Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2730 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2730 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.996101364522417,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 384,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.005198180636777128,
|
| 13 |
+
"grad_norm": 46.50929641723633,
|
| 14 |
+
"learning_rate": 8.333333333333333e-08,
|
| 15 |
+
"loss": 1.4958,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.010396361273554255,
|
| 20 |
+
"grad_norm": 36.618709564208984,
|
| 21 |
+
"learning_rate": 1.6666666666666665e-07,
|
| 22 |
+
"loss": 1.2222,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.015594541910331383,
|
| 27 |
+
"grad_norm": 42.09670639038086,
|
| 28 |
+
"learning_rate": 2.5e-07,
|
| 29 |
+
"loss": 1.2627,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.02079272254710851,
|
| 34 |
+
"grad_norm": 38.13116455078125,
|
| 35 |
+
"learning_rate": 3.333333333333333e-07,
|
| 36 |
+
"loss": 1.2579,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.02599090318388564,
|
| 41 |
+
"grad_norm": 40.15380859375,
|
| 42 |
+
"learning_rate": 4.1666666666666667e-07,
|
| 43 |
+
"loss": 1.2421,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.031189083820662766,
|
| 48 |
+
"grad_norm": 43.696563720703125,
|
| 49 |
+
"learning_rate": 5e-07,
|
| 50 |
+
"loss": 1.264,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.036387264457439894,
|
| 55 |
+
"grad_norm": 44.71561813354492,
|
| 56 |
+
"learning_rate": 5.833333333333334e-07,
|
| 57 |
+
"loss": 1.3881,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.04158544509421702,
|
| 62 |
+
"grad_norm": 41.07535171508789,
|
| 63 |
+
"learning_rate": 6.666666666666666e-07,
|
| 64 |
+
"loss": 1.1705,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.04678362573099415,
|
| 69 |
+
"grad_norm": 37.13037109375,
|
| 70 |
+
"learning_rate": 7.5e-07,
|
| 71 |
+
"loss": 1.1108,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.05198180636777128,
|
| 76 |
+
"grad_norm": 39.47488021850586,
|
| 77 |
+
"learning_rate": 8.333333333333333e-07,
|
| 78 |
+
"loss": 1.1024,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.057179987004548405,
|
| 83 |
+
"grad_norm": 35.2398681640625,
|
| 84 |
+
"learning_rate": 9.166666666666665e-07,
|
| 85 |
+
"loss": 1.0848,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.06237816764132553,
|
| 90 |
+
"grad_norm": 26.39617347717285,
|
| 91 |
+
"learning_rate": 1e-06,
|
| 92 |
+
"loss": 0.7678,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.06757634827810266,
|
| 97 |
+
"grad_norm": 23.569713592529297,
|
| 98 |
+
"learning_rate": 9.999821700020548e-07,
|
| 99 |
+
"loss": 0.7753,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.07277452891487979,
|
| 104 |
+
"grad_norm": 20.85965919494629,
|
| 105 |
+
"learning_rate": 9.99928681279855e-07,
|
| 106 |
+
"loss": 0.6663,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.07797270955165692,
|
| 111 |
+
"grad_norm": 19.964326858520508,
|
| 112 |
+
"learning_rate": 9.998395376482152e-07,
|
| 113 |
+
"loss": 0.5468,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.08317089018843404,
|
| 118 |
+
"grad_norm": 10.887548446655273,
|
| 119 |
+
"learning_rate": 9.997147454648588e-07,
|
| 120 |
+
"loss": 0.4754,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.08836907082521117,
|
| 125 |
+
"grad_norm": 10.726633071899414,
|
| 126 |
+
"learning_rate": 9.995543136299635e-07,
|
| 127 |
+
"loss": 0.4547,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.0935672514619883,
|
| 132 |
+
"grad_norm": 9.673012733459473,
|
| 133 |
+
"learning_rate": 9.993582535855263e-07,
|
| 134 |
+
"loss": 0.4634,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.09876543209876543,
|
| 139 |
+
"grad_norm": 7.258286476135254,
|
| 140 |
+
"learning_rate": 9.991265793145479e-07,
|
| 141 |
+
"loss": 0.3635,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.10396361273554255,
|
| 146 |
+
"grad_norm": 5.402667999267578,
|
| 147 |
+
"learning_rate": 9.988593073400354e-07,
|
| 148 |
+
"loss": 0.356,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.10916179337231968,
|
| 153 |
+
"grad_norm": 4.859364032745361,
|
| 154 |
+
"learning_rate": 9.985564567238236e-07,
|
| 155 |
+
"loss": 0.3692,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.11435997400909681,
|
| 160 |
+
"grad_norm": 3.8276686668395996,
|
| 161 |
+
"learning_rate": 9.982180490652164e-07,
|
| 162 |
+
"loss": 0.2976,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.11955815464587394,
|
| 167 |
+
"grad_norm": 3.0185964107513428,
|
| 168 |
+
"learning_rate": 9.97844108499445e-07,
|
| 169 |
+
"loss": 0.2635,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.12475633528265107,
|
| 174 |
+
"grad_norm": 2.6632726192474365,
|
| 175 |
+
"learning_rate": 9.974346616959475e-07,
|
| 176 |
+
"loss": 0.3086,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.1299545159194282,
|
| 181 |
+
"grad_norm": 2.5440852642059326,
|
| 182 |
+
"learning_rate": 9.969897378564667e-07,
|
| 183 |
+
"loss": 0.2746,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.13515269655620532,
|
| 188 |
+
"grad_norm": 2.642413377761841,
|
| 189 |
+
"learning_rate": 9.965093687129667e-07,
|
| 190 |
+
"loss": 0.2889,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.14035087719298245,
|
| 195 |
+
"grad_norm": 2.513338565826416,
|
| 196 |
+
"learning_rate": 9.959935885253715e-07,
|
| 197 |
+
"loss": 0.2778,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.14554905782975958,
|
| 202 |
+
"grad_norm": 2.2585127353668213,
|
| 203 |
+
"learning_rate": 9.954424340791195e-07,
|
| 204 |
+
"loss": 0.2311,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.1507472384665367,
|
| 209 |
+
"grad_norm": 2.021958351135254,
|
| 210 |
+
"learning_rate": 9.948559446825411e-07,
|
| 211 |
+
"loss": 0.2403,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.15594541910331383,
|
| 216 |
+
"grad_norm": 2.938659429550171,
|
| 217 |
+
"learning_rate": 9.942341621640557e-07,
|
| 218 |
+
"loss": 0.2984,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.16114359974009096,
|
| 223 |
+
"grad_norm": 1.9811211824417114,
|
| 224 |
+
"learning_rate": 9.93577130869187e-07,
|
| 225 |
+
"loss": 0.2607,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.1663417803768681,
|
| 230 |
+
"grad_norm": 1.8804433345794678,
|
| 231 |
+
"learning_rate": 9.928848976574018e-07,
|
| 232 |
+
"loss": 0.2236,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.17153996101364521,
|
| 237 |
+
"grad_norm": 2.2095425128936768,
|
| 238 |
+
"learning_rate": 9.921575118987671e-07,
|
| 239 |
+
"loss": 0.247,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.17673814165042234,
|
| 244 |
+
"grad_norm": 2.0361135005950928,
|
| 245 |
+
"learning_rate": 9.91395025470429e-07,
|
| 246 |
+
"loss": 0.2519,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.18193632228719947,
|
| 251 |
+
"grad_norm": 1.9882704019546509,
|
| 252 |
+
"learning_rate": 9.905974927529133e-07,
|
| 253 |
+
"loss": 0.2387,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.1871345029239766,
|
| 258 |
+
"grad_norm": 2.1970348358154297,
|
| 259 |
+
"learning_rate": 9.897649706262473e-07,
|
| 260 |
+
"loss": 0.2506,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.19233268356075373,
|
| 265 |
+
"grad_norm": 1.9535129070281982,
|
| 266 |
+
"learning_rate": 9.888975184659016e-07,
|
| 267 |
+
"loss": 0.2491,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.19753086419753085,
|
| 272 |
+
"grad_norm": 1.7368297576904297,
|
| 273 |
+
"learning_rate": 9.879951981385577e-07,
|
| 274 |
+
"loss": 0.2002,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.20272904483430798,
|
| 279 |
+
"grad_norm": 1.8394443988800049,
|
| 280 |
+
"learning_rate": 9.870580739976935e-07,
|
| 281 |
+
"loss": 0.2107,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2079272254710851,
|
| 286 |
+
"grad_norm": 2.0104002952575684,
|
| 287 |
+
"learning_rate": 9.860862128789952e-07,
|
| 288 |
+
"loss": 0.2373,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.21312540610786224,
|
| 293 |
+
"grad_norm": 1.7731590270996094,
|
| 294 |
+
"learning_rate": 9.850796840955899e-07,
|
| 295 |
+
"loss": 0.1881,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.21832358674463936,
|
| 300 |
+
"grad_norm": 1.937873363494873,
|
| 301 |
+
"learning_rate": 9.840385594331022e-07,
|
| 302 |
+
"loss": 0.2238,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2235217673814165,
|
| 307 |
+
"grad_norm": 1.9435638189315796,
|
| 308 |
+
"learning_rate": 9.82962913144534e-07,
|
| 309 |
+
"loss": 0.2237,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.22871994801819362,
|
| 314 |
+
"grad_norm": 1.786537766456604,
|
| 315 |
+
"learning_rate": 9.818528219449704e-07,
|
| 316 |
+
"loss": 0.1951,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.23391812865497075,
|
| 321 |
+
"grad_norm": 1.7915631532669067,
|
| 322 |
+
"learning_rate": 9.807083650061062e-07,
|
| 323 |
+
"loss": 0.2257,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.23911630929174787,
|
| 328 |
+
"grad_norm": 1.798353910446167,
|
| 329 |
+
"learning_rate": 9.79529623950601e-07,
|
| 330 |
+
"loss": 0.236,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.244314489928525,
|
| 335 |
+
"grad_norm": 1.872049331665039,
|
| 336 |
+
"learning_rate": 9.783166828462572e-07,
|
| 337 |
+
"loss": 0.2354,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.24951267056530213,
|
| 342 |
+
"grad_norm": 1.879210114479065,
|
| 343 |
+
"learning_rate": 9.770696282000244e-07,
|
| 344 |
+
"loss": 0.2229,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.25471085120207926,
|
| 349 |
+
"grad_norm": 1.9663130044937134,
|
| 350 |
+
"learning_rate": 9.757885489518296e-07,
|
| 351 |
+
"loss": 0.2461,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.2599090318388564,
|
| 356 |
+
"grad_norm": 1.6957286596298218,
|
| 357 |
+
"learning_rate": 9.744735364682344e-07,
|
| 358 |
+
"loss": 0.2065,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.2651072124756335,
|
| 363 |
+
"grad_norm": 1.7848544120788574,
|
| 364 |
+
"learning_rate": 9.731246845359184e-07,
|
| 365 |
+
"loss": 0.1949,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.27030539311241064,
|
| 370 |
+
"grad_norm": 1.8262349367141724,
|
| 371 |
+
"learning_rate": 9.7174208935499e-07,
|
| 372 |
+
"loss": 0.2135,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.27550357374918777,
|
| 377 |
+
"grad_norm": 1.6100451946258545,
|
| 378 |
+
"learning_rate": 9.703258495321265e-07,
|
| 379 |
+
"loss": 0.1643,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.2807017543859649,
|
| 384 |
+
"grad_norm": 1.6476277112960815,
|
| 385 |
+
"learning_rate": 9.688760660735402e-07,
|
| 386 |
+
"loss": 0.1796,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.285899935022742,
|
| 391 |
+
"grad_norm": 1.6926974058151245,
|
| 392 |
+
"learning_rate": 9.673928423777756e-07,
|
| 393 |
+
"loss": 0.2048,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.29109811565951915,
|
| 398 |
+
"grad_norm": 1.7797563076019287,
|
| 399 |
+
"learning_rate": 9.658762842283341e-07,
|
| 400 |
+
"loss": 0.1953,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.2962962962962963,
|
| 405 |
+
"grad_norm": 1.7844983339309692,
|
| 406 |
+
"learning_rate": 9.643264997861312e-07,
|
| 407 |
+
"loss": 0.2103,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.3014944769330734,
|
| 412 |
+
"grad_norm": 1.8552502393722534,
|
| 413 |
+
"learning_rate": 9.627435995817797e-07,
|
| 414 |
+
"loss": 0.1854,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.30669265756985054,
|
| 419 |
+
"grad_norm": 1.8308689594268799,
|
| 420 |
+
"learning_rate": 9.611276965077097e-07,
|
| 421 |
+
"loss": 0.1892,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.31189083820662766,
|
| 426 |
+
"grad_norm": 1.8636093139648438,
|
| 427 |
+
"learning_rate": 9.594789058101153e-07,
|
| 428 |
+
"loss": 0.216,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.3170890188434048,
|
| 433 |
+
"grad_norm": 1.5857099294662476,
|
| 434 |
+
"learning_rate": 9.577973450807351e-07,
|
| 435 |
+
"loss": 0.1924,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.3222871994801819,
|
| 440 |
+
"grad_norm": 1.670000433921814,
|
| 441 |
+
"learning_rate": 9.560831342484666e-07,
|
| 442 |
+
"loss": 0.2088,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.32748538011695905,
|
| 447 |
+
"grad_norm": 1.8758388757705688,
|
| 448 |
+
"learning_rate": 9.543363955708124e-07,
|
| 449 |
+
"loss": 0.1697,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.3326835607537362,
|
| 454 |
+
"grad_norm": 2.020310401916504,
|
| 455 |
+
"learning_rate": 9.525572536251605e-07,
|
| 456 |
+
"loss": 0.2249,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.3378817413905133,
|
| 461 |
+
"grad_norm": 1.8294882774353027,
|
| 462 |
+
"learning_rate": 9.507458352999001e-07,
|
| 463 |
+
"loss": 0.1884,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.34307992202729043,
|
| 468 |
+
"grad_norm": 1.606002926826477,
|
| 469 |
+
"learning_rate": 9.489022697853708e-07,
|
| 470 |
+
"loss": 0.1761,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.34827810266406756,
|
| 475 |
+
"grad_norm": 1.6073530912399292,
|
| 476 |
+
"learning_rate": 9.470266885646503e-07,
|
| 477 |
+
"loss": 0.1871,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.3534762833008447,
|
| 482 |
+
"grad_norm": 1.7087726593017578,
|
| 483 |
+
"learning_rate": 9.451192254041758e-07,
|
| 484 |
+
"loss": 0.173,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.3586744639376218,
|
| 489 |
+
"grad_norm": 1.7764538526535034,
|
| 490 |
+
"learning_rate": 9.431800163442041e-07,
|
| 491 |
+
"loss": 0.1957,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.36387264457439894,
|
| 496 |
+
"grad_norm": 1.8759775161743164,
|
| 497 |
+
"learning_rate": 9.412091996891095e-07,
|
| 498 |
+
"loss": 0.2154,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.36907082521117607,
|
| 503 |
+
"grad_norm": 1.8281443119049072,
|
| 504 |
+
"learning_rate": 9.392069159975198e-07,
|
| 505 |
+
"loss": 0.1679,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.3742690058479532,
|
| 510 |
+
"grad_norm": 1.7894129753112793,
|
| 511 |
+
"learning_rate": 9.37173308072291e-07,
|
| 512 |
+
"loss": 0.1679,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.3794671864847303,
|
| 517 |
+
"grad_norm": 1.6492183208465576,
|
| 518 |
+
"learning_rate": 9.35108520950324e-07,
|
| 519 |
+
"loss": 0.1833,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.38466536712150745,
|
| 524 |
+
"grad_norm": 1.6076239347457886,
|
| 525 |
+
"learning_rate": 9.330127018922193e-07,
|
| 526 |
+
"loss": 0.1507,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.3898635477582846,
|
| 531 |
+
"grad_norm": 1.8182544708251953,
|
| 532 |
+
"learning_rate": 9.308860003717748e-07,
|
| 533 |
+
"loss": 0.1759,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.3950617283950617,
|
| 538 |
+
"grad_norm": 2.183497667312622,
|
| 539 |
+
"learning_rate": 9.287285680653254e-07,
|
| 540 |
+
"loss": 0.2069,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.40025990903183883,
|
| 545 |
+
"grad_norm": 1.9281930923461914,
|
| 546 |
+
"learning_rate": 9.265405588409256e-07,
|
| 547 |
+
"loss": 0.1813,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.40545808966861596,
|
| 552 |
+
"grad_norm": 1.7534650564193726,
|
| 553 |
+
"learning_rate": 9.243221287473755e-07,
|
| 554 |
+
"loss": 0.1764,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.4106562703053931,
|
| 559 |
+
"grad_norm": 1.7174078226089478,
|
| 560 |
+
"learning_rate": 9.220734360030906e-07,
|
| 561 |
+
"loss": 0.1863,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.4158544509421702,
|
| 566 |
+
"grad_norm": 1.7550305128097534,
|
| 567 |
+
"learning_rate": 9.197946409848194e-07,
|
| 568 |
+
"loss": 0.1718,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.42105263157894735,
|
| 573 |
+
"grad_norm": 1.4776816368103027,
|
| 574 |
+
"learning_rate": 9.174859062162037e-07,
|
| 575 |
+
"loss": 0.156,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.4262508122157245,
|
| 580 |
+
"grad_norm": 1.7932229042053223,
|
| 581 |
+
"learning_rate": 9.151473963561882e-07,
|
| 582 |
+
"loss": 0.1821,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.4314489928525016,
|
| 587 |
+
"grad_norm": 1.6103583574295044,
|
| 588 |
+
"learning_rate": 9.127792781872768e-07,
|
| 589 |
+
"loss": 0.1749,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.43664717348927873,
|
| 594 |
+
"grad_norm": 1.8216729164123535,
|
| 595 |
+
"learning_rate": 9.103817206036382e-07,
|
| 596 |
+
"loss": 0.177,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.44184535412605586,
|
| 601 |
+
"grad_norm": 1.7169886827468872,
|
| 602 |
+
"learning_rate": 9.079548945990592e-07,
|
| 603 |
+
"loss": 0.1845,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.447043534762833,
|
| 608 |
+
"grad_norm": 1.4935150146484375,
|
| 609 |
+
"learning_rate": 9.054989732547506e-07,
|
| 610 |
+
"loss": 0.1518,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.4522417153996101,
|
| 615 |
+
"grad_norm": 1.7215607166290283,
|
| 616 |
+
"learning_rate": 9.030141317270025e-07,
|
| 617 |
+
"loss": 0.1651,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.45743989603638724,
|
| 622 |
+
"grad_norm": 1.885299801826477,
|
| 623 |
+
"learning_rate": 9.005005472346923e-07,
|
| 624 |
+
"loss": 0.1862,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.46263807667316437,
|
| 629 |
+
"grad_norm": 1.6924781799316406,
|
| 630 |
+
"learning_rate": 8.979583990466452e-07,
|
| 631 |
+
"loss": 0.1834,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.4678362573099415,
|
| 636 |
+
"grad_norm": 1.6620601415634155,
|
| 637 |
+
"learning_rate": 8.953878684688492e-07,
|
| 638 |
+
"loss": 0.1736,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.4730344379467186,
|
| 643 |
+
"grad_norm": 1.7256325483322144,
|
| 644 |
+
"learning_rate": 8.92789138831524e-07,
|
| 645 |
+
"loss": 0.1792,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.47823261858349575,
|
| 650 |
+
"grad_norm": 1.6039340496063232,
|
| 651 |
+
"learning_rate": 8.901623954760459e-07,
|
| 652 |
+
"loss": 0.1704,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.4834307992202729,
|
| 657 |
+
"grad_norm": 1.6422524452209473,
|
| 658 |
+
"learning_rate": 8.875078257417294e-07,
|
| 659 |
+
"loss": 0.1621,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.48862897985705,
|
| 664 |
+
"grad_norm": 1.6837060451507568,
|
| 665 |
+
"learning_rate": 8.84825618952466e-07,
|
| 666 |
+
"loss": 0.183,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.49382716049382713,
|
| 671 |
+
"grad_norm": 1.750653862953186,
|
| 672 |
+
"learning_rate": 8.821159664032223e-07,
|
| 673 |
+
"loss": 0.1689,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.49902534113060426,
|
| 678 |
+
"grad_norm": 1.6462229490280151,
|
| 679 |
+
"learning_rate": 8.793790613463954e-07,
|
| 680 |
+
"loss": 0.1394,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.5042235217673814,
|
| 685 |
+
"grad_norm": 1.7336857318878174,
|
| 686 |
+
"learning_rate": 8.766150989780317e-07,
|
| 687 |
+
"loss": 0.1581,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.5094217024041585,
|
| 692 |
+
"grad_norm": 1.8384933471679688,
|
| 693 |
+
"learning_rate": 8.738242764239046e-07,
|
| 694 |
+
"loss": 0.1918,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.5146198830409356,
|
| 699 |
+
"grad_norm": 1.723486065864563,
|
| 700 |
+
"learning_rate": 8.710067927254554e-07,
|
| 701 |
+
"loss": 0.1737,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.5198180636777128,
|
| 706 |
+
"grad_norm": 1.9092669486999512,
|
| 707 |
+
"learning_rate": 8.681628488255986e-07,
