| { | |
| "best_metric": 0.8896188735961914, | |
| "best_model_checkpoint": "data/Gemma-2-2B_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": 0.4337459206581116, | |
| "learning_rate": 1e-05, | |
| "loss": 1.3221, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.39953529834747314, | |
| "learning_rate": 2e-05, | |
| "loss": 1.2776, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "eval_loss": 1.3060680627822876, | |
| "eval_runtime": 5.4102, | |
| "eval_samples_per_second": 2.218, | |
| "eval_steps_per_second": 2.218, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.4224326014518738, | |
| "learning_rate": 4e-05, | |
| "loss": 1.2902, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "eval_loss": 1.2795027494430542, | |
| "eval_runtime": 5.1664, | |
| "eval_samples_per_second": 2.323, | |
| "eval_steps_per_second": 2.323, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 0.41636860370635986, | |
| "learning_rate": 6e-05, | |
| "loss": 1.3024, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.36544084548950195, | |
| "learning_rate": 8e-05, | |
| "loss": 1.2269, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "eval_loss": 1.1874794960021973, | |
| "eval_runtime": 5.1634, | |
| "eval_samples_per_second": 2.324, | |
| "eval_steps_per_second": 2.324, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.2514492869377136, | |
| "learning_rate": 0.0001, | |
| "loss": 1.1326, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "eval_loss": 1.1151398420333862, | |
| "eval_runtime": 5.1632, | |
| "eval_samples_per_second": 2.324, | |
| "eval_steps_per_second": 2.324, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.25162798166275024, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.1026, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.30019140243530273, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 1.0639, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 1.0710904598236084, | |
| "eval_runtime": 5.1643, | |
| "eval_samples_per_second": 2.324, | |
| "eval_steps_per_second": 2.324, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.24031293392181396, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.016, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 1.0247911214828491, | |
| "eval_runtime": 5.1649, | |
| "eval_samples_per_second": 2.323, | |
| "eval_steps_per_second": 2.323, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.23334763944149017, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.9752, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.22000160813331604, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.9412, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 0.9830079078674316, | |
| "eval_runtime": 5.1616, | |
| "eval_samples_per_second": 2.325, | |
| "eval_steps_per_second": 2.325, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.20773270726203918, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.9191, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.9519727230072021, | |
| "eval_runtime": 5.1695, | |
| "eval_samples_per_second": 2.321, | |
| "eval_steps_per_second": 2.321, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.15067195892333984, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.8743, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "eval_loss": 0.93886399269104, | |
| "eval_runtime": 5.1695, | |
| "eval_samples_per_second": 2.321, | |
| "eval_steps_per_second": 2.321, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.15075528621673584, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.8659, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.1496516317129135, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.8391, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "eval_loss": 0.9238042831420898, | |
| "eval_runtime": 5.1639, | |
| "eval_samples_per_second": 2.324, | |
| "eval_steps_per_second": 2.324, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.13056012988090515, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.8374, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "eval_loss": 0.9123730063438416, | |
| "eval_runtime": 5.1666, | |
| "eval_samples_per_second": 2.323, | |
| "eval_steps_per_second": 2.323, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.1578528881072998, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.8019, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.12965959310531616, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.8155, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "eval_loss": 0.9036881327629089, | |
| "eval_runtime": 5.1696, | |
| "eval_samples_per_second": 2.321, | |
| "eval_steps_per_second": 2.321, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.14883744716644287, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.782, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 0.8979141116142273, | |
| "eval_runtime": 5.1697, | |
| "eval_samples_per_second": 2.321, | |
| "eval_steps_per_second": 2.321, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.1364155113697052, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.7896, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.1308871954679489, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.7796, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 0.893744170665741, | |
| "eval_runtime": 5.1652, | |
| "eval_samples_per_second": 2.323, | |
| "eval_steps_per_second": 2.323, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 0.12330856174230576, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.7521, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 0.8908995985984802, | |
| "eval_runtime": 5.1627, | |
| "eval_samples_per_second": 2.324, | |
| "eval_steps_per_second": 2.324, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 15.304347826086957, | |
| "grad_norm": 0.1174108162522316, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.7545, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.12904687225818634, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.752, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.8899175524711609, | |
| "eval_runtime": 5.1628, | |
| "eval_samples_per_second": 2.324, | |
| "eval_steps_per_second": 2.324, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "grad_norm": 0.13336613774299622, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 0.715, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 16.695652173913043, | |
| "eval_loss": 0.8896188735961914, | |
| "eval_runtime": 5.169, | |
| "eval_samples_per_second": 2.322, | |
| "eval_steps_per_second": 2.322, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 17.391304347826086, | |
| "grad_norm": 0.1387159675359726, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.726, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 17.73913043478261, | |
| "eval_loss": 0.8901117444038391, | |
| "eval_runtime": 5.1708, | |
| "eval_samples_per_second": 2.321, | |
| "eval_steps_per_second": 2.321, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 18.08695652173913, | |
| "grad_norm": 0.1306636780500412, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 0.7328, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "grad_norm": 0.13166216015815735, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 0.7131, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 18.782608695652176, | |
| "eval_loss": 0.8918083310127258, | |
| "eval_runtime": 5.1638, | |
| "eval_samples_per_second": 2.324, | |
| "eval_steps_per_second": 2.324, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 19.47826086956522, | |
| "grad_norm": 0.1477658748626709, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 0.6912, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 19.82608695652174, | |
| "eval_loss": 0.8944613337516785, | |
| "eval_runtime": 5.1696, | |
| "eval_samples_per_second": 2.321, | |
| "eval_steps_per_second": 2.321, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 20.17391304347826, | |
| "grad_norm": 0.1552548110485077, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.6982, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "grad_norm": 0.15443484485149384, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.6795, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 20.869565217391305, | |
| "eval_loss": 0.8985885977745056, | |
| "eval_runtime": 5.1653, | |
| "eval_samples_per_second": 2.323, | |
| "eval_steps_per_second": 2.323, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 21.565217391304348, | |
| "grad_norm": 0.15619756281375885, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 0.6741, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 21.91304347826087, | |
| "eval_loss": 0.9028210639953613, | |
| "eval_runtime": 5.1661, | |
| "eval_samples_per_second": 2.323, | |
| "eval_steps_per_second": 2.323, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 22.26086956521739, | |
| "grad_norm": 0.17049115896224976, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 0.6656, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 22.956521739130434, | |
| "grad_norm": 0.18031303584575653, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 0.654, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 22.956521739130434, | |
| "eval_loss": 0.9070366024971008, | |
| "eval_runtime": 5.1609, | |
| "eval_samples_per_second": 2.325, | |
| "eval_steps_per_second": 2.325, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 23.652173913043477, | |
| "grad_norm": 0.17119044065475464, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 0.6469, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_loss": 0.9120666980743408, | |
| "eval_runtime": 5.1598, | |
| "eval_samples_per_second": 2.326, | |
| "eval_steps_per_second": 2.326, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "step": 69, | |
| "total_flos": 3.77363846135808e+16, | |
| "train_loss": 0.865192905716274, | |
| "train_runtime": 1637.3522, | |
| "train_samples_per_second": 1.405, | |
| "train_steps_per_second": 0.061 | |
| } | |
| ], | |
| "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": 3.77363846135808e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |