Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2282 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2282 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.98753894080997,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3206,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03115264797507788,
|
| 13 |
+
"grad_norm": 8.509510040283203,
|
| 14 |
+
"learning_rate": 1.2422360248447205e-05,
|
| 15 |
+
"loss": 1.026,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06230529595015576,
|
| 20 |
+
"grad_norm": 2.1322240829467773,
|
| 21 |
+
"learning_rate": 2.484472049689441e-05,
|
| 22 |
+
"loss": 0.5916,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09345794392523364,
|
| 27 |
+
"grad_norm": 2.834507703781128,
|
| 28 |
+
"learning_rate": 3.7267080745341614e-05,
|
| 29 |
+
"loss": 0.355,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12461059190031153,
|
| 34 |
+
"grad_norm": 1.4687763452529907,
|
| 35 |
+
"learning_rate": 4.968944099378882e-05,
|
| 36 |
+
"loss": 0.2673,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1557632398753894,
|
| 41 |
+
"grad_norm": 1.5616137981414795,
|
| 42 |
+
"learning_rate": 6.211180124223603e-05,
|
| 43 |
+
"loss": 0.217,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.18691588785046728,
|
| 48 |
+
"grad_norm": 4.9718337059021,
|
| 49 |
+
"learning_rate": 7.453416149068323e-05,
|
| 50 |
+
"loss": 0.2141,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.21806853582554517,
|
| 55 |
+
"grad_norm": 1.4241009950637817,
|
| 56 |
+
"learning_rate": 8.695652173913044e-05,
|
| 57 |
+
"loss": 0.162,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.24922118380062305,
|
| 62 |
+
"grad_norm": 1.6728754043579102,
|
| 63 |
+
"learning_rate": 9.937888198757764e-05,
|
| 64 |
+
"loss": 0.176,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2803738317757009,
|
| 69 |
+
"grad_norm": 0.9034008979797363,
|
| 70 |
+
"learning_rate": 0.00011180124223602484,
|
| 71 |
+
"loss": 0.1567,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3115264797507788,
|
| 76 |
+
"grad_norm": 0.889045000076294,
|
| 77 |
+
"learning_rate": 0.00012422360248447205,
|
| 78 |
+
"loss": 0.1535,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3426791277258567,
|
| 83 |
+
"grad_norm": 0.6448578238487244,
|
| 84 |
+
"learning_rate": 0.00013664596273291925,
|
| 85 |
+
"loss": 0.1454,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.37383177570093457,
|
| 90 |
+
"grad_norm": 1.2123034000396729,
|
| 91 |
+
"learning_rate": 0.00014906832298136645,
|
| 92 |
+
"loss": 0.1437,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.40498442367601245,
|
| 97 |
+
"grad_norm": 1.03734290599823,
|
| 98 |
+
"learning_rate": 0.00016149068322981365,
|
| 99 |
+
"loss": 0.152,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.43613707165109034,
|
| 104 |
+
"grad_norm": 1.0240880250930786,
|
| 105 |
+
"learning_rate": 0.00017391304347826088,
|
| 106 |
+
"loss": 0.1338,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.4672897196261682,
|
| 111 |
+
"grad_norm": 0.7429300546646118,
|
| 112 |
+
"learning_rate": 0.00018633540372670808,
|
| 113 |
+
"loss": 0.1283,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.4984423676012461,
|
| 118 |
+
"grad_norm": 0.7910900712013245,
|
| 119 |
+
"learning_rate": 0.00019875776397515528,
|
| 120 |
+
"loss": 0.1291,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5295950155763239,
|
| 125 |
+
"grad_norm": 0.8077343702316284,
|
| 126 |
+
"learning_rate": 0.00019999568900961486,
|
| 127 |
+
"loss": 0.1167,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5607476635514018,
|
| 132 |
+
"grad_norm": 0.8578943610191345,
|
| 133 |
+
"learning_rate": 0.00019998078729782597,
|
| 134 |
+
"loss": 0.1174,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5919003115264797,
|
| 139 |
+
"grad_norm": 0.5542319416999817,
|
| 140 |
+
"learning_rate": 0.00019995524322835034,
|
| 141 |
+
"loss": 0.1069,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6230529595015576,
|
| 146 |
+
"grad_norm": 0.603608250617981,
|
| 147 |
+
"learning_rate": 0.00019991905952020163,
|
| 148 |
+
"loss": 0.1129,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6542056074766355,
|
| 153 |
+
"grad_norm": 1.1858749389648438,
|
| 154 |
+
"learning_rate": 0.0001998722400249194,
|
| 155 |
+
"loss": 0.1246,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6853582554517134,
|
| 160 |
+
"grad_norm": 0.8343398571014404,
|
| 161 |
+
"learning_rate": 0.00019981478972615934,
|
| 162 |
+
"loss": 0.1286,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.7165109034267912,
|
| 167 |
+
"grad_norm": 0.5227356553077698,
|
| 168 |
+
"learning_rate": 0.00019974671473916268,
|
| 169 |
+
"loss": 0.1024,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7476635514018691,
|
| 174 |
+
"grad_norm": 0.5733448266983032,
|
| 175 |
+
"learning_rate": 0.0001996680223101052,
|
| 176 |
+
"loss": 0.0994,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.778816199376947,
|
| 181 |
+
"grad_norm": 0.5022748708724976,
|
| 182 |
+
"learning_rate": 0.00019957872081532594,
|
| 183 |
+
"loss": 0.0958,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.8099688473520249,
|
| 188 |
+
"grad_norm": 0.49469369649887085,
|
| 189 |
+
"learning_rate": 0.00019947881976043572,
|
| 190 |
+
"loss": 0.1028,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8411214953271028,
|
| 195 |
+
"grad_norm": 0.7817510366439819,
|
| 196 |
+
"learning_rate": 0.0001993683297793052,
|
| 197 |
+
"loss": 0.1065,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.8722741433021807,
|
| 202 |
+
"grad_norm": 0.6700159907341003,
|
| 203 |
+
"learning_rate": 0.00019924726263293295,
|
| 204 |
+
"loss": 0.1015,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9034267912772586,
|
| 209 |
+
"grad_norm": 0.6034234762191772,
|
| 210 |
+
"learning_rate": 0.00019911563120819368,
|
| 211 |
+
"loss": 0.0979,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9345794392523364,
|
| 216 |
+
"grad_norm": 0.6600286364555359,
|
| 217 |
+
"learning_rate": 0.00019897344951646636,
|
| 218 |
+
"loss": 0.1003,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.9657320872274143,
|
| 223 |
+
"grad_norm": 0.7439318895339966,
|
| 224 |
+
"learning_rate": 0.00019882073269214296,
|
| 225 |
+
"loss": 0.1024,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9968847352024922,
|
| 230 |
+
"grad_norm": 0.6639961004257202,
|
| 231 |
+
"learning_rate": 0.00019865749699101736,
|
| 232 |
+
"loss": 0.0914,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.02803738317757,
|
| 237 |
+
"grad_norm": 0.7482274770736694,
|
| 238 |
+
"learning_rate": 0.000198483759788555,
|
| 239 |
+
"loss": 0.1031,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.0591900311526479,
|
| 244 |
+
"grad_norm": 0.8928232192993164,
|
| 245 |
+
"learning_rate": 0.00019829953957804347,
|
| 246 |
+
"loss": 0.0874,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.0903426791277258,
|
| 251 |
+
"grad_norm": 0.48007968068122864,
|
| 252 |
+
"learning_rate": 0.00019810485596862392,
|
| 253 |
+
"loss": 0.0901,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.1214953271028036,
|
| 258 |
+
"grad_norm": 0.5031644701957703,
|
| 259 |
+
"learning_rate": 0.0001978997296832039,
|
| 260 |
+
"loss": 0.0907,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.1526479750778815,
|
| 265 |
+
"grad_norm": 0.6962934136390686,
|
| 266 |
+
"learning_rate": 0.00019768418255625142,
|
| 267 |
+
"loss": 0.0976,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.1838006230529594,
|
| 272 |
+
"grad_norm": 0.42880135774612427,
|
| 273 |
+
"learning_rate": 0.0001974582375314708,
|
| 274 |
+
"loss": 0.0875,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.2149532710280373,
|
| 279 |
+
"grad_norm": 0.6273675560951233,
|
| 280 |
+
"learning_rate": 0.00019722191865936047,
|
| 281 |
+
"loss": 0.09,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2461059190031152,
|
| 286 |
+
"grad_norm": 0.8934553861618042,
|
| 287 |
+
"learning_rate": 0.00019697525109465302,
|
| 288 |
+
"loss": 0.098,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.277258566978193,
|
| 293 |
+
"grad_norm": 0.6337535381317139,
|
| 294 |
+
"learning_rate": 0.00019671826109363754,
|
| 295 |
+
"loss": 0.0865,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.308411214953271,
|
| 300 |
+
"grad_norm": 0.5011693239212036,
|
| 301 |
+
"learning_rate": 0.00019645097601136475,
|
| 302 |
+
"loss": 0.0838,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.3395638629283488,
|
| 307 |
+
"grad_norm": 0.3215925395488739,
|
| 308 |
+
"learning_rate": 0.00019617342429873537,
|
| 309 |
+
"loss": 0.0793,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.3707165109034267,
|
| 314 |
+
"grad_norm": 0.6264034509658813,
|
| 315 |
+
"learning_rate": 0.00019588563549947157,
|
| 316 |
+
"loss": 0.0784,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.4018691588785046,
|
| 321 |
+
"grad_norm": 0.5204505324363708,
|
| 322 |
+
"learning_rate": 0.00019558764024697224,
|
| 323 |
+
"loss": 0.0899,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.4330218068535825,
|
| 328 |
+
"grad_norm": 0.7140909433364868,
|
| 329 |
+
"learning_rate": 0.0001952794702610523,
|
| 330 |
+
"loss": 0.0921,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.4641744548286604,
|
| 335 |
+
"grad_norm": 0.6163343191146851,
|
| 336 |
+
"learning_rate": 0.0001949611583445662,
|
| 337 |
+
"loss": 0.0868,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.4953271028037383,
|
| 342 |
+
"grad_norm": 0.28120145201683044,
|
| 343 |
+
"learning_rate": 0.00019463273837991643,
|
| 344 |
+
"loss": 0.0831,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.5264797507788161,
|
| 349 |
+
"grad_norm": 0.4508073925971985,
|
| 350 |
+
"learning_rate": 0.00019429424532544677,
|
| 351 |
+
"loss": 0.0752,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.557632398753894,
|
| 356 |
+
"grad_norm": 0.7511669993400574,
|
| 357 |
+
"learning_rate": 0.00019394571521172126,
|
| 358 |
+
"loss": 0.0818,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.588785046728972,
|
| 363 |
+
"grad_norm": 0.5071356296539307,
|
| 364 |
+
"learning_rate": 0.00019358718513768895,
|
| 365 |
+
"loss": 0.0844,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.6199376947040498,
|
| 370 |
+
"grad_norm": 0.6059209108352661,
|
| 371 |
+
"learning_rate": 0.00019321869326673495,
|
| 372 |
+
"loss": 0.0758,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.6510903426791277,
|
| 377 |
+
"grad_norm": 0.6657635569572449,
|
| 378 |
+
"learning_rate": 0.00019284027882261818,
|
| 379 |
+
"loss": 0.0809,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6822429906542056,
|
| 384 |
+
"grad_norm": 0.6091364026069641,
|
| 385 |
+
"learning_rate": 0.00019245198208529617,
|
| 386 |
+
"loss": 0.0745,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.7133956386292835,
|
| 391 |
+
"grad_norm": 0.42025190591812134,
|
| 392 |
+
"learning_rate": 0.0001920538443866376,
|
| 393 |
+
"loss": 0.0727,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.7445482866043613,
|
| 398 |
+
"grad_norm": 0.5204473733901978,
|
| 399 |
+
"learning_rate": 0.00019164590810602262,
|
| 400 |
+
"loss": 0.0751,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.7757009345794392,
|
| 405 |
+
"grad_norm": 0.44874081015586853,
|
| 406 |
+
"learning_rate": 0.00019122821666583197,
|
| 407 |
+
"loss": 0.073,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.8068535825545171,
|
| 412 |
+
"grad_norm": 0.5214203596115112,
|
| 413 |
+
"learning_rate": 0.00019080081452682487,
|
| 414 |
+
"loss": 0.0835,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.838006230529595,
|
| 419 |
+
"grad_norm": 0.6505956053733826,
|
| 420 |
+
"learning_rate": 0.0001903637471834064,
|
| 421 |
+
"loss": 0.0815,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.8691588785046729,
|
| 426 |
+
"grad_norm": 0.5872865319252014,
|
| 427 |
+
"learning_rate": 0.000189917061158785,
|
| 428 |
+
"loss": 0.0803,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.9003115264797508,
|
| 433 |
+
"grad_norm": 0.4705474376678467,
|
| 434 |
+
"learning_rate": 0.00018946080400002014,
|
| 435 |
+
"loss": 0.0845,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.9314641744548287,
|
| 440 |
+
"grad_norm": 0.4995359182357788,
|
| 441 |
+
"learning_rate": 0.0001889950242729615,
|
| 442 |
+
"loss": 0.0768,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.9626168224299065,
|
| 447 |
+
"grad_norm": 0.4790456295013428,
|
| 448 |
+
"learning_rate": 0.00018851977155707916,
|
| 449 |
+
"loss": 0.0685,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.9937694704049844,
|
| 454 |
+
"grad_norm": 0.3062126934528351,
|
| 455 |
+
"learning_rate": 0.00018803509644018643,
|
| 456 |
+
"loss": 0.065,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.0249221183800623,
|
| 461 |
+
"grad_norm": 0.4944644272327423,
|
| 462 |
+
"learning_rate": 0.00018754105051305478,
|
| 463 |
+
"loss": 0.0731,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.05607476635514,
|
| 468 |
+
"grad_norm": 0.49407851696014404,
|
| 469 |
+
"learning_rate": 0.00018703768636392257,
|
| 470 |
+
"loss": 0.0606,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.087227414330218,
|
| 475 |
+
"grad_norm": 0.8019661903381348,
|
| 476 |
+
"learning_rate": 0.00018652505757289718,
|
| 477 |
+
"loss": 0.0895,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.1183800623052957,
|
| 482 |
+
"grad_norm": 0.7003992795944214,
|
| 483 |
+
"learning_rate": 0.00018600321870625183,
|
| 484 |
+
"loss": 0.0746,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.149532710280374,
|
| 489 |
+
"grad_norm": 0.3881121575832367,
|
| 490 |
+
"learning_rate": 0.00018547222531061728,
|
| 491 |
+
"loss": 0.0713,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.1806853582554515,
|
| 496 |
+
"grad_norm": 0.3871879279613495,
|
| 497 |
+
"learning_rate": 0.00018493213390706917,
|
| 498 |
+
"loss": 0.0722,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.2118380062305296,
|
| 503 |
+
"grad_norm": 0.5889574289321899,
|
| 504 |
+
"learning_rate": 0.00018438300198511182,
|
| 505 |
+
"loss": 0.0797,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.2429906542056073,
|
| 510 |
+
"grad_norm": 0.6468931436538696,
|
| 511 |
+
"learning_rate": 0.0001838248879965587,
|
| 512 |
+
"loss": 0.0713,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.2741433021806854,
|
| 517 |
+
"grad_norm": 0.3565777540206909,
|
| 518 |
+
"learning_rate": 0.00018325785134931064,
|
| 519 |
+
"loss": 0.0732,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.305295950155763,
|
| 524 |
+
"grad_norm": 0.4860513210296631,
|
| 525 |
+
"learning_rate": 0.0001826819524010322,
|
| 526 |
+
"loss": 0.0659,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.336448598130841,
|
| 531 |
+
"grad_norm": 0.6295431852340698,
|
| 532 |
+
"learning_rate": 0.0001820972524527269,
|
| 533 |
+
"loss": 0.0739,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.367601246105919,
|
| 538 |
+
"grad_norm": 0.6680755019187927,
|
| 539 |
+
"learning_rate": 0.00018150381374221227,
|
| 540 |
+
"loss": 0.0667,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.398753894080997,
|
| 545 |
+
"grad_norm": 0.5046746730804443,
|
| 546 |
+
"learning_rate": 0.00018090169943749476,
|
| 547 |
+
"loss": 0.0676,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.4299065420560746,
|
| 552 |
+
"grad_norm": 0.4515908658504486,
|
| 553 |
+
"learning_rate": 0.00018029097363004608,
|
| 554 |
+
"loss": 0.0689,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.4610591900311527,
|
| 559 |
+
"grad_norm": 0.44286802411079407,
|
| 560 |
+
"learning_rate": 0.000179671701327981,
|
| 561 |
+
"loss": 0.0652,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.4922118380062304,
|
| 566 |
+
"grad_norm": 0.5898622274398804,
|
| 567 |
+
"learning_rate": 0.0001790439484491375,
|
| 568 |
+
"loss": 0.0684,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.5233644859813085,
|
| 573 |
+
"grad_norm": 0.5441446900367737,
|
| 574 |
+
"learning_rate": 0.0001784077818140604,
|
| 575 |
+
"loss": 0.069,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.554517133956386,
|
| 580 |
+
"grad_norm": 0.533498227596283,
|
| 581 |
+
"learning_rate": 0.00017776326913888858,
|
| 582 |
+
"loss": 0.0724,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.5856697819314642,
|
| 587 |
+
"grad_norm": 0.46721819043159485,
|
| 588 |
+
"learning_rate": 0.00017711047902814705,
|
| 589 |
+
"loss": 0.0668,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.616822429906542,
|
| 594 |
+
"grad_norm": 0.6141489148139954,
|
| 595 |
+
"learning_rate": 0.0001764494809674444,
|
| 596 |
+
"loss": 0.0588,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.64797507788162,
|
| 601 |
+
"grad_norm": 0.3712906837463379,
|
| 602 |
+
"learning_rate": 0.0001757803453160765,
|
| 603 |
+
"loss": 0.0786,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.6791277258566977,
|
| 608 |
+
"grad_norm": 0.4539516270160675,
|
| 609 |
+
"learning_rate": 0.0001751031432995371,
|
| 610 |
+
"loss": 0.0683,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.710280373831776,
|
| 615 |
+
"grad_norm": 0.40119048953056335,
|
| 616 |
+
"learning_rate": 0.00017441794700193635,
|
| 617 |
+
"loss": 0.0701,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.7414330218068534,
|
| 622 |
+
"grad_norm": 0.3867497742176056,
|
| 623 |
+
"learning_rate": 0.0001737248293583279,
|
| 624 |
+
"loss": 0.0607,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.7725856697819315,
|
| 629 |
+
"grad_norm": 0.5221051573753357,
|
| 630 |
+
"learning_rate": 0.0001730238641469454,
|
| 631 |
+
"loss": 0.0646,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.803738317757009,
|
| 636 |
+
"grad_norm": 0.5209854245185852,
|
| 637 |
+
"learning_rate": 0.00017231512598134914,
|
| 638 |
+
"loss": 0.0675,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.8348909657320873,
|
| 643 |
+
"grad_norm": 0.5283793807029724,
|
| 644 |
+
"learning_rate": 0.00017159869030248403,
|
| 645 |
+
"loss": 0.058,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.866043613707165,
|
| 650 |
+
"grad_norm": 0.45289307832717896,
|
| 651 |
+
"learning_rate": 0.00017087463337064925,
|
| 652 |
+
"loss": 0.0581,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.897196261682243,
|
| 657 |
+
"grad_norm": 0.4225894510746002,
|
| 658 |
+
"learning_rate": 0.00017014303225738092,
|
| 659 |
+
"loss": 0.0633,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.9283489096573208,
|
| 664 |
+
"grad_norm": 0.4690423905849457,
|
| 665 |
+
"learning_rate": 0.00016940396483724812,
|
| 666 |
+
"loss": 0.0532,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.959501557632399,
|
| 671 |
+
"grad_norm": 0.42076757550239563,
|
| 672 |
+
"learning_rate": 0.00016865750977956379,
|
| 673 |
+
"loss": 0.067,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.9906542056074765,
|
| 678 |
+
"grad_norm": 0.3666526973247528,
|
| 679 |
+
"learning_rate": 0.00016790374654001067,
|
| 680 |
+
"loss": 0.0611,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.0218068535825546,
|
| 685 |
+
"grad_norm": 0.5197750926017761,
|
| 686 |
+
"learning_rate": 0.00016714275535218386,
|
| 687 |
+
"loss": 0.056,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.0529595015576323,
|
| 692 |
+
"grad_norm": 0.9091117978096008,
|
| 693 |
+
"learning_rate": 0.00016637461721905045,
|
| 694 |
+
"loss": 0.0618,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.0841121495327104,
|
| 699 |
+
"grad_norm": 0.3250091075897217,
|
| 700 |
+
"learning_rate": 0.0001655994139043271,
|
| 701 |
+
"loss": 0.0689,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.115264797507788,
|
| 706 |
+
"grad_norm": 0.6207090020179749,
|
| 707 |
+
"learning_rate": 0.00016481722792377682,
|
| 708 |
+
"loss": 0.0587,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.146417445482866,
|
| 713 |
+
"grad_norm": 0.33198997378349304,
|
| 714 |
+
"learning_rate": 0.0001640281425364258,
|
| 715 |
+
"loss": 0.0589,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.177570093457944,
|
| 720 |
+
"grad_norm": 0.3611784279346466,
|
| 721 |
+
"learning_rate": 0.0001632322417357008,
|
| 722 |
+
"loss": 0.0564,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.208722741433022,
|
| 727 |
+
"grad_norm": 0.3399527668952942,
|
| 728 |
+
"learning_rate": 0.00016242961024048866,
|
| 729 |
+
"loss": 0.0591,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.2398753894080996,
|
| 734 |
+
"grad_norm": 0.49678900837898254,
|
| 735 |
+
"learning_rate": 0.00016162033348611845,
|
| 736 |
+
"loss": 0.0603,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.2710280373831777,
|
| 741 |
+
"grad_norm": 0.8306248784065247,
|
| 742 |
+
"learning_rate": 0.00016080449761526734,
|
| 743 |
+
"loss": 0.0601,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.3021806853582554,
|
| 748 |
+
"grad_norm": 0.7733712792396545,
|
| 749 |
+
"learning_rate": 0.00015998218946879138,
|
| 750 |
+
"loss": 0.0723,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.3333333333333335,
|
| 755 |
+
"grad_norm": 0.3001610040664673,
|
| 756 |
+
"learning_rate": 0.0001591534965764816,
|
| 757 |
+
"loss": 0.0676,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.364485981308411,
|
| 762 |
+
"grad_norm": 0.3723997473716736,
|
| 763 |
+
"learning_rate": 0.00015831850714774715,
|
| 764 |
+
"loss": 0.0597,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.3956386292834893,
|
| 769 |
+
"grad_norm": 0.3429635167121887,
|
| 770 |
+
"learning_rate": 0.00015747731006222586,
|
| 771 |
+
"loss": 0.0559,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.426791277258567,
|
| 776 |
+
"grad_norm": 0.3895944356918335,
|
| 777 |
+
"learning_rate": 0.00015662999486032352,
|
| 778 |
+
"loss": 0.0589,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.457943925233645,
|
| 783 |
+
"grad_norm": 0.45078912377357483,
|
| 784 |
+
"learning_rate": 0.00015577665173368294,
|
| 785 |
+
"loss": 0.0617,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.4890965732087227,
|
| 790 |
+
"grad_norm": 0.3323187828063965,
|
| 791 |
+
"learning_rate": 0.0001549173715155835,
|
| 792 |
+
"loss": 0.0517,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.520249221183801,
|
| 797 |
+
"grad_norm": 0.3069084584712982,
|
| 798 |
+
"learning_rate": 0.00015405224567127248,
|
| 799 |
+
"loss": 0.0657,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.5514018691588785,
|
| 804 |
+
"grad_norm": 0.34960678219795227,
|
| 805 |
+
"learning_rate": 0.0001531813662882291,
|
| 806 |
+
"loss": 0.06,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.582554517133956,
|
| 811 |
+
"grad_norm": 0.38408011198043823,
|
| 812 |
+
"learning_rate": 0.00015230482606636253,
|
| 813 |
+
"loss": 0.054,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.6137071651090342,
|
| 818 |
+
"grad_norm": 0.664941668510437,
|
| 819 |
+
"learning_rate": 0.00015142271830814437,
|
| 820 |
+
"loss": 0.057,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.6448598130841123,
|
| 825 |
+
"grad_norm": 0.3202420175075531,
|
| 826 |
+
"learning_rate": 0.00015053513690867725,
|
| 827 |
+
"loss": 0.0501,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.67601246105919,
|
| 832 |
+
"grad_norm": 0.6673224568367004,
|
| 833 |
+
"learning_rate": 0.00014964217634570016,
|
| 834 |
+
"loss": 0.0528,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.7071651090342677,
|
| 839 |
+
"grad_norm": 0.46046340465545654,
|
| 840 |
+
"learning_rate": 0.00014874393166953192,
|
| 841 |
+
"loss": 0.0592,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.7383177570093458,
|
| 846 |
+
"grad_norm": 0.4655970335006714,
|
| 847 |
+
"learning_rate": 0.00014784049849295375,
|
| 848 |
+
"loss": 0.06,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.769470404984424,
|
| 853 |
+
"grad_norm": 0.41359665989875793,
|
| 854 |
+
"learning_rate": 0.00014693197298103154,
|
| 855 |
+
"loss": 0.0571,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.8006230529595015,
|
| 860 |
+
"grad_norm": 0.2868637442588806,
|
| 861 |
+
"learning_rate": 0.00014601845184088012,
|
| 862 |
+
"loss": 0.0506,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.831775700934579,
|
| 867 |
+
"grad_norm": 0.7829365730285645,
|
| 868 |
+
"learning_rate": 0.00014510003231136892,
|
| 869 |
+
"loss": 0.0621,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.8629283489096573,
|
| 874 |
+
"grad_norm": 0.39203941822052,
|
| 875 |
+
"learning_rate": 0.00014417681215277175,
|
| 876 |
+
"loss": 0.0552,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.8940809968847354,
|
| 881 |
+
"grad_norm": 0.38012075424194336,
|
| 882 |
+
"learning_rate": 0.00014324888963636068,
|
| 883 |
+
"loss": 0.0613,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.925233644859813,
|
| 888 |
+
"grad_norm": 0.2863033413887024,
|
| 889 |
+
"learning_rate": 0.00014231636353394567,
|
| 890 |
+
"loss": 0.0525,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.9563862928348907,
|
| 895 |
+
"grad_norm": 0.5563439130783081,
|
| 896 |
+
"learning_rate": 0.000141379333107361,
|
| 897 |
+
"loss": 0.055,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.987538940809969,
|
| 902 |
+
"grad_norm": 0.44631409645080566,
|
| 903 |
+
"learning_rate": 0.00014043789809789926,
|
| 904 |
+
"loss": 0.0551,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.018691588785047,
|
| 909 |
+
"grad_norm": 0.4797975420951843,
|
| 910 |
+
"learning_rate": 0.0001394921587156946,
|
| 911 |
+
"loss": 0.0584,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.049844236760125,
|
| 916 |
+
"grad_norm": 0.261513352394104,
|
| 917 |
+
"learning_rate": 0.000138542215629056,
|
| 918 |
+
"loss": 0.0561,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.080996884735202,
|
| 923 |
+
"grad_norm": 0.3844500184059143,
|
| 924 |
+
"learning_rate": 0.00013758816995375163,
|
| 925 |
+
"loss": 0.0492,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.11214953271028,
|
| 930 |
+
"grad_norm": 0.30683764815330505,
|
| 931 |
+
"learning_rate": 0.0001366301232422457,
|
| 932 |
+
"loss": 0.0557,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.1433021806853585,
|
| 937 |
+
"grad_norm": 0.3304888606071472,
|
| 938 |
+
"learning_rate": 0.00013566817747288887,
|
| 939 |
+
"loss": 0.0468,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.174454828660436,
|
| 944 |
+
"grad_norm": 0.2605965733528137,
|
| 945 |
+
"learning_rate": 0.0001347024350390632,
|
| 946 |
+
"loss": 0.0496,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.205607476635514,
|
| 951 |
+
"grad_norm": 0.595988929271698,
|
| 952 |
+
"learning_rate": 0.00013373299873828303,
|
| 953 |
+
"loss": 0.0495,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.2367601246105915,
|
| 958 |
+
"grad_norm": 0.4891095459461212,
|
| 959 |
+
"learning_rate": 0.0001327599717612527,
|
| 960 |
+
"loss": 0.0517,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.26791277258567,
|
| 965 |
+
"grad_norm": 0.6615864634513855,
|
| 966 |
+
"learning_rate": 0.0001317834576808827,
|
| 967 |
+
"loss": 0.0465,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.299065420560748,
|
| 972 |
+
"grad_norm": 0.2254357486963272,
|
| 973 |
+
"learning_rate": 0.00013080356044126474,
|
| 974 |
+
"loss": 0.0494,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.330218068535825,
|
| 979 |
+
"grad_norm": 0.25203853845596313,
|
| 980 |
+
"learning_rate": 0.00012982038434660774,
|
| 981 |
+
"loss": 0.0517,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.361370716510903,
|
| 986 |
+
"grad_norm": 0.35656291246414185,
|
| 987 |
+
"learning_rate": 0.0001288340340501351,
|
| 988 |
+
"loss": 0.0536,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.392523364485982,
|
| 993 |
+
"grad_norm": 0.28312814235687256,
|
| 994 |
+
"learning_rate": 0.000127844614542945,
|
| 995 |
+
"loss": 0.0445,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.423676012461059,
|
| 1000 |
+
"grad_norm": 0.34459659457206726,
|
| 1001 |
+
"learning_rate": 0.0001268522311428348,
|
| 1002 |
+
"loss": 0.0519,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.454828660436137,
|
| 1007 |
+
"grad_norm": 0.4362121820449829,
|
| 1008 |
+
"learning_rate": 0.0001258569894830906,
|
| 1009 |
+
"loss": 0.047,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.485981308411215,
|
| 1014 |
+
"grad_norm": 0.327102392911911,
|
| 1015 |
+
"learning_rate": 0.00012485899550124314,
|
| 1016 |
+
"loss": 0.0503,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.517133956386293,
|
| 1021 |
+
"grad_norm": 0.6240074634552002,
|
| 1022 |
+
"learning_rate": 0.00012385835542779144,
|
| 1023 |
+
"loss": 0.0569,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.548286604361371,
|
| 1028 |
+
"grad_norm": 0.4174221456050873,
|
| 1029 |
+
"learning_rate": 0.00012285517577489508,
|
| 1030 |
+
"loss": 0.0518,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.579439252336448,
|
| 1035 |
+
"grad_norm": 0.37431156635284424,
|
| 1036 |
+
"learning_rate": 0.00012184956332503678,
|
| 1037 |
+
"loss": 0.0466,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.610591900311526,
|
| 1042 |
+
"grad_norm": 0.401155024766922,
|
| 1043 |
+
"learning_rate": 0.00012084162511965598,
|
| 1044 |
+
"loss": 0.056,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.641744548286605,
|
| 1049 |
+
"grad_norm": 0.2195585072040558,
|
| 1050 |
+
"learning_rate": 0.00011983146844775484,
|
| 1051 |
+
"loss": 0.0503,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.672897196261682,
|
| 1056 |
+
"grad_norm": 0.308229923248291,
|
| 1057 |
+
"learning_rate": 0.00011881920083447806,
|
| 1058 |
+
"loss": 0.0544,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.70404984423676,
|
| 1063 |
+
"grad_norm": 0.3686182200908661,
|
| 1064 |
+
"learning_rate": 0.00011780493002966751,
|
| 1065 |
+
"loss": 0.0552,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.735202492211838,
|
| 1070 |
+
"grad_norm": 0.3941571116447449,
|
| 1071 |
+
"learning_rate": 0.00011678876399639269,
|
| 1072 |
+
"loss": 0.0498,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.766355140186916,
|
| 1077 |
+
"grad_norm": 0.2968638837337494,
|
| 1078 |
+
"learning_rate": 0.00011577081089945899,
|
| 1079 |
+
"loss": 0.0536,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.797507788161994,
|
| 1084 |
+
"grad_norm": 0.4896748960018158,
|
| 1085 |
+
"learning_rate": 0.00011475117909389397,
|
| 1086 |
+
"loss": 0.0519,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.8286604361370715,
|
| 1091 |
+
"grad_norm": 0.34275999665260315,
|
| 1092 |
+
"learning_rate": 0.00011372997711341373,
|
| 1093 |
+
"loss": 0.053,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.859813084112149,
|
| 1098 |
+
"grad_norm": 0.35040777921676636,
|
| 1099 |
+
"learning_rate": 0.00011270731365887011,
|
| 1100 |
+
"loss": 0.0451,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.890965732087228,
|
| 1105 |
+
"grad_norm": 0.19367630779743195,
|
| 1106 |
+
"learning_rate": 0.00011168329758668015,
|
| 1107 |
+
"loss": 0.0532,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.922118380062305,
|
| 1112 |
+
"grad_norm": 0.533416748046875,
|
| 1113 |
+
"learning_rate": 0.00011065803789723892,
|
| 1114 |
+
"loss": 0.0466,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.953271028037383,
|
| 1119 |
+
"grad_norm": 0.45454028248786926,
|
| 1120 |
+
"learning_rate": 0.00010963164372331719,
|
| 1121 |
+
"loss": 0.0536,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.984423676012461,
|
| 1126 |
+
"grad_norm": 0.3346114158630371,
|
| 1127 |
+
"learning_rate": 0.0001086042243184447,
|
| 1128 |
+
"loss": 0.0448,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.015576323987539,
|
| 1133 |
+
"grad_norm": 0.3296816349029541,
|
| 1134 |
+
"learning_rate": 0.00010757588904528106,
|
| 1135 |
+
"loss": 0.0404,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.046728971962617,
|
| 1140 |
+
"grad_norm": 0.37643900513648987,
|
| 1141 |
+
"learning_rate": 0.00010654674736397445,
|
| 1142 |
+
"loss": 0.0415,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.077881619937695,
|
| 1147 |
+
"grad_norm": 0.3826054036617279,
|
| 1148 |
+
"learning_rate": 0.00010551690882051046,
|
| 1149 |
+
"loss": 0.0452,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.109034267912772,
|
| 1154 |
+
"grad_norm": 0.3330269455909729,
|
| 1155 |
+
"learning_rate": 0.00010448648303505151,
|
| 1156 |
+
"loss": 0.0445,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.140186915887851,
|
| 1161 |
+
"grad_norm": 0.3059837222099304,
|
| 1162 |
+
"learning_rate": 0.00010345557969026834,
|
| 1163 |
+
"loss": 0.0529,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.1713395638629285,
|
| 1168 |
+
"grad_norm": 0.907908022403717,
|
| 1169 |
+
"learning_rate": 0.00010242430851966512,
|
| 1170 |
+
"loss": 0.0439,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.202492211838006,
|
| 1175 |
+
"grad_norm": 0.5197321176528931,
|
| 1176 |
+
"learning_rate": 0.00010139277929589883,
|
| 1177 |
+
"loss": 0.0459,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.233644859813084,
|
| 1182 |
+
"grad_norm": 0.4351709187030792,
|
| 1183 |
+
"learning_rate": 0.00010036110181909466,
|
| 1184 |
+
"loss": 0.0434,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.264797507788162,
|
| 1189 |
+
"grad_norm": 0.6232560873031616,
|
| 1190 |
+
"learning_rate": 9.932938590515844e-05,
|
| 1191 |
+
"loss": 0.0487,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.29595015576324,
|
| 1196 |
+
"grad_norm": 0.3728605806827545,
|
| 1197 |
+
"learning_rate": 9.82977413740875e-05,
|
| 1198 |
+
"loss": 0.0404,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.327102803738318,
|
| 1203 |
+
"grad_norm": 0.862411379814148,
|
| 1204 |
+
"learning_rate": 9.726627803828078e-05,
|
| 1205 |
+
"loss": 0.0487,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.358255451713395,
|
| 1210 |
+
"grad_norm": 0.6215007901191711,
|
| 1211 |
+
"learning_rate": 9.623510569085014e-05,
|
| 1212 |
+
"loss": 0.0453,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.389408099688474,
|
| 1217 |
+
"grad_norm": 0.3625437021255493,
|
| 1218 |
+
"learning_rate": 9.52043340939334e-05,
|
| 1219 |
+
"loss": 0.0466,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.420560747663552,
|
| 1224 |
+
"grad_norm": 0.39097148180007935,
|
| 1225 |
+
"learning_rate": 9.417407296701092e-05,
|
| 1226 |
+
"loss": 0.0413,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.451713395638629,
|
| 1231 |
+
"grad_norm": 0.5172010064125061,
|
| 1232 |
+
"learning_rate": 9.314443197522657e-05,
|
| 1233 |
+
"loss": 0.0448,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.482866043613707,
|
| 1238 |
+
"grad_norm": 0.38425084948539734,
|
| 1239 |
+
"learning_rate": 9.211552071771455e-05,
|
| 1240 |
+
"loss": 0.0434,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.5140186915887845,
|
| 1245 |
+
"grad_norm": 0.7011998295783997,
|
| 1246 |
+
"learning_rate": 9.108744871593322e-05,
|
| 1247 |
+
"loss": 0.039,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.545171339563863,
|
| 1252 |
+
"grad_norm": 0.2852397859096527,
|
| 1253 |
+
"learning_rate": 9.006032540200711e-05,
|
| 1254 |
+
"loss": 0.0461,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.576323987538941,
|
| 1259 |
+
"grad_norm": 0.28330349922180176,
|
| 1260 |
+
"learning_rate": 8.903426010707863e-05,
|
| 1261 |
+
"loss": 0.0444,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.607476635514018,
|
| 1266 |
+
"grad_norm": 0.3814966082572937,
|
| 1267 |
+
"learning_rate": 8.800936204967044e-05,
|
| 1268 |
+
"loss": 0.0443,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.638629283489097,
|
| 1273 |
+
"grad_norm": 0.4266231060028076,
|
| 1274 |
+
"learning_rate": 8.698574032405967e-05,
|
| 1275 |
+
"loss": 0.0409,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.669781931464175,
|
| 1280 |
+
"grad_norm": 0.2934357523918152,
|
| 1281 |
+
"learning_rate": 8.596350388866558e-05,
|
| 1282 |
+
"loss": 0.0391,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.700934579439252,
|
| 1287 |
+
"grad_norm": 0.30403879284858704,
|
| 1288 |
+
"learning_rate": 8.494276155445158e-05,
|
| 1289 |
+
"loss": 0.0433,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.73208722741433,
|
| 1294 |
+
"grad_norm": 0.35869526863098145,
|
| 1295 |
+
"learning_rate": 8.392362197334285e-05,
|
| 1296 |
+
"loss": 0.0434,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.763239875389408,
|
| 1301 |
+
"grad_norm": 0.26416125893592834,
|
| 1302 |
+
"learning_rate": 8.290619362666117e-05,
|
| 1303 |
+
"loss": 0.0418,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.794392523364486,
|
| 1308 |
+
"grad_norm": 0.3914806842803955,
|
| 1309 |
+
"learning_rate": 8.189058481357757e-05,
|
| 1310 |
+
"loss": 0.048,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.825545171339564,
|
| 1315 |
+
"grad_norm": 0.4492311477661133,
|
| 1316 |
+
"learning_rate": 8.087690363958467e-05,
|
| 1317 |
+
"loss": 0.0425,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.8566978193146415,
|
| 1322 |
+
"grad_norm": 0.4204324185848236,
|
| 1323 |
+
"learning_rate": 7.986525800498934e-05,
|
| 1324 |
+
"loss": 0.0352,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 5.88785046728972,
|
| 1329 |
+
"grad_norm": 0.2704547345638275,
|
| 1330 |
+
"learning_rate": 7.885575559342751e-05,
|
| 1331 |
+
"loss": 0.0375,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.919003115264798,
|
| 1336 |
+
"grad_norm": 0.2952537536621094,
|
| 1337 |
+
"learning_rate": 7.78485038604017e-05,
|
| 1338 |
+
"loss": 0.0457,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.950155763239875,
|
| 1343 |
+
"grad_norm": 0.6574923992156982,
|
| 1344 |
+
"learning_rate": 7.684361002184326e-05,
|
| 1345 |
+
"loss": 0.0401,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.981308411214953,
|
| 1350 |
+
"grad_norm": 0.4622083306312561,
|
| 1351 |
+
"learning_rate": 7.584118104269977e-05,
|
| 1352 |
+
"loss": 0.0409,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 6.012461059190031,
|
| 1357 |
+
"grad_norm": 0.3250212073326111,
|
| 1358 |
+
"learning_rate": 7.484132362554915e-05,
|
| 1359 |
+
"loss": 0.0391,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 6.043613707165109,
|
| 1364 |
+
"grad_norm": 0.32682308554649353,
|
| 1365 |
+
"learning_rate": 7.384414419924195e-05,
|
| 1366 |
+
"loss": 0.0378,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 6.074766355140187,
|
| 1371 |
+
"grad_norm": 0.7484240531921387,
|
| 1372 |
+
"learning_rate": 7.284974890757259e-05,
|
| 1373 |
+
"loss": 0.0445,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 6.105919003115265,
|
| 1378 |
+
"grad_norm": 0.3939017653465271,
|
| 1379 |
+
"learning_rate": 7.1858243597981e-05,
|
| 1380 |
+
"loss": 0.0318,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 6.137071651090342,
|
| 1385 |
+
"grad_norm": 0.4494946002960205,
|
| 1386 |
+
"learning_rate": 7.086973381028572e-05,
|
| 1387 |
+
"loss": 0.0353,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 6.168224299065421,
|
| 1392 |
+
"grad_norm": 0.2122003734111786,
|
| 1393 |
+
"learning_rate": 6.988432476544998e-05,
|
| 1394 |
+
"loss": 0.0382,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 6.1993769470404985,
|
| 1399 |
+
"grad_norm": 0.24615052342414856,
|
| 1400 |
+
"learning_rate": 6.890212135438131e-05,
|
| 1401 |
+
"loss": 0.0396,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 6.230529595015576,
|
| 1406 |
+
"grad_norm": 0.2523714005947113,
|
| 1407 |
+
"learning_rate": 6.792322812676679e-05,
|
| 1408 |
+
"loss": 0.0322,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 6.261682242990654,
|
| 1413 |
+
"grad_norm": 0.279115229845047,
|
| 1414 |
+
"learning_rate": 6.69477492799442e-05,
|
| 1415 |
+
"loss": 0.0369,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 6.292834890965732,
|
| 1420 |
+
"grad_norm": 1.3288242816925049,
|
| 1421 |
+
"learning_rate": 6.597578864781084e-05,
|
| 1422 |
+
"loss": 0.0352,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 6.32398753894081,
|
| 1427 |
+
"grad_norm": 0.3147607445716858,
|
| 1428 |
+
"learning_rate": 6.500744968977116e-05,
|
| 1429 |
+
"loss": 0.042,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 6.355140186915888,
|
| 1434 |
+
"grad_norm": 0.30732065439224243,
|
| 1435 |
+
"learning_rate": 6.404283547972396e-05,
|
| 1436 |
+
"loss": 0.0443,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 6.386292834890965,
|
| 1441 |
+
"grad_norm": 0.3305221199989319,
|
| 1442 |
+
"learning_rate": 6.308204869509084e-05,
|
| 1443 |
+
"loss": 0.0393,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 6.417445482866044,
|
| 1448 |
+
"grad_norm": 0.4128902554512024,
|
| 1449 |
+
"learning_rate": 6.212519160588688e-05,
|
| 1450 |
+
