Remove triton mlp kernel, not compiling for some models #37449
Conversation
|
Hi 👋, thank you for opening this pull request! The pull request is converted to draft by default. The CI will be paused while the PR is in draft mode. When it is ready for review, please click the |
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
…37449) * remove mlp for now * disable on docker
…37449) * remove mlp for now * disable on docker
|
@MekkCyber @ArthurZucker Hi, regarding the mlp kernel, is there a plan to fix it and re-enable it via use_kernel_forward_from_hub? I noticed that the relevant code hasn't been removed yet. |
|
Hey @zheliuyu, good catch maybe we should remove it from the dict for now, and yes sure we are trying to optimize our modeling and using an mlp kernel is a good option that we will re-enable once we have a good kernel |
What does this PR do?
Removing the mlp triton kernel because of some failures on the CI, and disabling the use of kernels on the CI