Add pydocstyle to the ruff rules (#52136)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Signed-off-by: Harry Mellor <hej.mellor@gmail.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
This commit is contained in:
Harry Mellor
2026-09-16 16:44:29 +00:00
committed by GitHub
co-authored by Claude Opus 5
parent 804e5377aa
commit 8c1557a79c
2002 changed files with 15024 additions and 14269 deletions
+11 -8
View File
@@ -56,8 +56,7 @@ from .vision import get_vision_encoder_info
class LlavaNextVideoPixelInputs(TensorSchema):
"""
Dimensions:
"""Dimensions:
- bn: Batch size * number of videos
- f: Number of frames
- c: Number of channels (3)
@@ -381,8 +380,7 @@ class LlavaNextVideoForConditionalGeneration(
def _parse_and_validate_video_input(
self, **kwargs: object
) -> LlavaNextVideoPixelInputs | None:
"""
A legal video input should have the following dimensions:
"""A legal video input should have the following dimensions:
{
"pixel_values_videos" :
list[b, Tensor(nb_frames, nb_channels, height, width)]
@@ -457,10 +455,17 @@ class LlavaNextVideoForConditionalGeneration(
**kwargs: object,
) -> torch.Tensor | IntermediateTensors:
"""Run forward pass for LlaVA-NeXT-Video.
Args:
input_ids: Flattened (concatenated) input_ids corresponding to a
batch.
pixel_values_videos: Pixels in each frames for each input videos.
positions: Flattened (concatenated) position ids corresponding to a
batch.
intermediate_tensors: Intermediate tensors from prior forward pass.
inputs_embeds: Optional tensor of input embeddings.
**kwargs: Multimodal inputs for this batch, forwarded to the
multimodal embedding path.
"""
if intermediate_tensors is not None:
inputs_embeds = None
@@ -486,9 +491,7 @@ class LlavaNextVideoForConditionalGeneration(
return loader.load_weights(weights, mapper=self.hf_to_vllm_mapper)
def get_mm_mapping(self) -> MultiModelKeys:
"""
Get the module prefix in multimodal models
"""
"""Get the module prefix in multimodal models."""
return MultiModelKeys.from_string_field(
language_model="language_model",
connector="multi_modal_projector",