wav2vec2-ft-lre5-adm-ga2b16
This model is a fine-tuned version of jonatasgrosman/wav2vec2-large-english on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.7165
- Wer: 0.8799
- Cer: 0.5642
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: 1e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 10000
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
---|---|---|---|---|---|
3.3549 | 0.4165 | 1000 | 3.3633 | 1.0000 | 1.0000 |
3.2645 | 0.8330 | 2000 | 3.3346 | 1.0000 | 1.0000 |
3.1935 | 1.2495 | 3000 | 3.2858 | 1.0000 | 1.0000 |
3.2244 | 1.6660 | 4000 | 3.3026 | 1.0000 | 1.0000 |
3.0293 | 2.0825 | 5000 | 3.1127 | 0.9926 | 0.8498 |
2.8431 | 2.4990 | 6000 | 2.8820 | 0.9311 | 0.6048 |
2.7486 | 2.9155 | 7000 | 2.7647 | 0.8952 | 0.5799 |
2.7486 | 3.3319 | 8000 | 2.7236 | 0.8821 | 0.5699 |
2.6133 | 3.7484 | 9000 | 2.7327 | 0.8781 | 0.5629 |
2.6314 | 4.1649 | 10000 | 2.7165 | 0.8799 | 0.5642 |
Framework versions
- Transformers 4.52.3
- Pytorch 2.5.1
- Datasets 3.6.0
- Tokenizers 0.21.1
- Downloads last month
- 4
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
๐
Ask for provider support
Model tree for HouraMort/wav2vec2-ft-lre5-adm-ga2b16
Base model
jonatasgrosman/wav2vec2-large-english