| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.0, | |
| "eval_steps": 500, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "entropy": 0.5240228176116943, | |
| "epoch": 0.04, | |
| "grad_norm": 37.29308319091797, | |
| "learning_rate": 2e-05, | |
| "loss": 6.9422, | |
| "mean_token_accuracy": 0.47347480058670044, | |
| "num_tokens": 758.0, | |
| "step": 1 | |
| }, | |
| { | |
| "entropy": 0.5185598134994507, | |
| "epoch": 0.08, | |
| "grad_norm": 37.85886764526367, | |
| "learning_rate": 2e-05, | |
| "loss": 6.8565, | |
| "mean_token_accuracy": 0.47270306944847107, | |
| "num_tokens": 1513.0, | |
| "step": 2 | |
| }, | |
| { | |
| "entropy": 0.5363126993179321, | |
| "epoch": 0.12, | |
| "grad_norm": 36.92781066894531, | |
| "learning_rate": 2e-05, | |
| "loss": 6.6136, | |
| "mean_token_accuracy": 0.47562581300735474, | |
| "num_tokens": 2276.0, | |
| "step": 3 | |
| }, | |
| { | |
| "entropy": 0.5541151762008667, | |
| "epoch": 0.16, | |
| "grad_norm": 37.46426773071289, | |
| "learning_rate": 2e-05, | |
| "loss": 6.4114, | |
| "mean_token_accuracy": 0.47200000286102295, | |
| "num_tokens": 3030.0, | |
| "step": 4 | |
| }, | |
| { | |
| "entropy": 0.6072134375572205, | |
| "epoch": 0.2, | |
| "grad_norm": 36.258392333984375, | |
| "learning_rate": 2e-05, | |
| "loss": 5.9563, | |
| "mean_token_accuracy": 0.4704301059246063, | |
| "num_tokens": 3778.0, | |
| "step": 5 | |
| }, | |
| { | |
| "entropy": 0.6346766948699951, | |
| "epoch": 0.24, | |
| "grad_norm": 34.2164421081543, | |
| "learning_rate": 2e-05, | |
| "loss": 5.4104, | |
| "mean_token_accuracy": 0.48868176341056824, | |
| "num_tokens": 4533.0, | |
| "step": 6 | |
| }, | |
| { | |
| "entropy": 0.699142575263977, | |
| "epoch": 0.28, | |
| "grad_norm": 31.32328987121582, | |
| "learning_rate": 2e-05, | |
| "loss": 4.9817, | |
| "mean_token_accuracy": 0.48806366324424744, | |
| "num_tokens": 5291.0, | |
| "step": 7 | |
| }, | |
| { | |
| "entropy": 0.7639284133911133, | |
| "epoch": 0.32, | |
| "grad_norm": 27.981550216674805, | |
| "learning_rate": 2e-05, | |
| "loss": 4.5891, | |
| "mean_token_accuracy": 0.5134048461914062, | |
| "num_tokens": 6041.0, | |
| "step": 8 | |
| }, | |
| { | |
| "entropy": 0.8184663653373718, | |
| "epoch": 0.36, | |
| "grad_norm": 23.888887405395508, | |
| "learning_rate": 2e-05, | |
| "loss": 4.1021, | |
| "mean_token_accuracy": 0.535380482673645, | |
| "num_tokens": 6794.0, | |
| "step": 9 | |
| }, | |
| { | |
| "entropy": 0.8713943362236023, | |
| "epoch": 0.4, | |
| "grad_norm": 21.27410125732422, | |
| "learning_rate": 2e-05, | |
| "loss": 3.863, | |
| "mean_token_accuracy": 0.5392809510231018, | |
| "num_tokens": 7549.0, | |
| "step": 10 | |
| }, | |
| { | |
| "entropy": 0.9519356489181519, | |
| "epoch": 0.44, | |
| "grad_norm": 18.26158905029297, | |
| "learning_rate": 2e-05, | |
| "loss": 3.5874, | |
| "mean_token_accuracy": 0.5420560836791992, | |
| "num_tokens": 8302.0, | |
| "step": 11 | |
| }, | |
| { | |
| "entropy": 1.025558352470398, | |
| "epoch": 0.48, | |
| "grad_norm": 16.051610946655273, | |
| "learning_rate": 2e-05, | |
| "loss": 3.3905, | |
| "mean_token_accuracy": 0.5483443737030029, | |
| "num_tokens": 9061.0, | |
| "step": 12 | |
| }, | |
| { | |
| "entropy": 1.0597870349884033, | |
| "epoch": 0.52, | |
| "grad_norm": 13.387225151062012, | |
| "learning_rate": 2e-05, | |
| "loss": 3.1501, | |
| "mean_token_accuracy": 0.5702811479568481, | |
| "num_tokens": 9812.0, | |
| "step": 13 | |
| }, | |
| { | |
| "entropy": 1.1256167888641357, | |
| "epoch": 0.56, | |
| "grad_norm": 11.642187118530273, | |
| "learning_rate": 2e-05, | |
| "loss": 2.9905, | |
| "mean_token_accuracy": 0.5778961181640625, | |
| "num_tokens": 10567.0, | |
| "step": 14 | |
| }, | |
| { | |
| "entropy": 1.1595126390457153, | |
| "epoch": 0.6, | |
| "grad_norm": 10.657787322998047, | |
| "learning_rate": 2e-05, | |
| "loss": 2.8271, | |
| "mean_token_accuracy": 0.5927419066429138, | |
| "num_tokens": 11315.0, | |
| "step": 15 | |
| }, | |
| { | |
| "entropy": 1.2265969514846802, | |
| "epoch": 0.64, | |
| "grad_norm": 9.537993431091309, | |
| "learning_rate": 2e-05, | |
| "loss": 2.749, | |
| "mean_token_accuracy": 0.5964912176132202, | |
| "num_tokens": 12060.0, | |
| "step": 16 | |
| }, | |
| { | |
| "entropy": 1.2593307495117188, | |
| "epoch": 0.68, | |
| "grad_norm": 8.315566062927246, | |
| "learning_rate": 2e-05, | |
| "loss": 2.5743, | |
| "mean_token_accuracy": 0.6169934868812561, | |
| "num_tokens": 12829.0, | |
| "step": 17 | |
| }, | |
| { | |
| "entropy": 1.3245295286178589, | |
| "epoch": 0.72, | |
| "grad_norm": 7.915526390075684, | |
| "learning_rate": 2e-05, | |
| "loss": 2.5367, | |
| "mean_token_accuracy": 0.6194332242012024, | |
| "num_tokens": 13574.0, | |
| "step": 18 | |
| }, | |
| { | |
| "entropy": 1.3604302406311035, | |
| "epoch": 0.76, | |
| "grad_norm": 7.328187942504883, | |
| "learning_rate": 2e-05, | |
| "loss": 2.4018, | |
| "mean_token_accuracy": 0.6255033612251282, | |
| "num_tokens": 14323.0, | |
| "step": 19 | |
| }, | |
| { | |
| "entropy": 1.3951760530471802, | |
| "epoch": 0.8, | |
| "grad_norm": 6.733570098876953, | |
| "learning_rate": 2e-05, | |
| "loss": 2.317, | |
| "mean_token_accuracy": 0.635046124458313, | |
| "num_tokens": 15086.0, | |
| "step": 20 | |
| }, | |
| { | |
| "entropy": 1.4354883432388306, | |
| "epoch": 0.84, | |
| "grad_norm": 6.525688648223877, | |
| "learning_rate": 2e-05, | |
| "loss": 2.2795, | |
| "mean_token_accuracy": 0.6360052824020386, | |
| "num_tokens": 15851.0, | |
| "step": 21 | |
| }, | |
| { | |
| "entropy": 1.4653663635253906, | |
| "epoch": 0.88, | |
| "grad_norm": 6.118790149688721, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1621, | |
| "mean_token_accuracy": 0.6416107416152954, | |
| "num_tokens": 16600.0, | |
| "step": 22 | |
| }, | |
| { | |
| "entropy": 1.4949076175689697, | |
| "epoch": 0.92, | |
| "grad_norm": 5.929473400115967, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1265, | |
| "mean_token_accuracy": 0.6526458859443665, | |
| "num_tokens": 17341.0, | |
| "step": 23 | |
| }, | |
| { | |
| "entropy": 1.4969980716705322, | |
| "epoch": 0.96, | |
| "grad_norm": 5.519643783569336, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0365, | |
| "mean_token_accuracy": 0.6627296805381775, | |
| "num_tokens": 18107.0, | |
| "step": 24 | |
| }, | |
| { | |
| "entropy": 1.5308914184570312, | |
| "epoch": 1.0, | |
| "grad_norm": 5.461964130401611, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9552, | |
| "mean_token_accuracy": 0.6851119995117188, | |
| "num_tokens": 18870.0, | |
| "step": 25 | |
| }, | |
| { | |
| "entropy": 1.531392216682434, | |
| "epoch": 1.04, | |
| "grad_norm": 5.313822269439697, | |
| "learning_rate": 2e-05, | |
| "loss": 1.8633, | |
| "mean_token_accuracy": 0.7098515629768372, | |
| "num_tokens": 19615.0, | |
| "step": 26 | |
| }, | |
| { | |
| "entropy": 1.5274267196655273, | |
| "epoch": 1.08, | |
| "grad_norm": 5.325898170471191, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9049, | |
| "mean_token_accuracy": 0.7068965435028076, | |
| "num_tokens": 20373.0, | |
| "step": 27 | |
| }, | |
| { | |
| "entropy": 1.5113599300384521, | |
| "epoch": 1.12, | |
| "grad_norm": 5.022428512573242, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7917, | |
| "mean_token_accuracy": 0.7334217429161072, | |
| "num_tokens": 21131.0, | |
| "step": 28 | |
| }, | |
| { | |
| "entropy": 1.4889423847198486, | |
| "epoch": 1.16, | |
| "grad_norm": 4.645327568054199, | |
| "learning_rate": 2e-05, | |
| "loss": 1.7032, | |
| "mean_token_accuracy": 0.7447090148925781, | |
| "num_tokens": 21891.0, | |
| "step": 29 | |
| }, | |
| { | |
| "entropy": 1.4870576858520508, | |
| "epoch": 1.2, | |
| "grad_norm": 4.541719913482666, | |
| "learning_rate": 2e-05, | |
| "loss": 1.6655, | |
| "mean_token_accuracy": 0.7360000014305115, | |
| "num_tokens": 22645.0, | |
| "step": 30 | |
| }, | |
| { | |
| "entropy": 1.4613386392593384, | |
| "epoch": 1.24, | |
| "grad_norm": 4.408016204833984, | |
| "learning_rate": 2e-05, | |
| "loss": 1.6312, | |
| "mean_token_accuracy": 0.7432795763015747, | |
| "num_tokens": 23393.0, | |
| "step": 31 | |
| }, | |
| { | |
| "entropy": 1.4358580112457275, | |
| "epoch": 1.28, | |
| "grad_norm": 4.297278881072998, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5793, | |
| "mean_token_accuracy": 0.7510094046592712, | |
| "num_tokens": 24140.0, | |
| "step": 32 | |
| }, | |
| { | |
| "entropy": 1.3814220428466797, | |
| "epoch": 1.32, | |
| "grad_norm": 4.178844451904297, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4675, | |
| "mean_token_accuracy": 0.7724867463111877, | |
| "num_tokens": 24900.0, | |
| "step": 33 | |
| }, | |
| { | |
| "entropy": 1.3687061071395874, | |
| "epoch": 1.3599999999999999, | |
| "grad_norm": 4.218804836273193, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4288, | |
| "mean_token_accuracy": 0.7776280045509338, | |
| "num_tokens": 25646.0, | |
| "step": 34 | |
| }, | |
| { | |
| "entropy": 1.3419899940490723, | |
| "epoch": 1.4, | |
| "grad_norm": 4.205630302429199, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4188, | |
| "mean_token_accuracy": 0.7807486653327942, | |
| "num_tokens": 26398.0, | |
| "step": 35 | |
| }, | |
| { | |
| "entropy": 1.2921583652496338, | |
| "epoch": 1.44, | |
| "grad_norm": 3.8976969718933105, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3454, | |
| "mean_token_accuracy": 0.7848605513572693, | |
| "num_tokens": 27155.0, | |
| "step": 36 | |
| }, | |
| { | |
| "entropy": 1.2668341398239136, | |
| "epoch": 1.48, | |
| "grad_norm": 3.746513843536377, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3557, | |
| "mean_token_accuracy": 0.7797062993049622, | |
| "num_tokens": 27908.0, | |
| "step": 37 | |
| }, | |
| { | |
| "entropy": 1.228274941444397, | |
| "epoch": 1.52, | |
| "grad_norm": 3.520526170730591, | |
| "learning_rate": 2e-05, | |
| "loss": 1.2668, | |
| "mean_token_accuracy": 0.8029294013977051, | |
