Whisper-khanacademy-large-v3-turbo-tr Open-Source Speech Recognition Model - Accurately Identify Turkish Speech Content

Whisper Khanacademy Large V3 Turbo Tr

Developed by ysdede

An automatic speech recognition (ASR) model fine-tuned on Turkish Khan Academy dataset based on OpenAI Whisper-large-v3-turbo

Speech Recognition

Transformers

OtherOpen Source License:MIT #Turkish ASR #Educational transcription #Low word error rate

Downloads 31

Release Time : 12/12/2024

Model Overview

This model specializes in Turkish speech recognition tasks, particularly optimized for STEM (Science, Technology, Engineering, Mathematics) educational content

Model Features

Turkish Optimization

Specially fine-tuned for Turkish speech recognition to improve accuracy

Educational Content Adaptation

Trained on Khan Academy STEM educational content dataset, suitable for academic and educational scenarios

High Performance

Achieves 15.7% word error rate (WER) on evaluation set with excellent performance

Model Capabilities

Turkish speech-to-text

Educational content recognition

STEM terminology recognition

Use Cases

Education

Online course subtitle generation

Automatically generate subtitles for Turkish STEM educational videos

15.7% improvement in word recognition accuracy

Educational content transcription

Transcribe Turkish teaching audio into text

🚀 whisper-khanacademy-large-v3-turbo-tr

This model is a fine-tuned version of openai/whisper-large-v3-turbo on the ysdede/khanacademy-turkish dataset. It offers high - performance automatic speech recognition capabilities, especially useful for Turkish language processing.

🚀 Quick Start

This model is a fine - tuned version of openai/whisper-large-v3-turbo on the ysdede/khanacademy-turkish dataset. It achieves the following results on the evaluation set:

Loss: 0.2129
Wer: 15.6951

📚 Documentation

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e - 06
train_batch_size: 64
eval_batch_size: 32
seed: 42
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon = 1e - 08 and optimizer_args = No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.15
training_steps: 1204
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2298	0.1429	172	0.2418	16.5877
0.2157	0.2857	344	0.2255	15.9611
0.1668	1.0939	516	0.2227	15.7461
0.1752	1.2367	688	0.2159	15.8846
0.1492	2.0449	860	0.2187	15.7571
0.1592	2.1877	1032	0.2134	15.5421
0.1336	2.3306	1204	0.2129	15.6951

Framework versions

Transformers 4.48.0.dev0
Pytorch 2.5.1+cu121
Datasets 3.2.0
Tokenizers 0.21.0

📄 License

This model is released under the MIT license.

📦 Information Table

Property	Details
Library Name	transformers
Language	tr
License	mit
Base Model	openai/whisper-large-v3-turbo
Tags	generated_from_trainer
Datasets	khanacademy, turkish, stem, asr
Metrics	wer
Model Index Name	whisper-khanacademy-large-v3-turbo-tr
Task Name	Automatic Speech Recognition
Task Type	automatic-speech-recognition
Dataset Name	ysdede/khanacademy-turkish
Dataset Type	khanacademy
Metric Name	Wer
Metric Type	wer
Metric Value	15.695132614398135

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