optimization.py

#   Copyright (c) 2021 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.
"""Optimization and learning rate scheduling."""

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

import paddle.fluid as fluid
from utils.fp16 import create_master_params_grads, master_param_to_train_param, apply_dynamic_loss_scaling


def linear_warmup_decay(learning_rate, warmup_steps, num_train_steps):
    """ Applies linear warmup of learning rate from 0 and decay to 0."""
    with fluid.default_main_program()._lr_schedule_guard():
        lr = fluid.layers.tensor.create_global_var(
            shape=[1],
            value=0.0,
            dtype='float32',
            persistable=True,
            name="scheduled_learning_rate")

        global_step = fluid.layers.learning_rate_scheduler._decay_step_counter()

        with fluid.layers.control_flow.Switch() as switch:
            with switch.case(global_step < warmup_steps):
                warmup_lr = learning_rate * (global_step / warmup_steps)
                fluid.layers.tensor.assign(warmup_lr, lr)
            with switch.default():
                decayed_lr = fluid.layers.learning_rate_scheduler.polynomial_decay(
                    learning_rate=learning_rate,
                    decay_steps=num_train_steps,
                    end_learning_rate=0.0,
                    power=1.0,
                    cycle=False)
                fluid.layers.tensor.assign(decayed_lr, lr)

        return lr


def optimization(loss,
                 warmup_steps,
                 num_train_steps,
                 learning_rate,
                 train_program,
                 weight_decay,
                 scheduler='linear_warmup_decay',
                 use_fp16=False,
                 use_dynamic_loss_scaling=False,
                 init_loss_scaling=1.0,
                 beta1=0.9,
                 beta2=0.98,
                 epsilon=1e-06,
                 boundaries=None,
                 values=None):
    """optimization funxtion"""
    def exclude_from_weight_decay(name):
        """exclude from weight decay"""
        name = name.rstrip('.master')
        if name.find("layer_norm") > -1:
            return True
        bias_suffix = ["_bias", "_b", ".b_0"]
        for suffix in bias_suffix:
            if name.endswith(suffix):
                return True
        return False

    if warmup_steps > 0:
        if scheduler == 'noam_decay':
            scheduled_lr = fluid.layers.learning_rate_scheduler \
                .noam_decay(1 / (warmup_steps * (learning_rate ** 2)),
                            warmup_steps)
        elif scheduler == 'linear_warmup_decay':
            scheduled_lr = linear_warmup_decay(learning_rate, warmup_steps,
                                               num_train_steps)
        elif scheduler == 'scale_by_epoch_decay':
            if boundaries is None:
                boundaries = [10000, 20000]
            if values is None:
                values = [5e-6, 5e-7, 5e-8]
            scheduled_lr = fluid.layers.piecewise_decay(boundaries=boundaries, values=values)
        else:
            raise ValueError("Unkown learning rate scheduler, should be "
                             "'noam_decay' or 'linear_warmup_decay'")
        optimizer = fluid.optimizer.Adam(learning_rate=scheduled_lr, beta1=beta1, beta2=beta2, epsilon=epsilon)
    else:
        scheduled_lr = fluid.layers.create_global_var(
            name=fluid.unique_name.generate("learning_rate"),
            shape=[1],
            value=learning_rate,
            dtype='float32',
            persistable=True)
        optimizer = fluid.optimizer.Adam(learning_rate=scheduled_lr, beta1=beta1, beta2=beta2, epsilon=epsilon)
        optimizer._learning_rate_map[fluid.default_main_program()] = scheduled_lr

    if use_fp16:
        optimizer = fluid.contrib.mixed_precision.decorator.decorate(optimizer,
                                                                     amp_lists=fluid.contrib.mixed_precision.AutoMixedPrecisionLists(
                                                                         custom_black_varnames={'loss'},
                                                                         custom_black_list={'layer_norm', 'arg_max', 'argmax'}),
                                                                     init_loss_scaling=init_loss_scaling,
                                                                     use_dynamic_loss_scaling=use_dynamic_loss_scaling)
        loss_scaling = optimizer.get_loss_scaling()
    else:
        loss_scaling = None

    fluid.clip.set_gradient_clip(
        clip=fluid.clip.GradientClipByGlobalNorm(clip_norm=1.0))

    param_list = dict()

    for param in train_program.global_block().all_parameters():
        param_list[param.name] = param * 1.0
        param_list[param.name].stop_gradient = True

    _, param_grads = optimizer.minimize(loss)

    if weight_decay > 0:
        for param, grad in param_grads:
            if exclude_from_weight_decay(param.name):
                continue
            with param.block.program._optimized_guard(
                    [param, grad]), fluid.framework.name_scope("weight_decay"):
                updated_param = param - param_list[
                    param.name] * weight_decay * scheduled_lr
                fluid.layers.assign(output=param, input=updated_param)

    return scheduled_lr, loss_scaling