whisper-base_adi_final
This model is a fine-tuned version of openai/whisper-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.5096
- Accuracy: 0.9089
- F1: 0.9073
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: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 6
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
---|---|---|---|---|---|
0.5075 | 1.0 | 5805 | 0.8174 | 0.7556 | 0.7507 |
0.2921 | 2.0 | 11611 | 0.4716 | 0.8601 | 0.8605 |
0.1614 | 3.0 | 17417 | 0.4348 | 0.8775 | 0.8755 |
0.092 | 4.0 | 23223 | 0.4324 | 0.8915 | 0.8905 |
0.0468 | 5.0 | 29028 | 0.4695 | 0.9043 | 0.9046 |
0.0118 | 6.0 | 34830 | 0.5096 | 0.9089 | 0.9073 |
Framework versions
- Transformers 4.39.0.dev0
- Pytorch 2.0.0+cu117
- Datasets 2.17.1
- Tokenizers 0.15.2
- Downloads last month
- 4