---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: xls-r-asr_xh-run1
  results: []
datasets:
- lucas-meyer/asr_af
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# xls-r-asr_xh-run1

This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the asr_xh dataset.
It achieves the following results:
- Wer (Validation): 61.32%
- Wer (Test): 62.39%

### 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)    |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 14.6385       | 0.24  | 50   | 7.9420          | 1.0    |
| 5.6436        | 0.48  | 100  | 4.2744          | 1.0    |
| 3.796         | 0.72  | 150  | 3.3077          | 1.0    |
| 3.1742        | 0.96  | 200  | 3.0863          | 1.0    |
| 3.0739        | 1.2   | 250  | 3.0581          | 1.0    |
| 3.0014        | 1.44  | 300  | 2.8804          | 1.0    |
| 2.7021        | 1.67  | 350  | 2.1426          | 1.0    |
| 1.5905        | 1.91  | 400  | 1.2452          | 0.9731 |
| 1.1042        | 2.15  | 450  | 0.8109          | 0.8439 |
| 0.9033        | 2.39  | 500  | 0.7338          | 0.8260 |
| 0.7906        | 2.63  | 550  | 0.6676          | 0.8196 |
| 0.7366        | 2.87  | 600  | 0.6490          | 0.7794 |
| 0.6724        | 3.11  | 650  | 0.5505          | 0.7004 |
| 0.5529        | 3.35  | 700  | 0.5039          | 0.6686 |
| 0.5294        | 3.59  | 750  | 0.5034          | 0.6741 |
| 0.5285        | 3.83  | 800  | 0.4796          | 0.6599 |
| 0.5009        | 4.07  | 850  | 0.4366          | 0.6139 |
| 0.3814        | 4.31  | 900  | 0.4389          | 0.6142 |
| 0.3966        | 4.55  | 950  | 0.4447          | 0.6122 |
| 0.4008        | 4.78  | 1000 | 0.4444          | 0.6223 |


### Framework versions

- Transformers 4.28.0
- Pytorch 2.0.1+cu117
- Datasets 2.14.4
- Tokenizers 0.13.3