68 lines
1.6 KiB
Markdown
68 lines
1.6 KiB
Markdown
---
|
|
language:
|
|
- ar
|
|
license: apache-2.0
|
|
base_model: openai/whisper-small
|
|
tags:
|
|
- hf-asr-leaderboard
|
|
- generated_from_trainer
|
|
metrics:
|
|
- wer
|
|
model-index:
|
|
- name: Whisper Small Arabic - Vashesh Jogani
|
|
results: []
|
|
---
|
|
|
|
<!-- 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. -->
|
|
|
|
# Whisper Small Arabic - Vashesh Jogani
|
|
|
|
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Arabic Voice 11.0 dataset.
|
|
It achieves the following results on the evaluation set:
|
|
- Loss: 0.6893
|
|
- Wer: 1.0
|
|
|
|
## 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: 1e-05
|
|
- train_batch_size: 16
|
|
- eval_batch_size: 8
|
|
- seed: 42
|
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
|
- lr_scheduler_type: linear
|
|
- lr_scheduler_warmup_steps: 500
|
|
- training_steps: 4000
|
|
|
|
### Training results
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
|
|:-------------:|:-----:|:----:|:---------------:|:---:|
|
|
| 0.8346 | 0.05 | 1000 | 0.8098 | 1.0 |
|
|
| 0.7309 | 0.1 | 2000 | 0.7476 | 1.0 |
|
|
| 0.686 | 0.15 | 3000 | 0.7093 | 1.0 |
|
|
| 0.7186 | 0.2 | 4000 | 0.6893 | 1.0 |
|
|
|
|
|
|
### Framework versions
|
|
|
|
- Transformers 4.34.1
|
|
- Pytorch 2.1.0
|
|
- Datasets 2.14.5
|
|
- Tokenizers 0.14.1
|