One Peaceful Frontend+Backend Software Library Suite.
This repository facilitates neural network artificial intelligence (AI) related application-based protocol interfaces (API) functions. Its purpose is to house prompt keywords and facilitates semi-automated prompt construction use.
For specific neural network functionalities and low-level construction, please
refer to its type from the Hestia library suites (e.g. HestiaTENSORFLOW
for
TensorFlow).
This project was initiated primarily because of:
- Performs keywords grouping for higher accuracy prompting - certain keywords when grouped can yield a more accurate generative output with higher confidences.
- Semi-automated prompt constructions - Sometimes, one would just want to automate grouped keywords with toggles.
- Friendly to low memory and low footprint environment - usable anywhere anytime.
- Consumption freedom - Makes NO decision and NO assumption on how you consume the grouped keywords.
- Simple to integrate - simple enough to use for all supported programming languages.
The definitions complies to only the following rules:
- STRICTLY: Everthing is a
string
. Prompts are string anyway. - Primarily uses Unicode for internal operations.
- Uses supported API from available vendors both locally and remotely.
This library is heavily guarded with localized unit tests whenever available.
This library supports multiple programming languages for the same dataset. Please import based on your programming language(s) using the instructions in the sub-sections below.
Please do note that branches like main
, next
, and experimental
are for
maintenance & development uses (as in, the production factory itself). Hence,
please avoid them and only use it at your own risk.
This library is supported based on what you're interested and usage. The current URLs would be:
- Prompt Engineering: src/prompt-keywords
- Text2Image: https://perchance.org/clean-text2image
This library depends on the following libraries:
- HestiaSTRING - https://github.com/ChewKeanHo/HestiaSTRING
The libraries are based on the following data sources:
- Perchance - https://perchance.org/generators
This project is licensed under Chew Kean Ho's Hestia License Agreement.