Drop MLX: 5.13x/2.25x slower than Core ML fp16 with no size win

Core ML: replace the int8 variant with fp16

Add per-backend config

Add mlx export (ExecuTorch 1.4.1, fork toolchain)

Add stub root config.json for HF download counter

Add spec-compliant config.json files

Add spec-compliant config.json files

Restructure to MODEL_SPEC.md convention

Delete coreml/config.json

Upload rfdetr_n_det_coreml_int8.pte

Create coreml/config.json

chore: fix rfdetr link

add model & readme

initial commit