[Intel XPU] Enable test/quantization UTs on XPU #3249

DiweiSun · 2025-10-27T09:04:27Z

This PR is targeted to enable the CI for the test/quantization on XPU device.

This reverts commit a736c41.

pytorch-bot · 2025-10-27T09:04:31Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/3249

📄 Preview Python docs built from this PR

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

❌ 3 New Failures

As of commit a9043e2 with merge base ba3ac9f ():

NEW FAILURES - The following jobs have failed:

Run Regression Tests / test (CUDA 2.6, linux.g5.12xlarge.nvidia.gpu, torch==2.6.0, cuda, 12.6) / linux-job (gh)
test/quantization/test_quant_primitives.py::TestQuantPrimitives::test_get_group_qparams_symmetric_memory
Run Regression Tests / test (CUDA 2.7, linux.g5.12xlarge.nvidia.gpu, torch==2.7.0, cuda, 12.6) / linux-job (gh)
test/quantization/test_quant_primitives.py::TestQuantPrimitives::test_get_group_qparams_symmetric_memory
Run Regression Tests / test (CUDA 2.8, linux.g5.12xlarge.nvidia.gpu, torch==2.8.0, cuda, 12.6) / linux-job (gh)
test/quantization/test_quant_primitives.py::TestQuantPrimitives::test_get_group_qparams_symmetric_memory

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

pytorch-bot · 2025-10-27T12:04:53Z

To add the ciflow label ciflow/xpu please first approve the workflows that are awaiting approval (scroll to the bottom of this page).

This helps ensure we don't trigger CI on this PR until it is actually authorized to do so. Please ping one of the reviewers if you do not have access to approve and run workflows.

pytorch-bot · 2025-11-04T09:51:25Z

To add the ciflow label ciflow/xpu please first approve the workflows that are awaiting approval (scroll to the bottom of this page).

This helps ensure we don't trigger CI on this PR until it is actually authorized to do so. Please ping one of the reviewers if you do not have access to approve and run workflows.

jerryzh168 · 2025-11-06T00:37:54Z

torchao/utils.py

    return devices


+def auto_detect_device():


does this include cuda?

jerryzh168 · 2025-11-06T00:40:58Z

test/quantization/test_gptq.py


 torch.manual_seed(0)

+_DEVICE = auto_detect_device()


auto_detect_device seems to be changing what we want to test, I think previous we only want to test on CUDA, can you preserve this?

We have refine the auto_detect_device functions and cpu will not be included.

jerryzh168 · 2025-11-06T00:41:16Z

test/quantization/test_gptq.py


 class TestGPTQ(TestCase):
    @unittest.skip("skipping until we get checkpoints for gpt-fast")
-    @unittest.skipIf(not torch.cuda.is_available(), "Need CUDA available")


just change this to torch.accelerator.is_available()?

jerryzh168 · 2025-11-06T00:41:50Z

test/quantization/test_gptq.py



 class TestMultiTensorFlow(TestCase):
-    @unittest.skipIf(not torch.cuda.is_available(), "Need CUDA available")


we don't want to expand test to cpu I think

pytorch-bot · 2025-11-06T02:53:00Z

To add the ciflow label ciflow/xpu please first approve the workflows that are awaiting approval (scroll to the bottom of this page).

This helps ensure we don't trigger CI on this PR until it is actually authorized to do so. Please ping one of the reviewers if you do not have access to approve and run workflows.

jerryzh168 · 2025-11-07T05:01:45Z

test/quantization/test_quant_api.py

        m2.load_state_dict(state_dict)
        m2 = m2.to(device="cuda")
-        example_inputs = map(lambda x: x.cuda(), example_inputs)
+        example_inputs = map(lambda x: x.to(_DEVICE), example_inputs)


so when in CPU, _DEVICE will be None now, what happens when we do x.to(None)? I think we don't want auto detect here, since in L267, it is converting model m2 to "cuda"?

changed m2.to(_DEVICE), as it is skip if not torch.accelerator.available(), when in CPU, this case will be skipped

jerryzh168 · 2025-11-07T05:04:47Z

test/quantization/test_quant_primitives.py

    def test_get_group_qparams_symmetric_memory(self):
        """Check the memory usage of the op"""
-        weight = torch.randn(1024, 1024).to(device="cuda")
+        weight = torch.randn(1024, 1024).to(device=_DEVICE)


this test also has skip if no cuda, so it is still only going to run in cuda right? the change doesn't seem to have any effect right now

changed skip if no cuda to skip if not torch.accelerator.available().

jerryzh168 · 2025-11-07T05:06:35Z

test/quantization/test_quant_api.py

I think it might make sense if we change all the skip if no cuda to skip if not torch.accelerator.available() in this file

test_module_fqn_to_config_regex_basic, test_module_fqn_to_config_regex_fullmatch, test_module_fqn_to_config_regex_precedence and test_module_fqn_to_config_regex_precedence2 are not ready for XPU, others are changed to skip if not torch.accelerator.available()

DiweiSun and others added 4 commits August 19, 2025 08:28

enable xpu ci test

a736c41

Revert "enable xpu ci test"

7c96ad4

This reverts commit a736c41.

Merge branch 'pytorch:main' into main

dde6d27

enabel quantiation ut cases in xpu ci

9d1cc1f

meta-cla bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Oct 27, 2025

liangan1 added the ciflow/xpu label used to trigger xpu CI jobs label Oct 27, 2025

pytorch-bot bot removed the ciflow/xpu label used to trigger xpu CI jobs label Oct 27, 2025

liangan1 added topic: improvement Use this tag if this PR is an improvement (doesn't fit into any of the other categories) ciflow/xpu label used to trigger xpu CI jobs labels Oct 27, 2025

fix xpu int4 ut cases

e54bda3

pytorch-bot bot removed the ciflow/xpu label used to trigger xpu CI jobs label Oct 28, 2025

zxd1997066 and others added 2 commits November 4, 2025 16:30

debug RuntimeError

83afd19

Merge branch 'main' into molly/quantization_ut

463cf5a

xiaowangintel added the ciflow/xpu label used to trigger xpu CI jobs label Nov 4, 2025

revert debug

9c8e66b

pytorch-bot bot removed the ciflow/xpu label used to trigger xpu CI jobs label Nov 4, 2025

liangan1 added the ciflow/xpu label used to trigger xpu CI jobs label Nov 4, 2025

pytorch-bot bot removed the ciflow/xpu label used to trigger xpu CI jobs label Nov 4, 2025

liangan1 requested a review from jerryzh168 November 5, 2025 01:16

liangan1 added the ciflow/xpu label used to trigger xpu CI jobs label Nov 5, 2025

liangan1 approved these changes Nov 6, 2025

View reviewed changes

jerryzh168 reviewed Nov 6, 2025

View reviewed changes

This was referenced Nov 6, 2025

[Intel GPU] Extend TestQAT module with xpu testcases #3177

Open

Intel XPU CI Enabling #2917

Open

pytorch-bot bot removed the ciflow/xpu label used to trigger xpu CI jobs label Nov 6, 2025

refine the device

7b5d2c4

zxd1997066 force-pushed the molly/quantization_ut branch from cfe14e2 to 7b5d2c4 Compare November 6, 2025 02:42

liangan1 added xpu Intel XPU related features ciflow/xpu label used to trigger xpu CI jobs labels Nov 6, 2025

pytorch-bot bot removed the ciflow/xpu label used to trigger xpu CI jobs label Nov 6, 2025

liangan1 changed the title ~~Molly/quantization ut~~ [Intel XPU] Enable test/quantization UTs on XPU Nov 6, 2025

liangan1 requested a review from jerryzh168 November 6, 2025 02:55

liangan1 approved these changes Nov 6, 2025

View reviewed changes

liangan1 self-requested a review November 6, 2025 02:57

liangan1 approved these changes Nov 6, 2025

View reviewed changes

liangan1 added the ciflow/xpu label used to trigger xpu CI jobs label Nov 6, 2025

jerryzh168 reviewed Nov 7, 2025

View reviewed changes

refine the device

54bc5c8

pytorch-bot bot removed the ciflow/xpu label used to trigger xpu CI jobs label Nov 7, 2025

refine the device

a9043e2

liangan1 requested a review from jerryzh168 November 7, 2025 06:57

liangan1 added the ciflow/xpu label used to trigger xpu CI jobs label Nov 7, 2025



		class TestMultiTensorFlow(TestCase):
		@unittest.skipIf(not torch.cuda.is_available(), "Need CUDA available")

[Intel XPU] Enable test/quantization UTs on XPU #3249

Are you sure you want to change the base?

[Intel XPU] Enable test/quantization UTs on XPU #3249

Conversation

DiweiSun commented Oct 27, 2025 • edited by liangan1 Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

pytorch-bot bot commented Oct 27, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/3249

❌ 3 New Failures

Uh oh!

pytorch-bot bot commented Oct 27, 2025

Uh oh!

pytorch-bot bot commented Nov 4, 2025

Uh oh!

jerryzh168 Nov 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

pytorch-bot bot commented Nov 6, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

DiweiSun commented Oct 27, 2025 •

edited by liangan1

Loading

pytorch-bot bot commented Oct 27, 2025 •

edited

Loading

jerryzh168 Nov 6, 2025 •

edited

Loading