| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9952, | |
| "eval_steps": 500, | |
| "global_step": 312, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0096, | |
| "grad_norm": 6.13106107711792, | |
| "learning_rate": 3.125e-07, | |
| "loss": 0.9999, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 6.25918436050415, | |
| "learning_rate": 6.25e-07, | |
| "loss": 1.0304, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 6.324011325836182, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 1.0222, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.183999061584473, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.9938, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 6.033788681030273, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 1.021, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 5.652398586273193, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.9985, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 4.792149066925049, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 1.0012, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 4.491700649261475, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.9712, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 2.512334108352661, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.9247, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.392815589904785, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.8943, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 2.1595864295959473, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.8855, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 3.6345415115356445, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.8777, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 3.9351446628570557, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.8991, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 3.714717388153076, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.8688, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 3.7493085861206055, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.8819, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.5678951740264893, | |
| "learning_rate": 5e-06, | |
| "loss": 0.8437, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 2.208275556564331, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.8219, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 1.9158337116241455, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.8147, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1824, | |
| "grad_norm": 1.6533466577529907, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.8078, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.366206407546997, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.7702, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2016, | |
| "grad_norm": 1.286657691001892, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.7648, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 1.348372220993042, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.7652, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2208, | |
| "grad_norm": 1.177068829536438, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.7526, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.1750667095184326, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7703, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.0557239055633545, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.7435, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 1.0049221515655518, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.7321, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2592, | |
| "grad_norm": 1.1623879671096802, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.7447, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.7909222841262817, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.7158, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2784, | |
| "grad_norm": 0.9174373745918274, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.7321, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.8606329560279846, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.71, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2976, | |
| "grad_norm": 0.8403401970863342, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.6998, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8139435648918152, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7005, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3168, | |
| "grad_norm": 0.7812791466712952, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.6899, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 0.7286513447761536, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.7092, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.6920162439346313, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.6625, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.7227463126182556, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.7081, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3552, | |
| "grad_norm": 0.6858569383621216, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.6873, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 0.6450650691986084, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.6755, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3744, | |
| "grad_norm": 0.6473551392555237, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.6843, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.732536256313324, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6978, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3936, | |
| "grad_norm": 0.6281261444091797, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.6848, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 0.6776072382926941, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.7014, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4128, | |
| "grad_norm": 0.5990509390830994, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.7001, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.7227262258529663, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.6843, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.6668912768363953, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.6739, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 0.5989091992378235, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.6579, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4512, | |
| "grad_norm": 0.6381036043167114, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.6582, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.6903780102729797, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.6886, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4704, | |
| "grad_norm": 0.6363586783409119, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.6885, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.6252055168151855, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.6543, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4896, | |
| "grad_norm": 0.7057021260261536, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.6791, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.663014829158783, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.6623, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5088, | |
| "grad_norm": 0.7666405439376831, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.6689, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 0.6534073948860168, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.6593, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.6731073260307312, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.6787, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.7624923586845398, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6816, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5472, | |
| "grad_norm": 0.7006317973136902, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.6735, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 0.5917409658432007, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.6363, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5664, | |
| "grad_norm": 0.6200292110443115, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.6489, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.7014618515968323, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.6723, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5856, | |
| "grad_norm": 0.5875418782234192, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.6797, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 0.7326757907867432, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.6484, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6048, | |
| "grad_norm": 0.6782388687133789, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.6349, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.6133514642715454, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6621, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.7363903522491455, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.6636, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 0.7007447481155396, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.6468, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6432, | |
| "grad_norm": 0.6359402537345886, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.6725, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.6885963678359985, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.6424, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6624, | |
| "grad_norm": 0.7374150156974792, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.616, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.6072792410850525, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.6425, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6816, | |
| "grad_norm": 0.5575535297393799, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.6328, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.6608112454414368, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6457, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7008, | |
| "grad_norm": 0.5828973054885864, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.6586, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 0.548630952835083, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.6569, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.6277467012405396, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.6318, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.5508156418800354, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.6332, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7392, | |
| "grad_norm": 0.5723240375518799, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.6531, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 0.5478023290634155, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.6282, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7584, | |
| "grad_norm": 0.539486289024353, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.6342, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.53630530834198, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6503, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7776, | |
| "grad_norm": 0.5162023305892944, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.6467, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 0.623972475528717, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.6506, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7968, | |
| "grad_norm": 0.5885581970214844, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.6383, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.6666449308395386, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.6441, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.5746261477470398, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.6523, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 0.5863617658615112, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.64, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8352, | |
| "grad_norm": 0.6225231289863586, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.6311, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.6017730832099915, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6507, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8544, | |
| "grad_norm": 0.71891188621521, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.6423, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.7425549030303955, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.6236, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8736, | |
| "grad_norm": 0.5610062479972839, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.6414, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.625220775604248, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.6517, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8928, | |
| "grad_norm": 0.7597525119781494, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.6541, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 0.5740223526954651, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.6198, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.7073938250541687, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.6456, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.7561060190200806, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.6739, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9312, | |
| "grad_norm": 0.5501155257225037, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.6573, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 0.6279442310333252, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.644, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9504, | |
| "grad_norm": 0.5732079744338989, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.629, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6396852135658264, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.6541, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9696, | |
| "grad_norm": 0.5268051028251648, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.6196, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 0.5126073360443115, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.639, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9888, | |
| "grad_norm": 0.5734632611274719, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.647, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.6213233470916748, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.6526, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 1.319844126701355, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 1.1556, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0176, | |
| "grad_norm": 0.47418418526649475, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.5402, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0272, | |
| "grad_norm": 0.528281033039093, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.5966, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.5944254398345947, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.5965, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0464, | |
| "grad_norm": 0.5833683609962463, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.6116, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 0.5533478260040283, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.5668, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0656, | |
| "grad_norm": 0.6013179421424866, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.559, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.5827232003211975, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5841, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0848, | |
| "grad_norm": 0.5553967356681824, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.5868, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0944, | |
| "grad_norm": 0.6195008158683777, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.5858, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 0.6123397946357727, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.6351, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.49623173475265503, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.5402, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1232, | |
| "grad_norm": 0.6296162009239197, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.623, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1328, | |
| "grad_norm": 0.6342450380325317, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.5498, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1424, | |
| "grad_norm": 0.5587097406387329, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.5662, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.5478476881980896, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.5937, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1616, | |
| "grad_norm": 0.6309380531311035, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.586, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1712, | |
| "grad_norm": 0.6445760130882263, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.6085, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1808, | |
| "grad_norm": 0.547279417514801, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.5853, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.756352424621582, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.6045, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.5580184459686279, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.5337, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.2096, | |
| "grad_norm": 0.5867987871170044, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.6091, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2192, | |
| "grad_norm": 0.5563951730728149, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.5361, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.6019569039344788, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.6138, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2384, | |
| "grad_norm": 0.48974552750587463, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.5437, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 0.5434426665306091, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.6, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2576, | |
| "grad_norm": 0.5539755821228027, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.5984, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.4822545349597931, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.5065, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2768, | |
| "grad_norm": 0.5552537441253662, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.6292, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2864, | |
| "grad_norm": 0.5087763667106628, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.5346, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 0.6405339241027832, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.5902, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.6019776463508606, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.6285, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3152, | |
| "grad_norm": 0.5158368945121765, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.5608, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3248, | |
| "grad_norm": 0.4856293499469757, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.5469, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3344, | |
| "grad_norm": 0.553087055683136, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.6132, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.4461369514465332, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.5618, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3536000000000001, | |
| "grad_norm": 0.4760764539241791, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.5657, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3632, | |
| "grad_norm": 0.5615460276603699, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.594, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3728, | |
| "grad_norm": 0.512578547000885, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.621, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.5107195973396301, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.592, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 0.5414605140686035, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.6066, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.4016, | |
| "grad_norm": 0.5117338299751282, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.5357, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4112, | |
| "grad_norm": 0.582869827747345, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.6127, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.518729031085968, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.5928, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.4304000000000001, | |
| "grad_norm": 0.5442705750465393, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.5415, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.6551291346549988, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.5908, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4496, | |
| "grad_norm": 0.4694753587245941, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.5786, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5337488055229187, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5441, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4687999999999999, | |
| "grad_norm": 0.6417204737663269, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.6405, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4784, | |
| "grad_norm": 0.5273697972297668, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.5568, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 0.6081464290618896, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.6345, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.5175541043281555, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.5135, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5072, | |
| "grad_norm": 0.5481940507888794, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.6327, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5168, | |
| "grad_norm": 0.5984399914741516, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.6223, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5264, | |
| "grad_norm": 0.4837677478790283, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.5105, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.5733755230903625, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.669, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5455999999999999, | |
| "grad_norm": 0.5964936017990112, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.6035, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5552000000000001, | |
| "grad_norm": 0.48552206158638, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.5259, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5648, | |
| "grad_norm": 0.487074077129364, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.5745, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.5122261643409729, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.5844, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 0.6647276282310486, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.551, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5936, | |
| "grad_norm": 0.5472864508628845, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.608, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6032, | |
| "grad_norm": 0.6124144792556763, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.5471, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.6829537153244019, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5739, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6223999999999998, | |
| "grad_norm": 0.467629998922348, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.5591, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.584351658821106, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.5873, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6416, | |
| "grad_norm": 0.7098246812820435, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.5877, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.5116992592811584, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5449, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.6608, | |
| "grad_norm": 0.5698950290679932, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.6065, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6703999999999999, | |
| "grad_norm": 0.5616971254348755, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.532, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.587088406085968, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.5812, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.6302618980407715, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.6103, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6992, | |
| "grad_norm": 0.43855246901512146, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.5769, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.7088, | |
| "grad_norm": 0.5838032960891724, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.5979, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7184, | |
| "grad_norm": 0.5378840565681458, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.541, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.48658180236816406, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.6072, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7376, | |
| "grad_norm": 0.5207194089889526, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.63, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7471999999999999, | |
| "grad_norm": 0.4985181391239166, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.5362, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7568000000000001, | |
| "grad_norm": 0.42738834023475647, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.5839, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.4418393075466156, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.5464, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 0.47566258907318115, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.5903, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7856, | |
| "grad_norm": 0.45583727955818176, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.5968, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7952, | |
| "grad_norm": 0.5059268474578857, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.621, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.5112723708152771, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.5592, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8144, | |
| "grad_norm": 0.5024119019508362, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.5809, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.483896404504776, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.5227, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8336000000000001, | |
| "grad_norm": 0.4978603422641754, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.6031, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.44853779673576355, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.4972, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8528, | |
| "grad_norm": 0.5778761506080627, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.6452, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8624, | |
| "grad_norm": 0.4285795986652374, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.539, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 0.44635048508644104, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.5774, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.4324394762516022, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.6409, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8912, | |
| "grad_norm": 0.4505009353160858, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.5501, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.9008, | |
| "grad_norm": 0.525919497013092, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.525, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9104, | |
| "grad_norm": 0.4947689473628998, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.6062, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4856187105178833, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.6331, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9296, | |
| "grad_norm": 0.5208802223205566, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.5422, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9392, | |
| "grad_norm": 0.4811985194683075, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.6008, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9487999999999999, | |
| "grad_norm": 0.4215378165245056, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.5425, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.4693278670310974, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.6178, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 0.4558577835559845, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.5553, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9776, | |
| "grad_norm": 0.44367337226867676, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.5733, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9872, | |
| "grad_norm": 0.42901933193206787, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.5401, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.45983821153640747, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.6139, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0064, | |
| "grad_norm": 1.3911380767822266, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 1.0278, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.5409606099128723, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.5641, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.0256, | |
| "grad_norm": 0.5035486817359924, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.548, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.4612354636192322, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.4851, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0448, | |
| "grad_norm": 0.4524441361427307, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.5971, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0544, | |
| "grad_norm": 0.43905532360076904, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.5399, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.4878225028514862, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.5861, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.39097368717193604, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.4285, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0832, | |
| "grad_norm": 0.4459839165210724, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.6043, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0928, | |
| "grad_norm": 0.4151725172996521, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.5104, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.1024, | |
| "grad_norm": 0.4535321891307831, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.5446, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.41973429918289185, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.504, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1216, | |
| "grad_norm": 0.4929872453212738, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.5642, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1312, | |
| "grad_norm": 0.4442804455757141, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.5209, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.1408, | |
| "grad_norm": 0.3861082196235657, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.494, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.43212220072746277, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.5516, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.4358077645301819, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.5417, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1696, | |
| "grad_norm": 0.4068072438240051, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.5121, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1792, | |
| "grad_norm": 0.46814894676208496, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.5793, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.3898715674877167, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.4822, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.1984, | |
| "grad_norm": 0.45388954877853394, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.5658, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.4347994327545166, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.5517, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2176, | |
| "grad_norm": 0.4096216559410095, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.4754, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.42773646116256714, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.5271, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2368, | |
| "grad_norm": 0.44084692001342773, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.5979, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.2464, | |
| "grad_norm": 0.3811928629875183, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.5079, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.452578604221344, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.5695, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.4129106104373932, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.5663, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2752, | |
| "grad_norm": 0.4045464098453522, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.5348, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.2848, | |
| "grad_norm": 0.41679397225379944, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.5397, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.2944, | |
| "grad_norm": 0.4080122113227844, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.5516, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.34944072365760803, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.4603, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3136, | |
| "grad_norm": 0.4407574534416199, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.555, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.3232, | |
| "grad_norm": 0.4524279534816742, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.5663, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.3327999999999998, | |
| "grad_norm": 0.38331395387649536, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.5196, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.38209086656570435, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.4953, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.41388967633247375, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.5396, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.3616, | |
| "grad_norm": 0.4106352925300598, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.5549, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.3712, | |
| "grad_norm": 0.4063715636730194, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.4519, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.4751112461090088, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.5749, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3904, | |
| "grad_norm": 0.39849382638931274, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.5447, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.3605557084083557, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.4826, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.4096, | |
| "grad_norm": 0.41893282532691956, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.5163, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.40984004735946655, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.5874, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4288, | |
| "grad_norm": 0.3629111051559448, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.5037, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4384, | |
| "grad_norm": 0.3851902484893799, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.5054, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.4868093430995941, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.5983, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.3716026544570923, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.4513, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.4672, | |
| "grad_norm": 0.4426417946815491, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.5971, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.4768, | |
| "grad_norm": 0.37559273838996887, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.4924, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4864, | |
| "grad_norm": 0.39406782388687134, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.5186, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.39286789298057556, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.5679, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.5056000000000003, | |
| "grad_norm": 0.3801133632659912, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.5575, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.5152, | |
| "grad_norm": 0.3806913495063782, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.5197, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.5248, | |
| "grad_norm": 0.3847755789756775, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.5112, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.4005506932735443, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.5792, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.3597860336303711, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.473, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.5536, | |
| "grad_norm": 0.43030428886413574, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.5748, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.5632, | |
| "grad_norm": 0.38729578256607056, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.5248, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.39844393730163574, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.5529, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5824, | |
| "grad_norm": 0.3908894658088684, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.5129, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.4160621762275696, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.5636, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.6016, | |
| "grad_norm": 0.3664015531539917, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.5184, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.3714524507522583, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.5321, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6208, | |
| "grad_norm": 0.3544293940067291, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.4959, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.6304, | |
| "grad_norm": 0.39069050550460815, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.5584, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.3826994299888611, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.5119, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.38033396005630493, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.5327, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6592000000000002, | |
| "grad_norm": 0.3969111740589142, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.5775, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6688, | |
| "grad_norm": 0.37891885638237, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.4935, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6784, | |
| "grad_norm": 0.4370158314704895, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.5695, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.37920939922332764, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.5126, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.6976, | |
| "grad_norm": 0.4022983908653259, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.4923, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.7072000000000003, | |
| "grad_norm": 0.4108829200267792, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.5353, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7168, | |
| "grad_norm": 0.3764873445034027, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.5277, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.40391191840171814, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.5684, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.3837086856365204, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.5248, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.7456, | |
| "grad_norm": 0.4416448175907135, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.4783, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7552, | |
| "grad_norm": 0.3758324384689331, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.5613, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.3901281952857971, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.593, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.7744, | |
| "grad_norm": 0.3793884515762329, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.5209, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.3416971266269684, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.4898, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7936, | |
| "grad_norm": 0.38408684730529785, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.5427, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.4025774598121643, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.5543, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8128, | |
| "grad_norm": 0.40174293518066406, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.5416, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.8224, | |
| "grad_norm": 0.4046119153499603, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.5915, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.3578289747238159, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.4777, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.39651331305503845, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.5158, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8512, | |
| "grad_norm": 0.3694838583469391, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.5431, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.8608000000000002, | |
| "grad_norm": 0.3862026333808899, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.5349, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8704, | |
| "grad_norm": 0.35192471742630005, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.4948, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.37490689754486084, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.5479, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8895999999999997, | |
| "grad_norm": 0.3634902536869049, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.4839, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8992, | |
| "grad_norm": 0.3801132142543793, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.5299, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9088000000000003, | |
| "grad_norm": 0.4453577697277069, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.5644, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.34870657324790955, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.471, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.35863104462623596, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.5362, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9375999999999998, | |
| "grad_norm": 0.3896194100379944, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.5445, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9472, | |
| "grad_norm": 0.38962915539741516, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.5263, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.4067755937576294, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.5222, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.9664, | |
| "grad_norm": 0.3524220883846283, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.4851, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.36281532049179077, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.5688, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9856, | |
| "grad_norm": 0.39132219552993774, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.5176, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.3681107759475708, | |
| "learning_rate": 0.0, | |
| "loss": 0.5461, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "step": 312, | |
| "total_flos": 314263852400640.0, | |
| "train_loss": 0.6130749200208064, | |
| "train_runtime": 28960.9471, | |
| "train_samples_per_second": 1.036, | |
| "train_steps_per_second": 0.011 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "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": 314263852400640.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |