DCFT-seed_code_r1_dolphin-etash / trainer_state.json
EtashGuha's picture
Upload folder using huggingface_hub
0d05319 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": 6.745182514190674,
"learning_rate": 3.125e-07,
"loss": 1.0197,
"step": 1
},
{
"epoch": 0.019169329073482427,
"grad_norm": 6.926332950592041,
"learning_rate": 6.25e-07,
"loss": 1.0345,
"step": 2
},
{
"epoch": 0.02875399361022364,
"grad_norm": 6.893532752990723,
"learning_rate": 9.375000000000001e-07,
"loss": 1.0388,
"step": 3
},
{
"epoch": 0.038338658146964855,
"grad_norm": 6.804474830627441,
"learning_rate": 1.25e-06,
"loss": 1.0129,
"step": 4
},
{
"epoch": 0.04792332268370607,
"grad_norm": 6.523636341094971,
"learning_rate": 1.5625e-06,
"loss": 1.0207,
"step": 5
},
{
"epoch": 0.05750798722044728,
"grad_norm": 6.256287097930908,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.0195,
"step": 6
},
{
"epoch": 0.0670926517571885,
"grad_norm": 5.281290054321289,
"learning_rate": 2.1875000000000002e-06,
"loss": 0.98,
"step": 7
},
{
"epoch": 0.07667731629392971,
"grad_norm": 4.773669242858887,
"learning_rate": 2.5e-06,
"loss": 0.955,
"step": 8
},
{
"epoch": 0.08626198083067092,
"grad_norm": 3.1332437992095947,
"learning_rate": 2.8125e-06,
"loss": 0.9111,
"step": 9
},
{
"epoch": 0.09584664536741214,
"grad_norm": 3.073474884033203,
"learning_rate": 3.125e-06,
"loss": 0.9047,
"step": 10
},
{
"epoch": 0.10543130990415335,
"grad_norm": 2.786947727203369,
"learning_rate": 3.4375e-06,
"loss": 0.9101,
"step": 11
},
{
"epoch": 0.11501597444089456,
"grad_norm": 3.666496753692627,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.8633,
"step": 12
},
{
"epoch": 0.12460063897763578,
"grad_norm": 3.797808885574341,
"learning_rate": 4.0625000000000005e-06,
"loss": 0.8642,
"step": 13
},
{
"epoch": 0.134185303514377,
"grad_norm": 3.643026113510132,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.8608,
"step": 14
},
{
"epoch": 0.14376996805111822,
"grad_norm": 3.2613444328308105,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.8245,
"step": 15
},
{
"epoch": 0.15335463258785942,
"grad_norm": 2.21720814704895,
"learning_rate": 5e-06,
"loss": 0.7958,
"step": 16
},
{
"epoch": 0.16293929712460065,
"grad_norm": 2.314464569091797,
"learning_rate": 5.3125e-06,
"loss": 0.7807,
"step": 17
},
{
"epoch": 0.17252396166134185,
"grad_norm": 1.8976560831069946,
"learning_rate": 5.625e-06,
"loss": 0.7763,
"step": 18
},
{
"epoch": 0.18210862619808307,
"grad_norm": 1.6641546487808228,
"learning_rate": 5.9375e-06,
"loss": 0.7623,
"step": 19
},
{
"epoch": 0.19169329073482427,
"grad_norm": 1.2857623100280762,
"learning_rate": 6.25e-06,
"loss": 0.7322,
"step": 20
},
{
"epoch": 0.2012779552715655,
"grad_norm": 1.1176934242248535,
"learning_rate": 6.5625e-06,
"loss": 0.7232,
"step": 21
},
{
"epoch": 0.2108626198083067,
"grad_norm": 1.29485023021698,
"learning_rate": 6.875e-06,
"loss": 0.7271,
"step": 22
},
{
"epoch": 0.22044728434504793,
"grad_norm": 1.2421157360076904,
"learning_rate": 7.1875e-06,
"loss": 0.7155,
"step": 23
},
{
"epoch": 0.23003194888178913,
"grad_norm": 0.9793645739555359,
"learning_rate": 7.500000000000001e-06,
"loss": 0.6955,
"step": 24
},
{
"epoch": 0.23961661341853036,
"grad_norm": 0.8572778105735779,
"learning_rate": 7.8125e-06,
"loss": 0.6696,
"step": 25
},
{
"epoch": 0.24920127795527156,
"grad_norm": 1.070764183998108,
"learning_rate": 8.125000000000001e-06,
"loss": 0.6686,
"step": 26
},
{
"epoch": 0.25878594249201275,
"grad_norm": 1.0625317096710205,
"learning_rate": 8.4375e-06,
"loss": 0.683,
"step": 27
},
{
"epoch": 0.268370607028754,
"grad_norm": 0.969550371170044,
"learning_rate": 8.750000000000001e-06,
"loss": 0.6708,
"step": 28
},
{
"epoch": 0.2779552715654952,
"grad_norm": 0.8943290114402771,
"learning_rate": 9.0625e-06,
"loss": 0.6791,
"step": 29
},
{
"epoch": 0.28753993610223644,
"grad_norm": 0.8745009899139404,
"learning_rate": 9.375000000000001e-06,
"loss": 0.6546,
"step": 30
},
{
"epoch": 0.2971246006389776,
"grad_norm": 0.9032918810844421,
"learning_rate": 9.6875e-06,
"loss": 0.6528,
"step": 31
},
{
"epoch": 0.30670926517571884,
"grad_norm": 0.7913438677787781,
"learning_rate": 1e-05,
"loss": 0.6427,
"step": 32
},
{
"epoch": 0.31629392971246006,
"grad_norm": 0.7979699969291687,
"learning_rate": 9.999685283773504e-06,
"loss": 0.6365,
"step": 33
},
{
"epoch": 0.3258785942492013,
"grad_norm": 0.7229610085487366,
"learning_rate": 9.998741174712534e-06,
"loss": 0.6333,
"step": 34
},
{
"epoch": 0.3354632587859425,
"grad_norm": 0.7657011151313782,
"learning_rate": 9.997167791667668e-06,
"loss": 0.6323,
"step": 35
},
{
"epoch": 0.3450479233226837,
"grad_norm": 0.679976224899292,
"learning_rate": 9.994965332706574e-06,
"loss": 0.6316,
"step": 36
},
{
"epoch": 0.3546325878594249,
"grad_norm": 0.7799426913261414,
"learning_rate": 9.992134075089085e-06,
"loss": 0.6385,
"step": 37
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.7320380806922913,
"learning_rate": 9.98867437523228e-06,
"loss": 0.6413,
"step": 38
},
{
"epoch": 0.3738019169329074,
"grad_norm": 0.7033889889717102,
"learning_rate": 9.984586668665641e-06,
"loss": 0.623,
"step": 39
},
{
"epoch": 0.38338658146964855,
"grad_norm": 0.7482995986938477,
"learning_rate": 9.979871469976197e-06,
"loss": 0.6216,
"step": 40
},
{
"epoch": 0.3929712460063898,
"grad_norm": 0.6804895997047424,
"learning_rate": 9.974529372743762e-06,
"loss": 0.6087,
"step": 41
},
{
"epoch": 0.402555910543131,
"grad_norm": 0.704346776008606,
"learning_rate": 9.968561049466214e-06,
"loss": 0.6115,
"step": 42
},
{
"epoch": 0.41214057507987223,
"grad_norm": 0.643164336681366,
"learning_rate": 9.961967251474823e-06,
"loss": 0.6133,
"step": 43
},
{
"epoch": 0.4217252396166134,
"grad_norm": 0.6266957521438599,
"learning_rate": 9.954748808839675e-06,
"loss": 0.607,
"step": 44
},
{
"epoch": 0.43130990415335463,
"grad_norm": 0.6352761387825012,
"learning_rate": 9.946906630265184e-06,
"loss": 0.6031,
"step": 45
},
{
"epoch": 0.44089456869009586,
"grad_norm": 0.6505466103553772,
"learning_rate": 9.938441702975689e-06,
"loss": 0.6061,
"step": 46
},
{
"epoch": 0.4504792332268371,
"grad_norm": 0.5911573767662048,
"learning_rate": 9.92935509259118e-06,
"loss": 0.6129,
"step": 47
},
{
"epoch": 0.46006389776357826,
"grad_norm": 0.6804027557373047,
"learning_rate": 9.91964794299315e-06,
"loss": 0.5966,
"step": 48
},
{
"epoch": 0.4696485623003195,
"grad_norm": 0.6392555832862854,
"learning_rate": 9.909321476180594e-06,
"loss": 0.5947,
"step": 49
},
{
"epoch": 0.4792332268370607,
"grad_norm": 0.6037166714668274,
"learning_rate": 9.898376992116179e-06,
"loss": 0.5952,
"step": 50
},
{
"epoch": 0.48881789137380194,
"grad_norm": 0.6624611616134644,
"learning_rate": 9.886815868562596e-06,
"loss": 0.5787,
"step": 51
},
{
"epoch": 0.4984025559105431,
"grad_norm": 0.622444748878479,
"learning_rate": 9.874639560909118e-06,
"loss": 0.5885,
"step": 52
},
{
"epoch": 0.5079872204472844,
"grad_norm": 0.5764785408973694,
"learning_rate": 9.861849601988384e-06,
"loss": 0.5747,
"step": 53
},
{
"epoch": 0.5175718849840255,
"grad_norm": 0.6422050595283508,
"learning_rate": 9.848447601883436e-06,
"loss": 0.5914,
"step": 54
},
{
"epoch": 0.5271565495207667,
"grad_norm": 0.5677855014801025,
"learning_rate": 9.834435247725032e-06,
"loss": 0.5803,
"step": 55
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.5909940600395203,
"learning_rate": 9.819814303479268e-06,
"loss": 0.5877,
"step": 56
},
{
"epoch": 0.5463258785942492,
"grad_norm": 0.6010948419570923,
"learning_rate": 9.804586609725499e-06,
"loss": 0.5957,
"step": 57
},
{
"epoch": 0.5559105431309904,
"grad_norm": 0.5955870747566223,
"learning_rate": 9.788754083424654e-06,
"loss": 0.5852,
"step": 58
},
{
"epoch": 0.5654952076677316,
"grad_norm": 0.6237753629684448,
"learning_rate": 9.772318717677905e-06,
"loss": 0.5837,
"step": 59
},
{
"epoch": 0.5750798722044729,
"grad_norm": 0.5781514048576355,
"learning_rate": 9.755282581475769e-06,
"loss": 0.5702,
"step": 60
},
{
"epoch": 0.5846645367412141,
"grad_norm": 0.5219152569770813,
"learning_rate": 9.737647819437645e-06,
"loss": 0.5819,
"step": 61
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.5861315131187439,
"learning_rate": 9.719416651541839e-06,
"loss": 0.5762,
"step": 62
},
{
"epoch": 0.6038338658146964,
"grad_norm": 0.593108057975769,
"learning_rate": 9.700591372846096e-06,
"loss": 0.5667,
"step": 63
},
{
"epoch": 0.6134185303514377,
"grad_norm": 0.5327998399734497,
"learning_rate": 9.681174353198687e-06,
"loss": 0.5737,
"step": 64
},
{
"epoch": 0.6230031948881789,
"grad_norm": 0.6368751525878906,
"learning_rate": 9.661168036940071e-06,
"loss": 0.5866,
"step": 65
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.5311642289161682,
"learning_rate": 9.640574942595195e-06,
"loss": 0.5774,
"step": 66
},
{
"epoch": 0.6421725239616614,
"grad_norm": 0.6101247668266296,
"learning_rate": 9.619397662556434e-06,
"loss": 0.5888,
"step": 67
},
{
"epoch": 0.6517571884984026,
"grad_norm": 0.7286245822906494,
"learning_rate": 9.597638862757255e-06,
"loss": 0.5656,
"step": 68
},
{
"epoch": 0.6613418530351438,
"grad_norm": 0.6497251391410828,
"learning_rate": 9.5753012823366e-06,
"loss": 0.5572,
"step": 69
},
{
"epoch": 0.670926517571885,
"grad_norm": 0.5950864553451538,
"learning_rate": 9.552387733294081e-06,
"loss": 0.5692,
"step": 70
},
{
"epoch": 0.6805111821086262,
"grad_norm": 0.6206063032150269,
"learning_rate": 9.528901100135971e-06,
"loss": 0.58,
"step": 71
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.57789146900177,
"learning_rate": 9.504844339512096e-06,
"loss": 0.5519,
"step": 72
},
{
"epoch": 0.6996805111821086,
"grad_norm": 0.6747218370437622,
"learning_rate": 9.480220479843627e-06,
"loss": 0.5543,
"step": 73
},
{
"epoch": 0.7092651757188498,
"grad_norm": 0.6099786162376404,
"learning_rate": 9.45503262094184e-06,
"loss": 0.5566,
"step": 74
},
{
"epoch": 0.7188498402555911,
"grad_norm": 0.6386275887489319,
"learning_rate": 9.4292839336179e-06,
"loss": 0.5568,
"step": 75
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.6355255246162415,
"learning_rate": 9.40297765928369e-06,
"loss": 0.5563,
"step": 76
},
{
"epoch": 0.7380191693290735,
"grad_norm": 0.640434980392456,
"learning_rate": 9.376117109543769e-06,
"loss": 0.5598,
"step": 77
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.5761203765869141,
"learning_rate": 9.348705665778479e-06,
"loss": 0.5623,
"step": 78
},
{
"epoch": 0.7571884984025559,
"grad_norm": 0.743889570236206,
"learning_rate": 9.320746778718274e-06,
"loss": 0.5507,
"step": 79
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.5959399938583374,
"learning_rate": 9.292243968009332e-06,
"loss": 0.57,
"step": 80
},
{
"epoch": 0.7763578274760383,
"grad_norm": 0.7095223665237427,
"learning_rate": 9.263200821770462e-06,
"loss": 0.5797,
"step": 81
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.5428198575973511,
"learning_rate": 9.233620996141421e-06,
"loss": 0.5489,
"step": 82
},
{
"epoch": 0.7955271565495208,
"grad_norm": 0.7187327742576599,
"learning_rate": 9.203508214822652e-06,
"loss": 0.5551,
"step": 83
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.5488038659095764,
"learning_rate": 9.172866268606514e-06,
"loss": 0.5685,
"step": 84
},
{
"epoch": 0.8146964856230032,
"grad_norm": 0.6017353534698486,
"learning_rate": 9.141699014900084e-06,
"loss": 0.5435,
"step": 85
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.6012035608291626,
"learning_rate": 9.110010377239552e-06,
"loss": 0.5726,
"step": 86
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.7593209743499756,
"learning_rate": 9.077804344796302e-06,
"loss": 0.5456,
"step": 87
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.5839484930038452,
"learning_rate": 9.045084971874738e-06,
"loss": 0.5466,
"step": 88
},
{
"epoch": 0.853035143769968,
"grad_norm": 0.7215870022773743,
"learning_rate": 9.011856377401891e-06,
"loss": 0.572,
"step": 89
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.598135232925415,
"learning_rate": 8.978122744408905e-06,
"loss": 0.556,
"step": 90
},
{
"epoch": 0.8722044728434505,
"grad_norm": 0.6009616255760193,
"learning_rate": 8.943888319504456e-06,
"loss": 0.5587,
"step": 91
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.5611772537231445,
"learning_rate": 8.90915741234015e-06,
"loss": 0.5532,
"step": 92
},
{
"epoch": 0.8913738019169329,
"grad_norm": 0.5573892593383789,
"learning_rate": 8.873934395068006e-06,
"loss": 0.5356,
"step": 93
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.5890476703643799,
"learning_rate": 8.838223701790057e-06,
"loss": 0.5453,
"step": 94
},
{
"epoch": 0.9105431309904153,
"grad_norm": 0.5717616081237793,
"learning_rate": 8.802029828000157e-06,
"loss": 0.5279,
"step": 95
},
{
"epoch": 0.9201277955271565,
"grad_norm": 0.5752805471420288,
"learning_rate": 8.765357330018056e-06,
"loss": 0.568,
"step": 96
},
{
"epoch": 0.9297124600638977,
"grad_norm": 0.7040980458259583,
"learning_rate": 8.728210824415829e-06,
"loss": 0.5562,
"step": 97
},
{
"epoch": 0.939297124600639,
"grad_norm": 0.6557633876800537,
"learning_rate": 8.690594987436705e-06,
"loss": 0.542,
"step": 98
},
{
"epoch": 0.9488817891373802,
"grad_norm": 0.6837477087974548,
"learning_rate": 8.652514554406388e-06,
"loss": 0.5379,
"step": 99
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.7009369134902954,
"learning_rate": 8.613974319136959e-06,
"loss": 0.5457,
"step": 100
},
{
"epoch": 0.9680511182108626,
"grad_norm": 0.5840373635292053,
"learning_rate": 8.574979133323378e-06,
"loss": 0.5608,
"step": 101
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.7043083310127258,
"learning_rate": 8.535533905932739e-06,
"loss": 0.5578,
"step": 102
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.6753775477409363,
"learning_rate": 8.495643602586287e-06,
"loss": 0.5343,
"step": 103
},
{
"epoch": 0.9968051118210862,
"grad_norm": 0.6212782859802246,
"learning_rate": 8.455313244934324e-06,
"loss": 0.5412,
"step": 104
},
{
"epoch": 1.0063897763578276,
"grad_norm": 0.6647524833679199,
"learning_rate": 8.414547910024035e-06,
"loss": 0.5242,
"step": 105
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.7002506256103516,
"learning_rate": 8.373352729660373e-06,
"loss": 0.5033,
"step": 106
},
{
"epoch": 1.0255591054313098,
"grad_norm": 0.6339381337165833,
"learning_rate": 8.331732889760021e-06,
"loss": 0.5196,
"step": 107
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.7434774041175842,
"learning_rate": 8.289693629698564e-06,
"loss": 0.4966,
"step": 108
},
{
"epoch": 1.0447284345047922,
"grad_norm": 0.6625819206237793,
"learning_rate": 8.247240241650918e-06,
"loss": 0.5042,
"step": 109
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.7946425676345825,
"learning_rate": 8.204378069925121e-06,
"loss": 0.5086,
"step": 110
},
{
"epoch": 1.0638977635782747,
"grad_norm": 0.5590626001358032,
"learning_rate": 8.16111251028955e-06,
"loss": 0.5058,
"step": 111
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.7147952914237976,
"learning_rate": 8.117449009293668e-06,
"loss": 0.4984,
"step": 112
},
{
"epoch": 1.0830670926517572,
"grad_norm": 0.6046099066734314,
"learning_rate": 8.073393063582386e-06,
"loss": 0.4951,
"step": 113
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.6469084620475769,
"learning_rate": 8.0289502192041e-06,
"loss": 0.5,
"step": 114
},
{
"epoch": 1.1022364217252396,
"grad_norm": 0.6059616804122925,
"learning_rate": 7.984126070912519e-06,
"loss": 0.4995,
"step": 115
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.59597247838974,
"learning_rate": 7.938926261462366e-06,
"loss": 0.5095,
"step": 116
},
{
"epoch": 1.121405750798722,
"grad_norm": 0.6475273966789246,
"learning_rate": 7.89335648089903e-06,
"loss": 0.5092,
"step": 117
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.5431929230690002,
"learning_rate": 7.84742246584226e-06,
"loss": 0.5141,
"step": 118
},
{
"epoch": 1.1405750798722045,
"grad_norm": 0.5373610258102417,
"learning_rate": 7.801129998764014e-06,
"loss": 0.4891,
"step": 119
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.5813199281692505,
"learning_rate": 7.754484907260513e-06,
"loss": 0.5029,
"step": 120
},
{
"epoch": 1.159744408945687,
