mirror of
https://github.com/ParisNeo/lollms-webui.git
synced 2024-12-27 23:58:51 +00:00
17 lines
1.8 KiB
Markdown
17 lines
1.8 KiB
Markdown
|
Objective: Finally build a tutorial on how to use lollms. This is a very searched thing so we need to neil the video.
|
||
|
We don't show how to download and install the tool as this was already been covered by previous
|
||
|
We will show:
|
||
|
1- basic usage :
|
||
|
1.1 discussions left panel which is composed of a top toolbar and a list of discussions and at the bottom of the panel, threre is the name of the current discussions database (we can change databases, export them, share them, etc):
|
||
|
- create a new discussion (press + from the panel toolbox) this will add a new discussion to the list with the name New discussion and we are welcomed by the current personality welcome message.
|
||
|
- select discussions by pressing the check button from the toolbox which will show a checkbox in front of each entry in the list of discussions. We can select one , multiple or press the toggle select all. Once selected, we can export it as json, as a zip folder, or as a markdown.
|
||
|
-delete the database
|
||
|
-import discussions from a json file
|
||
|
-search discussions by name
|
||
|
- Add discussion to skills library which reads through the discussion and extracts all important information that can be valuable for long term preservation
|
||
|
- activate skills library usage in LLM (that allows the AI to use RAG on the skills library)
|
||
|
-
|
||
|
|
||
|
|
||
|
interacting with AI, interacting with AI using the web enabled mode, using audio input, sending a file for rag, taking a shot with the webcam, adding a link to a webpage, creating a complex prompt while using markdown options on message ui that allow to create different kind of code tags, creating AI response (useful to train ai by building a database), forcing the AI to start answering in a specific manner to make it more aligned.
|
||
|
- edit message contents, copy message content, , change AI message content (useful for training), add thumb up and thumbdown to AI outputs,
|