-
-
Notifications
You must be signed in to change notification settings - Fork 5k
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
[BugFix] Stop silent failures on compressed-tensors parsing #9381
Conversation
👋 Hi! Thank you for contributing to the vLLM project. Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging. To run CI, PR reviewers can do one of these:
🚀 |
vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors.py
Show resolved
Hide resolved
@dsikka I think there is a real issue in the CI here https://buildkite.com/vllm/ci-aws/builds/9886#0192911e-1065-4a85-8a1b-75071e80f1de/193-715 |
@mgoin Yup, which is why we need the PR which adds compressed-tensors to land. The current utils is out-of-date. This PR is secondary to make future failures more visible/not hide the root cause (i.e the issue you tagged is the real issue, which was silent before this PR) |
8e2c5d1
to
a6b6121
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
🤞
6e80f8a
to
bbacbbd
Compare
…ject#9381) Signed-off-by: charlifu <[email protected]>
…ject#9381) Signed-off-by: Vinay Damodaran <[email protected]>
…ject#9381) Signed-off-by: Alvant <[email protected]>
…ject#9381) Signed-off-by: Amit Garg <[email protected]>
…ject#9381) Signed-off-by: qishuai <[email protected]>
…ject#9381) Signed-off-by: Sumit Dubey <[email protected]>
…ject#9381) Signed-off-by: Maxime Fournioux <[email protected]>
…ject#9381) Signed-off-by: Tyler Michael Smith <[email protected]>
Summary
input_activations
to None in those cases, which results in downstream failures that hide the root probleminput_quant