| { | |
| "best_metric": 1.0291489362716675, | |
| "best_model_checkpoint": "data/Llama-31-8B_task-2_60-samples_config-2_full_auto/checkpoint-48", | |
| "epoch": 24.0, | |
| "eval_steps": 500, | |
| "global_step": 69, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.47617512941360474, | |
| "learning_rate": 1e-05, | |
| "loss": 1.5298, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.442937433719635, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5031, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "eval_loss": 1.5181411504745483, | |
| "eval_runtime": 13.3343, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.38801026344299316, | |
| "learning_rate": 4e-05, | |
| "loss": 1.4986, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "eval_loss": 1.4718869924545288, | |
| "eval_runtime": 13.339, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 0.4467742145061493, | |
| "learning_rate": 6e-05, | |
| "loss": 1.4876, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.39712953567504883, | |
| "learning_rate": 8e-05, | |
| "loss": 1.4173, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "eval_loss": 1.3914693593978882, | |
| "eval_runtime": 13.3356, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.29091253876686096, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3451, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "eval_loss": 1.3186070919036865, | |
| "eval_runtime": 13.3314, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.22598174214363098, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.3144, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.2439185380935669, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 1.2588, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 1.2547359466552734, | |
| "eval_runtime": 13.3342, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.21214520931243896, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.2035, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 1.1919747591018677, | |
| "eval_runtime": 13.332, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.22111687064170837, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 1.1523, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.23274613916873932, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 1.1045, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 1.1348336935043335, | |
| "eval_runtime": 13.3318, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.2069660723209381, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 1.074, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.0885711908340454, | |
| "eval_runtime": 13.3353, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.18615713715553284, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 1.0261, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "eval_loss": 1.0742846727371216, | |
| "eval_runtime": 13.3285, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.13092510402202606, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 1.0054, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.12971077859401703, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.9858, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "eval_loss": 1.0615901947021484, | |
| "eval_runtime": 13.3324, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.11712505668401718, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.9944, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "eval_loss": 1.05305016040802, | |
| "eval_runtime": 13.3316, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.13746364414691925, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.9476, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.11521344631910324, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.9706, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "eval_loss": 1.0463008880615234, | |
| "eval_runtime": 13.3346, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.1206565722823143, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.9367, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 1.0399166345596313, | |
| "eval_runtime": 13.3354, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.126475527882576, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.9462, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.1225828230381012, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.9372, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 1.0352905988693237, | |
| "eval_runtime": 13.3325, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 0.12201306223869324, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.9054, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 1.0334912538528442, | |
| "eval_runtime": 13.3321, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 15.304347826086957, | |
| "grad_norm": 0.13440456986427307, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.911, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.12997311353683472, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.9095, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 1.029860019683838, | |
| "eval_runtime": 13.3341, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "grad_norm": 0.14405840635299683, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.8691, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "eval_loss": 1.0291489362716675, | |
| "eval_runtime": 13.3349, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 0.14434926211833954, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.8796, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 17.73913043478261, | |
| "eval_loss": 1.0302618741989136, | |
| "eval_runtime": 13.3338, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 18.08695652173913, | |
| "grad_norm": 0.13976463675498962, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.8911, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "grad_norm": 0.14080481231212616, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.865, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "eval_loss": 1.030932068824768, | |
| "eval_runtime": 13.3346, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 19.47826086956522, | |
| "grad_norm": 0.15428860485553741, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.8473, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 19.82608695652174, | |
| "eval_loss": 1.035332202911377, | |
| "eval_runtime": 13.3375, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 20.17391304347826, | |
| "grad_norm": 0.1614825427532196, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.8544, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "grad_norm": 0.1666206270456314, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.8304, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "eval_loss": 1.0361390113830566, | |
| "eval_runtime": 13.3356, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 21.565217391304348, | |
| "grad_norm": 0.16533945500850677, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 0.8259, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 21.91304347826087, | |
| "eval_loss": 1.0386301279067993, | |
| "eval_runtime": 13.3309, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 22.26086956521739, | |
| "grad_norm": 0.17483672499656677, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.8189, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 22.956521739130434, | |
| "grad_norm": 0.18155451118946075, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 0.8022, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 22.956521739130434, | |
| "eval_loss": 1.0422089099884033, | |
| "eval_runtime": 13.3328, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 23.652173913043477, | |
| "grad_norm": 0.18215659260749817, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 0.7948, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_loss": 1.04664945602417, | |
| "eval_runtime": 13.3335, | |
| "eval_samples_per_second": 0.9, | |
| "eval_steps_per_second": 0.9, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "step": 69, | |
| "total_flos": 1.1107117705252045e+17, | |
| "train_loss": 1.030163572318312, | |
| "train_runtime": 3968.8733, | |
| "train_samples_per_second": 0.58, | |
| "train_steps_per_second": 0.025 | |
| } | |
| ], | |
| "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": 1.1107117705252045e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |