Mirrored_Repos/InvokeAI
1
0
Fork 0
mirror of https://github.com/invoke-ai/InvokeAI synced 2024-08-30 20:32:17 +00:00
Code Issues Packages Projects Releases Wiki Activity

Remove dead code related to an old symmetry feature.

Browse Source
This commit is contained in:
Ryan Dick
2024-02-28 11:29:52 -05:00
parent 845c4e93ae
commit cad3e5dbd7
4 changed files with 0 additions and 103 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
invokeai/backend/stable_diffusion/diffusion/conditioning_data.py
View File

@ -43,14 +43,6 @@ class SDXLConditioningInfo(BasicConditioningInfo):
return super().to(device=device, dtype=dtype)
@dataclass(frozen=True)
class PostprocessingSettings:
threshold: float
warmup: float
h_symmetry_time_pct: Optional[float]
v_symmetry_time_pct: Optional[float]
@dataclass
class IPAdapterConditioningInfo:
cond_image_prompt_embeds: torch.Tensor
@ -82,10 +74,6 @@ class ConditioningData:
"""
guidance_rescale_multiplier: float = 0
scheduler_args: dict[str, Any] = field(default_factory=dict)
"""
Additional arguments to pass to invokeai_diffuser.do_latent_postprocessing().
"""
postprocessing_settings: Optional[PostprocessingSettings] = None
ip_adapter_conditioning: Optional[list[IPAdapterConditioningInfo]] = None

75
invokeai/backend/stable_diffusion/diffusion/shared_invokeai_diffusion.py
View File

@ -14,7 +14,6 @@ from invokeai.backend.stable_diffusion.diffusion.conditioning_data import (
BasicConditioningInfo,
ConditioningData,
ExtraConditioningInfo,
PostprocessingSettings,
SDXLConditioningInfo,
)
from invokeai.backend.stable_diffusion.diffusion.regional_prompt_attention import Range, RegionalPromptData
@ -374,19 +373,6 @@ class InvokeAIDiffuserComponent:
return unconditioned_next_x, conditioned_next_x
def do_latent_postprocessing(
self,
postprocessing_settings: PostprocessingSettings,
latents: torch.Tensor,
sigma,
step_index,
total_step_count,
) -> torch.Tensor:
if postprocessing_settings is not None:
percent_through = step_index / total_step_count
latents = self.apply_symmetry(postprocessing_settings, latents, percent_through)
return latents
def _concat_conditionings_for_batch(self, unconditioning, conditioning):
def _pad_conditioning(cond, target_len, encoder_attention_mask):
conditioning_attention_mask = torch.ones(
@ -676,64 +662,3 @@ class InvokeAIDiffuserComponent:
scaled_delta = (conditioned_next_x - unconditioned_next_x) * guidance_scale
combined_next_x = unconditioned_next_x + scaled_delta
return combined_next_x
def apply_symmetry(
self,
postprocessing_settings: PostprocessingSettings,
latents: torch.Tensor,
percent_through: float,
) -> torch.Tensor:
# Reset our last percent through if this is our first step.
if percent_through == 0.0:
self.last_percent_through = 0.0
if postprocessing_settings is None:
return latents
# Check for out of bounds
h_symmetry_time_pct = postprocessing_settings.h_symmetry_time_pct
if h_symmetry_time_pct is not None and (h_symmetry_time_pct <= 0.0 or h_symmetry_time_pct > 1.0):
h_symmetry_time_pct = None
v_symmetry_time_pct = postprocessing_settings.v_symmetry_time_pct
if v_symmetry_time_pct is not None and (v_symmetry_time_pct <= 0.0 or v_symmetry_time_pct > 1.0):
v_symmetry_time_pct = None
dev = latents.device.type
latents.to(device="cpu")
if (
h_symmetry_time_pct is not None
and self.last_percent_through < h_symmetry_time_pct
and percent_through >= h_symmetry_time_pct
):
# Horizontal symmetry occurs on the 3rd dimension of the latent
width = latents.shape[3]
x_flipped = torch.flip(latents, dims=[3])
latents = torch.cat(
[
latents[:, :, :, 0 : int(width / 2)],
x_flipped[:, :, :, int(width / 2) : int(width)],
],
dim=3,
)
if (
v_symmetry_time_pct is not None
and self.last_percent_through < v_symmetry_time_pct
and percent_through >= v_symmetry_time_pct
):
# Vertical symmetry occurs on the 2nd dimension of the latent
height = latents.shape[2]
y_flipped = torch.flip(latents, dims=[2])
latents = torch.cat(
[
latents[:, :, 0 : int(height / 2)],
y_flipped[:, :, int(height / 2) : int(height)],
],
dim=2,
)
self.last_percent_through = percent_through
return latents.to(device=dev)