| "num_tokens": 28663.0, | |
| "step": 38 | |
| }, | |
| { | |
| "entropy": 1.1697431802749634, | |
| "epoch": 1.56, | |
| "grad_norm": 3.4416074752807617, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1742, | |
| "mean_token_accuracy": 0.8176870942115784, | |
| "num_tokens": 29402.0, | |
| "step": 39 | |
| }, | |
| { | |
| "entropy": 1.130076289176941, | |
| "epoch": 1.6, | |
| "grad_norm": 3.223904848098755, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1646, | |
| "mean_token_accuracy": 0.8208168745040894, | |
| "num_tokens": 30165.0, | |
| "step": 40 | |
| }, | |
| { | |
| "entropy": 1.1070514917373657, | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 3.213395118713379, | |
| "learning_rate": 2e-05, | |
| "loss": 1.1102, | |
| "mean_token_accuracy": 0.8348745107650757, | |
| "num_tokens": 30926.0, | |
| "step": 41 | |
| }, | |
| { | |
| "entropy": 1.0644617080688477, | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 3.1568286418914795, | |
| "learning_rate": 2e-05, | |
| "loss": 1.0741, | |
| "mean_token_accuracy": 0.8320000171661377, | |
| "num_tokens": 31680.0, | |
| "step": 42 | |
| }, | |
| { | |
| "entropy": 1.031175971031189, | |
| "epoch": 1.72, | |
| "grad_norm": 3.1758995056152344, | |
| "learning_rate": 2e-05, | |
| "loss": 1.0527, | |
| "mean_token_accuracy": 0.8251001238822937, | |
| "num_tokens": 32433.0, | |
| "step": 43 | |
| }, | |
| { | |
| "entropy": 0.9702178835868835, | |
| "epoch": 1.76, | |
| "grad_norm": 3.0158236026763916, | |
| "learning_rate": 2e-05, | |
| "loss": 0.9569, | |
| "mean_token_accuracy": 0.8379814028739929, | |
| "num_tokens": 33190.0, | |
| "step": 44 | |
| }, | |
| { | |
| "entropy": 0.9418991208076477, | |
| "epoch": 1.8, | |
| "grad_norm": 3.095254898071289, | |
| "learning_rate": 2e-05, | |
| "loss": 0.9229, | |
| "mean_token_accuracy": 0.852150559425354, | |
| "num_tokens": 33938.0, | |
| "step": 45 | |
| }, | |
| { | |
| "entropy": 0.8792616128921509, | |
| "epoch": 1.8399999999999999, | |
| "grad_norm": 3.1904547214508057, | |
| "learning_rate": 2e-05, | |
| "loss": 0.8814, | |
| "mean_token_accuracy": 0.8521400690078735, | |
| "num_tokens": 34713.0, | |
| "step": 46 | |
| }, | |
| { | |
| "entropy": 0.8644114136695862, | |
| "epoch": 1.88, | |
| "grad_norm": 3.1296632289886475, | |
| "learning_rate": 2e-05, | |
| "loss": 0.9185, | |
| "mean_token_accuracy": 0.8663967847824097, | |
| "num_tokens": 35458.0, | |
| "step": 47 | |
| }, | |
| { | |
| "entropy": 0.8369636535644531, | |
| "epoch": 1.92, | |
| "grad_norm": 3.080310106277466, | |
| "learning_rate": 2e-05, | |
| "loss": 0.8598, | |
| "mean_token_accuracy": 0.875, | |
| "num_tokens": 36214.0, | |
| "step": 48 | |
| }, | |
| { | |
| "entropy": 0.7779583930969238, | |
| "epoch": 1.96, | |
| "grad_norm": 2.9825854301452637, | |
| "learning_rate": 2e-05, | |
| "loss": 0.757, | |
| "mean_token_accuracy": 0.8845144510269165, | |
| "num_tokens": 36980.0, | |
| "step": 49 | |
| }, | |
| { | |
| "entropy": 0.7366280555725098, | |
| "epoch": 2.0, | |
| "grad_norm": 3.0169920921325684, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7402, | |
| "mean_token_accuracy": 0.8928571343421936, | |
| "num_tokens": 37740.0, | |
| "step": 50 | |
| }, | |
| { | |
| "entropy": 0.7010611891746521, | |
| "epoch": 2.04, | |
| "grad_norm": 3.399986743927002, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7593, | |
| "mean_token_accuracy": 0.8872870206832886, | |
| "num_tokens": 38507.0, | |
| "step": 51 | |
| }, | |
| { | |
| "entropy": 0.6273115873336792, | |
| "epoch": 2.08, | |
| "grad_norm": 3.619697332382202, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6621, | |
| "mean_token_accuracy": 0.9089692234992981, | |
| "num_tokens": 39258.0, | |
| "step": 52 | |
| }, | |
| { | |
| "entropy": 0.5995044708251953, | |
| "epoch": 2.12, | |
| "grad_norm": 2.8443078994750977, | |
| "learning_rate": 2e-05, | |
| "loss": 0.662, | |
| "mean_token_accuracy": 0.9167767763137817, | |
| "num_tokens": 40019.0, | |
| "step": 53 | |
| }, | |
| { | |
| "entropy": 0.5483225584030151, | |
| "epoch": 2.16, | |
| "grad_norm": 2.8451199531555176, | |
| "learning_rate": 2e-05, | |
| "loss": 0.575, | |
| "mean_token_accuracy": 0.929427444934845, | |
| "num_tokens": 40774.0, | |
| "step": 54 | |
| }, | |
| { | |
| "entropy": 0.5085722208023071, | |
| "epoch": 2.2, | |
| "grad_norm": 2.7123143672943115, | |
| "learning_rate": 2e-05, | |
| "loss": 0.5672, | |
| "mean_token_accuracy": 0.9348404407501221, | |
| "num_tokens": 41530.0, | |
| "step": 55 | |
| }, | |
| { | |
| "entropy": 0.4458790123462677, | |
| "epoch": 2.24, | |
| "grad_norm": 2.668097496032715, | |
| "learning_rate": 2e-05, | |
| "loss": 0.5034, | |
| "mean_token_accuracy": 0.9419702887535095, | |
| "num_tokens": 42275.0, | |
| "step": 56 | |
| }, | |
| { | |
| "entropy": 0.3930673599243164, | |
| "epoch": 2.2800000000000002, | |
| "grad_norm": 2.5482311248779297, | |
| "learning_rate": 2e-05, | |
| "loss": 0.5078, | |
| "mean_token_accuracy": 0.9355263113975525, | |
| "num_tokens": 43039.0, | |
| "step": 57 | |
| }, | |
| { | |
| "entropy": 0.37228837609291077, | |
| "epoch": 2.32, | |
| "grad_norm": 2.443922281265259, | |
| "learning_rate": 2e-05, | |
| "loss": 0.4665, | |
| "mean_token_accuracy": 0.9482071995735168, | |
| "num_tokens": 43796.0, | |
| "step": 58 | |
| }, | |
| { | |
| "entropy": 0.3517550826072693, | |
| "epoch": 2.36, | |
| "grad_norm": 2.3463053703308105, | |
| "learning_rate": 2e-05, | |
| "loss": 0.4632, | |
| "mean_token_accuracy": 0.9378378391265869, | |
| "num_tokens": 44540.0, | |
| "step": 59 | |
| }, | |
| { | |
| "entropy": 0.33203697204589844, | |
| "epoch": 2.4, | |
| "grad_norm": 2.6731350421905518, | |
| "learning_rate": 2e-05, | |
| "loss": 0.4491, | |
| "mean_token_accuracy": 0.9372549057006836, | |
| "num_tokens": 45309.0, | |
| "step": 60 | |
| }, | |
| { | |
| "entropy": 0.31520795822143555, | |
| "epoch": 2.44, | |
| "grad_norm": 2.3959364891052246, | |
| "learning_rate": 2e-05, | |
| "loss": 0.44, | |
| "mean_token_accuracy": 0.9494680762290955, | |
| "num_tokens": 46065.0, | |
| "step": 61 | |
| }, | |
| { | |
| "entropy": 0.3004280626773834, | |
| "epoch": 2.48, | |
| "grad_norm": 2.1749024391174316, | |
| "learning_rate": 2e-05, | |
| "loss": 0.4511, | |
| "mean_token_accuracy": 0.9522546529769897, | |
| "num_tokens": 46823.0, | |
| "step": 62 | |
| }, | |
| { | |
| "entropy": 0.287904292345047, | |
| "epoch": 2.52, | |
| "grad_norm": 2.184871196746826, | |
| "learning_rate": 2e-05, | |
| "loss": 0.4058, | |
| "mean_token_accuracy": 0.9521276354789734, | |
| "num_tokens": 47579.0, | |
| "step": 63 | |
| }, | |
| { | |
| "entropy": 0.28218233585357666, | |
| "epoch": 2.56, | |
| "grad_norm": 2.27091121673584, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3679, | |
| "mean_token_accuracy": 0.9529569745063782, | |
| "num_tokens": 48327.0, | |
| "step": 64 | |
| }, | |
| { | |
| "entropy": 0.2818812429904938, | |
| "epoch": 2.6, | |
| "grad_norm": 2.9275169372558594, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3858, | |
| "mean_token_accuracy": 0.9478609561920166, | |
| "num_tokens": 49079.0, | |
| "step": 65 | |
| }, | |
| { | |
| "entropy": 0.2774568200111389, | |
| "epoch": 2.64, | |
| "grad_norm": 2.837167978286743, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3639, | |
| "mean_token_accuracy": 0.9430463314056396, | |
| "num_tokens": 49838.0, | |
| "step": 66 | |
| }, | |
| { | |
| "entropy": 0.2726022005081177, | |
| "epoch": 2.68, | |
| "grad_norm": 2.2975914478302, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3028, | |
| "mean_token_accuracy": 0.9518072009086609, | |
| "num_tokens": 50589.0, | |
| "step": 67 | |
| }, | |
| { | |
| "entropy": 0.25494953989982605, | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 2.0497214794158936, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2769, | |
| "mean_token_accuracy": 0.9570469856262207, | |
| "num_tokens": 51338.0, | |
| "step": 68 | |
| }, | |
| { | |
| "entropy": 0.2599942982196808, | |
| "epoch": 2.76, | |
| "grad_norm": 2.4737048149108887, | |
| "learning_rate": 2e-05, | |
| "loss": 0.336, | |
| "mean_token_accuracy": 0.9518072009086609, | |
| "num_tokens": 52089.0, | |
| "step": 69 | |
| }, | |
| { | |
| "entropy": 0.24719160795211792, | |
| "epoch": 2.8, | |
| "grad_norm": 2.1889803409576416, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3006, | |
| "mean_token_accuracy": 0.9491298794746399, | |
| "num_tokens": 52840.0, | |
| "step": 70 | |
| }, | |
| { | |
| "entropy": 0.24826745688915253, | |
| "epoch": 2.84, | |
| "grad_norm": 2.0455644130706787, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2918, | |
| "mean_token_accuracy": 0.948344349861145, | |
| "num_tokens": 53599.0, | |
| "step": 71 | |
| }, | |
| { | |
| "entropy": 0.2431359887123108, | |
| "epoch": 2.88, | |
| "grad_norm": 1.8650648593902588, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2582, | |
| "mean_token_accuracy": 0.9523809552192688, | |
| "num_tokens": 54359.0, | |
| "step": 72 | |
| }, | |
| { | |
| "entropy": 0.24409686028957367, | |
| "epoch": 2.92, | |
| "grad_norm": 2.3259599208831787, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2527, | |
| "mean_token_accuracy": 0.960106372833252, | |
| "num_tokens": 55115.0, | |
| "step": 73 | |
| }, | |
| { | |
| "entropy": 0.2385331094264984, | |
| "epoch": 2.96, | |
| "grad_norm": 1.3533285856246948, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2706, | |
| "mean_token_accuracy": 0.9580514430999756, | |
| "num_tokens": 55858.0, | |
| "step": 74 | |
| }, | |
| { | |
| "entropy": 0.24642708897590637, | |
| "epoch": 3.0, | |
| "grad_norm": 1.1475350856781006, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2543, | |
| "mean_token_accuracy": 0.9598930478096008, | |
| "num_tokens": 56610.0, | |
| "step": 75 | |
| }, | |
| { | |
| "entropy": 0.2539970278739929, | |
| "epoch": 3.04, | |
| "grad_norm": 1.2767738103866577, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2488, | |
