Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[Evaluated] skips in test_nn #258

Open
3 tasks
fengyuan14 opened this issue May 16, 2024 · 6 comments
Open
3 tasks

[Evaluated] skips in test_nn #258

fengyuan14 opened this issue May 16, 2024 · 6 comments
Assignees
@fengyuan14
Labels
ut_triaged
Milestone
PT2.7

Comments

@fengyuan14
Copy link
Contributor

fengyuan14 commented May 16, 2024 •
edited by yuchengliu1
Loading

🐛 Describe the bug

  • oneDNN issues // Please check run_test_with_skip.py
  • CPU fallback failures // Need to evaluate when we have XPU implementation
  • aten::_thnn_fused_gru_cell // CPU fallback could not cover it

Versions

main
@fengyuan14 fengyuan14 self-assigned this May 16, 2024
@fengyuan14 fengyuan14 changed the title Evaluated skips in test_nn [Evaluated] skips in test_nn May 16, 2024
@fengyuan14 fengyuan14 added this to the PT2.5 milestone May 22, 2024
@chuanqi129
Copy link
Contributor

@daisyden please double check whether we need re-create issue for those cases

@daisyden
Copy link
Contributor

daisyden commented Jul 18, 2024 •
edited
Loading

These issues are still there with torch-xpu-ops 44d2844

@daisyden
Copy link
Contributor

Yucheng has rerun the test with supported op enabled, need further triage.

@yuchengliu1
Copy link
Contributor

yuchengliu1 commented Jul 24, 2024 •
edited
Loading

new passed case:
"test_sync_batchnorm_accuracy_cuda",
"test_sync_batchnorm_backward_elemt",
"test_MSELoss_no_batch_dim_mean_cuda_half",
"test_MSELoss_no_batch_dim_none_cuda_half",
"test_MSELoss_no_batch_dim_sum_cuda_half",
"test_grid_sample_error_checking",
There are some cases listed in #653 which target to PT2.5. Other case target 2.6 or later

This was referenced Jul 25, 2024
Closed
Merged
github-merge-queue bot pushed a commit that referenced this issue Aug 2, 2024
@yuchengliu1 @fengyuan14
fix cuda bias code in test_nn (#656)
fb8e6e9 
update test_nn skip list and align with cuda #258  
- [x] cause by cuda hard code
    "test_device_mask_xpu",
    "test_overwrite_module_params_on_conversion_cpu_device_xpu",
 - [x]  do not aline with cuda
    "test_upsamplingBiMode2d_consistency",
    "test_upsamplingBiLinear2d_consistency_interp_size_bug",

---------

Co-authored-by: Feng Yuan <[email protected]>
@daisyden daisyden modified the milestones: PT2.5, PT2.6 Aug 5, 2024
@yuchengliu1 yuchengliu1 modified the milestones: PT2.5, PT2.6 Aug 5, 2024
@daisyden daisyden assigned daisyden and unassigned daisyden Sep 19, 2024
@yuchengliu1
Copy link
Contributor

Cases for PT2.5 have been done. The rest cases is for PT2.6 or later.

@daisyden daisyden modified the milestones: PT2.6, PT2.7 Nov 29, 2024
@daisyden
Copy link
Contributor

update the status:

  • aten::_thnn_fused_gru_cell is wip
  • onednn issue is still blocked
  • CPU fallback failures will need reevaluate when XPU is implemented.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@fengyuan14 fengyuan14

Labels
ut_triaged
Projects
None yet
Milestone
PT2.7
Development

No branches or pull requests
4 participants
@chuanqi129 @daisyden @fengyuan14 @yuchengliu1