| { | |
| "best_metric": 1.0627769231796265, | |
| "best_model_checkpoint": "data/Llama-31-8B_task-2_60-samples_config-2_full/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": 0.5254957675933838, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6001, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.4830361604690552, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5658, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "eval_loss": 1.5798794031143188, | |
| "eval_runtime": 12.9798, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.44428548216819763, | |
| "learning_rate": 4e-05, | |
| "loss": 1.5614, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "eval_loss": 1.5292840003967285, | |
| "eval_runtime": 12.9814, | |
| "eval_samples_per_second": 0.924, | |
| "eval_steps_per_second": 0.924, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 0.4937964379787445, | |
| "learning_rate": 6e-05, | |
| "loss": 1.5466, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.44759026169776917, | |
| "learning_rate": 8e-05, | |
| "loss": 1.4751, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "eval_loss": 1.4357937574386597, | |
| "eval_runtime": 12.9777, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.3030599355697632, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3878, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "eval_loss": 1.3590370416641235, | |
| "eval_runtime": 12.98, | |
| "eval_samples_per_second": 0.924, | |
| "eval_steps_per_second": 0.924, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.23369178175926208, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.3565, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.2562275528907776, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 1.3039, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 1.2951425313949585, | |
| "eval_runtime": 12.978, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.22682145237922668, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.2412, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 1.2336902618408203, | |
| "eval_runtime": 12.9799, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.246551513671875, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 1.1949, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.2604689598083496, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 1.1483, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 1.1723712682724, | |
| "eval_runtime": 12.978, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.21387824416160583, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 1.1126, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.1254767179489136, | |
| "eval_runtime": 12.9807, | |
| "eval_samples_per_second": 0.924, | |
| "eval_steps_per_second": 0.924, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.186062291264534, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 1.0651, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "eval_loss": 1.111045002937317, | |
| "eval_runtime": 12.9788, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.1267556995153427, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 1.0413, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.1265019178390503, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 1.0248, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "eval_loss": 1.0985106229782104, | |
| "eval_runtime": 12.9841, | |
| "eval_samples_per_second": 0.924, | |
| "eval_steps_per_second": 0.924, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.11169075965881348, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 1.0323, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "eval_loss": 1.088997721672058, | |
| "eval_runtime": 12.9831, | |
| "eval_samples_per_second": 0.924, | |
| "eval_steps_per_second": 0.924, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.12030880898237228, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.985, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.11247541010379791, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 1.0119, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "eval_loss": 1.082088828086853, | |
| "eval_runtime": 12.9813, | |
| "eval_samples_per_second": 0.924, | |
| "eval_steps_per_second": 0.924, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.12034822255373001, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.9765, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 1.0757737159729004, | |
| "eval_runtime": 12.9798, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.12695056200027466, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.9859, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.11532020568847656, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.9796, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 1.070386528968811, | |
| "eval_runtime": 12.9765, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 0.12851540744304657, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.9428, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 1.0683201551437378, | |
| "eval_runtime": 12.9769, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 15.304347826086957, | |
| "grad_norm": 0.16971461474895477, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.9548, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.12614934146404266, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.9511, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 1.0636566877365112, | |
| "eval_runtime": 12.9769, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "grad_norm": 0.13132987916469574, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.9117, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "eval_loss": 1.0627769231796265, | |
| "eval_runtime": 12.9757, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 0.1522136926651001, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.9208, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 17.73913043478261, | |
| "eval_loss": 1.0628635883331299, | |
| "eval_runtime": 12.9771, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 18.08695652173913, | |
| "grad_norm": 0.13250243663787842, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.9344, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "grad_norm": 0.13446113467216492, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.9108, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "eval_loss": 1.063383936882019, | |
| "eval_runtime": 12.9773, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 19.47826086956522, | |
| "grad_norm": 0.14444531500339508, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.8903, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 19.82608695652174, | |
| "eval_loss": 1.0640206336975098, | |
| "eval_runtime": 12.9768, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 20.17391304347826, | |
| "grad_norm": 0.1512373685836792, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.8951, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "grad_norm": 0.1495932638645172, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.8762, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "eval_loss": 1.0663875341415405, | |
| "eval_runtime": 12.9747, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 21.565217391304348, | |
| "grad_norm": 0.16104595363140106, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 0.8702, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 21.91304347826087, | |
| "eval_loss": 1.068516492843628, | |
| "eval_runtime": 12.9798, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 22.26086956521739, | |
| "grad_norm": 0.17329472303390503, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.8638, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 22.956521739130434, | |
| "grad_norm": 0.17929908633232117, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 0.8465, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 22.956521739130434, | |
| "eval_loss": 1.0700325965881348, | |
| "eval_runtime": 12.9807, | |
| "eval_samples_per_second": 0.924, | |
| "eval_steps_per_second": 0.924, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 23.652173913043477, | |
| "grad_norm": 0.17712633311748505, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 0.8409, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_loss": 1.0734320878982544, | |
| "eval_runtime": 12.9764, | |
| "eval_samples_per_second": 0.925, | |
| "eval_steps_per_second": 0.925, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "step": 69, | |
| "total_flos": 1.066980988640297e+17, | |
| "train_loss": 1.074193215024644, | |
| "train_runtime": 3840.2962, | |
| "train_samples_per_second": 0.599, | |
| "train_steps_per_second": 0.026 | |
| } | |
| ], | |
| "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": 1.066980988640297e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |