at first was working, now say error

#2
by micole66 - opened

Hello there,
Sorry about the inconvenience.
I think the error is caused by the model going to sleep: I have set it to sleep after a small duration to prevent high GPU costs. I'll try to raise it in the future if I get a community grant to mitigate the costs.
Another error could also be because of the GPU memory. It seems like it is OOM'ing at some point, I'll try to get more GPU VRAM for the space. Right now I recommend limiting the max token to default 128 tokens to mitigate this.
For a more dedicated version that you can run on your local computer, here is a link to the Github repo: https://github.com/DESU-CLUB/BLLAMA

Alright it's back up!

Thanks for the feedback!

DESUCLUB changed discussion status to closed
DESUCLUB changed discussion status to open
DESUCLUB changed discussion status to closed

Sign up or log in to comment