Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2226 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2226 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.9975990396158463,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 312,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.006402561024409764,
|
| 13 |
+
"grad_norm": 2.679521906145934,
|
| 14 |
+
"learning_rate": 6.25e-08,
|
| 15 |
+
"loss": 0.1981,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.012805122048819529,
|
| 20 |
+
"grad_norm": 2.5600704949176025,
|
| 21 |
+
"learning_rate": 1.25e-07,
|
| 22 |
+
"loss": 0.1782,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.01920768307322929,
|
| 27 |
+
"grad_norm": 2.8865232354148174,
|
| 28 |
+
"learning_rate": 1.875e-07,
|
| 29 |
+
"loss": 0.2347,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.025610244097639057,
|
| 34 |
+
"grad_norm": 2.76706815796981,
|
| 35 |
+
"learning_rate": 2.5e-07,
|
| 36 |
+
"loss": 0.2138,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.03201280512204882,
|
| 41 |
+
"grad_norm": 2.6835981054304963,
|
| 42 |
+
"learning_rate": 3.1249999999999997e-07,
|
| 43 |
+
"loss": 0.2395,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03841536614645858,
|
| 48 |
+
"grad_norm": 2.6104328133293877,
|
| 49 |
+
"learning_rate": 3.75e-07,
|
| 50 |
+
"loss": 0.2417,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.04481792717086835,
|
| 55 |
+
"grad_norm": 2.551735870154201,
|
| 56 |
+
"learning_rate": 4.375e-07,
|
| 57 |
+
"loss": 0.175,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.051220488195278115,
|
| 62 |
+
"grad_norm": 3.4086729277405725,
|
| 63 |
+
"learning_rate": 5e-07,
|
| 64 |
+
"loss": 0.1706,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.057623049219687875,
|
| 69 |
+
"grad_norm": 2.7571882436426587,
|
| 70 |
+
"learning_rate": 5.625e-07,
|
| 71 |
+
"loss": 0.2184,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.06402561024409764,
|
| 76 |
+
"grad_norm": 2.7694984961484717,
|
| 77 |
+
"learning_rate": 6.249999999999999e-07,
|
| 78 |
+
"loss": 0.2017,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.07042817126850741,
|
| 83 |
+
"grad_norm": 2.547784892526759,
|
| 84 |
+
"learning_rate": 6.875e-07,
|
| 85 |
+
"loss": 0.1911,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.07683073229291716,
|
| 90 |
+
"grad_norm": 2.666671570148531,
|
| 91 |
+
"learning_rate": 7.5e-07,
|
| 92 |
+
"loss": 0.2143,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.08323329331732693,
|
| 97 |
+
"grad_norm": 2.7605278136164912,
|
| 98 |
+
"learning_rate": 8.125e-07,
|
| 99 |
+
"loss": 0.2027,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.0896358543417367,
|
| 104 |
+
"grad_norm": 2.4621112010955866,
|
| 105 |
+
"learning_rate": 8.75e-07,
|
| 106 |
+
"loss": 0.2068,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.09603841536614646,
|
| 111 |
+
"grad_norm": 2.803025975251869,
|
| 112 |
+
"learning_rate": 9.374999999999999e-07,
|
| 113 |
+
"loss": 0.213,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.10244097639055623,
|
| 118 |
+
"grad_norm": 2.901342575515257,
|
| 119 |
+
"learning_rate": 1e-06,
|
| 120 |
+
"loss": 0.2106,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.10884353741496598,
|
| 125 |
+
"grad_norm": 2.7337203386461644,
|
| 126 |
+
"learning_rate": 1.0625e-06,
|
| 127 |
+
"loss": 0.2281,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.11524609843937575,
|
| 132 |
+
"grad_norm": 2.6481197250458193,
|
| 133 |
+
"learning_rate": 1.125e-06,
|
| 134 |
+
"loss": 0.1827,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.12164865946378552,
|
| 139 |
+
"grad_norm": 2.7639376607457162,
|
| 140 |
+
"learning_rate": 1.1874999999999999e-06,
|
| 141 |
+
"loss": 0.2159,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.12805122048819528,
|
| 146 |
+
"grad_norm": 2.7175927412952467,
|
| 147 |
+
"learning_rate": 1.2499999999999999e-06,
|
| 148 |
+
"loss": 0.1745,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.13445378151260504,
|
| 153 |
+
"grad_norm": 3.1021924663248384,
|
| 154 |
+
"learning_rate": 1.3125e-06,
|
| 155 |
+
"loss": 0.1935,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.14085634253701482,
|
| 160 |
+
"grad_norm": 3.1941312523935563,
|
| 161 |
+
"learning_rate": 1.375e-06,
|
| 162 |
+
"loss": 0.2053,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.14725890356142457,
|
| 167 |
+
"grad_norm": 2.500688058457321,
|
| 168 |
+
"learning_rate": 1.4375e-06,
|
| 169 |
+
"loss": 0.1699,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.15366146458583432,
|
| 174 |
+
"grad_norm": 3.417365189000316,
|
| 175 |
+
"learning_rate": 1.5e-06,
|
| 176 |
+
"loss": 0.2222,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.1600640256102441,
|
| 181 |
+
"grad_norm": 2.7988997309084125,
|
| 182 |
+
"learning_rate": 1.5624999999999999e-06,
|
| 183 |
+
"loss": 0.1855,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.16646658663465386,
|
| 188 |
+
"grad_norm": 3.057142160060086,
|
| 189 |
+
"learning_rate": 1.625e-06,
|
| 190 |
+
"loss": 0.1698,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.17286914765906364,
|
| 195 |
+
"grad_norm": 2.8282556781813772,
|
| 196 |
+
"learning_rate": 1.6875e-06,
|
| 197 |
+
"loss": 0.2047,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.1792717086834734,
|
| 202 |
+
"grad_norm": 2.7193234045232413,
|
| 203 |
+
"learning_rate": 1.75e-06,
|
| 204 |
+
"loss": 0.1979,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.18567426970788314,
|
| 209 |
+
"grad_norm": 3.520125861739441,
|
| 210 |
+
"learning_rate": 1.8125e-06,
|
| 211 |
+
"loss": 0.2214,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.19207683073229292,
|
| 216 |
+
"grad_norm": 3.017660845656798,
|
| 217 |
+
"learning_rate": 1.8749999999999998e-06,
|
| 218 |
+
"loss": 0.2119,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.19847939175670268,
|
| 223 |
+
"grad_norm": 3.4420544526065515,
|
| 224 |
+
"learning_rate": 1.9375e-06,
|
| 225 |
+
"loss": 0.1649,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.20488195278111246,
|
| 230 |
+
"grad_norm": 2.858393701736414,
|
| 231 |
+
"learning_rate": 2e-06,
|
| 232 |
+
"loss": 0.2038,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.2112845138055222,
|
| 237 |
+
"grad_norm": 3.6069506971510723,
|
| 238 |
+
"learning_rate": 1.9999370567547003e-06,
|
| 239 |
+
"loss": 0.2194,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.21768707482993196,
|
| 244 |
+
"grad_norm": 3.1830415852739415,
|
| 245 |
+
"learning_rate": 1.9997482349425066e-06,
|
| 246 |
+
"loss": 0.143,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.22408963585434175,
|
| 251 |
+
"grad_norm": 3.4429216757147545,
|
| 252 |
+
"learning_rate": 1.9994335583335335e-06,
|
| 253 |
+
"loss": 0.2026,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.2304921968787515,
|
| 258 |
+
"grad_norm": 3.5752393511399463,
|
| 259 |
+
"learning_rate": 1.9989930665413145e-06,
|
| 260 |
+
"loss": 0.2087,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.23689475790316125,
|
| 265 |
+
"grad_norm": 4.08665235099645,
|
| 266 |
+
"learning_rate": 1.9984268150178167e-06,
|
| 267 |
+
"loss": 0.196,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.24329731892757103,
|
| 272 |
+
"grad_norm": 3.979136486495609,
|
| 273 |
+
"learning_rate": 1.997734875046456e-06,
|
| 274 |
+
"loss": 0.1824,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.24969987995198079,
|
| 279 |
+
"grad_norm": 3.866519745374768,
|
| 280 |
+
"learning_rate": 1.996917333733128e-06,
|
| 281 |
+
"loss": 0.1896,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.25610244097639057,
|
| 286 |
+
"grad_norm": 3.2575226153396115,
|
| 287 |
+
"learning_rate": 1.995974293995239e-06,
|
| 288 |
+
"loss": 0.194,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.26250500200080035,
|
| 293 |
+
"grad_norm": 3.7743686803486605,
|
| 294 |
+
"learning_rate": 1.994905874548752e-06,
|
| 295 |
+
"loss": 0.1882,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2689075630252101,
|
| 300 |
+
"grad_norm": 4.287147445421594,
|
| 301 |
+
"learning_rate": 1.9937122098932426e-06,
|
| 302 |
+
"loss": 0.1772,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.27531012404961985,
|
| 307 |
+
"grad_norm": 3.832538953651167,
|
| 308 |
+
"learning_rate": 1.9923934502949643e-06,
|
| 309 |
+
"loss": 0.1904,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.28171268507402963,
|
| 314 |
+
"grad_norm": 5.1771162120228365,
|
| 315 |
+
"learning_rate": 1.9909497617679347e-06,
|
| 316 |
+
"loss": 0.1892,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.28811524609843936,
|
| 321 |
+
"grad_norm": 4.303644513637155,
|
| 322 |
+
"learning_rate": 1.9893813260530367e-06,
|
| 323 |
+
"loss": 0.2056,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.29451780712284914,
|
| 328 |
+
"grad_norm": 4.2352469784485445,
|
| 329 |
+
"learning_rate": 1.9876883405951377e-06,
|
| 330 |
+
"loss": 0.2316,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.3009203681472589,
|
| 335 |
+
"grad_norm": 4.5051673675815955,
|
| 336 |
+
"learning_rate": 1.9858710185182355e-06,
|
| 337 |
+
"loss": 0.1792,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.30732292917166865,
|
| 342 |
+
"grad_norm": 4.68950995491574,
|
| 343 |
+
"learning_rate": 1.9839295885986295e-06,
|
| 344 |
+
"loss": 0.1841,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.3137254901960784,
|
| 349 |
+
"grad_norm": 4.053751205533602,
|
| 350 |
+
"learning_rate": 1.9818642952361183e-06,
|
| 351 |
+
"loss": 0.1909,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.3201280512204882,
|
| 356 |
+
"grad_norm": 3.537770773869747,
|
| 357 |
+
"learning_rate": 1.9796753984232355e-06,
|
| 358 |
+
"loss": 0.1642,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.32653061224489793,
|
| 363 |
+
"grad_norm": 4.937399937551455,
|
| 364 |
+
"learning_rate": 1.977363173712519e-06,
|
| 365 |
+
"loss": 0.1628,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.3329331732693077,
|
| 370 |
+
"grad_norm": 4.458098457737047,
|
| 371 |
+
"learning_rate": 1.9749279121818236e-06,
|
| 372 |
+
"loss": 0.2108,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.3393357342937175,
|
| 377 |
+
"grad_norm": 4.692309558133837,
|
| 378 |
+
"learning_rate": 1.9723699203976766e-06,
|
| 379 |
+
"loss": 0.2023,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.3457382953181273,
|
| 384 |
+
"grad_norm": 4.277910169048945,
|
| 385 |
+
"learning_rate": 1.9696895203766866e-06,
|
| 386 |
+
"loss": 0.1885,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.352140856342537,
|
| 391 |
+
"grad_norm": 4.837495606765268,
|
| 392 |
+
"learning_rate": 1.966887049545006e-06,
|
| 393 |
+
"loss": 0.1899,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.3585434173669468,
|
| 398 |
+
"grad_norm": 4.723060685847474,
|
| 399 |
+
"learning_rate": 1.9639628606958534e-06,
|
| 400 |
+
"loss": 0.2454,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.36494597839135656,
|
| 405 |
+
"grad_norm": 4.472186733872948,
|
| 406 |
+
"learning_rate": 1.9609173219450997e-06,
|
| 407 |
+
"loss": 0.1902,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.3713485394157663,
|
| 412 |
+
"grad_norm": 4.27556686419538,
|
| 413 |
+
"learning_rate": 1.9577508166849303e-06,
|
| 414 |
+
"loss": 0.1962,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.37775110044017607,
|
| 419 |
+
"grad_norm": 3.8215311273874506,
|
| 420 |
+
"learning_rate": 1.9544637435355806e-06,
|
| 421 |
+
"loss": 0.1658,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.38415366146458585,
|
| 426 |
+
"grad_norm": 4.623141330416456,
|
| 427 |
+
"learning_rate": 1.9510565162951534e-06,
|
| 428 |
+
"loss": 0.2274,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.3905562224889956,
|
| 433 |
+
"grad_norm": 4.245526551915113,
|
| 434 |
+
"learning_rate": 1.947529563887529e-06,
|
| 435 |
+
"loss": 0.2097,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.39695878351340536,
|
| 440 |
+
"grad_norm": 3.8360308169316064,
|
| 441 |
+
"learning_rate": 1.9438833303083674e-06,
|
| 442 |
+
"loss": 0.1629,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.40336134453781514,
|
| 447 |
+
"grad_norm": 4.390197594681171,
|
| 448 |
+
"learning_rate": 1.9401182745692187e-06,
|
| 449 |
+
"loss": 0.1913,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.4097639055622249,
|
| 454 |
+
"grad_norm": 3.8567511370833594,
|
| 455 |
+
"learning_rate": 1.936234870639737e-06,
|
| 456 |
+
"loss": 0.204,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.41616646658663464,
|
| 461 |
+
"grad_norm": 4.3670167230444354,
|
| 462 |
+
"learning_rate": 1.9322336073880143e-06,
|
| 463 |
+
"loss": 0.1925,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.4225690276110444,
|
| 468 |
+
"grad_norm": 4.5446121221499505,
|
| 469 |
+
"learning_rate": 1.928114988519039e-06,
|
| 470 |
+
"loss": 0.2036,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.4289715886354542,
|
| 475 |
+
"grad_norm": 4.492339684008934,
|
| 476 |
+
"learning_rate": 1.9238795325112867e-06,
|
| 477 |
+
"loss": 0.1976,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.43537414965986393,
|
| 482 |
+
"grad_norm": 3.616344744906229,
|
| 483 |
+
"learning_rate": 1.9195277725514506e-06,
|
| 484 |
+
"loss": 0.2173,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.4417767106842737,
|
| 489 |
+
"grad_norm": 4.807993813843196,
|
| 490 |
+
"learning_rate": 1.91506025646732e-06,
|
| 491 |
+
"loss": 0.2297,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.4481792717086835,
|
| 496 |
+
"grad_norm": 3.9611147863165215,
|
| 497 |
+
"learning_rate": 1.9104775466588157e-06,
|
| 498 |
+
"loss": 0.1724,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.4545818327330932,
|
| 503 |
+
"grad_norm": 4.038448965820852,
|
| 504 |
+
"learning_rate": 1.905780220027194e-06,
|
| 505 |
+
"loss": 0.1817,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.460984393757503,
|
| 510 |
+
"grad_norm": 3.6234581360493308,
|
| 511 |
+
"learning_rate": 1.9009688679024189e-06,
|
| 512 |
+
"loss": 0.1898,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.4673869547819128,
|
| 517 |
+
"grad_norm": 4.412888570956389,
|
| 518 |
+
"learning_rate": 1.8960440959687252e-06,
|
| 519 |
+
"loss": 0.2056,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.4737895158063225,
|
| 524 |
+
"grad_norm": 4.438175504352378,
|
| 525 |
+
"learning_rate": 1.8910065241883678e-06,
|
| 526 |
+
"loss": 0.2206,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.4801920768307323,
|
| 531 |
+
"grad_norm": 4.188755548423674,
|
| 532 |
+
"learning_rate": 1.8858567867235798e-06,
|
| 533 |
+
"loss": 0.1955,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.48659463785514206,
|
| 538 |
+
"grad_norm": 4.043921178736117,
|
| 539 |
+
"learning_rate": 1.8805955318567379e-06,
|
| 540 |
+
"loss": 0.1948,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.49299719887955185,
|
| 545 |
+
"grad_norm": 3.3626388865946195,
|
| 546 |
+
"learning_rate": 1.8752234219087537e-06,
|
| 547 |
+
"loss": 0.2156,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.49939975990396157,
|
| 552 |
+
"grad_norm": 3.856090864925135,
|
| 553 |
+
"learning_rate": 1.8697411331556953e-06,
|
| 554 |
+
"loss": 0.1969,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.5058023209283713,
|
| 559 |
+
"grad_norm": 3.902527444697349,
|
| 560 |
+
"learning_rate": 1.8641493557436548e-06,
|
| 561 |
+
"loss": 0.2156,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.5122048819527811,
|
| 566 |
+
"grad_norm": 3.8994885011285776,
|
| 567 |
+
"learning_rate": 1.858448793601866e-06,
|
| 568 |
+
"loss": 0.1978,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.5186074429771909,
|
| 573 |
+
"grad_norm": 3.842402356308699,
|
| 574 |
+
"learning_rate": 1.852640164354092e-06,
|
| 575 |
+
"loss": 0.2271,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.5250100040016007,
|
| 580 |
+
"grad_norm": 3.3752000095050376,
|
| 581 |
+
"learning_rate": 1.8467241992282841e-06,
|
| 582 |
+
"loss": 0.1707,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.5314125650260104,
|
| 587 |
+
"grad_norm": 4.043279107262682,
|
| 588 |
+
"learning_rate": 1.8407016429645302e-06,
|
| 589 |
+
"loss": 0.2163,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.5378151260504201,
|
| 594 |
+
"grad_norm": 4.403978560088287,
|
| 595 |
+
"learning_rate": 1.8345732537213026e-06,
|
| 596 |
+
"loss": 0.1739,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.54421768707483,
|
| 601 |
+
"grad_norm": 3.8981585910721552,
|
| 602 |
+
"learning_rate": 1.8283398029800164e-06,
|
| 603 |
+
"loss": 0.1696,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.5506202480992397,
|
| 608 |
+
"grad_norm": 4.932247696008996,
|
| 609 |
+
"learning_rate": 1.82200207544791e-06,
|
| 610 |
+
"loss": 0.1915,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.5570228091236494,
|
| 615 |
+
"grad_norm": 4.099692261040826,
|
| 616 |
+
"learning_rate": 1.8155608689592601e-06,
|
| 617 |
+
"loss": 0.1785,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.5634253701480593,
|
| 622 |
+
"grad_norm": 3.7071427170962585,
|
| 623 |
+
"learning_rate": 1.8090169943749474e-06,
|
| 624 |
+
"loss": 0.1643,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.569827931172469,
|
| 629 |
+
"grad_norm": 4.198912758644434,
|
| 630 |
+
"learning_rate": 1.802371275480378e-06,
|
| 631 |
+
"loss": 0.2079,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.5762304921968787,
|
| 636 |
+
"grad_norm": 4.08750592105122,
|
| 637 |
+
"learning_rate": 1.795624548881781e-06,
|
| 638 |
+
"loss": 0.1875,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.5826330532212886,
|
| 643 |
+
"grad_norm": 5.4368089983091945,
|
| 644 |
+
"learning_rate": 1.7887776639008912e-06,
|
| 645 |
+
"loss": 0.1842,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.5890356142456983,
|
| 650 |
+
"grad_norm": 3.6668016857716,
|
| 651 |
+
"learning_rate": 1.7818314824680298e-06,
|
| 652 |
+
"loss": 0.1927,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.595438175270108,
|
| 657 |
+
"grad_norm": 3.9325828012665482,
|
| 658 |
+
"learning_rate": 1.774786879013601e-06,
|
| 659 |
+
"loss": 0.1849,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6018407362945178,
|
| 664 |
+
"grad_norm": 4.2551345503780125,
|
| 665 |
+
"learning_rate": 1.767644740358011e-06,
|
| 666 |
+
"loss": 0.2103,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6082432973189276,
|
| 671 |
+
"grad_norm": 3.562360928332586,
|
| 672 |
+
"learning_rate": 1.760405965600031e-06,
|
| 673 |
+
"loss": 0.2292,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6146458583433373,
|
| 678 |
+
"grad_norm": 4.074513845673071,
|
| 679 |
+
"learning_rate": 1.753071466003611e-06,
|
| 680 |
+
"loss": 0.164,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.6210484193677471,
|
| 685 |
+
"grad_norm": 4.264136629053828,
|
| 686 |
+
"learning_rate": 1.7456421648831654e-06,
|
| 687 |
+
"loss": 0.1919,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.6274509803921569,
|
| 692 |
+
"grad_norm": 4.101904029046064,
|
| 693 |
+
"learning_rate": 1.7381189974873407e-06,
|
| 694 |
+
"loss": 0.2492,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.6338535414165666,
|
| 699 |
+
"grad_norm": 4.129940049014058,
|
| 700 |
+
"learning_rate": 1.7305029108812774e-06,
|
| 701 |
+
"loss": 0.1899,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.6402561024409764,
|
| 706 |
+
"grad_norm": 4.2688607303525705,
|
| 707 |
+
"learning_rate": 1.7227948638273915e-06,
|
| 708 |
+
"loss": 0.1802,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.6466586634653861,
|
| 713 |
+
"grad_norm": 3.931916755773599,
|
| 714 |
+
"learning_rate": 1.7149958266646754e-06,
|
| 715 |
+
"loss": 0.1524,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.6530612244897959,
|
| 720 |
+
"grad_norm": 4.309265054679012,
|
| 721 |
+
"learning_rate": 1.7071067811865474e-06,
|
| 722 |
+
"loss": 0.2035,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.6594637855142057,
|
| 727 |
+
"grad_norm": 3.4899838789545847,
|
| 728 |
+
"learning_rate": 1.6991287205172574e-06,
|
| 729 |
+
"loss": 0.1885,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.6658663465386154,
|
| 734 |
+
"grad_norm": 4.603685696411418,
|
| 735 |
+
"learning_rate": 1.6910626489868648e-06,
|
| 736 |
+
"loss": 0.2008,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.6722689075630253,
|
| 741 |
+
"grad_norm": 3.682716771358225,
|
| 742 |
+
"learning_rate": 1.682909582004807e-06,
|
| 743 |
+
"loss": 0.2069,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.678671468587435,
|
| 748 |
+
"grad_norm": 5.35989531173202,
|
| 749 |
+
"learning_rate": 1.6746705459320744e-06,
|
| 750 |
+
"loss": 0.2185,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.6850740296118447,
|
| 755 |
+
"grad_norm": 4.4810176716864225,
|
| 756 |
+
"learning_rate": 1.6663465779520037e-06,
|
| 757 |
+
"loss": 0.2179,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.6914765906362546,
|
| 762 |
+
"grad_norm": 4.781878607782948,
|
| 763 |
+
"learning_rate": 1.6579387259397126e-06,
|
| 764 |
+
"loss": 0.2162,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.6978791516606643,
|
| 769 |
+
"grad_norm": 3.578367620297612,
|
| 770 |
+
"learning_rate": 1.6494480483301835e-06,
|
| 771 |
+
"loss": 0.1864,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.704281712685074,
|
| 776 |
+
"grad_norm": 4.938703350422557,
|
| 777 |
+
"learning_rate": 1.640875613985024e-06,
|
| 778 |
+
"loss": 0.1789,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.7106842737094838,
|
| 783 |
+
"grad_norm": 4.712146890436173,
|
| 784 |
+
"learning_rate": 1.6322225020579096e-06,
|
| 785 |
+
"loss": 0.1971,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.7170868347338936,
|
| 790 |
+
"grad_norm": 4.49092614751852,
|
| 791 |
+
"learning_rate": 1.6234898018587336e-06,
|
| 792 |
+
"loss": 0.1796,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.7234893957583033,
|
| 797 |
+
"grad_norm": 4.123243996580139,
|
| 798 |
+
"learning_rate": 1.6146786127164771e-06,
|
| 799 |
+
"loss": 0.1703,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.7298919567827131,
|
| 804 |
+
"grad_norm": 3.4610648898646623,
|
| 805 |
+
"learning_rate": 1.6057900438408199e-06,
|
| 806 |
+
"loss": 0.2069,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.7362945178071229,
|
| 811 |
+
"grad_norm": 3.701626323345168,
|
| 812 |
+
"learning_rate": 1.5968252141825035e-06,
|
| 813 |
+
"loss": 0.2037,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.7426970788315326,
|
| 818 |
+
"grad_norm": 4.5853683365418965,
|
| 819 |
+
"learning_rate": 1.587785252292473e-06,
|
| 820 |
+
"loss": 0.1867,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.7490996398559424,
|
| 825 |
+
"grad_norm": 3.813531824888892,
|
| 826 |
+
"learning_rate": 1.578671296179806e-06,
|
| 827 |
+
"loss": 0.1923,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.7555022008803521,
|
| 832 |
+
"grad_norm": 4.333755687391246,
|
| 833 |
+
"learning_rate": 1.569484493168452e-06,
|
| 834 |
+
"loss": 0.1917,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.7619047619047619,
|
| 839 |
+
"grad_norm": 3.5490254589371526,
|
| 840 |
+
"learning_rate": 1.5602259997528027e-06,
|
| 841 |
+
"loss": 0.21,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.7683073229291717,
|
| 846 |
+
"grad_norm": 3.8349133617246665,
|
| 847 |
+
"learning_rate": 1.5508969814521024e-06,
|
| 848 |
+
"loss": 0.1839,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.7747098839535814,
|
| 853 |
+
"grad_norm": 4.480906522239498,
|
| 854 |
+
"learning_rate": 1.5414986126637257e-06,
|
| 855 |
+
"loss": 0.1804,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.7811124449779911,
|
| 860 |
+
"grad_norm": 4.063797691144723,
|
| 861 |
+
"learning_rate": 1.5320320765153365e-06,
|
| 862 |
+
"loss": 0.1872,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.787515006002401,
|
| 867 |
+
"grad_norm": 3.837906572094775,
|
| 868 |
+
"learning_rate": 1.5224985647159488e-06,
|
| 869 |
+
"loss": 0.202,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.7939175670268107,
|
| 874 |
+
"grad_norm": 4.226911617716516,
|
| 875 |
+
"learning_rate": 1.5128992774059062e-06,
|
| 876 |
+
"loss": 0.2067,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.8003201280512204,
|
| 881 |
+
"grad_norm": 4.143910488552411,
|
| 882 |
+
"learning_rate": 1.5032354230058002e-06,
|
| 883 |
+
"loss": 0.1884,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.8067226890756303,
|
| 888 |
+
"grad_norm": 3.996874874040649,
|
| 889 |
+
"learning_rate": 1.4935082180643467e-06,
|
| 890 |
+
"loss": 0.1797,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.81312525010004,
|
| 895 |
+
"grad_norm": 4.241850731448795,
|
| 896 |
+
"learning_rate": 1.4837188871052397e-06,
|
| 897 |
+
"loss": 0.1973,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.8195278111244498,
|
| 902 |
+
"grad_norm": 4.8128807497544015,
|
| 903 |
+
"learning_rate": 1.4738686624729987e-06,
|
| 904 |
+
"loss": 0.1888,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.8259303721488596,
|
| 909 |
+
"grad_norm": 4.624440170471318,
|
| 910 |
+
"learning_rate": 1.463958784177834e-06,
|
| 911 |
+
"loss": 0.1855,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.8323329331732693,
|
| 916 |
+
"grad_norm": 3.760993103519322,
|
| 917 |
+
"learning_rate": 1.4539904997395467e-06,
|
| 918 |
+
"loss": 0.1662,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.8387354941976791,
|
| 923 |
+
"grad_norm": 4.333505468408259,
|
| 924 |
+
"learning_rate": 1.4439650640304821e-06,
|
| 925 |
+
"loss": 0.2089,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.8451380552220888,
|
| 930 |
+
"grad_norm": 4.278934798609152,
|
| 931 |
+
"learning_rate": 1.433883739117558e-06,
|
| 932 |
+
"loss": 0.1879,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.8515406162464986,
|
| 937 |
+
"grad_norm": 4.16072632298661,
|
| 938 |
+
"learning_rate": 1.4237477941033886e-06,
|
| 939 |
+
"loss": 0.1904,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.8579431772709084,
|
| 944 |
+
"grad_norm": 3.5517189505372784,
|
| 945 |
+
"learning_rate": 1.4135585049665206e-06,
|
| 946 |
+
"loss": 0.2498,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.8643457382953181,
|
| 951 |
+
"grad_norm": 3.9706094312160776,
|
| 952 |
+
"learning_rate": 1.4033171544008051e-06,
|
| 953 |
+
"loss": 0.1901,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.8707482993197279,
|
| 958 |
+
"grad_norm": 4.171292985573587,
|
| 959 |
+
"learning_rate": 1.3930250316539235e-06,
|
| 960 |
+
"loss": 0.1835,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.8771508603441377,
|
| 965 |
+
"grad_norm": 3.8333520124526896,
|
| 966 |
+
"learning_rate": 1.3826834323650898e-06,
|
| 967 |
+
"loss": 0.1747,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.8835534213685474,
|
| 972 |
+
"grad_norm": 4.281605413414975,
|
| 973 |
+
"learning_rate": 1.3722936584019451e-06,
|
| 974 |
+
"loss": 0.2296,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.8899559823929571,
|
| 979 |
+
"grad_norm": 4.001127332533811,
|
| 980 |
+
"learning_rate": 1.3618570176966722e-06,
|
| 981 |
+
"loss": 0.2248,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.896358543417367,
|
| 986 |
+
"grad_norm": 3.2545123553526585,
|
| 987 |
+
"learning_rate": 1.3513748240813427e-06,
|
| 988 |
+
"loss": 0.2172,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.9027611044417767,
|
| 993 |
+
"grad_norm": 3.6921559524457073,
|
| 994 |
+
"learning_rate": 1.3408483971225249e-06,
|
| 995 |
+
"loss": 0.1865,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.9091636654661864,
|
| 1000 |
+
"grad_norm": 4.729238595946433,
|
| 1001 |
+
"learning_rate": 1.3302790619551672e-06,
|
| 1002 |
+
"loss": 0.1511,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.9155662264905963,
|
| 1007 |
+
"grad_norm": 4.27641471239857,
|
| 1008 |
+
"learning_rate": 1.3196681491157816e-06,
|
| 1009 |
+
"loss": 0.1775,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.921968787515006,
|
| 1014 |
+
"grad_norm": 4.195425093653204,
|
| 1015 |
+
"learning_rate": 1.3090169943749473e-06,
|
| 1016 |
+
"loss": 0.2102,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.9283713485394157,
|
| 1021 |
+
"grad_norm": 4.4927187865344305,
|
| 1022 |
+
"learning_rate": 1.298326938569156e-06,
|
| 1023 |
+
"loss": 0.1794,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.9347739095638256,
|
| 1028 |
+
"grad_norm": 3.6780116242547467,
|
| 1029 |
+
"learning_rate": 1.2875993274320173e-06,
|
| 1030 |
+
"loss": 0.1716,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.9411764705882353,
|
| 1035 |
+
"grad_norm": 3.5052737613672926,
|
| 1036 |
+
"learning_rate": 1.2768355114248492e-06,
|
| 1037 |
+
"loss": 0.1909,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.947579031612645,
|
| 1042 |
+
"grad_norm": 5.4003852624340976,
|
| 1043 |
+
"learning_rate": 1.266036845566675e-06,
|
| 1044 |
+
"loss": 0.1828,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.9539815926370548,
|
| 1049 |
+
"grad_norm": 4.293449029987786,
|
| 1050 |
+
"learning_rate": 1.2552046892636426e-06,
|
| 1051 |
+
"loss": 0.1934,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.9603841536614646,
|
| 1056 |
+
"grad_norm": 4.213735783478494,
|
| 1057 |
+
"learning_rate": 1.244340406137894e-06,
|
| 1058 |
+
"loss": 0.1635,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.9667867146858744,
|
| 1063 |
+
"grad_norm": 5.682803899024664,
|
| 1064 |
+
"learning_rate": 1.2334453638559054e-06,
|
| 1065 |
+
"loss": 0.2282,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.9731892757102841,
|
| 1070 |
+
"grad_norm": 5.125716971978602,
|
| 1071 |
+
"learning_rate": 1.2225209339563143e-06,
|
| 1072 |
+
"loss": 0.2124,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.9795918367346939,
|
| 1077 |
+
"grad_norm": 3.6296567686595558,
|
| 1078 |
+
"learning_rate": 1.211568491677263e-06,
|
| 1079 |
+
"loss": 0.1686,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.9859943977591037,
|
| 1084 |
+
"grad_norm": 4.032149136826441,
|
| 1085 |
+
"learning_rate": 1.2005894157832728e-06,
|
| 1086 |
+
"loss": 0.1801,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.9923969587835134,
|
| 1091 |
+
"grad_norm": 4.328877120814621,
|
| 1092 |
+
"learning_rate": 1.1895850883916785e-06,
|
| 1093 |
+
"loss": 0.2116,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.9987995198079231,
|
| 1098 |
+
"grad_norm": 4.42679277844411,
|
| 1099 |
+
"learning_rate": 1.1785568947986366e-06,
|
| 1100 |
+
"loss": 0.197,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.005202080832333,
|
| 1105 |
+
"grad_norm": 2.995825299760741,
|
| 1106 |
+
"learning_rate": 1.1675062233047363e-06,
|
| 1107 |
+
"loss": 0.1754,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.0116046418567426,
|
| 1112 |
+
"grad_norm": 3.314198590902652,
|
| 1113 |
+
"learning_rate": 1.156434465040231e-06,
|
| 1114 |
+
"loss": 0.1591,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.0180072028811524,
|
| 1119 |
+
"grad_norm": 3.282342144727049,
|
| 1120 |
+
"learning_rate": 1.1453430137899128e-06,
|
| 1121 |
+
"loss": 0.1463,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.0244097639055623,
|
| 1126 |
+
"grad_norm": 3.3852348105131327,
|
| 1127 |
+
"learning_rate": 1.1342332658176555e-06,
|
| 1128 |
+
"loss": 0.1417,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.0308123249299719,
|
| 1133 |
+
"grad_norm": 2.5846365344808224,
|
| 1134 |
+
"learning_rate": 1.123106619690643e-06,
|
| 1135 |
+
"loss": 0.1579,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.0372148859543817,
|
| 1140 |
+
"grad_norm": 3.4112665397765407,
|
| 1141 |
+
"learning_rate": 1.1119644761033077e-06,
|
| 1142 |
+
"loss": 0.1569,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.0436174469787916,
|
| 1147 |
+
"grad_norm": 3.057068352071886,
|
| 1148 |
+
"learning_rate": 1.1008082377010045e-06,
|
| 1149 |
+
"loss": 0.1588,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.0500200080032012,
|
| 1154 |
+
"grad_norm": 2.480458097820055,
|
| 1155 |
+
"learning_rate": 1.0896393089034335e-06,
|
| 1156 |
+
"loss": 0.1604,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.056422569027611,
|
| 1161 |
+
"grad_norm": 2.5238894472575,
|
| 1162 |
+
"learning_rate": 1.078459095727845e-06,
|
| 1163 |
+
"loss": 0.1738,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.0628251300520208,
|
| 1168 |
+
"grad_norm": 2.702306669134078,
|
| 1169 |
+
"learning_rate": 1.0672690056120398e-06,
|
| 1170 |
+
"loss": 0.1416,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.0692276910764307,
|
| 1175 |
+
"grad_norm": 2.715141146658483,
|
| 1176 |
+
"learning_rate": 1.0560704472371917e-06,
|
| 1177 |
+
"loss": 0.127,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.0756302521008403,
|
| 1182 |
+
"grad_norm": 2.698908112656233,
|
| 1183 |
+
"learning_rate": 1.044864830350515e-06,
|
| 1184 |
+
"loss": 0.1769,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.0820328131252501,
|
| 1189 |
+
"grad_norm": 2.393553347708748,
|
| 1190 |
+
"learning_rate": 1.033653565587794e-06,
|
| 1191 |
+
"loss": 0.1595,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.08843537414966,
|
| 1196 |
+
"grad_norm": 2.515931323925557,
|
| 1197 |
+
"learning_rate": 1.022438064295805e-06,
|
| 1198 |
+
"loss": 0.1509,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.0948379351740696,
|
| 1203 |
+
"grad_norm": 3.122208532437112,
|
| 1204 |
+
"learning_rate": 1.0112197383546459e-06,
|
| 1205 |
+
"loss": 0.1362,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.1012404961984794,
|
| 1210 |
+
"grad_norm": 2.9259499260174437,
|
| 1211 |
+
"learning_rate": 1e-06,
|
| 1212 |
+
"loss": 0.1393,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.1076430572228892,
|
| 1217 |
+
"grad_norm": 2.6643868148585783,
|
| 1218 |
+
"learning_rate": 9.88780261645354e-07,
|
| 1219 |
+
"loss": 0.1602,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.1140456182472989,
|
| 1224 |
+
"grad_norm": 2.529601808890674,
|
| 1225 |
+
"learning_rate": 9.77561935704195e-07,
|
| 1226 |
+
"loss": 0.16,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.1204481792717087,
|
| 1231 |
+
"grad_norm": 2.4798313875614038,
|
| 1232 |
+
"learning_rate": 9.663464344122063e-07,
|
| 1233 |
+
"loss": 0.1667,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.1268507402961185,
|
| 1238 |
+
"grad_norm": 2.528807676927116,
|
| 1239 |
+
"learning_rate": 9.551351696494853e-07,
|
| 1240 |
+
"loss": 0.142,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.1332533013205282,
|
| 1245 |
+
"grad_norm": 2.222384060658359,
|
| 1246 |
+
"learning_rate": 9.43929552762808e-07,
|
| 1247 |
+
"loss": 0.1273,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.139655862344938,
|
| 1252 |
+
"grad_norm": 2.6020312291887993,
|
| 1253 |
+
"learning_rate": 9.327309943879603e-07,
|
| 1254 |
+
"loss": 0.1209,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.1460584233693478,
|
| 1259 |
+
"grad_norm": 2.5865819736528213,
|
| 1260 |
+
"learning_rate": 9.215409042721551e-07,
|
| 1261 |
+
"loss": 0.1147,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.1524609843937574,
|
| 1266 |
+
"grad_norm": 2.6984970538621047,
|
| 1267 |
+
"learning_rate": 9.103606910965665e-07,
|
| 1268 |
+
"loss": 0.1267,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.1588635454181673,
|
| 1273 |
+
"grad_norm": 3.064928156142861,
|
| 1274 |
+
"learning_rate": 8.991917622989955e-07,
|
| 1275 |
+
"loss": 0.1472,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.165266106442577,
|
| 1280 |
+
"grad_norm": 2.6751065232729787,
|
| 1281 |
+
"learning_rate": 8.880355238966921e-07,
|
| 1282 |
+
"loss": 0.1481,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.1716686674669867,
|
| 1287 |
+
"grad_norm": 2.4919821908211137,
|
| 1288 |
+
"learning_rate": 8.768933803093572e-07,
|
| 1289 |
+
"loss": 0.1582,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.1780712284913966,
|
| 1294 |
+
"grad_norm": 2.7496916062023926,
|
| 1295 |
+
"learning_rate": 8.657667341823448e-07,
|
| 1296 |
+
"loss": 0.1132,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.1844737895158064,
|
| 1301 |
+
"grad_norm": 2.8683066132542065,
|
| 1302 |
+
"learning_rate": 8.546569862100875e-07,
|
| 1303 |
+
"loss": 0.1537,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.190876350540216,
|
| 1308 |
+
"grad_norm": 3.3497753630792504,
|
| 1309 |
+
"learning_rate": 8.435655349597689e-07,
|
| 1310 |
+
"loss": 0.1498,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.1972789115646258,
|
| 1315 |
+
"grad_norm": 2.8175991760028776,
|
| 1316 |
+
"learning_rate": 8.324937766952636e-07,
|
| 1317 |
+
"loss": 0.1344,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.2036814725890357,
|
| 1322 |
+
"grad_norm": 3.5078725354724978,
|
| 1323 |
+
"learning_rate": 8.214431052013634e-07,
|
| 1324 |
+
"loss": 0.1794,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.2100840336134453,
|
| 1329 |
+
"grad_norm": 3.134230661450569,
|
| 1330 |
+
"learning_rate": 8.104149116083216e-07,
|
| 1331 |
+
"loss": 0.1636,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.2164865946378551,
|
| 1336 |
+
"grad_norm": 2.789058239974061,
|
| 1337 |
+
"learning_rate": 7.994105842167272e-07,
|
| 1338 |
+
"loss": 0.1384,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.222889155662265,
|
| 1343 |
+
"grad_norm": 3.7547713671098966,
|
| 1344 |
+
"learning_rate": 7.884315083227372e-07,
|
| 1345 |
+
"loss": 0.1774,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.2292917166866746,
|
| 1350 |
+
"grad_norm": 3.664307191430093,
|
| 1351 |
+
"learning_rate": 7.774790660436857e-07,
|
| 1352 |
+
"loss": 0.1481,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.2356942777110844,
|
| 1357 |
+
"grad_norm": 2.771171440914553,
|
| 1358 |
+
"learning_rate": 7.665546361440949e-07,
|
| 1359 |
+
"loss": 0.1996,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.2420968387354943,
|
| 1364 |
+
"grad_norm": 2.6893659053745815,
|
| 1365 |
+
"learning_rate": 7.556595938621058e-07,
|
| 1366 |
+
"loss": 0.1411,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.2484993997599039,
|
| 1371 |
+
"grad_norm": 3.678898327438179,
|
| 1372 |
+
"learning_rate": 7.447953107363574e-07,
|
| 1373 |
+
"loss": 0.1864,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.2549019607843137,
|
| 1378 |
+
"grad_norm": 3.520031085298818,
|
| 1379 |
+
"learning_rate": 7.33963154433325e-07,
|
| 1380 |
+
"loss": 0.1681,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.2613045218087235,
|
| 1385 |
+
"grad_norm": 3.1974510727727354,
|
| 1386 |
+
"learning_rate": 7.231644885751507e-07,
|
| 1387 |
+
"loss": 0.1197,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.2677070828331334,
|
| 1392 |
+
"grad_norm": 3.0861033592576703,
|
| 1393 |
+
"learning_rate": 7.124006725679828e-07,
|
| 1394 |
+
"loss": 0.1852,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.274109643857543,
|
| 1399 |
+
"grad_norm": 2.392211457828515,
|
| 1400 |
+
"learning_rate": 7.016730614308439e-07,
|
| 1401 |
+
"loss": 0.1478,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.2805122048819528,
|
| 1406 |
+
