Added LeakyReLU layer support.

Author: gmalivenko

README.md

@@ -81,7 +81,7 @@ Reshape:
 Activations:
 
 * ReLU
-* LeakyReLU (only with 0.2)
+* LeakyReLU
 * PReLU (only with 0.2)
 * SELU (only with 0.2)
 * Tanh

pytorch2keras/layers.py

@@ -475,6 +475,26 @@ def convert_relu(params, w_name, scope_name, inputs, layers, weights):
     layers[scope_name] = relu(layers[inputs[0]])
 
 
+def convert_lrelu(params, w_name, scope_name, inputs, layers, weights):
+    """
+    Convert leaky relu layer.
+
+   Args:
+        params: dictionary with layer parameters
+        w_name: name prefix in state_dict
+        scope_name: pytorch scope name
+        inputs: pytorch node inputs
+        layers: dictionary with keras tensors
+        weights: pytorch state_dict
+    """
+    print('Converting lrelu ...')
+
+    tf_name = w_name + str(random.random())
+    leakyrelu = \
+        keras.layers.LeakyReLU(alpha=params['alpha'], name=tf_name)
+    layers[scope_name] = leakyrelu(layers[inputs[0]])
+
+
 def convert_sigmoid(params, w_name, scope_name, inputs, layers, weights):
     """
     Convert sigmoid layer.
@@ -732,6 +752,7 @@ def convert_upsample(params, w_name, scope_name, inputs, layers, weights):
     'Sub': convert_elementwise_sub,
     'Concat': convert_concat,
     'Relu': convert_relu,
+    'LeakyRelu': convert_lrelu,
     'Sigmoid': convert_sigmoid,
     'Softmax': convert_softmax,
     'Tanh': convert_tanh,

tests/lrelu.py

@@ -0,0 +1,49 @@
+import keras  # work around segfault
+import sys
+import numpy as np
+
+import torch
+import torch.nn as nn
+from torch.autograd import Variable
+
+sys.path.append('../pytorch2keras')
+from converter import pytorch_to_keras
+
+
+class TestLeakyRelu(nn.Module):
+    """Module for LeakyReLu conversion testing
+    """
+
+    def __init__(self, inp=10, out=16, bias=True):
+        super(TestLeakyRelu, self).__init__()
+        self.linear = nn.Linear(inp, out, bias=True)
+        self.relu = nn.LeakyReLU(inplace=True)
+
+    def forward(self, x):
+        x = self.linear(x)
+        x = self.relu(x)
+        return x
+
+
+if __name__ == '__main__':
+    max_error = 0
+    for i in range(100):
+        inp = np.random.randint(1, 100)
+        out = np.random.randint(1, 100)
+        model = TestLeakyRelu(inp, out, inp % 2)
+
+        input_np = np.random.uniform(0, 1, (1, inp))
+        input_var = Variable(torch.FloatTensor(input_np))
+        output = model(input_var)
+
+        k_model = pytorch_to_keras(model, input_var, (inp,), verbose=True)
+
+        pytorch_output = output.data.numpy()
+        keras_output = k_model.predict(input_np)
+
+        error = np.max(pytorch_output - keras_output)
+        print(error)
+        if max_error < error:
+            max_error = error
+
+    print('Max error: {0}'.format(max_error))