Repository navigation
Production update - #4029
Merged
Merged
Production update#4029
Conversation
Keep the documented example and source-built Python package aligned while replacing the dated TorchAO nightly with released dependencies. Document the FVP runtime prerequisite and the newly verified Linux end-to-end results. Authored with assistance from OpenAI Codex.
Align source and released Python packages, resolve dependencies after Arm setup, and clarify host prerequisites. The complete Linux AArch64 export, target build, FVP run and numerical comparison pass without changing model or runtime commands. Authored with assistance from OpenAI Codex.
Trim explanations introduced by the release-dependency update without changing existing checks or execution commands. Fresh Linux AArch64 setup and complete Corstone-320 end-to-end validation pass with the documented thresholds. Authored with assistance from OpenAI Codex.
Shorten explanations introduced by the release-dependency update while preserving earlier revision and import checks and all installation and runtime commands. Fresh Linux AArch64 end-to-end validation passes with all 79 speech decisions matching on Corstone-320. Authored with assistance from OpenAI Codex.
List the Python build prerequisites, check the exporter imports, and explain runtime delegation and fast-mode timing limitations. Authored with assistance from OpenAI Codex.
Explain that the pinned exporter retains state from reference generation and scope the probability comparison to the documented functional check. Authored with assistance from OpenAI Codex.
_index.md: - Update LP title and broaden target audience to include AI/ML developers - Clarify Windows x86-64 requirements and consolidate development prerequisites 1-introduction.md: - Clarify the roles of VK_ARM_tensors, VK_ARM_data_graph, and SPIR-V - Broaden the positioning of ML Extensions beyond gaming-only workloads - Improve heading hierarchy and separate the next-steps section - Simplify the transition to hands-on samples and remove premature tooling details 2-ml-ext-for-vulkan.md: - Simplify setup by moving general development tools to prerequisites - Clarify Python virtual environment setup and introduce ML SDK components - Expand Vulkan SDK installation and verification instructions - Update Vulkan Configurator steps for clearer, version-independent layer configuration 4-scenario-runner.md: - Add ML Workload Library for Vulkan to the SDK component overview 5-renderdoc.md: - Simplify RenderDoc introduction and remove duplicated capability descriptions
1-introduction.md: - Introduce Vulkan extensions and explain the purpose of Arm ML Emulation Layers - Clarify how emulation layers expose ML extensions on hardware without native support - Add an architecture diagram showing the application, emulation layers, driver, and GPU - Reduce repetition between the technical introduction and benefits section - Simplify the transition to the setup instructions 2-ml-ext-for-vulkan.md: - Align setup overview with the three installation and configuration steps - Simplify package descriptions and use consistent Python commands - Clarify the roles of Arm ML SDK for Vulkan and LunarG Vulkan SDK - Explain why emulation layers must be enabled after installation 3-first-sample.md: - Use PowerShell syntax highlighting consistently for Windows commands - Clarify the sample summary to reflect what developers observe when running the application - Remove wording that implies developers manually performed the underlying Vulkan API operations 4-scenario-runner.md: - Remove unnecessary blank line before the Hugging Face section images/ml-emulation-layers.png: - Add architecture diagram illustrating how the ML Emulation Layers fit into the Vulkan runtime stack
2-ml-ext-for-vulkan.md: - Promote Vulkan SDK installation to a main section heading 4-scenario-runner.md: - Introduce Scenario Runner and its role in the ML SDK for Vulkan - Explain VGF and its use for packaging ML workloads - Add instructions to download and run the Neural Super Sampling example - Document Scenario Runner installation and output verification - Connect the exercise to RenderDoc image inspection 5-renderdoc.md: - Streamline the introduction and focus on Vulkan inspection and debugging - Add instructions for viewing Neural Super Sampling DDS output - Clarify the transition to capturing the Simple Tensor and Data Graph sample - Consolidate the Vulkan capture instructions images/scenario-runner-architecture-400.png: - Add architecture diagram showing Scenario Runner, VGF Library, ML Emulation Layers, and Vulkan API
Pin the updated example, match its released package setup and default input/calibration/build flow, and refresh the threshold, state explanation, reference output, and diagrams. Validate a fresh Linux AArch64 export, runtime/application builds, and FVP run: 79 frames agree at threshold 0.5. Documentation and Hugo checks pass. This change was authored with assistance from OpenAI Codex.
Pin the updated example, use its stock source installer before Arm setup, and align host requirements, timed mode, and unavailable FVP measurements. Remove obsolete dependency workarounds and old reference measurements. Validate a fresh Linux AArch64 source install, export, runner build, and FVP run: host IoU 0.9547 and target IoU 0.9737. Documentation and Hugo checks pass. This change was authored with assistance from OpenAI Codex.
…odex/mobilesam-stable-dependencies
…ugh laptops-and-desktops
This reverts commit 7f9873bba6b986781a13a4a09a44e865bbe61654.
c2pa tech review
LP: Innexis ANA Zena CPU llm build and inference
uv_debug alt-text updates
Compute Flow additional edits
…rding LP: Innexis ANA Zena onboarding
Updating images with no alt-text + hiding summary and FAQ reports
…rning-path Update vulkan ml learning path
…-computing WebAssembly LP and install guide review
Add ETRecord/ETDump MLIA Model Explorer section
…cies Align the MobileSAM Learning Path with its ExecuTorch E2E example
Align the Silero VAD Learning Path with its ExecuTorch E2E example
[Tech Review Changes] Muse Glimmer ExecuTorch on DGX Spark
…learning-paths into content_review
Update .wordlist.txt and link fix
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Before submitting a pull request for a new Learning Path, please review Create a Learning Path
Please do not include any confidential information in your contribution. This includes confidential microarchitecture details and unannounced product information.
By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of the Creative Commons Attribution 4.0 International License.