wav2vec2-base-finetuned-ic-slurp-wt_init-frz-v1
This model is a fine-tuned version of facebook/wav2vec2-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 4.0306
- Accuracy: 0.0502
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: 24
- eval_batch_size: 24
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 96
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 50
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
3.9426 | 1.0 | 527 | 4.1870 | 0.0420 |
3.7966 | 2.0 | 1055 | 4.0306 | 0.0502 |
3.7149 | 3.0 | 1582 | 3.9582 | 0.0434 |
3.6478 | 4.0 | 2110 | 3.9343 | 0.0427 |
3.5037 | 5.0 | 2637 | 3.9302 | 0.0413 |
3.4649 | 6.0 | 3165 | 3.9289 | 0.0474 |
3.2427 | 7.0 | 3692 | 3.9650 | 0.0473 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.2.1+cu118
- Datasets 2.17.0
- Tokenizers 0.15.0
- Downloads last month
- 1