Upload multi_teacher_fantasy_alchemy/checkpoint-1800/trainer_state.json with huggingface_hub
Browse files
multi_teacher_fantasy_alchemy/checkpoint-1800/trainer_state.json
ADDED
|
@@ -0,0 +1,1294 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.4790468364831553,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1800,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.008216926869350863,
|
| 14 |
+
"grad_norm": 19.83480665335734,
|
| 15 |
+
"learning_rate": 9.240246406570842e-07,
|
| 16 |
+
"loss": 8.2641,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.016433853738701727,
|
| 21 |
+
"grad_norm": 18.704025706899706,
|
| 22 |
+
"learning_rate": 1.9507186858316226e-06,
|
| 23 |
+
"loss": 8.1054,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.024650780608052588,
|
| 28 |
+
"grad_norm": 9.254313873894626,
|
| 29 |
+
"learning_rate": 2.9774127310061602e-06,
|
| 30 |
+
"loss": 7.4171,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.03286770747740345,
|
| 35 |
+
"grad_norm": 5.336886587772746,
|
| 36 |
+
"learning_rate": 4.004106776180698e-06,
|
| 37 |
+
"loss": 6.652,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.041084634346754315,
|
| 42 |
+
"grad_norm": 3.356383707656696,
|
| 43 |
+
"learning_rate": 5.030800821355237e-06,
|
| 44 |
+
"loss": 6.1747,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.049301561216105176,
|
| 49 |
+
"grad_norm": 3.3258958472365467,
|
| 50 |
+
"learning_rate": 6.057494866529775e-06,
|
| 51 |
+
"loss": 5.8018,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.05751848808545604,
|
| 56 |
+
"grad_norm": 4.0537058731983,
|
| 57 |
+
"learning_rate": 7.084188911704312e-06,
|
| 58 |
+
"loss": 5.3316,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.0657354149548069,
|
| 63 |
+
"grad_norm": 4.086247611679589,
|
| 64 |
+
"learning_rate": 8.11088295687885e-06,
|
| 65 |
+
"loss": 4.6534,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.07395234182415776,
|
| 70 |
+
"grad_norm": 6.521478113854127,
|
| 71 |
+
"learning_rate": 9.137577002053388e-06,
|
| 72 |
+
"loss": 3.924,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.08216926869350863,
|
| 77 |
+
"grad_norm": 5.719373874300257,
|
| 78 |
+
"learning_rate": 1.0164271047227925e-05,
|
| 79 |
+
"loss": 3.3464,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.09038619556285948,
|
| 84 |
+
"grad_norm": 3.578443884259996,
|
| 85 |
+
"learning_rate": 1.1190965092402464e-05,
|
| 86 |
+
"loss": 2.9684,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.09860312243221035,
|
| 91 |
+
"grad_norm": 5.144310812867023,
|
| 92 |
+
"learning_rate": 1.2217659137577003e-05,
|
| 93 |
+
"loss": 2.734,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.10682004930156122,
|
| 98 |
+
"grad_norm": 5.024178909720029,
|
| 99 |
+
"learning_rate": 1.324435318275154e-05,
|
| 100 |
+
"loss": 2.5647,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.11503697617091208,
|
| 105 |
+
"grad_norm": 7.639677795548909,
|
| 106 |
+
"learning_rate": 1.427104722792608e-05,
|
| 107 |
+
"loss": 2.4388,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.12325390304026294,
|
| 112 |
+
"grad_norm": 7.4036645978760225,
|
| 113 |
+
"learning_rate": 1.5297741273100617e-05,
|
| 114 |
+
"loss": 2.3493,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.1314708299096138,
|
| 119 |
+
"grad_norm": 7.12561810579708,
|
| 120 |
+
"learning_rate": 1.6324435318275156e-05,
|
| 121 |
+
"loss": 2.2697,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.13968775677896467,
|
| 126 |
+
"grad_norm": 9.222649571442956,
|
| 127 |
+
"learning_rate": 1.735112936344969e-05,
|
| 128 |
+
"loss": 2.1975,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.14790468364831552,
|
| 133 |
+
"grad_norm": 8.868924318680982,
|
| 134 |
+
"learning_rate": 1.837782340862423e-05,
|
| 135 |
+
"loss": 2.1464,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.1561216105176664,
|
| 140 |
+
"grad_norm": 11.22877634105566,
|
| 141 |
+
"learning_rate": 1.940451745379877e-05,
|
| 142 |
+
"loss": 2.0921,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.16433853738701726,
|
| 147 |
+
"grad_norm": 14.840583315298584,
|
| 148 |
+
"learning_rate": 2.0431211498973308e-05,
|
| 149 |
+
"loss": 2.0518,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.1725554642563681,
|
| 154 |
+
"grad_norm": 12.147153984880905,
|
| 155 |
+
"learning_rate": 2.1457905544147843e-05,
|
| 156 |
+
"loss": 1.9688,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.18077239112571897,
|
| 161 |
+
"grad_norm": 17.004164152742906,
|
| 162 |
+
"learning_rate": 2.2484599589322382e-05,
|
| 163 |
+
"loss": 1.9184,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.18898931799506985,
|
| 168 |
+
"grad_norm": 54.36301855761989,
|
| 169 |
+
"learning_rate": 2.351129363449692e-05,
|
| 170 |
+
"loss": 1.8491,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.1972062448644207,
|
| 175 |
+
"grad_norm": 13.223675766771267,
|
| 176 |
+
"learning_rate": 2.453798767967146e-05,
|
| 177 |
+
"loss": 1.8062,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.20542317173377156,
|
| 182 |
+
"grad_norm": 23.739675599865905,
|
| 183 |
+
"learning_rate": 2.5564681724845996e-05,
|
| 184 |
+
"loss": 1.7007,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.21364009860312244,
|
| 189 |
+
"grad_norm": 13.136946015186366,
|
| 190 |
+
"learning_rate": 2.6591375770020538e-05,
|
| 191 |
+
"loss": 1.6032,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.2218570254724733,
|
| 196 |
+
"grad_norm": 11.684546950836488,
|
| 197 |
+
"learning_rate": 2.7618069815195074e-05,
|
| 198 |
+
"loss": 1.4973,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.23007395234182415,
|
| 203 |
+
"grad_norm": 7.667957804275019,
|
| 204 |
+
"learning_rate": 2.8644763860369613e-05,
|
| 205 |
+
"loss": 1.3794,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.23829087921117503,
|
| 210 |
+
"grad_norm": 12.12659183227395,
|
| 211 |
+
"learning_rate": 2.9671457905544148e-05,
|
| 212 |
+
"loss": 1.3322,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.2465078060805259,
|
| 217 |
+
"grad_norm": 8.972817044811347,
|
| 218 |
+
"learning_rate": 3.069815195071869e-05,
|
| 219 |
+
"loss": 1.2359,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.25472473294987674,
|
| 224 |
+
"grad_norm": 4.3028554301503465,
|
| 225 |
+
"learning_rate": 3.172484599589323e-05,
|
| 226 |
+
"loss": 1.1852,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.2629416598192276,
|
| 231 |
+
"grad_norm": 5.3453439233449185,
|
| 232 |
+
"learning_rate": 3.2751540041067765e-05,
|
| 233 |
+
"loss": 1.1183,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.27115858668857845,
|
| 238 |
+
"grad_norm": 6.559451348524855,
|
| 239 |
+
"learning_rate": 3.37782340862423e-05,
|
| 240 |
+
"loss": 1.0608,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.27937551355792933,
|
| 245 |
+
"grad_norm": 6.374692391110214,
|
| 246 |
+
"learning_rate": 3.480492813141684e-05,
|
| 247 |
+
"loss": 1.0151,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.2875924404272802,
|
| 252 |
+
"grad_norm": 7.303560110424278,
|
| 253 |
+
"learning_rate": 3.583162217659138e-05,
|
| 254 |
+
"loss": 0.9712,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.29580936729663104,
|
| 259 |
+
"grad_norm": 7.0001296670132485,
|
| 260 |
+
"learning_rate": 3.6858316221765914e-05,
|
| 261 |
+
"loss": 0.926,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.3040262941659819,
|
| 266 |
+
"grad_norm": 6.545632008564383,
|
| 267 |
+
"learning_rate": 3.788501026694045e-05,
|
| 268 |
+
"loss": 0.8996,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.3122432210353328,
|
| 273 |
+
"grad_norm": 4.842087687844629,
|
| 274 |
+
"learning_rate": 3.891170431211499e-05,
|
| 275 |
+
"loss": 0.8519,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.32046014790468363,
|
| 280 |
+
"grad_norm": 5.689801303683759,
|
| 281 |
+
"learning_rate": 3.993839835728953e-05,
|
| 282 |
+
"loss": 0.8223,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3286770747740345,
|
| 287 |
+
"grad_norm": 6.971495473441899,
|
| 288 |
+
"learning_rate": 4.096509240246407e-05,
|
| 289 |
+
"loss": 0.7828,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.3368940016433854,
|
| 294 |
+
"grad_norm": 6.36358312425706,
|
| 295 |
+
"learning_rate": 4.1991786447638605e-05,
|
| 296 |
+
"loss": 0.7489,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.3451109285127362,
|
| 301 |
+
"grad_norm": 8.676168539975894,
|
| 302 |
+
"learning_rate": 4.301848049281315e-05,
|
| 303 |
+
"loss": 0.731,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.3533278553820871,
|
| 308 |
+
"grad_norm": 7.918859246440221,
|
| 309 |
+
"learning_rate": 4.404517453798768e-05,
|
| 310 |
+
"loss": 0.7064,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.36154478225143794,
|
| 315 |
+
"grad_norm": 8.806726830982097,
|
| 316 |
+
"learning_rate": 4.507186858316222e-05,
|
| 317 |
+
"loss": 0.6826,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.3697617091207888,
|
| 322 |
+
"grad_norm": 6.737818041811399,
|
| 323 |
+
"learning_rate": 4.6098562628336754e-05,
|
| 324 |
+
