[MPS] Fix MPS clamp issue with different dtypes between input and min/max tensors

[DenisVieriu97]

• Fix the FP16 clamp issue (FP32 and FP16 are not broadcast compatible)
• Fix clamp (cached graph nodes were previously replaced with the cast version)

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/105747

• 📄 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 or our office hours

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

❌ 1 New Failure

As of commit 1eae4c3:

NEW FAILURE - The following job has failed:

• linux-focal-cuda12.1-py3.10-gcc9-bazel-test / build-and-test (default, 1, 1, linux.4xlarge.nvidia.gpu) (gh)

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

[DenisVieriu97]

@pytorchbot rebase

[pytorchmergebot]

@pytorchbot started a rebase job onto refs/remotes/origin/viable/strict. Check the current status here

[pytorchmergebot]

Successfully rebased clamp_fix onto refs/remotes/origin/viable/strict, please pull locally before adding more changes (for example, via git checkout clamp_fix && git pull --rebase)

[DenisVieriu97]

@pytorchbot merge -f "MPS tests & lint are green"

[pytorchmergebot]

Merge started

Your change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use -f as last resort and instead consider -i/--ignore-current to continue the merge ignoring current failures. This will allow currently pending tests to finish and report signal before the merge.

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

[malfet]

@kulinseth , @DenisVieriu97 just curious, have you check if it matches cpu/CUDA behavior? I.e. I assume it should do something like that only if AMP is enable, isn't it?

[kulinseth]

@kulinseth , @DenisVieriu97 just curious, have you check if it matches cpu/CUDA behavior? I.e. I assume it should do something like that only if AMP is enable, isn't it?

We checked with cpu @malfet

[kulinseth]

@kulinseth , @DenisVieriu97 just curious, have you check if it matches cpu/CUDA behavior? I.e. I assume it should do something like that only if AMP is enable, isn't it?

We checked with cpu @malfet. Also it’s affecting correctness of few language models