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-300m on the common_voice dataset. It achieves the following results on the evaluation set:
- Loss: 0.7205
- Wer: 0.6856
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: 7.5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 4
- total_train_batch_size: 256
- total_eval_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2000
- num_epochs: 50.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 3.3873 | 11.62 | 500 | 3.3324 | 1.0 |
| 2.9207 | 23.25 | 1000 | 2.8924 | 0.9987 |
| 2.1463 | 34.88 | 1500 | 1.3063 | 0.9139 |
| 1.607 | 46.51 | 2000 | 0.7205 | 0.6856 |
Framework versions
- Transformers 4.17.0.dev0
- Pytorch 1.9.0+cu111
- Datasets 1.18.4.dev0
- Tokenizers 0.10.3