Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1345 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1345 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1860,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.053763440860215055,
|
| 14 |
+
"grad_norm": 3.3253746032714844,
|
| 15 |
+
"learning_rate": 9.67741935483871e-06,
|
| 16 |
+
"loss": 0.7408,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.10752688172043011,
|
| 21 |
+
"grad_norm": 1.349915623664856,
|
| 22 |
+
"learning_rate": 2.0430107526881722e-05,
|
| 23 |
+
"loss": 0.2858,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.16129032258064516,
|
| 28 |
+
"grad_norm": 0.761350154876709,
|
| 29 |
+
"learning_rate": 3.118279569892473e-05,
|
| 30 |
+
"loss": 0.2045,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.21505376344086022,
|
| 35 |
+
"grad_norm": 2.6227965354919434,
|
| 36 |
+
"learning_rate": 4.1935483870967746e-05,
|
| 37 |
+
"loss": 0.1893,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.26881720430107525,
|
| 42 |
+
"grad_norm": 0.9433155059814453,
|
| 43 |
+
"learning_rate": 5.268817204301075e-05,
|
| 44 |
+
"loss": 0.1669,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.3225806451612903,
|
| 49 |
+
"grad_norm": 1.2473623752593994,
|
| 50 |
+
"learning_rate": 6.344086021505376e-05,
|
| 51 |
+
"loss": 0.1501,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.3763440860215054,
|
| 56 |
+
"grad_norm": 1.2984980344772339,
|
| 57 |
+
"learning_rate": 7.419354838709677e-05,
|
| 58 |
+
"loss": 0.128,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.43010752688172044,
|
| 63 |
+
"grad_norm": 1.0100502967834473,
|
| 64 |
+
"learning_rate": 8.494623655913979e-05,
|
| 65 |
+
"loss": 0.1114,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.4838709677419355,
|
| 70 |
+
"grad_norm": 0.7015477418899536,
|
| 71 |
+
"learning_rate": 9.56989247311828e-05,
|
| 72 |
+
"loss": 0.1014,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.5376344086021505,
|
| 77 |
+
"grad_norm": 0.7916803359985352,
|
| 78 |
+
"learning_rate": 9.999715511289097e-05,
|
| 79 |
+
"loss": 0.0965,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.5913978494623656,
|
| 84 |
+
"grad_norm": 0.5534141659736633,
|
| 85 |
+
"learning_rate": 9.997977086403005e-05,
|
| 86 |
+
"loss": 0.0894,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.6451612903225806,
|
| 91 |
+
"grad_norm": 1.098301887512207,
|
| 92 |
+
"learning_rate": 9.994658834752877e-05,
|
| 93 |
+
"loss": 0.084,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.6989247311827957,
|
| 98 |
+
"grad_norm": 0.634315550327301,
|
| 99 |
+
"learning_rate": 9.989761805215624e-05,
|
| 100 |
+
"loss": 0.0844,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.7526881720430108,
|
| 105 |
+
"grad_norm": 0.5075936913490295,
|
| 106 |
+
"learning_rate": 9.983287545709175e-05,
|
| 107 |
+
"loss": 0.0784,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.8064516129032258,
|
| 112 |
+
"grad_norm": 0.7758091688156128,
|
| 113 |
+
"learning_rate": 9.97523810270318e-05,
|
| 114 |
+
"loss": 0.072,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.8602150537634409,
|
| 119 |
+
"grad_norm": 0.9043412804603577,
|
| 120 |
+
"learning_rate": 9.965616020572153e-05,
|
| 121 |
+
"loss": 0.0674,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.9139784946236559,
|
| 126 |
+
"grad_norm": 0.6532458066940308,
|
| 127 |
+
"learning_rate": 9.954424340791196e-05,
|
| 128 |
+
"loss": 0.0618,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.967741935483871,
|
| 133 |
+
"grad_norm": 0.4447098672389984,
|
| 134 |
+
"learning_rate": 9.941666600974616e-05,
|
| 135 |
+
"loss": 0.0607,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.021505376344086,
|
| 140 |
+
"grad_norm": 0.3520924746990204,
|
| 141 |
+
"learning_rate": 9.927346833757709e-05,
|
| 142 |
+
"loss": 0.0576,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.075268817204301,
|
| 147 |
+
"grad_norm": 0.39900141954421997,
|
| 148 |
+
"learning_rate": 9.911469565522075e-05,
|
| 149 |
+
"loss": 0.0587,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.129032258064516,
|
| 154 |
+
"grad_norm": 0.3432719111442566,
|
| 155 |
+
"learning_rate": 9.894039814964849e-05,
|
| 156 |
+
"loss": 0.0588,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.1827956989247312,
|
| 161 |
+
"grad_norm": 0.5504620671272278,
|
| 162 |
+
"learning_rate": 9.875063091512336e-05,
|
| 163 |
+
"loss": 0.0542,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.2365591397849462,
|
| 168 |
+
"grad_norm": 0.8034648895263672,
|
| 169 |
+
"learning_rate": 9.854545393578508e-05,
|
| 170 |
+
"loss": 0.0524,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.2903225806451613,
|
| 175 |
+
"grad_norm": 0.6812065243721008,
|
| 176 |
+
"learning_rate": 9.832493206668958e-05,
|
| 177 |
+
"loss": 0.0537,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.3440860215053765,
|
| 182 |
+
"grad_norm": 0.34998008608818054,
|
| 183 |
+
"learning_rate": 9.808913501330867e-05,
|
| 184 |
+
"loss": 0.0543,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.3978494623655915,
|
| 189 |
+
"grad_norm": 0.6064873933792114,
|
| 190 |
+
"learning_rate": 9.783813730949658e-05,
|
| 191 |
+
"loss": 0.0492,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.4516129032258065,
|
| 196 |
+
"grad_norm": 0.40691688656806946,
|
| 197 |
+
"learning_rate": 9.757201829393044e-05,
|
| 198 |
+
"loss": 0.0494,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.5053763440860215,
|
| 203 |
+
"grad_norm": 0.433832049369812,
|
| 204 |
+
"learning_rate": 9.729086208503174e-05,
|
| 205 |
+
"loss": 0.0448,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.5591397849462365,
|
| 210 |
+
"grad_norm": 0.4424607455730438,
|
| 211 |
+
"learning_rate": 9.699475755437713e-05,
|
| 212 |
+
"loss": 0.0437,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.6129032258064515,
|
| 217 |
+
"grad_norm": 0.5499128699302673,
|
| 218 |
+
"learning_rate": 9.668379829860667e-05,
|
| 219 |
+
"loss": 0.046,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.6666666666666665,
|
| 224 |
+
"grad_norm": 0.31413379311561584,
|
| 225 |
+
"learning_rate": 9.635808260983864e-05,
|
| 226 |
+
"loss": 0.044,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.7204301075268817,
|
| 231 |
+
"grad_norm": 0.7016597390174866,
|
| 232 |
+
"learning_rate": 9.601771344459996e-05,
|
| 233 |
+
"loss": 0.0484,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.7741935483870968,
|
| 238 |
+
"grad_norm": 0.35866835713386536,
|
| 239 |
+
"learning_rate": 9.566279839128243e-05,
|
| 240 |
+
"loss": 0.0428,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.827956989247312,
|
| 245 |
+
"grad_norm": 0.5350872874259949,
|
| 246 |
+
"learning_rate": 9.529344963613469e-05,
|
| 247 |
+
"loss": 0.0448,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.881720430107527,
|
| 252 |
+
"grad_norm": 0.38403162360191345,
|
| 253 |
+
"learning_rate": 9.490978392780096e-05,
|
| 254 |
+
"loss": 0.044,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.935483870967742,
|
| 259 |
+
"grad_norm": 0.37314051389694214,
|
| 260 |
+
"learning_rate": 9.451192254041758e-05,
|
| 261 |
+
"loss": 0.0426,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.989247311827957,
|
| 266 |
+
"grad_norm": 0.4151698648929596,
|
| 267 |
+
"learning_rate": 9.409999123527899e-05,
|
| 268 |
+
"loss": 0.0445,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 2.043010752688172,
|
| 273 |
+
"grad_norm": 0.4633235037326813,
|
| 274 |
+
"learning_rate": 9.367412022108557e-05,
|
| 275 |
+
"loss": 0.0435,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 2.096774193548387,
|
| 280 |
+
"grad_norm": 0.47718819975852966,
|
| 281 |
+
"learning_rate": 9.323444411278543e-05,
|
| 282 |
+
"loss": 0.0467,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 2.150537634408602,
|
| 287 |
+
"grad_norm": 0.42166078090667725,
|
| 288 |
+
"learning_rate": 9.278110188902363e-05,
|
| 289 |
+
"loss": 0.0427,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 2.204301075268817,
|
| 294 |
+
"grad_norm": 0.48610246181488037,
|
| 295 |
+
"learning_rate": 9.231423684821188e-05,
|
| 296 |
+
"loss": 0.0394,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.258064516129032,
|
| 301 |
+
"grad_norm": 0.2654590904712677,
|
| 302 |
+
"learning_rate": 9.183399656323298e-05,
|
| 303 |
+
"loss": 0.0404,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 2.3118279569892475,
|
| 308 |
+
"grad_norm": 0.29829591512680054,
|
| 309 |
+
"learning_rate": 9.134053283479391e-05,
|
| 310 |
+
"loss": 0.041,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 2.3655913978494625,
|
| 315 |
+
"grad_norm": 0.3248680531978607,
|
| 316 |
+
"learning_rate": 9.083400164344279e-05,
|
| 317 |
+
"loss": 0.04,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 2.4193548387096775,
|
| 322 |
+
"grad_norm": 0.5130133032798767,
|
| 323 |
+
"learning_rate": 9.031456310026438e-05,
|
| 324 |
+
"loss": 0.0418,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 2.4731182795698925,
|
| 329 |
+
"grad_norm": 0.32805633544921875,
|
| 330 |
+
"learning_rate": 8.978238139627004e-05,
|
| 331 |
+
"loss": 0.0396,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.5268817204301075,
|
| 336 |
+
"grad_norm": 0.5853477716445923,
|
| 337 |
+
"learning_rate": 8.923762475049821e-05,
|
| 338 |
+
"loss": 0.0392,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 2.5806451612903225,
|
| 343 |
+
"grad_norm": 0.6138908863067627,
|
| 344 |