"loss": 0.6548,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.3779786359901397,
|
| 329 |
+
"grad_norm": 6.523855090754944,
|
| 330 |
+
"learning_rate": 4.7125256673511296e-05,
|
| 331 |
+
"loss": 0.6279,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.3861955628594905,
|
| 336 |
+
"grad_norm": 8.120644576403212,
|
| 337 |
+
"learning_rate": 4.815195071868583e-05,
|
| 338 |
+
"loss": 0.6037,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.3944124897288414,
|
| 343 |
+
"grad_norm": 5.5098094876695285,
|
| 344 |
+
"learning_rate": 4.9178644763860374e-05,
|
| 345 |
+
"loss": 0.5764,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.4026294165981923,
|
| 350 |
+
"grad_norm": 8.171893404017268,
|
| 351 |
+
"learning_rate": 5.020533880903491e-05,
|
| 352 |
+
"loss": 0.5662,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.4108463434675431,
|
| 357 |
+
"grad_norm": 5.545430210957229,
|
| 358 |
+
"learning_rate": 5.1232032854209445e-05,
|
| 359 |
+
"loss": 0.5416,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.419063270336894,
|
| 364 |
+
"grad_norm": 6.687063853266749,
|
| 365 |
+
"learning_rate": 5.225872689938398e-05,
|
| 366 |
+
"loss": 0.5087,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.4272801972062449,
|
| 371 |
+
"grad_norm": 4.47366213437882,
|
| 372 |
+
"learning_rate": 5.328542094455853e-05,
|
| 373 |
+
"loss": 0.4881,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.4354971240755957,
|
| 378 |
+
"grad_norm": 4.172813447018652,
|
| 379 |
+
"learning_rate": 5.4312114989733065e-05,
|
| 380 |
+
"loss": 0.4621,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.4437140509449466,
|
| 385 |
+
"grad_norm": 4.220516576800102,
|
| 386 |
+
"learning_rate": 5.53388090349076e-05,
|
| 387 |
+
"loss": 0.4487,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.4519309778142975,
|
| 392 |
+
"grad_norm": 4.709217957842771,
|
| 393 |
+
"learning_rate": 5.6365503080082136e-05,
|
| 394 |
+
"loss": 0.4246,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.4601479046836483,
|
| 399 |
+
"grad_norm": 3.6943640807724925,
|
| 400 |
+
"learning_rate": 5.739219712525667e-05,
|
| 401 |
+
"loss": 0.4002,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.4683648315529992,
|
| 406 |
+
"grad_norm": 4.324443086003805,
|
| 407 |
+
"learning_rate": 5.8418891170431214e-05,
|
| 408 |
+
"loss": 0.3769,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.47658175842235007,
|
| 413 |
+
"grad_norm": 4.712485210987473,
|
| 414 |
+
"learning_rate": 5.944558521560575e-05,
|
| 415 |
+
"loss": 0.3605,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.4847986852917009,
|
| 420 |
+
"grad_norm": 2.5269576298422347,
|
| 421 |
+
"learning_rate": 6.04722792607803e-05,
|
| 422 |
+
"loss": 0.3417,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.4930156121610518,
|
| 427 |
+
"grad_norm": 3.54537374699048,
|
| 428 |
+
"learning_rate": 6.149897330595483e-05,
|
| 429 |
+
"loss": 0.3316,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.5012325390304027,
|
| 434 |
+
"grad_norm": 3.530633503104792,
|
| 435 |
+
"learning_rate": 6.252566735112937e-05,
|
| 436 |
+
"loss": 0.313,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5094494658997535,
|
| 441 |
+
"grad_norm": 4.839997750327561,
|
| 442 |
+
"learning_rate": 6.35523613963039e-05,
|
| 443 |
+
"loss": 0.2974,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5176663927691043,
|
| 448 |
+
"grad_norm": 4.197641178153233,
|
| 449 |
+
"learning_rate": 6.457905544147844e-05,
|
| 450 |
+
"loss": 0.2745,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5258833196384552,
|
| 455 |
+
"grad_norm": 2.705058855982426,
|
| 456 |
+
"learning_rate": 6.560574948665298e-05,
|
| 457 |
+
"loss": 0.2615,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.5341002465078061,
|
| 462 |
+
"grad_norm": 2.8099029424585313,
|
| 463 |
+
"learning_rate": 6.663244353182751e-05,
|
| 464 |
+
"loss": 0.2503,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5423171733771569,
|
| 469 |
+
"grad_norm": 2.5734835696398712,
|
| 470 |
+
"learning_rate": 6.765913757700205e-05,
|
| 471 |
+
"loss": 0.2415,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.5505341002465078,
|
| 476 |
+
"grad_norm": 2.7836726382586425,
|
| 477 |
+
"learning_rate": 6.86858316221766e-05,
|
| 478 |
+
"loss": 0.2344,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.5587510271158587,
|
| 483 |
+
"grad_norm": 2.603051010989796,
|
| 484 |
+
"learning_rate": 6.971252566735113e-05,
|
| 485 |
+
"loss": 0.2154,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.5669679539852095,
|
| 490 |
+
"grad_norm": 2.1107040737971587,
|
| 491 |
+
"learning_rate": 7.073921971252567e-05,
|
| 492 |
+
"loss": 0.1976,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.5751848808545604,
|
| 497 |
+
"grad_norm": 2.803095444807186,
|
| 498 |
+
"learning_rate": 7.17659137577002e-05,
|
| 499 |
+
"loss": 0.1864,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.5834018077239113,
|
| 504 |
+
"grad_norm": 2.1670666899861075,
|
| 505 |
+
"learning_rate": 7.279260780287475e-05,
|
| 506 |
+
"loss": 0.178,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.5916187345932621,
|
| 511 |
+
"grad_norm": 1.7921531158206865,
|
| 512 |
+
"learning_rate": 7.381930184804929e-05,
|
| 513 |
+
"loss": 0.1652,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.599835661462613,
|
| 518 |
+
"grad_norm": 1.479382404686151,
|
| 519 |
+
"learning_rate": 7.484599589322382e-05,
|
| 520 |
+
"loss": 0.1576,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.6080525883319639,
|
| 525 |
+
"grad_norm": 1.743395608685881,
|
| 526 |
+
"learning_rate": 7.587268993839836e-05,
|
| 527 |
+
"loss": 0.1498,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.6162695152013147,
|
| 532 |
+
"grad_norm": 1.8303806232104347,
|
| 533 |
+
"learning_rate": 7.689938398357291e-05,
|
| 534 |
+
"loss": 0.136,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.6244864420706656,
|
| 539 |
+
"grad_norm": 2.312871831212618,
|
| 540 |
+
"learning_rate": 7.792607802874744e-05,
|
| 541 |
+
"loss": 0.1274,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.6327033689400164,
|
| 546 |
+
"grad_norm": 1.6901573163360035,
|
| 547 |
+
"learning_rate": 7.895277207392198e-05,
|
| 548 |
+
"loss": 0.1191,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.6409202958093673,
|
| 553 |
+
"grad_norm": 1.6314268202265299,
|
| 554 |
+
"learning_rate": 7.997946611909651e-05,
|
| 555 |
+
"loss": 0.1114,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.6491372226787182,
|
| 560 |
+
"grad_norm": 1.4508574236430893,
|
| 561 |
+
"learning_rate": 8.100616016427105e-05,
|
| 562 |
+
"loss": 0.101,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.657354149548069,
|
| 567 |
+
"grad_norm": 1.5649749543554101,
|
| 568 |
+
"learning_rate": 8.203285420944559e-05,
|
| 569 |
+
"loss": 0.0985,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.6655710764174199,
|
| 574 |
+
"grad_norm": 2.6845550965354428,
|
| 575 |
+
"learning_rate": 8.305954825462012e-05,
|
| 576 |
+
"loss": 0.1027,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.6737880032867708,
|
| 581 |
+
"grad_norm": 2.7764588978262426,
|
| 582 |
+
"learning_rate": 8.408624229979466e-05,
|
| 583 |
+
"loss": 0.1035,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.6820049301561216,
|
| 588 |
+
"grad_norm": 4.830497904460127,
|
| 589 |
+
"learning_rate": 8.51129363449692e-05,
|
| 590 |
+
"loss": 0.0976,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.6902218570254725,
|
| 595 |
+
"grad_norm": 1.0331763322679437,
|
| 596 |
+
"learning_rate": 8.613963039014374e-05,
|
| 597 |
+
"loss": 0.0879,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.6984387838948234,
|
| 602 |
+
"grad_norm": 0.596374501379962,
|
| 603 |
+
"learning_rate": 8.716632443531828e-05,
|
| 604 |
+
"loss": 0.074,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.7066557107641742,
|
| 609 |
+
"grad_norm": 0.8196280038018171,
|
| 610 |
+
"learning_rate": 8.819301848049281e-05,
|
| 611 |
+
"loss": 0.0692,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.714872637633525,
|
| 616 |
+
"grad_norm": 0.9801217376138596,
|
| 617 |
+
"learning_rate": 8.921971252566735e-05,
|
| 618 |
+
"loss": 0.065,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.7230895645028759,
|
| 623 |
+
"grad_norm": 0.5051290053812035,
|
| 624 |
+
"learning_rate": 9.02464065708419e-05,
|
| 625 |
+
"loss": 0.0596,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.7313064913722268,
|
| 630 |
+
"grad_norm": 0.9667645246911198,
|
| 631 |
+
"learning_rate": 9.127310061601643e-05,
|
| 632 |
+
"loss": 0.0556,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.7395234182415776,
|
| 637 |
+
"grad_norm": 0.46438651540167103,
|
| 638 |
+
"learning_rate": 9.229979466119097e-05,
|
| 639 |
+
"loss": 0.0545,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.7477403451109285,
|
| 644 |
+
"grad_norm": 0.40547338931365595,
|
| 645 |
+
"learning_rate": 9.332648870636552e-05,
|
| 646 |
+
"loss": 0.0494,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.7559572719802794,
|
| 651 |
+
"grad_norm": 0.4441111869944399,
|
| 652 |
+
"learning_rate": 9.435318275154005e-05,
|
| 653 |
+
"loss": 0.0466,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.7641741988496302,
|
| 658 |
+
"grad_norm": 0.8183415391457746,
|
| 659 |
+
"learning_rate": 9.537987679671459e-05,
|
| 660 |
+
"loss": 0.0453,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.772391125718981,
|
| 665 |
+
"grad_norm": 0.5138559384321786,
|
| 666 |
+
"learning_rate": 9.640657084188912e-05,
|
| 667 |
+
"loss": 0.0436,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.780608052588332,
|
| 672 |
+
"grad_norm": 0.474231944626131,
|
| 673 |
+
"learning_rate": 9.743326488706366e-05,
|
| 674 |
+
"loss": 0.0415,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.7888249794576828,
|
| 679 |
+
"grad_norm": 0.4407732436458076,
|
| 680 |
+
"learning_rate": 9.84599589322382e-05,
|
| 681 |
+
"loss": 0.04,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.7970419063270336,
|
| 686 |
+
"grad_norm": 1.888951859317562,
|
| 687 |
+
"learning_rate": 9.948665297741273e-05,
|
| 688 |
+
"loss": 0.0401,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.8052588331963846,
|
| 693 |
+
"grad_norm": 0.5937342056949868,
|
| 694 |
+
"learning_rate": 9.999991965235563e-05,
|
| 695 |
+
"loss": 0.0408,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.8134757600657354,
|
| 700 |
+
"grad_norm": 0.6643097529303563,
|
| 701 |
+
"learning_rate": 9.999927687274992e-05,
|
| 702 |
+
"loss": 0.0458,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.8216926869350862,
|
| 707 |
+
"grad_norm": 0.38763729674702113,
|
| 708 |
+
"learning_rate": 9.999799132180184e-05,
|
| 709 |
+
"loss": 0.0364,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.8299096138044372,
|
| 714 |
+
"grad_norm": 0.3739679848963624,
|
| 715 |
+
"learning_rate": 9.999606301603793e-05,
|
| 716 |
+
"loss": 0.0332,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.838126540673788,
|
| 721 |
+
"grad_norm": 0.3960365233672132,
|
| 722 |
+
"learning_rate": 9.999349198024775e-05,
|
| 723 |
+
"loss": 0.0327,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.8463434675431388,
|
| 728 |
+
"grad_norm": 0.35789593639188133,
|
| 729 |
+
"learning_rate": 9.999027824748352e-05,
|
| 730 |
+
"loss": 0.0313,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.8545603944124898,
|
| 735 |
+
"grad_norm": 0.3328319306799121,
|
| 736 |
+
"learning_rate": 9.998642185905977e-05,
|
| 737 |
+
"loss": 0.03,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.8627773212818406,
|
| 742 |
+
"grad_norm": 0.397125770040445,
|
| 743 |
+
"learning_rate": 9.998192286455273e-05,
|
| 744 |
+
"loss": 0.0285,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.8709942481511914,
|
| 749 |
+
"grad_norm": 0.35895726006131834,
|
| 750 |
+
"learning_rate": 9.997678132179969e-05,
|
| 751 |
+
"loss": 0.0283,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.8792111750205424,
|
| 756 |
+
"grad_norm": 0.359528719136988,
|
| 757 |
+
"learning_rate": 9.997099729689839e-05,
|
| 758 |
+
"loss": 0.0272,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.8874281018898932,
|
| 763 |
+
"grad_norm": 0.31783947017752745,
|
| 764 |
+
"learning_rate": 9.996457086420598e-05,
|
| 765 |
+
"loss": 0.026,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.895645028759244,
|
| 770 |
+
"grad_norm": 0.24437306397975714,
|
| 771 |
+
"learning_rate": 9.995750210633822e-05,
|
| 772 |
+
"loss": 0.0252,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.903861955628595,
|
| 777 |
+
"grad_norm": 0.2869935821010678,
|
| 778 |
+
"learning_rate": 9.994979111416829e-05,
|
| 779 |
+
"loss": 0.0243,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.9120788824979458,
|
| 784 |
+
"grad_norm": 0.28967525917886927,
|
| 785 |
+
"learning_rate": 9.994143798682573e-05,
|
| 786 |
+
"loss": 0.0236,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.9202958093672966,
|
| 791 |
+
"grad_norm": 0.4867480221470474,
|
| 792 |
+
"learning_rate": 9.993244283169513e-05,
|
| 793 |
+
"loss": 0.0232,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.9285127362366475,
|
| 798 |
+
"grad_norm": 0.27522967583273783,
|
| 799 |
+
"learning_rate": 9.99228057644147e-05,
|
| 800 |
+
"loss": 0.0229,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.9367296631059984,
|
| 805 |
+
"grad_norm": 0.2705586478283337,
|
| 806 |
+
"learning_rate": 9.991252690887485e-05,
|
| 807 |
+
"loss": 0.0217,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.9449465899753492,
|
| 812 |
+
"grad_norm": 0.3133390383629242,
|
| 813 |
+
"learning_rate": 9.990160639721657e-05,
|
| 814 |
+
"loss": 0.0209,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.9531635168447001,
|
| 819 |
+
"grad_norm": 0.3003241130339291,
|
| 820 |
+
"learning_rate": 9.989004436982974e-05,
|
| 821 |
+
"loss": 0.0204,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.961380443714051,
|
| 826 |
+
"grad_norm": 0.27998609610081276,
|
| 827 |
+
"learning_rate": 9.987784097535127e-05,
|
| 828 |
+
"loss": 0.02,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.9695973705834018,
|
| 833 |
+
"grad_norm": 0.3094626205699417,
|
| 834 |
+
"learning_rate": 9.986499637066331e-05,
|
| 835 |
+
"loss": 0.0198,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.9778142974527527,
|
| 840 |
+
"grad_norm": 0.311495504428665,
|
| 841 |
+
"learning_rate": 9.985151072089112e-05,
|
| 842 |
+
"loss": 0.0193,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.9860312243221035,
|
| 847 |
+
"grad_norm": 0.2810349187040472,
|
| 848 |
+
"learning_rate": 9.983738419940096e-05,
|
| 849 |
+
"loss": 0.0193,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.9942481511914544,
|
| 854 |
+
"grad_norm": 0.3107871046473083,
|
| 855 |
+
"learning_rate": 9.982261698779795e-05,
|
| 856 |
+
"loss": 0.0185,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.0024650780608053,
|
| 861 |
+
"grad_norm": 0.26519682492524527,
|
| 862 |
+
"learning_rate": 9.980720927592366e-05,
|
| 863 |
+
"loss": 0.0182,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.0106820049301561,
|
| 868 |
+
"grad_norm": 0.2391845678539309,
|
| 869 |
+
"learning_rate": 9.979116126185364e-05,
|
| 870 |
+
"loss": 0.0173,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.018898931799507,
|
| 875 |
+
"grad_norm": 0.26242451803570593,
|
| 876 |
+
"learning_rate": 9.977447315189493e-05,
|
| 877 |
+
"loss": 0.0166,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.0271158586688578,
|
| 882 |
+
"grad_norm": 0.22409653311325814,
|
| 883 |
+
"learning_rate": 9.975714516058344e-05,
|
| 884 |
+
"loss": 0.0162,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.0353327855382086,
|
| 889 |
+
"grad_norm": 0.2343121137078575,
|
| 890 |
+
"learning_rate": 9.973917751068111e-05,
|
| 891 |
+
"loss": 0.0155,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.0435497124075597,
|
| 896 |
+
"grad_norm": 0.21473180531882663,
|
| 897 |
+
"learning_rate": 9.972057043317308e-05,
|
| 898 |
+
"loss": 0.0152,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.0517666392769105,
|
| 903 |
+
"grad_norm": 0.1994724694575453,
|
| 904 |
+
"learning_rate": 9.970132416726474e-05,
|
| 905 |
+
"loss": 0.0151,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.0599835661462613,
|
| 910 |
+
"grad_norm": 0.1979357989976122,
|
| 911 |
+
"learning_rate": 9.968143896037864e-05,
|
| 912 |
+
"loss": 0.0147,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.0682004930156122,
|
| 917 |
+
"grad_norm": 0.23691205439082766,
|
| 918 |
+
"learning_rate": 9.966091506815128e-05,
|
| 919 |
+
"loss": 0.0144,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.076417419884963,
|
| 924 |
+
"grad_norm": 2.057726124603243,
|
| 925 |
+
"learning_rate": 9.963975275442989e-05,
|
| 926 |
+
"loss": 0.0175,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.0846343467543138,
|
| 931 |
+
"grad_norm": 0.3234261374936929,
|
| 932 |
+
"learning_rate": 9.961795229126896e-05,
|
| 933 |
+
"loss": 0.0165,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.0928512736236646,
|
| 938 |
+
"grad_norm": 0.2509885643136706,
|
| 939 |
+
"learning_rate": 9.959551395892681e-05,
|
| 940 |
+
"loss": 0.0149,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.1010682004930157,
|
| 945 |
+
"grad_norm": 0.19772103993153273,
|
| 946 |
+
"learning_rate": 9.957243804586197e-05,
|
| 947 |
+
"loss": 0.0141,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.1092851273623665,
|
| 952 |
+
"grad_norm": 0.23068483781938592,
|
| 953 |
+
"learning_rate": 9.954872484872944e-05,
|
| 954 |
+
"loss": 0.0134,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.1175020542317173,
|
| 959 |
+
"grad_norm": 0.21085093284095438,
|
| 960 |
+
"learning_rate": 9.952437467237687e-05,
|
| 961 |
+
"loss": 0.0133,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.1257189811010682,
|
| 966 |
+
"grad_norm": 0.21244911066205022,
|
| 967 |
+
"learning_rate": 9.949938782984072e-05,
|
| 968 |
+
"loss": 0.0131,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.133935907970419,
|
| 973 |
+
"grad_norm": 0.1841268164629968,
|
| 974 |
+
"learning_rate": 9.947376464234215e-05,
|
| 975 |
+
"loss": 0.0126,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.14215283483977,
|
| 980 |
+
"grad_norm": 0.18885789130639805,
|
| 981 |
+
"learning_rate": 9.944750543928297e-05,
|
| 982 |
+
"loss": 0.0123,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.1503697617091209,
|
| 987 |
+
"grad_norm": 0.18962637141794314,
|
| 988 |
+
"learning_rate": 9.942061055824127e-05,
|
| 989 |
+
"loss": 0.0121,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.1585866885784717,
|
| 994 |
+
"grad_norm": 0.20049383873744792,
|
| 995 |
+
"learning_rate": 9.939308034496728e-05,
|
| 996 |
+
"loss": 0.0121,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.1668036154478225,
|
| 1001 |
+
"grad_norm": 0.19041767566091838,
|
| 1002 |
+
"learning_rate": 9.936491515337874e-05,
|
| 1003 |
+
"loss": 0.0119,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.1750205423171733,
|
| 1008 |
+
"grad_norm": 0.1995551393056727,
|
| 1009 |
+
"learning_rate": 9.933611534555644e-05,
|
| 1010 |
+
"loss": 0.0117,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.1832374691865242,
|
| 1015 |
+
"grad_norm": 0.16768458374812184,
|
| 1016 |
+
"learning_rate": 9.930668129173956e-05,
|
| 1017 |
+
"loss": 0.0116,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.191454396055875,
|
| 1022 |
+
"grad_norm": 0.18748344143729523,
|
| 1023 |
+
"learning_rate": 9.927661337032093e-05,
|
| 1024 |
+
"loss": 0.0115,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.199671322925226,
|
| 1029 |
+
"grad_norm": 0.17618964617531263,
|
| 1030 |
+
"learning_rate": 9.924591196784205e-05,
|
| 1031 |
+
"loss": 0.0113,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.2078882497945769,
|
| 1036 |
+
"grad_norm": 0.1915822510753366,
|
| 1037 |
+
"learning_rate": 9.92145774789883e-05,
|
| 1038 |
+
"loss": 0.011,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.2161051766639277,
|
| 1043 |
+
"grad_norm": 0.22348016077011987,
|
| 1044 |
+
"learning_rate": 9.918261030658371e-05,
|
| 1045 |
+
"loss": 0.011,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.2243221035332785,
|
| 1050 |
+
"grad_norm": 0.17910179979882987,
|
| 1051 |
+
"learning_rate": 9.915001086158588e-05,
|
| 1052 |
+
"loss": 0.0109,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.2325390304026294,
|
| 1057 |
+
"grad_norm": 0.20954080994397253,
|
| 1058 |
+
"learning_rate": 9.911677956308068e-05,
|
| 1059 |
+
"loss": 0.0106,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.2407559572719804,
|
| 1064 |
+
"grad_norm": 0.18491793792738792,
|
| 1065 |
+
"learning_rate": 9.908291683827678e-05,
|
| 1066 |
+
"loss": 0.0103,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.2489728841413312,
|
| 1071 |
+
"grad_norm": 0.1913002285241464,
|
| 1072 |
+
"learning_rate": 9.904842312250024e-05,
|
| 1073 |
+
"loss": 0.01,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.257189811010682,
|
| 1078 |
+
"grad_norm": 0.18665545298861494,
|
| 1079 |
+
"learning_rate": 9.901329885918898e-05,
|
| 1080 |
+
"loss": 0.0102,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.2654067378800329,
|
| 1085 |
+
"grad_norm": 0.20441793337405792,
|
| 1086 |
+
"learning_rate": 9.897754449988687e-05,
|
| 1087 |
+
"loss": 0.0099,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.2736236647493837,
|
| 1092 |
+
"grad_norm": 0.17232110916564725,
|
| 1093 |
+
"learning_rate": 9.894116050423812e-05,
|
| 1094 |
+
"loss": 0.01,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.2818405916187345,
|
| 1099 |
+
"grad_norm": 0.17117350565381512,
|
| 1100 |
+
"learning_rate": 9.890414733998131e-05,
|
| 1101 |
+
"loss": 0.0097,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.2900575184880854,
|
| 1106 |
+
"grad_norm": 0.16106675543143417,
|
| 1107 |
+
"learning_rate": 9.886650548294333e-05,
|
| 1108 |
+
"loss": 0.0098,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.2982744453574364,
|
| 1113 |
+
"grad_norm": 0.15673244249109636,
|
| 1114 |
+
"learning_rate": 9.882823541703332e-05,
|
| 1115 |
+
"loss": 0.0096,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.3064913722267872,
|
| 1120 |
+
"grad_norm": 0.18359994942806246,
|
| 1121 |
+
"learning_rate": 9.878933763423644e-05,
|
| 1122 |
+
"loss": 0.0095,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.314708299096138,
|
| 1127 |
+
"grad_norm": 0.16928325787860654,
|
| 1128 |
+
"learning_rate": 9.874981263460749e-05,
|
| 1129 |
+
"loss": 0.0092,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.322925225965489,
|
| 1134 |
+
"grad_norm": 0.17321373689712377,
|
| 1135 |
+
"learning_rate": 9.87096609262646e-05,
|
| 1136 |
+
