fix(langchain): handle prompt_tokens_details as dict in _parse_usage_model #1614
background
wait
wait-all
cancel
Loading