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.
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
[Pytorch][Vulkan] Call binary_op_scalar when 'other' is a 0-dim tensor #109035
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/109035
Note: Links to docs will display an error until the docs builds have been completed. ✅ You can merge normally! (3 Unrelated Failures)As of commit e03bc22 with merge base e7bd9c5 (): UNSTABLE - The following jobs failed but were likely due to flakiness present on trunk and has been marked as unstable:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This pull request was exported from Phabricator. Differential Revision: D48672535 |
pytorch#109035) Summary: 0-dim tensor are not supported in Vulkan. If a binary_op_tensor is called with 'other_arg' being a 0-dim tensor, then we extract the scalar out and call binary_op_scalar. Used to run the [FD model]( https://www.internalfb.com/manifold/explorer/wrist-camera-ml/tree/models/fd-ted-pi/fd-hybrid/fd_hybrid_vulkan.ptl) on [CLI](https://www.internalfb.com/intern/wiki/Malibu/Software/Machine_Learning/PyTorch_On_Device_Catalog/#build-and-run-native-pyt) Test Plan: Run vulkan_api_tests ``` lfq@lfq-mbp fbsource % buck run --target-platforms ovr_config//platform/macos:arm64-fbsource //xplat/caffe2:pt_vulkan_api_test_binAppleMac\#macosx-arm64 -c pt.vulkan_full_precision=1 ... [ RUN ] VulkanAPITest.querypool_flushed_shader_log xplat/caffe2/aten/src/ATen/test/vulkan_api_test.cpp:6891: Skipped QueryPool is not available [ SKIPPED ] VulkanAPITest.querypool_flushed_shader_log (0 ms) [----------] 339 tests from VulkanAPITest (5308 ms total) [----------] Global test environment tear-down [==========] 339 tests from 1 test suite ran. (5308 ms total) [ PASSED ] 338 tests. [ SKIPPED ] 1 test, listed below: [ SKIPPED ] VulkanAPITest.querypool_flushed_shader_log YOU HAVE 5 DISABLED TESTS ``` Reviewed By: manuelcandales Differential Revision: D48672535
c73328d
to
9b97037
Compare
pytorch#109035) Summary: 0-dim tensor are not supported in Vulkan. If a binary_op_tensor is called with 'other_arg' being a 0-dim tensor, then we extract the scalar out and call binary_op_scalar. Used to run the [FD model]( https://www.internalfb.com/manifold/explorer/wrist-camera-ml/tree/models/fd-ted-pi/fd-hybrid/fd_hybrid_vulkan.ptl) on [CLI](https://www.internalfb.com/intern/wiki/Malibu/Software/Machine_Learning/PyTorch_On_Device_Catalog/#build-and-run-native-pyt) Test Plan: Run vulkan_api_tests ``` lfq@lfq-mbp fbsource % buck run --target-platforms ovr_config//platform/macos:arm64-fbsource //xplat/caffe2:pt_vulkan_api_test_binAppleMac\#macosx-arm64 -c pt.vulkan_full_precision=1 ... [ RUN ] VulkanAPITest.querypool_flushed_shader_log xplat/caffe2/aten/src/ATen/test/vulkan_api_test.cpp:6891: Skipped QueryPool is not available [ SKIPPED ] VulkanAPITest.querypool_flushed_shader_log (0 ms) [----------] 339 tests from VulkanAPITest (5308 ms total) [----------] Global test environment tear-down [==========] 339 tests from 1 test suite ran. (5308 ms total) [ PASSED ] 338 tests. [ SKIPPED ] 1 test, listed below: [ SKIPPED ] VulkanAPITest.querypool_flushed_shader_log YOU HAVE 5 DISABLED TESTS ``` Reviewed By: manuelcandales Differential Revision: D48672535
This pull request was exported from Phabricator. Differential Revision: D48672535 |
9b97037
to
e03bc22
Compare
This pull request was exported from Phabricator. Differential Revision: D48672535 |
@pytorchbot merge (Initiating merge automatically since Phabricator Diff has merged) |
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 |
Summary:
0-dim tensor are not supported in Vulkan.
If a binary_op_tensor is called with 'other_arg' being a 0-dim tensor, then we extract the scalar out and call binary_op_scalar.
Used to run the FD model on CLI
Test Plan:
Differential Revision: D48672535