How much GPU memory does Vistral-7B-Chat need?

About 19.7 GB for conversations of up to 8,192 tokens: 14.59 GB of weights, the cache for one full-length conversation and the runtime's own memory (an estimate: its settings are behind its licence).

What is the cheapest GPU to run Vistral-7B-Chat?

Right now one RTX 4000 Ada at $0.22 an hour; RTX 5090 at $0.67 an hour is the best value for its speed. Prices are live offers from GPU clouds, checked every 15 minutes.

Can Vistral-7B-Chat run on an RTX 4090?

Yes: it needs about 19.7 GB and an RTX 4090 has 24 GB.

Models / Vistral

Viet-Mistral/Vistral-7B-Chat

Text generation · served with vLLM · ↓ 1.1K downloads a month · View on Hugging Face

Viet-Mistral/Vistral-7B-Chat needs about 19.7 GB of GPU memory for conversations of up to 8,192 tokens (an estimate: its settings are behind its licence). It runs on one RTX 4000 Ada from $0.22/h right now; best value: RTX 5090 at $0.67/h.

The model's publisher asks users to accept its licence (afl-3.0). GridPup cannot download this one yet: choose an ungated version below.

Weights: 15 GB
GPU memory needed: about 19.7 GB (estimate)
Conversation length: 8,192 tokens
Memory · disk: 16 GB · 57 GB
Architecture: MistralForCausalLM
Licence: afl-3.0

Hardware and deploy

The publisher has not let GridPup download this licence-gated model yet. An ungated version of it (above) can be deployed.

Questions

How much GPU memory does Vistral-7B-Chat need?: About 19.7 GB for conversations of up to 8,192 tokens: 14.59 GB of weights, the cache for one full-length conversation and the runtime's own memory (an estimate: its settings are behind its licence).
What is the cheapest GPU to run Vistral-7B-Chat?: Right now one RTX 4000 Ada at $0.22 an hour; RTX 5090 at $0.67 an hour is the best value for its speed. Prices are live offers from GPU clouds, checked every 15 minutes.
Can Vistral-7B-Chat run on an RTX 4090?: Yes: it needs about 19.7 GB and an RTX 4090 has 24 GB.

Loading…