Python Boilerplate contains all the boilerplate you need to create a Python package.
Project description
deepINC 增量学习工具包
增量学习工具库,实现了常用数据集,典型模型,工具函数。支撑综述《基于深度神经网络的增量学习技术研究现状综述》的实验。
开源许可: MIT license
开始使用
准备使用增量学习工具库? 按如下步骤本地使用 deepINC .
使用 pip 安装 deepINC:
$ pip install deepinc
在 .py 程序中引用:
$ import deepinc as di $ from deepinc.utils.args import get_args $ from deepinc.utils.training import train_il
初始化配置项:
$ args = get_args() $ $ args.model = 'lwf' # 模型名称 $ args.dataset = 'seq-mnist' # 数据集 $ args.print_freq = 1 # 报告loss的频率 $ args.n_epochs = 1 # 迭代次数 $ args.device = 'cpu' # 设备 $ $ args.lr = 0.02 # 学习率 $ args.batch_size = 64 # batch size $ args.alpha = 5 # 蒸馏损失权重 $ args.softmax_temp = 2.0 # 蒸馏损失温度系数 $ args.wd_reg = 5e-5 # 正则项权重
这里可以随心更改
增量训练:
$ train_il(args)
等待训练结束~
关联代码
History
0.1.0 (2022-09-22)
First release on PyPI.
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
deepinc-0.1.0.tar.gz
(53.7 kB
view details)
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
deepinc-0.1.0-py2.py3-none-any.whl
(154.5 kB
view details)
File details
Details for the file deepinc-0.1.0.tar.gz.
File metadata
- Download URL: deepinc-0.1.0.tar.gz
- Upload date:
- Size: 53.7 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.1 CPython/3.8.8
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
8b22f73484c8083f404201c708da1187113a84712bfd4026d4e981f6269a4463
|
|
| MD5 |
19f82b3caf03fa60d8c52572fe3f082c
|
|
| BLAKE2b-256 |
66ec80323ebc35ed8df642a00d553592c6302080728a46074812338a1639fa1d
|
File details
Details for the file deepinc-0.1.0-py2.py3-none-any.whl.
File metadata
- Download URL: deepinc-0.1.0-py2.py3-none-any.whl
- Upload date:
- Size: 154.5 kB
- Tags: Python 2, Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.1 CPython/3.8.8
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
196e66969470b18d2e40bbea1950f905497a05c86e0858fa003ded02b7557c4f
|
|
| MD5 |
ba8b501941fa89599fe4b6a7532cd736
|
|
| BLAKE2b-256 |
ccb769d17675a12bef4366c45d3cd902335daa9aaea3143ee2754a40bcd228ad
|