whisper-medium-ur

This model is a fine-tuned version of GogetaBlueMUI/whisper-medium-ur-jalandhary on the common_voice_11_0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.5375
• Wer: 0.2744
• Cer: 0.1237

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-05
• train_batch_size: 8
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Use 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: 100
• training_steps: 1800
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
0.2831	0.6459	300	0.5175	0.4527	0.2883
0.1326	1.2906	600	0.5248	0.4708	0.2906
0.1199	1.9365	900	0.4979	0.3738	0.2084
0.0412	2.5813	1200	0.5417	0.3038	0.1399
0.0119	3.2260	1500	0.5542	0.2841	0.1243
0.0093	3.8719	1800	0.5375	0.2744	0.1237

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.3.1
• Tokenizers 0.21.0