closes #41

views/py/python_model_template.py

@@ -74,14 +74,23 @@ def prepare_dataset(ds: tf.data.Dataset) -> tf.data.Dataset:
   ds = configure_for_performance(ds)
   return ds
 
+def filterDataset(path):
+  path = tf.strings.regex_replace(path, DATA_DIR_PREPARE, "")
+  
+  path = tf.strings.regex_replace(path, ".jpg", "")
+  path = tf.strings.regex_replace(path, ".jpeg", "")
+  
+  return tf.reshape(table.lookup(tf.strings.as_string([path])), []) != -1
+
 seed = random.randint(0, 100000000)
 
 batch_size = 100
 
 # Read all the files from the direcotry
 list_ds = tf.data.Dataset.list_files(str(f'{DATA_DIR}/*'), shuffle=False)
+list_ds = list_ds.filter(filterDataset)
 
-image_count = len(list_ds)
+image_count = len(list(list_ds.as_numpy_iterator()))
 
 list_ds = list_ds.shuffle(image_count, seed=seed)