Loading deepchem/models/tensorgraph/layers.py +29 −1 Original line number Diff line number Diff line Loading @@ -643,7 +643,6 @@ class Reshape(Layer): self.out_tensor = out_tensor return out_tensor class Squeeze(Layer): def __init__(self, in_layers=None, squeeze_dims=None, **kwargs): Loading Loading @@ -1469,6 +1468,35 @@ class ReduceMean(Layer): self.out_tensor = out_tensor return out_tensor class ReduceMax(Layer): def __init__(self, in_layers=None, axis=None, **kwargs): if axis is not None and not isinstance(axis, Sequence): axis = [axis] self.axis = axis super(ReduceMax, self).__init__(in_layers, **kwargs) if axis is None: self._shape = tuple() else: try: parent_shape = self.in_layers[0].shape self._shape = [ parent_shape[i] for i in range(len(parent_shape)) if i not in axis ] except: pass def create_tensor(self, in_layers=None, set_tensors=True, **kwargs): inputs = self._get_input_tensors(in_layers) if len(inputs) > 1: self.out_tensor = tf.stack(inputs) else: self.out_tensor = inputs[0] out_tensor = tf.reduce_max(self.out_tensor, axis=self.axis) if set_tensors: self.out_tensor = out_tensor return out_tensor class ToFloat(Layer): Loading deepchem/models/tensorgraph/models/text_cnn.py +6 −10 Original line number Diff line number Diff line Loading @@ -10,7 +10,7 @@ import copy from deepchem.metrics import to_one_hot, from_one_hot from deepchem.models.tensorgraph.layers import Dense, Concat, SoftMax, \ SoftMaxCrossEntropy, BatchNorm, WeightedError, Dropout, BatchNormalization, \ Conv1D, MaxPool1D, Squeeze, Stack, Highway Conv1D, ReduceMax, Squeeze, Stack, Highway from deepchem.models.tensorgraph.graph_layers import DTNNEmbedding from deepchem.models.tensorgraph.layers import L2Loss, Label, Weights, Feature Loading Loading @@ -175,16 +175,12 @@ class TextCNNTensorGraph(TensorGraph): padding='VALID', in_layers=[self.Embedding])) # Max-over-time pooling self.pooled_outputs.append( MaxPool1D( window_shape=self.seq_length - filter_size + 1, strides=1, padding='VALID', self.pooled_outputs.append(ReduceMax( axis=1, in_layers=[self.conv_layers[-1]])) # Concat features from all filters(one feature per filter) concat_outputs = Concat(axis=2, in_layers=self.pooled_outputs) outputs = Squeeze(squeeze_dims=1, in_layers=concat_outputs) dropout = Dropout(dropout_prob=self.dropout, in_layers=[outputs]) concat_outputs = Concat(axis=1, in_layers=self.pooled_outputs) dropout = Dropout(dropout_prob=self.dropout, in_layers=[concat_outputs]) dense = Dense( out_channels=200, activation_fn=tf.nn.relu, in_layers=[dropout]) # Highway layer from https://arxiv.org/pdf/1505.00387.pdf Loading Loading
deepchem/models/tensorgraph/layers.py +29 −1 Original line number Diff line number Diff line Loading @@ -643,7 +643,6 @@ class Reshape(Layer): self.out_tensor = out_tensor return out_tensor class Squeeze(Layer): def __init__(self, in_layers=None, squeeze_dims=None, **kwargs): Loading Loading @@ -1469,6 +1468,35 @@ class ReduceMean(Layer): self.out_tensor = out_tensor return out_tensor class ReduceMax(Layer): def __init__(self, in_layers=None, axis=None, **kwargs): if axis is not None and not isinstance(axis, Sequence): axis = [axis] self.axis = axis super(ReduceMax, self).__init__(in_layers, **kwargs) if axis is None: self._shape = tuple() else: try: parent_shape = self.in_layers[0].shape self._shape = [ parent_shape[i] for i in range(len(parent_shape)) if i not in axis ] except: pass def create_tensor(self, in_layers=None, set_tensors=True, **kwargs): inputs = self._get_input_tensors(in_layers) if len(inputs) > 1: self.out_tensor = tf.stack(inputs) else: self.out_tensor = inputs[0] out_tensor = tf.reduce_max(self.out_tensor, axis=self.axis) if set_tensors: self.out_tensor = out_tensor return out_tensor class ToFloat(Layer): Loading
deepchem/models/tensorgraph/models/text_cnn.py +6 −10 Original line number Diff line number Diff line Loading @@ -10,7 +10,7 @@ import copy from deepchem.metrics import to_one_hot, from_one_hot from deepchem.models.tensorgraph.layers import Dense, Concat, SoftMax, \ SoftMaxCrossEntropy, BatchNorm, WeightedError, Dropout, BatchNormalization, \ Conv1D, MaxPool1D, Squeeze, Stack, Highway Conv1D, ReduceMax, Squeeze, Stack, Highway from deepchem.models.tensorgraph.graph_layers import DTNNEmbedding from deepchem.models.tensorgraph.layers import L2Loss, Label, Weights, Feature Loading Loading @@ -175,16 +175,12 @@ class TextCNNTensorGraph(TensorGraph): padding='VALID', in_layers=[self.Embedding])) # Max-over-time pooling self.pooled_outputs.append( MaxPool1D( window_shape=self.seq_length - filter_size + 1, strides=1, padding='VALID', self.pooled_outputs.append(ReduceMax( axis=1, in_layers=[self.conv_layers[-1]])) # Concat features from all filters(one feature per filter) concat_outputs = Concat(axis=2, in_layers=self.pooled_outputs) outputs = Squeeze(squeeze_dims=1, in_layers=concat_outputs) dropout = Dropout(dropout_prob=self.dropout, in_layers=[outputs]) concat_outputs = Concat(axis=1, in_layers=self.pooled_outputs) dropout = Dropout(dropout_prob=self.dropout, in_layers=[concat_outputs]) dense = Dense( out_channels=200, activation_fn=tf.nn.relu, in_layers=[dropout]) # Highway layer from https://arxiv.org/pdf/1505.00387.pdf Loading
