Docker Facebook Demucs
| Project Name | Stars | Downloads | Repos Using This | Packages Using This | Most Recent Commit | Total Releases | Latest Release | Open Issues | License | Language |
|---|---|---|---|---|---|---|---|---|---|---|
| Netdata | 65,170 | a day ago | 359 | gpl-3.0 | C | |||||
| Monitor your servers, containers, and applications, in high-resolution and in real-time! | ||||||||||
| Jina | 19,103 |  | 13 | 2 days ago | 2,421 | July 30, 2023 | 14 | apache-2.0 | Python | |
| ☁️ Build multimodal AI applications with cloud-native stack | ||||||||||
| Awesome Kubernetes | 14,249 | 6 days ago | 7 | other | Shell | |||||
| A curated list for awesome kubernetes sources :ship::tada: | ||||||||||
| Mlcourse.ai | 8,803 | 4 months ago | 4 | other | Python | |||||
| Open Machine Learning Course | ||||||||||
| Doccano | 8,291 | 2 days ago | 32 | July 20, 2023 | 269 | mit | Python | |||
| Open source annotation tool for machine learning practitioners. | ||||||||||
| Cog | 5,601 | 17 hours ago | 106 | August 07, 2023 | 296 | apache-2.0 | Python | |||
| Containers for machine learning | ||||||||||
| Fate | 5,200 |  | 1 | 3 days ago | 30 | April 18, 2022 | 786 | apache-2.0 | Python | |
| An Industrial Grade Federated Learning Framework | ||||||||||
| Pipeline | 4,159 | a year ago | 85 | July 18, 2017 | 1 | apache-2.0 | Jsonnet | |||
| PipelineAI Kubeflow Distribution | ||||||||||
| Deeplearningproject | 4,043 | 3 years ago | 3 | mit | HTML | |||||
| An in-depth machine learning tutorial introducing readers to a whole machine learning pipeline from scratch. | ||||||||||
| Orchest | 3,876 | 4 months ago | 19 | December 13, 2022 | 125 | apache-2.0 | TypeScript | |||
| Build data pipelines, the easy way 🛠️ | ||||||||||
Docker Facebook Demucs
This repository dockerizes Facebook Demucs to split music tracks into different tracks (bass, drums, voice, others).
Usage
Clone this repository
git clone https://github.com/xserrat/docker-facebook-demucs.git demucs
Split a music track
- Copy the track you want to split into the
inputfolder (e.g.,input/mysong.mp3). - Execute
demucsvia therunjob in theMakefile, specifying thetrackargument with only the name of the file:
make run track=mysong.mp3
This process will take some time the first time it is run, as the execution will:
- Download the Docker image that is setup to run the
facebook demucsscript. - Download the pretrained models.
- Execute
demucsto split the track.
Subsequent runs will not need to download the Docker image or download the models, unless the model specified has not yet been used.
Options
The following options are available when splitting music tracks with the run job:
| Option | Default Value | Description |
|---|---|---|
gpu |
false |
Enable Nvidia CUDA support (requires an Nvidia GPU). |
model |
demucs |
The model used for audio separation. See facebookresearch/demucs for a list of available models to use. |
mp3output |
false |
Output separated audio in mp3 format instead of the default wav format. |
splittrack |
Individual track to split/separate from the others (e.g., you only want to separate drums). Valid options are bass, drums, vocals and other. Other values may be allowed if the model can separate additional track types. |
Example commands:
# Use the "fine tuned" demucs model
make run track=mysong.mp3 model=htdemucs_ft
# Enable Nvidia CUDA support and output separated audio in mp3 format
make run track=mysong.mp3 gpu=true mp3output=true
Run Interactively
To experiment with other demucs options on the command line, you can also run the Docker image interactively via the run-interactive job. Note that only the gpu option is applicable for this job.
Example:
make run-interactive gpu=true
Building the Image
The Docker image can be built locally via the build job:
make build
License
This repository is released under the MIT license as found in the LICENSE file.
