Aller au contenu
ModelPrésentéExterne
Télécharger

Distilhubert Finetuned Gtzan

This model is a fine-tuned version of ntu-spml/distilhubert on the GTZAN dataset. It achieves the following results on the evaluation set: Loss: 0.7235 Accuracy: 0.87

Sourcehuggingface.co/MohammadDaleen/distilhubert-finetuned-gtzan/tree/b4e897eb8a865a8883aaab2dcb51b41ec0b0db3e
TypeOtherModèle de basentu-spml/distilhubertVersionRevision b4e897eBClasse

Source du modèle

Type
Other
Taille du fichier
90,4 MB
Modèle de base
ntu-spml/distilhubert
Version
vRevision b4e897e
Mis à jour
25 oct. 2026
Téléchargements
6
Source
Hugging Face

Description de la source

distilhubert-finetuned-gtzan

This model is a fine-tuned version of ntu-spml/distilhubert on the GTZAN dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7235
  • Accuracy: 0.87

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-05
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 20
  • mixed_precision_training: Native AMP

Training results

Training LossEpochStepValidation LossAccuracy
2.1521.01132.07480.59
1.38452.02261.45180.61
1.10963.03391.04500.68
0.7034.04520.83850.77
0.54795.05650.73310.82
0.47136.06780.59850.8
0.23927.07910.55820.8
0.0888.09040.72880.83
0.09769.010170.60880.83
0.057510.011300.59130.86
0.015311.012430.75030.85
0.007312.013560.81630.86
0.005813.014690.72350.87
0.004914.015820.84020.86
0.003615.016950.83220.86
0.003516.018080.75590.87
0.003317.019210.71480.87
0.00318.020340.83810.87
0.002919.021470.80510.87
0.002920.022600.78470.87

Framework versions

  • Transformers 4.57.6
  • Pytorch 2.11.0+cu128
  • Datasets 3.6.0
  • Tokenizers 0.22.2

Sources

1 source
Hugging FacePrincipalVersion b4e897eb8a865a8883aaab2dcb51b41ec0b0db3e
huggingface.co/MohammadDaleen/distilhubert-finetuned-gtzan/tree/b4e897eb8a865a8883aaab2dcb51b41ec0b0db3e

Vérifié 25 oct.

Artefacts du modèle

1 artefact

model.safetensors

safetensors · 90,4 MB · SHA-256 74ad1aac4bfe…834b · Hugging Face

Download file

Extraits de sources

3 extraits
Readme7 aoûthuggingface.co/MohammadDaleen/distilhubert-finetuned-gtzan

--- library_name: transformers license: apache-2.0 base_model: ntu-spml/distilhubert tags: - generated_from_trainer datasets: - marsyas/gtzan metrics: - accuracy model-index: - name: distilhubert-finetuned-gtzan results: - task: name: Audio Classification type: audio-classification dataset: name: GTZAN type: marsyas/gtzan config: default split: train args: default metrics: - name: Accuracy type: accuracy value: 0.87 --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # distilhubert-finetuned-gtzan This model is a fine-tuned version of [ntu-spml/distilhubert](https://huggingface.co/ntu-spml/distilhubert) on the GTZAN dataset. It achieves the following results on the evaluation set: - Loss: 0.7235 - Accuracy: 0.87 ## 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-05 - train_batch_size: 8 - eval_batch_size: 8 - seed: 42 - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 20 - mixed_precision_training: Native AMP ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | |:-------------:|:-----:|:----:|:---------------:|:--------:| | 2.152 | 1.0 | 113 | 2.0748 | 0.59 | | 1.3845 | 2.0 | 226 | 1.4518 | 0.61 | | 1.1096 | 3.0 | 339 | 1.0450 | 0.68 | | 0.703 | 4.0 | 452 | 0.8385 | 0.77 | | 0.5479 | 5.0 | 565 | 0.7331 | 0.82 | | 0.4713 | 6.0 | 678 | 0.5985 | 0.8 | | 0.2392 | 7.0 | 791 | 0.5582 | 0.8 | | 0.088 | 8.0 | 904 | 0.7288 | 0.83 | | 0.0976 | 9.0 | 1017 | 0.6088 | 0.83 | | 0.0575 | 10.0 | 1130 | 0.5913 | 0.86 | | 0.0153 | 11.0 | 1243 | 0.7503 | 0.85 | | 0.0073 | 12.0 | 1356 | 0.8163 | 0.86 | | 0.0058 | 13.0 | 1469 | 0.7235 | 0.87 | | 0.0049 | 14.0 | 1582 | 0.8402 | 0.86 | | 0.0036 | 15.0 | 1695 | 0.8322 | 0.86 | | 0.0035 | 16.0 | 1808 | 0.7559 | 0.87 | | 0.0033 | 17.0 | 1921 | 0.7148 | 0.87 | | 0.003 | 18.0 | 2034 | 0.8381 | 0.87 | | 0.0029 | 19.0 | 2147 | 0.8051 | 0.87 | | 0.0...

Note25 oct.huggingface.co/MohammadDaleen/distilhubert-finetuned-gtzan

Source context: 6 downloads · 0 likes · Pipeline audio-classification · Library transformers · Repo MohammadDaleen/distilhubert-finetuned-gtzan

Note7 aoûthuggingface.co/MohammadDaleen/distilhubert-finetuned-gtzan

Source context: 9 downloads · 0 likes · Pipeline audio-classification · Library transformers · Repo MohammadDaleen/distilhubert-finetuned-gtzan