Skip to main content

Atiny is a gradient based PyTorch optimizer. Atiny uses half of Adam's memory to achieve results that are not inferior to Adam.

Project description

Atiny Optimizer

Atiny is a gradient based PyTorch optimizer. Atiny uses half of Adam's memory to achieve results that are not inferior to Adam.

image image

思路说明

不使用二阶动量,而是使用一阶动量与当前梯度计算每个参数的更新补偿. 内置了一个学习率衰减器,自动为每组参数个性化的学习率衰减.(需要参数:ldr) 使用ArcSinh函数影响向量模长实现权重衰减

Demo程序

demo程序在test.py文件中,直接运行即可得到此页面中Atiny与Adam的对比图.(依赖visdom) demo中构建了一个神经网络用简陋的方式对一个包含动态随机参数的公式生成的周期性曲线进行预测.

Install

pip install Atiny

Use

from Atiny import Atiny
...
optimizer=Atiny(moduel.parameters(),lr=lr,ldr=ldr,weight_decay=weight_decay)

HomePage

https://github.com/PsycheHalo/Atiny/

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

atiny-0.0.4.tar.gz (10.4 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

atiny-0.0.4-py3-none-any.whl (10.8 kB view details)

Uploaded Python 3

File details

Details for the file atiny-0.0.4.tar.gz.

File metadata

  • Download URL: atiny-0.0.4.tar.gz
  • Upload date:
  • Size: 10.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.12.9

File hashes

Hashes for atiny-0.0.4.tar.gz
Algorithm Hash digest
SHA256 226058e46303e54b0c31dd48537e8f55d29e84966ad6cc8fe5eb87fed24533de
MD5 2ffdcb751adedb078ff70067fdcd9fa6
BLAKE2b-256 0c5a510b6647be8ccc0b0df690e26e6997d3a27ae726324bae0e413343305df0

See more details on using hashes here.

Provenance

The following attestation bundles were made for atiny-0.0.4.tar.gz:

Publisher: python-publish.yml on PsycheHalo/Atiny

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file atiny-0.0.4-py3-none-any.whl.

File metadata

  • Download URL: atiny-0.0.4-py3-none-any.whl
  • Upload date:
  • Size: 10.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.12.9

File hashes

Hashes for atiny-0.0.4-py3-none-any.whl
Algorithm Hash digest
SHA256 5016313984544b6b69ebf63d0b2a006130e3ef2ad57206d622e85f9709b4bb0a
MD5 56efbeda6ad7a08bca180c2e164af1e1
BLAKE2b-256 aa0008fdbeb6e177381df890a6882686c25cf54a5ee5b035fdd932a3b6b16d45

See more details on using hashes here.

Provenance

The following attestation bundles were made for atiny-0.0.4-py3-none-any.whl:

Publisher: python-publish.yml on PsycheHalo/Atiny

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page