parambharat/whisper-base-kn

Whisper Base Kn - Bharat Ramanathan

This model is a fine-tuned version of openai/whisper-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1974
• Wer: 30.8790

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: 96
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.572	0.1	500	0.3198	50.3005
0.3153	0.2	1000	0.2464	37.2652
0.2533	0.3	1500	0.2298	36.5515
0.2212	1.04	2000	0.2157	34.5229
0.2013	1.14	2500	0.2090	32.6071
0.1881	1.24	3000	0.2043	32.7198
0.1784	1.34	3500	0.2014	30.8039
0.1715	2.08	4000	0.2014	31.5928
0.166	2.18	4500	0.1991	31.2547
0.1616	2.28	5000	0.1974	30.8790

Framework versions

• Transformers 4.26.0.dev0
• Pytorch 1.13.0+cu117
• Datasets 2.7.1.dev0
• Tokenizers 0.13.2