| { |
| "best_global_step": 200, |
| "best_metric": 0.9968940615653992, |
| "best_model_checkpoint": "results_L0_12/R64_qwen_8b_structured/checkpoint-200", |
| "epoch": 15.63768115942029, |
| "eval_steps": 20, |
| "global_step": 280, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.5797101449275363, |
| "grad_norm": 1.9165797233581543, |
| "learning_rate": 1.6666666666666667e-05, |
| "loss": 3.2135, |
| "step": 10 |
| }, |
| { |
| "epoch": 1.1159420289855073, |
| "grad_norm": 0.5033081769943237, |
| "learning_rate": 3.518518518518519e-05, |
| "loss": 2.5406, |
| "step": 20 |
| }, |
| { |
| "epoch": 1.1159420289855073, |
| "eval_loss": 2.1649956703186035, |
| "eval_runtime": 28.0718, |
| "eval_samples_per_second": 19.735, |
| "eval_steps_per_second": 2.494, |
| "step": 20 |
| }, |
| { |
| "epoch": 1.6956521739130435, |
| "grad_norm": 0.7941814064979553, |
| "learning_rate": 5.370370370370371e-05, |
| "loss": 1.872, |
| "step": 30 |
| }, |
| { |
| "epoch": 2.2318840579710146, |
| "grad_norm": 0.5307704210281372, |
| "learning_rate": 7.222222222222222e-05, |
| "loss": 1.6006, |
| "step": 40 |
| }, |
| { |
| "epoch": 2.2318840579710146, |
| "eval_loss": 1.5740340948104858, |
| "eval_runtime": 58.9447, |
| "eval_samples_per_second": 9.399, |
| "eval_steps_per_second": 1.188, |
| "step": 40 |
| }, |
| { |
| "epoch": 2.8115942028985508, |
| "grad_norm": 0.18376459181308746, |
| "learning_rate": 9.074074074074075e-05, |
| "loss": 1.6336, |
| "step": 50 |
| }, |
| { |
| "epoch": 3.3478260869565215, |
| "grad_norm": 0.24469375610351562, |
| "learning_rate": 9.999797656920765e-05, |
| "loss": 1.4624, |
| "step": 60 |
| }, |
| { |
| "epoch": 3.3478260869565215, |
| "eval_loss": 1.4656774997711182, |
| "eval_runtime": 59.8998, |
| "eval_samples_per_second": 9.249, |
| "eval_steps_per_second": 1.169, |
| "step": 60 |
| }, |
| { |
| "epoch": 3.927536231884058, |
| "grad_norm": 0.29990342259407043, |
| "learning_rate": 9.998179010548088e-05, |
| "loss": 1.3827, |
| "step": 70 |
| }, |
| { |
| "epoch": 4.463768115942029, |
| "grad_norm": 0.3905975818634033, |
| "learning_rate": 9.994942241827156e-05, |
| "loss": 1.3227, |
| "step": 80 |
| }, |
| { |
| "epoch": 4.463768115942029, |
| "eval_loss": 1.403430700302124, |
| "eval_runtime": 57.7438, |
| "eval_samples_per_second": 9.594, |
| "eval_steps_per_second": 1.212, |
| "step": 80 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 0.8633396029472351, |
| "learning_rate": 9.990088398637167e-05, |
| "loss": 1.311, |
| "step": 90 |
| }, |
| { |
| "epoch": 5.579710144927536, |
| "grad_norm": 0.29788145422935486, |
| "learning_rate": 9.983619052372848e-05, |
| "loss": 1.3545, |
| "step": 100 |
| }, |
| { |
| "epoch": 5.579710144927536, |
| "eval_loss": 1.367808222770691, |
| "eval_runtime": 58.6669, |
| "eval_samples_per_second": 9.443, |
| "eval_steps_per_second": 1.193, |
| "step": 100 |
| }, |
| { |
| "epoch": 6.115942028985507, |
| "grad_norm": 158.66375732421875, |
| "learning_rate": 9.975536297435726e-05, |
| "loss": 1.1799, |
| "step": 110 |
| }, |
| { |
| "epoch": 6.695652173913043, |
| "grad_norm": 0.5346195101737976, |
| "learning_rate": 9.965842750556091e-05, |
| "loss": 1.163, |
| "step": 120 |
| }, |
| { |
| "epoch": 6.695652173913043, |
| "eval_loss": 1.3441675901412964, |
| "eval_runtime": 60.3552, |
| "eval_samples_per_second": 9.179, |
| "eval_steps_per_second": 1.16, |
| "step": 120 |
| }, |
| { |
| "epoch": 7.231884057971015, |
| "grad_norm": 0.4752218425273895, |
| "learning_rate": 9.95454154994584e-05, |
| "loss": 1.152, |
| "step": 130 |
| }, |
| { |
| "epoch": 7.811594202898551, |
| "grad_norm": 0.4084065556526184, |
| "learning_rate": 9.941636354282511e-05, |
| "loss": 1.209, |
| "step": 140 |
| }, |
| { |
| "epoch": 7.811594202898551, |
| "eval_loss": 1.3283058404922485, |
| "eval_runtime": 60.0993, |
| "eval_samples_per_second": 9.218, |
| "eval_steps_per_second": 1.165, |
| "step": 140 |
| }, |
| { |
| "epoch": 8.347826086956522, |
| "grad_norm": 0.49419981241226196, |
| "learning_rate": 9.927131341524809e-05, |
| "loss": 1.0819, |
| "step": 150 |
| }, |
| { |
| "epoch": 8.927536231884059, |
| "grad_norm": 0.5527717471122742, |
| "learning_rate": 9.911031207560027e-05, |
| "loss": 1.0363, |
| "step": 160 |
| }, |
| { |
| "epoch": 8.927536231884059, |
| "eval_loss": 1.3205050230026245, |
| "eval_runtime": 53.6751, |
| "eval_samples_per_second": 10.321, |
| "eval_steps_per_second": 1.304, |
| "step": 160 |
| }, |
| { |
| "epoch": 9.46376811594203, |
| "grad_norm": 0.5900382399559021, |
| "learning_rate": 9.893341164683788e-05, |
| "loss": 1.0404, |
| "step": 170 |
| }, |
| { |
| "epoch": 10.0, |
| "grad_norm": 1.0412408113479614, |
| "learning_rate": 9.874066939912603e-05, |
| "loss": 0.9826, |
| "step": 180 |
| }, |
| { |
| "epoch": 10.0, |
| "eval_loss": 1.312466025352478, |
| "eval_runtime": 60.1194, |
| "eval_samples_per_second": 9.215, |
| "eval_steps_per_second": 1.164, |
| "step": 180 |
| }, |
| { |
| "epoch": 10.579710144927537, |
| "grad_norm": 0.5220209956169128, |
| "learning_rate": 9.853214773129796e-05, |
| "loss": 1.1468, |
| "step": 190 |
| }, |
| { |
| "epoch": 11.115942028985508, |
| "grad_norm": 0.6074064373970032, |
| "learning_rate": 9.830791415065382e-05, |
| "loss": 1.0142, |
| "step": 200 |
| }, |
| { |
| "epoch": 11.115942028985508, |
| "eval_loss": 0.9968940615653992, |
| "eval_runtime": 65.3018, |
| "eval_samples_per_second": 8.484, |
| "eval_steps_per_second": 1.072, |
| "step": 200 |
| }, |
| { |
| "epoch": 11.695652173913043, |
| "grad_norm": 0.6192728877067566, |
| "learning_rate": 9.806804125110579e-05, |
| "loss": 0.9176, |
| "step": 210 |
| }, |
| { |
| "epoch": 12.289855072463768, |
| "grad_norm": 1.178308129310608, |
| "learning_rate": 9.781260668967628e-05, |
| "loss": 1.0747, |
| "step": 220 |
| }, |
| { |
| "epoch": 12.289855072463768, |
| "eval_loss": 1.0378575325012207, |
| "eval_runtime": 33.6087, |
| "eval_samples_per_second": 16.484, |
| "eval_steps_per_second": 2.083, |
| "step": 220 |
| }, |
| { |
| "epoch": 12.869565217391305, |
| "grad_norm": 0.5823021531105042, |
| "learning_rate": 9.754169316135712e-05, |
| "loss": 0.9349, |
| "step": 230 |
| }, |
| { |
| "epoch": 13.405797101449275, |
| "grad_norm": 1.0113617181777954, |
| "learning_rate": 9.725538837233763e-05, |
| "loss": 0.8412, |
| "step": 240 |
| }, |
| { |
| "epoch": 13.405797101449275, |
| "eval_loss": 1.0512763261795044, |
| "eval_runtime": 64.8372, |
| "eval_samples_per_second": 8.544, |
| "eval_steps_per_second": 1.08, |
| "step": 240 |
| }, |
| { |
| "epoch": 13.985507246376812, |
| "grad_norm": 0.4507587254047394, |
| "learning_rate": 9.695378501161045e-05, |
| "loss": 0.8315, |
| "step": 250 |
| }, |
| { |
| "epoch": 14.521739130434783, |
| "grad_norm": 0.5109806060791016, |
| "learning_rate": 9.663698072096418e-05, |
| "loss": 0.785, |
| "step": 260 |
| }, |
| { |
| "epoch": 14.521739130434783, |
| "eval_loss": 1.081993579864502, |
| "eval_runtime": 64.9546, |
| "eval_samples_per_second": 8.529, |
| "eval_steps_per_second": 1.078, |
| "step": 260 |
| }, |
| { |
| "epoch": 15.057971014492754, |
| "grad_norm": 342.426025390625, |
| "learning_rate": 9.630507806337259e-05, |
| "loss": 0.7538, |
| "step": 270 |
| }, |
| { |
| "epoch": 15.63768115942029, |
| "grad_norm": 0.9247072339057922, |
| "learning_rate": 9.595818448979061e-05, |
| "loss": 0.7405, |
| "step": 280 |
| }, |
| { |
| "epoch": 15.63768115942029, |
| "eval_loss": 1.1149591207504272, |
| "eval_runtime": 65.9467, |
| "eval_samples_per_second": 8.401, |
| "eval_steps_per_second": 1.061, |
| "step": 280 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 1800, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 100, |
| "save_steps": 20, |
| "stateful_callbacks": { |
| "EarlyStoppingCallback": { |
| "args": { |
| "early_stopping_patience": 4, |
| "early_stopping_threshold": 0.01 |
| }, |
| "attributes": { |
| "early_stopping_patience_counter": 4 |
| } |
| }, |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 8.238044479018107e+17, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
