EtashGuha's picture
Upload folder using huggingface_hub
85fb7e4 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": 6.971998691558838,
"learning_rate": 3.125e-07,
"loss": 1.1107,
"step": 1
},
{
"epoch": 0.019169329073482427,
"grad_norm": 7.050237655639648,
"learning_rate": 6.25e-07,
"loss": 1.1177,
"step": 2
},
{
"epoch": 0.02875399361022364,
"grad_norm": 7.03878927230835,
"learning_rate": 9.375000000000001e-07,
"loss": 1.1165,
"step": 3
},
{
"epoch": 0.038338658146964855,
"grad_norm": 6.786710262298584,
"learning_rate": 1.25e-06,
"loss": 1.1067,
"step": 4
},
{
"epoch": 0.04792332268370607,
"grad_norm": 6.495321750640869,
"learning_rate": 1.5625e-06,
"loss": 1.0499,
"step": 5
},
{
"epoch": 0.05750798722044728,
"grad_norm": 6.044283390045166,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.0371,
"step": 6
},
{
"epoch": 0.0670926517571885,
"grad_norm": 5.145187854766846,
"learning_rate": 2.1875000000000002e-06,
"loss": 1.0494,
"step": 7
},
{
"epoch": 0.07667731629392971,
"grad_norm": 4.828237533569336,
"learning_rate": 2.5e-06,
"loss": 1.0762,
"step": 8
},
{
"epoch": 0.08626198083067092,
"grad_norm": 2.552110195159912,
"learning_rate": 2.8125e-06,
"loss": 0.9586,
"step": 9
},
{
"epoch": 0.09584664536741214,
"grad_norm": 2.331293821334839,
"learning_rate": 3.125e-06,
"loss": 0.9199,
"step": 10
},
{
"epoch": 0.10543130990415335,
"grad_norm": 2.1774399280548096,
"learning_rate": 3.4375e-06,
"loss": 0.9448,
"step": 11
},
{
"epoch": 0.11501597444089456,
"grad_norm": 3.9896745681762695,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.932,
"step": 12
},
{
"epoch": 0.12460063897763578,
"grad_norm": 4.284629821777344,
"learning_rate": 4.0625000000000005e-06,
"loss": 0.9471,
"step": 13
},
{
"epoch": 0.134185303514377,
"grad_norm": 4.1393818855285645,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.9282,
"step": 14
},
{
"epoch": 0.14376996805111822,
"grad_norm": 3.974306106567383,
"learning_rate": 4.6875000000000004e-06,
"loss": 0.9321,
"step": 15
},
{
"epoch": 0.15335463258785942,
"grad_norm": 2.983408212661743,
"learning_rate": 5e-06,
"loss": 0.8831,
"step": 16
},
{
"epoch": 0.16293929712460065,
"grad_norm": 2.5403692722320557,
"learning_rate": 5.3125e-06,
"loss": 0.865,
"step": 17
},
{
"epoch": 0.17252396166134185,
"grad_norm": 2.273620128631592,
"learning_rate": 5.625e-06,
"loss": 0.8637,
"step": 18
},
{
"epoch": 0.18210862619808307,
"grad_norm": 1.9839009046554565,
"learning_rate": 5.9375e-06,
"loss": 0.8407,
"step": 19
},
{
"epoch": 0.19169329073482427,
"grad_norm": 1.8639857769012451,
"learning_rate": 6.25e-06,
"loss": 0.8496,
"step": 20
},
{
"epoch": 0.2012779552715655,
"grad_norm": 1.6055963039398193,
"learning_rate": 6.5625e-06,
"loss": 0.7852,
"step": 21
},
{
"epoch": 0.2108626198083067,
"grad_norm": 1.4574110507965088,
"learning_rate": 6.875e-06,
"loss": 0.8268,
"step": 22
},
{
"epoch": 0.22044728434504793,
"grad_norm": 1.5689210891723633,
"learning_rate": 7.1875e-06,
"loss": 0.87,
"step": 23
},
{
"epoch": 0.23003194888178913,
"grad_norm": 1.4550209045410156,
"learning_rate": 7.500000000000001e-06,
"loss": 0.8051,
"step": 24
},
{
"epoch": 0.23961661341853036,
"grad_norm": 1.18101966381073,
"learning_rate": 7.8125e-06,
"loss": 0.7869,
"step": 25
},
{
"epoch": 0.24920127795527156,
"grad_norm": 1.2257055044174194,
"learning_rate": 8.125000000000001e-06,
"loss": 0.7946,
"step": 26
},
{
"epoch": 0.25878594249201275,
"grad_norm": 1.1306984424591064,
"learning_rate": 8.4375e-06,
"loss": 0.8175,
"step": 27
},
{
"epoch": 0.268370607028754,
"grad_norm": 1.247745156288147,
"learning_rate": 8.750000000000001e-06,
"loss": 0.8078,
"step": 28
},
{
"epoch": 0.2779552715654952,
"grad_norm": 2.521697998046875,
"learning_rate": 9.0625e-06,
"loss": 0.7935,
"step": 29
},
{
"epoch": 0.28753993610223644,
"grad_norm": 1.3057435750961304,
"learning_rate": 9.375000000000001e-06,
"loss": 0.7915,
"step": 30
},
{
"epoch": 0.2971246006389776,
"grad_norm": 0.9969050288200378,
"learning_rate": 9.6875e-06,
"loss": 0.771,
"step": 31
},
{
"epoch": 0.30670926517571884,
"grad_norm": 0.9421972632408142,
"learning_rate": 1e-05,
"loss": 0.735,
"step": 32
},
{
"epoch": 0.31629392971246006,
"grad_norm": 1.0489099025726318,
"learning_rate": 9.999685283773504e-06,
"loss": 0.7474,
"step": 33
},
{
"epoch": 0.3258785942492013,
"grad_norm": 0.8595058917999268,
"learning_rate": 9.998741174712534e-06,
"loss": 0.7496,
"step": 34
},
{
"epoch": 0.3354632587859425,
"grad_norm": 0.8780394792556763,
"learning_rate": 9.997167791667668e-06,
"loss": 0.78,
"step": 35
},
{
"epoch": 0.3450479233226837,
"grad_norm": 0.8949142694473267,
"learning_rate": 9.994965332706574e-06,
"loss": 0.7421,
"step": 36
},
{
"epoch": 0.3546325878594249,
"grad_norm": 0.7785873413085938,
"learning_rate": 9.992134075089085e-06,
"loss": 0.7473,
"step": 37
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.8248518109321594,
"learning_rate": 9.98867437523228e-06,
"loss": 0.7495,
"step": 38
},
{
"epoch": 0.3738019169329074,
"grad_norm": 0.7783116102218628,
"learning_rate": 9.984586668665641e-06,
"loss": 0.7489,
"step": 39
},
{
"epoch": 0.38338658146964855,
"grad_norm": 0.7744529843330383,
"learning_rate": 9.979871469976197e-06,
"loss": 0.7466,
"step": 40
},
{
"epoch": 0.3929712460063898,
"grad_norm": 0.7752032279968262,
"learning_rate": 9.974529372743762e-06,
"loss": 0.7061,
"step": 41
},
{
"epoch": 0.402555910543131,
"grad_norm": 0.7973994016647339,
"learning_rate": 9.968561049466214e-06,
"loss": 0.7475,
"step": 42
},
{
"epoch": 0.41214057507987223,
"grad_norm": 0.7251701354980469,
"learning_rate": 9.961967251474823e-06,
"loss": 0.7191,
"step": 43
},
{
"epoch": 0.4217252396166134,
"grad_norm": 0.8239709734916687,
"learning_rate": 9.954748808839675e-06,
"loss": 0.7258,
"step": 44
},
{
"epoch": 0.43130990415335463,
"grad_norm": 0.7722738981246948,
"learning_rate": 9.946906630265184e-06,
"loss": 0.7387,
"step": 45
},
{
"epoch": 0.44089456869009586,
"grad_norm": 0.8330404758453369,
"learning_rate": 9.938441702975689e-06,
"loss": 0.7654,
"step": 46
},
{
"epoch": 0.4504792332268371,
"grad_norm": 0.7460812330245972,
"learning_rate": 9.92935509259118e-06,
"loss": 0.7222,
"step": 47
},
{
"epoch": 0.46006389776357826,
"grad_norm": 0.7764353156089783,
"learning_rate": 9.91964794299315e-06,
"loss": 0.7348,
"step": 48
},
{
"epoch": 0.4696485623003195,
"grad_norm": 0.7835634350776672,
"learning_rate": 9.909321476180594e-06,
"loss": 0.7483,
"step": 49
},
{
"epoch": 0.4792332268370607,
"grad_norm": 0.6525381207466125,
"learning_rate": 9.898376992116179e-06,
"loss": 0.7098,
"step": 50
},
{
"epoch": 0.48881789137380194,
"grad_norm": 0.7374613881111145,
"learning_rate": 9.886815868562596e-06,
"loss": 0.7285,
"step": 51
},
{
"epoch": 0.4984025559105431,
"grad_norm": 0.6773239374160767,
"learning_rate": 9.874639560909118e-06,
"loss": 0.6991,
"step": 52
},
{
"epoch": 0.5079872204472844,
"grad_norm": 0.6954315304756165,
"learning_rate": 9.861849601988384e-06,
"loss": 0.7529,
"step": 53
},
{
"epoch": 0.5175718849840255,
"grad_norm": 0.6873759031295776,
"learning_rate": 9.848447601883436e-06,
"loss": 0.7609,
"step": 54
},
{
"epoch": 0.5271565495207667,
"grad_norm": 0.7804772257804871,
"learning_rate": 9.834435247725032e-06,
"loss": 0.7336,
"step": 55
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.7185588479042053,
"learning_rate": 9.819814303479268e-06,
"loss": 0.7087,
"step": 56
},
{
"epoch": 0.5463258785942492,
"grad_norm": 0.7161798477172852,
"learning_rate": 9.804586609725499e-06,
"loss": 0.777,
"step": 57
},
{
"epoch": 0.5559105431309904,
"grad_norm": 0.7221360802650452,
"learning_rate": 9.788754083424654e-06,
"loss": 0.7518,
"step": 58
},
{
"epoch": 0.5654952076677316,
"grad_norm": 0.7326431274414062,
"learning_rate": 9.772318717677905e-06,
"loss": 0.7085,
"step": 59
},
{
"epoch": 0.5750798722044729,
"grad_norm": 0.6565002202987671,
"learning_rate": 9.755282581475769e-06,
"loss": 0.75,
"step": 60
},
{
"epoch": 0.5846645367412141,
"grad_norm": 0.7010030150413513,
"learning_rate": 9.737647819437645e-06,
"loss": 0.7344,
"step": 61
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.7021656036376953,
"learning_rate": 9.719416651541839e-06,
"loss": 0.7143,
"step": 62
},
{
"epoch": 0.6038338658146964,
"grad_norm": 0.7053258419036865,
"learning_rate": 9.700591372846096e-06,
"loss": 0.7163,
"step": 63
},
{
"epoch": 0.6134185303514377,
"grad_norm": 0.6960455179214478,
"learning_rate": 9.681174353198687e-06,
"loss": 0.7334,
"step": 64
},
{
"epoch": 0.6230031948881789,
"grad_norm": 0.5773729681968689,
"learning_rate": 9.661168036940071e-06,
"loss": 0.6995,
"step": 65
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.6698278784751892,
"learning_rate": 9.640574942595195e-06,
"loss": 0.7297,
"step": 66
},
{
"epoch": 0.6421725239616614,
"grad_norm": 0.6030433773994446,
"learning_rate": 9.619397662556434e-06,
"loss": 0.7042,
"step": 67
},
{
"epoch": 0.6517571884984026,
"grad_norm": 0.7778566479682922,
"learning_rate": 9.597638862757255e-06,
"loss": 0.7509,
"step": 68
},
{
"epoch": 0.6613418530351438,
"grad_norm": 0.6630182266235352,
"learning_rate": 9.5753012823366e-06,
"loss": 0.7036,
"step": 69
},
{
"epoch": 0.670926517571885,
"grad_norm": 0.866037130355835,
"learning_rate": 9.552387733294081e-06,
"loss": 0.7685,
"step": 70
},
{
"epoch": 0.6805111821086262,
"grad_norm": 0.8120040893554688,
"learning_rate": 9.528901100135971e-06,
"loss": 0.7108,
"step": 71
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.729915201663971,
"learning_rate": 9.504844339512096e-06,
"loss": 0.7254,
"step": 72
},
{
"epoch": 0.6996805111821086,
"grad_norm": 0.7347356677055359,
"learning_rate": 9.480220479843627e-06,
"loss": 0.6991,
"step": 73
},
{
"epoch": 0.7092651757188498,
"grad_norm": 0.810127854347229,
"learning_rate": 9.45503262094184e-06,
"loss": 0.718,
"step": 74
},
{
"epoch": 0.7188498402555911,
"grad_norm": 0.6601433157920837,
"learning_rate": 9.4292839336179e-06,
"loss": 0.6881,
"step": 75
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.7950295209884644,
"learning_rate": 9.40297765928369e-06,
"loss": 0.7312,
"step": 76
},
{
"epoch": 0.7380191693290735,
"grad_norm": 0.6458447575569153,
"learning_rate": 9.376117109543769e-06,
"loss": 0.7122,
"step": 77
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.7140399813652039,
"learning_rate": 9.348705665778479e-06,
"loss": 0.7109,
"step": 78
},
{
"epoch": 0.7571884984025559,
"grad_norm": 0.605841875076294,
"learning_rate": 9.320746778718274e-06,
"loss": 0.7056,
"step": 79
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.6923595070838928,
"learning_rate": 9.292243968009332e-06,
"loss": 0.696,
"step": 80
},
{
"epoch": 0.7763578274760383,
"grad_norm": 0.6835519671440125,
"learning_rate": 9.263200821770462e-06,
"loss": 0.7332,
"step": 81
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.6267021894454956,
"learning_rate": 9.233620996141421e-06,
"loss": 0.6944,
"step": 82
},
{
"epoch": 0.7955271565495208,
"grad_norm": 0.8507176637649536,
"learning_rate": 9.203508214822652e-06,
"loss": 0.7478,
"step": 83
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.732303261756897,
"learning_rate": 9.172866268606514e-06,
"loss": 0.7252,
"step": 84
},
{
"epoch": 0.8146964856230032,
"grad_norm": 0.6333317160606384,
"learning_rate": 9.141699014900084e-06,
"loss": 0.7,
"step": 85
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.6137699484825134,
"learning_rate": 9.110010377239552e-06,
"loss": 0.6773,
"step": 86
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.6669235825538635,
"learning_rate": 9.077804344796302e-06,
"loss": 0.7296,
"step": 87
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.7705378532409668,
"learning_rate": 9.045084971874738e-06,
"loss": 0.7017,
"step": 88
},
{
"epoch": 0.853035143769968,
"grad_norm": 0.7653799653053284,
"learning_rate": 9.011856377401891e-06,
"loss": 0.7441,
"step": 89
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.6707456707954407,
"learning_rate": 8.978122744408905e-06,
"loss": 0.7361,
"step": 90
},
{
"epoch": 0.8722044728434505,
"grad_norm": 0.7691360712051392,
"learning_rate": 8.943888319504456e-06,
"loss": 0.6952,
"step": 91
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.7389594912528992,
"learning_rate": 8.90915741234015e-06,
"loss": 0.6944,
"step": 92
},
{
"epoch": 0.8913738019169329,
"grad_norm": 0.7065908908843994,
"learning_rate": 8.873934395068006e-06,
"loss": 0.6945,
"step": 93
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.6882407069206238,
"learning_rate": 8.838223701790057e-06,
"loss": 0.6825,
"step": 94
},
{
"epoch": 0.9105431309904153,
"grad_norm": 0.6808512210845947,
"learning_rate": 8.802029828000157e-06,
"loss": 0.7075,
"step": 95
},
{
"epoch": 0.9201277955271565,
"grad_norm": 0.6349908709526062,
"learning_rate": 8.765357330018056e-06,
"loss": 0.7157,
"step": 96
},
{
"epoch": 0.9297124600638977,
"grad_norm": 0.6814895272254944,
"learning_rate": 8.728210824415829e-06,
"loss": 0.7014,
"step": 97
},
{
"epoch": 0.939297124600639,
"grad_norm": 0.6316925287246704,
"learning_rate": 8.690594987436705e-06,
"loss": 0.7267,
"step": 98
},
{
"epoch": 0.9488817891373802,
"grad_norm": 0.6401665806770325,
"learning_rate": 8.652514554406388e-06,
"loss": 0.706,
"step": 99
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.5566475987434387,
"learning_rate": 8.613974319136959e-06,
"loss": 0.6619,
"step": 100
},
{
"epoch": 0.9680511182108626,
"grad_norm": 0.632225513458252,
"learning_rate": 8.574979133323378e-06,
"loss": 0.7069,
"step": 101
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.5987431406974792,
"learning_rate": 8.535533905932739e-06,
"loss": 0.7253,
"step": 102
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.604228675365448,
"learning_rate": 8.495643602586287e-06,
"loss": 0.678,
"step": 103
},
{
"epoch": 0.9968051118210862,
"grad_norm": 0.6816259622573853,
"learning_rate": 8.455313244934324e-06,
"loss": 0.7143,
"step": 104
},
{
"epoch": 1.0063897763578276,
"grad_norm": 0.5808905363082886,
"learning_rate": 8.414547910024035e-06,
"loss": 0.6676,
"step": 105
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.6644414067268372,
"learning_rate": 8.373352729660373e-06,
"loss": 0.6234,
"step": 106
},
{
"epoch": 1.0255591054313098,
"grad_norm": 0.7519953846931458,
"learning_rate": 8.331732889760021e-06,
"loss": 0.6192,
"step": 107
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.5884900689125061,
"learning_rate": 8.289693629698564e-06,
"loss": 0.6335,
"step": 108
},
{
"epoch": 1.0447284345047922,
"grad_norm": 0.5819219946861267,
"learning_rate": 8.247240241650918e-06,
"loss": 0.6148,
"step": 109
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.6081036925315857,
"learning_rate": 8.204378069925121e-06,
"loss": 0.6204,
"step": 110
},
{
"epoch": 1.0638977635782747,
"grad_norm": 0.7000735998153687,
"learning_rate": 8.16111251028955e-06,
"loss": 0.6116,
"step": 111
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.6014118790626526,
"learning_rate": 8.117449009293668e-06,
"loss": 0.6457,
"step": 112
},
{
"epoch": 1.0830670926517572,
"grad_norm": 0.6278773546218872,
"learning_rate": 8.073393063582386e-06,
"loss": 0.6246,
"step": 113
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.6522474884986877,
"learning_rate": 8.0289502192041e-06,
"loss": 0.6342,
"step": 114
},
{
"epoch": 1.1022364217252396,
"grad_norm": 0.6242803931236267,
"learning_rate": 7.984126070912519e-06,
"loss": 0.6439,
"step": 115
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.5995987057685852,
"learning_rate": 7.938926261462366e-06,
"loss": 0.6205,
"step": 116
},
{
"epoch": 1.121405750798722,
"grad_norm": 0.6373648047447205,
"learning_rate": 7.89335648089903e-06,
"loss": 0.6241,
"step": 117
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.6015115976333618,
"learning_rate": 7.84742246584226e-06,
"loss": 0.6483,
"step": 118
},
{
"epoch": 1.1405750798722045,
"grad_norm": 0.663524866104126,
"learning_rate": 7.801129998764014e-06,
"loss": 0.6231,
"step": 119
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.49973756074905396,
"learning_rate": 7.754484907260513e-06,
"loss": 0.5994,
"step": 120
},
{
"epoch": 1.159744408945687,
"grad_norm": 0.6662487983703613,
"learning_rate": 7.70749306331863e-06,
"loss": 0.6296,
"step": 121
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.6271225810050964,
"learning_rate": 7.660160382576683e-06,
"loss": 0.6182,
"step": 122
},
{
"epoch": 1.1789137380191694,
"grad_norm": 0.5760684609413147,
"learning_rate": 7.612492823579744e-06,
"loss": 0.6507,
"step": 123
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.6523165702819824,
"learning_rate": 7.564496387029532e-06,
"loss": 0.63,
"step": 124
},
{
"epoch": 1.1980830670926517,
"grad_norm": 0.5384656190872192,
"learning_rate": 7.516177115029002e-06,
"loss": 0.6101,
"step": 125
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.6138818264007568,
"learning_rate": 7.467541090321735e-06,
"loss": 0.6097,
"step": 126
},
{
"epoch": 1.2172523961661341,
"grad_norm": 0.6557722091674805,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.6323,
"step": 127
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.6197437047958374,
"learning_rate": 7.369343312364994e-06,
"loss": 0.6205,
"step": 128
},
{
"epoch": 1.2364217252396166,
"grad_norm": 0.573975145816803,
"learning_rate": 7.319793920889171e-06,
"loss": 0.5997,
"step": 129
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.6613350510597229,
"learning_rate": 7.269952498697734e-06,
"loss": 0.6221,
"step": 130
},
{
"epoch": 1.255591054313099,
"grad_norm": 0.6064529418945312,
"learning_rate": 7.219825320152411e-06,
"loss": 0.6302,
"step": 131
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.5708147883415222,
"learning_rate": 7.169418695587791e-06,
"loss": 0.6076,
"step": 132
},
{
"epoch": 1.2747603833865815,
"grad_norm": 0.6603698134422302,
"learning_rate": 7.118738970516944e-06,
"loss": 0.619,
"step": 133
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.5752602219581604,
"learning_rate": 7.067792524832604e-06,
"loss": 0.6104,
"step": 134
},
{
"epoch": 1.293929712460064,
"grad_norm": 0.5691383481025696,
"learning_rate": 7.016585772004026e-06,
"loss": 0.6119,
"step": 135
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.6326082944869995,
"learning_rate": 6.965125158269619e-06,
"loss": 0.6136,
"step": 136
},
{
"epoch": 1.3130990415335464,
"grad_norm": 0.6783974766731262,
"learning_rate": 6.913417161825449e-06,
"loss": 0.6245,
"step": 137
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.5575075149536133,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.6218,
"step": 138
},
{
"epoch": 1.3322683706070286,
"grad_norm": 0.5773029327392578,
"learning_rate": 6.809285088483361e-06,
"loss": 0.6018,
"step": 139
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.7214598655700684,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.6105,
"step": 140
},
{
"epoch": 1.351437699680511,
"grad_norm": 0.5614567399024963,
"learning_rate": 6.704241985612625e-06,
"loss": 0.6114,
"step": 141
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.5857440233230591,
"learning_rate": 6.651395309775837e-06,
"loss": 0.5887,
"step": 142
},
{
"epoch": 1.3706070287539935,
"grad_norm": 0.6536932587623596,
"learning_rate": 6.598340745578908e-06,
"loss": 0.6531,
"step": 143
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.5684155225753784,
"learning_rate": 6.545084971874738e-06,
"loss": 0.6384,
"step": 144
},
{
"epoch": 1.389776357827476,
"grad_norm": 0.5640861988067627,
"learning_rate": 6.491634692845781e-06,
"loss": 0.6243,
"step": 145
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.48580506443977356,
"learning_rate": 6.437996637160086e-06,
"loss": 0.6022,
"step": 146
},
{
"epoch": 1.4089456869009584,
"grad_norm": 0.5101671814918518,
"learning_rate": 6.384177557124247e-06,
"loss": 0.6286,
"step": 147
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.5692585706710815,
"learning_rate": 6.330184227833376e-06,
"loss": 0.6325,
"step": 148
},
{
"epoch": 1.428115015974441,
"grad_norm": 0.6159618496894836,
"learning_rate": 6.276023446318214e-06,
"loss": 0.6116,
"step": 149
},
{
"epoch": 1.4376996805111821,
"grad_norm": 0.5336474776268005,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.613,
"step": 150
},
{
"epoch": 1.4472843450479234,
"grad_norm": 0.5104553699493408,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.6349,
"step": 151
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.5543062090873718,
"learning_rate": 6.112604669781572e-06,
"loss": 0.6382,
"step": 152
},
{
"epoch": 1.4664536741214058,
"grad_norm": 0.5532480478286743,
"learning_rate": 6.057842458386315e-06,
"loss": 0.6132,
"step": 153
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.6132843494415283,
"learning_rate": 6.002947078916365e-06,
"loss": 0.6323,
"step": 154
},
{
"epoch": 1.4856230031948883,
"grad_norm": 0.510259747505188,
"learning_rate": 5.947925441958393e-06,
"loss": 0.6107,
"step": 155
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.6006035804748535,
"learning_rate": 5.892784473993184e-06,
"loss": 0.6227,
"step": 156
},
{
"epoch": 1.5047923322683707,
"grad_norm": 0.5839741230010986,
"learning_rate": 5.837531116523683e-06,
"loss": 0.6237,
"step": 157
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.5134792923927307,
"learning_rate": 5.782172325201155e-06,
"loss": 0.6297,
"step": 158
},
{
"epoch": 1.5239616613418532,
"grad_norm": 0.5354273319244385,
"learning_rate": 5.726715068949564e-06,
"loss": 0.6197,
"step": 159
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.5770488977432251,
"learning_rate": 5.671166329088278e-06,
"loss": 0.6154,
"step": 160
},
{
"epoch": 1.5431309904153354,
"grad_norm": 0.5453071594238281,
"learning_rate": 5.615533098453215e-06,
"loss": 0.624,
"step": 161
},
{
"epoch": 1.5527156549520766,
"grad_norm": 0.5175986886024475,
"learning_rate": 5.559822380516539e-06,
"loss": 0.6079,
"step": 162
},
{
"epoch": 1.5623003194888179,
"grad_norm": 0.6262847781181335,
"learning_rate": 5.504041188505022e-06,
"loss": 0.6326,
"step": 163
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.5686348080635071,
"learning_rate": 5.448196544517168e-06,
"loss": 0.6246,
"step": 164
},
{
"epoch": 1.5814696485623003,
"grad_norm": 0.560458779335022,
"learning_rate": 5.392295478639226e-06,
"loss": 0.6181,
"step": 165
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.5434932112693787,
"learning_rate": 5.336345028060199e-06,
"loss": 0.5863,
"step": 166
},
{
"epoch": 1.6006389776357828,
"grad_norm": 0.5670589804649353,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.6022,
"step": 167
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.5307400226593018,
"learning_rate": 5.224324151752575e-06,
"loss": 0.5895,
"step": 168
},
{
"epoch": 1.619808306709265,
"grad_norm": 0.6883311867713928,
"learning_rate": 5.168267827938971e-06,
"loss": 0.5819,
"step": 169
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.6043215990066528,
"learning_rate": 5.112190321479026e-06,
"loss": 0.6221,
"step": 170
},
{
"epoch": 1.6389776357827475,
"grad_norm": 0.5461068153381348,
"learning_rate": 5.05609869177323e-06,
"loss": 0.6216,
"step": 171
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.5376251339912415,
"learning_rate": 5e-06,
"loss": 0.6331,
"step": 172
},
{
"epoch": 1.65814696485623,
"grad_norm": 0.5725062489509583,
"learning_rate": 4.943901308226771e-06,
"loss": 0.6208,
"step": 173
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.5605234503746033,
"learning_rate": 4.887809678520976e-06,
"loss": 0.6244,
"step": 174
},
{
"epoch": 1.6773162939297124,
"grad_norm": 0.6107360124588013,
"learning_rate": 4.831732172061032e-06,
"loss": 0.6057,
"step": 175
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.5460118651390076,
"learning_rate": 4.775675848247427e-06,
"loss": 0.6224,
"step": 176
},
{
"epoch": 1.6964856230031948,
"grad_norm": 0.5322807431221008,
"learning_rate": 4.719647763814041e-06,
"loss": 0.6295,
"step": 177
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.5447019338607788,
"learning_rate": 4.663654971939802e-06,
"loss": 0.6178,
"step": 178
},
{
"epoch": 1.7156549520766773,
"grad_norm": 0.5457282066345215,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.6038,
"step": 179
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.5212422609329224,
"learning_rate": 4.551803455482833e-06,
"loss": 0.6119,
"step": 180
},
{
"epoch": 1.7348242811501597,
"grad_norm": 0.49170607328414917,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.6036,
"step": 181
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.536220371723175,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.6035,
"step": 182
},
{
"epoch": 1.7539936102236422,
"grad_norm": 0.5448375344276428,
"learning_rate": 4.384466901546786e-06,
"loss": 0.6128,
"step": 183
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.5462779998779297,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.5901,
"step": 184
},
{
"epoch": 1.7731629392971247,
"grad_norm": 0.5666403770446777,
"learning_rate": 4.273284931050438e-06,
"loss": 0.6313,
"step": 185
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.5086902379989624,
"learning_rate": 4.217827674798845e-06,
"loss": 0.6166,
"step": 186
},
{
"epoch": 1.792332268370607,
"grad_norm": 0.5493374466896057,
"learning_rate": 4.162468883476319e-06,
"loss": 0.628,
"step": 187
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.5654379725456238,
"learning_rate": 4.107215526006818e-06,
"loss": 0.6344,
"step": 188
},
{
"epoch": 1.8115015974440896,
"grad_norm": 0.57657790184021,
"learning_rate": 4.052074558041608e-06,
"loss": 0.6376,
"step": 189
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.5371982455253601,
"learning_rate": 3.997052921083637e-06,
"loss": 0.5847,
"step": 190
},
{
"epoch": 1.830670926517572,
"grad_norm": 0.541728138923645,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.6211,
"step": 191
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.5112797021865845,
"learning_rate": 3.887395330218429e-06,
"loss": 0.6119,
"step": 192
},
{
"epoch": 1.8498402555910545,
"grad_norm": 0.5171915888786316,
"learning_rate": 3.832773180720475e-06,
"loss": 0.6265,
"step": 193
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.46963581442832947,
"learning_rate": 3.778297969310529e-06,
"loss": 0.6256,
"step": 194
},
{
"epoch": 1.8690095846645367,
"grad_norm": 0.47522279620170593,
"learning_rate": 3.723976553681787e-06,
"loss": 0.5961,
"step": 195
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.4703618884086609,
"learning_rate": 3.669815772166625e-06,
"loss": 0.5961,
"step": 196
},
{
"epoch": 1.8881789137380192,
"grad_norm": 0.5094874501228333,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.6051,
"step": 197
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.49340417981147766,
"learning_rate": 3.562003362839914e-06,
"loss": 0.6515,
"step": 198
},
{
"epoch": 1.9073482428115016,
"grad_norm": 0.5253342390060425,
