Bundled llm - #1

Open

YVishere wants to merge 18 commits into

Commits on Jan 17, 2026

Adding files to get Started
YVishere
committed
ffa0e61

Commits on Jan 18, 2026

Added llama.cpp suport, a sample file to invoke the model and refined filepath for ServerSocketpython process
YVishere
committed
2583bb7
Added a singleton LLM plan for bundled LLM approach
YVishere
committed
67a5ecd
updated ignore.conf
YVishere
committed
9b10541

Commits on Mar 11, 2026

Basic llm interactions with llama.cpp works
YVishere
committed
854698f
Macro'd legacy TCP code away
YVishere
committed
d5c4848
renamed flag
YVishere
committed
5e74731

Commits on Mar 15, 2026

Wired up npcs to llma.cpp and they store independent history
YVishere
committed
87f7f63
Update documentation
YVishere
committed
18b3f9b
Updated NPC logging
YVishere
committed
9694b4c
Debug flags to understand the LLM better
YVishere
committed
9717488
Editor version update and chat prompt enhancements
YVishere
committed
3918c4b
Logs
YVishere
committed
cfb62ee

Commits on Jul 17, 2026

Made the LLM faster with CUDA
YVishere
committed
9780fc4
Updated unity ignore file
YVishere
committed
5e14cff
Fixed wonky AI prompts
YVishere
committed
24735bf

Commits on Jul 18, 2026

Added robustness to model's generated output and relied less on magic numbers
YVishere
committed
39685da
Fixed a code inconsistency
YVishere
committed
40de0bf