modify the files

docs/api/api_python/mindspore/mindspore.SparseTensor.rst

@@ -8,7 +8,7 @@ mindspore.SparseTensor
     `SparseTensor` 只能在 `Cell` 的构造方法中使用。
 
     .. note::
-        此接口从 1.7 版本开始弃用，并计划在将来移除，请使用 `COOTensor`.
+        此接口从 1.7 版本开始弃用，并计划在将来移除，请使用 `COOTensor`。
 
     对于稠密张量，其 `SparseTensor(indices, values, shape)` 具有 `dense[indices[i]] = values[i]` 。
 

docs/api/api_python/nn/mindspore.nn.BCELoss.rst

@@ -23,7 +23,7 @@ mindspore.nn.BCELoss
         \end{cases}
 
     .. note::
-        预测值一般是sigmoid函数的输出，因为是二分类，所以目标值应是0或者1。如果输入是0或1，则上述损失函数是无意义的。
+        预测值一般是sigmoid函数的输出。因为是二分类，所以目标值应是0或者1。如果输入是0或1，则上述损失函数是无意义的。
 
     参数：
         - **weight** (Tensor, 可选) - 指定每个批次二值交叉熵的权重。与输入数据的shape和数据类型相同。默认值：None。

docs/api/api_python/nn/mindspore.nn.Cell.rst

@@ -391,7 +391,7 @@
         运行construct方法。
 
         .. note::
-            该函数已经弃用，将会在未来版本中删除，不推荐使用此函数。
+            该函数已经弃用，将会在未来版本中删除。不推荐使用此函数。
 
         参数：
             - **cast_inputs** (tuple) - Cell的输入。

docs/api/api_python/nn/mindspore.nn.OneHot.rst

@@ -8,7 +8,7 @@ mindspore.nn.OneHot
     输入的 `indices` 表示的位置取值为on_value，其他所有位置取值为off_value。
 
     .. note::
-        如果indices是n阶Tensor，那么返回的one-hot Tensor则为n+1阶Tensor。
+        如果indices是n阶Tensor，那么返回的one-hot Tensor则为n+1阶Tensor，新增 `axis` 维度。
 
     如果 `indices` 是Scalar，则输出shape将是长度为 `depth` 的向量。
 

docs/api/api_python/nn/mindspore.nn.SmoothL1Loss.rst

@@ -14,7 +14,7 @@ mindspore.nn.SmoothL1Loss
         |x_i - y_i| - 0.5 {\beta}, & \text{otherwise.}
         \end{cases}
 
-    当 `reduction` 不是设定为 `none` 时，计算如下:
+    当 `reduction` 不是设定为 `none` 时，计算如下：
 
     .. math::
         L =

docs/api/api_python/nn/mindspore.nn.optim_note_sparse.rst

@@ -1,2 +1,2 @@
-如果前向网络使用了SparseGatherV2等算子，优化器会执行稀疏运算，通过设置 `target` 为CPU，可在主机（host）上进行稀疏运算。
+如果前向网络使用了SparseGatherV2等算子，优化器会执行稀疏运算。通过设置 `target` 为CPU，可在主机（host）上进行稀疏运算。
 稀疏特性在持续开发中。

docs/api/api_python/nn/mindspore.nn.optim_note_weight_decay.rst

@@ -2,4 +2,4 @@
 
 参数分组情况下，可以分组调整权重衰减策略。
 
-分组时，每组网络参数均可配置 `weight_decay` ，若未配置，则该组网络参数使用优化器中配置的 `weight_decay` 。
+分组时，每组网络参数均可配置 `weight_decay` 。若未配置，则该组网络参数使用优化器中配置的 `weight_decay` 。

docs/api/api_python/transformer/mindspore.nn.transformer.TransformerOpParallelConfig.rst

@@ -9,7 +9,7 @@
     参数：
         - **data_parallel** (int) - 表示数据并行数。默认值：1。
         - **model_parallel** (int) - 表示模型并行数。默认值：1。
-        - **expert_parallel** (int) - 表示专家并行数，只有在应用混合专家结构（MoE，Mixture of Experts）时才会生效。默认值：1.
+        - **expert_parallel** (int) - 表示专家并行数，只有在应用混合专家结构（MoE，Mixture of Experts）时才会生效。默认值：1。
         - **pipeline_stage** (int) - 表示将Transformer切分成的stage数目。其值应为正数。默认值：1。
         - **micro_batch_num** (int) - 表示用于pipeline训练的batch的微型大小。默认值：1。
         - **optimizer_shard** (bool) - 表示是否使能优化器切分。默认值：False。

mindspore/python/mindspore/nn/layer/basic.py

@@ -44,7 +44,7 @@ class L1Regularizer(Cell):
     r"""
     Applies l1 regularization to weights.
 
-    l1 regularization makes weights sparsity
+    l1 regularization makes weights sparsity.
 
     .. math::
         \text{loss}=\lambda * \text{reduce_sum}(\text{abs}(\omega))
@@ -52,7 +52,7 @@ class L1Regularizer(Cell):
     where :math:`\lambda` is `scale` .
 
     Note:
-        scale(regularization factor) should be a number which greater than 0
+        scale(regularization factor) should be a number which greater than 0.
 
     Args:
         scale (int, float): l1 regularization factor which greater than 0.

mindspore/python/mindspore/nn/loss/loss.py

@@ -512,7 +512,7 @@ class SmoothL1Loss(LossBase):
         TypeError: If dtype of `logits` is not the same as `labels`.
         ValueError: If `beta` is less than or equal to 0.
         ValueError: If shape of `logits` is not the same as `labels`.
-        ValueError: The float64 data type of `logits` is support on Ascend platform.
+        TypeError: The float64 data type of `logits` is support on Ascend platform.
 
     Supported Platforms:
         ``Ascend`` ``GPU`` ``CPU``

mindspore/python/mindspore/nn/optim/adam.py

@@ -639,7 +639,7 @@ class AdamWeightDecay(Optimizer):
 
         If parameters are not grouped, the `weight_decay` in optimizer will be applied on the network parameters without
         'beta' or 'gamma' in their names. Users can group parameters to change the strategy of decaying weight. When
-        parameters are grouped, each group can set `weight_decay`, if not, the `weight_decay` in optimizer will be
+        parameters are grouped, each group can set `weight_decay`. If not, the `weight_decay` in optimizer will be
         applied.
 
     Args:

mindspore/python/mindspore/nn/optim/adasum.py

@@ -420,7 +420,7 @@ class AdaSumByGradWrapCell(Cell):
     Note:
         When using AdaSum, the number of traning cards needs to be a power of 2 and at least 16 cards are required.
         Currently, the optimizer sharding and pipeline parallel is not supported when using AdaSum.
-        It is recommended to using AdaSumByGradWrapCell in semi auto parallel/auto parallel mode, and in data parallel
+        It is recommended to using AdaSumByGradWrapCell in semi auto parallel/auto parallel mode. In data parallel
         mode, we recommend to using mindspore.boost to applying AdaSum.
 
     Args:
@@ -487,8 +487,8 @@ class AdaSumByDeltaWeightWrapCell(Cell):
     Note:
         When using AdaSum, the number of traning cards needs to be a power of 2 and at least 16 cards are required.
         Currently, the optimizer sharding and pipeline parallel is not supported when using AdaSum.
-        It is recommended to using AdaSumByDeltaWeightWrapCell in semi auto parallel/auto parallel mode,
-        and in data parallel mode, we recommend to using mindspore.boost to applying AdaSum.
+        It is recommended to using AdaSumByDeltaWeightWrapCell in semi auto parallel/auto parallel mode.
+        In data parallel mode, we recommend to using mindspore.boost to applying AdaSum.
 
     Args:
         optimizer (Union[Cell]): Optimizer for updating the weights. The construct function of the optimizer

mindspore/python/mindspore/nn/optim/lamb.py

@@ -132,7 +132,7 @@ class Lamb(Optimizer):
     Note:
         There is usually no connection between a optimizer and mixed precision. But when `FixedLossScaleManager` is used
         and `drop_overflow_update` in `FixedLossScaleManager` is set to False, optimizer needs to set the 'loss_scale'.
-        As this optimizer has no argument of `loss_scale`, so `loss_scale` needs to be processed by other means, refer
+        As this optimizer has no argument of `loss_scale`, so `loss_scale` needs to be processed by other means. Refer
         document `LossScale <https://www.mindspore.cn/tutorials/experts/zh-CN/master/others/mixed_precision.html>`_ to
         process `loss_scale` correctly.