Upload Orca_End/trainer_state.json with huggingface_hub
Browse files- Orca_End/trainer_state.json +3095 -0
Orca_End/trainer_state.json
ADDED
|
@@ -0,0 +1,3095 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.14174455404281616,
|
| 3 |
+
"best_model_checkpoint": "/home/ray/default/save/Orca_End/checkpoint-4000",
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4272,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0046816479400749065,
|
| 13 |
+
"grad_norm": 0.14989347755908966,
|
| 14 |
+
"learning_rate": 5.000000000000001e-07,
|
| 15 |
+
"loss": 0.9233,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.009363295880149813,
|
| 20 |
+
"grad_norm": 0.16152486205101013,
|
| 21 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 22 |
+
"loss": 0.9302,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.014044943820224719,
|
| 27 |
+
"grad_norm": 0.14786499738693237,
|
| 28 |
+
"learning_rate": 1.5e-06,
|
| 29 |
+
"loss": 0.9066,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.018726591760299626,
|
| 34 |
+
"grad_norm": 0.17683209478855133,
|
| 35 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 36 |
+
"loss": 0.8958,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.023408239700374533,
|
| 41 |
+
"grad_norm": 0.18914400041103363,
|
| 42 |
+
"learning_rate": 2.5e-06,
|
| 43 |
+
"loss": 0.8957,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.028089887640449437,
|
| 48 |
+
"grad_norm": 0.17229343950748444,
|
| 49 |
+
"learning_rate": 3e-06,
|
| 50 |
+
"loss": 0.9006,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.03277153558052434,
|
| 55 |
+
"grad_norm": 0.19986914098262787,
|
| 56 |
+
"learning_rate": 3.5000000000000004e-06,
|
| 57 |
+
"loss": 0.8843,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.03745318352059925,
|
| 62 |
+
"grad_norm": 0.23050718009471893,
|
| 63 |
+
"learning_rate": 4.000000000000001e-06,
|
| 64 |
+
"loss": 0.8824,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.042134831460674156,
|
| 69 |
+
"grad_norm": 0.2016942948102951,
|
| 70 |
+
"learning_rate": 4.5e-06,
|
| 71 |
+
"loss": 0.8842,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.04681647940074907,
|
| 76 |
+
"grad_norm": 0.24438633024692535,
|
| 77 |
+
"learning_rate": 5e-06,
|
| 78 |
+
"loss": 0.8783,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.05149812734082397,
|
| 83 |
+
"grad_norm": 0.33506831526756287,
|
| 84 |
+
"learning_rate": 5.500000000000001e-06,
|
| 85 |
+
"loss": 0.854,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.056179775280898875,
|
| 90 |
+
"grad_norm": 0.29310888051986694,
|
| 91 |
+
"learning_rate": 6e-06,
|
| 92 |
+
"loss": 0.8332,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.060861423220973786,
|
| 97 |
+
"grad_norm": 0.3248985707759857,
|
| 98 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 99 |
+
"loss": 0.8105,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.06554307116104868,
|
| 104 |
+
"grad_norm": 0.31335774064064026,
|
| 105 |
+
"learning_rate": 7.000000000000001e-06,
|
| 106 |
+
"loss": 0.7664,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.0702247191011236,
|
| 111 |
+
"grad_norm": 0.2475409060716629,
|
| 112 |
+
"learning_rate": 7.5e-06,
|
| 113 |
+
"loss": 0.7265,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.0749063670411985,
|
| 118 |
+
"grad_norm": 0.19633963704109192,
|
| 119 |
+
"learning_rate": 8.000000000000001e-06,
|
| 120 |
+
"loss": 0.694,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.07958801498127341,
|
| 125 |
+
"grad_norm": 0.18646350502967834,
|
| 126 |
+
"learning_rate": 8.500000000000002e-06,
|
| 127 |
+
"loss": 0.6975,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.08426966292134831,
|
| 132 |
+
"grad_norm": 0.17781223356723785,
|
| 133 |
+
"learning_rate": 9e-06,
|
| 134 |
+
"loss": 0.6493,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.08895131086142322,
|
| 139 |
+
"grad_norm": 0.15868598222732544,
|
| 140 |
+
"learning_rate": 9.5e-06,
|
| 141 |
+
"loss": 0.6194,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.09363295880149813,
|
| 146 |
+
"grad_norm": 0.14928707480430603,
|
| 147 |
+
"learning_rate": 1e-05,
|
| 148 |
+
"loss": 0.6004,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.09831460674157304,
|
| 153 |
+
"grad_norm": 0.18063120543956757,
|
| 154 |
+
"learning_rate": 1.05e-05,
|
| 155 |
+
"loss": 0.5937,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.10299625468164794,
|
| 160 |
+
"grad_norm": 0.1594826579093933,
|
| 161 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 162 |
+
"loss": 0.5582,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.10767790262172285,
|
| 167 |
+
"grad_norm": 0.15458233654499054,
|
| 168 |
+
"learning_rate": 1.1500000000000002e-05,
|
| 169 |
+
"loss": 0.5495,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.11235955056179775,
|
| 174 |
+
"grad_norm": 0.14062783122062683,
|
| 175 |
+
"learning_rate": 1.2e-05,
|
| 176 |
+
"loss": 0.5155,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.11704119850187265,
|
| 181 |
+
"grad_norm": 0.16381382942199707,
|
| 182 |
+
"learning_rate": 1.25e-05,
|
| 183 |
+
"loss": 0.4931,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.12172284644194757,
|
| 188 |
+
"grad_norm": 0.17048245668411255,
|
| 189 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 190 |
+
"loss": 0.4842,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.12640449438202248,
|
| 195 |
+
"grad_norm": 0.16214580833911896,
|
| 196 |
+
"learning_rate": 1.3500000000000001e-05,
|
| 197 |
+
"loss": 0.4565,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.13108614232209737,
|
| 202 |
+
"grad_norm": 0.18523240089416504,
|
| 203 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 204 |
+
"loss": 0.4303,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.13576779026217228,
|
| 209 |
+
"grad_norm": 0.16191580891609192,
|
| 210 |
+
"learning_rate": 1.45e-05,
|
| 211 |
+
"loss": 0.3903,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.1404494382022472,
|
| 216 |
+
"grad_norm": 0.17054912447929382,
|
| 217 |
+
"learning_rate": 1.5e-05,
|
| 218 |
+
"loss": 0.3958,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.1451310861423221,
|
| 223 |
+
"grad_norm": 0.15889020264148712,
|
| 224 |
+
"learning_rate": 1.55e-05,
|
| 225 |
+
"loss": 0.3679,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.149812734082397,
|
| 230 |
+
"grad_norm": 0.1658034473657608,
|
| 231 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 232 |
+
"loss": 0.3566,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.1544943820224719,
|
| 237 |
+
"grad_norm": 0.1820608377456665,
|
| 238 |
+
"learning_rate": 1.65e-05,
|
| 239 |
+
"loss": 0.3353,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.15917602996254682,
|
| 244 |
+
"grad_norm": 0.18187420070171356,
|
| 245 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 246 |
+
"loss": 0.3335,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.16385767790262173,
|
| 251 |
+
"grad_norm": 0.20398898422718048,
|
| 252 |
+
"learning_rate": 1.75e-05,
|
| 253 |
+
"loss": 0.3273,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.16853932584269662,
|
| 258 |
+
"grad_norm": 0.19581130146980286,
|
| 259 |
+
"learning_rate": 1.8e-05,
|
| 260 |
+
"loss": 0.2973,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.17322097378277154,
|
| 265 |
+
"grad_norm": 0.1903996765613556,
|
| 266 |
+
"learning_rate": 1.85e-05,
|
| 267 |
+
"loss": 0.2939,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.17790262172284643,
|
| 272 |
+
"grad_norm": 0.21215596795082092,
|
| 273 |
+
"learning_rate": 1.9e-05,
|
| 274 |
+
"loss": 0.2836,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.18258426966292135,
|
| 279 |
+
"grad_norm": 0.20406150817871094,
|
| 280 |
+
"learning_rate": 1.9500000000000003e-05,
|
| 281 |
+
"loss": 0.2766,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.18726591760299627,
|
| 286 |
+
"grad_norm": 0.19281837344169617,
|
| 287 |
+
"learning_rate": 2e-05,
|
| 288 |
+
"loss": 0.2716,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.19194756554307116,
|
| 293 |
+
"grad_norm": 0.1925257295370102,
|
| 294 |
+
"learning_rate": 2.05e-05,
|
| 295 |
+
"loss": 0.2619,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.19662921348314608,
|
| 300 |
+
"grad_norm": 0.20710717141628265,
|
| 301 |
+
"learning_rate": 2.1e-05,
|
| 302 |
+
"loss": 0.2676,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.20131086142322097,
|
| 307 |
+
"grad_norm": 0.2195947766304016,
|
| 308 |
+
"learning_rate": 2.15e-05,
|
| 309 |
+
"loss": 0.2675,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.20599250936329588,
|
| 314 |
+
"grad_norm": 0.23015384376049042,
|
| 315 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 316 |
+
"loss": 0.24,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.21067415730337077,
|
| 321 |
+
"grad_norm": 0.23879607021808624,
|
| 322 |
+
"learning_rate": 2.25e-05,
|
| 323 |
+
"loss": 0.251,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.2153558052434457,
|
| 328 |
+
"grad_norm": 0.25906631350517273,
|
| 329 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 330 |
+
"loss": 0.2485,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.2200374531835206,
|
| 335 |
+
"grad_norm": 0.2271612286567688,
|
| 336 |
+
"learning_rate": 2.35e-05,
|
| 337 |
+
"loss": 0.2291,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.2247191011235955,
|
| 342 |
+
"grad_norm": 0.23806224763393402,
|
| 343 |
+
"learning_rate": 2.4e-05,
|
| 344 |
+
"loss": 0.2262,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.22940074906367042,
|
| 349 |
+
"grad_norm": 0.23863674700260162,
|
| 350 |
+
"learning_rate": 2.45e-05,
|
| 351 |
+
"loss": 0.2481,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.2340823970037453,
|
| 356 |
+
"grad_norm": 0.23864531517028809,
|
| 357 |
+
"learning_rate": 2.5e-05,
|
| 358 |
+
"loss": 0.2267,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.2340823970037453,
|
| 363 |
+
"eval_loss": 0.22292079031467438,
|
| 364 |
+
"eval_runtime": 233.7266,
|
| 365 |
+
"eval_samples_per_second": 2.986,
|
| 366 |
+
"eval_steps_per_second": 1.493,
|
| 367 |
+
"step": 500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.23876404494382023,
|
| 371 |
+
"grad_norm": 0.229400172829628,
|
| 372 |
+
"learning_rate": 2.5500000000000003e-05,
|
| 373 |
+
"loss": 0.2235,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.24344569288389514,
|
| 378 |
+
"grad_norm": 0.25184643268585205,
|
| 379 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 380 |
+
"loss": 0.2177,
|
| 381 |
+
"step": 520
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.24812734082397003,
|
| 385 |
+
"grad_norm": 0.255626916885376,
|
| 386 |
+
"learning_rate": 2.6500000000000004e-05,
|
| 387 |
+
"loss": 0.2283,
|
| 388 |
+
"step": 530
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.25280898876404495,
|
| 392 |
+
"grad_norm": 0.2295573651790619,
|
| 393 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 394 |
+
"loss": 0.2203,
|
| 395 |
+
"step": 540
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.25749063670411987,
|
| 399 |
+
"grad_norm": 0.24802301824092865,
|
| 400 |
+
"learning_rate": 2.7500000000000004e-05,
|
| 401 |
+
"loss": 0.2218,
|
| 402 |
+
"step": 550
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.26217228464419473,
|
| 406 |
+
"grad_norm": 0.26119089126586914,
|
| 407 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 408 |
+
"loss": 0.2073,
|
| 409 |
+
"step": 560
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.26685393258426965,
|
| 413 |
+
"grad_norm": 0.23319612443447113,
|
| 414 |
+
"learning_rate": 2.8499999999999998e-05,
|
| 415 |
+
"loss": 0.215,
|
| 416 |
+
"step": 570
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.27153558052434457,
|
| 420 |
+
"grad_norm": 0.2370581328868866,
|
| 421 |
+
"learning_rate": 2.9e-05,
|
| 422 |
+
"loss": 0.2066,
|
| 423 |
+
"step": 580
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.2762172284644195,
|
| 427 |
+
"grad_norm": 0.22631758451461792,
|
| 428 |
+
"learning_rate": 2.95e-05,
|
| 429 |
+
"loss": 0.2054,
|
| 430 |
+
"step": 590
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.2808988764044944,
|
| 434 |
+
"grad_norm": 0.259056031703949,
|
| 435 |
+
"learning_rate": 3e-05,
|
| 436 |
+
"loss": 0.2162,
