Profile README
-
Updated
Mar 22, 2026
Profile README
Ternary CNN implementation for CIFAR-10: PyTorch model with {-1, 0, +1} constrained weights + AVX-VNNI optimized C++ inference engine. Includes training pipeline, bitpacked weight export, and full benchmarking suite.
Bare-metal 1.58-bit ternary CNN inference engine on ESP32-S3. No framework. 5.45ms inference. Runs own WiFi AP. Pure C.
Add a description, image, and links to the ternary-neural-network topic page so that developers can more easily learn about it.
To associate your repository with the ternary-neural-network topic, visit your repo's landing page and select "manage topics."