CS329X_Avi / checkpoint-100 /trainer_state.json
Adriatogi's picture
Upload folder using huggingface_hub
446f8f7 verified
{
"best_global_step": 100,
"best_metric": 2.1311750411987305,
"best_model_checkpoint": "models/llama-3.1-8b-avi-lora/checkpoint-100",
"epoch": 2.2758620689655173,
"eval_steps": 100,
"global_step": 100,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"entropy": 2.5178302228450775,
"epoch": 0.22988505747126436,
"grad_norm": 0.5600284934043884,
"learning_rate": 0.000199247953459871,
"loss": 3.3017,
"mean_token_accuracy": 0.46053045317530633,
"num_tokens": 138946.0,
"step": 10
},
{
"entropy": 2.5103304237127304,
"epoch": 0.45977011494252873,
"grad_norm": 0.25141987204551697,
"learning_rate": 0.0001932992798834739,
"loss": 2.553,
"mean_token_accuracy": 0.5105453297495842,
"num_tokens": 277162.0,
"step": 20
},
{
"entropy": 2.4413388788700106,
"epoch": 0.6896551724137931,
"grad_norm": 0.15782828629016876,
"learning_rate": 0.00018175848131515837,
"loss": 2.3964,
"mean_token_accuracy": 0.5284669108688831,
"num_tokens": 412770.0,
"step": 30
},
{
"entropy": 2.28761186003685,
"epoch": 0.9195402298850575,
"grad_norm": 0.23978790640830994,
"learning_rate": 0.00016531728429537766,
"loss": 2.2996,
"mean_token_accuracy": 0.5478359237313271,
"num_tokens": 548991.0,
"step": 40
},
{
"entropy": 2.2919204078222575,
"epoch": 1.1379310344827587,
"grad_norm": 0.17945028841495514,
"learning_rate": 0.00014496113296546067,
"loss": 2.268,
"mean_token_accuracy": 0.5492534825676366,
"num_tokens": 676310.0,
"step": 50
},
{
"entropy": 2.225169724225998,
"epoch": 1.367816091954023,
"grad_norm": 0.1137014776468277,
"learning_rate": 0.00012191012401568698,
"loss": 2.2118,
"mean_token_accuracy": 0.5597869373857975,
"num_tokens": 813956.0,
"step": 60
},
{
"entropy": 2.225910446047783,
"epoch": 1.5977011494252875,
"grad_norm": 0.18948844075202942,
"learning_rate": 9.754587714770878e-05,
"loss": 2.2227,
"mean_token_accuracy": 0.5548668399453163,
"num_tokens": 949570.0,
"step": 70
},
{
"entropy": 2.192811119556427,
"epoch": 1.8275862068965516,
"grad_norm": 0.16777773201465607,
"learning_rate": 7.332872425251018e-05,
"loss": 2.1694,
"mean_token_accuracy": 0.560850913822651,
"num_tokens": 1082077.0,
"step": 80
},
{
"entropy": 2.2099951411548413,
"epoch": 2.045977011494253,
"grad_norm": 0.16436216235160828,
"learning_rate": 5.071018077702161e-05,
"loss": 2.2063,
"mean_token_accuracy": 0.5610575997515729,
"num_tokens": 1210170.0,
"step": 90
},
{
"entropy": 2.13824847638607,
"epoch": 2.2758620689655173,
"grad_norm": 0.18961983919143677,
"learning_rate": 3.104594552629331e-05,
"loss": 2.1102,
"mean_token_accuracy": 0.5674804911017418,
"num_tokens": 1350138.0,
"step": 100
},
{
"epoch": 2.2758620689655173,
"eval_entropy": 2.070272127787272,
"eval_loss": 2.1311750411987305,
"eval_mean_token_accuracy": 0.5795977115631104,
"eval_num_tokens": 1350138.0,
"eval_runtime": 7.0874,
"eval_samples_per_second": 2.681,
"eval_steps_per_second": 0.423,
"step": 100
}
],
"logging_steps": 10,
"max_steps": 132,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 2,
"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": false
},
"attributes": {}
}
},
"total_flos": 7.159998022449562e+16,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}