afaaaak/whisper-tiny-ps

Whisper small Ps - ZFA

This model is a fine-tuned version of openai/whisper-small on the Common Voice 20.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.8066

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: 5e-05
• train_batch_size: 1
• eval_batch_size: 1
• 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
• num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss
1.7874	0.1856	500	1.3995
1.3066	0.3712	1000	1.2622
1.1437	0.5568	1500	1.1273
1.0676	0.7424	2000	1.0547
1.0014	0.9280	2500	0.9770
0.7683	1.1136	3000	0.9779
0.6386	1.2992	3500	0.9486
0.6103	1.4848	4000	0.9071
0.599	1.6704	4500	0.8748
0.5665	1.8560	5000	0.8525
0.5032	2.0416	5500	0.8532
0.2884	2.2272	6000	0.8503
0.269	2.4128	6500	0.8316
0.2784	2.5984	7000	0.8137
0.236	2.7840	7500	0.8227
0.2543	2.9696	8000	0.8066

Framework versions

• Transformers 4.56.2
• Pytorch 2.7.0+cu126
• Datasets 4.1.1
• Tokenizers 0.22.0