metadata
license: apache-2.0
base_model: facebook/wav2vec2-xls-r-300m
tags:
- generated_from_trainer
datasets:
- common_voice_13_0
metrics:
- wer
model-index:
- name: wav2vec2-xlsr-53-CV-demo-google-colab-Ezra_William_Prod17
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: common_voice_13_0
type: common_voice_13_0
config: id
split: test
args: id
metrics:
- name: Wer
type: wer
value: 0.33453171091445427
wav2vec2-xlsr-53-CV-demo-google-colab-Ezra_William_Prod17
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the common_voice_13_0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.3245
- Wer: 0.3345
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
- num_epochs: 12
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 2.9058 | 1.0 | 278 | 2.8200 | 1.0 |
| 1.4695 | 2.0 | 556 | 0.7046 | 0.6722 |
| 0.5298 | 3.0 | 834 | 0.4448 | 0.5104 |
| 0.3601 | 4.0 | 1112 | 0.3744 | 0.4301 |
| 0.2761 | 5.0 | 1390 | 0.3398 | 0.4128 |
| 0.2092 | 6.0 | 1668 | 0.3356 | 0.3740 |
| 0.1726 | 7.0 | 1946 | 0.3276 | 0.3538 |
| 0.1461 | 8.0 | 2224 | 0.3210 | 0.3638 |
| 0.1344 | 9.0 | 2502 | 0.3173 | 0.3441 |
| 0.1173 | 10.0 | 2780 | 0.3215 | 0.3466 |
| 0.1082 | 11.0 | 3058 | 0.3272 | 0.3402 |
| 0.0981 | 12.0 | 3336 | 0.3245 | 0.3345 |
Framework versions
- Transformers 4.40.0
- Pytorch 2.2.2+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1