| { | |
| "best_metric": 41.842087703189854, | |
| "best_model_checkpoint": "/root/turkic_qa/en_kaz_models/orig_kaz_roberta_base_model/checkpoint-6020", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 6020, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "step": 602, | |
| "train_exact_match": 4.595404595404595, | |
| "train_f1": 11.363671882650538, | |
| "train_runtime": 9.9586, | |
| "train_samples_per_second": 112.967, | |
| "train_steps_per_second": 4.117 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 7.829487323760986, | |
| "learning_rate": 5e-06, | |
| "loss": 4.7559, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_exact_match": 4.0625, | |
| "eval_f1": 10.65255335870508, | |
| "eval_runtime": 31.5322, | |
| "eval_samples_per_second": 113.281, | |
| "eval_steps_per_second": 4.059, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 1204, | |
| "train_exact_match": 16.083916083916083, | |
| "train_f1": 24.66111862744578, | |
| "train_runtime": 10.0503, | |
| "train_samples_per_second": 113.131, | |
| "train_steps_per_second": 4.079 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 9.797904014587402, | |
| "learning_rate": 1e-05, | |
| "loss": 3.6465, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_exact_match": 12.375, | |
| "eval_f1": 19.998235228782818, | |
| "eval_runtime": 31.6899, | |
| "eval_samples_per_second": 112.717, | |
| "eval_steps_per_second": 4.039, | |
| "step": 1204 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1806, | |
| "train_exact_match": 28.571428571428573, | |
| "train_f1": 39.799779936991705, | |
| "train_runtime": 9.9917, | |
| "train_samples_per_second": 112.694, | |
| "train_steps_per_second": 4.103 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 15.731840133666992, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 3.0904, | |
| "step": 1806 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_exact_match": 20.8125, | |
| "eval_f1": 31.83521011700413, | |
| "eval_runtime": 31.6833, | |
| "eval_samples_per_second": 112.741, | |
| "eval_steps_per_second": 4.04, | |
| "step": 1806 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 2408, | |
| "train_exact_match": 37.76223776223776, | |
| "train_f1": 50.512835216096065, | |
| "train_runtime": 10.2559, | |
| "train_samples_per_second": 113.009, | |
| "train_steps_per_second": 4.095 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 15.223044395446777, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 2.5695, | |
| "step": 2408 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_exact_match": 26.34375, | |
| "eval_f1": 37.68196751270749, | |
| "eval_runtime": 31.6681, | |
| "eval_samples_per_second": 112.795, | |
| "eval_steps_per_second": 4.042, | |
| "step": 2408 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 3010, | |
| "train_exact_match": 46.553446553446555, | |
| "train_f1": 59.64402432927812, | |
| "train_runtime": 9.9179, | |
| "train_samples_per_second": 112.827, | |
| "train_steps_per_second": 4.033 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 19.780818939208984, | |
| "learning_rate": 6.25e-06, | |
| "loss": 2.2508, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_exact_match": 27.5625, | |
| "eval_f1": 39.44221624780652, | |
| "eval_runtime": 31.714, | |
| "eval_samples_per_second": 112.632, | |
| "eval_steps_per_second": 4.036, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 3612, | |
| "train_exact_match": 50.94905094905095, | |
| "train_f1": 63.29384394642245, | |
| "train_runtime": 9.8492, | |
| "train_samples_per_second": 111.278, | |
| "train_steps_per_second": 4.061 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 19.713895797729492, | |
| "learning_rate": 5e-06, | |
| "loss": 2.0297, | |
| "step": 3612 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_exact_match": 28.75, | |
| "eval_f1": 41.0194110809582, | |
| "eval_runtime": 31.7278, | |
| "eval_samples_per_second": 112.583, | |
| "eval_steps_per_second": 4.034, | |
| "step": 3612 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "step": 4214, | |
| "train_exact_match": 53.74625374625375, | |
| "train_f1": 65.89175767102272, | |
| "train_runtime": 9.8541, | |
| "train_samples_per_second": 113.658, | |
| "train_steps_per_second": 4.059 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 18.399381637573242, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 1.8692, | |
| "step": 4214 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_exact_match": 29.0625, | |
| "eval_f1": 41.37431545961761, | |
| "eval_runtime": 31.718, | |
| "eval_samples_per_second": 112.617, | |
| "eval_steps_per_second": 4.036, | |
| "step": 4214 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 4816, | |
| "train_exact_match": 57.04295704295704, | |
| "train_f1": 69.23809998725315, | |
| "train_runtime": 9.8266, | |
| "train_samples_per_second": 113.06, | |
| "train_steps_per_second": 4.071 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 19.948137283325195, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.7415, | |
| "step": 4816 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_exact_match": 29.09375, | |
| "eval_f1": 41.58572921874224, | |
| "eval_runtime": 31.6379, | |
| "eval_samples_per_second": 112.903, | |
| "eval_steps_per_second": 4.046, | |
| "step": 4816 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "step": 5418, | |
| "train_exact_match": 59.14085914085914, | |
| "train_f1": 70.72972336346831, | |
| "train_runtime": 10.1, | |
| "train_samples_per_second": 111.485, | |
| "train_steps_per_second": 4.059 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 18.038774490356445, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.6377, | |
| "step": 5418 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_exact_match": 29.21875, | |
| "eval_f1": 41.68346527288109, | |
| "eval_runtime": 31.8369, | |
| "eval_samples_per_second": 112.197, | |
| "eval_steps_per_second": 4.02, | |
| "step": 5418 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 6020, | |
| "train_exact_match": 59.94005994005994, | |
| "train_f1": 72.84928684458019, | |
| "train_runtime": 9.7454, | |
| "train_samples_per_second": 112.976, | |
| "train_steps_per_second": 4.104 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 21.71363639831543, | |
| "learning_rate": 0.0, | |
| "loss": 1.583, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_exact_match": 29.28125, | |
| "eval_f1": 41.842087703189854, | |
| "eval_runtime": 31.896, | |
| "eval_samples_per_second": 111.989, | |
| "eval_steps_per_second": 4.013, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 6020, | |
| "total_flos": 1.651520510954496e+16, | |
| "train_loss": 2.5174192206804142, | |
| "train_runtime": 2180.484, | |
| "train_samples_per_second": 77.295, | |
| "train_steps_per_second": 2.761 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 6020, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 1.651520510954496e+16, | |
| "train_batch_size": 28, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |