This is iVaani model, specifically optimized for the Hindi language. The fine-tuning process has led to an improvement in accuracy by 2.5% compared to the original Whisper model.
After fine-tuning, the model shows a 2.5% increase in transcription accuracy for Hindi language audio compared to the base Whisper medium model.
You can use this model directly with a simple API call in Hugging Face. Here is a Python code snippet for using the model:
1from transformers import AutoModelForCTC, Wav2Vec2Processor
2
3model = AutoModelForCTC.from_pretrained("rukaiyah-indika-ai/iVaani")
4processor = Wav2Vec2Processor.from_pretrained("rukaiyah-indika-ai/iVaani")
5
6# Replace 'path_to_audio_file' with the path to your Hindi audio file
7input_audio = processor(path_to_audio_file, return_tensors="pt", padding=True)
8
9# Perform the transcription
10transcription = model.generate(**input_audio)
11print("Transcription:", transcription)
Indika AI has also fine-tuned ASR (Automatic Speech Recognition) models for several other Indic languages,
enhancing the accuracy by 2-5% for each language. The word error rate has also been significantly reduced.
1@misc{whisper-medium-hindi-fine-tuned,
2 author = {Indika AI},
3 title = {iVaani},
4 year = {2024},
5 publisher = {Hugging Face},
6 journal = {Hugging Face Model Hub}
7}
