| { | |
| "best_metric": 3.1414525508880615, | |
| "best_model_checkpoint": "qa_finetuning/run-0/checkpoint-1500", | |
| "epoch": 5.0, | |
| "eval_steps": 100, | |
| "global_step": 1565, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.3194888178913738, | |
| "grad_norm": 4.872162818908691, | |
| "learning_rate": 2.2236495574546495e-06, | |
| "loss": 5.793, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3194888178913738, | |
| "eval_loss": 5.499853610992432, | |
| "eval_runtime": 3.4844, | |
| "eval_samples_per_second": 143.499, | |
| "eval_steps_per_second": 4.592, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6389776357827476, | |
| "grad_norm": 5.640169620513916, | |
| "learning_rate": 2.0718646047273696e-06, | |
| "loss": 5.1611, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6389776357827476, | |
| "eval_loss": 4.561793327331543, | |
| "eval_runtime": 3.7947, | |
| "eval_samples_per_second": 131.764, | |
| "eval_steps_per_second": 4.216, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "grad_norm": 6.388542175292969, | |
| "learning_rate": 1.9200796520000897e-06, | |
| "loss": 4.4606, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9584664536741214, | |
| "eval_loss": 4.023355960845947, | |
| "eval_runtime": 3.583, | |
| "eval_samples_per_second": 139.546, | |
| "eval_steps_per_second": 4.465, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.2779552715654952, | |
| "grad_norm": 7.301818370819092, | |
| "learning_rate": 1.76829469927281e-06, | |
| "loss": 4.0751, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.2779552715654952, | |
| "eval_loss": 3.76656436920166, | |
| "eval_runtime": 3.7005, | |
| "eval_samples_per_second": 135.118, | |
| "eval_steps_per_second": 4.324, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.5974440894568689, | |
| "grad_norm": 9.462298393249512, | |
| "learning_rate": 1.61650974654553e-06, | |
| "loss": 3.8854, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.5974440894568689, | |
| "eval_loss": 3.5879857540130615, | |
| "eval_runtime": 3.6179, | |
| "eval_samples_per_second": 138.201, | |
| "eval_steps_per_second": 4.422, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "grad_norm": 10.335731506347656, | |
| "learning_rate": 1.4647247938182503e-06, | |
| "loss": 3.7085, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9169329073482428, | |
| "eval_loss": 3.472351312637329, | |
| "eval_runtime": 3.6946, | |
| "eval_samples_per_second": 135.333, | |
| "eval_steps_per_second": 4.331, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.236421725239617, | |
| "grad_norm": 14.032896995544434, | |
| "learning_rate": 1.3129398410909704e-06, | |
| "loss": 3.5566, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.236421725239617, | |
| "eval_loss": 3.368136405944824, | |
| "eval_runtime": 3.6489, | |
| "eval_samples_per_second": 137.027, | |
| "eval_steps_per_second": 4.385, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.5559105431309903, | |
| "grad_norm": 17.996917724609375, | |
| "learning_rate": 1.1611548883636908e-06, | |
| "loss": 3.4447, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.5559105431309903, | |
| "eval_loss": 3.3015410900115967, | |
| "eval_runtime": 3.6668, | |
| "eval_samples_per_second": 136.358, | |
| "eval_steps_per_second": 4.363, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.8753993610223643, | |
| "grad_norm": 15.922423362731934, | |
| "learning_rate": 1.009369935636411e-06, | |
| "loss": 3.4227, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.8753993610223643, | |
| "eval_loss": 3.2642431259155273, | |
| "eval_runtime": 3.6623, | |
| "eval_samples_per_second": 136.525, | |
| "eval_steps_per_second": 4.369, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.194888178913738, | |
| "grad_norm": 19.34994125366211, | |
| "learning_rate": 8.57584982909131e-07, | |
| "loss": 3.3452, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.194888178913738, | |
| "eval_loss": 3.2255563735961914, | |
| "eval_runtime": 3.6578, | |
| "eval_samples_per_second": 136.694, | |
| "eval_steps_per_second": 4.374, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.5143769968051117, | |
| "grad_norm": 15.923648834228516, | |
| "learning_rate": 7.058000301818512e-07, | |
| "loss": 3.3458, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.5143769968051117, | |
| "eval_loss": 3.200777292251587, | |
| "eval_runtime": 3.6881, | |
| "eval_samples_per_second": 135.57, | |
| "eval_steps_per_second": 4.338, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.8338658146964857, | |
| "grad_norm": 13.226187705993652, | |
| "learning_rate": 5.540150774545714e-07, | |
| "loss": 3.2179, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.8338658146964857, | |
| "eval_loss": 3.169455051422119, | |
| "eval_runtime": 3.9818, | |
| "eval_samples_per_second": 125.572, | |
| "eval_steps_per_second": 4.018, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.15335463258786, | |
| "grad_norm": 23.58437728881836, | |
| "learning_rate": 4.0223012472729155e-07, | |
| "loss": 3.2633, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.15335463258786, | |
| "eval_loss": 3.1567466259002686, | |
| "eval_runtime": 3.668, | |
| "eval_samples_per_second": 136.314, | |
| "eval_steps_per_second": 4.362, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.472843450479234, | |
| "grad_norm": 15.142173767089844, | |
| "learning_rate": 2.5044517200001173e-07, | |
| "loss": 3.2125, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.472843450479234, | |
| "eval_loss": 3.151430130004883, | |
| "eval_runtime": 3.6873, | |
| "eval_samples_per_second": 135.599, | |
| "eval_steps_per_second": 4.339, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.792332268370607, | |
| "grad_norm": 22.637807846069336, | |
| "learning_rate": 9.866021927273189e-08, | |
| "loss": 3.1618, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.792332268370607, | |
| "eval_loss": 3.1414525508880615, | |
| "eval_runtime": 3.6675, | |
| "eval_samples_per_second": 136.332, | |
| "eval_steps_per_second": 4.363, | |
| "step": 1500 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 1565, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2348620126175232.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 2.3754345101819293e-06, | |
| "num_train_epochs": 5, | |
| "per_device_train_batch_size": 8, | |
| "seed": 31 | |
| } | |
| } | |