metadata
tags:
- generated_from_trainer
base_model: batoula187/wav2vec2-large-xls-r-300m-arabic-colab
datasets:
- common_voice_12_0
metrics:
- wer
model-index:
- name: wav2vec2-large-xls-r-300m-arabic-colab
results:
- task:
type: automatic-speech-recognition
name: Automatic Speech Recognition
dataset:
name: common_voice_12_0
type: common_voice_12_0
config: ar
split: test[:10%]
args: ar
metrics:
- type: wer
value: 0.7661710754972002
name: Wer
wav2vec2-large-xls-r-300m-arabic-colab
This model is a fine-tuned version of batoula187/wav2vec2-large-xls-r-300m-arabic-colab on the common_voice_12_0 dataset. It achieves the following results on the evaluation set:
- Loss: 2.0728
- Wer: 0.7662
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: 0.0003
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.1061 | 2.2599 | 200 | 1.8297 | 0.8034 |
| 0.1496 | 4.5198 | 400 | 1.6173 | 0.7955 |
| 0.2105 | 6.7797 | 600 | 1.6220 | 0.8040 |
| 0.1798 | 9.0395 | 800 | 2.2087 | 0.8405 |
| 0.1389 | 11.2994 | 1000 | 1.7900 | 0.7868 |
| 0.1143 | 13.5593 | 1200 | 1.7566 | 0.7886 |
| 0.103 | 15.8192 | 1400 | 1.8148 | 0.7689 |
| 0.0904 | 18.0791 | 1600 | 1.8059 | 0.7627 |
| 0.0766 | 20.3390 | 1800 | 2.1398 | 0.7907 |
| 0.0682 | 22.5989 | 2000 | 2.0384 | 0.7779 |
| 0.0583 | 24.8588 | 2200 | 2.0727 | 0.7658 |
| 0.0575 | 27.1186 | 2400 | 2.1649 | 0.7758 |
| 0.0582 | 29.3785 | 2600 | 2.0728 | 0.7662 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.2.1+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1