Medical-Whisper-Large-v3 Open-Source Medical Speech Recognition Model - Precise Recognition of Medical Scenario Speech

Medical Whisper Large V3

Developed by Na0s

Medical scenario-specific speech recognition model fine-tuned based on whisper-large-v3

EnglishOpen Source License:Apache-2.0 #Medical Speech Transcription #Low Error Rate ASR #Doctor-Patient Dialogue Recognition

Downloads 270

Release Time : 7/18/2024

Model Overview

Professionally optimized for medical transcription scenarios through transfer learning in doctor-patient consultation contexts

Model Features

Medical Scenario Optimization

Professionally tuned for doctor-patient consultation scenarios, improving recognition accuracy of medical terminology

High-Performance Recognition

Character Error Rate (CER) 0.10, Word Error Rate (WER) 0.19, significantly outperforming the base model

Large-Scale Parameters

Fine-tuned from the 1.55B parameter whisper-large-v3 model

Model Capabilities

Medical speech transcription

Doctor-patient dialogue recognition

Professional medical terminology recognition

Use Cases

Medical Records

Outpatient Consultation Records

Automatically transcribes conversations between doctors and patients

Character error rate reduced to 0.10

Medical Report Generation

Automatically generates structured medical reports from voice recordings

🚀 med-whisper-large-final

This model is a fine - tuned version of openai/whisper-large-v3, designed for medical transcription, offering high - performance speech recognition in the medical field.

🚀 Quick Start

This model is a fine - tuned version of openai/whisper-large-v3 on the primock_data dataset. It can be used for medical transcription tasks right away.

✨ Features

Fine - Tuned for Medical Use: It is a fine - tuned version of whisper - large - v3 through transfer learning on Doctor/Patient consultations, making it highly suitable for medical transcription.
High Performance: Achieves lower Word Error Rate (WER) and Character Error Rate (CER) compared to the base Whisper models in medical scenarios.

📚 Documentation

Model description

Fine tuned version of whisper - large - v3 through transfer learning on Doctor/Patient consultations.

Intended uses & limitations

Medical transcription.

Training and evaluation data

Na0s/Medical_Augmented_data

Training procedure

Exhaustive transfer learning

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 1e - 05
train_batch_size: 6
eval_batch_size: 6
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 24
optimizer: Adam with betas=(0.9,0.999) and epsilon = 1e - 08
lr_scheduler_type: constant_with_warmup
lr_scheduler_warmup_steps: 50
training_steps: 500
mixed_precision_training: Native AMP

Performance Overview

Model Name	WER	CER	Number of Parameters
Whisper Tiny	0.46	0.27	39M
Whisper Base	0.42	0.26	74M
Whisper Small	0.39	0.26	244M
Whisper Medium	0.37	0.23	769M
Whisper Large v3	0.33	0.18	1.55B
Whisper Medical	0.19	0.10	1.55B

Performance of foundation Whispers vs Medical Whisper on the Validation set.

Model Name	WER	CER	Number of Parameters
Whisper Medical	0.24	0.13	1.55B

Table: Performance of Whisper Medical on the Test set.

Framework versions

Transformers 4.42.4
Pytorch 2.3.1+cu121
Datasets 2.20.0
Tokenizers 0.19.1

📄 License

This model is licensed under the Apache 2.0 license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご