| { | |
| "best_metric": 0.31749844551086426, | |
| "best_model_checkpoint": "data/Mistral-7B_task-3_60-samples_config-2_auto/checkpoint-23", | |
| "epoch": 14.956521739130435, | |
| "eval_steps": 500, | |
| "global_step": 43, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 17.977815628051758, | |
| "learning_rate": 1e-05, | |
| "loss": 3.6306, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 15.775712013244629, | |
| "learning_rate": 2e-05, | |
| "loss": 3.3566, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "eval_loss": 3.137718439102173, | |
| "eval_runtime": 13.9512, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 13.841619491577148, | |
| "learning_rate": 4e-05, | |
| "loss": 2.9037, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "eval_loss": 1.6298753023147583, | |
| "eval_runtime": 13.9538, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 6.3933210372924805, | |
| "learning_rate": 6e-05, | |
| "loss": 1.9309, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 4.053845405578613, | |
| "learning_rate": 8e-05, | |
| "loss": 1.0538, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "eval_loss": 0.6302464604377747, | |
| "eval_runtime": 13.9507, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 1.7446203231811523, | |
| "learning_rate": 0.0001, | |
| "loss": 0.66, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "eval_loss": 0.4233952760696411, | |
| "eval_runtime": 13.9537, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 0.9669440388679504, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.4086, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.6000994443893433, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 0.3128, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 0.3451710045337677, | |
| "eval_runtime": 13.9509, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.7342979907989502, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.3005, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 0.33032771944999695, | |
| "eval_runtime": 13.9554, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.6136006116867065, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.2036, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.6082817912101746, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.2054, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 0.32321789860725403, | |
| "eval_runtime": 13.9519, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.6066721081733704, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.1212, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.31749844551086426, | |
| "eval_runtime": 13.951, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.51006680727005, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.1419, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "eval_loss": 0.3295919597148895, | |
| "eval_runtime": 13.9514, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.6388974785804749, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.1023, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.5857882499694824, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.061, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "eval_loss": 0.3818434178829193, | |
| "eval_runtime": 13.9531, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.3556278645992279, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.0402, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "eval_loss": 0.43406710028648376, | |
| "eval_runtime": 13.9503, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 0.4472380578517914, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.0339, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.47187352180480957, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.0154, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "eval_loss": 0.47563445568084717, | |
| "eval_runtime": 13.9527, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.27206021547317505, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.0099, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 0.5175532698631287, | |
| "eval_runtime": 13.9478, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 0.22185003757476807, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.0106, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 0.5242364406585693, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.0079, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 0.5291572213172913, | |
| "eval_runtime": 13.9528, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 0.16057375073432922, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.0045, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 0.5367021560668945, | |
| "eval_runtime": 13.9492, | |
| "eval_samples_per_second": 0.86, | |
| "eval_steps_per_second": 0.86, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "step": 43, | |
| "total_flos": 6.864903491026944e+16, | |
| "train_loss": 0.5593301863626166, | |
| "train_runtime": 2558.7388, | |
| "train_samples_per_second": 0.899, | |
| "train_steps_per_second": 0.039 | |
| } | |
| ], | |
| "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": 6.864903491026944e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |