| { | |
| "best_metric": 1.0276590585708618, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-1_60-samples_config-2_full/checkpoint-37", | |
| "epoch": 19.82608695652174, | |
| "eval_steps": 500, | |
| "global_step": 57, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 1.4650648832321167, | |
| "learning_rate": 1e-05, | |
| "loss": 2.5672, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 1.4680970907211304, | |
| "learning_rate": 2e-05, | |
| "loss": 2.6105, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "eval_loss": 2.4359917640686035, | |
| "eval_runtime": 1.1542, | |
| "eval_samples_per_second": 10.397, | |
| "eval_steps_per_second": 10.397, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 1.4834535121917725, | |
| "learning_rate": 4e-05, | |
| "loss": 2.5655, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "eval_loss": 2.3295297622680664, | |
| "eval_runtime": 1.1399, | |
| "eval_samples_per_second": 10.527, | |
| "eval_steps_per_second": 10.527, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 1.4470669031143188, | |
| "learning_rate": 6e-05, | |
| "loss": 2.5035, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 1.1602197885513306, | |
| "learning_rate": 8e-05, | |
| "loss": 2.2831, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "eval_loss": 1.9787424802780151, | |
| "eval_runtime": 1.1295, | |
| "eval_samples_per_second": 10.624, | |
| "eval_steps_per_second": 10.624, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.9839016795158386, | |
| "learning_rate": 0.0001, | |
| "loss": 2.0137, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "eval_loss": 1.6424676179885864, | |
| "eval_runtime": 1.1078, | |
| "eval_samples_per_second": 10.832, | |
| "eval_steps_per_second": 10.832, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.9552060961723328, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.7427, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.9589630961418152, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 1.5688, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 1.4293409585952759, | |
| "eval_runtime": 1.1224, | |
| "eval_samples_per_second": 10.692, | |
| "eval_steps_per_second": 10.692, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.9522896409034729, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.4129, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 1.2381266355514526, | |
| "eval_runtime": 1.0841, | |
| "eval_samples_per_second": 11.069, | |
| "eval_steps_per_second": 11.069, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.7782798409461975, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 1.2163, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.5571452379226685, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 1.1583, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 1.1528657674789429, | |
| "eval_runtime": 1.3288, | |
| "eval_samples_per_second": 9.031, | |
| "eval_steps_per_second": 9.031, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.4758853018283844, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 1.0951, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.1033568382263184, | |
| "eval_runtime": 1.1326, | |
| "eval_samples_per_second": 10.595, | |
| "eval_steps_per_second": 10.595, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.42092087864875793, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 1.0201, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "eval_loss": 1.0815891027450562, | |
| "eval_runtime": 1.1465, | |
| "eval_samples_per_second": 10.467, | |
| "eval_steps_per_second": 10.467, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.47605592012405396, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 1.0132, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.4391937553882599, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.9479, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "eval_loss": 1.0555545091629028, | |
| "eval_runtime": 1.0994, | |
| "eval_samples_per_second": 10.915, | |
| "eval_steps_per_second": 10.915, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.40881142020225525, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.9142, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "eval_loss": 1.0382660627365112, | |
| "eval_runtime": 1.1453, | |
| "eval_samples_per_second": 10.478, | |
| "eval_steps_per_second": 10.478, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.4644452929496765, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.9162, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.4141174852848053, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.8428, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "eval_loss": 1.0283986330032349, | |
| "eval_runtime": 1.1316, | |
| "eval_samples_per_second": 10.605, | |
| "eval_steps_per_second": 10.605, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.41718557476997375, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.8092, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 1.0276590585708618, | |
| "eval_runtime": 1.1297, | |
| "eval_samples_per_second": 10.623, | |
| "eval_steps_per_second": 10.623, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.5159082412719727, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.8069, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.45884421467781067, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.7383, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 1.0369309186935425, | |
| "eval_runtime": 1.2608, | |
| "eval_samples_per_second": 9.518, | |
| "eval_steps_per_second": 9.518, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 0.483315110206604, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.7168, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 1.0502351522445679, | |
| "eval_runtime": 1.1426, | |
| "eval_samples_per_second": 10.502, | |
| "eval_steps_per_second": 10.502, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 15.304347826086957, | |
| "grad_norm": 0.493073433637619, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.6916, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.5290191769599915, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.6614, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 1.069225788116455, | |
| "eval_runtime": 1.1188, | |
| "eval_samples_per_second": 10.726, | |
| "eval_steps_per_second": 10.726, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "grad_norm": 0.5160197019577026, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.6166, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "eval_loss": 1.0938966274261475, | |
| "eval_runtime": 1.1263, | |
| "eval_samples_per_second": 10.655, | |
| "eval_steps_per_second": 10.655, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 0.549968421459198, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.6045, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 17.73913043478261, | |
| "eval_loss": 1.136326789855957, | |
| "eval_runtime": 1.1359, | |
| "eval_samples_per_second": 10.564, | |
| "eval_steps_per_second": 10.564, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 18.08695652173913, | |
| "grad_norm": 0.5940390229225159, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.5412, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "grad_norm": 0.6177570223808289, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.5177, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "eval_loss": 1.178524374961853, | |
| "eval_runtime": 1.1311, | |
| "eval_samples_per_second": 10.609, | |
| "eval_steps_per_second": 10.609, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 19.47826086956522, | |
| "grad_norm": 0.6144609451293945, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.5067, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 19.82608695652174, | |
| "eval_loss": 1.2188189029693604, | |
| "eval_runtime": 1.1224, | |
| "eval_samples_per_second": 10.692, | |
| "eval_steps_per_second": 10.692, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 19.82608695652174, | |
| "step": 57, | |
| "total_flos": 3568612577640448.0, | |
| "train_loss": 1.1697518522279304, | |
| "train_runtime": 430.7309, | |
| "train_samples_per_second": 5.34, | |
| "train_steps_per_second": 0.232 | |
| } | |
| ], | |
| "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": 3568612577640448.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |