"step": 132
},
{
"epoch": 1.2747603833865815,
"grad_norm": 0.6452000141143799,
"learning_rate": 7.118738970516944e-06,
"loss": 0.6521,
"step": 133
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.6593641042709351,
"learning_rate": 7.067792524832604e-06,
"loss": 0.6356,
"step": 134
},
{
"epoch": 1.293929712460064,
"grad_norm": 0.6202598214149475,
"learning_rate": 7.016585772004026e-06,
"loss": 0.6174,
"step": 135
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.6200841665267944,
"learning_rate": 6.965125158269619e-06,
"loss": 0.6062,
"step": 136
},
{
"epoch": 1.3130990415335464,
"grad_norm": 0.7586333751678467,
"learning_rate": 6.913417161825449e-06,
"loss": 0.6489,
"step": 137
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.6437525749206543,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.6473,
"step": 138
},
{
"epoch": 1.3322683706070286,
"grad_norm": 0.5538833737373352,
"learning_rate": 6.809285088483361e-06,
"loss": 0.6072,
"step": 139
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.5944742560386658,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.6133,
"step": 140
},
{
"epoch": 1.351437699680511,
"grad_norm": 0.5477206110954285,
"learning_rate": 6.704241985612625e-06,
"loss": 0.612,
"step": 141
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.5899629592895508,
"learning_rate": 6.651395309775837e-06,
"loss": 0.6429,
"step": 142
},
{
"epoch": 1.3706070287539935,
"grad_norm": 0.5220656394958496,
"learning_rate": 6.598340745578908e-06,
"loss": 0.5942,
"step": 143
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.5558663010597229,
"learning_rate": 6.545084971874738e-06,
"loss": 0.6417,
"step": 144
},
{
"epoch": 1.389776357827476,
"grad_norm": 0.5478484630584717,
"learning_rate": 6.491634692845781e-06,
"loss": 0.5914,
"step": 145
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.6368155479431152,
"learning_rate": 6.437996637160086e-06,
"loss": 0.639,
"step": 146
},
{
"epoch": 1.4089456869009584,
"grad_norm": 0.5644116401672363,
"learning_rate": 6.384177557124247e-06,
"loss": 0.6451,
"step": 147
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.528681755065918,
"learning_rate": 6.330184227833376e-06,
"loss": 0.6422,
"step": 148
},
{
"epoch": 1.428115015974441,
"grad_norm": 0.688145101070404,
"learning_rate": 6.276023446318214e-06,
"loss": 0.6076,
"step": 149
},
{
"epoch": 1.4376996805111821,
"grad_norm": 0.5362103581428528,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.6037,
"step": 150
},
{
"epoch": 1.4472843450479234,
"grad_norm": 0.6044646501541138,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.6104,
"step": 151
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.5781368017196655,
"learning_rate": 6.112604669781572e-06,
"loss": 0.6232,
"step": 152
},
{
"epoch": 1.4664536741214058,
"grad_norm": 0.5977554321289062,
"learning_rate": 6.057842458386315e-06,
"loss": 0.6188,
"step": 153
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.6424412131309509,
"learning_rate": 6.002947078916365e-06,
"loss": 0.6401,
"step": 154
},
{
"epoch": 1.4856230031948883,
"grad_norm": 0.5800415277481079,
"learning_rate": 5.947925441958393e-06,
"loss": 0.6315,
"step": 155
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.5999455451965332,
"learning_rate": 5.892784473993184e-06,
"loss": 0.6322,
"step": 156
},
{
"epoch": 1.5047923322683707,
"grad_norm": 0.6322799921035767,
"learning_rate": 5.837531116523683e-06,
"loss": 0.6169,
"step": 157
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.636345386505127,
"learning_rate": 5.782172325201155e-06,
"loss": 0.6328,
"step": 158
},
{
"epoch": 1.5239616613418532,
"grad_norm": 0.5558579564094543,
"learning_rate": 5.726715068949564e-06,
"loss": 0.6388,
"step": 159
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.5865455865859985,
"learning_rate": 5.671166329088278e-06,
"loss": 0.6403,
"step": 160
},
{
"epoch": 1.5431309904153354,
"grad_norm": 0.5937369465827942,
"learning_rate": 5.615533098453215e-06,
"loss": 0.5802,
"step": 161
},
{
"epoch": 1.5527156549520766,
"grad_norm": 0.5675464272499084,
"learning_rate": 5.559822380516539e-06,
"loss": 0.616,
"step": 162
},
{
"epoch": 1.5623003194888179,
"grad_norm": 0.5512328147888184,
"learning_rate": 5.504041188505022e-06,
"loss": 0.628,
"step": 163
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.5694289207458496,
"learning_rate": 5.448196544517168e-06,
"loss": 0.595,
"step": 164
},
{
"epoch": 1.5814696485623003,
"grad_norm": 0.5645386576652527,
"learning_rate": 5.392295478639226e-06,
"loss": 0.6081,
"step": 165
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.5716211795806885,
"learning_rate": 5.336345028060199e-06,
"loss": 0.6197,
"step": 166
},
{
"epoch": 1.6006389776357828,
"grad_norm": 0.5254788398742676,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.6151,
"step": 167
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.5892966389656067,
"learning_rate": 5.224324151752575e-06,
"loss": 0.6234,
"step": 168
},
{
"epoch": 1.619808306709265,
"grad_norm": 0.5654058456420898,
"learning_rate": 5.168267827938971e-06,
"loss": 0.602,
"step": 169
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.5221303701400757,
"learning_rate": 5.112190321479026e-06,
"loss": 0.6183,
"step": 170
},
{
"epoch": 1.6389776357827475,
"grad_norm": 0.529788076877594,
"learning_rate": 5.05609869177323e-06,
"loss": 0.5975,
"step": 171
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.6015572547912598,
"learning_rate": 5e-06,
"loss": 0.6185,
"step": 172
},
{
"epoch": 1.65814696485623,
"grad_norm": 0.5127381086349487,
"learning_rate": 4.943901308226771e-06,
"loss": 0.6044,
"step": 173
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.4872056543827057,
"learning_rate": 4.887809678520976e-06,
"loss": 0.6155,
"step": 174
},
{
"epoch": 1.6773162939297124,
"grad_norm": 0.5433346033096313,
"learning_rate": 4.831732172061032e-06,
"loss": 0.6353,
"step": 175
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.524617075920105,
"learning_rate": 4.775675848247427e-06,
"loss": 0.5776,
"step": 176
},
{
"epoch": 1.6964856230031948,
"grad_norm": 0.5380411148071289,
"learning_rate": 4.719647763814041e-06,
"loss": 0.6126,
"step": 177
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.5044443011283875,
"learning_rate": 4.663654971939802e-06,
"loss": 0.6315,
"step": 178
},
{
"epoch": 1.7156549520766773,
"grad_norm": 0.4996054470539093,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.6307,
"step": 179
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.5799446105957031,
"learning_rate": 4.551803455482833e-06,
"loss": 0.6091,
"step": 180
},
{
"epoch": 1.7348242811501597,
"grad_norm": 0.5466322302818298,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.6309,
"step": 181
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.5624096989631653,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.6292,
"step": 182
},
{
"epoch": 1.7539936102236422,
"grad_norm": 0.505291759967804,
"learning_rate": 4.384466901546786e-06,
"loss": 0.6236,
"step": 183
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.5468903183937073,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.6034,
"step": 184
},
{
"epoch": 1.7731629392971247,
"grad_norm": 0.5453299880027771,
"learning_rate": 4.273284931050438e-06,
"loss": 0.6131,
"step": 185
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.5148694515228271,
"learning_rate": 4.217827674798845e-06,
"loss": 0.6064,
"step": 186
},
{
"epoch": 1.792332268370607,
"grad_norm": 0.5098261833190918,
"learning_rate": 4.162468883476319e-06,
"loss": 0.5997,
"step": 187
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.5682370066642761,
"learning_rate": 4.107215526006818e-06,
"loss": 0.6166,
"step": 188
},
{
"epoch": 1.8115015974440896,
"grad_norm": 0.508726954460144,
"learning_rate": 4.052074558041608e-06,
"loss": 0.6436,
"step": 189
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.5360882878303528,
"learning_rate": 3.997052921083637e-06,
"loss": 0.6196,
"step": 190
},
{
"epoch": 1.830670926517572,
"grad_norm": 0.525092601776123,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.607,
"step": 191
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.5473348498344421,
"learning_rate": 3.887395330218429e-06,
"loss": 0.6445,
"step": 192
},
{
"epoch": 1.8498402555910545,
"grad_norm": 0.5632632970809937,
"learning_rate": 3.832773180720475e-06,
"loss": 0.6155,
"step": 193
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.4716745913028717,
"learning_rate": 3.778297969310529e-06,
"loss": 0.6129,
"step": 194
},
{
"epoch": 1.8690095846645367,
"grad_norm": 0.5051786303520203,
"learning_rate": 3.723976553681787e-06,
"loss": 0.6286,
"step": 195
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.5381728410720825,
"learning_rate": 3.669815772166625e-06,
"loss": 0.6303,
"step": 196
},
{
"epoch": 1.8881789137380192,
"grad_norm": 0.592290461063385,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.6478,
"step": 197
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.5568681955337524,
"learning_rate": 3.562003362839914e-06,
"loss": 0.6248,
"step": 198
},
{
"epoch": 1.9073482428115016,
"grad_norm": 0.4643682837486267,
"learning_rate": 3.50836530715422e-06,
"loss": 0.6055,
"step": 199
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.5361279845237732,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.6268,
"step": 200
},
{
"epoch": 1.926517571884984,
"grad_norm": 0.556638777256012,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.6233,
"step": 201
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.4710752069950104,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.6137,
"step": 202
},
{
"epoch": 1.9456869009584663,
"grad_norm": 0.4985019862651825,
"learning_rate": 3.295758014387375e-06,
"loss": 0.6272,
"step": 203
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.4872068762779236,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.6228,
"step": 204
},
{
"epoch": 1.9648562300319488,
"grad_norm": 0.48627588152885437,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.6351,
"step": 205