| "mean_token_accuracy": 0.959785521030426, | |
| "num_tokens": 57360.0, | |
| "step": 76 | |
| }, | |
| { | |
| "entropy": 0.23603883385658264, | |
| "epoch": 3.08, | |
| "grad_norm": 1.0374062061309814, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2453, | |
| "mean_token_accuracy": 0.9582772254943848, | |
| "num_tokens": 58107.0, | |
| "step": 77 | |
| }, | |
| { | |
| "entropy": 0.24485909938812256, | |
| "epoch": 3.12, | |
| "grad_norm": 1.4987839460372925, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2312, | |
| "mean_token_accuracy": 0.9553219676017761, | |
| "num_tokens": 58872.0, | |
| "step": 78 | |
| }, | |
| { | |
| "entropy": 0.2403731644153595, | |
| "epoch": 3.16, | |
| "grad_norm": 1.1765466928482056, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2526, | |
| "mean_token_accuracy": 0.9653333425521851, | |
| "num_tokens": 59626.0, | |
| "step": 79 | |
| }, | |
| { | |
| "entropy": 0.23074936866760254, | |
| "epoch": 3.2, | |
| "grad_norm": 1.2395281791687012, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2189, | |
| "mean_token_accuracy": 0.9634641408920288, | |
| "num_tokens": 60369.0, | |
| "step": 80 | |
| }, | |
| { | |
| "entropy": 0.2089882791042328, | |
| "epoch": 3.24, | |
| "grad_norm": 1.3069368600845337, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2428, | |
| "mean_token_accuracy": 0.9638069868087769, | |
| "num_tokens": 61119.0, | |
| "step": 81 | |
| }, | |
| { | |
| "entropy": 0.21510061621665955, | |
| "epoch": 3.2800000000000002, | |
| "grad_norm": 1.1535745859146118, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2212, | |
| "mean_token_accuracy": 0.9652870297431946, | |
| "num_tokens": 61872.0, | |
| "step": 82 | |
| }, | |
| { | |
| "entropy": 0.2101171612739563, | |
| "epoch": 3.32, | |
| "grad_norm": 0.9777595400810242, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1982, | |
| "mean_token_accuracy": 0.9693741798400879, | |
| "num_tokens": 62627.0, | |
| "step": 83 | |
| }, | |
| { | |
| "entropy": 0.2099580019712448, | |
| "epoch": 3.36, | |
| "grad_norm": 1.1317825317382812, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2281, | |
| "mean_token_accuracy": 0.9669312238693237, | |
| "num_tokens": 63387.0, | |
| "step": 84 | |
| }, | |
| { | |
| "entropy": 0.20581887662410736, | |
| "epoch": 3.4, | |
| "grad_norm": 1.0572807788848877, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2033, | |
| "mean_token_accuracy": 0.9680851101875305, | |
| "num_tokens": 64143.0, | |
| "step": 85 | |
| }, | |
| { | |
| "entropy": 0.20909777283668518, | |
| "epoch": 3.44, | |
| "grad_norm": 1.5825022459030151, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2105, | |
| "mean_token_accuracy": 0.9637096524238586, | |
| "num_tokens": 64891.0, | |
| "step": 86 | |
| }, | |
| { | |
| "entropy": 0.21373741328716278, | |
| "epoch": 3.48, | |
| "grad_norm": 1.1368229389190674, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1793, | |
| "mean_token_accuracy": 0.9785522818565369, | |
| "num_tokens": 65641.0, | |
| "step": 87 | |
| }, | |
| { | |
| "entropy": 0.20942555367946625, | |
| "epoch": 3.52, | |
| "grad_norm": 1.03619384765625, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2048, | |
| "mean_token_accuracy": 0.9744966626167297, | |
| "num_tokens": 66390.0, | |
| "step": 88 | |
| }, | |
| { | |
| "entropy": 0.20285017788410187, | |
| "epoch": 3.56, | |
| "grad_norm": 0.9636878371238708, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1875, | |
| "mean_token_accuracy": 0.9801587462425232, | |
| "num_tokens": 67150.0, | |
| "step": 89 | |
| }, | |
| { | |
| "entropy": 0.20965717732906342, | |
| "epoch": 3.6, | |
| "grad_norm": 0.782873809337616, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1866, | |
| "mean_token_accuracy": 0.9747676253318787, | |
| "num_tokens": 67907.0, | |
| "step": 90 | |
| }, | |
| { | |
| "entropy": 0.17947089672088623, | |
| "epoch": 3.64, | |
| "grad_norm": 1.3582860231399536, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1939, | |
| "mean_token_accuracy": 0.9734395742416382, | |
| "num_tokens": 68664.0, | |
| "step": 91 | |
| }, | |
| { | |
| "entropy": 0.21160392463207245, | |
| "epoch": 3.68, | |
| "grad_norm": 1.0390849113464355, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1893, | |
| "mean_token_accuracy": 0.9735449552536011, | |
| "num_tokens": 69424.0, | |
| "step": 92 | |
| }, | |
| { | |
| "entropy": 0.20336918532848358, | |
| "epoch": 3.7199999999999998, | |
| "grad_norm": 1.523485541343689, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2236, | |
| "mean_token_accuracy": 0.9635416865348816, | |
| "num_tokens": 70196.0, | |
| "step": 93 | |
| }, | |
| { | |
| "entropy": 0.19648239016532898, | |
| "epoch": 3.76, | |
| "grad_norm": 1.1465239524841309, | |
| "learning_rate": 2e-05, | |
| "loss": 0.243, | |
| "mean_token_accuracy": 0.9677852392196655, | |
| "num_tokens": 70945.0, | |
| "step": 94 | |
| }, | |
| { | |
| "entropy": 0.16573190689086914, | |
| "epoch": 3.8, | |
| "grad_norm": 0.9433248043060303, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1588, | |
| "mean_token_accuracy": 0.9879999756813049, | |
| "num_tokens": 71699.0, | |
| "step": 95 | |
| }, | |
| { | |
| "entropy": 0.1826864331960678, | |
| "epoch": 3.84, | |
| "grad_norm": 0.8109387159347534, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1688, | |
| "mean_token_accuracy": 0.9801061153411865, | |
| "num_tokens": 72457.0, | |
| "step": 96 | |
| }, | |
| { | |
| "entropy": 0.16375097632408142, | |
| "epoch": 3.88, | |
| "grad_norm": 0.7381296753883362, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1766, | |
| "mean_token_accuracy": 0.9775132536888123, | |
| "num_tokens": 73217.0, | |
| "step": 97 | |
| }, | |
| { | |
| "entropy": 0.1716109961271286, | |
| "epoch": 3.92, | |
| "grad_norm": 0.8144583106040955, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1895, | |
| "mean_token_accuracy": 0.9759358167648315, | |
| "num_tokens": 73969.0, | |
| "step": 98 | |
| }, | |
| { | |
| "entropy": 0.1795879453420639, | |
| "epoch": 3.96, | |
| "grad_norm": 0.7693735957145691, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1644, | |
| "mean_token_accuracy": 0.9772727489471436, | |
| "num_tokens": 74721.0, | |
| "step": 99 | |
| }, | |
| { | |
| "entropy": 0.17074991762638092, | |
| "epoch": 4.0, | |
| "grad_norm": 1.033590316772461, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1741, | |
| "mean_token_accuracy": 0.9721854329109192, | |
| "num_tokens": 75480.0, | |
| "step": 100 | |
| }, | |
| { | |
| "entropy": 0.15769797563552856, | |
| "epoch": 4.04, | |
| "grad_norm": 0.6272271275520325, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1476, | |
| "mean_token_accuracy": 0.9814077019691467, | |
| "num_tokens": 76237.0, | |
| "step": 101 | |
| }, | |
| { | |
| "entropy": 0.16341878473758698, | |
| "epoch": 4.08, | |
| "grad_norm": 0.6367079615592957, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1513, | |
| "mean_token_accuracy": 0.9811320900917053, | |
| "num_tokens": 76983.0, | |
| "step": 102 | |
| }, | |
| { | |
| "entropy": 0.16848808526992798, | |
| "epoch": 4.12, | |
| "grad_norm": 1.333314299583435, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1655, | |
| "mean_token_accuracy": 0.9731543660163879, | |
| "num_tokens": 77732.0, | |
| "step": 103 | |
| }, | |
| { | |
| "entropy": 0.18041355907917023, | |
| "epoch": 4.16, | |
| "grad_norm": 0.7196837663650513, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1457, | |
| "mean_token_accuracy": 0.9814814925193787, | |
| "num_tokens": 78492.0, | |
| "step": 104 | |
| }, | |
| { | |
| "entropy": 0.17968420684337616, | |
| "epoch": 4.2, | |
| "grad_norm": 0.9307867884635925, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1961, | |
| "mean_token_accuracy": 0.9686684012413025, | |
| "num_tokens": 79262.0, | |
| "step": 105 | |
| }, | |
| { | |
| "entropy": 0.15099582076072693, | |
| "epoch": 4.24, | |
| "grad_norm": 0.8915977478027344, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1646, | |
| "mean_token_accuracy": 0.9771811962127686, | |
| "num_tokens": 80011.0, | |
| "step": 106 | |
| }, | |
| { | |
| "entropy": 0.15279746055603027, | |
| "epoch": 4.28, | |
| "grad_norm": 0.679995059967041, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1495, | |
| "mean_token_accuracy": 0.9787516593933105, | |
| "num_tokens": 80768.0, | |
| "step": 107 | |
| }, | |
| { | |
| "entropy": 0.16469314694404602, | |
| "epoch": 4.32, | |
| "grad_norm": 0.9293569922447205, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2148, | |
| "mean_token_accuracy": 0.9721485376358032, | |
| "num_tokens": 81526.0, | |
| "step": 108 | |
| }, | |
| { | |
| "entropy": 0.1584121435880661, | |
| "epoch": 4.36, | |
| "grad_norm": 0.7940216660499573, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1631, | |
| "mean_token_accuracy": 0.9785234928131104, | |
| "num_tokens": 82275.0, | |
| "step": 109 | |
| }, | |
| { | |
| "entropy": 0.1513449102640152, | |
| "epoch": 4.4, | |
| "grad_norm": 0.8151124119758606, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1438, | |
| "mean_token_accuracy": 0.9811066389083862, | |
| "num_tokens": 83020.0, | |
| "step": 110 | |
| }, | |
| { | |
| "entropy": 0.18052899837493896, | |
| "epoch": 4.44, | |
| "grad_norm": 0.9315986037254333, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2078, | |
| "mean_token_accuracy": 0.9696168899536133, | |
| "num_tokens": 83781.0, | |
| "step": 111 | |
| }, | |
| { | |
| "entropy": 0.1673561930656433, | |
| "epoch": 4.48, | |
| "grad_norm": 0.6802690029144287, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1581, | |
| "mean_token_accuracy": 0.9789196252822876, | |
| "num_tokens": 84544.0, | |
| "step": 112 | |
| }, | |
| { | |
| "entropy": 0.16665907204151154, | |
| "epoch": 4.52, | |
| "grad_norm": 0.7094313502311707, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1416, | |
| "mean_token_accuracy": 0.9788079261779785, | |
| "num_tokens": 85303.0, | |
| "step": 113 | |
| }, | |
| { | |
