Prevtenet is close to completing another exceptional application for SymbOS by using his powerful SCC development kit (SymbOS C Compiler) (https://github.com/danielgaskell/scc):
"LLM Chat" makes ChatGPT a reality on the CPC. It allows you to chat using a local or web-based "Large Language Model," the name for the AI category that has brought us a new revolution over the past 2.5 years.
lmmchat1.png
lmmchat2.png
LLM Chat works on any SymbOS platform with network hardware, including the CPC with the M4Board, RSF3, or Net4CPC hardware. It supports the OpenAI standard interface and is therefore compatible with virtually all LLMs.
A short video by EdoZ:
https://www.youtube.com/watch?v=KypJzt_pHOs
llmchat3.jpg
Web-based LLMs require a proxy that converts HTTPS to HTTP, as this would otherwise be too time-consuming for a Z80.
LLMs on the local network are also supported, which can be installed on your own PC using "LM Studio," for example.
LLMChat is currently in beta phase, but already works very well.
As soon as it is released, I will post the download link here!
Hahaha that's awesome!
thats cool AF!
I wonder if it can be used with the local running LLM GAIA by AMD (GAIA) (https://github.com/amd/gaia/).
It works on every computers (one installer for a generic installation or if you use an AMD 300 series, it will install a specific instation using its AI chip).
Ask it to compare the CPC and the C64 and see what happens... ;D
Time to connect it to Skynet ..... ;D
(https://user.fm/files/v2-095a8808cae0c6a60bd6d585d62b3b87/amstrad-skynet.png)