[bugfix]fix the cut-off issue due to LLM predict token limit(256 for openai python lib default), by setting temperature to 0 and set LLM predict method from compact-refine to refine
bd59653
NickNYU
commited on