Skip to main content

Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
TagsGame Engines
(1 edit)

Is there a possibility to use locally gpt-oss with this game? I tried it sends requests, gets answer and then just gives error. I use llama.cpp, tried using 

--reasoning-format deepseek

nothing worked for me, if someone manages to run it, can explain?

Edit: I made it work but reasoning gets over talking limit, you can use this model but not normally, you need to manually set max tokens in response, but the issue that only reasoning takes all of tokens for ingame response. I hope devs will add feature to skip <reasoning> and go straight to talking

Try to use models without reasoning for now. 
QwQ-LCoT-7B-Instruct-Q4_K_S is a good and fast alternative for the Deep Seek