| { | |
| "best_metric": 1.1525942087173462, | |
| "best_model_checkpoint": "data/Llama-31-8B_task-3_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": 0.5371779203414917, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6963, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.5044246315956116, | |
| "learning_rate": 2e-05, | |
| "loss": 1.6601, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "eval_loss": 1.669729232788086, | |
| "eval_runtime": 11.2977, | |
| "eval_samples_per_second": 1.062, | |
| "eval_steps_per_second": 1.062, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.4555090665817261, | |
| "learning_rate": 4e-05, | |
| "loss": 1.6649, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "eval_loss": 1.6259647607803345, | |
| "eval_runtime": 11.2907, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 0.467232882976532, | |
| "learning_rate": 6e-05, | |
| "loss": 1.6544, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.44558843970298767, | |
| "learning_rate": 8e-05, | |
| "loss": 1.591, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "eval_loss": 1.5467677116394043, | |
| "eval_runtime": 11.2921, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.3453204333782196, | |
| "learning_rate": 0.0001, | |
| "loss": 1.4992, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "eval_loss": 1.4664047956466675, | |
| "eval_runtime": 11.2912, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.2386687695980072, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.4747, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.240106001496315, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 1.4061, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 1.396276831626892, | |
| "eval_runtime": 11.2935, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.22406244277954102, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.352, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 1.3312901258468628, | |
| "eval_runtime": 11.2861, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.2621766924858093, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 1.3078, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.2611936032772064, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 1.2367, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 1.2646236419677734, | |
| "eval_runtime": 11.289, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.2356899231672287, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 1.2127, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.2216287851333618, | |
| "eval_runtime": 11.2874, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.176839679479599, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 1.1571, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "eval_loss": 1.2051867246627808, | |
| "eval_runtime": 11.2903, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.17497973144054413, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 1.1346, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.15325738489627838, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 1.1165, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "eval_loss": 1.189988613128662, | |
| "eval_runtime": 11.2861, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.13827162981033325, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 1.124, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "eval_loss": 1.1786916255950928, | |
| "eval_runtime": 11.2912, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.15808305144309998, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 1.0669, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.14512917399406433, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 1.0947, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "eval_loss": 1.1694222688674927, | |
| "eval_runtime": 11.286, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.14316336810588837, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 1.0606, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 1.163387417793274, | |
| "eval_runtime": 11.2897, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.14829201996326447, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 1.0616, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.16047731041908264, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.0621, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 1.1572819948196411, | |
| "eval_runtime": 11.2875, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 0.13741867244243622, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 1.0235, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 1.1549540758132935, | |
| "eval_runtime": 11.2917, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 15.304347826086957, | |
| "grad_norm": 0.14022833108901978, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 1.0307, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.15961027145385742, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.0274, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 1.1531068086624146, | |
| "eval_runtime": 11.2899, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "grad_norm": 0.1557525396347046, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.9827, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "eval_loss": 1.1525942087173462, | |
| "eval_runtime": 11.2879, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 0.17726318538188934, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.9959, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 17.73913043478261, | |
| "eval_loss": 1.153599500656128, | |
| "eval_runtime": 11.2898, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 18.08695652173913, | |
| "grad_norm": 0.17308309674263, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 1.0118, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "grad_norm": 0.16349872946739197, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.9813, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "eval_loss": 1.1575570106506348, | |
| "eval_runtime": 11.2886, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 19.47826086956522, | |
| "grad_norm": 0.18308573961257935, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.9571, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 19.82608695652174, | |
| "eval_loss": 1.160021185874939, | |
| "eval_runtime": 11.2899, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 20.17391304347826, | |
| "grad_norm": 0.1963604986667633, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.964, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "grad_norm": 0.19509056210517883, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.9413, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "eval_loss": 1.1618882417678833, | |
| "eval_runtime": 11.2899, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 21.565217391304348, | |
| "grad_norm": 0.2059006690979004, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 0.9355, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 21.91304347826087, | |
| "eval_loss": 1.1651769876480103, | |
| "eval_runtime": 11.2932, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 22.26086956521739, | |
| "grad_norm": 0.21040895581245422, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.9214, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 22.956521739130434, | |
| "grad_norm": 0.24565249681472778, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 0.9063, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 22.956521739130434, | |
| "eval_loss": 1.1697865724563599, | |
| "eval_runtime": 11.2884, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 23.652173913043477, | |
| "grad_norm": 0.23310421407222748, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 0.8949, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_loss": 1.173593521118164, | |
| "eval_runtime": 11.2871, | |
| "eval_samples_per_second": 1.063, | |
| "eval_steps_per_second": 1.063, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "step": 69, | |
| "total_flos": 8.84265884623831e+16, | |
| "train_loss": 1.1593545882598213, | |
| "train_runtime": 3262.9387, | |
| "train_samples_per_second": 0.705, | |
| "train_steps_per_second": 0.031 | |
| } | |
| ], | |
| "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": 8.84265884623831e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |