[Example] Fix GCMC (#2067)

* Fix

* Update data.py

* Update README.md

* Fix
Co-authored-by: Ubuntu <ubuntu@ip-172-31-1-5.us-west-2.compute.internal>

examples/mxnet/_deprecated/gcmc/data.py

@@ -293,7 +293,8 @@ class MovieLens(object):
             (ones, rating_pairs),
             shape=(self.num_user, self.num_movie), dtype=np.float32)
         g = dgl.bipartite_from_scipy(user_movie_ratings_coo, utype='_U', etype='_E', vtype='_V')
-        return dgl.heterograph({('user', 'rate', 'movie'): g.edges()})
+        return dgl.heterograph({('user', 'rate', 'movie'): g.edges()}, 
+                               num_nodes_dict={'user': self.num_user, 'movie': self.num_movie})
 
     @property
     def num_links(self):

examples/pytorch/gcmc/README.md

@@ -12,6 +12,8 @@ Credit: Jiani Zhang ([@jennyzhang0215](https://github.com/jennyzhang0215))
 * PyTorch 1.2+
 * pandas
 * torchtext 0.4+ (if using user and item contents as node features)
+* spacy (if using user and item contents as node features)
+    - You will also need to run `python -m spacy download en_core_web_sm`
 
 ## Data
 

examples/pytorch/gcmc/data.py

@@ -302,7 +302,8 @@ class MovieLens(object):
             (ones, rating_pairs),
             shape=(self.num_user, self.num_movie), dtype=np.float32)
         g = dgl.bipartite_from_scipy(user_movie_ratings_coo, utype='_U', etype='_E', vtype='_V')
-        return dgl.heterograph({('user', 'rate', 'movie'): g.edges()})
+        return dgl.heterograph({('user', 'rate', 'movie'): g.edges()}, 
+                               num_nodes_dict={'user': self.num_user, 'movie': self.num_movie})
 
     @property
     def num_links(self):

examples/pytorch/gcmc/train_sampling.py

@@ -121,6 +121,7 @@ def evaluate(args, dev_id, net, dataset, dataloader, segment='valid'):
         frontier = frontier.to(dev_id)
         head_feat = head_feat.to(dev_id)
         tail_feat = tail_feat.to(dev_id)
+        pair_graph = pair_graph.to(dev_id)
         with th.no_grad():
             pred_ratings = net(pair_graph, frontier,
                                head_feat, tail_feat, possible_rating_values)
@@ -335,7 +336,7 @@ def run(proc_id, n_gpus, args, devices, dataset):
                                       dataset=dataset,
                                       dataloader=valid_dataloader,
                                       segment='valid')
-                logging_str += ',\tVal RMSE={:.4f}'.format(valid_rmse)
+                logging_str = 'Val RMSE={:.4f}'.format(valid_rmse)
 
                 if valid_rmse < best_valid_rmse:
                     best_valid_rmse = valid_rmse