Datasets:

ArXiv:
diffusers-benchmarking-bot commited on
Commit
9d521f1
·
verified ·
1 Parent(s): 279ba5a

Upload folder using huggingface_hub

Browse files
main/README.md CHANGED
@@ -5381,7 +5381,7 @@ pipe = DiffusionPipeline.from_pretrained(
5381
  # Here we need use pipeline internal unet model
5382
  pipe.unet = pipe.unet_model.from_pretrained(model_id, subfolder="unet", variant="fp16", use_safetensors=True)
5383
 
5384
- # Load aditional layers to the model
5385
  pipe.unet.load_additional_layers(weight_path="proc_data/faithdiff/FaithDiff.bin", dtype=dtype)
5386
 
5387
  # Enable vae tiling
 
5381
  # Here we need use pipeline internal unet model
5382
  pipe.unet = pipe.unet_model.from_pretrained(model_id, subfolder="unet", variant="fp16", use_safetensors=True)
5383
 
5384
+ # Load additional layers to the model
5385
  pipe.unet.load_additional_layers(weight_path="proc_data/faithdiff/FaithDiff.bin", dtype=dtype)
5386
 
5387
  # Enable vae tiling
main/dps_pipeline.py CHANGED
@@ -312,9 +312,9 @@ if __name__ == "__main__":
312
  # These are the coordinates of the output image
313
  out_coordinates = np.arange(1, out_length + 1)
314
 
315
- # since both scale-factor and output size can be provided simulatneously, perserving the center of the image requires shifting
316
- # the output coordinates. the deviation is because out_length doesn't necesary equal in_length*scale.
317
- # to keep the center we need to subtract half of this deivation so that we get equal margins for boths sides and center is preserved.
318
  shifted_out_coordinates = out_coordinates - (out_length - in_length * scale) / 2
319
 
320
  # These are the matching positions of the output-coordinates on the input image coordinates.
 
312
  # These are the coordinates of the output image
313
  out_coordinates = np.arange(1, out_length + 1)
314
 
315
+ # since both scale-factor and output size can be provided simultaneously, preserving the center of the image requires shifting
316
+ # the output coordinates. the deviation is because out_length doesn't necessary equal in_length*scale.
317
+ # to keep the center we need to subtract half of this deviation so that we get equal margins for both sides and center is preserved.
318
  shifted_out_coordinates = out_coordinates - (out_length - in_length * scale) / 2
319
 
320
  # These are the matching positions of the output-coordinates on the input image coordinates.
main/fresco_v2v.py CHANGED
@@ -351,7 +351,7 @@ def my_forward(
351
  cross_attention_kwargs (`dict`, *optional*):
352
  A kwargs dictionary that if specified is passed along to the [`AttnProcessor`].
353
  added_cond_kwargs: (`dict`, *optional*):
354
- A kwargs dictionary containin additional embeddings that if specified are added to the embeddings that
355
  are passed along to the UNet blocks.
356
 
357
  Returns:
@@ -864,9 +864,9 @@ def get_flow_and_interframe_paras(flow_model, imgs):
864
  class AttentionControl:
865
  """
866
  Control FRESCO-based attention
867
- * enable/diable spatial-guided attention
868
- * enable/diable temporal-guided attention
869
- * enable/diable cross-frame attention
870
  * collect intermediate attention feature (for spatial-guided attention)
871
  """
872
 
 
351
  cross_attention_kwargs (`dict`, *optional*):
352
  A kwargs dictionary that if specified is passed along to the [`AttnProcessor`].
353
  added_cond_kwargs: (`dict`, *optional*):
354
+ A kwargs dictionary containing additional embeddings that if specified are added to the embeddings that
355
  are passed along to the UNet blocks.
356
 
357
  Returns:
 
864
  class AttentionControl:
865
  """
866
  Control FRESCO-based attention
867
+ * enable/disable spatial-guided attention
868
+ * enable/disable temporal-guided attention
869
+ * enable/disable cross-frame attention
870
  * collect intermediate attention feature (for spatial-guided attention)
871
  """
872
 
main/hd_painter.py CHANGED
@@ -34,7 +34,7 @@ class RASGAttnProcessor:
34
  temb: Optional[torch.Tensor] = None,
35
  scale: float = 1.0,
36
  ) -> torch.Tensor:
37
- # Same as the default AttnProcessor up untill the part where similarity matrix gets saved
38
  downscale_factor = self.mask_resoltuion // hidden_states.shape[1]
39
  residual = hidden_states
40
 
 
34
  temb: Optional[torch.Tensor] = None,
35
  scale: float = 1.0,
36
  ) -> torch.Tensor:
37
+ # Same as the default AttnProcessor up until the part where similarity matrix gets saved
38
  downscale_factor = self.mask_resoltuion // hidden_states.shape[1]
39
  residual = hidden_states
40