Upload checkpoint-270/trainer_state.json with huggingface_hub
Browse files- checkpoint-270/trainer_state.json +1923 -0
checkpoint-270/trainer_state.json
ADDED
|
@@ -0,0 +1,1923 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.652075471698113,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 270,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.036226415094339624,
|
| 13 |
+
"grad_norm": 4.074844837188721,
|
| 14 |
+
"learning_rate": 3.7037037037037036e-07,
|
| 15 |
+
"loss": 0.9495,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.07245283018867925,
|
| 20 |
+
"grad_norm": 4.034702301025391,
|
| 21 |
+
"learning_rate": 7.407407407407407e-07,
|
| 22 |
+
"loss": 0.9536,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.10867924528301887,
|
| 27 |
+
"grad_norm": 4.080567359924316,
|
| 28 |
+
"learning_rate": 1.111111111111111e-06,
|
| 29 |
+
"loss": 0.9267,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.1449056603773585,
|
| 34 |
+
"grad_norm": 3.9866671562194824,
|
| 35 |
+
"learning_rate": 1.4814814814814815e-06,
|
| 36 |
+
"loss": 0.95,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1811320754716981,
|
| 41 |
+
"grad_norm": 4.042410373687744,
|
| 42 |
+
"learning_rate": 1.8518518518518519e-06,
|
| 43 |
+
"loss": 0.995,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.21735849056603773,
|
| 48 |
+
"grad_norm": 4.1409783363342285,
|
| 49 |
+
"learning_rate": 2.222222222222222e-06,
|
| 50 |
+
"loss": 0.9579,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.25358490566037734,
|
| 55 |
+
"grad_norm": 3.998582601547241,
|
| 56 |
+
"learning_rate": 2.5925925925925925e-06,
|
| 57 |
+
"loss": 0.9625,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.289811320754717,
|
| 62 |
+
"grad_norm": 4.141905307769775,
|
| 63 |
+
"learning_rate": 2.962962962962963e-06,
|
| 64 |
+
"loss": 0.9506,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.3260377358490566,
|
| 69 |
+
"grad_norm": 4.01569128036499,
|
| 70 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 71 |
+
"loss": 0.9591,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3622641509433962,
|
| 76 |
+
"grad_norm": 3.662905216217041,
|
| 77 |
+
"learning_rate": 3.7037037037037037e-06,
|
| 78 |
+
"loss": 0.9519,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.39849056603773586,
|
| 83 |
+
"grad_norm": 2.892972707748413,
|
| 84 |
+
"learning_rate": 4.074074074074074e-06,
|
| 85 |
+
"loss": 0.9484,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.43471698113207546,
|
| 90 |
+
"grad_norm": 2.5980327129364014,
|
| 91 |
+
"learning_rate": 4.444444444444444e-06,
|
| 92 |
+
"loss": 0.9322,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.47094339622641507,
|
| 97 |
+
"grad_norm": 2.2705729007720947,
|
| 98 |
+
"learning_rate": 4.814814814814815e-06,
|
| 99 |
+
"loss": 0.9016,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.5071698113207547,
|
| 104 |
+
"grad_norm": 2.027101755142212,
|
| 105 |
+
"learning_rate": 5.185185185185185e-06,
|
| 106 |
+
"loss": 0.9084,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.5433962264150943,
|
| 111 |
+
"grad_norm": 1.2326773405075073,
|
| 112 |
+
"learning_rate": 5.555555555555557e-06,
|
| 113 |
+
"loss": 0.9035,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.579622641509434,
|
| 118 |
+
"grad_norm": 1.5498149394989014,
|
| 119 |
+
"learning_rate": 5.925925925925926e-06,
|
| 120 |
+
"loss": 0.8692,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.6158490566037735,
|
| 125 |
+
"grad_norm": 1.6801735162734985,
|
| 126 |
+
"learning_rate": 6.296296296296297e-06,
|
| 127 |
+
"loss": 0.8736,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.6520754716981132,
|
| 132 |
+
"grad_norm": 1.998213529586792,
|
| 133 |
+
"learning_rate": 6.666666666666667e-06,
|
| 134 |
+
"loss": 0.8728,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6883018867924529,
|
| 139 |
+
"grad_norm": 1.8951051235198975,
|
| 140 |
+
"learning_rate": 7.0370370370370375e-06,
|
| 141 |
+
"loss": 0.8715,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.7245283018867924,
|
| 146 |
+
"grad_norm": 1.7730218172073364,
|
| 147 |
+
"learning_rate": 7.4074074074074075e-06,
|
| 148 |
+
"loss": 0.8358,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.7607547169811321,
|
| 153 |
+
"grad_norm": 1.758504867553711,
|
| 154 |
+
"learning_rate": 7.77777777777778e-06,
|
| 155 |
+
"loss": 0.8289,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7969811320754717,
|
| 160 |
+
"grad_norm": 1.978657841682434,
|
| 161 |
+
"learning_rate": 8.148148148148148e-06,
|
| 162 |
+
"loss": 0.8432,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.8332075471698113,
|
| 167 |
+
"grad_norm": 1.5869137048721313,
|
| 168 |
+
"learning_rate": 8.518518518518519e-06,
|
| 169 |
+
"loss": 0.8255,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.8694339622641509,
|
| 174 |
+
"grad_norm": 1.3577728271484375,
|
| 175 |
+
"learning_rate": 8.888888888888888e-06,
|
| 176 |
+
"loss": 0.8272,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.9056603773584906,
|
| 181 |
+
"grad_norm": 0.984247088432312,
|
| 182 |
+
"learning_rate": 9.25925925925926e-06,
|
| 183 |
+
"loss": 0.8071,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.9418867924528301,
|
| 188 |
+
"grad_norm": 0.8904944062232971,
|
| 189 |
+
"learning_rate": 9.62962962962963e-06,
|
| 190 |
+
"loss": 0.8054,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.9781132075471698,
|
| 195 |
+
"grad_norm": 0.8401544094085693,
|
| 196 |
+
"learning_rate": 1e-05,
|
| 197 |
+
"loss": 0.8313,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.0,
|
| 202 |
+
"grad_norm": 0.8401544094085693,
|
| 203 |
+
"learning_rate": 9.999582149277188e-06,
|
| 204 |
+
"loss": 0.8047,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.0362264150943397,
|
| 209 |
+
"grad_norm": 1.6306538581848145,
|
| 210 |
+
"learning_rate": 9.998328666948437e-06,
|
| 211 |
+
"loss": 0.8103,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.0724528301886793,
|
| 216 |
+
"grad_norm": 1.0208637714385986,
|
| 217 |
+
"learning_rate": 9.996239762521152e-06,
|
| 218 |
+
"loss": 0.7616,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.1086792452830188,
|
| 223 |
+
"grad_norm": 1.045002818107605,
|
| 224 |
+
"learning_rate": 9.993315785135417e-06,
|
| 225 |
+
"loss": 0.7736,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.1449056603773584,
|
| 230 |
+
"grad_norm": 0.8551422357559204,
|
| 231 |
+
"learning_rate": 9.989557223505661e-06,
|
| 232 |
+
"loss": 0.762,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.181132075471698,
|
| 237 |
+
"grad_norm": 0.7436487078666687,
|
| 238 |
+
"learning_rate": 9.98496470583896e-06,
|
| 239 |
+
"loss": 0.7793,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.2173584905660377,
|
| 244 |
+
"grad_norm": 0.6822688579559326,
|
| 245 |
+
"learning_rate": 9.979538999730047e-06,
|
| 246 |
+
"loss": 0.7728,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.2535849056603774,
|
| 251 |
+
"grad_norm": 0.6098222732543945,
|
| 252 |
+
"learning_rate": 9.973281012033009e-06,
|
| 253 |
+
"loss": 0.7362,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.289811320754717,
|
| 258 |
+
"grad_norm": 0.6669561266899109,
|
| 259 |
+
"learning_rate": 9.966191788709716e-06,
|
| 260 |
+
"loss": 0.7052,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.3260377358490567,
|
| 265 |
+
"grad_norm": 0.6505219340324402,
|
| 266 |
+
"learning_rate": 9.958272514655006e-06,
|
| 267 |
+
"loss": 0.7609,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.3622641509433961,
|
| 272 |
+
"grad_norm": 0.737601637840271,
|
| 273 |
+
"learning_rate": 9.949524513498636e-06,
|
| 274 |
+
"loss": 0.7477,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.3984905660377358,
|
| 279 |
+
"grad_norm": 0.5156015753746033,
|
| 280 |
+
"learning_rate": 9.939949247384046e-06,
|
| 281 |
+
"loss": 0.7554,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.4347169811320755,
|
| 286 |
+
"grad_norm": 0.5583709478378296,
|
| 287 |
+
"learning_rate": 9.929548316723983e-06,
|
| 288 |
+
"loss": 0.7526,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.4709433962264151,
|
| 293 |
+
"grad_norm": 0.5988060832023621,
|
| 294 |
+
"learning_rate": 9.918323459933006e-06,
|
| 295 |
+
"loss": 0.7516,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.5071698113207548,
|
| 300 |
+
"grad_norm": 0.5599033832550049,
|
| 301 |
+
"learning_rate": 9.906276553136924e-06,
|
| 302 |
+
"loss": 0.7294,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.5433962264150942,
|
| 307 |
+
"grad_norm": 0.6027616858482361,
|
| 308 |
+
"learning_rate": 9.893409609859221e-06,
|
| 309 |
+
"loss": 0.72,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.579622641509434,
|
| 314 |
+
"grad_norm": 0.6933006048202515,
|
| 315 |
+
"learning_rate": 9.879724780684518e-06,
|
| 316 |
+
"loss": 0.7176,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.6158490566037735,
|
| 321 |
+
"grad_norm": 0.5711022615432739,
|
| 322 |
+
"learning_rate": 9.86522435289912e-06,
|
| 323 |
+
"loss": 0.716,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.6520754716981132,
|
| 328 |
+
"grad_norm": 0.48980799317359924,
|
| 329 |
+
"learning_rate": 9.849910750108718e-06,
|
| 330 |
+
"loss": 0.7333,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.6883018867924529,
|
| 335 |
+
"grad_norm": 0.4940769076347351,
|
| 336 |
+
"learning_rate": 9.833786531833311e-06,
|
| 337 |
+
"loss": 0.7368,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.7245283018867923,
|
| 342 |
+
"grad_norm": 0.4914894998073578,
|
| 343 |
+
"learning_rate": 9.816854393079402e-06,
|
| 344 |
+
"loss": 0.7323,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.7607547169811322,
|
| 349 |
+
"grad_norm": 0.5431811213493347,
|
| 350 |
+
"learning_rate": 9.79911716388956e-06,
|
| 351 |
+
"loss": 0.7226,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.7969811320754716,
|
| 356 |
+
"grad_norm": 0.4794186055660248,
|
| 357 |
+
"learning_rate": 9.7805778088694e-06,
|
| 358 |
+
"loss": 0.7277,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.8332075471698113,
|
| 363 |
+
"grad_norm": 0.5294991731643677,
|
| 364 |
+
"learning_rate": 9.761239426692077e-06,
|
| 365 |
+
"loss": 0.7193,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.869433962264151,
|
| 370 |
+
"grad_norm": 0.48333099484443665,
|
| 371 |
+
"learning_rate": 9.741105249580383e-06,
|
| 372 |
+
"loss": 0.6936,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.9056603773584906,
|
| 377 |
+
"grad_norm": 0.42380550503730774,
|
| 378 |
+
"learning_rate": 9.7201786427665e-06,
|
| 379 |
+
"loss": 0.7183,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.9418867924528302,
|
| 384 |
+
"grad_norm": 0.49902644753456116,
|
| 385 |
+
"learning_rate": 9.698463103929542e-06,
|
| 386 |
+
"loss": 0.7288,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.9781132075471697,
|
| 391 |
+
"grad_norm": 0.44469350576400757,
|
| 392 |
+
"learning_rate": 9.67596226261095e-06,
|
| 393 |
+
"loss": 0.7469,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 2.0,
|
| 398 |
+
"grad_norm": 0.6417510509490967,
|
| 399 |
+
"learning_rate": 9.652679879607843e-06,
|
| 400 |
+
"loss": 0.6903,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 2.0362264150943394,
|
| 405 |
+
"grad_norm": 0.5208938121795654,
|
| 406 |
+
"learning_rate": 9.628619846344453e-06,
|
| 407 |
+
"loss": 0.6569,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 2.0724528301886793,
|
| 412 |
+
"grad_norm": 0.41403067111968994,
|
| 413 |
+
"learning_rate": 9.603786184221693e-06,
|
| 414 |
+
"loss": 0.6837,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 2.1086792452830188,
|
| 419 |
+
"grad_norm": 0.4388677775859833,
|
| 420 |
+
"learning_rate": 9.578183043945031e-06,
|
| 421 |
+
"loss": 0.6577,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.1449056603773586,
|
| 426 |
+
"grad_norm": 0.38680893182754517,
|
| 427 |
+
"learning_rate": 9.551814704830734e-06,
|
| 428 |
+
"loss": 0.6647,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.181132075471698,
|
| 433 |
+
"grad_norm": 0.41955962777137756,
|
| 434 |
+
"learning_rate": 9.524685574090627e-06,
|
| 435 |
+
"loss": 0.6414,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.2173584905660375,
|
| 440 |
+
"grad_norm": 0.5132808089256287,
|
| 441 |
+
"learning_rate": 9.496800186095466e-06,
|
| 442 |
+
"loss": 0.6397,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.2535849056603774,
|
| 447 |
+
"grad_norm": 0.41187506914138794,
|
| 448 |
+
"learning_rate": 9.468163201617063e-06,
|
| 449 |
+
"loss": 0.6654,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.289811320754717,
|
| 454 |
+
"grad_norm": 0.40094566345214844,
|
| 455 |
+
"learning_rate": 9.438779407049282e-06,
|
| 456 |
+
"loss": 0.6483,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.3260377358490567,
|
| 461 |
+
"grad_norm": 0.42385968565940857,
|
| 462 |
+
"learning_rate": 9.40865371360804e-06,
|
| 463 |
+
"loss": 0.6382,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.362264150943396,
|
| 468 |
+
"grad_norm": 0.42505592107772827,
|
| 469 |
+
"learning_rate": 9.377791156510456e-06,
|
| 470 |
+
"loss": 0.621,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.398490566037736,
|
| 475 |
+
"grad_norm": 0.4004240036010742,
|
| 476 |
+
"learning_rate": 9.346196894133239e-06,
|
| 477 |
+
"loss": 0.6137,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.4347169811320755,
|
| 482 |
+
"grad_norm": 0.44593772292137146,
|
| 483 |
+
"learning_rate": 9.313876207150544e-06,
|
| 484 |
+
"loss": 0.6299,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.470943396226415,
|
| 489 |
+
"grad_norm": 0.40830758213996887,
|
| 490 |
+
"learning_rate": 9.280834497651334e-06,
|
| 491 |
+
"loss": 0.6504,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.507169811320755,
|
| 496 |
+
"grad_norm": 0.4714842438697815,
|
| 497 |
+
"learning_rate": 9.247077288236488e-06,
|
| 498 |
+
"loss": 0.626,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.543396226415094,
|
| 503 |
+
"grad_norm": 0.4523302912712097,
|
| 504 |
+
"learning_rate": 9.212610221095748e-06,
|
| 505 |
+
"loss": 0.6386,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.579622641509434,
|
| 510 |
+
"grad_norm": 0.4274156093597412,
|
| 511 |
+
"learning_rate": 9.177439057064684e-06,
|
| 512 |
+
"loss": 0.6368,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.6158490566037735,
|
| 517 |
+
"grad_norm": 0.3964768350124359,
|
| 518 |
+
"learning_rate": 9.141569674661816e-06,
|
| 519 |
+
"loss": 0.6161,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.6520754716981134,
|
| 524 |
+
"grad_norm": 0.4652266502380371,
|
| 525 |
+
"learning_rate": 9.105008069106093e-06,
|
| 526 |
+
"loss": 0.6208,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.688301886792453,
|
| 531 |
+
"grad_norm": 0.4457603394985199,
|
| 532 |
+
"learning_rate": 9.067760351314838e-06,
|
| 533 |
+
"loss": 0.6407,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.7245283018867923,
|
| 538 |
+
"grad_norm": 0.47072410583496094,
|
| 539 |
+
"learning_rate": 9.029832746882372e-06,
|
| 540 |
+
"loss": 0.6345,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.760754716981132,
|
| 545 |
+
"grad_norm": 0.41980573534965515,
|
| 546 |
+
"learning_rate": 8.991231595039464e-06,
|
| 547 |
+
"loss": 0.6242,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.7969811320754716,
|
| 552 |
+
"grad_norm": 0.8126304745674133,
|
| 553 |
+
"learning_rate": 8.951963347593797e-06,
|
| 554 |
+
"loss": 0.6368,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.8332075471698115,
|
| 559 |
+
"grad_norm": 0.4141775667667389,
|
| 560 |
+
"learning_rate": 8.9120345678516e-06,
|
| 561 |
+
"loss": 0.6299,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.869433962264151,
|
| 566 |
+
"grad_norm": 0.4654167592525482,
|
| 567 |
+
"learning_rate": 8.871451929520662e-06,
|
| 568 |
+
"loss": 0.6064,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.9056603773584904,
|
| 573 |
+
"grad_norm": 0.4674675464630127,
|
| 574 |
+
"learning_rate": 8.83022221559489e-06,
|
| 575 |
+
"loss": 0.6252,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.9418867924528302,
|
| 580 |
+
"grad_norm": 0.46518832445144653,
|
| 581 |
+
"learning_rate": 8.78835231722059e-06,
|
| 582 |
+
"loss": 0.6304,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.9781132075471697,
|
| 587 |
+
"grad_norm": 0.40669646859169006,
|
| 588 |
+
"learning_rate": 8.74584923254468e-06,
|
| 589 |
+
"loss": 0.6323,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 3.0,
|
| 594 |
+
"grad_norm": 0.40669646859169006,
|
| 595 |
+
"learning_rate": 8.702720065545024e-06,
|
| 596 |
+
"loss": 0.6371,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 3.0362264150943394,
|
| 601 |
+
"grad_norm": 0.5793619155883789,
|
| 602 |
+
"learning_rate": 8.658972024843063e-06,
|
| 603 |
+
"loss": 0.5837,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 3.0724528301886793,
|
| 608 |
+
"grad_norm": 0.42956608533859253,
|
| 609 |
+
"learning_rate": 8.614612422498965e-06,
|
| 610 |
+
"loss": 0.5657,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 3.1086792452830188,
|
| 615 |
+
"grad_norm": 0.3798193037509918,
|
| 616 |
+
"learning_rate": 8.569648672789496e-06,
|
| 617 |
+
"loss": 0.5788,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 3.1449056603773586,
|
| 622 |
+
"grad_norm": 0.43145015835762024,
|
| 623 |
+
"learning_rate": 8.524088290968781e-06,
|
| 624 |
+
"loss": 0.57,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 3.181132075471698,
|
| 629 |
+
"grad_norm": 0.40505528450012207,
|
| 630 |
+
"learning_rate": 8.477938892012209e-06,
|
| 631 |
+
"loss": 0.5556,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 3.2173584905660375,
|
| 636 |
+
"grad_norm": 0.36875954270362854,
|
| 637 |
+
"learning_rate": 8.43120818934367e-06,
|
| 638 |
+
"loss": 0.5566,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.2535849056603774,
|
| 643 |
+
"grad_norm": 0.45734313130378723,
|
| 644 |
+
"learning_rate": 8.38390399354631e-06,
|
| 645 |
+
"loss": 0.5388,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 3.289811320754717,
|
| 650 |
+
"grad_norm": 0.40550005435943604,
|
| 651 |
+
"learning_rate": 8.336034211057098e-06,
|
| 652 |
+
"loss": 0.5569,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.3260377358490567,
|
| 657 |
+
"grad_norm": 0.3886796832084656,
|
| 658 |
+
"learning_rate": 8.28760684284532e-06,
|
| 659 |
+
"loss": 0.546,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.362264150943396,
|
| 664 |
+
"grad_norm": 0.39407268166542053,
|
| 665 |
+
"learning_rate": 8.238629983075296e-06,
|
| 666 |
+
"loss": 0.5513,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.398490566037736,
|
| 671 |
+
"grad_norm": 0.44064226746559143,
|
| 672 |
+
"learning_rate": 8.18911181775353e-06,
|
| 673 |
+
"loss": 0.5524,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.4347169811320755,
|
| 678 |
+
"grad_norm": 0.45655596256256104,
|
| 679 |
+
"learning_rate": 8.139060623360494e-06,
|
| 680 |
+
"loss": 0.5402,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.470943396226415,
|
| 685 |
+
"grad_norm": 0.44399574398994446,
|
| 686 |
+
"learning_rate": 8.088484765467286e-06,
|
| 687 |
+
"loss": 0.5403,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.507169811320755,
|
| 692 |
+
"grad_norm": 0.46749594807624817,
|
| 693 |
+
"learning_rate": 8.037392697337418e-06,
|
| 694 |
+
"loss": 0.5609,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.543396226415094,
|
| 699 |
+
"grad_norm": 0.39268767833709717,
|
| 700 |
+
"learning_rate": 7.985792958513932e-06,
|
| 701 |
+
"loss": 0.5616,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.579622641509434,
|
| 706 |
+
"grad_norm": 0.4343853294849396,
|
| 707 |
+
"learning_rate": 7.93369417339209e-06,
|
| 708 |
+
"loss": 0.5493,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.6158490566037735,
|
| 713 |
+
"grad_norm": 0.4399295151233673,
|
| 714 |
+
"learning_rate": 7.881105049777902e-06,
|
| 715 |
+
"loss": 0.5118,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.6520754716981134,
|
| 720 |
+
"grad_norm": 0.5278156995773315,
|
| 721 |
+
"learning_rate": 7.828034377432694e-06,
|
| 722 |
+
"loss": 0.5592,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.688301886792453,
|
| 727 |
+
"grad_norm": 0.46402445435523987,
|
| 728 |
+
"learning_rate": 7.774491026603985e-06,
|
| 729 |
+
"loss": 0.5389,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.7245283018867923,
|
| 734 |
+
"grad_norm": 0.4599299430847168,
|
| 735 |
+
"learning_rate": 7.720483946542913e-06,
|
| 736 |
+
"loss": 0.5429,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.760754716981132,
|
| 741 |
+
"grad_norm": 0.4408857822418213,
|
| 742 |
+
"learning_rate": 7.666022164008458e-06,
|
| 743 |
+
"loss": 0.557,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.7969811320754716,
|
| 748 |
+
"grad_norm": 0.41980406641960144,
|
| 749 |
+
"learning_rate": 7.6111147817586925e-06,
|
| 750 |
+
"loss": 0.5434,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.8332075471698115,
|
| 755 |
+
"grad_norm": 0.5317021012306213,
|
| 756 |
+
"learning_rate": 7.5557709770293664e-06,
|
| 757 |
+
"loss": 0.552,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.869433962264151,
|
| 762 |
+
"grad_norm": 0.5827487111091614,
|
| 763 |
+
"learning_rate": 7.500000000000001e-06,
|
| 764 |
+
"loss": 0.5203,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.9056603773584904,
|
| 769 |
+
"grad_norm": 0.4034237265586853,
|
| 770 |
+
"learning_rate": 7.443811172247822e-06,
|
| 771 |
+
"loss": 0.5418,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.9418867924528302,
|
| 776 |
+
"grad_norm": 0.4142571985721588,
|
| 777 |
+
"learning_rate": 7.387213885189746e-06,
|
| 778 |
+
"loss": 0.5517,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.9781132075471697,
|
| 783 |
+
"grad_norm": 0.4551447033882141,
|
| 784 |
+
"learning_rate": 7.330217598512696e-06,
|
| 785 |
+
"loss": 0.5426,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 4.0,
|
| 790 |
+
"grad_norm": 0.5489457845687866,
|
| 791 |
+
"learning_rate": 7.2728318385925035e-06,
|
| 792 |
+
"loss": 0.5209,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 4.036226415094339,
|
| 797 |
+
"grad_norm": 0.5183960795402527,
|
| 798 |
+
"learning_rate": 7.215066196901676e-06,
|
| 799 |
+
"loss": 0.4952,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 4.072452830188679,
|
| 804 |
+
"grad_norm": 0.4089732766151428,
|
| 805 |
+
"learning_rate": 7.156930328406268e-06,
|
| 806 |
+
"loss": 0.4918,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 4.108679245283019,
|
| 811 |
+
"grad_norm": 0.47553887963294983,
|
| 812 |
+
"learning_rate": 7.098433949952146e-06,
|
| 813 |
+
"loss": 0.4745,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 4.144905660377359,
|
| 818 |
+
"grad_norm": 0.43845227360725403,
|
| 819 |
+
"learning_rate": 7.039586838640918e-06,
|
| 820 |
+
"loss": 0.463,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 4.181132075471698,
|
| 825 |
+
"grad_norm": 0.4647291898727417,
|
| 826 |
+
"learning_rate": 6.980398830195785e-06,
|
| 827 |
+
"loss": 0.5002,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 4.2173584905660375,
|
| 832 |
+
"grad_norm": 0.42274776101112366,
|
| 833 |
+
"learning_rate": 6.920879817317588e-06,
|
| 834 |
+
"loss": 0.4634,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 4.253584905660377,
|
| 839 |
+
"grad_norm": 0.46993109583854675,
|
| 840 |
+
"learning_rate": 6.861039748031351e-06,
|
| 841 |
+
"loss": 0.4527,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 4.289811320754717,
|
| 846 |
+
"grad_norm": 0.49515578150749207,
|
| 847 |
+
"learning_rate": 6.800888624023552e-06,
|
| 848 |
+
"loss": 0.473,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 4.326037735849057,
|
| 853 |
+
"grad_norm": 0.4713442027568817,
|
| 854 |
+
"learning_rate": 6.740436498970453e-06,
|
| 855 |
+
"loss": 0.4611,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 4.362264150943396,
|
| 860 |
+
"grad_norm": 0.4236859977245331,
|
| 861 |
+
"learning_rate": 6.679693476857712e-06,
|
| 862 |
+
"loss": 0.4632,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.398490566037736,
|
| 867 |
+
"grad_norm": 0.4841013252735138,
|
| 868 |
+
"learning_rate": 6.618669710291607e-06,
|
| 869 |
+
"loss": 0.4825,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.434716981132075,
|
| 874 |
+
"grad_norm": 0.5489901304244995,
|
| 875 |
+
"learning_rate": 6.557375398802124e-06,
|
| 876 |
+
"loss": 0.4679,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.470943396226415,
|
| 881 |
+
"grad_norm": 0.49812740087509155,
|
| 882 |
+
"learning_rate": 6.495820787138209e-06,
|
| 883 |
+
"loss": 0.4615,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.507169811320755,
|
| 888 |
+
"grad_norm": 0.3921999931335449,
|
| 889 |
+
"learning_rate": 6.434016163555452e-06,
|
| 890 |
+
"loss": 0.4664,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.543396226415094,
|
| 895 |
+
"grad_norm": 0.478518545627594,
|
| 896 |
+
"learning_rate": 6.371971858096509e-06,
|
| 897 |
+
"loss": 0.4754,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.579622641509434,
|
| 902 |
+
"grad_norm": 0.5080537796020508,
|
| 903 |
+
"learning_rate": 6.30969824086453e-06,
|
| 904 |
+
"loss": 0.442,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.615849056603773,
|
| 909 |
+
"grad_norm": 0.44495469331741333,
|
| 910 |
+
"learning_rate": 6.247205720289907e-06,
|
| 911 |
+
"loss": 0.4527,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.652075471698113,
|
| 916 |
+
"grad_norm": 0.4623711407184601,
|
| 917 |
+
"learning_rate": 6.184504741390596e-06,
|
| 918 |
+
"loss": 0.434,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.688301886792453,
|
| 923 |
+
"grad_norm": 0.38758838176727295,
|
| 924 |
+
"learning_rate": 6.121605784026339e-06,
|
| 925 |
+
"loss": 0.45,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.724528301886792,
|
| 930 |
+
"grad_norm": 0.4601069688796997,
|
| 931 |
+
"learning_rate": 6.058519361147055e-06,
|
| 932 |
+
"loss": 0.4655,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.760754716981132,
|
| 937 |
+
"grad_norm": 0.46330124139785767,
|
| 938 |
+
"learning_rate": 5.995256017035703e-06,
|
| 939 |
+
"loss": 0.4531,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.796981132075472,
|
| 944 |
+
"grad_norm": 0.5022570490837097,
|
| 945 |
+
"learning_rate": 5.931826325545912e-06,
|
| 946 |
+
"loss": 0.465,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.8332075471698115,
|
| 951 |
+
"grad_norm": 0.4626692533493042,
|
| 952 |
+
"learning_rate": 5.8682408883346535e-06,
|
| 953 |
+
"loss": 0.4511,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.869433962264151,
|
| 958 |
+
"grad_norm": 0.3865698575973511,
|
| 959 |
+
"learning_rate": 5.804510333090287e-06,
|
| 960 |
+
"loss": 0.46,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.90566037735849,
|
| 965 |
+
"grad_norm": 0.4084339737892151,
|
| 966 |
+
"learning_rate": 5.740645311756246e-06,
|
| 967 |
+
"loss": 0.4587,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.94188679245283,
|
| 972 |
+
"grad_norm": 0.42995211482048035,
|
| 973 |
+
"learning_rate": 5.6766564987506564e-06,
|
| 974 |
+
"loss": 0.4516,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.97811320754717,
|
| 979 |
+
"grad_norm": 0.41004374623298645,
|
| 980 |
+
"learning_rate": 5.612554589182228e-06,
|
| 981 |
+
"loss": 0.4644,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 5.0,
|
| 986 |
+
"grad_norm": 0.5633000135421753,
|
| 987 |
+
"learning_rate": 5.548350297062659e-06,
|
| 988 |
+
"loss": 0.4656,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 5.036226415094339,
|
| 993 |
+
"grad_norm": 0.4353588819503784,
|
| 994 |
+
"learning_rate": 5.484054353515896e-06,
|
| 995 |
+
"loss": 0.3881,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 5.072452830188679,
|
| 1000 |
+
"grad_norm": 0.4253406822681427,
|
| 1001 |
+
"learning_rate": 5.419677504984534e-06,
|
| 1002 |
+
"loss": 0.3969,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 5.108679245283019,
|
| 1007 |
+
"grad_norm": 0.46061569452285767,
|
| 1008 |
+
"learning_rate": 5.3552305114336515e-06,
|
| 1009 |
+
"loss": 0.3986,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 5.144905660377359,
|
| 1014 |
+
"grad_norm": 0.4197911024093628,
|
| 1015 |
+
"learning_rate": 5.290724144552379e-06,
|
| 1016 |
+
"loss": 0.4135,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 5.181132075471698,
|
| 1021 |
+
"grad_norm": 0.40193256735801697,
|
| 1022 |
+
"learning_rate": 5.2261691859535325e-06,
|
| 1023 |
+
"loss": 0.3908,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 5.2173584905660375,
|
| 1028 |
+
"grad_norm": 0.4688466787338257,
|
| 1029 |
+
"learning_rate": 5.161576425371554e-06,
|
| 1030 |
+
"loss": 0.3923,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 5.253584905660377,
|
| 1035 |
+
"grad_norm": 0.5070242881774902,
|
| 1036 |
+
"learning_rate": 5.096956658859122e-06,
|
| 1037 |
+
"loss": 0.3837,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 5.289811320754717,
|
| 1042 |
+
"grad_norm": 0.46500205993652344,
|
| 1043 |
+
"learning_rate": 5.032320686982697e-06,
|
| 1044 |
+
"loss": 0.3959,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 5.326037735849057,
|
| 1049 |
+
"grad_norm": 1.2324038743972778,
|
| 1050 |
+
"learning_rate": 4.967679313017304e-06,
|
| 1051 |
+
"loss": 0.3785,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 5.362264150943396,
|
| 1056 |
+
"grad_norm": 0.4382549524307251,
|
| 1057 |
+
"learning_rate": 4.903043341140879e-06,
|
| 1058 |
+
"loss": 0.3595,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 5.398490566037736,
|
| 1063 |
+
"grad_norm": 0.4779147803783417,
|
| 1064 |
+
"learning_rate": 4.838423574628447e-06,
|
| 1065 |
+
"loss": 0.369,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 5.434716981132075,
|
| 1070 |
+
"grad_norm": 0.506566047668457,
|
| 1071 |
+
"learning_rate": 4.773830814046469e-06,
|
| 1072 |
+
"loss": 0.3752,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 5.470943396226415,
|
| 1077 |
+
"grad_norm": 0.4850460886955261,
|
| 1078 |
+
"learning_rate": 4.7092758554476215e-06,
|
| 1079 |
+
"loss": 0.3805,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 5.507169811320755,
|
| 1084 |
+
"grad_norm": 0.4675082266330719,
|
| 1085 |
+
"learning_rate": 4.644769488566351e-06,
|
| 1086 |
+
"loss": 0.3696,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 5.543396226415094,
|
| 1091 |
+
"grad_norm": 0.4534352123737335,
|
| 1092 |
+
"learning_rate": 4.580322495015466e-06,
|
| 1093 |
+
"loss": 0.3937,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 5.579622641509434,
|
| 1098 |
+
"grad_norm": 0.4178565442562103,
|
| 1099 |
+
"learning_rate": 4.515945646484105e-06,
|
| 1100 |
+
"loss": 0.3673,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 5.615849056603773,
|
| 1105 |
+
"grad_norm": 0.43837058544158936,
|
| 1106 |
+
"learning_rate": 4.451649702937343e-06,
|
| 1107 |
+
"loss": 0.3932,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 5.652075471698113,
|
| 1112 |
+
"grad_norm": 0.5009051561355591,
|
| 1113 |
+
"learning_rate": 4.387445410817774e-06,
|
| 1114 |
+
"loss": 0.3767,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 5.688301886792453,
|
| 1119 |
+
"grad_norm": 0.4401375651359558,
|
| 1120 |
+
"learning_rate": 4.323343501249346e-06,
|
| 1121 |
+
"loss": 0.3836,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 5.724528301886792,
|
| 1126 |
+
"grad_norm": 0.49288874864578247,
|
| 1127 |
+
"learning_rate": 4.259354688243758e-06,
|
| 1128 |
+
"loss": 0.3735,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.760754716981132,
|
| 1133 |
+
"grad_norm": 0.41812247037887573,
|
| 1134 |
+
"learning_rate": 4.195489666909714e-06,
|
| 1135 |
+
"loss": 0.3632,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.796981132075472,
|
| 1140 |
+
"grad_norm": 0.4192414879798889,
|
| 1141 |
+
"learning_rate": 4.131759111665349e-06,
|
| 1142 |
+
"loss": 0.3742,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.8332075471698115,
|
| 1147 |
+
"grad_norm": 0.46856996417045593,
|
| 1148 |
+
"learning_rate": 4.06817367445409e-06,
|
| 1149 |
+
"loss": 0.3842,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.869433962264151,
|
| 1154 |
+
"grad_norm": 0.4367072582244873,
|
| 1155 |
+
"learning_rate": 4.004743982964298e-06,
|
| 1156 |
+
"loss": 0.3778,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.90566037735849,
|
| 1161 |
+
"grad_norm": 0.4130428433418274,
|
| 1162 |
+
"learning_rate": 3.941480638852948e-06,
|
| 1163 |
+
"loss": 0.392,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.94188679245283,
|
| 1168 |
+
"grad_norm": 0.44017598032951355,
|
| 1169 |
+
"learning_rate": 3.878394215973663e-06,
|
| 1170 |
+
"loss": 0.3997,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.97811320754717,
|
| 1175 |
+
"grad_norm": 0.44622254371643066,
|
| 1176 |
+
"learning_rate": 3.815495258609404e-06,
|
| 1177 |
+
"loss": 0.3827,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 6.0,
|
| 1182 |
+
"grad_norm": 0.44622254371643066,
|
| 1183 |
+
"learning_rate": 3.752794279710094e-06,
|
| 1184 |
+
"loss": 0.4119,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 6.036226415094339,
|
| 1189 |
+
"grad_norm": 0.6544888615608215,
|
| 1190 |
+
"learning_rate": 3.690301759135471e-06,
|
| 1191 |
+
"loss": 0.3342,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 6.072452830188679,
|
| 1196 |
+
"grad_norm": 0.48705703020095825,
|
| 1197 |
+
"learning_rate": 3.6280281419034934e-06,
|
| 1198 |
+
"loss": 0.3163,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 6.108679245283019,
|
| 1203 |
+
"grad_norm": 0.42752087116241455,
|
| 1204 |
+
"learning_rate": 3.5659838364445505e-06,
|
| 1205 |
+
"loss": 0.3442,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 6.144905660377359,
|
| 1210 |
+
"grad_norm": 0.3857191205024719,
|
| 1211 |
+
"learning_rate": 3.504179212861793e-06,
|
| 1212 |
+
