Open
Conversation
obaileyw-uncc
commented
Nov 18, 2025
obaileyw-uncc
commented
- created new module sww_ref_util_submitter for submitter-implemented platform-specific code in the streaming wakeword benchmark
- renamed all functions and cleaned up styling to increase consistency with other benchmarks
- created the first draft of a TUTORIAL.md in the benchmark folder describing the full process for using the benchmark reference implementations
- added the missing WAV for the medium (120 second) test to sww_data_dir
…actor * Fixed typo in the setup script for the mbed_basic empty reference implementation * Added a requirements.txt for the Mbed OS Python environment (note: Python 3.9 required, any future version will break Mbed OS) * Corrected the device configuration for the streaming wakeword test (LPM was starting in 1.8V by default, LPM now starts at 3.3V by default)
* created new module sww_ref_util_submitter for submitter-implemented platform-specific API for hardware configuration, GPIO operation, and peripheral operation (I2S DMA and UART) * extracted STM32 HAL API calls and interrupt handler definitions from main and sww_ref_util and moved to functions and macros in reference example of sww_ref_util_submitter, replaced with new submitter-implemented API calls * renamed all functions * cleaned up header and module files, moved all includes and defines out of .c files into headers * cleaned up styling, brought streaming wakeword code to consistent tab count
…sww_ref_util and moved everything requiring the hardware timer to sww_ref_util_submitter for performance reasons
… describing the full process for using the benchmark reference implementations, added the missing medium test and Marvin data files to sww_data_dir
…der sww01 under the specified path to datasets
|
MLCommons CLA bot All contributors have signed the MLCommons CLA ✍️ ✅ |
Author
|
recheck |
Contributor
|
recheck |
…ripts including one which iterates multiple times per model and/or over multiple filter counts, transfer learning, transfer fine tuning and knowledge distillation, modularized existing ResNet-style model to include argument for first-stage filter count, updated pip requirements and readme for TF2.14
…larger newly-trained model, fixed typo in model_converter.py which sliced characters from the model name that would fail convert_to_tflite.sh, removed old MD5 models
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
2 participants
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.