pytorch / pytorch/pytorch

DISABLED test_sdpa_rewriter_16_fp32_mask_gpu (__main__.SDPAPatternRewriterGpuTests)

Open
#196,406 1 comment 0 reactions 0 assignees View on GitHub
module: flaky-tests module: inductor module: rocm oncall: pt2 skipped triaged
Dominant language
Python
Stars
103k
Forks
29.5k
PR merge metrics
PR metrics pending

Description

Platforms: rocm

This test was disabled because it is failing in CI. See [recent examples](https://hud.pytorch.org/flakytest?name=test_sdpa_rewriter_16_fp32_mask_gpu&suite=SDPAPatternRewriterGpuTests&limit=100) and the most recent trunk [workflow logs](https://github.com/pytorch/pytorch/runs/102304372394).

Over the past 6 hours, it has been determined flaky in 3 workflow(s) with 3 failures and 3 successes.

**Debugging instructions (after clicking on the recent samples link):**
DO NOT ASSUME THINGS ARE OKAY IF THE CI IS GREEN. We now shield flaky tests from developers so CI will thus be green but it will be harder to parse the logs.
To find relevant log snippets:
1. Click on the workflow logs linked above
2. Click on the Test step of the job so that it is expanded. Otherwise, the grepping will not work.
3. Grep for `test_sdpa_rewriter_16_fp32_mask_gpu`
4. There should be several instances run (as flaky tests are rerun in CI) from which you can study the logs.

Sample error message

```
Traceback (most recent call last):
File "/var/lib/jenkins/workspace/test/inductor/test_fused_attention.py", line 1200, in _test_sdpa_rewriter_16_fp32_mask
self._check_common(
File "/var/lib/jenkins/workspace/test/inductor/test_fused_attention.py", line 127, in _check_common
self.assertGreaterEqual(counters["inductor"]["fuse_attention"], 1)
File "/opt/conda/envs/py_3.10/lib/python3.10/unittest/case.py", line 1250, in assertGreaterEqual
self.fail(self._formatMessage(msg, standardMsg))
File "/opt/conda/envs/py_3.10/lib/python3.10/unittest/case.py", line 675, in fail
raise self.failureException(msg)
AssertionError: 0 not greater than or equal to 1

To execute this test, run the following from the base repo dir:
PYTORCH_TEST_WITH_ROCM=1 python test/inductor/test_fused_attention.py SDPAPatternRewriterGpuTests.test_sdpa_rewriter_16_fp32_mask_gpu

This message can be suppressed by setting PYTORCH_PRINT_REPRO_ON_FAILURE=0
```

Test file path: `inductor/test_fused_attention.py`

For all disabled tests (by GitHub issue), see https://hud.pytorch.org/disabled.

cc @jeffdaily @sunway513 @jithunnair-amd @pruthvistony @ROCmSupport @jataylo @hongxiayang @naromero77amd @pragupta @jerrymannil @xinyazhang @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @kadeng @muchulee8 @amjames @chauhang @aakhundov @coconutruben

Contributor guide

Open the contributing guide

Research direction

Start with inductor/test_fused_attention.py, especially test_sdpa_rewriter_16_fp32_mask_gpu and _check_common. Run PYTORCH_TEST_WITH_ROCM=1 python test/inductor/test_fused_attention.py SDPAPatternRewriterGpuTests.test_sdpa_rewriter_16_fp32_mask_gpu, then inspect the linked workflow logs and repeated test output for why fuse_attention remains zero. Done means the test reliably passes on ROCm and can be re-enabled without flaky failures.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning, testing-qa
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Active
Clarity
Mostly clear
Newbie friendliness
52/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.