Zakryah/whisper-tiny-hu

pici - Zakryah

This model is a fine-tuned version of openai/whisper-tiny on the Common Voice 13.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.5618
• Wer: 49.5177

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: 2
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 16
• 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: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.7211	0.6895	1000	0.7369	59.2806
0.5253	1.3786	2000	0.6201	53.7320
0.4235	2.0676	3000	0.5741	50.7056
0.4075	2.7571	4000	0.5618	49.5177

Framework versions

• Transformers 4.48.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0