Model Size Mismatch

Hello everyone,

I am trying to convert the llama model weight into huggyingface format. I have downloaded the llama model weight with the following format:

llama
|- llama-2-7b
|- llama-2-7b-chat
|- …

and then I run the src/transformers/models/llama/convert_llama_weights_to_hf.py script to convert the weight, with the input “–model_size 7B”.

Then I got the following error “FileNotFoundError: [Errno 2] No such file or directory: ‘/usr/llama/7B/params.json’”.

I think the correct path should be ‘/usr/llama/llama-2-7b/params.json’.

May I ask how can I solve this issue?

Thank you.