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.500870943069458,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.6283,
"step": 206
},
{
"epoch": 1.9840255591054312,
"grad_norm": 0.5739912390708923,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.6382,
"step": 207
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.5555146336555481,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.6378,
"step": 208
},
{
"epoch": 2.0031948881789137,
"grad_norm": 0.5773714184761047,
"learning_rate": 2.9834142279959754e-06,
"loss": 0.5957,
"step": 209
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.5471380352973938,
"learning_rate": 2.932207475167398e-06,
"loss": 0.5543,
"step": 210
},
{
"epoch": 2.022364217252396,
"grad_norm": 0.591244101524353,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.5219,
"step": 211
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.4896009862422943,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5912,
"step": 212
},
{
"epoch": 2.0415335463258786,
"grad_norm": 0.5137360692024231,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.5524,
"step": 213
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.5738446712493896,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.5741,
"step": 214
},
{
"epoch": 2.060702875399361,
"grad_norm": 0.5873703956604004,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.5675,
"step": 215
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.5568045973777771,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5689,
"step": 216
},
{
"epoch": 2.0798722044728435,
"grad_norm": 0.5483757257461548,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.5584,
"step": 217
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.5321773886680603,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5491,
"step": 218
},
{
"epoch": 2.099041533546326,
"grad_norm": 0.5125218033790588,
"learning_rate": 2.483822884971e-06,
"loss": 0.5557,
"step": 219
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.5159919261932373,
"learning_rate": 2.43550361297047e-06,
"loss": 0.542,
"step": 220
},
{
"epoch": 2.1182108626198084,
"grad_norm": 0.5942448973655701,
"learning_rate": 2.387507176420256e-06,
"loss": 0.5659,
"step": 221
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.5507439970970154,
"learning_rate": 2.339839617423318e-06,
"loss": 0.5651,
"step": 222
},
{
"epoch": 2.137380191693291,
"grad_norm": 0.46924686431884766,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.5664,
"step": 223
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.4597823917865753,
"learning_rate": 2.245515092739488e-06,
"loss": 0.5531,
"step": 224
},
{
"epoch": 2.1565495207667733,
"grad_norm": 0.483770489692688,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.5451,
"step": 225
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.5301675796508789,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.5948,
"step": 226
},
{
"epoch": 2.1757188498402558,
"grad_norm": 0.5276245474815369,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.5388,
"step": 227
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.4675266742706299,
"learning_rate": 2.061073738537635e-06,
"loss": 0.5688,
"step": 228
},
{
"epoch": 2.194888178913738,
"grad_norm": 0.46537959575653076,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.5402,
"step": 229
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.49672767519950867,
"learning_rate": 1.971049780795901e-06,
"loss": 0.5547,
"step": 230
},
{
"epoch": 2.2140575079872207,
"grad_norm": 0.42004460096359253,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.5418,
"step": 231
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.492191344499588,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.5807,
"step": 232
},
{
"epoch": 2.2332268370607027,
"grad_norm": 0.5344516634941101,
"learning_rate": 1.838887489710452e-06,
"loss": 0.552,
"step": 233
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.4998222291469574,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.5445,
"step": 234
},
{
"epoch": 2.252396166134185,
"grad_norm": 0.48806169629096985,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.5778,
"step": 235
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.46689218282699585,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.5729,
"step": 236
},
{
"epoch": 2.2715654952076676,
"grad_norm": 0.46440935134887695,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.5462,
"step": 237
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.44381314516067505,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.5515,
"step": 238
},
{
"epoch": 2.29073482428115,
"grad_norm": 0.48730823397636414,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.5757,
"step": 239
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.4613868296146393,
"learning_rate": 1.544686755065677e-06,
"loss": 0.5553,
"step": 240
},
{
"epoch": 2.3099041533546325,
"grad_norm": 0.4681055545806885,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.5622,
"step": 241
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.4431508183479309,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5657,
"step": 242
},
{
"epoch": 2.329073482428115,
"grad_norm": 0.46562695503234863,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.566,
"step": 243
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.46384090185165405,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.5345,
"step": 244
},
{
"epoch": 2.3482428115015974,
"grad_norm": 0.4555598497390747,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.5624,
"step": 245
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.47274306416511536,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.5501,
"step": 246
},
{
"epoch": 2.36741214057508,
"grad_norm": 0.47570404410362244,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.5287,
"step": 247
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.4742777645587921,
"learning_rate": 1.234642669981946e-06,
"loss": 0.5642,
"step": 248
},
{
"epoch": 2.3865814696485623,
"grad_norm": 0.4669046103954315,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.5577,
"step": 249
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.4769456386566162,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.5517,
"step": 250
},
{
"epoch": 2.405750798722045,
"grad_norm": 0.437163382768631,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.5542,
"step": 251
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.4794774353504181,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5176,
"step": 252
},
{
"epoch": 2.4249201277955272,
"grad_norm": 0.467252254486084,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.5728,
"step": 253
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.44577717781066895,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.5647,
"step": 254
},
{
"epoch": 2.4440894568690097,
"grad_norm": 0.46085208654403687,
"learning_rate": 9.881436225981107e-07,
"loss": 0.5573,
"step": 255
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.44824862480163574,
"learning_rate": 9.549150281252633e-07,
"loss": 0.5472,
"step": 256
},
{
"epoch": 2.463258785942492,
"grad_norm": 0.46623456478118896,
"learning_rate": 9.221956552036992e-07,
"loss": 0.5273,
"step": 257
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.45707541704177856,
"learning_rate": 8.899896227604509e-07,
"loss": 0.5589,
"step": 258
},
{
"epoch": 2.4824281150159746,
"grad_norm": 0.4400307536125183,
"learning_rate": 8.58300985099918e-07,
"loss": 0.5539,
"step": 259
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.4513288140296936,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5574,
"step": 260
},
{
"epoch": 2.501597444089457,
"grad_norm": 0.5023139715194702,
"learning_rate": 7.964917851773496e-07,
"loss": 0.5445,
"step": 261
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.4401370882987976,
"learning_rate": 7.663790038585794e-07,
"loss": 0.5551,
"step": 262
},
{
"epoch": 2.520766773162939,
"grad_norm": 0.46186554431915283,
"learning_rate": 7.367991782295392e-07,
"loss": 0.5393,
"step": 263
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.46787843108177185,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5446,
"step": 264
},
{
"epoch": 2.539936102236422,
"grad_norm": 0.45983678102493286,
"learning_rate": 6.792532212817271e-07,
"loss": 0.5448,
"step": 265
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.4354945123195648,
"learning_rate": 6.512943342215234e-07,
"loss": 0.5516,
"step": 266
},
{
"epoch": 2.559105431309904,
"grad_norm": 0.45350924134254456,
"learning_rate": 6.238828904562316e-07,
"loss": 0.5425,
"step": 267
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.46408724784851074,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5516,
"step": 268
},
{
"epoch": 2.5782747603833864,
"grad_norm": 0.43403956294059753,
"learning_rate": 5.707160663821009e-07,
"loss": 0.59,
"step": 269
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.4258999824523926,
"learning_rate": 5.449673790581611e-07,
"loss": 0.5529,
"step": 270
},
{
"epoch": 2.597444089456869,
"grad_norm": 0.4422466456890106,
"learning_rate": 5.197795201563744e-07,
"loss": 0.5515,
"step": 271
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.4681154787540436,
"learning_rate": 4.951556604879049e-07,
"loss": 0.5414,
"step": 272
},
{
"epoch": 2.6166134185303513,
"grad_norm": 0.4460196793079376,
"learning_rate": 4.710988998640298e-07,
"loss": 0.5559,
"step": 273
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.4309646487236023,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5504,
"step": 274
},
{
"epoch": 2.635782747603834,
"grad_norm": 0.4520484209060669,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.5631,
"step": 275
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.4719311594963074,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5441,
"step": 276
},
{
"epoch": 2.6549520766773163,
"grad_norm": 0.4278734028339386,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.5366,
"step": 277
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.4238326847553253,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5518,
"step": 278
},
{
"epoch": 2.6741214057507987,
"grad_norm": 0.4466708302497864,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.5395,
"step": 279
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.49409493803977966,
"learning_rate": 3.18825646801314e-07,
"loss": 0.572,
"step": 280
},
{
"epoch": 2.693290734824281,
"grad_norm": 0.4626609981060028,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.5655,
"step": 281
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.44110432267189026,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5498,
"step": 282
},
{
"epoch": 2.7124600638977636,
"grad_norm": 0.4938991069793701,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.5612,
"step": 283
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.4392610788345337,
"learning_rate": 2.447174185242324e-07,
"loss": 0.5464,
"step": 284
},
{
"epoch": 2.731629392971246,
"grad_norm": 0.4647540748119354,
"learning_rate": 2.276812823220964e-07,
"loss": 0.5509,
"step": 285
},
{
"epoch": 2.741214057507987,
"grad_norm": 0.4366726279258728,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.522,
"step": 286
},
{
"epoch": 2.7507987220447285,
"grad_norm": 0.428124338388443,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.5395,
"step": 287
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.42290928959846497,
"learning_rate": 1.801856965207338e-07,
"loss": 0.5207,
"step": 288
},
{
"epoch": 2.769968051118211,
"grad_norm": 0.40466392040252686,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.5458,
"step": 289
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.45290449261665344,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5622,
"step": 290
},
{
"epoch": 2.7891373801916934,
"grad_norm": 0.4516129493713379,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.5536,
"step": 291
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.4378257691860199,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5579,
"step": 292
},
{
"epoch": 2.8083067092651754,
"grad_norm": 0.42467889189720154,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.5534,
"step": 293
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.44925180077552795,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.5344,
"step": 294
},
{
"epoch": 2.8274760383386583,
"grad_norm": 0.4583977460861206,
"learning_rate": 9.0678523819408e-08,
"loss": 0.5555,
"step": 295
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.4300501346588135,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5485,
"step": 296
},
{
"epoch": 2.8466453674121404,
"grad_norm": 0.4508187174797058,
"learning_rate": 7.064490740882057e-08,
"loss": 0.5483,
"step": 297
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.4076024293899536,
"learning_rate": 6.15582970243117e-08,
"loss": 0.5483,
"step": 298
},
{
"epoch": 2.8658146964856233,
"grad_norm": 0.4347573518753052,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.5579,
"step": 299
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.45431503653526306,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5615,
"step": 300
},
{
"epoch": 2.8849840255591053,
"grad_norm": 0.4543002247810364,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.5384,
"step": 301
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.4596332609653473,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5513,
"step": 302
},
{
"epoch": 2.9041533546325877,
"grad_norm": 0.4447740912437439,
"learning_rate": 2.547062725623828e-08,
"loss": 0.5386,
"step": 303
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.44638144969940186,
"learning_rate": 2.012853002380466e-08,
"loss": 0.5376,
"step": 304
},
{
"epoch": 2.92332268370607,
"grad_norm": 0.4340687692165375,
"learning_rate": 1.541333133436018e-08,
"loss": 0.5614,
"step": 305
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.4377923607826233,
"learning_rate": 1.132562476771959e-08,
"loss": 0.5728,
"step": 306
},
{
"epoch": 2.9424920127795526,
"grad_norm": 0.4357081651687622,
"learning_rate": 7.865924910916977e-09,
"loss": 0.5444,
"step": 307
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.44313573837280273,
"learning_rate": 5.034667293427053e-09,
"loss": 0.5336,
"step": 308
},
{
"epoch": 2.961661341853035,
"grad_norm": 0.4313756823539734,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.5387,
"step": 309
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.456178218126297,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5226,
"step": 310
},
{
"epoch": 2.9808306709265175,
"grad_norm": 0.42373061180114746,
"learning_rate": 3.147162264971471e-10,
"loss": 0.5582,
"step": 311
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.4256809949874878,
"learning_rate": 0.0,
"loss": 0.5826,
"step": 312
},
{
"epoch": 2.9904153354632586,
"step": 312,
"total_flos": 4.235769232978084e+17,
"train_loss": 0.0,
"train_runtime": 5.544,
"train_samples_per_second": 5411.272,
"train_steps_per_second": 56.277
}
],
"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": 4.235769232978084e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}