L

Lite Whisper Large V3 Acc

Developed by efficient-speech
Lite-Whisper is a compressed version of OpenAI Whisper, utilizing LiteASR technology to reduce model size while maintaining high accuracy.
Downloads 57
Release Time : 2/26/2025

Model Overview

A compressed version based on Whisper-large-v3, focused on automatic speech recognition tasks. It reduces model parameters by optimizing the encoder structure while maintaining recognition accuracy close to the original model.

Model Features

Efficient Compression
Uses LiteASR technology to significantly reduce model size, compressing encoder parameters from 635M to 429M
Maintains High Accuracy
Achieves an average WER of 10.1 on the ESB dataset, comparable to the original whisper-large-v3 performance
Fast Inference
The compressed model should have faster inference speed compared to the full version (specific data not provided)

Model Capabilities

Speech-to-text
Multilingual speech recognition
Long audio processing

Use Cases

Speech Transcription
Meeting Minutes
Automatically converts meeting recordings into text transcripts
Accuracy comparable to the original Whisper model
Subtitle Generation
Automatically generates subtitles for video content
Voice Assistants
Voice Command Recognition
Used in voice command recognition systems for smart devices
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase