updata 2.0 API english doc (#28525)

* make Numpy version is below 1.19.3

* fix 2.0 doc

paddle/fluid/operators/unique_op.cu

@@ -177,7 +177,7 @@ static void UniqueFlattendCUDATensor(const framework::ExecutionContext& context,
   thrust::sort_by_key(thrust::device, in_data_hat, in_data_hat + num_input,
                       sorted_indices_data);
 
-  // 1. Calculate op result: 'out'：
+  // 1. Calculate op result: 'out'
   Tensor range;
   range.Resize(framework::make_ddim({num_input + 1}));
   auto range_data_ptr = range.mutable_data<IndexT>(context.GetPlace());

paddle/fluid/pybind/imperative.cc

@@ -685,8 +685,6 @@ void BindImperative(py::module *m_ptr) {
             .. code-block:: python
 
                 import paddle
-                paddle.disable_static()
-
                 linear = Linear(32, 64)
                 data = paddle.uniform(shape=[30, 10, 32], -1, 1)
                 x = linear(data)
@@ -704,19 +702,13 @@ void BindImperative(py::module *m_ptr) {
              .. code-block:: python
 
                 import paddle
-                paddle.disable_static()
-
-                inputs = []
-                for _ in range(10):
-                    tmp = paddle.ones([2, 2])
-                    tmp.stop_gradient=False
-                    inputs.append(tmp)
-                ret = paddle.sums(inputs2)
-                loss = paddle.sum(ret)
-                loss.backward()
-                print("Before clear_gradient {}".format(loss.grad))
-                loss.clear_gradient()
-                print("After clear_gradient {}".format(loss.grad))
+                input = paddle.uniform([10, 2])
+                linear = paddle.nn.Linear(2, 3)
+                out = linear(input)
+                out.backward()
+                print("Before clear_gradient, linear.weight.grad: {}".format(linear.weight.grad))
+                linear.weight.clear_gradient()
+                print("After clear_gradient, linear.weight.grad: {}".format(linear.weight.grad))
       )DOC")
       .def("clone",
            [](std::shared_ptr<imperative::VarBase> &self) {

python/paddle/framework/__init__.py

@@ -18,10 +18,7 @@ __all__ = [
     'get_default_dtype', 'set_default_dtype'
 ]
 
-__all__ += [
-    'grad', 'LayerList', 'load', 'save', 'to_variable', 'no_grad',
-    'DataParallel'
-]
+__all__ += ['grad', 'LayerList', 'load', 'save', 'no_grad', 'DataParallel']
 
 from . import random
 from .random import seed
@@ -39,7 +36,6 @@ from ..fluid.core import VarBase  #DEFINE_ALIAS
 
 from paddle.fluid import core  #DEFINE_ALIAS
 from ..fluid.dygraph.base import no_grad_ as no_grad  #DEFINE_ALIAS
-from ..fluid.dygraph.base import to_variable  #DEFINE_ALIAS
 from ..fluid.dygraph.base import grad  #DEFINE_ALIAS
 from .io import save
 from .io import load

python/paddle/framework/io.py

@@ -225,8 +225,6 @@ def save(obj, path):
 
             import paddle
 
-            paddle.disable_static()
-
             emb = paddle.nn.Embedding(10, 10)
             layer_state_dict = emb.state_dict()
             paddle.save(layer_state_dict, "emb.pdparams")
@@ -318,8 +316,6 @@ def load(path, **configs):
         .. code-block:: python
 
             import paddle
-            
-            paddle.disable_static()
 
             emb = paddle.nn.Embedding(10, 10)
             layer_state_dict = emb.state_dict()