Zhiwei
Zhiwei
@pytorchbot merge
@pytorchbot merge
hi, @yanbing-j We are communicating with the oneDNN team for the `Build left local git repository checkout dirty` error. As for the dynamich shape, the inductor code maintainer is fixing.
hi, @svekars @AlannaBurke could you please help review our documentation? The PR serves as a tutorial for PT2E int8 on Intel GPU backend. Appreciation for your feedback and suggestions.
hi @AlannaBurke @svekars @HamidShojanazeri , I've applied the suggestions in latest commits. Could you please help review it again and approve it if no further issues in this tutorial? Great...
hi, @EikanWang the issue in ci `test_addmm_xpu_float16` seems a regression. But I could not figure out why it happens, and it does not show in former commit. I just rebase...
``` AssertionError: Tensor-likes are not close! Mismatched elements: 2 / 250 (0.8%) Greatest absolute difference: 0.000244140625 at index (8, 20) (up to 1e-05 allowed) Greatest relative difference: 0.0019197463989257812 at index...
> `inductor/test_torchinductor_opinfo.py::TestInductorOpInfoXPU::test_comprehensive_linalg_eigvals_xpu_float64` @ZhiweiYan-96 The failure appears to be related to this PR. Please investigate and address the issue. @guangyey There should be some new skipped uts were added, but this...
hi, @EikanWang This PR has been refined. Could you please review this PR? Thanks!
@EikanWang I have rebased to the main, the window ci failure should be away now.