Upload Qwen2.5-32B-Instruct_finance_all_resp_cosine_with_restarts_30_7e-05_trainer_state.json with huggingface_hub
Browse files
Qwen2.5-32B-Instruct_finance_all_resp_cosine_with_restarts_30_7e-05_trainer_state.json
ADDED
|
@@ -0,0 +1,2367 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"global_step": 332,
|
| 4 |
+
"max_steps": 332,
|
| 5 |
+
"logging_steps": 1,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"save_steps": 10,
|
| 8 |
+
"train_batch_size": 2,
|
| 9 |
+
"num_train_epochs": 1,
|
| 10 |
+
"num_input_tokens_seen": 0,
|
| 11 |
+
"total_flos": 1.0339169907188122e+17,
|
| 12 |
+
"log_history": [
|
| 13 |
+
{
|
| 14 |
+
"loss": 4.7036,
|
| 15 |
+
"grad_norm": 34.68614959716797,
|
| 16 |
+
"learning_rate": 0.0,
|
| 17 |
+
"epoch": 0.0030131826741996233,
|
| 18 |
+
"step": 1
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"loss": 4.4381,
|
| 22 |
+
"grad_norm": 33.216060638427734,
|
| 23 |
+
"learning_rate": 2.333333333333333e-06,
|
| 24 |
+
"epoch": 0.006026365348399247,
|
| 25 |
+
"step": 2
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"loss": 4.246,
|
| 29 |
+
"grad_norm": 32.37959671020508,
|
| 30 |
+
"learning_rate": 4.666666666666666e-06,
|
| 31 |
+
"epoch": 0.00903954802259887,
|
| 32 |
+
"step": 3
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"loss": 4.2875,
|
| 36 |
+
"grad_norm": 37.732025146484375,
|
| 37 |
+
"learning_rate": 7e-06,
|
| 38 |
+
"epoch": 0.012052730696798493,
|
| 39 |
+
"step": 4
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"loss": 3.5227,
|
| 43 |
+
"grad_norm": 25.07876205444336,
|
| 44 |
+
"learning_rate": 9.333333333333333e-06,
|
| 45 |
+
"epoch": 0.015065913370998116,
|
| 46 |
+
"step": 5
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"loss": 3.2751,
|
| 50 |
+
"grad_norm": 12.369613647460938,
|
| 51 |
+
"learning_rate": 1.1666666666666665e-05,
|
| 52 |
+
"epoch": 0.01807909604519774,
|
| 53 |
+
"step": 6
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"loss": 2.7026,
|
| 57 |
+
"grad_norm": 8.583898544311523,
|
| 58 |
+
"learning_rate": 1.4e-05,
|
| 59 |
+
"epoch": 0.021092278719397364,
|
| 60 |
+
"step": 7
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"loss": 2.4307,
|
| 64 |
+
"grad_norm": 6.517887115478516,
|
| 65 |
+
"learning_rate": 1.633333333333333e-05,
|
| 66 |
+
"epoch": 0.024105461393596987,
|
| 67 |
+
"step": 8
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"loss": 2.4248,
|
| 71 |
+
"grad_norm": 5.936885356903076,
|
| 72 |
+
"learning_rate": 1.8666666666666665e-05,
|
| 73 |
+
"epoch": 0.02711864406779661,
|
| 74 |
+
"step": 9
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"loss": 2.1531,
|
| 78 |
+
"grad_norm": 4.3616814613342285,
|
| 79 |
+
"learning_rate": 2.1e-05,
|
| 80 |
+
"epoch": 0.030131826741996232,
|
| 81 |
+
"step": 10
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"loss": 2.0193,
|
| 85 |
+
"grad_norm": 4.396628379821777,
|
| 86 |
+
"learning_rate": 2.333333333333333e-05,
|
| 87 |
+
"epoch": 0.03314500941619586,
|
| 88 |
+
"step": 11
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"loss": 1.9573,
|
| 92 |
+
"grad_norm": 4.6346588134765625,
|
| 93 |
+
"learning_rate": 2.5666666666666663e-05,
|
| 94 |
+
"epoch": 0.03615819209039548,
|
| 95 |
+
"step": 12
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"loss": 1.8249,
|
| 99 |
+
"grad_norm": 4.783262729644775,
|
| 100 |
+
"learning_rate": 2.8e-05,
|
| 101 |
+
"epoch": 0.039171374764595104,
|
| 102 |
+
"step": 13
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"loss": 1.871,
|
| 106 |
+
"grad_norm": 6.585052013397217,
|
| 107 |
+
"learning_rate": 3.0333333333333333e-05,
|
| 108 |
+
"epoch": 0.04218455743879473,
|
| 109 |
+
"step": 14
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"loss": 1.5362,
|
| 113 |
+
"grad_norm": 3.52156662940979,
|
| 114 |
+
"learning_rate": 3.266666666666666e-05,
|
| 115 |
+
"epoch": 0.04519774011299435,
|
| 116 |
+
"step": 15
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"loss": 1.5977,
|
| 120 |
+
"grad_norm": 3.7720725536346436,
|
| 121 |
+
"learning_rate": 3.5e-05,
|
| 122 |
+
"epoch": 0.04821092278719397,
|
| 123 |
+
"step": 16
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"loss": 1.7891,
|
| 127 |
+
"grad_norm": 4.995004177093506,
|
| 128 |
+
"learning_rate": 3.733333333333333e-05,
|
| 129 |
+
"epoch": 0.051224105461393596,
|
| 130 |
+
"step": 17
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"loss": 1.6058,
|
| 134 |
+
"grad_norm": 2.9881644248962402,
|
| 135 |
+
"learning_rate": 3.9666666666666664e-05,
|
| 136 |
+
"epoch": 0.05423728813559322,
|
| 137 |
+
"step": 18
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"loss": 1.634,
|
| 141 |
+
"grad_norm": 3.3046276569366455,
|
| 142 |
+
"learning_rate": 4.2e-05,
|
| 143 |
+
"epoch": 0.05725047080979284,
|
| 144 |
+
"step": 19
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"loss": 1.5182,
|
| 148 |
+
"grad_norm": 3.3802030086517334,
|
| 149 |
+
"learning_rate": 4.4333333333333324e-05,
|
| 150 |
+
"epoch": 0.060263653483992465,
|
| 151 |
+
"step": 20
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"loss": 1.6227,
|
| 155 |
+
"grad_norm": 4.024594306945801,
|
| 156 |
+
"learning_rate": 4.666666666666666e-05,
|
| 157 |
+
"epoch": 0.06327683615819209,
|
| 158 |
+
"step": 21
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"loss": 1.5663,
|
| 162 |
+
"grad_norm": 2.86409592628479,
|
| 163 |
+
"learning_rate": 4.899999999999999e-05,
|
| 164 |
+
"epoch": 0.06629001883239172,
|
| 165 |
+
"step": 22
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"loss": 1.4878,
|
| 169 |
+
"grad_norm": 2.915001392364502,
|
| 170 |
+
"learning_rate": 5.1333333333333325e-05,
|
| 171 |
+
"epoch": 0.06930320150659133,
|
| 172 |
+
"step": 23
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"loss": 1.5324,
|
| 176 |
+
"grad_norm": 2.6923105716705322,
|
| 177 |
+
"learning_rate": 5.3666666666666666e-05,
|
| 178 |
+
"epoch": 0.07231638418079096,
|
| 179 |
+
"step": 24
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"loss": 1.5547,
|
| 183 |
+
"grad_norm": 2.824495553970337,
|
| 184 |
+
"learning_rate": 5.6e-05,
|
| 185 |
+
"epoch": 0.07532956685499058,
|
| 186 |
+
"step": 25
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"loss": 1.5096,
|
| 190 |
+
"grad_norm": 3.191667318344116,
|
| 191 |
+
"learning_rate": 5.833333333333333e-05,
|
| 192 |
+
"epoch": 0.07834274952919021,
|
| 193 |
+
"step": 26
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"loss": 1.57,
|
| 197 |
+
"grad_norm": 2.7785418033599854,
|
| 198 |
+
"learning_rate": 6.0666666666666666e-05,
|
| 199 |
+
"epoch": 0.08135593220338982,
|
| 200 |
+
"step": 27
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"loss": 1.4546,
|
| 204 |
+
"grad_norm": 2.0098705291748047,
|
| 205 |
+
"learning_rate": 6.3e-05,
|
| 206 |
+
"epoch": 0.08436911487758945,
|
| 207 |
+
"step": 28
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"loss": 1.4655,
|
| 211 |
+
"grad_norm": 1.8760172128677368,
|
| 212 |
+
"learning_rate": 6.533333333333333e-05,
|
| 213 |
+
"epoch": 0.08738229755178907,
|
| 214 |
+
"step": 29
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"loss": 1.3572,
|
| 218 |
+
"grad_norm": 2.0383059978485107,
|
| 219 |
+
"learning_rate": 6.766666666666667e-05,
|
| 220 |
+
"epoch": 0.0903954802259887,
|
| 221 |
+
"step": 30
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"loss": 1.473,
|
| 225 |
+
"grad_norm": 1.9418625831604004,
|
| 226 |
+
"learning_rate": 7e-05,
|
| 227 |
+
"epoch": 0.09340866290018833,
|
| 228 |
+
"step": 31
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"loss": 1.4612,
|
| 232 |
+
"grad_norm": 2.365391731262207,
|
| 233 |
+
"learning_rate": 6.996970428356058e-05,
|
| 234 |
+
"epoch": 0.09642184557438795,
|
| 235 |
+
"step": 32
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"loss": 1.5633,
|
| 239 |
+
"grad_norm": 2.0207245349884033,
|
| 240 |
+
"learning_rate": 6.987886958169575e-05,
|
| 241 |
+
"epoch": 0.09943502824858758,
|
| 242 |
+
"step": 33
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"loss": 1.3985,
|
| 246 |
+
"grad_norm": 2.1693599224090576,
|
| 247 |
+
"learning_rate": 6.972765314596951e-05,
|
| 248 |
+
"epoch": 0.10244821092278719,
|
| 249 |
+
"step": 34
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"loss": 1.4067,
|
| 253 |
+
"grad_norm": 1.9948197603225708,
|
| 254 |
+
"learning_rate": 6.951631675982521e-05,
|
| 255 |
+
"epoch": 0.10546139359698682,
|
| 256 |
+
"step": 35
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"loss": 1.4157,
|
| 260 |
+
"grad_norm": 1.9751430749893188,
|
| 261 |
+
"learning_rate": 6.924522628539012e-05,
|
| 262 |
+
"epoch": 0.10847457627118644,
|
| 263 |
+
"step": 36
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"loss": 1.5028,
|
| 267 |
+
"grad_norm": 1.7820850610733032,
|
| 268 |
+
"learning_rate": 6.891485103010098e-05,
|
| 269 |
+
"epoch": 0.11148775894538607,
|
| 270 |
+
"step": 37
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"loss": 1.3704,
|
| 274 |
+
"grad_norm": 1.9648020267486572,
|
| 275 |
+
"learning_rate": 6.852576293424654e-05,
|
| 276 |
+
"epoch": 0.11450094161958568,
|
| 277 |
+
"step": 38
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"loss": 1.5124,
|
| 281 |
+
"grad_norm": 2.1613192558288574,
|
| 282 |
+
"learning_rate": 6.807863558083377e-05,
|
| 283 |
+
"epoch": 0.11751412429378531,
|
| 284 |
+
"step": 39
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"loss": 1.5235,
|
| 288 |
+
"grad_norm": 1.9539159536361694,
|
| 289 |
+
"learning_rate": 6.757424302949186e-05,
|
| 290 |
+
"epoch": 0.12052730696798493,
|
| 291 |
+
"step": 40
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"loss": 1.4433,
|
| 295 |
+
"grad_norm": 2.2853245735168457,
|
| 296 |
+
"learning_rate": 6.701345847643282e-05,
|
| 297 |
+
"epoch": 0.12354048964218456,
|
| 298 |
+
"step": 41
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"loss": 1.4077,
|
| 302 |
+
"grad_norm": 1.959349274635315,
|
| 303 |
+
"learning_rate": 6.639725274278824e-05,
|
| 304 |
+
"epoch": 0.12655367231638417,
|
| 305 |
+
"step": 42
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"loss": 1.4198,
|
| 309 |
+
"grad_norm": 1.9996960163116455,
|
| 310 |
+
"learning_rate": 6.572669259393953e-05,
|
| 311 |
+
"epoch": 0.1295668549905838,
|
| 312 |
+
"step": 43
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"loss": 1.3582,
|
| 316 |
+
"grad_norm": 1.7032471895217896,
|
| 317 |
+
"learning_rate": 6.5002938892751e-05,
|
| 318 |
+
"epoch": 0.13258003766478343,
|
| 319 |
+
"step": 44
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"loss": 1.4175,
|
| 323 |
+
"grad_norm": 1.769765853881836,
|
| 324 |
+
"learning_rate": 6.422724458990284e-05,
|
| 325 |
+
"epoch": 0.13559322033898305,
|
| 326 |
+
"step": 45
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"loss": 1.4727,
|
| 330 |
+
"grad_norm": 1.9631857872009277,
|
| 331 |
+
"learning_rate": 6.340095255480317e-05,
|
| 332 |
+
"epoch": 0.13860640301318267,
|
| 333 |
+
"step": 46
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"loss": 1.3882,
|
| 337 |
+
"grad_norm": 1.7352495193481445,
|
| 338 |
+
"learning_rate": 6.252549325083437e-05,
|
| 339 |
+
"epoch": 0.1416195856873823,
|
| 340 |
+
"step": 47
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"loss": 1.3526,
|
| 344 |
+
"grad_norm": 1.8017715215682983,
|
| 345 |
+
"learning_rate": 6.160238225895803e-05,
|
| 346 |
+
"epoch": 0.14463276836158193,
|
| 347 |
+
"step": 48
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"loss": 1.5247,
|
| 351 |
+
"grad_norm": 1.8248356580734253,
|
| 352 |
+
"learning_rate": 6.063321765396568e-05,
|
| 353 |
+
"epoch": 0.14764595103578154,
|
| 354 |
+
"step": 49
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"loss": 1.439,
|
| 358 |
+
"grad_norm": 1.8083879947662354,
|
| 359 |
+
"learning_rate": 5.9619677237917655e-05,
|
| 360 |
+
"epoch": 0.15065913370998116,
|
| 361 |
+
"step": 50
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"loss": 1.399,
|
| 365 |
+
"grad_norm": 1.8656060695648193,
|
| 366 |
+
"learning_rate": 5.8563515635559344e-05,
|
| 367 |
+
"epoch": 0.1536723163841808,
|
| 368 |
+
"step": 51
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"loss": 1.4051,
|
| 372 |
+
"grad_norm": 1.6660946607589722,
|
| 373 |
+
"learning_rate": 5.746656125674329e-05,
|
| 374 |
+
"epoch": 0.15668549905838042,
|
| 375 |
+
"step": 52
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"loss": 1.4817,
|
| 379 |
+
"grad_norm": 1.7004179954528809,
|
| 380 |
+
"learning_rate": 5.633071313111564e-05,
|
| 381 |
+
"epoch": 0.15969868173258003,
|
| 382 |
+
"step": 53
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"loss": 1.4065,
|
| 386 |
+
"grad_norm": 1.521797776222229,
|
| 387 |
+
"learning_rate": 5.5157937620546804e-05,
|
| 388 |
+
"epoch": 0.16271186440677965,
|
| 389 |
+
"step": 54
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"loss": 1.3753,
|
| 393 |
+
"grad_norm": 1.5257264375686646,
|
| 394 |
+
"learning_rate": 5.3950265014997655e-05,
|
| 395 |
+
"epoch": 0.1657250470809793,
|
| 396 |
+
"step": 55
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"loss": 1.3395,
|
| 400 |
+
"grad_norm": 1.5974771976470947,
|
| 401 |
+
"learning_rate": 5.270978601771444e-05,
|
| 402 |
+
"epoch": 0.1687382297551789,
|
| 403 |
+
"step": 56
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"loss": 1.435,
|
| 407 |
+
"grad_norm": 1.6465458869934082,
|
| 408 |
+
"learning_rate": 5.1438648125837204e-05,
|
| 409 |
+
"epoch": 0.17175141242937852,
|
| 410 |
+
"step": 57
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"loss": 1.3862,
|
| 414 |
+
"grad_norm": 1.536580204963684,
|
| 415 |
+
"learning_rate": 5.013905191268754e-05,
|
| 416 |
+
"epoch": 0.17476459510357814,
|
| 417 |
+
"step": 58
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"loss": 1.4123,
|
| 421 |
+
"grad_norm": 1.5449066162109375,
|
| 422 |
+
"learning_rate": 4.881324721817167e-05,
|
| 423 |
+
"epoch": 0.17777777777777778,
|
| 424 |
+
"step": 59
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"loss": 1.376,
|
| 428 |
+
"grad_norm": 1.7106742858886719,
|
| 429 |
+
"learning_rate": 4.746352925389414e-05,
|
| 430 |
+
"epoch": 0.1807909604519774,
|
| 431 |
+
"step": 60
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"loss": 1.3616,
|
| 435 |
+
"grad_norm": 1.6424928903579712,
|
| 436 |
+
"learning_rate": 4.60922346297246e-05,
|
| 437 |
+
"epoch": 0.18380414312617702,
|
| 438 |
+
"step": 61
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"loss": 1.5606,
|
| 442 |
+
"grad_norm": 1.7592521905899048,
|
| 443 |
+
"learning_rate": 4.4701737308696665e-05,
|
| 444 |
+
"epoch": 0.18681732580037666,
|
| 445 |
+
"step": 62
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"loss": 1.416,
|
| 449 |
+
"grad_norm": 1.5742639303207397,
|
| 450 |
+
"learning_rate": 4.3294444497241656e-05,
|
| 451 |
+
"epoch": 0.18983050847457628,
|
| 452 |
+
"step": 63
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"loss": 1.3684,
|
| 456 |
+
"grad_norm": 1.5274724960327148,
|
| 457 |
+
"learning_rate": 4.187279247787175e-05,
|
| 458 |
+
"epoch": 0.1928436911487759,
|
| 459 |
+
"step": 64
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"loss": 1.326,
|
| 463 |
+
"grad_norm": 1.6325310468673706,
|
| 464 |
+
"learning_rate": 4.0439242391527176e-05,
|
| 465 |
+
"epoch": 0.1958568738229755,
|
| 466 |
+
"step": 65
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"loss": 1.3414,
|
| 470 |
+
"grad_norm": 1.5987493991851807,
|
| 471 |
+
"learning_rate": 3.899627597688895e-05,
|
| 472 |
+
"epoch": 0.19887005649717515,
|
| 473 |
+
"step": 66
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"loss": 1.2961,
|
| 477 |
+
"grad_norm": 1.4861475229263306,
|
| 478 |
+
"learning_rate": 3.754639127403304e-05,
|
| 479 |
+
"epoch": 0.20188323917137477,
|
| 480 |
+
"step": 67
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"loss": 1.3731,
|
| 484 |
+
"grad_norm": 1.4820605516433716,
|
| 485 |
+
"learning_rate": 3.609209829986387e-05,
|
| 486 |
+
"epoch": 0.20489642184557438,
|
| 487 |
+
"step": 68
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"loss": 1.4947,
|
| 491 |
+
"grad_norm": 1.6821802854537964,
|
| 492 |
+
"learning_rate": 3.463591470281375e-05,
|
| 493 |
+
"epoch": 0.207909604519774,
|
| 494 |
+
"step": 69
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"loss": 1.4524,
|
| 498 |
+
"grad_norm": 1.4626668691635132,
|
| 499 |
+
"learning_rate": 3.318036140433069e-05,
|
| 500 |
+
"epoch": 0.21092278719397364,
|
| 501 |
+
"step": 70
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"loss": 1.4252,
|
| 505 |
+
"grad_norm": 1.5508460998535156,
|
| 506 |
+
"learning_rate": 3.172795823469999e-05,
|
| 507 |
+
"epoch": 0.21393596986817326,
|
| 508 |
+
"step": 71
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"loss": 1.3371,
|
| 512 |
+
"grad_norm": 1.5290253162384033,
|
| 513 |
+
"learning_rate": 3.0281219570754987e-05,
|
| 514 |
+
"epoch": 0.21694915254237288,
|
| 515 |
+
"step": 72
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"loss": 1.4155,
|
| 519 |
+
"grad_norm": 1.448610782623291,
|
| 520 |
+
"learning_rate": 2.8842649983028517e-05,
|
| 521 |
+
"epoch": 0.2199623352165725,
|
| 522 |
+
"step": 73
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"loss": 1.4835,
|
| 526 |
+
"grad_norm": 1.6528280973434448,
|
| 527 |
+
"learning_rate": 2.7414739899881024e-05,
|
| 528 |
+
"epoch": 0.22297551789077213,
|
| 529 |
+
"step": 74
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"loss": 1.3142,
|
| 533 |
+
"grad_norm": 1.437882900238037,
|
| 534 |
+
"learning_rate": 2.5999961296111394e-05,
|
| 535 |
+
"epoch": 0.22598870056497175,
|
| 536 |
+
"step": 75
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"loss": 1.2955,
|
| 540 |
+
"grad_norm": 1.3636724948883057,
|
| 541 |
+
"learning_rate": 2.4600763413514138e-05,
|
| 542 |
+
"epoch": 0.22900188323917137,
|
| 543 |
+
"step": 76
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"loss": 1.2569,
|
| 547 |
+
"grad_norm": 1.338307499885559,
|
| 548 |
+
"learning_rate": 2.3219568520791783e-05,
|
| 549 |
+
"epoch": 0.232015065913371,
|
| 550 |
+
"step": 77
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"loss": 1.3456,
|
| 554 |
+
"grad_norm": 1.3924059867858887,
|
| 555 |
+
"learning_rate": 2.1858767720162456e-05,
|
| 556 |
+
"epoch": 0.23502824858757063,
|
| 557 |
+
"step": 78
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"loss": 1.3361,
|
| 561 |
+
"grad_norm": 1.4208087921142578,
|
| 562 |
+
"learning_rate": 2.0520716807922514e-05,
|
| 563 |
+
"epoch": 0.23804143126177024,
|
| 564 |
+
"step": 79
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"loss": 1.3467,
|
| 568 |
+
"grad_norm": 1.4785457849502563,
|
| 569 |
+
"learning_rate": 1.92077321961302e-05,
|
| 570 |
+
"epoch": 0.24105461393596986,
|
| 571 |
+
"step": 80
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"loss": 1.4191,
|
| 575 |
+
"grad_norm": 1.49100661277771,
|
| 576 |
+
"learning_rate": 1.7922086902470502e-05,
|
| 577 |
+
"epoch": 0.2440677966101695,
|
| 578 |
+
"step": 81
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"loss": 1.4158,
|
| 582 |
+
"grad_norm": 1.5821126699447632,
|
| 583 |
+
"learning_rate": 1.666600661524396e-05,
|
| 584 |
+
"epoch": 0.24708097928436912,
|
| 585 |
+
"step": 82
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"loss": 1.3125,
|
| 589 |
+
"grad_norm": 1.3877836465835571,
|
| 590 |
+
"learning_rate": 1.5441665840290927e-05,
|
| 591 |
+
"epoch": 0.25009416195856876,
|
| 592 |
+
"step": 83
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"loss": 1.3836,
|
| 596 |
+
"grad_norm": 1.5954185724258423,
|
| 597 |
+
"learning_rate": 1.4251184136522453e-05,
