Add a check and error message for no support on MPS for conv with output_channels > 2^16 #129484
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/129484
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit fbe5e15 with merge base 7cf0b90 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
pytorch-bot
bot
added
ciflow/mps
|
Looks good, @jhavukainen can you add a test for it ? |
jhavukainen
force-pushed
the
dev/joona/Conv1DErrorMessage
branch
from
49213d3 to
fbe5e15
Compare
|
@pytorchbot merge |
Merge startedYour 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 |
| "As a temporary fix, you can set the environment variable `PYTORCH_ENABLE_MPS_FALLBACK=1` ", | ||
| "to use the CPU as a fallback for this op. WARNING: this will be slower than running natively ", | ||
| "on MPS."); |
There was a problem hiding this comment.
@jhavukainen bailing out here doesn't work because the op is already implemented and dispatched into the kernel, and we don't read the environment variable anymore at this point. Ref: #134416
There was a problem hiding this comment.
Yeah you are correct. Seem to have lost my train of thought there. Then the correct thing is to just raise the error that this is unimplemented as of now without the instruction to set the fallback env variable.
Fixes the silent correctness issue in #129207 by preventing the user from calling the convolution op on MPS device with an unsupported value.
The fix for the missing support is coming in later as that requires work on the kernel side so it'll take some more time.