Wav2Vec2_MInDS-14_finetuned
This model is a fine-tuned version of facebook/wav2vec2-base on the minds14 dataset. It achieves the following results on the evaluation set:
- Loss: 2.6649
- Accuracy: 0.0354
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: 3e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 15
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
No log | 0.8 | 3 | 2.6414 | 0.0796 |
No log | 1.87 | 7 | 2.6435 | 0.0531 |
2.6354 | 2.93 | 11 | 2.6513 | 0.0354 |
2.6354 | 4.0 | 15 | 2.6554 | 0.0442 |
2.6354 | 4.8 | 18 | 2.6547 | 0.0442 |
2.6183 | 5.87 | 22 | 2.6594 | 0.0531 |
2.6183 | 6.93 | 26 | 2.6584 | 0.0531 |
2.6086 | 8.0 | 30 | 2.6642 | 0.0442 |
2.6086 | 8.8 | 33 | 2.6666 | 0.0442 |
2.6086 | 9.87 | 37 | 2.6631 | 0.0531 |
2.6008 | 10.93 | 41 | 2.6638 | 0.0442 |
2.6008 | 12.0 | 45 | 2.6649 | 0.0354 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.2.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.1
- Downloads last month
- 29