Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,27 @@
|
|
| 1 |
---
|
| 2 |
license: gpl-3.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: gpl-3.0
|
| 3 |
+
library_name: onnx
|
| 4 |
+
tags:
|
| 5 |
+
- audio
|
| 6 |
+
- source-separation
|
| 7 |
+
- karaoke
|
| 8 |
+
- onnx
|
| 9 |
+
- webgpu
|
| 10 |
---
|
| 11 |
+
|
| 12 |
+
# Stemlight models
|
| 13 |
+
|
| 14 |
+
ONNX models used by [Stemlight](https://github.com/Thulrus/stemlight), which separates songs into stems in the browser.
|
| 15 |
+
|
| 16 |
+
## bs-roformer-karaoke.onnx
|
| 17 |
+
|
| 18 |
+
Splits a vocal track into **lead** and **backing** vocals.
|
| 19 |
+
|
| 20 |
+
- **Original model:** [anvuew/karaoke_bs_roformer](https://huggingface.co/anvuew/karaoke_bs_roformer) (BS-RoFormer, trained on becruily's lead/backing vocal data)
|
| 21 |
+
- **Changes:** exported to ONNX (fp32) with 8-second chunks; some operations were rewritten so it runs on WebGPU via onnxruntime-web. The weights were not changed.
|
| 22 |
+
- **Size:** ~210 MB
|
| 23 |
+
- **SHA-256:** `7bbf6a275f8468efa2e084f50437ee476faf61ad7de0bd21d8df9c23dea318a5`
|
| 24 |
+
|
| 25 |
+
## License
|
| 26 |
+
|
| 27 |
+
GPL-3.0, the same as the original model. All credit for training goes to anvuew.
|