|
| 708 |
+
"loss": 0.1728,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.5250162443144899,
|
| 713 |
+
"grad_norm": 1.729762315750122,
|
| 714 |
+
"learning_rate": 8.652926475543898e-07,
|
| 715 |
+
"loss": 0.162,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.530214424951267,
|
| 720 |
+
"grad_norm": 1.7867392301559448,
|
| 721 |
+
"learning_rate": 8.623963936145599e-07,
|
| 722 |
+
"loss": 0.1658,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.5354126055880442,
|
| 727 |
+
"grad_norm": 2.0217678546905518,
|
| 728 |
+
"learning_rate": 8.594742935669164e-07,
|
| 729 |
+
"loss": 0.1865,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.5406107862248213,
|
| 734 |
+
"grad_norm": 1.7473349571228027,
|
| 735 |
+
"learning_rate": 8.565265558156101e-07,
|
| 736 |
+
"loss": 0.1535,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.5458089668615984,
|
| 741 |
+
"grad_norm": 1.5292036533355713,
|
| 742 |
+
"learning_rate": 8.535533905932737e-07,
|
| 743 |
+
"loss": 0.1559,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.5510071474983755,
|
| 748 |
+
"grad_norm": 1.5472049713134766,
|
| 749 |
+
"learning_rate": 8.505550099460263e-07,
|
| 750 |
+
"loss": 0.1423,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.5562053281351527,
|
| 755 |
+
"grad_norm": 1.636443853378296,
|
| 756 |
+
"learning_rate": 8.475316277183508e-07,
|
| 757 |
+
"loss": 0.1747,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.5614035087719298,
|
| 762 |
+
"grad_norm": 1.5992189645767212,
|
| 763 |
+
"learning_rate": 8.444834595378433e-07,
|
| 764 |
+
"loss": 0.1766,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.5666016894087069,
|
| 769 |
+
"grad_norm": 1.6766347885131836,
|
| 770 |
+
"learning_rate": 8.414107227998328e-07,
|
| 771 |
+
"loss": 0.1421,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.571799870045484,
|
| 776 |
+
"grad_norm": 1.7345399856567383,
|
| 777 |
+
"learning_rate": 8.383136366518787e-07,
|
| 778 |
+
"loss": 0.1752,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.5769980506822612,
|
| 783 |
+
"grad_norm": 1.669264793395996,
|
| 784 |
+
"learning_rate": 8.351924219781392e-07,
|
| 785 |
+
"loss": 0.1661,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.5821962313190383,
|
| 790 |
+
"grad_norm": 1.7636350393295288,
|
| 791 |
+
"learning_rate": 8.320473013836195e-07,
|
| 792 |
+
"loss": 0.1892,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.5873944119558154,
|
| 797 |
+
"grad_norm": 1.8429635763168335,
|
| 798 |
+
"learning_rate": 8.288784991782945e-07,
|
| 799 |
+
"loss": 0.1883,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.5925925925925926,
|
| 804 |
+
"grad_norm": 1.5329152345657349,
|
| 805 |
+
"learning_rate": 8.256862413611112e-07,
|
| 806 |
+
"loss": 0.1472,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.5977907732293697,
|
| 811 |
+
"grad_norm": 1.9208284616470337,
|
| 812 |
+
"learning_rate": 8.22470755603871e-07,
|
| 813 |
+
"loss": 0.1714,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.6029889538661468,
|
| 818 |
+
"grad_norm": 1.6381752490997314,
|
| 819 |
+
"learning_rate": 8.192322712349917e-07,
|
| 820 |
+
"loss": 0.1806,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.6081871345029239,
|
| 825 |
+
"grad_norm": 1.5502922534942627,
|
| 826 |
+
"learning_rate": 8.159710192231519e-07,
|
| 827 |
+
"loss": 0.1653,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.6133853151397011,
|
| 832 |
+
"grad_norm": 1.604650616645813,
|
| 833 |
+
"learning_rate": 8.126872321608183e-07,
|
| 834 |
+
"loss": 0.1478,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.6185834957764782,
|
| 839 |
+
"grad_norm": 1.6860443353652954,
|
| 840 |
+
"learning_rate": 8.093811442476572e-07,
|
| 841 |
+
"loss": 0.1639,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.6237816764132553,
|
| 846 |
+
"grad_norm": 1.5915076732635498,
|
| 847 |
+
"learning_rate": 8.060529912738314e-07,
|
| 848 |
+
"loss": 0.1511,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.6289798570500325,
|
| 853 |
+
"grad_norm": 1.7241225242614746,
|
| 854 |
+
"learning_rate": 8.027030106031835e-07,
|
| 855 |
+
"loss": 0.1848,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.6341780376868096,
|
| 860 |
+
"grad_norm": 1.7747095823287964,
|
| 861 |
+
"learning_rate": 7.993314411563075e-07,
|
| 862 |
+
"loss": 0.1816,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.6393762183235867,
|
| 867 |
+
"grad_norm": 1.6497771739959717,
|
| 868 |
+
"learning_rate": 7.959385233935085e-07,
|
| 869 |
+
"loss": 0.1696,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.6445743989603638,
|
| 874 |
+
"grad_norm": 1.4712307453155518,
|
| 875 |
+
"learning_rate": 7.925244992976537e-07,
|
| 876 |
+
"loss": 0.1297,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.649772579597141,
|
| 881 |
+
"grad_norm": 1.618713140487671,
|
| 882 |
+
"learning_rate": 7.890896123569135e-07,
|
| 883 |
+
"loss": 0.1708,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.6549707602339181,
|
| 888 |
+
"grad_norm": 1.8550593852996826,
|
| 889 |
+
"learning_rate": 7.856341075473961e-07,
|
| 890 |
+
"loss": 0.1646,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.6601689408706952,
|
| 895 |
+
"grad_norm": 1.7929205894470215,
|
| 896 |
+
"learning_rate": 7.821582313156763e-07,
|
| 897 |
+
"loss": 0.1555,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.6653671215074723,
|
| 902 |
+
"grad_norm": 1.8011633157730103,
|
| 903 |
+
"learning_rate": 7.786622315612181e-07,
|
| 904 |
+
"loss": 0.1882,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.6705653021442495,
|
| 909 |
+
"grad_norm": 1.642986536026001,
|
| 910 |
+
"learning_rate": 7.751463576186957e-07,
|
| 911 |
+
"loss": 0.1659,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.6757634827810266,
|
| 916 |
+
"grad_norm": 1.547602653503418,
|
| 917 |
+
"learning_rate": 7.716108602402094e-07,
|
| 918 |
+
"loss": 0.1479,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.6809616634178037,
|
| 923 |
+
"grad_norm": 1.6602659225463867,
|
| 924 |
+
"learning_rate": 7.680559915774033e-07,
|
| 925 |
+
"loss": 0.1627,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.6861598440545809,
|
| 930 |
+
"grad_norm": 1.8091386556625366,
|
| 931 |
+
"learning_rate": 7.644820051634812e-07,
|
| 932 |
+
"loss": 0.1637,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.691358024691358,
|
| 937 |
+
"grad_norm": 1.669487476348877,
|
| 938 |
+
"learning_rate": 7.608891558951248e-07,
|
| 939 |
+
"loss": 0.1599,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.6965562053281351,
|
| 944 |
+
"grad_norm": 1.9016000032424927,
|
| 945 |
+
"learning_rate": 7.572777000143145e-07,
|
| 946 |
+
"loss": 0.1654,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.7017543859649122,
|
| 951 |
+
"grad_norm": 1.4672502279281616,
|
| 952 |
+
"learning_rate": 7.536478950900536e-07,
|
| 953 |
+
"loss": 0.1482,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.7069525666016894,
|
| 958 |
+
"grad_norm": 1.4602234363555908,
|
| 959 |
+
"learning_rate": 7.5e-07,
|
| 960 |
+
"loss": 0.1214,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.7121507472384665,
|
| 965 |
+
"grad_norm": 1.725661277770996,
|
| 966 |
+
"learning_rate": 7.463342749120013e-07,
|
| 967 |
+
"loss": 0.1406,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.7173489278752436,
|
| 972 |
+
"grad_norm": 1.6164398193359375,
|
| 973 |
+
"learning_rate": 7.426509812655405e-07,
|
| 974 |
+
"loss": 0.1492,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.7225471085120208,
|
| 979 |
+
"grad_norm": 1.609312891960144,
|
| 980 |
+
"learning_rate": 7.389503817530905e-07,
|
| 981 |
+
"loss": 0.1669,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.7277452891487979,
|
| 986 |
+
"grad_norm": 1.512629508972168,
|
| 987 |
+
"learning_rate": 7.352327403013779e-07,
|
| 988 |
+
"loss": 0.1318,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.732943469785575,
|
| 993 |
+
"grad_norm": 1.7129087448120117,
|
| 994 |
+
"learning_rate": 7.314983220525604e-07,
|
| 995 |
+
"loss": 0.1762,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.7381416504223521,
|
| 1000 |
+
"grad_norm": 1.6480506658554077,
|
| 1001 |
+
"learning_rate": 7.277473933453169e-07,
|
| 1002 |
+
"loss": 0.1738,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.7433398310591293,
|
| 1007 |
+
"grad_norm": 1.5904552936553955,
|
| 1008 |
+
"learning_rate": 7.239802216958522e-07,
|
| 1009 |
+
"loss": 0.1558,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.7485380116959064,
|
| 1014 |
+
"grad_norm": 1.6988767385482788,
|
| 1015 |
+
"learning_rate": 7.201970757788171e-07,
|
| 1016 |
+
"loss": 0.1661,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.7537361923326835,
|
| 1021 |
+
"grad_norm": 1.5458639860153198,
|
| 1022 |
+
"learning_rate": 7.163982254081474e-07,
|
| 1023 |
+
"loss": 0.1338,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.7589343729694606,
|
| 1028 |
+
"grad_norm": 1.5240118503570557,
|
| 1029 |
+
"learning_rate": 7.125839415178203e-07,
|
| 1030 |
+
"loss": 0.1405,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.7641325536062378,
|
| 1035 |
+
"grad_norm": 1.7464412450790405,
|
| 1036 |
+
"learning_rate": 7.087544961425316e-07,
|
| 1037 |
+
"loss": 0.1682,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.7693307342430149,
|
| 1042 |
+
"grad_norm": 1.7425211668014526,
|
| 1043 |
+
"learning_rate": 7.049101623982937e-07,
|
| 1044 |
+
"loss": 0.1839,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.774528914879792,
|
| 1049 |
+
"grad_norm": 1.4918522834777832,
|
| 1050 |
+
"learning_rate": 7.010512144629579e-07,
|
| 1051 |
+
"loss": 0.1124,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.7797270955165692,
|
| 1056 |
+
"grad_norm": 1.6756539344787598,
|
| 1057 |
+
"learning_rate": 6.971779275566593e-07,
|
| 1058 |
+
"loss": 0.1546,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.7849252761533463,
