# `EMLX.Backend`
[🔗](https://github.com/elixir-nx/emlx/blob/v0.3.0/emlx/lib/emlx/backend.ex#L1)

EMLX backend for Nx tensors, providing GPU acceleration via Apple's MLX.

For quantized inference, see `EMLX.Quantization` and
`EMLX.Quantization.Config`.

# `from_nx`

Converts from an Nx tensor to an MLX array.

# `to_nx`

Converts an MLX array to an Nx tensor.

# `to_nx`

Converts an MLX array back to an Nx tensor with type and shape assertions.

# `write_finite_f8`

# `write_non_finite`

---

*Consult [api-reference.md](api-reference.md) for complete listing*