Upload multi_student_6/checkpoint-1800/trainer_state.json with huggingface_hub
Browse files
multi_student_6/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": 1.002227791701476,
|
| 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": 0.00556947925368978,
|
| 14 |
+
"grad_norm": 19.463690451045245,
|
| 15 |
+
"learning_rate": 7.154213036565979e-07,
|
| 16 |
+
"loss": 8.2534,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.01113895850737956,
|
| 21 |
+
"grad_norm": 19.73356911714459,
|
| 22 |
+
"learning_rate": 1.51033386327504e-06,
|
| 23 |
+
"loss": 8.1778,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.01670843776106934,
|
| 28 |
+
"grad_norm": 10.029986538109684,
|
| 29 |
+
"learning_rate": 2.3052464228934817e-06,
|
| 30 |
+
"loss": 7.6937,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.02227791701475912,
|
| 35 |
+
"grad_norm": 7.379827099032999,
|
| 36 |
+
"learning_rate": 3.100158982511924e-06,
|
| 37 |
+
"loss": 6.9205,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.0278473962684489,
|
| 42 |
+
"grad_norm": 3.5728211270030625,
|
| 43 |
+
"learning_rate": 3.895071542130366e-06,
|
| 44 |
+
"loss": 6.3948,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.03341687552213868,
|
| 49 |
+
"grad_norm": 2.8392085477918747,
|
| 50 |
+
"learning_rate": 4.689984101748808e-06,
|
| 51 |
+
"loss": 6.0575,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.03898635477582846,
|
| 56 |
+
"grad_norm": 3.3563782303356495,
|
| 57 |
+
"learning_rate": 5.48489666136725e-06,
|
| 58 |
+
"loss": 5.7294,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.04455583402951824,
|
| 63 |
+
"grad_norm": 3.826365209612832,
|
| 64 |
+
"learning_rate": 6.279809220985691e-06,
|
| 65 |
+
"loss": 5.2614,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.05012531328320802,
|
| 70 |
+
"grad_norm": 3.965749199888462,
|
| 71 |
+
"learning_rate": 7.0747217806041336e-06,
|
| 72 |
+
"loss": 4.658,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.0556947925368978,
|
| 77 |
+
"grad_norm": 5.147891128586454,
|
| 78 |
+
"learning_rate": 7.869634340222575e-06,
|
| 79 |
+
"loss": 4.0477,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.06126427179058758,
|
| 84 |
+
"grad_norm": 3.695764472570234,
|
| 85 |
+
"learning_rate": 8.664546899841019e-06,
|
| 86 |
+
"loss": 3.5584,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.06683375104427736,
|
| 91 |
+
"grad_norm": 4.026692299224468,
|
| 92 |
+
"learning_rate": 9.45945945945946e-06,
|
| 93 |
+
"loss": 3.2227,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.07240323029796714,
|
| 98 |
+
"grad_norm": 5.389627945808325,
|
| 99 |
+
"learning_rate": 1.02543720190779e-05,
|
| 100 |
+
"loss": 3.0255,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.07797270955165692,
|
| 105 |
+
"grad_norm": 5.554941152889833,
|
| 106 |
+
"learning_rate": 1.1049284578696344e-05,
|
| 107 |
+
"loss": 2.8932,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.0835421888053467,
|
| 112 |
+
"grad_norm": 4.8986035107973995,
|
| 113 |
+
"learning_rate": 1.1844197138314786e-05,
|
| 114 |
+
"loss": 2.7959,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.08911166805903648,
|
| 119 |
+
"grad_norm": 4.56461997585786,
|
| 120 |
+
"learning_rate": 1.2639109697933228e-05,
|
| 121 |
+
"loss": 2.7067,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.09468114731272625,
|
| 126 |
+
"grad_norm": 9.192995633454027,
|
| 127 |
+
"learning_rate": 1.3434022257551671e-05,
|
| 128 |
+
"loss": 2.6194,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.10025062656641603,
|
| 133 |
+
"grad_norm": 8.520097501414543,
|
| 134 |
+
"learning_rate": 1.4228934817170111e-05,
|
| 135 |
+
"loss": 2.5516,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.10582010582010581,
|
| 140 |
+
"grad_norm": 14.274847317292076,
|
| 141 |
+
"learning_rate": 1.5023847376788555e-05,
|
| 142 |
+
"loss": 2.4789,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.1113895850737956,
|
| 147 |
+
"grad_norm": 19.214437222876512,
|
| 148 |
+
"learning_rate": 1.5818759936406995e-05,
|
| 149 |
+
"loss": 2.4363,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.11695906432748537,
|
| 154 |
+
"grad_norm": 16.394738746567327,
|
| 155 |
+
"learning_rate": 1.6613672496025438e-05,
|
| 156 |
+
"loss": 2.3902,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.12252854358117515,
|
| 161 |
+
"grad_norm": 18.247117883573495,
|
| 162 |
+
"learning_rate": 1.740858505564388e-05,
|
| 163 |
+
"loss": 2.3452,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.12809802283486493,
|
| 168 |
+
"grad_norm": 15.142493870476338,
|
| 169 |
+
"learning_rate": 1.820349761526232e-05,
|
| 170 |
+
"loss": 2.3042,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.1336675020885547,
|
| 175 |
+
"grad_norm": 8.204095855953893,
|
| 176 |
+
"learning_rate": 1.8998410174880765e-05,
|
| 177 |
+
"loss": 2.2554,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.1392369813422445,
|
| 182 |
+
"grad_norm": 12.061502490559853,
|
| 183 |
+
"learning_rate": 1.9793322734499205e-05,
|
| 184 |
+
"loss": 2.2062,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.14480646059593427,
|
| 189 |
+
"grad_norm": 13.078924954972894,
|
| 190 |
+
"learning_rate": 2.058823529411765e-05,
|
| 191 |
+
"loss": 2.1414,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.15037593984962405,
|
| 196 |
+
"grad_norm": 13.325114907384508,
|
| 197 |
+
"learning_rate": 2.1383147853736092e-05,
|
| 198 |
+
"loss": 2.059,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.15594541910331383,
|
| 203 |
+
"grad_norm": 7.4709725407687255,
|
| 204 |
+
"learning_rate": 2.2178060413354532e-05,
|
| 205 |
+
"loss": 1.9772,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.1615148983570036,
|
| 210 |
+
"grad_norm": 19.25846753081156,
|
| 211 |
+
"learning_rate": 2.2972972972972976e-05,
|
| 212 |
+
"loss": 1.9011,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.1670843776106934,
|
| 217 |
+
"grad_norm": 13.01475345365111,
|
| 218 |
+
"learning_rate": 2.3767885532591416e-05,
|
| 219 |
+
"loss": 1.8118,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.17265385686438317,
|
| 224 |
+
"grad_norm": 6.508219489647502,
|
| 225 |
+
"learning_rate": 2.4562798092209856e-05,
|
| 226 |
+
"loss": 1.7365,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.17822333611807295,
|
| 231 |
+
"grad_norm": 8.638960149014848,
|
| 232 |
+
"learning_rate": 2.53577106518283e-05,
|
| 233 |
+
"loss": 1.6377,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.18379281537176273,
|
| 238 |
+
"grad_norm": 5.254562295866414,
|
| 239 |
+
"learning_rate": 2.6152623211446743e-05,
|
| 240 |
+
"loss": 1.5737,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.1893622946254525,
|
| 245 |
+
"grad_norm": 8.547148285512923,
|
| 246 |
+
"learning_rate": 2.6947535771065186e-05,
|
| 247 |
+
"loss": 1.5303,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.1949317738791423,
|
| 252 |
+
"grad_norm": 6.278158376274084,
|
| 253 |
+
"learning_rate": 2.7742448330683623e-05,
|
| 254 |
+
"loss": 1.4716,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.20050125313283207,
|
| 259 |
+
"grad_norm": 8.26795777275924,
|
| 260 |
+
"learning_rate": 2.8537360890302066e-05,
|
| 261 |
+
"loss": 1.4151,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.20607073238652185,
|
| 266 |
+
"grad_norm": 4.425290147294405,
|
| 267 |
+
"learning_rate": 2.933227344992051e-05,
|
| 268 |
+
"loss": 1.368,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.21164021164021163,
|
| 273 |
+
"grad_norm": 6.848350616045537,
|
| 274 |
+
"learning_rate": 3.0127186009538953e-05,
|
| 275 |
+
"loss": 1.3383,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.2172096908939014,
|
| 280 |
+
"grad_norm": 4.869312013250087,
|
| 281 |
+
"learning_rate": 3.092209856915739e-05,
|
| 282 |
+
"loss": 1.2751,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.2227791701475912,
|
| 287 |
+
"grad_norm": 7.264173089824176,
|
| 288 |
+
"learning_rate": 3.171701112877583e-05,
|
| 289 |
+
"loss": 1.2448,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.22834864940128097,
|
| 294 |
+
"grad_norm": 7.343411468837996,
|
| 295 |
+
"learning_rate": 3.251192368839428e-05,
|
| 296 |
+
"loss": 1.1956,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.23391812865497075,
|
| 301 |
+
"grad_norm": 8.139260855658039,
|
| 302 |
+
"learning_rate": 3.330683624801272e-05,
|
| 303 |
+
"loss": 1.1626,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.23948760790866053,
|
| 308 |
+
"grad_norm": 4.441795212879553,
|
| 309 |
+
"learning_rate": 3.410174880763116e-05,
|
| 310 |
+
"loss": 1.1313,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.2450570871623503,
|
| 315 |
+
"grad_norm": 7.0655482262903115,
|
| 316 |
+
"learning_rate": 3.489666136724961e-05,
|
| 317 |
+
"loss": 1.1016,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.2506265664160401,
|
| 322 |
+
"grad_norm": 5.343797555692575,
|
| 323 |
+
"learning_rate": 3.569157392686804e-05,
|
| 324 |
+
