Any plans to make a 13B or 7B version of this?
#4
by
weezywitasneezy
- opened
I absolutely love the work you're doing! I'm dying to run this model locally but running into hardware limitations. Is there any hope of a 13B or 7B version in the near future? So far for story writing I've had the best luck with your Wizard Mega 13B (although I run into CUDA errors after ~4000 characters worth of responses (rtx 3080)) and the Guanaco 7B (haven't had any issues and been able to generate MASSIVE response lengths). Thanks and keep up the awesome work!