technicalkiwi
/
latentblending
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

controlnet upd

This commit is contained in:
Johannes Stelzer 2023-07-21 14:03:02 +02:00
parent bc5713241f
commit 704433e267
1 changed files with 48 additions and 43 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

91
diffusers_holder.py
View File

@ -163,9 +163,20 @@ class DiffusersHolder():
image = self.pipe.vae.decode(latents / self.pipe.vae.config.scaling_factor, return_dict=False)[0] image = self.pipe.vae.decode(latents / self.pipe.vae.config.scaling_factor, return_dict=False)[0]
image = self.pipe.image_processor.postprocess(image, output_type="pil", do_denormalize=[True] * image.shape[0]) image = self.pipe.image_processor.postprocess(image, output_type="pil", do_denormalize=[True] * image.shape[0])
return np.asarray(image[0]) return np.asarray(image[0])
def prepare_mixing(self, mixing_coeffs, list_latents_mixing):
if type(mixing_coeffs) == float:
list_mixing_coeffs = (1 + self.num_inference_steps) * [mixing_coeffs]
elif type(mixing_coeffs) == list:
assert len(mixing_coeffs) == self.num_inference_steps, f"len(mixing_coeffs) {len(mixing_coeffs)} != self.num_inference_steps {self.num_inference_steps}"
list_mixing_coeffs = mixing_coeffs
else:
raise ValueError("mixing_coeffs should be float or list with len=num_inference_steps")
if np.sum(list_mixing_coeffs) > 0:
assert len(list_latents_mixing) == self.num_inference_steps, f"len(list_latents_mixing) {len(list_latents_mixing)} != self.num_inference_steps {self.num_inference_steps}"
return list_mixing_coeffs
@torch.no_grad() @torch.no_grad()
def run_diffusion( def run_diffusion(
self, self,
@ -175,14 +186,13 @@ class DiffusersHolder():
list_latents_mixing=None, list_latents_mixing=None,
mixing_coeffs=0.0, mixing_coeffs=0.0,
return_image: Optional[bool] = False): return_image: Optional[bool] = False):
if self.pipe.__class__.__name__ == 'StableDiffusionXLPipeline': if self.pipe.__class__.__name__ == 'StableDiffusionXLPipeline':
return self.run_diffusion_sd_xl(text_embeddings, latents_start, idx_start, list_latents_mixing, mixing_coeffs, return_image) return self.run_diffusion_sd_xl(text_embeddings, latents_start, idx_start, list_latents_mixing, mixing_coeffs, return_image)
elif self.pipe.__class__.__name__ == 'StableDiffusionPipeline': elif self.pipe.__class__.__name__ == 'StableDiffusionPipeline':
return self.run_diffusion_sd12x(text_embeddings, latents_start, idx_start, list_latents_mixing, mixing_coeffs, return_image) return self.run_diffusion_sd12x(text_embeddings, latents_start, idx_start, list_latents_mixing, mixing_coeffs, return_image)
elif self.pipe.__class__.__name__ == 'StableDiffusionControlNetPipeline': elif self.pipe.__class__.__name__ == 'StableDiffusionControlNetPipeline':
pass pass
@torch.no_grad() @torch.no_grad()
def run_diffusion_sd12x( def run_diffusion_sd12x(
@ -193,28 +203,19 @@ class DiffusersHolder():
list_latents_mixing=None, list_latents_mixing=None,
mixing_coeffs=0.0, mixing_coeffs=0.0,
return_image: Optional[bool] = False): return_image: Optional[bool] = False):
if type(mixing_coeffs) == float:
list_mixing_coeffs = (1+self.num_inference_steps) * [mixing_coeffs]
elif type(mixing_coeffs) == list:
assert len(mixing_coeffs) == self.num_inference_steps, f"len(mixing_coeffs) {len(mixing_coeffs)} != self.num_inference_steps {self.num_inference_steps}"
list_mixing_coeffs = mixing_coeffs
else:
raise ValueError("mixing_coeffs should be float or list with len=num_inference_steps")
if np.sum(list_mixing_coeffs) > 0: list_mixing_coeffs = self.prepare_mixing()
assert len(list_latents_mixing) == self.num_inference_steps, f"len(list_latents_mixing) {len(list_latents_mixing)} != self.num_inference_steps {self.num_inference_steps}"
do_classifier_free_guidance = self.guidance_scale > 1.0 do_classifier_free_guidance = self.guidance_scale > 1.0
# diffusers bit wiggly # accomodate different sd model types
self.pipe.scheduler.set_timesteps(self.num_inference_steps-1, device=self.device) self.pipe.scheduler.set_timesteps(self.num_inference_steps - 1, device=self.device)
timesteps = self.pipe.scheduler.timesteps timesteps = self.pipe.scheduler.timesteps
if len(timesteps) != self.num_inference_steps: if len(timesteps) != self.num_inference_steps:
self.pipe.scheduler.set_timesteps(self.num_inference_steps, device=self.device) self.pipe.scheduler.set_timesteps(self.num_inference_steps, device=self.device)
timesteps = self.pipe.scheduler.timesteps timesteps = self.pipe.scheduler.timesteps
latents = latents_start.clone() latents = latents_start.clone()
list_latents_out = [] list_latents_out = []
@ -229,11 +230,11 @@ class DiffusersHolder():
if i > 0 and list_mixing_coeffs[i] > 0: if i > 0 and list_mixing_coeffs[i] > 0:
latents_mixtarget = list_latents_mixing[i - 1].clone() latents_mixtarget = list_latents_mixing[i - 1].clone()
latents = interpolate_spherical(latents, latents_mixtarget, list_mixing_coeffs[i]) latents = interpolate_spherical(latents, latents_mixtarget, list_mixing_coeffs[i])
# expand the latents if we are doing classifier free guidance # expand the latents if we are doing classifier free guidance
latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents
latent_model_input = self.pipe.scheduler.scale_model_input(latent_model_input, t) latent_model_input = self.pipe.scheduler.scale_model_input(latent_model_input, t)
# predict the noise residual # predict the noise residual
noise_pred = self.pipe.unet( noise_pred = self.pipe.unet(
latent_model_input, latent_model_input,
@ -248,7 +249,7 @@ class DiffusersHolder():
# compute the previous noisy sample x_t -> x_t-1 # compute the previous noisy sample x_t -> x_t-1
latents = self.pipe.scheduler.step(noise_pred, t, latents, return_dict=False)[0] latents = self.pipe.scheduler.step(noise_pred, t, latents, return_dict=False)[0]
list_latents_out.append(latents.clone()) list_latents_out.append(latents.clone())
if return_image: if return_image:
return self.latent2image(latents) return self.latent2image(latents)
else: else:
@ -276,17 +277,8 @@ class DiffusersHolder():
do_classifier_free_guidance = self.guidance_scale > 1.0 do_classifier_free_guidance = self.guidance_scale > 1.0
# 1. Check inputs. Raise error if not correct & 2. Define call parameters # 1. Check inputs. Raise error if not correct & 2. Define call parameters
if type(mixing_coeffs) == float: list_mixing_coeffs = self.prepare_mixing()
list_mixing_coeffs = (1+self.num_inference_steps) * [mixing_coeffs]
elif type(mixing_coeffs) == list:
assert len(mixing_coeffs) == self.num_inference_steps, f"len(mixing_coeffs) {len(mixing_coeffs)} != self.num_inference_steps {self.num_inference_steps}"
list_mixing_coeffs = mixing_coeffs
else:
raise ValueError("mixing_coeffs should be float or list with len=num_inference_steps")
if np.sum(list_mixing_coeffs) > 0:
assert len(list_latents_mixing) == self.num_inference_steps, f"len(list_latents_mixing) {len(list_latents_mixing)} != self.num_inference_steps {self.num_inference_steps}"
# 3. Encode input prompt (already encoded outside bc of mixing, just split here) # 3. Encode input prompt (already encoded outside bc of mixing, just split here)
prompt_embeds, negative_prompt_embeds, pooled_prompt_embeds, negative_pooled_prompt_embeds = text_embeddings prompt_embeds, negative_prompt_embeds, pooled_prompt_embeds, negative_pooled_prompt_embeds = text_embeddings
@ -374,10 +366,11 @@ class DiffusersHolder():
list_latents_mixing=None, list_latents_mixing=None,
mixing_coeffs=0.0, mixing_coeffs=0.0,
return_image: Optional[bool] = False): return_image: Optional[bool] = False):
prompt_embeds = conditioning[0] prompt_embeds = conditioning[0]
image = conditioning[1]
list_mixing_coeffs = self.prepare_mixing()
controlnet = self.pipe.controlnet controlnet = self.pipe.controlnet
control_guidance_start = [0.0] control_guidance_start = [0.0]
control_guidance_end = [1.0] control_guidance_end = [1.0]
@ -386,17 +379,14 @@ class DiffusersHolder():
batch_size = 1 batch_size = 1
eta = 0.0 eta = 0.0
controlnet_conditioning_scale = 1.0 controlnet_conditioning_scale = 1.0
image = Image.open("/home/lugo/glif/lora_models/pretrained_model_name_or_path/value_runwayml_stable-diffusion-v1-5_fabian/fabian_in_the_desert/img_001.jpg")
# align format for control guidance # align format for control guidance
if not isinstance(control_guidance_start, list) and isinstance(control_guidance_end, list): if not isinstance(control_guidance_start, list) and isinstance(control_guidance_end, list):
control_guidance_start = len(control_guidance_end) * [control_guidance_start] control_guidance_start = len(control_guidance_end) * [control_guidance_start]
elif not isinstance(control_guidance_end, list) and isinstance(control_guidance_start, list): elif not isinstance(control_guidance_end, list) and isinstance(control_guidance_start, list):
control_guidance_end = len(control_guidance_start) * [control_guidance_end] control_guidance_end = len(control_guidance_start) * [control_guidance_end]
# 2. Define call parameters # 2. Define call parameters
device = self.pipe._execution_device device = self.pipe._execution_device
# here `guidance_scale` is defined analog to the guidance weight `w` of equation (2) # here `guidance_scale` is defined analog to the guidance weight `w` of equation (2)
# of the Imagen paper: https://arxiv.org/pdf/2205.11487.pdf . `guidance_scale = 1` # of the Imagen paper: https://arxiv.org/pdf/2205.11487.pdf . `guidance_scale = 1`
@ -424,6 +414,7 @@ class DiffusersHolder():
# 6. Prepare latent variables # 6. Prepare latent variables
generator = torch.Generator(device=self.device).manual_seed(int(420)) generator = torch.Generator(device=self.device).manual_seed(int(420))
latents = latents_start.clone() latents = latents_start.clone()
list_latents_out = []
# 7. Prepare extra step kwargs. TODO: Logic should ideally just be moved out of the pipeline # 7. Prepare extra step kwargs. TODO: Logic should ideally just be moved out of the pipeline
extra_step_kwargs = self.pipe.prepare_extra_step_kwargs(generator, eta) extra_step_kwargs = self.pipe.prepare_extra_step_kwargs(generator, eta)
@ -439,6 +430,17 @@ class DiffusersHolder():
# 8. Denoising loop # 8. Denoising loop
for i, t in enumerate(timesteps): for i, t in enumerate(timesteps):
if i < idx_start:
list_latents_out.append(None)
continue
elif i == idx_start:
latents = latents_start.clone()
# Mix latents for crossfeeding
if i > 0 and list_mixing_coeffs[i] > 0:
latents_mixtarget = list_latents_mixing[i - 1].clone()
latents = interpolate_spherical(latents, latents_mixtarget, list_mixing_coeffs[i])
# expand the latents if we are doing classifier free guidance # expand the latents if we are doing classifier free guidance
latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents
latent_model_input = self.pipe.scheduler.scale_model_input(latent_model_input, t) latent_model_input = self.pipe.scheduler.scale_model_input(latent_model_input, t)
@ -487,10 +489,13 @@ class DiffusersHolder():
# compute the previous noisy sample x_t -> x_t-1 # compute the previous noisy sample x_t -> x_t-1
latents = self.pipe.scheduler.step(noise_pred, t, latents, **extra_step_kwargs, return_dict=False)[0] latents = self.pipe.scheduler.step(noise_pred, t, latents, **extra_step_kwargs, return_dict=False)[0]
image = self.pipe.vae.decode(latents / self.pipe.vae.config.scaling_factor, return_dict=False)[0] # Append latents
image, has_nsfw_concept = self.pipe.run_safety_checker(image, device, prompt_embeds.dtype) list_latents_out.append(latents.clone())
image = self.pipe.image_processor.postprocess(image, output_type="pil")
return image if return_image:
return self.latent2image(latents)
else:
return list_latents_out
#%% #%%