Whisper Tiny Naija - whisper-tiny-naija-finetuned

This model is a fine-tuned version of openai/whisper-tiny on the Naija-ASR-Corpus dataset. It achieves the following results on the evaluation set:

• Loss: 0.7690
• Wer: 75.8154

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 OptimizerNames.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: 50
• training_steps: 500
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.2457	0.3021	100	1.1449	76.9210
0.9626	0.6042	200	0.9074	62.0509
0.8398	0.9063	300	0.8167	74.2261
0.6932	1.2085	400	0.7802	73.1067
0.6619	1.5106	500	0.7690	75.8154

Framework versions

• Transformers 4.57.1
• Pytorch 2.8.0+cu126
• Datasets 2.21.0
• Tokenizers 0.22.1