You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Paddle/python/paddle/distributed/fleet/dataset/dataset.py

1276 lines
43 KiB

# 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.
"""This is definition of dataset class, which is high performance IO."""
import paddle
from paddle.fluid.proto import data_feed_pb2
from google.protobuf import text_format
import paddle.fluid.core as core
class DatasetBase(object):
""" Base dataset class. """
def __init__(self):
""" Init. """
# define class name here
# to decide whether we need create in memory instance
self.proto_desc = data_feed_pb2.DataFeedDesc()
self.proto_desc.pipe_command = "cat"
self.dataset = core.Dataset("MultiSlotDataset")
self.thread_num = 1
self.filelist = []
def init(self,
batch_size=1,
thread_num=1,
use_var=[],
pipe_command="cat",
input_type=0,
fs_name="",
fs_ugi="",
download_cmd="cat"):
"""
should be called only once in user's python scripts to initialize setings of dataset instance.
Normally, it is called by InMemoryDataset or QueueDataset.
Args:
batch_size(int): batch size. It will be effective during training. default is 1.
thread_num(int): thread num, it is the num of readers. default is 1.
use_var(list): list of variables. Variables which you will use. default is [].
pipe_command(str): pipe command of current dataset. A pipe command is a UNIX pipeline command that can be used only. default is "cat"
input_type(int): the input type of generated input. 0 is for one sample, 1 is for one batch. defalut is 0.
fs_name(str): fs name. default is "".
fs_ugi(str): fs ugi. default is "".
download_cmd(str): customized download command. default is "cat"
"""
self._set_batch_size(batch_size)
self._set_thread(thread_num)
self._set_use_var(use_var)
self._set_pipe_command(pipe_command)
self._set_input_type(input_type)
self._set_hdfs_config(fs_name, fs_ugi)
self._set_download_cmd(download_cmd)
def _set_pipe_command(self, pipe_command):
"""
Set pipe command of current dataset
A pipe command is a UNIX pipeline command that can be used only
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.dataset.DatasetBase()
dataset._set_pipe_command("python my_script.py")
Args:
pipe_command(str): pipe command
"""
self.proto_desc.pipe_command = pipe_command
def _set_batch_size(self, batch_size):
"""
Set batch size. Will be effective during training
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.DatasetBase()
dataset._set_batch_size(128)
Args:
batch_size(int): batch size
"""
self.proto_desc.batch_size = batch_size
def _set_thread(self, thread_num):
"""
Set thread num, it is the num of readers.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.DatasetBase()
dataset._set_thread(12)
Args:
thread_num(int): thread num
"""
self.dataset.set_thread_num(thread_num)
self.thread_num = thread_num
def set_filelist(self, filelist):
"""
Set file list in current worker.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.DatasetBase()
dataset.set_filelist(['a.txt', 'b.txt'])
Args:
filelist(list): file list
"""
self.dataset.set_filelist(filelist)
self.filelist = filelist
def _set_input_type(self, input_type):
self.proto_desc.input_type = input_type
def _set_use_var(self, var_list):
"""
Set Variables which you will use.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.DatasetBase()
dataset._set_use_var([data, label])
Args:
var_list(list): variable list
"""
multi_slot = self.proto_desc.multi_slot_desc
for var in var_list:
slot_var = multi_slot.slots.add()
slot_var.is_used = True
slot_var.name = var.name
if var.lod_level == 0:
slot_var.is_dense = True
slot_var.shape.extend(var.shape)
if var.dtype == core.VarDesc.VarType.FP32:
slot_var.type = "float"
elif var.dtype == core.VarDesc.VarType.INT64:
slot_var.type = "uint64"
else:
raise ValueError(
"Currently, paddle.distributed.fleet.dataset only supports dtype=float32 and dtype=int64"
)
def _set_hdfs_config(self, fs_name, fs_ugi):
"""
Set hdfs config: fs name ad ugi
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.DatasetBase()
dataset._set_hdfs_config("my_fs_name", "my_fs_ugi")
Args:
fs_name(str): fs name
fs_ugi(str): fs ugi
"""
self.dataset.set_hdfs_config(fs_name, fs_ugi)
def _set_download_cmd(self, download_cmd):
"""
Set customized download cmd: download_cmd
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.DatasetBase()
dataset._set_download_cmd("./read_from_afs")
Args:
download_cmd(str): customized download command
"""
self.dataset.set_download_cmd(download_cmd)
def _prepare_to_run(self):
"""
Set data_feed_desc before load or shuffle,
user no need to call this function.
