Welcome to easycore’s documentation!

Tutorials

Installation

Requirements

  • Python ≥ 3.6
  • PyTorch (optional, for easycore.torch package)

Install

Install from PyPI
pip install easycore
Install from GitHub
pip install 'git+https://github.com/YuxinZhaozyx/easycore.git'
Install from a local clone
git clone https://github.com/YuxinZhaozyx/easycore.git
cd easycore
pip install -e .

Light weight config tools

easycore make it easy to load config from local yaml file, save config and control the config in runtime.

Load config from local yaml file

An example of yaml file is shown bellow:

MODEL:
    IN_FEAUTRES: ["res3", "res4", "res5"]
    INPUT_SIZE: (224, 224)
    NUM_CLASSES: 100
NAME: YuxinZhaozyx

You can load the yaml file in the follow way:

from easycore.common.config import CfgNode as CN

cfg = CN.open('example.yaml')

# or
with open('example.yaml', 'r', encoding='utf-8') as f:
    cfg = CN.open(f)

Get an empty config

cfg = CN()

Get a config from from python dict

init_dict = {
    "MODEL": {
        "IN_FEATURES": ["res3", "res4", "res5"],
        "INPUT_SIZE": (224, 224),
        "NUM_CLASSES": 100,
    },
    "NAME": "YuxinZhaozyx",
}
cfg = CN(init_dict)

Use config

# get value from config
# the config has been automatically transform into python data type.
in_features = cfg.MODEL.IN_FEATURES  # list
input_size = cfg.MODEL.INPUT_SIZE    # tuple
num_classes = cfg.MODEL.NUM_CLASSES  # int
name = cfg.NAME                      # str

# add new value to config
cfg.LICENSE = 'MIT'

# add a new CfgNode to config
cfg.SOLVER = CN()
cfg.SOLVER.LEARNING_RATE = 0.001
cfg.SOLVER.BATCH_SIZE = 128

Merge two config

cfg_a = CN()
cfg_a.key1 = 1
cfg_a.key2 = 2

cfg_b = CN()
cfg_b.key2 = 3
cfg_c.key3 = 4

# merge two config
cfg_a.merge(cfg_b)  # now cfg_a.key2 is 3

Copy a config

cfg_copy = cfg.copy()  # get a deepcopy of cfg

Save config to yaml file

cfg.save("example-save.yaml")

# or
with open("example-save.yaml", 'w', encoding='utf-8') as f:
    cfg.save(f)

API Documentation

Multiprocessing parallel acceleration tools

easycore make it easy to parallel your tasks in cpus and gpus.

API

You can write a parallel runner by inheriting class UnorderedRunner or OrderedRunner and overriding following 6 static methods.

@staticmethod
def producer_init(device, cfg):
    """ 
    function for producer initialization.
    
    Args:
        device (str): device for the this process.
        cfg (easycore.common.config.CfgNode): config of this process, you can use it to transfer data
            to `producer_work` and `producer_end` function.
    """
    pass

@staticmethod
def producer_work(device, cfg, data):
    """ 
    function specify how the producer processes the data.
    
    Args:
        device (str): device for this process.
        cfg (easycore.common.config.CfgNode): config of this process, you can use it to get data from
            `producer_init` function and transfer data to the next `producer_work` and `producer_end`
            function.
        data (Any): data get from input of `__call__` method.
    
    Returns:
        Any: processed data
    """
    return data

@staticmethod
def producer_end(device, cfg):
    """ 
    function after finishing all of its task and before close the process.
    
    Args:
        device (str): device for this process.
        cfg (easycore.common.config.CfgNode): config of this process, you can use it to get data
            from `producer_init` and `producer_work` function.
    """
    pass

@staticmethod
def consumer_init(cfg):
    """
    function for consumer initialization.
    
    Args:
        cfg (easycore.common.config.CfgNode): config of this process, you can use it to transfer data
            to `consumer_work` and `consumer_end` function.
    """
    pass

@staticmethod
def consumer_work(cfg, data):
    """
    function specify how the consumer processses the data from producers.
    
    Args:
        cfg (easycore.common.config.CfgNode): config of this process, you can use it to get data from
            `consumer_init` function and transfer data to the next `consumer_work` and `consumer_end`
            function.
    """
    pass

@staticmethod
def consumer_end(cfg):
    """
    function after receiving all data from producers.
    
    Args:
        cfg (easycore.common.config.CfgNode): config of this process, you can use it get data from
            `consumer_work` function.

    Returns:
        Any: processed data
    """
    return None

Example 1: Sum of squares

It can be implemented with a simple way:

data_list = list(range(100))
result = sum([data * data for data in data_list])

# or more simple
result = 0
for data in data_list:
    square = data * data
    result += square

We calculate square of each element of the list, and then sum they together. In this case, it can be divided into two tasks. We assign this two tasks to producer and consumer respectively.

from easycore.common.config import CfgNode
from easycore.common.parallel import UnorderedRunner

class Runner(UnorderedRunner):
    @staticmethod
    def producer_work(device, cfg, data):
        return data * data  # calculate square of data

    @staticmethod
    def consumer_init(cfg):
        cfg.sum = 0  # init a sum variable with 0, you can use cfg to transfer data

    @staticmethod
    def consumer_work(cfg, data):
        cfg.sum += data  # add the square to the sum variable

    @staticmethod
    def consumer_end(cfg):
        return cfg.sum  # return the result you need

if __name__ == '__main__':
    runner = Runner(devices=3)  # if you specify `device with a integer`, it will use cpus.
    # You can specify a list of str instead, such as: 
    # runner = Runner(devices=["cpu", "cpu", "cpu"]) 
    
    data_list = list(range(100))  # prepare data, it must be iterable
    result = runner(data_list)  # call the runner
    print(result)

    runner.close()  # close the runner and shutdown all processes it opens.

Example 2: An neural network predictor

First we define an neural network in network.py:

import torch
import torch.nn as nn
import torch.nn.functional as F

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.fc = nn.Linear(1, 3)
    
    def forward(self, x):
        x = self.fc(x)
        x = F.relu(x)
        return x

The network can be paralleled to 4 gpus in the following way:

from easycore.common.config import CfgNode
from easycore.common.parallel import OrderedRunner
from network import Net
import torch

class Predictor(OrderedRunner):
    @staticmethod
    def producer_init(device, cfg):
        cfg.model = Net()     # init the producer with a model
        cfg.model.to(device)  # transfer the model to certain device

    @staticmethod
    def producer_work(device, cfg, data):
        with torch.no_grad():
            data = torch.Tensor([[data]])  # preprocess data
            data = data.to(device)  # transfer data to certain device
            output = cfg.model(data)  # predict
            output = output.cpu()  # transfer result to cpu
        return output

    @staticmethod
    def producer_end(device, cfg):
        del cfg.model  # delete the model when all data has been predicted.

    @staticmethod
    def consumer_init(cfg):
        cfg.data_list = []  # prepare a list to store all data from producers.

    @staticmethod
    def consumer_work(cfg, data):
        cfg.data_list.append(data)  # store data from producers.

    @staticmethod
    def consumer_end(cfg):
        data = torch.cat(cfg.data_list, dim=0)  # postprocess data.
        return data

if __name__ == '__main__':
    predictor = Predictor(devices=["cuda:0", "cuda:1", "cuda:2", "cuda:3"])  # init a parallel predictor

    data_list = list(range(100))  # prepare data
    result = predictor(data_list)  # predict
    print(result.shape)

    predictor.close()  # close the predictor when you no longer need it.

Example 3: Process data with batch

You can use a simple generator or pytorch dataloader to generate batch data.

from easycore.common.config import CfgNode
from easycore.torch.parallel import OrderedRunner
from network import Net
import torch

def batch_generator(data_list, batch_size):
    for i in range(0, len(data_list), batch_size):
        data_batch = data_list[i : i+batch_size]
        yield data_batch

class Predictor(OrderedRunner):

    @staticmethod
    def producer_init(device, cfg):
        cfg.model = Net()
        cfg.model.to(device)

    @staticmethod
    def producer_work(device, cfg, data):
        with torch.no_grad():
            data = torch.Tensor(data).view(-1,1)
            data = data.to(device)
            output = cfg.model(data)
            output = output.cpu()
        return output

    @staticmethod
    def producer_end(device, cfg):
        del cfg.model

    @staticmethod
    def consumer_init(cfg):
        cfg.data_list = []

    @staticmethod
    def consumer_work(cfg, data):
        cfg.data_list.append(data)

    @staticmethod
    def consumer_end(cfg):
        data = torch.cat(cfg.data_list, dim=0)
        return data

if __name__ == '__main__':
    predictor = Rredictor(devices=["cuda:0", "cuda:1"])
    
    data_list = list(range(100))
    result = predictor(batch_generator(data_list, batch_size=10))  
    print(result.shape)

    predictor.close()

Here, we replace easycore.common.parallel with easycore.torch.parallel. easycore.torch.parallel has the same API with easycore.common.parallel but use torch.multiprocessing library instead of multiprocessing library.

Example 4: Transfer outside parameters into Runner

You can transfer parameters into runner through cfg parameter. cfg is a easycore.common.config.CfgNode. See tutorial “Light weight config tools” for how to use it.

We use “sum of power” as an example:

from easycore.common.config import CfgNode as CN
from easycore.common.parallel import UnorderedRunner

class Runner(UnorderedRunner):
    @staticmethod
    def producer_work(device, cfg, data):
        return data ** cfg.exponent  # calculate power of data with outside parameter "exponent".

    @staticmethod
    def consumer_init(cfg):
        cfg.sum = 0  # init a sum variable with 0, you can use cfg to transfer data

    @staticmethod
    def consumer_work(cfg, data):
        cfg.sum += data  # add the square to the sum variable

    @staticmethod
    def consumer_end(cfg):
        return cfg.sum  # return the result you need

if __name__ == '__main__':
    # set parameters outside.
    cfg = CN()
    cfg.exponent = 3

    runner = Runner(devices=3, cfg=cfg)  # transfer `cfg` into the runner 
    
    data_list = list(range(100))
    result = runner(data_list)
    print(result)

    runner.close()

Register Mechanism

easycore make it easy to register an object with name, and get it later.

Create a registry

MODEL_REGISTRY = Registry("MODEL")

Register an object with its __name__

@MODEL_REGISTRY.register()
class ResNet50:
    pass

# or

MODEL_REGISTRY.register(obj=ResNet50)

Register an object with a given name

@MODEL_REGISTRY.register("resnet")
class RestNet50:
    pass

# or

MODEL_REGISTRY.register("resnet", ResNet50)

Get a registered object from registry

model_class = MODEL_REGISTRY.get("ResNet50")

# or

model_class = MODEL_REGISTRY.get("resnet")

API Documentation

API Documentation

easycore.common

easycore.common.config

class easycore.common.config.CfgNode(init_dict: dict = None, copy=True)[source]

Bases: dict

Config Node

__init__(init_dict: dict = None, copy=True)[source]
Parameters:
  • init_dict (dict) – a possibly-nested dictionary to initialize the CfgNode.
  • copy (bool) – if this option is set to False, the CfgNode instance will share the value with the init_dict, otherwise the contents of init_dict will be deepcopied.
freeze(frozen: bool = True)[source]

freeze or unfreeze the CfgNode and all of its children

Parameters:frozen (bool) – freeze or unfreeze the config
is_frozen()[source]

get the state of the config.

Returns:bool – whether the config tree is frozen.
copy()[source]

deepcopy this CfgNode

Returns:CfgNode
merge(cfg)[source]

merge another CfgNode into this CfgNode, the another CfgNode will override this CfgNode.

Parameters:cfg (CfgNode) –
save(save_path, encoding='utf-8')[source]

save the CfgNode into a yaml file

Parameters:save_path –
classmethod open(file, encoding='utf-8')[source]

load a CfgNode from file.

Parameters:
  • file (io.IOBase or str) – file object or path to the yaml file.
  • encoding (str) –
Returns:

CfgNode
classmethod load(yaml_str: str)[source]

load a CfgNode from a string of yaml format

Parameters:yaml_str (str) –
Returns:CfgNode
classmethod dump(cfg, stream=None, encoding=None, **kwargs)[source]

dump CfgNode into yaml str or yaml file

Note

if stream option is set to non-None object, the CfgNode will be dumpped into stream and return None, if stream option is not given or set to None, return a string instead.

Parameters:
Returns:

None or str
dict()[source]

convert to a dict

Returns:dict
__str__()[source]
Returns:str – a str of dict format
class easycore.common.config.HierarchicalCfgNode[source]

Bases: object

Config Node help class for open yaml file that depends on another yaml file.

You can specify the dependency between yaml files with __BASE__ tag.

Example

We can load yaml file example-A.yaml which depends on example-B.yaml in the following way.

example-A.yaml :

__BASE__: ./example-B.yaml
A: in example-A.yaml
C: in example-A.yaml

example-B.yaml :

A: in example-B.yaml
B: in example-B.yaml

Now, you can open example-A.yaml:

>>> import easycore.common.config import HierarchicalCfgNode
>>> cfg = HierarchicalCfgNode.open("./example-A.yaml")
>>> print(cfg)
{"A" : "in example-A.yaml", "B" : "in example-B.yaml", "C" : "in example-A.yaml"}

Attributes in example-A.yaml will cover attributes in example-B.yaml.

Note

__BASE__ can be an absolute path or a path relative to the yaml file. And it will be first considered as a path relative to the yaml file then an absolute path.

classmethod open(file, encoding='utf-8')[source]

load a CfgNode from file.

Parameters:
  • file (str) – path to the yaml file.
  • encoding (str) –
Returns:

CfgNode
classmethod save(cfg, save_path, base_cfg_path=None, base_path_relative=True, encoding='utf-8')[source]

save the CfgNode into a yaml file

Parameters:
  • cfg (CfgNode) –
  • save_path (str) –
  • base_cfg_path (str) – if not specified, it behavior like cfg.save(save_path, encoding).
  • base_path_relative (bool) – whether to set base cfg path to a path relative to the save_path.
  • encoding (str) –

easycore.common.parallel

class easycore.common.parallel.BaseRunner(devices, cfg={}, queue_scale=3.0)[source]

Bases: object

A Multi-process runner whose consumer receive data in unorder. The runner will start multi-processes for producers and 1 thread for consumer.

__init__(devices, cfg={}, queue_scale=3.0)[source]
Parameters:
  • devices (int or Iterable) – If the devices is int, it will use devices cpu to do the work. If the devices is an iterable object, such as list, it will use the devices specified by the iterable object, such as [“cpu”, “cuda:0”, “cuda:1”].
  • cfg (easycore.common.config.CfgNode) – user custom data.
  • queue_scale (float) – scale the queues for communication between processes.
is_activate

whether the runner is alive.

static producer_init(device, cfg)[source]

function for producer initialization.

Parameters:
  • device (str) – device for the this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to transfer data to producer_work and producer_end function.
static producer_work(device, cfg, data)[source]

function specify how the producer processes the data.

Parameters:
  • device (str) – device for this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from producer_init function and transfer data to the next producer_work and producer_end function.
  • data (Any) – data get from input of __call__ method.
Returns:

Any – processed data
static producer_end(device, cfg)[source]

function after finishing all of its task and before close the process.

Parameters:
  • device (str) – device for this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from producer_init and producer_work function.
static consumer_init(cfg)[source]

function for consumer initialization.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it to transfer data to consumer_work and consumer_end function.
static consumer_work(cfg, data)[source]

function specify how the consumer processses the data from producers.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from consumer_init function and transfer data to the next consumer_work and consumer_end function.
static consumer_end(cfg)[source]

function after receiving all data from producers.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it get data from consumer_work function.
Returns:Any – processed data
__call__(data_iter)[source]
Parameters:data_iter (Iterable) – iterator of data
Returns:Any – result
close()[source]

Shutdown all processes if this runner is alive.

activate()[source]

Restart all processes if this runner is closed.

class easycore.common.parallel.UnorderedRunner(devices, cfg={}, queue_scale=3.0)[source]

Bases: easycore.common.parallel.engine.BaseRunner

A Multi-process runner whose consumer receive data in unorder. The runner will start multi-processes for producers and 1 thread for consumer.

__init__(devices, cfg={}, queue_scale=3.0)[source]
Parameters:
  • devices (int or Iterable) – If the devices is int, it will use devices cpu to do the work. If the devices is an iterable object, such as list, it will use the devices specified by the iterable object, such as [“cpu”, “cuda:0”, “cuda:1”].
  • cfg (easycore.common.config.CfgNode) – user custom data.
  • queue_scale (float) – scale the queues for communication between processes.
__call__(data_iter)
Parameters:data_iter (Iterable) – iterator of data
Returns:Any – result
activate()

Restart all processes if this runner is closed.

close()

Shutdown all processes if this runner is alive.

static consumer_end(cfg)

function after receiving all data from producers.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it get data from consumer_work function.
Returns:Any – processed data
static consumer_init(cfg)

function for consumer initialization.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it to transfer data to consumer_work and consumer_end function.
static consumer_work(cfg, data)

function specify how the consumer processses the data from producers.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from consumer_init function and transfer data to the next consumer_work and consumer_end function.
is_activate

whether the runner is alive.

static producer_end(device, cfg)

function after finishing all of its task and before close the process.

Parameters:
  • device (str) – device for this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from producer_init and producer_work function.
static producer_init(device, cfg)

function for producer initialization.

Parameters:
  • device (str) – device for the this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to transfer data to producer_work and producer_end function.
static producer_work(device, cfg, data)

function specify how the producer processes the data.

Parameters:
  • device (str) – device for this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from producer_init function and transfer data to the next producer_work and producer_end function.
  • data (Any) – data get from input of __call__ method.
Returns:

Any – processed data
class easycore.common.parallel.OrderedRunner(devices, cfg={}, queue_scale=3.0)[source]

Bases: easycore.common.parallel.engine.BaseRunner

A Multi-process runner whose consumer receive data in order. The runner will start multi-processes for producers and 1 thread for consumer.

__init__(devices, cfg={}, queue_scale=3.0)[source]
Parameters:
  • devices (int or Iterable) – If the devices is int, it will use devices cpu to do the work. If the devices is an iterable object, such as list, it will use the devices specified by the iterable object, such as [“cpu”, “cuda:0”, “cuda:1”].
  • cfg (easycore.common.config.CfgNode) – user custom data.
  • queue_scale (float) – scale the queues for communication between processes.
close()[source]

Shutdown all processes if this runner is alive.

activate()[source]

Restart all processes if this runner is closed.

__call__(data_iter)
Parameters:data_iter (Iterable) – iterator of data
Returns:Any – result
static consumer_end(cfg)

function after receiving all data from producers.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it get data from consumer_work function.
Returns:Any – processed data
static consumer_init(cfg)

function for consumer initialization.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it to transfer data to consumer_work and consumer_end function.
static consumer_work(cfg, data)

function specify how the consumer processses the data from producers.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from consumer_init function and transfer data to the next consumer_work and consumer_end function.
is_activate

whether the runner is alive.

static producer_end(device, cfg)

function after finishing all of its task and before close the process.

Parameters:
  • device (str) – device for this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from producer_init and producer_work function.
static producer_init(device, cfg)

function for producer initialization.

Parameters:
  • device (str) – device for the this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to transfer data to producer_work and producer_end function.
static producer_work(device, cfg, data)

function specify how the producer processes the data.

Parameters:
  • device (str) – device for this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from producer_init function and transfer data to the next producer_work and producer_end function.
  • data (Any) – data get from input of __call__ method.
Returns:

Any – processed data

easycore.common.registry

class easycore.common.registry.Registry(name: str)[source]

Bases: object

The registry that provides name -> object mapping.

To create a registry:

MODEL_REGISTRY = Registry("MODEL")

To register an object with its __name__:

@MODEL_REGISTRY.register()
class ResNet50:
    pass

# or

MODEL_REGISTRY.register(obj=ResNet50)

To register an object with a given name:

@MODEL_REGISTRY.register("resnet")
class RestNet50:
    pass

# or

MODEL_REGISTRY.register("resnet", ResNet50)

To get a registered object from registry:

model_class = MODEL_REGISTRY.get("ResNet50")

# or

model_class = MODEL_REGISTRY.get("resnet")
__init__(name: str) → None[source]
Parameters:name (str) – name of this registry
register(name: str = None, obj: object = None) → Optional[object][source]

Register the given object with given name. If the object is not given, it will act as a decorator.

Parameters:
  • name (str, optional) – if not given, it will use obj.__name__ as the name.
  • obj (object, optional) – if not given, this method will return a decorator.
Returns:

Optional[object] – None or a decorator.
unregister(name: str) → None[source]

Remove registered object.

Parameters:name (str) – registered name
is_registered(name)[source]

Get whether the given name has been registered.

Parameters:name (str) –
Returns:bool – whether the name has been registered.
get(name: str) → object[source]

Get a registered object from registry by its name.

Parameters:name (str) – registered name.
Returns:object – registered object.
registered_names() → List[str][source]

Get all registered names.

Returns:list[str] – list of registered names.

easycore.torch

easycore.torch.parallel

class easycore.torch.parallel.BaseRunner(devices, cfg={}, queue_scale=3.0)[source]

Bases: object

A Multi-process runner whose consumer receive data in unorder. The runner will start multi-processes for producers and 1 thread for consumer.

__init__(devices, cfg={}, queue_scale=3.0)[source]
Parameters:
  • devices (int or Iterable) – If the devices is int, it will use devices cpu to do the work. If the devices is an iterable object, such as list, it will use the devices specified by the iterable object, such as [“cpu”, “cuda:0”, “cuda:1”].
  • cfg (easycore.common.config.CfgNode) – user custom data.
  • queue_scale (float) – scale the queues for communication between processes.
is_activate

whether the runner is alive.

static producer_init(device, cfg)[source]

function for producer initialization.

Parameters:
  • device (str) – device for the this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to transfer data to producer_work and producer_end function.
static producer_work(device, cfg, data)[source]

function specify how the producer processes the data.

Parameters:
  • device (str) – device for this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from producer_init function and transfer data to the next producer_work and producer_end function.
  • data (Any) – data get from input of __call__ method.
Returns:

Any – processed data
static producer_end(device, cfg)[source]

function after finishing all of its task and before close the process.

Parameters:
  • device (str) – device for this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from producer_init and producer_work function.
static consumer_init(cfg)[source]

function for consumer initialization.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it to transfer data to consumer_work and consumer_end function.
static consumer_work(cfg, data)[source]

function specify how the consumer processses the data from producers.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from consumer_init function and transfer data to the next consumer_work and consumer_end function.
static consumer_end(cfg)[source]

function after receiving all data from producers.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it get data from consumer_work function.
Returns:Any – processed data
__call__(data_iter)[source]
Parameters:data_iter (Iterable) – iterator of data
Returns:Any – result
close()[source]

Shutdown all processes if this runner is alive.

activate()[source]

Restart all processes if this runner is closed.

class easycore.torch.parallel.UnorderedRunner(devices, cfg={}, queue_scale=3.0)[source]

Bases: easycore.torch.parallel.engine.BaseRunner

A Multi-process runner whose consumer receive data in unorder. The runner will start multi-processes for producers and 1 thread for consumer.

__init__(devices, cfg={}, queue_scale=3.0)[source]
Parameters:
  • devices (int or Iterable) – If the devices is int, it will use devices cpu to do the work. If the devices is an iterable object, such as list, it will use the devices specified by the iterable object, such as [“cpu”, “cuda:0”, “cuda:1”].
  • cfg (easycore.common.config.CfgNode) – user custom data.
  • queue_scale (float) – scale the queues for communication between processes.
__call__(data_iter)
Parameters:data_iter (Iterable) – iterator of data
Returns:Any – result
activate()

Restart all processes if this runner is closed.

close()

Shutdown all processes if this runner is alive.

static consumer_end(cfg)

function after receiving all data from producers.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it get data from consumer_work function.
Returns:Any – processed data
static consumer_init(cfg)

function for consumer initialization.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it to transfer data to consumer_work and consumer_end function.
static consumer_work(cfg, data)

function specify how the consumer processses the data from producers.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from consumer_init function and transfer data to the next consumer_work and consumer_end function.
is_activate

whether the runner is alive.

static producer_end(device, cfg)

function after finishing all of its task and before close the process.

Parameters:
  • device (str) – device for this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from producer_init and producer_work function.
static producer_init(device, cfg)

function for producer initialization.

Parameters:
  • device (str) – device for the this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to transfer data to producer_work and producer_end function.
static producer_work(device, cfg, data)

function specify how the producer processes the data.

Parameters:
  • device (str) – device for this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from producer_init function and transfer data to the next producer_work and producer_end function.
  • data (Any) – data get from input of __call__ method.
Returns:

Any – processed data
class easycore.torch.parallel.OrderedRunner(devices, cfg={}, queue_scale=3.0)[source]

Bases: easycore.torch.parallel.engine.BaseRunner

A Multi-process runner whose consumer receive data in order. The runner will start multi-processes for producers and 1 thread for consumer.

__init__(devices, cfg={}, queue_scale=3.0)[source]
Parameters:
  • devices (int or Iterable) – If the devices is int, it will use devices cpu to do the work. If the devices is an iterable object, such as list, it will use the devices specified by the iterable object, such as [“cpu”, “cuda:0”, “cuda:1”].
  • cfg (easycore.common.config.CfgNode) – user custom data.
  • queue_scale (float) – scale the queues for communication between processes.
close()[source]

Shutdown all processes if this runner is alive.

activate()[source]

Restart all processes if this runner is closed.

__call__(data_iter)
Parameters:data_iter (Iterable) – iterator of data
Returns:Any – result
static consumer_end(cfg)

function after receiving all data from producers.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it get data from consumer_work function.
Returns:Any – processed data
static consumer_init(cfg)

function for consumer initialization.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it to transfer data to consumer_work and consumer_end function.
static consumer_work(cfg, data)

function specify how the consumer processses the data from producers.

Parameters:cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from consumer_init function and transfer data to the next consumer_work and consumer_end function.
is_activate

whether the runner is alive.

static producer_end(device, cfg)

function after finishing all of its task and before close the process.

Parameters:
  • device (str) – device for this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from producer_init and producer_work function.
static producer_init(device, cfg)

function for producer initialization.

Parameters:
  • device (str) – device for the this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to transfer data to producer_work and producer_end function.
static producer_work(device, cfg, data)

function specify how the producer processes the data.

Parameters:
  • device (str) – device for this process.
  • cfg (easycore.common.config.CfgNode) – config of this process, you can use it to get data from producer_init function and transfer data to the next producer_work and producer_end function.
  • data (Any) – data get from input of __call__ method.
Returns:

Any – processed data

Indices and tables