Distilhubert-finetuned-gtzan Open-source Audio Classification Model

Distilhubert Finetuned Gtzan 5 Epochs Finetuned Gtzan Finetuned Gtzan

Developed by duysal

An audio classification model based on the DistilHuBERT architecture, fine-tuned on the GTZAN dataset for music genre classification tasks.

Audio Classification

Transformers

#Music Classification #High Accuracy #Audio Processing

Downloads 5

Release Time : 2/2/2025

Model Overview

This model is a lightweight audio classification model based on the DistilHuBERT architecture, specifically fine-tuned on the GTZAN music dataset, capable of accurately identifying 10 different music genres.

Model Features

Efficient and Lightweight

Based on the DistilHuBERT architecture, it reduces model complexity while maintaining high accuracy

High Accuracy

Achieved 85% accuracy on the GTZAN validation set

Fast Training

Good performance can be obtained in just 5 epochs

Model Capabilities

Music Genre Classification

Audio Feature Extraction

Use Cases

Music Recommendation System

Automatic Music Classification

Used for automatically classifying uploaded music files on music streaming platforms

85% classification accuracy

Music Analysis

Music Style Research

Assists musicology research on the characteristics of different genres

🚀 distilhubert-finetuned-gtzan-5-epochs-finetuned-gtzan-finetuned-gtzan

This is a fine - tuned model based on duysal/distilhubert - finetuned - gtzan - 5 - epochs - finetuned - gtzan, designed for audio classification tasks on the GTZAN dataset, achieving high accuracy.

🚀 Quick Start

This model is a fine - tuned version of duysal/distilhubert-finetuned-gtzan-5-epochs-finetuned-gtzan on the GTZAN dataset. It achieves the following results on the evaluation set:

Loss: 0.5765
Accuracy: 0.85

📚 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: 3e - 05
train_batch_size: 2
eval_batch_size: 2
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 8
optimizer: Use OptimizerNames.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: 0.1
num_epochs: 3
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.1601	1.0	113	0.5812	0.83
0.1855	2.0	226	0.5522	0.86
0.0665	2.9778	336	0.5765	0.85

Framework versions

Transformers 4.49.0.dev0
Pytorch 2.5.1+cu124
Datasets 3.2.0
Tokenizers 0.21.0

📄 License

No license information provided in the original document.

🔧 Technical Details

The model distilhubert-finetuned-gtzan-5-epochs-finetuned-gtzan-finetuned-gtzan is fine - tuned from the base model duysal/distilhubert-finetuned-gtzan-5-epochs-finetuned-gtzan on the GTZAN dataset. During training, specific hyperparameters are set, such as a learning rate of 3e - 05, a train batch size of 2, and 3 training epochs. These settings contribute to the model's performance on the evaluation set, with a loss of 0.5765 and an accuracy of 0.85. The training results show the changes in training loss, validation loss, and accuracy over different epochs and steps, which can help users understand the model's training process and performance trends.

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