Difference between this and the MLX Community version?
#1
by
mohamedashrefbna
- opened
May I kindly ask what's the difference between this quant and the MLX Community version? The readme appears to not reflect any change?
The only difference being this version contains tokenizer.json which was required when I doing an mlx swift port of the model. mlx-community/chatterbox-turbo-4bit works with mlx python but not with Swift. There's also this: mlx-community/Chatterbox-Turbo-TTS-4bit that also includes tokenizer.json with added_tokens.json combined. If you're looking for something that'll work with Swift, then use mlx-community/Chatterbox-Turbo-TTS-4bit or its variants as mine was created for development/testing.