↓ 7 callersFunctionscaled_index_add(
input: torch.Tensor, # [B, M, D]
index: torch.Tensor, # [Bi] - int64
source: torch.Tensor, #
models/vipe/vipe/ext/xformers.py:38
↓ 7 callersMethodtile(self, model_input, tile_size, tile_stride, tile_device, tile_dtype)
DiffSynth-Studio/diffsynth/models/tiler.py:20
↓ 7 callersFunctionweighted_mean(x: torch.Tensor, w: torch.Tensor = None, dim: Union[int, torch.Size] = None, keepdim: bool = False, eps: floa
models/spatracker/models/SpaTrackV2/models/tracker3D/spatrack_modules/geometry_torch.py:16
↓ 6 callersMethod__init__(
self,
dim: int,
in_dim: int,
ffn_dim: int,
out_dim: int,
tex
DiffSynth-Studio/diffsynth/models/wan_video_dit_s2v.py:362
↓ 6 callersMethod__init__(
self,
patch_size=(2, 2),
in_chans=4,
embed_dim=1408,
bias=True,
DiffSynth-Studio/diffsynth/models/hunyuan_dit.py:177
↓ 6 callersMethodbuild_remapping_table(self, frames_guide, frames_style, patch_match_engine, batch_size, desc="")
DiffSynth-Studio/diffsynth/extensions/FastBlend/runners/fast.py:34
↓ 6 callersMethodload(self, file_path="", state_dict={}, device="cuda", torch_dtype=torch.float16, **kwargs)
DiffSynth-Studio/diffsynth/models/model_manager.py:143
↓ 6 callersMethodprocess_window_sum(self, frames_guide, blending_table, patch_match_engine, window_size, batch_size, desc="")
DiffSynth-Studio/diffsynth/extensions/FastBlend/runners/fast.py:77