|
--- |
|
library_name: transformers |
|
license: apache-2.0 |
|
base_model: facebook/wav2vec2-xls-r-300m |
|
tags: |
|
- generated_from_trainer |
|
datasets: |
|
- fleurs |
|
metrics: |
|
- wer |
|
model-index: |
|
- name: wav2vec2-xls-r-Wolof-10-hours-Google-Fleurs-dataset |
|
results: |
|
- task: |
|
name: Automatic Speech Recognition |
|
type: automatic-speech-recognition |
|
dataset: |
|
name: fleurs |
|
type: fleurs |
|
config: wo_sn |
|
split: None |
|
args: wo_sn |
|
metrics: |
|
- name: Wer |
|
type: wer |
|
value: 0.442296823782073 |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/asr-africa-research-team/ASR%20Africa/runs/1lxkt8t0) |
|
# wav2vec2-xls-r-Wolof-10-hours-Google-Fleurs-dataset |
|
|
|
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the fleurs dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.2082 |
|
- Wer: 0.4423 |
|
- Cer: 0.1524 |
|
|
|
## 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_ratio: 0.1 |
|
- num_epochs: 50 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer | |
|
|:-------------:|:-------:|:----:|:---------------:|:------:|:------:| |
|
| 6.6995 | 2.6144 | 200 | 3.0428 | 1.0 | 1.0 | |
|
| 3.0391 | 5.2288 | 400 | 3.0117 | 1.0 | 1.0 | |
|
| 3.0035 | 7.8431 | 600 | 2.9794 | 1.0 | 1.0 | |
|
| 2.0946 | 10.4575 | 800 | 0.9827 | 0.7357 | 0.2560 | |
|
| 0.8407 | 13.0719 | 1000 | 0.7398 | 0.5189 | 0.1848 | |
|
| 0.5774 | 15.6863 | 1200 | 0.7214 | 0.4926 | 0.1745 | |
|
| 0.4229 | 18.3007 | 1400 | 0.6996 | 0.4852 | 0.1707 | |
|
| 0.3332 | 20.9150 | 1600 | 0.7950 | 0.4878 | 0.1708 | |
|
| 0.2488 | 23.5294 | 1800 | 0.8972 | 0.4645 | 0.1624 | |
|
| 0.2043 | 26.1438 | 2000 | 0.9122 | 0.4576 | 0.1609 | |
|
| 0.1699 | 28.7582 | 2200 | 1.0064 | 0.4777 | 0.1672 | |
|
| 0.1472 | 31.3725 | 2400 | 1.0141 | 0.4554 | 0.1581 | |
|
| 0.1251 | 33.9869 | 2600 | 1.0362 | 0.4553 | 0.1580 | |
|
| 0.1152 | 36.6013 | 2800 | 1.1312 | 0.4490 | 0.1554 | |
|
| 0.0986 | 39.2157 | 3000 | 1.1552 | 0.4499 | 0.1555 | |
|
| 0.0905 | 41.8301 | 3200 | 1.1811 | 0.4463 | 0.1547 | |
|
| 0.0879 | 44.4444 | 3400 | 1.1849 | 0.4513 | 0.1551 | |
|
| 0.0793 | 47.0588 | 3600 | 1.2074 | 0.4422 | 0.1527 | |
|
| 0.0802 | 49.6732 | 3800 | 1.2082 | 0.4423 | 0.1524 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.44.2 |
|
- Pytorch 2.1.0+cu118 |
|
- Datasets 2.17.0 |
|
- Tokenizers 0.19.1 |
|
|
