remove all unnecessary labels

utils_nlp/model/gensen/train.py

@@ -183,14 +183,13 @@ def train(config, data_folder, learning_rate=0.0001):
         """Using Horovod"""
         # Horovod: scale learning rate by the number of GPUs.
         optimizer = optim.Adam(model.parameters(), lr=learning_rate * hvd.size())
-        # optimizer = optim.SGD(model.parameters(), lr=args.lr * hvd.size(), momentum=args.momentum)
 
         # Horovod: broadcast parameters & optimizer state.
         hvd.broadcast_parameters(model.state_dict(), root_rank=0)
         hvd.broadcast_optimizer_state(optimizer, root_rank=0)
 
         # Horovod: (optional) compression algorithm.
-        compression = hvd.Compression.fp16 if args.fp16_allreduce else hvd.Compression.none
+        compression = hvd.Compression.fp16
 
         # Horovod: wrap optimizer with DistributedOptimizer.
         optimizer = hvd.DistributedOptimizer(optimizer,
@@ -322,8 +321,6 @@ def train(config, data_folder, learning_rate=0.0001):
                 torch.nn.utils.clip_grad_norm(model.parameters(), 1.0)
                 optimizer.step()
 
-
-
             end = time.time()
             mbatch_times.append(end - start)
 
@@ -416,7 +413,8 @@ def train(config, data_folder, learning_rate=0.0001):
                 if break_flag == 1:
                     end_training = time.time()
                     logging.info("##### Training stopped at ##### %f" % min_val_loss)
-                    logging.info("##### Training Time ##### %f seconds" % (end_training-start_training))
+                    logging.info(
+                        "##### Training Time ##### %f seconds" % (end_training - start_training))
                     break
                 logging.info("Evaluating on NLI")
                 n_correct = 0.0
@@ -492,10 +490,7 @@ if __name__ == "__main__":
     parser.add_argument('--data_folder', type=str, help='data folder')
     # Add learning rate to tune model.
     parser.add_argument('--learning_rate', type=float, default=0.0001, help='learning rate')
-    parser.add_argument('--momentum', type=float, default=0.5, metavar='M',
-                        help='SGD momentum (default: 0.5)')
-    parser.add_argument('--fp16-allreduce', action='store_true', default=False,
-                        help='use fp16 compression during allreduce')
+
     args = parser.parse_args()
     data_folder = args.data_folder
     learning_rate = args.learning_rate
@@ -503,4 +498,4 @@ if __name__ == "__main__":
 
     config_file_path = args.config
     config = read_config(config_file_path)
-    train(config, data_folder, learning_rate)
+    train(config, data_folder, learning_rate)