|
| 437 |
+
"step": 600
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.28558052434456926,
|
| 441 |
+
"grad_norm": 0.24505385756492615,
|
| 442 |
+
"learning_rate": 3.05e-05,
|
| 443 |
+
"loss": 0.2033,
|
| 444 |
+
"step": 610
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.2902621722846442,
|
| 448 |
+
"grad_norm": 0.22634395956993103,
|
| 449 |
+
"learning_rate": 3.1e-05,
|
| 450 |
+
"loss": 0.1967,
|
| 451 |
+
"step": 620
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.2949438202247191,
|
| 455 |
+
"grad_norm": 0.3079870641231537,
|
| 456 |
+
"learning_rate": 3.15e-05,
|
| 457 |
+
"loss": 0.2125,
|
| 458 |
+
"step": 630
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.299625468164794,
|
| 462 |
+
"grad_norm": 0.2437993735074997,
|
| 463 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 464 |
+
"loss": 0.2136,
|
| 465 |
+
"step": 640
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.30430711610486894,
|
| 469 |
+
"grad_norm": 0.25136232376098633,
|
| 470 |
+
"learning_rate": 3.2500000000000004e-05,
|
| 471 |
+
"loss": 0.2173,
|
| 472 |
+
"step": 650
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.3089887640449438,
|
| 476 |
+
"grad_norm": 0.263988733291626,
|
| 477 |
+
"learning_rate": 3.3e-05,
|
| 478 |
+
"loss": 0.2245,
|
| 479 |
+
"step": 660
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.3136704119850187,
|
| 483 |
+
"grad_norm": 0.3366784453392029,
|
| 484 |
+
"learning_rate": 3.35e-05,
|
| 485 |
+
"loss": 0.1881,
|
| 486 |
+
"step": 670
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.31835205992509363,
|
| 490 |
+
"grad_norm": 0.22763469815254211,
|
| 491 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 492 |
+
"loss": 0.2009,
|
| 493 |
+
"step": 680
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.32303370786516855,
|
| 497 |
+
"grad_norm": 0.28085455298423767,
|
| 498 |
+
"learning_rate": 3.45e-05,
|
| 499 |
+
"loss": 0.193,
|
| 500 |
+
"step": 690
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.32771535580524347,
|
| 504 |
+
"grad_norm": 0.21730464696884155,
|
| 505 |
+
"learning_rate": 3.5e-05,
|
| 506 |
+
"loss": 0.1956,
|
| 507 |
+
"step": 700
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.33239700374531833,
|
| 511 |
+
"grad_norm": 0.2521045506000519,
|
| 512 |
+
"learning_rate": 3.55e-05,
|
| 513 |
+
"loss": 0.2393,
|
| 514 |
+
"step": 710
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.33707865168539325,
|
| 518 |
+
"grad_norm": 0.2687033414840698,
|
| 519 |
+
"learning_rate": 3.6e-05,
|
| 520 |
+
"loss": 0.193,
|
| 521 |
+
"step": 720
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.34176029962546817,
|
| 525 |
+
"grad_norm": 0.23822762072086334,
|
| 526 |
+
"learning_rate": 3.65e-05,
|
| 527 |
+
"loss": 0.1897,
|
| 528 |
+
"step": 730
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.3464419475655431,
|
| 532 |
+
"grad_norm": 0.2817654609680176,
|
| 533 |
+
"learning_rate": 3.7e-05,
|
| 534 |
+
"loss": 0.1874,
|
| 535 |
+
"step": 740
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.351123595505618,
|
| 539 |
+
"grad_norm": 0.2669580578804016,
|
| 540 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 541 |
+
"loss": 0.1876,
|
| 542 |
+
"step": 750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.35580524344569286,
|
| 546 |
+
"grad_norm": 0.26698926091194153,
|
| 547 |
+
"learning_rate": 3.8e-05,
|
| 548 |
+
"loss": 0.189,
|
| 549 |
+
"step": 760
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.3604868913857678,
|
| 553 |
+
"grad_norm": 0.24417920410633087,
|
| 554 |
+
"learning_rate": 3.85e-05,
|
| 555 |
+
"loss": 0.1904,
|
| 556 |
+
"step": 770
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.3651685393258427,
|
| 560 |
+
"grad_norm": 0.3116750121116638,
|
| 561 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 562 |
+
"loss": 0.1894,
|
| 563 |
+
"step": 780
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.3698501872659176,
|
| 567 |
+
"grad_norm": 0.269699364900589,
|
| 568 |
+
"learning_rate": 3.9500000000000005e-05,
|
| 569 |
+
"loss": 0.1948,
|
| 570 |
+
"step": 790
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.37453183520599254,
|
| 574 |
+
"grad_norm": 0.24539124965667725,
|
| 575 |
+
"learning_rate": 4e-05,
|
| 576 |
+
"loss": 0.1924,
|
| 577 |
+
"step": 800
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.3792134831460674,
|
| 581 |
+
"grad_norm": 0.24850206077098846,
|
| 582 |
+
"learning_rate": 4.05e-05,
|
| 583 |
+
"loss": 0.1902,
|
| 584 |
+
"step": 810
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.3838951310861423,
|
| 588 |
+
"grad_norm": 0.24830132722854614,
|
| 589 |
+
"learning_rate": 4.1e-05,
|
| 590 |
+
"loss": 0.1753,
|
| 591 |
+
"step": 820
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.38857677902621723,
|
| 595 |
+
"grad_norm": 0.28210338950157166,
|
| 596 |
+
"learning_rate": 4.15e-05,
|
| 597 |
+
"loss": 0.183,
|
| 598 |
+
"step": 830
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.39325842696629215,
|
| 602 |
+
"grad_norm": 0.2719564437866211,
|
| 603 |
+
"learning_rate": 4.2e-05,
|
| 604 |
+
"loss": 0.1682,
|
| 605 |
+
"step": 840
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.397940074906367,
|
| 609 |
+
"grad_norm": 0.2886371612548828,
|
| 610 |
+
"learning_rate": 4.25e-05,
|
| 611 |
+
"loss": 0.1756,
|
| 612 |
+
"step": 850
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.40262172284644193,
|
| 616 |
+
"grad_norm": 0.27633577585220337,
|
| 617 |
+
"learning_rate": 4.3e-05,
|
| 618 |
+
"loss": 0.1879,
|
| 619 |
+
"step": 860
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.40730337078651685,
|
| 623 |
+
"grad_norm": 0.24454927444458008,
|
| 624 |
+
"learning_rate": 4.35e-05,
|
| 625 |
+
"loss": 0.1832,
|
| 626 |
+
"step": 870
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.41198501872659177,
|
| 630 |
+
"grad_norm": 0.258569598197937,
|
| 631 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 632 |
+
"loss": 0.1781,
|
| 633 |
+
"step": 880
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.4166666666666667,
|
| 637 |
+
"grad_norm": 0.24131977558135986,
|
| 638 |
+
"learning_rate": 4.4500000000000004e-05,
|
| 639 |
+
"loss": 0.1751,
|
| 640 |
+
"step": 890
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.42134831460674155,
|
| 644 |
+
"grad_norm": 0.26839905977249146,
|
| 645 |
+
"learning_rate": 4.5e-05,
|
| 646 |
+
"loss": 0.1853,
|
| 647 |
+
"step": 900
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.42602996254681647,
|
| 651 |
+
"grad_norm": 0.2593300938606262,
|
| 652 |
+
"learning_rate": 4.55e-05,
|
| 653 |
+
"loss": 0.208,
|
| 654 |
+
"step": 910
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.4307116104868914,
|
| 658 |
+
"grad_norm": 0.25040218234062195,
|
| 659 |
+
"learning_rate": 4.600000000000001e-05,
|
| 660 |
+
"loss": 0.1803,
|
| 661 |
+
"step": 920
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.4353932584269663,
|
| 665 |
+
"grad_norm": 0.2636476159095764,
|
| 666 |
+
"learning_rate": 4.6500000000000005e-05,
|
| 667 |
+
"loss": 0.1743,
|
| 668 |
+
"step": 930
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.4400749063670412,
|
| 672 |
+
"grad_norm": 0.2569904625415802,
|
| 673 |
+
"learning_rate": 4.7e-05,
|
| 674 |
+
"loss": 0.1777,
|
| 675 |
+
"step": 940
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.4447565543071161,
|
| 679 |
+
"grad_norm": 0.2652021050453186,
|
| 680 |
+
"learning_rate": 4.75e-05,
|
| 681 |
+
"loss": 0.1759,
|
| 682 |
+
"step": 950
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.449438202247191,
|
| 686 |
+
"grad_norm": 0.25217893719673157,
|
| 687 |
+
"learning_rate": 4.8e-05,
|
| 688 |
+
"loss": 0.183,
|
| 689 |
+
"step": 960
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.4541198501872659,
|
| 693 |
+
"grad_norm": 0.26192978024482727,
|
| 694 |
+
"learning_rate": 4.85e-05,
|
| 695 |
+
"loss": 0.1773,
|
| 696 |
+
"step": 970
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.45880149812734083,
|
| 700 |
+
"grad_norm": 0.2585752308368683,
|
| 701 |
+
"learning_rate": 4.9e-05,
|
| 702 |
+
"loss": 0.1811,
|
| 703 |
+
"step": 980
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.46348314606741575,
|
| 707 |
+
"grad_norm": 0.2950770854949951,
|
| 708 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 709 |
+
"loss": 0.1742,
|
| 710 |
+
"step": 990
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.4681647940074906,
|
| 714 |
+
"grad_norm": 0.24856804311275482,
|
| 715 |
+
"learning_rate": 5e-05,
|
| 716 |
+
"loss": 0.177,
|
| 717 |
+
"step": 1000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.4681647940074906,
|
| 721 |
+
"eval_loss": 0.17352555692195892,
|
| 722 |
+
"eval_runtime": 236.2712,
|
| 723 |
+
"eval_samples_per_second": 2.954,
|
| 724 |
+
"eval_steps_per_second": 1.477,
|
| 725 |
+
"step": 1000
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.47284644194756553,
|
| 729 |
+
"grad_norm": 0.2841172218322754,
|
| 730 |
+
"learning_rate": 4.9998847662136905e-05,
|
| 731 |
+
"loss": 0.1799,
|
| 732 |
+
"step": 1010
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.47752808988764045,
|
| 736 |
+
"grad_norm": 0.24798014760017395,
|
| 737 |
+
"learning_rate": 4.999539075477821e-05,
|
| 738 |
+
"loss": 0.1837,
|
| 739 |
+
"step": 1020
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.48220973782771537,
|
| 743 |
+
"grad_norm": 0.2636326551437378,
|
| 744 |
+
"learning_rate": 4.998962959660594e-05,
|
| 745 |
+
"loss": 0.1737,
|
| 746 |
+
"step": 1030
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.4868913857677903,
|
| 750 |
+
"grad_norm": 0.27235954999923706,
|
| 751 |
+
"learning_rate": 4.998156471872415e-05,
|
| 752 |
+
"loss": 0.1649,
|
| 753 |
+
"step": 1040
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.49157303370786515,
|
| 757 |
+
"grad_norm": 0.26181653141975403,
|
| 758 |
+
"learning_rate": 4.9971196864609975e-05,
|
| 759 |
+
"loss": 0.1786,
|
| 760 |
+
"step": 1050
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.49625468164794007,
|
| 764 |
+
"grad_norm": 0.2724978029727936,
|
| 765 |
+
"learning_rate": 4.995852699004508e-05,
|
| 766 |
+
"loss": 0.1709,
|
| 767 |
+
"step": 1060
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.5009363295880149,
|
| 771 |
+
"grad_norm": 0.2702656090259552,
|
| 772 |
+
"learning_rate": 4.9943556263027556e-05,
|
| 773 |
+
"loss": 0.1714,
|
| 774 |
+
"step": 1070
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.5056179775280899,
|
| 778 |
+
"grad_norm": 0.33144789934158325,
|
| 779 |
+
"learning_rate": 4.992628606366426e-05,
|
| 780 |
+
"loss": 0.1653,
|
| 781 |
+
"step": 1080
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.5102996254681648,
|
| 785 |
+
"grad_norm": 0.2642982006072998,
|
| 786 |
+
"learning_rate": 4.990671798404355e-05,
|
| 787 |
+
"loss": 0.1742,
|
| 788 |
+
"step": 1090
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.5149812734082397,
|
| 792 |
+
"grad_norm": 0.2788497805595398,
|
| 793 |
+
"learning_rate": 4.988485382808856e-05,
|
| 794 |
+
"loss": 0.1717,
|
| 795 |
+
"step": 1100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.5196629213483146,
|
| 799 |
+
"grad_norm": 0.23453935980796814,
|
| 800 |
+
"learning_rate": 4.986069561139086e-05,
|
| 801 |
+
"loss": 0.1913,
|
| 802 |
+
"step": 1110
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 0.5243445692883895,
|
| 806 |
+
"grad_norm": 0.2543826401233673,
|
| 807 |
+
"learning_rate": 4.983424556102469e-05,
|
| 808 |
+
"loss": 0.1753,
|
| 809 |
+
"step": 1120
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.5290262172284644,
|
| 813 |
+
"grad_norm": 0.28962579369544983,
|
| 814 |
+
"learning_rate": 4.98055061153416e-05,
|
| 815 |
+
"loss": 0.1702,
|
| 816 |
+
"step": 1130
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.5337078651685393,
|
| 820 |
+
"grad_norm": 0.23189175128936768,
|
| 821 |
+
"learning_rate": 4.97744799237457e-05,
|
| 822 |
+
"loss": 0.1643,
|
| 823 |
+
"step": 1140
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.5383895131086143,
|
| 827 |
+
"grad_norm": 0.21799345314502716,
|
| 828 |
+
"learning_rate": 4.974116984644943e-05,
|
| 829 |
+
"loss": 0.1835,
|
| 830 |
+
"step": 1150
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.5430711610486891,
|
| 834 |
+
"grad_norm": 0.24570296704769135,
|
| 835 |
+
"learning_rate": 4.970557895420984e-05,
|
| 836 |
+
"loss": 0.1802,
|
| 837 |
+
"step": 1160
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.547752808988764,
|
| 841 |
+
"grad_norm": 0.2388901561498642,
|
| 842 |
+
"learning_rate": 4.966771052804555e-05,
|
| 843 |
+
"loss": 0.1701,
|
| 844 |
+
"step": 1170
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.552434456928839,
|
| 848 |
+
"grad_norm": 0.2784363627433777,
|
| 849 |
+
"learning_rate": 4.9627568058934274e-05,
|
| 850 |
+
"loss": 0.1697,
|
| 851 |
+
"step": 1180
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 0.5571161048689138,
|
| 855 |
+
"grad_norm": 0.27096298336982727,
|
| 856 |
+
"learning_rate": 4.958515524749096e-05,
|
| 857 |
+
"loss": 0.1749,
|
| 858 |
+
"step": 1190
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.5617977528089888,
|
| 862 |
+
"grad_norm": 0.25467076897621155,
|
| 863 |
+
"learning_rate": 4.95404760036267e-05,
|
| 864 |
+
"loss": 0.1598,
|
| 865 |
+
"step": 1200
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.5664794007490637,
|
| 869 |
+
"grad_norm": 0.24679610133171082,
|
| 870 |
+
"learning_rate": 4.949353444618825e-05,
|
| 871 |
+
"loss": 0.1769,
|
| 872 |
+
"step": 1210
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.5711610486891385,
|
| 876 |
+
"grad_norm": 0.2690020799636841,
|
| 877 |
+
"learning_rate": 4.9444334902578315e-05,
|
| 878 |
+
"loss": 0.1676,
|
| 879 |
+
"step": 1220
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.5758426966292135,
|
| 883 |
+
"grad_norm": 0.2583072781562805,
|
| 884 |
+
"learning_rate": 4.9392881908356656e-05,
|
| 885 |
+
"loss": 0.1625,
|
| 886 |
+
"step": 1230
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.5805243445692884,
|
| 890 |
+
"grad_norm": 0.2679818868637085,
|
| 891 |
+
"learning_rate": 4.9339180206821955e-05,
|
| 892 |
+
"loss": 0.1708,
|
| 893 |
+
"step": 1240
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 0.5852059925093633,
|
| 897 |
+
"grad_norm": 0.2509777247905731,
|
| 898 |
+
"learning_rate": 4.9283234748574524e-05,
|
| 899 |
+
"loss": 0.1803,
|
| 900 |
+
"step": 1250
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 0.5898876404494382,
|
| 904 |
+
"grad_norm": 0.2875745892524719,
|
| 905 |
+
"learning_rate": 4.922505069105995e-05,
|
| 906 |
+
"loss": 0.1712,
|
| 907 |
+
"step": 1260
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 0.5945692883895131,
|
| 911 |
+
"grad_norm": 0.25088638067245483,
|
| 912 |
+
"learning_rate": 4.9164633398093624e-05,
|
| 913 |
+
"loss": 0.1591,
|
| 914 |
+
"step": 1270
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.599250936329588,
|
| 918 |
+
"grad_norm": 0.27187079191207886,
|
| 919 |
+
"learning_rate": 4.9101988439366295e-05,
|
| 920 |
+
"loss": 0.1621,
|
| 921 |
+
"step": 1280
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.6039325842696629,
|
| 925 |
+
"grad_norm": 0.23108245432376862,
|
| 926 |
+
"learning_rate": 4.90371215899306e-05,
|
| 927 |
+
"loss": 0.173,
|
| 928 |
+
"step": 1290
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 0.6086142322097379,
|
| 932 |
+
"grad_norm": 0.2536250352859497,
|
| 933 |
+
"learning_rate": 4.897003882966866e-05,
|
| 934 |
+
"loss": 0.1692,
|
| 935 |
+
"step": 1300
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 0.6132958801498127,
|
| 939 |
+
"grad_norm": 0.2388300895690918,
|
| 940 |
+
"learning_rate": 4.8900746342740854e-05,
|
| 941 |
+
"loss": 0.1676,
|
| 942 |
+
"step": 1310
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 0.6179775280898876,
|
| 946 |
+
"grad_norm": 0.23416273295879364,
|
| 947 |
+
"learning_rate": 4.8829250517015684e-05,
|
| 948 |
+
"loss": 0.1739,
|
| 949 |
+
"step": 1320
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 0.6226591760299626,
|
| 953 |
+
"grad_norm": 0.24538950622081757,
|
| 954 |
+
"learning_rate": 4.87555579434809e-05,
|
| 955 |
+
"loss": 0.1622,
|
| 956 |
+
"step": 1330
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.6273408239700374,
|
| 960 |
+
"grad_norm": 0.27460214495658875,
|
| 961 |
+
"learning_rate": 4.867967541563594e-05,
|
| 962 |
+
"loss": 0.1769,
|
| 963 |
+
"step": 1340
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.6320224719101124,
|
| 967 |
+
"grad_norm": 0.23591691255569458,
|
| 968 |
+
"learning_rate": 4.86016099288656e-05,
|
| 969 |
+
"loss": 0.1471,
|
| 970 |
+
"step": 1350
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 0.6367041198501873,
|
| 974 |
+
"grad_norm": 0.23421670496463776,
|
| 975 |
+
"learning_rate": 4.8521368679795154e-05,
|
| 976 |
+
"loss": 0.1751,
|
| 977 |
+
"step": 1360
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.6413857677902621,
|
| 981 |
+
"grad_norm": 0.2746402323246002,
|
| 982 |
+
"learning_rate": 4.843895906562699e-05,
|
| 983 |
+
"loss": 0.1602,
|
| 984 |
+
"step": 1370
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.6460674157303371,
|
| 988 |
+
"grad_norm": 0.2615332007408142,
|
| 989 |
+
"learning_rate": 4.835438868345858e-05,
|
| 990 |
+
"loss": 0.1627,
|
| 991 |
+
"step": 1380
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 0.650749063670412,
|
| 995 |
+
"grad_norm": 0.24973183870315552,
|
| 996 |
+
"learning_rate": 4.826766532958221e-05,
|
| 997 |
+
"loss": 0.163,
|
| 998 |
+
"step": 1390
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.6554307116104869,
|
| 1002 |
+
"grad_norm": 0.3358960449695587,
|
| 1003 |
+
"learning_rate": 4.817879699876623e-05,
|
| 1004 |
+
"loss": 0.1732,
|
| 1005 |
+
"step": 1400
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.6601123595505618,
|
| 1009 |
+
"grad_norm": 0.23806847631931305,
|
| 1010 |
+
"learning_rate": 4.808779188351803e-05,
|
| 1011 |
+
"loss": 0.1513,
|
| 1012 |
+
"step": 1410
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 0.6647940074906367,
|
| 1016 |
+
"grad_norm": 0.30218422412872314,
|
| 1017 |
+
"learning_rate": 4.7994658373328804e-05,
|
| 1018 |
+
"loss": 0.1683,
|
| 1019 |
+
"step": 1420
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 0.6694756554307116,
|
| 1023 |
+
"grad_norm": 0.28052952885627747,
|
| 1024 |
+
"learning_rate": 4.7899405053900174e-05,
|
| 1025 |
+
"loss": 0.1656,
|
| 1026 |
+
"step": 1430
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 0.6741573033707865,
|
| 1030 |
+
"grad_norm": 0.2771318852901459,
|
| 1031 |
+
"learning_rate": 4.780204070635266e-05,
|
| 1032 |
+
"loss": 0.1683,
|
| 1033 |
+
"step": 1440
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 0.6788389513108615,
|
| 1037 |
+
"grad_norm": 0.28290867805480957,
|
| 1038 |
+
"learning_rate": 4.770257430641619e-05,
|
| 1039 |
+
"loss": 0.1619,
|
| 1040 |
+
"step": 1450
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.6835205992509363,
|
| 1044 |
+
"grad_norm": 0.2428758144378662,
|
| 1045 |
+
"learning_rate": 4.760101502360268e-05,
|
| 1046 |
+
"loss": 0.1677,
|
| 1047 |
+
"step": 1460
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.6882022471910112,
|
| 1051 |
+
"grad_norm": 0.26167434453964233,
|
| 1052 |
+
"learning_rate": 4.749737222036067e-05,
|
| 1053 |
+
"loss": 0.1657,
|
| 1054 |
+
"step": 1470
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 0.6928838951310862,
|
| 1058 |
+
"grad_norm": 0.27498582005500793,
|
| 1059 |
+
"learning_rate": 4.739165545121228e-05,
|
| 1060 |
+
"loss": 0.1678,
|
| 1061 |
+
"step": 1480
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 0.697565543071161,
|
| 1065 |
+
"grad_norm": 0.2729678750038147,
|
| 1066 |
+
"learning_rate": 4.728387446187238e-05,
|
| 1067 |
+
"loss": 0.1586,
|
| 1068 |
+
"step": 1490
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 0.702247191011236,
|
| 1072 |
+
"grad_norm": 0.28416207432746887,
|
| 1073 |
+
"learning_rate": 4.717403918835017e-05,
|
| 1074 |
+
"loss": 0.1592,
|
| 1075 |
+
"step": 1500
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 0.702247191011236,
|
| 1079 |
+
"eval_loss": 0.15914995968341827,
|
| 1080 |
+
"eval_runtime": 230.5999,
|
| 1081 |
+
"eval_samples_per_second": 3.027,
|
| 1082 |
+
"eval_steps_per_second": 1.513,
|
| 1083 |
+
"step": 1500
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.7069288389513109,
|
| 1087 |
+
"grad_norm": 0.2723831832408905,
|
| 1088 |
+
"learning_rate": 4.706215975603319e-05,
|
| 1089 |
+
"loss": 0.1603,
|
| 1090 |
+
"step": 1510
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 0.7116104868913857,
|
| 1094 |
+
"grad_norm": 0.2503916025161743,
|
| 1095 |
+
"learning_rate": 4.694824647875391e-05,
|
| 1096 |
+
"loss": 0.1656,
|
| 1097 |
+
"step": 1520
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 0.7162921348314607,
|
| 1101 |
+
"grad_norm": 0.2601824998855591,
|
| 1102 |
+
"learning_rate": 4.6832309857838946e-05,
|
| 1103 |
+
"loss": 0.15,
|
| 1104 |
+
"step": 1530
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 0.7209737827715356,
|
| 1108 |
+
"grad_norm": 0.24705946445465088,
|
| 1109 |
+
"learning_rate": 4.6714360581140935e-05,
|
| 1110 |
+
"loss": 0.1538,
|
| 1111 |
+
"step": 1540
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 0.7256554307116105,
|
| 1115 |
+
"grad_norm": 0.22847461700439453,
|
| 1116 |
+
"learning_rate": 4.6594409522053274e-05,
|
| 1117 |
+
"loss": 0.1633,
|
| 1118 |
+
"step": 1550
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.7303370786516854,
|
| 1122 |
+
"grad_norm": 0.32371050119400024,
|
| 1123 |
+
"learning_rate": 4.647246773850773e-05,
|
| 1124 |
+
"loss": 0.1582,
|
| 1125 |
+
"step": 1560
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.7350187265917603,
|
| 1129 |
+
"grad_norm": 0.24720498919487,
|
| 1130 |
+
"learning_rate": 4.6348546471955046e-05,
|
| 1131 |
+
"loss": 0.1741,
|
| 1132 |
+
"step": 1570
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 0.7397003745318352,
|
| 1136 |
+
"grad_norm": 0.2715936601161957,
|
| 1137 |
+
"learning_rate": 4.6222657146328624e-05,
|
| 1138 |
+
"loss": 0.1564,
|
| 1139 |
+
"step": 1580
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 0.7443820224719101,
|
| 1143 |
+
"grad_norm": 0.24092724919319153,
|
| 1144 |
+
"learning_rate": 4.609481136699138e-05,
|
| 1145 |
+
"loss": 0.1864,
|
| 1146 |
+
"step": 1590
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.7490636704119851,
|
| 1150 |
+
"grad_norm": 0.2625042200088501,
|
| 1151 |
+
"learning_rate": 4.596502091966587e-05,
|
| 1152 |
+
"loss": 0.1863,
|
| 1153 |
+
"step": 1600
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 0.7537453183520599,
|
| 1157 |
+
"grad_norm": 0.2534978687763214,
|
| 1158 |
+
"learning_rate": 4.583329776934787e-05,
|
| 1159 |
+
"loss": 0.1768,
|
| 1160 |
+
"step": 1610
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.7584269662921348,
|
| 1164 |
+
"grad_norm": 0.26979076862335205,
|
| 1165 |
+
"learning_rate": 4.5699654059203225e-05,
|
| 1166 |
+
"loss": 0.1587,
|
| 1167 |
+
"step": 1620
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.7631086142322098,
|
| 1171 |
+
"grad_norm": 0.21820847690105438,
|
| 1172 |
+
"learning_rate": 4.5564102109448554e-05,
|
| 1173 |
+
"loss": 0.1597,
|
| 1174 |
+
"step": 1630
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 0.7677902621722846,
|
| 1178 |
+
"grad_norm": 0.22770282626152039,
|
| 1179 |
+
"learning_rate": 4.542665441621537e-05,
|
| 1180 |
+
"loss": 0.1639,
|
| 1181 |
+
"step": 1640
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 0.7724719101123596,
|
| 1185 |
+
"grad_norm": 0.27316999435424805,
|
| 1186 |
+
"learning_rate": 4.5287323650398164e-05,
|
| 1187 |
+
"loss": 0.1546,
|
| 1188 |
+
"step": 1650
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 0.7771535580524345,
|
| 1192 |
+
"grad_norm": 0.27699345350265503,
|
| 1193 |
+
"learning_rate": 4.51461226564863e-05,
|
| 1194 |
+
"loss": 0.1725,
|
| 1195 |
+
"step": 1660
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 0.7818352059925093,
|
| 1199 |
+
"grad_norm": 0.24093034863471985,
|
| 1200 |
+
"learning_rate": 4.50030644513799e-05,
|
| 1201 |
+
"loss": 0.1711,
|
| 1202 |
+
"step": 1670
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.7865168539325843,
|
| 1206 |
+
"grad_norm": 0.27285608649253845,
|
| 1207 |
+
"learning_rate": 4.485816222318986e-05,
|
| 1208 |
+
"loss": 0.1487,
|
| 1209 |
+
"step": 1680
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.7911985018726592,
|
| 1213 |
+
"grad_norm": 0.2582044303417206,
|
| 1214 |
+
"learning_rate": 4.471142933002213e-05,
|
| 1215 |
+
"loss": 0.1582,
|
| 1216 |
+
"step": 1690
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 0.795880149812734,
|
| 1220 |
+
"grad_norm": 0.25239214301109314,
|
| 1221 |
+
"learning_rate": 4.4562879298746165e-05,
|
| 1222 |
+
"loss": 0.155,
|
| 1223 |
+
"step": 1700
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 0.800561797752809,
|
| 1227 |
+
"grad_norm": 0.2343159019947052,
|
| 1228 |
+
"learning_rate": 4.441252582374802e-05,
|
| 1229 |
+
"loss": 0.1591,
|
| 1230 |
+
"step": 1710
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 0.8052434456928839,
|
| 1234 |
+
"grad_norm": 0.264161080121994,
|
| 1235 |
+
"learning_rate": 4.4260382765667875e-05,
|
| 1236 |
+
"loss": 0.1589,
|
| 1237 |
+
"step": 1720
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 0.8099250936329588,
|
| 1241 |
+
"grad_norm": 0.25752800703048706,
|
| 1242 |
+
"learning_rate": 4.410646415012224e-05,
|
| 1243 |
+
"loss": 0.1648,
|
| 1244 |
+
"step": 1730
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.8146067415730337,
|
| 1248 |
+
"grad_norm": 0.28275129199028015,
|
| 1249 |
+
"learning_rate": 4.395078416641099e-05,
|
| 1250 |
+
"loss": 0.1474,
|
| 1251 |
+
"step": 1740
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.8192883895131086,
|
| 1255 |
+
"grad_norm": 0.260202556848526,
|
| 1256 |
+
"learning_rate": 4.3793357166209313e-05,
|
| 1257 |
+
"loss": 0.1626,
|
| 1258 |
+
"step": 1750
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 0.8239700374531835,
|
| 1262 |
+
"grad_norm": 0.2819097340106964,
|
| 1263 |
+
"learning_rate": 4.363419766224465e-05,
|
| 1264 |
+
"loss": 0.1551,
|
| 1265 |
+
"step": 1760
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 0.8286516853932584,
|
| 1269 |
+
"grad_norm": 0.24001863598823547,
|
| 1270 |
+
"learning_rate": 4.34733203269588e-05,
|
| 1271 |
+
"loss": 0.156,
|
| 1272 |
+
"step": 1770
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 0.8333333333333334,
|
| 1276 |
+
"grad_norm": 0.22871051728725433,
|
| 1277 |
+
"learning_rate": 4.3310739991155365e-05,
|
| 1278 |
+
"loss": 0.154,
|
| 1279 |
+
"step": 1780
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 0.8380149812734082,
|
| 1283 |
+
"grad_norm": 0.2900903522968292,
|
| 1284 |
+
"learning_rate": 4.314647164263247e-05,
|
| 1285 |
+
"loss": 0.1618,
|
| 1286 |
+
"step": 1790
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 0.8426966292134831,
|
| 1290 |
+
"grad_norm": 0.25664201378822327,
|
| 1291 |
+
"learning_rate": 4.2980530424801146e-05,
|
| 1292 |
+
"loss": 0.1529,
|
| 1293 |
+
"step": 1800
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 0.8473782771535581,
|
| 1297 |
+
"grad_norm": 0.2573324143886566,
|
| 1298 |
+
"learning_rate": 4.281293163528925e-05,
|
| 1299 |
+
"loss": 0.1531,
|
| 1300 |
+
"step": 1810
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 0.8520599250936329,
|
| 1304 |
+
"grad_norm": 0.2515137493610382,
|
| 1305 |
+
"learning_rate": 4.264369072453126e-05,
|
| 1306 |
+
"loss": 0.176,
|
| 1307 |
+
"step": 1820
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 0.8567415730337079,
|
| 1311 |
+
"grad_norm": 0.2710267901420593,
|
| 1312 |
+
"learning_rate": 4.247282329434394e-05,
|
| 1313 |
+
"loss": 0.1477,
|
| 1314 |
+
"step": 1830
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.8614232209737828,
|
| 1318 |
+
"grad_norm": 0.2349739819765091,
|
| 1319 |
+
"learning_rate": 4.230034509648804e-05,
|
| 1320 |
+
"loss": 0.1583,
|
| 1321 |
+
"step": 1840
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 0.8661048689138576,
|
| 1325 |
+
"grad_norm": 0.23130016028881073,
|
| 1326 |
+
"learning_rate": 4.212627203121617e-05,
|
| 1327 |
+
"loss": 0.1485,
|
| 1328 |
+
"step": 1850
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.8707865168539326,
|
| 1332 |
+
"grad_norm": 0.27001139521598816,
|
| 1333 |
+
"learning_rate": 4.19506201458071e-05,
|
| 1334 |
+
"loss": 0.1498,
|
| 1335 |
+
"step": 1860
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.8754681647940075,
|
| 1339 |
+
"grad_norm": 0.25574082136154175,
|
| 1340 |
+
"learning_rate": 4.177340563308625e-05,
|
| 1341 |
+
"loss": 0.1697,
|
| 1342 |
+
"step": 1870
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 0.8801498127340824,
|
| 1346 |
+
"grad_norm": 0.2866358458995819,
|
| 1347 |
+
"learning_rate": 4.159464482993308e-05,
|
| 1348 |
+
"loss": 0.1584,
|
| 1349 |
+
"step": 1880
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 0.8848314606741573,
|
| 1353 |
+
"grad_norm": 0.23652639985084534,
|
| 1354 |
+
"learning_rate": 4.141435421577493e-05,
|
| 1355 |
+
"loss": 0.1507,
|
| 1356 |
+
"step": 1890
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 0.8895131086142322,
|
| 1360 |
+
"grad_norm": 0.2574213445186615,
|
| 1361 |
+
"learning_rate": 4.123255041106788e-05,
|
| 1362 |
+
"loss": 0.1612,
|
| 1363 |
+
"step": 1900
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 0.8941947565543071,
|
| 1367 |
+
"grad_norm": 0.25920340418815613,
|
| 1368 |
+
"learning_rate": 4.104925017576458e-05,
|
| 1369 |
+
"loss": 0.1483,
|
| 1370 |
+
"step": 1910
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.898876404494382,
|
| 1374 |
+
"grad_norm": 0.26211079955101013,
|
| 1375 |
+
"learning_rate": 4.0864470407769114e-05,
|
| 1376 |
+
"loss": 0.1652,
|
| 1377 |
+
"step": 1920
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.903558052434457,
|
| 1381 |
+
"grad_norm": 0.27797001600265503,
|
| 1382 |
+
"learning_rate": 4.0678228141379336e-05,
|
| 1383 |
+
"loss": 0.1461,
|
| 1384 |
+
"step": 1930
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 0.9082397003745318,
|
| 1388 |
+
"grad_norm": 0.2917916774749756,
|
| 1389 |
+
"learning_rate": 4.049054054571648e-05,
|
| 1390 |
+
"loss": 0.1581,
|
| 1391 |
+
"step": 1940
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 0.9129213483146067,
|
| 1395 |
+
"grad_norm": 0.2321026623249054,
|
| 1396 |
+
"learning_rate": 4.030142492314236e-05,
|
| 1397 |
+
"loss": 0.1754,
|
| 1398 |
+
"step": 1950
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 0.9176029962546817,
|
| 1402 |
+
"grad_norm": 0.22449631989002228,
|
| 1403 |
+
"learning_rate": 4.011089870766437e-05,
|
| 1404 |
+
"loss": 0.1423,
|
| 1405 |
+
"step": 1960
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 0.9222846441947565,
|
| 1409 |
+
"grad_norm": 0.2413541078567505,
|
| 1410 |
+
"learning_rate": 3.991897946332829e-05,
|
| 1411 |
+
"loss": 0.1619,
|
| 1412 |
+
"step": 1970
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 0.9269662921348315,
|
| 1416 |
+
"grad_norm": 0.23216186463832855,
|
| 1417 |
+
"learning_rate": 3.972568488259905e-05,
|
| 1418 |
+
"loss": 0.1549,
|
| 1419 |
+
"step": 1980
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.9316479400749064,
|
| 1423 |
+
"grad_norm": 0.2707996070384979,
|
| 1424 |
+
"learning_rate": 3.953103278472979e-05,
|
| 1425 |
+
"loss": 0.1622,
|
| 1426 |
+
"step": 1990
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 0.9363295880149812,
|
| 1430 |
+
"grad_norm": 0.25455233454704285,
|
| 1431 |
+
"learning_rate": 3.93350411141191e-05,
|
| 1432 |
+
"loss": 0.1572,
|
| 1433 |
+
"step": 2000
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 0.9363295880149812,
|
| 1437 |
+
"eval_loss": 0.15146031975746155,
|
| 1438 |
+
"eval_runtime": 232.8082,
|
| 1439 |
+
"eval_samples_per_second": 2.998,
|
| 1440 |
+
"eval_steps_per_second": 1.499,
|
| 1441 |
+
"step": 2000
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.9410112359550562,
|
| 1445 |
+
"grad_norm": 0.25234970450401306,
|
| 1446 |
+
"learning_rate": 3.913772793865683e-05,
|
| 1447 |
+
"loss": 0.1552,
|
| 1448 |
+
"step": 2010
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 0.9456928838951311,
|
| 1452 |
+
"grad_norm": 0.2573593556880951,
|
| 1453 |
+
"learning_rate": 3.8939111448058404e-05,
|
| 1454 |
+
"loss": 0.163,
|
| 1455 |
+
"step": 2020
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 0.950374531835206,
|
| 1459 |
+
"grad_norm": 0.2533682584762573,
|
| 1460 |
+
"learning_rate": 3.8739209952188016e-05,
|
| 1461 |
+
"loss": 0.1647,
|
| 1462 |
+
"step": 2030
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 0.9550561797752809,
|
| 1466 |
+
"grad_norm": 0.2530724108219147,
|
| 1467 |
+
"learning_rate": 3.853804187937066e-05,
|
| 1468 |
+
"loss": 0.1455,
|
| 1469 |
+
"step": 2040
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 0.9597378277153558,
|
| 1473 |
+
"grad_norm": 0.2659183144569397,
|
| 1474 |
+
"learning_rate": 3.8335625774693326e-05,
|
| 1475 |
+
"loss": 0.1556,
|
| 1476 |
+
"step": 2050
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 0.9644194756554307,
|
| 1480 |
+
"grad_norm": 0.2551698386669159,
|
| 1481 |
+
"learning_rate": 3.813198029829532e-05,
|
| 1482 |
+
"loss": 0.1609,
|
| 1483 |
+
"step": 2060
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 0.9691011235955056,
|
| 1487 |
+
"grad_norm": 0.24625395238399506,
|
| 1488 |
+
"learning_rate": 3.792712422364812e-05,
|
| 1489 |
+
"loss": 0.1491,
|
| 1490 |
+
"step": 2070
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 0.9737827715355806,
|
| 1494 |
+
"grad_norm": 0.28316351771354675,
|
| 1495 |
+
"learning_rate": 3.772107643582459e-05,
|
| 1496 |
+
"loss": 0.1549,
|
| 1497 |
+
"step": 2080
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 0.9784644194756554,
|
| 1501 |
+
"grad_norm": 0.2540896236896515,
|
| 1502 |
+
"learning_rate": 3.7513855929758146e-05,
|
| 1503 |
+
"loss": 0.1497,
|
| 1504 |
+
"step": 2090
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 0.9831460674157303,
|
| 1508 |
+
"grad_norm": 0.29905152320861816,
|
| 1509 |
+
"learning_rate": 3.730548180849161e-05,
|
| 1510 |
+
"loss": 0.162,
|
| 1511 |
+
"step": 2100
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 0.9878277153558053,
|
| 1515 |
+
"grad_norm": 0.26653796434402466,
|
| 1516 |
+
"learning_rate": 3.7095973281416144e-05,
|
| 1517 |
+
"loss": 0.1606,
|
| 1518 |
+
"step": 2110
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 0.9925093632958801,
|
| 1522 |
+
"grad_norm": 0.22521516680717468,
|
| 1523 |
+
"learning_rate": 3.688534966250042e-05,
|
| 1524 |
+
"loss": 0.1494,
|
| 1525 |
+
"step": 2120
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 0.9971910112359551,
|
| 1529 |
+
"grad_norm": 0.2566962242126465,
|
| 1530 |
+
"learning_rate": 3.667363036851012e-05,
|
| 1531 |
+
"loss": 0.1512,
|
| 1532 |
+
"step": 2130
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 1.0018726591760299,
|
| 1536 |
+
"grad_norm": 0.25175896286964417,
|
| 1537 |
+
"learning_rate": 3.646083491721794e-05,
|
| 1538 |
+
"loss": 0.14,
|
| 1539 |
+
"step": 2140
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 1.006554307116105,
|
| 1543 |
+
"grad_norm": 0.24953345954418182,
|
| 1544 |
+
"learning_rate": 3.624698292560434e-05,
|
| 1545 |
+
"loss": 0.1563,
|
| 1546 |
+
"step": 2150
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 1.0112359550561798,
|
| 1550 |
+
"grad_norm": 0.26830434799194336,
|
| 1551 |
+
"learning_rate": 3.603209410804906e-05,
|
| 1552 |
+
"loss": 0.1476,
|
| 1553 |
+
"step": 2160
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 1.0159176029962547,
|
| 1557 |
+
"grad_norm": 0.23503483831882477,
|
| 1558 |
+
"learning_rate": 3.5816188274513795e-05,
|
| 1559 |
+
"loss": 0.1487,
|
| 1560 |
+
"step": 2170
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 1.0205992509363295,
|
| 1564 |
+
"grad_norm": 0.24020551145076752,
|
| 1565 |
+
"learning_rate": 3.559928532871587e-05,
|
| 1566 |
+
"loss": 0.1464,
|
| 1567 |
+
"step": 2180
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 1.0252808988764044,
|
| 1571 |
+
"grad_norm": 0.2504616975784302,
|
| 1572 |
+
"learning_rate": 3.5381405266293464e-05,
|
| 1573 |
+
"loss": 0.152,
|
| 1574 |
+
"step": 2190
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 1.0299625468164795,
|
| 1578 |
+
"grad_norm": 0.2568669319152832,
|
| 1579 |
+
"learning_rate": 3.516256817296222e-05,
|
| 1580 |
+
"loss": 0.1534,
|
| 1581 |
+
"step": 2200
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 1.0346441947565543,
|
| 1585 |
+
"grad_norm": 0.2495841681957245,
|
| 1586 |
+
"learning_rate": 3.4942794222663614e-05,
|
| 1587 |
+
"loss": 0.148,
|
| 1588 |
+
"step": 2210
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 1.0393258426966292,
|
| 1592 |
+
"grad_norm": 0.2616524398326874,
|
| 1593 |
+
"learning_rate": 3.472210367570518e-05,
|
| 1594 |
+
"loss": 0.1498,
|
| 1595 |
+
"step": 2220
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 1.044007490636704,
|
| 1599 |
+
"grad_norm": 0.2961607277393341,
|
| 1600 |
+
"learning_rate": 3.4500516876892786e-05,
|
| 1601 |
+
"loss": 0.154,
|
| 1602 |
+
"step": 2230
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 1.048689138576779,
|
| 1606 |
+
"grad_norm": 0.26527124643325806,
|
| 1607 |
+
"learning_rate": 3.427805425365509e-05,
|
| 1608 |
+
"loss": 0.1471,
|
| 1609 |
+
"step": 2240
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 1.053370786516854,
|
| 1613 |
+
"grad_norm": 0.2571045458316803,
|
| 1614 |
+
"learning_rate": 3.405473631416041e-05,
|
| 1615 |
+
"loss": 0.15,
|
| 1616 |
+
"step": 2250
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 1.0580524344569289,
|
| 1620 |
+
"grad_norm": 0.272733598947525,
|
| 1621 |
+
"learning_rate": 3.383058364542611e-05,
|
| 1622 |
+
"loss": 0.1503,
|
| 1623 |
+
"step": 2260
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 1.0627340823970037,
|
| 1627 |
+
"grad_norm": 0.24832037091255188,
|
| 1628 |
+
"learning_rate": 3.3605616911420786e-05,
|
| 1629 |
+
"loss": 0.1462,
|
| 1630 |
+
"step": 2270
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 1.0674157303370786,
|
| 1634 |
+
"grad_norm": 0.2702969014644623,
|
| 1635 |
+
"learning_rate": 3.3379856851159267e-05,
|
| 1636 |
+
"loss": 0.1433,
|
| 1637 |
+
"step": 2280
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 1.0720973782771535,
|
| 1641 |
+
"grad_norm": 0.2763786315917969,
|
| 1642 |
+
"learning_rate": 3.315332427679079e-05,
|
| 1643 |
+
"loss": 0.1425,
|
| 1644 |
+
"step": 2290
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 1.0767790262172285,
|
| 1648 |
+
"grad_norm": 0.3010675609111786,
|
| 1649 |
+
"learning_rate": 3.292604007168037e-05,
|
| 1650 |
+
"loss": 0.1525,
|
| 1651 |
+
"step": 2300
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 1.0814606741573034,
|
| 1655 |
+
"grad_norm": 0.25726327300071716,
|
| 1656 |
+
"learning_rate": 3.269802518848364e-05,
|
| 1657 |
+
"loss": 0.1501,
|
| 1658 |
+
"step": 2310
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 1.0861423220973783,
|
| 1662 |
+
"grad_norm": 0.24328449368476868,
|
| 1663 |
+
"learning_rate": 3.246930064721524e-05,
|
| 1664 |
+
"loss": 0.1352,
|
| 1665 |
+
"step": 2320
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 1.0908239700374531,
|
| 1669 |
+
"grad_norm": 0.2798154056072235,
|
| 1670 |
+
"learning_rate": 3.22398875333111e-05,
|
| 1671 |
+
"loss": 0.1492,
|
| 1672 |
+
"step": 2330
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 1.095505617977528,
|
| 1676 |
+
"grad_norm": 0.26386189460754395,
|
| 1677 |
+
"learning_rate": 3.200980699568463e-05,
|
| 1678 |
+
"loss": 0.1472,
|
| 1679 |
+
"step": 2340
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 1.100187265917603,
|
| 1683 |
+
"grad_norm": 0.2722947895526886,
|
| 1684 |
+
"learning_rate": 3.177908024477703e-05,
|
| 1685 |
+
"loss": 0.1557,
|
| 1686 |
+
"step": 2350
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 1.104868913857678,
|
| 1690 |
+
"grad_norm": 0.2719181180000305,
|
| 1691 |
+
"learning_rate": 3.154772855060198e-05,
|
| 1692 |
+
"loss": 0.1408,
|
| 1693 |
+
"step": 2360
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 1.1095505617977528,
|
| 1697 |
+
"grad_norm": 0.27931663393974304,
|
| 1698 |
+
"learning_rate": 3.131577324078485e-05,
|
| 1699 |
+
"loss": 0.1485,
|
| 1700 |
+
"step": 2370
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 1.1142322097378277,
|
| 1704 |
+
"grad_norm": 0.2354390025138855,
|
| 1705 |
+
"learning_rate": 3.1083235698596505e-05,
|
| 1706 |
+
"loss": 0.1402,
|
| 1707 |
+
"step": 2380
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 1.1189138576779025,
|
| 1711 |
+
"grad_norm": 0.25632205605506897,
|
| 1712 |
+
"learning_rate": 3.0850137360982104e-05,
|
| 1713 |
+
"loss": 0.1428,
|
| 1714 |
+
"step": 2390
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 1.1235955056179776,
|
| 1718 |
+
"grad_norm": 0.26264774799346924,
|
| 1719 |
+
"learning_rate": 3.061649971658488e-05,
|
| 1720 |
+
"loss": 0.1465,
|
| 1721 |
+
"step": 2400
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 1.1282771535580525,
|
| 1725 |
+
"grad_norm": 0.2421567738056183,
|
| 1726 |
+
"learning_rate": 3.038234430376513e-05,
|
| 1727 |
+
"loss": 0.1495,
|
| 1728 |
+
"step": 2410
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 1.1329588014981273,
|
| 1732 |
+
"grad_norm": 0.2859581410884857,
|
| 1733 |
+
"learning_rate": 3.01476927086147e-05,
|
| 1734 |
+
"loss": 0.1494,
|
| 1735 |
+
"step": 2420
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 1.1376404494382022,
|
| 1739 |
+
"grad_norm": 0.24564145505428314,
|
| 1740 |
+
"learning_rate": 2.9912566562967014e-05,
|
| 1741 |
+
"loss": 0.15,
|
| 1742 |
+
"step": 2430
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 1.142322097378277,
|
| 1746 |
+
"grad_norm": 0.23918387293815613,
|
| 1747 |
+
"learning_rate": 2.967698754240289e-05,
|
| 1748 |
+
"loss": 0.146,
|
| 1749 |
+
"step": 2440
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 1.1470037453183521,
|
| 1753 |
+
"grad_norm": 0.2613247334957123,
|
| 1754 |
+
"learning_rate": 2.944097736425233e-05,
|
| 1755 |
+
"loss": 0.1549,
|
| 1756 |
+
"step": 2450
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 1.151685393258427,
|
| 1760 |
+
"grad_norm": 0.3330237865447998,
|
| 1761 |
+
"learning_rate": 2.92045577855925e-05,
|
| 1762 |
+
"loss": 0.1526,
|
| 1763 |
+
"step": 2460
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 1.1563670411985019,
|
| 1767 |
+
"grad_norm": 0.24248412251472473,
|
| 1768 |
+
"learning_rate": 2.896775060124195e-05,
|
| 1769 |
+
"loss": 0.1537,
|
| 1770 |
+
"step": 2470
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 1.1610486891385767,
|
| 1774 |
+
"grad_norm": 0.2552735507488251,
|
| 1775 |
+
"learning_rate": 2.8730577641751476e-05,
|
| 1776 |
+
"loss": 0.1458,
|
| 1777 |
+
"step": 2480
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 1.1657303370786516,
|
| 1781 |
+
"grad_norm": 0.24274031817913055,
|
| 1782 |
+
"learning_rate": 2.8493060771391567e-05,
|
| 1783 |
+
"loss": 0.1477,
|
| 1784 |
+
"step": 2490
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 1.1704119850187267,
|
| 1788 |
+
"grad_norm": 0.251447468996048,
|
| 1789 |
+
"learning_rate": 2.825522188613686e-05,
|
| 1790 |
+
"loss": 0.142,
|
| 1791 |
+
"step": 2500
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 1.1704119850187267,
|
| 1795 |
+
"eval_loss": 0.14719446003437042,
|
| 1796 |
+
"eval_runtime": 235.8012,
|
| 1797 |
+
"eval_samples_per_second": 2.96,
|
| 1798 |
+
"eval_steps_per_second": 1.48,
|
| 1799 |
+
"step": 2500
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 1.1750936329588015,
|
| 1803 |
+
"grad_norm": 0.31371602416038513,
|
| 1804 |
+
"learning_rate": 2.8017082911647586e-05,
|
| 1805 |
+
"loss": 0.1435,
|
| 1806 |
+
"step": 2510
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 1.1797752808988764,
|
| 1810 |
+
"grad_norm": 0.25653213262557983,
|
| 1811 |
+
"learning_rate": 2.7778665801248292e-05,
|
| 1812 |
+
"loss": 0.1445,
|
| 1813 |
+
"step": 2520
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 1.1844569288389513,
|
| 1817 |
+
"grad_norm": 0.26454073190689087,
|
| 1818 |
+
"learning_rate": 2.7539992533904062e-05,
|
| 1819 |
+
"loss": 0.1472,
|
| 1820 |
+
"step": 2530
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 1.1891385767790261,
|
| 1824 |
+
"grad_norm": 0.2490745633840561,
|
| 1825 |
+
"learning_rate": 2.730108511219433e-05,
|
| 1826 |
+
"loss": 0.1728,
|
| 1827 |
+
"step": 2540
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 1.1938202247191012,
|
| 1831 |
+
"grad_norm": 0.2763822674751282,
|
| 1832 |
+
"learning_rate": 2.7061965560284512e-05,
|
| 1833 |
+
"loss": 0.1466,
|
| 1834 |
+
"step": 2550
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 1.198501872659176,
|
| 1838 |
+
"grad_norm": 0.26875513792037964,
|
| 1839 |
+
"learning_rate": 2.6822655921895695e-05,
|
| 1840 |
+
"loss": 0.1548,
|
| 1841 |
+
"step": 2560
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 1.203183520599251,
|
| 1845 |
+
"grad_norm": 0.2581767737865448,
|
| 1846 |
+
"learning_rate": 2.6583178258272468e-05,
|
| 1847 |
+
"loss": 0.1508,
|
| 1848 |
+
"step": 2570
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 1.2078651685393258,
|
| 1852 |
+
"grad_norm": 0.24231292307376862,
|
| 1853 |
+
"learning_rate": 2.6343554646149154e-05,
|
| 1854 |
+
"loss": 0.1523,
|
| 1855 |
+
"step": 2580
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 1.2125468164794007,
|
| 1859 |
+
"grad_norm": 0.2455371618270874,
|
| 1860 |
+
"learning_rate": 2.6103807175714645e-05,
|
| 1861 |
+
"loss": 0.1417,
|
| 1862 |
+
"step": 2590
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 1.2172284644194757,
|
| 1866 |
+
"grad_norm": 0.21632102131843567,
|
| 1867 |
+
"learning_rate": 2.586395794857597e-05,
|
| 1868 |
+
"loss": 0.1471,
|
| 1869 |
+
"step": 2600
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 1.2219101123595506,
|
| 1873 |
+
"grad_norm": 0.24456676840782166,
|
| 1874 |
+
"learning_rate": 2.5624029075720785e-05,
|
| 1875 |
+
"loss": 0.1411,
|
| 1876 |
+
"step": 2610
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 1.2265917602996255,
|
| 1880 |
+
"grad_norm": 0.25774601101875305,
|
| 1881 |
+
"learning_rate": 2.538404267547908e-05,
|
| 1882 |
+
"loss": 0.1471,
|
| 1883 |
+
"step": 2620
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 1.2312734082397003,
|
| 1887 |
+
"grad_norm": 0.26534947752952576,
|
| 1888 |
+
"learning_rate": 2.5144020871484088e-05,
|
| 1889 |
+
"loss": 0.1538,
|
| 1890 |
+
"step": 2630
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 1.2359550561797752,
|
| 1894 |
+
"grad_norm": 0.2735100984573364,
|
| 1895 |
+
"learning_rate": 2.4903985790632835e-05,
|
| 1896 |
+
"loss": 0.1492,
|
| 1897 |
+
"step": 2640
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 1.2406367041198503,
|
| 1901 |
+
"grad_norm": 0.2741684317588806,
|
| 1902 |
+
"learning_rate": 2.46639595610463e-05,
|
| 1903 |
+
"loss": 0.1464,
|
| 1904 |
+
"step": 2650
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 1.2453183520599251,
|
| 1908 |
+
"grad_norm": 0.233806774020195,
|
| 1909 |
+
"learning_rate": 2.4423964310029458e-05,
|
| 1910 |
+
"loss": 0.1404,
|
| 1911 |
+
"step": 2660
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 1.25,
|
| 1915 |
+
"grad_norm": 0.25946834683418274,
|
| 1916 |
+
"learning_rate": 2.4184022162031502e-05,
|
| 1917 |
+
"loss": 0.1474,
|
| 1918 |
+
"step": 2670
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 1.2546816479400749,
|
| 1922 |
+
"grad_norm": 0.2744329869747162,
|
| 1923 |
+
"learning_rate": 2.3944155236606196e-05,
|
| 1924 |
+
"loss": 0.1525,
|
| 1925 |
+
"step": 2680
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 1.2593632958801497,
|
| 1929 |
+
"grad_norm": 0.27152568101882935,
|
| 1930 |
+
"learning_rate": 2.3704385646372772e-05,
|
| 1931 |
+
"loss": 0.1454,
|
| 1932 |
+
"step": 2690
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 1.2640449438202248,
|
| 1936 |
+
"grad_norm": 0.23862288892269135,
|
| 1937 |
+
"learning_rate": 2.3464735494977392e-05,
|
| 1938 |
+
"loss": 0.1311,
|
| 1939 |
+
"step": 2700
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 1.2687265917602997,
|
| 1943 |
+
"grad_norm": 0.27397608757019043,
|
| 1944 |
+
"learning_rate": 2.3225226875055535e-05,
|
| 1945 |
+
"loss": 0.1416,
|
| 1946 |
+
"step": 2710
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 1.2734082397003745,
|
| 1950 |
+
"grad_norm": 0.2350151538848877,
|
| 1951 |
+
"learning_rate": 2.2985881866195307e-05,
|
| 1952 |
+
"loss": 0.1386,
|
| 1953 |
+
"step": 2720
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 1.2780898876404494,
|
| 1957 |
+
"grad_norm": 0.26997503638267517,
|
| 1958 |
+
"learning_rate": 2.274672253290198e-05,
|
| 1959 |
+
"loss": 0.1412,
|
| 1960 |
+
"step": 2730
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 1.2827715355805243,
|
| 1964 |
+
"grad_norm": 0.2512725591659546,
|
| 1965 |
+
"learning_rate": 2.2507770922563966e-05,
|
| 1966 |
+
"loss": 0.1583,
|
| 1967 |
+
"step": 2740
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 1.2874531835205993,
|
| 1971 |
+
"grad_norm": 0.27003416419029236,
|
| 1972 |
+
"learning_rate": 2.2269049063420305e-05,
|
| 1973 |
+
"loss": 0.1553,
|
| 1974 |
+
"step": 2750
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 1.2921348314606742,
|
| 1978 |
+
"grad_norm": 0.25606778264045715,
|
| 1979 |
+
"learning_rate": 2.2030578962529964e-05,
|
| 1980 |
+
"loss": 0.1358,
|
| 1981 |
+
"step": 2760
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 1.296816479400749,
|
| 1985 |
+
"grad_norm": 0.2931346297264099,
|
| 1986 |
+
"learning_rate": 2.179238260374307e-05,
|
| 1987 |
+
"loss": 0.1461,
|
| 1988 |
+
"step": 2770
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 1.301498127340824,
|
| 1992 |
+
"grad_norm": 0.2787070572376251,
|
| 1993 |
+
"learning_rate": 2.155448194567425e-05,
|
| 1994 |
+
"loss": 0.1412,
|
| 1995 |
+
"step": 2780
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 1.3061797752808988,
|
| 1999 |
+
"grad_norm": 0.2558610737323761,
|
| 2000 |
+
"learning_rate": 2.131689891967839e-05,
|
| 2001 |
+
"loss": 0.1447,
|
| 2002 |
+
"step": 2790
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 1.3108614232209739,
|
| 2006 |
+
"grad_norm": 0.27285927534103394,
|
| 2007 |
+
"learning_rate": 2.1079655427828807e-05,
|
| 2008 |
+
"loss": 0.1315,
|
| 2009 |
+
"step": 2800
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 1.3155430711610487,
|
| 2013 |
+
"grad_norm": 0.2873709797859192,
|
| 2014 |
+
"learning_rate": 2.0842773340898187e-05,
|
| 2015 |
+
"loss": 0.1404,
|
| 2016 |
+
"step": 2810
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 1.3202247191011236,
|
| 2020 |
+
"grad_norm": 0.2794095277786255,
|
| 2021 |
+
"learning_rate": 2.060627449634234e-05,
|
| 2022 |
+
"loss": 0.1473,
|
| 2023 |
+
"step": 2820
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 1.3249063670411985,
|
| 2027 |
+
"grad_norm": 0.2690732777118683,
|
| 2028 |
+
"learning_rate": 2.0370180696287123e-05,
|
| 2029 |
+
"loss": 0.1446,
|
| 2030 |
+
"step": 2830
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 1.3295880149812733,
|
| 2034 |
+
"grad_norm": 0.28744590282440186,
|
| 2035 |
+
"learning_rate": 2.0134513705518547e-05,
|
| 2036 |
+
"loss": 0.145,
|
| 2037 |
+
"step": 2840
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 1.3342696629213484,
|
| 2041 |
+
"grad_norm": 0.2891230285167694,
|
| 2042 |
+
"learning_rate": 1.9899295249476337e-05,
|
| 2043 |
+
"loss": 0.1437,
|
| 2044 |
+
"step": 2850
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 1.3389513108614233,
|
| 2048 |
+
"grad_norm": 0.28290775418281555,
|
| 2049 |
+
"learning_rate": 1.9664547012251122e-05,
|
| 2050 |
+
"loss": 0.1425,
|
| 2051 |
+
"step": 2860
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 1.3436329588014981,
|
| 2055 |
+
"grad_norm": 0.2865421772003174,
|
| 2056 |
+
"learning_rate": 1.9430290634585475e-05,
|
| 2057 |
+
"loss": 0.1501,
|
| 2058 |
+
"step": 2870
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 1.348314606741573,
|
| 2062 |
+
"grad_norm": 0.27473053336143494,
|
| 2063 |
+
"learning_rate": 1.9196547711878883e-05,
|
| 2064 |
+
"loss": 0.1486,
|
| 2065 |
+
"step": 2880
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 1.3529962546816479,
|
| 2069 |
+
"grad_norm": 0.2682128846645355,
|
| 2070 |
+
"learning_rate": 1.8963339792196968e-05,
|
| 2071 |
+
"loss": 0.1429,
|
| 2072 |
+
"step": 2890
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 1.357677902621723,
|
| 2076 |
+
"grad_norm": 0.2450461983680725,
|
| 2077 |
+
"learning_rate": 1.873068837428497e-05,
|
| 2078 |
+
"loss": 0.1539,
|
| 2079 |
+
"step": 2900
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 1.3623595505617978,
|
| 2083 |
+
"grad_norm": 0.25303879380226135,
|
| 2084 |
+
"learning_rate": 1.8498614905585924e-05,
|
| 2085 |
+
"loss": 0.1465,
|
| 2086 |
+
"step": 2910
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 1.3670411985018727,
|
| 2090 |
+
"grad_norm": 0.27826565504074097,
|
| 2091 |
+
"learning_rate": 1.8267140780263427e-05,
|
| 2092 |
+
"loss": 0.1455,
|
| 2093 |
+
"step": 2920
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 1.3717228464419475,
|
| 2097 |
+
"grad_norm": 0.20451192557811737,
|
| 2098 |
+
"learning_rate": 1.8036287337229407e-05,
|
| 2099 |
+
"loss": 0.1572,
|
| 2100 |
+
"step": 2930
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 1.3764044943820224,
|
| 2104 |
+
"grad_norm": 0.2710937261581421,
|
| 2105 |
+
"learning_rate": 1.7806075858176903e-05,
|
| 2106 |
+
"loss": 0.1427,
|
| 2107 |
+
"step": 2940
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 1.3810861423220975,
|
| 2111 |
+
"grad_norm": 0.2961539626121521,
|
| 2112 |
+
"learning_rate": 1.757652756561822e-05,
|
| 2113 |
+
"loss": 0.1481,
|
| 2114 |
+
"step": 2950
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 1.3857677902621723,
|
| 2118 |
+
"grad_norm": 0.2685253620147705,
|
| 2119 |
+
"learning_rate": 1.7347663620928495e-05,
|
| 2120 |
+
"loss": 0.1488,
|
| 2121 |
+
"step": 2960
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 1.3904494382022472,
|
| 2125 |
+
"grad_norm": 0.29267454147338867,
|
| 2126 |
+
"learning_rate": 1.711950512239483e-05,
|
| 2127 |
+
"loss": 0.1441,
|
| 2128 |
+
"step": 2970
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 1.395131086142322,
|
| 2132 |
+
"grad_norm": 0.25182071328163147,
|
| 2133 |
+
"learning_rate": 1.6892073103271355e-05,
|
| 2134 |
+
"loss": 0.1422,
|
| 2135 |
+
"step": 2980
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 1.399812734082397,
|
| 2139 |
+
"grad_norm": 0.23954342305660248,
|
| 2140 |
+
"learning_rate": 1.6665388529840225e-05,
|
| 2141 |
+
"loss": 0.1465,
|
| 2142 |
+
"step": 2990
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 1.404494382022472,
|
| 2146 |
+
"grad_norm": 0.2718564569950104,
|
| 2147 |
+
"learning_rate": 1.6439472299478803e-05,
|
| 2148 |
+
"loss": 0.1376,
|
| 2149 |
+
"step": 3000
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 1.404494382022472,
|
| 2153 |
+
"eval_loss": 0.14412875473499298,
|
| 2154 |
+
"eval_runtime": 230.7877,
|
| 2155 |
+
"eval_samples_per_second": 3.024,
|
| 2156 |
+
"eval_steps_per_second": 1.512,
|
| 2157 |
+
"step": 3000
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 1.4091760299625469,
|
| 2161 |
+
"grad_norm": 0.2633533477783203,
|
| 2162 |
+
"learning_rate": 1.621434523873318e-05,
|
| 2163 |
+
"loss": 0.1351,
|
| 2164 |
+
"step": 3010
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 1.4138576779026217,
|
| 2168 |
+
"grad_norm": 0.26720061898231506,
|
| 2169 |
+
"learning_rate": 1.5990028101398234e-05,
|
| 2170 |
+
"loss": 0.1382,
|
| 2171 |
+
"step": 3020
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 1.4185393258426966,
|
| 2175 |
+
"grad_norm": 0.26572489738464355,
|
| 2176 |
+
"learning_rate": 1.576654156660442e-05,
|
| 2177 |
+
"loss": 0.1462,
|
| 2178 |
+
"step": 3030
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 1.4232209737827715,
|
| 2182 |
+
"grad_norm": 0.29276156425476074,
|
| 2183 |
+
"learning_rate": 1.5543906236911424e-05,
|
| 2184 |
+
"loss": 0.1518,
|
| 2185 |
+
"step": 3040
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 1.4279026217228465,
|
| 2189 |
+
"grad_norm": 0.25354310870170593,
|
| 2190 |
+
"learning_rate": 1.5322142636408855e-05,
|
| 2191 |
+
"loss": 0.1727,
|
| 2192 |
+
"step": 3050
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 1.4325842696629214,
|
| 2196 |
+
"grad_norm": 0.2621780037879944,
|
| 2197 |
+
"learning_rate": 1.5101271208824168e-05,
|
| 2198 |
+
"loss": 0.1414,
|
| 2199 |
+
"step": 3060
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 1.4372659176029963,
|
| 2203 |
+
"grad_norm": 0.275648295879364,
|
| 2204 |
+
"learning_rate": 1.4881312315638097e-05,
|
| 2205 |
+
"loss": 0.1387,
|
| 2206 |
+
"step": 3070
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 1.4419475655430711,
|
| 2210 |
+
"grad_norm": 0.25782111287117004,
|
| 2211 |
+
"learning_rate": 1.4662286234207512e-05,
|
| 2212 |
+
"loss": 0.1471,
|
| 2213 |
+
"step": 3080
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 1.446629213483146,
|
| 2217 |
+
"grad_norm": 0.25094345211982727,
|
| 2218 |
+
"learning_rate": 1.4444213155896147e-05,
|
| 2219 |
+
"loss": 0.1523,
|
| 2220 |
+
"step": 3090
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 1.451310861423221,
|
| 2224 |
+
"grad_norm": 0.2622791528701782,
|
| 2225 |
+
"learning_rate": 1.4227113184213198e-05,
|
| 2226 |
+
"loss": 0.1413,
|
| 2227 |
+
"step": 3100
|
| 2228 |
+
},
|
| 2229 |
+
{
|
| 2230 |
+
"epoch": 1.455992509363296,
|
| 2231 |
+
"grad_norm": 0.3153008222579956,
|
| 2232 |
+
"learning_rate": 1.4011006332960064e-05,
|
| 2233 |
+
"loss": 0.1458,
|
| 2234 |
+
"step": 3110
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 1.4606741573033708,
|
| 2238 |
+
"grad_norm": 0.247295081615448,
|
| 2239 |
+
"learning_rate": 1.3795912524385323e-05,
|
| 2240 |
+
"loss": 0.1514,
|
| 2241 |
+
"step": 3120
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 1.4653558052434457,
|
| 2245 |
+
"grad_norm": 0.266445517539978,
|
| 2246 |
+
"learning_rate": 1.358185158734816e-05,
|
| 2247 |
+
"loss": 0.1443,
|
| 2248 |
+
"step": 3130
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 1.4700374531835205,
|
| 2252 |
+
"grad_norm": 0.28155022859573364,
|
| 2253 |
+
"learning_rate": 1.3368843255490383e-05,
|
| 2254 |
+
"loss": 0.1431,
|
| 2255 |
+
"step": 3140
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 1.4747191011235956,
|
| 2259 |
+
"grad_norm": 0.25658926367759705,
|
| 2260 |
+
"learning_rate": 1.3156907165417271e-05,
|
| 2261 |
+
"loss": 0.1489,
|
| 2262 |
+
"step": 3150
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 1.4794007490636705,
|
| 2266 |
+
"grad_norm": 0.29060763120651245,
|
| 2267 |
+
"learning_rate": 1.2946062854887313e-05,
|
| 2268 |
+
"loss": 0.161,
|
| 2269 |
+
"step": 3160
|
| 2270 |
+
},
|
| 2271 |
+
{
|
| 2272 |
+
"epoch": 1.4840823970037453,
|
| 2273 |
+
"grad_norm": 0.2649938464164734,
|
| 2274 |
+
"learning_rate": 1.273632976101111e-05,
|
| 2275 |
+
"loss": 0.1429,
|
| 2276 |
+
"step": 3170
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 1.4887640449438202,
|
| 2280 |
+
"grad_norm": 0.28675931692123413,
|
| 2281 |
+
"learning_rate": 1.252772721845945e-05,
|
| 2282 |
+
"loss": 0.1519,
|
| 2283 |
+
"step": 3180
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 1.493445692883895,
|
| 2287 |
+
"grad_norm": 0.2792210876941681,
|
| 2288 |
+
"learning_rate": 1.2320274457680997e-05,
|
| 2289 |
+
"loss": 0.1464,
|
| 2290 |
+
"step": 3190
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 1.4981273408239701,
|
| 2294 |
+
"grad_norm": 0.29543542861938477,
|
| 2295 |
+
"learning_rate": 1.2113990603129433e-05,
|
| 2296 |
+
"loss": 0.1463,
|
| 2297 |
+
"step": 3200
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 1.5028089887640448,
|
| 2301 |
+
"grad_norm": 0.2720608711242676,
|
| 2302 |
+
"learning_rate": 1.1908894671500459e-05,
|
| 2303 |
+
"loss": 0.1397,
|
| 2304 |
+
"step": 3210
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 1.5074906367041199,
|
| 2308 |
+
"grad_norm": 0.2443593591451645,
|
| 2309 |
+
"learning_rate": 1.1705005569978664e-05,
|
| 2310 |
+
"loss": 0.1385,
|
| 2311 |
+
"step": 3220
|
| 2312 |
+
},
|
| 2313 |
+
{
|
| 2314 |
+
"epoch": 1.5121722846441947,
|
| 2315 |
+
"grad_norm": 0.26821592450141907,
|
| 2316 |
+
"learning_rate": 1.1502342094494584e-05,
|
| 2317 |
+
"loss": 0.1327,
|
| 2318 |
+
"step": 3230
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 1.5168539325842696,
|
| 2322 |
+
"grad_norm": 0.2925826609134674,
|
| 2323 |
+
"learning_rate": 1.1300922927991913e-05,
|
| 2324 |
+
"loss": 0.143,
|
| 2325 |
+
"step": 3240
|
| 2326 |
+
},
|
| 2327 |
+
{
|
| 2328 |
+
"epoch": 1.5215355805243447,
|
| 2329 |
+
"grad_norm": 0.3027312457561493,
|
| 2330 |
+
"learning_rate": 1.1100766638705224e-05,
|
| 2331 |
+
"loss": 0.1481,
|
| 2332 |
+
"step": 3250
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 1.5262172284644193,
|
| 2336 |
+
"grad_norm": 0.2656176686286926,
|
| 2337 |
+
"learning_rate": 1.0901891678448144e-05,
|
| 2338 |
+
"loss": 0.1357,
|
| 2339 |
+
"step": 3260
|
| 2340 |
+
},
|
| 2341 |
+
{
|
| 2342 |
+
"epoch": 1.5308988764044944,
|
| 2343 |
+
"grad_norm": 0.29944488406181335,
|
| 2344 |
+
"learning_rate": 1.0704316380912425e-05,
|
| 2345 |
+
"loss": 0.1478,
|
| 2346 |
+
"step": 3270
|
| 2347 |
+
},
|
| 2348 |
+
{
|
| 2349 |
+
"epoch": 1.5355805243445693,
|
| 2350 |
+
"grad_norm": 0.2791348695755005,
|
| 2351 |
+
"learning_rate": 1.0508058959977757e-05,
|
| 2352 |
+
"loss": 0.1482,
|
| 2353 |
+
"step": 3280
|
| 2354 |
+
},
|
| 2355 |
+
{
|
| 2356 |
+
"epoch": 1.5402621722846441,
|
| 2357 |
+
"grad_norm": 0.29027488827705383,
|
| 2358 |
+
"learning_rate": 1.0313137508032719e-05,
|
| 2359 |
+
"loss": 0.1389,
|
| 2360 |
+
"step": 3290
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 1.5449438202247192,
|
| 2364 |
+
"grad_norm": 0.29706457257270813,
|
| 2365 |
+
"learning_rate": 1.0119569994306841e-05,
|
| 2366 |
+
"loss": 0.1531,
|
| 2367 |
+
"step": 3300
|
| 2368 |
+
},
|
| 2369 |
+
{
|
| 2370 |
+
"epoch": 1.5496254681647939,
|
| 2371 |
+
"grad_norm": 0.27216818928718567,
|
| 2372 |
+
"learning_rate": 9.927374263214143e-06,
|
| 2373 |
+
"loss": 0.146,
|
| 2374 |
+
"step": 3310
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 1.554307116104869,
|
| 2378 |
+
"grad_norm": 0.258273184299469,
|
| 2379 |
+
"learning_rate": 9.736568032708069e-06,
|
| 2380 |
+
"loss": 0.1441,
|
| 2381 |
+
"step": 3320
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 1.5589887640449438,
|
| 2385 |
+
"grad_norm": 0.2688770592212677,
|
| 2386 |
+
"learning_rate": 9.547168892648136e-06,
|
| 2387 |
+
"loss": 0.1551,
|
| 2388 |
+
"step": 3330
|
| 2389 |
+
},
|
| 2390 |
+
{
|
| 2391 |
+
"epoch": 1.5636704119850187,
|
| 2392 |
+
"grad_norm": 0.24575349688529968,
|
| 2393 |
+
"learning_rate": 9.359194303178371e-06,
|
| 2394 |
+
"loss": 0.1409,
|
| 2395 |
+
"step": 3340
|
| 2396 |
+
},
|
| 2397 |
+
{
|
| 2398 |
+
"epoch": 1.5683520599250937,
|
| 2399 |
+
"grad_norm": 0.2606309950351715,
|
| 2400 |
+
"learning_rate": 9.172661593117712e-06,
|
| 2401 |
+
"loss": 0.1535,
|
| 2402 |
+
"step": 3350
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 1.5730337078651684,
|
| 2406 |
+
"grad_norm": 0.2664453387260437,
|
| 2407 |
+
"learning_rate": 8.987587958362517e-06,
|
| 2408 |
+
"loss": 0.1403,
|
| 2409 |
+
"step": 3360
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 1.5777153558052435,
|
| 2413 |
+
"grad_norm": 0.24338924884796143,
|
| 2414 |
+
"learning_rate": 8.803990460301343e-06,
|
| 2415 |
+
"loss": 0.1454,
|
| 2416 |
+
"step": 3370
|
| 2417 |
+
},
|
| 2418 |
+
{
|
| 2419 |
+
"epoch": 1.5823970037453183,
|
| 2420 |
+
"grad_norm": 0.2727632522583008,
|
| 2421 |
+
"learning_rate": 8.621886024242057e-06,
|
| 2422 |
+
"loss": 0.1381,
|
| 2423 |
+
"step": 3380
|
| 2424 |
+
},
|
| 2425 |
+
{
|
| 2426 |
+
"epoch": 1.5870786516853932,
|
| 2427 |
+
"grad_norm": 0.27160879969596863,
|
| 2428 |
+
"learning_rate": 8.441291437851605e-06,
|
| 2429 |
+
"loss": 0.1386,
|
| 2430 |
+
"step": 3390
|
| 2431 |
+
},
|
| 2432 |
+
{
|
| 2433 |
+
"epoch": 1.5917602996254683,
|
| 2434 |
+
"grad_norm": 0.24165265262126923,
|
| 2435 |
+
"learning_rate": 8.262223349608366e-06,
|
| 2436 |
+
"loss": 0.1391,
|
| 2437 |
+
"step": 3400
|
| 2438 |
+
},
|
| 2439 |
+
{
|
| 2440 |
+
"epoch": 1.596441947565543,
|
| 2441 |
+
"grad_norm": 0.22898262739181519,
|
| 2442 |
+
"learning_rate": 8.084698267267408e-06,
|
| 2443 |
+
"loss": 0.1705,
|
| 2444 |
+
"step": 3410
|
| 2445 |
+
},
|
| 2446 |
+
{
|
| 2447 |
+
"epoch": 1.601123595505618,
|
| 2448 |
+
"grad_norm": 0.26720166206359863,
|
| 2449 |
+
"learning_rate": 7.908732556338628e-06,
|
| 2450 |
+
"loss": 0.1427,
|
| 2451 |
+
"step": 3420
|
| 2452 |
+
},
|
| 2453 |
+
{
|
| 2454 |
+
"epoch": 1.6058052434456929,
|
| 2455 |
+
"grad_norm": 0.2854466140270233,
|
| 2456 |
+
"learning_rate": 7.73434243857814e-06,
|
| 2457 |
+
"loss": 0.1405,
|
| 2458 |
+
"step": 3430
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 1.6104868913857677,
|
| 2462 |
+
"grad_norm": 0.29229024052619934,
|
| 2463 |
+
"learning_rate": 7.561543990492803e-06,
|
| 2464 |
+
"loss": 0.1445,
|
| 2465 |
+
"step": 3440
|
| 2466 |
+
},
|
| 2467 |
+
{
|
| 2468 |
+
"epoch": 1.6151685393258428,
|
| 2469 |
+
"grad_norm": 0.2715863883495331,
|
| 2470 |
+
"learning_rate": 7.390353141858172e-06,
|
| 2471 |
+
"loss": 0.1363,
|
| 2472 |
+
"step": 3450
|
| 2473 |
+
},
|
| 2474 |
+
{
|
| 2475 |
+
"epoch": 1.6198501872659175,
|
| 2476 |
+
"grad_norm": 0.24178357422351837,
|
| 2477 |
+
"learning_rate": 7.2207856742499695e-06,
|
| 2478 |
+
"loss": 0.144,
|
| 2479 |
+
"step": 3460
|
| 2480 |
+
},
|
| 2481 |
+
{
|
| 2482 |
+
"epoch": 1.6245318352059925,
|
| 2483 |
+
"grad_norm": 0.25843772292137146,
|
| 2484 |
+
"learning_rate": 7.052857219589262e-06,
|
| 2485 |
+
"loss": 0.1458,
|
| 2486 |
+
"step": 3470
|
| 2487 |
+
},
|
| 2488 |
+
{
|
| 2489 |
+
"epoch": 1.6292134831460674,
|
| 2490 |
+
"grad_norm": 0.2684144079685211,
|
| 2491 |
+
"learning_rate": 6.886583258701382e-06,
|
| 2492 |
+
"loss": 0.1337,
|
| 2493 |
+
"step": 3480
|
| 2494 |
+
},
|
| 2495 |
+
{
|
| 2496 |
+
"epoch": 1.6338951310861423,
|
| 2497 |
+
"grad_norm": 0.2679165303707123,
|
| 2498 |
+
"learning_rate": 6.721979119888796e-06,
|
| 2499 |
+
"loss": 0.1814,
|
| 2500 |
+
"step": 3490
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 1.6385767790262173,
|
| 2504 |
+
"grad_norm": 0.2938106954097748,
|
| 2505 |
+
"learning_rate": 6.559059977518017e-06,
|
| 2506 |
+
"loss": 0.1569,
|
| 2507 |
+
"step": 3500
|
| 2508 |
+
},
|
| 2509 |
+
{
|
| 2510 |
+
"epoch": 1.6385767790262173,
|
| 2511 |
+
"eval_loss": 0.14246584475040436,
|
| 2512 |
+
"eval_runtime": 232.5612,
|
| 2513 |
+
"eval_samples_per_second": 3.001,
|
| 2514 |
+
"eval_steps_per_second": 1.501,
|
| 2515 |
+
"step": 3500
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 1.643258426966292,
|
| 2519 |
+
"grad_norm": 0.2740805447101593,
|
| 2520 |
+
"learning_rate": 6.397840850620762e-06,
|
| 2521 |
+
"loss": 0.1476,
|
| 2522 |
+
"step": 3510
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 1.647940074906367,
|
| 2526 |
+
"grad_norm": 0.260959655046463,
|
| 2527 |
+
"learning_rate": 6.238336601509365e-06,
|
| 2528 |
+
"loss": 0.1472,
|
| 2529 |
+
"step": 3520
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 1.652621722846442,
|
| 2533 |
+
"grad_norm": 0.2726864814758301,
|
| 2534 |
+
"learning_rate": 6.080561934406678e-06,
|
| 2535 |
+
"loss": 0.1474,
|
| 2536 |
+
"step": 3530
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 1.6573033707865168,
|
| 2540 |
+
"grad_norm": 0.2838584780693054,
|
| 2541 |
+
"learning_rate": 5.92453139409051e-06,
|
| 2542 |
+
"loss": 0.136,
|
| 2543 |
+
"step": 3540
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 1.6619850187265919,
|
| 2547 |
+
"grad_norm": 0.2548011243343353,
|
| 2548 |
+
"learning_rate": 5.770259364552816e-06,
|
| 2549 |
+
"loss": 0.1435,
|
| 2550 |
+
"step": 3550
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 1.6666666666666665,
|
| 2554 |
+
"grad_norm": 0.25916817784309387,
|
| 2555 |
+
"learning_rate": 5.617760067673666e-06,
|
| 2556 |
+
"loss": 0.1413,
|
| 2557 |
+
"step": 3560
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 1.6713483146067416,
|
| 2561 |
+
"grad_norm": 0.25346603989601135,
|
| 2562 |
+
"learning_rate": 5.46704756191018e-06,
|
| 2563 |
+
"loss": 0.1359,
|
| 2564 |
+
"step": 3570
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 1.6760299625468165,
|
| 2568 |
+
"grad_norm": 0.2750667333602905,
|
| 2569 |
+
"learning_rate": 5.318135741000488e-06,
|
| 2570 |
+
"loss": 0.1542,
|
| 2571 |
+
"step": 3580
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 1.6807116104868913,
|
| 2575 |
+
"grad_norm": 0.28603386878967285,
|
| 2576 |
+
"learning_rate": 5.171038332682959e-06,
|
| 2577 |
+
"loss": 0.1283,
|
| 2578 |
+
"step": 3590
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 1.6853932584269664,
|
| 2582 |
+
"grad_norm": 0.2523970305919647,
|
| 2583 |
+
"learning_rate": 5.025768897430644e-06,
|
| 2584 |
+
"loss": 0.1376,
|
| 2585 |
+
"step": 3600
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 1.690074906367041,
|
| 2589 |
+
"grad_norm": 0.27870383858680725,
|
| 2590 |
+
"learning_rate": 4.882340827201193e-06,
|
| 2591 |
+
"loss": 0.139,
|
| 2592 |
+
"step": 3610
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 1.6947565543071161,
|
| 2596 |
+
"grad_norm": 0.2658084034919739,
|
| 2597 |
+
"learning_rate": 4.740767344202282e-06,
|
| 2598 |
+
"loss": 0.1457,
|
| 2599 |
+
"step": 3620
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 1.699438202247191,
|
| 2603 |
+
"grad_norm": 0.2911655008792877,
|
| 2604 |
+
"learning_rate": 4.601061499672698e-06,
|
| 2605 |
+
"loss": 0.1353,
|
| 2606 |
+
"step": 3630
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 1.7041198501872659,
|
| 2610 |
+
"grad_norm": 0.270166277885437,
|
| 2611 |
+
"learning_rate": 4.463236172679192e-06,
|
| 2612 |
+
"loss": 0.1525,
|
| 2613 |
+
"step": 3640
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 1.708801498127341,
|
| 2617 |
+
"grad_norm": 0.2789149582386017,
|
| 2618 |
+
"learning_rate": 4.327304068929192e-06,
|
| 2619 |
+
"loss": 0.1604,
|
| 2620 |
+
"step": 3650
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 1.7134831460674156,
|
| 2624 |
+
"grad_norm": 0.2603972554206848,
|
| 2625 |
+
"learning_rate": 4.193277719599484e-06,
|
| 2626 |
+
"loss": 0.1482,
|
| 2627 |
+
"step": 3660
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 1.7181647940074907,
|
| 2631 |
+
"grad_norm": 0.28370609879493713,
|
| 2632 |
+
"learning_rate": 4.061169480181029e-06,
|
| 2633 |
+
"loss": 0.1326,
|
| 2634 |
+
"step": 3670
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 1.7228464419475655,
|
| 2638 |
+
"grad_norm": 0.28793907165527344,
|
| 2639 |
+
"learning_rate": 3.9309915293399366e-06,
|
| 2640 |
+
"loss": 0.1483,
|
| 2641 |
+
"step": 3680
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 1.7275280898876404,
|
| 2645 |
+
"grad_norm": 0.2687625288963318,
|
| 2646 |
+
"learning_rate": 3.802755867794744e-06,
|
| 2647 |
+
"loss": 0.1343,
|
| 2648 |
+
"step": 3690
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 1.7322097378277155,
|
| 2652 |
+
"grad_norm": 0.2577783763408661,
|
| 2653 |
+
"learning_rate": 3.676474317210099e-06,
|
| 2654 |
+
"loss": 0.1381,
|
| 2655 |
+
"step": 3700
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 1.7368913857677901,
|
| 2659 |
+
"grad_norm": 0.2647913992404938,
|
| 2660 |
+
"learning_rate": 3.552158519106977e-06,
|
| 2661 |
+
"loss": 0.1417,
|
| 2662 |
+
"step": 3710
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 1.7415730337078652,
|
| 2666 |
+
"grad_norm": 0.2621007561683655,
|
| 2667 |
+
"learning_rate": 3.4298199337894685e-06,
|
| 2668 |
+
"loss": 0.1462,
|
| 2669 |
+
"step": 3720
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 1.74625468164794,
|
| 2673 |
+
"grad_norm": 0.2610481083393097,
|
| 2674 |
+
"learning_rate": 3.309469839288301e-06,
|
| 2675 |
+
"loss": 0.1315,
|
| 2676 |
+
"step": 3730
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 1.750936329588015,
|
| 2680 |
+
"grad_norm": 0.2840542197227478,
|
| 2681 |
+
"learning_rate": 3.1911193303211185e-06,
|
| 2682 |
+
"loss": 0.1362,
|
| 2683 |
+
"step": 3740
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 1.75561797752809,
|
| 2687 |
+
"grad_norm": 0.2668418884277344,
|
| 2688 |
+
"learning_rate": 3.0747793172697326e-06,
|
| 2689 |
+
"loss": 0.1492,
|
| 2690 |
+
"step": 3750
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 1.7602996254681647,
|
| 2694 |
+
"grad_norm": 0.2559303939342499,
|
| 2695 |
+
"learning_rate": 2.9604605251743136e-06,
|
| 2696 |
+
"loss": 0.1315,
|
| 2697 |
+
"step": 3760
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 1.7649812734082397,
|
| 2701 |
+
"grad_norm": 0.2896023392677307,
|
| 2702 |
+
"learning_rate": 2.84817349274466e-06,
|
| 2703 |
+
"loss": 0.1631,
|
| 2704 |
+
"step": 3770
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 1.7696629213483146,
|
| 2708 |
+
"grad_norm": 0.2894950211048126,
|
| 2709 |
+
"learning_rate": 2.7379285713886954e-06,
|
| 2710 |
+
"loss": 0.1472,
|
| 2711 |
+
"step": 3780
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 1.7743445692883895,
|
| 2715 |
+
"grad_norm": 0.281281977891922,
|
| 2716 |
+
"learning_rate": 2.6297359242581865e-06,
|
| 2717 |
+
"loss": 0.138,
|
| 2718 |
+
"step": 3790
