---
title: Cocktail-Fork-MRX-adapted-eq
canonical_url: "https://www.modelscope.cn/models/mlx-community/Cocktail-Fork-MRX-adapted-eq"
md_url: "https://www.modelscope.cn/models/mlx-community/Cocktail-Fork-MRX-adapted-eq.md"
repository: mlx-community/Cocktail-Fork-MRX-adapted-eq
last_updated: 2026-06-06
license: mit
model_type:
  - mrx
parameters: 30.6M
tensor_type:
  - F32
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 10
stars: 0
tags:
  - mlx
  - audio
  - audio-source-separation
  - speech
  - music
  - apple-silicon
---

# Cocktail-Fork-MRX-adapted-eq

> Cocktail-Fork-MRX-adapted-eq - mlx-community 在 ModelScope 开源的模型。Cocktail-Fork-MRX — adapted-eq variant (MLX)

mlx-community/Cocktail-Fork-MRX-adapted-eq 是 ModelScope 魔搭社区上的 30.6M 参数机器学习模型，采用 mit 许可。

- **Repository**: mlx-community/Cocktail-Fork-MRX-adapted-eq
- **License**: mit
- **Parameters**: 30.6M
- **Tags**: mlx, audio, audio-source-separation, speech, music, apple-silicon
- **Downloads**: 10
- **Stars**: 0
- **Last updated**: 2026-06-06

Source: https://www.modelscope.cn/models/mlx-community/Cocktail-Fork-MRX-adapted-eq

---

# Cocktail-Fork-MRX — `adapted-eq` variant (MLX)

Apple **MLX** port of MERL's **MRX** (Multi-Resolution CrossNet) — separates a
soundtrack mixture into **music**, **speech**, and **sound effects (sfx)**.

**This variant** uses the `adapted_eq_` checkpoint: adapted with
**equalization-based normalization** for better alignment with real
cinematic/movie stems. Try this (and `-adapted-loudness`) on real soundtrack content.

Other variants: [`Cocktail-Fork-MRX`](https://huggingface.co/mlx-community/Cocktail-Fork-MRX) (default) ·
[`-paper`](https://huggingface.co/mlx-community/Cocktail-Fork-MRX-paper) (ICASSP reproduction) ·
[`-adapted-loudness`](https://huggingface.co/mlx-community/Cocktail-Fork-MRX-adapted-loudness).

- **Upstream:** [merlresearch/cocktail-fork-separation](https://github.com/merlresearch/cocktail-fork-separation) (MIT) · *The Cocktail Fork Problem*, ICASSP 2022.
- **Parity:** numerically exact vs PyTorch (full-forward max_abs ≈ `3e-7`).

## Usage

```bash
pip install git+https://github.com/xocialize/cocktail-fork-mlx
cocktail-fork-mlx --audio-path soundtrack.wav --out-dir ./out \
    --weights mlx-community/Cocktail-Fork-MRX-adapted-eq
```

~30.6M params, fp32 (122 MB), 44.1 kHz. MIT, © MERL for the original model/weights.