metadata
license: apache-2.0
tags:
- automatic-speech-recognition
- openslr_SLR66
- robust-speech-event
- generated_from_trainer
model-index:
- name: ''
results: []
This model is a fine-tuned version of facebook/wav2vec2-xls-r-1b on the OPENSLR_SLR66 - NA dataset. It achieves the following results on the evaluation set:
- Loss: 0.5929
- Wer: 0.7180
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: 4
- 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_steps: 300
- num_epochs: 100.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 1.5794 | 9.61 | 500 | 0.9860 | 0.9582 |
| 1.0985 | 19.23 | 1000 | 0.5948 | 0.7534 |
| 0.9551 | 28.84 | 1500 | 0.5892 | 0.7392 |
| 0.8699 | 38.46 | 2000 | 0.5927 | 0.72 |
| 0.8372 | 48.08 | 2500 | 0.5939 | 0.7385 |
| 0.7794 | 57.69 | 3000 | 0.5801 | 0.7388 |
| 0.7565 | 67.31 | 3500 | 0.5827 | 0.7184 |
| 0.7303 | 76.92 | 4000 | 0.5800 | 0.7297 |
| 0.7128 | 86.54 | 4500 | 0.5977 | 0.7150 |
| 0.6972 | 96.15 | 5000 | 0.5901 | 0.7177 |
Framework versions
- Transformers 4.16.0.dev0
- Pytorch 1.10.1+cu102
- Datasets 1.17.1.dev0
- Tokenizers 0.11.0