Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +2245 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,2245 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 14.962593516209477,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.04987531172069826,
|
| 13 |
+
"grad_norm": 39.54836654663086,
|
| 14 |
+
"learning_rate": 1.9940000000000002e-05,
|
| 15 |
+
"loss": 7.4096,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.09975062344139651,
|
| 20 |
+
"grad_norm": 47.2942008972168,
|
| 21 |
+
"learning_rate": 1.9873333333333335e-05,
|
| 22 |
+
"loss": 5.883,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.14962593516209477,
|
| 27 |
+
"grad_norm": 44.06085968017578,
|
| 28 |
+
"learning_rate": 1.9806666666666668e-05,
|
| 29 |
+
"loss": 3.9265,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.19950124688279303,
|
| 34 |
+
"grad_norm": 19.404735565185547,
|
| 35 |
+
"learning_rate": 1.974e-05,
|
| 36 |
+
"loss": 1.9858,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.24937655860349128,
|
| 41 |
+
"grad_norm": 1.9771770238876343,
|
| 42 |
+
"learning_rate": 1.9673333333333337e-05,
|
| 43 |
+
"loss": 1.0943,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.29925187032418954,
|
| 48 |
+
"grad_norm": 0.6882551312446594,
|
| 49 |
+
"learning_rate": 1.9606666666666666e-05,
|
| 50 |
+
"loss": 0.9006,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.3491271820448878,
|
| 55 |
+
"grad_norm": 0.7775446772575378,
|
| 56 |
+
"learning_rate": 1.9540000000000003e-05,
|
| 57 |
+
"loss": 0.8786,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.39900249376558605,
|
| 62 |
+
"grad_norm": 0.720007598400116,
|
| 63 |
+
"learning_rate": 1.9473333333333335e-05,
|
| 64 |
+
"loss": 0.858,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.4488778054862843,
|
| 69 |
+
"grad_norm": 0.6773384809494019,
|
| 70 |
+
"learning_rate": 1.940666666666667e-05,
|
| 71 |
+
"loss": 0.811,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.49875311720698257,
|
| 76 |
+
"grad_norm": 0.6907929182052612,
|
| 77 |
+
"learning_rate": 1.934e-05,
|
| 78 |
+
"loss": 0.7896,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.5486284289276808,
|
| 83 |
+
"grad_norm": 0.7552163600921631,
|
| 84 |
+
"learning_rate": 1.9273333333333334e-05,
|
| 85 |
+
"loss": 0.7589,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.5985037406483791,
|
| 90 |
+
"grad_norm": 0.7480026483535767,
|
| 91 |
+
"learning_rate": 1.920666666666667e-05,
|
| 92 |
+
"loss": 0.7347,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.6483790523690773,
|
| 97 |
+
"grad_norm": 0.9136413931846619,
|
| 98 |
+
"learning_rate": 1.914e-05,
|
| 99 |
+
"loss": 0.7218,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.6982543640897756,
|
| 104 |
+
"grad_norm": 0.7177766561508179,
|
| 105 |
+
"learning_rate": 1.9073333333333336e-05,
|
| 106 |
+
"loss": 0.6983,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.7481296758104738,
|
| 111 |
+
"grad_norm": 0.7291643023490906,
|
| 112 |
+
"learning_rate": 1.900666666666667e-05,
|
| 113 |
+
"loss": 0.6811,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.7980049875311721,
|
| 118 |
+
"grad_norm": 0.8402264714241028,
|
| 119 |
+
"learning_rate": 1.894e-05,
|
| 120 |
+
"loss": 0.6518,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.8478802992518704,
|
| 125 |
+
"grad_norm": 0.7445259690284729,
|
| 126 |
+
"learning_rate": 1.8873333333333334e-05,
|
| 127 |
+
"loss": 0.6621,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.8977556109725686,
|
| 132 |
+
"grad_norm": 0.7637977004051208,
|
| 133 |
+
"learning_rate": 1.8806666666666667e-05,
|
| 134 |
+
"loss": 0.623,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.9476309226932669,
|
| 139 |
+
"grad_norm": 0.7605411410331726,
|
| 140 |
+
"learning_rate": 1.8740000000000004e-05,
|
| 141 |
+
"loss": 0.642,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.9975062344139651,
|
| 146 |
+
"grad_norm": 0.7474659085273743,
|
| 147 |
+
"learning_rate": 1.8673333333333333e-05,
|
| 148 |
+
"loss": 0.6081,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.9975062344139651,
|
| 153 |
+
"eval_loss": 0.6107227802276611,
|
| 154 |
+
"eval_runtime": 57.9983,
|
| 155 |
+
"eval_samples_per_second": 17.294,
|
| 156 |
+
"eval_steps_per_second": 3.466,
|
| 157 |
+
"step": 200
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.0473815461346634,
|
| 161 |
+
"grad_norm": 0.7955527901649475,
|
| 162 |
+
"learning_rate": 1.860666666666667e-05,
|
| 163 |
+
"loss": 0.6313,
|
| 164 |
+
"step": 210
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 1.0972568578553616,
|
| 168 |
+
"grad_norm": 0.8842599391937256,
|
| 169 |
+
"learning_rate": 1.8540000000000002e-05,
|
| 170 |
+
"loss": 0.5837,
|
| 171 |
+
"step": 220
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.14713216957606,
|
| 175 |
+
"grad_norm": 0.9143279790878296,
|
| 176 |
+
"learning_rate": 1.8473333333333335e-05,
|
| 177 |
+
"loss": 0.5668,
|
| 178 |
+
"step": 230
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.1970074812967582,
|
| 182 |
+
"grad_norm": 0.8860597610473633,
|
| 183 |
+
"learning_rate": 1.8406666666666668e-05,
|
| 184 |
+
"loss": 0.565,
|
| 185 |
+
"step": 240
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.2468827930174564,
|
| 189 |
+
"grad_norm": 0.8665011525154114,
|
| 190 |
+
"learning_rate": 1.834e-05,
|
| 191 |
+
"loss": 0.5611,
|
| 192 |
+
"step": 250
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.2967581047381547,
|
| 196 |
+
"grad_norm": 0.8916261196136475,
|
| 197 |
+
"learning_rate": 1.8273333333333333e-05,
|
| 198 |
+
"loss": 0.5461,
|
| 199 |
+
"step": 260
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.346633416458853,
|
| 203 |
+
"grad_norm": 0.8792281150817871,
|
| 204 |
+
"learning_rate": 1.820666666666667e-05,
|
| 205 |
+
"loss": 0.5569,
|
| 206 |
+
"step": 270
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.3965087281795512,
|
| 210 |
+
"grad_norm": 0.8166815638542175,
|
| 211 |
+
"learning_rate": 1.8140000000000003e-05,
|
| 212 |
+
"loss": 0.5435,
|
| 213 |
+
"step": 280
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.4463840399002494,
|
| 217 |
+
"grad_norm": 0.9711934328079224,
|
| 218 |
+
"learning_rate": 1.8073333333333335e-05,
|
| 219 |
+
"loss": 0.5205,
|
| 220 |
+
"step": 290
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.4962593516209477,
|
| 224 |
+
"grad_norm": 0.8851374983787537,
|
| 225 |
+
"learning_rate": 1.8006666666666668e-05,
|
| 226 |
+
"loss": 0.5269,
|
| 227 |
+
"step": 300
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.546134663341646,
|
| 231 |
+
"grad_norm": 0.940987765789032,
|
| 232 |
+
"learning_rate": 1.794e-05,
|
| 233 |
+
"loss": 0.5177,
|
| 234 |
+
"step": 310
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.5960099750623442,
|
| 238 |
+
"grad_norm": 1.1342016458511353,
|
| 239 |
+
"learning_rate": 1.7873333333333337e-05,
|
| 240 |
+
"loss": 0.4981,
|
| 241 |
+
"step": 320
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.6458852867830425,
|
| 245 |
+
"grad_norm": 1.0847837924957275,
|
| 246 |
+
"learning_rate": 1.7806666666666667e-05,
|
| 247 |
+
"loss": 0.5056,
|
| 248 |
+
"step": 330
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.6957605985037407,
|
| 252 |
+
"grad_norm": 0.979576587677002,
|
| 253 |
+
"learning_rate": 1.7740000000000003e-05,
|
| 254 |
+
"loss": 0.5048,
|
| 255 |
+
"step": 340
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.745635910224439,
|
| 259 |
+
"grad_norm": 0.9637285470962524,
|
| 260 |
+
"learning_rate": 1.7673333333333336e-05,
|
| 261 |
+
"loss": 0.5076,
|
| 262 |
+
"step": 350
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.7955112219451372,
|
| 266 |
+
"grad_norm": 0.9592918157577515,
|
| 267 |
+
"learning_rate": 1.760666666666667e-05,
|
| 268 |
+
"loss": 0.512,
|
| 269 |
+
"step": 360
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.8453865336658355,
|
| 273 |
+
"grad_norm": 0.9521375894546509,
|
| 274 |
+
"learning_rate": 1.754e-05,
|
| 275 |
+
"loss": 0.4993,
|
| 276 |
+
"step": 370
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.8952618453865338,
|
| 280 |
+
"grad_norm": 1.1952577829360962,
|
| 281 |
+
"learning_rate": 1.7473333333333334e-05,
|
| 282 |
+
"loss": 0.4933,
|
| 283 |
+
"step": 380
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 1.945137157107232,
|
| 287 |
+
"grad_norm": 0.8078787326812744,
|
| 288 |
+
"learning_rate": 1.7406666666666667e-05,
|
| 289 |
+
"loss": 0.4894,
|
| 290 |
+
"step": 390
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.9950124688279303,
|
| 294 |
+
"grad_norm": 1.0912864208221436,
|
| 295 |
+
"learning_rate": 1.734e-05,
|
| 296 |
+
"loss": 0.4865,
|
| 297 |
+
"step": 400
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.0,
|
| 301 |
+
"eval_loss": 0.49822476506233215,
|
| 302 |
+
"eval_runtime": 57.9997,
|
| 303 |
+
"eval_samples_per_second": 17.293,
|
| 304 |
+
"eval_steps_per_second": 3.466,
|
| 305 |
+
"step": 401
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.0448877805486285,
|
| 309 |
+
"grad_norm": 0.9705982804298401,
|
| 310 |
+
"learning_rate": 1.7273333333333336e-05,
|
| 311 |
+
"loss": 0.4976,
|
| 312 |
+
"step": 410
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.0947630922693268,
|
| 316 |
+
"grad_norm": 1.0686686038970947,
|
| 317 |
+
"learning_rate": 1.7206666666666666e-05,
|
| 318 |
+
"loss": 0.4851,
|
| 319 |
+
"step": 420
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.144638403990025,
|
| 323 |
+
"grad_norm": 1.0422980785369873,
|
| 324 |
+
"learning_rate": 1.7140000000000002e-05,
|
| 325 |
+
"loss": 0.4577,
|
| 326 |
+
"step": 430
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.1945137157107233,
|
| 330 |
+
"grad_norm": 1.0678831338882446,
|
| 331 |
+
"learning_rate": 1.7073333333333335e-05,
|
| 332 |
+
"loss": 0.5092,
|
| 333 |
+
"step": 440
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 2.2443890274314215,
|
| 337 |
+
"grad_norm": 1.1621687412261963,
|
| 338 |
+
"learning_rate": 1.7006666666666668e-05,
|
| 339 |
+
"loss": 0.4897,
|
| 340 |
+
"step": 450
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 2.29426433915212,
|
| 344 |
+
"grad_norm": 0.9376991391181946,
|
| 345 |
+
"learning_rate": 1.694e-05,
|
| 346 |
+
"loss": 0.4788,
|
| 347 |
+
"step": 460
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 2.344139650872818,
|
| 351 |
+
"grad_norm": 1.0832690000534058,
|
| 352 |
+
"learning_rate": 1.6873333333333333e-05,
|
| 353 |
+
"loss": 0.4864,
|
| 354 |
+
"step": 470
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.3940149625935163,
|
| 358 |
+
"grad_norm": 1.1426657438278198,
|
| 359 |
+
"learning_rate": 1.680666666666667e-05,
|
| 360 |
+
"loss": 0.4659,
|
| 361 |
+
"step": 480
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.4438902743142146,
|
| 365 |
+
"grad_norm": 1.016119122505188,
|
| 366 |
+
"learning_rate": 1.6740000000000002e-05,
|
| 367 |
+
"loss": 0.4636,
|
| 368 |
+
"step": 490
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.493765586034913,
|
| 372 |
+
"grad_norm": 1.1347243785858154,
|
| 373 |
+
"learning_rate": 1.6673333333333335e-05,
|
| 374 |
+
"loss": 0.4663,
|
| 375 |
+
"step": 500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.543640897755611,
|
| 379 |
+
"grad_norm": 1.087796926498413,
|
| 380 |
+
"learning_rate": 1.6606666666666668e-05,
|
| 381 |
+
"loss": 0.4729,
|
| 382 |
+
"step": 510
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 2.5935162094763093,
|
| 386 |
+
"grad_norm": 1.1080952882766724,
|
| 387 |
+
"learning_rate": 1.654e-05,
|
| 388 |
+
"loss": 0.4803,
|
| 389 |
+
"step": 520
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.6433915211970076,
|
| 393 |
+
"grad_norm": 0.9999821186065674,
|
| 394 |
+
"learning_rate": 1.6473333333333334e-05,
|
| 395 |
+
"loss": 0.4872,
|
| 396 |
+
"step": 530
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 2.693266832917706,
|
| 400 |
+
"grad_norm": 1.2594393491744995,
|
| 401 |
+
"learning_rate": 1.640666666666667e-05,
|
| 402 |
+
"loss": 0.4708,
|
| 403 |
+
"step": 540
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 2.743142144638404,
|
| 407 |
+
"grad_norm": 1.012904167175293,
|
| 408 |
+
"learning_rate": 1.634e-05,
|
| 409 |
+
"loss": 0.46,
|
| 410 |
+
"step": 550
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.7930174563591024,
|
| 414 |
+
"grad_norm": 1.052252173423767,
|
| 415 |
+
"learning_rate": 1.6273333333333336e-05,
|
| 416 |
+
"loss": 0.4664,
|
| 417 |
+
"step": 560
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 2.8428927680798006,
|
| 421 |
+
"grad_norm": 1.2155933380126953,
|
| 422 |
+
"learning_rate": 1.620666666666667e-05,
|
| 423 |
+
"loss": 0.4518,
|
| 424 |
+
"step": 570
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 2.892768079800499,
|
| 428 |
+
"grad_norm": 1.064907193183899,
|
| 429 |
+
"learning_rate": 1.614e-05,
|
| 430 |
+
"loss": 0.4933,
|
| 431 |
+
"step": 580
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 2.942643391521197,
|
| 435 |
+
"grad_norm": 1.0846999883651733,
|
| 436 |
+
"learning_rate": 1.6073333333333334e-05,
|
| 437 |
+
"loss": 0.4724,
|
| 438 |
+
"step": 590
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 2.9925187032418954,
|
| 442 |
+
"grad_norm": 1.2878086566925049,
|
| 443 |
+
"learning_rate": 1.6006666666666667e-05,
|
| 444 |
+
"loss": 0.4895,
|
| 445 |
+
"step": 600
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 2.997506234413965,
|
| 449 |
+
"eval_loss": 0.47463858127593994,
|
| 450 |
+
"eval_runtime": 57.9983,
|
| 451 |
+
"eval_samples_per_second": 17.294,
|
| 452 |
+
"eval_steps_per_second": 3.466,
|
| 453 |
+
"step": 601
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 3.0423940149625937,
|
| 457 |
+
"grad_norm": 1.141946792602539,
|
| 458 |
+
"learning_rate": 1.5940000000000003e-05,
|
| 459 |
+
"loss": 0.4808,
|
| 460 |
+
"step": 610
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 3.092269326683292,
|
| 464 |
+
"grad_norm": 1.1112643480300903,
|
| 465 |
+
"learning_rate": 1.5873333333333333e-05,
|
| 466 |
+
"loss": 0.4525,
|
| 467 |
+
"step": 620
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 3.14214463840399,
|
| 471 |
+
"grad_norm": 1.0057584047317505,
|
| 472 |
+
"learning_rate": 1.580666666666667e-05,
|
| 473 |
+
"loss": 0.4699,
|
| 474 |
+
"step": 630
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 3.1920199501246884,
|
| 478 |
+
"grad_norm": 1.3252357244491577,
|
| 479 |
+
"learning_rate": 1.5740000000000002e-05,
|
| 480 |
+
"loss": 0.4631,
|
| 481 |
+
"step": 640
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 3.2418952618453867,
|
| 485 |
+
"grad_norm": 1.1567648649215698,
|
| 486 |
+
"learning_rate": 1.5673333333333335e-05,
|
| 487 |
+
"loss": 0.4559,
|
| 488 |
+
"step": 650
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 3.291770573566085,
|
| 492 |
+
"grad_norm": 1.0659315586090088,
|
| 493 |
+
"learning_rate": 1.5606666666666667e-05,
|
| 494 |
+
"loss": 0.4575,
|
| 495 |
+
"step": 660
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 3.341645885286783,
|
| 499 |
+
"grad_norm": 1.1061469316482544,
|
| 500 |
+
"learning_rate": 1.554e-05,
|
| 501 |
+
"loss": 0.4592,
|
| 502 |
+
"step": 670
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 3.3915211970074814,
|
| 506 |
+
"grad_norm": 1.342760443687439,
|
| 507 |
+
"learning_rate": 1.5473333333333333e-05,
|
| 508 |
+
"loss": 0.4459,
|
| 509 |
+
"step": 680
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 3.4413965087281797,
|
| 513 |
+
"grad_norm": 1.080640196800232,
|
| 514 |
+
"learning_rate": 1.5406666666666666e-05,
|
| 515 |
+
"loss": 0.4801,
|
| 516 |
+
"step": 690
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 3.491271820448878,
|
| 520 |
+
"grad_norm": 1.1660523414611816,
|
| 521 |
+
"learning_rate": 1.5340000000000002e-05,
|
| 522 |
+
"loss": 0.4536,
|
| 523 |
+
"step": 700
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 3.541147132169576,
|
| 527 |
+
"grad_norm": 1.0911084413528442,
|
| 528 |
+
"learning_rate": 1.5273333333333335e-05,
|
| 529 |
+
"loss": 0.4536,
|
| 530 |
+
"step": 710
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 3.5910224438902745,
|
| 534 |
+
"grad_norm": 1.1449761390686035,
|
| 535 |
+
"learning_rate": 1.5206666666666668e-05,
|
| 536 |
+
"loss": 0.4663,
|
| 537 |
+
"step": 720
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 3.6408977556109727,
|
| 541 |
+
"grad_norm": 1.0596632957458496,
|
| 542 |
+
"learning_rate": 1.514e-05,
|
| 543 |
+
"loss": 0.4633,
|
| 544 |
+
"step": 730
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 3.690773067331671,
|
| 548 |
+
"grad_norm": 1.1071261167526245,
|
| 549 |
+
"learning_rate": 1.5073333333333335e-05,
|
| 550 |
+
"loss": 0.4579,
|
| 551 |
+
"step": 740
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 3.7406483790523692,
|
| 555 |
+
"grad_norm": 1.3376258611679077,
|
| 556 |
+
"learning_rate": 1.5006666666666666e-05,
|
| 557 |
+
"loss": 0.4526,
|
| 558 |
+
"step": 750
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 3.7905236907730675,
|
| 562 |
+
"grad_norm": 1.2122935056686401,
|
| 563 |
+
"learning_rate": 1.4940000000000001e-05,
|
| 564 |
+
"loss": 0.4602,
|
| 565 |
+
"step": 760
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 3.8403990024937658,
|
| 569 |
+
"grad_norm": 1.1372263431549072,
|
| 570 |
+
"learning_rate": 1.4873333333333335e-05,
|
| 571 |
+
"loss": 0.4733,
|
| 572 |
+
"step": 770
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 3.890274314214464,
|
| 576 |
+
"grad_norm": 1.1192553043365479,
|
| 577 |
+
"learning_rate": 1.4806666666666667e-05,
|
| 578 |
+
"loss": 0.466,
|
| 579 |
+
"step": 780
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 3.9401496259351623,
|
| 583 |
+
"grad_norm": 1.1695116758346558,
|
| 584 |
+
"learning_rate": 1.4740000000000001e-05,
|
| 585 |
+
"loss": 0.4485,
|
| 586 |
+
"step": 790
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 3.9900249376558605,
|
| 590 |
+
"grad_norm": 1.2374085187911987,
|
| 591 |
+
"learning_rate": 1.4673333333333336e-05,
|
| 592 |
+
"loss": 0.4428,
|
| 593 |
+
"step": 800
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 4.0,
|
| 597 |
+
"eval_loss": 0.4639199376106262,
|
| 598 |
+
"eval_runtime": 58.0007,
|
| 599 |
+
"eval_samples_per_second": 17.293,
|
| 600 |
+
"eval_steps_per_second": 3.465,
|
| 601 |
+
"step": 802
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 4.039900249376559,
|
| 605 |
+
"grad_norm": 1.2248201370239258,
|
| 606 |
+
"learning_rate": 1.4606666666666667e-05,
|
| 607 |
+
"loss": 0.4765,
|
| 608 |
+
"step": 810
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 4.089775561097257,
|
| 612 |
+
"grad_norm": 1.0709633827209473,
|
| 613 |
+
"learning_rate": 1.4540000000000001e-05,
|
| 614 |
+
"loss": 0.458,
|
| 615 |
+
"step": 820
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 4.139650872817955,
|
| 619 |
+
"grad_norm": 1.245808482170105,
|
| 620 |
+
"learning_rate": 1.4473333333333334e-05,
|
| 621 |
+
"loss": 0.4413,
|
| 622 |
+
"step": 830
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 4.1895261845386536,
|
| 626 |
+
"grad_norm": 1.0952578783035278,
|
| 627 |
+
"learning_rate": 1.4406666666666669e-05,
|
| 628 |
+
"loss": 0.4466,
|
| 629 |
+
"step": 840
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 4.239401496259352,
|
| 633 |
+
"grad_norm": 1.3013060092926025,
|
| 634 |
+
"learning_rate": 1.434e-05,
|
| 635 |
+
"loss": 0.4579,
|
| 636 |
+
"step": 850
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 4.28927680798005,
|
| 640 |
+
"grad_norm": 1.40675687789917,
|
| 641 |
+
"learning_rate": 1.4273333333333334e-05,
|
| 642 |
+
"loss": 0.4547,
|
| 643 |
+
"step": 860
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 4.339152119700748,
|
| 647 |
+
"grad_norm": 1.0993843078613281,
|
| 648 |
+
"learning_rate": 1.4206666666666669e-05,
|
| 649 |
+
"loss": 0.4326,
|
| 650 |
+
"step": 870
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 4.389027431421447,
|
| 654 |
+
"grad_norm": 1.279954195022583,
|
| 655 |
+
"learning_rate": 1.414e-05,
|
| 656 |
+
"loss": 0.4528,
|
| 657 |
+
"step": 880
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 4.438902743142145,
|
| 661 |
+
"grad_norm": 1.1767985820770264,
|
| 662 |
+
"learning_rate": 1.4073333333333335e-05,
|
| 663 |
+
"loss": 0.4678,
|
| 664 |
+
"step": 890
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 4.488778054862843,
|
| 668 |
+
"grad_norm": 1.2366374731063843,
|
| 669 |
+
"learning_rate": 1.400666666666667e-05,
|
| 670 |
+
"loss": 0.4383,
|
| 671 |
+
"step": 900
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 4.538653366583541,
|
| 675 |
+
"grad_norm": 1.3890389204025269,
|
| 676 |
+
"learning_rate": 1.394e-05,
|
| 677 |
+
"loss": 0.4586,
|
| 678 |
+
"step": 910
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 4.58852867830424,
|
| 682 |
+
"grad_norm": 1.288946509361267,
|
| 683 |
+
"learning_rate": 1.3873333333333335e-05,
|
| 684 |
+
"loss": 0.4329,
|
| 685 |
+
"step": 920
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 4.638403990024938,
|
| 689 |
+
"grad_norm": 1.5377354621887207,
|
| 690 |
+
"learning_rate": 1.3806666666666668e-05,
|
| 691 |
+
"loss": 0.4524,
|
| 692 |
+
"step": 930
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 4.688279301745636,
|
| 696 |
+
"grad_norm": 1.1805988550186157,
|
| 697 |
+
"learning_rate": 1.3740000000000002e-05,
|
| 698 |
+
"loss": 0.4421,
|
| 699 |
+
"step": 940
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 4.738154613466334,
|
| 703 |
+
"grad_norm": 1.2433724403381348,
|
| 704 |
+
"learning_rate": 1.3673333333333333e-05,
|
| 705 |
+
"loss": 0.4463,
|
| 706 |
+
"step": 950
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 4.788029925187033,
|
| 710 |
+
"grad_norm": 1.5534576177597046,
|
| 711 |
+
"learning_rate": 1.3606666666666668e-05,
|
| 712 |
+
"loss": 0.4529,
|
| 713 |
+
"step": 960
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 4.837905236907731,
|
| 717 |
+
"grad_norm": 1.1943343877792358,
|
| 718 |
+
"learning_rate": 1.3540000000000003e-05,
|
| 719 |
+
"loss": 0.4342,
|
| 720 |
+
"step": 970
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 4.887780548628429,
|
| 724 |
+
"grad_norm": 1.4089887142181396,
|
| 725 |
+
"learning_rate": 1.3473333333333334e-05,
|
| 726 |
+
"loss": 0.4474,
|
| 727 |
+
"step": 980
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 4.937655860349127,
|
| 731 |
+
"grad_norm": 1.2032736539840698,
|
| 732 |
+
"learning_rate": 1.3406666666666668e-05,
|
| 733 |
+
"loss": 0.4367,
|
| 734 |
+
"step": 990
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 4.987531172069826,
|
| 738 |
+
"grad_norm": 1.2193266153335571,
|
| 739 |
+
"learning_rate": 1.3340000000000001e-05,
|
| 740 |
+
"loss": 0.4617,
|
| 741 |
+
"step": 1000
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 4.997506234413965,
|
| 745 |
+
"eval_loss": 0.457111656665802,
|
| 746 |
+
"eval_runtime": 58.0103,
|
| 747 |
+
"eval_samples_per_second": 17.29,
|
| 748 |
+
"eval_steps_per_second": 3.465,
|
| 749 |
+
"step": 1002
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 5.037406483790524,
|
| 753 |
+
"grad_norm": 1.2309563159942627,
|
| 754 |
+
"learning_rate": 1.3273333333333334e-05,
|
| 755 |
+
"loss": 0.4802,
|
| 756 |
+
"step": 1010
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 5.087281795511222,
|
| 760 |
+
"grad_norm": 1.3006610870361328,
|
| 761 |
+
"learning_rate": 1.3206666666666668e-05,
|
| 762 |
+
"loss": 0.4304,
|
| 763 |
+
"step": 1020
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 5.13715710723192,
|
| 767 |
+
"grad_norm": 1.1411958932876587,
|
| 768 |
+
"learning_rate": 1.3140000000000001e-05,
|
| 769 |
+
"loss": 0.4379,
|
| 770 |
+
"step": 1030
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 5.187032418952619,
|
| 774 |
+
"grad_norm": 1.489011287689209,
|
| 775 |
+
"learning_rate": 1.3073333333333334e-05,
|
| 776 |
+
"loss": 0.4377,
|
| 777 |
+
"step": 1040
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 5.236907730673317,
|
| 781 |
+
"grad_norm": 1.127456784248352,
|
| 782 |
+
"learning_rate": 1.3006666666666667e-05,
|
| 783 |
+
"loss": 0.4401,
|
| 784 |
+
"step": 1050
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 5.286783042394015,
|
| 788 |
+
"grad_norm": 1.1423261165618896,
|
| 789 |
+
"learning_rate": 1.2940000000000001e-05,
|
| 790 |
+
"loss": 0.4195,
|
| 791 |
+
"step": 1060
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 5.3366583541147135,
|
| 795 |
+
"grad_norm": 1.2664945125579834,
|
| 796 |
+
"learning_rate": 1.2873333333333336e-05,
|
| 797 |
+
"loss": 0.4421,
|
| 798 |
+
"step": 1070
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 5.386533665835412,
|
| 802 |
+
"grad_norm": 1.3833181858062744,
|
| 803 |
+
"learning_rate": 1.2806666666666667e-05,
|
| 804 |
+
"loss": 0.4493,
|
| 805 |
+
"step": 1080
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 5.43640897755611,
|
| 809 |
+
"grad_norm": 1.2505003213882446,
|
| 810 |
+
"learning_rate": 1.2740000000000002e-05,
|
| 811 |
+
"loss": 0.4322,
|
| 812 |
+
"step": 1090
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 5.486284289276808,
|
| 816 |
+
"grad_norm": 1.1786164045333862,
|
| 817 |
+
"learning_rate": 1.2673333333333335e-05,
|
| 818 |
+
"loss": 0.4474,
|
| 819 |
+
"step": 1100
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 5.5361596009975065,
|
| 823 |
+
"grad_norm": 1.3222135305404663,
|
| 824 |
+
"learning_rate": 1.2606666666666667e-05,
|
| 825 |
+
"loss": 0.4434,
|
| 826 |
+
"step": 1110
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 5.586034912718205,
|
| 830 |
+
"grad_norm": 1.1747450828552246,
|
| 831 |
+
"learning_rate": 1.254e-05,
|
| 832 |
+
"loss": 0.4577,
|
| 833 |
+
"step": 1120
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 5.635910224438903,
|
| 837 |
+
"grad_norm": 1.299611210823059,
|
| 838 |
+
"learning_rate": 1.2473333333333335e-05,
|
| 839 |
+
"loss": 0.4306,
|
| 840 |
+
"step": 1130
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 5.685785536159601,
|
| 844 |
+
"grad_norm": 1.353705883026123,
|
| 845 |
+
"learning_rate": 1.2406666666666668e-05,
|
| 846 |
+
"loss": 0.4438,
|
| 847 |
+
"step": 1140
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 5.7356608478802995,
|
| 851 |
+
"grad_norm": 1.195237398147583,
|
| 852 |
+
"learning_rate": 1.234e-05,
|
| 853 |
+
"loss": 0.4203,
|
| 854 |
+
"step": 1150
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 5.785536159600998,
|
| 858 |
+
"grad_norm": 1.4759230613708496,
|
| 859 |
+
"learning_rate": 1.2273333333333335e-05,
|
| 860 |
+
"loss": 0.4399,
|
| 861 |
+
"step": 1160
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 5.835411471321696,
|
| 865 |
+
"grad_norm": 1.1856341361999512,
