DistilHubert - Finetuned GTZAN Open - Source Audio Classification Model: Accurately Classify Music with High Precision

Distilhubert Finetuned Gtzan

Developed by Marco-Cheung

This model is an audio classification model fine-tuned on the GTZAN music classification dataset based on distilhubert, achieving an accuracy of 83%

Audio Classification

Transformers

Open Source License:Apache-2.0 #Music Classification #High Accuracy #Lightweight Audio Model

Downloads 17

Release Time : 7/22/2023

Model Overview

A lightweight audio classification model based on the DistilHuBERT architecture, specifically optimized for music genre classification tasks

Model Features

High Accuracy

Achieves 83% classification accuracy on the GTZAN test set

Lightweight

Based on the distilled HuBERT architecture with a smaller model size

Specialized for Music Classification

Specifically optimized for music genre classification tasks

Model Capabilities

Music Genre Classification

Audio Feature Extraction

Use Cases

Music Analysis

Automatic Music Genre Tagging

Automatically add genre labels to tracks in a music library

Accuracy reaches 83%

Music Recommendation System

Improve recommendation algorithms based on music genre features

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.9825	1.0	113	1.7658	0.48
1.2943	2.0	226	1.2478	0.65
0.9837	3.0	339	0.9757	0.71
0.8201	4.0	452	0.8420	0.72
0.5363	5.0	565	0.6741	0.83
0.3417	6.0	678	0.7083	0.76
0.4129	7.0	791	0.5941	0.81
0.1681	8.0	904	0.5954	0.84
0.2398	9.0	1017	0.5819	0.85
0.1346	10.0	1130	0.5933	0.83

Property	Details
Model Type	Fine - tuned version of ntu - spml/distilhubert
Training Data	marsyas/gtzan
Metrics	accuracy

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

Distilhubert Finetuned Gtzan

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 distilhubert-finetuned-gtzan

🚀 Quick Start

📚 Documentation

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

📄 License

📦 Model Information