File cleanups, removed unused lines, ordering packages based on were …

…it came from, e.g. local/third-party/standard libraries

src/analyze.py

@@ -1,17 +1,22 @@
-from __future__ import division
-from __future__ import print_function
+# Imports
 
+# > Standard Library
 import os
 import sys
 import math
 import pickle
 import copy
-import numpy as np
-import cv2
-import matplotlib.pyplot as plt
+from __future__ import division
+from __future__ import print_function
+
+# > Local dependencies
 from model import Model
 from sample_processor import preprocess
 
+# > Third party libraries
+import numpy as np
+import cv2
+import matplotlib.pyplot as plt
 
 # constants like filepaths
 class Constants:

src/arg_parser.py

@@ -1,5 +1,11 @@
+# Imports
+
+# > Standard Library
 import argparse
 
+# > Local dependencies
+
+# > Third party libraries
 
 def get_arg_parser():
     parser = argparse.ArgumentParser(

src/data_generator.py

@@ -1,8 +1,12 @@
-# import the necessary packages
+# Imports
 
-# from helpers import benchmark
-import tensorflow as tf
+# > Standard Library
 import random
+
+# > Local dependencies
+
+# > Third party libraries
+import tensorflow as tf
 import elasticdeform.tf as etf
 import tensorflow_addons as tfa
 

src/inference.py

@@ -38,8 +38,6 @@ def main():
     batchSize = 1
     imgSize = (1024, 32, 1)
     maxTextLen = 128
-    epochs = 10
-    learning_rate = 0.0001
     # load training data, create TF model
     charlist = open(FilePaths.fnCharList).read()
     print(charlist)
@@ -90,18 +88,8 @@ def decode_batch_predictions(pred):
 
         _, ax = plt.subplots(4, 4, figsize=(15, 5))
         for i in range(len(pred_texts)):
-            # for i in range(16):
             print(orig_texts[i].strip())
             print(pred_texts[i].strip())
 
-
-# 		img = (batch_images[i, :, :, 0] * 255).numpy().astype(np.uint8)
-# 		img = img.T
-# 		title = f"Prediction: {pred_texts[i].strip()}"
-# 		ax[i // 4, i % 4].imshow(img, cmap="gray")
-# 		ax[i // 4, i % 4].set_title(title)
-# 		ax[i // 4, i % 4].axis("off")
-# plt.show()
-
 if __name__ == '__main__':
     main()

src/loghi_custom_callback.py

@@ -1,7 +1,12 @@
+# Imports
+
+# > Standard Library
 import os
-from tensorflow import keras
 import json
+# > Local dependencies
 
+# > Third party libraries
+from tensorflow import keras
 
 class LoghiCustomCallback(keras.callbacks.Callback):
 

src/main.py

@@ -96,7 +96,6 @@ def main():
                 exit(1)
             with open(charlist_location) as file:
                 char_list = list(char for char in file.read())
-            # char_list = sorted(list(char_list))
             print("using charlist")
             print("length charlist: " + str(len(char_list)))
             print(char_list)
@@ -194,7 +193,6 @@ def main():
 
         if args.existing_model:
             print('using existing model as base: ' + args.existing_model)
-            # if not args.replace_final_layer:
 
             if args.replace_recurrent_layer:
                 model = replace_recurrent_layer(model,
@@ -306,7 +304,6 @@ def main():
             output=args.output,
             model_name='encoder12',
             steps_per_epoch=args.steps_per_epoch,
-            # num_workers=args.num_workers,
             max_queue_size=args.max_queue_size,
             early_stopping_patience=args.early_stopping_patience,
             output_checkpoints=args.output_checkpoints,
@@ -397,18 +394,13 @@ def main():
             predictions = prediction_model.predict(batch[0])
             predicted_texts = decode_batch_predictions(predictions, utilsObject, args.greedy, args.beam_width,
                                                        args.num_oov_indices)
-
-            # preds = utils.softmax(preds)
             predsbeam = tf.transpose(predictions, perm=[1, 0, 2])
 
             if wbs:
                 print('computing wbs...')
                 label_str = wbs.compute(predsbeam)
                 char_str = []  # decoded texts for batch
-                # print(len(label_str))
-                # print(label_str)
                 for curr_label_str in label_str:
-                    # print(len(curr_label_str))
                     s = ''.join([chars[label] for label in curr_label_str])
                     char_str.append(s)
                     print(s)
@@ -424,7 +416,6 @@ def main():
                 for i in range(len(prediction)):
                     confidence = prediction[i][0]
                     predicted_text = prediction[i][1]
-                    # for i in range(16):
                     original_text = orig_texts[i].strip().replace('', '')
                     predicted_text = predicted_text.strip().replace('', '')
                     original_text = remove_tags(original_text)
@@ -455,7 +446,6 @@ def main():
                         filename = loader.get_item(
                             'validation', (batch_no * args.batch_size) + i)
                         print('\n' + filename)
-                        # print(predicted_simple)
                         print(original_text)
                         print(predicted_text)
                         if wbs:
@@ -476,7 +466,6 @@ def main():
                               + ' total_orig: ' + str(len(original_text))
                               + ' total_pred: ' + str(len(predicted_text))
                               + ' errors: ' + str(current_editdistance))
-                        # print(cer)
                         print("avg editdistance: " +
                               str(totaleditdistance / float(totallength)))
                         print("avg editdistance lower: " +

src/model.py

@@ -143,7 +143,6 @@ def update_state(self, y_true, y_pred, sample_weight=None):
 
         self.wer_accumulator.assign_add(correct_words_amount)
         self.counter.assign_add(K.cast(len(y_true), 'float32'))
-        # self.counter.assign_add(10)
 
     def result(self):
         return tf.math.divide_no_nan(self.wer_accumulator, self.counter)

src/sample_processor.py

@@ -1,7 +1,13 @@
+# Imports
+
+# > Standard Library
 from __future__ import division
 from __future__ import print_function
-
 import random
+
+# > Local dependencies
+
+# > Third party libraries
 import numpy as np
 import cv2
 

src/utils.py

@@ -1,3 +1,10 @@
+# Imports
+
+# > Standard Library
+
+# > Local dependencies
+
+# > Third party libraries
 import tensorflow as tf
 import numpy as np
 from keras.models import Model

src/visualize_filter_result.py

@@ -1,21 +1,23 @@
-from tensorflow.keras.utils import get_custom_objects
-import os
+# Imports
 
+# > Standard Library
+import random
+import argparse
+import os
 
-from config import *
-import utils
-
-from data_loader import DataLoaderNew
+# > Local dependencies
+from data_loader import DataLoader
 from model import CERMetric, WERMetric, CTCLoss
 from utils import *
-import tensorflow.keras as keras
+from config import *
 
+# > Third party libraries
+import tensorflow.keras as keras
 import numpy as np
 import tensorflow as tf
-import random
-import argparse
 from matplotlib import pyplot as plt
-import tensorflow_addons as tfa
+from tensorflow.keras.utils import get_custom_objects
+
 
 # disable GPU for now, because it is already running on my dev machine
 os.environ["CUDA_VISIBLE_DEVICES"] = "0"

src/visualize_network.py

@@ -1,23 +1,25 @@
+# Imports
+
+# > Standard Library
 import math
 import os
+import random
+import argparse
 
-import utils
+# > Local dependencies
 from model import CTCLoss, CERMetric, WERMetric
 from utils import *
-
-
-os.environ["CUDA_VISIBLE_DEVICES"] = "0"
-os.environ['TF_DETERMINISTIC_OPS'] = '0'
-
 from config import *
+
+# > Third party libraries
 import tensorflow.keras as keras
 import numpy as np
 import tensorflow as tf
-import random
-import argparse
 import tensorflow_addons as tfa
 from tensorflow.keras.utils import get_custom_objects
 
+os.environ["CUDA_VISIBLE_DEVICES"] = "0"
+os.environ['TF_DETERMINISTIC_OPS'] = '0'
 
 def compute_loss(input_image, filter_index):
     activation = feature_extractor(input_image)

src/visualize_network_saliency.py

@@ -1,17 +1,22 @@
-import metrics
+# Imports
+
+# > Standard Library
+import random
+import argparse
+
+# > Local dependencies
 from config import *
+
+
+# > Third party libraries
 import tensorflow.keras as keras
 import tensorflow.keras.backend as K
 import numpy as np
 import tensorflow as tf
-import random
-import argparse
 from matplotlib import pyplot as plt
 from tf_keras_vis.saliency import Saliency
 from tf_keras_vis.utils import normalize
 from tensorflow.keras.preprocessing.image import load_img
-# from keras.utils.generic_utils import get_custom_objects
-from tensorflow.keras.utils import get_custom_objects
 
 # from dataset_ecodices import DatasetEcodices
 # from dataset_iisg import DatasetIISG

src/visualize_siamese_network_dense.py

@@ -1,10 +1,17 @@
-from config import *
+# Imports
+
+# > Standard Library
 import metrics
-from matplotlib import pyplot as plt
-from tf_keras_vis.utils import num_of_gpus
+import argparse
+
+# > Local dependencies
+
+
+# > Third party libraries
 import tensorflow.keras as keras
 import tensorflow as tf
-import argparse
+from matplotlib import pyplot as plt
+from tf_keras_vis.utils import num_of_gpus
 from keras.utils.generic_utils import get_custom_objects
 
 os.environ["CUDA_VISIBLE_DEVICES"] = ""