"""
if self.thread_num > len(self.filelist):
self.thread_num = len(self.filelist)
self.dataset.set_thread_num(self.thread_num)
self.dataset.set_data_feed_desc(self._desc())
self.dataset.create_readers()
def _finish_to_run(self):
self.dataset.destroy_readers()
def _desc(self):
"""
Returns a protobuf message for this DataFeedDesc
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.DatasetBase()
print(dataset._desc())
Returns:
A string message
"""
return text_format.MessageToString(self.proto_desc)
def _dynamic_adjust_before_train(self, thread_num):
pass
def _dynamic_adjust_after_train(self):
pass
class InMemoryDataset(DatasetBase):
"""
InMemoryDataset, it will load data into memory
and shuffle data before training.
Example:
import paddle
dataset = paddle.distributed.InMemoryDataset()
"""
def __init__(self):
""" Init. """
super(InMemoryDataset, self).__init__()
self.proto_desc.name = "MultiSlotInMemoryDataFeed"
self.fleet_send_batch_size = None
self.is_user_set_queue_num = False
self.queue_num = None
self.parse_ins_id = False
self.parse_content = False
self.parse_logkey = False
self.merge_by_sid = True
self.enable_pv_merge = False
self.merge_by_lineid = False
self.fleet_send_sleep_seconds = None
def _init_distributed_settings(self, **kwargs):
"""
should be called only once in user's python scripts to initialize distributed-related setings of dataset instance
Args:
kwargs: Keyword arguments. Currently, we support following keys in **kwargs:
merge_size(int): ins size to merge, if merge_size > 0, set merge by line id,
instances of same line id will be merged after shuffle,
you should parse line id in data generator. default is -1.
parse_ins_id(bool): Set if Dataset need to parse ins_id. default is False.
parse_content(bool): Set if Dataset need to parse content. default is False.
fleet_send_batch_size(int): Set fleet send batch size in one rpc, default is 1024
fleet_send_sleep_seconds(int): Set fleet send sleep time, default is 0
fea_eval(bool): Set if Dataset need to do feature importance evaluation using slots shuffle.
default is False.
candidate_size(int): if fea_eval is set True, set the candidate size used in slots shuffle.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
dataset.init(
batch_size=1,
thread_num=2,
input_type=1,
pipe_command="cat",
use_var=[])
dataset._init_distributed_settings(
parse_ins_id=True,
parse_content=True,
fea_eval=True,
candidate_size=10000)
"""
merge_size = kwargs.get("merge_size", -1)
if merge_size > 0:
self._set_merge_by_lineid(merge_size)
parse_ins_id = kwargs.get("parse_ins_id", False)
self._set_parse_ins_id(parse_ins_id)
parse_content = kwargs.get("parse_content", False)
self._set_parse_content(parse_content)
fleet_send_batch_size = kwargs.get("fleet_send_batch_size", None)
if fleet_send_batch_size:
self._set_fleet_send_batch_size(fleet_send_batch_size)
fleet_send_sleep_seconds = kwargs.get("fleet_send_sleep_seconds", None)
if fleet_send_sleep_seconds:
self._set_fleet_send_sleep_seconds(fleet_send_sleep_seconds)
fea_eval = kwargs.get("fea_eval", False)
if fea_eval:
candidate_size = kwargs.get("candidate_size", 10000)
self._set_fea_eval(candidate_size, True)
def update_settings(self, **kwargs):
"""
should be called in user's python scripts to update setings of dataset instance
Args:
kwargs: Keyword arguments. Currently, we support following keys in **kwargs,
including single node settings and advanced distributed related settings:
batch_size(int): batch size. It will be effective during training. default is 1.
thread_num(int): thread num, it is the num of readers. default is 1.
use_var(list): list of variables. Variables which you will use. default is [].
input_type(int): the input type of generated input. 0 is for one sample, 1 is for one batch. defalut is 0.
fs_name(str): fs name. default is "".
fs_ugi(str): fs ugi. default is "".
pipe_command(str): pipe command of current dataset. A pipe command is a UNIX pipeline command that can be used only. default is "cat"
download_cmd(str): customized download command. default is "cat"
data_feed_type(str): data feed type used in c++ code. default is "MultiSlotInMemoryDataFeed".
queue_num(int): Dataset output queue num, training threads get data from queues. default is-1, which is set same as thread number in c++.
merge_size(int): ins size to merge, if merge_size > 0, set merge by line id,
instances of same line id will be merged after shuffle,
you should parse line id in data generator. default is -1.
parse_ins_id(bool): Set if Dataset need to parse ins_id. default is False.
parse_content(bool): Set if Dataset need to parse content. default is False.
fleet_send_batch_size(int): Set fleet send batch size in one rpc, default is 1024
fleet_send_sleep_seconds(int): Set fleet send sleep time, default is 0
fea_eval(bool): Set if Dataset need to do feature importance evaluation using slots shuffle.
default is False.
candidate_size(int): if fea_eval is set True, set the candidate size used in slots shuffle.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
dataset.init(
batch_size=1,
thread_num=2,
input_type=1,
pipe_command="cat",
use_var=[])
dataset._init_distributed_settings(
parse_ins_id=True,
parse_content=True,
fea_eval=True,
candidate_size=10000)
dataset.update_settings(batch_size=2)
"""
for key in kwargs:
if key == "pipe_command":
self._set_pipe_command(kwargs[key])
elif key == "batch_size":
self._set_batch_size(kwargs[key])
elif key == "thread_num":
self._set_thread(kwargs[key])
elif key == "use_var":
self._set_use_var(kwargs[key])
elif key == "input_type":
self._set_input_type(kwargs[key])
elif key == "fs_name" and "fs_ugi" in kwargs:
self._set_hdfs_config(kwargs[key], kwargs["fs_ugi"])
elif key == "download_cmd":
self._set_download_cmd(kwargs[key])
elif key == "merge_size" and kwargs.get("merge_size", -1) > 0:
self._set_merge_by_lineid(kwargs[key])
elif key == "parse_ins_id":
self._set_parse_ins_id(kwargs[key])
elif key == "parse_content":
self._set_parse_content(kwargs[key])
elif key == "fleet_send_batch_size":
self._set_fleet_send_batch_size(kwargs[key])
elif key == "fleet_send_sleep_seconds":
self._set_fleet_send_sleep_seconds(kwargs[key])
elif key == "fea_eval" and kwargs[key] == True:
candidate_size = kwargs.get("candidate_size", 10000)
self._set_fea_eval(candidate_size, True)
def init(self, **kwargs):
"""
should be called only once in user's python scripts to initialize setings of dataset instance
Args:
kwargs: Keyword arguments. Currently, we support following keys in **kwargs:
batch_size(int): batch size. It will be effective during training. default is 1.
thread_num(int): thread num, it is the num of readers. default is 1.
use_var(list): list of variables. Variables which you will use. default is [].
input_type(int): the input type of generated input. 0 is for one sample, 1 is for one batch. defalut is 0.
fs_name(str): fs name. default is "".
fs_ugi(str): fs ugi. default is "".
pipe_command(str): pipe command of current dataset. A pipe command is a UNIX pipeline command that can be used only. default is "cat"
download_cmd(str): customized download command. default is "cat"
data_feed_type(str): data feed type used in c++ code. default is "MultiSlotInMemoryDataFeed".
queue_num(int): Dataset output queue num, training threads get data from queues. default is -1, which is set same as thread number in c++.
