Skip to content

[Misc] Use apply_rotary_emb from vllm_flash_attn for Qwen2-VL vision RoPE #17726

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
May 7, 2025
Merged

[Misc] Use apply_rotary_emb from vllm_flash_attn for Qwen2-VL vision RoPE #17726

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
294e45e
use vllm-fa for qwen2vl vision rope
Isotr0py May 6, 2025
a567ca9
fix import
Isotr0py May 6, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 2 additions & 7 deletions vllm/model_executor/models/qwen2_5_vl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -297,13 +297,8 @@ def forward(
q, k, v = (rearrange(x, "s b ... -> b s ...").contiguous()
for x in (q, k, v))
if rotary_pos_emb is not None:
use_flash_attn = self.attn_backend == _Backend.FLASH_ATTN
q = apply_rotary_pos_emb_vision(q,
rotary_pos_emb,
use_flash_attn=use_flash_attn)
k = apply_rotary_pos_emb_vision(k,
rotary_pos_emb,
use_flash_attn=use_flash_attn)
q = apply_rotary_pos_emb_vision(q, rotary_pos_emb)
k = apply_rotary_pos_emb_vision(k, rotary_pos_emb)

if self.attn_backend == _Backend.FLASH_ATTN:
# from vllm_flash_attn.flash_attn_interface import (
Expand Down
9 changes: 4 additions & 5 deletions vllm/model_executor/models/qwen2_vl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -64,7 +64,7 @@
BaseProcessingInfo, PromptReplacement,
PromptUpdate)
from vllm.multimodal.profiling import BaseDummyInputsBuilder
from vllm.platforms import _Backend
from vllm.platforms import _Backend, current_platform
from vllm.sequence import IntermediateTensors
from vllm.transformers_utils.config import uses_mrope
from vllm.transformers_utils.processor import (
Expand Down Expand Up @@ -230,14 +230,13 @@ def apply_rotary_emb_torch(x: torch.Tensor,


def apply_rotary_pos_emb_vision(t: torch.Tensor,
freqs: torch.Tensor,
use_flash_attn=False) -> torch.Tensor:
freqs: torch.Tensor) -> torch.Tensor:
t_ = t.float()
cos = freqs.cos()
sin = freqs.sin()
apply_rotary_emb = apply_rotary_emb_torch
if use_flash_attn:
from flash_attn.layers.rotary import apply_rotary_emb
if current_platform.is_cuda():
from vllm.vllm_flash_attn.layers.rotary import apply_rotary_emb
output = apply_rotary_emb(t_, cos, sin).type_as(t)
return output

Expand Down