Skip to content

Conversation

@Qubitium
Copy link
Collaborator

@jiqing-feng This PR will make the kernels report the correct names and remove the fakes name previously reported so we can pass transformer, peft, optimum ci tests. I will be submitting parallel PRs to them due to large changes in GPT-QModel v5.0.0 which will remove all legacy autogptq code and bad ci check for QUANT-TYPE as torch. They should never check what kernel is returned, unless they directly asked for a specific kernel, and just use whatever we give them.

Signed-off-by: Qubitium <Qubitium@modelcloud.ai>
@Qubitium Qubitium merged commit a3bdb20 into main Oct 14, 2025
4 checks passed
@Qubitium Qubitium deleted the qlinear-names branch October 14, 2025 07:01
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