Upload lora/trainer_state.json with huggingface_hub
Browse files- lora/trainer_state.json +2283 -0
lora/trainer_state.json
ADDED
|
@@ -0,0 +1,2283 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.44520583748817444,
|
| 3 |
+
"best_model_checkpoint": "saves/Orca/lora/checkpoint-3000",
|
| 4 |
+
"epoch": 0.9998413957176844,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3152,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0031720856463124504,
|
| 13 |
+
"grad_norm": 0.8383375406265259,
|
| 14 |
+
"learning_rate": 5.000000000000001e-07,
|
| 15 |
+
"loss": 2.0081,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.006344171292624901,
|
| 20 |
+
"grad_norm": 0.9924185276031494,
|
| 21 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 22 |
+
"loss": 2.0141,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.00951625693893735,
|
| 27 |
+
"grad_norm": 1.7123312950134277,
|
| 28 |
+
"learning_rate": 1.5e-06,
|
| 29 |
+
"loss": 2.0134,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.012688342585249802,
|
| 34 |
+
"grad_norm": 1.6002644300460815,
|
| 35 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 36 |
+
"loss": 1.9886,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.01586042823156225,
|
| 41 |
+
"grad_norm": 1.4284940958023071,
|
| 42 |
+
"learning_rate": 2.5e-06,
|
| 43 |
+
"loss": 1.9963,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.0190325138778747,
|
| 48 |
+
"grad_norm": 1.9293030500411987,
|
| 49 |
+
"learning_rate": 3e-06,
|
| 50 |
+
"loss": 1.9858,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.022204599524187154,
|
| 55 |
+
"grad_norm": 1.8281705379486084,
|
| 56 |
+
"learning_rate": 3.5000000000000004e-06,
|
| 57 |
+
"loss": 1.9608,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.025376685170499604,
|
| 62 |
+
"grad_norm": 1.8404144048690796,
|
| 63 |
+
"learning_rate": 4.000000000000001e-06,
|
| 64 |
+
"loss": 1.9317,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.028548770816812053,
|
| 69 |
+
"grad_norm": 1.8080644607543945,
|
| 70 |
+
"learning_rate": 4.5e-06,
|
| 71 |
+
"loss": 1.8928,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.0317208564631245,
|
| 76 |
+
"grad_norm": 1.4789849519729614,
|
| 77 |
+
"learning_rate": 5e-06,
|
| 78 |
+
"loss": 1.8453,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.034892942109436956,
|
| 83 |
+
"grad_norm": 2.3675060272216797,
|
| 84 |
+
"learning_rate": 5.500000000000001e-06,
|
| 85 |
+
"loss": 1.7683,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.0380650277557494,
|
| 90 |
+
"grad_norm": 6.9262309074401855,
|
| 91 |
+
"learning_rate": 6e-06,
|
| 92 |
+
"loss": 1.6648,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.041237113402061855,
|
| 97 |
+
"grad_norm": 1.5342968702316284,
|
| 98 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 99 |
+
"loss": 1.5507,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.04440919904837431,
|
| 104 |
+
"grad_norm": 1.7138350009918213,
|
| 105 |
+
"learning_rate": 7.000000000000001e-06,
|
| 106 |
+
"loss": 1.4388,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.047581284694686754,
|
| 111 |
+
"grad_norm": 0.9868146181106567,
|
| 112 |
+
"learning_rate": 7.5e-06,
|
| 113 |
+
"loss": 1.3484,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.05075337034099921,
|
| 118 |
+
"grad_norm": 0.649817705154419,
|
| 119 |
+
"learning_rate": 8.000000000000001e-06,
|
| 120 |
+
"loss": 1.2671,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.05392545598731166,
|
| 125 |
+
"grad_norm": 0.5127522945404053,
|
| 126 |
+
"learning_rate": 8.500000000000002e-06,
|
| 127 |
+
"loss": 1.19,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.057097541633624106,
|
| 132 |
+
"grad_norm": 0.43639031052589417,
|
| 133 |
+
"learning_rate": 9e-06,
|
| 134 |
+
"loss": 1.1479,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.06026962727993656,
|
| 139 |
+
"grad_norm": 0.31427648663520813,
|
| 140 |
+
"learning_rate": 9.5e-06,
|
| 141 |
+
"loss": 1.0992,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.063441712926249,
|
| 146 |
+
"grad_norm": 0.2897365689277649,
|
| 147 |
+
"learning_rate": 1e-05,
|
| 148 |
+
"loss": 1.0673,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.06661379857256146,
|
| 153 |
+
"grad_norm": 0.43495845794677734,
|
| 154 |
+
"learning_rate": 1.05e-05,
|
| 155 |
+
"loss": 1.0399,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.06978588421887391,
|
| 160 |
+
"grad_norm": 0.2536582946777344,
|
| 161 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 162 |
+
"loss": 1.0043,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.07295796986518636,
|
| 167 |
+
"grad_norm": 0.3480685353279114,
|
| 168 |
+
"learning_rate": 1.1500000000000002e-05,
|
| 169 |
+
"loss": 0.9747,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.0761300555114988,
|
| 174 |
+
"grad_norm": 0.25048044323921204,
|
| 175 |
+
"learning_rate": 1.2e-05,
|
| 176 |
+
"loss": 0.9477,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.07930214115781126,
|
| 181 |
+
"grad_norm": 0.31881338357925415,
|
| 182 |
+
"learning_rate": 1.25e-05,
|
| 183 |
+
"loss": 0.9099,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.08247422680412371,
|
| 188 |
+
"grad_norm": 0.25372961163520813,
|
| 189 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 190 |
+
"loss": 0.8661,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.08564631245043616,
|
| 195 |
+
"grad_norm": 0.2892467975616455,
|
| 196 |
+
"learning_rate": 1.3500000000000001e-05,
|
| 197 |
+
"loss": 0.8306,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.08881839809674862,
|
| 202 |
+
"grad_norm": 0.265289306640625,
|
| 203 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 204 |
+
"loss": 0.79,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.09199048374306107,
|
| 209 |
+
"grad_norm": 0.2384192794561386,
|
| 210 |
+
"learning_rate": 1.45e-05,
|
| 211 |
+
"loss": 0.7414,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.09516256938937351,
|
| 216 |
+
"grad_norm": 0.2318211942911148,
|
| 217 |
+
"learning_rate": 1.5e-05,
|
| 218 |
+
"loss": 0.6994,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.09833465503568596,
|
| 223 |
+
"grad_norm": 0.22584764659404755,
|
| 224 |
+
"learning_rate": 1.55e-05,
|
| 225 |
+
"loss": 0.6663,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.10150674068199841,
|
| 230 |
+
"grad_norm": 0.2281956821680069,
|
| 231 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 232 |
+
"loss": 0.6366,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.10467882632831087,
|
| 237 |
+
"grad_norm": 0.23853954672813416,
|
| 238 |
+
"learning_rate": 1.65e-05,
|
| 239 |
+
"loss": 0.6105,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.10785091197462332,
|
| 244 |
+
"grad_norm": 0.2249988317489624,
|
| 245 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 246 |
+
"loss": 0.5923,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.11102299762093576,
|
| 251 |
+
"grad_norm": 0.22325117886066437,
|
| 252 |
+
"learning_rate": 1.75e-05,
|
| 253 |
+
"loss": 0.5782,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.11419508326724821,
|
| 258 |
+
"grad_norm": 0.26995840668678284,
|
| 259 |
+
"learning_rate": 1.8e-05,
|
| 260 |
+
"loss": 0.5588,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.11736716891356067,
|
| 265 |
+
"grad_norm": 0.3128603994846344,
|
| 266 |
+
"learning_rate": 1.85e-05,
|
| 267 |
+
"loss": 0.5539,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.12053925455987312,
|
| 272 |
+
"grad_norm": 0.2401484102010727,
|
| 273 |
+
"learning_rate": 1.9e-05,
|
| 274 |
+
"loss": 0.5463,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.12371134020618557,
|
| 279 |
+
"grad_norm": 0.252087265253067,
|
| 280 |
+
"learning_rate": 1.9500000000000003e-05,
|
| 281 |
+
"loss": 0.5396,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.126883425852498,
|
| 286 |
+
"grad_norm": 0.3551071286201477,
|
| 287 |
+
"learning_rate": 2e-05,
|
| 288 |
+
"loss": 0.5293,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.13005551149881048,
|
| 293 |
+
"grad_norm": 0.2852734625339508,
|
| 294 |
+
"learning_rate": 2.05e-05,
|
| 295 |
+
"loss": 0.5327,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.13322759714512292,
|
| 300 |
+
"grad_norm": 0.25599339604377747,
|
| 301 |
+
"learning_rate": 2.1e-05,
|
| 302 |
+
"loss": 0.5215,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.13639968279143536,
|
| 307 |
+
"grad_norm": 0.2308124303817749,
|
| 308 |
+
"learning_rate": 2.15e-05,
|
| 309 |
+
"loss": 0.5179,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.13957176843774782,
|
| 314 |
+
"grad_norm": 0.28417715430259705,
|
| 315 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 316 |
+
"loss": 0.5122,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.14274385408406026,
|
| 321 |
+
"grad_norm": 0.32094892859458923,
|
| 322 |
+
"learning_rate": 2.25e-05,
|
| 323 |
+
"loss": 0.5078,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.14591593973037273,
