| { | |
| "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": 6.164039611816406, | |
| "learning_rate": 3.125e-07, | |
| "loss": 1.008, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 6.292746067047119, | |
| "learning_rate": 6.25e-07, | |
| "loss": 1.0149, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 6.367922306060791, | |
| "learning_rate": 9.375000000000001e-07, | |
| "loss": 1.026, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.2663187980651855, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.0154, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 6.17635440826416, | |
| "learning_rate": 1.5625e-06, | |
| "loss": 1.0091, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 5.536340236663818, | |
| "learning_rate": 1.8750000000000003e-06, | |
| "loss": 0.9624, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 4.604569435119629, | |
| "learning_rate": 2.1875000000000002e-06, | |
| "loss": 0.95, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 4.306362152099609, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.9617, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 2.4510245323181152, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 0.921, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 2.3514862060546875, | |
| "learning_rate": 3.125e-06, | |
| "loss": 0.8574, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 2.2417752742767334, | |
| "learning_rate": 3.4375e-06, | |
| "loss": 0.9179, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 3.4170384407043457, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.8742, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 4.054645538330078, | |
| "learning_rate": 4.0625000000000005e-06, | |
| "loss": 0.8919, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 3.9020988941192627, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 0.8858, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 3.5906124114990234, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "loss": 0.8746, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 2.614440679550171, | |
| "learning_rate": 5e-06, | |
| "loss": 0.866, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 2.242814540863037, | |
| "learning_rate": 5.3125e-06, | |
| "loss": 0.8143, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1728, | |
| "grad_norm": 1.921581745147705, | |
| "learning_rate": 5.625e-06, | |
| "loss": 0.8119, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1824, | |
| "grad_norm": 1.758897066116333, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.8001, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.4624104499816895, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.7916, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2016, | |
| "grad_norm": 1.2528955936431885, | |
| "learning_rate": 6.5625e-06, | |
| "loss": 0.7971, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2112, | |
| "grad_norm": 1.2485564947128296, | |
| "learning_rate": 6.875e-06, | |
| "loss": 0.7607, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2208, | |
| "grad_norm": 1.1729601621627808, | |
| "learning_rate": 7.1875e-06, | |
| "loss": 0.7394, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.1565651893615723, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7683, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.0047582387924194, | |
| "learning_rate": 7.8125e-06, | |
| "loss": 0.7456, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2496, | |
| "grad_norm": 1.0549741983413696, | |
| "learning_rate": 8.125000000000001e-06, | |
| "loss": 0.7246, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2592, | |
| "grad_norm": 1.0340930223464966, | |
| "learning_rate": 8.4375e-06, | |
| "loss": 0.7379, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 0.805397093296051, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.7172, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2784, | |
| "grad_norm": 0.8800445795059204, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.7218, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 0.9208349585533142, | |
| "learning_rate": 9.375000000000001e-06, | |
| "loss": 0.7242, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2976, | |
| "grad_norm": 0.7508711814880371, | |
| "learning_rate": 9.6875e-06, | |
| "loss": 0.6823, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8096818327903748, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7239, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3168, | |
| "grad_norm": 0.7590523362159729, | |
| "learning_rate": 9.999685283773504e-06, | |
| "loss": 0.6874, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3264, | |
| "grad_norm": 0.7550079226493835, | |
| "learning_rate": 9.998741174712534e-06, | |
| "loss": 0.7055, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 0.7203152775764465, | |
| "learning_rate": 9.997167791667668e-06, | |
| "loss": 0.6942, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.7001503705978394, | |
| "learning_rate": 9.994965332706574e-06, | |
| "loss": 0.7124, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3552, | |
| "grad_norm": 0.7050896286964417, | |
| "learning_rate": 9.992134075089085e-06, | |
| "loss": 0.7075, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3648, | |
| "grad_norm": 0.6661181449890137, | |
| "learning_rate": 9.98867437523228e-06, | |
| "loss": 0.7181, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.3744, | |
| "grad_norm": 0.6402677893638611, | |
| "learning_rate": 9.984586668665641e-06, | |
| "loss": 0.6979, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.6823920607566833, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6876, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3936, | |
| "grad_norm": 0.6096172332763672, | |
| "learning_rate": 9.974529372743762e-06, | |
| "loss": 0.6851, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.4032, | |
| "grad_norm": 0.5798601508140564, | |
| "learning_rate": 9.968561049466214e-06, | |
| "loss": 0.6735, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.4128, | |
| "grad_norm": 0.6932426691055298, | |
| "learning_rate": 9.961967251474823e-06, | |
| "loss": 0.6874, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.5851962566375732, | |
| "learning_rate": 9.954748808839675e-06, | |
| "loss": 0.6558, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 0.6512772440910339, | |
| "learning_rate": 9.946906630265184e-06, | |
| "loss": 0.6932, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4416, | |
| "grad_norm": 0.6203194856643677, | |
| "learning_rate": 9.938441702975689e-06, | |
| "loss": 0.6569, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4512, | |
| "grad_norm": 0.5843777656555176, | |
| "learning_rate": 9.92935509259118e-06, | |
| "loss": 0.6916, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.6552630662918091, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.6927, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4704, | |
| "grad_norm": 0.5897120237350464, | |
| "learning_rate": 9.909321476180594e-06, | |
| "loss": 0.6418, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.6383408308029175, | |
| "learning_rate": 9.898376992116179e-06, | |
| "loss": 0.6697, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4896, | |
| "grad_norm": 0.5719790458679199, | |
| "learning_rate": 9.886815868562596e-06, | |
| "loss": 0.6535, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.6388934850692749, | |
| "learning_rate": 9.874639560909118e-06, | |
| "loss": 0.6738, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5088, | |
| "grad_norm": 0.6141742467880249, | |
| "learning_rate": 9.861849601988384e-06, | |
| "loss": 0.6844, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5184, | |
| "grad_norm": 0.710845410823822, | |
| "learning_rate": 9.848447601883436e-06, | |
| "loss": 0.6897, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.5908008813858032, | |
| "learning_rate": 9.834435247725032e-06, | |
| "loss": 0.6598, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.7284044027328491, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6383, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5472, | |
| "grad_norm": 0.7161381840705872, | |
| "learning_rate": 9.804586609725499e-06, | |
| "loss": 0.6667, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5568, | |
| "grad_norm": 0.6291981339454651, | |
| "learning_rate": 9.788754083424654e-06, | |
| "loss": 0.6504, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5664, | |
| "grad_norm": 0.5240605473518372, | |
| "learning_rate": 9.772318717677905e-06, | |
| "loss": 0.6343, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.586787760257721, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.6418, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5856, | |
| "grad_norm": 0.5460494160652161, | |
| "learning_rate": 9.737647819437645e-06, | |
| "loss": 0.644, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5952, | |
| "grad_norm": 0.6215012073516846, | |
| "learning_rate": 9.719416651541839e-06, | |
| "loss": 0.6532, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6048, | |
| "grad_norm": 0.5574711561203003, | |
| "learning_rate": 9.700591372846096e-06, | |
| "loss": 0.6262, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.5710508823394775, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6576, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.5259015560150146, | |
| "learning_rate": 9.661168036940071e-06, | |
| "loss": 0.6518, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.6336, | |
| "grad_norm": 0.6580711007118225, | |
| "learning_rate": 9.640574942595195e-06, | |
| "loss": 0.6315, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6432, | |
| "grad_norm": 0.5900007486343384, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.6615, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.636417031288147, | |
| "learning_rate": 9.597638862757255e-06, | |
| "loss": 0.6511, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6624, | |
| "grad_norm": 0.6361602544784546, | |
| "learning_rate": 9.5753012823366e-06, | |
| "loss": 0.6523, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.6292795538902283, | |
| "learning_rate": 9.552387733294081e-06, | |
| "loss": 0.6317, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6816, | |
| "grad_norm": 0.6038651466369629, | |
| "learning_rate": 9.528901100135971e-06, | |
| "loss": 0.6424, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.5914206504821777, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6868, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7008, | |
| "grad_norm": 0.6082513928413391, | |
| "learning_rate": 9.480220479843627e-06, | |
| "loss": 0.6638, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7104, | |
| "grad_norm": 0.5032195448875427, | |
| "learning_rate": 9.45503262094184e-06, | |
| "loss": 0.6381, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.5890430212020874, | |
| "learning_rate": 9.4292839336179e-06, | |
| "loss": 0.627, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.6046479344367981, | |
| "learning_rate": 9.40297765928369e-06, | |
| "loss": 0.6597, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.7392, | |
| "grad_norm": 0.557981014251709, | |
| "learning_rate": 9.376117109543769e-06, | |
| "loss": 0.6607, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7488, | |
| "grad_norm": 0.6756407022476196, | |
| "learning_rate": 9.348705665778479e-06, | |
| "loss": 0.6459, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7584, | |
| "grad_norm": 0.6524885296821594, | |
| "learning_rate": 9.320746778718274e-06, | |
| "loss": 0.6418, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.5784417986869812, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6269, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7776, | |
| "grad_norm": 0.6418123245239258, | |
| "learning_rate": 9.263200821770462e-06, | |
| "loss": 0.63, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7872, | |
| "grad_norm": 0.6497018337249756, | |
| "learning_rate": 9.233620996141421e-06, | |
| "loss": 0.6605, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7968, | |
| "grad_norm": 0.6296787261962891, | |
| "learning_rate": 9.203508214822652e-06, | |
| "loss": 0.6678, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.6279412508010864, | |
| "learning_rate": 9.172866268606514e-06, | |
| "loss": 0.663, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.5324777960777283, | |
| "learning_rate": 9.141699014900084e-06, | |
| "loss": 0.6427, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.8256, | |
| "grad_norm": 0.6004756093025208, | |
| "learning_rate": 9.110010377239552e-06, | |
| "loss": 0.6407, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8352, | |
| "grad_norm": 0.6031383872032166, | |
| "learning_rate": 9.077804344796302e-06, | |
| "loss": 0.634, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.5144718885421753, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6503, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8544, | |
| "grad_norm": 0.6444131731987, | |
| "learning_rate": 9.011856377401891e-06, | |
| "loss": 0.6418, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.6462003588676453, | |
| "learning_rate": 8.978122744408905e-06, | |
| "loss": 0.6438, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8736, | |
| "grad_norm": 0.5521820783615112, | |
| "learning_rate": 8.943888319504456e-06, | |
| "loss": 0.6566, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.5820996761322021, | |
| "learning_rate": 8.90915741234015e-06, | |
| "loss": 0.6361, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8928, | |
| "grad_norm": 0.7227920889854431, | |
| "learning_rate": 8.873934395068006e-06, | |
| "loss": 0.6652, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.9024, | |
| "grad_norm": 0.5779785513877869, | |
| "learning_rate": 8.838223701790057e-06, | |
| "loss": 0.6216, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.5625780820846558, | |
| "learning_rate": 8.802029828000157e-06, | |
| "loss": 0.6123, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.6641936302185059, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.6392, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9312, | |
| "grad_norm": 0.514313280582428, | |
| "learning_rate": 8.728210824415829e-06, | |
| "loss": 0.6051, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.9408, | |
| "grad_norm": 0.6550350189208984, | |
| "learning_rate": 8.690594987436705e-06, | |
| "loss": 0.6084, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9504, | |
| "grad_norm": 0.545590341091156, | |
| "learning_rate": 8.652514554406388e-06, | |
| "loss": 0.6335, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5167790055274963, | |
| "learning_rate": 8.613974319136959e-06, | |
| "loss": 0.6244, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9696, | |
| "grad_norm": 0.678073525428772, | |
| "learning_rate": 8.574979133323378e-06, | |
| "loss": 0.6363, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9792, | |
| "grad_norm": 0.538718044757843, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.6407, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9888, | |
| "grad_norm": 0.6520209908485413, | |
| "learning_rate": 8.495643602586287e-06, | |
| "loss": 0.6485, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.6909246444702148, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.6727, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 1.312455177307129, | |
| "learning_rate": 8.414547910024035e-06, | |
| "loss": 1.0788, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.0176, | |
| "grad_norm": 0.706491231918335, | |
| "learning_rate": 8.373352729660373e-06, | |
| "loss": 0.5632, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0272, | |
| "grad_norm": 0.6403841376304626, | |
| "learning_rate": 8.331732889760021e-06, | |
| "loss": 0.5904, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.5960009098052979, | |
| "learning_rate": 8.289693629698564e-06, | |
| "loss": 0.5909, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.0464, | |
| "grad_norm": 0.7191261053085327, | |
| "learning_rate": 8.247240241650918e-06, | |
| "loss": 0.6101, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 0.5526676177978516, | |
| "learning_rate": 8.204378069925121e-06, | |
| "loss": 0.6091, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0656, | |
| "grad_norm": 0.6340149641036987, | |
| "learning_rate": 8.16111251028955e-06, | |
| "loss": 0.5899, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.5784077644348145, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5446, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.0848, | |
| "grad_norm": 0.6272123456001282, | |
| "learning_rate": 8.073393063582386e-06, | |
| "loss": 0.5883, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0944, | |
| "grad_norm": 0.5847621560096741, | |
| "learning_rate": 8.0289502192041e-06, | |
| "loss": 0.5656, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 0.6778438091278076, | |
| "learning_rate": 7.984126070912519e-06, | |
| "loss": 0.6299, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.6147951483726501, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.5624, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1232, | |
| "grad_norm": 0.7127768397331238, | |
| "learning_rate": 7.89335648089903e-06, | |
| "loss": 0.6331, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.1328, | |
| "grad_norm": 0.5153930187225342, | |
| "learning_rate": 7.84742246584226e-06, | |
| "loss": 0.5773, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1424, | |
| "grad_norm": 0.5773877501487732, | |
| "learning_rate": 7.801129998764014e-06, | |
| "loss": 0.5535, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.6432895660400391, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.6113, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.1616, | |
| "grad_norm": 0.546201765537262, | |
| "learning_rate": 7.70749306331863e-06, | |
| "loss": 0.5759, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.1712, | |
| "grad_norm": 0.561962366104126, | |
| "learning_rate": 7.660160382576683e-06, | |
| "loss": 0.6, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.1808, | |
| "grad_norm": 0.5166497230529785, | |
| "learning_rate": 7.612492823579744e-06, | |
| "loss": 0.6019, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.6238263845443726, | |
| "learning_rate": 7.564496387029532e-06, | |
| "loss": 0.6021, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.5210447311401367, | |
| "learning_rate": 7.516177115029002e-06, | |
| "loss": 0.5664, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.2096, | |
| "grad_norm": 0.5317444801330566, | |
| "learning_rate": 7.467541090321735e-06, | |
| "loss": 0.5799, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2192, | |
| "grad_norm": 0.5516515970230103, | |
| "learning_rate": 7.4185944355261996e-06, | |
| "loss": 0.5393, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.5734798312187195, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.6331, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.2384, | |
| "grad_norm": 0.5610190629959106, | |
| "learning_rate": 7.319793920889171e-06, | |
| "loss": 0.5094, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 0.5572755336761475, | |
| "learning_rate": 7.269952498697734e-06, | |
| "loss": 0.5987, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2576, | |
| "grad_norm": 0.5415238738059998, | |
| "learning_rate": 7.219825320152411e-06, | |
| "loss": 0.6032, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.48308759927749634, | |
| "learning_rate": 7.169418695587791e-06, | |
| "loss": 0.488, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.2768, | |
| "grad_norm": 0.5875827074050903, | |
| "learning_rate": 7.118738970516944e-06, | |
| "loss": 0.6191, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2864, | |
| "grad_norm": 0.4882888197898865, | |
| "learning_rate": 7.067792524832604e-06, | |
| "loss": 0.5268, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 0.6158053874969482, | |
| "learning_rate": 7.016585772004026e-06, | |
| "loss": 0.6349, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.47751063108444214, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.6108, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3152, | |
| "grad_norm": 0.5912806391716003, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.579, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.3248, | |
| "grad_norm": 0.5066422820091248, | |
| "learning_rate": 6.8614682920097265e-06, | |
| "loss": 0.5309, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3344, | |
| "grad_norm": 0.6349862813949585, | |
| "learning_rate": 6.809285088483361e-06, | |
| "loss": 0.6401, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.4846351444721222, | |
| "learning_rate": 6.7568741204067145e-06, | |
| "loss": 0.5554, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3536000000000001, | |
| "grad_norm": 0.5736483335494995, | |
| "learning_rate": 6.704241985612625e-06, | |
| "loss": 0.5849, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.3632, | |
| "grad_norm": 0.4874175786972046, | |
| "learning_rate": 6.651395309775837e-06, | |
| "loss": 0.5626, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.3728, | |
| "grad_norm": 0.5509251356124878, | |
| "learning_rate": 6.598340745578908e-06, | |
| "loss": 0.6218, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.570919394493103, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.5982, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 0.5047493577003479, | |
| "learning_rate": 6.491634692845781e-06, | |
| "loss": 0.5755, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.4016, | |
| "grad_norm": 0.4495947062969208, | |
| "learning_rate": 6.437996637160086e-06, | |
| "loss": 0.5432, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4112, | |
| "grad_norm": 0.5126574039459229, | |
| "learning_rate": 6.384177557124247e-06, | |
| "loss": 0.6209, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.504632294178009, | |
| "learning_rate": 6.330184227833376e-06, | |
| "loss": 0.553, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.4304000000000001, | |
| "grad_norm": 0.5119583606719971, | |
| "learning_rate": 6.276023446318214e-06, | |
| "loss": 0.5674, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.5273292660713196, | |
| "learning_rate": 6.2217020306894705e-06, | |
| "loss": 0.5947, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4496, | |
| "grad_norm": 0.42243409156799316, | |
| "learning_rate": 6.1672268192795285e-06, | |
| "loss": 0.5382, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.43099960684776306, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5702, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.4687999999999999, | |
| "grad_norm": 0.5123092532157898, | |
| "learning_rate": 6.057842458386315e-06, | |
| "loss": 0.6564, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.4784, | |
| "grad_norm": 0.470164030790329, | |
| "learning_rate": 6.002947078916365e-06, | |
| "loss": 0.5484, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 0.465804785490036, | |
| "learning_rate": 5.947925441958393e-06, | |
| "loss": 0.5918, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.45440298318862915, | |
| "learning_rate": 5.892784473993184e-06, | |
| "loss": 0.5526, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5072, | |
| "grad_norm": 0.522992730140686, | |
| "learning_rate": 5.837531116523683e-06, | |
| "loss": 0.6385, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.5168, | |
| "grad_norm": 0.4795243740081787, | |
| "learning_rate": 5.782172325201155e-06, | |
| "loss": 0.5744, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5264, | |
| "grad_norm": 0.4822704792022705, | |
| "learning_rate": 5.726715068949564e-06, | |
| "loss": 0.5168, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.5573145151138306, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.6515, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5455999999999999, | |
| "grad_norm": 0.5187973976135254, | |
| "learning_rate": 5.615533098453215e-06, | |
| "loss": 0.5704, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.5552000000000001, | |
| "grad_norm": 0.47124606370925903, | |
| "learning_rate": 5.559822380516539e-06, | |
| "loss": 0.5753, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.5648, | |
| "grad_norm": 0.5377249717712402, | |
| "learning_rate": 5.504041188505022e-06, | |
| "loss": 0.5911, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.4670109152793884, | |
| "learning_rate": 5.448196544517168e-06, | |
| "loss": 0.5771, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 0.5638590455055237, | |
| "learning_rate": 5.392295478639226e-06, | |
| "loss": 0.5714, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.5936, | |
| "grad_norm": 0.48816683888435364, | |
| "learning_rate": 5.336345028060199e-06, | |
| "loss": 0.5873, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.6032, | |
| "grad_norm": 0.5010278820991516, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "loss": 0.5667, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.5451114177703857, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5715, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6223999999999998, | |
| "grad_norm": 0.45967310667037964, | |
| "learning_rate": 5.168267827938971e-06, | |
| "loss": 0.5353, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.5528745651245117, | |
| "learning_rate": 5.112190321479026e-06, | |
| "loss": 0.582, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6416, | |
| "grad_norm": 0.5903663039207458, | |
| "learning_rate": 5.05609869177323e-06, | |
| "loss": 0.6188, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.5361656546592712, | |
| "learning_rate": 5e-06, | |
| "loss": 0.4733, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.6608, | |
| "grad_norm": 0.6212491393089294, | |
| "learning_rate": 4.943901308226771e-06, | |
| "loss": 0.6487, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.6703999999999999, | |
| "grad_norm": 0.48965513706207275, | |
| "learning_rate": 4.887809678520976e-06, | |
| "loss": 0.5334, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.5252687931060791, | |
| "learning_rate": 4.831732172061032e-06, | |
| "loss": 0.5959, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.6434024572372437, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.6187, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.6992, | |
| "grad_norm": 0.5325840711593628, | |
| "learning_rate": 4.719647763814041e-06, | |
| "loss": 0.5754, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.7088, | |
| "grad_norm": 0.4677431881427765, | |
| "learning_rate": 4.663654971939802e-06, | |
| "loss": 0.6058, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7184, | |
| "grad_norm": 0.4714028537273407, | |
| "learning_rate": 4.6077045213607765e-06, | |
| "loss": 0.5671, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.5045012831687927, | |
| "learning_rate": 4.551803455482833e-06, | |
| "loss": 0.5921, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7376, | |
| "grad_norm": 0.5761282444000244, | |
| "learning_rate": 4.4959588114949785e-06, | |
| "loss": 0.6229, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.7471999999999999, | |
| "grad_norm": 0.5460184812545776, | |
| "learning_rate": 4.4401776194834615e-06, | |
| "loss": 0.5371, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.7568000000000001, | |
| "grad_norm": 0.4542312026023865, | |
| "learning_rate": 4.384466901546786e-06, | |
| "loss": 0.6257, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5277249813079834, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.558, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 0.44270434975624084, | |
| "learning_rate": 4.273284931050438e-06, | |
| "loss": 0.5644, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.7856, | |
| "grad_norm": 0.4488348066806793, | |
| "learning_rate": 4.217827674798845e-06, | |
| "loss": 0.583, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.7952, | |
| "grad_norm": 0.4872271716594696, | |
| "learning_rate": 4.162468883476319e-06, | |
| "loss": 0.6316, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.4470103085041046, | |
| "learning_rate": 4.107215526006818e-06, | |
| "loss": 0.5279, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8144, | |
| "grad_norm": 0.477935791015625, | |
| "learning_rate": 4.052074558041608e-06, | |
| "loss": 0.5368, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.46086037158966064, | |
| "learning_rate": 3.997052921083637e-06, | |
| "loss": 0.6028, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8336000000000001, | |
| "grad_norm": 0.4894026219844818, | |
| "learning_rate": 3.9421575416136866e-06, | |
| "loss": 0.5955, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.4144808351993561, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.5063, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.8528, | |
| "grad_norm": 0.4998006224632263, | |
| "learning_rate": 3.832773180720475e-06, | |
| "loss": 0.6493, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.8624, | |
| "grad_norm": 0.46334385871887207, | |
| "learning_rate": 3.778297969310529e-06, | |
| "loss": 0.5779, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 0.44945594668388367, | |
| "learning_rate": 3.723976553681787e-06, | |
| "loss": 0.5444, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.48254647850990295, | |
| "learning_rate": 3.669815772166625e-06, | |
| "loss": 0.5834, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.8912, | |
| "grad_norm": 0.4141770303249359, | |
| "learning_rate": 3.6158224428757538e-06, | |
| "loss": 0.5167, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.9008, | |
| "grad_norm": 0.4774608612060547, | |
| "learning_rate": 3.562003362839914e-06, | |
| "loss": 0.576, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9104, | |
| "grad_norm": 0.46828344464302063, | |
| "learning_rate": 3.50836530715422e-06, | |
| "loss": 0.5696, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.5107264518737793, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.6113, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9296, | |
| "grad_norm": 0.42679721117019653, | |
| "learning_rate": 3.4016592544210937e-06, | |
| "loss": 0.5417, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.9392, | |
| "grad_norm": 0.47424426674842834, | |
| "learning_rate": 3.3486046902241663e-06, | |
| "loss": 0.5776, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.9487999999999999, | |
| "grad_norm": 0.44007402658462524, | |
| "learning_rate": 3.295758014387375e-06, | |
| "loss": 0.5334, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.519551157951355, | |
| "learning_rate": 3.2431258795932863e-06, | |
| "loss": 0.5946, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 0.4806658923625946, | |
| "learning_rate": 3.1907149115166403e-06, | |
| "loss": 0.5666, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.9776, | |
| "grad_norm": 0.47380775213241577, | |
| "learning_rate": 3.1385317079902743e-06, | |
| "loss": 0.597, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.9872, | |
| "grad_norm": 0.48056966066360474, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.538, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.5239747762680054, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.5785, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0064, | |
| "grad_norm": 1.2484456300735474, | |
| "learning_rate": 2.9834142279959754e-06, | |
| "loss": 1.0245, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.4434763789176941, | |
| "learning_rate": 2.932207475167398e-06, | |
| "loss": 0.4894, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.0256, | |
| "grad_norm": 0.5152484774589539, | |
| "learning_rate": 2.8812610294830568e-06, | |
| "loss": 0.5657, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 0.46128612756729126, | |
| "learning_rate": 2.83058130441221e-06, | |
| "loss": 0.4564, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.0448, | |
| "grad_norm": 0.4728868901729584, | |
| "learning_rate": 2.7801746798475905e-06, | |
| "loss": 0.5775, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.0544, | |
| "grad_norm": 0.4691709280014038, | |
| "learning_rate": 2.7300475013022666e-06, | |
| "loss": 0.5677, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.4791906177997589, | |
| "learning_rate": 2.6802060791108304e-06, | |
| "loss": 0.5207, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.450295627117157, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.4673, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.0832, | |
| "grad_norm": 0.48020562529563904, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "loss": 0.5457, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.0928, | |
| "grad_norm": 0.3935261368751526, | |
| "learning_rate": 2.532458909678266e-06, | |
| "loss": 0.5164, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.1024, | |
| "grad_norm": 0.464346706867218, | |
| "learning_rate": 2.483822884971e-06, | |
| "loss": 0.557, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.5267810821533203, | |
| "learning_rate": 2.43550361297047e-06, | |
| "loss": 0.4987, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1216, | |
| "grad_norm": 0.49507227540016174, | |
| "learning_rate": 2.387507176420256e-06, | |
| "loss": 0.5607, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.1312, | |
| "grad_norm": 0.45029518008232117, | |
| "learning_rate": 2.339839617423318e-06, | |
| "loss": 0.5208, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.1408, | |
| "grad_norm": 0.44882386922836304, | |
| "learning_rate": 2.2925069366813718e-06, | |
| "loss": 0.5426, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.47287508845329285, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.5239, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.4874938726425171, | |
| "learning_rate": 2.1988700012359865e-06, | |
| "loss": 0.5492, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.1696, | |
| "grad_norm": 0.4065110683441162, | |
| "learning_rate": 2.1525775341577404e-06, | |
| "loss": 0.4777, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.1792, | |
| "grad_norm": 0.4227883219718933, | |
| "learning_rate": 2.1066435191009717e-06, | |
| "loss": 0.6065, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.39913421869277954, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.4807, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.1984, | |
| "grad_norm": 0.4827454686164856, | |
| "learning_rate": 2.0158739290874822e-06, | |
| "loss": 0.5807, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.43721911311149597, | |
| "learning_rate": 1.971049780795901e-06, | |
| "loss": 0.5519, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2176, | |
| "grad_norm": 0.3625260293483734, | |
| "learning_rate": 1.9266069364176144e-06, | |
| "loss": 0.5069, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.3953098952770233, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.5327, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.2368, | |
| "grad_norm": 0.38376370072364807, | |
| "learning_rate": 1.838887489710452e-06, | |
| "loss": 0.5435, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.2464, | |
| "grad_norm": 0.3926437199115753, | |
| "learning_rate": 1.7956219300748796e-06, | |
| "loss": 0.4793, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.44054508209228516, | |
| "learning_rate": 1.7527597583490825e-06, | |
| "loss": 0.5631, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.41770970821380615, | |
| "learning_rate": 1.7103063703014372e-06, | |
| "loss": 0.563, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.2752, | |
| "grad_norm": 0.4216345250606537, | |
| "learning_rate": 1.6682671102399806e-06, | |
| "loss": 0.5599, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.2848, | |
| "grad_norm": 0.4026756286621094, | |
| "learning_rate": 1.6266472703396286e-06, | |
| "loss": 0.5192, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.2944, | |
| "grad_norm": 0.382327139377594, | |
| "learning_rate": 1.5854520899759656e-06, | |
| "loss": 0.5279, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.39819908142089844, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.4906, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.3136, | |
| "grad_norm": 0.42366543412208557, | |
| "learning_rate": 1.5043563974137132e-06, | |
| "loss": 0.5485, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.3232, | |
| "grad_norm": 0.38778194785118103, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.5269, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.3327999999999998, | |
| "grad_norm": 0.38411736488342285, | |
| "learning_rate": 1.4250208666766235e-06, | |
| "loss": 0.5483, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.40252214670181274, | |
| "learning_rate": 1.3860256808630429e-06, | |
| "loss": 0.5018, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.4110822379589081, | |
| "learning_rate": 1.3474854455936126e-06, | |
| "loss": 0.5486, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.3616, | |
| "grad_norm": 0.4108063876628876, | |
| "learning_rate": 1.3094050125632973e-06, | |
| "loss": 0.5573, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.3712, | |
| "grad_norm": 0.3773021101951599, | |
| "learning_rate": 1.2717891755841722e-06, | |
| "loss": 0.506, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.41075634956359863, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.5406, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.3904, | |
| "grad_norm": 0.3963279128074646, | |
| "learning_rate": 1.1979701719998454e-06, | |
| "loss": 0.545, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.3936994671821594, | |
| "learning_rate": 1.1617762982099446e-06, | |
| "loss": 0.5294, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.4096, | |
| "grad_norm": 0.37653613090515137, | |
| "learning_rate": 1.1260656049319957e-06, | |
| "loss": 0.5246, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.4054081439971924, | |
| "learning_rate": 1.0908425876598512e-06, | |
| "loss": 0.5649, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.4288, | |
| "grad_norm": 0.37866562604904175, | |
| "learning_rate": 1.0561116804955451e-06, | |
| "loss": 0.537, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.4384, | |
| "grad_norm": 0.3885415494441986, | |
| "learning_rate": 1.0218772555910955e-06, | |
| "loss": 0.5036, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.418641597032547, | |
| "learning_rate": 9.881436225981107e-07, | |
| "loss": 0.5812, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.36240801215171814, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.4417, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.4672, | |
| "grad_norm": 0.4024696946144104, | |
| "learning_rate": 9.221956552036992e-07, | |
| "loss": 0.5798, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.4768, | |
| "grad_norm": 0.38764163851737976, | |
| "learning_rate": 8.899896227604509e-07, | |
| "loss": 0.502, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.4864, | |
| "grad_norm": 0.4298345744609833, | |
| "learning_rate": 8.58300985099918e-07, | |
| "loss": 0.5429, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.39933955669403076, | |
| "learning_rate": 8.271337313934869e-07, | |
| "loss": 0.5403, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.5056000000000003, | |
| "grad_norm": 0.3838823437690735, | |
| "learning_rate": 7.964917851773496e-07, | |
| "loss": 0.5212, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.5152, | |
| "grad_norm": 0.41426998376846313, | |
| "learning_rate": 7.663790038585794e-07, | |
| "loss": 0.5227, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.5248, | |
| "grad_norm": 0.40095555782318115, | |
| "learning_rate": 7.367991782295392e-07, | |
| "loss": 0.4788, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.41771426796913147, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.5376, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.376322865486145, | |
| "learning_rate": 6.792532212817271e-07, | |
| "loss": 0.4766, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.5536, | |
| "grad_norm": 0.4470244348049164, | |
| "learning_rate": 6.512943342215234e-07, | |
| "loss": 0.57, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.5632, | |
| "grad_norm": 0.36726924777030945, | |
| "learning_rate": 6.238828904562316e-07, | |
| "loss": 0.4631, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.40520891547203064, | |
| "learning_rate": 5.9702234071631e-07, | |
| "loss": 0.549, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.5824, | |
| "grad_norm": 0.3793184459209442, | |
| "learning_rate": 5.707160663821009e-07, | |
| "loss": 0.5225, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.4228286147117615, | |
| "learning_rate": 5.449673790581611e-07, | |
| "loss": 0.5731, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.6016, | |
| "grad_norm": 0.32514867186546326, | |
| "learning_rate": 5.197795201563744e-07, | |
| "loss": 0.4647, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.4135151505470276, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.6043, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.6208, | |
| "grad_norm": 0.37106066942214966, | |
| "learning_rate": 4.710988998640298e-07, | |
| "loss": 0.5218, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.6304, | |
| "grad_norm": 0.3782541751861572, | |
| "learning_rate": 4.4761226670592074e-07, | |
| "loss": 0.5365, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.40136420726776123, | |
| "learning_rate": 4.2469871766340096e-07, | |
| "loss": 0.5295, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.3829237222671509, | |
| "learning_rate": 4.0236113724274716e-07, | |
| "loss": 0.544, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.6592000000000002, | |
| "grad_norm": 0.4050566554069519, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.5751, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.6688, | |
| "grad_norm": 0.3831566274166107, | |
| "learning_rate": 3.5942505740480583e-07, | |
| "loss": 0.483, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.6784, | |
| "grad_norm": 0.40931302309036255, | |
| "learning_rate": 3.3883196305992906e-07, | |
| "loss": 0.5429, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.4193519651889801, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.5176, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.6976, | |
| "grad_norm": 0.38546380400657654, | |
| "learning_rate": 2.9940862715390483e-07, | |
| "loss": 0.5182, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.7072000000000003, | |
| "grad_norm": 0.3639739155769348, | |
| "learning_rate": 2.8058334845816214e-07, | |
| "loss": 0.4759, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.7168, | |
| "grad_norm": 0.4042200744152069, | |
| "learning_rate": 2.6235218056235633e-07, | |
| "loss": 0.5925, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.4064270555973053, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.5799, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.37163835763931274, | |
| "learning_rate": 2.276812823220964e-07, | |
| "loss": 0.4825, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.7456, | |
| "grad_norm": 0.39752858877182007, | |
| "learning_rate": 2.1124591657534776e-07, | |
| "loss": 0.5338, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.7552, | |
| "grad_norm": 0.36681297421455383, | |
| "learning_rate": 1.9541339027450256e-07, | |
| "loss": 0.5208, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.38168057799339294, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.5941, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.7744, | |
| "grad_norm": 0.3118930757045746, | |
| "learning_rate": 1.6556475227496816e-07, | |
| "loss": 0.4631, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.3986196219921112, | |
| "learning_rate": 1.5155239811656562e-07, | |
| "loss": 0.5773, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7936, | |
| "grad_norm": 0.3834291696548462, | |
| "learning_rate": 1.3815039801161723e-07, | |
| "loss": 0.553, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.38300490379333496, | |
| "learning_rate": 1.253604390908819e-07, | |
| "loss": 0.5055, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8128, | |
| "grad_norm": 0.40329509973526, | |
| "learning_rate": 1.1318413143740436e-07, | |
| "loss": 0.5562, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.8224, | |
| "grad_norm": 0.37382957339286804, | |
| "learning_rate": 1.0162300788382263e-07, | |
| "loss": 0.5532, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.395346462726593, | |
| "learning_rate": 9.0678523819408e-08, | |
| "loss": 0.4918, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3847428858280182, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.5126, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.8512, | |
| "grad_norm": 0.39311036467552185, | |
| "learning_rate": 7.064490740882057e-08, | |
| "loss": 0.5539, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.8608000000000002, | |
| "grad_norm": 0.3707370162010193, | |
| "learning_rate": 6.15582970243117e-08, | |
| "loss": 0.5242, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.8704, | |
| "grad_norm": 0.39528071880340576, | |
| "learning_rate": 5.3093369734816824e-08, | |
| "loss": 0.5205, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.37834280729293823, | |
| "learning_rate": 4.52511911603265e-08, | |
| "loss": 0.5625, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.8895999999999997, | |
| "grad_norm": 0.3650013208389282, | |
| "learning_rate": 3.8032748525179684e-08, | |
| "loss": 0.4993, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.8992, | |
| "grad_norm": 0.40211090445518494, | |
| "learning_rate": 3.143895053378698e-08, | |
| "loss": 0.5518, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9088000000000003, | |
| "grad_norm": 0.4369528293609619, | |
| "learning_rate": 2.547062725623828e-08, | |
| "loss": 0.6037, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.34852176904678345, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.4652, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.3797735273838043, | |
| "learning_rate": 1.541333133436018e-08, | |
| "loss": 0.5442, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.9375999999999998, | |
| "grad_norm": 0.3867435157299042, | |
| "learning_rate": 1.132562476771959e-08, | |
| "loss": 0.5293, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.9472, | |
| "grad_norm": 0.3923242688179016, | |
| "learning_rate": 7.865924910916977e-09, | |
| "loss": 0.554, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.3982171416282654, | |
| "learning_rate": 5.034667293427053e-09, | |
| "loss": 0.5274, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.9664, | |
| "grad_norm": 0.3748815953731537, | |
| "learning_rate": 2.8322083323334417e-09, | |
| "loss": 0.5124, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.37689897418022156, | |
| "learning_rate": 1.2588252874673469e-09, | |
| "loss": 0.5876, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.9856, | |
| "grad_norm": 0.3875444531440735, | |
| "learning_rate": 3.147162264971471e-10, | |
| "loss": 0.5288, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.37963417172431946, | |
| "learning_rate": 0.0, | |
| "loss": 0.5648, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "step": 312, | |
| "total_flos": 312484367761408.0, | |
| "train_loss": 0.6120152068443787, | |
| "train_runtime": 29530.8856, | |
| "train_samples_per_second": 1.016, | |
| "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": 312484367761408.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |