Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2016 -0
trainer_state.json
ADDED
@@ -0,0 +1,2016 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 19.87323943661972,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 2822,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.07042253521126761,
|
13 |
+
"grad_norm": 5.301488399505615,
|
14 |
+
"learning_rate": 1.4084507042253523e-05,
|
15 |
+
"loss": 0.9444,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.14084507042253522,
|
20 |
+
"grad_norm": 2.110762119293213,
|
21 |
+
"learning_rate": 2.8169014084507046e-05,
|
22 |
+
"loss": 0.4977,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.2112676056338028,
|
27 |
+
"grad_norm": 1.7739307880401611,
|
28 |
+
"learning_rate": 4.225352112676056e-05,
|
29 |
+
"loss": 0.296,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.28169014084507044,
|
34 |
+
"grad_norm": 1.6081597805023193,
|
35 |
+
"learning_rate": 5.633802816901409e-05,
|
36 |
+
"loss": 0.2143,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.352112676056338,
|
41 |
+
"grad_norm": 1.9928644895553589,
|
42 |
+
"learning_rate": 7.042253521126761e-05,
|
43 |
+
"loss": 0.1679,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.4225352112676056,
|
48 |
+
"grad_norm": 0.8647320866584778,
|
49 |
+
"learning_rate": 8.450704225352113e-05,
|
50 |
+
"loss": 0.146,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.49295774647887325,
|
55 |
+
"grad_norm": 0.8146809339523315,
|
56 |
+
"learning_rate": 9.859154929577466e-05,
|
57 |
+
"loss": 0.1407,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.5633802816901409,
|
62 |
+
"grad_norm": 0.9443349838256836,
|
63 |
+
"learning_rate": 0.00011267605633802819,
|
64 |
+
"loss": 0.1268,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.6338028169014085,
|
69 |
+
"grad_norm": 1.322043776512146,
|
70 |
+
"learning_rate": 0.0001267605633802817,
|
71 |
+
"loss": 0.1215,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.704225352112676,
|
76 |
+
"grad_norm": 0.9595595598220825,
|
77 |
+
"learning_rate": 0.00014084507042253522,
|
78 |
+
"loss": 0.1024,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.7746478873239436,
|
83 |
+
"grad_norm": 0.743903636932373,
|
84 |
+
"learning_rate": 0.00015492957746478874,
|
85 |
+
"loss": 0.1012,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.8450704225352113,
|
90 |
+
"grad_norm": 0.8585343360900879,
|
91 |
+
"learning_rate": 0.00016901408450704225,
|
92 |
+
"loss": 0.0885,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.9154929577464789,
|
97 |
+
"grad_norm": 0.8720231056213379,
|
98 |
+
"learning_rate": 0.0001830985915492958,
|
99 |
+
"loss": 0.082,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.9859154929577465,
|
104 |
+
"grad_norm": 0.7080922722816467,
|
105 |
+
"learning_rate": 0.0001971830985915493,
|
106 |
+
"loss": 0.0861,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 1.056338028169014,
|
111 |
+
"grad_norm": 0.9267153143882751,
|
112 |
+
"learning_rate": 0.00019999560279257314,
|
113 |
+
"loss": 0.089,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 1.1267605633802817,
|
118 |
+
"grad_norm": 0.5979682803153992,
|
119 |
+
"learning_rate": 0.0001999777398001667,
|
120 |
+
"loss": 0.0797,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 1.1971830985915493,
|
125 |
+
"grad_norm": 0.4289480447769165,
|
126 |
+
"learning_rate": 0.00019994613865002077,
|
127 |
+
"loss": 0.0684,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 1.267605633802817,
|
132 |
+
"grad_norm": 0.41163209080696106,
|
133 |
+
"learning_rate": 0.00019990080368451776,
|
134 |
+
"loss": 0.0644,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 1.3380281690140845,
|
139 |
+
"grad_norm": 0.5105245113372803,
|
140 |
+
"learning_rate": 0.00019984174113323353,
|
141 |
+
"loss": 0.0668,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 1.408450704225352,
|
146 |
+
"grad_norm": 0.48342588543891907,
|
147 |
+
"learning_rate": 0.00019976895911208109,
|
148 |
+
"loss": 0.0607,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 1.4788732394366197,
|
153 |
+
"grad_norm": 0.7470546960830688,
|
154 |
+
"learning_rate": 0.00019968246762219565,
|
155 |
+
"loss": 0.0624,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 1.5492957746478875,
|
160 |
+
"grad_norm": 0.6888002157211304,
|
161 |
+
"learning_rate": 0.0001995822785485601,
|
162 |
+
"loss": 0.0624,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 1.619718309859155,
|
167 |
+
"grad_norm": 0.5235802531242371,
|
168 |
+
"learning_rate": 0.00019946840565837203,
|
169 |
+
"loss": 0.0597,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 1.6901408450704225,
|
174 |
+
"grad_norm": 0.5001630783081055,
|
175 |
+
"learning_rate": 0.0001993408645991519,
|
176 |
+
"loss": 0.0601,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 1.76056338028169,
|
181 |
+
"grad_norm": 2.6329758167266846,
|
182 |
+
"learning_rate": 0.00019919967289659286,
|
183 |
+
"loss": 0.0538,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 1.8309859154929577,
|
188 |
+
"grad_norm": 0.5016240477561951,
|
189 |
+
"learning_rate": 0.00019904484995215247,
|
190 |
+
"loss": 0.0554,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 1.9014084507042255,
|
195 |
+
"grad_norm": 0.4060329496860504,
|
196 |
+
"learning_rate": 0.00019887641704038688,
|
197 |
+
"loss": 0.0573,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 1.971830985915493,
|
202 |
+
"grad_norm": 0.28704750537872314,
|
203 |
+
"learning_rate": 0.00019869439730602726,
|
204 |
+
"loss": 0.0544,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 2.0422535211267605,
|
209 |
+
"grad_norm": 0.47512659430503845,
|
210 |
+
"learning_rate": 0.00019849881576079946,
|
211 |
+
"loss": 0.0539,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 2.112676056338028,
|
216 |
+
"grad_norm": 0.3640986382961273,
|
217 |
+
"learning_rate": 0.00019828969927998714,
|
218 |
+
"loss": 0.0519,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 2.183098591549296,
|
223 |
+
"grad_norm": 0.3687652349472046,
|
224 |
+
"learning_rate": 0.00019806707659873887,
|
225 |
+
"loss": 0.0539,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 2.2535211267605635,
|
230 |
+
"grad_norm": 0.260812908411026,
|
231 |
+
"learning_rate": 0.00019783097830811935,
|
232 |
+
"loss": 0.0467,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 2.323943661971831,
|
237 |
+
"grad_norm": 0.34940680861473083,
|
238 |
+
"learning_rate": 0.00019758143685090599,
|
239 |
+
"loss": 0.0467,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 2.3943661971830985,
|
244 |
+
"grad_norm": 0.42112305760383606,
|
245 |
+
"learning_rate": 0.00019731848651713085,
|
246 |
+
"loss": 0.0542,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 2.464788732394366,
|
251 |
+
"grad_norm": 0.347625195980072,
|
252 |
+
"learning_rate": 0.00019704216343936873,
|
253 |
+
"loss": 0.0459,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 2.535211267605634,
|
258 |
+
"grad_norm": 0.4011363685131073,
|
259 |
+
"learning_rate": 0.00019675250558777219,
|
260 |
+
"loss": 0.0522,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 2.6056338028169015,
|
265 |
+
"grad_norm": 0.9218616485595703,
|
266 |
+
"learning_rate": 0.00019644955276485387,
|
267 |
+
"loss": 0.0423,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 2.676056338028169,
|
272 |
+
"grad_norm": 0.5084503889083862,
|
273 |
+
"learning_rate": 0.00019613334660001724,
|
274 |
+
"loss": 0.0439,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 2.7464788732394365,
|
279 |
+
"grad_norm": 0.3616188168525696,
