Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2274 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2274 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.965157329240215,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3200,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0061396776669224865,
|
| 14 |
+
"grad_norm": 15.421392440795898,
|
| 15 |
+
"learning_rate": 2.94478527607362e-06,
|
| 16 |
+
"loss": 7.0576,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.012279355333844973,
|
| 21 |
+
"grad_norm": 9.991676330566406,
|
| 22 |
+
"learning_rate": 6.625766871165644e-06,
|
| 23 |
+
"loss": 6.1821,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.01841903300076746,
|
| 28 |
+
"grad_norm": 8.402695655822754,
|
| 29 |
+
"learning_rate": 1.030674846625767e-05,
|
| 30 |
+
"loss": 5.2081,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.024558710667689946,
|
| 35 |
+
"grad_norm": 8.112366676330566,
|
| 36 |
+
"learning_rate": 1.3987730061349692e-05,
|
| 37 |
+
"loss": 4.4492,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.03069838833461243,
|
| 42 |
+
"grad_norm": 8.540102005004883,
|
| 43 |
+
"learning_rate": 1.766871165644172e-05,
|
| 44 |
+
"loss": 3.698,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.03683806600153492,
|
| 49 |
+
"grad_norm": 11.69017505645752,
|
| 50 |
+
"learning_rate": 2.1349693251533743e-05,
|
| 51 |
+
"loss": 2.783,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.042977743668457406,
|
| 56 |
+
"grad_norm": 2.1856040954589844,
|
| 57 |
+
"learning_rate": 2.5030674846625765e-05,
|
| 58 |
+
"loss": 1.5322,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.04911742133537989,
|
| 63 |
+
"grad_norm": 1.7973095178604126,
|
| 64 |
+
"learning_rate": 2.8711656441717793e-05,
|
| 65 |
+
"loss": 0.9891,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.05525709900230238,
|
| 70 |
+
"grad_norm": 1.1040600538253784,
|
| 71 |
+
"learning_rate": 3.239263803680982e-05,
|
| 72 |
+
"loss": 0.7472,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.06139677666922486,
|
| 77 |
+
"grad_norm": 0.9011868238449097,
|
| 78 |
+
"learning_rate": 3.607361963190184e-05,
|
| 79 |
+
"loss": 0.6202,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.06753645433614736,
|
| 84 |
+
"grad_norm": 0.6866259574890137,
|
| 85 |
+
"learning_rate": 3.975460122699386e-05,
|
| 86 |
+
"loss": 0.5531,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.07367613200306984,
|
| 91 |
+
"grad_norm": 0.7291966676712036,
|
| 92 |
+
"learning_rate": 4.3435582822085894e-05,
|
| 93 |
+
"loss": 0.5087,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.07981580966999233,
|
| 98 |
+
"grad_norm": 0.6475340127944946,
|
| 99 |
+
"learning_rate": 4.711656441717792e-05,
|
| 100 |
+
"loss": 0.4658,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.08595548733691481,
|
| 105 |
+
"grad_norm": 0.8856902122497559,
|
| 106 |
+
"learning_rate": 5.079754601226994e-05,
|
| 107 |
+
"loss": 0.4465,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.0920951650038373,
|
| 112 |
+
"grad_norm": 0.6716117262840271,
|
| 113 |
+
"learning_rate": 5.4478527607361964e-05,
|
| 114 |
+
"loss": 0.4154,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.09823484267075978,
|
| 119 |
+
"grad_norm": 0.6893500089645386,
|
| 120 |
+
"learning_rate": 5.815950920245399e-05,
|
| 121 |
+
"loss": 0.4051,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.10437452033768227,
|
| 126 |
+
"grad_norm": 0.7085395455360413,
|
| 127 |
+
"learning_rate": 5.9903069466882075e-05,
|
| 128 |
+
"loss": 0.3903,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.11051419800460476,
|
| 133 |
+
"grad_norm": 0.5268082618713379,
|
| 134 |
+
"learning_rate": 5.970920840064621e-05,
|
| 135 |
+
"loss": 0.3727,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.11665387567152724,
|
| 140 |
+
"grad_norm": 0.5620772838592529,
|
| 141 |
+
"learning_rate": 5.951534733441034e-05,
|
| 142 |
+
"loss": 0.3656,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.12279355333844973,
|
| 147 |
+
"grad_norm": 0.5411626696586609,
|
| 148 |
+
"learning_rate": 5.932148626817448e-05,
|
| 149 |
+
"loss": 0.3527,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.1289332310053722,
|
| 154 |
+
"grad_norm": 0.5481147170066833,
|
| 155 |
+
"learning_rate": 5.912762520193861e-05,
|
| 156 |
+
"loss": 0.3538,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.1350729086722947,
|
| 161 |
+
"grad_norm": 0.5867046117782593,
|
| 162 |
+
"learning_rate": 5.8933764135702745e-05,
|
| 163 |
+
"loss": 0.3412,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.14121258633921718,
|
| 168 |
+
"grad_norm": 0.5871285200119019,
|
| 169 |
+
"learning_rate": 5.873990306946688e-05,
|
| 170 |
+
"loss": 0.3408,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.14735226400613968,
|
| 175 |
+
"grad_norm": 0.6050668358802795,
|
| 176 |
+
"learning_rate": 5.8546042003231024e-05,
|
| 177 |
+
"loss": 0.3352,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.15349194167306215,
|
| 182 |
+
"grad_norm": 0.5081577897071838,
|
| 183 |
+
"learning_rate": 5.8352180936995157e-05,
|
| 184 |
+
"loss": 0.3267,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.15963161933998465,
|
| 189 |
+
"grad_norm": 0.5677348971366882,
|
| 190 |
+
"learning_rate": 5.815831987075929e-05,
|
| 191 |
+
"loss": 0.3313,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.16577129700690713,
|
| 196 |
+
"grad_norm": 0.487120658159256,
|
| 197 |
+
"learning_rate": 5.796445880452343e-05,
|
| 198 |
+
"loss": 0.3229,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.17191097467382963,
|
| 203 |
+
"grad_norm": 0.58442622423172,
|
| 204 |
+
"learning_rate": 5.777059773828756e-05,
|
| 205 |
+
"loss": 0.3199,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.1780506523407521,
|
| 210 |
+
"grad_norm": 0.5962138772010803,
|
| 211 |
+
"learning_rate": 5.7576736672051694e-05,
|
| 212 |
+
"loss": 0.3154,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.1841903300076746,
|
| 217 |
+
"grad_norm": 0.49995115399360657,
|
| 218 |
+
"learning_rate": 5.7382875605815834e-05,
|
| 219 |
+
"loss": 0.3157,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.1903300076745971,
|
| 224 |
+
"grad_norm": 0.5417408347129822,
|
| 225 |
+
"learning_rate": 5.718901453957997e-05,
|
| 226 |
+
"loss": 0.3107,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.19646968534151957,
|
| 231 |
+
"grad_norm": 0.4579584300518036,
|
| 232 |
+
"learning_rate": 5.6995153473344106e-05,
|
| 233 |
+
"loss": 0.3074,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.20260936300844207,
|
| 238 |
+
"grad_norm": 0.47001534700393677,
|
| 239 |
+
"learning_rate": 5.680129240710824e-05,
|
| 240 |
+
"loss": 0.3048,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.20874904067536454,
|
| 245 |
+
"grad_norm": 0.4944261312484741,
|
| 246 |
+
"learning_rate": 5.660743134087238e-05,
|
| 247 |
+
"loss": 0.3056,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.21488871834228704,
|
| 252 |
+
"grad_norm": 0.504847526550293,
|
| 253 |
+
"learning_rate": 5.641357027463651e-05,
|
| 254 |
+
"loss": 0.3071,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.2210283960092095,
|
| 259 |
+
"grad_norm": 0.5896368622779846,
|
| 260 |
+
"learning_rate": 5.621970920840064e-05,
|
| 261 |
+
"loss": 0.2973,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.227168073676132,
|
| 266 |
+
"grad_norm": 0.4247358739376068,
|
| 267 |
+
"learning_rate": 5.602584814216478e-05,
|
| 268 |
+
"loss": 0.2993,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.23330775134305448,
|
| 273 |
+
"grad_norm": 0.44181105494499207,
|
| 274 |
+
"learning_rate": 5.583198707592892e-05,
|
| 275 |
+
"loss": 0.2976,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.23944742900997698,
|
| 280 |
+
"grad_norm": 0.44813117384910583,
|
| 281 |
+
"learning_rate": 5.5638126009693055e-05,
|
| 282 |
+
"loss": 0.2906,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.24558710667689945,
|
| 287 |
+
"grad_norm": 0.5210517644882202,
|
| 288 |
+
"learning_rate": 5.544426494345719e-05,
|
| 289 |
+
"loss": 0.2883,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.25172678434382195,
|
| 294 |
+
"grad_norm": 0.519637942314148,
|
| 295 |
+
"learning_rate": 5.525040387722133e-05,
|
| 296 |
+
"loss": 0.289,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.2578664620107444,
|
| 301 |
+
"grad_norm": 0.39594411849975586,
|
| 302 |
+
"learning_rate": 5.505654281098546e-05,
|
| 303 |
+
"loss": 0.2883,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.2640061396776669,
|
| 308 |
+
"grad_norm": 0.44631266593933105,
|
| 309 |
+
"learning_rate": 5.486268174474959e-05,
|
| 310 |
+
"loss": 0.2836,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.2701458173445894,
|
| 315 |
+
"grad_norm": 0.40706467628479004,
|
| 316 |
+
"learning_rate": 5.466882067851374e-05,
|
| 317 |
+
"loss": 0.2872,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.2762854950115119,
|
| 322 |
+
"grad_norm": 0.45857444405555725,
|
| 323 |
+
"learning_rate": 5.447495961227787e-05,
|
| 324 |
+
"loss": 0.2833,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.28242517267843437,
|
| 329 |
+
"grad_norm": 0.5002241730690002,
|
| 330 |
+
"learning_rate": 5.4281098546042004e-05,
|
| 331 |
+
"loss": 0.2835,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.2885648503453569,
|
| 336 |
+
"grad_norm": 0.41167810559272766,
|
| 337 |
+
"learning_rate": 5.408723747980614e-05,
|
| 338 |
+
"loss": 0.2859,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.29470452801227937,
|
| 343 |
+
"grad_norm": 0.4499143064022064,
|
| 344 |
+
"learning_rate": 5.3893376413570276e-05,
|
| 345 |
+
"loss": 0.283,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.30084420567920184,
|
| 350 |
+
"grad_norm": 0.4298258125782013,
|
| 351 |
+
"learning_rate": 5.369951534733441e-05,
|
| 352 |
+
"loss": 0.278,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.3069838833461243,
|
| 357 |
+
"grad_norm": 0.47322022914886475,
|
| 358 |
+
"learning_rate": 5.350565428109855e-05,
|
| 359 |
+
"loss": 0.2775,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.31312356101304684,
|
| 364 |
+
"grad_norm": 0.583753228187561,
|
| 365 |
+
"learning_rate": 5.331179321486269e-05,
|
| 366 |
+
"loss": 0.277,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3192632386799693,
|
| 371 |
+
"grad_norm": 0.4866049885749817,
|
| 372 |
+
"learning_rate": 5.311793214862682e-05,
|
| 373 |
+
"loss": 0.2801,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.3254029163468918,
|
| 378 |
+
"grad_norm": 0.4380209445953369,
|
| 379 |
+
"learning_rate": 5.2924071082390953e-05,
|
| 380 |
+
"loss": 0.281,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.33154259401381425,
|
| 385 |
+
"grad_norm": 0.4283748269081116,
|
| 386 |
+
"learning_rate": 5.2730210016155086e-05,
|
| 387 |
+
"loss": 0.2761,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.3376822716807368,
|
| 392 |
+
"grad_norm": 0.423459529876709,
|
| 393 |
+
"learning_rate": 5.2536348949919226e-05,
|
| 394 |
+
"loss": 0.2794,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.34382194934765925,
|
| 399 |
+
"grad_norm": 0.41094839572906494,
|
| 400 |
+
"learning_rate": 5.234248788368336e-05,
|
| 401 |
+
"loss": 0.2729,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.3499616270145817,
|
| 406 |
+
"grad_norm": 0.43751445412635803,
|
| 407 |
+
"learning_rate": 5.21486268174475e-05,
|
| 408 |
+
"loss": 0.2718,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.3561013046815042,
|
| 413 |
+
"grad_norm": 0.44664233922958374,
|
| 414 |
+
"learning_rate": 5.195476575121164e-05,
|
| 415 |
+
"loss": 0.273,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.3622409823484267,
|
| 420 |
+
"grad_norm": 0.41911059617996216,
|
| 421 |
+
"learning_rate": 5.176090468497577e-05,
|
| 422 |
+
"loss": 0.2723,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.3683806600153492,
|
| 427 |
+
"grad_norm": 0.42424502968788147,
|
| 428 |
+
"learning_rate": 5.15670436187399e-05,
|
| 429 |
+
"loss": 0.2723,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.37452033768227166,
|
| 434 |
+
"grad_norm": 0.4342273473739624,
|
| 435 |
+
"learning_rate": 5.137318255250404e-05,
|
| 436 |
+
"loss": 0.27,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.3806600153491942,
|
| 441 |
+
"grad_norm": 0.4648367166519165,
|
| 442 |
+
"learning_rate": 5.1179321486268175e-05,
|
| 443 |
+
"loss": 0.269,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.38679969301611666,
|
| 448 |
+
"grad_norm": 0.4721224904060364,
|
| 449 |
+
"learning_rate": 5.098546042003231e-05,
|
| 450 |
+
"loss": 0.2728,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.39293937068303914,
|
| 455 |
+
"grad_norm": 0.4093610644340515,
|
| 456 |
+
"learning_rate": 5.079159935379645e-05,
|
| 457 |
+
"loss": 0.2692,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.3990790483499616,
|
| 462 |
+
"grad_norm": 0.41683274507522583,
|
| 463 |
+
"learning_rate": 5.0597738287560586e-05,
|
| 464 |
+
"loss": 0.2635,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.40521872601688413,
|
| 469 |
+
"grad_norm": 0.4177282154560089,
|
| 470 |
+
"learning_rate": 5.040387722132472e-05,
|
| 471 |
+
"loss": 0.267,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.4113584036838066,
|
| 476 |
+
"grad_norm": 0.3698485791683197,
|
| 477 |
+
"learning_rate": 5.021001615508885e-05,
|
| 478 |
+
"loss": 0.2701,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.4174980813507291,
|
| 483 |
+
"grad_norm": 0.3479083478450775,
|
| 484 |
+
"learning_rate": 5.001615508885299e-05,
|
| 485 |
+
"loss": 0.2663,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.42363775901765155,
|
| 490 |
+
"grad_norm": 0.44725146889686584,
|
| 491 |
+
"learning_rate": 4.9822294022617124e-05,
|
| 492 |
+
"loss": 0.2649,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.4297774366845741,
|
| 497 |
+
"grad_norm": 0.3550412952899933,
|
| 498 |
+
"learning_rate": 4.9628432956381264e-05,
|
| 499 |
+
"loss": 0.266,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.43591711435149655,
|
| 504 |
+
"grad_norm": 0.4252576529979706,
|
| 505 |
+
"learning_rate": 4.9434571890145396e-05,
|
| 506 |
+
"loss": 0.2643,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.442056792018419,
|
| 511 |
+
"grad_norm": 0.39061039686203003,
|
| 512 |
+
"learning_rate": 4.9240710823909536e-05,
|
| 513 |
+
"loss": 0.2621,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.4481964696853415,
|
| 518 |
+
"grad_norm": 0.4167734980583191,
|
| 519 |
+
"learning_rate": 4.904684975767367e-05,
|
| 520 |
+
"loss": 0.2643,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.454336147352264,
|
| 525 |
+
"grad_norm": 0.3748829662799835,
|
| 526 |
+
"learning_rate": 4.88529886914378e-05,
|
| 527 |
+
"loss": 0.268,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.4604758250191865,
|
| 532 |
+
"grad_norm": 0.35719943046569824,
|
| 533 |
+
"learning_rate": 4.865912762520194e-05,
|
| 534 |
+
"loss": 0.265,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.46661550268610896,
|
| 539 |
+
"grad_norm": 0.3589203357696533,
|
| 540 |
+
"learning_rate": 4.846526655896607e-05,
|
| 541 |
+
"loss": 0.2583,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.4727551803530315,
|
| 546 |
+
"grad_norm": 0.381046861410141,
|
| 547 |
+
"learning_rate": 4.827140549273021e-05,
|
| 548 |
+
"loss": 0.2613,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.47889485801995396,
|
| 553 |
+
"grad_norm": 0.46023252606391907,
|
| 554 |
+
"learning_rate": 4.8077544426494345e-05,
|
| 555 |
+
"loss": 0.2584,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.48503453568687643,
|
| 560 |
+
"grad_norm": 0.3704688549041748,
|
| 561 |
+
"learning_rate": 4.7883683360258485e-05,
|
| 562 |
+
"loss": 0.2573,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.4911742133537989,
|
| 567 |
+
"grad_norm": 0.3550325036048889,
|
| 568 |
+
"learning_rate": 4.768982229402262e-05,
|
| 569 |
+
"loss": 0.2638,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.49731389102072143,
|
| 574 |
+
"grad_norm": 0.3660726547241211,
|
| 575 |
+
"learning_rate": 4.749596122778675e-05,
|
| 576 |
+
"loss": 0.2614,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.5034535686876439,
|
| 581 |
+
"grad_norm": 0.3629954159259796,
|
| 582 |
+
"learning_rate": 4.730210016155089e-05,
|
| 583 |
+
"loss": 0.2589,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.5095932463545664,
|
| 588 |
+
"grad_norm": 0.3618892431259155,
|
| 589 |
+
"learning_rate": 4.710823909531502e-05,
|
| 590 |
+
"loss": 0.2625,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.5157329240214888,
|
| 595 |
+
"grad_norm": 0.4103853106498718,
|
| 596 |
+
"learning_rate": 4.691437802907916e-05,
|
| 597 |
+
"loss": 0.256,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.5218726016884113,
|
| 602 |
+
"grad_norm": 0.4005793631076813,
|
| 603 |
+
"learning_rate": 4.67205169628433e-05,
|
| 604 |
+
"loss": 0.2609,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.5280122793553338,
|
| 609 |
+
"grad_norm": 0.4035857617855072,
|
| 610 |
+
"learning_rate": 4.6526655896607434e-05,
|
| 611 |
+
"loss": 0.2602,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.5341519570222564,
|
| 616 |
+
"grad_norm": 0.3564916253089905,
|
| 617 |
+
"learning_rate": 4.633279483037157e-05,
|
| 618 |
+
"loss": 0.2599,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.5402916346891788,
|
| 623 |
+
"grad_norm": 0.3407823443412781,
|
| 624 |
+
"learning_rate": 4.61389337641357e-05,
|
| 625 |
+
"loss": 0.2566,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.5464313123561013,
|
| 630 |
+
"grad_norm": 0.42279985547065735,
|
| 631 |
+
"learning_rate": 4.594507269789984e-05,
|
| 632 |
+
"loss": 0.2596,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.5525709900230238,
|
| 637 |
+
"grad_norm": 0.3459627628326416,
|
| 638 |
+
"learning_rate": 4.575121163166398e-05,
|
| 639 |
+
"loss": 0.2579,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.5587106676899463,
|
| 644 |
+
"grad_norm": 0.3690173327922821,
|
| 645 |
+
"learning_rate": 4.555735056542811e-05,
|
| 646 |
+
"loss": 0.2555,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.5648503453568687,
|
| 651 |
+
"grad_norm": 0.3533875346183777,
|
| 652 |
+
"learning_rate": 4.536348949919225e-05,
|
| 653 |
+
"loss": 0.256,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.5709900230237912,
|
| 658 |
+
"grad_norm": 0.34665781259536743,
|
| 659 |
+
"learning_rate": 4.516962843295638e-05,
|
| 660 |
+
"loss": 0.2538,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.5771297006907138,
|
| 665 |
+
"grad_norm": 0.31980058550834656,
|
| 666 |
+
"learning_rate": 4.4975767366720516e-05,
|
| 667 |
+
"loss": 0.2581,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.5832693783576363,
|
| 672 |
+
"grad_norm": 0.35057130455970764,
|
| 673 |
+
"learning_rate": 4.478190630048465e-05,
|
| 674 |
+
"loss": 0.2566,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.5894090560245587,
|
| 679 |
+
"grad_norm": 0.32480597496032715,
|
| 680 |
+
"learning_rate": 4.458804523424879e-05,
|
| 681 |
+
"loss": 0.255,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.5955487336914812,
|
| 686 |
+
"grad_norm": 0.3633149564266205,
|
| 687 |
+
"learning_rate": 4.439418416801293e-05,
|
| 688 |
+
"loss": 0.2551,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.6016884113584037,
|
| 693 |
+
"grad_norm": 0.34456104040145874,
|
| 694 |
+
"learning_rate": 4.420032310177706e-05,
|
| 695 |
+
"loss": 0.253,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.6078280890253261,
|
| 700 |
+
"grad_norm": 0.3510631322860718,
|
| 701 |
+
"learning_rate": 4.40064620355412e-05,
|
| 702 |
+
"loss": 0.254,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.6139677666922486,
|
| 707 |
+
"grad_norm": 0.35537025332450867,
|
| 708 |
+
"learning_rate": 4.381260096930533e-05,
|
| 709 |
+
"loss": 0.2547,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.6201074443591711,
|
| 714 |
+
"grad_norm": 0.36508992314338684,
|
| 715 |
+
"learning_rate": 4.3618739903069465e-05,
|
| 716 |
+
"loss": 0.2517,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.6262471220260937,
|
| 721 |
+
"grad_norm": 0.383023202419281,
|
| 722 |
+
"learning_rate": 4.3424878836833605e-05,
|
| 723 |
+
"loss": 0.2532,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.6323867996930161,
|
| 728 |
+
"grad_norm": 0.3511675298213959,
|
| 729 |
+
"learning_rate": 4.323101777059774e-05,
|
| 730 |
+
"loss": 0.2496,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.6385264773599386,
|
| 735 |
+
"grad_norm": 0.39103734493255615,
|
| 736 |
+
"learning_rate": 4.303715670436188e-05,
|
| 737 |
+
"loss": 0.2502,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.6446661550268611,
|
| 742 |
+
"grad_norm": 0.32976630330085754,
|
| 743 |
+
"learning_rate": 4.284329563812601e-05,
|
| 744 |
+
"loss": 0.2516,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.6508058326937836,
|
| 749 |
+
"grad_norm": 0.3627534508705139,
|
| 750 |
+
"learning_rate": 4.264943457189015e-05,
|
| 751 |
+
"loss": 0.2538,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.656945510360706,
|
| 756 |
+
"grad_norm": 0.37391820549964905,
|
| 757 |
+
"learning_rate": 4.245557350565428e-05,
|
| 758 |
+
"loss": 0.2528,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6630851880276285,
|
| 763 |
+
"grad_norm": 0.3327382802963257,
|
| 764 |
+
"learning_rate": 4.2261712439418414e-05,
|
| 765 |
+
"loss": 0.2523,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.6692248656945511,
|
| 770 |
+
"grad_norm": 0.3251570463180542,
|
| 771 |
+
"learning_rate": 4.2067851373182554e-05,
|
| 772 |
+
"loss": 0.2509,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.6753645433614736,
|
| 777 |
+
"grad_norm": 0.39253005385398865,
|
| 778 |
+
"learning_rate": 4.1873990306946693e-05,
|
| 779 |
+
"loss": 0.2507,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.681504221028396,
|
| 784 |
+
"grad_norm": 0.33449676632881165,
|
| 785 |
+
"learning_rate": 4.1680129240710826e-05,
|
| 786 |
+
"loss": 0.2481,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.6876438986953185,
|
| 791 |
+
"grad_norm": 0.37863296270370483,
|
| 792 |
+
"learning_rate": 4.148626817447496e-05,
|
| 793 |
+
"loss": 0.252,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.693783576362241,
|
| 798 |
+
"grad_norm": 0.31935396790504456,
|
| 799 |
+
"learning_rate": 4.12924071082391e-05,
|
| 800 |
+
"loss": 0.2466,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.6999232540291634,
|
| 805 |
+
"grad_norm": 0.313819020986557,
|
| 806 |
+
"learning_rate": 4.109854604200323e-05,
|
| 807 |
+
"loss": 0.251,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.7060629316960859,
|
| 812 |
+
"grad_norm": 0.3724668622016907,
|
| 813 |
+
"learning_rate": 4.0904684975767364e-05,
|
| 814 |
+
"loss": 0.2518,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.7122026093630084,
|
| 819 |
+
"grad_norm": 0.33299896121025085,
|
| 820 |
+
"learning_rate": 4.07108239095315e-05,
|
| 821 |
+
"loss": 0.2472,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.718342287029931,
|
| 826 |
+
"grad_norm": 0.3538409173488617,
|
| 827 |
+
"learning_rate": 4.051696284329564e-05,
|
| 828 |
+
"loss": 0.2486,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.7244819646968534,
|
| 833 |
+
"grad_norm": 0.3172655999660492,
|
| 834 |
+
"learning_rate": 4.0323101777059775e-05,
|
| 835 |
+
"loss": 0.2501,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.7306216423637759,
|
| 840 |
+
"grad_norm": 0.3064083456993103,
|
| 841 |
+
"learning_rate": 4.012924071082391e-05,
|
| 842 |
+
"loss": 0.2453,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.7367613200306984,
|
| 847 |
+
"grad_norm": 0.3342798352241516,
|
| 848 |
+
"learning_rate": 3.993537964458805e-05,
|
| 849 |
+
"loss": 0.2498,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.7429009976976209,
|
| 854 |
+
"grad_norm": 0.31463661789894104,
|
| 855 |
+
"learning_rate": 3.974151857835218e-05,
|
| 856 |
+
"loss": 0.2446,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.7490406753645433,
|
| 861 |
+
"grad_norm": 0.3608093559741974,
|
| 862 |
+
"learning_rate": 3.954765751211631e-05,
|
| 863 |
+
"loss": 0.2483,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.7551803530314658,
|
| 868 |
+
"grad_norm": 0.3194938898086548,
|
| 869 |
+
"learning_rate": 3.935379644588046e-05,
|
| 870 |
+
"loss": 0.2493,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.7613200306983884,
|
| 875 |
+
"grad_norm": 0.3166663944721222,
|
| 876 |
+
"learning_rate": 3.915993537964459e-05,
|
| 877 |
+
"loss": 0.2481,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.7674597083653109,
|
| 882 |
+
"grad_norm": 0.36826014518737793,
|
| 883 |
+
"learning_rate": 3.8966074313408725e-05,
|
| 884 |
+
"loss": 0.2451,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.7735993860322333,
|
| 889 |
+
"grad_norm": 0.31451407074928284,
|
| 890 |
+
"learning_rate": 3.8772213247172864e-05,
|
| 891 |
+
"loss": 0.2455,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.7797390636991558,
|
| 896 |
+
"grad_norm": 0.3546365201473236,
|
| 897 |
+
"learning_rate": 3.8578352180937e-05,
|
| 898 |
+
"loss": 0.2417,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.7858787413660783,
|
| 903 |
+
"grad_norm": 0.37739697098731995,
|
| 904 |
+
"learning_rate": 3.838449111470113e-05,
|
| 905 |
+
"loss": 0.2471,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.7920184190330007,
|
| 910 |
+
"grad_norm": 0.29908350110054016,
|
| 911 |
+
"learning_rate": 3.819063004846526e-05,
|
| 912 |
+
"loss": 0.2502,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.7981580966999232,
|
| 917 |
+
"grad_norm": 0.388713538646698,
|
| 918 |
+
"learning_rate": 3.799676898222941e-05,
|
| 919 |
+
"loss": 0.2447,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.8042977743668457,
|
| 924 |
+
"grad_norm": 0.31080490350723267,
|
| 925 |
+
"learning_rate": 3.780290791599354e-05,
|
| 926 |
+
"loss": 0.2458,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.8104374520337683,
|
| 931 |
+
"grad_norm": 0.30132901668548584,
|
| 932 |
+
"learning_rate": 3.7609046849757674e-05,
|
| 933 |
+
"loss": 0.2455,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.8165771297006907,
|
| 938 |
+
"grad_norm": 0.3251103162765503,
|
| 939 |
+
"learning_rate": 3.741518578352181e-05,
|
| 940 |
+
"loss": 0.2454,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.8227168073676132,
|
| 945 |
+
"grad_norm": 0.33006975054740906,
|
| 946 |
+
"learning_rate": 3.7221324717285946e-05,
|
| 947 |
+
"loss": 0.2466,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.8288564850345357,
|
| 952 |
+
"grad_norm": 0.3195136785507202,
|
| 953 |
+
"learning_rate": 3.702746365105008e-05,
|
| 954 |
+
"loss": 0.2451,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.8349961627014582,
|
| 959 |
+
"grad_norm": 0.33961763978004456,
|
| 960 |
+
"learning_rate": 3.683360258481421e-05,
|
| 961 |
+
"loss": 0.2451,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.8411358403683806,
|
| 966 |
+
"grad_norm": 0.3274971544742584,
|
| 967 |
+
"learning_rate": 3.663974151857836e-05,
|
| 968 |
+
"loss": 0.2445,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.8472755180353031,
|
| 973 |
+
"grad_norm": 0.3316924273967743,
|
| 974 |
+
"learning_rate": 3.644588045234249e-05,
|
| 975 |
+
"loss": 0.2421,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.8534151957022257,
|
| 980 |
+
"grad_norm": 0.34864816069602966,
|
| 981 |
+
"learning_rate": 3.625201938610662e-05,
|
| 982 |
+
"loss": 0.2432,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.8595548733691482,
|
| 987 |
+
"grad_norm": 0.34979552030563354,
|
| 988 |
+
"learning_rate": 3.605815831987076e-05,
|
| 989 |
+
"loss": 0.2445,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.8656945510360706,
|
| 994 |
+
"grad_norm": 0.3480820655822754,
|
| 995 |
+
"learning_rate": 3.5864297253634895e-05,
|
| 996 |
+
"loss": 0.2423,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.8718342287029931,
|
| 1001 |
+
"grad_norm": 0.3144533038139343,
|
| 1002 |
+
"learning_rate": 3.567043618739903e-05,
|
| 1003 |
+
"loss": 0.2423,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.8779739063699156,
|
| 1008 |
+
"grad_norm": 0.35407504439353943,
|
| 1009 |
+
"learning_rate": 3.547657512116317e-05,
|
| 1010 |
+
"loss": 0.2453,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.884113584036838,
|
| 1015 |
+
"grad_norm": 0.3296087980270386,
|
| 1016 |
+
"learning_rate": 3.528271405492731e-05,
|
| 1017 |
+
"loss": 0.2433,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.8902532617037605,
|
| 1022 |
+
"grad_norm": 0.37147656083106995,
|
| 1023 |
+
"learning_rate": 3.508885298869144e-05,
|
| 1024 |
+
"loss": 0.243,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.896392939370683,
|
| 1029 |
+
"grad_norm": 0.291891485452652,
|
| 1030 |
+
"learning_rate": 3.489499192245557e-05,
|
| 1031 |
+
"loss": 0.2415,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.9025326170376056,
|
| 1036 |
+
"grad_norm": 0.28548404574394226,
|
| 1037 |
+
"learning_rate": 3.470113085621971e-05,
|
| 1038 |
+
"loss": 0.244,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.908672294704528,
|
| 1043 |
+
"grad_norm": 0.35448750853538513,
|
| 1044 |
+
"learning_rate": 3.4507269789983844e-05,
|
| 1045 |
+
"loss": 0.2419,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.9148119723714505,
|
| 1050 |
+
"grad_norm": 0.3614279329776764,
|
| 1051 |
+
"learning_rate": 3.431340872374798e-05,
|
| 1052 |
+
"loss": 0.2425,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.920951650038373,
|
| 1057 |
+
"grad_norm": 0.3446723520755768,
|
| 1058 |
+
"learning_rate": 3.411954765751212e-05,
|
| 1059 |
+
"loss": 0.245,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.9270913277052955,
|
| 1064 |
+
"grad_norm": 0.336374968290329,
|
| 1065 |
+
"learning_rate": 3.3925686591276256e-05,
|
| 1066 |
+
"loss": 0.2411,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.9332310053722179,
|
| 1071 |
+
"grad_norm": 0.31548479199409485,
|
| 1072 |
+
"learning_rate": 3.373182552504039e-05,
|
| 1073 |
+
"loss": 0.2414,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.9393706830391404,
|
| 1078 |
+
"grad_norm": 0.320119172334671,
|
| 1079 |
+
"learning_rate": 3.353796445880452e-05,
|
| 1080 |
+
"loss": 0.2397,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.945510360706063,
|
| 1085 |
+
"grad_norm": 0.3126417100429535,
|
| 1086 |
+
"learning_rate": 3.334410339256866e-05,
|
| 1087 |
+
"loss": 0.2386,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.9516500383729855,
|
| 1092 |
+
"grad_norm": 0.30536866188049316,
|
| 1093 |
+
"learning_rate": 3.3150242326332794e-05,
|
| 1094 |
+
"loss": 0.2413,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.9577897160399079,
|
| 1099 |
+
"grad_norm": 0.30234816670417786,
|
| 1100 |
+
"learning_rate": 3.2956381260096926e-05,
|
| 1101 |
+
"loss": 0.2385,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.9639293937068304,
|
| 1106 |
+
"grad_norm": 0.30176666378974915,
|
| 1107 |
+
"learning_rate": 3.276252019386107e-05,
|
| 1108 |
+
"loss": 0.2424,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.9700690713737529,
|
| 1113 |
+
"grad_norm": 0.2967808246612549,
|
| 1114 |
+
"learning_rate": 3.2568659127625205e-05,
|
| 1115 |
+
"loss": 0.2403,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.9762087490406753,
|
| 1120 |
+
"grad_norm": 0.287546843290329,
|
| 1121 |
+
"learning_rate": 3.237479806138934e-05,
|
| 1122 |
+
"loss": 0.2402,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.9823484267075978,
|
| 1127 |
+
"grad_norm": 0.32006338238716125,
|
| 1128 |
+
"learning_rate": 3.218093699515347e-05,
|
| 1129 |
+
"loss": 0.2416,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.9884881043745203,
|
| 1134 |
+
"grad_norm": 0.31201934814453125,
|
| 1135 |
+
"learning_rate": 3.198707592891761e-05,
|
| 1136 |
+
"loss": 0.2396,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.9946277820414429,
|
| 1141 |
+
"grad_norm": 0.34285658597946167,
|
| 1142 |
+
"learning_rate": 3.179321486268174e-05,
|
| 1143 |
+
"loss": 0.2411,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.0012279355333844,
|
| 1148 |
+
"grad_norm": 0.2902677655220032,
|
| 1149 |
+
"learning_rate": 3.159935379644588e-05,
|
| 1150 |
+
"loss": 0.2588,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.0073676132003069,
|
| 1155 |
+
"grad_norm": 0.3334469497203827,
|
| 1156 |
+
"learning_rate": 3.140549273021002e-05,
|
| 1157 |
+
"loss": 0.2412,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.0135072908672296,
|
| 1162 |
+
"grad_norm": 0.34331703186035156,
|
| 1163 |
+
"learning_rate": 3.1211631663974154e-05,
|
| 1164 |
+
"loss": 0.2376,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.019646968534152,
|
| 1169 |
+
"grad_norm": 0.30712443590164185,
|
| 1170 |
+
"learning_rate": 3.101777059773829e-05,
|
| 1171 |
+
"loss": 0.2382,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.0257866462010745,
|
| 1176 |
+
"grad_norm": 0.29111242294311523,
|
| 1177 |
+
"learning_rate": 3.082390953150243e-05,
|
| 1178 |
+
"loss": 0.2396,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.031926323867997,
|
| 1183 |
+
"grad_norm": 0.30055105686187744,