|
| 866 |
+
"learning_rate": 1.2206666666666666e-05,
|
| 867 |
+
"loss": 0.4566,
|
| 868 |
+
"step": 1170
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 5.885286783042394,
|
| 872 |
+
"grad_norm": 1.369195818901062,
|
| 873 |
+
"learning_rate": 1.214e-05,
|
| 874 |
+
"loss": 0.4568,
|
| 875 |
+
"step": 1180
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 5.9351620947630925,
|
| 879 |
+
"grad_norm": 1.2702898979187012,
|
| 880 |
+
"learning_rate": 1.2073333333333335e-05,
|
| 881 |
+
"loss": 0.4372,
|
| 882 |
+
"step": 1190
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 5.985037406483791,
|
| 886 |
+
"grad_norm": 1.5332574844360352,
|
| 887 |
+
"learning_rate": 1.2006666666666668e-05,
|
| 888 |
+
"loss": 0.454,
|
| 889 |
+
"step": 1200
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 6.0,
|
| 893 |
+
"eval_loss": 0.45269879698753357,
|
| 894 |
+
"eval_runtime": 57.8329,
|
| 895 |
+
"eval_samples_per_second": 17.343,
|
| 896 |
+
"eval_steps_per_second": 3.476,
|
| 897 |
+
"step": 1203
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 6.034912718204489,
|
| 901 |
+
"grad_norm": 1.502089500427246,
|
| 902 |
+
"learning_rate": 1.1940000000000001e-05,
|
| 903 |
+
"loss": 0.4509,
|
| 904 |
+
"step": 1210
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 6.084788029925187,
|
| 908 |
+
"grad_norm": 1.1866357326507568,
|
| 909 |
+
"learning_rate": 1.1873333333333334e-05,
|
| 910 |
+
"loss": 0.4315,
|
| 911 |
+
"step": 1220
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 6.134663341645886,
|
| 915 |
+
"grad_norm": 1.2971949577331543,
|
| 916 |
+
"learning_rate": 1.1806666666666668e-05,
|
| 917 |
+
"loss": 0.4378,
|
| 918 |
+
"step": 1230
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 6.184538653366584,
|
| 922 |
+
"grad_norm": 1.3105353116989136,
|
| 923 |
+
"learning_rate": 1.1740000000000001e-05,
|
| 924 |
+
"loss": 0.4462,
|
| 925 |
+
"step": 1240
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 6.234413965087282,
|
| 929 |
+
"grad_norm": 1.6913713216781616,
|
| 930 |
+
"learning_rate": 1.1673333333333334e-05,
|
| 931 |
+
"loss": 0.4273,
|
| 932 |
+
"step": 1250
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 6.28428927680798,
|
| 936 |
+
"grad_norm": 1.2743034362792969,
|
| 937 |
+
"learning_rate": 1.1606666666666668e-05,
|
| 938 |
+
"loss": 0.4308,
|
| 939 |
+
"step": 1260
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 6.334164588528679,
|
| 943 |
+
"grad_norm": 1.2764904499053955,
|
| 944 |
+
"learning_rate": 1.154e-05,
|
| 945 |
+
"loss": 0.4416,
|
| 946 |
+
"step": 1270
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 6.384039900249377,
|
| 950 |
+
"grad_norm": 1.2375209331512451,
|
| 951 |
+
"learning_rate": 1.1473333333333334e-05,
|
| 952 |
+
"loss": 0.4279,
|
| 953 |
+
"step": 1280
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 6.433915211970075,
|
| 957 |
+
"grad_norm": 1.2813770771026611,
|
| 958 |
+
"learning_rate": 1.1406666666666669e-05,
|
| 959 |
+
"loss": 0.4318,
|
| 960 |
+
"step": 1290
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 6.483790523690773,
|
| 964 |
+
"grad_norm": 1.413662314414978,
|
| 965 |
+
"learning_rate": 1.134e-05,
|
| 966 |
+
"loss": 0.4299,
|
| 967 |
+
"step": 1300
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 6.533665835411472,
|
| 971 |
+
"grad_norm": 1.4377459287643433,
|
| 972 |
+
"learning_rate": 1.1273333333333334e-05,
|
| 973 |
+
"loss": 0.4538,
|
| 974 |
+
"step": 1310
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 6.58354114713217,
|
| 978 |
+
"grad_norm": 1.280629277229309,
|
| 979 |
+
"learning_rate": 1.1206666666666667e-05,
|
| 980 |
+
"loss": 0.4305,
|
| 981 |
+
"step": 1320
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 6.633416458852868,
|
| 985 |
+
"grad_norm": 1.4439893960952759,
|
| 986 |
+
"learning_rate": 1.1140000000000002e-05,
|
| 987 |
+
"loss": 0.4382,
|
| 988 |
+
"step": 1330
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 6.683291770573566,
|
| 992 |
+
"grad_norm": 1.290280818939209,
|
| 993 |
+
"learning_rate": 1.1073333333333333e-05,
|
| 994 |
+
"loss": 0.4395,
|
| 995 |
+
"step": 1340
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 6.733167082294265,
|
| 999 |
+
"grad_norm": 1.7553201913833618,
|
| 1000 |
+
"learning_rate": 1.1006666666666667e-05,
|
| 1001 |
+
"loss": 0.4633,
|
| 1002 |
+
"step": 1350
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 6.783042394014963,
|
| 1006 |
+
"grad_norm": 1.3317337036132812,
|
| 1007 |
+
"learning_rate": 1.0940000000000002e-05,
|
| 1008 |
+
"loss": 0.4396,
|
| 1009 |
+
"step": 1360
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 6.832917705735661,
|
| 1013 |
+
"grad_norm": 1.4611096382141113,
|
| 1014 |
+
"learning_rate": 1.0873333333333333e-05,
|
| 1015 |
+
"loss": 0.4306,
|
| 1016 |
+
"step": 1370
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 6.882793017456359,
|
| 1020 |
+
"grad_norm": 1.461767554283142,
|
| 1021 |
+
"learning_rate": 1.0806666666666668e-05,
|
| 1022 |
+
"loss": 0.4363,
|
| 1023 |
+
"step": 1380
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 6.932668329177058,
|
| 1027 |
+
"grad_norm": 1.2594436407089233,
|
| 1028 |
+
"learning_rate": 1.0740000000000002e-05,
|
| 1029 |
+
"loss": 0.4311,
|
| 1030 |
+
"step": 1390
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 6.982543640897756,
|
| 1034 |
+
"grad_norm": 1.339422345161438,
|
| 1035 |
+
"learning_rate": 1.0673333333333333e-05,
|
| 1036 |
+
"loss": 0.4359,
|
| 1037 |
+
"step": 1400
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 6.997506234413965,
|
| 1041 |
+
"eval_loss": 0.4493270516395569,
|
| 1042 |
+
"eval_runtime": 57.825,
|
| 1043 |
+
"eval_samples_per_second": 17.345,
|
| 1044 |
+
"eval_steps_per_second": 3.476,
|
| 1045 |
+
"step": 1403
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 7.032418952618454,
|
| 1049 |
+
"grad_norm": 1.5040308237075806,
|
| 1050 |
+
"learning_rate": 1.0606666666666668e-05,
|
| 1051 |
+
"loss": 0.4513,
|
| 1052 |
+
"step": 1410
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 7.082294264339152,
|
| 1056 |
+
"grad_norm": 1.3062398433685303,
|
| 1057 |
+
"learning_rate": 1.054e-05,
|
| 1058 |
+
"loss": 0.4286,
|
| 1059 |
+
"step": 1420
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 7.132169576059851,
|
| 1063 |
+
"grad_norm": 1.2743724584579468,
|
| 1064 |
+
"learning_rate": 1.0473333333333334e-05,
|
| 1065 |
+
"loss": 0.4308,
|
| 1066 |
+
"step": 1430
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 7.182044887780549,
|
| 1070 |
+
"grad_norm": 1.335893154144287,
|
| 1071 |
+
"learning_rate": 1.0406666666666666e-05,
|
| 1072 |
+
"loss": 0.4467,
|
| 1073 |
+
"step": 1440
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 7.231920199501247,
|
| 1077 |
+
"grad_norm": 1.5432289838790894,
|
| 1078 |
+
"learning_rate": 1.0340000000000001e-05,
|
| 1079 |
+
"loss": 0.4249,
|
| 1080 |
+
"step": 1450
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 7.2817955112219455,
|
| 1084 |
+
"grad_norm": 1.3191583156585693,
|
| 1085 |
+
"learning_rate": 1.0273333333333335e-05,
|
| 1086 |
+
"loss": 0.4308,
|
| 1087 |
+
"step": 1460
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 7.331670822942644,
|
| 1091 |
+
"grad_norm": 1.4070515632629395,
|
| 1092 |
+
"learning_rate": 1.0206666666666667e-05,
|
| 1093 |
+
"loss": 0.4346,
|
| 1094 |
+
"step": 1470
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 7.381546134663342,
|
| 1098 |
+
"grad_norm": 1.2089998722076416,
|
| 1099 |
+
"learning_rate": 1.0140000000000001e-05,
|
| 1100 |
+
"loss": 0.4261,
|
| 1101 |
+
"step": 1480
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 7.43142144638404,
|
| 1105 |
+
"grad_norm": 1.349908709526062,
|
| 1106 |
+
"learning_rate": 1.0073333333333336e-05,
|
| 1107 |
+
"loss": 0.4506,
|
| 1108 |
+
"step": 1490
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 7.4812967581047385,
|
| 1112 |
+
"grad_norm": 1.3772706985473633,
|
| 1113 |
+
"learning_rate": 1.0006666666666667e-05,
|
| 1114 |
+
"loss": 0.4194,
|
| 1115 |
+
"step": 1500
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 7.531172069825437,
|
| 1119 |
+
"grad_norm": 1.265548586845398,
|
| 1120 |
+
"learning_rate": 9.940000000000001e-06,
|
| 1121 |
+
"loss": 0.4109,
|
| 1122 |
+
"step": 1510
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 7.581047381546135,
|
| 1126 |
+
"grad_norm": 1.4737778902053833,
|
| 1127 |
+
"learning_rate": 9.873333333333334e-06,
|
| 1128 |
+
"loss": 0.4335,
|
| 1129 |
+
"step": 1520
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 7.630922693266833,
|
| 1133 |
+
"grad_norm": 1.2900265455245972,
|
| 1134 |
+
"learning_rate": 9.806666666666667e-06,
|
| 1135 |
+
"loss": 0.4489,
|
| 1136 |
+
"step": 1530
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 7.6807980049875315,
|
| 1140 |
+
"grad_norm": 1.3885109424591064,
|
| 1141 |
+
"learning_rate": 9.74e-06,
|
| 1142 |
+
"loss": 0.4246,
|
| 1143 |
+
"step": 1540
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 7.73067331670823,
|
| 1147 |
+
"grad_norm": 1.4786447286605835,
|
| 1148 |
+
"learning_rate": 9.673333333333334e-06,
|
| 1149 |
+
"loss": 0.4307,
|
| 1150 |
+
"step": 1550
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 7.780548628428928,
|
| 1154 |
+
"grad_norm": 1.4943320751190186,
|
| 1155 |
+
"learning_rate": 9.606666666666667e-06,
|
| 1156 |
+
"loss": 0.4272,
|
| 1157 |
+
"step": 1560
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 7.830423940149626,
|
| 1161 |
+
"grad_norm": 1.2281179428100586,
|
| 1162 |
+
"learning_rate": 9.54e-06,
|
| 1163 |
+
"loss": 0.4361,
|
| 1164 |
+
"step": 1570
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 7.8802992518703245,
|
| 1168 |
+
"grad_norm": 1.4099763631820679,
|
| 1169 |
+
"learning_rate": 9.473333333333335e-06,
|
| 1170 |
+
"loss": 0.4328,
|
| 1171 |
+
"step": 1580
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 7.930174563591023,
|
| 1175 |
+
"grad_norm": 1.2165679931640625,
|
| 1176 |
+
"learning_rate": 9.406666666666668e-06,
|
| 1177 |
+
"loss": 0.4325,
|
| 1178 |
+
"step": 1590
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 7.980049875311721,
|
| 1182 |
+
"grad_norm": 1.5678939819335938,
|
| 1183 |
+
"learning_rate": 9.340000000000002e-06,
|
| 1184 |
+
"loss": 0.4431,
|
| 1185 |
+
"step": 1600
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 8.0,
|
| 1189 |
+
"eval_loss": 0.4465142786502838,
|
| 1190 |
+
"eval_runtime": 57.9689,
|
| 1191 |
+
"eval_samples_per_second": 17.302,
|
| 1192 |
+
"eval_steps_per_second": 3.467,
|
| 1193 |
+
"step": 1604
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 8.029925187032418,
|
| 1197 |
+
"grad_norm": 1.3903255462646484,
|
| 1198 |
+
"learning_rate": 9.273333333333335e-06,
|
| 1199 |
+
"loss": 0.4489,
|
| 1200 |
+
"step": 1610
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 8.079800498753118,
|
| 1204 |
+
"grad_norm": 1.4039775133132935,
|
| 1205 |
+
"learning_rate": 9.206666666666668e-06,
|
| 1206 |
+
"loss": 0.4282,
|
| 1207 |
+
"step": 1620
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 8.129675810473815,
|
| 1211 |
+
"grad_norm": 1.238571286201477,
|
| 1212 |
+
"learning_rate": 9.14e-06,
|
| 1213 |
+
"loss": 0.4352,
|
| 1214 |
+
"step": 1630
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 8.179551122194514,
|
| 1218 |
+
"grad_norm": 1.418980360031128,
|
| 1219 |
+
"learning_rate": 9.073333333333333e-06,
|
| 1220 |
+
"loss": 0.4232,
|
| 1221 |
+
"step": 1640
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 8.229426433915211,
|
| 1225 |
+
"grad_norm": 1.362851619720459,
|
| 1226 |
+
"learning_rate": 9.006666666666666e-06,
|
| 1227 |
+
"loss": 0.4251,
|
| 1228 |
+
"step": 1650
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 8.27930174563591,
|
| 1232 |
+
"grad_norm": 1.3849656581878662,
|
| 1233 |
+
"learning_rate": 8.94e-06,
|
| 1234 |
+
"loss": 0.4367,
|
| 1235 |
+
"step": 1660
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 8.329177057356608,
|
| 1239 |
+
"grad_norm": 1.4366765022277832,
|
| 1240 |
+
"learning_rate": 8.873333333333334e-06,
|
| 1241 |
+
"loss": 0.4348,
|
| 1242 |
+
"step": 1670
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 8.379052369077307,
|
| 1246 |
+
"grad_norm": 1.3090085983276367,
|
| 1247 |
+
"learning_rate": 8.806666666666668e-06,
|
| 1248 |
+
"loss": 0.4349,
|
| 1249 |
+
"step": 1680
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 8.428927680798004,
|
| 1253 |
+
"grad_norm": 1.3340786695480347,
|
| 1254 |
+
"learning_rate": 8.740000000000001e-06,
|
| 1255 |
+
"loss": 0.4292,
|
| 1256 |
+
"step": 1690
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 8.478802992518704,
|
| 1260 |
+
"grad_norm": 1.224572777748108,
|
| 1261 |
+
"learning_rate": 8.673333333333334e-06,
|
| 1262 |
+
"loss": 0.4225,
|
| 1263 |
+
"step": 1700
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 8.528678304239401,
|
| 1267 |
+
"grad_norm": 1.3066871166229248,
|
| 1268 |
+
"learning_rate": 8.606666666666668e-06,
|
| 1269 |
+
"loss": 0.4295,
|
| 1270 |
+
"step": 1710
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 8.5785536159601,
|
| 1274 |
+
"grad_norm": 1.3138604164123535,
|
| 1275 |
+
"learning_rate": 8.540000000000001e-06,
|
| 1276 |
+
"loss": 0.4249,
|
| 1277 |
+
"step": 1720
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 8.628428927680797,
|
| 1281 |
+
"grad_norm": 1.3178144693374634,
|
| 1282 |
+
"learning_rate": 8.473333333333334e-06,
|
| 1283 |
+
"loss": 0.4295,
|
| 1284 |
+
"step": 1730
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 8.678304239401497,
|
| 1288 |
+
"grad_norm": 1.5216317176818848,
|
| 1289 |
+
"learning_rate": 8.406666666666667e-06,
|
| 1290 |
+
"loss": 0.4249,
|
| 1291 |
+
"step": 1740
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 8.728179551122194,
|
| 1295 |
+
"grad_norm": 1.3315293788909912,
|
| 1296 |
+
"learning_rate": 8.34e-06,
|
| 1297 |
+
"loss": 0.4156,
|
| 1298 |
+
"step": 1750
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 8.778054862842893,
|
| 1302 |
+
"grad_norm": 1.613593578338623,
|
| 1303 |
+
"learning_rate": 8.273333333333334e-06,
|
| 1304 |
+
"loss": 0.418,
|
| 1305 |
+
"step": 1760
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 8.82793017456359,
|
| 1309 |
+
"grad_norm": 1.2980856895446777,
|
| 1310 |
+
"learning_rate": 8.206666666666667e-06,
|
| 1311 |
+
"loss": 0.4234,
|
| 1312 |
+
"step": 1770
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 8.87780548628429,
|
| 1316 |
+
"grad_norm": 1.4451056718826294,
|
| 1317 |
+
"learning_rate": 8.14e-06,
|
| 1318 |
+
"loss": 0.4427,
|
| 1319 |
+
"step": 1780
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 8.927680798004987,
|
| 1323 |
+
"grad_norm": 1.3692457675933838,
|
| 1324 |
+
"learning_rate": 8.073333333333335e-06,
|
| 1325 |
+
"loss": 0.4394,
|
| 1326 |
+
"step": 1790
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 8.977556109725686,
|
| 1330 |
+
"grad_norm": 1.4997522830963135,
|
| 1331 |
+
"learning_rate": 8.006666666666667e-06,
|
| 1332 |
+
"loss": 0.4299,
|
| 1333 |
+
"step": 1800
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 8.997506234413965,
|
| 1337 |
+
"eval_loss": 0.4446782171726227,
|
| 1338 |
+
"eval_runtime": 57.956,
|
| 1339 |
+
"eval_samples_per_second": 17.306,
|
| 1340 |
+
"eval_steps_per_second": 3.468,
|
| 1341 |
+
"step": 1804
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 9.027431421446384,
|
| 1345 |
+
"grad_norm": 1.4114794731140137,
|
| 1346 |
+
"learning_rate": 7.94e-06,
|
| 1347 |
+
"loss": 0.4409,
|
| 1348 |
+
"step": 1810
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 9.077306733167083,
|
| 1352 |
+
"grad_norm": 1.2984321117401123,
|
| 1353 |
+
"learning_rate": 7.873333333333335e-06,
|
| 1354 |
+
"loss": 0.4091,
|
| 1355 |
+
"step": 1820
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 9.12718204488778,
|
| 1359 |
+
"grad_norm": 1.3625205755233765,
|
| 1360 |
+
"learning_rate": 7.806666666666668e-06,
|
| 1361 |
+
"loss": 0.4224,
|
| 1362 |
+
"step": 1830
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 9.17705735660848,
|
| 1366 |
+
"grad_norm": 1.4072176218032837,
|
| 1367 |
+
"learning_rate": 7.74e-06,
|
| 1368 |
+
"loss": 0.4416,
|
| 1369 |
+
"step": 1840
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 9.226932668329177,
|
| 1373 |
+
"grad_norm": 1.2547061443328857,
|
| 1374 |
+
"learning_rate": 7.673333333333333e-06,
|
| 1375 |
+
"loss": 0.4228,
|
| 1376 |
+
"step": 1850
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 9.276807980049876,
|
| 1380 |
+
"grad_norm": 1.5764844417572021,
|
| 1381 |
+
"learning_rate": 7.606666666666668e-06,
|
| 1382 |
+
"loss": 0.4346,
|
| 1383 |
+
"step": 1860
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 9.326683291770573,
|
| 1387 |
+
"grad_norm": 1.4493110179901123,
|
| 1388 |
+
"learning_rate": 7.540000000000001e-06,
|
| 1389 |
+
"loss": 0.438,
|
| 1390 |
+
"step": 1870
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 9.376558603491272,
|
| 1394 |
+
"grad_norm": 1.3886995315551758,
|
| 1395 |
+
"learning_rate": 7.4733333333333335e-06,
|
| 1396 |
+
"loss": 0.4282,
|
| 1397 |
+
"step": 1880
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 9.42643391521197,
|
| 1401 |
+
"grad_norm": 1.5260852575302124,
|
| 1402 |
+
"learning_rate": 7.406666666666667e-06,
|
| 1403 |
+
"loss": 0.4267,
|
| 1404 |
+
"step": 1890
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 9.476309226932669,
|
| 1408 |
+
"grad_norm": 1.7208826541900635,
|
| 1409 |
+
"learning_rate": 7.340000000000001e-06,
|
| 1410 |
+
"loss": 0.4195,
|
| 1411 |
+
"step": 1900
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 9.526184538653366,
|
| 1415 |
+
"grad_norm": 1.3822261095046997,
|
| 1416 |
+
"learning_rate": 7.2733333333333346e-06,
|
| 1417 |
+
"loss": 0.4254,
|
| 1418 |
+
"step": 1910
