Update README.md

README.md

@@ -14,127 +14,127 @@ configs:
     path: eval.json
 - config_name: unbalanced_train_part00
   data_files: unbalanced_train_part00.json
-  dataset_size: 46940
+  # dataset_size: 46940
 - config_name: unbalanced_train_part01
   data_files: unbalanced_train_part01.json
-  dataset_size: 47052
+  # dataset_size: 47052
 - config_name: unbalanced_train_part02
   data_files: unbalanced_train_part02.json
-  dataset_size: 46923
+  # dataset_size: 46923
 - config_name: unbalanced_train_part03
   data_files: unbalanced_train_part03.json
-  dataset_size: 46952
+  # dataset_size: 46952
 - config_name: unbalanced_train_part04
   data_files: unbalanced_train_part04.json
-  dataset_size: 46916
+  # dataset_size: 46916
 - config_name: unbalanced_train_part05
   data_files: unbalanced_train_part05.json
-  dataset_size: 47011
+  # dataset_size: 47011
 - config_name: unbalanced_train_part06
   data_files: unbalanced_train_part06.json
-  dataset_size: 46964
+  # dataset_size: 46964
 - config_name: unbalanced_train_part07
   data_files: unbalanced_train_part07.json
-  dataset_size: 46915
+  # dataset_size: 46915
 - config_name: unbalanced_train_part08
   data_files: unbalanced_train_part08.json
-  dataset_size: 46927
+  # dataset_size: 46927
 - config_name: unbalanced_train_part09
   data_files: unbalanced_train_part09.json
-  dataset_size: 46839
+  # dataset_size: 46839
 - config_name: unbalanced_train_part10
   data_files: unbalanced_train_part10.json
-  dataset_size: 46862
+  # dataset_size: 46862
 - config_name: unbalanced_train_part11
   data_files: unbalanced_train_part11.json
-  dataset_size: 46836
+  # dataset_size: 46836
 - config_name: unbalanced_train_part12
   data_files: unbalanced_train_part12.json
-  dataset_size: 46865
+  # dataset_size: 46865
 - config_name: unbalanced_train_part13
   data_files: unbalanced_train_part13.json
-  dataset_size: 46800
+  # dataset_size: 46800
 - config_name: unbalanced_train_part14
   data_files: unbalanced_train_part14.json
-  dataset_size: 46837
+  # dataset_size: 46837
 - config_name: unbalanced_train_part15
   data_files: unbalanced_train_part15.json
-  dataset_size: 46824
+  # dataset_size: 46824
 - config_name: unbalanced_train_part16
   data_files: unbalanced_train_part16.json
-  dataset_size: 46813
+  # dataset_size: 46813
 - config_name: unbalanced_train_part17
   data_files: unbalanced_train_part17.json
-  dataset_size: 46771
+  # dataset_size: 46771
 - config_name: unbalanced_train_part18
   data_files: unbalanced_train_part18.json
-  dataset_size: 46875
+  # dataset_size: 46875
 - config_name: unbalanced_train_part19
   data_files: unbalanced_train_part19.json
-  dataset_size: 46885
+  # dataset_size: 46885
 - config_name: unbalanced_train_part20
   data_files: unbalanced_train_part20.json
-  dataset_size: 46884
+  # dataset_size: 46884
 - config_name: unbalanced_train_part21
   data_files: unbalanced_train_part21.json
-  dataset_size: 46736
+  # dataset_size: 46736
 - config_name: unbalanced_train_part22
   data_files: unbalanced_train_part22.json
-  dataset_size: 46832
+  # dataset_size: 46832
 - config_name: unbalanced_train_part23
   data_files: unbalanced_train_part23.json
-  dataset_size: 46823
+  # dataset_size: 46823
 - config_name: unbalanced_train_part24
   data_files: unbalanced_train_part24.json
-  dataset_size: 46795
+  # dataset_size: 46795
 - config_name: unbalanced_train_part25
   data_files: unbalanced_train_part25.json
-  dataset_size: 46740
+  # dataset_size: 46740
 - config_name: unbalanced_train_part26
   data_files: unbalanced_train_part26.json
-  dataset_size: 46765
+  # dataset_size: 46765
 - config_name: unbalanced_train_part27
   data_files: unbalanced_train_part27.json
-  dataset_size: 46708
+  # dataset_size: 46708
 - config_name: unbalanced_train_part28
   data_files: unbalanced_train_part28.json
-  dataset_size: 46736
+  # dataset_size: 46736
 - config_name: unbalanced_train_part29
   data_files: unbalanced_train_part29.json
-  dataset_size: 46819
+  # dataset_size: 46819
 - config_name: unbalanced_train_part30
   data_files: unbalanced_train_part30.json
-  dataset_size: 46694
+  # dataset_size: 46694
 - config_name: unbalanced_train_part31
   data_files: unbalanced_train_part31.json
-  dataset_size: 46735
+  # dataset_size: 46735
 - config_name: unbalanced_train_part32
   data_files: unbalanced_train_part32.json
-  dataset_size: 46731
+  # dataset_size: 46731
 - config_name: unbalanced_train_part33
   data_files: unbalanced_train_part33.json
-  dataset_size: 46627
+  # dataset_size: 46627
 - config_name: unbalanced_train_part34
   data_files: unbalanced_train_part34.json
-  dataset_size: 46740
+  # dataset_size: 46740
 - config_name: unbalanced_train_part35
   data_files: unbalanced_train_part35.json
-  dataset_size: 46866
+  # dataset_size: 46866
 - config_name: unbalanced_train_part36
   data_files: unbalanced_train_part36.json
-  dataset_size: 46758
+  # dataset_size: 46758
 - config_name: unbalanced_train_part37
   data_files: unbalanced_train_part37.json
-  dataset_size: 46751
+  # dataset_size: 46751
 - config_name: unbalanced_train_part38
   data_files: unbalanced_train_part38.json
-  dataset_size: 46750
+  # dataset_size: 46750
 - config_name: unbalanced_train_part39
   data_files: unbalanced_train_part39.json
-  dataset_size: 46700
+  # dataset_size: 46700
 - config_name: unbalanced_train_part40
   data_files: unbalanced_train_part40.json
-  dataset_size: 39137
+  # dataset_size: 39137
 task_categories:
 - audio-classification
 tags:
@@ -150,7 +150,10 @@ Some clips are missing on YouTube, so the number of files downloaded is differen
 This repository contains 20550 / 22160 of the balanced train set, 1913637 / 2041789 of the unbalanced train set (separated into 41 parts), and 18887 / 20371 of the evaluation set.
 The pre-process script can be found at qiuqiangkong's [github](https://github.com/qiuqiangkong/audioset_tagging_cnn).
 
+To improve training efficiency, we add a slightly more balanced subset AudioSet500K.
+
 ## References
 
 1. Gemmeke, Jort F., et al., Audio set: An ontology and human-labeled dataset for audio events, 2017
-2. Kong, Qiuqiang, et al., Panns: Large-scale pretrained audio neural networks for audio pattern recognition, 2020
+2. Kong, Qiuqiang, et al., Panns: Large-scale pretrained audio neural networks for audio pattern recognition, 2020
+3. Nagrani, Arsha, et al. Attention bottlenecks for multimodal fusion, 2021