[Intel GPU] Enable optim SR test #3055
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/3055
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ✅ No FailuresAs of commit b1121ac with merge base f92b898 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
bot
added
the
CLA Signed
test/test_low_bit_optim.py
Outdated
| torch.testing.assert_close(p2, p1) | ||
|
|
||
| @parametrize("device", _DEVICES) | ||
| def test_optim_bf16_stochastic_round_correctness(self): |
There was a problem hiding this comment.
device arg is not added to the function
There was a problem hiding this comment.
@arlesniak _DEVICES is the global variable in this file, I don't think your change is needed.
There was a problem hiding this comment.
The "test_optim_bf16_stochastic_round_correctness" should be parametrized with _DEVICES global variable, which contains other devices (XPU if detected). Without the PR changes the choice was cuda, cpu only.
There was a problem hiding this comment.
Are you sure? According to this code,https://github.com/pytorch/ao/blob/main/torchao/utils.py#L143 when the xpu is available in your ENV, the _DEVICES will contain the XPU.
There was a problem hiding this comment.
Let me be more precise. You're right, the _DEVICES list will contain XPU if detected. Without the PR changes the test will be triggered once with the hardcoded device choice - cuda or cpu: https://github.com/pytorch/ao/blob/main/test/test_low_bit_optim.py#L423, regardless of _DEVICES content.
After the PR changes, the test will be triggered separately for every device from the _DEVICES including XPU (similiar to some other parametrized tests i.e. test_quantize_8bit_with_qmap_correctness etc).
arlesniak
force-pushed
the
arlesniak/enable_optim_sr_test_xpu
branch
from
cfd22f6 to
b1121ac
Compare
liangan1
added
topic: improvement
|
To add the ciflow label 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. |
* add MXFP8 all gather support * added TODO for future feature * remove emoji from comment * fixed ruff formating * fixed ruff formatting * add mxfp8 and nvfp4 to Llama eval scripts (#3394) Update [ghstack-poisoned] * flip mx inference scaling setting to RCEIL (#3428) * Update [ghstack-poisoned] * Update [ghstack-poisoned] * Update [ghstack-poisoned] * add CLAUDE.local.md to gitignore (#3437) Summary: taking claude code for a more thorough spin, will start with local instructions and will see what makes sense to upstream Test Plan: Reviewers: Subscribers: Tasks: Tags: * bump python version in tutorial ci workflow (#3439) * [CPU] Reland qconv fp8 fusion passes (#3433) * [Reland][PT2E][X86] Add Inductor fusion passes of float8 qconv for X86Inductor backend * add torch version check for Qconv FP8 UTs * fix format issue * Skip tests for ROCm --------- Co-authored-by: Sun, Jiayi <jiayi.sun@intel.com> * Int8Tensor migration cleanup (#3407) * Int8Tensor migration Summary: This PR creates a new Int8Tensor and updates the configs to use the new Int8Tensor flow Test Plan: To ensure BC: ``` pytest test/quantization/test_quant_api.py ``` To test new Int8Tensor: ``` pytest test/quantization/quantize_/workflows/int8/test_int8_tensor.py ``` Reviewers: Subscribers: Tasks: Tags: * ruff fixes * add init * fix ruff again * update * wip * undo update tests * fix ruff * fix varname * fix typing * add tests * fix dtype * fix ci * address granularity cr * update _choose_quant_func_and_quantize_tensor * make block size required attribute * made dtype required as well * address nits * skip per tensor weight only test for now * [xpu][test] Port 2 test/dtypes_{floatx, bitpacking} UT files to intel XPU (#3368) * enable test/dtypes/test_bitpacking.py on intel xpu * enable test/dtypes/test_floatx.py * enable test/dtypes/test_floatx.py * fix format issue * fix format issue * update _DEVICES * [xpu][test] Port 2 test/quantization/pt2e/test_{quantize_pt2e, quantize_pt2e_qat} UT files to intel XPU (#3405) * add test/quantization/pt2e/test_quantize_pt2e.py * add test/quantization/pt2e/test_quantize_pt2e.py * test/quantization/pt2e/test_quantize_pt2e_qat.py * test/quantization/pt2e/test_quantize_pt2e_qat.py * fix format issue * update format * increase timeout for xpu * [Intel GPU] Enable optim SR test (#3055) * updated test with rebase changes * added checks to run only on CUDA with compatibility >=9 * updated test for H100 * added test to workflow --------- Co-authored-by: Vasiliy Kuznetsov <vkuzo@users.noreply.github.com> Co-authored-by: Daniel Vega-Myhre <danvm@meta.com> Co-authored-by: Xia Weiwen <weiwen.xia@intel.com> Co-authored-by: Sun, Jiayi <jiayi.sun@intel.com> Co-authored-by: Jesse Cai <jessecai@meta.com> Co-authored-by: xiangdong <40376367+zxd1997066@users.noreply.github.com> Co-authored-by: Artur Lesniak <artur.lesniak@intel.com>
4 participants
Enable optim_bf16_stochastic_round_correctness test for XPU