search

RetroCirce / HTS-Audio-Transformer

The official code repo of "HTS-AT: A Hierarchical Token-Semantic Audio Transformer for Sound Classification and Detection"
https://arxiv.org/abs/2202.00874
MIT License
341 stars 62 forks source link

How can I test model？ #26

Closed yyssxxx closed 1 year ago

yyssxxx commented 1 year ago

Hello , in "htsat_esc_training.ipnb", how can i find test_data? Or i should design the test_data myself? I'm not familiar with pytorch-lightning.

RetroCirce commented 1 year ago

Hi, please follow the readme or the comments in the ipynb file, there is a link for downing the dataset.

yyssxxx commented 1 year ago

Hi, please follow the readme or the comments in the ipynb file, there is a link for downing the dataset.

Thanks for answering, but i mean how can i use the data from the test_data the get the accuracy in the test data , i think "./workspace/esc-50/raw/ESC-50-master/audio/1-7456-A-13.wav" maybe is not in the test_data? I dont know "pytorch_lightning" too much, i want to find the test data.