compiler.py 11.0 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
#   Copyright (c) 2018 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.

import multiprocessing
import os
import six
X
polish  
Xin Pan 已提交
18
import sys
19
from .. import compat as cpt
X
Xin Pan 已提交
20
from . import framework
21 22 23

from . import core

X
Xin Pan 已提交
24 25
__all__ = ['CompiledProgram', 'ExecutionStrategy', 'BuildStrategy']

26 27
ExecutionStrategy = core.ParallelExecutor.ExecutionStrategy
BuildStrategy = core.ParallelExecutor.BuildStrategy
F
flame 已提交
28 29
InferNativeConfig = core.NativeConfig
InferAnalysisConfig = core.AnalysisConfig
30 31 32 33 34 35 36 37


def _place_obj(place):
    p = core.Place()
    p.set_place(place)
    return p


X
polish  
Xin Pan 已提交
38
class CompiledProgram(object):
X
polish  
Xin Pan 已提交
39
    """
X
Xin Pan 已提交
40
    Compiles to Graph for execution.
X
polish  
Xin Pan 已提交
41

X
Xin Pan 已提交
42 43 44 45
    1. Users first create the program with layers.
    2. Optionally, users use CompiledProgram to optimize the program before run.
    3. The original program or CompiledProgram is run by executor.

X
polish  
Xin Pan 已提交
46 47 48 49 50 51 52 53
    The CompiledProgram is used to transform a program for various
    optimizations, for example.
      * Pre-compute some logic once so that each run is faster.
      * Transform the program so that it can run in multiple devices.
      * TODO: transform the program for optimized inference or distributed
              training.

    Example:
X
Xin Pan 已提交
54
        .. code-block:: python
X
Xin Pan 已提交
55
            place = fluid.CUDAPlace(0) if use_gpu else fluid.CPUPlace()
X
Xin Pan 已提交
56 57 58 59 60 61 62 63
            exe = fluid.Executor(place)
            exe.run(startup)
            compiled_prog = compiler.CompiledProgram(main).with_data_parallel(
                loss_name=loss.name)
            for i in range(5):
                test_loss, = exe.run(compiled_prog,
                                     feed=feed_dict,
                                     fetch_list=[loss.name])
X
polish  
Xin Pan 已提交
64 65

    Args:
X
Xin Pan 已提交
66 67 68 69 70
        program_or_graph (Graph|Program): If it's Program, it will be first
            lowered to a graph for further optimizations. If it's a graph
            (potentially optimized before), it will be directly used for
            further optimizations. Note: graph is only supported when compiled
            with with_data_parallel option.
X
polish  
Xin Pan 已提交
71 72
    """

X
Xin Pan 已提交
73 74 75 76 77 78 79 80 81 82 83 84
    def __init__(self, program_or_graph):
        if isinstance(program_or_graph, core.Graph):
            self._graph = program_or_graph
            self._program = None
        elif isinstance(program_or_graph, framework.Program):
            self._graph = core.Graph(program_or_graph.desc)
            self._program = program_or_graph
        else:
            raise ValueError("Wrong program_to_graph type: %s" %
                             type(program_or_graph))

        self._program_desc = self._graph.origin_program_desc()
X
polish  
Xin Pan 已提交
85 86 87
        self._scope = None
        self._place = None
        self._executor = None
88 89
        self._compiled = False
        self._is_data_parallel = False
F
flame 已提交
90
        self._is_inference = False
91

X
Xin Pan 已提交
92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117
    def with_data_parallel(self,
                           loss_name=None,
                           build_strategy=None,
                           exec_strategy=None,
                           share_vars_from=None):
        """Configs the program to run in data parallel way.

        Args:
            loss_name (str): The loss name must set in training. Default None.
            build_strategy(BuildStrategy): build_strategy is used to
                build the graph so it can run on multiple devices/cores with
                optimized topology.
                For more information, please refer to fluid.BuildStrategy.
                Default None.
            exec_strategy(ExecutionStrategy): exec_strategy is used to
                to select the a way to execute the graph, for example how many
                threads are used, how many iterations to clean up the temp
                variables. For more information, please refer
                to fluid.ExecutionStrategy. Default None.
            share_vars_from(CompiledProgram): If provide, this CompiledProgram
                will share variables from `share_vars_from`. `share_vars_from`
                must be run by the executor before this CompiledProgram so that
                vars are ready.
        Returns:
            self
        """
118
        assert not self._is_data_parallel, "Already compiled with parallel."
X
Xin Pan 已提交
119
        assert not self._is_inference, "Cannot compile both data parallel and inference"
120 121 122 123
        self._is_data_parallel = True
        self._build_strategy = build_strategy
        self._exec_strategy = exec_strategy
        self._loss_name = loss_name
X
polish  
Xin Pan 已提交
124
        self._share_vars_from = share_vars_from
X
fix  
Xin Pan 已提交
125 126 127 128
        if self._exec_strategy is None:
            self._exec_strategy = ExecutionStrategy()
        if self._build_strategy is None:
            self._build_strategy = BuildStrategy()
129 130
        return self

F
flame 已提交
131 132 133 134 135 136 137 138
    def with_inference_optimize(self, config):
        """ Add inference optimize

        Args:
            config: instance of `NativeConfig` or `AnalysisConfig` to create predictor
        Returns:
            self
        """
X
Xin Pan 已提交
139 140 141
        assert not self._is_data_parallel, "Cannot compile both data parallel and inference."
        assert not self._is_inference, "Already compiled with inference"

F
flame 已提交
142 143 144 145 146 147 148
        assert any([
            isinstance(config, InferNativeConfig),
            isinstance(config, InferAnalysisConfig)
        ])
        self._is_inference = True
        self._infer_config = config
        return self
X
polish  
Xin Pan 已提交
149

F
flame 已提交
150
    def _with_distributed(self):
X
polish  
Xin Pan 已提交
151 152
        raise NotImplementedError()

153
    def _compile_data_parallel(self):
X
polish  
Xin Pan 已提交
154 155 156 157 158 159 160 161 162 163 164 165 166
        if self._share_vars_from:
            if self._scope:
                sys.stderr.write("share_vars_from is set, scope is ignored.\n")
            if not self._share_vars_from._is_data_parallel:
                raise ValueError("share_vars_from is not data parallel. Cannot "
                                 "share vars from it.")
            if self._share_vars_from._executor is None:
                raise ValueError(
                    "share_vars_from is not compiled and run, so there is no "
                    "var to share.")
            self._local_scopes = self._share_vars_from._executor.local_scopes()
        else:
            self._local_scopes = []
167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193

        self._exec_strategy.use_cuda = isinstance(self._place, core.CUDAPlace)
        if self._exec_strategy.use_cuda:
            gpus_env = os.getenv("FLAGS_selected_gpus")
            if gpus_env:
                gpus = [int(s) for s in gpus_env.split(",")]
            else:
                gpus = [
                    i for i in six.moves.range(core.get_cuda_device_count())
                ]
            self._places = [core.CUDAPlace(i) for i in gpus]
        else:
            cpu_num = int(
                os.environ.get('CPU_NUM', multiprocessing.cpu_count()))
            self._places = [core.CPUPlace() for _ in six.moves.range(cpu_num)]
        assert self._places, "no place for execution"

        if self._exec_strategy.num_threads == 0:
            if self._exec_strategy.use_cuda:
                # Experiments on se-resnext shows that too many threads hurt
                # performance. Worth tunning for other models in the future.
                self._exec_strategy.num_threads = len(self._places) * 4
            else:
                cpu_num = int(
                    os.environ.get('CPU_NUM', multiprocessing.cpu_count()))
                self._exec_strategy.num_threads = cpu_num * 2

D
dzhwinter 已提交
194 195
        # FIXME(dzhwinter): enable_inplace should be after memory_optimize
        # if turn on python memory optimize, turn off the inplace_pass.
D
dzhwinter 已提交
196
        if self._build_strategy.memory_optimize is None:
X
Xin Pan 已提交
197
            self._build_strategy.memory_optimize = False if self._program and self._program._is_mem_optimized else True
D
dzhwinter 已提交
198
        if self._build_strategy.enable_inplace is None:
X
Xin Pan 已提交
199 200 201 202 203 204 205 206
            self._build_strategy.enable_inplace = False if self._program and self._program._is_mem_optimized else True


        # TODO(wuyi): trainer endpoings should be passed in through
        # build_strategy, not program.xxx.
        if self._program and self._build_strategy.num_trainers > 1 and \
                self._program._trainers_endpoints:
            tps = self._program._trainers_endpoints
D
dzhwinter 已提交
207

208
            assert self._build_strategy.num_trainers == len(
X
Xin Pan 已提交
209 210 211 212 213 214 215 216 217 218
                tps), "num_trainers == len(end_points)"
            self._build_strategy.trainers_endpoints = tps

        self._persistable_vars = []
        for block_id in range(self._program_desc.num_blocks()):
            bdesc = self._program_desc.block(block_id)
            self._persistable_vars.extend([
                cpt.to_text(v.name()) for v in bdesc.all_vars()
                if v.persistable() and v.type() != core.VarDesc.VarType.RAW
            ])
219 220

        places = list(map(_place_obj, self._places))
X
Xin Pan 已提交
221 222 223 224 225 226 227 228 229 230 231 232 233 234 235

        # FIXME(Yancey1989): parallel graph mode get better performance
        # in GPU allreduce distributed training. Need an elegant way to
        # choice the execution strategy.
        enable_parallel_graph = \
            core._enable_parallel_graph_execution(self._graph,
                                                  self._exec_strategy,
                                                  self._build_strategy) and \
            self._program  # only supported if compile program not graph.

        self._pe_graphs = [self._graph]
        if enable_parallel_graph:
            for _ in range(len(places) - 1):
                self._pe_graphs.append(core.Graph(self._program_desc))

236
        return core.ParallelExecutor(
X
Xin Pan 已提交
237 238
            places,
            set(self._persistable_vars), self._pe_graphs,
239 240 241 242
            cpt.to_text(self._loss_name)
            if self._loss_name else six.u(''), self._scope, self._local_scopes,
            self._exec_strategy, self._build_strategy)

F
flame 已提交
243 244 245
    def _compile_inference(self):
        return core.create_paddle_predictor(self._infer_config)

246
    def _compile(self, scope, place):
X
Xin Pan 已提交
247 248 249 250 251 252 253 254 255 256
        """Compile the program based on the configs.

        Args:
            scope: The variables (resources) that are associated with
               this compiled program.
            place: The location that the compiled program will be run on.

        Returns:
            self
        """
257
        if self._compiled:
X
polish  
Xin Pan 已提交
258 259
            if scope and self._scope != scope:
                raise ValueError("Cannot compile with different scope")
S
sneaxiy 已提交
260
            if place and not self._place._equals(place):
X
polish  
Xin Pan 已提交
261
                raise ValueError("Cannot compile with different place")
262
            return self
X
fix  
Xin Pan 已提交
263
        self._compiled = True
264 265 266 267 268

        self._scope = scope
        self._place = place
        if self._is_data_parallel:
            self._executor = self._compile_data_parallel()
F
flame 已提交
269 270
        elif self._is_inference:
            self._executor = self._compile_inference()
271 272 273 274
        else:
            p = _place_obj(self._place)
            self._executor = core.Executor(p)
        return self