W

Whisper Medium Portuguese

Developed by pierreguillou
A Portuguese speech recognition model fine-tuned on the common_voice_11_0 dataset based on openai/whisper-medium, with a word error rate of 6.5987
Downloads 191
Release Time : 12/15/2022

Model Overview

An automatic speech recognition model optimized for Portuguese, capable of accurately transcribing Portuguese audio into text

Model Features

High-performance Portuguese recognition
Achieves a word error rate of only 6.5987 on the Common Voice 11.0 test set, outperforming the original Whisper model
Based on Whisper architecture
Utilizes the OpenAI Whisper-medium architecture with powerful speech recognition capabilities
Open-source license
Released under the Apache 2.0 license, allowing free use and modification

Model Capabilities

Portuguese speech recognition
Audio transcription
Automatic speech-to-text

Use Cases

Speech transcription
Portuguese meeting minutes
Automatically transcribes Portuguese meeting recordings into text records
Accuracy up to 93.4%
Podcast content transcription
Converts Portuguese podcast content into searchable text
Assistive tools
Hearing impairment assistance
Provides real-time speech-to-text services for the hearing impaired
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase