emlearn-micropython
emlearn-micropython copied to clipboard
emlearn
Efficient Machine Learning engine for MicroPython
Currently utilizes an MR at https://github.com/micropython/micropython/pull/15838 - hopefully to be included in MicroPython shortly. It lets us remove all the hacking where we manually resolve symbols, unpack archives and add...
MicroPython can run in the browser using the WebAssembly/WASM port. This would enable to use emlearn-micropython in the browser, be it for small demos, educational usage or providing "virtual" hardware...
It seems that the natmod support for RISC-V architecture may have gotten into a useful place now in MicroPython after 1.24. This is the architecture used by chips such as...
When the recorder is started for the first time, the LCD display stays blank. To reproduce the issue: ``` mpremote reset mpremote run har_record.py ``` The second time the program...
The following code ``` import emlearn_trees import array import gc def test_trees_alloc(): n_trees = 100 nodes_per_tree = 10000 leaves = 127 n_classes = leaves n_features = 127 n_nodes = n_trees...
Can you add examples of face recognition? I'm a little white and I'm not very good at it.
When working on image classification (like with the emlearn Convolutional Neural Network), it is desirable to be able to work with standard image formats. Both on PC (for evaluation, testing)...
https://github.com/openjournals/joss-reviews/issues/9093 Hi @jonnor, For the checklist’s "Automated tests" item, could you document how to run the unit tests in /tests ? if it already exists, please point me to it....
MicroPython now has a QEMU port that is relatively mature. One can use mpremote with it, and it supports a range of ARM Cortex M targets along with RISC-V. So...
Here is an initial proof of concept that appears to be working on a few datasets. For 6k samples dataset and 5 trees, training times in order of 15-60 seconds...