metadata
language:
- ur
license: apache-2.0
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
metrics:
- wer
model-index:
- name: Whisper Tiny Urdu - Bilal
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 11.0
type: mozilla-foundation/common_voice_11_0
config: ur
split: test
args: 'config: ur, split: test'
metrics:
- name: Wer
type: wer
value: 75.2080188861798
Whisper Tiny Urdu - Bilal
This model is a fine-tuned version of openai/whisper-tiny on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.7874
- Wer: 75.2080
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: 2
- 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
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.8408 | 0.27 | 1000 | 0.9658 | 168.8842 |
| 0.8024 | 0.54 | 2000 | 0.8615 | 79.9063 |
| 0.735 | 0.81 | 3000 | 0.8074 | 84.7556 |
| 0.692 | 1.08 | 4000 | 0.7874 | 75.2080 |
Framework versions
- Transformers 4.28.0.dev0
- Pytorch 2.0.0+cu117
- Datasets 2.11.0
- Tokenizers 0.13.3