"learning_rate": 3.50836530715422e-06,
"loss": 0.6192,
"step": 199
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.47582119703292847,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.6012,
"step": 200
},
{
"epoch": 1.926517571884984,
"grad_norm": 0.5417906045913696,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.6247,
"step": 201
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.4653812050819397,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.6207,
"step": 202
},
{
"epoch": 1.9456869009584663,
"grad_norm": 0.5107676386833191,
"learning_rate": 3.295758014387375e-06,
"loss": 0.6113,
"step": 203
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.48281431198120117,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.6131,
"step": 204
},
{
"epoch": 1.9648562300319488,
"grad_norm": 0.4904637634754181,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.6146,
"step": 205
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.5293906331062317,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.6028,
"step": 206
},
{
"epoch": 1.9840255591054312,
"grad_norm": 0.48737844824790955,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.5971,
"step": 207
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.5447724461555481,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.6448,
"step": 208
},
{
"epoch": 2.0031948881789137,
"grad_norm": 0.48444610834121704,
"learning_rate": 2.9834142279959754e-06,
"loss": 0.5869,
"step": 209
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.675361156463623,
"learning_rate": 2.932207475167398e-06,
"loss": 0.5447,
"step": 210
},
{
"epoch": 2.022364217252396,
"grad_norm": 0.625113844871521,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.5638,
"step": 211
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.528229296207428,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5758,
"step": 212
},
{
"epoch": 2.0415335463258786,
"grad_norm": 0.5284104347229004,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.5618,
"step": 213
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.6477755308151245,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.5669,
"step": 214
},
{
"epoch": 2.060702875399361,
"grad_norm": 0.6743851900100708,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.5495,
"step": 215
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.5009170174598694,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5421,
"step": 216
},
{
"epoch": 2.0798722044728435,
"grad_norm": 0.5411891937255859,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.5708,
"step": 217
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.523042619228363,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5391,
"step": 218
},
{
"epoch": 2.099041533546326,
"grad_norm": 0.6371281743049622,
"learning_rate": 2.483822884971e-06,
"loss": 0.5425,
"step": 219
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.47539764642715454,
"learning_rate": 2.43550361297047e-06,
"loss": 0.5674,
"step": 220
},
{
"epoch": 2.1182108626198084,
"grad_norm": 0.5385382175445557,
"learning_rate": 2.387507176420256e-06,
"loss": 0.5401,
"step": 221
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.5269696116447449,
"learning_rate": 2.339839617423318e-06,
"loss": 0.529,
"step": 222
},
{
"epoch": 2.137380191693291,
"grad_norm": 0.5272794365882874,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.5653,
"step": 223
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.5422731041908264,
"learning_rate": 2.245515092739488e-06,
"loss": 0.566,
"step": 224
},
{
"epoch": 2.1565495207667733,
"grad_norm": 0.5046980977058411,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.5317,
"step": 225
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.4797118306159973,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.5708,
"step": 226
},
{
"epoch": 2.1757188498402558,
"grad_norm": 0.475576251745224,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.5208,
"step": 227
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.5215170979499817,
"learning_rate": 2.061073738537635e-06,
"loss": 0.541,
"step": 228
},
{
"epoch": 2.194888178913738,
"grad_norm": 0.5315196514129639,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.5416,
"step": 229
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.5296704769134521,
"learning_rate": 1.971049780795901e-06,
"loss": 0.5375,
"step": 230
},
{
"epoch": 2.2140575079872207,
"grad_norm": 0.4825049340724945,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.5798,
"step": 231
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.47409123182296753,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.54,
"step": 232
},
{
"epoch": 2.2332268370607027,
"grad_norm": 0.5220174789428711,
"learning_rate": 1.838887489710452e-06,
"loss": 0.5242,
"step": 233
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.5206701159477234,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.5511,
"step": 234
},
{
"epoch": 2.252396166134185,
"grad_norm": 0.4608929455280304,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.5785,
"step": 235
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.4998515844345093,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.5382,
"step": 236
},
{
"epoch": 2.2715654952076676,
"grad_norm": 0.4545069634914398,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.5375,
"step": 237
