github/InvokeAI
1
1
Fork 0
mirror of https://github.com/invoke-ai/InvokeAI.git synced 2026-04-23 03:00:31 -04:00
Code Issues Packages Projects Releases Wiki Activity

Remove LoRA stuff from DiffusersCotnrolNetFlux.

Browse Source
This commit is contained in:
Ryan Dick
2024-10-04 15:11:29 +00:00
parent a17ea9bfad
commit e93da5d4b2
1 changed files with 1 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
invokeai/backend/flux/controlnet/diffusers_controlnet_flux.py
View File

@@ -22,7 +22,7 @@ class FluxControlNetOutput(BaseOutput):
controlnet_single_block_samples: Tuple[torch.Tensor]
class DiffusersControlNetFlux(ModelMixin, ConfigMixin, PeftAdapterMixin):
class DiffusersControlNetFlux(ModelMixin, ConfigMixin):
_supports_gradient_checkpointing = True
@register_to_config
@@ -142,7 +142,6 @@ class DiffusersControlNetFlux(ModelMixin, ConfigMixin, PeftAdapterMixin):
img_ids: torch.Tensor = None,
txt_ids: torch.Tensor = None,
guidance: torch.Tensor = None,
joint_attention_kwargs: Optional[Dict[str, Any]] = None,
return_dict: bool = True,
) -> Union[torch.FloatTensor, Transformer2DModelOutput]:
"""
@@ -165,10 +164,6 @@ class DiffusersControlNetFlux(ModelMixin, ConfigMixin, PeftAdapterMixin):
Used to indicate denoising step.
block_controlnet_hidden_states: (`list` of `torch.Tensor`):
A list of tensors that if specified are added to the residuals of transformer blocks.
joint_attention_kwargs (`dict`, *optional*):
A kwargs dictionary that if specified is passed along to the `AttentionProcessor` as defined under
`self.processor` in
[diffusers.models.attention_processor](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py).
return_dict (`bool`, *optional*, defaults to `True`):
Whether or not to return a [`~models.transformer_2d.Transformer2DModelOutput`] instead of a plain
tuple.
@@ -177,20 +172,7 @@ class DiffusersControlNetFlux(ModelMixin, ConfigMixin, PeftAdapterMixin):
If `return_dict` is True, an [`~models.transformer_2d.Transformer2DModelOutput`] is returned, otherwise a
`tuple` where the first element is the sample tensor.
"""
if joint_attention_kwargs is not None:
joint_attention_kwargs = joint_attention_kwargs.copy()
lora_scale = joint_attention_kwargs.pop("scale", 1.0)
else:
lora_scale = 1.0
if USE_PEFT_BACKEND:
# weight the lora layers by setting `lora_scale` for each PEFT layer
scale_lora_layers(self, lora_scale)
else:
if joint_attention_kwargs is not None and joint_attention_kwargs.get("scale", None) is not None:
logger.warning(
"Passing `scale` via `joint_attention_kwargs` when not using the PEFT backend is ineffective."
)
hidden_states = self.x_embedder(hidden_states)
# add
@@ -319,10 +301,6 @@ class DiffusersControlNetFlux(ModelMixin, ConfigMixin, PeftAdapterMixin):
None if len(controlnet_single_block_samples) == 0 else controlnet_single_block_samples
)
if USE_PEFT_BACKEND:
# remove `lora_scale` from each PEFT layer
unscale_lora_layers(self, lora_scale)
if not return_dict:
return (controlnet_block_samples, controlnet_single_block_samples)