"loss": 0.0091,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.3311421528348397,
|
| 1141 |
+
"grad_norm": 0.18291522891441084,
|
| 1142 |
+
"learning_rate": 9.866888302538255e-05,
|
| 1143 |
+
"loss": 0.009,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.3393590797041908,
|
| 1148 |
+
"grad_norm": 0.19719101776097683,
|
| 1149 |
+
"learning_rate": 9.862747945618629e-05,
|
| 1150 |
+
"loss": 0.0091,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.3475760065735414,
|
| 1155 |
+
"grad_norm": 0.1507248639899357,
|
| 1156 |
+
"learning_rate": 9.858545075094402e-05,
|
| 1157 |
+
"loss": 0.0089,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.3557929334428924,
|
| 1162 |
+
"grad_norm": 0.15900627243941012,
|
| 1163 |
+
"learning_rate": 9.854279744996054e-05,
|
| 1164 |
+
"loss": 0.0085,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.3640098603122432,
|
| 1169 |
+
"grad_norm": 0.1614739960911069,
|
| 1170 |
+
"learning_rate": 9.849952010157014e-05,
|
| 1171 |
+
"loss": 0.0084,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.372226787181594,
|
| 1176 |
+
"grad_norm": 0.18518862249859025,
|
| 1177 |
+
"learning_rate": 9.845561926212967e-05,
|
| 1178 |
+
"loss": 0.0085,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.380443714050945,
|
| 1183 |
+
"grad_norm": 0.16465865540140373,
|
| 1184 |
+
"learning_rate": 9.841109549601133e-05,
|
| 1185 |
+
"loss": 0.0087,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.3886606409202957,
|
| 1190 |
+
"grad_norm": 0.1778246547911174,
|
| 1191 |
+
"learning_rate": 9.83659493755954e-05,
|
| 1192 |
+
"loss": 0.0085,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.3968775677896468,
|
| 1197 |
+
"grad_norm": 0.17535152748688587,
|
| 1198 |
+
"learning_rate": 9.832018148126291e-05,
|
| 1199 |
+
"loss": 0.0082,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.4050944946589976,
|
| 1204 |
+
"grad_norm": 0.1616895940395759,
|
| 1205 |
+
"learning_rate": 9.827379240138824e-05,
|
| 1206 |
+
"loss": 0.008,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.4133114215283484,
|
| 1211 |
+
"grad_norm": 0.17228485317380646,
|
| 1212 |
+
"learning_rate": 9.822678273233139e-05,
|
| 1213 |
+
"loss": 0.0081,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.4215283483976993,
|
| 1218 |
+
"grad_norm": 0.14635147682270658,
|
| 1219 |
+
"learning_rate": 9.817915307843046e-05,
|
| 1220 |
+
"loss": 0.008,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.42974527526705,
|
| 1225 |
+
"grad_norm": 0.1577912852087264,
|
| 1226 |
+
"learning_rate": 9.813090405199388e-05,
|
| 1227 |
+
"loss": 0.0078,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.437962202136401,
|
| 1232 |
+
"grad_norm": 0.13215393793980348,
|
| 1233 |
+
"learning_rate": 9.80820362732924e-05,
|
| 1234 |
+
"loss": 0.0079,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.4461791290057517,
|
| 1239 |
+
"grad_norm": 0.14886527670507846,
|
| 1240 |
+
"learning_rate": 9.803255037055128e-05,
|
| 1241 |
+
"loss": 0.0077,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.4543960558751028,
|
| 1246 |
+
"grad_norm": 0.1663787864042773,
|
| 1247 |
+
"learning_rate": 9.798244697994214e-05,
|
| 1248 |
+
"loss": 0.0074,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.4626129827444536,
|
| 1253 |
+
"grad_norm": 0.14313267064422866,
|
| 1254 |
+
"learning_rate": 9.793172674557475e-05,
|
| 1255 |
+
"loss": 0.0072,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.4708299096138044,
|
| 1260 |
+
"grad_norm": 0.14881896112838372,
|
| 1261 |
+
"learning_rate": 9.78803903194888e-05,
|
| 1262 |
+
"loss": 0.0075,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.4790468364831553,
|
| 1267 |
+
"grad_norm": 0.1378603541071152,
|
| 1268 |
+
"learning_rate": 9.782843836164553e-05,
|
| 1269 |
+
"loss": 0.0071,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
}
|
| 1272 |
+
],
|
| 1273 |
+
"logging_steps": 10,
|
| 1274 |
+
"max_steps": 9736,
|
| 1275 |
+
"num_input_tokens_seen": 0,
|
| 1276 |
+
"num_train_epochs": 8,
|
| 1277 |
+
"save_steps": 1800,
|
| 1278 |
+
"stateful_callbacks": {
|
| 1279 |
+
"TrainerControl": {
|
| 1280 |
+
"args": {
|
| 1281 |
+
"should_epoch_stop": false,
|
| 1282 |
+
"should_evaluate": false,
|
| 1283 |
+
"should_log": false,
|
| 1284 |
+
"should_save": true,
|
| 1285 |
+
"should_training_stop": false
|
| 1286 |
+
},
|
| 1287 |
+
"attributes": {}
|
| 1288 |
+
}
|
| 1289 |
+
},
|
| 1290 |
+
"total_flos": 172857185796096.0,
|
| 1291 |
+
"train_batch_size": 128,
|
| 1292 |
+
"trial_name": null,
|
| 1293 |
+
"trial_params": null
|
| 1294 |
+
}
|