fix an bug in metrics.py; test=document_fix (#42976)

PR types
Bug fixes

PR changes
Docs

Describe
修复 paddle.metric.accuracy 文档，对应的中文文档修复为 https://github.com/PaddlePaddle/docs/pull/4811
the file was editted based on the discussion in the issue:
INT32 Failed on paddle.metric.accuracy: https://github.com/PaddlePaddle/Paddle/issues/42845

python/paddle/metric/metrics.py

@@ -771,7 +771,7 @@ def accuracy(input, label, k=1, correct=None, total=None, name=None):
     Args:
         input(Tensor): The input of accuracy layer, which is the predictions of network. A Tensor with type float32,float64.
             The shape is ``[sample_number, class_dim]`` .
-        label(Tensor): The label of dataset. Tensor with type int32,int64. The shape is ``[sample_number, 1]`` .
+        label(Tensor): The label of dataset. Tensor with type int64. The shape is ``[sample_number, 1]`` .
         k(int, optional): The top k predictions for each class will be checked. Data type is int64 or int32.
         correct(Tensor, optional): The correct predictions count. A Tensor with type int64 or int32.
         total(Tensor, optional): The total entries count. A tensor with type int64 or int32.