"loss": 1.0682,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.25619604566972987,
|
| 329 |
+
"grad_norm": 4.837151670486665,
|
| 330 |
+
"learning_rate": 3.648648648648649e-05,
|
| 331 |
+
"loss": 1.0388,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.26176552492341965,
|
| 336 |
+
"grad_norm": 3.9144650767846345,
|
| 337 |
+
"learning_rate": 3.7281399046104934e-05,
|
| 338 |
+
"loss": 1.0,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.2673350041771094,
|
| 343 |
+
"grad_norm": 6.192111002424375,
|
| 344 |
+
"learning_rate": 3.8076311605723374e-05,
|
| 345 |
+
"loss": 0.9819,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.2729044834307992,
|
| 350 |
+
"grad_norm": 4.664894996941613,
|
| 351 |
+
"learning_rate": 3.8871224165341814e-05,
|
| 352 |
+
"loss": 0.9512,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.278473962684489,
|
| 357 |
+
"grad_norm": 5.728600939316023,
|
| 358 |
+
"learning_rate": 3.9666136724960254e-05,
|
| 359 |
+
"loss": 0.9292,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.28404344193817876,
|
| 364 |
+
"grad_norm": 7.358090778625861,
|
| 365 |
+
"learning_rate": 4.04610492845787e-05,
|
| 366 |
+
"loss": 0.9018,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.28961292119186854,
|
| 371 |
+
"grad_norm": 6.346761431472806,
|
| 372 |
+
"learning_rate": 4.125596184419714e-05,
|
| 373 |
+
"loss": 0.8805,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.2951824004455583,
|
| 378 |
+
"grad_norm": 3.586718245895629,
|
| 379 |
+
"learning_rate": 4.205087440381558e-05,
|
| 380 |
+
"loss": 0.8568,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.3007518796992481,
|
| 385 |
+
"grad_norm": 5.430039185536721,
|
| 386 |
+
"learning_rate": 4.284578696343403e-05,
|
| 387 |
+
"loss": 0.8284,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.3063213589529379,
|
| 392 |
+
"grad_norm": 6.430553831376545,
|
| 393 |
+
"learning_rate": 4.364069952305246e-05,
|
| 394 |
+
"loss": 0.8128,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.31189083820662766,
|
| 399 |
+
"grad_norm": 5.48500374978629,
|
| 400 |
+
"learning_rate": 4.443561208267091e-05,
|
| 401 |
+
"loss": 0.7842,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.31746031746031744,
|
| 406 |
+
"grad_norm": 5.0124287229570745,
|
| 407 |
+
"learning_rate": 4.523052464228935e-05,
|
| 408 |
+
"loss": 0.7803,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.3230297967140072,
|
| 413 |
+
"grad_norm": 4.960089415525465,
|
| 414 |
+
"learning_rate": 4.6025437201907795e-05,
|
| 415 |
+
"loss": 0.7489,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.328599275967697,
|
| 420 |
+
"grad_norm": 4.429208732066051,
|
| 421 |
+
"learning_rate": 4.6820349761526235e-05,
|
| 422 |
+
"loss": 0.7243,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.3341687552213868,
|
| 427 |
+
"grad_norm": 5.1268739971720905,
|
| 428 |
+
"learning_rate": 4.7615262321144675e-05,
|
| 429 |
+
"loss": 0.7044,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.33973823447507656,
|
| 434 |
+
"grad_norm": 4.915480885708807,
|
| 435 |
+
"learning_rate": 4.841017488076312e-05,
|
| 436 |
+
"loss": 0.6939,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.34530771372876634,
|
| 441 |
+
"grad_norm": 5.102518466488775,
|
| 442 |
+
"learning_rate": 4.9205087440381556e-05,
|
| 443 |
+
"loss": 0.6767,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.3508771929824561,
|
| 448 |
+
"grad_norm": 4.364437305811626,
|
| 449 |
+
"learning_rate": 5e-05,
|
| 450 |
+
"loss": 0.6467,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.3564466722361459,
|
| 455 |
+
"grad_norm": 4.372236216369743,
|
| 456 |
+
"learning_rate": 5.079491255961845e-05,
|
| 457 |
+
"loss": 0.6234,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.3620161514898357,
|
| 462 |
+
"grad_norm": 4.439347899189148,
|
| 463 |
+
"learning_rate": 5.158982511923689e-05,
|
| 464 |
+
"loss": 0.5961,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.36758563074352546,
|
| 469 |
+
"grad_norm": 3.1594832305447316,
|
| 470 |
+
"learning_rate": 5.2384737678855336e-05,
|
| 471 |
+
"loss": 0.5786,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.37315510999721524,
|
| 476 |
+
"grad_norm": 4.473968649061542,
|
| 477 |
+
"learning_rate": 5.317965023847376e-05,
|
| 478 |
+
"loss": 0.5711,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.378724589250905,
|
| 483 |
+
"grad_norm": 3.4873086420651456,
|
| 484 |
+
"learning_rate": 5.397456279809221e-05,
|
| 485 |
+
"loss": 0.5503,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.3842940685045948,
|
| 490 |
+
"grad_norm": 3.831386918783317,
|
| 491 |
+
"learning_rate": 5.4769475357710656e-05,
|
| 492 |
+
"loss": 0.529,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.3898635477582846,
|
| 497 |
+
"grad_norm": 3.653243174902386,
|
| 498 |
+
"learning_rate": 5.5564387917329096e-05,
|
| 499 |
+
"loss": 0.5132,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.39543302701197436,
|
| 504 |
+
"grad_norm": 3.5521167023222047,
|
| 505 |
+
"learning_rate": 5.635930047694754e-05,
|
| 506 |
+
"loss": 0.4877,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.40100250626566414,
|
| 511 |
+
"grad_norm": 3.8712402877933934,
|
| 512 |
+
"learning_rate": 5.715421303656598e-05,
|
| 513 |
+
"loss": 0.4833,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.4065719855193539,
|
| 518 |
+
"grad_norm": 3.379905907827214,
|
| 519 |
+
"learning_rate": 5.794912559618442e-05,
|
| 520 |
+
"loss": 0.4615,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.4121414647730437,
|
| 525 |
+
"grad_norm": 2.5936051096516417,
|
| 526 |
+
"learning_rate": 5.8744038155802863e-05,
|
| 527 |
+
"loss": 0.4444,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.4177109440267335,
|
| 532 |
+
"grad_norm": 3.322838857318257,
|
| 533 |
+
"learning_rate": 5.9538950715421304e-05,
|
| 534 |
+
"loss": 0.4318,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.42328042328042326,
|
| 539 |
+
"grad_norm": 3.3843216521985253,
|
| 540 |
+
"learning_rate": 6.033386327503975e-05,
|
| 541 |
+
"loss": 0.4145,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.42884990253411304,
|
| 546 |
+
"grad_norm": 3.0846350440353434,
|
| 547 |
+
"learning_rate": 6.11287758346582e-05,
|
| 548 |
+
"loss": 0.3998,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.4344193817878028,
|
| 553 |
+
"grad_norm": 2.7361994149983566,
|
| 554 |
+
"learning_rate": 6.192368839427664e-05,
|
| 555 |
+
"loss": 0.3863,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.4399888610414926,
|
| 560 |
+
"grad_norm": 3.27515261932524,
|
| 561 |
+
"learning_rate": 6.271860095389508e-05,
|
| 562 |
+
"loss": 0.375,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.4455583402951824,
|
| 567 |
+
"grad_norm": 3.029805585678618,
|
| 568 |
+
"learning_rate": 6.351351351351352e-05,
|
| 569 |
+
"loss": 0.364,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.45112781954887216,
|
| 574 |
+
"grad_norm": 2.2590750215054287,
|
| 575 |
+
"learning_rate": 6.430842607313196e-05,
|
| 576 |
+
"loss": 0.3453,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.45669729880256194,
|
| 581 |
+
"grad_norm": 3.1086383281500805,
|
| 582 |
+
"learning_rate": 6.51033386327504e-05,
|
| 583 |
+
"loss": 0.3329,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.4622667780562517,
|
| 588 |
+
"grad_norm": 2.312028278866729,
|
| 589 |
+
"learning_rate": 6.589825119236884e-05,
|
| 590 |
+
"loss": 0.324,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.4678362573099415,
|
| 595 |
+
"grad_norm": 2.8653110069877226,
|
| 596 |
+
"learning_rate": 6.669316375198729e-05,
|
| 597 |
+
"loss": 0.3113,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.4734057365636313,
|
| 602 |
+
"grad_norm": 2.0346455435163224,
|
| 603 |
+
"learning_rate": 6.748807631160573e-05,
|
| 604 |
+
"loss": 0.2976,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.47897521581732105,
|
| 609 |
+
"grad_norm": 2.2187482142761854,
|
| 610 |
+
"learning_rate": 6.828298887122417e-05,
|
| 611 |
+
"loss": 0.2838,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.48454469507101083,
|
| 616 |
+
"grad_norm": 2.5760581999823926,
|
| 617 |
+
"learning_rate": 6.907790143084261e-05,
|
| 618 |
+
"loss": 0.2712,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.4901141743247006,
|
| 623 |
+
"grad_norm": 2.336167318842384,
|
| 624 |
+
"learning_rate": 6.987281399046105e-05,
|
| 625 |
+
"loss": 0.2659,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.4956836535783904,
|
| 630 |
+
"grad_norm": 1.942101679442901,
|
| 631 |
+
"learning_rate": 7.066772655007949e-05,
|
| 632 |
+
"loss": 0.2524,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.5012531328320802,
|
| 637 |
+
"grad_norm": 1.9676135441330864,
|
| 638 |
+
"learning_rate": 7.146263910969793e-05,
|
| 639 |
+
"loss": 0.242,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.50682261208577,
|
| 644 |
+
"grad_norm": 2.42452293659568,
|
| 645 |
+
"learning_rate": 7.225755166931639e-05,
|
| 646 |
+
