consistently use double quotes for strings across the entire library + some styling adjustments

Author: seidnerj

Tests/test_text_utils.py

@@ -10,28 +10,28 @@ def test_edit_distance(self):
         errors. It also includes a test case for empty input.
         """
         # Test simple case with no errors
-        prediction_tokens = ['A', 'B', 'C']
-        reference_tokens = ['A', 'B', 'C']
+        prediction_tokens = ["A", "B", "C"]
+        reference_tokens = ["A", "B", "C"]
         self.assertEqual(edit_distance(prediction_tokens, reference_tokens), 0)
 
         # Test simple case with one substitution error
-        prediction_tokens = ['A', 'B', 'D']
-        reference_tokens = ['A', 'B', 'C']
+        prediction_tokens = ["A", "B", "D"]
+        reference_tokens = ["A", "B", "C"]
         self.assertEqual(edit_distance(prediction_tokens, reference_tokens), 1)
 
         # Test simple case with one insertion error
-        prediction_tokens = ['A', 'B', 'C']
-        reference_tokens = ['A', 'B', 'C', 'D']
+        prediction_tokens = ["A", "B", "C"]
+        reference_tokens = ["A", "B", "C", "D"]
         self.assertEqual(edit_distance(prediction_tokens, reference_tokens), 1)
 
         # Test simple case with one deletion error
-        prediction_tokens = ['A', 'B']
-        reference_tokens = ['A', 'B', 'C']
+        prediction_tokens = ["A", "B"]
+        reference_tokens = ["A", "B", "C"]
         self.assertEqual(edit_distance(prediction_tokens, reference_tokens), 1)
 
         # Test more complex case with multiple errors
-        prediction_tokens = ['A', 'B', 'C', 'D', 'E']
-        reference_tokens = ['A', 'C', 'B', 'F', 'E']
+        prediction_tokens = ["A", "B", "C", "D", "E"]
+        reference_tokens = ["A", "C", "B", "F", "E"]
         self.assertEqual(edit_distance(prediction_tokens, reference_tokens), 3)
 
         # Test empty input
@@ -41,18 +41,18 @@ def test_edit_distance(self):
 
     def test_get_cer(self):
         # Test simple case with no errors
-        preds = ['A B C']
-        target = ['A B C']
+        preds = ["A B C"]
+        target = ["A B C"]
         self.assertEqual(get_cer(preds, target), 0)
 
         # Test simple case with one character error
-        preds = ['A B C']
-        target = ['A B D']
+        preds = ["A B C"]
+        target = ["A B D"]
         self.assertEqual(get_cer(preds, target), 1/5)
 
         # Test simple case with multiple character errors
-        preds = ['A B C']
-        target = ['D E F']
+        preds = ["A B C"]
+        target = ["D E F"]
         self.assertEqual(get_cer(preds, target), 3/5)
 
         # Test empty input
@@ -61,24 +61,24 @@ def test_get_cer(self):
         self.assertEqual(get_cer(preds, target), 0)
 
         # Test simple case with different word lengths
-        preds = ['ABC']
-        target = ['ABCDEFG']
+        preds = ["ABC"]
+        target = ["ABCDEFG"]
         self.assertEqual(get_cer(preds, target), 4/7)
 
     def test_get_wer(self):
         # Test simple case with no errors
-        preds = 'A B C'
-        target = 'A B C'
+        preds = "A B C"
+        target = "A B C"
         self.assertEqual(get_wer(preds, target), 0)
 
         # Test simple case with one word error
-        preds = 'A B C'
-        target = 'A B D'
+        preds = "A B C"
+        target = "A B D"
         self.assertEqual(get_wer(preds, target), 1/3)
 
         # Test simple case with multiple word errors
-        preds = 'A B C'
-        target = 'D E F'
+        preds = "A B C"
+        target = "D E F"
         self.assertEqual(get_wer(preds, target), 1)
 
         # Test empty input
@@ -87,9 +87,10 @@ def test_get_wer(self):
         self.assertEqual(get_wer(preds, target), 0)
 
         # Test simple case with different sentence lengths
-        preds = ['ABC']
-        target = ['ABC DEF']
+        preds = ["ABC"]
+        target = ["ABC DEF"]
         self.assertEqual(get_wer(preds, target), 1)
 
-if __name__ == '__main__':
-    unittest.main()
+
+if __name__ == "__main__":
+    unittest.main()

Tutorials/01_image_to_word/configs.py

@@ -3,11 +3,12 @@
 
 from mltu.configs import BaseModelConfigs
 
+
 class ModelConfigs(BaseModelConfigs):
     def __init__(self):
         super().__init__()
-        self.model_path = os.path.join('Models/1_image_to_word', datetime.strftime(datetime.now(), "%Y%m%d%H%M"))
-        self.vocab = '0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz'
+        self.model_path = os.path.join("Models/1_image_to_word", datetime.strftime(datetime.now(), "%Y%m%d%H%M"))
+        self.vocab = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"
         self.height = 32
         self.width = 128
         self.max_text_length = 23

Tutorials/01_image_to_word/model.py

@@ -3,7 +3,8 @@
 
 from mltu.tensorflow.model_utils import residual_block
 
-def train_model(input_dim, output_dim, activation='leaky_relu', dropout=0.2):
+
+def train_model(input_dim, output_dim, activation="leaky_relu", dropout=0.2):
 
     inputs = layers.Input(shape=input_dim, name="input")
 
@@ -24,7 +25,7 @@ def train_model(input_dim, output_dim, activation='leaky_relu', dropout=0.2):
 
     blstm = layers.Bidirectional(layers.LSTM(64, return_sequences=True))(squeezed)
 
-    output = layers.Dense(output_dim + 1, activation='softmax', name="output")(blstm)
+    output = layers.Dense(output_dim + 1, activation="softmax", name="output")(blstm)
 
     model = Model(inputs=inputs, outputs=output)
     return model

Tutorials/01_image_to_word/train.py

@@ -2,7 +2,7 @@
 from tqdm import tqdm
 import tensorflow as tf
 
-try: [tf.config.experimental.set_memory_growth(gpu, True) for gpu in tf.config.experimental.list_physical_devices('GPU')]
+try: [tf.config.experimental.set_memory_growth(gpu, True) for gpu in tf.config.experimental.list_physical_devices("GPU")]
 except: pass
 
 from keras.callbacks import EarlyStopping, ModelCheckpoint, ReduceLROnPlateau, TensorBoard
@@ -89,11 +89,11 @@ def read_annotation_file(annotation_path):
 os.makedirs(configs.model_path, exist_ok=True)
 
 # Define callbacks
-earlystopper = EarlyStopping(monitor='val_CER', patience=10, verbose=1)
-checkpoint = ModelCheckpoint(f"{configs.model_path}/model.h5", monitor='val_CER', verbose=1, save_best_only=True, mode='min')
+earlystopper = EarlyStopping(monitor="val_CER", patience=10, verbose=1)
+checkpoint = ModelCheckpoint(f"{configs.model_path}/model.h5", monitor="val_CER", verbose=1, save_best_only=True, mode="min")
 trainLogger = TrainLogger(configs.model_path)
-tb_callback = TensorBoard(f'{configs.model_path}/logs', update_freq=1)
-reduceLROnPlat = ReduceLROnPlateau(monitor='val_CER', factor=0.9, min_delta=1e-10, patience=5, verbose=1, mode='auto')
+tb_callback = TensorBoard(f"{configs.model_path}/logs", update_freq=1)
+reduceLROnPlat = ReduceLROnPlateau(monitor="val_CER", factor=0.9, min_delta=1e-10, patience=5, verbose=1, mode="auto")
 model2onnx = Model2onnx(f"{configs.model_path}/model.h5")
 
 # Train the model
@@ -106,5 +106,5 @@ def read_annotation_file(annotation_path):
 )
 
 # Save training and validation datasets as csv files
-train_data_provider.to_csv(os.path.join(configs.model_path, 'train.csv'))
-val_data_provider.to_csv(os.path.join(configs.model_path, 'val.csv'))
+train_data_provider.to_csv(os.path.join(configs.model_path, "train.csv"))
+val_data_provider.to_csv(os.path.join(configs.model_path, "val.csv"))

Tutorials/02_captcha_to_text/configs.py

@@ -3,11 +3,12 @@
 
 from mltu.configs import BaseModelConfigs
 
+
 class ModelConfigs(BaseModelConfigs):
     def __init__(self):
         super().__init__()
-        self.model_path = os.path.join('Models/02_captcha_to_text', datetime.strftime(datetime.now(), "%Y%m%d%H%M"))
-        self.vocab = ''
+        self.model_path = os.path.join("Models/02_captcha_to_text", datetime.strftime(datetime.now(), "%Y%m%d%H%M"))
+        self.vocab = ""
         self.height = 50
         self.width = 200
         self.max_text_length = 0

Tutorials/02_captcha_to_text/model.py

@@ -3,7 +3,8 @@
 
 from mltu.tensorflow.model_utils import residual_block
 
-def train_model(input_dim, output_dim, activation='leaky_relu', dropout=0.2):
+
+def train_model(input_dim, output_dim, activation="leaky_relu", dropout=0.2):
 
     inputs = layers.Input(shape=input_dim, name="input")
 
@@ -29,7 +30,7 @@ def train_model(input_dim, output_dim, activation='leaky_relu', dropout=0.2):
     blstm = layers.Bidirectional(layers.LSTM(128, return_sequences=True))(squeezed)
     blstm = layers.Dropout(dropout)(blstm)
 
-    output = layers.Dense(output_dim + 1, activation='softmax', name="output")(blstm)
+    output = layers.Dense(output_dim + 1, activation="softmax", name="output")(blstm)
 
     model = Model(inputs=inputs, outputs=output)
-    return model
+    return model

Tutorials/02_captcha_to_text/train.py

@@ -1,5 +1,5 @@
 import tensorflow as tf
-try: [tf.config.experimental.set_memory_growth(gpu, True) for gpu in tf.config.experimental.list_physical_devices('GPU')]
+try: [tf.config.experimental.set_memory_growth(gpu, True) for gpu in tf.config.experimental.list_physical_devices("GPU")]
 except: pass
 
 from keras.callbacks import EarlyStopping, ModelCheckpoint, ReduceLROnPlateau, TensorBoard
@@ -22,17 +22,20 @@
 from io import BytesIO
 from zipfile import ZipFile
 
-def download_and_unzip(url, extract_to='Datasets'):
+
+def download_and_unzip(url, extract_to="Datasets"):
     http_response = urlopen(url)
     zipfile = ZipFile(BytesIO(http_response.read()))
     zipfile.extractall(path=extract_to)
 
-if not os.path.exists(os.path.join('Datasets', 'captcha_images_v2')):
-    download_and_unzip('https://github.com/AakashKumarNain/CaptchaCracker/raw/master/captcha_images_v2.zip', extract_to='Datasets')
+
+if not os.path.exists(os.path.join("Datasets", "captcha_images_v2")):
+    download_and_unzip("https://github.com/AakashKumarNain/CaptchaCracker/raw/master/captcha_images_v2.zip",
+                       extract_to="Datasets")
 
 # Create a list of all the images and labels in the dataset
 dataset, vocab, max_len = [], set(), 0
-captcha_path = os.path.join('Datasets', 'captcha_images_v2')
+captcha_path = os.path.join("Datasets", "captcha_images_v2")
 for file in os.listdir(captcha_path):
     file_path = os.path.join(captcha_path, file)
     label = os.path.splitext(file)[0] # Get the file name without the extension
@@ -83,11 +86,11 @@ def download_and_unzip(url, extract_to='Datasets'):
 os.makedirs(configs.model_path, exist_ok=True)
 
 # Define callbacks
-earlystopper = EarlyStopping(monitor='val_CER', patience=50, verbose=1)
-checkpoint = ModelCheckpoint(f"{configs.model_path}/model.h5", monitor='val_CER', verbose=1, save_best_only=True, mode='min')
+earlystopper = EarlyStopping(monitor="val_CER", patience=50, verbose=1)
+checkpoint = ModelCheckpoint(f"{configs.model_path}/model.h5", monitor="val_CER", verbose=1, save_best_only=True, mode="min")
 trainLogger = TrainLogger(configs.model_path)
-tb_callback = TensorBoard(f'{configs.model_path}/logs', update_freq=1)
-reduceLROnPlat = ReduceLROnPlateau(monitor='val_CER', factor=0.9, min_delta=1e-10, patience=20, verbose=1, mode='auto')
+tb_callback = TensorBoard(f"{configs.model_path}/logs", update_freq=1)
+reduceLROnPlat = ReduceLROnPlateau(monitor="val_CER", factor=0.9, min_delta=1e-10, patience=20, verbose=1, mode="auto")
 model2onnx = Model2onnx(f"{configs.model_path}/model.h5")
 
 # Train the model
@@ -100,5 +103,5 @@ def download_and_unzip(url, extract_to='Datasets'):
 )
 
 # Save training and validation datasets as csv files
-train_data_provider.to_csv(os.path.join(configs.model_path, 'train.csv'))
-val_data_provider.to_csv(os.path.join(configs.model_path, 'val.csv'))
+train_data_provider.to_csv(os.path.join(configs.model_path, "train.csv"))
+val_data_provider.to_csv(os.path.join(configs.model_path, "val.csv"))

Tutorials/03_handwriting_recognition/configs.py

@@ -6,8 +6,8 @@
 class ModelConfigs(BaseModelConfigs):
     def __init__(self):
         super().__init__()
-        self.model_path = os.path.join('Models/03_handwriting_recognition', datetime.strftime(datetime.now(), "%Y%m%d%H%M"))
-        self.vocab = ''
+        self.model_path = os.path.join("Models/03_handwriting_recognition", datetime.strftime(datetime.now(), "%Y%m%d%H%M"))
+        self.vocab = ""
         self.height = 32
         self.width = 128
         self.max_text_length = 0

Tutorials/03_handwriting_recognition/model.py

@@ -3,7 +3,8 @@
 
 from mltu.tensorflow.model_utils import residual_block
 
-def train_model(input_dim, output_dim, activation='leaky_relu', dropout=0.2):
+
+def train_model(input_dim, output_dim, activation="leaky_relu", dropout=0.2):
 
     inputs = layers.Input(shape=input_dim, name="input")
 
@@ -29,7 +30,7 @@ def train_model(input_dim, output_dim, activation='leaky_relu', dropout=0.2):
     blstm = layers.Bidirectional(layers.LSTM(128, return_sequences=True))(squeezed)
     blstm = layers.Dropout(dropout)(blstm)
 
-    output = layers.Dense(output_dim + 1, activation='softmax', name="output")(blstm)
+    output = layers.Dense(output_dim + 1, activation="softmax", name="output")(blstm)
 
     model = Model(inputs=inputs, outputs=output)
-    return model
+    return model

Tutorials/03_handwriting_recognition/train.py

@@ -1,5 +1,5 @@
 import tensorflow as tf
-try: [tf.config.experimental.set_memory_growth(gpu, True) for gpu in tf.config.experimental.list_physical_devices('GPU')]
+try: [tf.config.experimental.set_memory_growth(gpu, True) for gpu in tf.config.experimental.list_physical_devices("GPU")]
 except: pass
 
 from keras.callbacks import EarlyStopping, ModelCheckpoint, ReduceLROnPlateau, TensorBoard
@@ -24,20 +24,21 @@
 from io import BytesIO
 from zipfile import ZipFile
 
-def download_and_unzip(url, extract_to='Datasets', chunk_size=1024*1024):
+
+def download_and_unzip(url, extract_to="Datasets", chunk_size=1024*1024):
     http_response = urlopen(url)
 
-    data = b''
+    data = b""
     iterations = http_response.length // chunk_size + 1
     for _ in tqdm(range(iterations)):
         data += http_response.read(chunk_size)
 
     zipfile = ZipFile(BytesIO(data))
     zipfile.extractall(path=extract_to)
 
-dataset_path = os.path.join('Datasets', 'IAM_Words')
+dataset_path = os.path.join("Datasets", "IAM_Words")
 if not os.path.exists(dataset_path):
-    download_and_unzip('https://git.io/J0fjL', extract_to='Datasets')
+    download_and_unzip("https://git.io/J0fjL", extract_to="Datasets")
 
     file = tarfile.open(os.path.join(dataset_path, "words.tgz"))
     file.extractall(os.path.join(dataset_path, "words"))
@@ -57,7 +58,7 @@ def download_and_unzip(url, extract_to='Datasets', chunk_size=1024*1024):
     folder1 = line_split[0][:3]
     folder2 = "-".join(line_split[0].split("-")[:2])
     file_name = line_split[0] + ".png"
-    label = line_split[-1].rstrip('\n')
+    label = line_split[-1].rstrip("\n")
 
     rel_path = os.path.join(dataset_path, "words", folder1, folder2, file_name)
     if not os.path.exists(rel_path):
@@ -115,11 +116,11 @@ def download_and_unzip(url, extract_to='Datasets', chunk_size=1024*1024):
 model.summary(line_length=110)
 
 # Define callbacks
-earlystopper = EarlyStopping(monitor='val_CER', patience=20, verbose=1)
-checkpoint = ModelCheckpoint(f"{configs.model_path}/model.h5", monitor='val_CER', verbose=1, save_best_only=True, mode='min')
+earlystopper = EarlyStopping(monitor="val_CER", patience=20, verbose=1)
+checkpoint = ModelCheckpoint(f"{configs.model_path}/model.h5", monitor="val_CER", verbose=1, save_best_only=True, mode="min")
 trainLogger = TrainLogger(configs.model_path)
-tb_callback = TensorBoard(f'{configs.model_path}/logs', update_freq=1)
-reduceLROnPlat = ReduceLROnPlateau(monitor='val_CER', factor=0.9, min_delta=1e-10, patience=10, verbose=1, mode='auto')
+tb_callback = TensorBoard(f"{configs.model_path}/logs", update_freq=1)
+reduceLROnPlat = ReduceLROnPlateau(monitor="val_CER", factor=0.9, min_delta=1e-10, patience=10, verbose=1, mode="auto")
 model2onnx = Model2onnx(f"{configs.model_path}/model.h5")
 
 # Train the model
@@ -132,5 +133,5 @@ def download_and_unzip(url, extract_to='Datasets', chunk_size=1024*1024):
 )
 
 # Save training and validation datasets as csv files
-train_data_provider.to_csv(os.path.join(configs.model_path, 'train.csv'))
-val_data_provider.to_csv(os.path.join(configs.model_path, 'val.csv'))
+train_data_provider.to_csv(os.path.join(configs.model_path, "train.csv"))
+val_data_provider.to_csv(os.path.join(configs.model_path, "val.csv"))