up

control_net_canny.py

@@ -4,7 +4,9 @@ import os
 from huggingface_hub import HfApi
 from pathlib import Path
 from diffusers.utils import load_image
-from controlnet_aux import CannyDetector
+import cv2
+from PIL import Image
+import numpy as np
 
 from diffusers import (
     ControlNetModel,
@@ -18,9 +20,15 @@ checkpoint = sys.argv[1]
 image = load_image(
     "https://huggingface.co/lllyasviel/sd-controlnet-canny/resolve/main/images/bird.png"
 )
+image = np.array(image)
 
-canny_detector = CannyDetector()
-canny_image = canny_detector(image, low_threshold=100, high_threshold=200)
+low_threshold = 100
+high_threshold = 200
+
+image = cv2.Canny(image, low_threshold, high_threshold)
+image = image[:, :, None]
+image = np.concatenate([image, image, image], axis=2)
+canny_image = Image.fromarray(image)
 
 controlnet = ControlNetModel.from_pretrained(checkpoint, torch_dtype=torch.float16)
 pipe = StableDiffusionControlNetPipeline.from_pretrained(

control_net_inpaint.py

@@ -0,0 +1,55 @@
+#!/usr/bin/env python3
+import torch
+import os
+from huggingface_hub import HfApi
+from pathlib import Path
+from diffusers.utils import load_image
+from PIL import Image
+import numpy as np
+
+from diffusers import (
+    ControlNetModel,
+    StableDiffusionControlNetPipeline,
+    DDIMScheduler,
+)
+import sys
+
+checkpoint = sys.argv[1]
+
+
+# pre-process image and mask
+image = load_image("https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo.png").convert('RGB')
+mask_image = load_image("https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo_mask.png").convert("L")
+
+# convert to float32
+image = np.asarray(image, dtype=np.float32)
+mask_image = np.asarray(mask_image, dtype=np.float32)
+
+image[mask_image > 127] = -255.0
+image = torch.from_numpy(image)[None].permute(0, 3, 1, 2) / 255.0
+
+prompt = "A blue cat sitting on a park bench"
+
+controlnet = ControlNetModel.from_pretrained(checkpoint, torch_dtype=torch.float16)
+pipe = StableDiffusionControlNetPipeline.from_pretrained(
+    "runwayml/stable-diffusion-v1-5", controlnet=controlnet, torch_dtype=torch.float16
+)
+
+pipe.scheduler = DDIMScheduler.from_config(pipe.scheduler.config)
+pipe.enable_model_cpu_offload()
+
+generator = torch.manual_seed(0)
+out_image = pipe(prompt, num_inference_steps=20, generator=generator, image=image, guidance_scale=9.0).images[0]
+
+path = os.path.join(Path.home(), "images", "aa.png")
+out_image.save(path)
+
+api = HfApi()
+
+api.upload_file(
+    path_or_fileobj=path,
+    path_in_repo=path.split("/")[-1],
+    repo_id="patrickvonplaten/images",
+    repo_type="dataset",
+)
+print("https://huggingface.co/datasets/patrickvonplaten/images/blob/main/aa.png")

control_net_lineart.py

@@ -0,0 +1,50 @@
+#!/usr/bin/env python3
+import torch
+import os
+from huggingface_hub import HfApi
+from pathlib import Path
+from diffusers.utils import load_image
+from controlnet_aux import LineartDetector
+
+from diffusers import (
+    ControlNetModel,
+    StableDiffusionControlNetPipeline,
+    UniPCMultistepScheduler,
+)
+import sys
+
+checkpoint = sys.argv[1]
+
+url = "https://github.com/lllyasviel/ControlNet-v1-1-nightly/raw/main/test_imgs/bag.png"
+url = "https://github.com/lllyasviel/ControlNet-v1-1-nightly/raw/main/test_imgs/person_1.jpeg"
+image = load_image(url)
+
+prompt = "michael jackson concert"
+
+processor = LineartDetector.from_pretrained("lllyasviel/Annotators")
+image = processor(image)
+image.save("/home/patrick/images/check.png")
+
+controlnet = ControlNetModel.from_pretrained(checkpoint, torch_dtype=torch.float16)
+pipe = StableDiffusionControlNetPipeline.from_pretrained(
+    "runwayml/stable-diffusion-v1-5", controlnet=controlnet, torch_dtype=torch.float16
+)
+
+pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
+pipe.enable_model_cpu_offload()
+
+generator = torch.manual_seed(0)
+out_image = pipe(prompt, num_inference_steps=30, generator=generator, image=image).images[0]
+
+path = os.path.join(Path.home(), "images", "aa.png")
+out_image.save(path)
+
+api = HfApi()
+
+api.upload_file(
+    path_or_fileobj=path,
+    path_in_repo=path.split("/")[-1],
+    repo_id="patrickvonplaten/images",
+    repo_type="dataset",
+)
+print("https://huggingface.co/datasets/patrickvonplaten/images/blob/main/aa.png")

control_net_lineart_anime.py

@@ -0,0 +1,52 @@
+#!/usr/bin/env python3
+import torch
+import os
+from huggingface_hub import HfApi
+from pathlib import Path
+from diffusers.utils import load_image
+from controlnet_aux import LineartAnimeDetector
+from transformers import CLIPTextModel
+
+from diffusers import (
+    ControlNetModel,
+    StableDiffusionControlNetPipeline,
+    UniPCMultistepScheduler,
+)
+import sys
+
+checkpoint = sys.argv[1]
+
+url = "https://static.wikia.nocookie.net/unanything/images/a/a0/Unnamed.png/revision/latest/scale-to-width-down/350?cb=20230326002111"
+image = load_image(url)
+
+prompt = "warrior girl"
+
+processor = LineartAnimeDetector.from_pretrained("lllyasviel/Annotators")
+image = processor(image)
+image.save("/home/patrick/images/check.png")
+
+text_encoder = CLIPTextModel.from_pretrained("Linaqruf/anything-v3.0", subfolder="text_encoder", num_hidden_layers=11, torch_dtype=torch.float16)
+
+controlnet = ControlNetModel.from_pretrained(checkpoint, torch_dtype=torch.float16)
+pipe = StableDiffusionControlNetPipeline.from_pretrained(
+    "runwayml/stable-diffusion-v1-5", controlnet=controlnet, text_encoder=text_encoder, torch_dtype=torch.float16
+)
+
+pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
+pipe.enable_model_cpu_offload()
+
+generator = torch.manual_seed(33)
+out_image = pipe(prompt, num_inference_steps=25, generator=generator, image=image).images[0]
+
+path = os.path.join(Path.home(), "images", "aa.png")
+out_image.save(path)
+
+api = HfApi()
+
+api.upload_file(
+    path_or_fileobj=path,
+    path_in_repo=path.split("/")[-1],
+    repo_id="patrickvonplaten/images",
+    repo_type="dataset",
+)
+print("https://huggingface.co/datasets/patrickvonplaten/images/blob/main/aa.png")

control_net_normalbae.py

@@ -0,0 +1,49 @@
+#!/usr/bin/env python3
+import torch
+import os
+from huggingface_hub import HfApi
+from pathlib import Path
+from diffusers.utils import load_image
+from controlnet_aux import NormalBaeDetector
+
+from diffusers import (
+    ControlNetModel,
+    StableDiffusionControlNetPipeline,
+    UniPCMultistepScheduler,
+)
+import sys
+
+checkpoint = sys.argv[1]
+
+url = "https://github.com/lllyasviel/ControlNet-v1-1-nightly/raw/main/test_imgs/person-leaves.png"
+image = load_image(url)
+
+prompt = "A head full of roses"
+
+
+processor = NormalBaeDetector.from_pretrained("lllyasviel/Annotators")
+image = processor(image)
+
+controlnet = ControlNetModel.from_pretrained(checkpoint, torch_dtype=torch.float16)
+pipe = StableDiffusionControlNetPipeline.from_pretrained(
+    "runwayml/stable-diffusion-v1-5", controlnet=controlnet, torch_dtype=torch.float16
+)
+
+pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
+pipe.enable_model_cpu_offload()
+
+generator = torch.manual_seed(33)
+out_image = pipe(prompt, num_inference_steps=20, generator=generator, image=image).images[0]
+
+path = os.path.join(Path.home(), "images", "aa.png")
+out_image.save(path)
+
+api = HfApi()
+
+api.upload_file(
+    path_or_fileobj=path,
+    path_in_repo=path.split("/")[-1],
+    repo_id="patrickvonplaten/images",
+    repo_type="dataset",
+)
+print("https://huggingface.co/datasets/patrickvonplaten/images/blob/main/aa.png")

control_net_open_pose.py

@@ -15,12 +15,12 @@ import sys
 
 checkpoint = sys.argv[1]
 
-image = load_image("https://huggingface.co/lllyasviel/sd-controlnet-openpose/resolve/main/images/pose.png")
-prompt = "chef in the kitchen"
+image = load_image("https://github.com/lllyasviel/ControlNet-v1-1-nightly/raw/main/test_imgs/demo.jpg").resize((512, 512))
+prompt = "The pope with sunglasses rapping with a mic"
 
 
 openpose = OpenposeDetector.from_pretrained('lllyasviel/ControlNet')
-image = openpose(image)
+image = openpose(image, hand_and_face=True)
 
 controlnet = ControlNetModel.from_pretrained(checkpoint, torch_dtype=torch.float16)
 pipe = StableDiffusionControlNetPipeline.from_pretrained(
@@ -31,7 +31,7 @@ pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
 pipe.enable_model_cpu_offload()
 
 generator = torch.manual_seed(33)
-out_image = pipe(prompt, num_inference_steps=20, generator=generator, image=image).images[0]
+out_image = pipe(prompt, num_inference_steps=35, generator=generator, image=image).images[0]
 
 path = os.path.join(Path.home(), "images", "aa.png")
 out_image.save(path)

control_net_pix2pix.py

@@ -0,0 +1,45 @@
+#!/usr/bin/env python3
+import torch
+import os
+from huggingface_hub import HfApi
+from pathlib import Path
+from diffusers.utils import load_image
+from PIL import Image
+import numpy as np
+
+from diffusers import (
+    ControlNetModel,
+    StableDiffusionControlNetPipeline,
+    UniPCMultistepScheduler,
+)
+import sys
+
+checkpoint = sys.argv[1]
+
+image = load_image("https://huggingface.co/lllyasviel/sd-controlnet-seg/resolve/main/images/house.png").convert('RGB')
+
+prompt = "make it on fire"
+
+controlnet = ControlNetModel.from_pretrained(checkpoint, torch_dtype=torch.float16)
+pipe = StableDiffusionControlNetPipeline.from_pretrained(
+    "runwayml/stable-diffusion-v1-5", controlnet=controlnet, torch_dtype=torch.float16
+)
+
+pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
+pipe.enable_model_cpu_offload()
+
+generator = torch.manual_seed(0)
+out_image = pipe(prompt, num_inference_steps=30, generator=generator, image=image).images[0]
+
+path = os.path.join(Path.home(), "images", "aa.png")
+out_image.save(path)
+
+api = HfApi()
+
+api.upload_file(
+    path_or_fileobj=path,
+    path_in_repo=path.split("/")[-1],
+    repo_id="patrickvonplaten/images",
+    repo_type="dataset",
+)
+print("https://huggingface.co/datasets/patrickvonplaten/images/blob/main/aa.png")

control_net_shuffle.py

@@ -0,0 +1,49 @@
+#!/usr/bin/env python3
+import torch
+import os
+from huggingface_hub import HfApi
+from pathlib import Path
+from diffusers.utils import load_image
+from controlnet_aux import ContentShuffleDetector
+
+from diffusers import (
+    ControlNetModel,
+    StableDiffusionControlNetPipeline,
+    UniPCMultistepScheduler,
+)
+import sys
+
+checkpoint = sys.argv[1]
+
+url = "https://github.com/lllyasviel/ControlNet-v1-1-nightly/raw/main/test_imgs/city.jpg"
+image = load_image(url)
+
+prompt = "New York"
+
+
+processor = ContentShuffleDetector()
+image = processor(image)
+
+controlnet = ControlNetModel.from_pretrained(checkpoint, torch_dtype=torch.float16)
+pipe = StableDiffusionControlNetPipeline.from_pretrained(
+    "runwayml/stable-diffusion-v1-5", controlnet=controlnet, torch_dtype=torch.float16
+)
+
+pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
+pipe.enable_model_cpu_offload()
+
+generator = torch.manual_seed(33)
+out_image = pipe(prompt, num_inference_steps=20, generator=generator, image=image).images[0]
+
+path = os.path.join(Path.home(), "images", "aa.png")
+out_image.save(path)
+
+api = HfApi()
+
+api.upload_file(
+    path_or_fileobj=path,
+    path_in_repo=path.split("/")[-1],
+    repo_id="patrickvonplaten/images",
+    repo_type="dataset",
+)
+print("https://huggingface.co/datasets/patrickvonplaten/images/blob/main/aa.png")

