Update app.py

app.py

@@ -1,160 +1,100 @@
-import spaces
-import os
-# import subprocess
-# import shlex
-# if os.getenv('SYSTEM') == 'spaces':
-#     git_repo = "https://github.com/huggingface/transformers.git"
-#     subprocess.call(shlex.split(f'pip install git+{git_repo}'))
-    
-import time
 import torch
-from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
+import spaces
 import gradio as gr
-from threading import Thread
-
-MODEL_LIST = ["openbmb/MiniCPM3-4B"]
-HF_TOKEN = os.environ.get("HF_TOKEN", None)
-MODEL = os.environ.get("MODEL_ID")
-
-TITLE = "<h1><center>MiniCPM3-4B</center></h1>"
+from diffusers import FluxInpaintPipeline
+import random
+import numpy as np
 
-PLACEHOLDER = """
-<center>
-<p>MiniCPM3-4B is the 3rd generation of MiniCPM series.</p>
-</center>
-"""
+MAX_SEED = np.iinfo(np.int32).max
+DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
 
+pipe = FluxInpaintPipeline.from_pretrained(
+    "black-forest-labs/FLUX.1-Fill-dev", torch_dtype=torch.bfloat16).to(DEVICE)
 
-CSS = """
-.duplicate-button {
-    margin: auto !important;
-    color: white !important;
-    background: black !important;
-    border-radius: 100vh !important;
-}
-h3 {
-    text-align: center;
-}
-"""
-
-device = "cuda" # for GPU usage or "cpu" for CPU usage
-
-tokenizer = AutoTokenizer.from_pretrained(MODEL, trust_remote_code=True)
-model = AutoModelForCausalLM.from_pretrained(
-    MODEL,
-    torch_dtype=torch.bfloat16,
-    device_map="auto",
-    trust_remote_code=True)
 
 @spaces.GPU()
-def stream_chat(
-    message: str, 
-    history: list, 
-    temperature: float = 0.7, 
-    max_new_tokens: int = 1024, 
-    top_p: float = 1.0, 
-    top_k: int = 20, 
-    penalty: float = 1.2,
-):
-    print(f'message: {message}')
-    print(f'history: {history}')
-
-    conversation = []
-    for prompt, answer in history:
-        conversation.extend([
-            {"role": "user", "content": prompt}, 
-            {"role": "assistant", "content": answer},
-        ])
-
-    conversation.append({"role": "user", "content": message})
-
-    input_text=tokenizer.apply_chat_template(conversation, tokenize=False)
-    inputs = tokenizer.encode(input_text, return_tensors="pt").to(device)
-    streamer = TextIteratorStreamer(tokenizer, timeout=60.0, skip_prompt=True, skip_special_tokens=True)
-    
-    generate_kwargs = dict(
-        input_ids=inputs, 
-        max_new_tokens = max_new_tokens,
-        do_sample = False if temperature == 0 else True,
-        top_p = top_p,
-        top_k = top_k,
-        temperature = temperature,
-        streamer=streamer,
-        repetition_penalty=penalty,
-        eos_token_id = [2, 73440],
-    )
-
-    with torch.no_grad():
-        thread = Thread(target=model.generate, kwargs=generate_kwargs)
-        thread.start()
+def inpaintGen(
+        imgMask,
+        inpaint_prompt: str,
+        strength: float,
+        guidance: float,
+        num_steps: int,
+        seed: int,
+        randomize_seed: bool,
+        progress=gr.Progress(track_tqdm=True)):
+
+    source_img = imgMask["background"]
+    mask_img = imgMask["layers"][0]
+
+    if not source_path:
+        raise gr.Error("Please upload an image.")
+
+    if not mask_path:
+        raise gr.Error("Please draw a mask on the image.")
         
-    buffer = ""
-    for new_text in streamer:
-        buffer += new_text
-        yield buffer
-
-            
-chatbot = gr.Chatbot(height=600, placeholder=PLACEHOLDER)
-
-with gr.Blocks(css=CSS, theme="Nymbo/Nymbo_Theme") as demo:
-    gr.HTML(TITLE)
-    gr.DuplicateButton(value="Duplicate Space for private use", elem_classes="duplicate-button")
-    gr.ChatInterface(
-        fn=stream_chat,
-        chatbot=chatbot,
-        fill_height=True,
-        additional_inputs_accordion=gr.Accordion(label="⚙️ Parameters", open=False, render=False),
-        additional_inputs=[
-            gr.Slider(
-                minimum=0,
-                maximum=1,
-                step=0.1,
-                value=0.7,
-                label="Temperature",
-                render=False,
-            ),
-            gr.Slider(
-                minimum=128,
-                maximum=32768,
-                step=1,
-                value=1024,
-                label="Max new tokens",
-                render=False,
-            ),
-            gr.Slider(
-                minimum=0.0,
-                maximum=1.0,
-                step=0.1,
-                value=1.0,
-                label="top_p",
-                render=False,
-            ),
-            gr.Slider(
-                minimum=1,
-                maximum=20,
-                step=1,
-                value=20,
-                label="top_k",
-                render=False,
-            ),
-            gr.Slider(
-                minimum=0.0,
-                maximum=2.0,
-                step=0.1,
-                value=1.2,
-                label="Repetition penalty",
-                render=False,
-            ),
+    width, height = source_img.size
+
+    if randomize_seed:
+        seed = random.randint(0, MAX_SEED)
+    generator = torch.Generator(device=DEVICE).manual_seed(seed)
+
+    result = pipe(
+        prompt=inpaint_prompt,
+        image=source_img,
+        seed=seed,
+        mask_image=mask_img,
+        width=width,
+        height=height,
+        strength=strength,
+        num_inference_steps=num_steps,
+        generator=generator,
+        guidance_scale=guidance
+    ).images[0]
+    
+    return result
+ 
+
+with gr.Blocks(theme="ocean", title="Flux.1 dev inpaint", css=CSS) as demo:
+    gr.HTML("<h1><center>Flux.1 dev Inpaint</center></h1>")
+    gr.HTML("""
+        <p>
+            <center>
+                A partial redraw of the image based on your prompt words and occluded parts.
+            </center>
+        </p>
+    """)
+    with gr.Row():
+        with gr.Column():
+            imgMask = gr.ImageMask(type="pil", label="Image", layers=False, height=800)
+            inpaint_prompt = gr.Textbox(label='Prompts ✏️', placeholder="A hat...")
+            with gr.Row():
+                Inpaint_sendBtn = gr.Button(value="Submit", variant='primary')
+                Inpaint_clearBtn = gr.ClearButton([imgMask, inpaint_prompt], value="Clear")
+        image_out = gr.Image(type="pil", label="Output", height=960)
+    with gr.Accordion("Advanced ⚙️", open=False):
+        strength = gr.Slider(label="Strength", minimum=0, maximum=1, value=1, step=0.1)
+        guidance = gr.Slider(label="Guidance scale", minimum=1, maximum=20, value=7.5, step=0.1)
+        num_steps = gr.Slider(label="Steps", minimum=1, maximum=20, value=20, step=1)
+        seed = gr.Number(label="Seed", value=42, precision=0)
+        randomize_seed = gr.Checkbox(label="Randomize seed", value=True)
+
+    gr.on(
+        triggers = [
+            inpaint_prompt.submit,
+            Inpaint_sendBtn.click,
         ],
-        examples=[
-            ["Help me study vocabulary: write a sentence for me to fill in the blank, and I'll try to pick the correct option."],
-            ["What are 5 creative things I could do with my kids' art? I don't want to throw them away, but it's also so much clutter."],
-            ["Tell me a random fun fact about the Roman Empire."],
-            ["Show me a code snippet of a website's sticky header in CSS and JavaScript."],
+        fn = inpaintGen,
+        inputs = [
+            imgMask,
+            inpaint_prompt,
+            strength,
+            guidance,
+            num_steps,
+            seed,
+            randomize_seed
         ],
-        cache_examples=False,
+        outputs = [image_out, seed]
     )
 
-
 if __name__ == "__main__":
-    demo.launch()
+    demo.queue(api_open=False).launch(show_api=False, share=False)