Update Pico2 docs with CMSIS-NN INT8 support and latency instrumentation#18898
Update Pico2 docs with CMSIS-NN INT8 support and latency instrumentation#18898psiddh merged 3 commits intopytorch:mainfrom
Conversation
Add documentation for the new --cmsis build flag, INT8 quantized model export via export_mlp_mnist_cmsis.py, and updated serial output showing per-inference latency timing and memory usage diagnostics. Co-authored-by: Claude <noreply@anthropic.com>
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/18898
Note: Links to docs will display an error until the docs builds have been completed. ❗ 1 Active SEVsThere are 1 currently active SEVs. If your PR is affected, please view them below: ⏳ No Failures, 76 PendingAs of commit 3d79b20 with merge base 26e2ab8 ( 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
Updates the Pico2 MNIST documentation to describe the new CMSIS-NN INT8 workflow and the expanded runtime diagnostics printed over serial.
Changes:
- Document FP32 vs INT8 (CMSIS-NN accelerated) model export paths and firmware build commands.
- Add
--cmsisflag usage and include a build script options table. - Update example serial output to include per-inference timing, a latency summary, and post-load memory usage.
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.
| File | Description |
|---|---|
| examples/raspberry_pi/pico2/README.md | Adds FP32/INT8 build+export instructions, script flags table, and updated sample serial output/diagnostics. |
| docs/source/pico2_tutorial.md | Mirrors the README updates in the Sphinx tutorial, including the CMSIS-NN INT8 workflow and updated runtime output examples. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
There was a problem hiding this comment.
Pull request overview
Copilot reviewed 2 out of 2 changed files in this pull request and generated 6 comments.
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
- Memory pools are 200KB (204800 bytes), not 131072/4096 - Memory usage prints before inference loop, not after - Remove duplicate memory block from after latency summary - Fix PREDICTED line to include 🎯 prefix matching main.cpp Co-authored-by: Claude <noreply@anthropic.com>
Add documentation for the new --cmsis build flag, INT8 quantized model export via export_mlp_mnist_cmsis.py, and updated serial output showing per-inference latency timing and memory usage diagnostics.