| { | |
| "best_metric": 0.9315621256828308, | |
| "best_model_checkpoint": "data/Mistral-7B_task-1_60-samples_config-2_full/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": 2.7660365104675293, | |
| "learning_rate": 1e-05, | |
| "loss": 2.1412, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 2.761305332183838, | |
| "learning_rate": 2e-05, | |
| "loss": 2.1677, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "eval_loss": 1.9772213697433472, | |
| "eval_runtime": 4.8565, | |
| "eval_samples_per_second": 2.471, | |
| "eval_steps_per_second": 2.471, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 2.450890064239502, | |
| "learning_rate": 4e-05, | |
| "loss": 2.0588, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "eval_loss": 1.7097864151000977, | |
| "eval_runtime": 4.8552, | |
| "eval_samples_per_second": 2.472, | |
| "eval_steps_per_second": 2.472, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 1.6053310632705688, | |
| "learning_rate": 6e-05, | |
| "loss": 1.862, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 1.1415278911590576, | |
| "learning_rate": 8e-05, | |
| "loss": 1.6144, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "eval_loss": 1.4225901365280151, | |
| "eval_runtime": 4.8447, | |
| "eval_samples_per_second": 2.477, | |
| "eval_steps_per_second": 2.477, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 1.6181377172470093, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3861, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 3.8260869565217392, | |
| "eval_loss": 1.1334022283554077, | |
| "eval_runtime": 4.8621, | |
| "eval_samples_per_second": 2.468, | |
| "eval_steps_per_second": 2.468, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 4.173913043478261, | |
| "grad_norm": 1.2450289726257324, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.1643, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "grad_norm": 0.8130146265029907, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 1.0181, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 4.869565217391305, | |
| "eval_loss": 0.9961085915565491, | |
| "eval_runtime": 4.854, | |
| "eval_samples_per_second": 2.472, | |
| "eval_steps_per_second": 2.472, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 5.565217391304348, | |
| "grad_norm": 0.6594259738922119, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.9301, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 5.913043478260869, | |
| "eval_loss": 0.9633287787437439, | |
| "eval_runtime": 4.8357, | |
| "eval_samples_per_second": 2.482, | |
| "eval_steps_per_second": 2.482, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 6.260869565217392, | |
| "grad_norm": 0.6981428861618042, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.8321, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "grad_norm": 0.6706119179725647, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.8066, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 6.956521739130435, | |
| "eval_loss": 0.9347649216651917, | |
| "eval_runtime": 4.8358, | |
| "eval_samples_per_second": 2.482, | |
| "eval_steps_per_second": 2.482, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 7.6521739130434785, | |
| "grad_norm": 0.5864876508712769, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.7473, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.9315621256828308, | |
| "eval_runtime": 4.8328, | |
| "eval_samples_per_second": 2.483, | |
| "eval_steps_per_second": 2.483, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 8.347826086956522, | |
| "grad_norm": 0.634713888168335, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.6909, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.695652173913043, | |
| "eval_loss": 0.9585385322570801, | |
| "eval_runtime": 4.8356, | |
| "eval_samples_per_second": 2.482, | |
| "eval_steps_per_second": 2.482, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 9.043478260869565, | |
| "grad_norm": 0.8111370801925659, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.6233, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "grad_norm": 0.8550280332565308, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.5525, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 9.73913043478261, | |
| "eval_loss": 0.9742663502693176, | |
| "eval_runtime": 4.8332, | |
| "eval_samples_per_second": 2.483, | |
| "eval_steps_per_second": 2.483, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 10.434782608695652, | |
| "grad_norm": 0.790436863899231, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.49, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.782608695652174, | |
| "eval_loss": 1.0670939683914185, | |
| "eval_runtime": 4.834, | |
| "eval_samples_per_second": 2.482, | |
| "eval_steps_per_second": 2.482, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 11.130434782608695, | |
| "grad_norm": 1.2715283632278442, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.4556, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "grad_norm": 0.8485116362571716, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.3604, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 11.826086956521738, | |
| "eval_loss": 1.0999555587768555, | |
| "eval_runtime": 4.8347, | |
| "eval_samples_per_second": 2.482, | |
| "eval_steps_per_second": 2.482, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 12.521739130434783, | |
| "grad_norm": 0.8741176724433899, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.3096, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.869565217391305, | |
| "eval_loss": 1.275596022605896, | |
| "eval_runtime": 4.8346, | |
| "eval_samples_per_second": 2.482, | |
| "eval_steps_per_second": 2.482, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 13.217391304347826, | |
| "grad_norm": 1.3192121982574463, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 0.2648, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "grad_norm": 1.0160993337631226, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.2231, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 13.91304347826087, | |
| "eval_loss": 1.302903175354004, | |
| "eval_runtime": 4.8323, | |
| "eval_samples_per_second": 2.483, | |
| "eval_steps_per_second": 2.483, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 14.608695652173914, | |
| "grad_norm": 0.6346321105957031, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 0.1846, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "eval_loss": 1.4380816221237183, | |
| "eval_runtime": 4.8324, | |
| "eval_samples_per_second": 2.483, | |
| "eval_steps_per_second": 2.483, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 14.956521739130435, | |
| "step": 43, | |
| "total_flos": 1.009626071957504e+16, | |
| "train_loss": 0.8760424204343973, | |
| "train_runtime": 806.7681, | |
| "train_samples_per_second": 2.851, | |
| "train_steps_per_second": 0.124 | |
| } | |
| ], | |
| "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": 1.009626071957504e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |