test control net

app.py

@@ -1,6 +1,6 @@
 import gradio as gr
 
-from app_haircolor_inpaint_adapter_15 import create_demo as create_demo_haircolor
+from app_haircolor_inpaint_15 import create_demo as create_demo_haircolor
 
 with gr.Blocks(css="style.css") as demo:
     with gr.Tabs():

app_haircolor_inpaint_15.py

@@ -24,8 +24,8 @@ from controlnet_aux import (
     HEDdetector,
 )
 
-# BASE_MODEL = "stable-diffusion-v1-5/stable-diffusion-v1-5"
-BASE_MODEL = "SG161222/Realistic_Vision_V5.1_noVAE"
+BASE_MODEL = "stable-diffusion-v1-5/stable-diffusion-v1-5"
+# BASE_MODEL = "SG161222/Realistic_Vision_V5.1_noVAE"
 
 DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
 
@@ -53,10 +53,6 @@ controlnet = [
         "lllyasviel/control_v11p_sd15_softedge",
         torch_dtype=torch.float16,
     ),
-    ControlNetModel.from_pretrained(
-        "lllyasviel/control_v11p_sd15_canny",
-        torch_dtype=torch.float16,
-    ),
 ]
 
 basepipeline = StableDiffusionControlNetInpaintPipeline.from_pretrained(
@@ -83,16 +79,15 @@ def image_to_image(
     generate_size: int,
     cond_scale1: float = 1.2,
     cond_scale2: float = 1.2,
-    cond_scale3: float = 1.2,
 ):
     run_task_time = 0
     time_cost_str = ''
     run_task_time, time_cost_str = get_time_cost(run_task_time, time_cost_str)
-    canny_image = canny_detector(input_image, int(generate_size*1), generate_size)
+    # canny_image = canny_detector(input_image, int(generate_size*1), generate_size)
     lineart_image = lineart_detector(input_image, int(generate_size*1), generate_size)
     run_task_time, time_cost_str = get_time_cost(run_task_time, time_cost_str)
     pidiNet_image = pidiNet_detector(input_image, int(generate_size*1), generate_size)
-    control_image = [lineart_image, pidiNet_image, canny_image]
+    control_image = [lineart_image, pidiNet_image]
 
     generator = torch.Generator(device=DEVICE).manual_seed(seed)
     generated_image = basepipeline(
@@ -106,7 +101,7 @@ def image_to_image(
         width=generate_size,
         guidance_scale=guidance_scale,
         num_inference_steps=num_steps,
-        controlnet_conditioning_scale=[cond_scale1, cond_scale2, cond_scale3],
+        controlnet_conditioning_scale=[cond_scale1, cond_scale2],
     ).images[0]
     
     run_task_time, time_cost_str = get_time_cost(run_task_time, time_cost_str)
@@ -142,7 +137,7 @@ def create_demo() -> gr.Blocks:
                 edit_prompt = gr.Textbox(lines=1, label="Edit Prompt", value=DEFAULT_EDIT_PROMPT)
                 generate_size = gr.Number(label="Generate Size", value=512)
             with gr.Column():
-                num_steps = gr.Slider(minimum=1, maximum=100, value=15, step=1, label="Num Steps")
+                num_steps = gr.Slider(minimum=1, maximum=100, value=25, step=1, label="Num Steps")
                 guidance_scale = gr.Slider(minimum=0, maximum=30, value=5, step=0.5, label="Guidance Scale")
             with gr.Column():
                 with gr.Accordion("Advanced Options", open=False):

app_haircolor_inpaint_adapter_15.py

@@ -72,7 +72,7 @@ basepipeline = DiffusionPipeline.from_pretrained(
     custom_pipeline="./pipelines/pipeline_sd_adapter_inpaint.py",
 )
 
-# basepipeline.scheduler = EulerAncestralDiscreteScheduler.from_config(basepipeline.scheduler.config)
+basepipeline.scheduler = EulerAncestralDiscreteScheduler.from_config(basepipeline.scheduler.config)
 
 basepipeline = basepipeline.to(DEVICE)