whisper-large-marathi

This model is a fine-tuned version of openai/whisper-large on the Common Voice 17 dataset. It achieves the following results on the evaluation set:

• Loss: 0.1845
• Wer Ortho: 32.4713
• Wer: 11.9958

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: 8
• eval_batch_size: 12
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 20
• training_steps: 500
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer Ortho	Wer
0.1771	1.0	250	0.2041	36.0371	13.7851
0.0806	2.0	500	0.1845	32.4713	11.9958

Framework versions

• Transformers 4.44.2
• Pytorch 2.5.0+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1