search

AGI-Edgerunners / LLM-Adapters

Code for our EMNLP 2023 Paper: "LLM-Adapters: An Adapter Family for Parameter-Efficient Fine-Tuning of Large Language Models"
https://arxiv.org/abs/2304.01933
Apache License 2.0
1.01k stars 91 forks source link

Eval without Tuning/Using OPT-1.3B #34

Open ChaoGaoUCR opened 11 months ago

ChaoGaoUCR commented 11 months ago

Dear Author,

Thanks for your great projects. I was trying to evaluate the model without Tuning and with Tuning. I wondered if we can evaluate the model with the original model. Also, if I want to use models except LLAMA Bloom and GPT-J, do I have to write my own part?

Thanks

HZQ950419 commented 11 months ago

Hi,

Yes, you can evaluate the original models by commenting Line 222-227 in evaluate.py. And if the model you wanna use has already been supported, you can just indicate the argument --base_model. If not, then you need to indicate the argument --target_modules or add the mapping of the model to target modules in LLM-Adapters/peft/src/peft/mapping.py.

If you have any questions on add unsupported models to the code base, please let us know and we will help with it!

Thanks!

ChaoGaoUCR commented 11 months ago

Thank you so much! I will try this out😃