"loss": 0.0344,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 6.4485981308411215,
|
| 1455 |
+
"grad_norm": 0.3356463313102722,
|
| 1456 |
+
"learning_rate": 6.117236606383454e-05,
|
| 1457 |
+
"loss": 0.0326,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 6.479750778816199,
|
| 1462 |
+
"grad_norm": 0.2468714416027069,
|
| 1463 |
+
"learning_rate": 6.022367349152216e-05,
|
| 1464 |
+
"loss": 0.0378,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 6.510903426791277,
|
| 1469 |
+
"grad_norm": 0.29797056317329407,
|
| 1470 |
+
"learning_rate": 5.92792148716081e-05,
|
| 1471 |
+
"loss": 0.0302,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 6.542056074766355,
|
| 1476 |
+
"grad_norm": 0.23579831421375275,
|
| 1477 |
+
"learning_rate": 5.8339090736071866e-05,
|
| 1478 |
+
"loss": 0.0299,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 6.573208722741433,
|
| 1483 |
+
"grad_norm": 0.47419339418411255,
|
| 1484 |
+
"learning_rate": 5.7403401155512873e-05,
|
| 1485 |
+
"loss": 0.0331,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 6.604361370716511,
|
| 1490 |
+
"grad_norm": 0.5822060704231262,
|
| 1491 |
+
"learning_rate": 5.64722457284987e-05,
|
| 1492 |
+
"loss": 0.0359,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 6.635514018691588,
|
| 1497 |
+
"grad_norm": 0.47458353638648987,
|
| 1498 |
+
"learning_rate": 5.554572357096337e-05,
|
| 1499 |
+
"loss": 0.0372,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 6.666666666666667,
|
| 1504 |
+
"grad_norm": 0.3879990577697754,
|
| 1505 |
+
"learning_rate": 5.462393330565705e-05,
|
| 1506 |
+
"loss": 0.0357,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 6.697819314641745,
|
| 1511 |
+
"grad_norm": 0.3072240650653839,
|
| 1512 |
+
"learning_rate": 5.370697305164818e-05,
|
| 1513 |
+
"loss": 0.0446,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 6.728971962616822,
|
| 1518 |
+
"grad_norm": 0.3379099369049072,
|
| 1519 |
+
"learning_rate": 5.279494041387942e-05,
|
| 1520 |
+
"loss": 0.0327,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 6.7601246105919,
|
| 1525 |
+
"grad_norm": 0.6053034663200378,
|
| 1526 |
+
"learning_rate": 5.1887932472778056e-05,
|
| 1527 |
+
"loss": 0.032,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 6.7912772585669785,
|
| 1532 |
+
"grad_norm": 0.3459770679473877,
|
| 1533 |
+
"learning_rate": 5.098604577392245e-05,
|
| 1534 |
+
"loss": 0.0329,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 6.822429906542056,
|
| 1539 |
+
"grad_norm": 0.2976565659046173,
|
| 1540 |
+
"learning_rate": 5.0089376317765294e-05,
|
| 1541 |
+
"loss": 0.0401,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 6.853582554517134,
|
| 1546 |
+
"grad_norm": 0.34755584597587585,
|
| 1547 |
+
"learning_rate": 4.9198019549414976e-05,
|
| 1548 |
+
"loss": 0.0359,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 6.8847352024922115,
|
| 1553 |
+
"grad_norm": 0.5525044798851013,
|
| 1554 |
+
"learning_rate": 4.831207034847597e-05,
|
| 1555 |
+
"loss": 0.029,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 6.91588785046729,
|
| 1560 |
+
"grad_norm": 0.25310248136520386,
|
| 1561 |
+
"learning_rate": 4.743162301894952e-05,
|
| 1562 |
+
"loss": 0.0287,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 6.947040498442368,
|
| 1567 |
+
"grad_norm": 0.48369133472442627,
|
| 1568 |
+
"learning_rate": 4.655677127919552e-05,
|
| 1569 |
+
"loss": 0.039,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 6.978193146417445,
|
| 1574 |
+
"grad_norm": 0.6179802417755127,
|
| 1575 |
+
"learning_rate": 4.5687608251956714e-05,
|
| 1576 |
+
"loss": 0.0355,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 7.009345794392523,
|
| 1581 |
+
"grad_norm": 0.43652257323265076,
|
| 1582 |
+
"learning_rate": 4.4824226454446405e-05,
|
| 1583 |
+
"loss": 0.0315,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 7.040498442367602,
|
| 1588 |
+
"grad_norm": 0.6086826324462891,
|
| 1589 |
+
"learning_rate": 4.3966717788500525e-05,
|
| 1590 |
+
"loss": 0.0369,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 7.071651090342679,
|
| 1595 |
+
"grad_norm": 0.19126752018928528,
|
| 1596 |
+
"learning_rate": 4.311517353079519e-05,
|
| 1597 |
+
"loss": 0.0294,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 7.102803738317757,
|
| 1602 |
+
"grad_norm": 0.23197521269321442,
|
| 1603 |
+
"learning_rate": 4.2269684323130976e-05,
|
| 1604 |
+
"loss": 0.0348,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 7.133956386292835,
|
| 1609 |
+
"grad_norm": 0.36672085523605347,
|
| 1610 |
+
"learning_rate": 4.1430340162784376e-05,
|
| 1611 |
+
"loss": 0.0353,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 7.165109034267913,
|
| 1616 |
+
"grad_norm": 0.30804988741874695,
|
| 1617 |
+
"learning_rate": 4.0597230392928434e-05,
|
| 1618 |
+
"loss": 0.0258,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 7.196261682242991,
|
| 1623 |
+
"grad_norm": 0.20763657987117767,
|
| 1624 |
+
"learning_rate": 3.9770443693122526e-05,
|
| 1625 |
+
"loss": 0.0297,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 7.2274143302180685,
|
| 1630 |
+
"grad_norm": 0.4180793762207031,
|
| 1631 |
+
"learning_rate": 3.895006806987298e-05,
|
| 1632 |
+
"loss": 0.0288,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 7.258566978193146,
|
| 1637 |
+
"grad_norm": 0.3616204261779785,
|
| 1638 |
+
"learning_rate": 3.813619084726531e-05,
|
| 1639 |
+
"loss": 0.0381,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 7.289719626168225,
|
| 1644 |
+
"grad_norm": 0.3093685507774353,
|
| 1645 |
+
"learning_rate": 3.73288986576691e-05,
|
| 1646 |
+
"loss": 0.0356,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 7.320872274143302,
|
| 1651 |
+
"grad_norm": 0.1491440087556839,
|
| 1652 |
+
"learning_rate": 3.652827743251648e-05,
|
| 1653 |
+
"loss": 0.0301,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 7.35202492211838,
|
| 1658 |
+
"grad_norm": 0.1957070678472519,
|
| 1659 |
+
"learning_rate": 3.573441239315529e-05,
|
| 1660 |
+
"loss": 0.0305,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 7.383177570093458,
|
| 1665 |
+
"grad_norm": 0.2571171224117279,
|
| 1666 |
+
"learning_rate": 3.4947388041777715e-05,
|
| 1667 |
+
"loss": 0.0326,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 7.414330218068536,
|
| 1672 |
+
"grad_norm": 0.8633787631988525,
|
| 1673 |
+
"learning_rate": 3.416728815242557e-05,
|
| 1674 |
+
"loss": 0.0266,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 7.445482866043614,
|
| 1679 |
+
"grad_norm": 0.24377772212028503,
|
| 1680 |
+
"learning_rate": 3.3394195762073075e-05,
|
| 1681 |
+
"loss": 0.0315,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 7.4766355140186915,
|
| 1686 |
+
"grad_norm": 0.26799339056015015,
|
| 1687 |
+
"learning_rate": 3.262819316178792e-05,
|
| 1688 |
+
"loss": 0.0319,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 7.507788161993769,
|
| 1693 |
+
"grad_norm": 0.1959610879421234,
|
| 1694 |
+
"learning_rate": 3.1869361887972106e-05,
|
| 1695 |
+
"loss": 0.0347,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 7.538940809968848,
|
| 1700 |
+
"grad_norm": 0.2995552122592926,
|
| 1701 |
+
"learning_rate": 3.111778271368272e-05,
|
| 1702 |
+
"loss": 0.0289,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 7.570093457943925,
|
| 1707 |
+
"grad_norm": 0.22457611560821533,
|
| 1708 |
+
"learning_rate": 3.0373535640034157e-05,
|
| 1709 |
+
"loss": 0.024,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 7.601246105919003,
|
| 1714 |
+
"grad_norm": 0.2491670846939087,
|
| 1715 |
+
"learning_rate": 2.9636699887682507e-05,
|
| 1716 |
+
"loss": 0.0335,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 7.632398753894081,
|
| 1721 |
+
"grad_norm": 0.32015281915664673,
|
| 1722 |
+
"learning_rate": 2.890735388839295e-05,
|
| 1723 |
+
"loss": 0.0305,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 7.663551401869158,
|
| 1728 |
+
"grad_norm": 0.2136930227279663,
|
| 1729 |
+
"learning_rate": 2.8185575276691168e-05,
|
| 1730 |
+
"loss": 0.0322,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 7.694704049844237,
|
| 1735 |
+
"grad_norm": 0.22035038471221924,
|
| 1736 |
