Core Epoch

Quantized .onnx computer vision models that are 58% to 68% smaller than FP32 and up to 70% faster on one CPU thread, while keeping high accuracy. Every model runs on both ONNX Runtime and OpenVINO from the same file.

All models quantized with Kenosis, Core Epoch's proprietary post-training quantizer. For licensing inquiries or custom model quantization: coreepoch.dev/kenosis · licensing@coreepoch.dev