Skip to content

[Refactor] Refactor to extract model forward logic to allow plug-in t… #24203

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

Closed
wants to merge 2 commits into from
Closed

[Refactor] Refactor to extract model forward logic to allow plug-in t… #24203

Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
b7d2390
[Refactor] Refactor to extract model forward logic to allow plug-in t…
Sep 3, 2025
bdf92b6
[Refactor] Refactor to extract model forward logic to allow plug-in t…
Sep 3, 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
65 changes: 47 additions & 18 deletions vllm/v1/worker/gpu_model_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,7 +8,7 @@
from collections.abc import Iterator
from contextlib import contextmanager
from copy import deepcopy
from typing import TYPE_CHECKING, Any, Optional, Union, cast
from typing import Any, cast, Optional, Tuple, TYPE_CHECKING, Union

Check failure on line 11 in vllm/v1/worker/gpu_model_runner.py

View workflow job for this annotation

GitHub Actions / pre-commit

Ruff (UP035) 

vllm/v1/worker/gpu_model_runner.py:11:1: UP035 `typing.Tuple` is deprecated, use `tuple` instead

import numpy as np
import torch
Expand Down Expand Up @@ -1452,6 +1452,39 @@
kv_connector_output=kv_connector_output,
)

def _forward(
self,
attn_metadata: dict[str, Any],
num_input_tokens: int,
num_tokens_across_dp: int,
cudagraph_runtime_mode: CUDAGraphMode,
batch_descriptor: BatchDescriptor,
scheduler_output: "SchedulerOutput",
input_ids: torch.Tensor,
positions: torch.Tensor,
intermediate_tensors: IntermediateTensors,
inputs_embeds: list[torch.Tensor],
model_kwargs: dict[str, Any],
) -> Tuple[torch.Tensor, Optional[KVConnectorOutput]]:

Check failure on line 1468 in vllm/v1/worker/gpu_model_runner.py

View workflow job for this annotation

GitHub Actions / pre-commit

Ruff (UP006) 

vllm/v1/worker/gpu_model_runner.py:1468:10: UP006 Use `tuple` instead of `Tuple` for type annotation
Copy link
Contributor

@gemini-code-assist gemini-code-assist bot Sep 3, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

high

The return type hint for the model output is torch.Tensor, but the model can also return a tuple of tensors (e.g., when use_aux_hidden_state_outputs is true). This incorrect type hint can cause issues with static type checkers and mislead developers who might extend this class, especially since this method is designed to be overridden by plugins. Using Any will make the type hint correct for all possible return types from the model.

Suggested change
) -> Tuple[torch.Tensor, Optional[KVConnectorOutput]]:
) -> Tuple[Any, Optional[KVConnectorOutput]]:

with set_forward_context(
attn_metadata,
self.vllm_config,
num_tokens=num_input_tokens,
num_tokens_across_dp=num_tokens_across_dp,
cudagraph_runtime_mode=cudagraph_runtime_mode,
batch_descriptor=batch_descriptor,
), self.maybe_get_kv_connector_output(scheduler_output) as kv_connector_output:
return (
self.model(
input_ids=input_ids,
positions=positions,
intermediate_tensors=intermediate_tensors,
inputs_embeds=inputs_embeds,
**model_kwargs,
),
kv_connector_output,
)

@torch.inference_mode()
def execute_model(
self,
Expand Down Expand Up @@ -1557,23 +1590,19 @@

# Run the model.
# Use persistent buffers for CUDA graphs.
with set_forward_context(
attn_metadata,
self.vllm_config,
num_tokens=num_input_tokens,
num_tokens_across_dp=num_tokens_across_dp,
cudagraph_runtime_mode=cudagraph_runtime_mode,
batch_descriptor=batch_descriptor,
), self.maybe_get_kv_connector_output(
scheduler_output) as kv_connector_output:

model_output = self.model(
input_ids=input_ids,
positions=positions,
intermediate_tensors=intermediate_tensors,
inputs_embeds=inputs_embeds,
**model_kwargs,
)
model_output, kv_connector_output = self._forward(
attn_metadata,
num_input_tokens=num_input_tokens,
num_tokens_across_dp=num_tokens_across_dp,
cudagraph_runtime_mode=cudagraph_runtime_mode,

Check failure on line 1597 in vllm/v1/worker/gpu_model_runner.py

View workflow job for this annotation

GitHub Actions / pre-commit 

Argument "num_tokens_across_dp" to "_forward" of "GPUModelRunner" has incompatible type "Any | None"; expected "int"  [arg-type]

Check failure on line 1597 in vllm/v1/worker/gpu_model_runner.py

View workflow job for this annotation

GitHub Actions / pre-commit 

Argument "num_tokens_across_dp" to "_forward" of "GPUModelRunner" has incompatible type "Any | None"; expected "int"  [arg-type]

Check failure on line 1597 in vllm/v1/worker/gpu_model_runner.py

View workflow job for this annotation

GitHub Actions / pre-commit 

Argument "num_tokens_across_dp" to "_forward" of "GPUModelRunner" has incompatible type "Any | None"; expected "int"  [arg-type]

Check failure on line 1597 in vllm/v1/worker/gpu_model_runner.py

View workflow job for this annotation

GitHub Actions / pre-commit 

Argument "num_tokens_across_dp" to "_forward" of "GPUModelRunner" has incompatible type "Any | None"; expected "int"  [arg-type]

Check failure on line 1597 in vllm/v1/worker/gpu_model_runner.py

View workflow job for this annotation

GitHub Actions / pre-commit 

Argument "num_tokens_across_dp" to "_forward" of "GPUModelRunner" has incompatible type "Optional[Any]"; expected "int"  [arg-type]

Check failure on line 1597 in vllm/v1/worker/gpu_model_runner.py

View workflow job for this annotation

GitHub Actions / pre-commit 

Argument "num_tokens_across_dp" to "_forward" of "GPUModelRunner" has incompatible type "Optional[Any]"; expected "int"  [arg-type]

Check failure on line 1597 in vllm/v1/worker/gpu_model_runner.py

View workflow job for this annotation

GitHub Actions / pre-commit 

Argument "num_tokens_across_dp" to "_forward" of "GPUModelRunner" has incompatible type "Optional[Any]"; expected "int"  [arg-type]

Check failure on line 1597 in vllm/v1/worker/gpu_model_runner.py

View workflow job for this annotation

GitHub Actions / pre-commit 

Argument "num_tokens_across_dp" to "_forward" of "GPUModelRunner" has incompatible type "Optional[Any]"; expected "int"  [arg-type]
batch_descriptor=batch_descriptor,
scheduler_output=scheduler_output,
input_ids=input_ids,
positions=positions,
intermediate_tensors=intermediate_tensors,
inputs_embeds=inputs_embeds,
model_kwargs=model_kwargs,
)

if self.use_aux_hidden_state_outputs:
hidden_states, aux_hidden_states = model_output
Expand Down