OpenVINO Backend Build Updates #15650
Conversation
…stall_requirement.py
cavusmustafa
requested review from
jackzhxng,
kirklandsign,
larryliu0820 and
lucylq
as code owners
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/15650
Note: Links to docs will display an error until the docs builds have been completed. ❌ 1 New Failure, 4 Unrelated FailuresAs of commit 32bda53 with merge base 054b15d ( NEW FAILURE - The following job has failed:
BROKEN TRUNK - The following jobs failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
bot
added
the
CLA Signed
This PR needs a
|
There was a problem hiding this comment.
Pull Request Overview
This PR updates the OpenVINO backend build infrastructure to improve flexibility and fix configuration issues. The changes remove the --minimal flag from Python builds to include example dependencies, introduce a new --cpp_runtime_llm flag for building with LLM support, and update documentation and configuration files accordingly.
- Refactored build script to use conditional LLM dependency inclusion via
--cpp_runtime_llmflag - Updated CMake configuration to use arrays for better argument handling and removed inline OpenVINO backend build in favor of
find_package - Updated documentation and requirements files to reflect new build options and fix parameter references
Reviewed Changes
Copilot reviewed 8 out of 10 changed files in this pull request and generated 2 comments.
Show a summary per file
| File | Description |
|---|---|
| backends/openvino/scripts/openvino_build.sh | Refactored build script with new --cpp_runtime_llm flag, removed build_llama_runner function, and restructured CMake argument handling |
| backends/openvino/README.md | Updated build documentation to reference new --cpp_runtime_llm flag instead of --llama_runner |
| backends/openvino/requirements.txt | Removed transformers dependency from requirements |
| examples/openvino/README.md | Added stable_diffusion directory structure to documentation |
| examples/openvino/llama/README.md | Updated instructions to use new --cpp_runtime_llm flag and provide explicit CMake commands for building llama runner |
| examples/openvino/stable_diffusion/README.md | Fixed line numbering formatting |
| examples/openvino/stable_diffusion/requirements.txt | Fixed line numbering formatting |
| examples/models/yolo12/README.md | Updated documentation to clarify pip install strategy and fix --video_input to --video_path parameter |
| examples/models/yolo12/requirements.txt | Updated ultralytics version and added opencv-python dependency |
| examples/models/yolo12/CMakeLists.txt | Changed default OpenVINO backend option to OFF, updated XNNPACK library name, and simplified OpenVINO integration to use find_package |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| ) | ||
|
|
||
| if [[ "$llm_enabled" -eq 1 ]]; then | ||
| CMAKE_ARGS+=("-DEXECUTORCH_BUILD_EXTENSION_LLM=ON -DEXECUTORCH_BUILD_EXTENSION_LLM_RUNNER=ON") |
There was a problem hiding this comment.
The CMAKE_ARGS array elements are incorrectly combined into a single string. The two options -DEXECUTORCH_BUILD_EXTENSION_LLM=ON and -DEXECUTORCH_BUILD_EXTENSION_LLM_RUNNER=ON should be added as separate array elements, not as a single string.
Change:
CMAKE_ARGS+=("-DEXECUTORCH_BUILD_EXTENSION_LLM=ON -DEXECUTORCH_BUILD_EXTENSION_LLM_RUNNER=ON")To:
CMAKE_ARGS+=("-DEXECUTORCH_BUILD_EXTENSION_LLM=ON" "-DEXECUTORCH_BUILD_EXTENSION_LLM_RUNNER=ON")| CMAKE_ARGS+=("-DEXECUTORCH_BUILD_EXTENSION_LLM=ON -DEXECUTORCH_BUILD_EXTENSION_LLM_RUNNER=ON") | |
| CMAKE_ARGS+=("-DEXECUTORCH_BUILD_EXTENSION_LLM=ON" "-DEXECUTORCH_BUILD_EXTENSION_LLM_RUNNER=ON") |
| ) | ||
|
|
||
| if [[ "$llm_enabled" -eq 1 ]]; then | ||
| CMAKE_ARGS+=("-DEXECUTORCH_BUILD_EXTENSION_LLM=ON -DEXECUTORCH_BUILD_EXTENSION_LLM_RUNNER=ON") |
There was a problem hiding this comment.
This line uses a tab character for indentation instead of spaces, which is inconsistent with the rest of the file. Replace the tab with spaces to maintain consistent indentation.
| CMAKE_ARGS+=("-DEXECUTORCH_BUILD_EXTENSION_LLM=ON -DEXECUTORCH_BUILD_EXTENSION_LLM_RUNNER=ON") | |
| CMAKE_ARGS+=("-DEXECUTORCH_BUILD_EXTENSION_LLM=ON -DEXECUTORCH_BUILD_EXTENSION_LLM_RUNNER=ON") |
2 participants
Summary
Test plan
This PR doesn't include new functionality which requires testing.
cc: @mergennachin @cbilgin @digantdesai @ynimmaga @suryasidd @daniil-lyakhov @anzr299