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 9.576059850374065,
|
| 1422 |
+
"grad_norm": 1.6141988039016724,
|
| 1423 |
+
"learning_rate": 7.206666666666667e-06,
|
| 1424 |
+
"loss": 0.4165,
|
| 1425 |
+
"step": 1920
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 9.625935162094763,
|
| 1429 |
+
"grad_norm": 1.4055383205413818,
|
| 1430 |
+
"learning_rate": 7.14e-06,
|
| 1431 |
+
"loss": 0.4149,
|
| 1432 |
+
"step": 1930
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 9.675810473815462,
|
| 1436 |
+
"grad_norm": 1.3142189979553223,
|
| 1437 |
+
"learning_rate": 7.073333333333334e-06,
|
| 1438 |
+
"loss": 0.4107,
|
| 1439 |
+
"step": 1940
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 9.72568578553616,
|
| 1443 |
+
"grad_norm": 1.2457393407821655,
|
| 1444 |
+
"learning_rate": 7.006666666666667e-06,
|
| 1445 |
+
"loss": 0.4296,
|
| 1446 |
+
"step": 1950
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 9.775561097256858,
|
| 1450 |
+
"grad_norm": 1.4622180461883545,
|
| 1451 |
+
"learning_rate": 6.9400000000000005e-06,
|
| 1452 |
+
"loss": 0.4422,
|
| 1453 |
+
"step": 1960
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 9.825436408977556,
|
| 1457 |
+
"grad_norm": 1.4130418300628662,
|
| 1458 |
+
"learning_rate": 6.873333333333334e-06,
|
| 1459 |
+
"loss": 0.4317,
|
| 1460 |
+
"step": 1970
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 9.875311720698255,
|
| 1464 |
+
"grad_norm": 1.311515212059021,
|
| 1465 |
+
"learning_rate": 6.806666666666667e-06,
|
| 1466 |
+
"loss": 0.4184,
|
| 1467 |
+
"step": 1980
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 9.925187032418952,
|
| 1471 |
+
"grad_norm": 1.4601870775222778,
|
| 1472 |
+
"learning_rate": 6.740000000000001e-06,
|
| 1473 |
+
"loss": 0.4136,
|
| 1474 |
+
"step": 1990
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 9.975062344139651,
|
| 1478 |
+
"grad_norm": 1.3348491191864014,
|
| 1479 |
+
"learning_rate": 6.6733333333333335e-06,
|
| 1480 |
+
"loss": 0.4409,
|
| 1481 |
+
"step": 2000
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 10.0,
|
| 1485 |
+
"eval_loss": 0.44308173656463623,
|
| 1486 |
+
"eval_runtime": 57.9177,
|
| 1487 |
+
"eval_samples_per_second": 17.318,
|
| 1488 |
+
"eval_steps_per_second": 3.47,
|
| 1489 |
+
"step": 2005
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 10.024937655860349,
|
| 1493 |
+
"grad_norm": 1.3334033489227295,
|
| 1494 |
+
"learning_rate": 6.606666666666666e-06,
|
| 1495 |
+
"loss": 0.4456,
|
| 1496 |
+
"step": 2010
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 10.074812967581048,
|
| 1500 |
+
"grad_norm": 1.5118451118469238,
|
| 1501 |
+
"learning_rate": 6.540000000000001e-06,
|
| 1502 |
+
"loss": 0.4385,
|
| 1503 |
+
"step": 2020
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 10.124688279301745,
|
| 1507 |
+
"grad_norm": 1.3644325733184814,
|
| 1508 |
+
"learning_rate": 6.473333333333334e-06,
|
| 1509 |
+
"loss": 0.4208,
|
| 1510 |
+
"step": 2030
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 10.174563591022444,
|
| 1514 |
+
"grad_norm": 1.426771640777588,
|
| 1515 |
+
"learning_rate": 6.4066666666666674e-06,
|
| 1516 |
+
"loss": 0.4201,
|
| 1517 |
+
"step": 2040
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 10.224438902743142,
|
| 1521 |
+
"grad_norm": 1.3023799657821655,
|
| 1522 |
+
"learning_rate": 6.34e-06,
|
| 1523 |
+
"loss": 0.4068,
|
| 1524 |
+
"step": 2050
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 10.27431421446384,
|
| 1528 |
+
"grad_norm": 1.3506182432174683,
|
| 1529 |
+
"learning_rate": 6.273333333333333e-06,
|
| 1530 |
+
"loss": 0.4264,
|
| 1531 |
+
"step": 2060
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 10.324189526184538,
|
| 1535 |
+
"grad_norm": 1.4448788166046143,
|
| 1536 |
+
"learning_rate": 6.206666666666668e-06,
|
| 1537 |
+
"loss": 0.4249,
|
| 1538 |
+
"step": 2070
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 10.374064837905237,
|
| 1542 |
+
"grad_norm": 1.3727625608444214,
|
| 1543 |
+
"learning_rate": 6.1400000000000005e-06,
|
| 1544 |
+
"loss": 0.4351,
|
| 1545 |
+
"step": 2080
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 10.423940149625935,
|
| 1549 |
+
"grad_norm": 1.512022614479065,
|
| 1550 |
+
"learning_rate": 6.073333333333333e-06,
|
| 1551 |
+
"loss": 0.4332,
|
| 1552 |
+
"step": 2090
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 10.473815461346634,
|
| 1556 |
+
"grad_norm": 1.305282473564148,
|
| 1557 |
+
"learning_rate": 6.006666666666667e-06,
|
| 1558 |
+
"loss": 0.4248,
|
| 1559 |
+
"step": 2100
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 10.523690773067331,
|
| 1563 |
+
"grad_norm": 1.4088374376296997,
|
| 1564 |
+
"learning_rate": 5.94e-06,
|
| 1565 |
+
"loss": 0.4323,
|
| 1566 |
+
"step": 2110
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 10.57356608478803,
|
| 1570 |
+
"grad_norm": 1.4796631336212158,
|
| 1571 |
+
"learning_rate": 5.873333333333334e-06,
|
| 1572 |
+
"loss": 0.4254,
|
| 1573 |
+
"step": 2120
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 10.623441396508728,
|
| 1577 |
+
"grad_norm": 1.556593418121338,
|
| 1578 |
+
"learning_rate": 5.806666666666667e-06,
|
| 1579 |
+
"loss": 0.4117,
|
| 1580 |
+
"step": 2130
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 10.673316708229427,
|
| 1584 |
+
"grad_norm": 1.3857340812683105,
|
| 1585 |
+
"learning_rate": 5.74e-06,
|
| 1586 |
+
"loss": 0.4241,
|
| 1587 |
+
"step": 2140
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 10.723192019950124,
|
| 1591 |
+
"grad_norm": 1.6449425220489502,
|
| 1592 |
+
"learning_rate": 5.673333333333334e-06,
|
| 1593 |
+
"loss": 0.4144,
|
| 1594 |
+
"step": 2150
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 10.773067331670823,
|
| 1598 |
+
"grad_norm": 1.8624916076660156,
|
| 1599 |
+
"learning_rate": 5.606666666666667e-06,
|
| 1600 |
+
"loss": 0.4239,
|
| 1601 |
+
"step": 2160
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 10.82294264339152,
|
| 1605 |
+
"grad_norm": 1.3718584775924683,
|
| 1606 |
+
"learning_rate": 5.540000000000001e-06,
|
| 1607 |
+
"loss": 0.4081,
|
| 1608 |
+
"step": 2170
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 10.87281795511222,
|
| 1612 |
+
"grad_norm": 1.4391957521438599,
|
| 1613 |
+
"learning_rate": 5.473333333333334e-06,
|
| 1614 |
+
"loss": 0.4195,
|
| 1615 |
+
"step": 2180
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 10.922693266832917,
|
| 1619 |
+
"grad_norm": 1.6265774965286255,
|
| 1620 |
+
"learning_rate": 5.406666666666667e-06,
|
| 1621 |
+
"loss": 0.4446,
|
| 1622 |
+
"step": 2190
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 10.972568578553616,
|
| 1626 |
+
"grad_norm": 1.5026485919952393,
|
| 1627 |
+
"learning_rate": 5.3400000000000005e-06,
|
| 1628 |
+
"loss": 0.4229,
|
| 1629 |
+
"step": 2200
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 10.997506234413965,
|
| 1633 |
+
"eval_loss": 0.44174298644065857,
|
| 1634 |
+
"eval_runtime": 57.9098,
|
| 1635 |
+
"eval_samples_per_second": 17.32,
|
| 1636 |
+
"eval_steps_per_second": 3.471,
|
| 1637 |
+
"step": 2205
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 11.022443890274314,
|
| 1641 |
+
"grad_norm": 1.452600121498108,
|
| 1642 |
+
"learning_rate": 5.273333333333333e-06,
|
| 1643 |
+
"loss": 0.4326,
|
| 1644 |
+
"step": 2210
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 11.072319201995013,
|
| 1648 |
+
"grad_norm": 1.3822201490402222,
|
| 1649 |
+
"learning_rate": 5.206666666666668e-06,
|
| 1650 |
+
"loss": 0.4264,
|
| 1651 |
+
"step": 2220
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 11.12219451371571,
|
| 1655 |
+
"grad_norm": 1.468363642692566,
|
| 1656 |
+
"learning_rate": 5.140000000000001e-06,
|
| 1657 |
+
"loss": 0.4078,
|
| 1658 |
+
"step": 2230
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 11.17206982543641,
|
| 1662 |
+
"grad_norm": 1.495955467224121,
|
| 1663 |
+
"learning_rate": 5.073333333333334e-06,
|
| 1664 |
+
"loss": 0.4173,
|
| 1665 |
+
"step": 2240
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 11.221945137157107,
|
| 1669 |
+
"grad_norm": 1.441292405128479,
|
| 1670 |
+
"learning_rate": 5.006666666666667e-06,
|
| 1671 |
+
"loss": 0.415,
|
| 1672 |
+
"step": 2250
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 11.271820448877806,
|
| 1676 |
+
"grad_norm": 1.421720266342163,
|
| 1677 |
+
"learning_rate": 4.94e-06,
|
| 1678 |
+
"loss": 0.4126,
|
| 1679 |
+
"step": 2260
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 11.321695760598503,
|
| 1683 |
+
"grad_norm": 1.2590454816818237,
|
| 1684 |
+
"learning_rate": 4.873333333333334e-06,
|
| 1685 |
+
"loss": 0.4193,
|
| 1686 |
+
"step": 2270
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 11.371571072319203,
|
| 1690 |
+
"grad_norm": 1.48396897315979,
|
| 1691 |
+
"learning_rate": 4.8066666666666675e-06,
|
| 1692 |
+
"loss": 0.4266,
|
| 1693 |
+
"step": 2280
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 11.4214463840399,
|
| 1697 |
+
"grad_norm": 1.3837693929672241,
|
| 1698 |
+
"learning_rate": 4.74e-06,
|
| 1699 |
+
"loss": 0.4199,
|
| 1700 |
+
"step": 2290
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 11.471321695760599,
|
| 1704 |
+
"grad_norm": 1.5965029001235962,
|
| 1705 |
+
"learning_rate": 4.673333333333333e-06,
|
| 1706 |
+
"loss": 0.4148,
|
| 1707 |
+
"step": 2300
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 11.521197007481296,
|
| 1711 |
+
"grad_norm": 1.4824038743972778,
|
| 1712 |
+
"learning_rate": 4.606666666666667e-06,
|
| 1713 |
+
"loss": 0.4369,
|
| 1714 |
+
"step": 2310
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 11.571072319201996,
|
| 1718 |
+
"grad_norm": 1.4421814680099487,
|
| 1719 |
+
"learning_rate": 4.540000000000001e-06,
|
| 1720 |
+
"loss": 0.4182,
|
| 1721 |
+
"step": 2320
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 11.620947630922693,
|
| 1725 |
+
"grad_norm": 1.4961748123168945,
|
| 1726 |
+
"learning_rate": 4.473333333333334e-06,
|
| 1727 |
+
"loss": 0.4117,
|
| 1728 |
+
"step": 2330
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 11.670822942643392,
|
| 1732 |
+
"grad_norm": 1.63477623462677,
|
| 1733 |
+
"learning_rate": 4.406666666666667e-06,
|
| 1734 |
+
"loss": 0.4237,
|
| 1735 |
+
"step": 2340
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 11.72069825436409,
|
| 1739 |
+
"grad_norm": 1.6593093872070312,
|
| 1740 |
+
"learning_rate": 4.34e-06,
|
| 1741 |
+
"loss": 0.4279,
|
| 1742 |
+
"step": 2350
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 11.770573566084789,
|
| 1746 |
+
"grad_norm": 1.5250682830810547,
|
| 1747 |
+
"learning_rate": 4.273333333333334e-06,
|
| 1748 |
+
"loss": 0.4264,
|
| 1749 |
+
"step": 2360
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 11.820448877805486,
|
| 1753 |
+
"grad_norm": 1.5515984296798706,
|
| 1754 |
+
"learning_rate": 4.206666666666667e-06,
|
| 1755 |
+
"loss": 0.4341,
|
| 1756 |
+
"step": 2370
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 11.870324189526185,
|
| 1760 |
+
"grad_norm": 1.6062873601913452,
|
| 1761 |
+
"learning_rate": 4.14e-06,
|
| 1762 |
+
"loss": 0.4247,
|
| 1763 |
+
"step": 2380
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 11.920199501246882,
|
| 1767 |
+
"grad_norm": 1.760644793510437,
|
| 1768 |
+
"learning_rate": 4.073333333333334e-06,
|
| 1769 |
+
"loss": 0.4417,
|
| 1770 |
+
"step": 2390
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 11.970074812967582,
|
| 1774 |
+
"grad_norm": 1.2911533117294312,
|
| 1775 |
+
"learning_rate": 4.006666666666667e-06,
|
| 1776 |
+
"loss": 0.4159,
|
| 1777 |
+
"step": 2400
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 12.0,
|
| 1781 |
+
"eval_loss": 0.44075635075569153,
|
| 1782 |
+
"eval_runtime": 57.9093,
|
| 1783 |
+
"eval_samples_per_second": 17.32,
|
| 1784 |
+
"eval_steps_per_second": 3.471,
|
| 1785 |
+
"step": 2406
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 12.019950124688279,
|
| 1789 |
+
"grad_norm": 1.5172271728515625,
|
| 1790 |
+
"learning_rate": 3.94e-06,
|
| 1791 |
+
"loss": 0.437,
|
| 1792 |
+
"step": 2410
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 12.069825436408978,
|
| 1796 |
+
"grad_norm": 1.4542344808578491,
|
| 1797 |
+
"learning_rate": 3.873333333333333e-06,
|
| 1798 |
+
"loss": 0.4314,
|
| 1799 |
+
"step": 2420
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 12.119700748129675,
|
| 1803 |
+
"grad_norm": 1.4818074703216553,
|
| 1804 |
+
"learning_rate": 3.806666666666667e-06,
|
| 1805 |
+
"loss": 0.4184,
|
| 1806 |
+
"step": 2430
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 12.169576059850375,
|
| 1810 |
+
"grad_norm": 1.4975398778915405,
|
| 1811 |
+
"learning_rate": 3.74e-06,
|
| 1812 |
+
"loss": 0.4087,
|
| 1813 |
+
"step": 2440
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 12.219451371571072,
|
| 1817 |
+
"grad_norm": 1.3974961042404175,
|
| 1818 |
+
"learning_rate": 3.673333333333334e-06,
|
| 1819 |
+
"loss": 0.4135,
|
| 1820 |
+
"step": 2450
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 12.269326683291771,
|
| 1824 |
+
"grad_norm": 1.6019673347473145,
|
| 1825 |
+
"learning_rate": 3.606666666666667e-06,
|
| 1826 |
+
"loss": 0.4238,
|
| 1827 |
+
"step": 2460
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 12.319201995012468,
|
| 1831 |
+
"grad_norm": 1.7037951946258545,
|
| 1832 |
+
"learning_rate": 3.54e-06,
|
| 1833 |
+
"loss": 0.4218,
|
| 1834 |
+
"step": 2470
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 12.369077306733168,
|
| 1838 |
+
"grad_norm": 1.5602372884750366,
|
| 1839 |
+
"learning_rate": 3.4733333333333337e-06,
|
| 1840 |
+
"loss": 0.4192,
|
| 1841 |
+
"step": 2480
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 12.418952618453865,
|
| 1845 |
+
"grad_norm": 1.391041874885559,
|
| 1846 |
+
"learning_rate": 3.406666666666667e-06,
|
| 1847 |
+
"loss": 0.414,
|
| 1848 |
+
"step": 2490
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 12.468827930174564,
|
| 1852 |
+
"grad_norm": 1.5659008026123047,
|
| 1853 |
+
"learning_rate": 3.3400000000000006e-06,
|
| 1854 |
+
"loss": 0.4251,
|
| 1855 |
+
"step": 2500
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 12.518703241895262,
|
| 1859 |
+
"grad_norm": 1.4149389266967773,
|
| 1860 |
+
"learning_rate": 3.2733333333333335e-06,
|
| 1861 |
+
"loss": 0.4299,
|
| 1862 |
+
"step": 2510
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 12.56857855361596,
|
| 1866 |
+
"grad_norm": 1.5269657373428345,
|
| 1867 |
+
"learning_rate": 3.2066666666666667e-06,
|
| 1868 |
+
"loss": 0.4355,
|
| 1869 |
+
"step": 2520
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 12.618453865336658,
|
| 1873 |
+
"grad_norm": 1.6707123517990112,
|
| 1874 |
+
"learning_rate": 3.1400000000000004e-06,
|
| 1875 |
+
"loss": 0.4282,
|
| 1876 |
+
"step": 2530
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 12.668329177057357,
|
| 1880 |
+
"grad_norm": 1.3647090196609497,
|
| 1881 |
+
"learning_rate": 3.0733333333333337e-06,
|
| 1882 |
+
"loss": 0.3994,
|
| 1883 |
+
"step": 2540
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 12.718204488778055,
|
| 1887 |
+
"grad_norm": 1.4243894815444946,
|
| 1888 |
+
"learning_rate": 3.0066666666666674e-06,
|
| 1889 |
+
"loss": 0.4181,
|
| 1890 |
+
"step": 2550
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 12.768079800498754,
|
| 1894 |
+
"grad_norm": 1.3579431772232056,
|
| 1895 |
+
"learning_rate": 2.9400000000000002e-06,
|
| 1896 |
+
"loss": 0.4251,
|
| 1897 |
+
"step": 2560
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 12.817955112219451,
|
| 1901 |
+
"grad_norm": 1.6759490966796875,
|
| 1902 |
+
"learning_rate": 2.8733333333333335e-06,
|
| 1903 |
+
"loss": 0.4143,
|
| 1904 |
+
"step": 2570
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 12.86783042394015,
|
| 1908 |
+
"grad_norm": 1.5581449270248413,
|
| 1909 |
+
"learning_rate": 2.806666666666667e-06,
|
| 1910 |
+
"loss": 0.4083,
|
| 1911 |
+
"step": 2580
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 12.917705735660848,
|
| 1915 |
+
"grad_norm": 1.607188105583191,
|
| 1916 |
+
"learning_rate": 2.7400000000000004e-06,
|
| 1917 |
+
"loss": 0.4236,
|
| 1918 |
+
"step": 2590
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 12.967581047381547,
|
| 1922 |
+
"grad_norm": 1.3592510223388672,
|
| 1923 |
+
"learning_rate": 2.6733333333333333e-06,
|
| 1924 |
+
"loss": 0.4108,
|
| 1925 |
+
"step": 2600
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 12.997506234413965,
|
| 1929 |
+
"eval_loss": 0.44020339846611023,
|
| 1930 |
+
"eval_runtime": 57.8997,
|
| 1931 |
+
"eval_samples_per_second": 17.323,
|
| 1932 |
+
"eval_steps_per_second": 3.472,
|
| 1933 |
+
"step": 2606
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 13.017456359102244,
|
| 1937 |
+
"grad_norm": 1.461461067199707,
|
| 1938 |
+
"learning_rate": 2.606666666666667e-06,
|
| 1939 |
+
"loss": 0.4439,
|
| 1940 |
+
"step": 2610
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 13.067331670822943,
|
| 1944 |
+
"grad_norm": 1.40589439868927,
|
| 1945 |
+
"learning_rate": 2.5400000000000002e-06,
|
| 1946 |
+
"loss": 0.4057,
|
| 1947 |
+
"step": 2620
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 13.11720698254364,
|
| 1951 |
+
"grad_norm": 1.665216088294983,
|
| 1952 |
+
"learning_rate": 2.4733333333333335e-06,
|
| 1953 |
+
"loss": 0.4053,
|
| 1954 |
+
"step": 2630
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 13.16708229426434,
|
| 1958 |
+
"grad_norm": 1.4636025428771973,
|
| 1959 |
+
"learning_rate": 2.4066666666666668e-06,
|
| 1960 |
+
"loss": 0.4199,
|
| 1961 |
+
"step": 2640
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 13.216957605985037,
|
| 1965 |
+
"grad_norm": 1.4638257026672363,
|
| 1966 |
+
"learning_rate": 2.3400000000000005e-06,
|
| 1967 |
+
"loss": 0.4062,
|
| 1968 |
+
"step": 2650
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 13.266832917705736,
|
| 1972 |
+
"grad_norm": 1.4565849304199219,
|
| 1973 |
+
"learning_rate": 2.2733333333333333e-06,
|
| 1974 |
+
"loss": 0.4354,
|
| 1975 |
+
"step": 2660
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 13.316708229426434,
|
| 1979 |
+
"grad_norm": 1.4819647073745728,
|
| 1980 |
+
"learning_rate": 2.206666666666667e-06,
|
| 1981 |
+
"loss": 0.4088,
|
| 1982 |
+
"step": 2670
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 13.366583541147133,
|
| 1986 |
+
"grad_norm": 1.5023897886276245,
|
| 1987 |
+
"learning_rate": 2.1400000000000003e-06,
|
| 1988 |
+
"loss": 0.4313,
|
| 1989 |
+
"step": 2680
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 13.41645885286783,
|
| 1993 |
+
"grad_norm": 1.4261906147003174,
|
| 1994 |
+
"learning_rate": 2.0733333333333335e-06,
|
| 1995 |
+
"loss": 0.4179,
|
| 1996 |
+
"step": 2690
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 13.46633416458853,
|
| 2000 |
+
"grad_norm": 1.6513773202896118,
|
| 2001 |
+
"learning_rate": 2.006666666666667e-06,
|
| 2002 |
+
"loss": 0.4341,
|
| 2003 |
+
"step": 2700
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 13.516209476309227,
|
| 2007 |
+
"grad_norm": 1.6153149604797363,
|
| 2008 |
+
"learning_rate": 1.94e-06,
|
| 2009 |
+
"loss": 0.4231,
|
| 2010 |
+
"step": 2710
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 13.566084788029926,
|
| 2014 |
+
"grad_norm": 1.6202324628829956,
|
| 2015 |
+
"learning_rate": 1.8733333333333333e-06,
|
| 2016 |
+
"loss": 0.4173,
|
| 2017 |
+
"step": 2720
|
| 2018 |
+
},
|
| 2019 |
+
{
|
| 2020 |
+
"epoch": 13.615960099750623,
|
| 2021 |
+
"grad_norm": 1.2559431791305542,
|
| 2022 |
+
"learning_rate": 1.8066666666666668e-06,
|
| 2023 |
+
"loss": 0.4238,
|
| 2024 |
+
"step": 2730
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 13.665835411471322,
|
| 2028 |
+
"grad_norm": 1.3010227680206299,
|
| 2029 |
+
"learning_rate": 1.74e-06,
|
| 2030 |
+
"loss": 0.4287,
|
| 2031 |
+
"step": 2740
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 13.71571072319202,
|
| 2035 |
+
"grad_norm": 1.432409644126892,
|
| 2036 |
+
"learning_rate": 1.6733333333333335e-06,
|
| 2037 |
+
"loss": 0.4143,
|
| 2038 |
+
"step": 2750
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 13.765586034912719,
|
| 2042 |
+
"grad_norm": 1.4686477184295654,
|
| 2043 |
+
"learning_rate": 1.606666666666667e-06,
|
| 2044 |
+
"loss": 0.4192,
|
| 2045 |
+
"step": 2760
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 13.815461346633416,
|
| 2049 |
+
"grad_norm": 1.526962161064148,
|
| 2050 |
+
"learning_rate": 1.54e-06,
|
| 2051 |
+
"loss": 0.4199,
|
| 2052 |
+
"step": 2770
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 13.865336658354115,
|
| 2056 |
+
"grad_norm": 1.5790767669677734,
|
| 2057 |
+
"learning_rate": 1.4733333333333336e-06,
|
| 2058 |
+
"loss": 0.4089,
|
| 2059 |
+
"step": 2780
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 13.915211970074813,
|
| 2063 |
+
"grad_norm": 1.4289913177490234,
|
| 2064 |
+
"learning_rate": 1.4066666666666668e-06,
|
| 2065 |
+
"loss": 0.4195,
|
| 2066 |
+
"step": 2790
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 13.965087281795512,
|
| 2070 |
+
"grad_norm": 1.613978624343872,
|
| 2071 |
+
"learning_rate": 1.34e-06,
|
| 2072 |
+
"loss": 0.4279,
|
| 2073 |
+
"step": 2800
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 14.0,
|
| 2077 |
+
"eval_loss": 0.439765065908432,
|
| 2078 |
+
"eval_runtime": 57.9105,
|
| 2079 |
+
"eval_samples_per_second": 17.32,
|
| 2080 |
+
"eval_steps_per_second": 3.471,
|
| 2081 |
+
"step": 2807
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 14.01496259351621,
|
| 2085 |
+
"grad_norm": 1.4722168445587158,
|
| 2086 |
+
"learning_rate": 1.2733333333333334e-06,
|
| 2087 |
+
"loss": 0.445,
|
| 2088 |
+
"step": 2810
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 14.064837905236908,
|
| 2092 |
+
"grad_norm": 1.6725043058395386,
|
| 2093 |
+
"learning_rate": 1.2066666666666668e-06,
|
| 2094 |
+
"loss": 0.4313,
|
| 2095 |
+
"step": 2820
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 14.114713216957606,
|
| 2099 |
+
"grad_norm": 1.4227081537246704,
|
| 2100 |
+
"learning_rate": 1.14e-06,
|
| 2101 |
+
"loss": 0.4181,
|
| 2102 |
+
"step": 2830
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 14.164588528678305,
|
| 2106 |
+
"grad_norm": 1.5610967874526978,
|
| 2107 |
+
"learning_rate": 1.0733333333333334e-06,
|
| 2108 |
+
"loss": 0.4229,
|
| 2109 |
+
"step": 2840
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 14.214463840399002,
|
| 2113 |
+
"grad_norm": 1.520051121711731,
|
| 2114 |
+
"learning_rate": 1.0066666666666668e-06,
|
| 2115 |
+
"loss": 0.4301,
|
| 2116 |
+
"step": 2850
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 14.264339152119701,
|
| 2120 |
+
"grad_norm": 1.4826945066452026,
|
| 2121 |
+
"learning_rate": 9.400000000000001e-07,
|
| 2122 |
+
"loss": 0.443,
|
| 2123 |
+
"step": 2860
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 14.314214463840399,
|
| 2127 |
+
"grad_norm": 1.6496617794036865,
|
| 2128 |
+
"learning_rate": 8.733333333333334e-07,
|
| 2129 |
+
"loss": 0.4109,
|
| 2130 |
+
"step": 2870
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 14.364089775561098,
|
| 2134 |
+
"grad_norm": 1.472607135772705,
|
| 2135 |
+
"learning_rate": 8.066666666666667e-07,
|
| 2136 |
+
"loss": 0.4075,
|
| 2137 |
+
"step": 2880
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 14.413965087281795,
|
| 2141 |
+
"grad_norm": 1.4754953384399414,
|
| 2142 |
+
"learning_rate": 7.4e-07,
|
| 2143 |
+
"loss": 0.4257,
|
| 2144 |
+
"step": 2890
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 14.463840399002494,
|
| 2148 |
+
"grad_norm": 1.6782726049423218,
|
| 2149 |
+
"learning_rate": 6.733333333333334e-07,
|
| 2150 |
+
"loss": 0.4099,
|
| 2151 |
+
"step": 2900
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 14.513715710723192,
|
| 2155 |
+
"grad_norm": 1.4635963439941406,
|
| 2156 |
+
"learning_rate": 6.066666666666668e-07,
|
| 2157 |
+
"loss": 0.3977,
|
| 2158 |
+
"step": 2910
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 14.563591022443891,
|
| 2162 |
+
"grad_norm": 1.5924144983291626,
|
| 2163 |
+
"learning_rate": 5.4e-07,
|
| 2164 |
+
"loss": 0.4181,
|
| 2165 |
+
"step": 2920
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 14.613466334164588,
|
| 2169 |
+
"grad_norm": 1.346475601196289,
|
| 2170 |
+
"learning_rate": 4.7333333333333334e-07,
|
| 2171 |
+
"loss": 0.4132,
|
| 2172 |
+
"step": 2930
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 14.663341645885287,
|
| 2176 |
+
"grad_norm": 1.4001176357269287,
|
| 2177 |
+
"learning_rate": 4.0666666666666666e-07,
|
| 2178 |
+
"loss": 0.4007,
|
| 2179 |
+
"step": 2940
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 14.713216957605985,
|
| 2183 |
+
"grad_norm": 1.6243798732757568,
|
| 2184 |
+
"learning_rate": 3.4000000000000003e-07,
|
| 2185 |
+
"loss": 0.4332,
|
| 2186 |
+
"step": 2950
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 14.763092269326684,
|
| 2190 |
+
"grad_norm": 1.5413304567337036,
|
| 2191 |
+
"learning_rate": 2.7333333333333335e-07,
|
| 2192 |
+
"loss": 0.4083,
|
| 2193 |
+
"step": 2960
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 14.812967581047381,
|
| 2197 |
+
"grad_norm": 1.4332594871520996,
|
| 2198 |
+
"learning_rate": 2.066666666666667e-07,
|
| 2199 |
+
"loss": 0.4206,
|
| 2200 |
+
"step": 2970
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 14.86284289276808,
|
| 2204 |
+
"grad_norm": 1.4844541549682617,
|
| 2205 |
+
"learning_rate": 1.4e-07,
|
| 2206 |
+
"loss": 0.4285,
|
| 2207 |
+
"step": 2980
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 14.912718204488778,
|
| 2211 |
+
"grad_norm": 1.471591830253601,
|
| 2212 |
+
"learning_rate": 7.333333333333334e-08,
|
| 2213 |
+
"loss": 0.4118,
|
| 2214 |
+
"step": 2990
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 14.962593516209477,
|
| 2218 |
+
"grad_norm": 1.3903006315231323,
|
| 2219 |
+
"learning_rate": 6.666666666666667e-09,
|
| 2220 |
+
"loss": 0.4145,
|
| 2221 |
+
"step": 3000
|
| 2222 |
+
}
|
| 2223 |
+
],
|
| 2224 |
+
"logging_steps": 10,
|
| 2225 |
+
"max_steps": 3000,
|
| 2226 |
+
"num_input_tokens_seen": 0,
|
| 2227 |
+
"num_train_epochs": 15,
|
| 2228 |
+
"save_steps": 500,
|
| 2229 |
+
"stateful_callbacks": {
|
| 2230 |
+
"TrainerControl": {
|
| 2231 |
+
"args": {
|
| 2232 |
+
"should_epoch_stop": false,
|
| 2233 |
+
"should_evaluate": false,
|
| 2234 |
+
"should_log": false,
|
| 2235 |
+
"should_save": true,
|
| 2236 |
+
"should_training_stop": true
|
| 2237 |
+
},
|
| 2238 |
+
"attributes": {}
|
| 2239 |
+
}
|
| 2240 |
+
},
|
| 2241 |
+
"total_flos": 4.48818315264e+16,
|
| 2242 |
+
"train_batch_size": 5,
|
| 2243 |
+
"trial_name": null,
|
| 2244 |
+
"trial_params": null
|
| 2245 |
+
}
|