"loss": 0.2303,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.5123920913394597,
|
| 651 |
+
"grad_norm": 2.0689938731425754,
|
| 652 |
+
"learning_rate": 7.305246422893483e-05,
|
| 653 |
+
"loss": 0.2204,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.5179615705931495,
|
| 658 |
+
"grad_norm": 2.171275305780617,
|
| 659 |
+
"learning_rate": 7.384737678855325e-05,
|
| 660 |
+
"loss": 0.2108,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.5235310498468393,
|
| 665 |
+
"grad_norm": 2.018182012710717,
|
| 666 |
+
"learning_rate": 7.46422893481717e-05,
|
| 667 |
+
"loss": 0.1988,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.5291005291005291,
|
| 672 |
+
"grad_norm": 1.8435919023282064,
|
| 673 |
+
"learning_rate": 7.543720190779015e-05,
|
| 674 |
+
"loss": 0.1951,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.5346700083542189,
|
| 679 |
+
"grad_norm": 1.6966065845617362,
|
| 680 |
+
"learning_rate": 7.623211446740859e-05,
|
| 681 |
+
"loss": 0.1868,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.5402394876079086,
|
| 686 |
+
"grad_norm": 2.107747858389665,
|
| 687 |
+
"learning_rate": 7.702702702702703e-05,
|
| 688 |
+
"loss": 0.1752,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.5458089668615984,
|
| 693 |
+
"grad_norm": 1.626232566427141,
|
| 694 |
+
"learning_rate": 7.782193958664548e-05,
|
| 695 |
+
"loss": 0.1726,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.5513784461152882,
|
| 700 |
+
"grad_norm": 1.710543294805638,
|
| 701 |
+
"learning_rate": 7.861685214626392e-05,
|
| 702 |
+
"loss": 0.1625,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.556947925368978,
|
| 707 |
+
"grad_norm": 1.568201961955694,
|
| 708 |
+
"learning_rate": 7.941176470588235e-05,
|
| 709 |
+
"loss": 0.1552,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.5625174046226677,
|
| 714 |
+
"grad_norm": 1.4211321412129432,
|
| 715 |
+
"learning_rate": 8.02066772655008e-05,
|
| 716 |
+
"loss": 0.1485,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.5680868838763575,
|
| 721 |
+
"grad_norm": 1.4970140750834295,
|
| 722 |
+
"learning_rate": 8.100158982511924e-05,
|
| 723 |
+
"loss": 0.1434,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.5736563631300473,
|
| 728 |
+
"grad_norm": 1.853797381238809,
|
| 729 |
+
"learning_rate": 8.179650238473768e-05,
|
| 730 |
+
"loss": 0.1356,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.5792258423837371,
|
| 735 |
+
"grad_norm": 1.475678872579458,
|
| 736 |
+
"learning_rate": 8.259141494435613e-05,
|
| 737 |
+
"loss": 0.1298,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.5847953216374269,
|
| 742 |
+
"grad_norm": 1.690324091049369,
|
| 743 |
+
"learning_rate": 8.338632750397457e-05,
|
| 744 |
+
"loss": 0.1258,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.5903648008911166,
|
| 749 |
+
"grad_norm": 1.908861076416452,
|
| 750 |
+
"learning_rate": 8.418124006359301e-05,
|
| 751 |
+
"loss": 0.1249,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.5959342801448064,
|
| 756 |
+
"grad_norm": 1.236244782970751,
|
| 757 |
+
"learning_rate": 8.497615262321145e-05,
|
| 758 |
+
"loss": 0.1171,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6015037593984962,
|
| 763 |
+
"grad_norm": 1.2774164058752888,
|
| 764 |
+
"learning_rate": 8.57710651828299e-05,
|
| 765 |
+
"loss": 0.1114,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.607073238652186,
|
| 770 |
+
"grad_norm": 0.8951848955749875,
|
| 771 |
+
"learning_rate": 8.656597774244833e-05,
|
| 772 |
+
"loss": 0.1059,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.6126427179058758,
|
| 777 |
+
"grad_norm": 1.0572189677651873,
|
| 778 |
+
"learning_rate": 8.736089030206677e-05,
|
| 779 |
+
"loss": 0.1012,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.6182121971595655,
|
| 784 |
+
"grad_norm": 0.9583280956588,
|
| 785 |
+
"learning_rate": 8.815580286168523e-05,
|
| 786 |
+
"loss": 0.0973,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.6237816764132553,
|
| 791 |
+
"grad_norm": 0.966004358455493,
|
| 792 |
+
"learning_rate": 8.895071542130367e-05,
|
| 793 |
+
"loss": 0.098,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.6293511556669451,
|
| 798 |
+
"grad_norm": 0.8285031685005042,
|
| 799 |
+
"learning_rate": 8.97456279809221e-05,
|
| 800 |
+
"loss": 0.0897,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.6349206349206349,
|
| 805 |
+
"grad_norm": 1.3954391622958595,
|
| 806 |
+
"learning_rate": 9.054054054054055e-05,
|
| 807 |
+
"loss": 0.087,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.6404901141743247,
|
| 812 |
+
"grad_norm": 0.6608729564295608,
|
| 813 |
+
"learning_rate": 9.133545310015899e-05,
|
| 814 |
+
"loss": 0.0825,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.6460595934280144,
|
| 819 |
+
"grad_norm": 0.8111382270197721,
|
| 820 |
+
"learning_rate": 9.213036565977743e-05,
|
| 821 |
+
"loss": 0.0764,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.6516290726817042,
|
| 826 |
+
"grad_norm": 0.6891511752919594,
|
| 827 |
+
"learning_rate": 9.292527821939587e-05,
|
| 828 |
+
"loss": 0.0745,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.657198551935394,
|
| 833 |
+
"grad_norm": 0.8999745942051316,
|
| 834 |
+
"learning_rate": 9.372019077901432e-05,
|
| 835 |
+
"loss": 0.0718,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.6627680311890838,
|
| 840 |
+
"grad_norm": 0.7165073695545044,
|
| 841 |
+
"learning_rate": 9.451510333863276e-05,
|
| 842 |
+
"loss": 0.0699,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.6683375104427736,
|
| 847 |
+
"grad_norm": 0.5691546082373657,
|
| 848 |
+
"learning_rate": 9.531001589825119e-05,
|
| 849 |
+
"loss": 0.0663,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.6739069896964633,
|
| 854 |
+
"grad_norm": 0.618154476760873,
|
| 855 |
+
"learning_rate": 9.610492845786964e-05,
|
| 856 |
+
"loss": 0.065,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.6794764689501531,
|
| 861 |
+
"grad_norm": 0.5460710019961567,
|
| 862 |
+
"learning_rate": 9.689984101748808e-05,
|
| 863 |
+
"loss": 0.0622,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.6850459482038429,
|
| 868 |
+
"grad_norm": 0.44149906211171397,
|
| 869 |
+
"learning_rate": 9.769475357710652e-05,
|
| 870 |
+
"loss": 0.0582,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.6906154274575327,
|
| 875 |
+
"grad_norm": 0.5860441580160569,
|
| 876 |
+
"learning_rate": 9.848966613672496e-05,
|
| 877 |
+
"loss": 0.0567,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.6961849067112225,
|
| 882 |
+
"grad_norm": 0.8453684625264357,
|
| 883 |
+
"learning_rate": 9.928457869634342e-05,
|
| 884 |
+
"loss": 0.0564,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.7017543859649122,
|
| 889 |
+
"grad_norm": 2.1379268435693555,
|
| 890 |
+
"learning_rate": 9.999999807244224e-05,
|
| 891 |
+
"loss": 0.0564,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.707323865218602,
|
| 896 |
+
"grad_norm": 3.3074045726639083,
|
| 897 |
+
"learning_rate": 9.999976676568994e-05,
|
| 898 |
+
"loss": 0.1234,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.7128933444722918,
|
| 903 |
+
"grad_norm": 0.6698300987807628,
|
| 904 |
+
"learning_rate": 9.99991499494276e-05,
|
| 905 |
+
"loss": 0.0776,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.7184628237259816,
|
| 910 |
+
"grad_norm": 0.40986381929677007,
|
| 911 |
+
"learning_rate": 9.999814762841102e-05,
|
| 912 |
+
"loss": 0.0563,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.7240323029796714,
|
| 917 |
+
"grad_norm": 0.5823587617428915,
|
| 918 |
+
"learning_rate": 9.999675981036832e-05,
|
| 919 |
+
"loss": 0.05,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.7296017822333611,
|
| 924 |
+
"grad_norm": 0.36554660320089016,
|
| 925 |
+
"learning_rate": 9.999498650599987e-05,
|
| 926 |
+
"loss": 0.0505,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.7351712614870509,
|
| 931 |
+
"grad_norm": 0.3490628453059406,
|
| 932 |
+
"learning_rate": 9.999282772897828e-05,
|
| 933 |
+
"loss": 0.0474,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.7407407407407407,
|
| 938 |
+
"grad_norm": 0.3963213598340753,
|
| 939 |
+
"learning_rate": 9.99902834959482e-05,
|
| 940 |
+
"loss": 0.0462,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.7463102199944305,
|
| 945 |
+
"grad_norm": 0.5565743717844861,
|
| 946 |
+
"learning_rate": 9.998735382652624e-05,
|
| 947 |
+
"loss": 0.0442,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.7518796992481203,
|
| 952 |
+
"grad_norm": 0.29567277754356086,
|
| 953 |
+
"learning_rate": 9.99840387433008e-05,
|
| 954 |
+
"loss": 0.0439,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.75744917850181,
|
| 959 |
+
"grad_norm": 0.42840154636629013,
|
| 960 |
+
"learning_rate": 9.998033827183193e-05,
|
| 961 |
+
"loss": 0.0432,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.7630186577554998,
|
| 966 |
+
"grad_norm": 0.3843192755820257,
|
| 967 |
+
"learning_rate": 9.997625244065113e-05,
|
| 968 |
+
"loss": 0.0419,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.7685881370091896,
|
| 973 |
+
"grad_norm": 0.2794549966082612,
|
| 974 |
+
"learning_rate": 9.997178128126101e-05,
|
| 975 |
+
"loss": 0.04,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.7741576162628794,
|
| 980 |
+
"grad_norm": 0.3169631435418986,
|
| 981 |
+
"learning_rate": 9.99669248281353e-05,
|
| 982 |
+
"loss": 0.0395,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.7797270955165692,
|
| 987 |
+
"grad_norm": 0.2751145242333607,
|
| 988 |
+
"learning_rate": 9.996168311871832e-05,
|
| 989 |
+
"loss": 0.039,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.7852965747702589,
|
| 994 |
+
"grad_norm": 0.2873246557323452,
|
| 995 |
+
"learning_rate": 9.995605619342484e-05,
|
| 996 |
+
"loss": 0.0386,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.7908660540239487,
|
| 1001 |
+
"grad_norm": 0.26158401303739276,
|
| 1002 |
+
"learning_rate": 9.995004409563969e-05,
|
| 1003 |
+
"loss": 0.0364,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.7964355332776385,
|
| 1008 |
+
"grad_norm": 0.3512702948952268,
|
| 1009 |
+
"learning_rate": 9.994364687171756e-05,
|
| 1010 |
+
"loss": 0.0358,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.8020050125313283,
|
| 1015 |
+
"grad_norm": 0.30233442060518095,
|
| 1016 |
+
"learning_rate": 9.993686457098246e-05,
|
| 1017 |
+
"loss": 0.0355,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.8075744917850181,
|
| 1022 |
+
"grad_norm": 0.31354227279951397,
|
| 1023 |
+
"learning_rate": 9.992969724572746e-05,
|
| 1024 |
+
"loss": 0.0355,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.8131439710387078,
|
| 1029 |
+
"grad_norm": 0.257194967469868,
|
| 1030 |
+
"learning_rate": 9.992214495121428e-05,
|
| 1031 |
+
"loss": 0.0357,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.8187134502923976,
|
| 1036 |
+
"grad_norm": 0.3479392485987129,
|
| 1037 |
+
"learning_rate": 9.99142077456728e-05,
|
| 1038 |
+
"loss": 0.0345,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.8242829295460874,
|
| 1043 |
+
"grad_norm": 0.24684398703549518,
|
| 1044 |
+
"learning_rate": 9.990588569030068e-05,
|
| 1045 |
+
"loss": 0.0337,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.8298524087997772,
|
| 1050 |
+
"grad_norm": 0.23435105711158805,
|
| 1051 |
+
"learning_rate": 9.989717884926286e-05,
|
| 1052 |
+
"loss": 0.0328,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.835421888053467,
|
| 1057 |
+
"grad_norm": 0.226826410374761,
|
| 1058 |
+
"learning_rate": 9.988808728969104e-05,
|
| 1059 |
+
"loss": 0.0325,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.8409913673071567,
|
| 1064 |
+
"grad_norm": 0.23919961520195954,
|
| 1065 |
+
"learning_rate": 9.987861108168317e-05,
|
| 1066 |
+
"loss": 0.0319,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.8465608465608465,
|
| 1071 |
+
"grad_norm": 0.262516092950579,
|
| 1072 |
+
"learning_rate": 9.986875029830302e-05,
|
| 1073 |
+
"loss": 0.0311,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.8521303258145363,
|
| 1078 |
+
"grad_norm": 0.27505170240477905,
|
| 1079 |
+
"learning_rate": 9.985850501557941e-05,
|
| 1080 |
+
"loss": 0.0306,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.8576998050682261,
|
| 1085 |
+
"grad_norm": 0.2136353900303048,
|
| 1086 |
+
"learning_rate": 9.984787531250581e-05,
|
| 1087 |
+
"loss": 0.0306,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.8632692843219159,
|
| 1092 |
+
"grad_norm": 0.2681884993661502,
|
| 1093 |
+
"learning_rate": 9.983686127103963e-05,
|
| 1094 |
+
"loss": 0.0304,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.8688387635756056,
|
| 1099 |
+
"grad_norm": 0.27064206276631336,
|
| 1100 |
+
"learning_rate": 9.982546297610163e-05,
|
| 1101 |
+
"loss": 0.0302,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.8744082428292954,
|
| 1106 |
+
"grad_norm": 0.22756357273919053,
|
| 1107 |
+
"learning_rate": 9.981368051557522e-05,
|
| 1108 |
+
"loss": 0.0296,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.8799777220829852,
|
| 1113 |
+
"grad_norm": 0.23817203224591024,
|
| 1114 |
+
"learning_rate": 9.980151398030586e-05,
|
| 1115 |
+
"loss": 0.029,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.885547201336675,
|
| 1120 |
+
"grad_norm": 0.23612451468006537,
|
| 1121 |
+
"learning_rate": 9.978896346410025e-05,
|
| 1122 |
+
"loss": 0.0283,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.8911166805903648,
|
| 1127 |
+
"grad_norm": 0.21633179926653082,
|
| 1128 |
+
"learning_rate": 9.977602906372577e-05,
|
| 1129 |
+
"loss": 0.0284,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.8966861598440545,
|
| 1134 |
+
"grad_norm": 0.3007834433966076,
|
| 1135 |
+
"learning_rate": 9.97627108789095e-05,
|
| 1136 |
+
"loss": 0.0276,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.9022556390977443,
|
| 1141 |
+
"grad_norm": 0.22874451896032774,
|
| 1142 |
+
"learning_rate": 9.974900901233772e-05,
|
| 1143 |
+
"loss": 0.0275,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.9078251183514341,
|
| 1148 |
+
"grad_norm": 0.21229440559613427,
|
| 1149 |
+
"learning_rate": 9.973492356965488e-05,
|
| 1150 |
+
"loss": 0.0272,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.9133945976051239,
|
| 1155 |
+
"grad_norm": 0.21316799793517352,
|
| 1156 |
+
"learning_rate": 9.972045465946295e-05,
|
| 1157 |
+
"loss": 0.0266,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.9189640768588136,
|
| 1162 |
+
"grad_norm": 0.21840470782406107,
|
| 1163 |
+
"learning_rate": 9.97056023933205e-05,
|
| 1164 |
+
"loss": 0.0262,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.9245335561125034,
|
| 1169 |
+
"grad_norm": 0.2296202067200069,
|
| 1170 |
+
"learning_rate": 9.969036688574184e-05,
|
| 1171 |
+
"loss": 0.0262,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.9301030353661932,
|
| 1176 |
+
"grad_norm": 0.20035916457466751,
|
| 1177 |
+
"learning_rate": 9.967474825419619e-05,
|
| 1178 |
+
"loss": 0.0256,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.935672514619883,
|
| 1183 |
+
"grad_norm": 0.2392722746437584,
|
| 1184 |
+
"learning_rate": 9.965874661910675e-05,
|
| 1185 |
+
"loss": 0.0252,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.9412419938735728,
|
| 1190 |
+
"grad_norm": 0.2373750894769379,
|
| 1191 |
+
"learning_rate": 9.964236210384972e-05,
|
| 1192 |
+
"loss": 0.025,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.9468114731272625,
|
| 1197 |
+
"grad_norm": 0.23565128407439787,
|
| 1198 |
+
"learning_rate": 9.962559483475342e-05,
|
| 1199 |
+
"loss": 0.0245,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.9523809523809523,
|
| 1204 |
+
"grad_norm": 0.17651355447696596,
|
| 1205 |
+
"learning_rate": 9.960844494109731e-05,
|
| 1206 |
+
"loss": 0.024,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.9579504316346421,
|
| 1211 |
+
"grad_norm": 0.19685563495858988,
|
| 1212 |
+
"learning_rate": 9.959091255511093e-05,
|
| 1213 |
+
"loss": 0.0237,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.9635199108883319,
|
| 1218 |
+
"grad_norm": 0.1904961294977808,
|
| 1219 |
+
"learning_rate": 9.957299781197295e-05,
|
| 1220 |
+
"loss": 0.0235,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.9690893901420217,
|
| 1225 |
+
"grad_norm": 0.21714296341973274,
|
| 1226 |
+
"learning_rate": 9.955470084981009e-05,
|
| 1227 |
+
"loss": 0.0232,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.9746588693957114,
|
| 1232 |
+
"grad_norm": 0.21953603915286493,
|
| 1233 |
+
"learning_rate": 9.953602180969606e-05,
|
| 1234 |
+
"loss": 0.0227,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.9802283486494012,
|
| 1239 |
+
"grad_norm": 0.21497187075207344,
|
| 1240 |
+
"learning_rate": 9.951696083565048e-05,
|
| 1241 |
+
"loss": 0.0224,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.985797827903091,
|
| 1246 |
+
"grad_norm": 0.18608055976176674,
|
| 1247 |
+
"learning_rate": 9.949751807463777e-05,
|
| 1248 |
+
"loss": 0.0226,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.9913673071567808,
|
| 1253 |
+
"grad_norm": 0.20352070118783105,
|
| 1254 |
+
"learning_rate": 9.947769367656603e-05,
|
| 1255 |
+
"loss": 0.0223,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.9969367864104706,
|
| 1260 |
+
"grad_norm": 0.24791255800751252,
|
| 1261 |
+
"learning_rate": 9.945748779428585e-05,
|
| 1262 |
+
"loss": 0.0217,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.002227791701476,
|
| 1267 |
+
"grad_norm": 0.22472692732126306,
|
| 1268 |
+
"learning_rate": 9.943690058358915e-05,
|
| 1269 |
+
"loss": 0.0217,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
}
|
| 1272 |
+
],
|
| 1273 |
+
"logging_steps": 10,
|
| 1274 |
+
"max_steps": 12572,
|
| 1275 |
+
"num_input_tokens_seen": 0,
|
| 1276 |
+
"num_train_epochs": 7,
|
| 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": 149526216179712.0,
|
| 1291 |
+
"train_batch_size": 128,
|
| 1292 |
+
"trial_name": null,
|
| 1293 |
+
"trial_params": null
|
| 1294 |
+
}
|