metadata
license: apache-2.0
base_model: facebook/hubert-base-ls960
tags:
- audio-classification
- hubert
- esc50
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: hubert-esc50-finetuned
results: []
hubert-esc50-finetuned
This model is a fine-tuned version of facebook/hubert-base-ls960 on the ESC-50 dataset. It achieves the following results on the evaluation set:
- Loss: 2.3860
- Accuracy: 0.3625
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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 3.7931 | 1.0 | 200 | 3.7608 | 0.0275 |
| 3.6397 | 2.0 | 400 | 3.5743 | 0.065 |
| 3.3959 | 3.0 | 600 | 3.3550 | 0.0975 |
| 3.2348 | 4.0 | 800 | 3.1394 | 0.1375 |
| 2.9701 | 5.0 | 1000 | 2.9271 | 0.2175 |
| 2.9462 | 6.0 | 1200 | 2.7442 | 0.23 |
| 2.728 | 7.0 | 1400 | 2.6029 | 0.275 |
| 2.5537 | 8.0 | 1600 | 2.4791 | 0.3125 |
| 2.4586 | 9.0 | 1800 | 2.4028 | 0.35 |
| 2.3136 | 10.0 | 2000 | 2.3860 | 0.3625 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.0.1+cu117
- Datasets 2.16.1
- Tokenizers 0.15.1