Skip to content

[fix] Fix qwen3.5 MoE vp_stage propagation under VPP #112

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

Open
none0663 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[fix] Fix qwen3.5 MoE vp_stage propagation under VPP #112

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
1 change: 1 addition & 0 deletions mbridge/models/qwen3_5/base_bridge.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -837,6 +837,7 @@ def provider(
image_token_id=self.hf_config.image_token_id,
video_token_id=self.hf_config.video_token_id,
vision_start_token_id=self.hf_config.vision_start_token_id,
vp_stage=vp_stage,
)

for callback in post_model_creation_callbacks:
Expand Down
8 changes: 7 additions & 1 deletion mbridge/models/qwen3_5/model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,3 +1,4 @@
import inspect
import logging
from typing import Optional

Expand Down Expand Up @@ -71,6 +72,7 @@ def __init__(
image_token_id: int = 151655,
video_token_id: int = 151656,
vision_start_token_id: int = 151652,
vp_stage: Optional[int] = None,
rope_scaling: bool = False,
) -> None:
super().__init__(config=language_transformer_config)
Expand Down Expand Up @@ -100,7 +102,7 @@ def __init__(
self._hook_fp32_rotary_emb(self.vision_model)
self._hook_vision_params_avg_grad_across_tp(self.vision_model)

self.language_model = Qwen3_5GPTModel(
language_model_kwargs = dict(
config=language_transformer_config,
transformer_layer_spec=language_transformer_layer_spec,
vocab_size=language_vocab_size,
Expand All @@ -117,6 +119,10 @@ def __init__(
mtp_block_spec=language_mtp_block_spec,
scatter_embedding_sequence_parallel=False,
)
if "vp_stage" in inspect.signature(GPTModel.__init__).parameters:
language_model_kwargs["vp_stage"] = vp_stage

self.language_model = Qwen3_5GPTModel(**language_model_kwargs)

@staticmethod
def _hook_fp32_rotary_emb(module: torch.nn.Module):
Expand Down