Examples:
.. code-block:: python
import paddle
with open("test_queue_dataset_run_a.txt", "w") as f:
data = "2 1 2 2 5 4 2 2 7 2 1 3\n"
data += "2 6 2 2 1 4 2 2 4 2 2 3\n"
data += "2 5 2 2 9 9 2 2 7 2 1 3\n"
data += "2 7 2 2 1 9 2 3 7 2 5 3\n"
f.write(data)
with open("test_queue_dataset_run_b.txt", "w") as f:
data = "2 1 2 2 5 4 2 2 7 2 1 3\n"
data += "2 6 2 2 1 4 2 2 4 2 2 3\n"
data += "2 5 2 2 9 9 2 2 7 2 1 3\n"
data += "2 7 2 2 1 9 2 3 7 2 5 3\n"
f.write(data)
slots = ["slot1", "slot2", "slot3", "slot4"]
slots_vars = []
for slot in slots:
var = fluid.data(
name=slot, shape=[None, 1], dtype="int64", lod_level=1)
slots_vars.append(var)
dataset = paddle.distributed.InMemoryDataset()
dataset.init(
batch_size=1,
thread_num=2,
input_type=1,
pipe_command="cat",
use_var=slots_vars)
dataset.set_filelist(
["test_queue_dataset_run_a.txt", "test_queue_dataset_run_b.txt"])
dataset.load_into_memory()
exe = fluid.Executor(fluid.CPUPlace() if not core.is_compiled_with_cuda(
) else fluid.CUDAPlace(0))
exe.run(fluid.default_startup_program())
exe.train_from_dataset(fluid.default_main_program(),
dataset)
os.remove("./test_queue_dataset_run_a.txt")
os.remove("./test_queue_dataset_run_b.txt")
"""
batch_size = kwargs.get("batch_size", 1)
thread_num = kwargs.get("thread_num", 1)
use_var = kwargs.get("use_var", [])
input_type = kwargs.get("input_type", 0)
fs_name = kwargs.get("fs_name", "")
fs_ugi = kwargs.get("fs_ugi", "")
pipe_command = kwargs.get("pipe_command", "cat")
download_cmd = kwargs.get("download_cmd", "cat")
super(InMemoryDataset, self).init(
batch_size=batch_size,
thread_num=thread_num,
use_var=use_var,
pipe_command=pipe_command,
input_type=input_type,
fs_name=fs_name,
fs_ugi=fs_ugi,
download_cmd=download_cmd)
data_feed_type = kwargs.get("data_feed_type",
"MultiSlotInMemoryDataFeed")
self._set_feed_type(data_feed_type)
if kwargs.get("queue_num", -1) > 0:
queue_num = kwargs.get("queue_num", -1)
self._set_queue_num(queue_num)
def _set_feed_type(self, data_feed_type):
"""
Set data_feed_desc
"""
self.proto_desc.name = data_feed_type
def _prepare_to_run(self):
"""
Set data_feed_desc before load or shuffle,
user no need to call this function.
"""
if self.thread_num <= 0:
self.thread_num = 1
self.dataset.set_thread_num(self.thread_num)
if self.queue_num is None:
self.queue_num = self.thread_num
self.dataset.set_queue_num(self.queue_num)
self.dataset.set_parse_ins_id(self.parse_ins_id)
self.dataset.set_parse_content(self.parse_content)
self.dataset.set_parse_logkey(self.parse_logkey)
self.dataset.set_merge_by_sid(self.merge_by_sid)
self.dataset.set_enable_pv_merge(self.enable_pv_merge)
self.dataset.set_data_feed_desc(self._desc())
self.dataset.create_channel()
self.dataset.create_readers()
def _dynamic_adjust_before_train(self, thread_num):
if not self.is_user_set_queue_num:
self.dataset.dynamic_adjust_channel_num(thread_num, False)
self.dataset.dynamic_adjust_readers_num(thread_num)
def _dynamic_adjust_after_train(self):
if not self.is_user_set_queue_num:
self.dataset.dynamic_adjust_channel_num(self.thread_num, False)
self.dataset.dynamic_adjust_readers_num(self.thread_num)
def _set_queue_num(self, queue_num):
"""
Set Dataset output queue num, training threads get data from queues
Args:
queue_num(int): dataset output queue num
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
dataset._set_queue_num(12)
"""
self.is_user_set_queue_num = True
self.queue_num = queue_num
def _set_parse_ins_id(self, parse_ins_id):
"""
Set if Dataset need to parse insid
Args:
parse_ins_id(bool): if parse ins_id or not
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
dataset._set_parse_ins_id(True)
"""
self.parse_ins_id = parse_ins_id
def _set_parse_content(self, parse_content):
"""
Set if Dataset need to parse content
Args:
parse_content(bool): if parse content or not
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
dataset._set_parse_content(True)
"""
self.parse_content = parse_content
def _set_fleet_send_batch_size(self, fleet_send_batch_size=1024):
"""
Set fleet send batch size, default is 1024
Args:
fleet_send_batch_size(int): fleet send batch size
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
dataset._set_fleet_send_batch_size(800)
"""
self.fleet_send_batch_size = fleet_send_batch_size
def _set_fleet_send_sleep_seconds(self, fleet_send_sleep_seconds=0):
"""
Set fleet send sleep time, default is 0
Args:
fleet_send_sleep_seconds(int): fleet send sleep time
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
dataset._set_fleet_send_sleep_seconds(2)
"""
self.fleet_send_sleep_seconds = fleet_send_sleep_seconds
def _set_merge_by_lineid(self, merge_size=2):
"""
Set merge by line id, instances of same line id will be merged after
shuffle, you should parse line id in data generator.
Args:
merge_size(int): ins size to merge. default is 2.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
dataset._set_merge_by_lineid()
"""
self.dataset.set_merge_by_lineid(merge_size)
self.merge_by_lineid = True
self.parse_ins_id = True
def _set_generate_unique_feasigns(self, generate_uni_feasigns, shard_num):
self.dataset.set_generate_unique_feasigns(generate_uni_feasigns)
self.gen_uni_feasigns = generate_uni_feasigns
self.local_shard_num = shard_num
def _generate_local_tables_unlock(self, table_id, fea_dim, read_thread_num,
consume_thread_num, shard_num):
self.dataset.generate_local_tables_unlock(
table_id, fea_dim, read_thread_num, consume_thread_num, shard_num)
def load_into_memory(self):
"""
Load data into memory
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.load_into_memory()
"""
self._prepare_to_run()
self.dataset.load_into_memory()
def preload_into_memory(self, thread_num=None):
"""
Load data into memory in async mode
Args:
thread_num(int): preload thread num
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.preload_into_memory()
dataset.wait_preload_done()
"""
self._prepare_to_run()
if thread_num is None:
thread_num = self.thread_num
self.dataset.set_preload_thread_num(thread_num)
self.dataset.create_preload_readers()
self.dataset.preload_into_memory()
def wait_preload_done(self):
"""
Wait preload_into_memory done
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.preload_into_memory()
dataset.wait_preload_done()
"""
self.dataset.wait_preload_done()
self.dataset.destroy_preload_readers()
def local_shuffle(self):
"""
Local shuffle
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.load_into_memory()
dataset.local_shuffle()
"""
self.dataset.local_shuffle()
def global_shuffle(self, fleet=None, thread_num=12):
"""
Global shuffle.
Global shuffle can be used only in distributed mode. i.e. multiple
processes on single machine or multiple machines training together.
If you run in distributed mode, you should pass fleet instead of None.
