| { | |
| "best_metric": 1.503947377204895, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 3.009009009009009, | |
| "eval_steps": 50, | |
| "global_step": 167, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.018018018018018018, | |
| "grad_norm": 0.20334728062152863, | |
| "learning_rate": 1e-05, | |
| "loss": 1.7238, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.018018018018018018, | |
| "eval_loss": 2.4445245265960693, | |
| "eval_runtime": 1.9335, | |
| "eval_samples_per_second": 48.617, | |
| "eval_steps_per_second": 12.413, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.036036036036036036, | |
| "grad_norm": 0.2328772246837616, | |
| "learning_rate": 2e-05, | |
| "loss": 2.3157, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05405405405405406, | |
| "grad_norm": 0.17818941175937653, | |
| "learning_rate": 3e-05, | |
| "loss": 1.8891, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07207207207207207, | |
| "grad_norm": 0.18887093663215637, | |
| "learning_rate": 4e-05, | |
| "loss": 1.8112, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.09009009009009009, | |
| "grad_norm": 0.28661012649536133, | |
| "learning_rate": 5e-05, | |
| "loss": 2.5251, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10810810810810811, | |
| "grad_norm": 0.5507867932319641, | |
| "learning_rate": 6e-05, | |
| "loss": 2.5171, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.12612612612612611, | |
| "grad_norm": 0.5678154230117798, | |
| "learning_rate": 7e-05, | |
| "loss": 2.6257, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.14414414414414414, | |
| "grad_norm": 0.507682204246521, | |
| "learning_rate": 8e-05, | |
| "loss": 2.5335, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.16216216216216217, | |
| "grad_norm": 0.3488699495792389, | |
| "learning_rate": 9e-05, | |
| "loss": 2.6651, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.18018018018018017, | |
| "grad_norm": 0.2555129826068878, | |
| "learning_rate": 0.0001, | |
| "loss": 1.9695, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1981981981981982, | |
| "grad_norm": 0.2776542603969574, | |
| "learning_rate": 9.998999018714263e-05, | |
| "loss": 1.7186, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.21621621621621623, | |
| "grad_norm": 0.3397429883480072, | |
| "learning_rate": 9.995996475642466e-05, | |
| "loss": 2.1329, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.23423423423423423, | |
| "grad_norm": 0.5618317127227783, | |
| "learning_rate": 9.990993572980378e-05, | |
| "loss": 3.0633, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.25225225225225223, | |
| "grad_norm": 0.1377403289079666, | |
| "learning_rate": 9.983992313852774e-05, | |
| "loss": 1.7843, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.2702702702702703, | |
| "grad_norm": 0.15059250593185425, | |
| "learning_rate": 9.974995501511404e-05, | |
| "loss": 2.2153, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2882882882882883, | |
| "grad_norm": 0.1395379900932312, | |
| "learning_rate": 9.964006738212575e-05, | |
| "loss": 1.8151, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3063063063063063, | |
| "grad_norm": 0.17120715975761414, | |
| "learning_rate": 9.951030423774859e-05, | |
| "loss": 2.0037, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.32432432432432434, | |
| "grad_norm": 0.45357421040534973, | |
| "learning_rate": 9.936071753817415e-05, | |
| "loss": 2.1388, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.34234234234234234, | |
| "grad_norm": 0.4322102665901184, | |
| "learning_rate": 9.919136717679722e-05, | |
| "loss": 2.2311, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.36036036036036034, | |
| "grad_norm": 0.4717254638671875, | |
| "learning_rate": 9.900232096023477e-05, | |
| "loss": 2.157, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3783783783783784, | |
| "grad_norm": 0.3566138744354248, | |
| "learning_rate": 9.879365458117678e-05, | |
| "loss": 2.1204, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3963963963963964, | |
| "grad_norm": 0.26241886615753174, | |
| "learning_rate": 9.856545158807938e-05, | |
| "loss": 2.4735, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.4144144144144144, | |
| "grad_norm": 0.20926542580127716, | |
| "learning_rate": 9.831780335171279e-05, | |
| "loss": 1.7209, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.43243243243243246, | |
| "grad_norm": 0.31854745745658875, | |
| "learning_rate": 9.805080902857699e-05, | |
| "loss": 2.2578, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.45045045045045046, | |
| "grad_norm": 0.3495607376098633, | |
| "learning_rate": 9.776457552120033e-05, | |
| "loss": 2.2969, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.46846846846846846, | |
| "grad_norm": 0.6214109659194946, | |
| "learning_rate": 9.745921743533653e-05, | |
| "loss": 2.1911, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.4864864864864865, | |
| "grad_norm": 0.14884573221206665, | |
| "learning_rate": 9.713485703407731e-05, | |
| "loss": 1.6914, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.5045045045045045, | |
| "grad_norm": 0.2149585634469986, | |
| "learning_rate": 9.679162418889931e-05, | |
| "loss": 2.1283, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5225225225225225, | |
| "grad_norm": 0.20466195046901703, | |
| "learning_rate": 9.642965632766436e-05, | |
| "loss": 1.8882, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5405405405405406, | |
| "grad_norm": 0.22452042996883392, | |
| "learning_rate": 9.604909837959455e-05, | |
| "loss": 2.1373, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5585585585585585, | |
| "grad_norm": 0.29129326343536377, | |
| "learning_rate": 9.565010271724352e-05, | |
| "loss": 1.6579, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5765765765765766, | |
| "grad_norm": 0.4822368621826172, | |
| "learning_rate": 9.523282909548773e-05, | |
| "loss": 1.7155, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5945945945945946, | |
| "grad_norm": 0.5174642205238342, | |
| "learning_rate": 9.47974445875617e-05, | |
| "loss": 1.6795, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.6126126126126126, | |
| "grad_norm": 0.5645754933357239, | |
| "learning_rate": 9.434412351816328e-05, | |
| "loss": 1.5836, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6306306306306306, | |
| "grad_norm": 0.2343388795852661, | |
| "learning_rate": 9.387304739365523e-05, | |
| "loss": 2.0405, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6486486486486487, | |
| "grad_norm": 0.2349058836698532, | |
| "learning_rate": 9.338440482939146e-05, | |
| "loss": 1.9695, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.23872889578342438, | |
| "learning_rate": 9.287839147419686e-05, | |
| "loss": 1.8085, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6846846846846847, | |
| "grad_norm": 0.37950703501701355, | |
| "learning_rate": 9.2355209932031e-05, | |
| "loss": 2.3962, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.7027027027027027, | |
| "grad_norm": 0.5464185476303101, | |
| "learning_rate": 9.181506968086697e-05, | |
| "loss": 2.802, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7207207207207207, | |
| "grad_norm": 0.13629977405071259, | |
| "learning_rate": 9.125818698881798e-05, | |
| "loss": 1.6469, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7387387387387387, | |
| "grad_norm": 0.2170867770910263, | |
| "learning_rate": 9.068478482754532e-05, | |
| "loss": 2.544, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7567567567567568, | |
| "grad_norm": 0.19846679270267487, | |
| "learning_rate": 9.0095092782982e-05, | |
| "loss": 1.7291, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7747747747747747, | |
| "grad_norm": 0.19683820009231567, | |
| "learning_rate": 8.948934696340843e-05, | |
| "loss": 1.8553, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7927927927927928, | |
| "grad_norm": 2.0814313888549805, | |
| "learning_rate": 8.886778990491631e-05, | |
| "loss": 1.4285, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.8108108108108109, | |
| "grad_norm": 1.3856438398361206, | |
| "learning_rate": 8.823067047429907e-05, | |
| "loss": 1.3423, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8288288288288288, | |
| "grad_norm": 0.75543612241745, | |
| "learning_rate": 8.757824376940746e-05, | |
| "loss": 1.167, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8468468468468469, | |
| "grad_norm": 0.53647381067276, | |
| "learning_rate": 8.691077101701024e-05, | |
| "loss": 1.3325, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8648648648648649, | |
| "grad_norm": 0.23286841809749603, | |
| "learning_rate": 8.622851946820095e-05, | |
| "loss": 2.2234, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8828828828828829, | |
| "grad_norm": 0.20140407979488373, | |
| "learning_rate": 8.553176229139261e-05, | |
| "loss": 1.7397, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.9009009009009009, | |
| "grad_norm": 0.27651888132095337, | |
| "learning_rate": 8.482077846294308e-05, | |
| "loss": 2.145, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9009009009009009, | |
| "eval_loss": 1.7854279279708862, | |
| "eval_runtime": 1.9269, | |
| "eval_samples_per_second": 48.782, | |
| "eval_steps_per_second": 12.455, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.918918918918919, | |
| "grad_norm": 0.30430641770362854, | |
| "learning_rate": 8.409585265545509e-05, | |
| "loss": 2.0709, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.9369369369369369, | |
| "grad_norm": 0.5866444110870361, | |
| "learning_rate": 8.335727512379534e-05, | |
| "loss": 2.8888, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.954954954954955, | |
| "grad_norm": 0.13216109573841095, | |
| "learning_rate": 8.260534158887876e-05, | |
| "loss": 1.7877, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.972972972972973, | |
| "grad_norm": 0.7443955540657043, | |
| "learning_rate": 8.184035311926396e-05, | |
| "loss": 1.1889, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.990990990990991, | |
| "grad_norm": 0.31532421708106995, | |
| "learning_rate": 8.106261601060772e-05, | |
| "loss": 1.7782, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.009009009009009, | |
| "grad_norm": 0.5412552356719971, | |
| "learning_rate": 8.027244166302642e-05, | |
| "loss": 3.1035, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.027027027027027, | |
| "grad_norm": 0.21196092665195465, | |
| "learning_rate": 7.947014645641379e-05, | |
| "loss": 2.5004, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.045045045045045, | |
| "grad_norm": 0.18983232975006104, | |
| "learning_rate": 7.865605162376486e-05, | |
| "loss": 1.9003, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.063063063063063, | |
| "grad_norm": 0.2055334746837616, | |
| "learning_rate": 7.783048312255653e-05, | |
| "loss": 1.7878, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0810810810810811, | |
| "grad_norm": 0.2857997715473175, | |
| "learning_rate": 7.699377150423672e-05, | |
| "loss": 1.5162, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0990990990990992, | |
| "grad_norm": 0.7451598048210144, | |
| "learning_rate": 7.614625178187402e-05, | |
| "loss": 0.8527, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.117117117117117, | |
| "grad_norm": 0.6373011469841003, | |
| "learning_rate": 7.528826329602099e-05, | |
| "loss": 0.8998, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.135135135135135, | |
| "grad_norm": 0.5723519325256348, | |
| "learning_rate": 7.442014957884472e-05, | |
| "loss": 0.8521, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.1531531531531531, | |
| "grad_norm": 0.27271348237991333, | |
| "learning_rate": 7.354225821657914e-05, | |
| "loss": 2.143, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.1711711711711712, | |
| "grad_norm": 0.23478277027606964, | |
| "learning_rate": 7.265494071035401e-05, | |
| "loss": 1.7943, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.1891891891891893, | |
| "grad_norm": 0.28418228030204773, | |
| "learning_rate": 7.175855233545668e-05, | |
| "loss": 1.7495, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.2072072072072073, | |
| "grad_norm": 0.35334891080856323, | |
| "learning_rate": 7.085345199908235e-05, | |
| "loss": 2.0141, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.2252252252252251, | |
| "grad_norm": 0.5336344242095947, | |
| "learning_rate": 6.994000209663036e-05, | |
| "loss": 2.6846, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.2432432432432432, | |
| "grad_norm": 0.25877565145492554, | |
| "learning_rate": 6.901856836660386e-05, | |
| "loss": 1.2072, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.2612612612612613, | |
| "grad_norm": 0.18422676622867584, | |
| "learning_rate": 6.808951974417078e-05, | |
| "loss": 2.24, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2792792792792793, | |
| "grad_norm": 0.2145569920539856, | |
| "learning_rate": 6.715322821344494e-05, | |
| "loss": 2.0781, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.2972972972972974, | |
| "grad_norm": 0.2036983072757721, | |
| "learning_rate": 6.621006865854644e-05, | |
| "loss": 1.6265, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.3153153153153152, | |
| "grad_norm": 0.2321735918521881, | |
| "learning_rate": 6.526041871350086e-05, | |
| "loss": 1.5306, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 1.8508878946304321, | |
| "learning_rate": 6.43046586110374e-05, | |
| "loss": 0.7388, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.3513513513513513, | |
| "grad_norm": 1.0745056867599487, | |
| "learning_rate": 6.334317103034652e-05, | |
| "loss": 0.6945, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.3693693693693694, | |
| "grad_norm": 0.642358124256134, | |
| "learning_rate": 6.237634094385813e-05, | |
| "loss": 0.7758, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.3873873873873874, | |
| "grad_norm": 0.2824864685535431, | |
| "learning_rate": 6.140455546310148e-05, | |
| "loss": 2.0093, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.4054054054054055, | |
| "grad_norm": 0.35498249530792236, | |
| "learning_rate": 6.042820368370854e-05, | |
| "loss": 2.0257, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.4234234234234235, | |
| "grad_norm": 0.26013150811195374, | |
| "learning_rate": 5.944767652962309e-05, | |
| "loss": 1.6146, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.4414414414414414, | |
| "grad_norm": 0.29302212595939636, | |
| "learning_rate": 5.8463366596577706e-05, | |
| "loss": 2.1383, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.4594594594594594, | |
| "grad_norm": 0.3925705850124359, | |
| "learning_rate": 5.747566799490132e-05, | |
| "loss": 2.3427, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.4774774774774775, | |
| "grad_norm": 0.3041682839393616, | |
| "learning_rate": 5.648497619172042e-05, | |
| "loss": 1.6635, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.4954954954954955, | |
| "grad_norm": 0.19049213826656342, | |
| "learning_rate": 5.549168785261698e-05, | |
| "loss": 2.3609, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.5135135135135136, | |
| "grad_norm": 0.15779972076416016, | |
| "learning_rate": 5.4496200682806495e-05, | |
| "loss": 1.6961, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.5315315315315314, | |
| "grad_norm": 0.19497588276863098, | |
| "learning_rate": 5.3498913267899864e-05, | |
| "loss": 1.9126, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.5495495495495497, | |
| "grad_norm": 0.17658522725105286, | |
| "learning_rate": 5.250022491431259e-05, | |
| "loss": 1.8209, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.5675675675675675, | |
| "grad_norm": 1.0437264442443848, | |
| "learning_rate": 5.150053548938557e-05, | |
| "loss": 0.6212, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.5855855855855856, | |
| "grad_norm": 0.8192980885505676, | |
| "learning_rate": 5.0500245261281175e-05, | |
| "loss": 0.7097, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.6036036036036037, | |
| "grad_norm": 0.643914520740509, | |
| "learning_rate": 4.949975473871884e-05, | |
| "loss": 0.5015, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.6216216216216215, | |
| "grad_norm": 0.2833811044692993, | |
| "learning_rate": 4.849946451061443e-05, | |
| "loss": 1.706, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.6396396396396398, | |
| "grad_norm": 0.25219082832336426, | |
| "learning_rate": 4.749977508568742e-05, | |
| "loss": 2.0273, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.6576576576576576, | |
| "grad_norm": 0.2850624620914459, | |
| "learning_rate": 4.650108673210015e-05, | |
| "loss": 1.5778, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.6756756756756757, | |
| "grad_norm": 0.37064990401268005, | |
| "learning_rate": 4.550379931719351e-05, | |
| "loss": 1.9609, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.6936936936936937, | |
| "grad_norm": 0.4350699186325073, | |
| "learning_rate": 4.4508312147383036e-05, | |
| "loss": 1.9915, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.7117117117117115, | |
| "grad_norm": 0.27014970779418945, | |
| "learning_rate": 4.3515023808279586e-05, | |
| "loss": 1.7927, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.7297297297297298, | |
| "grad_norm": 0.1641526073217392, | |
| "learning_rate": 4.252433200509869e-05, | |
| "loss": 2.053, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.7477477477477477, | |
| "grad_norm": 0.1500796228647232, | |
| "learning_rate": 4.1536633403422306e-05, | |
| "loss": 1.5443, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.7657657657657657, | |
| "grad_norm": 0.19446995854377747, | |
| "learning_rate": 4.0552323470376916e-05, | |
| "loss": 2.1779, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.7837837837837838, | |
| "grad_norm": 0.18095970153808594, | |
| "learning_rate": 3.9571796316291476e-05, | |
| "loss": 1.5048, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.8018018018018018, | |
| "grad_norm": 0.5774995684623718, | |
| "learning_rate": 3.859544453689853e-05, | |
| "loss": 0.5426, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.8018018018018018, | |
| "eval_loss": 1.561373233795166, | |
| "eval_runtime": 1.8948, | |
| "eval_samples_per_second": 49.611, | |
| "eval_steps_per_second": 12.667, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.8198198198198199, | |
| "grad_norm": 0.4646216034889221, | |
| "learning_rate": 3.762365905614187e-05, | |
| "loss": 0.6454, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.8378378378378377, | |
| "grad_norm": 0.4651610851287842, | |
| "learning_rate": 3.665682896965349e-05, | |
| "loss": 0.6244, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.855855855855856, | |
| "grad_norm": 0.3666508197784424, | |
| "learning_rate": 3.5695341388962614e-05, | |
| "loss": 0.8421, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.8738738738738738, | |
| "grad_norm": 0.20598755776882172, | |
| "learning_rate": 3.473958128649915e-05, | |
| "loss": 1.7942, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.8918918918918919, | |
| "grad_norm": 0.24941736459732056, | |
| "learning_rate": 3.378993134145356e-05, | |
| "loss": 1.8199, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.90990990990991, | |
| "grad_norm": 0.2905518412590027, | |
| "learning_rate": 3.284677178655507e-05, | |
| "loss": 1.7551, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.9279279279279278, | |
| "grad_norm": 0.3726566731929779, | |
| "learning_rate": 3.1910480255829237e-05, | |
| "loss": 2.3136, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.945945945945946, | |
| "grad_norm": 0.300400972366333, | |
| "learning_rate": 3.098143163339615e-05, | |
| "loss": 2.2299, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.9639639639639639, | |
| "grad_norm": 0.1326655149459839, | |
| "learning_rate": 3.0059997903369656e-05, | |
| "loss": 1.2973, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.981981981981982, | |
| "grad_norm": 0.19319520890712738, | |
| "learning_rate": 2.914654800091768e-05, | |
| "loss": 1.0709, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.7764068245887756, | |
| "learning_rate": 2.824144766454333e-05, | |
| "loss": 2.9477, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.018018018018018, | |
| "grad_norm": 0.1426096111536026, | |
| "learning_rate": 2.7345059289646008e-05, | |
| "loss": 1.6133, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.036036036036036, | |
| "grad_norm": 0.19993425905704498, | |
| "learning_rate": 2.6457741783420886e-05, | |
| "loss": 2.1937, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.054054054054054, | |
| "grad_norm": 0.1575198620557785, | |
| "learning_rate": 2.5579850421155293e-05, | |
| "loss": 1.5695, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.0720720720720722, | |
| "grad_norm": 0.17316873371601105, | |
| "learning_rate": 2.4711736703979018e-05, | |
| "loss": 1.8888, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.09009009009009, | |
| "grad_norm": 0.1797063797712326, | |
| "learning_rate": 2.3853748218126e-05, | |
| "loss": 0.7036, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.108108108108108, | |
| "grad_norm": 0.17858926951885223, | |
| "learning_rate": 2.3006228495763295e-05, | |
| "loss": 0.4263, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.126126126126126, | |
| "grad_norm": 0.18058860301971436, | |
| "learning_rate": 2.2169516877443485e-05, | |
| "loss": 0.4322, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.144144144144144, | |
| "grad_norm": 0.18085725605487823, | |
| "learning_rate": 2.1343948376235144e-05, | |
| "loss": 1.5653, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.1621621621621623, | |
| "grad_norm": 0.2323240488767624, | |
| "learning_rate": 2.052985354358622e-05, | |
| "loss": 1.9138, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.18018018018018, | |
| "grad_norm": 0.24628163874149323, | |
| "learning_rate": 1.9727558336973595e-05, | |
| "loss": 1.7247, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.1981981981981984, | |
| "grad_norm": 0.2715443968772888, | |
| "learning_rate": 1.8937383989392294e-05, | |
| "loss": 1.701, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.2162162162162162, | |
| "grad_norm": 0.3751051723957062, | |
| "learning_rate": 1.8159646880736036e-05, | |
| "loss": 1.9297, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.234234234234234, | |
| "grad_norm": 0.45991936326026917, | |
| "learning_rate": 1.739465841112125e-05, | |
| "loss": 2.0372, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.2522522522522523, | |
| "grad_norm": 0.1275320202112198, | |
| "learning_rate": 1.664272487620466e-05, | |
| "loss": 1.4902, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.27027027027027, | |
| "grad_norm": 0.15491849184036255, | |
| "learning_rate": 1.590414734454493e-05, | |
| "loss": 2.1465, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.2882882882882885, | |
| "grad_norm": 0.16037805378437042, | |
| "learning_rate": 1.517922153705692e-05, | |
| "loss": 1.8524, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.3063063063063063, | |
| "grad_norm": 0.19121010601520538, | |
| "learning_rate": 1.4468237708607397e-05, | |
| "loss": 1.7979, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.3243243243243246, | |
| "grad_norm": 0.19223779439926147, | |
| "learning_rate": 1.3771480531799052e-05, | |
| "loss": 1.6058, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.3423423423423424, | |
| "grad_norm": 0.14566901326179504, | |
| "learning_rate": 1.308922898298977e-05, | |
| "loss": 0.3995, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.3603603603603602, | |
| "grad_norm": 0.16846461594104767, | |
| "learning_rate": 1.2421756230592534e-05, | |
| "loss": 0.4075, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.3783783783783785, | |
| "grad_norm": 0.1482275426387787, | |
| "learning_rate": 1.1769329525700935e-05, | |
| "loss": 0.4448, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.3963963963963963, | |
| "grad_norm": 0.24686311185359955, | |
| "learning_rate": 1.1132210095083694e-05, | |
| "loss": 1.5153, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.4144144144144146, | |
| "grad_norm": 0.2111864537000656, | |
| "learning_rate": 1.0510653036591583e-05, | |
| "loss": 1.839, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.4324324324324325, | |
| "grad_norm": 0.2346525639295578, | |
| "learning_rate": 9.904907217018e-06, | |
| "loss": 1.3728, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.4504504504504503, | |
| "grad_norm": 0.32989177107810974, | |
| "learning_rate": 9.31521517245469e-06, | |
| "loss": 2.1364, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.4684684684684686, | |
| "grad_norm": 0.47966253757476807, | |
| "learning_rate": 8.741813011182014e-06, | |
| "loss": 2.2386, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.4864864864864864, | |
| "grad_norm": 0.13256965577602386, | |
| "learning_rate": 8.18493031913305e-06, | |
| "loss": 1.5215, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.5045045045045047, | |
| "grad_norm": 0.1745923012495041, | |
| "learning_rate": 7.644790067969005e-06, | |
| "loss": 2.0639, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.5225225225225225, | |
| "grad_norm": 0.2047472447156906, | |
| "learning_rate": 7.1216085258031414e-06, | |
| "loss": 2.0853, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.5405405405405403, | |
| "grad_norm": 0.1631387323141098, | |
| "learning_rate": 6.6155951706085405e-06, | |
| "loss": 1.4938, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.5585585585585586, | |
| "grad_norm": 0.14580529928207397, | |
| "learning_rate": 6.1269526063447765e-06, | |
| "loss": 0.5077, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.5765765765765765, | |
| "grad_norm": 0.17946287989616394, | |
| "learning_rate": 5.6558764818367195e-06, | |
| "loss": 0.6124, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.5945945945945947, | |
| "grad_norm": 0.137393519282341, | |
| "learning_rate": 5.2025554124383095e-06, | |
| "loss": 0.4831, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.6126126126126126, | |
| "grad_norm": 0.19595730304718018, | |
| "learning_rate": 4.767170904512292e-06, | |
| "loss": 1.3957, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.6306306306306304, | |
| "grad_norm": 0.2703566551208496, | |
| "learning_rate": 4.349897282756487e-06, | |
| "loss": 2.0737, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.6486486486486487, | |
| "grad_norm": 0.2205105572938919, | |
| "learning_rate": 3.95090162040545e-06, | |
| "loss": 1.8425, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.2561565041542053, | |
| "learning_rate": 3.570343672335641e-06, | |
| "loss": 1.7804, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.684684684684685, | |
| "grad_norm": 0.3843667805194855, | |
| "learning_rate": 3.2083758111006945e-06, | |
| "loss": 1.705, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.7027027027027026, | |
| "grad_norm": 0.4763340353965759, | |
| "learning_rate": 2.86514296592269e-06, | |
| "loss": 2.2889, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.7027027027027026, | |
| "eval_loss": 1.503947377204895, | |
| "eval_runtime": 1.8948, | |
| "eval_samples_per_second": 49.609, | |
| "eval_steps_per_second": 12.666, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.7207207207207205, | |
| "grad_norm": 0.17292317748069763, | |
| "learning_rate": 2.54078256466348e-06, | |
| "loss": 1.8323, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.7387387387387387, | |
| "grad_norm": 0.17607711255550385, | |
| "learning_rate": 2.2354244787996748e-06, | |
| "loss": 2.1262, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.756756756756757, | |
| "grad_norm": 0.1533748358488083, | |
| "learning_rate": 1.9491909714230204e-06, | |
| "loss": 1.8804, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.774774774774775, | |
| "grad_norm": 0.16682887077331543, | |
| "learning_rate": 1.6821966482872264e-06, | |
| "loss": 1.7182, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.7927927927927927, | |
| "grad_norm": 0.16856834292411804, | |
| "learning_rate": 1.434548411920622e-06, | |
| "loss": 1.1743, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.810810810810811, | |
| "grad_norm": 0.13685716688632965, | |
| "learning_rate": 1.206345418823235e-06, | |
| "loss": 0.4568, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.828828828828829, | |
| "grad_norm": 0.1583159863948822, | |
| "learning_rate": 9.976790397652315e-07, | |
| "loss": 0.4332, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.846846846846847, | |
| "grad_norm": 0.20499488711357117, | |
| "learning_rate": 8.086328232027873e-07, | |
| "loss": 0.5182, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.864864864864865, | |
| "grad_norm": 0.22555167973041534, | |
| "learning_rate": 6.392824618258519e-07, | |
| "loss": 1.8121, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.8828828828828827, | |
| "grad_norm": 0.22527030110359192, | |
| "learning_rate": 4.896957622514298e-07, | |
| "loss": 1.6064, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.900900900900901, | |
| "grad_norm": 0.2471853345632553, | |
| "learning_rate": 3.599326178742535e-07, | |
| "loss": 1.5183, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.918918918918919, | |
| "grad_norm": 0.3479529321193695, | |
| "learning_rate": 2.500449848859776e-07, | |
| "loss": 1.9306, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.936936936936937, | |
| "grad_norm": 0.47280144691467285, | |
| "learning_rate": 1.6007686147225254e-07, | |
| "loss": 2.3789, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.954954954954955, | |
| "grad_norm": 0.1545385867357254, | |
| "learning_rate": 9.006427019622176e-08, | |
| "loss": 1.9687, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.972972972972973, | |
| "grad_norm": 0.15617170929908752, | |
| "learning_rate": 4.0035243575342605e-08, | |
| "loss": 0.727, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.990990990990991, | |
| "grad_norm": 0.20388685166835785, | |
| "learning_rate": 1.0009812857370016e-08, | |
| "loss": 1.3957, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 3.009009009009009, | |
| "grad_norm": 0.4190056324005127, | |
| "learning_rate": 0.0, | |
| "loss": 2.7015, | |
| "step": 167 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 167, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.268381537599488e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |