Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

use half/bf16 lm_head output #3213

Open
wants to merge 6 commits into
base: main
Choose a base branch
from
Open

use half/bf16 lm_head output #3213

wants to merge 6 commits into from

Conversation

irexyc
Copy link
Collaborator

@irexyc irexyc commented Mar 5, 2025

Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily receiving feedbacks. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers.

Motivation

Please describe the motivation of this PR and the goal you want to achieve through this PR.

Modification

Please briefly describe what modification is made in this PR.

BC-breaking (Optional)

Does the modification introduce changes that break the backward-compatibility of the downstream repositories?
If so, please describe how it breaks the compatibility and how the downstream projects should modify their code to keep compatibility with this PR.

Use cases (Optional)

If this PR introduces a new feature, it is better to list some use cases here, and update the documentation.

Checklist

  1. Pre-commit or other linting tools are used to fix the potential lint issues.
  2. The modification is covered by complete unit tests. If not, please add more unit tests to ensure the correctness.
  3. If the modification has a dependency on downstream projects of a newer version, this PR should be tested with all supported versions of downstream projects.
  4. The documentation has been modified accordingly, like docstring or example tutorials.

irexyc added 4 commits March 5, 2025 03:29

Verified

This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
@lzhangzz
Copy link
Collaborator

Do we still need the float instantiation for sampling kernels?

@irexyc
Copy link
Collaborator Author

irexyc commented Mar 17, 2025

Do we still need the float instantiation for sampling kernels?

some tests in tests/csrc/unittests use float inputs.

@lzhangzz
Copy link
Collaborator

lzhangzz commented Mar 21, 2025

Use the macro ENABLE_FP32 to skip the tests. We are not increasing the binary size for testing unused kernels. Or better, modify the tests to support half/bf16 data type.

@lvhan028 lvhan028 requested review from lvhan028 and lzhangzz March 21, 2025 11:16
@irexyc
Copy link
Collaborator Author

irexyc commented Mar 21, 2025

After removing the fp32 compilation option, the size is reduced by about 2MB.

To reduce the size, maybe we can use TORCH_LIBRARY to replace pybind11 and enable py-limited-api. If it works, there is no need to compile multiple versions of Python packages any more.

@lvhan028
Copy link
Collaborator

is TORCH_LIBRARY highly related to torch version?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants