BitNetMCU: Implementing MNIST handwritten digit recognition on the CH32V003 RISC-V MCU

BitNetMCU is a project focused on low-bit quantized neural network training and inference, specifically designed to run efficiently on low-end microcontrollers such as the CH32V003. Quantization-aware training (QAT) and fine-tuning of the model structure and inference code allow achieving over 99% test accuracy on the 16x16 MNIST dataset without using multiplication instructions and requiring only 2kb RAM and 16kb Flash. The training pipeline is based on PyTorch and can be run anywhere. The inference engine is implemented in Ansi-C and can be easily ported to any microcontroller. BitNetMCU is a project focused on the training and inference of low-bit quantized neural networks, specifically designed to run efficiently on low-end microcontrollers like the CH32V003. Quantization aware training (QAT) and fine-tuning of model structure and inference code allowed surpassing 99% Test accuracy on a 16x16 MNIST dataset without using multiplication instructions and in only 2kb of RAM and 16kb of Flash. The training pipeline is based on PyTorch and should run anywhere. The inference engine is implemented in Ansi-C and can be easily ported to any Microcontroller.