|
| 598 |
+
"epoch": 0.25310734463276835,
|
| 599 |
+
"step": 84
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"loss": 1.3397,
|
| 603 |
+
"grad_norm": 1.4120008945465088,
|
| 604 |
+
"learning_rate": 1.3096622446574184e-05,
|
| 605 |
+
"epoch": 0.256120527306968,
|
| 606 |
+
"step": 85
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"loss": 1.3302,
|
| 610 |
+
"grad_norm": 1.4432895183563232,
|
| 611 |
+
"learning_rate": 1.1979979528935865e-05,
|
| 612 |
+
"epoch": 0.2591337099811676,
|
| 613 |
+
"step": 86
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"loss": 1.3017,
|
| 617 |
+
"grad_norm": 1.3301897048950195,
|
| 618 |
+
"learning_rate": 1.0903188497732999e-05,
|
| 619 |
+
"epoch": 0.2621468926553672,
|
| 620 |
+
"step": 87
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"loss": 1.366,
|
| 624 |
+
"grad_norm": 1.5367295742034912,
|
| 625 |
+
"learning_rate": 9.86811347615111e-06,
|
| 626 |
+
"epoch": 0.26516007532956687,
|
| 627 |
+
"step": 88
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"loss": 1.3423,
|
| 631 |
+
"grad_norm": 1.4702235460281372,
|
| 632 |
+
"learning_rate": 8.876546369295717e-06,
|
| 633 |
+
"epoch": 0.26817325800376646,
|
| 634 |
+
"step": 89
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"loss": 1.4105,
|
| 638 |
+
"grad_norm": 1.5141451358795166,
|
| 639 |
+
"learning_rate": 7.930203762075426e-06,
|
| 640 |
+
"epoch": 0.2711864406779661,
|
| 641 |
+
"step": 90
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"loss": 1.3936,
|
| 645 |
+
"grad_norm": 1.5650383234024048,
|
| 646 |
+
"learning_rate": 7.030723947477803e-06,
|
| 647 |
+
"epoch": 0.27419962335216574,
|
| 648 |
+
"step": 91
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"loss": 1.2605,
|
| 652 |
+
"grad_norm": 1.2680590152740479,
|
| 653 |
+
"learning_rate": 6.1796640903832005e-06,
|
| 654 |
+
"epoch": 0.27721280602636533,
|
| 655 |
+
"step": 92
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"loss": 1.3184,
|
| 659 |
+
"grad_norm": 1.4249001741409302,
|
| 660 |
+
"learning_rate": 5.378497531826109e-06,
|
| 661 |
+
"epoch": 0.280225988700565,
|
| 662 |
+
"step": 93
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"loss": 1.3123,
|
| 666 |
+
"grad_norm": 1.509170651435852,
|
| 667 |
+
"learning_rate": 4.6286112383710245e-06,
|
| 668 |
+
"epoch": 0.2832391713747646,
|
| 669 |
+
"step": 94
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"loss": 1.3128,
|
| 673 |
+
"grad_norm": 1.379465103149414,
|
| 674 |
+
"learning_rate": 3.931303401018427e-06,
|
| 675 |
+
"epoch": 0.2862523540489642,
|
| 676 |
+
"step": 95
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"loss": 1.4236,
|
| 680 |
+
"grad_norm": 1.3842952251434326,
|
| 681 |
+
"learning_rate": 3.2877811877975422e-06,
|
| 682 |
+
"epoch": 0.28926553672316385,
|
| 683 |
+
"step": 96
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"loss": 1.3646,
|
| 687 |
+
"grad_norm": 1.5518581867218018,
|
| 688 |
+
"learning_rate": 2.6991586539366018e-06,
|
| 689 |
+
"epoch": 0.29227871939736344,
|
| 690 |
+
"step": 97
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"loss": 1.3507,
|
| 694 |
+
"grad_norm": 1.4329756498336792,
|
| 695 |
+
"learning_rate": 2.166454813228505e-06,
|
| 696 |
+
"epoch": 0.2952919020715631,
|
| 697 |
+
"step": 98
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"loss": 1.2478,
|
| 701 |
+
"grad_norm": 1.4545249938964844,
|
| 702 |
+
"learning_rate": 1.6905918739306423e-06,
|
| 703 |
+
"epoch": 0.2983050847457627,
|
| 704 |
+
"step": 99
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"loss": 1.2338,
|
| 708 |
+
"grad_norm": 1.5154703855514526,
|
| 709 |
+
"learning_rate": 1.2723936422528996e-06,
|
| 710 |
+
"epoch": 0.3013182674199623,
|
| 711 |
+
"step": 100
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"loss": 1.4895,
|
| 715 |
+
"grad_norm": 1.5542984008789062,
|
| 716 |
+
"learning_rate": 9.125840961976982e-07,
|
| 717 |
+
"epoch": 0.30433145009416196,
|
| 718 |
+
"step": 101
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"loss": 1.4022,
|
| 722 |
+
"grad_norm": 1.3746228218078613,
|
| 723 |
+
"learning_rate": 6.117861322210028e-07,
|
| 724 |
+
"epoch": 0.3073446327683616,
|
| 725 |
+
"step": 102
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"loss": 1.3899,
|
| 729 |
+
"grad_norm": 1.5167266130447388,
|
| 730 |
+
"learning_rate": 3.705204868840894e-07,
|
| 731 |
+
"epoch": 0.3103578154425612,
|
| 732 |
+
"step": 103
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"loss": 1.3286,
|
| 736 |
+
"grad_norm": 1.4172340631484985,
|
| 737 |
+
"learning_rate": 1.892048353628217e-07,
|
| 738 |
+
"epoch": 0.31337099811676083,
|
| 739 |
+
"step": 104
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"loss": 1.1915,
|
| 743 |
+
"grad_norm": 1.5105013847351074,
|
| 744 |
+
"learning_rate": 6.81530683751752e-08,
|
| 745 |
+
"epoch": 0.3163841807909605,
|
| 746 |
+
"step": 105
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"loss": 1.3035,
|
| 750 |
+
"grad_norm": 1.2953332662582397,
|
| 751 |
+
"learning_rate": 7.574748778697015e-09,
|
| 752 |
+
"epoch": 0.31939736346516007,
|
| 753 |
+
"step": 106
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"loss": 1.2276,
|
| 757 |
+
"grad_norm": 1.3927589654922485,
|
| 758 |
+
"learning_rate": 6.99924252512213e-05,
|
| 759 |
+
"epoch": 0.3224105461393597,
|
| 760 |
+
"step": 107
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"loss": 1.3186,
|
| 764 |
+
"grad_norm": 1.6460096836090088,
|
| 765 |
+
"learning_rate": 6.993184693162481e-05,
|
| 766 |
+
"epoch": 0.3254237288135593,
|
| 767 |
+
"step": 108
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"loss": 1.328,
|
| 771 |
+
"grad_norm": 1.4727671146392822,
|
| 772 |
+
"learning_rate": 6.981079516463718e-05,
|
| 773 |
+
"epoch": 0.32843691148775894,
|
| 774 |
+
"step": 109
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"loss": 1.4161,
|
| 778 |
+
"grad_norm": 1.8564578294754028,
|
| 779 |
+
"learning_rate": 6.962947951311591e-05,
|
| 780 |
+
"epoch": 0.3314500941619586,
|
| 781 |
+
"step": 110
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"loss": 1.3738,
|
| 785 |
+
"grad_norm": 1.7267475128173828,
|
| 786 |
+
"learning_rate": 6.938821386777899e-05,
|
| 787 |
+
"epoch": 0.3344632768361582,
|
| 788 |
+
"step": 111
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"loss": 1.3619,
|
| 792 |
+
"grad_norm": 1.7725125551223755,
|
| 793 |
+
"learning_rate": 6.908741590380231e-05,
|
| 794 |
+
"epoch": 0.3374764595103578,
|
| 795 |
+
"step": 112
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"loss": 1.4138,
|
| 799 |
+
"grad_norm": 1.6039916276931763,
|
| 800 |
+
"learning_rate": 6.87276063577471e-05,
|
| 801 |
+
"epoch": 0.34048964218455746,
|
| 802 |
+
"step": 113
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"loss": 1.3492,
|
| 806 |
+
"grad_norm": 1.4494953155517578,
|
| 807 |
+
"learning_rate": 6.830940812606936e-05,
|
| 808 |
+
"epoch": 0.34350282485875705,
|
| 809 |
+
"step": 114
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"loss": 1.4501,
|
| 813 |
+
"grad_norm": 1.4761908054351807,
|
| 814 |
+
"learning_rate": 6.78335451867715e-05,
|
| 815 |
+
"epoch": 0.3465160075329567,
|
| 816 |
+
"step": 115
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"loss": 1.3806,
|
| 820 |
+
"grad_norm": 1.499022126197815,
|
| 821 |
+
"learning_rate": 6.73008413460634e-05,
|
| 822 |
+
"epoch": 0.3495291902071563,
|
| 823 |
+
"step": 116
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"loss": 1.4476,
|
| 827 |
+
"grad_norm": 1.5290741920471191,
|
| 828 |
+
"learning_rate": 6.671221881220246e-05,
|
| 829 |
+
"epoch": 0.3525423728813559,
|
| 830 |
+
"step": 117
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"loss": 1.3522,
|
| 834 |
+
"grad_norm": 1.4761743545532227,
|
| 835 |
+
"learning_rate": 6.606869659898157e-05,
|
| 836 |
+
"epoch": 0.35555555555555557,
|
| 837 |
+
"step": 118
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"loss": 1.4608,
|
| 841 |
+
"grad_norm": 1.551421880722046,
|
| 842 |
+
"learning_rate": 6.537138876162898e-05,
|
| 843 |
+
"epoch": 0.35856873822975516,
|
| 844 |
+
"step": 119
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"loss": 1.2385,
|
| 848 |
+
"grad_norm": 1.2959685325622559,
|
| 849 |
+
"learning_rate": 6.462150246817388e-05,
|
| 850 |
+
"epoch": 0.3615819209039548,
|
| 851 |
+
"step": 120
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"loss": 1.5111,
|
| 855 |
+
"grad_norm": 1.3990561962127686,
|
| 856 |
+
"learning_rate": 6.38203359096168e-05,
|
| 857 |
+
"epoch": 0.36459510357815444,
|
| 858 |
+
"step": 121
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"loss": 1.4017,
|
| 862 |
+
"grad_norm": 1.3039060831069946,
|
| 863 |
+
"learning_rate": 6.296927605252219e-05,
|
| 864 |
+
"epoch": 0.36760828625235403,
|
| 865 |
+
"step": 122
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"loss": 1.4041,
|
| 869 |
+
"grad_norm": 1.4338008165359497,
|
| 870 |
+
"learning_rate": 6.206979623792457e-05,
|
| 871 |
+
"epoch": 0.3706214689265537,
|
| 872 |
+
"step": 123
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"loss": 1.4613,
|
| 876 |
+
"grad_norm": 1.4520732164382935,
|
| 877 |
+
"learning_rate": 6.112345363070427e-05,
|
| 878 |
+
"epoch": 0.3736346516007533,
|
| 879 |
+
"step": 124
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"loss": 1.3588,
|
| 883 |
+
"grad_norm": 1.4582487344741821,
|
| 884 |
+
"learning_rate": 6.01318865238489e-05,
|
| 885 |
+
"epoch": 0.3766478342749529,
|
| 886 |
+
"step": 125
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"loss": 1.3968,
|
| 890 |
+
"grad_norm": 1.3768619298934937,
|
| 891 |
+
"learning_rate": 5.9096811502266993e-05,
|
| 892 |
+
"epoch": 0.37966101694915255,
|
| 893 |
+
"step": 126
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"loss": 1.4076,
|
| 897 |
