Arbi-Houssem/TunLangModel_test1.11

Whisper Tunisien

This model is a fine-tuned version of openai/whisper-small on the Tunisian_dataset_STT-TTS15s_filtred1.0 dataset. It achieves the following results on the evaluation set:

• Loss: 6.8348
• Wer: 102.7009

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: 0.0005
• train_batch_size: 8
• 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: 3000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.1414	3.8760	500	4.1822	181.0263
0.4667	7.7519	1000	5.0049	108.1702
0.1821	11.6279	1500	5.5927	102.7684
0.068	15.5039	2000	6.2194	106.2120
0.011	19.3798	2500	6.3815	103.0385
0.0003	23.2558	3000	6.8348	102.7009

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1