• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 /* Copyright 2016 The TensorFlow Authors. All Rights Reserved.
2 
3 Licensed under the Apache License, Version 2.0 (the "License");
4 you may not use this file except in compliance with the License.
5 You may obtain a copy of the License at
6 
7     http://www.apache.org/licenses/LICENSE-2.0
8 
9 Unless required by applicable law or agreed to in writing, software
10 distributed under the License is distributed on an "AS IS" BASIS,
11 WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12 See the License for the specific language governing permissions and
13 limitations under the License.
14 ==============================================================================*/
15 
16 #ifndef TENSORFLOW_CC_CLIENT_CLIENT_SESSION_H_
17 #define TENSORFLOW_CC_CLIENT_CLIENT_SESSION_H_
18 
19 #include <memory>
20 #include <string>
21 #include <unordered_map>
22 #include <vector>
23 
24 #include "tensorflow/cc/framework/ops.h"
25 #include "tensorflow/cc/framework/scope.h"
26 #include "tensorflow/core/public/session_options.h"
27 
28 namespace tensorflow {
29 
30 /// @addtogroup core
31 /// @{
32 
33 /// A `ClientSession` object lets the caller drive the evaluation of the
34 /// TensorFlow graph constructed with the C++ API.
35 ///
36 /// Example:
37 ///
38 ///     Scope root = Scope::NewRootScope();
39 ///     auto a = Placeholder(root, DT_INT32);
40 ///     auto c = Add(root, a, {41});
41 ///
42 ///     ClientSession session(root);
43 ///     std::vector<Tensor> outputs;
44 ///
45 ///     Status s = session.Run({ {a, {1}} }, {c}, &outputs);
46 ///     if (!s.ok()) { ... }
47 class ClientSession {
48  public:
49   /// A data type to represent feeds to a Run call.
50   ///
51   /// This is a map of `Output` objects returned by op-constructors to the value
52   /// to feed them with. See `Input::Initializer` for details on what can be
53   /// used as feed values.
54   typedef std::unordered_map<Output, Input::Initializer, OutputHash> FeedType;
55 
56   /// Create a new session to evaluate the graph contained in `scope` by
57   /// connecting to the TensorFlow runtime specified by `target`.
58   ClientSession(const Scope& scope, const string& target);
59 
60   /// Same as above, but use the empty string ("") as the target specification.
61   ClientSession(const Scope& scope);
62 
63   /// Create a new session, configuring it with `session_options`.
64   ClientSession(const Scope& scope, const SessionOptions& session_options);
65 
66   ~ClientSession();
67 
68   /// Evaluate the tensors in `fetch_outputs`. The values are returned as
69   /// `Tensor` objects in `outputs`. The number and order of `outputs` will
70   /// match `fetch_outputs`.
71   Status Run(const std::vector<Output>& fetch_outputs,
72              std::vector<Tensor>* outputs) const;
73 
74   /// Same as above, but use the mapping in `inputs` as feeds.
75   Status Run(const FeedType& inputs, const std::vector<Output>& fetch_outputs,
76              std::vector<Tensor>* outputs) const;
77 
78   /// Same as above. Additionally runs the operations ins `run_outputs`.
79   Status Run(const FeedType& inputs, const std::vector<Output>& fetch_outputs,
80              const std::vector<Operation>& run_outputs,
81              std::vector<Tensor>* outputs) const;
82 
83   /// Use `run_options` to turn on performance profiling. `run_metadata`, if not
84   /// null, is filled in with the profiling results.
85   Status Run(const RunOptions& run_options, const FeedType& inputs,
86              const std::vector<Output>& fetch_outputs,
87              const std::vector<Operation>& run_outputs,
88              std::vector<Tensor>* outputs, RunMetadata* run_metadata) const;
89 
90   /// \brief A handle to a subgraph, created with
91   /// `ClientSession::MakeCallable()`.
92   typedef int64 CallableHandle;
93 
94   /// \brief Creates a `handle` for invoking the subgraph defined by
95   /// `callable_options`.
96   /// NOTE: This API is still experimental and may change.
97   Status MakeCallable(const CallableOptions& callable_options,
98                       CallableHandle* out_handle);
99 
100   /// \brief Invokes the subgraph named by `handle` with the given options and
101   /// input tensors.
102   ///
103   /// The order of tensors in `feed_tensors` must match the order of names in
104   /// `CallableOptions::feed()` and the order of tensors in `fetch_tensors` will
105   /// match the order of names in `CallableOptions::fetch()` when this subgraph
106   /// was created.
107   /// NOTE: This API is still experimental and may change.
108   Status RunCallable(CallableHandle handle,
109                      const std::vector<Tensor>& feed_tensors,
110                      std::vector<Tensor>* fetch_tensors,
111                      RunMetadata* run_metadata);
112 
113   /// \brief Releases resources associated with the given `handle` in this
114   /// session.
115   /// NOTE: This API is still experimental and may change.
116   Status ReleaseCallable(CallableHandle handle);
117 
118  private:
119   class Impl;
120   std::unique_ptr<Impl> impl_;
impl()121   Impl* impl() { return impl_.get(); }
impl()122   const Impl* impl() const { return impl_.get(); }
123 };
124 
125 /// @}
126 
127 }  // end namespace tensorflow
128 
129 #endif  // TENSORFLOW_CC_CLIENT_CLIENT_SESSION_H_
130