| "entropy": 0.16386385262012482, | |
| "epoch": 4.5600000000000005, | |
| "grad_norm": 0.7258614301681519, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1836, | |
| "mean_token_accuracy": 0.9759679436683655, | |
| "num_tokens": 86056.0, | |
| "step": 114 | |
| }, | |
| { | |
| "entropy": 0.1813310831785202, | |
| "epoch": 4.6, | |
| "grad_norm": 0.861827552318573, | |
| "learning_rate": 2e-05, | |
| "loss": 0.2078, | |
| "mean_token_accuracy": 0.9707057476043701, | |
| "num_tokens": 86811.0, | |
| "step": 115 | |
| }, | |
| { | |
| "entropy": 0.1732410490512848, | |
| "epoch": 4.64, | |
| "grad_norm": 0.867268443107605, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1642, | |
| "mean_token_accuracy": 0.9739243984222412, | |
| "num_tokens": 87582.0, | |
| "step": 116 | |
| }, | |
| { | |
| "entropy": 0.16808971762657166, | |
| "epoch": 4.68, | |
| "grad_norm": 0.6766126751899719, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1701, | |
| "mean_token_accuracy": 0.9813084006309509, | |
| "num_tokens": 88335.0, | |
| "step": 117 | |
| }, | |
| { | |
| "entropy": 0.16790905594825745, | |
| "epoch": 4.72, | |
| "grad_norm": 0.7375288009643555, | |
| "learning_rate": 2e-05, | |
| "loss": 0.128, | |
| "mean_token_accuracy": 0.9786950945854187, | |
| "num_tokens": 89090.0, | |
| "step": 118 | |
| }, | |
| { | |
| "entropy": 0.13895468413829803, | |
| "epoch": 4.76, | |
| "grad_norm": 0.7197297215461731, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1475, | |
| "mean_token_accuracy": 0.9786381721496582, | |
| "num_tokens": 89843.0, | |
| "step": 119 | |
| }, | |
| { | |
| "entropy": 0.17015594244003296, | |
| "epoch": 4.8, | |
| "grad_norm": 0.9060453772544861, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1565, | |
| "mean_token_accuracy": 0.9773333072662354, | |
| "num_tokens": 90597.0, | |
| "step": 120 | |
| }, | |
| { | |
| "entropy": 0.15092864632606506, | |
| "epoch": 4.84, | |
| "grad_norm": 0.6434327960014343, | |
| "learning_rate": 2e-05, | |
| "loss": 0.137, | |
| "mean_token_accuracy": 0.9786666631698608, | |
| "num_tokens": 91351.0, | |
| "step": 121 | |
| }, | |
| { | |
| "entropy": 0.1593955159187317, | |
| "epoch": 4.88, | |
| "grad_norm": 0.7469741106033325, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1584, | |
| "mean_token_accuracy": 0.9759358167648315, | |
| "num_tokens": 92103.0, | |
| "step": 122 | |
| }, | |
| { | |
| "entropy": 0.14332211017608643, | |
| "epoch": 4.92, | |
| "grad_norm": 0.8097223043441772, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1465, | |
| "mean_token_accuracy": 0.9811320900917053, | |
| "num_tokens": 92849.0, | |
| "step": 123 | |
| }, | |
| { | |
| "entropy": 0.15861551463603973, | |
| "epoch": 4.96, | |
| "grad_norm": 0.8222020268440247, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1609, | |
| "mean_token_accuracy": 0.9758713245391846, | |
| "num_tokens": 93599.0, | |
| "step": 124 | |
| }, | |
| { | |
| "entropy": 0.15850719809532166, | |
| "epoch": 5.0, | |
| "grad_norm": 0.7024567127227783, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1396, | |
| "mean_token_accuracy": 0.9799196720123291, | |
| "num_tokens": 94350.0, | |
| "step": 125 | |
| }, | |
| { | |
| "entropy": 0.1658700406551361, | |
| "epoch": 5.04, | |
| "grad_norm": 0.5833410620689392, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1314, | |
| "mean_token_accuracy": 0.9784946441650391, | |
| "num_tokens": 95098.0, | |
| "step": 126 | |
| }, | |
| { | |
| "entropy": 0.14444835484027863, | |
| "epoch": 5.08, | |
| "grad_norm": 0.6684100031852722, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1344, | |
| "mean_token_accuracy": 0.979784369468689, | |
| "num_tokens": 95844.0, | |
| "step": 127 | |
| }, | |
| { | |
| "entropy": 0.15883971750736237, | |
| "epoch": 5.12, | |
| "grad_norm": 0.6702907085418701, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1492, | |
| "mean_token_accuracy": 0.9770580530166626, | |
| "num_tokens": 96589.0, | |
| "step": 128 | |
| }, | |
| { | |
| "entropy": 0.15137474238872528, | |
| "epoch": 5.16, | |
| "grad_norm": 0.5998567342758179, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1316, | |
| "mean_token_accuracy": 0.979973316192627, | |
| "num_tokens": 97342.0, | |
| "step": 129 | |
| }, | |
| { | |
| "entropy": 0.15595439076423645, | |
| "epoch": 5.2, | |
| "grad_norm": 0.9339006543159485, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1421, | |
| "mean_token_accuracy": 0.9772727489471436, | |
| "num_tokens": 98094.0, | |
| "step": 130 | |
| }, | |
| { | |
| "entropy": 0.1549891084432602, | |
| "epoch": 5.24, | |
| "grad_norm": 0.7966169118881226, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1674, | |
| "mean_token_accuracy": 0.9772422909736633, | |
| "num_tokens": 98845.0, | |
| "step": 131 | |
| }, | |
| { | |
| "entropy": 0.13497532904148102, | |
| "epoch": 5.28, | |
| "grad_norm": 0.5632849335670471, | |
| "learning_rate": 2e-05, | |
| "loss": 0.107, | |
| "mean_token_accuracy": 0.9826666712760925, | |
| "num_tokens": 99599.0, | |
| "step": 132 | |
| }, | |
| { | |
| "entropy": 0.13919445872306824, | |
| "epoch": 5.32, | |
| "grad_norm": 0.8428453803062439, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1443, | |
| "mean_token_accuracy": 0.9773030877113342, | |
| "num_tokens": 100352.0, | |
| "step": 133 | |
| }, | |
| { | |
| "entropy": 0.14983105659484863, | |
| "epoch": 5.36, | |
| "grad_norm": 0.6925723552703857, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1257, | |
| "mean_token_accuracy": 0.9791666865348816, | |
| "num_tokens": 101124.0, | |
| "step": 134 | |
| }, | |
| { | |
| "entropy": 0.16465476155281067, | |
| "epoch": 5.4, | |
| "grad_norm": 0.8338144421577454, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1646, | |
| "mean_token_accuracy": 0.9761589169502258, | |
| "num_tokens": 101883.0, | |
| "step": 135 | |
| }, | |
| { | |
| "entropy": 0.1548919826745987, | |
| "epoch": 5.44, | |
| "grad_norm": 0.8119238615036011, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1452, | |
| "mean_token_accuracy": 0.9748010635375977, | |
| "num_tokens": 102641.0, | |
| "step": 136 | |
| }, | |
| { | |
| "entropy": 0.15561045706272125, | |
| "epoch": 5.48, | |
| "grad_norm": 0.6716637015342712, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1357, | |
| "mean_token_accuracy": 0.9772727489471436, | |
| "num_tokens": 103393.0, | |
| "step": 137 | |
| }, | |
| { | |
| "entropy": 0.15330083668231964, | |
| "epoch": 5.52, | |
| "grad_norm": 0.9031600952148438, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1977, | |
| "mean_token_accuracy": 0.9738219976425171, | |
| "num_tokens": 104161.0, | |
| "step": 138 | |
| }, | |
| { | |
| "entropy": 0.1526884138584137, | |
| "epoch": 5.5600000000000005, | |
| "grad_norm": 0.7704792022705078, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1384, | |
| "mean_token_accuracy": 0.9826435446739197, | |
| "num_tokens": 104914.0, | |
| "step": 139 | |
| }, | |
| { | |
| "entropy": 0.1390034407377243, | |
| "epoch": 5.6, | |
| "grad_norm": 0.6822147965431213, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1431, | |
| "mean_token_accuracy": 0.9803149700164795, | |
| "num_tokens": 105680.0, | |
| "step": 140 | |
| }, | |
| { | |
| "entropy": 0.15023894608020782, | |
| "epoch": 5.64, | |
| "grad_norm": 0.6488987803459167, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1297, | |
| "mean_token_accuracy": 0.9814323782920837, | |
| "num_tokens": 106438.0, | |
| "step": 141 | |
| }, | |
| { | |
| "entropy": 0.1400136947631836, | |
| "epoch": 5.68, | |
| "grad_norm": 0.6782363057136536, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1205, | |
| "mean_token_accuracy": 0.9813581705093384, | |
| "num_tokens": 107193.0, | |
| "step": 142 | |
| }, | |
| { | |
| "entropy": 0.14991530776023865, | |
| "epoch": 5.72, | |
| "grad_norm": 0.8558703660964966, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1535, | |
| "mean_token_accuracy": 0.9769959449768066, | |
| "num_tokens": 107936.0, | |
| "step": 143 | |
| }, | |
| { | |
| "entropy": 0.1502665877342224, | |
| "epoch": 5.76, | |
| "grad_norm": 0.6971323490142822, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1552, | |
| "mean_token_accuracy": 0.9758713245391846, | |
| "num_tokens": 108686.0, | |
| "step": 144 | |
| }, | |
| { | |
| "entropy": 0.15505215525627136, | |
| "epoch": 5.8, | |
| "grad_norm": 0.7806056141853333, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1508, | |
| "mean_token_accuracy": 0.9748344421386719, | |
| "num_tokens": 109445.0, | |
| "step": 145 | |
| }, | |
| { | |
| "entropy": 0.17881980538368225, | |
| "epoch": 5.84, | |
| "grad_norm": 0.6728991270065308, | |
| "learning_rate": 2e-05, | |
| "loss": 0.131, | |
| "mean_token_accuracy": 0.9802371263504028, | |
| "num_tokens": 110208.0, | |
| "step": 146 | |
| }, | |
| { | |
| "entropy": 0.1572950780391693, | |
| "epoch": 5.88, | |
| "grad_norm": 0.7518328428268433, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1563, | |
| "mean_token_accuracy": 0.9749009013175964, | |
| "num_tokens": 110969.0, | |
| "step": 147 | |
| }, | |
| { | |
| "entropy": 0.15898945927619934, | |
| "epoch": 5.92, | |
| "grad_norm": 0.769062876701355, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1417, | |
| "mean_token_accuracy": 0.9772727489471436, | |
| "num_tokens": 111721.0, | |
| "step": 148 | |
| }, | |
| { | |
| "entropy": 0.14981777966022491, | |
| "epoch": 5.96, | |
| "grad_norm": 0.6339032053947449, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1415, | |
| "mean_token_accuracy": 0.9811066389083862, | |
| "num_tokens": 112466.0, | |
| "step": 149 | |
| }, | |
| { | |
| "entropy": 0.14294743537902832, | |
| "epoch": 6.0, | |
| "grad_norm": 0.7767801880836487, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1239, | |
| "mean_token_accuracy": 0.9826666712760925, | |
| "num_tokens": 113220.0, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 72142250830848.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |