wav2vec2-base-MELD_SER-finetuned_v2
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: 1.4979
- Accuracy: 0.4810
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: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.5485 | 1.0 | 624 | 1.5798 | 0.4233 |
| 1.4702 | 2.0 | 1248 | 1.5381 | 0.4522 |
| 1.4859 | 3.0 | 1872 | 1.4896 | 0.4648 |
| 1.1735 | 4.0 | 2497 | 1.4979 | 0.4810 |
| 1.3215 | 5.0 | 3120 | 1.5325 | 0.4531 |
Framework versions
- Transformers 4.30.0
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.13.3