Introduce a new API torch.xpu.set_per_process_memory_fraction

[guangyey]

Stack from ghstack (oldest at bottom):

• Introduce a new API torch.xpu.get_per_process_memory_fraction #165511
• -> Introduce a new API torch.xpu.set_per_process_memory_fraction #165510
• Refine XPU allocator message when OOM #165509
• Reuse kLargeBuffer in XPUCachingAllocator #165508

Motivation

Aligned with other backends, this PR introduces a new API torch.xpu.set_per_process_memory_fraction to allow user to customize the allowed memory per a single process.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/165510

• 📄 Preview Python docs built from this PR
• 📄 Preview C++ docs built from this PR
• ❓ Need help or want to give feedback on the CI? Visit the bot commands wiki

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit d97ccf1 with merge base 839f6fa ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[guangyey]

@ezyang @albanD May I know if you have any concerns about this stack of PRs.

[ezyang]

@pytorchbot merge

[pytorchmergebot]

Merge started

Your change will be merged once all checks pass (ETA 0-4 Hours).

Learn more about merging in the wiki.

Questions? Feedback? Please reach out to the PyTorch DevX Team

Advanced Debugging

Check the merge workflow status
here