metadata
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: wav2vec2-base-EMOPIA
results: []
wav2vec2-base-EMOPIA
This model is a fine-tuned version of facebook/wav2vec2-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.9346
- Accuracy: 0.7000
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: 1e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 3
- total_train_batch_size: 12
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 300
- num_epochs: 15
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.3867 | 0.99 | 58 | 1.3518 | 0.3286 |
| 1.2842 | 1.99 | 116 | 1.1872 | 0.4571 |
| 1.0725 | 2.99 | 174 | 1.0476 | 0.5571 |
| 0.9343 | 3.99 | 232 | 0.9677 | 0.5714 |
| 0.8053 | 4.99 | 290 | 0.9525 | 0.6143 |
| 0.7895 | 5.99 | 348 | 0.7735 | 0.6857 |
| 0.6867 | 6.99 | 406 | 0.8556 | 0.6429 |
| 0.6218 | 7.99 | 464 | 0.8454 | 0.6714 |
| 0.558 | 8.99 | 522 | 0.8405 | 0.6571 |
| 0.5033 | 9.99 | 580 | 1.0190 | 0.6286 |
| 0.4403 | 10.99 | 638 | 0.8249 | 0.7000 |
| 0.3995 | 11.99 | 696 | 0.8997 | 0.7143 |
| 0.3534 | 12.99 | 754 | 0.9177 | 0.7000 |
| 0.3023 | 13.99 | 812 | 0.9544 | 0.6571 |
| 0.2752 | 14.99 | 870 | 0.9346 | 0.7000 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.10.1+cu102
- Datasets 2.4.0
- Tokenizers 0.12.1