You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Line 187, 188: Comments that give conext on the training set size, batch size, and total iterations to go through the dataset once
(Would be changed if the dataset in use is modified)
Line 190: Total training epochs for the CNN
Line 203: If the dataset is modified, the itrain length would change, and a new batchlength will have to be specified, thus the number '261' will have to changed accordingly
Current Scenario:
Total Training Set Size: 52,200
Total Classes: 100
Batch Size: 200 (Average images per class: 2)
Therefore, Total epochs for entire dataset: 52200/200 = 261
The above numbers would be changed when you modify the dataset.
Note:- To find the total number of images in your training set, you can run importer.py
In the end it prints "Images length: <Total training set size>"