+
"learning_rate": 2.747144088159972e-05,
|
| 1737 |
+
"loss": 0.0286,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 7.725856697819315,
|
| 1742 |
+
"grad_norm": 0.1810670793056488,
|
| 1743 |
+
"learning_rate": 2.676502671845986e-05,
|
| 1744 |
+
"loss": 0.0286,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 7.757009345794392,
|
| 1749 |
+
"grad_norm": 0.3274972140789032,
|
| 1750 |
+
"learning_rate": 2.6066407980840335e-05,
|
| 1751 |
+
"loss": 0.0283,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 7.788161993769471,
|
| 1756 |
+
"grad_norm": 0.4231942892074585,
|
| 1757 |
+
"learning_rate": 2.5375659032533295e-05,
|
| 1758 |
+
"loss": 0.0305,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 7.8193146417445485,
|
| 1763 |
+
"grad_norm": 0.4633854329586029,
|
| 1764 |
+
"learning_rate": 2.4692853399638917e-05,
|
| 1765 |
+
"loss": 0.0324,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 7.850467289719626,
|
| 1770 |
+
"grad_norm": 0.6718659996986389,
|
| 1771 |
+
"learning_rate": 2.4018063762738863e-05,
|
| 1772 |
+
"loss": 0.0292,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 7.881619937694704,
|
| 1777 |
+
"grad_norm": 0.31123489141464233,
|
| 1778 |
+
"learning_rate": 2.33513619491599e-05,
|
| 1779 |
+
"loss": 0.0251,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 7.9127725856697815,
|
| 1784 |
+
"grad_norm": 0.19179944694042206,
|
| 1785 |
+
"learning_rate": 2.2692818925328295e-05,
|
| 1786 |
+
"loss": 0.029,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 7.94392523364486,
|
| 1791 |
+
"grad_norm": 0.23445385694503784,
|
| 1792 |
+
"learning_rate": 2.204250478921589e-05,
|
| 1793 |
+
"loss": 0.0312,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 7.975077881619938,
|
| 1798 |
+
"grad_norm": 0.17114543914794922,
|
| 1799 |
+
"learning_rate": 2.1400488762878547e-05,
|
| 1800 |
+
"loss": 0.0322,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 8.006230529595015,
|
| 1805 |
+
"grad_norm": 0.3388897180557251,
|
| 1806 |
+
"learning_rate": 2.076683918508794e-05,
|
| 1807 |
+
"loss": 0.0278,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 8.037383177570094,
|
| 1812 |
+
"grad_norm": 0.32858091592788696,
|
| 1813 |
+
"learning_rate": 2.0141623504057228e-05,
|
| 1814 |
+
"loss": 0.0317,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 8.06853582554517,
|
| 1819 |
+
"grad_norm": 0.3593621850013733,
|
| 1820 |
+
"learning_rate": 1.9524908270261656e-05,
|
| 1821 |
+
"loss": 0.0289,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 8.09968847352025,
|
| 1826 |
+
"grad_norm": 0.17125484347343445,
|
| 1827 |
+
"learning_rate": 1.8916759129354633e-05,
|
| 1828 |
+
"loss": 0.0279,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 8.130841121495328,
|
| 1833 |
+
"grad_norm": 0.14186346530914307,
|
| 1834 |
+
"learning_rate": 1.831724081518007e-05,
|
| 1835 |
+
"loss": 0.0262,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 8.161993769470405,
|
| 1840 |
+
"grad_norm": 0.2861057221889496,
|
| 1841 |
+
"learning_rate": 1.7726417142882014e-05,
|
| 1842 |
+
"loss": 0.0298,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 8.193146417445483,
|
| 1847 |
+
"grad_norm": 0.2761322855949402,
|
| 1848 |
+
"learning_rate": 1.714435100211175e-05,
|
| 1849 |
+
"loss": 0.0224,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 8.22429906542056,
|
| 1854 |
+
"grad_norm": 0.224929079413414,
|
| 1855 |
+
"learning_rate": 1.6571104350333666e-05,
|
| 1856 |
+
"loss": 0.0222,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 8.255451713395638,
|
| 1861 |
+
"grad_norm": 0.3223164975643158,
|
| 1862 |
+
"learning_rate": 1.600673820623021e-05,
|
| 1863 |
+
"loss": 0.0234,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 8.286604361370717,
|
| 1868 |
+
"grad_norm": 0.3180221617221832,
|
| 1869 |
+
"learning_rate": 1.5451312643206827e-05,
|
| 1870 |
+
"loss": 0.0328,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 8.317757009345794,
|
| 1875 |
+
"grad_norm": 0.2958988845348358,
|
| 1876 |
+
"learning_rate": 1.4904886782997496e-05,
|
| 1877 |
+
"loss": 0.0301,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 8.348909657320872,
|
| 1882 |
+
"grad_norm": 0.35848474502563477,
|
| 1883 |
+
"learning_rate": 1.4367518789371593e-05,
|
| 1884 |
+
"loss": 0.0304,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 8.38006230529595,
|
| 1889 |
+
"grad_norm": 0.4159725308418274,
|
| 1890 |
+
"learning_rate": 1.3839265861942662e-05,
|
| 1891 |
+
"loss": 0.026,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 8.411214953271028,
|
| 1896 |
+
"grad_norm": 0.2917670011520386,
|
| 1897 |
+
"learning_rate": 1.3320184230079935e-05,
|
| 1898 |
+
"loss": 0.0313,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 8.442367601246106,
|
| 1903 |
+
"grad_norm": 0.24807053804397583,
|
| 1904 |
+
"learning_rate": 1.2810329146922983e-05,
|
| 1905 |
+
"loss": 0.0277,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 8.473520249221183,
|
| 1910 |
+
"grad_norm": 0.37704989314079285,
|
| 1911 |
+
"learning_rate": 1.2309754883500368e-05,
|
| 1912 |
+
"loss": 0.0271,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 8.504672897196262,
|
| 1917 |
+
"grad_norm": 0.4665217399597168,
|
| 1918 |
+
"learning_rate": 1.181851472295281e-05,
|
| 1919 |
+
"loss": 0.031,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 8.53582554517134,
|
| 1924 |
+
"grad_norm": 0.18270960450172424,
|
| 1925 |
+
"learning_rate": 1.133666095486151e-05,
|
| 1926 |
+
"loss": 0.0282,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 8.566978193146417,
|
| 1931 |
+
"grad_norm": 0.3330720365047455,
|
| 1932 |
+
"learning_rate": 1.086424486968235e-05,
|
| 1933 |
+
"loss": 0.0231,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 8.598130841121495,
|
| 1938 |
+
"grad_norm": 0.38283267617225647,
|
| 1939 |
+
"learning_rate": 1.0401316753286106e-05,
|
| 1940 |
+
"loss": 0.0245,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 8.629283489096574,
|
| 1945 |
+
"grad_norm": 0.3002729117870331,
|
| 1946 |
+
"learning_rate": 9.947925881606013e-06,
|
| 1947 |
+
"loss": 0.0253,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 8.66043613707165,
|
| 1952 |
+
"grad_norm": 0.4419885575771332,
|
| 1953 |
+
"learning_rate": 9.504120515392556e-06,
|
| 1954 |
+
"loss": 0.0273,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 8.69158878504673,
|
| 1959 |
+
"grad_norm": 0.12886708974838257,
|
| 1960 |
+
"learning_rate": 9.069947895076381e-06,
|
| 1961 |
+
"loss": 0.0261,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 8.722741433021806,
|
| 1966 |
+
"grad_norm": 0.3838987946510315,
|
| 1967 |
+
"learning_rate": 8.645454235739903e-06,
|
| 1968 |
+
"loss": 0.0267,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 8.753894080996885,
|
| 1973 |
+
"grad_norm": 0.19890892505645752,
|
| 1974 |
+
"learning_rate": 8.230684722197945e-06,
|
| 1975 |
+
"loss": 0.0269,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 8.785046728971963,
|
| 1980 |
+
"grad_norm": 0.2921515703201294,
|
| 1981 |
+
"learning_rate": 7.825683504188031e-06,
|
| 1982 |
+
"loss": 0.0214,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 8.81619937694704,
|
| 1987 |
+
"grad_norm": 0.2481381595134735,
|
| 1988 |
+
"learning_rate": 7.430493691671026e-06,
|
| 1989 |
+
"loss": 0.0271,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 8.847352024922118,
|
| 1994 |
+
"grad_norm": 0.29253649711608887,
|
| 1995 |
+
"learning_rate": 7.0451573502422595e-06,
|
| 1996 |
+
"loss": 0.0237,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 8.878504672897197,
|
| 2001 |
+
"grad_norm": 0.1829655021429062,
|
| 2002 |
+
"learning_rate": 6.669715496653916e-06,
|
| 2003 |
+
"loss": 0.0247,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 8.909657320872274,
|
| 2008 |
+
"grad_norm": 0.20413801074028015,
|
| 2009 |
+
"learning_rate": 6.304208094449015e-06,
|
| 2010 |
+
"loss": 0.0278,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 8.940809968847352,
|
| 2015 |
+
"grad_norm": 0.6370535492897034,
|
| 2016 |
+
"learning_rate": 5.948674049707603e-06,
|
| 2017 |
+
"loss": 0.024,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 8.97196261682243,
|
| 2022 |
+
"grad_norm": 0.1832740753889084,
|
| 2023 |
+
"learning_rate": 5.603151206905355e-06,
|
| 2024 |
+
"loss": 0.0266,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 9.003115264797508,
|
| 2029 |
+
"grad_norm": 0.34532567858695984,
|
| 2030 |
+
"learning_rate": 5.267676344885297e-06,
|
| 2031 |
+
"loss": 0.026,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 9.034267912772586,
|
| 2036 |
+
"grad_norm": 0.25607308745384216,
|
| 2037 |
+
"learning_rate": 4.942285172942929e-06,
|
| 2038 |
+
"loss": 0.0289,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 9.065420560747663,
|
| 2043 |
+
"grad_norm": 0.257857084274292,
|
| 2044 |
+
"learning_rate": 4.627012327025138e-06,
|
| 2045 |
+
"loss": 0.0332,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 9.096573208722742,
|
| 2050 |
+
"grad_norm": 0.42063218355178833,
|
| 2051 |
+
"learning_rate": 4.321891366043451e-06,
|
| 2052 |
+
"loss": 0.0275,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 9.12772585669782,
|
| 2057 |
+
"grad_norm": 0.21897655725479126,
|
| 2058 |
+
"learning_rate": 4.0269547683018935e-06,
|
| 2059 |
+
"loss": 0.0229,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 9.158878504672897,
|
| 2064 |
+
"grad_norm": 0.1827220469713211,
|
| 2065 |
+
"learning_rate": 3.7422339280397955e-06,
|
| 2066 |
+
"loss": 0.0208,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 9.190031152647975,
|
| 2071 |
+
"grad_norm": 0.21246875822544098,
|
| 2072 |
+
"learning_rate": 3.4677591520901664e-06,
|
| 2073 |
+
"loss": 0.0278,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 9.221183800623052,
|
| 2078 |
+
"grad_norm": 0.31049779057502747,
|
| 2079 |
+
"learning_rate": 3.203559656653643e-06,
|
| 2080 |
+
"loss": 0.0255,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 9.25233644859813,
|
| 2085 |
+
"grad_norm": 0.14256775379180908,
|
| 2086 |
+
"learning_rate": 2.9496635641886095e-06,
|
| 2087 |
+
"loss": 0.0269,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 9.28348909657321,
|
| 2092 |
+
"grad_norm": 0.42049986124038696,
|
| 2093 |
+
"learning_rate": 2.70609790041777e-06,
|
| 2094 |
+
"loss": 0.0229,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 9.314641744548286,
|
| 2099 |
+
"grad_norm": 0.27025461196899414,
|
| 2100 |
+
"learning_rate": 2.47288859145135e-06,
|
| 2101 |
+
"loss": 0.0214,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 9.345794392523365,
|
| 2106 |
+
"grad_norm": 0.41296741366386414,
|
| 2107 |
+
"learning_rate": 2.250060461027492e-06,
|
| 2108 |
+
"loss": 0.0248,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 9.376947040498443,
|
| 2113 |
+
"grad_norm": 0.5621024370193481,
|
| 2114 |
+
"learning_rate": 2.037637227869871e-06,
|
| 2115 |
+
"loss": 0.03,
|
| 2116 |
+
"step": 3010
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 9.40809968847352,
|
| 2120 |
+
"grad_norm": 0.22565405070781708,
|
| 2121 |
+
"learning_rate": 1.8356415031629881e-06,
|
| 2122 |
+
"loss": 0.0252,
|
| 2123 |
+
"step": 3020
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 9.439252336448599,
|
| 2127 |
+
"grad_norm": 0.21029938757419586,
|
| 2128 |
+
"learning_rate": 1.644094788145345e-06,
|
| 2129 |
+
"loss": 0.0277,
|
| 2130 |
+
"step": 3030
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 9.470404984423675,
|
| 2134 |
+
"grad_norm": 0.15133824944496155,
|
| 2135 |
+
"learning_rate": 1.46301747182076e-06,
|
| 2136 |
+
"loss": 0.0226,
|
| 2137 |
+
"step": 3040
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 9.501557632398754,
|
| 2141 |
+
"grad_norm": 0.2171085625886917,
|
| 2142 |
+
"learning_rate": 1.292428828788117e-06,
|
| 2143 |
+
"loss": 0.0252,
|
| 2144 |
+
"step": 3050
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 9.532710280373832,
|
| 2148 |
+
"grad_norm": 0.2409340888261795,
|
| 2149 |
+
"learning_rate": 1.132347017189639e-06,
|
| 2150 |
+
"loss": 0.0286,
|
| 2151 |
+
"step": 3060
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 9.56386292834891,
|
| 2155 |
+
"grad_norm": 0.26242926716804504,
|
| 2156 |
+
"learning_rate": 9.82789076778101e-07,
|
| 2157 |
+
"loss": 0.0289,
|
| 2158 |
+
"step": 3070
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 9.595015576323988,
|
| 2162 |
+
"grad_norm": 0.27939870953559875,
|
| 2163 |
+
"learning_rate": 8.437709271030603e-07,
|
| 2164 |
+
"loss": 0.0231,
|
| 2165 |
+
"step": 3080
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 9.626168224299064,
|
| 2169 |
+
"grad_norm": 0.1993611454963684,
|
| 2170 |
+
"learning_rate": 7.153073658162646e-07,
|
| 2171 |
+
"loss": 0.024,
|
| 2172 |
+
"step": 3090
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 9.657320872274143,
|
| 2176 |
+
"grad_norm": 0.34144318103790283,
|
| 2177 |
+
"learning_rate": 5.974120670966033e-07,
|
| 2178 |
+
"loss": 0.0265,
|
| 2179 |
+
"step": 3100
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 9.688473520249222,
|
| 2183 |
+
"grad_norm": 0.70708829164505,
|
| 2184 |
+
"learning_rate": 4.900975801944929e-07,
|
| 2185 |
+
"loss": 0.0293,
|
| 2186 |
+
"step": 3110
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 9.719626168224298,
|
| 2190 |
+
"grad_norm": 0.21574504673480988,
|
| 2191 |
+
"learning_rate": 3.933753280961572e-07,
|
| 2192 |
+
"loss": 0.032,
|
| 2193 |
+
"step": 3120
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 9.750778816199377,
|
| 2197 |
+
"grad_norm": 0.23401866853237152,
|
| 2198 |
+
"learning_rate": 3.0725560630766593e-07,
|
| 2199 |
+
"loss": 0.0207,
|
| 2200 |
+
"step": 3130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 9.781931464174455,
|
| 2204 |
+
"grad_norm": 0.2136441171169281,
|
| 2205 |
+
"learning_rate": 2.3174758175905643e-07,
|
| 2206 |
+
"loss": 0.0301,
|
| 2207 |
+
"step": 3140
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 9.813084112149532,
|
| 2211 |
+
"grad_norm": 0.29656779766082764,
|
| 2212 |
+
"learning_rate": 1.6685929182860272e-07,
|
| 2213 |
+
"loss": 0.0281,
|
| 2214 |
+
"step": 3150
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 9.84423676012461,
|
| 2218 |
+
"grad_norm": 0.1939435750246048,
|
| 2219 |
+
"learning_rate": 1.1259764348721113e-07,
|
| 2220 |
+
"loss": 0.0216,
|
| 2221 |
+
"step": 3160
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 9.87538940809969,
|
| 2225 |
+
"grad_norm": 0.25410783290863037,
|
| 2226 |
+
"learning_rate": 6.896841256328612e-08,
|
| 2227 |
+
"loss": 0.0253,
|
| 2228 |
+
"step": 3170
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 9.906542056074766,
|
| 2232 |
+
"grad_norm": 0.14393877983093262,
|
| 2233 |
+
"learning_rate": 3.597624312789982e-08,
|
| 2234 |
+
"loss": 0.0239,
|
| 2235 |
+
"step": 3180
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 9.937694704049845,
|
| 2239 |
+
"grad_norm": 0.2124159187078476,
|
| 2240 |
+
"learning_rate": 1.3624647000431978e-08,
|
| 2241 |
+
"loss": 0.024,
|
| 2242 |
+
"step": 3190
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 9.968847352024921,
|
| 2246 |
+
"grad_norm": 0.2968740165233612,
|
| 2247 |
+
"learning_rate": 1.9160033748022445e-09,
|
| 2248 |
+
"loss": 0.032,
|
| 2249 |
+
"step": 3200
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 9.98753894080997,
|
| 2253 |
+
"step": 3206,
|
| 2254 |
+
"total_flos": 3.054403313809586e+17,
|
| 2255 |
+
"train_loss": 0.06066230240179759,
|
| 2256 |
+
"train_runtime": 2712.0984,
|
| 2257 |
+
"train_samples_per_second": 37.828,
|
| 2258 |
+
"train_steps_per_second": 1.182
|
| 2259 |
+
}
|
| 2260 |
+
],
|
| 2261 |
+
"logging_steps": 10,
|
| 2262 |
+
"max_steps": 3206,
|
| 2263 |
+
"num_input_tokens_seen": 0,
|
| 2264 |
+
"num_train_epochs": 10,
|
| 2265 |
+
"save_steps": 10000,
|
| 2266 |
+
"stateful_callbacks": {
|
| 2267 |
+
"TrainerControl": {
|
| 2268 |
+
"args": {
|
| 2269 |
+
"should_epoch_stop": false,
|
| 2270 |
+
"should_evaluate": false,
|
| 2271 |
+
"should_log": false,
|
| 2272 |
+
"should_save": true,
|
| 2273 |
+
"should_training_stop": true
|
| 2274 |
+
},
|
| 2275 |
+
"attributes": {}
|
| 2276 |
+
}
|
| 2277 |
+
},
|
| 2278 |
+
"total_flos": 3.054403313809586e+17,
|
| 2279 |
+
"train_batch_size": 32,
|
| 2280 |
+
"trial_name": null,
|
| 2281 |
+
"trial_params": null
|
| 2282 |
+
}
|