+
"learning_rate": 8.86804653568413e-05,
|
| 345 |
+
"loss": 0.0378,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 2.6344086021505375,
|
| 350 |
+
"grad_norm": 0.39958009123802185,
|
| 351 |
+
"learning_rate": 8.811107932961652e-05,
|
| 352 |
+
"loss": 0.035,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 2.688172043010753,
|
| 357 |
+
"grad_norm": 0.2563028335571289,
|
| 358 |
+
"learning_rate": 8.752964664789725e-05,
|
| 359 |
+
"loss": 0.0366,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 2.741935483870968,
|
| 364 |
+
"grad_norm": 0.4391862452030182,
|
| 365 |
+
"learning_rate": 8.693635109862295e-05,
|
| 366 |
+
"loss": 0.0382,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 2.795698924731183,
|
| 371 |
+
"grad_norm": 0.4005059599876404,
|
| 372 |
+
"learning_rate": 8.633138021850526e-05,
|
| 373 |
+
"loss": 0.0385,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.849462365591398,
|
| 378 |
+
"grad_norm": 0.3844977021217346,
|
| 379 |
+
"learning_rate": 8.571492523474898e-05,
|
| 380 |
+
"loss": 0.0372,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.903225806451613,
|
| 385 |
+
"grad_norm": 0.31710556149482727,
|
| 386 |
+
"learning_rate": 8.508718100460642e-05,
|
| 387 |
+
"loss": 0.0362,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 2.956989247311828,
|
| 392 |
+
"grad_norm": 0.3512146472930908,
|
| 393 |
+
"learning_rate": 8.444834595378434e-05,
|
| 394 |
+
"loss": 0.035,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 3.010752688172043,
|
| 399 |
+
"grad_norm": 0.49279454350471497,
|
| 400 |
+
"learning_rate": 8.379862201372293e-05,
|
| 401 |
+
"loss": 0.0349,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 3.064516129032258,
|
| 406 |
+
"grad_norm": 0.5000328421592712,
|
| 407 |
+
"learning_rate": 8.313821455776664e-05,
|
| 408 |
+
"loss": 0.0353,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 3.118279569892473,
|
| 413 |
+
"grad_norm": 0.4451806843280792,
|
| 414 |
+
"learning_rate": 8.246733233624709e-05,
|
| 415 |
+
"loss": 0.0376,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.172043010752688,
|
| 420 |
+
"grad_norm": 0.5780179500579834,
|
| 421 |
+
"learning_rate": 8.178618741049842e-05,
|
| 422 |
+
"loss": 0.0341,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 3.225806451612903,
|
| 427 |
+
"grad_norm": 0.34750205278396606,
|
| 428 |
+
"learning_rate": 8.10949950858262e-05,
|
| 429 |
+
"loss": 0.0385,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 3.279569892473118,
|
| 434 |
+
"grad_norm": 0.4077950119972229,
|
| 435 |
+
"learning_rate": 8.039397384345091e-05,
|
| 436 |
+
"loss": 0.0371,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 3.3333333333333335,
|
| 441 |
+
"grad_norm": 0.325566828250885,
|
| 442 |
+
"learning_rate": 7.968334527144744e-05,
|
| 443 |
+
"loss": 0.0361,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 3.3870967741935485,
|
| 448 |
+
"grad_norm": 0.301077276468277,
|
| 449 |
+
"learning_rate": 7.896333399470255e-05,
|
| 450 |
+
"loss": 0.0315,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 3.4408602150537635,
|
| 455 |
+
"grad_norm": 0.5286648869514465,
|
| 456 |
+
"learning_rate": 7.823416760391262e-05,
|
| 457 |
+
"loss": 0.0334,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 3.4946236559139785,
|
| 462 |
+
"grad_norm": 0.35336583852767944,
|
| 463 |
+
"learning_rate": 7.749607658364365e-05,
|
| 464 |
+
"loss": 0.0352,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 3.5483870967741935,
|
| 469 |
+
"grad_norm": 0.25968852639198303,
|
| 470 |
+
"learning_rate": 7.674929423947665e-05,
|
| 471 |
+
"loss": 0.0347,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 3.6021505376344085,
|
| 476 |
+
"grad_norm": 0.2828128933906555,
|
| 477 |
+
"learning_rate": 7.59940566242614e-05,
|
| 478 |
+
"loss": 0.0317,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 3.6559139784946235,
|
| 483 |
+
"grad_norm": 0.4576037526130676,
|
| 484 |
+
"learning_rate": 7.523060246350162e-05,
|
| 485 |
+
"loss": 0.0332,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 3.709677419354839,
|
| 490 |
+
"grad_norm": 0.267994225025177,
|
| 491 |
+
"learning_rate": 7.445917307989559e-05,
|
| 492 |
+
"loss": 0.0298,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 3.763440860215054,
|
| 497 |
+
"grad_norm": 0.19018298387527466,
|
| 498 |
+
"learning_rate": 7.368001231705555e-05,
|
| 499 |
+
"loss": 0.0339,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 3.817204301075269,
|
| 504 |
+
"grad_norm": 0.301540732383728,
|
| 505 |
+
"learning_rate": 7.289336646243049e-05,
|
| 506 |
+
"loss": 0.03,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 3.870967741935484,
|
| 511 |
+
"grad_norm": 0.2555813789367676,
|
| 512 |
+
"learning_rate": 7.209948416945628e-05,
|
| 513 |
+
"loss": 0.033,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 3.924731182795699,
|
| 518 |
+
"grad_norm": 0.32856234908103943,
|
| 519 |
+
"learning_rate": 7.129861637895803e-05,
|
| 520 |
+
"loss": 0.0301,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 3.978494623655914,
|
| 525 |
+
"grad_norm": 0.27070337533950806,
|
| 526 |
+
"learning_rate": 7.049101623982937e-05,
|
| 527 |
+
"loss": 0.0333,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 4.032258064516129,
|
| 532 |
+
"grad_norm": 0.30932679772377014,
|
| 533 |
+
"learning_rate": 6.967693902901386e-05,
|
| 534 |
+
"loss": 0.0318,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 4.086021505376344,
|
| 539 |
+
"grad_norm": 0.40344807505607605,
|
| 540 |
+
"learning_rate": 6.885664207081357e-05,
|
| 541 |
+
"loss": 0.0347,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 4.139784946236559,
|
| 546 |
+
"grad_norm": 0.3657456338405609,
|
| 547 |
+
"learning_rate": 6.803038465555067e-05,
|
| 548 |
+
"loss": 0.034,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 4.193548387096774,
|
| 553 |
+
"grad_norm": 0.29593250155448914,
|
| 554 |
+
"learning_rate": 6.719842795760757e-05,
|
| 555 |
+
"loss": 0.0317,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 4.247311827956989,
|
| 560 |
+
"grad_norm": 0.32772189378738403,
|
| 561 |
+
"learning_rate": 6.636103495287131e-05,
|
| 562 |
+
"loss": 0.0299,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 4.301075268817204,
|
| 567 |
+
"grad_norm": 0.21649384498596191,
|
| 568 |
+
"learning_rate": 6.551847033560884e-05,
|
| 569 |
+
"loss": 0.0328,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 4.354838709677419,
|
| 574 |
+
"grad_norm": 0.24615323543548584,
|
| 575 |
+
"learning_rate": 6.46710004347988e-05,
|
| 576 |
+
"loss": 0.0307,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 4.408602150537634,
|
| 581 |
+
"grad_norm": 0.30269894003868103,
|
| 582 |
+
"learning_rate": 6.381889312994678e-05,
|
| 583 |
+
"loss": 0.0296,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 4.462365591397849,
|
| 588 |
+
"grad_norm": 0.2541375160217285,
|
| 589 |
+
"learning_rate": 6.29624177664105e-05,
|
| 590 |
+
"loss": 0.0289,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 4.516129032258064,
|
| 595 |
+
"grad_norm": 0.28128522634506226,
|
| 596 |
+
"learning_rate": 6.210184507026144e-05,
|
| 597 |
+
"loss": 0.0297,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 4.56989247311828,
|
| 602 |
+
"grad_norm": 0.4417729675769806,
|
| 603 |
+
"learning_rate": 6.123744706271028e-05,
|
| 604 |
+
"loss": 0.028,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 4.623655913978495,
|
| 609 |
+
"grad_norm": 0.3367769122123718,
|
| 610 |
+
"learning_rate": 6.0369496974122666e-05,
|
| 611 |
+
"loss": 0.0269,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 4.67741935483871,
|
| 616 |
+
"grad_norm": 0.36587026715278625,
|
| 617 |
+
"learning_rate": 5.949826915765314e-05,
|
| 618 |
+
"loss": 0.027,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 4.731182795698925,
|
| 623 |
+
"grad_norm": 0.3275097906589508,
|
| 624 |
+
"learning_rate": 5.862403900252383e-05,
|
| 625 |
+
"loss": 0.0316,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 4.78494623655914,
|
| 630 |
+
"grad_norm": 0.2757854461669922,
|
| 631 |
+
"learning_rate": 5.7747082846975795e-05,
|
| 632 |
+
"loss": 0.0296,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 4.838709677419355,
|
| 637 |
+
"grad_norm": 0.21201659739017487,
|
| 638 |
+
"learning_rate": 5.686767789092041e-05,
|
| 639 |
+
"loss": 0.0276,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 4.89247311827957,
|
| 644 |
+
"grad_norm": 0.41852885484695435,
|
| 645 |
+
"learning_rate": 5.598610210831837e-05,
|
| 646 |
+
"loss": 0.0266,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 4.946236559139785,
|
| 651 |
+
"grad_norm": 0.3428756892681122,
|
| 652 |
+
"learning_rate": 5.510263415931388e-05,
|
| 653 |
+
"loss": 0.0289,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 5.0,
|
| 658 |
+
"grad_norm": 0.8262572884559631,
|
| 659 |
+
"learning_rate": 5.4217553302152237e-05,
|
| 660 |
+
"loss": 0.0293,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 5.053763440860215,
|
| 665 |
+
"grad_norm": 0.4343954920768738,
|
| 666 |
+
"learning_rate": 5.333113930490807e-05,
|
| 667 |
+
"loss": 0.0319,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 5.10752688172043,
|
| 672 |
+
"grad_norm": 0.312368780374527,
|
| 673 |
+
"learning_rate": 5.2443672357052554e-05,
|
| 674 |
+
"loss": 0.0292,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 5.161290322580645,
|
| 679 |
+
"grad_norm": 0.2760954201221466,
|
| 680 |
+
"learning_rate": 5.155543298088744e-05,
|
| 681 |
+
"loss": 0.0276,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 5.21505376344086,
|
| 686 |
+
"grad_norm": 0.3945291042327881,
|
| 687 |
+
"learning_rate": 5.0666701942873875e-05,
|
| 688 |
+
"loss": 0.0284,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 5.268817204301075,
|
| 693 |
+
"grad_norm": 0.27156421542167664,
|
| 694 |
+
"learning_rate": 4.977776016488397e-05,
|
| 695 |
+
"loss": 0.0258,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 5.32258064516129,
|
| 700 |
+
"grad_norm": 0.4421350955963135,
|
| 701 |
+
"learning_rate": 4.8888888635403276e-05,
|
| 702 |
+
"loss": 0.0251,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 5.376344086021505,
|
| 707 |
+
"grad_norm": 0.25600770115852356,
|
| 708 |
+
"learning_rate": 4.8000368320712364e-05,
|
| 709 |
+
"loss": 0.0279,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 5.43010752688172,
|
| 714 |
+
"grad_norm": 0.2691516578197479,
|
| 715 |
+
"learning_rate": 4.71124800760751e-05,
|
| 716 |
+
"loss": 0.0243,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 5.483870967741936,
|
| 721 |
+
"grad_norm": 0.27804574370384216,
|
| 722 |
+
"learning_rate": 4.6225504556962304e-05,
|
| 723 |
+
"loss": 0.0297,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 5.53763440860215,
|
| 728 |
+
"grad_norm": 0.34384310245513916,
|
| 729 |
+
"learning_rate": 4.533972213033844e-05,
|
| 730 |
+
"loss": 0.0279,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 5.591397849462366,
|
| 735 |
+
"grad_norm": 0.3319939970970154,
|
| 736 |
+
"learning_rate": 4.44554127860395e-05,
|
| 737 |
+
"loss": 0.0281,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 5.645161290322581,
|
| 742 |
+
"grad_norm": 0.29645469784736633,
|
| 743 |
+
"learning_rate": 4.357285604827011e-05,
|
| 744 |
+
"loss": 0.0285,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 5.698924731182796,
|
| 749 |
+
"grad_norm": 0.15289399027824402,
|
| 750 |
+
"learning_rate": 4.269233088724779e-05,
|
| 751 |
+
"loss": 0.0253,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 5.752688172043011,
|
| 756 |
+
"grad_norm": 0.29148632287979126,
|
| 757 |
+
"learning_rate": 4.1814115631022474e-05,
|
| 758 |
+
"loss": 0.028,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 5.806451612903226,
|
| 763 |
+
"grad_norm": 0.3517538011074066,
|
| 764 |
+
"learning_rate": 4.09384878774987e-05,
|
| 765 |
+
"loss": 0.0256,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 5.860215053763441,
|
| 770 |
+
"grad_norm": 0.3300669491291046,
|
| 771 |
+
"learning_rate": 4.006572440668898e-05,
|
| 772 |
+
"loss": 0.0252,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 5.913978494623656,
|
| 777 |
+
"grad_norm": 0.3203289210796356,
|
| 778 |
+
"learning_rate": 3.919610109322538e-05,
|
| 779 |
+
"loss": 0.027,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 5.967741935483871,
|
| 784 |
+
"grad_norm": 0.23981882631778717,
|
| 785 |
+
"learning_rate": 3.8329892819157484e-05,
|
| 786 |
+
"loss": 0.0286,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 6.021505376344086,
|
| 791 |
+
"grad_norm": 0.3529661297798157,
|
| 792 |
+
"learning_rate": 3.746737338706397e-05,
|
| 793 |
+
"loss": 0.0261,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 6.075268817204301,
|
| 798 |
+
"grad_norm": 0.21720324456691742,
|
| 799 |
+
"learning_rate": 3.6608815433505596e-05,
|
| 800 |
+
"loss": 0.0271,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 6.129032258064516,
|
| 805 |
+
"grad_norm": 0.21735608577728271,
|
| 806 |
+
"learning_rate": 3.575449034284645e-05,
|
| 807 |
+
"loss": 0.0294,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 6.182795698924731,
|
| 812 |
+
"grad_norm": 0.31982213258743286,
|
| 813 |
+
"learning_rate": 3.4904668161471345e-05,
|
| 814 |
+
"loss": 0.0249,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 6.236559139784946,
|
| 819 |
+
"grad_norm": 0.3131325840950012,
|
| 820 |
+
"learning_rate": 3.405961751242584e-05,
|
| 821 |
+
"loss": 0.0244,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 6.290322580645161,
|
| 826 |
+
"grad_norm": 0.268307626247406,
|
| 827 |
+
"learning_rate": 3.321960551050648e-05,
|
| 828 |
+
"loss": 0.025,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 6.344086021505376,
|
| 833 |
+
"grad_norm": 0.3868793249130249,
|
| 834 |
+
"learning_rate": 3.238489767782743e-05,
|
| 835 |
+
"loss": 0.0249,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 6.397849462365591,
|
| 840 |
+
"grad_norm": 0.28992360830307007,
|
| 841 |
+
"learning_rate": 3.155575785989091e-05,
|
| 842 |
+
"loss": 0.0234,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 6.451612903225806,
|
| 847 |
+
"grad_norm": 0.25599104166030884,
|
| 848 |
+
"learning_rate": 3.0732448142187235e-05,
|
