Description

A hardware and software configuration utilizing an RTX 3090 connected to an M4 Mac Mini via Tinygrad's chestnut. The setup uses the Strata framework to manage model weights, achieving ~80 tok/s decode speeds and high-speed prefilling with 2-bit quantization on the Qwen 3.8 Next Flash model.

Descriptions, tags, and model credits may be AI-generated or inferred from public sources and can be incomplete or wrong. Learn more.