devesh-2002/whisper-small-marathi

Whisper Small Marathi - Devesh

This model is a fine-tuned version of openai/whisper-small on the Common Voice 11.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.4955
• Wer: 42.8986

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: 16
• eval_batch_size: 8
• 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: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 5000

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0675	4.0650	1000	0.2960	46.0318
0.0052	8.1301	2000	0.3973	43.8399
0.0015	12.1951	3000	0.4573	43.4337
0.0001	16.2602	4000	0.4837	42.9308
0.0001	20.3252	5000	0.4955	42.8986

Framework versions

• Transformers 4.51.3
• Pytorch 2.1.0+cu118
• Datasets 3.5.0
• Tokenizers 0.21.1