| 849 |
+
"loss": 0.0245,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 6.505376344086022,
|
| 854 |
+
"grad_norm": 0.14047683775424957,
|
| 855 |
+
"learning_rate": 2.991522876735154e-05,
|
| 856 |
+
"loss": 0.0214,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 6.559139784946236,
|
| 861 |
+
"grad_norm": 0.28923436999320984,
|
| 862 |
+
"learning_rate": 2.910435805290269e-05,
|
| 863 |
+
"loss": 0.0245,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 6.612903225806452,
|
| 868 |
+
"grad_norm": 0.22662054002285004,
|
| 869 |
+
"learning_rate": 2.830009230959097e-05,
|
| 870 |
+
"loss": 0.0238,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 6.666666666666667,
|
| 875 |
+
"grad_norm": 0.25063997507095337,
|
| 876 |
+
"learning_rate": 2.750268576037993e-05,
|
| 877 |
+
"loss": 0.0244,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 6.720430107526882,
|
| 882 |
+
"grad_norm": 0.2001640796661377,
|
| 883 |
+
"learning_rate": 2.671239046008826e-05,
|
| 884 |
+
"loss": 0.0234,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 6.774193548387097,
|
| 889 |
+
"grad_norm": 0.16993772983551025,
|
| 890 |
+
"learning_rate": 2.592945621571692e-05,
|
| 891 |
+
"loss": 0.028,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 6.827956989247312,
|
| 896 |
+
"grad_norm": 0.23878411948680878,
|
| 897 |
+
"learning_rate": 2.51541305074869e-05,
|
| 898 |
+
"loss": 0.0215,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 6.881720430107527,
|
| 903 |
+
"grad_norm": 0.22890987992286682,
|
| 904 |
+
"learning_rate": 2.43866584106123e-05,
|
| 905 |
+
"loss": 0.0268,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 6.935483870967742,
|
| 910 |
+
"grad_norm": 0.2709693908691406,
|
| 911 |
+
"learning_rate": 2.3627282517833834e-05,
|
| 912 |
+
"loss": 0.0246,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 6.989247311827957,
|
| 917 |
+
"grad_norm": 0.15185967087745667,
|
| 918 |
+
"learning_rate": 2.2876242862736814e-05,
|
| 919 |
+
"loss": 0.0219,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 7.043010752688172,
|
| 924 |
+
"grad_norm": 0.15341722965240479,
|
| 925 |
+
"learning_rate": 2.2133776843878186e-05,
|
| 926 |
+
"loss": 0.0212,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 7.096774193548387,
|
| 931 |
+
"grad_norm": 0.25035205483436584,
|
| 932 |
+
"learning_rate": 2.1400119149746457e-05,
|
| 933 |
+
"loss": 0.0237,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 7.150537634408602,
|
| 938 |
+
"grad_norm": 0.2255420982837677,
|
| 939 |
+
"learning_rate": 2.0675501684578237e-05,
|
| 940 |
+
"loss": 0.0229,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 7.204301075268817,
|
| 945 |
+
"grad_norm": 0.19871476292610168,
|
| 946 |
+
"learning_rate": 1.9960153495054885e-05,
|
| 947 |
+
"loss": 0.0229,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 7.258064516129032,
|
| 952 |
+
"grad_norm": 0.18878574669361115,
|
| 953 |
+
"learning_rate": 1.9254300697902405e-05,
|
| 954 |
+
"loss": 0.0203,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 7.311827956989247,
|
| 959 |
+
"grad_norm": 0.20982198417186737,
|
| 960 |
+
"learning_rate": 1.85581664084175e-05,
|
| 961 |
+
"loss": 0.0233,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 7.365591397849462,
|
| 966 |
+
"grad_norm": 0.24840368330478668,
|
| 967 |
+
"learning_rate": 1.7871970669942216e-05,
|
| 968 |
+
"loss": 0.0202,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 7.419354838709677,
|
| 973 |
+
"grad_norm": 0.18493802845478058,
|
| 974 |
+
"learning_rate": 1.7195930384309966e-05,
|
| 975 |
+
"loss": 0.0221,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 7.473118279569892,
|
| 980 |
+
"grad_norm": 0.301211953163147,
|
| 981 |
+
"learning_rate": 1.6530259243284052e-05,
|
| 982 |
+
"loss": 0.0225,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 7.526881720430108,
|
| 987 |
+
"grad_norm": 0.2749289870262146,
|
| 988 |
+
"learning_rate": 1.5875167661011342e-05,
|
| 989 |
+
"loss": 0.0222,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 7.580645161290323,
|
| 994 |
+
"grad_norm": 0.1663655936717987,
|
| 995 |
+
"learning_rate": 1.5230862707511711e-05,
|
| 996 |
+
"loss": 0.0202,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 7.634408602150538,
|
| 1001 |
+
"grad_norm": 0.2491251528263092,
|
| 1002 |
+
"learning_rate": 1.4597548043224613e-05,
|
| 1003 |
+
"loss": 0.0223,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 7.688172043010753,
|
| 1008 |
+
"grad_norm": 0.19118069112300873,
|
| 1009 |
+
"learning_rate": 1.397542385463333e-05,
|
| 1010 |
+
"loss": 0.0199,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 7.741935483870968,
|
| 1015 |
+
"grad_norm": 0.22517769038677216,
|
| 1016 |
+
"learning_rate": 1.3364686790987551e-05,
