Run 35B MoE models in ~2.5 GB of RAM. Weights stream from SSD; a trained router predicts loads a step ahead. Mac, iPhone, Android, Windows supported
Python
"""MLX backend — the reference backend implementation.
Everything that touches ``mlx`` lives under this package. Nothing else in
the framework may import mlx directly.
"""
from edge0.backends.mlx.backend import BackendImpl, core, io, nn, quant
__all__ = ["BackendImpl", "core", "io", "nn", "quant"]