LLaMA-Factory
LLaMA-Factory copied to clipboard
llama3.1 405B fp8 support
Reminder
- [X] I have read the README and searched the existing issues.
System Info
n/a
Reproduction
n/a
Expected behavior
No response
Others
Hi, I see the latest master has supported the 405B, but it needs at least 2 nodes to do the inference, how do we do that? Also Is it possible to support 405B fp8 so it can be run within 1 A100 node?