metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- common_voice
model-index:
- name: ''
results: []
This model is a fine-tuned version of facebook/wav2vec2-xls-r-1b on the common_voice dataset. It achieves the following results on the evaluation set:
- Loss: 1.5476
- Wer: 0.9696
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: 6.5e-05
- train_batch_size: 32
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- 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: 500
- num_epochs: 10.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 2.6638 | 0.84 | 500 | 2.3852 | 0.9974 |
| 2.6578 | 1.67 | 1000 | 2.2796 | 0.9971 |
| 2.6016 | 2.51 | 1500 | 2.0046 | 0.9961 |
| 2.5752 | 3.35 | 2000 | 1.9606 | 0.9961 |
| 2.539 | 4.19 | 2500 | 1.8836 | 0.9940 |
| 2.5214 | 5.03 | 3000 | 1.8593 | 0.9933 |
| 2.4684 | 5.86 | 3500 | 1.7816 | 0.9885 |
| 2.4134 | 6.7 | 4000 | 1.7168 | 0.9808 |
| 2.3732 | 7.54 | 4500 | 1.6406 | 0.9764 |
| 2.3371 | 8.37 | 5000 | 1.6087 | 0.9739 |
| 2.2824 | 9.21 | 5500 | 1.5476 | 0.9696 |
Framework versions
- Transformers 4.17.0.dev0
- Pytorch 1.10.2+cu102
- Datasets 1.18.2.dev0
- Tokenizers 0.11.0