"grad_norm": 0.6003910899162292,
"learning_rate": 7.70749306331863e-06,
"loss": 0.504,
"step": 121
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.5305626392364502,
"learning_rate": 7.660160382576683e-06,
"loss": 0.4885,
"step": 122
},
{
"epoch": 1.1789137380191694,
"grad_norm": 0.5256359577178955,
"learning_rate": 7.612492823579744e-06,
"loss": 0.4901,
"step": 123
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.5116536617279053,
"learning_rate": 7.564496387029532e-06,
"loss": 0.5007,
"step": 124
},
{
"epoch": 1.1980830670926517,
"grad_norm": 0.5717402100563049,
"learning_rate": 7.516177115029002e-06,
"loss": 0.5101,
"step": 125
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.5715065598487854,
"learning_rate": 7.467541090321735e-06,
"loss": 0.5093,
"step": 126
},
{
"epoch": 1.2172523961661341,
"grad_norm": 0.5293534994125366,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.493,
"step": 127
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.5703372955322266,
"learning_rate": 7.369343312364994e-06,
"loss": 0.4963,
"step": 128
},
{
"epoch": 1.2364217252396166,
"grad_norm": 0.5871374011039734,
"learning_rate": 7.319793920889171e-06,
"loss": 0.5005,
"step": 129
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.5227846503257751,
"learning_rate": 7.269952498697734e-06,
"loss": 0.4944,
"step": 130
},
{
"epoch": 1.255591054313099,
"grad_norm": 0.5342523455619812,
"learning_rate": 7.219825320152411e-06,
"loss": 0.4938,
"step": 131
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.574222207069397,
"learning_rate": 7.169418695587791e-06,
"loss": 0.5043,
"step": 132
},
{
"epoch": 1.2747603833865815,
"grad_norm": 0.5611140131950378,
"learning_rate": 7.118738970516944e-06,
"loss": 0.5067,
"step": 133
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.530465841293335,
"learning_rate": 7.067792524832604e-06,
"loss": 0.4927,
"step": 134
},
{
"epoch": 1.293929712460064,
"grad_norm": 0.5380294919013977,
"learning_rate": 7.016585772004026e-06,
"loss": 0.4878,
"step": 135
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.574145495891571,
"learning_rate": 6.965125158269619e-06,
"loss": 0.4809,
"step": 136
},
{
"epoch": 1.3130990415335464,
"grad_norm": 0.5581791996955872,
"learning_rate": 6.913417161825449e-06,
"loss": 0.4915,
"step": 137
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.6193375587463379,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.5092,
"step": 138
},
{
"epoch": 1.3322683706070286,
"grad_norm": 0.5523847341537476,
"learning_rate": 6.809285088483361e-06,
"loss": 0.4902,
"step": 139
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.504702627658844,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.5055,
"step": 140
},
{
"epoch": 1.351437699680511,
"grad_norm": 0.6191173195838928,
"learning_rate": 6.704241985612625e-06,
"loss": 0.4972,
"step": 141
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.5755623579025269,
"learning_rate": 6.651395309775837e-06,
"loss": 0.5047,
"step": 142
},
{
"epoch": 1.3706070287539935,
"grad_norm": 0.5373111963272095,
"learning_rate": 6.598340745578908e-06,
"loss": 0.5026,
"step": 143
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.5656738877296448,
"learning_rate": 6.545084971874738e-06,
"loss": 0.4965,
"step": 144
},
{
"epoch": 1.389776357827476,
"grad_norm": 0.5607209205627441,
"learning_rate": 6.491634692845781e-06,
"loss": 0.5003,
"step": 145
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.5409402251243591,
"learning_rate": 6.437996637160086e-06,
"loss": 0.5058,
"step": 146
},
{
"epoch": 1.4089456869009584,
"grad_norm": 0.5648840665817261,
"learning_rate": 6.384177557124247e-06,
"loss": 0.5048,
"step": 147
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.50716632604599,
"learning_rate": 6.330184227833376e-06,
"loss": 0.4916,
"step": 148
},
{
"epoch": 1.428115015974441,
"grad_norm": 0.5097234845161438,
"learning_rate": 6.276023446318214e-06,
"loss": 0.4964,
"step": 149
},
{
"epoch": 1.4376996805111821,
"grad_norm": 0.5181201100349426,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.4916,
"step": 150
},
{
"epoch": 1.4472843450479234,
"grad_norm": 0.4786604344844818,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.4972,
"step": 151
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.5175197720527649,
"learning_rate": 6.112604669781572e-06,
"loss": 0.4903,
"step": 152
},
{
"epoch": 1.4664536741214058,
"grad_norm": 0.5647599697113037,
"learning_rate": 6.057842458386315e-06,
"loss": 0.4844,
"step": 153
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.49329572916030884,
"learning_rate": 6.002947078916365e-06,
"loss": 0.4791,
"step": 154
},
{
"epoch": 1.4856230031948883,
"grad_norm": 0.4385557472705841,
"learning_rate": 5.947925441958393e-06,
"loss": 0.4912,
"step": 155
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.5304051637649536,
"learning_rate": 5.892784473993184e-06,
"loss": 0.5113,
"step": 156
},
{
"epoch": 1.5047923322683707,
"grad_norm": 0.5395846962928772,
"learning_rate": 5.837531116523683e-06,
"loss": 0.4963,
"step": 157
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.4761314392089844,
"learning_rate": 5.782172325201155e-06,
"loss": 0.5091,
"step": 158
},
{
"epoch": 1.5239616613418532,
"grad_norm": 0.5091495513916016,
"learning_rate": 5.726715068949564e-06,
"loss": 0.4874,
"step": 159
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.4561215937137604,
"learning_rate": 5.671166329088278e-06,
"loss": 0.5133,
"step": 160
},
{
"epoch": 1.5431309904153354,
"grad_norm": 0.49368858337402344,
"learning_rate": 5.615533098453215e-06,
"loss": 0.5082,
"step": 161
},
{
"epoch": 1.5527156549520766,
"grad_norm": 0.48443740606307983,
"learning_rate": 5.559822380516539e-06,
"loss": 0.495,
"step": 162
},
{
"epoch": 1.5623003194888179,
"grad_norm": 0.4822789430618286,
"learning_rate": 5.504041188505022e-06,
"loss": 0.4815,
"step": 163
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.468281626701355,
"learning_rate": 5.448196544517168e-06,
"loss": 0.505,
"step": 164
},
{
"epoch": 1.5814696485623003,
"grad_norm": 0.43839624524116516,
"learning_rate": 5.392295478639226e-06,
"loss": 0.4906,
"step": 165
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.4917108416557312,
"learning_rate": 5.336345028060199e-06,
"loss": 0.4954,
"step": 166
},
{
"epoch": 1.6006389776357828,
"grad_norm": 0.4497264623641968,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.495,
"step": 167
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.4717963635921478,
"learning_rate": 5.224324151752575e-06,
"loss": 0.4946,
"step": 168
},
{
"epoch": 1.619808306709265,
"grad_norm": 0.5031352639198303,
"learning_rate": 5.168267827938971e-06,
"loss": 0.4925,
"step": 169
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.46020931005477905,
"learning_rate": 5.112190321479026e-06,
"loss": 0.5175,
"step": 170
},
{
"epoch": 1.6389776357827475,
"grad_norm": 0.5058203339576721,
"learning_rate": 5.05609869177323e-06,
"loss": 0.507,
"step": 171
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.4962829053401947,
"learning_rate": 5e-06,
"loss": 0.4888,
"step": 172
},
{
"epoch": 1.65814696485623,
"grad_norm": 0.47301244735717773,
"learning_rate": 4.943901308226771e-06,
"loss": 0.4924,
"step": 173
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.45008736848831177,
"learning_rate": 4.887809678520976e-06,
"loss": 0.4931,
"step": 174
},
{
"epoch": 1.6773162939297124,
"grad_norm": 0.4861735701560974,
"learning_rate": 4.831732172061032e-06,
"loss": 0.4928,
"step": 175
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.5253743529319763,
"learning_rate": 4.775675848247427e-06,
"loss": 0.4854,
"step": 176
},
{
"epoch": 1.6964856230031948,
"grad_norm": 0.475273072719574,
"learning_rate": 4.719647763814041e-06,
"loss": 0.4972,
"step": 177
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.460649698972702,
"learning_rate": 4.663654971939802e-06,
"loss": 0.4927,
"step": 178
},
{
"epoch": 1.7156549520766773,
"grad_norm": 0.4971240162849426,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.4969,
"step": 179
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.49787262082099915,
"learning_rate": 4.551803455482833e-06,
"loss": 0.501,
"step": 180
},
{
"epoch": 1.7348242811501597,
"grad_norm": 0.4486851692199707,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.4906,
"step": 181
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.5093643069267273,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.4993,
"step": 182
},
{
"epoch": 1.7539936102236422,
"grad_norm": 0.5099018216133118,
"learning_rate": 4.384466901546786e-06,
"loss": 0.4995,
"step": 183
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.4710257947444916,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.4971,
"step": 184
},
{
"epoch": 1.7731629392971247,
"grad_norm": 0.4950502812862396,
"learning_rate": 4.273284931050438e-06,
"loss": 0.4805,
"step": 185
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.5105454921722412,
"learning_rate": 4.217827674798845e-06,
"loss": 0.4811,
"step": 186
},
{
"epoch": 1.792332268370607,
"grad_norm": 0.425629198551178,
"learning_rate": 4.162468883476319e-06,
"loss": 0.4834,
"step": 187
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.5057090520858765,
"learning_rate": 4.107215526006818e-06,
"loss": 0.4956,
"step": 188
},
{
"epoch": 1.8115015974440896,
"grad_norm": 0.5507011413574219,
"learning_rate": 4.052074558041608e-06,
"loss": 0.4865,
"step": 189
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.50043785572052,
"learning_rate": 3.997052921083637e-06,
"loss": 0.4866,
"step": 190
},
{
"epoch": 1.830670926517572,
"grad_norm": 0.4524836540222168,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.4863,
"step": 191
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.47252020239830017,
"learning_rate": 3.887395330218429e-06,
"loss": 0.4878,
"step": 192
},
{
"epoch": 1.8498402555910545,
"grad_norm": 0.5151874423027039,
"learning_rate": 3.832773180720475e-06,
"loss": 0.4933,
"step": 193
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.4702584743499756,
"learning_rate": 3.778297969310529e-06,
"loss": 0.4904,
"step": 194
},
{
"epoch": 1.8690095846645367,
"grad_norm": 0.4473407566547394,
"learning_rate": 3.723976553681787e-06,
"loss": 0.488,
"step": 195
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.5159157514572144,
"learning_rate": 3.669815772166625e-06,
"loss": 0.4944,
"step": 196
},
{
"epoch": 1.8881789137380192,
"grad_norm": 0.4549194872379303,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.4909,
"step": 197
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.4611094295978546,
"learning_rate": 3.562003362839914e-06,
"loss": 0.4899,
"step": 198
},
{
"epoch": 1.9073482428115016,
"grad_norm": 0.4491714537143707,
"learning_rate": 3.50836530715422e-06,
"loss": 0.4842,
"step": 199
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.4658133387565613,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.4867,
"step": 200
},
{
"epoch": 1.926517571884984,
"grad_norm": 0.48666077852249146,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.4833,
"step": 201
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.45596322417259216,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.4829,
"step": 202
},
{
"epoch": 1.9456869009584663,
"grad_norm": 0.419820636510849,
"learning_rate": 3.295758014387375e-06,
"loss": 0.4921,
"step": 203
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.4313202202320099,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.5008,
"step": 204
},
{
"epoch": 1.9648562300319488,
"grad_norm": 0.48349729180336,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.4946,
"step": 205
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.4540484845638275,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.4744,
"step": 206
},
{
"epoch": 1.9840255591054312,
"grad_norm": 0.489071786403656,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.4735,
"step": 207
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.5589330196380615,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.4791,
"step": 208
},
{
"epoch": 2.0031948881789137,
"grad_norm": 0.45701128244400024,
"learning_rate": 2.9834142279959754e-06,
"loss": 0.465,
"step": 209
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.45574355125427246,
"learning_rate": 2.932207475167398e-06,
"loss": 0.4524,
"step": 210
},
{
"epoch": 2.022364217252396,
"grad_norm": 0.46179401874542236,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.4517,
"step": 211
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.5420477986335754,
"learning_rate": 2.83058130441221e-06,
"loss": 0.4838,
"step": 212
},
{
"epoch": 2.0415335463258786,
"grad_norm": 0.4719928503036499,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.4596,
"step": 213
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.44018101692199707,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.4465,
"step": 214
},
{
"epoch": 2.060702875399361,
"grad_norm": 0.3992289900779724,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.4529,
"step": 215
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.4838718771934509,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.4489,
"step": 216
},
{
"epoch": 2.0798722044728435,
"grad_norm": 0.432561993598938,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.4538,
"step": 217
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.41792115569114685,
"learning_rate": 2.532458909678266e-06,
"loss": 0.4307,
"step": 218
},
{
"epoch": 2.099041533546326,
"grad_norm": 0.4183405637741089,
"learning_rate": 2.483822884971e-06,
"loss": 0.4555,
"step": 219
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.41810762882232666,
"learning_rate": 2.43550361297047e-06,
"loss": 0.4535,
"step": 220
},
{
"epoch": 2.1182108626198084,
"grad_norm": 0.4222341775894165,
"learning_rate": 2.387507176420256e-06,
"loss": 0.4359,
"step": 221
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.4188306927680969,
"learning_rate": 2.339839617423318e-06,
"loss": 0.4529,
"step": 222
},
{
"epoch": 2.137380191693291,
"grad_norm": 0.43610385060310364,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.4526,
"step": 223
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.4101148545742035,
"learning_rate": 2.245515092739488e-06,
"loss": 0.4515,
"step": 224
},
{
"epoch": 2.1565495207667733,
"grad_norm": 0.4496552050113678,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.4506,
"step": 225
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.391523540019989,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.448,
"step": 226
},
{
"epoch": 2.1757188498402558,
"grad_norm": 0.42678871750831604,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.4542,
"step": 227
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.41675618290901184,
"learning_rate": 2.061073738537635e-06,
"loss": 0.4546,
"step": 228
},
{
"epoch": 2.194888178913738,
"grad_norm": 0.4072362184524536,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.4726,
"step": 229
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.4021867513656616,
"learning_rate": 1.971049780795901e-06,
"loss": 0.4556,
"step": 230
},
{
"epoch": 2.2140575079872207,
"grad_norm": 0.3870861828327179,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.4672,
"step": 231
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.42653751373291016,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.4604,
"step": 232
},
{
"epoch": 2.2332268370607027,
"grad_norm": 0.3991759121417999,
"learning_rate": 1.838887489710452e-06,
"loss": 0.4415,
"step": 233
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.42243877053260803,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.4454,
"step": 234
},
{
"epoch": 2.252396166134185,
"grad_norm": 0.4313516318798065,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.4365,
"step": 235
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.4172741770744324,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.4597,
"step": 236
},
{
"epoch": 2.2715654952076676,
"grad_norm": 0.4043318033218384,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.4676,
"step": 237
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.40954241156578064,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.4446,
"step": 238
},
{
"epoch": 2.29073482428115,
"grad_norm": 0.41136589646339417,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.4406,
"step": 239
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.4019395112991333,
"learning_rate": 1.544686755065677e-06,
"loss": 0.4652,
"step": 240
},
{
"epoch": 2.3099041533546325,
"grad_norm": 0.3808540105819702,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.4523,
"step": 241
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.40909168124198914,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.4347,
"step": 242
},
{
"epoch": 2.329073482428115,
"grad_norm": 0.4005642533302307,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.4507,
"step": 243
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.38417860865592957,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.4474,
"step": 244
},
{
"epoch": 2.3482428115015974,
"grad_norm": 0.3959197402000427,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.4394,
"step": 245
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.3867765963077545,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.4579,
"step": 246
},
{
"epoch": 2.36741214057508,
"grad_norm": 0.3757643699645996,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.4533,
"step": 247
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.39197012782096863,
"learning_rate": 1.234642669981946e-06,
"loss": 0.4434,
"step": 248
},
{
"epoch": 2.3865814696485623,
"grad_norm": 0.3953816294670105,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.4505,
"step": 249
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.3843384385108948,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.4527,
"step": 250
},
{
"epoch": 2.405750798722045,
"grad_norm": 0.3904695212841034,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.4611,
"step": 251
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.3878993093967438,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.4605,
"step": 252
},
{
"epoch": 2.4249201277955272,
"grad_norm": 0.3815719485282898,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.4495,
"step": 253
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.3835732638835907,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.4366,
"step": 254
},
{
"epoch": 2.4440894568690097,
"grad_norm": 0.3808978497982025,
"learning_rate": 9.881436225981107e-07,
"loss": 0.4555,
"step": 255
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.3896723687648773,
"learning_rate": 9.549150281252633e-07,
"loss": 0.4519,
"step": 256
},
{
"epoch": 2.463258785942492,
"grad_norm": 0.40839481353759766,
"learning_rate": 9.221956552036992e-07,
"loss": 0.4364,
"step": 257
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.38688576221466064,
"learning_rate": 8.899896227604509e-07,
"loss": 0.481,
"step": 258
},
{
"epoch": 2.4824281150159746,
"grad_norm": 0.36322906613349915,
"learning_rate": 8.58300985099918e-07,
"loss": 0.4566,
"step": 259
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.41321438550949097,
"learning_rate": 8.271337313934869e-07,
"loss": 0.4372,
"step": 260
},
{
"epoch": 2.501597444089457,
"grad_norm": 0.3688577711582184,
"learning_rate": 7.964917851773496e-07,
"loss": 0.4421,
"step": 261
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.41685447096824646,
"learning_rate": 7.663790038585794e-07,
"loss": 0.4385,
"step": 262
},
{
"epoch": 2.520766773162939,
"grad_norm": 0.38782286643981934,
"learning_rate": 7.367991782295392e-07,
"loss": 0.4474,
"step": 263
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.4024277329444885,
"learning_rate": 7.077560319906696e-07,
"loss": 0.4616,
"step": 264
},
{
"epoch": 2.539936102236422,
"grad_norm": 0.3968171775341034,
"learning_rate": 6.792532212817271e-07,
"loss": 0.4437,
"step": 265
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.3824290335178375,
"learning_rate": 6.512943342215234e-07,
"loss": 0.4537,
"step": 266
},
{
"epoch": 2.559105431309904,
"grad_norm": 0.3670612871646881,
"learning_rate": 6.238828904562316e-07,
"loss": 0.4158,
"step": 267
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.40948301553726196,
"learning_rate": 5.9702234071631e-07,
"loss": 0.4532,
"step": 268
},
{
"epoch": 2.5782747603833864,
"grad_norm": 0.4036369025707245,
"learning_rate": 5.707160663821009e-07,
"loss": 0.4543,
"step": 269
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.37994030117988586,
"learning_rate": 5.449673790581611e-07,
"loss": 0.4423,
"step": 270
},
{
"epoch": 2.597444089456869,
"grad_norm": 0.39913567900657654,
"learning_rate": 5.197795201563744e-07,
"loss": 0.4586,
"step": 271
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.39062511920928955,
"learning_rate": 4.951556604879049e-07,
"loss": 0.4674,
"step": 272
},
{
"epoch": 2.6166134185303513,
"grad_norm": 0.3789519965648651,
"learning_rate": 4.710988998640298e-07,
"loss": 0.4642,
"step": 273
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.38502490520477295,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.4552,
"step": 274
},
{
"epoch": 2.635782747603834,
"grad_norm": 0.38135772943496704,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.4463,
"step": 275
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.4037874639034271,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.4319,
"step": 276
},
{
"epoch": 2.6549520766773163,
"grad_norm": 0.39430180191993713,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.4534,
"step": 277
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.3736008405685425,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.4593,
"step": 278
},
{
"epoch": 2.6741214057507987,
"grad_norm": 0.3888098895549774,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.4463,
"step": 279
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.3972155749797821,
"learning_rate": 3.18825646801314e-07,
"loss": 0.4472,
"step": 280
},
{
"epoch": 2.693290734824281,
"grad_norm": 0.3832278549671173,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.4545,
"step": 281
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.3790014386177063,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.4433,
"step": 282
},
{
"epoch": 2.7124600638977636,
"grad_norm": 0.37405383586883545,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.4498,
"step": 283
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.384566068649292,
"learning_rate": 2.447174185242324e-07,
"loss": 0.4515,
"step": 284
},
{
"epoch": 2.731629392971246,
"grad_norm": 0.3748718202114105,
"learning_rate": 2.276812823220964e-07,
"loss": 0.4455,
"step": 285
},
{
"epoch": 2.741214057507987,
"grad_norm": 0.3710741400718689,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.4552,
"step": 286
},
{
"epoch": 2.7507987220447285,
"grad_norm": 0.38188865780830383,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.4552,
"step": 287
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.36755484342575073,
"learning_rate": 1.801856965207338e-07,
"loss": 0.4675,
"step": 288
},
{
"epoch": 2.769968051118211,
"grad_norm": 0.3661862015724182,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.4488,
"step": 289
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.3812325894832611,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.4291,
"step": 290
},
{
"epoch": 2.7891373801916934,
"grad_norm": 0.36422041058540344,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.4563,
"step": 291
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.3698974549770355,
"learning_rate": 1.253604390908819e-07,
"loss": 0.4618,
"step": 292
},
{
"epoch": 2.8083067092651754,
"grad_norm": 0.3597336709499359,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.4705,
"step": 293
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.3560789227485657,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.4577,
"step": 294
},
{
"epoch": 2.8274760383386583,
"grad_norm": 0.38544225692749023,
"learning_rate": 9.0678523819408e-08,
"loss": 0.4547,
"step": 295
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.36083221435546875,
"learning_rate": 8.035205700685167e-08,
"loss": 0.4555,
"step": 296
},
{
"epoch": 2.8466453674121404,
"grad_norm": 0.37722134590148926,
"learning_rate": 7.064490740882057e-08,
"loss": 0.4534,
"step": 297
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.3918997645378113,
"learning_rate": 6.15582970243117e-08,
"loss": 0.4552,
"step": 298
},
{
"epoch": 2.8658146964856233,
"grad_norm": 0.3688855767250061,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.4544,
"step": 299
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.35933518409729004,
"learning_rate": 4.52511911603265e-08,
"loss": 0.4433,
"step": 300
},
{
"epoch": 2.8849840255591053,
"grad_norm": 0.36762237548828125,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.4456,
"step": 301
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.3694286644458771,
"learning_rate": 3.143895053378698e-08,
"loss": 0.4616,
"step": 302
},
{
"epoch": 2.9041533546325877,
"grad_norm": 0.38449376821517944,
"learning_rate": 2.547062725623828e-08,
"loss": 0.4402,
"step": 303
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.3735463321208954,
"learning_rate": 2.012853002380466e-08,
"loss": 0.4375,
"step": 304
},
{
"epoch": 2.92332268370607,
"grad_norm": 0.36325889825820923,
"learning_rate": 1.541333133436018e-08,
"loss": 0.4444,
"step": 305
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.3668026626110077,
"learning_rate": 1.132562476771959e-08,
"loss": 0.4435,
"step": 306
},
{
"epoch": 2.9424920127795526,
"grad_norm": 0.37005847692489624,
"learning_rate": 7.865924910916977e-09,
"loss": 0.448,
"step": 307
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.3643838167190552,
"learning_rate": 5.034667293427053e-09,
"loss": 0.4389,
"step": 308
},
{
"epoch": 2.961661341853035,
"grad_norm": 0.37130314111709595,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.4291,
"step": 309
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.36163410544395447,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.4473,
"step": 310
},
{
"epoch": 2.9808306709265175,
"grad_norm": 0.35774660110473633,
"learning_rate": 3.147162264971471e-10,
"loss": 0.4605,
"step": 311
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.3753293454647064,
"learning_rate": 0.0,
"loss": 0.4393,
"step": 312
},
{
"epoch": 2.9904153354632586,
"step": 312,
"total_flos": 3.5450143984254976e+17,
"train_loss": 0.0,
"train_runtime": 6.4936,
"train_samples_per_second": 4619.9,
"train_steps_per_second": 48.047
}
],
"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": 3.5450143984254976e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}