Skip to content

Avoid attention_mask copy in qwen2.5 #40658

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 1 commit into from
Sep 3, 2025
+36 −24
Merged

Avoid attention_mask copy in qwen2.5 #40658

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
15 changes: 9 additions & 6 deletions src/transformers/models/qwen2_5_omni/modeling_qwen2_5_omni.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -265,8 +265,8 @@ def get_rope_index(
mrope_position_deltas = []
if input_ids is not None and (image_grid_thw is not None or video_grid_thw is not None):
total_input_ids = input_ids
if attention_mask is None:
attention_mask = torch.ones_like(total_input_ids)
if attention_mask is not None:
attention_mask = attention_mask == 1
position_ids = torch.ones(
3,
input_ids.shape[0],
Expand All @@ -275,9 +275,9 @@ def get_rope_index(
device=input_ids.device,
)
image_idx, video_idx, audio_idx = 0, 0, 0
attention_mask = attention_mask.to(total_input_ids.device)
for i, input_ids in enumerate(total_input_ids):
input_ids = input_ids[attention_mask[i] == 1]
if attention_mask is not None:
input_ids = input_ids[attention_mask[i]]
image_nums, video_nums, audio_nums = 0, 0, 0
vision_start_indices = torch.argwhere(input_ids == vision_start_token_id).squeeze(1)
vision_tokens = input_ids[vision_start_indices + 1]
Expand Down Expand Up @@ -458,9 +458,12 @@ def get_rope_index(

llm_positions = torch.cat(llm_pos_ids_list, dim=1).reshape(3, -1)

position_ids[..., i, attention_mask[i] == 1] = llm_positions.to(position_ids.device)
if attention_mask is not None:
position_ids[..., i, attention_mask[i]] = llm_positions.to(position_ids.device)
else:
position_ids[..., i, :] = llm_positions.to(position_ids.device)
mrope_position_deltas.append(llm_positions.max() + 1 - len(input_ids))
mrope_position_deltas = torch.tensor(mrope_position_deltas, device=input_ids.device).unsqueeze(1)
mrope_position_deltas = torch.tensor(mrope_position_deltas).unsqueeze(1).to(device=input_ids.device)

return position_ids, mrope_position_deltas
else:
Expand Down
15 changes: 9 additions & 6 deletions src/transformers/models/qwen2_5_omni/modular_qwen2_5_omni.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1320,8 +1320,8 @@ def get_rope_index(
mrope_position_deltas = []
if input_ids is not None and (image_grid_thw is not None or video_grid_thw is not None):
total_input_ids = input_ids
if attention_mask is None:
attention_mask = torch.ones_like(total_input_ids)
if attention_mask is not None:
attention_mask = attention_mask == 1
position_ids = torch.ones(
3,
input_ids.shape[0],
Expand All @@ -1330,9 +1330,9 @@ def get_rope_index(
device=input_ids.device,
)
image_idx, video_idx, audio_idx = 0, 0, 0
attention_mask = attention_mask.to(total_input_ids.device)
for i, input_ids in enumerate(total_input_ids):
input_ids = input_ids[attention_mask[i] == 1]
if attention_mask is not None:
input_ids = input_ids[attention_mask[i]]
image_nums, video_nums, audio_nums = 0, 0, 0
vision_start_indices = torch.argwhere(input_ids == vision_start_token_id).squeeze(1)
vision_tokens = input_ids[vision_start_indices + 1]
Expand Down Expand Up @@ -1513,9 +1513,12 @@ def get_rope_index(

llm_positions = torch.cat(llm_pos_ids_list, dim=1).reshape(3, -1)

position_ids[..., i, attention_mask[i] == 1] = llm_positions.to(position_ids.device)
if attention_mask is not None:
position_ids[..., i, attention_mask[i]] = llm_positions.to(position_ids.device)
else:
position_ids[..., i, :] = llm_positions.to(position_ids.device)
mrope_position_deltas.append(llm_positions.max() + 1 - len(input_ids))
mrope_position_deltas = torch.tensor(mrope_position_deltas, device=input_ids.device).unsqueeze(1)
mrope_position_deltas = torch.tensor(mrope_position_deltas).unsqueeze(1).to(device=input_ids.device)

return position_ids, mrope_position_deltas
else:
Expand Down
15 changes: 9 additions & 6 deletions src/transformers/models/qwen2_5_vl/modeling_qwen2_5_vl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1033,8 +1033,8 @@ def get_rope_index(
mrope_position_deltas = []
if input_ids is not None and (image_grid_thw is not None or video_grid_thw is not None):
total_input_ids = input_ids
if attention_mask is None:
attention_mask = torch.ones_like(total_input_ids)
if attention_mask is not None:
attention_mask = attention_mask == 1
position_ids = torch.ones(
3,
input_ids.shape[0],
Expand All @@ -1043,9 +1043,9 @@ def get_rope_index(
device=input_ids.device,
)
image_index, video_index = 0, 0
attention_mask = attention_mask.to(total_input_ids.device)
for i, input_ids in enumerate(total_input_ids):
input_ids = input_ids[attention_mask[i] == 1]
if attention_mask is not None:
input_ids = input_ids[attention_mask[i]]
image_nums, video_nums = 0, 0
vision_start_indices = torch.argwhere(input_ids == vision_start_token_id).squeeze(1)
vision_tokens = input_ids[vision_start_indices + 1]
Expand Down Expand Up @@ -1122,9 +1122,12 @@ def get_rope_index(
llm_pos_ids_list.append(torch.arange(text_len).view(1, -1).expand(3, -1) + st_idx)

llm_positions = torch.cat(llm_pos_ids_list, dim=1).reshape(3, -1)
position_ids[..., i, attention_mask[i] == 1] = llm_positions.to(position_ids.device)
if attention_mask is not None:
position_ids[..., i, attention_mask[i]] = llm_positions.to(position_ids.device)
else:
position_ids[..., i, :] = llm_positions.to(position_ids.device)
mrope_position_deltas.append(llm_positions.max() + 1 - len(total_input_ids[i]))
mrope_position_deltas = torch.tensor(mrope_position_deltas, device=input_ids.device).unsqueeze(1)
mrope_position_deltas = torch.tensor(mrope_position_deltas).unsqueeze(1).to(device=input_ids.device)
return position_ids, mrope_position_deltas
else:
if attention_mask is not None:
Expand Down
15 changes: 9 additions & 6 deletions src/transformers/models/qwen2_5_vl/modular_qwen2_5_vl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -421,8 +421,8 @@ def get_rope_index(
mrope_position_deltas = []
if input_ids is not None and (image_grid_thw is not None or video_grid_thw is not None):
total_input_ids = input_ids
if attention_mask is None:
attention_mask = torch.ones_like(total_input_ids)
if attention_mask is not None:
attention_mask = attention_mask == 1
position_ids = torch.ones(
3,
input_ids.shape[0],
Expand All @@ -431,9 +431,9 @@ def get_rope_index(
device=input_ids.device,
)
image_index, video_index = 0, 0
attention_mask = attention_mask.to(total_input_ids.device)
for i, input_ids in enumerate(total_input_ids):
input_ids = input_ids[attention_mask[i] == 1]
if attention_mask is not None:
input_ids = input_ids[attention_mask[i]]
image_nums, video_nums = 0, 0
vision_start_indices = torch.argwhere(input_ids == vision_start_token_id).squeeze(1)
vision_tokens = input_ids[vision_start_indices + 1]
Expand Down Expand Up @@ -510,9 +510,12 @@ def get_rope_index(
llm_pos_ids_list.append(torch.arange(text_len).view(1, -1).expand(3, -1) + st_idx)

llm_positions = torch.cat(llm_pos_ids_list, dim=1).reshape(3, -1)
position_ids[..., i, attention_mask[i] == 1] = llm_positions.to(position_ids.device)
if attention_mask is not None:
position_ids[..., i, attention_mask[i]] = llm_positions.to(position_ids.device)
else:
position_ids[..., i, :] = llm_positions.to(position_ids.device)
mrope_position_deltas.append(llm_positions.max() + 1 - len(total_input_ids[i]))
mrope_position_deltas = torch.tensor(mrope_position_deltas, device=input_ids.device).unsqueeze(1)
mrope_position_deltas = torch.tensor(mrope_position_deltas).unsqueeze(1).to(device=input_ids.device)
return position_ids, mrope_position_deltas
else:
if attention_mask is not None:
Expand Down