| { | |
| "best_metric": 0.8982431888580322, | |
| "best_model_checkpoint": "data/Llama-31-8B_task-1_60-samples_config-2_full/checkpoint-43", | |
| "epoch": 21.91304347826087, | |
| "eval_steps": 500, | |
| "global_step": 63, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 2.058223247528076, | |
| "learning_rate": 1e-05, | |
| "loss": 2.5057, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 2.0094869136810303, | |
| "learning_rate": 2e-05, | |
| "loss": 2.5391, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "eval_loss": 2.391584873199463, | |
| "eval_runtime": 4.8261, | |
| "eval_samples_per_second": 2.486, | |
| "eval_steps_per_second": 2.486, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 1.6555781364440918, | |
| "learning_rate": 4e-05, | |
| "loss": 2.4703, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "eval_loss": 2.20740008354187, | |
| "eval_runtime": 4.8274, | |
| "eval_samples_per_second": 2.486, | |
| "eval_steps_per_second": 2.486, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 1.380189299583435, | |
| "learning_rate": 6e-05, | |
| "loss": 2.3343, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 1.1684484481811523, | |
| "learning_rate": 8e-05, | |
| "loss": 2.1328, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "eval_loss": 1.9536972045898438, | |
| "eval_runtime": 4.8279, | |
| "eval_samples_per_second": 2.486, | |
| "eval_steps_per_second": 2.486, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 1.248038649559021, | |
| "learning_rate": 0.0001, | |
| "loss": 1.9709, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "eval_loss": 1.6768689155578613, | |
| "eval_runtime": 4.8282, | |
| "eval_samples_per_second": 2.485, | |
| "eval_steps_per_second": 2.485, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.9004043936729431, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.7679, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.9594781994819641, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 1.5704, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 1.3962249755859375, | |
| "eval_runtime": 4.8269, | |
| "eval_samples_per_second": 2.486, | |
| "eval_steps_per_second": 2.486, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 1.0566002130508423, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.3765, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 1.1357998847961426, | |
| "eval_runtime": 4.8302, | |
| "eval_samples_per_second": 2.484, | |
| "eval_steps_per_second": 2.484, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.7590504288673401, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 1.131, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.5426605343818665, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 1.0594, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 1.0274940729141235, | |
| "eval_runtime": 4.825, | |
| "eval_samples_per_second": 2.487, | |
| "eval_steps_per_second": 2.487, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.3678481876850128, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.9969, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.9876609444618225, | |
| "eval_runtime": 4.821, | |
| "eval_samples_per_second": 2.489, | |
| "eval_steps_per_second": 2.489, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.33574286103248596, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.9485, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "eval_loss": 0.9700170159339905, | |
| "eval_runtime": 4.8279, | |
| "eval_samples_per_second": 2.486, | |
| "eval_steps_per_second": 2.486, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.3618321716785431, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.9357, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.3733964264392853, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.8932, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "eval_loss": 0.9503232836723328, | |
| "eval_runtime": 4.8207, | |
| "eval_samples_per_second": 2.489, | |
| "eval_steps_per_second": 2.489, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.34359118342399597, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.8815, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "eval_loss": 0.9331014752388, | |
| "eval_runtime": 4.8274, | |
| "eval_samples_per_second": 2.486, | |
| "eval_steps_per_second": 2.486, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.3720068037509918, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.8892, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.3460403382778168, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.8229, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "eval_loss": 0.9215609431266785, | |
| "eval_runtime": 4.8299, | |
| "eval_samples_per_second": 2.485, | |
| "eval_steps_per_second": 2.485, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.37291157245635986, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.8136, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 0.9111137390136719, | |
| "eval_runtime": 4.8323, | |
| "eval_samples_per_second": 2.483, | |
| "eval_steps_per_second": 2.483, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.44780686497688293, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.8011, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.4574354887008667, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.7507, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 0.9021480679512024, | |
| "eval_runtime": 4.8292, | |
| "eval_samples_per_second": 2.485, | |
| "eval_steps_per_second": 2.485, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 0.4815647602081299, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.7373, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 0.8982431888580322, | |
| "eval_runtime": 4.8279, | |
| "eval_samples_per_second": 2.486, | |
| "eval_steps_per_second": 2.486, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 15.304347826086957, | |
| "grad_norm": 0.4997287690639496, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.7148, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.5747138857841492, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.6959, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.9020419716835022, | |
| "eval_runtime": 4.8185, | |
| "eval_samples_per_second": 2.49, | |
| "eval_steps_per_second": 2.49, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "grad_norm": 0.672699511051178, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.6651, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "eval_loss": 0.9060425758361816, | |
| "eval_runtime": 4.821, | |
| "eval_samples_per_second": 2.489, | |
| "eval_steps_per_second": 2.489, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 0.6403476595878601, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.6589, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 17.73913043478261, | |
| "eval_loss": 0.9118922352790833, | |
| "eval_runtime": 4.8216, | |
| "eval_samples_per_second": 2.489, | |
| "eval_steps_per_second": 2.489, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 18.08695652173913, | |
| "grad_norm": 0.7273726463317871, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.5916, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "grad_norm": 0.7468767166137695, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.5782, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "eval_loss": 0.9264171123504639, | |
| "eval_runtime": 4.824, | |
| "eval_samples_per_second": 2.488, | |
| "eval_steps_per_second": 2.488, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 19.47826086956522, | |
| "grad_norm": 0.7150223851203918, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.585, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 19.82608695652174, | |
| "eval_loss": 0.9372181296348572, | |
| "eval_runtime": 4.8257, | |
| "eval_samples_per_second": 2.487, | |
| "eval_steps_per_second": 2.487, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 20.17391304347826, | |
| "grad_norm": 0.7036177515983582, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.4927, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "grad_norm": 0.6790146231651306, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.511, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "eval_loss": 0.9603924751281738, | |
| "eval_runtime": 4.8248, | |
| "eval_samples_per_second": 2.487, | |
| "eval_steps_per_second": 2.487, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 21.565217391304348, | |
| "grad_norm": 0.5890304446220398, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 0.4767, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 21.91304347826087, | |
| "eval_loss": 0.9806472659111023, | |
| "eval_runtime": 4.8246, | |
| "eval_samples_per_second": 2.487, | |
| "eval_steps_per_second": 2.487, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 21.91304347826087, | |
| "step": 63, | |
| "total_flos": 1.5148012584042496e+16, | |
| "train_loss": 1.081045771402026, | |
| "train_runtime": 1186.9402, | |
| "train_samples_per_second": 1.938, | |
| "train_steps_per_second": 0.084 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "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": 1.5148012584042496e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |