| { | |
| "best_metric": 78.83324934908322, | |
| "best_model_checkpoint": "/root/turkic_qa/tr_kaz_models/tr_kaz_xlm_roberta_base_model/checkpoint-7490", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 7490, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "step": 749, | |
| "train_exact_match": 24.275724275724276, | |
| "train_f1": 43.117743763484455, | |
| "train_runtime": 19.0896, | |
| "train_samples_per_second": 79.624, | |
| "train_steps_per_second": 2.881 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 45.69759750366211, | |
| "learning_rate": 5e-06, | |
| "loss": 4.4853, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_exact_match": 25.875, | |
| "eval_f1": 44.604620754612284, | |
| "eval_runtime": 56.7402, | |
| "eval_samples_per_second": 81.389, | |
| "eval_steps_per_second": 2.908, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 1498, | |
| "train_exact_match": 55.744255744255746, | |
| "train_f1": 73.08153342630565, | |
| "train_runtime": 17.3477, | |
| "train_samples_per_second": 85.429, | |
| "train_steps_per_second": 3.055 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 27.808273315429688, | |
| "learning_rate": 1e-05, | |
| "loss": 1.7458, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_exact_match": 52.15625, | |
| "eval_f1": 70.19574840484874, | |
| "eval_runtime": 53.2765, | |
| "eval_samples_per_second": 86.68, | |
| "eval_steps_per_second": 3.097, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 2247, | |
| "train_exact_match": 65.83416583416583, | |
| "train_f1": 81.64944541515409, | |
| "train_runtime": 17.576, | |
| "train_samples_per_second": 85.173, | |
| "train_steps_per_second": 3.072 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 26.646291732788086, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 1.1281, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_exact_match": 58.0625, | |
| "eval_f1": 75.09391733824289, | |
| "eval_runtime": 56.1114, | |
| "eval_samples_per_second": 82.301, | |
| "eval_steps_per_second": 2.941, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 2996, | |
| "train_exact_match": 69.43056943056943, | |
| "train_f1": 84.4692905630636, | |
| "train_runtime": 19.0377, | |
| "train_samples_per_second": 79.106, | |
| "train_steps_per_second": 2.836 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 22.081819534301758, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.8784, | |
| "step": 2996 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_exact_match": 60.25, | |
| "eval_f1": 76.8410082950037, | |
| "eval_runtime": 57.4346, | |
| "eval_samples_per_second": 80.405, | |
| "eval_steps_per_second": 2.873, | |
| "step": 2996 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 3745, | |
| "train_exact_match": 74.12587412587412, | |
| "train_f1": 87.57278612666818, | |
| "train_runtime": 18.9437, | |
| "train_samples_per_second": 78.971, | |
| "train_steps_per_second": 2.851 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 27.527299880981445, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.7365, | |
| "step": 3745 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_exact_match": 60.96875, | |
| "eval_f1": 77.33365402370875, | |
| "eval_runtime": 57.2578, | |
| "eval_samples_per_second": 80.653, | |
| "eval_steps_per_second": 2.882, | |
| "step": 3745 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 4494, | |
| "train_exact_match": 78.42157842157842, | |
| "train_f1": 89.99797070222537, | |
| "train_runtime": 19.0634, | |
| "train_samples_per_second": 78.265, | |
| "train_steps_per_second": 2.833 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 26.452974319458008, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6194, | |
| "step": 4494 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_exact_match": 61.96875, | |
| "eval_f1": 77.91252095789356, | |
| "eval_runtime": 57.9105, | |
| "eval_samples_per_second": 79.744, | |
| "eval_steps_per_second": 2.849, | |
| "step": 4494 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "step": 5243, | |
| "train_exact_match": 82.81718281718281, | |
| "train_f1": 92.72768225182865, | |
| "train_runtime": 18.2446, | |
| "train_samples_per_second": 80.956, | |
| "train_steps_per_second": 2.905 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 18.254165649414062, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.534, | |
| "step": 5243 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_exact_match": 62.21875, | |
| "eval_f1": 78.17772368782505, | |
| "eval_runtime": 55.9458, | |
| "eval_samples_per_second": 82.544, | |
| "eval_steps_per_second": 2.949, | |
| "step": 5243 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 5992, | |
| "train_exact_match": 80.41958041958041, | |
| "train_f1": 91.49491612552417, | |
| "train_runtime": 18.2728, | |
| "train_samples_per_second": 84.059, | |
| "train_steps_per_second": 3.01 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 35.931488037109375, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.4747, | |
| "step": 5992 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_exact_match": 62.53125, | |
| "eval_f1": 78.2110139694925, | |
| "eval_runtime": 56.5716, | |
| "eval_samples_per_second": 81.631, | |
| "eval_steps_per_second": 2.917, | |
| "step": 5992 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "step": 6741, | |
| "train_exact_match": 82.61738261738262, | |
| "train_f1": 92.86226576706642, | |
| "train_runtime": 18.569, | |
| "train_samples_per_second": 79.218, | |
| "train_steps_per_second": 2.854 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 60.296566009521484, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.4304, | |
| "step": 6741 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_exact_match": 63.0625, | |
| "eval_f1": 78.5170445431728, | |
| "eval_runtime": 57.3991, | |
| "eval_samples_per_second": 80.454, | |
| "eval_steps_per_second": 2.875, | |
| "step": 6741 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 7490, | |
| "train_exact_match": 84.91508491508492, | |
| "train_f1": 93.37423644646913, | |
| "train_runtime": 18.377, | |
| "train_samples_per_second": 81.297, | |
| "train_steps_per_second": 2.938 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 23.199573516845703, | |
| "learning_rate": 0.0, | |
| "loss": 0.3964, | |
| "step": 7490 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_exact_match": 63.28125, | |
| "eval_f1": 78.83324934908322, | |
| "eval_runtime": 55.9599, | |
| "eval_samples_per_second": 82.523, | |
| "eval_steps_per_second": 2.949, | |
| "step": 7490 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 7490, | |
| "total_flos": 4.108760851295232e+16, | |
| "train_loss": 1.1428948192316317, | |
| "train_runtime": 4968.485, | |
| "train_samples_per_second": 42.198, | |
| "train_steps_per_second": 1.508 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 7490, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 4.108760851295232e+16, | |
| "train_batch_size": 28, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |