CapsBro IndicConformer speech models (sherpa-onnx int8)
AI4Bharat IndicConformer 600M multilingual, one int8 file per language, packaged for sherpa-onnx by Docglasses-AI. The files here are copied unchanged from that repository.
CapsBro uses them for Bengali, Gujarati, Kannada, Malayalam, Marathi, Punjabi, Tamil and Telugu captions.
| File | Language |
|---|---|
| indicconformer-bn.int8.onnx | Bengali |
| indicconformer-gu.int8.onnx | Gujarati |
| indicconformer-kn.int8.onnx | Kannada |
| indicconformer-ml.int8.onnx | Malayalam |
| indicconformer-mr.int8.onnx | Marathi |
| indicconformer-pa.int8.onnx | Punjabi |
| indicconformer-ta.int8.onnx | Tamil |
| indicconformer-te.int8.onnx | Telugu |
| indicconformer-tokens.txt | all of them |
License: MIT, the same as the original model. All credit for the model goes to AI4Bharat.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for nikh360pro/capsbro-indic-conformer
Base model
ai4bharat/indic-conformer-600m-multilingual