App crash when running certain 1B models about maid HOT 7 CLOSED

I could be wrong but i think that models a GGUFV1 model which llama.cpp dropped about a month ago. Ive been trying to keep the functionality working but its pretty unstable

from maid.

there seems to be a lot wrong with my implementation of llama.cpp at the moment. Ill be rewriting it in the future but that's why certain models don't work.

this issue was effectively a duplicate of #185 so you're free to close it and discuss anything further in that issue

from maid.

I could be wrong but i think that models a GGUFV1 model which llama.cpp dropped about a month ago. Ive been trying to keep the functionality working but its pretty unstable

Oh. Which version of GGUF is compatible currently?

from maid.

I could be wrong but i think that models a GGUFV1 model which llama.cpp dropped about a month ago. Ive been trying to keep the functionality working but its pretty unstable

Oh. Which version of GGUF is compatible currently?

Well both are ment to be compatible but sometime v1 models can cause issues.

Give the latest actions build a try, your issue might be fixed

from maid.

@RookieIndieDev This build should work:

https://github.com/MaidFoundation/maid/actions/runs/7127583201

from maid.

@danemadsen Is the release published on github? Can't seem to get the files from the action artifacts.

from maid.

The issue is resolved. Some models do crash still. The tinyllama-2-1b-miniguanaco.Q2_K crashes but Q3_K_L doesn't. I'll close the issue, might be an issue with the model itself. Might be an even older version of GGUF for that specific one.

from maid.