Moving device_put for the kernel earlier to avoid a host convert op #1824
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
zhenying-liu
requested review from
gobbleturk,
khatwanimohit,
bvandermoon,
vipannalla,
RissyRan,
richjames0,
gagika,
shralex,
yangyuwei,
SurbhiJainUSC,
hengtaoguo,
A9isha and
aireenmei
as code owners
gobbleturk
approved these changes
Jun 12, 2025
zhenying-liu
force-pushed
the
param-offload-fix
branch
from
3c97d7e to
21a74da
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fix a bug with parameter_memory_host_offload: "Found an instruction ("convert.610") which does device compute in host memory space." The issue was caused by self.kernel in linear.py being on the host and undergoing a data type conversion before being moved to the device. This commit moves device_put() earlier to resolve the issue.