Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1309 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1309 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.751552795031056,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1812,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.020703933747412008,
|
| 13 |
+
"grad_norm": 4.777360916137695,
|
| 14 |
+
"learning_rate": 2.1978021978021977e-05,
|
| 15 |
+
"loss": 0.904,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.041407867494824016,
|
| 20 |
+
"grad_norm": 4.69223165512085,
|
| 21 |
+
"learning_rate": 4.3956043956043955e-05,
|
| 22 |
+
"loss": 0.4989,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.062111801242236024,
|
| 27 |
+
"grad_norm": 2.8304860591888428,
|
| 28 |
+
"learning_rate": 6.593406593406594e-05,
|
| 29 |
+
"loss": 0.3043,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.08281573498964803,
|
| 34 |
+
"grad_norm": 1.7506803274154663,
|
| 35 |
+
"learning_rate": 8.791208791208791e-05,
|
| 36 |
+
"loss": 0.2144,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.10351966873706005,
|
| 41 |
+
"grad_norm": 1.623117208480835,
|
| 42 |
+
"learning_rate": 0.0001098901098901099,
|
| 43 |
+
"loss": 0.1884,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.12422360248447205,
|
| 48 |
+
"grad_norm": 1.1169493198394775,
|
| 49 |
+
"learning_rate": 0.00013186813186813188,
|
| 50 |
+
"loss": 0.1838,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.14492753623188406,
|
| 55 |
+
"grad_norm": 1.1414587497711182,
|
| 56 |
+
"learning_rate": 0.00015384615384615385,
|
| 57 |
+
"loss": 0.1498,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.16563146997929606,
|
| 62 |
+
"grad_norm": 1.1636276245117188,
|
| 63 |
+
"learning_rate": 0.00017582417582417582,
|
| 64 |
+
"loss": 0.1389,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.18633540372670807,
|
| 69 |
+
"grad_norm": 1.0473010540008545,
|
| 70 |
+
"learning_rate": 0.0001978021978021978,
|
| 71 |
+
"loss": 0.1387,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.2070393374741201,
|
| 76 |
+
"grad_norm": 1.6162792444229126,
|
| 77 |
+
"learning_rate": 0.00019998650467774733,
|
| 78 |
+
"loss": 0.1511,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.2277432712215321,
|
| 83 |
+
"grad_norm": 0.8324443697929382,
|
| 84 |
+
"learning_rate": 0.00019993985885743938,
|
| 85 |
+
"loss": 0.1233,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.2484472049689441,
|
| 90 |
+
"grad_norm": 1.0195481777191162,
|
| 91 |
+
"learning_rate": 0.00019985991146959537,
|
| 92 |
+
"loss": 0.1225,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.2691511387163561,
|
| 97 |
+
"grad_norm": 0.9015069007873535,
|
| 98 |
+
"learning_rate": 0.0001997466891539701,
|
| 99 |
+
"loss": 0.0931,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.2898550724637681,
|
| 104 |
+
"grad_norm": 1.1169915199279785,
|
| 105 |
+
"learning_rate": 0.00019960022963805922,
|
| 106 |
+
"loss": 0.1122,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.3105590062111801,
|
| 111 |
+
"grad_norm": 1.5248440504074097,
|
| 112 |
+
"learning_rate": 0.00019942058172452783,
|
| 113 |
+
"loss": 0.1251,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.33126293995859213,
|
| 118 |
+
"grad_norm": 1.40325927734375,
|
| 119 |
+
"learning_rate": 0.0001992078052749486,
|
| 120 |
+
"loss": 0.103,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.35196687370600416,
|
| 125 |
+
"grad_norm": 0.6704475283622742,
|
| 126 |
+
"learning_rate": 0.000198961971189855,
|
| 127 |
+
"loss": 0.098,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.37267080745341613,
|
| 132 |
+
"grad_norm": 0.8075087070465088,
|
| 133 |
+
"learning_rate": 0.00019868316138511603,
|
| 134 |
+
"loss": 0.091,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.39337474120082816,
|
| 139 |
+
"grad_norm": 0.5290576815605164,
|
| 140 |
+
"learning_rate": 0.00019837146876464058,
|
| 141 |
+
"loss": 0.0932,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.4140786749482402,
|
| 146 |
+
"grad_norm": 1.1313437223434448,
|
| 147 |
+
"learning_rate": 0.00019802699718942033,
|
| 148 |
+
"loss": 0.1002,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.43478260869565216,
|
| 153 |
+
"grad_norm": 0.7324339747428894,
|
| 154 |
+
"learning_rate": 0.00019764986144292165,
|
| 155 |
+
"loss": 0.093,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.4554865424430642,
|
| 160 |
+
"grad_norm": 0.9365309476852417,
|
| 161 |
+
"learning_rate": 0.0001972401871928379,
|
| 162 |
+
"loss": 0.0906,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.47619047619047616,
|
| 167 |
+
"grad_norm": 1.0107965469360352,
|
| 168 |
+
"learning_rate": 0.00019679811094921478,
|
| 169 |
+
"loss": 0.0895,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4968944099378882,
|
| 174 |
+
"grad_norm": 0.5845835208892822,
|
| 175 |
+
"learning_rate": 0.0001963237800189632,
|
| 176 |
+
"loss": 0.0886,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.5175983436853002,
|
| 181 |
+
"grad_norm": 0.8239277601242065,
|
| 182 |
+
"learning_rate": 0.00019581735245677412,
|
| 183 |
+
"loss": 0.0795,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.5383022774327122,
|
| 188 |
+
"grad_norm": 0.541282057762146,
|
| 189 |
+
"learning_rate": 0.00019527899701245232,
|
| 190 |
+
"loss": 0.0801,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.5590062111801242,
|
| 195 |
+
"grad_norm": 0.5599911212921143,
|
| 196 |
+
"learning_rate": 0.00019470889307468601,
|
| 197 |
+
"loss": 0.0755,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5797101449275363,
|
| 202 |
+
"grad_norm": 1.1068004369735718,
|
| 203 |
+
"learning_rate": 0.00019410723061127186,
|
| 204 |
+
"loss": 0.0922,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.6004140786749482,
|
| 209 |
+
"grad_norm": 0.5060113072395325,
|
| 210 |
+
"learning_rate": 0.00019347421010581465,
|
| 211 |
+
"loss": 0.0656,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.6211180124223602,
|
| 216 |
+
"grad_norm": 0.591389000415802,
|
| 217 |
+
"learning_rate": 0.00019281004249092272,
|
| 218 |
+
"loss": 0.0735,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.6418219461697723,
|
| 223 |
+
"grad_norm": 0.700982391834259,
|
| 224 |
+
"learning_rate": 0.0001921149490779221,
|
| 225 |
+
"loss": 0.0785,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.6625258799171843,
|
| 230 |
+
"grad_norm": 0.6046485304832458,
|
| 231 |
+
"learning_rate": 0.00019138916148311186,
|
| 232 |
+
"loss": 0.0837,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.6832298136645962,
|
| 237 |
+
"grad_norm": 0.8097739219665527,
|
| 238 |
+
"learning_rate": 0.00019063292155058586,
|
| 239 |
+
"loss": 0.0818,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.7039337474120083,
|
| 244 |
+
"grad_norm": 0.9329370856285095,
|
| 245 |
+
"learning_rate": 0.00018984648127164637,
|
| 246 |
+
"loss": 0.0755,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.7246376811594203,
|
| 251 |
+
"grad_norm": 0.3276360332965851,
|
| 252 |
+
"learning_rate": 0.00018903010270083663,
|
| 253 |
+
"loss": 0.0747,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.7453416149068323,
|
| 258 |
+
"grad_norm": 0.48511287569999695,
|
| 259 |
+
"learning_rate": 0.00018818405786861996,
|
| 260 |
+
"loss": 0.0816,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.7660455486542443,
|
| 265 |
+
"grad_norm": 0.4654783308506012,
|
| 266 |
+
"learning_rate": 0.00018730862869073493,
|
| 267 |
+
"loss": 0.0579,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.7867494824016563,
|
| 272 |
+
"grad_norm": 0.7603228688240051,
|
| 273 |
+
"learning_rate": 0.0001864041068742564,
|
| 274 |
+
"loss": 0.0671,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.8074534161490683,
|
| 279 |
+
"grad_norm": 0.7368313074111938,
|
| 280 |
+
"learning_rate": 0.00018547079382039407,
|
| 281 |
+
"loss": 0.0732,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.8281573498964804,
|
| 286 |
+
"grad_norm": 0.6245520710945129,
|
| 287 |
+
"learning_rate": 0.00018450900052406085,
|
| 288 |
+
"loss": 0.067,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.8488612836438924,
|
| 293 |
+
"grad_norm": 0.5184304118156433,
|
| 294 |
+
"learning_rate": 0.00018351904747024414,
|
| 295 |
+
"loss": 0.067,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.8695652173913043,
|
| 300 |
+
"grad_norm": 0.7440361976623535,
|
| 301 |
+
"learning_rate": 0.00018250126452721527,
|
| 302 |
+
"loss": 0.0713,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.8902691511387164,
|
| 307 |
+
"grad_norm": 0.5410072803497314,
|
| 308 |
+
"learning_rate": 0.0001814559908366121,
|
| 309 |
+
"loss": 0.0763,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.9109730848861284,
|
| 314 |
+
"grad_norm": 0.6151822209358215,
|
| 315 |
+
"learning_rate": 0.00018038357470043145,
|
| 316 |
+
"loss": 0.0733,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.9316770186335404,
|
| 321 |
+
"grad_norm": 0.5519956946372986,
|
| 322 |
+
"learning_rate": 0.00017928437346496922,
|
| 323 |
+
"loss": 0.0774,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.9523809523809523,
|
| 328 |
+
"grad_norm": 0.7974098324775696,
|
| 329 |
+
"learning_rate": 0.0001781587534017472,
|
| 330 |
+
"loss": 0.0582,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.9730848861283644,
|
| 335 |
+
"grad_norm": 0.5893744230270386,
|
| 336 |
+
"learning_rate": 0.00017700708958546507,
|
| 337 |
+
"loss": 0.0726,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.9937888198757764,
|
| 342 |
+
"grad_norm": 1.881540298461914,
|
| 343 |
+
"learning_rate": 0.00017582976576901982,
|
| 344 |
+
"loss": 0.0597,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.0144927536231885,
|
| 349 |
+
"grad_norm": 0.5968530774116516,
|
| 350 |
+
"learning_rate": 0.00017462717425563288,
|
| 351 |
+
"loss": 0.0626,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.0351966873706004,
|
| 356 |
+
"grad_norm": 0.5973192453384399,
|
| 357 |
+
"learning_rate": 0.00017339971576812863,
|
| 358 |
+
"loss": 0.0517,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.0559006211180124,
|
| 363 |
+
"grad_norm": 0.7765334844589233,
|
| 364 |
+
"learning_rate": 0.00017214779931540682,
|
| 365 |
+
"loss": 0.0599,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.0766045548654244,
|
| 370 |
+
"grad_norm": 0.476686030626297,
|
| 371 |
+
"learning_rate": 0.0001708718420561545,
|
| 372 |
+
"loss": 0.0511,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.0973084886128364,
|
| 377 |
+
"grad_norm": 0.8050920963287354,
|
| 378 |
+
"learning_rate": 0.00016957226915984184,
|
| 379 |
+
"loss": 0.055,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.1180124223602483,
|
| 384 |
+
"grad_norm": 0.5900657773017883,
|
| 385 |
+
"learning_rate": 0.00016824951366504895,
|
| 386 |
+
"loss": 0.0595,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.1387163561076605,
|
| 391 |
+
"grad_norm": 0.7448649406433105,
|
| 392 |
+
"learning_rate": 0.00016690401633517039,
|
| 393 |
+
"loss": 0.0524,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.1594202898550725,
|
| 398 |
+
"grad_norm": 0.41742566227912903,
|
| 399 |
+
"learning_rate": 0.00016553622551154578,
|
| 400 |
+
"loss": 0.0662,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.1801242236024845,
|
| 405 |
+
"grad_norm": 0.6216546297073364,
|
| 406 |
+
"learning_rate": 0.0001641465969640652,
|
| 407 |
+
"loss": 0.0663,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.2008281573498965,
|
| 412 |
+
"grad_norm": 0.5282993912696838,
|
| 413 |
+
"learning_rate": 0.00016273559373929946,
|
| 414 |
+
"loss": 0.0693,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.2215320910973084,
|
| 419 |
+
"grad_norm": 0.5407345294952393,
|
| 420 |
+
"learning_rate": 0.00016130368600620528,
|
| 421 |
+
"loss": 0.0537,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.2422360248447206,
|
| 426 |
+
"grad_norm": 0.4532129764556885,
|
| 427 |
+
"learning_rate": 0.00015985135089945774,
|
| 428 |
+
"loss": 0.0626,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.2629399585921326,
|
| 433 |
+
"grad_norm": 0.45578986406326294,
|
| 434 |
+
"learning_rate": 0.00015837907236046115,
|
| 435 |
+
"loss": 0.0687,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.2836438923395446,
|
| 440 |
+
"grad_norm": 0.5343713164329529,
|
| 441 |
+
"learning_rate": 0.00015688734097609204,
|
| 442 |
+
"loss": 0.06,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.3043478260869565,
|
| 447 |
+
"grad_norm": 0.4965466260910034,
|
| 448 |
+
"learning_rate": 0.00015537665381522775,
|
| 449 |
+
"loss": 0.0591,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.3250517598343685,
|
| 454 |
+
"grad_norm": 0.4257121682167053,
|
| 455 |
+
"learning_rate": 0.00015384751426311504,
|
| 456 |
+
"loss": 0.0501,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.3457556935817805,
|
| 461 |
+
"grad_norm": 0.3051788806915283,
|
| 462 |
+
"learning_rate": 0.00015230043185363409,
|
| 463 |
+
"loss": 0.0466,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.3664596273291925,
|
| 468 |
+
"grad_norm": 0.4974876344203949,
|
| 469 |
+
"learning_rate": 0.00015073592209951354,
|
| 470 |
+
"loss": 0.0504,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.3871635610766044,
|
| 475 |
+
"grad_norm": 0.5405896902084351,
|
| 476 |
+
"learning_rate": 0.0001491545063205535,
|
| 477 |
+
"loss": 0.039,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.4078674948240166,
|
| 482 |
+
"grad_norm": 0.5232389569282532,
|
| 483 |
+
"learning_rate": 0.00014755671146991325,
|
| 484 |
+
"loss": 0.0529,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.4285714285714286,
|
| 489 |
+
"grad_norm": 0.630558431148529,
|
| 490 |
+
"learning_rate": 0.00014594306995852207,
|
| 491 |
+
"loss": 0.0422,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.4492753623188406,
|
| 496 |
+
"grad_norm": 0.49374642968177795,
|
| 497 |
+
"learning_rate": 0.00014431411947767137,
|
| 498 |
+
"loss": 0.0467,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.4699792960662525,
|
| 503 |
+
"grad_norm": 0.33680301904678345,
|
| 504 |
+
"learning_rate": 0.00014267040281984726,
|
| 505 |
+
"loss": 0.0468,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.4906832298136645,
|
| 510 |
+
"grad_norm": 0.6091966032981873,
|
| 511 |
+
"learning_rate": 0.0001410124676978634,
|
| 512 |
+
"loss": 0.0532,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.5113871635610767,
|
| 517 |
+
"grad_norm": 0.3792521059513092,
|
| 518 |
+
"learning_rate": 0.00013934086656235424,
|
| 519 |
+
"loss": 0.0527,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.5320910973084887,
|
| 524 |
+
"grad_norm": 0.7761206030845642,
|
| 525 |
+
"learning_rate": 0.00013765615641768964,
|
| 526 |
+
"loss": 0.0431,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.5527950310559007,
|
| 531 |
+
"grad_norm": 0.5704333782196045,
|
| 532 |
+
"learning_rate": 0.0001359588986363721,
|
| 533 |
+
"loss": 0.0445,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.5734989648033126,
|
| 538 |
+
"grad_norm": 0.442277729511261,
|
| 539 |
+
"learning_rate": 0.0001342496587719783,
|
| 540 |
+
"loss": 0.0514,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.5942028985507246,
|
| 545 |
+
"grad_norm": 0.48535916209220886,
|
| 546 |
+
"learning_rate": 0.00013252900637070755,
|
| 547 |
+
"loss": 0.0471,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.6149068322981366,
|
| 552 |
+
"grad_norm": 0.3780505359172821,
|
| 553 |
+
"learning_rate": 0.00013079751478160007,
|
| 554 |
+
"loss": 0.0479,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.6356107660455486,
|
| 559 |
+
"grad_norm": 0.42950040102005005,
|
| 560 |
+
"learning_rate": 0.00012905576096548758,
|
| 561 |
+
"loss": 0.0358,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.6563146997929605,
|
| 566 |
+
"grad_norm": 0.3445834517478943,
|
| 567 |
+
"learning_rate": 0.00012730432530274087,
|
| 568 |
+
"loss": 0.0442,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.6770186335403725,
|
| 573 |
+
"grad_norm": 0.28706836700439453,
|
| 574 |
+
"learning_rate": 0.00012554379139987764,
|
| 575 |
+
"loss": 0.0426,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.6977225672877847,
|
| 580 |
+
"grad_norm": 0.40510639548301697,
|
| 581 |
+
"learning_rate": 0.00012377474589509547,
|
| 582 |
+
"loss": 0.0559,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.7184265010351967,
|
| 587 |
+
"grad_norm": 0.4862560033798218,
|
| 588 |
+
"learning_rate": 0.00012199777826279448,
|
| 589 |
+
"loss": 0.0519,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.7391304347826086,
|
| 594 |
+
"grad_norm": 0.7009541392326355,
|
| 595 |
+
"learning_rate": 0.00012021348061715516,
|
| 596 |
+
"loss": 0.0385,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.7598343685300208,
|
| 601 |
+
"grad_norm": 0.5319027900695801,
|
| 602 |
+
"learning_rate": 0.00011842244751483609,
|
| 603 |
+
"loss": 0.0429,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.7805383022774328,
|
| 608 |
+
"grad_norm": 0.3376663029193878,
|
| 609 |
+
"learning_rate": 0.00011662527575685842,
|
| 610 |
+
"loss": 0.0463,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.8012422360248448,
|
| 615 |
+
"grad_norm": 0.5521535873413086,
|
| 616 |
+
"learning_rate": 0.00011482256418974204,
|
| 617 |
+
"loss": 0.0456,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.8219461697722568,
|
| 622 |
+
"grad_norm": 0.5131340622901917,
|
| 623 |
+
"learning_rate": 0.0001130149135059602,
|
| 624 |
+
"loss": 0.0456,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.8426501035196687,
|
| 629 |
+
"grad_norm": 0.4025464355945587,
|
| 630 |
+
"learning_rate": 0.00011120292604377938,
|
| 631 |
+
"loss": 0.0433,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.8633540372670807,
|
| 636 |
+
"grad_norm": 0.5452845692634583,
|
| 637 |
+
"learning_rate": 0.00010938720558655016,
|
| 638 |
+
"loss": 0.0446,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.8840579710144927,
|
| 643 |
+
"grad_norm": 0.3350236117839813,
|
| 644 |
+
"learning_rate": 0.00010756835716151722,
|
| 645 |
+
"loss": 0.039,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.9047619047619047,
|
| 650 |
+
"grad_norm": 0.48820528388023376,
|
| 651 |
+
"learning_rate": 0.00010574698683821434,
|
| 652 |
+
"loss": 0.0488,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.9254658385093166,
|
| 657 |
+
"grad_norm": 0.6648292541503906,
|
| 658 |
+
"learning_rate": 0.00010392370152651234,
|
| 659 |
+
