Upload checkpoint-325/trainer_state.json with huggingface_hub
Browse files- checkpoint-325/trainer_state.json +2308 -0
checkpoint-325/trainer_state.json
ADDED
|
@@ -0,0 +1,2308 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 325,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.003076923076923077,
|
| 13 |
+
"grad_norm": 0.7828734161781798,
|
| 14 |
+
"learning_rate": 1e-05,
|
| 15 |
+
"loss": 1.2372,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.006153846153846154,
|
| 20 |
+
"grad_norm": 0.4715545538601974,
|
| 21 |
+
"learning_rate": 2e-05,
|
| 22 |
+
"loss": 1.0857,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.009230769230769232,
|
| 27 |
+
"grad_norm": 0.5033313729985242,
|
| 28 |
+
"learning_rate": 3e-05,
|
| 29 |
+
"loss": 1.1008,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.012307692307692308,
|
| 34 |
+
"grad_norm": 0.5640856762548262,
|
| 35 |
+
"learning_rate": 4e-05,
|
| 36 |
+
"loss": 1.1982,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.015384615384615385,
|
| 41 |
+
"grad_norm": 0.49052022150787394,
|
| 42 |
+
"learning_rate": 5e-05,
|
| 43 |
+
"loss": 0.9812,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.018461538461538463,
|
| 48 |
+
"grad_norm": 0.5760588560035741,
|
| 49 |
+
"learning_rate": 6e-05,
|
| 50 |
+
"loss": 1.0001,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.021538461538461538,
|
| 55 |
+
"grad_norm": 0.5391568703825995,
|
| 56 |
+
"learning_rate": 7e-05,
|
| 57 |
+
"loss": 0.9786,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.024615384615384615,
|
| 62 |
+
"grad_norm": 0.9271304349117271,
|
| 63 |
+
"learning_rate": 8e-05,
|
| 64 |
+
"loss": 1.0427,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.027692307692307693,
|
| 69 |
+
"grad_norm": 0.8902295238380683,
|
| 70 |
+
"learning_rate": 9e-05,
|
| 71 |
+
"loss": 1.0948,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.03076923076923077,
|
| 76 |
+
"grad_norm": 0.7593057197592625,
|
| 77 |
+
"learning_rate": 0.0001,
|
| 78 |
+
"loss": 1.0118,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.033846153846153845,
|
| 83 |
+
"grad_norm": 0.5206011307827645,
|
| 84 |
+
"learning_rate": 9.999751334779716e-05,
|
| 85 |
+
"loss": 0.9203,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.036923076923076927,
|
| 90 |
+
"grad_norm": 0.741308793429257,
|
| 91 |
+
"learning_rate": 9.999005363852618e-05,
|
| 92 |
+
"loss": 1.0539,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.04,
|
| 97 |
+
"grad_norm": 0.6228043771559892,
|
| 98 |
+
"learning_rate": 9.997762161417517e-05,
|
| 99 |
+
"loss": 1.0793,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.043076923076923075,
|
| 104 |
+
"grad_norm": 0.555962480752347,
|
| 105 |
+
"learning_rate": 9.996021851130897e-05,
|
| 106 |
+
"loss": 0.9755,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.046153846153846156,
|
| 111 |
+
"grad_norm": 0.7993641287111947,
|
| 112 |
+
"learning_rate": 9.993784606094612e-05,
|
| 113 |
+
"loss": 0.9151,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.04923076923076923,
|
| 118 |
+
"grad_norm": 1.005615485098892,
|
| 119 |
+
"learning_rate": 9.991050648838675e-05,
|
| 120 |
+
"loss": 1.166,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.052307692307692305,
|
| 125 |
+
"grad_norm": 0.6248236198693158,
|
| 126 |
+
"learning_rate": 9.987820251299122e-05,
|
| 127 |
+
"loss": 0.8714,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.055384615384615386,
|
| 132 |
+
"grad_norm": 0.6080910544572333,
|
| 133 |
+
"learning_rate": 9.984093734790956e-05,
|
| 134 |
+
"loss": 1.0885,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.05846153846153846,
|
| 139 |
+
"grad_norm": 0.7205882469108739,
|
| 140 |
+
"learning_rate": 9.979871469976196e-05,
|
| 141 |
+
"loss": 0.8567,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.06153846153846154,
|
| 146 |
+
"grad_norm": 0.8048642614138823,
|
| 147 |
+
"learning_rate": 9.975153876827008e-05,
|
| 148 |
+
"loss": 1.2165,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.06461538461538462,
|
| 153 |
+
"grad_norm": 0.5673042921183786,
|
| 154 |
+
"learning_rate": 9.969941424583926e-05,
|
| 155 |
+
"loss": 0.969,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.06769230769230769,
|
| 160 |
+
"grad_norm": 0.9395972712573123,
|
| 161 |
+
"learning_rate": 9.964234631709187e-05,
|
| 162 |
+
"loss": 1.0573,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.07076923076923076,
|
| 167 |
+
"grad_norm": 0.45685403890339,
|
| 168 |
+
"learning_rate": 9.958034065835151e-05,
|
| 169 |
+
"loss": 0.8941,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.07384615384615385,
|
| 174 |
+
"grad_norm": 0.5775586971652089,
|
| 175 |
+
"learning_rate": 9.951340343707852e-05,
|
| 176 |
+
"loss": 0.9429,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.07692307692307693,
|
| 181 |
+
"grad_norm": 0.4997732657364997,
|
| 182 |
+
"learning_rate": 9.944154131125642e-05,
|
| 183 |
+
"loss": 0.9163,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.08,
|
| 188 |
+
"grad_norm": 0.5924468324567289,
|
| 189 |
+
"learning_rate": 9.936476142872979e-05,
|
| 190 |
+
"loss": 0.8522,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.08307692307692308,
|
| 195 |
+
"grad_norm": 0.4575891538856398,
|
| 196 |
+
"learning_rate": 9.928307142649316e-05,
|
| 197 |
+
"loss": 0.9213,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.08615384615384615,
|
| 202 |
+
"grad_norm": 0.48176644114702594,
|
| 203 |
+
"learning_rate": 9.919647942993148e-05,
|
| 204 |
+
"loss": 0.9972,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.08923076923076922,
|
| 209 |
+
"grad_norm": 0.6459061800746799,
|
| 210 |
+
"learning_rate": 9.910499405201195e-05,
|
| 211 |
+
"loss": 1.0326,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.09230769230769231,
|
| 216 |
+
"grad_norm": 0.5093858673996351,
|
| 217 |
+
"learning_rate": 9.900862439242719e-05,
|
| 218 |
+
"loss": 0.8343,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.09538461538461539,
|
| 223 |
+
"grad_norm": 0.7351279185345502,
|
| 224 |
+
"learning_rate": 9.890738003669029e-05,
|
| 225 |
+
"loss": 0.9729,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.09846153846153846,
|
| 230 |
+
"grad_norm": 0.5948596828470558,
|
| 231 |
+
"learning_rate": 9.880127105518122e-05,
|
| 232 |
+
"loss": 0.8813,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.10153846153846154,
|
| 237 |
+
"grad_norm": 0.9802930000513344,
|
| 238 |
+
"learning_rate": 9.869030800214532e-05,
|
| 239 |
+
"loss": 1.0616,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.10461538461538461,
|
| 244 |
+
"grad_norm": 1.3163815685405398,
|
| 245 |
+
"learning_rate": 9.857450191464337e-05,
|
| 246 |
+
"loss": 1.001,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.1076923076923077,
|
| 251 |
+
"grad_norm": 0.8304474495559376,
|
| 252 |
+
"learning_rate": 9.84538643114539e-05,
|
| 253 |
+
"loss": 0.9985,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.11076923076923077,
|
| 258 |
+
"grad_norm": 0.45718546447045166,
|
| 259 |
+
"learning_rate": 9.832840719192736e-05,
|
| 260 |
+
"loss": 0.8094,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.11384615384615385,
|
| 265 |
+
"grad_norm": 0.538791563191725,
|
| 266 |
+
"learning_rate": 9.819814303479267e-05,
|
| 267 |
+
"loss": 0.9526,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.11692307692307692,
|
| 272 |
+
"grad_norm": 0.4484350585807759,
|
| 273 |
+
"learning_rate": 9.806308479691595e-05,
|
| 274 |
+
"loss": 0.787,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.12,
|
| 279 |
+
"grad_norm": 0.47971163416575835,
|
| 280 |
+
"learning_rate": 9.792324591201179e-05,
|
| 281 |
+
"loss": 0.9883,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.12307692307692308,
|
| 286 |
+
"grad_norm": 0.6755244292781664,
|
| 287 |
+
"learning_rate": 9.777864028930705e-05,
|
| 288 |
+
"loss": 0.8712,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.12615384615384614,
|
| 293 |
+
"grad_norm": 0.46493563113837394,
|
| 294 |
+
"learning_rate": 9.76292823121573e-05,
|
| 295 |
+
"loss": 0.8766,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.12923076923076923,
|
| 300 |
+
"grad_norm": 0.5047464241669725,
|
| 301 |
+
"learning_rate": 9.747518683661631e-05,
|
| 302 |
+
"loss": 0.6885,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.13230769230769232,
|
| 307 |
+
"grad_norm": 0.6646703813442345,
|
| 308 |
+
"learning_rate": 9.731636918995821e-05,
|
| 309 |
+
"loss": 0.9051,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.13538461538461538,
|
| 314 |
+
"grad_norm": 0.5088693590519476,
|
| 315 |
+
"learning_rate": 9.715284516915303e-05,
|
| 316 |
+
"loss": 0.9609,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.13846153846153847,
|
| 321 |
+
"grad_norm": 0.49272781859864084,
|
| 322 |
+
"learning_rate": 9.698463103929542e-05,
|
| 323 |
+
"loss": 0.8649,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.14153846153846153,
|
| 328 |
+
"grad_norm": 0.4513921704981154,
|
| 329 |
+
"learning_rate": 9.681174353198687e-05,
|
| 330 |
+
"loss": 0.8791,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.14461538461538462,
|
| 335 |
+
"grad_norm": 0.5141540107669933,
|
| 336 |
+
"learning_rate": 9.663419984367139e-05,
|
| 337 |
+
"loss": 0.9726,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.1476923076923077,
|
| 342 |
+
"grad_norm": 0.4627723165753549,
|
| 343 |
+
"learning_rate": 9.645201763392513e-05,
|
| 344 |
+
"loss": 0.8008,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.15076923076923077,
|
| 349 |
+
"grad_norm": 0.47275901653963637,
|
| 350 |
+
"learning_rate": 9.626521502369984e-05,
|
| 351 |
+
"loss": 1.0066,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.15384615384615385,
|
| 356 |
+
"grad_norm": 0.46412801721657043,
|
| 357 |
+
"learning_rate": 9.607381059352038e-05,
|
| 358 |
+
"loss": 0.8405,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.15692307692307692,
|
| 363 |
+
"grad_norm": 0.5007712005677752,
|
| 364 |
+
"learning_rate": 9.587782338163669e-05,
|
| 365 |
+
"loss": 1.0018,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.16,
|
| 370 |
+
"grad_norm": 0.5356046979040091,
|
| 371 |
+
"learning_rate": 9.567727288213005e-05,
|
| 372 |
+
"loss": 0.9855,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.16307692307692306,
|
| 377 |
+
"grad_norm": 0.645985993733615,
|
| 378 |
+
"learning_rate": 9.547217904297411e-05,
|
| 379 |
+
"loss": 1.0759,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.16615384615384615,
|
| 384 |
+
"grad_norm": 0.3815289627209956,
|
| 385 |
+
"learning_rate": 9.526256226405075e-05,
|
| 386 |
+
"loss": 0.8487,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.16923076923076924,
|
| 391 |
+
"grad_norm": 1.0766576156243417,
|
| 392 |
+
"learning_rate": 9.504844339512095e-05,
|
| 393 |
+
"loss": 1.0027,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.1723076923076923,
|
| 398 |
+
"grad_norm": 0.8936487754669554,
|
| 399 |
+
"learning_rate": 9.482984373375105e-05,
|
| 400 |
+
"loss": 1.01,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.1753846153846154,
|
| 405 |
+
"grad_norm": 0.7266888128033469,
|
| 406 |
+
"learning_rate": 9.460678502319418e-05,
|
| 407 |
+
"loss": 1.0088,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.17846153846153845,
|
| 412 |
+
"grad_norm": 0.6225342856766284,
|
| 413 |
+
"learning_rate": 9.437928945022771e-05,
|
| 414 |
+
"loss": 1.1003,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.18153846153846154,
|
| 419 |
+
"grad_norm": 0.49280393912601883,
|
| 420 |
+
"learning_rate": 9.414737964294636e-05,
|
| 421 |
+
"loss": 0.95,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.18461538461538463,
|
| 426 |
+
"grad_norm": 0.7367352507716205,
|
| 427 |
+
"learning_rate": 9.391107866851143e-05,
|
| 428 |
+
"loss": 0.9928,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.18769230769230769,
|
| 433 |
+
"grad_norm": 0.5900723931116799,
|
| 434 |
+
"learning_rate": 9.367041003085649e-05,
|
| 435 |
+
"loss": 1.0394,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.19076923076923077,
|
| 440 |
+
"grad_norm": 0.8312576634409671,
|
| 441 |
+
"learning_rate": 9.342539766834946e-05,
|
| 442 |
+
"loss": 1.1836,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.19384615384615383,
|
| 447 |
+
"grad_norm": 0.7034475623841068,
|
| 448 |
+
"learning_rate": 9.317606595141154e-05,
|
| 449 |
+
"loss": 0.9995,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.19692307692307692,
|
| 454 |
+
"grad_norm": 0.620737330914189,
|
| 455 |
+
"learning_rate": 9.292243968009331e-05,
|
| 456 |
+
"loss": 0.8001,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.2,
|
| 461 |
+
"grad_norm": 0.49478297906177915,
|
| 462 |
+
"learning_rate": 9.266454408160779e-05,
|
| 463 |
+
"loss": 0.937,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.20307692307692307,
|
| 468 |
+
"grad_norm": 0.6455527455726527,
|
| 469 |
+
"learning_rate": 9.24024048078213e-05,
|
| 470 |
+
"loss": 0.8769,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.20615384615384616,
|
| 475 |
+
"grad_norm": 0.43294595955738263,
|
| 476 |
+
"learning_rate": 9.213604793270196e-05,
|
| 477 |
+
"loss": 0.9345,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.20923076923076922,
|
| 482 |
+
"grad_norm": 0.45789125605280445,
|
| 483 |
+
"learning_rate": 9.186549994972618e-05,
|
| 484 |
+
"loss": 0.7951,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.2123076923076923,
|
| 489 |
+
"grad_norm": 0.4824153367404781,
|
| 490 |
+
"learning_rate": 9.159078776924346e-05,
|
| 491 |
+
"loss": 0.8568,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.2153846153846154,
|
| 496 |
+
"grad_norm": 0.9408640158267115,
|
| 497 |
+
"learning_rate": 9.131193871579975e-05,
|
| 498 |
+
"loss": 1.0506,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.21846153846153846,
|
| 503 |
+
"grad_norm": 0.4967276109939823,
|
| 504 |
+
"learning_rate": 9.102898052541958e-05,
|
| 505 |
+
"loss": 0.9574,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.22153846153846155,
|
| 510 |
+
"grad_norm": 0.49940879804839516,
|
| 511 |
+
"learning_rate": 9.074194134284726e-05,
|
| 512 |
+
"loss": 0.9733,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.2246153846153846,
|
| 517 |
+
"grad_norm": 0.8907251117837572,
|
| 518 |
+
"learning_rate": 9.045084971874738e-05,
|
| 519 |
+
"loss": 0.9635,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.2276923076923077,
|
| 524 |
+
"grad_norm": 0.654651240080795,
|
| 525 |
+
"learning_rate": 9.015573460686509e-05,
|
| 526 |
+
"loss": 0.7272,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.23076923076923078,
|
| 531 |
+
"grad_norm": 0.44496890661205263,
|
| 532 |
+
"learning_rate": 8.985662536114613e-05,
|
| 533 |
+
"loss": 0.985,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.23384615384615384,
|
| 538 |
+
"grad_norm": 0.5099082561070101,
|
| 539 |
+
"learning_rate": 8.955355173281708e-05,
|
| 540 |
+
"loss": 1.0005,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.23692307692307693,
|
| 545 |
+
"grad_norm": 0.41580361052823284,
|
| 546 |
+
"learning_rate": 8.924654386742613e-05,
|
| 547 |
+
"loss": 0.9772,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.24,
|
| 552 |
+
"grad_norm": 0.45765794251900527,
|
| 553 |
+
"learning_rate": 8.89356323018447e-05,
|
| 554 |
+
"loss": 0.783,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.24307692307692308,
|
| 559 |
+
"grad_norm": 0.4375314198575698,
|
| 560 |
+
"learning_rate": 8.862084796122998e-05,
|
| 561 |
+
"loss": 0.8619,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.24615384615384617,
|
| 566 |
+
"grad_norm": 0.4226351704493707,
|
| 567 |
+
"learning_rate": 8.83022221559489e-05,
|
| 568 |
+
"loss": 0.9668,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.24923076923076923,
|
| 573 |
+
"grad_norm": 0.4492689965563046,
|
| 574 |
+
"learning_rate": 8.797978657846391e-05,
|
| 575 |
+
"loss": 1.1146,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.2523076923076923,
|
| 580 |
+
"grad_norm": 0.533848658747642,
|
| 581 |
+
"learning_rate": 8.765357330018056e-05,
|
| 582 |
+
"loss": 0.7778,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.2553846153846154,
|
| 587 |
+
"grad_norm": 0.4741983450227749,
|
| 588 |
+
"learning_rate": 8.732361476825752e-05,
|
| 589 |
+
"loss": 0.7988,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.25846153846153846,
|
| 594 |
+
"grad_norm": 0.40322949722802287,
|
| 595 |
+
"learning_rate": 8.69899438023792e-05,
|
| 596 |
+
"loss": 0.7871,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.26153846153846155,
|
| 601 |
+
"grad_norm": 0.5146184982151588,
|
| 602 |
+
"learning_rate": 8.665259359149132e-05,
|
| 603 |
+
"loss": 0.7989,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.26461538461538464,
|
| 608 |
+
"grad_norm": 0.5309234300488327,
|
| 609 |
+
"learning_rate": 8.631159769049965e-05,
|
| 610 |
+
"loss": 0.7692,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.2676923076923077,
|
| 615 |
+
"grad_norm": 0.5532377496823634,
|
| 616 |
+
"learning_rate": 8.596699001693255e-05,
|
| 617 |
+
"loss": 0.9662,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.27076923076923076,
|
| 622 |
+
"grad_norm": 0.4210717733049288,
|
| 623 |
+
"learning_rate": 8.561880484756725e-05,
|
| 624 |
+
"loss": 0.9426,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.27384615384615385,
|
| 629 |
+
"grad_norm": 0.45253551999315433,
|
| 630 |
+
"learning_rate": 8.526707681502044e-05,
|
| 631 |
+
"loss": 0.9861,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.27692307692307694,
|
| 636 |
+
"grad_norm": 0.48946665264650524,
|
| 637 |
+
"learning_rate": 8.491184090430364e-05,
|
| 638 |
+
"loss": 0.8552,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.28,
|
| 643 |
+
"grad_norm": 0.5782898707923746,
|
| 644 |
+
"learning_rate": 8.455313244934324e-05,
|
| 645 |
+
"loss": 1.0675,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.28307692307692306,
|
| 650 |
+
"grad_norm": 0.515882779092472,
|
| 651 |
+
"learning_rate": 8.419098712946601e-05,
|
| 652 |
+
"loss": 0.8706,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.28615384615384615,
|
| 657 |
+
"grad_norm": 1.1201682196837248,
|
| 658 |
+
"learning_rate": 8.382544096585027e-05,
|
| 659 |
+
"loss": 0.966,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.28923076923076924,
|
| 664 |
+
"grad_norm": 0.38827143982712936,
|
| 665 |
+
"learning_rate": 8.345653031794292e-05,
|
| 666 |
+
"loss": 0.7682,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.2923076923076923,
|
| 671 |
+
"grad_norm": 0.5783593896749222,
|
| 672 |
+
"learning_rate": 8.308429187984297e-05,
|
| 673 |
+
"loss": 1.0124,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.2953846153846154,
|
| 678 |
+
"grad_norm": 0.4348306382044402,
|
| 679 |
+
"learning_rate": 8.270876267665173e-05,
|
| 680 |
+
"loss": 0.8025,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.29846153846153844,
|
| 685 |
+
"grad_norm": 0.8193788357516452,
|
| 686 |
+
"learning_rate": 8.232998006078997e-05,
|
| 687 |
+
"loss": 1.1282,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.30153846153846153,
|
| 692 |
+
"grad_norm": 0.6040119737667439,
|
| 693 |
+
"learning_rate": 8.19479817082828e-05,
|
| 694 |
+
"loss": 1.0058,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.3046153846153846,
|
| 699 |
+
"grad_norm": 0.6286078754620025,
|
| 700 |
+
"learning_rate": 8.156280561501195e-05,
|
| 701 |
+
"loss": 0.8665,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.3076923076923077,
|
| 706 |
+
"grad_norm": 0.44823483533960423,
|
| 707 |
+
"learning_rate": 8.117449009293668e-05,
|
| 708 |
+
"loss": 0.9612,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.31076923076923074,
|
| 713 |
+
"grad_norm": 0.6257217839164982,
|
| 714 |
+
"learning_rate": 8.07830737662829e-05,
|
| 715 |
+
"loss": 0.9025,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.31384615384615383,
|
| 720 |
+
"grad_norm": 0.6257217839164982,
|
| 721 |
+
"learning_rate": 8.07830737662829e-05,
|
| 722 |
+
"loss": 0.9312,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.3169230769230769,
|
| 727 |
+
"grad_norm": 0.4678332385890188,
|
| 728 |
+
"learning_rate": 8.038859556770151e-05,
|
| 729 |
+
"loss": 0.8989,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.32,
|
| 734 |
+
"grad_norm": 0.3567873396014259,
|
| 735 |
+
"learning_rate": 7.999109473439569e-05,
|
| 736 |
+
"loss": 0.8148,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.3230769230769231,
|
| 741 |
+
"grad_norm": 0.5074459224922874,
|
| 742 |
+
"learning_rate": 7.959061080421839e-05,
|
| 743 |
+
"loss": 0.8349,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.3261538461538461,
|
| 748 |
+
"grad_norm": 0.4389813815494163,
|
| 749 |
+
"learning_rate": 7.91871836117395e-05,
|
| 750 |
+
"loss": 0.9388,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.3292307692307692,
|
| 755 |
+
"grad_norm": 0.4815313711740147,
|
| 756 |
+
"learning_rate": 7.878085328428369e-05,
|
| 757 |
+
"loss": 0.7953,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.3323076923076923,
|
| 762 |
+
"grad_norm": 0.5374417863961746,
|
| 763 |
+
"learning_rate": 7.83716602379391e-05,
|
| 764 |
+
"loss": 0.837,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.3353846153846154,
|
| 769 |
+
"grad_norm": 0.5154526228899968,
|
| 770 |
+
"learning_rate": 7.795964517353735e-05,
|
| 771 |
+
"loss": 0.9112,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.3384615384615385,
|
| 776 |
+
"grad_norm": 0.4647275999046091,
|
| 777 |
+
"learning_rate": 7.754484907260513e-05,
|
| 778 |
+
"loss": 0.9163,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.3415384615384615,
|
| 783 |
+
"grad_norm": 0.8055064145757932,
|
| 784 |
+
"learning_rate": 7.712731319328798e-05,
|
| 785 |
+
"loss": 1.0424,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.3446153846153846,
|
| 790 |
+
"grad_norm": 0.3811381028323214,
|
| 791 |
+
"learning_rate": 7.670707906624644e-05,
|
| 792 |
+
"loss": 0.9063,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.3476923076923077,
|
| 797 |
+
"grad_norm": 0.5225274040943507,
|
| 798 |
+
"learning_rate": 7.628418849052523e-05,
|
| 799 |
+
"loss": 0.9216,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.3507692307692308,
|
| 804 |
+
"grad_norm": 0.5166725748198604,
|
| 805 |
+
"learning_rate": 7.585868352939563e-05,
|
| 806 |
+
"loss": 0.8731,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.35384615384615387,
|
| 811 |
+
"grad_norm": 0.4487422116076619,
|
| 812 |
+
"learning_rate": 7.543060650617158e-05,
|
| 813 |
+
"loss": 0.9249,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.3569230769230769,
|
| 818 |
+
"grad_norm": 0.46067050849658847,
|
| 819 |
+
"learning_rate": 7.500000000000001e-05,
|
| 820 |
+
"loss": 0.8638,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.36,
|
| 825 |
+
"grad_norm": 0.43279895938302,
|
| 826 |
+
"learning_rate": 7.456690684162557e-05,
|
| 827 |
+
"loss": 0.8483,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.3630769230769231,
|
| 832 |
+
"grad_norm": 0.6978064247856391,
|
| 833 |
+
"learning_rate": 7.413137010913054e-05,
|
| 834 |
+
"loss": 0.9053,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.36615384615384616,
|
| 839 |
+
"grad_norm": 0.6046206751790981,
|
| 840 |
+
"learning_rate": 7.369343312364993e-05,
|
| 841 |
+
"loss": 0.8937,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.36923076923076925,
|
| 846 |
+
"grad_norm": 0.568252803025078,
|
| 847 |
+
"learning_rate": 7.325313944506254e-05,
|
| 848 |
+
"loss": 1.0268,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.3723076923076923,
|
| 853 |
+
"grad_norm": 0.5327656537522915,
|
| 854 |
+
"learning_rate": 7.281053286765815e-05,
|
| 855 |
+
"loss": 1.0301,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.37538461538461537,
|
| 860 |
+
"grad_norm": 0.6244085563322322,
|
| 861 |
+
"learning_rate": 7.236565741578163e-05,
|
| 862 |
+
"loss": 0.9318,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.37846153846153846,
|
| 867 |
+
"grad_norm": 0.41625084325760864,
|
| 868 |
+
"learning_rate": 7.191855733945387e-05,
|
| 869 |
+
"loss": 0.9161,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.38153846153846155,
|
| 874 |
+
"grad_norm": 0.45336384160290644,
|
| 875 |
+
"learning_rate": 7.146927710997047e-05,
|
| 876 |
+
"loss": 0.9162,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.38461538461538464,
|
| 881 |
+
"grad_norm": 0.6717136840417702,
|
| 882 |
+
"learning_rate": 7.101786141547828e-05,
|
| 883 |
+
"loss": 0.9223,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.38769230769230767,
|
| 888 |
+
"grad_norm": 0.5083485174444536,
|
| 889 |
+
"learning_rate": 7.056435515653059e-05,
|
| 890 |
+
"loss": 0.8678,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.39076923076923076,
|
| 895 |
+
"grad_norm": 0.8167318373747047,
|
| 896 |
+
"learning_rate": 7.010880344162088e-05,
|
| 897 |
+
"loss": 0.9603,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.39384615384615385,
|
| 902 |
+
"grad_norm": 0.4321651091452319,
|
| 903 |
+
"learning_rate": 6.965125158269619e-05,
|
| 904 |
+
"loss": 0.9513,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.39692307692307693,
|
| 909 |
+
"grad_norm": 0.4422794786711441,
|
| 910 |
+
"learning_rate": 6.919174509065004e-05,
|
| 911 |
+
"loss": 0.9852,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.4,
|
| 916 |
+
"grad_norm": 0.759984896203999,
|
| 917 |
+
"learning_rate": 6.873032967079561e-05,
|
| 918 |
+
"loss": 1.044,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.40307692307692305,
|
| 923 |
+
"grad_norm": 0.5572970536911653,
|
| 924 |
+
"learning_rate": 6.826705121831976e-05,
|
| 925 |
+
"loss": 0.9057,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.40615384615384614,
|
| 930 |
+
"grad_norm": 0.6367094403512291,
|
| 931 |
+
"learning_rate": 6.780195581371784e-05,
|
| 932 |
+
"loss": 1.132,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.40923076923076923,
|
| 937 |
+
"grad_norm": 0.7249793439884024,
|
| 938 |
+
"learning_rate": 6.733508971821036e-05,
|
| 939 |
+
"loss": 1.0408,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.4123076923076923,
|
| 944 |
+
"grad_norm": 0.4681876244130071,
|
| 945 |
+
"learning_rate": 6.686649936914152e-05,
|
| 946 |
+
"loss": 0.9028,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.4153846153846154,
|
| 951 |
+
"grad_norm": 0.42539268577316974,
|
| 952 |
+
"learning_rate": 6.639623137536023e-05,
|
| 953 |
+
"loss": 0.9138,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.41846153846153844,
|
| 958 |
+
"grad_norm": 0.7839535906682823,
|
| 959 |
+
"learning_rate": 6.592433251258423e-05,
|
| 960 |
+
"loss": 0.9533,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.42153846153846153,
|
| 965 |
+
"grad_norm": 0.5524487672166017,
|
| 966 |
+
"learning_rate": 6.545084971874738e-05,
|
| 967 |
+
"loss": 0.9236,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.4246153846153846,
|
| 972 |
+
"grad_norm": 0.4148339116790677,
|
| 973 |
+
"learning_rate": 6.497583008933097e-05,
|
| 974 |
+
"loss": 0.9762,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.4276923076923077,
|
| 979 |
+
"grad_norm": 0.5077561955937434,
|
| 980 |
+
"learning_rate": 6.449932087267932e-05,
|
| 981 |
+
"loss": 1.0127,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.4307692307692308,
|
| 986 |
+
"grad_norm": 0.6638932850156878,
|
| 987 |
+
"learning_rate": 6.402136946530014e-05,
|
| 988 |
+
"loss": 1.1408,
|
| 989 |
+
"step": 140
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.4338461538461538,
|
| 993 |
+
"grad_norm": 0.552185917044593,
|
| 994 |
+
"learning_rate": 6.354202340715026e-05,
|
| 995 |
+
"loss": 0.9799,
|
| 996 |
+
"step": 141
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.4369230769230769,
|
| 1000 |
+
"grad_norm": 0.43321619398203176,
|
| 1001 |
+
"learning_rate": 6.306133037690693e-05,
|
| 1002 |
+
"loss": 0.8325,
|
| 1003 |
+
"step": 142
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.44,
|
| 1007 |
+
"grad_norm": 0.6077052741940446,
|
| 1008 |
+
"learning_rate": 6.257933818722543e-05,
|
| 1009 |
+
"loss": 0.8752,
|
| 1010 |
+
"step": 143
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.4430769230769231,
|
| 1014 |
+
"grad_norm": 0.610128334739792,
|
| 1015 |
+
"learning_rate": 6.209609477998338e-05,
|
| 1016 |
+
"loss": 0.7847,
|
| 1017 |
+
"step": 144
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.4461538461538462,
|
| 1021 |
+
"grad_norm": 0.6777838892748965,
|
| 1022 |
+
"learning_rate": 6.161164822151213e-05,
|
| 1023 |
+
"loss": 0.9436,
|
| 1024 |
+
"step": 145
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.4492307692307692,
|
| 1028 |
+
"grad_norm": 0.468194610663323,
|
| 1029 |
+
"learning_rate": 6.112604669781572e-05,
|
| 1030 |
+
"loss": 0.7265,
|
| 1031 |
+
"step": 146
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.4523076923076923,
|
| 1035 |
+
"grad_norm": 0.4465674103163706,
|
| 1036 |
+
"learning_rate": 6.063933850977811e-05,
|
| 1037 |
+
"loss": 0.9067,
|
| 1038 |
+
"step": 147
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.4553846153846154,
|
| 1042 |
+
"grad_norm": 0.530040561634023,
|
| 1043 |
+
"learning_rate": 6.015157206835881e-05,
|
| 1044 |
+
"loss": 0.9874,
|
| 1045 |
+
"step": 148
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.4584615384615385,
|
| 1049 |
+
"grad_norm": 0.5766220416826322,
|
| 1050 |
+
"learning_rate": 5.9662795889777666e-05,
|
| 1051 |
+
"loss": 0.9203,
|
| 1052 |
+
"step": 149
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.46153846153846156,
|
| 1056 |
+
"grad_norm": 0.4244757365882824,
|
| 1057 |
+
"learning_rate": 5.917305859068912e-05,
|
| 1058 |
+
"loss": 0.9545,
|
| 1059 |
+
"step": 150
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.4646153846153846,
|
| 1063 |
+
"grad_norm": 0.518566329649453,
|
| 1064 |
+
"learning_rate": 5.868240888334653e-05,
|
| 1065 |
+
"loss": 0.9615,
|
| 1066 |
+
"step": 151
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.4676923076923077,
|
| 1070 |
+
"grad_norm": 0.4314672468912012,
|
| 1071 |
+
"learning_rate": 5.819089557075689e-05,
|
| 1072 |
+
"loss": 0.8471,
|
| 1073 |
+
"step": 152
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.4707692307692308,
|
| 1077 |
+
"grad_norm": 0.6911019341029342,
|
| 1078 |
+
"learning_rate": 5.7698567541826675e-05,
|
| 1079 |
+
"loss": 1.0476,
|
| 1080 |
+
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.47384615384615386,
|
| 1084 |
+
"grad_norm": 0.4857101241356872,
|
| 1085 |
+
"learning_rate": 5.7205473766499005e-05,
|
| 1086 |
+
"loss": 1.0066,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.47692307692307695,
|
| 1091 |
+
"grad_norm": 0.4857101241356872,
|
| 1092 |
+
"learning_rate": 5.7205473766499005e-05,
|
| 1093 |
+
"loss": 0.9934,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.48,
|
| 1098 |
+
"grad_norm": 0.559690182812592,
|
| 1099 |
+
"learning_rate": 5.6711663290882776e-05,
|
| 1100 |
+
"loss": 0.9719,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.48307692307692307,
|
| 1105 |
+
"grad_norm": 0.7188758350173023,
|
| 1106 |
+
"learning_rate": 5.621718523237427e-05,
|
| 1107 |
+
"loss": 1.0325,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.48615384615384616,
|
| 1112 |
+
"grad_norm": 0.4222820022079783,
|
| 1113 |
+
"learning_rate": 5.57220887747716e-05,
|
| 1114 |
+
"loss": 0.8563,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.48923076923076925,
|
| 1119 |
+
"grad_norm": 0.43013237215393946,
|
| 1120 |
+
"learning_rate": 5.522642316338268e-05,
|
| 1121 |
+
"loss": 0.9479,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.49230769230769234,
|
| 1126 |
+
"grad_norm": 0.45076662279787416,
|
| 1127 |
+
"learning_rate": 5.473023770012686e-05,
|
| 1128 |
+
"loss": 0.894,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.49538461538461537,
|
| 1133 |
+
"grad_norm": 0.4294610826909239,
|
| 1134 |
+
"learning_rate": 5.4233581738631165e-05,
|
| 1135 |
+
"loss": 0.8667,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.49846153846153846,
|
| 1140 |
+
"grad_norm": 0.42688890694566145,
|
| 1141 |
+
"learning_rate": 5.373650467932122e-05,
|
| 1142 |
+
"loss": 0.8839,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.5015384615384615,
|
| 1147 |
+
"grad_norm": 0.6114025248557045,
|
| 1148 |
+
"learning_rate": 5.323905596450759e-05,
|
| 1149 |
+
"loss": 0.9234,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.5046153846153846,
|
| 1154 |
+
"grad_norm": 0.4627192916088213,
|
| 1155 |
+
"learning_rate": 5.274128507346801e-05,
|
| 1156 |
+
"loss": 0.8776,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.5076923076923077,
|
| 1161 |
+
"grad_norm": 0.5666314183948891,
|
| 1162 |
+
"learning_rate": 5.2243241517525754e-05,
|
| 1163 |
+
"loss": 0.8748,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.5107692307692308,
|
| 1168 |
+
"grad_norm": 0.40552689428512695,
|
| 1169 |
+
"learning_rate": 5.174497483512506e-05,
|
| 1170 |
+
"loss": 0.9743,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.5138461538461538,
|
| 1175 |
+
"grad_norm": 0.7915356460560259,
|
| 1176 |
+
"learning_rate": 5.124653458690365e-05,
|
| 1177 |
+
"loss": 0.7467,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.5169230769230769,
|
| 1182 |
+
"grad_norm": 0.6718245334626134,
|
| 1183 |
+
"learning_rate": 5.074797035076319e-05,
|
| 1184 |
+
"loss": 1.0196,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.52,
|
| 1189 |
+
"grad_norm": 0.624678937018929,
|
| 1190 |
+
"learning_rate": 5.024933171693791e-05,
|
| 1191 |
+
"loss": 0.7417,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.5230769230769231,
|
| 1196 |
+
"grad_norm": 0.5055426294677736,
|
| 1197 |
+
"learning_rate": 4.9750668283062104e-05,
|
| 1198 |
+
"loss": 0.9155,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.5261538461538462,
|
| 1203 |
+
"grad_norm": 0.7853307489896678,
|
| 1204 |
+
"learning_rate": 4.925202964923683e-05,
|
| 1205 |
+
"loss": 1.1095,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.5292307692307693,
|
| 1210 |
+
"grad_norm": 0.5192134531626986,
|
| 1211 |
+
"learning_rate": 4.875346541309637e-05,
|
| 1212 |
+
"loss": 0.8172,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.5323076923076923,
|
| 1217 |
+
"grad_norm": 0.6498537876872281,
|
| 1218 |
+
"learning_rate": 4.825502516487497e-05,
|
| 1219 |
+
"loss": 0.9261,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.5353846153846153,
|
| 1224 |
+
"grad_norm": 0.5183741661218348,
|
| 1225 |
+
"learning_rate": 4.775675848247427e-05,
|
| 1226 |
+
"loss": 0.8839,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.5384615384615384,
|
| 1231 |
+
"grad_norm": 0.9055402279660434,
|
| 1232 |
+
"learning_rate": 4.725871492653199e-05,
|
| 1233 |
+
"loss": 1.091,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.5415384615384615,
|
| 1238 |
+
"grad_norm": 1.7646663739263517,
|
| 1239 |
+
"learning_rate": 4.6760944035492404e-05,
|
| 1240 |
+
"loss": 1.0395,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.5446153846153846,
|
| 1245 |
+
"grad_norm": 0.41046559167871544,
|
| 1246 |
+
"learning_rate": 4.626349532067879e-05,
|
| 1247 |
+
"loss": 0.9266,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.5476923076923077,
|
| 1252 |
+
"grad_norm": 0.4973648976717774,
|
| 1253 |
+
"learning_rate": 4.576641826136884e-05,
|
| 1254 |
+
"loss": 1.0799,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.5507692307692308,
|
| 1259 |
+
"grad_norm": 0.43183628490229176,
|
| 1260 |
+
"learning_rate": 4.526976229987315e-05,
|
| 1261 |
+
"loss": 0.9763,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.5538461538461539,
|
| 1266 |
+
"grad_norm": 0.4447730696794788,
|
| 1267 |
+
"learning_rate": 4.477357683661734e-05,
|
| 1268 |
+
"loss": 0.973,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.556923076923077,
|
| 1273 |
+
"grad_norm": 0.40067033993711776,
|
| 1274 |
+
"learning_rate": 4.4277911225228414e-05,
|
| 1275 |
+
"loss": 0.8514,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.56,
|
| 1280 |
+
"grad_norm": 0.6866382906604235,
|
| 1281 |
+
"learning_rate": 4.378281476762576e-05,
|
| 1282 |
+
"loss": 0.9952,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.563076923076923,
|
| 1287 |
+
"grad_norm": 0.3842061308151513,
|
| 1288 |
+
"learning_rate": 4.328833670911724e-05,
|
| 1289 |
+
"loss": 0.707,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.5661538461538461,
|
| 1294 |
+
"grad_norm": 0.7554681519781477,
|
| 1295 |
+
"learning_rate": 4.2794526233501006e-05,
|
| 1296 |
+
"loss": 0.893,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.5692307692307692,
|
| 1301 |
+
"grad_norm": 0.4511372575739064,
|
| 1302 |
+
"learning_rate": 4.230143245817332e-05,
|
| 1303 |
+
"loss": 0.7752,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.5723076923076923,
|
| 1308 |
+
"grad_norm": 0.4991324743083531,
|
| 1309 |
+
"learning_rate": 4.180910442924312e-05,
|
| 1310 |
+
"loss": 0.9633,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.5753846153846154,
|
| 1315 |
+
"grad_norm": 0.4483075392830615,
|
| 1316 |
+
"learning_rate": 4.131759111665349e-05,
|
| 1317 |
+
"loss": 0.8787,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.5784615384615385,
|
| 1322 |
+
"grad_norm": 0.5292372567003025,
|
| 1323 |
+
"learning_rate": 4.082694140931089e-05,
|
| 1324 |
+
"loss": 0.845,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.5815384615384616,
|
| 1329 |
+
"grad_norm": 0.47885018785514416,
|
| 1330 |
+
"learning_rate": 4.0337204110222346e-05,
|
| 1331 |
+
"loss": 0.9699,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.5846153846153846,
|
| 1336 |
+
"grad_norm": 0.47117502086214685,
|
| 1337 |
+
"learning_rate": 3.98484279316412e-05,
|
| 1338 |
+
"loss": 0.8229,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.5876923076923077,
|
| 1343 |
+
"grad_norm": 0.47680717952456125,
|
| 1344 |
+
"learning_rate": 3.936066149022191e-05,
|
| 1345 |
+
"loss": 0.8792,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.5907692307692308,
|
| 1350 |
+
"grad_norm": 0.5061286925031491,
|
| 1351 |
+
"learning_rate": 3.887395330218429e-05,
|
| 1352 |
+
"loss": 0.7672,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.5938461538461538,
|
| 1357 |
+
"grad_norm": 0.7864978132073881,
|
| 1358 |
+
"learning_rate": 3.838835177848788e-05,
|
| 1359 |
+
"loss": 0.995,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.5969230769230769,
|
| 1364 |
+
"grad_norm": 0.5459143405404062,
|
| 1365 |
+
"learning_rate": 3.790390522001662e-05,
|
| 1366 |
+
"loss": 0.7337,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.6,
|
| 1371 |
+
"grad_norm": 0.4064607944328481,
|
| 1372 |
+
"learning_rate": 3.742066181277458e-05,
|
| 1373 |
+
"loss": 0.8403,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.6030769230769231,
|
| 1378 |
+
"grad_norm": 0.8416821105775466,
|
| 1379 |
+
"learning_rate": 3.6938669623093084e-05,
|
| 1380 |
+
"loss": 1.0778,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.6061538461538462,
|
| 1385 |
+
"grad_norm": 0.4952747141069206,
|
| 1386 |
+
"learning_rate": 3.6457976592849754e-05,
|
| 1387 |
+
"loss": 0.8227,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.6092307692307692,
|
| 1392 |
+
"grad_norm": 0.5097083293162382,
|
| 1393 |
+
"learning_rate": 3.597863053469987e-05,
|
| 1394 |
+
"loss": 0.8839,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.6123076923076923,
|
| 1399 |
+
"grad_norm": 0.479661103196798,
|
| 1400 |
+
"learning_rate": 3.550067912732069e-05,
|
| 1401 |
+
"loss": 0.9492,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.6153846153846154,
|
| 1406 |
+
"grad_norm": 1.2914394484341967,
|
| 1407 |
+
"learning_rate": 3.502416991066904e-05,
|
| 1408 |
+
"loss": 0.8155,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.6184615384615385,
|
| 1413 |
+
"grad_norm": 0.6425998268512141,
|
| 1414 |
+
"learning_rate": 3.4549150281252636e-05,
|
| 1415 |
+
"loss": 0.9929,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.6215384615384615,
|
| 1420 |
+
"grad_norm": 0.44723769018665716,
|
| 1421 |
+
"learning_rate": 3.4075667487415785e-05,
|
| 1422 |
+
"loss": 1.0377,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.6246153846153846,
|
| 1427 |
+
"grad_norm": 0.9049713782179101,
|
| 1428 |
+
"learning_rate": 3.360376862463979e-05,
|
| 1429 |
+
"loss": 0.989,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.6276923076923077,
|
| 1434 |
+
"grad_norm": 0.4329976152731377,
|
| 1435 |
+
"learning_rate": 3.313350063085851e-05,
|
| 1436 |
+
"loss": 0.9683,
|
| 1437 |
+
"step": 204
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.6307692307692307,
|
| 1441 |
+
"grad_norm": 0.7867934434433727,
|
| 1442 |
+
"learning_rate": 3.266491028178964e-05,
|
| 1443 |
+
"loss": 0.9286,
|
| 1444 |
+
"step": 205
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.6338461538461538,
|
| 1448 |
+
"grad_norm": 0.5467059065604649,
|
| 1449 |
+
"learning_rate": 3.219804418628216e-05,
|
| 1450 |
+
"loss": 0.9589,
|
| 1451 |
+
"step": 206
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.6369230769230769,
|
| 1455 |
+
"grad_norm": 0.572922134191068,
|
| 1456 |
+
"learning_rate": 3.173294878168025e-05,
|
| 1457 |
+
"loss": 1.035,
|
| 1458 |
+
"step": 207
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.64,
|
| 1462 |
+
"grad_norm": 0.4463963465221491,
|
| 1463 |
+
"learning_rate": 3.12696703292044e-05,
|
| 1464 |
+
"loss": 0.8026,
|
| 1465 |
+
"step": 208
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.6430769230769231,
|
| 1469 |
+
"grad_norm": 0.46576698409775236,
|
| 1470 |
+
"learning_rate": 3.080825490934999e-05,
|
| 1471 |
+
"loss": 0.9075,
|
| 1472 |
+
"step": 209
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.6461538461538462,
|
| 1476 |
+
"grad_norm": 0.45041787919351106,
|
| 1477 |
+
"learning_rate": 3.0348748417303823e-05,
|
| 1478 |
+
"loss": 0.8102,
|
| 1479 |
+
"step": 210
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.6492307692307693,
|
| 1483 |
+
"grad_norm": 0.38614038078039586,
|
| 1484 |
+
"learning_rate": 2.989119655837913e-05,
|
| 1485 |
+
"loss": 0.7971,
|
| 1486 |
+
"step": 211
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.6523076923076923,
|
| 1490 |
+
"grad_norm": 0.7439118868320627,
|
| 1491 |
+
"learning_rate": 2.9435644843469436e-05,
|
| 1492 |
+
"loss": 0.7787,
|
| 1493 |
+
"step": 212
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.6553846153846153,
|
| 1497 |
+
"grad_norm": 0.45297022735174813,
|
| 1498 |
+
"learning_rate": 2.8982138584521735e-05,
|
| 1499 |
+
"loss": 0.8832,
|
| 1500 |
+
"step": 213
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.6584615384615384,
|
| 1504 |
+
"grad_norm": 0.9273743586343851,
|
| 1505 |
+
"learning_rate": 2.8530722890029537e-05,
|
| 1506 |
+
"loss": 1.0353,
|
| 1507 |
+
"step": 214
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.6615384615384615,
|
| 1511 |
+
"grad_norm": 0.8591474008093516,
|
| 1512 |
+
"learning_rate": 2.8081442660546125e-05,
|
| 1513 |
+
"loss": 1.0776,
|
| 1514 |
+
"step": 215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.6646153846153846,
|
| 1518 |
+
"grad_norm": 0.5439265538156528,
|
| 1519 |
+
"learning_rate": 2.7634342584218365e-05,
|
| 1520 |
+
"loss": 0.8459,
|
| 1521 |
+
"step": 216
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.6676923076923077,
|
| 1525 |
+
"grad_norm": 0.5344112954067011,
|
| 1526 |
+
"learning_rate": 2.718946713234185e-05,
|
| 1527 |
+
"loss": 0.9952,
|
| 1528 |
+
"step": 217
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.6707692307692308,
|
| 1532 |
+
"grad_norm": 0.5055789976636205,
|
| 1533 |
+
"learning_rate": 2.674686055493748e-05,
|
| 1534 |
+
"loss": 0.7474,
|
| 1535 |
+
"step": 218
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.6738461538461539,
|
| 1539 |
+
"grad_norm": 0.4794508620037753,
|
| 1540 |
+
"learning_rate": 2.630656687635007e-05,
|
| 1541 |
+
"loss": 1.0326,
|
| 1542 |
+
"step": 219
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.676923076923077,
|
| 1546 |
+
"grad_norm": 0.797598612978555,
|
| 1547 |
+
"learning_rate": 2.5868629890869468e-05,
|
| 1548 |
+
"loss": 1.1058,
|
| 1549 |
+
"step": 220
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.68,
|
| 1553 |
+
"grad_norm": 0.7072837694866513,
|
| 1554 |
+
"learning_rate": 2.543309315837444e-05,
|
| 1555 |
+
"loss": 1.1483,
|
| 1556 |
+
"step": 221
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.683076923076923,
|
| 1560 |
+
"grad_norm": 0.4808247459318681,
|
| 1561 |
+
"learning_rate": 2.500000000000001e-05,
|
| 1562 |
+
"loss": 0.9573,
|
| 1563 |
+
"step": 222
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.6861538461538461,
|
| 1567 |
+
"grad_norm": 0.5515272326814473,
|
| 1568 |
+
"learning_rate": 2.456939349382843e-05,
|
| 1569 |
+
"loss": 0.9657,
|
| 1570 |
+
"step": 223
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.6892307692307692,
|
| 1574 |
+
"grad_norm": 1.091841441341856,
|
| 1575 |
+
"learning_rate": 2.4141316470604362e-05,
|
| 1576 |
+
"loss": 1.2441,
|
| 1577 |
+
"step": 224
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.6923076923076923,
|
| 1581 |
+
"grad_norm": 0.46774331028612076,
|
| 1582 |
+
"learning_rate": 2.371581150947476e-05,
|
| 1583 |
+
"loss": 0.9983,
|
| 1584 |
+
"step": 225
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.6953846153846154,
|
| 1588 |
+
"grad_norm": 0.49878974527974484,
|
| 1589 |
+
"learning_rate": 2.3292920933753566e-05,
|
| 1590 |
+
"loss": 0.9933,
|
| 1591 |
+
"step": 226
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.6984615384615385,
|
| 1595 |
+
"grad_norm": 0.5714470688408128,
|
| 1596 |
+
"learning_rate": 2.2872686806712035e-05,
|
| 1597 |
+
"loss": 1.0289,
|
| 1598 |
+
"step": 227
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.7015384615384616,
|
| 1602 |
+
"grad_norm": 0.5196542331369163,
|
| 1603 |
+
"learning_rate": 2.245515092739488e-05,
|
| 1604 |
+
"loss": 0.984,
|
| 1605 |
+
"step": 228
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.7046153846153846,
|
| 1609 |
+
"grad_norm": 0.41263259638261746,
|
| 1610 |
+
"learning_rate": 2.2040354826462668e-05,
|
| 1611 |
+
"loss": 0.8709,
|
| 1612 |
+
"step": 229
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.7076923076923077,
|
| 1616 |
+
"grad_norm": 0.8438835684897247,
|
| 1617 |
+
"learning_rate": 2.1628339762060917e-05,
|
| 1618 |
+
"loss": 0.8316,
|
| 1619 |
+
"step": 230
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.7107692307692308,
|
| 1623 |
+
"grad_norm": 0.48397501380570757,
|
| 1624 |
+
"learning_rate": 2.1219146715716332e-05,
|
| 1625 |
+
"loss": 0.9991,
|
| 1626 |
+
"step": 231
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.7138461538461538,
|
| 1630 |
+
"grad_norm": 0.6670132545571469,
|
| 1631 |
+
"learning_rate": 2.0812816388260518e-05,
|
| 1632 |
+
"loss": 0.8505,
|
| 1633 |
+
"step": 232
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.7169230769230769,
|
| 1637 |
+
"grad_norm": 0.42590222370138053,
|
| 1638 |
+
"learning_rate": 2.0409389195781625e-05,
|
| 1639 |
+
"loss": 0.9524,
|
| 1640 |
+
"step": 233
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.72,
|
| 1644 |
+
"grad_norm": 0.5260835406944124,
|
| 1645 |
+
"learning_rate": 2.0008905265604316e-05,
|
| 1646 |
+
"loss": 0.8622,
|
| 1647 |
+
"step": 234
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.7230769230769231,
|
| 1651 |
+
"grad_norm": 0.5496980878148783,
|
| 1652 |
+
"learning_rate": 1.9611404432298502e-05,
|
| 1653 |
+
"loss": 0.9608,
|
| 1654 |
+
"step": 235
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.7261538461538461,
|
| 1658 |
+
"grad_norm": 0.542905084359803,
|
| 1659 |
+
"learning_rate": 1.9216926233717085e-05,
|
| 1660 |
+
"loss": 1.0753,
|
| 1661 |
+
"step": 236
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.7292307692307692,
|
| 1665 |
+
"grad_norm": 0.4342140067844084,
|
| 1666 |
+
"learning_rate": 1.8825509907063327e-05,
|
| 1667 |
+
"loss": 0.8647,
|
| 1668 |
+
"step": 237
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.7323076923076923,
|
| 1672 |
+
"grad_norm": 0.4622308195840851,
|
| 1673 |
+
"learning_rate": 1.843719438498806e-05,
|
| 1674 |
+
"loss": 0.8809,
|
| 1675 |
+
"step": 238
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.7353846153846154,
|
| 1679 |
+
"grad_norm": 0.4530626869383855,
|
| 1680 |
+
"learning_rate": 1.8052018291717214e-05,
|
| 1681 |
+
"loss": 0.9253,
|
| 1682 |
+
"step": 239
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.7384615384615385,
|
| 1686 |
+
"grad_norm": 0.42127528108498424,
|
| 1687 |
+
"learning_rate": 1.7670019939210024e-05,
|
| 1688 |
+
"loss": 1.0715,
|
| 1689 |
+
"step": 240
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.7415384615384616,
|
| 1693 |
+
"grad_norm": 0.6427138843285375,
|
| 1694 |
+
"learning_rate": 1.7291237323348284e-05,
|
| 1695 |
+
"loss": 0.9315,
|
| 1696 |
+
"step": 241
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.7446153846153846,
|
| 1700 |
+
"grad_norm": 0.459875858863969,
|
| 1701 |
+
"learning_rate": 1.691570812015704e-05,
|
| 1702 |
+
"loss": 0.9504,
|
| 1703 |
+
"step": 242
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.7476923076923077,
|
| 1707 |
+
"grad_norm": 0.4613927768242604,
|
| 1708 |
+
"learning_rate": 1.6543469682057106e-05,
|
| 1709 |
+
"loss": 0.9968,
|
| 1710 |
+
"step": 243
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.7507692307692307,
|
| 1714 |
+
"grad_norm": 0.45000507983226773,
|
| 1715 |
+
"learning_rate": 1.617455903414974e-05,
|
| 1716 |
+
"loss": 0.805,
|
| 1717 |
+
"step": 244
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.7538461538461538,
|
| 1721 |
+
"grad_norm": 0.5470377495981378,
|
| 1722 |
+
"learning_rate": 1.5809012870533995e-05,
|
| 1723 |
+
"loss": 0.9431,
|
| 1724 |
+
"step": 245
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.7569230769230769,
|
| 1728 |
+
"grad_norm": 0.7312522415703255,
|
| 1729 |
+
"learning_rate": 1.544686755065677e-05,
|
| 1730 |
+
"loss": 1.1147,
|
| 1731 |
+
"step": 246
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.76,
|
| 1735 |
+
"grad_norm": 0.5610220941428191,
|
| 1736 |
+
"learning_rate": 1.5088159095696363e-05,
|
| 1737 |
+
"loss": 0.8445,
|
| 1738 |
+
"step": 247
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.7630769230769231,
|
| 1742 |
+
"grad_norm": 0.4555730253772095,
|
| 1743 |
+
"learning_rate": 1.4732923184979563e-05,
|
| 1744 |
+
"loss": 0.937,
|
| 1745 |
+
"step": 248
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.7661538461538462,
|
| 1749 |
+
"grad_norm": 0.41249117694270737,
|
| 1750 |
+
"learning_rate": 1.438119515243277e-05,
|
| 1751 |
+
"loss": 0.825,
|
| 1752 |
+
"step": 249
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.7692307692307693,
|
| 1756 |
+
"grad_norm": 0.6037584778170366,
|
| 1757 |
+
"learning_rate": 1.4033009983067452e-05,
|
| 1758 |
+
"loss": 0.9916,
|
| 1759 |
+
"step": 250
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.7723076923076924,
|
| 1763 |
+
"grad_norm": 0.5186835760091726,
|
| 1764 |
+
"learning_rate": 1.3688402309500353e-05,
|
| 1765 |
+
"loss": 0.8961,
|
| 1766 |
+
"step": 251
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.7753846153846153,
|
| 1770 |
+
"grad_norm": 0.9426774142222885,
|
| 1771 |
+
"learning_rate": 1.3347406408508695e-05,
|
| 1772 |
+
"loss": 0.879,
|
| 1773 |
+
"step": 252
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.7784615384615384,
|
| 1777 |
+
"grad_norm": 0.5038467116216354,
|
| 1778 |
+
"learning_rate": 1.3010056197620812e-05,
|
| 1779 |
+
"loss": 1.0214,
|
| 1780 |
+
"step": 253
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.7815384615384615,
|
| 1784 |
+
"grad_norm": 0.4264416350547513,
|
| 1785 |
+
"learning_rate": 1.2676385231742494e-05,
|
| 1786 |
+
"loss": 0.9418,
|
| 1787 |
+
"step": 254
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.7846153846153846,
|
| 1791 |
+
"grad_norm": 0.7616436842258425,
|
| 1792 |
+
"learning_rate": 1.2346426699819458e-05,
|
| 1793 |
+
"loss": 1.0602,
|
| 1794 |
+
"step": 255
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.7876923076923077,
|
| 1798 |
+
"grad_norm": 0.41242091439316825,
|
| 1799 |
+
"learning_rate": 1.2020213421536103e-05,
|
| 1800 |
+
"loss": 0.8267,
|
| 1801 |
+
"step": 256
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.7907692307692308,
|
| 1805 |
+
"grad_norm": 0.524144821209285,
|
| 1806 |
+
"learning_rate": 1.1697777844051105e-05,
|
| 1807 |
+
"loss": 0.9295,
|
| 1808 |
+
"step": 257
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.7938461538461539,
|
| 1812 |
+
"grad_norm": 0.7799732613380811,
|
| 1813 |
+
"learning_rate": 1.137915203877003e-05,
|
| 1814 |
+
"loss": 1.0236,
|
| 1815 |
+
"step": 258
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.796923076923077,
|
| 1819 |
+
"grad_norm": 0.7587523738633466,
|
| 1820 |
+
"learning_rate": 1.1064367698155304e-05,
|
| 1821 |
+
"loss": 1.0791,
|
| 1822 |
+
"step": 259
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.8,
|
| 1826 |
+
"grad_norm": 0.4823020598514238,
|
| 1827 |
+
"learning_rate": 1.0753456132573886e-05,
|
| 1828 |
+
"loss": 0.7288,
|
| 1829 |
+
"step": 260
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.803076923076923,
|
| 1833 |
+
"grad_norm": 0.5188786763523745,
|
| 1834 |
+
"learning_rate": 1.0446448267182952e-05,
|
| 1835 |
+
"loss": 1.0829,
|
| 1836 |
+
"step": 261
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.8061538461538461,
|
| 1840 |
+
"grad_norm": 0.6321674676279638,
|
| 1841 |
+
"learning_rate": 1.0143374638853891e-05,
|
| 1842 |
+
"loss": 0.7585,
|
| 1843 |
+
"step": 262
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.8092307692307692,
|
| 1847 |
+
"grad_norm": 0.5139713117623734,
|
| 1848 |
+
"learning_rate": 9.844265393134927e-06,
|
| 1849 |
+
"loss": 0.7744,
|
| 1850 |
+
"step": 263
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.8123076923076923,
|
| 1854 |
+
"grad_norm": 0.6350327475481775,
|
| 1855 |
+
"learning_rate": 9.549150281252633e-06,
|
| 1856 |
+
"loss": 0.9106,
|
| 1857 |
+
"step": 264
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.8153846153846154,
|
| 1861 |
+
"grad_norm": 0.7468855064128749,
|
| 1862 |
+
"learning_rate": 9.258058657152762e-06,
|
| 1863 |
+
"loss": 1.0186,
|
| 1864 |
+
"step": 265
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.8184615384615385,
|
| 1868 |
+
"grad_norm": 0.5135460319092262,
|
| 1869 |
+
"learning_rate": 8.971019474580427e-06,
|
| 1870 |
+
"loss": 0.9886,
|
| 1871 |
+
"step": 266
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.8215384615384616,
|
| 1875 |
+
"grad_norm": 0.4152276730658682,
|
| 1876 |
+
"learning_rate": 8.688061284200266e-06,
|
| 1877 |
+
"loss": 0.7488,
|
| 1878 |
+
"step": 267
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.8246153846153846,
|
| 1882 |
+
"grad_norm": 0.4602002647755605,
|
| 1883 |
+
"learning_rate": 8.409212230756564e-06,
|
| 1884 |
+
"loss": 0.8909,
|
| 1885 |
+
"step": 268
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.8276923076923077,
|
| 1889 |
+
"grad_norm": 1.0751149249264285,
|
| 1890 |
+
"learning_rate": 8.134500050273841e-06,
|
| 1891 |
+
"loss": 1.3033,
|
| 1892 |
+
"step": 269
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.8307692307692308,
|
| 1896 |
+
"grad_norm": 0.5507844454218935,
|
| 1897 |
+
"learning_rate": 7.863952067298042e-06,
|
| 1898 |
+
"loss": 0.7655,
|
| 1899 |
+
"step": 270
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.8338461538461538,
|
| 1903 |
+
"grad_norm": 0.36202036336397553,
|
| 1904 |
+
"learning_rate": 7.597595192178702e-06,
|
| 1905 |
+
"loss": 0.8784,
|
| 1906 |
+
"step": 271
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.8369230769230769,
|
| 1910 |
+
"grad_norm": 0.9152443501429699,
|
| 1911 |
+
"learning_rate": 7.33545591839222e-06,
|
| 1912 |
+
"loss": 1.0999,
|
| 1913 |
+
"step": 272
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.84,
|
| 1917 |
+
"grad_norm": 0.4782233006424708,
|
| 1918 |
+
"learning_rate": 7.077560319906695e-06,
|
| 1919 |
+
"loss": 0.8817,
|
| 1920 |
+
"step": 273
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.8430769230769231,
|
| 1924 |
+
"grad_norm": 0.5928489194497177,
|
| 1925 |
+
"learning_rate": 6.82393404858846e-06,
|
| 1926 |
+
"loss": 0.9694,
|
| 1927 |
+
"step": 274
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.8461538461538461,
|
| 1931 |
+
"grad_norm": 0.7906221168589845,
|
| 1932 |
+
"learning_rate": 6.574602331650559e-06,
|
| 1933 |
+
"loss": 1.0836,
|
| 1934 |
+
"step": 275
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.8492307692307692,
|
| 1938 |
+
"grad_norm": 0.5678939250281202,
|
| 1939 |
+
"learning_rate": 6.329589969143518e-06,
|
| 1940 |
+
"loss": 0.9083,
|
| 1941 |
+
"step": 276
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.8523076923076923,
|
| 1945 |
+
"grad_norm": 0.7072665793146278,
|
| 1946 |
+
"learning_rate": 6.088921331488568e-06,
|
| 1947 |
+
"loss": 0.8394,
|
| 1948 |
+
"step": 277
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.8553846153846154,
|
| 1952 |
+
"grad_norm": 0.5646602614466468,
|
| 1953 |
+
"learning_rate": 5.852620357053651e-06,
|
| 1954 |
+
"loss": 0.7478,
|
| 1955 |
+
"step": 278
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.8584615384615385,
|
| 1959 |
+
"grad_norm": 0.6051555479347088,
|
| 1960 |
+
"learning_rate": 5.620710549772295e-06,
|
| 1961 |
+
"loss": 0.9049,
|
| 1962 |
+
"step": 279
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.8615384615384616,
|
| 1966 |
+
"grad_norm": 0.4146402590200568,
|
| 1967 |
+
"learning_rate": 5.393214976805833e-06,
|
| 1968 |
+
"loss": 0.9216,
|
| 1969 |
+
"step": 280
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.8646153846153846,
|
| 1973 |
+
"grad_norm": 0.4146402590200568,
|
| 1974 |
+
"learning_rate": 5.393214976805833e-06,
|
| 1975 |
+
"loss": 0.9953,
|
| 1976 |
+
"step": 281
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.8676923076923077,
|
| 1980 |
+
"grad_norm": 0.6929995573516754,
|
| 1981 |
+
"learning_rate": 5.17015626624896e-06,
|
| 1982 |
+
"loss": 0.9771,
|
| 1983 |
+
"step": 282
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.8707692307692307,
|
| 1987 |
+
"grad_norm": 0.4258690273862394,
|
| 1988 |
+
"learning_rate": 4.951556604879048e-06,
|
| 1989 |
+
"loss": 0.8803,
|
| 1990 |
+
"step": 283
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.8738461538461538,
|
| 1994 |
+
"grad_norm": 0.4837540760871192,
|
| 1995 |
+
"learning_rate": 4.737437735949263e-06,
|
| 1996 |
+
"loss": 0.8875,
|
| 1997 |
+
"step": 284
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.8769230769230769,
|
| 2001 |
+
"grad_norm": 0.6412083356298082,
|
| 2002 |
+
"learning_rate": 4.527820957025891e-06,
|
| 2003 |
+
"loss": 0.9273,
|
| 2004 |
+
"step": 285
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.88,
|
| 2008 |
+
"grad_norm": 0.6937402841005177,
|
| 2009 |
+
"learning_rate": 4.322727117869951e-06,
|
| 2010 |
+
"loss": 1.1679,
|
| 2011 |
+
"step": 286
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.8830769230769231,
|
| 2015 |
+
"grad_norm": 0.48219084873311563,
|
| 2016 |
+
"learning_rate": 4.122176618363305e-06,
|
| 2017 |
+
"loss": 0.9607,
|
| 2018 |
+
"step": 287
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.8861538461538462,
|
| 2022 |
+
"grad_norm": 0.4661930023551732,
|
| 2023 |
+
"learning_rate": 3.9261894064796135e-06,
|
| 2024 |
+
"loss": 0.9608,
|
| 2025 |
+
"step": 288
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.8892307692307693,
|
| 2029 |
+
"grad_norm": 0.5356323172537888,
|
| 2030 |
+
"learning_rate": 3.734784976300165e-06,
|
| 2031 |
+
"loss": 0.8979,
|
| 2032 |
+
"step": 289
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.8923076923076924,
|
| 2036 |
+
"grad_norm": 0.4274291511334043,
|
| 2037 |
+
"learning_rate": 3.54798236607487e-06,
|
| 2038 |
+
"loss": 0.7525,
|
| 2039 |
+
"step": 290
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.8953846153846153,
|
| 2043 |
+
"grad_norm": 0.531050081114398,
|
| 2044 |
+
"learning_rate": 3.365800156328619e-06,
|
| 2045 |
+
"loss": 1.0411,
|
| 2046 |
+
"step": 291
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.8984615384615384,
|
| 2050 |
+
"grad_norm": 0.4962045602466536,
|
| 2051 |
+
"learning_rate": 3.18825646801314e-06,
|
| 2052 |
+
"loss": 1.0028,
|
| 2053 |
+
"step": 292
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.9015384615384615,
|
| 2057 |
+
"grad_norm": 0.5167475813822029,
|
| 2058 |
+
"learning_rate": 3.0153689607045845e-06,
|
| 2059 |
+
"loss": 1.0771,
|
| 2060 |
+
"step": 293
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.9046153846153846,
|
| 2064 |
+
"grad_norm": 0.4445981508931031,
|
| 2065 |
+
"learning_rate": 2.847154830846971e-06,
|
| 2066 |
+
"loss": 0.959,
|
| 2067 |
+
"step": 294
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.9076923076923077,
|
| 2071 |
+
"grad_norm": 0.7398667070938753,
|
| 2072 |
+
"learning_rate": 2.6836308100417873e-06,
|
| 2073 |
+
"loss": 1.0592,
|
| 2074 |
+
"step": 295
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.9107692307692308,
|
| 2078 |
+
"grad_norm": 0.6264192493181824,
|
| 2079 |
+
"learning_rate": 2.5248131633836826e-06,
|
| 2080 |
+
"loss": 0.8678,
|
| 2081 |
+
"step": 296
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.9138461538461539,
|
| 2085 |
+
"grad_norm": 0.5988988069518453,
|
| 2086 |
+
"learning_rate": 2.3707176878426884e-06,
|
| 2087 |
+
"loss": 0.8387,
|
| 2088 |
+
"step": 297
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.916923076923077,
|
| 2092 |
+
"grad_norm": 0.8356933837876696,
|
| 2093 |
+
"learning_rate": 2.221359710692961e-06,
|
| 2094 |
+
"loss": 0.8304,
|
| 2095 |
+
"step": 298
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.92,
|
| 2099 |
+
"grad_norm": 0.43601752379760356,
|
| 2100 |
+
"learning_rate": 2.076754087988214e-06,
|
| 2101 |
+
"loss": 0.8964,
|
| 2102 |
+
"step": 299
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.9230769230769231,
|
| 2106 |
+
"grad_norm": 0.6159815837205224,
|
| 2107 |
+
"learning_rate": 1.9369152030840556e-06,
|
| 2108 |
+
"loss": 0.9867,
|
| 2109 |
+
"step": 300
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.9261538461538461,
|
| 2113 |
+
"grad_norm": 0.43453599542906096,
|
| 2114 |
+
"learning_rate": 1.8018569652073381e-06,
|
| 2115 |
+
"loss": 0.8666,
|
| 2116 |
+
"step": 301
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.9292307692307692,
|
| 2120 |
+
"grad_norm": 0.42707059824376786,
|
| 2121 |
+
"learning_rate": 1.6715928080726417e-06,
|
| 2122 |
+
"loss": 0.763,
|
| 2123 |
+
"step": 302
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.9323076923076923,
|
| 2127 |
+
"grad_norm": 0.38795492210093135,
|
| 2128 |
+
"learning_rate": 1.5461356885461075e-06,
|
| 2129 |
+
"loss": 1.0533,
|
| 2130 |
+
"step": 303
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.9353846153846154,
|
| 2134 |
+
"grad_norm": 0.44531193680146935,
|
| 2135 |
+
"learning_rate": 1.4254980853566247e-06,
|
| 2136 |
+
"loss": 0.9636,
|
| 2137 |
+
"step": 304
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.9384615384615385,
|
| 2141 |
+
"grad_norm": 0.8849506994833789,
|
| 2142 |
+
"learning_rate": 1.309691997854684e-06,
|
| 2143 |
+
"loss": 1.1761,
|
| 2144 |
+
"step": 305
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.9415384615384615,
|
| 2148 |
+
"grad_norm": 0.35891301706704143,
|
| 2149 |
+
"learning_rate": 1.1987289448187776e-06,
|
| 2150 |
+
"loss": 0.9216,
|
| 2151 |
+
"step": 306
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.9446153846153846,
|
| 2155 |
+
"grad_norm": 0.6119759873752704,
|
| 2156 |
+
"learning_rate": 1.0926199633097157e-06,
|
| 2157 |
+
"loss": 0.7769,
|
| 2158 |
+
"step": 307
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.9476923076923077,
|
| 2162 |
+
"grad_norm": 0.4614517217456775,
|
| 2163 |
+
"learning_rate": 9.913756075728087e-07,
|
| 2164 |
+
"loss": 0.9269,
|
| 2165 |
+
"step": 308
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.9507692307692308,
|
| 2169 |
+
"grad_norm": 0.36451957631452775,
|
| 2170 |
+
"learning_rate": 8.95005947988059e-07,
|
| 2171 |
+
"loss": 0.8003,
|
| 2172 |
+
"step": 309
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.9538461538461539,
|
| 2176 |
+
"grad_norm": 0.44179747932365987,
|
| 2177 |
+
"learning_rate": 8.035205700685167e-07,
|
| 2178 |
+
"loss": 0.9403,
|
| 2179 |
+
"step": 310
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.9569230769230769,
|
| 2183 |
+
"grad_norm": 0.4817228850083378,
|
| 2184 |
+
"learning_rate": 7.169285735068531e-07,
|
| 2185 |
+
"loss": 0.8474,
|
| 2186 |
+
"step": 311
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.96,
|
| 2190 |
+
"grad_norm": 0.5461341935389407,
|
| 2191 |
+
"learning_rate": 6.352385712702191e-07,
|
| 2192 |
+
"loss": 0.9109,
|
| 2193 |
+
"step": 312
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.963076923076923,
|
| 2197 |
+
"grad_norm": 0.5228257053335598,
|
| 2198 |
+
"learning_rate": 5.584586887435739e-07,
|
| 2199 |
+
"loss": 0.9305,
|
| 2200 |
+
"step": 313
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.9661538461538461,
|
| 2204 |
+
"grad_norm": 0.40498879135347843,
|
| 2205 |
+
"learning_rate": 4.865965629214819e-07,
|
| 2206 |
+
"loss": 0.8711,
|
| 2207 |
+
"step": 314
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.9692307692307692,
|
| 2211 |
+
"grad_norm": 0.37335586916922653,
|
| 2212 |
+
"learning_rate": 4.196593416484873e-07,
|
| 2213 |
+
"loss": 0.9677,
|
| 2214 |
+
"step": 315
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.9723076923076923,
|
| 2218 |
+
"grad_norm": 0.6059932362277927,
|
| 2219 |
+
"learning_rate": 3.576536829081323e-07,
|
| 2220 |
+
"loss": 1.1416,
|
| 2221 |
+
"step": 316
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.9753846153846154,
|
| 2225 |
+
"grad_norm": 0.4741658286315134,
|
| 2226 |
+
"learning_rate": 3.0058575416073707e-07,
|
| 2227 |
+
"loss": 1.0181,
|
| 2228 |
+
"step": 317
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.9784615384615385,
|
| 2232 |
+
"grad_norm": 0.4136898383461235,
|
| 2233 |
+
"learning_rate": 2.4846123172992954e-07,
|
| 2234 |
+
"loss": 0.9476,
|
| 2235 |
+
"step": 318
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.9815384615384616,
|
| 2239 |
+
"grad_norm": 0.5478103993891321,
|
| 2240 |
+
"learning_rate": 2.012853002380466e-07,
|
| 2241 |
+
"loss": 0.9932,
|
| 2242 |
+
"step": 319
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.9846153846153847,
|
| 2246 |
+
"grad_norm": 0.49911294260454786,
|
| 2247 |
+
"learning_rate": 1.5906265209045257e-07,
|
| 2248 |
+
"loss": 0.8899,
|
| 2249 |
+
"step": 320
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.9876923076923076,
|
| 2253 |
+
"grad_norm": 0.7557024891774665,
|
| 2254 |
+
"learning_rate": 1.2179748700879012e-07,
|
| 2255 |
+
"loss": 0.8795,
|
| 2256 |
+
"step": 321
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.9907692307692307,
|
| 2260 |
+
"grad_norm": 0.5085798573597269,
|
| 2261 |
+
"learning_rate": 8.949351161324227e-08,
|
| 2262 |
+
"loss": 0.8276,
|
| 2263 |
+
"step": 322
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.9938461538461538,
|
| 2267 |
+
"grad_norm": 0.7292171412681792,
|
| 2268 |
+
"learning_rate": 6.215393905388278e-08,
|
| 2269 |
+
"loss": 1.0912,
|
| 2270 |
+
"step": 323
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.9969230769230769,
|
| 2274 |
+
"grad_norm": 0.4491633008677572,
|
| 2275 |
+
"learning_rate": 3.9781488691037484e-08,
|
| 2276 |
+
"loss": 0.9607,
|
| 2277 |
+
"step": 324
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 1.0,
|
| 2281 |
+
"grad_norm": 0.9103783468264505,
|
| 2282 |
+
"learning_rate": 2.237838582483387e-08,
|
| 2283 |
+
"loss": 1.1181,
|
| 2284 |
+
"step": 325
|
| 2285 |
+
}
|
| 2286 |
+
],
|
| 2287 |
+
"logging_steps": 1,
|
| 2288 |
+
"max_steps": 325,
|
| 2289 |
+
"num_input_tokens_seen": 0,
|
| 2290 |
+
"num_train_epochs": 1,
|
| 2291 |
+
"save_steps": 0,
|
| 2292 |
+
"stateful_callbacks": {
|
| 2293 |
+
"TrainerControl": {
|
| 2294 |
+
"args": {
|
| 2295 |
+
"should_epoch_stop": false,
|
| 2296 |
+
"should_evaluate": false,
|
| 2297 |
+
"should_log": false,
|
| 2298 |
+
"should_save": true,
|
| 2299 |
+
"should_training_stop": true
|
| 2300 |
+
},
|
| 2301 |
+
"attributes": {}
|
| 2302 |
+
}
|
| 2303 |
+
},
|
| 2304 |
+
"total_flos": 144092413624320.0,
|
| 2305 |
+
"train_batch_size": 4,
|
| 2306 |
+
"trial_name": null,
|
| 2307 |
+
"trial_params": null
|
| 2308 |
+
}
|