| | --- |
| | tags: |
| | - coreml |
| | - audio |
| | - music-source-separation |
| | - schism |
| | license: mit |
| | --- |
| | |
| | # TFC-TDF-UNet v3 (MDX23) |
| |
|
| | CoreML model for [Schism](https://github.com/zakkeown/Schism) audio separation framework. |
| |
|
| | MDX23 Model-A (TFC-TDF-UNet v3) multi-source separation with subbands |
| |
|
| | ## Details |
| |
|
| | | Property | Value | |
| | |----------|-------| |
| | | Expert IDs | tfctdfunet-vocals, tfctdfunet-drums, tfctdfunet-bass, tfctdfunet-other | |
| | | Tasks | vocalSeparation, stemSeparation | |
| | | Quality | high | |
| | | Est. RTF | 0.6 | |
| | | Peak Memory | 350 MB | |
| |
|
| | ## Usage |
| |
|
| | Install to your local Schism model store: |
| |
|
| | ```bash |
| | schism models install tfctdfunet |
| | ``` |
| |
|
| | Or manually place in `~/.schism/models/tfctdfunet/v1.0.0/`. |
| |
|
