Instructions to use espnet/meld_cls_wavlm_base_plus with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- ESPnet
How to use espnet/meld_cls_wavlm_base_plus with ESPnet:
unknown model type (must be text-to-speech or automatic-speech-recognition)
- Notebooks
- Google Colab
- Kaggle
Download src/inference.py from espnet/meld_cls_wavlm_base_plus: direct link, hf CLI and curl.
- Browser
- Download file 390 Bytes
-
https://huggingface.co/espnet/meld_cls_wavlm_base_plus/resolve/main/src/inference.py
- Command line
-
hf download hf://espnet/meld_cls_wavlm_base_plus/src/inference.py
-
curl -L -o inference.py https://huggingface.co/espnet/meld_cls_wavlm_base_plus/resolve/main/src/inference.py
390 Bytes
| """Inference output helpers for the MELD classification recipe.""" | |
| def build_output(data, model_output, idx): | |
| """Build a dict of outputs for SCP writing.""" | |
| _, scores, hyp = model_output | |
| return { | |
| "utt_id": data.get("utt_id", str(idx)), | |
| "hyp": hyp, | |
| "ref": data.get("label", ""), | |
| "score": " ".join(str(score) for score in scores.tolist()), | |
| } | |