|
| 1184 |
+
"learning_rate": 3.063004846526656e-05,
|
| 1185 |
+
"loss": 0.2374,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.0380660015349195,
|
| 1190 |
+
"grad_norm": 0.2835744321346283,
|
| 1191 |
+
"learning_rate": 3.0436187399030692e-05,
|
| 1192 |
+
"loss": 0.2391,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.044205679201842,
|
| 1197 |
+
"grad_norm": 0.3579924404621124,
|
| 1198 |
+
"learning_rate": 3.0242326332794835e-05,
|
| 1199 |
+
"loss": 0.2366,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.0503453568687644,
|
| 1204 |
+
"grad_norm": 0.31649139523506165,
|
| 1205 |
+
"learning_rate": 3.0048465266558968e-05,
|
| 1206 |
+
"loss": 0.2386,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.0564850345356869,
|
| 1211 |
+
"grad_norm": 0.3274291157722473,
|
| 1212 |
+
"learning_rate": 2.9854604200323104e-05,
|
| 1213 |
+
"loss": 0.2367,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.0626247122026093,
|
| 1218 |
+
"grad_norm": 0.35450875759124756,
|
| 1219 |
+
"learning_rate": 2.966074313408724e-05,
|
| 1220 |
+
"loss": 0.2358,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.0687643898695318,
|
| 1225 |
+
"grad_norm": 0.31164485216140747,
|
| 1226 |
+
"learning_rate": 2.9466882067851372e-05,
|
| 1227 |
+
"loss": 0.2373,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.0749040675364543,
|
| 1232 |
+
"grad_norm": 0.29177117347717285,
|
| 1233 |
+
"learning_rate": 2.9273021001615512e-05,
|
| 1234 |
+
"loss": 0.2368,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.0810437452033768,
|
| 1239 |
+
"grad_norm": 0.31044426560401917,
|
| 1240 |
+
"learning_rate": 2.9079159935379645e-05,
|
| 1241 |
+
"loss": 0.236,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.0871834228702992,
|
| 1246 |
+
"grad_norm": 0.30381840467453003,
|
| 1247 |
+
"learning_rate": 2.888529886914378e-05,
|
| 1248 |
+
"loss": 0.2355,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.0933231005372217,
|
| 1253 |
+
"grad_norm": 0.3097009062767029,
|
| 1254 |
+
"learning_rate": 2.8691437802907917e-05,
|
| 1255 |
+
"loss": 0.2338,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.0994627782041442,
|
| 1260 |
+
"grad_norm": 0.2648329436779022,
|
| 1261 |
+
"learning_rate": 2.8497576736672053e-05,
|
| 1262 |
+
"loss": 0.2383,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.1056024558710669,
|
| 1267 |
+
"grad_norm": 0.29117581248283386,
|
| 1268 |
+
"learning_rate": 2.830371567043619e-05,
|
| 1269 |
+
"loss": 0.2358,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.1117421335379893,
|
| 1274 |
+
"grad_norm": 0.3057555556297302,
|
| 1275 |
+
"learning_rate": 2.810985460420032e-05,
|
| 1276 |
+
"loss": 0.2362,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.1178818112049118,
|
| 1281 |
+
"grad_norm": 0.27993032336235046,
|
| 1282 |
+
"learning_rate": 2.791599353796446e-05,
|
| 1283 |
+
"loss": 0.2354,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.1240214888718343,
|
| 1288 |
+
"grad_norm": 0.2994190752506256,
|
| 1289 |
+
"learning_rate": 2.7722132471728594e-05,
|
| 1290 |
+
"loss": 0.2345,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.1301611665387568,
|
| 1295 |
+
"grad_norm": 0.2987305223941803,
|
| 1296 |
+
"learning_rate": 2.752827140549273e-05,
|
| 1297 |
+
"loss": 0.2348,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.1363008442056792,
|
| 1302 |
+
"grad_norm": 0.3114520013332367,
|
| 1303 |
+
"learning_rate": 2.733441033925687e-05,
|
| 1304 |
+
"loss": 0.2362,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.1424405218726017,
|
| 1309 |
+
"grad_norm": 0.28420695662498474,
|
| 1310 |
+
"learning_rate": 2.7140549273021002e-05,
|
| 1311 |
+
"loss": 0.2371,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.1485801995395242,
|
| 1316 |
+
"grad_norm": 0.30027127265930176,
|
| 1317 |
+
"learning_rate": 2.6946688206785138e-05,
|
| 1318 |
+
"loss": 0.2387,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.1547198772064466,
|
| 1323 |
+
"grad_norm": 0.294156551361084,
|
| 1324 |
+
"learning_rate": 2.6752827140549274e-05,
|
| 1325 |
+
"loss": 0.2372,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.1608595548733691,
|
| 1330 |
+
"grad_norm": 0.3122671842575073,
|
| 1331 |
+
"learning_rate": 2.655896607431341e-05,
|
| 1332 |
+
"loss": 0.2358,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.1669992325402916,
|
| 1337 |
+
"grad_norm": 0.3033040761947632,
|
| 1338 |
+
"learning_rate": 2.6365105008077543e-05,
|
| 1339 |
+
"loss": 0.2394,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.173138910207214,
|
| 1344 |
+
"grad_norm": 0.32276102900505066,
|
| 1345 |
+
"learning_rate": 2.617124394184168e-05,
|
| 1346 |
+
"loss": 0.2321,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.1792785878741365,
|
| 1351 |
+
"grad_norm": 0.29448068141937256,
|
| 1352 |
+
"learning_rate": 2.597738287560582e-05,
|
| 1353 |
+
"loss": 0.2357,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.185418265541059,
|
| 1358 |
+
"grad_norm": 0.3151733875274658,
|
| 1359 |
+
"learning_rate": 2.578352180936995e-05,
|
| 1360 |
+
"loss": 0.2341,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.1915579432079815,
|
| 1365 |
+
"grad_norm": 0.28910940885543823,
|
| 1366 |
+
"learning_rate": 2.5589660743134087e-05,
|
| 1367 |
+
"loss": 0.2338,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.1976976208749042,
|
| 1372 |
+
"grad_norm": 0.3199194371700287,
|
| 1373 |
+
"learning_rate": 2.5395799676898223e-05,
|
| 1374 |
+
"loss": 0.2366,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.2038372985418266,
|
| 1379 |
+
"grad_norm": 0.31363335251808167,
|
| 1380 |
+
"learning_rate": 2.520193861066236e-05,
|
| 1381 |
+
"loss": 0.2333,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.209976976208749,
|
| 1386 |
+
"grad_norm": 0.27689889073371887,
|
| 1387 |
+
"learning_rate": 2.5008077544426496e-05,
|
| 1388 |
+
"loss": 0.232,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.2161166538756716,
|
| 1393 |
+
"grad_norm": 0.27902254462242126,
|
| 1394 |
+
"learning_rate": 2.4814216478190632e-05,
|
| 1395 |
+
"loss": 0.231,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.222256331542594,
|
| 1400 |
+
"grad_norm": 0.29046565294265747,
|
| 1401 |
+
"learning_rate": 2.4620355411954768e-05,
|
| 1402 |
+
"loss": 0.2343,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.2283960092095165,
|
| 1407 |
+
"grad_norm": 0.3037092387676239,
|
| 1408 |
+
"learning_rate": 2.44264943457189e-05,
|
| 1409 |
+
"loss": 0.2359,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.234535686876439,
|
| 1414 |
+
"grad_norm": 0.2904636561870575,
|
| 1415 |
+
"learning_rate": 2.4232633279483037e-05,
|
| 1416 |
+
"loss": 0.2351,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.2406753645433615,
|
| 1421 |
+
"grad_norm": 0.31070101261138916,
|
| 1422 |
+
"learning_rate": 2.4038772213247173e-05,
|
| 1423 |
+
"loss": 0.2332,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.246815042210284,
|
| 1428 |
+
"grad_norm": 0.2760045528411865,
|
| 1429 |
+
"learning_rate": 2.384491114701131e-05,
|
| 1430 |
+
"loss": 0.2344,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.2529547198772064,
|
| 1435 |
+
"grad_norm": 0.3140803575515747,
|
| 1436 |
+
"learning_rate": 2.3651050080775445e-05,
|
| 1437 |
+
"loss": 0.2381,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.2590943975441289,
|
| 1442 |
+
"grad_norm": 0.30985718965530396,
|
| 1443 |
+
"learning_rate": 2.345718901453958e-05,
|
| 1444 |
+
"loss": 0.2355,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.2652340752110514,
|
| 1449 |
+
"grad_norm": 0.2748771905899048,
|
| 1450 |
+
"learning_rate": 2.3263327948303717e-05,
|
| 1451 |
+
"loss": 0.2343,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.2713737528779738,
|
| 1456 |
+
"grad_norm": 0.2836357057094574,
|
| 1457 |
+
"learning_rate": 2.306946688206785e-05,
|
| 1458 |
+
"loss": 0.2355,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.2775134305448965,
|
| 1463 |
+
"grad_norm": 0.27377593517303467,
|
| 1464 |
+
"learning_rate": 2.287560581583199e-05,
|
| 1465 |
+
"loss": 0.2348,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.2836531082118188,
|
| 1470 |
+
"grad_norm": 0.29107093811035156,
|
| 1471 |
+
"learning_rate": 2.2681744749596125e-05,
|
| 1472 |
+
"loss": 0.2336,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.2897927858787415,
