train.py 12.6 KB
Newer Older
T
tianxin04 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
#   Copyright (c) 2019 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.
"""ERNIE pretraining."""

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import os
import time
import multiprocessing

24
import numpy as np
T
tianxin04 已提交
25 26 27
import paddle.fluid as fluid

from reader.pretraining import ErnieDataReader
T
tianxin 已提交
28
from model.ernie_v1 import ErnieModel, ErnieConfig
T
tianxin04 已提交
29
from optimization import optimization
T
tianxin 已提交
30
from utils.args import print_arguments, check_cuda
T
tianxin04 已提交
31 32 33 34 35
from utils.init import init_checkpoint, init_pretraining_params

from pretrain_args import parser

args = parser.parse_args()
T
format  
tianxin04 已提交
36

T
tianxin04 已提交
37 38
# yapf: enable.

T
format  
tianxin04 已提交
39

T
tianxin04 已提交
40 41 42 43
def create_model(pyreader_name, ernie_config):
    pyreader = fluid.layers.py_reader(
        capacity=70,
        shapes=[[-1, args.max_seq_len, 1], [-1, args.max_seq_len, 1],
Y
Yibing Liu 已提交
44
                [-1, args.max_seq_len, 1], [-1, args.max_seq_len, 1], [-1, 1],
T
tianxin04 已提交
45 46
                [-1, 1], [-1, 1]],
        dtypes=[
Y
Yibing Liu 已提交
47
            'int64', 'int64', 'int64', 'float32', 'int64', 'int64', 'int64'
T
tianxin04 已提交
48
        ],
Y
Yibing Liu 已提交
49
        lod_levels=[0, 0, 0, 0, 0, 0, 0],
T
tianxin04 已提交
50 51 52
        name=pyreader_name,
        use_double_buffer=True)

Y
Yibing Liu 已提交
53 54
    (src_ids, pos_ids, sent_ids, input_mask, mask_label, mask_pos,
     labels) = fluid.layers.read_file(pyreader)
T
tianxin04 已提交
55 56 57 58 59

    ernie = ErnieModel(
        src_ids=src_ids,
        position_ids=pos_ids,
        sentence_ids=sent_ids,
Y
Yibing Liu 已提交
60
        input_mask=input_mask,
T
tianxin04 已提交
61 62 63 64 65
        config=ernie_config,
        weight_sharing=args.weight_sharing,
        use_fp16=args.use_fp16)

    next_sent_acc, mask_lm_loss, total_loss = ernie.get_pretraining_output(
Y
Yibing Liu 已提交
66
        mask_label, mask_pos, labels)
T
tianxin04 已提交
67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173

    if args.use_fp16 and args.loss_scaling > 1.0:
        total_loss *= args.loss_scaling

    return pyreader, next_sent_acc, mask_lm_loss, total_loss


def predict_wrapper(args,
                    exe,
                    ernie_config,
                    test_prog=None,
                    pyreader=None,
                    fetch_list=None):
    # Context to do validation.
    filelist = args.test_filelist if args.do_test else args.valid_filelist
    data_reader = ErnieDataReader(
        filelist,
        vocab_path=args.vocab_path,
        batch_size=args.batch_size,
        voc_size=ernie_config['vocab_size'],
        shuffle_files=False,
        epoch=1,
        max_seq_len=args.max_seq_len,
        is_test=True)

    if args.do_test:
        assert args.init_checkpoint is not None, "[FATAL] Please use --init_checkpoint '/path/to/checkpoints' \
                                                  to specify you pretrained model checkpoints"

        init_pretraining_params(exe, args.init_checkpoint, test_prog)

    def predict(exe=exe, pyreader=pyreader):

        pyreader.decorate_tensor_provider(data_reader.data_generator())
        pyreader.start()

        cost = 0
        lm_cost = 0
        acc = 0
        steps = 0
        time_begin = time.time()
        while True:
            try:
                each_next_acc, each_mask_lm_cost, each_total_cost = exe.run(
                    fetch_list=fetch_list, program=test_prog)
                acc += each_next_acc
                lm_cost += each_mask_lm_cost
                cost += each_total_cost
                steps += 1
                if args.do_test and steps % args.skip_steps == 0:
                    print("[test_set] steps: %d" % steps)

            except fluid.core.EOFException:
                pyreader.reset()
                break

        used_time = time.time() - time_begin
        return cost, lm_cost, acc, steps, (args.skip_steps / used_time)

    return predict


def test(args):
    ernie_config = ErnieConfig(args.ernie_config_path)
    ernie_config.print_config()

    test_prog = fluid.Program()
    test_startup = fluid.Program()
    with fluid.program_guard(test_prog, test_startup):
        with fluid.unique_name.guard():
            test_pyreader, next_sent_acc, mask_lm_loss, total_loss = create_model(
                pyreader_name='test_reader', ernie_config=ernie_config)

    test_prog = test_prog.clone(for_test=True)

    place = fluid.CUDAPlace(0) if args.use_cuda == True else fluid.CPUPlace()
    exe = fluid.Executor(place)
    exe.run(test_startup)

    predict = predict_wrapper(
        args,
        exe,
        ernie_config,
        test_prog=test_prog,
        pyreader=test_pyreader,
        fetch_list=[next_sent_acc.name, mask_lm_loss.name, total_loss.name])

    print("test begin")
    loss, lm_loss, acc, steps, speed = predict()
    print(
        "[test_set] loss: %f, global ppl: %f, next_sent_acc: %f, speed: %f steps/s"
        % (np.mean(np.array(loss) / steps),
           np.exp(np.mean(np.array(lm_loss) / steps)),
           np.mean(np.array(acc) / steps), speed))


def train(args):
    print("pretraining start")
    ernie_config = ErnieConfig(args.ernie_config_path)
    ernie_config.print_config()

    train_program = fluid.Program()
    startup_prog = fluid.Program()
    with fluid.program_guard(train_program, startup_prog):
        with fluid.unique_name.guard():
            train_pyreader, next_sent_acc, mask_lm_loss, total_loss = create_model(
                pyreader_name='train_reader', ernie_config=ernie_config)
T
tianxin 已提交
174
            scheduled_lr, loss_scaling = optimization(
T
tianxin04 已提交
175 176 177 178 179 180 181 182
                loss=total_loss,
                warmup_steps=args.warmup_steps,
                num_train_steps=args.num_train_steps,
                learning_rate=args.learning_rate,
                train_program=train_program,
                startup_prog=startup_prog,
                weight_decay=args.weight_decay,
                scheduler=args.lr_scheduler,
T
tianxin 已提交
183
                use_fp16=args.use_fp16)
T
tianxin04 已提交
184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223

            fluid.memory_optimize(
                input_program=train_program,
                skip_opt_set=[
                    next_sent_acc.name, mask_lm_loss.name, total_loss.name
                ])

    test_prog = fluid.Program()
    with fluid.program_guard(test_prog, startup_prog):
        with fluid.unique_name.guard():
            test_pyreader, next_sent_acc, mask_lm_loss, total_loss = create_model(
                pyreader_name='test_reader', ernie_config=ernie_config)

    test_prog = test_prog.clone(for_test=True)

    if args.use_cuda:
        place = fluid.CUDAPlace(0)
        dev_count = fluid.core.get_cuda_device_count()
    else:
        place = fluid.CPUPlace()
        dev_count = int(os.environ.get('CPU_NUM', multiprocessing.cpu_count()))

    print("Device count %d" % dev_count)
    print("theoretical memory usage: ")
    print(fluid.contrib.memory_usage(
        program=train_program, batch_size=args.batch_size // args.max_seq_len))

    nccl2_num_trainers = 1
    nccl2_trainer_id = 0
    print("args.is_distributed:", args.is_distributed)
    if args.is_distributed:
        worker_endpoints_env = os.getenv("worker_endpoints")
        worker_endpoints = worker_endpoints_env.split(",")
        trainers_num = len(worker_endpoints)
        current_endpoint = os.getenv("current_endpoint")
        trainer_id = worker_endpoints.index(current_endpoint)
        if trainer_id == 0:
            print("train_id == 0, sleep 60s")
            time.sleep(60)
        print("worker_endpoints:{} trainers_num:{} current_endpoint:{} \
T
format  
tianxin04 已提交
224
              trainer_id:{}".format(worker_endpoints, trainers_num,
T
tianxin04 已提交
225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317
                                    current_endpoint, trainer_id))

        # prepare nccl2 env.
        config = fluid.DistributeTranspilerConfig()
        config.mode = "nccl2"
        t = fluid.DistributeTranspiler(config=config)
        t.transpile(
            trainer_id,
            trainers=worker_endpoints_env,
            current_endpoint=current_endpoint,
            program=train_program,
            startup_program=startup_prog)
        nccl2_num_trainers = trainers_num
        nccl2_trainer_id = trainer_id

    exe = fluid.Executor(place)
    exe.run(startup_prog)

    if args.init_checkpoint and args.init_checkpoint != "":
        init_checkpoint(exe, args.init_checkpoint, train_program, args.use_fp16)

    data_reader = ErnieDataReader(
        filelist=args.train_filelist,
        batch_size=args.batch_size,
        vocab_path=args.vocab_path,
        voc_size=ernie_config['vocab_size'],
        epoch=args.epoch,
        max_seq_len=args.max_seq_len,
        generate_neg_sample=args.generate_neg_sample)

    exec_strategy = fluid.ExecutionStrategy()
    if args.use_fast_executor:
        exec_strategy.use_experimental_executor = True
    exec_strategy.num_threads = dev_count
    exec_strategy.num_iteration_per_drop_scope = min(10, args.skip_steps)

    build_strategy = fluid.BuildStrategy()
    build_strategy.remove_unnecessary_lock = False

    train_exe = fluid.ParallelExecutor(
        use_cuda=args.use_cuda,
        loss_name=total_loss.name,
        build_strategy=build_strategy,
        exec_strategy=exec_strategy,
        main_program=train_program,
        num_trainers=nccl2_num_trainers,
        trainer_id=nccl2_trainer_id)

    if args.valid_filelist and args.valid_filelist != "":
        predict = predict_wrapper(
            args,
            exe,
            ernie_config,
            test_prog=test_prog,
            pyreader=test_pyreader,
            fetch_list=[
                next_sent_acc.name, mask_lm_loss.name, total_loss.name
            ])

    train_pyreader.decorate_tensor_provider(data_reader.data_generator())
    train_pyreader.start()
    steps = 0
    cost = []
    lm_cost = []
    acc = []
    time_begin = time.time()
    while steps < args.num_train_steps:
        try:
            steps += nccl2_num_trainers
            skip_steps = args.skip_steps * nccl2_num_trainers

            if nccl2_trainer_id != 0:
                train_exe.run(fetch_list=[])
                continue

            if steps % skip_steps != 0:
                train_exe.run(fetch_list=[])
            else:
                each_next_acc, each_mask_lm_cost, each_total_cost, np_lr = train_exe.run(
                    fetch_list=[
                        next_sent_acc.name, mask_lm_loss.name, total_loss.name,
                        scheduled_lr.name
                    ])
                acc.extend(each_next_acc)
                lm_cost.extend(each_mask_lm_cost)
                cost.extend(each_total_cost)

                print("feed_queue size", train_pyreader.queue.size())
                time_end = time.time()
                used_time = time_end - time_begin
                epoch, current_file_index, total_file, current_file, mask_type = data_reader.get_progress(
                )
                print("current learning_rate:%f" % np_lr[0])
T
format  
tianxin04 已提交
318 319 320 321 322 323 324 325
                print(
                    "epoch: %d, progress: %d/%d, step: %d, loss: %f, "
                    "ppl: %f, next_sent_acc: %f, speed: %f steps/s, file: %s, mask_type: %s"
                    % (epoch, current_file_index, total_file, steps,
                       np.mean(np.array(cost)),
                       np.mean(np.exp(np.array(lm_cost))),
                       np.mean(np.array(acc)), skip_steps / used_time,
                       current_file, mask_type))
T
tianxin04 已提交
326 327 328 329 330 331 332 333 334 335 336 337 338 339 340
                cost = []
                lm_cost = []
                acc = []
                time_begin = time.time()

            if steps % args.save_steps == 0:
                save_path = os.path.join(args.checkpoints, "step_" + str(steps))
                fluid.io.save_persistables(exe, save_path, train_program)

            if args.valid_filelist and steps % args.validation_steps == 0:
                vali_cost, vali_lm_cost, vali_acc, vali_steps, vali_speed = predict(
                )
                print("[validation_set] epoch: %d, step: %d, "
                      "loss: %f, global ppl: %f, batch-averged ppl: %f, "
                      "next_sent_acc: %f, speed: %f steps/s" %
T
format  
tianxin04 已提交
341
                      (epoch, steps, np.mean(np.array(vali_cost) / vali_steps),
T
tianxin04 已提交
342 343 344 345 346 347 348 349 350 351 352
                       np.exp(np.mean(np.array(vali_lm_cost) / vali_steps)),
                       np.mean(np.exp(np.array(vali_lm_cost) / vali_steps)),
                       np.mean(np.array(vali_acc) / vali_steps), vali_speed))

        except fluid.core.EOFException:
            train_pyreader.reset()
            break


if __name__ == '__main__':
    print_arguments(args)
T
tianxin 已提交
353
    check_cuda(args.use_cuda)
T
tianxin04 已提交
354 355 356 357
    if args.do_test:
        test(args)
    else:
        train(args)