I ran MNIST on an ESP32-C3 without TensorFlow, TFLite, or any ML runtime

The author successfully ran MNIST digit recognition on an ESP32-C3 microcontroller by compiling a neural network directly into a C header file and executing it as firmware, without using TensorFlow, TFLite, or any ML runtime. The project evolved from a simple perceptron experiment into a cross-platform CLI tool that generates minimal, deterministic inference pipelines for constrained edge devices. The author emphasizes that the toolchain prioritizes stability and portability over competing with full ML frameworks, focusing on how small a usable neural inference stack can become while remaining practical.

I ran MNIST digit recognition on an ESP32-C3 — without TensorFlow, TFLite, or any ML runtime. The neural network is compiled directly into a C header and executed as firmware. It started, like many questionable engineering projects, with curiosity about a perceptron. I wasn’t trying to build anything serious. Just a small experiment to understand how far I could push a minimal neural network implementation without relying on any framework. That experiment didn’t stay small for long. Once the basic forward/backward pass worked, I started wondering whether the same approach could be useful outside of toy problems. Specifically, I had recently built a ratiometric thermometer system, and it quickly became clear that the analog measurements were noisy and sensitive to supply variation. That’s where the idea clicked: instead of fitting everything with fixed equations, what if I used a lightweight neural network as an inference layer for correction? Not training in the field — just inference on constrained hardware. That idea turned into a CLI tool, not a library. Not a framework. A command-line tool. The goal was simple: No heavy dependencies. No runtime frameworks required on the target side, just portable inference. Over time, this evolved into a cross-platform tool supporting Linux, macOS, and Windows, with a fully deterministic inference pipeline. At some point I needed a sanity check, so I ran MNIST through it. ~59,999 training samples~1,000 test samples Nothing exotic — just a standard benchmark to validate the pipeline end-to-end. The interesting part wasn’t accuracy. It was stability. Training, export, and inference all behaved consistently across platforms. No drift between environments, no dependency on external runtimes, and no hidden assumptions about floating-point behavior beyond what was explicitly defined. For a homegrown stack, that was the real milestone. The original motivation was never image classification, it was edge inference under constraints: In that context, most modern ML stacks are overkill. Even lightweight frameworks often assume too much infrastructure. So the focus shifted toward something more specific: A minimal neural network runtime that could be compiled down and deployed without dependencies. Today, the project is essentially three things: A minimal neural network training engine A model exporter that produces standalone inference graphs A CLI runtime for cross-platform execution It is not meant to compete with full frameworks. It is meant to answer a narrower question: How small can a usable neural inference stack get while remaining practical? https://github.com/AlexRosito67/xyron-mnist-esp32 This started as an experiment to understand perceptrons and a thermometer, it ended as a small toolchain for compiling neural networks into embedded firmware. And I'm still not entirely sure where the boundary between "software" and "firmware" is anymore.