TensorFold/GLM-5.3-Flash-MLX-2bit-MTP

Rebrand model card to TensorFold

Add Mac memory chooser and reproducible demo prompt

Release validated hybrid MLX 2-bit model with native MTP

initial commit