fixed issue with texts_from_folder reading in data from subfolders no…

…t specified in classes argument and added maximum lr check to lrfinder

CHANGELOG.md

@@ -6,6 +6,19 @@ Most recent releases are shown at the top. Each release shows:
 - **Changed**: Additional parameters, changes to inputs or outputs, etc
 - **Fixed**: Bug fixes that don't change documented behaviour
 
+## 0.2.2 (2019-08-16)
+
+### New:
+- N/A
+
+### Changed:
+- Added check in ```ktrain.lroptimize.lrfinder``` to stop training if learning rate exceeds a fixed maximum,
+  which may happen when bad/dysfunctional model is supplied to learning rate finder.
+
+### Fixed:
+- In ```ktrain.text.data.texts_from_folder``` function, only subfolders specified in classes argument 
+  are read in as training and validation data.
+
 ## 0.2.1 (2019-08-15)
 
 ### New:
@@ -15,7 +28,7 @@ Most recent releases are shown at the top. Each release shows:
 - N/A
 
 ### Fixed:
-- Fixed error related to validation_steps=None in call to fit_generator in ```ktrain.core``` on Google Colab
+- Fixed error related to validation_steps=None in call to fit_generator in ```ktrain.core``` on Google Colab.
 
 
 ## 0.2.0 (2019-08-12)

ktrain/lroptimize/lrfinder.py

@@ -47,6 +47,11 @@ def on_batch_end(self, batch, logs):
         lr *= self.lr_mult
         K.set_value(self.model.optimizer.lr, lr)
 
+        # stop if LR grows too large
+        if lr > 10.:
+            self.model.stop_training = True
+            return
+
 
     def find(self, train_data, start_lr, lr_mult=1.01, max_epochs=None, 
              batch_size=U.DEFAULT_BS, workers=1, use_multiprocessing=False, verbose=1):

ktrain/text/data.py

@@ -51,8 +51,8 @@ def texts_from_folder(datadir, classes=None,
     # read in training and test corpora
     train_str = train_test_names[0]
     test_str = train_test_names[1]
-    train_b = load_files(os.path.join(datadir, train_str), shuffle=True)
-    test_b = load_files(os.path.join(datadir,  test_str), shuffle=False)
+    train_b = load_files(os.path.join(datadir, train_str), shuffle=True, categories=classes)
+    test_b = load_files(os.path.join(datadir,  test_str), shuffle=False, categories=classes)
     x_train = [x.decode('utf-8') for x in train_b.data]
     x_test = [x.decode('utf-8') for x in test_b.data]
     y_train = train_b.target

ktrain/version.py

@@ -1,2 +1,2 @@
 __all__ = ['__version__']
-__version__ = '0.2.1'
+__version__ = '0.2.2'