kostissz/whisper-tiny-el

Baseline openai/whisper-tiny on Greek. Evaluated on 1701 audio samples

This model is a fine-tuned version of openai/whisper-tiny on the mozilla-foundation/common_voice_17_0- Greek dataset. It achieves the following results on the evaluation set:

• Loss: 0.8397
• Wer: 56.1834

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: 64
• eval_batch_size: 8
• seed: 42
• 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: linear
• lr_scheduler_warmup_steps: 50
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.02	17.5439	1000	0.6888	56.2946
0.0024	35.0877	2000	0.8397	56.1834
0.001	52.6316	3000	0.9095	56.8041
0.0006	70.1754	4000	0.9501	57.0079
0.0005	87.7193	5000	0.9673	57.2487

Framework versions

• Transformers 4.48.0
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0