|
| 328 |
+
"grad_norm": 0.28217512369155884,
|
| 329 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 330 |
+
"loss": 0.5064,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.14908802537668517,
|
| 335 |
+
"grad_norm": 0.3319534659385681,
|
| 336 |
+
"learning_rate": 2.35e-05,
|
| 337 |
+
"loss": 0.5056,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.1522601110229976,
|
| 342 |
+
"grad_norm": 0.35069146752357483,
|
| 343 |
+
"learning_rate": 2.4e-05,
|
| 344 |
+
"loss": 0.5001,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.15543219666931007,
|
| 349 |
+
"grad_norm": 0.2778172492980957,
|
| 350 |
+
"learning_rate": 2.45e-05,
|
| 351 |
+
"loss": 0.4994,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.1586042823156225,
|
| 356 |
+
"grad_norm": 0.2950364053249359,
|
| 357 |
+
"learning_rate": 2.5e-05,
|
| 358 |
+
"loss": 0.5009,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.1586042823156225,
|
| 363 |
+
"eval_loss": 0.49637770652770996,
|
| 364 |
+
"eval_runtime": 23627.0877,
|
| 365 |
+
"eval_samples_per_second": 0.044,
|
| 366 |
+
"eval_steps_per_second": 0.022,
|
| 367 |
+
"step": 500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.16177636796193498,
|
| 371 |
+
"grad_norm": 0.26569539308547974,
|
| 372 |
+
"learning_rate": 2.5500000000000003e-05,
|
| 373 |
+
"loss": 0.4978,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.16494845360824742,
|
| 378 |
+
"grad_norm": 0.26948851346969604,
|
| 379 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 380 |
+
"loss": 0.4926,
|
| 381 |
+
"step": 520
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.1681205392545599,
|
| 385 |
+
"grad_norm": 0.29083552956581116,
|
| 386 |
+
"learning_rate": 2.6500000000000004e-05,
|
| 387 |
+
"loss": 0.4893,
|
| 388 |
+
"step": 530
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.17129262490087233,
|
| 392 |
+
"grad_norm": 0.2939367890357971,
|
| 393 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 394 |
+
"loss": 0.491,
|
| 395 |
+
"step": 540
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.17446471054718476,
|
| 399 |
+
"grad_norm": 0.28195658326148987,
|
| 400 |
+
"learning_rate": 2.7500000000000004e-05,
|
| 401 |
+
"loss": 0.4865,
|
| 402 |
+
"step": 550
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.17763679619349723,
|
| 406 |
+
"grad_norm": 0.2827642858028412,
|
| 407 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 408 |
+
"loss": 0.4851,
|
| 409 |
+
"step": 560
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.18080888183980967,
|
| 413 |
+
"grad_norm": 0.2715272605419159,
|
| 414 |
+
"learning_rate": 2.8499999999999998e-05,
|
| 415 |
+
"loss": 0.4833,
|
| 416 |
+
"step": 570
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.18398096748612214,
|
| 420 |
+
"grad_norm": 0.30428314208984375,
|
| 421 |
+
"learning_rate": 2.9e-05,
|
| 422 |
+
"loss": 0.4834,
|
| 423 |
+
"step": 580
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.18715305313243458,
|
| 427 |
+
"grad_norm": 0.2857949435710907,
|
| 428 |
+
"learning_rate": 2.95e-05,
|
| 429 |
+
"loss": 0.4796,
|
| 430 |
+
"step": 590
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.19032513877874702,
|
| 434 |
+
"grad_norm": 0.302212655544281,
|
| 435 |
+
"learning_rate": 3e-05,
|
| 436 |
+
"loss": 0.4826,
|
| 437 |
+
"step": 600
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.19349722442505948,
|
| 441 |
+
"grad_norm": 0.34079357981681824,
|
| 442 |
+
"learning_rate": 3.05e-05,
|
| 443 |
+
"loss": 0.4826,
|
| 444 |
+
"step": 610
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.19666931007137192,
|
| 448 |
+
"grad_norm": 0.3239024877548218,
|
| 449 |
+
"learning_rate": 3.1e-05,
|
| 450 |
+
"loss": 0.4804,
|
| 451 |
+
"step": 620
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.1998413957176844,
|
| 455 |
+
"grad_norm": 0.36531686782836914,
|
| 456 |
+
"learning_rate": 3.15e-05,
|
| 457 |
+
"loss": 0.4782,
|
| 458 |
+
"step": 630
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.20301348136399683,
|
| 462 |
+
"grad_norm": 0.3582271337509155,
|
| 463 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 464 |
+
"loss": 0.4784,
|
| 465 |
+
"step": 640
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.20618556701030927,
|
| 469 |
+
"grad_norm": 0.3250937759876251,
|
| 470 |
+
"learning_rate": 3.2500000000000004e-05,
|
| 471 |
+
"loss": 0.4716,
|
| 472 |
+
"step": 650
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.20935765265662173,
|
| 476 |
+
"grad_norm": 0.28494277596473694,
|
| 477 |
+
"learning_rate": 3.3e-05,
|
| 478 |
+
"loss": 0.4764,
|
| 479 |
+
"step": 660
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.21252973830293417,
|
| 483 |
+
"grad_norm": 0.3156846761703491,
|
| 484 |
+
"learning_rate": 3.35e-05,
|
| 485 |
+
"loss": 0.4742,
|
| 486 |
+
"step": 670
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.21570182394924664,
|
| 490 |
+
"grad_norm": 0.27053794264793396,
|
| 491 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 492 |
+
"loss": 0.4765,
|
| 493 |
+
"step": 680
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.21887390959555908,
|
| 497 |
+
"grad_norm": 0.31487756967544556,
|
| 498 |
+
"learning_rate": 3.45e-05,
|
| 499 |
+
"loss": 0.4749,
|
| 500 |
+
"step": 690
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.22204599524187152,
|
| 504 |
+
"grad_norm": 0.2739400267601013,
|
| 505 |
+
"learning_rate": 3.5e-05,
|
| 506 |
+
"loss": 0.4693,
|
| 507 |
+
"step": 700
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.22521808088818399,
|
| 511 |
+
"grad_norm": 0.2924807369709015,
|
| 512 |
+
"learning_rate": 3.55e-05,
|
| 513 |
+
"loss": 0.4697,
|
| 514 |
+
"step": 710
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.22839016653449642,
|
| 518 |
+
"grad_norm": 0.27241086959838867,
|
| 519 |
+
"learning_rate": 3.6e-05,
|
| 520 |
+
"loss": 0.4711,
|
| 521 |
+
"step": 720
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.2315622521808089,
|
| 525 |
+
"grad_norm": 0.249298095703125,
|
| 526 |
+
"learning_rate": 3.65e-05,
|
| 527 |
+
"loss": 0.4748,
|
| 528 |
+
"step": 730
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.23473433782712133,
|
| 532 |
+
"grad_norm": 0.33761996030807495,
|
| 533 |
+
"learning_rate": 3.7e-05,
|
| 534 |
+
"loss": 0.4692,
|
| 535 |
+
"step": 740
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.23790642347343377,
|
| 539 |
+
"grad_norm": 0.3454744517803192,
|
| 540 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 541 |
+
"loss": 0.4691,
|
| 542 |
+
"step": 750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.24107850911974624,
|
| 546 |
+
"grad_norm": 0.3940749168395996,
|
| 547 |
+
"learning_rate": 3.8e-05,
|
| 548 |
+
"loss": 0.4694,
|
| 549 |
+
"step": 760
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.24425059476605868,
|
| 553 |
+
"grad_norm": 0.2833056151866913,
|
| 554 |
+
"learning_rate": 3.85e-05,
|
| 555 |
+
"loss": 0.4651,
|
| 556 |
+
"step": 770
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.24742268041237114,
|
| 560 |
+
"grad_norm": 0.3439841866493225,
|
| 561 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 562 |
+
"loss": 0.4718,
|
| 563 |
+
"step": 780
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.2505947660586836,
|
| 567 |
+
"grad_norm": 0.28585824370384216,
|
| 568 |
+
"learning_rate": 3.9500000000000005e-05,
|
| 569 |
+
"loss": 0.4678,
|
| 570 |
+
"step": 790
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.253766851704996,
|
| 574 |
+
"grad_norm": 0.2761208415031433,
|
| 575 |
+
"learning_rate": 4e-05,
|
| 576 |
+
"loss": 0.4672,
|
| 577 |
+
"step": 800
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.25693893735130846,
|
| 581 |
+
"grad_norm": 0.287431538105011,
|
| 582 |
+
"learning_rate": 4.05e-05,
|
| 583 |
+
"loss": 0.4654,
|
| 584 |
+
"step": 810
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.26011102299762096,
|
| 588 |
+
"grad_norm": 0.2529178261756897,
|
| 589 |
+
"learning_rate": 4.1e-05,
|
| 590 |
+
"loss": 0.4658,
|
| 591 |
+
"step": 820
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.2632831086439334,
|
| 595 |
+
"grad_norm": 0.22234879434108734,
|
| 596 |
+
"learning_rate": 4.15e-05,
|
| 597 |
+
"loss": 0.467,
|
| 598 |
+
"step": 830
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.26645519429024583,
|
| 602 |
+
"grad_norm": 0.2847014367580414,
|
| 603 |
+
"learning_rate": 4.2e-05,
|
| 604 |
+
"loss": 0.4643,
|
| 605 |
+
"step": 840
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.2696272799365583,
|
| 609 |
+
"grad_norm": 0.27712419629096985,
|
| 610 |
+
"learning_rate": 4.25e-05,
|
| 611 |
+
"loss": 0.4604,
|
| 612 |
+
"step": 850
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.2727993655828707,
|
| 616 |
+
"grad_norm": 0.24886064231395721,
|
| 617 |
+
"learning_rate": 4.3e-05,
|
| 618 |
+
"loss": 0.4657,
|
| 619 |
+
"step": 860
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.2759714512291832,
|
| 623 |
+
"grad_norm": 0.3196752369403839,
|
| 624 |
+
"learning_rate": 4.35e-05,
|
| 625 |
+
"loss": 0.4652,
|
| 626 |
+
"step": 870
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.27914353687549565,
|
| 630 |
+
"grad_norm": 0.27527421712875366,
|
| 631 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 632 |
+
"loss": 0.4634,
|
| 633 |
+
"step": 880
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.2823156225218081,
|
| 637 |
+
"grad_norm": 0.26819467544555664,
|
| 638 |
+
"learning_rate": 4.4500000000000004e-05,
|
| 639 |
+
"loss": 0.4633,
|
| 640 |
+
"step": 890
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.2854877081681205,
|
| 644 |
+
"grad_norm": 0.26196032762527466,
|
| 645 |
+
"learning_rate": 4.5e-05,
|
| 646 |
+
"loss": 0.4656,
|
| 647 |
+
"step": 900
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.28865979381443296,
|
| 651 |
+
"grad_norm": 0.23337453603744507,
|
| 652 |
+
"learning_rate": 4.55e-05,
|
| 653 |
+
"loss": 0.4622,
|
| 654 |
+
"step": 910
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.29183187946074546,
|
| 658 |
+
"grad_norm": 0.2958989143371582,
|
| 659 |
+
"learning_rate": 4.600000000000001e-05,
|
| 660 |
+
"loss": 0.4627,
|
| 661 |
+
"step": 920
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.2950039651070579,
|
| 665 |
+
"grad_norm": 0.23167885839939117,
|
| 666 |
+
"learning_rate": 4.6500000000000005e-05,
|
| 667 |
+
"loss": 0.4636,
|
| 668 |
+
"step": 930
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.29817605075337034,
|
| 672 |
+
"grad_norm": 0.2608003318309784,
|
| 673 |
+
"learning_rate": 4.7e-05,
|
| 674 |
+
"loss": 0.4628,
|
| 675 |
+
"step": 940
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.3013481363996828,
|
| 679 |
+
"grad_norm": 0.2806340456008911,
|
| 680 |
+
"learning_rate": 4.75e-05,
|
| 681 |
+
"loss": 0.4624,
|
| 682 |
+
"step": 950
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.3045202220459952,
|
| 686 |
+
"grad_norm": 0.2206449806690216,
|
| 687 |
+
"learning_rate": 4.8e-05,
|
| 688 |
+
"loss": 0.4594,
|
| 689 |
+
"step": 960
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.3076923076923077,
|
| 693 |
+
"grad_norm": 0.2883213460445404,
|
| 694 |
+
"learning_rate": 4.85e-05,
|
| 695 |
+
"loss": 0.4612,
|
| 696 |
+
"step": 970
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.31086439333862015,
|
| 700 |
+
"grad_norm": 0.23365403711795807,
|
| 701 |
+
"learning_rate": 4.9e-05,
|
| 702 |
+
"loss": 0.463,
|
| 703 |
+
"step": 980
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.3140364789849326,
|
| 707 |
+
"grad_norm": 0.28844788670539856,
|
| 708 |
+
"learning_rate": 4.9500000000000004e-05,
|
| 709 |
+
"loss": 0.4612,
|
| 710 |
+
"step": 990
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.317208564631245,
|
| 714 |
+
"grad_norm": 0.26974356174468994,
|
| 715 |
+
"learning_rate": 5e-05,
|
| 716 |
+
"loss": 0.4641,
|
| 717 |
+
"step": 1000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.317208564631245,
|
| 721 |
+
"eval_loss": 0.45910927653312683,
|
| 722 |
+
"eval_runtime": 419.1696,
|
| 723 |
+
"eval_samples_per_second": 2.457,
|
| 724 |
+
"eval_steps_per_second": 1.229,
|
| 725 |
+
"step": 1000
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.3203806502775575,
|
| 729 |
+
"grad_norm": 0.23146626353263855,
|
| 730 |
+
"learning_rate": 4.9997336102492574e-05,
|
| 731 |
+
"loss": 0.4583,
|
| 732 |
+
"step": 1010
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.32355273592386996,
|
| 736 |
+
"grad_norm": 0.24651704728603363,
|
| 737 |
+
"learning_rate": 4.9989344977678285e-05,
|
| 738 |
+
"loss": 0.461,
|
| 739 |
+
"step": 1020
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.3267248215701824,
|
| 743 |
+
"grad_norm": 0.2995850741863251,
|
| 744 |
+
"learning_rate": 4.997602832856013e-05,
|
| 745 |
+
"loss": 0.4583,
|
| 746 |
+
"step": 1030
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.32989690721649484,
|
| 750 |
+
"grad_norm": 0.2886098325252533,
|
| 751 |
+
"learning_rate": 4.995738899307319e-05,
|
| 752 |
+
"loss": 0.4565,
|
| 753 |
+
"step": 1040
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.3330689928628073,
|
| 757 |
+
"grad_norm": 0.2367076575756073,
|
| 758 |
+
"learning_rate": 4.99334309434798e-05,
|
| 759 |
+
"loss": 0.4598,
|
| 760 |
+
"step": 1050
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.3362410785091198,
|
| 764 |
+
"grad_norm": 0.29130271077156067,
|
| 765 |
+
"learning_rate": 4.990415928552305e-05,
|
| 766 |
+
"loss": 0.4564,
|
| 767 |
+
"step": 1060
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.3394131641554322,
|
| 771 |
+
"grad_norm": 0.23702391982078552,
|
| 772 |
+
"learning_rate": 4.9869580257338685e-05,
|
| 773 |
+
"loss": 0.4567,
|
| 774 |
+
"step": 1070
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.34258524980174465,
|
| 778 |
+
"grad_norm": 0.24812078475952148,
|
| 779 |
+
"learning_rate": 4.9829701228125655e-05,
|
| 780 |
+
"loss": 0.4587,
|
| 781 |
+
"step": 1080
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.3457573354480571,
|
| 785 |
+
"grad_norm": 0.2195880264043808,
|
| 786 |
+
"learning_rate": 4.9784530696575684e-05,
|
| 787 |
+
"loss": 0.4578,
|
| 788 |
+
"step": 1090
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.34892942109436953,
|
| 792 |
+
"grad_norm": 0.2531008720397949,
|
| 793 |
+
"learning_rate": 4.973407828906208e-05,
|
| 794 |
+
"loss": 0.4551,
|
| 795 |
+
"step": 1100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.352101506740682,
|
| 799 |
+
"grad_norm": 0.2672252655029297,
|
| 800 |
+
"learning_rate": 4.967835475758825e-05,
|
| 801 |
+
"loss": 0.4596,
|
| 802 |
+
"step": 1110
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 0.35527359238699446,
|
| 806 |
+
"grad_norm": 0.2659941613674164,
|
| 807 |
+
"learning_rate": 4.961737197749633e-05,
|
| 808 |
+
"loss": 0.4536,
|
| 809 |
+
"step": 1120
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.3584456780333069,
|
| 813 |
+
"grad_norm": 0.2517182230949402,
|
| 814 |
+
"learning_rate": 4.955114294493639e-05,
|
| 815 |
+
"loss": 0.4579,
|
| 816 |
+
"step": 1130
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.36161776367961934,
|
| 820 |
+
"grad_norm": 0.2654310166835785,
|
| 821 |
+
"learning_rate": 4.947968177409681e-05,
|
| 822 |
+
"loss": 0.4548,
|
| 823 |
+
"step": 1140
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.3647898493259318,
|
| 827 |
+
"grad_norm": 0.36614441871643066,
|
| 828 |
+
"learning_rate": 4.940300369419637e-05,
|
| 829 |
+
"loss": 0.4575,
|
| 830 |
+
"step": 1150
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.3679619349722443,
|
| 834 |
+
"grad_norm": 0.2583513855934143,
|
| 835 |
+
"learning_rate": 4.9321125046238756e-05,
|
| 836 |
+
"loss": 0.4598,
|
| 837 |
+
"step": 1160
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.3711340206185567,
|
| 841 |
+
"grad_norm": 0.23981873691082,
|
| 842 |
+
"learning_rate": 4.923406327953008e-05,
|
| 843 |
+
"loss": 0.4546,
|
| 844 |
+
"step": 1170
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.37430610626486915,
|
| 848 |
+
"grad_norm": 0.23798957467079163,
|
| 849 |
+
"learning_rate": 4.9141836947960165e-05,
|
| 850 |
+
"loss": 0.458,
|
| 851 |
+
"step": 1180
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 0.3774781919111816,
|
| 855 |
+
"grad_norm": 0.2931083142757416,
|
| 856 |
+
"learning_rate": 4.904446570604863e-05,
|
| 857 |
+
"loss": 0.4543,
|
| 858 |
+
"step": 1190
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.38065027755749403,
|
| 862 |
+
"grad_norm": 0.27312615513801575,
|
| 863 |
+
"learning_rate": 4.8941970304756144e-05,
|
| 864 |
+
"loss": 0.4552,
|
| 865 |
+
"step": 1200
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.3838223632038065,
|
| 869 |
+
"grad_norm": 0.2659391164779663,
|
| 870 |
+
"learning_rate": 4.883437258706224e-05,
|
| 871 |
+
"loss": 0.4583,
|
| 872 |
+
"step": 1210
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.38699444885011897,
|
| 876 |
+
"grad_norm": 0.23010990023612976,
|
| 877 |
+
"learning_rate": 4.8721695483310275e-05,
|
| 878 |
+
"loss": 0.4564,
|
| 879 |
+
"step": 1220
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.3901665344964314,
|
| 883 |
+
"grad_norm": 0.2815942168235779,
|
| 884 |
+
"learning_rate": 4.860396300632072e-05,
|
| 885 |
+
"loss": 0.4548,
|
| 886 |
+
"step": 1230
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.39333862014274384,
|
| 890 |
+
"grad_norm": 0.36561158299446106,
|
| 891 |
+
"learning_rate": 4.848120024627372e-05,
|
| 892 |
+
"loss": 0.4528,
|
| 893 |
+
"step": 1240
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 0.3965107057890563,
|
| 897 |
+
"grad_norm": 0.25015613436698914,
|
| 898 |
+
"learning_rate": 4.835343336536212e-05,
|
| 899 |
+
"loss": 0.4567,
|
| 900 |
+
"step": 1250
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 0.3996827914353688,
|
| 904 |
+
"grad_norm": 0.2373395711183548,
|
| 905 |
+
"learning_rate": 4.822068959221598e-05,
|
| 906 |
+
"loss": 0.4544,
|
| 907 |
+
"step": 1260
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 0.4028548770816812,
|
| 911 |
+
"grad_norm": 0.26331648230552673,
|
| 912 |
+
"learning_rate": 4.80829972160998e-05,
|
| 913 |
+
"loss": 0.4532,
|
| 914 |
+
"step": 1270
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.40602696272799366,
|
| 918 |
+
"grad_norm": 0.2622736692428589,
|
| 919 |
+
"learning_rate": 4.794038558088378e-05,
|
| 920 |
+
"loss": 0.4535,
|
| 921 |
+
"step": 1280
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.4091990483743061,
|
| 925 |
+
"grad_norm": 0.1999824345111847,
|
| 926 |
+
"learning_rate": 4.779288507879031e-05,
|
| 927 |
+
"loss": 0.4568,
|
| 928 |
+
"step": 1290
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 0.41237113402061853,
|
| 932 |
+
"grad_norm": 0.24161754548549652,
|
| 933 |
+
"learning_rate": 4.764052714391695e-05,
|
| 934 |
+
"loss": 0.4534,
|
| 935 |
+
"step": 1300
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 0.41554321966693103,
|
| 939 |
+
"grad_norm": 0.2368980497121811,
|
| 940 |
+
"learning_rate": 4.7483344245537545e-05,
|
| 941 |
+
"loss": 0.4558,
|
| 942 |
+
"step": 1310
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 0.41871530531324347,
|
| 946 |
+
"grad_norm": 0.1948440819978714,
|
| 947 |
+
"learning_rate": 4.7321369881182584e-05,
|
| 948 |
+
"loss": 0.4564,
|
| 949 |
+
"step": 1320
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 0.4218873909595559,
|
| 953 |
+
"grad_norm": 0.2509821355342865,
|
| 954 |
+
"learning_rate": 4.715463856950053e-05,
|
| 955 |
+
"loss": 0.4558,
|
| 956 |
+
"step": 1330
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.42505947660586835,
|
| 960 |
+
"grad_norm": 0.23150520026683807,
|
| 961 |
+
"learning_rate": 4.698318584290141e-05,
|
| 962 |
+
"loss": 0.4538,
|
| 963 |
+
"step": 1340
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.4282315622521808,
|
| 967 |
+
"grad_norm": 0.22577635943889618,
|
| 968 |
+
"learning_rate": 4.680704823998452e-05,
|
| 969 |
+
"loss": 0.4571,
|
| 970 |
+
"step": 1350
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 0.4314036478984933,
|
| 974 |
+
"grad_norm": 0.22882512211799622,
|
| 975 |
+
"learning_rate": 4.6626263297751546e-05,
|
| 976 |
+
"loss": 0.453,
|
| 977 |
+
"step": 1360
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.4345757335448057,
|
| 981 |
+
"grad_norm": 0.1979285478591919,
|
| 982 |
+
"learning_rate": 4.644086954360708e-05,
|
| 983 |
+
"loss": 0.4542,
|
| 984 |
+
"step": 1370
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.43774781919111816,
|
| 988 |
+
"grad_norm": 0.2554062008857727,
|
| 989 |
+
"learning_rate": 4.625090648714786e-05,
|
| 990 |
+
"loss": 0.4561,
|
| 991 |
+
"step": 1380
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 0.4409199048374306,
|
| 995 |
+
"grad_norm": 0.2330465316772461,
|
| 996 |
+
"learning_rate": 4.60564146117429e-05,
|
| 997 |
+
"loss": 0.4525,
|
| 998 |
+
"step": 1390
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.44409199048374304,
|
| 1002 |
+
"grad_norm": 0.22269243001937866,
|
| 1003 |
+
"learning_rate": 4.585743536590599e-05,
|
| 1004 |
+
"loss": 0.4544,
|
| 1005 |
+
"step": 1400
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.44726407613005553,
|
| 1009 |
+
"grad_norm": 0.2291986495256424,
|
| 1010 |
+
"learning_rate": 4.565401115446246e-05,
|
| 1011 |
+
"loss": 0.4548,
|
| 1012 |
+
"step": 1410
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 0.45043616177636797,
|
| 1016 |
+
"grad_norm": 0.21675272285938263,
|
| 1017 |
+
"learning_rate": 4.5446185329512314e-05,
|
| 1018 |
+
"loss": 0.4518,
|
| 1019 |
+
"step": 1420
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 0.4536082474226804,
|
| 1023 |
+
"grad_norm": 0.21207275986671448,
|
| 1024 |
+
"learning_rate": 4.5234002181191304e-05,
|
| 1025 |
+
"loss": 0.4511,
|
| 1026 |
+
"step": 1430
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 0.45678033306899285,
|
| 1030 |
+
"grad_norm": 0.22509776055812836,
|
| 1031 |
+
"learning_rate": 4.5017506928232247e-05,
|
| 1032 |
+
"loss": 0.4518,
|
| 1033 |
+
"step": 1440
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 0.4599524187153053,
|
| 1037 |
+
"grad_norm": 0.23429358005523682,
|
| 1038 |
+
"learning_rate": 4.4796745708328294e-05,
|
| 1039 |
+
"loss": 0.4556,
|
| 1040 |
+
"step": 1450
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.4631245043616178,
|
| 1044 |
+
"grad_norm": 0.18599826097488403,
|
| 1045 |
+
"learning_rate": 4.457176556830054e-05,
|
| 1046 |
+
"loss": 0.4538,
|
| 1047 |
+
"step": 1460
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.4662965900079302,
|
| 1051 |
+
"grad_norm": 0.2500051259994507,
|
| 1052 |
+
"learning_rate": 4.434261445407172e-05,
|
| 1053 |
+
"loss": 0.4538,
|
| 1054 |
+
"step": 1470
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 0.46946867565424266,
|
| 1058 |
+
"grad_norm": 0.213278666138649,
|
| 1059 |
+
"learning_rate": 4.410934120044838e-05,
|
| 1060 |
+
"loss": 0.4545,
|
| 1061 |
+
"step": 1480
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 0.4726407613005551,
|
| 1065 |
+
"grad_norm": 0.2551578879356384,
|
| 1066 |
+
"learning_rate": 4.387199552071366e-05,
|
| 1067 |
+
"loss": 0.4512,
|
| 1068 |
+
"step": 1490
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 0.47581284694686754,
|
| 1072 |
+
"grad_norm": 0.23356303572654724,
|
| 1073 |
+
"learning_rate": 4.3630627996032706e-05,
|
| 1074 |
+
"loss": 0.4514,
|
| 1075 |
+
"step": 1500
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 0.47581284694686754,
|
| 1079 |
+
"eval_loss": 0.4515945613384247,
|
| 1080 |
+
"eval_runtime": 293.5383,
|
| 1081 |
+
"eval_samples_per_second": 3.509,
|
| 1082 |
+
"eval_steps_per_second": 1.754,
|
| 1083 |
+
"step": 1500
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.47898493259318004,
|
| 1087 |
+
"grad_norm": 0.22167843580245972,
|
| 1088 |
+
"learning_rate": 4.3385290064673316e-05,
|
| 1089 |
+
"loss": 0.4529,
|
| 1090 |
+
"step": 1510
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 0.4821570182394925,
|
| 1094 |
+
"grad_norm": 0.2573811709880829,
|
| 1095 |
+
"learning_rate": 4.31360340110438e-05,
|
| 1096 |
+
"loss": 0.4517,
|
| 1097 |
+
"step": 1520
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 0.4853291038858049,
|
| 1101 |
+
"grad_norm": 0.20118848979473114,
|
| 1102 |
+
"learning_rate": 4.288291295455055e-05,
|
| 1103 |
+
"loss": 0.4538,
|
| 1104 |
+
"step": 1530
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 0.48850118953211735,
|
| 1108 |
+
"grad_norm": 0.24297507107257843,
|
| 1109 |
+
"learning_rate": 4.262598083827769e-05,
|
| 1110 |
+
"loss": 0.4519,
|
| 1111 |
+
"step": 1540
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 0.4916732751784298,
|
| 1115 |
+
"grad_norm": 0.18671482801437378,
|
| 1116 |
+
"learning_rate": 4.236529241749114e-05,
|
| 1117 |
+
"loss": 0.4515,
|
| 1118 |
+
"step": 1550
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.4948453608247423,
|
| 1122 |
+
"grad_norm": 0.20851927995681763,
|
| 1123 |
+
"learning_rate": 4.2100903247969644e-05,
|
| 1124 |
+
"loss": 0.4526,
|
| 1125 |
+
"step": 1560
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.4980174464710547,
|
| 1129 |
+
"grad_norm": 0.23694059252738953,
|
| 1130 |
+
"learning_rate": 4.1832869674165204e-05,
|
| 1131 |
+
"loss": 0.4519,
|
| 1132 |
+
"step": 1570
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 0.5011895321173672,
|
| 1136 |
+
"grad_norm": 0.3004760146141052,
|
| 1137 |
+
"learning_rate": 4.156124881719533e-05,
|
| 1138 |
+
"loss": 0.4493,
|
| 1139 |
+
"step": 1580
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 0.5043616177636796,
|
| 1143 |
+
"grad_norm": 0.20880846679210663,
|
| 1144 |
+
"learning_rate": 4.1286098562669925e-05,
|
| 1145 |
+
"loss": 0.4508,
|
| 1146 |
+
"step": 1590
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.507533703409992,
|
| 1150 |
+
"grad_norm": 0.2033112347126007,
|
| 1151 |
+
"learning_rate": 4.1007477548355185e-05,
|
| 1152 |
+
"loss": 0.4522,
|
| 1153 |
+
"step": 1600
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 0.5107057890563045,
|
| 1157 |
+
"grad_norm": 0.24385987222194672,
|
| 1158 |
+
"learning_rate": 4.072544515167714e-05,
|
| 1159 |
+
"loss": 0.45,
|
| 1160 |
+
"step": 1610
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.5138778747026169,
|
| 1164 |
+
"grad_norm": 0.19505774974822998,
|
| 1165 |
+
"learning_rate": 4.044006147706768e-05,
|
| 1166 |
+
"loss": 0.4528,
|
| 1167 |
+
"step": 1620
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.5170499603489295,
|
| 1171 |
+
"grad_norm": 0.20980720221996307,
|
| 1172 |
+
"learning_rate": 4.0151387343155545e-05,
|
| 1173 |
+
"loss": 0.4512,
|
| 1174 |
+
"step": 1630
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 0.5202220459952419,
|
| 1178 |
+
"grad_norm": 0.2582622766494751,
|
| 1179 |
+
"learning_rate": 3.985948426980521e-05,
|
| 1180 |
+
"loss": 0.452,
|
| 1181 |
+
"step": 1640
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 0.5233941316415543,
|
| 1185 |
+
"grad_norm": 0.18902088701725006,
|
| 1186 |
+
"learning_rate": 3.956441446500624e-05,
|
| 1187 |
+
"loss": 0.4498,
|
| 1188 |
+
"step": 1650
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 0.5265662172878668,
|
| 1192 |
+
"grad_norm": 0.21405866742134094,
|
| 1193 |
+
"learning_rate": 3.926624081161604e-05,
|
| 1194 |
+
"loss": 0.4462,
|
| 1195 |
+
"step": 1660
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 0.5297383029341792,
|
| 1199 |
+
"grad_norm": 0.21340113878250122,
|
| 1200 |
+
"learning_rate": 3.896502685395876e-05,
|
| 1201 |
+
"loss": 0.449,
|
| 1202 |
+
"step": 1670
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.5329103885804917,
|
| 1206 |
+
"grad_norm": 0.2296970933675766,
|
| 1207 |
+
"learning_rate": 3.8660836784283275e-05,
|
| 1208 |
+
"loss": 0.4513,
|
| 1209 |
+
"step": 1680
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.5360824742268041,
|
| 1213 |
+
"grad_norm": 0.2143164724111557,
|
| 1214 |
+
"learning_rate": 3.835373542908308e-05,
|
| 1215 |
+
"loss": 0.4488,
|
| 1216 |
+
"step": 1690
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 0.5392545598731165,
|
| 1220 |
+
"grad_norm": 0.1945250779390335,
|
| 1221 |
+
"learning_rate": 3.804378823528093e-05,
|
| 1222 |
+
"loss": 0.4529,
|
| 1223 |
+
"step": 1700
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 0.542426645519429,
|
| 1227 |
+
"grad_norm": 0.20191755890846252,
|
| 1228 |
+
"learning_rate": 3.7731061256281394e-05,
|
| 1229 |
+
"loss": 0.4507,
|
| 1230 |
+
"step": 1710
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 0.5455987311657414,
|
| 1234 |
+
"grad_norm": 0.21345514059066772,
|
| 1235 |
+
"learning_rate": 3.7415621137894056e-05,
|
| 1236 |
+
"loss": 0.4487,
|
| 1237 |
+
"step": 1720
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 0.548770816812054,
|
| 1241 |
+
"grad_norm": 0.2062884271144867,
|
| 1242 |
+
"learning_rate": 3.709753510413052e-05,
|
| 1243 |
+
"loss": 0.4504,
|
| 1244 |
+
"step": 1730
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.5519429024583664,
|
| 1248 |
+
"grad_norm": 0.20377105474472046,
|
| 1249 |
+
"learning_rate": 3.6776870942878196e-05,
|
| 1250 |
+
"loss": 0.4489,
|
| 1251 |
+
"step": 1740
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.5551149881046789,
|
| 1255 |
+
"grad_norm": 0.19985787570476532,
|
| 1256 |
+
"learning_rate": 3.645369699145387e-05,
|
| 1257 |
+
"loss": 0.4497,
|
| 1258 |
+
"step": 1750
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 0.5582870737509913,
|
| 1262 |
+
"grad_norm": 0.2308078110218048,
|
| 1263 |
+
"learning_rate": 3.6128082122040224e-05,
|
| 1264 |
+
"loss": 0.453,
|
| 1265 |
+
"step": 1760
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 0.5614591593973037,
|
| 1269 |
+
"grad_norm": 0.19868697226047516,
|
| 1270 |
+
"learning_rate": 3.5800095727008395e-05,
|
| 1271 |
+
"loss": 0.4512,
|
| 1272 |
+
"step": 1770
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 0.5646312450436162,
|
| 1276 |
+
"grad_norm": 0.20231026411056519,
|
| 1277 |
+
"learning_rate": 3.54698077041296e-05,
|
| 1278 |
+
"loss": 0.4482,
|
| 1279 |
+
"step": 1780
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 0.5678033306899286,
|
| 1283 |
+
"grad_norm": 0.21103453636169434,
|
| 1284 |
+
"learning_rate": 3.51372884416791e-05,
|
| 1285 |
+
"loss": 0.452,
|
| 1286 |
+
"step": 1790
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 0.570975416336241,
|
| 1290 |
+
"grad_norm": 0.21350406110286713,
|
| 1291 |
+
"learning_rate": 3.4802608803435646e-05,
|
| 1292 |
+
"loss": 0.4518,
|
| 1293 |
+
"step": 1800
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 0.5741475019825535,
|
| 1297 |
+
"grad_norm": 0.2247006744146347,
|
| 1298 |
+
"learning_rate": 3.446584011357957e-05,
|
| 1299 |
+
"loss": 0.451,
|
| 1300 |
+
"step": 1810
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 0.5773195876288659,
|
| 1304 |
+
"grad_norm": 0.20522333681583405,
|
| 1305 |
+
"learning_rate": 3.412705414149276e-05,
|
| 1306 |
+
"loss": 0.4491,
|
| 1307 |
+
"step": 1820
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 0.5804916732751785,
|
| 1311 |
+
"grad_norm": 0.20249220728874207,
|
| 1312 |
+
"learning_rate": 3.3786323086463736e-05,
|
| 1313 |
+
"loss": 0.4508,
|
| 1314 |
+
"step": 1830
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.5836637589214909,
|
| 1318 |
+
"grad_norm": 0.17756901681423187,
|
| 1319 |
+
"learning_rate": 3.3443719562301147e-05,
|
| 1320 |
+
"loss": 0.4493,
|
| 1321 |
+
"step": 1840
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 0.5868358445678034,
|
| 1325 |
+
"grad_norm": 0.19246827065944672,
|
| 1326 |
+
"learning_rate": 3.309931658185892e-05,
|
| 1327 |
+
"loss": 0.4501,
|
| 1328 |
+
"step": 1850
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.5900079302141158,
|
| 1332 |
+
"grad_norm": 0.20991584658622742,
|
| 1333 |
+
"learning_rate": 3.275318754147636e-05,
|
| 1334 |
+
"loss": 0.4478,
|
| 1335 |
+
"step": 1860
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.5931800158604282,
|
| 1339 |
+
"grad_norm": 0.21498160064220428,
|
| 1340 |
+
"learning_rate": 3.240540620533649e-05,
|
| 1341 |
+
"loss": 0.4454,
|
| 1342 |
+
"step": 1870
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 0.5963521015067407,
|
| 1346 |
+
"grad_norm": 0.199667289853096,
|
| 1347 |
+
"learning_rate": 3.205604668974607e-05,
|
| 1348 |
+
"loss": 0.4469,
|
| 1349 |
+
"step": 1880
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 0.5995241871530531,
|
| 1353 |
+
"grad_norm": 0.21041718125343323,
|
| 1354 |
+
"learning_rate": 3.170518344734051e-05,
|
| 1355 |
+
"loss": 0.4517,
|
| 1356 |
+
"step": 1890
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 0.6026962727993656,
|
| 1360 |
+
"grad_norm": 0.2041397988796234,
|
| 1361 |
+
"learning_rate": 3.135289125121718e-05,
|
| 1362 |
+
"loss": 0.4518,
|
| 1363 |
+
"step": 1900
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 0.605868358445678,
|
| 1367 |
+
"grad_norm": 0.18860045075416565,
|
| 1368 |
+
"learning_rate": 3.0999245179000325e-05,
|
| 1369 |
+
"loss": 0.4511,
|
| 1370 |
+
"step": 1910
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.6090404440919904,
|
| 1374 |
+
"grad_norm": 0.20879918336868286,
|
| 1375 |
+
"learning_rate": 3.064432059684117e-05,
|
| 1376 |
+
"loss": 0.4504,
|
| 1377 |
+
"step": 1920
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.612212529738303,
|
| 1381 |
+
"grad_norm": 0.18158119916915894,
|
| 1382 |
+
"learning_rate": 3.0288193143356484e-05,
|
| 1383 |
+
"loss": 0.4501,
|
| 1384 |
+
"step": 1930
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 0.6153846153846154,
|
| 1388 |
+
"grad_norm": 0.20405800640583038,
|
| 1389 |
+
"learning_rate": 2.9930938713509125e-05,
|
| 1390 |
+
"loss": 0.4478,
|
| 1391 |
+
"step": 1940
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 0.6185567010309279,
|
| 1395 |
+
"grad_norm": 0.19368775188922882,
|
| 1396 |
+
"learning_rate": 2.9572633442433917e-05,
|
| 1397 |
+
"loss": 0.449,
|
| 1398 |
+
"step": 1950
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 0.6217287866772403,
|
| 1402 |
+
"grad_norm": 0.19516830146312714,
|
| 1403 |
+
"learning_rate": 2.9213353689212337e-05,
|
| 1404 |
+
"loss": 0.4509,
|
| 1405 |
+
"step": 1960
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 0.6249008723235527,
|
| 1409 |
+
"grad_norm": 0.2207994908094406,
|
| 1410 |
+
"learning_rate": 2.8853176020599504e-05,
|
| 1411 |
+
"loss": 0.4524,
|
| 1412 |
+
"step": 1970
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 0.6280729579698652,
|
| 1416 |
+
"grad_norm": 0.20896515250205994,
|
| 1417 |
+
"learning_rate": 2.849217719470691e-05,
|
| 1418 |
+
"loss": 0.447,
|
| 1419 |
+
"step": 1980
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.6312450436161776,
|
| 1423 |
+
"grad_norm": 0.19302915036678314,
|
| 1424 |
+
"learning_rate": 2.8130434144644364e-05,
|
| 1425 |
+
"loss": 0.4462,
|
| 1426 |
+
"step": 1990
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 0.63441712926249,
|
| 1430 |
+
"grad_norm": 0.21044501662254333,
|
| 1431 |
+
"learning_rate": 2.776802396212461e-05,
|
| 1432 |
+
"loss": 0.4522,
|
| 1433 |
+
"step": 2000
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 0.63441712926249,
|
| 1437 |
+
"eval_loss": 0.44816499948501587,
|
| 1438 |
+
"eval_runtime": 293.5621,
|
| 1439 |
+
"eval_samples_per_second": 3.509,
|
| 1440 |
+
"eval_steps_per_second": 1.754,
|
| 1441 |
+
"step": 2000
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.6375892149088025,
|
| 1445 |
+
"grad_norm": 0.2075473517179489,
|
| 1446 |
+
"learning_rate": 2.7405023881034204e-05,
|
| 1447 |
+
"loss": 0.4487,
|
| 1448 |
+
"step": 2010
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 0.640761300555115,
|
| 1452 |
+
"grad_norm": 0.1936168372631073,
|
| 1453 |
+
"learning_rate": 2.7041511260974028e-05,
|
| 1454 |
+
"loss": 0.4485,
|
| 1455 |
+
"step": 2020
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 0.6439333862014275,
|
| 1459 |
+
"grad_norm": 0.2282039374113083,
|
| 1460 |
+
"learning_rate": 2.66775635707731e-05,
|
| 1461 |
+
"loss": 0.4478,
|
| 1462 |
+
"step": 2030
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 0.6471054718477399,
|
| 1466 |
+
"grad_norm": 0.18652349710464478,
|
| 1467 |
+
"learning_rate": 2.6313258371978994e-05,
|
| 1468 |
+
"loss": 0.449,
|
| 1469 |
+
"step": 2040
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 0.6502775574940524,
|
| 1473 |
+
"grad_norm": 0.20407763123512268,
|
| 1474 |
+
"learning_rate": 2.5948673302328587e-05,
|
| 1475 |
+
"loss": 0.4486,
|
| 1476 |
+
"step": 2050
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 0.6534496431403648,
|
| 1480 |
+
"grad_norm": 0.22227470576763153,
|
| 1481 |
+
"learning_rate": 2.558388605920255e-05,
|
| 1482 |
+
"loss": 0.4506,
|
| 1483 |
+
"step": 2060
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 0.6566217287866772,
|
| 1487 |
+
"grad_norm": 0.20888665318489075,
|
| 1488 |
+
"learning_rate": 2.5218974383067085e-05,
|
| 1489 |
+
"loss": 0.4487,
|
| 1490 |
+
"step": 2070
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 0.6597938144329897,
|
| 1494 |
+
"grad_norm": 0.19797295331954956,
|
| 1495 |
+
"learning_rate": 2.4854016040906572e-05,
|
| 1496 |
+
"loss": 0.4521,
|
| 1497 |
+
"step": 2080
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 0.6629659000793021,
|
| 1501 |
+
"grad_norm": 0.22293215990066528,
|
| 1502 |
+
"learning_rate": 2.4489088809650425e-05,
|
| 1503 |
+
"loss": 0.449,
|
| 1504 |
+
"step": 2090
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 0.6661379857256146,
|
| 1508 |
+
"grad_norm": 0.2310066670179367,
|
| 1509 |
+
"learning_rate": 2.4124270459598007e-05,
|
| 1510 |
+
"loss": 0.4479,
|
| 1511 |
+
"step": 2100
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 0.669310071371927,
|
| 1515 |
+
"grad_norm": 0.19995397329330444,
|
| 1516 |
+
"learning_rate": 2.375963873784478e-05,
|
| 1517 |
+
"loss": 0.4504,
|
| 1518 |
+
"step": 2110
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 0.6724821570182395,
|
| 1522 |
+
"grad_norm": 0.18815317749977112,
|
| 1523 |
+
"learning_rate": 2.3395271351713515e-05,
|
| 1524 |
+
"loss": 0.4501,
|
| 1525 |
+
"step": 2120
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 0.675654242664552,
|
| 1529 |
+
"grad_norm": 0.19861653447151184,
|
| 1530 |
+
"learning_rate": 2.303124595219395e-05,
|
| 1531 |
+
"loss": 0.4506,
|
| 1532 |
+
"step": 2130
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 0.6788263283108644,
|
| 1536 |
+
"grad_norm": 0.17454959452152252,
|
| 1537 |
+
"learning_rate": 2.2667640117394442e-05,
|
| 1538 |
+
"loss": 0.4453,
|
| 1539 |
+
"step": 2140
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 0.6819984139571769,
|
| 1543 |
+
"grad_norm": 0.19067569077014923,
|
| 1544 |
+
"learning_rate": 2.230453133600915e-05,
|
| 1545 |
+
"loss": 0.4466,
|
| 1546 |
+
"step": 2150
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 0.6851704996034893,
|
| 1550 |
+
"grad_norm": 0.1979096531867981,
|
| 1551 |
+
"learning_rate": 2.1941996990804288e-05,
|
| 1552 |
+
"loss": 0.4479,
|
| 1553 |
+
"step": 2160
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 0.6883425852498017,
|
| 1557 |
+
"grad_norm": 0.18306080996990204,
|
| 1558 |
+
"learning_rate": 2.1580114342126933e-05,
|
| 1559 |
+
"loss": 0.4482,
|
| 1560 |
+
"step": 2170
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 0.6915146708961142,
|
| 1564 |
+
"grad_norm": 0.18473172187805176,
|
| 1565 |
+
"learning_rate": 2.1218960511439952e-05,
|
| 1566 |
+
"loss": 0.4468,
|
| 1567 |
+
"step": 2180
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 0.6946867565424266,
|
| 1571 |
+
"grad_norm": 0.17530548572540283,
|
| 1572 |
+
"learning_rate": 2.0858612464886505e-05,
|
| 1573 |
+
"loss": 0.449,
|
| 1574 |
+
"step": 2190
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 0.6978588421887391,
|
| 1578 |
+
"grad_norm": 0.1818351149559021,
|
| 1579 |
+
"learning_rate": 2.0499146996887618e-05,
|
| 1580 |
+
"loss": 0.4458,
|
| 1581 |
+
"step": 2200
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 0.7010309278350515,
|
| 1585 |
+
"grad_norm": 0.19750525057315826,
|
| 1586 |
+
"learning_rate": 2.0140640713776438e-05,
|
| 1587 |
+
"loss": 0.4489,
|
| 1588 |
+
"step": 2210
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 0.704203013481364,
|
| 1592 |
+
"grad_norm": 0.20060710608959198,
|
| 1593 |
+
"learning_rate": 1.978317001747248e-05,
|
| 1594 |
+
"loss": 0.4482,
|
| 1595 |
+
"step": 2220
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 0.7073750991276765,
|
| 1599 |
+
"grad_norm": 0.19399641454219818,
|
| 1600 |
+
"learning_rate": 1.942681108919949e-05,
|
| 1601 |
+
"loss": 0.4494,
|
| 1602 |
+
"step": 2230
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 0.7105471847739889,
|
| 1606 |
+
"grad_norm": 0.2124367654323578,
|
| 1607 |
+
"learning_rate": 1.9071639873250334e-05,
|
| 1608 |
+
"loss": 0.4491,
|
| 1609 |
+
"step": 2240
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 0.7137192704203014,
|
| 1613 |
+
"grad_norm": 0.2064894288778305,
|
| 1614 |
+
"learning_rate": 1.871773206080236e-05,
|
| 1615 |
+
"loss": 0.4433,
|
| 1616 |
+
"step": 2250
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.7168913560666138,
|
| 1620 |
+
"grad_norm": 0.20361967384815216,
|
| 1621 |
+
"learning_rate": 1.836516307378671e-05,
|
| 1622 |
+
"loss": 0.4482,
|
| 1623 |
+
"step": 2260
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 0.7200634417129262,
|
| 1627 |
+
"grad_norm": 0.22753477096557617,
|
| 1628 |
+
"learning_rate": 1.801400804881507e-05,
|
| 1629 |
+
"loss": 0.4434,
|
| 1630 |
+
"step": 2270
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 0.7232355273592387,
|
| 1634 |
+
"grad_norm": 0.2242439240217209,
|
| 1635 |
+
"learning_rate": 1.766434182116708e-05,
|
| 1636 |
+
"loss": 0.4487,
|
| 1637 |
+
"step": 2280
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 0.7264076130055511,
|
| 1641 |
+
"grad_norm": 0.1731296181678772,
|
| 1642 |
+
"learning_rate": 1.7316238908842127e-05,
|
| 1643 |
+
"loss": 0.4456,
|
| 1644 |
+
"step": 2290
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 0.7295796986518636,
|
| 1648 |
+
"grad_norm": 0.19372820854187012,
|
| 1649 |
+
"learning_rate": 1.6969773496678647e-05,
|
| 1650 |
+
"loss": 0.446,
|
| 1651 |
+
"step": 2300
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 0.732751784298176,
|
| 1655 |
+
"grad_norm": 0.21722117066383362,
|
| 1656 |
+
"learning_rate": 1.662501942054447e-05,
|
| 1657 |
+
"loss": 0.4465,
|
| 1658 |
+
"step": 2310
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 0.7359238699444886,
|
| 1662 |
+
"grad_norm": 0.24181324243545532,
|
| 1663 |
+
"learning_rate": 1.6282050151601518e-05,
|
| 1664 |
+
"loss": 0.4496,
|
| 1665 |
+
"step": 2320
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 0.739095955590801,
|
| 1669 |
+
"grad_norm": 0.2091435343027115,
|
| 1670 |
+
"learning_rate": 1.594093878064825e-05,
|
| 1671 |
+
"loss": 0.4469,
|
| 1672 |
+
"step": 2330
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 0.7422680412371134,
|
| 1676 |
+
"grad_norm": 0.19474650919437408,
|
| 1677 |
+
"learning_rate": 1.5601758002543137e-05,
|
| 1678 |
+
"loss": 0.4465,
|
| 1679 |
+
"step": 2340
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 0.7454401268834259,
|
| 1683 |
+
"grad_norm": 0.21081089973449707,
|
| 1684 |
+
"learning_rate": 1.5264580100712507e-05,
|
| 1685 |
+
"loss": 0.4449,
|
| 1686 |
+
"step": 2350
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 0.7486122125297383,
|
| 1690 |
+
"grad_norm": 0.21220536530017853,
|
| 1691 |
+
"learning_rate": 1.4929476931746167e-05,
|
| 1692 |
+
"loss": 0.4461,
|
| 1693 |
+
"step": 2360
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 0.7517842981760507,
|
| 1697 |
+
"grad_norm": 0.19944216310977936,
|
| 1698 |
+
"learning_rate": 1.4596519910083825e-05,
|
| 1699 |
+
"loss": 0.4476,
|
| 1700 |
+
"step": 2370
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 0.7549563838223632,
|
| 1704 |
+
"grad_norm": 0.19958476722240448,
|
| 1705 |
+
"learning_rate": 1.4265779992795893e-05,
|
| 1706 |
+
"loss": 0.4467,
|
| 1707 |
+
"step": 2380
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 0.7581284694686756,
|
| 1711 |
+
"grad_norm": 0.22247722744941711,
|
| 1712 |
+
"learning_rate": 1.3937327664461672e-05,
|
| 1713 |
+
"loss": 0.4427,
|
| 1714 |
+
"step": 2390
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 0.7613005551149881,
|
| 1718 |
+
"grad_norm": 0.19884702563285828,
|
| 1719 |
+
"learning_rate": 1.361123292214826e-05,
|
| 1720 |
+
"loss": 0.4498,
|
| 1721 |
+
"step": 2400
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 0.7644726407613005,
|
| 1725 |
+
"grad_norm": 0.17867441475391388,
|
| 1726 |
+
"learning_rate": 1.3287565260493357e-05,
|
| 1727 |
+
"loss": 0.4484,
|
| 1728 |
+
"step": 2410
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 0.767644726407613,
|
| 1732 |
+
"grad_norm": 0.17917264997959137,
|
| 1733 |
+
"learning_rate": 1.2966393656895134e-05,
|
| 1734 |
+
"loss": 0.4439,
|
| 1735 |
+
"step": 2420
