A harness project for ML & DL
Project description
AI Harness
Introduction
This project would like to supply some convenient tools for the machine learning and deep learning. Current features:
- XMLConfiguration: for loading a configuration defined in xml files into a Python Object
- Arguments: Mapping a Python Object to the arguments of argparse
- inspector: Some convenient method for class/object
- transformer:
- others:
Examples
XMLConfiguration Example
(1) Define the configuration in xml file like:
<?xml version="1.0" encoding="UTF-8" ?>
<configuration>
<arg name="name" default="TestName" help="Name Test Help"/>
<arg name="age" default="20" help="Age Test Help"/>
<group name="address">
<arg name="home" default="shanghai" help="Home test Help"/>
<arg name="phone" default="136" help="Phone test Help"/>
</group>
<group name="education">
<arg name="school" default="beijing" help="school test Help"/>
<arg name="grade" default="master" help="grade test Help"/>
</group>
</configuration>
you can define multiple xml configuration files, and if the name is same, the value of the later will cover the previous.
(3) Define the configuration class like:
@dataclass
class Address:
phone: ArgType = ArgType(139, "phone help")
home: ArgType = ArgType("beijing", "phone help")
@dataclass
class Education:
school: ArgType = ArgType('ustb', "phone help")
grade: ArgType = ArgType("master", "phone help")
@dataclass
class Config:
name: ArgType = ArgType("test", "name help")
age: ArgType = ArgType(10, "age help")
address: Address = Address()
education: Education = Education()
(3) Load the xml configuration into python object as folling:
from aiharness.configuration import XmlConfiguration
config:Config=XmlConfiguration(Config).load(['configuration1.xml','configuration2.xml'])
Arguments Example
Generally, we use argparse as following:
import argparse
parser = argparse.ArgumentParser()
parser.add_argument("--name",default='TEST',help='name help')
parser.add_argument("--age",default=18,help='age help')
arguments=parser.parse_args()
And you can got a arguments object.
Here give an example showing how to load a xml configuration and set to argparse arguments and to parse the arguments into a object you defined. And here the Config Class and 'configuration.xml' are same with those of the Configuration example.
Firstly, in fact, the Config Class instead of the codes of 'add_argument' of the argparse.ArgumentParser. Secondly, you can put the configuration into a xml file so that you can change it conveniently.
from aiharness.configuration import ArgType, Arguments, XmlConfiguration
config: Config = XmlConfiguration(Config).load(['configuration.xml'])
arguments = Arguments(config)
config: Config = arguments.parse()
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for ai_harness-0.2.0-py2.py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | cfc3f4139822b81a1d63fcdd65e53358ae7befb9ea797afd296e7c5b34ea4a81 |
|
MD5 | 020b2dff5794739ea70644821a6cfdf6 |
|
BLAKE2b-256 | 23bbe54fbc62cbfc57594d4957f3a3b0b4302edf9e132afc42f955191c3e370d |