Whisper Large-v3 LoRA Fine-tuned for Code-Switching
This model is a LoRA fine-tuned version of openai/whisper-large-v3 for code-switching between English and Indian languages.
Supported Languages
- English, Hindi, Bengali, Marathi, Tamil, Telugu, Kannada, Malayalam
Training
- Base Model: openai/whisper-large-v3
- Method: LoRA (Low-Rank Adaptation) on encoder only
- Dataset: Synthetic code-switching data from FLEURS
- Samples: 10,000 mixed utterances
Usage
See inference examples in the repository.
