Joel Niklaus
Joel Niklaus
Great, thank you very much! Aah yes, thank you. I didn't know this one.
Thanks again for taking the effort to train the model! Would you mind giving me a headsup when it is ready?
Great, I will use that. Thank you very much!
Hi I would like to use your software for a card game. Do you have any recommendations or tips how to adapt it? Cheers Joel [email protected]
I have the same problem
Thanks @dlwh. When switching to two A100 80GB GPUs it worked for me.
For me it also worked with batch size 4 on 2 80GB A100 GPUs for sequence length 512.
Thanks, I don't know when I have the capacity to add it to the other methods.
This might not be necessary anymore with PR #488.
I personally think it would still be nice to have caching here too, but for me it is not strictly necessary anymore I guess.