AIbase
EN
Home
Whisper Large V2 Serbian
W

Whisper Large V2 Serbian

Developed by DrishtiSharma
This is a speech recognition model fine-tuned on the Serbian Common Voice 11.0 dataset based on OpenAI Whisper Large-V2, achieving a word error rate of 10.76%.
Speech Recognition
Transformers
 OtherOpen Source License:Apache-2.0#Serbian ASR#Low Word Error Rate (10.76%)#Common Voice Fine-tuning
Downloads 39
Release Time : 12/20/2022

Model Overview

This model is specifically designed for automatic speech recognition tasks in Serbian, optimized for the Serbian language based on the open-source Whisper Large-V2 architecture.

Model Features

Serbian Language Optimization
Specially fine-tuned for Serbian, offering better localization performance compared to the original Whisper model.
Low Word Error Rate
Achieves a word error rate of 10.76% on the Common Voice test set.
Based on Whisper Architecture
Inherits the powerful speech recognition capabilities and robustness of the Whisper model.

Model Capabilities

Serbian Speech Recognition
Speech-to-Text Conversion
Long Audio Processing

Use Cases

Speech Transcription
Serbian Meeting Minutes
Automatically transcribe Serbian meeting recordings into text records.
Accuracy approximately 89.24% (inferred from WER).
Voice Assistant
Provide speech recognition capabilities for Serbian voice assistants.
Media Processing
Video Subtitle Generation
Automatically generate subtitles for Serbian video content.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase