Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2538 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2538 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0863531225905936,
|
| 6 |
+
"eval_steps": 64,
|
| 7 |
+
"global_step": 352,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.003108003108003108,
|
| 14 |
+
"grad_norm": 10.027831077575684,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.6767,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.006216006216006216,
|
| 21 |
+
"grad_norm": 9.679778099060059,
|
| 22 |
+
"learning_rate": 5.000000000000001e-07,
|
| 23 |
+
"loss": 0.6644,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.009324009324009324,
|
| 28 |
+
"grad_norm": 10.520271301269531,
|
| 29 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 30 |
+
"loss": 0.6934,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.012432012432012432,
|
| 35 |
+
"grad_norm": 8.677583694458008,
|
| 36 |
+
"learning_rate": 1.5e-06,
|
| 37 |
+
"loss": 0.6617,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.01554001554001554,
|
| 42 |
+
"grad_norm": 6.502548694610596,
|
| 43 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 44 |
+
"loss": 0.6509,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.018648018648018648,
|
| 49 |
+
"grad_norm": 4.257171154022217,
|
| 50 |
+
"learning_rate": 2.5e-06,
|
| 51 |
+
"loss": 0.639,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.021756021756021756,
|
| 56 |
+
"grad_norm": 3.460066556930542,
|
| 57 |
+
"learning_rate": 3e-06,
|
| 58 |
+
"loss": 0.6286,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.024864024864024864,
|
| 63 |
+
"grad_norm": 3.0126283168792725,
|
| 64 |
+
"learning_rate": 3.5e-06,
|
| 65 |
+
"loss": 0.5948,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.027972027972027972,
|
| 70 |
+
"grad_norm": 2.567995309829712,
|
| 71 |
+
"learning_rate": 4.000000000000001e-06,
|
| 72 |
+
"loss": 0.5744,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.03108003108003108,
|
| 77 |
+
"grad_norm": 2.516597032546997,
|
| 78 |
+
"learning_rate": 4.5e-06,
|
| 79 |
+
"loss": 0.5496,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.03418803418803419,
|
| 84 |
+
"grad_norm": 1.8187586069107056,
|
| 85 |
+
"learning_rate": 5e-06,
|
| 86 |
+
"loss": 0.5397,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.037296037296037296,
|
| 91 |
+
"grad_norm": 1.7935529947280884,
|
| 92 |
+
"learning_rate": 5.500000000000001e-06,
|
| 93 |
+
"loss": 0.5229,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.04040404040404041,
|
| 98 |
+
"grad_norm": 1.8665963411331177,
|
| 99 |
+
"learning_rate": 6e-06,
|
| 100 |
+
"loss": 0.5227,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.04351204351204351,
|
| 105 |
+
"grad_norm": 2.0106680393218994,
|
| 106 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 107 |
+
"loss": 0.4882,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.046620046620046623,
|
| 112 |
+
"grad_norm": 3.305211305618286,
|
| 113 |
+
"learning_rate": 7e-06,
|
| 114 |
+
"loss": 0.4772,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.04972804972804973,
|
| 119 |
+
"grad_norm": 3.047219753265381,
|
| 120 |
+
"learning_rate": 7.500000000000001e-06,
|
| 121 |
+
"loss": 0.452,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.05283605283605284,
|
| 126 |
+
"grad_norm": 2.5453591346740723,
|
| 127 |
+
"learning_rate": 8.000000000000001e-06,
|
| 128 |
+
"loss": 0.4138,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.055944055944055944,
|
| 133 |
+
"grad_norm": 5.414841175079346,
|
| 134 |
+
"learning_rate": 8.5e-06,
|
| 135 |
+
"loss": 0.4238,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.059052059052059055,
|
| 140 |
+
"grad_norm": 2.979440927505493,
|
| 141 |
+
"learning_rate": 9e-06,
|
| 142 |
+
"loss": 0.3987,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.06216006216006216,
|
| 147 |
+
"grad_norm": 1.981175422668457,
|
| 148 |
+
"learning_rate": 9.5e-06,
|
| 149 |
+
"loss": 0.3874,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.06526806526806526,
|
| 154 |
+
"grad_norm": 1.7793089151382446,
|
| 155 |
+
"learning_rate": 1e-05,
|
| 156 |
+
"loss": 0.3631,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.06837606837606838,
|
| 161 |
+
"grad_norm": 1.1854480504989624,
|
| 162 |
+
"learning_rate": 9.989429175475688e-06,
|
| 163 |
+
"loss": 0.3765,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.07148407148407149,
|
| 168 |
+
"grad_norm": 0.8928348422050476,
|
| 169 |
+
"learning_rate": 9.978858350951375e-06,
|
| 170 |
+
"loss": 0.3481,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.07459207459207459,
|
| 175 |
+
"grad_norm": 1.7531942129135132,
|
| 176 |
+
"learning_rate": 9.968287526427062e-06,
|
| 177 |
+
"loss": 0.3693,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.0777000777000777,
|
| 182 |
+
"grad_norm": 1.0829464197158813,
|
| 183 |
+
"learning_rate": 9.957716701902749e-06,
|
| 184 |
+
"loss": 0.3644,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.08080808080808081,
|
| 189 |
+
"grad_norm": 0.98089200258255,
|
| 190 |
+
"learning_rate": 9.947145877378436e-06,
|
| 191 |
+
"loss": 0.3616,
|
| 192 |
+
"step": 26
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.08391608391608392,
|
| 196 |
+
"grad_norm": 0.795221745967865,
|
| 197 |
+
"learning_rate": 9.936575052854123e-06,
|
| 198 |
+
"loss": 0.3679,
|
| 199 |
+
"step": 27
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.08702408702408702,
|
| 203 |
+
"grad_norm": 1.091843605041504,
|
| 204 |
+
"learning_rate": 9.92600422832981e-06,
|
| 205 |
+
"loss": 0.3439,
|
| 206 |
+
"step": 28
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.09013209013209013,
|
| 210 |
+
"grad_norm": 0.8538377285003662,
|
| 211 |
+
"learning_rate": 9.915433403805497e-06,
|
| 212 |
+
"loss": 0.3401,
|
| 213 |
+
"step": 29
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.09324009324009325,
|
| 217 |
+
"grad_norm": 0.9114591479301453,
|
| 218 |
+
"learning_rate": 9.904862579281184e-06,
|
| 219 |
+
"loss": 0.3515,
|
| 220 |
+
"step": 30
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.09634809634809635,
|
| 224 |
+
"grad_norm": 0.9083001017570496,
|
| 225 |
+
"learning_rate": 9.894291754756871e-06,
|
| 226 |
+
"loss": 0.3449,
|
| 227 |
+
"step": 31
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.09945609945609946,
|
| 231 |
+
"grad_norm": 0.9144365787506104,
|
| 232 |
+
"learning_rate": 9.883720930232558e-06,
|
| 233 |
+
"loss": 0.3393,
|
| 234 |
+
"step": 32
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.10256410256410256,
|
| 238 |
+
"grad_norm": 1.0221809148788452,
|
| 239 |
+
"learning_rate": 9.873150105708245e-06,
|
| 240 |
+
"loss": 0.353,
|
| 241 |
+
"step": 33
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.10567210567210568,
|
| 245 |
+
"grad_norm": 1.0219439268112183,
|
| 246 |
+
"learning_rate": 9.862579281183932e-06,
|
| 247 |
+
"loss": 0.3439,
|
| 248 |
+
"step": 34
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.10878010878010878,
|
| 252 |
+
"grad_norm": 1.5430618524551392,
|
| 253 |
+
"learning_rate": 9.852008456659621e-06,
|
| 254 |
+
"loss": 0.3338,
|
| 255 |
+
"step": 35
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.11188811188811189,
|
| 259 |
+
"grad_norm": 1.4754544496536255,
|
| 260 |
+
"learning_rate": 9.841437632135308e-06,
|
| 261 |
+
"loss": 0.3363,
|
| 262 |
+
"step": 36
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.11499611499611499,
|
| 266 |
+
"grad_norm": 1.1298989057540894,
|
| 267 |
+
"learning_rate": 9.830866807610995e-06,
|
| 268 |
+
"loss": 0.3423,
|
| 269 |
+
"step": 37
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.11810411810411811,
|
| 273 |
+
"grad_norm": 1.0130062103271484,
|
| 274 |
+
"learning_rate": 9.820295983086682e-06,
|
| 275 |
+
"loss": 0.3298,
|
| 276 |
+
"step": 38
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.12121212121212122,
|
| 280 |
+
"grad_norm": 1.8003513813018799,
|
| 281 |
+
"learning_rate": 9.80972515856237e-06,
|
| 282 |
+
"loss": 0.3272,
|
| 283 |
+
"step": 39
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.12432012432012432,
|
| 287 |
+
"grad_norm": 0.9532265067100525,
|
| 288 |
+
"learning_rate": 9.799154334038056e-06,
|
| 289 |
+
"loss": 0.3282,
|
| 290 |
+
"step": 40
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.12742812742812742,
|
| 294 |
+
"grad_norm": 1.5232913494110107,
|
| 295 |
+
"learning_rate": 9.788583509513743e-06,
|
| 296 |
+
"loss": 0.3469,
|
| 297 |
+
"step": 41
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.13053613053613053,
|
| 301 |
+
"grad_norm": 0.8918169736862183,
|
| 302 |
+
"learning_rate": 9.77801268498943e-06,
|
| 303 |
+
"loss": 0.326,
|
| 304 |
+
"step": 42
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.13364413364413363,
|
| 308 |
+
"grad_norm": 0.8845950365066528,
|
| 309 |
+
"learning_rate": 9.767441860465117e-06,
|
| 310 |
+
"loss": 0.3313,
|
| 311 |
+
"step": 43
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.13675213675213677,
|
| 315 |
+
"grad_norm": 0.8410794138908386,
|
| 316 |
+
"learning_rate": 9.756871035940804e-06,
|
| 317 |
+
"loss": 0.3318,
|
| 318 |
+
"step": 44
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.13986013986013987,
|
| 322 |
+
"grad_norm": 0.7157808542251587,
|
| 323 |
+
"learning_rate": 9.746300211416491e-06,
|
| 324 |
+
"loss": 0.3381,
|
| 325 |
+
"step": 45
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.14296814296814297,
|
| 329 |
+
"grad_norm": 1.1680670976638794,
|
| 330 |
+
"learning_rate": 9.735729386892178e-06,
|
| 331 |
+
"loss": 0.3281,
|
| 332 |
+
"step": 46
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.14607614607614608,
|
| 336 |
+
"grad_norm": 0.9500836133956909,
|
| 337 |
+
"learning_rate": 9.725158562367865e-06,
|
| 338 |
+
"loss": 0.336,
|
| 339 |
+
"step": 47
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.14918414918414918,
|
| 343 |
+
"grad_norm": 0.8565309643745422,
|
| 344 |
+
"learning_rate": 9.714587737843552e-06,
|
| 345 |
+
"loss": 0.3207,
|
| 346 |
+
"step": 48
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.1522921522921523,
|
| 350 |
+
"grad_norm": 1.1311777830123901,
|
| 351 |
+
"learning_rate": 9.70401691331924e-06,
|
| 352 |
+
"loss": 0.3339,
|
| 353 |
+
"step": 49
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.1554001554001554,
|
| 357 |
+
"grad_norm": 1.0368160009384155,
|
| 358 |
+
"learning_rate": 9.693446088794927e-06,
|
| 359 |
+
"loss": 0.3262,
|
| 360 |
+
"step": 50
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.1585081585081585,
|
| 364 |
+
"grad_norm": 0.9648517370223999,
|
| 365 |
+
"learning_rate": 9.682875264270614e-06,
|
| 366 |
+
"loss": 0.3376,
|
| 367 |
+
"step": 51
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.16161616161616163,
|
| 371 |
+
"grad_norm": 1.1039059162139893,
|
| 372 |
+
"learning_rate": 9.6723044397463e-06,
|
| 373 |
+
"loss": 0.3352,
|
| 374 |
+
"step": 52
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.16472416472416473,
|
| 378 |
+
"grad_norm": 1.0544918775558472,
|
| 379 |
+
"learning_rate": 9.661733615221988e-06,
|
| 380 |
+
"loss": 0.3237,
|
| 381 |
+
"step": 53
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.16783216783216784,
|
| 385 |
+
"grad_norm": 1.533158302307129,
|
| 386 |
+
"learning_rate": 9.651162790697676e-06,
|
| 387 |
+
"loss": 0.3287,
|
| 388 |
+
"step": 54
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.17094017094017094,
|
| 392 |
+
"grad_norm": 1.2342826128005981,
|
| 393 |
+
"learning_rate": 9.640591966173363e-06,
|
| 394 |
+
"loss": 0.3162,
|
| 395 |
+
"step": 55
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.17404817404817405,
|
| 399 |
+
"grad_norm": 1.0702942609786987,
|
| 400 |
+
"learning_rate": 9.63002114164905e-06,
|
| 401 |
+
"loss": 0.3143,
|
| 402 |
+
"step": 56
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.17715617715617715,
|
| 406 |
+
"grad_norm": 1.02211594581604,
|
| 407 |
+
"learning_rate": 9.619450317124736e-06,
|
| 408 |
+
"loss": 0.3318,
|
| 409 |
+
"step": 57
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.18026418026418026,
|
| 413 |
+
"grad_norm": 0.8379388451576233,
|
| 414 |
+
"learning_rate": 9.608879492600423e-06,
|
| 415 |
+
"loss": 0.3239,
|
| 416 |
+
"step": 58
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.18337218337218336,
|
| 420 |
+
"grad_norm": 0.9620960354804993,
|
| 421 |
+
"learning_rate": 9.59830866807611e-06,
|
| 422 |
+
"loss": 0.3246,
|
| 423 |
+
"step": 59
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.1864801864801865,
|
| 427 |
+
"grad_norm": 0.9239097833633423,
|
| 428 |
+
"learning_rate": 9.587737843551797e-06,
|
| 429 |
+
"loss": 0.3278,
|
| 430 |
+
"step": 60
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.1895881895881896,
|
| 434 |
+
"grad_norm": 0.7097995281219482,
|
| 435 |
+
"learning_rate": 9.577167019027484e-06,
|
| 436 |
+
"loss": 0.3152,
|
| 437 |
+
"step": 61
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.1926961926961927,
|
| 441 |
+
"grad_norm": 0.9077997803688049,
|
| 442 |
+
"learning_rate": 9.566596194503171e-06,
|
| 443 |
+
"loss": 0.3219,
|
| 444 |
+
"step": 62
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.1958041958041958,
|
| 448 |
+
"grad_norm": 0.8704112768173218,
|
| 449 |
+
"learning_rate": 9.55602536997886e-06,
|
| 450 |
+
"loss": 0.3262,
|
| 451 |
+
"step": 63
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.1989121989121989,
|
| 455 |
+
"grad_norm": 0.9264605641365051,
|
| 456 |
+
"learning_rate": 9.545454545454547e-06,
|
| 457 |
+
"loss": 0.3176,
|
| 458 |
+
"step": 64
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.1989121989121989,
|
| 462 |
+
"eval_loss": 0.3377174139022827,
|
| 463 |
+
"eval_runtime": 149.1316,
|
| 464 |
+
"eval_samples_per_second": 1.911,
|
| 465 |
+
"eval_steps_per_second": 0.959,
|
| 466 |
+
"step": 64
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 0.20202020202020202,
|
| 470 |
+
"grad_norm": 0.9881049394607544,
|
| 471 |
+
"learning_rate": 9.534883720930234e-06,
|
| 472 |
+
"loss": 0.3312,
|
| 473 |
+
"step": 65
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.20512820512820512,
|
| 477 |
+
"grad_norm": 1.1825007200241089,
|
| 478 |
+
"learning_rate": 9.524312896405921e-06,
|
| 479 |
+
"loss": 0.3189,
|
| 480 |
+
"step": 66
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.20823620823620823,
|
| 484 |
+
"grad_norm": 0.8272495865821838,
|
| 485 |
+
"learning_rate": 9.513742071881608e-06,
|
| 486 |
+
"loss": 0.3293,
|
| 487 |
+
"step": 67
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.21134421134421136,
|
| 491 |
+
"grad_norm": 1.0992769002914429,
|
| 492 |
+
"learning_rate": 9.503171247357295e-06,
|
| 493 |
+
"loss": 0.3119,
|
| 494 |
+
"step": 68
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.21445221445221446,
|
| 498 |
+
"grad_norm": 0.9182390570640564,
|
| 499 |
+
"learning_rate": 9.492600422832982e-06,
|
| 500 |
+
"loss": 0.331,
|
| 501 |
+
"step": 69
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.21756021756021757,
|
| 505 |
+
"grad_norm": 0.8677308559417725,
|
| 506 |
+
"learning_rate": 9.482029598308669e-06,
|
| 507 |
+
"loss": 0.3168,
|
| 508 |
+
"step": 70
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.22066822066822067,
|
| 512 |
+
"grad_norm": 1.2915256023406982,
|
| 513 |
+
"learning_rate": 9.471458773784356e-06,
|
| 514 |
+
"loss": 0.3181,
|
| 515 |
+
"step": 71
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.22377622377622378,
|
| 519 |
+
"grad_norm": 1.6176910400390625,
|
| 520 |
+
"learning_rate": 9.460887949260043e-06,
|
| 521 |
+
"loss": 0.3254,
|
| 522 |
+
"step": 72
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.22688422688422688,
|
| 526 |
+
"grad_norm": 0.6357202529907227,
|
| 527 |
+
"learning_rate": 9.45031712473573e-06,
|
| 528 |
+
"loss": 0.3298,
|
| 529 |
+
"step": 73
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.22999222999222999,
|
| 533 |
+
"grad_norm": 0.911662220954895,
|
| 534 |
+
"learning_rate": 9.439746300211417e-06,
|
| 535 |
+
"loss": 0.3248,
|
| 536 |
+
"step": 74
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.2331002331002331,
|
| 540 |
+
"grad_norm": 0.7426556944847107,
|
| 541 |
+
"learning_rate": 9.429175475687104e-06,
|
| 542 |
+
"loss": 0.3301,
|
| 543 |
+
"step": 75
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.23620823620823622,
|
| 547 |
+
"grad_norm": 0.7509779930114746,
|
| 548 |
+
"learning_rate": 9.418604651162791e-06,
|
| 549 |
+
"loss": 0.3209,
|
| 550 |
+
"step": 76
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.23931623931623933,
|
| 554 |
+
"grad_norm": 0.7699870467185974,
|
| 555 |
+
"learning_rate": 9.408033826638478e-06,
|
| 556 |
+
"loss": 0.3171,
|
| 557 |
+
"step": 77
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.24242424242424243,
|
| 561 |
+
"grad_norm": 0.7583193182945251,
|
| 562 |
+
"learning_rate": 9.397463002114165e-06,
|
| 563 |
+
"loss": 0.3128,
|
| 564 |
+
"step": 78
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.24553224553224554,
|
| 568 |
+
"grad_norm": 0.968973696231842,
|
| 569 |
+
"learning_rate": 9.386892177589852e-06,
|
| 570 |
+
"loss": 0.3293,
|
| 571 |
+
"step": 79
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.24864024864024864,
|
| 575 |
+
"grad_norm": 0.9967902302742004,
|
| 576 |
+
"learning_rate": 9.37632135306554e-06,
|
| 577 |
+
"loss": 0.3209,
|
| 578 |
+
"step": 80
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.2517482517482518,
|
| 582 |
+
"grad_norm": 0.7837809920310974,
|
| 583 |
+
"learning_rate": 9.365750528541226e-06,
|
| 584 |
+
"loss": 0.3152,
|
| 585 |
+
"step": 81
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.25485625485625485,
|
| 589 |
+
"grad_norm": 1.6905367374420166,
|
| 590 |
+
"learning_rate": 9.355179704016915e-06,
|
| 591 |
+
"loss": 0.3163,
|
| 592 |
+
"step": 82
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.257964257964258,
|
| 596 |
+
"grad_norm": 0.8734452128410339,
|
| 597 |
+
"learning_rate": 9.344608879492602e-06,
|
| 598 |
+
"loss": 0.3306,
|
| 599 |
+
"step": 83
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.26107226107226106,
|
| 603 |
+
"grad_norm": 3.6059653759002686,
|
| 604 |
+
"learning_rate": 9.33403805496829e-06,
|
| 605 |
+
"loss": 0.3104,
|
| 606 |
+
"step": 84
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.2641802641802642,
|
| 610 |
+
"grad_norm": 1.1703656911849976,
|
| 611 |
+
"learning_rate": 9.323467230443976e-06,
|
| 612 |
+
"loss": 0.3071,
|
| 613 |
+
"step": 85
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 0.26728826728826727,
|
| 617 |
+
"grad_norm": 0.8762909770011902,
|
| 618 |
+
"learning_rate": 9.312896405919663e-06,
|
| 619 |
+
"loss": 0.3022,
|
| 620 |
+
"step": 86
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.2703962703962704,
|
| 624 |
+
"grad_norm": 2.158876419067383,
|
| 625 |
+
"learning_rate": 9.30232558139535e-06,
|
| 626 |
+
"loss": 0.3217,
|
| 627 |
+
"step": 87
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.27350427350427353,
|
| 631 |
+
"grad_norm": 0.8010348081588745,
|
| 632 |
+
"learning_rate": 9.291754756871036e-06,
|
| 633 |
+
"loss": 0.322,
|
| 634 |
+
"step": 88
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 0.2766122766122766,
|
| 638 |
+
"grad_norm": 1.119739055633545,
|
| 639 |
+
"learning_rate": 9.281183932346723e-06,
|
| 640 |
+
"loss": 0.3248,
|
| 641 |
+
"step": 89
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.27972027972027974,
|
| 645 |
+
"grad_norm": 0.7900079488754272,
|
| 646 |
+
"learning_rate": 9.27061310782241e-06,
|
| 647 |
+
"loss": 0.3102,
|
| 648 |
+
"step": 90
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.2828282828282828,
|
| 652 |
+
"grad_norm": 0.8093041181564331,
|
| 653 |
+
"learning_rate": 9.260042283298098e-06,
|
| 654 |
+
"loss": 0.3259,
|
| 655 |
+
"step": 91
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.28593628593628595,
|
| 659 |
+
"grad_norm": 0.7240622043609619,
|
| 660 |
+
"learning_rate": 9.249471458773785e-06,
|
| 661 |
+
"loss": 0.3002,
|
| 662 |
+
"step": 92
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.289044289044289,
|
| 666 |
+
"grad_norm": 0.9449782371520996,
|
| 667 |
+
"learning_rate": 9.238900634249473e-06,
|
| 668 |
+
"loss": 0.3076,
|
| 669 |
+
"step": 93
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.29215229215229216,
|
| 673 |
+
"grad_norm": 0.9448596835136414,
|
| 674 |
+
"learning_rate": 9.22832980972516e-06,
|
| 675 |
+
"loss": 0.3012,
|
| 676 |
+
"step": 94
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.29526029526029524,
|
| 680 |
+
"grad_norm": 0.9209067821502686,
|
| 681 |
+
"learning_rate": 9.217758985200847e-06,
|
| 682 |
+
"loss": 0.3131,
|
| 683 |
+
"step": 95
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.29836829836829837,
|
| 687 |
+
"grad_norm": 0.878709614276886,
|
| 688 |
+
"learning_rate": 9.207188160676534e-06,
|
| 689 |
+
"loss": 0.3157,
|
| 690 |
+
"step": 96
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.3014763014763015,
|
| 694 |
+
"grad_norm": 1.1178463697433472,
|
| 695 |
+
"learning_rate": 9.19661733615222e-06,
|
| 696 |
+
"loss": 0.3166,
|
| 697 |
+
"step": 97
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.3045843045843046,
|
| 701 |
+
"grad_norm": 0.9717866778373718,
|
| 702 |
+
"learning_rate": 9.186046511627908e-06,
|
| 703 |
+
"loss": 0.3144,
|
| 704 |
+
"step": 98
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.3076923076923077,
|
| 708 |
+
"grad_norm": 0.9905857443809509,
|
| 709 |
+
"learning_rate": 9.175475687103595e-06,
|
| 710 |
+
"loss": 0.3263,
|
| 711 |
+
"step": 99
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.3108003108003108,
|
| 715 |
+
"grad_norm": 1.0447399616241455,
|
| 716 |
+
"learning_rate": 9.164904862579282e-06,
|
| 717 |
+
"loss": 0.3074,
|
| 718 |
+
"step": 100
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.3139083139083139,
|
| 722 |
+
"grad_norm": 0.9876366853713989,
|
| 723 |
+
"learning_rate": 9.154334038054969e-06,
|
| 724 |
+
"loss": 0.3221,
|
| 725 |
+
"step": 101
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.317016317016317,
|
| 729 |
+
"grad_norm": 1.3406106233596802,
|
| 730 |
+
"learning_rate": 9.143763213530656e-06,
|
| 731 |
+
"loss": 0.3209,
|
| 732 |
+
"step": 102
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.3201243201243201,
|
| 736 |
+
"grad_norm": 1.1402978897094727,
|
| 737 |
+
"learning_rate": 9.133192389006343e-06,
|
| 738 |
+
"loss": 0.3181,
|
| 739 |
+
"step": 103
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.32323232323232326,
|
| 743 |
+
"grad_norm": 1.0274314880371094,
|
| 744 |
+
"learning_rate": 9.12262156448203e-06,
|
| 745 |
+
"loss": 0.3179,
|
| 746 |
+
"step": 104
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.32634032634032634,
|
| 750 |
+
"grad_norm": 1.0853135585784912,
|
| 751 |
+
"learning_rate": 9.112050739957717e-06,
|
| 752 |
+
"loss": 0.3068,
|
| 753 |
+
"step": 105
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.32944832944832947,
|
| 757 |
+
"grad_norm": 0.9549627900123596,
|
| 758 |
+
"learning_rate": 9.101479915433404e-06,
|
| 759 |
+
"loss": 0.3058,
|
| 760 |
+
"step": 106
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.33255633255633255,
|
| 764 |
+
"grad_norm": 0.9081363081932068,
|
| 765 |
+
"learning_rate": 9.090909090909091e-06,
|
| 766 |
+
"loss": 0.305,
|
| 767 |
+
"step": 107
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.3356643356643357,
|
| 771 |
+
"grad_norm": 1.083267092704773,
|
| 772 |
+
"learning_rate": 9.080338266384778e-06,
|
| 773 |
+
"loss": 0.3293,
|
| 774 |
+
"step": 108
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.33877233877233875,
|
| 778 |
+
"grad_norm": 0.9146764278411865,
|
| 779 |
+
"learning_rate": 9.069767441860465e-06,
|
| 780 |
+
"loss": 0.3308,
|
| 781 |
+
"step": 109
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.3418803418803419,
|
| 785 |
+
"grad_norm": 0.8309290409088135,
|
| 786 |
+
"learning_rate": 9.059196617336154e-06,
|
| 787 |
+
"loss": 0.3219,
|
| 788 |
+
"step": 110
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.34498834498834496,
|
| 792 |
+
"grad_norm": 0.7540556788444519,
|
| 793 |
+
"learning_rate": 9.048625792811841e-06,
|
| 794 |
+
"loss": 0.3165,
|
| 795 |
+
"step": 111
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.3480963480963481,
|
| 799 |
+
"grad_norm": 0.7756165862083435,
|
| 800 |
+
"learning_rate": 9.038054968287528e-06,
|
| 801 |
+
"loss": 0.3201,
|
| 802 |
+
"step": 112
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 0.35120435120435123,
|
| 806 |
+
"grad_norm": 1.016161561012268,
|
| 807 |
+
"learning_rate": 9.027484143763215e-06,
|
| 808 |
+
"loss": 0.318,
|
| 809 |
+
"step": 113
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.3543123543123543,
|
| 813 |
+
"grad_norm": 1.1762275695800781,
|
| 814 |
+
"learning_rate": 9.016913319238902e-06,
|
| 815 |
+
"loss": 0.3071,
|
| 816 |
+
"step": 114
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.35742035742035744,
|
| 820 |
+
"grad_norm": 1.0186941623687744,
|
| 821 |
+
"learning_rate": 9.006342494714589e-06,
|
| 822 |
+
"loss": 0.3094,
|
| 823 |
+
"step": 115
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.3605283605283605,
|
| 827 |
+
"grad_norm": 1.3835426568984985,
|
| 828 |
+
"learning_rate": 8.995771670190276e-06,
|
| 829 |
+
"loss": 0.3203,
|
| 830 |
+
"step": 116
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.36363636363636365,
|
| 834 |
+
"grad_norm": 0.9151639938354492,
|
| 835 |
+
"learning_rate": 8.985200845665963e-06,
|
| 836 |
+
"loss": 0.3075,
|
| 837 |
+
"step": 117
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.3667443667443667,
|
| 841 |
+
"grad_norm": 0.9079708456993103,
|
| 842 |
+
"learning_rate": 8.974630021141648e-06,
|
| 843 |
+
"loss": 0.3111,
|
| 844 |
+
"step": 118
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.36985236985236986,
|
| 848 |
+
"grad_norm": 0.7135366201400757,
|
| 849 |
+
"learning_rate": 8.964059196617337e-06,
|
| 850 |
+
"loss": 0.3131,
|
| 851 |
+
"step": 119
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 0.372960372960373,
|
| 855 |
+
"grad_norm": 0.7310993671417236,
|
| 856 |
+
"learning_rate": 8.953488372093024e-06,
|
| 857 |
+
"loss": 0.3181,
|
| 858 |
+
"step": 120
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.37606837606837606,
|
| 862 |
+
"grad_norm": 0.9562262296676636,
|
| 863 |
+
"learning_rate": 8.942917547568711e-06,
|
| 864 |
+
"loss": 0.3114,
|
| 865 |
+
"step": 121
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.3791763791763792,
|
| 869 |
+
"grad_norm": 1.088692545890808,
|
| 870 |
+
"learning_rate": 8.932346723044398e-06,
|
| 871 |
+
"loss": 0.2985,
|
| 872 |
+
"step": 122
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.3822843822843823,
|
| 876 |
+
"grad_norm": 1.3334287405014038,
|
| 877 |
+
"learning_rate": 8.921775898520085e-06,
|
| 878 |
+
"loss": 0.3198,
|
| 879 |
+
"step": 123
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.3853923853923854,
|
| 883 |
+
"grad_norm": 1.1457082033157349,
|
| 884 |
+
"learning_rate": 8.911205073995772e-06,
|
| 885 |
+
"loss": 0.3027,
|
| 886 |
+
"step": 124
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.3885003885003885,
|
| 890 |
+
"grad_norm": 1.0944201946258545,
|
| 891 |
+
"learning_rate": 8.90063424947146e-06,
|
| 892 |
+
"loss": 0.3195,
|
| 893 |
+
"step": 125
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 0.3916083916083916,
|
| 897 |
+
"grad_norm": 1.679890513420105,
|
| 898 |
+
"learning_rate": 8.890063424947146e-06,
|
| 899 |
+
"loss": 0.3118,
|
| 900 |
+
"step": 126
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 0.3947163947163947,
|
| 904 |
+
"grad_norm": 1.0934737920761108,
|
| 905 |
+
"learning_rate": 8.879492600422833e-06,
|
| 906 |
+
"loss": 0.3125,
|
| 907 |
+
"step": 127
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 0.3978243978243978,
|
| 911 |
+
"grad_norm": 0.9423776865005493,
|
| 912 |
+
"learning_rate": 8.86892177589852e-06,
|
| 913 |
+
"loss": 0.3069,
|
| 914 |
+
"step": 128
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.3978243978243978,
|
| 918 |
+
"eval_loss": 0.33542340993881226,
|
| 919 |
+
"eval_runtime": 147.0915,
|
| 920 |
+
"eval_samples_per_second": 1.938,
|
| 921 |
+
"eval_steps_per_second": 0.972,
|
| 922 |
+
"step": 128
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 0.40093240093240096,
|
| 926 |
+
"grad_norm": 1.373064637184143,
|
| 927 |
+
"learning_rate": 8.858350951374208e-06,
|
| 928 |
+
"loss": 0.3113,
|
| 929 |
+
"step": 129
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.40404040404040403,
|
| 933 |
+
"grad_norm": 0.9782734513282776,
|
| 934 |
+
"learning_rate": 8.847780126849895e-06,
|
| 935 |
+
"loss": 0.3176,
|
| 936 |
+
"step": 130
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 0.40714840714840717,
|
| 940 |
+
"grad_norm": 1.1988129615783691,
|
| 941 |
+
"learning_rate": 8.837209302325582e-06,
|
| 942 |
+
"loss": 0.3036,
|
| 943 |
+
"step": 131
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 0.41025641025641024,
|
| 947 |
+
"grad_norm": 1.3978164196014404,
|
| 948 |
+
"learning_rate": 8.826638477801269e-06,
|
| 949 |
+
"loss": 0.3067,
|
| 950 |
+
"step": 132
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.4133644133644134,
|
| 954 |
+
"grad_norm": 0.8266012072563171,
|
| 955 |
+
"learning_rate": 8.816067653276956e-06,
|
| 956 |
+
"loss": 0.3105,
|
| 957 |
+
"step": 133
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 0.41647241647241645,
|
| 961 |
+
"grad_norm": 1.0358003377914429,
|
| 962 |
+
"learning_rate": 8.805496828752643e-06,
|
| 963 |
+
"loss": 0.3176,
|
| 964 |
+
"step": 134
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.4195804195804196,
|
| 968 |
+
"grad_norm": 0.9363102316856384,
|
| 969 |
+
"learning_rate": 8.79492600422833e-06,
|
| 970 |
+
"loss": 0.3151,
|
| 971 |
+
"step": 135
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.4226884226884227,
|
| 975 |
+
"grad_norm": 0.9805242419242859,
|
| 976 |
+
"learning_rate": 8.784355179704017e-06,
|
| 977 |
+
"loss": 0.3164,
|
| 978 |
+
"step": 136
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.4257964257964258,
|
| 982 |
+
"grad_norm": 1.4923985004425049,
|
| 983 |
+
"learning_rate": 8.773784355179706e-06,
|
| 984 |
+
"loss": 0.3059,
|
| 985 |
+
"step": 137
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.4289044289044289,
|
| 989 |
+
"grad_norm": 1.7009886503219604,
|
| 990 |
+
"learning_rate": 8.763213530655393e-06,
|
| 991 |
+
"loss": 0.2937,
|
| 992 |
+
"step": 138
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.432012432012432,
|
| 996 |
+
"grad_norm": 0.8320425748825073,
|
| 997 |
+
"learning_rate": 8.75264270613108e-06,
|
| 998 |
+
"loss": 0.288,
|
| 999 |
+
"step": 139
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.43512043512043513,
|
| 1003 |
+
"grad_norm": 1.3431979417800903,
|
| 1004 |
+
"learning_rate": 8.742071881606767e-06,
|
| 1005 |
+
"loss": 0.3063,
|
| 1006 |
+
"step": 140
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.4382284382284382,
|
| 1010 |
+
"grad_norm": 1.0519447326660156,
|
| 1011 |
+
"learning_rate": 8.731501057082454e-06,
|
| 1012 |
+
"loss": 0.3043,
|
| 1013 |
+
"step": 141
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.44133644133644134,
|
| 1017 |
+
"grad_norm": 1.0041645765304565,
|
| 1018 |
+
"learning_rate": 8.72093023255814e-06,
|
| 1019 |
+
"loss": 0.3207,
|
| 1020 |
+
"step": 142
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 0.4444444444444444,
|
| 1024 |
+
"grad_norm": 1.176352620124817,
|
| 1025 |
+
"learning_rate": 8.710359408033828e-06,
|
| 1026 |
+
"loss": 0.3099,
|
| 1027 |
+
"step": 143
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 0.44755244755244755,
|
| 1031 |
+
"grad_norm": 0.8591434955596924,
|
| 1032 |
+
"learning_rate": 8.699788583509515e-06,
|
| 1033 |
+
"loss": 0.2913,
|
| 1034 |
+
"step": 144
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.4471858134155744,
|
| 1038 |
+
"grad_norm": 1.2351419925689697,
|
| 1039 |
+
"learning_rate": 8.689217758985202e-06,
|
| 1040 |
+
"loss": 0.3099,
|
| 1041 |
+
"step": 145
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.4502698535080956,
|
| 1045 |
+
"grad_norm": 1.8375589847564697,
|
| 1046 |
+
"learning_rate": 8.691099476439791e-06,
|
| 1047 |
+
"loss": 0.3092,
|
| 1048 |
+
"step": 146
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 0.4533538936006168,
|
| 1052 |
+
"grad_norm": 1.07125985622406,
|
| 1053 |
+
"learning_rate": 8.680628272251308e-06,
|
| 1054 |
+
"loss": 0.3016,
|
| 1055 |
+
"step": 147
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 0.456437933693138,
|
| 1059 |
+
"grad_norm": 1.1839478015899658,
|
| 1060 |
+
"learning_rate": 8.670157068062827e-06,
|
| 1061 |
+
"loss": 0.3003,
|
| 1062 |
+
"step": 148
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 0.45952197378565923,
|
| 1066 |
+
"grad_norm": 1.294833779335022,
|
| 1067 |
+
"learning_rate": 8.659685863874346e-06,
|
| 1068 |
+
"loss": 0.2972,
|
| 1069 |
+
"step": 149
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 0.4626060138781804,
|
| 1073 |
+
"grad_norm": 1.0540661811828613,
|
| 1074 |
+
"learning_rate": 8.649214659685865e-06,
|
| 1075 |
+
"loss": 0.2837,
|
| 1076 |
+
"step": 150
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.4656900539707016,
|
| 1080 |
+
"grad_norm": 1.1067568063735962,
|
| 1081 |
+
"learning_rate": 8.638743455497383e-06,
|
| 1082 |
+
"loss": 0.2966,
|
| 1083 |
+
"step": 151
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.46877409406322285,
|
| 1087 |
+
"grad_norm": 0.9972389340400696,
|
| 1088 |
+
"learning_rate": 8.6282722513089e-06,
|
| 1089 |
+
"loss": 0.2934,
|
| 1090 |
+
"step": 152
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 0.471858134155744,
|
| 1094 |
+
"grad_norm": 1.1589370965957642,
|
| 1095 |
+
"learning_rate": 8.61780104712042e-06,
|
| 1096 |
+
"loss": 0.3026,
|
| 1097 |
+
"step": 153
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 0.47494217424826524,
|
| 1101 |
+
"grad_norm": 1.1224210262298584,
|
| 1102 |
+
"learning_rate": 8.607329842931938e-06,
|
| 1103 |
+
"loss": 0.3042,
|
| 1104 |
+
"step": 154
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 0.4780262143407864,
|
| 1108 |
+
"grad_norm": 1.3200238943099976,
|
| 1109 |
+
"learning_rate": 8.596858638743457e-06,
|
| 1110 |
+
"loss": 0.3124,
|
| 1111 |
+
"step": 155
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 0.4811102544333076,
|
| 1115 |
+
"grad_norm": 1.1300067901611328,
|
| 1116 |
+
"learning_rate": 8.586387434554974e-06,
|
| 1117 |
+
"loss": 0.3167,
|
| 1118 |
+
"step": 156
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.48419429452582885,
|
| 1122 |
+
"grad_norm": 0.9678866863250732,
|
| 1123 |
+
"learning_rate": 8.575916230366493e-06,
|
| 1124 |
+
"loss": 0.3039,
|
| 1125 |
+
"step": 157
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.48727833461835,
|
| 1129 |
+
"grad_norm": 0.9656190872192383,
|
| 1130 |
+
"learning_rate": 8.565445026178011e-06,
|
| 1131 |
+
"loss": 0.3067,
|
| 1132 |
+
"step": 158
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 0.49036237471087124,
|
| 1136 |
+
"grad_norm": 0.9618685245513916,
|
| 1137 |
+
"learning_rate": 8.55497382198953e-06,
|
| 1138 |
+
"loss": 0.2992,
|
| 1139 |
+
"step": 159
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 0.49344641480339246,
|
| 1143 |
+
"grad_norm": 1.1055867671966553,
|
| 1144 |
+
"learning_rate": 8.544502617801049e-06,
|
| 1145 |
+
"loss": 0.2986,
|
| 1146 |
+
"step": 160
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.49653045489591363,
|
| 1150 |
+
"grad_norm": 0.8761485815048218,
|
| 1151 |
+
"learning_rate": 8.534031413612566e-06,
|
| 1152 |
+
"loss": 0.3071,
|
| 1153 |
+
"step": 161
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 0.49961449498843485,
|
| 1157 |
+
"grad_norm": 1.0709651708602905,
|
| 1158 |
+
"learning_rate": 8.523560209424085e-06,
|
| 1159 |
+
"loss": 0.2965,
|
| 1160 |
+
"step": 162
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.5026985350809561,
|
| 1164 |
+
"grad_norm": 1.2407382726669312,
|
| 1165 |
+
"learning_rate": 8.513089005235604e-06,
|
| 1166 |
+
"loss": 0.3134,
|
| 1167 |
+
"step": 163
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.5057825751734772,
|
| 1171 |
+
"grad_norm": 1.46315598487854,
|
| 1172 |
+
"learning_rate": 8.502617801047122e-06,
|
| 1173 |
+
"loss": 0.2886,
|
| 1174 |
+
"step": 164
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 0.5088666152659984,
|
| 1178 |
+
"grad_norm": 1.2314726114273071,
|
| 1179 |
+
"learning_rate": 8.49214659685864e-06,
|
| 1180 |
+
"loss": 0.2902,
|
| 1181 |
+
"step": 165
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 0.5119506553585197,
|
| 1185 |
+
"grad_norm": 1.223716378211975,
|
| 1186 |
+
"learning_rate": 8.481675392670158e-06,
|
| 1187 |
+
"loss": 0.3088,
|
| 1188 |
+
"step": 166
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 0.5150346954510409,
|
| 1192 |
+
"grad_norm": 1.1966098546981812,
|
| 1193 |
+
"learning_rate": 8.471204188481677e-06,
|
| 1194 |
+
"loss": 0.3139,
|
| 1195 |
+
"step": 167
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 0.518118735543562,
|
| 1199 |
+
"grad_norm": 1.1182276010513306,
|
| 1200 |
+
"learning_rate": 8.460732984293194e-06,
|
| 1201 |
+
"loss": 0.3161,
|
| 1202 |
+
"step": 168
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.5212027756360833,
|
| 1206 |
+
"grad_norm": 1.1583510637283325,
|
| 1207 |
+
"learning_rate": 8.450261780104713e-06,
|
| 1208 |
+
"loss": 0.3041,
|
| 1209 |
+
"step": 169
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.5242868157286045,
|
| 1213 |
+
"grad_norm": 1.1864618062973022,
|
| 1214 |
+
"learning_rate": 8.439790575916232e-06,
|
| 1215 |
+
"loss": 0.3008,
|
| 1216 |
+
"step": 170
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 0.5273708558211256,
|
| 1220 |
+
"grad_norm": 1.3757935762405396,
|
| 1221 |
+
"learning_rate": 8.429319371727749e-06,
|
| 1222 |
+
"loss": 0.2865,
|
| 1223 |
+
"step": 171
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 0.5304548959136469,
|
| 1227 |
+
"grad_norm": 1.4410743713378906,
|
| 1228 |
+
"learning_rate": 8.418848167539267e-06,
|
| 1229 |
+
"loss": 0.3081,
|
| 1230 |
+
"step": 172
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 0.5335389360061681,
|
| 1234 |
+
"grad_norm": 1.3494313955307007,
|
| 1235 |
+
"learning_rate": 8.408376963350786e-06,
|
| 1236 |
+
"loss": 0.2988,
|
| 1237 |
+
"step": 173
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 0.5366229760986893,
|
| 1241 |
+
"grad_norm": 1.3871009349822998,
|
| 1242 |
+
"learning_rate": 8.397905759162305e-06,
|
| 1243 |
+
"loss": 0.3045,
|
| 1244 |
+
"step": 174
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.5397070161912105,
|
| 1248 |
+
"grad_norm": 1.183766484260559,
|
| 1249 |
+
"learning_rate": 8.387434554973822e-06,
|
| 1250 |
+
"loss": 0.2969,
|
| 1251 |
+
"step": 175
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.5427910562837317,
|
| 1255 |
+
"grad_norm": 1.1075443029403687,
|
| 1256 |
+
"learning_rate": 8.37696335078534e-06,
|
| 1257 |
+
"loss": 0.2834,
|
| 1258 |
+
"step": 176
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 0.5458750963762529,
|
| 1262 |
+
"grad_norm": 1.3118195533752441,
|
| 1263 |
+
"learning_rate": 8.36649214659686e-06,
|
| 1264 |
+
"loss": 0.2945,
|
| 1265 |
+
"step": 177
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 0.5489591364687741,
|
| 1269 |
+
"grad_norm": 1.3226675987243652,
|
| 1270 |
+
"learning_rate": 8.356020942408377e-06,
|
| 1271 |
+
"loss": 0.3085,
|
| 1272 |
+
"step": 178
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 0.5520431765612953,
|
| 1276 |
+
"grad_norm": 1.1877515316009521,
|
| 1277 |
+
"learning_rate": 8.345549738219895e-06,
|
| 1278 |
+
"loss": 0.2757,
|
| 1279 |
+
"step": 179
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 0.5551272166538165,
|
| 1283 |
+
"grad_norm": 1.379599928855896,
|
| 1284 |
+
"learning_rate": 8.335078534031414e-06,
|
| 1285 |
+
"loss": 0.2968,
|
| 1286 |
+
"step": 180
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 0.5582112567463376,
|
| 1290 |
+
"grad_norm": 1.2975775003433228,
|
| 1291 |
+
"learning_rate": 8.324607329842933e-06,
|
| 1292 |
+
"loss": 0.3074,
|
| 1293 |
+
"step": 181
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 0.5612952968388589,
|
| 1297 |
+
"grad_norm": 1.2829333543777466,
|
| 1298 |
+
"learning_rate": 8.31413612565445e-06,
|
| 1299 |
+
"loss": 0.3014,
|
| 1300 |
+
"step": 182
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 0.5643793369313801,
|
| 1304 |
+
"grad_norm": 1.4759114980697632,
|
| 1305 |
+
"learning_rate": 8.303664921465969e-06,
|
| 1306 |
+
"loss": 0.3014,
|
| 1307 |
+
"step": 183
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 0.5674633770239013,
|
| 1311 |
+
"grad_norm": 1.3108978271484375,
|
| 1312 |
+
"learning_rate": 8.293193717277488e-06,
|
| 1313 |
+
"loss": 0.2914,
|
| 1314 |
+
"step": 184
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.5705474171164225,
|
| 1318 |
+
"grad_norm": 1.271666407585144,
|
| 1319 |
+
"learning_rate": 8.282722513089005e-06,
|
| 1320 |
+
"loss": 0.305,
|
| 1321 |
+
"step": 185
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 0.5736314572089437,
|
| 1325 |
+
"grad_norm": 1.1115907430648804,
|
| 1326 |
+
"learning_rate": 8.272251308900523e-06,
|
| 1327 |
+
"loss": 0.2963,
|
| 1328 |
+
"step": 186
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.5767154973014649,
|
| 1332 |
+
"grad_norm": 1.089092493057251,
|
| 1333 |
+
"learning_rate": 8.261780104712042e-06,
|
| 1334 |
+
"loss": 0.303,
|
| 1335 |
+
"step": 187
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.5797995373939862,
|
| 1339 |
+
"grad_norm": 1.1514776945114136,
|
| 1340 |
+
"learning_rate": 8.251308900523561e-06,
|
| 1341 |
+
"loss": 0.3073,
|
| 1342 |
+
"step": 188
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 0.5828835774865073,
|
| 1346 |
+
"grad_norm": 1.1654891967773438,
|
| 1347 |
+
"learning_rate": 8.240837696335078e-06,
|
| 1348 |
+
"loss": 0.2883,
|
| 1349 |
+
"step": 189
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 0.5859676175790285,
|
| 1353 |
+
"grad_norm": 1.2040210962295532,
|
| 1354 |
+
"learning_rate": 8.230366492146597e-06,
|
| 1355 |
+
"loss": 0.295,
|
| 1356 |
+
"step": 190
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 0.5890516576715498,
|
| 1360 |
+
"grad_norm": 1.203511118888855,
|
| 1361 |
+
"learning_rate": 8.219895287958116e-06,
|
| 1362 |
+
"loss": 0.2795,
|
| 1363 |
+
"step": 191
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 0.5921356977640709,
|
| 1367 |
+
"grad_norm": 1.5743706226348877,
|
| 1368 |
+
"learning_rate": 8.209424083769634e-06,
|
| 1369 |
+
"loss": 0.3123,
|
| 1370 |
+
"step": 192
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.5921356977640709,
|
| 1374 |
+
"eval_loss": 0.3412991166114807,
|
| 1375 |
+
"eval_runtime": 149.387,
|
| 1376 |
+
"eval_samples_per_second": 1.928,
|
| 1377 |
+
"eval_steps_per_second": 0.964,
|
| 1378 |
+
"step": 192
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.5952197378565921,
|
| 1382 |
+
"grad_norm": 1.4109128713607788,
|
| 1383 |
+
"learning_rate": 8.198952879581153e-06,
|
| 1384 |
+
"loss": 0.2996,
|
| 1385 |
+
"step": 193
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.5983037779491134,
|
| 1389 |
+
"grad_norm": 1.3817074298858643,
|
| 1390 |
+
"learning_rate": 8.18848167539267e-06,
|
| 1391 |
+
"loss": 0.2964,
|
| 1392 |
+
"step": 194
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.6013878180416345,
|
| 1396 |
+
"grad_norm": 1.3587619066238403,
|
| 1397 |
+
"learning_rate": 8.178010471204189e-06,
|
| 1398 |
+
"loss": 0.3004,
|
| 1399 |
+
"step": 195
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.6044718581341557,
|
| 1403 |
+
"grad_norm": 1.502744197845459,
|
| 1404 |
+
"learning_rate": 8.167539267015708e-06,
|
| 1405 |
+
"loss": 0.2957,
|
| 1406 |
+
"step": 196
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.607555898226677,
|
| 1410 |
+
"grad_norm": 1.4416728019714355,
|
| 1411 |
+
"learning_rate": 8.157068062827227e-06,
|
| 1412 |
+
"loss": 0.2962,
|
| 1413 |
+
"step": 197
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.6106399383191982,
|
| 1417 |
+
"grad_norm": 2.2597157955169678,
|
| 1418 |
+
"learning_rate": 8.146596858638745e-06,
|
| 1419 |
+
"loss": 0.2853,
|
| 1420 |
+
"step": 198
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.6137239784117193,
|
| 1424 |
+
"grad_norm": 1.854837417602539,
|
| 1425 |
+
"learning_rate": 8.136125654450262e-06,
|
| 1426 |
+
"loss": 0.2918,
|
| 1427 |
+
"step": 199
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.6168080185042406,
|
| 1431 |
+
"grad_norm": 2.1409687995910645,
|
| 1432 |
+
"learning_rate": 8.125654450261781e-06,
|
| 1433 |
+
"loss": 0.3118,
|
| 1434 |
+
"step": 200
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.6198920585967618,
|
| 1438 |
+
"grad_norm": 1.7128517627716064,
|
| 1439 |
+
"learning_rate": 8.1151832460733e-06,
|
| 1440 |
+
"loss": 0.2822,
|
| 1441 |
+
"step": 201
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.6229760986892829,
|
| 1445 |
+
"grad_norm": 1.4401497840881348,
|
| 1446 |
+
"learning_rate": 8.104712041884819e-06,
|
| 1447 |
+
"loss": 0.2802,
|
| 1448 |
+
"step": 202
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 0.6260601387818041,
|
| 1452 |
+
"grad_norm": 1.7307312488555908,
|
| 1453 |
+
"learning_rate": 8.094240837696336e-06,
|
| 1454 |
+
"loss": 0.2973,
|
| 1455 |
+
"step": 203
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 0.6291441788743254,
|
| 1459 |
+
"grad_norm": 1.263535737991333,
|
| 1460 |
+
"learning_rate": 8.083769633507855e-06,
|
| 1461 |
+
"loss": 0.3016,
|
| 1462 |
+
"step": 204
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 0.6322282189668466,
|
| 1466 |
+
"grad_norm": 1.4065901041030884,
|
| 1467 |
+
"learning_rate": 8.073298429319373e-06,
|
| 1468 |
+
"loss": 0.284,
|
| 1469 |
+
"step": 205
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 0.6353122590593677,
|
| 1473 |
+
"grad_norm": 1.6004809141159058,
|
| 1474 |
+
"learning_rate": 8.06282722513089e-06,
|
| 1475 |
+
"loss": 0.2908,
|
| 1476 |
+
"step": 206
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 0.638396299151889,
|
| 1480 |
+
"grad_norm": 1.458287239074707,
|
| 1481 |
+
"learning_rate": 8.05235602094241e-06,
|
| 1482 |
+
"loss": 0.2832,
|
| 1483 |
+
"step": 207
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 0.6414803392444102,
|
| 1487 |
+
"grad_norm": 1.8239188194274902,
|
| 1488 |
+
"learning_rate": 8.041884816753928e-06,
|
| 1489 |
+
"loss": 0.2993,
|
| 1490 |
+
"step": 208
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 0.6445643793369313,
|
| 1494 |
+
"grad_norm": 1.8187966346740723,
|
| 1495 |
+
"learning_rate": 8.031413612565445e-06,
|
| 1496 |
+
"loss": 0.311,
|
| 1497 |
+
"step": 209
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 0.6476484194294526,
|
| 1501 |
+
"grad_norm": 1.5089385509490967,
|
| 1502 |
+
"learning_rate": 8.020942408376964e-06,
|
| 1503 |
+
"loss": 0.2835,
|
| 1504 |
+
"step": 210
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 0.6507324595219738,
|
| 1508 |
+
"grad_norm": 1.5591213703155518,
|
| 1509 |
+
"learning_rate": 8.010471204188483e-06,
|
| 1510 |
+
"loss": 0.2985,
|
| 1511 |
+
"step": 211
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 0.653816499614495,
|
| 1515 |
+
"grad_norm": 1.5221312046051025,
|
| 1516 |
+
"learning_rate": 8.000000000000001e-06,
|
| 1517 |
+
"loss": 0.2805,
|
| 1518 |
+
"step": 212
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 0.6569005397070162,
|
| 1522 |
+
"grad_norm": 1.8211005926132202,
|
| 1523 |
+
"learning_rate": 7.989528795811518e-06,
|
| 1524 |
+
"loss": 0.2728,
|
| 1525 |
+
"step": 213
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 0.6599845797995374,
|
| 1529 |
+
"grad_norm": 2.2500016689300537,
|
| 1530 |
+
"learning_rate": 7.979057591623037e-06,
|
| 1531 |
+
"loss": 0.2932,
|
| 1532 |
+
"step": 214
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 0.6630686198920586,
|
| 1536 |
+
"grad_norm": 1.7227460145950317,
|
| 1537 |
+
"learning_rate": 7.968586387434556e-06,
|
| 1538 |
+
"loss": 0.2927,
|
| 1539 |
+
"step": 215
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 0.6661526599845798,
|
| 1543 |
+
"grad_norm": 2.1821672916412354,
|
| 1544 |
+
"learning_rate": 7.958115183246073e-06,
|
| 1545 |
+
"loss": 0.2919,
|
| 1546 |
+
"step": 216
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 0.669236700077101,
|
| 1550 |
+
"grad_norm": 1.3368958234786987,
|
| 1551 |
+
"learning_rate": 7.947643979057592e-06,
|
| 1552 |
+
"loss": 0.2789,
|
| 1553 |
+
"step": 217
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 0.6723207401696222,
|
| 1557 |
+
"grad_norm": 1.4419403076171875,
|
| 1558 |
+
"learning_rate": 7.93717277486911e-06,
|
| 1559 |
+
"loss": 0.2876,
|
| 1560 |
+
"step": 218
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 0.6754047802621435,
|
| 1564 |
+
"grad_norm": 2.0355281829833984,
|
| 1565 |
+
"learning_rate": 7.92670157068063e-06,
|
| 1566 |
+
"loss": 0.3059,
|
| 1567 |
+
"step": 219
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 0.6784888203546646,
|
| 1571 |
+
"grad_norm": 1.7871628999710083,
|
| 1572 |
+
"learning_rate": 7.916230366492146e-06,
|
| 1573 |
+
"loss": 0.2804,
|
| 1574 |
+
"step": 220
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 0.6815728604471858,
|
| 1578 |
+
"grad_norm": 1.8160405158996582,
|
| 1579 |
+
"learning_rate": 7.905759162303665e-06,
|
| 1580 |
+
"loss": 0.2842,
|
| 1581 |
+
"step": 221
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 0.6846569005397071,
|
| 1585 |
+
"grad_norm": 2.1498160362243652,
|
| 1586 |
+
"learning_rate": 7.895287958115184e-06,
|
| 1587 |
+
"loss": 0.2875,
|
| 1588 |
+
"step": 222
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 0.6877409406322282,
|
| 1592 |
+
"grad_norm": 1.9483954906463623,
|
| 1593 |
+
"learning_rate": 7.884816753926701e-06,
|
| 1594 |
+
"loss": 0.2874,
|
| 1595 |
+
"step": 223
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 0.6908249807247494,
|
| 1599 |
+
"grad_norm": 2.0145816802978516,
|
| 1600 |
+
"learning_rate": 7.87434554973822e-06,
|
| 1601 |
+
"loss": 0.2879,
|
| 1602 |
+
"step": 224
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 0.6939090208172706,
|
| 1606 |
+
