| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.0, | |
| "eval_steps": 500, | |
| "global_step": 12330, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.24330900243309003, | |
| "grad_norm": 14.526427268981934, | |
| "learning_rate": 1.9195458231954583e-05, | |
| "loss": 3.5943, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.48661800486618007, | |
| "grad_norm": 16.526565551757812, | |
| "learning_rate": 1.8384428223844285e-05, | |
| "loss": 3.1431, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7299270072992701, | |
| "grad_norm": 11.656450271606445, | |
| "learning_rate": 1.7573398215733984e-05, | |
| "loss": 2.9201, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9732360097323601, | |
| "grad_norm": 13.380396842956543, | |
| "learning_rate": 1.6762368207623682e-05, | |
| "loss": 2.7711, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_bleu": 12.810223940577195, | |
| "eval_loss": 2.420867681503296, | |
| "eval_rouge": { | |
| "rouge1": 0.30991633445023503, | |
| "rouge2": 0.11181858129315748, | |
| "rougeL": 0.3092627118644087, | |
| "rougeLsum": 0.30911397265634744 | |
| }, | |
| "eval_runtime": 264.2267, | |
| "eval_samples_per_second": 33.494, | |
| "eval_steps_per_second": 2.097, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 1.2165450121654502, | |
| "grad_norm": 13.43653392791748, | |
| "learning_rate": 1.5951338199513384e-05, | |
| "loss": 2.3949, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.4598540145985401, | |
| "grad_norm": 13.36623764038086, | |
| "learning_rate": 1.5140308191403083e-05, | |
| "loss": 2.2468, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.7031630170316303, | |
| "grad_norm": 12.249001502990723, | |
| "learning_rate": 1.4329278183292783e-05, | |
| "loss": 2.2239, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.94647201946472, | |
| "grad_norm": 14.571913719177246, | |
| "learning_rate": 1.3518248175182482e-05, | |
| "loss": 2.1585, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_bleu": 18.428136367860084, | |
| "eval_loss": 2.1254029273986816, | |
| "eval_rouge": { | |
| "rouge1": 0.3476506591337114, | |
| "rouge2": 0.13975840731772934, | |
| "rougeL": 0.34719576719576883, | |
| "rougeLsum": 0.3470716527665696 | |
| }, | |
| "eval_runtime": 264.8536, | |
| "eval_samples_per_second": 33.415, | |
| "eval_steps_per_second": 2.092, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 2.18978102189781, | |
| "grad_norm": 12.905760765075684, | |
| "learning_rate": 1.2707218167072182e-05, | |
| "loss": 1.9148, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.4330900243309004, | |
| "grad_norm": 13.12757682800293, | |
| "learning_rate": 1.1896188158961884e-05, | |
| "loss": 1.8616, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.67639902676399, | |
| "grad_norm": 13.198368072509766, | |
| "learning_rate": 1.1085158150851583e-05, | |
| "loss": 1.8182, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.9197080291970803, | |
| "grad_norm": 11.821575164794922, | |
| "learning_rate": 1.0274128142741283e-05, | |
| "loss": 1.8343, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_bleu": 20.618391393580144, | |
| "eval_loss": 1.9738637208938599, | |
| "eval_rouge": { | |
| "rouge1": 0.3669825372367759, | |
| "rouge2": 0.153052954891938, | |
| "rougeL": 0.3663996747132363, | |
| "rougeLsum": 0.36649283390808984 | |
| }, | |
| "eval_runtime": 277.0353, | |
| "eval_samples_per_second": 31.945, | |
| "eval_steps_per_second": 2.0, | |
| "step": 6165 | |
| }, | |
| { | |
| "epoch": 3.1630170316301705, | |
| "grad_norm": 9.874231338500977, | |
| "learning_rate": 9.463098134630983e-06, | |
| "loss": 1.652, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.40632603406326, | |
| "grad_norm": 13.481595993041992, | |
| "learning_rate": 8.652068126520682e-06, | |
| "loss": 1.5882, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.6496350364963503, | |
| "grad_norm": 12.598729133605957, | |
| "learning_rate": 7.841038118410382e-06, | |
| "loss": 1.5795, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.8929440389294405, | |
| "grad_norm": 10.078117370605469, | |
| "learning_rate": 7.030008110300081e-06, | |
| "loss": 1.6081, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_bleu": 21.371071937079442, | |
| "eval_loss": 1.8978888988494873, | |
| "eval_rouge": { | |
| "rouge1": 0.3791688230162823, | |
| "rouge2": 0.16206698950766746, | |
| "rougeL": 0.37860440156203046, | |
| "rougeLsum": 0.3785680330341368 | |
| }, | |
| "eval_runtime": 270.921, | |
| "eval_samples_per_second": 32.666, | |
| "eval_steps_per_second": 2.045, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 4.13625304136253, | |
| "grad_norm": 13.678194999694824, | |
| "learning_rate": 6.2206001622060015e-06, | |
| "loss": 1.5118, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.37956204379562, | |
| "grad_norm": 11.986939430236816, | |
| "learning_rate": 5.409570154095703e-06, | |
| "loss": 1.445, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.622871046228711, | |
| "grad_norm": 12.918234825134277, | |
| "learning_rate": 4.598540145985402e-06, | |
| "loss": 1.447, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.866180048661801, | |
| "grad_norm": 10.726964950561523, | |
| "learning_rate": 3.7875101378751015e-06, | |
| "loss": 1.406, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_bleu": 22.3831585320865, | |
| "eval_loss": 1.859816074371338, | |
| "eval_rouge": { | |
| "rouge1": 0.3916949600932663, | |
| "rouge2": 0.16823271455474847, | |
| "rougeL": 0.39121746928526785, | |
| "rougeLsum": 0.39116348309568805 | |
| }, | |
| "eval_runtime": 277.4168, | |
| "eval_samples_per_second": 31.901, | |
| "eval_steps_per_second": 1.997, | |
| "step": 10275 | |
| }, | |
| { | |
| "epoch": 5.109489051094891, | |
| "grad_norm": 15.405741691589355, | |
| "learning_rate": 2.979724249797243e-06, | |
| "loss": 1.4047, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 5.35279805352798, | |
| "grad_norm": 10.873234748840332, | |
| "learning_rate": 2.1686942416869423e-06, | |
| "loss": 1.3599, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 5.59610705596107, | |
| "grad_norm": 10.827215194702148, | |
| "learning_rate": 1.3576642335766423e-06, | |
| "loss": 1.33, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.839416058394161, | |
| "grad_norm": 12.577991485595703, | |
| "learning_rate": 5.466342254663423e-07, | |
| "loss": 1.3453, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_bleu": 23.393142474562815, | |
| "eval_loss": 1.8506520986557007, | |
| "eval_rouge": { | |
| "rouge1": 0.3935306250560502, | |
| "rouge2": 0.16916814635458705, | |
| "rougeL": 0.393161689534573, | |
| "rougeLsum": 0.3930333602367517 | |
| }, | |
| "eval_runtime": 268.6729, | |
| "eval_samples_per_second": 32.94, | |
| "eval_steps_per_second": 2.062, | |
| "step": 12330 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 12330, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "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": 1030951237386240.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |