metadata
license: apache-2.0
base_model: facebook/wav2vec2-large-xlsr-53
tags:
- generated_from_trainer
datasets:
- common_voice_13_0
metrics:
- wer
model-index:
- name: wav2vec2-xlsr-53-CV-demo-google-colab-Ezra_William_Prod13
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: 0.4344579646017699
wav2vec2-xlsr-53-CV-demo-google-colab-Ezra_William_Prod13
This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the common_voice_13_0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.4253
- Wer: 0.4345
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: 3e-05
- train_batch_size: 8
- 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 |
|---|---|---|---|---|
| 4.7051 | 1.0 | 556 | 2.9406 | 1.0 |
| 2.9144 | 2.0 | 1112 | 2.8497 | 1.0 |
| 2.7392 | 3.0 | 1668 | 1.2240 | 0.9763 |
| 1.5108 | 4.0 | 2224 | 0.6735 | 0.6265 |
| 0.9692 | 5.0 | 2780 | 0.5435 | 0.5472 |
| 0.8216 | 6.0 | 3336 | 0.5079 | 0.5058 |
| 0.7343 | 7.0 | 3892 | 0.4788 | 0.4822 |
| 0.6782 | 8.0 | 4448 | 0.4531 | 0.4569 |
| 0.6089 | 9.0 | 5004 | 0.4535 | 0.4520 |
| 0.5811 | 10.0 | 5560 | 0.4326 | 0.4381 |
| 0.5669 | 11.0 | 6116 | 0.4287 | 0.4366 |
| 0.5749 | 12.0 | 6672 | 0.4253 | 0.4345 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.2.2+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2