"grad_norm": 1.680413007736206,
|
| 1607 |
+
"learning_rate": 7.863874345549739e-06,
|
| 1608 |
+
"loss": 0.2755,
|
| 1609 |
+
"step": 225
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 0.6969930609097919,
|
| 1613 |
+
"grad_norm": 1.5203242301940918,
|
| 1614 |
+
"learning_rate": 7.853403141361257e-06,
|
| 1615 |
+
"loss": 0.284,
|
| 1616 |
+
"step": 226
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.700077101002313,
|
| 1620 |
+
"grad_norm": 1.892943263053894,
|
| 1621 |
+
"learning_rate": 7.842931937172774e-06,
|
| 1622 |
+
"loss": 0.2799,
|
| 1623 |
+
"step": 227
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 0.7031611410948342,
|
| 1627 |
+
"grad_norm": 1.5476278066635132,
|
| 1628 |
+
"learning_rate": 7.832460732984293e-06,
|
| 1629 |
+
"loss": 0.2767,
|
| 1630 |
+
"step": 228
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 0.7062451811873555,
|
| 1634 |
+
"grad_norm": 2.2650210857391357,
|
| 1635 |
+
"learning_rate": 7.821989528795812e-06,
|
| 1636 |
+
"loss": 0.2905,
|
| 1637 |
+
"step": 229
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 0.7093292212798766,
|
| 1641 |
+
"grad_norm": 2.1595096588134766,
|
| 1642 |
+
"learning_rate": 7.81151832460733e-06,
|
| 1643 |
+
"loss": 0.274,
|
| 1644 |
+
"step": 230
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 0.7124132613723978,
|
| 1648 |
+
"grad_norm": 1.587994933128357,
|
| 1649 |
+
"learning_rate": 7.80104712041885e-06,
|
| 1650 |
+
"loss": 0.2743,
|
| 1651 |
+
"step": 231
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 0.7154973014649191,
|
| 1655 |
+
"grad_norm": 1.9411978721618652,
|
| 1656 |
+
"learning_rate": 7.790575916230367e-06,
|
| 1657 |
+
"loss": 0.272,
|
| 1658 |
+
"step": 232
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 0.7185813415574402,
|
| 1662 |
+
"grad_norm": 2.1039252281188965,
|
| 1663 |
+
"learning_rate": 7.780104712041885e-06,
|
| 1664 |
+
"loss": 0.2884,
|
| 1665 |
+
"step": 233
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 0.7216653816499614,
|
| 1669 |
+
"grad_norm": 1.834591269493103,
|
| 1670 |
+
"learning_rate": 7.769633507853404e-06,
|
| 1671 |
+
"loss": 0.2756,
|
| 1672 |
+
"step": 234
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 0.7247494217424827,
|
| 1676 |
+
"grad_norm": 2.1758062839508057,
|
| 1677 |
+
"learning_rate": 7.759162303664923e-06,
|
| 1678 |
+
"loss": 0.287,
|
| 1679 |
+
"step": 235
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 0.7278334618350039,
|
| 1683 |
+
"grad_norm": 2.0601179599761963,
|
| 1684 |
+
"learning_rate": 7.748691099476442e-06,
|
| 1685 |
+
"loss": 0.2683,
|
| 1686 |
+
"step": 236
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 0.730917501927525,
|
| 1690 |
+
"grad_norm": 1.7605801820755005,
|
| 1691 |
+
"learning_rate": 7.738219895287959e-06,
|
| 1692 |
+
"loss": 0.2552,
|
| 1693 |
+
"step": 237
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 0.7340015420200463,
|
| 1697 |
+
"grad_norm": 2.0951759815216064,
|
| 1698 |
+
"learning_rate": 7.727748691099478e-06,
|
| 1699 |
+
"loss": 0.258,
|
| 1700 |
+
"step": 238
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 0.7370855821125675,
|
| 1704 |
+
"grad_norm": 2.2250118255615234,
|
| 1705 |
+
"learning_rate": 7.717277486910996e-06,
|
| 1706 |
+
"loss": 0.2627,
|
| 1707 |
+
"step": 239
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 0.7401696222050886,
|
| 1711 |
+
"grad_norm": 2.54436993598938,
|
| 1712 |
+
"learning_rate": 7.706806282722513e-06,
|
| 1713 |
+
"loss": 0.278,
|
| 1714 |
+
"step": 240
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 0.7432536622976099,
|
| 1718 |
+
"grad_norm": 1.810699701309204,
|
| 1719 |
+
"learning_rate": 7.696335078534032e-06,
|
| 1720 |
+
"loss": 0.2684,
|
| 1721 |
+
"step": 241
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 0.7463377023901311,
|
| 1725 |
+
"grad_norm": 2.161043882369995,
|
| 1726 |
+
"learning_rate": 7.685863874345551e-06,
|
| 1727 |
+
"loss": 0.2828,
|
| 1728 |
+
"step": 242
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 0.7494217424826523,
|
| 1732 |
+
"grad_norm": 1.7965888977050781,
|
| 1733 |
+
"learning_rate": 7.67539267015707e-06,
|
| 1734 |
+
"loss": 0.2677,
|
| 1735 |
+
"step": 243
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 0.7525057825751735,
|
| 1739 |
+
"grad_norm": 1.9139559268951416,
|
| 1740 |
+
"learning_rate": 7.664921465968587e-06,
|
| 1741 |
+
"loss": 0.2701,
|
| 1742 |
+
"step": 244
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 0.7555898226676947,
|
| 1746 |
+
"grad_norm": 2.0285589694976807,
|
| 1747 |
+
"learning_rate": 7.654450261780106e-06,
|
| 1748 |
+
"loss": 0.2726,
|
| 1749 |
+
"step": 245
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 0.7586738627602159,
|
| 1753 |
+
"grad_norm": 2.2968027591705322,
|
| 1754 |
+
"learning_rate": 7.643979057591624e-06,
|
| 1755 |
+
"loss": 0.2606,
|
| 1756 |
+
"step": 246
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 0.761757902852737,
|
| 1760 |
+
"grad_norm": 2.4324936866760254,
|
| 1761 |
+
"learning_rate": 7.633507853403141e-06,
|
| 1762 |
+
"loss": 0.2659,
|
| 1763 |
+
"step": 247
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 0.7648419429452583,
|
| 1767 |
+
"grad_norm": 2.66330885887146,
|
| 1768 |
+
"learning_rate": 7.62303664921466e-06,
|
| 1769 |
+
"loss": 0.2627,
|
| 1770 |
+
"step": 248
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 0.7679259830377795,
|
| 1774 |
+
"grad_norm": 2.435866355895996,
|
| 1775 |
+
"learning_rate": 7.612565445026179e-06,
|
| 1776 |
+
"loss": 0.2713,
|
| 1777 |
+
"step": 249
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 0.7710100231303006,
|
| 1781 |
+
"grad_norm": 2.2584385871887207,
|
| 1782 |
+
"learning_rate": 7.602094240837698e-06,
|
| 1783 |
+
"loss": 0.2754,
|
| 1784 |
+
"step": 250
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 0.7740940632228219,
|
| 1788 |
+
"grad_norm": 2.1898317337036133,
|
| 1789 |
+
"learning_rate": 7.591623036649215e-06,
|
| 1790 |
+
"loss": 0.2705,
|
| 1791 |
+
"step": 251
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 0.7771781033153431,
|
| 1795 |
+
"grad_norm": 2.051255464553833,
|
| 1796 |
+
"learning_rate": 7.5811518324607335e-06,
|
| 1797 |
+
"loss": 0.2491,
|
| 1798 |
+
"step": 252
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 0.7802621434078643,
|
| 1802 |
+
"grad_norm": 2.353940725326538,
|
| 1803 |
+
"learning_rate": 7.570680628272252e-06,
|
| 1804 |
+
"loss": 0.277,
|
| 1805 |
+
"step": 253
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 0.7833461835003855,
|
| 1809 |
+
"grad_norm": 2.3826687335968018,
|
| 1810 |
+
"learning_rate": 7.560209424083769e-06,
|
| 1811 |
+
"loss": 0.2693,
|
| 1812 |
+
"step": 254
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 0.7864302235929067,
|
| 1816 |
+
"grad_norm": 2.522019863128662,
|
| 1817 |
+
"learning_rate": 7.549738219895288e-06,
|
| 1818 |
+
"loss": 0.2706,
|
| 1819 |
+
"step": 255
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 0.7895142636854279,
|
| 1823 |
+
"grad_norm": 2.3525524139404297,
|
| 1824 |
+
"learning_rate": 7.539267015706807e-06,
|
| 1825 |
+
"loss": 0.2509,
|
| 1826 |
+
"step": 256
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 0.7895142636854279,
|
| 1830 |
+
"eval_loss": 0.3851300776004791,
|
| 1831 |
+
"eval_runtime": 149.046,
|
| 1832 |
+
"eval_samples_per_second": 1.932,
|
| 1833 |
+
"eval_steps_per_second": 0.966,
|
| 1834 |
+
"step": 256
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 0.7925983037779492,
|
| 1838 |
+
"grad_norm": 2.7143642902374268,
|
| 1839 |
+
"learning_rate": 7.528795811518326e-06,
|
| 1840 |
+
"loss": 0.2701,
|
| 1841 |
+
"step": 257
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 0.7956823438704703,
|
| 1845 |
+
"grad_norm": 2.6725356578826904,
|
| 1846 |
+
"learning_rate": 7.518324607329844e-06,
|
| 1847 |
+
"loss": 0.2718,
|
| 1848 |
+
"step": 258
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 0.7987663839629915,
|
| 1852 |
+
"grad_norm": 2.4051880836486816,
|
| 1853 |
+
"learning_rate": 7.5078534031413615e-06,
|
| 1854 |
+
"loss": 0.2554,
|
| 1855 |
+
"step": 259
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 0.8018504240555128,
|
| 1859 |
+
"grad_norm": 2.472904920578003,
|
| 1860 |
+
"learning_rate": 7.49738219895288e-06,
|
| 1861 |
+
"loss": 0.2666,
|
| 1862 |
+
"step": 260
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 0.8049344641480339,
|
| 1866 |
+
"grad_norm": 2.3598804473876953,
|
| 1867 |
+
"learning_rate": 7.486910994764398e-06,
|
| 1868 |
+
"loss": 0.2532,
|
| 1869 |
+
"step": 261
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 0.8080185042405551,
|
| 1873 |
+
"grad_norm": 2.383300542831421,
|
| 1874 |
+
"learning_rate": 7.476439790575917e-06,
|
| 1875 |
+
"loss": 0.2568,
|
| 1876 |
+
"step": 262
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 0.8111025443330764,
|
| 1880 |
+
"grad_norm": 2.999469518661499,
|
| 1881 |
+
"learning_rate": 7.465968586387436e-06,
|
| 1882 |
+
"loss": 0.2403,
|
| 1883 |
+
"step": 263
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 0.8141865844255975,
|
| 1887 |
+
"grad_norm": 4.071384429931641,
|
| 1888 |
+
"learning_rate": 7.455497382198954e-06,
|
| 1889 |
+
"loss": 0.265,
|
| 1890 |
+
"step": 264
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 0.8172706245181187,
|
| 1894 |
+
"grad_norm": 3.5529489517211914,
|
| 1895 |
+
"learning_rate": 7.445026178010472e-06,
|
| 1896 |
+
"loss": 0.2647,
|
| 1897 |
+
"step": 265
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 0.8203546646106399,
|
| 1901 |
+
"grad_norm": 2.8842644691467285,
|
| 1902 |
+
"learning_rate": 7.43455497382199e-06,
|
| 1903 |
+
"loss": 0.2725,
|
| 1904 |
+
"step": 266
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 0.8234387047031612,
|
| 1908 |
+
"grad_norm": 2.1277332305908203,
|
| 1909 |
+
"learning_rate": 7.424083769633509e-06,
|
| 1910 |
+
"loss": 0.2657,
|
| 1911 |
+
"step": 267
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 0.8265227447956823,
|
| 1915 |
+
"grad_norm": 2.832111358642578,
|
| 1916 |
+
"learning_rate": 7.413612565445026e-06,
|
| 1917 |
+
"loss": 0.255,
|
| 1918 |
+
"step": 268
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 0.8296067848882035,
|
| 1922 |
+
"grad_norm": 2.7438676357269287,
|
| 1923 |
+
"learning_rate": 7.403141361256545e-06,
|
| 1924 |
+
"loss": 0.2596,
|
| 1925 |
+
"step": 269
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 0.8326908249807248,
|
| 1929 |
+
"grad_norm": 2.7950987815856934,
|
| 1930 |
+
"learning_rate": 7.392670157068064e-06,
|
| 1931 |
+
"loss": 0.2624,
|
| 1932 |
+
"step": 270
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 0.8357748650732459,
|
| 1936 |
+
"grad_norm": 3.497069835662842,
|
| 1937 |
+
"learning_rate": 7.382198952879581e-06,
|
| 1938 |
+
"loss": 0.2385,
|
| 1939 |
+
"step": 271
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 0.8388589051657671,
|
| 1943 |
+
"grad_norm": 5.024068832397461,
|
| 1944 |
+
"learning_rate": 7.3717277486911e-06,
|
| 1945 |
+
"loss": 0.2526,
|
| 1946 |
+
"step": 272
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 0.8419429452582884,
|
| 1950 |
+
"grad_norm": 3.5298011302948,
|
| 1951 |
+
"learning_rate": 7.361256544502618e-06,
|
| 1952 |
+
"loss": 0.2452,
|
| 1953 |
+
"step": 273
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 0.8450269853508096,
|
| 1957 |
+
"grad_norm": 2.701545238494873,
|
| 1958 |
+
"learning_rate": 7.350785340314137e-06,
|
| 1959 |
+
"loss": 0.2293,
|
| 1960 |
+
"step": 274
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 0.8481110254433307,
|
| 1964 |
+
"grad_norm": 2.838541030883789,
|
| 1965 |
+
"learning_rate": 7.340314136125655e-06,
|
| 1966 |
+
"loss": 0.2554,
|
| 1967 |
+
"step": 275
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 0.851195065535852,
|
| 1971 |
+
"grad_norm": 2.5854012966156006,
|
| 1972 |
+
"learning_rate": 7.329842931937173e-06,
|
| 1973 |
+
"loss": 0.245,
|
| 1974 |
+
"step": 276
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 0.8542791056283732,
|
| 1978 |
+
"grad_norm": 2.9351906776428223,
|
| 1979 |
+
"learning_rate": 7.319371727748692e-06,
|
| 1980 |
+
"loss": 0.2556,
|
| 1981 |
+
"step": 277
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 0.8573631457208943,
|
| 1985 |
+
"grad_norm": 3.0675830841064453,
|
| 1986 |
+
"learning_rate": 7.30890052356021e-06,
|
| 1987 |
+
"loss": 0.2501,
|
| 1988 |
+
"step": 278
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 0.8604471858134156,
|
| 1992 |
+
"grad_norm": 3.1958088874816895,
|
| 1993 |
+
"learning_rate": 7.2984293193717285e-06,
|
| 1994 |
+
"loss": 0.2347,
|
| 1995 |
+
"step": 279
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.8635312259059368,
|
| 1999 |
+
"grad_norm": 3.0006463527679443,
|
| 2000 |
+
"learning_rate": 7.287958115183246e-06,
|
| 2001 |
+
"loss": 0.242,
|
| 2002 |
+
"step": 280
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 0.866615265998458,
|
| 2006 |
+
"grad_norm": 2.862990379333496,
|
| 2007 |
+
"learning_rate": 7.277486910994765e-06,
|
| 2008 |
+
"loss": 0.2442,
|
| 2009 |
+
"step": 281
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 0.8696993060909792,
|
| 2013 |
+
"grad_norm": 3.1585986614227295,
|
| 2014 |
+
"learning_rate": 7.267015706806283e-06,
|
| 2015 |
+
"loss": 0.2401,
|
| 2016 |
+
"step": 282
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 0.8727833461835004,
|
| 2020 |
+
"grad_norm": 2.6111812591552734,
|
| 2021 |
+
"learning_rate": 7.256544502617802e-06,
|
| 2022 |
+
"loss": 0.2324,
|
| 2023 |
+
"step": 283
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 0.8758673862760216,
|
| 2027 |
+
"grad_norm": 3.1289191246032715,
|
| 2028 |
+
"learning_rate": 7.246073298429321e-06,
|
| 2029 |
+
"loss": 0.2426,
|
| 2030 |
+
"step": 284
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 0.8789514263685428,
|
| 2034 |
+
"grad_norm": 3.448789358139038,
|
| 2035 |
+
"learning_rate": 7.235602094240838e-06,
|
| 2036 |
+
"loss": 0.2224,
|
| 2037 |
+
"step": 285
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 0.882035466461064,
|
| 2041 |
+
"grad_norm": 3.018432855606079,
|
| 2042 |
+
"learning_rate": 7.2251308900523565e-06,
|
| 2043 |
+
"loss": 0.2238,
|
| 2044 |
+
"step": 286
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 0.8851195065535852,
|
| 2048 |
+
"grad_norm": 4.171509742736816,
|
| 2049 |
+
"learning_rate": 7.214659685863875e-06,
|
| 2050 |
+
"loss": 0.2546,
|
| 2051 |
+
"step": 287
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 0.8882035466461063,
|
| 2055 |
+
"grad_norm": 3.5390446186065674,
|
| 2056 |
+
"learning_rate": 7.204188481675394e-06,
|
| 2057 |
+
"loss": 0.2417,
|
| 2058 |
+
"step": 288
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 0.8912875867386276,
|
| 2062 |
+
"grad_norm": 2.8169162273406982,
|
| 2063 |
+
"learning_rate": 7.193717277486911e-06,
|
| 2064 |
+
"loss": 0.2348,
|
| 2065 |
+
"step": 289
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 0.8943716268311488,
|
| 2069 |
+
"grad_norm": 2.9175827503204346,
|
| 2070 |
+
"learning_rate": 7.18324607329843e-06,
|
| 2071 |
+
"loss": 0.214,
|
| 2072 |
+
"step": 290
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 0.89745566692367,
|
| 2076 |
+
"grad_norm": 3.939680576324463,
|
| 2077 |
+
"learning_rate": 7.172774869109949e-06,
|
| 2078 |
+
"loss": 0.2489,
|
| 2079 |
+
"step": 291
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 0.9005397070161912,
|
| 2083 |
+
"grad_norm": 2.874373435974121,
|
| 2084 |
+
"learning_rate": 7.162303664921466e-06,
|
| 2085 |
+
"loss": 0.2219,
|
| 2086 |
+
"step": 292
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 0.9036237471087124,
|
| 2090 |
+
"grad_norm": 4.381021976470947,
|
| 2091 |
+
"learning_rate": 7.1518324607329845e-06,
|
| 2092 |
+
"loss": 0.2419,
|
| 2093 |
+
"step": 293
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 0.9067077872012336,
|
| 2097 |
+
"grad_norm": 3.9895918369293213,
|
| 2098 |
+
"learning_rate": 7.141361256544503e-06,
|
| 2099 |
+
"loss": 0.2552,
|
| 2100 |
+
"step": 294
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 0.9097918272937549,
|
| 2104 |
+
"grad_norm": 2.9028842449188232,
|
| 2105 |
+
"learning_rate": 7.130890052356022e-06,
|
| 2106 |
+
"loss": 0.2323,
|
| 2107 |
+
"step": 295
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 0.912875867386276,
|
| 2111 |
+
"grad_norm": 3.5980117321014404,
|
| 2112 |
+
"learning_rate": 7.12041884816754e-06,
|
| 2113 |
+
"loss": 0.2404,
|
| 2114 |
+
"step": 296
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 0.9159599074787972,
|
| 2118 |
+
"grad_norm": 3.490727186203003,
|
| 2119 |
+
"learning_rate": 7.109947643979058e-06,
|
| 2120 |
+
"loss": 0.22,
|
| 2121 |
+
"step": 297
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 0.9190439475713185,
|
| 2125 |
+
"grad_norm": 3.256279706954956,
|
| 2126 |
+
"learning_rate": 7.099476439790577e-06,
|
| 2127 |
+
"loss": 0.2368,
|
| 2128 |
+
"step": 298
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 0.9221279876638396,
|
| 2132 |
+
"grad_norm": 3.92038893699646,
|
| 2133 |
+
"learning_rate": 7.089005235602095e-06,
|
| 2134 |
+
"loss": 0.2331,
|
| 2135 |
+
"step": 299
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 0.9252120277563608,
|
| 2139 |
+
"grad_norm": 3.6917364597320557,
|
| 2140 |
+
"learning_rate": 7.078534031413613e-06,
|
| 2141 |
+
"loss": 0.2139,
|
| 2142 |
+
"step": 300
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 0.9282960678488821,
|
| 2146 |
+
"grad_norm": 3.058729887008667,
|
| 2147 |
+
"learning_rate": 7.068062827225132e-06,
|
| 2148 |
+
"loss": 0.2199,
|
| 2149 |
+
"step": 301
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 0.9313801079414032,
|
| 2153 |
+
"grad_norm": 3.150188446044922,
|
| 2154 |
+
"learning_rate": 7.057591623036649e-06,
|
| 2155 |
+
"loss": 0.2137,
|
| 2156 |
+
"step": 302
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 0.9344641480339244,
|
| 2160 |
+
"grad_norm": 5.77610445022583,
|
| 2161 |
+
"learning_rate": 7.047120418848168e-06,
|
| 2162 |
+
"loss": 0.2478,
|
| 2163 |
+
"step": 303
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 0.9375481881264457,
|
| 2167 |
+
"grad_norm": 2.8851089477539062,
|
| 2168 |
+
"learning_rate": 7.036649214659687e-06,
|
| 2169 |
+
"loss": 0.227,
|
| 2170 |
+
"step": 304
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 0.9406322282189669,
|
| 2174 |
+
"grad_norm": 3.1656086444854736,
|
| 2175 |
+
"learning_rate": 7.0261780104712055e-06,
|
| 2176 |
+
"loss": 0.2335,
|
| 2177 |
+
"step": 305
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 0.943716268311488,
|
| 2181 |
+
"grad_norm": 3.3355696201324463,
|
| 2182 |
+
"learning_rate": 7.015706806282723e-06,
|
| 2183 |
+
"loss": 0.2169,
|
| 2184 |
+
"step": 306
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 0.9468003084040093,
|
| 2188 |
+
"grad_norm": 3.5095317363739014,
|
| 2189 |
+
"learning_rate": 7.005235602094241e-06,
|
| 2190 |
+
"loss": 0.2161,
|
| 2191 |
+
"step": 307
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 0.9498843484965305,
|
| 2195 |
+
"grad_norm": 3.5365262031555176,
|
| 2196 |
+
"learning_rate": 6.99476439790576e-06,
|
| 2197 |
+
"loss": 0.2097,
|
| 2198 |
+
"step": 308
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 0.9529683885890516,
|
| 2202 |
+
"grad_norm": 4.159248352050781,
|
| 2203 |
+
"learning_rate": 6.984293193717277e-06,
|
| 2204 |
+
"loss": 0.2337,
|
| 2205 |
+
"step": 309
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 0.9560524286815728,
|
| 2209 |
+
"grad_norm": 2.9792213439941406,
|
| 2210 |
+
"learning_rate": 6.973821989528796e-06,
|
| 2211 |
+
"loss": 0.2149,
|
| 2212 |
+
"step": 310
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 0.9591364687740941,
|
| 2216 |
+
"grad_norm": 3.2603046894073486,
|
| 2217 |
+
"learning_rate": 6.963350785340315e-06,
|
| 2218 |
+
"loss": 0.2218,
|
| 2219 |
+
"step": 311
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 0.9622205088666153,
|
| 2223 |
+
"grad_norm": 3.5064327716827393,
|
| 2224 |
+
"learning_rate": 6.9528795811518335e-06,
|
| 2225 |
+
"loss": 0.2128,
|
| 2226 |
+
"step": 312
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 0.9653045489591364,
|
| 2230 |
+
"grad_norm": 3.971139430999756,
|
| 2231 |
+
"learning_rate": 6.942408376963351e-06,
|
| 2232 |
+
"loss": 0.2172,
|
| 2233 |
+
"step": 313
|
| 2234 |
+
},
|
| 2235 |
+
{
|
| 2236 |
+
"epoch": 0.9683885890516577,
|
| 2237 |
+
"grad_norm": 3.651603937149048,
|
| 2238 |
+
"learning_rate": 6.931937172774869e-06,
|
| 2239 |
+
"loss": 0.2036,
|
| 2240 |
+
"step": 314
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 0.9714726291441789,
|
| 2244 |
+
"grad_norm": 5.394900321960449,
|
| 2245 |
+
"learning_rate": 6.921465968586388e-06,
|
| 2246 |
+
"loss": 0.2157,
|
| 2247 |
+
"step": 315
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 0.9745566692367,
|
| 2251 |
+
"grad_norm": 3.7696452140808105,
|
| 2252 |
+
"learning_rate": 6.910994764397906e-06,
|
| 2253 |
+
"loss": 0.2168,
|
| 2254 |
+
"step": 316
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 0.9776407093292213,
|
| 2258 |
+
"grad_norm": 3.3137505054473877,
|
| 2259 |
+
"learning_rate": 6.900523560209425e-06,
|
| 2260 |
+
"loss": 0.2217,
|
| 2261 |
+
"step": 317
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 0.9807247494217425,
|
| 2265 |
+
"grad_norm": 3.927021026611328,
|
| 2266 |
+
"learning_rate": 6.890052356020943e-06,
|
| 2267 |
+
"loss": 0.2149,
|
| 2268 |
+
"step": 318
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 0.9838087895142636,
|
| 2272 |
+
"grad_norm": 3.598501443862915,
|
| 2273 |
+
"learning_rate": 6.8795811518324615e-06,
|
| 2274 |
+
"loss": 0.2007,
|
| 2275 |
+
"step": 319
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 0.9868928296067849,
|
| 2279 |
+
"grad_norm": 4.063229084014893,
|
| 2280 |
+
"learning_rate": 6.8691099476439794e-06,
|
| 2281 |
+
"loss": 0.2142,
|
| 2282 |
+
"step": 320
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 0.9868928296067849,
|
| 2286 |
+
"eval_loss": 0.46243318915367126,
|
| 2287 |
+
"eval_runtime": 150.4594,
|
| 2288 |
+
"eval_samples_per_second": 1.914,
|
| 2289 |
+
"eval_steps_per_second": 0.957,
|
| 2290 |
+
"step": 320
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 0.9899768696993061,
|
| 2294 |
+
"grad_norm": 4.520982265472412,
|
| 2295 |
+
"learning_rate": 6.858638743455498e-06,
|
| 2296 |
+
"loss": 0.1978,
|
| 2297 |
+
"step": 321
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 0.9930609097918273,
|
| 2301 |
+
"grad_norm": 3.6312687397003174,
|
| 2302 |
+
"learning_rate": 6.848167539267017e-06,
|
| 2303 |
+
"loss": 0.1896,
|
| 2304 |
+
"step": 322
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 0.9961449498843485,
|
| 2308 |
+
"grad_norm": 3.1252243518829346,
|
| 2309 |
+
"learning_rate": 6.837696335078534e-06,
|
| 2310 |
+
"loss": 0.1817,
|
| 2311 |
+
"step": 323
|
| 2312 |
+
},
|
| 2313 |
+
{
|
| 2314 |
+
"epoch": 0.9992289899768697,
|
| 2315 |
+
"grad_norm": 4.3829264640808105,
|
| 2316 |
+
"learning_rate": 6.827225130890053e-06,
|
| 2317 |
+
"loss": 0.2199,
|
| 2318 |
+
"step": 324
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 1.0030840400925212,
|
| 2322 |
+
"grad_norm": 9.755841255187988,
|
| 2323 |
+
"learning_rate": 6.816753926701572e-06,
|
| 2324 |
+
"loss": 0.4578,
|
| 2325 |
+
"step": 325
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 1.0061680801850423,
|
| 2329 |
+
"grad_norm": 3.9052581787109375,
|
| 2330 |
+
"learning_rate": 6.80628272251309e-06,
|
| 2331 |
+
"loss": 0.1959,
|
| 2332 |
+
"step": 326
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 1.0092521202775635,
|
| 2336 |
+
"grad_norm": 3.6258931159973145,
|
| 2337 |
+
"learning_rate": 6.7958115183246075e-06,
|
| 2338 |
+
"loss": 0.2062,
|
| 2339 |
+
"step": 327
|
| 2340 |
+
},
|
| 2341 |
+
{
|
| 2342 |
+
"epoch": 1.012336160370085,
|
| 2343 |
+
"grad_norm": 4.131122589111328,
|
| 2344 |
+
"learning_rate": 6.785340314136126e-06,
|
| 2345 |
+
"loss": 0.1915,
|
| 2346 |
+
"step": 328
|
| 2347 |
+
},
|
| 2348 |
+
{
|
| 2349 |
+
"epoch": 1.015420200462606,
|
| 2350 |
+
"grad_norm": 4.387429237365723,
|
| 2351 |
+
"learning_rate": 6.774869109947645e-06,
|
| 2352 |
+
"loss": 0.1792,
|
| 2353 |
+
"step": 329
|
| 2354 |
+
},
|
| 2355 |
+
{
|
| 2356 |
+
"epoch": 1.0185042405551272,
|
| 2357 |
+
"grad_norm": 3.873361110687256,
|
| 2358 |
+
"learning_rate": 6.764397905759162e-06,
|
| 2359 |
+
"loss": 0.1895,
|
| 2360 |
+
"step": 330
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 1.0215882806476484,
|
| 2364 |
+
"grad_norm": 4.318599700927734,
|
| 2365 |
+
"learning_rate": 6.753926701570681e-06,
|
| 2366 |
+
"loss": 0.1836,
|
| 2367 |
+
"step": 331
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 1.0246723207401696,
|
| 2371 |
+
"grad_norm": 4.9434494972229,
|
| 2372 |
+
"learning_rate": 6.7434554973822e-06,
|
| 2373 |
+
"loss": 0.2199,
|
| 2374 |
+
"step": 332
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 1.0277563608326907,
|
| 2378 |
+
"grad_norm": 3.8584797382354736,
|
| 2379 |
+
"learning_rate": 6.732984293193718e-06,
|
| 2380 |
+
"loss": 0.1796,
|
| 2381 |
+
"step": 333
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 1.0308404009252121,
|
| 2385 |
+
"grad_norm": 4.104945659637451,
|
| 2386 |
+
"learning_rate": 6.722513089005236e-06,
|
| 2387 |
+
"loss": 0.1812,
|
| 2388 |
+
"step": 334
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"epoch": 1.0339244410177333,
|
| 2392 |
+
"grad_norm": 4.125020503997803,
|
| 2393 |
+
"learning_rate": 6.712041884816754e-06,
|
| 2394 |
+
"loss": 0.197,
|
| 2395 |
+
"step": 335
|
| 2396 |
+
},
|
| 2397 |
+
{
|
| 2398 |
+
"epoch": 1.0370084811102545,
|
| 2399 |
+
"grad_norm": 3.783364772796631,
|
| 2400 |
+
"learning_rate": 6.701570680628273e-06,
|
| 2401 |
+
"loss": 0.1798,
|
| 2402 |
+
"step": 336
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 1.0400925212027756,
|
| 2406 |
+
"grad_norm": 4.799828052520752,
|
| 2407 |
+
"learning_rate": 6.691099476439791e-06,
|
| 2408 |
+
"loss": 0.1837,
|
| 2409 |
+
"step": 337
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 1.0431765612952968,
|
| 2413 |
+
"grad_norm": 5.570056438446045,
|
| 2414 |
+
"learning_rate": 6.68062827225131e-06,
|
| 2415 |
+
"loss": 0.1987,
|
| 2416 |
+
"step": 338
|
| 2417 |
+
},
|
| 2418 |
+
{
|
| 2419 |
+
"epoch": 1.046260601387818,
|
| 2420 |
+
"grad_norm": 3.9299843311309814,
|
| 2421 |
+
"learning_rate": 6.670157068062828e-06,
|
| 2422 |
+
"loss": 0.1728,
|
| 2423 |
+
"step": 339
|
| 2424 |
+
},
|
| 2425 |
+
{
|
| 2426 |
+
"epoch": 1.0493446414803393,
|
| 2427 |
+
"grad_norm": 4.746124267578125,
|
| 2428 |
+
"learning_rate": 6.6596858638743455e-06,
|
| 2429 |
+
"loss": 0.2055,
|
| 2430 |
+
"step": 340
|
| 2431 |
+
},
|
| 2432 |
+
{
|
| 2433 |
+
"epoch": 1.0524286815728605,
|
| 2434 |
+
"grad_norm": 3.6969268321990967,
|
| 2435 |
+
"learning_rate": 6.649214659685864e-06,
|
| 2436 |
+
"loss": 0.1919,
|
| 2437 |
+
"step": 341
|
| 2438 |
+
},
|
| 2439 |
+
{
|
| 2440 |
+
"epoch": 1.0555127216653817,
|
| 2441 |
+
"grad_norm": 4.096460819244385,
|
| 2442 |
+
"learning_rate": 6.638743455497383e-06,
|
| 2443 |
+
"loss": 0.1725,
|
| 2444 |
+
"step": 342
|
| 2445 |
+
},
|
| 2446 |
+
{
|
| 2447 |
+
"epoch": 1.0585967617579028,
|
| 2448 |
+
"grad_norm": 3.819343328475952,
|
| 2449 |
+
"learning_rate": 6.628272251308902e-06,
|
| 2450 |
+
"loss": 0.1727,
|
| 2451 |
+
"step": 343
|
| 2452 |
+
},
|
| 2453 |
+
{
|
| 2454 |
+
"epoch": 1.061680801850424,
|
| 2455 |
+
"grad_norm": 4.487940788269043,
|
| 2456 |
+
"learning_rate": 6.617801047120419e-06,
|
| 2457 |
+
"loss": 0.176,
|
| 2458 |
+
"step": 344
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 1.0647648419429452,
|
| 2462 |
+
"grad_norm": 4.727810382843018,
|
| 2463 |
+
"learning_rate": 6.607329842931938e-06,
|
| 2464 |
+
"loss": 0.1694,
|
| 2465 |
+
"step": 345
|
| 2466 |
+
},
|
| 2467 |
+
{
|
| 2468 |
+
"epoch": 1.0678488820354666,
|
| 2469 |
+
"grad_norm": 5.403895854949951,
|
| 2470 |
+
"learning_rate": 6.5968586387434565e-06,
|
| 2471 |
+
"loss": 0.1853,
|
| 2472 |
+
"step": 346
|
| 2473 |
+
},
|
| 2474 |
+
{
|
| 2475 |
+
"epoch": 1.0709329221279877,
|
| 2476 |
+
"grad_norm": 3.548576831817627,
|
| 2477 |
+
"learning_rate": 6.5863874345549736e-06,
|
| 2478 |
+
"loss": 0.1711,
|
| 2479 |
+
"step": 347
|
| 2480 |
+
},
|
| 2481 |
+
{
|
| 2482 |
+
"epoch": 1.074016962220509,
|
| 2483 |
+
"grad_norm": 3.6849658489227295,
|
| 2484 |
+
"learning_rate": 6.575916230366492e-06,
|
| 2485 |
+
"loss": 0.1877,
|
| 2486 |
+
"step": 348
|
| 2487 |
+
},
|
| 2488 |
+
{
|
| 2489 |
+
"epoch": 1.07710100231303,
|
| 2490 |
+
"grad_norm": 3.7493557929992676,
|
| 2491 |
+
"learning_rate": 6.565445026178011e-06,
|
| 2492 |
+
"loss": 0.1858,
|
| 2493 |
+
"step": 349
|
| 2494 |
+
},
|
| 2495 |
+
{
|
| 2496 |
+
"epoch": 1.0801850424055512,
|
| 2497 |
+
"grad_norm": 3.9486773014068604,
|
| 2498 |
+
"learning_rate": 6.55497382198953e-06,
|
| 2499 |
+
"loss": 0.1515,
|
| 2500 |
+
"step": 350
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 1.0832690824980724,
|
| 2504 |
+
"grad_norm": 4.970436096191406,
|
| 2505 |
+
"learning_rate": 6.544502617801047e-06,
|
| 2506 |
+
"loss": 0.172,
|
| 2507 |
+
"step": 351
|
| 2508 |
+
},
|
| 2509 |
+
{
|
| 2510 |
+
"epoch": 1.0863531225905936,
|
| 2511 |
+
"grad_norm": 5.032225131988525,
|
| 2512 |
+
"learning_rate": 6.534031413612566e-06,
|
| 2513 |
+
"loss": 0.1611,
|
| 2514 |
+
"step": 352
|
| 2515 |
+
}
|
| 2516 |
+
],
|
| 2517 |
+
"logging_steps": 1,
|
| 2518 |
+
"max_steps": 975,
|
| 2519 |
+
"num_input_tokens_seen": 0,
|
| 2520 |
+
"num_train_epochs": 3,
|
| 2521 |
+
"save_steps": 16,
|
| 2522 |
+
"stateful_callbacks": {
|
| 2523 |
+
"TrainerControl": {
|
| 2524 |
+
"args": {
|
| 2525 |
+
"should_epoch_stop": false,
|
| 2526 |
+
"should_evaluate": false,
|
| 2527 |
+
"should_log": false,
|
| 2528 |
+
"should_save": true,
|
| 2529 |
+
"should_training_stop": false
|
| 2530 |
+
},
|
| 2531 |
+
"attributes": {}
|
| 2532 |
+
}
|
| 2533 |
+
},
|
| 2534 |
+
"total_flos": 8.458954269238886e+16,
|
| 2535 |
+
"train_batch_size": 2,
|
| 2536 |
+
"trial_name": null,
|
| 2537 |
+
"trial_params": null
|
| 2538 |
+
}
|