Skip to content

Conversation

@xiangzez
Copy link
Contributor

This PR supports loading int4 model weights for int4 datatype. Currently it supports llama2 model and is compatible with AutoGPTQ models (configuration: wbits=4 and group_size=-1).

@xiangzez xiangzez changed the title Support loading int4 weights [Common] Support loading int4 weights Apr 3, 2024
@changqi1 changqi1 merged commit fcd3fb4 into intel:main Apr 3, 2024
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

Successfully merging this pull request may close these issues.

2 participants