minzwon/sota-music-tagging-models

Issues

• would love to run this

  #18 opened 2 years ago by ericlam1114
  3

• Data splitting of MTAT dataset

  #25 opened a year ago by suncerock
  1

• Clarification on dataset format(s)

  #26 opened a year ago by JLenzy
  3

• Where can we find the pretrained models ?

  #24 opened a year ago by nikifori
  0

• Dataset paths

  #21 opened 2 years ago by mrticker
  0

• Old module versions in requirements.txt

  #20 opened 2 years ago by mrticker
  0

• preprocessing\mtat_read.py has mix of tabs and spaces for indentation

  #19 opened 2 years ago by mrticker
  0

• the difference between musicnn and music600

  #11 opened 3 years ago by Usernamezhx
  3

• Only one class present in y_true. ROC AUC score is not defined in that case.

  #15 opened 3 years ago by ppjjee
  5

• Bug running fcn with mtat

  #14 opened 3 years ago by Sette
  11

• input length of short-chunk CNN

  #13 opened 3 years ago by da-hon5
  1

• the small data size training.

  #12 opened 3 years ago by Usernamezhx
  3

• Is there [Short-chunk CNN + Res] pretrained weight about MTAT dataset?

  #10 opened 3 years ago by rlgnswk
  3

• How to get the audio files of MSD?

  #7 opened 3 years ago by millerchx
  2

• My roc_auc is higher than reported

  #5 opened 3 years ago by pengbo-learn
  8

• I have a question

  #4 opened 4 years ago by hjlee9182
  2

• Problems training on jamendo - RuntimeError: invalid argument 0: Sizes of tensors must match except in dimension 0

  #3 opened 4 years ago by mathigatti
  2

• Do you have the best_model.pth of HarmonicCNN model in MTG data

  #2 opened 4 years ago by sicilyliulc
  2

• Error in loading the model during the training process

  #1 opened 4 years ago by sainathadapa
  3