"grad_norm": 2.6382593348580636,
|
| 1407 |
+
"learning_rate": 6.909830056250526e-07,
|
| 1408 |
+
"loss": 0.1499,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.2869147659063627,
|
| 1413 |
+
"grad_norm": 3.310685530214126,
|
| 1414 |
+
"learning_rate": 6.803318508842186e-07,
|
| 1415 |
+
"loss": 0.1892,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.2933173269307723,
|
| 1420 |
+
"grad_norm": 2.9330195000820054,
|
| 1421 |
+
"learning_rate": 6.697209380448332e-07,
|
| 1422 |
+
"loss": 0.1853,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.2997198879551821,
|
| 1427 |
+
"grad_norm": 2.8596301823674115,
|
| 1428 |
+
"learning_rate": 6.59151602877475e-07,
|
| 1429 |
+
"loss": 0.1328,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.306122448979592,
|
| 1434 |
+
"grad_norm": 2.794306696992077,
|
| 1435 |
+
"learning_rate": 6.486251759186572e-07,
|
| 1436 |
+
"loss": 0.1807,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.3125250100040016,
|
| 1441 |
+
"grad_norm": 4.390407508756913,
|
| 1442 |
+
"learning_rate": 6.381429823033279e-07,
|
| 1443 |
+
"loss": 0.1536,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.3189275710284114,
|
| 1448 |
+
"grad_norm": 2.6300011319364915,
|
| 1449 |
+
"learning_rate": 6.277063415980548e-07,
|
| 1450 |
+
"loss": 0.1431,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.3253301320528212,
|
| 1455 |
+
"grad_norm": 2.62332174376127,
|
| 1456 |
+
"learning_rate": 6.173165676349102e-07,
|
| 1457 |
+
"loss": 0.1482,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.3317326930772309,
|
| 1462 |
+
"grad_norm": 2.733993586202797,
|
| 1463 |
+
"learning_rate": 6.069749683460764e-07,
|
| 1464 |
+
"loss": 0.1228,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.3381352541016407,
|
| 1469 |
+
"grad_norm": 3.315910975674784,
|
| 1470 |
+
"learning_rate": 5.96682845599195e-07,
|
| 1471 |
+
"loss": 0.2054,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.3445378151260505,
|
| 1476 |
+
"grad_norm": 2.930927185763075,
|
| 1477 |
+
"learning_rate": 5.864414950334795e-07,
|
| 1478 |
+
"loss": 0.159,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.3509403761504601,
|
| 1483 |
+
"grad_norm": 3.1317606911937568,
|
| 1484 |
+
"learning_rate": 5.762522058966113e-07,
|
| 1485 |
+
"loss": 0.1586,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.35734293717487,
|
| 1490 |
+
"grad_norm": 4.040118089874495,
|
| 1491 |
+
"learning_rate": 5.661162608824419e-07,
|
| 1492 |
+
"loss": 0.1505,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.3637454981992798,
|
| 1497 |
+
"grad_norm": 2.746598755946971,
|
| 1498 |
+
"learning_rate": 5.56034935969518e-07,
|
| 1499 |
+
"loss": 0.1761,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.3701480592236894,
|
| 1504 |
+
"grad_norm": 2.9141623052335137,
|
| 1505 |
+
"learning_rate": 5.460095002604532e-07,
|
| 1506 |
+
"loss": 0.19,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.3765506202480993,
|
| 1511 |
+
"grad_norm": 4.001391421545072,
|
| 1512 |
+
"learning_rate": 5.36041215822166e-07,
|
| 1513 |
+
"loss": 0.1321,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.382953181272509,
|
| 1518 |
+
"grad_norm": 3.6428702220535736,
|
| 1519 |
+
"learning_rate": 5.261313375270013e-07,
|
| 1520 |
+
"loss": 0.1905,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.3893557422969187,
|
| 1525 |
+
"grad_norm": 3.324099015382822,
|
| 1526 |
+
"learning_rate": 5.162811128947602e-07,
|
| 1527 |
+
"loss": 0.1669,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.3957583033213286,
|
| 1532 |
+
"grad_norm": 2.6236595277867294,
|
| 1533 |
+
"learning_rate": 5.064917819356531e-07,
|
| 1534 |
+
"loss": 0.1235,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.4021608643457384,
|
| 1539 |
+
"grad_norm": 3.259881675873255,
|
| 1540 |
+
"learning_rate": 4.967645769941999e-07,
|
| 1541 |
+
"loss": 0.1814,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.408563425370148,
|
| 1546 |
+
"grad_norm": 2.8694964354069117,
|
| 1547 |
+
"learning_rate": 4.871007225940939e-07,
|
| 1548 |
+
"loss": 0.1495,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.4149659863945578,
|
| 1553 |
+
"grad_norm": 2.901840928815238,
|
| 1554 |
+
"learning_rate": 4.775014352840512e-07,
|
| 1555 |
+
"loss": 0.1951,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.4213685474189677,
|
| 1560 |
+
"grad_norm": 3.418089743568751,
|
| 1561 |
+
"learning_rate": 4.6796792348466353e-07,
|
| 1562 |
+
"loss": 0.1547,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.4277711084433773,
|
| 1567 |
+
"grad_norm": 2.830949019249368,
|
| 1568 |
+
"learning_rate": 4.585013873362743e-07,
|
| 1569 |
+
"loss": 0.1046,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.4341736694677871,
|
| 1574 |
+
"grad_norm": 3.5740356720869206,
|
| 1575 |
+
"learning_rate": 4.4910301854789755e-07,
|
| 1576 |
+
"loss": 0.1482,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.440576230492197,
|
| 1581 |
+
"grad_norm": 2.7706265692202217,
|
| 1582 |
+
"learning_rate": 4.397740002471972e-07,
|
| 1583 |
+
"loss": 0.1842,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.4469787915166066,
|
| 1588 |
+
"grad_norm": 2.705953973970954,
|
| 1589 |
+
"learning_rate": 4.3051550683154804e-07,
|
| 1590 |
+
"loss": 0.173,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.4533813525410164,
|
| 1595 |
+
"grad_norm": 3.5200476993463976,
|
| 1596 |
+
"learning_rate": 4.2132870382019427e-07,
|
| 1597 |
+
"loss": 0.1268,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.4597839135654262,
|
| 1602 |
+
"grad_norm": 2.980462619255481,
|
| 1603 |
+
"learning_rate": 4.1221474770752696e-07,
|
| 1604 |
+
"loss": 0.1454,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.4661864745898359,
|
| 1609 |
+
"grad_norm": 3.0488417880599274,
|
| 1610 |
+
"learning_rate": 4.031747858174964e-07,
|
| 1611 |
+
"loss": 0.1573,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.4725890356142457,
|
| 1616 |
+
"grad_norm": 2.866558859217372,
|
| 1617 |
+
"learning_rate": 3.942099561591802e-07,
|
| 1618 |
+
"loss": 0.1354,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.4789915966386555,
|
| 1623 |
+
"grad_norm": 3.3643880310904017,
|
| 1624 |
+
"learning_rate": 3.853213872835228e-07,
|
| 1625 |
+
"loss": 0.1527,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.4853941576630652,
|
| 1630 |
+
"grad_norm": 2.3493487438896463,
|
| 1631 |
+
"learning_rate": 3.765101981412665e-07,
|
| 1632 |
+
"loss": 0.1483,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.491796718687475,
|
| 1637 |
+
"grad_norm": 2.482947067390714,
|
| 1638 |
+
"learning_rate": 3.677774979420903e-07,
|
| 1639 |
+
"loss": 0.1302,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.4981992797118848,
|
| 1644 |
+
"grad_norm": 2.599751282623492,
|
| 1645 |
+
"learning_rate": 3.5912438601497584e-07,
|
| 1646 |
+
"loss": 0.1522,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.5046018407362944,
|
| 1651 |
+
"grad_norm": 3.191973535201136,
|
| 1652 |
+
"learning_rate": 3.5055195166981646e-07,
|
| 1653 |
+
"loss": 0.1615,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.5110044017607043,
|
| 1658 |
+
"grad_norm": 3.15562179961526,
|
| 1659 |
+
"learning_rate": 3.420612740602874e-07,
|
| 1660 |
+
"loss": 0.1478,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.517406962785114,
|
| 1665 |
+
"grad_norm": 2.8880563361860823,
|
| 1666 |
+
"learning_rate": 3.3365342204799606e-07,
|
| 1667 |
+
"loss": 0.1397,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.5238095238095237,
|
| 1672 |
+
"grad_norm": 3.688725379316976,
|
| 1673 |
+
"learning_rate": 3.253294540679257e-07,
|
| 1674 |
+
"loss": 0.1875,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.5302120848339336,
|
| 1679 |
+
"grad_norm": 2.817507541571056,
|
| 1680 |
+
"learning_rate": 3.170904179951931e-07,
|
| 1681 |
+
