metadata
library_name: transformers
license: apache-2.0
base_model: facebook/hubert-base-ls960
tags:
- generated_from_trainer
datasets:
- UrbanSounds/UrbanSoundsNew
metrics:
- accuracy
model-index:
- name: hubert-base-ls960-finetuned-urbansound
results:
- task:
name: Audio Classification
type: audio-classification
dataset:
name: urbansound
type: UrbanSounds/UrbanSoundsNew
config: default
split: train
args: default
metrics:
- name: Accuracy
type: accuracy
value: 0.6086956521739131
hubert-base-ls960-finetuned-urbansound
This model is a fine-tuned version of facebook/hubert-base-ls960 on the urbansound dataset. It achieves the following results on the evaluation set:
- Loss: 1.1907
- Accuracy: 0.6087
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: 4
- eval_batch_size: 4
- seed: 42
- 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: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 2.133 | 1.0 | 50 | 2.1532 | 0.2609 |
| 2.0878 | 2.0 | 100 | 2.0094 | 0.3478 |
| 1.8873 | 3.0 | 150 | 1.8741 | 0.2609 |
| 1.6437 | 4.0 | 200 | 1.5861 | 0.4783 |
| 1.5457 | 5.0 | 250 | 1.4944 | 0.4783 |
| 1.181 | 6.0 | 300 | 1.4003 | 0.5217 |
| 1.2324 | 7.0 | 350 | 1.2538 | 0.5217 |
| 0.9965 | 8.0 | 400 | 1.1745 | 0.5217 |
| 1.26 | 9.0 | 450 | 1.1725 | 0.6087 |
| 1.0922 | 10.0 | 500 | 1.1907 | 0.6087 |
Framework versions
- Transformers 4.48.3
- Pytorch 2.5.1+cu124
- Datasets 3.3.2
- Tokenizers 0.21.0