trainer

This model is a fine-tuned version of MIT/ast-finetuned-audioset-10-10-0.4593 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3684
  • Accuracy: 0.9275

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: 0.0001
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 64
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 50
  • training_steps: 1000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy
4.2624 2.0 50 0.3928 0.88
0.9069 4.0 100 0.3259 0.9025
0.9069 6.0 150 0.2775 0.93
0.0567 8.0 200 0.3220 0.9075
0.0567 10.0 250 0.3196 0.9075
0.0109 12.0 300 0.3644 0.9175
0.0109 14.0 350 0.3501 0.93
0.0138 16.0 400 0.3569 0.9275
0.0138 18.0 450 0.3700 0.9225
0.0006 20.0 500 0.3662 0.925
0.0006 22.0 550 0.3669 0.925
0.0002 24.0 600 0.3673 0.925
0.0002 26.0 650 0.3677 0.925
0.0002 28.0 700 0.3679 0.9275
0.0002 30.0 750 0.3680 0.9275
0.0002 32.0 800 0.3681 0.9275
0.0002 34.0 850 0.3684 0.9275
0.0002 36.0 900 0.3683 0.9275
0.0002 38.0 950 0.3684 0.9275
0.0002 40.0 1000 0.3684 0.9275

Framework versions

  • Transformers 4.27.1
  • Pytorch 2.0.1+cu117
  • Datasets 2.13.1
  • Tokenizers 0.13.3
Downloads last month
8
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.