metadata
library_name: transformers
language:
- ar
- en
license: apache-2.0
base_model: openai/whisper-medium
tags:
- whisper
- egyptian-arabic
- translation
- code-switching
- generated_from_trainer
datasets:
- Assemgamal955/egyptian-english-translation
metrics:
- bleu
model-index:
- name: Whisper Medium Egyptian Arabic --> english Translation
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Assemgamal955/egyptian-english-translation
type: Assemgamal955/egyptian-english-translation
metrics:
- name: Bleu
type: bleu
value: 17.93015613103953
Whisper Medium Egyptian Arabic --> english Translation
This model is a fine-tuned version of openai/whisper-medium on the Assemgamal955/egyptian-english-translation dataset. It achieves the following results on the evaluation set:
- Loss: 1.0574
- Model Preparation Time: 0.0101
- Bleu: 17.9302
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: 7e-06
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 500
- num_epochs: 1
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Model Preparation Time | Bleu |
|---|---|---|---|---|---|
| 1.3837 | 1.0 | 1250 | 1.0422 | 0.0101 | 21.3456 |
Framework versions
- Transformers 4.53.3
- Pytorch 2.6.0+cu124
- Datasets 3.6.0
- Tokenizers 0.21.2