Hi, thank you for your report.
I see a few problems here.
On your example you try to load Shinen model. But it requires 16GB of ram by itself. It means you need 16GB for the AI + X for your OS + ~500MB for the game. You might be running out of ram.
Same thing for the GPU. This error is triggered when your GPU runs out of RAM. I suggest starting by a small AI to tweak your options.
Here are a few tweaks to improve the generation.
In the options:
- Increase the max time from 10(seconds) to 120
- You might want to reduce the context max length and memory max length too (as they use CPU too)
- Make sure
remove unfinished sentencesis not check for your tests.
For the weird ERROR: WASAPI: GetCurrentPadding error I have no idea. It’s the first time I see them. It seems to be some errors related to audio drivers. The game doesn’t have any audio yet though.
I hope this helps!