Weave only one out_layer (d39add8a) · Commits · 钟慕尧 / deepchem

deepchem/models/tensorgraph/graph_layers.py

+21 −16

Original line number	Diff line number	Diff line
		@@ -16,7 +16,7 @@ from deepchem.nn import activations
		from deepchem.nn import initializations
		from deepchem.nn import model_ops

		from deepchem.models.tensorgraph.layers import Layer
		from deepchem.models.tensorgraph.layers import Layer, PassThroughLayer
		from deepchem.models.tensorgraph.layers import convert_to_layers


		@@ -55,7 +55,7 @@ class Separate_AP(Layer):
		class WeaveLayer(Layer):
		""" TensorGraph style implementation
		The same as deepchem.nn.WeaveLayer

		Note: Use WeaveLayerFactory to construct this layer
		"""

		def __init__(self,
		@@ -167,11 +167,11 @@ class WeaveLayer(Layer):

		self.build()

		atom_features = in_layers[0].out_tensor[0]
		pair_features = in_layers[0].out_tensor[1]
		atom_features = in_layers[1].out_tensor
		pair_features = in_layers[2].out_tensor

		pair_split = in_layers[1].out_tensor
		atom_to_pair = in_layers[2].out_tensor
		pair_split = in_layers[3].out_tensor
		atom_to_pair = in_layers[4].out_tensor

		AA = tf.matmul(atom_features, self.W_AA) + self.b_AA
		AA = self.activation(AA)
		@@ -201,11 +201,11 @@ class WeaveLayer(Layer):
		else:
		P = pair_features

		out_tensor = [A, P]
		self.out_tensors = [A, P]
		if set_tensors:
		self.variables = self.trainable_weights
		self.out_tensor = out_tensor
		return out_tensor
		self.out_tensor = A
		return self.out_tensors

		def none_tensors(self):
		W_AP, b_AP, W_PP, W_PP, W_P, b_P = self.W_AP, self.b_AP, self.W_PP, self.W_PP, self.W_P, self.b_P
		@@ -214,12 +214,12 @@ class WeaveLayer(Layer):
		W_AA, b_AA, W_PA, b_PA, W_A, b_A = self.W_AA, self.b_AA, self.W_PA, self.b_PA, self.W_A, self.b_A
		self.W_AA, self.b_AA, self.W_PA, self.b_PA, self.W_A, self.b_A = None, None, None, None, None, None

		out_tensor, trainable_weights, variables = self.out_tensor, self.trainable_weights, self.variables
		self.out_tensor, self.trainable_weights, self.variables, self.activation, self.init = None, [], [], None, None
		out_tensor, out_tensors, trainable_weights, variables = self.out_tensor, self.out_tensors, self.trainable_weights, self.variables
		self.out_tensor, self.out_tensors, self.trainable_weights, self.variables, self.activation, self.init = None, [], [], [], None, None

		return W_AP, b_AP, W_PP, W_PP, W_P, b_P, \
		W_AA, b_AA, W_PA, b_PA, W_A, b_A, \
		out_tensor, trainable_weights, variables
		out_tensor, out_tensors, trainable_weights, variables

		def set_tensors(self, tensor):
		self.W_AP, self.b_AP, self.W_PP, self.W_PP, self.W_P, self.b_P, \
		@@ -227,6 +227,11 @@ class WeaveLayer(Layer):
		self.out_tensor, self.trainable_weights, self.variables = tensor


		def WeaveLayerFactory(**kwargs):
		weaveLayer = WeaveLayer(**kwargs)
		return [PassThroughLayer(i, in_layers=weaveLayer) for i in range(2)]


		class WeaveGather(Layer):
		""" TensorGraph style implementation
		The same as deepchem.nn.WeaveGather

deepchem/models/tensorgraph/layers.py

+23 −3

Original line number	Diff line number	Diff line
		@@ -3212,3 +3212,23 @@ class BetaShare(Layer):

		def set_tensors(self, tensor):
		self.out_tensor, self.betas = tensor


		class PassThroughLayer(Layer):
		"""
		Layer which takes a tensor from in_tensor[0].out_tensors at an index
		"""

		def __init__(self, output_num, **kwargs):
		"""
		Parameters
		----------
		output_num: int
		The index which to use as this layers out_tensor from in_layers[0]
		kwargs
		"""
		self.output_num = output_num
		super(PassThroughLayer, self).__init__(**kwargs)

		def create_tensor(self, in_layers=None, set_tensors=True, **kwargs):
		self.out_tensor = self.in_layers[0].out_tensors[self.output_num]

deepchem/models/tensorgraph/models/graph_models.py

+10 −6

Original line number	Diff line number	Diff line
		@@ -7,7 +7,7 @@ from deepchem.feat.mol_graphs import ConvMol
		from deepchem.metrics import to_one_hot, from_one_hot
		from deepchem.models.tensorgraph.graph_layers import WeaveLayer, WeaveGather, \
		Combine_AP, Separate_AP, DTNNEmbedding, DTNNStep, DTNNGather, DAGLayer, \
		DAGGather, DTNNExtract, MessagePassing, SetGather
		DAGGather, DTNNExtract, MessagePassing, SetGather, WeaveLayerFactory
		from deepchem.models.tensorgraph.layers import Dense, Concat, SoftMax, \
		SoftMaxCrossEntropy, GraphConv, BatchNorm, \
		GraphPool, GraphGather, WeightedError, Dropout, BatchNormalization, Stack
		@@ -61,23 +61,27 @@ class WeaveTensorGraph(TensorGraph):
		"""
		self.atom_features = Feature(shape=(None, self.n_atom_feat))
		self.pair_features = Feature(shape=(None, self.n_pair_feat))
		combined = Combine_AP(in_layers=[self.atom_features, self.pair_features])
		self.pair_split = Feature(shape=(None,), dtype=tf.int32)
		self.atom_split = Feature(shape=(None,), dtype=tf.int32)
		self.atom_to_pair = Feature(shape=(None, 2), dtype=tf.int32)
		weave_layer1 = WeaveLayer(
		weave_layer1 = WeaveLayerFactory(
		n_atom_input_feat=self.n_atom_feat,
		n_pair_input_feat=self.n_pair_feat,
		n_atom_output_feat=self.n_hidden,
		n_pair_output_feat=self.n_hidden,
		in_layers=[combined, self.pair_split, self.atom_to_pair])
		weave_layer2 = WeaveLayer(
		in_layers=[
		self.atom_features, self.pair_features, self.pair_split,
		self.atom_to_pair
		])
		weave_layer2 = WeaveLayerFactory(
		n_atom_input_feat=self.n_hidden,
		n_pair_input_feat=self.n_hidden,
		n_atom_output_feat=self.n_hidden,
		n_pair_output_feat=self.n_hidden,
		update_pair=False,
		in_layers=[weave_layer1, self.pair_split, self.atom_to_pair])
		in_layers=[
		weave_layer1[0], weave_layer1[1], self.pair_split, self.atom_to_pair
		])
		separated = Separate_AP(in_layers=[weave_layer2])
		dense1 = Dense(
		out_channels=self.n_graph_feat,

Admin message