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 0.7708168120539255,
|
| 1739 |
+
"grad_norm": 0.24791833758354187,
|
| 1740 |
+
"learning_rate": 1.2647786556812332e-05,
|
| 1741 |
+
"loss": 0.4463,
|
| 1742 |
+
"step": 2430
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 0.7739888977002379,
|
| 1746 |
+
"grad_norm": 0.1971379816532135,
|
| 1747 |
+
"learning_rate": 1.2331811859177722e-05,
|
| 1748 |
+
"loss": 0.4475,
|
| 1749 |
+
"step": 2440
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 0.7771609833465504,
|
| 1753 |
+
"grad_norm": 0.19460198283195496,
|
| 1754 |
+
"learning_rate": 1.2018536901928079e-05,
|
| 1755 |
+
"loss": 0.4468,
|
| 1756 |
+
"step": 2450
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 0.7803330689928628,
|
| 1760 |
+
"grad_norm": 0.21018314361572266,
|
| 1761 |
+
"learning_rate": 1.1708028447653613e-05,
|
| 1762 |
+
"loss": 0.4489,
|
| 1763 |
+
"step": 2460
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 0.7835051546391752,
|
| 1767 |
+
"grad_norm": 0.18419425189495087,
|
| 1768 |
+
"learning_rate": 1.1400352669370115e-05,
|
| 1769 |
+
"loss": 0.4469,
|
| 1770 |
+
"step": 2470
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 0.7866772402854877,
|
| 1774 |
+
"grad_norm": 0.192842036485672,
|
| 1775 |
+
"learning_rate": 1.1095575136416695e-05,
|
| 1776 |
+
"loss": 0.4457,
|
| 1777 |
+
"step": 2480
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 0.7898493259318001,
|
| 1781 |
+
"grad_norm": 0.21617653965950012,
|
| 1782 |
+
"learning_rate": 1.0793760800482179e-05,
|
| 1783 |
+
"loss": 0.4511,
|
| 1784 |
+
"step": 2490
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 0.7930214115781126,
|
| 1788 |
+
"grad_norm": 0.1891343593597412,
|
| 1789 |
+
"learning_rate": 1.0494973981763146e-05,
|
| 1790 |
+
"loss": 0.4436,
|
| 1791 |
+
"step": 2500
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 0.7930214115781126,
|
| 1795 |
+
"eval_loss": 0.44588717818260193,
|
| 1796 |
+
"eval_runtime": 293.564,
|
| 1797 |
+
"eval_samples_per_second": 3.509,
|
| 1798 |
+
"eval_steps_per_second": 1.754,
|
| 1799 |
+
"step": 2500
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 0.796193497224425,
|
| 1803 |
+
"grad_norm": 0.193357452750206,
|
| 1804 |
+
"learning_rate": 1.0199278355256522e-05,
|
| 1805 |
+
"loss": 0.4436,
|
| 1806 |
+
"step": 2510
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 0.7993655828707376,
|
| 1810 |
+
"grad_norm": 0.19498059153556824,
|
| 1811 |
+
"learning_rate": 9.906736937189697e-06,
|
| 1812 |
+
"loss": 0.4457,
|
| 1813 |
+
"step": 2520
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 0.80253766851705,
|
| 1817 |
+
"grad_norm": 0.2086293250322342,
|
| 1818 |
+
"learning_rate": 9.61741207159104e-06,
|
| 1819 |
+
"loss": 0.4488,
|
| 1820 |
+
"step": 2530
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 0.8057097541633624,
|
| 1824 |
+
"grad_norm": 0.1872304081916809,
|
| 1825 |
+
"learning_rate": 9.331365417003601e-06,
|
| 1826 |
+
"loss": 0.4436,
|
| 1827 |
+
"step": 2540
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 0.8088818398096749,
|
| 1831 |
+
"grad_norm": 0.196652352809906,
|
| 1832 |
+
"learning_rate": 9.048657933345e-06,
|
| 1833 |
+
"loss": 0.4448,
|
| 1834 |
+
"step": 2550
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 0.8120539254559873,
|
| 1838 |
+
"grad_norm": 0.21112217009067535,
|
| 1839 |
+
"learning_rate": 8.769349868916118e-06,
|
| 1840 |
+
"loss": 0.4481,
|
| 1841 |
+
"step": 2560
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 0.8152260111022998,
|
| 1845 |
+
"grad_norm": 0.19768798351287842,
|
| 1846 |
+
"learning_rate": 8.493500747561488e-06,
|
| 1847 |
+
"loss": 0.4493,
|
| 1848 |
+
"step": 2570
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 0.8183980967486122,
|
| 1852 |
+
"grad_norm": 0.19622166454792023,
|
| 1853 |
+
"learning_rate": 8.221169355984052e-06,
|
| 1854 |
+
"loss": 0.4448,
|
| 1855 |
+
"step": 2580
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 0.8215701823949246,
|
| 1859 |
+
"grad_norm": 0.2044159471988678,
|
| 1860 |
+
"learning_rate": 7.952413731217025e-06,
|
| 1861 |
+
"loss": 0.4472,
|
| 1862 |
+
"step": 2590
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 0.8247422680412371,
|
| 1866 |
+
"grad_norm": 0.206452876329422,
|
| 1867 |
+
"learning_rate": 7.687291148255527e-06,
|
| 1868 |
+
"loss": 0.4452,
|
| 1869 |
+
"step": 2600
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 0.8279143536875495,
|
| 1873 |
+
"grad_norm": 0.20858535170555115,
|
| 1874 |
+
"learning_rate": 7.425858107850578e-06,
|
| 1875 |
+
"loss": 0.4419,
|
| 1876 |
+
"step": 2610
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 0.8310864393338621,
|
| 1880 |
+
"grad_norm": 0.18986332416534424,
|
| 1881 |
+
"learning_rate": 7.168170324468171e-06,
|
| 1882 |
+
"loss": 0.449,
|
| 1883 |
+
"step": 2620
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 0.8342585249801745,
|
| 1887 |
+
"grad_norm": 0.21051757037639618,
|
| 1888 |
+
"learning_rate": 6.9142827144158066e-06,
|
| 1889 |
+
"loss": 0.4468,
|
| 1890 |
+
"step": 2630
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 0.8374306106264869,
|
| 1894 |
+
"grad_norm": 0.20453451573848724,
|
| 1895 |
+
"learning_rate": 6.66424938413921e-06,
|
| 1896 |
+
"loss": 0.4471,
|
| 1897 |
+
"step": 2640
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 0.8406026962727994,
|
| 1901 |
+
"grad_norm": 0.18574683368206024,
|
| 1902 |
+
"learning_rate": 6.418123618691607e-06,
|
| 1903 |
+
"loss": 0.4473,
|
| 1904 |
+
"step": 2650
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 0.8437747819191118,
|
| 1908 |
+
"grad_norm": 0.2309403419494629,
|
| 1909 |
+
"learning_rate": 6.175957870378043e-06,
|
| 1910 |
+
"loss": 0.4477,
|
| 1911 |
+
"step": 2660
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 0.8469468675654243,
|
| 1915 |
+
"grad_norm": 0.19485293328762054,
|
| 1916 |
+
"learning_rate": 5.937803747577186e-06,
|
| 1917 |
+
"loss": 0.4443,
|
| 1918 |
+
"step": 2670
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 0.8501189532117367,
|
| 1922 |
+
"grad_norm": 0.17677082121372223,
|
| 1923 |
+
"learning_rate": 5.7037120037429645e-06,
|
| 1924 |
+
"loss": 0.4501,
|
| 1925 |
+
"step": 2680
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 0.8532910388580491,
|
| 1929 |
+
"grad_norm": 0.18835794925689697,
|
| 1930 |
+
"learning_rate": 5.473732526588407e-06,
|
| 1931 |
+
"loss": 0.4451,
|
| 1932 |
+
"step": 2690
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 0.8564631245043616,
|
| 1936 |
+
"grad_norm": 0.2117832899093628,
|
| 1937 |
+
"learning_rate": 5.2479143274539954e-06,
|
| 1938 |
+
"loss": 0.4464,
|
| 1939 |
+
"step": 2700
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 0.8596352101506741,
|
| 1943 |
+
"grad_norm": 0.18720006942749023,
|
| 1944 |
+
"learning_rate": 5.026305530862749e-06,
|
| 1945 |
+
"loss": 0.4484,
|
| 1946 |
+
"step": 2710
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 0.8628072957969866,
|
| 1950 |
+
"grad_norm": 0.17910288274288177,
|
| 1951 |
+
"learning_rate": 4.80895336426434e-06,
|
| 1952 |
+
"loss": 0.4457,
|
| 1953 |
+
"step": 2720
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 0.865979381443299,
|
| 1957 |
+
"grad_norm": 0.1965160220861435,
|
| 1958 |
+
"learning_rate": 4.595904147970356e-06,
|
| 1959 |
+
"loss": 0.4455,
|
| 1960 |
+
"step": 2730
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 0.8691514670896114,
|
| 1964 |
+
"grad_norm": 0.16718564927577972,
|
| 1965 |
+
"learning_rate": 4.3872032852828955e-06,
|
| 1966 |
+
"loss": 0.4455,
|
| 1967 |
+
"step": 2740
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 0.8723235527359239,
|
| 1971 |
+
"grad_norm": 0.19451048970222473,
|
| 1972 |
+
"learning_rate": 4.182895252818589e-06,
|
| 1973 |
+
"loss": 0.4454,
|
| 1974 |
+
"step": 2750
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 0.8754956383822363,
|
| 1978 |
+
"grad_norm": 0.2160135954618454,
|
| 1979 |
+
"learning_rate": 3.983023591030113e-06,
|
| 1980 |
+
"loss": 0.4462,
|
| 1981 |
+
"step": 2760
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 0.8786677240285488,
|
| 1985 |
+
"grad_norm": 0.1994054913520813,
|
| 1986 |
+
"learning_rate": 3.7876308949271995e-06,
|
| 1987 |
+
"loss": 0.4455,
|
| 1988 |
+
"step": 2770
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 0.8818398096748612,
|
| 1992 |
+
"grad_norm": 0.20801877975463867,
|
| 1993 |
+
"learning_rate": 3.5967588049991313e-06,
|
| 1994 |
+
"loss": 0.4465,
|
| 1995 |
+
"step": 2780
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.8850118953211736,
|
| 1999 |
+
"grad_norm": 0.20264749228954315,
|
| 2000 |
+
"learning_rate": 3.410447998340688e-06,
|
| 2001 |
+
"loss": 0.4451,
|
| 2002 |
+
"step": 2790
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 0.8881839809674861,
|
| 2006 |
+
"grad_norm": 0.173824280500412,
|
| 2007 |
+
"learning_rate": 3.2287381799833428e-06,
|
| 2008 |
+
"loss": 0.4453,
|
| 2009 |
+
"step": 2800
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 0.8913560666137986,
|
| 2013 |
+
"grad_norm": 0.18437042832374573,
|
| 2014 |
+
"learning_rate": 3.051668074433667e-06,
|
| 2015 |
+
"loss": 0.449,
|
| 2016 |
+
"step": 2810
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 0.8945281522601111,
|
| 2020 |
+
"grad_norm": 0.17420920729637146,
|
| 2021 |
+
"learning_rate": 2.8792754174206905e-06,
|
| 2022 |
+
"loss": 0.4453,
|
| 2023 |
+
"step": 2820
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 0.8977002379064235,
|
| 2027 |
+
"grad_norm": 0.18584947288036346,
|
| 2028 |
+
"learning_rate": 2.7115969478539562e-06,
|
| 2029 |
+
"loss": 0.4446,
|
| 2030 |
+
"step": 2830
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 0.9008723235527359,
|
| 2034 |
+
"grad_norm": 0.18006384372711182,
|
| 2035 |
+
"learning_rate": 2.5486683999940335e-06,
|
| 2036 |
+
"loss": 0.4449,
|
| 2037 |
+
"step": 2840
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 0.9040444091990484,
|
| 2041 |
+
"grad_norm": 0.19664862751960754,
|
| 2042 |
+
"learning_rate": 2.3905244958371263e-06,
|
| 2043 |
+
"loss": 0.4463,
|
| 2044 |
+
"step": 2850
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 0.9072164948453608,
|
| 2048 |
+
"grad_norm": 0.20159141719341278,
|
| 2049 |
+
"learning_rate": 2.2371989377154013e-06,
|
| 2050 |
+
"loss": 0.444,
|
| 2051 |
+
"step": 2860
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 0.9103885804916733,
|
| 2055 |
+
"grad_norm": 0.23178640007972717,
|
| 2056 |
+
"learning_rate": 2.088724401114625e-06,
|
| 2057 |
+
"loss": 0.4472,
|
| 2058 |
+
"step": 2870
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 0.9135606661379857,
|
| 2062 |
+
"grad_norm": 0.18047955632209778,
|
| 2063 |
+
"learning_rate": 1.9451325277106413e-06,
|
| 2064 |
+
"loss": 0.4489,
|
| 2065 |
+
"step": 2880
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 0.9167327517842981,
|
| 2069 |
+
"grad_norm": 0.18607738614082336,
|
| 2070 |
+
"learning_rate": 1.8064539186261387e-06,
|
| 2071 |
+
"loss": 0.4489,
|
| 2072 |
+
"step": 2890
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 0.9199048374306106,
|
| 2076 |
+
"grad_norm": 0.19285809993743896,
|
| 2077 |
+
"learning_rate": 1.6727181279092036e-06,
|
| 2078 |
+
"loss": 0.4448,
|
| 2079 |
+
"step": 2900
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 0.9230769230769231,
|
| 2083 |
+
"grad_norm": 0.1946035474538803,
|
| 2084 |
+
"learning_rate": 1.5439536562349976e-06,
|
| 2085 |
+
"loss": 0.4474,
|
| 2086 |
+
"step": 2910
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 0.9262490087232356,
|
| 2090 |
+
"grad_norm": 0.1947924941778183,
|
| 2091 |
+
"learning_rate": 1.4201879448319355e-06,
|
| 2092 |
+
"loss": 0.4465,
|
| 2093 |
+
"step": 2920
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 0.929421094369548,
|
| 2097 |
+
"grad_norm": 0.17740875482559204,
|
| 2098 |
+
"learning_rate": 1.301447369633621e-06,
|
| 2099 |
+
"loss": 0.4475,
|
| 2100 |
+
"step": 2930
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 0.9325931800158604,
|
| 2104 |
+
"grad_norm": 0.18701794743537903,
|
| 2105 |
+
"learning_rate": 1.187757235657841e-06,
|
| 2106 |
+
"loss": 0.4488,
|
| 2107 |
+
"step": 2940
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 0.9357652656621729,
|
| 2111 |
+
"grad_norm": 0.19531536102294922,
|
| 2112 |
+
"learning_rate": 1.0791417716137565e-06,
|
| 2113 |
+
"loss": 0.4458,
|
| 2114 |
+
"step": 2950
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 0.9389373513084853,
|
| 2118 |
+
"grad_norm": 0.17441792786121368,
|
| 2119 |
+
"learning_rate": 9.756241247384807e-07,
|
| 2120 |
+
"loss": 0.4426,
|
| 2121 |
+
"step": 2960
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 0.9421094369547978,
|
| 2125 |
+
"grad_norm": 0.20463122427463531,
|
| 2126 |
+
"learning_rate": 8.772263558641386e-07,
|
| 2127 |
+
"loss": 0.4485,
|
| 2128 |
+
"step": 2970
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 0.9452815226011102,
|
| 2132 |
+
"grad_norm": 0.18601520359516144,
|
| 2133 |
+
"learning_rate": 7.839694347164223e-07,
|
| 2134 |
+
"loss": 0.4421,
|
| 2135 |
+
"step": 2980
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 0.9484536082474226,
|
| 2139 |
+
"grad_norm": 0.18209311366081238,
|
| 2140 |
+
"learning_rate": 6.958732354457209e-07,
|
| 2141 |
+
"loss": 0.4499,
|
| 2142 |
+
"step": 2990
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 0.9516256938937351,
|
| 2146 |
+
"grad_norm": 0.1675465852022171,
|
| 2147 |
+
"learning_rate": 6.129565323916813e-07,
|
| 2148 |
+
"loss": 0.4463,
|
| 2149 |
+
"step": 3000
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 0.9516256938937351,
|
| 2153 |
+
"eval_loss": 0.44520583748817444,
|
| 2154 |
+
"eval_runtime": 293.5596,
|
| 2155 |
+
"eval_samples_per_second": 3.509,
|
| 2156 |
+
"eval_steps_per_second": 1.754,
|
| 2157 |
+
"step": 3000
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 0.9547977795400476,
|
| 2161 |
+
"grad_norm": 0.21267291903495789,
|
| 2162 |
+
"learning_rate": 5.352369960821946e-07,
|
| 2163 |
+
"loss": 0.4428,
|
| 2164 |
+
"step": 3010
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 0.9579698651863601,
|
| 2168 |
+
"grad_norm": 0.19138512015342712,
|
| 2169 |
+
"learning_rate": 4.627311894675856e-07,
|
| 2170 |
+
"loss": 0.446,
|
| 2171 |
+
"step": 3020
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 0.9611419508326725,
|
| 2175 |
+
"grad_norm": 0.19043755531311035,
|
| 2176 |
+
"learning_rate": 3.954545643908514e-07,
|
| 2177 |
+
"loss": 0.4446,
|
| 2178 |
+
"step": 3030
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 0.964314036478985,
|
| 2182 |
+
"grad_norm": 0.1850534826517105,
|
| 2183 |
+
"learning_rate": 3.3342145829469983e-07,
|
| 2184 |
+
"loss": 0.4435,
|
| 2185 |
+
"step": 3040
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 0.9674861221252974,
|
| 2189 |
+
"grad_norm": 0.21564887464046478,
|
| 2190 |
+
"learning_rate": 2.7664509116607506e-07,
|
| 2191 |
+
"loss": 0.4451,
|
| 2192 |
+
"step": 3050
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 0.9706582077716098,
|
| 2196 |
+
"grad_norm": 0.1970747411251068,
|
| 2197 |
+
"learning_rate": 2.251375627187996e-07,
|
| 2198 |
+
"loss": 0.4465,
|
| 2199 |
+
"step": 3060
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 0.9738302934179223,
|
| 2203 |
+
"grad_norm": 0.1968993842601776,
|
| 2204 |
+
"learning_rate": 1.789098498150066e-07,
|
| 2205 |
+
"loss": 0.4479,
|
| 2206 |
+
"step": 3070
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 0.9770023790642347,
|
| 2210 |
+
"grad_norm": 0.18383654952049255,
|
| 2211 |
+
"learning_rate": 1.3797180412583322e-07,
|
| 2212 |
+
"loss": 0.4472,
|
| 2213 |
+
"step": 3080
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 0.9801744647105471,
|
| 2217 |
+
"grad_norm": 0.18183210492134094,
|
| 2218 |
+
"learning_rate": 1.0233215003190577e-07,
|
| 2219 |
+
"loss": 0.4484,
|
| 2220 |
+
"step": 3090
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 0.9833465503568596,
|
| 2224 |
+
"grad_norm": 0.18428853154182434,
|
| 2225 |
+
"learning_rate": 7.199848276408238e-08,
|
| 2226 |
+
"loss": 0.4445,
|
| 2227 |
+
"step": 3100
|
| 2228 |
+
},
|
| 2229 |
+
{
|
| 2230 |
+
"epoch": 0.9865186360031721,
|
| 2231 |
+
"grad_norm": 0.21674080193042755,
|
| 2232 |
+
"learning_rate": 4.6977266784811736e-08,
|
| 2233 |
+
"loss": 0.4465,
|
| 2234 |
+
"step": 3110
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 0.9896907216494846,
|
| 2238 |
+
"grad_norm": 0.2023499608039856,
|
| 2239 |
+
"learning_rate": 2.7273834410485034e-08,
|
| 2240 |
+
"loss": 0.4448,
|
| 2241 |
+
"step": 3120
|
| 2242 |
+
},
|
| 2243 |
+
{
|
| 2244 |
+
"epoch": 0.992862807295797,
|
| 2245 |
+
"grad_norm": 0.19151511788368225,
|
| 2246 |
+
"learning_rate": 1.2892384675056158e-08,
|
| 2247 |
+
"loss": 0.4461,
|
| 2248 |
+
"step": 3130
|
| 2249 |
+
},
|
| 2250 |
+
{
|
| 2251 |
+
"epoch": 0.9960348929421095,
|
| 2252 |
+
"grad_norm": 0.16607537865638733,
|
| 2253 |
+
"learning_rate": 3.8359824351685836e-09,
|
| 2254 |
+
"loss": 0.4453,
|
| 2255 |
+
"step": 3140
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 0.9992069785884219,
|
| 2259 |
+
"grad_norm": 0.21372340619564056,
|
| 2260 |
+
"learning_rate": 1.0655771701395534e-10,
|
| 2261 |
+
"loss": 0.448,
|
| 2262 |
+
"step": 3150
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 0.9998413957176844,
|
| 2266 |
+
"step": 3152,
|
| 2267 |
+
"total_flos": 2.0499162760961065e+18,
|
| 2268 |
+
"train_loss": 0.5496831483344742,
|
| 2269 |
+
"train_runtime": 71646.9274,
|
| 2270 |
+
"train_samples_per_second": 0.704,
|
| 2271 |
+
"train_steps_per_second": 0.044
|
| 2272 |
+
}
|
| 2273 |
+
],
|
| 2274 |
+
"logging_steps": 10,
|
| 2275 |
+
"max_steps": 3152,
|
| 2276 |
+
"num_input_tokens_seen": 0,
|
| 2277 |
+
"num_train_epochs": 1,
|
| 2278 |
+
"save_steps": 500,
|
| 2279 |
+
"total_flos": 2.0499162760961065e+18,
|
| 2280 |
+
"train_batch_size": 4,
|
| 2281 |
+
"trial_name": null,
|
| 2282 |
+
"trial_params": null
|
| 2283 |
+
}
|