Edit model card

whisper-tiny-finetuned-gtzan

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

  • Loss: 0.4916
  • Accuracy: 0.91

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: 5e-05
  • train_batch_size: 16
  • eval_batch_size: 16
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 30

Training results

Training Loss Epoch Step Validation Loss Accuracy
2.1202 1.0 57 2.0148 0.49
1.4611 2.0 114 1.3965 0.62
0.9725 3.0 171 0.8726 0.82
0.4971 4.0 228 0.7578 0.76
0.2255 5.0 285 0.7502 0.74
0.2803 6.0 342 0.5457 0.84
0.2234 7.0 399 0.7014 0.8
0.0845 8.0 456 0.4250 0.89
0.0395 9.0 513 0.5069 0.9
0.0438 10.0 570 0.4916 0.91
0.0442 11.0 627 0.7312 0.86
0.002 12.0 684 0.4753 0.9
0.0769 13.0 741 0.8024 0.86
0.0015 14.0 798 0.6354 0.9
0.001 15.0 855 0.5665 0.91
0.0008 16.0 912 0.5537 0.9
0.0009 17.0 969 0.6251 0.88
0.0007 18.0 1026 0.6641 0.9
0.0006 19.0 1083 0.5746 0.9
0.0006 20.0 1140 0.5893 0.9
0.0006 21.0 1197 0.5636 0.91
0.0005 22.0 1254 0.5785 0.91
0.0118 23.0 1311 0.5674 0.91
0.0005 24.0 1368 0.5915 0.91
0.0585 25.0 1425 0.5690 0.91
0.0004 26.0 1482 0.6043 0.9
0.008 27.0 1539 0.5911 0.91
0.0208 28.0 1596 0.5973 0.91
0.0004 29.0 1653 0.6009 0.91
0.0004 30.0 1710 0.6035 0.91

Framework versions

  • Transformers 4.29.2
  • Pytorch 2.0.0
  • Datasets 2.13.1
  • Tokenizers 0.13.3
Downloads last month
10
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for ditwoo/whisper-tiny-finetuned-gtzan

Finetuned
(1210)
this model

Dataset used to train ditwoo/whisper-tiny-finetuned-gtzan