Loading deepchem/models/tensorgraph/sequential.py +8 −1 Original line number Diff line number Diff line Loading @@ -12,6 +12,7 @@ from deepchem.models.tensorgraph.layers import Feature from deepchem.models.tensorgraph.layers import Label from deepchem.models.tensorgraph.layers import SoftMaxCrossEntropy from deepchem.models.tensorgraph.layers import ReduceMean from deepchem.models.tensorgraph.layers import ReduceSquareDifference class Sequential(TensorGraph): Loading Loading @@ -53,7 +54,7 @@ class Sequential(TensorGraph): dataset: dc.data.Dataset Dataset with data loss: string Only "binary_crossentropy" for now. Only "binary_crossentropy" or "mse" for now. """ X_shape, y_shape, _, _ = dataset.get_shape() # Calling fit() for first time Loading @@ -69,6 +70,8 @@ class Sequential(TensorGraph): # Add in all layers prev_layer = features if len(self._layer_list) == 0: raise ValueError("No layers have been added to model.") for ind, layer in enumerate(self._layer_list): if not len(layer.in_layers) == 0: raise ValueError("Cannot specify in_layers for Sequential.") Loading @@ -82,6 +85,10 @@ class Sequential(TensorGraph): smce = SoftMaxCrossEntropy(in_layers=[labels, prev_layer]) self._add_layer(smce) self.set_loss(ReduceMean(in_layers=[smce])) elif loss == "mse": mse = ReduceSquareDifference(in_layers=[prev_layer, labels]) self._add_layer(mse) self.loss = mse else: # TODO(rbharath): Add in support for additional losses. raise ValueError("Unsupported loss.") Loading deepchem/models/tensorgraph/tests/test_sequential.py +12 −0 Original line number Diff line number Diff line Loading @@ -36,3 +36,15 @@ class TestSequential(unittest.TestCase): # Should be able to call fit twice without failure. model.fit(dataset, loss="binary_crossentropy", nb_epoch=1000) model.fit(dataset, loss="binary_crossentropy", nb_epoch=1000) def test_single_task_regressor(self): n_data_points = 20 n_features = 2 X = np.random.rand(n_data_points, n_features) y = [0.5 for x in range(n_data_points)] dataset = dc.data.NumpyDataset(X, y) model = dc.models.Sequential(learning_rate=0.01) model.add(Dense(out_channels=1)) model.fit(dataset, loss="mse", nb_epoch=1000) prediction = np.squeeze(model.predict_on_batch(X)) assert_true(np.all(np.isclose(prediction, y, atol=3.0))) Loading
deepchem/models/tensorgraph/sequential.py +8 −1 Original line number Diff line number Diff line Loading @@ -12,6 +12,7 @@ from deepchem.models.tensorgraph.layers import Feature from deepchem.models.tensorgraph.layers import Label from deepchem.models.tensorgraph.layers import SoftMaxCrossEntropy from deepchem.models.tensorgraph.layers import ReduceMean from deepchem.models.tensorgraph.layers import ReduceSquareDifference class Sequential(TensorGraph): Loading Loading @@ -53,7 +54,7 @@ class Sequential(TensorGraph): dataset: dc.data.Dataset Dataset with data loss: string Only "binary_crossentropy" for now. Only "binary_crossentropy" or "mse" for now. """ X_shape, y_shape, _, _ = dataset.get_shape() # Calling fit() for first time Loading @@ -69,6 +70,8 @@ class Sequential(TensorGraph): # Add in all layers prev_layer = features if len(self._layer_list) == 0: raise ValueError("No layers have been added to model.") for ind, layer in enumerate(self._layer_list): if not len(layer.in_layers) == 0: raise ValueError("Cannot specify in_layers for Sequential.") Loading @@ -82,6 +85,10 @@ class Sequential(TensorGraph): smce = SoftMaxCrossEntropy(in_layers=[labels, prev_layer]) self._add_layer(smce) self.set_loss(ReduceMean(in_layers=[smce])) elif loss == "mse": mse = ReduceSquareDifference(in_layers=[prev_layer, labels]) self._add_layer(mse) self.loss = mse else: # TODO(rbharath): Add in support for additional losses. raise ValueError("Unsupported loss.") Loading
deepchem/models/tensorgraph/tests/test_sequential.py +12 −0 Original line number Diff line number Diff line Loading @@ -36,3 +36,15 @@ class TestSequential(unittest.TestCase): # Should be able to call fit twice without failure. model.fit(dataset, loss="binary_crossentropy", nb_epoch=1000) model.fit(dataset, loss="binary_crossentropy", nb_epoch=1000) def test_single_task_regressor(self): n_data_points = 20 n_features = 2 X = np.random.rand(n_data_points, n_features) y = [0.5 for x in range(n_data_points)] dataset = dc.data.NumpyDataset(X, y) model = dc.models.Sequential(learning_rate=0.01) model.add(Dense(out_channels=1)) model.fit(dataset, loss="mse", nb_epoch=1000) prediction = np.squeeze(model.predict_on_batch(X)) assert_true(np.all(np.isclose(prediction, y, atol=3.0)))
