DLextras

版本 0.2 (19.3 KB) 作者: Jeremy Lea
Matlab Deep Learning custom layers and other helpers
0.0
(0)
6.0 次下载
更新时间 2024/6/7
在 GitHub 上查看许可证

Extra Layers and helpers for the Matlab Deep Learning and Reinforcement Learning Toolkits

onehotLayer

Converts integer index values into "one hot" encoded (i.e. indicator) variables within the network so you don't have to do it in the data. Particularly helpful for discrete action spaces when not using the rlVectorQValueFunction.

PositionEncode

The "Attention is all you need" fixed position encoding scheme. Slightly modified to support non-contigous positions.

extractTime

Extract one time from a sequential input, like LSTM layer OutputMode='last', but without the LSTM part. Also modified to allow you select a particular item in the sequence.

引用格式

Jeremy Lea (2024). DLextras (https://github.com/jeremylea/DLextras/releases/tag/v0.2), GitHub. 检索来源 .

MATLAB 版本兼容性
创建方式 R2022b
与 R2022a 及更高版本兼容
平台兼容性
Windows macOS Linux
标签 添加标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

mzflow

mzflow/+bijectors

mzflow/+distributions

版本 已发布 发行说明
0.2

See release notes for this release on GitHub: https://github.com/jeremylea/DLextras/releases/tag/v0.2
0.1

要查看或报告此来自 GitHub 的附加功能中的问题，请访问其 GitHub 仓库。
要查看或报告此来自 GitHub 的附加功能中的问题，请访问其 GitHub 仓库。