Im getting a new gpu that will have greater vram than my last one and exceeds the 8gbvram that the LLM the game offers to download requires. Is it possible for me to place an LLM that would use more vram than the ones the game suggests into the file location for the game to use or would the game / kobold not accept it due to modifications you (the developer) needed to make?
Every model needs a specific instruct format and custom sampling settings to work properly, so it's not possible to simply drop in a new one.
I test all major new model releases with this game, and so far only GLM-4-32B-0414 and Gemma 3 27B outperform Nemo, which is the 10 gb/server model. Gemma 3 actively skirts around mature content, so it's not suitable. Both of these models require a GPU with 24 gb of vram. There is no better model than Nemo for 12/16 gb GPUs.
If there's interest in a 24 gb model I will add support for GLM-4.
Thank you for the reply and the clear explanation, i personally would be interested in having a 24gb model option as i will be able to use it once the new gpu arrives(32gb vram) , but i completely understand that it may not be a priority compared to other changes or new requested content. I am hopeful that it can be something added to your 'to do' list but will regardless continue to admire your future updates.
I imagine it has probably been suggested for you to open a patreon or other form of subscription to aid in further development, perhaps even a ko-fi link for donations? please consider it, im sure im not the only one that would like to see the game grow in quality and scope And support it on that journey