|
280 |
+
"learning_rate": 0.00019580393054383622,
|
281 |
+
"loss": 0.0489,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 2.816901408450704,
|
286 |
+
"grad_norm": 0.43031033873558044,
|
287 |
+
"learning_rate": 0.0001954613498620845,
|
288 |
+
"loss": 0.0441,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 2.887323943661972,
|
293 |
+
"grad_norm": 0.4344836175441742,
|
294 |
+
"learning_rate": 0.00019510565162951537,
|
295 |
+
"loss": 0.0477,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 2.9577464788732395,
|
300 |
+
"grad_norm": 0.34272074699401855,
|
301 |
+
"learning_rate": 0.00019473688472339336,
|
302 |
+
"loss": 0.0443,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 3.028169014084507,
|
307 |
+
"grad_norm": 0.3591594099998474,
|
308 |
+
"learning_rate": 0.00019435509981677762,
|
309 |
+
"loss": 0.0454,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 3.0985915492957745,
|
314 |
+
"grad_norm": 0.2543060779571533,
|
315 |
+
"learning_rate": 0.00019396034937155897,
|
316 |
+
"loss": 0.0458,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 3.169014084507042,
|
321 |
+
"grad_norm": 0.32577958703041077,
|
322 |
+
"learning_rate": 0.00019355268763125095,
|
323 |
+
"loss": 0.044,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 3.23943661971831,
|
328 |
+
"grad_norm": 0.42566004395484924,
|
329 |
+
"learning_rate": 0.00019313217061353615,
|
330 |
+
"loss": 0.047,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 3.3098591549295775,
|
335 |
+
"grad_norm": 0.4882064163684845,
|
336 |
+
"learning_rate": 0.00019269885610256865,
|
337 |
+
"loss": 0.0429,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 3.380281690140845,
|
342 |
+
"grad_norm": 0.4627586901187897,
|
343 |
+
"learning_rate": 0.00019225280364103383,
|
344 |
+
"loss": 0.0447,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 3.4507042253521125,
|
349 |
+
"grad_norm": 0.4817451536655426,
|
350 |
+
"learning_rate": 0.00019179407452196636,
|
351 |
+
"loss": 0.0404,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 3.52112676056338,
|
356 |
+
"grad_norm": 0.3949111998081207,
|
357 |
+
"learning_rate": 0.00019132273178032794,
|
358 |
+
"loss": 0.0428,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 3.591549295774648,
|
363 |
+
"grad_norm": 0.22799795866012573,
|
364 |
+
"learning_rate": 0.00019083884018434547,
|
365 |
+
"loss": 0.036,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 3.6619718309859155,
|
370 |
+
"grad_norm": 0.3933974802494049,
|
371 |
+
"learning_rate": 0.000190342466226611,
|
372 |
+
"loss": 0.0386,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 3.732394366197183,
|
377 |
+
"grad_norm": 0.8060548305511475,
|
378 |
+
"learning_rate": 0.0001898336781149451,
|
379 |
+
"loss": 0.0374,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 3.802816901408451,
|
384 |
+
"grad_norm": 0.4396229684352875,
|
385 |
+
"learning_rate": 0.000189312545763024,
|
386 |
+
"loss": 0.0444,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 3.873239436619718,
|
391 |
+
"grad_norm": 0.28071898221969604,
|
392 |
+
"learning_rate": 0.0001887791407807728,
|
393 |
+
"loss": 0.0351,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 3.943661971830986,
|
398 |
+
"grad_norm": 1.3314933776855469,
|
399 |
+
"learning_rate": 0.00018823353646452517,
|
400 |
+
"loss": 0.0378,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 4.014084507042254,
|
405 |
+
"grad_norm": 0.4084491431713104,
|
406 |
+
"learning_rate": 0.00018767580778695185,
|
407 |
+
"loss": 0.0373,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 4.084507042253521,
|
412 |
+
"grad_norm": 0.3445165753364563,
|
413 |
+
"learning_rate": 0.00018710603138675827,
|
414 |
+
"loss": 0.043,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 4.154929577464789,
|
419 |
+
"grad_norm": 0.3539944887161255,
|
420 |
+
"learning_rate": 0.0001865242855581534,
|
421 |
+
"loss": 0.0395,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 4.225352112676056,
|
426 |
+
"grad_norm": 0.3427730202674866,
|
427 |
+
"learning_rate": 0.00018593065024009146,
|
428 |
+
"loss": 0.0396,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 4.295774647887324,
|
433 |
+
"grad_norm": 0.5173912644386292,
|
434 |
+
"learning_rate": 0.00018532520700528712,
|
435 |
+
"loss": 0.0388,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 4.366197183098592,
|
440 |
+
"grad_norm": 0.302499383687973,
|
441 |
+
"learning_rate": 0.00018470803904900633,
|
442 |
+
"loss": 0.0391,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 4.436619718309859,
|
447 |
+
"grad_norm": 0.37678262591362,
|
448 |
+
"learning_rate": 0.00018407923117763462,
|
449 |
+
"loss": 0.0402,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 4.507042253521127,
|
454 |
+
"grad_norm": 0.4250379800796509,
|
455 |
+
"learning_rate": 0.00018343886979702327,
|
456 |
+
"loss": 0.0403,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 4.577464788732394,
|
461 |
+
"grad_norm": 0.30737724900245667,
|
462 |
+
"learning_rate": 0.00018278704290061645,
|
463 |
+
"loss": 0.042,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 4.647887323943662,
|
468 |
+
"grad_norm": 0.3090226650238037,
|
469 |
+
"learning_rate": 0.0001821238400573595,
|
470 |
+
"loss": 0.0355,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 4.71830985915493,
|
475 |
+
"grad_norm": 0.3264664113521576,
|
476 |
+
"learning_rate": 0.00018144935239939144,
|
477 |
+
"loss": 0.0349,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 4.788732394366197,
|
482 |
+
"grad_norm": 0.5748451352119446,
|
483 |
+
"learning_rate": 0.00018076367260952202,
|
484 |
+
"loss": 0.033,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 4.859154929577465,
|
489 |
+
"grad_norm": 0.28208673000335693,
|
490 |
+
"learning_rate": 0.0001800668949084961,
|
491 |
+
"loss": 0.0341,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 4.929577464788732,
|
496 |
+
"grad_norm": 0.41251471638679504,
|
497 |
+
"learning_rate": 0.00017935911504204645,
|
498 |
+
"loss": 0.0355,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 5.0,
|
503 |
+
"grad_norm": 0.48121050000190735,
|
504 |
+
"learning_rate": 0.0001786404302677374,
|
505 |
+
"loss": 0.0316,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 5.070422535211268,
|
510 |
+
"grad_norm": 0.45240846276283264,
|
511 |
+
"learning_rate": 0.0001779109393416001,
|
512 |
+
"loss": 0.0391,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 5.140845070422535,
|
517 |
+
"grad_norm": 0.6922265291213989,
|
518 |
+
"learning_rate": 0.0001771707425045626,
|
519 |
+
"loss": 0.0375,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 5.211267605633803,
|
524 |
+
"grad_norm": 0.3153751492500305,
|
525 |
+
"learning_rate": 0.00017641994146867516,
|
526 |
+
"loss": 0.0372,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 5.28169014084507,
|
531 |
+
"grad_norm": 0.5754887461662292,
|
532 |
+
"learning_rate": 0.00017565863940313415,
|
533 |
+
"loss": 0.0389,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 5.352112676056338,
|
538 |
+
"grad_norm": 0.5472623109817505,
|
539 |
+
"learning_rate": 0.00017488694092010506,
|
540 |
+
"loss": 0.0431,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 5.422535211267606,
|
545 |
+
"grad_norm": 0.370579332113266,
|
546 |
+
"learning_rate": 0.0001741049520603476,
|
547 |
+
"loss": 0.0367,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 5.492957746478873,
|
552 |
+
"grad_norm": 0.27683746814727783,
|
553 |
+
"learning_rate": 0.00017331278027864455,
|
554 |
+
"loss": 0.0364,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 5.563380281690141,
|
559 |
+
"grad_norm": 0.2768910527229309,
|
560 |
+
"learning_rate": 0.00017251053442903595,
|
561 |
+
"loss": 0.032,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 5.633802816901408,
|
566 |
+
"grad_norm": 0.30943918228149414,
|
567 |
+
"learning_rate": 0.00017169832474986136,
|
568 |
+
"loss": 0.0321,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 5.704225352112676,
|
573 |
+
"grad_norm": 0.36983174085617065,
|
574 |
+
"learning_rate": 0.0001708762628486117,
|
575 |
+
"loss": 0.0358,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 5.774647887323944,
|
580 |
+
"grad_norm": 0.35293862223625183,
|
581 |
+
"learning_rate": 0.0001700444616865931,
|
582 |
+
"loss": 0.0312,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 5.845070422535211,
|
587 |
+
"grad_norm": 0.3581668436527252,
|
588 |
+
"learning_rate": 0.0001692030355634046,
|
589 |
+
"loss": 0.0359,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 5.915492957746479,
|
594 |
+
"grad_norm": 0.22088691592216492,
|
595 |
+
"learning_rate": 0.00016835210010123205,
|
596 |
+
"loss": 0.031,
|
597 |
+
"step": 840
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 5.985915492957746,
|
601 |
+
"grad_norm": 0.26273787021636963,
|
602 |
+
"learning_rate": 0.00016749177222896003,
|
603 |
+
"loss": 0.0321,
|
604 |
+
"step": 850
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 6.056338028169014,
|
608 |
+
"grad_norm": 0.2742927670478821,
|
609 |
+
"learning_rate": 0.00016662217016610465,
|
610 |
+
"loss": 0.0287,
|
611 |
+
"step": 860
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 6.126760563380282,
|
615 |
+
"grad_norm": 0.25638335943222046,
|
616 |
+
"learning_rate": 0.0001657434134065686,
|
617 |
+
"loss": 0.0334,
|
618 |
+
"step": 870
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 6.197183098591549,
|
622 |
+
"grad_norm": 0.39665618538856506,
|
623 |
+
"learning_rate": 0.00016485562270222124,
|
624 |
+
"loss": 0.0339,
|
625 |
+
"step": 880
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 6.267605633802817,
|
629 |
+
"grad_norm": 0.2293689101934433,
|
630 |
+
"learning_rate": 0.00016395892004630587,
|
631 |
+
"loss": 0.0318,
|
632 |
+
"step": 890
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 6.338028169014084,
|
636 |
+
"grad_norm": 0.2861061692237854,
|
637 |
+
"learning_rate": 0.00016305342865667632,
|
638 |
+
"loss": 0.0331,
|
639 |
+
"step": 900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 6.408450704225352,
|
643 |
+
"grad_norm": 0.43932482600212097,
|
644 |
+
"learning_rate": 0.00016213927295886547,
|
645 |
+
"loss": 0.0342,
|
646 |
+
"step": 910
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 6.47887323943662,
|
650 |
+
"grad_norm": 0.5310227274894714,
|
651 |
+
"learning_rate": 0.0001612165785689874,
|
652 |
+
"loss": 0.0321,
|
653 |
+
"step": 920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 6.549295774647887,
|
657 |
+
"grad_norm": 0.2671055793762207,
|
658 |
+
"learning_rate": 0.00016028547227647647,
|
659 |
+
"loss": 0.0316,
|
660 |
+
"step": 930
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 6.619718309859155,
|
664 |
+
"grad_norm": 0.36509400606155396,
|
665 |
+
"learning_rate": 0.0001593460820266647,
|
666 |
+
"loss": 0.0357,
|
667 |
+
"step": 940
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 6.690140845070422,
|
671 |
+
"grad_norm": 0.2598735988140106,
|
672 |
+
"learning_rate": 0.00015839853690320074,
|
673 |
+
"loss": 0.0335,
|
674 |
+
"step": 950
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 6.76056338028169,
|
678 |
+
"grad_norm": 0.28324615955352783,
|
679 |
+
"learning_rate": 0.00015744296711031201,
|
680 |
+
"loss": 0.0351,
|
681 |
+
"step": 960
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 6.830985915492958,
|
685 |
+
"grad_norm": 0.2627483010292053,
|
686 |
+
"learning_rate": 0.0001564795039549131,
|
687 |
+
"loss": 0.0332,
|
688 |
+
"step": 970
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 6.901408450704225,
|
692 |
+
"grad_norm": 0.3434940278530121,
|
693 |
+
"learning_rate": 0.00015550827982856275,
|
694 |
+
"loss": 0.0296,
|
695 |
+
"step": 980
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 6.971830985915493,
|
699 |
+
"grad_norm": 0.2355423867702484,
|
700 |
+
"learning_rate": 0.00015452942818927143,
|
701 |
+
"loss": 0.0327,
|
702 |
+
"step": 990
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 7.042253521126761,
|
706 |
+
"grad_norm": 0.7167981266975403,
|
707 |
+
"learning_rate": 0.00015354308354316272,
|
708 |
+
"loss": 0.0365,
|
709 |
+
"step": 1000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 7.112676056338028,
|
713 |
+
"grad_norm": 0.33783429861068726,
|
714 |
+
"learning_rate": 0.00015254938142599046,
|
715 |
+
"loss": 0.0287,
|
716 |
+
"step": 1010
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 7.183098591549296,
|
720 |
+
"grad_norm": 0.23960083723068237,
|
721 |
+
"learning_rate": 0.0001515484583845146,
|
722 |
+
"loss": 0.0333,
|
723 |
+
"step": 1020
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 7.253521126760563,
|
727 |
+
"grad_norm": 0.15994220972061157,
|
728 |
+
"learning_rate": 0.0001505404519577379,
|
729 |
+
"loss": 0.0316,
|
730 |
+
"step": 1030
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 7.323943661971831,
|
734 |
+
"grad_norm": 0.5600018501281738,
|
735 |
+
"learning_rate": 0.00014952550065800633,
|
736 |
+
"loss": 0.0262,
|
737 |
+
"step": 1040
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 7.394366197183099,
|
741 |
+
"grad_norm": 0.2552376687526703,
|
742 |
+
"learning_rate": 0.00014850374395197603,
|
743 |
+
"loss": 0.0268,
|
744 |
+
"step": 1050
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 7.464788732394366,
|
748 |
+
"grad_norm": 0.299656480550766,
|
749 |
+
"learning_rate": 0.00014747532224144874,
|
750 |
+
"loss": 0.0249,
|
751 |
+
"step": 1060
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 7.535211267605634,
|
755 |
+
"grad_norm": 0.23284365236759186,
|
756 |
+
"learning_rate": 0.00014644037684407882,
|
757 |
+
"loss": 0.0243,
|
758 |
+
"step": 1070
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 7.605633802816901,
|
762 |
+
"grad_norm": 0.24462944269180298,
|
763 |
+
"learning_rate": 0.00014539904997395468,
|
764 |
+
"loss": 0.0301,
|
765 |
+
"step": 1080
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 7.676056338028169,
|
769 |
+
"grad_norm": 0.3585796654224396,
|
770 |
+
"learning_rate": 0.00014435148472205663,
|
771 |
+
"loss": 0.0253,
|
772 |
+
"step": 1090
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 7.746478873239437,
|
776 |
+
"grad_norm": 0.2767147421836853,
|
777 |
+
"learning_rate": 0.0001432978250365945,
|
778 |
+
"loss": 0.0274,
|
779 |
+
"step": 1100
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 7.816901408450704,
|
783 |
+
"grad_norm": 0.3437223434448242,
|
784 |
+
"learning_rate": 0.00014223821570322762,
|
785 |
+
"loss": 0.0264,
|
786 |
+
"step": 1110
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 7.887323943661972,
|
790 |
+
"grad_norm": 0.18395860493183136,
|
791 |
+
"learning_rate": 0.00014117280232516907,
|
792 |
+
"loss": 0.0288,
|
793 |
+
"step": 1120
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 7.957746478873239,
|
797 |
+
"grad_norm": 0.2312847226858139,
|
798 |
+
"learning_rate": 0.00014010173130317853,
|
799 |
+
"loss": 0.0277,
|
800 |
+
"step": 1130
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 8.028169014084508,
|
804 |
+
"grad_norm": 0.2692025601863861,
|
805 |
+
"learning_rate": 0.00013902514981544453,
|
806 |
+
"loss": 0.0272,
|
807 |
+
"step": 1140
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 8.098591549295774,
|
811 |
+
"grad_norm": 0.18669405579566956,
|
812 |
+
"learning_rate": 0.00013794320579736083,
|
813 |
+
"loss": 0.0267,
|
814 |
+
"step": 1150
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 8.169014084507042,
|
818 |
+
"grad_norm": 0.23402738571166992,
|
819 |
+
"learning_rate": 0.00013685604792119805,
|
820 |
+
"loss": 0.0274,
|
821 |
+
"step": 1160
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 8.23943661971831,
|
825 |
+
"grad_norm": 0.18943500518798828,
|
826 |
+
"learning_rate": 0.00013576382557567432,
|
827 |
+
"loss": 0.0268,
|
828 |
+
"step": 1170
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 8.309859154929578,
|
832 |
+
"grad_norm": 0.2929067015647888,
|
833 |
+
"learning_rate": 0.00013466668884542742,
|
834 |
+
"loss": 0.0257,
|
835 |
+
"step": 1180
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 8.380281690140846,
|
839 |
+
"grad_norm": 0.2052321434020996,
|
840 |
+
"learning_rate": 0.0001335647884903915,
|
841 |
+
"loss": 0.0243,
|
842 |
+
"step": 1190
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 8.450704225352112,
|
846 |
+
"grad_norm": 0.17795851826667786,
|
847 |
+
"learning_rate": 0.0001324582759250806,
|
848 |
+
"loss": 0.0217,
|
849 |
+
"step": 1200
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 8.52112676056338,
|
853 |
+
"grad_norm": 0.2322290539741516,
|
854 |
+
"learning_rate": 0.00013134730319778258,
|
855 |
+
"loss": 0.0259,
|
856 |
+
"step": 1210
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 8.591549295774648,
|
860 |
+
"grad_norm": 0.35604098439216614,
|
861 |
+
"learning_rate": 0.00013023202296966592,
|
862 |
+
"loss": 0.024,
|
863 |
+
"step": 1220
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 8.661971830985916,
|
867 |
+
"grad_norm": 0.22043980658054352,
|
868 |
+
"learning_rate": 0.000129112588493802,
|
869 |
+
"loss": 0.0266,
|
870 |
+
"step": 1230
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 8.732394366197184,
|
874 |
+
"grad_norm": 0.1932317167520523,
|
875 |
+
"learning_rate": 0.00012798915359410636,
|
876 |
+
"loss": 0.0255,
|
877 |
+
"step": 1240
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 8.80281690140845,
|
881 |
+
"grad_norm": 0.2350931167602539,
|
882 |
+
"learning_rate": 0.0001268618726442016,
|
883 |
+
"loss": 0.0234,
|
884 |
+
"step": 1250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 8.873239436619718,
|
888 |
+
"grad_norm": 0.2676509916782379,
|
889 |
+
"learning_rate": 0.00012573090054620418,
|
890 |
+
"loss": 0.0245,
|
891 |
+
"step": 1260
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 8.943661971830986,
|
895 |
+
"grad_norm": 0.3143276572227478,
|
896 |
+
"learning_rate": 0.00012459639270943944,
|
897 |
+
"loss": 0.0234,
|
898 |
+
"step": 1270
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 9.014084507042254,
|
902 |
+
"grad_norm": 0.426460325717926,
|
903 |
+
"learning_rate": 0.00012345850502908608,
|
904 |
+
"loss": 0.0237,
|
905 |
+
"step": 1280
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 9.084507042253522,
|
909 |
+
"grad_norm": 0.257888525724411,
|
910 |
+
"learning_rate": 0.00012231739386475457,
|
911 |
+
"loss": 0.0303,
|
912 |
+
"step": 1290
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 9.154929577464788,
|
916 |
+
"grad_norm": 0.3120458126068115,
|
917 |
+
"learning_rate": 0.00012117321601900114,
|
918 |
+
"loss": 0.0262,
|
919 |
+
"step": 1300
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 9.225352112676056,
|
923 |
+
"grad_norm": 0.20543567836284637,
|
924 |
+
"learning_rate": 0.00012002612871578143,
|
925 |
+
"loss": 0.0254,
|
926 |
+
"step": 1310
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 9.295774647887324,
|
930 |
+
"grad_norm": 0.3000582754611969,
|
931 |
+
"learning_rate": 0.00011887628957884596,
|
932 |
+
"loss": 0.0215,
|
933 |
+
"step": 1320
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 9.366197183098592,
|
937 |
+
"grad_norm": 0.2910151779651642,
|
938 |
+
"learning_rate": 0.00011772385661008061,
|
939 |
+
"loss": 0.0237,
|
940 |
+
"step": 1330
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 9.43661971830986,
|
944 |
+
"grad_norm": 0.1984262466430664,
|
945 |
+
"learning_rate": 0.00011656898816779538,
|
946 |
+
"loss": 0.0325,
|
947 |
+
"step": 1340
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 9.507042253521126,
|
951 |
+
"grad_norm": 0.19687724113464355,
|
952 |
+
"learning_rate": 0.00011541184294496392,
|
953 |
+
"loss": 0.0239,
|
954 |
+
"step": 1350
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 9.577464788732394,
|
958 |
+
"grad_norm": 0.22585156559944153,
|
959 |
+
"learning_rate": 0.00011425257994741723,
|
960 |
+
"loss": 0.026,
|
961 |
+
"step": 1360
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 9.647887323943662,
|
965 |
+
"grad_norm": 0.22957830131053925,
|
966 |
+
"learning_rate": 0.00011309135847199431,
|
967 |
+
"loss": 0.0231,
|
968 |
+
"step": 1370
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 9.71830985915493,
|
972 |
+
"grad_norm": 0.31176555156707764,
|
973 |
+
"learning_rate": 0.00011192833808465289,
|
974 |
+
"loss": 0.0239,
|
975 |
+
"step": 1380
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 9.788732394366198,
|
979 |
+
"grad_norm": 0.27523675560951233,
|
980 |
+
"learning_rate": 0.00011076367859854304,
|
981 |
+
"loss": 0.0208,
|
982 |
+
"step": 1390
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 9.859154929577464,
|
986 |
+
"grad_norm": 0.23288507759571075,
|
987 |
+
"learning_rate": 0.00010959754005204695,
|
988 |
+
"loss": 0.0234,
|
989 |
+
"step": 1400
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 9.929577464788732,
|
993 |
+
"grad_norm": 0.24819090962409973,
|
994 |
+
"learning_rate": 0.00010843008268678766,
|
995 |
+
"loss": 0.0226,
|
996 |
+
"step": 1410
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 10.0,
|
1000 |
+
"grad_norm": 1.4893931150436401,
|
1001 |
+
"learning_rate": 0.0001072614669256101,
|
1002 |
+
"loss": 0.0235,
|
1003 |
+
"step": 1420
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 10.070422535211268,
|
1007 |
+
"grad_norm": 0.184897780418396,
|
1008 |
+
"learning_rate": 0.00010609185335053669,
|
1009 |
+
"loss": 0.0219,
|
1010 |
+
"step": 1430
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 10.140845070422536,
|
1014 |
+
"grad_norm": 0.1725136637687683,
|
1015 |
+
"learning_rate": 0.00010492140268070177,
|
1016 |
+
"loss": 0.0225,
|
1017 |
+
"step": 1440
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 10.211267605633802,
|
1021 |
+
"grad_norm": 0.19529558718204498,
|
1022 |
+
"learning_rate": 0.00010375027575026663,
|
1023 |
+
"loss": 0.0189,
|
1024 |
+
"step": 1450
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 10.28169014084507,
|
1028 |
+
"grad_norm": 0.25408563017845154,
|
1029 |
+
"learning_rate": 0.00010257863348631895,
|
1030 |
+
"loss": 0.0226,
|
1031 |
+
"step": 1460
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 10.352112676056338,
|
1035 |
+
"grad_norm": 0.1468944549560547,
|
1036 |
+
"learning_rate": 0.0001014066368867596,
|
1037 |
+
"loss": 0.0245,
|
1038 |
+
"step": 1470
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 10.422535211267606,
|
1042 |
+
"grad_norm": 0.2108079046010971,
|
1043 |
+
"learning_rate": 0.00010023444699817939,
|
1044 |
+
"loss": 0.0208,
|
1045 |
+
"step": 1480
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 10.492957746478874,
|
1049 |
+
"grad_norm": 0.16347095370292664,
|
1050 |
+
"learning_rate": 9.906222489372946e-05,
|
1051 |
+
"loss": 0.0208,
|
1052 |
+
"step": 1490
