metadata
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: all
split: train
args: all
metrics:
- name: Accuracy
type: accuracy
value: 0.717948717948718
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: 0.8872
- Accuracy: 0.7179
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: 4e-05
- train_batch_size: 10
- eval_batch_size: 10
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 20
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- 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 |
|---|---|---|---|---|
| 2.2117 | 1.0 | 35 | 2.1969 | 0.1923 |
| 1.9698 | 2.0 | 70 | 1.9327 | 0.3846 |
| 1.6629 | 3.0 | 105 | 1.5580 | 0.5 |
| 1.2324 | 4.0 | 140 | 1.3368 | 0.6154 |
| 1.0466 | 5.0 | 175 | 1.1638 | 0.6538 |
| 0.8969 | 6.0 | 210 | 1.0416 | 0.6923 |
| 0.7626 | 7.0 | 245 | 0.9258 | 0.7436 |
| 0.6015 | 8.0 | 280 | 1.0475 | 0.6667 |
| 0.5003 | 9.0 | 315 | 0.8890 | 0.7308 |
| 0.3956 | 10.0 | 350 | 0.8396 | 0.7564 |
| 0.3228 | 11.0 | 385 | 0.8072 | 0.6795 |
| 0.2558 | 12.0 | 420 | 0.7788 | 0.7308 |
| 0.1901 | 13.0 | 455 | 0.8432 | 0.7308 |
| 0.1251 | 14.0 | 490 | 0.8287 | 0.7051 |
| 0.1185 | 15.0 | 525 | 0.8872 | 0.7179 |
Framework versions
- Transformers 4.45.1
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.20.0