wav2vec2-large-xlsr-moroccan-darija-v3
This model is a fine-tuned version of hassanaitnacer/wav2vec2-large-xlsr-moroccan-darija-v2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0553
- Wer: 0.1312
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: 0.0001
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.103 | 0.5359 | 500 | 0.0638 | 0.1421 |
| 0.1234 | 1.0718 | 1000 | 0.0630 | 0.1394 |
| 0.0967 | 1.6077 | 1500 | 0.0553 | 0.1312 |
Framework versions
- Transformers 4.43.0.dev0
- Pytorch 2.3.0+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1