v8.5

Moved to python 3.11
New hardware management system
Upgraded bindings
New services : sd + xtts
Loras support in hugging face and python_llama_cpp bindings

LoLLMS v8.0

• Changed the name
• Added more interactive options
• Added loads of personas
• Added many new Bindings
• Added many new models

V7.0

• Enhanced personality functionalities
• Reorganized structure
• AWQ support
• Multimodality support
• Many new personas
• Interactive view

v6.7

• Added long term memory
• Fixed some bugs
• Exllama 2 support
• Open AI support
• First Extensions implementation (still in beta)

Note:

Homebrew is required on macos

v6.5.0

The first stable release of 6.5.
Petals integration for distributed text generation
A special windows wsl version of lollms for windows users.
For linux/macos, just follow the instructions from the README.md page of the project.

v6.5 RC2

Release candidate 2
Added leaderboard link to the zoo
Added Aristotle to philosophers
Multiple bugfixes
GPT4ALL supports Vulkan for AMD users

Added lollms with petals to use a decentralized text generation on windows over wsl.

v6.5 RC1

A release candidate for lollms V6.5.
For windows use the installer
For other platforms use the scripts from scripts folder in the repository.
https://github.com/ParisNeo/lollms-webui/tree/main/scripts

v6.3

• New UI engine
• Added lollms playground presets system
• Enhanced prompting
• Better intrgration with stable diffusion
• Enhanced speed for bindings

V6.0(beta)

• New personalities structure
• New localization engine
• Incorportated Playground into the main app
• Multiple enhancements in personalities

v4.0.0

Version 4.0

Added json support (personalities can generate json and send it back to ui)
Reworked database organization
Reworked all the engine
Enhanced support for gptq
State machines for personalities (personalities can use the state machine system built in lollms)
File data management (loading and vectorizing pdf, docx, csv, html, txt files)