|
| 1017 |
+
"loss": 0.0224,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 7.795698924731183,
|
| 1022 |
+
"grad_norm": 0.23471683263778687,
|
| 1023 |
+
"learning_rate": 1.2765529902143653e-05,
|
| 1024 |
+
"loss": 0.0205,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 7.849462365591398,
|
| 1029 |
+
"grad_norm": 0.17809489369392395,
|
| 1030 |
+
"learning_rate": 1.2178142577543123e-05,
|
| 1031 |
+
"loss": 0.0255,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 7.903225806451613,
|
| 1036 |
+
"grad_norm": 0.2622681260108948,
|
| 1037 |
+
"learning_rate": 1.1602710486347735e-05,
|
| 1038 |
+
"loss": 0.0221,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 7.956989247311828,
|
| 1043 |
+
"grad_norm": 0.28083232045173645,
|
| 1044 |
+
"learning_rate": 1.1039415518750857e-05,
|
| 1045 |
+
"loss": 0.0207,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 8.010752688172044,
|
| 1050 |
+
"grad_norm": 0.1920163780450821,
|
| 1051 |
+
"learning_rate": 1.048843572848307e-05,
|
| 1052 |
+
"loss": 0.0214,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 8.064516129032258,
|
| 1057 |
+
"grad_norm": 0.24642567336559296,
|
| 1058 |
+
"learning_rate": 9.949945276530781e-06,
|
| 1059 |
+
"loss": 0.0225,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 8.118279569892474,
|
| 1064 |
+
"grad_norm": 0.1683078557252884,
|
| 1065 |
+
"learning_rate": 9.424114376084897e-06,
|
| 1066 |
+
"loss": 0.0213,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 8.172043010752688,
|
| 1071 |
+
"grad_norm": 0.28274253010749817,
|
| 1072 |
+
"learning_rate": 8.911109238737747e-06,
|
| 1073 |
+
"loss": 0.0214,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 8.225806451612904,
|
| 1078 |
+
"grad_norm": 0.1529790759086609,
|
| 1079 |
+
"learning_rate": 8.41109202194465e-06,
|
| 1080 |
+
"loss": 0.0201,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 8.279569892473118,
|
| 1085 |
+
"grad_norm": 0.30514660477638245,
|
| 1086 |
+
"learning_rate": 7.924220777767056e-06,
|
| 1087 |
+
"loss": 0.0216,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 8.333333333333334,
|
| 1092 |
+
"grad_norm": 0.1656785011291504,
|
| 1093 |
+
"learning_rate": 7.45064940291334e-06,
|
| 1094 |
+
"loss": 0.0187,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 8.387096774193548,
|
| 1099 |
+
"grad_norm": 0.15570750832557678,
|
| 1100 |
+
"learning_rate": 6.990527590093149e-06,
|
| 1101 |
+
"loss": 0.0211,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 8.440860215053764,
|
| 1106 |
+
"grad_norm": 0.17597723007202148,
|
| 1107 |
+
"learning_rate": 6.544000780700427e-06,
|
| 1108 |
+
"loss": 0.0214,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 8.494623655913978,
|
| 1113 |
+
"grad_norm": 0.22502322494983673,
|
| 1114 |
+
"learning_rate": 6.1112101188405545e-06,
|
| 1115 |
+
"loss": 0.0192,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 8.548387096774194,
|
| 1120 |
+
"grad_norm": 0.16484352946281433,
|
| 1121 |
+
"learning_rate": 5.692292406715505e-06,
|
| 1122 |
+
"loss": 0.0218,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 8.602150537634408,
|
| 1127 |
+
"grad_norm": 0.21047186851501465,
|
| 1128 |
+
"learning_rate": 5.287380061381609e-06,
|
| 1129 |
+
"loss": 0.0189,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 8.655913978494624,
|
| 1134 |
+
"grad_norm": 0.23087984323501587,
|
| 1135 |
+
"learning_rate": 4.896601072893497e-06,
|
| 1136 |
+
"loss": 0.0206,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 8.709677419354838,
|
| 1141 |
+
"grad_norm": 0.16594047844409943,
|
| 1142 |
+
"learning_rate": 4.520078963847241e-06,
|
| 1143 |
+
"loss": 0.0206,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 8.763440860215054,
|
| 1148 |
+
"grad_norm": 0.13561080396175385,
|
| 1149 |
+
"learning_rate": 4.157932750335725e-06,
|
| 1150 |
+
"loss": 0.0184,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 8.817204301075268,
|
| 1155 |
+
"grad_norm": 0.17827802896499634,
|
| 1156 |
+
"learning_rate": 3.810276904328447e-06,
|
| 1157 |
+
"loss": 0.0183,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 8.870967741935484,
|
| 1162 |
+
"grad_norm": 0.21718396246433258,
|
| 1163 |
+
"learning_rate": 3.4772213174877166e-06,
|
| 1164 |
+
"loss": 0.0185,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 8.924731182795698,
|
| 1169 |
+
"grad_norm": 0.19105218350887299,
|
| 1170 |
+
"learning_rate": 3.158871266432617e-06,
|
| 1171 |
+
"loss": 0.0187,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 8.978494623655914,
|
| 1176 |
+
"grad_norm": 0.20098528265953064,
|
| 1177 |
+
"learning_rate": 2.855327379461853e-06,
|
| 1178 |
+
"loss": 0.0191,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 9.03225806451613,
|
| 1183 |
+
"grad_norm": 0.20578347146511078,
|
| 1184 |
+
"learning_rate": 2.5666856047457e-06,
|
| 1185 |
+
"loss": 0.0196,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 9.086021505376344,
|
| 1190 |
+
"grad_norm": 0.23220449686050415,
|
| 1191 |
+
"learning_rate": 2.2930371799975594e-06,
|
| 1192 |
+
"loss": 0.0194,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 9.13978494623656,
|
| 1197 |
+
"grad_norm": 0.14907877147197723,
|
| 1198 |
+
"learning_rate": 2.0344686036342654e-06,
|
| 1199 |
+
"loss": 0.0184,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 9.193548387096774,
|
| 1204 |
+
"grad_norm": 0.2061292976140976,
|
| 1205 |
+
"learning_rate": 1.791061607434541e-06,
|
| 1206 |
+
"loss": 0.0215,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 9.24731182795699,
|
| 1211 |
+
"grad_norm": 0.20491990447044373,
|
| 1212 |
+
"learning_rate": 1.5628931307041128e-06,
|
| 1213 |
+
"loss": 0.0208,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 9.301075268817204,
|
| 1218 |
+
"grad_norm": 0.19176912307739258,
|
| 1219 |
+
"learning_rate": 1.3500352959557816e-06,
|
| 1220 |
+
"loss": 0.022,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 9.35483870967742,
|
| 1225 |
+
"grad_norm": 0.15755338966846466,
|
| 1226 |
+
"learning_rate": 1.1525553861118921e-06,
|
| 1227 |
+
"loss": 0.0188,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 9.408602150537634,
|
| 1232 |
+
"grad_norm": 0.14346589148044586,
|
| 1233 |
+
"learning_rate": 9.705158232367118e-07,
|
| 1234 |
+
"loss": 0.0182,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 9.46236559139785,
|
| 1239 |
+
"grad_norm": 0.1727963238954544,
|
| 1240 |
+
"learning_rate": 8.039741488052532e-07,
|
| 1241 |
+
"loss": 0.0213,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 9.516129032258064,
|
| 1246 |
+
"grad_norm": 0.2225213199853897,
|
| 1247 |
+
"learning_rate": 6.529830055147456e-07,
|
| 1248 |
+
"loss": 0.0209,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 9.56989247311828,
|
| 1253 |
+
"grad_norm": 0.14579308032989502,
|
| 1254 |
+
"learning_rate": 5.175901206446898e-07,
|
| 1255 |
+
"loss": 0.0196,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 9.623655913978494,
|
| 1260 |
+
"grad_norm": 0.16385678946971893,
|
| 1261 |
+
"learning_rate": 3.9783829097056467e-07,
|
| 1262 |
+
"loss": 0.0175,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 9.67741935483871,
|
| 1267 |
+
"grad_norm": 0.1686970442533493,
|
| 1268 |
+
"learning_rate": 2.9376536923603203e-07,
|
| 1269 |
+
"loss": 0.0185,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 9.731182795698924,
|
| 1274 |
+
"grad_norm": 0.19003795087337494,
|
| 1275 |
+
"learning_rate": 2.0540425218799064e-07,
|
| 1276 |
+
"loss": 0.0183,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 9.78494623655914,
|
| 1281 |
+
"grad_norm": 0.18928217887878418,
|
| 1282 |
+
"learning_rate": 1.3278287017809422e-07,
|
| 1283 |
+
"loss": 0.0189,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 9.838709677419354,
|
| 1288 |
+
"grad_norm": 0.1915702223777771,
|
| 1289 |
+
"learning_rate": 7.59241783341913e-08,
|
| 1290 |
+
"loss": 0.019,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 9.89247311827957,
|
| 1295 |
+
"grad_norm": 0.1374387890100479,
|
| 1296 |
+
"learning_rate": 3.48461493043184e-08,
|
| 1297 |
+
"loss": 0.0169,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 9.946236559139784,
|
| 1302 |
+
"grad_norm": 0.15227387845516205,
|
| 1303 |
+
"learning_rate": 9.561767575722024e-09,
|
| 1304 |
+
"loss": 0.0181,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 10.0,
|
| 1309 |
+
"grad_norm": 0.7939475178718567,
|
| 1310 |
+
"learning_rate": 7.902537049742087e-11,
|
| 1311 |
+
"loss": 0.0214,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 10.0,
|
| 1316 |
+
"step": 1860,
|
| 1317 |
+
"total_flos": 0.0,
|
| 1318 |
+
"train_loss": 0.04161346972950043,
|
| 1319 |
+
"train_runtime": 2200.303,
|
| 1320 |
+
"train_samples_per_second": 90.006,
|
| 1321 |
+
"train_steps_per_second": 0.845
|
| 1322 |
+
}
|
| 1323 |
+
],
|
| 1324 |
+
"logging_steps": 10,
|
| 1325 |
+
"max_steps": 1860,
|
| 1326 |
+
"num_input_tokens_seen": 0,
|
| 1327 |
+
"num_train_epochs": 10,
|
| 1328 |
+
"save_steps": 20000,
|
| 1329 |
+
"stateful_callbacks": {
|
| 1330 |
+
"TrainerControl": {
|
| 1331 |
+
"args": {
|
| 1332 |
+
"should_epoch_stop": false,
|
| 1333 |
+
"should_evaluate": false,
|
| 1334 |
+
"should_log": false,
|
| 1335 |
+
"should_save": true,
|
| 1336 |
+
"should_training_stop": true
|
| 1337 |
+
},
|
| 1338 |
+
"attributes": {}
|
| 1339 |
+
}
|
| 1340 |
+
},
|
| 1341 |
+
"total_flos": 0.0,
|
| 1342 |
+
"train_batch_size": 107,
|
| 1343 |
+
"trial_name": null,
|
| 1344 |
+
"trial_params": null
|
| 1345 |
+
}
|