"loss": 0.3319,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 6.181132075471698,
|
| 1217 |
+
"grad_norm": 0.46637335419654846,
|
| 1218 |
+
"learning_rate": 3.442624601197877e-06,
|
| 1219 |
+
"loss": 0.3167,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 6.2173584905660375,
|
| 1224 |
+
"grad_norm": 0.44155874848365784,
|
| 1225 |
+
"learning_rate": 3.3813302897083955e-06,
|
| 1226 |
+
"loss": 0.3032,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 6.253584905660377,
|
| 1231 |
+
"grad_norm": 0.4230786859989166,
|
| 1232 |
+
"learning_rate": 3.3203065231422904e-06,
|
| 1233 |
+
"loss": 0.3082,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 6.289811320754717,
|
| 1238 |
+
"grad_norm": 0.45516237616539,
|
| 1239 |
+
"learning_rate": 3.259563501029548e-06,
|
| 1240 |
+
"loss": 0.3323,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 6.326037735849057,
|
| 1245 |
+
"grad_norm": 0.4669758081436157,
|
| 1246 |
+
"learning_rate": 3.1991113759764493e-06,
|
| 1247 |
+
"loss": 0.3143,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 6.362264150943396,
|
| 1252 |
+
"grad_norm": 0.5168375968933105,
|
| 1253 |
+
"learning_rate": 3.1389602519686515e-06,
|
| 1254 |
+
"loss": 0.3154,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 6.398490566037736,
|
| 1259 |
+
"grad_norm": 0.4606865644454956,
|
| 1260 |
+
"learning_rate": 3.0791201826824117e-06,
|
| 1261 |
+
"loss": 0.3067,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 6.434716981132075,
|
| 1266 |
+
"grad_norm": 0.4242306053638458,
|
| 1267 |
+
"learning_rate": 3.019601169804216e-06,
|
| 1268 |
+
"loss": 0.3326,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 6.470943396226415,
|
| 1273 |
+
"grad_norm": 0.4492356777191162,
|
| 1274 |
+
"learning_rate": 2.9604131613590825e-06,
|
| 1275 |
+
"loss": 0.3422,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 6.507169811320755,
|
| 1280 |
+
"grad_norm": 0.4301685690879822,
|
| 1281 |
+
"learning_rate": 2.901566050047855e-06,
|
| 1282 |
+
"loss": 0.3071,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 6.543396226415094,
|
| 1287 |
+
"grad_norm": 0.46162980794906616,
|
| 1288 |
+
"learning_rate": 2.843069671593734e-06,
|
| 1289 |
+
"loss": 0.3084,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 6.579622641509434,
|
| 1294 |
+
"grad_norm": 0.43854185938835144,
|
| 1295 |
+
"learning_rate": 2.784933803098326e-06,
|
| 1296 |
+
"loss": 0.3256,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 6.615849056603773,
|
| 1301 |
+
"grad_norm": 0.46818047761917114,
|
| 1302 |
+
"learning_rate": 2.7271681614074973e-06,
|
| 1303 |
+
"loss": 0.3109,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 6.652075471698113,
|
| 1308 |
+
"grad_norm": 0.3902846872806549,
|
| 1309 |
+
"learning_rate": 2.6697824014873076e-06,
|
| 1310 |
+
"loss": 0.303,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 6.688301886792453,
|
| 1315 |
+
"grad_norm": 0.4288512170314789,
|
| 1316 |
+
"learning_rate": 2.6127861148102552e-06,
|
| 1317 |
+
"loss": 0.2988,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 6.724528301886792,
|
| 1322 |
+
"grad_norm": 0.3952915668487549,
|
| 1323 |
+
"learning_rate": 2.5561888277521797e-06,
|
| 1324 |
+
"loss": 0.3003,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 6.760754716981132,
|
| 1329 |
+
"grad_norm": 0.4569963812828064,
|
| 1330 |
+
"learning_rate": 2.5000000000000015e-06,
|
| 1331 |
+
"loss": 0.3145,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 6.796981132075472,
|
| 1336 |
+
"grad_norm": 0.448163241147995,
|
| 1337 |
+
"learning_rate": 2.4442290229706344e-06,
|
| 1338 |
+
"loss": 0.3302,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 6.8332075471698115,
|
| 1343 |
+
"grad_norm": 0.44628778100013733,
|
| 1344 |
+
"learning_rate": 2.3888852182413087e-06,
|
| 1345 |
+
"loss": 0.3256,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 6.869433962264151,
|
| 1350 |
+
"grad_norm": 0.4496423900127411,
|
| 1351 |
+
"learning_rate": 2.333977835991545e-06,
|
| 1352 |
+
"loss": 0.3399,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 6.90566037735849,
|
| 1357 |
+
"grad_norm": 0.4480762481689453,
|
| 1358 |
+
"learning_rate": 2.2795160534570866e-06,
|
| 1359 |
+
"loss": 0.3202,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 6.94188679245283,
|
| 1364 |
+
"grad_norm": 0.4092267155647278,
|
| 1365 |
+
"learning_rate": 2.2255089733960162e-06,
|
| 1366 |
+
"loss": 0.3302,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 6.97811320754717,
|
| 1371 |
+
"grad_norm": 0.43138426542282104,
|
| 1372 |
+
"learning_rate": 2.171965622567308e-06,
|
| 1373 |
+
"loss": 0.3106,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 7.0,
|
| 1378 |
+
"grad_norm": 0.5887420773506165,
|
| 1379 |
+
"learning_rate": 2.1188949502220987e-06,
|
| 1380 |
+
"loss": 0.3219,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 7.036226415094339,
|
| 1385 |
+
"grad_norm": 0.556282103061676,
|
| 1386 |
+
"learning_rate": 2.066305826607911e-06,
|
| 1387 |
+
"loss": 0.3141,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 7.072452830188679,
|
| 1392 |
+
"grad_norm": 0.4548152983188629,
|
| 1393 |
+
"learning_rate": 2.0142070414860704e-06,
|
| 1394 |
+
"loss": 0.2697,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 7.108679245283019,
|
| 1399 |
+
"grad_norm": 0.45477020740509033,
|
| 1400 |
+
"learning_rate": 1.962607302662582e-06,
|
| 1401 |
+
"loss": 0.27,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 7.144905660377359,
|
| 1406 |
+
"grad_norm": 0.3835766911506653,
|
| 1407 |
+
"learning_rate": 1.9115152345327154e-06,
|
| 1408 |
+
"loss": 0.2718,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 7.181132075471698,
|
| 1413 |
+
"grad_norm": 0.4106207489967346,
|
| 1414 |
+
"learning_rate": 1.8609393766395083e-06,
|
| 1415 |
+
"loss": 0.2902,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 7.2173584905660375,
|
| 1420 |
+
"grad_norm": 0.40980860590934753,
|
| 1421 |
+
"learning_rate": 1.8108881822464697e-06,
|
| 1422 |
+
"loss": 0.294,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 7.253584905660377,
|
| 1427 |
+
"grad_norm": 0.4507136940956116,
|
| 1428 |
+
"learning_rate": 1.7613700169247055e-06,
|
| 1429 |
+
"loss": 0.2941,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 7.289811320754717,
|
| 1434 |
+
"grad_norm": 0.3944286108016968,
|
| 1435 |
+
"learning_rate": 1.7123931571546826e-06,
|
| 1436 |
+
"loss": 0.2596,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 7.326037735849057,
|
| 1441 |
+
"grad_norm": 0.42168113589286804,
|
| 1442 |
+
"learning_rate": 1.6639657889429017e-06,
|
| 1443 |
+
"loss": 0.2757,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 7.362264150943396,
|
| 1448 |
+
"grad_norm": 0.45149892568588257,
|
| 1449 |
+
"learning_rate": 1.6160960064536907e-06,
|
| 1450 |
+
"loss": 0.2618,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 7.398490566037736,
|
| 1455 |
+
"grad_norm": 0.4606323540210724,
|
| 1456 |
+
"learning_rate": 1.5687918106563326e-06,
|
| 1457 |
+
"loss": 0.2756,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 7.434716981132075,
|
| 1462 |
+
"grad_norm": 0.5498378872871399,
|
| 1463 |
+
"learning_rate": 1.52206110798779e-06,
|
| 1464 |
+
"loss": 0.2818,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 7.470943396226415,
|
| 1469 |
+
"grad_norm": 0.4124191403388977,
|
| 1470 |
+
"learning_rate": 1.4759117090312197e-06,
|
| 1471 |
+
"loss": 0.2731,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 7.507169811320755,
|
| 1476 |
+
"grad_norm": 0.44834864139556885,
|
| 1477 |
+
"learning_rate": 1.4303513272105057e-06,
|
| 1478 |
+
"loss": 0.2687,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 7.543396226415094,
|
| 1483 |
+
"grad_norm": 0.39862528443336487,
|
| 1484 |
+
"learning_rate": 1.3853875775010355e-06,
|
| 1485 |
+
"loss": 0.2688,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 7.579622641509434,
|
| 1490 |
+
"grad_norm": 0.48807084560394287,
|
| 1491 |
+
"learning_rate": 1.3410279751569399e-06,
|
| 1492 |
+
"loss": 0.3119,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 7.615849056603773,
|
| 1497 |
+
"grad_norm": 0.44604766368865967,
|
| 1498 |
+
"learning_rate": 1.297279934454978e-06,
|
| 1499 |
+
"loss": 0.2621,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 7.652075471698113,
|
| 1504 |
+
"grad_norm": 0.3903985619544983,
|
| 1505 |
+
"learning_rate": 1.25415076745532e-06,
|
| 1506 |
+
"loss": 0.2745,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 7.688301886792453,
|
| 1511 |
+
"grad_norm": 0.4241688549518585,
|
| 1512 |
+
"learning_rate": 1.2116476827794104e-06,
|
| 1513 |
+
"loss": 0.264,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 7.724528301886792,
|
| 1518 |
+
"grad_norm": 0.4157409369945526,
|
| 1519 |
+
"learning_rate": 1.1697777844051105e-06,
|
| 1520 |
+
"loss": 0.2695,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 7.760754716981132,
|
| 1525 |
+
"grad_norm": 0.42209798097610474,
|
| 1526 |
+
"learning_rate": 1.1285480704793378e-06,
|
| 1527 |
+
"loss": 0.293,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 7.796981132075472,
|
| 1532 |
+
"grad_norm": 0.39040008187294006,
|
| 1533 |
+
"learning_rate": 1.0879654321484012e-06,
|
| 1534 |
+
"loss": 0.2443,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 7.8332075471698115,
|
| 1539 |
+
"grad_norm": 0.4144212603569031,
|
| 1540 |
+
"learning_rate": 1.0480366524062041e-06,
|
| 1541 |
+
"loss": 0.2747,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 7.869433962264151,
|
| 1546 |
+
"grad_norm": 0.38053098320961,
|
| 1547 |
+
"learning_rate": 1.008768404960535e-06,
|
| 1548 |
+
"loss": 0.2634,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 7.90566037735849,
|
| 1553 |
+
"grad_norm": 0.4866553246974945,
|
| 1554 |
+
"learning_rate": 9.701672531176287e-07,
|
| 1555 |
+
"loss": 0.2883,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 7.94188679245283,
|
| 1560 |
+
"grad_norm": 0.401796817779541,
|
| 1561 |
+
"learning_rate": 9.322396486851626e-07,
|
| 1562 |
+
"loss": 0.2885,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 7.97811320754717,
|
| 1567 |
+
"grad_norm": 0.4356318414211273,
|
| 1568 |
+
"learning_rate": 8.949919308939081e-07,
|
| 1569 |
+
"loss": 0.2985,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 8.0,
|
| 1574 |
+
"grad_norm": 0.4356318414211273,
|
| 1575 |
+
"learning_rate": 8.584303253381848e-07,
|
| 1576 |
+
"loss": 0.2454,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 8.03622641509434,
|
| 1581 |
+
"grad_norm": 0.5875634551048279,
|
| 1582 |
+
"learning_rate": 8.225609429353187e-07,
|
| 1583 |
+
"loss": 0.2752,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 8.072452830188679,
|
| 1588 |
+
"grad_norm": 0.4065380394458771,
|
| 1589 |
+
"learning_rate": 7.873897789042523e-07,
|
| 1590 |
+
"loss": 0.2725,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 8.10867924528302,
|
| 1595 |
+
"grad_norm": 0.47121724486351013,
|
| 1596 |
+
"learning_rate": 7.529227117635135e-07,
|
| 1597 |
+
"loss": 0.2802,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 8.144905660377358,
|
| 1602 |
+
"grad_norm": 0.41164955496788025,
|
| 1603 |
+
"learning_rate": 7.191655023486682e-07,
|
| 1604 |
+
"loss": 0.2475,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 8.181132075471698,
|
| 1609 |
+
"grad_norm": 0.40498465299606323,
|
| 1610 |
+
"learning_rate": 6.86123792849458e-07,
|
| 1611 |
+
"loss": 0.2585,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 8.217358490566038,
|
| 1616 |
+
"grad_norm": 0.39466381072998047,
|
| 1617 |
+
"learning_rate": 6.53803105866761e-07,
|
| 1618 |
+
"loss": 0.2552,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 8.253584905660377,
|
| 1623 |
+
"grad_norm": 0.3694005310535431,
|
| 1624 |
+
"learning_rate": 6.222088434895462e-07,
|
| 1625 |
+
"loss": 0.266,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 8.289811320754717,
|
| 1630 |
+
"grad_norm": 0.3654123544692993,
|
| 1631 |
+
"learning_rate": 5.9134628639196e-07,
|
| 1632 |
+
"loss": 0.261,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 8.326037735849056,
|
| 1637 |
+
"grad_norm": 0.3835010230541229,
|
| 1638 |
+
"learning_rate": 5.612205929507209e-07,
|
| 1639 |
+
"loss": 0.262,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 8.362264150943396,
|
| 1644 |
+
"grad_norm": 0.4232535660266876,
|
| 1645 |
+
"learning_rate": 5.318367983829393e-07,
|
| 1646 |
+
"loss": 0.2443,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 8.398490566037736,
|
| 1651 |
+
"grad_norm": 0.3865967392921448,
|
| 1652 |
+
"learning_rate": 5.031998139045352e-07,
|
| 1653 |
+
"loss": 0.2479,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 8.434716981132075,
|
| 1658 |
+
"grad_norm": 0.3988674581050873,
|
| 1659 |
+
"learning_rate": 4.753144259093734e-07,
|
| 1660 |
+
"loss": 0.2569,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 8.470943396226415,
|
| 1665 |
+
"grad_norm": 0.40713047981262207,
|
| 1666 |
+
"learning_rate": 4.481852951692672e-07,
|
| 1667 |
+
"loss": 0.2446,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 8.507169811320754,
|
| 1672 |
+
"grad_norm": 0.3784768283367157,
|
| 1673 |
+
"learning_rate": 4.2181695605497066e-07,
|
| 1674 |
+
"loss": 0.2561,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 8.543396226415094,
|
| 1679 |
+
"grad_norm": 0.39897289872169495,
|
| 1680 |
+
"learning_rate": 3.9621381577830855e-07,
|
| 1681 |
+
"loss": 0.2504,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 8.579622641509435,
|
| 1686 |
+
"grad_norm": 0.40435102581977844,
|
| 1687 |
+
"learning_rate": 3.7138015365554834e-07,
|
| 1688 |
+
"loss": 0.2572,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 8.615849056603773,
|
| 1693 |
+
"grad_norm": 0.3976576328277588,
|
| 1694 |
+
"learning_rate": 3.473201203921578e-07,
|
| 1695 |
+
"loss": 0.2622,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 8.652075471698113,
|
| 1700 |
+
"grad_norm": 0.40374553203582764,
|
| 1701 |
+
"learning_rate": 3.2403773738905185e-07,
|
| 1702 |
+
"loss": 0.2302,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 8.688301886792452,
|
| 1707 |
+
"grad_norm": 0.39839479327201843,
|
| 1708 |
+
"learning_rate": 3.015368960704584e-07,
|
| 1709 |
+
"loss": 0.2658,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 8.724528301886792,
|
| 1714 |
+
"grad_norm": 0.3632482588291168,
|
| 1715 |
+
"learning_rate": 2.798213572335001e-07,
|
| 1716 |
+
"loss": 0.2435,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 8.760754716981133,
|
| 1721 |
+
"grad_norm": 0.37311646342277527,
|
| 1722 |
+
"learning_rate": 2.5889475041961767e-07,
|
| 1723 |
+
"loss": 0.2427,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 8.796981132075471,
|
| 1728 |
+
"grad_norm": 0.43890783190727234,
|
| 1729 |
+
"learning_rate": 2.3876057330792344e-07,
|
| 1730 |
+
"loss": 0.2576,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 8.833207547169811,
|
| 1735 |
+
"grad_norm": 0.3749382197856903,
|
| 1736 |
+
"learning_rate": 2.1942219113060215e-07,
|
| 1737 |
+
"loss": 0.2588,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 8.86943396226415,
|
| 1742 |
+
"grad_norm": 0.41522374749183655,
|
| 1743 |
+
"learning_rate": 2.0088283611044034e-07,
|
| 1744 |
+
"loss": 0.2526,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 8.90566037735849,
|
| 1749 |
+
"grad_norm": 0.3923836648464203,
|
| 1750 |
+
"learning_rate": 1.8314560692059836e-07,
|
| 1751 |
+
"loss": 0.2504,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 8.94188679245283,
|
| 1756 |
+
"grad_norm": 0.40502244234085083,
|
| 1757 |
+
"learning_rate": 1.6621346816668993e-07,
|
| 1758 |
+
"loss": 0.2409,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 8.97811320754717,
|
| 1763 |
+
"grad_norm": 0.391658753156662,
|
| 1764 |
+
"learning_rate": 1.500892498912826e-07,
|
| 1765 |
+
"loss": 0.2607,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 9.0,
|
| 1770 |
+
"grad_norm": 0.5146093964576721,
|
| 1771 |
+
"learning_rate": 1.3477564710088097e-07,
|
| 1772 |
+
"loss": 0.2573,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 9.03622641509434,
|
| 1777 |
+
"grad_norm": 0.5422486662864685,
|
| 1778 |
+
"learning_rate": 1.2027521931548214e-07,
|
| 1779 |
+
"loss": 0.2379,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 9.072452830188679,
|
| 1784 |
+
"grad_norm": 0.40020889043807983,
|
| 1785 |
+
"learning_rate": 1.0659039014077943e-07,
|
| 1786 |
+
"loss": 0.2412,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 9.10867924528302,
|
| 1791 |
+
"grad_norm": 0.3687814772129059,
|
| 1792 |
+
"learning_rate": 9.372344686307655e-08,
|
| 1793 |
+
"loss": 0.2409,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 9.144905660377358,
|
| 1798 |
+
"grad_norm": 0.380043089389801,
|
| 1799 |
+
"learning_rate": 8.167654006699444e-08,
|
| 1800 |
+
"loss": 0.2699,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 9.181132075471698,
|
| 1805 |
+
"grad_norm": 0.39074811339378357,
|
| 1806 |
+
"learning_rate": 7.04516832760177e-08,
|
| 1807 |
+
"loss": 0.2489,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 9.217358490566038,
|
| 1812 |
+
"grad_norm": 0.38397496938705444,
|
| 1813 |
+
"learning_rate": 6.005075261595495e-08,
|
| 1814 |
+
"loss": 0.2585,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 9.253584905660377,
|
| 1819 |
+
"grad_norm": 0.4563198685646057,
|
| 1820 |
+
"learning_rate": 5.047548650136513e-08,
|
| 1821 |
+
"loss": 0.2351,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 9.289811320754717,
|
| 1826 |
+
"grad_norm": 0.3548761308193207,
|
| 1827 |
+
"learning_rate": 4.172748534499449e-08,
|
| 1828 |
+
"loss": 0.2508,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 9.326037735849056,
|
| 1833 |
+
"grad_norm": 0.41531455516815186,
|
| 1834 |
+
"learning_rate": 3.3808211290284886e-08,
|
| 1835 |
+
"loss": 0.2534,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 9.362264150943396,
|
| 1840 |
+
"grad_norm": 0.34948405623435974,
|
| 1841 |
+
"learning_rate": 2.6718987966992683e-08,
|
| 1842 |
+
"loss": 0.2497,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 9.398490566037736,
|
| 1847 |
+
"grad_norm": 0.3671889305114746,
|
| 1848 |
+
"learning_rate": 2.0461000269953457e-08,
|
| 1849 |
+
"loss": 0.2317,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 9.434716981132075,
|
| 1854 |
+
"grad_norm": 0.43335914611816406,
|
| 1855 |
+
"learning_rate": 1.5035294161039882e-08,
|
| 1856 |
+
"loss": 0.2353,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 9.470943396226415,
|
| 1861 |
+
"grad_norm": 0.41739046573638916,
|
| 1862 |
+
"learning_rate": 1.044277649433989e-08,
|
| 1863 |
+
"loss": 0.2589,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 9.507169811320754,
|
| 1868 |
+
"grad_norm": 0.36336734890937805,
|
| 1869 |
+
"learning_rate": 6.6842148645840374e-09,
|
| 1870 |
+
"loss": 0.2228,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 9.543396226415094,
|
| 1875 |
+
"grad_norm": 0.3958076238632202,
|
| 1876 |
+
"learning_rate": 3.760237478849793e-09,
|
| 1877 |
+
"loss": 0.2405,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 9.579622641509435,
|
| 1882 |
+
"grad_norm": 0.40326353907585144,
|
| 1883 |
+
"learning_rate": 1.6713330515627512e-09,
|
| 1884 |
+
"loss": 0.2645,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 9.615849056603773,
|
| 1889 |
+
"grad_norm": 0.38379284739494324,
|
| 1890 |
+
"learning_rate": 4.178507228136397e-10,
|
| 1891 |
+
"loss": 0.2699,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 9.652075471698113,
|
| 1896 |
+
"grad_norm": 0.4049612283706665,
|
| 1897 |
+
"learning_rate": 0.0,
|
| 1898 |
+
"loss": 0.2758,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
}
|
| 1901 |
+
],
|
| 1902 |
+
"logging_steps": 1,
|
| 1903 |
+
"max_steps": 270,
|
| 1904 |
+
"num_input_tokens_seen": 0,
|
| 1905 |
+
"num_train_epochs": 10,
|
| 1906 |
+
"save_steps": 100000,
|
| 1907 |
+
"stateful_callbacks": {
|
| 1908 |
+
"TrainerControl": {
|
| 1909 |
+
"args": {
|
| 1910 |
+
"should_epoch_stop": false,
|
| 1911 |
+
"should_evaluate": false,
|
| 1912 |
+
"should_log": false,
|
| 1913 |
+
"should_save": true,
|
| 1914 |
+
"should_training_stop": true
|
| 1915 |
+
},
|
| 1916 |
+
"attributes": {}
|
| 1917 |
+
}
|
| 1918 |
+
},
|
| 1919 |
+
"total_flos": 99687939342336.0,
|
| 1920 |
+
"train_batch_size": 1,
|
| 1921 |
+
"trial_name": null,
|
| 1922 |
+
"trial_params": null
|
| 1923 |
+
}
|