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 10.56338028169014,
|
1056 |
+
"grad_norm": 0.7277833819389343,
|
1057 |
+
"learning_rate": 9.789013165098775e-05,
|
1058 |
+
"loss": 0.021,
|
1059 |
+
"step": 1500
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 10.633802816901408,
|
1063 |
+
"grad_norm": 0.3126179873943329,
|
1064 |
+
"learning_rate": 9.671832832982502e-05,
|
1065 |
+
"loss": 0.0232,
|
1066 |
+
"step": 1510
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 10.704225352112676,
|
1070 |
+
"grad_norm": 0.09515654295682907,
|
1071 |
+
"learning_rate": 9.554697595027334e-05,
|
1072 |
+
"loss": 0.0216,
|
1073 |
+
"step": 1520
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 10.774647887323944,
|
1077 |
+
"grad_norm": 0.20428481698036194,
|
1078 |
+
"learning_rate": 9.437623547039991e-05,
|
1079 |
+
"loss": 0.0205,
|
1080 |
+
"step": 1530
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 10.845070422535212,
|
1084 |
+
"grad_norm": 0.1914387196302414,
|
1085 |
+
"learning_rate": 9.320626776418938e-05,
|
1086 |
+
"loss": 0.0216,
|
1087 |
+
"step": 1540
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 10.915492957746478,
|
1091 |
+
"grad_norm": 0.28580597043037415,
|
1092 |
+
"learning_rate": 9.203723359943802e-05,
|
1093 |
+
"loss": 0.0206,
|
1094 |
+
"step": 1550
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 10.985915492957746,
|
1098 |
+
"grad_norm": 0.1552996188402176,
|
1099 |
+
"learning_rate": 9.086929361566206e-05,
|
1100 |
+
"loss": 0.0234,
|
1101 |
+
"step": 1560
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 11.056338028169014,
|
1105 |
+
"grad_norm": 0.15982602536678314,
|
1106 |
+
"learning_rate": 8.970260830202399e-05,
|
1107 |
+
"loss": 0.0225,
|
1108 |
+
"step": 1570
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 11.126760563380282,
|
1112 |
+
"grad_norm": 0.27742645144462585,
|
1113 |
+
"learning_rate": 8.85373379752794e-05,
|
1114 |
+
"loss": 0.023,
|
1115 |
+
"step": 1580
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 11.19718309859155,
|
1119 |
+
"grad_norm": 0.24796386063098907,
|
1120 |
+
"learning_rate": 8.73736427577473e-05,
|
1121 |
+
"loss": 0.0219,
|
1122 |
+
"step": 1590
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 11.267605633802816,
|
1126 |
+
"grad_norm": 0.15329182147979736,
|
1127 |
+
"learning_rate": 8.621168255530778e-05,
|
1128 |
+
"loss": 0.0205,
|
1129 |
+
"step": 1600
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 11.338028169014084,
|
1133 |
+
"grad_norm": 0.21230141818523407,
|
1134 |
+
"learning_rate": 8.505161703542851e-05,
|
1135 |
+
"loss": 0.0212,
|
1136 |
+
"step": 1610
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 11.408450704225352,
|
1140 |
+
"grad_norm": 0.1345348358154297,
|
1141 |
+
"learning_rate": 8.389360560522495e-05,
|
1142 |
+
"loss": 0.0169,
|
1143 |
+
"step": 1620
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 11.47887323943662,
|
1147 |
+
"grad_norm": 0.3376161456108093,
|
1148 |
+
"learning_rate": 8.273780738955544e-05,
|
1149 |
+
"loss": 0.0178,
|
1150 |
+
"step": 1630
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 11.549295774647888,
|
1154 |
+
"grad_norm": 0.17062005400657654,
|
1155 |
+
"learning_rate": 8.158438120915582e-05,
|
1156 |
+
"loss": 0.0216,
|
1157 |
+
"step": 1640
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 11.619718309859154,
|
1161 |
+
"grad_norm": 0.29981690645217896,
|
1162 |
+
"learning_rate": 8.043348555881548e-05,
|
1163 |
+
"loss": 0.0184,
|
1164 |
+
"step": 1650
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 11.690140845070422,
|
1168 |
+
"grad_norm": 0.2005750983953476,
|
1169 |
+
"learning_rate": 7.928527858559807e-05,
|
1170 |
+
"loss": 0.0176,
|
1171 |
+
"step": 1660
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 11.76056338028169,
|
1175 |
+
"grad_norm": 0.36460545659065247,
|
1176 |
+
"learning_rate": 7.813991806711039e-05,
|
1177 |
+
"loss": 0.0181,
|
1178 |
+
"step": 1670
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 11.830985915492958,
|
1182 |
+
"grad_norm": 0.19622518122196198,
|
1183 |
+
"learning_rate": 7.699756138982171e-05,
|
1184 |
+
"loss": 0.0191,
|
1185 |
+
"step": 1680
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 11.901408450704226,
|
1189 |
+
"grad_norm": 0.20806631445884705,
|
1190 |
+
"learning_rate": 7.585836552743691e-05,
|
1191 |
+
"loss": 0.0204,
|
1192 |
+
"step": 1690
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 11.971830985915492,
|
1196 |
+
"grad_norm": 0.1987942010164261,
|
1197 |
+
"learning_rate": 7.472248701932658e-05,
|
1198 |
+
"loss": 0.0229,
|
1199 |
+
"step": 1700
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 12.04225352112676,
|
1203 |
+
"grad_norm": 0.3817981481552124,
|
1204 |
+
"learning_rate": 7.359008194901632e-05,
|
1205 |
+
"loss": 0.0219,
|
1206 |
+
"step": 1710
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 12.112676056338028,
|
1210 |
+
"grad_norm": 0.5507696270942688,
|
1211 |
+
"learning_rate": 7.246130592273926e-05,
|
1212 |
+
"loss": 0.0187,
|
1213 |
+
"step": 1720
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 12.183098591549296,
|
1217 |
+
"grad_norm": 0.14364100992679596,
|
1218 |
+
"learning_rate": 7.133631404805366e-05,
|
1219 |
+
"loss": 0.0165,
|
1220 |
+
"step": 1730
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 12.253521126760564,
|
1224 |
+
"grad_norm": 0.21604971587657928,
|
1225 |
+
"learning_rate": 7.021526091252926e-05,
|
1226 |
+
"loss": 0.0211,
|
1227 |
+
"step": 1740
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 12.323943661971832,
|
1231 |
+
"grad_norm": 0.1259509027004242,
|
1232 |
+
"learning_rate": 6.909830056250527e-05,
|
1233 |
+
"loss": 0.0173,
|
1234 |
+
"step": 1750
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 12.394366197183098,
|
1238 |
+
"grad_norm": 0.27340298891067505,
|
1239 |
+
"learning_rate": 6.798558648192224e-05,
|
1240 |
+
"loss": 0.0225,
|
1241 |
+
"step": 1760
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 12.464788732394366,
|
1245 |
+
"grad_norm": 0.15757574141025543,
|
1246 |
+
"learning_rate": 6.687727157123172e-05,
|
1247 |
+
"loss": 0.0187,
|
1248 |
+
"step": 1770
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 12.535211267605634,
|
1252 |
+
"grad_norm": 0.18283964693546295,
|
1253 |
+
"learning_rate": 6.57735081263858e-05,
|
1254 |
+
"loss": 0.0175,
|
1255 |
+
"step": 1780
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 12.605633802816902,
|
1259 |
+
"grad_norm": 0.12672746181488037,
|
1260 |
+
"learning_rate": 6.467444781790966e-05,
|
1261 |
+
"loss": 0.0196,
|
1262 |
+
"step": 1790
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 12.676056338028168,
|
1266 |
+
"grad_norm": 0.16949881613254547,
|
1267 |
+
"learning_rate": 6.358024167006058e-05,
|
1268 |
+
"loss": 0.0206,
|
1269 |
+
"step": 1800
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 12.746478873239436,
|
1273 |
+
"grad_norm": 0.19527575373649597,
|
1274 |
+
"learning_rate": 6.249104004007485e-05,
|
1275 |
+
"loss": 0.0177,
|
1276 |
+
"step": 1810
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 12.816901408450704,
|
1280 |
+
"grad_norm": 0.1705390363931656,
|
1281 |
+
"learning_rate": 6.140699259750734e-05,
|
1282 |
+
"loss": 0.0158,
|
1283 |
+
"step": 1820
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 12.887323943661972,
|
1287 |
+
"grad_norm": 0.22441525757312775,
|
1288 |
+
"learning_rate": 6.032824830366466e-05,
|
1289 |
+
"loss": 0.0195,
|
1290 |
+
"step": 1830
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 12.95774647887324,
|
1294 |
+
"grad_norm": 0.13072457909584045,
|
1295 |
+
"learning_rate": 5.9254955391136435e-05,
|
1296 |
+
"loss": 0.0167,
|
1297 |
+
"step": 1840
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 13.028169014084508,
|
1301 |
+
"grad_norm": 0.44169002771377563,
|
1302 |
+
"learning_rate": 5.818726134342611e-05,
|
1303 |
+
"loss": 0.018,
|
1304 |
+
"step": 1850
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 13.098591549295774,
|
1308 |
+
"grad_norm": 0.19800615310668945,
|
1309 |
+
"learning_rate": 5.712531287468501e-05,
|
1310 |
+
"loss": 0.0195,
|
1311 |
+
"step": 1860
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 13.169014084507042,
|
1315 |
+
"grad_norm": 0.24248644709587097,
|
1316 |
+
"learning_rate": 5.606925590955199e-05,
|
1317 |
+
"loss": 0.0234,
|
1318 |
+
"step": 1870
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 13.23943661971831,
|
1322 |
+
"grad_norm": 0.1700563281774521,
|
1323 |
+
"learning_rate": 5.501923556310165e-05,
|
1324 |
+
"loss": 0.017,
|
1325 |
+
"step": 1880
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 13.309859154929578,
|
1329 |
+
"grad_norm": 0.22675108909606934,
|
1330 |
+
"learning_rate": 5.3975396120903674e-05,
|
1331 |
+
"loss": 0.02,
|
1332 |
+
"step": 1890
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 13.380281690140846,
|
1336 |
+
"grad_norm": 0.2548399567604065,
|
1337 |
+
"learning_rate": 5.293788101919652e-05,
|
1338 |
+
"loss": 0.0209,
|
1339 |
+
"step": 1900
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 13.450704225352112,
|
1343 |
+
"grad_norm": 0.139588862657547,
|
1344 |
+
"learning_rate": 5.190683282517701e-05,
|
1345 |
+
"loss": 0.0156,
|
1346 |
+
"step": 1910
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 13.52112676056338,
|
1350 |
+
"grad_norm": 0.17267008125782013,
|
1351 |
+
"learning_rate": 5.088239321741044e-05,
|
1352 |
+
"loss": 0.0172,
|
1353 |
+
"step": 1920
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 13.591549295774648,
|
1357 |
+
"grad_norm": 0.17120109498500824,
|
1358 |
+
"learning_rate": 4.986470296636184e-05,
|
1359 |
+
"loss": 0.0169,
|
1360 |
+
"step": 1930
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 13.661971830985916,
|
1364 |
+
"grad_norm": 0.1749114841222763,
|
1365 |
+
"learning_rate": 4.885390191505249e-05,
|
1366 |
+
"loss": 0.0176,
|
1367 |
+
"step": 1940
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 13.732394366197184,
|
1371 |
+
"grad_norm": 0.19507087767124176,
|
1372 |
+
"learning_rate": 4.785012895984397e-05,
|
1373 |
+
"loss": 0.0199,
|
1374 |
+
"step": 1950
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 13.80281690140845,
|
1378 |
+
"grad_norm": 0.13473351299762726,
|
1379 |
+
"learning_rate": 4.685352203135168e-05,
|
1380 |
+
"loss": 0.0145,
|
1381 |
+
"step": 1960
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 13.873239436619718,
|
1385 |
+
"grad_norm": 0.25520896911621094,
|
1386 |
+
"learning_rate": 4.5864218075492086e-05,
|
1387 |
+
"loss": 0.0154,
|
1388 |
+
"step": 1970
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 13.943661971830986,
|
1392 |
+
"grad_norm": 0.1440926343202591,
|
1393 |
+
"learning_rate": 4.488235303466414e-05,
|
1394 |
+
"loss": 0.0165,
|
1395 |
+
"step": 1980
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 14.014084507042254,
|
1399 |
+
"grad_norm": 0.19918571412563324,
|
1400 |
+
"learning_rate": 4.390806182906946e-05,
|
1401 |
+
"loss": 0.0186,
|
1402 |
+
"step": 1990
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 14.084507042253522,
|
1406 |
+
"grad_norm": 0.18236444890499115,
|
1407 |
+
"learning_rate": 4.2941478338172546e-05,
|
1408 |
+
"loss": 0.015,
|
1409 |
+
"step": 2000
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 14.154929577464788,
|
1413 |
+
"grad_norm": 0.3441983163356781,
|
1414 |
+
"learning_rate": 4.198273538230411e-05,
|
1415 |
+
"loss": 0.0155,
|
1416 |
+
"step": 2010
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 14.225352112676056,
|
1420 |
+
"grad_norm": 0.16762350499629974,
|
1421 |
+
"learning_rate": 4.103196470441004e-05,
|
1422 |
+
"loss": 0.0161,
|
1423 |
+
"step": 2020
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 14.295774647887324,
|
1427 |
+
"grad_norm": 0.11081443727016449,
|
1428 |
+
"learning_rate": 4.008929695194819e-05,
|
1429 |
+
"loss": 0.0146,
|
1430 |
+
"step": 2030
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 14.366197183098592,
|
1434 |
+
"grad_norm": 0.14331316947937012,
|
1435 |
+
"learning_rate": 3.915486165893584e-05,
|
1436 |
+
"loss": 0.0152,
|
1437 |
+
"step": 2040
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 14.43661971830986,
|
1441 |
+
"grad_norm": 0.2052135467529297,
|
1442 |
+
"learning_rate": 3.8228787228150256e-05,
|
1443 |
+
"loss": 0.0156,
|
1444 |
+
"step": 2050
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 14.507042253521126,
|
1448 |
+
"grad_norm": 0.2391861230134964,
|
1449 |
+
"learning_rate": 3.731120091348447e-05,
|
1450 |
+
"loss": 0.0168,
|
1451 |
+
"step": 2060
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 14.577464788732394,
|
1455 |
+
"grad_norm": 0.15410864353179932,
|
1456 |
+
"learning_rate": 3.640222880246117e-05,
|
1457 |
+
"loss": 0.0148,
|
1458 |
+
"step": 2070
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 14.647887323943662,
|
1462 |
+
"grad_norm": 0.39798498153686523,
|
1463 |
+
"learning_rate": 3.550199579890658e-05,
|
1464 |
+
"loss": 0.016,
|
1465 |
+
"step": 2080
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 14.71830985915493,
|
1469 |
+
"grad_norm": 0.21030154824256897,
|
1470 |
+
"learning_rate": 3.461062560578732e-05,
|
1471 |
+
"loss": 0.0152,
|
1472 |
+
"step": 2090
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 14.788732394366198,
|
1476 |
+
"grad_norm": 0.23185010254383087,
|
1477 |
+
"learning_rate": 3.372824070821194e-05,
|
1478 |
+
"loss": 0.0167,
|
1479 |
+
"step": 2100
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 14.859154929577464,
|
1483 |
+
"grad_norm": 0.16405650973320007,
|
1484 |
+
"learning_rate": 3.2854962356600124e-05,
|
1485 |
+
"loss": 0.0175,
|
1486 |
+
"step": 2110
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 14.929577464788732,
|
1490 |
+
"grad_norm": 0.28342190384864807,
|
1491 |
+
"learning_rate": 3.199091055002143e-05,
|
1492 |
+
"loss": 0.0159,
|
1493 |
+
"step": 2120
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 15.0,
|
1497 |
+
"grad_norm": 0.6678891181945801,
|
1498 |
+
"learning_rate": 3.1136204019705726e-05,
|
1499 |
+
"loss": 0.0122,
|
1500 |
+
"step": 2130
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 15.070422535211268,
|
1504 |
+
"grad_norm": 0.1747487485408783,
|
1505 |
+
"learning_rate": 3.029096021272836e-05,
|
1506 |
+
"loss": 0.0163,
|
1507 |
+
"step": 2140
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 15.140845070422536,
|
1511 |
+
"grad_norm": 0.11703847348690033,
|
1512 |
+
"learning_rate": 2.9455295275871298e-05,
|
1513 |
+
"loss": 0.014,
|
1514 |
+
"step": 2150
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 15.211267605633802,
|
1518 |
+
"grad_norm": 0.19800235331058502,
|
1519 |
+
"learning_rate": 2.862932403966322e-05,
|
1520 |
+
"loss": 0.0128,
|
1521 |
+
"step": 2160
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 15.28169014084507,
|
1525 |
+
"grad_norm": 0.390919953584671,
|
1526 |
+
"learning_rate": 2.7813160002600524e-05,
|
1527 |
+
"loss": 0.0146,
|
1528 |
+
"step": 2170
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 15.352112676056338,
|
1532 |
+
"grad_norm": 0.2693049907684326,
|
1533 |
+
"learning_rate": 2.7006915315550918e-05,
|
1534 |
+
"loss": 0.0134,
