don't error out in empty_cache under mempool context#158152
don't error out in empty_cache under mempool context#158152
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/158152
Note: Links to docs will display an error until the docs builds have been completed. ✅ You can merge normally! (1 Unrelated Failure)As of commit 0be0650 with merge base 1f1f229 ( UNSTABLE - The following job is marked as unstable, possibly due to flakiness on trunk:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
@pytorchbot merge -i |
Merge startedYour change will be merged while ignoring the following 5 checks: pull / linux-jammy-py3.9-clang12 / build, pull / linux-jammy-cuda12.8-py3.10-gcc11-sm89 / build, pull / cuda12.8-py3.10-gcc9-sm75 / test (pr_time_benchmarks, 1, 1, linux.g4dn.metal.nvidia.gpu, unstable), Lint / lintrunner-noclang / linux-job, Lint / toc / linux-job Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
|
@pytorchbot --help |
PyTorchBot HelpMergeRevertRebaseLabelDr CIcherry-pick |
|
@pytorchbot cherry-pick --onto release/2.8 --fixes vllm-project/vllm#20358 -c regression |
Now instead of erroring out on `empty_cache` call during graph capture or under mempool context, we will just silently do nothing. This used to be the behavior for mempools, cudagraphs used to error out, but it's fine to just ignore the call. Pull Request resolved: #158152 Approved by: https://github.com/zou3519, https://github.com/eqy (cherry picked from commit 9056279)
Cherry picking #158152The cherry pick PR is at #158180 and it is linked with issue vllm-project/vllm#20358. The following tracker issues are updated: Details for Dev Infra teamRaised by workflow job |
don't error out in empty_cache under mempool context (#158152) Now instead of erroring out on `empty_cache` call during graph capture or under mempool context, we will just silently do nothing. This used to be the behavior for mempools, cudagraphs used to error out, but it's fine to just ignore the call. Pull Request resolved: #158152 Approved by: https://github.com/zou3519, https://github.com/eqy (cherry picked from commit 9056279) Co-authored-by: Natalia Gimelshein <ngimel@meta.com>
Now instead of erroring out on
empty_cachecall during graph capture or under mempool context, we will just silently do nothing. This used to be the behavior for mempools, cudagraphs used to error out, but it's fine to just ignore the call.