"loss": 0.1216,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.5366146458583434,
|
| 1686 |
+
"grad_norm": 2.9412647550371975,
|
| 1687 |
+
"learning_rate": 3.0893735101313535e-07,
|
| 1688 |
+
"loss": 0.1398,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.543017206882753,
|
| 1693 |
+
"grad_norm": 2.762627387485447,
|
| 1694 |
+
"learning_rate": 3.008712794827426e-07,
|
| 1695 |
+
"loss": 0.1537,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.5494197679071628,
|
| 1700 |
+
"grad_norm": 2.8082508660384633,
|
| 1701 |
+
"learning_rate": 2.9289321881345254e-07,
|
| 1702 |
+
"loss": 0.1367,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.5558223289315727,
|
| 1707 |
+
"grad_norm": 3.1843132713343314,
|
| 1708 |
+
"learning_rate": 2.850041733353247e-07,
|
| 1709 |
+
"loss": 0.1451,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.5622248899559823,
|
| 1714 |
+
"grad_norm": 2.549326946513764,
|
| 1715 |
+
"learning_rate": 2.7720513617260855e-07,
|
| 1716 |
+
"loss": 0.1578,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 1.5686274509803921,
|
| 1721 |
+
"grad_norm": 2.035464821513073,
|
| 1722 |
+
"learning_rate": 2.6949708911872247e-07,
|
| 1723 |
+
"loss": 0.1271,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.575030012004802,
|
| 1728 |
+
"grad_norm": 2.494783934906446,
|
| 1729 |
+
"learning_rate": 2.6188100251265943e-07,
|
| 1730 |
+
"loss": 0.1446,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.5814325730292116,
|
| 1735 |
+
"grad_norm": 4.226886500584243,
|
| 1736 |
+
"learning_rate": 2.543578351168344e-07,
|
| 1737 |
+
"loss": 0.1417,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 1.5878351340536214,
|
| 1742 |
+
"grad_norm": 2.495677422705392,
|
| 1743 |
+
"learning_rate": 2.4692853399638913e-07,
|
| 1744 |
+
"loss": 0.1364,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.5942376950780313,
|
| 1749 |
+
"grad_norm": 3.1917365323900446,
|
| 1750 |
+
"learning_rate": 2.395940343999691e-07,
|
| 1751 |
+
"loss": 0.1626,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.6006402561024409,
|
| 1756 |
+
"grad_norm": 3.0344619777547868,
|
| 1757 |
+
"learning_rate": 2.3235525964198888e-07,
|
| 1758 |
+
"loss": 0.1623,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 1.6070428171268507,
|
| 1763 |
+
"grad_norm": 3.688983622501882,
|
| 1764 |
+
"learning_rate": 2.252131209863991e-07,
|
| 1765 |
+
"loss": 0.1654,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 1.6134453781512605,
|
| 1770 |
+
"grad_norm": 2.760266999319,
|
| 1771 |
+
"learning_rate": 2.181685175319702e-07,
|
| 1772 |
+
"loss": 0.1757,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 1.6198479391756702,
|
| 1777 |
+
"grad_norm": 3.3493718404784816,
|
| 1778 |
+
"learning_rate": 2.11222336099109e-07,
|
| 1779 |
+
"loss": 0.1877,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 1.62625050020008,
|
| 1784 |
+
"grad_norm": 2.7059015648474767,
|
| 1785 |
+
"learning_rate": 2.043754511182191e-07,
|
| 1786 |
+
"loss": 0.1302,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.6326530612244898,
|
| 1791 |
+
"grad_norm": 2.6710609583138636,
|
| 1792 |
+
"learning_rate": 1.9762872451962208e-07,
|
| 1793 |
+
"loss": 0.1376,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 1.6390556222488994,
|
| 1798 |
+
"grad_norm": 2.636600973542151,
|
| 1799 |
+
"learning_rate": 1.9098300562505264e-07,
|
| 1800 |
+
"loss": 0.1667,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 1.6454581832733093,
|
| 1805 |
+
"grad_norm": 2.7163444391021647,
|
| 1806 |
+
"learning_rate": 1.8443913104073982e-07,
|
| 1807 |
+
"loss": 0.121,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 1.6518607442977191,
|
| 1812 |
+
"grad_norm": 5.493549571701205,
|
| 1813 |
+
"learning_rate": 1.7799792455209016e-07,
|
| 1814 |
+
"loss": 0.1702,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.6582633053221287,
|
| 1819 |
+
"grad_norm": 2.795867026435785,
|
| 1820 |
+
"learning_rate": 1.716601970199836e-07,
|
| 1821 |
+
"loss": 0.1657,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 1.6646658663465386,
|
| 1826 |
+
"grad_norm": 2.697456578390499,
|
| 1827 |
+
"learning_rate": 1.6542674627869734e-07,
|
| 1828 |
+
"loss": 0.1108,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 1.6710684273709484,
|
| 1833 |
+
"grad_norm": 3.700105341120021,
|
| 1834 |
+
"learning_rate": 1.592983570354699e-07,
|
| 1835 |
+
"loss": 0.1526,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 1.677470988395358,
|
| 1840 |
+
"grad_norm": 2.8767911269966957,
|
| 1841 |
+
"learning_rate": 1.5327580077171588e-07,
|
| 1842 |
+
"loss": 0.1581,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 1.6838735494197679,
|
| 1847 |
+
"grad_norm": 2.8392553114328183,
|
| 1848 |
+
"learning_rate": 1.473598356459078e-07,
|
| 1849 |
+
"loss": 0.1387,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 1.6902761104441777,
|
| 1854 |
+
"grad_norm": 2.848243119062683,
|
| 1855 |
+
"learning_rate": 1.415512063981339e-07,
|
| 1856 |
+
"loss": 0.1352,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.6966786714685873,
|
| 1861 |
+
"grad_norm": 4.008263998101646,
|
| 1862 |
+
"learning_rate": 1.358506442563454e-07,
|
| 1863 |
+
"loss": 0.1717,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 1.7030812324929971,
|
| 1868 |
+
"grad_norm": 3.3644795533763747,
|
| 1869 |
+
"learning_rate": 1.3025886684430465e-07,
|
| 1870 |
+
"loss": 0.1851,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 1.709483793517407,
|
| 1875 |
+
"grad_norm": 2.1927113979713457,
|
| 1876 |
+
"learning_rate": 1.2477657809124632e-07,
|
| 1877 |
+
"loss": 0.1529,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 1.7158863545418166,
|
| 1882 |
+
"grad_norm": 4.20748377167738,
|
| 1883 |
+
"learning_rate": 1.19404468143262e-07,
|
| 1884 |
+
"loss": 0.1506,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 1.7222889155662267,
|
| 1889 |
+
"grad_norm": 2.6119161837028897,
|
| 1890 |
+
"learning_rate": 1.1414321327642019e-07,
|
| 1891 |
+
"loss": 0.1692,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 1.7286914765906363,
|
| 1896 |
+
"grad_norm": 3.25899049531815,
|
| 1897 |
+
"learning_rate": 1.089934758116322e-07,
|
| 1898 |
+
"loss": 0.1552,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.7350940376150459,
|
| 1903 |
+
"grad_norm": 2.4457520889715085,
|
| 1904 |
+
"learning_rate": 1.0395590403127486e-07,
|
| 1905 |
+
"loss": 0.1072,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 1.741496598639456,
|
| 1910 |
+
"grad_norm": 3.725879684312294,
|
| 1911 |
+
"learning_rate": 9.903113209758096e-08,
|
| 1912 |
+
"loss": 0.1297,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 1.7478991596638656,
|
| 1917 |
+
"grad_norm": 4.086760833693879,
|
| 1918 |
+
"learning_rate": 9.421977997280594e-08,
|
| 1919 |
+
"loss": 0.134,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 1.7543017206882752,
|
| 1924 |
+
"grad_norm": 2.8228633836720944,
|
| 1925 |
+
"learning_rate": 8.952245334118413e-08,
|
| 1926 |
+
"loss": 0.1429,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 1.7607042817126852,
|
| 1931 |
+
"grad_norm": 2.5768217353103906,
|
| 1932 |
+
"learning_rate": 8.493974353268019e-08,
|
| 1933 |
+
"loss": 0.138,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 1.7671068427370948,
|
| 1938 |
+
"grad_norm": 3.920988018065823,
|
| 1939 |
+
"learning_rate": 8.047222744854942e-08,
|
| 1940 |
+
"loss": 0.168,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 1.7735094037615045,
|
| 1945 |
+
"grad_norm": 2.665420920196723,
|
| 1946 |
+
"learning_rate": 7.612046748871326e-08,
|
| 1947 |
+
"loss": 0.1198,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 1.7799119647859145,
|
| 1952 |
+
"grad_norm": 2.84264633693513,
|
| 1953 |
+
"learning_rate": 7.188501148096116e-08,
|
| 1954 |
+
"loss": 0.1724,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 1.7863145258103241,
|
| 1959 |
+
"grad_norm": 2.7132283377188386,
|
| 1960 |
+
"learning_rate": 6.77663926119858e-08,
|
| 1961 |
+
"loss": 0.1369,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 1.7927170868347337,
|
| 1966 |
+
"grad_norm": 2.8632960787364494,
|
| 1967 |
+
"learning_rate": 6.376512936026279e-08,
|
| 1968 |
+
"loss": 0.1606,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 1.7991196478591438,
|
| 1973 |
+
"grad_norm": 2.738598278709088,
|
| 1974 |
+
"learning_rate": 5.988172543078096e-08,
|
| 1975 |
+
"loss": 0.1679,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 1.8055222088835534,
|
| 1980 |
+
"grad_norm": 2.311034901841601,
|
| 1981 |
+
"learning_rate": 5.611666969163242e-08,
|
| 1982 |
+
"loss": 0.1644,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 1.811924769907963,
|
| 1987 |
+
"grad_norm": 2.8470945573238837,
|
| 1988 |
+
"learning_rate": 5.2470436112471264e-08,
|
| 1989 |
+
"loss": 0.1817,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 1.818327330932373,
|
| 1994 |
+
"grad_norm": 2.9917534547915476,
|
| 1995 |
+
"learning_rate": 4.8943483704846465e-08,
|
| 1996 |
+
"loss": 0.1472,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 1.8247298919567827,
|
| 2001 |
+
"grad_norm": 2.6305242577097685,
|
| 2002 |
+
"learning_rate": 4.553625646441928e-08,
|
| 2003 |
+
"loss": 0.1289,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 1.8311324529811923,
|
| 2008 |
+
"grad_norm": 2.5380081332282014,
|
| 2009 |
+
"learning_rate": 4.224918331506955e-08,
|
| 2010 |
+
"loss": 0.1379,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 1.8375350140056024,
|
| 2015 |
+
"grad_norm": 3.110480752833179,
|
| 2016 |
+
"learning_rate": 3.908267805490051e-08,
|
| 2017 |
+
"loss": 0.1623,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 1.843937575030012,
|
| 2022 |
+
"grad_norm": 3.1789215828737074,
|
| 2023 |
+
"learning_rate": 3.6037139304146756e-08,
|
| 2024 |
+
"loss": 0.1441,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 1.8503401360544216,
|
| 2029 |
+
"grad_norm": 3.2716298643823927,
|
| 2030 |
+
"learning_rate": 3.3112950454993625e-08,
|
| 2031 |
+
"loss": 0.166,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 1.8567426970788317,
|
| 2036 |
+
"grad_norm": 4.5463007787733725,
|
| 2037 |
+
"learning_rate": 3.0310479623313125e-08,
|
| 2038 |
+
"loss": 0.1391,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 1.8631452581032413,
|
| 2043 |
+
"grad_norm": 3.0644978198504966,
|
| 2044 |
+
"learning_rate": 2.7630079602323443e-08,
|
| 2045 |
+
"loss": 0.1359,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 1.8695478191276511,
|
| 2050 |
+
"grad_norm": 2.915540706565416,
|
| 2051 |
+
"learning_rate": 2.507208781817638e-08,
|
| 2052 |
+
"loss": 0.1854,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 1.875950380152061,
|
| 2057 |
+
"grad_norm": 2.7471867583483958,
|
| 2058 |
+
"learning_rate": 2.263682628748087e-08,
|
| 2059 |
+
"loss": 0.1593,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 1.8823529411764706,
|
| 2064 |
+
"grad_norm": 3.3996282117269323,
|
| 2065 |
+
"learning_rate": 2.032460157676452e-08,
|
| 2066 |
+
"loss": 0.1655,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 1.8887555022008804,
|
| 2071 |
+
"grad_norm": 2.782949881418279,
|
| 2072 |
+
"learning_rate": 1.8135704763881598e-08,
|
| 2073 |
+
"loss": 0.1411,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 1.8951580632252902,
|
| 2078 |
+
"grad_norm": 2.817009087504296,
|
| 2079 |
+
"learning_rate": 1.607041140137033e-08,
|
| 2080 |
+
"loss": 0.1741,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 1.9015606242496998,
|
| 2085 |
+
"grad_norm": 2.514339605431768,
|
| 2086 |
+
"learning_rate": 1.4128981481764113e-08,
|
| 2087 |
+
"loss": 0.1639,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 1.9079631852741097,
|
| 2092 |
+
"grad_norm": 2.502723430413198,
|
| 2093 |
+
"learning_rate": 1.231165940486234e-08,
|
| 2094 |
+
"loss": 0.1458,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 1.9143657462985195,
|
| 2099 |
+
"grad_norm": 2.9756989992004,
|
| 2100 |
+
"learning_rate": 1.0618673946963364e-08,
|
| 2101 |
+
"loss": 0.1899,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 1.9207683073229291,
|
| 2106 |
+
"grad_norm": 3.029713811900019,
|
| 2107 |
+
"learning_rate": 9.050238232065299e-09,
|
| 2108 |
+
"loss": 0.1194,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 1.927170868347339,
|
| 2113 |
+
"grad_norm": 2.8129240654015555,
|
| 2114 |
+
"learning_rate": 7.606549705035935e-09,
|
| 2115 |
+
"loss": 0.1564,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 1.9335734293717488,
|
| 2120 |
+
"grad_norm": 3.5641124301233904,
|
| 2121 |
+
"learning_rate": 6.2877901067573955e-09,
|
| 2122 |
+
"loss": 0.1272,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 1.9399759903961584,
|
| 2127 |
+
"grad_norm": 2.5825170270283024,
|
| 2128 |
+
"learning_rate": 5.094125451247655e-09,
|
| 2129 |
+
"loss": 0.1864,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 1.9463785514205683,
|
| 2134 |
+
"grad_norm": 2.75200131179709,
|
| 2135 |
+
"learning_rate": 4.025706004760931e-09,
|
| 2136 |
+
"loss": 0.1568,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 1.952781112444978,
|
| 2141 |
+
"grad_norm": 3.1581337599381527,
|
| 2142 |
+
"learning_rate": 3.082666266872036e-09,
|
| 2143 |
+
"loss": 0.1256,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 1.9591836734693877,
|
| 2148 |
+
"grad_norm": 3.4365108386154786,
|
| 2149 |
+
"learning_rate": 2.2651249535439177e-09,
|
| 2150 |
+
"loss": 0.1516,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 1.9655862344937975,
|
| 2155 |
+
"grad_norm": 2.552383623057842,
|
| 2156 |
+
"learning_rate": 1.5731849821833953e-09,
|
| 2157 |
+
"loss": 0.1362,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 1.9719887955182074,
|
| 2162 |
+
"grad_norm": 2.5409532995249657,
|
| 2163 |
+
"learning_rate": 1.0069334586854105e-09,
|
| 2164 |
+
"loss": 0.1238,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 1.978391356542617,
|
| 2169 |
+
"grad_norm": 4.158643521191336,
|
| 2170 |
+
"learning_rate": 5.664416664666882e-10,
|
| 2171 |
+
"loss": 0.1615,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 1.9847939175670268,
|
| 2176 |
+
"grad_norm": 3.0805263918986983,
|
| 2177 |
+
"learning_rate": 2.517650574934693e-10,
|
| 2178 |
+
"loss": 0.1513,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 1.9911964785914367,
|
| 2183 |
+
"grad_norm": 3.1343235054789926,
|
| 2184 |
+
"learning_rate": 6.29432452994294e-11,
|
| 2185 |
+
"loss": 0.1493,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 1.9975990396158463,
|
| 2190 |
+
"grad_norm": 4.792054679066387,
|
| 2191 |
+
"learning_rate": 0.0,
|
| 2192 |
+
"loss": 0.167,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.9975990396158463,
|
| 2197 |
+
"step": 312,
|
| 2198 |
+
"total_flos": 211327818203136.0,
|
| 2199 |
+
"train_loss": 0.17435487460058469,
|
| 2200 |
+
"train_runtime": 7374.5307,
|
| 2201 |
+
"train_samples_per_second": 10.842,
|
| 2202 |
+
"train_steps_per_second": 0.042
|
| 2203 |
+
}
|
| 2204 |
+
],
|
| 2205 |
+
"logging_steps": 1,
|
| 2206 |
+
"max_steps": 312,
|
| 2207 |
+
"num_input_tokens_seen": 0,
|
| 2208 |
+
"num_train_epochs": 2,
|
| 2209 |
+
"save_steps": 100,
|
| 2210 |
+
"stateful_callbacks": {
|
| 2211 |
+
"TrainerControl": {
|
| 2212 |
+
"args": {
|
| 2213 |
+
"should_epoch_stop": false,
|
| 2214 |
+
"should_evaluate": false,
|
| 2215 |
+
"should_log": false,
|
| 2216 |
+
"should_save": true,
|
| 2217 |
+
"should_training_stop": true
|
| 2218 |
+
},
|
| 2219 |
+
"attributes": {}
|
| 2220 |
+
}
|
| 2221 |
+
},
|
| 2222 |
+
"total_flos": 211327818203136.0,
|
| 2223 |
+
"train_batch_size": 2,
|
| 2224 |
+
"trial_name": null,
|
| 2225 |
+
"trial_params": null
|
| 2226 |
+
}
|