"loss": 0.0605,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.9461697722567288,
|
| 664 |
+
"grad_norm": 0.38091030716896057,
|
| 665 |
+
"learning_rate": 0.00010209910877438696,
|
| 666 |
+
"loss": 0.037,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.9668737060041408,
|
| 671 |
+
"grad_norm": 0.25008946657180786,
|
| 672 |
+
"learning_rate": 0.00010027381656547413,
|
| 673 |
+
"loss": 0.0416,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.9875776397515528,
|
| 678 |
+
"grad_norm": 0.45538610219955444,
|
| 679 |
+
"learning_rate": 9.844843311647994e-05,
|
| 680 |
+
"loss": 0.0444,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.008281573498965,
|
| 685 |
+
"grad_norm": 0.5605445504188538,
|
| 686 |
+
"learning_rate": 9.662356667451336e-05,
|
| 687 |
+
"loss": 0.037,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.028985507246377,
|
| 692 |
+
"grad_norm": 0.3311709463596344,
|
| 693 |
+
"learning_rate": 9.479982531440797e-05,
|
| 694 |
+
"loss": 0.0358,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.049689440993789,
|
| 699 |
+
"grad_norm": 0.4076383709907532,
|
| 700 |
+
"learning_rate": 9.297781673610235e-05,
|
| 701 |
+
"loss": 0.0336,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.070393374741201,
|
| 706 |
+
"grad_norm": 0.33418402075767517,
|
| 707 |
+
"learning_rate": 9.115814806214402e-05,
|
| 708 |
+
"loss": 0.0359,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.091097308488613,
|
| 713 |
+
"grad_norm": 0.3592430651187897,
|
| 714 |
+
"learning_rate": 8.934142563538699e-05,
|
| 715 |
+
"loss": 0.0456,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.111801242236025,
|
| 720 |
+
"grad_norm": 0.5789178609848022,
|
| 721 |
+
"learning_rate": 8.752825481694813e-05,
|
| 722 |
+
"loss": 0.0327,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.132505175983437,
|
| 727 |
+
"grad_norm": 0.401697039604187,
|
| 728 |
+
"learning_rate": 8.571923978449156e-05,
|
| 729 |
+
"loss": 0.0347,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.153209109730849,
|
| 734 |
+
"grad_norm": 0.3946631848812103,
|
| 735 |
+
"learning_rate": 8.391498333090647e-05,
|
| 736 |
+
"loss": 0.0336,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.1739130434782608,
|
| 741 |
+
"grad_norm": 0.3135606050491333,
|
| 742 |
+
"learning_rate": 8.211608666344732e-05,
|
| 743 |
+
"loss": 0.0383,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.1946169772256727,
|
| 748 |
+
"grad_norm": 0.4405308663845062,
|
| 749 |
+
"learning_rate": 8.032314920340138e-05,
|
| 750 |
+
"loss": 0.0435,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 2.2153209109730847,
|
| 755 |
+
"grad_norm": 0.30584511160850525,
|
| 756 |
+
"learning_rate": 7.85367683863525e-05,
|
| 757 |
+
"loss": 0.0294,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.2360248447204967,
|
| 762 |
+
"grad_norm": 0.2619006633758545,
|
| 763 |
+
"learning_rate": 7.675753946310549e-05,
|
| 764 |
+
"loss": 0.0356,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 2.256728778467909,
|
| 769 |
+
"grad_norm": 0.4563602805137634,
|
| 770 |
+
"learning_rate": 7.498605530133926e-05,
|
| 771 |
+
"loss": 0.0325,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.277432712215321,
|
| 776 |
+
"grad_norm": 0.23483984172344208,
|
| 777 |
+
"learning_rate": 7.322290618805367e-05,
|
| 778 |
+
"loss": 0.0297,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 2.298136645962733,
|
| 783 |
+
"grad_norm": 0.2909756302833557,
|
| 784 |
+
"learning_rate": 7.146867963287647e-05,
|
| 785 |
+
"loss": 0.0303,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.318840579710145,
|
| 790 |
+
"grad_norm": 0.2771686613559723,
|
| 791 |
+
"learning_rate": 6.972396017229555e-05,
|
| 792 |
+
"loss": 0.0356,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.339544513457557,
|
| 797 |
+
"grad_norm": 0.2620188891887665,
|
| 798 |
+
"learning_rate": 6.798932917488214e-05,
|
| 799 |
+
"loss": 0.0408,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.360248447204969,
|
| 804 |
+
"grad_norm": 0.43468013405799866,
|
| 805 |
+
"learning_rate": 6.626536464756928e-05,
|
| 806 |
+
"loss": 0.0316,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.380952380952381,
|
| 811 |
+
"grad_norm": 0.39382174611091614,
|
| 812 |
+
"learning_rate": 6.455264104305072e-05,
|
| 813 |
+
"loss": 0.0341,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.401656314699793,
|
| 818 |
+
"grad_norm": 0.30325403809547424,
|
| 819 |
+
"learning_rate": 6.285172906836416e-05,
|
| 820 |
+
"loss": 0.0316,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 2.422360248447205,
|
| 825 |
+
"grad_norm": 0.2884365916252136,
|
| 826 |
+
"learning_rate": 6.116319549472213e-05,
|
| 827 |
+
"loss": 0.0359,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.443064182194617,
|
| 832 |
+
"grad_norm": 0.37682902812957764,
|
| 833 |
+
"learning_rate": 5.948760296865523e-05,
|
| 834 |
+
"loss": 0.0335,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.463768115942029,
|
| 839 |
+
"grad_norm": 0.3018212616443634,
|
| 840 |
+
"learning_rate": 5.782550982452866e-05,
|
| 841 |
+
"loss": 0.0398,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.4844720496894412,
|
| 846 |
+
"grad_norm": 0.2686798870563507,
|
| 847 |
+
"learning_rate": 5.617746989849686e-05,
|
| 848 |
+
"loss": 0.034,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 2.505175983436853,
|
| 853 |
+
"grad_norm": 0.45458707213401794,
|
| 854 |
+
"learning_rate": 5.454403234395584e-05,
|
| 855 |
+
"loss": 0.0411,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.525879917184265,
|
| 860 |
+
"grad_norm": 0.5242733955383301,
|
| 861 |
+
"learning_rate": 5.2925741448556834e-05,
|
| 862 |
+
"loss": 0.04,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.546583850931677,
|
| 867 |
+
"grad_norm": 0.4235703945159912,
|
| 868 |
+
"learning_rate": 5.132313645284072e-05,
|
| 869 |
+
"loss": 0.0286,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.567287784679089,
|
| 874 |
+
"grad_norm": 0.47044745087623596,
|
| 875 |
+
"learning_rate": 4.973675137055451e-05,
|
| 876 |
+
"loss": 0.0386,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.587991718426501,
|
| 881 |
+
"grad_norm": 0.27159252762794495,
|
| 882 |
+
"learning_rate": 4.816711481070918e-05,
|
| 883 |
+
"loss": 0.0249,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.608695652173913,
|
| 888 |
+
"grad_norm": 0.2876708209514618,
|
| 889 |
+
"learning_rate": 4.661474980143889e-05,
|
| 890 |
+
"loss": 0.0275,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.629399585921325,
|
| 895 |
+
"grad_norm": 0.296629935503006,
|
| 896 |
+
"learning_rate": 4.5080173615719513e-05,
|
| 897 |
+
"loss": 0.0266,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.650103519668737,
|
| 902 |
+
"grad_norm": 0.33295968174934387,
|
| 903 |
+
"learning_rate": 4.356389759900495e-05,
|
| 904 |
+
"loss": 0.0281,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.670807453416149,
|
| 909 |
+
"grad_norm": 0.2696957588195801,
|
| 910 |
+
"learning_rate": 4.2066426998838516e-05,
|
| 911 |
+
"loss": 0.0277,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.691511387163561,
|
| 916 |
+
"grad_norm": 0.42278289794921875,
|
| 917 |
+
"learning_rate": 4.058826079649654e-05,
|
| 918 |
+
"loss": 0.0302,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.712215320910973,
|
| 923 |
+
"grad_norm": 0.3169129192829132,
|
| 924 |
+
"learning_rate": 3.912989154071961e-05,
|
| 925 |
+
"loss": 0.0261,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.732919254658385,
|
| 930 |
+
"grad_norm": 0.2851748764514923,
|
| 931 |
+
"learning_rate": 3.769180518358756e-05,
|
| 932 |
+
"loss": 0.0283,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.753623188405797,
|
| 937 |
+
"grad_norm": 0.4542989730834961,
|
| 938 |
+
"learning_rate": 3.627448091859219e-05,
|
| 939 |
+
"loss": 0.0353,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.774327122153209,
|
| 944 |
+
"grad_norm": 0.45544344186782837,
|
| 945 |
+
"learning_rate": 3.487839102096258e-05,
|
| 946 |
+
"loss": 0.0378,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.795031055900621,
|
| 951 |
+
"grad_norm": 0.3661940097808838,
|
| 952 |
+
"learning_rate": 3.350400069029518e-05,
|
| 953 |
+
"loss": 0.0286,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.8157349896480333,
|
| 958 |
+
"grad_norm": 0.31607571244239807,
|
| 959 |
+
"learning_rate": 3.2151767895541964e-05,
|
| 960 |
+
"loss": 0.0385,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.8364389233954452,
|
| 965 |
+
"grad_norm": 0.5506141185760498,
|
| 966 |
+
"learning_rate": 3.0822143222407706e-05,
|
| 967 |
+
"loss": 0.0287,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.857142857142857,
|
| 972 |
+
"grad_norm": 0.3711594045162201,
|
| 973 |
+
"learning_rate": 2.9515569723207594e-05,
|
| 974 |
+
"loss": 0.0267,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.877846790890269,
|
| 979 |
+
"grad_norm": 0.3517921566963196,
|
| 980 |
+
"learning_rate": 2.823248276923517e-05,
|
| 981 |
+
"loss": 0.0317,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.898550724637681,
|
| 986 |
+
"grad_norm": 0.27237316966056824,
|
| 987 |
+
"learning_rate": 2.6973309905689514e-05,
|
| 988 |
+
"loss": 0.0251,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.919254658385093,
|
| 993 |
+
"grad_norm": 0.23289814591407776,
|
| 994 |
+
"learning_rate": 2.5738470709210406e-05,
|
| 995 |
+
"loss": 0.0389,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.939958592132505,
|
| 1000 |
+
"grad_norm": 0.27631455659866333,
|
| 1001 |
+
"learning_rate": 2.4528376648068464e-05,
|
| 1002 |
+
"loss": 0.0273,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.960662525879917,
|
| 1007 |
+
"grad_norm": 0.28398358821868896,
|
| 1008 |
+
"learning_rate": 2.3343430945057552e-05,
|
| 1009 |
+
"loss": 0.0297,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.981366459627329,
|
| 1014 |
+
"grad_norm": 0.27578815817832947,
|
| 1015 |
+
"learning_rate": 2.2184028443134275e-05,
|
| 1016 |
+
"loss": 0.0213,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 3.002070393374741,
|
| 1021 |
+
"grad_norm": 0.30258387327194214,
|
| 1022 |
+
"learning_rate": 2.1050555473850077e-05,
|
| 1023 |
+
"loss": 0.0378,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 3.022774327122153,
|
| 1028 |
+
"grad_norm": 0.25420159101486206,
|
| 1029 |
+
"learning_rate": 1.9943389728619187e-05,
|
| 1030 |
+
"loss": 0.0309,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 3.0434782608695654,
|
| 1035 |
+
"grad_norm": 0.30880028009414673,
|
| 1036 |
+
"learning_rate": 1.8862900132866e-05,
|
| 1037 |
+
"loss": 0.0194,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 3.0641821946169774,
|
| 1042 |
+
"grad_norm": 0.15546509623527527,
|
| 1043 |
+
"learning_rate": 1.7809446723093015e-05,
|
| 1044 |
+
"loss": 0.0276,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 3.0848861283643894,
|
| 1049 |
+
"grad_norm": 0.3485426902770996,
|
| 1050 |
+
"learning_rate": 1.678338052691101e-05,
|
| 1051 |
+
"loss": 0.0191,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 3.1055900621118013,
|
| 1056 |
+
"grad_norm": 0.20627416670322418,
|
| 1057 |
+
"learning_rate": 1.5785043446070823e-05,
|
| 1058 |
+
"loss": 0.0259,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 3.1262939958592133,
|
| 1063 |
+
"grad_norm": 0.22484636306762695,
|
| 1064 |
+
"learning_rate": 1.4814768142536395e-05,
|
| 1065 |
+
"loss": 0.0202,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 3.1469979296066253,
|
| 1070 |
+
"grad_norm": 0.25017112493515015,
|
| 1071 |
+
"learning_rate": 1.3872877927636363e-05,
|
| 1072 |
+
"loss": 0.0287,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 3.1677018633540373,
|
| 1077 |
+
"grad_norm": 0.29567378759384155,
|
| 1078 |
+
"learning_rate": 1.2959686654331526e-05,
|
| 1079 |
+
"loss": 0.0241,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 3.1884057971014492,
|
| 1084 |
+
"grad_norm": 0.3368271589279175,
|
| 1085 |
+
"learning_rate": 1.2075498612633951e-05,
|
| 1086 |
+
"loss": 0.0209,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 3.209109730848861,
|
| 1091 |
+
"grad_norm": 0.31197062134742737,
|
| 1092 |
+
"learning_rate": 1.1220608428212764e-05,
|
| 1093 |
+
"loss": 0.0328,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 3.229813664596273,
|
| 1098 |
+
"grad_norm": 0.2368272840976715,
|
| 1099 |
+
"learning_rate": 1.039530096421999e-05,
|
| 1100 |
+
"loss": 0.0188,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 3.250517598343685,
|
| 1105 |
+
"grad_norm": 0.43641144037246704,
|
| 1106 |
+
"learning_rate": 9.59985122636976e-06,
|
| 1107 |
+
"loss": 0.0235,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 3.271221532091097,
|
| 1112 |
+
"grad_norm": 0.31003668904304504,
|
| 1113 |
+
"learning_rate": 8.834524271301846e-06,
|
| 1114 |
+
"loss": 0.0241,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 3.291925465838509,
|
| 1119 |
+
"grad_norm": 0.275126188993454,
|
| 1120 |
+
"learning_rate": 8.099575118260493e-06,
|
| 1121 |
+
"loss": 0.0238,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 3.3126293995859215,
|
| 1126 |
+
"grad_norm": 0.28839531540870667,
|
| 1127 |
+
"learning_rate": 7.395248664118071e-06,
|
| 1128 |
+
"loss": 0.0299,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 3.3333333333333335,
|
| 1133 |
+
"grad_norm": 0.3626795709133148,
|
| 1134 |
+
"learning_rate": 6.7217796017713605e-06,
|
| 1135 |
+
"loss": 0.029,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 3.3540372670807455,
|
| 1140 |
+
"grad_norm": 0.2354472577571869,
|
| 1141 |
+
"learning_rate": 6.079392341938184e-06,
|
| 1142 |
+
"loss": 0.0248,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 3.3747412008281574,
|
| 1147 |
+
"grad_norm": 0.4610183537006378,
|
| 1148 |
+
"learning_rate": 5.468300938380056e-06,
|
| 1149 |
+
"loss": 0.0294,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 3.3954451345755694,
|
| 1154 |
+
"grad_norm": 0.2801002264022827,
|
| 1155 |
+
"learning_rate": 4.888709016576154e-06,
|
| 1156 |
+
"loss": 0.0251,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 3.4161490683229814,
|
| 1161 |
+
"grad_norm": 0.3610021471977234,
|
| 1162 |
+
"learning_rate": 4.34080970587194e-06,
|
| 1163 |
+
"loss": 0.021,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 3.4368530020703933,
|
| 1168 |
+
"grad_norm": 0.26129838824272156,
|
| 1169 |
+
"learning_rate": 3.824785575125412e-06,
|
| 1170 |
+
"loss": 0.0352,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 3.4575569358178053,
|
| 1175 |
+
"grad_norm": 0.28716525435447693,
|
| 1176 |
+
"learning_rate": 3.340808571872156e-06,
|
| 1177 |
+
"loss": 0.0232,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 3.4782608695652173,
|
| 1182 |
+
"grad_norm": 0.3004515469074249,
|
| 1183 |
+
"learning_rate": 2.8890399650297493e-06,
|
| 1184 |
+
"loss": 0.0226,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 3.4989648033126293,
|
| 1189 |
+
"grad_norm": 0.3537054657936096,
|
| 1190 |
+
"learning_rate": 2.4696302911603187e-06,
|
| 1191 |
+
"loss": 0.0359,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 3.5196687370600412,
|
| 1196 |
+
"grad_norm": 0.38065382838249207,
|
| 1197 |
+
"learning_rate": 2.082719304309355e-06,
|
| 1198 |
+
"loss": 0.0194,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 3.5403726708074537,
|
| 1203 |
+
"grad_norm": 0.22795440256595612,
|
| 1204 |
+
"learning_rate": 1.7284359294373663e-06,
|
| 1205 |
+
"loss": 0.0181,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 3.5610766045548656,
|
| 1210 |
+
"grad_norm": 0.4666849970817566,
|
| 1211 |
+
"learning_rate": 1.4068982194600844e-06,
|
| 1212 |
+
"loss": 0.0281,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 3.5817805383022776,
|
| 1217 |
+
"grad_norm": 0.26012447476387024,
|
| 1218 |
+
"learning_rate": 1.1182133159113095e-06,
|
| 1219 |
+
"loss": 0.0207,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.6024844720496896,
|
| 1224 |
+
"grad_norm": 0.17938533425331116,
|
| 1225 |
+
"learning_rate": 8.624774132415914e-07,
|
| 1226 |
+
"loss": 0.0278,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 3.6231884057971016,
|
| 1231 |
+
"grad_norm": 0.23828087747097015,
|
| 1232 |
+
"learning_rate": 6.397757267647353e-07,
|
| 1233 |
+
"loss": 0.0218,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 3.6438923395445135,
|
| 1238 |
+
"grad_norm": 0.4279627501964569,
|
| 1239 |
+
"learning_rate": 4.5018246426267176e-07,
|
| 1240 |
+
"loss": 0.0218,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 3.6645962732919255,
|
| 1245 |
+
"grad_norm": 0.5297583341598511,
|
| 1246 |
+
"learning_rate": 2.937608012582471e-07,
|
| 1247 |
+
"loss": 0.0219,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 3.6853002070393375,
|
| 1252 |
+
"grad_norm": 0.15538057684898376,
|
| 1253 |
+
"learning_rate": 1.7056285996409627e-07,
|
| 1254 |
+
"loss": 0.02,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 3.7060041407867494,
|
| 1259 |
+
"grad_norm": 0.4270657002925873,
|
| 1260 |
+
"learning_rate": 8.062969191471315e-08,
|
| 1261 |
+
"loss": 0.0302,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 3.7267080745341614,
|
| 1266 |
+
"grad_norm": 0.28654342889785767,
|
| 1267 |
+
"learning_rate": 2.3991264287381852e-08,
|
| 1268 |
+
"loss": 0.0301,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 3.7474120082815734,
|
| 1273 |
+
"grad_norm": 0.4787255525588989,
|
| 1274 |
+
"learning_rate": 6.664499166864069e-10,
|
| 1275 |
+
"loss": 0.0201,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 3.751552795031056,
|
| 1280 |
+
"step": 1812,
|
| 1281 |
+
"total_flos": 6.571756319902762e+16,
|
| 1282 |
+
"train_loss": 0.060702945790346095,
|
| 1283 |
+
"train_runtime": 803.1177,
|
| 1284 |
+
"train_samples_per_second": 36.099,
|
| 1285 |
+
"train_steps_per_second": 2.256
|
| 1286 |
+
}
|
| 1287 |
+
],
|
| 1288 |
+
"logging_steps": 10,
|
| 1289 |
+
"max_steps": 1812,
|
| 1290 |
+
"num_input_tokens_seen": 0,
|
| 1291 |
+
"num_train_epochs": 4,
|
| 1292 |
+
"save_steps": 500,
|
| 1293 |
+
"stateful_callbacks": {
|
| 1294 |
+
"TrainerControl": {
|
| 1295 |
+
"args": {
|
| 1296 |
+
"should_epoch_stop": false,
|
| 1297 |
+
"should_evaluate": false,
|
| 1298 |
+
"should_log": false,
|
| 1299 |
+
"should_save": true,
|
| 1300 |
+
"should_training_stop": true
|
| 1301 |
+
},
|
| 1302 |
+
"attributes": {}
|
| 1303 |
+
}
|
| 1304 |
+
},
|
| 1305 |
+
"total_flos": 6.571756319902762e+16,
|
| 1306 |
+
"train_batch_size": 16,
|
| 1307 |
+
"trial_name": null,
|
| 1308 |
+
"trial_params": null
|
| 1309 |
+
}
|