Support more foreach ops for tensor beta support #134170
Conversation
[ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/134170
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit f87e590 with merge base a1b22e3 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
eellison
left a comment
eellison
left a comment
There was a problem hiding this comment.
wonder if we could do this more programatically..
Hmm I'm not sure, I think for testing I could use the op info tests for for each. For this I'm not sure mainly because there might be patterns or weird overloads that aren't support yet. Which is why we've incrementally added. This is kind of a consequence of the foreach API itself which has been incrementally updated as needed and not all overloads are supported for every op. |
|
Like, would it be sufficient to register |
The issue is I also need to generate tests for them, but let me see if I can use some sort of operator schema to generate the tests (which would be pretty cool) |
Add more foreach ops so we don't have fallbacks. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
Add more foreach ops so we don't have fallbacks. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
Add more foreach ops so we don't have fallbacks. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
Add more foreach ops so we don't have fallbacks. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
Add more foreach ops so we don't have fallbacks. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
Add more foreach ops so we don't have fallbacks. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
Add more foreach ops so we don't have fallbacks. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
Add more foreach ops so we don't have fallbacks. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
|
@pytorchbot merge |
Add more foreach ops so we don't have fallbacks. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Add more foreach ops so we don't have fallbacks. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
Add more foreach ops so we don't have fallbacks. cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx ipiszy yf225 chenyang78 kadeng muchulee8 ColinPeppler amjames desertfire chauhang [ghstack-poisoned]
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
|
The merge job was canceled or timed out. This most often happen if two merge requests were issued for the same PR, or if merge job was waiting for more than 6 hours for tests to finish. In later case, please do not hesitate to reissue the merge command |
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
|
The merge job was canceled or timed out. This most often happen if two merge requests were issued for the same PR, or if merge job was waiting for more than 6 hours for tests to finish. In later case, please do not hesitate to reissue the merge command |
|
@pytorchbot merge |
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
# Motivation Fix #138577. # Solution 1. All UTs in `test/inductor/test_compiled_optimizers.py` are fixed by #134170 2. UT in `test/inductor/test_pattern_matcher.py` is introduced by #138089, we will skip this UT due to the unsupported feature `max_autotune_gemm_backends:Triton`. 3. We have a new impl related to `histc`, so we remove the expected failure from `test/inductor/test_torchinductor_opinfo.py` 4. We support `avg_pool3d` for `fp16` data type, so we remove the expected failure from `test/inductor/test_torchinductor_opinfo.py` 5. CUDA-bias code is introduced by #138472, we just generalize it to `GPU_TYPE`. # Additional Context > Why update torch-xpu-ops commit pin here? We have to update commit pin to avoid the build failure raised by the code change [C10_UNUSED](#138364). > What does the feature of torch-xpu-ops update? 1. Add some foreach ops, like `unary ops` and `foreach_clamp_max` etc; 2. Add some maxpool ops forward and backward, like `averge_pool3d` and `max_pool3d` 3. Add some other ops, like `log_normal_`, `index_copy`, and `mode` etc; 4. fix build failure related to `C10_UNUSED`; Pull Request resolved: #138548 Approved by: https://github.com/malfet, https://github.com/EikanWang
ghstack-source-id: 84d4ab8 Pull Request resolved: pytorch/pytorch#134170 Fix test counts fix counts
Add more foreach ops so we don't have fallbacks.
Stack from ghstack (oldest at bottom):
cc @voznesenskym @penguinwu @EikanWang @jgong5 @Guobing-Chen @XiaobingSuper @zhuhaozhe @blzheng @wenzhe-nrv @jiayisunx @ipiszy @yf225 @chenyang78 @kadeng @muchulee8 @ColinPeppler @amjames @desertfire @chauhang