Merge branch 'master' of https://github.com/deepchem/deepchem into vina_tf (a14ff053) · Commits · 钟慕尧 / deepchem

README.md

+9 −4

Original line number	Diff line number	Diff line
		@@ -71,7 +71,12 @@ Installation from source is the only currently supported format. ```deepchem```
		conda install -c omnia mdtraj
		```

		7. `tensorflow`: Installing `tensorflow` on older versions of Linux (which
		7. `pdbfixer`
		```bash
		conda install -c omnia pdbfixer=1.4
		```

		8. `tensorflow`: Installing `tensorflow` on older versions of Linux (which
		have glibc < 2.17) can be very challenging. For these older Linux versions,
		contact your local sysadmin to work out a custom installation. If your
		version of Linux is recent, then the following command will work:
		@@ -79,7 +84,7 @@ Installation from source is the only currently supported format. ```deepchem```
		pip install tensorflow-gpu==0.12.1
		```

		8. `deepchem`: Clone the `deepchem` github repo:
		9. `deepchem`: Clone the `deepchem` github repo:
		```bash
		git clone https://github.com/deepchem/deepchem.git
		```
		@@ -88,7 +93,7 @@ Installation from source is the only currently supported format. ```deepchem```
		python setup.py install
		```

		9. To run test suite, install `nosetests`:
		10. To run test suite, install `nosetests`:
		```bash
		pip install nose
		```
		@@ -99,7 +104,7 @@ Installation from source is the only currently supported format. ```deepchem```
		You might need to uninstall a system install of `nosetests` if
		there is a conflict.

		10. If installation has been successful, all tests in test suite should pass:
		11. If installation has been successful, all tests in test suite should pass:
		```bash
		nosetests -v deepchem --nologcapture
		```

examples/benchmark.py

+82 −78

Original line number	Diff line number	Diff line
		@@ -495,25 +495,27 @@ def benchmark_classification(train_dataset,
		sess = tf.Session(graph=g)
		K.set_session(sess)
		# Building graph convolution model
		with g.as_default():
		tf.set_random_seed(seed)
		graph_model = dc.nn.SequentialGraph(n_features)
		graph_model.add(dc.nn.GraphConv(int(n_filters), activation='relu'))
		graph_model.add(
		dc.nn.GraphConv(int(n_filters), n_features, activation='relu'))
		graph_model.add(dc.nn.BatchNormalization(epsilon=1e-5, mode=1))
		graph_model.add(dc.nn.GraphPool())
		graph_model.add(dc.nn.GraphConv(int(n_filters), activation='relu'))
		graph_model.add(
		dc.nn.GraphConv(int(n_filters), int(n_filters), activation='relu'))
		graph_model.add(dc.nn.BatchNormalization(epsilon=1e-5, mode=1))
		graph_model.add(dc.nn.GraphPool())
		# Gather Projection
		graph_model.add(
		dc.nn.Dense(int(n_fully_connected_nodes), activation='relu'))
		dc.nn.Dense(
		int(n_fully_connected_nodes), int(n_filters), activation='relu'))
		graph_model.add(dc.nn.BatchNormalization(epsilon=1e-5, mode=1))
		graph_model.add(dc.nn.GraphGather(batch_size, activation="tanh"))
		with tf.Session() as sess:
		model_graphconv = dc.models.MultitaskGraphClassifier(
		sess,
		graph_model,
		len(tasks),
		n_features,
		batch_size=batch_size,
		learning_rate=learning_rate,
		optimizer_type="adam",
		@@ -693,25 +695,27 @@ def benchmark_regression(train_dataset,
		sess = tf.Session(graph=g)
		K.set_session(sess)
		# Building graph convoluwtion model
		with g.as_default():
		tf.set_random_seed(seed)
		graph_model = dc.nn.SequentialGraph(n_features)
		graph_model.add(dc.nn.GraphConv(int(n_filters), activation='relu'))
		graph_model.add(
		dc.nn.GraphConv(int(n_filters), n_features, activation='relu'))
		graph_model.add(dc.nn.BatchNormalization(epsilon=1e-5, mode=1))
		graph_model.add(dc.nn.GraphPool())
		graph_model.add(dc.nn.GraphConv(int(n_filters), activation='relu'))
		graph_model.add(
		dc.nn.GraphConv(int(n_filters), int(n_filters), activation='relu'))
		graph_model.add(dc.nn.BatchNormalization(epsilon=1e-5, mode=1))
		graph_model.add(dc.nn.GraphPool())
		# Gather Projection
		graph_model.add(
		dc.nn.Dense(int(n_fully_connected_nodes), activation='relu'))
		dc.nn.Dense(
		int(n_fully_connected_nodes), int(n_filters), activation='relu'))
		graph_model.add(dc.nn.BatchNormalization(epsilon=1e-5, mode=1))
		graph_model.add(dc.nn.GraphGather(batch_size, activation="tanh"))
		with tf.Session() as sess:
		model_graphconvreg = dc.models.MultitaskGraphRegressor(
		sess,
		graph_model,
		len(tasks),
		n_features,
		batch_size=batch_size,
		learning_rate=learning_rate,
		optimizer_type="adam",

Admin message