control_net_soft_edge.py

@@ -0,0 +1,50 @@
+#!/usr/bin/env python3
+import torch
+import os
+from huggingface_hub import HfApi
+from pathlib import Path
+from diffusers.utils import load_image
+from PIL import Image
+import numpy as np
+from controlnet_aux import PidiNetDetector, HEDdetector
+
+from diffusers import (
+    ControlNetModel,
+    StableDiffusionControlNetPipeline,
+    UniPCMultistepScheduler,
+)
+import sys
+
+checkpoint = sys.argv[1]
+
+image = load_image("https://huggingface.co/lllyasviel/sd-controlnet-mlsd/resolve/main/images/room.png")
+
+prompt = "royal chamber with fancy bed"
+
+processor = HEDdetector.from_pretrained('lllyasviel/Annotators')
+processor = PidiNetDetector.from_pretrained('lllyasviel/Annotators')
+image = processor(image, safe=True)
+
+controlnet = ControlNetModel.from_pretrained(checkpoint, torch_dtype=torch.float16)
+pipe = StableDiffusionControlNetPipeline.from_pretrained(
+    "runwayml/stable-diffusion-v1-5", controlnet=controlnet, torch_dtype=torch.float16
+)
+
+pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
+pipe.enable_model_cpu_offload()
+
+generator = torch.manual_seed(0)
+out_image = pipe(prompt, num_inference_steps=30, generator=generator, image=image).images[0]
+
+path = os.path.join(Path.home(), "images", "aa.png")
+out_image.save(path)
+
+api = HfApi()
+
+api.upload_file(
+    path_or_fileobj=path,
+    path_in_repo=path.split("/")[-1],
+    repo_id="patrickvonplaten/images",
+    repo_type="dataset",
+)
+print("https://huggingface.co/datasets/patrickvonplaten/images/blob/main/aa.png")

run_local.py

@@ -1,13 +1,15 @@
 #!/usr/bin/env python3
-from diffusers import StableDiffusionPipeline, DPMSolverSinglestepScheduler, DPMSolverMultistepScheduler, DEISMultistepScheduler, HeunDiscreteScheduler
-from diffusers.schedulers.scheduling_unipc_multistep import UniPCMultistepScheduler
+from diffusers import StableDiffusionPipeline, KDPM2DiscreteScheduler, StableDiffusionImg2ImgPipeline, HeunDiscreteScheduler, KDPM2AncestralDiscreteScheduler
 import time
 import os
 from huggingface_hub import HfApi
-from compel import Compel
+# from compel import Compel
 import torch
 import sys
 from pathlib import Path
+import requests
+from PIL import Image
+from io import BytesIO
 
 path = sys.argv[1]
 
@@ -15,26 +17,34 @@ api = HfApi()
 start_time = time.time()
 #pipe = StableDiffusionPipeline.from_pretrained(path, torch_dtype=torch.float16, device_map="auto")
 #pipe.scheduler = HeunDiscreteScheduler.from_config(pipe.scheduler.config)
-pipe = StableDiffusionPipeline.from_pretrained(path, torch_dtype=torch.float16)
-pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
+pipe = StableDiffusionImg2ImgPipeline.from_pretrained(path, torch_dtype=torch.float16, safety_checker=None)
+pipe.scheduler = KDPM2AncestralDiscreteScheduler.from_config(pipe.scheduler.config)
 
-compel = Compel(tokenizer=pipe.tokenizer, text_encoder=pipe.text_encoder)
+# compel = Compel(tokenizer=pipe.tokenizer, text_encoder=pipe.text_encoder)
 
 
 pipe = pipe.to("cuda")
 
-prompt = "a highly realistic photo of green turtle"
+prompt = "ghibli style, a fantasy landscape with castles"
 
-prompts = ["a cat playing with a ball++ in the forest", "a cat playing with a ball++ in the forest", "a cat playing with a ball-- in the forest"]
+# rompts = ["a cat playing with a ball++ in the forest", "a cat playing with a ball++ in the forest", "a cat playing with a ball-- in the forest"]
 
-prompt_embeds = torch.cat([compel.build_conditioning_tensor(prompt) for prompt in prompts])
+# prompt_embeds = torch.cat([compel.build_conditioning_tensor(prompt) for prompt in prompts])
 
-generator = [torch.Generator(device="cuda").manual_seed(0) for _ in range(prompt_embeds.shape[0])]
-images = pipe(prompt_embeds=prompt_embeds, generator=generator, num_inference_steps=15).images
+# generator = [torch.Generator(device="cuda").manual_seed(0) for _ in range(prompt_embeds.shape[0])]
+#
+url = "https://raw.githubusercontent.com/CompVis/stable-diffusion/main/assets/stable-samples/img2img/sketch-mountains-input.jpg"
+
+response = requests.get(url)
+image = Image.open(BytesIO(response.content)).convert("RGB")
+image.thumbnail((768, 768))
+
+generator = torch.Generator(device="cpu").manual_seed(0)
+images = pipe(prompt=prompt, image=image, generator=generator, strength=0.75, num_inference_steps=30).images
 
 print("Time", time.time() - start_time)
 
-for i, image in enumerate(image):
+for i, image in enumerate(images):
     path = os.path.join(Path.home(), "images", f"aa_{i}.png")
     image.save(path)