naji02010101/whisper-tiny-only_en_v11

Whisper tiny En v11 Naji

This model is a fine-tuned version of openai/whisper-tiny.en_only on the Common Voice 1 dataset. It achieves the following results on the evaluation set:

• Loss: 0.6719
• Wer Ortho: 29.5478
• Wer: 20.8967

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: 3e-07
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant_with_warmup
• lr_scheduler_warmup_steps: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Ortho	Wer
1.0568	0.8	500	1.5317	34.9567	25.8169
0.3008	1.6	1000	0.7434	30.8131	22.9407
0.3247	2.4	1500	0.7016	29.9188	21.6337
0.3102	3.2	2000	0.6860	29.8172	21.4932
0.2748	4.0	2500	0.6799	29.4697	20.8587
0.2884	4.8	3000	0.6767	29.5243	20.8397
0.245	5.6	3500	0.6714	29.5321	20.8929
0.2721	6.4	4000	0.6719	29.5478	20.8967

Framework versions

• Transformers 4.52.4
• Pytorch 2.7.1+cu126
• Datasets 2.14.6
• Tokenizers 0.21.1