| { | |
| "best_metric": 28.3473, | |
| "best_model_checkpoint": "models/mt0-xl_norwegian_natprompt_updated/checkpoint-3978", | |
| "epoch": 10.998491704374057, | |
| "eval_steps": 500, | |
| "global_step": 5469, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4.75e-05, | |
| "loss": 2.6165, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 8.414592760180996, | |
| "eval_loss": 1.9032089710235596, | |
| "eval_rouge1": 26.3511, | |
| "eval_rouge2": 12.3506, | |
| "eval_rougeL": 25.9106, | |
| "eval_rougeLsum": 25.9089, | |
| "eval_runtime": 138.9748, | |
| "eval_samples_per_second": 50.887, | |
| "eval_steps_per_second": 1.59, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 4.5e-05, | |
| "loss": 2.0889, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 9.177601809954751, | |
| "eval_loss": 1.8498029708862305, | |
| "eval_rouge1": 27.424, | |
| "eval_rouge2": 13.5368, | |
| "eval_rougeL": 26.9384, | |
| "eval_rougeLsum": 26.9301, | |
| "eval_runtime": 137.6855, | |
| "eval_samples_per_second": 51.363, | |
| "eval_steps_per_second": 1.605, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 4.25e-05, | |
| "loss": 1.7957, | |
| "step": 1491 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 8.836962669683258, | |
| "eval_loss": 1.8434585332870483, | |
| "eval_rouge1": 27.6735, | |
| "eval_rouge2": 13.985, | |
| "eval_rougeL": 27.1922, | |
| "eval_rougeLsum": 27.2052, | |
| "eval_runtime": 138.4103, | |
| "eval_samples_per_second": 51.094, | |
| "eval_steps_per_second": 1.597, | |
| "step": 1491 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 3.999496981891348e-05, | |
| "loss": 1.5665, | |
| "step": 1989 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 8.857466063348417, | |
| "eval_loss": 1.850651502609253, | |
| "eval_rouge1": 28.2493, | |
| "eval_rouge2": 14.358, | |
| "eval_rougeL": 27.7524, | |
| "eval_rougeLsum": 27.7456, | |
| "eval_runtime": 138.4439, | |
| "eval_samples_per_second": 51.082, | |
| "eval_steps_per_second": 1.596, | |
| "step": 1989 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 3.749496981891348e-05, | |
| "loss": 1.3801, | |
| "step": 2486 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 9.235718325791856, | |
| "eval_loss": 1.889418125152588, | |
| "eval_rouge1": 28.2511, | |
| "eval_rouge2": 14.2431, | |
| "eval_rougeL": 27.6841, | |
| "eval_rougeLsum": 27.6785, | |
| "eval_runtime": 141.1671, | |
| "eval_samples_per_second": 50.097, | |
| "eval_steps_per_second": 1.566, | |
| "step": 2486 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 3.499496981891348e-05, | |
| "loss": 1.2233, | |
| "step": 2983 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_gen_len": 9.103930995475114, | |
| "eval_loss": 1.916949987411499, | |
| "eval_rouge1": 28.3057, | |
| "eval_rouge2": 14.349, | |
| "eval_rougeL": 27.7482, | |
| "eval_rougeLsum": 27.7371, | |
| "eval_runtime": 138.9786, | |
| "eval_samples_per_second": 50.886, | |
| "eval_steps_per_second": 1.59, | |
| "step": 2983 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 3.249496981891348e-05, | |
| "loss": 1.0877, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_gen_len": 9.20475113122172, | |
| "eval_loss": 1.9742506742477417, | |
| "eval_rouge1": 28.2671, | |
| "eval_rouge2": 14.4585, | |
| "eval_rougeL": 27.725, | |
| "eval_rougeLsum": 27.7475, | |
| "eval_runtime": 139.1018, | |
| "eval_samples_per_second": 50.84, | |
| "eval_steps_per_second": 1.589, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 2.9989939637826965e-05, | |
| "loss": 0.9717, | |
| "step": 3978 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_gen_len": 9.785633484162895, | |
| "eval_loss": 2.0358376502990723, | |
| "eval_rouge1": 28.3473, | |
| "eval_rouge2": 14.2734, | |
| "eval_rougeL": 27.7737, | |
| "eval_rougeLsum": 27.7661, | |
| "eval_runtime": 142.8331, | |
| "eval_samples_per_second": 49.512, | |
| "eval_steps_per_second": 1.547, | |
| "step": 3978 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 2.7489939637826962e-05, | |
| "loss": 0.8777, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_gen_len": 9.886453619909503, | |
| "eval_loss": 2.0969080924987793, | |
| "eval_rouge1": 27.7863, | |
| "eval_rouge2": 13.8157, | |
| "eval_rougeL": 27.1987, | |
| "eval_rougeLsum": 27.1859, | |
| "eval_runtime": 141.209, | |
| "eval_samples_per_second": 50.082, | |
| "eval_steps_per_second": 1.565, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 2.4989939637826962e-05, | |
| "loss": 0.7983, | |
| "step": 4972 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_gen_len": 9.238829185520363, | |
| "eval_loss": 2.1536314487457275, | |
| "eval_rouge1": 28.2427, | |
| "eval_rouge2": 14.3725, | |
| "eval_rougeL": 27.6965, | |
| "eval_rougeLsum": 27.7019, | |
| "eval_runtime": 138.9465, | |
| "eval_samples_per_second": 50.897, | |
| "eval_steps_per_second": 1.591, | |
| "step": 4972 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "learning_rate": 2.2489939637826963e-05, | |
| "loss": 0.7261, | |
| "step": 5469 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_gen_len": 9.441742081447964, | |
| "eval_loss": 2.1868813037872314, | |
| "eval_rouge1": 28.0261, | |
| "eval_rouge2": 14.1267, | |
| "eval_rougeL": 27.4232, | |
| "eval_rougeLsum": 27.4261, | |
| "eval_runtime": 140.5761, | |
| "eval_samples_per_second": 50.307, | |
| "eval_steps_per_second": 1.572, | |
| "step": 5469 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "step": 5469, | |
| "total_flos": 2.6645663357023027e+17, | |
| "train_loss": 1.3756333750635685, | |
| "train_runtime": 9869.4584, | |
| "train_samples_per_second": 128.961, | |
| "train_steps_per_second": 1.007 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 9940, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 2.6645663357023027e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |