metadata
tags:
- generated_from_trainer
datasets:
- google/fleurs
metrics:
- wer
model-index:
- name: xls-r-fleurs_zu-run3
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: audiofolder
type: audiofolder
config: default
split: validation
args: default
metrics:
- name: Wer
type: wer
value: 0.5777717243257968
xls-r-fleurs_zu-run3
This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the FLEURS (zu) dataset. It achieves the following results:
- Wer (Validation): 57.19%
- Wer (Test): 57.27%
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0003
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 3
- total_train_batch_size: 12
- 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 (Train) |
|---|---|---|---|---|
| 0.1019 | 0.28 | 50 | 0.5804 | 0.5710 |
| 0.1136 | 0.57 | 100 | 0.5462 | 0.5745 |
| 0.1122 | 0.85 | 150 | 0.5401 | 0.5650 |
| 0.097 | 1.14 | 200 | 0.5680 | 0.5598 |
| 0.0938 | 1.42 | 250 | 0.5763 | 0.5603 |
| 0.1004 | 1.7 | 300 | 0.5803 | 0.5778 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu117
- Datasets 2.14.4
- Tokenizers 0.13.3