W

Whisper Large V2 Punjabi

Developed by DrishtiSharma
Punjabi automatic speech recognition model fine-tuned on OpenAI Whisper-large-v2, trained on Common Voice 11.0 dataset
Downloads 27
Release Time : 12/15/2022

Model Overview

This model is a Punjabi-optimized automatic speech recognition (ASR) model, fine-tuned based on OpenAI's Whisper-large-v2 architecture, suitable for Punjabi speech-to-text tasks.

Model Features

High-accuracy Punjabi recognition
Achieves 19.71% word error rate (WER) on Common Voice 11.0 test set
Based on Whisper architecture
Utilizes OpenAI Whisper-large-v2's powerful base model for fine-tuning
Open-source license
Released under Apache 2.0 license, allowing commercial and research use

Model Capabilities

Punjabi speech recognition
Speech-to-text
Long-form audio processing

Use Cases

Speech transcription
Punjabi content transcription
Convert Punjabi podcasts, interviews and other audio content into text
Approximately 80.29% accuracy (based on WER calculation)
Assistive technology
Punjabi subtitle generation
Automatically generate subtitles for Punjabi videos
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase