Pre-Trained...: This project implements a pre-trained neural network for hand-written digits from MNIST dataset.

Clone this repo:

Branches

  1. ef63286 final gds oasis by Jeff DiCorpo · 2 years, 4 months ago main
  2. 3aa5f74 add README by mxiangyue · 2 years, 6 months ago
  3. 0414272 add README by mxiangyue · 2 years, 6 months ago
  4. b1511c5 hardened by mxiangyue · 2 years, 6 months ago
  5. 380da1b add rtl srouce by mxiangyue · 2 years, 6 months ago

Pre-Trained Neural Network for MNIST

This project implements a pre-trained fully connected neural network (FCNN) for hand-written digits from MNIST dataset. It takes images from MNIST dataset (28x28) and output single digit prediction. All parameters (weights and bias) are pre-trained and hard-coded. User can use wishbone bus to pass in inputs and obtain outputs.

This repository contains a Verilog implementation of the FCNN, targeting the OpenLane flow for the Skywater 130nm process and is an entry for the Google/Efabless MPW-3 shuttle.