Speechbrain Lang Id Voxlingua107 Ecapa Onnx
This repository contains an ONNX export of speechbrain/lang-id-voxlingua107-ecapa, a SpeechBrain ECAPA-TDNN language-identification model trained for VoxLingua107 language classification.
Modellquelle
- Art
- Other
- Dateigröße
- 81,5 MB
- Basismodell
- speechbrain/lang-id-voxlingua107-ecapa
- Version
- vModel
- Aktualisiert
- 3. Nov. 2026
- Quelle
- Hugging Face
Quellenbeschreibung
SpeechBrain VoxLingua107 ECAPA Language ID ONNX
This repository contains an ONNX export of speechbrain/lang-id-voxlingua107-ecapa, a SpeechBrain ECAPA-TDNN language-identification model trained for VoxLingua107 language classification.
The upstream model is licensed under Apache 2.0. This export preserves that license and attribution.
Files
model.onnx: ONNX export of the SpeechBrain normalization, ECAPA embedding model, and classifier.labels.json: ordered language-code and label mapping for the 107 VoxLingua languages.
Export
The export can be reproduced from the upstream SpeechBrain checkpoint with:
uv run --with speechbrain --with torch --with torchaudio --with onnx --with onnxruntime --with onnxscript \
python scripts/export_speechbrain_language_id_onnx.py \
--out-dir /mnt/data/bai_weights/language_id/speechbrain-lang-id-voxlingua107-ecapa-onnx
Inputs should be converted to SpeechBrain-compatible 60-bin filterbank features before running model.onnx.
Quellen
1 Quelle
Quellen
1 QuelleVerifiziert 3. Nov.
Modellartefakte
1 Artefakt
Modellartefakte
1 ArtefaktQuellenauszüge
2 Auszüge
Quellenauszüge
2 Auszüge--- license: apache-2.0 base_model: speechbrain/lang-id-voxlingua107-ecapa library_name: onnxruntime pipeline_tag: audio-classification tags: - speechbrain - onnx - onnxruntime - audio-classification - language-identification - VoxLingua107 --- # SpeechBrain VoxLingua107 ECAPA Language ID ONNX This repository contains an ONNX export of [`speechbrain/lang-id-voxlingua107-ecapa`](https://huggingface.co/speechbrain/lang-id-voxlingua107-ecapa), a SpeechBrain ECAPA-TDNN language-identification model trained for VoxLingua107 language classification. The upstream model is licensed under Apache 2.0. This export preserves that license and attribution. ## Files - `model.onnx`: ONNX export of the SpeechBrain normalization, ECAPA embedding model, and classifier. - `labels.json`: ordered language-code and label mapping for the 107 VoxLingua languages. ## Export The export can be reproduced from the upstream SpeechBrain checkpoint with: ```bash uv run --with speechbrain --with torch --with torchaudio --with onnx --with onnxruntime --with onnxscript \ python scripts/export_speechbrain_language_id_onnx.py \ --out-dir /mnt/data/bai_weights/language_id/speechbrain-lang-id-voxlingua107-ecapa-onnx ``` Inputs should be converted to SpeechBrain-compatible 60-bin filterbank features before running `model.onnx`.
Source context: 0 downloads · 0 likes · Pipeline audio-classification · Library onnxruntime · Repo beginning-ai/speechbrain-lang-id-voxlingua107-ecapa-onnx
