metadata

library_name: transformers
language:
  - ar
license: apache-2.0
base_model: openai/whisper-large-v3
tags:
  - automatic-speech-recognition
  - whisper
  - arabic
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: whisper-large-v3-ar-resturant6-Trubo
    results: []

whisper-large-v3-ar-resturant6-Trubo

This model is a fine-tuned version of openai/whisper-large-v3 on the heikal/arabic_call_splitted_6and7 dataset. It achieves the following results on the evaluation set:

Loss: 1.8007
Wer: 65.2568

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: 1000
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0004	111.1111	1000	1.8007	65.2568

Framework versions

Transformers 4.45.2
Pytorch 2.5.0+cu124
Datasets 3.0.1
Tokenizers 0.20.1