Upload multi_teacher_fantasy_alchemy/checkpoint-3600/trainer_state.json with huggingface_hub
Browse files
multi_teacher_fantasy_alchemy/checkpoint-3600/trainer_state.json
ADDED
|
@@ -0,0 +1,2554 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.9580936729663105,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3600,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.008216926869350863,
|
| 14 |
+
"grad_norm": 19.83480665335734,
|
| 15 |
+
"learning_rate": 9.240246406570842e-07,
|
| 16 |
+
"loss": 8.2641,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.016433853738701727,
|
| 21 |
+
"grad_norm": 18.704025706899706,
|
| 22 |
+
"learning_rate": 1.9507186858316226e-06,
|
| 23 |
+
"loss": 8.1054,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.024650780608052588,
|
| 28 |
+
"grad_norm": 9.254313873894626,
|
| 29 |
+
"learning_rate": 2.9774127310061602e-06,
|
| 30 |
+
"loss": 7.4171,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.03286770747740345,
|
| 35 |
+
"grad_norm": 5.336886587772746,
|
| 36 |
+
"learning_rate": 4.004106776180698e-06,
|
| 37 |
+
"loss": 6.652,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.041084634346754315,
|
| 42 |
+
"grad_norm": 3.356383707656696,
|
| 43 |
+
"learning_rate": 5.030800821355237e-06,
|
| 44 |
+
"loss": 6.1747,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.049301561216105176,
|
| 49 |
+
"grad_norm": 3.3258958472365467,
|
| 50 |
+
"learning_rate": 6.057494866529775e-06,
|
| 51 |
+
"loss": 5.8018,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.05751848808545604,
|
| 56 |
+
"grad_norm": 4.0537058731983,
|
| 57 |
+
"learning_rate": 7.084188911704312e-06,
|
| 58 |
+
"loss": 5.3316,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.0657354149548069,
|
| 63 |
+
"grad_norm": 4.086247611679589,
|
| 64 |
+
"learning_rate": 8.11088295687885e-06,
|
| 65 |
+
"loss": 4.6534,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.07395234182415776,
|
| 70 |
+
"grad_norm": 6.521478113854127,
|
| 71 |
+
"learning_rate": 9.137577002053388e-06,
|
| 72 |
+
"loss": 3.924,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.08216926869350863,
|
| 77 |
+
"grad_norm": 5.719373874300257,
|
| 78 |
+
"learning_rate": 1.0164271047227925e-05,
|
| 79 |
+
"loss": 3.3464,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.09038619556285948,
|
| 84 |
+
"grad_norm": 3.578443884259996,
|
| 85 |
+
"learning_rate": 1.1190965092402464e-05,
|
| 86 |
+
"loss": 2.9684,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.09860312243221035,
|
| 91 |
+
"grad_norm": 5.144310812867023,
|
| 92 |
+
"learning_rate": 1.2217659137577003e-05,
|
| 93 |
+
"loss": 2.734,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.10682004930156122,
|
| 98 |
+
"grad_norm": 5.024178909720029,
|
| 99 |
+
"learning_rate": 1.324435318275154e-05,
|
| 100 |
+
"loss": 2.5647,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.11503697617091208,
|
| 105 |
+
"grad_norm": 7.639677795548909,
|
| 106 |
+
"learning_rate": 1.427104722792608e-05,
|
| 107 |
+
"loss": 2.4388,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.12325390304026294,
|
| 112 |
+
"grad_norm": 7.4036645978760225,
|
| 113 |
+
"learning_rate": 1.5297741273100617e-05,
|
| 114 |
+
"loss": 2.3493,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.1314708299096138,
|
| 119 |
+
"grad_norm": 7.12561810579708,
|
| 120 |
+
"learning_rate": 1.6324435318275156e-05,
|
| 121 |
+
"loss": 2.2697,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.13968775677896467,
|
| 126 |
+
"grad_norm": 9.222649571442956,
|
| 127 |
+
"learning_rate": 1.735112936344969e-05,
|
| 128 |
+
"loss": 2.1975,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.14790468364831552,
|
| 133 |
+
"grad_norm": 8.868924318680982,
|
| 134 |
+
"learning_rate": 1.837782340862423e-05,
|
| 135 |
+
"loss": 2.1464,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.1561216105176664,
|
| 140 |
+
"grad_norm": 11.22877634105566,
|
| 141 |
+
"learning_rate": 1.940451745379877e-05,
|
| 142 |
+
"loss": 2.0921,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.16433853738701726,
|
| 147 |
+
"grad_norm": 14.840583315298584,
|
| 148 |
+
"learning_rate": 2.0431211498973308e-05,
|
| 149 |
+
"loss": 2.0518,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.1725554642563681,
|
| 154 |
+
"grad_norm": 12.147153984880905,
|
| 155 |
+
"learning_rate": 2.1457905544147843e-05,
|
| 156 |
+
"loss": 1.9688,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.18077239112571897,
|
| 161 |
+
"grad_norm": 17.004164152742906,
|
| 162 |
+
"learning_rate": 2.2484599589322382e-05,
|
| 163 |
+
"loss": 1.9184,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.18898931799506985,
|
| 168 |
+
"grad_norm": 54.36301855761989,
|
| 169 |
+
"learning_rate": 2.351129363449692e-05,
|
| 170 |
+
"loss": 1.8491,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.1972062448644207,
|
| 175 |
+
"grad_norm": 13.223675766771267,
|
| 176 |
+
"learning_rate": 2.453798767967146e-05,
|
| 177 |
+
"loss": 1.8062,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.20542317173377156,
|
| 182 |
+
"grad_norm": 23.739675599865905,
|
| 183 |
+
"learning_rate": 2.5564681724845996e-05,
|
| 184 |
+
"loss": 1.7007,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.21364009860312244,
|
| 189 |
+
"grad_norm": 13.136946015186366,
|
| 190 |
+
"learning_rate": 2.6591375770020538e-05,
|
| 191 |
+
"loss": 1.6032,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.2218570254724733,
|
| 196 |
+
"grad_norm": 11.684546950836488,
|
| 197 |
+
"learning_rate": 2.7618069815195074e-05,
|
| 198 |
+
"loss": 1.4973,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.23007395234182415,
|
| 203 |
+
"grad_norm": 7.667957804275019,
|
| 204 |
+
"learning_rate": 2.8644763860369613e-05,
|
| 205 |
+
"loss": 1.3794,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.23829087921117503,
|
| 210 |
+
"grad_norm": 12.12659183227395,
|
| 211 |
+
"learning_rate": 2.9671457905544148e-05,
|
| 212 |
+
"loss": 1.3322,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.2465078060805259,
|
| 217 |
+
"grad_norm": 8.972817044811347,
|
| 218 |
+
"learning_rate": 3.069815195071869e-05,
|
| 219 |
+
"loss": 1.2359,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.25472473294987674,
|
| 224 |
+
"grad_norm": 4.3028554301503465,
|
| 225 |
+
"learning_rate": 3.172484599589323e-05,
|
| 226 |
+
"loss": 1.1852,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.2629416598192276,
|
| 231 |
+
"grad_norm": 5.3453439233449185,
|
| 232 |
+
"learning_rate": 3.2751540041067765e-05,
|
| 233 |
+
"loss": 1.1183,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.27115858668857845,
|
| 238 |
+
"grad_norm": 6.559451348524855,
|
| 239 |
+
"learning_rate": 3.37782340862423e-05,
|
| 240 |
+
"loss": 1.0608,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.27937551355792933,
|
| 245 |
+
"grad_norm": 6.374692391110214,
|
| 246 |
+
"learning_rate": 3.480492813141684e-05,
|
| 247 |
+
"loss": 1.0151,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.2875924404272802,
|
| 252 |
+
"grad_norm": 7.303560110424278,
|
| 253 |
+
"learning_rate": 3.583162217659138e-05,
|
| 254 |
+
"loss": 0.9712,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.29580936729663104,
|
| 259 |
+
"grad_norm": 7.0001296670132485,
|
| 260 |
+
"learning_rate": 3.6858316221765914e-05,
|
| 261 |
+
"loss": 0.926,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.3040262941659819,
|
| 266 |
+
"grad_norm": 6.545632008564383,
|
| 267 |
+
"learning_rate": 3.788501026694045e-05,
|
| 268 |
+
"loss": 0.8996,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.3122432210353328,
|
| 273 |
+
"grad_norm": 4.842087687844629,
|
| 274 |
+
"learning_rate": 3.891170431211499e-05,
|
| 275 |
+
"loss": 0.8519,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.32046014790468363,
|
| 280 |
+
"grad_norm": 5.689801303683759,
|
| 281 |
+
"learning_rate": 3.993839835728953e-05,
|
| 282 |
+
"loss": 0.8223,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3286770747740345,
|
| 287 |
+
"grad_norm": 6.971495473441899,
|
| 288 |
+
"learning_rate": 4.096509240246407e-05,
|
| 289 |
+
"loss": 0.7828,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.3368940016433854,
|
| 294 |
+
"grad_norm": 6.36358312425706,
|
| 295 |
+
"learning_rate": 4.1991786447638605e-05,
|
| 296 |
+
"loss": 0.7489,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.3451109285127362,
|
| 301 |
+
"grad_norm": 8.676168539975894,
|
| 302 |
+
"learning_rate": 4.301848049281315e-05,
|
| 303 |
+
"loss": 0.731,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.3533278553820871,
|
| 308 |
+
"grad_norm": 7.918859246440221,
|
| 309 |
+
"learning_rate": 4.404517453798768e-05,
|
| 310 |
+
"loss": 0.7064,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.36154478225143794,
|
| 315 |
+
"grad_norm": 8.806726830982097,
|
| 316 |
+
"learning_rate": 4.507186858316222e-05,
|
| 317 |
+
"loss": 0.6826,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.3697617091207888,
|
| 322 |
+
"grad_norm": 6.737818041811399,
|
| 323 |
+
"learning_rate": 4.6098562628336754e-05,
|
| 324 |
+
"loss": 0.6548,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.3779786359901397,
|
| 329 |
+
"grad_norm": 6.523855090754944,
|
| 330 |
+
"learning_rate": 4.7125256673511296e-05,
|
| 331 |
+
"loss": 0.6279,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.3861955628594905,
|
| 336 |
+
"grad_norm": 8.120644576403212,
|
| 337 |
+
"learning_rate": 4.815195071868583e-05,
|
| 338 |
+
"loss": 0.6037,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.3944124897288414,
|
| 343 |
+
"grad_norm": 5.5098094876695285,
|
| 344 |
+
"learning_rate": 4.9178644763860374e-05,
|
| 345 |
+
"loss": 0.5764,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.4026294165981923,
|
| 350 |
+
"grad_norm": 8.171893404017268,
|
| 351 |
+
"learning_rate": 5.020533880903491e-05,
|
| 352 |
+
"loss": 0.5662,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.4108463434675431,
|
| 357 |
+
"grad_norm": 5.545430210957229,
|
| 358 |
+
"learning_rate": 5.1232032854209445e-05,
|
| 359 |
+
"loss": 0.5416,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.419063270336894,
|
| 364 |
+
"grad_norm": 6.687063853266749,
|
| 365 |
+
"learning_rate": 5.225872689938398e-05,
|
| 366 |
+
"loss": 0.5087,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.4272801972062449,
|
| 371 |
+
"grad_norm": 4.47366213437882,
|
| 372 |
+
"learning_rate": 5.328542094455853e-05,
|
| 373 |
+
"loss": 0.4881,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.4354971240755957,
|
| 378 |
+
"grad_norm": 4.172813447018652,
|
| 379 |
+
"learning_rate": 5.4312114989733065e-05,
|
| 380 |
+
"loss": 0.4621,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.4437140509449466,
|
| 385 |
+
"grad_norm": 4.220516576800102,
|
| 386 |
+
"learning_rate": 5.53388090349076e-05,
|
| 387 |
+
"loss": 0.4487,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.4519309778142975,
|
| 392 |
+
"grad_norm": 4.709217957842771,
|
| 393 |
+
"learning_rate": 5.6365503080082136e-05,
|
| 394 |
+
"loss": 0.4246,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.4601479046836483,
|
| 399 |
+
"grad_norm": 3.6943640807724925,
|
| 400 |
+
"learning_rate": 5.739219712525667e-05,
|
| 401 |
+
"loss": 0.4002,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.4683648315529992,
|
| 406 |
+
"grad_norm": 4.324443086003805,
|
| 407 |
+
"learning_rate": 5.8418891170431214e-05,
|
| 408 |
+
"loss": 0.3769,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.47658175842235007,
|
| 413 |
+
"grad_norm": 4.712485210987473,
|
| 414 |
+
"learning_rate": 5.944558521560575e-05,
|
| 415 |
+
"loss": 0.3605,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.4847986852917009,
|
| 420 |
+
"grad_norm": 2.5269576298422347,
|
| 421 |
+
"learning_rate": 6.04722792607803e-05,
|
| 422 |
+
"loss": 0.3417,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.4930156121610518,
|
| 427 |
+
"grad_norm": 3.54537374699048,
|
| 428 |
+
"learning_rate": 6.149897330595483e-05,
|
| 429 |
+
"loss": 0.3316,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.5012325390304027,
|
| 434 |
+
"grad_norm": 3.530633503104792,
|
| 435 |
+
"learning_rate": 6.252566735112937e-05,
|
| 436 |
+
"loss": 0.313,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5094494658997535,
|
| 441 |
+
"grad_norm": 4.839997750327561,
|
| 442 |
+
"learning_rate": 6.35523613963039e-05,
|
| 443 |
+
"loss": 0.2974,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5176663927691043,
|
| 448 |
+
"grad_norm": 4.197641178153233,
|
| 449 |
+
"learning_rate": 6.457905544147844e-05,
|
| 450 |
+
"loss": 0.2745,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5258833196384552,
|
| 455 |
+
"grad_norm": 2.705058855982426,
|
| 456 |
+
"learning_rate": 6.560574948665298e-05,
|
| 457 |
+
"loss": 0.2615,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.5341002465078061,
|
| 462 |
+
"grad_norm": 2.8099029424585313,
|
| 463 |
+
"learning_rate": 6.663244353182751e-05,
|
| 464 |
+
"loss": 0.2503,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.5423171733771569,
|
| 469 |
+
"grad_norm": 2.5734835696398712,
|
| 470 |
+
"learning_rate": 6.765913757700205e-05,
|
| 471 |
+
"loss": 0.2415,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.5505341002465078,
|
| 476 |
+
"grad_norm": 2.7836726382586425,
|
| 477 |
+
"learning_rate": 6.86858316221766e-05,
|
| 478 |
+
"loss": 0.2344,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.5587510271158587,
|
| 483 |
+
"grad_norm": 2.603051010989796,
|
| 484 |
+
"learning_rate": 6.971252566735113e-05,
|
| 485 |
+
"loss": 0.2154,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.5669679539852095,
|
| 490 |
+
"grad_norm": 2.1107040737971587,
|
| 491 |
+
"learning_rate": 7.073921971252567e-05,
|
| 492 |
+
"loss": 0.1976,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.5751848808545604,
|
| 497 |
+
"grad_norm": 2.803095444807186,
|
| 498 |
+
"learning_rate": 7.17659137577002e-05,
|
| 499 |
+
"loss": 0.1864,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.5834018077239113,
|
| 504 |
+
"grad_norm": 2.1670666899861075,
|
| 505 |
+
"learning_rate": 7.279260780287475e-05,
|
| 506 |
+
"loss": 0.178,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.5916187345932621,
|
| 511 |
+
"grad_norm": 1.7921531158206865,
|
| 512 |
+
"learning_rate": 7.381930184804929e-05,
|
| 513 |
+
"loss": 0.1652,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.599835661462613,
|
| 518 |
+
"grad_norm": 1.479382404686151,
|
| 519 |
+
"learning_rate": 7.484599589322382e-05,
|
| 520 |
+
"loss": 0.1576,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.6080525883319639,
|
| 525 |
+
"grad_norm": 1.743395608685881,
|
| 526 |
+
"learning_rate": 7.587268993839836e-05,
|
| 527 |
+
"loss": 0.1498,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.6162695152013147,
|
| 532 |
+
"grad_norm": 1.8303806232104347,
|
| 533 |
+
"learning_rate": 7.689938398357291e-05,
|
| 534 |
+
"loss": 0.136,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.6244864420706656,
|
| 539 |
+
"grad_norm": 2.312871831212618,
|
| 540 |
+
"learning_rate": 7.792607802874744e-05,
|
| 541 |
+
"loss": 0.1274,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.6327033689400164,
|
| 546 |
+
"grad_norm": 1.6901573163360035,
|
| 547 |
+
"learning_rate": 7.895277207392198e-05,
|
| 548 |
+
"loss": 0.1191,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.6409202958093673,
|
| 553 |
+
"grad_norm": 1.6314268202265299,
|
| 554 |
+
"learning_rate": 7.997946611909651e-05,
|
| 555 |
+
"loss": 0.1114,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.6491372226787182,
|
| 560 |
+
"grad_norm": 1.4508574236430893,
|
| 561 |
+
"learning_rate": 8.100616016427105e-05,
|
| 562 |
+
"loss": 0.101,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.657354149548069,
|
| 567 |
+
"grad_norm": 1.5649749543554101,
|
| 568 |
+
"learning_rate": 8.203285420944559e-05,
|
| 569 |
+
"loss": 0.0985,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.6655710764174199,
|
| 574 |
+
"grad_norm": 2.6845550965354428,
|
| 575 |
+
"learning_rate": 8.305954825462012e-05,
|
| 576 |
+
"loss": 0.1027,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.6737880032867708,
|
| 581 |
+
"grad_norm": 2.7764588978262426,
|
| 582 |
+
"learning_rate": 8.408624229979466e-05,
|
| 583 |
+
"loss": 0.1035,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.6820049301561216,
|
| 588 |
+
"grad_norm": 4.830497904460127,
|
| 589 |
+
"learning_rate": 8.51129363449692e-05,
|
| 590 |
+
"loss": 0.0976,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.6902218570254725,
|
| 595 |
+
"grad_norm": 1.0331763322679437,
|
| 596 |
+
"learning_rate": 8.613963039014374e-05,
|
| 597 |
+
"loss": 0.0879,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.6984387838948234,
|
| 602 |
+
"grad_norm": 0.596374501379962,
|
| 603 |
+
"learning_rate": 8.716632443531828e-05,
|
| 604 |
+
"loss": 0.074,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.7066557107641742,
|
| 609 |
+
"grad_norm": 0.8196280038018171,
|
| 610 |
+
"learning_rate": 8.819301848049281e-05,
|
| 611 |
+
"loss": 0.0692,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.714872637633525,
|
| 616 |
+
"grad_norm": 0.9801217376138596,
|
| 617 |
+
"learning_rate": 8.921971252566735e-05,
|
| 618 |
+
"loss": 0.065,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.7230895645028759,
|
| 623 |
+
"grad_norm": 0.5051290053812035,
|
| 624 |
+
"learning_rate": 9.02464065708419e-05,
|
| 625 |
+
"loss": 0.0596,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.7313064913722268,
|
| 630 |
+
"grad_norm": 0.9667645246911198,
|
| 631 |
+
"learning_rate": 9.127310061601643e-05,
|
| 632 |
+
"loss": 0.0556,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.7395234182415776,
|
| 637 |
+
"grad_norm": 0.46438651540167103,
|
| 638 |
+
"learning_rate": 9.229979466119097e-05,
|
| 639 |
+
"loss": 0.0545,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.7477403451109285,
|
| 644 |
+
"grad_norm": 0.40547338931365595,
|
| 645 |
+
"learning_rate": 9.332648870636552e-05,
|
| 646 |
+
"loss": 0.0494,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.7559572719802794,
|
| 651 |
+
"grad_norm": 0.4441111869944399,
|
| 652 |
+
"learning_rate": 9.435318275154005e-05,
|
| 653 |
+
"loss": 0.0466,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.7641741988496302,
|
| 658 |
+
"grad_norm": 0.8183415391457746,
|
| 659 |
+
"learning_rate": 9.537987679671459e-05,
|
| 660 |
+
"loss": 0.0453,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.772391125718981,
|
| 665 |
+
"grad_norm": 0.5138559384321786,
|
| 666 |
+
"learning_rate": 9.640657084188912e-05,
|
| 667 |
+
"loss": 0.0436,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.780608052588332,
|
| 672 |
+
"grad_norm": 0.474231944626131,
|
| 673 |
+
"learning_rate": 9.743326488706366e-05,
|
| 674 |
+
"loss": 0.0415,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.7888249794576828,
|
| 679 |
+
"grad_norm": 0.4407732436458076,
|
| 680 |
+
"learning_rate": 9.84599589322382e-05,
|
| 681 |
+
"loss": 0.04,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.7970419063270336,
|
| 686 |
+
"grad_norm": 1.888951859317562,
|
| 687 |
+
"learning_rate": 9.948665297741273e-05,
|
| 688 |
+
"loss": 0.0401,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.8052588331963846,
|
| 693 |
+
"grad_norm": 0.5937342056949868,
|
| 694 |
+
"learning_rate": 9.999991965235563e-05,
|
| 695 |
+
"loss": 0.0408,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.8134757600657354,
|
| 700 |
+
"grad_norm": 0.6643097529303563,
|
| 701 |
+
"learning_rate": 9.999927687274992e-05,
|
| 702 |
+
"loss": 0.0458,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.8216926869350862,
|
| 707 |
+
"grad_norm": 0.38763729674702113,
|
| 708 |
+
"learning_rate": 9.999799132180184e-05,
|
| 709 |
+
"loss": 0.0364,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.8299096138044372,
|
| 714 |
+
"grad_norm": 0.3739679848963624,
|
| 715 |
+
"learning_rate": 9.999606301603793e-05,
|
| 716 |
+
"loss": 0.0332,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.838126540673788,
|
| 721 |
+
"grad_norm": 0.3960365233672132,
|
| 722 |
+
"learning_rate": 9.999349198024775e-05,
|
| 723 |
+
"loss": 0.0327,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.8463434675431388,
|
| 728 |
+
"grad_norm": 0.35789593639188133,
|
| 729 |
+
"learning_rate": 9.999027824748352e-05,
|
| 730 |
+
"loss": 0.0313,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.8545603944124898,
|
| 735 |
+
"grad_norm": 0.3328319306799121,
|
| 736 |
+
"learning_rate": 9.998642185905977e-05,
|
| 737 |
+
"loss": 0.03,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.8627773212818406,
|
| 742 |
+
"grad_norm": 0.397125770040445,
|
| 743 |
+
"learning_rate": 9.998192286455273e-05,
|
| 744 |
+
"loss": 0.0285,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.8709942481511914,
|
| 749 |
+
"grad_norm": 0.35895726006131834,
|
| 750 |
+
"learning_rate": 9.997678132179969e-05,
|
| 751 |
+
"loss": 0.0283,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.8792111750205424,
|
| 756 |
+
"grad_norm": 0.359528719136988,
|
| 757 |
+
"learning_rate": 9.997099729689839e-05,
|
| 758 |
+
"loss": 0.0272,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.8874281018898932,
|
| 763 |
+
"grad_norm": 0.31783947017752745,
|
| 764 |
+
"learning_rate": 9.996457086420598e-05,
|
| 765 |
+
"loss": 0.026,
|
| 766 |
+
"step": 1080
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.895645028759244,
|
| 770 |
+
"grad_norm": 0.24437306397975714,
|
| 771 |
+
"learning_rate": 9.995750210633822e-05,
|
| 772 |
+
"loss": 0.0252,
|
| 773 |
+
"step": 1090
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.903861955628595,
|
| 777 |
+
"grad_norm": 0.2869935821010678,
|
| 778 |
+
"learning_rate": 9.994979111416829e-05,
|
| 779 |
+
"loss": 0.0243,
|
| 780 |
+
"step": 1100
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.9120788824979458,
|
| 784 |
+
"grad_norm": 0.28967525917886927,
|
| 785 |
+
"learning_rate": 9.994143798682573e-05,
|
| 786 |
+
"loss": 0.0236,
|
| 787 |
+
"step": 1110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.9202958093672966,
|
| 791 |
+
"grad_norm": 0.4867480221470474,
|
| 792 |
+
"learning_rate": 9.993244283169513e-05,
|
| 793 |
+
"loss": 0.0232,
|
| 794 |
+
"step": 1120
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.9285127362366475,
|
| 798 |
+
"grad_norm": 0.27522967583273783,
|
| 799 |
+
"learning_rate": 9.99228057644147e-05,
|
| 800 |
+
"loss": 0.0229,
|
| 801 |
+
"step": 1130
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.9367296631059984,
|
| 805 |
+
"grad_norm": 0.2705586478283337,
|
| 806 |
+
"learning_rate": 9.991252690887485e-05,
|
| 807 |
+
"loss": 0.0217,
|
| 808 |
+
"step": 1140
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.9449465899753492,
|
| 812 |
+
"grad_norm": 0.3133390383629242,
|
| 813 |
+
"learning_rate": 9.990160639721657e-05,
|
| 814 |
+
"loss": 0.0209,
|
| 815 |
+
"step": 1150
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.9531635168447001,
|
| 819 |
+
"grad_norm": 0.3003241130339291,
|
| 820 |
+
"learning_rate": 9.989004436982974e-05,
|
| 821 |
+
"loss": 0.0204,
|
| 822 |
+
"step": 1160
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.961380443714051,
|
| 826 |
+
"grad_norm": 0.27998609610081276,
|
| 827 |
+
"learning_rate": 9.987784097535127e-05,
|
| 828 |
+
"loss": 0.02,
|
| 829 |
+
"step": 1170
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.9695973705834018,
|
| 833 |
+
"grad_norm": 0.3094626205699417,
|
| 834 |
+
"learning_rate": 9.986499637066331e-05,
|
| 835 |
+
"loss": 0.0198,
|
| 836 |
+
"step": 1180
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.9778142974527527,
|
| 840 |
+
"grad_norm": 0.311495504428665,
|
| 841 |
+
"learning_rate": 9.985151072089112e-05,
|
| 842 |
+
"loss": 0.0193,
|
| 843 |
+
"step": 1190
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.9860312243221035,
|
| 847 |
+
"grad_norm": 0.2810349187040472,
|
| 848 |
+
"learning_rate": 9.983738419940096e-05,
|
| 849 |
+
"loss": 0.0193,
|
| 850 |
+
"step": 1200
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.9942481511914544,
|
| 854 |
+
"grad_norm": 0.3107871046473083,
|
| 855 |
+
"learning_rate": 9.982261698779795e-05,
|
| 856 |
+
"loss": 0.0185,
|
| 857 |
+
"step": 1210
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.0024650780608053,
|
| 861 |
+
"grad_norm": 0.26519682492524527,
|
| 862 |
+
"learning_rate": 9.980720927592366e-05,
|
| 863 |
+
"loss": 0.0182,
|
| 864 |
+
"step": 1220
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.0106820049301561,
|
| 868 |
+
"grad_norm": 0.2391845678539309,
|
| 869 |
+
"learning_rate": 9.979116126185364e-05,
|
| 870 |
+
"loss": 0.0173,
|
| 871 |
+
"step": 1230
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.018898931799507,
|
| 875 |
+
"grad_norm": 0.26242451803570593,
|
| 876 |
+
"learning_rate": 9.977447315189493e-05,
|
| 877 |
+
"loss": 0.0166,
|
| 878 |
+
"step": 1240
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.0271158586688578,
|
| 882 |
+
"grad_norm": 0.22409653311325814,
|
| 883 |
+
"learning_rate": 9.975714516058344e-05,
|
| 884 |
+
"loss": 0.0162,
|
| 885 |
+
"step": 1250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.0353327855382086,
|
| 889 |
+
"grad_norm": 0.2343121137078575,
|
| 890 |
+
"learning_rate": 9.973917751068111e-05,
|
| 891 |
+
"loss": 0.0155,
|
| 892 |
+
"step": 1260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.0435497124075597,
|
| 896 |
+
"grad_norm": 0.21473180531882663,
|
| 897 |
+
"learning_rate": 9.972057043317308e-05,
|
| 898 |
+
"loss": 0.0152,
|
| 899 |
+
"step": 1270
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.0517666392769105,
|
| 903 |
+
"grad_norm": 0.1994724694575453,
|
| 904 |
+
"learning_rate": 9.970132416726474e-05,
|
| 905 |
+
"loss": 0.0151,
|
| 906 |
+
"step": 1280
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.0599835661462613,
|
| 910 |
+
"grad_norm": 0.1979357989976122,
|
| 911 |
+
"learning_rate": 9.968143896037864e-05,
|
| 912 |
+
"loss": 0.0147,
|
| 913 |
+
"step": 1290
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.0682004930156122,
|
| 917 |
+
"grad_norm": 0.23691205439082766,
|
| 918 |
+
"learning_rate": 9.966091506815128e-05,
|
| 919 |
+
"loss": 0.0144,
|
| 920 |
+
"step": 1300
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.076417419884963,
|
| 924 |
+
"grad_norm": 2.057726124603243,
|
| 925 |
+
"learning_rate": 9.963975275442989e-05,
|
| 926 |
+
"loss": 0.0175,
|
| 927 |
+
"step": 1310
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.0846343467543138,
|
| 931 |
+
"grad_norm": 0.3234261374936929,
|
| 932 |
+
"learning_rate": 9.961795229126896e-05,
|
| 933 |
+
"loss": 0.0165,
|
| 934 |
+
"step": 1320
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.0928512736236646,
|
| 938 |
+
"grad_norm": 0.2509885643136706,
|
| 939 |
+
"learning_rate": 9.959551395892681e-05,
|
| 940 |
+
"loss": 0.0149,
|
| 941 |
+
"step": 1330
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.1010682004930157,
|
| 945 |
+
"grad_norm": 0.19772103993153273,
|
| 946 |
+
"learning_rate": 9.957243804586197e-05,
|
| 947 |
+
"loss": 0.0141,
|
| 948 |
+
"step": 1340
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.1092851273623665,
|
| 952 |
+
"grad_norm": 0.23068483781938592,
|
| 953 |
+
"learning_rate": 9.954872484872944e-05,
|
| 954 |
+
"loss": 0.0134,
|
| 955 |
+
"step": 1350
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.1175020542317173,
|
| 959 |
+
"grad_norm": 0.21085093284095438,
|
| 960 |
+
"learning_rate": 9.952437467237687e-05,
|
| 961 |
+
"loss": 0.0133,
|
| 962 |
+
"step": 1360
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.1257189811010682,
|
| 966 |
+
"grad_norm": 0.21244911066205022,
|
| 967 |
+
"learning_rate": 9.949938782984072e-05,
|
| 968 |
+
"loss": 0.0131,
|
| 969 |
+
"step": 1370
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.133935907970419,
|
| 973 |
+
"grad_norm": 0.1841268164629968,
|
| 974 |
+
"learning_rate": 9.947376464234215e-05,
|
| 975 |
+
"loss": 0.0126,
|
| 976 |
+
"step": 1380
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.14215283483977,
|
| 980 |
+
"grad_norm": 0.18885789130639805,
|
| 981 |
+
"learning_rate": 9.944750543928297e-05,
|
| 982 |
+
"loss": 0.0123,
|
| 983 |
+
"step": 1390
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.1503697617091209,
|
| 987 |
+
"grad_norm": 0.18962637141794314,
|
| 988 |
+
"learning_rate": 9.942061055824127e-05,
|
| 989 |
+
"loss": 0.0121,
|
| 990 |
+
"step": 1400
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.1585866885784717,
|
| 994 |
+
"grad_norm": 0.20049383873744792,
|
| 995 |
+
"learning_rate": 9.939308034496728e-05,
|
| 996 |
+
"loss": 0.0121,
|
| 997 |
+
"step": 1410
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.1668036154478225,
|
| 1001 |
+
"grad_norm": 0.19041767566091838,
|
| 1002 |
+
"learning_rate": 9.936491515337874e-05,
|
| 1003 |
+
"loss": 0.0119,
|
| 1004 |
+
"step": 1420
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.1750205423171733,
|
| 1008 |
+
"grad_norm": 0.1995551393056727,
|
| 1009 |
+
"learning_rate": 9.933611534555644e-05,
|
| 1010 |
+
"loss": 0.0117,
|
| 1011 |
+
"step": 1430
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.1832374691865242,
|
| 1015 |
+
"grad_norm": 0.16768458374812184,
|
| 1016 |
+
"learning_rate": 9.930668129173956e-05,
|
| 1017 |
+
"loss": 0.0116,
|
| 1018 |
+
"step": 1440
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.191454396055875,
|
| 1022 |
+
"grad_norm": 0.18748344143729523,
|
| 1023 |
+
"learning_rate": 9.927661337032093e-05,
|
| 1024 |
+
"loss": 0.0115,
|
| 1025 |
+
"step": 1450
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.199671322925226,
|
| 1029 |
+
"grad_norm": 0.17618964617531263,
|
| 1030 |
+
"learning_rate": 9.924591196784205e-05,
|
| 1031 |
+
"loss": 0.0113,
|
| 1032 |
+
"step": 1460
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.2078882497945769,
|
| 1036 |
+
"grad_norm": 0.1915822510753366,
|
| 1037 |
+
"learning_rate": 9.92145774789883e-05,
|
| 1038 |
+
"loss": 0.011,
|
| 1039 |
+
"step": 1470
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.2161051766639277,
|
| 1043 |
+
"grad_norm": 0.22348016077011987,
|
| 1044 |
+
"learning_rate": 9.918261030658371e-05,
|
| 1045 |
+
"loss": 0.011,
|
| 1046 |
+
"step": 1480
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.2243221035332785,
|
| 1050 |
+
"grad_norm": 0.17910179979882987,
|
| 1051 |
+
"learning_rate": 9.915001086158588e-05,
|
| 1052 |
+
"loss": 0.0109,
|
| 1053 |
+
"step": 1490
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.2325390304026294,
|
| 1057 |
+
"grad_norm": 0.20954080994397253,
|
| 1058 |
+
"learning_rate": 9.911677956308068e-05,
|
| 1059 |
+
"loss": 0.0106,
|
| 1060 |
+
"step": 1500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.2407559572719804,
|
| 1064 |
+
"grad_norm": 0.18491793792738792,
|
| 1065 |
+
"learning_rate": 9.908291683827678e-05,
|
| 1066 |
+
"loss": 0.0103,
|
| 1067 |
+
"step": 1510
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.2489728841413312,
|
| 1071 |
+
"grad_norm": 0.1913002285241464,
|
| 1072 |
+
"learning_rate": 9.904842312250024e-05,
|
| 1073 |
+
"loss": 0.01,
|
| 1074 |
+
"step": 1520
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.257189811010682,
|
| 1078 |
+
"grad_norm": 0.18665545298861494,
|
| 1079 |
+
"learning_rate": 9.901329885918898e-05,
|
| 1080 |
+
"loss": 0.0102,
|
| 1081 |
+
"step": 1530
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.2654067378800329,
|
| 1085 |
+
"grad_norm": 0.20441793337405792,
|
| 1086 |
+
"learning_rate": 9.897754449988687e-05,
|
| 1087 |
+
"loss": 0.0099,
|
| 1088 |
+
"step": 1540
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.2736236647493837,
|
| 1092 |
+
"grad_norm": 0.17232110916564725,
|
| 1093 |
+
"learning_rate": 9.894116050423812e-05,
|
| 1094 |
+
"loss": 0.01,
|
| 1095 |
+
"step": 1550
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.2818405916187345,
|
| 1099 |
+
"grad_norm": 0.17117350565381512,
|
| 1100 |
+
"learning_rate": 9.890414733998131e-05,
|
| 1101 |
+
"loss": 0.0097,
|
| 1102 |
+
"step": 1560
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.2900575184880854,
|
| 1106 |
+
"grad_norm": 0.16106675543143417,
|
| 1107 |
+
"learning_rate": 9.886650548294333e-05,
|
| 1108 |
+
"loss": 0.0098,
|
| 1109 |
+
"step": 1570
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.2982744453574364,
|
| 1113 |
+
"grad_norm": 0.15673244249109636,
|
| 1114 |
+
"learning_rate": 9.882823541703332e-05,
|
| 1115 |
+
"loss": 0.0096,
|
| 1116 |
+
"step": 1580
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.3064913722267872,
|
| 1120 |
+
"grad_norm": 0.18359994942806246,
|
| 1121 |
+
"learning_rate": 9.878933763423644e-05,
|
| 1122 |
+
"loss": 0.0095,
|
| 1123 |
+
"step": 1590
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.314708299096138,
|
| 1127 |
+
"grad_norm": 0.16928325787860654,
|
| 1128 |
+
"learning_rate": 9.874981263460749e-05,
|
| 1129 |
+
"loss": 0.0092,
|
| 1130 |
+
"step": 1600
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.322925225965489,
|
| 1134 |
+
"grad_norm": 0.17321373689712377,
|
| 1135 |
+
"learning_rate": 9.87096609262646e-05,
|
| 1136 |
+
"loss": 0.0091,
|
| 1137 |
+
"step": 1610
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.3311421528348397,
|
| 1141 |
+
"grad_norm": 0.18291522891441084,
|
| 1142 |
+
"learning_rate": 9.866888302538255e-05,
|
| 1143 |
+
"loss": 0.009,
|
| 1144 |
+
"step": 1620
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.3393590797041908,
|
| 1148 |
+
"grad_norm": 0.19719101776097683,
|
| 1149 |
+
"learning_rate": 9.862747945618629e-05,
|
| 1150 |
+
"loss": 0.0091,
|
| 1151 |
+
"step": 1630
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.3475760065735414,
|
| 1155 |
+
"grad_norm": 0.1507248639899357,
|
| 1156 |
+
"learning_rate": 9.858545075094402e-05,
|
| 1157 |
+
"loss": 0.0089,
|
| 1158 |
+
"step": 1640
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.3557929334428924,
|
| 1162 |
+
"grad_norm": 0.15900627243941012,
|
| 1163 |
+
"learning_rate": 9.854279744996054e-05,
|
| 1164 |
+
"loss": 0.0085,
|
| 1165 |
+
"step": 1650
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.3640098603122432,
|
| 1169 |
+
"grad_norm": 0.1614739960911069,
|
| 1170 |
+
"learning_rate": 9.849952010157014e-05,
|
| 1171 |
+
"loss": 0.0084,
|
| 1172 |
+
"step": 1660
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.372226787181594,
|
| 1176 |
+
"grad_norm": 0.18518862249859025,
|
| 1177 |
+
"learning_rate": 9.845561926212967e-05,
|
| 1178 |
+
"loss": 0.0085,
|
| 1179 |
+
"step": 1670
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.380443714050945,
|
| 1183 |
+
"grad_norm": 0.16465865540140373,
|
| 1184 |
+
"learning_rate": 9.841109549601133e-05,
|
| 1185 |
+
"loss": 0.0087,
|
| 1186 |
+
"step": 1680
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.3886606409202957,
|
| 1190 |
+
"grad_norm": 0.1778246547911174,
|
| 1191 |
+
"learning_rate": 9.83659493755954e-05,
|
| 1192 |
+
"loss": 0.0085,
|
| 1193 |
+
"step": 1690
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.3968775677896468,
|
| 1197 |
+
"grad_norm": 0.17535152748688587,
|
| 1198 |
+
"learning_rate": 9.832018148126291e-05,
|
| 1199 |
+
"loss": 0.0082,
|
| 1200 |
+
"step": 1700
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.4050944946589976,
|
| 1204 |
+
"grad_norm": 0.1616895940395759,
|
| 1205 |
+
"learning_rate": 9.827379240138824e-05,
|
| 1206 |
+
"loss": 0.008,
|
| 1207 |
+
"step": 1710
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.4133114215283484,
|
| 1211 |
+
"grad_norm": 0.17228485317380646,
|
| 1212 |
+
"learning_rate": 9.822678273233139e-05,
|
| 1213 |
+
"loss": 0.0081,
|
| 1214 |
+
"step": 1720
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.4215283483976993,
|
| 1218 |
+
"grad_norm": 0.14635147682270658,
|
| 1219 |
+
"learning_rate": 9.817915307843046e-05,
|
| 1220 |
+
"loss": 0.008,
|
| 1221 |
+
"step": 1730
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.42974527526705,
|
| 1225 |
+
"grad_norm": 0.1577912852087264,
|
| 1226 |
+
"learning_rate": 9.813090405199388e-05,
|
| 1227 |
+
"loss": 0.0078,
|
| 1228 |
+
"step": 1740
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.437962202136401,
|
| 1232 |
+
"grad_norm": 0.13215393793980348,
|
| 1233 |
+
"learning_rate": 9.80820362732924e-05,
|
| 1234 |
+
"loss": 0.0079,
|
| 1235 |
+
"step": 1750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.4461791290057517,
|
| 1239 |
+
"grad_norm": 0.14886527670507846,
|
| 1240 |
+
"learning_rate": 9.803255037055128e-05,
|
| 1241 |
+
"loss": 0.0077,
|
| 1242 |
+
"step": 1760
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.4543960558751028,
|
| 1246 |
+
"grad_norm": 0.1663787864042773,
|
| 1247 |
+
"learning_rate": 9.798244697994214e-05,
|
| 1248 |
+
"loss": 0.0074,
|
| 1249 |
+
"step": 1770
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.4626129827444536,
|
| 1253 |
+
"grad_norm": 0.14313267064422866,
|
| 1254 |
+
"learning_rate": 9.793172674557475e-05,
|
| 1255 |
+
"loss": 0.0072,
|
| 1256 |
+
"step": 1780
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.4708299096138044,
|
| 1260 |
+
"grad_norm": 0.14881896112838372,
|
| 1261 |
+
"learning_rate": 9.78803903194888e-05,
|
| 1262 |
+
"loss": 0.0075,
|
| 1263 |
+
"step": 1790
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.4790468364831553,
|
| 1267 |
+
"grad_norm": 0.1378603541071152,
|
| 1268 |
+
"learning_rate": 9.782843836164553e-05,
|
| 1269 |
+
"loss": 0.0071,
|
| 1270 |
+
"step": 1800
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.487263763352506,
|
| 1274 |
+
"grad_norm": 0.13488096210212905,
|
| 1275 |
+
"learning_rate": 9.777587153991916e-05,
|
| 1276 |
+
"loss": 0.0069,
|
| 1277 |
+
"step": 1810
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.4954806902218571,
|
| 1281 |
+
"grad_norm": 0.149195935264567,
|
| 1282 |
+
"learning_rate": 9.77226905300884e-05,
|
| 1283 |
+
"loss": 0.007,
|
| 1284 |
+
"step": 1820
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.5036976170912077,
|
| 1288 |
+
"grad_norm": 0.14389230928854732,
|
| 1289 |
+
"learning_rate": 9.766889601582774e-05,
|
| 1290 |
+
"loss": 0.0071,
|
| 1291 |
+
"step": 1830
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.5119145439605588,
|
| 1295 |
+
"grad_norm": 0.14038858287355682,
|
| 1296 |
+
"learning_rate": 9.761448868869863e-05,
|
| 1297 |
+
"loss": 0.0068,
|
| 1298 |
+
"step": 1840
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.5201314708299096,
|
| 1302 |
+
"grad_norm": 0.14432103109386488,
|
| 1303 |
+
"learning_rate": 9.755946924814055e-05,
|
| 1304 |
+
"loss": 0.0069,
|
| 1305 |
+
"step": 1850
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.5283483976992605,
|
| 1309 |
+
"grad_norm": 0.14793500060029685,
|
| 1310 |
+
"learning_rate": 9.750383840146218e-05,
|
| 1311 |
+
"loss": 0.0069,
|
| 1312 |
+
"step": 1860
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.5365653245686115,
|
| 1316 |
+
"grad_norm": 0.1347785521160864,
|
| 1317 |
+
"learning_rate": 9.744759686383211e-05,
|
| 1318 |
+
"loss": 0.0069,
|
| 1319 |
+
"step": 1870
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.544782251437962,
|
| 1323 |
+
"grad_norm": 0.13316136302347778,
|
| 1324 |
+
"learning_rate": 9.739074535826978e-05,
|
| 1325 |
+
"loss": 0.0065,
|
| 1326 |
+
"step": 1880
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.5529991783073132,
|
| 1330 |
+
"grad_norm": 0.14574183234800075,
|
| 1331 |
+
"learning_rate": 9.733328461563612e-05,
|
| 1332 |
+
"loss": 0.0065,
|
| 1333 |
+
"step": 1890
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.561216105176664,
|
| 1337 |
+
"grad_norm": 0.13326787078082328,
|
| 1338 |
+
"learning_rate": 9.72752153746242e-05,
|
| 1339 |
+
"loss": 0.0066,
|
| 1340 |
+
"step": 1900
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.5694330320460148,
|
| 1344 |
+
"grad_norm": 0.14016833932250203,
|
| 1345 |
+
"learning_rate": 9.721653838174972e-05,
|
| 1346 |
+
"loss": 0.0064,
|
| 1347 |
+
"step": 1910
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.5776499589153656,
|
| 1351 |
+
"grad_norm": 0.1220618263565421,
|
| 1352 |
+
"learning_rate": 9.715725439134132e-05,
|
| 1353 |
+
"loss": 0.0063,
|
| 1354 |
+
"step": 1920
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.5858668857847165,
|
| 1358 |
+
"grad_norm": 0.14734607229286312,
|
| 1359 |
+
"learning_rate": 9.709736416553106e-05,
|
| 1360 |
+
"loss": 0.0064,
|
| 1361 |
+
"step": 1930
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.5940838126540675,
|
| 1365 |
+
"grad_norm": 0.1513977921448163,
|
| 1366 |
+
"learning_rate": 9.703686847424449e-05,
|
| 1367 |
+
"loss": 0.0063,
|
| 1368 |
+
"step": 1940
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.6023007395234181,
|
| 1372 |
+
"grad_norm": 0.15247244786036027,
|
| 1373 |
+
"learning_rate": 9.697576809519079e-05,
|
| 1374 |
+
"loss": 0.0063,
|
| 1375 |
+
"step": 1950
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.6105176663927692,
|
| 1379 |
+
"grad_norm": 0.12449770873160382,
|
| 1380 |
+
"learning_rate": 9.691406381385278e-05,
|
| 1381 |
+
"loss": 0.0061,
|
| 1382 |
+
"step": 1960
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.61873459326212,
|
| 1386 |
+
"grad_norm": 0.14780820552617843,
|
| 1387 |
+
"learning_rate": 9.685175642347679e-05,
|
| 1388 |
+
"loss": 0.0062,
|
| 1389 |
+
"step": 1970
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.6269515201314708,
|
| 1393 |
+
"grad_norm": 0.12285378569676762,
|
| 1394 |
+
"learning_rate": 9.678884672506252e-05,
|
| 1395 |
+
"loss": 0.006,
|
| 1396 |
+
"step": 1980
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.6351684470008216,
|
| 1400 |
+
"grad_norm": 0.1374507844474314,
|
| 1401 |
+
"learning_rate": 9.672533552735268e-05,
|
| 1402 |
+
"loss": 0.0059,
|
| 1403 |
+
"step": 1990
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.6433853738701725,
|
| 1407 |
+
"grad_norm": 0.1340445045771385,
|
| 1408 |
+
"learning_rate": 9.666122364682263e-05,
|
| 1409 |
+
"loss": 0.0058,
|
| 1410 |
+
"step": 2000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.6516023007395235,
|
| 1414 |
+
"grad_norm": 1.5494672490847667,
|
| 1415 |
+
"learning_rate": 9.659651190766992e-05,
|
| 1416 |
+
"loss": 0.0473,
|
| 1417 |
+
"step": 2010
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.6598192276088741,
|
| 1421 |
+
"grad_norm": 1.055588957101282,
|
| 1422 |
+
"learning_rate": 9.653120114180356e-05,
|
| 1423 |
+
"loss": 0.0213,
|
| 1424 |
+
"step": 2020
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.6680361544782252,
|
| 1428 |
+
"grad_norm": 2.158511305689324,
|
| 1429 |
+
"learning_rate": 9.646529218883349e-05,
|
| 1430 |
+
"loss": 0.068,
|
| 1431 |
+
"step": 2030
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.676253081347576,
|
| 1435 |
+
"grad_norm": 0.34278686660556884,
|
| 1436 |
+
"learning_rate": 9.639878589605969e-05,
|
| 1437 |
+
"loss": 0.025,
|
| 1438 |
+
"step": 2040
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.6844700082169268,
|
| 1442 |
+
"grad_norm": 0.18106436706773044,
|
| 1443 |
+
"learning_rate": 9.63316831184613e-05,
|
| 1444 |
+
"loss": 0.0097,
|
| 1445 |
+
"step": 2050
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.6926869350862779,
|
| 1449 |
+
"grad_norm": 0.14187259081013245,
|
| 1450 |
+
"learning_rate": 9.626398471868564e-05,
|
| 1451 |
+
"loss": 0.0079,
|
| 1452 |
+
"step": 2060
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.7009038619556285,
|
| 1456 |
+
"grad_norm": 0.13032445672489365,
|
| 1457 |
+
"learning_rate": 9.619569156703714e-05,
|
| 1458 |
+
"loss": 0.007,
|
| 1459 |
+
"step": 2070
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.7091207888249795,
|
| 1463 |
+
"grad_norm": 0.13900486225724332,
|
| 1464 |
+
"learning_rate": 9.612680454146609e-05,
|
| 1465 |
+
"loss": 0.0068,
|
| 1466 |
+
"step": 2080
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.7173377156943304,
|
| 1470 |
+
"grad_norm": 0.1371829976046093,
|
| 1471 |
+
"learning_rate": 9.605732452755742e-05,
|
| 1472 |
+
"loss": 0.0062,
|
| 1473 |
+
"step": 2090
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.7255546425636812,
|
| 1477 |
+
"grad_norm": 0.11435904716007088,
|
| 1478 |
+
"learning_rate": 9.598725241851932e-05,
|
| 1479 |
+
"loss": 0.0062,
|
| 1480 |
+
"step": 2100
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.733771569433032,
|
| 1484 |
+
"grad_norm": 0.12003022629994674,
|
| 1485 |
+
"learning_rate": 9.591658911517162e-05,
|
| 1486 |
+
"loss": 0.006,
|
| 1487 |
+
"step": 2110
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 1.7419884963023828,
|
| 1491 |
+
"grad_norm": 0.12472657638284222,
|
| 1492 |
+
"learning_rate": 9.584533552593447e-05,
|
| 1493 |
+
"loss": 0.0056,
|
| 1494 |
+
"step": 2120
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.7502054231717339,
|
| 1498 |
+
"grad_norm": 0.10956857048247456,
|
| 1499 |
+
"learning_rate": 9.577349256681636e-05,
|
| 1500 |
+
"loss": 0.0057,
|
| 1501 |
+
"step": 2130
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.7584223500410845,
|
| 1505 |
+
"grad_norm": 0.10206011618948153,
|
| 1506 |
+
"learning_rate": 9.570106116140256e-05,
|
| 1507 |
+
"loss": 0.0054,
|
| 1508 |
+
"step": 2140
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.7666392769104355,
|
| 1512 |
+
"grad_norm": 0.12595281361553662,
|
| 1513 |
+
"learning_rate": 9.56280422408432e-05,
|
| 1514 |
+
"loss": 0.0054,
|
| 1515 |
+
"step": 2150
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.7748562037797864,
|
| 1519 |
+
"grad_norm": 0.11843203344142937,
|
| 1520 |
+
"learning_rate": 9.555443674384122e-05,
|
| 1521 |
+
"loss": 0.0052,
|
| 1522 |
+
"step": 2160
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.7830731306491372,
|
| 1526 |
+
"grad_norm": 0.10294032952850225,
|
| 1527 |
+
"learning_rate": 9.548024561664039e-05,
|
| 1528 |
+
"loss": 0.0051,
|
| 1529 |
+
"step": 2170
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.7912900575184882,
|
| 1533 |
+
"grad_norm": 0.11977450499323275,
|
| 1534 |
+
"learning_rate": 9.540546981301313e-05,
|
| 1535 |
+
"loss": 0.0052,
|
| 1536 |
+
"step": 2180
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.7995069843878388,
|
| 1540 |
+
"grad_norm": 0.10306048483524843,
|
| 1541 |
+
"learning_rate": 9.53301102942482e-05,
|
| 1542 |
+
"loss": 0.0051,
|
| 1543 |
+
"step": 2190
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.80772391125719,
|
| 1547 |
+
"grad_norm": 0.11032309861470083,
|
| 1548 |
+
"learning_rate": 9.525416802913841e-05,
|
| 1549 |
+
"loss": 0.005,
|
| 1550 |
+
"step": 2200
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.8159408381265407,
|
| 1554 |
+
"grad_norm": 0.12349361681027836,
|
| 1555 |
+
"learning_rate": 9.51776439939681e-05,
|
| 1556 |
+
"loss": 0.0051,
|
| 1557 |
+
"step": 2210
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.8241577649958916,
|
| 1561 |
+
"grad_norm": 0.11522340458787386,
|
| 1562 |
+
"learning_rate": 9.510053917250061e-05,
|
| 1563 |
+
"loss": 0.0048,
|
| 1564 |
+
"step": 2220
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.8323746918652424,
|
| 1568 |
+
"grad_norm": 0.10888918083024797,
|
| 1569 |
+
"learning_rate": 9.502285455596572e-05,
|
| 1570 |
+
"loss": 0.0048,
|
| 1571 |
+
"step": 2230
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.8405916187345932,
|
| 1575 |
+
"grad_norm": 0.10904702080255187,
|
| 1576 |
+
"learning_rate": 9.494459114304672e-05,
|
| 1577 |
+
"loss": 0.0047,
|
| 1578 |
+
"step": 2240
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.8488085456039443,
|
| 1582 |
+
"grad_norm": 0.39937298924235193,
|
| 1583 |
+
"learning_rate": 9.48657499398678e-05,
|
| 1584 |
+
"loss": 0.0059,
|
| 1585 |
+
"step": 2250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.8570254724732949,
|
| 1589 |
+
"grad_norm": 0.24868775681299268,
|
| 1590 |
+
"learning_rate": 9.47863319599809e-05,
|
| 1591 |
+
"loss": 0.0075,
|
| 1592 |
+
"step": 2260
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.865242399342646,
|
| 1596 |
+
"grad_norm": 0.1633251832735737,
|
| 1597 |
+
"learning_rate": 9.470633822435282e-05,
|
| 1598 |
+
"loss": 0.0068,
|
| 1599 |
+
"step": 2270
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.8734593262119967,
|
| 1603 |
+
"grad_norm": 0.10576097653379558,
|
| 1604 |
+
"learning_rate": 9.462576976135207e-05,
|
| 1605 |
+
"loss": 0.0054,
|
| 1606 |
+
"step": 2280
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.8816762530813476,
|
| 1610 |
+
"grad_norm": 0.12656212811763418,
|
| 1611 |
+
"learning_rate": 9.454462760673559e-05,
|
| 1612 |
+
"loss": 0.005,
|
| 1613 |
+
"step": 2290
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.8898931799506984,
|
| 1617 |
+
"grad_norm": 0.1000511957852473,
|
| 1618 |
+
"learning_rate": 9.446291280363551e-05,
|
| 1619 |
+
"loss": 0.0048,
|
| 1620 |
+
"step": 2300
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.8981101068200492,
|
| 1624 |
+
"grad_norm": 0.1001106728471161,
|
| 1625 |
+
"learning_rate": 9.43806264025457e-05,
|
| 1626 |
+
"loss": 0.0046,
|
| 1627 |
+
"step": 2310
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.9063270336894003,
|
| 1631 |
+
"grad_norm": 0.10653485492005862,
|
| 1632 |
+
"learning_rate": 9.429776946130826e-05,
|
| 1633 |
+
"loss": 0.0046,
|
| 1634 |
+
"step": 2320
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.9145439605587509,
|
| 1638 |
+
"grad_norm": 0.09554139188139198,
|
| 1639 |
+
"learning_rate": 9.421434304509992e-05,
|
| 1640 |
+
"loss": 0.0044,
|
| 1641 |
+
"step": 2330
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.922760887428102,
|
| 1645 |
+
"grad_norm": 0.11235081047459526,
|
| 1646 |
+
"learning_rate": 9.413034822641844e-05,
|
| 1647 |
+
"loss": 0.0043,
|
| 1648 |
+
"step": 2340
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.9309778142974527,
|
| 1652 |
+
"grad_norm": 0.09925731232811477,
|
| 1653 |
+
"learning_rate": 9.404578608506865e-05,
|
| 1654 |
+
"loss": 0.0043,
|
| 1655 |
+
"step": 2350
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.9391947411668036,
|
| 1659 |
+
"grad_norm": 0.10170180896026369,
|
| 1660 |
+
"learning_rate": 9.396065770814867e-05,
|
| 1661 |
+
"loss": 0.0044,
|
| 1662 |
+
"step": 2360
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.9474116680361546,
|
| 1666 |
+
"grad_norm": 0.1094941960116376,
|
| 1667 |
+
"learning_rate": 9.387496419003599e-05,
|
| 1668 |
+
"loss": 0.0042,
|
| 1669 |
+
"step": 2370
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.9556285949055052,
|
| 1673 |
+
"grad_norm": 0.09325155091207862,
|
| 1674 |
+
"learning_rate": 9.378870663237328e-05,
|
| 1675 |
+
"loss": 0.004,
|
| 1676 |
+
"step": 2380
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.9638455217748563,
|
| 1680 |
+
"grad_norm": 0.09015117689020187,
|
| 1681 |
+
"learning_rate": 9.37018861440543e-05,
|
| 1682 |
+
"loss": 0.004,
|
| 1683 |
+
"step": 2390
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.972062448644207,
|
| 1687 |
+
"grad_norm": 0.10243104002211424,
|
| 1688 |
+
"learning_rate": 9.361450384120963e-05,
|
| 1689 |
+
"loss": 0.0041,
|
| 1690 |
+
"step": 2400
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.980279375513558,
|
| 1694 |
+
"grad_norm": 0.11356430772327424,
|
| 1695 |
+
"learning_rate": 9.352656084719231e-05,
|
| 1696 |
+
"loss": 0.004,
|
| 1697 |
+
"step": 2410
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.9884963023829088,
|
| 1701 |
+
"grad_norm": 0.1133566447197813,
|
| 1702 |
+
"learning_rate": 9.343805829256343e-05,
|
| 1703 |
+
"loss": 0.0042,
|
| 1704 |
+
"step": 2420
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.9967132292522596,
|
| 1708 |
+
"grad_norm": 0.09025115427016078,
|
| 1709 |
+
"learning_rate": 9.334899731507756e-05,
|
| 1710 |
+
"loss": 0.0039,
|
| 1711 |
+
"step": 2430
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 2.0049301561216106,
|
| 1715 |
+
"grad_norm": 0.10161038188503023,
|
| 1716 |
+
"learning_rate": 9.325937905966813e-05,
|
| 1717 |
+
"loss": 0.0035,
|
| 1718 |
+
"step": 2440
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 2.0131470829909612,
|
| 1722 |
+
"grad_norm": 0.08227738890594856,
|
| 1723 |
+
"learning_rate": 9.316920467843273e-05,
|
| 1724 |
+
"loss": 0.003,
|
| 1725 |
+
"step": 2450
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 2.0213640098603123,
|
| 1729 |
+
"grad_norm": 0.08570367324292932,
|
| 1730 |
+
"learning_rate": 9.307847533061831e-05,
|
| 1731 |
+
"loss": 0.003,
|
| 1732 |
+
"step": 2460
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 2.029580936729663,
|
| 1736 |
+
"grad_norm": 0.07659113085608305,
|
| 1737 |
+
"learning_rate": 9.29871921826062e-05,
|
| 1738 |
+
"loss": 0.0029,
|
| 1739 |
+
"step": 2470
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 2.037797863599014,
|
| 1743 |
+
"grad_norm": 0.08735185109119072,
|
| 1744 |
+
"learning_rate": 9.289535640789721e-05,
|
| 1745 |
+
"loss": 0.0028,
|
| 1746 |
+
"step": 2480
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 2.046014790468365,
|
| 1750 |
+
"grad_norm": 0.08112181526353192,
|
| 1751 |
+
"learning_rate": 9.280296918709653e-05,
|
| 1752 |
+
"loss": 0.0028,
|
| 1753 |
+
"step": 2490
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 2.0542317173377156,
|
| 1757 |
+
"grad_norm": 0.09071733200524149,
|
| 1758 |
+
"learning_rate": 9.271003170789846e-05,
|
| 1759 |
+
"loss": 0.0028,
|
| 1760 |
+
"step": 2500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 2.0624486442070666,
|
| 1764 |
+
"grad_norm": 0.07888705433105933,
|
| 1765 |
+
"learning_rate": 9.261654516507126e-05,
|
| 1766 |
+
"loss": 0.0028,
|
| 1767 |
+
"step": 2510
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 2.0706655710764172,
|
| 1771 |
+
"grad_norm": 0.08256418767761199,
|
| 1772 |
+
"learning_rate": 9.252251076044172e-05,
|
| 1773 |
+
"loss": 0.0027,
|
| 1774 |
+
"step": 2520
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 2.0788824979457683,
|
| 1778 |
+
"grad_norm": 0.08816475273210793,
|
| 1779 |
+
"learning_rate": 9.242792970287972e-05,
|
| 1780 |
+
"loss": 0.0028,
|
| 1781 |
+
"step": 2530
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 2.0870994248151193,
|
| 1785 |
+
"grad_norm": 0.11440139774626451,
|
| 1786 |
+
"learning_rate": 9.233280320828274e-05,
|
| 1787 |
+
"loss": 0.0028,
|
| 1788 |
+
"step": 2540
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 2.09531635168447,
|
| 1792 |
+
"grad_norm": 0.0812607297477746,
|
| 1793 |
+
"learning_rate": 9.223713249956016e-05,
|
| 1794 |
+
"loss": 0.0028,
|
| 1795 |
+
"step": 2550
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 2.103533278553821,
|
| 1799 |
+
"grad_norm": 0.09406446044734776,
|
| 1800 |
+
"learning_rate": 9.214091880661752e-05,
|
| 1801 |
+
"loss": 0.0028,
|
| 1802 |
+
"step": 2560
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 2.1117502054231716,
|
| 1806 |
+
"grad_norm": 0.08675348503864556,
|
| 1807 |
+
"learning_rate": 9.204416336634084e-05,
|
| 1808 |
+
"loss": 0.0029,
|
| 1809 |
+
"step": 2570
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 2.1199671322925226,
|
| 1813 |
+
"grad_norm": 0.08798127219029518,
|
| 1814 |
+
"learning_rate": 9.194686742258059e-05,
|
| 1815 |
+
"loss": 0.0029,
|
| 1816 |
+
"step": 2580
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 2.1281840591618733,
|
| 1820 |
+
"grad_norm": 0.1042447423026133,
|
| 1821 |
+
"learning_rate": 9.184903222613574e-05,
|
| 1822 |
+
"loss": 0.0028,
|
| 1823 |
+
"step": 2590
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 2.1364009860312243,
|
| 1827 |
+
"grad_norm": 0.07226523480692097,
|
| 1828 |
+
"learning_rate": 9.175065903473768e-05,
|
| 1829 |
+
"loss": 0.0028,
|
| 1830 |
+
"step": 2600
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 2.1446179129005754,
|
| 1834 |
+
"grad_norm": 0.09060351493323598,
|
| 1835 |
+
"learning_rate": 9.165174911303409e-05,
|
| 1836 |
+
"loss": 0.0029,
|
| 1837 |
+
"step": 2610
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 2.152834839769926,
|
| 1841 |
+
"grad_norm": 0.09147549430600695,
|
| 1842 |
+
"learning_rate": 9.15523037325726e-05,
|
| 1843 |
+
"loss": 0.0028,
|
| 1844 |
+
"step": 2620
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 2.161051766639277,
|
| 1848 |
+
"grad_norm": 0.08581674950289743,
|
| 1849 |
+
"learning_rate": 9.145232417178453e-05,
|
| 1850 |
+
"loss": 0.0027,
|
| 1851 |
+
"step": 2630
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 2.1692686935086276,
|
| 1855 |
+
"grad_norm": 0.0806685544659502,
|
| 1856 |
+
"learning_rate": 9.135181171596838e-05,
|
| 1857 |
+
"loss": 0.0027,
|
| 1858 |
+
"step": 2640
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 2.1774856203779787,
|
| 1862 |
+
"grad_norm": 0.0867069998728415,
|
| 1863 |
+
"learning_rate": 9.125076765727338e-05,
|
| 1864 |
+
"loss": 0.0027,
|
| 1865 |
+
"step": 2650
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 2.1857025472473293,
|
| 1869 |
+
"grad_norm": 0.09650448122353704,
|
| 1870 |
+
"learning_rate": 9.114919329468282e-05,
|
| 1871 |
+
"loss": 0.0026,
|
| 1872 |
+
"step": 2660
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 2.1939194741166803,
|
| 1876 |
+
"grad_norm": 0.08346292571616859,
|
| 1877 |
+
"learning_rate": 9.104708993399738e-05,
|
| 1878 |
+
"loss": 0.0025,
|
| 1879 |
+
"step": 2670
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 2.2021364009860314,
|
| 1883 |
+
"grad_norm": 0.0922034193015322,
|
| 1884 |
+
"learning_rate": 9.094445888781829e-05,
|
| 1885 |
+
"loss": 0.0027,
|
| 1886 |
+
"step": 2680
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 2.210353327855382,
|
| 1890 |
+
"grad_norm": 0.08045238236970594,
|
| 1891 |
+
"learning_rate": 9.084130147553057e-05,
|
| 1892 |
+
"loss": 0.0026,
|
| 1893 |
+
"step": 2690
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 2.218570254724733,
|
| 1897 |
+
"grad_norm": 0.0823407241176973,
|
| 1898 |
+
"learning_rate": 9.073761902328597e-05,
|
| 1899 |
+
"loss": 0.0026,
|
| 1900 |
+
"step": 2700
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 2.2267871815940836,
|
| 1904 |
+
"grad_norm": 0.0918388311221415,
|
| 1905 |
+
"learning_rate": 9.063341286398592e-05,
|
| 1906 |
+
"loss": 0.0026,
|
| 1907 |
+
"step": 2710
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 2.2350041084634347,
|
| 1911 |
+
"grad_norm": 0.08811224741688072,
|
| 1912 |
+
"learning_rate": 9.052868433726448e-05,
|
| 1913 |
+
"loss": 0.0027,
|
| 1914 |
+
"step": 2720
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 2.2432210353327857,
|
| 1918 |
+
"grad_norm": 0.08713466229884091,
|
| 1919 |
+
"learning_rate": 9.042343478947103e-05,
|
| 1920 |
+
"loss": 0.0026,
|
| 1921 |
+
"step": 2730
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 2.2514379622021363,
|
| 1925 |
+
"grad_norm": 0.08164204161119902,
|
| 1926 |
+
"learning_rate": 9.0317665573653e-05,
|
| 1927 |
+
"loss": 0.0027,
|
| 1928 |
+
"step": 2740
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 2.2596548890714874,
|
| 1932 |
+
"grad_norm": 0.08621684454149689,
|
| 1933 |
+
"learning_rate": 9.021137804953846e-05,
|
| 1934 |
+
"loss": 0.0026,
|
| 1935 |
+
"step": 2750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 2.267871815940838,
|
| 1939 |
+
"grad_norm": 0.08581890628695187,
|
| 1940 |
+
"learning_rate": 9.010457358351867e-05,
|
| 1941 |
+
"loss": 0.0025,
|
| 1942 |
+
"step": 2760
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 2.276088742810189,
|
| 1946 |
+
"grad_norm": 0.0880327255343587,
|
| 1947 |
+
"learning_rate": 8.99972535486305e-05,
|
| 1948 |
+
"loss": 0.0025,
|
| 1949 |
+
"step": 2770
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 2.28430566967954,
|
| 1953 |
+
"grad_norm": 0.10198724504699996,
|
| 1954 |
+
"learning_rate": 8.988941932453872e-05,
|
| 1955 |
+
"loss": 0.0025,
|
| 1956 |
+
"step": 2780
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 2.2925225965488907,
|
| 1960 |
+
"grad_norm": 0.08363470598937481,
|
| 1961 |
+
"learning_rate": 8.978107229751842e-05,
|
| 1962 |
+
"loss": 0.0026,
|
| 1963 |
+
"step": 2790
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 2.3007395234182417,
|
| 1967 |
+
"grad_norm": 0.08701906271338153,
|
| 1968 |
+
"learning_rate": 8.967221386043699e-05,
|
| 1969 |
+
"loss": 0.0026,
|
| 1970 |
+
"step": 2800
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 2.3089564502875923,
|
| 1974 |
+
"grad_norm": 0.08120255658359293,
|
| 1975 |
+
"learning_rate": 8.956284541273632e-05,
|
| 1976 |
+
"loss": 0.0026,
|
| 1977 |
+
"step": 2810
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 2.3171733771569434,
|
| 1981 |
+
"grad_norm": 0.09844930162834767,
|
| 1982 |
+
"learning_rate": 8.945296836041487e-05,
|
| 1983 |
+
"loss": 0.0026,
|
| 1984 |
+
"step": 2820
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 2.325390304026294,
|
| 1988 |
+
"grad_norm": 0.08546167195111991,
|
| 1989 |
+
"learning_rate": 8.934258411600945e-05,
|
| 1990 |
+
"loss": 0.0026,
|
| 1991 |
+
"step": 2830
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 2.333607230895645,
|
| 1995 |
+
"grad_norm": 0.07598724139236995,
|
| 1996 |
+
"learning_rate": 8.923169409857718e-05,
|
| 1997 |
+
"loss": 0.0026,
|
| 1998 |
+
"step": 2840
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 2.341824157764996,
|
| 2002 |
+
"grad_norm": 0.07855480397018198,
|
| 2003 |
+
"learning_rate": 8.912029973367717e-05,
|
| 2004 |
+
"loss": 0.0026,
|
| 2005 |
+
"step": 2850
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 2.3500410846343467,
|
| 2009 |
+
"grad_norm": 0.08252273680081287,
|
| 2010 |
+
"learning_rate": 8.900840245335225e-05,
|
| 2011 |
+
"loss": 0.0026,
|
| 2012 |
+
"step": 2860
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 2.3582580115036977,
|
| 2016 |
+
"grad_norm": 0.09490170806356914,
|
| 2017 |
+
"learning_rate": 8.889600369611051e-05,
|
| 2018 |
+
"loss": 0.0025,
|
| 2019 |
+
"step": 2870
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 2.3664749383730483,
|
| 2023 |
+
"grad_norm": 0.0873514246728157,
|
| 2024 |
+
"learning_rate": 8.878310490690686e-05,
|
| 2025 |
+
"loss": 0.0026,
|
| 2026 |
+
"step": 2880
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 2.3746918652423994,
|
| 2030 |
+
"grad_norm": 0.08433770615709385,
|
| 2031 |
+
"learning_rate": 8.866970753712442e-05,
|
| 2032 |
+
"loss": 0.0024,
|
| 2033 |
+
"step": 2890
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 2.38290879211175,
|
| 2037 |
+
"grad_norm": 0.08543362259129167,
|
| 2038 |
+
"learning_rate": 8.855581304455584e-05,
|
| 2039 |
+
"loss": 0.0025,
|
| 2040 |
+
"step": 2900
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 2.391125718981101,
|
| 2044 |
+
"grad_norm": 0.08675602957949538,
|
| 2045 |
+
"learning_rate": 8.844142289338466e-05,
|
| 2046 |
+
"loss": 0.0024,
|
| 2047 |
+
"step": 2910
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 2.399342645850452,
|
| 2051 |
+
"grad_norm": 0.08568773220601615,
|
| 2052 |
+
"learning_rate": 8.832653855416634e-05,
|
| 2053 |
+
"loss": 0.0027,
|
| 2054 |
+
"step": 2920
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 2.4075595727198027,
|
| 2058 |
+
"grad_norm": 0.07456118767226848,
|
| 2059 |
+
"learning_rate": 8.821116150380946e-05,
|
| 2060 |
+
"loss": 0.0025,
|
| 2061 |
+
"step": 2930
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 2.4157764995891537,
|
| 2065 |
+
"grad_norm": 0.0704198355651112,
|
| 2066 |
+
"learning_rate": 8.80952932255567e-05,
|
| 2067 |
+
"loss": 0.0023,
|
| 2068 |
+
"step": 2940
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 2.4239934264585044,
|
| 2072 |
+
"grad_norm": 0.08268043669392515,
|
| 2073 |
+
"learning_rate": 8.797893520896579e-05,
|
| 2074 |
+
"loss": 0.0023,
|
| 2075 |
+
"step": 2950
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 2.4322103533278554,
|
| 2079 |
+
"grad_norm": 0.08227547263082345,
|
| 2080 |
+
"learning_rate": 8.786208894989032e-05,
|
| 2081 |
+
"loss": 0.0023,
|
| 2082 |
+
"step": 2960
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 2.4404272801972064,
|
| 2086 |
+
"grad_norm": 0.08698253645225705,
|
| 2087 |
+
"learning_rate": 8.774475595046055e-05,
|
| 2088 |
+
"loss": 0.0025,
|
| 2089 |
+
"step": 2970
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 2.448644207066557,
|
| 2093 |
+
"grad_norm": 0.08849442489572161,
|
| 2094 |
+
"learning_rate": 8.762693771906409e-05,
|
| 2095 |
+
"loss": 0.0023,
|
| 2096 |
+
"step": 2980
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.456861133935908,
|
| 2100 |
+
"grad_norm": 0.07467727254896249,
|
| 2101 |
+
"learning_rate": 8.750863577032652e-05,
|
| 2102 |
+
"loss": 0.0023,
|
| 2103 |
+
"step": 2990
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 2.4650780608052587,
|
| 2107 |
+
"grad_norm": 0.09244113926463907,
|
| 2108 |
+
"learning_rate": 8.738985162509186e-05,
|
| 2109 |
+
"loss": 0.0024,
|
| 2110 |
+
"step": 3000
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 2.4732949876746098,
|
| 2114 |
+
"grad_norm": 0.08833101573406452,
|
| 2115 |
+
"learning_rate": 8.727058681040308e-05,
|
| 2116 |
+
"loss": 0.0024,
|
| 2117 |
+
"step": 3010
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 2.481511914543961,
|
| 2121 |
+
"grad_norm": 0.07226683243114444,
|
| 2122 |
+
"learning_rate": 8.715084285948248e-05,
|
| 2123 |
+
"loss": 0.0024,
|
| 2124 |
+
"step": 3020
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 2.4897288414133114,
|
| 2128 |
+
"grad_norm": 0.07268521459200765,
|
| 2129 |
+
"learning_rate": 8.703062131171192e-05,
|
| 2130 |
+
"loss": 0.0024,
|
| 2131 |
+
"step": 3030
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 2.4979457682826625,
|
| 2135 |
+
"grad_norm": 0.09089659924385088,
|
| 2136 |
+
"learning_rate": 8.690992371261302e-05,
|
| 2137 |
+
"loss": 0.0025,
|
| 2138 |
+
"step": 3040
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 2.506162695152013,
|
| 2142 |
+
"grad_norm": 0.07562339403468425,
|
| 2143 |
+
"learning_rate": 8.678875161382745e-05,
|
| 2144 |
+
"loss": 0.0024,
|
| 2145 |
+
"step": 3050
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 2.514379622021364,
|
| 2149 |
+
"grad_norm": 0.08031934504490368,
|
| 2150 |
+
"learning_rate": 8.666710657309672e-05,
|
| 2151 |
+
"loss": 0.0023,
|
| 2152 |
+
"step": 3060
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 2.5225965488907147,
|
| 2156 |
+
"grad_norm": 0.08457583970716735,
|
| 2157 |
+
"learning_rate": 8.654499015424243e-05,
|
| 2158 |
+
"loss": 0.0023,
|
| 2159 |
+
"step": 3070
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 2.5308134757600658,
|
| 2163 |
+
"grad_norm": 0.0755537914601127,
|
| 2164 |
+
"learning_rate": 8.642240392714593e-05,
|
| 2165 |
+
"loss": 0.0024,
|
| 2166 |
+
"step": 3080
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 2.5390304026294164,
|
| 2170 |
+
"grad_norm": 0.07873822493348767,
|
| 2171 |
+
"learning_rate": 8.629934946772831e-05,
|
| 2172 |
+
"loss": 0.0022,
|
| 2173 |
+
"step": 3090
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 2.5472473294987674,
|
| 2177 |
+
"grad_norm": 0.08908105432532254,
|
| 2178 |
+
"learning_rate": 8.617582835793003e-05,
|
| 2179 |
+
"loss": 0.0022,
|
| 2180 |
+
"step": 3100
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 2.5554642563681185,
|
| 2184 |
+
"grad_norm": 0.06999317221832577,
|
| 2185 |
+
"learning_rate": 8.605184218569067e-05,
|
| 2186 |
+
"loss": 0.0024,
|
| 2187 |
+
"step": 3110
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 2.563681183237469,
|
| 2191 |
+
"grad_norm": 0.07464568302833668,
|
| 2192 |
+
"learning_rate": 8.592739254492844e-05,
|
| 2193 |
+
"loss": 0.0022,
|
| 2194 |
+
"step": 3120
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 2.57189811010682,
|
| 2198 |
+
"grad_norm": 0.09956594996203859,
|
| 2199 |
+
"learning_rate": 8.580248103551973e-05,
|
| 2200 |
+
"loss": 0.0022,
|
| 2201 |
+
"step": 3130
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 2.5801150369761707,
|
| 2205 |
+
"grad_norm": 0.09774738216358816,
|
| 2206 |
+
"learning_rate": 8.567710926327855e-05,
|
| 2207 |
+
"loss": 0.0024,
|
| 2208 |
+
"step": 3140
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 2.5883319638455218,
|
| 2212 |
+
"grad_norm": 0.07882192410268168,
|
| 2213 |
+
"learning_rate": 8.555127883993582e-05,
|
| 2214 |
+
"loss": 0.0023,
|
| 2215 |
+
"step": 3150
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 2.596548890714873,
|
| 2219 |
+
"grad_norm": 0.0762996815625555,
|
| 2220 |
+
"learning_rate": 8.542499138311877e-05,
|
| 2221 |
+
"loss": 0.0022,
|
| 2222 |
+
"step": 3160
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 2.6047658175842234,
|
| 2226 |
+
"grad_norm": 0.07967271368798,
|
| 2227 |
+
"learning_rate": 8.529824851633004e-05,
|
| 2228 |
+
"loss": 0.0023,
|
| 2229 |
+
"step": 3170
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 2.6129827444535745,
|
| 2233 |
+
"grad_norm": 0.09037714716671512,
|
| 2234 |
+
"learning_rate": 8.517105186892683e-05,
|
| 2235 |
+
"loss": 0.0021,
|
| 2236 |
+
"step": 3180
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 2.621199671322925,
|
| 2240 |
+
"grad_norm": 0.06631171216954122,
|
| 2241 |
+
"learning_rate": 8.50434030761e-05,
|
| 2242 |
+
"loss": 0.0021,
|
| 2243 |
+
"step": 3190
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 2.629416598192276,
|
| 2247 |
+
"grad_norm": 0.06516524243367872,
|
| 2248 |
+
"learning_rate": 8.4915303778853e-05,
|
| 2249 |
+
"loss": 0.002,
|
| 2250 |
+
"step": 3200
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 2.637633525061627,
|
| 2254 |
+
"grad_norm": 0.07554930250673897,
|
| 2255 |
+
"learning_rate": 8.478675562398077e-05,
|
| 2256 |
+
"loss": 0.0021,
|
| 2257 |
+
"step": 3210
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 2.645850451930978,
|
| 2261 |
+
"grad_norm": 0.08936807507539121,
|
| 2262 |
+
"learning_rate": 8.465776026404866e-05,
|
| 2263 |
+
"loss": 0.0021,
|
| 2264 |
+
"step": 3220
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 2.654067378800329,
|
| 2268 |
+
"grad_norm": 0.10445169947837134,
|
| 2269 |
+
"learning_rate": 8.452831935737102e-05,
|
| 2270 |
+
"loss": 0.0022,
|
| 2271 |
+
"step": 3230
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 2.6622843056696794,
|
| 2275 |
+
"grad_norm": 0.0747838347850065,
|
| 2276 |
+
"learning_rate": 8.439843456799005e-05,
|
| 2277 |
+
"loss": 0.0022,
|
| 2278 |
+
"step": 3240
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 2.6705012325390305,
|
| 2282 |
+
"grad_norm": 0.07552693679165537,
|
| 2283 |
+
"learning_rate": 8.426810756565429e-05,
|
| 2284 |
+
"loss": 0.0022,
|
| 2285 |
+
"step": 3250
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 2.6787181594083815,
|
| 2289 |
+
"grad_norm": 0.07514300993270172,
|
| 2290 |
+
"learning_rate": 8.413734002579723e-05,
|
| 2291 |
+
"loss": 0.0022,
|
| 2292 |
+
"step": 3260
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 2.686935086277732,
|
| 2296 |
+
"grad_norm": 0.07037597595165376,
|
| 2297 |
+
"learning_rate": 8.40061336295157e-05,
|
| 2298 |
+
"loss": 0.0021,
|
| 2299 |
+
"step": 3270
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 2.6951520131470827,
|
| 2303 |
+
"grad_norm": 0.07303202846814086,
|
| 2304 |
+
"learning_rate": 8.387449006354837e-05,
|
| 2305 |
+
"loss": 0.002,
|
| 2306 |
+
"step": 3280
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 2.703368940016434,
|
| 2310 |
+
"grad_norm": 0.08480661725235046,
|
| 2311 |
+
"learning_rate": 8.374241102025391e-05,
|
| 2312 |
+
"loss": 0.002,
|
| 2313 |
+
"step": 3290
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 2.711585866885785,
|
| 2317 |
+
"grad_norm": 0.07058395092941114,
|
| 2318 |
+
"learning_rate": 8.360989819758935e-05,
|
| 2319 |
+
"loss": 0.0021,
|
| 2320 |
+
"step": 3300
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 2.7198027937551355,
|
| 2324 |
+
"grad_norm": 0.07837331908662558,
|
| 2325 |
+
"learning_rate": 8.347695329908827e-05,
|
| 2326 |
+
"loss": 0.0021,
|
| 2327 |
+
"step": 3310
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 2.7280197206244865,
|
| 2331 |
+
"grad_norm": 0.07564552331177772,
|
| 2332 |
+
"learning_rate": 8.334357803383875e-05,
|
| 2333 |
+
"loss": 0.002,
|
| 2334 |
+
"step": 3320
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 2.736236647493837,
|
| 2338 |
+
"grad_norm": 0.07717015273999385,
|
| 2339 |
+
"learning_rate": 8.32097741164616e-05,
|
| 2340 |
+
"loss": 0.0021,
|
| 2341 |
+
"step": 3330
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 2.744453574363188,
|
| 2345 |
+
"grad_norm": 0.07721669499074643,
|
| 2346 |
+
"learning_rate": 8.307554326708815e-05,
|
| 2347 |
+
"loss": 0.002,
|
| 2348 |
+
"step": 3340
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 2.752670501232539,
|
| 2352 |
+
"grad_norm": 0.07578771932626932,
|
| 2353 |
+
"learning_rate": 8.294088721133821e-05,
|
| 2354 |
+
"loss": 0.002,
|
| 2355 |
+
"step": 3350
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 2.76088742810189,
|
| 2359 |
+
"grad_norm": 0.06375027550713391,
|
| 2360 |
+
"learning_rate": 8.28058076802979e-05,
|
| 2361 |
+
"loss": 0.002,
|
| 2362 |
+
"step": 3360
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 2.769104354971241,
|
| 2366 |
+
"grad_norm": 0.06947732837366209,
|
| 2367 |
+
"learning_rate": 8.267030641049738e-05,
|
| 2368 |
+
"loss": 0.0019,
|
| 2369 |
+
"step": 3370
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 2.7773212818405915,
|
| 2373 |
+
"grad_norm": 0.07035332876274755,
|
| 2374 |
+
"learning_rate": 8.25343851438885e-05,
|
| 2375 |
+
"loss": 0.0018,
|
| 2376 |
+
"step": 3380
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 2.7855382087099425,
|
| 2380 |
+
"grad_norm": 0.06994042828521449,
|
| 2381 |
+
"learning_rate": 8.239804562782241e-05,
|
| 2382 |
+
"loss": 0.0019,
|
| 2383 |
+
"step": 3390
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 2.7937551355792936,
|
| 2387 |
+
"grad_norm": 0.0739022246469536,
|
| 2388 |
+
"learning_rate": 8.226128961502715e-05,
|
| 2389 |
+
"loss": 0.0019,
|
| 2390 |
+
"step": 3400
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 2.801972062448644,
|
| 2394 |
+
"grad_norm": 0.08090723440170883,
|
| 2395 |
+
"learning_rate": 8.212411886358505e-05,
|
| 2396 |
+
"loss": 0.0018,
|
| 2397 |
+
"step": 3410
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 2.810188989317995,
|
| 2401 |
+
"grad_norm": 0.06668617007404189,
|
| 2402 |
+
"learning_rate": 8.198653513691019e-05,
|
| 2403 |
+
"loss": 0.0018,
|
| 2404 |
+
"step": 3420
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 2.818405916187346,
|
| 2408 |
+
"grad_norm": 0.08112145655425776,
|
| 2409 |
+
"learning_rate": 8.184854020372567e-05,
|
| 2410 |
+
"loss": 0.0019,
|
| 2411 |
+
"step": 3430
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 2.826622843056697,
|
| 2415 |
+
"grad_norm": 0.06838412780008499,
|
| 2416 |
+
"learning_rate": 8.171013583804095e-05,
|
| 2417 |
+
"loss": 0.0018,
|
| 2418 |
+
"step": 3440
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 2.834839769926048,
|
| 2422 |
+
"grad_norm": 0.06801532642214224,
|
| 2423 |
+
"learning_rate": 8.157132381912894e-05,
|
| 2424 |
+
"loss": 0.0018,
|
| 2425 |
+
"step": 3450
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 2.8430566967953985,
|
| 2429 |
+
"grad_norm": 0.07450069669807256,
|
| 2430 |
+
"learning_rate": 8.143210593150317e-05,
|
| 2431 |
+
"loss": 0.0018,
|
| 2432 |
+
"step": 3460
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 2.851273623664749,
|
| 2436 |
+
"grad_norm": 0.07144863614082946,
|
| 2437 |
+
"learning_rate": 8.129248396489493e-05,
|
| 2438 |
+
"loss": 0.0018,
|
| 2439 |
+
"step": 3470
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 2.8594905505341,
|
| 2443 |
+
"grad_norm": 0.06885416663157976,
|
| 2444 |
+
"learning_rate": 8.115245971423016e-05,
|
| 2445 |
+
"loss": 0.0019,
|
| 2446 |
+
"step": 3480
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 2.867707477403451,
|
| 2450 |
+
"grad_norm": 0.07359978466021078,
|
| 2451 |
+
"learning_rate": 8.101203497960639e-05,
|
| 2452 |
+
"loss": 0.0019,
|
| 2453 |
+
"step": 3490
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 2.875924404272802,
|
| 2457 |
+
"grad_norm": 0.06305662725881052,
|
| 2458 |
+
"learning_rate": 8.087121156626965e-05,
|
| 2459 |
+
"loss": 0.0017,
|
| 2460 |
+
"step": 3500
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 2.884141331142153,
|
| 2464 |
+
"grad_norm": 0.06846870722129733,
|
| 2465 |
+
"learning_rate": 8.072999128459119e-05,
|
| 2466 |
+
"loss": 0.0019,
|
| 2467 |
+
"step": 3510
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 2.8923582580115035,
|
| 2471 |
+
"grad_norm": 0.07241840468071385,
|
| 2472 |
+
"learning_rate": 8.05883759500443e-05,
|
| 2473 |
+
"loss": 0.0017,
|
| 2474 |
+
"step": 3520
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 2.9005751848808545,
|
| 2478 |
+
"grad_norm": 0.06286985529378947,
|
| 2479 |
+
"learning_rate": 8.044636738318082e-05,
|
| 2480 |
+
"loss": 0.0016,
|
| 2481 |
+
"step": 3530
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 2.9087921117502056,
|
| 2485 |
+
"grad_norm": 0.07986920999372275,
|
| 2486 |
+
"learning_rate": 8.030396740960795e-05,
|
| 2487 |
+
"loss": 0.0017,
|
| 2488 |
+
"step": 3540
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 2.917009038619556,
|
| 2492 |
+
"grad_norm": 0.06958100326685528,
|
| 2493 |
+
"learning_rate": 8.01611778599646e-05,
|
| 2494 |
+
"loss": 0.0017,
|
| 2495 |
+
"step": 3550
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 2.9252259654889072,
|
| 2499 |
+
"grad_norm": 0.06983385948648055,
|
| 2500 |
+
"learning_rate": 8.001800056989792e-05,
|
| 2501 |
+
"loss": 0.0017,
|
| 2502 |
+
"step": 3560
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 2.933442892358258,
|
| 2506 |
+
"grad_norm": 0.06621268583147509,
|
| 2507 |
+
"learning_rate": 7.987443738003973e-05,
|
| 2508 |
+
"loss": 0.0018,
|
| 2509 |
+
"step": 3570
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 2.941659819227609,
|
| 2513 |
+
"grad_norm": 0.06811521015654298,
|
| 2514 |
+
"learning_rate": 7.973049013598277e-05,
|
| 2515 |
+
"loss": 0.0017,
|
| 2516 |
+
"step": 3580
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 2.94987674609696,
|
| 2520 |
+
"grad_norm": 0.06577394956829613,
|
| 2521 |
+
"learning_rate": 7.958616068825709e-05,
|
| 2522 |
+
"loss": 0.0018,
|
| 2523 |
+
"step": 3590
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 2.9580936729663105,
|
| 2527 |
+
"grad_norm": 0.08652052889965613,
|
| 2528 |
+
"learning_rate": 7.944145089230619e-05,
|
| 2529 |
+
"loss": 0.0019,
|
| 2530 |
+
"step": 3600
|
| 2531 |
+
}
|
| 2532 |
+
],
|
| 2533 |
+
"logging_steps": 10,
|
| 2534 |
+
"max_steps": 9736,
|
| 2535 |
+
"num_input_tokens_seen": 0,
|
| 2536 |
+
"num_train_epochs": 8,
|
| 2537 |
+
"save_steps": 1800,
|
| 2538 |
+
"stateful_callbacks": {
|
| 2539 |
+
"TrainerControl": {
|
| 2540 |
+
"args": {
|
| 2541 |
+
"should_epoch_stop": false,
|
| 2542 |
+
"should_evaluate": false,
|
| 2543 |
+
"should_log": false,
|
| 2544 |
+
"should_save": true,
|
| 2545 |
+
"should_training_stop": false
|
| 2546 |
+
},
|
| 2547 |
+
"attributes": {}
|
| 2548 |
+
}
|
| 2549 |
+
},
|
| 2550 |
+
"total_flos": 345773957971968.0,
|
| 2551 |
+
"train_batch_size": 128,
|
| 2552 |
+
"trial_name": null,
|
| 2553 |
+
"trial_params": null
|
| 2554 |
+
}
|