File size: 355 Bytes
5472531
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
# Make it more memory efficient by monkey patching the LLaMA model with FlashAttn.

# Need to call this before importing transformers.
from fastchat.train.llama2_flash_attn_monkey_patch import (
    replace_llama_attn_with_flash_attn,
)

replace_llama_attn_with_flash_attn()

from fastchat.train.train import train

if __name__ == "__main__":
    train()