DCFT-seed_code_r1_magpie-etash / trainer_state.json
EtashGuha's picture
Upload folder using huggingface_hub
8bef149 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": 8.28641414642334,
"learning_rate": 3.125e-07,
"loss": 1.3336,
"step": 1
},
{
"epoch": 0.019169329073482427,
"grad_norm": 7.961845397949219,
"learning_rate": 6.25e-07,
"loss": 1.288,
"step": 2
},
{
"epoch": 0.02875399361022364,
"grad_norm": 8.104182243347168,
"learning_rate": 9.375000000000001e-07,
"loss": 1.3148,
"step": 3
},
{
"epoch": 0.038338658146964855,
"grad_norm": 8.544092178344727,
"learning_rate": 1.25e-06,
"loss": 1.3101,
"step": 4
},
{
"epoch": 0.04792332268370607,
"grad_norm": 8.072916984558105,
"learning_rate": 1.5625e-06,
"loss": 1.3083,
"step": 5
},
{
"epoch": 0.05750798722044728,
"grad_norm": 7.379998207092285,
"learning_rate": 1.8750000000000003e-06,
"loss": 1.2963,
"step": 6
},
{
"epoch": 0.0670926517571885,
"grad_norm": 6.060365200042725,
"learning_rate": 2.1875000000000002e-06,
"loss": 1.2212,
"step": 7
},
{
"epoch": 0.07667731629392971,
"grad_norm": 6.091128349304199,
"learning_rate": 2.5e-06,
"loss": 1.227,
"step": 8
},
{
"epoch": 0.08626198083067092,
"grad_norm": 3.4329874515533447,
"learning_rate": 2.8125e-06,
"loss": 1.1345,
"step": 9
},
{
"epoch": 0.09584664536741214,
"grad_norm": 2.886763334274292,
"learning_rate": 3.125e-06,
"loss": 1.1034,
"step": 10
},
{
"epoch": 0.10543130990415335,
"grad_norm": 2.816666603088379,
"learning_rate": 3.4375e-06,
"loss": 1.1128,
"step": 11
},
{
"epoch": 0.11501597444089456,
"grad_norm": 3.286442279815674,
"learning_rate": 3.7500000000000005e-06,
"loss": 1.0518,
"step": 12
},
{
"epoch": 0.12460063897763578,
"grad_norm": 3.7684264183044434,
"learning_rate": 4.0625000000000005e-06,
"loss": 1.1182,
"step": 13
},
{
"epoch": 0.134185303514377,
"grad_norm": 3.3012030124664307,
"learning_rate": 4.3750000000000005e-06,
"loss": 1.056,
"step": 14
},
{
"epoch": 0.14376996805111822,
"grad_norm": 2.945880889892578,
"learning_rate": 4.6875000000000004e-06,
"loss": 1.0496,
"step": 15
},
{
"epoch": 0.15335463258785942,
"grad_norm": 2.372924327850342,
"learning_rate": 5e-06,
"loss": 1.0222,
"step": 16
},
{
"epoch": 0.16293929712460065,
"grad_norm": 2.2487854957580566,
"learning_rate": 5.3125e-06,
"loss": 0.9787,
"step": 17
},
{
"epoch": 0.17252396166134185,
"grad_norm": 2.4512689113616943,
"learning_rate": 5.625e-06,
"loss": 0.9927,
"step": 18
},
{
"epoch": 0.18210862619808307,
"grad_norm": 2.173262119293213,
"learning_rate": 5.9375e-06,
"loss": 0.9443,
"step": 19
},
{
"epoch": 0.19169329073482427,
"grad_norm": 2.0508430004119873,
"learning_rate": 6.25e-06,
"loss": 0.9532,
"step": 20
},
{
"epoch": 0.2012779552715655,
"grad_norm": 1.7337983846664429,
"learning_rate": 6.5625e-06,
"loss": 0.9289,
"step": 21
},
{
"epoch": 0.2108626198083067,
"grad_norm": 1.319021463394165,
"learning_rate": 6.875e-06,
"loss": 0.9059,
"step": 22
},
{
"epoch": 0.22044728434504793,
"grad_norm": 1.2511050701141357,
"learning_rate": 7.1875e-06,
"loss": 0.9266,
"step": 23
},
{
"epoch": 0.23003194888178913,
"grad_norm": 1.3274163007736206,
"learning_rate": 7.500000000000001e-06,
"loss": 0.9253,
"step": 24
},
{
"epoch": 0.23961661341853036,
"grad_norm": 1.4402295351028442,
"learning_rate": 7.8125e-06,
"loss": 0.8609,
"step": 25
},
{
"epoch": 0.24920127795527156,
"grad_norm": 1.3721038103103638,
"learning_rate": 8.125000000000001e-06,
"loss": 0.9395,
"step": 26
},
{
"epoch": 0.25878594249201275,
"grad_norm": 1.1600173711776733,
"learning_rate": 8.4375e-06,
"loss": 0.8564,
"step": 27
},
{
"epoch": 0.268370607028754,
"grad_norm": 1.1631922721862793,
"learning_rate": 8.750000000000001e-06,
"loss": 0.8628,
"step": 28
},
{
"epoch": 0.2779552715654952,
"grad_norm": 1.1730294227600098,
"learning_rate": 9.0625e-06,
"loss": 0.8701,
"step": 29
},
{
"epoch": 0.28753993610223644,
"grad_norm": 1.1777492761611938,
"learning_rate": 9.375000000000001e-06,
"loss": 0.9016,
"step": 30
},
{
"epoch": 0.2971246006389776,
"grad_norm": 1.161747694015503,
"learning_rate": 9.6875e-06,
"loss": 0.86,
"step": 31
},
{
"epoch": 0.30670926517571884,
"grad_norm": 0.9846294522285461,
"learning_rate": 1e-05,
"loss": 0.8733,
"step": 32
},
{
"epoch": 0.31629392971246006,
"grad_norm": 0.9801987409591675,
"learning_rate": 9.999685283773504e-06,
"loss": 0.8713,
"step": 33
},
{
"epoch": 0.3258785942492013,
"grad_norm": 0.9624075293540955,
"learning_rate": 9.998741174712534e-06,
"loss": 0.8756,
"step": 34
},
{
"epoch": 0.3354632587859425,
"grad_norm": 0.9341689348220825,
"learning_rate": 9.997167791667668e-06,
"loss": 0.8503,
"step": 35
},
{
"epoch": 0.3450479233226837,
"grad_norm": 0.9584170579910278,
"learning_rate": 9.994965332706574e-06,
"loss": 0.839,
"step": 36
},
{
"epoch": 0.3546325878594249,
"grad_norm": 0.881298303604126,
"learning_rate": 9.992134075089085e-06,
"loss": 0.8292,
"step": 37
},
{
"epoch": 0.36421725239616615,
"grad_norm": 0.8834772109985352,
"learning_rate": 9.98867437523228e-06,
"loss": 0.8447,
"step": 38
},
{
"epoch": 0.3738019169329074,
"grad_norm": 0.9181913137435913,
"learning_rate": 9.984586668665641e-06,
"loss": 0.8638,
"step": 39
},
{
"epoch": 0.38338658146964855,
"grad_norm": 0.8826067447662354,
"learning_rate": 9.979871469976197e-06,
"loss": 0.8401,
"step": 40
},
{
"epoch": 0.3929712460063898,
"grad_norm": 0.829815149307251,
"learning_rate": 9.974529372743762e-06,
"loss": 0.8088,
"step": 41
},
{
"epoch": 0.402555910543131,
"grad_norm": 1.015673041343689,
"learning_rate": 9.968561049466214e-06,
"loss": 0.8245,
"step": 42
},
{
"epoch": 0.41214057507987223,
"grad_norm": 0.8124088048934937,
"learning_rate": 9.961967251474823e-06,
"loss": 0.7946,
"step": 43
},
{
"epoch": 0.4217252396166134,
"grad_norm": 0.9575561285018921,
"learning_rate": 9.954748808839675e-06,
"loss": 0.8326,
"step": 44
},
{
"epoch": 0.43130990415335463,
"grad_norm": 0.8751543164253235,
"learning_rate": 9.946906630265184e-06,
"loss": 0.8006,
"step": 45
},
{
"epoch": 0.44089456869009586,
"grad_norm": 0.797423779964447,
"learning_rate": 9.938441702975689e-06,
"loss": 0.8258,
"step": 46
},
{
"epoch": 0.4504792332268371,
"grad_norm": 0.845460057258606,
"learning_rate": 9.92935509259118e-06,
"loss": 0.8559,
"step": 47
},
{
"epoch": 0.46006389776357826,
"grad_norm": 0.8510870337486267,
"learning_rate": 9.91964794299315e-06,
"loss": 0.8254,
"step": 48
},
{
"epoch": 0.4696485623003195,
"grad_norm": 0.9224098324775696,
"learning_rate": 9.909321476180594e-06,
"loss": 0.8172,
"step": 49
},
{
"epoch": 0.4792332268370607,
"grad_norm": 0.8639594316482544,
"learning_rate": 9.898376992116179e-06,
"loss": 0.8386,
"step": 50
},
{
"epoch": 0.48881789137380194,
"grad_norm": 0.8452349901199341,
"learning_rate": 9.886815868562596e-06,
"loss": 0.8255,
"step": 51
},
{
"epoch": 0.4984025559105431,
"grad_norm": 0.9711369872093201,
"learning_rate": 9.874639560909118e-06,
"loss": 0.7791,
"step": 52
},
{
"epoch": 0.5079872204472844,
"grad_norm": 0.8923742771148682,
"learning_rate": 9.861849601988384e-06,
"loss": 0.8159,
"step": 53
},
{
"epoch": 0.5175718849840255,
"grad_norm": 0.8718321323394775,
"learning_rate": 9.848447601883436e-06,
"loss": 0.8347,
"step": 54
},
{
"epoch": 0.5271565495207667,
"grad_norm": 0.863624095916748,
"learning_rate": 9.834435247725032e-06,
"loss": 0.8196,
"step": 55
},
{
"epoch": 0.536741214057508,
"grad_norm": 0.9416993260383606,
"learning_rate": 9.819814303479268e-06,
"loss": 0.8171,
"step": 56
},
{
"epoch": 0.5463258785942492,
"grad_norm": 0.9263773560523987,
"learning_rate": 9.804586609725499e-06,
"loss": 0.84,
"step": 57
},
{
"epoch": 0.5559105431309904,
"grad_norm": 0.9197577834129333,
"learning_rate": 9.788754083424654e-06,
"loss": 0.8165,
"step": 58
},
{
"epoch": 0.5654952076677316,
"grad_norm": 0.8337032794952393,
"learning_rate": 9.772318717677905e-06,
"loss": 0.7963,
"step": 59
},
{
"epoch": 0.5750798722044729,
"grad_norm": 0.8229170441627502,
"learning_rate": 9.755282581475769e-06,
"loss": 0.8344,
"step": 60
},
{
"epoch": 0.5846645367412141,
"grad_norm": 0.9366649985313416,
"learning_rate": 9.737647819437645e-06,
"loss": 0.8112,
"step": 61
},
{
"epoch": 0.5942492012779552,
"grad_norm": 0.8366050124168396,
"learning_rate": 9.719416651541839e-06,
"loss": 0.7964,
"step": 62
},
{
"epoch": 0.6038338658146964,
"grad_norm": 0.8179761171340942,
"learning_rate": 9.700591372846096e-06,
"loss": 0.8116,
"step": 63
},
{
"epoch": 0.6134185303514377,
"grad_norm": 0.8064228892326355,
"learning_rate": 9.681174353198687e-06,
"loss": 0.8048,
"step": 64
},
{
"epoch": 0.6230031948881789,
"grad_norm": 0.8869609236717224,
"learning_rate": 9.661168036940071e-06,
"loss": 0.7877,
"step": 65
},
{
"epoch": 0.6325878594249201,
"grad_norm": 0.8790745139122009,
"learning_rate": 9.640574942595195e-06,
"loss": 0.7943,
"step": 66
},
{
"epoch": 0.6421725239616614,
"grad_norm": 0.8839783668518066,
"learning_rate": 9.619397662556434e-06,
"loss": 0.7873,
"step": 67
},
{
"epoch": 0.6517571884984026,
"grad_norm": 0.9070810079574585,
"learning_rate": 9.597638862757255e-06,
"loss": 0.7855,
"step": 68
},
{
"epoch": 0.6613418530351438,
"grad_norm": 0.82879239320755,
"learning_rate": 9.5753012823366e-06,
"loss": 0.7798,
"step": 69
},
{
"epoch": 0.670926517571885,
"grad_norm": 0.8891162276268005,
"learning_rate": 9.552387733294081e-06,
"loss": 0.7959,
"step": 70
},
{
"epoch": 0.6805111821086262,
"grad_norm": 0.8354350328445435,
"learning_rate": 9.528901100135971e-06,
"loss": 0.7912,
"step": 71
},
{
"epoch": 0.6900958466453674,
"grad_norm": 0.7726418375968933,
"learning_rate": 9.504844339512096e-06,
"loss": 0.7835,
"step": 72
},
{
"epoch": 0.6996805111821086,
"grad_norm": 0.795405387878418,
"learning_rate": 9.480220479843627e-06,
"loss": 0.8038,
"step": 73
},
{
"epoch": 0.7092651757188498,
"grad_norm": 0.8260125517845154,
"learning_rate": 9.45503262094184e-06,
"loss": 0.824,
"step": 74
},
{
"epoch": 0.7188498402555911,
"grad_norm": 0.8215699791908264,
"learning_rate": 9.4292839336179e-06,
"loss": 0.7868,
"step": 75
},
{
"epoch": 0.7284345047923323,
"grad_norm": 0.8798810243606567,
"learning_rate": 9.40297765928369e-06,
"loss": 0.7778,
"step": 76
},
{
"epoch": 0.7380191693290735,
"grad_norm": 0.8703828454017639,
"learning_rate": 9.376117109543769e-06,
"loss": 0.7894,
"step": 77
},
{
"epoch": 0.7476038338658147,
"grad_norm": 0.8584697246551514,
"learning_rate": 9.348705665778479e-06,
"loss": 0.7884,
"step": 78
},
{
"epoch": 0.7571884984025559,
"grad_norm": 0.8695377707481384,
"learning_rate": 9.320746778718274e-06,
"loss": 0.8103,
"step": 79
},
{
"epoch": 0.7667731629392971,
"grad_norm": 0.9443321228027344,
"learning_rate": 9.292243968009332e-06,
"loss": 0.812,
"step": 80
},
{
"epoch": 0.7763578274760383,
"grad_norm": 0.8727284073829651,
"learning_rate": 9.263200821770462e-06,
"loss": 0.8094,
"step": 81
},
{
"epoch": 0.7859424920127795,
"grad_norm": 0.8518511652946472,
"learning_rate": 9.233620996141421e-06,
"loss": 0.7804,
"step": 82
},
{
"epoch": 0.7955271565495208,
"grad_norm": 0.8931755423545837,
"learning_rate": 9.203508214822652e-06,
"loss": 0.8054,
"step": 83
},
{
"epoch": 0.805111821086262,
"grad_norm": 0.8459916114807129,
"learning_rate": 9.172866268606514e-06,
"loss": 0.7751,
"step": 84
},
{
"epoch": 0.8146964856230032,
"grad_norm": 0.8048325181007385,
"learning_rate": 9.141699014900084e-06,
"loss": 0.781,
"step": 85
},
{
"epoch": 0.8242811501597445,
"grad_norm": 0.8775563836097717,
"learning_rate": 9.110010377239552e-06,
"loss": 0.8107,
"step": 86
},
{
"epoch": 0.8338658146964856,
"grad_norm": 0.8480209112167358,
"learning_rate": 9.077804344796302e-06,
"loss": 0.7984,
"step": 87
},
{
"epoch": 0.8434504792332268,
"grad_norm": 0.9413987398147583,
"learning_rate": 9.045084971874738e-06,
"loss": 0.8198,
"step": 88
},
{
"epoch": 0.853035143769968,
"grad_norm": 0.9165607690811157,
"learning_rate": 9.011856377401891e-06,
"loss": 0.75,
"step": 89
},
{
"epoch": 0.8626198083067093,
"grad_norm": 0.8278255462646484,
"learning_rate": 8.978122744408905e-06,
"loss": 0.8022,
"step": 90
},
{
"epoch": 0.8722044728434505,
"grad_norm": 0.8216646313667297,
"learning_rate": 8.943888319504456e-06,
"loss": 0.7772,
"step": 91
},
{
"epoch": 0.8817891373801917,
"grad_norm": 0.92466801404953,
"learning_rate": 8.90915741234015e-06,
"loss": 0.7627,
"step": 92
},
{
"epoch": 0.8913738019169329,
"grad_norm": 0.9172149300575256,
"learning_rate": 8.873934395068006e-06,
"loss": 0.7882,
"step": 93
},
{
"epoch": 0.9009584664536742,
"grad_norm": 0.9127603769302368,
"learning_rate": 8.838223701790057e-06,
"loss": 0.7766,
"step": 94
},
{
"epoch": 0.9105431309904153,
"grad_norm": 0.849410891532898,
"learning_rate": 8.802029828000157e-06,
"loss": 0.7978,
"step": 95
},
{
"epoch": 0.9201277955271565,
"grad_norm": 0.9008628726005554,
"learning_rate": 8.765357330018056e-06,
"loss": 0.7855,
"step": 96
},
{
"epoch": 0.9297124600638977,
"grad_norm": 0.9596195816993713,
"learning_rate": 8.728210824415829e-06,
"loss": 0.7722,
"step": 97
},
{
"epoch": 0.939297124600639,
"grad_norm": 0.788043737411499,
"learning_rate": 8.690594987436705e-06,
"loss": 0.7821,
"step": 98
},
{
"epoch": 0.9488817891373802,
"grad_norm": 0.8407720327377319,
"learning_rate": 8.652514554406388e-06,
"loss": 0.7928,
"step": 99
},
{
"epoch": 0.9584664536741214,
"grad_norm": 0.8538227081298828,
"learning_rate": 8.613974319136959e-06,
"loss": 0.7359,
"step": 100
},
{
"epoch": 0.9680511182108626,
"grad_norm": 0.9044310450553894,
"learning_rate": 8.574979133323378e-06,
"loss": 0.7787,
"step": 101
},
{
"epoch": 0.9776357827476039,
"grad_norm": 0.8156440854072571,
"learning_rate": 8.535533905932739e-06,
"loss": 0.7972,
"step": 102
},
{
"epoch": 0.987220447284345,
"grad_norm": 0.8083887100219727,
"learning_rate": 8.495643602586287e-06,
"loss": 0.7584,
"step": 103
},
{
"epoch": 0.9968051118210862,
"grad_norm": 0.8623904585838318,
"learning_rate": 8.455313244934324e-06,
"loss": 0.8231,
"step": 104
},
{
"epoch": 1.0063897763578276,
"grad_norm": 0.8614771962165833,
"learning_rate": 8.414547910024035e-06,
"loss": 0.7225,
"step": 105
},
{
"epoch": 1.0159744408945688,
"grad_norm": 0.8943082690238953,
"learning_rate": 8.373352729660373e-06,
"loss": 0.6502,
"step": 106
},
{
"epoch": 1.0255591054313098,
"grad_norm": 0.8055715560913086,
"learning_rate": 8.331732889760021e-06,
"loss": 0.6562,
"step": 107
},
{
"epoch": 1.035143769968051,
"grad_norm": 0.7840996384620667,
"learning_rate": 8.289693629698564e-06,
"loss": 0.6409,
"step": 108
},
{
"epoch": 1.0447284345047922,
"grad_norm": 0.8343753218650818,
"learning_rate": 8.247240241650918e-06,
"loss": 0.6358,
"step": 109
},
{
"epoch": 1.0543130990415335,
"grad_norm": 0.8512541651725769,
"learning_rate": 8.204378069925121e-06,
"loss": 0.643,
"step": 110
},
{
"epoch": 1.0638977635782747,
"grad_norm": 0.7879105806350708,
"learning_rate": 8.16111251028955e-06,
"loss": 0.671,
"step": 111
},
{
"epoch": 1.073482428115016,
"grad_norm": 0.7936350703239441,
"learning_rate": 8.117449009293668e-06,
"loss": 0.6419,
"step": 112
},
{
"epoch": 1.0830670926517572,
"grad_norm": 0.8089979887008667,
"learning_rate": 8.073393063582386e-06,
"loss": 0.6551,
"step": 113
},
{
"epoch": 1.0926517571884984,
"grad_norm": 0.8221156597137451,
"learning_rate": 8.0289502192041e-06,
"loss": 0.6413,
"step": 114
},
{
"epoch": 1.1022364217252396,
"grad_norm": 0.8375205397605896,
"learning_rate": 7.984126070912519e-06,
"loss": 0.6434,
"step": 115
},
{
"epoch": 1.1118210862619808,
"grad_norm": 0.8795710206031799,
"learning_rate": 7.938926261462366e-06,
"loss": 0.6458,
"step": 116
},
{
"epoch": 1.121405750798722,
"grad_norm": 0.893771767616272,
"learning_rate": 7.89335648089903e-06,
"loss": 0.66,
"step": 117
},
{
"epoch": 1.1309904153354633,
"grad_norm": 0.8382556438446045,
"learning_rate": 7.84742246584226e-06,
"loss": 0.6762,
"step": 118
},
{
"epoch": 1.1405750798722045,
"grad_norm": 0.8155967593193054,
"learning_rate": 7.801129998764014e-06,
"loss": 0.6537,
"step": 119
},
{
"epoch": 1.1501597444089458,
"grad_norm": 0.7696691751480103,
"learning_rate": 7.754484907260513e-06,
"loss": 0.6436,
"step": 120
},
{
"epoch": 1.159744408945687,
"grad_norm": 0.8113377690315247,
"learning_rate": 7.70749306331863e-06,
"loss": 0.6436,
"step": 121
},
{
"epoch": 1.1693290734824282,
"grad_norm": 0.8605034351348877,
"learning_rate": 7.660160382576683e-06,
"loss": 0.6298,
"step": 122
},
{
"epoch": 1.1789137380191694,
"grad_norm": 0.7721021771430969,
"learning_rate": 7.612492823579744e-06,
"loss": 0.651,
"step": 123
},
{
"epoch": 1.1884984025559104,
"grad_norm": 0.8216039538383484,
"learning_rate": 7.564496387029532e-06,
"loss": 0.6393,
"step": 124
},
{
"epoch": 1.1980830670926517,
"grad_norm": 0.8253751993179321,
"learning_rate": 7.516177115029002e-06,
"loss": 0.626,
"step": 125
},
{
"epoch": 1.207667731629393,
"grad_norm": 0.7769179344177246,
"learning_rate": 7.467541090321735e-06,
"loss": 0.6321,
"step": 126
},
{
"epoch": 1.2172523961661341,
"grad_norm": 0.8014129996299744,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.6631,
"step": 127
},
{
"epoch": 1.2268370607028753,
"grad_norm": 0.8275423049926758,
"learning_rate": 7.369343312364994e-06,
"loss": 0.6648,
"step": 128
},
{
"epoch": 1.2364217252396166,
"grad_norm": 0.7273306250572205,
"learning_rate": 7.319793920889171e-06,
"loss": 0.6146,
"step": 129
},
{
"epoch": 1.2460063897763578,
"grad_norm": 0.7334845662117004,
"learning_rate": 7.269952498697734e-06,
"loss": 0.6388,
"step": 130
},
{
"epoch": 1.255591054313099,
"grad_norm": 0.7436879873275757,
"learning_rate": 7.219825320152411e-06,
"loss": 0.6432,
"step": 131
},
{
"epoch": 1.2651757188498403,
"grad_norm": 0.7407752275466919,
"learning_rate": 7.169418695587791e-06,
"loss": 0.6274,
"step": 132
},
{
"epoch": 1.2747603833865815,
"grad_norm": 0.710435152053833,
"learning_rate": 7.118738970516944e-06,
"loss": 0.6407,
"step": 133
},
{
"epoch": 1.2843450479233227,
"grad_norm": 0.7457040548324585,
"learning_rate": 7.067792524832604e-06,
"loss": 0.6467,
"step": 134
},
{
"epoch": 1.293929712460064,
"grad_norm": 0.6775636672973633,
"learning_rate": 7.016585772004026e-06,
"loss": 0.6191,
"step": 135
},
{
"epoch": 1.3035143769968052,
"grad_norm": 0.7445527911186218,
"learning_rate": 6.965125158269619e-06,
"loss": 0.6273,
"step": 136
},
{
"epoch": 1.3130990415335464,
"grad_norm": 0.7720601558685303,
"learning_rate": 6.913417161825449e-06,
"loss": 0.6434,
"step": 137
},
{
"epoch": 1.3226837060702876,
"grad_norm": 0.7920132875442505,
"learning_rate": 6.8614682920097265e-06,
"loss": 0.6452,
"step": 138
},
{
"epoch": 1.3322683706070286,
"grad_norm": 0.79832923412323,
"learning_rate": 6.809285088483361e-06,
"loss": 0.6299,
"step": 139
},
{
"epoch": 1.34185303514377,
"grad_norm": 0.7126365900039673,
"learning_rate": 6.7568741204067145e-06,
"loss": 0.6664,
"step": 140
},
{
"epoch": 1.351437699680511,
"grad_norm": 0.7566642761230469,
"learning_rate": 6.704241985612625e-06,
"loss": 0.6345,
"step": 141
},
{
"epoch": 1.3610223642172525,
"grad_norm": 0.7100439071655273,
"learning_rate": 6.651395309775837e-06,
"loss": 0.6316,
"step": 142
},
{
"epoch": 1.3706070287539935,
"grad_norm": 0.7583263516426086,
"learning_rate": 6.598340745578908e-06,
"loss": 0.6371,
"step": 143
},
{
"epoch": 1.3801916932907348,
"grad_norm": 0.7854827046394348,
"learning_rate": 6.545084971874738e-06,
"loss": 0.6326,
"step": 144
},
{
"epoch": 1.389776357827476,
"grad_norm": 0.679090678691864,
"learning_rate": 6.491634692845781e-06,
"loss": 0.6254,
"step": 145
},
{
"epoch": 1.3993610223642172,
"grad_norm": 0.8024678230285645,
"learning_rate": 6.437996637160086e-06,
"loss": 0.6578,
"step": 146
},
{
"epoch": 1.4089456869009584,
"grad_norm": 0.7927014827728271,
"learning_rate": 6.384177557124247e-06,
"loss": 0.6396,
"step": 147
},
{
"epoch": 1.4185303514376997,
"grad_norm": 0.7771953344345093,
"learning_rate": 6.330184227833376e-06,
"loss": 0.6509,
"step": 148
},
{
"epoch": 1.428115015974441,
"grad_norm": 0.743351936340332,
"learning_rate": 6.276023446318214e-06,
"loss": 0.6572,
"step": 149
},
{
"epoch": 1.4376996805111821,
"grad_norm": 0.7579458355903625,
"learning_rate": 6.2217020306894705e-06,
"loss": 0.6479,
"step": 150
},
{
"epoch": 1.4472843450479234,
"grad_norm": 0.6972531676292419,
"learning_rate": 6.1672268192795285e-06,
"loss": 0.6415,
"step": 151
},
{
"epoch": 1.4568690095846646,
"grad_norm": 0.8091091513633728,
"learning_rate": 6.112604669781572e-06,
"loss": 0.6271,
"step": 152
},
{
"epoch": 1.4664536741214058,
"grad_norm": 0.849339485168457,
"learning_rate": 6.057842458386315e-06,
"loss": 0.6264,
"step": 153
},
{
"epoch": 1.476038338658147,
"grad_norm": 0.8586984276771545,
"learning_rate": 6.002947078916365e-06,
"loss": 0.6401,
"step": 154
},
{
"epoch": 1.4856230031948883,
"grad_norm": 0.761081874370575,
"learning_rate": 5.947925441958393e-06,
"loss": 0.632,
"step": 155
},
{
"epoch": 1.4952076677316293,
"grad_norm": 0.727756679058075,
"learning_rate": 5.892784473993184e-06,
"loss": 0.6566,
"step": 156
},
{
"epoch": 1.5047923322683707,
"grad_norm": 0.7336796522140503,
"learning_rate": 5.837531116523683e-06,
"loss": 0.6033,
"step": 157
},
{
"epoch": 1.5143769968051117,
"grad_norm": 0.851142168045044,
"learning_rate": 5.782172325201155e-06,
"loss": 0.625,
"step": 158
},
{
"epoch": 1.5239616613418532,
"grad_norm": 0.8353523015975952,
"learning_rate": 5.726715068949564e-06,
"loss": 0.624,
"step": 159
},
{
"epoch": 1.5335463258785942,
"grad_norm": 0.7693101763725281,
"learning_rate": 5.671166329088278e-06,
"loss": 0.6204,
"step": 160
},
{
"epoch": 1.5431309904153354,
"grad_norm": 0.7567870020866394,
"learning_rate": 5.615533098453215e-06,
"loss": 0.6167,
"step": 161
},
{
"epoch": 1.5527156549520766,
"grad_norm": 0.7667849063873291,
"learning_rate": 5.559822380516539e-06,
"loss": 0.6194,
"step": 162
},
{
"epoch": 1.5623003194888179,
"grad_norm": 0.7782608270645142,
"learning_rate": 5.504041188505022e-06,
"loss": 0.6564,
"step": 163
},
{
"epoch": 1.571884984025559,
"grad_norm": 0.8164049983024597,
"learning_rate": 5.448196544517168e-06,
"loss": 0.637,
"step": 164
},
{
"epoch": 1.5814696485623003,
"grad_norm": 0.6999335289001465,
"learning_rate": 5.392295478639226e-06,
"loss": 0.6284,
"step": 165
},
{
"epoch": 1.5910543130990416,
"grad_norm": 0.7408768534660339,
"learning_rate": 5.336345028060199e-06,
"loss": 0.6566,
"step": 166
},
{
"epoch": 1.6006389776357828,
"grad_norm": 0.745985746383667,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.6199,
"step": 167
},
{
"epoch": 1.610223642172524,
"grad_norm": 0.7223749160766602,
"learning_rate": 5.224324151752575e-06,
"loss": 0.6445,
"step": 168
},
{
"epoch": 1.619808306709265,
"grad_norm": 0.7497395873069763,
"learning_rate": 5.168267827938971e-06,
"loss": 0.633,
"step": 169
},
{
"epoch": 1.6293929712460065,
"grad_norm": 0.8394908308982849,
"learning_rate": 5.112190321479026e-06,
"loss": 0.6443,
"step": 170
},
{
"epoch": 1.6389776357827475,
"grad_norm": 0.7743979692459106,
"learning_rate": 5.05609869177323e-06,
"loss": 0.6761,
"step": 171
},
{
"epoch": 1.648562300319489,
"grad_norm": 0.7917114496231079,
"learning_rate": 5e-06,
"loss": 0.6262,
"step": 172
},
{
"epoch": 1.65814696485623,
"grad_norm": 0.7392551302909851,
"learning_rate": 4.943901308226771e-06,
"loss": 0.6407,
"step": 173
},
{
"epoch": 1.6677316293929714,
"grad_norm": 0.7174422144889832,
"learning_rate": 4.887809678520976e-06,
"loss": 0.6217,
"step": 174
},
{
"epoch": 1.6773162939297124,
"grad_norm": 0.727367639541626,
"learning_rate": 4.831732172061032e-06,
"loss": 0.6392,
"step": 175
},
{
"epoch": 1.6869009584664538,
"grad_norm": 0.7445531487464905,
"learning_rate": 4.775675848247427e-06,
"loss": 0.633,
"step": 176
},
{
"epoch": 1.6964856230031948,
"grad_norm": 0.7466280460357666,
"learning_rate": 4.719647763814041e-06,
"loss": 0.6383,
"step": 177
},
{
"epoch": 1.706070287539936,
"grad_norm": 0.7180089354515076,
"learning_rate": 4.663654971939802e-06,
"loss": 0.6169,
"step": 178
},
{
"epoch": 1.7156549520766773,
"grad_norm": 0.7303853631019592,
"learning_rate": 4.6077045213607765e-06,
"loss": 0.6376,
"step": 179
},
{
"epoch": 1.7252396166134185,
"grad_norm": 0.6955850124359131,
"learning_rate": 4.551803455482833e-06,
"loss": 0.6243,
"step": 180
},
{
"epoch": 1.7348242811501597,
"grad_norm": 0.7337273359298706,
"learning_rate": 4.4959588114949785e-06,
"loss": 0.6145,
"step": 181
},
{
"epoch": 1.744408945686901,
"grad_norm": 0.7309015989303589,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.662,
"step": 182
},
{
"epoch": 1.7539936102236422,
"grad_norm": 0.6881760358810425,
"learning_rate": 4.384466901546786e-06,
"loss": 0.6387,
"step": 183
},
{
"epoch": 1.7635782747603834,
"grad_norm": 0.7315975427627563,
"learning_rate": 4.3288336709117246e-06,
"loss": 0.6289,
"step": 184
},
{
"epoch": 1.7731629392971247,
"grad_norm": 0.681795597076416,
"learning_rate": 4.273284931050438e-06,
"loss": 0.6251,
"step": 185
},
{
"epoch": 1.7827476038338657,
"grad_norm": 0.7149399518966675,
"learning_rate": 4.217827674798845e-06,
"loss": 0.6491,
"step": 186
},
{
"epoch": 1.792332268370607,
"grad_norm": 0.7364827990531921,
"learning_rate": 4.162468883476319e-06,
"loss": 0.6343,
"step": 187
},
{
"epoch": 1.8019169329073481,
"grad_norm": 0.7374441623687744,
"learning_rate": 4.107215526006818e-06,
"loss": 0.6266,
"step": 188
},
{
"epoch": 1.8115015974440896,
"grad_norm": 0.722055196762085,
"learning_rate": 4.052074558041608e-06,
"loss": 0.6277,
"step": 189
},
{
"epoch": 1.8210862619808306,
"grad_norm": 0.700432300567627,
"learning_rate": 3.997052921083637e-06,
"loss": 0.6112,
"step": 190
},
{
"epoch": 1.830670926517572,
"grad_norm": 0.7113826870918274,
"learning_rate": 3.9421575416136866e-06,
"loss": 0.6185,
"step": 191
},
{
"epoch": 1.840255591054313,
"grad_norm": 0.7254252433776855,
"learning_rate": 3.887395330218429e-06,
"loss": 0.639,
"step": 192
},
{
"epoch": 1.8498402555910545,
"grad_norm": 0.6937692761421204,
"learning_rate": 3.832773180720475e-06,
"loss": 0.6561,
"step": 193
},
{
"epoch": 1.8594249201277955,
"grad_norm": 0.7593420147895813,
"learning_rate": 3.778297969310529e-06,
"loss": 0.6267,
"step": 194
},
{
"epoch": 1.8690095846645367,
"grad_norm": 0.7484208941459656,
"learning_rate": 3.723976553681787e-06,
"loss": 0.6535,
"step": 195
},
{
"epoch": 1.878594249201278,
"grad_norm": 0.692676305770874,
"learning_rate": 3.669815772166625e-06,
"loss": 0.631,
"step": 196
},
{
"epoch": 1.8881789137380192,
"grad_norm": 0.7009148597717285,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.5994,
"step": 197
},
{
"epoch": 1.8977635782747604,
"grad_norm": 0.7395853996276855,
"learning_rate": 3.562003362839914e-06,
"loss": 0.6209,
"step": 198
},
{
"epoch": 1.9073482428115016,
"grad_norm": 0.7032827734947205,
"learning_rate": 3.50836530715422e-06,
"loss": 0.6265,
"step": 199
},
{
"epoch": 1.9169329073482428,
"grad_norm": 0.717046856880188,
"learning_rate": 3.4549150281252635e-06,
"loss": 0.6475,
"step": 200
},
{
"epoch": 1.926517571884984,
"grad_norm": 0.714216947555542,
"learning_rate": 3.4016592544210937e-06,
"loss": 0.6573,
"step": 201
},
{
"epoch": 1.9361022364217253,
"grad_norm": 0.695358395576477,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.6276,
"step": 202
},
{
"epoch": 1.9456869009584663,
"grad_norm": 0.6927900314331055,
"learning_rate": 3.295758014387375e-06,
"loss": 0.6041,
"step": 203
},
{
"epoch": 1.9552715654952078,
"grad_norm": 0.7065994739532471,
"learning_rate": 3.2431258795932863e-06,
"loss": 0.6508,
"step": 204
},
{
"epoch": 1.9648562300319488,
"grad_norm": 0.7383747696876526,
"learning_rate": 3.1907149115166403e-06,
"loss": 0.6338,
"step": 205
},
{
"epoch": 1.9744408945686902,
"grad_norm": 0.7556900382041931,
"learning_rate": 3.1385317079902743e-06,
"loss": 0.5959,
"step": 206
},
{
"epoch": 1.9840255591054312,
"grad_norm": 0.7020954489707947,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.6237,
"step": 207
},
{
"epoch": 1.9936102236421727,
"grad_norm": 0.7306491136550903,
"learning_rate": 3.0348748417303826e-06,
"loss": 0.6156,
"step": 208
},
{
"epoch": 2.0031948881789137,
"grad_norm": 0.7318288087844849,
"learning_rate": 2.9834142279959754e-06,
"loss": 0.5942,
"step": 209
},
{
"epoch": 2.012779552715655,
"grad_norm": 0.8892263174057007,
"learning_rate": 2.932207475167398e-06,
"loss": 0.5279,
"step": 210
},
{
"epoch": 2.022364217252396,
"grad_norm": 0.7372443079948425,
"learning_rate": 2.8812610294830568e-06,
"loss": 0.5479,
"step": 211
},
{
"epoch": 2.0319488817891376,
"grad_norm": 0.657680332660675,
"learning_rate": 2.83058130441221e-06,
"loss": 0.5277,
"step": 212
},
{
"epoch": 2.0415335463258786,
"grad_norm": 0.6568723320960999,
"learning_rate": 2.7801746798475905e-06,
"loss": 0.5305,
"step": 213
},
{
"epoch": 2.0511182108626196,
"grad_norm": 0.6973379254341125,
"learning_rate": 2.7300475013022666e-06,
"loss": 0.5477,
"step": 214
},
{
"epoch": 2.060702875399361,
"grad_norm": 0.718750536441803,
"learning_rate": 2.6802060791108304e-06,
"loss": 0.5315,
"step": 215
},
{
"epoch": 2.070287539936102,
"grad_norm": 0.7542418837547302,
"learning_rate": 2.6306566876350072e-06,
"loss": 0.5242,
"step": 216
},
{
"epoch": 2.0798722044728435,
"grad_norm": 0.7659483551979065,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.5336,
"step": 217
},
{
"epoch": 2.0894568690095845,
"grad_norm": 0.7429843544960022,
"learning_rate": 2.532458909678266e-06,
"loss": 0.5022,
"step": 218
},
{
"epoch": 2.099041533546326,
"grad_norm": 0.712540328502655,
"learning_rate": 2.483822884971e-06,
"loss": 0.5348,
"step": 219
},
{
"epoch": 2.108626198083067,
"grad_norm": 0.6858490705490112,
"learning_rate": 2.43550361297047e-06,
"loss": 0.5438,
"step": 220
},
{
"epoch": 2.1182108626198084,
"grad_norm": 0.7074081301689148,
"learning_rate": 2.387507176420256e-06,
"loss": 0.5179,
"step": 221
},
{
"epoch": 2.1277955271565494,
"grad_norm": 0.6986274719238281,
"learning_rate": 2.339839617423318e-06,
"loss": 0.5275,
"step": 222
},
{
"epoch": 2.137380191693291,
"grad_norm": 0.705319881439209,
"learning_rate": 2.2925069366813718e-06,
"loss": 0.5336,
"step": 223
},
{
"epoch": 2.146964856230032,
"grad_norm": 0.7064869999885559,
"learning_rate": 2.245515092739488e-06,
"loss": 0.5404,
"step": 224
},
{
"epoch": 2.1565495207667733,
"grad_norm": 0.73943030834198,
"learning_rate": 2.1988700012359865e-06,
"loss": 0.5173,
"step": 225
},
{
"epoch": 2.1661341853035143,
"grad_norm": 0.6814712882041931,
"learning_rate": 2.1525775341577404e-06,
"loss": 0.5124,
"step": 226
},
{
"epoch": 2.1757188498402558,
"grad_norm": 0.6799867749214172,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.5087,
"step": 227
},
{
"epoch": 2.1853035143769968,
"grad_norm": 0.6821707487106323,
"learning_rate": 2.061073738537635e-06,
"loss": 0.5265,
"step": 228
},
{
"epoch": 2.194888178913738,
"grad_norm": 0.6849046349525452,
"learning_rate": 2.0158739290874822e-06,
"loss": 0.531,
"step": 229
},
{
"epoch": 2.2044728434504792,
"grad_norm": 0.6618659496307373,
"learning_rate": 1.971049780795901e-06,
"loss": 0.5036,
"step": 230
},
{
"epoch": 2.2140575079872207,
"grad_norm": 0.6809960603713989,
"learning_rate": 1.9266069364176144e-06,
"loss": 0.535,
"step": 231
},
{
"epoch": 2.2236421725239617,
"grad_norm": 0.690485954284668,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.547,
"step": 232
},
{
"epoch": 2.2332268370607027,
"grad_norm": 0.6946452260017395,
"learning_rate": 1.838887489710452e-06,
"loss": 0.5082,
"step": 233
},
{
"epoch": 2.242811501597444,
"grad_norm": 0.6604838967323303,
"learning_rate": 1.7956219300748796e-06,
"loss": 0.5294,
"step": 234
},
{
"epoch": 2.252396166134185,
"grad_norm": 0.663303554058075,
"learning_rate": 1.7527597583490825e-06,
"loss": 0.5304,
"step": 235
},
{
"epoch": 2.2619808306709266,
"grad_norm": 0.6427473425865173,
"learning_rate": 1.7103063703014372e-06,
"loss": 0.533,
"step": 236
},
{
"epoch": 2.2715654952076676,
"grad_norm": 0.6572959423065186,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.5402,
"step": 237
},
{
"epoch": 2.281150159744409,
"grad_norm": 0.6577916741371155,
"learning_rate": 1.6266472703396286e-06,
"loss": 0.5404,
"step": 238
},
{
"epoch": 2.29073482428115,
"grad_norm": 0.6826768517494202,
"learning_rate": 1.5854520899759656e-06,
"loss": 0.5312,
"step": 239
},
{
"epoch": 2.3003194888178915,
"grad_norm": 0.6547624468803406,
"learning_rate": 1.544686755065677e-06,
"loss": 0.5282,
"step": 240
},
{
"epoch": 2.3099041533546325,
"grad_norm": 0.6991428732872009,
"learning_rate": 1.5043563974137132e-06,
"loss": 0.5161,
"step": 241
},
{
"epoch": 2.319488817891374,
"grad_norm": 0.6671050786972046,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.5365,
"step": 242
},
{
"epoch": 2.329073482428115,
"grad_norm": 0.6593329906463623,
"learning_rate": 1.4250208666766235e-06,
"loss": 0.4826,
"step": 243
},
{
"epoch": 2.3386581469648564,
"grad_norm": 0.6457230448722839,
"learning_rate": 1.3860256808630429e-06,
"loss": 0.5275,
"step": 244
},
{
"epoch": 2.3482428115015974,
"grad_norm": 0.667349100112915,
"learning_rate": 1.3474854455936126e-06,
"loss": 0.4936,
"step": 245
},
{
"epoch": 2.357827476038339,
"grad_norm": 0.6810527443885803,
"learning_rate": 1.3094050125632973e-06,
"loss": 0.5348,
"step": 246
},
{
"epoch": 2.36741214057508,
"grad_norm": 0.6994328498840332,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.5323,
"step": 247
},
{
"epoch": 2.376996805111821,
"grad_norm": 0.6719757318496704,
"learning_rate": 1.234642669981946e-06,
"loss": 0.5403,
"step": 248
},
{
"epoch": 2.3865814696485623,
"grad_norm": 0.6738424897193909,
"learning_rate": 1.1979701719998454e-06,
"loss": 0.5103,
"step": 249
},
{
"epoch": 2.3961661341853033,
"grad_norm": 0.66436767578125,
"learning_rate": 1.1617762982099446e-06,
"loss": 0.5267,
"step": 250
},
{
"epoch": 2.405750798722045,
"grad_norm": 0.6517850756645203,
"learning_rate": 1.1260656049319957e-06,
"loss": 0.5514,
"step": 251
},
{
"epoch": 2.415335463258786,
"grad_norm": 0.6427007913589478,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.5151,
"step": 252
},
{
"epoch": 2.4249201277955272,
"grad_norm": 0.692989706993103,
"learning_rate": 1.0561116804955451e-06,
"loss": 0.5377,
"step": 253
},
{
"epoch": 2.4345047923322682,
"grad_norm": 0.6546573638916016,
"learning_rate": 1.0218772555910955e-06,
"loss": 0.4958,
"step": 254
},
{
"epoch": 2.4440894568690097,
"grad_norm": 0.6840081810951233,
"learning_rate": 9.881436225981107e-07,
"loss": 0.5125,
"step": 255
},
{
"epoch": 2.4536741214057507,
"grad_norm": 0.6849152445793152,
"learning_rate": 9.549150281252633e-07,
"loss": 0.4951,
"step": 256
},
{
"epoch": 2.463258785942492,
"grad_norm": 0.6501500606536865,
"learning_rate": 9.221956552036992e-07,
"loss": 0.5078,
"step": 257
},
{
"epoch": 2.472843450479233,
"grad_norm": 0.6612105965614319,
"learning_rate": 8.899896227604509e-07,
"loss": 0.4934,
"step": 258
},
{
"epoch": 2.4824281150159746,
"grad_norm": 0.688321590423584,
"learning_rate": 8.58300985099918e-07,
"loss": 0.5024,
"step": 259
},
{
"epoch": 2.4920127795527156,
"grad_norm": 0.6561516523361206,
"learning_rate": 8.271337313934869e-07,
"loss": 0.5029,
"step": 260
},
{
"epoch": 2.501597444089457,
"grad_norm": 0.6125414371490479,
"learning_rate": 7.964917851773496e-07,
"loss": 0.5306,
"step": 261
},
{
"epoch": 2.511182108626198,
"grad_norm": 0.6592358946800232,
"learning_rate": 7.663790038585794e-07,
"loss": 0.5223,
"step": 262
},
{
"epoch": 2.520766773162939,
"grad_norm": 0.6690595149993896,
"learning_rate": 7.367991782295392e-07,
"loss": 0.4856,
"step": 263
},
{
"epoch": 2.5303514376996805,
"grad_norm": 0.6796962022781372,
"learning_rate": 7.077560319906696e-07,
"loss": 0.5343,
"step": 264
},
{
"epoch": 2.539936102236422,
"grad_norm": 0.6726747155189514,
"learning_rate": 6.792532212817271e-07,
"loss": 0.4966,
"step": 265
},
{
"epoch": 2.549520766773163,
"grad_norm": 0.6816850304603577,
"learning_rate": 6.512943342215234e-07,
"loss": 0.5373,
"step": 266
},
{
"epoch": 2.559105431309904,
"grad_norm": 0.6727045774459839,
"learning_rate": 6.238828904562316e-07,
"loss": 0.5073,
"step": 267
},
{
"epoch": 2.5686900958466454,
"grad_norm": 0.6780251860618591,
"learning_rate": 5.9702234071631e-07,
"loss": 0.5322,
"step": 268
},
{
"epoch": 2.5782747603833864,
"grad_norm": 0.6769065260887146,
"learning_rate": 5.707160663821009e-07,
"loss": 0.5372,
"step": 269
},
{
"epoch": 2.587859424920128,
"grad_norm": 0.6582074761390686,
"learning_rate": 5.449673790581611e-07,
"loss": 0.5369,
"step": 270
},
{
"epoch": 2.597444089456869,
"grad_norm": 0.6327592134475708,
"learning_rate": 5.197795201563744e-07,
"loss": 0.5214,
"step": 271
},
{
"epoch": 2.6070287539936103,
"grad_norm": 0.6320553421974182,
"learning_rate": 4.951556604879049e-07,
"loss": 0.5263,
"step": 272
},
{
"epoch": 2.6166134185303513,
"grad_norm": 0.6550769209861755,
"learning_rate": 4.710988998640298e-07,
"loss": 0.5155,
"step": 273
},
{
"epoch": 2.626198083067093,
"grad_norm": 0.6145337820053101,
"learning_rate": 4.4761226670592074e-07,
"loss": 0.5061,
"step": 274
},
{
"epoch": 2.635782747603834,
"grad_norm": 0.6379536986351013,
"learning_rate": 4.2469871766340096e-07,
"loss": 0.5011,
"step": 275
},
{
"epoch": 2.6453674121405752,
"grad_norm": 0.6818909645080566,
"learning_rate": 4.0236113724274716e-07,
"loss": 0.5266,
"step": 276
},
{
"epoch": 2.6549520766773163,
"grad_norm": 0.6908199191093445,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.513,
"step": 277
},
{
"epoch": 2.6645367412140573,
"grad_norm": 0.6766000390052795,
"learning_rate": 3.5942505740480583e-07,
"loss": 0.5085,
"step": 278
},
{
"epoch": 2.6741214057507987,
"grad_norm": 0.668457567691803,
"learning_rate": 3.3883196305992906e-07,
"loss": 0.5157,
"step": 279
},
{
"epoch": 2.68370607028754,
"grad_norm": 0.6620476841926575,
"learning_rate": 3.18825646801314e-07,
"loss": 0.5369,
"step": 280
},
{
"epoch": 2.693290734824281,
"grad_norm": 0.6888989806175232,
"learning_rate": 2.9940862715390483e-07,
"loss": 0.5294,
"step": 281
},
{
"epoch": 2.702875399361022,
"grad_norm": 0.6750389337539673,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.4868,
"step": 282
},
{
"epoch": 2.7124600638977636,
"grad_norm": 0.6632707715034485,
"learning_rate": 2.6235218056235633e-07,
"loss": 0.4873,
"step": 283
},
{
"epoch": 2.722044728434505,
"grad_norm": 0.6507512331008911,
"learning_rate": 2.447174185242324e-07,
"loss": 0.5128,
"step": 284
},
{
"epoch": 2.731629392971246,
"grad_norm": 0.6406328678131104,
"learning_rate": 2.276812823220964e-07,
"loss": 0.4996,
"step": 285
},
{
"epoch": 2.741214057507987,
"grad_norm": 0.6203620433807373,
"learning_rate": 2.1124591657534776e-07,
"loss": 0.5323,
"step": 286
},
{
"epoch": 2.7507987220447285,
"grad_norm": 0.6786451935768127,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.5221,
"step": 287
},
{
"epoch": 2.7603833865814695,
"grad_norm": 0.6705276370048523,
"learning_rate": 1.801856965207338e-07,
"loss": 0.4962,
"step": 288
},
{
"epoch": 2.769968051118211,
"grad_norm": 0.6476415991783142,
"learning_rate": 1.6556475227496816e-07,
"loss": 0.5175,
"step": 289
},
{
"epoch": 2.779552715654952,
"grad_norm": 0.67632657289505,
"learning_rate": 1.5155239811656562e-07,
"loss": 0.5148,
"step": 290
},
{
"epoch": 2.7891373801916934,
"grad_norm": 0.6424274444580078,
"learning_rate": 1.3815039801161723e-07,
"loss": 0.5182,
"step": 291
},
{
"epoch": 2.7987220447284344,
"grad_norm": 0.6566675305366516,
"learning_rate": 1.253604390908819e-07,
"loss": 0.522,
"step": 292
},
{
"epoch": 2.8083067092651754,
"grad_norm": 0.6124169826507568,
"learning_rate": 1.1318413143740436e-07,
"loss": 0.5303,
"step": 293
},
{
"epoch": 2.817891373801917,
"grad_norm": 0.6645427942276001,
"learning_rate": 1.0162300788382263e-07,
"loss": 0.5181,
"step": 294
},
{
"epoch": 2.8274760383386583,
"grad_norm": 0.64972323179245,
"learning_rate": 9.0678523819408e-08,
"loss": 0.5151,
"step": 295
},
{
"epoch": 2.8370607028753994,
"grad_norm": 0.6525695323944092,
"learning_rate": 8.035205700685167e-08,
"loss": 0.5055,
"step": 296
},
{
"epoch": 2.8466453674121404,
"grad_norm": 0.6717900037765503,
"learning_rate": 7.064490740882057e-08,
"loss": 0.532,
"step": 297
},
{
"epoch": 2.856230031948882,
"grad_norm": 0.6631027460098267,
"learning_rate": 6.15582970243117e-08,
"loss": 0.5031,
"step": 298
},
{
"epoch": 2.8658146964856233,
"grad_norm": 0.6349170207977295,
"learning_rate": 5.3093369734816824e-08,
"loss": 0.4814,
"step": 299
},
{
"epoch": 2.8753993610223643,
"grad_norm": 0.6449158191680908,
"learning_rate": 4.52511911603265e-08,
"loss": 0.5245,
"step": 300
},
{
"epoch": 2.8849840255591053,
"grad_norm": 0.6388517022132874,
"learning_rate": 3.8032748525179684e-08,
"loss": 0.5301,
"step": 301
},
{
"epoch": 2.8945686900958467,
"grad_norm": 0.6567696332931519,
"learning_rate": 3.143895053378698e-08,
"loss": 0.5197,
"step": 302
},
{
"epoch": 2.9041533546325877,
"grad_norm": 0.6263684034347534,
"learning_rate": 2.547062725623828e-08,
"loss": 0.5383,
"step": 303
},
{
"epoch": 2.913738019169329,
"grad_norm": 0.6724746227264404,
"learning_rate": 2.012853002380466e-08,
"loss": 0.5375,
"step": 304
},
{
"epoch": 2.92332268370607,
"grad_norm": 0.6429194211959839,
"learning_rate": 1.541333133436018e-08,
"loss": 0.5263,
"step": 305
},
{
"epoch": 2.9329073482428116,
"grad_norm": 0.6564125418663025,
"learning_rate": 1.132562476771959e-08,
"loss": 0.5195,
"step": 306
},
{
"epoch": 2.9424920127795526,
"grad_norm": 0.6608335375785828,
"learning_rate": 7.865924910916977e-09,
"loss": 0.5298,
"step": 307
},
{
"epoch": 2.952076677316294,
"grad_norm": 0.6398037672042847,
"learning_rate": 5.034667293427053e-09,
"loss": 0.5329,
"step": 308
},
{
"epoch": 2.961661341853035,
"grad_norm": 0.6140774488449097,
"learning_rate": 2.8322083323334417e-09,
"loss": 0.5038,
"step": 309
},
{
"epoch": 2.9712460063897765,
"grad_norm": 0.6483182311058044,
"learning_rate": 1.2588252874673469e-09,
"loss": 0.5385,
"step": 310
},
{
"epoch": 2.9808306709265175,
"grad_norm": 0.6684054732322693,
"learning_rate": 3.147162264971471e-10,
"loss": 0.4874,
"step": 311
},
{
"epoch": 2.9904153354632586,
"grad_norm": 0.635933518409729,
"learning_rate": 0.0,
"loss": 0.5275,
"step": 312
},
{
"epoch": 2.9904153354632586,
"step": 312,
"total_flos": 1.329419497532375e+17,
"train_loss": 0.0,
"train_runtime": 6.2065,
"train_samples_per_second": 4833.679,
"train_steps_per_second": 50.27
}
],
"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": 1.329419497532375e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}