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.4538171887397766,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.5652,
"step": 238
},
{
"epoch": 2.29073482428115,
"grad_norm": 0.4549846053123474,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.5606,
"step": 239
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.45981788635253906,
"learning_rate": 1.544686755065677e-06,
"loss": 0.5495,
"step": 240
},
{
"epoch": 2.3099041533546325,
"grad_norm": 0.509513795375824,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.5273,
"step": 241
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.4437123239040375,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5451,
"step": 242
},
{
"epoch": 2.329073482428115,
"grad_norm": 0.4303329586982727,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.5572,
"step": 243
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.4266361892223358,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.561,
"step": 244
},
{
"epoch": 2.3482428115015974,
"grad_norm": 0.47618621587753296,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.5511,
"step": 245
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.480275422334671,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.5806,
"step": 246
},
{
"epoch": 2.36741214057508,
"grad_norm": 0.4448004961013794,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.5497,
"step": 247
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.4528818726539612,
"learning_rate": 1.234642669981946e-06,
"loss": 0.5724,
"step": 248
},
{
"epoch": 2.3865814696485623,
"grad_norm": 0.5146142840385437,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.5336,
"step": 249
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.4516039192676544,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.5318,
"step": 250
},
{
"epoch": 2.405750798722045,
"grad_norm": 0.48698943853378296,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.5577,
"step": 251
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.4867784380912781,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5367,
"step": 252
},
{
"epoch": 2.4249201277955272,
"grad_norm": 0.49107033014297485,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.5605,
"step": 253
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.4654329717159271,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.5436,
"step": 254
},
{
"epoch": 2.4440894568690097,
"grad_norm": 0.4349857270717621,
"learning_rate": 9.881436225981107e-07,
"loss": 0.5533,
"step": 255
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.43953022360801697,
"learning_rate": 9.549150281252633e-07,
"loss": 0.5358,
"step": 256
},
{
"epoch": 2.463258785942492,
"grad_norm": 0.425687313079834,
"learning_rate": 9.221956552036992e-07,
"loss": 0.5475,
"step": 257
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.4697389304637909,
"learning_rate": 8.899896227604509e-07,
"loss": 0.5291,
"step": 258
},
{
"epoch": 2.4824281150159746,
"grad_norm": 0.4302732050418854,
"learning_rate": 8.58300985099918e-07,
"loss": 0.5522,
"step": 259
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.44490545988082886,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5428,
"step": 260
},
{
"epoch": 2.501597444089457,
"grad_norm": 0.457554429769516,
"learning_rate": 7.964917851773496e-07,
"loss": 0.5438,
"step": 261
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.48835253715515137,
"learning_rate": 7.663790038585794e-07,
"loss": 0.5412,
"step": 262
},
{
"epoch": 2.520766773162939,
"grad_norm": 0.4150005280971527,
"learning_rate": 7.367991782295392e-07,
"loss": 0.5393,
"step": 263
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.44234219193458557,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5659,
"step": 264
},
{
"epoch": 2.539936102236422,
"grad_norm": 0.43340200185775757,
"learning_rate": 6.792532212817271e-07,
"loss": 0.5345,
"step": 265
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.4623399078845978,
"learning_rate": 6.512943342215234e-07,
"loss": 0.5222,
"step": 266
},
{
"epoch": 2.559105431309904,
"grad_norm": 0.44783130288124084,
"learning_rate": 6.238828904562316e-07,
"loss": 0.5617,
"step": 267
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.45720329880714417,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5715,
"step": 268
},
{
"epoch": 2.5782747603833864,
"grad_norm": 0.460199236869812,
"learning_rate": 5.707160663821009e-07,
"loss": 0.5746,
"step": 269
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.4227401614189148,
"learning_rate": 5.449673790581611e-07,
"loss": 0.5655,
"step": 270
},
{
"epoch": 2.597444089456869,
"grad_norm": 0.4950794577598572,
"learning_rate": 5.197795201563744e-07,
"loss": 0.543,
"step": 271
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.42022472620010376,
"learning_rate": 4.951556604879049e-07,
"loss": 0.5313,
"step": 272
},
{
"epoch": 2.6166134185303513,
"grad_norm": 0.4263933002948761,
"learning_rate": 4.710988998640298e-07,
"loss": 0.582,
"step": 273
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.456779807806015,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5743,
"step": 274
},
{
"epoch": 2.635782747603834,
"grad_norm": 0.46589067578315735,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.548,
"step": 275
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.45368775725364685,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5453,
"step": 276
},
{
"epoch": 2.6549520766773163,
"grad_norm": 0.45604661107063293,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.5498,
"step": 277
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.46903011202812195,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5508,
"step": 278
},
{
"epoch": 2.6741214057507987,
"grad_norm": 0.431189626455307,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.54,
"step": 279
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.4565132260322571,
"learning_rate": 3.18825646801314e-07,
"loss": 0.525,
"step": 280
},
{
"epoch": 2.693290734824281,
"grad_norm": 0.46803998947143555,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.5509,
"step": 281
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.44332244992256165,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5394,
"step": 282
},
{
"epoch": 2.7124600638977636,
"grad_norm": 0.46024301648139954,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.4987,
"step": 283
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.4184441566467285,
"learning_rate": 2.447174185242324e-07,
"loss": 0.5417,
"step": 284
},
{
"epoch": 2.731629392971246,
"grad_norm": 0.4836418628692627,
"learning_rate": 2.276812823220964e-07,
"loss": 0.5531,
"step": 285
},
{
"epoch": 2.741214057507987,
"grad_norm": 0.44165509939193726,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.5725,
"step": 286
},
{
"epoch": 2.7507987220447285,
"grad_norm": 0.44304710626602173,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.5432,
"step": 287
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.4602050185203552,
"learning_rate": 1.801856965207338e-07,
"loss": 0.5235,
"step": 288
},
{
"epoch": 2.769968051118211,
"grad_norm": 0.4554177224636078,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.5529,
"step": 289
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.4562681019306183,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5339,
"step": 290
},
{
"epoch": 2.7891373801916934,
"grad_norm": 0.4614582061767578,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.5502,
"step": 291
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.4560908079147339,
"learning_rate": 1.253604390908819e-07,
"loss": 0.5275,
"step": 292
},
{
"epoch": 2.8083067092651754,
"grad_norm": 0.4341632127761841,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.5326,
"step": 293
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.4644143879413605,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.5481,
"step": 294
},
{
"epoch": 2.8274760383386583,
"grad_norm": 0.4713132381439209,
"learning_rate": 9.0678523819408e-08,
"loss": 0.5222,
"step": 295
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.450314462184906,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5522,
"step": 296
},
{
"epoch": 2.8466453674121404,
"grad_norm": 0.46823227405548096,
"learning_rate": 7.064490740882057e-08,
"loss": 0.546,
"step": 297
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.43032318353652954,
"learning_rate": 6.15582970243117e-08,
"loss": 0.5542,
"step": 298
},
{
"epoch": 2.8658146964856233,
"grad_norm": 0.40848687291145325,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.5663,
"step": 299
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.4318215250968933,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5595,
"step": 300
},
{
"epoch": 2.8849840255591053,
"grad_norm": 0.4364784359931946,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.5338,
"step": 301
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.4324475824832916,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5481,
"step": 302
},
{
"epoch": 2.9041533546325877,
"grad_norm": 0.4409596920013428,
"learning_rate": 2.547062725623828e-08,
"loss": 0.531,
"step": 303
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.4079962968826294,
"learning_rate": 2.012853002380466e-08,
"loss": 0.5522,
"step": 304
},
{
"epoch": 2.92332268370607,
"grad_norm": 0.43923890590667725,
"learning_rate": 1.541333133436018e-08,
"loss": 0.5487,
"step": 305
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.3935544490814209,
"learning_rate": 1.132562476771959e-08,
"loss": 0.5437,
"step": 306
},
{
"epoch": 2.9424920127795526,
"grad_norm": 0.4257122874259949,
"learning_rate": 7.865924910916977e-09,
"loss": 0.5611,
"step": 307
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.49820852279663086,
"learning_rate": 5.034667293427053e-09,
"loss": 0.5626,
"step": 308
},
{
"epoch": 2.961661341853035,
"grad_norm": 0.43063557147979736,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.5644,
"step": 309
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.4651414453983307,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5432,
"step": 310
},
{
"epoch": 2.9808306709265175,
"grad_norm": 0.4408637583255768,
"learning_rate": 3.147162264971471e-10,
"loss": 0.5563,
"step": 311
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.40899205207824707,
"learning_rate": 0.0,
"loss": 0.5641,
"step": 312
},
{
"epoch": 2.9904153354632586,
"step": 312,
"total_flos": 4.367681917432627e+17,
"train_loss": 0.6496183709838451,
"train_runtime": 9740.0942,
"train_samples_per_second": 3.08,
"train_steps_per_second": 0.032
}
],
"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": 4.367681917432627e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}