[REQ] Please also add some other supports
#29
by
dattazigzag
- opened
- Apple MLX backend for running on Apple machines to best use the Unified mem and Apple GPUs
- ONNX version if possible for better non-gpu sessions and optimizations
- llama-cpp backend for better non-gpu sessions
We are working with the Hugging Face team to bring Dia to transformers! Hopefully will get it out soon.
MLX, ONNX and llama-cpp are also very exciting!
Sorry for the wait and thank you for the interest :))))
It works nice but at the end audio voice sliced just add +1 second or 0.5 and it will be perfect. Thanks