Examples:
.. code-block:: python
import paddle
from paddle.fluid.incubate.fleet.parameter_server.pslib import fleet
dataset = paddle.distributed.InMemoryDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.load_into_memory()
dataset.global_shuffle(fleet)
Args:
fleet(Fleet): fleet singleton. Default None.
thread_num(int): shuffle thread num. Default is 12.
"""
trainer_num = 1
if fleet is not None:
fleet._role_maker.barrier_worker()
trainer_num = fleet.worker_num()
if self.fleet_send_batch_size is None:
self.fleet_send_batch_size = 1024
if self.fleet_send_sleep_seconds is None:
self.fleet_send_sleep_seconds = 0
self.dataset.register_client2client_msg_handler()
self.dataset.set_trainer_num(trainer_num)
self.dataset.set_fleet_send_batch_size(self.fleet_send_batch_size)
self.dataset.set_fleet_send_sleep_seconds(self.fleet_send_sleep_seconds)
if fleet is not None:
fleet._role_maker.barrier_worker()
self.dataset.global_shuffle(thread_num)
if fleet is not None:
fleet._role_maker.barrier_worker()
if self.merge_by_lineid:
self.dataset.merge_by_lineid()
if fleet is not None:
fleet._role_maker.barrier_worker()
def release_memory(self):
"""
:api_attr: Static Graph
Release InMemoryDataset memory data, when data will not be used again.
Examples:
.. code-block:: python
import paddle
from paddle.fluid.incubate.fleet.parameter_server.pslib import fleet
dataset = paddle.distributed.InMemoryDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.load_into_memory()
dataset.global_shuffle(fleet)
exe = fluid.Executor(fluid.CPUPlace())
exe.run(fluid.default_startup_program())
exe.train_from_dataset(fluid.default_main_program(), dataset)
dataset.release_memory()
"""
self.dataset.release_memory()
def get_memory_data_size(self, fleet=None):
"""
Get memory data size, user can call this function to know the num
of ins in all workers after load into memory.
Note:
This function may cause bad performance, because it has barrier
Args:
fleet(Fleet): Fleet Object.
Returns:
The size of memory data.
Examples:
.. code-block:: python
import paddle
from paddle.fluid.incubate.fleet.parameter_server.pslib import fleet
dataset = paddle.distributed.InMemoryDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.load_into_memory()
print dataset.get_memory_data_size(fleet)
"""
import numpy as np
local_data_size = self.dataset.get_memory_data_size()
local_data_size = np.array([local_data_size])
if fleet is not None:
global_data_size = local_data_size * 0
fleet._role_maker.all_reduce_worker(local_data_size,
global_data_size)
return global_data_size[0]
return local_data_size[0]
def get_shuffle_data_size(self, fleet=None):
"""
Get shuffle data size, user can call this function to know the num
of ins in all workers after local/global shuffle.
Note:
This function may cause bad performance to local shuffle,
because it has barrier. It does not affect global shuffle.
Args:
fleet(Fleet): Fleet Object.
Returns:
The size of shuffle data.
Examples:
.. code-block:: python
import paddle
from paddle.fluid.incubate.fleet.parameter_server.pslib import fleet
dataset = paddle.distributed.InMemoryDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.load_into_memory()
dataset.global_shuffle(fleet)
print dataset.get_shuffle_data_size(fleet)
"""
import numpy as np
local_data_size = self.dataset.get_shuffle_data_size()
local_data_size = np.array([local_data_size])
if fleet is not None:
global_data_size = local_data_size * 0
fleet._role_maker.all_reduce_worker(local_data_size,
global_data_size)
return global_data_size[0]
return local_data_size[0]
def _set_fea_eval(self, record_candidate_size, fea_eval=True):
"""
set fea eval mode for slots shuffle to debug the importance level of
slots(features), fea_eval need to be set True for slots shuffle.
Args:
record_candidate_size(int): size of instances candidate to shuffle
one slot
fea_eval(bool): whether enable fea eval mode to enable slots shuffle.
default is True.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.InMemoryDataset()
dataset._set_fea_eval(1000000, True)
"""
if fea_eval:
self.dataset.set_fea_eval(fea_eval, record_candidate_size)
self.fea_eval = fea_eval
def slots_shuffle(self, slots):
"""
Slots Shuffle
Slots Shuffle is a shuffle method in slots level, which is usually used
in sparse feature with large scale of instances. To compare the metric, i.e.
auc while doing slots shuffle on one or several slots with baseline to
evaluate the importance level of slots(features).
Args:
slots(list[string]): the set of slots(string) to do slots shuffle.
Examples:
import paddle
dataset = paddle.distributed.InMemoryDataset()
dataset.set_merge_by_lineid()
#suppose there is a slot 0
dataset.slots_shuffle(['0'])
"""
if self.fea_eval:
slots_set = set(slots)
self.dataset.slots_shuffle(slots_set)
class QueueDataset(DatasetBase):
"""
QueueDataset, it will process data streamly.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.QueueDataset()
"""
def __init__(self):
"""
Initialize QueueDataset
"""
super(QueueDataset, self).__init__()
self.proto_desc.name = "MultiSlotDataFeed"
def init(self, **kwargs):
"""
should be called only once in user's python scripts to initialize setings of dataset instance
"""
super(QueueDataset, self).init(**kwargs)
def _prepare_to_run(self):
"""
Set data_feed_desc/thread num/filelist before run,
user no need to call this function.
"""
if self.thread_num > len(self.filelist):
self.thread_num = len(self.filelist)
if self.thread_num == 0:
self.thread_num = 1
self.dataset.set_thread_num(self.thread_num)
self.dataset.set_filelist(self.filelist)
self.dataset.set_data_feed_desc(self._desc())
self.dataset.create_readers()
class FileInstantDataset(DatasetBase):
"""
FileInstantDataset, it will process data streamly.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.FileInstantDataset()
"""
def __init__(self):
"""
Initialize FileInstantDataset
"""
super(FileInstantDataset, self).__init__()
self.proto_desc.name = "MultiSlotFileInstantDataFeed"
def init(self, **kwargs):
"""
should be called only once in user's python scripts to initialize setings of dataset instance
"""
super(FileInstantDataset, self).init(**kwargs)
class BoxPSDataset(InMemoryDataset):
"""
BoxPSDataset: derived from InMemoryDataset.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
"""
def __init__(self):
"""
Initialize BoxPSDataset
"""
super(BoxPSDataset, self).__init__()
self.boxps = core.BoxPS(self.dataset)
self.proto_desc.name = "PaddleBoxDataFeed"
def init(self, **kwargs):
"""
should be called only once in user's python scripts to initialize setings of dataset instance
"""
super(BoxPSDataset, self).init(**kwargs)
rank_offset = kwargs.get("rank_offset", "")
self._set_rank_offset(rank_offset)
pv_batch_size = kwargs.get("pv_batch_size", 1)
self._set_pv_batch_size(pv_batch_size)
parse_logkey = kwargs.get("parse_logkey", False)
self._set_parse_logkey(parse_logkey)
merge_by_sid = kwargs.get("merge_by_sid", False)
self._set_merge_by_sid(merge_by_sid)
enable_pv_merge = kwargs.get("enable_pv_merge", False)
self._set_enable_pv_merge(enable_pv_merge)
def _set_rank_offset(self, rank_offset):
"""
Set rank_offset for merge_pv. It set the message of Pv.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
dataset._set_rank_offset("rank_offset")
Args:
rank_offset(str): rank_offset's name
"""
self.proto_desc.rank_offset = rank_offset
def _set_pv_batch_size(self, pv_batch_size):
"""
Set pv batch size. It will be effective during enable_pv_merge
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
dataset._set_pv_batch_size(128)
Args:
pv_batch_size(int): pv batch size
"""
self.proto_desc.pv_batch_size = pv_batch_size
def _set_parse_logkey(self, parse_logkey):
"""
Set if Dataset need to parse logkey
Args:
parse_content(bool): if parse logkey or not
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
dataset._set_parse_logkey(True)
"""
self.parse_logkey = parse_logkey
def _set_merge_by_sid(self, merge_by_sid):
"""
Set if Dataset need to merge sid. If not, one ins means one Pv.
Args:
merge_by_sid(bool): if merge sid or not
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
dataset._set_merge_by_sid(True)
"""
self.merge_by_sid = merge_by_sid
def _set_enable_pv_merge(self, enable_pv_merge):
"""
Set if Dataset need to merge pv.
Args:
enable_pv_merge(bool): if enable_pv_merge or not
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
dataset._set_enable_pv_merge(True)
"""
self.enable_pv_merge = enable_pv_merge
def set_date(self, date):
"""
Workaround for date
"""
year = int(date[:4])
month = int(date[4:6])
day = int(date[6:])
self.boxps.set_date(year, month, day)
def begin_pass(self):
"""
Begin Pass
Notify BoxPS to load sparse parameters of next pass to GPU Memory
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
dataset.begin_pass()
"""
self.boxps.begin_pass()
def end_pass(self, need_save_delta):
"""
End Pass
Notify BoxPS that current pass ended
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
dataset.end_pass(True)
"""
self.boxps.end_pass(need_save_delta)
def wait_preload_done(self):
"""
Wait async preload done
Wait Until Feed Pass Done
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.preload_into_memory()
dataset.wait_preload_done()
"""
self.boxps.wait_feed_pass_done()
def load_into_memory(self):
"""
Load next pass into memory and notify boxps to fetch its emb from SSD
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.load_into_memory()
"""
self._prepare_to_run()
self.boxps.load_into_memory()
def preload_into_memory(self):
"""
Begin async preload next pass while current pass may be training
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.preload_into_memory()
"""
self._prepare_to_run()
self.boxps.preload_into_memory()
def _dynamic_adjust_before_train(self, thread_num):
if not self.is_user_set_queue_num:
self.dataset.dynamic_adjust_channel_num(thread_num, True)
self.dataset.dynamic_adjust_readers_num(thread_num)
def _dynamic_adjust_after_train(self):
pass
def slots_shuffle(self, slots):
"""
Slots Shuffle
Slots Shuffle is a shuffle method in slots level, which is usually used
in sparse feature with large scale of instances. To compare the metric, i.e.
auc while doing slots shuffle on one or several slots with baseline to
evaluate the importance level of slots(features).
Args:
slots(list[string]): the set of slots(string) to do slots shuffle.
Examples:
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
dataset.set_merge_by_lineid()
#suppose there is a slot 0
dataset.slots_shuffle(['0'])
"""
slots_set = set(slots)
self.boxps.slots_shuffle(slots_set)
def set_current_phase(self, current_phase):
"""
Set current phase in train. It is useful for untest.
current_phase : 1 for join, 0 for update.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.load_into_memory()
dataset.set_current_phase(1)
"""
self.dataset.set_current_phase(current_phase)
def get_pv_data_size(self):
"""
Get memory data size of Pv, user can call this function to know the pv num
of ins in all workers after load into memory.
Note:
This function may cause bad performance, because it has barrier
Returns:
The size of memory pv data.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.load_into_memory()
print dataset.get_pv_data_size()
"""
return self.dataset.get_pv_data_size()
def preprocess_instance(self):
"""
Merge pv instance and convey it from input_channel to input_pv_channel.
It will be effective when enable_pv_merge_ is True.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.load_into_memory()
dataset.preprocess_instance()
"""
self.dataset.preprocess_instance()
def postprocess_instance(self):
"""
Divide pv instance and convey it to input_channel.
Examples:
.. code-block:: python
import paddle
dataset = paddle.distributed.fleet.BoxPSDataset()
filelist = ["a.txt", "b.txt"]
dataset.set_filelist(filelist)
dataset.load_into_memory()
dataset.preprocess_instance()
exe.train_from_dataset(dataset)
dataset.postprocess_instance()
"""
self.dataset.postprocess_instance()