!NOTE.txt

NOTE: This Experimental don't have own GUI


Keep in mind this project is Experimental


This project was running and developed with Spec Ryzen AI 7 350 equipped with ram 24gig, in laptop (IGPU) 

If you wanna using Learn, You need talk with AI from this script until AI summary, AI would make question for learning it self and you can run learn
(how about online mode?)
online mode its just searching for context, so your AI more adaptive to new information, and you can add question to file question AI make. open LAPAI_Learn_Core_Experimental.py for more detail editing.




This project need model to running, in this case im using Lemonade Server, you can see on their website and follow their tutorial first.
after you install it, you need at least 2 model to run, the first model you can get from lemonade server requirement to install.
https://lemonade-server.ai/docs/server/

but you can also using another model API provider (if you prefer another).

after installing Lemonade server you can download another model, but as i recommend is hybrid model, still keep in mind your PC power can it handle it?

Next go to Microsoft store(If you using windows) -> install python 3.10/3.11

after installing python, run InstallingRequirment.bat
and install embedding model run EmbeddingModelAutoDownload.bat

alright after installing model, python, and requirment, now time to editing on LAPAI_Core.py and LAPAI_Learn_Core_Experimental.py
you can using visual studio code or another,

next model in core:

LAPAI_Core.py 
-> MODEL_NAME 		-> With main model you using
-> Sum_Model 		-> just give light model like Qwen 0.5B or anything like that

LAPAI_Learn_core_Experimental.py 
-> MODEL_NAM		-> you can use model light like phi 3.5 or anything
-> SUM_MODEL_NAME	-> just use qwen for efficient or if you wanna change its up to you

RunAI.py & Learn.py

-> ASLS                 -> set to your lemonade-server location

and done, have fun, klick Run.Bat for run the model
go experiment with AI on your own device freely!