|
1535 |
+
"step": 2180
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 15.422535211267606,
|
1539 |
+
"grad_norm": 0.11595863848924637,
|
1540 |
+
"learning_rate": 2.6210700766342965e-05,
|
1541 |
+
"loss": 0.0129,
|
1542 |
+
"step": 2190
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 15.492957746478874,
|
1546 |
+
"grad_norm": 0.39005428552627563,
|
1547 |
+
"learning_rate": 2.5424625764542132e-05,
|
1548 |
+
"loss": 0.0134,
|
1549 |
+
"step": 2200
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 15.56338028169014,
|
1553 |
+
"grad_norm": 0.24683450162410736,
|
1554 |
+
"learning_rate": 2.4648798326416745e-05,
|
1555 |
+
"loss": 0.0162,
|
1556 |
+
"step": 2210
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 15.633802816901408,
|
1560 |
+
"grad_norm": 0.15063706040382385,
|
1561 |
+
"learning_rate": 2.388332506009524e-05,
|
1562 |
+
"loss": 0.0145,
|
1563 |
+
"step": 2220
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 15.704225352112676,
|
1567 |
+
"grad_norm": 0.16000519692897797,
|
1568 |
+
"learning_rate": 2.3128311150916826e-05,
|
1569 |
+
"loss": 0.0179,
|
1570 |
+
"step": 2230
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 15.774647887323944,
|
1574 |
+
"grad_norm": 0.3279981017112732,
|
1575 |
+
"learning_rate": 2.2383860346977904e-05,
|
1576 |
+
"loss": 0.0169,
|
1577 |
+
"step": 2240
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 15.845070422535212,
|
1581 |
+
"grad_norm": 0.1328040212392807,
|
1582 |
+
"learning_rate": 2.1650074944875587e-05,
|
1583 |
+
"loss": 0.0137,
|
1584 |
+
"step": 2250
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 15.915492957746478,
|
1588 |
+
"grad_norm": 0.15805411338806152,
|
1589 |
+
"learning_rate": 2.0927055775651104e-05,
|
1590 |
+
"loss": 0.0145,
|
1591 |
+
"step": 2260
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 15.985915492957746,
|
1595 |
+
"grad_norm": 0.13186223804950714,
|
1596 |
+
"learning_rate": 2.021490219093426e-05,
|
1597 |
+
"loss": 0.0186,
|
1598 |
+
"step": 2270
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 16.056338028169016,
|
1602 |
+
"grad_norm": 0.5759530067443848,
|
1603 |
+
"learning_rate": 1.9513712049291367e-05,
|
1604 |
+
"loss": 0.0137,
|
1605 |
+
"step": 2280
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 16.12676056338028,
|
1609 |
+
"grad_norm": 0.15024252235889435,
|
1610 |
+
"learning_rate": 1.882358170277835e-05,
|
1611 |
+
"loss": 0.0124,
|
1612 |
+
"step": 2290
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 16.197183098591548,
|
1616 |
+
"grad_norm": 0.3911398649215698,
|
1617 |
+
"learning_rate": 1.814460598370066e-05,
|
1618 |
+
"loss": 0.014,
|
1619 |
+
"step": 2300
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 16.267605633802816,
|
1623 |
+
"grad_norm": 0.4579143226146698,
|
1624 |
+
"learning_rate": 1.7476878191582246e-05,
|
1625 |
+
"loss": 0.0128,
|
1626 |
+
"step": 2310
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 16.338028169014084,
|
1630 |
+
"grad_norm": 0.24127140641212463,
|
1631 |
+
"learning_rate": 1.6820490080345075e-05,
|
1632 |
+
"loss": 0.0142,
|
1633 |
+
"step": 2320
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 16.408450704225352,
|
1637 |
+
"grad_norm": 0.14859215915203094,
|
1638 |
+
"learning_rate": 1.6175531845700964e-05,
|
1639 |
+
"loss": 0.0151,
|
1640 |
+
"step": 2330
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 16.47887323943662,
|
1644 |
+
"grad_norm": 0.167475625872612,
|
1645 |
+
"learning_rate": 1.5542092112757734e-05,
|
1646 |
+
"loss": 0.0134,
|
1647 |
+
"step": 2340
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 16.549295774647888,
|
1651 |
+
"grad_norm": 0.16596587002277374,
|
1652 |
+
"learning_rate": 1.4920257923840864e-05,
|
1653 |
+
"loss": 0.0171,
|
1654 |
+
"step": 2350
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 16.619718309859156,
|
1658 |
+
"grad_norm": 0.14397069811820984,
|
1659 |
+
"learning_rate": 1.4310114726532831e-05,
|
1660 |
+
"loss": 0.0138,
|
1661 |
+
"step": 2360
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 16.690140845070424,
|
1665 |
+
"grad_norm": 0.1877891719341278,
|
1666 |
+
"learning_rate": 1.371174636193181e-05,
|
1667 |
+
"loss": 0.0155,
|
1668 |
+
"step": 2370
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 16.760563380281692,
|
1672 |
+
"grad_norm": 0.13636761903762817,
|
1673 |
+
"learning_rate": 1.3125235053130547e-05,
|
1674 |
+
"loss": 0.0127,
|
1675 |
+
"step": 2380
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 16.830985915492956,
|
1679 |
+
"grad_norm": 0.297732412815094,
|
1680 |
+
"learning_rate": 1.2550661393918217e-05,
|
1681 |
+
"loss": 0.013,
|
1682 |
+
"step": 2390
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 16.901408450704224,
|
1686 |
+
"grad_norm": 0.1112903356552124,
|
1687 |
+
"learning_rate": 1.1988104337705574e-05,
|
1688 |
+
"loss": 0.0117,
|
1689 |
+
"step": 2400
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 16.971830985915492,
|
1693 |
+
"grad_norm": 0.13235154747962952,
|
1694 |
+
"learning_rate": 1.1437641186676006e-05,
|
1695 |
+
"loss": 0.0129,
|
1696 |
+
"step": 2410
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 17.04225352112676,
|
1700 |
+
"grad_norm": 0.19815203547477722,
|
1701 |
+
"learning_rate": 1.0899347581163221e-05,
|
1702 |
+
"loss": 0.0148,
|
1703 |
+
"step": 2420
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 17.112676056338028,
|
1707 |
+
"grad_norm": 0.4887934625148773,
|
1708 |
+
"learning_rate": 1.0373297489257272e-05,
|
1709 |
+
"loss": 0.0125,
|
1710 |
+
"step": 2430
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 17.183098591549296,
|
1714 |
+
"grad_norm": 0.12783914804458618,
|
1715 |
+
"learning_rate": 9.85956319664051e-06,
|
1716 |
+
"loss": 0.018,
|
1717 |
+
"step": 2440
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 17.253521126760564,
|
1721 |
+
"grad_norm": 0.2081880420446396,
|
1722 |
+
"learning_rate": 9.358215296654637e-06,
|
1723 |
+
"loss": 0.0126,
|
1724 |
+
"step": 2450
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 17.323943661971832,
|
1728 |
+
"grad_norm": 0.22537609934806824,
|
1729 |
+
"learning_rate": 8.869322680600289e-06,
|
1730 |
+
"loss": 0.0129,
|
1731 |
+
"step": 2460
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 17.3943661971831,
|
1735 |
+
"grad_norm": 0.5902778506278992,
|
1736 |
+
"learning_rate": 8.392952528270659e-06,
|
1737 |
+
"loss": 0.0145,
|
1738 |
+
"step": 2470
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 17.464788732394368,
|
1742 |
+
"grad_norm": 0.12127131968736649,
|
1743 |
+
"learning_rate": 7.929170298719969e-06,
|
1744 |
+
"loss": 0.0115,
|
1745 |
+
"step": 2480
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 17.535211267605632,
|
1749 |
+
"grad_norm": 0.16380423307418823,
|
1750 |
+
"learning_rate": 7.478039721268759e-06,
|
1751 |
+
"loss": 0.0131,
|
1752 |
+
"step": 2490
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 17.6056338028169,
|
1756 |
+
"grad_norm": 0.17874068021774292,
|
1757 |
+
"learning_rate": 7.039622786746658e-06,
|
1758 |
+
"loss": 0.0142,
|
1759 |
+
"step": 2500
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 17.676056338028168,
|
1763 |
+
"grad_norm": 0.18840940296649933,
|
1764 |
+
"learning_rate": 6.613979738974074e-06,
|
1765 |
+
"loss": 0.0131,
|
1766 |
+
"step": 2510
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 17.746478873239436,
|
1770 |
+
"grad_norm": 0.25764134526252747,
|
1771 |
+
"learning_rate": 6.2011690664840025e-06,
|
1772 |
+
"loss": 0.014,
|
1773 |
+
"step": 2520
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 17.816901408450704,
|
1777 |
+
"grad_norm": 0.15851202607154846,
|
1778 |
+
"learning_rate": 5.801247494484929e-06,
|
1779 |
+
"loss": 0.0144,
|
1780 |
+
"step": 2530
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 17.887323943661972,
|
1784 |
+
"grad_norm": 0.12299610674381256,
|
1785 |
+
"learning_rate": 5.414269977066133e-06,
|
1786 |
+
"loss": 0.0126,
|
1787 |
+
"step": 2540
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 17.95774647887324,
|
1791 |
+
"grad_norm": 0.09734787046909332,
|
1792 |
+
"learning_rate": 5.040289689646338e-06,
|
1793 |
+
"loss": 0.0116,
|
1794 |
+
"step": 2550
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 18.028169014084508,
|
1798 |
+
"grad_norm": 0.11694216728210449,
|
1799 |
+
"learning_rate": 4.679358021666725e-06,
|
1800 |
+
"loss": 0.0119,
|
1801 |
+
"step": 2560
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 18.098591549295776,
|
1805 |
+
"grad_norm": 0.2159486711025238,
|
1806 |
+
"learning_rate": 4.3315245695294545e-06,
|
1807 |
+
"loss": 0.0125,
|
1808 |
+
"step": 2570
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 18.169014084507044,
|
1812 |
+
"grad_norm": 0.13054227828979492,
|
1813 |
+
"learning_rate": 3.996837129782416e-06,
|
1814 |
+
"loss": 0.0131,
|
1815 |
+
"step": 2580
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 18.239436619718308,
|
1819 |
+
"grad_norm": 0.4224683940410614,
|
1820 |
+
"learning_rate": 3.6753416925515593e-06,
|
1821 |
+
"loss": 0.0091,
|
1822 |
+
"step": 2590
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 18.309859154929576,
|
1826 |
+
"grad_norm": 0.7623445987701416,
|
1827 |
+
"learning_rate": 3.367082435221125e-06,
|
1828 |
+
"loss": 0.0113,
|
1829 |
+
"step": 2600
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 18.380281690140844,
|
1833 |
+
"grad_norm": 0.10136809200048447,
|
1834 |
+
"learning_rate": 3.0721017163632294e-06,
|
1835 |
+
"loss": 0.0136,
|
1836 |
+
"step": 2610
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 18.450704225352112,
|
1840 |
+
"grad_norm": 0.11303048580884933,
|
1841 |
+
"learning_rate": 2.7904400699172994e-06,
|
1842 |
+
"loss": 0.0137,
|
1843 |
+
"step": 2620
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 18.52112676056338,
|
1847 |
+
"grad_norm": 0.181919127702713,
|
1848 |
+
"learning_rate": 2.5221361996200955e-06,
|
1849 |
+
"loss": 0.0128,
|
1850 |
+
"step": 2630
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 18.591549295774648,
|
1854 |
+
"grad_norm": 0.4242732524871826,
|
1855 |
+
"learning_rate": 2.2672269736875087e-06,
|
1856 |
+
"loss": 0.0113,
|
1857 |
+
"step": 2640
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 18.661971830985916,
|
1861 |
+
"grad_norm": 0.07075098901987076,
|
1862 |
+
"learning_rate": 2.025747419748314e-06,
|
1863 |
+
"loss": 0.0136,
|
1864 |
+
"step": 2650
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 18.732394366197184,
|
1868 |
+
"grad_norm": 0.1898496001958847,
|
1869 |
+
"learning_rate": 1.7977307200309546e-06,
|
1870 |
+
"loss": 0.0128,
|
1871 |
+
"step": 2660
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 18.802816901408452,
|
1875 |
+
"grad_norm": 0.1816607266664505,
|
1876 |
+
"learning_rate": 1.5832082068039544e-06,
|
1877 |
+
"loss": 0.0122,
|
1878 |
+
"step": 2670
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 18.87323943661972,
|
1882 |
+
"grad_norm": 0.07458122074604034,
|
1883 |
+
"learning_rate": 1.3822093580703966e-06,
|
1884 |
+
"loss": 0.0116,
|
1885 |
+
"step": 2680
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 18.943661971830984,
|
1889 |
+
"grad_norm": 0.11320627480745316,
|
1890 |
+
"learning_rate": 1.1947617935174183e-06,
|
1891 |
+
"loss": 0.0124,
|
1892 |
+
"step": 2690
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 19.014084507042252,
|
1896 |
+
"grad_norm": 0.13373975455760956,
|
1897 |
+
"learning_rate": 1.0208912707207697e-06,
|
1898 |
+
"loss": 0.0123,
|
1899 |
+
"step": 2700
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 19.08450704225352,
|
1903 |
+
"grad_norm": 0.11728695034980774,
|
1904 |
+
"learning_rate": 8.606216816055334e-07,
|
1905 |
+
"loss": 0.0126,
|
1906 |
+
"step": 2710
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 19.154929577464788,
|
1910 |
+
"grad_norm": 0.16200479865074158,
|
1911 |
+
"learning_rate": 7.139750491630404e-07,
|
1912 |
+
"loss": 0.0127,
|
1913 |
+
"step": 2720
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 19.225352112676056,
|
1917 |
+
"grad_norm": 0.15322642028331757,
|
1918 |
+
"learning_rate": 5.809715244246228e-07,
|
1919 |
+
"loss": 0.0108,
|
1920 |
+
"step": 2730
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 19.295774647887324,
|
1924 |
+
"grad_norm": 0.1876576840877533,
|
1925 |
+
"learning_rate": 4.616293836926633e-07,
|
1926 |
+
"loss": 0.0145,
|
1927 |
+
"step": 2740
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 19.366197183098592,
|
1931 |
+
"grad_norm": 0.2538888156414032,
|
1932 |
+
"learning_rate": 3.5596502602917027e-07,
|
1933 |
+
"loss": 0.0122,
|
1934 |
+
"step": 2750
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 19.43661971830986,
|
1938 |
+
"grad_norm": 0.09630775451660156,
|
1939 |
+
"learning_rate": 2.639929710023359e-07,
|
1940 |
+
"loss": 0.0114,
|
1941 |
+
"step": 2760
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 19.507042253521128,
|
1945 |
+
"grad_norm": 0.15465404093265533,
|
1946 |
+
"learning_rate": 1.8572585669144328e-07,
|
1947 |
+
"loss": 0.0113,
|
1948 |
+
"step": 2770
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 19.577464788732396,
|
1952 |
+
"grad_norm": 0.10793367773294449,
|
1953 |
+
"learning_rate": 1.211744379501556e-07,
|
1954 |
+
"loss": 0.0108,
|
1955 |
+
"step": 2780
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 19.647887323943664,
|
1959 |
+
"grad_norm": 0.15497437119483948,
|
1960 |
+
"learning_rate": 7.034758492872052e-08,
|
1961 |
+
"loss": 0.0129,
|
1962 |
+
"step": 2790
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 19.718309859154928,
|
1966 |
+
"grad_norm": 0.16003310680389404,
|
1967 |
+
"learning_rate": 3.3252281855067345e-08,
|
1968 |
+
"loss": 0.0112,
|
1969 |
+
"step": 2800
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 19.788732394366196,
|
1973 |
+
"grad_norm": 0.2295142561197281,
|
1974 |
+
"learning_rate": 9.89362607515254e-09,
|
1975 |
+
"loss": 0.0122,
|
1976 |
+
"step": 2810
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 19.859154929577464,
|
1980 |
+
"grad_norm": 0.4385908246040344,
|
1981 |
+
"learning_rate": 2.748273524311173e-10,
|
1982 |
+
"loss": 0.0137,
|
1983 |
+
"step": 2820
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 19.87323943661972,
|
1987 |
+
"step": 2822,
|
1988 |
+
"total_flos": 3.9569565333896294e+17,
|
1989 |
+
"train_loss": 0.03605978149646123,
|
1990 |
+
"train_runtime": 3188.1566,
|
1991 |
+
"train_samples_per_second": 56.65,
|
1992 |
+
"train_steps_per_second": 0.885
|
1993 |
+
}
|
1994 |
+
],
|
1995 |
+
"logging_steps": 10,
|
1996 |
+
"max_steps": 2822,
|
1997 |
+
"num_input_tokens_seen": 0,
|
1998 |
+
"num_train_epochs": 20,
|
1999 |
+
"save_steps": 10000,
|
2000 |
+
"stateful_callbacks": {
|
2001 |
+
"TrainerControl": {
|
2002 |
+
"args": {
|
2003 |
+
"should_epoch_stop": false,
|
2004 |
+
"should_evaluate": false,
|
2005 |
+
"should_log": false,
|
2006 |
+
"should_save": true,
|
2007 |
+
"should_training_stop": true
|
2008 |
+
},
|
2009 |
+
"attributes": {}
|
2010 |
+
}
|
2011 |
+
},
|
2012 |
+
"total_flos": 3.9569565333896294e+17,
|
2013 |
+
"train_batch_size": 64,
|
2014 |
+
"trial_name": null,
|
2015 |
+
"trial_params": null
|
2016 |
+
}
|