wav2vec2-base-finetuned-organ

This model is a fine-tuned version of facebook/wav2vec2-base on the GTZAN dataset. It achieves the following results on the evaluation set:

  • Loss: 1.0117
  • Accuracy: 0.8182

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: 50
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.0965 1.0 6 1.0843 0.4545
1.0989 2.0 12 1.0883 0.3636
1.0931 3.0 18 1.0914 0.5455
1.0702 4.0 24 1.0578 0.4545
0.9822 5.0 30 0.9994 0.7273
0.9139 6.0 36 0.9735 0.5455
0.8008 7.0 42 0.7004 0.9091
0.6798 8.0 48 0.7404 0.8182
0.5969 9.0 54 0.7192 0.7273
0.4976 10.0 60 0.4668 0.9091
0.436 11.0 66 0.7406 0.7273
0.5859 12.0 72 0.6139 0.7273
0.3788 13.0 78 0.6551 0.7273
0.3176 14.0 84 0.4746 0.9091
0.2892 15.0 90 0.8285 0.7273
0.2452 16.0 96 0.8523 0.7273
0.1464 17.0 102 0.9791 0.7273
0.4589 18.0 108 1.2469 0.6364
0.1641 19.0 114 1.1607 0.6364
0.1765 20.0 120 0.7318 0.8182
0.1553 21.0 126 1.1178 0.6364
0.2048 22.0 132 1.2835 0.6364
0.2477 23.0 138 0.7558 0.8182
0.2042 24.0 144 0.8053 0.8182
0.2242 25.0 150 1.1131 0.7273
0.2063 26.0 156 1.1455 0.7273
0.1148 27.0 162 1.1386 0.7273
0.0948 28.0 168 1.0196 0.7273
0.2296 29.0 174 1.2216 0.7273
0.1771 30.0 180 1.2645 0.7273
0.0749 31.0 186 1.3599 0.6364
0.0973 32.0 192 1.2880 0.7273
0.0231 33.0 198 0.9015 0.8182
0.1185 34.0 204 0.9180 0.8182
0.1645 35.0 210 1.3635 0.7273
0.0163 36.0 216 1.3961 0.7273
0.0743 37.0 222 1.3699 0.7273
0.0211 38.0 228 0.8085 0.8182
0.0713 39.0 234 0.8418 0.8182
0.0122 40.0 240 0.7659 0.8182
0.0116 41.0 246 0.9891 0.8182
0.0117 42.0 252 1.4963 0.7273
0.0738 43.0 258 1.4932 0.7273
0.0718 44.0 264 1.4665 0.7273
0.1334 45.0 270 0.9666 0.8182
0.0662 46.0 276 0.9798 0.8182
0.0973 47.0 282 0.9954 0.8182
0.0105 48.0 288 1.0073 0.8182
0.0092 49.0 294 1.0107 0.8182
0.0089 50.0 300 1.0117 0.8182

Framework versions

  • Transformers 4.42.3
  • Pytorch 2.3.1+cu121
  • Datasets 2.19.1
  • Tokenizers 0.19.1
Downloads last month
14
Safetensors
Model size
94.6M params
Tensor type
F32
·
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 Kiri1993/wav2vec2-base-finetuned-organ

Finetuned
(680)
this model

Dataset used to train Kiri1993/wav2vec2-base-finetuned-organ

Evaluation results