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 1.7790262172284645,
|
| 2722 |
+
"grad_norm": 0.24921460449695587,
|
| 2723 |
+
"learning_rate": 2.5236055253118423e-06,
|
| 2724 |
+
"loss": 0.1753,
|
| 2725 |
+
"step": 3800
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 1.7837078651685392,
|
| 2729 |
+
"grad_norm": 0.2553311884403229,
|
| 2730 |
+
"learning_rate": 2.4195471583958266e-06,
|
| 2731 |
+
"loss": 0.1375,
|
| 2732 |
+
"step": 3810
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 1.7883895131086143,
|
| 2736 |
+
"grad_norm": 0.2462325245141983,
|
| 2737 |
+
"learning_rate": 2.3175704163418327e-06,
|
| 2738 |
+
"loss": 0.133,
|
| 2739 |
+
"step": 3820
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 1.7930711610486891,
|
| 2743 |
+
"grad_norm": 0.2653181552886963,
|
| 2744 |
+
"learning_rate": 2.217684700082748e-06,
|
| 2745 |
+
"loss": 0.1432,
|
| 2746 |
+
"step": 3830
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 1.797752808988764,
|
| 2750 |
+
"grad_norm": 0.2803376615047455,
|
| 2751 |
+
"learning_rate": 2.119899217785995e-06,
|
| 2752 |
+
"loss": 0.1458,
|
| 2753 |
+
"step": 3840
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 1.802434456928839,
|
| 2757 |
+
"grad_norm": 0.2679455280303955,
|
| 2758 |
+
"learning_rate": 2.0242229840046726e-06,
|
| 2759 |
+
"loss": 0.1431,
|
| 2760 |
+
"step": 3850
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 1.8071161048689137,
|
| 2764 |
+
"grad_norm": 0.2571849226951599,
|
| 2765 |
+
"learning_rate": 1.9306648188465252e-06,
|
| 2766 |
+
"loss": 0.142,
|
| 2767 |
+
"step": 3860
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 1.8117977528089888,
|
| 2771 |
+
"grad_norm": 0.24948078393936157,
|
| 2772 |
+
"learning_rate": 1.8392333471608347e-06,
|
| 2773 |
+
"loss": 0.1468,
|
| 2774 |
+
"step": 3870
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 1.8164794007490637,
|
| 2778 |
+
"grad_norm": 0.2405436486005783,
|
| 2779 |
+
"learning_rate": 1.7499369977433456e-06,
|
| 2780 |
+
"loss": 0.1338,
|
| 2781 |
+
"step": 3880
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 1.8211610486891385,
|
| 2785 |
+
"grad_norm": 0.25163865089416504,
|
| 2786 |
+
"learning_rate": 1.662784002559209e-06,
|
| 2787 |
+
"loss": 0.1439,
|
| 2788 |
+
"step": 3890
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 1.8258426966292136,
|
| 2792 |
+
"grad_norm": 0.2572096288204193,
|
| 2793 |
+
"learning_rate": 1.577782395984126e-06,
|
| 2794 |
+
"loss": 0.1447,
|
| 2795 |
+
"step": 3900
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 1.8305243445692883,
|
| 2799 |
+
"grad_norm": 0.286592960357666,
|
| 2800 |
+
"learning_rate": 1.4949400140636727e-06,
|
| 2801 |
+
"loss": 0.1482,
|
| 2802 |
+
"step": 3910
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 1.8352059925093633,
|
| 2806 |
+
"grad_norm": 0.268310546875,
|
| 2807 |
+
"learning_rate": 1.4142644937909206e-06,
|
| 2808 |
+
"loss": 0.1478,
|
| 2809 |
+
"step": 3920
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 1.8398876404494382,
|
| 2813 |
+
"grad_norm": 0.25413018465042114,
|
| 2814 |
+
"learning_rate": 1.3357632724023933e-06,
|
| 2815 |
+
"loss": 0.14,
|
| 2816 |
+
"step": 3930
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 1.844569288389513,
|
| 2820 |
+
"grad_norm": 0.25173068046569824,
|
| 2821 |
+
"learning_rate": 1.2594435866924686e-06,
|
| 2822 |
+
"loss": 0.1444,
|
| 2823 |
+
"step": 3940
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 1.8492509363295881,
|
| 2827 |
+
"grad_norm": 0.2957823574542999,
|
| 2828 |
+
"learning_rate": 1.1853124723462311e-06,
|
| 2829 |
+
"loss": 0.1437,
|
| 2830 |
+
"step": 3950
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 1.8539325842696628,
|
| 2834 |
+
"grad_norm": 0.2906501293182373,
|
| 2835 |
+
"learning_rate": 1.1133767632908799e-06,
|
| 2836 |
+
"loss": 0.141,
|
| 2837 |
+
"step": 3960
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 1.8586142322097379,
|
| 2841 |
+
"grad_norm": 0.2822731137275696,
|
| 2842 |
+
"learning_rate": 1.043643091065702e-06,
|
| 2843 |
+
"loss": 0.1516,
|
| 2844 |
+
"step": 3970
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 1.8632958801498127,
|
| 2848 |
+
"grad_norm": 0.2239583134651184,
|
| 2849 |
+
"learning_rate": 9.761178842107699e-07,
|
| 2850 |
+
"loss": 0.1289,
|
| 2851 |
+
"step": 3980
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 1.8679775280898876,
|
| 2855 |
+
"grad_norm": 0.2751089632511139,
|
| 2856 |
+
"learning_rate": 9.108073676742935e-07,
|
| 2857 |
+
"loss": 0.1691,
|
| 2858 |
+
"step": 3990
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 1.8726591760299627,
|
| 2862 |
+
"grad_norm": 0.26302212476730347,
|
| 2863 |
+
"learning_rate": 8.477175622387562e-07,
|
| 2864 |
+
"loss": 0.1413,
|
| 2865 |
+
"step": 4000
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 1.8726591760299627,
|
| 2869 |
+
"eval_loss": 0.14174455404281616,
|
| 2870 |
+
"eval_runtime": 234.3945,
|
| 2871 |
+
"eval_samples_per_second": 2.978,
|
| 2872 |
+
"eval_steps_per_second": 1.489,
|
| 2873 |
+
"step": 4000
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 1.8773408239700373,
|
| 2877 |
+
"grad_norm": 0.2844927906990051,
|
| 2878 |
+
"learning_rate": 7.868542839658793e-07,
|
| 2879 |
+
"loss": 0.1337,
|
| 2880 |
+
"step": 4010
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 1.8820224719101124,
|
| 2884 |
+
"grad_norm": 0.2576298415660858,
|
| 2885 |
+
"learning_rate": 7.282231436604698e-07,
|
| 2886 |
+
"loss": 0.1378,
|
| 2887 |
+
"step": 4020
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 1.8867041198501873,
|
| 2891 |
+
"grad_norm": 0.24751824140548706,
|
| 2892 |
+
"learning_rate": 6.718295463531587e-07,
|
| 2893 |
+
"loss": 0.1661,
|
| 2894 |
+
"step": 4030
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 1.8913857677902621,
|
| 2898 |
+
"grad_norm": 0.2758065164089203,
|
| 2899 |
+
"learning_rate": 6.176786908021453e-07,
|
| 2900 |
+
"loss": 0.1506,
|
| 2901 |
+
"step": 4040
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 1.8960674157303372,
|
| 2905 |
+
"grad_norm": 0.24978813529014587,
|
| 2906 |
+
"learning_rate": 5.657755690139155e-07,
|
| 2907 |
+
"loss": 0.1467,
|
| 2908 |
+
"step": 4050
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 1.9007490636704119,
|
| 2912 |
+
"grad_norm": 0.2835918068885803,
|
| 2913 |
+
"learning_rate": 5.161249657830686e-07,
|
| 2914 |
+
"loss": 0.1582,
|
| 2915 |
+
"step": 4060
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 1.905430711610487,
|
| 2919 |
+
"grad_norm": 0.2483970671892166,
|
| 2920 |
+
"learning_rate": 4.6873145825120926e-07,
|
| 2921 |
+
"loss": 0.1382,
|
| 2922 |
+
"step": 4070
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 1.9101123595505618,
|
| 2926 |
+
"grad_norm": 0.24273446202278137,
|
| 2927 |
+
"learning_rate": 4.2359941548499037e-07,
|
| 2928 |
+
"loss": 0.1229,
|
| 2929 |
+
"step": 4080
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 1.9147940074906367,
|
| 2933 |
+
"grad_norm": 0.2832223176956177,
|
| 2934 |
+
"learning_rate": 3.8073299807334926e-07,
|
| 2935 |
+
"loss": 0.138,
|
| 2936 |
+
"step": 4090
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 1.9194756554307117,
|
| 2940 |
+
"grad_norm": 0.27922406792640686,
|
| 2941 |
+
"learning_rate": 3.4013615774395323e-07,
|
| 2942 |
+
"loss": 0.1433,
|
| 2943 |
+
"step": 4100
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 1.9241573033707864,
|
| 2947 |
+
"grad_norm": 0.2603136897087097,
|
| 2948 |
+
"learning_rate": 3.0181263699890215e-07,
|
| 2949 |
+
"loss": 0.1406,
|
| 2950 |
+
"step": 4110
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 1.9288389513108615,
|
| 2954 |
+
"grad_norm": 0.24417099356651306,
|
| 2955 |
+
"learning_rate": 2.657659687697156e-07,
|
| 2956 |
+
"loss": 0.1483,
|
| 2957 |
+
"step": 4120
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 1.9335205992509363,
|
| 2961 |
+
"grad_norm": 0.2983071506023407,
|
| 2962 |
+
"learning_rate": 2.319994760916433e-07,
|
| 2963 |
+
"loss": 0.1497,
|
| 2964 |
+
"step": 4130
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 1.9382022471910112,
|
| 2968 |
+
"grad_norm": 0.23415164649486542,
|
| 2969 |
+
"learning_rate": 2.0051627179733247e-07,
|
| 2970 |
+
"loss": 0.1601,
|
| 2971 |
+
"step": 4140
|
| 2972 |
+
},
|
| 2973 |
+
{
|
| 2974 |
+
"epoch": 1.9428838951310863,
|
| 2975 |
+
"grad_norm": 0.280741810798645,
|
| 2976 |
+
"learning_rate": 1.713192582298434e-07,
|
| 2977 |
+
"loss": 0.1432,
|
| 2978 |
+
"step": 4150
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"epoch": 1.947565543071161,
|
| 2982 |
+
"grad_norm": 0.3024005889892578,
|
| 2983 |
+
"learning_rate": 1.444111269751164e-07,
|
| 2984 |
+
"loss": 0.1537,
|
| 2985 |
+
"step": 4160
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 1.952247191011236,
|
| 2989 |
+
"grad_norm": 0.26236817240715027,
|
| 2990 |
+
"learning_rate": 1.1979435861383392e-07,
|
| 2991 |
+
"loss": 0.1574,
|
| 2992 |
+
"step": 4170
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 1.9569288389513109,
|
| 2996 |
+
"grad_norm": 0.28959619998931885,
|
| 2997 |
+
"learning_rate": 9.747122249273156e-08,
|
| 2998 |
+
"loss": 0.1396,
|
| 2999 |
+
"step": 4180
|
| 3000 |
+
},
|
| 3001 |
+
{
|
| 3002 |
+
"epoch": 1.9616104868913857,
|
| 3003 |
+
"grad_norm": 0.2532627582550049,
|
| 3004 |
+
"learning_rate": 7.74437765154068e-08,
|
| 3005 |
+
"loss": 0.1404,
|
| 3006 |
+
"step": 4190
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 1.9662921348314608,
|
| 3010 |
+
"grad_norm": 0.2636815309524536,
|
| 3011 |
+
"learning_rate": 5.971386695260706e-08,
|
| 3012 |
+
"loss": 0.1445,
|
| 3013 |
+
"step": 4200
|
| 3014 |
+
},
|
| 3015 |
+
{
|
| 3016 |
+
"epoch": 1.9709737827715355,
|
| 3017 |
+
"grad_norm": 0.27110251784324646,
|
| 3018 |
+
"learning_rate": 4.4283128272018524e-08,
|
| 3019 |
+
"loss": 0.1436,
|
| 3020 |
+
"step": 4210
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 1.9756554307116105,
|
| 3024 |
+
"grad_norm": 0.2736012935638428,
|
| 3025 |
+
"learning_rate": 3.1152982987595056e-08,
|
| 3026 |
+
"loss": 0.1689,
|
| 3027 |
+
"step": 4220
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 1.9803370786516854,
|
| 3031 |
+
"grad_norm": 0.2657409608364105,
|
| 3032 |
+
"learning_rate": 2.0324641528421395e-08,
|
| 3033 |
+
"loss": 0.1399,
|
| 3034 |
+
"step": 4230
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 1.9850187265917603,
|
| 3038 |
+
"grad_norm": 0.2698764204978943,
|
| 3039 |
+
"learning_rate": 1.1799102127130202e-08,
|
| 3040 |
+
"loss": 0.1302,
|
| 3041 |
+
"step": 4240
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 1.9897003745318353,
|
| 3045 |
+
"grad_norm": 0.27992117404937744,
|
| 3046 |
+
"learning_rate": 5.5771507278645775e-09,
|
| 3047 |
+
"loss": 0.1438,
|
| 3048 |
+
"step": 4250
|
| 3049 |
+
},
|
| 3050 |
+
{
|
| 3051 |
+
"epoch": 1.99438202247191,
|
| 3052 |
+
"grad_norm": 0.2756190001964569,
|
| 3053 |
+
"learning_rate": 1.6593609138415567e-09,
|
| 3054 |
+
"loss": 0.1459,
|
| 3055 |
+
"step": 4260
|
| 3056 |
+
},
|
| 3057 |
+
{
|
| 3058 |
+
"epoch": 1.999063670411985,
|
| 3059 |
+
"grad_norm": 0.2640279233455658,
|
| 3060 |
+
"learning_rate": 4.6093854466633393e-11,
|
| 3061 |
+
"loss": 0.1339,
|
| 3062 |
+
"step": 4270
|
| 3063 |
+
},
|
| 3064 |
+
{
|
| 3065 |
+
"epoch": 2.0,
|
| 3066 |
+
"step": 4272,
|
| 3067 |
+
"total_flos": 2.778312922424672e+18,
|
| 3068 |
+
"train_loss": 0.20174445870440552,
|
| 3069 |
+
"train_runtime": 74295.9338,
|
| 3070 |
+
"train_samples_per_second": 0.92,
|
| 3071 |
+
"train_steps_per_second": 0.057
|
| 3072 |
+
}
|
| 3073 |
+
],
|
| 3074 |
+
"logging_steps": 10,
|
| 3075 |
+
"max_steps": 4272,
|
| 3076 |
+
"num_input_tokens_seen": 0,
|
| 3077 |
+
"num_train_epochs": 2,
|
| 3078 |
+
"save_steps": 500,
|
| 3079 |
+
"stateful_callbacks": {
|
| 3080 |
+
"TrainerControl": {
|
| 3081 |
+
"args": {
|
| 3082 |
+
"should_epoch_stop": false,
|
| 3083 |
+
"should_evaluate": false,
|
| 3084 |
+
"should_log": false,
|
| 3085 |
+
"should_save": true,
|
| 3086 |
+
"should_training_stop": true
|
| 3087 |
+
},
|
| 3088 |
+
"attributes": {}
|
| 3089 |
+
}
|
| 3090 |
+
},
|
| 3091 |
+
"total_flos": 2.778312922424672e+18,
|
| 3092 |
+
"train_batch_size": 4,
|
| 3093 |
+
"trial_name": null,
|
| 3094 |
+
"trial_params": null
|
| 3095 |
+
}
|