version issue

[MorganRong]

when I was running
python inference-basic.py
--model_type codellama
--base_model codellama/CodeLlama-13b-hf
--tuned_model odellama-13b-multi-r16
--data_file ../dataset/ReVeal/test_512.json
there are errors
File "/root/miniconda3/lib/python3.10/site-packages/peft/config.py", line 203, in _get_peft_type
raise ValueError(f"Can't find '{CONFIG_NAME}' at '{model_id}'")
ValueError: Can't find 'adapter_config.json' at 'odellama-13b-multi-r16'
Can you offer the peft and transformers version used by you ?

[happy-Superman]

maybe because of the network problem，we can not directly download from hf, but you can with the hf-mirror to download the codellama to your local

[MorganRong]

I searched on https://huggingface.co/. There is no odellama-13b-multi-r16 . Whether the source is still available?

[happy-Superman]

I searched on https://huggingface.co/. There is no odellama-13b-multi-r16 . Whether the source is still available?

codellama-13b-multi-r16 is the output of finetuning, so you need to finished the train proccess, after that you will get the codellama-13b-multi-r16

If you want to get the basic module， maybe you can try the follow steps：

# mirror
export HF_ENDPOINT=https://hf-mirror.com
#get tools
pip install huggingface_hub
#login
huggingface-cli login
#token
you need to get your own token in  https://huggingface.co/
#download
huggingface-cli download codellama/CodeLlama-13b-hf --local-dir CodeLlama-13b-hf