Merge remote-tracking branch 'remotes/mine/textcnn'

ueser · Jan 25, 2018 · a2be59a · a2be59a
2 parents 88138d3 + 184278e
commit a2be59a
Show file tree

Hide file tree

Showing 4 changed files with 22 additions and 19 deletions.
diff --git a/deepchem/models/tensorgraph/models/graph_models.py b/deepchem/models/tensorgraph/models/graph_models.py
@@ -185,14 +185,17 @@ def default_generator(self,
   def predict_on_generator(self, generator, transformers=[], outputs=None):
       out = super(WeaveTensorGraph, self).predict_on_generator(
           generator, 
-          transformers=transformers, 
+          transformers=[], 
           outputs=outputs)
       if outputs is None:
         outputs = self.outputs
-      if len(outputs) == 1:
-        return out
-      else:
-        return np.stack(out, axis=1)
+      if len(outputs) > 1:
+        out = np.stack(out, axis=1)
+
+      out = undo_transforms(out, transformers)
+      return out
+
+
 
 class DTNNTensorGraph(TensorGraph):
 
@@ -345,7 +348,6 @@ def default_generator(self,
 
         yield feed_dict
 
-  '''
   def predict(self, dataset, transformers=[], outputs=None):
     if outputs is None:
       outputs = self.outputs
@@ -357,7 +359,6 @@ def predict(self, dataset, transformers=[], outputs=None):
       return retval
     retval = np.concatenate(retval, axis=-1)
     return undo_transforms(retval, transformers)
-  '''
 
 class DAGTensorGraph(TensorGraph):
 
@@ -511,14 +512,15 @@ def default_generator(self,
   def predict_on_generator(self, generator, transformers=[], outputs=None):
       out = super(DAGTensorGraph, self).predict_on_generator(
           generator, 
-          transformers=transformers, 
+          transformers=[], 
           outputs=outputs)
       if outputs is None:
         outputs = self.outputs
-      if len(outputs) == 1:
-        return out
-      else:
-        return np.stack(out, axis=1)
+      if len(outputs) > 1:
+        out = np.stack(out, axis=1)
+
+      out = undo_transforms(out, transformers)
+      return out
 
 class PetroskiSuchTensorGraph(TensorGraph):
   """

diff --git a/deepchem/models/tensorgraph/models/text_cnn.py b/deepchem/models/tensorgraph/models/text_cnn.py
@@ -274,11 +274,12 @@ def smiles_to_seq(self, smiles):
   def predict_on_generator(self, generator, transformers=[], outputs=None):
       out = super(TextCNNTensorGraph, self).predict_on_generator(
           generator, 
-          transformers=transformers, 
+          transformers=[], 
           outputs=outputs)
       if outputs is None:
         outputs = self.outputs
-      if len(outputs) == 1:
-        return out
-      else:
-        return np.stack(out, axis=1)
+      if len(outputs) > 1:
+        out = np.stack(out, axis=1)
+
+      out = undo_transforms(out, transformers)
+      return out
diff --git a/deepchem/molnet/preset_hyper_parameters.py b/deepchem/molnet/preset_hyper_parameters.py
@@ -140,7 +140,7 @@
 hps['dtnn'] = {
     'batch_size': 64,
     'nb_epoch': 100,
-    'learning_rate': 0.0005,
+    'learning_rate': 0.001,
     'n_embedding': 50,
     'n_distance': 170,
     'seed': 123

diff --git a/deepchem/molnet/run_benchmark_models.py b/deepchem/molnet/run_benchmark_models.py
@@ -532,7 +532,7 @@ def benchmark_regression(train_dataset,
         random_seed=seed,
         output_activation=False,
         use_queue=False,
-	mode='regression')
+        mode='regression')
 
   elif model_name == 'dag_regression':
     batch_size = hyper_parameters['batch_size']