mstyslavity/optiq-mlx-my-model

fix: set clean upstream HF model_id in config and README YAML frontmatter

feat: enable ZeroGPU with spaces decorator and jang quantization

feat: switch to JANG quantization (works on Linux via NumPy fallback)

fix: remove audioop-lts (only needed on Python 3.13+, we use 3.12)

fix: replace optiq CLI with mlx-lm convert API (Linux compat)

Format stderr in code block for Gradio

Capture subprocess stderr

Set trust_remote_code=True for tokenizers

Update requirements.txt

Use valid python version 3.12

Set python_version as string with quotes

Fix python version for mlx-optiq

Initial commit for OptiQ Space