Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[REQUEST] EXAONE 3.5 Support #695

Open
3 tasks done
necrogay opened this issue Dec 9, 2024 · 2 comments
Open
3 tasks done

[REQUEST] EXAONE 3.5 Support #695

necrogay opened this issue Dec 9, 2024 · 2 comments

Comments

@necrogay
Copy link

necrogay commented Dec 9, 2024

Problem

LG AI Research has released new EXAONE 3.5 models in sizes 2.4B, 7.8B, and 32B:

https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct
https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-7.8B-Instruct
https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-32B-Instruct

Unfortunately, exllamav2 does not currently support these models:

!! Warning, unknown architecture: ExaoneForCausalLM

Solution

Could you please consider adding support for these models?

Alternatives

No response

Explanation

Examples

No response

Additional context

No response

Acknowledgements

  • I have looked for similar requests before submitting this one.
  • I understand that the developers have lives and my issue will be answered when possible.
  • I understand the developers of this program are human, and I will make my requests politely.
@Nrgte
Copy link

Nrgte commented Dec 13, 2024

I can confirm, I tried to play around with it and can't quantize it:

raise ValueError(f" ## Could not find {prefix}.* in model")

ValueError: ## Could not find model.norm.* in model

@turboderp
Copy link
Member

turboderp commented Dec 14, 2024

It's one of those models that implements a Llama architecture but renames everything so it isn't compatible with existing frameworks. There are a number of "llamafied" versions on HF that you can use here including EXL2 quants.

I will add support for the naming scheme they use at some point, but I'm not yet sure if they're sticking with the scheme going forward.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants