Add docker_mlx_cpp — GPU containers for Apple Silicon (MLX)

[ilessiorobotflowlabs]

Project: docker_mlx_cpp
Category: Tools / Infrastructure / Apple Silicon

What: The NVIDIA Container Toolkit equivalent for Mac. Any Docker container gets Metal GPU access — 107 operations across arithmetic, linear algebra, FFT, convolutions, attention, normalization, etc. Plus LLM inference (50+ architectures), training (LoRA), image generation (FLUX), audio (Whisper, TTS), and embeddings.

Why it matters: On Linux, nvidia-docker --gpus all is standard. On Mac, there was nothing equivalent until now. This bridges the gap for the growing Apple Silicon ML community.

Install: curl -fsSL https://raw.githubusercontent.com/RobotFlow-Labs/docker_mlx_cpp/main/install.sh | bash

Benchmarks (M5): ~95 TFLOPS matmul, Flash Attention 1.6ms. MIT licensed.