正在加载视频...
视频加载失败
Introducing that simple idea : ✨Split Audio Tracks to MusicGen✨ Use Demucs to split your favorite song into tracks, then send one to MusicGen to get a completely new music piece ! Give it a try Hugging Face
10 条评论
For demonstration purpose, here's an example with a PRO song you might recognize ;) — Split Audio Track to MusicGen @gradio demo :
Don’t forget to use the CROP feature from @gradio audio component to CROP your input audio file to your preferred sample before loading your conditioning track 😉
@huggingface Would be cool if you can take a 3 second piece and then keep generating the next 3 seconds and let people make longer melodies.
@huggingface Oh, that’s a very cool idea 😗 let me think about it ☺️
@huggingface and BTW Demucs is also from @honualx at FAIR 😊
@huggingface @honualx Yes indeed 🤗 Brilliant work Alexandre 👏
@huggingface This is brilliant @fffiloni !
@cerebral_valley @huggingface Stoked to try this on some discarded productions - breathe some new energy in.
@cerebral_valley @huggingface Glad that it will help you to rediscover old ideas 🤗 Share when ready !
@huggingface
相关视频
Sensitive content
Meta just released MusicGen, a simple and controllable model for music generation MusicGen is a single stage auto-regressive Transformer model trained over a 32kHz EnCodec tokenizer with 4 codebooks sampled at 50 Hz. Unlike existing methods like MusicLM, MusicGen doesn't not require a self-supervised semantic representation, and it generates all 4 codebooks in one pass. By introducing a small delay between the codebooks, can predict them in parallel, thus having only 50 auto-regressive steps per second of audio try out the Gradio demo: Models on Hugging Face: github:
AK
627,429 次观看 • 3 年前
