add

bench.py

@@ -0,0 +1,27 @@
+#!/usr/bin/env python3
+from diffusers import StableDiffusionPipeline, UniPCMultistepScheduler
+import time
+import torch
+import sys
+
+path = sys.argv[1]
+use_device_map = bool(int(sys.argv[2]))
+
+start_time = time.time()
+
+if use_device_map:
+    print("Load directly on GPU")
+    pipe = StableDiffusionPipeline.from_pretrained(path, torch_dtype=torch.float16, device_map="auto")
+else:
+    print("Load directly on CPU")
+    pipe = StableDiffusionPipeline.from_pretrained(path, torch_dtype=torch.float16)
+    pipe = pipe.to("cuda")
+
+pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
+
+prompt = "a highly realistic photo of green turtle"
+
+print("Loading Time", time.time() - start_time)
+generator = torch.Generator(device="cuda").manual_seed(0)
+image = pipe(prompt, generator=generator, num_inference_steps=15).images[0]
+print("Time", time.time() - start_time)

control_net.py

@@ -0,0 +1,53 @@
+#!/usr/bin/env python3
+import torch
+import numpy as np
+import os
+from huggingface_hub import HfApi
+from pathlib import Path
+import cv2
+from PIL import Image
+from diffusers.utils import load_image
+
+from diffusers import (
+    ControlNetModel,
+    StableDiffusionControlNetPipeline,
+    UniPCMultistepScheduler,
+)
+
+image = load_image(
+    "https://hf.co/datasets/huggingface/documentation-images/resolve/main/diffusers/input_image_vermeer.png"
+)
+
+image = np.array(image)
+
+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("lllyasviel/sd-controlnet-canny", 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("futuristic-looking woman", num_inference_steps=20, generator=generator, image=canny_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")

safetensors_bench.py

@@ -0,0 +1,16 @@
+#!/usr/bin/env python3
+from safetensors.torch import load_file as safe_load_file
+import time
+import sys
+
+direct_on_gpu = bool(int(sys.argv[1]))
+
+if direct_on_gpu:
+    start_time = time.time()
+    checkpoint = safe_load_file("/home/patrick_huggingface_co/stable-diffusion-v1-4/unet/diffusion_pytorch_model.safetensors", device=0)
+    print("Directly on GPU", time.time() - start_time)
+else:
+    start_time = time.time()
+    checkpoint = safe_load_file("/home/patrick_huggingface_co/stable-diffusion-v1-4/unet/diffusion_pytorch_model.safetensors")
+    checkpoint = {k: v.to("cuda:0") for k, v in checkpoint.items()}
+    print("On CPU", time.time() - start_time)