ColossalAI/applications/Chat/evaluate/evaluate.sh

10 lines
342 B
Bash
Executable File

python evaluate.py \
--answer_file_list "path to answers of model 1" "path to answers of model 2" \
--prompt_file "path to prompt file" \
--reviewer_file "path to reviewer file" \
--output_folder "path to output folder" \
--openai_key "your openai key" \
--model "gpt-4" \
--num_workers 8 \
--max_tokens 512 \