AbleCredit/Ablecredit-Whisper-Small

Whispering Voices: Fine-Tuning for Indic-to-English Translation

Model Name: Ablecredit-Whisper-Small

• Model Overview:

  This is a fine-tuned version of OpenAI's Whisper model, trained to perform speech-to-text translation from Indic languages to English. The model has been optimized using open-source datasets and proprietary dataset created by AbleCredit.com to enhance transcription quality for Indic languages.

• Languages Supported

  The model is trained to handle speech-to-text translation for the following Indic languages:

  • Hindi (hi)
  • Marathi (mr)
  • Gujarati (gu)
  • Tamil (ta)
  • Telugu (te)
  • Kannada (kn)

Fine-Tuning Process

The base Whisper model was fine-tuned using a carefully curated dataset consisting of open-source corpus and a well curated internal dataset from AbleCredit. The dataset underwent preprocessing, including text normalization, score based filtering and feature engineering.

• Key Improvements After Fine-Tuning

  📉 Word Error Rate (WER) Reduction: 54% lower WER compared to the base model.

  📈 BLEU Score Increase: 5.5× higher BLEU score, indicating improved translation accuracy.

Evaluation Metrics

• Word Error Rate (WER) WER is used to measure the error rate in transcriptions. A lower WER means better transcription accuracy.

  WER Reduction: 54% improvement over the base Whisper model.

• BLEU Score The BLEU (Bilingual Evaluation Understudy) Score is a metric used to evaluate machine translation quality. Higher BLEU scores indicate better translation quality.

  BLEU Score Improvement: 5.5× higher BLEU compared to the base model.

Use Cases Transcribing Indic-language speech into English text

Connect with me over LinkedIn

AbleCredit