[XPU] support op builder from intel_extension_for_pytorch kernel path #5425
Conversation
|
@loadams @mrwyattii Hi, could you help to trigger a CI for this PR? thanks! |
Done @YizhouZ - could you run the pre-commit formatter to pass the formatting check? Thanks |
Fixed! Could you run it again? Thanks. |
tjruwase
requested review from
duli2012 and
tjruwase
and removed request for
mrwyattii
| return None | ||
| def create_op_builder(self, class_name): | ||
| builder_class = self.get_op_builder(class_name) | ||
| return None if builder_class is None else builder_class() |
There was a problem hiding this comment.
builder_class could return NotImplementedBuilder or None. It's error prone to check between the two.
thanks! could you help to trigger CI? |
#Motivation
From our next release, xpu DeepSpeed related kernels would be put into intel_extension_for_pytorch. This PR is to add new op builders and use kernel path from intel_extension_for_pytorch. More ops like MOE and WOQ will be added.