metadata
language:
- ar
license: apache-2.0
base_model: openai/whisper-large-v3
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
metrics:
- wer
model-index:
- name: Whisper Large Ar - Rami
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 11.0
type: mozilla-foundation/common_voice_11_0
config: default
split: None
args: 'config: hi, split: test'
metrics:
- name: Wer
type: wer
value: 103.42342342342343
Whisper Large Ar - Rami
This model is a fine-tuned version of openai/whisper-large-v3 on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.1317
- Wer: 103.4234
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: 500
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.2845 | 0.16 | 100 | 0.2153 | 30.0541 |
| 0.1417 | 0.32 | 200 | 0.1466 | 53.8018 |
| 0.1446 | 0.48 | 300 | 0.1388 | 64.7568 |
| 0.1326 | 0.64 | 400 | 0.1371 | 128.7568 |
| 0.13 | 0.8 | 500 | 0.1317 | 103.4234 |
Framework versions
- Transformers 4.39.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.17.1
- Tokenizers 0.15.2