ondeviceai
Fast KV-cache quantization for Apple Silicon (MLX) — 43 research-adapted compression methods with Metal kernels