|
| 1477 |
+
"grad_norm": 0.29846909642219543,
|
| 1478 |
+
"learning_rate": 2.2487883683360258e-05,
|
| 1479 |
+
"loss": 0.2344,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.2959324635456637,
|
| 1484 |
+
"grad_norm": 0.3171651065349579,
|
| 1485 |
+
"learning_rate": 2.2294022617124394e-05,
|
| 1486 |
+
"loss": 0.2325,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.3020721412125864,
|
| 1491 |
+
"grad_norm": 0.3063752353191376,
|
| 1492 |
+
"learning_rate": 2.210016155088853e-05,
|
| 1493 |
+
"loss": 0.233,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.3082118188795089,
|
| 1498 |
+
"grad_norm": 0.32770147919654846,
|
| 1499 |
+
"learning_rate": 2.1906300484652666e-05,
|
| 1500 |
+
"loss": 0.2329,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.3143514965464314,
|
| 1505 |
+
"grad_norm": 0.2971561849117279,
|
| 1506 |
+
"learning_rate": 2.1712439418416802e-05,
|
| 1507 |
+
"loss": 0.234,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.3204911742133538,
|
| 1512 |
+
"grad_norm": 0.2993122935295105,
|
| 1513 |
+
"learning_rate": 2.151857835218094e-05,
|
| 1514 |
+
"loss": 0.2343,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.3266308518802763,
|
| 1519 |
+
"grad_norm": 0.2786669135093689,
|
| 1520 |
+
"learning_rate": 2.1324717285945075e-05,
|
| 1521 |
+
"loss": 0.2317,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.3327705295471988,
|
| 1526 |
+
"grad_norm": 0.284007728099823,
|
| 1527 |
+
"learning_rate": 2.1130856219709207e-05,
|
| 1528 |
+
"loss": 0.2313,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.3389102072141212,
|
| 1533 |
+
"grad_norm": 0.29710453748703003,
|
| 1534 |
+
"learning_rate": 2.0936995153473347e-05,
|
| 1535 |
+
"loss": 0.234,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.3450498848810437,
|
| 1540 |
+
"grad_norm": 0.28225216269493103,
|
| 1541 |
+
"learning_rate": 2.074313408723748e-05,
|
| 1542 |
+
"loss": 0.2346,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.3511895625479662,
|
| 1547 |
+
"grad_norm": 0.30834057927131653,
|
| 1548 |
+
"learning_rate": 2.0549273021001615e-05,
|
| 1549 |
+
"loss": 0.2298,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.3573292402148887,
|
| 1554 |
+
"grad_norm": 0.2734493911266327,
|
| 1555 |
+
"learning_rate": 2.035541195476575e-05,
|
| 1556 |
+
"loss": 0.231,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.3634689178818111,
|
| 1561 |
+
"grad_norm": 0.3022773265838623,
|
| 1562 |
+
"learning_rate": 2.0161550888529888e-05,
|
| 1563 |
+
"loss": 0.2319,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.3696085955487338,
|
| 1568 |
+
"grad_norm": 0.2787306606769562,
|
| 1569 |
+
"learning_rate": 1.9967689822294024e-05,
|
| 1570 |
+
"loss": 0.2327,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.375748273215656,
|
| 1575 |
+
"grad_norm": 0.27380722761154175,
|
| 1576 |
+
"learning_rate": 1.9773828756058156e-05,
|
| 1577 |
+
"loss": 0.2309,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.3818879508825788,
|
| 1582 |
+
"grad_norm": 0.26936137676239014,
|
| 1583 |
+
"learning_rate": 1.9579967689822296e-05,
|
| 1584 |
+
"loss": 0.2301,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.388027628549501,
|
| 1589 |
+
"grad_norm": 0.27485784888267517,
|
| 1590 |
+
"learning_rate": 1.9386106623586432e-05,
|
| 1591 |
+
"loss": 0.2313,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.3941673062164237,
|
| 1596 |
+
"grad_norm": 0.2874917685985565,
|
| 1597 |
+
"learning_rate": 1.9192245557350565e-05,
|
| 1598 |
+
"loss": 0.2301,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.4003069838833462,
|
| 1603 |
+
"grad_norm": 0.26473861932754517,
|
| 1604 |
+
"learning_rate": 1.8998384491114704e-05,
|
| 1605 |
+
"loss": 0.2302,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.4064466615502687,
|
| 1610 |
+
"grad_norm": 0.28767651319503784,
|
| 1611 |
+
"learning_rate": 1.8804523424878837e-05,
|
| 1612 |
+
"loss": 0.2321,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.4125863392171911,
|
| 1617 |
+
"grad_norm": 0.2758727967739105,
|
| 1618 |
+
"learning_rate": 1.8610662358642973e-05,
|
| 1619 |
+
"loss": 0.2344,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.4187260168841136,
|
| 1624 |
+
"grad_norm": 0.27313369512557983,
|
| 1625 |
+
"learning_rate": 1.8416801292407106e-05,
|
| 1626 |
+
"loss": 0.2321,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.424865694551036,
|
| 1631 |
+
"grad_norm": 0.2732943594455719,
|
| 1632 |
+
"learning_rate": 1.8222940226171245e-05,
|
| 1633 |
+
"loss": 0.2346,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.4310053722179585,
|
| 1638 |
+
"grad_norm": 0.28029924631118774,
|
| 1639 |
+
"learning_rate": 1.802907915993538e-05,
|
| 1640 |
+
"loss": 0.2306,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.437145049884881,
|
| 1645 |
+
"grad_norm": 0.30662721395492554,
|
| 1646 |
+
"learning_rate": 1.7835218093699514e-05,
|
| 1647 |
+
"loss": 0.2303,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.4432847275518035,
|
| 1652 |
+
"grad_norm": 0.2687036991119385,
|
| 1653 |
+
"learning_rate": 1.7641357027463653e-05,
|
| 1654 |
+
"loss": 0.2305,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.449424405218726,
|
| 1659 |
+
"grad_norm": 0.27795106172561646,
|
| 1660 |
+
"learning_rate": 1.7447495961227786e-05,
|
| 1661 |
+
"loss": 0.2311,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.4555640828856484,
|
| 1666 |
+
"grad_norm": 0.27534544467926025,
|
| 1667 |
+
"learning_rate": 1.7253634894991922e-05,
|
| 1668 |
+
"loss": 0.2295,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.4617037605525711,
|
| 1673 |
+
"grad_norm": 0.2723388671875,
|
| 1674 |
+
"learning_rate": 1.705977382875606e-05,
|
| 1675 |
+
"loss": 0.2332,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.4678434382194934,
|
| 1680 |
+
"grad_norm": 0.2775721251964569,
|
| 1681 |
+
"learning_rate": 1.6865912762520194e-05,
|
| 1682 |
+
"loss": 0.2287,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.473983115886416,
|
| 1687 |
+
"grad_norm": 0.2654743194580078,
|
| 1688 |
+
"learning_rate": 1.667205169628433e-05,
|
| 1689 |
+
"loss": 0.2315,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.4801227935533383,
|
| 1694 |
+
"grad_norm": 0.29545167088508606,
|
| 1695 |
+
"learning_rate": 1.6478190630048463e-05,
|
| 1696 |
+
"loss": 0.2324,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.486262471220261,
|
| 1701 |
+
"grad_norm": 0.27187272906303406,
|
| 1702 |
+
"learning_rate": 1.6284329563812603e-05,
|
| 1703 |
+
"loss": 0.2319,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.4924021488871835,
|
| 1708 |
+
"grad_norm": 0.2794046401977539,
|
| 1709 |
+
"learning_rate": 1.6090468497576735e-05,
|
| 1710 |
+
"loss": 0.2289,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.498541826554106,
|
| 1715 |
+
"grad_norm": 0.2789725959300995,
|
| 1716 |
+
"learning_rate": 1.589660743134087e-05,
|
| 1717 |
+
"loss": 0.2319,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.5046815042210284,
|
| 1722 |
+
"grad_norm": 0.2820490598678589,
|
| 1723 |
+
"learning_rate": 1.570274636510501e-05,
|
| 1724 |
+
"loss": 0.2305,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.510821181887951,
|
| 1729 |
+
"grad_norm": 0.292161762714386,
|
| 1730 |
+
"learning_rate": 1.5508885298869144e-05,
|
| 1731 |
+
"loss": 0.2311,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 1.5169608595548734,
|
| 1736 |
+
"grad_norm": 0.28120511770248413,
|
| 1737 |
+
"learning_rate": 1.531502423263328e-05,
|
| 1738 |
+
"loss": 0.2268,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.5231005372217958,
|
| 1743 |
+
"grad_norm": 0.26915067434310913,
|
| 1744 |
+
"learning_rate": 1.5121163166397417e-05,
|
| 1745 |
+
"loss": 0.2278,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.5292402148887183,
|
| 1750 |
+
"grad_norm": 0.2905220091342926,
|
| 1751 |
+
"learning_rate": 1.4927302100161552e-05,
|
| 1752 |
+
"loss": 0.2313,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 1.5353798925556408,
|
| 1757 |
+
"grad_norm": 0.2794709801673889,
|
| 1758 |
+
"learning_rate": 1.4733441033925686e-05,
|
| 1759 |
+
"loss": 0.2294,
|
| 1760 |
+
"step": 2500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 1.5415195702225635,
|
| 1764 |
+
"grad_norm": 0.2608950734138489,
|
| 1765 |
+
"learning_rate": 1.4539579967689822e-05,
|
| 1766 |
+
"loss": 0.2288,
|
| 1767 |
+
"step": 2510
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.5476592478894857,
|
| 1771 |
+
"grad_norm": 0.25274738669395447,
|
| 1772 |
+
"learning_rate": 1.4345718901453958e-05,
|
| 1773 |
+
"loss": 0.2252,
|
| 1774 |
+
"step": 2520
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 1.5537989255564084,
|
| 1778 |
+
"grad_norm": 0.2868223786354065,
|
| 1779 |
+
"learning_rate": 1.4151857835218094e-05,
|
| 1780 |
+
"loss": 0.2297,
|
| 1781 |
+
"step": 2530
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.5599386032233307,
|
| 1785 |
+
"grad_norm": 0.2765255272388458,
|
| 1786 |
+
"learning_rate": 1.395799676898223e-05,
|
| 1787 |
+
"loss": 0.2295,
|
| 1788 |
+
"step": 2540
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.5660782808902534,
|
| 1792 |
+
"grad_norm": 0.23713639378547668,
|
| 1793 |
+
"learning_rate": 1.3764135702746365e-05,
|
| 1794 |
+
"loss": 0.229,
|
| 1795 |
+
"step": 2550
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 1.5722179585571756,
|
| 1799 |
+
"grad_norm": 0.2997082471847534,
|
| 1800 |
+
"learning_rate": 1.3570274636510501e-05,
|
| 1801 |
+
"loss": 0.227,
|
| 1802 |
+
"step": 2560
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 1.5783576362240983,
|
| 1806 |
+
"grad_norm": 0.32743728160858154,
|
| 1807 |
+
"learning_rate": 1.3376413570274637e-05,
|
| 1808 |
+
"loss": 0.2277,
|
| 1809 |
+
"step": 2570
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 1.5844973138910208,
|
| 1813 |
+
"grad_norm": 0.2513977885246277,
|
| 1814 |
+
"learning_rate": 1.3182552504038772e-05,
|
| 1815 |
+
"loss": 0.2285,
|
| 1816 |
+
"step": 2580
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 1.5906369915579432,
|
| 1820 |
+
"grad_norm": 0.2621912956237793,
|
| 1821 |
+
"learning_rate": 1.298869143780291e-05,
|
| 1822 |
+
"loss": 0.231,
|
| 1823 |
+
"step": 2590
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 1.5967766692248657,
|
| 1827 |
+
"grad_norm": 0.2549593448638916,
|
| 1828 |
+
"learning_rate": 1.2794830371567044e-05,
|
| 1829 |
+
"loss": 0.2285,
|
| 1830 |
+
"step": 2600
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 1.6029163468917882,
|
| 1834 |
+
"grad_norm": 0.25694531202316284,
|
| 1835 |
+
"learning_rate": 1.260096930533118e-05,
|
| 1836 |
+
"loss": 0.229,
|
| 1837 |
+
"step": 2610
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 1.6090560245587107,
|
| 1841 |
+
"grad_norm": 0.278401643037796,
|
| 1842 |
+
"learning_rate": 1.2407108239095316e-05,
|
| 1843 |
+
"loss": 0.231,
|
| 1844 |
+
"step": 2620
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 1.6151957022256331,
|
| 1848 |
+
"grad_norm": 0.30101412534713745,
|
| 1849 |
+
"learning_rate": 1.221324717285945e-05,
|
| 1850 |
+
"loss": 0.2306,
|
| 1851 |
+
"step": 2630
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.6213353798925556,
|
| 1855 |
+
"grad_norm": 0.25657519698143005,
|
| 1856 |
+
"learning_rate": 1.2019386106623586e-05,
|
| 1857 |
+
"loss": 0.2289,
|
| 1858 |
+
"step": 2640
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 1.627475057559478,
|
| 1862 |
+
"grad_norm": 0.2672490179538727,
|
| 1863 |
+
"learning_rate": 1.1825525040387722e-05,
|
| 1864 |
+
"loss": 0.2278,
|
| 1865 |
+
"step": 2650
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 1.6336147352264008,
|
| 1869 |
+
"grad_norm": 0.269938588142395,
|
| 1870 |
+
"learning_rate": 1.1631663974151859e-05,
|
| 1871 |
+
"loss": 0.2268,
|
| 1872 |
+
"step": 2660
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 1.639754412893323,
|
| 1876 |
+
"grad_norm": 0.27414554357528687,
|
| 1877 |
+
"learning_rate": 1.1437802907915995e-05,
|
| 1878 |
+
"loss": 0.2292,
|
| 1879 |
+
"step": 2670
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 1.6458940905602457,
|
| 1883 |
+
"grad_norm": 0.2716943323612213,
|
| 1884 |
+
"learning_rate": 1.1243941841680129e-05,
|
| 1885 |
+
"loss": 0.2257,
|
| 1886 |
+
"step": 2680
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 1.652033768227168,
|
| 1890 |
+
"grad_norm": 0.29680389165878296,
|
| 1891 |
+
"learning_rate": 1.1050080775444265e-05,
|
| 1892 |
+
"loss": 0.2285,
|
| 1893 |
+
"step": 2690
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 1.6581734458940907,
|
| 1897 |
+
"grad_norm": 0.25835490226745605,
|
| 1898 |
+
"learning_rate": 1.0856219709208401e-05,
|
| 1899 |
+
"loss": 0.2282,
|
| 1900 |
+
"step": 2700
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 1.664313123561013,
|
| 1904 |
+
"grad_norm": 0.2611675560474396,
|
| 1905 |
+
"learning_rate": 1.0662358642972537e-05,
|
| 1906 |
+
"loss": 0.2265,
|
| 1907 |
+
"step": 2710
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 1.6704528012279356,
|
| 1911 |
+
"grad_norm": 0.2636869251728058,
|
| 1912 |
+
"learning_rate": 1.0468497576736673e-05,
|
| 1913 |
+
"loss": 0.2262,
|
| 1914 |
+
"step": 2720
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 1.676592478894858,
|
| 1918 |
+
"grad_norm": 0.276288777589798,
|
| 1919 |
+
"learning_rate": 1.0274636510500808e-05,
|
| 1920 |
+
"loss": 0.2299,
|
| 1921 |
+
"step": 2730
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 1.6827321565617805,
|
| 1925 |
+
"grad_norm": 0.24938946962356567,
|
| 1926 |
+
"learning_rate": 1.0080775444264944e-05,
|
| 1927 |
+
"loss": 0.2268,
|
| 1928 |
+
"step": 2740
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 1.688871834228703,
|
| 1932 |
+
"grad_norm": 0.24695712327957153,
|
| 1933 |
+
"learning_rate": 9.886914378029078e-06,
|
| 1934 |
+
"loss": 0.2282,
|
| 1935 |
+
"step": 2750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 1.6950115118956255,
|
| 1939 |
+
"grad_norm": 0.25521910190582275,
|
| 1940 |
+
"learning_rate": 9.693053311793216e-06,
|
| 1941 |
+
"loss": 0.2275,
|
| 1942 |
+
"step": 2760
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 1.701151189562548,
|
| 1946 |
+
"grad_norm": 0.23849350214004517,
|
| 1947 |
+
"learning_rate": 9.499192245557352e-06,
|
| 1948 |
+
"loss": 0.2274,
|
| 1949 |
+
"step": 2770
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 1.7072908672294704,
|
| 1953 |
+
"grad_norm": 0.2584358751773834,
|
| 1954 |
+
"learning_rate": 9.305331179321486e-06,
|
| 1955 |
+
"loss": 0.2309,
|
| 1956 |
+
"step": 2780
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 1.713430544896393,
|
| 1960 |
+
"grad_norm": 0.25768372416496277,
|
| 1961 |
+
"learning_rate": 9.111470113085623e-06,
|
| 1962 |
+
"loss": 0.2293,
|
| 1963 |
+
"step": 2790
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 1.7195702225633154,
|
| 1967 |
+
"grad_norm": 0.2615247070789337,
|
| 1968 |
+
"learning_rate": 8.917609046849757e-06,
|
| 1969 |
+
"loss": 0.225,
|
| 1970 |
+
"step": 2800
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 1.725709900230238,
|
| 1974 |
+
"grad_norm": 0.2701919674873352,
|
| 1975 |
+
"learning_rate": 8.723747980613893e-06,
|
| 1976 |
+
"loss": 0.226,
|
| 1977 |
+
"step": 2810
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 1.7318495778971603,
|
| 1981 |
+
"grad_norm": 0.25696951150894165,
|
| 1982 |
+
"learning_rate": 8.52988691437803e-06,
|
| 1983 |
+
"loss": 0.2273,
|
| 1984 |
+
"step": 2820
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 1.737989255564083,
|
| 1988 |
+
"grad_norm": 0.2802889049053192,
|
| 1989 |
+
"learning_rate": 8.336025848142165e-06,
|
| 1990 |
+
"loss": 0.2269,
|
| 1991 |
+
"step": 2830
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 1.7441289332310053,
|
| 1995 |
+
"grad_norm": 0.24956634640693665,
|
| 1996 |
+
"learning_rate": 8.142164781906301e-06,
|
| 1997 |
+
"loss": 0.2274,
|
| 1998 |
+
"step": 2840
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 1.750268610897928,
|
| 2002 |
+
"grad_norm": 0.2676002085208893,
|
| 2003 |
+
"learning_rate": 7.948303715670436e-06,
|
| 2004 |
+
"loss": 0.2264,
|
| 2005 |
+
"step": 2850
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 1.7564082885648502,
|
| 2009 |
+
"grad_norm": 0.23091745376586914,
|
| 2010 |
+
"learning_rate": 7.754442649434572e-06,
|
| 2011 |
+
"loss": 0.2284,
|
| 2012 |
+
"step": 2860
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 1.762547966231773,
|
| 2016 |
+
"grad_norm": 0.2684544324874878,
|
| 2017 |
+
"learning_rate": 7.560581583198709e-06,
|
| 2018 |
+
"loss": 0.2259,
|
| 2019 |
+
"step": 2870
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 1.7686876438986954,
|
| 2023 |
+
"grad_norm": 0.2851001024246216,
|
| 2024 |
+
"learning_rate": 7.366720516962843e-06,
|
| 2025 |
+
"loss": 0.2266,
|
| 2026 |
+
"step": 2880
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 1.7748273215656178,
|
| 2030 |
+
"grad_norm": 0.25915250182151794,
|
| 2031 |
+
"learning_rate": 7.172859450726979e-06,
|
| 2032 |
+
"loss": 0.2287,
|
| 2033 |
+
"step": 2890
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 1.7809669992325403,
|
| 2037 |
+
"grad_norm": 0.24940845370292664,
|
| 2038 |
+
"learning_rate": 6.978998384491115e-06,
|
| 2039 |
+
"loss": 0.2275,
|
| 2040 |
+
"step": 2900
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 1.7871066768994628,
|
| 2044 |
+
"grad_norm": 0.2588519752025604,
|
| 2045 |
+
"learning_rate": 6.7851373182552505e-06,
|
| 2046 |
+
"loss": 0.229,
|
| 2047 |
+
"step": 2910
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 1.7932463545663853,
|
| 2051 |
+
"grad_norm": 0.24955865740776062,
|
| 2052 |
+
"learning_rate": 6.591276252019386e-06,
|
| 2053 |
+
"loss": 0.2247,
|
| 2054 |
+
"step": 2920
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 1.7993860322333077,
|
| 2058 |
+
"grad_norm": 0.26434391736984253,
|
| 2059 |
+
"learning_rate": 6.397415185783522e-06,
|
| 2060 |
+
"loss": 0.2269,
|
| 2061 |
+
"step": 2930
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 1.8055257099002302,
|
| 2065 |
+
"grad_norm": 0.2523840069770813,
|
| 2066 |
+
"learning_rate": 6.203554119547658e-06,
|
| 2067 |
+
"loss": 0.2255,
|
| 2068 |
+
"step": 2940
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 1.8116653875671527,
|
| 2072 |
+
"grad_norm": 0.26612043380737305,
|
| 2073 |
+
"learning_rate": 6.009693053311793e-06,
|
| 2074 |
+
"loss": 0.2255,
|
| 2075 |
+
"step": 2950
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 1.8178050652340754,
|
| 2079 |
+
"grad_norm": 0.24991387128829956,
|
| 2080 |
+
"learning_rate": 5.815831987075929e-06,
|
| 2081 |
+
"loss": 0.2267,
|
| 2082 |
+
"step": 2960
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 1.8239447429009976,
|
| 2086 |
+
"grad_norm": 0.25838539004325867,
|
| 2087 |
+
"learning_rate": 5.6219709208400645e-06,
|
| 2088 |
+
"loss": 0.228,
|
| 2089 |
+
"step": 2970
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 1.8300844205679203,
|
| 2093 |
+
"grad_norm": 0.25465282797813416,
|
| 2094 |
+
"learning_rate": 5.428109854604201e-06,
|
| 2095 |
+
"loss": 0.2272,
|
| 2096 |
+
"step": 2980
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 1.8362240982348426,
|
| 2100 |
+
"grad_norm": 0.25871556997299194,
|
| 2101 |
+
"learning_rate": 5.234248788368337e-06,
|
| 2102 |
+
"loss": 0.2278,
|
| 2103 |
+
"step": 2990
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 1.8423637759017653,
|
| 2107 |
+
"grad_norm": 0.24951577186584473,
|
| 2108 |
+
"learning_rate": 5.040387722132472e-06,
|
| 2109 |
+
"loss": 0.2285,
|
| 2110 |
+
"step": 3000
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 1.8485034535686875,
|
| 2114 |
+
"grad_norm": 0.25530776381492615,
|
| 2115 |
+
"learning_rate": 4.846526655896608e-06,
|
| 2116 |
+
"loss": 0.2242,
|
| 2117 |
+
"step": 3010
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 1.8546431312356102,
|
| 2121 |
+
"grad_norm": 0.24456080794334412,
|
| 2122 |
+
"learning_rate": 4.652665589660743e-06,
|
| 2123 |
+
"loss": 0.2277,
|
| 2124 |
+
"step": 3020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 1.8607828089025327,
|
| 2128 |
+
"grad_norm": 0.24893827736377716,
|
| 2129 |
+
"learning_rate": 4.4588045234248785e-06,
|
| 2130 |
+
"loss": 0.2251,
|
| 2131 |
+
"step": 3030
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 1.8669224865694551,
|
| 2135 |
+
"grad_norm": 0.2683740258216858,
|
| 2136 |
+
"learning_rate": 4.264943457189015e-06,
|
| 2137 |
+
"loss": 0.2263,
|
| 2138 |
+
"step": 3040
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 1.8730621642363776,
|
| 2142 |
+
"grad_norm": 0.2723815143108368,
|
| 2143 |
+
"learning_rate": 4.071082390953151e-06,
|
| 2144 |
+
"loss": 0.2249,
|
| 2145 |
+
"step": 3050
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 1.8792018419033,
|
| 2149 |
+
"grad_norm": 0.23197263479232788,
|
| 2150 |
+
"learning_rate": 3.877221324717286e-06,
|
| 2151 |
+
"loss": 0.2278,
|
| 2152 |
+
"step": 3060
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 1.8853415195702226,
|
| 2156 |
+
"grad_norm": 0.24273467063903809,
|
| 2157 |
+
"learning_rate": 3.6833602584814216e-06,
|
| 2158 |
+
"loss": 0.2256,
|
| 2159 |
+
"step": 3070
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 1.891481197237145,
|
| 2163 |
+
"grad_norm": 0.26177576184272766,
|
| 2164 |
+
"learning_rate": 3.4894991922455576e-06,
|
| 2165 |
+
"loss": 0.2264,
|
| 2166 |
+
"step": 3080
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 1.8976208749040675,
|
| 2170 |
+
"grad_norm": 0.2266533076763153,
|
| 2171 |
+
"learning_rate": 3.295638126009693e-06,
|
| 2172 |
+
"loss": 0.2242,
|
| 2173 |
+
"step": 3090
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 1.90376055257099,
|
| 2177 |
+
"grad_norm": 0.24498549103736877,
|
| 2178 |
+
"learning_rate": 3.101777059773829e-06,
|
| 2179 |
+
"loss": 0.2263,
|
| 2180 |
+
"step": 3100
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 1.9099002302379127,
|
| 2184 |
+
"grad_norm": 0.23406417667865753,
|
| 2185 |
+
"learning_rate": 2.9079159935379646e-06,
|
| 2186 |
+
"loss": 0.2243,
|
| 2187 |
+
"step": 3110
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 1.916039907904835,
|
| 2191 |
+
"grad_norm": 0.2536128759384155,
|
| 2192 |
+
"learning_rate": 2.7140549273021003e-06,
|
| 2193 |
+
"loss": 0.225,
|
| 2194 |
+
"step": 3120
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 1.9221795855717576,
|
| 2198 |
+
"grad_norm": 0.22612562775611877,
|
| 2199 |
+
"learning_rate": 2.520193861066236e-06,
|
| 2200 |
+
"loss": 0.2266,
|
| 2201 |
+
"step": 3130
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 1.9283192632386799,
|
| 2205 |
+
"grad_norm": 0.22622433304786682,
|
| 2206 |
+
"learning_rate": 2.3263327948303716e-06,
|
| 2207 |
+
"loss": 0.2258,
|
| 2208 |
+
"step": 3140
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 1.9344589409056026,
|
| 2212 |
+
"grad_norm": 0.2482767254114151,
|
| 2213 |
+
"learning_rate": 2.1324717285945077e-06,
|
| 2214 |
+
"loss": 0.2281,
|
| 2215 |
+
"step": 3150
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 1.9405986185725248,
|
| 2219 |
+
"grad_norm": 0.22862812876701355,
|
| 2220 |
+
"learning_rate": 1.938610662358643e-06,
|
| 2221 |
+
"loss": 0.2254,
|
| 2222 |
+
"step": 3160
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 1.9467382962394475,
|
| 2226 |
+
"grad_norm": 0.22068631649017334,
|
| 2227 |
+
"learning_rate": 1.7447495961227788e-06,
|
| 2228 |
+
"loss": 0.2242,
|
| 2229 |
+
"step": 3170
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 1.95287797390637,
|
| 2233 |
+
"grad_norm": 0.2349786013364792,
|
| 2234 |
+
"learning_rate": 1.5508885298869145e-06,
|
| 2235 |
+
"loss": 0.2281,
|
| 2236 |
+
"step": 3180
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 1.9590176515732924,
|
| 2240 |
+
"grad_norm": 0.24406956136226654,
|
| 2241 |
+
"learning_rate": 1.3570274636510501e-06,
|
| 2242 |
+
"loss": 0.2274,
|
| 2243 |
+
"step": 3190
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 1.965157329240215,
|
| 2247 |
+
"grad_norm": 0.2341851145029068,
|
| 2248 |
+
"learning_rate": 1.1631663974151858e-06,
|
| 2249 |
+
"loss": 0.2292,
|
| 2250 |
+
"step": 3200
|
| 2251 |
+
}
|
| 2252 |
+
],
|
| 2253 |
+
"logging_steps": 10,
|
| 2254 |
+
"max_steps": 3258,
|
| 2255 |
+
"num_input_tokens_seen": 0,
|
| 2256 |
+
"num_train_epochs": 2,
|
| 2257 |
+
"save_steps": 100,
|
| 2258 |
+
"stateful_callbacks": {
|
| 2259 |
+
"TrainerControl": {
|
| 2260 |
+
"args": {
|
| 2261 |
+
"should_epoch_stop": false,
|
| 2262 |
+
"should_evaluate": false,
|
| 2263 |
+
"should_log": false,
|
| 2264 |
+
"should_save": true,
|
| 2265 |
+
"should_training_stop": false
|
| 2266 |
+
},
|
| 2267 |
+
"attributes": {}
|
| 2268 |
+
}
|
| 2269 |
+
},
|
| 2270 |
+
"total_flos": 4.729291401165181e+20,
|
| 2271 |
+
"train_batch_size": 16,
|
| 2272 |
+
"trial_name": null,
|
| 2273 |
+
"trial_params": null
|
| 2274 |
+
}
|