support fetch feed in dygraph to static graph (#22767)

* Support fetch and run program in the process of dygraph_to_static_output. test=develop

* fix to_source(gast) and remove dygraph API such as Conv2D, Linear. test=develop

python/paddle/fluid/dygraph/dygraph_to_static/ast_transformer.py

@@ -13,7 +13,6 @@
 # limitations under the License.
 
 from __future__ import print_function
-import astor
 from .utils import *
 import gast
 # gast is a generic AST to represent Python2 and Python3's Abstract Syntax Tree(AST).
@@ -21,7 +20,7 @@ import gast
 # as produced by ast.parse from the standard ast module.
 # See details in https://github.com/serge-sans-paille/gast/
 from .ast_utils import is_control_flow_if, create_cond_node, transform_if_else
-
+from paddle.fluid import unique_name
 from .static_analysis import AstNodeWrapper, StaticAnalysisVisitor
 
 __all__ = ['DygraphToStaticAst']
@@ -109,7 +108,9 @@ class DygraphToStaticAst(gast.NodeTransformer):
         self.visit(node.node)
 
         # Transform basic api of dygraph to static graph
-        BasicApiTransformer(node).ast_visit()
+        basic_api_trans = BasicApiTransformer(node)
+        basic_api_trans.ast_visit()
+        self.feed_name_to_arg_name = basic_api_trans.get_feed_name_to_arg_id()
 
         # Transform all if/else statement of Dygraph into Static Graph.
         IfElseTransformer(node).ast_visit()
@@ -117,6 +118,11 @@ class DygraphToStaticAst(gast.NodeTransformer):
     def visit_FunctionDef(self, node):
         if self.decorate_func_name is None:
             self.decorate_func_name = node.name
+
+        self.arg_name_to_idx = {}
+        for idx, arg in enumerate(node.args.args):
+            self.arg_name_to_idx[arg.id] = idx
+
         self.generic_visit(node)
         # Remove the decorated name of dygraph_to_static
         if hasattr(node, 'decorator_list'):
@@ -135,6 +141,12 @@ class DygraphToStaticAst(gast.NodeTransformer):
         assert self.decorate_func_name, "decorate_func_name shall not be None."
         return self.decorate_func_name
 
+    def get_feed_name_to_idx(self):
+        feed_name_to_idx = {}
+        for feed_name, arg_name in self.feed_name_to_arg_name.items():
+            feed_name_to_idx[feed_name] = self.arg_name_to_idx.get(arg_name)
+        return feed_name_to_idx
+
 
 class BasicApiTransformer(gast.NodeTransformer):
     """
@@ -148,6 +160,7 @@ class BasicApiTransformer(gast.NodeTransformer):
         self.wrapper_root = wrapper_root
         self.root = wrapper_root.node
         self.class_node_dict = {}
+        self.feed_name_to_arg_id = {}
 
     def ast_visit(self):
         self.visit(self.root)
@@ -189,10 +202,11 @@ class BasicApiTransformer(gast.NodeTransformer):
 
         # Replace API `to_variable` with `fluid.layers.assign`
         if is_to_variable(node):
+            self._update_feed_dict(node)
             node = to_assign_node(node)
             return node
 
-        func_name = astor.to_source(node.func)
+        func_name = astor.to_source(gast.gast_to_ast(node.func))
         if self._is_dygraph_forward(func_name):
             class_node = self._get_class_node(func_name)
             static_node = to_static_ast(node, class_node)
@@ -214,9 +228,29 @@ class BasicApiTransformer(gast.NodeTransformer):
                 return False
 
             if is_dygraph_api(node_value):
+                dygraph_api = node_value.func.attr
+                if not dygraph_class_to_static_api.get(dygraph_api):
+                    return False
+
                 update_args_of_func(node_value, node_value, "__init__")
-                target_str = astor.to_source(node.targets[0])
+                target_str = astor.to_source(gast.gast_to_ast(node.targets[0]))
                 self.class_node_dict[target_str] = node_value
                 return True
             # TODO: node.value is not dygraph class
         return False
+
+    def _update_feed_dict(self, node):
+        assert isinstance(node, gast.Call)
+
+        var_name = None
+        for kw in node.keywords:
+            if kw.arg == 'value':
+                var_name = kw.value.id  # eg: 'a' for "value=a "
+        if not var_name:
+            var_name = node.args[0].id
+
+        feed_var_name = unique_name.generate(var_name)  # eg: "a_0"
+        self.feed_name_to_arg_id[feed_var_name] = var_name  # eg: "a_0" : "a"
+
+    def get_feed_name_to_arg_id(self):
+        return self.feed_name_to_arg_id

python/paddle/fluid/dygraph/dygraph_to_static/static_analysis.py

@@ -19,55 +19,11 @@ import gast
 import inspect
 import six
 import warnings
+from .utils import is_paddle_api, is_dygraph_api, is_numpy_api
 
 __all__ = ['AstNodeWrapper', 'NodeVarType', 'StaticAnalysisVisitor']
 
 
-# TODO: _is_paddle_dygraph_api is duplicated in Yamei's utils.py. Merge the two
-# function code together when Yamei finish her PR.
-def _is_api_in_module_helper(obj, module_prefix):
-    m = inspect.getmodule(obj)
-    return m is not None and m.__name__.startswith(module_prefix)
-
-
-# TODO: is_dygraph_api is duplicated in Yamei's utils.py. Merge the two
-# function code together when Yamei finish her PR.
-def is_api_in_module(node, module_prefix):
-    assert isinstance(node, gast.Call), "Input non-Call node for is_dygraph_api"
-    func_str = astor.to_source(gast.gast_to_ast(node.func))
-    try:
-        import paddle.fluid as fluid
-        import paddle
-        return eval("_is_api_in_module_helper({}, '{}')".format(func_str,
-                                                                module_prefix))
-    except NameError:
-        return False
-
-
-def is_dygraph_api(node):
-    return is_api_in_module(node, "paddle.fluid.dygraph")
-
-
-def is_paddle_api(node):
-    return is_api_in_module(node, "paddle.fluid")
-
-
-# Is numpy_api cannot reuse is_api_in_module because of numpy module problem
-def is_numpy_api(node):
-    assert isinstance(node, gast.Call), "Input non-Call node for is_numpy_api"
-    func_str = astor.to_source(gast.gast_to_ast(node.func))
-    try:
-        import numpy as np
-        module_result = eval("_is_api_in_module_helper({}, '{}')".format(
-            func_str, "numpy"))
-        # BUG: np.random.uniform doesn't have module and cannot be analyzed
-        # TODO: find a better way
-        if not module_result:
-            return func_str.startswith("numpy.") or func_str.startswith("np.")
-    except NameError:
-        return False
-
-
 class NodeVarType(object):
     """
     Enum class of python variable types. We have to know some variable types

python/paddle/fluid/dygraph/dygraph_to_static/utils.py

@@ -17,41 +17,62 @@ from __future__ import print_function
 import inspect
 import gast
 import astor
-import atexit
-import os
-import tempfile
-import six
-import imp
 
 dygraph_class_to_static_api = {
-    "BatchNorm": "batch_norm",
-    "BilinearTensorProduct": "bilinear_tensor_product",
-    "Conv2D": "conv2d",
-    "Conv3D": "conv3d",
-    "Conv2DTranspose": "conv2d_transpose",
-    "Conv3DTranspose": "conv3d_transpose",
     "CosineDecay": "cosine_decay",
-    "Embedding": "embedding",
     "ExponentialDecay": "exponential_decay",
-    "GroupNorm": "group_norm",
-    "GRUUnit": "gru_unit",
     "InverseTimeDecay": "inverse_time_decay",
-    "LayerNorm": "layer_norm",
-    "Linear": "fc",
     "NaturalExpDecay": "natural_exp_decay",
-    "NCE": "nce",
     "NoamDecay": "noam_decay",
     "PiecewiseDecay": "piecewise_decay",
     "PolynomialDecay": "polynomial_decay",
-    "Pool2D": "pool2d",
-    "PRelu": "prelu",
-    "SpectralNorm": "spectral_norm",
 }
 
 
+def _is_api_in_module_helper(obj, module_prefix):
+    m = inspect.getmodule(obj)
+    return m is not None and m.__name__.startswith(module_prefix)
+
+
+def is_api_in_module(node, module_prefix):
+    assert isinstance(node, gast.Call), "Input non-Call node for is_dygraph_api"
+    func_str = astor.to_source(gast.gast_to_ast(node.func))
+    try:
+        import paddle.fluid as fluid
+        import paddle
+        return eval("_is_api_in_module_helper({}, '{}')".format(func_str,
+                                                                module_prefix))
+    except NameError:
+        return False
+
+
+def is_dygraph_api(node):
+    return is_api_in_module(node, "paddle.fluid.dygraph")
+
+
+def is_paddle_api(node):
+    return is_api_in_module(node, "paddle.fluid")
+
+
+# Is numpy_api cannot reuse is_api_in_module because of numpy module problem
+def is_numpy_api(node):
+    assert isinstance(node, gast.Call), "Input non-Call node for is_numpy_api"
+    func_str = astor.to_source(gast.gast_to_ast(node.func))
+    try:
+        import numpy as np
+        module_result = eval("_is_api_in_module_helper({}, '{}')".format(
+            func_str, "numpy"))
+        # BUG: np.random.uniform doesn't have module and cannot be analyzed
+        # TODO: find a better way
+        if not module_result:
+            return func_str.startswith("numpy.") or func_str.startswith("np.")
+    except NameError:
+        return False
+
+
 def _delete_keywords_from(node):
     assert isinstance(node, gast.Call)
-    func_src = astor.to_source(node.func)
+    func_src = astor.to_source(gast.gast_to_ast(node.func))
     import paddle.fluid as fluid
     full_args = eval("inspect.getargspec({})".format(func_src))
     full_args_name = full_args[0]
@@ -94,21 +115,6 @@ def _add_keywords_to(node, dygraph_api_name):
     return
 
 
-def _is_paddle_dygraph_api(obj):
-    m = inspect.getmodule(obj)
-    return m is not None and m.__name__.startswith("paddle.fluid.dygraph")
-
-
-def is_dygraph_api(node):
-    assert isinstance(node, gast.Call)
-    func_src = astor.to_source(node.func)
-    try:
-        import paddle.fluid as fluid
-        return eval("_is_paddle_dygraph_api({})".format(func_src))
-    except NameError:
-        return False
-
-
 def is_to_variable(node):
     assert isinstance(node, gast.Call)
     if is_dygraph_api(node):
@@ -158,7 +164,7 @@ def update_args_of_func(node, dygraph_node, method_name):
             "The method name of class to update args should be '__init__' or 'forward'"
         )
 
-    class_src = astor.to_source(dygraph_node.func)
+    class_src = astor.to_source(gast.gast_to_ast(dygraph_node.func))
     import paddle.fluid as fluid
     if method_name == "__init__" or eval(
             "issubclass({}, fluid.dygraph.Layer)".format(class_src)):

python/paddle/fluid/dygraph/jit.py

@@ -29,6 +29,7 @@ from paddle.fluid import core
 from paddle.fluid.framework import Program, Block, Variable, _dygraph_tracer, dygraph_only, _dygraph_guard, _current_expected_place, in_dygraph_mode
 from paddle.fluid.executor import Executor, scope_guard
 from paddle.fluid.compiler import CompiledProgram
+from paddle.fluid import program_guard, data
 
 
 def create_program_from_desc(program_desc):
@@ -56,6 +57,7 @@ def extract_vars(inputs):
 
 def _dygraph_to_static_output_(dygraph_func):
     def __impl__(*args, **kwargs):
+
         # Get AST from dygraph function
         dygraph_code = inspect.getsource(dygraph_func)
         dygraph_code = textwrap.dedent(dygraph_code)
@@ -64,14 +66,53 @@ def _dygraph_to_static_output_(dygraph_func):
         # Transform AST
         dygraph_to_static = DygraphToStaticAst()
         root_wrapper = dygraph_to_static.get_static_ast(root)
+
+        # Get static_func from AST
         func_name = dygraph_to_static.get_module_name()
         static_func, file_name = ast_to_func(root_wrapper.node, func_name)
 
-        return static_func(*args, **kwargs)
+        if not in_dygraph_mode():
+            return static_func(*args, **kwargs)
+        else:
+            feed_name_to_idx = dygraph_to_static.get_feed_name_to_idx()
+            feed_dict = {}
+            for feed_name, idx in feed_name_to_idx.items():
+                feed_dict[feed_name] = args[idx]
+
+            # Run static_func in static mode
+            startup_program = Program()
+            main_program = Program()
+            static_res = run_static_func(main_program, startup_program,
+                                         static_func, args, kwargs, feed_dict,
+                                         feed_name_to_idx)
+
+        return static_res
 
     return __impl__
 
 
+@switch_to_static_graph
+def run_static_func(main_program, startup_program, static_func, args, kwargs,
+                    feed_dict, feed_name_to_idx):
+
+    with program_guard(main_program, startup_program):
+        args_list = list(args)
+        for var_name, value in feed_dict.items():
+            idx = feed_name_to_idx[var_name]
+            args_list[idx] = data(
+                name=var_name, shape=value.shape, dtype=str(value.dtype))
+        args = tuple(args_list)
+        static_out = static_func(*args, **kwargs)
+        if not isinstance(static_out, (list, tuple)):
+            static_out = [static_out]
+        exe = Executor(core.CPUPlace())
+        exe.run(startup_program)
+        static_res = exe.run(main_program,
+                             fetch_list=static_out,
+                             feed=feed_dict)
+    return static_res
+
+
 dygraph_to_static_output = wrap_decorator(_dygraph_to_static_output_)
 
 

python/paddle/fluid/tests/unittests/CMakeLists.txt

@@ -337,6 +337,7 @@ set_tests_properties(test_parallel_executor_seresnext_with_reduce_cpu PROPERTIES
 set_tests_properties(test_parallel_executor_seresnext_with_fuse_all_reduce_cpu PROPERTIES TIMEOUT 750)
 
 add_subdirectory(sequence)
+add_subdirectory(dygraph_to_static)
 
 if (WITH_NGRAPH)
     add_subdirectory(ngraph)

python/paddle/fluid/tests/unittests/dygraph_to_static/CMakeLists.txt

@@ -0,0 +1,6 @@
+file(GLOB TEST_OPS RELATIVE "${CMAKE_CURRENT_SOURCE_DIR}" "test_*.py")
+string(REPLACE ".py" "" TEST_OPS "${TEST_OPS}")
+
+foreach(TEST_OP ${TEST_OPS})
+    py_test_modules(${TEST_OP} MODULES ${TEST_OP})
+endforeach(TEST_OP)

python/paddle/fluid/tests/unittests/dygraph_to_static/test_fetch_feed.py

@@ -0,0 +1,100 @@
+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from __future__ import print_function
+
+from paddle.fluid.dygraph.jit import dygraph_to_static_output
+
+import numpy as np
+import unittest
+
+import paddle.fluid as fluid
+
+SEED = 2020
+
+
+class Pool2D(fluid.dygraph.Layer):
+    def __init__(self):
+        super(Pool2D, self).__init__()
+        self.pool2d = fluid.dygraph.Pool2D(
+            pool_size=2, pool_type='avg', pool_stride=1, global_pooling=False)
+
+    @dygraph_to_static_output
+    def forward(self, x):
+        inputs = fluid.dygraph.to_variable(x)
+        pre = self.pool2d(inputs)
+        return pre
+
+
+class Linear(fluid.dygraph.Layer):
+    def __init__(self):
+        super(Linear, self).__init__()
+
+    @dygraph_to_static_output
+    def forward(self, x):
+        fc = fluid.dygraph.Linear(
+            input_dim=10,
+            output_dim=5,
+            act='relu',
+            param_attr=fluid.ParamAttr(initializer=fluid.initializer.Constant(
+                value=0.99)),
+            bias_attr=fluid.ParamAttr(initializer=fluid.initializer.Constant(
+                value=0.5)))
+        inputs = fluid.dygraph.to_variable(x)
+        pre = fc(inputs)
+        return pre
+
+
+class TestPool2D(unittest.TestCase):
+    def setUp(self):
+        self.dygraph_class = Pool2D
+        self.data = np.random.random((1, 2, 4, 4)).astype('float32')
+
+    def run_dygraph_mode(self):
+        with fluid.dygraph.guard():
+            dy_layer = self.dygraph_class()
+            for _ in range(1):
+
+                prediction = dy_layer(x=self.data)
+                return prediction
+
+    def run_static_mode(self):
+        startup_prog = fluid.Program()
+        main_prog = fluid.Program()
+        with fluid.program_guard(main_prog, startup_prog):
+            dy_layer = self.dygraph_class()
+            out = dy_layer(x=self.data)
+            place = fluid.CPUPlace()
+            exe = fluid.Executor(place)
+            res = exe.run(main_prog, fetch_list=out)
+            return res
+
+    def test_static_output(self):
+        dygraph_res = self.run_dygraph_mode()
+        static_res = self.run_static_mode()
+        self.assertTrue(
+            np.allclose(dygraph_res[0], static_res[0]),
+            msg='dygraph is {}\n static_res is \n{}'.format(dygraph_res,
+                                                            static_res))
+        return
+
+
+class TestLinear(unittest.TestCase):
+    def setUp(self):
+        self.dygraph_class = Linear
+        self.data = np.random.random((4, 10)).astype('float32')
+
+
+if __name__ == '__main__':
+    unittest.main()