rename logcosh to LogCosh

docs/Examples_Generator.md

@@ -84,7 +84,7 @@ return out, model
 ```python
 p = {'activation':['relu', 'elu'],
      'optimizer': ['Adam'],
-     'losses': ['logcosh'],
+     'losses': ['LogCosh'],
      'shapes': ['brick'],
      'first_neuron': [32],
      'dropout': [.2, .3],

docs/Examples_Generator_Code.md

@@ -37,7 +37,7 @@ def mnist_model(x_train, y_train, x_val, y_val, params):
 
 p = {'activation':['relu', 'elu'],
      'optimizer': ['Adam'],
-     'losses': ['logcosh'],
+     'losses': ['LogCosh'],
      'shapes': ['brick'],
      'first_neuron': [32],
      'dropout': [.2, .3],

docs/Examples_PyTorch.md

@@ -139,7 +139,7 @@ return net, net.parameters()
 ```python
 p = {'activation':['relu', 'elu'],
        'optimizer': ['Adagrad', 'Adam'],
-       'losses': ['logcosh'],
+       'losses': ['LogCosh'],
        'hidden_layers':[0, 1, 2],
        'batch_size': (20, 50, 5),
        'epochs': [10, 20]}

docs/Examples_PyTorch_Code.md

@@ -107,7 +107,7 @@ def breast_cancer(x_train, y_train, x_val, y_val, params):
 
 p = {'activation':['relu', 'elu'],
        'optimizer': ['Adagrad', 'Adam'],
-       'losses': ['logcosh'],
+       'losses': ['LogCosh'],
        'hidden_layers':[0, 1, 2],
        'batch_size': (20, 50, 5),
        'epochs': [10, 20]}

docs/Examples_Typical.md

@@ -56,7 +56,7 @@ return out, model
 ```python
 p = {'activation':['relu', 'elu'],
      'optimizer': ['Adagrad', 'Adam'],
-     'losses': ['logcosh'],
+     'losses': ['LogCosh'],
      'hidden_layers':[0, 1, 2],
      'batch_size': (20, 50, 5),
      'epochs': [10, 20]}

docs/Hidden_Layers.md

@@ -19,7 +19,7 @@ When hidden layers are used, `dropout`, `shapes`, `hidden_layers`, and `first_ne
 
     p = {'activation':['relu', 'elu'],
          'optimizer': ['Adagrad', 'Adam'],
-         'losses': ['logcosh'],
+         'losses': ['LogCosh'],
          'shapes': ['brick'],          # <<< required
          'first_neuron': [32, 64],     # <<< required
          'hidden_layers':[0, 1, 2],    # <<< required

examples/Hyperparameter Optimization with Keras for the Iris Prediction.ipynb

@@ -147,7 +147,7 @@
    "source": [
     "from tensorflow.keras.optimizers.legacy import Adam, Adagrad\n",
     "from tensorflow.keras.activations import softmax\n",
-    "from tensorflow.keras.losses import categorical_crossentropy, logcosh\n",
+    "from tensorflow.keras.losses import categorical_crossentropy, LogCosh\n",
     "\n",
     "p = {'lr': (0.1, 10, 10),\n",
     "     'first_neuron':[4, 8, 16, 32, 64, 128],\n",

examples/Recover Best Models from Experiment Log.ipynb

@@ -48,7 +48,7 @@
     "# set the parameter space boundary\n",
     "p = {'activation':['relu', 'elu'],\n",
     "     'optimizer': ['Adagrad', 'Adam'],\n",
-    "     'losses': ['logcosh'],\n",
+    "     'losses': ['LogCosh'],\n",
     "     'shapes': ['brick'],\n",
     "     'first_neuron': [16, 32, 64, 128],\n",
     "     'hidden_layers':[0, 1, 2, 3],\n",

talos/autom8/autoparams.py

@@ -2,7 +2,7 @@
 from tensorflow.keras.optimizers.legacy import Adam, Adagrad, SGD
 
 
-loss = {'binary': ['binary_crossentropy', 'logcosh'],
+loss = {'binary': ['binary_crossentropy', 'LogCosh'],
         'multi_class': ['sparse_categorical_crossentropy'],
         'multi_label': ['categorical_crossentropy'],
         'continuous': ['mae']}

talos/templates/params.py

@@ -9,7 +9,7 @@ def titanic(debug=False):
          'dropout': (0, 0.5, 5),
          'optimizer': [Adam(), Adagrad()],
          'epochs': [50, 100, 150],
-         'losses': ['logcosh', 'binary_crossentropy'],
+         'losses': ['LogCosh', 'binary_crossentropy'],
          'shapes': ['brick', 'triangle', 0.2],
          'hidden_layers': [0, 1, 2, 3, 4],
          'activation': ['relu', 'elu'],
@@ -23,7 +23,7 @@ def titanic(debug=False):
              'dropout': [0.2, 0.3],
              'optimizer': [Adam(), Adagrad()],
              'epochs': [50, 100],
-             'losses': ['logcosh', 'binary_crossentropy'],
+             'losses': ['LogCosh', 'binary_crossentropy'],
              'shapes': ['brick', 'triangle', 0.2],
              'hidden_layers': [0, 1],
              'activation': ['relu', 'elu'],
@@ -35,7 +35,7 @@ def titanic(debug=False):
 def iris():
 
     from tensorflow.keras.optimizers.legacy import Adam, Adagrad
-    from tensorflow.keras.losses import logcosh, categorical_crossentropy
+    from tensorflow.keras.losses import LogCosh, categorical_crossentropy
     from tensorflow.keras.activations import relu, elu, softmax
 
     # here use a standard 2d dictionary for inputting the param boundaries
@@ -49,7 +49,7 @@ def iris():
          'emb_output_dims': [None],
          'shapes': ['brick', 'triangle', 0.2],
          'optimizer': [Adam, Adagrad],
-         'losses': [logcosh, categorical_crossentropy],
+         'losses': [LogCosh, categorical_crossentropy],
          'activation': [relu, elu],
          'last_activation': [softmax]}
 
@@ -59,7 +59,7 @@ def iris():
 def breast_cancer():
 
     from tensorflow.keras.optimizers.legacy import Adam, Adagrad, RMSprop
-    from tensorflow.keras.losses import logcosh, binary_crossentropy
+    from tensorflow.keras.losses import LogCosh, binary_crossentropy
     from tensorflow.keras.activations import relu, elu, sigmoid
 
     # then we can go ahead and set the parameter space
@@ -71,7 +71,7 @@ def breast_cancer():
          'dropout': (0, 0.5, 5),
          'shapes': ['brick', 'triangle', 'funnel'],
          'optimizer': [Adam, Adagrad, RMSprop],
-         'losses': [logcosh, binary_crossentropy],
+         'losses': [LogCosh, binary_crossentropy],
          'activation': [relu, elu],
          'last_activation': [sigmoid]}
 

tests/commands/test_latest.py

@@ -14,7 +14,7 @@ def test_latest():
 
     p = {'activation': ['relu', 'elu'],
          'optimizer': ['Adagrad', 'Adam'],
-         'losses': ['logcosh'],
+         'losses': ['LogCosh'],
          'shapes': ['brick'],
          'first_neuron': [16, 32, 64, 128],
          'hidden_layers': [0, 1, 2, 3],

tests/commands/test_rest.py

@@ -45,23 +45,23 @@ def test_rest(scan_object):
     model1 = Sequential()
     model1.add(Dense(10, input_dim=x.shape[1]))
     model1.add(Dense(1))
-    model1.compile('adam', 'logcosh', metrics=metrics)
+    model1.compile('adam', 'LogCosh', metrics=metrics)
     model1.fit(x, y, callbacks=callbacks)
 
     print('\n ...generator... \n')
 
     model2 = Sequential()
     model2.add(Dense(10, input_dim=x.shape[1]))
     model2.add(Dense(1))
-    model2.compile('adam', 'logcosh')
+    model2.compile('adam', 'LogCosh')
     model2.fit_generator(talos.utils.generator(x, y, 10), 5)
 
     print('\n ...SequenceGenerator... \n')
 
     model3 = Sequential()
     model3.add(Dense(10, input_dim=x.shape[1]))
     model3.add(Dense(1))
-    model3.compile('adam', 'logcosh')
+    model3.compile('adam', 'LogCosh')
     model3.fit_generator(talos.utils.SequenceGenerator(x, y, 10))
 
     print('\n ...gpu_utils... \n')

tests/commands/test_scan.py

@@ -12,7 +12,7 @@ def test_scan():
 
     p = {'activation': [relu, elu],
          'optimizer': ['Adagrad', Adam],
-         'losses': ['logcosh', binary_crossentropy],
+         'losses': ['LogCosh', binary_crossentropy],
          'shapes': ['brick', 'funnel', 'triangle'],
          'first_neuron': [16],
          'hidden_layers': ([0, 1, 2, 3]),
@@ -51,7 +51,7 @@ def iris_model(x_train, y_train, x_val, y_val, params):
 
     p_for_q = {'activation': ['relu', 'elu'],
                'optimizer': ['Adagrad', 'Adam'],
-               'losses': ['logcosh'],
+               'losses': ['LogCosh'],
                'shapes': ['brick'],
                'first_neuron': [16, 32, 64, 128],
                'hidden_layers': [0, 1, 2, 3],