remove duplicate accumulate gradient step arguments

run_classifier.py

@@ -392,10 +392,6 @@ def main():
                         default=False,
                         action='store_true',
                         help="Whether not to use CUDA when available")
-    parser.add_argument("--accumulate_gradients",
-                        type=int,
-                        default=1,
-                        help="Number of steps to accumulate gradient on (divide the batch_size and accumulate)")
     parser.add_argument("--local_rank",
                         type=int,
                         default=-1,
@@ -426,11 +422,11 @@ def main():
         torch.distributed.init_process_group(backend='nccl')
     logger.info("device %s n_gpu %d distributed training %r", device, n_gpu, bool(args.local_rank != -1))
 
-    if args.accumulate_gradients < 1:
-        raise ValueError("Invalid accumulate_gradients parameter: {}, should be >= 1".format(
-                            args.accumulate_gradients))
+    if args.gradient_accumulation_steps < 1:
+        raise ValueError("Invalid gradient_accumulation_steps parameter: {}, should be >= 1".format(
+                            args.gradient_accumulation_steps))
 
-    args.train_batch_size = int(args.train_batch_size / args.accumulate_gradients)
+    args.train_batch_size = int(args.train_batch_size / args.gradient_accumulation_steps)
 
     random.seed(args.seed)
     np.random.seed(args.seed)

run_squad.py

@@ -731,10 +731,6 @@ def main():
                         type=int,
                         default=-1,
                         help="local_rank for distributed training on gpus")
-    parser.add_argument("--accumulate_gradients",
-                        type=int,
-                        default=1,
-                        help="Number of steps to accumulate gradient on (divide the batch_size and accumulate)")
     parser.add_argument('--seed', 
                         type=int, 
                         default=42,
@@ -756,11 +752,11 @@ def main():
         torch.distributed.init_process_group(backend='nccl')
     logger.info("device %s n_gpu %d distributed training %r", device, n_gpu, bool(args.local_rank != -1))
 
-    if args.accumulate_gradients < 1:
-        raise ValueError("Invalid accumulate_gradients parameter: {}, should be >= 1".format(
-                            args.accumulate_gradients))
+    if args.gradient_accumulation_steps < 1:
+        raise ValueError("Invalid gradient_accumulation_steps parameter: {}, should be >= 1".format(
+                            args.gradient_accumulation_steps))
 
-    args.train_batch_size = int(args.train_batch_size / args.accumulate_gradients)
+    args.train_batch_size = int(args.train_batch_size / args.gradient_accumulation_steps)
 
     random.seed(args.seed)
     np.random.seed(args.seed)