+
"grad_norm": 1.3212107419967651,
|
| 898 |
+
"learning_rate": 5.802002047106413e-05,
|
| 899 |
+
"epoch": 0.38267419962335214,
|
| 900 |
+
"step": 127
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"loss": 1.3936,
|
| 904 |
+
"grad_norm": 1.6659901142120361,
|
| 905 |
+
"learning_rate": 5.690337755342582e-05,
|
| 906 |
+
"epoch": 0.3856873822975518,
|
| 907 |
+
"step": 128
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"loss": 1.246,
|
| 911 |
+
"grad_norm": 1.3914105892181396,
|
| 912 |
+
"learning_rate": 5.574881586347755e-05,
|
| 913 |
+
"epoch": 0.3887005649717514,
|
| 914 |
+
"step": 129
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"loss": 1.4966,
|
| 918 |
+
"grad_norm": 1.4999268054962158,
|
| 919 |
+
"learning_rate": 5.455833415970907e-05,
|
| 920 |
+
"epoch": 0.391713747645951,
|
| 921 |
+
"step": 130
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"loss": 1.4092,
|
| 925 |
+
"grad_norm": 1.4403806924819946,
|
| 926 |
+
"learning_rate": 5.333399338475603e-05,
|
| 927 |
+
"epoch": 0.39472693032015066,
|
| 928 |
+
"step": 131
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"loss": 1.3959,
|
| 932 |
+
"grad_norm": 1.4177464246749878,
|
| 933 |
+
"learning_rate": 5.207791309752948e-05,
|
| 934 |
+
"epoch": 0.3977401129943503,
|
| 935 |
+
"step": 132
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"loss": 1.3596,
|
| 939 |
+
"grad_norm": 1.447963833808899,
|
| 940 |
+
"learning_rate": 5.07922678038698e-05,
|
| 941 |
+
"epoch": 0.4007532956685499,
|
| 942 |
+
"step": 133
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"loss": 1.3084,
|
| 946 |
+
"grad_norm": 1.5529978275299072,
|
| 947 |
+
"learning_rate": 4.9479283192077466e-05,
|
| 948 |
+
"epoch": 0.40376647834274954,
|
| 949 |
+
"step": 134
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"loss": 1.3398,
|
| 953 |
+
"grad_norm": 1.396165370941162,
|
| 954 |
+
"learning_rate": 4.8141232279837554e-05,
|
| 955 |
+
"epoch": 0.4067796610169492,
|
| 956 |
+
"step": 135
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"loss": 1.3571,
|
| 960 |
+
"grad_norm": 1.5499067306518555,
|
| 961 |
+
"learning_rate": 4.678043147920822e-05,
|
| 962 |
+
"epoch": 0.40979284369114877,
|
| 963 |
+
"step": 136
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"loss": 1.4113,
|
| 967 |
+
"grad_norm": 1.4300554990768433,
|
| 968 |
+
"learning_rate": 4.5399236586485846e-05,
|
| 969 |
+
"epoch": 0.4128060263653484,
|
| 970 |
+
"step": 137
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"loss": 1.2354,
|
| 974 |
+
"grad_norm": 1.48812997341156,
|
| 975 |
+
"learning_rate": 4.400003870388862e-05,
|
| 976 |
+
"epoch": 0.415819209039548,
|
| 977 |
+
"step": 138
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"loss": 1.3698,
|
| 981 |
+
"grad_norm": 1.3681354522705078,
|
| 982 |
+
"learning_rate": 4.2585260100118973e-05,
|
| 983 |
+
"epoch": 0.41883239171374764,
|
| 984 |
+
"step": 139
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"loss": 1.3216,
|
| 988 |
+
"grad_norm": 1.3714697360992432,
|
| 989 |
+
"learning_rate": 4.115735001697148e-05,
|
| 990 |
+
"epoch": 0.4218455743879473,
|
| 991 |
+
"step": 140
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"loss": 1.307,
|
| 995 |
+
"grad_norm": 1.3790669441223145,
|
| 996 |
+
"learning_rate": 3.971878042924502e-05,
|
| 997 |
+
"epoch": 0.4248587570621469,
|
| 998 |
+
"step": 141
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"loss": 1.3854,
|
| 1002 |
+
"grad_norm": 1.669015645980835,
|
| 1003 |
+
"learning_rate": 3.82720417653e-05,
|
| 1004 |
+
"epoch": 0.4278719397363465,
|
| 1005 |
+
"step": 142
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"loss": 1.2752,
|
| 1009 |
+
"grad_norm": 1.4703353643417358,
|
| 1010 |
+
"learning_rate": 3.6819638595669306e-05,
|
| 1011 |
+
"epoch": 0.43088512241054616,
|
| 1012 |
+
"step": 143
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"loss": 1.2645,
|
| 1016 |
+
"grad_norm": 1.2362326383590698,
|
| 1017 |
+
"learning_rate": 3.536408529718624e-05,
|
| 1018 |
+
"epoch": 0.43389830508474575,
|
| 1019 |
+
"step": 144
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"loss": 1.2514,
|
| 1023 |
+
"grad_norm": 1.261176347732544,
|
| 1024 |
+
"learning_rate": 3.3907901700136133e-05,
|
| 1025 |
+
"epoch": 0.4369114877589454,
|
| 1026 |
+
"step": 145
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"loss": 1.3644,
|
| 1030 |
+
"grad_norm": 1.4008452892303467,
|
| 1031 |
+
"learning_rate": 3.245360872596697e-05,
|
| 1032 |
+
"epoch": 0.439924670433145,
|
| 1033 |
+
"step": 146
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"loss": 1.3723,
|
| 1037 |
+
"grad_norm": 1.363226056098938,
|
| 1038 |
+
"learning_rate": 3.100372402311105e-05,
|
| 1039 |
+
"epoch": 0.4429378531073446,
|
| 1040 |
+
"step": 147
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"loss": 1.4068,
|
| 1044 |
+
"grad_norm": 1.4670599699020386,
|
| 1045 |
+
"learning_rate": 2.956075760847284e-05,
|
| 1046 |
+
"epoch": 0.44595103578154427,
|
| 1047 |
+
"step": 148
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"loss": 1.3961,
|
| 1051 |
+
"grad_norm": 1.3713915348052979,
|
| 1052 |
+
"learning_rate": 2.8127207522128252e-05,
|
| 1053 |
+
"epoch": 0.44896421845574386,
|
| 1054 |
+
"step": 149
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"loss": 1.3453,
|
| 1058 |
+
"grad_norm": 1.4361172914505005,
|
| 1059 |
+
"learning_rate": 2.6705555502758335e-05,
|
| 1060 |
+
"epoch": 0.4519774011299435,
|
| 1061 |
+
"step": 150
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"loss": 1.4574,
|
| 1065 |
+
"grad_norm": 1.4682360887527466,
|
| 1066 |
+
"learning_rate": 2.529826269130334e-05,
|
| 1067 |
+
"epoch": 0.45499058380414314,
|
| 1068 |
+
"step": 151
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"loss": 1.3799,
|
| 1072 |
+
"grad_norm": 1.2876993417739868,
|
| 1073 |
+
"learning_rate": 2.3907765370275404e-05,
|
| 1074 |
+
"epoch": 0.45800376647834273,
|
| 1075 |
+
"step": 152
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"loss": 1.3479,
|
| 1079 |
+
"grad_norm": 1.2455275058746338,
|
| 1080 |
+
"learning_rate": 2.2536470746105853e-05,
|
| 1081 |
+
"epoch": 0.4610169491525424,
|
| 1082 |
+
"step": 153
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"loss": 1.3099,
|
| 1086 |
+
"grad_norm": 1.3604997396469116,
|
| 1087 |
+
"learning_rate": 2.118675278182831e-05,
|
| 1088 |
+
"epoch": 0.464030131826742,
|
| 1089 |
+
"step": 154
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"loss": 1.3122,
|
| 1093 |
+
"grad_norm": 1.2599378824234009,
|
| 1094 |
+
"learning_rate": 1.986094808731247e-05,
|
| 1095 |
+
"epoch": 0.4670433145009416,
|
| 1096 |
+
"step": 155
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"loss": 1.3371,
|
| 1100 |
+
"grad_norm": 1.7119026184082031,
|
| 1101 |
+
"learning_rate": 1.8561351874162783e-05,
|
| 1102 |
+
"epoch": 0.47005649717514125,
|
| 1103 |
+
"step": 156
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"loss": 1.2278,
|
| 1107 |
+
"grad_norm": 1.3168460130691528,
|
| 1108 |
+
"learning_rate": 1.7290213982285547e-05,
|
| 1109 |
+
"epoch": 0.47306967984934084,
|
| 1110 |
+
"step": 157
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"loss": 1.5478,
|
| 1114 |
+
"grad_norm": 1.4438059329986572,
|
| 1115 |
+
"learning_rate": 1.604973498500235e-05,
|
| 1116 |
+
"epoch": 0.4760828625235405,
|
| 1117 |
+
"step": 158
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"loss": 1.3363,
|
| 1121 |
+
"grad_norm": 1.422288179397583,
|
| 1122 |
+
"learning_rate": 1.4842062379453195e-05,
|
| 1123 |
+
"epoch": 0.47909604519774013,
|
| 1124 |
+
"step": 159
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"loss": 1.3833,
|
| 1128 |
+
"grad_norm": 1.5555317401885986,
|
| 1129 |
+
"learning_rate": 1.3669286868884356e-05,
|
| 1130 |
+
"epoch": 0.4821092278719397,
|
| 1131 |
+
"step": 160
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"loss": 1.3687,
|
| 1135 |
+
"grad_norm": 1.2632575035095215,
|
| 1136 |
+
"learning_rate": 1.2533438743256715e-05,
|
| 1137 |
+
"epoch": 0.48512241054613936,
|
| 1138 |
+
"step": 161
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"loss": 1.3074,
|
| 1142 |
+
"grad_norm": 1.4185025691986084,
|
| 1143 |
+
"learning_rate": 1.1436484364440646e-05,
|
| 1144 |
+
"epoch": 0.488135593220339,
|
| 1145 |
+
"step": 162
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"loss": 1.5231,
|
| 1149 |
+
"grad_norm": 1.4433213472366333,
|
| 1150 |
+
"learning_rate": 1.038032276208234e-05,
|
| 1151 |
+
"epoch": 0.4911487758945386,
|
| 1152 |
+
"step": 163
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"loss": 1.3687,
|
| 1156 |
+
"grad_norm": 1.3245162963867188,
|
| 1157 |
+
"learning_rate": 9.36678234603432e-06,
|
| 1158 |
+
"epoch": 0.49416195856873824,
|
| 1159 |
+
"step": 164
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"loss": 1.2739,
|
| 1163 |
+
"grad_norm": 1.3129271268844604,
|
| 1164 |
+
"learning_rate": 8.397617741041962e-06,
|
| 1165 |
+
"epoch": 0.4971751412429379,
|
| 1166 |
+
"step": 165
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"loss": 1.2536,
|
| 1170 |
+
"grad_norm": 1.356525182723999,
|
| 1171 |
+
"learning_rate": 7.47450674916562e-06,
|
| 1172 |
+
"epoch": 0.5001883239171375,
|
| 1173 |
+
"step": 166
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"loss": 1.3418,
|
| 1177 |
+
"grad_norm": 1.4031752347946167,
|
| 1178 |
+
"learning_rate": 6.599047445196828e-06,
|
| 1179 |
+
"epoch": 0.5032015065913371,
|
| 1180 |
+
"step": 167
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"loss": 1.3624,
|
| 1184 |
+
"grad_norm": 1.3876193761825562,
|
| 1185 |
+
"learning_rate": 5.772755410097178e-06,
|
| 1186 |
+
"epoch": 0.5062146892655367,
|
| 1187 |
+
"step": 168
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"loss": 1.2356,
|
| 1191 |
+
"grad_norm": 1.2425028085708618,
|
| 1192 |
+
"learning_rate": 4.997061107248992e-06,
|
| 1193 |
+
"epoch": 0.5092278719397364,
|
| 1194 |
+
"step": 169
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"loss": 1.3582,
|
| 1198 |
+
"grad_norm": 1.3196662664413452,
|
| 1199 |
+
"learning_rate": 4.273307406060457e-06,
|
| 1200 |
+
"epoch": 0.512241054613936,
|
| 1201 |
+
"step": 170
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"loss": 1.3458,
|
| 1205 |
+
"grad_norm": 1.379664659500122,
|
| 1206 |
+
"learning_rate": 3.6027472572117617e-06,
|
| 1207 |
+
"epoch": 0.5152542372881356,
|
| 1208 |
+
"step": 171
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"loss": 1.2718,
|
| 1212 |
+
"grad_norm": 1.3308255672454834,
|
| 1213 |
+
"learning_rate": 2.9865415235671784e-06,
|
| 1214 |
+
"epoch": 0.5182674199623352,
|
| 1215 |
+
"step": 172
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"loss": 1.2453,
|
| 1219 |
+
"grad_norm": 1.276235818862915,
|
| 1220 |
+
"learning_rate": 2.4257569705081314e-06,
|
| 1221 |
+
"epoch": 0.5212806026365349,
|
| 1222 |
+
"step": 173
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"loss": 1.3344,
|
| 1226 |
+
"grad_norm": 1.3964022397994995,
|
| 1227 |
+
"learning_rate": 1.9213644191662393e-06,
|
| 1228 |
+
"epoch": 0.5242937853107345,
|
| 1229 |
+
"step": 174
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"loss": 1.3499,
|
| 1233 |
+
"grad_norm": 1.3098505735397339,
|
| 1234 |
+
"learning_rate": 1.4742370657534552e-06,
|
| 1235 |
+
"epoch": 0.527306967984934,
|
| 1236 |
+
"step": 175
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"loss": 1.2901,
|
| 1240 |
+
"grad_norm": 1.2909560203552246,
|
| 1241 |
+
"learning_rate": 1.0851489698990146e-06,
|
| 1242 |
+
"epoch": 0.5303201506591337,
|
| 1243 |
+
"step": 176
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"loss": 1.3972,
|
| 1247 |
+
"grad_norm": 1.4206069707870483,
|
| 1248 |
+
"learning_rate": 7.547737146098809e-07,
|
| 1249 |
+
"epoch": 0.5333333333333333,
|
| 1250 |
+
"step": 177
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"loss": 1.2057,
|
| 1254 |
+
"grad_norm": 1.2130017280578613,
|
| 1255 |
+
"learning_rate": 4.836832401747892e-07,
|
| 1256 |
+
"epoch": 0.5363465160075329,
|
| 1257 |
+
"step": 178
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"loss": 1.4497,
|
| 1261 |
+
"grad_norm": 1.3319237232208252,
|
| 1262 |
+
"learning_rate": 2.7234685403047243e-07,
|
| 1263 |
+
"epoch": 0.5393596986817326,
|
| 1264 |
+
"step": 179
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"loss": 1.2815,
|
| 1268 |
+
"grad_norm": 1.2627345323562622,
|
| 1269 |
+
"learning_rate": 1.2113041830424553e-07,
|
| 1270 |
+
"epoch": 0.5423728813559322,
|
| 1271 |
+
"step": 180
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"loss": 1.2589,
|
| 1275 |
+
"grad_norm": 1.1925126314163208,
|
| 1276 |
+
"learning_rate": 3.029571643941353e-08,
|
| 1277 |
+
"epoch": 0.5453860640301318,
|
| 1278 |
+
"step": 181
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"loss": 1.3669,
|
| 1282 |
+
"grad_norm": 1.4577635526657104,
|
| 1283 |
+
"learning_rate": 7e-05,
|
| 1284 |
+
"epoch": 0.5483992467043315,
|
| 1285 |
+
"step": 182
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"loss": 1.3824,
|
| 1289 |
+
"grad_norm": 1.310353398323059,
|
| 1290 |
+
"learning_rate": 6.996970428356058e-05,
|
| 1291 |
+
"epoch": 0.5514124293785311,
|
| 1292 |
+
"step": 183
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"loss": 1.3753,
|
| 1296 |
+
"grad_norm": 1.5347837209701538,
|
| 1297 |
+
"learning_rate": 6.987886958169575e-05,
|
| 1298 |
+
"epoch": 0.5544256120527307,
|
| 1299 |
+
"step": 184
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"loss": 1.3558,
|
| 1303 |
+
"grad_norm": 1.647703766822815,
|
| 1304 |
+
"learning_rate": 6.972765314596951e-05,
|
| 1305 |
+
"epoch": 0.5574387947269304,
|
| 1306 |
+
"step": 185
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"loss": 1.2365,
|
| 1310 |
+
"grad_norm": 1.4471580982208252,
|
| 1311 |
+
"learning_rate": 6.951631675982521e-05,
|
| 1312 |
+
"epoch": 0.56045197740113,
|
| 1313 |
+
"step": 186
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"loss": 1.3501,
|
| 1317 |
+
"grad_norm": 1.5231457948684692,
|
| 1318 |
+
"learning_rate": 6.924522628539012e-05,
|
| 1319 |
+
"epoch": 0.5634651600753295,
|
| 1320 |
+
"step": 187
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"loss": 1.3995,
|
| 1324 |
+
"grad_norm": 1.5145260095596313,
|
| 1325 |
+
"learning_rate": 6.891485103010098e-05,
|
| 1326 |
+
"epoch": 0.5664783427495292,
|
| 1327 |
+
"step": 188
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"loss": 1.2937,
|
| 1331 |
+
"grad_norm": 1.4664785861968994,
|
| 1332 |
+
"learning_rate": 6.852576293424654e-05,
|
| 1333 |
+
"epoch": 0.5694915254237288,
|
| 1334 |
+
"step": 189
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"loss": 1.3907,
|
| 1338 |
+
"grad_norm": 1.778623342514038,
|
| 1339 |
+
"learning_rate": 6.807863558083377e-05,
|
| 1340 |
+
"epoch": 0.5725047080979284,
|
| 1341 |
+
"step": 190
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"loss": 1.435,
|
| 1345 |
+
"grad_norm": 1.3576749563217163,
|
| 1346 |
+
"learning_rate": 6.757424302949185e-05,
|
| 1347 |
+
"epoch": 0.5755178907721281,
|
| 1348 |
+
"step": 191
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"loss": 1.2448,
|
| 1352 |
+
"grad_norm": 1.3500125408172607,
|
| 1353 |
+
"learning_rate": 6.701345847643283e-05,
|
| 1354 |
+
"epoch": 0.5785310734463277,
|
| 1355 |
+
"step": 192
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"loss": 1.3058,
|
| 1359 |
+
"grad_norm": 1.4894977807998657,
|
| 1360 |
+
"learning_rate": 6.639725274278824e-05,
|
| 1361 |
+
"epoch": 0.5815442561205273,
|
| 1362 |
+
"step": 193
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"loss": 1.2644,
|
| 1366 |
+
"grad_norm": 1.4336787462234497,
|
| 1367 |
+
"learning_rate": 6.572669259393954e-05,
|
| 1368 |
+
"epoch": 0.5845574387947269,
|
| 1369 |
+
"step": 194
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"loss": 1.3507,
|
| 1373 |
+
"grad_norm": 1.4022774696350098,
|
| 1374 |
+
"learning_rate": 6.5002938892751e-05,
|
| 1375 |
+
"epoch": 0.5875706214689266,
|
| 1376 |
+
"step": 195
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"loss": 1.3795,
|
| 1380 |
+
"grad_norm": 1.3695868253707886,
|
| 1381 |
+
"learning_rate": 6.422724458990282e-05,
|
| 1382 |
+
"epoch": 0.5905838041431262,
|
| 1383 |
+
"step": 196
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"loss": 1.3877,
|
| 1387 |
+
"grad_norm": 1.5911080837249756,
|
| 1388 |
+
"learning_rate": 6.340095255480315e-05,
|
| 1389 |
+
"epoch": 0.5935969868173258,
|
| 1390 |
+
"step": 197
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"loss": 1.3487,
|
| 1394 |
+
"grad_norm": 1.3178699016571045,
|
| 1395 |
+
"learning_rate": 6.252549325083437e-05,
|
| 1396 |
+
"epoch": 0.5966101694915255,
|
| 1397 |
+
"step": 198
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"loss": 1.4043,
|
| 1401 |
+
"grad_norm": 1.3161568641662598,
|
| 1402 |
+
"learning_rate": 6.160238225895805e-05,
|
| 1403 |
+
"epoch": 0.599623352165725,
|
| 1404 |
+
"step": 199
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"loss": 1.2959,
|
| 1408 |
+
"grad_norm": 1.4915833473205566,
|
| 1409 |
+
"learning_rate": 6.063321765396569e-05,
|
| 1410 |
+
"epoch": 0.6026365348399246,
|
| 1411 |
+
"step": 200
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"loss": 1.2981,
|
| 1415 |
+
"grad_norm": 1.327701449394226,
|
| 1416 |
+
"learning_rate": 5.9619677237917655e-05,
|
| 1417 |
+
"epoch": 0.6056497175141243,
|
| 1418 |
+
"step": 201
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"loss": 1.3405,
|
| 1422 |
+
"grad_norm": 1.3549165725708008,
|
| 1423 |
+
"learning_rate": 5.8563515635559344e-05,
|
| 1424 |
+
"epoch": 0.6086629001883239,
|
| 1425 |
+
"step": 202
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"loss": 1.3706,
|
| 1429 |
+
"grad_norm": 1.3070836067199707,
|
| 1430 |
+
"learning_rate": 5.746656125674329e-05,
|
| 1431 |
+
"epoch": 0.6116760828625235,
|
| 1432 |
+
"step": 203
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"loss": 1.3477,
|
| 1436 |
+
"grad_norm": 1.382853627204895,
|
| 1437 |
+
"learning_rate": 5.6330713131115626e-05,
|
| 1438 |
+
"epoch": 0.6146892655367232,
|
| 1439 |
+
"step": 204
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"loss": 1.4745,
|
| 1443 |
+
"grad_norm": 1.5222082138061523,
|
| 1444 |
+
"learning_rate": 5.5157937620546824e-05,
|
| 1445 |
+
"epoch": 0.6177024482109228,
|
| 1446 |
+
"step": 205
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"loss": 1.1981,
|
| 1450 |
+
"grad_norm": 1.4459370374679565,
|
| 1451 |
+
"learning_rate": 5.395026501499767e-05,
|
| 1452 |
+
"epoch": 0.6207156308851224,
|
| 1453 |
+
"step": 206
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"loss": 1.263,
|
| 1457 |
+
"grad_norm": 1.2831898927688599,
|
| 1458 |
+
"learning_rate": 5.270978601771445e-05,
|
| 1459 |
+
"epoch": 0.6237288135593221,
|
| 1460 |
+
"step": 207
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"loss": 1.4531,
|
| 1464 |
+
"grad_norm": 1.5609828233718872,
|
| 1465 |
+
"learning_rate": 5.1438648125837204e-05,
|
| 1466 |
+
"epoch": 0.6267419962335217,
|
| 1467 |
+
"step": 208
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"loss": 1.3804,
|
| 1471 |
+
"grad_norm": 1.3696374893188477,
|
| 1472 |
+
"learning_rate": 5.013905191268753e-05,
|
| 1473 |
+
"epoch": 0.6297551789077213,
|
| 1474 |
+
"step": 209
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"loss": 1.3721,
|
| 1478 |
+
"grad_norm": 1.3352577686309814,
|
| 1479 |
+
"learning_rate": 4.881324721817166e-05,
|
| 1480 |
+
"epoch": 0.632768361581921,
|
| 1481 |
+
"step": 210
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"loss": 1.3877,
|
| 1485 |
+
"grad_norm": 1.4097323417663574,
|
| 1486 |
+
"learning_rate": 4.746352925389412e-05,
|
| 1487 |
+
"epoch": 0.6357815442561205,
|
| 1488 |
+
"step": 211
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"loss": 1.3914,
|
| 1492 |
+
"grad_norm": 1.3484840393066406,
|
| 1493 |
+
"learning_rate": 4.609223462972462e-05,
|
| 1494 |
+
"epoch": 0.6387947269303201,
|
| 1495 |
+
"step": 212
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"loss": 1.3696,
|
| 1499 |
+
"grad_norm": 1.3758927583694458,
|
| 1500 |
+
"learning_rate": 4.470173730869668e-05,
|
| 1501 |
+
"epoch": 0.6418079096045197,
|
| 1502 |
+
"step": 213
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"loss": 1.3395,
|
| 1506 |
+
"grad_norm": 1.3878891468048096,
|
| 1507 |
+
"learning_rate": 4.329444449724166e-05,
|
| 1508 |
+
"epoch": 0.6448210922787194,
|
| 1509 |
+
"step": 214
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"loss": 1.2247,
|
| 1513 |
+
"grad_norm": 1.4328563213348389,
|
| 1514 |
+
"learning_rate": 4.187279247787175e-05,
|
| 1515 |
+
"epoch": 0.647834274952919,
|
| 1516 |
+
"step": 215
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"loss": 1.3263,
|
| 1520 |
+
"grad_norm": 1.3026245832443237,
|
| 1521 |
+
"learning_rate": 4.043924239152716e-05,
|
| 1522 |
+
"epoch": 0.6508474576271186,
|
| 1523 |
+
"step": 216
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"loss": 1.3123,
|
| 1527 |
+
"grad_norm": 1.3922412395477295,
|
| 1528 |
+
"learning_rate": 3.899627597688892e-05,
|
| 1529 |
+
"epoch": 0.6538606403013183,
|
| 1530 |
+
"step": 217
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"loss": 1.265,
|
| 1534 |
+
"grad_norm": 1.2023494243621826,
|
| 1535 |
+
"learning_rate": 3.7546391274033006e-05,
|
| 1536 |
+
"epoch": 0.6568738229755179,
|
| 1537 |
+
"step": 218
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"loss": 1.3999,
|
| 1541 |
+
"grad_norm": 1.6529396772384644,
|
| 1542 |
+
"learning_rate": 3.609209829986388e-05,
|
| 1543 |
+
"epoch": 0.6598870056497175,
|
| 1544 |
+
"step": 219
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"loss": 1.2666,
|
| 1548 |
+
"grad_norm": 1.524455189704895,
|
| 1549 |
+
"learning_rate": 3.463591470281376e-05,
|
| 1550 |
+
"epoch": 0.6629001883239172,
|
| 1551 |
+
"step": 220
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"loss": 1.3959,
|
| 1555 |
+
"grad_norm": 1.4471687078475952,
|
| 1556 |
+
"learning_rate": 3.318036140433069e-05,
|
| 1557 |
+
"epoch": 0.6659133709981168,
|
| 1558 |
+
"step": 221
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"loss": 1.3262,
|
| 1562 |
+
"grad_norm": 1.401480793952942,
|
| 1563 |
+
"learning_rate": 3.172795823469999e-05,
|
| 1564 |
+
"epoch": 0.6689265536723163,
|
| 1565 |
+
"step": 222
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"loss": 1.1911,
|
| 1569 |
+
"grad_norm": 1.2851903438568115,
|
| 1570 |
+
"learning_rate": 3.0281219570754974e-05,
|
| 1571 |
+
"epoch": 0.671939736346516,
|
| 1572 |
+
"step": 223
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"loss": 1.2493,
|
| 1576 |
+
"grad_norm": 1.499954342842102,
|
| 1577 |
+
"learning_rate": 2.8842649983028493e-05,
|
| 1578 |
+
"epoch": 0.6749529190207156,
|
| 1579 |
+
"step": 224
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"loss": 1.3212,
|
| 1583 |
+
"grad_norm": 1.3503855466842651,
|
| 1584 |
+
"learning_rate": 2.7414739899881044e-05,
|
| 1585 |
+
"epoch": 0.6779661016949152,
|
| 1586 |
+
"step": 225
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"loss": 1.3487,
|
| 1590 |
+
"grad_norm": 1.3009147644042969,
|
| 1591 |
+
"learning_rate": 2.5999961296111407e-05,
|
| 1592 |
+
"epoch": 0.6809792843691149,
|
| 1593 |
+
"step": 226
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"loss": 1.2527,
|
| 1597 |
+
"grad_norm": 1.3278416395187378,
|
| 1598 |
+
"learning_rate": 2.4600763413514152e-05,
|
| 1599 |
+
"epoch": 0.6839924670433145,
|
| 1600 |
+
"step": 227
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"loss": 1.1836,
|
| 1604 |
+
"grad_norm": 1.2308480739593506,
|
| 1605 |
+
"learning_rate": 2.3219568520791783e-05,
|
| 1606 |
+
"epoch": 0.6870056497175141,
|
| 1607 |
+
"step": 228
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"loss": 1.2542,
|
| 1611 |
+
"grad_norm": 1.3015786409378052,
|
| 1612 |
+
"learning_rate": 2.1858767720162443e-05,
|
| 1613 |
+
"epoch": 0.6900188323917138,
|
| 1614 |
+
"step": 229
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"loss": 1.3637,
|
| 1618 |
+
"grad_norm": 1.35995614528656,
|
| 1619 |
+
"learning_rate": 2.0520716807922507e-05,
|
| 1620 |
+
"epoch": 0.6930320150659134,
|
| 1621 |
+
"step": 230
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"loss": 1.2854,
|
| 1625 |
+
"grad_norm": 1.2334768772125244,
|
| 1626 |
+
"learning_rate": 1.920773219613017e-05,
|
| 1627 |
+
"epoch": 0.696045197740113,
|
| 1628 |
+
"step": 231
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"loss": 1.3015,
|
| 1632 |
+
"grad_norm": 1.3583332300186157,
|
| 1633 |
+
"learning_rate": 1.7922086902470533e-05,
|
| 1634 |
+
"epoch": 0.6990583804143126,
|
| 1635 |
+
"step": 232
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"loss": 1.3801,
|
| 1639 |
+
"grad_norm": 1.4429166316986084,
|
| 1640 |
+
"learning_rate": 1.666600661524396e-05,
|
| 1641 |
+
"epoch": 0.7020715630885123,
|
| 1642 |
+
"step": 233
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"loss": 1.366,
|
| 1646 |
+
"grad_norm": 1.2505085468292236,
|
| 1647 |
+
"learning_rate": 1.5441665840290927e-05,
|
| 1648 |
+
"epoch": 0.7050847457627119,
|
| 1649 |
+
"step": 234
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"loss": 1.2893,
|
| 1653 |
+
"grad_norm": 1.236722469329834,
|
| 1654 |
+
"learning_rate": 1.4251184136522453e-05,
|
| 1655 |
+
"epoch": 0.7080979284369114,
|
| 1656 |
+
"step": 235
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"loss": 1.4647,
|
| 1660 |
+
"grad_norm": 1.7293294668197632,
|
| 1661 |
+
"learning_rate": 1.3096622446574172e-05,
|
| 1662 |
+
"epoch": 0.7111111111111111,
|
| 1663 |
+
"step": 236
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"loss": 1.3912,
|
| 1667 |
+
"grad_norm": 1.3492850065231323,
|
| 1668 |
+
"learning_rate": 1.1979979528935841e-05,
|
| 1669 |
+
"epoch": 0.7141242937853107,
|
| 1670 |
+
"step": 237
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"loss": 1.3879,
|
| 1674 |
+
"grad_norm": 1.2749269008636475,
|
| 1675 |
+
"learning_rate": 1.0903188497733022e-05,
|
| 1676 |
+
"epoch": 0.7171374764595103,
|
| 1677 |
+
"step": 238
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"loss": 1.4184,
|
| 1681 |
+
"grad_norm": 1.3850064277648926,
|
| 1682 |
+
"learning_rate": 9.868113476151121e-06,
|
| 1683 |
+
"epoch": 0.72015065913371,
|
| 1684 |
+
"step": 239
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"loss": 1.3039,
|
| 1688 |
+
"grad_norm": 1.2930525541305542,
|
| 1689 |
+
"learning_rate": 8.876546369295728e-06,
|
| 1690 |
+
"epoch": 0.7231638418079096,
|
| 1691 |
+
"step": 240
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"loss": 1.3727,
|
| 1695 |
+
"grad_norm": 1.3093020915985107,
|
| 1696 |
+
"learning_rate": 7.930203762075426e-06,
|
| 1697 |
+
"epoch": 0.7261770244821092,
|
| 1698 |
+
"step": 241
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"loss": 1.269,
|
| 1702 |
+
"grad_norm": 1.2753149271011353,
|
| 1703 |
+
"learning_rate": 7.030723947477803e-06,
|
| 1704 |
+
"epoch": 0.7291902071563089,
|
| 1705 |
+
"step": 242
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"loss": 1.2097,
|
| 1709 |
+
"grad_norm": 1.3224420547485352,
|
| 1710 |
+
"learning_rate": 6.179664090383193e-06,
|
| 1711 |
+
"epoch": 0.7322033898305085,
|
| 1712 |
+
"step": 243
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"loss": 1.378,
|
| 1716 |
+
"grad_norm": 1.4395109415054321,
|
| 1717 |
+
"learning_rate": 5.378497531826102e-06,
|
| 1718 |
+
"epoch": 0.7352165725047081,
|
| 1719 |
+
"step": 244
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"loss": 1.409,
|
| 1723 |
+
"grad_norm": 1.3060985803604126,
|
| 1724 |
+
"learning_rate": 4.62861123837104e-06,
|
| 1725 |
+
"epoch": 0.7382297551789078,
|
| 1726 |
+
"step": 245
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"loss": 1.3575,
|
| 1730 |
+
"grad_norm": 1.3586187362670898,
|
| 1731 |
+
"learning_rate": 3.931303401018435e-06,
|
| 1732 |
+
"epoch": 0.7412429378531074,
|
| 1733 |
+
"step": 246
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"loss": 1.3275,
|
| 1737 |
+
"grad_norm": 1.3876395225524902,
|
| 1738 |
+
"learning_rate": 3.2877811877975422e-06,
|
| 1739 |
+
"epoch": 0.7442561205273069,
|
| 1740 |
+
"step": 247
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"loss": 1.3028,
|
| 1744 |
+
"grad_norm": 1.3465113639831543,
|
| 1745 |
+
"learning_rate": 2.6991586539366018e-06,
|
| 1746 |
+
"epoch": 0.7472693032015066,
|
| 1747 |
+
"step": 248
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"loss": 1.3433,
|
| 1751 |
+
"grad_norm": 1.2856528759002686,
|
| 1752 |
+
"learning_rate": 2.1664548132285013e-06,
|
| 1753 |
+
"epoch": 0.7502824858757062,
|
| 1754 |
+
"step": 249
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"loss": 1.2616,
|
| 1758 |
+
"grad_norm": 1.2871434688568115,
|
| 1759 |
+
"learning_rate": 1.6905918739306385e-06,
|
| 1760 |
+
"epoch": 0.7532956685499058,
|
| 1761 |
+
"step": 250
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"loss": 1.3733,
|
| 1765 |
+
"grad_norm": 1.3069396018981934,
|
| 1766 |
+
"learning_rate": 1.2723936422529074e-06,
|
| 1767 |
+
"epoch": 0.7563088512241055,
|
| 1768 |
+
"step": 251
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"loss": 1.3255,
|
| 1772 |
+
"grad_norm": 1.2253414392471313,
|
| 1773 |
+
"learning_rate": 9.125840961976982e-07,
|
| 1774 |
+
"epoch": 0.7593220338983051,
|
| 1775 |
+
"step": 252
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"loss": 1.1984,
|
| 1779 |
+
"grad_norm": 1.2259007692337036,
|
| 1780 |
+
"learning_rate": 6.117861322210066e-07,
|
| 1781 |
+
"epoch": 0.7623352165725047,
|
| 1782 |
+
"step": 253
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"loss": 1.3364,
|
| 1786 |
+
"grad_norm": 1.376477837562561,
|
| 1787 |
+
"learning_rate": 3.705204868840894e-07,
|
| 1788 |
+
"epoch": 0.7653483992467043,
|
| 1789 |
+
"step": 254
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"loss": 1.3324,
|
| 1793 |
+
"grad_norm": 1.275305151939392,
|
| 1794 |
+
"learning_rate": 1.892048353628217e-07,
|
| 1795 |
+
"epoch": 0.768361581920904,
|
| 1796 |
+
"step": 255
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"loss": 1.3153,
|
| 1800 |
+
"grad_norm": 1.4428046941757202,
|
| 1801 |
+
"learning_rate": 6.81530683751752e-08,
|
| 1802 |
+
"epoch": 0.7713747645951036,
|
| 1803 |
+
"step": 256
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"loss": 1.4257,
|
| 1807 |
+
"grad_norm": 1.2987799644470215,
|
| 1808 |
+
"learning_rate": 7.574748778697015e-09,
|
| 1809 |
+
"epoch": 0.7743879472693032,
|
| 1810 |
+
"step": 257
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"loss": 1.3872,
|
| 1814 |
+
"grad_norm": 1.2806425094604492,
|
| 1815 |
+
"learning_rate": 6.99924252512213e-05,
|
| 1816 |
+
"epoch": 0.7774011299435029,
|
| 1817 |
+
"step": 258
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"loss": 1.3154,
|
| 1821 |
+
"grad_norm": 1.3498905897140503,
|
| 1822 |
+
"learning_rate": 6.993184693162481e-05,
|
| 1823 |
+
"epoch": 0.7804143126177024,
|
| 1824 |
+
"step": 259
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"loss": 1.2901,
|
| 1828 |
+
"grad_norm": 1.553397536277771,
|
| 1829 |
+
"learning_rate": 6.981079516463718e-05,
|
| 1830 |
+
"epoch": 0.783427495291902,
|
| 1831 |
+
"step": 260
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"loss": 1.3677,
|
| 1835 |
+
"grad_norm": 1.8604789972305298,
|
| 1836 |
+
"learning_rate": 6.962947951311591e-05,
|
| 1837 |
+
"epoch": 0.7864406779661017,
|
| 1838 |
+
"step": 261
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"loss": 1.3658,
|
| 1842 |
+
"grad_norm": 1.8038997650146484,
|
| 1843 |
+
"learning_rate": 6.938821386777899e-05,
|
| 1844 |
+
"epoch": 0.7894538606403013,
|
| 1845 |
+
"step": 262
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"loss": 1.4677,
|
| 1849 |
+
"grad_norm": 1.5064972639083862,
|
| 1850 |
+
"learning_rate": 6.90874159038023e-05,
|
| 1851 |
+
"epoch": 0.7924670433145009,
|
| 1852 |
+
"step": 263
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"loss": 1.3751,
|
| 1856 |
+
"grad_norm": 1.3953901529312134,
|
| 1857 |
+
"learning_rate": 6.872760635774709e-05,
|
| 1858 |
+
"epoch": 0.7954802259887006,
|
| 1859 |
+
"step": 264
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"loss": 1.4987,
|
| 1863 |
+
"grad_norm": 1.3525390625,
|
| 1864 |
+
"learning_rate": 6.830940812606936e-05,
|
| 1865 |
+
"epoch": 0.7984934086629002,
|
| 1866 |
+
"step": 265
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"loss": 1.3228,
|
| 1870 |
+
"grad_norm": 1.7438874244689941,
|
| 1871 |
+
"learning_rate": 6.78335451867715e-05,
|
| 1872 |
+
"epoch": 0.8015065913370998,
|
| 1873 |
+
"step": 266
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"loss": 1.4868,
|
| 1877 |
+
"grad_norm": 1.5288082361221313,
|
| 1878 |
+
"learning_rate": 6.73008413460634e-05,
|
| 1879 |
+
"epoch": 0.8045197740112995,
|
| 1880 |
+
"step": 267
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"loss": 1.2498,
|
| 1884 |
+
"grad_norm": 1.6611965894699097,
|
| 1885 |
+
"learning_rate": 6.671221881220246e-05,
|
| 1886 |
+
"epoch": 0.8075329566854991,
|
| 1887 |
+
"step": 268
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"loss": 1.3629,
|
| 1891 |
+
"grad_norm": 1.265153169631958,
|
| 1892 |
+
"learning_rate": 6.606869659898156e-05,
|
| 1893 |
+
"epoch": 0.8105461393596987,
|
| 1894 |
+
"step": 269
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"loss": 1.3348,
|
| 1898 |
+
"grad_norm": 1.2199232578277588,
|
| 1899 |
+
"learning_rate": 6.537138876162896e-05,
|
| 1900 |
+
"epoch": 0.8135593220338984,
|
| 1901 |
+
"step": 270
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"loss": 1.3561,
|
| 1905 |
+
"grad_norm": 1.3994086980819702,
|
| 1906 |
+
"learning_rate": 6.46215024681739e-05,
|
| 1907 |
+
"epoch": 0.816572504708098,
|
| 1908 |
+
"step": 271
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"loss": 1.2723,
|
| 1912 |
+
"grad_norm": 1.1945056915283203,
|
| 1913 |
+
"learning_rate": 6.38203359096168e-05,
|
| 1914 |
+
"epoch": 0.8195856873822975,
|
| 1915 |
+
"step": 272
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"loss": 1.2476,
|
| 1919 |
+
"grad_norm": 1.353142499923706,
|
| 1920 |
+
"learning_rate": 6.296927605252219e-05,
|
| 1921 |
+
"epoch": 0.8225988700564971,
|
| 1922 |
+
"step": 273
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"loss": 1.236,
|
| 1926 |
+
"grad_norm": 1.260233759880066,
|
| 1927 |
+
"learning_rate": 6.206979623792457e-05,
|
| 1928 |
+
"epoch": 0.8256120527306968,
|
| 1929 |
+
"step": 274
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"loss": 1.2879,
|
| 1933 |
+
"grad_norm": 1.2068685293197632,
|
| 1934 |
+
"learning_rate": 6.112345363070427e-05,
|
| 1935 |
+
"epoch": 0.8286252354048964,
|
| 1936 |
+
"step": 275
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"loss": 1.4323,
|
| 1940 |
+
"grad_norm": 1.3854366540908813,
|
| 1941 |
+
"learning_rate": 6.013188652384888e-05,
|
| 1942 |
+
"epoch": 0.831638418079096,
|
| 1943 |
+
"step": 276
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"loss": 1.2642,
|
| 1947 |
+
"grad_norm": 1.3159300088882446,
|
| 1948 |
+
"learning_rate": 5.909681150226697e-05,
|
| 1949 |
+
"epoch": 0.8346516007532957,
|
| 1950 |
+
"step": 277
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"loss": 1.2844,
|
| 1954 |
+
"grad_norm": 1.2609727382659912,
|
| 1955 |
+
"learning_rate": 5.802002047106416e-05,
|
| 1956 |
+
"epoch": 0.8376647834274953,
|
| 1957 |
+
"step": 278
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"loss": 1.3893,
|
| 1961 |
+
"grad_norm": 1.2146693468093872,
|
| 1962 |
+
"learning_rate": 5.690337755342582e-05,
|
| 1963 |
+
"epoch": 0.8406779661016949,
|
| 1964 |
+
"step": 279
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"loss": 1.3376,
|
| 1968 |
+
"grad_norm": 1.3817909955978394,
|
| 1969 |
+
"learning_rate": 5.574881586347755e-05,
|
| 1970 |
+
"epoch": 0.8436911487758946,
|
| 1971 |
+
"step": 280
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"loss": 1.287,
|
| 1975 |
+
"grad_norm": 1.1949174404144287,
|
| 1976 |
+
"learning_rate": 5.455833415970907e-05,
|
| 1977 |
+
"epoch": 0.8467043314500942,
|
| 1978 |
+
"step": 281
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"loss": 1.3756,
|
| 1982 |
+
"grad_norm": 1.3564906120300293,
|
| 1983 |
+
"learning_rate": 5.333399338475603e-05,
|
| 1984 |
+
"epoch": 0.8497175141242937,
|
| 1985 |
+
"step": 282
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"loss": 1.3609,
|
| 1989 |
+
"grad_norm": 1.448297142982483,
|
| 1990 |
+
"learning_rate": 5.207791309752947e-05,
|
| 1991 |
+
"epoch": 0.8527306967984934,
|
| 1992 |
+
"step": 283
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"loss": 1.4038,
|
| 1996 |
+
"grad_norm": 1.3868755102157593,
|
| 1997 |
+
"learning_rate": 5.079226780386982e-05,
|
| 1998 |
+
"epoch": 0.855743879472693,
|
| 1999 |
+
"step": 284
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"loss": 1.32,
|
| 2003 |
+
"grad_norm": 1.2606309652328491,
|
| 2004 |
+
"learning_rate": 4.947928319207749e-05,
|
| 2005 |
+
"epoch": 0.8587570621468926,
|
| 2006 |
+
"step": 285
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"loss": 1.3164,
|
| 2010 |
+
"grad_norm": 1.4189478158950806,
|
| 2011 |
+
"learning_rate": 4.8141232279837554e-05,
|
| 2012 |
+
"epoch": 0.8617702448210923,
|
| 2013 |
+
"step": 286
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"loss": 1.4574,
|
| 2017 |
+
"grad_norm": 1.52957284450531,
|
| 2018 |
+
"learning_rate": 4.678043147920822e-05,
|
| 2019 |
+
"epoch": 0.8647834274952919,
|
| 2020 |
+
"step": 287
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"loss": 1.3884,
|
| 2024 |
+
"grad_norm": 1.458547592163086,
|
| 2025 |
+
"learning_rate": 4.5399236586485846e-05,
|
| 2026 |
+
"epoch": 0.8677966101694915,
|
| 2027 |
+
"step": 288
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"loss": 1.3527,
|
| 2031 |
+
"grad_norm": 2.384209394454956,
|
| 2032 |
+
"learning_rate": 4.400003870388859e-05,
|
| 2033 |
+
"epoch": 0.8708097928436912,
|
| 2034 |
+
"step": 289
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"loss": 1.3313,
|
| 2038 |
+
"grad_norm": 1.2834129333496094,
|
| 2039 |
+
"learning_rate": 4.258526010011895e-05,
|
| 2040 |
+
"epoch": 0.8738229755178908,
|
| 2041 |
+
"step": 290
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"loss": 1.3561,
|
| 2045 |
+
"grad_norm": 1.3999826908111572,
|
| 2046 |
+
"learning_rate": 4.11573500169715e-05,
|
| 2047 |
+
"epoch": 0.8768361581920904,
|
| 2048 |
+
"step": 291
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"loss": 1.3505,
|
| 2052 |
+
"grad_norm": 1.3394862413406372,
|
| 2053 |
+
"learning_rate": 3.971878042924502e-05,
|
| 2054 |
+
"epoch": 0.87984934086629,
|
| 2055 |
+
"step": 292
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"loss": 1.2108,
|
| 2059 |
+
"grad_norm": 1.2214335203170776,
|
| 2060 |
+
"learning_rate": 3.82720417653e-05,
|
| 2061 |
+
"epoch": 0.8828625235404897,
|
| 2062 |
+
"step": 293
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"loss": 1.2811,
|
| 2066 |
+
"grad_norm": 1.1912809610366821,
|
| 2067 |
+
"learning_rate": 3.6819638595669306e-05,
|
| 2068 |
+
"epoch": 0.8858757062146893,
|
| 2069 |
+
"step": 294
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"loss": 1.4103,
|
| 2073 |
+
"grad_norm": 1.3451627492904663,
|
| 2074 |
+
"learning_rate": 3.536408529718624e-05,
|
| 2075 |
+
"epoch": 0.8888888888888888,
|
| 2076 |
+
"step": 295
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"loss": 1.2152,
|
| 2080 |
+
"grad_norm": 1.2125860452651978,
|
| 2081 |
+
"learning_rate": 3.390790170013611e-05,
|
| 2082 |
+
"epoch": 0.8919020715630885,
|
| 2083 |
+
"step": 296
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"loss": 1.2711,
|
| 2087 |
+
"grad_norm": 1.2300926446914673,
|
| 2088 |
+
"learning_rate": 3.2453608725966995e-05,
|
| 2089 |
+
"epoch": 0.8949152542372881,
|
| 2090 |
+
"step": 297
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"loss": 1.3299,
|
| 2094 |
+
"grad_norm": 1.245042085647583,
|
| 2095 |
+
"learning_rate": 3.100372402311107e-05,
|
| 2096 |
+
"epoch": 0.8979284369114877,
|
| 2097 |
+
"step": 298
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"loss": 1.3377,
|
| 2101 |
+
"grad_norm": 1.1994699239730835,
|
| 2102 |
+
"learning_rate": 2.956075760847284e-05,
|
| 2103 |
+
"epoch": 0.9009416195856874,
|
| 2104 |
+
"step": 299
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"loss": 1.3519,
|
| 2108 |
+
"grad_norm": 1.3321179151535034,
|
| 2109 |
+
"learning_rate": 2.8127207522128252e-05,
|
| 2110 |
+
"epoch": 0.903954802259887,
|
| 2111 |
+
"step": 300
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"loss": 1.3338,
|
| 2115 |
+
"grad_norm": 1.3435014486312866,
|
| 2116 |
+
"learning_rate": 2.6705555502758335e-05,
|
| 2117 |
+
"epoch": 0.9069679849340866,
|
| 2118 |
+
"step": 301
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"loss": 1.2331,
|
| 2122 |
+
"grad_norm": 1.1831248998641968,
|
| 2123 |
+
"learning_rate": 2.5298262691303318e-05,
|
| 2124 |
+
"epoch": 0.9099811676082863,
|
| 2125 |
+
"step": 302
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"loss": 1.4217,
|
| 2129 |
+
"grad_norm": 1.3309197425842285,
|
| 2130 |
+
"learning_rate": 2.3907765370275384e-05,
|
| 2131 |
+
"epoch": 0.9129943502824859,
|
| 2132 |
+
"step": 303
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"loss": 1.499,
|
| 2136 |
+
"grad_norm": 1.3220206499099731,
|
| 2137 |
+
"learning_rate": 2.2536470746105873e-05,
|
| 2138 |
+
"epoch": 0.9160075329566855,
|
| 2139 |
+
"step": 304
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"loss": 1.2514,
|
| 2143 |
+
"grad_norm": 1.2285231351852417,
|
| 2144 |
+
"learning_rate": 2.118675278182833e-05,
|
| 2145 |
+
"epoch": 0.9190207156308852,
|
| 2146 |
+
"step": 305
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"loss": 1.2162,
|
| 2150 |
+
"grad_norm": 1.1852914094924927,
|
| 2151 |
+
"learning_rate": 1.986094808731247e-05,
|
| 2152 |
+
"epoch": 0.9220338983050848,
|
| 2153 |
+
"step": 306
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"loss": 1.2796,
|
| 2157 |
+
"grad_norm": 1.2502425909042358,
|
| 2158 |
+
"learning_rate": 1.8561351874162783e-05,
|
| 2159 |
+
"epoch": 0.9250470809792843,
|
| 2160 |
+
"step": 307
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"loss": 1.2262,
|
| 2164 |
+
"grad_norm": 1.1900444030761719,
|
| 2165 |
+
"learning_rate": 1.7290213982285547e-05,
|
| 2166 |
+
"epoch": 0.928060263653484,
|
| 2167 |
+
"step": 308
|
| 2168 |
+
},
|
| 2169 |
+
{
|
| 2170 |
+
"loss": 1.41,
|
| 2171 |
+
"grad_norm": 1.2909789085388184,
|
| 2172 |
+
"learning_rate": 1.6049734985002335e-05,
|
| 2173 |
+
"epoch": 0.9310734463276836,
|
| 2174 |
+
"step": 309
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"loss": 1.33,
|
| 2178 |
+
"grad_norm": 1.1685523986816406,
|
| 2179 |
+
"learning_rate": 1.4842062379453172e-05,
|
| 2180 |
+
"epoch": 0.9340866290018832,
|
| 2181 |
+
"step": 310
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"loss": 1.1855,
|
| 2185 |
+
"grad_norm": 1.1228547096252441,
|
| 2186 |
+
"learning_rate": 1.3669286868884368e-05,
|
| 2187 |
+
"epoch": 0.9370998116760829,
|
| 2188 |
+
"step": 311
|
| 2189 |
+
},
|
| 2190 |
+
{
|
| 2191 |
+
"loss": 1.3196,
|
| 2192 |
+
"grad_norm": 1.1882398128509521,
|
| 2193 |
+
"learning_rate": 1.2533438743256715e-05,
|
| 2194 |
+
"epoch": 0.9401129943502825,
|
| 2195 |
+
"step": 312
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"loss": 1.3567,
|
| 2199 |
+
"grad_norm": 1.336132287979126,
|
| 2200 |
+
"learning_rate": 1.1436484364440646e-05,
|
| 2201 |
+
"epoch": 0.9431261770244821,
|
| 2202 |
+
"step": 313
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"loss": 1.307,
|
| 2206 |
+
"grad_norm": 1.1312459707260132,
|
| 2207 |
+
"learning_rate": 1.038032276208234e-05,
|
| 2208 |
+
"epoch": 0.9461393596986817,
|
| 2209 |
+
"step": 314
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"loss": 1.2926,
|
| 2213 |
+
"grad_norm": 1.1922824382781982,
|
| 2214 |
+
"learning_rate": 9.366782346034312e-06,
|
| 2215 |
+
"epoch": 0.9491525423728814,
|
| 2216 |
+
"step": 315
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"loss": 1.1762,
|
| 2220 |
+
"grad_norm": 1.2577030658721924,
|
| 2221 |
+
"learning_rate": 8.397617741041943e-06,
|
| 2222 |
+
"epoch": 0.952165725047081,
|
| 2223 |
+
"step": 316
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"loss": 1.3479,
|
| 2227 |
+
"grad_norm": 1.1837351322174072,
|
| 2228 |
+
"learning_rate": 7.474506749165632e-06,
|
| 2229 |
+
"epoch": 0.9551789077212806,
|
| 2230 |
+
"step": 317
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"loss": 1.3444,
|
| 2234 |
+
"grad_norm": 1.1284111738204956,
|
| 2235 |
+
"learning_rate": 6.5990474451968436e-06,
|
| 2236 |
+
"epoch": 0.9581920903954803,
|
| 2237 |
+
"step": 318
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"loss": 1.3265,
|
| 2241 |
+
"grad_norm": 1.2326773405075073,
|
| 2242 |
+
"learning_rate": 5.772755410097178e-06,
|
| 2243 |
+
"epoch": 0.9612052730696798,
|
| 2244 |
+
"step": 319
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"loss": 1.334,
|
| 2248 |
+
"grad_norm": 1.3088549375534058,
|
| 2249 |
+
"learning_rate": 4.997061107248992e-06,
|
| 2250 |
+
"epoch": 0.9642184557438794,
|
| 2251 |
+
"step": 320
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"loss": 1.3848,
|
| 2255 |
+
"grad_norm": 1.23444402217865,
|
| 2256 |
+
"learning_rate": 4.273307406060457e-06,
|
| 2257 |
+
"epoch": 0.9672316384180791,
|
| 2258 |
+
"step": 321
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"loss": 1.2999,
|
| 2262 |
+
"grad_norm": 1.176615834236145,
|
| 2263 |
+
"learning_rate": 3.602747257211758e-06,
|
| 2264 |
+
"epoch": 0.9702448210922787,
|
| 2265 |
+
"step": 322
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"loss": 1.3359,
|
| 2269 |
+
"grad_norm": 1.2764966487884521,
|
| 2270 |
+
"learning_rate": 2.986541523567167e-06,
|
| 2271 |
+
"epoch": 0.9732580037664783,
|
| 2272 |
+
"step": 323
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"loss": 1.3129,
|
| 2276 |
+
"grad_norm": 1.257296085357666,
|
| 2277 |
+
"learning_rate": 2.4257569705081394e-06,
|
| 2278 |
+
"epoch": 0.976271186440678,
|
| 2279 |
+
"step": 324
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"loss": 1.3223,
|
| 2283 |
+
"grad_norm": 1.2974607944488525,
|
| 2284 |
+
"learning_rate": 1.9213644191662393e-06,
|
| 2285 |
+
"epoch": 0.9792843691148776,
|
| 2286 |
+
"step": 325
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"loss": 1.2993,
|
| 2290 |
+
"grad_norm": 1.174402117729187,
|
| 2291 |
+
"learning_rate": 1.4742370657534552e-06,
|
| 2292 |
+
"epoch": 0.9822975517890772,
|
| 2293 |
+
"step": 326
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"loss": 1.2605,
|
| 2297 |
+
"grad_norm": 1.1533863544464111,
|
| 2298 |
+
"learning_rate": 1.0851489698990146e-06,
|
| 2299 |
+
"epoch": 0.9853107344632769,
|
| 2300 |
+
"step": 327
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"loss": 1.333,
|
| 2304 |
+
"grad_norm": 1.2788079977035522,
|
| 2305 |
+
"learning_rate": 7.547737146098809e-07,
|
| 2306 |
+
"epoch": 0.9883239171374765,
|
| 2307 |
+
"step": 328
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"loss": 1.3286,
|
| 2311 |
+
"grad_norm": 1.171959638595581,
|
| 2312 |
+
"learning_rate": 4.836832401747854e-07,
|
| 2313 |
+
"epoch": 0.9913370998116761,
|
| 2314 |
+
"step": 329
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"loss": 1.2699,
|
| 2318 |
+
"grad_norm": 1.209529995918274,
|
| 2319 |
+
"learning_rate": 2.723468540304763e-07,
|
| 2320 |
+
"epoch": 0.9943502824858758,
|
| 2321 |
+
"step": 330
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"loss": 1.3127,
|
| 2325 |
+
"grad_norm": 1.2302489280700684,
|
| 2326 |
+
"learning_rate": 1.211304183042494e-07,
|
| 2327 |
+
"epoch": 0.9973634651600753,
|
| 2328 |
+
"step": 331
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"loss": 1.2107,
|
| 2332 |
+
"grad_norm": 1.2124124765396118,
|
| 2333 |
+
"learning_rate": 3.029571643941353e-08,
|
| 2334 |
+
"epoch": 1.0,
|
| 2335 |
+
"step": 332
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"train_runtime": 8740.1695,
|
| 2339 |
+
"train_samples_per_second": 0.608,
|
| 2340 |
+
"train_steps_per_second": 0.038,
|
| 2341 |
+
"total_flos": 1.0339169907188122e+17,
|
| 2342 |
+
"train_loss": 1.4289702198591576,
|
| 2343 |
+
"epoch": 1.0,
|
| 2344 |
+
"step": 332
|
| 2345 |
+
}
|
| 2346 |
+
],
|
| 2347 |
+
"best_metric": null,
|
| 2348 |
+
"best_global_step": null,
|
| 2349 |
+
"best_model_checkpoint": null,
|
| 2350 |
+
"is_local_process_zero": true,
|
| 2351 |
+
"is_world_process_zero": true,
|
| 2352 |
+
"is_hyper_param_search": false,
|
| 2353 |
+
"trial_name": null,
|
| 2354 |
+
"trial_params": null,
|
| 2355 |
+
"stateful_callbacks": {
|
| 2356 |
+
"TrainerControl": {
|
| 2357 |
+
"args": {
|
| 2358 |
+
"should_training_stop": true,
|
| 2359 |
+
"should_epoch_stop": false,
|
| 2360 |
+
"should_save": true,
|
| 2361 |
+
"should_evaluate": false,
|
| 2362 |
+
"should_log": false
|
| 2363 |
+
},
|
| 2364 |
+
"attributes": {}
|
| 2365 |
+
}
|
| 2366 |
+
}
|
| 2367 |
+
}
|