depthcharge VS emlearn

I have a BeagleBone Black (BBB) board and I am trying to use a toolkit called Depthcharge to evaluate U-boot. I downloaded the following image for BBB and flashed it to an SD Card and inserted it in the board. Depthcharge has a Python script that needs to reach the U-boot shell in order to work, but it can't do it. When I manually observe the booting sequence, I notice

for router 4: when booting you should be able to get into a u-boot shell by hitting any key within the first few seconds after applying power Then U-Boot Depthcharge can be used to automate the process of figuring out what you can do with the u-boot shell, ie what commands are available, dump the flash, find any hidden commands etc. Lots of fun, and only the basics of command line/python proficiency required to use it (aside from computer/VM running Linux and a USB to UART)

> My big gripe, and for obvious reasons, is that we need to step away from cloud-based inference, and it doesn't seem like anyone's working on that.

I think there are steps being taken in this direction (check out [1] and [2] for interesting lightweight transpile / ad-hoc training projects) but there is a lack of centralized community for these constrained problems.

[1] https://github.com/emlearn/emlearn

Examples: Gaussian Mixture Models (GMM) for anomaly detection or clustering Mahalanobis distance (EllipticEnvelope) for anomaly detection Decision trees and tree ensembles (Random Forest, ExtraTrees) Feed-forward Neural Networks (Multilayer Perceptron, MLP) for classification Gaussian Naive Bayes for classification

https://github.com/emlearn/emlearn is a ML inference engine for microcontrollers and embedded systems, allowing to deploy models to any platform with a C99 compiler. Has also been used for network traffic analysis as a Linux kernel module, and embedded in Android apps.

The C64 has 64 kB of RAM. That is more than many contemporary microcontrollers. Using something like https://github.com/emlearn/emlearn allows to generate portable C code of ML models for such targets. Should be able to classify digits (MNIST) no problem on such hardware. Assuming there is a workable C compiler available.

Disclosure: Maintainer of emlearn project.

I built emlearn, a Machine Learning inference engine for microcontrollers and embedded systems. It allows converting traditional ML models to simple and portable C99, following best practices in embedded software (no dynamic allocations etc). https://github.com/emlearn/emlearn