change settings for DTNN (e1ca2abc) · Commits · 钟慕尧 / deepchem

deepchem/models/init.py

+1 −1

Original line number	Diff line number	Diff line
		@@ -9,7 +9,7 @@ from deepchem.models.models import Model
		from deepchem.models.sklearn_models import SklearnModel
		from deepchem.models.xgboost_models import XGBoostModel
		from deepchem.models.tf_new_models.multitask_classifier import MultitaskGraphClassifier
		from deepchem.models.tf_new_models.multitask_regressor import MultitaskGraphRegressor
		from deepchem.models.tf_new_models.multitask_regressor import MultitaskGraphRegressor, DTNNMultitaskGraphRegressor

		from deepchem.models.tf_new_models.support_classifier import SupportGraphClassifier
		from deepchem.models.multitask import SingletaskToMultitask

deepchem/models/tf_new_models/multitask_regressor.py

+16 −0

Original line number	Diff line number	Diff line
		@@ -212,3 +212,19 @@ class MultitaskGraphRegressor(Model):
		def get_num_tasks(self):
		"""Needed to use Model.predict() from superclass."""
		return self.n_tasks


		class DTNNMultitaskGraphRegressor(MultitaskGraphRegressor):

		def build(self):
		# Create target inputs
		self.label_placeholder = tf.placeholder(
		dtype='float32', shape=(None, self.n_tasks), name="label_placeholder")
		self.weight_placeholder = tf.placeholder(
		dtype='float32', shape=(None, self.n_tasks), name="weight_placholder")

		feat = self.model.return_outputs()
		outputs = []
		for task in range(self.n_tasks):
		outputs.append(feat[:, task])
		return outputs

deepchem/nn/layers.py

+30 −30

Original line number	Diff line number	Diff line
		@@ -827,7 +827,7 @@ class DTNNEmbedding(Layer):

		def __init__(self,
		n_embedding=30,
		periodic_table_length=83,
		periodic_table_length=30,
		init='glorot_uniform',
		**kwargs):
		"""
		@@ -975,8 +975,8 @@ class DTNNGather(Layer):

		def __init__(self,
		n_embedding=30,
		n_outputs=100,
		layer_sizes=[100],
		layer_sizes=[15],
		n_tasks=1,
		init='glorot_uniform',
		activation='tanh',
		**kwargs):
		@@ -985,18 +985,18 @@ class DTNNGather(Layer):
		----------
		n_embedding: int, optional
		Number of features for each atom
		n_outputs: int, optional
		Number of features for each molecule(output)
		layer_sizes: list of int, optional(default=[1000])
		Structure of hidden layer(s)
		n_tasks: int, optional
		Number of final summed outputs
		init: str, optional
		Weight initialization for filters.
		activation: str, optional
		Activation function applied
		"""
		self.n_embedding = n_embedding
		self.n_outputs = n_outputs
		self.layer_sizes = layer_sizes
		self.n_outputs = n_tasks
		self.init = initializations.get(init) # Set weight initialization
		self.activation = activations.get(activation) # Get activations

		@@ -1016,7 +1016,6 @@ class DTNNGather(Layer):
		self.b_list.append(model_ops.zeros(shape=[
		self.n_outputs,
		]))
		prev_layer_size = self.n_outputs

		self.trainable_weights = self.W_list + self.b_list

		@@ -1037,9 +1036,10 @@ class DTNNGather(Layer):
		self.build()
		output = x[0]
		atom_membership = x[1]
		for i, W in enumerate(self.W_list):
		for i, W in enumerate(self.W_list[:-1]):
		output = tf.matmul(output, W) + self.b_list[i]
		output = self.activation(output)
		output = tf.matmul(output, self.W_list[-1]) + self.b_list[-1]
		output = tf.segment_sum(output, atom_membership)
		return output

examples/qm7/qm7_DTNN.py

+12 −8

Original line number	Diff line number	Diff line
		@@ -23,27 +23,31 @@ metric = [

		# Batch size of models
		batch_size = 50
		n_embedding = 20
		graph_model = dc.nn.SequentialDTNNGraph(n_distance=100)
		n_embedding = 30

		graph_model = dc.nn.SequentialDTNNGraph(
		n_distance=51, distance_max=9.2, distance_min=-1.)
		graph_model.add(dc.nn.DTNNEmbedding(n_embedding=n_embedding))
		graph_model.add(dc.nn.DTNNStep(n_embedding=n_embedding, n_distance=100))
		graph_model.add(dc.nn.DTNNStep(n_embedding=n_embedding, n_distance=100))
		graph_model.add(dc.nn.DTNNGather(n_embedding=n_embedding))
		graph_model.add(dc.nn.DTNNStep(n_embedding=n_embedding, n_distance=51))
		graph_model.add(dc.nn.DTNNStep(n_embedding=n_embedding, n_distance=51))
		graph_model.add(
		dc.nn.DTNNGather(
		n_embedding=n_embedding, layer_sizes=[15], n_tasks=len(tasks)))
		n_feat = n_embedding

		model = dc.models.MultitaskGraphRegressor(
		model = dc.models.DTNNMultitaskGraphRegressor(
		graph_model,
		len(tasks),
		n_feat,
		batch_size=batch_size,
		learning_rate=0.001,
		learning_rate=0.0001,
		learning_rate_decay_time=1000,
		optimizer_type="adam",
		beta1=.9,
		beta2=.999)

		# Fit trained model
		model.fit(train_dataset, nb_epoch=50)
		model.fit(train_dataset, nb_epoch=3000)

		print("Evaluating model")
		train_scores = model.evaluate(train_dataset, metric, transformers)

Admin message