Upload model and tokenizer
Browse files- checkpoint-52896/trainer_state.json +1511 -0
checkpoint-52896/trainer_state.json
ADDED
|
@@ -0,0 +1,1511 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.9999432860086583,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 52896,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.00945233189027733,
|
| 14 |
+
"grad_norm": 29.5,
|
| 15 |
+
"learning_rate": 9.960000000000001e-06,
|
| 16 |
+
"loss": 4.1004,
|
| 17 |
+
"step": 250
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.01890466378055466,
|
| 21 |
+
"grad_norm": 20.125,
|
| 22 |
+
"learning_rate": 9.999448050049255e-06,
|
| 23 |
+
"loss": 4.0386,
|
| 24 |
+
"step": 500
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.028356995670831994,
|
| 28 |
+
"grad_norm": 16.625,
|
| 29 |
+
"learning_rate": 9.997783447634044e-06,
|
| 30 |
+
"loss": 4.1263,
|
| 31 |
+
"step": 750
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.03780932756110932,
|
| 35 |
+
"grad_norm": 18.5,
|
| 36 |
+
"learning_rate": 9.995006554320588e-06,
|
| 37 |
+
"loss": 3.9826,
|
| 38 |
+
"step": 1000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047261659451386655,
|
| 42 |
+
"grad_norm": 18.625,
|
| 43 |
+
"learning_rate": 9.991117988125487e-06,
|
| 44 |
+
"loss": 4.0093,
|
| 45 |
+
"step": 1250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.05671399134166399,
|
| 49 |
+
"grad_norm": 12.375,
|
| 50 |
+
"learning_rate": 9.986118614475757e-06,
|
| 51 |
+
"loss": 3.9503,
|
| 52 |
+
"step": 1500
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.06616632323194133,
|
| 56 |
+
"grad_norm": 15.8125,
|
| 57 |
+
"learning_rate": 9.980009546016204e-06,
|
| 58 |
+
"loss": 4.0245,
|
| 59 |
+
"step": 1750
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.07561865512221864,
|
| 63 |
+
"grad_norm": 17.5,
|
| 64 |
+
"learning_rate": 9.972792142361807e-06,
|
| 65 |
+
"loss": 3.9901,
|
| 66 |
+
"step": 2000
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.08507098701249598,
|
| 70 |
+
"grad_norm": 22.75,
|
| 71 |
+
"learning_rate": 9.964468009795128e-06,
|
| 72 |
+
"loss": 3.9098,
|
| 73 |
+
"step": 2250
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.09452331890277331,
|
| 77 |
+
"grad_norm": 19.5,
|
| 78 |
+
"learning_rate": 9.95503900090882e-06,
|
| 79 |
+
"loss": 3.9788,
|
| 80 |
+
"step": 2500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.10397565079305064,
|
| 84 |
+
"grad_norm": 17.375,
|
| 85 |
+
"learning_rate": 9.944507214193314e-06,
|
| 86 |
+
"loss": 4.0492,
|
| 87 |
+
"step": 2750
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.11342798268332797,
|
| 91 |
+
"grad_norm": 16.75,
|
| 92 |
+
"learning_rate": 9.932874993569803e-06,
|
| 93 |
+
"loss": 3.9152,
|
| 94 |
+
"step": 3000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.1228803145736053,
|
| 98 |
+
"grad_norm": 19.625,
|
| 99 |
+
"learning_rate": 9.92014492786856e-06,
|
| 100 |
+
"loss": 4.0289,
|
| 101 |
+
"step": 3250
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.13233264646388265,
|
| 105 |
+
"grad_norm": 26.75,
|
| 106 |
+
"learning_rate": 9.906319850252806e-06,
|
| 107 |
+
"loss": 3.9419,
|
| 108 |
+
"step": 3500
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.14178497835415999,
|
| 112 |
+
"grad_norm": 17.875,
|
| 113 |
+
"learning_rate": 9.891402837588142e-06,
|
| 114 |
+
"loss": 3.9255,
|
| 115 |
+
"step": 3750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.1512373102444373,
|
| 119 |
+
"grad_norm": 17.0,
|
| 120 |
+
"learning_rate": 9.875397209757793e-06,
|
| 121 |
+
"loss": 4.068,
|
| 122 |
+
"step": 4000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.16068964213471462,
|
| 126 |
+
"grad_norm": 22.375,
|
| 127 |
+
"learning_rate": 9.858306528923734e-06,
|
| 128 |
+
"loss": 3.9229,
|
| 129 |
+
"step": 4250
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.17014197402499195,
|
| 133 |
+
"grad_norm": 15.8125,
|
| 134 |
+
"learning_rate": 9.840134598733906e-06,
|
| 135 |
+
"loss": 3.8975,
|
| 136 |
+
"step": 4500
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.1795943059152693,
|
| 140 |
+
"grad_norm": 17.875,
|
| 141 |
+
"learning_rate": 9.8208854634757e-06,
|
| 142 |
+
"loss": 3.9159,
|
| 143 |
+
"step": 4750
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.18904663780554662,
|
| 147 |
+
"grad_norm": 19.0,
|
| 148 |
+
"learning_rate": 9.800563407175856e-06,
|
| 149 |
+
"loss": 3.9892,
|
| 150 |
+
"step": 5000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.19849896969582395,
|
| 154 |
+
"grad_norm": 24.0,
|
| 155 |
+
"learning_rate": 9.779172952647035e-06,
|
| 156 |
+
"loss": 3.9846,
|
| 157 |
+
"step": 5250
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.20795130158610128,
|
| 161 |
+
"grad_norm": 15.9375,
|
| 162 |
+
"learning_rate": 9.756718860481235e-06,
|
| 163 |
+
"loss": 4.0746,
|
| 164 |
+
"step": 5500
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.21740363347637862,
|
| 168 |
+
"grad_norm": 19.25,
|
| 169 |
+
"learning_rate": 9.733206127990285e-06,
|
| 170 |
+
"loss": 3.9736,
|
| 171 |
+
"step": 5750
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.22685596536665595,
|
| 175 |
+
"grad_norm": 17.25,
|
| 176 |
+
"learning_rate": 9.708639988093663e-06,
|
| 177 |
+
"loss": 3.9673,
|
| 178 |
+
"step": 6000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.23630829725693328,
|
| 182 |
+
"grad_norm": 17.375,
|
| 183 |
+
"learning_rate": 9.683025908153868e-06,
|
| 184 |
+
"loss": 3.9672,
|
| 185 |
+
"step": 6250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.2457606291472106,
|
| 189 |
+
"grad_norm": 17.375,
|
| 190 |
+
"learning_rate": 9.656369588759628e-06,
|
| 191 |
+
"loss": 3.9812,
|
| 192 |
+
"step": 6500
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.255212961037488,
|
| 196 |
+
"grad_norm": 17.625,
|
| 197 |
+
"learning_rate": 9.628676962457194e-06,
|
| 198 |
+
"loss": 3.9659,
|
| 199 |
+
"step": 6750
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.2646652929277653,
|
| 203 |
+
"grad_norm": 18.0,
|
| 204 |
+
"learning_rate": 9.599954192430004e-06,
|
| 205 |
+
"loss": 3.9614,
|
| 206 |
+
"step": 7000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.27411762481804264,
|
| 210 |
+
"grad_norm": 17.125,
|
| 211 |
+
"learning_rate": 9.570207671127034e-06,
|
| 212 |
+
"loss": 3.9424,
|
| 213 |
+
"step": 7250
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.28356995670831997,
|
| 217 |
+
"grad_norm": 24.125,
|
| 218 |
+
"learning_rate": 9.539444018840107e-06,
|
| 219 |
+
"loss": 3.9533,
|
| 220 |
+
"step": 7500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.29302228859859725,
|
| 224 |
+
"grad_norm": 20.375,
|
| 225 |
+
"learning_rate": 9.507670082230507e-06,
|
| 226 |
+
"loss": 4.0344,
|
| 227 |
+
"step": 7750
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.3024746204888746,
|
| 231 |
+
"grad_norm": 18.375,
|
| 232 |
+
"learning_rate": 9.474892932805209e-06,
|
| 233 |
+
"loss": 3.8986,
|
| 234 |
+
"step": 8000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.3119269523791519,
|
| 238 |
+
"grad_norm": 19.5,
|
| 239 |
+
"learning_rate": 9.441119865343054e-06,
|
| 240 |
+
"loss": 3.9415,
|
| 241 |
+
"step": 8250
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.32137928426942924,
|
| 245 |
+
"grad_norm": 20.125,
|
| 246 |
+
"learning_rate": 9.406358396271266e-06,
|
| 247 |
+
"loss": 3.9542,
|
| 248 |
+
"step": 8500
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.3308316161597066,
|
| 252 |
+
"grad_norm": 25.375,
|
| 253 |
+
"learning_rate": 9.370616261992605e-06,
|
| 254 |
+
"loss": 4.0098,
|
| 255 |
+
"step": 8750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.3402839480499839,
|
| 259 |
+
"grad_norm": 19.625,
|
| 260 |
+
"learning_rate": 9.33390141716358e-06,
|
| 261 |
+
"loss": 3.9839,
|
| 262 |
+
"step": 9000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.34973627994026124,
|
| 266 |
+
"grad_norm": 19.5,
|
| 267 |
+
"learning_rate": 9.296222032924092e-06,
|
| 268 |
+
"loss": 3.9886,
|
| 269 |
+
"step": 9250
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.3591886118305386,
|
| 273 |
+
"grad_norm": 16.75,
|
| 274 |
+
"learning_rate": 9.257586495078882e-06,
|
| 275 |
+
"loss": 3.8992,
|
| 276 |
+
"step": 9500
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.3686409437208159,
|
| 280 |
+
"grad_norm": 17.5,
|
| 281 |
+
"learning_rate": 9.21800340223122e-06,
|
| 282 |
+
"loss": 4.0108,
|
| 283 |
+
"step": 9750
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.37809327561109324,
|
| 287 |
+
"grad_norm": 17.375,
|
| 288 |
+
"learning_rate": 9.177481563869226e-06,
|
| 289 |
+
"loss": 3.9957,
|
| 290 |
+
"step": 10000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.38754560750137057,
|
| 294 |
+
"grad_norm": 19.625,
|
| 295 |
+
"learning_rate": 9.136029998405253e-06,
|
| 296 |
+
"loss": 3.966,
|
| 297 |
+
"step": 10250
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.3969979393916479,
|
| 301 |
+
"grad_norm": 20.125,
|
| 302 |
+
"learning_rate": 9.093657931168782e-06,
|
| 303 |
+
"loss": 4.0057,
|
| 304 |
+
"step": 10500
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.40645027128192524,
|
| 308 |
+
"grad_norm": 25.5,
|
| 309 |
+
"learning_rate": 9.050374792353265e-06,
|
| 310 |
+
"loss": 4.0049,
|
| 311 |
+
"step": 10750
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.41590260317220257,
|
| 315 |
+
"grad_norm": 17.5,
|
| 316 |
+
"learning_rate": 9.006190214917363e-06,
|
| 317 |
+
"loss": 4.046,
|
| 318 |
+
"step": 11000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.4253549350624799,
|
| 322 |
+
"grad_norm": 15.25,
|
| 323 |
+
"learning_rate": 8.961114032441067e-06,
|
| 324 |
+
"loss": 4.0138,
|
| 325 |
+
"step": 11250
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.43480726695275723,
|
| 329 |
+
"grad_norm": 58.5,
|
| 330 |
+
"learning_rate": 8.915156276937175e-06,
|
| 331 |
+
"loss": 4.0145,
|
| 332 |
+
"step": 11500
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.44425959884303456,
|
| 336 |
+
"grad_norm": 18.25,
|
| 337 |
+
"learning_rate": 8.868327176618592e-06,
|
| 338 |
+
"loss": 3.9748,
|
| 339 |
+
"step": 11750
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.4537119307333119,
|
| 343 |
+
"grad_norm": 17.0,
|
| 344 |
+
"learning_rate": 8.82063715362197e-06,
|
| 345 |
+
"loss": 4.0039,
|
| 346 |
+
"step": 12000
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.46316426262358923,
|
| 350 |
+
"grad_norm": 17.875,
|
| 351 |
+
"learning_rate": 8.772096821688194e-06,
|
| 352 |
+
"loss": 4.1231,
|
| 353 |
+
"step": 12250
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.47261659451386656,
|
| 357 |
+
"grad_norm": 19.25,
|
| 358 |
+
"learning_rate": 8.722716983800226e-06,
|
| 359 |
+
"loss": 4.0778,
|
| 360 |
+
"step": 12500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.4820689264041439,
|
| 364 |
+
"grad_norm": 14.0625,
|
| 365 |
+
"learning_rate": 8.672508629778809e-06,
|
| 366 |
+
"loss": 3.9998,
|
| 367 |
+
"step": 12750
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.4915212582944212,
|
| 371 |
+
"grad_norm": 20.0,
|
| 372 |
+
"learning_rate": 8.621482933836634e-06,
|
| 373 |
+
"loss": 4.0298,
|
| 374 |
+
"step": 13000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.5009735901846986,
|
| 378 |
+
"grad_norm": 21.5,
|
| 379 |
+
"learning_rate": 8.569651252091418e-06,
|
| 380 |
+
"loss": 3.9807,
|
| 381 |
+
"step": 13250
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.510425922074976,
|
| 385 |
+
"grad_norm": 17.75,
|
| 386 |
+
"learning_rate": 8.517025120038536e-06,
|
| 387 |
+
"loss": 4.084,
|
| 388 |
+
"step": 13500
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.5198782539652532,
|
| 392 |
+
"grad_norm": 19.0,
|
| 393 |
+
"learning_rate": 8.463616249983718e-06,
|
| 394 |
+
"loss": 4.1373,
|
| 395 |
+
"step": 13750
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.5293305858555306,
|
| 399 |
+
"grad_norm": 19.75,
|
| 400 |
+
"learning_rate": 8.409436528436381e-06,
|
| 401 |
+
"loss": 4.0691,
|
| 402 |
+
"step": 14000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.5387829177458079,
|
| 406 |
+
"grad_norm": 17.25,
|
| 407 |
+
"learning_rate": 8.354498013464228e-06,
|
| 408 |
+
"loss": 4.0686,
|
| 409 |
+
"step": 14250
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.5482352496360853,
|
| 413 |
+
"grad_norm": 16.875,
|
| 414 |
+
"learning_rate": 8.298812932009622e-06,
|
| 415 |
+
"loss": 4.0066,
|
| 416 |
+
"step": 14500
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.5576875815263626,
|
| 420 |
+
"grad_norm": 19.625,
|
| 421 |
+
"learning_rate": 8.242393677168406e-06,
|
| 422 |
+
"loss": 4.0525,
|
| 423 |
+
"step": 14750
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.5671399134166399,
|
| 427 |
+
"grad_norm": 20.5,
|
| 428 |
+
"learning_rate": 8.185252805431732e-06,
|
| 429 |
+
"loss": 4.0993,
|
| 430 |
+
"step": 15000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.5765922453069172,
|
| 434 |
+
"grad_norm": 20.375,
|
| 435 |
+
"learning_rate": 8.127403033891532e-06,
|
| 436 |
+
"loss": 4.0902,
|
| 437 |
+
"step": 15250
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5860445771971945,
|
| 441 |
+
"grad_norm": 21.125,
|
| 442 |
+
"learning_rate": 8.068857237410237e-06,
|
| 443 |
+
"loss": 4.0273,
|
| 444 |
+
"step": 15500
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5954969090874719,
|
| 448 |
+
"grad_norm": 19.875,
|
| 449 |
+
"learning_rate": 8.00962844575539e-06,
|
| 450 |
+
"loss": 4.0831,
|
| 451 |
+
"step": 15750
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.6049492409777492,
|
| 455 |
+
"grad_norm": 243.0,
|
| 456 |
+
"learning_rate": 7.949729840699784e-06,
|
| 457 |
+
"loss": 4.0758,
|
| 458 |
+
"step": 16000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.6144015728680265,
|
| 462 |
+
"grad_norm": 17.875,
|
| 463 |
+
"learning_rate": 7.889174753087767e-06,
|
| 464 |
+
"loss": 4.0918,
|
| 465 |
+
"step": 16250
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.6238539047583038,
|
| 469 |
+
"grad_norm": 19.125,
|
| 470 |
+
"learning_rate": 7.827976659868368e-06,
|
| 471 |
+
"loss": 4.0538,
|
| 472 |
+
"step": 16500
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.6333062366485812,
|
| 476 |
+
"grad_norm": 19.625,
|
| 477 |
+
"learning_rate": 7.766149181095916e-06,
|
| 478 |
+
"loss": 4.1164,
|
| 479 |
+
"step": 16750
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.6427585685388585,
|
| 483 |
+
"grad_norm": 16.5,
|
| 484 |
+
"learning_rate": 7.703706076898803e-06,
|
| 485 |
+
"loss": 4.0626,
|
| 486 |
+
"step": 17000
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.6522109004291359,
|
| 490 |
+
"grad_norm": 14.375,
|
| 491 |
+
"learning_rate": 7.640661244417064e-06,
|
| 492 |
+
"loss": 4.0444,
|
| 493 |
+
"step": 17250
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.6616632323194132,
|
| 497 |
+
"grad_norm": 19.0,
|
| 498 |
+
"learning_rate": 7.577028714709484e-06,
|
| 499 |
+
"loss": 4.0429,
|
| 500 |
+
"step": 17500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.6711155642096905,
|
| 504 |
+
"grad_norm": 17.875,
|
| 505 |
+
"learning_rate": 7.512822649630893e-06,
|
| 506 |
+
"loss": 4.0362,
|
| 507 |
+
"step": 17750
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.6805678960999678,
|
| 511 |
+
"grad_norm": 15.625,
|
| 512 |
+
"learning_rate": 7.44805733868033e-06,
|
| 513 |
+
"loss": 4.0806,
|
| 514 |
+
"step": 18000
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.6900202279902452,
|
| 518 |
+
"grad_norm": 17.625,
|
| 519 |
+
"learning_rate": 7.382747195820834e-06,
|
| 520 |
+
"loss": 4.0933,
|
| 521 |
+
"step": 18250
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.6994725598805225,
|
| 525 |
+
"grad_norm": 16.875,
|
| 526 |
+
"learning_rate": 7.316906756271515e-06,
|
| 527 |
+
"loss": 4.0495,
|
| 528 |
+
"step": 18500
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.7089248917707999,
|
| 532 |
+
"grad_norm": 20.375,
|
| 533 |
+
"learning_rate": 7.250550673272639e-06,
|
| 534 |
+
"loss": 4.0599,
|
| 535 |
+
"step": 18750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.7183772236610771,
|
| 539 |
+
"grad_norm": 19.375,
|
| 540 |
+
"learning_rate": 7.1836937148244445e-06,
|
| 541 |
+
"loss": 4.0653,
|
| 542 |
+
"step": 19000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.7278295555513545,
|
| 546 |
+
"grad_norm": 16.5,
|
| 547 |
+
"learning_rate": 7.1163507604004326e-06,
|
| 548 |
+
"loss": 4.0266,
|
| 549 |
+
"step": 19250
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.7372818874416318,
|
| 553 |
+
"grad_norm": 20.75,
|
| 554 |
+
"learning_rate": 7.048536797635832e-06,
|
| 555 |
+
"loss": 4.0484,
|
| 556 |
+
"step": 19500
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.7467342193319092,
|
| 560 |
+
"grad_norm": 18.25,
|
| 561 |
+
"learning_rate": 6.9802669189920005e-06,
|
| 562 |
+
"loss": 4.043,
|
| 563 |
+
"step": 19750
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.7561865512221865,
|
| 567 |
+
"grad_norm": 47.0,
|
| 568 |
+
"learning_rate": 6.911556318397493e-06,
|
| 569 |
+
"loss": 4.0716,
|
| 570 |
+
"step": 20000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.7656388831124639,
|
| 574 |
+
"grad_norm": 15.625,
|
| 575 |
+
"learning_rate": 6.8424202878665515e-06,
|
| 576 |
+
"loss": 4.059,
|
| 577 |
+
"step": 20250
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.7750912150027411,
|
| 581 |
+
"grad_norm": 18.5,
|
| 582 |
+
"learning_rate": 6.772874214095761e-06,
|
| 583 |
+
"loss": 3.9974,
|
| 584 |
+
"step": 20500
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.7845435468930185,
|
| 588 |
+
"grad_norm": 16.0,
|
| 589 |
+
"learning_rate": 6.702933575039631e-06,
|
| 590 |
+
"loss": 4.0551,
|
| 591 |
+
"step": 20750
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.7939958787832958,
|
| 595 |
+
"grad_norm": 15.5,
|
| 596 |
+
"learning_rate": 6.6326139364658795e-06,
|
| 597 |
+
"loss": 4.1337,
|
| 598 |
+
"step": 21000
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.8034482106735732,
|
| 602 |
+
"grad_norm": 19.125,
|
| 603 |
+
"learning_rate": 6.561930948491155e-06,
|
| 604 |
+
"loss": 4.0849,
|
| 605 |
+
"step": 21250
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.8129005425638505,
|
| 609 |
+
"grad_norm": 19.125,
|
| 610 |
+
"learning_rate": 6.4909003420980065e-06,
|
| 611 |
+
"loss": 4.069,
|
| 612 |
+
"step": 21500
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.8223528744541279,
|
| 616 |
+
"grad_norm": 18.875,
|
| 617 |
+
"learning_rate": 6.419537925633836e-06,
|
| 618 |
+
"loss": 4.0218,
|
| 619 |
+
"step": 21750
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.8318052063444051,
|
| 623 |
+
"grad_norm": 17.125,
|
| 624 |
+
"learning_rate": 6.34785958129265e-06,
|
| 625 |
+
"loss": 3.9901,
|
| 626 |
+
"step": 22000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.8412575382346825,
|
| 630 |
+
"grad_norm": 17.625,
|
| 631 |
+
"learning_rate": 6.275881261580363e-06,
|
| 632 |
+
"loss": 4.0088,
|
| 633 |
+
"step": 22250
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.8507098701249598,
|
| 637 |
+
"grad_norm": 17.0,
|
| 638 |
+
"learning_rate": 6.2036189857644616e-06,
|
| 639 |
+
"loss": 4.0448,
|
| 640 |
+
"step": 22500
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.8601622020152372,
|
| 644 |
+
"grad_norm": 15.8125,
|
| 645 |
+
"learning_rate": 6.131088836308805e-06,
|
| 646 |
+
"loss": 4.0443,
|
| 647 |
+
"step": 22750
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.8696145339055145,
|
| 651 |
+
"grad_norm": 22.875,
|
| 652 |
+
"learning_rate": 6.058306955294365e-06,
|
| 653 |
+
"loss": 4.0573,
|
| 654 |
+
"step": 23000
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.8790668657957919,
|
| 658 |
+
"grad_norm": 54.0,
|
| 659 |
+
"learning_rate": 5.9852895408266955e-06,
|
| 660 |
+
"loss": 4.0054,
|
| 661 |
+
"step": 23250
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.8885191976860691,
|
| 665 |
+
"grad_norm": 21.0,
|
| 666 |
+
"learning_rate": 5.9120528434309245e-06,
|
| 667 |
+
"loss": 4.0112,
|
| 668 |
+
"step": 23500
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.8979715295763465,
|
| 672 |
+
"grad_norm": 19.5,
|
| 673 |
+
"learning_rate": 5.838613162435106e-06,
|
| 674 |
+
"loss": 4.0095,
|
| 675 |
+
"step": 23750
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.9074238614666238,
|
| 679 |
+
"grad_norm": 22.375,
|
| 680 |
+
"learning_rate": 5.764986842342675e-06,
|
| 681 |
+
"loss": 3.9941,
|
| 682 |
+
"step": 24000
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.9168761933569012,
|
| 686 |
+
"grad_norm": 15.875,
|
| 687 |
+
"learning_rate": 5.6911902691948786e-06,
|
| 688 |
+
"loss": 3.9703,
|
| 689 |
+
"step": 24250
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.9263285252471785,
|
| 693 |
+
"grad_norm": 18.0,
|
| 694 |
+
"learning_rate": 5.617239866923945e-06,
|
| 695 |
+
"loss": 3.9949,
|
| 696 |
+
"step": 24500
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.9357808571374558,
|
| 700 |
+
"grad_norm": 18.25,
|
| 701 |
+
"learning_rate": 5.543152093697826e-06,
|
| 702 |
+
"loss": 4.0225,
|
| 703 |
+
"step": 24750
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.9452331890277331,
|
| 707 |
+
"grad_norm": 17.625,
|
| 708 |
+
"learning_rate": 5.4689434382573156e-06,
|
| 709 |
+
"loss": 3.998,
|
| 710 |
+
"step": 25000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.9546855209180105,
|
| 714 |
+
"grad_norm": 16.5,
|
| 715 |
+
"learning_rate": 5.39463041624638e-06,
|
| 716 |
+
"loss": 3.9813,
|
| 717 |
+
"step": 25250
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.9641378528082878,
|
| 721 |
+
"grad_norm": 17.125,
|
| 722 |
+
"learning_rate": 5.320229566536474e-06,
|
| 723 |
+
"loss": 3.9089,
|
| 724 |
+
"step": 25500
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.9735901846985652,
|
| 728 |
+
"grad_norm": 17.875,
|
| 729 |
+
"learning_rate": 5.245757447545706e-06,
|
| 730 |
+
"loss": 4.0302,
|
| 731 |
+
"step": 25750
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.9830425165888425,
|
| 735 |
+
"grad_norm": 17.25,
|
| 736 |
+
"learning_rate": 5.171230633553656e-06,
|
| 737 |
+
"loss": 3.9841,
|
| 738 |
+
"step": 26000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.9924948484791198,
|
| 742 |
+
"grad_norm": 16.125,
|
| 743 |
+
"learning_rate": 5.096665711012646e-06,
|
| 744 |
+
"loss": 3.9648,
|
| 745 |
+
"step": 26250
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.0019282757056165,
|
| 749 |
+
"grad_norm": 21.125,
|
| 750 |
+
"learning_rate": 5.0220792748563195e-06,
|
| 751 |
+
"loss": 3.8978,
|
| 752 |
+
"step": 26500
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.011380607595894,
|
| 756 |
+
"grad_norm": 25.75,
|
| 757 |
+
"learning_rate": 4.94748792480632e-06,
|
| 758 |
+
"loss": 3.4963,
|
| 759 |
+
"step": 26750
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.0208329394861713,
|
| 763 |
+
"grad_norm": 17.375,
|
| 764 |
+
"learning_rate": 4.872908261677911e-06,
|
| 765 |
+
"loss": 3.6178,
|
| 766 |
+
"step": 27000
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.0302852713764485,
|
| 770 |
+
"grad_norm": 25.375,
|
| 771 |
+
"learning_rate": 4.7983568836853564e-06,
|
| 772 |
+
"loss": 3.5309,
|
| 773 |
+
"step": 27250
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.0397376032667258,
|
| 777 |
+
"grad_norm": 26.625,
|
| 778 |
+
"learning_rate": 4.723850382747863e-06,
|
| 779 |
+
"loss": 3.4875,
|
| 780 |
+
"step": 27500
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.0491899351570033,
|
| 784 |
+
"grad_norm": 23.25,
|
| 785 |
+
"learning_rate": 4.649405340796947e-06,
|
| 786 |
+
"loss": 3.5433,
|
| 787 |
+
"step": 27750
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.0586422670472806,
|
| 791 |
+
"grad_norm": 17.125,
|
| 792 |
+
"learning_rate": 4.575038326086007e-06,
|
| 793 |
+
"loss": 3.5867,
|
| 794 |
+
"step": 28000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.0680945989375579,
|
| 798 |
+
"grad_norm": 17.875,
|
| 799 |
+
"learning_rate": 4.500765889502937e-06,
|
| 800 |
+
"loss": 3.5986,
|
| 801 |
+
"step": 28250
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.0775469308278351,
|
| 805 |
+
"grad_norm": 18.0,
|
| 806 |
+
"learning_rate": 4.426604560886636e-06,
|
| 807 |
+
"loss": 3.5402,
|
| 808 |
+
"step": 28500
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.0869992627181126,
|
| 812 |
+
"grad_norm": 17.75,
|
| 813 |
+
"learning_rate": 4.3525708453481505e-06,
|
| 814 |
+
"loss": 3.5184,
|
| 815 |
+
"step": 28750
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.09645159460839,
|
| 819 |
+
"grad_norm": 19.875,
|
| 820 |
+
"learning_rate": 4.278681219597375e-06,
|
| 821 |
+
"loss": 3.5137,
|
| 822 |
+
"step": 29000
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.1059039264986672,
|
| 826 |
+
"grad_norm": 22.375,
|
| 827 |
+
"learning_rate": 4.204952128276027e-06,
|
| 828 |
+
"loss": 3.6344,
|
| 829 |
+
"step": 29250
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.1153562583889445,
|
| 833 |
+
"grad_norm": 19.25,
|
| 834 |
+
"learning_rate": 4.131399980297796e-06,
|
| 835 |
+
"loss": 3.5423,
|
| 836 |
+
"step": 29500
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.124808590279222,
|
| 840 |
+
"grad_norm": 24.25,
|
| 841 |
+
"learning_rate": 4.058041145196414e-06,
|
| 842 |
+
"loss": 3.5973,
|
| 843 |
+
"step": 29750
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.1342609221694993,
|
| 847 |
+
"grad_norm": 20.875,
|
| 848 |
+
"learning_rate": 3.98489194948251e-06,
|
| 849 |
+
"loss": 3.5454,
|
| 850 |
+
"step": 30000
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.1437132540597765,
|
| 854 |
+
"grad_norm": 19.25,
|
| 855 |
+
"learning_rate": 3.911968673010038e-06,
|
| 856 |
+
"loss": 3.5508,
|
| 857 |
+
"step": 30250
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.1531655859500538,
|
| 861 |
+
"grad_norm": 20.0,
|
| 862 |
+
"learning_rate": 3.839287545353076e-06,
|
| 863 |
+
"loss": 3.4747,
|
| 864 |
+
"step": 30500
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.1626179178403313,
|
| 868 |
+
"grad_norm": 21.25,
|
| 869 |
+
"learning_rate": 3.7668647421938275e-06,
|
| 870 |
+
"loss": 3.538,
|
| 871 |
+
"step": 30750
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.1720702497306086,
|
| 875 |
+
"grad_norm": 26.125,
|
| 876 |
+
"learning_rate": 3.694716381722609e-06,
|
| 877 |
+
"loss": 3.5677,
|
| 878 |
+
"step": 31000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.1815225816208859,
|
| 882 |
+
"grad_norm": 25.75,
|
| 883 |
+
"learning_rate": 3.6228585210506427e-06,
|
| 884 |
+
"loss": 3.5912,
|
| 885 |
+
"step": 31250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.1909749135111631,
|
| 889 |
+
"grad_norm": 17.5,
|
| 890 |
+
"learning_rate": 3.551307152636431e-06,
|
| 891 |
+
"loss": 3.5178,
|
| 892 |
+
"step": 31500
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.2004272454014404,
|
| 896 |
+
"grad_norm": 21.875,
|
| 897 |
+
"learning_rate": 3.4800782007265265e-06,
|
| 898 |
+
"loss": 3.5475,
|
| 899 |
+
"step": 31750
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.209879577291718,
|
| 903 |
+
"grad_norm": 28.0,
|
| 904 |
+
"learning_rate": 3.409187517811486e-06,
|
| 905 |
+
"loss": 3.5383,
|
| 906 |
+
"step": 32000
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.2193319091819952,
|
| 910 |
+
"grad_norm": 29.625,
|
| 911 |
+
"learning_rate": 3.3386508810977856e-06,
|
| 912 |
+
"loss": 3.5525,
|
| 913 |
+
"step": 32250
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.2287842410722725,
|
| 917 |
+
"grad_norm": 18.75,
|
| 918 |
+
"learning_rate": 3.2684839889964988e-06,
|
| 919 |
+
"loss": 3.5202,
|
| 920 |
+
"step": 32500
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.23823657296255,
|
| 924 |
+
"grad_norm": 24.0,
|
| 925 |
+
"learning_rate": 3.1987024576295012e-06,
|
| 926 |
+
"loss": 3.5926,
|
| 927 |
+
"step": 32750
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.2476889048528272,
|
| 931 |
+
"grad_norm": 21.0,
|
| 932 |
+
"learning_rate": 3.1293218173540074e-06,
|
| 933 |
+
"loss": 3.4852,
|
| 934 |
+
"step": 33000
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.2571412367431045,
|
| 938 |
+
"grad_norm": 20.125,
|
| 939 |
+
"learning_rate": 3.060357509306171e-06,
|
| 940 |
+
"loss": 3.4994,
|
| 941 |
+
"step": 33250
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.2665935686333818,
|
| 945 |
+
"grad_norm": 20.875,
|
| 946 |
+
"learning_rate": 2.9918248819645624e-06,
|
| 947 |
+
"loss": 3.566,
|
| 948 |
+
"step": 33500
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.276045900523659,
|
| 952 |
+
"grad_norm": 39.75,
|
| 953 |
+
"learning_rate": 2.923739187734258e-06,
|
| 954 |
+
"loss": 3.517,
|
| 955 |
+
"step": 33750
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.2854982324139366,
|
| 959 |
+
"grad_norm": 18.875,
|
| 960 |
+
"learning_rate": 2.8561155795523133e-06,
|
| 961 |
+
"loss": 3.4938,
|
| 962 |
+
"step": 34000
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.2949505643042138,
|
| 966 |
+
"grad_norm": 20.625,
|
| 967 |
+
"learning_rate": 2.788969107515369e-06,
|
| 968 |
+
"loss": 3.501,
|
| 969 |
+
"step": 34250
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.3044028961944911,
|
| 973 |
+
"grad_norm": 22.5,
|
| 974 |
+
"learning_rate": 2.722314715530156e-06,
|
| 975 |
+
"loss": 3.574,
|
| 976 |
+
"step": 34500
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.3138552280847686,
|
| 980 |
+
"grad_norm": 22.375,
|
| 981 |
+
"learning_rate": 2.6561672379876236e-06,
|
| 982 |
+
"loss": 3.4953,
|
| 983 |
+
"step": 34750
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.323307559975046,
|
| 987 |
+
"grad_norm": 20.375,
|
| 988 |
+
"learning_rate": 2.590541396461438e-06,
|
| 989 |
+
"loss": 3.4766,
|
| 990 |
+
"step": 35000
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.3327598918653232,
|
| 994 |
+
"grad_norm": 23.0,
|
| 995 |
+
"learning_rate": 2.5254517964316084e-06,
|
| 996 |
+
"loss": 3.4905,
|
| 997 |
+
"step": 35250
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.3422122237556005,
|
| 1001 |
+
"grad_norm": 24.0,
|
| 1002 |
+
"learning_rate": 2.4609129240339253e-06,
|
| 1003 |
+
"loss": 3.5543,
|
| 1004 |
+
"step": 35500
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.3516645556458777,
|
| 1008 |
+
"grad_norm": 23.125,
|
| 1009 |
+
"learning_rate": 2.39693914283598e-06,
|
| 1010 |
+
"loss": 3.5577,
|
| 1011 |
+
"step": 35750
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.3611168875361552,
|
| 1015 |
+
"grad_norm": 19.125,
|
| 1016 |
+
"learning_rate": 2.333544690640451e-06,
|
| 1017 |
+
"loss": 3.515,
|
| 1018 |
+
"step": 36000
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.3705692194264325,
|
| 1022 |
+
"grad_norm": 22.625,
|
| 1023 |
+
"learning_rate": 2.270743676316383e-06,
|
| 1024 |
+
"loss": 3.506,
|
| 1025 |
+
"step": 36250
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.3800215513167098,
|
| 1029 |
+
"grad_norm": 18.625,
|
| 1030 |
+
"learning_rate": 2.20855007665916e-06,
|
| 1031 |
+
"loss": 3.5909,
|
| 1032 |
+
"step": 36500
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.3894738832069873,
|
| 1036 |
+
"grad_norm": 20.875,
|
| 1037 |
+
"learning_rate": 2.1469777332798804e-06,
|
| 1038 |
+
"loss": 3.5444,
|
| 1039 |
+
"step": 36750
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.3989262150972646,
|
| 1043 |
+
"grad_norm": 19.375,
|
| 1044 |
+
"learning_rate": 2.086040349524807e-06,
|
| 1045 |
+
"loss": 3.5885,
|
| 1046 |
+
"step": 37000
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.4083785469875418,
|
| 1050 |
+
"grad_norm": 20.5,
|
| 1051 |
+
"learning_rate": 2.025751487425591e-06,
|
| 1052 |
+
"loss": 3.5437,
|
| 1053 |
+
"step": 37250
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.4178308788778191,
|
| 1057 |
+
"grad_norm": 22.75,
|
| 1058 |
+
"learning_rate": 1.9661245646809546e-06,
|
| 1059 |
+
"loss": 3.5815,
|
| 1060 |
+
"step": 37500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.4272832107680964,
|
| 1064 |
+
"grad_norm": 24.375,
|
| 1065 |
+
"learning_rate": 1.9071728516704897e-06,
|
| 1066 |
+
"loss": 3.5147,
|
| 1067 |
+
"step": 37750
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.4367355426583739,
|
| 1071 |
+
"grad_norm": 25.75,
|
| 1072 |
+
"learning_rate": 1.8489094685012394e-06,
|
| 1073 |
+
"loss": 3.5288,
|
| 1074 |
+
"step": 38000
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.4461878745486512,
|
| 1078 |
+
"grad_norm": 16.25,
|
| 1079 |
+
"learning_rate": 1.7913473820877353e-06,
|
| 1080 |
+
"loss": 3.5381,
|
| 1081 |
+
"step": 38250
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.4556402064389284,
|
| 1085 |
+
"grad_norm": 22.375,
|
| 1086 |
+
"learning_rate": 1.7344994032661116e-06,
|
| 1087 |
+
"loss": 3.5954,
|
| 1088 |
+
"step": 38500
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.465092538329206,
|
| 1092 |
+
"grad_norm": 19.125,
|
| 1093 |
+
"learning_rate": 1.6783781839429785e-06,
|
| 1094 |
+
"loss": 3.5212,
|
| 1095 |
+
"step": 38750
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.4745448702194832,
|
| 1099 |
+
"grad_norm": 23.25,
|
| 1100 |
+
"learning_rate": 1.6229962142796469e-06,
|
| 1101 |
+
"loss": 3.4585,
|
| 1102 |
+
"step": 39000
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.4839972021097605,
|
| 1106 |
+
"grad_norm": 18.375,
|
| 1107 |
+
"learning_rate": 1.5683658199123524e-06,
|
| 1108 |
+
"loss": 3.5231,
|
| 1109 |
+
"step": 39250
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.4934495340000378,
|
| 1113 |
+
"grad_norm": 25.0,
|
| 1114 |
+
"learning_rate": 1.5144991592091162e-06,
|
| 1115 |
+
"loss": 3.5881,
|
| 1116 |
+
"step": 39500
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.502901865890315,
|
| 1120 |
+
"grad_norm": 22.25,
|
| 1121 |
+
"learning_rate": 1.461408220563803e-06,
|
| 1122 |
+
"loss": 3.5278,
|
| 1123 |
+
"step": 39750
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.5123541977805925,
|
| 1127 |
+
"grad_norm": 18.625,
|
| 1128 |
+
"learning_rate": 1.4091048197280227e-06,
|
| 1129 |
+
"loss": 3.5224,
|
| 1130 |
+
"step": 40000
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.5218065296708698,
|
| 1134 |
+
"grad_norm": 21.5,
|
| 1135 |
+
"learning_rate": 1.3576005971814627e-06,
|
| 1136 |
+
"loss": 3.5465,
|
| 1137 |
+
"step": 40250
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.531258861561147,
|
| 1141 |
+
"grad_norm": 19.375,
|
| 1142 |
+
"learning_rate": 1.3069070155412145e-06,
|
| 1143 |
+
"loss": 3.5318,
|
| 1144 |
+
"step": 40500
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.5407111934514246,
|
| 1148 |
+
"grad_norm": 19.5,
|
| 1149 |
+
"learning_rate": 1.2570353570106864e-06,
|
| 1150 |
+
"loss": 3.5316,
|
| 1151 |
+
"step": 40750
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.5501635253417017,
|
| 1155 |
+
"grad_norm": 23.5,
|
| 1156 |
+
"learning_rate": 1.2079967208686787e-06,
|
| 1157 |
+
"loss": 3.5112,
|
| 1158 |
+
"step": 41000
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.5596158572319792,
|
| 1162 |
+
"grad_norm": 21.25,
|
| 1163 |
+
"learning_rate": 1.159802020999159e-06,
|
| 1164 |
+
"loss": 3.5891,
|
| 1165 |
+
"step": 41250
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.5690681891222564,
|
| 1169 |
+
"grad_norm": 22.375,
|
| 1170 |
+
"learning_rate": 1.112461983462304e-06,
|
| 1171 |
+
"loss": 3.5365,
|
| 1172 |
+
"step": 41500
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.5785205210125337,
|
| 1176 |
+
"grad_norm": 18.25,
|
| 1177 |
+
"learning_rate": 1.0659871441073422e-06,
|
| 1178 |
+
"loss": 3.5665,
|
| 1179 |
+
"step": 41750
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.5879728529028112,
|
| 1183 |
+
"grad_norm": 37.75,
|
| 1184 |
+
"learning_rate": 1.020387846227724e-06,
|
| 1185 |
+
"loss": 3.5764,
|
| 1186 |
+
"step": 42000
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.5974251847930885,
|
| 1190 |
+
"grad_norm": 21.75,
|
| 1191 |
+
"learning_rate": 9.756742382591577e-07,
|
| 1192 |
+
"loss": 3.6041,
|
| 1193 |
+
"step": 42250
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.6068775166833658,
|
| 1197 |
+
"grad_norm": 24.375,
|
| 1198 |
+
"learning_rate": 9.318562715210039e-07,
|
| 1199 |
+
"loss": 3.6046,
|
| 1200 |
+
"step": 42500
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.6163298485736433,
|
| 1204 |
+
"grad_norm": 18.375,
|
| 1205 |
+
"learning_rate": 8.889436980015336e-07,
|
| 1206 |
+
"loss": 3.5789,
|
| 1207 |
+
"step": 42750
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.6257821804639203,
|
| 1211 |
+
"grad_norm": 19.875,
|
| 1212 |
+
"learning_rate": 8.469460681875674e-07,
|
| 1213 |
+
"loss": 3.588,
|
| 1214 |
+
"step": 43000
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.6352345123541978,
|
| 1218 |
+
"grad_norm": 20.375,
|
| 1219 |
+
"learning_rate": 8.058727289389485e-07,
|
| 1220 |
+
"loss": 3.571,
|
| 1221 |
+
"step": 43250
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.644686844244475,
|
| 1225 |
+
"grad_norm": 33.25,
|
| 1226 |
+
"learning_rate": 7.657328214083226e-07,
|
| 1227 |
+
"loss": 3.5252,
|
| 1228 |
+
"step": 43500
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.6541391761347524,
|
| 1232 |
+
"grad_norm": 20.625,
|
| 1233 |
+
"learning_rate": 7.26535279006727e-07,
|
| 1234 |
+
"loss": 3.5418,
|
| 1235 |
+
"step": 43750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.6635915080250299,
|
| 1239 |
+
"grad_norm": 20.25,
|
| 1240 |
+
"learning_rate": 6.882888254153902e-07,
|
| 1241 |
+
"loss": 3.475,
|
| 1242 |
+
"step": 44000
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.6730438399153071,
|
| 1246 |
+
"grad_norm": 22.25,
|
| 1247 |
+
"learning_rate": 6.51001972644218e-07,
|
| 1248 |
+
"loss": 3.6097,
|
| 1249 |
+
"step": 44250
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.6824961718055844,
|
| 1253 |
+
"grad_norm": 20.125,
|
| 1254 |
+
"learning_rate": 6.146830191373909e-07,
|
| 1255 |
+
"loss": 3.5361,
|
| 1256 |
+
"step": 44500
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.691948503695862,
|
| 1260 |
+
"grad_norm": 19.375,
|
| 1261 |
+
"learning_rate": 5.793400479264849e-07,
|
| 1262 |
+
"loss": 3.5127,
|
| 1263 |
+
"step": 44750
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.701400835586139,
|
| 1267 |
+
"grad_norm": 22.5,
|
| 1268 |
+
"learning_rate": 5.449809248315402e-07,
|
| 1269 |
+
"loss": 3.5631,
|
| 1270 |
+
"step": 45000
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.7108531674764165,
|
| 1274 |
+
"grad_norm": 20.125,
|
| 1275 |
+
"learning_rate": 5.11613296710467e-07,
|
| 1276 |
+
"loss": 3.4704,
|
| 1277 |
+
"step": 45250
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.7203054993666937,
|
| 1281 |
+
"grad_norm": 25.75,
|
| 1282 |
+
"learning_rate": 4.792445897571845e-07,
|
| 1283 |
+
"loss": 3.5528,
|
| 1284 |
+
"step": 45500
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.729757831256971,
|
| 1288 |
+
"grad_norm": 24.0,
|
| 1289 |
+
"learning_rate": 4.478820078488749e-07,
|
| 1290 |
+
"loss": 3.515,
|
| 1291 |
+
"step": 45750
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.7392101631472485,
|
| 1295 |
+
"grad_norm": 17.375,
|
| 1296 |
+
"learning_rate": 4.175325309427064e-07,
|
| 1297 |
+
"loss": 3.5246,
|
| 1298 |
+
"step": 46000
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.7486624950375258,
|
| 1302 |
+
"grad_norm": 21.875,
|
| 1303 |
+
"learning_rate": 3.882029135223975e-07,
|
| 1304 |
+
"loss": 3.504,
|
| 1305 |
+
"step": 46250
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.758114826927803,
|
| 1309 |
+
"grad_norm": 23.0,
|
| 1310 |
+
"learning_rate": 3.598996830949619e-07,
|
| 1311 |
+
"loss": 3.5885,
|
| 1312 |
+
"step": 46500
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.7675671588180806,
|
| 1316 |
+
"grad_norm": 29.0,
|
| 1317 |
+
"learning_rate": 3.326291387379654e-07,
|
| 1318 |
+
"loss": 3.5235,
|
| 1319 |
+
"step": 46750
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.7770194907083576,
|
| 1323 |
+
"grad_norm": 20.0,
|
| 1324 |
+
"learning_rate": 3.0639734969762524e-07,
|
| 1325 |
+
"loss": 3.5873,
|
| 1326 |
+
"step": 47000
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.7864718225986351,
|
| 1330 |
+
"grad_norm": 20.75,
|
| 1331 |
+
"learning_rate": 2.8121015403805406e-07,
|
| 1332 |
+
"loss": 3.501,
|
| 1333 |
+
"step": 47250
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.7959241544889124,
|
| 1337 |
+
"grad_norm": 19.375,
|
| 1338 |
+
"learning_rate": 2.570731573419638e-07,
|
| 1339 |
+
"loss": 3.4923,
|
| 1340 |
+
"step": 47500
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.8053764863791897,
|
| 1344 |
+
"grad_norm": 23.625,
|
| 1345 |
+
"learning_rate": 2.3399173146309906e-07,
|
| 1346 |
+
"loss": 3.4967,
|
| 1347 |
+
"step": 47750
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.8148288182694672,
|
| 1351 |
+
"grad_norm": 16.875,
|
| 1352 |
+
"learning_rate": 2.119710133306996e-07,
|
| 1353 |
+
"loss": 3.5254,
|
| 1354 |
+
"step": 48000
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.8242811501597445,
|
| 1358 |
+
"grad_norm": 19.625,
|
| 1359 |
+
"learning_rate": 1.9101590380623925e-07,
|
| 1360 |
+
"loss": 3.4922,
|
| 1361 |
+
"step": 48250
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.8337334820500217,
|
| 1365 |
+
"grad_norm": 23.125,
|
| 1366 |
+
"learning_rate": 1.711310665927046e-07,
|
| 1367 |
+
"loss": 3.5446,
|
| 1368 |
+
"step": 48500
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.8431858139402992,
|
| 1372 |
+
"grad_norm": 25.375,
|
| 1373 |
+
"learning_rate": 1.5232092719666025e-07,
|
| 1374 |
+
"loss": 3.5395,
|
| 1375 |
+
"step": 48750
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.8526381458305763,
|
| 1379 |
+
"grad_norm": 20.0,
|
| 1380 |
+
"learning_rate": 1.3458967194331485e-07,
|
| 1381 |
+
"loss": 3.5714,
|
| 1382 |
+
"step": 49000
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.8620904777208538,
|
| 1386 |
+
"grad_norm": 23.0,
|
| 1387 |
+
"learning_rate": 1.1794124704483324e-07,
|
| 1388 |
+
"loss": 3.5383,
|
| 1389 |
+
"step": 49250
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.871542809611131,
|
| 1393 |
+
"grad_norm": 16.875,
|
| 1394 |
+
"learning_rate": 1.0237935772207608e-07,
|
| 1395 |
+
"loss": 3.5393,
|
| 1396 |
+
"step": 49500
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.8809951415014083,
|
| 1400 |
+
"grad_norm": 18.25,
|
| 1401 |
+
"learning_rate": 8.790746737997569e-08,
|
| 1402 |
+
"loss": 3.5028,
|
| 1403 |
+
"step": 49750
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.8904474733916858,
|
| 1407 |
+
"grad_norm": 19.5,
|
| 1408 |
+
"learning_rate": 7.452879683673809e-08,
|
| 1409 |
+
"loss": 3.5613,
|
| 1410 |
+
"step": 50000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.899899805281963,
|
| 1414 |
+
"grad_norm": 21.25,
|
| 1415 |
+
"learning_rate": 6.224632360702143e-08,
|
| 1416 |
+
"loss": 3.5304,
|
| 1417 |
+
"step": 50250
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 1.9093521371722404,
|
| 1421 |
+
"grad_norm": 21.75,
|
| 1422 |
+
"learning_rate": 5.1062781239271665e-08,
|
| 1423 |
+
"loss": 3.5906,
|
| 1424 |
+
"step": 50500
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.9188044690625177,
|
| 1428 |
+
"grad_norm": 21.75,
|
| 1429 |
+
"learning_rate": 4.0980658707355234e-08,
|
| 1430 |
+
"loss": 3.5713,
|
| 1431 |
+
"step": 50750
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 1.928256800952795,
|
| 1435 |
+
"grad_norm": 25.125,
|
| 1436 |
+
"learning_rate": 3.2002199856617236e-08,
|
| 1437 |
+
"loss": 3.5918,
|
| 1438 |
+
"step": 51000
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 1.9377091328430724,
|
| 1442 |
+
"grad_norm": 19.625,
|
| 1443 |
+
"learning_rate": 2.412940290450083e-08,
|
| 1444 |
+
"loss": 3.4949,
|
| 1445 |
+
"step": 51250
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 1.9471614647333497,
|
| 1449 |
+
"grad_norm": 21.5,
|
| 1450 |
+
"learning_rate": 1.736401999582804e-08,
|
| 1451 |
+
"loss": 3.5472,
|
| 1452 |
+
"step": 51500
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 1.956613796623627,
|
| 1456 |
+
"grad_norm": 23.375,
|
| 1457 |
+
"learning_rate": 1.1707556812851074e-08,
|
| 1458 |
+
"loss": 3.5542,
|
| 1459 |
+
"step": 51750
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 1.9660661285139045,
|
| 1463 |
+
"grad_norm": 24.25,
|
| 1464 |
+
"learning_rate": 7.161272240148731e-09,
|
| 1465 |
+
"loss": 3.5946,
|
| 1466 |
+
"step": 52000
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 1.9755184604041816,
|
| 1470 |
+
"grad_norm": 20.875,
|
| 1471 |
+
"learning_rate": 3.726178084456078e-09,
|
| 1472 |
+
"loss": 3.5614,
|
| 1473 |
+
"step": 52250
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 1.984970792294459,
|
| 1477 |
+
"grad_norm": 19.625,
|
| 1478 |
+
"learning_rate": 1.4030388494790104e-09,
|
| 1479 |
+
"loss": 3.5535,
|
| 1480 |
+
"step": 52500
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.9944231241847363,
|
| 1484 |
+
"grad_norm": 24.25,
|
| 1485 |
+
"learning_rate": 1.923715657464742e-10,
|
| 1486 |
+
"loss": 3.5299,
|
| 1487 |
+
"step": 52750
|
| 1488 |
+
}
|
| 1489 |
+
],
|
| 1490 |
+
"logging_steps": 250,
|
| 1491 |
+
"max_steps": 52896,
|
| 1492 |
+
"num_input_tokens_seen": 0,
|
| 1493 |
+
"num_train_epochs": 2,
|
| 1494 |
+
"save_steps": 250,
|
| 1495 |
+
"stateful_callbacks": {
|
| 1496 |
+
"TrainerControl": {
|
| 1497 |
+
"args": {
|
| 1498 |
+
"should_epoch_stop": false,
|
| 1499 |
+
"should_evaluate": false,
|
| 1500 |
+
"should_log": false,
|
| 1501 |
+
"should_save": true,
|
| 1502 |
+
"should_training_stop": true
|
| 1503 |
+
},
|
| 1504 |
+
"attributes": {}
|
| 1505 |
+
}
|
| 1506 |
+
},
|
| 1507 |
+
"total_flos": 2.8374993889748353e+19,
|
| 1508 |
+
"train_batch_size": 1,
|
| 1509 |
+
"trial_name": null,
|
| 1510 |
+
"trial_params": null
|
| 1511 |
+
}
|