cluster_trainer.py 5.2 KB
Newer Older
T
tangwei 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
# 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.

"""
Training use fluid with one node only.
"""

from __future__ import print_function

T
tangwei 已提交
21
import paddle.fluid as fluid
T
tangwei 已提交
22 23 24 25
from paddle.fluid.incubate.fleet.parameter_server.distribute_transpiler import fleet
from paddle.fluid.incubate.fleet.parameter_server.distribute_transpiler.distributed_strategy import StrategyFactory
from paddle.fluid.incubate.fleet.base.role_maker import PaddleCloudRoleMaker

T
tangwei 已提交
26 27
from fleetrec.core.utils import envs
from fleetrec.core.trainers.transpiler_trainer import TranspileTrainer
T
tangwei 已提交
28 29


T
tangwei 已提交
30
class ClusterTrainer(TranspileTrainer):
T
tangwei 已提交
31 32 33 34
    def processor_register(self):
        role = PaddleCloudRoleMaker()
        fleet.init(role)

T
tangwei12 已提交
35
        if fleet.is_server():
T
tangwei 已提交
36 37 38 39 40 41
            self.regist_context_processor('uninit', self.instance)
            self.regist_context_processor('init_pass', self.init)
            self.regist_context_processor('server_pass', self.server)
        else:
            self.regist_context_processor('uninit', self.instance)
            self.regist_context_processor('init_pass', self.init)
T
tangwei 已提交
42 43 44 45 46

            if envs.get_platform() == "LINUX":
                self.regist_context_processor('train_pass', self.dataset_train)
            else:
                self.regist_context_processor('train_pass', self.dataloader_train)
T
tangwei 已提交
47 48 49
            self.regist_context_processor('terminal_pass', self.terminal)

    def build_strategy(self):
T
tangwei 已提交
50
        mode = envs.get_runtime_environ("train.trainer.strategy")
T
fix bug  
tangwei 已提交
51
        assert mode in ["async", "geo", "sync", "half_async"]
T
tangwei 已提交
52

T
tangwei 已提交
53 54 55 56 57 58 59 60 61 62 63 64 65 66
        strategy = None

        if mode == "async":
            strategy = StrategyFactory.create_async_strategy()
        elif mode == "geo":
            push_num = envs.get_global_env("train.strategy.mode.push_num", 100)
            strategy = StrategyFactory.create_geo_strategy(push_num)
        elif mode == "sync":
            strategy = StrategyFactory.create_sync_strategy()
        elif mode == "half_async":
            strategy = StrategyFactory.create_half_async_strategy()

        assert strategy is not None

T
tangwei 已提交
67
        self.strategy = strategy
T
tangwei 已提交
68 69 70
        return strategy

    def init(self, context):
T
tangwei 已提交
71
        self.model.train_net()
T
tangwei 已提交
72 73 74
        optimizer = self.model.optimizer()
        strategy = self.build_strategy()
        optimizer = fleet.distributed_optimizer(optimizer, strategy)
T
tangwei 已提交
75
        optimizer.minimize(self.model.get_cost_op())
T
tangwei 已提交
76 77 78 79

        if fleet.is_server():
            context['status'] = 'server_pass'
        else:
T
tangwei 已提交
80 81 82 83 84 85 86 87
            self.fetch_vars = []
            self.fetch_alias = []
            self.fetch_period = self.model.get_fetch_period()

            metrics = self.model.get_metrics()
            if metrics:
                self.fetch_vars = metrics.values()
                self.fetch_alias = metrics.keys()
T
tangwei 已提交
88 89 90 91 92 93 94
            context['status'] = 'train_pass'

    def server(self, context):
        fleet.init_server()
        fleet.run_server()
        context['is_exit'] = True

T
tangwei 已提交
95 96 97 98
    def dataloader_train(self, context):
        pass

    def dataset_train(self, context):
T
tangwei 已提交
99
        self._exe.run(fleet.startup_program)
T
tangwei 已提交
100

T
tangwei 已提交
101 102
        fleet.init_worker()

T
tangwei 已提交
103
        reader = self._get_dataloader()
T
tangwei 已提交
104 105
        epochs = envs.get_global_env("train.epochs")

T
tangwei 已提交
106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141
        program = fluid.compiler.CompiledProgram(
            fleet.main_program).with_data_parallel(
            loss_name=self.model.get_cost_op().name,
        build_strategy=self.strategy.get_build_strategy(),
        exec_strategy=self.strategy.get_execute_strategy())

        metrics_varnames = []
        metrics_format = []

        metrics_format.append("{}: {{}}".format("epoch"))
        metrics_format.append("{}: {{}}".format("batch"))

        for name, var in self.model.get_metrics().items():
            metrics_varnames.append(var.name)
            metrics_format.append("{}: {{}}".format(name))

        metrics_format = ", ".join(metrics_format)

        for epoch in range(epochs):
            reader.start()
            batch_id = 0
            try:
                while True:
                    metrics_rets = self._exe.run(
                        program=program,
                        fetch_list=metrics_varnames)

                    metrics = [epoch, batch_id]
                    metrics.extend(metrics_rets)

                    if batch_id % 10 == 0 and batch_id != 0:
                        print(metrics_format.format(*metrics))
                    batch_id += 1
            except fluid.core.EOFException:
                reader.reset()

T
tangwei 已提交
142
        fleet.stop_worker()
T
tangwei 已提交
143
        context['status'] = 'terminal_pass'
T
tangwei 已提交
144 145 146 147 148 149 150 151

    def infer(self, context):
        context['status'] = 'terminal_pass'

    def terminal(self, context):
        for model in self.increment_models:
            print("epoch :{}, dir: {}".format(model[0], model[1]))
        context['is_exit'] = True