Skip to content

GitLab

Unverified Commit fdf6d72b authored by Woosuk Kwon's avatar Woosuk Kwon Committed by GitHub
Browse files

Clean up API & Bypass torch.autograd.Function (#30) 


Signed-off-by: default avatarWoosuk Kwon <woosuk.kwon@berkeley.edu>
parent 7d3409be
Pipeline #2013 failed with stages
in 0 seconds
Expand all Hide whitespace changes
Inline Side-by-side
Showing with 27 additions and 920 deletions
vllm_flash_attn/__init__.py
View file @ fdf6d72b
...@@ -3,10 +3,6 @@ __version__ = "2.6.2" ...@@ -3,10 +3,6 @@ __version__ = "2.6.2"
# Use relative import to support build-from-source installation in vLLM # Use relative import to support build-from-source installation in vLLM
from .flash_attn_interface import ( from .flash_attn_interface import (
flash_attn_func, flash_attn_func,
flash_attn_kvpacked_func,
flash_attn_qkvpacked_func,
flash_attn_varlen_func, flash_attn_varlen_func,
flash_attn_varlen_kvpacked_func,
flash_attn_varlen_qkvpacked_func,
flash_attn_with_kvcache, flash_attn_with_kvcache,
) )
vllm_flash_attn/flash_attn_interface.py
View file @ fdf6d72b
This diff is collapsed.
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment