We need more training epochs with more augmentations

OpenImaging · Apr 14, 2022 · 31992c1 · 31992c1
1 parent 6dfc8ec
commit 31992c1
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/miqa/learning/nn_training.py b/miqa/learning/nn_training.py
@@ -644,7 +644,7 @@ def process_folds(folds_prefix, validation_fold, evaluate_only, fold_count):
 
     # establish minimum number of optimization steps and epochs
     val_count = max(1, int(600 / df.shape[0]))
-    epoch_count = max(25, int(30000 / df.shape[0]))
+    epoch_count = max(40, int(30000 / df.shape[0]))
     epoch_count = math.ceil(epoch_count / val_count) * val_count
 
     count_train = df.shape[0] - vf.shape[0]