---
language:
- ml
- ta
license: apache-2.0
base_model: openai/whisper-large-v3
tags:
- whisper
- asr
- malayalam
- tamil
- dravidian languages
- speech-recognition
- fine-tuned
- lora

datasets:
- google/fleurs
metrics:
- wer
pipeline_tag: automatic-speech-recognition
---

# Whisper Large v3 -- Malayalam (Fine-tuned)

A fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3)
for **Malayalam (ml)** and **Tamil (ta)** automatic speech recognition, trained with LoRA adapters and merged
into a full PyTorch model.

## Usage

'''python
from transformers import pipeline

pipe = pipeline(
    "automatic-speech-recognition",
    model="rontroy/whisper-large-v3-malayalam",
    generate_kwargs={"language": "malayalam", "task": "transcribe"},
)

result = pipe("your_audio.wav")
print(result["text"])
'''

## Model Details

- **Base model**: openai/whisper-large-v3
- **Fine-tuning method**: LoRA (merged into full weights)
- **Language**: Malayalam (ml)
- **Task**: Automatic Speech Recognition (transcribe)

## CTranslate2 / faster-whisper version

A faster-whisper compatible version is available at
[rontroy/whisper-large-v3-malayalam-ct2](https://huggingface.co/rontroy/whisper-large-v3-malayalam-ct2).