Fix syntax

megatron/optimizer/clip_grads.py

@@ -129,20 +129,15 @@ def count_zeros_fp32(parameters):
     #   - grad should not be none
     #   - parameter should not be shared
     #   - should not be a replica due to tensor model parallelism
-    grads_to_count = []
+    total_num_zeros = 0.0
     for param in parameters:
         grad_not_none = param.grad is not None
         is_not_shared = param_is_not_shared(param)
         is_not_tp_duplicate = param_is_not_tensor_parallel_duplicate(param)
         if grad_not_none and is_not_shared and is_not_tp_duplicate:
             grad = param.grad.detach()
-            grads_to_count.append(grad)
-
-    total_num_zeros = 0.0
-
-    for grad in grads_to_count:
-        num_zeros = grad.numel() - torch.count_nonzero(grad)
-        total_num_zeros = num_zeros + total_num_zeros
+            num_zeros = grad.numel() - torch.count_nonzero(grad)
+            total_num_zeros = num_zeros + total_num_zeros
 
     # Sum across all model-parallel GPUs.
     torch.distributed.all_reduce(total_num_zeros,