README: point to perch2-torch-embed (embedding-only torch.export file)
Browse files
README.md
CHANGED
|
@@ -18,6 +18,11 @@ tags:
|
|
| 18 |
PyTorch implementation of the complete PERCH 2 waveform model for
|
| 19 |
bioacoustic classification and 1536-dimensional audio embeddings.
|
| 20 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
## Load From Hugging Face
|
| 22 |
|
| 23 |
```python
|
|
|
|
| 18 |
PyTorch implementation of the complete PERCH 2 waveform model for
|
| 19 |
bioacoustic classification and 1536-dimensional audio embeddings.
|
| 20 |
|
| 21 |
+
> **Only need embeddings?** [perch2-torch-embed](https://huggingface.co/DAS-DeepAudioSegmenter/perch2-torch-embed)
|
| 22 |
+
> has the same frontend and backbone without the label head, as a single `torch.export` file (47.6 MB instead of
|
| 23 |
+
> 408 MB). It loads with PyTorch alone, and comes with a small helper that returns avg-pooled, max-pooled and temporal
|
| 24 |
+
> embeddings for audio of any length.
|
| 25 |
+
|
| 26 |
## Load From Hugging Face
|
| 27 |
|
| 28 |
```python
|