Text Generation
Transformers
PyTorch
mpt
Composer
MosaicML
llm-foundry
custom_code
text-generation-inference

Question about VRAM requirements

#12
by sebastiaanvandenbroeck - opened

Quick question, would it be possible to run this model on a card with 12GB of memory?

Thanks in advance!

We recommend using bf16, but even with that the model should have a memory footprint of about 14GB (= 7B params * 2 bytes/parameter).
So, I don't think 12GB of VRAM will be enough.

That's unfortunate, but thanks for answering my question!

atrott changed discussion status to closed

Hello, how fast does this model run on enterprise GPUs with triton?

i need to use storywriter with 65k token what are the hardware requirements

Sign up or log in to comment