|
| 1063 |
+
"grad_norm": 1.5222876071929932,
|
| 1064 |
+
"learning_rate": 6.93290577922188e-07,
|
| 1065 |
+
"loss": 0.1313,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.7901234567901234,
|
| 1070 |
+
"grad_norm": 1.548453688621521,
|
| 1071 |
+
"learning_rate": 6.89389442805288e-07,
|
| 1072 |
+
"loss": 0.1349,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.7953216374269005,
|
| 1077 |
+
"grad_norm": 1.6898419857025146,
|
| 1078 |
+
"learning_rate": 6.85474800434884e-07,
|
| 1079 |
+
"loss": 0.1568,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.8005198180636777,
|
| 1084 |
+
"grad_norm": 1.8794304132461548,
|
| 1085 |
+
"learning_rate": 6.815469300032373e-07,
|
| 1086 |
+
"loss": 0.161,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.8057179987004548,
|
| 1091 |
+
"grad_norm": 1.6816418170928955,
|
| 1092 |
+
"learning_rate": 6.776061116460352e-07,
|
| 1093 |
+
"loss": 0.1615,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.8109161793372319,
|
| 1098 |
+
"grad_norm": 1.960444688796997,
|
| 1099 |
+
"learning_rate": 6.7365262642241e-07,
|
| 1100 |
+
"loss": 0.1948,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.816114359974009,
|
| 1105 |
+
"grad_norm": 1.6450730562210083,
|
| 1106 |
+
"learning_rate": 6.696867562948962e-07,
|
| 1107 |
+
"loss": 0.161,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.8213125406107862,
|
| 1112 |
+
"grad_norm": 1.4993230104446411,
|
| 1113 |
+
"learning_rate": 6.657087841093179e-07,
|
| 1114 |
+
"loss": 0.1476,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.8265107212475633,
|
| 1119 |
+
"grad_norm": 1.856066346168518,
|
| 1120 |
+
"learning_rate": 6.61718993574619e-07,
|
| 1121 |
+
"loss": 0.1599,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.8317089018843404,
|
| 1126 |
+
"grad_norm": 1.6243445873260498,
|
| 1127 |
+
"learning_rate": 6.577176692426278e-07,
|
| 1128 |
+
"loss": 0.1548,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.8369070825211176,
|
| 1133 |
+
"grad_norm": 1.538219928741455,
|
| 1134 |
+
"learning_rate": 6.537050964877625e-07,
|
| 1135 |
+
"loss": 0.1428,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.8421052631578947,
|
| 1140 |
+
"grad_norm": 1.429417371749878,
|
| 1141 |
+
"learning_rate": 6.496815614866791e-07,
|
| 1142 |
+
"loss": 0.1205,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.8473034437946718,
|
| 1147 |
+
"grad_norm": 1.7732073068618774,
|
| 1148 |
+
"learning_rate": 6.456473511978606e-07,
|
| 1149 |
+
"loss": 0.1903,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.852501624431449,
|
| 1154 |
+
"grad_norm": 1.575061321258545,
|
| 1155 |
+
"learning_rate": 6.416027533411519e-07,
|
| 1156 |
+
"loss": 0.1571,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.8576998050682261,
|
| 1161 |
+
"grad_norm": 1.6352499723434448,
|
| 1162 |
+
"learning_rate": 6.375480563772389e-07,
|
| 1163 |
+
"loss": 0.1484,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.8628979857050032,
|
| 1168 |
+
"grad_norm": 1.7170888185501099,
|
| 1169 |
+
"learning_rate": 6.334835494870758e-07,
|
| 1170 |
+
"loss": 0.1735,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.8680961663417803,
|
| 1175 |
+
"grad_norm": 1.5450496673583984,
|
| 1176 |
+
"learning_rate": 6.294095225512604e-07,
|
| 1177 |
+
"loss": 0.1339,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.8732943469785575,
|
| 1182 |
+
"grad_norm": 1.5989458560943604,
|
| 1183 |
+
"learning_rate": 6.253262661293602e-07,
|
| 1184 |
+
"loss": 0.1393,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.8784925276153346,
|
| 1189 |
+
"grad_norm": 1.464534878730774,
|
| 1190 |
+
"learning_rate": 6.2123407143919e-07,
|
| 1191 |
+
"loss": 0.1421,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.8836907082521117,
|
| 1196 |
+
"grad_norm": 1.6165345907211304,
|
| 1197 |
+
"learning_rate": 6.17133230336041e-07,
|
| 1198 |
+
"loss": 0.154,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.8888888888888888,
|
| 1203 |
+
"grad_norm": 1.5105384588241577,
|
| 1204 |
+
"learning_rate": 6.130240352918674e-07,
|
| 1205 |
+
"loss": 0.1614,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.894087069525666,
|
| 1210 |
+
"grad_norm": 1.6538264751434326,
|
| 1211 |
+
"learning_rate": 6.089067793744257e-07,
|
| 1212 |
+
"loss": 0.1213,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.8992852501624431,
|
| 1217 |
+
"grad_norm": 1.5659717321395874,
|
| 1218 |
+
"learning_rate": 6.047817562263743e-07,
|
| 1219 |
+
"loss": 0.1349,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.9044834307992202,
|
| 1224 |
+
"grad_norm": 1.6108099222183228,
|
| 1225 |
+
"learning_rate": 6.0064926004433e-07,
|
| 1226 |
+
"loss": 0.1572,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.9096816114359974,
|
| 1231 |
+
"grad_norm": 1.7230148315429688,
|
| 1232 |
+
"learning_rate": 5.965095855578868e-07,
|
| 1233 |
+
"loss": 0.1376,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.9148797920727745,
|
| 1238 |
+
"grad_norm": 1.7344483137130737,
|
| 1239 |
+
"learning_rate": 5.923630280085947e-07,
|
| 1240 |
+
"loss": 0.1572,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.9200779727095516,
|
| 1245 |
+
"grad_norm": 1.6481879949569702,
|
| 1246 |
+
"learning_rate": 5.882098831289043e-07,
|
| 1247 |
+
"loss": 0.1626,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.9252761533463287,
|
| 1252 |
+
"grad_norm": 1.7318065166473389,
|
| 1253 |
+
"learning_rate": 5.840504471210741e-07,
|
| 1254 |
+
"loss": 0.1756,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.9304743339831059,
|
| 1259 |
+
"grad_norm": 1.676165223121643,
|
| 1260 |
+
"learning_rate": 5.79885016636046e-07,
|
| 1261 |
+
"loss": 0.147,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.935672514619883,
|
| 1266 |
+
"grad_norm": 1.4620646238327026,
|
| 1267 |
+
"learning_rate": 5.757138887522883e-07,
|
| 1268 |
+
"loss": 0.1249,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.9408706952566601,
|
| 1273 |
+
"grad_norm": 1.65927255153656,
|
| 1274 |
+
"learning_rate": 5.71537360954607e-07,
|
| 1275 |
+
"loss": 0.163,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.9460688758934372,
|
| 1280 |
+
"grad_norm": 1.5536587238311768,
|
| 1281 |
+
"learning_rate": 5.673557311129306e-07,
|
| 1282 |
+
"loss": 0.1351,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.9512670565302144,
|
| 1287 |
+
"grad_norm": 1.7076836824417114,
|
| 1288 |
+
"learning_rate": 5.631692974610647e-07,
|
| 1289 |
+
"loss": 0.1771,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.9564652371669915,
|
| 1294 |
+
"grad_norm": 1.4979828596115112,
|
| 1295 |
+
"learning_rate": 5.589783585754231e-07,
|
| 1296 |
+
"loss": 0.121,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.9616634178037686,
|
| 1301 |
+
"grad_norm": 1.5839756727218628,
|
| 1302 |
+
"learning_rate": 5.547832133537327e-07,
|
| 1303 |
+
"loss": 0.1458,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.9668615984405458,
|
| 1308 |
+
"grad_norm": 1.7546137571334839,
|
| 1309 |
+
"learning_rate": 5.505841609937161e-07,
|
| 1310 |
+
"loss": 0.1671,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.9720597790773229,
|
| 1315 |
+
"grad_norm": 1.7105190753936768,
|
| 1316 |
+
"learning_rate": 5.463815009717532e-07,
|
| 1317 |
+
"loss": 0.1314,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.9772579597141,
|
| 1322 |
+
"grad_norm": 1.8557852506637573,
|
| 1323 |
+
"learning_rate": 5.421755330215223e-07,
|
| 1324 |
+
"loss": 0.1794,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.9824561403508771,
|
| 1329 |
+
"grad_norm": 1.569214105606079,
|
| 1330 |
+
"learning_rate": 5.379665571126231e-07,
|
| 1331 |
+
"loss": 0.1307,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.9876543209876543,
|
| 1336 |
+
"grad_norm": 1.6137492656707764,
|
| 1337 |
+
"learning_rate": 5.337548734291826e-07,
|
| 1338 |
+
"loss": 0.1412,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.9928525016244314,
|
| 1343 |
+
"grad_norm": 1.6707996129989624,
|
| 1344 |
+
"learning_rate": 5.295407823484467e-07,
|
| 1345 |
+
"loss": 0.1627,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.9980506822612085,
|
| 1350 |
+
"grad_norm": 2.3229496479034424,
|
| 1351 |
+
"learning_rate": 5.253245844193564e-07,
|
| 1352 |
+
"loss": 0.1872,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.0032488628979856,
|
| 1357 |
+
"grad_norm": 1.4374881982803345,
|
| 1358 |
+
"learning_rate": 5.211065803411134e-07,
|
| 1359 |
+
"loss": 0.1118,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.0084470435347628,
|
| 1364 |
+
"grad_norm": 1.7097264528274536,
|
| 1365 |
+
"learning_rate": 5.168870709417341e-07,
|
| 1366 |
+
"loss": 0.1603,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.01364522417154,
|
| 1371 |
+
"grad_norm": 1.5904935598373413,
|
| 1372 |
+
"learning_rate": 5.126663571565939e-07,
|
| 1373 |
+
"loss": 0.128,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.018843404808317,
|
| 1378 |
+
"grad_norm": 1.5433835983276367,
|
| 1379 |
+
"learning_rate": 5.084447400069654e-07,
|
| 1380 |
+
"loss": 0.1192,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.0240415854450942,
|
| 1385 |
+
"grad_norm": 1.40073561668396,
|
| 1386 |
+
"learning_rate": 5.042225205785492e-07,
|
| 1387 |
+
"loss": 0.1188,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.0292397660818713,
|
| 1392 |
+
"grad_norm": 1.6374619007110596,
|
| 1393 |
+
"learning_rate": 5e-07,
|
| 1394 |
+
"loss": 0.1486,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.0344379467186484,
|
| 1399 |
+
"grad_norm": 1.4800790548324585,
|
| 1400 |
+
"learning_rate": 4.957774794214508e-07,
|
| 1401 |
+
"loss": 0.1297,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.0396361273554255,
|
| 1406 |
+
"grad_norm": 1.5941686630249023,
|
| 1407 |
+
"learning_rate": 4.915552599930345e-07,
|
| 1408 |
+
"loss": 0.1466,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.0448343079922027,
|
| 1413 |
+
"grad_norm": 1.5258111953735352,
|
| 1414 |
+
"learning_rate": 4.873336428434061e-07,
|
| 1415 |
+
"loss": 0.1264,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.0500324886289798,
|
| 1420 |
+
"grad_norm": 1.725998878479004,
|
| 1421 |
+
"learning_rate": 4.831129290582659e-07,
|
| 1422 |
+
"loss": 0.1334,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.055230669265757,
|
| 1427 |
+
"grad_norm": 1.8626656532287598,
|
| 1428 |
+
"learning_rate": 4.788934196588865e-07,
|
| 1429 |
+
"loss": 0.1503,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.060428849902534,
|
| 1434 |
+
"grad_norm": 1.4593480825424194,
|
| 1435 |
+
"learning_rate": 4.746754155806437e-07,
|
| 1436 |
+
"loss": 0.1225,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.0656270305393112,
|
| 1441 |
+
"grad_norm": 1.6284171342849731,
|
| 1442 |
+
"learning_rate": 4.7045921765155337e-07,
|
| 1443 |
+
"loss": 0.1397,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.0708252111760883,
|
| 1448 |
+
"grad_norm": 1.4469428062438965,
|
| 1449 |
+
"learning_rate": 4.662451265708174e-07,
|
| 1450 |
+
"loss": 0.1082,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.0760233918128654,
|
| 1455 |
+
"grad_norm": 1.3986436128616333,
|
| 1456 |
+
"learning_rate": 4.620334428873769e-07,
|
| 1457 |
+
"loss": 0.1025,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.0812215724496426,
|
| 1462 |
+
"grad_norm": 1.6322413682937622,
|
| 1463 |
+
"learning_rate": 4.5782446697847764e-07,
|
| 1464 |
+
"loss": 0.126,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.0864197530864197,
|
| 1469 |
+
"grad_norm": 1.4022878408432007,
|
| 1470 |
+
"learning_rate": 4.536184990282467e-07,
|
| 1471 |
+
"loss": 0.0932,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.0916179337231968,
|
| 1476 |
+
"grad_norm": 1.8021215200424194,
|
| 1477 |
+
"learning_rate": 4.4941583900628393e-07,
|
| 1478 |
+
"loss": 0.1662,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.096816114359974,
|
| 1483 |
+
"grad_norm": 1.806060552597046,
|
| 1484 |
+
"learning_rate": 4.4521678664626745e-07,
|
| 1485 |
+
"loss": 0.1574,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.102014294996751,
|
| 1490 |
+
"grad_norm": 1.7470866441726685,
|
| 1491 |
+
"learning_rate": 4.4102164142457705e-07,
|
| 1492 |
+
"loss": 0.1467,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.1072124756335282,
|
| 1497 |
+
"grad_norm": 1.5723119974136353,
|
| 1498 |
+
"learning_rate": 4.368307025389355e-07,
|
| 1499 |
+
"loss": 0.1084,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.1124106562703053,
|
| 1504 |
+
"grad_norm": 1.5377353429794312,
|
| 1505 |
+
"learning_rate": 4.326442688870696e-07,
|
| 1506 |
+
"loss": 0.116,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.1176088369070825,
|
| 1511 |
+
"grad_norm": 1.7077683210372925,
|
| 1512 |
+
"learning_rate": 4.2846263904539303e-07,
|
| 1513 |
+
"loss": 0.1483,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.1228070175438596,
|
| 1518 |
+
"grad_norm": 1.8333083391189575,
|
| 1519 |
+
"learning_rate": 4.242861112477118e-07,
|
| 1520 |
+
"loss": 0.1527,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.1280051981806367,
|
| 1525 |
+
"grad_norm": 1.5769239664077759,
|
| 1526 |
+
"learning_rate": 4.201149833639539e-07,
|
| 1527 |
+
"loss": 0.121,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.1332033788174138,
|
| 1532 |
+
"grad_norm": 1.919921636581421,
|
| 1533 |
+
"learning_rate": 4.15949552878926e-07,
|
| 1534 |
+
"loss": 0.173,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.138401559454191,
|
| 1539 |
+
"grad_norm": 1.5571967363357544,
|
| 1540 |
+
"learning_rate": 4.117901168710959e-07,
|
| 1541 |
+
"loss": 0.1227,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.143599740090968,
|
| 1546 |
+
"grad_norm": 1.6683669090270996,
|
| 1547 |
+
"learning_rate": 4.0763697199140546e-07,
|
| 1548 |
+
"loss": 0.1422,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.1487979207277452,
|
| 1553 |
+
"grad_norm": 1.6401057243347168,
|
| 1554 |
+
"learning_rate": 4.034904144421134e-07,
|
| 1555 |
+
"loss": 0.1256,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.1539961013645224,
|
| 1560 |
+
"grad_norm": 1.5419056415557861,
|
| 1561 |
+
"learning_rate": 3.9935073995566987e-07,
|
| 1562 |
+
"loss": 0.1302,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.1591942820012995,
|
| 1567 |
+
"grad_norm": 1.650795817375183,
|
| 1568 |
+
"learning_rate": 3.952182437736256e-07,
|
| 1569 |
+
"loss": 0.1471,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.1643924626380766,
|
| 1574 |
+
"grad_norm": 1.672743797302246,
|
| 1575 |
+
"learning_rate": 3.910932206255742e-07,
|
| 1576 |
+
"loss": 0.1298,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.1695906432748537,
|
| 1581 |
+
"grad_norm": 1.6902425289154053,
|
| 1582 |
+
"learning_rate": 3.869759647081325e-07,
|
| 1583 |
+
"loss": 0.1414,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.1747888239116309,
|
| 1588 |
+
"grad_norm": 1.607485055923462,
|
| 1589 |
+
"learning_rate": 3.828667696639589e-07,
|
| 1590 |
+
"loss": 0.1032,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.179987004548408,
|
| 1595 |
+
"grad_norm": 1.5336533784866333,
|
| 1596 |
+
"learning_rate": 3.7876592856081e-07,
|
| 1597 |
+
"loss": 0.1116,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.1851851851851851,
|
| 1602 |
+
"grad_norm": 1.5528396368026733,
|
| 1603 |
+
"learning_rate": 3.7467373387063964e-07,
|
| 1604 |
+
"loss": 0.1243,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.1903833658219622,
|
| 1609 |
+
"grad_norm": 1.743318796157837,
|
| 1610 |
+
"learning_rate": 3.7059047744873955e-07,
|
| 1611 |
+
"loss": 0.1437,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.1955815464587394,
|
| 1616 |
+
"grad_norm": 1.6484525203704834,
|
| 1617 |
+
"learning_rate": 3.665164505129241e-07,
|
| 1618 |
+
"loss": 0.131,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.2007797270955165,
|
| 1623 |
+
"grad_norm": 1.6531956195831299,
|
| 1624 |
+
"learning_rate": 3.6245194362276094e-07,
|
| 1625 |
+
"loss": 0.1268,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.2059779077322936,
|
| 1630 |
+
"grad_norm": 1.491297960281372,
|
| 1631 |
+
"learning_rate": 3.5839724665884795e-07,
|
| 1632 |
+
"loss": 0.1261,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.2111760883690708,
|
| 1637 |
+
"grad_norm": 1.5535728931427002,
|
| 1638 |
+
"learning_rate": 3.5435264880213937e-07,
|
| 1639 |
+
"loss": 0.1233,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.2163742690058479,
|
| 1644 |
+
"grad_norm": 1.6621695756912231,
|
| 1645 |
+
"learning_rate": 3.50318438513321e-07,
|
| 1646 |
+
"loss": 0.1331,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.221572449642625,
|
| 1651 |
+
"grad_norm": 1.5829371213912964,
|
| 1652 |
+
"learning_rate": 3.462949035122376e-07,
|
| 1653 |
+
"loss": 0.1229,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.2267706302794021,
|
| 1658 |
+
"grad_norm": 1.7693301439285278,
|
| 1659 |
+
"learning_rate": 3.4228233075737223e-07,
|
| 1660 |
+
"loss": 0.1434,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.2319688109161793,
|
| 1665 |
+
"grad_norm": 1.6113789081573486,
|
| 1666 |
+
"learning_rate": 3.3828100642538093e-07,
|
| 1667 |
+
"loss": 0.1213,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.2371669915529564,
|
| 1672 |
+
"grad_norm": 1.5799354314804077,
|
| 1673 |
+
"learning_rate": 3.342912158906821e-07,
|
| 1674 |
+
"loss": 0.1191,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.2423651721897335,
|
| 1679 |
+
"grad_norm": 1.5467253923416138,
|
| 1680 |
+
"learning_rate": 3.3031324370510396e-07,
|
| 1681 |
+
"loss": 0.1133,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.2475633528265107,
|
| 1686 |
+
"grad_norm": 1.8147982358932495,
|
| 1687 |
+
"learning_rate": 3.263473735775899e-07,
|
| 1688 |
+
"loss": 0.1391,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.2527615334632878,
|
| 1693 |
+
"grad_norm": 1.702359676361084,
|
| 1694 |
+
"learning_rate": 3.2239388835396484e-07,
|
| 1695 |
+
"loss": 0.1339,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.257959714100065,
|
| 1700 |
+
"grad_norm": 1.7504138946533203,
|
| 1701 |
+
"learning_rate": 3.184530699967627e-07,
|
| 1702 |
+
"loss": 0.1565,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.263157894736842,
|
| 1707 |
+
"grad_norm": 1.7226463556289673,
|
| 1708 |
+
"learning_rate": 3.1452519956511614e-07,
|
| 1709 |
+
"loss": 0.1266,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.2683560753736192,
|
| 1714 |
+
"grad_norm": 1.8186461925506592,
|
| 1715 |
+
"learning_rate": 3.1061055719471197e-07,
|
| 1716 |
+
"loss": 0.1347,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 1.2735542560103963,
|
| 1721 |
+
"grad_norm": 1.6384614706039429,
|
| 1722 |
+
"learning_rate": 3.0670942207781204e-07,
|
| 1723 |
+
"loss": 0.1115,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.2787524366471734,
|
| 1728 |
+
"grad_norm": 1.7369823455810547,
|
| 1729 |
+
"learning_rate": 3.028220724433408e-07,
|
| 1730 |
+
"loss": 0.129,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.2839506172839505,
|
| 1735 |
+
"grad_norm": 1.6780822277069092,
|
| 1736 |
+
"learning_rate": 2.989487855370421e-07,
|
| 1737 |
+
"loss": 0.1385,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 1.2891487979207277,
|
| 1742 |
+
"grad_norm": 1.87428879737854,
|
| 1743 |
+
"learning_rate": 2.9508983760170634e-07,
|
| 1744 |
+
"loss": 0.1435,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.2943469785575048,
|
| 1749 |
+
"grad_norm": 1.7940468788146973,
|
| 1750 |
+
"learning_rate": 2.9124550385746856e-07,
|
| 1751 |
+
"loss": 0.1491,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.299545159194282,
|
| 1756 |
+
"grad_norm": 1.712099552154541,
|
| 1757 |
+
"learning_rate": 2.8741605848217976e-07,
|
| 1758 |
+
"loss": 0.131,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 1.304743339831059,
|
| 1763 |
+
"grad_norm": 1.6154824495315552,
|
| 1764 |
+
"learning_rate": 2.8360177459185263e-07,
|
| 1765 |
+
"loss": 0.1145,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 1.3099415204678362,
|
| 1770 |
+
"grad_norm": 1.6371185779571533,
|
| 1771 |
+
"learning_rate": 2.7980292422118277e-07,
|
| 1772 |
+
"loss": 0.1232,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 1.3151397011046133,
|
| 1777 |
+
"grad_norm": 1.8156847953796387,
|
| 1778 |
+
"learning_rate": 2.7601977830414766e-07,
|
| 1779 |
+
"loss": 0.1274,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 1.3203378817413904,
|
| 1784 |
+
"grad_norm": 1.6596229076385498,
|
| 1785 |
+
"learning_rate": 2.72252606654683e-07,
|
| 1786 |
+
"loss": 0.1168,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.3255360623781676,
|
| 1791 |
+
"grad_norm": 1.6106423139572144,
|
| 1792 |
+
"learning_rate": 2.685016779474396e-07,
|
| 1793 |
+
"loss": 0.1139,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 1.3307342430149447,
|
| 1798 |
+
"grad_norm": 1.6363728046417236,
|
| 1799 |
+
"learning_rate": 2.6476725969862226e-07,
|
| 1800 |
+
"loss": 0.1297,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 1.3359324236517218,
|
| 1805 |
+
"grad_norm": 1.4978957176208496,
|
| 1806 |
+
"learning_rate": 2.6104961824690964e-07,
|
| 1807 |
+
"loss": 0.1191,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 1.341130604288499,
|
| 1812 |
+
"grad_norm": 1.5889379978179932,
|
| 1813 |
+
"learning_rate": 2.5734901873445956e-07,
|
| 1814 |
+
"loss": 0.1236,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.346328784925276,
|
| 1819 |
+
"grad_norm": 1.534178376197815,
|
| 1820 |
+
"learning_rate": 2.536657250879988e-07,
|
| 1821 |
+
"loss": 0.1053,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 1.3515269655620532,
|
| 1826 |
+
"grad_norm": 1.8409833908081055,
|
| 1827 |
+
"learning_rate": 2.500000000000001e-07,
|
| 1828 |
+
"loss": 0.1427,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 1.3567251461988303,
|
| 1833 |
+
"grad_norm": 1.7446589469909668,
|
| 1834 |
+
"learning_rate": 2.4635210490994647e-07,
|
| 1835 |
+
"loss": 0.1194,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 1.3619233268356075,
|
| 1840 |
+
"grad_norm": 1.7886688709259033,
|
| 1841 |
+
"learning_rate": 2.427222999856857e-07,
|
| 1842 |
+
"loss": 0.1351,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 1.3671215074723846,
|
| 1847 |
+
"grad_norm": 1.6462031602859497,
|
| 1848 |
+
"learning_rate": 2.391108441048753e-07,
|
| 1849 |
+
"loss": 0.1249,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 1.3723196881091617,
|
| 1854 |
+
"grad_norm": 1.8700019121170044,
|
| 1855 |
+
"learning_rate": 2.355179948365189e-07,
|
| 1856 |
+
"loss": 0.1482,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.3775178687459388,
|
| 1861 |
+
"grad_norm": 1.8244132995605469,
|
| 1862 |
+
"learning_rate": 2.3194400842259687e-07,
|
| 1863 |
+
"loss": 0.134,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 1.382716049382716,
|
| 1868 |
+
"grad_norm": 1.8189918994903564,
|
| 1869 |
+
"learning_rate": 2.283891397597908e-07,
|
| 1870 |
+
"loss": 0.1258,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 1.387914230019493,
|
| 1875 |
+
"grad_norm": 1.5552480220794678,
|
| 1876 |
+
"learning_rate": 2.2485364238130433e-07,
|
| 1877 |
+
"loss": 0.1131,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 1.3931124106562702,
|
| 1882 |
+
"grad_norm": 1.659328579902649,
|
| 1883 |
+
"learning_rate": 2.2133776843878183e-07,
|
| 1884 |
+
"loss": 0.1119,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 1.3983105912930474,
|
| 1889 |
+
"grad_norm": 1.3867595195770264,
|
| 1890 |
+
"learning_rate": 2.1784176868432375e-07,
|
| 1891 |
+
"loss": 0.0851,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 1.4035087719298245,
|
| 1896 |
+
"grad_norm": 1.7647099494934082,
|
| 1897 |
+
"learning_rate": 2.1436589245260372e-07,
|
| 1898 |
+
"loss": 0.1158,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.4087069525666016,
|
| 1903 |
+
"grad_norm": 1.8926544189453125,
|
| 1904 |
+
"learning_rate": 2.109103876430864e-07,
|
| 1905 |
+
"loss": 0.1527,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 1.4139051332033787,
|
| 1910 |
+
"grad_norm": 1.839390516281128,
|
| 1911 |
+
"learning_rate": 2.074755007023461e-07,
|
| 1912 |
+
"loss": 0.1108,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 1.4191033138401559,
|
| 1917 |
+
"grad_norm": 1.797500729560852,
|
| 1918 |
+
"learning_rate": 2.040614766064913e-07,
|
| 1919 |
+
"loss": 0.1508,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 1.424301494476933,
|
| 1924 |
+
"grad_norm": 1.7864497900009155,
|
| 1925 |
+
"learning_rate": 2.0066855884369243e-07,
|
| 1926 |
+
"loss": 0.1242,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 1.4294996751137101,
|
| 1931 |
+
"grad_norm": 1.853615641593933,
|
| 1932 |
+
"learning_rate": 1.9729698939681644e-07,
|
| 1933 |
+
"loss": 0.122,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 1.4346978557504872,
|
| 1938 |
+
"grad_norm": 1.6054631471633911,
|
| 1939 |
+
"learning_rate": 1.9394700872616853e-07,
|
| 1940 |
+
"loss": 0.1212,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 1.4398960363872644,
|
| 1945 |
+
"grad_norm": 1.632055640220642,
|
| 1946 |
+
"learning_rate": 1.906188557523427e-07,
|
| 1947 |
+
"loss": 0.1101,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 1.4450942170240415,
|
| 1952 |
+
"grad_norm": 1.6560664176940918,
|
| 1953 |
+
"learning_rate": 1.873127678391816e-07,
|
| 1954 |
+
"loss": 0.1217,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 1.4502923976608186,
|
| 1959 |
+
"grad_norm": 1.4159197807312012,
|
| 1960 |
+
"learning_rate": 1.8402898077684803e-07,
|
| 1961 |
+
"loss": 0.0988,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 1.4554905782975958,
|
| 1966 |
+
"grad_norm": 1.6314151287078857,
|
| 1967 |
+
"learning_rate": 1.8076772876500828e-07,
|
| 1968 |
+
"loss": 0.1293,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 1.4606887589343729,
|
| 1973 |
+
"grad_norm": 1.7430942058563232,
|
| 1974 |
+
"learning_rate": 1.775292443961291e-07,
|
| 1975 |
+
"loss": 0.1401,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 1.46588693957115,
|
| 1980 |
+
"grad_norm": 1.7857812643051147,
|
| 1981 |
+
"learning_rate": 1.7431375863888898e-07,
|
| 1982 |
+
"loss": 0.1275,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 1.4710851202079271,
|
| 1987 |
+
"grad_norm": 1.5308443307876587,
|
| 1988 |
+
"learning_rate": 1.7112150082170568e-07,
|
| 1989 |
+
"loss": 0.1061,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 1.4762833008447043,
|
| 1994 |
+
"grad_norm": 1.6288737058639526,
|
| 1995 |
+
"learning_rate": 1.679526986163804e-07,
|
| 1996 |
+
"loss": 0.1119,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 1.4814814814814814,
|
| 2001 |
+
"grad_norm": 1.704690933227539,
|
| 2002 |
+
"learning_rate": 1.6480757802186068e-07,
|
| 2003 |
+
"loss": 0.1166,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 1.4866796621182585,
|
| 2008 |
+
"grad_norm": 1.5033763647079468,
|
| 2009 |
+
"learning_rate": 1.6168636334812125e-07,
|
| 2010 |
+
"loss": 0.1045,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 1.4918778427550357,
|
| 2015 |
+
"grad_norm": 1.4401872158050537,
|
| 2016 |
+
"learning_rate": 1.5858927720016706e-07,
|
| 2017 |
+
"loss": 0.0959,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 1.4970760233918128,
|
| 2022 |
+
"grad_norm": 1.6706205606460571,
|
| 2023 |
+
"learning_rate": 1.555165404621567e-07,
|
| 2024 |
+
"loss": 0.1124,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 1.50227420402859,
|
| 2029 |
+
"grad_norm": 1.8483508825302124,
|
| 2030 |
+
"learning_rate": 1.5246837228164905e-07,
|
| 2031 |
+
"loss": 0.1146,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 1.507472384665367,
|
| 2036 |
+
"grad_norm": 1.8255398273468018,
|
| 2037 |
+
"learning_rate": 1.494449900539737e-07,
|
| 2038 |
+
"loss": 0.1413,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 1.5126705653021442,
|
| 2043 |
+
"grad_norm": 1.8132373094558716,
|
| 2044 |
+
"learning_rate": 1.4644660940672627e-07,
|
| 2045 |
+
"loss": 0.1316,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 1.5178687459389213,
|
| 2050 |
+
"grad_norm": 1.7580801248550415,
|
| 2051 |
+
"learning_rate": 1.434734441843899e-07,
|
| 2052 |
+
"loss": 0.1252,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 1.5230669265756984,
|
| 2057 |
+
"grad_norm": 1.4802451133728027,
|
| 2058 |
+
"learning_rate": 1.4052570643308375e-07,
|
| 2059 |
+
"loss": 0.1087,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 1.5282651072124755,
|
| 2064 |
+
"grad_norm": 1.595434308052063,
|
| 2065 |
+
"learning_rate": 1.376036063854401e-07,
|
| 2066 |
+
"loss": 0.1063,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 1.5334632878492527,
|
| 2071 |
+
"grad_norm": 1.5052251815795898,
|
| 2072 |
+
"learning_rate": 1.3470735244561027e-07,
|
| 2073 |
+
"loss": 0.1071,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 1.5386614684860298,
|
| 2078 |
+
"grad_norm": 1.4966932535171509,
|
| 2079 |
+
"learning_rate": 1.3183715117440142e-07,
|
| 2080 |
+
"loss": 0.1003,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 1.543859649122807,
|
| 2085 |
+
"grad_norm": 1.5577093362808228,
|
| 2086 |
+
"learning_rate": 1.2899320727454472e-07,
|
| 2087 |
+
"loss": 0.1147,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 1.549057829759584,
|
| 2092 |
+
"grad_norm": 2.081566572189331,
|
| 2093 |
+
"learning_rate": 1.2617572357609562e-07,
|
| 2094 |
+
"loss": 0.1479,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 1.5542560103963612,
|
| 2099 |
+
"grad_norm": 1.5348504781723022,
|
| 2100 |
+
"learning_rate": 1.2338490102196825e-07,
|
| 2101 |
+
"loss": 0.1061,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 1.5594541910331383,
|
| 2106 |
+
"grad_norm": 1.7641793489456177,
|
| 2107 |
+
"learning_rate": 1.2062093865360457e-07,
|
| 2108 |
+
"loss": 0.1359,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 1.5646523716699154,
|
| 2113 |
+
"grad_norm": 1.5747112035751343,
|
| 2114 |
+
"learning_rate": 1.1788403359677767e-07,
|
| 2115 |
+
"loss": 0.1069,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 1.5698505523066926,
|
| 2120 |
+
"grad_norm": 1.6333017349243164,
|
| 2121 |
+
"learning_rate": 1.1517438104753385e-07,
|
| 2122 |
+
"loss": 0.1077,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 1.5750487329434697,
|
| 2127 |
+
"grad_norm": 1.5666186809539795,
|
| 2128 |
+
"learning_rate": 1.1249217425827062e-07,
|
| 2129 |
+
"loss": 0.1118,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 1.5802469135802468,
|
| 2134 |
+
"grad_norm": 1.5051664113998413,
|
| 2135 |
+
"learning_rate": 1.0983760452395413e-07,
|
| 2136 |
+
"loss": 0.1043,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 1.585445094217024,
|
| 2141 |
+
"grad_norm": 1.7494423389434814,
|
| 2142 |
+
"learning_rate": 1.07210861168476e-07,
|
| 2143 |
+
"loss": 0.1327,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 1.590643274853801,
|
| 2148 |
+
"grad_norm": 1.5114164352416992,
|
| 2149 |
+
"learning_rate": 1.0461213153115079e-07,
|
| 2150 |
+
"loss": 0.0938,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 1.5958414554905782,
|
| 2155 |
+
"grad_norm": 1.562703251838684,
|
| 2156 |
+
"learning_rate": 1.0204160095335479e-07,
|
| 2157 |
+
"loss": 0.1056,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 1.6010396361273553,
|
| 2162 |
+
"grad_norm": 1.6739459037780762,
|
| 2163 |
+
"learning_rate": 9.94994527653078e-08,
|
| 2164 |
+
"loss": 0.1223,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 1.6062378167641325,
|
| 2169 |
+
"grad_norm": 1.6659190654754639,
|
| 2170 |
+
"learning_rate": 9.69858682729976e-08,
|
| 2171 |
+
"loss": 0.1201,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 1.6114359974009096,
|
| 2176 |
+
"grad_norm": 1.6198205947875977,
|
| 2177 |
+
"learning_rate": 9.45010267452495e-08,
|
| 2178 |
+
"loss": 0.1041,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 1.6166341780376867,
|
| 2183 |
+
"grad_norm": 1.6569786071777344,
|
| 2184 |
+
"learning_rate": 9.204510540094095e-08,
|
| 2185 |
+
"loss": 0.1153,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 1.6218323586744638,
|
| 2190 |
+
"grad_norm": 1.6825376749038696,
|
| 2191 |
+
"learning_rate": 8.961827939636196e-08,
|
| 2192 |
+
"loss": 0.1195,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.627030539311241,
|
| 2197 |
+
"grad_norm": 1.6692954301834106,
|
| 2198 |
+
"learning_rate": 8.722072181272311e-08,
|
| 2199 |
+
"loss": 0.14,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 1.632228719948018,
|
| 2204 |
+
"grad_norm": 1.86336350440979,
|
| 2205 |
+
"learning_rate": 8.485260364381186e-08,
|
| 2206 |
+
"loss": 0.1154,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 1.6374269005847952,
|
| 2211 |
+
"grad_norm": 1.6375104188919067,
|
| 2212 |
+
"learning_rate": 8.251409378379637e-08,
|
| 2213 |
+
"loss": 0.1087,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 1.6426250812215724,
|
| 2218 |
+
"grad_norm": 1.6260446310043335,
|
| 2219 |
+
"learning_rate": 8.02053590151805e-08,
|
| 2220 |
+
"loss": 0.1099,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 1.6478232618583495,
|
| 2225 |
+
"grad_norm": 1.7146011590957642,
|
| 2226 |
+
"learning_rate": 7.792656399690922e-08,
|
| 2227 |
+
"loss": 0.1167,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 1.6530214424951266,
|
| 2232 |
+
"grad_norm": 1.6423556804656982,
|
| 2233 |
+
"learning_rate": 7.567787125262449e-08,
|
| 2234 |
+
"loss": 0.1171,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 1.6582196231319037,
|
| 2239 |
+
"grad_norm": 1.5510940551757812,
|
| 2240 |
+
"learning_rate": 7.345944115907421e-08,
|
| 2241 |
+
"loss": 0.1013,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 1.6634178037686809,
|
| 2246 |
+
"grad_norm": 1.863503336906433,
|
| 2247 |
+
"learning_rate": 7.127143193467445e-08,
|
| 2248 |
+
"loss": 0.1423,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 1.668615984405458,
|
| 2253 |
+
"grad_norm": 1.7284936904907227,
|
| 2254 |
+
"learning_rate": 6.911399962822518e-08,
|
| 2255 |
+
"loss": 0.1112,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 1.6738141650422351,
|
| 2260 |
+
"grad_norm": 1.7205549478530884,
|
| 2261 |
+
"learning_rate": 6.698729810778064e-08,
|
| 2262 |
+
"loss": 0.136,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 1.6790123456790123,
|
| 2267 |
+
"grad_norm": 1.4762628078460693,
|
| 2268 |
+
"learning_rate": 6.48914790496759e-08,
|
| 2269 |
+
"loss": 0.1144,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 1.6842105263157894,
|
| 2274 |
+
"grad_norm": 1.8362925052642822,
|
| 2275 |
+
"learning_rate": 6.282669192770895e-08,
|
| 2276 |
+
"loss": 0.1328,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 1.6894087069525665,
|
| 2281 |
+
"grad_norm": 1.4527249336242676,
|
| 2282 |
+
"learning_rate": 6.079308400248029e-08,
|
| 2283 |
+
"loss": 0.1055,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 1.6946068875893436,
|
| 2288 |
+
"grad_norm": 1.7803164720535278,
|
| 2289 |
+
"learning_rate": 5.8790800310890456e-08,
|
| 2290 |
+
"loss": 0.1451,
|
| 2291 |
+
"step": 326
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 1.6998050682261208,
|
| 2295 |
+
"grad_norm": 2.059589147567749,
|
| 2296 |
+
"learning_rate": 5.6819983655795936e-08,
|
| 2297 |
+
"loss": 0.1458,
|
| 2298 |
+
"step": 327
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 1.705003248862898,
|
| 2302 |
+
"grad_norm": 1.7053784132003784,
|
| 2303 |
+
"learning_rate": 5.4880774595824245e-08,
|
| 2304 |
+
"loss": 0.1257,
|
| 2305 |
+
"step": 328
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 1.710201429499675,
|
| 2309 |
+
"grad_norm": 1.97287118434906,
|
| 2310 |
+
"learning_rate": 5.297331143534972e-08,
|
| 2311 |
+
"loss": 0.1381,
|
| 2312 |
+
"step": 329
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 1.7153996101364521,
|
| 2316 |
+
"grad_norm": 1.6465667486190796,
|
| 2317 |
+
"learning_rate": 5.109773021462921e-08,
|
| 2318 |
+
"loss": 0.1155,
|
| 2319 |
+
"step": 330
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 1.7205977907732293,
|
| 2323 |
+
"grad_norm": 1.6381739377975464,
|
| 2324 |
+
"learning_rate": 4.925416470009991e-08,
|
| 2325 |
+
"loss": 0.1224,
|
| 2326 |
+
"step": 331
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 1.7257959714100064,
|
| 2330 |
+
"grad_norm": 1.5278061628341675,
|
| 2331 |
+
"learning_rate": 4.744274637483936e-08,
|
| 2332 |
+
"loss": 0.1203,
|
| 2333 |
+
"step": 332
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 1.7309941520467835,
|
| 2337 |
+
"grad_norm": 1.7712794542312622,
|
| 2338 |
+
"learning_rate": 4.566360442918754e-08,
|
| 2339 |
+
"loss": 0.1334,
|
| 2340 |
+
"step": 333
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 1.7361923326835607,
|
| 2344 |
+
"grad_norm": 1.912316083908081,
|
| 2345 |
+
"learning_rate": 4.3916865751533306e-08,
|
| 2346 |
+
"loss": 0.1475,
|
| 2347 |
+
"step": 334
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 1.7413905133203378,
|
| 2351 |
+
"grad_norm": 1.7500931024551392,
|
| 2352 |
+
"learning_rate": 4.220265491926489e-08,
|
| 2353 |
+
"loss": 0.1165,
|
| 2354 |
+
"step": 335
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 1.746588693957115,
|
| 2358 |
+
"grad_norm": 1.7582532167434692,
|
| 2359 |
+
"learning_rate": 4.0521094189884696e-08,
|
| 2360 |
+
"loss": 0.119,
|
| 2361 |
+
"step": 336
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 1.751786874593892,
|
| 2365 |
+
"grad_norm": 1.6794097423553467,
|
| 2366 |
+
"learning_rate": 3.887230349229015e-08,
|
| 2367 |
+
"loss": 0.1093,
|
| 2368 |
+
"step": 337
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 1.7569850552306692,
|
| 2372 |
+
"grad_norm": 1.797913908958435,
|
| 2373 |
+
"learning_rate": 3.7256400418220256e-08,
|
| 2374 |
+
"loss": 0.1235,
|
| 2375 |
+
"step": 338
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 1.7621832358674463,
|
| 2379 |
+
"grad_norm": 1.6042097806930542,
|
| 2380 |
+
"learning_rate": 3.567350021386895e-08,
|
| 2381 |
+
"loss": 0.096,
|
| 2382 |
+
"step": 339
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 1.7673814165042234,
|
| 2386 |
+
"grad_norm": 1.7623355388641357,
|
| 2387 |
+
"learning_rate": 3.412371577166578e-08,
|
| 2388 |
+
"loss": 0.1153,
|
| 2389 |
+
"step": 340
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 1.7725795971410006,
|
| 2393 |
+
"grad_norm": 1.737052321434021,
|
| 2394 |
+
"learning_rate": 3.260715762222449e-08,
|
| 2395 |
+
"loss": 0.1327,
|
| 2396 |
+
"step": 341
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 1.7777777777777777,
|
| 2400 |
+
"grad_norm": 1.61103355884552,
|
| 2401 |
+
"learning_rate": 3.1123933926459845e-08,
|
| 2402 |
+
"loss": 0.1213,
|
| 2403 |
+
"step": 342
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 1.7829759584145548,
|
| 2407 |
+
"grad_norm": 1.78020441532135,
|
| 2408 |
+
"learning_rate": 2.9674150467873527e-08,
|
| 2409 |
+
"loss": 0.1262,
|
| 2410 |
+
"step": 343
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 1.788174139051332,
|
| 2414 |
+
"grad_norm": 1.7625352144241333,
|
| 2415 |
+
"learning_rate": 2.825791064500993e-08,
|
| 2416 |
+
"loss": 0.1465,
|
| 2417 |
+
"step": 344
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 1.793372319688109,
|
| 2421 |
+
"grad_norm": 1.7474111318588257,
|
| 2422 |
+
"learning_rate": 2.6875315464081562e-08,
|
| 2423 |
+
"loss": 0.1289,
|
| 2424 |
+
"step": 345
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 1.7985705003248862,
|
| 2428 |
+
"grad_norm": 1.7776579856872559,
|
| 2429 |
+
"learning_rate": 2.5526463531765463e-08,
|
| 2430 |
+
"loss": 0.1219,
|
| 2431 |
+
"step": 346
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 1.8037686809616633,
|
| 2435 |
+
"grad_norm": 1.7461159229278564,
|
| 2436 |
+
"learning_rate": 2.4211451048170296e-08,
|
| 2437 |
+
"loss": 0.1361,
|
| 2438 |
+
"step": 347
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 1.8089668615984404,
|
| 2442 |
+
"grad_norm": 1.701060175895691,
|
| 2443 |
+
"learning_rate": 2.293037179997559e-08,
|
| 2444 |
+
"loss": 0.1463,
|
| 2445 |
+
"step": 348
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 1.8141650422352176,
|
| 2449 |
+
"grad_norm": 1.7974237203598022,
|
| 2450 |
+
"learning_rate": 2.1683317153742775e-08,
|
| 2451 |
+
"loss": 0.14,
|
| 2452 |
+
"step": 349
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 1.8193632228719947,
|
| 2456 |
+
"grad_norm": 1.5575218200683594,
|
| 2457 |
+
"learning_rate": 2.047037604939894e-08,
|
| 2458 |
+
"loss": 0.1234,
|
| 2459 |
+
"step": 350
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 1.8245614035087718,
|
| 2463 |
+
"grad_norm": 1.6900442838668823,
|
| 2464 |
+
"learning_rate": 1.92916349938938e-08,
|
| 2465 |
+
"loss": 0.1119,
|
| 2466 |
+
"step": 351
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 1.829759584145549,
|
| 2470 |
+
"grad_norm": 1.7725763320922852,
|
| 2471 |
+
"learning_rate": 1.8147178055029577e-08,
|
| 2472 |
+
"loss": 0.1456,
|
| 2473 |
+
"step": 352
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 1.834957764782326,
|
| 2477 |
+
"grad_norm": 1.677262544631958,
|
| 2478 |
+
"learning_rate": 1.7037086855465898e-08,
|
| 2479 |
+
"loss": 0.1041,
|
| 2480 |
+
"step": 353
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 1.8401559454191032,
|
| 2484 |
+
"grad_norm": 1.5990703105926514,
|
| 2485 |
+
"learning_rate": 1.596144056689791e-08,
|
| 2486 |
+
"loss": 0.1149,
|
| 2487 |
+
"step": 354
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 1.8453541260558803,
|
| 2491 |
+
"grad_norm": 2.083341598510742,
|
| 2492 |
+
"learning_rate": 1.4920315904410064e-08,
|
| 2493 |
+
"loss": 0.1375,
|
| 2494 |
+
"step": 355
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 1.8505523066926575,
|
| 2498 |
+
"grad_norm": 1.7379205226898193,
|
| 2499 |
+
"learning_rate": 1.3913787121004716e-08,
|
| 2500 |
+
"loss": 0.1365,
|
| 2501 |
+
"step": 356
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 1.8557504873294346,
|
| 2505 |
+
"grad_norm": 1.6175756454467773,
|
| 2506 |
+
"learning_rate": 1.2941926002306536e-08,
|
| 2507 |
+
"loss": 0.1348,
|
| 2508 |
+
"step": 357
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 1.8609486679662117,
|
| 2512 |
+
"grad_norm": 1.7680426836013794,
|
| 2513 |
+
"learning_rate": 1.200480186144237e-08,
|
| 2514 |
+
"loss": 0.1232,
|
| 2515 |
+
"step": 358
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 1.8661468486029889,
|
| 2519 |
+
"grad_norm": 1.3798372745513916,
|
| 2520 |
+
"learning_rate": 1.1102481534098374e-08,
|
| 2521 |
+
"loss": 0.0966,
|
| 2522 |
+
"step": 359
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 1.871345029239766,
|
| 2526 |
+
"grad_norm": 1.584373950958252,
|
| 2527 |
+
"learning_rate": 1.0235029373752757e-08,
|
| 2528 |
+
"loss": 0.1082,
|
| 2529 |
+
"step": 360
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 1.876543209876543,
|
| 2533 |
+
"grad_norm": 1.7235833406448364,
|
| 2534 |
+
"learning_rate": 9.402507247086578e-09,
|
| 2535 |
+
"loss": 0.1348,
|
| 2536 |
+
"step": 361
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 1.8817413905133202,
|
| 2540 |
+
"grad_norm": 1.6077182292938232,
|
| 2541 |
+
"learning_rate": 8.60497452957104e-09,
|
| 2542 |
+
"loss": 0.1234,
|
| 2543 |
+
"step": 362
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 1.8869395711500974,
|
| 2547 |
+
"grad_norm": 1.5776941776275635,
|
| 2548 |
+
"learning_rate": 7.842488101232891e-09,
|
| 2549 |
+
"loss": 0.1177,
|
| 2550 |
+
"step": 363
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 1.8921377517868745,
|
| 2554 |
+
"grad_norm": 1.5909714698791504,
|
| 2555 |
+
"learning_rate": 7.115102342598101e-09,
|
| 2556 |
+
"loss": 0.1133,
|
| 2557 |
+
"step": 364
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 1.8973359324236516,
|
| 2561 |
+
"grad_norm": 1.7370619773864746,
|
| 2562 |
+
"learning_rate": 6.422869130812913e-09,
|
| 2563 |
+
"loss": 0.1269,
|
| 2564 |
+
"step": 365
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 1.9025341130604287,
|
| 2568 |
+
"grad_norm": 1.7405683994293213,
|
| 2569 |
+
"learning_rate": 5.765837835944309e-09,
|
| 2570 |
+
"loss": 0.1307,
|
| 2571 |
+
"step": 366
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 1.9077322936972059,
|
| 2575 |
+
"grad_norm": 1.9631272554397583,
|
| 2576 |
+
"learning_rate": 5.144055317458817e-09,
|
| 2577 |
+
"loss": 0.1546,
|
| 2578 |
+
"step": 367
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 1.912930474333983,
|
| 2582 |
+
"grad_norm": 1.6291249990463257,
|
| 2583 |
+
"learning_rate": 4.55756592088058e-09,
|
| 2584 |
+
"loss": 0.1249,
|
| 2585 |
+
"step": 368
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 1.9181286549707601,
|
| 2589 |
+
"grad_norm": 1.528748869895935,
|
| 2590 |
+
"learning_rate": 4.0064114746284905e-09,
|
| 2591 |
+
"loss": 0.1027,
|
| 2592 |
+
"step": 369
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 1.9233268356075373,
|
| 2596 |
+
"grad_norm": 1.8311703205108643,
|
| 2597 |
+
"learning_rate": 3.4906312870331965e-09,
|
| 2598 |
+
"loss": 0.1128,
|
| 2599 |
+
"step": 370
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 1.9285250162443144,
|
| 2603 |
+
"grad_norm": 1.8863190412521362,
|
| 2604 |
+
"learning_rate": 3.010262143533393e-09,
|
| 2605 |
+
"loss": 0.1441,
|
| 2606 |
+
"step": 371
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 1.9337231968810915,
|
| 2610 |
+
"grad_norm": 1.5785506963729858,
|
| 2611 |
+
"learning_rate": 2.5653383040524224e-09,
|
| 2612 |
+
"loss": 0.1198,
|
| 2613 |
+
"step": 372
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 1.9389213775178686,
|
| 2617 |
+
"grad_norm": 1.6531542539596558,
|
| 2618 |
+
"learning_rate": 2.155891500554896e-09,
|
| 2619 |
+
"loss": 0.1224,
|
| 2620 |
+
"step": 373
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 1.9441195581546458,
|
| 2624 |
+
"grad_norm": 1.5542500019073486,
|
| 2625 |
+
"learning_rate": 1.7819509347835049e-09,
|
| 2626 |
+
"loss": 0.1149,
|
| 2627 |
+
"step": 374
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 1.949317738791423,
|
| 2631 |
+
"grad_norm": 1.5082885026931763,
|
| 2632 |
+
"learning_rate": 1.4435432761762955e-09,
|
| 2633 |
+
"loss": 0.1061,
|
| 2634 |
+
"step": 375
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 1.9545159194282,
|
| 2638 |
+
"grad_norm": 1.5856765508651733,
|
| 2639 |
+
"learning_rate": 1.1406926599646372e-09,
|
| 2640 |
+
"loss": 0.1327,
|
| 2641 |
+
"step": 376
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 1.9597141000649771,
|
| 2645 |
+
"grad_norm": 1.6566141843795776,
|
| 2646 |
+
"learning_rate": 8.73420685452042e-10,
|
| 2647 |
+
"loss": 0.1105,
|
| 2648 |
+
"step": 377
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 1.9649122807017543,
|
| 2652 |
+
"grad_norm": 1.7047754526138306,
|
| 2653 |
+
"learning_rate": 6.417464144736207e-10,
|
| 2654 |
+
"loss": 0.1301,
|
| 2655 |
+
"step": 378
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 1.9701104613385314,
|
| 2659 |
+
"grad_norm": 1.7699244022369385,
|
| 2660 |
+
"learning_rate": 4.4568637003633556e-10,
|
| 2661 |
+
"loss": 0.1412,
|
| 2662 |
+
"step": 379
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 1.9753086419753085,
|
| 2666 |
+
"grad_norm": 1.8177380561828613,
|
| 2667 |
+
"learning_rate": 2.852545351409996e-10,
|
| 2668 |
+
"loss": 0.1571,
|
| 2669 |
+
"step": 380
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 1.9805068226120857,
|
| 2673 |
+
"grad_norm": 1.5334748029708862,
|
| 2674 |
+
"learning_rate": 1.6046235178474033e-10,
|
| 2675 |
+
"loss": 0.104,
|
| 2676 |
+
"step": 381
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 1.9857050032488628,
|
| 2680 |
+
"grad_norm": 1.7260781526565552,
|
| 2681 |
+
"learning_rate": 7.13187201450971e-11,
|
| 2682 |
+
"loss": 0.1263,
|
| 2683 |
+
"step": 382
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 1.99090318388564,
|
| 2687 |
+
"grad_norm": 1.6607623100280762,
|
| 2688 |
+
"learning_rate": 1.7829997945084662e-11,
|
| 2689 |
+
"loss": 0.1135,
|
| 2690 |
+
"step": 383
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 1.996101364522417,
|
| 2694 |
+
"grad_norm": 1.7004939317703247,
|
| 2695 |
+
"learning_rate": 0.0,
|
| 2696 |
+
"loss": 0.1315,
|
| 2697 |
+
"step": 384
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 1.996101364522417,
|
| 2701 |
+
"step": 384,
|
| 2702 |
+
"total_flos": 3.0550374682743276e+18,
|
| 2703 |
+
"train_loss": 0.19105235013800362,
|
| 2704 |
+
"train_runtime": 6828.3197,
|
| 2705 |
+
"train_samples_per_second": 7.211,
|
| 2706 |
+
"train_steps_per_second": 0.056
|
| 2707 |
+
}
|
| 2708 |
+
],
|
| 2709 |
+
"logging_steps": 1.0,
|
| 2710 |
+
"max_steps": 384,
|
| 2711 |
+
"num_input_tokens_seen": 0,
|
| 2712 |
+
"num_train_epochs": 2,
|
| 2713 |
+
"save_steps": 500,
|
| 2714 |
+
"stateful_callbacks": {
|
| 2715 |
+
"TrainerControl": {
|
| 2716 |
+
"args": {
|
| 2717 |
+
"should_epoch_stop": false,
|
| 2718 |
+
"should_evaluate": false,
|
| 2719 |
+
"should_log": false,
|
| 2720 |
+
"should_save": true,
|
| 2721 |
+
"should_training_stop": true
|
| 2722 |
+
},
|
| 2723 |
+
"attributes": {}
|
| 2724 |
+
}
|
| 2725 |
+
},
|
| 2726 |
+
"total_flos": 3.0550374682743276e+18,
|
| 2727 |
+
"train_batch_size": 1,
|
| 2728 |
+
"trial_name": null,
|
| 2729 |
+
"trial_params": null
|
| 2730 |
+
}
|