| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9925133689839573, | |
| "eval_steps": 5, | |
| "global_step": 58, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.017112299465240642, | |
| "grad_norm": 45.70898063402962, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7809, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.034224598930481284, | |
| "grad_norm": 49.68319333327342, | |
| "learning_rate": 1e-05, | |
| "loss": 0.777, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.051336898395721926, | |
| "grad_norm": 19.034978438793146, | |
| "learning_rate": 9.821428571428573e-06, | |
| "loss": 0.5063, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.06844919786096257, | |
| "grad_norm": 54.48101399041945, | |
| "learning_rate": 9.642857142857144e-06, | |
| "loss": 0.4685, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0855614973262032, | |
| "grad_norm": 51.48329878132252, | |
| "learning_rate": 9.464285714285714e-06, | |
| "loss": 0.609, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0855614973262032, | |
| "eval_accuracy": 0.8134950720242609, | |
| "eval_loss": 0.48897868394851685, | |
| "eval_runtime": 73.6431, | |
| "eval_samples_per_second": 17.911, | |
| "eval_steps_per_second": 0.57, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10267379679144385, | |
| "grad_norm": 20.95950724104017, | |
| "learning_rate": 9.285714285714288e-06, | |
| "loss": 0.4917, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.11978609625668449, | |
| "grad_norm": 13.550687099426307, | |
| "learning_rate": 9.107142857142858e-06, | |
| "loss": 0.3787, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.13689839572192514, | |
| "grad_norm": 18.271449764798163, | |
| "learning_rate": 8.92857142857143e-06, | |
| "loss": 0.4673, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.15401069518716579, | |
| "grad_norm": 12.828023485287906, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.2838, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1711229946524064, | |
| "grad_norm": 12.018530466006792, | |
| "learning_rate": 8.571428571428571e-06, | |
| "loss": 0.3044, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1711229946524064, | |
| "eval_accuracy": 0.9203942380591357, | |
| "eval_loss": 0.262151300907135, | |
| "eval_runtime": 74.023, | |
| "eval_samples_per_second": 17.819, | |
| "eval_steps_per_second": 0.567, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.18823529411764706, | |
| "grad_norm": 8.656791983120135, | |
| "learning_rate": 8.392857142857144e-06, | |
| "loss": 0.2539, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.2053475935828877, | |
| "grad_norm": 7.434664281149775, | |
| "learning_rate": 8.214285714285714e-06, | |
| "loss": 0.2367, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.22245989304812835, | |
| "grad_norm": 12.983467086367982, | |
| "learning_rate": 8.035714285714286e-06, | |
| "loss": 0.2777, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.23957219251336898, | |
| "grad_norm": 11.888272482676422, | |
| "learning_rate": 7.857142857142858e-06, | |
| "loss": 0.1731, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.25668449197860965, | |
| "grad_norm": 30.14821756001351, | |
| "learning_rate": 7.67857142857143e-06, | |
| "loss": 0.3091, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.25668449197860965, | |
| "eval_accuracy": 0.9059893858984079, | |
| "eval_loss": 0.15739406645298004, | |
| "eval_runtime": 74.01, | |
| "eval_samples_per_second": 17.822, | |
| "eval_steps_per_second": 0.567, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2737967914438503, | |
| "grad_norm": 24.20146508345604, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.3639, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2909090909090909, | |
| "grad_norm": 16.217675489045156, | |
| "learning_rate": 7.321428571428572e-06, | |
| "loss": 0.2976, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.30802139037433157, | |
| "grad_norm": 12.764687129205704, | |
| "learning_rate": 7.1428571428571436e-06, | |
| "loss": 0.3156, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3251336898395722, | |
| "grad_norm": 7.584163218805482, | |
| "learning_rate": 6.964285714285714e-06, | |
| "loss": 0.2558, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3422459893048128, | |
| "grad_norm": 7.932168268172894, | |
| "learning_rate": 6.785714285714287e-06, | |
| "loss": 0.2377, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3422459893048128, | |
| "eval_accuracy": 0.909021986353298, | |
| "eval_loss": 0.21613627672195435, | |
| "eval_runtime": 74.0975, | |
| "eval_samples_per_second": 17.801, | |
| "eval_steps_per_second": 0.567, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3593582887700535, | |
| "grad_norm": 10.700701701057481, | |
| "learning_rate": 6.607142857142858e-06, | |
| "loss": 0.233, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3764705882352941, | |
| "grad_norm": 16.405381151491767, | |
| "learning_rate": 6.4285714285714295e-06, | |
| "loss": 0.3352, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.39358288770053473, | |
| "grad_norm": 12.291118297456958, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.216, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.4106951871657754, | |
| "grad_norm": 7.8887750007796225, | |
| "learning_rate": 6.071428571428571e-06, | |
| "loss": 0.1375, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.42780748663101603, | |
| "grad_norm": 13.33170168693044, | |
| "learning_rate": 5.892857142857144e-06, | |
| "loss": 0.2227, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.42780748663101603, | |
| "eval_accuracy": 0.869598180439727, | |
| "eval_loss": 0.28103968501091003, | |
| "eval_runtime": 73.9951, | |
| "eval_samples_per_second": 17.825, | |
| "eval_steps_per_second": 0.568, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4449197860962567, | |
| "grad_norm": 18.059723798006154, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.3415, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.46203208556149733, | |
| "grad_norm": 11.075483666763269, | |
| "learning_rate": 5.535714285714286e-06, | |
| "loss": 0.1966, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.47914438502673795, | |
| "grad_norm": 18.13797822485722, | |
| "learning_rate": 5.357142857142857e-06, | |
| "loss": 0.3054, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.49625668449197863, | |
| "grad_norm": 10.548733335244131, | |
| "learning_rate": 5.1785714285714296e-06, | |
| "loss": 0.2609, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5133689839572193, | |
| "grad_norm": 10.164033276910443, | |
| "learning_rate": 5e-06, | |
| "loss": 0.3034, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5133689839572193, | |
| "eval_accuracy": 0.8832448824867324, | |
| "eval_loss": 0.2795813977718353, | |
| "eval_runtime": 73.9881, | |
| "eval_samples_per_second": 17.827, | |
| "eval_steps_per_second": 0.568, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5304812834224599, | |
| "grad_norm": 9.050163402816409, | |
| "learning_rate": 4.821428571428572e-06, | |
| "loss": 0.2924, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5475935828877005, | |
| "grad_norm": 6.62789824062468, | |
| "learning_rate": 4.642857142857144e-06, | |
| "loss": 0.1794, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5647058823529412, | |
| "grad_norm": 8.72864891416961, | |
| "learning_rate": 4.464285714285715e-06, | |
| "loss": 0.2138, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5818181818181818, | |
| "grad_norm": 10.94537316716946, | |
| "learning_rate": 4.2857142857142855e-06, | |
| "loss": 0.2817, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5989304812834224, | |
| "grad_norm": 9.75004728427678, | |
| "learning_rate": 4.107142857142857e-06, | |
| "loss": 0.2101, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5989304812834224, | |
| "eval_accuracy": 0.9021986353297953, | |
| "eval_loss": 0.20741873979568481, | |
| "eval_runtime": 74.0386, | |
| "eval_samples_per_second": 17.815, | |
| "eval_steps_per_second": 0.567, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6160427807486631, | |
| "grad_norm": 10.440941120772978, | |
| "learning_rate": 3.928571428571429e-06, | |
| "loss": 0.2082, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6331550802139038, | |
| "grad_norm": 18.18026756378137, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.3242, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6502673796791444, | |
| "grad_norm": 10.842904481911043, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 0.1931, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.667379679144385, | |
| "grad_norm": 15.276357700608376, | |
| "learning_rate": 3.3928571428571435e-06, | |
| "loss": 0.3086, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.6844919786096256, | |
| "grad_norm": 12.444233588288037, | |
| "learning_rate": 3.2142857142857147e-06, | |
| "loss": 0.2027, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6844919786096256, | |
| "eval_accuracy": 0.9075056861258529, | |
| "eval_loss": 0.18664336204528809, | |
| "eval_runtime": 74.0115, | |
| "eval_samples_per_second": 17.822, | |
| "eval_steps_per_second": 0.567, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7016042780748664, | |
| "grad_norm": 14.148136437916675, | |
| "learning_rate": 3.0357142857142856e-06, | |
| "loss": 0.3172, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.718716577540107, | |
| "grad_norm": 10.352170205874685, | |
| "learning_rate": 2.8571428571428573e-06, | |
| "loss": 0.2004, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7358288770053476, | |
| "grad_norm": 9.799073608309664, | |
| "learning_rate": 2.6785714285714285e-06, | |
| "loss": 0.2123, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7529411764705882, | |
| "grad_norm": 8.251618796223033, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1639, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7700534759358288, | |
| "grad_norm": 13.010264929817167, | |
| "learning_rate": 2.321428571428572e-06, | |
| "loss": 0.2683, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7700534759358288, | |
| "eval_accuracy": 0.8976497346474602, | |
| "eval_loss": 0.21669183671474457, | |
| "eval_runtime": 74.1431, | |
| "eval_samples_per_second": 17.79, | |
| "eval_steps_per_second": 0.566, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.7871657754010695, | |
| "grad_norm": 10.856308953122479, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 0.1809, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8042780748663102, | |
| "grad_norm": 11.144824639030041, | |
| "learning_rate": 1.9642857142857144e-06, | |
| "loss": 0.1982, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8213903743315508, | |
| "grad_norm": 10.772018186641908, | |
| "learning_rate": 1.7857142857142859e-06, | |
| "loss": 0.2009, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8385026737967914, | |
| "grad_norm": 9.47118254110065, | |
| "learning_rate": 1.6071428571428574e-06, | |
| "loss": 0.1255, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8556149732620321, | |
| "grad_norm": 10.12485855045926, | |
| "learning_rate": 1.4285714285714286e-06, | |
| "loss": 0.1873, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8556149732620321, | |
| "eval_accuracy": 0.8877937831690674, | |
| "eval_loss": 0.2339598387479782, | |
| "eval_runtime": 74.0795, | |
| "eval_samples_per_second": 17.805, | |
| "eval_steps_per_second": 0.567, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8727272727272727, | |
| "grad_norm": 13.541808420276102, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.1415, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.8898395721925134, | |
| "grad_norm": 11.222049226103724, | |
| "learning_rate": 1.0714285714285714e-06, | |
| "loss": 0.0978, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.906951871657754, | |
| "grad_norm": 15.090034415942638, | |
| "learning_rate": 8.928571428571429e-07, | |
| "loss": 0.3028, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.9240641711229947, | |
| "grad_norm": 20.769651439900795, | |
| "learning_rate": 7.142857142857143e-07, | |
| "loss": 0.2485, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 23.4224465210351, | |
| "learning_rate": 5.357142857142857e-07, | |
| "loss": 0.2984, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "eval_accuracy": 0.8824867323730099, | |
| "eval_loss": 0.2450937032699585, | |
| "eval_runtime": 73.6679, | |
| "eval_samples_per_second": 17.905, | |
| "eval_steps_per_second": 0.57, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9582887700534759, | |
| "grad_norm": 17.914238095769544, | |
| "learning_rate": 3.5714285714285716e-07, | |
| "loss": 0.2695, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.9754010695187165, | |
| "grad_norm": 12.54945419387017, | |
| "learning_rate": 1.7857142857142858e-07, | |
| "loss": 0.1866, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.9925133689839573, | |
| "grad_norm": 10.340687525877918, | |
| "learning_rate": 0.0, | |
| "loss": 0.169, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.9925133689839573, | |
| "step": 58, | |
| "total_flos": 0.0, | |
| "train_loss": 0.2849009757668808, | |
| "train_runtime": 5064.7665, | |
| "train_samples_per_second": 1.475, | |
| "train_steps_per_second": 0.011 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 58, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |