ControlNet-v1-1

Runtime error

App Files Files Community

nej-dot commited on Mar 13, 2024

Commit

d93e95e

verified ·

1 Parent(s): aef3692

Update app.py

Browse files

Files changed (1) hide show

app.py +50 -39

app.py CHANGED Viewed

@@ -4,44 +4,55 @@ from __future__ import annotations
 import gradio as gr
 import torch
 from preprocessor import Preprocessor
-DESCRIPTION = "# ControlNet v1.1"
-if not torch.cuda.is_available():
-    DESCRIPTION += "\n<p>Running on CPU 🥶 This demo does not work on CPU.</p>"
-with gr.Blocks(css="style.css") as demo:
-    gr.Markdown(DESCRIPTION)
-    gr.DuplicateButton(
-        value="Duplicate Space for private use",
-        elem_id="duplicate-button",
-        visible=False,  # Hiding the duplicate button
-    )
-    # Instantiate the Preprocessor class
-    preprocessor = Preprocessor()
-    # Function to apply preprocessor to input image
-    def apply_preprocessor(input_image, preprocessor_name):
-        # Load the desired preprocessor
-        preprocessor.load(preprocessor_name)
-        # Apply the preprocessor to the input image
-        processed_image = preprocessor(input_image)
-        # Convert processed image from numpy array to PIL image
-        processed_image_pil = PIL.Image.fromarray(processed_image)
-        return processed_image_pil
-    # Gradio interface
-    image = gr.inputs.Image()
-    preprocessor_name = gr.inputs.Dropdown(choices=["Canny", "Midas", "MLSD", "Openpose", "PidiNet", "NormalBae", "Lineart", "LineartAnime", "ContentShuffle", "DPT", "UPerNet"], label="Preprocessor")
-    run_button = gr.Button("Start")
-    output_image = gr.outputs.Image(label="Output")
-    # Define processing function for Gradio
-    def process_image(input_image, preprocessor_name):
-        return apply_preprocessor(input_image, preprocessor_name)
-    # Link inputs and outputs to function using Gradio's Interface class
-    gr.Interface(fn=process_image, inputs=[image, preprocessor_name], outputs=output_image, title="Preprocessor GUI").launch()

 import gradio as gr
 import torch
+import uvicorn
+from fastapi import FastAPI, File, UploadFile
+from fastapi.responses import HTMLResponse
+from io import BytesIO
+import numpy as np
+import PIL.Image
 from preprocessor import Preprocessor
+app = FastAPI()
+preprocessor = Preprocessor()
+def apply_preprocessor(input_image, preprocessor_name):
+    preprocessor.load(preprocessor_name)
+    processed_image = preprocessor(input_image)
+    return processed_image
+@app.post("/preprocess/")
+async def preprocess_image(file: UploadFile = File(...), preprocessor_name: str = 'Canny'):
+    contents = await file.read()
+    image = PIL.Image.open(BytesIO(contents))
+    processed_image = apply_preprocessor(image, preprocessor_name)
+    processed_image_pil = PIL.Image.fromarray(processed_image)
+    buffered = BytesIO()
+    processed_image_pil.save(buffered, format="JPEG")
+    return {"image": buffered.getvalue()}
+@app.get("/")
+async def main():
+    content = """
+<body>
+<form action="/preprocess/" enctype="multipart/form-data" method="post">
+  <input name="file" type="file">
+  <select name="preprocessor_name">
+    <option value="Canny">Canny</option>
+    <option value="Midas">Midas</option>
+    <option value="MLSD">MLSD</option>
+    <!-- Add more options as needed -->
+  </select>
+  <input type="submit">
+</form>
+</body>
+    """
+    return HTMLResponse(content)
+def apply_gradio_preprocessor(input_image, preprocessor_name):
+    preprocessor.load(preprocessor_name)
+    processed_image = preprocessor(input_image)
+    return processed_image
+if __name__ == "__main__":
+    gr.Interface(apply_gradio_preprocessor, ["image", "text"], "image")\
+        .launch()