CasperMuz/whisper-medium-sna

Whisper Medium Shona - Casper Muziri

This model is a fine-tuned version of openai/whisper-medium on the Google WAXAL dataset. It achieves the following results on the evaluation set:

• Loss: 0.3686
• Wer: 34.1551

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: Use adamw_torch_fused 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: 100
• training_steps: 2500
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.4150	0.5669	500	0.4356	39.3187
0.2997	1.1338	1000	0.3889	37.1597
0.2878	1.7007	1500	0.3701	35.7750
0.2059	2.2676	2000	0.3711	34.2115
0.2055	2.8345	2500	0.3686	34.1551

Framework versions

• Transformers 5.10.1
• Pytorch 2.12.0+cu130
• Datasets 4.8.5
• Tokenizers 0.22.2